{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.25, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.5804758071899414, "learning_rate": 2.25e-07, "loss": 1.3241, "step": 10 }, { "grad_norm": 1.8717567920684814, "learning_rate": 4.75e-07, "loss": 1.3195, "step": 20 }, { "grad_norm": 1.5315630435943604, "learning_rate": 7.25e-07, "loss": 1.3166, "step": 30 }, { "grad_norm": 1.2373793125152588, "learning_rate": 9.75e-07, "loss": 1.3178, "step": 40 }, { "grad_norm": 1.0853451490402222, "learning_rate": 1.2250000000000001e-06, "loss": 1.2908, "step": 50 }, { "grad_norm": 0.612554669380188, "learning_rate": 1.475e-06, "loss": 1.2667, "step": 60 }, { "grad_norm": 0.5330291390419006, "learning_rate": 1.7250000000000002e-06, "loss": 1.2497, "step": 70 }, { "grad_norm": 0.3932086229324341, "learning_rate": 1.975e-06, "loss": 1.2441, "step": 80 }, { "grad_norm": 0.36236467957496643, "learning_rate": 2.225e-06, "loss": 1.2364, "step": 90 }, { "grad_norm": 0.3068911135196686, "learning_rate": 2.4750000000000004e-06, "loss": 1.2446, "step": 100 }, { "grad_norm": 0.48526155948638916, "learning_rate": 2.725e-06, "loss": 1.2193, "step": 110 }, { "grad_norm": 0.5729174613952637, "learning_rate": 2.975e-06, "loss": 1.2086, "step": 120 }, { "grad_norm": 0.7855035066604614, "learning_rate": 3.225e-06, "loss": 1.1626, "step": 130 }, { "grad_norm": 0.5046129822731018, "learning_rate": 3.4750000000000006e-06, "loss": 1.1224, "step": 140 }, { "grad_norm": 0.5061249732971191, "learning_rate": 3.725e-06, "loss": 1.0954, "step": 150 }, { "grad_norm": 0.5324543118476868, "learning_rate": 3.975e-06, "loss": 1.0829, "step": 160 }, { "grad_norm": 0.4080473482608795, "learning_rate": 4.225e-06, "loss": 1.0811, "step": 170 }, { "grad_norm": 0.37840619683265686, "learning_rate": 4.475e-06, "loss": 1.0724, "step": 180 }, { "grad_norm": 0.48704662919044495, "learning_rate": 4.7250000000000005e-06, "loss": 1.0603, "step": 190 }, { "grad_norm": 0.4169861972332001, "learning_rate": 4.975000000000001e-06, "loss": 1.0546, "step": 200 }, { "grad_norm": 0.5208290219306946, "learning_rate": 5.225e-06, "loss": 1.0535, "step": 210 }, { "grad_norm": 0.524604856967926, "learning_rate": 5.475e-06, "loss": 1.0537, "step": 220 }, { "grad_norm": 0.4449763596057892, "learning_rate": 5.725e-06, "loss": 1.0449, "step": 230 }, { "grad_norm": 0.49080708622932434, "learning_rate": 5.975e-06, "loss": 1.0422, "step": 240 }, { "grad_norm": 0.5068081021308899, "learning_rate": 6.2250000000000005e-06, "loss": 1.0394, "step": 250 }, { "grad_norm": 0.4316968321800232, "learning_rate": 6.475000000000001e-06, "loss": 1.0446, "step": 260 }, { "grad_norm": 0.4551021158695221, "learning_rate": 6.725000000000001e-06, "loss": 1.0416, "step": 270 }, { "grad_norm": 0.3824114501476288, "learning_rate": 6.975000000000001e-06, "loss": 1.0402, "step": 280 }, { "grad_norm": 0.4912269115447998, "learning_rate": 7.2249999999999994e-06, "loss": 1.0319, "step": 290 }, { "grad_norm": 0.4402122497558594, "learning_rate": 7.4750000000000004e-06, "loss": 1.0323, "step": 300 }, { "grad_norm": 0.5344911217689514, "learning_rate": 7.725e-06, "loss": 1.0347, "step": 310 }, { "grad_norm": 0.5090672969818115, "learning_rate": 7.975e-06, "loss": 1.0334, "step": 320 }, { "grad_norm": 0.6098300218582153, "learning_rate": 8.225e-06, "loss": 1.0209, "step": 330 }, { "grad_norm": 0.7419542670249939, "learning_rate": 8.475000000000001e-06, "loss": 1.0021, "step": 340 }, { "grad_norm": 1.1041414737701416, "learning_rate": 8.725e-06, "loss": 0.9955, "step": 350 }, { "grad_norm": 1.7459360361099243, "learning_rate": 8.975e-06, "loss": 0.9665, "step": 360 }, { "grad_norm": 1.0827856063842773, "learning_rate": 9.225e-06, "loss": 0.952, "step": 370 }, { "grad_norm": 0.8258600234985352, "learning_rate": 9.475e-06, "loss": 0.9445, "step": 380 }, { "grad_norm": 0.9553149938583374, "learning_rate": 9.725000000000001e-06, "loss": 0.9302, "step": 390 }, { "grad_norm": 1.535204291343689, "learning_rate": 9.975e-06, "loss": 0.8962, "step": 400 }, { "grad_norm": 1.6667358875274658, "learning_rate": 1.0225e-05, "loss": 0.8592, "step": 410 }, { "grad_norm": 1.6375443935394287, "learning_rate": 1.0475e-05, "loss": 0.8122, "step": 420 }, { "grad_norm": 1.6874339580535889, "learning_rate": 1.0725e-05, "loss": 0.7872, "step": 430 }, { "grad_norm": 1.56001615524292, "learning_rate": 1.0975e-05, "loss": 0.751, "step": 440 }, { "grad_norm": 1.6908584833145142, "learning_rate": 1.1225e-05, "loss": 0.7269, "step": 450 }, { "grad_norm": 1.6333063840866089, "learning_rate": 1.1475000000000001e-05, "loss": 0.6882, "step": 460 }, { "grad_norm": 1.6692426204681396, "learning_rate": 1.1725e-05, "loss": 0.6605, "step": 470 }, { "grad_norm": 1.8694617748260498, "learning_rate": 1.1975e-05, "loss": 0.6209, "step": 480 }, { "grad_norm": 1.3084651231765747, "learning_rate": 1.2225e-05, "loss": 0.5866, "step": 490 }, { "grad_norm": 1.5429617166519165, "learning_rate": 1.2475e-05, "loss": 0.5642, "step": 500 }, { "grad_norm": 1.674414038658142, "learning_rate": 1.2725000000000001e-05, "loss": 0.532, "step": 510 }, { "grad_norm": 1.3731719255447388, "learning_rate": 1.2975e-05, "loss": 0.5052, "step": 520 }, { "grad_norm": 2.012216091156006, "learning_rate": 1.3225000000000001e-05, "loss": 0.4744, "step": 530 }, { "grad_norm": 1.848905324935913, "learning_rate": 1.3475000000000002e-05, "loss": 0.4497, "step": 540 }, { "grad_norm": 1.5704342126846313, "learning_rate": 1.3725000000000002e-05, "loss": 0.4273, "step": 550 }, { "grad_norm": 1.6261579990386963, "learning_rate": 1.3975000000000003e-05, "loss": 0.4125, "step": 560 }, { "grad_norm": 1.88864004611969, "learning_rate": 1.4225e-05, "loss": 0.4013, "step": 570 }, { "grad_norm": 2.1141104698181152, "learning_rate": 1.4475e-05, "loss": 0.3807, "step": 580 }, { "grad_norm": 2.2621655464172363, "learning_rate": 1.4725e-05, "loss": 0.3562, "step": 590 }, { "grad_norm": 1.5682884454727173, "learning_rate": 1.4975e-05, "loss": 0.3506, "step": 600 }, { "grad_norm": 1.804977297782898, "learning_rate": 1.5225e-05, "loss": 0.3284, "step": 610 }, { "grad_norm": 1.7495858669281006, "learning_rate": 1.5475e-05, "loss": 0.3249, "step": 620 }, { "grad_norm": 2.168168067932129, "learning_rate": 1.5725e-05, "loss": 0.2988, "step": 630 }, { "grad_norm": 2.179145097732544, "learning_rate": 1.5975000000000002e-05, "loss": 0.2707, "step": 640 }, { "grad_norm": 2.426710367202759, "learning_rate": 1.6225e-05, "loss": 0.2569, "step": 650 }, { "grad_norm": 2.249342441558838, "learning_rate": 1.6475e-05, "loss": 0.2383, "step": 660 }, { "grad_norm": 1.7724189758300781, "learning_rate": 1.6725000000000003e-05, "loss": 0.2155, "step": 670 }, { "grad_norm": 1.7804030179977417, "learning_rate": 1.6975000000000003e-05, "loss": 0.1947, "step": 680 }, { "grad_norm": 2.3450708389282227, "learning_rate": 1.7225e-05, "loss": 0.1867, "step": 690 }, { "grad_norm": 1.6324489116668701, "learning_rate": 1.7475e-05, "loss": 0.1652, "step": 700 }, { "grad_norm": 2.6608893871307373, "learning_rate": 1.7725e-05, "loss": 0.1712, "step": 710 }, { "grad_norm": 1.6525903940200806, "learning_rate": 1.7975e-05, "loss": 0.1533, "step": 720 }, { "grad_norm": 2.136901617050171, "learning_rate": 1.8225e-05, "loss": 0.1312, "step": 730 }, { "grad_norm": 2.1403656005859375, "learning_rate": 1.8475000000000002e-05, "loss": 0.1429, "step": 740 }, { "grad_norm": 2.2409255504608154, "learning_rate": 1.8725e-05, "loss": 0.1424, "step": 750 }, { "grad_norm": 1.8224451541900635, "learning_rate": 1.8975e-05, "loss": 0.1438, "step": 760 }, { "grad_norm": 1.8351229429244995, "learning_rate": 1.9225e-05, "loss": 0.1337, "step": 770 }, { "grad_norm": 1.8898205757141113, "learning_rate": 1.9475000000000002e-05, "loss": 0.1332, "step": 780 }, { "grad_norm": 1.8654084205627441, "learning_rate": 1.9725000000000002e-05, "loss": 0.1309, "step": 790 }, { "grad_norm": 1.5830415487289429, "learning_rate": 1.9975e-05, "loss": 0.1287, "step": 800 }, { "grad_norm": 2.4884510040283203, "learning_rate": 2.0225000000000004e-05, "loss": 0.1279, "step": 810 }, { "grad_norm": 1.9810264110565186, "learning_rate": 2.0475e-05, "loss": 0.1235, "step": 820 }, { "grad_norm": 1.2656575441360474, "learning_rate": 2.0725e-05, "loss": 0.123, "step": 830 }, { "grad_norm": 1.218823790550232, "learning_rate": 2.0975e-05, "loss": 0.1163, "step": 840 }, { "grad_norm": 1.6116087436676025, "learning_rate": 2.1225e-05, "loss": 0.1239, "step": 850 }, { "grad_norm": 1.736755609512329, "learning_rate": 2.1475e-05, "loss": 0.1275, "step": 860 }, { "grad_norm": 1.3718782663345337, "learning_rate": 2.1725e-05, "loss": 0.1126, "step": 870 }, { "grad_norm": 1.6815882921218872, "learning_rate": 2.1975000000000002e-05, "loss": 0.1193, "step": 880 }, { "grad_norm": 1.7289360761642456, "learning_rate": 2.2225e-05, "loss": 0.1078, "step": 890 }, { "grad_norm": 1.62739896774292, "learning_rate": 2.2475e-05, "loss": 0.1063, "step": 900 }, { "grad_norm": 1.3754408359527588, "learning_rate": 2.2725000000000003e-05, "loss": 0.1046, "step": 910 }, { "grad_norm": 1.4291608333587646, "learning_rate": 2.2975000000000003e-05, "loss": 0.0991, "step": 920 }, { "grad_norm": 1.6611374616622925, "learning_rate": 2.3225000000000002e-05, "loss": 0.1048, "step": 930 }, { "grad_norm": 1.599436640739441, "learning_rate": 2.3475e-05, "loss": 0.1004, "step": 940 }, { "grad_norm": 1.556668758392334, "learning_rate": 2.3725e-05, "loss": 0.1048, "step": 950 }, { "grad_norm": 1.4808510541915894, "learning_rate": 2.3975e-05, "loss": 0.1048, "step": 960 }, { "grad_norm": 1.568935751914978, "learning_rate": 2.4225e-05, "loss": 0.1048, "step": 970 }, { "grad_norm": 1.7636423110961914, "learning_rate": 2.4475000000000002e-05, "loss": 0.1044, "step": 980 }, { "grad_norm": 1.374710202217102, "learning_rate": 2.4725e-05, "loss": 0.1025, "step": 990 }, { "grad_norm": 1.6602052450180054, "learning_rate": 2.4975e-05, "loss": 0.1055, "step": 1000 }, { "grad_norm": 1.4446256160736084, "learning_rate": 2.5225e-05, "loss": 0.0987, "step": 1010 }, { "grad_norm": 1.3204998970031738, "learning_rate": 2.5475e-05, "loss": 0.1005, "step": 1020 }, { "grad_norm": 1.3235571384429932, "learning_rate": 2.5725e-05, "loss": 0.105, "step": 1030 }, { "grad_norm": 1.5824202299118042, "learning_rate": 2.5974999999999998e-05, "loss": 0.0966, "step": 1040 }, { "grad_norm": 1.6798382997512817, "learning_rate": 2.6225e-05, "loss": 0.0913, "step": 1050 }, { "grad_norm": 1.3357754945755005, "learning_rate": 2.6475e-05, "loss": 0.0918, "step": 1060 }, { "grad_norm": 1.4864839315414429, "learning_rate": 2.6725e-05, "loss": 0.0925, "step": 1070 }, { "grad_norm": 1.4565593004226685, "learning_rate": 2.6975000000000002e-05, "loss": 0.0862, "step": 1080 }, { "grad_norm": 1.758023738861084, "learning_rate": 2.7225e-05, "loss": 0.0904, "step": 1090 }, { "grad_norm": 1.8773068189620972, "learning_rate": 2.7475e-05, "loss": 0.0997, "step": 1100 }, { "grad_norm": 1.2083154916763306, "learning_rate": 2.7725e-05, "loss": 0.0845, "step": 1110 }, { "grad_norm": 1.3486703634262085, "learning_rate": 2.7975000000000002e-05, "loss": 0.0966, "step": 1120 }, { "grad_norm": 1.534922480583191, "learning_rate": 2.8225e-05, "loss": 0.0921, "step": 1130 }, { "grad_norm": 1.4647705554962158, "learning_rate": 2.8475e-05, "loss": 0.0922, "step": 1140 }, { "grad_norm": 1.4060485363006592, "learning_rate": 2.8725e-05, "loss": 0.0948, "step": 1150 }, { "grad_norm": 1.51994788646698, "learning_rate": 2.8975000000000003e-05, "loss": 0.0855, "step": 1160 }, { "grad_norm": 1.33531653881073, "learning_rate": 2.9225000000000002e-05, "loss": 0.0903, "step": 1170 }, { "grad_norm": 1.6646572351455688, "learning_rate": 2.9475e-05, "loss": 0.0952, "step": 1180 }, { "grad_norm": 1.41112220287323, "learning_rate": 2.9725000000000004e-05, "loss": 0.0868, "step": 1190 }, { "grad_norm": 1.2162317037582397, "learning_rate": 2.9975000000000004e-05, "loss": 0.091, "step": 1200 }, { "grad_norm": 1.1530526876449585, "learning_rate": 3.0225000000000003e-05, "loss": 0.0837, "step": 1210 }, { "grad_norm": 1.087131381034851, "learning_rate": 3.0475000000000002e-05, "loss": 0.0783, "step": 1220 }, { "grad_norm": 0.9916775822639465, "learning_rate": 3.0725e-05, "loss": 0.0809, "step": 1230 }, { "grad_norm": 1.5904446840286255, "learning_rate": 3.0975e-05, "loss": 0.0914, "step": 1240 }, { "grad_norm": 1.2189592123031616, "learning_rate": 3.122500000000001e-05, "loss": 0.0852, "step": 1250 }, { "grad_norm": 1.201271653175354, "learning_rate": 3.1475e-05, "loss": 0.0734, "step": 1260 }, { "grad_norm": 1.0126718282699585, "learning_rate": 3.1725e-05, "loss": 0.0894, "step": 1270 }, { "grad_norm": 1.4852685928344727, "learning_rate": 3.1975e-05, "loss": 0.0861, "step": 1280 }, { "grad_norm": 1.4800124168395996, "learning_rate": 3.2225e-05, "loss": 0.0889, "step": 1290 }, { "grad_norm": 1.4371424913406372, "learning_rate": 3.2474999999999997e-05, "loss": 0.0848, "step": 1300 }, { "grad_norm": 1.0644303560256958, "learning_rate": 3.2725e-05, "loss": 0.076, "step": 1310 }, { "grad_norm": 0.9836310744285583, "learning_rate": 3.2975e-05, "loss": 0.0906, "step": 1320 }, { "grad_norm": 1.8215713500976562, "learning_rate": 3.3225e-05, "loss": 0.0784, "step": 1330 }, { "grad_norm": 1.2189385890960693, "learning_rate": 3.3475e-05, "loss": 0.0752, "step": 1340 }, { "grad_norm": 1.1196247339248657, "learning_rate": 3.3725e-05, "loss": 0.0839, "step": 1350 }, { "grad_norm": 1.3622581958770752, "learning_rate": 3.3975e-05, "loss": 0.0823, "step": 1360 }, { "grad_norm": 1.0857255458831787, "learning_rate": 3.4225e-05, "loss": 0.0743, "step": 1370 }, { "grad_norm": 1.053144931793213, "learning_rate": 3.4475000000000005e-05, "loss": 0.0755, "step": 1380 }, { "grad_norm": 1.0505937337875366, "learning_rate": 3.4725000000000004e-05, "loss": 0.0697, "step": 1390 }, { "grad_norm": 1.0389550924301147, "learning_rate": 3.4975e-05, "loss": 0.0709, "step": 1400 }, { "grad_norm": 1.38200843334198, "learning_rate": 3.5225e-05, "loss": 0.072, "step": 1410 }, { "grad_norm": 1.0021291971206665, "learning_rate": 3.5475e-05, "loss": 0.0721, "step": 1420 }, { "grad_norm": 1.2702776193618774, "learning_rate": 3.5725e-05, "loss": 0.0872, "step": 1430 }, { "grad_norm": 1.5002415180206299, "learning_rate": 3.5975e-05, "loss": 0.0733, "step": 1440 }, { "grad_norm": 0.9429214596748352, "learning_rate": 3.6225000000000006e-05, "loss": 0.0678, "step": 1450 }, { "grad_norm": 1.4249478578567505, "learning_rate": 3.6475000000000006e-05, "loss": 0.0775, "step": 1460 }, { "grad_norm": 1.377415657043457, "learning_rate": 3.6725000000000005e-05, "loss": 0.0718, "step": 1470 }, { "grad_norm": 1.0675086975097656, "learning_rate": 3.6975000000000004e-05, "loss": 0.0705, "step": 1480 }, { "grad_norm": 1.2412070035934448, "learning_rate": 3.7225000000000004e-05, "loss": 0.0749, "step": 1490 }, { "grad_norm": 1.0937161445617676, "learning_rate": 3.7475e-05, "loss": 0.0711, "step": 1500 }, { "grad_norm": 1.202326774597168, "learning_rate": 3.7725e-05, "loss": 0.076, "step": 1510 }, { "grad_norm": 0.9144443273544312, "learning_rate": 3.7975e-05, "loss": 0.0713, "step": 1520 }, { "grad_norm": 1.1995528936386108, "learning_rate": 3.8225e-05, "loss": 0.0675, "step": 1530 }, { "grad_norm": 1.1961479187011719, "learning_rate": 3.8475e-05, "loss": 0.075, "step": 1540 }, { "grad_norm": 0.8805493116378784, "learning_rate": 3.8725e-05, "loss": 0.0711, "step": 1550 }, { "grad_norm": 1.279342770576477, "learning_rate": 3.8975e-05, "loss": 0.0733, "step": 1560 }, { "grad_norm": 1.0989315509796143, "learning_rate": 3.9225e-05, "loss": 0.0718, "step": 1570 }, { "grad_norm": 1.0184385776519775, "learning_rate": 3.9475000000000004e-05, "loss": 0.0666, "step": 1580 }, { "grad_norm": 1.127661943435669, "learning_rate": 3.9725e-05, "loss": 0.0724, "step": 1590 }, { "grad_norm": 0.7816567420959473, "learning_rate": 3.9975e-05, "loss": 0.0751, "step": 1600 }, { "grad_norm": 0.8313968181610107, "learning_rate": 4.0225e-05, "loss": 0.0673, "step": 1610 }, { "grad_norm": 1.028823733329773, "learning_rate": 4.0475e-05, "loss": 0.0667, "step": 1620 }, { "grad_norm": 1.1914780139923096, "learning_rate": 4.0725e-05, "loss": 0.0718, "step": 1630 }, { "grad_norm": 1.2256585359573364, "learning_rate": 4.0975e-05, "loss": 0.0743, "step": 1640 }, { "grad_norm": 0.9711947441101074, "learning_rate": 4.1225e-05, "loss": 0.064, "step": 1650 }, { "grad_norm": 0.9852539300918579, "learning_rate": 4.1475000000000005e-05, "loss": 0.0655, "step": 1660 }, { "grad_norm": 1.0790479183197021, "learning_rate": 4.1725000000000005e-05, "loss": 0.0725, "step": 1670 }, { "grad_norm": 0.9619845151901245, "learning_rate": 4.1975000000000004e-05, "loss": 0.068, "step": 1680 }, { "grad_norm": 0.9753040671348572, "learning_rate": 4.2225e-05, "loss": 0.0618, "step": 1690 }, { "grad_norm": 0.9900170564651489, "learning_rate": 4.2475e-05, "loss": 0.0787, "step": 1700 }, { "grad_norm": 1.0293396711349487, "learning_rate": 4.2725e-05, "loss": 0.0648, "step": 1710 }, { "grad_norm": 1.0182240009307861, "learning_rate": 4.2975e-05, "loss": 0.0696, "step": 1720 }, { "grad_norm": 0.7806861996650696, "learning_rate": 4.322500000000001e-05, "loss": 0.0665, "step": 1730 }, { "grad_norm": 0.96494460105896, "learning_rate": 4.3475000000000006e-05, "loss": 0.0622, "step": 1740 }, { "grad_norm": 0.813382625579834, "learning_rate": 4.3725000000000006e-05, "loss": 0.062, "step": 1750 }, { "grad_norm": 0.7419213056564331, "learning_rate": 4.3975e-05, "loss": 0.0589, "step": 1760 }, { "grad_norm": 1.1195895671844482, "learning_rate": 4.4225e-05, "loss": 0.0746, "step": 1770 }, { "grad_norm": 0.7494054436683655, "learning_rate": 4.4475e-05, "loss": 0.0695, "step": 1780 }, { "grad_norm": 0.9094904661178589, "learning_rate": 4.4725e-05, "loss": 0.0661, "step": 1790 }, { "grad_norm": 0.8741764426231384, "learning_rate": 4.4975e-05, "loss": 0.0725, "step": 1800 }, { "grad_norm": 0.8735498785972595, "learning_rate": 4.5225e-05, "loss": 0.0643, "step": 1810 }, { "grad_norm": 1.215312123298645, "learning_rate": 4.5475e-05, "loss": 0.0682, "step": 1820 }, { "grad_norm": 1.1337857246398926, "learning_rate": 4.5725e-05, "loss": 0.0645, "step": 1830 }, { "grad_norm": 0.9647021889686584, "learning_rate": 4.5975e-05, "loss": 0.0598, "step": 1840 }, { "grad_norm": 1.049381971359253, "learning_rate": 4.6225e-05, "loss": 0.0594, "step": 1850 }, { "grad_norm": 0.8957306146621704, "learning_rate": 4.6475000000000005e-05, "loss": 0.059, "step": 1860 }, { "grad_norm": 1.0318583250045776, "learning_rate": 4.6725000000000004e-05, "loss": 0.0619, "step": 1870 }, { "grad_norm": 1.258488416671753, "learning_rate": 4.6975000000000003e-05, "loss": 0.0613, "step": 1880 }, { "grad_norm": 0.9142686724662781, "learning_rate": 4.7225e-05, "loss": 0.0561, "step": 1890 }, { "grad_norm": 1.0278676748275757, "learning_rate": 4.7475e-05, "loss": 0.0636, "step": 1900 }, { "grad_norm": 0.8370326161384583, "learning_rate": 4.7725e-05, "loss": 0.0648, "step": 1910 }, { "grad_norm": 0.6601892113685608, "learning_rate": 4.7975e-05, "loss": 0.0564, "step": 1920 }, { "grad_norm": 0.951420247554779, "learning_rate": 4.822500000000001e-05, "loss": 0.0604, "step": 1930 }, { "grad_norm": 0.9954196810722351, "learning_rate": 4.8475000000000006e-05, "loss": 0.0725, "step": 1940 }, { "grad_norm": 0.6554648280143738, "learning_rate": 4.8725000000000005e-05, "loss": 0.0605, "step": 1950 }, { "grad_norm": 0.5898730158805847, "learning_rate": 4.8975000000000005e-05, "loss": 0.057, "step": 1960 }, { "grad_norm": 0.8766753673553467, "learning_rate": 4.9225000000000004e-05, "loss": 0.0591, "step": 1970 }, { "grad_norm": 0.6528785228729248, "learning_rate": 4.9475e-05, "loss": 0.065, "step": 1980 }, { "grad_norm": 0.8809645771980286, "learning_rate": 4.9725e-05, "loss": 0.0606, "step": 1990 }, { "grad_norm": 0.8587789535522461, "learning_rate": 4.9975e-05, "loss": 0.0605, "step": 2000 }, { "grad_norm": 1.0797219276428223, "learning_rate": 5.0225e-05, "loss": 0.0603, "step": 2010 }, { "grad_norm": 1.0983003377914429, "learning_rate": 5.047500000000001e-05, "loss": 0.0626, "step": 2020 }, { "grad_norm": 0.8176846504211426, "learning_rate": 5.0725e-05, "loss": 0.056, "step": 2030 }, { "grad_norm": 0.7047812342643738, "learning_rate": 5.0975000000000006e-05, "loss": 0.0619, "step": 2040 }, { "grad_norm": 1.0492463111877441, "learning_rate": 5.1225e-05, "loss": 0.0644, "step": 2050 }, { "grad_norm": 0.9031375050544739, "learning_rate": 5.1475000000000004e-05, "loss": 0.0626, "step": 2060 }, { "grad_norm": 0.8540308475494385, "learning_rate": 5.1725000000000004e-05, "loss": 0.0611, "step": 2070 }, { "grad_norm": 0.8157968521118164, "learning_rate": 5.197500000000001e-05, "loss": 0.0628, "step": 2080 }, { "grad_norm": 0.7345325350761414, "learning_rate": 5.2225e-05, "loss": 0.0622, "step": 2090 }, { "grad_norm": 1.0019264221191406, "learning_rate": 5.247500000000001e-05, "loss": 0.0591, "step": 2100 }, { "grad_norm": 1.0575833320617676, "learning_rate": 5.2725e-05, "loss": 0.0584, "step": 2110 }, { "grad_norm": 0.9776384234428406, "learning_rate": 5.297500000000001e-05, "loss": 0.0569, "step": 2120 }, { "grad_norm": 1.0658453702926636, "learning_rate": 5.3225e-05, "loss": 0.0623, "step": 2130 }, { "grad_norm": 0.720639705657959, "learning_rate": 5.3475e-05, "loss": 0.0622, "step": 2140 }, { "grad_norm": 1.065430998802185, "learning_rate": 5.3725000000000005e-05, "loss": 0.0606, "step": 2150 }, { "grad_norm": 0.7559114694595337, "learning_rate": 5.3975e-05, "loss": 0.0494, "step": 2160 }, { "grad_norm": 0.8789105415344238, "learning_rate": 5.4225000000000003e-05, "loss": 0.0575, "step": 2170 }, { "grad_norm": 0.8627957701683044, "learning_rate": 5.4474999999999996e-05, "loss": 0.059, "step": 2180 }, { "grad_norm": 0.6651662588119507, "learning_rate": 5.4725e-05, "loss": 0.0579, "step": 2190 }, { "grad_norm": 0.7759892344474792, "learning_rate": 5.4975e-05, "loss": 0.0535, "step": 2200 }, { "grad_norm": 1.0674357414245605, "learning_rate": 5.522500000000001e-05, "loss": 0.0538, "step": 2210 }, { "grad_norm": 0.8384922742843628, "learning_rate": 5.5475e-05, "loss": 0.0548, "step": 2220 }, { "grad_norm": 0.5815379023551941, "learning_rate": 5.5725000000000006e-05, "loss": 0.0554, "step": 2230 }, { "grad_norm": 0.5822318196296692, "learning_rate": 5.5975e-05, "loss": 0.056, "step": 2240 }, { "grad_norm": 0.8641577959060669, "learning_rate": 5.6225000000000005e-05, "loss": 0.0643, "step": 2250 }, { "grad_norm": 0.7832859754562378, "learning_rate": 5.6475e-05, "loss": 0.0654, "step": 2260 }, { "grad_norm": 0.8979800939559937, "learning_rate": 5.6725e-05, "loss": 0.0633, "step": 2270 }, { "grad_norm": 0.6623832583427429, "learning_rate": 5.6975e-05, "loss": 0.0566, "step": 2280 }, { "grad_norm": 0.6127943992614746, "learning_rate": 5.722500000000001e-05, "loss": 0.0519, "step": 2290 }, { "grad_norm": 0.7150406241416931, "learning_rate": 5.7475e-05, "loss": 0.0577, "step": 2300 }, { "grad_norm": 0.6027668714523315, "learning_rate": 5.772500000000001e-05, "loss": 0.0583, "step": 2310 }, { "grad_norm": 0.6096724271774292, "learning_rate": 5.7975e-05, "loss": 0.0567, "step": 2320 }, { "grad_norm": 0.9102227687835693, "learning_rate": 5.8225000000000006e-05, "loss": 0.061, "step": 2330 }, { "grad_norm": 0.5551831722259521, "learning_rate": 5.8475000000000005e-05, "loss": 0.0564, "step": 2340 }, { "grad_norm": 0.887827455997467, "learning_rate": 5.8725000000000004e-05, "loss": 0.0538, "step": 2350 }, { "grad_norm": 0.7618852257728577, "learning_rate": 5.8975000000000004e-05, "loss": 0.0572, "step": 2360 }, { "grad_norm": 0.6520123481750488, "learning_rate": 5.922500000000001e-05, "loss": 0.0567, "step": 2370 }, { "grad_norm": 0.6751309037208557, "learning_rate": 5.9475e-05, "loss": 0.0552, "step": 2380 }, { "grad_norm": 0.7604072690010071, "learning_rate": 5.9724999999999995e-05, "loss": 0.0559, "step": 2390 }, { "grad_norm": 0.6734346151351929, "learning_rate": 5.9975e-05, "loss": 0.056, "step": 2400 }, { "grad_norm": 0.7062649726867676, "learning_rate": 6.0225e-05, "loss": 0.049, "step": 2410 }, { "grad_norm": 0.9381898641586304, "learning_rate": 6.0475000000000006e-05, "loss": 0.0552, "step": 2420 }, { "grad_norm": 0.6055729389190674, "learning_rate": 6.0725e-05, "loss": 0.0554, "step": 2430 }, { "grad_norm": 0.6736180782318115, "learning_rate": 6.0975000000000005e-05, "loss": 0.0519, "step": 2440 }, { "grad_norm": 0.6537895202636719, "learning_rate": 6.1225e-05, "loss": 0.0526, "step": 2450 }, { "grad_norm": 0.72300124168396, "learning_rate": 6.1475e-05, "loss": 0.0594, "step": 2460 }, { "grad_norm": 0.7536320090293884, "learning_rate": 6.1725e-05, "loss": 0.052, "step": 2470 }, { "grad_norm": 0.7169305682182312, "learning_rate": 6.1975e-05, "loss": 0.0575, "step": 2480 }, { "grad_norm": 0.7931415438652039, "learning_rate": 6.2225e-05, "loss": 0.0506, "step": 2490 }, { "grad_norm": 0.7386366724967957, "learning_rate": 6.2475e-05, "loss": 0.0554, "step": 2500 }, { "grad_norm": 0.8231036067008972, "learning_rate": 6.2725e-05, "loss": 0.0575, "step": 2510 }, { "grad_norm": 0.6468790769577026, "learning_rate": 6.297500000000001e-05, "loss": 0.0525, "step": 2520 }, { "grad_norm": 0.7281479835510254, "learning_rate": 6.3225e-05, "loss": 0.0509, "step": 2530 }, { "grad_norm": 0.6865285634994507, "learning_rate": 6.347500000000001e-05, "loss": 0.0502, "step": 2540 }, { "grad_norm": 0.7730660438537598, "learning_rate": 6.3725e-05, "loss": 0.0565, "step": 2550 }, { "grad_norm": 0.6887521147727966, "learning_rate": 6.397500000000001e-05, "loss": 0.0505, "step": 2560 }, { "grad_norm": 0.8000773787498474, "learning_rate": 6.4225e-05, "loss": 0.0529, "step": 2570 }, { "grad_norm": 0.581964910030365, "learning_rate": 6.447500000000001e-05, "loss": 0.0547, "step": 2580 }, { "grad_norm": 0.6635438799858093, "learning_rate": 6.4725e-05, "loss": 0.0501, "step": 2590 }, { "grad_norm": 0.6221058964729309, "learning_rate": 6.497500000000001e-05, "loss": 0.0544, "step": 2600 }, { "grad_norm": 0.8988073468208313, "learning_rate": 6.5225e-05, "loss": 0.0545, "step": 2610 }, { "grad_norm": 0.8067904114723206, "learning_rate": 6.5475e-05, "loss": 0.0506, "step": 2620 }, { "grad_norm": 0.6335830092430115, "learning_rate": 6.5725e-05, "loss": 0.0541, "step": 2630 }, { "grad_norm": 0.6546992063522339, "learning_rate": 6.5975e-05, "loss": 0.0514, "step": 2640 }, { "grad_norm": 0.660797655582428, "learning_rate": 6.6225e-05, "loss": 0.0515, "step": 2650 }, { "grad_norm": 0.6269710063934326, "learning_rate": 6.6475e-05, "loss": 0.05, "step": 2660 }, { "grad_norm": 0.6809478402137756, "learning_rate": 6.672500000000001e-05, "loss": 0.0532, "step": 2670 }, { "grad_norm": 0.7377316355705261, "learning_rate": 6.6975e-05, "loss": 0.0506, "step": 2680 }, { "grad_norm": 0.7506793141365051, "learning_rate": 6.722500000000001e-05, "loss": 0.0603, "step": 2690 }, { "grad_norm": 0.6821800470352173, "learning_rate": 6.7475e-05, "loss": 0.0536, "step": 2700 }, { "grad_norm": 0.5513445734977722, "learning_rate": 6.7725e-05, "loss": 0.0571, "step": 2710 }, { "grad_norm": 0.5924985408782959, "learning_rate": 6.7975e-05, "loss": 0.0478, "step": 2720 }, { "grad_norm": 0.5621650218963623, "learning_rate": 6.8225e-05, "loss": 0.0537, "step": 2730 }, { "grad_norm": 0.6558359265327454, "learning_rate": 6.8475e-05, "loss": 0.0544, "step": 2740 }, { "grad_norm": 0.6878412961959839, "learning_rate": 6.8725e-05, "loss": 0.0513, "step": 2750 }, { "grad_norm": 0.691437840461731, "learning_rate": 6.8975e-05, "loss": 0.0541, "step": 2760 }, { "grad_norm": 0.5907464027404785, "learning_rate": 6.9225e-05, "loss": 0.0445, "step": 2770 }, { "grad_norm": 0.706429123878479, "learning_rate": 6.9475e-05, "loss": 0.0495, "step": 2780 }, { "grad_norm": 0.6383002996444702, "learning_rate": 6.9725e-05, "loss": 0.052, "step": 2790 }, { "grad_norm": 0.6239849925041199, "learning_rate": 6.997500000000001e-05, "loss": 0.0489, "step": 2800 }, { "grad_norm": 0.7653884291648865, "learning_rate": 7.022500000000001e-05, "loss": 0.0472, "step": 2810 }, { "grad_norm": 0.5935457944869995, "learning_rate": 7.0475e-05, "loss": 0.0425, "step": 2820 }, { "grad_norm": 0.5883824229240417, "learning_rate": 7.072500000000001e-05, "loss": 0.0514, "step": 2830 }, { "grad_norm": 0.44607630372047424, "learning_rate": 7.0975e-05, "loss": 0.0526, "step": 2840 }, { "grad_norm": 0.6327741146087646, "learning_rate": 7.122500000000001e-05, "loss": 0.0475, "step": 2850 }, { "grad_norm": 0.7613495588302612, "learning_rate": 7.1475e-05, "loss": 0.0475, "step": 2860 }, { "grad_norm": 0.5484658479690552, "learning_rate": 7.172500000000001e-05, "loss": 0.0486, "step": 2870 }, { "grad_norm": 0.38672545552253723, "learning_rate": 7.1975e-05, "loss": 0.0463, "step": 2880 }, { "grad_norm": 0.49973148107528687, "learning_rate": 7.2225e-05, "loss": 0.0451, "step": 2890 }, { "grad_norm": 0.6536918878555298, "learning_rate": 7.2475e-05, "loss": 0.0458, "step": 2900 }, { "grad_norm": 0.4710186719894409, "learning_rate": 7.272499999999999e-05, "loss": 0.0452, "step": 2910 }, { "grad_norm": 0.6256555318832397, "learning_rate": 7.2975e-05, "loss": 0.046, "step": 2920 }, { "grad_norm": 0.6633718013763428, "learning_rate": 7.3225e-05, "loss": 0.0535, "step": 2930 }, { "grad_norm": 0.554173469543457, "learning_rate": 7.347500000000001e-05, "loss": 0.0532, "step": 2940 }, { "grad_norm": 0.5239824056625366, "learning_rate": 7.3725e-05, "loss": 0.0521, "step": 2950 }, { "grad_norm": 0.6194398999214172, "learning_rate": 7.397500000000001e-05, "loss": 0.0565, "step": 2960 }, { "grad_norm": 0.5731562972068787, "learning_rate": 7.4225e-05, "loss": 0.0484, "step": 2970 }, { "grad_norm": 0.6060215830802917, "learning_rate": 7.447500000000001e-05, "loss": 0.0462, "step": 2980 }, { "grad_norm": 0.5943408608436584, "learning_rate": 7.4725e-05, "loss": 0.0469, "step": 2990 }, { "grad_norm": 0.5700511932373047, "learning_rate": 7.4975e-05, "loss": 0.0499, "step": 3000 }, { "grad_norm": 0.7378332614898682, "learning_rate": 7.5225e-05, "loss": 0.0557, "step": 3010 }, { "grad_norm": 0.5845997929573059, "learning_rate": 7.5475e-05, "loss": 0.0498, "step": 3020 }, { "grad_norm": 0.6922839283943176, "learning_rate": 7.5725e-05, "loss": 0.0472, "step": 3030 }, { "grad_norm": 0.7374712228775024, "learning_rate": 7.5975e-05, "loss": 0.0449, "step": 3040 }, { "grad_norm": 0.5667315721511841, "learning_rate": 7.6225e-05, "loss": 0.0483, "step": 3050 }, { "grad_norm": 0.5987017154693604, "learning_rate": 7.6475e-05, "loss": 0.0475, "step": 3060 }, { "grad_norm": 0.6497403979301453, "learning_rate": 7.672500000000001e-05, "loss": 0.0501, "step": 3070 }, { "grad_norm": 0.551892876625061, "learning_rate": 7.697500000000001e-05, "loss": 0.0484, "step": 3080 }, { "grad_norm": 0.41989222168922424, "learning_rate": 7.722500000000001e-05, "loss": 0.0485, "step": 3090 }, { "grad_norm": 0.6634812951087952, "learning_rate": 7.747500000000001e-05, "loss": 0.0512, "step": 3100 }, { "grad_norm": 0.5761085152626038, "learning_rate": 7.7725e-05, "loss": 0.0547, "step": 3110 }, { "grad_norm": 0.49064913392066956, "learning_rate": 7.797500000000001e-05, "loss": 0.0424, "step": 3120 }, { "grad_norm": 0.7838964462280273, "learning_rate": 7.8225e-05, "loss": 0.0558, "step": 3130 }, { "grad_norm": 0.4178016483783722, "learning_rate": 7.8475e-05, "loss": 0.0459, "step": 3140 }, { "grad_norm": 0.612350583076477, "learning_rate": 7.8725e-05, "loss": 0.052, "step": 3150 }, { "grad_norm": 0.3996430039405823, "learning_rate": 7.8975e-05, "loss": 0.0421, "step": 3160 }, { "grad_norm": 0.4704982042312622, "learning_rate": 7.9225e-05, "loss": 0.0459, "step": 3170 }, { "grad_norm": 0.5812197923660278, "learning_rate": 7.9475e-05, "loss": 0.0473, "step": 3180 }, { "grad_norm": 0.6637822389602661, "learning_rate": 7.9725e-05, "loss": 0.05, "step": 3190 }, { "grad_norm": 0.7082248330116272, "learning_rate": 7.9975e-05, "loss": 0.0444, "step": 3200 }, { "grad_norm": 0.6479303240776062, "learning_rate": 8.022500000000001e-05, "loss": 0.0468, "step": 3210 }, { "grad_norm": 0.5674563050270081, "learning_rate": 8.0475e-05, "loss": 0.0496, "step": 3220 }, { "grad_norm": 0.5263784527778625, "learning_rate": 8.072500000000001e-05, "loss": 0.0527, "step": 3230 }, { "grad_norm": 0.7040343880653381, "learning_rate": 8.0975e-05, "loss": 0.0478, "step": 3240 }, { "grad_norm": 0.5729039311408997, "learning_rate": 8.122500000000001e-05, "loss": 0.0493, "step": 3250 }, { "grad_norm": 0.5437662601470947, "learning_rate": 8.1475e-05, "loss": 0.0505, "step": 3260 }, { "grad_norm": 0.6828826069831848, "learning_rate": 8.172500000000001e-05, "loss": 0.0485, "step": 3270 }, { "grad_norm": 0.4766859710216522, "learning_rate": 8.1975e-05, "loss": 0.0436, "step": 3280 }, { "grad_norm": 0.4283410608768463, "learning_rate": 8.2225e-05, "loss": 0.0479, "step": 3290 }, { "grad_norm": 0.5730196237564087, "learning_rate": 8.2475e-05, "loss": 0.0473, "step": 3300 }, { "grad_norm": 0.5976664423942566, "learning_rate": 8.2725e-05, "loss": 0.0467, "step": 3310 }, { "grad_norm": 0.6283445358276367, "learning_rate": 8.2975e-05, "loss": 0.0518, "step": 3320 }, { "grad_norm": 0.5029792785644531, "learning_rate": 8.3225e-05, "loss": 0.0474, "step": 3330 }, { "grad_norm": 0.5867703557014465, "learning_rate": 8.347500000000001e-05, "loss": 0.0466, "step": 3340 }, { "grad_norm": 0.4986271560192108, "learning_rate": 8.3725e-05, "loss": 0.0462, "step": 3350 }, { "grad_norm": 0.5508722066879272, "learning_rate": 8.397500000000001e-05, "loss": 0.0437, "step": 3360 }, { "grad_norm": 0.4374672770500183, "learning_rate": 8.422500000000001e-05, "loss": 0.0463, "step": 3370 }, { "grad_norm": 0.5155274271965027, "learning_rate": 8.447500000000001e-05, "loss": 0.0433, "step": 3380 }, { "grad_norm": 0.6851692795753479, "learning_rate": 8.4725e-05, "loss": 0.048, "step": 3390 }, { "grad_norm": 0.4729194641113281, "learning_rate": 8.4975e-05, "loss": 0.0502, "step": 3400 }, { "grad_norm": 0.5451651811599731, "learning_rate": 8.5225e-05, "loss": 0.0523, "step": 3410 }, { "grad_norm": 0.47126802802085876, "learning_rate": 8.5475e-05, "loss": 0.044, "step": 3420 }, { "grad_norm": 0.5168582201004028, "learning_rate": 8.5725e-05, "loss": 0.04, "step": 3430 }, { "grad_norm": 0.6073234677314758, "learning_rate": 8.5975e-05, "loss": 0.0465, "step": 3440 }, { "grad_norm": 0.6091145873069763, "learning_rate": 8.6225e-05, "loss": 0.0393, "step": 3450 }, { "grad_norm": 0.4942414164543152, "learning_rate": 8.6475e-05, "loss": 0.046, "step": 3460 }, { "grad_norm": 0.495524138212204, "learning_rate": 8.672500000000001e-05, "loss": 0.0511, "step": 3470 }, { "grad_norm": 0.44508063793182373, "learning_rate": 8.6975e-05, "loss": 0.0469, "step": 3480 }, { "grad_norm": 0.5521207451820374, "learning_rate": 8.7225e-05, "loss": 0.0449, "step": 3490 }, { "grad_norm": 0.5172187089920044, "learning_rate": 8.747500000000001e-05, "loss": 0.0422, "step": 3500 }, { "grad_norm": 0.6109887957572937, "learning_rate": 8.7725e-05, "loss": 0.0481, "step": 3510 }, { "grad_norm": 0.6200289726257324, "learning_rate": 8.797500000000001e-05, "loss": 0.0468, "step": 3520 }, { "grad_norm": 0.6270468831062317, "learning_rate": 8.8225e-05, "loss": 0.048, "step": 3530 }, { "grad_norm": 0.48467984795570374, "learning_rate": 8.847500000000001e-05, "loss": 0.0464, "step": 3540 }, { "grad_norm": 0.5796993374824524, "learning_rate": 8.8725e-05, "loss": 0.045, "step": 3550 }, { "grad_norm": 0.5684071183204651, "learning_rate": 8.897500000000001e-05, "loss": 0.0467, "step": 3560 }, { "grad_norm": 0.5523401498794556, "learning_rate": 8.9225e-05, "loss": 0.0484, "step": 3570 }, { "grad_norm": 0.6100823879241943, "learning_rate": 8.9475e-05, "loss": 0.0467, "step": 3580 }, { "grad_norm": 0.5646355152130127, "learning_rate": 8.9725e-05, "loss": 0.043, "step": 3590 }, { "grad_norm": 0.4322081208229065, "learning_rate": 8.9975e-05, "loss": 0.0458, "step": 3600 }, { "grad_norm": 0.5350585579872131, "learning_rate": 9.0225e-05, "loss": 0.0441, "step": 3610 }, { "grad_norm": 0.42050138115882874, "learning_rate": 9.0475e-05, "loss": 0.0425, "step": 3620 }, { "grad_norm": 0.6589758396148682, "learning_rate": 9.072500000000001e-05, "loss": 0.0468, "step": 3630 }, { "grad_norm": 0.35618093609809875, "learning_rate": 9.0975e-05, "loss": 0.0417, "step": 3640 }, { "grad_norm": 0.5839056968688965, "learning_rate": 9.122500000000001e-05, "loss": 0.0408, "step": 3650 }, { "grad_norm": 0.44553500413894653, "learning_rate": 9.1475e-05, "loss": 0.0426, "step": 3660 }, { "grad_norm": 0.4031362533569336, "learning_rate": 9.172500000000001e-05, "loss": 0.0404, "step": 3670 }, { "grad_norm": 0.542816162109375, "learning_rate": 9.1975e-05, "loss": 0.0425, "step": 3680 }, { "grad_norm": 0.45747488737106323, "learning_rate": 9.2225e-05, "loss": 0.051, "step": 3690 }, { "grad_norm": 0.5859243273735046, "learning_rate": 9.2475e-05, "loss": 0.0404, "step": 3700 }, { "grad_norm": 0.5099524259567261, "learning_rate": 9.2725e-05, "loss": 0.0414, "step": 3710 }, { "grad_norm": 0.4879830479621887, "learning_rate": 9.2975e-05, "loss": 0.0421, "step": 3720 }, { "grad_norm": 0.4224315881729126, "learning_rate": 9.3225e-05, "loss": 0.0477, "step": 3730 }, { "grad_norm": 0.4144099950790405, "learning_rate": 9.3475e-05, "loss": 0.0455, "step": 3740 }, { "grad_norm": 0.4595228433609009, "learning_rate": 9.3725e-05, "loss": 0.043, "step": 3750 }, { "grad_norm": 0.6182426810264587, "learning_rate": 9.397500000000001e-05, "loss": 0.0469, "step": 3760 }, { "grad_norm": 0.4187731146812439, "learning_rate": 9.422500000000001e-05, "loss": 0.0393, "step": 3770 }, { "grad_norm": 0.45657241344451904, "learning_rate": 9.4475e-05, "loss": 0.045, "step": 3780 }, { "grad_norm": 0.4843043088912964, "learning_rate": 9.472500000000001e-05, "loss": 0.0394, "step": 3790 }, { "grad_norm": 0.45717254281044006, "learning_rate": 9.4975e-05, "loss": 0.0456, "step": 3800 }, { "grad_norm": 0.5439549088478088, "learning_rate": 9.522500000000001e-05, "loss": 0.0491, "step": 3810 }, { "grad_norm": 0.40307188034057617, "learning_rate": 9.5475e-05, "loss": 0.0491, "step": 3820 }, { "grad_norm": 0.48558807373046875, "learning_rate": 9.572500000000001e-05, "loss": 0.0401, "step": 3830 }, { "grad_norm": 0.45158901810646057, "learning_rate": 9.5975e-05, "loss": 0.0419, "step": 3840 }, { "grad_norm": 0.5567598938941956, "learning_rate": 9.622500000000001e-05, "loss": 0.0476, "step": 3850 }, { "grad_norm": 0.4925945997238159, "learning_rate": 9.6475e-05, "loss": 0.0445, "step": 3860 }, { "grad_norm": 0.4812849164009094, "learning_rate": 9.6725e-05, "loss": 0.0395, "step": 3870 }, { "grad_norm": 0.4819734990596771, "learning_rate": 9.6975e-05, "loss": 0.0417, "step": 3880 }, { "grad_norm": 0.46075639128685, "learning_rate": 9.7225e-05, "loss": 0.0453, "step": 3890 }, { "grad_norm": 0.5041094422340393, "learning_rate": 9.747500000000001e-05, "loss": 0.0476, "step": 3900 }, { "grad_norm": 0.46115848422050476, "learning_rate": 9.7725e-05, "loss": 0.0446, "step": 3910 }, { "grad_norm": 0.6490151286125183, "learning_rate": 9.797500000000001e-05, "loss": 0.0531, "step": 3920 }, { "grad_norm": 0.4663606584072113, "learning_rate": 9.8225e-05, "loss": 0.0417, "step": 3930 }, { "grad_norm": 0.4671683609485626, "learning_rate": 9.847500000000001e-05, "loss": 0.0446, "step": 3940 }, { "grad_norm": 0.37579095363616943, "learning_rate": 9.8725e-05, "loss": 0.0446, "step": 3950 }, { "grad_norm": 0.4108712375164032, "learning_rate": 9.897500000000001e-05, "loss": 0.0439, "step": 3960 }, { "grad_norm": 0.40677133202552795, "learning_rate": 9.9225e-05, "loss": 0.0404, "step": 3970 }, { "grad_norm": 0.46062856912612915, "learning_rate": 9.9475e-05, "loss": 0.0403, "step": 3980 }, { "grad_norm": 0.3810589909553528, "learning_rate": 9.9725e-05, "loss": 0.0422, "step": 3990 }, { "grad_norm": 0.4931447505950928, "learning_rate": 9.9975e-05, "loss": 0.0454, "step": 4000 }, { "grad_norm": 0.48716166615486145, "learning_rate": 9.999999653982884e-05, "loss": 0.0403, "step": 4010 }, { "grad_norm": 0.4851459860801697, "learning_rate": 9.999998457874392e-05, "loss": 0.0393, "step": 4020 }, { "grad_norm": 0.43392226099967957, "learning_rate": 9.999996407402913e-05, "loss": 0.044, "step": 4030 }, { "grad_norm": 0.45107170939445496, "learning_rate": 9.999993502568801e-05, "loss": 0.0394, "step": 4040 }, { "grad_norm": 0.5803298354148865, "learning_rate": 9.999989743372548e-05, "loss": 0.0445, "step": 4050 }, { "grad_norm": 0.35279059410095215, "learning_rate": 9.999985129814798e-05, "loss": 0.0409, "step": 4060 }, { "grad_norm": 0.47739607095718384, "learning_rate": 9.99997966189634e-05, "loss": 0.0356, "step": 4070 }, { "grad_norm": 0.43369925022125244, "learning_rate": 9.999973339618107e-05, "loss": 0.0391, "step": 4080 }, { "grad_norm": 0.40336859226226807, "learning_rate": 9.999966162981179e-05, "loss": 0.038, "step": 4090 }, { "grad_norm": 0.3650773763656616, "learning_rate": 9.999958131986784e-05, "loss": 0.0377, "step": 4100 }, { "grad_norm": 0.5107702016830444, "learning_rate": 9.999949246636293e-05, "loss": 0.0459, "step": 4110 }, { "grad_norm": 0.3754749894142151, "learning_rate": 9.999939506931224e-05, "loss": 0.0455, "step": 4120 }, { "grad_norm": 0.4532296061515808, "learning_rate": 9.999928912873243e-05, "loss": 0.0448, "step": 4130 }, { "grad_norm": 0.5211286544799805, "learning_rate": 9.999917464464159e-05, "loss": 0.0402, "step": 4140 }, { "grad_norm": 0.4890589118003845, "learning_rate": 9.999905161705929e-05, "loss": 0.0437, "step": 4150 }, { "grad_norm": 0.3461669683456421, "learning_rate": 9.999892004600653e-05, "loss": 0.041, "step": 4160 }, { "grad_norm": 0.513232409954071, "learning_rate": 9.999877993150581e-05, "loss": 0.0444, "step": 4170 }, { "grad_norm": 0.39680346846580505, "learning_rate": 9.999863127358108e-05, "loss": 0.043, "step": 4180 }, { "grad_norm": 0.3761858344078064, "learning_rate": 9.999847407225773e-05, "loss": 0.0443, "step": 4190 }, { "grad_norm": 0.4674254059791565, "learning_rate": 9.999830832756262e-05, "loss": 0.0386, "step": 4200 }, { "grad_norm": 0.4501320421695709, "learning_rate": 9.999813403952407e-05, "loss": 0.0382, "step": 4210 }, { "grad_norm": 0.43404489755630493, "learning_rate": 9.999795120817187e-05, "loss": 0.0373, "step": 4220 }, { "grad_norm": 0.31401917338371277, "learning_rate": 9.999775983353725e-05, "loss": 0.037, "step": 4230 }, { "grad_norm": 0.4827536642551422, "learning_rate": 9.999755991565292e-05, "loss": 0.038, "step": 4240 }, { "grad_norm": 0.40263837575912476, "learning_rate": 9.999735145455303e-05, "loss": 0.0404, "step": 4250 }, { "grad_norm": 0.33777299523353577, "learning_rate": 9.99971344502732e-05, "loss": 0.0414, "step": 4260 }, { "grad_norm": 0.4570449888706207, "learning_rate": 9.999690890285053e-05, "loss": 0.0408, "step": 4270 }, { "grad_norm": 0.42940354347229004, "learning_rate": 9.999667481232356e-05, "loss": 0.0398, "step": 4280 }, { "grad_norm": 0.4839484989643097, "learning_rate": 9.999643217873225e-05, "loss": 0.0381, "step": 4290 }, { "grad_norm": 0.4431973397731781, "learning_rate": 9.999618100211809e-05, "loss": 0.0358, "step": 4300 }, { "grad_norm": 0.5034391283988953, "learning_rate": 9.999592128252402e-05, "loss": 0.0375, "step": 4310 }, { "grad_norm": 0.5814909934997559, "learning_rate": 9.999565301999437e-05, "loss": 0.0422, "step": 4320 }, { "grad_norm": 0.5034936666488647, "learning_rate": 9.999537621457502e-05, "loss": 0.0382, "step": 4330 }, { "grad_norm": 0.4933886229991913, "learning_rate": 9.999509086631323e-05, "loss": 0.0405, "step": 4340 }, { "grad_norm": 0.4717528522014618, "learning_rate": 9.99947969752578e-05, "loss": 0.0463, "step": 4350 }, { "grad_norm": 0.3782670497894287, "learning_rate": 9.999449454145891e-05, "loss": 0.0385, "step": 4360 }, { "grad_norm": 0.40590450167655945, "learning_rate": 9.999418356496827e-05, "loss": 0.0398, "step": 4370 }, { "grad_norm": 0.4562370777130127, "learning_rate": 9.999386404583899e-05, "loss": 0.0419, "step": 4380 }, { "grad_norm": 0.32026341557502747, "learning_rate": 9.999353598412568e-05, "loss": 0.0471, "step": 4390 }, { "grad_norm": 0.49280548095703125, "learning_rate": 9.999319937988442e-05, "loss": 0.0386, "step": 4400 }, { "grad_norm": 0.5130825638771057, "learning_rate": 9.999285423317268e-05, "loss": 0.0397, "step": 4410 }, { "grad_norm": 0.36568334698677063, "learning_rate": 9.999250054404947e-05, "loss": 0.0407, "step": 4420 }, { "grad_norm": 0.3084375560283661, "learning_rate": 9.99921383125752e-05, "loss": 0.0377, "step": 4430 }, { "grad_norm": 0.5656561255455017, "learning_rate": 9.99917675388118e-05, "loss": 0.0396, "step": 4440 }, { "grad_norm": 0.40321871638298035, "learning_rate": 9.99913882228226e-05, "loss": 0.0416, "step": 4450 }, { "grad_norm": 0.4154428541660309, "learning_rate": 9.999100036467242e-05, "loss": 0.0369, "step": 4460 }, { "grad_norm": 0.35509753227233887, "learning_rate": 9.999060396442753e-05, "loss": 0.0366, "step": 4470 }, { "grad_norm": 0.4621882140636444, "learning_rate": 9.999019902215566e-05, "loss": 0.0398, "step": 4480 }, { "grad_norm": 0.4398028552532196, "learning_rate": 9.998978553792602e-05, "loss": 0.0436, "step": 4490 }, { "grad_norm": 0.33019211888313293, "learning_rate": 9.998936351180926e-05, "loss": 0.0356, "step": 4500 }, { "grad_norm": 0.3726966679096222, "learning_rate": 9.998893294387747e-05, "loss": 0.0353, "step": 4510 }, { "grad_norm": 0.5215499401092529, "learning_rate": 9.998849383420426e-05, "loss": 0.0354, "step": 4520 }, { "grad_norm": 0.6213363409042358, "learning_rate": 9.998804618286465e-05, "loss": 0.0395, "step": 4530 }, { "grad_norm": 0.3779863119125366, "learning_rate": 9.99875899899351e-05, "loss": 0.0393, "step": 4540 }, { "grad_norm": 0.322689026594162, "learning_rate": 9.99871252554936e-05, "loss": 0.0343, "step": 4550 }, { "grad_norm": 0.5080601572990417, "learning_rate": 9.998665197961955e-05, "loss": 0.0377, "step": 4560 }, { "grad_norm": 0.37500789761543274, "learning_rate": 9.998617016239379e-05, "loss": 0.0357, "step": 4570 }, { "grad_norm": 0.36697426438331604, "learning_rate": 9.998567980389869e-05, "loss": 0.0416, "step": 4580 }, { "grad_norm": 0.4403620958328247, "learning_rate": 9.998518090421802e-05, "loss": 0.043, "step": 4590 }, { "grad_norm": 0.49090543389320374, "learning_rate": 9.998467346343703e-05, "loss": 0.0391, "step": 4600 }, { "grad_norm": 0.5613012909889221, "learning_rate": 9.998415748164243e-05, "loss": 0.0432, "step": 4610 }, { "grad_norm": 0.48424169421195984, "learning_rate": 9.998363295892238e-05, "loss": 0.0377, "step": 4620 }, { "grad_norm": 0.4136826694011688, "learning_rate": 9.998309989536652e-05, "loss": 0.0389, "step": 4630 }, { "grad_norm": 0.36063486337661743, "learning_rate": 9.998255829106593e-05, "loss": 0.0405, "step": 4640 }, { "grad_norm": 0.30049610137939453, "learning_rate": 9.998200814611316e-05, "loss": 0.0364, "step": 4650 }, { "grad_norm": 0.4097042679786682, "learning_rate": 9.998144946060219e-05, "loss": 0.0395, "step": 4660 }, { "grad_norm": 0.4151292145252228, "learning_rate": 9.998088223462852e-05, "loss": 0.0376, "step": 4670 }, { "grad_norm": 0.32112443447113037, "learning_rate": 9.998030646828905e-05, "loss": 0.035, "step": 4680 }, { "grad_norm": 0.32498112320899963, "learning_rate": 9.997972216168217e-05, "loss": 0.0366, "step": 4690 }, { "grad_norm": 0.4590705335140228, "learning_rate": 9.997912931490771e-05, "loss": 0.0464, "step": 4700 }, { "grad_norm": 0.49022001028060913, "learning_rate": 9.9978527928067e-05, "loss": 0.0374, "step": 4710 }, { "grad_norm": 0.38668593764305115, "learning_rate": 9.997791800126277e-05, "loss": 0.0343, "step": 4720 }, { "grad_norm": 0.4539598226547241, "learning_rate": 9.997729953459927e-05, "loss": 0.036, "step": 4730 }, { "grad_norm": 0.4395278990268707, "learning_rate": 9.997667252818214e-05, "loss": 0.0369, "step": 4740 }, { "grad_norm": 0.4104231297969818, "learning_rate": 9.997603698211855e-05, "loss": 0.0359, "step": 4750 }, { "grad_norm": 0.3561823070049286, "learning_rate": 9.99753928965171e-05, "loss": 0.0397, "step": 4760 }, { "grad_norm": 0.3691237270832062, "learning_rate": 9.997474027148781e-05, "loss": 0.0431, "step": 4770 }, { "grad_norm": 0.31150445342063904, "learning_rate": 9.997407910714223e-05, "loss": 0.0363, "step": 4780 }, { "grad_norm": 0.42776718735694885, "learning_rate": 9.997340940359332e-05, "loss": 0.0398, "step": 4790 }, { "grad_norm": 0.43336281180381775, "learning_rate": 9.997273116095552e-05, "loss": 0.0399, "step": 4800 }, { "grad_norm": 0.411251962184906, "learning_rate": 9.997204437934473e-05, "loss": 0.0351, "step": 4810 }, { "grad_norm": 0.3477552533149719, "learning_rate": 9.997134905887829e-05, "loss": 0.0362, "step": 4820 }, { "grad_norm": 0.3760080635547638, "learning_rate": 9.997064519967501e-05, "loss": 0.0348, "step": 4830 }, { "grad_norm": 0.3541578948497772, "learning_rate": 9.996993280185517e-05, "loss": 0.0348, "step": 4840 }, { "grad_norm": 0.34484386444091797, "learning_rate": 9.99692118655405e-05, "loss": 0.0348, "step": 4850 }, { "grad_norm": 0.3778262436389923, "learning_rate": 9.996848239085417e-05, "loss": 0.033, "step": 4860 }, { "grad_norm": 0.3373708724975586, "learning_rate": 9.996774437792085e-05, "loss": 0.0398, "step": 4870 }, { "grad_norm": 0.37428897619247437, "learning_rate": 9.996699782686664e-05, "loss": 0.0358, "step": 4880 }, { "grad_norm": 0.5135112404823303, "learning_rate": 9.996624273781909e-05, "loss": 0.0379, "step": 4890 }, { "grad_norm": 0.4462299942970276, "learning_rate": 9.996547911090725e-05, "loss": 0.0356, "step": 4900 }, { "grad_norm": 0.3338138163089752, "learning_rate": 9.996470694626157e-05, "loss": 0.0337, "step": 4910 }, { "grad_norm": 0.41377365589141846, "learning_rate": 9.996392624401403e-05, "loss": 0.0379, "step": 4920 }, { "grad_norm": 0.3938923180103302, "learning_rate": 9.996313700429801e-05, "loss": 0.0394, "step": 4930 }, { "grad_norm": 0.4879482090473175, "learning_rate": 9.996233922724836e-05, "loss": 0.0335, "step": 4940 }, { "grad_norm": 0.3863908648490906, "learning_rate": 9.996153291300141e-05, "loss": 0.0381, "step": 4950 }, { "grad_norm": 0.4217034578323364, "learning_rate": 9.996071806169494e-05, "loss": 0.0373, "step": 4960 }, { "grad_norm": 0.378920316696167, "learning_rate": 9.995989467346817e-05, "loss": 0.0304, "step": 4970 }, { "grad_norm": 0.3604336977005005, "learning_rate": 9.995906274846183e-05, "loss": 0.0402, "step": 4980 }, { "grad_norm": 0.3891845643520355, "learning_rate": 9.995822228681803e-05, "loss": 0.0383, "step": 4990 }, { "grad_norm": 0.35356947779655457, "learning_rate": 9.99573732886804e-05, "loss": 0.0356, "step": 5000 }, { "grad_norm": 0.35341525077819824, "learning_rate": 9.995651575419402e-05, "loss": 0.0347, "step": 5010 }, { "grad_norm": 0.4237957298755646, "learning_rate": 9.995564968350541e-05, "loss": 0.0401, "step": 5020 }, { "grad_norm": 0.3126814663410187, "learning_rate": 9.995477507676256e-05, "loss": 0.032, "step": 5030 }, { "grad_norm": 0.3962535858154297, "learning_rate": 9.995389193411493e-05, "loss": 0.0324, "step": 5040 }, { "grad_norm": 0.41021761298179626, "learning_rate": 9.995300025571339e-05, "loss": 0.0341, "step": 5050 }, { "grad_norm": 0.34523773193359375, "learning_rate": 9.995210004171034e-05, "loss": 0.0386, "step": 5060 }, { "grad_norm": 0.3886120319366455, "learning_rate": 9.995119129225956e-05, "loss": 0.0349, "step": 5070 }, { "grad_norm": 0.2725774943828583, "learning_rate": 9.995027400751637e-05, "loss": 0.0316, "step": 5080 }, { "grad_norm": 0.3511434495449066, "learning_rate": 9.994934818763751e-05, "loss": 0.036, "step": 5090 }, { "grad_norm": 0.43635377287864685, "learning_rate": 9.994841383278115e-05, "loss": 0.0359, "step": 5100 }, { "grad_norm": 0.3380202353000641, "learning_rate": 9.994747094310695e-05, "loss": 0.0369, "step": 5110 }, { "grad_norm": 0.3579171597957611, "learning_rate": 9.994651951877604e-05, "loss": 0.0333, "step": 5120 }, { "grad_norm": 0.4262107014656067, "learning_rate": 9.994555955995099e-05, "loss": 0.0354, "step": 5130 }, { "grad_norm": 0.5168253779411316, "learning_rate": 9.994459106679581e-05, "loss": 0.0371, "step": 5140 }, { "grad_norm": 0.3941894471645355, "learning_rate": 9.994361403947603e-05, "loss": 0.0358, "step": 5150 }, { "grad_norm": 0.33130091428756714, "learning_rate": 9.994262847815854e-05, "loss": 0.0356, "step": 5160 }, { "grad_norm": 0.33550360798835754, "learning_rate": 9.99416343830118e-05, "loss": 0.0327, "step": 5170 }, { "grad_norm": 0.2731087803840637, "learning_rate": 9.994063175420565e-05, "loss": 0.0307, "step": 5180 }, { "grad_norm": 0.40401625633239746, "learning_rate": 9.99396205919114e-05, "loss": 0.0368, "step": 5190 }, { "grad_norm": 0.3560211658477783, "learning_rate": 9.993860089630185e-05, "loss": 0.0324, "step": 5200 }, { "grad_norm": 0.40533947944641113, "learning_rate": 9.993757266755123e-05, "loss": 0.0393, "step": 5210 }, { "grad_norm": 0.304386168718338, "learning_rate": 9.993653590583522e-05, "loss": 0.0415, "step": 5220 }, { "grad_norm": 0.35559573769569397, "learning_rate": 9.993549061133102e-05, "loss": 0.0364, "step": 5230 }, { "grad_norm": 0.27958977222442627, "learning_rate": 9.993443678421719e-05, "loss": 0.031, "step": 5240 }, { "grad_norm": 0.4003497362136841, "learning_rate": 9.993337442467384e-05, "loss": 0.0355, "step": 5250 }, { "grad_norm": 0.3873649835586548, "learning_rate": 9.993230353288248e-05, "loss": 0.0351, "step": 5260 }, { "grad_norm": 0.3913843333721161, "learning_rate": 9.993122410902608e-05, "loss": 0.0348, "step": 5270 }, { "grad_norm": 0.358053982257843, "learning_rate": 9.993013615328912e-05, "loss": 0.0361, "step": 5280 }, { "grad_norm": 0.4676128923892975, "learning_rate": 9.992903966585747e-05, "loss": 0.0364, "step": 5290 }, { "grad_norm": 0.3467820882797241, "learning_rate": 9.992793464691852e-05, "loss": 0.0389, "step": 5300 }, { "grad_norm": 0.3816692531108856, "learning_rate": 9.992682109666105e-05, "loss": 0.0405, "step": 5310 }, { "grad_norm": 0.380704790353775, "learning_rate": 9.992569901527538e-05, "loss": 0.0423, "step": 5320 }, { "grad_norm": 0.363036572933197, "learning_rate": 9.99245684029532e-05, "loss": 0.0356, "step": 5330 }, { "grad_norm": 0.3765076696872711, "learning_rate": 9.992342925988774e-05, "loss": 0.0375, "step": 5340 }, { "grad_norm": 0.3486805260181427, "learning_rate": 9.992228158627361e-05, "loss": 0.0375, "step": 5350 }, { "grad_norm": 0.38034600019454956, "learning_rate": 9.992112538230693e-05, "loss": 0.0328, "step": 5360 }, { "grad_norm": 0.3038675785064697, "learning_rate": 9.991996064818527e-05, "loss": 0.0325, "step": 5370 }, { "grad_norm": 0.3357524275779724, "learning_rate": 9.991878738410768e-05, "loss": 0.0323, "step": 5380 }, { "grad_norm": 0.41991445422172546, "learning_rate": 9.991760559027457e-05, "loss": 0.0331, "step": 5390 }, { "grad_norm": 0.35563498735427856, "learning_rate": 9.991641526688793e-05, "loss": 0.0304, "step": 5400 }, { "grad_norm": 0.46026307344436646, "learning_rate": 9.991521641415113e-05, "loss": 0.0311, "step": 5410 }, { "grad_norm": 0.30938202142715454, "learning_rate": 9.991400903226904e-05, "loss": 0.0325, "step": 5420 }, { "grad_norm": 0.37668007612228394, "learning_rate": 9.991279312144794e-05, "loss": 0.0371, "step": 5430 }, { "grad_norm": 0.3141757845878601, "learning_rate": 9.991156868189564e-05, "loss": 0.0319, "step": 5440 }, { "grad_norm": 0.37040385603904724, "learning_rate": 9.991033571382131e-05, "loss": 0.0351, "step": 5450 }, { "grad_norm": 0.40831753611564636, "learning_rate": 9.990909421743569e-05, "loss": 0.0346, "step": 5460 }, { "grad_norm": 0.37922415137290955, "learning_rate": 9.990784419295085e-05, "loss": 0.0335, "step": 5470 }, { "grad_norm": 0.38443723320961, "learning_rate": 9.990658564058044e-05, "loss": 0.0332, "step": 5480 }, { "grad_norm": 0.36643046140670776, "learning_rate": 9.990531856053948e-05, "loss": 0.0343, "step": 5490 }, { "grad_norm": 0.32703396677970886, "learning_rate": 9.99040429530445e-05, "loss": 0.0317, "step": 5500 }, { "grad_norm": 0.4024263918399811, "learning_rate": 9.990275881831346e-05, "loss": 0.0374, "step": 5510 }, { "grad_norm": 0.41268572211265564, "learning_rate": 9.990146615656577e-05, "loss": 0.0372, "step": 5520 }, { "grad_norm": 0.30443331599235535, "learning_rate": 9.990016496802233e-05, "loss": 0.0348, "step": 5530 }, { "grad_norm": 0.24607008695602417, "learning_rate": 9.989885525290548e-05, "loss": 0.0349, "step": 5540 }, { "grad_norm": 0.4181927442550659, "learning_rate": 9.989753701143897e-05, "loss": 0.0345, "step": 5550 }, { "grad_norm": 0.3667057752609253, "learning_rate": 9.989621024384812e-05, "loss": 0.0338, "step": 5560 }, { "grad_norm": 0.3519102931022644, "learning_rate": 9.989487495035959e-05, "loss": 0.0336, "step": 5570 }, { "grad_norm": 0.31576403975486755, "learning_rate": 9.989353113120156e-05, "loss": 0.0311, "step": 5580 }, { "grad_norm": 0.33985912799835205, "learning_rate": 9.989217878660366e-05, "loss": 0.0344, "step": 5590 }, { "grad_norm": 0.3794171214103699, "learning_rate": 9.989081791679695e-05, "loss": 0.0322, "step": 5600 }, { "grad_norm": 0.3017905056476593, "learning_rate": 9.988944852201397e-05, "loss": 0.0318, "step": 5610 }, { "grad_norm": 0.37784573435783386, "learning_rate": 9.988807060248873e-05, "loss": 0.0311, "step": 5620 }, { "grad_norm": 0.31930088996887207, "learning_rate": 9.988668415845665e-05, "loss": 0.0295, "step": 5630 }, { "grad_norm": 0.3441274166107178, "learning_rate": 9.988528919015466e-05, "loss": 0.0295, "step": 5640 }, { "grad_norm": 0.29257112741470337, "learning_rate": 9.988388569782112e-05, "loss": 0.0304, "step": 5650 }, { "grad_norm": 0.3475782573223114, "learning_rate": 9.988247368169583e-05, "loss": 0.0383, "step": 5660 }, { "grad_norm": 0.4307859539985657, "learning_rate": 9.988105314202007e-05, "loss": 0.0315, "step": 5670 }, { "grad_norm": 0.3964911103248596, "learning_rate": 9.987962407903659e-05, "loss": 0.0339, "step": 5680 }, { "grad_norm": 0.2994661331176758, "learning_rate": 9.987818649298957e-05, "loss": 0.0287, "step": 5690 }, { "grad_norm": 0.28415563702583313, "learning_rate": 9.987674038412465e-05, "loss": 0.033, "step": 5700 }, { "grad_norm": 0.38855230808258057, "learning_rate": 9.987528575268891e-05, "loss": 0.0311, "step": 5710 }, { "grad_norm": 0.3901574909687042, "learning_rate": 9.987382259893095e-05, "loss": 0.0334, "step": 5720 }, { "grad_norm": 0.32090842723846436, "learning_rate": 9.987235092310074e-05, "loss": 0.0309, "step": 5730 }, { "grad_norm": 0.33864298462867737, "learning_rate": 9.987087072544978e-05, "loss": 0.0356, "step": 5740 }, { "grad_norm": 0.3641158640384674, "learning_rate": 9.9869382006231e-05, "loss": 0.0342, "step": 5750 }, { "grad_norm": 0.3902667462825775, "learning_rate": 9.986788476569875e-05, "loss": 0.0315, "step": 5760 }, { "grad_norm": 0.26436519622802734, "learning_rate": 9.986637900410887e-05, "loss": 0.0364, "step": 5770 }, { "grad_norm": 0.34543678164482117, "learning_rate": 9.986486472171869e-05, "loss": 0.0331, "step": 5780 }, { "grad_norm": 0.4165075421333313, "learning_rate": 9.986334191878692e-05, "loss": 0.0349, "step": 5790 }, { "grad_norm": 0.3214016854763031, "learning_rate": 9.986181059557378e-05, "loss": 0.0325, "step": 5800 }, { "grad_norm": 0.3641844689846039, "learning_rate": 9.986027075234094e-05, "loss": 0.0299, "step": 5810 }, { "grad_norm": 0.3222180902957916, "learning_rate": 9.985872238935152e-05, "loss": 0.0346, "step": 5820 }, { "grad_norm": 0.3920753300189972, "learning_rate": 9.985716550687008e-05, "loss": 0.0357, "step": 5830 }, { "grad_norm": 0.40709930658340454, "learning_rate": 9.985560010516264e-05, "loss": 0.0343, "step": 5840 }, { "grad_norm": 0.29426005482673645, "learning_rate": 9.985402618449668e-05, "loss": 0.0343, "step": 5850 }, { "grad_norm": 0.43747836351394653, "learning_rate": 9.985244374514118e-05, "loss": 0.0353, "step": 5860 }, { "grad_norm": 0.4103662967681885, "learning_rate": 9.985085278736651e-05, "loss": 0.0383, "step": 5870 }, { "grad_norm": 0.49300602078437805, "learning_rate": 9.984925331144452e-05, "loss": 0.0345, "step": 5880 }, { "grad_norm": 0.4642360806465149, "learning_rate": 9.984764531764851e-05, "loss": 0.0308, "step": 5890 }, { "grad_norm": 0.42237722873687744, "learning_rate": 9.984602880625326e-05, "loss": 0.0348, "step": 5900 }, { "grad_norm": 0.33068668842315674, "learning_rate": 9.9844403777535e-05, "loss": 0.0348, "step": 5910 }, { "grad_norm": 0.3703913688659668, "learning_rate": 9.984277023177135e-05, "loss": 0.0312, "step": 5920 }, { "grad_norm": 0.37593597173690796, "learning_rate": 9.984112816924148e-05, "loss": 0.03, "step": 5930 }, { "grad_norm": 0.2675609290599823, "learning_rate": 9.983947759022596e-05, "loss": 0.0357, "step": 5940 }, { "grad_norm": 0.36951160430908203, "learning_rate": 9.983781849500682e-05, "loss": 0.031, "step": 5950 }, { "grad_norm": 0.3635440468788147, "learning_rate": 9.98361508838676e-05, "loss": 0.032, "step": 5960 }, { "grad_norm": 0.28543999791145325, "learning_rate": 9.98344747570932e-05, "loss": 0.0338, "step": 5970 }, { "grad_norm": 0.33879971504211426, "learning_rate": 9.983279011497004e-05, "loss": 0.0315, "step": 5980 }, { "grad_norm": 0.33814799785614014, "learning_rate": 9.983109695778596e-05, "loss": 0.0333, "step": 5990 }, { "grad_norm": 0.36611610651016235, "learning_rate": 9.982939528583032e-05, "loss": 0.0358, "step": 6000 }, { "grad_norm": 0.2786788046360016, "learning_rate": 9.982768509939385e-05, "loss": 0.0284, "step": 6010 }, { "grad_norm": 0.41717275977134705, "learning_rate": 9.982596639876879e-05, "loss": 0.0294, "step": 6020 }, { "grad_norm": 0.2580466568470001, "learning_rate": 9.982423918424881e-05, "loss": 0.0283, "step": 6030 }, { "grad_norm": 0.30569028854370117, "learning_rate": 9.982250345612908e-05, "loss": 0.033, "step": 6040 }, { "grad_norm": 0.26355358958244324, "learning_rate": 9.982075921470611e-05, "loss": 0.031, "step": 6050 }, { "grad_norm": 0.21114680171012878, "learning_rate": 9.981900646027802e-05, "loss": 0.0301, "step": 6060 }, { "grad_norm": 0.2657208740711212, "learning_rate": 9.981724519314425e-05, "loss": 0.0323, "step": 6070 }, { "grad_norm": 0.2643822431564331, "learning_rate": 9.981547541360581e-05, "loss": 0.0348, "step": 6080 }, { "grad_norm": 0.3083733916282654, "learning_rate": 9.981369712196508e-05, "loss": 0.0274, "step": 6090 }, { "grad_norm": 0.36244767904281616, "learning_rate": 9.981191031852592e-05, "loss": 0.0354, "step": 6100 }, { "grad_norm": 0.3607291877269745, "learning_rate": 9.981011500359362e-05, "loss": 0.0313, "step": 6110 }, { "grad_norm": 0.3287528157234192, "learning_rate": 9.9808311177475e-05, "loss": 0.037, "step": 6120 }, { "grad_norm": 0.305624783039093, "learning_rate": 9.980649884047826e-05, "loss": 0.0316, "step": 6130 }, { "grad_norm": 0.3737472593784332, "learning_rate": 9.980467799291307e-05, "loss": 0.0298, "step": 6140 }, { "grad_norm": 0.32851657271385193, "learning_rate": 9.980284863509058e-05, "loss": 0.0341, "step": 6150 }, { "grad_norm": 0.345829039812088, "learning_rate": 9.980101076732334e-05, "loss": 0.0301, "step": 6160 }, { "grad_norm": 0.300635427236557, "learning_rate": 9.979916438992544e-05, "loss": 0.0336, "step": 6170 }, { "grad_norm": 0.301094651222229, "learning_rate": 9.979730950321237e-05, "loss": 0.0327, "step": 6180 }, { "grad_norm": 0.2443520873785019, "learning_rate": 9.979544610750104e-05, "loss": 0.0288, "step": 6190 }, { "grad_norm": 0.3393642008304596, "learning_rate": 9.97935742031099e-05, "loss": 0.0314, "step": 6200 }, { "grad_norm": 0.4151686429977417, "learning_rate": 9.979169379035878e-05, "loss": 0.0319, "step": 6210 }, { "grad_norm": 0.39623329043388367, "learning_rate": 9.978980486956899e-05, "loss": 0.0356, "step": 6220 }, { "grad_norm": 0.31326496601104736, "learning_rate": 9.978790744106332e-05, "loss": 0.0329, "step": 6230 }, { "grad_norm": 0.3928259015083313, "learning_rate": 9.978600150516594e-05, "loss": 0.0307, "step": 6240 }, { "grad_norm": 0.34548911452293396, "learning_rate": 9.978408706220259e-05, "loss": 0.0358, "step": 6250 }, { "grad_norm": 0.35732609033584595, "learning_rate": 9.978216411250032e-05, "loss": 0.0377, "step": 6260 }, { "grad_norm": 0.3739506006240845, "learning_rate": 9.978023265638778e-05, "loss": 0.0295, "step": 6270 }, { "grad_norm": 0.28990718722343445, "learning_rate": 9.977829269419495e-05, "loss": 0.0325, "step": 6280 }, { "grad_norm": 0.34217458963394165, "learning_rate": 9.977634422625335e-05, "loss": 0.031, "step": 6290 }, { "grad_norm": 0.2887698709964752, "learning_rate": 9.97743872528959e-05, "loss": 0.0369, "step": 6300 }, { "grad_norm": 0.29507142305374146, "learning_rate": 9.9772421774457e-05, "loss": 0.0337, "step": 6310 }, { "grad_norm": 0.2605436444282532, "learning_rate": 9.977044779127252e-05, "loss": 0.0315, "step": 6320 }, { "grad_norm": 0.2866348326206207, "learning_rate": 9.976846530367971e-05, "loss": 0.0339, "step": 6330 }, { "grad_norm": 0.3067980110645294, "learning_rate": 9.976647431201735e-05, "loss": 0.0362, "step": 6340 }, { "grad_norm": 0.3955288827419281, "learning_rate": 9.976447481662568e-05, "loss": 0.0356, "step": 6350 }, { "grad_norm": 0.35509157180786133, "learning_rate": 9.976246681784629e-05, "loss": 0.0356, "step": 6360 }, { "grad_norm": 0.34650588035583496, "learning_rate": 9.976045031602234e-05, "loss": 0.035, "step": 6370 }, { "grad_norm": 0.4421693682670593, "learning_rate": 9.975842531149837e-05, "loss": 0.0356, "step": 6380 }, { "grad_norm": 0.2946693003177643, "learning_rate": 9.975639180462043e-05, "loss": 0.035, "step": 6390 }, { "grad_norm": 0.35178568959236145, "learning_rate": 9.975434979573596e-05, "loss": 0.0308, "step": 6400 }, { "grad_norm": 0.3600434362888336, "learning_rate": 9.97522992851939e-05, "loss": 0.0367, "step": 6410 }, { "grad_norm": 0.3031100034713745, "learning_rate": 9.975024027334461e-05, "loss": 0.0351, "step": 6420 }, { "grad_norm": 0.33404844999313354, "learning_rate": 9.974817276053993e-05, "loss": 0.0317, "step": 6430 }, { "grad_norm": 0.292591392993927, "learning_rate": 9.974609674713315e-05, "loss": 0.0315, "step": 6440 }, { "grad_norm": 0.3303113579750061, "learning_rate": 9.9744012233479e-05, "loss": 0.0311, "step": 6450 }, { "grad_norm": 0.36271142959594727, "learning_rate": 9.974191921993366e-05, "loss": 0.031, "step": 6460 }, { "grad_norm": 0.21289807558059692, "learning_rate": 9.973981770685474e-05, "loss": 0.0303, "step": 6470 }, { "grad_norm": 0.28503507375717163, "learning_rate": 9.97377076946014e-05, "loss": 0.0331, "step": 6480 }, { "grad_norm": 0.3316730260848999, "learning_rate": 9.973558918353412e-05, "loss": 0.0324, "step": 6490 }, { "grad_norm": 0.3115486800670624, "learning_rate": 9.973346217401494e-05, "loss": 0.03, "step": 6500 }, { "grad_norm": 0.3007924556732178, "learning_rate": 9.973132666640726e-05, "loss": 0.0296, "step": 6510 }, { "grad_norm": 0.35446974635124207, "learning_rate": 9.972918266107602e-05, "loss": 0.0286, "step": 6520 }, { "grad_norm": 0.31375303864479065, "learning_rate": 9.972703015838756e-05, "loss": 0.031, "step": 6530 }, { "grad_norm": 0.388107031583786, "learning_rate": 9.97248691587097e-05, "loss": 0.0299, "step": 6540 }, { "grad_norm": 0.392443984746933, "learning_rate": 9.972269966241166e-05, "loss": 0.0305, "step": 6550 }, { "grad_norm": 0.5582070350646973, "learning_rate": 9.972052166986417e-05, "loss": 0.0335, "step": 6560 }, { "grad_norm": 0.2978241443634033, "learning_rate": 9.971833518143938e-05, "loss": 0.0364, "step": 6570 }, { "grad_norm": 0.28746116161346436, "learning_rate": 9.971614019751093e-05, "loss": 0.0287, "step": 6580 }, { "grad_norm": 0.33122915029525757, "learning_rate": 9.971393671845383e-05, "loss": 0.0317, "step": 6590 }, { "grad_norm": 0.2545918822288513, "learning_rate": 9.971172474464464e-05, "loss": 0.0359, "step": 6600 }, { "grad_norm": 0.29709136486053467, "learning_rate": 9.97095042764613e-05, "loss": 0.036, "step": 6610 }, { "grad_norm": 0.285858154296875, "learning_rate": 9.970727531428324e-05, "loss": 0.0267, "step": 6620 }, { "grad_norm": 0.30100950598716736, "learning_rate": 9.970503785849132e-05, "loss": 0.0352, "step": 6630 }, { "grad_norm": 0.4385583996772766, "learning_rate": 9.970279190946788e-05, "loss": 0.0337, "step": 6640 }, { "grad_norm": 0.28529447317123413, "learning_rate": 9.970053746759667e-05, "loss": 0.0307, "step": 6650 }, { "grad_norm": 0.31889936327934265, "learning_rate": 9.969827453326292e-05, "loss": 0.0357, "step": 6660 }, { "grad_norm": 0.26854726672172546, "learning_rate": 9.969600310685332e-05, "loss": 0.0333, "step": 6670 }, { "grad_norm": 0.3059445321559906, "learning_rate": 9.969372318875596e-05, "loss": 0.0334, "step": 6680 }, { "grad_norm": 0.40667131543159485, "learning_rate": 9.969143477936043e-05, "loss": 0.0337, "step": 6690 }, { "grad_norm": 0.2750890552997589, "learning_rate": 9.968913787905775e-05, "loss": 0.032, "step": 6700 }, { "grad_norm": 0.2651737630367279, "learning_rate": 9.968683248824045e-05, "loss": 0.0319, "step": 6710 }, { "grad_norm": 0.2615031898021698, "learning_rate": 9.968451860730238e-05, "loss": 0.0301, "step": 6720 }, { "grad_norm": 0.29281651973724365, "learning_rate": 9.968219623663896e-05, "loss": 0.0278, "step": 6730 }, { "grad_norm": 0.2853200435638428, "learning_rate": 9.967986537664702e-05, "loss": 0.0296, "step": 6740 }, { "grad_norm": 0.27219313383102417, "learning_rate": 9.967752602772483e-05, "loss": 0.0276, "step": 6750 }, { "grad_norm": 0.3905573785305023, "learning_rate": 9.967517819027212e-05, "loss": 0.0289, "step": 6760 }, { "grad_norm": 0.2805691361427307, "learning_rate": 9.967282186469009e-05, "loss": 0.0292, "step": 6770 }, { "grad_norm": 0.28261953592300415, "learning_rate": 9.967045705138135e-05, "loss": 0.0303, "step": 6780 }, { "grad_norm": 0.25201138854026794, "learning_rate": 9.966808375074998e-05, "loss": 0.0292, "step": 6790 }, { "grad_norm": 0.2911285161972046, "learning_rate": 9.966570196320154e-05, "loss": 0.0298, "step": 6800 }, { "grad_norm": 0.37716567516326904, "learning_rate": 9.966331168914299e-05, "loss": 0.0316, "step": 6810 }, { "grad_norm": 0.3129914700984955, "learning_rate": 9.966091292898277e-05, "loss": 0.0335, "step": 6820 }, { "grad_norm": 0.3183538615703583, "learning_rate": 9.965850568313076e-05, "loss": 0.0289, "step": 6830 }, { "grad_norm": 0.2631528377532959, "learning_rate": 9.965608995199827e-05, "loss": 0.0309, "step": 6840 }, { "grad_norm": 0.3151519000530243, "learning_rate": 9.965366573599812e-05, "loss": 0.0332, "step": 6850 }, { "grad_norm": 0.3650720715522766, "learning_rate": 9.965123303554453e-05, "loss": 0.0348, "step": 6860 }, { "grad_norm": 0.2541489899158478, "learning_rate": 9.964879185105317e-05, "loss": 0.0272, "step": 6870 }, { "grad_norm": 0.23737789690494537, "learning_rate": 9.964634218294119e-05, "loss": 0.033, "step": 6880 }, { "grad_norm": 0.3223654329776764, "learning_rate": 9.964388403162714e-05, "loss": 0.0258, "step": 6890 }, { "grad_norm": 0.26257339119911194, "learning_rate": 9.96414173975311e-05, "loss": 0.0285, "step": 6900 }, { "grad_norm": 0.38703134655952454, "learning_rate": 9.963894228107451e-05, "loss": 0.0284, "step": 6910 }, { "grad_norm": 0.3771122395992279, "learning_rate": 9.963645868268032e-05, "loss": 0.0302, "step": 6920 }, { "grad_norm": 0.31571948528289795, "learning_rate": 9.963396660277289e-05, "loss": 0.0314, "step": 6930 }, { "grad_norm": 0.26772892475128174, "learning_rate": 9.963146604177807e-05, "loss": 0.0274, "step": 6940 }, { "grad_norm": 0.3277246057987213, "learning_rate": 9.962895700012311e-05, "loss": 0.0281, "step": 6950 }, { "grad_norm": 0.3613256514072418, "learning_rate": 9.962643947823677e-05, "loss": 0.0312, "step": 6960 }, { "grad_norm": 0.21844875812530518, "learning_rate": 9.962391347654921e-05, "loss": 0.0268, "step": 6970 }, { "grad_norm": 0.2649064362049103, "learning_rate": 9.962137899549204e-05, "loss": 0.0302, "step": 6980 }, { "grad_norm": 0.3500897288322449, "learning_rate": 9.961883603549835e-05, "loss": 0.0298, "step": 6990 }, { "grad_norm": 0.2769932746887207, "learning_rate": 9.961628459700267e-05, "loss": 0.03, "step": 7000 }, { "grad_norm": 0.3259471356868744, "learning_rate": 9.961372468044095e-05, "loss": 0.0327, "step": 7010 }, { "grad_norm": 0.3110877275466919, "learning_rate": 9.961115628625062e-05, "loss": 0.031, "step": 7020 }, { "grad_norm": 0.3188609182834625, "learning_rate": 9.960857941487056e-05, "loss": 0.0261, "step": 7030 }, { "grad_norm": 0.30640456080436707, "learning_rate": 9.960599406674106e-05, "loss": 0.0285, "step": 7040 }, { "grad_norm": 0.26176437735557556, "learning_rate": 9.960340024230393e-05, "loss": 0.028, "step": 7050 }, { "grad_norm": 0.35196366906166077, "learning_rate": 9.960079794200232e-05, "loss": 0.0301, "step": 7060 }, { "grad_norm": 0.40819481015205383, "learning_rate": 9.959818716628096e-05, "loss": 0.028, "step": 7070 }, { "grad_norm": 0.23419378697872162, "learning_rate": 9.95955679155859e-05, "loss": 0.0283, "step": 7080 }, { "grad_norm": 0.2530882954597473, "learning_rate": 9.959294019036472e-05, "loss": 0.0295, "step": 7090 }, { "grad_norm": 0.2924431264400482, "learning_rate": 9.959030399106646e-05, "loss": 0.0264, "step": 7100 }, { "grad_norm": 0.26737111806869507, "learning_rate": 9.958765931814153e-05, "loss": 0.0289, "step": 7110 }, { "grad_norm": 0.26453226804733276, "learning_rate": 9.958500617204184e-05, "loss": 0.0292, "step": 7120 }, { "grad_norm": 0.3721925914287567, "learning_rate": 9.958234455322075e-05, "loss": 0.0257, "step": 7130 }, { "grad_norm": 0.34627965092658997, "learning_rate": 9.957967446213308e-05, "loss": 0.0301, "step": 7140 }, { "grad_norm": 0.2805823087692261, "learning_rate": 9.957699589923501e-05, "loss": 0.0247, "step": 7150 }, { "grad_norm": 0.3742091953754425, "learning_rate": 9.957430886498431e-05, "loss": 0.028, "step": 7160 }, { "grad_norm": 0.2237827181816101, "learning_rate": 9.957161335984008e-05, "loss": 0.0258, "step": 7170 }, { "grad_norm": 0.28122249245643616, "learning_rate": 9.956890938426291e-05, "loss": 0.0286, "step": 7180 }, { "grad_norm": 0.3202076852321625, "learning_rate": 9.956619693871482e-05, "loss": 0.0279, "step": 7190 }, { "grad_norm": 0.29190629720687866, "learning_rate": 9.956347602365934e-05, "loss": 0.0288, "step": 7200 }, { "grad_norm": 0.2795129716396332, "learning_rate": 9.956074663956135e-05, "loss": 0.0284, "step": 7210 }, { "grad_norm": 0.34091615676879883, "learning_rate": 9.955800878688726e-05, "loss": 0.0294, "step": 7220 }, { "grad_norm": 0.28620803356170654, "learning_rate": 9.955526246610489e-05, "loss": 0.0303, "step": 7230 }, { "grad_norm": 0.365091472864151, "learning_rate": 9.955250767768349e-05, "loss": 0.0274, "step": 7240 }, { "grad_norm": 0.3387274742126465, "learning_rate": 9.95497444220938e-05, "loss": 0.0268, "step": 7250 }, { "grad_norm": 0.290875107049942, "learning_rate": 9.954697269980797e-05, "loss": 0.028, "step": 7260 }, { "grad_norm": 0.30702725052833557, "learning_rate": 9.954419251129962e-05, "loss": 0.0266, "step": 7270 }, { "grad_norm": 0.17307442426681519, "learning_rate": 9.95414038570438e-05, "loss": 0.0233, "step": 7280 }, { "grad_norm": 0.2371431589126587, "learning_rate": 9.953860673751703e-05, "loss": 0.0268, "step": 7290 }, { "grad_norm": 0.3226504921913147, "learning_rate": 9.953580115319725e-05, "loss": 0.0276, "step": 7300 }, { "grad_norm": 0.2972102165222168, "learning_rate": 9.953298710456387e-05, "loss": 0.0276, "step": 7310 }, { "grad_norm": 0.2684117257595062, "learning_rate": 9.953016459209771e-05, "loss": 0.0279, "step": 7320 }, { "grad_norm": 0.28155723214149475, "learning_rate": 9.952733361628108e-05, "loss": 0.0328, "step": 7330 }, { "grad_norm": 0.3225802779197693, "learning_rate": 9.952449417759772e-05, "loss": 0.0297, "step": 7340 }, { "grad_norm": 0.36575207114219666, "learning_rate": 9.952164627653279e-05, "loss": 0.0315, "step": 7350 }, { "grad_norm": 0.2081121802330017, "learning_rate": 9.951878991357292e-05, "loss": 0.0267, "step": 7360 }, { "grad_norm": 0.32313939929008484, "learning_rate": 9.951592508920622e-05, "loss": 0.0278, "step": 7370 }, { "grad_norm": 0.3537410795688629, "learning_rate": 9.951305180392219e-05, "loss": 0.0279, "step": 7380 }, { "grad_norm": 0.3098139464855194, "learning_rate": 9.951017005821178e-05, "loss": 0.03, "step": 7390 }, { "grad_norm": 0.3148583471775055, "learning_rate": 9.95072798525674e-05, "loss": 0.028, "step": 7400 }, { "grad_norm": 0.30219322443008423, "learning_rate": 9.950438118748293e-05, "loss": 0.031, "step": 7410 }, { "grad_norm": 0.3134860098361969, "learning_rate": 9.950147406345366e-05, "loss": 0.0267, "step": 7420 }, { "grad_norm": 0.2550756633281708, "learning_rate": 9.949855848097635e-05, "loss": 0.0364, "step": 7430 }, { "grad_norm": 0.26177558302879333, "learning_rate": 9.949563444054916e-05, "loss": 0.0307, "step": 7440 }, { "grad_norm": 0.35159537196159363, "learning_rate": 9.949270194267178e-05, "loss": 0.03, "step": 7450 }, { "grad_norm": 0.35326099395751953, "learning_rate": 9.948976098784526e-05, "loss": 0.0303, "step": 7460 }, { "grad_norm": 0.3067980706691742, "learning_rate": 9.948681157657213e-05, "loss": 0.0353, "step": 7470 }, { "grad_norm": 0.2734922170639038, "learning_rate": 9.948385370935638e-05, "loss": 0.0287, "step": 7480 }, { "grad_norm": 0.3073256313800812, "learning_rate": 9.94808873867034e-05, "loss": 0.0305, "step": 7490 }, { "grad_norm": 0.3056360185146332, "learning_rate": 9.947791260912009e-05, "loss": 0.03, "step": 7500 }, { "grad_norm": 0.26718515157699585, "learning_rate": 9.947492937711474e-05, "loss": 0.0256, "step": 7510 }, { "grad_norm": 0.28077831864356995, "learning_rate": 9.947193769119707e-05, "loss": 0.0253, "step": 7520 }, { "grad_norm": 0.40254315733909607, "learning_rate": 9.946893755187834e-05, "loss": 0.0293, "step": 7530 }, { "grad_norm": 0.2525235414505005, "learning_rate": 9.946592895967115e-05, "loss": 0.0274, "step": 7540 }, { "grad_norm": 0.279502809047699, "learning_rate": 9.94629119150896e-05, "loss": 0.0288, "step": 7550 }, { "grad_norm": 0.3983485996723175, "learning_rate": 9.94598864186492e-05, "loss": 0.029, "step": 7560 }, { "grad_norm": 0.22508955001831055, "learning_rate": 9.945685247086696e-05, "loss": 0.0251, "step": 7570 }, { "grad_norm": 0.2255309522151947, "learning_rate": 9.945381007226129e-05, "loss": 0.0313, "step": 7580 }, { "grad_norm": 0.24029585719108582, "learning_rate": 9.945075922335203e-05, "loss": 0.0242, "step": 7590 }, { "grad_norm": 0.27566906809806824, "learning_rate": 9.944769992466049e-05, "loss": 0.0261, "step": 7600 }, { "grad_norm": 0.223931685090065, "learning_rate": 9.944463217670945e-05, "loss": 0.028, "step": 7610 }, { "grad_norm": 0.4113691449165344, "learning_rate": 9.944155598002307e-05, "loss": 0.0283, "step": 7620 }, { "grad_norm": 0.30404648184776306, "learning_rate": 9.943847133512701e-05, "loss": 0.0264, "step": 7630 }, { "grad_norm": 0.3519235849380493, "learning_rate": 9.943537824254834e-05, "loss": 0.0343, "step": 7640 }, { "grad_norm": 0.27482321858406067, "learning_rate": 9.943227670281559e-05, "loss": 0.0304, "step": 7650 }, { "grad_norm": 0.2678792178630829, "learning_rate": 9.942916671645873e-05, "loss": 0.0227, "step": 7660 }, { "grad_norm": 0.3117668032646179, "learning_rate": 9.942604828400916e-05, "loss": 0.0274, "step": 7670 }, { "grad_norm": 0.2566882073879242, "learning_rate": 9.942292140599975e-05, "loss": 0.0284, "step": 7680 }, { "grad_norm": 0.27179381251335144, "learning_rate": 9.94197860829648e-05, "loss": 0.0285, "step": 7690 }, { "grad_norm": 0.30688121914863586, "learning_rate": 9.941664231544004e-05, "loss": 0.0279, "step": 7700 }, { "grad_norm": 0.22601372003555298, "learning_rate": 9.941349010396264e-05, "loss": 0.0291, "step": 7710 }, { "grad_norm": 0.3158596158027649, "learning_rate": 9.941032944907125e-05, "loss": 0.0239, "step": 7720 }, { "grad_norm": 0.23848508298397064, "learning_rate": 9.940716035130596e-05, "loss": 0.0277, "step": 7730 }, { "grad_norm": 0.204639732837677, "learning_rate": 9.940398281120821e-05, "loss": 0.0264, "step": 7740 }, { "grad_norm": 0.2862238883972168, "learning_rate": 9.940079682932102e-05, "loss": 0.0268, "step": 7750 }, { "grad_norm": 0.20541976392269135, "learning_rate": 9.939760240618877e-05, "loss": 0.0273, "step": 7760 }, { "grad_norm": 0.3079169988632202, "learning_rate": 9.939439954235729e-05, "loss": 0.0255, "step": 7770 }, { "grad_norm": 0.4076707363128662, "learning_rate": 9.939118823837387e-05, "loss": 0.0281, "step": 7780 }, { "grad_norm": 0.3606453835964203, "learning_rate": 9.938796849478725e-05, "loss": 0.0269, "step": 7790 }, { "grad_norm": 0.42018550634384155, "learning_rate": 9.938474031214755e-05, "loss": 0.0293, "step": 7800 }, { "grad_norm": 0.2228974848985672, "learning_rate": 9.938150369100643e-05, "loss": 0.0301, "step": 7810 }, { "grad_norm": 0.30721309781074524, "learning_rate": 9.93782586319169e-05, "loss": 0.0263, "step": 7820 }, { "grad_norm": 0.39208361506462097, "learning_rate": 9.937500513543348e-05, "loss": 0.0273, "step": 7830 }, { "grad_norm": 0.32347774505615234, "learning_rate": 9.937174320211207e-05, "loss": 0.0279, "step": 7840 }, { "grad_norm": 0.28271690011024475, "learning_rate": 9.936847283251009e-05, "loss": 0.0229, "step": 7850 }, { "grad_norm": 0.22819893062114716, "learning_rate": 9.936519402718632e-05, "loss": 0.0241, "step": 7860 }, { "grad_norm": 0.35827377438545227, "learning_rate": 9.936190678670102e-05, "loss": 0.0252, "step": 7870 }, { "grad_norm": 0.279680073261261, "learning_rate": 9.935861111161593e-05, "loss": 0.0301, "step": 7880 }, { "grad_norm": 0.2553001940250397, "learning_rate": 9.935530700249416e-05, "loss": 0.0254, "step": 7890 }, { "grad_norm": 0.3010421097278595, "learning_rate": 9.935199445990028e-05, "loss": 0.0295, "step": 7900 }, { "grad_norm": 0.27936625480651855, "learning_rate": 9.934867348440033e-05, "loss": 0.0317, "step": 7910 }, { "grad_norm": 0.3152134418487549, "learning_rate": 9.934534407656176e-05, "loss": 0.0264, "step": 7920 }, { "grad_norm": 0.2799608111381531, "learning_rate": 9.93420062369535e-05, "loss": 0.0296, "step": 7930 }, { "grad_norm": 0.2615707516670227, "learning_rate": 9.933865996614589e-05, "loss": 0.0306, "step": 7940 }, { "grad_norm": 0.22891958057880402, "learning_rate": 9.933530526471068e-05, "loss": 0.0257, "step": 7950 }, { "grad_norm": 0.3250536620616913, "learning_rate": 9.933194213322114e-05, "loss": 0.0284, "step": 7960 }, { "grad_norm": 0.2896082401275635, "learning_rate": 9.932857057225192e-05, "loss": 0.0277, "step": 7970 }, { "grad_norm": 0.2832154631614685, "learning_rate": 9.932519058237912e-05, "loss": 0.0294, "step": 7980 }, { "grad_norm": 0.2537767291069031, "learning_rate": 9.932180216418032e-05, "loss": 0.0248, "step": 7990 }, { "grad_norm": 0.27973246574401855, "learning_rate": 9.931840531823446e-05, "loss": 0.0234, "step": 8000 }, { "grad_norm": 0.3351617753505707, "learning_rate": 9.9315000045122e-05, "loss": 0.027, "step": 8010 }, { "grad_norm": 0.21468253433704376, "learning_rate": 9.931158634542481e-05, "loss": 0.03, "step": 8020 }, { "grad_norm": 0.26580989360809326, "learning_rate": 9.930816421972617e-05, "loss": 0.0255, "step": 8030 }, { "grad_norm": 0.33854180574417114, "learning_rate": 9.930473366861086e-05, "loss": 0.0286, "step": 8040 }, { "grad_norm": 0.24350763857364655, "learning_rate": 9.930129469266505e-05, "loss": 0.0251, "step": 8050 }, { "grad_norm": 0.2687734365463257, "learning_rate": 9.929784729247638e-05, "loss": 0.0231, "step": 8060 }, { "grad_norm": 0.30856743454933167, "learning_rate": 9.929439146863389e-05, "loss": 0.0249, "step": 8070 }, { "grad_norm": 0.5552591681480408, "learning_rate": 9.92909272217281e-05, "loss": 0.0275, "step": 8080 }, { "grad_norm": 0.2625058591365814, "learning_rate": 9.928745455235097e-05, "loss": 0.0273, "step": 8090 }, { "grad_norm": 0.2155928611755371, "learning_rate": 9.928397346109588e-05, "loss": 0.028, "step": 8100 }, { "grad_norm": 0.29587146639823914, "learning_rate": 9.928048394855762e-05, "loss": 0.0308, "step": 8110 }, { "grad_norm": 0.24402636289596558, "learning_rate": 9.92769860153325e-05, "loss": 0.0283, "step": 8120 }, { "grad_norm": 0.26942381262779236, "learning_rate": 9.927347966201819e-05, "loss": 0.0261, "step": 8130 }, { "grad_norm": 0.37884223461151123, "learning_rate": 9.926996488921383e-05, "loss": 0.0276, "step": 8140 }, { "grad_norm": 0.33005890250205994, "learning_rate": 9.926644169752001e-05, "loss": 0.0295, "step": 8150 }, { "grad_norm": 0.29457855224609375, "learning_rate": 9.926291008753875e-05, "loss": 0.0274, "step": 8160 }, { "grad_norm": 0.2743181884288788, "learning_rate": 9.92593700598735e-05, "loss": 0.0267, "step": 8170 }, { "grad_norm": 0.22575163841247559, "learning_rate": 9.925582161512915e-05, "loss": 0.0264, "step": 8180 }, { "grad_norm": 0.23500016331672668, "learning_rate": 9.925226475391205e-05, "loss": 0.0266, "step": 8190 }, { "grad_norm": 0.2989623546600342, "learning_rate": 9.924869947682993e-05, "loss": 0.023, "step": 8200 }, { "grad_norm": 0.2595002353191376, "learning_rate": 9.924512578449204e-05, "loss": 0.0242, "step": 8210 }, { "grad_norm": 0.28166505694389343, "learning_rate": 9.924154367750901e-05, "loss": 0.0236, "step": 8220 }, { "grad_norm": 0.3096007704734802, "learning_rate": 9.923795315649293e-05, "loss": 0.0241, "step": 8230 }, { "grad_norm": 0.2706878185272217, "learning_rate": 9.92343542220573e-05, "loss": 0.0236, "step": 8240 }, { "grad_norm": 0.2554178237915039, "learning_rate": 9.92307468748171e-05, "loss": 0.0228, "step": 8250 }, { "grad_norm": 0.2104376256465912, "learning_rate": 9.922713111538873e-05, "loss": 0.0254, "step": 8260 }, { "grad_norm": 0.2673639953136444, "learning_rate": 9.922350694439003e-05, "loss": 0.0297, "step": 8270 }, { "grad_norm": 0.33514103293418884, "learning_rate": 9.921987436244024e-05, "loss": 0.0289, "step": 8280 }, { "grad_norm": 0.32839831709861755, "learning_rate": 9.921623337016008e-05, "loss": 0.0286, "step": 8290 }, { "grad_norm": 0.27490898966789246, "learning_rate": 9.921258396817172e-05, "loss": 0.0239, "step": 8300 }, { "grad_norm": 0.27165845036506653, "learning_rate": 9.920892615709874e-05, "loss": 0.0289, "step": 8310 }, { "grad_norm": 0.26114630699157715, "learning_rate": 9.920525993756612e-05, "loss": 0.027, "step": 8320 }, { "grad_norm": 0.26886820793151855, "learning_rate": 9.920158531020036e-05, "loss": 0.0238, "step": 8330 }, { "grad_norm": 0.27751654386520386, "learning_rate": 9.919790227562933e-05, "loss": 0.024, "step": 8340 }, { "grad_norm": 0.2137313336133957, "learning_rate": 9.919421083448237e-05, "loss": 0.0244, "step": 8350 }, { "grad_norm": 0.2834407687187195, "learning_rate": 9.919051098739022e-05, "loss": 0.0239, "step": 8360 }, { "grad_norm": 0.24640581011772156, "learning_rate": 9.918680273498514e-05, "loss": 0.0235, "step": 8370 }, { "grad_norm": 0.3255692422389984, "learning_rate": 9.918308607790072e-05, "loss": 0.0255, "step": 8380 }, { "grad_norm": 0.3716834485530853, "learning_rate": 9.917936101677205e-05, "loss": 0.0248, "step": 8390 }, { "grad_norm": 0.27407440543174744, "learning_rate": 9.917562755223564e-05, "loss": 0.0226, "step": 8400 }, { "grad_norm": 0.2612476348876953, "learning_rate": 9.917188568492944e-05, "loss": 0.0237, "step": 8410 }, { "grad_norm": 0.24177910387516022, "learning_rate": 9.916813541549283e-05, "loss": 0.0245, "step": 8420 }, { "grad_norm": 0.2847660779953003, "learning_rate": 9.916437674456663e-05, "loss": 0.0232, "step": 8430 }, { "grad_norm": 0.25172290205955505, "learning_rate": 9.916060967279308e-05, "loss": 0.0263, "step": 8440 }, { "grad_norm": 0.3184906542301178, "learning_rate": 9.91568342008159e-05, "loss": 0.027, "step": 8450 }, { "grad_norm": 0.373081237077713, "learning_rate": 9.915305032928019e-05, "loss": 0.0283, "step": 8460 }, { "grad_norm": 0.2341998815536499, "learning_rate": 9.914925805883253e-05, "loss": 0.0295, "step": 8470 }, { "grad_norm": 0.3236059844493866, "learning_rate": 9.914545739012088e-05, "loss": 0.0257, "step": 8480 }, { "grad_norm": 0.28731751441955566, "learning_rate": 9.91416483237947e-05, "loss": 0.0252, "step": 8490 }, { "grad_norm": 0.3083130717277527, "learning_rate": 9.913783086050485e-05, "loss": 0.0264, "step": 8500 }, { "grad_norm": 0.284101665019989, "learning_rate": 9.913400500090364e-05, "loss": 0.0264, "step": 8510 }, { "grad_norm": 0.23619093000888824, "learning_rate": 9.913017074564479e-05, "loss": 0.0276, "step": 8520 }, { "grad_norm": 0.3078193664550781, "learning_rate": 9.912632809538348e-05, "loss": 0.0287, "step": 8530 }, { "grad_norm": 0.35077035427093506, "learning_rate": 9.912247705077629e-05, "loss": 0.0296, "step": 8540 }, { "grad_norm": 0.2386656105518341, "learning_rate": 9.911861761248127e-05, "loss": 0.0267, "step": 8550 }, { "grad_norm": 0.33767643570899963, "learning_rate": 9.91147497811579e-05, "loss": 0.0288, "step": 8560 }, { "grad_norm": 0.33517271280288696, "learning_rate": 9.911087355746709e-05, "loss": 0.0236, "step": 8570 }, { "grad_norm": 0.2958509027957916, "learning_rate": 9.910698894207117e-05, "loss": 0.0252, "step": 8580 }, { "grad_norm": 0.34860607981681824, "learning_rate": 9.910309593563392e-05, "loss": 0.0261, "step": 8590 }, { "grad_norm": 0.2849890887737274, "learning_rate": 9.909919453882057e-05, "loss": 0.0245, "step": 8600 }, { "grad_norm": 0.2825022339820862, "learning_rate": 9.90952847522977e-05, "loss": 0.025, "step": 8610 }, { "grad_norm": 0.2755861282348633, "learning_rate": 9.909136657673346e-05, "loss": 0.0246, "step": 8620 }, { "grad_norm": 0.26101917028427124, "learning_rate": 9.908744001279731e-05, "loss": 0.0209, "step": 8630 }, { "grad_norm": 0.27829205989837646, "learning_rate": 9.90835050611602e-05, "loss": 0.0226, "step": 8640 }, { "grad_norm": 0.2305421233177185, "learning_rate": 9.90795617224945e-05, "loss": 0.0254, "step": 8650 }, { "grad_norm": 0.24428695440292358, "learning_rate": 9.907560999747405e-05, "loss": 0.0254, "step": 8660 }, { "grad_norm": 0.2561878561973572, "learning_rate": 9.907164988677408e-05, "loss": 0.0258, "step": 8670 }, { "grad_norm": 0.3324437439441681, "learning_rate": 9.906768139107124e-05, "loss": 0.026, "step": 8680 }, { "grad_norm": 0.24381320178508759, "learning_rate": 9.906370451104367e-05, "loss": 0.0266, "step": 8690 }, { "grad_norm": 0.238735169172287, "learning_rate": 9.905971924737088e-05, "loss": 0.0238, "step": 8700 }, { "grad_norm": 0.2624504566192627, "learning_rate": 9.905572560073387e-05, "loss": 0.0227, "step": 8710 }, { "grad_norm": 0.2958982586860657, "learning_rate": 9.905172357181501e-05, "loss": 0.0204, "step": 8720 }, { "grad_norm": 0.2829151451587677, "learning_rate": 9.904771316129817e-05, "loss": 0.0271, "step": 8730 }, { "grad_norm": 0.21042770147323608, "learning_rate": 9.904369436986862e-05, "loss": 0.0242, "step": 8740 }, { "grad_norm": 0.2611141502857208, "learning_rate": 9.903966719821303e-05, "loss": 0.0269, "step": 8750 }, { "grad_norm": 0.28635498881340027, "learning_rate": 9.903563164701956e-05, "loss": 0.0234, "step": 8760 }, { "grad_norm": 0.2497563511133194, "learning_rate": 9.903158771697778e-05, "loss": 0.0214, "step": 8770 }, { "grad_norm": 0.3599894344806671, "learning_rate": 9.902753540877867e-05, "loss": 0.0237, "step": 8780 }, { "grad_norm": 0.2163272500038147, "learning_rate": 9.902347472311466e-05, "loss": 0.0215, "step": 8790 }, { "grad_norm": 0.23424963653087616, "learning_rate": 9.901940566067962e-05, "loss": 0.0266, "step": 8800 }, { "grad_norm": 0.2432343065738678, "learning_rate": 9.901532822216883e-05, "loss": 0.0239, "step": 8810 }, { "grad_norm": 0.3356296718120575, "learning_rate": 9.901124240827904e-05, "loss": 0.0259, "step": 8820 }, { "grad_norm": 0.3217085003852844, "learning_rate": 9.900714821970835e-05, "loss": 0.0298, "step": 8830 }, { "grad_norm": 0.26747724413871765, "learning_rate": 9.900304565715641e-05, "loss": 0.027, "step": 8840 }, { "grad_norm": 0.2695053517818451, "learning_rate": 9.899893472132419e-05, "loss": 0.0292, "step": 8850 }, { "grad_norm": 0.28775542974472046, "learning_rate": 9.899481541291415e-05, "loss": 0.0247, "step": 8860 }, { "grad_norm": 0.22356580197811127, "learning_rate": 9.899068773263016e-05, "loss": 0.0263, "step": 8870 }, { "grad_norm": 0.2822844386100769, "learning_rate": 9.898655168117754e-05, "loss": 0.0228, "step": 8880 }, { "grad_norm": 0.36259081959724426, "learning_rate": 9.898240725926302e-05, "loss": 0.0265, "step": 8890 }, { "grad_norm": 0.22561979293823242, "learning_rate": 9.897825446759478e-05, "loss": 0.021, "step": 8900 }, { "grad_norm": 0.2395559549331665, "learning_rate": 9.897409330688241e-05, "loss": 0.0235, "step": 8910 }, { "grad_norm": 0.22173504531383514, "learning_rate": 9.896992377783692e-05, "loss": 0.0215, "step": 8920 }, { "grad_norm": 0.2772115468978882, "learning_rate": 9.89657458811708e-05, "loss": 0.0235, "step": 8930 }, { "grad_norm": 0.26467689871788025, "learning_rate": 9.896155961759792e-05, "loss": 0.0241, "step": 8940 }, { "grad_norm": 0.3599916100502014, "learning_rate": 9.895736498783361e-05, "loss": 0.0274, "step": 8950 }, { "grad_norm": 0.316063791513443, "learning_rate": 9.895316199259462e-05, "loss": 0.0308, "step": 8960 }, { "grad_norm": 0.2285536229610443, "learning_rate": 9.894895063259909e-05, "loss": 0.0253, "step": 8970 }, { "grad_norm": 0.2411327213048935, "learning_rate": 9.894473090856667e-05, "loss": 0.0266, "step": 8980 }, { "grad_norm": 0.3267722725868225, "learning_rate": 9.894050282121839e-05, "loss": 0.0268, "step": 8990 }, { "grad_norm": 0.18914009630680084, "learning_rate": 9.893626637127668e-05, "loss": 0.03, "step": 9000 }, { "grad_norm": 0.2581503093242645, "learning_rate": 9.893202155946546e-05, "loss": 0.0278, "step": 9010 }, { "grad_norm": 0.2972189784049988, "learning_rate": 9.892776838651006e-05, "loss": 0.0222, "step": 9020 }, { "grad_norm": 0.3232499659061432, "learning_rate": 9.892350685313722e-05, "loss": 0.0268, "step": 9030 }, { "grad_norm": 0.23038552701473236, "learning_rate": 9.891923696007513e-05, "loss": 0.026, "step": 9040 }, { "grad_norm": 0.25076839327812195, "learning_rate": 9.891495870805336e-05, "loss": 0.0224, "step": 9050 }, { "grad_norm": 0.2655271291732788, "learning_rate": 9.891067209780298e-05, "loss": 0.0224, "step": 9060 }, { "grad_norm": 0.2816680669784546, "learning_rate": 9.890637713005646e-05, "loss": 0.0243, "step": 9070 }, { "grad_norm": 0.2317475825548172, "learning_rate": 9.890207380554767e-05, "loss": 0.0247, "step": 9080 }, { "grad_norm": 0.2968375086784363, "learning_rate": 9.889776212501196e-05, "loss": 0.0259, "step": 9090 }, { "grad_norm": 0.3022483289241791, "learning_rate": 9.889344208918605e-05, "loss": 0.0249, "step": 9100 }, { "grad_norm": 0.26926133036613464, "learning_rate": 9.888911369880812e-05, "loss": 0.0256, "step": 9110 }, { "grad_norm": 0.2868924140930176, "learning_rate": 9.888477695461777e-05, "loss": 0.0258, "step": 9120 }, { "grad_norm": 0.29715394973754883, "learning_rate": 9.888043185735607e-05, "loss": 0.0229, "step": 9130 }, { "grad_norm": 0.2629702687263489, "learning_rate": 9.887607840776542e-05, "loss": 0.0237, "step": 9140 }, { "grad_norm": 0.23283155262470245, "learning_rate": 9.887171660658975e-05, "loss": 0.0208, "step": 9150 }, { "grad_norm": 0.25454244017601013, "learning_rate": 9.886734645457435e-05, "loss": 0.0232, "step": 9160 }, { "grad_norm": 0.24043768644332886, "learning_rate": 9.886296795246597e-05, "loss": 0.0254, "step": 9170 }, { "grad_norm": 0.2633034288883209, "learning_rate": 9.885858110101276e-05, "loss": 0.0223, "step": 9180 }, { "grad_norm": 0.27087798714637756, "learning_rate": 9.885418590096434e-05, "loss": 0.0249, "step": 9190 }, { "grad_norm": 0.25369948148727417, "learning_rate": 9.88497823530717e-05, "loss": 0.0248, "step": 9200 }, { "grad_norm": 0.31804272532463074, "learning_rate": 9.884537045808732e-05, "loss": 0.0265, "step": 9210 }, { "grad_norm": 0.2958488166332245, "learning_rate": 9.884095021676502e-05, "loss": 0.0275, "step": 9220 }, { "grad_norm": 0.26592618227005005, "learning_rate": 9.883652162986017e-05, "loss": 0.0242, "step": 9230 }, { "grad_norm": 0.30972805619239807, "learning_rate": 9.883208469812943e-05, "loss": 0.0273, "step": 9240 }, { "grad_norm": 0.20547877252101898, "learning_rate": 9.882763942233098e-05, "loss": 0.0226, "step": 9250 }, { "grad_norm": 0.2545967102050781, "learning_rate": 9.882318580322441e-05, "loss": 0.0222, "step": 9260 }, { "grad_norm": 0.3048904240131378, "learning_rate": 9.881872384157067e-05, "loss": 0.0244, "step": 9270 }, { "grad_norm": 0.2371196150779724, "learning_rate": 9.881425353813225e-05, "loss": 0.0257, "step": 9280 }, { "grad_norm": 0.23665012419223785, "learning_rate": 9.880977489367296e-05, "loss": 0.0233, "step": 9290 }, { "grad_norm": 0.30012771487236023, "learning_rate": 9.88052879089581e-05, "loss": 0.0251, "step": 9300 }, { "grad_norm": 0.20596519112586975, "learning_rate": 9.880079258475434e-05, "loss": 0.0214, "step": 9310 }, { "grad_norm": 0.28430840373039246, "learning_rate": 9.879628892182985e-05, "loss": 0.0204, "step": 9320 }, { "grad_norm": 0.25995075702667236, "learning_rate": 9.879177692095416e-05, "loss": 0.0212, "step": 9330 }, { "grad_norm": 0.26969489455223083, "learning_rate": 9.878725658289825e-05, "loss": 0.0204, "step": 9340 }, { "grad_norm": 0.272656112909317, "learning_rate": 9.878272790843454e-05, "loss": 0.023, "step": 9350 }, { "grad_norm": 0.2743775248527527, "learning_rate": 9.877819089833682e-05, "loss": 0.0222, "step": 9360 }, { "grad_norm": 0.22833365201950073, "learning_rate": 9.877364555338038e-05, "loss": 0.0198, "step": 9370 }, { "grad_norm": 0.2777186930179596, "learning_rate": 9.876909187434186e-05, "loss": 0.0236, "step": 9380 }, { "grad_norm": 0.3321308195590973, "learning_rate": 9.876452986199939e-05, "loss": 0.021, "step": 9390 }, { "grad_norm": 0.2805924117565155, "learning_rate": 9.875995951713248e-05, "loss": 0.0228, "step": 9400 }, { "grad_norm": 0.3278806805610657, "learning_rate": 9.875538084052207e-05, "loss": 0.0192, "step": 9410 }, { "grad_norm": 0.29617685079574585, "learning_rate": 9.875079383295053e-05, "loss": 0.0226, "step": 9420 }, { "grad_norm": 0.268395334482193, "learning_rate": 9.874619849520167e-05, "loss": 0.025, "step": 9430 }, { "grad_norm": 0.32225704193115234, "learning_rate": 9.874159482806069e-05, "loss": 0.021, "step": 9440 }, { "grad_norm": 0.31073135137557983, "learning_rate": 9.873698283231426e-05, "loss": 0.0235, "step": 9450 }, { "grad_norm": 0.24044065177440643, "learning_rate": 9.87323625087504e-05, "loss": 0.0216, "step": 9460 }, { "grad_norm": 0.2841043770313263, "learning_rate": 9.872773385815863e-05, "loss": 0.0239, "step": 9470 }, { "grad_norm": 0.314447283744812, "learning_rate": 9.872309688132986e-05, "loss": 0.0256, "step": 9480 }, { "grad_norm": 0.21195895969867706, "learning_rate": 9.871845157905639e-05, "loss": 0.0248, "step": 9490 }, { "grad_norm": 0.2903771996498108, "learning_rate": 9.871379795213201e-05, "loss": 0.0244, "step": 9500 }, { "grad_norm": 0.2178994119167328, "learning_rate": 9.87091360013519e-05, "loss": 0.0254, "step": 9510 }, { "grad_norm": 0.23022277653217316, "learning_rate": 9.870446572751262e-05, "loss": 0.0206, "step": 9520 }, { "grad_norm": 0.29495418071746826, "learning_rate": 9.869978713141224e-05, "loss": 0.0187, "step": 9530 }, { "grad_norm": 0.25011539459228516, "learning_rate": 9.869510021385016e-05, "loss": 0.0226, "step": 9540 }, { "grad_norm": 0.2388850301504135, "learning_rate": 9.869040497562727e-05, "loss": 0.0267, "step": 9550 }, { "grad_norm": 0.25706472992897034, "learning_rate": 9.868570141754587e-05, "loss": 0.0238, "step": 9560 }, { "grad_norm": 0.26016682386398315, "learning_rate": 9.868098954040965e-05, "loss": 0.0223, "step": 9570 }, { "grad_norm": 0.2529788017272949, "learning_rate": 9.867626934502374e-05, "loss": 0.0249, "step": 9580 }, { "grad_norm": 0.28706181049346924, "learning_rate": 9.86715408321947e-05, "loss": 0.026, "step": 9590 }, { "grad_norm": 0.31187182664871216, "learning_rate": 9.86668040027305e-05, "loss": 0.0207, "step": 9600 }, { "grad_norm": 0.21323713660240173, "learning_rate": 9.866205885744053e-05, "loss": 0.0233, "step": 9610 }, { "grad_norm": 0.21219390630722046, "learning_rate": 9.865730539713563e-05, "loss": 0.0188, "step": 9620 }, { "grad_norm": 0.26316219568252563, "learning_rate": 9.8652543622628e-05, "loss": 0.0206, "step": 9630 }, { "grad_norm": 0.2655700743198395, "learning_rate": 9.864777353473132e-05, "loss": 0.0255, "step": 9640 }, { "grad_norm": 0.2587810158729553, "learning_rate": 9.864299513426068e-05, "loss": 0.0247, "step": 9650 }, { "grad_norm": 0.2956332862377167, "learning_rate": 9.863820842203254e-05, "loss": 0.0238, "step": 9660 }, { "grad_norm": 0.23464709520339966, "learning_rate": 9.863341339886483e-05, "loss": 0.0196, "step": 9670 }, { "grad_norm": 0.22518271207809448, "learning_rate": 9.86286100655769e-05, "loss": 0.0197, "step": 9680 }, { "grad_norm": 0.2332381159067154, "learning_rate": 9.862379842298953e-05, "loss": 0.0235, "step": 9690 }, { "grad_norm": 0.3114089369773865, "learning_rate": 9.861897847192485e-05, "loss": 0.022, "step": 9700 }, { "grad_norm": 0.3150867819786072, "learning_rate": 9.86141502132065e-05, "loss": 0.0219, "step": 9710 }, { "grad_norm": 0.2645142078399658, "learning_rate": 9.860931364765946e-05, "loss": 0.0221, "step": 9720 }, { "grad_norm": 0.3106870949268341, "learning_rate": 9.860446877611021e-05, "loss": 0.0218, "step": 9730 }, { "grad_norm": 0.23792165517807007, "learning_rate": 9.859961559938655e-05, "loss": 0.0231, "step": 9740 }, { "grad_norm": 0.304152250289917, "learning_rate": 9.85947541183178e-05, "loss": 0.025, "step": 9750 }, { "grad_norm": 0.22986872494220734, "learning_rate": 9.858988433373463e-05, "loss": 0.0228, "step": 9760 }, { "grad_norm": 0.2393023669719696, "learning_rate": 9.858500624646918e-05, "loss": 0.0205, "step": 9770 }, { "grad_norm": 0.2337121218442917, "learning_rate": 9.858011985735497e-05, "loss": 0.0196, "step": 9780 }, { "grad_norm": 0.2813776433467865, "learning_rate": 9.857522516722693e-05, "loss": 0.0207, "step": 9790 }, { "grad_norm": 0.22624768316745758, "learning_rate": 9.857032217692145e-05, "loss": 0.0207, "step": 9800 }, { "grad_norm": 0.23102273046970367, "learning_rate": 9.856541088727631e-05, "loss": 0.0291, "step": 9810 }, { "grad_norm": 0.2624845504760742, "learning_rate": 9.856049129913072e-05, "loss": 0.0238, "step": 9820 }, { "grad_norm": 0.22061438858509064, "learning_rate": 9.85555634133253e-05, "loss": 0.0248, "step": 9830 }, { "grad_norm": 0.19422325491905212, "learning_rate": 9.855062723070208e-05, "loss": 0.0233, "step": 9840 }, { "grad_norm": 0.3682219088077545, "learning_rate": 9.854568275210454e-05, "loss": 0.0263, "step": 9850 }, { "grad_norm": 0.25819966197013855, "learning_rate": 9.854072997837754e-05, "loss": 0.0234, "step": 9860 }, { "grad_norm": 0.23041947185993195, "learning_rate": 9.853576891036737e-05, "loss": 0.018, "step": 9870 }, { "grad_norm": 0.2413894683122635, "learning_rate": 9.853079954892177e-05, "loss": 0.0204, "step": 9880 }, { "grad_norm": 0.2752460241317749, "learning_rate": 9.852582189488983e-05, "loss": 0.0198, "step": 9890 }, { "grad_norm": 0.25823408365249634, "learning_rate": 9.852083594912212e-05, "loss": 0.0203, "step": 9900 }, { "grad_norm": 0.27497896552085876, "learning_rate": 9.851584171247058e-05, "loss": 0.0222, "step": 9910 }, { "grad_norm": 0.1870100349187851, "learning_rate": 9.851083918578863e-05, "loss": 0.0196, "step": 9920 }, { "grad_norm": 0.22232802212238312, "learning_rate": 9.850582836993103e-05, "loss": 0.0212, "step": 9930 }, { "grad_norm": 0.22527024149894714, "learning_rate": 9.850080926575397e-05, "loss": 0.0177, "step": 9940 }, { "grad_norm": 0.25716665387153625, "learning_rate": 9.849578187411515e-05, "loss": 0.0272, "step": 9950 }, { "grad_norm": 0.26628026366233826, "learning_rate": 9.849074619587354e-05, "loss": 0.0226, "step": 9960 }, { "grad_norm": 0.20828302204608917, "learning_rate": 9.848570223188964e-05, "loss": 0.0203, "step": 9970 }, { "grad_norm": 0.23057571053504944, "learning_rate": 9.848064998302531e-05, "loss": 0.0169, "step": 9980 }, { "grad_norm": 0.2186940312385559, "learning_rate": 9.847558945014386e-05, "loss": 0.0198, "step": 9990 }, { "grad_norm": 0.2939648926258087, "learning_rate": 9.847052063410996e-05, "loss": 0.0204, "step": 10000 }, { "grad_norm": 0.27488598227500916, "learning_rate": 9.846544353578977e-05, "loss": 0.0192, "step": 10010 }, { "grad_norm": 0.22887849807739258, "learning_rate": 9.846035815605081e-05, "loss": 0.0209, "step": 10020 }, { "grad_norm": 0.2603186070919037, "learning_rate": 9.845526449576204e-05, "loss": 0.0223, "step": 10030 }, { "grad_norm": 0.2629733383655548, "learning_rate": 9.845016255579383e-05, "loss": 0.0194, "step": 10040 }, { "grad_norm": 0.23498961329460144, "learning_rate": 9.844505233701794e-05, "loss": 0.0205, "step": 10050 }, { "grad_norm": 0.20002023875713348, "learning_rate": 9.843993384030757e-05, "loss": 0.0194, "step": 10060 }, { "grad_norm": 0.22101539373397827, "learning_rate": 9.843480706653737e-05, "loss": 0.0193, "step": 10070 }, { "grad_norm": 0.2326389104127884, "learning_rate": 9.84296720165833e-05, "loss": 0.0269, "step": 10080 }, { "grad_norm": 0.28777652978897095, "learning_rate": 9.842452869132286e-05, "loss": 0.0206, "step": 10090 }, { "grad_norm": 0.3235422670841217, "learning_rate": 9.841937709163489e-05, "loss": 0.0209, "step": 10100 }, { "grad_norm": 0.2043895423412323, "learning_rate": 9.841421721839962e-05, "loss": 0.0188, "step": 10110 }, { "grad_norm": 0.2644239068031311, "learning_rate": 9.840904907249879e-05, "loss": 0.0214, "step": 10120 }, { "grad_norm": 0.25609707832336426, "learning_rate": 9.840387265481545e-05, "loss": 0.0222, "step": 10130 }, { "grad_norm": 0.21332962810993195, "learning_rate": 9.839868796623411e-05, "loss": 0.0192, "step": 10140 }, { "grad_norm": 0.2738405764102936, "learning_rate": 9.839349500764072e-05, "loss": 0.0184, "step": 10150 }, { "grad_norm": 0.3147866725921631, "learning_rate": 9.83882937799226e-05, "loss": 0.0208, "step": 10160 }, { "grad_norm": 0.288007527589798, "learning_rate": 9.838308428396849e-05, "loss": 0.0227, "step": 10170 }, { "grad_norm": 0.30472269654273987, "learning_rate": 9.837786652066854e-05, "loss": 0.0231, "step": 10180 }, { "grad_norm": 0.2523372769355774, "learning_rate": 9.837264049091437e-05, "loss": 0.0224, "step": 10190 }, { "grad_norm": 0.2573886811733246, "learning_rate": 9.836740619559893e-05, "loss": 0.0222, "step": 10200 }, { "grad_norm": 0.2550314664840698, "learning_rate": 9.836216363561659e-05, "loss": 0.0257, "step": 10210 }, { "grad_norm": 0.32082676887512207, "learning_rate": 9.835691281186322e-05, "loss": 0.0268, "step": 10220 }, { "grad_norm": 0.3452287018299103, "learning_rate": 9.8351653725236e-05, "loss": 0.0237, "step": 10230 }, { "grad_norm": 0.2591940760612488, "learning_rate": 9.83463863766336e-05, "loss": 0.021, "step": 10240 }, { "grad_norm": 0.41475915908813477, "learning_rate": 9.834111076695602e-05, "loss": 0.0223, "step": 10250 }, { "grad_norm": 0.2581203281879425, "learning_rate": 9.833582689710477e-05, "loss": 0.0259, "step": 10260 }, { "grad_norm": 0.26569345593452454, "learning_rate": 9.833053476798268e-05, "loss": 0.0255, "step": 10270 }, { "grad_norm": 0.30389857292175293, "learning_rate": 9.832523438049404e-05, "loss": 0.0259, "step": 10280 }, { "grad_norm": 0.2335321456193924, "learning_rate": 9.831992573554454e-05, "loss": 0.0197, "step": 10290 }, { "grad_norm": 0.19948722422122955, "learning_rate": 9.831460883404128e-05, "loss": 0.0218, "step": 10300 }, { "grad_norm": 0.23150113224983215, "learning_rate": 9.830928367689278e-05, "loss": 0.0186, "step": 10310 }, { "grad_norm": 0.20933547616004944, "learning_rate": 9.830395026500896e-05, "loss": 0.0175, "step": 10320 }, { "grad_norm": 0.2277691513299942, "learning_rate": 9.829860859930115e-05, "loss": 0.021, "step": 10330 }, { "grad_norm": 0.20719978213310242, "learning_rate": 9.829325868068212e-05, "loss": 0.0186, "step": 10340 }, { "grad_norm": 0.23335568606853485, "learning_rate": 9.8287900510066e-05, "loss": 0.0244, "step": 10350 }, { "grad_norm": 0.2935087978839874, "learning_rate": 9.828253408836834e-05, "loss": 0.0236, "step": 10360 }, { "grad_norm": 0.21509674191474915, "learning_rate": 9.827715941650615e-05, "loss": 0.0202, "step": 10370 }, { "grad_norm": 0.2749275863170624, "learning_rate": 9.82717764953978e-05, "loss": 0.0183, "step": 10380 }, { "grad_norm": 0.29683718085289, "learning_rate": 9.826638532596308e-05, "loss": 0.0199, "step": 10390 }, { "grad_norm": 0.2656606137752533, "learning_rate": 9.82609859091232e-05, "loss": 0.0201, "step": 10400 }, { "grad_norm": 0.27659276127815247, "learning_rate": 9.825557824580076e-05, "loss": 0.0203, "step": 10410 }, { "grad_norm": 0.25727754831314087, "learning_rate": 9.82501623369198e-05, "loss": 0.0162, "step": 10420 }, { "grad_norm": 0.31244784593582153, "learning_rate": 9.824473818340574e-05, "loss": 0.0214, "step": 10430 }, { "grad_norm": 0.21961535513401031, "learning_rate": 9.823930578618541e-05, "loss": 0.02, "step": 10440 }, { "grad_norm": 0.21688979864120483, "learning_rate": 9.823386514618709e-05, "loss": 0.0193, "step": 10450 }, { "grad_norm": 0.20757648348808289, "learning_rate": 9.82284162643404e-05, "loss": 0.0203, "step": 10460 }, { "grad_norm": 0.2294018417596817, "learning_rate": 9.822295914157642e-05, "loss": 0.0198, "step": 10470 }, { "grad_norm": 0.26489317417144775, "learning_rate": 9.821749377882763e-05, "loss": 0.0188, "step": 10480 }, { "grad_norm": 0.24768415093421936, "learning_rate": 9.821202017702791e-05, "loss": 0.0193, "step": 10490 }, { "grad_norm": 0.33312082290649414, "learning_rate": 9.820653833711253e-05, "loss": 0.0197, "step": 10500 }, { "grad_norm": 0.25030675530433655, "learning_rate": 9.820104826001822e-05, "loss": 0.0217, "step": 10510 }, { "grad_norm": 0.2725224792957306, "learning_rate": 9.819554994668305e-05, "loss": 0.0227, "step": 10520 }, { "grad_norm": 0.1868051290512085, "learning_rate": 9.819004339804654e-05, "loss": 0.0202, "step": 10530 }, { "grad_norm": 0.209153950214386, "learning_rate": 9.818452861504961e-05, "loss": 0.0207, "step": 10540 }, { "grad_norm": 0.22917388379573822, "learning_rate": 9.81790055986346e-05, "loss": 0.0238, "step": 10550 }, { "grad_norm": 0.274032860994339, "learning_rate": 9.817347434974523e-05, "loss": 0.0197, "step": 10560 }, { "grad_norm": 0.2764982581138611, "learning_rate": 9.816793486932664e-05, "loss": 0.0209, "step": 10570 }, { "grad_norm": 0.29091471433639526, "learning_rate": 9.816238715832538e-05, "loss": 0.0187, "step": 10580 }, { "grad_norm": 0.25286200642585754, "learning_rate": 9.815683121768939e-05, "loss": 0.0214, "step": 10590 }, { "grad_norm": 0.3519565463066101, "learning_rate": 9.815126704836804e-05, "loss": 0.0217, "step": 10600 }, { "grad_norm": 0.22415044903755188, "learning_rate": 9.81456946513121e-05, "loss": 0.0229, "step": 10610 }, { "grad_norm": 0.26799091696739197, "learning_rate": 9.814011402747373e-05, "loss": 0.0268, "step": 10620 }, { "grad_norm": 0.1925109475851059, "learning_rate": 9.813452517780651e-05, "loss": 0.0209, "step": 10630 }, { "grad_norm": 0.18273292481899261, "learning_rate": 9.81289281032654e-05, "loss": 0.0229, "step": 10640 }, { "grad_norm": 0.15176481008529663, "learning_rate": 9.812332280480683e-05, "loss": 0.0199, "step": 10650 }, { "grad_norm": 0.28883445262908936, "learning_rate": 9.811770928338854e-05, "loss": 0.0202, "step": 10660 }, { "grad_norm": 0.18953080475330353, "learning_rate": 9.811208753996979e-05, "loss": 0.0184, "step": 10670 }, { "grad_norm": 0.21895815432071686, "learning_rate": 9.810645757551113e-05, "loss": 0.0156, "step": 10680 }, { "grad_norm": 0.21101528406143188, "learning_rate": 9.810081939097459e-05, "loss": 0.0176, "step": 10690 }, { "grad_norm": 0.2178855985403061, "learning_rate": 9.809517298732356e-05, "loss": 0.0212, "step": 10700 }, { "grad_norm": 0.21389643847942352, "learning_rate": 9.80895183655229e-05, "loss": 0.017, "step": 10710 }, { "grad_norm": 0.2984575629234314, "learning_rate": 9.808385552653877e-05, "loss": 0.0227, "step": 10720 }, { "grad_norm": 0.19534827768802643, "learning_rate": 9.807818447133886e-05, "loss": 0.0226, "step": 10730 }, { "grad_norm": 0.2492058426141739, "learning_rate": 9.807250520089215e-05, "loss": 0.0184, "step": 10740 }, { "grad_norm": 0.2333592027425766, "learning_rate": 9.806681771616908e-05, "loss": 0.019, "step": 10750 }, { "grad_norm": 0.24145835638046265, "learning_rate": 9.80611220181415e-05, "loss": 0.0198, "step": 10760 }, { "grad_norm": 0.19235028326511383, "learning_rate": 9.805541810778264e-05, "loss": 0.0206, "step": 10770 }, { "grad_norm": 0.1690615564584732, "learning_rate": 9.804970598606716e-05, "loss": 0.0175, "step": 10780 }, { "grad_norm": 0.24259620904922485, "learning_rate": 9.804398565397106e-05, "loss": 0.0245, "step": 10790 }, { "grad_norm": 0.32438504695892334, "learning_rate": 9.803825711247183e-05, "loss": 0.02, "step": 10800 }, { "grad_norm": 0.2625963091850281, "learning_rate": 9.803252036254831e-05, "loss": 0.0163, "step": 10810 }, { "grad_norm": 0.29829689860343933, "learning_rate": 9.802677540518076e-05, "loss": 0.0212, "step": 10820 }, { "grad_norm": 0.3267967402935028, "learning_rate": 9.802102224135081e-05, "loss": 0.0223, "step": 10830 }, { "grad_norm": 0.22496075928211212, "learning_rate": 9.801526087204155e-05, "loss": 0.0204, "step": 10840 }, { "grad_norm": 0.22666290402412415, "learning_rate": 9.800949129823743e-05, "loss": 0.0188, "step": 10850 }, { "grad_norm": 0.24709749221801758, "learning_rate": 9.80037135209243e-05, "loss": 0.0219, "step": 10860 }, { "grad_norm": 0.18393516540527344, "learning_rate": 9.799792754108946e-05, "loss": 0.0206, "step": 10870 }, { "grad_norm": 0.23902221024036407, "learning_rate": 9.799213335972152e-05, "loss": 0.0201, "step": 10880 }, { "grad_norm": 0.22036035358905792, "learning_rate": 9.798633097781058e-05, "loss": 0.0233, "step": 10890 }, { "grad_norm": 0.16045823693275452, "learning_rate": 9.79805203963481e-05, "loss": 0.0174, "step": 10900 }, { "grad_norm": 0.25288039445877075, "learning_rate": 9.797470161632697e-05, "loss": 0.02, "step": 10910 }, { "grad_norm": 0.19493606686592102, "learning_rate": 9.796887463874145e-05, "loss": 0.0221, "step": 10920 }, { "grad_norm": 0.20628218352794647, "learning_rate": 9.796303946458718e-05, "loss": 0.0201, "step": 10930 }, { "grad_norm": 0.21774351596832275, "learning_rate": 9.795719609486127e-05, "loss": 0.0204, "step": 10940 }, { "grad_norm": 0.20591984689235687, "learning_rate": 9.795134453056219e-05, "loss": 0.0175, "step": 10950 }, { "grad_norm": 0.23972801864147186, "learning_rate": 9.794548477268979e-05, "loss": 0.0178, "step": 10960 }, { "grad_norm": 0.21541334688663483, "learning_rate": 9.793961682224537e-05, "loss": 0.0164, "step": 10970 }, { "grad_norm": 0.2511148750782013, "learning_rate": 9.793374068023156e-05, "loss": 0.0208, "step": 10980 }, { "grad_norm": 0.25219643115997314, "learning_rate": 9.792785634765247e-05, "loss": 0.0221, "step": 10990 }, { "grad_norm": 0.24498599767684937, "learning_rate": 9.792196382551357e-05, "loss": 0.0203, "step": 11000 }, { "grad_norm": 0.24652181565761566, "learning_rate": 9.791606311482171e-05, "loss": 0.0172, "step": 11010 }, { "grad_norm": 0.1450411081314087, "learning_rate": 9.791015421658518e-05, "loss": 0.018, "step": 11020 }, { "grad_norm": 0.2582523226737976, "learning_rate": 9.790423713181362e-05, "loss": 0.0205, "step": 11030 }, { "grad_norm": 0.16003862023353577, "learning_rate": 9.789831186151814e-05, "loss": 0.0178, "step": 11040 }, { "grad_norm": 0.26337435841560364, "learning_rate": 9.789237840671118e-05, "loss": 0.0199, "step": 11050 }, { "grad_norm": 0.20410926640033722, "learning_rate": 9.78864367684066e-05, "loss": 0.0179, "step": 11060 }, { "grad_norm": 0.24945813417434692, "learning_rate": 9.788048694761968e-05, "loss": 0.0189, "step": 11070 }, { "grad_norm": 0.23861199617385864, "learning_rate": 9.787452894536709e-05, "loss": 0.0201, "step": 11080 }, { "grad_norm": 0.26044031977653503, "learning_rate": 9.786856276266685e-05, "loss": 0.0178, "step": 11090 }, { "grad_norm": 0.24907591938972473, "learning_rate": 9.786258840053845e-05, "loss": 0.0219, "step": 11100 }, { "grad_norm": 0.23105935752391815, "learning_rate": 9.785660586000273e-05, "loss": 0.0199, "step": 11110 }, { "grad_norm": 0.29685473442077637, "learning_rate": 9.785061514208196e-05, "loss": 0.0184, "step": 11120 }, { "grad_norm": 0.34990617632865906, "learning_rate": 9.784461624779977e-05, "loss": 0.0205, "step": 11130 }, { "grad_norm": 0.23200103640556335, "learning_rate": 9.783860917818123e-05, "loss": 0.0212, "step": 11140 }, { "grad_norm": 0.2587655782699585, "learning_rate": 9.783259393425277e-05, "loss": 0.0211, "step": 11150 }, { "grad_norm": 0.1354336142539978, "learning_rate": 9.782657051704221e-05, "loss": 0.0159, "step": 11160 }, { "grad_norm": 0.209062859416008, "learning_rate": 9.782053892757883e-05, "loss": 0.0175, "step": 11170 }, { "grad_norm": 0.2994021773338318, "learning_rate": 9.781449916689324e-05, "loss": 0.0208, "step": 11180 }, { "grad_norm": 0.3047916889190674, "learning_rate": 9.780845123601746e-05, "loss": 0.0186, "step": 11190 }, { "grad_norm": 0.2150055468082428, "learning_rate": 9.780239513598492e-05, "loss": 0.0192, "step": 11200 }, { "grad_norm": 0.43727704882621765, "learning_rate": 9.779633086783047e-05, "loss": 0.0238, "step": 11210 }, { "grad_norm": 0.29295340180397034, "learning_rate": 9.779025843259031e-05, "loss": 0.0186, "step": 11220 }, { "grad_norm": 0.2038799226284027, "learning_rate": 9.778417783130204e-05, "loss": 0.0182, "step": 11230 }, { "grad_norm": 0.21761058270931244, "learning_rate": 9.777808906500468e-05, "loss": 0.0172, "step": 11240 }, { "grad_norm": 0.1858905851840973, "learning_rate": 9.777199213473862e-05, "loss": 0.0188, "step": 11250 }, { "grad_norm": 0.23797602951526642, "learning_rate": 9.77658870415457e-05, "loss": 0.0218, "step": 11260 }, { "grad_norm": 0.33256128430366516, "learning_rate": 9.775977378646906e-05, "loss": 0.0253, "step": 11270 }, { "grad_norm": 0.2266288548707962, "learning_rate": 9.775365237055331e-05, "loss": 0.0177, "step": 11280 }, { "grad_norm": 0.30771198868751526, "learning_rate": 9.774752279484445e-05, "loss": 0.0177, "step": 11290 }, { "grad_norm": 0.19915752112865448, "learning_rate": 9.774138506038984e-05, "loss": 0.0179, "step": 11300 }, { "grad_norm": 0.2763212323188782, "learning_rate": 9.773523916823826e-05, "loss": 0.0228, "step": 11310 }, { "grad_norm": 0.2468460500240326, "learning_rate": 9.772908511943986e-05, "loss": 0.0163, "step": 11320 }, { "grad_norm": 0.1837548315525055, "learning_rate": 9.77229229150462e-05, "loss": 0.0174, "step": 11330 }, { "grad_norm": 0.23041242361068726, "learning_rate": 9.771675255611024e-05, "loss": 0.0201, "step": 11340 }, { "grad_norm": 0.24687428772449493, "learning_rate": 9.771057404368632e-05, "loss": 0.0201, "step": 11350 }, { "grad_norm": 0.21581122279167175, "learning_rate": 9.770438737883018e-05, "loss": 0.0172, "step": 11360 }, { "grad_norm": 0.21284006536006927, "learning_rate": 9.769819256259898e-05, "loss": 0.0189, "step": 11370 }, { "grad_norm": 0.3252716362476349, "learning_rate": 9.769198959605119e-05, "loss": 0.0215, "step": 11380 }, { "grad_norm": 0.26655328273773193, "learning_rate": 9.768577848024678e-05, "loss": 0.0214, "step": 11390 }, { "grad_norm": 0.30771878361701965, "learning_rate": 9.767955921624702e-05, "loss": 0.0202, "step": 11400 }, { "grad_norm": 0.3264099359512329, "learning_rate": 9.767333180511465e-05, "loss": 0.0222, "step": 11410 }, { "grad_norm": 0.2874441146850586, "learning_rate": 9.766709624791373e-05, "loss": 0.0248, "step": 11420 }, { "grad_norm": 0.22969099879264832, "learning_rate": 9.766085254570975e-05, "loss": 0.0239, "step": 11430 }, { "grad_norm": 0.3268851637840271, "learning_rate": 9.76546006995696e-05, "loss": 0.0212, "step": 11440 }, { "grad_norm": 0.23701435327529907, "learning_rate": 9.764834071056155e-05, "loss": 0.0178, "step": 11450 }, { "grad_norm": 0.2926957309246063, "learning_rate": 9.764207257975526e-05, "loss": 0.0185, "step": 11460 }, { "grad_norm": 0.2133902758359909, "learning_rate": 9.763579630822179e-05, "loss": 0.0205, "step": 11470 }, { "grad_norm": 0.2138747274875641, "learning_rate": 9.762951189703356e-05, "loss": 0.0179, "step": 11480 }, { "grad_norm": 0.20924456417560577, "learning_rate": 9.762321934726442e-05, "loss": 0.0161, "step": 11490 }, { "grad_norm": 0.20569387078285217, "learning_rate": 9.761691865998959e-05, "loss": 0.0206, "step": 11500 }, { "grad_norm": 0.2268054485321045, "learning_rate": 9.76106098362857e-05, "loss": 0.0184, "step": 11510 }, { "grad_norm": 0.25970736145973206, "learning_rate": 9.760429287723072e-05, "loss": 0.0206, "step": 11520 }, { "grad_norm": 0.22137345373630524, "learning_rate": 9.759796778390406e-05, "loss": 0.0182, "step": 11530 }, { "grad_norm": 0.18656471371650696, "learning_rate": 9.759163455738653e-05, "loss": 0.0203, "step": 11540 }, { "grad_norm": 0.22846952080726624, "learning_rate": 9.75852931987603e-05, "loss": 0.0163, "step": 11550 }, { "grad_norm": 0.19582560658454895, "learning_rate": 9.757894370910891e-05, "loss": 0.0215, "step": 11560 }, { "grad_norm": 0.24684782326221466, "learning_rate": 9.757258608951733e-05, "loss": 0.0202, "step": 11570 }, { "grad_norm": 0.23838120698928833, "learning_rate": 9.75662203410719e-05, "loss": 0.0167, "step": 11580 }, { "grad_norm": 0.27070775628089905, "learning_rate": 9.755984646486034e-05, "loss": 0.0159, "step": 11590 }, { "grad_norm": 0.2429112046957016, "learning_rate": 9.75534644619718e-05, "loss": 0.0155, "step": 11600 }, { "grad_norm": 0.28102198243141174, "learning_rate": 9.754707433349676e-05, "loss": 0.0192, "step": 11610 }, { "grad_norm": 0.2159099578857422, "learning_rate": 9.754067608052715e-05, "loss": 0.0227, "step": 11620 }, { "grad_norm": 0.19809971749782562, "learning_rate": 9.753426970415622e-05, "loss": 0.0161, "step": 11630 }, { "grad_norm": 0.2691294252872467, "learning_rate": 9.752785520547868e-05, "loss": 0.0182, "step": 11640 }, { "grad_norm": 0.279301255941391, "learning_rate": 9.752143258559056e-05, "loss": 0.0186, "step": 11650 }, { "grad_norm": 0.21483710408210754, "learning_rate": 9.751500184558933e-05, "loss": 0.0181, "step": 11660 }, { "grad_norm": 0.26444026827812195, "learning_rate": 9.750856298657383e-05, "loss": 0.0159, "step": 11670 }, { "grad_norm": 0.2858373820781708, "learning_rate": 9.750211600964428e-05, "loss": 0.0166, "step": 11680 }, { "grad_norm": 0.2655319571495056, "learning_rate": 9.749566091590226e-05, "loss": 0.0221, "step": 11690 }, { "grad_norm": 0.20442770421504974, "learning_rate": 9.748919770645083e-05, "loss": 0.0214, "step": 11700 }, { "grad_norm": 0.22846493124961853, "learning_rate": 9.748272638239432e-05, "loss": 0.0188, "step": 11710 }, { "grad_norm": 0.22037361562252045, "learning_rate": 9.747624694483855e-05, "loss": 0.0216, "step": 11720 }, { "grad_norm": 0.18531198799610138, "learning_rate": 9.746975939489065e-05, "loss": 0.0149, "step": 11730 }, { "grad_norm": 0.18271471560001373, "learning_rate": 9.746326373365918e-05, "loss": 0.017, "step": 11740 }, { "grad_norm": 0.274687796831131, "learning_rate": 9.745675996225403e-05, "loss": 0.0176, "step": 11750 }, { "grad_norm": 0.18602631986141205, "learning_rate": 9.745024808178657e-05, "loss": 0.0191, "step": 11760 }, { "grad_norm": 0.25692635774612427, "learning_rate": 9.744372809336947e-05, "loss": 0.0183, "step": 11770 }, { "grad_norm": 0.1961025595664978, "learning_rate": 9.743719999811682e-05, "loss": 0.0232, "step": 11780 }, { "grad_norm": 0.2340070605278015, "learning_rate": 9.743066379714412e-05, "loss": 0.0162, "step": 11790 }, { "grad_norm": 0.22419200837612152, "learning_rate": 9.74241194915682e-05, "loss": 0.0236, "step": 11800 }, { "grad_norm": 0.25545060634613037, "learning_rate": 9.741756708250731e-05, "loss": 0.0175, "step": 11810 }, { "grad_norm": 0.22371838986873627, "learning_rate": 9.741100657108109e-05, "loss": 0.0211, "step": 11820 }, { "grad_norm": 0.3355225920677185, "learning_rate": 9.740443795841054e-05, "loss": 0.0205, "step": 11830 }, { "grad_norm": 0.2511918246746063, "learning_rate": 9.739786124561805e-05, "loss": 0.0191, "step": 11840 }, { "grad_norm": 0.15254716575145721, "learning_rate": 9.73912764338274e-05, "loss": 0.0205, "step": 11850 }, { "grad_norm": 0.22108015418052673, "learning_rate": 9.738468352416377e-05, "loss": 0.0207, "step": 11860 }, { "grad_norm": 0.2089259922504425, "learning_rate": 9.737808251775369e-05, "loss": 0.0224, "step": 11870 }, { "grad_norm": 0.21650834381580353, "learning_rate": 9.737147341572512e-05, "loss": 0.0183, "step": 11880 }, { "grad_norm": 0.20833459496498108, "learning_rate": 9.736485621920735e-05, "loss": 0.0196, "step": 11890 }, { "grad_norm": 0.2317640781402588, "learning_rate": 9.735823092933108e-05, "loss": 0.0271, "step": 11900 }, { "grad_norm": 0.23941192030906677, "learning_rate": 9.735159754722838e-05, "loss": 0.0191, "step": 11910 }, { "grad_norm": 0.3287619948387146, "learning_rate": 9.734495607403275e-05, "loss": 0.0196, "step": 11920 }, { "grad_norm": 0.24900713562965393, "learning_rate": 9.733830651087901e-05, "loss": 0.0196, "step": 11930 }, { "grad_norm": 0.22000187635421753, "learning_rate": 9.733164885890338e-05, "loss": 0.0167, "step": 11940 }, { "grad_norm": 0.2324901819229126, "learning_rate": 9.732498311924349e-05, "loss": 0.0215, "step": 11950 }, { "grad_norm": 0.20142823457717896, "learning_rate": 9.731830929303833e-05, "loss": 0.0196, "step": 11960 }, { "grad_norm": 0.19510656595230103, "learning_rate": 9.731162738142827e-05, "loss": 0.0175, "step": 11970 }, { "grad_norm": 0.211184561252594, "learning_rate": 9.730493738555506e-05, "loss": 0.0224, "step": 11980 }, { "grad_norm": 0.2330097258090973, "learning_rate": 9.729823930656186e-05, "loss": 0.0203, "step": 11990 }, { "grad_norm": 0.23414331674575806, "learning_rate": 9.729153314559316e-05, "loss": 0.0198, "step": 12000 }, { "grad_norm": 0.22416187822818756, "learning_rate": 9.728481890379486e-05, "loss": 0.0221, "step": 12010 }, { "grad_norm": 0.22886383533477783, "learning_rate": 9.727809658231428e-05, "loss": 0.0202, "step": 12020 }, { "grad_norm": 0.2826075851917267, "learning_rate": 9.727136618230003e-05, "loss": 0.0204, "step": 12030 }, { "grad_norm": 0.25740987062454224, "learning_rate": 9.726462770490219e-05, "loss": 0.0204, "step": 12040 }, { "grad_norm": 0.25582218170166016, "learning_rate": 9.725788115127214e-05, "loss": 0.0167, "step": 12050 }, { "grad_norm": 0.2289591282606125, "learning_rate": 9.725112652256274e-05, "loss": 0.0175, "step": 12060 }, { "grad_norm": 0.19923928380012512, "learning_rate": 9.724436381992812e-05, "loss": 0.0182, "step": 12070 }, { "grad_norm": 0.22635547816753387, "learning_rate": 9.723759304452387e-05, "loss": 0.0187, "step": 12080 }, { "grad_norm": 0.21469198167324066, "learning_rate": 9.72308141975069e-05, "loss": 0.0166, "step": 12090 }, { "grad_norm": 0.17345571517944336, "learning_rate": 9.722402728003557e-05, "loss": 0.0182, "step": 12100 }, { "grad_norm": 0.19827012717723846, "learning_rate": 9.721723229326953e-05, "loss": 0.0169, "step": 12110 }, { "grad_norm": 0.3330991566181183, "learning_rate": 9.721042923836992e-05, "loss": 0.021, "step": 12120 }, { "grad_norm": 0.24948695302009583, "learning_rate": 9.720361811649914e-05, "loss": 0.018, "step": 12130 }, { "grad_norm": 0.19254659116268158, "learning_rate": 9.719679892882106e-05, "loss": 0.0149, "step": 12140 }, { "grad_norm": 0.2414981871843338, "learning_rate": 9.718997167650085e-05, "loss": 0.0167, "step": 12150 }, { "grad_norm": 0.21447575092315674, "learning_rate": 9.718313636070515e-05, "loss": 0.0192, "step": 12160 }, { "grad_norm": 0.2529369592666626, "learning_rate": 9.717629298260192e-05, "loss": 0.0195, "step": 12170 }, { "grad_norm": 0.27506980299949646, "learning_rate": 9.716944154336047e-05, "loss": 0.0169, "step": 12180 }, { "grad_norm": 0.1967649608850479, "learning_rate": 9.716258204415157e-05, "loss": 0.0176, "step": 12190 }, { "grad_norm": 0.24737775325775146, "learning_rate": 9.715571448614728e-05, "loss": 0.0219, "step": 12200 }, { "grad_norm": 0.27130183577537537, "learning_rate": 9.71488388705211e-05, "loss": 0.0185, "step": 12210 }, { "grad_norm": 0.281478613615036, "learning_rate": 9.714195519844788e-05, "loss": 0.0191, "step": 12220 }, { "grad_norm": 0.27363917231559753, "learning_rate": 9.713506347110386e-05, "loss": 0.017, "step": 12230 }, { "grad_norm": 0.20127913355827332, "learning_rate": 9.712816368966662e-05, "loss": 0.0168, "step": 12240 }, { "grad_norm": 0.3368346691131592, "learning_rate": 9.712125585531517e-05, "loss": 0.0207, "step": 12250 }, { "grad_norm": 0.22935083508491516, "learning_rate": 9.711433996922988e-05, "loss": 0.0193, "step": 12260 }, { "grad_norm": 0.21330472826957703, "learning_rate": 9.710741603259245e-05, "loss": 0.0217, "step": 12270 }, { "grad_norm": 0.20881007611751556, "learning_rate": 9.710048404658603e-05, "loss": 0.0186, "step": 12280 }, { "grad_norm": 0.17342618107795715, "learning_rate": 9.709354401239508e-05, "loss": 0.0183, "step": 12290 }, { "grad_norm": 0.23784968256950378, "learning_rate": 9.708659593120546e-05, "loss": 0.0192, "step": 12300 }, { "grad_norm": 0.23157134652137756, "learning_rate": 9.707963980420443e-05, "loss": 0.0207, "step": 12310 }, { "grad_norm": 0.172146737575531, "learning_rate": 9.707267563258058e-05, "loss": 0.0158, "step": 12320 }, { "grad_norm": 0.216389998793602, "learning_rate": 9.70657034175239e-05, "loss": 0.0173, "step": 12330 }, { "grad_norm": 0.20704789459705353, "learning_rate": 9.705872316022577e-05, "loss": 0.0208, "step": 12340 }, { "grad_norm": 0.20111405849456787, "learning_rate": 9.705173486187891e-05, "loss": 0.0179, "step": 12350 }, { "grad_norm": 0.25501254200935364, "learning_rate": 9.704473852367741e-05, "loss": 0.0247, "step": 12360 }, { "grad_norm": 0.23284225165843964, "learning_rate": 9.70377341468168e-05, "loss": 0.0191, "step": 12370 }, { "grad_norm": 0.24420854449272156, "learning_rate": 9.703072173249389e-05, "loss": 0.0185, "step": 12380 }, { "grad_norm": 0.22483965754508972, "learning_rate": 9.702370128190693e-05, "loss": 0.0161, "step": 12390 }, { "grad_norm": 0.19778388738632202, "learning_rate": 9.701667279625552e-05, "loss": 0.0213, "step": 12400 }, { "grad_norm": 0.19247694313526154, "learning_rate": 9.700963627674065e-05, "loss": 0.0169, "step": 12410 }, { "grad_norm": 0.2351117879152298, "learning_rate": 9.700259172456466e-05, "loss": 0.0172, "step": 12420 }, { "grad_norm": 0.16455838084220886, "learning_rate": 9.699553914093124e-05, "loss": 0.0157, "step": 12430 }, { "grad_norm": 0.18576236069202423, "learning_rate": 9.698847852704553e-05, "loss": 0.0165, "step": 12440 }, { "grad_norm": 0.22169683873653412, "learning_rate": 9.6981409884114e-05, "loss": 0.0249, "step": 12450 }, { "grad_norm": 0.4429340064525604, "learning_rate": 9.697433321334443e-05, "loss": 0.0177, "step": 12460 }, { "grad_norm": 0.2589166462421417, "learning_rate": 9.696724851594607e-05, "loss": 0.0219, "step": 12470 }, { "grad_norm": 0.1737583875656128, "learning_rate": 9.696015579312952e-05, "loss": 0.0167, "step": 12480 }, { "grad_norm": 0.20093324780464172, "learning_rate": 9.695305504610668e-05, "loss": 0.0162, "step": 12490 }, { "grad_norm": 0.22829070687294006, "learning_rate": 9.694594627609092e-05, "loss": 0.0164, "step": 12500 }, { "grad_norm": 0.2258264422416687, "learning_rate": 9.693882948429691e-05, "loss": 0.017, "step": 12510 }, { "grad_norm": 0.20375555753707886, "learning_rate": 9.693170467194071e-05, "loss": 0.0184, "step": 12520 }, { "grad_norm": 0.17012882232666016, "learning_rate": 9.692457184023977e-05, "loss": 0.0195, "step": 12530 }, { "grad_norm": 0.18955504894256592, "learning_rate": 9.691743099041291e-05, "loss": 0.017, "step": 12540 }, { "grad_norm": 0.25962018966674805, "learning_rate": 9.691028212368027e-05, "loss": 0.0167, "step": 12550 }, { "grad_norm": 0.1951056867837906, "learning_rate": 9.690312524126342e-05, "loss": 0.0151, "step": 12560 }, { "grad_norm": 0.19832569360733032, "learning_rate": 9.689596034438527e-05, "loss": 0.0162, "step": 12570 }, { "grad_norm": 0.20500408113002777, "learning_rate": 9.688878743427012e-05, "loss": 0.015, "step": 12580 }, { "grad_norm": 0.271946519613266, "learning_rate": 9.688160651214359e-05, "loss": 0.016, "step": 12590 }, { "grad_norm": 0.21125653386116028, "learning_rate": 9.687441757923273e-05, "loss": 0.0176, "step": 12600 }, { "grad_norm": 0.22015267610549927, "learning_rate": 9.68672206367659e-05, "loss": 0.0161, "step": 12610 }, { "grad_norm": 0.2658013105392456, "learning_rate": 9.686001568597291e-05, "loss": 0.0201, "step": 12620 }, { "grad_norm": 0.24626858532428741, "learning_rate": 9.685280272808486e-05, "loss": 0.0166, "step": 12630 }, { "grad_norm": 0.26217472553253174, "learning_rate": 9.684558176433424e-05, "loss": 0.0166, "step": 12640 }, { "grad_norm": 0.2128893882036209, "learning_rate": 9.683835279595495e-05, "loss": 0.0171, "step": 12650 }, { "grad_norm": 0.18179072439670563, "learning_rate": 9.683111582418216e-05, "loss": 0.0133, "step": 12660 }, { "grad_norm": 0.2710782587528229, "learning_rate": 9.682387085025254e-05, "loss": 0.0148, "step": 12670 }, { "grad_norm": 0.23871701955795288, "learning_rate": 9.681661787540401e-05, "loss": 0.0173, "step": 12680 }, { "grad_norm": 0.2869615852832794, "learning_rate": 9.680935690087593e-05, "loss": 0.0191, "step": 12690 }, { "grad_norm": 0.2417600452899933, "learning_rate": 9.680208792790901e-05, "loss": 0.0183, "step": 12700 }, { "grad_norm": 0.2003248929977417, "learning_rate": 9.679481095774529e-05, "loss": 0.0159, "step": 12710 }, { "grad_norm": 0.21244405210018158, "learning_rate": 9.678752599162822e-05, "loss": 0.0172, "step": 12720 }, { "grad_norm": 0.26162102818489075, "learning_rate": 9.678023303080259e-05, "loss": 0.0158, "step": 12730 }, { "grad_norm": 0.19478337466716766, "learning_rate": 9.677293207651459e-05, "loss": 0.0164, "step": 12740 }, { "grad_norm": 0.2510935068130493, "learning_rate": 9.676562313001173e-05, "loss": 0.0165, "step": 12750 }, { "grad_norm": 0.19088700413703918, "learning_rate": 9.675830619254293e-05, "loss": 0.0176, "step": 12760 }, { "grad_norm": 0.20341435074806213, "learning_rate": 9.675098126535843e-05, "loss": 0.0195, "step": 12770 }, { "grad_norm": 0.21900177001953125, "learning_rate": 9.674364834970988e-05, "loss": 0.0205, "step": 12780 }, { "grad_norm": 0.25362831354141235, "learning_rate": 9.673630744685028e-05, "loss": 0.0156, "step": 12790 }, { "grad_norm": 0.20551882684230804, "learning_rate": 9.672895855803397e-05, "loss": 0.0182, "step": 12800 }, { "grad_norm": 0.2756454646587372, "learning_rate": 9.672160168451667e-05, "loss": 0.0158, "step": 12810 }, { "grad_norm": 0.2876286506652832, "learning_rate": 9.671423682755549e-05, "loss": 0.0188, "step": 12820 }, { "grad_norm": 0.18926411867141724, "learning_rate": 9.670686398840888e-05, "loss": 0.0171, "step": 12830 }, { "grad_norm": 0.1908763200044632, "learning_rate": 9.669948316833664e-05, "loss": 0.0162, "step": 12840 }, { "grad_norm": 0.2439303696155548, "learning_rate": 9.669209436859997e-05, "loss": 0.0162, "step": 12850 }, { "grad_norm": 0.23167113959789276, "learning_rate": 9.66846975904614e-05, "loss": 0.0198, "step": 12860 }, { "grad_norm": 0.22150014340877533, "learning_rate": 9.667729283518483e-05, "loss": 0.016, "step": 12870 }, { "grad_norm": 0.18948450684547424, "learning_rate": 9.666988010403557e-05, "loss": 0.0157, "step": 12880 }, { "grad_norm": 0.2069055587053299, "learning_rate": 9.66624593982802e-05, "loss": 0.0165, "step": 12890 }, { "grad_norm": 0.23214726150035858, "learning_rate": 9.665503071918675e-05, "loss": 0.017, "step": 12900 }, { "grad_norm": 0.22907349467277527, "learning_rate": 9.664759406802456e-05, "loss": 0.0153, "step": 12910 }, { "grad_norm": 0.17405131459236145, "learning_rate": 9.664014944606437e-05, "loss": 0.0168, "step": 12920 }, { "grad_norm": 0.19061607122421265, "learning_rate": 9.663269685457822e-05, "loss": 0.0167, "step": 12930 }, { "grad_norm": 0.21375314891338348, "learning_rate": 9.662523629483962e-05, "loss": 0.0181, "step": 12940 }, { "grad_norm": 0.23798643052577972, "learning_rate": 9.661776776812333e-05, "loss": 0.0152, "step": 12950 }, { "grad_norm": 0.22009873390197754, "learning_rate": 9.661029127570553e-05, "loss": 0.0167, "step": 12960 }, { "grad_norm": 0.24503712356090546, "learning_rate": 9.660280681886373e-05, "loss": 0.0162, "step": 12970 }, { "grad_norm": 0.20816203951835632, "learning_rate": 9.659531439887685e-05, "loss": 0.0176, "step": 12980 }, { "grad_norm": 0.2793334722518921, "learning_rate": 9.658781401702511e-05, "loss": 0.0242, "step": 12990 }, { "grad_norm": 0.2006528377532959, "learning_rate": 9.658030567459015e-05, "loss": 0.0151, "step": 13000 }, { "grad_norm": 0.23950548470020294, "learning_rate": 9.65727893728549e-05, "loss": 0.0175, "step": 13010 }, { "grad_norm": 0.28782758116722107, "learning_rate": 9.656526511310375e-05, "loss": 0.0169, "step": 13020 }, { "grad_norm": 0.2134096622467041, "learning_rate": 9.655773289662233e-05, "loss": 0.0166, "step": 13030 }, { "grad_norm": 0.19827748835086823, "learning_rate": 9.655019272469772e-05, "loss": 0.0218, "step": 13040 }, { "grad_norm": 0.2226150929927826, "learning_rate": 9.654264459861832e-05, "loss": 0.0198, "step": 13050 }, { "grad_norm": 0.1740843802690506, "learning_rate": 9.653508851967391e-05, "loss": 0.0161, "step": 13060 }, { "grad_norm": 0.23059391975402832, "learning_rate": 9.65275244891556e-05, "loss": 0.0145, "step": 13070 }, { "grad_norm": 0.22842340171337128, "learning_rate": 9.651995250835591e-05, "loss": 0.0143, "step": 13080 }, { "grad_norm": 0.24035169184207916, "learning_rate": 9.651237257856862e-05, "loss": 0.0156, "step": 13090 }, { "grad_norm": 0.22183674573898315, "learning_rate": 9.6504784701089e-05, "loss": 0.0132, "step": 13100 }, { "grad_norm": 0.27995991706848145, "learning_rate": 9.649718887721357e-05, "loss": 0.0184, "step": 13110 }, { "grad_norm": 0.27742472290992737, "learning_rate": 9.648958510824028e-05, "loss": 0.0159, "step": 13120 }, { "grad_norm": 0.1380978524684906, "learning_rate": 9.648197339546837e-05, "loss": 0.0167, "step": 13130 }, { "grad_norm": 0.17858648300170898, "learning_rate": 9.647435374019851e-05, "loss": 0.0182, "step": 13140 }, { "grad_norm": 0.30735474824905396, "learning_rate": 9.646672614373266e-05, "loss": 0.0185, "step": 13150 }, { "grad_norm": 0.27172085642814636, "learning_rate": 9.645909060737418e-05, "loss": 0.0164, "step": 13160 }, { "grad_norm": 0.20482251048088074, "learning_rate": 9.645144713242778e-05, "loss": 0.0195, "step": 13170 }, { "grad_norm": 0.23945070803165436, "learning_rate": 9.64437957201995e-05, "loss": 0.0157, "step": 13180 }, { "grad_norm": 0.22851334512233734, "learning_rate": 9.643613637199678e-05, "loss": 0.0181, "step": 13190 }, { "grad_norm": 0.16484220325946808, "learning_rate": 9.642846908912839e-05, "loss": 0.0177, "step": 13200 }, { "grad_norm": 0.15400375425815582, "learning_rate": 9.642079387290444e-05, "loss": 0.016, "step": 13210 }, { "grad_norm": 0.1980152130126953, "learning_rate": 9.641311072463644e-05, "loss": 0.0191, "step": 13220 }, { "grad_norm": 0.19746384024620056, "learning_rate": 9.640541964563722e-05, "loss": 0.0188, "step": 13230 }, { "grad_norm": 0.22499318420886993, "learning_rate": 9.639772063722096e-05, "loss": 0.0202, "step": 13240 }, { "grad_norm": 0.2118482142686844, "learning_rate": 9.639001370070324e-05, "loss": 0.0163, "step": 13250 }, { "grad_norm": 0.21345260739326477, "learning_rate": 9.638229883740095e-05, "loss": 0.0168, "step": 13260 }, { "grad_norm": 0.22634746134281158, "learning_rate": 9.637457604863233e-05, "loss": 0.0149, "step": 13270 }, { "grad_norm": 0.2709628641605377, "learning_rate": 9.636684533571703e-05, "loss": 0.0162, "step": 13280 }, { "grad_norm": 0.23433049023151398, "learning_rate": 9.635910669997599e-05, "loss": 0.0133, "step": 13290 }, { "grad_norm": 0.1543884128332138, "learning_rate": 9.635136014273154e-05, "loss": 0.0179, "step": 13300 }, { "grad_norm": 0.2005540132522583, "learning_rate": 9.634360566530735e-05, "loss": 0.0162, "step": 13310 }, { "grad_norm": 0.2269100397825241, "learning_rate": 9.633584326902845e-05, "loss": 0.0162, "step": 13320 }, { "grad_norm": 0.19906538724899292, "learning_rate": 9.632807295522124e-05, "loss": 0.0166, "step": 13330 }, { "grad_norm": 0.28860315680503845, "learning_rate": 9.632029472521342e-05, "loss": 0.0177, "step": 13340 }, { "grad_norm": 0.1586405336856842, "learning_rate": 9.631250858033409e-05, "loss": 0.0164, "step": 13350 }, { "grad_norm": 0.18872053921222687, "learning_rate": 9.630471452191371e-05, "loss": 0.0183, "step": 13360 }, { "grad_norm": 0.2351226806640625, "learning_rate": 9.629691255128405e-05, "loss": 0.0168, "step": 13370 }, { "grad_norm": 0.2073543220758438, "learning_rate": 9.628910266977825e-05, "loss": 0.0148, "step": 13380 }, { "grad_norm": 0.1721937507390976, "learning_rate": 9.628128487873083e-05, "loss": 0.0136, "step": 13390 }, { "grad_norm": 0.21049675345420837, "learning_rate": 9.627345917947761e-05, "loss": 0.0173, "step": 13400 }, { "grad_norm": 0.18327344954013824, "learning_rate": 9.626562557335579e-05, "loss": 0.015, "step": 13410 }, { "grad_norm": 0.19546419382095337, "learning_rate": 9.625778406170393e-05, "loss": 0.0152, "step": 13420 }, { "grad_norm": 0.24785923957824707, "learning_rate": 9.624993464586193e-05, "loss": 0.0182, "step": 13430 }, { "grad_norm": 0.2629808485507965, "learning_rate": 9.624207732717105e-05, "loss": 0.0194, "step": 13440 }, { "grad_norm": 0.20664054155349731, "learning_rate": 9.623421210697386e-05, "loss": 0.0174, "step": 13450 }, { "grad_norm": 0.21396569907665253, "learning_rate": 9.622633898661434e-05, "loss": 0.0184, "step": 13460 }, { "grad_norm": 0.2185993641614914, "learning_rate": 9.621845796743778e-05, "loss": 0.0208, "step": 13470 }, { "grad_norm": 0.2630806267261505, "learning_rate": 9.621056905079082e-05, "loss": 0.0173, "step": 13480 }, { "grad_norm": 0.23326046764850616, "learning_rate": 9.620267223802149e-05, "loss": 0.0158, "step": 13490 }, { "grad_norm": 0.17136557400226593, "learning_rate": 9.619476753047911e-05, "loss": 0.0145, "step": 13500 }, { "grad_norm": 0.17010322213172913, "learning_rate": 9.618685492951438e-05, "loss": 0.0172, "step": 13510 }, { "grad_norm": 0.19387157261371613, "learning_rate": 9.617893443647938e-05, "loss": 0.0135, "step": 13520 }, { "grad_norm": 0.19256310164928436, "learning_rate": 9.617100605272746e-05, "loss": 0.015, "step": 13530 }, { "grad_norm": 0.2246532440185547, "learning_rate": 9.616306977961338e-05, "loss": 0.0139, "step": 13540 }, { "grad_norm": 0.1905011087656021, "learning_rate": 9.615512561849326e-05, "loss": 0.0145, "step": 13550 }, { "grad_norm": 0.2382698953151703, "learning_rate": 9.61471735707245e-05, "loss": 0.0199, "step": 13560 }, { "grad_norm": 0.24660086631774902, "learning_rate": 9.613921363766592e-05, "loss": 0.0156, "step": 13570 }, { "grad_norm": 0.19265654683113098, "learning_rate": 9.613124582067763e-05, "loss": 0.0146, "step": 13580 }, { "grad_norm": 0.1838933229446411, "learning_rate": 9.612327012112112e-05, "loss": 0.0155, "step": 13590 }, { "grad_norm": 0.25861412286758423, "learning_rate": 9.611528654035921e-05, "loss": 0.0138, "step": 13600 }, { "grad_norm": 0.18514253199100494, "learning_rate": 9.610729507975611e-05, "loss": 0.017, "step": 13610 }, { "grad_norm": 0.1880900114774704, "learning_rate": 9.609929574067731e-05, "loss": 0.0159, "step": 13620 }, { "grad_norm": 0.15703365206718445, "learning_rate": 9.609128852448967e-05, "loss": 0.0159, "step": 13630 }, { "grad_norm": 0.16377493739128113, "learning_rate": 9.608327343256143e-05, "loss": 0.0158, "step": 13640 }, { "grad_norm": 0.2304820865392685, "learning_rate": 9.607525046626216e-05, "loss": 0.0158, "step": 13650 }, { "grad_norm": 0.18753643333911896, "learning_rate": 9.606721962696272e-05, "loss": 0.0143, "step": 13660 }, { "grad_norm": 0.23071718215942383, "learning_rate": 9.60591809160354e-05, "loss": 0.0161, "step": 13670 }, { "grad_norm": 0.2683578431606293, "learning_rate": 9.605113433485378e-05, "loss": 0.0157, "step": 13680 }, { "grad_norm": 0.2092842161655426, "learning_rate": 9.604307988479279e-05, "loss": 0.0166, "step": 13690 }, { "grad_norm": 0.1711520403623581, "learning_rate": 9.603501756722876e-05, "loss": 0.0151, "step": 13700 }, { "grad_norm": 0.20006150007247925, "learning_rate": 9.602694738353927e-05, "loss": 0.0169, "step": 13710 }, { "grad_norm": 0.2190553843975067, "learning_rate": 9.601886933510331e-05, "loss": 0.0172, "step": 13720 }, { "grad_norm": 0.22339503467082977, "learning_rate": 9.60107834233012e-05, "loss": 0.0155, "step": 13730 }, { "grad_norm": 0.2253050059080124, "learning_rate": 9.60026896495146e-05, "loss": 0.0187, "step": 13740 }, { "grad_norm": 0.19820579886436462, "learning_rate": 9.599458801512652e-05, "loss": 0.0144, "step": 13750 }, { "grad_norm": 0.24450229108333588, "learning_rate": 9.598647852152129e-05, "loss": 0.0143, "step": 13760 }, { "grad_norm": 0.19496899843215942, "learning_rate": 9.597836117008462e-05, "loss": 0.0171, "step": 13770 }, { "grad_norm": 0.19759394228458405, "learning_rate": 9.597023596220356e-05, "loss": 0.0186, "step": 13780 }, { "grad_norm": 0.2045399397611618, "learning_rate": 9.596210289926643e-05, "loss": 0.0153, "step": 13790 }, { "grad_norm": 0.2786710560321808, "learning_rate": 9.5953961982663e-05, "loss": 0.0154, "step": 13800 }, { "grad_norm": 0.3687841594219208, "learning_rate": 9.594581321378431e-05, "loss": 0.0165, "step": 13810 }, { "grad_norm": 0.20075836777687073, "learning_rate": 9.593765659402276e-05, "loss": 0.0177, "step": 13820 }, { "grad_norm": 0.23547106981277466, "learning_rate": 9.59294921247721e-05, "loss": 0.015, "step": 13830 }, { "grad_norm": 0.19620366394519806, "learning_rate": 9.59213198074274e-05, "loss": 0.0141, "step": 13840 }, { "grad_norm": 0.21660323441028595, "learning_rate": 9.59131396433851e-05, "loss": 0.0163, "step": 13850 }, { "grad_norm": 0.26219087839126587, "learning_rate": 9.590495163404297e-05, "loss": 0.0178, "step": 13860 }, { "grad_norm": 0.17897824943065643, "learning_rate": 9.589675578080009e-05, "loss": 0.016, "step": 13870 }, { "grad_norm": 0.23164260387420654, "learning_rate": 9.588855208505694e-05, "loss": 0.0163, "step": 13880 }, { "grad_norm": 0.2048715353012085, "learning_rate": 9.588034054821529e-05, "loss": 0.0145, "step": 13890 }, { "grad_norm": 0.20711378753185272, "learning_rate": 9.587212117167826e-05, "loss": 0.0162, "step": 13900 }, { "grad_norm": 0.22472336888313293, "learning_rate": 9.586389395685033e-05, "loss": 0.0138, "step": 13910 }, { "grad_norm": 0.22538676857948303, "learning_rate": 9.585565890513733e-05, "loss": 0.0146, "step": 13920 }, { "grad_norm": 0.20942215621471405, "learning_rate": 9.584741601794636e-05, "loss": 0.0147, "step": 13930 }, { "grad_norm": 0.21139447391033173, "learning_rate": 9.58391652966859e-05, "loss": 0.0152, "step": 13940 }, { "grad_norm": 0.21437665820121765, "learning_rate": 9.583090674276583e-05, "loss": 0.0171, "step": 13950 }, { "grad_norm": 0.2563517689704895, "learning_rate": 9.582264035759726e-05, "loss": 0.0156, "step": 13960 }, { "grad_norm": 0.2916884124279022, "learning_rate": 9.58143661425927e-05, "loss": 0.0167, "step": 13970 }, { "grad_norm": 0.17540308833122253, "learning_rate": 9.580608409916601e-05, "loss": 0.0179, "step": 13980 }, { "grad_norm": 0.26157665252685547, "learning_rate": 9.579779422873233e-05, "loss": 0.0193, "step": 13990 }, { "grad_norm": 0.2449590265750885, "learning_rate": 9.578949653270819e-05, "loss": 0.021, "step": 14000 }, { "grad_norm": 0.20943903923034668, "learning_rate": 9.578119101251144e-05, "loss": 0.0137, "step": 14010 }, { "grad_norm": 0.18112343549728394, "learning_rate": 9.577287766956127e-05, "loss": 0.0163, "step": 14020 }, { "grad_norm": 0.24193121492862701, "learning_rate": 9.57645565052782e-05, "loss": 0.0152, "step": 14030 }, { "grad_norm": 0.26600030064582825, "learning_rate": 9.575622752108407e-05, "loss": 0.018, "step": 14040 }, { "grad_norm": 0.22042126953601837, "learning_rate": 9.57478907184021e-05, "loss": 0.0195, "step": 14050 }, { "grad_norm": 0.25403931736946106, "learning_rate": 9.573954609865681e-05, "loss": 0.0176, "step": 14060 }, { "grad_norm": 0.24770650267601013, "learning_rate": 9.573119366327408e-05, "loss": 0.0194, "step": 14070 }, { "grad_norm": 0.27592337131500244, "learning_rate": 9.57228334136811e-05, "loss": 0.0173, "step": 14080 }, { "grad_norm": 0.2219957858324051, "learning_rate": 9.571446535130641e-05, "loss": 0.0186, "step": 14090 }, { "grad_norm": 0.25445759296417236, "learning_rate": 9.570608947757988e-05, "loss": 0.0173, "step": 14100 }, { "grad_norm": 0.2438957393169403, "learning_rate": 9.569770579393274e-05, "loss": 0.0166, "step": 14110 }, { "grad_norm": 0.17989687621593475, "learning_rate": 9.56893143017975e-05, "loss": 0.0168, "step": 14120 }, { "grad_norm": 0.24896939098834991, "learning_rate": 9.568091500260806e-05, "loss": 0.0151, "step": 14130 }, { "grad_norm": 0.17997905611991882, "learning_rate": 9.567250789779961e-05, "loss": 0.0143, "step": 14140 }, { "grad_norm": 0.2202969789505005, "learning_rate": 9.566409298880872e-05, "loss": 0.0157, "step": 14150 }, { "grad_norm": 0.23241761326789856, "learning_rate": 9.565567027707326e-05, "loss": 0.0179, "step": 14160 }, { "grad_norm": 0.26811039447784424, "learning_rate": 9.56472397640324e-05, "loss": 0.0169, "step": 14170 }, { "grad_norm": 0.2038077414035797, "learning_rate": 9.563880145112675e-05, "loss": 0.0202, "step": 14180 }, { "grad_norm": 0.16342124342918396, "learning_rate": 9.563035533979814e-05, "loss": 0.0154, "step": 14190 }, { "grad_norm": 0.1986149400472641, "learning_rate": 9.562190143148981e-05, "loss": 0.0193, "step": 14200 }, { "grad_norm": 0.15543045103549957, "learning_rate": 9.561343972764627e-05, "loss": 0.0133, "step": 14210 }, { "grad_norm": 0.16018593311309814, "learning_rate": 9.560497022971343e-05, "loss": 0.0155, "step": 14220 }, { "grad_norm": 0.2365039736032486, "learning_rate": 9.559649293913847e-05, "loss": 0.0149, "step": 14230 }, { "grad_norm": 0.21893291175365448, "learning_rate": 9.558800785736993e-05, "loss": 0.0178, "step": 14240 }, { "grad_norm": 0.24009978771209717, "learning_rate": 9.557951498585767e-05, "loss": 0.0177, "step": 14250 }, { "grad_norm": 0.22301085293293, "learning_rate": 9.557101432605293e-05, "loss": 0.0158, "step": 14260 }, { "grad_norm": 0.2044178694486618, "learning_rate": 9.556250587940818e-05, "loss": 0.0163, "step": 14270 }, { "grad_norm": 0.23526006937026978, "learning_rate": 9.555398964737734e-05, "loss": 0.0162, "step": 14280 }, { "grad_norm": 0.25124457478523254, "learning_rate": 9.554546563141555e-05, "loss": 0.0155, "step": 14290 }, { "grad_norm": 0.2508791387081146, "learning_rate": 9.553693383297937e-05, "loss": 0.0192, "step": 14300 }, { "grad_norm": 0.19767458736896515, "learning_rate": 9.552839425352663e-05, "loss": 0.0148, "step": 14310 }, { "grad_norm": 0.20059888064861298, "learning_rate": 9.551984689451652e-05, "loss": 0.0162, "step": 14320 }, { "grad_norm": 0.2128722220659256, "learning_rate": 9.551129175740953e-05, "loss": 0.0166, "step": 14330 }, { "grad_norm": 0.14695122838020325, "learning_rate": 9.550272884366754e-05, "loss": 0.0142, "step": 14340 }, { "grad_norm": 0.1504315286874771, "learning_rate": 9.549415815475369e-05, "loss": 0.0155, "step": 14350 }, { "grad_norm": 0.15290725231170654, "learning_rate": 9.548557969213247e-05, "loss": 0.0143, "step": 14360 }, { "grad_norm": 0.1714729517698288, "learning_rate": 9.547699345726972e-05, "loss": 0.0123, "step": 14370 }, { "grad_norm": 0.19310979545116425, "learning_rate": 9.546839945163257e-05, "loss": 0.0133, "step": 14380 }, { "grad_norm": 0.2227582186460495, "learning_rate": 9.545979767668953e-05, "loss": 0.0165, "step": 14390 }, { "grad_norm": 0.19209016859531403, "learning_rate": 9.54511881339104e-05, "loss": 0.0157, "step": 14400 }, { "grad_norm": 0.1955484300851822, "learning_rate": 9.54425708247663e-05, "loss": 0.017, "step": 14410 }, { "grad_norm": 0.1926402449607849, "learning_rate": 9.543394575072972e-05, "loss": 0.0157, "step": 14420 }, { "grad_norm": 0.21464917063713074, "learning_rate": 9.542531291327441e-05, "loss": 0.0143, "step": 14430 }, { "grad_norm": 0.16287541389465332, "learning_rate": 9.541667231387552e-05, "loss": 0.0132, "step": 14440 }, { "grad_norm": 0.20301847159862518, "learning_rate": 9.540802395400949e-05, "loss": 0.0143, "step": 14450 }, { "grad_norm": 0.1699981838464737, "learning_rate": 9.539936783515406e-05, "loss": 0.0146, "step": 14460 }, { "grad_norm": 0.19603444635868073, "learning_rate": 9.539070395878835e-05, "loss": 0.0171, "step": 14470 }, { "grad_norm": 0.21420486271381378, "learning_rate": 9.538203232639277e-05, "loss": 0.0137, "step": 14480 }, { "grad_norm": 0.1256834715604782, "learning_rate": 9.537335293944907e-05, "loss": 0.0149, "step": 14490 }, { "grad_norm": 0.1961212009191513, "learning_rate": 9.536466579944032e-05, "loss": 0.0141, "step": 14500 }, { "grad_norm": 0.1681392639875412, "learning_rate": 9.535597090785091e-05, "loss": 0.0158, "step": 14510 }, { "grad_norm": 0.17317494750022888, "learning_rate": 9.534726826616656e-05, "loss": 0.0141, "step": 14520 }, { "grad_norm": 0.17453011870384216, "learning_rate": 9.53385578758743e-05, "loss": 0.0143, "step": 14530 }, { "grad_norm": 0.15404562652111053, "learning_rate": 9.532983973846252e-05, "loss": 0.016, "step": 14540 }, { "grad_norm": 0.20784296095371246, "learning_rate": 9.53211138554209e-05, "loss": 0.0137, "step": 14550 }, { "grad_norm": 0.18411390483379364, "learning_rate": 9.531238022824047e-05, "loss": 0.014, "step": 14560 }, { "grad_norm": 0.19282034039497375, "learning_rate": 9.530363885841355e-05, "loss": 0.0162, "step": 14570 }, { "grad_norm": 0.17514410614967346, "learning_rate": 9.52948897474338e-05, "loss": 0.0168, "step": 14580 }, { "grad_norm": 0.19982695579528809, "learning_rate": 9.528613289679622e-05, "loss": 0.0145, "step": 14590 }, { "grad_norm": 0.17274288833141327, "learning_rate": 9.52773683079971e-05, "loss": 0.0143, "step": 14600 }, { "grad_norm": 0.1900903433561325, "learning_rate": 9.526859598253407e-05, "loss": 0.0153, "step": 14610 }, { "grad_norm": 0.2061288058757782, "learning_rate": 9.525981592190609e-05, "loss": 0.0164, "step": 14620 }, { "grad_norm": 0.18867795169353485, "learning_rate": 9.525102812761342e-05, "loss": 0.0158, "step": 14630 }, { "grad_norm": 0.17402228713035583, "learning_rate": 9.524223260115768e-05, "loss": 0.0148, "step": 14640 }, { "grad_norm": 0.277385413646698, "learning_rate": 9.523342934404175e-05, "loss": 0.014, "step": 14650 }, { "grad_norm": 0.1722254753112793, "learning_rate": 9.522461835776989e-05, "loss": 0.0185, "step": 14660 }, { "grad_norm": 0.20834419131278992, "learning_rate": 9.521579964384764e-05, "loss": 0.0152, "step": 14670 }, { "grad_norm": 0.17455589771270752, "learning_rate": 9.52069732037819e-05, "loss": 0.016, "step": 14680 }, { "grad_norm": 0.2282034307718277, "learning_rate": 9.519813903908083e-05, "loss": 0.0145, "step": 14690 }, { "grad_norm": 0.2551082968711853, "learning_rate": 9.5189297151254e-05, "loss": 0.0161, "step": 14700 }, { "grad_norm": 0.21686170995235443, "learning_rate": 9.518044754181218e-05, "loss": 0.0139, "step": 14710 }, { "grad_norm": 0.17962917685508728, "learning_rate": 9.51715902122676e-05, "loss": 0.0161, "step": 14720 }, { "grad_norm": 0.1655292510986328, "learning_rate": 9.516272516413368e-05, "loss": 0.0159, "step": 14730 }, { "grad_norm": 0.18939705193042755, "learning_rate": 9.515385239892525e-05, "loss": 0.0188, "step": 14740 }, { "grad_norm": 0.19590482115745544, "learning_rate": 9.514497191815839e-05, "loss": 0.0159, "step": 14750 }, { "grad_norm": 0.25577613711357117, "learning_rate": 9.513608372335055e-05, "loss": 0.0135, "step": 14760 }, { "grad_norm": 0.18215744197368622, "learning_rate": 9.512718781602045e-05, "loss": 0.0158, "step": 14770 }, { "grad_norm": 0.19325025379657745, "learning_rate": 9.511828419768823e-05, "loss": 0.0155, "step": 14780 }, { "grad_norm": 0.19891586899757385, "learning_rate": 9.510937286987521e-05, "loss": 0.0154, "step": 14790 }, { "grad_norm": 0.1641089767217636, "learning_rate": 9.510045383410408e-05, "loss": 0.0135, "step": 14800 }, { "grad_norm": 0.21450623869895935, "learning_rate": 9.509152709189892e-05, "loss": 0.0139, "step": 14810 }, { "grad_norm": 0.223328098654747, "learning_rate": 9.508259264478504e-05, "loss": 0.0155, "step": 14820 }, { "grad_norm": 0.22708944976329803, "learning_rate": 9.507365049428909e-05, "loss": 0.0154, "step": 14830 }, { "grad_norm": 0.19422335922718048, "learning_rate": 9.506470064193902e-05, "loss": 0.0163, "step": 14840 }, { "grad_norm": 0.21020230650901794, "learning_rate": 9.505574308926414e-05, "loss": 0.0156, "step": 14850 }, { "grad_norm": 0.16628028452396393, "learning_rate": 9.504677783779505e-05, "loss": 0.0125, "step": 14860 }, { "grad_norm": 0.17815561592578888, "learning_rate": 9.503780488906365e-05, "loss": 0.0133, "step": 14870 }, { "grad_norm": 0.22450663149356842, "learning_rate": 9.502882424460319e-05, "loss": 0.0162, "step": 14880 }, { "grad_norm": 0.1820373386144638, "learning_rate": 9.501983590594821e-05, "loss": 0.0169, "step": 14890 }, { "grad_norm": 0.1957058310508728, "learning_rate": 9.501083987463455e-05, "loss": 0.015, "step": 14900 }, { "grad_norm": 0.22598977386951447, "learning_rate": 9.500183615219942e-05, "loss": 0.0136, "step": 14910 }, { "grad_norm": 0.2029353678226471, "learning_rate": 9.49928247401813e-05, "loss": 0.019, "step": 14920 }, { "grad_norm": 0.2294318825006485, "learning_rate": 9.498380564011997e-05, "loss": 0.0168, "step": 14930 }, { "grad_norm": 0.23529984056949615, "learning_rate": 9.497477885355656e-05, "loss": 0.0185, "step": 14940 }, { "grad_norm": 0.2705339789390564, "learning_rate": 9.496574438203353e-05, "loss": 0.0151, "step": 14950 }, { "grad_norm": 0.25373366475105286, "learning_rate": 9.495670222709459e-05, "loss": 0.0133, "step": 14960 }, { "grad_norm": 0.2222069352865219, "learning_rate": 9.494765239028483e-05, "loss": 0.02, "step": 14970 }, { "grad_norm": 0.17430450022220612, "learning_rate": 9.493859487315057e-05, "loss": 0.0146, "step": 14980 }, { "grad_norm": 0.16832736134529114, "learning_rate": 9.492952967723953e-05, "loss": 0.0132, "step": 14990 }, { "grad_norm": 0.18448369204998016, "learning_rate": 9.492045680410068e-05, "loss": 0.0161, "step": 15000 }, { "grad_norm": 0.18256011605262756, "learning_rate": 9.491137625528436e-05, "loss": 0.0143, "step": 15010 }, { "grad_norm": 0.19558802247047424, "learning_rate": 9.490228803234215e-05, "loss": 0.0149, "step": 15020 }, { "grad_norm": 0.18421515822410583, "learning_rate": 9.489319213682701e-05, "loss": 0.0132, "step": 15030 }, { "grad_norm": 0.21926437318325043, "learning_rate": 9.488408857029316e-05, "loss": 0.0119, "step": 15040 }, { "grad_norm": 0.2009773701429367, "learning_rate": 9.487497733429616e-05, "loss": 0.0145, "step": 15050 }, { "grad_norm": 0.2238122522830963, "learning_rate": 9.486585843039286e-05, "loss": 0.013, "step": 15060 }, { "grad_norm": 0.17851389944553375, "learning_rate": 9.485673186014143e-05, "loss": 0.0132, "step": 15070 }, { "grad_norm": 0.14462709426879883, "learning_rate": 9.484759762510137e-05, "loss": 0.0146, "step": 15080 }, { "grad_norm": 0.19499407708644867, "learning_rate": 9.483845572683346e-05, "loss": 0.0141, "step": 15090 }, { "grad_norm": 0.21804137527942657, "learning_rate": 9.48293061668998e-05, "loss": 0.0143, "step": 15100 }, { "grad_norm": 0.22167813777923584, "learning_rate": 9.48201489468638e-05, "loss": 0.0157, "step": 15110 }, { "grad_norm": 0.18958468735218048, "learning_rate": 9.481098406829016e-05, "loss": 0.0142, "step": 15120 }, { "grad_norm": 0.2194022834300995, "learning_rate": 9.480181153274495e-05, "loss": 0.0156, "step": 15130 }, { "grad_norm": 0.18684031069278717, "learning_rate": 9.479263134179548e-05, "loss": 0.0177, "step": 15140 }, { "grad_norm": 0.20384439826011658, "learning_rate": 9.478344349701039e-05, "loss": 0.0177, "step": 15150 }, { "grad_norm": 0.13189858198165894, "learning_rate": 9.477424799995964e-05, "loss": 0.0158, "step": 15160 }, { "grad_norm": 0.16131140291690826, "learning_rate": 9.476504485221448e-05, "loss": 0.0153, "step": 15170 }, { "grad_norm": 0.2511102855205536, "learning_rate": 9.475583405534748e-05, "loss": 0.015, "step": 15180 }, { "grad_norm": 0.2235811948776245, "learning_rate": 9.474661561093251e-05, "loss": 0.0175, "step": 15190 }, { "grad_norm": 0.1730978786945343, "learning_rate": 9.473738952054478e-05, "loss": 0.014, "step": 15200 }, { "grad_norm": 0.2336443066596985, "learning_rate": 9.472815578576073e-05, "loss": 0.0161, "step": 15210 }, { "grad_norm": 0.15602189302444458, "learning_rate": 9.471891440815817e-05, "loss": 0.0137, "step": 15220 }, { "grad_norm": 0.15883241593837738, "learning_rate": 9.470966538931621e-05, "loss": 0.0133, "step": 15230 }, { "grad_norm": 0.18022103607654572, "learning_rate": 9.470040873081525e-05, "loss": 0.0154, "step": 15240 }, { "grad_norm": 0.17435993254184723, "learning_rate": 9.469114443423698e-05, "loss": 0.014, "step": 15250 }, { "grad_norm": 0.24051645398139954, "learning_rate": 9.468187250116445e-05, "loss": 0.0161, "step": 15260 }, { "grad_norm": 0.21319904923439026, "learning_rate": 9.467259293318197e-05, "loss": 0.0145, "step": 15270 }, { "grad_norm": 0.23522554337978363, "learning_rate": 9.466330573187514e-05, "loss": 0.0167, "step": 15280 }, { "grad_norm": 0.17915168404579163, "learning_rate": 9.46540108988309e-05, "loss": 0.0143, "step": 15290 }, { "grad_norm": 0.2052653729915619, "learning_rate": 9.46447084356375e-05, "loss": 0.0132, "step": 15300 }, { "grad_norm": 0.122018001973629, "learning_rate": 9.463539834388447e-05, "loss": 0.0127, "step": 15310 }, { "grad_norm": 0.20774081349372864, "learning_rate": 9.462608062516263e-05, "loss": 0.0152, "step": 15320 }, { "grad_norm": 0.15646332502365112, "learning_rate": 9.461675528106413e-05, "loss": 0.0132, "step": 15330 }, { "grad_norm": 0.22152385115623474, "learning_rate": 9.460742231318244e-05, "loss": 0.0155, "step": 15340 }, { "grad_norm": 0.23049704730510712, "learning_rate": 9.459808172311229e-05, "loss": 0.0151, "step": 15350 }, { "grad_norm": 0.21023215353488922, "learning_rate": 9.458873351244972e-05, "loss": 0.0165, "step": 15360 }, { "grad_norm": 0.20612594485282898, "learning_rate": 9.457937768279211e-05, "loss": 0.0193, "step": 15370 }, { "grad_norm": 0.22063352167606354, "learning_rate": 9.45700142357381e-05, "loss": 0.0181, "step": 15380 }, { "grad_norm": 0.2048928141593933, "learning_rate": 9.456064317288765e-05, "loss": 0.0157, "step": 15390 }, { "grad_norm": 0.2310309112071991, "learning_rate": 9.455126449584201e-05, "loss": 0.0173, "step": 15400 }, { "grad_norm": 0.2834858298301697, "learning_rate": 9.454187820620375e-05, "loss": 0.0158, "step": 15410 }, { "grad_norm": 0.23790660500526428, "learning_rate": 9.453248430557673e-05, "loss": 0.015, "step": 15420 }, { "grad_norm": 0.13044805824756622, "learning_rate": 9.452308279556611e-05, "loss": 0.0156, "step": 15430 }, { "grad_norm": 0.18065245449543, "learning_rate": 9.451367367777835e-05, "loss": 0.016, "step": 15440 }, { "grad_norm": 0.17809069156646729, "learning_rate": 9.450425695382122e-05, "loss": 0.0136, "step": 15450 }, { "grad_norm": 0.1890561729669571, "learning_rate": 9.449483262530375e-05, "loss": 0.0157, "step": 15460 }, { "grad_norm": 0.15700289607048035, "learning_rate": 9.448540069383633e-05, "loss": 0.0169, "step": 15470 }, { "grad_norm": 0.26946744322776794, "learning_rate": 9.447596116103061e-05, "loss": 0.0172, "step": 15480 }, { "grad_norm": 0.23468106985092163, "learning_rate": 9.446651402849955e-05, "loss": 0.0211, "step": 15490 }, { "grad_norm": 0.27216005325317383, "learning_rate": 9.44570592978574e-05, "loss": 0.0193, "step": 15500 }, { "grad_norm": 0.23088940978050232, "learning_rate": 9.444759697071972e-05, "loss": 0.0167, "step": 15510 }, { "grad_norm": 0.1846339851617813, "learning_rate": 9.443812704870336e-05, "loss": 0.0157, "step": 15520 }, { "grad_norm": 0.2194863110780716, "learning_rate": 9.442864953342649e-05, "loss": 0.0164, "step": 15530 }, { "grad_norm": 0.17453725636005402, "learning_rate": 9.441916442650852e-05, "loss": 0.0139, "step": 15540 }, { "grad_norm": 0.18926174938678741, "learning_rate": 9.440967172957023e-05, "loss": 0.0156, "step": 15550 }, { "grad_norm": 0.22595082223415375, "learning_rate": 9.440017144423364e-05, "loss": 0.0142, "step": 15560 }, { "grad_norm": 0.191427543759346, "learning_rate": 9.439066357212209e-05, "loss": 0.014, "step": 15570 }, { "grad_norm": 0.20061276853084564, "learning_rate": 9.438114811486022e-05, "loss": 0.0128, "step": 15580 }, { "grad_norm": 0.20556987822055817, "learning_rate": 9.4371625074074e-05, "loss": 0.0148, "step": 15590 }, { "grad_norm": 0.13929864764213562, "learning_rate": 9.436209445139059e-05, "loss": 0.0155, "step": 15600 }, { "grad_norm": 0.18576376140117645, "learning_rate": 9.435255624843855e-05, "loss": 0.0149, "step": 15610 }, { "grad_norm": 0.16214491426944733, "learning_rate": 9.43430104668477e-05, "loss": 0.013, "step": 15620 }, { "grad_norm": 0.19655998051166534, "learning_rate": 9.433345710824914e-05, "loss": 0.0148, "step": 15630 }, { "grad_norm": 0.17354176938533783, "learning_rate": 9.432389617427529e-05, "loss": 0.0158, "step": 15640 }, { "grad_norm": 0.1801779568195343, "learning_rate": 9.431432766655984e-05, "loss": 0.0155, "step": 15650 }, { "grad_norm": 0.21585814654827118, "learning_rate": 9.430475158673778e-05, "loss": 0.016, "step": 15660 }, { "grad_norm": 0.2738094925880432, "learning_rate": 9.429516793644542e-05, "loss": 0.0159, "step": 15670 }, { "grad_norm": 0.13534900546073914, "learning_rate": 9.428557671732034e-05, "loss": 0.0159, "step": 15680 }, { "grad_norm": 0.16867917776107788, "learning_rate": 9.42759779310014e-05, "loss": 0.0129, "step": 15690 }, { "grad_norm": 0.17517364025115967, "learning_rate": 9.426637157912879e-05, "loss": 0.0121, "step": 15700 }, { "grad_norm": 0.23626117408275604, "learning_rate": 9.425675766334397e-05, "loss": 0.0148, "step": 15710 }, { "grad_norm": 0.2137906700372696, "learning_rate": 9.424713618528968e-05, "loss": 0.0135, "step": 15720 }, { "grad_norm": 0.23525677621364594, "learning_rate": 9.423750714661e-05, "loss": 0.0162, "step": 15730 }, { "grad_norm": 0.20610226690769196, "learning_rate": 9.422787054895022e-05, "loss": 0.015, "step": 15740 }, { "grad_norm": 0.21256107091903687, "learning_rate": 9.4218226393957e-05, "loss": 0.0153, "step": 15750 }, { "grad_norm": 0.2055271565914154, "learning_rate": 9.420857468327828e-05, "loss": 0.0179, "step": 15760 }, { "grad_norm": 0.20985589921474457, "learning_rate": 9.419891541856323e-05, "loss": 0.0165, "step": 15770 }, { "grad_norm": 0.240744948387146, "learning_rate": 9.41892486014624e-05, "loss": 0.0147, "step": 15780 }, { "grad_norm": 0.24006356298923492, "learning_rate": 9.417957423362756e-05, "loss": 0.0153, "step": 15790 }, { "grad_norm": 0.19599591195583344, "learning_rate": 9.416989231671178e-05, "loss": 0.015, "step": 15800 }, { "grad_norm": 0.20046544075012207, "learning_rate": 9.416020285236946e-05, "loss": 0.0132, "step": 15810 }, { "grad_norm": 0.17914868891239166, "learning_rate": 9.415050584225626e-05, "loss": 0.0183, "step": 15820 }, { "grad_norm": 0.18282847106456757, "learning_rate": 9.414080128802914e-05, "loss": 0.0119, "step": 15830 }, { "grad_norm": 0.24273057281970978, "learning_rate": 9.413108919134632e-05, "loss": 0.0174, "step": 15840 }, { "grad_norm": 0.13225342333316803, "learning_rate": 9.412136955386734e-05, "loss": 0.0147, "step": 15850 }, { "grad_norm": 0.21628102660179138, "learning_rate": 9.411164237725303e-05, "loss": 0.0157, "step": 15860 }, { "grad_norm": 0.17797453701496124, "learning_rate": 9.41019076631655e-05, "loss": 0.0142, "step": 15870 }, { "grad_norm": 0.2675248980522156, "learning_rate": 9.409216541326815e-05, "loss": 0.0165, "step": 15880 }, { "grad_norm": 0.1430657058954239, "learning_rate": 9.408241562922564e-05, "loss": 0.0164, "step": 15890 }, { "grad_norm": 0.16653312742710114, "learning_rate": 9.407265831270395e-05, "loss": 0.018, "step": 15900 }, { "grad_norm": 0.2622269093990326, "learning_rate": 9.406289346537035e-05, "loss": 0.0164, "step": 15910 }, { "grad_norm": 0.23229163885116577, "learning_rate": 9.405312108889339e-05, "loss": 0.0209, "step": 15920 }, { "grad_norm": 0.2314019799232483, "learning_rate": 9.404334118494288e-05, "loss": 0.0162, "step": 15930 }, { "grad_norm": 0.21511061489582062, "learning_rate": 9.403355375518995e-05, "loss": 0.0155, "step": 15940 }, { "grad_norm": 0.15595567226409912, "learning_rate": 9.4023758801307e-05, "loss": 0.0178, "step": 15950 }, { "grad_norm": 0.17034125328063965, "learning_rate": 9.401395632496774e-05, "loss": 0.0156, "step": 15960 }, { "grad_norm": 0.17916478216648102, "learning_rate": 9.400414632784711e-05, "loss": 0.0154, "step": 15970 }, { "grad_norm": 0.1689840406179428, "learning_rate": 9.39943288116214e-05, "loss": 0.0137, "step": 15980 }, { "grad_norm": 0.1237712875008583, "learning_rate": 9.398450377796815e-05, "loss": 0.0148, "step": 15990 }, { "grad_norm": 0.18790259957313538, "learning_rate": 9.397467122856616e-05, "loss": 0.0182, "step": 16000 }, { "grad_norm": 0.16724517941474915, "learning_rate": 9.396483116509558e-05, "loss": 0.0158, "step": 16010 }, { "grad_norm": 0.1509292721748352, "learning_rate": 9.39549835892378e-05, "loss": 0.016, "step": 16020 }, { "grad_norm": 0.1854977011680603, "learning_rate": 9.39451285026755e-05, "loss": 0.0137, "step": 16030 }, { "grad_norm": 0.20644846558570862, "learning_rate": 9.393526590709262e-05, "loss": 0.0134, "step": 16040 }, { "grad_norm": 0.1723942905664444, "learning_rate": 9.392539580417444e-05, "loss": 0.0138, "step": 16050 }, { "grad_norm": 0.19489523768424988, "learning_rate": 9.391551819560747e-05, "loss": 0.0148, "step": 16060 }, { "grad_norm": 0.19925473630428314, "learning_rate": 9.390563308307955e-05, "loss": 0.0129, "step": 16070 }, { "grad_norm": 0.20470963418483734, "learning_rate": 9.389574046827974e-05, "loss": 0.0156, "step": 16080 }, { "grad_norm": 0.23102861642837524, "learning_rate": 9.388584035289845e-05, "loss": 0.0137, "step": 16090 }, { "grad_norm": 0.20399601757526398, "learning_rate": 9.387593273862732e-05, "loss": 0.014, "step": 16100 }, { "grad_norm": 0.19173862040042877, "learning_rate": 9.386601762715929e-05, "loss": 0.0141, "step": 16110 }, { "grad_norm": 0.19492694735527039, "learning_rate": 9.38560950201886e-05, "loss": 0.013, "step": 16120 }, { "grad_norm": 0.34695786237716675, "learning_rate": 9.384616491941071e-05, "loss": 0.0156, "step": 16130 }, { "grad_norm": 0.16394686698913574, "learning_rate": 9.383622732652245e-05, "loss": 0.0143, "step": 16140 }, { "grad_norm": 0.1930679827928543, "learning_rate": 9.382628224322187e-05, "loss": 0.0146, "step": 16150 }, { "grad_norm": 0.22488871216773987, "learning_rate": 9.381632967120829e-05, "loss": 0.0149, "step": 16160 }, { "grad_norm": 0.20802849531173706, "learning_rate": 9.380636961218235e-05, "loss": 0.0142, "step": 16170 }, { "grad_norm": 0.2235526442527771, "learning_rate": 9.379640206784597e-05, "loss": 0.0152, "step": 16180 }, { "grad_norm": 0.150466188788414, "learning_rate": 9.378642703990229e-05, "loss": 0.0144, "step": 16190 }, { "grad_norm": 0.19228458404541016, "learning_rate": 9.37764445300558e-05, "loss": 0.0138, "step": 16200 }, { "grad_norm": 0.18313848972320557, "learning_rate": 9.376645454001222e-05, "loss": 0.0141, "step": 16210 }, { "grad_norm": 0.13457289338111877, "learning_rate": 9.375645707147858e-05, "loss": 0.0148, "step": 16220 }, { "grad_norm": 0.18288718163967133, "learning_rate": 9.374645212616316e-05, "loss": 0.0136, "step": 16230 }, { "grad_norm": 0.22269238531589508, "learning_rate": 9.373643970577555e-05, "loss": 0.0155, "step": 16240 }, { "grad_norm": 0.1905863881111145, "learning_rate": 9.372641981202659e-05, "loss": 0.0118, "step": 16250 }, { "grad_norm": 0.21878916025161743, "learning_rate": 9.37163924466284e-05, "loss": 0.0174, "step": 16260 }, { "grad_norm": 0.17946727573871613, "learning_rate": 9.370635761129438e-05, "loss": 0.0132, "step": 16270 }, { "grad_norm": 0.21154817938804626, "learning_rate": 9.36963153077392e-05, "loss": 0.0174, "step": 16280 }, { "grad_norm": 0.15327388048171997, "learning_rate": 9.368626553767888e-05, "loss": 0.0126, "step": 16290 }, { "grad_norm": 0.2562628388404846, "learning_rate": 9.367620830283057e-05, "loss": 0.0153, "step": 16300 }, { "grad_norm": 0.18431724607944489, "learning_rate": 9.366614360491281e-05, "loss": 0.016, "step": 16310 }, { "grad_norm": 0.21832621097564697, "learning_rate": 9.365607144564539e-05, "loss": 0.0153, "step": 16320 }, { "grad_norm": 0.19833441078662872, "learning_rate": 9.364599182674934e-05, "loss": 0.0136, "step": 16330 }, { "grad_norm": 0.17081749439239502, "learning_rate": 9.3635904749947e-05, "loss": 0.0127, "step": 16340 }, { "grad_norm": 0.18572725355625153, "learning_rate": 9.362581021696202e-05, "loss": 0.0147, "step": 16350 }, { "grad_norm": 0.17651677131652832, "learning_rate": 9.361570822951921e-05, "loss": 0.0163, "step": 16360 }, { "grad_norm": 0.2028864473104477, "learning_rate": 9.360559878934476e-05, "loss": 0.0129, "step": 16370 }, { "grad_norm": 0.14670483767986298, "learning_rate": 9.359548189816611e-05, "loss": 0.0121, "step": 16380 }, { "grad_norm": 0.20284074544906616, "learning_rate": 9.358535755771193e-05, "loss": 0.0125, "step": 16390 }, { "grad_norm": 0.1511525958776474, "learning_rate": 9.357522576971221e-05, "loss": 0.0174, "step": 16400 }, { "grad_norm": 0.23163187503814697, "learning_rate": 9.356508653589819e-05, "loss": 0.0152, "step": 16410 }, { "grad_norm": 0.20781131088733673, "learning_rate": 9.355493985800237e-05, "loss": 0.0157, "step": 16420 }, { "grad_norm": 0.212296262383461, "learning_rate": 9.354478573775857e-05, "loss": 0.0128, "step": 16430 }, { "grad_norm": 0.21016065776348114, "learning_rate": 9.353462417690186e-05, "loss": 0.013, "step": 16440 }, { "grad_norm": 0.18680572509765625, "learning_rate": 9.352445517716853e-05, "loss": 0.0163, "step": 16450 }, { "grad_norm": 0.21422000229358673, "learning_rate": 9.351427874029621e-05, "loss": 0.0175, "step": 16460 }, { "grad_norm": 0.16948416829109192, "learning_rate": 9.350409486802379e-05, "loss": 0.0133, "step": 16470 }, { "grad_norm": 0.2630597651004791, "learning_rate": 9.349390356209138e-05, "loss": 0.0139, "step": 16480 }, { "grad_norm": 0.19730514287948608, "learning_rate": 9.348370482424042e-05, "loss": 0.0132, "step": 16490 }, { "grad_norm": 0.22803492844104767, "learning_rate": 9.347349865621357e-05, "loss": 0.0138, "step": 16500 }, { "grad_norm": 0.17060816287994385, "learning_rate": 9.346328505975481e-05, "loss": 0.0112, "step": 16510 }, { "grad_norm": 0.20220130681991577, "learning_rate": 9.345306403660936e-05, "loss": 0.0145, "step": 16520 }, { "grad_norm": 0.22310589253902435, "learning_rate": 9.344283558852371e-05, "loss": 0.0148, "step": 16530 }, { "grad_norm": 0.17131102085113525, "learning_rate": 9.343259971724563e-05, "loss": 0.0128, "step": 16540 }, { "grad_norm": 0.14745129644870758, "learning_rate": 9.342235642452413e-05, "loss": 0.0108, "step": 16550 }, { "grad_norm": 0.2612754702568054, "learning_rate": 9.341210571210954e-05, "loss": 0.0151, "step": 16560 }, { "grad_norm": 0.16644613444805145, "learning_rate": 9.340184758175338e-05, "loss": 0.0156, "step": 16570 }, { "grad_norm": 0.13887174427509308, "learning_rate": 9.339158203520854e-05, "loss": 0.0134, "step": 16580 }, { "grad_norm": 0.15758036077022552, "learning_rate": 9.338130907422908e-05, "loss": 0.0139, "step": 16590 }, { "grad_norm": 0.19017046689987183, "learning_rate": 9.337102870057037e-05, "loss": 0.0139, "step": 16600 }, { "grad_norm": 0.1715335100889206, "learning_rate": 9.336074091598907e-05, "loss": 0.0122, "step": 16610 }, { "grad_norm": 0.2065221667289734, "learning_rate": 9.335044572224306e-05, "loss": 0.0133, "step": 16620 }, { "grad_norm": 0.17273688316345215, "learning_rate": 9.334014312109151e-05, "loss": 0.0159, "step": 16630 }, { "grad_norm": 0.17008918523788452, "learning_rate": 9.332983311429486e-05, "loss": 0.0138, "step": 16640 }, { "grad_norm": 0.14901453256607056, "learning_rate": 9.33195157036148e-05, "loss": 0.0196, "step": 16650 }, { "grad_norm": 0.22838228940963745, "learning_rate": 9.330919089081432e-05, "loss": 0.0134, "step": 16660 }, { "grad_norm": 0.20237226784229279, "learning_rate": 9.32988586776576e-05, "loss": 0.0154, "step": 16670 }, { "grad_norm": 0.1727929413318634, "learning_rate": 9.328851906591016e-05, "loss": 0.0128, "step": 16680 }, { "grad_norm": 0.14901335537433624, "learning_rate": 9.327817205733875e-05, "loss": 0.012, "step": 16690 }, { "grad_norm": 0.16281628608703613, "learning_rate": 9.326781765371142e-05, "loss": 0.0117, "step": 16700 }, { "grad_norm": 0.1181926280260086, "learning_rate": 9.325745585679741e-05, "loss": 0.0141, "step": 16710 }, { "grad_norm": 0.16505885124206543, "learning_rate": 9.32470866683673e-05, "loss": 0.0125, "step": 16720 }, { "grad_norm": 0.22310473024845123, "learning_rate": 9.323671009019288e-05, "loss": 0.0127, "step": 16730 }, { "grad_norm": 0.18609865009784698, "learning_rate": 9.322632612404725e-05, "loss": 0.0168, "step": 16740 }, { "grad_norm": 0.18702957034111023, "learning_rate": 9.321593477170471e-05, "loss": 0.0148, "step": 16750 }, { "grad_norm": 0.1731112152338028, "learning_rate": 9.320553603494088e-05, "loss": 0.0155, "step": 16760 }, { "grad_norm": 0.18251252174377441, "learning_rate": 9.319512991553261e-05, "loss": 0.0168, "step": 16770 }, { "grad_norm": 0.22470144927501678, "learning_rate": 9.318471641525803e-05, "loss": 0.015, "step": 16780 }, { "grad_norm": 0.20350950956344604, "learning_rate": 9.317429553589652e-05, "loss": 0.013, "step": 16790 }, { "grad_norm": 0.16841988265514374, "learning_rate": 9.316386727922873e-05, "loss": 0.0148, "step": 16800 }, { "grad_norm": 0.2101232260465622, "learning_rate": 9.315343164703656e-05, "loss": 0.0111, "step": 16810 }, { "grad_norm": 0.2091200053691864, "learning_rate": 9.314298864110316e-05, "loss": 0.0133, "step": 16820 }, { "grad_norm": 0.2014889419078827, "learning_rate": 9.313253826321295e-05, "loss": 0.0149, "step": 16830 }, { "grad_norm": 0.2174564003944397, "learning_rate": 9.312208051515165e-05, "loss": 0.0128, "step": 16840 }, { "grad_norm": 0.22000622749328613, "learning_rate": 9.311161539870618e-05, "loss": 0.0156, "step": 16850 }, { "grad_norm": 0.19429561495780945, "learning_rate": 9.310114291566474e-05, "loss": 0.013, "step": 16860 }, { "grad_norm": 0.25182807445526123, "learning_rate": 9.309066306781679e-05, "loss": 0.0136, "step": 16870 }, { "grad_norm": 0.21879269182682037, "learning_rate": 9.308017585695306e-05, "loss": 0.0146, "step": 16880 }, { "grad_norm": 0.16775402426719666, "learning_rate": 9.306968128486552e-05, "loss": 0.014, "step": 16890 }, { "grad_norm": 0.21015304327011108, "learning_rate": 9.30591793533474e-05, "loss": 0.0146, "step": 16900 }, { "grad_norm": 0.17685796320438385, "learning_rate": 9.304867006419321e-05, "loss": 0.0126, "step": 16910 }, { "grad_norm": 0.18142317235469818, "learning_rate": 9.303815341919868e-05, "loss": 0.0181, "step": 16920 }, { "grad_norm": 0.15678000450134277, "learning_rate": 9.302762942016084e-05, "loss": 0.0123, "step": 16930 }, { "grad_norm": 0.15896353125572205, "learning_rate": 9.301709806887792e-05, "loss": 0.0132, "step": 16940 }, { "grad_norm": 0.1651400774717331, "learning_rate": 9.300655936714948e-05, "loss": 0.0158, "step": 16950 }, { "grad_norm": 0.15916705131530762, "learning_rate": 9.299601331677627e-05, "loss": 0.0169, "step": 16960 }, { "grad_norm": 0.21911557018756866, "learning_rate": 9.298545991956033e-05, "loss": 0.017, "step": 16970 }, { "grad_norm": 0.1588379144668579, "learning_rate": 9.297489917730493e-05, "loss": 0.0136, "step": 16980 }, { "grad_norm": 0.19507203996181488, "learning_rate": 9.296433109181464e-05, "loss": 0.0122, "step": 16990 }, { "grad_norm": 0.12231962382793427, "learning_rate": 9.295375566489523e-05, "loss": 0.0149, "step": 17000 }, { "grad_norm": 0.2651425898075104, "learning_rate": 9.294317289835379e-05, "loss": 0.0192, "step": 17010 }, { "grad_norm": 0.1801925003528595, "learning_rate": 9.293258279399859e-05, "loss": 0.0169, "step": 17020 }, { "grad_norm": 0.15901073813438416, "learning_rate": 9.292198535363919e-05, "loss": 0.014, "step": 17030 }, { "grad_norm": 0.1665562242269516, "learning_rate": 9.291138057908641e-05, "loss": 0.016, "step": 17040 }, { "grad_norm": 0.24653851985931396, "learning_rate": 9.290076847215234e-05, "loss": 0.016, "step": 17050 }, { "grad_norm": 0.17401795089244843, "learning_rate": 9.289014903465025e-05, "loss": 0.0156, "step": 17060 }, { "grad_norm": 0.21735665202140808, "learning_rate": 9.287952226839475e-05, "loss": 0.0145, "step": 17070 }, { "grad_norm": 0.15272307395935059, "learning_rate": 9.286888817520164e-05, "loss": 0.0116, "step": 17080 }, { "grad_norm": 0.1428499072790146, "learning_rate": 9.285824675688803e-05, "loss": 0.0163, "step": 17090 }, { "grad_norm": 0.13992807269096375, "learning_rate": 9.28475980152722e-05, "loss": 0.012, "step": 17100 }, { "grad_norm": 0.20532168447971344, "learning_rate": 9.283694195217379e-05, "loss": 0.0117, "step": 17110 }, { "grad_norm": 0.19014383852481842, "learning_rate": 9.282627856941356e-05, "loss": 0.0135, "step": 17120 }, { "grad_norm": 0.22833752632141113, "learning_rate": 9.281560786881363e-05, "loss": 0.0137, "step": 17130 }, { "grad_norm": 0.18526598811149597, "learning_rate": 9.280492985219733e-05, "loss": 0.0155, "step": 17140 }, { "grad_norm": 0.19062937796115875, "learning_rate": 9.279424452138924e-05, "loss": 0.0142, "step": 17150 }, { "grad_norm": 0.22262446582317352, "learning_rate": 9.278355187821517e-05, "loss": 0.0156, "step": 17160 }, { "grad_norm": 0.17228619754314423, "learning_rate": 9.277285192450224e-05, "loss": 0.0122, "step": 17170 }, { "grad_norm": 0.1610649973154068, "learning_rate": 9.276214466207875e-05, "loss": 0.0125, "step": 17180 }, { "grad_norm": 0.1859932839870453, "learning_rate": 9.275143009277427e-05, "loss": 0.0174, "step": 17190 }, { "grad_norm": 0.20759321749210358, "learning_rate": 9.274070821841964e-05, "loss": 0.0117, "step": 17200 }, { "grad_norm": 0.17737212777137756, "learning_rate": 9.272997904084696e-05, "loss": 0.0112, "step": 17210 }, { "grad_norm": 0.17654649913311005, "learning_rate": 9.271924256188951e-05, "loss": 0.013, "step": 17220 }, { "grad_norm": 0.25891897082328796, "learning_rate": 9.270849878338189e-05, "loss": 0.0166, "step": 17230 }, { "grad_norm": 0.18855370581150055, "learning_rate": 9.269774770715991e-05, "loss": 0.0167, "step": 17240 }, { "grad_norm": 0.2651379108428955, "learning_rate": 9.268698933506061e-05, "loss": 0.0123, "step": 17250 }, { "grad_norm": 0.15031670033931732, "learning_rate": 9.267622366892235e-05, "loss": 0.0151, "step": 17260 }, { "grad_norm": 0.19041188061237335, "learning_rate": 9.266545071058465e-05, "loss": 0.0139, "step": 17270 }, { "grad_norm": 0.13770002126693726, "learning_rate": 9.265467046188833e-05, "loss": 0.0131, "step": 17280 }, { "grad_norm": 0.13993360102176666, "learning_rate": 9.264388292467543e-05, "loss": 0.0118, "step": 17290 }, { "grad_norm": 0.18237127363681793, "learning_rate": 9.263308810078926e-05, "loss": 0.0134, "step": 17300 }, { "grad_norm": 0.1838529258966446, "learning_rate": 9.262228599207434e-05, "loss": 0.0129, "step": 17310 }, { "grad_norm": 0.20635727047920227, "learning_rate": 9.261147660037647e-05, "loss": 0.0136, "step": 17320 }, { "grad_norm": 0.20753958821296692, "learning_rate": 9.26006599275427e-05, "loss": 0.0165, "step": 17330 }, { "grad_norm": 0.23016618192195892, "learning_rate": 9.258983597542124e-05, "loss": 0.0148, "step": 17340 }, { "grad_norm": 0.21711838245391846, "learning_rate": 9.257900474586167e-05, "loss": 0.0164, "step": 17350 }, { "grad_norm": 0.18427689373493195, "learning_rate": 9.256816624071471e-05, "loss": 0.0146, "step": 17360 }, { "grad_norm": 0.22144655883312225, "learning_rate": 9.25573204618324e-05, "loss": 0.0126, "step": 17370 }, { "grad_norm": 0.17013725638389587, "learning_rate": 9.254646741106796e-05, "loss": 0.0134, "step": 17380 }, { "grad_norm": 0.19174036383628845, "learning_rate": 9.253560709027589e-05, "loss": 0.0139, "step": 17390 }, { "grad_norm": 0.17933440208435059, "learning_rate": 9.252473950131192e-05, "loss": 0.0157, "step": 17400 }, { "grad_norm": 0.1593896746635437, "learning_rate": 9.251386464603302e-05, "loss": 0.0137, "step": 17410 }, { "grad_norm": 0.13276706635951996, "learning_rate": 9.250298252629741e-05, "loss": 0.0119, "step": 17420 }, { "grad_norm": 0.24178029596805573, "learning_rate": 9.249209314396454e-05, "loss": 0.0132, "step": 17430 }, { "grad_norm": 0.17739956080913544, "learning_rate": 9.248119650089513e-05, "loss": 0.0139, "step": 17440 }, { "grad_norm": 0.20257575809955597, "learning_rate": 9.247029259895108e-05, "loss": 0.0155, "step": 17450 }, { "grad_norm": 0.12598265707492828, "learning_rate": 9.24593814399956e-05, "loss": 0.0117, "step": 17460 }, { "grad_norm": 0.21242749691009521, "learning_rate": 9.244846302589309e-05, "loss": 0.0153, "step": 17470 }, { "grad_norm": 0.16382955014705658, "learning_rate": 9.243753735850923e-05, "loss": 0.0139, "step": 17480 }, { "grad_norm": 0.2097787708044052, "learning_rate": 9.24266044397109e-05, "loss": 0.0135, "step": 17490 }, { "grad_norm": 0.19653743505477905, "learning_rate": 9.241566427136624e-05, "loss": 0.0143, "step": 17500 }, { "grad_norm": 0.1797575056552887, "learning_rate": 9.240471685534463e-05, "loss": 0.0133, "step": 17510 }, { "grad_norm": 0.23303283751010895, "learning_rate": 9.239376219351667e-05, "loss": 0.011, "step": 17520 }, { "grad_norm": 0.1612747311592102, "learning_rate": 9.238280028775425e-05, "loss": 0.0136, "step": 17530 }, { "grad_norm": 0.2138129323720932, "learning_rate": 9.237183113993041e-05, "loss": 0.012, "step": 17540 }, { "grad_norm": 0.40228506922721863, "learning_rate": 9.236085475191952e-05, "loss": 0.0117, "step": 17550 }, { "grad_norm": 0.21797437965869904, "learning_rate": 9.234987112559709e-05, "loss": 0.0136, "step": 17560 }, { "grad_norm": 0.16454502940177917, "learning_rate": 9.233888026283999e-05, "loss": 0.0144, "step": 17570 }, { "grad_norm": 0.18037325143814087, "learning_rate": 9.232788216552619e-05, "loss": 0.0146, "step": 17580 }, { "grad_norm": 0.12091096490621567, "learning_rate": 9.231687683553502e-05, "loss": 0.0147, "step": 17590 }, { "grad_norm": 0.1627086102962494, "learning_rate": 9.230586427474698e-05, "loss": 0.0121, "step": 17600 }, { "grad_norm": 0.17902030050754547, "learning_rate": 9.229484448504379e-05, "loss": 0.0136, "step": 17610 }, { "grad_norm": 0.13853690028190613, "learning_rate": 9.228381746830843e-05, "loss": 0.0124, "step": 17620 }, { "grad_norm": 0.16326120495796204, "learning_rate": 9.227278322642514e-05, "loss": 0.0142, "step": 17630 }, { "grad_norm": 0.16619136929512024, "learning_rate": 9.226174176127937e-05, "loss": 0.0126, "step": 17640 }, { "grad_norm": 0.18257743120193481, "learning_rate": 9.22506930747578e-05, "loss": 0.0108, "step": 17650 }, { "grad_norm": 0.18720301985740662, "learning_rate": 9.223963716874831e-05, "loss": 0.0109, "step": 17660 }, { "grad_norm": 0.15280914306640625, "learning_rate": 9.222857404514012e-05, "loss": 0.0129, "step": 17670 }, { "grad_norm": 0.16900137066841125, "learning_rate": 9.221750370582355e-05, "loss": 0.0159, "step": 17680 }, { "grad_norm": 0.20187008380889893, "learning_rate": 9.220642615269028e-05, "loss": 0.0132, "step": 17690 }, { "grad_norm": 0.16697333753108978, "learning_rate": 9.219534138763311e-05, "loss": 0.0137, "step": 17700 }, { "grad_norm": 0.1659562885761261, "learning_rate": 9.218424941254613e-05, "loss": 0.0117, "step": 17710 }, { "grad_norm": 0.16744118928909302, "learning_rate": 9.217315022932468e-05, "loss": 0.0156, "step": 17720 }, { "grad_norm": 0.18823124468326569, "learning_rate": 9.216204383986528e-05, "loss": 0.0123, "step": 17730 }, { "grad_norm": 0.19633594155311584, "learning_rate": 9.215093024606574e-05, "loss": 0.0169, "step": 17740 }, { "grad_norm": 0.20865464210510254, "learning_rate": 9.213980944982506e-05, "loss": 0.011, "step": 17750 }, { "grad_norm": 0.2081579715013504, "learning_rate": 9.212868145304346e-05, "loss": 0.0193, "step": 17760 }, { "grad_norm": 0.2152722030878067, "learning_rate": 9.211754625762241e-05, "loss": 0.016, "step": 17770 }, { "grad_norm": 0.16820742189884186, "learning_rate": 9.210640386546463e-05, "loss": 0.0156, "step": 17780 }, { "grad_norm": 0.15621528029441833, "learning_rate": 9.209525427847405e-05, "loss": 0.0145, "step": 17790 }, { "grad_norm": 0.19536343216896057, "learning_rate": 9.208409749855583e-05, "loss": 0.0131, "step": 17800 }, { "grad_norm": 0.22496946156024933, "learning_rate": 9.207293352761633e-05, "loss": 0.0166, "step": 17810 }, { "grad_norm": 0.1796385794878006, "learning_rate": 9.206176236756319e-05, "loss": 0.0149, "step": 17820 }, { "grad_norm": 0.14013731479644775, "learning_rate": 9.205058402030525e-05, "loss": 0.0119, "step": 17830 }, { "grad_norm": 0.1799067258834839, "learning_rate": 9.203939848775259e-05, "loss": 0.0139, "step": 17840 }, { "grad_norm": 0.22663234174251556, "learning_rate": 9.202820577181652e-05, "loss": 0.0136, "step": 17850 }, { "grad_norm": 0.21341551840305328, "learning_rate": 9.201700587440953e-05, "loss": 0.0152, "step": 17860 }, { "grad_norm": 0.21235667169094086, "learning_rate": 9.200579879744544e-05, "loss": 0.0166, "step": 17870 }, { "grad_norm": 0.20847250521183014, "learning_rate": 9.199458454283918e-05, "loss": 0.0131, "step": 17880 }, { "grad_norm": 0.24244031310081482, "learning_rate": 9.198336311250697e-05, "loss": 0.0164, "step": 17890 }, { "grad_norm": 0.17664207518100739, "learning_rate": 9.197213450836626e-05, "loss": 0.0142, "step": 17900 }, { "grad_norm": 0.20240342617034912, "learning_rate": 9.19608987323357e-05, "loss": 0.0148, "step": 17910 }, { "grad_norm": 0.15881052613258362, "learning_rate": 9.194965578633517e-05, "loss": 0.0148, "step": 17920 }, { "grad_norm": 0.14502176642417908, "learning_rate": 9.193840567228582e-05, "loss": 0.0118, "step": 17930 }, { "grad_norm": 0.19050702452659607, "learning_rate": 9.192714839210994e-05, "loss": 0.0144, "step": 17940 }, { "grad_norm": 0.1580478549003601, "learning_rate": 9.19158839477311e-05, "loss": 0.0106, "step": 17950 }, { "grad_norm": 0.24122609198093414, "learning_rate": 9.190461234107411e-05, "loss": 0.0162, "step": 17960 }, { "grad_norm": 0.1892397701740265, "learning_rate": 9.189333357406496e-05, "loss": 0.0149, "step": 17970 }, { "grad_norm": 0.2164130061864853, "learning_rate": 9.188204764863089e-05, "loss": 0.0135, "step": 17980 }, { "grad_norm": 0.17519353330135345, "learning_rate": 9.187075456670033e-05, "loss": 0.0145, "step": 17990 }, { "grad_norm": 0.16445520520210266, "learning_rate": 9.1859454330203e-05, "loss": 0.011, "step": 18000 }, { "grad_norm": 0.1517212688922882, "learning_rate": 9.18481469410698e-05, "loss": 0.0143, "step": 18010 }, { "grad_norm": 0.15926885604858398, "learning_rate": 9.183683240123281e-05, "loss": 0.0118, "step": 18020 }, { "grad_norm": 0.20616121590137482, "learning_rate": 9.182551071262541e-05, "loss": 0.0117, "step": 18030 }, { "grad_norm": 0.17007237672805786, "learning_rate": 9.181418187718218e-05, "loss": 0.0122, "step": 18040 }, { "grad_norm": 0.15647763013839722, "learning_rate": 9.180284589683888e-05, "loss": 0.0146, "step": 18050 }, { "grad_norm": 0.22422470152378082, "learning_rate": 9.17915027735325e-05, "loss": 0.0151, "step": 18060 }, { "grad_norm": 0.1501070261001587, "learning_rate": 9.178015250920133e-05, "loss": 0.0159, "step": 18070 }, { "grad_norm": 0.18095903098583221, "learning_rate": 9.176879510578477e-05, "loss": 0.0122, "step": 18080 }, { "grad_norm": 0.1992620974779129, "learning_rate": 9.17574305652235e-05, "loss": 0.0147, "step": 18090 }, { "grad_norm": 0.17437763512134552, "learning_rate": 9.174605888945942e-05, "loss": 0.0137, "step": 18100 }, { "grad_norm": 0.15271982550621033, "learning_rate": 9.173468008043564e-05, "loss": 0.0119, "step": 18110 }, { "grad_norm": 0.18603858351707458, "learning_rate": 9.172329414009648e-05, "loss": 0.0169, "step": 18120 }, { "grad_norm": 0.12792974710464478, "learning_rate": 9.171190107038747e-05, "loss": 0.0137, "step": 18130 }, { "grad_norm": 0.137007474899292, "learning_rate": 9.170050087325541e-05, "loss": 0.0106, "step": 18140 }, { "grad_norm": 0.15156742930412292, "learning_rate": 9.168909355064824e-05, "loss": 0.0109, "step": 18150 }, { "grad_norm": 0.17367218434810638, "learning_rate": 9.167767910451519e-05, "loss": 0.0128, "step": 18160 }, { "grad_norm": 0.1794963926076889, "learning_rate": 9.166625753680669e-05, "loss": 0.0121, "step": 18170 }, { "grad_norm": 0.16084620356559753, "learning_rate": 9.165482884947431e-05, "loss": 0.0126, "step": 18180 }, { "grad_norm": 0.14923535287380219, "learning_rate": 9.164339304447098e-05, "loss": 0.0133, "step": 18190 }, { "grad_norm": 0.21801245212554932, "learning_rate": 9.163195012375072e-05, "loss": 0.015, "step": 18200 }, { "grad_norm": 0.16680410504341125, "learning_rate": 9.16205000892688e-05, "loss": 0.0122, "step": 18210 }, { "grad_norm": 0.11541450023651123, "learning_rate": 9.160904294298175e-05, "loss": 0.0126, "step": 18220 }, { "grad_norm": 0.1710604876279831, "learning_rate": 9.159757868684727e-05, "loss": 0.0159, "step": 18230 }, { "grad_norm": 0.1606317162513733, "learning_rate": 9.15861073228243e-05, "loss": 0.0153, "step": 18240 }, { "grad_norm": 0.1500059962272644, "learning_rate": 9.157462885287296e-05, "loss": 0.0138, "step": 18250 }, { "grad_norm": 0.16559025645256042, "learning_rate": 9.156314327895461e-05, "loss": 0.014, "step": 18260 }, { "grad_norm": 0.1564033180475235, "learning_rate": 9.155165060303185e-05, "loss": 0.0145, "step": 18270 }, { "grad_norm": 0.16176164150238037, "learning_rate": 9.154015082706841e-05, "loss": 0.0136, "step": 18280 }, { "grad_norm": 0.19588464498519897, "learning_rate": 9.152864395302936e-05, "loss": 0.0117, "step": 18290 }, { "grad_norm": 0.1216060221195221, "learning_rate": 9.151712998288085e-05, "loss": 0.0109, "step": 18300 }, { "grad_norm": 0.19247251749038696, "learning_rate": 9.150560891859031e-05, "loss": 0.0175, "step": 18310 }, { "grad_norm": 0.1400294154882431, "learning_rate": 9.14940807621264e-05, "loss": 0.0118, "step": 18320 }, { "grad_norm": 0.20920270681381226, "learning_rate": 9.148254551545894e-05, "loss": 0.0159, "step": 18330 }, { "grad_norm": 0.24419790506362915, "learning_rate": 9.147100318055901e-05, "loss": 0.0143, "step": 18340 }, { "grad_norm": 0.21323910355567932, "learning_rate": 9.145945375939888e-05, "loss": 0.0137, "step": 18350 }, { "grad_norm": 0.19295136630535126, "learning_rate": 9.144789725395203e-05, "loss": 0.0132, "step": 18360 }, { "grad_norm": 0.33407384157180786, "learning_rate": 9.14363336661931e-05, "loss": 0.0118, "step": 18370 }, { "grad_norm": 0.22013448178768158, "learning_rate": 9.142476299809806e-05, "loss": 0.0141, "step": 18380 }, { "grad_norm": 0.213907852768898, "learning_rate": 9.1413185251644e-05, "loss": 0.0115, "step": 18390 }, { "grad_norm": 0.18205253779888153, "learning_rate": 9.140160042880923e-05, "loss": 0.0121, "step": 18400 }, { "grad_norm": 0.170466348528862, "learning_rate": 9.139000853157327e-05, "loss": 0.0116, "step": 18410 }, { "grad_norm": 0.15316399931907654, "learning_rate": 9.137840956191688e-05, "loss": 0.0123, "step": 18420 }, { "grad_norm": 0.16568119823932648, "learning_rate": 9.136680352182199e-05, "loss": 0.0148, "step": 18430 }, { "grad_norm": 0.1825479120016098, "learning_rate": 9.135519041327177e-05, "loss": 0.0149, "step": 18440 }, { "grad_norm": 0.17678920924663544, "learning_rate": 9.134357023825058e-05, "loss": 0.0138, "step": 18450 }, { "grad_norm": 0.18914444744586945, "learning_rate": 9.133194299874398e-05, "loss": 0.0131, "step": 18460 }, { "grad_norm": 0.1719539761543274, "learning_rate": 9.132030869673876e-05, "loss": 0.0146, "step": 18470 }, { "grad_norm": 0.1666962057352066, "learning_rate": 9.130866733422288e-05, "loss": 0.0131, "step": 18480 }, { "grad_norm": 0.17296157777309418, "learning_rate": 9.129701891318556e-05, "loss": 0.0161, "step": 18490 }, { "grad_norm": 0.2073615938425064, "learning_rate": 9.128536343561718e-05, "loss": 0.0142, "step": 18500 }, { "grad_norm": 0.20233534276485443, "learning_rate": 9.127370090350934e-05, "loss": 0.0138, "step": 18510 }, { "grad_norm": 0.19519734382629395, "learning_rate": 9.126203131885487e-05, "loss": 0.0141, "step": 18520 }, { "grad_norm": 0.19951848685741425, "learning_rate": 9.125035468364775e-05, "loss": 0.0133, "step": 18530 }, { "grad_norm": 0.25807660818099976, "learning_rate": 9.123867099988322e-05, "loss": 0.0158, "step": 18540 }, { "grad_norm": 0.22377526760101318, "learning_rate": 9.122698026955769e-05, "loss": 0.0161, "step": 18550 }, { "grad_norm": 0.19101595878601074, "learning_rate": 9.12152824946688e-05, "loss": 0.0141, "step": 18560 }, { "grad_norm": 0.16517424583435059, "learning_rate": 9.120357767721538e-05, "loss": 0.0133, "step": 18570 }, { "grad_norm": 0.15414109826087952, "learning_rate": 9.119186581919745e-05, "loss": 0.0121, "step": 18580 }, { "grad_norm": 0.16478826105594635, "learning_rate": 9.118014692261624e-05, "loss": 0.013, "step": 18590 }, { "grad_norm": 0.17089414596557617, "learning_rate": 9.116842098947422e-05, "loss": 0.0116, "step": 18600 }, { "grad_norm": 0.19202743470668793, "learning_rate": 9.115668802177499e-05, "loss": 0.0142, "step": 18610 }, { "grad_norm": 0.19099971652030945, "learning_rate": 9.114494802152342e-05, "loss": 0.0146, "step": 18620 }, { "grad_norm": 0.16103799641132355, "learning_rate": 9.113320099072555e-05, "loss": 0.0143, "step": 18630 }, { "grad_norm": 0.1427232027053833, "learning_rate": 9.112144693138864e-05, "loss": 0.0109, "step": 18640 }, { "grad_norm": 0.15710587799549103, "learning_rate": 9.110968584552111e-05, "loss": 0.0116, "step": 18650 }, { "grad_norm": 0.1425226479768753, "learning_rate": 9.109791773513264e-05, "loss": 0.0112, "step": 18660 }, { "grad_norm": 0.19688542187213898, "learning_rate": 9.108614260223403e-05, "loss": 0.0117, "step": 18670 }, { "grad_norm": 0.21799279749393463, "learning_rate": 9.107436044883738e-05, "loss": 0.0142, "step": 18680 }, { "grad_norm": 0.1813950538635254, "learning_rate": 9.10625712769559e-05, "loss": 0.0127, "step": 18690 }, { "grad_norm": 0.17664691805839539, "learning_rate": 9.105077508860406e-05, "loss": 0.0121, "step": 18700 }, { "grad_norm": 0.1390695869922638, "learning_rate": 9.103897188579751e-05, "loss": 0.0125, "step": 18710 }, { "grad_norm": 0.1558862328529358, "learning_rate": 9.102716167055308e-05, "loss": 0.0118, "step": 18720 }, { "grad_norm": 0.20968446135520935, "learning_rate": 9.10153444448888e-05, "loss": 0.016, "step": 18730 }, { "grad_norm": 0.1652984619140625, "learning_rate": 9.100352021082393e-05, "loss": 0.0126, "step": 18740 }, { "grad_norm": 0.2140059471130371, "learning_rate": 9.099168897037891e-05, "loss": 0.0121, "step": 18750 }, { "grad_norm": 0.12128559499979019, "learning_rate": 9.097985072557538e-05, "loss": 0.0088, "step": 18760 }, { "grad_norm": 0.2227984070777893, "learning_rate": 9.096800547843615e-05, "loss": 0.0106, "step": 18770 }, { "grad_norm": 0.11223388463258743, "learning_rate": 9.095615323098526e-05, "loss": 0.0116, "step": 18780 }, { "grad_norm": 0.18303819000720978, "learning_rate": 9.094429398524795e-05, "loss": 0.0105, "step": 18790 }, { "grad_norm": 0.20908120274543762, "learning_rate": 9.093242774325061e-05, "loss": 0.0138, "step": 18800 }, { "grad_norm": 0.19817358255386353, "learning_rate": 9.092055450702088e-05, "loss": 0.0121, "step": 18810 }, { "grad_norm": 0.16806824505329132, "learning_rate": 9.090867427858756e-05, "loss": 0.0175, "step": 18820 }, { "grad_norm": 0.18254899978637695, "learning_rate": 9.089678705998066e-05, "loss": 0.0126, "step": 18830 }, { "grad_norm": 0.21638166904449463, "learning_rate": 9.088489285323139e-05, "loss": 0.0127, "step": 18840 }, { "grad_norm": 0.21583996713161469, "learning_rate": 9.087299166037212e-05, "loss": 0.0134, "step": 18850 }, { "grad_norm": 0.18967284262180328, "learning_rate": 9.086108348343647e-05, "loss": 0.0108, "step": 18860 }, { "grad_norm": 0.15250705182552338, "learning_rate": 9.08491683244592e-05, "loss": 0.0105, "step": 18870 }, { "grad_norm": 0.1655692309141159, "learning_rate": 9.08372461854763e-05, "loss": 0.0127, "step": 18880 }, { "grad_norm": 0.1682489663362503, "learning_rate": 9.082531706852492e-05, "loss": 0.0147, "step": 18890 }, { "grad_norm": 0.15783154964447021, "learning_rate": 9.081338097564342e-05, "loss": 0.012, "step": 18900 }, { "grad_norm": 0.1730799525976181, "learning_rate": 9.080143790887137e-05, "loss": 0.0128, "step": 18910 }, { "grad_norm": 0.19487155973911285, "learning_rate": 9.07894878702495e-05, "loss": 0.0126, "step": 18920 }, { "grad_norm": 0.15859703719615936, "learning_rate": 9.077753086181974e-05, "loss": 0.0141, "step": 18930 }, { "grad_norm": 0.1806102991104126, "learning_rate": 9.076556688562524e-05, "loss": 0.0128, "step": 18940 }, { "grad_norm": 0.1629876047372818, "learning_rate": 9.075359594371029e-05, "loss": 0.0141, "step": 18950 }, { "grad_norm": 0.2116788923740387, "learning_rate": 9.07416180381204e-05, "loss": 0.0149, "step": 18960 }, { "grad_norm": 0.21932393312454224, "learning_rate": 9.072963317090228e-05, "loss": 0.0127, "step": 18970 }, { "grad_norm": 0.17755873501300812, "learning_rate": 9.071764134410382e-05, "loss": 0.0128, "step": 18980 }, { "grad_norm": 0.16978396475315094, "learning_rate": 9.070564255977407e-05, "loss": 0.0134, "step": 18990 }, { "grad_norm": 0.15558291971683502, "learning_rate": 9.06936368199633e-05, "loss": 0.0144, "step": 19000 }, { "grad_norm": 0.21362631022930145, "learning_rate": 9.0681624126723e-05, "loss": 0.0131, "step": 19010 }, { "grad_norm": 0.15842536091804504, "learning_rate": 9.066960448210576e-05, "loss": 0.0128, "step": 19020 }, { "grad_norm": 0.21659120917320251, "learning_rate": 9.065757788816543e-05, "loss": 0.0145, "step": 19030 }, { "grad_norm": 0.20719420909881592, "learning_rate": 9.064554434695705e-05, "loss": 0.0115, "step": 19040 }, { "grad_norm": 0.1853378266096115, "learning_rate": 9.063350386053677e-05, "loss": 0.0122, "step": 19050 }, { "grad_norm": 0.16207851469516754, "learning_rate": 9.062145643096202e-05, "loss": 0.012, "step": 19060 }, { "grad_norm": 0.18837369978427887, "learning_rate": 9.060940206029136e-05, "loss": 0.0148, "step": 19070 }, { "grad_norm": 0.2358926236629486, "learning_rate": 9.059734075058457e-05, "loss": 0.0159, "step": 19080 }, { "grad_norm": 0.20649117231369019, "learning_rate": 9.058527250390257e-05, "loss": 0.0132, "step": 19090 }, { "grad_norm": 0.15739327669143677, "learning_rate": 9.057319732230752e-05, "loss": 0.0131, "step": 19100 }, { "grad_norm": 0.11631253361701965, "learning_rate": 9.056111520786273e-05, "loss": 0.0116, "step": 19110 }, { "grad_norm": 0.1558752954006195, "learning_rate": 9.054902616263268e-05, "loss": 0.0137, "step": 19120 }, { "grad_norm": 0.18798238039016724, "learning_rate": 9.05369301886831e-05, "loss": 0.0114, "step": 19130 }, { "grad_norm": 0.21035727858543396, "learning_rate": 9.052482728808083e-05, "loss": 0.0114, "step": 19140 }, { "grad_norm": 0.17892824113368988, "learning_rate": 9.051271746289391e-05, "loss": 0.0117, "step": 19150 }, { "grad_norm": 0.15115466713905334, "learning_rate": 9.050060071519162e-05, "loss": 0.0103, "step": 19160 }, { "grad_norm": 0.14821623265743256, "learning_rate": 9.048847704704437e-05, "loss": 0.011, "step": 19170 }, { "grad_norm": 0.15745961666107178, "learning_rate": 9.047634646052376e-05, "loss": 0.0138, "step": 19180 }, { "grad_norm": 0.13289059698581696, "learning_rate": 9.046420895770256e-05, "loss": 0.0126, "step": 19190 }, { "grad_norm": 0.15205134451389313, "learning_rate": 9.045206454065473e-05, "loss": 0.0119, "step": 19200 }, { "grad_norm": 0.12743744254112244, "learning_rate": 9.043991321145546e-05, "loss": 0.0099, "step": 19210 }, { "grad_norm": 0.14787529408931732, "learning_rate": 9.042775497218105e-05, "loss": 0.0095, "step": 19220 }, { "grad_norm": 0.213703453540802, "learning_rate": 9.041558982490901e-05, "loss": 0.012, "step": 19230 }, { "grad_norm": 0.1238865777850151, "learning_rate": 9.040341777171805e-05, "loss": 0.012, "step": 19240 }, { "grad_norm": 0.15548966825008392, "learning_rate": 9.039123881468802e-05, "loss": 0.0139, "step": 19250 }, { "grad_norm": 0.13444682955741882, "learning_rate": 9.037905295589998e-05, "loss": 0.0098, "step": 19260 }, { "grad_norm": 0.14257341623306274, "learning_rate": 9.036686019743617e-05, "loss": 0.0116, "step": 19270 }, { "grad_norm": 0.1959122270345688, "learning_rate": 9.035466054137997e-05, "loss": 0.0139, "step": 19280 }, { "grad_norm": 0.16291335225105286, "learning_rate": 9.0342453989816e-05, "loss": 0.0133, "step": 19290 }, { "grad_norm": 0.15779830515384674, "learning_rate": 9.033024054483e-05, "loss": 0.0097, "step": 19300 }, { "grad_norm": 0.16115352511405945, "learning_rate": 9.031802020850894e-05, "loss": 0.0098, "step": 19310 }, { "grad_norm": 0.1699976623058319, "learning_rate": 9.030579298294092e-05, "loss": 0.0129, "step": 19320 }, { "grad_norm": 0.19057117402553558, "learning_rate": 9.029355887021524e-05, "loss": 0.0157, "step": 19330 }, { "grad_norm": 0.2208966612815857, "learning_rate": 9.028131787242238e-05, "loss": 0.0139, "step": 19340 }, { "grad_norm": 0.15297353267669678, "learning_rate": 9.026906999165399e-05, "loss": 0.0121, "step": 19350 }, { "grad_norm": 0.14919699728488922, "learning_rate": 9.025681523000291e-05, "loss": 0.012, "step": 19360 }, { "grad_norm": 0.1497335582971573, "learning_rate": 9.024455358956315e-05, "loss": 0.0112, "step": 19370 }, { "grad_norm": 0.16103237867355347, "learning_rate": 9.023228507242984e-05, "loss": 0.012, "step": 19380 }, { "grad_norm": 0.15775282680988312, "learning_rate": 9.022000968069937e-05, "loss": 0.0141, "step": 19390 }, { "grad_norm": 0.2044057697057724, "learning_rate": 9.020772741646928e-05, "loss": 0.0137, "step": 19400 }, { "grad_norm": 0.186885267496109, "learning_rate": 9.019543828183826e-05, "loss": 0.015, "step": 19410 }, { "grad_norm": 0.14713314175605774, "learning_rate": 9.018314227890616e-05, "loss": 0.0115, "step": 19420 }, { "grad_norm": 0.2185310423374176, "learning_rate": 9.017083940977408e-05, "loss": 0.0109, "step": 19430 }, { "grad_norm": 0.18490713834762573, "learning_rate": 9.015852967654422e-05, "loss": 0.0119, "step": 19440 }, { "grad_norm": 0.2125953733921051, "learning_rate": 9.014621308131996e-05, "loss": 0.0118, "step": 19450 }, { "grad_norm": 0.18960866332054138, "learning_rate": 9.01338896262059e-05, "loss": 0.0106, "step": 19460 }, { "grad_norm": 0.14531509578227997, "learning_rate": 9.012155931330777e-05, "loss": 0.0129, "step": 19470 }, { "grad_norm": 0.18266703188419342, "learning_rate": 9.010922214473246e-05, "loss": 0.0138, "step": 19480 }, { "grad_norm": 0.14042457938194275, "learning_rate": 9.009687812258808e-05, "loss": 0.0116, "step": 19490 }, { "grad_norm": 0.1970616728067398, "learning_rate": 9.00845272489839e-05, "loss": 0.0101, "step": 19500 }, { "grad_norm": 0.16763289272785187, "learning_rate": 9.007216952603031e-05, "loss": 0.0139, "step": 19510 }, { "grad_norm": 0.1574927717447281, "learning_rate": 9.005980495583894e-05, "loss": 0.0111, "step": 19520 }, { "grad_norm": 0.20359042286872864, "learning_rate": 9.004743354052252e-05, "loss": 0.0118, "step": 19530 }, { "grad_norm": 0.12863026559352875, "learning_rate": 9.003505528219503e-05, "loss": 0.0106, "step": 19540 }, { "grad_norm": 0.2274320125579834, "learning_rate": 9.002267018297154e-05, "loss": 0.014, "step": 19550 }, { "grad_norm": 0.1863194704055786, "learning_rate": 9.001027824496834e-05, "loss": 0.0123, "step": 19560 }, { "grad_norm": 0.15403300523757935, "learning_rate": 8.999787947030287e-05, "loss": 0.0128, "step": 19570 }, { "grad_norm": 0.20692487061023712, "learning_rate": 8.998547386109376e-05, "loss": 0.0095, "step": 19580 }, { "grad_norm": 0.21347887814044952, "learning_rate": 8.997306141946073e-05, "loss": 0.0141, "step": 19590 }, { "grad_norm": 0.18330919742584229, "learning_rate": 8.996064214752481e-05, "loss": 0.0144, "step": 19600 }, { "grad_norm": 0.15147873759269714, "learning_rate": 8.994821604740806e-05, "loss": 0.0114, "step": 19610 }, { "grad_norm": 0.21634122729301453, "learning_rate": 8.993578312123377e-05, "loss": 0.0131, "step": 19620 }, { "grad_norm": 0.16738319396972656, "learning_rate": 8.992334337112639e-05, "loss": 0.0122, "step": 19630 }, { "grad_norm": 0.2196042239665985, "learning_rate": 8.991089679921154e-05, "loss": 0.0153, "step": 19640 }, { "grad_norm": 0.186531201004982, "learning_rate": 8.989844340761599e-05, "loss": 0.0128, "step": 19650 }, { "grad_norm": 0.17217138409614563, "learning_rate": 8.988598319846768e-05, "loss": 0.0114, "step": 19660 }, { "grad_norm": 0.12963542342185974, "learning_rate": 8.987351617389574e-05, "loss": 0.0119, "step": 19670 }, { "grad_norm": 0.18053196370601654, "learning_rate": 8.98610423360304e-05, "loss": 0.0142, "step": 19680 }, { "grad_norm": 0.17507526278495789, "learning_rate": 8.984856168700317e-05, "loss": 0.0109, "step": 19690 }, { "grad_norm": 0.18384018540382385, "learning_rate": 8.983607422894658e-05, "loss": 0.0101, "step": 19700 }, { "grad_norm": 0.19164909422397614, "learning_rate": 8.982357996399442e-05, "loss": 0.0138, "step": 19710 }, { "grad_norm": 0.13584749400615692, "learning_rate": 8.981107889428164e-05, "loss": 0.011, "step": 19720 }, { "grad_norm": 0.19808608293533325, "learning_rate": 8.979857102194428e-05, "loss": 0.0154, "step": 19730 }, { "grad_norm": 0.15770451724529266, "learning_rate": 8.978605634911968e-05, "loss": 0.012, "step": 19740 }, { "grad_norm": 0.15566027164459229, "learning_rate": 8.977353487794616e-05, "loss": 0.0116, "step": 19750 }, { "grad_norm": 0.12780232727527618, "learning_rate": 8.976100661056334e-05, "loss": 0.0097, "step": 19760 }, { "grad_norm": 0.153254434466362, "learning_rate": 8.974847154911197e-05, "loss": 0.0126, "step": 19770 }, { "grad_norm": 0.16228877007961273, "learning_rate": 8.973592969573393e-05, "loss": 0.0111, "step": 19780 }, { "grad_norm": 0.13248680531978607, "learning_rate": 8.972338105257228e-05, "loss": 0.0117, "step": 19790 }, { "grad_norm": 0.16334618628025055, "learning_rate": 8.971082562177125e-05, "loss": 0.0102, "step": 19800 }, { "grad_norm": 0.22023648023605347, "learning_rate": 8.96982634054762e-05, "loss": 0.0124, "step": 19810 }, { "grad_norm": 0.1974465250968933, "learning_rate": 8.96856944058337e-05, "loss": 0.0121, "step": 19820 }, { "grad_norm": 0.22594384849071503, "learning_rate": 8.967311862499144e-05, "loss": 0.0144, "step": 19830 }, { "grad_norm": 0.1827198565006256, "learning_rate": 8.966053606509825e-05, "loss": 0.0134, "step": 19840 }, { "grad_norm": 0.19004255533218384, "learning_rate": 8.964794672830417e-05, "loss": 0.0114, "step": 19850 }, { "grad_norm": 0.15893465280532837, "learning_rate": 8.963535061676038e-05, "loss": 0.0145, "step": 19860 }, { "grad_norm": 0.16078674793243408, "learning_rate": 8.962274773261918e-05, "loss": 0.0146, "step": 19870 }, { "grad_norm": 0.15072260797023773, "learning_rate": 8.961013807803409e-05, "loss": 0.0146, "step": 19880 }, { "grad_norm": 0.24211744964122772, "learning_rate": 8.959752165515973e-05, "loss": 0.013, "step": 19890 }, { "grad_norm": 0.16286133229732513, "learning_rate": 8.958489846615193e-05, "loss": 0.0136, "step": 19900 }, { "grad_norm": 0.19395215809345245, "learning_rate": 8.957226851316762e-05, "loss": 0.0149, "step": 19910 }, { "grad_norm": 0.19108660519123077, "learning_rate": 8.955963179836493e-05, "loss": 0.0123, "step": 19920 }, { "grad_norm": 0.20050570368766785, "learning_rate": 8.954698832390312e-05, "loss": 0.0113, "step": 19930 }, { "grad_norm": 0.13333995640277863, "learning_rate": 8.953433809194263e-05, "loss": 0.0121, "step": 19940 }, { "grad_norm": 0.1517324447631836, "learning_rate": 8.9521681104645e-05, "loss": 0.0097, "step": 19950 }, { "grad_norm": 0.17090235650539398, "learning_rate": 8.9509017364173e-05, "loss": 0.0131, "step": 19960 }, { "grad_norm": 0.23141241073608398, "learning_rate": 8.949634687269052e-05, "loss": 0.0144, "step": 19970 }, { "grad_norm": 0.18972906470298767, "learning_rate": 8.948366963236259e-05, "loss": 0.0113, "step": 19980 }, { "grad_norm": 0.12762348353862762, "learning_rate": 8.947098564535538e-05, "loss": 0.0104, "step": 19990 }, { "grad_norm": 0.17370954155921936, "learning_rate": 8.945829491383627e-05, "loss": 0.0127, "step": 20000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }