{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.1411043405532837, "learning_rate": 2.25e-07, "loss": 1.5009, "step": 10 }, { "grad_norm": 1.1378308534622192, "learning_rate": 4.75e-07, "loss": 1.5282, "step": 20 }, { "grad_norm": 1.1929908990859985, "learning_rate": 7.25e-07, "loss": 1.5056, "step": 30 }, { "grad_norm": 0.9630304574966431, "learning_rate": 9.75e-07, "loss": 1.5199, "step": 40 }, { "grad_norm": 0.9362477660179138, "learning_rate": 1.2250000000000001e-06, "loss": 1.4897, "step": 50 }, { "grad_norm": 1.5471577644348145, "learning_rate": 1.475e-06, "loss": 1.4781, "step": 60 }, { "grad_norm": 2.098245859146118, "learning_rate": 1.7250000000000002e-06, "loss": 1.3914, "step": 70 }, { "grad_norm": 2.8530704975128174, "learning_rate": 1.975e-06, "loss": 1.3526, "step": 80 }, { "grad_norm": 4.912765979766846, "learning_rate": 2.225e-06, "loss": 1.3343, "step": 90 }, { "grad_norm": 1.669730305671692, "learning_rate": 2.4750000000000004e-06, "loss": 1.2544, "step": 100 }, { "grad_norm": 2.822474479675293, "learning_rate": 2.725e-06, "loss": 1.2903, "step": 110 }, { "grad_norm": 2.854647159576416, "learning_rate": 2.975e-06, "loss": 1.314, "step": 120 }, { "grad_norm": 2.0382018089294434, "learning_rate": 3.225e-06, "loss": 1.2963, "step": 130 }, { "grad_norm": 1.9853246212005615, "learning_rate": 3.4750000000000006e-06, "loss": 1.299, "step": 140 }, { "grad_norm": 1.8031747341156006, "learning_rate": 3.725e-06, "loss": 1.2535, "step": 150 }, { "grad_norm": 2.3167059421539307, "learning_rate": 3.975e-06, "loss": 1.2322, "step": 160 }, { "grad_norm": 2.6719274520874023, "learning_rate": 4.225e-06, "loss": 1.2418, "step": 170 }, { "grad_norm": 3.3177082538604736, "learning_rate": 4.475e-06, "loss": 1.2245, "step": 180 }, { "grad_norm": 1.9509849548339844, "learning_rate": 4.7250000000000005e-06, "loss": 1.1921, "step": 190 }, { "grad_norm": 2.8562092781066895, "learning_rate": 4.975000000000001e-06, "loss": 1.1698, "step": 200 }, { "grad_norm": 2.2511706352233887, "learning_rate": 5.225e-06, "loss": 1.1444, "step": 210 }, { "grad_norm": 1.7650641202926636, "learning_rate": 5.475e-06, "loss": 1.1722, "step": 220 }, { "grad_norm": 1.9069126844406128, "learning_rate": 5.725e-06, "loss": 1.1479, "step": 230 }, { "grad_norm": 1.7101662158966064, "learning_rate": 5.975e-06, "loss": 1.1639, "step": 240 }, { "grad_norm": 2.559478282928467, "learning_rate": 6.2250000000000005e-06, "loss": 1.1113, "step": 250 }, { "grad_norm": 2.0006308555603027, "learning_rate": 6.475000000000001e-06, "loss": 1.1351, "step": 260 }, { "grad_norm": 1.9709502458572388, "learning_rate": 6.725000000000001e-06, "loss": 1.1315, "step": 270 }, { "grad_norm": 2.0163707733154297, "learning_rate": 6.975000000000001e-06, "loss": 1.1461, "step": 280 }, { "grad_norm": 2.0533976554870605, "learning_rate": 7.2249999999999994e-06, "loss": 1.1498, "step": 290 }, { "grad_norm": 2.052779197692871, "learning_rate": 7.4750000000000004e-06, "loss": 1.1214, "step": 300 }, { "grad_norm": 1.8817071914672852, "learning_rate": 7.725e-06, "loss": 1.1402, "step": 310 }, { "grad_norm": 2.2815325260162354, "learning_rate": 7.975e-06, "loss": 1.1241, "step": 320 }, { "grad_norm": 1.95901620388031, "learning_rate": 8.225e-06, "loss": 1.1099, "step": 330 }, { "grad_norm": 2.3284692764282227, "learning_rate": 8.475000000000001e-06, "loss": 1.1055, "step": 340 }, { "grad_norm": 2.582881450653076, "learning_rate": 8.725e-06, "loss": 1.1139, "step": 350 }, { "grad_norm": 2.4560179710388184, "learning_rate": 8.975e-06, "loss": 1.1047, "step": 360 }, { "grad_norm": 3.305924654006958, "learning_rate": 9.225e-06, "loss": 1.0885, "step": 370 }, { "grad_norm": 1.8011983633041382, "learning_rate": 9.475e-06, "loss": 1.116, "step": 380 }, { "grad_norm": 3.822506904602051, "learning_rate": 9.725000000000001e-06, "loss": 1.1359, "step": 390 }, { "grad_norm": 2.1991934776306152, "learning_rate": 9.975e-06, "loss": 1.0812, "step": 400 }, { "grad_norm": 2.68794846534729, "learning_rate": 1.0225e-05, "loss": 1.1057, "step": 410 }, { "grad_norm": 1.784138560295105, "learning_rate": 1.0475e-05, "loss": 1.1195, "step": 420 }, { "grad_norm": 1.9934443235397339, "learning_rate": 1.0725e-05, "loss": 1.0893, "step": 430 }, { "grad_norm": 1.7036008834838867, "learning_rate": 1.0975e-05, "loss": 1.0932, "step": 440 }, { "grad_norm": 1.9385591745376587, "learning_rate": 1.1225e-05, "loss": 1.076, "step": 450 }, { "grad_norm": 1.73807954788208, "learning_rate": 1.1475000000000001e-05, "loss": 1.0933, "step": 460 }, { "grad_norm": 1.9219274520874023, "learning_rate": 1.1725e-05, "loss": 1.0574, "step": 470 }, { "grad_norm": 1.7374770641326904, "learning_rate": 1.1975e-05, "loss": 1.0949, "step": 480 }, { "grad_norm": 1.7972736358642578, "learning_rate": 1.2225e-05, "loss": 1.0866, "step": 490 }, { "grad_norm": 1.3599379062652588, "learning_rate": 1.2475e-05, "loss": 1.0731, "step": 500 }, { "grad_norm": 1.7818571329116821, "learning_rate": 1.2725000000000001e-05, "loss": 1.0877, "step": 510 }, { "grad_norm": 1.2559229135513306, "learning_rate": 1.2975e-05, "loss": 1.0685, "step": 520 }, { "grad_norm": 1.505492925643921, "learning_rate": 1.3225000000000001e-05, "loss": 1.0868, "step": 530 }, { "grad_norm": 1.5542023181915283, "learning_rate": 1.3475000000000002e-05, "loss": 1.0924, "step": 540 }, { "grad_norm": 1.9233747720718384, "learning_rate": 1.3725000000000002e-05, "loss": 1.1065, "step": 550 }, { "grad_norm": 1.9977113008499146, "learning_rate": 1.3975000000000003e-05, "loss": 1.1102, "step": 560 }, { "grad_norm": 1.6313462257385254, "learning_rate": 1.4225e-05, "loss": 1.082, "step": 570 }, { "grad_norm": 1.6658507585525513, "learning_rate": 1.4475e-05, "loss": 1.0513, "step": 580 }, { "grad_norm": 1.6478042602539062, "learning_rate": 1.4725e-05, "loss": 1.0718, "step": 590 }, { "grad_norm": 1.7092751264572144, "learning_rate": 1.4975e-05, "loss": 1.0786, "step": 600 }, { "grad_norm": 1.5405710935592651, "learning_rate": 1.5225e-05, "loss": 1.0615, "step": 610 }, { "grad_norm": 2.211653709411621, "learning_rate": 1.5475e-05, "loss": 1.0645, "step": 620 }, { "grad_norm": 2.179666042327881, "learning_rate": 1.5725e-05, "loss": 1.0768, "step": 630 }, { "grad_norm": 1.692984938621521, "learning_rate": 1.5975000000000002e-05, "loss": 1.0449, "step": 640 }, { "grad_norm": 2.203728199005127, "learning_rate": 1.6225e-05, "loss": 1.0434, "step": 650 }, { "grad_norm": 1.990425944328308, "learning_rate": 1.6475e-05, "loss": 1.014, "step": 660 }, { "grad_norm": 2.342724323272705, "learning_rate": 1.6725000000000003e-05, "loss": 0.9934, "step": 670 }, { "grad_norm": 2.312938928604126, "learning_rate": 1.6975000000000003e-05, "loss": 0.9954, "step": 680 }, { "grad_norm": 1.9315762519836426, "learning_rate": 1.7225e-05, "loss": 0.9658, "step": 690 }, { "grad_norm": 2.7544121742248535, "learning_rate": 1.7475e-05, "loss": 0.9719, "step": 700 }, { "grad_norm": 3.937480926513672, "learning_rate": 1.7725e-05, "loss": 0.9468, "step": 710 }, { "grad_norm": 2.0535695552825928, "learning_rate": 1.7975e-05, "loss": 0.9068, "step": 720 }, { "grad_norm": 2.432188034057617, "learning_rate": 1.8225e-05, "loss": 0.8915, "step": 730 }, { "grad_norm": 2.0674047470092773, "learning_rate": 1.8475000000000002e-05, "loss": 0.8659, "step": 740 }, { "grad_norm": 2.2209222316741943, "learning_rate": 1.8725e-05, "loss": 0.853, "step": 750 }, { "grad_norm": 3.0573770999908447, "learning_rate": 1.8975e-05, "loss": 0.8145, "step": 760 }, { "grad_norm": 2.6392455101013184, "learning_rate": 1.9225e-05, "loss": 0.8518, "step": 770 }, { "grad_norm": 3.2313618659973145, "learning_rate": 1.9475000000000002e-05, "loss": 0.7965, "step": 780 }, { "grad_norm": 2.5969715118408203, "learning_rate": 1.9725000000000002e-05, "loss": 0.7767, "step": 790 }, { "grad_norm": 3.815755844116211, "learning_rate": 1.9975e-05, "loss": 0.7982, "step": 800 }, { "grad_norm": 2.506633996963501, "learning_rate": 2.0225000000000004e-05, "loss": 0.7738, "step": 810 }, { "grad_norm": 2.1760659217834473, "learning_rate": 2.0475e-05, "loss": 0.7176, "step": 820 }, { "grad_norm": 3.0994036197662354, "learning_rate": 2.0725e-05, "loss": 0.7271, "step": 830 }, { "grad_norm": 3.205294132232666, "learning_rate": 2.0975e-05, "loss": 0.6934, "step": 840 }, { "grad_norm": 2.8753888607025146, "learning_rate": 2.1225e-05, "loss": 0.7157, "step": 850 }, { "grad_norm": 2.774038314819336, "learning_rate": 2.1475e-05, "loss": 0.6369, "step": 860 }, { "grad_norm": 2.6242783069610596, "learning_rate": 2.1725e-05, "loss": 0.6445, "step": 870 }, { "grad_norm": 2.728069305419922, "learning_rate": 2.1975000000000002e-05, "loss": 0.6065, "step": 880 }, { "grad_norm": 2.8003652095794678, "learning_rate": 2.2225e-05, "loss": 0.5917, "step": 890 }, { "grad_norm": 2.4765989780426025, "learning_rate": 2.2475e-05, "loss": 0.5648, "step": 900 }, { "grad_norm": 2.1017401218414307, "learning_rate": 2.2725000000000003e-05, "loss": 0.5774, "step": 910 }, { "grad_norm": 2.596670389175415, "learning_rate": 2.2975000000000003e-05, "loss": 0.5726, "step": 920 }, { "grad_norm": 2.542613983154297, "learning_rate": 2.3225000000000002e-05, "loss": 0.5722, "step": 930 }, { "grad_norm": 3.9585816860198975, "learning_rate": 2.3475e-05, "loss": 0.5505, "step": 940 }, { "grad_norm": 2.834043264389038, "learning_rate": 2.3725e-05, "loss": 0.5597, "step": 950 }, { "grad_norm": 2.56421160697937, "learning_rate": 2.3975e-05, "loss": 0.5038, "step": 960 }, { "grad_norm": 2.7280869483947754, "learning_rate": 2.4225e-05, "loss": 0.5212, "step": 970 }, { "grad_norm": 2.9279253482818604, "learning_rate": 2.4475000000000002e-05, "loss": 0.478, "step": 980 }, { "grad_norm": 3.620800495147705, "learning_rate": 2.4725e-05, "loss": 0.4766, "step": 990 }, { "grad_norm": 3.21756911277771, "learning_rate": 2.4975e-05, "loss": 0.417, "step": 1000 }, { "grad_norm": 3.6428780555725098, "learning_rate": 2.5225e-05, "loss": 0.4569, "step": 1010 }, { "grad_norm": 2.7199370861053467, "learning_rate": 2.5475e-05, "loss": 0.4544, "step": 1020 }, { "grad_norm": 2.9267494678497314, "learning_rate": 2.5725e-05, "loss": 0.4492, "step": 1030 }, { "grad_norm": 2.5317327976226807, "learning_rate": 2.5974999999999998e-05, "loss": 0.4054, "step": 1040 }, { "grad_norm": 3.1197638511657715, "learning_rate": 2.6225e-05, "loss": 0.452, "step": 1050 }, { "grad_norm": 2.872746229171753, "learning_rate": 2.6475e-05, "loss": 0.3763, "step": 1060 }, { "grad_norm": 2.5092437267303467, "learning_rate": 2.6725e-05, "loss": 0.3874, "step": 1070 }, { "grad_norm": 2.4246182441711426, "learning_rate": 2.6975000000000002e-05, "loss": 0.2995, "step": 1080 }, { "grad_norm": 2.8341257572174072, "learning_rate": 2.7225e-05, "loss": 0.3879, "step": 1090 }, { "grad_norm": 3.221226930618286, "learning_rate": 2.7475e-05, "loss": 0.355, "step": 1100 }, { "grad_norm": 2.8217883110046387, "learning_rate": 2.7725e-05, "loss": 0.3612, "step": 1110 }, { "grad_norm": 3.155372381210327, "learning_rate": 2.7975000000000002e-05, "loss": 0.3842, "step": 1120 }, { "grad_norm": 2.0959062576293945, "learning_rate": 2.8225e-05, "loss": 0.3442, "step": 1130 }, { "grad_norm": 2.3488192558288574, "learning_rate": 2.8475e-05, "loss": 0.3106, "step": 1140 }, { "grad_norm": 2.35886549949646, "learning_rate": 2.8725e-05, "loss": 0.2983, "step": 1150 }, { "grad_norm": 2.276590347290039, "learning_rate": 2.8975000000000003e-05, "loss": 0.3086, "step": 1160 }, { "grad_norm": 2.8849775791168213, "learning_rate": 2.9225000000000002e-05, "loss": 0.3286, "step": 1170 }, { "grad_norm": 2.6560330390930176, "learning_rate": 2.9475e-05, "loss": 0.3078, "step": 1180 }, { "grad_norm": 1.9796392917633057, "learning_rate": 2.9725000000000004e-05, "loss": 0.3126, "step": 1190 }, { "grad_norm": 2.858863592147827, "learning_rate": 2.9975000000000004e-05, "loss": 0.3094, "step": 1200 }, { "grad_norm": 2.387202739715576, "learning_rate": 3.0225000000000003e-05, "loss": 0.303, "step": 1210 }, { "grad_norm": 2.746229410171509, "learning_rate": 3.0475000000000002e-05, "loss": 0.284, "step": 1220 }, { "grad_norm": 2.354336977005005, "learning_rate": 3.0725e-05, "loss": 0.2993, "step": 1230 }, { "grad_norm": 2.460106134414673, "learning_rate": 3.0975e-05, "loss": 0.2913, "step": 1240 }, { "grad_norm": 2.765320301055908, "learning_rate": 3.122500000000001e-05, "loss": 0.2955, "step": 1250 }, { "grad_norm": 2.916910409927368, "learning_rate": 3.1475e-05, "loss": 0.2956, "step": 1260 }, { "grad_norm": 2.1253433227539062, "learning_rate": 3.1725e-05, "loss": 0.2611, "step": 1270 }, { "grad_norm": 2.73580002784729, "learning_rate": 3.1975e-05, "loss": 0.2842, "step": 1280 }, { "grad_norm": 2.893096446990967, "learning_rate": 3.2225e-05, "loss": 0.259, "step": 1290 }, { "grad_norm": 2.845269203186035, "learning_rate": 3.2474999999999997e-05, "loss": 0.2769, "step": 1300 }, { "grad_norm": 2.0829920768737793, "learning_rate": 3.2725e-05, "loss": 0.2859, "step": 1310 }, { "grad_norm": 2.9410297870635986, "learning_rate": 3.2975e-05, "loss": 0.3132, "step": 1320 }, { "grad_norm": 2.4630508422851562, "learning_rate": 3.3225e-05, "loss": 0.2467, "step": 1330 }, { "grad_norm": 1.9745759963989258, "learning_rate": 3.3475e-05, "loss": 0.2448, "step": 1340 }, { "grad_norm": 2.766003370285034, "learning_rate": 3.3725e-05, "loss": 0.2589, "step": 1350 }, { "grad_norm": 2.8162121772766113, "learning_rate": 3.3975e-05, "loss": 0.2848, "step": 1360 }, { "grad_norm": 1.9521484375, "learning_rate": 3.4225e-05, "loss": 0.2559, "step": 1370 }, { "grad_norm": 2.18934965133667, "learning_rate": 3.4475000000000005e-05, "loss": 0.2749, "step": 1380 }, { "grad_norm": 3.1683433055877686, "learning_rate": 3.4725000000000004e-05, "loss": 0.2606, "step": 1390 }, { "grad_norm": 1.8014354705810547, "learning_rate": 3.4975e-05, "loss": 0.2718, "step": 1400 }, { "grad_norm": 1.6640677452087402, "learning_rate": 3.5225e-05, "loss": 0.221, "step": 1410 }, { "grad_norm": 2.5776188373565674, "learning_rate": 3.5475e-05, "loss": 0.2821, "step": 1420 }, { "grad_norm": 2.946040391921997, "learning_rate": 3.5725e-05, "loss": 0.268, "step": 1430 }, { "grad_norm": 2.4129226207733154, "learning_rate": 3.5975e-05, "loss": 0.2639, "step": 1440 }, { "grad_norm": 2.116281270980835, "learning_rate": 3.6225000000000006e-05, "loss": 0.2561, "step": 1450 }, { "grad_norm": 3.1807713508605957, "learning_rate": 3.6475000000000006e-05, "loss": 0.2672, "step": 1460 }, { "grad_norm": 1.929166316986084, "learning_rate": 3.6725000000000005e-05, "loss": 0.2175, "step": 1470 }, { "grad_norm": 1.709700345993042, "learning_rate": 3.6975000000000004e-05, "loss": 0.2149, "step": 1480 }, { "grad_norm": 1.8080477714538574, "learning_rate": 3.7225000000000004e-05, "loss": 0.2819, "step": 1490 }, { "grad_norm": 1.754961609840393, "learning_rate": 3.7475e-05, "loss": 0.2337, "step": 1500 }, { "grad_norm": 2.4286534786224365, "learning_rate": 3.7725e-05, "loss": 0.2531, "step": 1510 }, { "grad_norm": 1.8299020528793335, "learning_rate": 3.7975e-05, "loss": 0.2345, "step": 1520 }, { "grad_norm": 2.0831751823425293, "learning_rate": 3.8225e-05, "loss": 0.2374, "step": 1530 }, { "grad_norm": 2.351322650909424, "learning_rate": 3.8475e-05, "loss": 0.2772, "step": 1540 }, { "grad_norm": 2.010362148284912, "learning_rate": 3.8725e-05, "loss": 0.3039, "step": 1550 }, { "grad_norm": 2.2971041202545166, "learning_rate": 3.8975e-05, "loss": 0.273, "step": 1560 }, { "grad_norm": 1.8951433897018433, "learning_rate": 3.9225e-05, "loss": 0.2247, "step": 1570 }, { "grad_norm": 1.5891499519348145, "learning_rate": 3.9475000000000004e-05, "loss": 0.2341, "step": 1580 }, { "grad_norm": 1.973850131034851, "learning_rate": 3.9725e-05, "loss": 0.2554, "step": 1590 }, { "grad_norm": 2.354398250579834, "learning_rate": 3.9975e-05, "loss": 0.2019, "step": 1600 }, { "grad_norm": 2.4019672870635986, "learning_rate": 4.0225e-05, "loss": 0.2266, "step": 1610 }, { "grad_norm": 1.8630294799804688, "learning_rate": 4.0475e-05, "loss": 0.2408, "step": 1620 }, { "grad_norm": 2.2136218547821045, "learning_rate": 4.0725e-05, "loss": 0.2161, "step": 1630 }, { "grad_norm": 1.9379124641418457, "learning_rate": 4.0975e-05, "loss": 0.2306, "step": 1640 }, { "grad_norm": 2.5303971767425537, "learning_rate": 4.1225e-05, "loss": 0.2661, "step": 1650 }, { "grad_norm": 1.960793375968933, "learning_rate": 4.1475000000000005e-05, "loss": 0.2298, "step": 1660 }, { "grad_norm": 1.8725700378417969, "learning_rate": 4.1725000000000005e-05, "loss": 0.1993, "step": 1670 }, { "grad_norm": 2.361534595489502, "learning_rate": 4.1975000000000004e-05, "loss": 0.2455, "step": 1680 }, { "grad_norm": 1.956240177154541, "learning_rate": 4.2225e-05, "loss": 0.2149, "step": 1690 }, { "grad_norm": 1.8507776260375977, "learning_rate": 4.2475e-05, "loss": 0.2279, "step": 1700 }, { "grad_norm": 1.8275498151779175, "learning_rate": 4.2725e-05, "loss": 0.176, "step": 1710 }, { "grad_norm": 1.9867188930511475, "learning_rate": 4.2975e-05, "loss": 0.2044, "step": 1720 }, { "grad_norm": 2.268561840057373, "learning_rate": 4.322500000000001e-05, "loss": 0.2273, "step": 1730 }, { "grad_norm": 1.9439998865127563, "learning_rate": 4.3475000000000006e-05, "loss": 0.2023, "step": 1740 }, { "grad_norm": 2.100299835205078, "learning_rate": 4.3725000000000006e-05, "loss": 0.1939, "step": 1750 }, { "grad_norm": 1.4146021604537964, "learning_rate": 4.3975e-05, "loss": 0.1827, "step": 1760 }, { "grad_norm": 2.51370906829834, "learning_rate": 4.4225e-05, "loss": 0.1949, "step": 1770 }, { "grad_norm": 1.8884246349334717, "learning_rate": 4.4475e-05, "loss": 0.1932, "step": 1780 }, { "grad_norm": 1.737455129623413, "learning_rate": 4.4725e-05, "loss": 0.1978, "step": 1790 }, { "grad_norm": 1.9068185091018677, "learning_rate": 4.4975e-05, "loss": 0.2089, "step": 1800 }, { "grad_norm": 2.191763162612915, "learning_rate": 4.5225e-05, "loss": 0.1951, "step": 1810 }, { "grad_norm": 1.6701914072036743, "learning_rate": 4.5475e-05, "loss": 0.2023, "step": 1820 }, { "grad_norm": 2.3558690547943115, "learning_rate": 4.5725e-05, "loss": 0.1878, "step": 1830 }, { "grad_norm": 1.9164665937423706, "learning_rate": 4.5975e-05, "loss": 0.1987, "step": 1840 }, { "grad_norm": 2.0304572582244873, "learning_rate": 4.6225e-05, "loss": 0.1874, "step": 1850 }, { "grad_norm": 2.116255283355713, "learning_rate": 4.6475000000000005e-05, "loss": 0.2299, "step": 1860 }, { "grad_norm": 1.4288641214370728, "learning_rate": 4.6725000000000004e-05, "loss": 0.193, "step": 1870 }, { "grad_norm": 2.3144893646240234, "learning_rate": 4.6975000000000003e-05, "loss": 0.2004, "step": 1880 }, { "grad_norm": 1.5949140787124634, "learning_rate": 4.7225e-05, "loss": 0.1935, "step": 1890 }, { "grad_norm": 1.884854793548584, "learning_rate": 4.7475e-05, "loss": 0.1896, "step": 1900 }, { "grad_norm": 1.5797553062438965, "learning_rate": 4.7725e-05, "loss": 0.2005, "step": 1910 }, { "grad_norm": 1.166224479675293, "learning_rate": 4.7975e-05, "loss": 0.162, "step": 1920 }, { "grad_norm": 1.939835786819458, "learning_rate": 4.822500000000001e-05, "loss": 0.2247, "step": 1930 }, { "grad_norm": 2.037465810775757, "learning_rate": 4.8475000000000006e-05, "loss": 0.1809, "step": 1940 }, { "grad_norm": 2.0143465995788574, "learning_rate": 4.8725000000000005e-05, "loss": 0.1741, "step": 1950 }, { "grad_norm": 1.9073787927627563, "learning_rate": 4.8975000000000005e-05, "loss": 0.2076, "step": 1960 }, { "grad_norm": 1.7466120719909668, "learning_rate": 4.9225000000000004e-05, "loss": 0.1745, "step": 1970 }, { "grad_norm": 2.1410043239593506, "learning_rate": 4.9475e-05, "loss": 0.1513, "step": 1980 }, { "grad_norm": 1.9507335424423218, "learning_rate": 4.9725e-05, "loss": 0.1849, "step": 1990 }, { "grad_norm": 1.7274888753890991, "learning_rate": 4.9975e-05, "loss": 0.1673, "step": 2000 }, { "grad_norm": 1.3407832384109497, "learning_rate": 5.0225e-05, "loss": 0.1649, "step": 2010 }, { "grad_norm": 1.7138310670852661, "learning_rate": 5.047500000000001e-05, "loss": 0.179, "step": 2020 }, { "grad_norm": 1.929917812347412, "learning_rate": 5.0725e-05, "loss": 0.2084, "step": 2030 }, { "grad_norm": 1.8077236413955688, "learning_rate": 5.0975000000000006e-05, "loss": 0.1833, "step": 2040 }, { "grad_norm": 2.7358601093292236, "learning_rate": 5.1225e-05, "loss": 0.165, "step": 2050 }, { "grad_norm": 1.2704538106918335, "learning_rate": 5.1475000000000004e-05, "loss": 0.1828, "step": 2060 }, { "grad_norm": 1.5390723943710327, "learning_rate": 5.1725000000000004e-05, "loss": 0.1618, "step": 2070 }, { "grad_norm": 1.5960735082626343, "learning_rate": 5.197500000000001e-05, "loss": 0.183, "step": 2080 }, { "grad_norm": 1.801072120666504, "learning_rate": 5.2225e-05, "loss": 0.1864, "step": 2090 }, { "grad_norm": 1.2170137166976929, "learning_rate": 5.247500000000001e-05, "loss": 0.1683, "step": 2100 }, { "grad_norm": 1.4243305921554565, "learning_rate": 5.2725e-05, "loss": 0.1807, "step": 2110 }, { "grad_norm": 1.5292835235595703, "learning_rate": 5.297500000000001e-05, "loss": 0.2044, "step": 2120 }, { "grad_norm": 1.9178295135498047, "learning_rate": 5.3225e-05, "loss": 0.2156, "step": 2130 }, { "grad_norm": 1.3854104280471802, "learning_rate": 5.3475e-05, "loss": 0.183, "step": 2140 }, { "grad_norm": 1.3958868980407715, "learning_rate": 5.3725000000000005e-05, "loss": 0.2054, "step": 2150 }, { "grad_norm": 1.2939783334732056, "learning_rate": 5.3975e-05, "loss": 0.2177, "step": 2160 }, { "grad_norm": 1.9570180177688599, "learning_rate": 5.4225000000000003e-05, "loss": 0.1733, "step": 2170 }, { "grad_norm": 1.4811259508132935, "learning_rate": 5.4474999999999996e-05, "loss": 0.185, "step": 2180 }, { "grad_norm": 1.365875482559204, "learning_rate": 5.4725e-05, "loss": 0.1629, "step": 2190 }, { "grad_norm": 1.4355844259262085, "learning_rate": 5.4975e-05, "loss": 0.1715, "step": 2200 }, { "grad_norm": 1.7018288373947144, "learning_rate": 5.522500000000001e-05, "loss": 0.2065, "step": 2210 }, { "grad_norm": 1.9805059432983398, "learning_rate": 5.5475e-05, "loss": 0.239, "step": 2220 }, { "grad_norm": 1.4645628929138184, "learning_rate": 5.5725000000000006e-05, "loss": 0.1586, "step": 2230 }, { "grad_norm": 1.2981822490692139, "learning_rate": 5.5975e-05, "loss": 0.1348, "step": 2240 }, { "grad_norm": 1.4443062543869019, "learning_rate": 5.6225000000000005e-05, "loss": 0.172, "step": 2250 }, { "grad_norm": 1.2464312314987183, "learning_rate": 5.6475e-05, "loss": 0.1889, "step": 2260 }, { "grad_norm": 1.7011979818344116, "learning_rate": 5.6725e-05, "loss": 0.2272, "step": 2270 }, { "grad_norm": 2.3886332511901855, "learning_rate": 5.6975e-05, "loss": 0.2326, "step": 2280 }, { "grad_norm": 1.9743016958236694, "learning_rate": 5.722500000000001e-05, "loss": 0.2131, "step": 2290 }, { "grad_norm": 1.2327812910079956, "learning_rate": 5.7475e-05, "loss": 0.2093, "step": 2300 }, { "grad_norm": 1.9890296459197998, "learning_rate": 5.772500000000001e-05, "loss": 0.1672, "step": 2310 }, { "grad_norm": 1.2514920234680176, "learning_rate": 5.7975e-05, "loss": 0.1898, "step": 2320 }, { "grad_norm": 1.7855437994003296, "learning_rate": 5.8225000000000006e-05, "loss": 0.1682, "step": 2330 }, { "grad_norm": 1.347716212272644, "learning_rate": 5.8475000000000005e-05, "loss": 0.1935, "step": 2340 }, { "grad_norm": 1.1115901470184326, "learning_rate": 5.8725000000000004e-05, "loss": 0.159, "step": 2350 }, { "grad_norm": 1.6505149602890015, "learning_rate": 5.8975000000000004e-05, "loss": 0.156, "step": 2360 }, { "grad_norm": 1.9659488201141357, "learning_rate": 5.922500000000001e-05, "loss": 0.1698, "step": 2370 }, { "grad_norm": 1.5143674612045288, "learning_rate": 5.9475e-05, "loss": 0.165, "step": 2380 }, { "grad_norm": 1.1423747539520264, "learning_rate": 5.9724999999999995e-05, "loss": 0.1552, "step": 2390 }, { "grad_norm": 1.8133280277252197, "learning_rate": 5.9975e-05, "loss": 0.2106, "step": 2400 }, { "grad_norm": 1.4063150882720947, "learning_rate": 6.0225e-05, "loss": 0.225, "step": 2410 }, { "grad_norm": 2.581613540649414, "learning_rate": 6.0475000000000006e-05, "loss": 0.1804, "step": 2420 }, { "grad_norm": 1.3815596103668213, "learning_rate": 6.0725e-05, "loss": 0.1767, "step": 2430 }, { "grad_norm": 2.2186145782470703, "learning_rate": 6.0975000000000005e-05, "loss": 0.1988, "step": 2440 }, { "grad_norm": 1.7920396327972412, "learning_rate": 6.1225e-05, "loss": 0.2254, "step": 2450 }, { "grad_norm": 1.6490808725357056, "learning_rate": 6.1475e-05, "loss": 0.1497, "step": 2460 }, { "grad_norm": 1.3957897424697876, "learning_rate": 6.1725e-05, "loss": 0.2371, "step": 2470 }, { "grad_norm": 1.9014972448349, "learning_rate": 6.1975e-05, "loss": 0.17, "step": 2480 }, { "grad_norm": 1.561697244644165, "learning_rate": 6.2225e-05, "loss": 0.1789, "step": 2490 }, { "grad_norm": 1.3052376508712769, "learning_rate": 6.2475e-05, "loss": 0.1739, "step": 2500 }, { "grad_norm": 1.3185702562332153, "learning_rate": 6.2725e-05, "loss": 0.1865, "step": 2510 }, { "grad_norm": 1.882246732711792, "learning_rate": 6.297500000000001e-05, "loss": 0.1815, "step": 2520 }, { "grad_norm": 1.228444218635559, "learning_rate": 6.3225e-05, "loss": 0.1614, "step": 2530 }, { "grad_norm": 1.5859390497207642, "learning_rate": 6.347500000000001e-05, "loss": 0.1633, "step": 2540 }, { "grad_norm": 1.4030574560165405, "learning_rate": 6.3725e-05, "loss": 0.1916, "step": 2550 }, { "grad_norm": 1.3214472532272339, "learning_rate": 6.397500000000001e-05, "loss": 0.1882, "step": 2560 }, { "grad_norm": 1.4843155145645142, "learning_rate": 6.4225e-05, "loss": 0.2118, "step": 2570 }, { "grad_norm": 1.0910192728042603, "learning_rate": 6.447500000000001e-05, "loss": 0.1591, "step": 2580 }, { "grad_norm": 1.3498103618621826, "learning_rate": 6.4725e-05, "loss": 0.1852, "step": 2590 }, { "grad_norm": 1.0905286073684692, "learning_rate": 6.497500000000001e-05, "loss": 0.172, "step": 2600 }, { "grad_norm": 1.2152172327041626, "learning_rate": 6.5225e-05, "loss": 0.1709, "step": 2610 }, { "grad_norm": 1.5024681091308594, "learning_rate": 6.5475e-05, "loss": 0.1681, "step": 2620 }, { "grad_norm": 1.112623691558838, "learning_rate": 6.5725e-05, "loss": 0.1697, "step": 2630 }, { "grad_norm": 1.2200078964233398, "learning_rate": 6.5975e-05, "loss": 0.1835, "step": 2640 }, { "grad_norm": 1.2298283576965332, "learning_rate": 6.6225e-05, "loss": 0.1696, "step": 2650 }, { "grad_norm": 1.0633589029312134, "learning_rate": 6.6475e-05, "loss": 0.1571, "step": 2660 }, { "grad_norm": 1.1219624280929565, "learning_rate": 6.672500000000001e-05, "loss": 0.1718, "step": 2670 }, { "grad_norm": 1.6035257577896118, "learning_rate": 6.6975e-05, "loss": 0.1988, "step": 2680 }, { "grad_norm": 1.2176463603973389, "learning_rate": 6.722500000000001e-05, "loss": 0.1707, "step": 2690 }, { "grad_norm": 0.9701326489448547, "learning_rate": 6.7475e-05, "loss": 0.1995, "step": 2700 }, { "grad_norm": 1.6673688888549805, "learning_rate": 6.7725e-05, "loss": 0.2039, "step": 2710 }, { "grad_norm": 1.6591843366622925, "learning_rate": 6.7975e-05, "loss": 0.1796, "step": 2720 }, { "grad_norm": 1.092907190322876, "learning_rate": 6.8225e-05, "loss": 0.1458, "step": 2730 }, { "grad_norm": 1.5041040182113647, "learning_rate": 6.8475e-05, "loss": 0.1589, "step": 2740 }, { "grad_norm": 1.0825576782226562, "learning_rate": 6.8725e-05, "loss": 0.1348, "step": 2750 }, { "grad_norm": 1.1705235242843628, "learning_rate": 6.8975e-05, "loss": 0.1578, "step": 2760 }, { "grad_norm": 1.2275880575180054, "learning_rate": 6.9225e-05, "loss": 0.1655, "step": 2770 }, { "grad_norm": 1.304119348526001, "learning_rate": 6.9475e-05, "loss": 0.1824, "step": 2780 }, { "grad_norm": 1.5699948072433472, "learning_rate": 6.9725e-05, "loss": 0.1517, "step": 2790 }, { "grad_norm": 1.2170822620391846, "learning_rate": 6.997500000000001e-05, "loss": 0.1707, "step": 2800 }, { "grad_norm": 1.0545234680175781, "learning_rate": 7.022500000000001e-05, "loss": 0.1709, "step": 2810 }, { "grad_norm": 1.0934690237045288, "learning_rate": 7.0475e-05, "loss": 0.1702, "step": 2820 }, { "grad_norm": 1.3907052278518677, "learning_rate": 7.072500000000001e-05, "loss": 0.1619, "step": 2830 }, { "grad_norm": 1.6014525890350342, "learning_rate": 7.0975e-05, "loss": 0.206, "step": 2840 }, { "grad_norm": 1.5065544843673706, "learning_rate": 7.122500000000001e-05, "loss": 0.1442, "step": 2850 }, { "grad_norm": 1.5349990129470825, "learning_rate": 7.1475e-05, "loss": 0.1508, "step": 2860 }, { "grad_norm": 1.7157320976257324, "learning_rate": 7.172500000000001e-05, "loss": 0.1809, "step": 2870 }, { "grad_norm": 1.4333992004394531, "learning_rate": 7.1975e-05, "loss": 0.1693, "step": 2880 }, { "grad_norm": 1.1227190494537354, "learning_rate": 7.2225e-05, "loss": 0.2169, "step": 2890 }, { "grad_norm": 1.4129832983016968, "learning_rate": 7.2475e-05, "loss": 0.178, "step": 2900 }, { "grad_norm": 1.08963942527771, "learning_rate": 7.272499999999999e-05, "loss": 0.1938, "step": 2910 }, { "grad_norm": 0.9895902872085571, "learning_rate": 7.2975e-05, "loss": 0.1536, "step": 2920 }, { "grad_norm": 1.5164794921875, "learning_rate": 7.3225e-05, "loss": 0.1684, "step": 2930 }, { "grad_norm": 1.1421613693237305, "learning_rate": 7.347500000000001e-05, "loss": 0.1612, "step": 2940 }, { "grad_norm": 1.2887755632400513, "learning_rate": 7.3725e-05, "loss": 0.1612, "step": 2950 }, { "grad_norm": 1.0467487573623657, "learning_rate": 7.397500000000001e-05, "loss": 0.1453, "step": 2960 }, { "grad_norm": 1.5273332595825195, "learning_rate": 7.4225e-05, "loss": 0.1481, "step": 2970 }, { "grad_norm": 1.1094087362289429, "learning_rate": 7.447500000000001e-05, "loss": 0.1607, "step": 2980 }, { "grad_norm": 1.0182894468307495, "learning_rate": 7.4725e-05, "loss": 0.148, "step": 2990 }, { "grad_norm": 0.9867938756942749, "learning_rate": 7.4975e-05, "loss": 0.1688, "step": 3000 }, { "grad_norm": 1.203399419784546, "learning_rate": 7.5225e-05, "loss": 0.1295, "step": 3010 }, { "grad_norm": 1.147312045097351, "learning_rate": 7.5475e-05, "loss": 0.1634, "step": 3020 }, { "grad_norm": 0.881496012210846, "learning_rate": 7.5725e-05, "loss": 0.1448, "step": 3030 }, { "grad_norm": 1.5371818542480469, "learning_rate": 7.5975e-05, "loss": 0.1712, "step": 3040 }, { "grad_norm": 1.0979211330413818, "learning_rate": 7.6225e-05, "loss": 0.127, "step": 3050 }, { "grad_norm": 0.6949929594993591, "learning_rate": 7.6475e-05, "loss": 0.1755, "step": 3060 }, { "grad_norm": 1.2048890590667725, "learning_rate": 7.672500000000001e-05, "loss": 0.1961, "step": 3070 }, { "grad_norm": 0.9711630940437317, "learning_rate": 7.697500000000001e-05, "loss": 0.1511, "step": 3080 }, { "grad_norm": 1.0730910301208496, "learning_rate": 7.722500000000001e-05, "loss": 0.1529, "step": 3090 }, { "grad_norm": 1.4278315305709839, "learning_rate": 7.747500000000001e-05, "loss": 0.1633, "step": 3100 }, { "grad_norm": 1.2676589488983154, "learning_rate": 7.7725e-05, "loss": 0.1983, "step": 3110 }, { "grad_norm": 0.9549498558044434, "learning_rate": 7.797500000000001e-05, "loss": 0.1757, "step": 3120 }, { "grad_norm": 0.8361928462982178, "learning_rate": 7.8225e-05, "loss": 0.1281, "step": 3130 }, { "grad_norm": 1.0449154376983643, "learning_rate": 7.8475e-05, "loss": 0.1394, "step": 3140 }, { "grad_norm": 1.2515077590942383, "learning_rate": 7.8725e-05, "loss": 0.1567, "step": 3150 }, { "grad_norm": 1.0329166650772095, "learning_rate": 7.8975e-05, "loss": 0.1682, "step": 3160 }, { "grad_norm": 1.3104019165039062, "learning_rate": 7.9225e-05, "loss": 0.1631, "step": 3170 }, { "grad_norm": 1.096264123916626, "learning_rate": 7.9475e-05, "loss": 0.178, "step": 3180 }, { "grad_norm": 1.0527808666229248, "learning_rate": 7.9725e-05, "loss": 0.1526, "step": 3190 }, { "grad_norm": 1.0387372970581055, "learning_rate": 7.9975e-05, "loss": 0.1597, "step": 3200 }, { "grad_norm": 1.1987032890319824, "learning_rate": 8.022500000000001e-05, "loss": 0.1381, "step": 3210 }, { "grad_norm": 0.9307028651237488, "learning_rate": 8.0475e-05, "loss": 0.1372, "step": 3220 }, { "grad_norm": 1.7746381759643555, "learning_rate": 8.072500000000001e-05, "loss": 0.1631, "step": 3230 }, { "grad_norm": 1.0137569904327393, "learning_rate": 8.0975e-05, "loss": 0.1604, "step": 3240 }, { "grad_norm": 1.144358515739441, "learning_rate": 8.122500000000001e-05, "loss": 0.187, "step": 3250 }, { "grad_norm": 1.025037407875061, "learning_rate": 8.1475e-05, "loss": 0.2201, "step": 3260 }, { "grad_norm": 0.8855072855949402, "learning_rate": 8.172500000000001e-05, "loss": 0.1601, "step": 3270 }, { "grad_norm": 0.9119786620140076, "learning_rate": 8.1975e-05, "loss": 0.1671, "step": 3280 }, { "grad_norm": 1.0451996326446533, "learning_rate": 8.2225e-05, "loss": 0.1624, "step": 3290 }, { "grad_norm": 0.8755694627761841, "learning_rate": 8.2475e-05, "loss": 0.1684, "step": 3300 }, { "grad_norm": 1.2489674091339111, "learning_rate": 8.2725e-05, "loss": 0.1544, "step": 3310 }, { "grad_norm": 1.2031806707382202, "learning_rate": 8.2975e-05, "loss": 0.1907, "step": 3320 }, { "grad_norm": 1.210959553718567, "learning_rate": 8.3225e-05, "loss": 0.151, "step": 3330 }, { "grad_norm": 1.0400811433792114, "learning_rate": 8.347500000000001e-05, "loss": 0.1372, "step": 3340 }, { "grad_norm": 1.0443756580352783, "learning_rate": 8.3725e-05, "loss": 0.1615, "step": 3350 }, { "grad_norm": 1.1860963106155396, "learning_rate": 8.397500000000001e-05, "loss": 0.1641, "step": 3360 }, { "grad_norm": 1.3811856508255005, "learning_rate": 8.422500000000001e-05, "loss": 0.1755, "step": 3370 }, { "grad_norm": 0.975093424320221, "learning_rate": 8.447500000000001e-05, "loss": 0.158, "step": 3380 }, { "grad_norm": 1.018692970275879, "learning_rate": 8.4725e-05, "loss": 0.1604, "step": 3390 }, { "grad_norm": 1.3373653888702393, "learning_rate": 8.4975e-05, "loss": 0.171, "step": 3400 }, { "grad_norm": 1.1561702489852905, "learning_rate": 8.5225e-05, "loss": 0.1441, "step": 3410 }, { "grad_norm": 1.3654937744140625, "learning_rate": 8.5475e-05, "loss": 0.2057, "step": 3420 }, { "grad_norm": 0.874896228313446, "learning_rate": 8.5725e-05, "loss": 0.1763, "step": 3430 }, { "grad_norm": 1.136811375617981, "learning_rate": 8.5975e-05, "loss": 0.1739, "step": 3440 }, { "grad_norm": 1.1401480436325073, "learning_rate": 8.6225e-05, "loss": 0.174, "step": 3450 }, { "grad_norm": 1.199530005455017, "learning_rate": 8.6475e-05, "loss": 0.1683, "step": 3460 }, { "grad_norm": 1.3311399221420288, "learning_rate": 8.672500000000001e-05, "loss": 0.1738, "step": 3470 }, { "grad_norm": 1.1482478380203247, "learning_rate": 8.6975e-05, "loss": 0.1506, "step": 3480 }, { "grad_norm": 1.6440980434417725, "learning_rate": 8.7225e-05, "loss": 0.1694, "step": 3490 }, { "grad_norm": 1.590854525566101, "learning_rate": 8.747500000000001e-05, "loss": 0.1913, "step": 3500 }, { "grad_norm": 1.6173045635223389, "learning_rate": 8.7725e-05, "loss": 0.1672, "step": 3510 }, { "grad_norm": 1.3954148292541504, "learning_rate": 8.797500000000001e-05, "loss": 0.168, "step": 3520 }, { "grad_norm": 1.2949190139770508, "learning_rate": 8.8225e-05, "loss": 0.1478, "step": 3530 }, { "grad_norm": 1.060381531715393, "learning_rate": 8.847500000000001e-05, "loss": 0.1476, "step": 3540 }, { "grad_norm": 1.3913969993591309, "learning_rate": 8.8725e-05, "loss": 0.1814, "step": 3550 }, { "grad_norm": 0.8121777176856995, "learning_rate": 8.897500000000001e-05, "loss": 0.1776, "step": 3560 }, { "grad_norm": 1.1602299213409424, "learning_rate": 8.9225e-05, "loss": 0.1504, "step": 3570 }, { "grad_norm": 1.3946528434753418, "learning_rate": 8.9475e-05, "loss": 0.1592, "step": 3580 }, { "grad_norm": 1.2881159782409668, "learning_rate": 8.9725e-05, "loss": 0.1819, "step": 3590 }, { "grad_norm": 1.3772467374801636, "learning_rate": 8.9975e-05, "loss": 0.1831, "step": 3600 }, { "grad_norm": 1.0651711225509644, "learning_rate": 9.0225e-05, "loss": 0.1746, "step": 3610 }, { "grad_norm": 1.1944496631622314, "learning_rate": 9.0475e-05, "loss": 0.1415, "step": 3620 }, { "grad_norm": 1.1176005601882935, "learning_rate": 9.072500000000001e-05, "loss": 0.1419, "step": 3630 }, { "grad_norm": 1.700946569442749, "learning_rate": 9.0975e-05, "loss": 0.1845, "step": 3640 }, { "grad_norm": 1.4069651365280151, "learning_rate": 9.122500000000001e-05, "loss": 0.155, "step": 3650 }, { "grad_norm": 1.420601725578308, "learning_rate": 9.1475e-05, "loss": 0.1772, "step": 3660 }, { "grad_norm": 0.8058393001556396, "learning_rate": 9.172500000000001e-05, "loss": 0.1949, "step": 3670 }, { "grad_norm": 1.1936455965042114, "learning_rate": 9.1975e-05, "loss": 0.1464, "step": 3680 }, { "grad_norm": 1.133561372756958, "learning_rate": 9.2225e-05, "loss": 0.2283, "step": 3690 }, { "grad_norm": 1.2361021041870117, "learning_rate": 9.2475e-05, "loss": 0.1883, "step": 3700 }, { "grad_norm": 1.2071659564971924, "learning_rate": 9.2725e-05, "loss": 0.155, "step": 3710 }, { "grad_norm": 0.9159539937973022, "learning_rate": 9.2975e-05, "loss": 0.2008, "step": 3720 }, { "grad_norm": 0.987783670425415, "learning_rate": 9.3225e-05, "loss": 0.1417, "step": 3730 }, { "grad_norm": 1.3011003732681274, "learning_rate": 9.3475e-05, "loss": 0.161, "step": 3740 }, { "grad_norm": 1.0498920679092407, "learning_rate": 9.3725e-05, "loss": 0.1303, "step": 3750 }, { "grad_norm": 1.277905821800232, "learning_rate": 9.397500000000001e-05, "loss": 0.1492, "step": 3760 }, { "grad_norm": 0.9760408401489258, "learning_rate": 9.422500000000001e-05, "loss": 0.1718, "step": 3770 }, { "grad_norm": 1.7972551584243774, "learning_rate": 9.4475e-05, "loss": 0.1466, "step": 3780 }, { "grad_norm": 0.9655901193618774, "learning_rate": 9.472500000000001e-05, "loss": 0.1546, "step": 3790 }, { "grad_norm": 0.926257848739624, "learning_rate": 9.4975e-05, "loss": 0.1447, "step": 3800 }, { "grad_norm": 0.7397300601005554, "learning_rate": 9.522500000000001e-05, "loss": 0.16, "step": 3810 }, { "grad_norm": 0.9154761433601379, "learning_rate": 9.5475e-05, "loss": 0.196, "step": 3820 }, { "grad_norm": 1.221725344657898, "learning_rate": 9.572500000000001e-05, "loss": 0.1793, "step": 3830 }, { "grad_norm": 1.2185050249099731, "learning_rate": 9.5975e-05, "loss": 0.1757, "step": 3840 }, { "grad_norm": 0.9588304162025452, "learning_rate": 9.622500000000001e-05, "loss": 0.1545, "step": 3850 }, { "grad_norm": 1.0506041049957275, "learning_rate": 9.6475e-05, "loss": 0.151, "step": 3860 }, { "grad_norm": 1.1889941692352295, "learning_rate": 9.6725e-05, "loss": 0.1519, "step": 3870 }, { "grad_norm": 1.652692437171936, "learning_rate": 9.6975e-05, "loss": 0.1686, "step": 3880 }, { "grad_norm": 1.28211510181427, "learning_rate": 9.7225e-05, "loss": 0.1539, "step": 3890 }, { "grad_norm": 1.1183631420135498, "learning_rate": 9.747500000000001e-05, "loss": 0.1323, "step": 3900 }, { "grad_norm": 0.9137545824050903, "learning_rate": 9.7725e-05, "loss": 0.1954, "step": 3910 }, { "grad_norm": 1.2751446962356567, "learning_rate": 9.797500000000001e-05, "loss": 0.1855, "step": 3920 }, { "grad_norm": 1.209147334098816, "learning_rate": 9.8225e-05, "loss": 0.1392, "step": 3930 }, { "grad_norm": 1.2782448530197144, "learning_rate": 9.847500000000001e-05, "loss": 0.1501, "step": 3940 }, { "grad_norm": 1.0401923656463623, "learning_rate": 9.8725e-05, "loss": 0.1592, "step": 3950 }, { "grad_norm": 1.0987681150436401, "learning_rate": 9.897500000000001e-05, "loss": 0.1791, "step": 3960 }, { "grad_norm": 0.7724909782409668, "learning_rate": 9.9225e-05, "loss": 0.1392, "step": 3970 }, { "grad_norm": 1.603715419769287, "learning_rate": 9.9475e-05, "loss": 0.1584, "step": 3980 }, { "grad_norm": 1.1438980102539062, "learning_rate": 9.9725e-05, "loss": 0.1654, "step": 3990 }, { "grad_norm": 0.8095149397850037, "learning_rate": 9.9975e-05, "loss": 0.1173, "step": 4000 }, { "grad_norm": 1.025301456451416, "learning_rate": 9.999999653982884e-05, "loss": 0.1725, "step": 4010 }, { "grad_norm": 1.1890023946762085, "learning_rate": 9.999998457874392e-05, "loss": 0.2058, "step": 4020 }, { "grad_norm": 0.8732126355171204, "learning_rate": 9.999996407402913e-05, "loss": 0.1561, "step": 4030 }, { "grad_norm": 1.063908576965332, "learning_rate": 9.999993502568801e-05, "loss": 0.1929, "step": 4040 }, { "grad_norm": 1.2216095924377441, "learning_rate": 9.999989743372548e-05, "loss": 0.1283, "step": 4050 }, { "grad_norm": 1.1684223413467407, "learning_rate": 9.999985129814798e-05, "loss": 0.1605, "step": 4060 }, { "grad_norm": 0.8945369720458984, "learning_rate": 9.99997966189634e-05, "loss": 0.1395, "step": 4070 }, { "grad_norm": 1.179211139678955, "learning_rate": 9.999973339618107e-05, "loss": 0.1987, "step": 4080 }, { "grad_norm": 1.6644705533981323, "learning_rate": 9.999966162981179e-05, "loss": 0.1582, "step": 4090 }, { "grad_norm": 1.1816574335098267, "learning_rate": 9.999958131986784e-05, "loss": 0.1611, "step": 4100 }, { "grad_norm": 1.5145540237426758, "learning_rate": 9.999949246636293e-05, "loss": 0.175, "step": 4110 }, { "grad_norm": 1.3198506832122803, "learning_rate": 9.999939506931224e-05, "loss": 0.1731, "step": 4120 }, { "grad_norm": 1.1429126262664795, "learning_rate": 9.999928912873243e-05, "loss": 0.1438, "step": 4130 }, { "grad_norm": 1.172878384590149, "learning_rate": 9.999917464464159e-05, "loss": 0.1424, "step": 4140 }, { "grad_norm": 1.2027417421340942, "learning_rate": 9.999905161705929e-05, "loss": 0.1405, "step": 4150 }, { "grad_norm": 1.5733821392059326, "learning_rate": 9.999892004600653e-05, "loss": 0.1756, "step": 4160 }, { "grad_norm": 1.096047282218933, "learning_rate": 9.999877993150581e-05, "loss": 0.1313, "step": 4170 }, { "grad_norm": 0.8980554342269897, "learning_rate": 9.999863127358108e-05, "loss": 0.173, "step": 4180 }, { "grad_norm": 1.0166395902633667, "learning_rate": 9.999847407225773e-05, "loss": 0.1255, "step": 4190 }, { "grad_norm": 1.053863763809204, "learning_rate": 9.999830832756262e-05, "loss": 0.1678, "step": 4200 }, { "grad_norm": 0.975598931312561, "learning_rate": 9.999813403952407e-05, "loss": 0.1728, "step": 4210 }, { "grad_norm": 0.8927367925643921, "learning_rate": 9.999795120817187e-05, "loss": 0.1737, "step": 4220 }, { "grad_norm": 1.2994307279586792, "learning_rate": 9.999775983353725e-05, "loss": 0.152, "step": 4230 }, { "grad_norm": 1.0626661777496338, "learning_rate": 9.999755991565292e-05, "loss": 0.1416, "step": 4240 }, { "grad_norm": 0.9267320036888123, "learning_rate": 9.999735145455303e-05, "loss": 0.1753, "step": 4250 }, { "grad_norm": 0.9127563238143921, "learning_rate": 9.99971344502732e-05, "loss": 0.1361, "step": 4260 }, { "grad_norm": 0.9845833778381348, "learning_rate": 9.999690890285053e-05, "loss": 0.1237, "step": 4270 }, { "grad_norm": 1.192184567451477, "learning_rate": 9.999667481232356e-05, "loss": 0.1269, "step": 4280 }, { "grad_norm": 1.0745301246643066, "learning_rate": 9.999643217873225e-05, "loss": 0.1655, "step": 4290 }, { "grad_norm": 0.972606897354126, "learning_rate": 9.999618100211809e-05, "loss": 0.1475, "step": 4300 }, { "grad_norm": 0.8432597517967224, "learning_rate": 9.999592128252402e-05, "loss": 0.1519, "step": 4310 }, { "grad_norm": 0.8370494246482849, "learning_rate": 9.999565301999437e-05, "loss": 0.1471, "step": 4320 }, { "grad_norm": 1.066065788269043, "learning_rate": 9.999537621457502e-05, "loss": 0.153, "step": 4330 }, { "grad_norm": 1.1362944841384888, "learning_rate": 9.999509086631323e-05, "loss": 0.116, "step": 4340 }, { "grad_norm": 1.3521140813827515, "learning_rate": 9.99947969752578e-05, "loss": 0.1449, "step": 4350 }, { "grad_norm": 0.9327712655067444, "learning_rate": 9.999449454145891e-05, "loss": 0.1563, "step": 4360 }, { "grad_norm": 1.2750622034072876, "learning_rate": 9.999418356496827e-05, "loss": 0.1358, "step": 4370 }, { "grad_norm": 1.0172017812728882, "learning_rate": 9.999386404583899e-05, "loss": 0.1384, "step": 4380 }, { "grad_norm": 1.2895245552062988, "learning_rate": 9.999353598412568e-05, "loss": 0.1601, "step": 4390 }, { "grad_norm": 0.7912449240684509, "learning_rate": 9.999319937988442e-05, "loss": 0.1331, "step": 4400 }, { "grad_norm": 1.0768111944198608, "learning_rate": 9.999285423317268e-05, "loss": 0.1538, "step": 4410 }, { "grad_norm": 0.6195421814918518, "learning_rate": 9.999250054404947e-05, "loss": 0.1421, "step": 4420 }, { "grad_norm": 1.2882484197616577, "learning_rate": 9.99921383125752e-05, "loss": 0.1417, "step": 4430 }, { "grad_norm": 1.024831771850586, "learning_rate": 9.99917675388118e-05, "loss": 0.138, "step": 4440 }, { "grad_norm": 0.7856372594833374, "learning_rate": 9.99913882228226e-05, "loss": 0.1294, "step": 4450 }, { "grad_norm": 1.0359054803848267, "learning_rate": 9.999100036467242e-05, "loss": 0.1441, "step": 4460 }, { "grad_norm": 1.1032922267913818, "learning_rate": 9.999060396442753e-05, "loss": 0.1321, "step": 4470 }, { "grad_norm": 1.2156585454940796, "learning_rate": 9.999019902215566e-05, "loss": 0.1341, "step": 4480 }, { "grad_norm": 0.9167007207870483, "learning_rate": 9.998978553792602e-05, "loss": 0.1321, "step": 4490 }, { "grad_norm": 1.1567178964614868, "learning_rate": 9.998936351180926e-05, "loss": 0.1661, "step": 4500 }, { "grad_norm": 1.132697582244873, "learning_rate": 9.998893294387747e-05, "loss": 0.1451, "step": 4510 }, { "grad_norm": 1.086759090423584, "learning_rate": 9.998849383420426e-05, "loss": 0.1255, "step": 4520 }, { "grad_norm": 0.8024387359619141, "learning_rate": 9.998804618286465e-05, "loss": 0.1078, "step": 4530 }, { "grad_norm": 0.8866140842437744, "learning_rate": 9.99875899899351e-05, "loss": 0.1355, "step": 4540 }, { "grad_norm": 1.1190747022628784, "learning_rate": 9.99871252554936e-05, "loss": 0.1325, "step": 4550 }, { "grad_norm": 0.851883590221405, "learning_rate": 9.998665197961955e-05, "loss": 0.1232, "step": 4560 }, { "grad_norm": 0.6642021536827087, "learning_rate": 9.998617016239379e-05, "loss": 0.1418, "step": 4570 }, { "grad_norm": 1.1634786128997803, "learning_rate": 9.998567980389869e-05, "loss": 0.1641, "step": 4580 }, { "grad_norm": 2.115309000015259, "learning_rate": 9.998518090421802e-05, "loss": 0.1355, "step": 4590 }, { "grad_norm": 1.1101535558700562, "learning_rate": 9.998467346343703e-05, "loss": 0.1333, "step": 4600 }, { "grad_norm": 0.9310990571975708, "learning_rate": 9.998415748164243e-05, "loss": 0.1679, "step": 4610 }, { "grad_norm": 1.01819908618927, "learning_rate": 9.998363295892238e-05, "loss": 0.1555, "step": 4620 }, { "grad_norm": 1.005719780921936, "learning_rate": 9.998309989536652e-05, "loss": 0.1324, "step": 4630 }, { "grad_norm": 0.802505373954773, "learning_rate": 9.998255829106593e-05, "loss": 0.1326, "step": 4640 }, { "grad_norm": 1.2888091802597046, "learning_rate": 9.998200814611316e-05, "loss": 0.1816, "step": 4650 }, { "grad_norm": 1.153633952140808, "learning_rate": 9.998144946060219e-05, "loss": 0.1734, "step": 4660 }, { "grad_norm": 1.2500653266906738, "learning_rate": 9.998088223462852e-05, "loss": 0.1631, "step": 4670 }, { "grad_norm": 1.0315016508102417, "learning_rate": 9.998030646828905e-05, "loss": 0.1417, "step": 4680 }, { "grad_norm": 0.8797957897186279, "learning_rate": 9.997972216168217e-05, "loss": 0.1739, "step": 4690 }, { "grad_norm": 0.7511084079742432, "learning_rate": 9.997912931490771e-05, "loss": 0.1263, "step": 4700 }, { "grad_norm": 0.9544983506202698, "learning_rate": 9.9978527928067e-05, "loss": 0.1427, "step": 4710 }, { "grad_norm": 1.3710229396820068, "learning_rate": 9.997791800126277e-05, "loss": 0.1551, "step": 4720 }, { "grad_norm": 0.7991692423820496, "learning_rate": 9.997729953459927e-05, "loss": 0.1636, "step": 4730 }, { "grad_norm": 1.1116526126861572, "learning_rate": 9.997667252818214e-05, "loss": 0.1403, "step": 4740 }, { "grad_norm": 0.8908485770225525, "learning_rate": 9.997603698211855e-05, "loss": 0.1459, "step": 4750 }, { "grad_norm": 0.833859920501709, "learning_rate": 9.99753928965171e-05, "loss": 0.1593, "step": 4760 }, { "grad_norm": 1.0384949445724487, "learning_rate": 9.997474027148781e-05, "loss": 0.1396, "step": 4770 }, { "grad_norm": 0.7987760305404663, "learning_rate": 9.997407910714223e-05, "loss": 0.1483, "step": 4780 }, { "grad_norm": 1.155945897102356, "learning_rate": 9.997340940359332e-05, "loss": 0.1589, "step": 4790 }, { "grad_norm": 0.841317892074585, "learning_rate": 9.997273116095552e-05, "loss": 0.1329, "step": 4800 }, { "grad_norm": 1.0301356315612793, "learning_rate": 9.997204437934473e-05, "loss": 0.1416, "step": 4810 }, { "grad_norm": 0.8531216979026794, "learning_rate": 9.997134905887829e-05, "loss": 0.1345, "step": 4820 }, { "grad_norm": 0.7975953817367554, "learning_rate": 9.997064519967501e-05, "loss": 0.1844, "step": 4830 }, { "grad_norm": 0.7506058812141418, "learning_rate": 9.996993280185517e-05, "loss": 0.1461, "step": 4840 }, { "grad_norm": 1.4432958364486694, "learning_rate": 9.99692118655405e-05, "loss": 0.1571, "step": 4850 }, { "grad_norm": 1.1131513118743896, "learning_rate": 9.996848239085417e-05, "loss": 0.1474, "step": 4860 }, { "grad_norm": 1.2007877826690674, "learning_rate": 9.996774437792085e-05, "loss": 0.1496, "step": 4870 }, { "grad_norm": 0.8291547894477844, "learning_rate": 9.996699782686664e-05, "loss": 0.138, "step": 4880 }, { "grad_norm": 0.8840456008911133, "learning_rate": 9.996624273781909e-05, "loss": 0.1322, "step": 4890 }, { "grad_norm": 0.9060882329940796, "learning_rate": 9.996547911090725e-05, "loss": 0.1222, "step": 4900 }, { "grad_norm": 0.9155063033103943, "learning_rate": 9.996470694626157e-05, "loss": 0.1209, "step": 4910 }, { "grad_norm": 0.8165646195411682, "learning_rate": 9.996392624401403e-05, "loss": 0.1034, "step": 4920 }, { "grad_norm": 0.9330236315727234, "learning_rate": 9.996313700429801e-05, "loss": 0.1311, "step": 4930 }, { "grad_norm": 1.048215389251709, "learning_rate": 9.996233922724836e-05, "loss": 0.118, "step": 4940 }, { "grad_norm": 0.8646103143692017, "learning_rate": 9.996153291300141e-05, "loss": 0.1375, "step": 4950 }, { "grad_norm": 0.9851760268211365, "learning_rate": 9.996071806169494e-05, "loss": 0.1481, "step": 4960 }, { "grad_norm": 1.021305799484253, "learning_rate": 9.995989467346817e-05, "loss": 0.1058, "step": 4970 }, { "grad_norm": 0.7294428944587708, "learning_rate": 9.995906274846183e-05, "loss": 0.1229, "step": 4980 }, { "grad_norm": 0.9191893339157104, "learning_rate": 9.995822228681803e-05, "loss": 0.1197, "step": 4990 }, { "grad_norm": 0.8386814594268799, "learning_rate": 9.99573732886804e-05, "loss": 0.1151, "step": 5000 }, { "grad_norm": 0.8202689290046692, "learning_rate": 9.995651575419402e-05, "loss": 0.1044, "step": 5010 }, { "grad_norm": 0.9217038154602051, "learning_rate": 9.995564968350541e-05, "loss": 0.0971, "step": 5020 }, { "grad_norm": 0.7822023630142212, "learning_rate": 9.995477507676256e-05, "loss": 0.1315, "step": 5030 }, { "grad_norm": 0.7836185693740845, "learning_rate": 9.995389193411493e-05, "loss": 0.117, "step": 5040 }, { "grad_norm": 0.8877218961715698, "learning_rate": 9.995300025571339e-05, "loss": 0.1196, "step": 5050 }, { "grad_norm": 0.9763333201408386, "learning_rate": 9.995210004171034e-05, "loss": 0.1166, "step": 5060 }, { "grad_norm": 0.86274254322052, "learning_rate": 9.995119129225956e-05, "loss": 0.1267, "step": 5070 }, { "grad_norm": 0.6666380167007446, "learning_rate": 9.995027400751637e-05, "loss": 0.1012, "step": 5080 }, { "grad_norm": 0.83863365650177, "learning_rate": 9.994934818763751e-05, "loss": 0.1551, "step": 5090 }, { "grad_norm": 0.7664852738380432, "learning_rate": 9.994841383278115e-05, "loss": 0.1072, "step": 5100 }, { "grad_norm": 0.8742172718048096, "learning_rate": 9.994747094310695e-05, "loss": 0.1499, "step": 5110 }, { "grad_norm": 1.108948826789856, "learning_rate": 9.994651951877604e-05, "loss": 0.1273, "step": 5120 }, { "grad_norm": 0.9516023397445679, "learning_rate": 9.994555955995099e-05, "loss": 0.1249, "step": 5130 }, { "grad_norm": 0.8547828793525696, "learning_rate": 9.994459106679581e-05, "loss": 0.1366, "step": 5140 }, { "grad_norm": 0.8744113445281982, "learning_rate": 9.994361403947603e-05, "loss": 0.1489, "step": 5150 }, { "grad_norm": 0.836616575717926, "learning_rate": 9.994262847815854e-05, "loss": 0.1222, "step": 5160 }, { "grad_norm": 1.5337283611297607, "learning_rate": 9.99416343830118e-05, "loss": 0.1655, "step": 5170 }, { "grad_norm": 0.7944661378860474, "learning_rate": 9.994063175420565e-05, "loss": 0.1128, "step": 5180 }, { "grad_norm": 0.9599660634994507, "learning_rate": 9.99396205919114e-05, "loss": 0.1556, "step": 5190 }, { "grad_norm": 0.8506893515586853, "learning_rate": 9.993860089630185e-05, "loss": 0.1214, "step": 5200 }, { "grad_norm": 1.0653305053710938, "learning_rate": 9.993757266755123e-05, "loss": 0.128, "step": 5210 }, { "grad_norm": 1.1201932430267334, "learning_rate": 9.993653590583522e-05, "loss": 0.1552, "step": 5220 }, { "grad_norm": 0.7808775305747986, "learning_rate": 9.993549061133102e-05, "loss": 0.1189, "step": 5230 }, { "grad_norm": 0.7363652586936951, "learning_rate": 9.993443678421719e-05, "loss": 0.0977, "step": 5240 }, { "grad_norm": 0.7646018862724304, "learning_rate": 9.993337442467384e-05, "loss": 0.1107, "step": 5250 }, { "grad_norm": 1.3279634714126587, "learning_rate": 9.993230353288248e-05, "loss": 0.1472, "step": 5260 }, { "grad_norm": 0.9952992796897888, "learning_rate": 9.993122410902608e-05, "loss": 0.1239, "step": 5270 }, { "grad_norm": 0.9590794444084167, "learning_rate": 9.993013615328912e-05, "loss": 0.1528, "step": 5280 }, { "grad_norm": 1.0887807607650757, "learning_rate": 9.992903966585747e-05, "loss": 0.1361, "step": 5290 }, { "grad_norm": 0.855718731880188, "learning_rate": 9.992793464691852e-05, "loss": 0.1244, "step": 5300 }, { "grad_norm": 0.7498486042022705, "learning_rate": 9.992682109666105e-05, "loss": 0.1465, "step": 5310 }, { "grad_norm": 0.7074082493782043, "learning_rate": 9.992569901527538e-05, "loss": 0.1404, "step": 5320 }, { "grad_norm": 1.065004825592041, "learning_rate": 9.99245684029532e-05, "loss": 0.1455, "step": 5330 }, { "grad_norm": 0.8893193602561951, "learning_rate": 9.992342925988774e-05, "loss": 0.1627, "step": 5340 }, { "grad_norm": 0.8236026763916016, "learning_rate": 9.992228158627361e-05, "loss": 0.1719, "step": 5350 }, { "grad_norm": 0.7739737033843994, "learning_rate": 9.992112538230693e-05, "loss": 0.1642, "step": 5360 }, { "grad_norm": 0.7970970273017883, "learning_rate": 9.991996064818527e-05, "loss": 0.1421, "step": 5370 }, { "grad_norm": 0.6811489462852478, "learning_rate": 9.991878738410768e-05, "loss": 0.1249, "step": 5380 }, { "grad_norm": 0.7706612944602966, "learning_rate": 9.991760559027457e-05, "loss": 0.1293, "step": 5390 }, { "grad_norm": 1.0343260765075684, "learning_rate": 9.991641526688793e-05, "loss": 0.1253, "step": 5400 }, { "grad_norm": 1.0623701810836792, "learning_rate": 9.991521641415113e-05, "loss": 0.1248, "step": 5410 }, { "grad_norm": 0.7375714182853699, "learning_rate": 9.991400903226904e-05, "loss": 0.1407, "step": 5420 }, { "grad_norm": 0.8264533877372742, "learning_rate": 9.991279312144794e-05, "loss": 0.1265, "step": 5430 }, { "grad_norm": 0.9134120941162109, "learning_rate": 9.991156868189564e-05, "loss": 0.1673, "step": 5440 }, { "grad_norm": 0.8146782517433167, "learning_rate": 9.991033571382131e-05, "loss": 0.1433, "step": 5450 }, { "grad_norm": 1.1135351657867432, "learning_rate": 9.990909421743569e-05, "loss": 0.1164, "step": 5460 }, { "grad_norm": 0.9828506708145142, "learning_rate": 9.990784419295085e-05, "loss": 0.1197, "step": 5470 }, { "grad_norm": 0.8659884929656982, "learning_rate": 9.990658564058044e-05, "loss": 0.1388, "step": 5480 }, { "grad_norm": 0.9322956204414368, "learning_rate": 9.990531856053948e-05, "loss": 0.1317, "step": 5490 }, { "grad_norm": 1.1889233589172363, "learning_rate": 9.99040429530445e-05, "loss": 0.1546, "step": 5500 }, { "grad_norm": 0.6421924829483032, "learning_rate": 9.990275881831346e-05, "loss": 0.1071, "step": 5510 }, { "grad_norm": 0.7866718769073486, "learning_rate": 9.990146615656577e-05, "loss": 0.1306, "step": 5520 }, { "grad_norm": 0.8308515548706055, "learning_rate": 9.990016496802233e-05, "loss": 0.1154, "step": 5530 }, { "grad_norm": 0.9110240340232849, "learning_rate": 9.989885525290548e-05, "loss": 0.143, "step": 5540 }, { "grad_norm": 0.8795160055160522, "learning_rate": 9.989753701143897e-05, "loss": 0.1227, "step": 5550 }, { "grad_norm": 0.7415983080863953, "learning_rate": 9.989621024384812e-05, "loss": 0.1258, "step": 5560 }, { "grad_norm": 0.951055109500885, "learning_rate": 9.989487495035959e-05, "loss": 0.1283, "step": 5570 }, { "grad_norm": 0.8232394456863403, "learning_rate": 9.989353113120156e-05, "loss": 0.122, "step": 5580 }, { "grad_norm": 0.8520922660827637, "learning_rate": 9.989217878660366e-05, "loss": 0.1152, "step": 5590 }, { "grad_norm": 0.9450141787528992, "learning_rate": 9.989081791679695e-05, "loss": 0.1238, "step": 5600 }, { "grad_norm": 1.0242787599563599, "learning_rate": 9.988944852201397e-05, "loss": 0.1341, "step": 5610 }, { "grad_norm": 0.8783028721809387, "learning_rate": 9.988807060248873e-05, "loss": 0.1981, "step": 5620 }, { "grad_norm": 0.9969605207443237, "learning_rate": 9.988668415845665e-05, "loss": 0.1268, "step": 5630 }, { "grad_norm": 0.9255369901657104, "learning_rate": 9.988528919015466e-05, "loss": 0.1409, "step": 5640 }, { "grad_norm": 1.0274205207824707, "learning_rate": 9.988388569782112e-05, "loss": 0.1126, "step": 5650 }, { "grad_norm": 1.1021430492401123, "learning_rate": 9.988247368169583e-05, "loss": 0.126, "step": 5660 }, { "grad_norm": 1.0439176559448242, "learning_rate": 9.988105314202007e-05, "loss": 0.1438, "step": 5670 }, { "grad_norm": 0.6601024866104126, "learning_rate": 9.987962407903659e-05, "loss": 0.1456, "step": 5680 }, { "grad_norm": 0.9163444638252258, "learning_rate": 9.987818649298957e-05, "loss": 0.1388, "step": 5690 }, { "grad_norm": 0.9859479665756226, "learning_rate": 9.987674038412465e-05, "loss": 0.1269, "step": 5700 }, { "grad_norm": 0.8820024728775024, "learning_rate": 9.987528575268891e-05, "loss": 0.1294, "step": 5710 }, { "grad_norm": 1.0348855257034302, "learning_rate": 9.987382259893095e-05, "loss": 0.1384, "step": 5720 }, { "grad_norm": 0.932634711265564, "learning_rate": 9.987235092310074e-05, "loss": 0.1316, "step": 5730 }, { "grad_norm": 0.8185351490974426, "learning_rate": 9.987087072544978e-05, "loss": 0.1108, "step": 5740 }, { "grad_norm": 0.9516857266426086, "learning_rate": 9.9869382006231e-05, "loss": 0.1149, "step": 5750 }, { "grad_norm": 0.882323682308197, "learning_rate": 9.986788476569875e-05, "loss": 0.1074, "step": 5760 }, { "grad_norm": 0.770130455493927, "learning_rate": 9.986637900410887e-05, "loss": 0.1239, "step": 5770 }, { "grad_norm": 0.6937487721443176, "learning_rate": 9.986486472171869e-05, "loss": 0.1177, "step": 5780 }, { "grad_norm": 1.1810415983200073, "learning_rate": 9.986334191878692e-05, "loss": 0.1106, "step": 5790 }, { "grad_norm": 0.9455435872077942, "learning_rate": 9.986181059557378e-05, "loss": 0.1637, "step": 5800 }, { "grad_norm": 0.7010846734046936, "learning_rate": 9.986027075234094e-05, "loss": 0.1246, "step": 5810 }, { "grad_norm": 1.1263712644577026, "learning_rate": 9.985872238935152e-05, "loss": 0.1219, "step": 5820 }, { "grad_norm": 0.7397040724754333, "learning_rate": 9.985716550687008e-05, "loss": 0.1323, "step": 5830 }, { "grad_norm": 0.9382083415985107, "learning_rate": 9.985560010516264e-05, "loss": 0.1288, "step": 5840 }, { "grad_norm": 0.8046036958694458, "learning_rate": 9.985402618449668e-05, "loss": 0.1045, "step": 5850 }, { "grad_norm": 0.830643892288208, "learning_rate": 9.985244374514118e-05, "loss": 0.119, "step": 5860 }, { "grad_norm": 1.2969461679458618, "learning_rate": 9.985085278736651e-05, "loss": 0.1363, "step": 5870 }, { "grad_norm": 0.9422857165336609, "learning_rate": 9.984925331144452e-05, "loss": 0.1255, "step": 5880 }, { "grad_norm": 1.3058439493179321, "learning_rate": 9.984764531764851e-05, "loss": 0.1097, "step": 5890 }, { "grad_norm": 0.7233833074569702, "learning_rate": 9.984602880625326e-05, "loss": 0.1114, "step": 5900 }, { "grad_norm": 0.8567097783088684, "learning_rate": 9.9844403777535e-05, "loss": 0.1348, "step": 5910 }, { "grad_norm": 0.7800244688987732, "learning_rate": 9.984277023177135e-05, "loss": 0.1005, "step": 5920 }, { "grad_norm": 0.9684569835662842, "learning_rate": 9.984112816924148e-05, "loss": 0.1014, "step": 5930 }, { "grad_norm": 1.1680833101272583, "learning_rate": 9.983947759022596e-05, "loss": 0.1376, "step": 5940 }, { "grad_norm": 0.7962195873260498, "learning_rate": 9.983781849500682e-05, "loss": 0.1233, "step": 5950 }, { "grad_norm": 1.0889116525650024, "learning_rate": 9.98361508838676e-05, "loss": 0.1536, "step": 5960 }, { "grad_norm": 1.0807454586029053, "learning_rate": 9.98344747570932e-05, "loss": 0.1043, "step": 5970 }, { "grad_norm": 1.1222904920578003, "learning_rate": 9.983279011497004e-05, "loss": 0.1498, "step": 5980 }, { "grad_norm": 1.253632664680481, "learning_rate": 9.983109695778596e-05, "loss": 0.1493, "step": 5990 }, { "grad_norm": 0.6595586538314819, "learning_rate": 9.982939528583032e-05, "loss": 0.1141, "step": 6000 }, { "grad_norm": 0.810352623462677, "learning_rate": 9.982768509939385e-05, "loss": 0.1071, "step": 6010 }, { "grad_norm": 0.9264913201332092, "learning_rate": 9.982596639876879e-05, "loss": 0.098, "step": 6020 }, { "grad_norm": 0.76678466796875, "learning_rate": 9.982423918424881e-05, "loss": 0.1236, "step": 6030 }, { "grad_norm": 0.6576812863349915, "learning_rate": 9.982250345612908e-05, "loss": 0.098, "step": 6040 }, { "grad_norm": 1.0613311529159546, "learning_rate": 9.982075921470611e-05, "loss": 0.1116, "step": 6050 }, { "grad_norm": 0.8158771991729736, "learning_rate": 9.981900646027802e-05, "loss": 0.1148, "step": 6060 }, { "grad_norm": 0.8875443339347839, "learning_rate": 9.981724519314425e-05, "loss": 0.1147, "step": 6070 }, { "grad_norm": 0.9362044334411621, "learning_rate": 9.981547541360581e-05, "loss": 0.1391, "step": 6080 }, { "grad_norm": 0.7608993649482727, "learning_rate": 9.981369712196508e-05, "loss": 0.1079, "step": 6090 }, { "grad_norm": 1.1125072240829468, "learning_rate": 9.981191031852592e-05, "loss": 0.1145, "step": 6100 }, { "grad_norm": 0.7686514854431152, "learning_rate": 9.981011500359362e-05, "loss": 0.1296, "step": 6110 }, { "grad_norm": 0.7016928791999817, "learning_rate": 9.9808311177475e-05, "loss": 0.1117, "step": 6120 }, { "grad_norm": 0.6476864218711853, "learning_rate": 9.980649884047826e-05, "loss": 0.1055, "step": 6130 }, { "grad_norm": 1.10622239112854, "learning_rate": 9.980467799291307e-05, "loss": 0.1063, "step": 6140 }, { "grad_norm": 0.912352442741394, "learning_rate": 9.980284863509058e-05, "loss": 0.116, "step": 6150 }, { "grad_norm": 0.7677583694458008, "learning_rate": 9.980101076732334e-05, "loss": 0.142, "step": 6160 }, { "grad_norm": 0.9247700572013855, "learning_rate": 9.979916438992544e-05, "loss": 0.1265, "step": 6170 }, { "grad_norm": 0.8300501704216003, "learning_rate": 9.979730950321237e-05, "loss": 0.1649, "step": 6180 }, { "grad_norm": 0.6083024144172668, "learning_rate": 9.979544610750104e-05, "loss": 0.1155, "step": 6190 }, { "grad_norm": 0.7055187821388245, "learning_rate": 9.97935742031099e-05, "loss": 0.1082, "step": 6200 }, { "grad_norm": 0.8078483939170837, "learning_rate": 9.979169379035878e-05, "loss": 0.1173, "step": 6210 }, { "grad_norm": 0.8004080653190613, "learning_rate": 9.978980486956899e-05, "loss": 0.1082, "step": 6220 }, { "grad_norm": 1.0123612880706787, "learning_rate": 9.978790744106332e-05, "loss": 0.1239, "step": 6230 }, { "grad_norm": 0.7738893032073975, "learning_rate": 9.978600150516594e-05, "loss": 0.091, "step": 6240 }, { "grad_norm": 1.5990909337997437, "learning_rate": 9.978408706220259e-05, "loss": 0.1134, "step": 6250 }, { "grad_norm": 0.8633920550346375, "learning_rate": 9.978216411250032e-05, "loss": 0.1124, "step": 6260 }, { "grad_norm": 0.7525709271430969, "learning_rate": 9.978023265638778e-05, "loss": 0.1046, "step": 6270 }, { "grad_norm": 0.8637456297874451, "learning_rate": 9.977829269419495e-05, "loss": 0.1179, "step": 6280 }, { "grad_norm": 0.9911900758743286, "learning_rate": 9.977634422625335e-05, "loss": 0.1168, "step": 6290 }, { "grad_norm": 0.7146417498588562, "learning_rate": 9.97743872528959e-05, "loss": 0.1148, "step": 6300 }, { "grad_norm": 0.6926625967025757, "learning_rate": 9.9772421774457e-05, "loss": 0.1075, "step": 6310 }, { "grad_norm": 0.7240981459617615, "learning_rate": 9.977044779127252e-05, "loss": 0.1375, "step": 6320 }, { "grad_norm": 0.6746109127998352, "learning_rate": 9.976846530367971e-05, "loss": 0.1095, "step": 6330 }, { "grad_norm": 0.8054181933403015, "learning_rate": 9.976647431201735e-05, "loss": 0.1181, "step": 6340 }, { "grad_norm": 0.8811993598937988, "learning_rate": 9.976447481662568e-05, "loss": 0.1329, "step": 6350 }, { "grad_norm": 0.8074590563774109, "learning_rate": 9.976246681784629e-05, "loss": 0.1188, "step": 6360 }, { "grad_norm": 0.7757161259651184, "learning_rate": 9.976045031602234e-05, "loss": 0.1229, "step": 6370 }, { "grad_norm": 0.6142646074295044, "learning_rate": 9.975842531149837e-05, "loss": 0.1163, "step": 6380 }, { "grad_norm": 0.7247372269630432, "learning_rate": 9.975639180462043e-05, "loss": 0.0983, "step": 6390 }, { "grad_norm": 0.8098075985908508, "learning_rate": 9.975434979573596e-05, "loss": 0.1295, "step": 6400 }, { "grad_norm": 0.7826366424560547, "learning_rate": 9.97522992851939e-05, "loss": 0.1077, "step": 6410 }, { "grad_norm": 0.8838618397712708, "learning_rate": 9.975024027334461e-05, "loss": 0.1045, "step": 6420 }, { "grad_norm": 1.2137314081192017, "learning_rate": 9.974817276053993e-05, "loss": 0.1265, "step": 6430 }, { "grad_norm": 0.8405417203903198, "learning_rate": 9.974609674713315e-05, "loss": 0.1544, "step": 6440 }, { "grad_norm": 0.996895968914032, "learning_rate": 9.9744012233479e-05, "loss": 0.1366, "step": 6450 }, { "grad_norm": 0.8866038918495178, "learning_rate": 9.974191921993366e-05, "loss": 0.1142, "step": 6460 }, { "grad_norm": 0.7615402340888977, "learning_rate": 9.973981770685474e-05, "loss": 0.1145, "step": 6470 }, { "grad_norm": 0.8584765791893005, "learning_rate": 9.97377076946014e-05, "loss": 0.0891, "step": 6480 }, { "grad_norm": 0.8230586051940918, "learning_rate": 9.973558918353412e-05, "loss": 0.1062, "step": 6490 }, { "grad_norm": 0.7276773452758789, "learning_rate": 9.973346217401494e-05, "loss": 0.1432, "step": 6500 }, { "grad_norm": 1.1038553714752197, "learning_rate": 9.973132666640726e-05, "loss": 0.1216, "step": 6510 }, { "grad_norm": 0.5747677087783813, "learning_rate": 9.972918266107602e-05, "loss": 0.0988, "step": 6520 }, { "grad_norm": 0.885498046875, "learning_rate": 9.972703015838756e-05, "loss": 0.1145, "step": 6530 }, { "grad_norm": 0.8599845170974731, "learning_rate": 9.97248691587097e-05, "loss": 0.1219, "step": 6540 }, { "grad_norm": 1.0393537282943726, "learning_rate": 9.972269966241166e-05, "loss": 0.125, "step": 6550 }, { "grad_norm": 0.7361873388290405, "learning_rate": 9.972052166986417e-05, "loss": 0.1138, "step": 6560 }, { "grad_norm": 0.8738608360290527, "learning_rate": 9.971833518143938e-05, "loss": 0.1142, "step": 6570 }, { "grad_norm": 1.0855869054794312, "learning_rate": 9.971614019751093e-05, "loss": 0.1142, "step": 6580 }, { "grad_norm": 0.6838251352310181, "learning_rate": 9.971393671845383e-05, "loss": 0.1043, "step": 6590 }, { "grad_norm": 0.8201761245727539, "learning_rate": 9.971172474464464e-05, "loss": 0.1017, "step": 6600 }, { "grad_norm": 0.6259744167327881, "learning_rate": 9.97095042764613e-05, "loss": 0.1236, "step": 6610 }, { "grad_norm": 1.5338947772979736, "learning_rate": 9.970727531428324e-05, "loss": 0.1393, "step": 6620 }, { "grad_norm": 0.8114010095596313, "learning_rate": 9.970503785849132e-05, "loss": 0.1412, "step": 6630 }, { "grad_norm": 0.6135753393173218, "learning_rate": 9.970279190946788e-05, "loss": 0.1285, "step": 6640 }, { "grad_norm": 0.7453290224075317, "learning_rate": 9.970053746759667e-05, "loss": 0.0977, "step": 6650 }, { "grad_norm": 1.1737252473831177, "learning_rate": 9.969827453326292e-05, "loss": 0.1107, "step": 6660 }, { "grad_norm": 1.0376927852630615, "learning_rate": 9.969600310685332e-05, "loss": 0.1151, "step": 6670 }, { "grad_norm": 0.774487316608429, "learning_rate": 9.969372318875596e-05, "loss": 0.1203, "step": 6680 }, { "grad_norm": 0.753231406211853, "learning_rate": 9.969143477936043e-05, "loss": 0.1175, "step": 6690 }, { "grad_norm": 0.6993904709815979, "learning_rate": 9.968913787905775e-05, "loss": 0.1102, "step": 6700 }, { "grad_norm": 0.7620125412940979, "learning_rate": 9.968683248824045e-05, "loss": 0.102, "step": 6710 }, { "grad_norm": 0.7769166827201843, "learning_rate": 9.968451860730238e-05, "loss": 0.1072, "step": 6720 }, { "grad_norm": 0.8868599534034729, "learning_rate": 9.968219623663896e-05, "loss": 0.1012, "step": 6730 }, { "grad_norm": 1.1127817630767822, "learning_rate": 9.967986537664702e-05, "loss": 0.1155, "step": 6740 }, { "grad_norm": 1.3266777992248535, "learning_rate": 9.967752602772483e-05, "loss": 0.1446, "step": 6750 }, { "grad_norm": 0.7172905802726746, "learning_rate": 9.967517819027212e-05, "loss": 0.1618, "step": 6760 }, { "grad_norm": 0.7120369672775269, "learning_rate": 9.967282186469009e-05, "loss": 0.1078, "step": 6770 }, { "grad_norm": 0.6581645607948303, "learning_rate": 9.967045705138135e-05, "loss": 0.1164, "step": 6780 }, { "grad_norm": 0.9294963479042053, "learning_rate": 9.966808375074998e-05, "loss": 0.1258, "step": 6790 }, { "grad_norm": 0.7974227070808411, "learning_rate": 9.966570196320154e-05, "loss": 0.1115, "step": 6800 }, { "grad_norm": 0.7187811136245728, "learning_rate": 9.966331168914299e-05, "loss": 0.1152, "step": 6810 }, { "grad_norm": 0.9120466113090515, "learning_rate": 9.966091292898277e-05, "loss": 0.0988, "step": 6820 }, { "grad_norm": 0.9255893230438232, "learning_rate": 9.965850568313076e-05, "loss": 0.0958, "step": 6830 }, { "grad_norm": 0.7845327258110046, "learning_rate": 9.965608995199827e-05, "loss": 0.0981, "step": 6840 }, { "grad_norm": 0.7065679430961609, "learning_rate": 9.965366573599812e-05, "loss": 0.1013, "step": 6850 }, { "grad_norm": 0.665264904499054, "learning_rate": 9.965123303554453e-05, "loss": 0.1041, "step": 6860 }, { "grad_norm": 0.7344934344291687, "learning_rate": 9.964879185105317e-05, "loss": 0.1038, "step": 6870 }, { "grad_norm": 0.6537965536117554, "learning_rate": 9.964634218294119e-05, "loss": 0.1205, "step": 6880 }, { "grad_norm": 1.3955504894256592, "learning_rate": 9.964388403162714e-05, "loss": 0.1049, "step": 6890 }, { "grad_norm": 0.7237362861633301, "learning_rate": 9.96414173975311e-05, "loss": 0.108, "step": 6900 }, { "grad_norm": 0.7465169429779053, "learning_rate": 9.963894228107451e-05, "loss": 0.1011, "step": 6910 }, { "grad_norm": 0.7251818180084229, "learning_rate": 9.963645868268032e-05, "loss": 0.0938, "step": 6920 }, { "grad_norm": 0.8837078213691711, "learning_rate": 9.963396660277289e-05, "loss": 0.1315, "step": 6930 }, { "grad_norm": 0.8700870275497437, "learning_rate": 9.963146604177807e-05, "loss": 0.1028, "step": 6940 }, { "grad_norm": 0.821173369884491, "learning_rate": 9.962895700012311e-05, "loss": 0.1286, "step": 6950 }, { "grad_norm": 1.256909728050232, "learning_rate": 9.962643947823677e-05, "loss": 0.1151, "step": 6960 }, { "grad_norm": 0.7935903668403625, "learning_rate": 9.962391347654921e-05, "loss": 0.1262, "step": 6970 }, { "grad_norm": 0.9867850542068481, "learning_rate": 9.962137899549204e-05, "loss": 0.1266, "step": 6980 }, { "grad_norm": 0.8138741254806519, "learning_rate": 9.961883603549835e-05, "loss": 0.1097, "step": 6990 }, { "grad_norm": 0.7445108890533447, "learning_rate": 9.961628459700267e-05, "loss": 0.0941, "step": 7000 }, { "grad_norm": 0.6308836340904236, "learning_rate": 9.961372468044095e-05, "loss": 0.1016, "step": 7010 }, { "grad_norm": 0.7410215139389038, "learning_rate": 9.961115628625062e-05, "loss": 0.1013, "step": 7020 }, { "grad_norm": 0.7475342750549316, "learning_rate": 9.960857941487056e-05, "loss": 0.1028, "step": 7030 }, { "grad_norm": 0.7124561667442322, "learning_rate": 9.960599406674106e-05, "loss": 0.1072, "step": 7040 }, { "grad_norm": 0.7836161255836487, "learning_rate": 9.960340024230393e-05, "loss": 0.1117, "step": 7050 }, { "grad_norm": 0.8014628291130066, "learning_rate": 9.960079794200232e-05, "loss": 0.1017, "step": 7060 }, { "grad_norm": 0.6862634420394897, "learning_rate": 9.959818716628096e-05, "loss": 0.0988, "step": 7070 }, { "grad_norm": 0.7875009179115295, "learning_rate": 9.95955679155859e-05, "loss": 0.1196, "step": 7080 }, { "grad_norm": 1.014989972114563, "learning_rate": 9.959294019036472e-05, "loss": 0.0982, "step": 7090 }, { "grad_norm": 1.0152451992034912, "learning_rate": 9.959030399106646e-05, "loss": 0.1286, "step": 7100 }, { "grad_norm": 0.8593723773956299, "learning_rate": 9.958765931814153e-05, "loss": 0.1124, "step": 7110 }, { "grad_norm": 1.0139557123184204, "learning_rate": 9.958500617204184e-05, "loss": 0.0882, "step": 7120 }, { "grad_norm": 0.6508366465568542, "learning_rate": 9.958234455322075e-05, "loss": 0.115, "step": 7130 }, { "grad_norm": 0.7161981463432312, "learning_rate": 9.957967446213308e-05, "loss": 0.1341, "step": 7140 }, { "grad_norm": 0.8595768213272095, "learning_rate": 9.957699589923501e-05, "loss": 0.1393, "step": 7150 }, { "grad_norm": 0.7890865802764893, "learning_rate": 9.957430886498431e-05, "loss": 0.1147, "step": 7160 }, { "grad_norm": 1.1860958337783813, "learning_rate": 9.957161335984008e-05, "loss": 0.1384, "step": 7170 }, { "grad_norm": 0.9399498701095581, "learning_rate": 9.956890938426291e-05, "loss": 0.1155, "step": 7180 }, { "grad_norm": 0.6774275302886963, "learning_rate": 9.956619693871482e-05, "loss": 0.1037, "step": 7190 }, { "grad_norm": 0.8287143707275391, "learning_rate": 9.956347602365934e-05, "loss": 0.1243, "step": 7200 }, { "grad_norm": 0.8744297027587891, "learning_rate": 9.956074663956135e-05, "loss": 0.1002, "step": 7210 }, { "grad_norm": 1.023193359375, "learning_rate": 9.955800878688726e-05, "loss": 0.109, "step": 7220 }, { "grad_norm": 0.7338206171989441, "learning_rate": 9.955526246610489e-05, "loss": 0.0918, "step": 7230 }, { "grad_norm": 0.7712762355804443, "learning_rate": 9.955250767768349e-05, "loss": 0.0887, "step": 7240 }, { "grad_norm": 0.5705710649490356, "learning_rate": 9.95497444220938e-05, "loss": 0.1016, "step": 7250 }, { "grad_norm": 0.7000652551651001, "learning_rate": 9.954697269980797e-05, "loss": 0.092, "step": 7260 }, { "grad_norm": 0.6856899857521057, "learning_rate": 9.954419251129962e-05, "loss": 0.1147, "step": 7270 }, { "grad_norm": 0.9384810924530029, "learning_rate": 9.95414038570438e-05, "loss": 0.1004, "step": 7280 }, { "grad_norm": 0.5617523789405823, "learning_rate": 9.953860673751703e-05, "loss": 0.0906, "step": 7290 }, { "grad_norm": 0.9578689932823181, "learning_rate": 9.953580115319725e-05, "loss": 0.1061, "step": 7300 }, { "grad_norm": 0.7129742503166199, "learning_rate": 9.953298710456387e-05, "loss": 0.1259, "step": 7310 }, { "grad_norm": 0.7357267141342163, "learning_rate": 9.953016459209771e-05, "loss": 0.0936, "step": 7320 }, { "grad_norm": 0.6878396272659302, "learning_rate": 9.952733361628108e-05, "loss": 0.1242, "step": 7330 }, { "grad_norm": 1.0572378635406494, "learning_rate": 9.952449417759772e-05, "loss": 0.1126, "step": 7340 }, { "grad_norm": 0.6893327832221985, "learning_rate": 9.952164627653279e-05, "loss": 0.1145, "step": 7350 }, { "grad_norm": 0.7767282724380493, "learning_rate": 9.951878991357292e-05, "loss": 0.1191, "step": 7360 }, { "grad_norm": 1.208797574043274, "learning_rate": 9.951592508920622e-05, "loss": 0.0947, "step": 7370 }, { "grad_norm": 0.701353132724762, "learning_rate": 9.951305180392219e-05, "loss": 0.1118, "step": 7380 }, { "grad_norm": 0.7363319993019104, "learning_rate": 9.951017005821178e-05, "loss": 0.1245, "step": 7390 }, { "grad_norm": 0.5643223524093628, "learning_rate": 9.95072798525674e-05, "loss": 0.0978, "step": 7400 }, { "grad_norm": 0.8420168161392212, "learning_rate": 9.950438118748293e-05, "loss": 0.1099, "step": 7410 }, { "grad_norm": 0.8253527879714966, "learning_rate": 9.950147406345366e-05, "loss": 0.1047, "step": 7420 }, { "grad_norm": 0.8036839962005615, "learning_rate": 9.949855848097635e-05, "loss": 0.1153, "step": 7430 }, { "grad_norm": 0.7101341485977173, "learning_rate": 9.949563444054916e-05, "loss": 0.108, "step": 7440 }, { "grad_norm": 1.1065846681594849, "learning_rate": 9.949270194267178e-05, "loss": 0.1099, "step": 7450 }, { "grad_norm": 0.7688115239143372, "learning_rate": 9.948976098784526e-05, "loss": 0.0831, "step": 7460 }, { "grad_norm": 0.99811851978302, "learning_rate": 9.948681157657213e-05, "loss": 0.0942, "step": 7470 }, { "grad_norm": 0.6049559712409973, "learning_rate": 9.948385370935638e-05, "loss": 0.0848, "step": 7480 }, { "grad_norm": 0.6737817525863647, "learning_rate": 9.94808873867034e-05, "loss": 0.0973, "step": 7490 }, { "grad_norm": 0.6779716610908508, "learning_rate": 9.947791260912009e-05, "loss": 0.1158, "step": 7500 }, { "grad_norm": 0.819254994392395, "learning_rate": 9.947492937711474e-05, "loss": 0.1123, "step": 7510 }, { "grad_norm": 0.6397066116333008, "learning_rate": 9.947193769119707e-05, "loss": 0.08, "step": 7520 }, { "grad_norm": 0.7305687665939331, "learning_rate": 9.946893755187834e-05, "loss": 0.0902, "step": 7530 }, { "grad_norm": 0.9969472289085388, "learning_rate": 9.946592895967115e-05, "loss": 0.1229, "step": 7540 }, { "grad_norm": 0.7615848779678345, "learning_rate": 9.94629119150896e-05, "loss": 0.115, "step": 7550 }, { "grad_norm": 1.111546277999878, "learning_rate": 9.94598864186492e-05, "loss": 0.1006, "step": 7560 }, { "grad_norm": 0.9529621005058289, "learning_rate": 9.945685247086696e-05, "loss": 0.1097, "step": 7570 }, { "grad_norm": 0.6072466373443604, "learning_rate": 9.945381007226129e-05, "loss": 0.0767, "step": 7580 }, { "grad_norm": 0.7061586976051331, "learning_rate": 9.945075922335203e-05, "loss": 0.1269, "step": 7590 }, { "grad_norm": 0.6291716694831848, "learning_rate": 9.944769992466049e-05, "loss": 0.1158, "step": 7600 }, { "grad_norm": 0.7787336707115173, "learning_rate": 9.944463217670945e-05, "loss": 0.0957, "step": 7610 }, { "grad_norm": 0.6391757726669312, "learning_rate": 9.944155598002307e-05, "loss": 0.0764, "step": 7620 }, { "grad_norm": 0.6986787915229797, "learning_rate": 9.943847133512701e-05, "loss": 0.1039, "step": 7630 }, { "grad_norm": 1.0146549940109253, "learning_rate": 9.943537824254834e-05, "loss": 0.1052, "step": 7640 }, { "grad_norm": 0.8019434213638306, "learning_rate": 9.943227670281559e-05, "loss": 0.0937, "step": 7650 }, { "grad_norm": 0.7482237815856934, "learning_rate": 9.942916671645873e-05, "loss": 0.0887, "step": 7660 }, { "grad_norm": 0.7824860215187073, "learning_rate": 9.942604828400916e-05, "loss": 0.0814, "step": 7670 }, { "grad_norm": 0.773445188999176, "learning_rate": 9.942292140599975e-05, "loss": 0.0999, "step": 7680 }, { "grad_norm": 0.7170068025588989, "learning_rate": 9.94197860829648e-05, "loss": 0.1116, "step": 7690 }, { "grad_norm": 0.9642508625984192, "learning_rate": 9.941664231544004e-05, "loss": 0.1147, "step": 7700 }, { "grad_norm": 0.7127233147621155, "learning_rate": 9.941349010396264e-05, "loss": 0.0886, "step": 7710 }, { "grad_norm": 0.5265669822692871, "learning_rate": 9.941032944907125e-05, "loss": 0.0903, "step": 7720 }, { "grad_norm": 0.7604942321777344, "learning_rate": 9.940716035130596e-05, "loss": 0.0916, "step": 7730 }, { "grad_norm": 0.6837223172187805, "learning_rate": 9.940398281120821e-05, "loss": 0.088, "step": 7740 }, { "grad_norm": 0.6408495903015137, "learning_rate": 9.940079682932102e-05, "loss": 0.0848, "step": 7750 }, { "grad_norm": 0.6775439381599426, "learning_rate": 9.939760240618877e-05, "loss": 0.0789, "step": 7760 }, { "grad_norm": 0.6268342733383179, "learning_rate": 9.939439954235729e-05, "loss": 0.1081, "step": 7770 }, { "grad_norm": 0.6981508731842041, "learning_rate": 9.939118823837387e-05, "loss": 0.102, "step": 7780 }, { "grad_norm": 0.6043279767036438, "learning_rate": 9.938796849478725e-05, "loss": 0.0999, "step": 7790 }, { "grad_norm": 0.7542682886123657, "learning_rate": 9.938474031214755e-05, "loss": 0.1138, "step": 7800 }, { "grad_norm": 0.9983211755752563, "learning_rate": 9.938150369100643e-05, "loss": 0.1195, "step": 7810 }, { "grad_norm": 0.5696719288825989, "learning_rate": 9.93782586319169e-05, "loss": 0.1091, "step": 7820 }, { "grad_norm": 0.7054195404052734, "learning_rate": 9.937500513543348e-05, "loss": 0.1043, "step": 7830 }, { "grad_norm": 0.7730304002761841, "learning_rate": 9.937174320211207e-05, "loss": 0.11, "step": 7840 }, { "grad_norm": 0.6365925073623657, "learning_rate": 9.936847283251009e-05, "loss": 0.0924, "step": 7850 }, { "grad_norm": 0.8969640731811523, "learning_rate": 9.936519402718632e-05, "loss": 0.0845, "step": 7860 }, { "grad_norm": 1.0982089042663574, "learning_rate": 9.936190678670102e-05, "loss": 0.1128, "step": 7870 }, { "grad_norm": 0.7079673409461975, "learning_rate": 9.935861111161593e-05, "loss": 0.1144, "step": 7880 }, { "grad_norm": 0.6608090400695801, "learning_rate": 9.935530700249416e-05, "loss": 0.1254, "step": 7890 }, { "grad_norm": 0.6472672820091248, "learning_rate": 9.935199445990028e-05, "loss": 0.0879, "step": 7900 }, { "grad_norm": 0.7408196330070496, "learning_rate": 9.934867348440033e-05, "loss": 0.137, "step": 7910 }, { "grad_norm": 1.0766024589538574, "learning_rate": 9.934534407656176e-05, "loss": 0.0919, "step": 7920 }, { "grad_norm": 0.8330156207084656, "learning_rate": 9.93420062369535e-05, "loss": 0.0973, "step": 7930 }, { "grad_norm": 0.8514344096183777, "learning_rate": 9.933865996614589e-05, "loss": 0.0975, "step": 7940 }, { "grad_norm": 0.7301719784736633, "learning_rate": 9.933530526471068e-05, "loss": 0.123, "step": 7950 }, { "grad_norm": 0.7888876795768738, "learning_rate": 9.933194213322114e-05, "loss": 0.0939, "step": 7960 }, { "grad_norm": 0.7365117073059082, "learning_rate": 9.932857057225192e-05, "loss": 0.1056, "step": 7970 }, { "grad_norm": 1.1767672300338745, "learning_rate": 9.932519058237912e-05, "loss": 0.1048, "step": 7980 }, { "grad_norm": 1.4088003635406494, "learning_rate": 9.932180216418032e-05, "loss": 0.1352, "step": 7990 }, { "grad_norm": 0.9643948674201965, "learning_rate": 9.931840531823446e-05, "loss": 0.1018, "step": 8000 }, { "grad_norm": 0.6391529440879822, "learning_rate": 9.9315000045122e-05, "loss": 0.0743, "step": 8010 }, { "grad_norm": 0.8416100144386292, "learning_rate": 9.931158634542481e-05, "loss": 0.1155, "step": 8020 }, { "grad_norm": 1.0444942712783813, "learning_rate": 9.930816421972617e-05, "loss": 0.1025, "step": 8030 }, { "grad_norm": 0.6078028678894043, "learning_rate": 9.930473366861086e-05, "loss": 0.0953, "step": 8040 }, { "grad_norm": 0.6837689876556396, "learning_rate": 9.930129469266505e-05, "loss": 0.1227, "step": 8050 }, { "grad_norm": 0.8167172074317932, "learning_rate": 9.929784729247638e-05, "loss": 0.1297, "step": 8060 }, { "grad_norm": 0.8705393075942993, "learning_rate": 9.929439146863389e-05, "loss": 0.0928, "step": 8070 }, { "grad_norm": 0.7137790322303772, "learning_rate": 9.92909272217281e-05, "loss": 0.0895, "step": 8080 }, { "grad_norm": 0.5586733818054199, "learning_rate": 9.928745455235097e-05, "loss": 0.1142, "step": 8090 }, { "grad_norm": 1.0835925340652466, "learning_rate": 9.928397346109588e-05, "loss": 0.1128, "step": 8100 }, { "grad_norm": 0.6174253225326538, "learning_rate": 9.928048394855762e-05, "loss": 0.0956, "step": 8110 }, { "grad_norm": 0.9007217884063721, "learning_rate": 9.92769860153325e-05, "loss": 0.1053, "step": 8120 }, { "grad_norm": 0.592047929763794, "learning_rate": 9.927347966201819e-05, "loss": 0.0969, "step": 8130 }, { "grad_norm": 0.8758342266082764, "learning_rate": 9.926996488921383e-05, "loss": 0.0857, "step": 8140 }, { "grad_norm": 0.7759428024291992, "learning_rate": 9.926644169752001e-05, "loss": 0.1057, "step": 8150 }, { "grad_norm": 0.6389051675796509, "learning_rate": 9.926291008753875e-05, "loss": 0.1126, "step": 8160 }, { "grad_norm": 0.6763360500335693, "learning_rate": 9.92593700598735e-05, "loss": 0.1115, "step": 8170 }, { "grad_norm": 0.6885762214660645, "learning_rate": 9.925582161512915e-05, "loss": 0.101, "step": 8180 }, { "grad_norm": 0.7368013858795166, "learning_rate": 9.925226475391205e-05, "loss": 0.0879, "step": 8190 }, { "grad_norm": 0.6874315738677979, "learning_rate": 9.924869947682993e-05, "loss": 0.086, "step": 8200 }, { "grad_norm": 0.9073324203491211, "learning_rate": 9.924512578449204e-05, "loss": 0.0784, "step": 8210 }, { "grad_norm": 0.6629902720451355, "learning_rate": 9.924154367750901e-05, "loss": 0.1133, "step": 8220 }, { "grad_norm": 1.1312520503997803, "learning_rate": 9.923795315649293e-05, "loss": 0.1107, "step": 8230 }, { "grad_norm": 0.6981549859046936, "learning_rate": 9.92343542220573e-05, "loss": 0.0996, "step": 8240 }, { "grad_norm": 0.6657299995422363, "learning_rate": 9.92307468748171e-05, "loss": 0.0947, "step": 8250 }, { "grad_norm": 0.7647214531898499, "learning_rate": 9.922713111538873e-05, "loss": 0.1033, "step": 8260 }, { "grad_norm": 0.7697203755378723, "learning_rate": 9.922350694439003e-05, "loss": 0.0903, "step": 8270 }, { "grad_norm": 0.6522473096847534, "learning_rate": 9.921987436244024e-05, "loss": 0.0878, "step": 8280 }, { "grad_norm": 0.727695643901825, "learning_rate": 9.921623337016008e-05, "loss": 0.1053, "step": 8290 }, { "grad_norm": 0.770412027835846, "learning_rate": 9.921258396817172e-05, "loss": 0.1133, "step": 8300 }, { "grad_norm": 0.7488704323768616, "learning_rate": 9.920892615709874e-05, "loss": 0.0984, "step": 8310 }, { "grad_norm": 0.9381898641586304, "learning_rate": 9.920525993756612e-05, "loss": 0.0867, "step": 8320 }, { "grad_norm": 0.6742661595344543, "learning_rate": 9.920158531020036e-05, "loss": 0.0848, "step": 8330 }, { "grad_norm": 0.7703860998153687, "learning_rate": 9.919790227562933e-05, "loss": 0.0964, "step": 8340 }, { "grad_norm": 0.5086979866027832, "learning_rate": 9.919421083448237e-05, "loss": 0.08, "step": 8350 }, { "grad_norm": 0.6900819540023804, "learning_rate": 9.919051098739022e-05, "loss": 0.0838, "step": 8360 }, { "grad_norm": 0.7550523281097412, "learning_rate": 9.918680273498514e-05, "loss": 0.0906, "step": 8370 }, { "grad_norm": 0.6301543712615967, "learning_rate": 9.918308607790072e-05, "loss": 0.0923, "step": 8380 }, { "grad_norm": 0.7694907188415527, "learning_rate": 9.917936101677205e-05, "loss": 0.0878, "step": 8390 }, { "grad_norm": 0.7481673359870911, "learning_rate": 9.917562755223564e-05, "loss": 0.093, "step": 8400 }, { "grad_norm": 1.0336837768554688, "learning_rate": 9.917188568492944e-05, "loss": 0.1019, "step": 8410 }, { "grad_norm": 0.8113382458686829, "learning_rate": 9.916813541549283e-05, "loss": 0.1039, "step": 8420 }, { "grad_norm": 0.7963823676109314, "learning_rate": 9.916437674456663e-05, "loss": 0.0995, "step": 8430 }, { "grad_norm": 0.8127685785293579, "learning_rate": 9.916060967279308e-05, "loss": 0.1119, "step": 8440 }, { "grad_norm": 0.8741199374198914, "learning_rate": 9.91568342008159e-05, "loss": 0.1083, "step": 8450 }, { "grad_norm": 0.9192792773246765, "learning_rate": 9.915305032928019e-05, "loss": 0.1324, "step": 8460 }, { "grad_norm": 0.7139813899993896, "learning_rate": 9.914925805883253e-05, "loss": 0.105, "step": 8470 }, { "grad_norm": 1.0152907371520996, "learning_rate": 9.914545739012088e-05, "loss": 0.1144, "step": 8480 }, { "grad_norm": 0.7375652194023132, "learning_rate": 9.91416483237947e-05, "loss": 0.1199, "step": 8490 }, { "grad_norm": 0.75989830493927, "learning_rate": 9.913783086050485e-05, "loss": 0.1044, "step": 8500 }, { "grad_norm": 0.7298781275749207, "learning_rate": 9.913400500090364e-05, "loss": 0.0996, "step": 8510 }, { "grad_norm": 0.7127362489700317, "learning_rate": 9.913017074564479e-05, "loss": 0.0963, "step": 8520 }, { "grad_norm": 0.9116458296775818, "learning_rate": 9.912632809538348e-05, "loss": 0.104, "step": 8530 }, { "grad_norm": 0.8405850529670715, "learning_rate": 9.912247705077629e-05, "loss": 0.1119, "step": 8540 }, { "grad_norm": 0.5536789298057556, "learning_rate": 9.911861761248127e-05, "loss": 0.0823, "step": 8550 }, { "grad_norm": 0.8474905490875244, "learning_rate": 9.91147497811579e-05, "loss": 0.0896, "step": 8560 }, { "grad_norm": 0.5413193106651306, "learning_rate": 9.911087355746709e-05, "loss": 0.0838, "step": 8570 }, { "grad_norm": 0.9396812915802002, "learning_rate": 9.910698894207117e-05, "loss": 0.097, "step": 8580 }, { "grad_norm": 0.7692996263504028, "learning_rate": 9.910309593563392e-05, "loss": 0.0925, "step": 8590 }, { "grad_norm": 0.8004506826400757, "learning_rate": 9.909919453882057e-05, "loss": 0.0807, "step": 8600 }, { "grad_norm": 0.8942788243293762, "learning_rate": 9.90952847522977e-05, "loss": 0.1033, "step": 8610 }, { "grad_norm": 0.6693552732467651, "learning_rate": 9.909136657673346e-05, "loss": 0.0983, "step": 8620 }, { "grad_norm": 0.9465764164924622, "learning_rate": 9.908744001279731e-05, "loss": 0.1765, "step": 8630 }, { "grad_norm": 1.0752190351486206, "learning_rate": 9.90835050611602e-05, "loss": 0.1035, "step": 8640 }, { "grad_norm": 0.8245248198509216, "learning_rate": 9.90795617224945e-05, "loss": 0.081, "step": 8650 }, { "grad_norm": 0.48409295082092285, "learning_rate": 9.907560999747405e-05, "loss": 0.1009, "step": 8660 }, { "grad_norm": 0.5335037708282471, "learning_rate": 9.907164988677408e-05, "loss": 0.0857, "step": 8670 }, { "grad_norm": 0.6860232353210449, "learning_rate": 9.906768139107124e-05, "loss": 0.0901, "step": 8680 }, { "grad_norm": 0.6556084156036377, "learning_rate": 9.906370451104367e-05, "loss": 0.086, "step": 8690 }, { "grad_norm": 0.9430599212646484, "learning_rate": 9.905971924737088e-05, "loss": 0.0804, "step": 8700 }, { "grad_norm": 0.5941507816314697, "learning_rate": 9.905572560073387e-05, "loss": 0.0956, "step": 8710 }, { "grad_norm": 0.8090370893478394, "learning_rate": 9.905172357181501e-05, "loss": 0.0736, "step": 8720 }, { "grad_norm": 0.737673282623291, "learning_rate": 9.904771316129817e-05, "loss": 0.0831, "step": 8730 }, { "grad_norm": 0.8645185232162476, "learning_rate": 9.904369436986862e-05, "loss": 0.102, "step": 8740 }, { "grad_norm": 0.7550544738769531, "learning_rate": 9.903966719821303e-05, "loss": 0.0806, "step": 8750 }, { "grad_norm": 0.597295343875885, "learning_rate": 9.903563164701956e-05, "loss": 0.0978, "step": 8760 }, { "grad_norm": 0.8297408819198608, "learning_rate": 9.903158771697778e-05, "loss": 0.0885, "step": 8770 }, { "grad_norm": 0.5743157863616943, "learning_rate": 9.902753540877867e-05, "loss": 0.0712, "step": 8780 }, { "grad_norm": 0.6694906949996948, "learning_rate": 9.902347472311466e-05, "loss": 0.1035, "step": 8790 }, { "grad_norm": 0.5136216878890991, "learning_rate": 9.901940566067962e-05, "loss": 0.0771, "step": 8800 }, { "grad_norm": 0.64524906873703, "learning_rate": 9.901532822216883e-05, "loss": 0.0966, "step": 8810 }, { "grad_norm": 0.8081063032150269, "learning_rate": 9.901124240827904e-05, "loss": 0.0997, "step": 8820 }, { "grad_norm": 0.5997750163078308, "learning_rate": 9.900714821970835e-05, "loss": 0.1117, "step": 8830 }, { "grad_norm": 0.6414975523948669, "learning_rate": 9.900304565715641e-05, "loss": 0.1076, "step": 8840 }, { "grad_norm": 1.0077311992645264, "learning_rate": 9.899893472132419e-05, "loss": 0.1133, "step": 8850 }, { "grad_norm": 0.7404696345329285, "learning_rate": 9.899481541291415e-05, "loss": 0.0858, "step": 8860 }, { "grad_norm": 0.7821084260940552, "learning_rate": 9.899068773263016e-05, "loss": 0.1096, "step": 8870 }, { "grad_norm": 0.9412806034088135, "learning_rate": 9.898655168117754e-05, "loss": 0.1114, "step": 8880 }, { "grad_norm": 0.9501787424087524, "learning_rate": 9.898240725926302e-05, "loss": 0.1203, "step": 8890 }, { "grad_norm": 0.7674968242645264, "learning_rate": 9.897825446759478e-05, "loss": 0.1297, "step": 8900 }, { "grad_norm": 0.5977182388305664, "learning_rate": 9.897409330688241e-05, "loss": 0.1098, "step": 8910 }, { "grad_norm": 0.5921242833137512, "learning_rate": 9.896992377783692e-05, "loss": 0.0782, "step": 8920 }, { "grad_norm": 0.6298778057098389, "learning_rate": 9.89657458811708e-05, "loss": 0.1075, "step": 8930 }, { "grad_norm": 0.5018866062164307, "learning_rate": 9.896155961759792e-05, "loss": 0.0759, "step": 8940 }, { "grad_norm": 0.9076637029647827, "learning_rate": 9.895736498783361e-05, "loss": 0.0905, "step": 8950 }, { "grad_norm": 0.7666672468185425, "learning_rate": 9.895316199259462e-05, "loss": 0.1047, "step": 8960 }, { "grad_norm": 0.6622706055641174, "learning_rate": 9.894895063259909e-05, "loss": 0.0936, "step": 8970 }, { "grad_norm": 0.8840289115905762, "learning_rate": 9.894473090856667e-05, "loss": 0.1096, "step": 8980 }, { "grad_norm": 0.848022997379303, "learning_rate": 9.894050282121839e-05, "loss": 0.0946, "step": 8990 }, { "grad_norm": 0.8205865025520325, "learning_rate": 9.893626637127668e-05, "loss": 0.0919, "step": 9000 }, { "grad_norm": 0.8004144430160522, "learning_rate": 9.893202155946546e-05, "loss": 0.1312, "step": 9010 }, { "grad_norm": 0.5940009355545044, "learning_rate": 9.892776838651006e-05, "loss": 0.1236, "step": 9020 }, { "grad_norm": 0.7674715518951416, "learning_rate": 9.892350685313722e-05, "loss": 0.0852, "step": 9030 }, { "grad_norm": 0.525499701499939, "learning_rate": 9.891923696007513e-05, "loss": 0.0942, "step": 9040 }, { "grad_norm": 0.7983026504516602, "learning_rate": 9.891495870805336e-05, "loss": 0.0933, "step": 9050 }, { "grad_norm": 0.6947644948959351, "learning_rate": 9.891067209780298e-05, "loss": 0.0978, "step": 9060 }, { "grad_norm": 0.7130889892578125, "learning_rate": 9.890637713005646e-05, "loss": 0.1155, "step": 9070 }, { "grad_norm": 1.1110644340515137, "learning_rate": 9.890207380554767e-05, "loss": 0.1012, "step": 9080 }, { "grad_norm": 0.6771550178527832, "learning_rate": 9.889776212501196e-05, "loss": 0.1274, "step": 9090 }, { "grad_norm": 0.6680861115455627, "learning_rate": 9.889344208918605e-05, "loss": 0.0909, "step": 9100 }, { "grad_norm": 0.734569787979126, "learning_rate": 9.888911369880812e-05, "loss": 0.1034, "step": 9110 }, { "grad_norm": 0.7226519584655762, "learning_rate": 9.888477695461777e-05, "loss": 0.1057, "step": 9120 }, { "grad_norm": 0.6582194566726685, "learning_rate": 9.888043185735607e-05, "loss": 0.0855, "step": 9130 }, { "grad_norm": 0.6821518540382385, "learning_rate": 9.887607840776542e-05, "loss": 0.105, "step": 9140 }, { "grad_norm": 0.6727258563041687, "learning_rate": 9.887171660658975e-05, "loss": 0.0986, "step": 9150 }, { "grad_norm": 0.7533292174339294, "learning_rate": 9.886734645457435e-05, "loss": 0.0985, "step": 9160 }, { "grad_norm": 0.7913766503334045, "learning_rate": 9.886296795246597e-05, "loss": 0.0915, "step": 9170 }, { "grad_norm": 0.8045887351036072, "learning_rate": 9.885858110101276e-05, "loss": 0.1197, "step": 9180 }, { "grad_norm": 0.8079291582107544, "learning_rate": 9.885418590096434e-05, "loss": 0.1194, "step": 9190 }, { "grad_norm": 0.7106922268867493, "learning_rate": 9.88497823530717e-05, "loss": 0.0886, "step": 9200 }, { "grad_norm": 0.7645319104194641, "learning_rate": 9.884537045808732e-05, "loss": 0.0988, "step": 9210 }, { "grad_norm": 0.9265239834785461, "learning_rate": 9.884095021676502e-05, "loss": 0.1014, "step": 9220 }, { "grad_norm": 0.8393725752830505, "learning_rate": 9.883652162986017e-05, "loss": 0.072, "step": 9230 }, { "grad_norm": 0.5898089408874512, "learning_rate": 9.883208469812943e-05, "loss": 0.0973, "step": 9240 }, { "grad_norm": 0.6945863366127014, "learning_rate": 9.882763942233098e-05, "loss": 0.0921, "step": 9250 }, { "grad_norm": 0.7651230692863464, "learning_rate": 9.882318580322441e-05, "loss": 0.1149, "step": 9260 }, { "grad_norm": 0.7192394733428955, "learning_rate": 9.881872384157067e-05, "loss": 0.0912, "step": 9270 }, { "grad_norm": 0.7364445328712463, "learning_rate": 9.881425353813225e-05, "loss": 0.0998, "step": 9280 }, { "grad_norm": 0.875902533531189, "learning_rate": 9.880977489367296e-05, "loss": 0.0973, "step": 9290 }, { "grad_norm": 0.8561074137687683, "learning_rate": 9.88052879089581e-05, "loss": 0.0831, "step": 9300 }, { "grad_norm": 0.5125272274017334, "learning_rate": 9.880079258475434e-05, "loss": 0.093, "step": 9310 }, { "grad_norm": 0.8611241579055786, "learning_rate": 9.879628892182985e-05, "loss": 0.098, "step": 9320 }, { "grad_norm": 0.5624474883079529, "learning_rate": 9.879177692095416e-05, "loss": 0.0947, "step": 9330 }, { "grad_norm": 1.3578287363052368, "learning_rate": 9.878725658289825e-05, "loss": 0.1038, "step": 9340 }, { "grad_norm": 0.6461309194564819, "learning_rate": 9.878272790843454e-05, "loss": 0.1052, "step": 9350 }, { "grad_norm": 0.7540466785430908, "learning_rate": 9.877819089833682e-05, "loss": 0.0933, "step": 9360 }, { "grad_norm": 0.7225292921066284, "learning_rate": 9.877364555338038e-05, "loss": 0.0964, "step": 9370 }, { "grad_norm": 0.619891881942749, "learning_rate": 9.876909187434186e-05, "loss": 0.0982, "step": 9380 }, { "grad_norm": 0.6901982426643372, "learning_rate": 9.876452986199939e-05, "loss": 0.0921, "step": 9390 }, { "grad_norm": 0.6367073655128479, "learning_rate": 9.875995951713248e-05, "loss": 0.0795, "step": 9400 }, { "grad_norm": 0.6851145029067993, "learning_rate": 9.875538084052207e-05, "loss": 0.0791, "step": 9410 }, { "grad_norm": 0.9175860285758972, "learning_rate": 9.875079383295053e-05, "loss": 0.0955, "step": 9420 }, { "grad_norm": 1.0815314054489136, "learning_rate": 9.874619849520167e-05, "loss": 0.0945, "step": 9430 }, { "grad_norm": 0.8292188048362732, "learning_rate": 9.874159482806069e-05, "loss": 0.0973, "step": 9440 }, { "grad_norm": 0.7016839385032654, "learning_rate": 9.873698283231426e-05, "loss": 0.0951, "step": 9450 }, { "grad_norm": 0.8508594036102295, "learning_rate": 9.87323625087504e-05, "loss": 0.1288, "step": 9460 }, { "grad_norm": 0.6824504733085632, "learning_rate": 9.872773385815863e-05, "loss": 0.1214, "step": 9470 }, { "grad_norm": 0.7380742430686951, "learning_rate": 9.872309688132986e-05, "loss": 0.0759, "step": 9480 }, { "grad_norm": 0.707511305809021, "learning_rate": 9.871845157905639e-05, "loss": 0.0739, "step": 9490 }, { "grad_norm": 0.8848876357078552, "learning_rate": 9.871379795213201e-05, "loss": 0.0921, "step": 9500 }, { "grad_norm": 0.6103163957595825, "learning_rate": 9.87091360013519e-05, "loss": 0.0793, "step": 9510 }, { "grad_norm": 0.9768627285957336, "learning_rate": 9.870446572751262e-05, "loss": 0.1223, "step": 9520 }, { "grad_norm": 0.5544093251228333, "learning_rate": 9.869978713141224e-05, "loss": 0.0835, "step": 9530 }, { "grad_norm": 0.8516059517860413, "learning_rate": 9.869510021385016e-05, "loss": 0.0811, "step": 9540 }, { "grad_norm": 0.6709721684455872, "learning_rate": 9.869040497562727e-05, "loss": 0.1184, "step": 9550 }, { "grad_norm": 1.0999774932861328, "learning_rate": 9.868570141754587e-05, "loss": 0.093, "step": 9560 }, { "grad_norm": 0.6385477185249329, "learning_rate": 9.868098954040965e-05, "loss": 0.0877, "step": 9570 }, { "grad_norm": 0.7550693154335022, "learning_rate": 9.867626934502374e-05, "loss": 0.0932, "step": 9580 }, { "grad_norm": 0.8801603317260742, "learning_rate": 9.86715408321947e-05, "loss": 0.0955, "step": 9590 }, { "grad_norm": 0.5227881669998169, "learning_rate": 9.86668040027305e-05, "loss": 0.0918, "step": 9600 }, { "grad_norm": 0.5952836275100708, "learning_rate": 9.866205885744053e-05, "loss": 0.0832, "step": 9610 }, { "grad_norm": 0.6427983045578003, "learning_rate": 9.865730539713563e-05, "loss": 0.0823, "step": 9620 }, { "grad_norm": 0.6607542037963867, "learning_rate": 9.8652543622628e-05, "loss": 0.0902, "step": 9630 }, { "grad_norm": 0.8905013203620911, "learning_rate": 9.864777353473132e-05, "loss": 0.1161, "step": 9640 }, { "grad_norm": 0.7888710498809814, "learning_rate": 9.864299513426068e-05, "loss": 0.0955, "step": 9650 }, { "grad_norm": 0.5642648339271545, "learning_rate": 9.863820842203254e-05, "loss": 0.1002, "step": 9660 }, { "grad_norm": 0.741557240486145, "learning_rate": 9.863341339886483e-05, "loss": 0.109, "step": 9670 }, { "grad_norm": 0.7492069005966187, "learning_rate": 9.86286100655769e-05, "loss": 0.0837, "step": 9680 }, { "grad_norm": 0.7735142707824707, "learning_rate": 9.862379842298953e-05, "loss": 0.09, "step": 9690 }, { "grad_norm": 0.8834574818611145, "learning_rate": 9.861897847192485e-05, "loss": 0.0828, "step": 9700 }, { "grad_norm": 0.7069301605224609, "learning_rate": 9.86141502132065e-05, "loss": 0.0986, "step": 9710 }, { "grad_norm": 0.8231694102287292, "learning_rate": 9.860931364765946e-05, "loss": 0.0833, "step": 9720 }, { "grad_norm": 1.153098225593567, "learning_rate": 9.860446877611021e-05, "loss": 0.1046, "step": 9730 }, { "grad_norm": 1.2448912858963013, "learning_rate": 9.859961559938655e-05, "loss": 0.0904, "step": 9740 }, { "grad_norm": 0.6442114114761353, "learning_rate": 9.85947541183178e-05, "loss": 0.0918, "step": 9750 }, { "grad_norm": 0.747956395149231, "learning_rate": 9.858988433373463e-05, "loss": 0.0997, "step": 9760 }, { "grad_norm": 0.6450518369674683, "learning_rate": 9.858500624646918e-05, "loss": 0.1055, "step": 9770 }, { "grad_norm": 0.7873722314834595, "learning_rate": 9.858011985735497e-05, "loss": 0.1104, "step": 9780 }, { "grad_norm": 0.6798747777938843, "learning_rate": 9.857522516722693e-05, "loss": 0.1006, "step": 9790 }, { "grad_norm": 0.6361938118934631, "learning_rate": 9.857032217692145e-05, "loss": 0.0828, "step": 9800 }, { "grad_norm": 0.5791428089141846, "learning_rate": 9.856541088727631e-05, "loss": 0.0826, "step": 9810 }, { "grad_norm": 0.5901641249656677, "learning_rate": 9.856049129913072e-05, "loss": 0.0941, "step": 9820 }, { "grad_norm": 0.8980714082717896, "learning_rate": 9.85555634133253e-05, "loss": 0.1048, "step": 9830 }, { "grad_norm": 1.3309249877929688, "learning_rate": 9.855062723070208e-05, "loss": 0.1032, "step": 9840 }, { "grad_norm": 0.7717546820640564, "learning_rate": 9.854568275210454e-05, "loss": 0.0971, "step": 9850 }, { "grad_norm": 0.8004527688026428, "learning_rate": 9.854072997837754e-05, "loss": 0.0802, "step": 9860 }, { "grad_norm": 0.7926846742630005, "learning_rate": 9.853576891036737e-05, "loss": 0.0799, "step": 9870 }, { "grad_norm": 0.7610250115394592, "learning_rate": 9.853079954892177e-05, "loss": 0.0763, "step": 9880 }, { "grad_norm": 0.8393798470497131, "learning_rate": 9.852582189488983e-05, "loss": 0.116, "step": 9890 }, { "grad_norm": 0.8220639824867249, "learning_rate": 9.852083594912212e-05, "loss": 0.0864, "step": 9900 }, { "grad_norm": 0.7035698890686035, "learning_rate": 9.851584171247058e-05, "loss": 0.0915, "step": 9910 }, { "grad_norm": 0.6369820237159729, "learning_rate": 9.851083918578863e-05, "loss": 0.0948, "step": 9920 }, { "grad_norm": 0.9123503565788269, "learning_rate": 9.850582836993103e-05, "loss": 0.1197, "step": 9930 }, { "grad_norm": 0.7967870831489563, "learning_rate": 9.850080926575397e-05, "loss": 0.1041, "step": 9940 }, { "grad_norm": 0.7368808388710022, "learning_rate": 9.849578187411515e-05, "loss": 0.106, "step": 9950 }, { "grad_norm": 1.0037055015563965, "learning_rate": 9.849074619587354e-05, "loss": 0.1013, "step": 9960 }, { "grad_norm": 0.851769208908081, "learning_rate": 9.848570223188964e-05, "loss": 0.0941, "step": 9970 }, { "grad_norm": 0.7383456230163574, "learning_rate": 9.848064998302531e-05, "loss": 0.098, "step": 9980 }, { "grad_norm": 0.758791983127594, "learning_rate": 9.847558945014386e-05, "loss": 0.0877, "step": 9990 }, { "grad_norm": 0.6572126150131226, "learning_rate": 9.847052063410996e-05, "loss": 0.1094, "step": 10000 }, { "grad_norm": 0.6283993124961853, "learning_rate": 9.846544353578977e-05, "loss": 0.0989, "step": 10010 }, { "grad_norm": 1.0276449918746948, "learning_rate": 9.846035815605081e-05, "loss": 0.1036, "step": 10020 }, { "grad_norm": 0.6520408987998962, "learning_rate": 9.845526449576204e-05, "loss": 0.0889, "step": 10030 }, { "grad_norm": 0.8403255343437195, "learning_rate": 9.845016255579383e-05, "loss": 0.1054, "step": 10040 }, { "grad_norm": 0.9149771332740784, "learning_rate": 9.844505233701794e-05, "loss": 0.1014, "step": 10050 }, { "grad_norm": 0.740679144859314, "learning_rate": 9.843993384030757e-05, "loss": 0.0837, "step": 10060 }, { "grad_norm": 1.1945258378982544, "learning_rate": 9.843480706653737e-05, "loss": 0.0855, "step": 10070 }, { "grad_norm": 0.4197392761707306, "learning_rate": 9.84296720165833e-05, "loss": 0.0896, "step": 10080 }, { "grad_norm": 0.6170247197151184, "learning_rate": 9.842452869132286e-05, "loss": 0.1134, "step": 10090 }, { "grad_norm": 0.9212870001792908, "learning_rate": 9.841937709163489e-05, "loss": 0.1013, "step": 10100 }, { "grad_norm": 0.6818482279777527, "learning_rate": 9.841421721839962e-05, "loss": 0.0865, "step": 10110 }, { "grad_norm": 0.5348016619682312, "learning_rate": 9.840904907249879e-05, "loss": 0.0783, "step": 10120 }, { "grad_norm": 1.1092932224273682, "learning_rate": 9.840387265481545e-05, "loss": 0.1126, "step": 10130 }, { "grad_norm": 0.8289865851402283, "learning_rate": 9.839868796623411e-05, "loss": 0.1119, "step": 10140 }, { "grad_norm": 0.7064892053604126, "learning_rate": 9.839349500764072e-05, "loss": 0.1124, "step": 10150 }, { "grad_norm": 0.6445887684822083, "learning_rate": 9.83882937799226e-05, "loss": 0.0826, "step": 10160 }, { "grad_norm": 0.7297993898391724, "learning_rate": 9.838308428396849e-05, "loss": 0.0925, "step": 10170 }, { "grad_norm": 0.5782002210617065, "learning_rate": 9.837786652066854e-05, "loss": 0.0899, "step": 10180 }, { "grad_norm": 0.6109523177146912, "learning_rate": 9.837264049091437e-05, "loss": 0.0923, "step": 10190 }, { "grad_norm": 0.9153347015380859, "learning_rate": 9.836740619559893e-05, "loss": 0.0852, "step": 10200 }, { "grad_norm": 0.8436773419380188, "learning_rate": 9.836216363561659e-05, "loss": 0.0803, "step": 10210 }, { "grad_norm": 0.5406210422515869, "learning_rate": 9.835691281186322e-05, "loss": 0.1053, "step": 10220 }, { "grad_norm": 0.7716276049613953, "learning_rate": 9.8351653725236e-05, "loss": 0.1058, "step": 10230 }, { "grad_norm": 0.6573424935340881, "learning_rate": 9.83463863766336e-05, "loss": 0.1149, "step": 10240 }, { "grad_norm": 0.8041984438896179, "learning_rate": 9.834111076695602e-05, "loss": 0.0934, "step": 10250 }, { "grad_norm": 0.7506484985351562, "learning_rate": 9.833582689710477e-05, "loss": 0.1231, "step": 10260 }, { "grad_norm": 0.7998095154762268, "learning_rate": 9.833053476798268e-05, "loss": 0.0937, "step": 10270 }, { "grad_norm": 0.8202142715454102, "learning_rate": 9.832523438049404e-05, "loss": 0.0792, "step": 10280 }, { "grad_norm": 0.6752286553382874, "learning_rate": 9.831992573554454e-05, "loss": 0.0864, "step": 10290 }, { "grad_norm": 0.6092173457145691, "learning_rate": 9.831460883404128e-05, "loss": 0.0733, "step": 10300 }, { "grad_norm": 0.7817274332046509, "learning_rate": 9.830928367689278e-05, "loss": 0.1142, "step": 10310 }, { "grad_norm": 0.569685161113739, "learning_rate": 9.830395026500896e-05, "loss": 0.0882, "step": 10320 }, { "grad_norm": 0.5811476707458496, "learning_rate": 9.829860859930115e-05, "loss": 0.0959, "step": 10330 }, { "grad_norm": 0.6566359996795654, "learning_rate": 9.829325868068212e-05, "loss": 0.0804, "step": 10340 }, { "grad_norm": 1.0354037284851074, "learning_rate": 9.8287900510066e-05, "loss": 0.0962, "step": 10350 }, { "grad_norm": 0.6254611015319824, "learning_rate": 9.828253408836834e-05, "loss": 0.0941, "step": 10360 }, { "grad_norm": 0.6863200068473816, "learning_rate": 9.827715941650615e-05, "loss": 0.088, "step": 10370 }, { "grad_norm": 0.8394700288772583, "learning_rate": 9.82717764953978e-05, "loss": 0.1256, "step": 10380 }, { "grad_norm": 0.9677095413208008, "learning_rate": 9.826638532596308e-05, "loss": 0.0921, "step": 10390 }, { "grad_norm": 0.7142594456672668, "learning_rate": 9.82609859091232e-05, "loss": 0.0771, "step": 10400 }, { "grad_norm": 0.6801044940948486, "learning_rate": 9.825557824580076e-05, "loss": 0.0928, "step": 10410 }, { "grad_norm": 0.8823993802070618, "learning_rate": 9.82501623369198e-05, "loss": 0.1003, "step": 10420 }, { "grad_norm": 0.6956878900527954, "learning_rate": 9.824473818340574e-05, "loss": 0.086, "step": 10430 }, { "grad_norm": 0.5787699222564697, "learning_rate": 9.823930578618541e-05, "loss": 0.0882, "step": 10440 }, { "grad_norm": 0.5563122034072876, "learning_rate": 9.823386514618709e-05, "loss": 0.0853, "step": 10450 }, { "grad_norm": 0.8095725774765015, "learning_rate": 9.82284162643404e-05, "loss": 0.088, "step": 10460 }, { "grad_norm": 0.8790398240089417, "learning_rate": 9.822295914157642e-05, "loss": 0.0996, "step": 10470 }, { "grad_norm": 0.5778751969337463, "learning_rate": 9.821749377882763e-05, "loss": 0.0754, "step": 10480 }, { "grad_norm": 0.6322532296180725, "learning_rate": 9.821202017702791e-05, "loss": 0.1216, "step": 10490 }, { "grad_norm": 0.9385680556297302, "learning_rate": 9.820653833711253e-05, "loss": 0.0993, "step": 10500 }, { "grad_norm": 0.9714288115501404, "learning_rate": 9.820104826001822e-05, "loss": 0.0924, "step": 10510 }, { "grad_norm": 0.922282874584198, "learning_rate": 9.819554994668305e-05, "loss": 0.0967, "step": 10520 }, { "grad_norm": 0.587931752204895, "learning_rate": 9.819004339804654e-05, "loss": 0.0927, "step": 10530 }, { "grad_norm": 0.7242797017097473, "learning_rate": 9.818452861504961e-05, "loss": 0.0787, "step": 10540 }, { "grad_norm": 0.8893179297447205, "learning_rate": 9.81790055986346e-05, "loss": 0.0979, "step": 10550 }, { "grad_norm": 0.8386508226394653, "learning_rate": 9.817347434974523e-05, "loss": 0.0976, "step": 10560 }, { "grad_norm": 0.5388476848602295, "learning_rate": 9.816793486932664e-05, "loss": 0.1087, "step": 10570 }, { "grad_norm": 0.7070854306221008, "learning_rate": 9.816238715832538e-05, "loss": 0.0704, "step": 10580 }, { "grad_norm": 0.5376454591751099, "learning_rate": 9.815683121768939e-05, "loss": 0.0827, "step": 10590 }, { "grad_norm": 0.5355557799339294, "learning_rate": 9.815126704836804e-05, "loss": 0.0689, "step": 10600 }, { "grad_norm": 0.5122017860412598, "learning_rate": 9.81456946513121e-05, "loss": 0.073, "step": 10610 }, { "grad_norm": 0.7329176664352417, "learning_rate": 9.814011402747373e-05, "loss": 0.0853, "step": 10620 }, { "grad_norm": 0.7286433577537537, "learning_rate": 9.813452517780651e-05, "loss": 0.0937, "step": 10630 }, { "grad_norm": 0.6416440010070801, "learning_rate": 9.81289281032654e-05, "loss": 0.084, "step": 10640 }, { "grad_norm": 0.7991395592689514, "learning_rate": 9.812332280480683e-05, "loss": 0.1044, "step": 10650 }, { "grad_norm": 1.0954334735870361, "learning_rate": 9.811770928338854e-05, "loss": 0.0933, "step": 10660 }, { "grad_norm": 0.955048143863678, "learning_rate": 9.811208753996979e-05, "loss": 0.0856, "step": 10670 }, { "grad_norm": 0.9542484879493713, "learning_rate": 9.810645757551113e-05, "loss": 0.1124, "step": 10680 }, { "grad_norm": 0.6558322310447693, "learning_rate": 9.810081939097459e-05, "loss": 0.0924, "step": 10690 }, { "grad_norm": 0.6028797626495361, "learning_rate": 9.809517298732356e-05, "loss": 0.0826, "step": 10700 }, { "grad_norm": 0.6233251690864563, "learning_rate": 9.80895183655229e-05, "loss": 0.0849, "step": 10710 }, { "grad_norm": 0.6124404668807983, "learning_rate": 9.808385552653877e-05, "loss": 0.0956, "step": 10720 }, { "grad_norm": 0.6022828221321106, "learning_rate": 9.807818447133886e-05, "loss": 0.0741, "step": 10730 }, { "grad_norm": 0.6868501901626587, "learning_rate": 9.807250520089215e-05, "loss": 0.0941, "step": 10740 }, { "grad_norm": 0.6662160158157349, "learning_rate": 9.806681771616908e-05, "loss": 0.0897, "step": 10750 }, { "grad_norm": 0.567113995552063, "learning_rate": 9.80611220181415e-05, "loss": 0.0771, "step": 10760 }, { "grad_norm": 0.5871113538742065, "learning_rate": 9.805541810778264e-05, "loss": 0.0789, "step": 10770 }, { "grad_norm": 0.7718914747238159, "learning_rate": 9.804970598606716e-05, "loss": 0.0821, "step": 10780 }, { "grad_norm": 0.684146523475647, "learning_rate": 9.804398565397106e-05, "loss": 0.0902, "step": 10790 }, { "grad_norm": 0.6355236172676086, "learning_rate": 9.803825711247183e-05, "loss": 0.0795, "step": 10800 }, { "grad_norm": 0.679836094379425, "learning_rate": 9.803252036254831e-05, "loss": 0.0828, "step": 10810 }, { "grad_norm": 0.4686671197414398, "learning_rate": 9.802677540518076e-05, "loss": 0.0772, "step": 10820 }, { "grad_norm": 0.5727829337120056, "learning_rate": 9.802102224135081e-05, "loss": 0.0855, "step": 10830 }, { "grad_norm": 0.6217652559280396, "learning_rate": 9.801526087204155e-05, "loss": 0.0784, "step": 10840 }, { "grad_norm": 0.6697045564651489, "learning_rate": 9.800949129823743e-05, "loss": 0.0689, "step": 10850 }, { "grad_norm": 0.5614807605743408, "learning_rate": 9.80037135209243e-05, "loss": 0.0871, "step": 10860 }, { "grad_norm": 0.6213507056236267, "learning_rate": 9.799792754108946e-05, "loss": 0.1171, "step": 10870 }, { "grad_norm": 0.6573527455329895, "learning_rate": 9.799213335972152e-05, "loss": 0.0895, "step": 10880 }, { "grad_norm": 0.6356884241104126, "learning_rate": 9.798633097781058e-05, "loss": 0.0744, "step": 10890 }, { "grad_norm": 0.6982664465904236, "learning_rate": 9.79805203963481e-05, "loss": 0.0742, "step": 10900 }, { "grad_norm": 1.171636700630188, "learning_rate": 9.797470161632697e-05, "loss": 0.0908, "step": 10910 }, { "grad_norm": 0.6255581974983215, "learning_rate": 9.796887463874145e-05, "loss": 0.1067, "step": 10920 }, { "grad_norm": 0.7604122757911682, "learning_rate": 9.796303946458718e-05, "loss": 0.0735, "step": 10930 }, { "grad_norm": 0.6268395781517029, "learning_rate": 9.795719609486127e-05, "loss": 0.0752, "step": 10940 }, { "grad_norm": 0.937559962272644, "learning_rate": 9.795134453056219e-05, "loss": 0.0864, "step": 10950 }, { "grad_norm": 0.8054167628288269, "learning_rate": 9.794548477268979e-05, "loss": 0.0947, "step": 10960 }, { "grad_norm": 0.6701606512069702, "learning_rate": 9.793961682224537e-05, "loss": 0.0801, "step": 10970 }, { "grad_norm": 0.6064708828926086, "learning_rate": 9.793374068023156e-05, "loss": 0.0813, "step": 10980 }, { "grad_norm": 0.921841025352478, "learning_rate": 9.792785634765247e-05, "loss": 0.0996, "step": 10990 }, { "grad_norm": 0.6610098481178284, "learning_rate": 9.792196382551357e-05, "loss": 0.0842, "step": 11000 }, { "grad_norm": 0.8937835097312927, "learning_rate": 9.791606311482171e-05, "loss": 0.1026, "step": 11010 }, { "grad_norm": 0.6071370840072632, "learning_rate": 9.791015421658518e-05, "loss": 0.0853, "step": 11020 }, { "grad_norm": 0.7747380137443542, "learning_rate": 9.790423713181362e-05, "loss": 0.0906, "step": 11030 }, { "grad_norm": 0.642665684223175, "learning_rate": 9.789831186151814e-05, "loss": 0.0761, "step": 11040 }, { "grad_norm": 0.5344647169113159, "learning_rate": 9.789237840671118e-05, "loss": 0.084, "step": 11050 }, { "grad_norm": 0.8099713921546936, "learning_rate": 9.78864367684066e-05, "loss": 0.0958, "step": 11060 }, { "grad_norm": 0.6606696844100952, "learning_rate": 9.788048694761968e-05, "loss": 0.0752, "step": 11070 }, { "grad_norm": 0.6719430088996887, "learning_rate": 9.787452894536709e-05, "loss": 0.083, "step": 11080 }, { "grad_norm": 0.7039771676063538, "learning_rate": 9.786856276266685e-05, "loss": 0.0829, "step": 11090 }, { "grad_norm": 0.7471461892127991, "learning_rate": 9.786258840053845e-05, "loss": 0.0835, "step": 11100 }, { "grad_norm": 0.7259395122528076, "learning_rate": 9.785660586000273e-05, "loss": 0.0815, "step": 11110 }, { "grad_norm": 0.8947005867958069, "learning_rate": 9.785061514208196e-05, "loss": 0.0842, "step": 11120 }, { "grad_norm": 0.7393448352813721, "learning_rate": 9.784461624779977e-05, "loss": 0.0986, "step": 11130 }, { "grad_norm": 0.7009161114692688, "learning_rate": 9.783860917818123e-05, "loss": 0.0817, "step": 11140 }, { "grad_norm": 0.5223697423934937, "learning_rate": 9.783259393425277e-05, "loss": 0.091, "step": 11150 }, { "grad_norm": 1.092195987701416, "learning_rate": 9.782657051704221e-05, "loss": 0.0997, "step": 11160 }, { "grad_norm": 0.551846981048584, "learning_rate": 9.782053892757883e-05, "loss": 0.0796, "step": 11170 }, { "grad_norm": 0.6890935897827148, "learning_rate": 9.781449916689324e-05, "loss": 0.0743, "step": 11180 }, { "grad_norm": 0.7900750637054443, "learning_rate": 9.780845123601746e-05, "loss": 0.0811, "step": 11190 }, { "grad_norm": 0.6850210428237915, "learning_rate": 9.780239513598492e-05, "loss": 0.0975, "step": 11200 }, { "grad_norm": 0.7038272023200989, "learning_rate": 9.779633086783047e-05, "loss": 0.0954, "step": 11210 }, { "grad_norm": 0.6152286529541016, "learning_rate": 9.779025843259031e-05, "loss": 0.0778, "step": 11220 }, { "grad_norm": 0.8660353422164917, "learning_rate": 9.778417783130204e-05, "loss": 0.0839, "step": 11230 }, { "grad_norm": 0.7089241147041321, "learning_rate": 9.777808906500468e-05, "loss": 0.0816, "step": 11240 }, { "grad_norm": 0.5499656796455383, "learning_rate": 9.777199213473862e-05, "loss": 0.0778, "step": 11250 }, { "grad_norm": 0.548965573310852, "learning_rate": 9.77658870415457e-05, "loss": 0.0891, "step": 11260 }, { "grad_norm": 0.7914564609527588, "learning_rate": 9.775977378646906e-05, "loss": 0.0833, "step": 11270 }, { "grad_norm": 0.6174649596214294, "learning_rate": 9.775365237055331e-05, "loss": 0.0759, "step": 11280 }, { "grad_norm": 0.657505989074707, "learning_rate": 9.774752279484445e-05, "loss": 0.0796, "step": 11290 }, { "grad_norm": 0.4709673821926117, "learning_rate": 9.774138506038984e-05, "loss": 0.1265, "step": 11300 }, { "grad_norm": 1.0291234254837036, "learning_rate": 9.773523916823826e-05, "loss": 0.0905, "step": 11310 }, { "grad_norm": 0.7348482608795166, "learning_rate": 9.772908511943986e-05, "loss": 0.1002, "step": 11320 }, { "grad_norm": 0.8203378319740295, "learning_rate": 9.77229229150462e-05, "loss": 0.0826, "step": 11330 }, { "grad_norm": 0.7839202880859375, "learning_rate": 9.771675255611024e-05, "loss": 0.0765, "step": 11340 }, { "grad_norm": 0.6977891325950623, "learning_rate": 9.771057404368632e-05, "loss": 0.0764, "step": 11350 }, { "grad_norm": 0.6852166652679443, "learning_rate": 9.770438737883018e-05, "loss": 0.1021, "step": 11360 }, { "grad_norm": 0.7206506729125977, "learning_rate": 9.769819256259898e-05, "loss": 0.0825, "step": 11370 }, { "grad_norm": 0.666120707988739, "learning_rate": 9.769198959605119e-05, "loss": 0.1107, "step": 11380 }, { "grad_norm": 0.779377818107605, "learning_rate": 9.768577848024678e-05, "loss": 0.0892, "step": 11390 }, { "grad_norm": 0.5830934643745422, "learning_rate": 9.767955921624702e-05, "loss": 0.0875, "step": 11400 }, { "grad_norm": 0.7492455840110779, "learning_rate": 9.767333180511465e-05, "loss": 0.0904, "step": 11410 }, { "grad_norm": 0.9089920520782471, "learning_rate": 9.766709624791373e-05, "loss": 0.1028, "step": 11420 }, { "grad_norm": 0.7851070165634155, "learning_rate": 9.766085254570975e-05, "loss": 0.0872, "step": 11430 }, { "grad_norm": 0.5381270051002502, "learning_rate": 9.76546006995696e-05, "loss": 0.0885, "step": 11440 }, { "grad_norm": 0.8484400510787964, "learning_rate": 9.764834071056155e-05, "loss": 0.0824, "step": 11450 }, { "grad_norm": 0.6160448789596558, "learning_rate": 9.764207257975526e-05, "loss": 0.0747, "step": 11460 }, { "grad_norm": 0.8076842427253723, "learning_rate": 9.763579630822179e-05, "loss": 0.0961, "step": 11470 }, { "grad_norm": 0.5728793740272522, "learning_rate": 9.762951189703356e-05, "loss": 0.102, "step": 11480 }, { "grad_norm": 0.659589409828186, "learning_rate": 9.762321934726442e-05, "loss": 0.1125, "step": 11490 }, { "grad_norm": 0.48778924345970154, "learning_rate": 9.761691865998959e-05, "loss": 0.1029, "step": 11500 }, { "grad_norm": 0.7686329483985901, "learning_rate": 9.76106098362857e-05, "loss": 0.0971, "step": 11510 }, { "grad_norm": 0.7148563265800476, "learning_rate": 9.760429287723072e-05, "loss": 0.0907, "step": 11520 }, { "grad_norm": 0.5220844149589539, "learning_rate": 9.759796778390406e-05, "loss": 0.0783, "step": 11530 }, { "grad_norm": 0.5528750419616699, "learning_rate": 9.759163455738653e-05, "loss": 0.0952, "step": 11540 }, { "grad_norm": 0.799903929233551, "learning_rate": 9.75852931987603e-05, "loss": 0.1183, "step": 11550 }, { "grad_norm": 1.2908307313919067, "learning_rate": 9.757894370910891e-05, "loss": 0.1032, "step": 11560 }, { "grad_norm": 0.9768847823143005, "learning_rate": 9.757258608951733e-05, "loss": 0.0794, "step": 11570 }, { "grad_norm": 1.0441935062408447, "learning_rate": 9.75662203410719e-05, "loss": 0.0827, "step": 11580 }, { "grad_norm": 0.723759651184082, "learning_rate": 9.755984646486034e-05, "loss": 0.1071, "step": 11590 }, { "grad_norm": 0.8871656656265259, "learning_rate": 9.75534644619718e-05, "loss": 0.0852, "step": 11600 }, { "grad_norm": 0.6714484095573425, "learning_rate": 9.754707433349676e-05, "loss": 0.0907, "step": 11610 }, { "grad_norm": 0.7165369987487793, "learning_rate": 9.754067608052715e-05, "loss": 0.0769, "step": 11620 }, { "grad_norm": 0.7909830808639526, "learning_rate": 9.753426970415622e-05, "loss": 0.0698, "step": 11630 }, { "grad_norm": 0.5649897456169128, "learning_rate": 9.752785520547868e-05, "loss": 0.0814, "step": 11640 }, { "grad_norm": 0.6388845443725586, "learning_rate": 9.752143258559056e-05, "loss": 0.0866, "step": 11650 }, { "grad_norm": 0.6707885265350342, "learning_rate": 9.751500184558933e-05, "loss": 0.0924, "step": 11660 }, { "grad_norm": 1.2035521268844604, "learning_rate": 9.750856298657383e-05, "loss": 0.0877, "step": 11670 }, { "grad_norm": 0.6005570888519287, "learning_rate": 9.750211600964428e-05, "loss": 0.0723, "step": 11680 }, { "grad_norm": 0.50471431016922, "learning_rate": 9.749566091590226e-05, "loss": 0.0827, "step": 11690 }, { "grad_norm": 0.9525503516197205, "learning_rate": 9.748919770645083e-05, "loss": 0.0718, "step": 11700 }, { "grad_norm": 0.6899144649505615, "learning_rate": 9.748272638239432e-05, "loss": 0.0929, "step": 11710 }, { "grad_norm": 0.6362553834915161, "learning_rate": 9.747624694483855e-05, "loss": 0.0771, "step": 11720 }, { "grad_norm": 0.8728675246238708, "learning_rate": 9.746975939489065e-05, "loss": 0.0843, "step": 11730 }, { "grad_norm": 0.9800493717193604, "learning_rate": 9.746326373365918e-05, "loss": 0.0987, "step": 11740 }, { "grad_norm": 0.8145681023597717, "learning_rate": 9.745675996225403e-05, "loss": 0.0956, "step": 11750 }, { "grad_norm": 0.7997785806655884, "learning_rate": 9.745024808178657e-05, "loss": 0.0947, "step": 11760 }, { "grad_norm": 0.6246140599250793, "learning_rate": 9.744372809336947e-05, "loss": 0.1039, "step": 11770 }, { "grad_norm": 0.6649565100669861, "learning_rate": 9.743719999811682e-05, "loss": 0.075, "step": 11780 }, { "grad_norm": 1.6336804628372192, "learning_rate": 9.743066379714412e-05, "loss": 0.0998, "step": 11790 }, { "grad_norm": 0.7954223155975342, "learning_rate": 9.74241194915682e-05, "loss": 0.1239, "step": 11800 }, { "grad_norm": 0.6929610967636108, "learning_rate": 9.741756708250731e-05, "loss": 0.0937, "step": 11810 }, { "grad_norm": 0.6825568079948425, "learning_rate": 9.741100657108109e-05, "loss": 0.0858, "step": 11820 }, { "grad_norm": 0.8115597367286682, "learning_rate": 9.740443795841054e-05, "loss": 0.0827, "step": 11830 }, { "grad_norm": 0.6701542735099792, "learning_rate": 9.739786124561805e-05, "loss": 0.0855, "step": 11840 }, { "grad_norm": 0.49665114283561707, "learning_rate": 9.73912764338274e-05, "loss": 0.0885, "step": 11850 }, { "grad_norm": 2.2583751678466797, "learning_rate": 9.738468352416377e-05, "loss": 0.1025, "step": 11860 }, { "grad_norm": 0.6335291266441345, "learning_rate": 9.737808251775369e-05, "loss": 0.0842, "step": 11870 }, { "grad_norm": 0.8086139559745789, "learning_rate": 9.737147341572512e-05, "loss": 0.0906, "step": 11880 }, { "grad_norm": 0.7077919840812683, "learning_rate": 9.736485621920735e-05, "loss": 0.0963, "step": 11890 }, { "grad_norm": 0.6755766868591309, "learning_rate": 9.735823092933108e-05, "loss": 0.083, "step": 11900 }, { "grad_norm": 0.6919364929199219, "learning_rate": 9.735159754722838e-05, "loss": 0.08, "step": 11910 }, { "grad_norm": 0.6277778744697571, "learning_rate": 9.734495607403275e-05, "loss": 0.0856, "step": 11920 }, { "grad_norm": 0.5448897480964661, "learning_rate": 9.733830651087901e-05, "loss": 0.0665, "step": 11930 }, { "grad_norm": 0.6729269027709961, "learning_rate": 9.733164885890338e-05, "loss": 0.0958, "step": 11940 }, { "grad_norm": 0.7037299275398254, "learning_rate": 9.732498311924349e-05, "loss": 0.0859, "step": 11950 }, { "grad_norm": 0.6043156385421753, "learning_rate": 9.731830929303833e-05, "loss": 0.0815, "step": 11960 }, { "grad_norm": 0.644117534160614, "learning_rate": 9.731162738142827e-05, "loss": 0.076, "step": 11970 }, { "grad_norm": 0.6703755855560303, "learning_rate": 9.730493738555506e-05, "loss": 0.0985, "step": 11980 }, { "grad_norm": 0.6795238256454468, "learning_rate": 9.729823930656186e-05, "loss": 0.0796, "step": 11990 }, { "grad_norm": 0.5866633653640747, "learning_rate": 9.729153314559316e-05, "loss": 0.1231, "step": 12000 }, { "grad_norm": 0.6001140475273132, "learning_rate": 9.728481890379486e-05, "loss": 0.0853, "step": 12010 }, { "grad_norm": 0.6275665760040283, "learning_rate": 9.727809658231428e-05, "loss": 0.0815, "step": 12020 }, { "grad_norm": 0.3820286691188812, "learning_rate": 9.727136618230003e-05, "loss": 0.0891, "step": 12030 }, { "grad_norm": 1.666040301322937, "learning_rate": 9.726462770490219e-05, "loss": 0.0909, "step": 12040 }, { "grad_norm": 0.7648060917854309, "learning_rate": 9.725788115127214e-05, "loss": 0.084, "step": 12050 }, { "grad_norm": 0.5852405428886414, "learning_rate": 9.725112652256274e-05, "loss": 0.0828, "step": 12060 }, { "grad_norm": 1.013217568397522, "learning_rate": 9.724436381992812e-05, "loss": 0.0822, "step": 12070 }, { "grad_norm": 0.7738102078437805, "learning_rate": 9.723759304452387e-05, "loss": 0.0889, "step": 12080 }, { "grad_norm": 0.7887134552001953, "learning_rate": 9.72308141975069e-05, "loss": 0.0942, "step": 12090 }, { "grad_norm": 0.6560775637626648, "learning_rate": 9.722402728003557e-05, "loss": 0.0837, "step": 12100 }, { "grad_norm": 0.7321577072143555, "learning_rate": 9.721723229326953e-05, "loss": 0.0767, "step": 12110 }, { "grad_norm": 0.6801449060440063, "learning_rate": 9.721042923836992e-05, "loss": 0.0716, "step": 12120 }, { "grad_norm": 0.6904383897781372, "learning_rate": 9.720361811649914e-05, "loss": 0.0783, "step": 12130 }, { "grad_norm": 0.7963318824768066, "learning_rate": 9.719679892882106e-05, "loss": 0.094, "step": 12140 }, { "grad_norm": 1.123914122581482, "learning_rate": 9.718997167650085e-05, "loss": 0.0962, "step": 12150 }, { "grad_norm": 0.773665726184845, "learning_rate": 9.718313636070515e-05, "loss": 0.0961, "step": 12160 }, { "grad_norm": 0.6553956866264343, "learning_rate": 9.717629298260192e-05, "loss": 0.0932, "step": 12170 }, { "grad_norm": 0.47950342297554016, "learning_rate": 9.716944154336047e-05, "loss": 0.0848, "step": 12180 }, { "grad_norm": 0.8133123517036438, "learning_rate": 9.716258204415157e-05, "loss": 0.1149, "step": 12190 }, { "grad_norm": 0.8845427632331848, "learning_rate": 9.715571448614728e-05, "loss": 0.1095, "step": 12200 }, { "grad_norm": 0.7739129066467285, "learning_rate": 9.71488388705211e-05, "loss": 0.0713, "step": 12210 }, { "grad_norm": 0.7392635941505432, "learning_rate": 9.714195519844788e-05, "loss": 0.0754, "step": 12220 }, { "grad_norm": 0.5532587766647339, "learning_rate": 9.713506347110386e-05, "loss": 0.0805, "step": 12230 }, { "grad_norm": 0.645677387714386, "learning_rate": 9.712816368966662e-05, "loss": 0.1063, "step": 12240 }, { "grad_norm": 0.7800450325012207, "learning_rate": 9.712125585531517e-05, "loss": 0.0862, "step": 12250 }, { "grad_norm": 0.7396237850189209, "learning_rate": 9.711433996922988e-05, "loss": 0.0758, "step": 12260 }, { "grad_norm": 0.7439005970954895, "learning_rate": 9.710741603259245e-05, "loss": 0.0826, "step": 12270 }, { "grad_norm": 0.5282379984855652, "learning_rate": 9.710048404658603e-05, "loss": 0.0906, "step": 12280 }, { "grad_norm": 0.7391471862792969, "learning_rate": 9.709354401239508e-05, "loss": 0.1085, "step": 12290 }, { "grad_norm": 0.5466798543930054, "learning_rate": 9.708659593120546e-05, "loss": 0.0945, "step": 12300 }, { "grad_norm": 0.88286292552948, "learning_rate": 9.707963980420443e-05, "loss": 0.0897, "step": 12310 }, { "grad_norm": 0.6586604714393616, "learning_rate": 9.707267563258058e-05, "loss": 0.1207, "step": 12320 }, { "grad_norm": 0.6634289622306824, "learning_rate": 9.70657034175239e-05, "loss": 0.1295, "step": 12330 }, { "grad_norm": 0.8884888291358948, "learning_rate": 9.705872316022577e-05, "loss": 0.0913, "step": 12340 }, { "grad_norm": 0.7719091773033142, "learning_rate": 9.705173486187891e-05, "loss": 0.0884, "step": 12350 }, { "grad_norm": 0.6233705282211304, "learning_rate": 9.704473852367741e-05, "loss": 0.0679, "step": 12360 }, { "grad_norm": 0.7880741357803345, "learning_rate": 9.70377341468168e-05, "loss": 0.0917, "step": 12370 }, { "grad_norm": 0.5748524069786072, "learning_rate": 9.703072173249389e-05, "loss": 0.0935, "step": 12380 }, { "grad_norm": 0.5720374584197998, "learning_rate": 9.702370128190693e-05, "loss": 0.085, "step": 12390 }, { "grad_norm": 0.6051774024963379, "learning_rate": 9.701667279625552e-05, "loss": 0.0533, "step": 12400 }, { "grad_norm": 0.856592059135437, "learning_rate": 9.700963627674065e-05, "loss": 0.0829, "step": 12410 }, { "grad_norm": 0.8560513854026794, "learning_rate": 9.700259172456466e-05, "loss": 0.0842, "step": 12420 }, { "grad_norm": 0.5023458003997803, "learning_rate": 9.699553914093124e-05, "loss": 0.0609, "step": 12430 }, { "grad_norm": 0.736875593662262, "learning_rate": 9.698847852704553e-05, "loss": 0.0704, "step": 12440 }, { "grad_norm": 0.6154307126998901, "learning_rate": 9.6981409884114e-05, "loss": 0.1001, "step": 12450 }, { "grad_norm": 0.7155463099479675, "learning_rate": 9.697433321334443e-05, "loss": 0.0745, "step": 12460 }, { "grad_norm": 0.6349183320999146, "learning_rate": 9.696724851594607e-05, "loss": 0.0823, "step": 12470 }, { "grad_norm": 0.5962799787521362, "learning_rate": 9.696015579312952e-05, "loss": 0.0738, "step": 12480 }, { "grad_norm": 0.6827492713928223, "learning_rate": 9.695305504610668e-05, "loss": 0.0929, "step": 12490 }, { "grad_norm": 0.49394625425338745, "learning_rate": 9.694594627609092e-05, "loss": 0.0957, "step": 12500 }, { "grad_norm": 0.6343525052070618, "learning_rate": 9.693882948429691e-05, "loss": 0.0774, "step": 12510 }, { "grad_norm": 0.895384669303894, "learning_rate": 9.693170467194071e-05, "loss": 0.0826, "step": 12520 }, { "grad_norm": 0.8163570165634155, "learning_rate": 9.692457184023977e-05, "loss": 0.071, "step": 12530 }, { "grad_norm": 0.6187010407447815, "learning_rate": 9.691743099041291e-05, "loss": 0.1024, "step": 12540 }, { "grad_norm": 0.652799129486084, "learning_rate": 9.691028212368027e-05, "loss": 0.0679, "step": 12550 }, { "grad_norm": 0.8703932166099548, "learning_rate": 9.690312524126342e-05, "loss": 0.0938, "step": 12560 }, { "grad_norm": 0.6588616967201233, "learning_rate": 9.689596034438527e-05, "loss": 0.0748, "step": 12570 }, { "grad_norm": 0.637446403503418, "learning_rate": 9.688878743427012e-05, "loss": 0.0828, "step": 12580 }, { "grad_norm": 0.6540963649749756, "learning_rate": 9.688160651214359e-05, "loss": 0.0823, "step": 12590 }, { "grad_norm": 0.6550186276435852, "learning_rate": 9.687441757923273e-05, "loss": 0.0664, "step": 12600 }, { "grad_norm": 0.5559945106506348, "learning_rate": 9.68672206367659e-05, "loss": 0.0768, "step": 12610 }, { "grad_norm": 0.6224814653396606, "learning_rate": 9.686001568597291e-05, "loss": 0.08, "step": 12620 }, { "grad_norm": 0.4125194549560547, "learning_rate": 9.685280272808486e-05, "loss": 0.075, "step": 12630 }, { "grad_norm": 0.8510065078735352, "learning_rate": 9.684558176433424e-05, "loss": 0.078, "step": 12640 }, { "grad_norm": 0.7208478450775146, "learning_rate": 9.683835279595495e-05, "loss": 0.0815, "step": 12650 }, { "grad_norm": 0.7043651342391968, "learning_rate": 9.683111582418216e-05, "loss": 0.0828, "step": 12660 }, { "grad_norm": 1.0787413120269775, "learning_rate": 9.682387085025254e-05, "loss": 0.1121, "step": 12670 }, { "grad_norm": 0.8452022075653076, "learning_rate": 9.681661787540401e-05, "loss": 0.0691, "step": 12680 }, { "grad_norm": 0.6160442233085632, "learning_rate": 9.680935690087593e-05, "loss": 0.0894, "step": 12690 }, { "grad_norm": 0.8252565860748291, "learning_rate": 9.680208792790901e-05, "loss": 0.0823, "step": 12700 }, { "grad_norm": 0.7023848295211792, "learning_rate": 9.679481095774529e-05, "loss": 0.1028, "step": 12710 }, { "grad_norm": 0.8090103268623352, "learning_rate": 9.678752599162822e-05, "loss": 0.0684, "step": 12720 }, { "grad_norm": 0.6213136911392212, "learning_rate": 9.678023303080259e-05, "loss": 0.0632, "step": 12730 }, { "grad_norm": 0.6142275929450989, "learning_rate": 9.677293207651459e-05, "loss": 0.0918, "step": 12740 }, { "grad_norm": 1.2832489013671875, "learning_rate": 9.676562313001173e-05, "loss": 0.0748, "step": 12750 }, { "grad_norm": 0.8260116577148438, "learning_rate": 9.675830619254293e-05, "loss": 0.085, "step": 12760 }, { "grad_norm": 0.7194324135780334, "learning_rate": 9.675098126535843e-05, "loss": 0.0786, "step": 12770 }, { "grad_norm": 0.7700023651123047, "learning_rate": 9.674364834970988e-05, "loss": 0.0868, "step": 12780 }, { "grad_norm": 0.7628956437110901, "learning_rate": 9.673630744685028e-05, "loss": 0.0667, "step": 12790 }, { "grad_norm": 0.7963407039642334, "learning_rate": 9.672895855803397e-05, "loss": 0.0978, "step": 12800 }, { "grad_norm": 0.8194965124130249, "learning_rate": 9.672160168451667e-05, "loss": 0.058, "step": 12810 }, { "grad_norm": 0.6004250049591064, "learning_rate": 9.671423682755549e-05, "loss": 0.0782, "step": 12820 }, { "grad_norm": 0.5190635323524475, "learning_rate": 9.670686398840888e-05, "loss": 0.0943, "step": 12830 }, { "grad_norm": 0.5681372284889221, "learning_rate": 9.669948316833664e-05, "loss": 0.068, "step": 12840 }, { "grad_norm": 0.851018488407135, "learning_rate": 9.669209436859997e-05, "loss": 0.0762, "step": 12850 }, { "grad_norm": 0.5621426105499268, "learning_rate": 9.66846975904614e-05, "loss": 0.084, "step": 12860 }, { "grad_norm": 0.5798401832580566, "learning_rate": 9.667729283518483e-05, "loss": 0.0901, "step": 12870 }, { "grad_norm": 1.0014231204986572, "learning_rate": 9.666988010403557e-05, "loss": 0.0846, "step": 12880 }, { "grad_norm": 0.6992645263671875, "learning_rate": 9.66624593982802e-05, "loss": 0.0872, "step": 12890 }, { "grad_norm": 0.6920861601829529, "learning_rate": 9.665503071918675e-05, "loss": 0.0859, "step": 12900 }, { "grad_norm": 0.6865988373756409, "learning_rate": 9.664759406802456e-05, "loss": 0.0818, "step": 12910 }, { "grad_norm": 0.7727065682411194, "learning_rate": 9.664014944606437e-05, "loss": 0.0758, "step": 12920 }, { "grad_norm": 0.5609363317489624, "learning_rate": 9.663269685457822e-05, "loss": 0.0755, "step": 12930 }, { "grad_norm": 0.6995246410369873, "learning_rate": 9.662523629483962e-05, "loss": 0.0842, "step": 12940 }, { "grad_norm": 0.6755458116531372, "learning_rate": 9.661776776812333e-05, "loss": 0.0792, "step": 12950 }, { "grad_norm": 0.6864222884178162, "learning_rate": 9.661029127570553e-05, "loss": 0.0733, "step": 12960 }, { "grad_norm": 0.6682247519493103, "learning_rate": 9.660280681886373e-05, "loss": 0.0753, "step": 12970 }, { "grad_norm": 0.5994102954864502, "learning_rate": 9.659531439887685e-05, "loss": 0.0963, "step": 12980 }, { "grad_norm": 0.6375612616539001, "learning_rate": 9.658781401702511e-05, "loss": 0.0706, "step": 12990 }, { "grad_norm": 0.5576258897781372, "learning_rate": 9.658030567459015e-05, "loss": 0.0769, "step": 13000 }, { "grad_norm": 0.5285448431968689, "learning_rate": 9.65727893728549e-05, "loss": 0.0748, "step": 13010 }, { "grad_norm": 0.6113888025283813, "learning_rate": 9.656526511310375e-05, "loss": 0.0748, "step": 13020 }, { "grad_norm": 0.9723451137542725, "learning_rate": 9.655773289662233e-05, "loss": 0.0815, "step": 13030 }, { "grad_norm": 0.7847538590431213, "learning_rate": 9.655019272469772e-05, "loss": 0.0797, "step": 13040 }, { "grad_norm": 0.49591124057769775, "learning_rate": 9.654264459861832e-05, "loss": 0.066, "step": 13050 }, { "grad_norm": 0.5445635914802551, "learning_rate": 9.653508851967391e-05, "loss": 0.0757, "step": 13060 }, { "grad_norm": 0.563317060470581, "learning_rate": 9.65275244891556e-05, "loss": 0.1043, "step": 13070 }, { "grad_norm": 0.5903027653694153, "learning_rate": 9.651995250835591e-05, "loss": 0.0916, "step": 13080 }, { "grad_norm": 0.5763533711433411, "learning_rate": 9.651237257856862e-05, "loss": 0.0648, "step": 13090 }, { "grad_norm": 0.8482250571250916, "learning_rate": 9.6504784701089e-05, "loss": 0.0897, "step": 13100 }, { "grad_norm": 0.6904575228691101, "learning_rate": 9.649718887721357e-05, "loss": 0.0768, "step": 13110 }, { "grad_norm": 0.692891538143158, "learning_rate": 9.648958510824028e-05, "loss": 0.0857, "step": 13120 }, { "grad_norm": 0.783759593963623, "learning_rate": 9.648197339546837e-05, "loss": 0.0848, "step": 13130 }, { "grad_norm": 0.5447341203689575, "learning_rate": 9.647435374019851e-05, "loss": 0.0998, "step": 13140 }, { "grad_norm": 0.4139499068260193, "learning_rate": 9.646672614373266e-05, "loss": 0.0648, "step": 13150 }, { "grad_norm": 0.6931167840957642, "learning_rate": 9.645909060737418e-05, "loss": 0.0733, "step": 13160 }, { "grad_norm": 0.6999537944793701, "learning_rate": 9.645144713242778e-05, "loss": 0.0826, "step": 13170 }, { "grad_norm": 0.8372579216957092, "learning_rate": 9.64437957201995e-05, "loss": 0.1028, "step": 13180 }, { "grad_norm": 0.8494738936424255, "learning_rate": 9.643613637199678e-05, "loss": 0.0889, "step": 13190 }, { "grad_norm": 0.6257893443107605, "learning_rate": 9.642846908912839e-05, "loss": 0.1049, "step": 13200 }, { "grad_norm": 0.5293910503387451, "learning_rate": 9.642079387290444e-05, "loss": 0.0691, "step": 13210 }, { "grad_norm": 0.6867719292640686, "learning_rate": 9.641311072463644e-05, "loss": 0.079, "step": 13220 }, { "grad_norm": 0.7743628025054932, "learning_rate": 9.640541964563722e-05, "loss": 0.0842, "step": 13230 }, { "grad_norm": 0.45451071858406067, "learning_rate": 9.639772063722096e-05, "loss": 0.0564, "step": 13240 }, { "grad_norm": 1.1665418148040771, "learning_rate": 9.639001370070324e-05, "loss": 0.0838, "step": 13250 }, { "grad_norm": 0.7424172163009644, "learning_rate": 9.638229883740095e-05, "loss": 0.0728, "step": 13260 }, { "grad_norm": 0.7220883965492249, "learning_rate": 9.637457604863233e-05, "loss": 0.0751, "step": 13270 }, { "grad_norm": 0.7057815790176392, "learning_rate": 9.636684533571703e-05, "loss": 0.0738, "step": 13280 }, { "grad_norm": 0.9164227843284607, "learning_rate": 9.635910669997599e-05, "loss": 0.0786, "step": 13290 }, { "grad_norm": 0.5496073961257935, "learning_rate": 9.635136014273154e-05, "loss": 0.0991, "step": 13300 }, { "grad_norm": 0.7915287017822266, "learning_rate": 9.634360566530735e-05, "loss": 0.0865, "step": 13310 }, { "grad_norm": 0.6377527713775635, "learning_rate": 9.633584326902845e-05, "loss": 0.0685, "step": 13320 }, { "grad_norm": 0.6015055775642395, "learning_rate": 9.632807295522124e-05, "loss": 0.0913, "step": 13330 }, { "grad_norm": 0.825514018535614, "learning_rate": 9.632029472521342e-05, "loss": 0.0749, "step": 13340 }, { "grad_norm": 1.3103246688842773, "learning_rate": 9.631250858033409e-05, "loss": 0.0884, "step": 13350 }, { "grad_norm": 0.6769656538963318, "learning_rate": 9.630471452191371e-05, "loss": 0.0971, "step": 13360 }, { "grad_norm": 0.6393900513648987, "learning_rate": 9.629691255128405e-05, "loss": 0.0831, "step": 13370 }, { "grad_norm": 0.7450215220451355, "learning_rate": 9.628910266977825e-05, "loss": 0.0849, "step": 13380 }, { "grad_norm": 0.5188366770744324, "learning_rate": 9.628128487873083e-05, "loss": 0.0782, "step": 13390 }, { "grad_norm": 0.6073042750358582, "learning_rate": 9.627345917947761e-05, "loss": 0.0759, "step": 13400 }, { "grad_norm": 0.6448907256126404, "learning_rate": 9.626562557335579e-05, "loss": 0.0719, "step": 13410 }, { "grad_norm": 0.590133786201477, "learning_rate": 9.625778406170393e-05, "loss": 0.0733, "step": 13420 }, { "grad_norm": 0.8306886553764343, "learning_rate": 9.624993464586193e-05, "loss": 0.1133, "step": 13430 }, { "grad_norm": 0.518078625202179, "learning_rate": 9.624207732717105e-05, "loss": 0.0742, "step": 13440 }, { "grad_norm": 0.9031630158424377, "learning_rate": 9.623421210697386e-05, "loss": 0.072, "step": 13450 }, { "grad_norm": 0.9205825328826904, "learning_rate": 9.622633898661434e-05, "loss": 0.0789, "step": 13460 }, { "grad_norm": 1.0711936950683594, "learning_rate": 9.621845796743778e-05, "loss": 0.1185, "step": 13470 }, { "grad_norm": 0.5700156092643738, "learning_rate": 9.621056905079082e-05, "loss": 0.0909, "step": 13480 }, { "grad_norm": 0.6577034592628479, "learning_rate": 9.620267223802149e-05, "loss": 0.0644, "step": 13490 }, { "grad_norm": 0.6392770409584045, "learning_rate": 9.619476753047911e-05, "loss": 0.0676, "step": 13500 }, { "grad_norm": 0.6168755292892456, "learning_rate": 9.618685492951438e-05, "loss": 0.0769, "step": 13510 }, { "grad_norm": 0.7427413463592529, "learning_rate": 9.617893443647938e-05, "loss": 0.0799, "step": 13520 }, { "grad_norm": 0.5498666167259216, "learning_rate": 9.617100605272746e-05, "loss": 0.0655, "step": 13530 }, { "grad_norm": 0.4039534032344818, "learning_rate": 9.616306977961338e-05, "loss": 0.0658, "step": 13540 }, { "grad_norm": 0.6552375555038452, "learning_rate": 9.615512561849326e-05, "loss": 0.0862, "step": 13550 }, { "grad_norm": 0.9737327694892883, "learning_rate": 9.61471735707245e-05, "loss": 0.0895, "step": 13560 }, { "grad_norm": 0.6628707051277161, "learning_rate": 9.613921363766592e-05, "loss": 0.0743, "step": 13570 }, { "grad_norm": 0.657723069190979, "learning_rate": 9.613124582067763e-05, "loss": 0.0606, "step": 13580 }, { "grad_norm": 0.6277823448181152, "learning_rate": 9.612327012112112e-05, "loss": 0.0633, "step": 13590 }, { "grad_norm": 0.9074448943138123, "learning_rate": 9.611528654035921e-05, "loss": 0.109, "step": 13600 }, { "grad_norm": 0.8644704222679138, "learning_rate": 9.610729507975611e-05, "loss": 0.0668, "step": 13610 }, { "grad_norm": 0.9379439949989319, "learning_rate": 9.609929574067731e-05, "loss": 0.0761, "step": 13620 }, { "grad_norm": 0.5967087745666504, "learning_rate": 9.609128852448967e-05, "loss": 0.0703, "step": 13630 }, { "grad_norm": 0.6629801988601685, "learning_rate": 9.608327343256143e-05, "loss": 0.1004, "step": 13640 }, { "grad_norm": 0.8515738248825073, "learning_rate": 9.607525046626216e-05, "loss": 0.0801, "step": 13650 }, { "grad_norm": 0.8518524765968323, "learning_rate": 9.606721962696272e-05, "loss": 0.0903, "step": 13660 }, { "grad_norm": 0.6941203474998474, "learning_rate": 9.60591809160354e-05, "loss": 0.0856, "step": 13670 }, { "grad_norm": 0.8370322585105896, "learning_rate": 9.605113433485378e-05, "loss": 0.0757, "step": 13680 }, { "grad_norm": 0.7501068115234375, "learning_rate": 9.604307988479279e-05, "loss": 0.078, "step": 13690 }, { "grad_norm": 0.6434438824653625, "learning_rate": 9.603501756722876e-05, "loss": 0.078, "step": 13700 }, { "grad_norm": 0.5806750059127808, "learning_rate": 9.602694738353927e-05, "loss": 0.0734, "step": 13710 }, { "grad_norm": 0.7243499159812927, "learning_rate": 9.601886933510331e-05, "loss": 0.0751, "step": 13720 }, { "grad_norm": 1.135637640953064, "learning_rate": 9.60107834233012e-05, "loss": 0.1091, "step": 13730 }, { "grad_norm": 0.7017584443092346, "learning_rate": 9.60026896495146e-05, "loss": 0.0911, "step": 13740 }, { "grad_norm": 0.670221745967865, "learning_rate": 9.599458801512652e-05, "loss": 0.0708, "step": 13750 }, { "grad_norm": 0.6104877591133118, "learning_rate": 9.598647852152129e-05, "loss": 0.0926, "step": 13760 }, { "grad_norm": 0.5687328577041626, "learning_rate": 9.597836117008462e-05, "loss": 0.0686, "step": 13770 }, { "grad_norm": 0.8421403765678406, "learning_rate": 9.597023596220356e-05, "loss": 0.0814, "step": 13780 }, { "grad_norm": 0.4438593089580536, "learning_rate": 9.596210289926643e-05, "loss": 0.0687, "step": 13790 }, { "grad_norm": 0.5733187794685364, "learning_rate": 9.5953961982663e-05, "loss": 0.0929, "step": 13800 }, { "grad_norm": 0.5304587483406067, "learning_rate": 9.594581321378431e-05, "loss": 0.0785, "step": 13810 }, { "grad_norm": 0.6502103805541992, "learning_rate": 9.593765659402276e-05, "loss": 0.0877, "step": 13820 }, { "grad_norm": 1.1100345849990845, "learning_rate": 9.59294921247721e-05, "loss": 0.0755, "step": 13830 }, { "grad_norm": 1.1977829933166504, "learning_rate": 9.59213198074274e-05, "loss": 0.0972, "step": 13840 }, { "grad_norm": 0.7946622371673584, "learning_rate": 9.59131396433851e-05, "loss": 0.0952, "step": 13850 }, { "grad_norm": 0.7967789769172668, "learning_rate": 9.590495163404297e-05, "loss": 0.1039, "step": 13860 }, { "grad_norm": 0.7585741281509399, "learning_rate": 9.589675578080009e-05, "loss": 0.0714, "step": 13870 }, { "grad_norm": 0.7570009827613831, "learning_rate": 9.588855208505694e-05, "loss": 0.0708, "step": 13880 }, { "grad_norm": 1.0319782495498657, "learning_rate": 9.588034054821529e-05, "loss": 0.0716, "step": 13890 }, { "grad_norm": 0.7002356052398682, "learning_rate": 9.587212117167826e-05, "loss": 0.0762, "step": 13900 }, { "grad_norm": 0.6630069613456726, "learning_rate": 9.586389395685033e-05, "loss": 0.0774, "step": 13910 }, { "grad_norm": 0.6402719020843506, "learning_rate": 9.585565890513733e-05, "loss": 0.0783, "step": 13920 }, { "grad_norm": 0.9756448268890381, "learning_rate": 9.584741601794636e-05, "loss": 0.0878, "step": 13930 }, { "grad_norm": 0.9329896569252014, "learning_rate": 9.58391652966859e-05, "loss": 0.0807, "step": 13940 }, { "grad_norm": 0.6754884719848633, "learning_rate": 9.583090674276583e-05, "loss": 0.0665, "step": 13950 }, { "grad_norm": 0.7546695470809937, "learning_rate": 9.582264035759726e-05, "loss": 0.1056, "step": 13960 }, { "grad_norm": 0.5454154014587402, "learning_rate": 9.58143661425927e-05, "loss": 0.0621, "step": 13970 }, { "grad_norm": 0.9440450072288513, "learning_rate": 9.580608409916601e-05, "loss": 0.0768, "step": 13980 }, { "grad_norm": 0.7118135094642639, "learning_rate": 9.579779422873233e-05, "loss": 0.0813, "step": 13990 }, { "grad_norm": 1.1806044578552246, "learning_rate": 9.578949653270819e-05, "loss": 0.0776, "step": 14000 }, { "grad_norm": 0.45402249693870544, "learning_rate": 9.578119101251144e-05, "loss": 0.0768, "step": 14010 }, { "grad_norm": 0.5206577181816101, "learning_rate": 9.577287766956127e-05, "loss": 0.0873, "step": 14020 }, { "grad_norm": 0.731568455696106, "learning_rate": 9.57645565052782e-05, "loss": 0.0763, "step": 14030 }, { "grad_norm": 0.7055201530456543, "learning_rate": 9.575622752108407e-05, "loss": 0.0778, "step": 14040 }, { "grad_norm": 0.6101962924003601, "learning_rate": 9.57478907184021e-05, "loss": 0.1162, "step": 14050 }, { "grad_norm": 0.7384747266769409, "learning_rate": 9.573954609865681e-05, "loss": 0.0872, "step": 14060 }, { "grad_norm": 0.8797985911369324, "learning_rate": 9.573119366327408e-05, "loss": 0.0812, "step": 14070 }, { "grad_norm": 0.7289772629737854, "learning_rate": 9.57228334136811e-05, "loss": 0.0867, "step": 14080 }, { "grad_norm": 0.7966951727867126, "learning_rate": 9.571446535130641e-05, "loss": 0.0755, "step": 14090 }, { "grad_norm": 0.8815261125564575, "learning_rate": 9.570608947757988e-05, "loss": 0.0774, "step": 14100 }, { "grad_norm": 0.7932762503623962, "learning_rate": 9.569770579393274e-05, "loss": 0.0785, "step": 14110 }, { "grad_norm": 0.5792227387428284, "learning_rate": 9.56893143017975e-05, "loss": 0.069, "step": 14120 }, { "grad_norm": 0.4850488603115082, "learning_rate": 9.568091500260806e-05, "loss": 0.0725, "step": 14130 }, { "grad_norm": 0.8078652620315552, "learning_rate": 9.567250789779961e-05, "loss": 0.0879, "step": 14140 }, { "grad_norm": 0.6485117673873901, "learning_rate": 9.566409298880872e-05, "loss": 0.0962, "step": 14150 }, { "grad_norm": 0.6813474297523499, "learning_rate": 9.565567027707326e-05, "loss": 0.1143, "step": 14160 }, { "grad_norm": 0.6466482877731323, "learning_rate": 9.56472397640324e-05, "loss": 0.0912, "step": 14170 }, { "grad_norm": 0.8076621294021606, "learning_rate": 9.563880145112675e-05, "loss": 0.0906, "step": 14180 }, { "grad_norm": 0.5676119327545166, "learning_rate": 9.563035533979814e-05, "loss": 0.0542, "step": 14190 }, { "grad_norm": 0.5986106991767883, "learning_rate": 9.562190143148981e-05, "loss": 0.0863, "step": 14200 }, { "grad_norm": 0.7050231695175171, "learning_rate": 9.561343972764627e-05, "loss": 0.1007, "step": 14210 }, { "grad_norm": 0.6827389597892761, "learning_rate": 9.560497022971343e-05, "loss": 0.0793, "step": 14220 }, { "grad_norm": 0.47096535563468933, "learning_rate": 9.559649293913847e-05, "loss": 0.073, "step": 14230 }, { "grad_norm": 0.6462118029594421, "learning_rate": 9.558800785736993e-05, "loss": 0.0792, "step": 14240 }, { "grad_norm": 0.8936969041824341, "learning_rate": 9.557951498585767e-05, "loss": 0.0946, "step": 14250 }, { "grad_norm": 0.6224512457847595, "learning_rate": 9.557101432605293e-05, "loss": 0.0795, "step": 14260 }, { "grad_norm": 0.8518701791763306, "learning_rate": 9.556250587940818e-05, "loss": 0.0955, "step": 14270 }, { "grad_norm": 0.5995000004768372, "learning_rate": 9.555398964737734e-05, "loss": 0.1029, "step": 14280 }, { "grad_norm": 0.8535947799682617, "learning_rate": 9.554546563141555e-05, "loss": 0.1019, "step": 14290 }, { "grad_norm": 0.7550173997879028, "learning_rate": 9.553693383297937e-05, "loss": 0.0863, "step": 14300 }, { "grad_norm": 0.5667080879211426, "learning_rate": 9.552839425352663e-05, "loss": 0.0854, "step": 14310 }, { "grad_norm": 0.9780660271644592, "learning_rate": 9.551984689451652e-05, "loss": 0.0878, "step": 14320 }, { "grad_norm": 1.0402464866638184, "learning_rate": 9.551129175740953e-05, "loss": 0.0922, "step": 14330 }, { "grad_norm": 0.7319589853286743, "learning_rate": 9.550272884366754e-05, "loss": 0.0686, "step": 14340 }, { "grad_norm": 0.6247016191482544, "learning_rate": 9.549415815475369e-05, "loss": 0.0794, "step": 14350 }, { "grad_norm": 0.6704735159873962, "learning_rate": 9.548557969213247e-05, "loss": 0.0742, "step": 14360 }, { "grad_norm": 0.8140187859535217, "learning_rate": 9.547699345726972e-05, "loss": 0.0729, "step": 14370 }, { "grad_norm": 1.0276079177856445, "learning_rate": 9.546839945163257e-05, "loss": 0.0981, "step": 14380 }, { "grad_norm": 0.7884044051170349, "learning_rate": 9.545979767668953e-05, "loss": 0.0747, "step": 14390 }, { "grad_norm": 0.8345138430595398, "learning_rate": 9.54511881339104e-05, "loss": 0.0792, "step": 14400 }, { "grad_norm": 0.6338111162185669, "learning_rate": 9.54425708247663e-05, "loss": 0.0572, "step": 14410 }, { "grad_norm": 0.5765150189399719, "learning_rate": 9.543394575072972e-05, "loss": 0.0961, "step": 14420 }, { "grad_norm": 0.727873682975769, "learning_rate": 9.542531291327441e-05, "loss": 0.0734, "step": 14430 }, { "grad_norm": 1.0524723529815674, "learning_rate": 9.541667231387552e-05, "loss": 0.0792, "step": 14440 }, { "grad_norm": 0.8421058654785156, "learning_rate": 9.540802395400949e-05, "loss": 0.0734, "step": 14450 }, { "grad_norm": 0.798353374004364, "learning_rate": 9.539936783515406e-05, "loss": 0.0839, "step": 14460 }, { "grad_norm": 0.770632803440094, "learning_rate": 9.539070395878835e-05, "loss": 0.0981, "step": 14470 }, { "grad_norm": 0.5439642071723938, "learning_rate": 9.538203232639277e-05, "loss": 0.0764, "step": 14480 }, { "grad_norm": 0.958779513835907, "learning_rate": 9.537335293944907e-05, "loss": 0.0907, "step": 14490 }, { "grad_norm": 0.7024015784263611, "learning_rate": 9.536466579944032e-05, "loss": 0.0823, "step": 14500 }, { "grad_norm": 0.6981980800628662, "learning_rate": 9.535597090785091e-05, "loss": 0.077, "step": 14510 }, { "grad_norm": 0.747570276260376, "learning_rate": 9.534726826616656e-05, "loss": 0.0726, "step": 14520 }, { "grad_norm": 0.5411996245384216, "learning_rate": 9.53385578758743e-05, "loss": 0.0639, "step": 14530 }, { "grad_norm": 0.33851343393325806, "learning_rate": 9.532983973846252e-05, "loss": 0.0619, "step": 14540 }, { "grad_norm": 0.6012274026870728, "learning_rate": 9.53211138554209e-05, "loss": 0.0652, "step": 14550 }, { "grad_norm": 0.7115229368209839, "learning_rate": 9.531238022824047e-05, "loss": 0.0753, "step": 14560 }, { "grad_norm": 0.6191178560256958, "learning_rate": 9.530363885841355e-05, "loss": 0.0763, "step": 14570 }, { "grad_norm": 0.6051804423332214, "learning_rate": 9.52948897474338e-05, "loss": 0.0801, "step": 14580 }, { "grad_norm": 0.6996073722839355, "learning_rate": 9.528613289679622e-05, "loss": 0.0607, "step": 14590 }, { "grad_norm": 0.7267373204231262, "learning_rate": 9.52773683079971e-05, "loss": 0.0896, "step": 14600 }, { "grad_norm": 0.6680135130882263, "learning_rate": 9.526859598253407e-05, "loss": 0.0871, "step": 14610 }, { "grad_norm": 0.6042401790618896, "learning_rate": 9.525981592190609e-05, "loss": 0.085, "step": 14620 }, { "grad_norm": 0.5203402638435364, "learning_rate": 9.525102812761342e-05, "loss": 0.0603, "step": 14630 }, { "grad_norm": 1.0228294134140015, "learning_rate": 9.524223260115768e-05, "loss": 0.0737, "step": 14640 }, { "grad_norm": 0.5539608001708984, "learning_rate": 9.523342934404175e-05, "loss": 0.0808, "step": 14650 }, { "grad_norm": 0.6142928004264832, "learning_rate": 9.522461835776989e-05, "loss": 0.0713, "step": 14660 }, { "grad_norm": 0.6928886771202087, "learning_rate": 9.521579964384764e-05, "loss": 0.0868, "step": 14670 }, { "grad_norm": 0.7587027549743652, "learning_rate": 9.52069732037819e-05, "loss": 0.0964, "step": 14680 }, { "grad_norm": 0.6360909938812256, "learning_rate": 9.519813903908083e-05, "loss": 0.0641, "step": 14690 }, { "grad_norm": 0.632002055644989, "learning_rate": 9.5189297151254e-05, "loss": 0.0709, "step": 14700 }, { "grad_norm": 0.5618894100189209, "learning_rate": 9.518044754181218e-05, "loss": 0.0627, "step": 14710 }, { "grad_norm": 0.7249349355697632, "learning_rate": 9.51715902122676e-05, "loss": 0.0692, "step": 14720 }, { "grad_norm": 0.6475852727890015, "learning_rate": 9.516272516413368e-05, "loss": 0.0692, "step": 14730 }, { "grad_norm": 0.6063041090965271, "learning_rate": 9.515385239892525e-05, "loss": 0.081, "step": 14740 }, { "grad_norm": 0.8426134586334229, "learning_rate": 9.514497191815839e-05, "loss": 0.0699, "step": 14750 }, { "grad_norm": 0.8328513503074646, "learning_rate": 9.513608372335055e-05, "loss": 0.0751, "step": 14760 }, { "grad_norm": 0.8343555331230164, "learning_rate": 9.512718781602045e-05, "loss": 0.0743, "step": 14770 }, { "grad_norm": 0.6872228384017944, "learning_rate": 9.511828419768823e-05, "loss": 0.0694, "step": 14780 }, { "grad_norm": 0.6885570287704468, "learning_rate": 9.510937286987521e-05, "loss": 0.0711, "step": 14790 }, { "grad_norm": 0.9486514329910278, "learning_rate": 9.510045383410408e-05, "loss": 0.0774, "step": 14800 }, { "grad_norm": 0.5434894561767578, "learning_rate": 9.509152709189892e-05, "loss": 0.0703, "step": 14810 }, { "grad_norm": 0.7242141962051392, "learning_rate": 9.508259264478504e-05, "loss": 0.0836, "step": 14820 }, { "grad_norm": 1.0026028156280518, "learning_rate": 9.507365049428909e-05, "loss": 0.0883, "step": 14830 }, { "grad_norm": 0.7863060832023621, "learning_rate": 9.506470064193902e-05, "loss": 0.0893, "step": 14840 }, { "grad_norm": 0.7985230684280396, "learning_rate": 9.505574308926414e-05, "loss": 0.0919, "step": 14850 }, { "grad_norm": 0.6662471890449524, "learning_rate": 9.504677783779505e-05, "loss": 0.08, "step": 14860 }, { "grad_norm": 0.8593547344207764, "learning_rate": 9.503780488906365e-05, "loss": 0.0704, "step": 14870 }, { "grad_norm": 0.842815101146698, "learning_rate": 9.502882424460319e-05, "loss": 0.0682, "step": 14880 }, { "grad_norm": 0.7733100652694702, "learning_rate": 9.501983590594821e-05, "loss": 0.0715, "step": 14890 }, { "grad_norm": 0.6326672434806824, "learning_rate": 9.501083987463455e-05, "loss": 0.0612, "step": 14900 }, { "grad_norm": 0.8505397439002991, "learning_rate": 9.500183615219942e-05, "loss": 0.0959, "step": 14910 }, { "grad_norm": 0.615458071231842, "learning_rate": 9.49928247401813e-05, "loss": 0.0867, "step": 14920 }, { "grad_norm": 0.7827132344245911, "learning_rate": 9.498380564011997e-05, "loss": 0.0956, "step": 14930 }, { "grad_norm": 0.9593434929847717, "learning_rate": 9.497477885355656e-05, "loss": 0.085, "step": 14940 }, { "grad_norm": 0.5956190824508667, "learning_rate": 9.496574438203353e-05, "loss": 0.0951, "step": 14950 }, { "grad_norm": 0.48094943165779114, "learning_rate": 9.495670222709459e-05, "loss": 0.0888, "step": 14960 }, { "grad_norm": 0.6410210728645325, "learning_rate": 9.494765239028483e-05, "loss": 0.0765, "step": 14970 }, { "grad_norm": 0.47315189242362976, "learning_rate": 9.493859487315057e-05, "loss": 0.0756, "step": 14980 }, { "grad_norm": 0.6514699459075928, "learning_rate": 9.492952967723953e-05, "loss": 0.0645, "step": 14990 }, { "grad_norm": 0.6883151531219482, "learning_rate": 9.492045680410068e-05, "loss": 0.0862, "step": 15000 }, { "grad_norm": 0.5902198553085327, "learning_rate": 9.491137625528436e-05, "loss": 0.0622, "step": 15010 }, { "grad_norm": 0.8874081969261169, "learning_rate": 9.490228803234215e-05, "loss": 0.0791, "step": 15020 }, { "grad_norm": 0.5284709334373474, "learning_rate": 9.489319213682701e-05, "loss": 0.0713, "step": 15030 }, { "grad_norm": 0.6734861731529236, "learning_rate": 9.488408857029316e-05, "loss": 0.0607, "step": 15040 }, { "grad_norm": 0.496870756149292, "learning_rate": 9.487497733429616e-05, "loss": 0.0602, "step": 15050 }, { "grad_norm": 0.5430782437324524, "learning_rate": 9.486585843039286e-05, "loss": 0.0707, "step": 15060 }, { "grad_norm": 0.8268352150917053, "learning_rate": 9.485673186014143e-05, "loss": 0.0612, "step": 15070 }, { "grad_norm": 0.7095949053764343, "learning_rate": 9.484759762510137e-05, "loss": 0.0739, "step": 15080 }, { "grad_norm": 0.644654393196106, "learning_rate": 9.483845572683346e-05, "loss": 0.0669, "step": 15090 }, { "grad_norm": 0.8168236613273621, "learning_rate": 9.48293061668998e-05, "loss": 0.0784, "step": 15100 }, { "grad_norm": 0.5246550440788269, "learning_rate": 9.48201489468638e-05, "loss": 0.0744, "step": 15110 }, { "grad_norm": 0.4817655384540558, "learning_rate": 9.481098406829016e-05, "loss": 0.0802, "step": 15120 }, { "grad_norm": 0.5443965196609497, "learning_rate": 9.480181153274495e-05, "loss": 0.0938, "step": 15130 }, { "grad_norm": 0.6018293499946594, "learning_rate": 9.479263134179548e-05, "loss": 0.0776, "step": 15140 }, { "grad_norm": 0.8605936765670776, "learning_rate": 9.478344349701039e-05, "loss": 0.0634, "step": 15150 }, { "grad_norm": 0.7707449197769165, "learning_rate": 9.477424799995964e-05, "loss": 0.0961, "step": 15160 }, { "grad_norm": 0.7584964632987976, "learning_rate": 9.476504485221448e-05, "loss": 0.0897, "step": 15170 }, { "grad_norm": 0.7077351808547974, "learning_rate": 9.475583405534748e-05, "loss": 0.0639, "step": 15180 }, { "grad_norm": 0.701171338558197, "learning_rate": 9.474661561093251e-05, "loss": 0.0928, "step": 15190 }, { "grad_norm": 0.4919525682926178, "learning_rate": 9.473738952054478e-05, "loss": 0.0728, "step": 15200 }, { "grad_norm": 0.6840315461158752, "learning_rate": 9.472815578576073e-05, "loss": 0.1003, "step": 15210 }, { "grad_norm": 0.6391716599464417, "learning_rate": 9.471891440815817e-05, "loss": 0.0607, "step": 15220 }, { "grad_norm": 0.7113944292068481, "learning_rate": 9.470966538931621e-05, "loss": 0.0944, "step": 15230 }, { "grad_norm": 3.007077693939209, "learning_rate": 9.470040873081525e-05, "loss": 0.067, "step": 15240 }, { "grad_norm": 0.7921971082687378, "learning_rate": 9.469114443423698e-05, "loss": 0.0891, "step": 15250 }, { "grad_norm": 0.9309539794921875, "learning_rate": 9.468187250116445e-05, "loss": 0.0642, "step": 15260 }, { "grad_norm": 0.5892938375473022, "learning_rate": 9.467259293318197e-05, "loss": 0.0617, "step": 15270 }, { "grad_norm": 0.6553553342819214, "learning_rate": 9.466330573187514e-05, "loss": 0.0864, "step": 15280 }, { "grad_norm": 0.4436100125312805, "learning_rate": 9.46540108988309e-05, "loss": 0.0638, "step": 15290 }, { "grad_norm": 0.6866828799247742, "learning_rate": 9.46447084356375e-05, "loss": 0.0683, "step": 15300 }, { "grad_norm": 0.6907343864440918, "learning_rate": 9.463539834388447e-05, "loss": 0.0661, "step": 15310 }, { "grad_norm": 0.5542011857032776, "learning_rate": 9.462608062516263e-05, "loss": 0.0693, "step": 15320 }, { "grad_norm": 0.5611538887023926, "learning_rate": 9.461675528106413e-05, "loss": 0.068, "step": 15330 }, { "grad_norm": 0.6375567317008972, "learning_rate": 9.460742231318244e-05, "loss": 0.065, "step": 15340 }, { "grad_norm": 1.428890347480774, "learning_rate": 9.459808172311229e-05, "loss": 0.0816, "step": 15350 }, { "grad_norm": 0.5993049740791321, "learning_rate": 9.458873351244972e-05, "loss": 0.0724, "step": 15360 }, { "grad_norm": 0.6403233408927917, "learning_rate": 9.457937768279211e-05, "loss": 0.0699, "step": 15370 }, { "grad_norm": 0.7608993649482727, "learning_rate": 9.45700142357381e-05, "loss": 0.0683, "step": 15380 }, { "grad_norm": 0.6542829871177673, "learning_rate": 9.456064317288765e-05, "loss": 0.0716, "step": 15390 }, { "grad_norm": 1.0628000497817993, "learning_rate": 9.455126449584201e-05, "loss": 0.0675, "step": 15400 }, { "grad_norm": 0.678351879119873, "learning_rate": 9.454187820620375e-05, "loss": 0.0671, "step": 15410 }, { "grad_norm": 0.6790980696678162, "learning_rate": 9.453248430557673e-05, "loss": 0.0988, "step": 15420 }, { "grad_norm": 0.5056710839271545, "learning_rate": 9.452308279556611e-05, "loss": 0.0746, "step": 15430 }, { "grad_norm": 0.5545510053634644, "learning_rate": 9.451367367777835e-05, "loss": 0.072, "step": 15440 }, { "grad_norm": 0.5873265862464905, "learning_rate": 9.450425695382122e-05, "loss": 0.0736, "step": 15450 }, { "grad_norm": 1.0247142314910889, "learning_rate": 9.449483262530375e-05, "loss": 0.083, "step": 15460 }, { "grad_norm": 0.8747562766075134, "learning_rate": 9.448540069383633e-05, "loss": 0.0713, "step": 15470 }, { "grad_norm": 0.6881130933761597, "learning_rate": 9.447596116103061e-05, "loss": 0.082, "step": 15480 }, { "grad_norm": 0.5861718058586121, "learning_rate": 9.446651402849955e-05, "loss": 0.0685, "step": 15490 }, { "grad_norm": 0.6231937408447266, "learning_rate": 9.44570592978574e-05, "loss": 0.0682, "step": 15500 }, { "grad_norm": 0.6045047640800476, "learning_rate": 9.444759697071972e-05, "loss": 0.0927, "step": 15510 }, { "grad_norm": 0.6832506060600281, "learning_rate": 9.443812704870336e-05, "loss": 0.077, "step": 15520 }, { "grad_norm": 0.7358971834182739, "learning_rate": 9.442864953342649e-05, "loss": 0.0809, "step": 15530 }, { "grad_norm": 0.5821779370307922, "learning_rate": 9.441916442650852e-05, "loss": 0.0771, "step": 15540 }, { "grad_norm": 0.7942155003547668, "learning_rate": 9.440967172957023e-05, "loss": 0.0714, "step": 15550 }, { "grad_norm": 0.7508469223976135, "learning_rate": 9.440017144423364e-05, "loss": 0.0909, "step": 15560 }, { "grad_norm": 0.6888314485549927, "learning_rate": 9.439066357212209e-05, "loss": 0.0726, "step": 15570 }, { "grad_norm": 0.7227113842964172, "learning_rate": 9.438114811486022e-05, "loss": 0.0887, "step": 15580 }, { "grad_norm": 0.9954414963722229, "learning_rate": 9.4371625074074e-05, "loss": 0.0766, "step": 15590 }, { "grad_norm": 0.5395188927650452, "learning_rate": 9.436209445139059e-05, "loss": 0.1052, "step": 15600 }, { "grad_norm": 0.8848686218261719, "learning_rate": 9.435255624843855e-05, "loss": 0.077, "step": 15610 }, { "grad_norm": 0.6394565105438232, "learning_rate": 9.43430104668477e-05, "loss": 0.0674, "step": 15620 }, { "grad_norm": 0.7391993403434753, "learning_rate": 9.433345710824914e-05, "loss": 0.0615, "step": 15630 }, { "grad_norm": 0.5996522903442383, "learning_rate": 9.432389617427529e-05, "loss": 0.0617, "step": 15640 }, { "grad_norm": 0.7100884318351746, "learning_rate": 9.431432766655984e-05, "loss": 0.0795, "step": 15650 }, { "grad_norm": 0.5532001256942749, "learning_rate": 9.430475158673778e-05, "loss": 0.0668, "step": 15660 }, { "grad_norm": 0.6939322352409363, "learning_rate": 9.429516793644542e-05, "loss": 0.0679, "step": 15670 }, { "grad_norm": 0.665383517742157, "learning_rate": 9.428557671732034e-05, "loss": 0.0734, "step": 15680 }, { "grad_norm": 0.6363072395324707, "learning_rate": 9.42759779310014e-05, "loss": 0.0792, "step": 15690 }, { "grad_norm": 0.6583414673805237, "learning_rate": 9.426637157912879e-05, "loss": 0.0738, "step": 15700 }, { "grad_norm": 0.6186015009880066, "learning_rate": 9.425675766334397e-05, "loss": 0.0902, "step": 15710 }, { "grad_norm": 0.6585507988929749, "learning_rate": 9.424713618528968e-05, "loss": 0.0739, "step": 15720 }, { "grad_norm": 0.6297289729118347, "learning_rate": 9.423750714661e-05, "loss": 0.0623, "step": 15730 }, { "grad_norm": 0.7380022406578064, "learning_rate": 9.422787054895022e-05, "loss": 0.0809, "step": 15740 }, { "grad_norm": 0.5366931557655334, "learning_rate": 9.4218226393957e-05, "loss": 0.0784, "step": 15750 }, { "grad_norm": 0.8004036545753479, "learning_rate": 9.420857468327828e-05, "loss": 0.064, "step": 15760 }, { "grad_norm": 0.4552052617073059, "learning_rate": 9.419891541856323e-05, "loss": 0.0586, "step": 15770 }, { "grad_norm": 0.45883774757385254, "learning_rate": 9.41892486014624e-05, "loss": 0.0755, "step": 15780 }, { "grad_norm": 0.5700733065605164, "learning_rate": 9.417957423362756e-05, "loss": 0.0817, "step": 15790 }, { "grad_norm": 0.6832616329193115, "learning_rate": 9.416989231671178e-05, "loss": 0.0836, "step": 15800 }, { "grad_norm": 0.6775122284889221, "learning_rate": 9.416020285236946e-05, "loss": 0.0861, "step": 15810 }, { "grad_norm": 0.624035656452179, "learning_rate": 9.415050584225626e-05, "loss": 0.0781, "step": 15820 }, { "grad_norm": 0.641247570514679, "learning_rate": 9.414080128802914e-05, "loss": 0.1041, "step": 15830 }, { "grad_norm": 0.663452684879303, "learning_rate": 9.413108919134632e-05, "loss": 0.1067, "step": 15840 }, { "grad_norm": 0.6143090724945068, "learning_rate": 9.412136955386734e-05, "loss": 0.0772, "step": 15850 }, { "grad_norm": 0.6110151410102844, "learning_rate": 9.411164237725303e-05, "loss": 0.0754, "step": 15860 }, { "grad_norm": 1.069148063659668, "learning_rate": 9.41019076631655e-05, "loss": 0.0889, "step": 15870 }, { "grad_norm": 1.334079623222351, "learning_rate": 9.409216541326815e-05, "loss": 0.0912, "step": 15880 }, { "grad_norm": 0.9490678906440735, "learning_rate": 9.408241562922564e-05, "loss": 0.0764, "step": 15890 }, { "grad_norm": 0.7688511610031128, "learning_rate": 9.407265831270395e-05, "loss": 0.0745, "step": 15900 }, { "grad_norm": 0.6292468309402466, "learning_rate": 9.406289346537035e-05, "loss": 0.0709, "step": 15910 }, { "grad_norm": 0.7593712210655212, "learning_rate": 9.405312108889339e-05, "loss": 0.0694, "step": 15920 }, { "grad_norm": 0.8257141709327698, "learning_rate": 9.404334118494288e-05, "loss": 0.0678, "step": 15930 }, { "grad_norm": 0.5935477614402771, "learning_rate": 9.403355375518995e-05, "loss": 0.0788, "step": 15940 }, { "grad_norm": 0.5769801735877991, "learning_rate": 9.4023758801307e-05, "loss": 0.0725, "step": 15950 }, { "grad_norm": 0.7526648640632629, "learning_rate": 9.401395632496774e-05, "loss": 0.0668, "step": 15960 }, { "grad_norm": 0.9936874508857727, "learning_rate": 9.400414632784711e-05, "loss": 0.0604, "step": 15970 }, { "grad_norm": 0.5707525610923767, "learning_rate": 9.39943288116214e-05, "loss": 0.074, "step": 15980 }, { "grad_norm": 0.6321322321891785, "learning_rate": 9.398450377796815e-05, "loss": 0.062, "step": 15990 }, { "grad_norm": 0.7826822996139526, "learning_rate": 9.397467122856616e-05, "loss": 0.0662, "step": 16000 }, { "grad_norm": 0.7763966917991638, "learning_rate": 9.396483116509558e-05, "loss": 0.077, "step": 16010 }, { "grad_norm": 0.536696195602417, "learning_rate": 9.39549835892378e-05, "loss": 0.0626, "step": 16020 }, { "grad_norm": 0.6743658185005188, "learning_rate": 9.39451285026755e-05, "loss": 0.0761, "step": 16030 }, { "grad_norm": 0.5012525320053101, "learning_rate": 9.393526590709262e-05, "loss": 0.0775, "step": 16040 }, { "grad_norm": 0.5693168044090271, "learning_rate": 9.392539580417444e-05, "loss": 0.0597, "step": 16050 }, { "grad_norm": 0.38974374532699585, "learning_rate": 9.391551819560747e-05, "loss": 0.0657, "step": 16060 }, { "grad_norm": 0.9330602288246155, "learning_rate": 9.390563308307955e-05, "loss": 0.0755, "step": 16070 }, { "grad_norm": 0.8787240982055664, "learning_rate": 9.389574046827974e-05, "loss": 0.0636, "step": 16080 }, { "grad_norm": 0.599503755569458, "learning_rate": 9.388584035289845e-05, "loss": 0.0617, "step": 16090 }, { "grad_norm": 0.7474706768989563, "learning_rate": 9.387593273862732e-05, "loss": 0.0731, "step": 16100 }, { "grad_norm": 0.6824368238449097, "learning_rate": 9.386601762715929e-05, "loss": 0.0667, "step": 16110 }, { "grad_norm": 0.6646455526351929, "learning_rate": 9.38560950201886e-05, "loss": 0.061, "step": 16120 }, { "grad_norm": 0.5987110137939453, "learning_rate": 9.384616491941071e-05, "loss": 0.0694, "step": 16130 }, { "grad_norm": 0.6277573704719543, "learning_rate": 9.383622732652245e-05, "loss": 0.0667, "step": 16140 }, { "grad_norm": 0.5630729794502258, "learning_rate": 9.382628224322187e-05, "loss": 0.0963, "step": 16150 }, { "grad_norm": 0.4159252345561981, "learning_rate": 9.381632967120829e-05, "loss": 0.0792, "step": 16160 }, { "grad_norm": 1.071775197982788, "learning_rate": 9.380636961218235e-05, "loss": 0.0772, "step": 16170 }, { "grad_norm": 0.8122494220733643, "learning_rate": 9.379640206784597e-05, "loss": 0.0752, "step": 16180 }, { "grad_norm": 0.6798321008682251, "learning_rate": 9.378642703990229e-05, "loss": 0.0882, "step": 16190 }, { "grad_norm": 0.6226637959480286, "learning_rate": 9.37764445300558e-05, "loss": 0.0731, "step": 16200 }, { "grad_norm": 0.46472635865211487, "learning_rate": 9.376645454001222e-05, "loss": 0.0905, "step": 16210 }, { "grad_norm": 0.8534212112426758, "learning_rate": 9.375645707147858e-05, "loss": 0.0698, "step": 16220 }, { "grad_norm": 0.6098300218582153, "learning_rate": 9.374645212616316e-05, "loss": 0.0793, "step": 16230 }, { "grad_norm": 0.76595538854599, "learning_rate": 9.373643970577555e-05, "loss": 0.0633, "step": 16240 }, { "grad_norm": 0.6432155966758728, "learning_rate": 9.372641981202659e-05, "loss": 0.0743, "step": 16250 }, { "grad_norm": 0.7093824148178101, "learning_rate": 9.37163924466284e-05, "loss": 0.1042, "step": 16260 }, { "grad_norm": 0.5749117732048035, "learning_rate": 9.370635761129438e-05, "loss": 0.0997, "step": 16270 }, { "grad_norm": 0.5865365862846375, "learning_rate": 9.36963153077392e-05, "loss": 0.0657, "step": 16280 }, { "grad_norm": 0.6819198131561279, "learning_rate": 9.368626553767888e-05, "loss": 0.074, "step": 16290 }, { "grad_norm": 0.7370285391807556, "learning_rate": 9.367620830283057e-05, "loss": 0.0832, "step": 16300 }, { "grad_norm": 0.5008688569068909, "learning_rate": 9.366614360491281e-05, "loss": 0.0979, "step": 16310 }, { "grad_norm": 0.5820056200027466, "learning_rate": 9.365607144564539e-05, "loss": 0.065, "step": 16320 }, { "grad_norm": 0.6934221386909485, "learning_rate": 9.364599182674934e-05, "loss": 0.0804, "step": 16330 }, { "grad_norm": 0.671903133392334, "learning_rate": 9.3635904749947e-05, "loss": 0.0709, "step": 16340 }, { "grad_norm": 1.0549120903015137, "learning_rate": 9.362581021696202e-05, "loss": 0.0744, "step": 16350 }, { "grad_norm": 0.47098883986473083, "learning_rate": 9.361570822951921e-05, "loss": 0.079, "step": 16360 }, { "grad_norm": 0.42835941910743713, "learning_rate": 9.360559878934476e-05, "loss": 0.0635, "step": 16370 }, { "grad_norm": 0.7006904482841492, "learning_rate": 9.359548189816611e-05, "loss": 0.0899, "step": 16380 }, { "grad_norm": 0.776009738445282, "learning_rate": 9.358535755771193e-05, "loss": 0.0727, "step": 16390 }, { "grad_norm": 0.6556136608123779, "learning_rate": 9.357522576971221e-05, "loss": 0.0744, "step": 16400 }, { "grad_norm": 0.7619614601135254, "learning_rate": 9.356508653589819e-05, "loss": 0.0767, "step": 16410 }, { "grad_norm": 0.6292936205863953, "learning_rate": 9.355493985800237e-05, "loss": 0.0596, "step": 16420 }, { "grad_norm": 0.6237704753875732, "learning_rate": 9.354478573775857e-05, "loss": 0.0779, "step": 16430 }, { "grad_norm": 0.589681088924408, "learning_rate": 9.353462417690186e-05, "loss": 0.0702, "step": 16440 }, { "grad_norm": 0.7242327928543091, "learning_rate": 9.352445517716853e-05, "loss": 0.071, "step": 16450 }, { "grad_norm": 0.6089664101600647, "learning_rate": 9.351427874029621e-05, "loss": 0.0625, "step": 16460 }, { "grad_norm": 0.600731611251831, "learning_rate": 9.350409486802379e-05, "loss": 0.0703, "step": 16470 }, { "grad_norm": 0.5495620369911194, "learning_rate": 9.349390356209138e-05, "loss": 0.063, "step": 16480 }, { "grad_norm": 0.5943922996520996, "learning_rate": 9.348370482424042e-05, "loss": 0.077, "step": 16490 }, { "grad_norm": 0.5901444554328918, "learning_rate": 9.347349865621357e-05, "loss": 0.0652, "step": 16500 }, { "grad_norm": 0.5697475671768188, "learning_rate": 9.346328505975481e-05, "loss": 0.0531, "step": 16510 }, { "grad_norm": 0.5590641498565674, "learning_rate": 9.345306403660936e-05, "loss": 0.056, "step": 16520 }, { "grad_norm": 0.5388385057449341, "learning_rate": 9.344283558852371e-05, "loss": 0.0648, "step": 16530 }, { "grad_norm": 1.274574637413025, "learning_rate": 9.343259971724563e-05, "loss": 0.0706, "step": 16540 }, { "grad_norm": 0.5543491840362549, "learning_rate": 9.342235642452413e-05, "loss": 0.0768, "step": 16550 }, { "grad_norm": 0.8544610142707825, "learning_rate": 9.341210571210954e-05, "loss": 0.0853, "step": 16560 }, { "grad_norm": 0.484580397605896, "learning_rate": 9.340184758175338e-05, "loss": 0.0505, "step": 16570 }, { "grad_norm": 0.5812819600105286, "learning_rate": 9.339158203520854e-05, "loss": 0.0576, "step": 16580 }, { "grad_norm": 0.4933837056159973, "learning_rate": 9.338130907422908e-05, "loss": 0.0787, "step": 16590 }, { "grad_norm": 0.5746811032295227, "learning_rate": 9.337102870057037e-05, "loss": 0.0791, "step": 16600 }, { "grad_norm": 0.5675414204597473, "learning_rate": 9.336074091598907e-05, "loss": 0.0619, "step": 16610 }, { "grad_norm": 0.8254327178001404, "learning_rate": 9.335044572224306e-05, "loss": 0.065, "step": 16620 }, { "grad_norm": 0.5763334631919861, "learning_rate": 9.334014312109151e-05, "loss": 0.076, "step": 16630 }, { "grad_norm": 0.5126423835754395, "learning_rate": 9.332983311429486e-05, "loss": 0.0543, "step": 16640 }, { "grad_norm": 0.5849089026451111, "learning_rate": 9.33195157036148e-05, "loss": 0.0673, "step": 16650 }, { "grad_norm": 0.5897670388221741, "learning_rate": 9.330919089081432e-05, "loss": 0.0724, "step": 16660 }, { "grad_norm": 0.4470441937446594, "learning_rate": 9.32988586776576e-05, "loss": 0.0648, "step": 16670 }, { "grad_norm": 0.566578209400177, "learning_rate": 9.328851906591016e-05, "loss": 0.0611, "step": 16680 }, { "grad_norm": 0.5331544876098633, "learning_rate": 9.327817205733875e-05, "loss": 0.0858, "step": 16690 }, { "grad_norm": 0.515958845615387, "learning_rate": 9.326781765371142e-05, "loss": 0.0589, "step": 16700 }, { "grad_norm": 0.7542361617088318, "learning_rate": 9.325745585679741e-05, "loss": 0.0641, "step": 16710 }, { "grad_norm": 0.6847577691078186, "learning_rate": 9.32470866683673e-05, "loss": 0.0607, "step": 16720 }, { "grad_norm": 0.5219243764877319, "learning_rate": 9.323671009019288e-05, "loss": 0.0626, "step": 16730 }, { "grad_norm": 0.5405434966087341, "learning_rate": 9.322632612404725e-05, "loss": 0.0681, "step": 16740 }, { "grad_norm": 0.39269521832466125, "learning_rate": 9.321593477170471e-05, "loss": 0.0622, "step": 16750 }, { "grad_norm": 0.6315184831619263, "learning_rate": 9.320553603494088e-05, "loss": 0.0654, "step": 16760 }, { "grad_norm": 0.8724428415298462, "learning_rate": 9.319512991553261e-05, "loss": 0.0639, "step": 16770 }, { "grad_norm": 0.7898690104484558, "learning_rate": 9.318471641525803e-05, "loss": 0.0712, "step": 16780 }, { "grad_norm": 0.5399903655052185, "learning_rate": 9.317429553589652e-05, "loss": 0.0677, "step": 16790 }, { "grad_norm": 0.5564744472503662, "learning_rate": 9.316386727922873e-05, "loss": 0.0631, "step": 16800 }, { "grad_norm": 0.7935240268707275, "learning_rate": 9.315343164703656e-05, "loss": 0.0774, "step": 16810 }, { "grad_norm": 0.689720094203949, "learning_rate": 9.314298864110316e-05, "loss": 0.0556, "step": 16820 }, { "grad_norm": 0.6161953210830688, "learning_rate": 9.313253826321295e-05, "loss": 0.0547, "step": 16830 }, { "grad_norm": 0.702714204788208, "learning_rate": 9.312208051515165e-05, "loss": 0.0615, "step": 16840 }, { "grad_norm": 0.6338838934898376, "learning_rate": 9.311161539870618e-05, "loss": 0.0563, "step": 16850 }, { "grad_norm": 0.7200518846511841, "learning_rate": 9.310114291566474e-05, "loss": 0.0665, "step": 16860 }, { "grad_norm": 0.579155445098877, "learning_rate": 9.309066306781679e-05, "loss": 0.0713, "step": 16870 }, { "grad_norm": 0.7300761342048645, "learning_rate": 9.308017585695306e-05, "loss": 0.0657, "step": 16880 }, { "grad_norm": 0.5957217812538147, "learning_rate": 9.306968128486552e-05, "loss": 0.0874, "step": 16890 }, { "grad_norm": 0.8344043493270874, "learning_rate": 9.30591793533474e-05, "loss": 0.0634, "step": 16900 }, { "grad_norm": 0.7404932379722595, "learning_rate": 9.304867006419321e-05, "loss": 0.0778, "step": 16910 }, { "grad_norm": 0.9957110285758972, "learning_rate": 9.303815341919868e-05, "loss": 0.0642, "step": 16920 }, { "grad_norm": 0.6837548017501831, "learning_rate": 9.302762942016084e-05, "loss": 0.0577, "step": 16930 }, { "grad_norm": 0.5460843443870544, "learning_rate": 9.301709806887792e-05, "loss": 0.0791, "step": 16940 }, { "grad_norm": 0.7579642534255981, "learning_rate": 9.300655936714948e-05, "loss": 0.076, "step": 16950 }, { "grad_norm": 0.7987999320030212, "learning_rate": 9.299601331677627e-05, "loss": 0.078, "step": 16960 }, { "grad_norm": 0.7193232178688049, "learning_rate": 9.298545991956033e-05, "loss": 0.0738, "step": 16970 }, { "grad_norm": 1.0515720844268799, "learning_rate": 9.297489917730493e-05, "loss": 0.0814, "step": 16980 }, { "grad_norm": 0.7081000804901123, "learning_rate": 9.296433109181464e-05, "loss": 0.0876, "step": 16990 }, { "grad_norm": 0.6570872664451599, "learning_rate": 9.295375566489523e-05, "loss": 0.081, "step": 17000 }, { "grad_norm": 0.7327709794044495, "learning_rate": 9.294317289835379e-05, "loss": 0.0635, "step": 17010 }, { "grad_norm": 0.5996992588043213, "learning_rate": 9.293258279399859e-05, "loss": 0.0635, "step": 17020 }, { "grad_norm": 0.4687156677246094, "learning_rate": 9.292198535363919e-05, "loss": 0.0548, "step": 17030 }, { "grad_norm": 0.518194854259491, "learning_rate": 9.291138057908641e-05, "loss": 0.0582, "step": 17040 }, { "grad_norm": 0.6312122941017151, "learning_rate": 9.290076847215234e-05, "loss": 0.0548, "step": 17050 }, { "grad_norm": 0.6033991575241089, "learning_rate": 9.289014903465025e-05, "loss": 0.0711, "step": 17060 }, { "grad_norm": 0.6626973748207092, "learning_rate": 9.287952226839475e-05, "loss": 0.0668, "step": 17070 }, { "grad_norm": 0.6841105818748474, "learning_rate": 9.286888817520164e-05, "loss": 0.072, "step": 17080 }, { "grad_norm": 0.7762273550033569, "learning_rate": 9.285824675688803e-05, "loss": 0.0792, "step": 17090 }, { "grad_norm": 0.599150538444519, "learning_rate": 9.28475980152722e-05, "loss": 0.0882, "step": 17100 }, { "grad_norm": 0.5428541302680969, "learning_rate": 9.283694195217379e-05, "loss": 0.0788, "step": 17110 }, { "grad_norm": 1.3617397546768188, "learning_rate": 9.282627856941356e-05, "loss": 0.0927, "step": 17120 }, { "grad_norm": 0.7251827120780945, "learning_rate": 9.281560786881363e-05, "loss": 0.1399, "step": 17130 }, { "grad_norm": 0.7499848008155823, "learning_rate": 9.280492985219733e-05, "loss": 0.0919, "step": 17140 }, { "grad_norm": 0.5669058561325073, "learning_rate": 9.279424452138924e-05, "loss": 0.0708, "step": 17150 }, { "grad_norm": 0.7033649682998657, "learning_rate": 9.278355187821517e-05, "loss": 0.0917, "step": 17160 }, { "grad_norm": 0.7719815373420715, "learning_rate": 9.277285192450224e-05, "loss": 0.0803, "step": 17170 }, { "grad_norm": 0.6424931287765503, "learning_rate": 9.276214466207875e-05, "loss": 0.0695, "step": 17180 }, { "grad_norm": 0.7072197198867798, "learning_rate": 9.275143009277427e-05, "loss": 0.0535, "step": 17190 }, { "grad_norm": 0.5419362187385559, "learning_rate": 9.274070821841964e-05, "loss": 0.0661, "step": 17200 }, { "grad_norm": 0.5057399868965149, "learning_rate": 9.272997904084696e-05, "loss": 0.0538, "step": 17210 }, { "grad_norm": 0.6834180355072021, "learning_rate": 9.271924256188951e-05, "loss": 0.0652, "step": 17220 }, { "grad_norm": 0.6254364848136902, "learning_rate": 9.270849878338189e-05, "loss": 0.0695, "step": 17230 }, { "grad_norm": 0.727160632610321, "learning_rate": 9.269774770715991e-05, "loss": 0.0821, "step": 17240 }, { "grad_norm": 0.6479933261871338, "learning_rate": 9.268698933506061e-05, "loss": 0.0569, "step": 17250 }, { "grad_norm": 1.3039463758468628, "learning_rate": 9.267622366892235e-05, "loss": 0.0741, "step": 17260 }, { "grad_norm": 0.48201149702072144, "learning_rate": 9.266545071058465e-05, "loss": 0.0611, "step": 17270 }, { "grad_norm": 0.7526757121086121, "learning_rate": 9.265467046188833e-05, "loss": 0.0805, "step": 17280 }, { "grad_norm": 0.49889928102493286, "learning_rate": 9.264388292467543e-05, "loss": 0.0745, "step": 17290 }, { "grad_norm": 0.4950762093067169, "learning_rate": 9.263308810078926e-05, "loss": 0.0841, "step": 17300 }, { "grad_norm": 0.9587177038192749, "learning_rate": 9.262228599207434e-05, "loss": 0.074, "step": 17310 }, { "grad_norm": 0.5123556852340698, "learning_rate": 9.261147660037647e-05, "loss": 0.0734, "step": 17320 }, { "grad_norm": 0.8310051560401917, "learning_rate": 9.26006599275427e-05, "loss": 0.0593, "step": 17330 }, { "grad_norm": 0.6417874693870544, "learning_rate": 9.258983597542124e-05, "loss": 0.0584, "step": 17340 }, { "grad_norm": 0.5956642031669617, "learning_rate": 9.257900474586167e-05, "loss": 0.0618, "step": 17350 }, { "grad_norm": 0.6503071188926697, "learning_rate": 9.256816624071471e-05, "loss": 0.0706, "step": 17360 }, { "grad_norm": 0.7910024523735046, "learning_rate": 9.25573204618324e-05, "loss": 0.0752, "step": 17370 }, { "grad_norm": 0.9835769534111023, "learning_rate": 9.254646741106796e-05, "loss": 0.072, "step": 17380 }, { "grad_norm": 0.6603435277938843, "learning_rate": 9.253560709027589e-05, "loss": 0.0685, "step": 17390 }, { "grad_norm": 0.48644357919692993, "learning_rate": 9.252473950131192e-05, "loss": 0.0673, "step": 17400 }, { "grad_norm": 0.7426921129226685, "learning_rate": 9.251386464603302e-05, "loss": 0.0779, "step": 17410 }, { "grad_norm": 0.7397546172142029, "learning_rate": 9.250298252629741e-05, "loss": 0.0808, "step": 17420 }, { "grad_norm": 0.5007892847061157, "learning_rate": 9.249209314396454e-05, "loss": 0.073, "step": 17430 }, { "grad_norm": 0.7987591624259949, "learning_rate": 9.248119650089513e-05, "loss": 0.1028, "step": 17440 }, { "grad_norm": 1.100940465927124, "learning_rate": 9.247029259895108e-05, "loss": 0.0699, "step": 17450 }, { "grad_norm": 0.9373079538345337, "learning_rate": 9.24593814399956e-05, "loss": 0.0637, "step": 17460 }, { "grad_norm": 0.5014302134513855, "learning_rate": 9.244846302589309e-05, "loss": 0.0669, "step": 17470 }, { "grad_norm": 0.6397873759269714, "learning_rate": 9.243753735850923e-05, "loss": 0.0665, "step": 17480 }, { "grad_norm": 0.6034424901008606, "learning_rate": 9.24266044397109e-05, "loss": 0.0649, "step": 17490 }, { "grad_norm": 0.44877129793167114, "learning_rate": 9.241566427136624e-05, "loss": 0.0799, "step": 17500 }, { "grad_norm": 0.8321662545204163, "learning_rate": 9.240471685534463e-05, "loss": 0.0586, "step": 17510 }, { "grad_norm": 0.6019365787506104, "learning_rate": 9.239376219351667e-05, "loss": 0.0712, "step": 17520 }, { "grad_norm": 0.7574284672737122, "learning_rate": 9.238280028775425e-05, "loss": 0.063, "step": 17530 }, { "grad_norm": 0.6610057353973389, "learning_rate": 9.237183113993041e-05, "loss": 0.0656, "step": 17540 }, { "grad_norm": 0.7484449148178101, "learning_rate": 9.236085475191952e-05, "loss": 0.0578, "step": 17550 }, { "grad_norm": 0.7416619062423706, "learning_rate": 9.234987112559709e-05, "loss": 0.0989, "step": 17560 }, { "grad_norm": 0.54360032081604, "learning_rate": 9.233888026283999e-05, "loss": 0.0663, "step": 17570 }, { "grad_norm": 0.5185120701789856, "learning_rate": 9.232788216552619e-05, "loss": 0.0585, "step": 17580 }, { "grad_norm": 1.0894781351089478, "learning_rate": 9.231687683553502e-05, "loss": 0.0636, "step": 17590 }, { "grad_norm": 0.5332756042480469, "learning_rate": 9.230586427474698e-05, "loss": 0.0504, "step": 17600 }, { "grad_norm": 0.6351644396781921, "learning_rate": 9.229484448504379e-05, "loss": 0.0774, "step": 17610 }, { "grad_norm": 0.5695013999938965, "learning_rate": 9.228381746830843e-05, "loss": 0.0716, "step": 17620 }, { "grad_norm": 0.5249639749526978, "learning_rate": 9.227278322642514e-05, "loss": 0.0651, "step": 17630 }, { "grad_norm": 0.736652672290802, "learning_rate": 9.226174176127937e-05, "loss": 0.0755, "step": 17640 }, { "grad_norm": 0.7108548283576965, "learning_rate": 9.22506930747578e-05, "loss": 0.0711, "step": 17650 }, { "grad_norm": 0.526066780090332, "learning_rate": 9.223963716874831e-05, "loss": 0.0649, "step": 17660 }, { "grad_norm": 0.6391469836235046, "learning_rate": 9.222857404514012e-05, "loss": 0.0611, "step": 17670 }, { "grad_norm": 0.8302865028381348, "learning_rate": 9.221750370582355e-05, "loss": 0.0627, "step": 17680 }, { "grad_norm": 0.5381010174751282, "learning_rate": 9.220642615269028e-05, "loss": 0.0542, "step": 17690 }, { "grad_norm": 0.557544469833374, "learning_rate": 9.219534138763311e-05, "loss": 0.0557, "step": 17700 }, { "grad_norm": 0.6720629334449768, "learning_rate": 9.218424941254613e-05, "loss": 0.0651, "step": 17710 }, { "grad_norm": 0.5908384919166565, "learning_rate": 9.217315022932468e-05, "loss": 0.0713, "step": 17720 }, { "grad_norm": 0.6481565237045288, "learning_rate": 9.216204383986528e-05, "loss": 0.0648, "step": 17730 }, { "grad_norm": 0.8378984928131104, "learning_rate": 9.215093024606574e-05, "loss": 0.0832, "step": 17740 }, { "grad_norm": 0.559981644153595, "learning_rate": 9.213980944982506e-05, "loss": 0.0808, "step": 17750 }, { "grad_norm": 0.8135169744491577, "learning_rate": 9.212868145304346e-05, "loss": 0.0653, "step": 17760 }, { "grad_norm": 0.6312233805656433, "learning_rate": 9.211754625762241e-05, "loss": 0.0844, "step": 17770 }, { "grad_norm": 0.41362157464027405, "learning_rate": 9.210640386546463e-05, "loss": 0.051, "step": 17780 }, { "grad_norm": 0.6621851325035095, "learning_rate": 9.209525427847405e-05, "loss": 0.0716, "step": 17790 }, { "grad_norm": 0.4675804376602173, "learning_rate": 9.208409749855583e-05, "loss": 0.0676, "step": 17800 }, { "grad_norm": 0.7536871433258057, "learning_rate": 9.207293352761633e-05, "loss": 0.058, "step": 17810 }, { "grad_norm": 0.71511310338974, "learning_rate": 9.206176236756319e-05, "loss": 0.0614, "step": 17820 }, { "grad_norm": 0.8188609480857849, "learning_rate": 9.205058402030525e-05, "loss": 0.0773, "step": 17830 }, { "grad_norm": 0.6178667545318604, "learning_rate": 9.203939848775259e-05, "loss": 0.0783, "step": 17840 }, { "grad_norm": 0.6935093998908997, "learning_rate": 9.202820577181652e-05, "loss": 0.0645, "step": 17850 }, { "grad_norm": 0.6644685864448547, "learning_rate": 9.201700587440953e-05, "loss": 0.063, "step": 17860 }, { "grad_norm": 0.5330386161804199, "learning_rate": 9.200579879744544e-05, "loss": 0.059, "step": 17870 }, { "grad_norm": 1.0231297016143799, "learning_rate": 9.199458454283918e-05, "loss": 0.0777, "step": 17880 }, { "grad_norm": 0.5582821369171143, "learning_rate": 9.198336311250697e-05, "loss": 0.0628, "step": 17890 }, { "grad_norm": 0.488650918006897, "learning_rate": 9.197213450836626e-05, "loss": 0.0556, "step": 17900 }, { "grad_norm": 0.43195268511772156, "learning_rate": 9.19608987323357e-05, "loss": 0.0498, "step": 17910 }, { "grad_norm": 0.6495391130447388, "learning_rate": 9.194965578633517e-05, "loss": 0.0758, "step": 17920 }, { "grad_norm": 0.3986833095550537, "learning_rate": 9.193840567228582e-05, "loss": 0.0666, "step": 17930 }, { "grad_norm": 0.5848480463027954, "learning_rate": 9.192714839210994e-05, "loss": 0.0562, "step": 17940 }, { "grad_norm": 0.6070983409881592, "learning_rate": 9.19158839477311e-05, "loss": 0.0726, "step": 17950 }, { "grad_norm": 0.5679516196250916, "learning_rate": 9.190461234107411e-05, "loss": 0.058, "step": 17960 }, { "grad_norm": 0.5620606541633606, "learning_rate": 9.189333357406496e-05, "loss": 0.0858, "step": 17970 }, { "grad_norm": 0.5182449221611023, "learning_rate": 9.188204764863089e-05, "loss": 0.0944, "step": 17980 }, { "grad_norm": 0.7627137899398804, "learning_rate": 9.187075456670033e-05, "loss": 0.0738, "step": 17990 }, { "grad_norm": 0.5172449946403503, "learning_rate": 9.1859454330203e-05, "loss": 0.0643, "step": 18000 }, { "grad_norm": 0.6183227896690369, "learning_rate": 9.18481469410698e-05, "loss": 0.0636, "step": 18010 }, { "grad_norm": 0.7873265743255615, "learning_rate": 9.183683240123281e-05, "loss": 0.0659, "step": 18020 }, { "grad_norm": 0.502373218536377, "learning_rate": 9.182551071262541e-05, "loss": 0.0629, "step": 18030 }, { "grad_norm": 0.8073650598526001, "learning_rate": 9.181418187718218e-05, "loss": 0.0928, "step": 18040 }, { "grad_norm": 0.6816155910491943, "learning_rate": 9.180284589683888e-05, "loss": 0.0814, "step": 18050 }, { "grad_norm": 0.6740767359733582, "learning_rate": 9.17915027735325e-05, "loss": 0.0849, "step": 18060 }, { "grad_norm": 0.612311601638794, "learning_rate": 9.178015250920133e-05, "loss": 0.0687, "step": 18070 }, { "grad_norm": 0.588982105255127, "learning_rate": 9.176879510578477e-05, "loss": 0.0735, "step": 18080 }, { "grad_norm": 0.6549825072288513, "learning_rate": 9.17574305652235e-05, "loss": 0.0671, "step": 18090 }, { "grad_norm": 0.8065651655197144, "learning_rate": 9.174605888945942e-05, "loss": 0.0554, "step": 18100 }, { "grad_norm": 0.6746642589569092, "learning_rate": 9.173468008043564e-05, "loss": 0.075, "step": 18110 }, { "grad_norm": 1.205604076385498, "learning_rate": 9.172329414009648e-05, "loss": 0.0716, "step": 18120 }, { "grad_norm": 0.3789193630218506, "learning_rate": 9.171190107038747e-05, "loss": 0.0553, "step": 18130 }, { "grad_norm": 0.548145055770874, "learning_rate": 9.170050087325541e-05, "loss": 0.0629, "step": 18140 }, { "grad_norm": 0.49613240361213684, "learning_rate": 9.168909355064824e-05, "loss": 0.0645, "step": 18150 }, { "grad_norm": 0.4528392553329468, "learning_rate": 9.167767910451519e-05, "loss": 0.0557, "step": 18160 }, { "grad_norm": 0.5489740371704102, "learning_rate": 9.166625753680669e-05, "loss": 0.0769, "step": 18170 }, { "grad_norm": 0.46654003858566284, "learning_rate": 9.165482884947431e-05, "loss": 0.0584, "step": 18180 }, { "grad_norm": 0.7458412647247314, "learning_rate": 9.164339304447098e-05, "loss": 0.0847, "step": 18190 }, { "grad_norm": 0.7647210955619812, "learning_rate": 9.163195012375072e-05, "loss": 0.0724, "step": 18200 }, { "grad_norm": 0.5673498511314392, "learning_rate": 9.16205000892688e-05, "loss": 0.07, "step": 18210 }, { "grad_norm": 0.581778347492218, "learning_rate": 9.160904294298175e-05, "loss": 0.0746, "step": 18220 }, { "grad_norm": 0.5468534827232361, "learning_rate": 9.159757868684727e-05, "loss": 0.0644, "step": 18230 }, { "grad_norm": 0.6522334814071655, "learning_rate": 9.15861073228243e-05, "loss": 0.073, "step": 18240 }, { "grad_norm": 0.7419313192367554, "learning_rate": 9.157462885287296e-05, "loss": 0.0784, "step": 18250 }, { "grad_norm": 1.3909600973129272, "learning_rate": 9.156314327895461e-05, "loss": 0.0715, "step": 18260 }, { "grad_norm": 0.47850123047828674, "learning_rate": 9.155165060303185e-05, "loss": 0.057, "step": 18270 }, { "grad_norm": 0.9331872463226318, "learning_rate": 9.154015082706841e-05, "loss": 0.0719, "step": 18280 }, { "grad_norm": 0.5866044759750366, "learning_rate": 9.152864395302936e-05, "loss": 0.1, "step": 18290 }, { "grad_norm": 0.4155329167842865, "learning_rate": 9.151712998288085e-05, "loss": 0.0614, "step": 18300 }, { "grad_norm": 0.5615982413291931, "learning_rate": 9.150560891859031e-05, "loss": 0.0765, "step": 18310 }, { "grad_norm": 0.5436238050460815, "learning_rate": 9.14940807621264e-05, "loss": 0.0742, "step": 18320 }, { "grad_norm": 0.8014198541641235, "learning_rate": 9.148254551545894e-05, "loss": 0.0753, "step": 18330 }, { "grad_norm": 0.892208456993103, "learning_rate": 9.147100318055901e-05, "loss": 0.0655, "step": 18340 }, { "grad_norm": 0.9294843077659607, "learning_rate": 9.145945375939888e-05, "loss": 0.0667, "step": 18350 }, { "grad_norm": 0.8562403917312622, "learning_rate": 9.144789725395203e-05, "loss": 0.068, "step": 18360 }, { "grad_norm": 0.5646378993988037, "learning_rate": 9.14363336661931e-05, "loss": 0.057, "step": 18370 }, { "grad_norm": 0.5166459679603577, "learning_rate": 9.142476299809806e-05, "loss": 0.0457, "step": 18380 }, { "grad_norm": 0.48105689883232117, "learning_rate": 9.1413185251644e-05, "loss": 0.0656, "step": 18390 }, { "grad_norm": 0.5014415979385376, "learning_rate": 9.140160042880923e-05, "loss": 0.0615, "step": 18400 }, { "grad_norm": 0.7565469145774841, "learning_rate": 9.139000853157327e-05, "loss": 0.0612, "step": 18410 }, { "grad_norm": 1.0423513650894165, "learning_rate": 9.137840956191688e-05, "loss": 0.0904, "step": 18420 }, { "grad_norm": 0.6839625239372253, "learning_rate": 9.136680352182199e-05, "loss": 0.0724, "step": 18430 }, { "grad_norm": 0.4098643958568573, "learning_rate": 9.135519041327177e-05, "loss": 0.0664, "step": 18440 }, { "grad_norm": 0.6941704154014587, "learning_rate": 9.134357023825058e-05, "loss": 0.0711, "step": 18450 }, { "grad_norm": 0.639042317867279, "learning_rate": 9.133194299874398e-05, "loss": 0.0654, "step": 18460 }, { "grad_norm": 0.6694461703300476, "learning_rate": 9.132030869673876e-05, "loss": 0.0554, "step": 18470 }, { "grad_norm": 0.5214896202087402, "learning_rate": 9.130866733422288e-05, "loss": 0.09, "step": 18480 }, { "grad_norm": 0.5501320362091064, "learning_rate": 9.129701891318556e-05, "loss": 0.05, "step": 18490 }, { "grad_norm": 1.1004339456558228, "learning_rate": 9.128536343561718e-05, "loss": 0.0718, "step": 18500 }, { "grad_norm": 0.6503961682319641, "learning_rate": 9.127370090350934e-05, "loss": 0.0649, "step": 18510 }, { "grad_norm": 0.6239789724349976, "learning_rate": 9.126203131885487e-05, "loss": 0.0526, "step": 18520 }, { "grad_norm": 0.5334919095039368, "learning_rate": 9.125035468364775e-05, "loss": 0.0528, "step": 18530 }, { "grad_norm": 1.6336486339569092, "learning_rate": 9.123867099988322e-05, "loss": 0.0691, "step": 18540 }, { "grad_norm": 0.6579279899597168, "learning_rate": 9.122698026955769e-05, "loss": 0.0607, "step": 18550 }, { "grad_norm": 0.6652407646179199, "learning_rate": 9.12152824946688e-05, "loss": 0.0719, "step": 18560 }, { "grad_norm": 0.9226453900337219, "learning_rate": 9.120357767721538e-05, "loss": 0.0774, "step": 18570 }, { "grad_norm": 1.0385191440582275, "learning_rate": 9.119186581919745e-05, "loss": 0.0811, "step": 18580 }, { "grad_norm": 0.5616375207901001, "learning_rate": 9.118014692261624e-05, "loss": 0.0624, "step": 18590 }, { "grad_norm": 0.6352896690368652, "learning_rate": 9.116842098947422e-05, "loss": 0.0873, "step": 18600 }, { "grad_norm": 0.5591168999671936, "learning_rate": 9.115668802177499e-05, "loss": 0.0603, "step": 18610 }, { "grad_norm": 0.689087450504303, "learning_rate": 9.114494802152342e-05, "loss": 0.0724, "step": 18620 }, { "grad_norm": 0.5710930824279785, "learning_rate": 9.113320099072555e-05, "loss": 0.0867, "step": 18630 }, { "grad_norm": 0.9113106727600098, "learning_rate": 9.112144693138864e-05, "loss": 0.0912, "step": 18640 }, { "grad_norm": 0.617483913898468, "learning_rate": 9.110968584552111e-05, "loss": 0.0635, "step": 18650 }, { "grad_norm": 0.5129151344299316, "learning_rate": 9.109791773513264e-05, "loss": 0.0651, "step": 18660 }, { "grad_norm": 0.6729016900062561, "learning_rate": 9.108614260223403e-05, "loss": 0.0728, "step": 18670 }, { "grad_norm": 0.7316334843635559, "learning_rate": 9.107436044883738e-05, "loss": 0.0723, "step": 18680 }, { "grad_norm": 0.551874577999115, "learning_rate": 9.10625712769559e-05, "loss": 0.0757, "step": 18690 }, { "grad_norm": 0.532172441482544, "learning_rate": 9.105077508860406e-05, "loss": 0.0811, "step": 18700 }, { "grad_norm": 0.7629595994949341, "learning_rate": 9.103897188579751e-05, "loss": 0.0602, "step": 18710 }, { "grad_norm": 1.0134624242782593, "learning_rate": 9.102716167055308e-05, "loss": 0.0727, "step": 18720 }, { "grad_norm": 0.789374589920044, "learning_rate": 9.10153444448888e-05, "loss": 0.0716, "step": 18730 }, { "grad_norm": 0.5738869905471802, "learning_rate": 9.100352021082393e-05, "loss": 0.0841, "step": 18740 }, { "grad_norm": 0.5561072826385498, "learning_rate": 9.099168897037891e-05, "loss": 0.0813, "step": 18750 }, { "grad_norm": 0.4836503565311432, "learning_rate": 9.097985072557538e-05, "loss": 0.058, "step": 18760 }, { "grad_norm": 0.8240982890129089, "learning_rate": 9.096800547843615e-05, "loss": 0.0785, "step": 18770 }, { "grad_norm": 0.6446333527565002, "learning_rate": 9.095615323098526e-05, "loss": 0.0663, "step": 18780 }, { "grad_norm": 0.7202029228210449, "learning_rate": 9.094429398524795e-05, "loss": 0.0825, "step": 18790 }, { "grad_norm": 0.5976616740226746, "learning_rate": 9.093242774325061e-05, "loss": 0.0866, "step": 18800 }, { "grad_norm": 0.7401134967803955, "learning_rate": 9.092055450702088e-05, "loss": 0.0765, "step": 18810 }, { "grad_norm": 0.8090812563896179, "learning_rate": 9.090867427858756e-05, "loss": 0.0641, "step": 18820 }, { "grad_norm": 0.6402300000190735, "learning_rate": 9.089678705998066e-05, "loss": 0.0561, "step": 18830 }, { "grad_norm": 1.008621335029602, "learning_rate": 9.088489285323139e-05, "loss": 0.0738, "step": 18840 }, { "grad_norm": 0.6189914345741272, "learning_rate": 9.087299166037212e-05, "loss": 0.0563, "step": 18850 }, { "grad_norm": 0.613132894039154, "learning_rate": 9.086108348343647e-05, "loss": 0.0649, "step": 18860 }, { "grad_norm": 0.7253468632698059, "learning_rate": 9.08491683244592e-05, "loss": 0.0688, "step": 18870 }, { "grad_norm": 0.6816182732582092, "learning_rate": 9.08372461854763e-05, "loss": 0.0641, "step": 18880 }, { "grad_norm": 0.715958833694458, "learning_rate": 9.082531706852492e-05, "loss": 0.0695, "step": 18890 }, { "grad_norm": 0.584312379360199, "learning_rate": 9.081338097564342e-05, "loss": 0.0467, "step": 18900 }, { "grad_norm": 1.021510124206543, "learning_rate": 9.080143790887137e-05, "loss": 0.0691, "step": 18910 }, { "grad_norm": 0.5726742148399353, "learning_rate": 9.07894878702495e-05, "loss": 0.062, "step": 18920 }, { "grad_norm": 0.5472241044044495, "learning_rate": 9.077753086181974e-05, "loss": 0.0692, "step": 18930 }, { "grad_norm": 0.5128340721130371, "learning_rate": 9.076556688562524e-05, "loss": 0.0988, "step": 18940 }, { "grad_norm": 0.6205823421478271, "learning_rate": 9.075359594371029e-05, "loss": 0.0769, "step": 18950 }, { "grad_norm": 0.4973224699497223, "learning_rate": 9.07416180381204e-05, "loss": 0.0633, "step": 18960 }, { "grad_norm": 0.5464394092559814, "learning_rate": 9.072963317090228e-05, "loss": 0.0594, "step": 18970 }, { "grad_norm": 0.59962397813797, "learning_rate": 9.071764134410382e-05, "loss": 0.0579, "step": 18980 }, { "grad_norm": 0.40656936168670654, "learning_rate": 9.070564255977407e-05, "loss": 0.0542, "step": 18990 }, { "grad_norm": 0.8604210019111633, "learning_rate": 9.06936368199633e-05, "loss": 0.0588, "step": 19000 }, { "grad_norm": 0.6150001883506775, "learning_rate": 9.0681624126723e-05, "loss": 0.0607, "step": 19010 }, { "grad_norm": 0.5745644569396973, "learning_rate": 9.066960448210576e-05, "loss": 0.072, "step": 19020 }, { "grad_norm": 0.3817529082298279, "learning_rate": 9.065757788816543e-05, "loss": 0.0705, "step": 19030 }, { "grad_norm": 0.6998336911201477, "learning_rate": 9.064554434695705e-05, "loss": 0.0722, "step": 19040 }, { "grad_norm": 0.5630521774291992, "learning_rate": 9.063350386053677e-05, "loss": 0.0636, "step": 19050 }, { "grad_norm": 0.5452964305877686, "learning_rate": 9.062145643096202e-05, "loss": 0.0642, "step": 19060 }, { "grad_norm": 0.6934813261032104, "learning_rate": 9.060940206029136e-05, "loss": 0.0645, "step": 19070 }, { "grad_norm": 1.0174561738967896, "learning_rate": 9.059734075058457e-05, "loss": 0.0913, "step": 19080 }, { "grad_norm": 0.5574613213539124, "learning_rate": 9.058527250390257e-05, "loss": 0.0661, "step": 19090 }, { "grad_norm": 0.6196295619010925, "learning_rate": 9.057319732230752e-05, "loss": 0.0623, "step": 19100 }, { "grad_norm": 0.6079989075660706, "learning_rate": 9.056111520786273e-05, "loss": 0.0883, "step": 19110 }, { "grad_norm": 0.739511251449585, "learning_rate": 9.054902616263268e-05, "loss": 0.0581, "step": 19120 }, { "grad_norm": 0.9058651924133301, "learning_rate": 9.05369301886831e-05, "loss": 0.0863, "step": 19130 }, { "grad_norm": 0.5616799592971802, "learning_rate": 9.052482728808083e-05, "loss": 0.0806, "step": 19140 }, { "grad_norm": 1.0385823249816895, "learning_rate": 9.051271746289391e-05, "loss": 0.07, "step": 19150 }, { "grad_norm": 0.7867032289505005, "learning_rate": 9.050060071519162e-05, "loss": 0.0648, "step": 19160 }, { "grad_norm": 0.5076068043708801, "learning_rate": 9.048847704704437e-05, "loss": 0.0605, "step": 19170 }, { "grad_norm": 0.7703844308853149, "learning_rate": 9.047634646052376e-05, "loss": 0.0749, "step": 19180 }, { "grad_norm": 0.5855392217636108, "learning_rate": 9.046420895770256e-05, "loss": 0.0602, "step": 19190 }, { "grad_norm": 0.5034085512161255, "learning_rate": 9.045206454065473e-05, "loss": 0.0578, "step": 19200 }, { "grad_norm": 0.46184706687927246, "learning_rate": 9.043991321145546e-05, "loss": 0.0571, "step": 19210 }, { "grad_norm": 0.6857430934906006, "learning_rate": 9.042775497218105e-05, "loss": 0.0683, "step": 19220 }, { "grad_norm": 0.4010189175605774, "learning_rate": 9.041558982490901e-05, "loss": 0.055, "step": 19230 }, { "grad_norm": 0.5072759985923767, "learning_rate": 9.040341777171805e-05, "loss": 0.0618, "step": 19240 }, { "grad_norm": 0.706986665725708, "learning_rate": 9.039123881468802e-05, "loss": 0.0675, "step": 19250 }, { "grad_norm": 0.5342660546302795, "learning_rate": 9.037905295589998e-05, "loss": 0.0529, "step": 19260 }, { "grad_norm": 0.7831840515136719, "learning_rate": 9.036686019743617e-05, "loss": 0.0619, "step": 19270 }, { "grad_norm": 0.6232349872589111, "learning_rate": 9.035466054137997e-05, "loss": 0.0679, "step": 19280 }, { "grad_norm": 0.5597758293151855, "learning_rate": 9.0342453989816e-05, "loss": 0.0718, "step": 19290 }, { "grad_norm": 0.7500550746917725, "learning_rate": 9.033024054483e-05, "loss": 0.0562, "step": 19300 }, { "grad_norm": 0.5185874700546265, "learning_rate": 9.031802020850894e-05, "loss": 0.0773, "step": 19310 }, { "grad_norm": 0.6141298413276672, "learning_rate": 9.030579298294092e-05, "loss": 0.0648, "step": 19320 }, { "grad_norm": 0.5261078476905823, "learning_rate": 9.029355887021524e-05, "loss": 0.0614, "step": 19330 }, { "grad_norm": 0.7702931761741638, "learning_rate": 9.028131787242238e-05, "loss": 0.0821, "step": 19340 }, { "grad_norm": 0.9259278178215027, "learning_rate": 9.026906999165399e-05, "loss": 0.0654, "step": 19350 }, { "grad_norm": 0.6216511130332947, "learning_rate": 9.025681523000291e-05, "loss": 0.0539, "step": 19360 }, { "grad_norm": 0.6657760739326477, "learning_rate": 9.024455358956315e-05, "loss": 0.063, "step": 19370 }, { "grad_norm": 0.6683436036109924, "learning_rate": 9.023228507242984e-05, "loss": 0.061, "step": 19380 }, { "grad_norm": 0.6515514850616455, "learning_rate": 9.022000968069937e-05, "loss": 0.0565, "step": 19390 }, { "grad_norm": 0.5634434819221497, "learning_rate": 9.020772741646928e-05, "loss": 0.0509, "step": 19400 }, { "grad_norm": 0.5073848366737366, "learning_rate": 9.019543828183826e-05, "loss": 0.0631, "step": 19410 }, { "grad_norm": 0.6418549418449402, "learning_rate": 9.018314227890616e-05, "loss": 0.0554, "step": 19420 }, { "grad_norm": 0.4963427782058716, "learning_rate": 9.017083940977408e-05, "loss": 0.0466, "step": 19430 }, { "grad_norm": 0.6499828696250916, "learning_rate": 9.015852967654422e-05, "loss": 0.0747, "step": 19440 }, { "grad_norm": 0.7275366187095642, "learning_rate": 9.014621308131996e-05, "loss": 0.0794, "step": 19450 }, { "grad_norm": 0.5305042862892151, "learning_rate": 9.01338896262059e-05, "loss": 0.0704, "step": 19460 }, { "grad_norm": 0.8322250247001648, "learning_rate": 9.012155931330777e-05, "loss": 0.0537, "step": 19470 }, { "grad_norm": 0.6101081371307373, "learning_rate": 9.010922214473246e-05, "loss": 0.0656, "step": 19480 }, { "grad_norm": 0.676109790802002, "learning_rate": 9.009687812258808e-05, "loss": 0.0655, "step": 19490 }, { "grad_norm": 0.631348729133606, "learning_rate": 9.00845272489839e-05, "loss": 0.0588, "step": 19500 }, { "grad_norm": 0.5720543265342712, "learning_rate": 9.007216952603031e-05, "loss": 0.0699, "step": 19510 }, { "grad_norm": 0.6668919324874878, "learning_rate": 9.005980495583894e-05, "loss": 0.0606, "step": 19520 }, { "grad_norm": 0.5726596713066101, "learning_rate": 9.004743354052252e-05, "loss": 0.0608, "step": 19530 }, { "grad_norm": 0.6352741122245789, "learning_rate": 9.003505528219503e-05, "loss": 0.0684, "step": 19540 }, { "grad_norm": 0.6942742466926575, "learning_rate": 9.002267018297154e-05, "loss": 0.0728, "step": 19550 }, { "grad_norm": 0.658370852470398, "learning_rate": 9.001027824496834e-05, "loss": 0.0581, "step": 19560 }, { "grad_norm": 0.5530434250831604, "learning_rate": 8.999787947030287e-05, "loss": 0.0527, "step": 19570 }, { "grad_norm": 0.5185115933418274, "learning_rate": 8.998547386109376e-05, "loss": 0.0578, "step": 19580 }, { "grad_norm": 0.7476016283035278, "learning_rate": 8.997306141946073e-05, "loss": 0.0596, "step": 19590 }, { "grad_norm": 0.46581441164016724, "learning_rate": 8.996064214752481e-05, "loss": 0.0436, "step": 19600 }, { "grad_norm": 0.5403585433959961, "learning_rate": 8.994821604740806e-05, "loss": 0.0479, "step": 19610 }, { "grad_norm": 0.6916559934616089, "learning_rate": 8.993578312123377e-05, "loss": 0.0655, "step": 19620 }, { "grad_norm": 0.7994701266288757, "learning_rate": 8.992334337112639e-05, "loss": 0.0796, "step": 19630 }, { "grad_norm": 0.6984329223632812, "learning_rate": 8.991089679921154e-05, "loss": 0.0501, "step": 19640 }, { "grad_norm": 0.6193528175354004, "learning_rate": 8.989844340761599e-05, "loss": 0.0792, "step": 19650 }, { "grad_norm": 0.49505603313446045, "learning_rate": 8.988598319846768e-05, "loss": 0.0649, "step": 19660 }, { "grad_norm": 0.7565456628799438, "learning_rate": 8.987351617389574e-05, "loss": 0.0796, "step": 19670 }, { "grad_norm": 0.7267931699752808, "learning_rate": 8.98610423360304e-05, "loss": 0.0716, "step": 19680 }, { "grad_norm": 0.508873462677002, "learning_rate": 8.984856168700317e-05, "loss": 0.0765, "step": 19690 }, { "grad_norm": 0.5508507490158081, "learning_rate": 8.983607422894658e-05, "loss": 0.0666, "step": 19700 }, { "grad_norm": 0.6335818767547607, "learning_rate": 8.982357996399442e-05, "loss": 0.0554, "step": 19710 }, { "grad_norm": 0.636352002620697, "learning_rate": 8.981107889428164e-05, "loss": 0.0626, "step": 19720 }, { "grad_norm": 0.6604889631271362, "learning_rate": 8.979857102194428e-05, "loss": 0.0669, "step": 19730 }, { "grad_norm": 0.539260745048523, "learning_rate": 8.978605634911968e-05, "loss": 0.0493, "step": 19740 }, { "grad_norm": 0.8864102363586426, "learning_rate": 8.977353487794616e-05, "loss": 0.0665, "step": 19750 }, { "grad_norm": 0.8190581798553467, "learning_rate": 8.976100661056334e-05, "loss": 0.0734, "step": 19760 }, { "grad_norm": 0.9584612250328064, "learning_rate": 8.974847154911197e-05, "loss": 0.0625, "step": 19770 }, { "grad_norm": 0.5473459362983704, "learning_rate": 8.973592969573393e-05, "loss": 0.0626, "step": 19780 }, { "grad_norm": 0.566771924495697, "learning_rate": 8.972338105257228e-05, "loss": 0.0742, "step": 19790 }, { "grad_norm": 0.6515986323356628, "learning_rate": 8.971082562177125e-05, "loss": 0.0613, "step": 19800 }, { "grad_norm": 0.9233022332191467, "learning_rate": 8.96982634054762e-05, "loss": 0.0727, "step": 19810 }, { "grad_norm": 0.4280672073364258, "learning_rate": 8.96856944058337e-05, "loss": 0.0631, "step": 19820 }, { "grad_norm": 1.728558897972107, "learning_rate": 8.967311862499144e-05, "loss": 0.0824, "step": 19830 }, { "grad_norm": 0.4378092288970947, "learning_rate": 8.966053606509825e-05, "loss": 0.0663, "step": 19840 }, { "grad_norm": 0.6194527745246887, "learning_rate": 8.964794672830417e-05, "loss": 0.0725, "step": 19850 }, { "grad_norm": 0.9608381390571594, "learning_rate": 8.963535061676038e-05, "loss": 0.054, "step": 19860 }, { "grad_norm": 0.629043459892273, "learning_rate": 8.962274773261918e-05, "loss": 0.0607, "step": 19870 }, { "grad_norm": 0.43277695775032043, "learning_rate": 8.961013807803409e-05, "loss": 0.0672, "step": 19880 }, { "grad_norm": 0.7821388840675354, "learning_rate": 8.959752165515973e-05, "loss": 0.0539, "step": 19890 }, { "grad_norm": 0.7430356740951538, "learning_rate": 8.958489846615193e-05, "loss": 0.0662, "step": 19900 }, { "grad_norm": 0.5809447169303894, "learning_rate": 8.957226851316762e-05, "loss": 0.0665, "step": 19910 }, { "grad_norm": 0.5347902178764343, "learning_rate": 8.955963179836493e-05, "loss": 0.0506, "step": 19920 }, { "grad_norm": 0.6485026478767395, "learning_rate": 8.954698832390312e-05, "loss": 0.0595, "step": 19930 }, { "grad_norm": 0.5500348210334778, "learning_rate": 8.953433809194263e-05, "loss": 0.079, "step": 19940 }, { "grad_norm": 0.8676353096961975, "learning_rate": 8.9521681104645e-05, "loss": 0.0731, "step": 19950 }, { "grad_norm": 0.8306151032447815, "learning_rate": 8.9509017364173e-05, "loss": 0.0633, "step": 19960 }, { "grad_norm": 0.7735332250595093, "learning_rate": 8.949634687269052e-05, "loss": 0.0798, "step": 19970 }, { "grad_norm": 0.8389769196510315, "learning_rate": 8.948366963236259e-05, "loss": 0.0587, "step": 19980 }, { "grad_norm": 0.6740069389343262, "learning_rate": 8.947098564535538e-05, "loss": 0.071, "step": 19990 }, { "grad_norm": 0.7869225144386292, "learning_rate": 8.945829491383627e-05, "loss": 0.0853, "step": 20000 }, { "grad_norm": 0.6057875156402588, "learning_rate": 8.944559743997374e-05, "loss": 0.0647, "step": 20010 }, { "grad_norm": 0.606755793094635, "learning_rate": 8.943289322593746e-05, "loss": 0.0537, "step": 20020 }, { "grad_norm": 0.47947022318840027, "learning_rate": 8.942018227389821e-05, "loss": 0.053, "step": 20030 }, { "grad_norm": 0.5114125609397888, "learning_rate": 8.940746458602795e-05, "loss": 0.0657, "step": 20040 }, { "grad_norm": 0.5708791017532349, "learning_rate": 8.939474016449979e-05, "loss": 0.0569, "step": 20050 }, { "grad_norm": 0.5546413660049438, "learning_rate": 8.938200901148799e-05, "loss": 0.0598, "step": 20060 }, { "grad_norm": 0.596352756023407, "learning_rate": 8.936927112916795e-05, "loss": 0.0716, "step": 20070 }, { "grad_norm": 0.7176905870437622, "learning_rate": 8.935652651971622e-05, "loss": 0.0721, "step": 20080 }, { "grad_norm": 0.44379574060440063, "learning_rate": 8.934377518531052e-05, "loss": 0.0721, "step": 20090 }, { "grad_norm": 0.8951563239097595, "learning_rate": 8.933101712812967e-05, "loss": 0.0726, "step": 20100 }, { "grad_norm": 0.43140947818756104, "learning_rate": 8.931825235035374e-05, "loss": 0.0779, "step": 20110 }, { "grad_norm": 0.6755140423774719, "learning_rate": 8.930548085416382e-05, "loss": 0.0783, "step": 20120 }, { "grad_norm": 0.5001846551895142, "learning_rate": 8.92927026417422e-05, "loss": 0.0572, "step": 20130 }, { "grad_norm": 0.6551517248153687, "learning_rate": 8.92799177152724e-05, "loss": 0.055, "step": 20140 }, { "grad_norm": 0.6733238697052002, "learning_rate": 8.926712607693895e-05, "loss": 0.0631, "step": 20150 }, { "grad_norm": 0.9029421210289001, "learning_rate": 8.925432772892762e-05, "loss": 0.0638, "step": 20160 }, { "grad_norm": 0.5927464365959167, "learning_rate": 8.924152267342529e-05, "loss": 0.0602, "step": 20170 }, { "grad_norm": 2.2757132053375244, "learning_rate": 8.922871091261998e-05, "loss": 0.0711, "step": 20180 }, { "grad_norm": 0.5432871580123901, "learning_rate": 8.92158924487009e-05, "loss": 0.0571, "step": 20190 }, { "grad_norm": 0.5988584160804749, "learning_rate": 8.920306728385834e-05, "loss": 0.0666, "step": 20200 }, { "grad_norm": 0.6654399633407593, "learning_rate": 8.919023542028379e-05, "loss": 0.0509, "step": 20210 }, { "grad_norm": 0.5903798341751099, "learning_rate": 8.917739686016988e-05, "loss": 0.0544, "step": 20220 }, { "grad_norm": 0.44830262660980225, "learning_rate": 8.916455160571033e-05, "loss": 0.0518, "step": 20230 }, { "grad_norm": 0.4107033312320709, "learning_rate": 8.915169965910008e-05, "loss": 0.063, "step": 20240 }, { "grad_norm": 0.6507498025894165, "learning_rate": 8.913884102253514e-05, "loss": 0.0681, "step": 20250 }, { "grad_norm": 0.8558493256568909, "learning_rate": 8.912597569821273e-05, "loss": 0.0926, "step": 20260 }, { "grad_norm": 0.7425901889801025, "learning_rate": 8.911310368833118e-05, "loss": 0.0752, "step": 20270 }, { "grad_norm": 0.67109614610672, "learning_rate": 8.910022499508994e-05, "loss": 0.0699, "step": 20280 }, { "grad_norm": 0.6483206748962402, "learning_rate": 8.908733962068965e-05, "loss": 0.0659, "step": 20290 }, { "grad_norm": 0.5277379751205444, "learning_rate": 8.907444756733207e-05, "loss": 0.0712, "step": 20300 }, { "grad_norm": 0.5739470720291138, "learning_rate": 8.906154883722006e-05, "loss": 0.0671, "step": 20310 }, { "grad_norm": 0.5832505226135254, "learning_rate": 8.904864343255773e-05, "loss": 0.0702, "step": 20320 }, { "grad_norm": 0.6309769749641418, "learning_rate": 8.90357313555502e-05, "loss": 0.0562, "step": 20330 }, { "grad_norm": 0.9424728751182556, "learning_rate": 8.90228126084038e-05, "loss": 0.0718, "step": 20340 }, { "grad_norm": 0.5392253398895264, "learning_rate": 8.900988719332601e-05, "loss": 0.0739, "step": 20350 }, { "grad_norm": 0.676449716091156, "learning_rate": 8.899695511252542e-05, "loss": 0.0842, "step": 20360 }, { "grad_norm": 0.6247406601905823, "learning_rate": 8.898401636821176e-05, "loss": 0.069, "step": 20370 }, { "grad_norm": 0.7943164706230164, "learning_rate": 8.897107096259593e-05, "loss": 0.1017, "step": 20380 }, { "grad_norm": 0.6418213844299316, "learning_rate": 8.895811889788994e-05, "loss": 0.0697, "step": 20390 }, { "grad_norm": 0.6225273609161377, "learning_rate": 8.894516017630692e-05, "loss": 0.0626, "step": 20400 }, { "grad_norm": 0.6037925481796265, "learning_rate": 8.893219480006118e-05, "loss": 0.0675, "step": 20410 }, { "grad_norm": 0.7001067996025085, "learning_rate": 8.891922277136817e-05, "loss": 0.0664, "step": 20420 }, { "grad_norm": 0.5324599146842957, "learning_rate": 8.890624409244441e-05, "loss": 0.0715, "step": 20430 }, { "grad_norm": 0.4434494972229004, "learning_rate": 8.889325876550763e-05, "loss": 0.0551, "step": 20440 }, { "grad_norm": 0.8757927417755127, "learning_rate": 8.888026679277666e-05, "loss": 0.063, "step": 20450 }, { "grad_norm": 0.6437824368476868, "learning_rate": 8.886726817647147e-05, "loss": 0.0737, "step": 20460 }, { "grad_norm": 0.8455667495727539, "learning_rate": 8.885426291881319e-05, "loss": 0.0656, "step": 20470 }, { "grad_norm": 0.5341448187828064, "learning_rate": 8.884125102202401e-05, "loss": 0.0581, "step": 20480 }, { "grad_norm": 0.8262205123901367, "learning_rate": 8.882823248832736e-05, "loss": 0.0661, "step": 20490 }, { "grad_norm": 0.6991628408432007, "learning_rate": 8.881520731994772e-05, "loss": 0.0614, "step": 20500 }, { "grad_norm": 0.3689647614955902, "learning_rate": 8.880217551911077e-05, "loss": 0.051, "step": 20510 }, { "grad_norm": 0.47688931226730347, "learning_rate": 8.878913708804323e-05, "loss": 0.0569, "step": 20520 }, { "grad_norm": 0.4432177245616913, "learning_rate": 8.877609202897308e-05, "loss": 0.0604, "step": 20530 }, { "grad_norm": 0.620155930519104, "learning_rate": 8.876304034412933e-05, "loss": 0.058, "step": 20540 }, { "grad_norm": 0.6744645833969116, "learning_rate": 8.874998203574214e-05, "loss": 0.0776, "step": 20550 }, { "grad_norm": 0.8464251756668091, "learning_rate": 8.873691710604284e-05, "loss": 0.0576, "step": 20560 }, { "grad_norm": 0.703877866268158, "learning_rate": 8.872384555726387e-05, "loss": 0.0823, "step": 20570 }, { "grad_norm": 0.5238223671913147, "learning_rate": 8.871076739163878e-05, "loss": 0.0604, "step": 20580 }, { "grad_norm": 0.6763887405395508, "learning_rate": 8.86976826114023e-05, "loss": 0.0578, "step": 20590 }, { "grad_norm": 0.691879153251648, "learning_rate": 8.868459121879023e-05, "loss": 0.0636, "step": 20600 }, { "grad_norm": 0.819402277469635, "learning_rate": 8.867149321603956e-05, "loss": 0.0596, "step": 20610 }, { "grad_norm": 0.7950870394706726, "learning_rate": 8.865838860538835e-05, "loss": 0.0697, "step": 20620 }, { "grad_norm": 0.7069525718688965, "learning_rate": 8.864527738907585e-05, "loss": 0.0871, "step": 20630 }, { "grad_norm": 0.7114612460136414, "learning_rate": 8.863215956934239e-05, "loss": 0.0681, "step": 20640 }, { "grad_norm": 0.563555121421814, "learning_rate": 8.861903514842947e-05, "loss": 0.0703, "step": 20650 }, { "grad_norm": 0.7068413496017456, "learning_rate": 8.860590412857966e-05, "loss": 0.0643, "step": 20660 }, { "grad_norm": 0.5794710516929626, "learning_rate": 8.85927665120367e-05, "loss": 0.053, "step": 20670 }, { "grad_norm": 0.6379481554031372, "learning_rate": 8.857962230104546e-05, "loss": 0.0857, "step": 20680 }, { "grad_norm": 0.5357361435890198, "learning_rate": 8.856647149785193e-05, "loss": 0.0674, "step": 20690 }, { "grad_norm": 0.4803427755832672, "learning_rate": 8.855331410470322e-05, "loss": 0.0772, "step": 20700 }, { "grad_norm": 0.5786077976226807, "learning_rate": 8.854015012384756e-05, "loss": 0.0573, "step": 20710 }, { "grad_norm": 0.7600740194320679, "learning_rate": 8.852697955753433e-05, "loss": 0.0443, "step": 20720 }, { "grad_norm": 0.7207571268081665, "learning_rate": 8.851380240801399e-05, "loss": 0.0659, "step": 20730 }, { "grad_norm": 0.4711228609085083, "learning_rate": 8.850061867753818e-05, "loss": 0.05, "step": 20740 }, { "grad_norm": 0.5766306519508362, "learning_rate": 8.848742836835963e-05, "loss": 0.0533, "step": 20750 }, { "grad_norm": 0.7335509657859802, "learning_rate": 8.847423148273221e-05, "loss": 0.0648, "step": 20760 }, { "grad_norm": 0.6545851826667786, "learning_rate": 8.846102802291092e-05, "loss": 0.0626, "step": 20770 }, { "grad_norm": 0.48209527134895325, "learning_rate": 8.844781799115183e-05, "loss": 0.0592, "step": 20780 }, { "grad_norm": 0.6661927103996277, "learning_rate": 8.84346013897122e-05, "loss": 0.0639, "step": 20790 }, { "grad_norm": 0.8646058440208435, "learning_rate": 8.842137822085038e-05, "loss": 0.0607, "step": 20800 }, { "grad_norm": 0.7088509202003479, "learning_rate": 8.840814848682585e-05, "loss": 0.055, "step": 20810 }, { "grad_norm": 0.4976431131362915, "learning_rate": 8.83949121898992e-05, "loss": 0.0587, "step": 20820 }, { "grad_norm": 0.5076876282691956, "learning_rate": 8.838166933233217e-05, "loss": 0.0537, "step": 20830 }, { "grad_norm": 0.4580915570259094, "learning_rate": 8.83684199163876e-05, "loss": 0.0648, "step": 20840 }, { "grad_norm": 0.6445158123970032, "learning_rate": 8.835516394432943e-05, "loss": 0.0561, "step": 20850 }, { "grad_norm": 0.6131120920181274, "learning_rate": 8.834190141842276e-05, "loss": 0.0511, "step": 20860 }, { "grad_norm": 0.5507746934890747, "learning_rate": 8.83286323409338e-05, "loss": 0.0579, "step": 20870 }, { "grad_norm": 1.1263558864593506, "learning_rate": 8.831535671412986e-05, "loss": 0.0636, "step": 20880 }, { "grad_norm": 0.630460798740387, "learning_rate": 8.830207454027938e-05, "loss": 0.051, "step": 20890 }, { "grad_norm": 0.3123641312122345, "learning_rate": 8.828878582165192e-05, "loss": 0.0674, "step": 20900 }, { "grad_norm": 0.49918290972709656, "learning_rate": 8.827549056051818e-05, "loss": 0.0573, "step": 20910 }, { "grad_norm": 0.6659142971038818, "learning_rate": 8.826218875914993e-05, "loss": 0.0567, "step": 20920 }, { "grad_norm": 0.5034124851226807, "learning_rate": 8.82488804198201e-05, "loss": 0.0566, "step": 20930 }, { "grad_norm": 0.4721842408180237, "learning_rate": 8.82355655448027e-05, "loss": 0.0576, "step": 20940 }, { "grad_norm": 0.6606333255767822, "learning_rate": 8.822224413637293e-05, "loss": 0.0675, "step": 20950 }, { "grad_norm": 0.6437013745307922, "learning_rate": 8.820891619680697e-05, "loss": 0.0503, "step": 20960 }, { "grad_norm": 0.4467027187347412, "learning_rate": 8.819558172838227e-05, "loss": 0.0493, "step": 20970 }, { "grad_norm": 0.5131641626358032, "learning_rate": 8.818224073337731e-05, "loss": 0.0587, "step": 20980 }, { "grad_norm": 0.5664061307907104, "learning_rate": 8.816889321407169e-05, "loss": 0.0463, "step": 20990 }, { "grad_norm": 0.4545719027519226, "learning_rate": 8.815553917274615e-05, "loss": 0.0539, "step": 21000 }, { "grad_norm": 0.542976975440979, "learning_rate": 8.81421786116825e-05, "loss": 0.0556, "step": 21010 }, { "grad_norm": 0.46539801359176636, "learning_rate": 8.812881153316373e-05, "loss": 0.0794, "step": 21020 }, { "grad_norm": 0.6342597603797913, "learning_rate": 8.81154379394739e-05, "loss": 0.0584, "step": 21030 }, { "grad_norm": 0.48193690180778503, "learning_rate": 8.810205783289818e-05, "loss": 0.0508, "step": 21040 }, { "grad_norm": 0.6625446677207947, "learning_rate": 8.808867121572286e-05, "loss": 0.0759, "step": 21050 }, { "grad_norm": 0.49563512206077576, "learning_rate": 8.807527809023537e-05, "loss": 0.0577, "step": 21060 }, { "grad_norm": 0.6518958210945129, "learning_rate": 8.80618784587242e-05, "loss": 0.0646, "step": 21070 }, { "grad_norm": 0.5806330442428589, "learning_rate": 8.804847232347902e-05, "loss": 0.0525, "step": 21080 }, { "grad_norm": 0.6550595760345459, "learning_rate": 8.803505968679054e-05, "loss": 0.0523, "step": 21090 }, { "grad_norm": 0.649861216545105, "learning_rate": 8.802164055095061e-05, "loss": 0.0481, "step": 21100 }, { "grad_norm": 0.8291419744491577, "learning_rate": 8.80082149182522e-05, "loss": 0.0618, "step": 21110 }, { "grad_norm": 0.6029055118560791, "learning_rate": 8.79947827909894e-05, "loss": 0.0651, "step": 21120 }, { "grad_norm": 0.44255632162094116, "learning_rate": 8.798134417145738e-05, "loss": 0.0552, "step": 21130 }, { "grad_norm": 0.6024697422981262, "learning_rate": 8.796789906195243e-05, "loss": 0.0465, "step": 21140 }, { "grad_norm": 0.5101638436317444, "learning_rate": 8.795444746477195e-05, "loss": 0.0582, "step": 21150 }, { "grad_norm": 0.5395312309265137, "learning_rate": 8.794098938221446e-05, "loss": 0.065, "step": 21160 }, { "grad_norm": 0.38761043548583984, "learning_rate": 8.792752481657957e-05, "loss": 0.0652, "step": 21170 }, { "grad_norm": 0.6721885800361633, "learning_rate": 8.791405377016802e-05, "loss": 0.057, "step": 21180 }, { "grad_norm": 0.9779784083366394, "learning_rate": 8.790057624528163e-05, "loss": 0.0663, "step": 21190 }, { "grad_norm": 0.48443636298179626, "learning_rate": 8.788709224422333e-05, "loss": 0.0853, "step": 21200 }, { "grad_norm": 0.5812711715698242, "learning_rate": 8.787360176929717e-05, "loss": 0.0965, "step": 21210 }, { "grad_norm": 0.4564872980117798, "learning_rate": 8.786010482280834e-05, "loss": 0.0582, "step": 21220 }, { "grad_norm": 0.7626149654388428, "learning_rate": 8.784660140706306e-05, "loss": 0.0823, "step": 21230 }, { "grad_norm": 0.7428777813911438, "learning_rate": 8.783309152436872e-05, "loss": 0.0719, "step": 21240 }, { "grad_norm": 0.9371791481971741, "learning_rate": 8.781957517703375e-05, "loss": 0.0596, "step": 21250 }, { "grad_norm": 0.6503125429153442, "learning_rate": 8.780605236736776e-05, "loss": 0.0425, "step": 21260 }, { "grad_norm": 0.5830585360527039, "learning_rate": 8.779252309768142e-05, "loss": 0.0576, "step": 21270 }, { "grad_norm": 0.8852337002754211, "learning_rate": 8.777898737028652e-05, "loss": 0.0697, "step": 21280 }, { "grad_norm": 0.5833387970924377, "learning_rate": 8.776544518749591e-05, "loss": 0.075, "step": 21290 }, { "grad_norm": 1.0284799337387085, "learning_rate": 8.775189655162364e-05, "loss": 0.0588, "step": 21300 }, { "grad_norm": 0.8630209565162659, "learning_rate": 8.773834146498474e-05, "loss": 0.0697, "step": 21310 }, { "grad_norm": 0.4796529710292816, "learning_rate": 8.772477992989545e-05, "loss": 0.0577, "step": 21320 }, { "grad_norm": 0.8897953033447266, "learning_rate": 8.771121194867304e-05, "loss": 0.094, "step": 21330 }, { "grad_norm": 0.7616614103317261, "learning_rate": 8.769763752363589e-05, "loss": 0.0701, "step": 21340 }, { "grad_norm": 0.44400855898857117, "learning_rate": 8.768405665710352e-05, "loss": 0.0694, "step": 21350 }, { "grad_norm": 0.6301707625389099, "learning_rate": 8.767046935139655e-05, "loss": 0.0625, "step": 21360 }, { "grad_norm": 0.4091760516166687, "learning_rate": 8.765687560883666e-05, "loss": 0.0558, "step": 21370 }, { "grad_norm": 0.6515419483184814, "learning_rate": 8.764327543174664e-05, "loss": 0.0485, "step": 21380 }, { "grad_norm": 0.4735148251056671, "learning_rate": 8.762966882245038e-05, "loss": 0.0575, "step": 21390 }, { "grad_norm": 0.5910167098045349, "learning_rate": 8.761605578327291e-05, "loss": 0.0917, "step": 21400 }, { "grad_norm": 1.1118321418762207, "learning_rate": 8.76024363165403e-05, "loss": 0.0687, "step": 21410 }, { "grad_norm": 0.5241365432739258, "learning_rate": 8.758881042457976e-05, "loss": 0.0525, "step": 21420 }, { "grad_norm": 0.419109582901001, "learning_rate": 8.757517810971957e-05, "loss": 0.0477, "step": 21430 }, { "grad_norm": 0.6869828104972839, "learning_rate": 8.756153937428913e-05, "loss": 0.0565, "step": 21440 }, { "grad_norm": 0.7188447117805481, "learning_rate": 8.754789422061889e-05, "loss": 0.0613, "step": 21450 }, { "grad_norm": 0.5494624972343445, "learning_rate": 8.753424265104052e-05, "loss": 0.0504, "step": 21460 }, { "grad_norm": 0.6266254782676697, "learning_rate": 8.752058466788659e-05, "loss": 0.0622, "step": 21470 }, { "grad_norm": 0.7226842641830444, "learning_rate": 8.750692027349097e-05, "loss": 0.0539, "step": 21480 }, { "grad_norm": 0.6526488661766052, "learning_rate": 8.749324947018847e-05, "loss": 0.0817, "step": 21490 }, { "grad_norm": 0.7056708335876465, "learning_rate": 8.747957226031507e-05, "loss": 0.0647, "step": 21500 }, { "grad_norm": 0.6318761706352234, "learning_rate": 8.746588864620787e-05, "loss": 0.0912, "step": 21510 }, { "grad_norm": 0.7108086943626404, "learning_rate": 8.745219863020498e-05, "loss": 0.063, "step": 21520 }, { "grad_norm": 0.4607839286327362, "learning_rate": 8.743850221464564e-05, "loss": 0.0576, "step": 21530 }, { "grad_norm": 0.7331581115722656, "learning_rate": 8.742479940187026e-05, "loss": 0.0843, "step": 21540 }, { "grad_norm": 0.5033329725265503, "learning_rate": 8.74110901942202e-05, "loss": 0.0707, "step": 21550 }, { "grad_norm": 0.6073421239852905, "learning_rate": 8.739737459403803e-05, "loss": 0.0725, "step": 21560 }, { "grad_norm": 0.5877411961555481, "learning_rate": 8.738365260366737e-05, "loss": 0.0671, "step": 21570 }, { "grad_norm": 0.5229293704032898, "learning_rate": 8.736992422545292e-05, "loss": 0.079, "step": 21580 }, { "grad_norm": 0.9872821569442749, "learning_rate": 8.73561894617405e-05, "loss": 0.083, "step": 21590 }, { "grad_norm": 0.4954920709133148, "learning_rate": 8.734244831487697e-05, "loss": 0.0569, "step": 21600 }, { "grad_norm": 0.7638189196586609, "learning_rate": 8.732870078721035e-05, "loss": 0.0656, "step": 21610 }, { "grad_norm": 0.7086394429206848, "learning_rate": 8.731494688108972e-05, "loss": 0.0557, "step": 21620 }, { "grad_norm": 0.5221805572509766, "learning_rate": 8.730118659886523e-05, "loss": 0.061, "step": 21630 }, { "grad_norm": 0.41627758741378784, "learning_rate": 8.728741994288814e-05, "loss": 0.0603, "step": 21640 }, { "grad_norm": 0.43166106939315796, "learning_rate": 8.727364691551079e-05, "loss": 0.0621, "step": 21650 }, { "grad_norm": 0.49378684163093567, "learning_rate": 8.725986751908661e-05, "loss": 0.0515, "step": 21660 }, { "grad_norm": 0.9400157332420349, "learning_rate": 8.724608175597016e-05, "loss": 0.0886, "step": 21670 }, { "grad_norm": 0.6347653865814209, "learning_rate": 8.723228962851699e-05, "loss": 0.0524, "step": 21680 }, { "grad_norm": 0.5511806607246399, "learning_rate": 8.721849113908385e-05, "loss": 0.0615, "step": 21690 }, { "grad_norm": 0.684083104133606, "learning_rate": 8.720468629002848e-05, "loss": 0.0616, "step": 21700 }, { "grad_norm": 0.6148364543914795, "learning_rate": 8.719087508370978e-05, "loss": 0.0583, "step": 21710 }, { "grad_norm": 0.5593186020851135, "learning_rate": 8.717705752248772e-05, "loss": 0.0588, "step": 21720 }, { "grad_norm": 0.6406545639038086, "learning_rate": 8.71632336087233e-05, "loss": 0.0608, "step": 21730 }, { "grad_norm": 0.6105580925941467, "learning_rate": 8.71494033447787e-05, "loss": 0.0621, "step": 21740 }, { "grad_norm": 0.9141104221343994, "learning_rate": 8.713556673301708e-05, "loss": 0.0829, "step": 21750 }, { "grad_norm": 0.5340642333030701, "learning_rate": 8.712172377580278e-05, "loss": 0.0643, "step": 21760 }, { "grad_norm": 0.7627577781677246, "learning_rate": 8.710787447550114e-05, "loss": 0.0653, "step": 21770 }, { "grad_norm": 0.5944981575012207, "learning_rate": 8.70940188344787e-05, "loss": 0.0622, "step": 21780 }, { "grad_norm": 0.7839078903198242, "learning_rate": 8.708015685510293e-05, "loss": 0.0752, "step": 21790 }, { "grad_norm": 0.7201610207557678, "learning_rate": 8.706628853974252e-05, "loss": 0.0603, "step": 21800 }, { "grad_norm": 0.48590564727783203, "learning_rate": 8.705241389076715e-05, "loss": 0.06, "step": 21810 }, { "grad_norm": 0.44295307993888855, "learning_rate": 8.703853291054764e-05, "loss": 0.0568, "step": 21820 }, { "grad_norm": 0.7783655524253845, "learning_rate": 8.702464560145587e-05, "loss": 0.0619, "step": 21830 }, { "grad_norm": 0.679247260093689, "learning_rate": 8.701075196586476e-05, "loss": 0.0645, "step": 21840 }, { "grad_norm": 0.7385334372520447, "learning_rate": 8.699685200614842e-05, "loss": 0.0882, "step": 21850 }, { "grad_norm": 0.6342458724975586, "learning_rate": 8.698294572468193e-05, "loss": 0.0642, "step": 21860 }, { "grad_norm": 0.8245880007743835, "learning_rate": 8.696903312384148e-05, "loss": 0.0552, "step": 21870 }, { "grad_norm": 0.5612295866012573, "learning_rate": 8.695511420600439e-05, "loss": 0.0795, "step": 21880 }, { "grad_norm": 0.7008810043334961, "learning_rate": 8.694118897354901e-05, "loss": 0.0587, "step": 21890 }, { "grad_norm": 0.5021196007728577, "learning_rate": 8.692725742885478e-05, "loss": 0.0595, "step": 21900 }, { "grad_norm": 0.6580286026000977, "learning_rate": 8.691331957430221e-05, "loss": 0.0531, "step": 21910 }, { "grad_norm": 0.6395254731178284, "learning_rate": 8.68993754122729e-05, "loss": 0.0634, "step": 21920 }, { "grad_norm": 0.752317488193512, "learning_rate": 8.688542494514955e-05, "loss": 0.0514, "step": 21930 }, { "grad_norm": 0.5294877290725708, "learning_rate": 8.68714681753159e-05, "loss": 0.0614, "step": 21940 }, { "grad_norm": 0.54995197057724, "learning_rate": 8.685750510515676e-05, "loss": 0.0508, "step": 21950 }, { "grad_norm": 0.43665140867233276, "learning_rate": 8.684353573705805e-05, "loss": 0.0494, "step": 21960 }, { "grad_norm": 0.6583656072616577, "learning_rate": 8.682956007340677e-05, "loss": 0.0499, "step": 21970 }, { "grad_norm": 0.8045487999916077, "learning_rate": 8.681557811659095e-05, "loss": 0.0636, "step": 21980 }, { "grad_norm": 0.6253414154052734, "learning_rate": 8.680158986899974e-05, "loss": 0.0593, "step": 21990 }, { "grad_norm": 1.0657979249954224, "learning_rate": 8.678759533302335e-05, "loss": 0.0605, "step": 22000 }, { "grad_norm": 0.8009856939315796, "learning_rate": 8.677359451105308e-05, "loss": 0.0668, "step": 22010 }, { "grad_norm": 1.017295241355896, "learning_rate": 8.675958740548123e-05, "loss": 0.0553, "step": 22020 }, { "grad_norm": 0.6354982852935791, "learning_rate": 8.674557401870129e-05, "loss": 0.0694, "step": 22030 }, { "grad_norm": 0.9770088195800781, "learning_rate": 8.673155435310775e-05, "loss": 0.0739, "step": 22040 }, { "grad_norm": 0.5454126000404358, "learning_rate": 8.671752841109617e-05, "loss": 0.0578, "step": 22050 }, { "grad_norm": 0.6810362339019775, "learning_rate": 8.670349619506321e-05, "loss": 0.0595, "step": 22060 }, { "grad_norm": 0.6219758987426758, "learning_rate": 8.66894577074066e-05, "loss": 0.0552, "step": 22070 }, { "grad_norm": 0.5482583045959473, "learning_rate": 8.667541295052513e-05, "loss": 0.0582, "step": 22080 }, { "grad_norm": 0.6311449408531189, "learning_rate": 8.666136192681865e-05, "loss": 0.0572, "step": 22090 }, { "grad_norm": 0.6595725417137146, "learning_rate": 8.664730463868811e-05, "loss": 0.0498, "step": 22100 }, { "grad_norm": 0.6485617756843567, "learning_rate": 8.663324108853552e-05, "loss": 0.0667, "step": 22110 }, { "grad_norm": 0.7292831540107727, "learning_rate": 8.661917127876395e-05, "loss": 0.0672, "step": 22120 }, { "grad_norm": 0.5259402990341187, "learning_rate": 8.660509521177754e-05, "loss": 0.0544, "step": 22130 }, { "grad_norm": 0.48807236552238464, "learning_rate": 8.65910128899815e-05, "loss": 0.0518, "step": 22140 }, { "grad_norm": 0.47974318265914917, "learning_rate": 8.657692431578214e-05, "loss": 0.0509, "step": 22150 }, { "grad_norm": 0.3997730016708374, "learning_rate": 8.656282949158679e-05, "loss": 0.0677, "step": 22160 }, { "grad_norm": 0.618710994720459, "learning_rate": 8.654872841980388e-05, "loss": 0.0532, "step": 22170 }, { "grad_norm": 0.5739135146141052, "learning_rate": 8.653462110284289e-05, "loss": 0.0612, "step": 22180 }, { "grad_norm": 0.535825788974762, "learning_rate": 8.652050754311437e-05, "loss": 0.0461, "step": 22190 }, { "grad_norm": 0.5005944967269897, "learning_rate": 8.650638774302995e-05, "loss": 0.0624, "step": 22200 }, { "grad_norm": 1.8459248542785645, "learning_rate": 8.649226170500233e-05, "loss": 0.0757, "step": 22210 }, { "grad_norm": 0.48926281929016113, "learning_rate": 8.647812943144524e-05, "loss": 0.0655, "step": 22220 }, { "grad_norm": 0.5903764367103577, "learning_rate": 8.646399092477351e-05, "loss": 0.0545, "step": 22230 }, { "grad_norm": 0.5273777842521667, "learning_rate": 8.644984618740301e-05, "loss": 0.067, "step": 22240 }, { "grad_norm": 0.640543520450592, "learning_rate": 8.643569522175073e-05, "loss": 0.068, "step": 22250 }, { "grad_norm": 0.4362531900405884, "learning_rate": 8.642153803023463e-05, "loss": 0.0539, "step": 22260 }, { "grad_norm": 0.6334659457206726, "learning_rate": 8.640737461527383e-05, "loss": 0.0747, "step": 22270 }, { "grad_norm": 0.5102505683898926, "learning_rate": 8.639320497928845e-05, "loss": 0.055, "step": 22280 }, { "grad_norm": 0.38217008113861084, "learning_rate": 8.637902912469969e-05, "loss": 0.0419, "step": 22290 }, { "grad_norm": 0.5329813957214355, "learning_rate": 8.636484705392982e-05, "loss": 0.0527, "step": 22300 }, { "grad_norm": 0.616836428642273, "learning_rate": 8.635065876940216e-05, "loss": 0.0615, "step": 22310 }, { "grad_norm": 0.5697592496871948, "learning_rate": 8.633646427354112e-05, "loss": 0.05, "step": 22320 }, { "grad_norm": 0.6220093965530396, "learning_rate": 8.632226356877213e-05, "loss": 0.0554, "step": 22330 }, { "grad_norm": 0.5301994681358337, "learning_rate": 8.630805665752173e-05, "loss": 0.0517, "step": 22340 }, { "grad_norm": 0.6766137480735779, "learning_rate": 8.629384354221748e-05, "loss": 0.0622, "step": 22350 }, { "grad_norm": 0.47454679012298584, "learning_rate": 8.627962422528797e-05, "loss": 0.0583, "step": 22360 }, { "grad_norm": 0.5274328589439392, "learning_rate": 8.626539870916296e-05, "loss": 0.0723, "step": 22370 }, { "grad_norm": 0.7489970922470093, "learning_rate": 8.625116699627317e-05, "loss": 0.0544, "step": 22380 }, { "grad_norm": 0.5408119559288025, "learning_rate": 8.623692908905041e-05, "loss": 0.066, "step": 22390 }, { "grad_norm": 0.48033106327056885, "learning_rate": 8.622268498992755e-05, "loss": 0.0666, "step": 22400 }, { "grad_norm": 0.5901774764060974, "learning_rate": 8.620843470133851e-05, "loss": 0.0623, "step": 22410 }, { "grad_norm": 0.7365714311599731, "learning_rate": 8.619417822571829e-05, "loss": 0.0656, "step": 22420 }, { "grad_norm": 0.47011446952819824, "learning_rate": 8.617991556550292e-05, "loss": 0.0488, "step": 22430 }, { "grad_norm": 0.47887349128723145, "learning_rate": 8.616564672312952e-05, "loss": 0.0616, "step": 22440 }, { "grad_norm": 0.43724051117897034, "learning_rate": 8.61513717010362e-05, "loss": 0.0781, "step": 22450 }, { "grad_norm": 0.5377855896949768, "learning_rate": 8.613709050166221e-05, "loss": 0.0628, "step": 22460 }, { "grad_norm": 0.5465762615203857, "learning_rate": 8.61228031274478e-05, "loss": 0.0579, "step": 22470 }, { "grad_norm": 0.4662528932094574, "learning_rate": 8.610850958083431e-05, "loss": 0.055, "step": 22480 }, { "grad_norm": 0.41332006454467773, "learning_rate": 8.609420986426409e-05, "loss": 0.0569, "step": 22490 }, { "grad_norm": 0.5750963687896729, "learning_rate": 8.60799039801806e-05, "loss": 0.0636, "step": 22500 }, { "grad_norm": 0.4513961970806122, "learning_rate": 8.606559193102828e-05, "loss": 0.0387, "step": 22510 }, { "grad_norm": 0.7728062868118286, "learning_rate": 8.605127371925273e-05, "loss": 0.0616, "step": 22520 }, { "grad_norm": 0.7800347805023193, "learning_rate": 8.603694934730047e-05, "loss": 0.0674, "step": 22530 }, { "grad_norm": 0.4931769073009491, "learning_rate": 8.602261881761919e-05, "loss": 0.0478, "step": 22540 }, { "grad_norm": 0.5016546845436096, "learning_rate": 8.600828213265759e-05, "loss": 0.0493, "step": 22550 }, { "grad_norm": 0.7416279315948486, "learning_rate": 8.599393929486539e-05, "loss": 0.0788, "step": 22560 }, { "grad_norm": 0.5476364493370056, "learning_rate": 8.59795903066934e-05, "loss": 0.0565, "step": 22570 }, { "grad_norm": 0.540161669254303, "learning_rate": 8.596523517059347e-05, "loss": 0.048, "step": 22580 }, { "grad_norm": 0.4555046558380127, "learning_rate": 8.59508738890185e-05, "loss": 0.0485, "step": 22590 }, { "grad_norm": 0.530731737613678, "learning_rate": 8.593650646442246e-05, "loss": 0.0526, "step": 22600 }, { "grad_norm": 0.8381444215774536, "learning_rate": 8.59221328992603e-05, "loss": 0.064, "step": 22610 }, { "grad_norm": 0.6784690618515015, "learning_rate": 8.590775319598813e-05, "loss": 0.0604, "step": 22620 }, { "grad_norm": 0.5536420941352844, "learning_rate": 8.589336735706301e-05, "loss": 0.0554, "step": 22630 }, { "grad_norm": 1.0545117855072021, "learning_rate": 8.587897538494307e-05, "loss": 0.0696, "step": 22640 }, { "grad_norm": 0.6391228437423706, "learning_rate": 8.586457728208756e-05, "loss": 0.0504, "step": 22650 }, { "grad_norm": 0.5731256604194641, "learning_rate": 8.585017305095667e-05, "loss": 0.0593, "step": 22660 }, { "grad_norm": 0.44805672764778137, "learning_rate": 8.583576269401173e-05, "loss": 0.0454, "step": 22670 }, { "grad_norm": 0.6206335425376892, "learning_rate": 8.582134621371504e-05, "loss": 0.0598, "step": 22680 }, { "grad_norm": 0.6848139762878418, "learning_rate": 8.580692361253e-05, "loss": 0.0831, "step": 22690 }, { "grad_norm": 0.5603808164596558, "learning_rate": 8.579249489292104e-05, "loss": 0.074, "step": 22700 }, { "grad_norm": 0.647843062877655, "learning_rate": 8.577806005735363e-05, "loss": 0.0549, "step": 22710 }, { "grad_norm": 0.6809065937995911, "learning_rate": 8.576361910829429e-05, "loss": 0.0663, "step": 22720 }, { "grad_norm": 0.5712376832962036, "learning_rate": 8.574917204821057e-05, "loss": 0.0545, "step": 22730 }, { "grad_norm": 0.4088905155658722, "learning_rate": 8.57347188795711e-05, "loss": 0.0552, "step": 22740 }, { "grad_norm": 0.5490099787712097, "learning_rate": 8.572025960484551e-05, "loss": 0.0558, "step": 22750 }, { "grad_norm": 0.606931746006012, "learning_rate": 8.57057942265045e-05, "loss": 0.0551, "step": 22760 }, { "grad_norm": 0.5400859117507935, "learning_rate": 8.569132274701984e-05, "loss": 0.0812, "step": 22770 }, { "grad_norm": 0.7126957178115845, "learning_rate": 8.567684516886427e-05, "loss": 0.0555, "step": 22780 }, { "grad_norm": 0.8061597943305969, "learning_rate": 8.56623614945116e-05, "loss": 0.0573, "step": 22790 }, { "grad_norm": 0.48083367943763733, "learning_rate": 8.564787172643675e-05, "loss": 0.0701, "step": 22800 }, { "grad_norm": 0.6128987669944763, "learning_rate": 8.563337586711559e-05, "loss": 0.051, "step": 22810 }, { "grad_norm": 0.49508023262023926, "learning_rate": 8.561887391902506e-05, "loss": 0.0572, "step": 22820 }, { "grad_norm": 0.6172971725463867, "learning_rate": 8.560436588464316e-05, "loss": 0.0801, "step": 22830 }, { "grad_norm": 0.3375830054283142, "learning_rate": 8.55898517664489e-05, "loss": 0.0511, "step": 22840 }, { "grad_norm": 0.500312864780426, "learning_rate": 8.557533156692236e-05, "loss": 0.0611, "step": 22850 }, { "grad_norm": 0.6619619131088257, "learning_rate": 8.556080528854467e-05, "loss": 0.0634, "step": 22860 }, { "grad_norm": 0.3917389512062073, "learning_rate": 8.554627293379791e-05, "loss": 0.0471, "step": 22870 }, { "grad_norm": 0.39983516931533813, "learning_rate": 8.553173450516531e-05, "loss": 0.0492, "step": 22880 }, { "grad_norm": 0.4189278781414032, "learning_rate": 8.551719000513108e-05, "loss": 0.0376, "step": 22890 }, { "grad_norm": 0.5596674680709839, "learning_rate": 8.550263943618049e-05, "loss": 0.0544, "step": 22900 }, { "grad_norm": 0.533229649066925, "learning_rate": 8.54880828007998e-05, "loss": 0.08, "step": 22910 }, { "grad_norm": 0.9373416900634766, "learning_rate": 8.547352010147637e-05, "loss": 0.0529, "step": 22920 }, { "grad_norm": 0.4270705580711365, "learning_rate": 8.545895134069855e-05, "loss": 0.0511, "step": 22930 }, { "grad_norm": 0.5895360708236694, "learning_rate": 8.544437652095576e-05, "loss": 0.0637, "step": 22940 }, { "grad_norm": 0.6563377976417542, "learning_rate": 8.542979564473843e-05, "loss": 0.0606, "step": 22950 }, { "grad_norm": 0.6123634576797485, "learning_rate": 8.541520871453802e-05, "loss": 0.045, "step": 22960 }, { "grad_norm": 0.6372371912002563, "learning_rate": 8.540061573284705e-05, "loss": 0.0596, "step": 22970 }, { "grad_norm": 0.7557340860366821, "learning_rate": 8.538601670215906e-05, "loss": 0.0595, "step": 22980 }, { "grad_norm": 0.4791291654109955, "learning_rate": 8.537141162496864e-05, "loss": 0.043, "step": 22990 }, { "grad_norm": 0.5186588168144226, "learning_rate": 8.535680050377137e-05, "loss": 0.0641, "step": 23000 }, { "grad_norm": 0.42371290922164917, "learning_rate": 8.534218334106391e-05, "loss": 0.057, "step": 23010 }, { "grad_norm": 0.8624916672706604, "learning_rate": 8.532756013934393e-05, "loss": 0.0531, "step": 23020 }, { "grad_norm": 0.6940524578094482, "learning_rate": 8.531293090111012e-05, "loss": 0.0546, "step": 23030 }, { "grad_norm": 0.5938053131103516, "learning_rate": 8.529829562886225e-05, "loss": 0.0433, "step": 23040 }, { "grad_norm": 0.5955674052238464, "learning_rate": 8.528365432510105e-05, "loss": 0.0504, "step": 23050 }, { "grad_norm": 0.3767929971218109, "learning_rate": 8.526900699232833e-05, "loss": 0.0476, "step": 23060 }, { "grad_norm": 0.5335680842399597, "learning_rate": 8.525435363304695e-05, "loss": 0.0517, "step": 23070 }, { "grad_norm": 0.810087263584137, "learning_rate": 8.523969424976072e-05, "loss": 0.057, "step": 23080 }, { "grad_norm": 0.6162288188934326, "learning_rate": 8.522502884497457e-05, "loss": 0.0526, "step": 23090 }, { "grad_norm": 0.7123993039131165, "learning_rate": 8.521035742119437e-05, "loss": 0.058, "step": 23100 }, { "grad_norm": 0.49907925724983215, "learning_rate": 8.519567998092712e-05, "loss": 0.058, "step": 23110 }, { "grad_norm": 0.5160807371139526, "learning_rate": 8.518099652668075e-05, "loss": 0.0549, "step": 23120 }, { "grad_norm": 0.5104130506515503, "learning_rate": 8.516630706096429e-05, "loss": 0.0582, "step": 23130 }, { "grad_norm": 0.6174226999282837, "learning_rate": 8.515161158628773e-05, "loss": 0.0498, "step": 23140 }, { "grad_norm": 0.599396824836731, "learning_rate": 8.513691010516216e-05, "loss": 0.0574, "step": 23150 }, { "grad_norm": 0.687974214553833, "learning_rate": 8.512220262009966e-05, "loss": 0.0594, "step": 23160 }, { "grad_norm": 0.5687412023544312, "learning_rate": 8.510748913361332e-05, "loss": 0.0623, "step": 23170 }, { "grad_norm": 0.7477232813835144, "learning_rate": 8.509276964821726e-05, "loss": 0.0614, "step": 23180 }, { "grad_norm": 0.5941533446311951, "learning_rate": 8.507804416642669e-05, "loss": 0.0592, "step": 23190 }, { "grad_norm": 0.7408114671707153, "learning_rate": 8.506331269075774e-05, "loss": 0.0551, "step": 23200 }, { "grad_norm": 0.9461615085601807, "learning_rate": 8.504857522372765e-05, "loss": 0.0746, "step": 23210 }, { "grad_norm": 0.5058380365371704, "learning_rate": 8.503383176785461e-05, "loss": 0.0574, "step": 23220 }, { "grad_norm": 0.7406257390975952, "learning_rate": 8.501908232565792e-05, "loss": 0.0626, "step": 23230 }, { "grad_norm": 1.0045541524887085, "learning_rate": 8.50043268996578e-05, "loss": 0.0723, "step": 23240 }, { "grad_norm": 0.8496467471122742, "learning_rate": 8.498956549237562e-05, "loss": 0.0726, "step": 23250 }, { "grad_norm": 0.4219854772090912, "learning_rate": 8.497479810633366e-05, "loss": 0.0937, "step": 23260 }, { "grad_norm": 0.5064714550971985, "learning_rate": 8.496002474405525e-05, "loss": 0.0611, "step": 23270 }, { "grad_norm": 0.6411758065223694, "learning_rate": 8.494524540806478e-05, "loss": 0.0647, "step": 23280 }, { "grad_norm": 0.477873295545578, "learning_rate": 8.493046010088761e-05, "loss": 0.0621, "step": 23290 }, { "grad_norm": 0.5170873999595642, "learning_rate": 8.491566882505018e-05, "loss": 0.0589, "step": 23300 }, { "grad_norm": 0.613539457321167, "learning_rate": 8.490087158307988e-05, "loss": 0.0596, "step": 23310 }, { "grad_norm": 0.3841843008995056, "learning_rate": 8.488606837750518e-05, "loss": 0.0532, "step": 23320 }, { "grad_norm": 0.6658738851547241, "learning_rate": 8.487125921085552e-05, "loss": 0.0574, "step": 23330 }, { "grad_norm": 0.6622893214225769, "learning_rate": 8.485644408566141e-05, "loss": 0.0524, "step": 23340 }, { "grad_norm": 0.47709038853645325, "learning_rate": 8.484162300445431e-05, "loss": 0.0682, "step": 23350 }, { "grad_norm": 0.5488954186439514, "learning_rate": 8.482679596976676e-05, "loss": 0.0547, "step": 23360 }, { "grad_norm": 0.7186574339866638, "learning_rate": 8.48119629841323e-05, "loss": 0.0626, "step": 23370 }, { "grad_norm": 0.4950767159461975, "learning_rate": 8.479712405008547e-05, "loss": 0.0652, "step": 23380 }, { "grad_norm": 0.6520513892173767, "learning_rate": 8.478227917016184e-05, "loss": 0.0611, "step": 23390 }, { "grad_norm": 0.5747976303100586, "learning_rate": 8.476742834689801e-05, "loss": 0.0556, "step": 23400 }, { "grad_norm": 0.6600662469863892, "learning_rate": 8.475257158283157e-05, "loss": 0.0673, "step": 23410 }, { "grad_norm": 0.6407651305198669, "learning_rate": 8.473770888050112e-05, "loss": 0.0596, "step": 23420 }, { "grad_norm": 0.8242456316947937, "learning_rate": 8.47228402424463e-05, "loss": 0.0565, "step": 23430 }, { "grad_norm": 0.5660633444786072, "learning_rate": 8.470796567120775e-05, "loss": 0.0566, "step": 23440 }, { "grad_norm": 0.7098948359489441, "learning_rate": 8.469308516932714e-05, "loss": 0.0539, "step": 23450 }, { "grad_norm": 0.7133013010025024, "learning_rate": 8.467819873934714e-05, "loss": 0.0528, "step": 23460 }, { "grad_norm": 0.4496493339538574, "learning_rate": 8.466330638381143e-05, "loss": 0.0581, "step": 23470 }, { "grad_norm": 0.9842579364776611, "learning_rate": 8.464840810526469e-05, "loss": 0.0688, "step": 23480 }, { "grad_norm": 0.6148020625114441, "learning_rate": 8.463350390625264e-05, "loss": 0.0401, "step": 23490 }, { "grad_norm": 0.7450087666511536, "learning_rate": 8.4618593789322e-05, "loss": 0.0854, "step": 23500 }, { "grad_norm": 0.3545961081981659, "learning_rate": 8.46036777570205e-05, "loss": 0.0577, "step": 23510 }, { "grad_norm": 0.5919432044029236, "learning_rate": 8.458875581189688e-05, "loss": 0.0567, "step": 23520 }, { "grad_norm": 0.6337258815765381, "learning_rate": 8.457382795650092e-05, "loss": 0.0666, "step": 23530 }, { "grad_norm": 0.6382520198822021, "learning_rate": 8.455889419338335e-05, "loss": 0.0493, "step": 23540 }, { "grad_norm": 0.42704975605010986, "learning_rate": 8.454395452509593e-05, "loss": 0.049, "step": 23550 }, { "grad_norm": 0.7916044592857361, "learning_rate": 8.452900895419146e-05, "loss": 0.0528, "step": 23560 }, { "grad_norm": 1.2121002674102783, "learning_rate": 8.451405748322376e-05, "loss": 0.0469, "step": 23570 }, { "grad_norm": 0.3551985025405884, "learning_rate": 8.449910011474759e-05, "loss": 0.0607, "step": 23580 }, { "grad_norm": 0.41923609375953674, "learning_rate": 8.448413685131876e-05, "loss": 0.0527, "step": 23590 }, { "grad_norm": 0.41523975133895874, "learning_rate": 8.446916769549407e-05, "loss": 0.045, "step": 23600 }, { "grad_norm": 0.5005254745483398, "learning_rate": 8.445419264983136e-05, "loss": 0.0775, "step": 23610 }, { "grad_norm": 0.6331596374511719, "learning_rate": 8.443921171688947e-05, "loss": 0.0637, "step": 23620 }, { "grad_norm": 0.5982899069786072, "learning_rate": 8.442422489922819e-05, "loss": 0.0699, "step": 23630 }, { "grad_norm": 0.4725753664970398, "learning_rate": 8.440923219940838e-05, "loss": 0.0422, "step": 23640 }, { "grad_norm": 0.561942458152771, "learning_rate": 8.439423361999189e-05, "loss": 0.0498, "step": 23650 }, { "grad_norm": 0.6000518202781677, "learning_rate": 8.437922916354155e-05, "loss": 0.0866, "step": 23660 }, { "grad_norm": 0.5412598848342896, "learning_rate": 8.436421883262123e-05, "loss": 0.0567, "step": 23670 }, { "grad_norm": 0.5163037776947021, "learning_rate": 8.434920262979577e-05, "loss": 0.07, "step": 23680 }, { "grad_norm": 0.9385006427764893, "learning_rate": 8.433418055763104e-05, "loss": 0.0615, "step": 23690 }, { "grad_norm": 0.4287985563278198, "learning_rate": 8.431915261869389e-05, "loss": 0.0553, "step": 23700 }, { "grad_norm": 0.6356627941131592, "learning_rate": 8.43041188155522e-05, "loss": 0.0675, "step": 23710 }, { "grad_norm": 0.5439733862876892, "learning_rate": 8.428907915077481e-05, "loss": 0.0461, "step": 23720 }, { "grad_norm": 0.6801834106445312, "learning_rate": 8.42740336269316e-05, "loss": 0.0537, "step": 23730 }, { "grad_norm": 0.33847254514694214, "learning_rate": 8.425898224659345e-05, "loss": 0.0527, "step": 23740 }, { "grad_norm": 0.41197073459625244, "learning_rate": 8.42439250123322e-05, "loss": 0.0668, "step": 23750 }, { "grad_norm": 0.6165210604667664, "learning_rate": 8.422886192672076e-05, "loss": 0.0494, "step": 23760 }, { "grad_norm": 0.5924456119537354, "learning_rate": 8.421379299233297e-05, "loss": 0.0493, "step": 23770 }, { "grad_norm": 0.4775460660457611, "learning_rate": 8.419871821174371e-05, "loss": 0.0509, "step": 23780 }, { "grad_norm": 0.6084215044975281, "learning_rate": 8.418363758752884e-05, "loss": 0.048, "step": 23790 }, { "grad_norm": 0.41029995679855347, "learning_rate": 8.416855112226523e-05, "loss": 0.0489, "step": 23800 }, { "grad_norm": 0.6424959301948547, "learning_rate": 8.415345881853075e-05, "loss": 0.0643, "step": 23810 }, { "grad_norm": 0.5804826021194458, "learning_rate": 8.413836067890426e-05, "loss": 0.0614, "step": 23820 }, { "grad_norm": 0.49525755643844604, "learning_rate": 8.41232567059656e-05, "loss": 0.0474, "step": 23830 }, { "grad_norm": 0.5810272693634033, "learning_rate": 8.410814690229565e-05, "loss": 0.0577, "step": 23840 }, { "grad_norm": 0.5094470977783203, "learning_rate": 8.409303127047626e-05, "loss": 0.0598, "step": 23850 }, { "grad_norm": 0.4458054006099701, "learning_rate": 8.407790981309028e-05, "loss": 0.0499, "step": 23860 }, { "grad_norm": 1.049494743347168, "learning_rate": 8.406278253272153e-05, "loss": 0.0568, "step": 23870 }, { "grad_norm": 0.5025987029075623, "learning_rate": 8.404764943195487e-05, "loss": 0.0629, "step": 23880 }, { "grad_norm": 0.5502660870552063, "learning_rate": 8.403251051337613e-05, "loss": 0.0447, "step": 23890 }, { "grad_norm": 0.3926081955432892, "learning_rate": 8.401736577957214e-05, "loss": 0.0484, "step": 23900 }, { "grad_norm": 0.6813040375709534, "learning_rate": 8.40022152331307e-05, "loss": 0.0451, "step": 23910 }, { "grad_norm": 0.6084557771682739, "learning_rate": 8.398705887664064e-05, "loss": 0.0588, "step": 23920 }, { "grad_norm": 0.7227733731269836, "learning_rate": 8.397189671269177e-05, "loss": 0.0503, "step": 23930 }, { "grad_norm": 0.3557947278022766, "learning_rate": 8.395672874387488e-05, "loss": 0.0587, "step": 23940 }, { "grad_norm": 0.5816057324409485, "learning_rate": 8.394155497278177e-05, "loss": 0.065, "step": 23950 }, { "grad_norm": 0.5364567637443542, "learning_rate": 8.392637540200523e-05, "loss": 0.0583, "step": 23960 }, { "grad_norm": 0.7094352841377258, "learning_rate": 8.391119003413902e-05, "loss": 0.0571, "step": 23970 }, { "grad_norm": 0.7960895299911499, "learning_rate": 8.38959988717779e-05, "loss": 0.0632, "step": 23980 }, { "grad_norm": 0.6374249458312988, "learning_rate": 8.388080191751764e-05, "loss": 0.0667, "step": 23990 }, { "grad_norm": 0.5979352593421936, "learning_rate": 8.386559917395496e-05, "loss": 0.0595, "step": 24000 }, { "grad_norm": 0.5819076299667358, "learning_rate": 8.385039064368761e-05, "loss": 0.0586, "step": 24010 }, { "grad_norm": 0.6150839924812317, "learning_rate": 8.383517632931431e-05, "loss": 0.054, "step": 24020 }, { "grad_norm": 0.6069976091384888, "learning_rate": 8.381995623343477e-05, "loss": 0.0627, "step": 24030 }, { "grad_norm": 0.41790568828582764, "learning_rate": 8.380473035864968e-05, "loss": 0.0511, "step": 24040 }, { "grad_norm": 0.7155181169509888, "learning_rate": 8.378949870756076e-05, "loss": 0.0646, "step": 24050 }, { "grad_norm": 0.5896844863891602, "learning_rate": 8.377426128277063e-05, "loss": 0.0468, "step": 24060 }, { "grad_norm": 0.8448485732078552, "learning_rate": 8.375901808688298e-05, "loss": 0.0597, "step": 24070 }, { "grad_norm": 0.5231429934501648, "learning_rate": 8.374376912250246e-05, "loss": 0.0633, "step": 24080 }, { "grad_norm": 0.5201070308685303, "learning_rate": 8.372851439223468e-05, "loss": 0.054, "step": 24090 }, { "grad_norm": 0.5607741475105286, "learning_rate": 8.371325389868627e-05, "loss": 0.0575, "step": 24100 }, { "grad_norm": 0.5615660548210144, "learning_rate": 8.369798764446482e-05, "loss": 0.0614, "step": 24110 }, { "grad_norm": 0.5345473289489746, "learning_rate": 8.368271563217893e-05, "loss": 0.0592, "step": 24120 }, { "grad_norm": 0.4966062307357788, "learning_rate": 8.366743786443817e-05, "loss": 0.0675, "step": 24130 }, { "grad_norm": 0.47758936882019043, "learning_rate": 8.365215434385309e-05, "loss": 0.0828, "step": 24140 }, { "grad_norm": 0.7443968653678894, "learning_rate": 8.36368650730352e-05, "loss": 0.0598, "step": 24150 }, { "grad_norm": 0.5556560754776001, "learning_rate": 8.362157005459705e-05, "loss": 0.0677, "step": 24160 }, { "grad_norm": 0.42503994703292847, "learning_rate": 8.360626929115213e-05, "loss": 0.0597, "step": 24170 }, { "grad_norm": 0.720920205116272, "learning_rate": 8.359096278531492e-05, "loss": 0.0551, "step": 24180 }, { "grad_norm": 0.7873311042785645, "learning_rate": 8.357565053970088e-05, "loss": 0.0638, "step": 24190 }, { "grad_norm": 0.512458324432373, "learning_rate": 8.356033255692647e-05, "loss": 0.0635, "step": 24200 }, { "grad_norm": 0.5894606113433838, "learning_rate": 8.354500883960911e-05, "loss": 0.0619, "step": 24210 }, { "grad_norm": 0.5696331858634949, "learning_rate": 8.352967939036717e-05, "loss": 0.054, "step": 24220 }, { "grad_norm": 0.43779686093330383, "learning_rate": 8.35143442118201e-05, "loss": 0.0465, "step": 24230 }, { "grad_norm": 0.6410816311836243, "learning_rate": 8.349900330658819e-05, "loss": 0.0534, "step": 24240 }, { "grad_norm": 1.0555534362792969, "learning_rate": 8.348365667729284e-05, "loss": 0.05, "step": 24250 }, { "grad_norm": 0.46443089842796326, "learning_rate": 8.346830432655633e-05, "loss": 0.0729, "step": 24260 }, { "grad_norm": 0.5957905054092407, "learning_rate": 8.345294625700195e-05, "loss": 0.0604, "step": 24270 }, { "grad_norm": 0.6315454840660095, "learning_rate": 8.343758247125402e-05, "loss": 0.0483, "step": 24280 }, { "grad_norm": 0.5120254158973694, "learning_rate": 8.342221297193776e-05, "loss": 0.0543, "step": 24290 }, { "grad_norm": 0.5312827229499817, "learning_rate": 8.34068377616794e-05, "loss": 0.0505, "step": 24300 }, { "grad_norm": 0.6026142835617065, "learning_rate": 8.339145684310615e-05, "loss": 0.0606, "step": 24310 }, { "grad_norm": 0.43746066093444824, "learning_rate": 8.337607021884618e-05, "loss": 0.0439, "step": 24320 }, { "grad_norm": 0.7536316514015198, "learning_rate": 8.336067789152867e-05, "loss": 0.0608, "step": 24330 }, { "grad_norm": 0.4394803047180176, "learning_rate": 8.334527986378369e-05, "loss": 0.0555, "step": 24340 }, { "grad_norm": 0.4925103485584259, "learning_rate": 8.332987613824239e-05, "loss": 0.0492, "step": 24350 }, { "grad_norm": 0.7338787317276001, "learning_rate": 8.331446671753685e-05, "loss": 0.0807, "step": 24360 }, { "grad_norm": 0.4037250876426697, "learning_rate": 8.329905160430007e-05, "loss": 0.0519, "step": 24370 }, { "grad_norm": 0.7647895216941833, "learning_rate": 8.328363080116611e-05, "loss": 0.0673, "step": 24380 }, { "grad_norm": 0.6616246700286865, "learning_rate": 8.326820431076997e-05, "loss": 0.0815, "step": 24390 }, { "grad_norm": 0.5129621028900146, "learning_rate": 8.325277213574759e-05, "loss": 0.0582, "step": 24400 }, { "grad_norm": 0.42188164591789246, "learning_rate": 8.32373342787359e-05, "loss": 0.0843, "step": 24410 }, { "grad_norm": 0.5605700612068176, "learning_rate": 8.322189074237285e-05, "loss": 0.0523, "step": 24420 }, { "grad_norm": 0.5618084669113159, "learning_rate": 8.32064415292973e-05, "loss": 0.0506, "step": 24430 }, { "grad_norm": 0.5317719578742981, "learning_rate": 8.319098664214907e-05, "loss": 0.0571, "step": 24440 }, { "grad_norm": 0.7799975275993347, "learning_rate": 8.3175526083569e-05, "loss": 0.064, "step": 24450 }, { "grad_norm": 0.5844513773918152, "learning_rate": 8.316005985619889e-05, "loss": 0.0581, "step": 24460 }, { "grad_norm": 0.5896006226539612, "learning_rate": 8.314458796268147e-05, "loss": 0.0579, "step": 24470 }, { "grad_norm": 0.5555753707885742, "learning_rate": 8.312911040566047e-05, "loss": 0.0673, "step": 24480 }, { "grad_norm": 0.6063089370727539, "learning_rate": 8.31136271877806e-05, "loss": 0.0771, "step": 24490 }, { "grad_norm": 0.41066625714302063, "learning_rate": 8.309813831168748e-05, "loss": 0.0605, "step": 24500 }, { "grad_norm": 0.3992278575897217, "learning_rate": 8.308264378002777e-05, "loss": 0.0805, "step": 24510 }, { "grad_norm": 0.4804646968841553, "learning_rate": 8.306714359544906e-05, "loss": 0.0587, "step": 24520 }, { "grad_norm": 0.596579372882843, "learning_rate": 8.30516377605999e-05, "loss": 0.0627, "step": 24530 }, { "grad_norm": 0.5118683576583862, "learning_rate": 8.30361262781298e-05, "loss": 0.0527, "step": 24540 }, { "grad_norm": 0.3461325466632843, "learning_rate": 8.302060915068924e-05, "loss": 0.0479, "step": 24550 }, { "grad_norm": 1.1047111749649048, "learning_rate": 8.300508638092972e-05, "loss": 0.0519, "step": 24560 }, { "grad_norm": 0.5190373659133911, "learning_rate": 8.298955797150361e-05, "loss": 0.059, "step": 24570 }, { "grad_norm": 0.6369159817695618, "learning_rate": 8.297402392506433e-05, "loss": 0.0553, "step": 24580 }, { "grad_norm": 0.6546006798744202, "learning_rate": 8.295848424426617e-05, "loss": 0.0446, "step": 24590 }, { "grad_norm": 0.5808834433555603, "learning_rate": 8.29429389317645e-05, "loss": 0.0429, "step": 24600 }, { "grad_norm": 0.5568984150886536, "learning_rate": 8.292738799021556e-05, "loss": 0.0601, "step": 24610 }, { "grad_norm": 0.5537000298500061, "learning_rate": 8.291183142227656e-05, "loss": 0.0432, "step": 24620 }, { "grad_norm": 0.6946290731430054, "learning_rate": 8.289626923060572e-05, "loss": 0.0585, "step": 24630 }, { "grad_norm": 0.5257387161254883, "learning_rate": 8.288070141786218e-05, "loss": 0.0484, "step": 24640 }, { "grad_norm": 0.4373616576194763, "learning_rate": 8.286512798670605e-05, "loss": 0.0387, "step": 24650 }, { "grad_norm": 1.044184684753418, "learning_rate": 8.284954893979842e-05, "loss": 0.0537, "step": 24660 }, { "grad_norm": 0.5219815969467163, "learning_rate": 8.283396427980131e-05, "loss": 0.0555, "step": 24670 }, { "grad_norm": 0.5425310134887695, "learning_rate": 8.281837400937771e-05, "loss": 0.0615, "step": 24680 }, { "grad_norm": 0.6285703182220459, "learning_rate": 8.28027781311916e-05, "loss": 0.0589, "step": 24690 }, { "grad_norm": 0.5796401500701904, "learning_rate": 8.278717664790785e-05, "loss": 0.0635, "step": 24700 }, { "grad_norm": 0.5572880506515503, "learning_rate": 8.277156956219234e-05, "loss": 0.0731, "step": 24710 }, { "grad_norm": 0.8660666346549988, "learning_rate": 8.275595687671189e-05, "loss": 0.0489, "step": 24720 }, { "grad_norm": 0.49050426483154297, "learning_rate": 8.27403385941343e-05, "loss": 0.0515, "step": 24730 }, { "grad_norm": 0.8671559691429138, "learning_rate": 8.272471471712828e-05, "loss": 0.0487, "step": 24740 }, { "grad_norm": 0.5232312679290771, "learning_rate": 8.270908524836355e-05, "loss": 0.0537, "step": 24750 }, { "grad_norm": 0.5793764591217041, "learning_rate": 8.269345019051074e-05, "loss": 0.0505, "step": 24760 }, { "grad_norm": 0.8399308919906616, "learning_rate": 8.267780954624147e-05, "loss": 0.0559, "step": 24770 }, { "grad_norm": 0.5741423964500427, "learning_rate": 8.266216331822827e-05, "loss": 0.0616, "step": 24780 }, { "grad_norm": 0.6861943602561951, "learning_rate": 8.264651150914469e-05, "loss": 0.0716, "step": 24790 }, { "grad_norm": 0.6702383756637573, "learning_rate": 8.263085412166517e-05, "loss": 0.0661, "step": 24800 }, { "grad_norm": 0.46188899874687195, "learning_rate": 8.261519115846514e-05, "loss": 0.0537, "step": 24810 }, { "grad_norm": 0.6334993839263916, "learning_rate": 8.259952262222096e-05, "loss": 0.0656, "step": 24820 }, { "grad_norm": 0.43874335289001465, "learning_rate": 8.258384851560997e-05, "loss": 0.0426, "step": 24830 }, { "grad_norm": 0.567783772945404, "learning_rate": 8.256816884131044e-05, "loss": 0.0459, "step": 24840 }, { "grad_norm": 0.5634710192680359, "learning_rate": 8.255248360200159e-05, "loss": 0.0569, "step": 24850 }, { "grad_norm": 0.8724627494812012, "learning_rate": 8.253679280036359e-05, "loss": 0.0792, "step": 24860 }, { "grad_norm": 0.6418820023536682, "learning_rate": 8.252109643907762e-05, "loss": 0.0696, "step": 24870 }, { "grad_norm": 0.7649175524711609, "learning_rate": 8.250539452082569e-05, "loss": 0.0575, "step": 24880 }, { "grad_norm": 0.7244877815246582, "learning_rate": 8.248968704829087e-05, "loss": 0.0567, "step": 24890 }, { "grad_norm": 0.46329736709594727, "learning_rate": 8.247397402415714e-05, "loss": 0.0531, "step": 24900 }, { "grad_norm": 0.5810317993164062, "learning_rate": 8.24582554511094e-05, "loss": 0.0474, "step": 24910 }, { "grad_norm": 0.4989296793937683, "learning_rate": 8.244253133183355e-05, "loss": 0.0717, "step": 24920 }, { "grad_norm": 0.551935613155365, "learning_rate": 8.24268016690164e-05, "loss": 0.0521, "step": 24930 }, { "grad_norm": 0.5189571976661682, "learning_rate": 8.241106646534571e-05, "loss": 0.0474, "step": 24940 }, { "grad_norm": 0.4157083034515381, "learning_rate": 8.23953257235102e-05, "loss": 0.0419, "step": 24950 }, { "grad_norm": 1.1579639911651611, "learning_rate": 8.237957944619956e-05, "loss": 0.0441, "step": 24960 }, { "grad_norm": 0.497502863407135, "learning_rate": 8.236382763610437e-05, "loss": 0.052, "step": 24970 }, { "grad_norm": 0.5402418375015259, "learning_rate": 8.234807029591619e-05, "loss": 0.0469, "step": 24980 }, { "grad_norm": 0.5708376169204712, "learning_rate": 8.233230742832752e-05, "loss": 0.0613, "step": 24990 }, { "grad_norm": 0.7377601861953735, "learning_rate": 8.231653903603178e-05, "loss": 0.0523, "step": 25000 }, { "grad_norm": 0.7631035447120667, "learning_rate": 8.23007651217234e-05, "loss": 0.05, "step": 25010 }, { "grad_norm": 0.7632380127906799, "learning_rate": 8.228498568809769e-05, "loss": 0.0754, "step": 25020 }, { "grad_norm": 0.8439436554908752, "learning_rate": 8.22692007378509e-05, "loss": 0.0633, "step": 25030 }, { "grad_norm": 0.4925805330276489, "learning_rate": 8.225341027368028e-05, "loss": 0.0567, "step": 25040 }, { "grad_norm": 0.6060549020767212, "learning_rate": 8.223761429828399e-05, "loss": 0.0612, "step": 25050 }, { "grad_norm": 0.3503388464450836, "learning_rate": 8.22218128143611e-05, "loss": 0.0564, "step": 25060 }, { "grad_norm": 0.3683084547519684, "learning_rate": 8.220600582461166e-05, "loss": 0.076, "step": 25070 }, { "grad_norm": 0.44881200790405273, "learning_rate": 8.219019333173668e-05, "loss": 0.062, "step": 25080 }, { "grad_norm": 0.6626182198524475, "learning_rate": 8.217437533843805e-05, "loss": 0.0641, "step": 25090 }, { "grad_norm": 0.6252796053886414, "learning_rate": 8.215855184741867e-05, "loss": 0.0494, "step": 25100 }, { "grad_norm": 0.4115295112133026, "learning_rate": 8.21427228613823e-05, "loss": 0.0571, "step": 25110 }, { "grad_norm": 0.6875037550926208, "learning_rate": 8.21268883830337e-05, "loss": 0.0591, "step": 25120 }, { "grad_norm": 0.5652124285697937, "learning_rate": 8.211104841507855e-05, "loss": 0.0594, "step": 25130 }, { "grad_norm": 0.5039250254631042, "learning_rate": 8.209520296022346e-05, "loss": 0.0596, "step": 25140 }, { "grad_norm": 0.6998279690742493, "learning_rate": 8.207935202117599e-05, "loss": 0.0545, "step": 25150 }, { "grad_norm": 0.5065485239028931, "learning_rate": 8.206349560064463e-05, "loss": 0.0478, "step": 25160 }, { "grad_norm": 0.5937326550483704, "learning_rate": 8.204763370133881e-05, "loss": 0.0785, "step": 25170 }, { "grad_norm": 0.7849785685539246, "learning_rate": 8.203176632596892e-05, "loss": 0.0431, "step": 25180 }, { "grad_norm": 0.5316458940505981, "learning_rate": 8.20158934772462e-05, "loss": 0.0526, "step": 25190 }, { "grad_norm": 1.0476276874542236, "learning_rate": 8.200001515788294e-05, "loss": 0.0599, "step": 25200 }, { "grad_norm": 0.5064131617546082, "learning_rate": 8.198413137059228e-05, "loss": 0.0408, "step": 25210 }, { "grad_norm": 0.620439887046814, "learning_rate": 8.196824211808835e-05, "loss": 0.0581, "step": 25220 }, { "grad_norm": 0.5878111720085144, "learning_rate": 8.195234740308617e-05, "loss": 0.0544, "step": 25230 }, { "grad_norm": 0.6113716959953308, "learning_rate": 8.193644722830171e-05, "loss": 0.0569, "step": 25240 }, { "grad_norm": 0.689369797706604, "learning_rate": 8.19205415964519e-05, "loss": 0.0517, "step": 25250 }, { "grad_norm": 0.7091355323791504, "learning_rate": 8.190463051025456e-05, "loss": 0.0502, "step": 25260 }, { "grad_norm": 0.7642145752906799, "learning_rate": 8.188871397242843e-05, "loss": 0.0633, "step": 25270 }, { "grad_norm": 0.5595517754554749, "learning_rate": 8.187279198569326e-05, "loss": 0.0767, "step": 25280 }, { "grad_norm": 0.8051742315292358, "learning_rate": 8.185686455276966e-05, "loss": 0.0575, "step": 25290 }, { "grad_norm": 0.59253990650177, "learning_rate": 8.184093167637921e-05, "loss": 0.0459, "step": 25300 }, { "grad_norm": 0.4244091212749481, "learning_rate": 8.182499335924437e-05, "loss": 0.0341, "step": 25310 }, { "grad_norm": 0.5342060923576355, "learning_rate": 8.18090496040886e-05, "loss": 0.0711, "step": 25320 }, { "grad_norm": 0.5839393138885498, "learning_rate": 8.179310041363621e-05, "loss": 0.0541, "step": 25330 }, { "grad_norm": 0.8687857389450073, "learning_rate": 8.17771457906125e-05, "loss": 0.064, "step": 25340 }, { "grad_norm": 0.4406833350658417, "learning_rate": 8.176118573774371e-05, "loss": 0.0516, "step": 25350 }, { "grad_norm": 0.4493158459663391, "learning_rate": 8.174522025775692e-05, "loss": 0.0764, "step": 25360 }, { "grad_norm": 0.5779921412467957, "learning_rate": 8.172924935338022e-05, "loss": 0.0513, "step": 25370 }, { "grad_norm": 0.6505154371261597, "learning_rate": 8.171327302734262e-05, "loss": 0.0675, "step": 25380 }, { "grad_norm": 0.6119747161865234, "learning_rate": 8.169729128237401e-05, "loss": 0.0529, "step": 25390 }, { "grad_norm": 0.5877448916435242, "learning_rate": 8.168130412120525e-05, "loss": 0.0515, "step": 25400 }, { "grad_norm": 0.5675383806228638, "learning_rate": 8.16653115465681e-05, "loss": 0.0479, "step": 25410 }, { "grad_norm": 0.7369139194488525, "learning_rate": 8.164931356119526e-05, "loss": 0.0555, "step": 25420 }, { "grad_norm": 0.6528488993644714, "learning_rate": 8.163331016782032e-05, "loss": 0.0492, "step": 25430 }, { "grad_norm": 0.7589999437332153, "learning_rate": 8.161730136917785e-05, "loss": 0.0481, "step": 25440 }, { "grad_norm": 0.5671908855438232, "learning_rate": 8.160128716800333e-05, "loss": 0.0381, "step": 25450 }, { "grad_norm": 0.579562246799469, "learning_rate": 8.158526756703313e-05, "loss": 0.0549, "step": 25460 }, { "grad_norm": 0.8204660415649414, "learning_rate": 8.156924256900455e-05, "loss": 0.0578, "step": 25470 }, { "grad_norm": 0.345951646566391, "learning_rate": 8.155321217665584e-05, "loss": 0.0538, "step": 25480 }, { "grad_norm": 0.7872212529182434, "learning_rate": 8.153717639272614e-05, "loss": 0.0518, "step": 25490 }, { "grad_norm": 0.4326138198375702, "learning_rate": 8.152113521995555e-05, "loss": 0.0559, "step": 25500 }, { "grad_norm": 0.8012291789054871, "learning_rate": 8.150508866108505e-05, "loss": 0.0736, "step": 25510 }, { "grad_norm": 1.1143429279327393, "learning_rate": 8.148903671885657e-05, "loss": 0.0784, "step": 25520 }, { "grad_norm": 0.6157563924789429, "learning_rate": 8.147297939601292e-05, "loss": 0.0789, "step": 25530 }, { "grad_norm": 0.3589949905872345, "learning_rate": 8.145691669529792e-05, "loss": 0.052, "step": 25540 }, { "grad_norm": 0.7005326151847839, "learning_rate": 8.144084861945618e-05, "loss": 0.0513, "step": 25550 }, { "grad_norm": 0.5552548170089722, "learning_rate": 8.142477517123333e-05, "loss": 0.0603, "step": 25560 }, { "grad_norm": 0.7772814035415649, "learning_rate": 8.140869635337586e-05, "loss": 0.0761, "step": 25570 }, { "grad_norm": 0.4270360767841339, "learning_rate": 8.139261216863123e-05, "loss": 0.0416, "step": 25580 }, { "grad_norm": 0.45653074979782104, "learning_rate": 8.137652261974776e-05, "loss": 0.0487, "step": 25590 }, { "grad_norm": 0.6909067630767822, "learning_rate": 8.136042770947472e-05, "loss": 0.0571, "step": 25600 }, { "grad_norm": 0.5330705046653748, "learning_rate": 8.134432744056228e-05, "loss": 0.0513, "step": 25610 }, { "grad_norm": 0.48973187804222107, "learning_rate": 8.132822181576158e-05, "loss": 0.0564, "step": 25620 }, { "grad_norm": 0.415033757686615, "learning_rate": 8.131211083782459e-05, "loss": 0.0542, "step": 25630 }, { "grad_norm": 0.397207111120224, "learning_rate": 8.129599450950424e-05, "loss": 0.0514, "step": 25640 }, { "grad_norm": 0.6505996584892273, "learning_rate": 8.127987283355438e-05, "loss": 0.0609, "step": 25650 }, { "grad_norm": 0.9559144377708435, "learning_rate": 8.126374581272976e-05, "loss": 0.0508, "step": 25660 }, { "grad_norm": 0.5805586576461792, "learning_rate": 8.124761344978605e-05, "loss": 0.0462, "step": 25670 }, { "grad_norm": 0.5086453557014465, "learning_rate": 8.12314757474798e-05, "loss": 0.0452, "step": 25680 }, { "grad_norm": 0.5719674825668335, "learning_rate": 8.121533270856856e-05, "loss": 0.0503, "step": 25690 }, { "grad_norm": 0.5023298263549805, "learning_rate": 8.119918433581069e-05, "loss": 0.044, "step": 25700 }, { "grad_norm": 0.7558231949806213, "learning_rate": 8.118303063196551e-05, "loss": 0.0736, "step": 25710 }, { "grad_norm": 0.5997403264045715, "learning_rate": 8.116687159979326e-05, "loss": 0.043, "step": 25720 }, { "grad_norm": 0.4242359399795532, "learning_rate": 8.115070724205508e-05, "loss": 0.044, "step": 25730 }, { "grad_norm": 0.7596864104270935, "learning_rate": 8.113453756151296e-05, "loss": 0.0467, "step": 25740 }, { "grad_norm": 0.849144458770752, "learning_rate": 8.111836256092995e-05, "loss": 0.0553, "step": 25750 }, { "grad_norm": 0.6943283081054688, "learning_rate": 8.110218224306985e-05, "loss": 0.0455, "step": 25760 }, { "grad_norm": 0.3396516442298889, "learning_rate": 8.108599661069745e-05, "loss": 0.0489, "step": 25770 }, { "grad_norm": 0.4671943485736847, "learning_rate": 8.106980566657845e-05, "loss": 0.0475, "step": 25780 }, { "grad_norm": 0.5660430788993835, "learning_rate": 8.10536094134794e-05, "loss": 0.0662, "step": 25790 }, { "grad_norm": 0.3550562858581543, "learning_rate": 8.103740785416783e-05, "loss": 0.0485, "step": 25800 }, { "grad_norm": 1.0315572023391724, "learning_rate": 8.102120099141212e-05, "loss": 0.0541, "step": 25810 }, { "grad_norm": 0.3892517387866974, "learning_rate": 8.100498882798163e-05, "loss": 0.045, "step": 25820 }, { "grad_norm": 0.5457439422607422, "learning_rate": 8.09887713666465e-05, "loss": 0.0518, "step": 25830 }, { "grad_norm": 0.5537508726119995, "learning_rate": 8.09725486101779e-05, "loss": 0.0468, "step": 25840 }, { "grad_norm": 0.6502188444137573, "learning_rate": 8.095632056134784e-05, "loss": 0.0474, "step": 25850 }, { "grad_norm": 0.6175304651260376, "learning_rate": 8.094008722292925e-05, "loss": 0.0485, "step": 25860 }, { "grad_norm": 0.4410676956176758, "learning_rate": 8.092384859769598e-05, "loss": 0.0429, "step": 25870 }, { "grad_norm": 0.6440999507904053, "learning_rate": 8.090760468842275e-05, "loss": 0.0491, "step": 25880 }, { "grad_norm": 0.41992390155792236, "learning_rate": 8.089135549788521e-05, "loss": 0.0464, "step": 25890 }, { "grad_norm": 0.33282071352005005, "learning_rate": 8.087510102885987e-05, "loss": 0.0366, "step": 25900 }, { "grad_norm": 0.46372783184051514, "learning_rate": 8.085884128412422e-05, "loss": 0.0447, "step": 25910 }, { "grad_norm": 0.775658369064331, "learning_rate": 8.084257626645659e-05, "loss": 0.0467, "step": 25920 }, { "grad_norm": 0.6846922039985657, "learning_rate": 8.08263059786362e-05, "loss": 0.0603, "step": 25930 }, { "grad_norm": 0.5267491340637207, "learning_rate": 8.081003042344325e-05, "loss": 0.0442, "step": 25940 }, { "grad_norm": 0.40600845217704773, "learning_rate": 8.079374960365872e-05, "loss": 0.0441, "step": 25950 }, { "grad_norm": 0.6155325770378113, "learning_rate": 8.077746352206463e-05, "loss": 0.0512, "step": 25960 }, { "grad_norm": 0.4192092716693878, "learning_rate": 8.076117218144377e-05, "loss": 0.0461, "step": 25970 }, { "grad_norm": 0.45426151156425476, "learning_rate": 8.074487558457991e-05, "loss": 0.05, "step": 25980 }, { "grad_norm": 0.7073601484298706, "learning_rate": 8.072857373425768e-05, "loss": 0.063, "step": 25990 }, { "grad_norm": 0.4850989878177643, "learning_rate": 8.071226663326264e-05, "loss": 0.0602, "step": 26000 }, { "grad_norm": 0.6916990280151367, "learning_rate": 8.069595428438121e-05, "loss": 0.058, "step": 26010 }, { "grad_norm": 0.7496178150177002, "learning_rate": 8.067963669040072e-05, "loss": 0.0723, "step": 26020 }, { "grad_norm": 0.7130172252655029, "learning_rate": 8.066331385410942e-05, "loss": 0.0645, "step": 26030 }, { "grad_norm": 0.6493472456932068, "learning_rate": 8.064698577829641e-05, "loss": 0.0724, "step": 26040 }, { "grad_norm": 0.5044482350349426, "learning_rate": 8.063065246575175e-05, "loss": 0.0698, "step": 26050 }, { "grad_norm": 0.5563294291496277, "learning_rate": 8.061431391926631e-05, "loss": 0.0543, "step": 26060 }, { "grad_norm": 0.6421453952789307, "learning_rate": 8.059797014163195e-05, "loss": 0.0638, "step": 26070 }, { "grad_norm": 0.49996599555015564, "learning_rate": 8.058162113564133e-05, "loss": 0.0451, "step": 26080 }, { "grad_norm": 0.5408265590667725, "learning_rate": 8.056526690408806e-05, "loss": 0.0459, "step": 26090 }, { "grad_norm": 0.4688023328781128, "learning_rate": 8.054890744976666e-05, "loss": 0.0473, "step": 26100 }, { "grad_norm": 0.5121863484382629, "learning_rate": 8.053254277547248e-05, "loss": 0.0505, "step": 26110 }, { "grad_norm": 0.5921395421028137, "learning_rate": 8.051617288400182e-05, "loss": 0.0571, "step": 26120 }, { "grad_norm": 0.5164998173713684, "learning_rate": 8.049979777815182e-05, "loss": 0.0523, "step": 26130 }, { "grad_norm": 0.6244220733642578, "learning_rate": 8.048341746072054e-05, "loss": 0.0744, "step": 26140 }, { "grad_norm": 0.486751914024353, "learning_rate": 8.046703193450696e-05, "loss": 0.0568, "step": 26150 }, { "grad_norm": 0.8173201680183411, "learning_rate": 8.04506412023109e-05, "loss": 0.0626, "step": 26160 }, { "grad_norm": 0.4858076274394989, "learning_rate": 8.043424526693306e-05, "loss": 0.0426, "step": 26170 }, { "grad_norm": 0.4927389621734619, "learning_rate": 8.04178441311751e-05, "loss": 0.0644, "step": 26180 }, { "grad_norm": 0.42447003722190857, "learning_rate": 8.04014377978395e-05, "loss": 0.0554, "step": 26190 }, { "grad_norm": 0.5180978178977966, "learning_rate": 8.038502626972967e-05, "loss": 0.0566, "step": 26200 }, { "grad_norm": 1.1698781251907349, "learning_rate": 8.036860954964989e-05, "loss": 0.0511, "step": 26210 }, { "grad_norm": 0.6209810972213745, "learning_rate": 8.035218764040531e-05, "loss": 0.0508, "step": 26220 }, { "grad_norm": 0.46394097805023193, "learning_rate": 8.033576054480199e-05, "loss": 0.0485, "step": 26230 }, { "grad_norm": 0.6911510229110718, "learning_rate": 8.031932826564688e-05, "loss": 0.0507, "step": 26240 }, { "grad_norm": 0.5354512929916382, "learning_rate": 8.030289080574782e-05, "loss": 0.073, "step": 26250 }, { "grad_norm": 0.5517635941505432, "learning_rate": 8.028644816791349e-05, "loss": 0.0598, "step": 26260 }, { "grad_norm": 0.975607693195343, "learning_rate": 8.027000035495351e-05, "loss": 0.08, "step": 26270 }, { "grad_norm": 0.6135662198066711, "learning_rate": 8.025354736967836e-05, "loss": 0.0564, "step": 26280 }, { "grad_norm": 0.5381144881248474, "learning_rate": 8.023708921489941e-05, "loss": 0.0567, "step": 26290 }, { "grad_norm": 0.8082989454269409, "learning_rate": 8.022062589342887e-05, "loss": 0.0583, "step": 26300 }, { "grad_norm": 0.4968417286872864, "learning_rate": 8.020415740807993e-05, "loss": 0.0439, "step": 26310 }, { "grad_norm": 0.4220190644264221, "learning_rate": 8.018768376166656e-05, "loss": 0.0499, "step": 26320 }, { "grad_norm": 0.5346596240997314, "learning_rate": 8.017120495700368e-05, "loss": 0.059, "step": 26330 }, { "grad_norm": 0.5149128437042236, "learning_rate": 8.015472099690704e-05, "loss": 0.0499, "step": 26340 }, { "grad_norm": 0.38555577397346497, "learning_rate": 8.013823188419332e-05, "loss": 0.0679, "step": 26350 }, { "grad_norm": 0.38350653648376465, "learning_rate": 8.012173762168006e-05, "loss": 0.0568, "step": 26360 }, { "grad_norm": 0.4608142673969269, "learning_rate": 8.010523821218567e-05, "loss": 0.0454, "step": 26370 }, { "grad_norm": 0.6579099297523499, "learning_rate": 8.008873365852945e-05, "loss": 0.0414, "step": 26380 }, { "grad_norm": 0.450034499168396, "learning_rate": 8.007222396353157e-05, "loss": 0.04, "step": 26390 }, { "grad_norm": 0.7500739097595215, "learning_rate": 8.00557091300131e-05, "loss": 0.0632, "step": 26400 }, { "grad_norm": 0.6741234660148621, "learning_rate": 8.003918916079597e-05, "loss": 0.0481, "step": 26410 }, { "grad_norm": 0.5242570042610168, "learning_rate": 8.002266405870298e-05, "loss": 0.0445, "step": 26420 }, { "grad_norm": 0.83516526222229, "learning_rate": 8.000613382655782e-05, "loss": 0.0523, "step": 26430 }, { "grad_norm": 0.6697637438774109, "learning_rate": 7.998959846718505e-05, "loss": 0.046, "step": 26440 }, { "grad_norm": 0.7715635895729065, "learning_rate": 7.997305798341012e-05, "loss": 0.071, "step": 26450 }, { "grad_norm": 0.5768121480941772, "learning_rate": 7.995651237805937e-05, "loss": 0.0457, "step": 26460 }, { "grad_norm": 0.8722366690635681, "learning_rate": 7.993996165395996e-05, "loss": 0.0726, "step": 26470 }, { "grad_norm": 0.8389999270439148, "learning_rate": 7.992340581393996e-05, "loss": 0.0447, "step": 26480 }, { "grad_norm": 0.6837485432624817, "learning_rate": 7.990684486082831e-05, "loss": 0.0493, "step": 26490 }, { "grad_norm": 0.5285077095031738, "learning_rate": 7.989027879745482e-05, "loss": 0.0513, "step": 26500 }, { "grad_norm": 0.6387868523597717, "learning_rate": 7.98737076266502e-05, "loss": 0.0452, "step": 26510 }, { "grad_norm": 0.7334091663360596, "learning_rate": 7.985713135124598e-05, "loss": 0.0626, "step": 26520 }, { "grad_norm": 0.603907585144043, "learning_rate": 7.98405499740746e-05, "loss": 0.0494, "step": 26530 }, { "grad_norm": 0.6050036549568176, "learning_rate": 7.98239634979694e-05, "loss": 0.0412, "step": 26540 }, { "grad_norm": 0.5763862133026123, "learning_rate": 7.98073719257645e-05, "loss": 0.0493, "step": 26550 }, { "grad_norm": 0.6842754483222961, "learning_rate": 7.979077526029499e-05, "loss": 0.0462, "step": 26560 }, { "grad_norm": 0.4982607960700989, "learning_rate": 7.977417350439675e-05, "loss": 0.0415, "step": 26570 }, { "grad_norm": 0.545243501663208, "learning_rate": 7.97575666609066e-05, "loss": 0.0497, "step": 26580 }, { "grad_norm": 0.33405911922454834, "learning_rate": 7.974095473266216e-05, "loss": 0.0424, "step": 26590 }, { "grad_norm": 0.43756282329559326, "learning_rate": 7.972433772250198e-05, "loss": 0.048, "step": 26600 }, { "grad_norm": 0.4951668381690979, "learning_rate": 7.970771563326544e-05, "loss": 0.0404, "step": 26610 }, { "grad_norm": 0.6982136368751526, "learning_rate": 7.96910884677928e-05, "loss": 0.0654, "step": 26620 }, { "grad_norm": 0.7548884153366089, "learning_rate": 7.967445622892523e-05, "loss": 0.066, "step": 26630 }, { "grad_norm": 0.6213887929916382, "learning_rate": 7.965781891950465e-05, "loss": 0.0668, "step": 26640 }, { "grad_norm": 0.5271270871162415, "learning_rate": 7.964117654237397e-05, "loss": 0.0604, "step": 26650 }, { "grad_norm": 0.6246338486671448, "learning_rate": 7.962452910037692e-05, "loss": 0.0552, "step": 26660 }, { "grad_norm": 0.423369437456131, "learning_rate": 7.96078765963581e-05, "loss": 0.0674, "step": 26670 }, { "grad_norm": 0.5835240483283997, "learning_rate": 7.95912190331629e-05, "loss": 0.0535, "step": 26680 }, { "grad_norm": 0.4674685597419739, "learning_rate": 7.957455641363772e-05, "loss": 0.0488, "step": 26690 }, { "grad_norm": 0.6252361536026001, "learning_rate": 7.955788874062968e-05, "loss": 0.0516, "step": 26700 }, { "grad_norm": 0.5147507786750793, "learning_rate": 7.95412160169869e-05, "loss": 0.0584, "step": 26710 }, { "grad_norm": 0.6773943901062012, "learning_rate": 7.952453824555824e-05, "loss": 0.0563, "step": 26720 }, { "grad_norm": 0.5749679207801819, "learning_rate": 7.95078554291935e-05, "loss": 0.0576, "step": 26730 }, { "grad_norm": 1.0687108039855957, "learning_rate": 7.94911675707433e-05, "loss": 0.0736, "step": 26740 }, { "grad_norm": 0.5592588782310486, "learning_rate": 7.947447467305915e-05, "loss": 0.0447, "step": 26750 }, { "grad_norm": 0.43148475885391235, "learning_rate": 7.94577767389934e-05, "loss": 0.0724, "step": 26760 }, { "grad_norm": 0.9956825971603394, "learning_rate": 7.944107377139928e-05, "loss": 0.0572, "step": 26770 }, { "grad_norm": 0.6334615349769592, "learning_rate": 7.942436577313088e-05, "loss": 0.0592, "step": 26780 }, { "grad_norm": 0.7017940282821655, "learning_rate": 7.940765274704312e-05, "loss": 0.06, "step": 26790 }, { "grad_norm": 0.5159381031990051, "learning_rate": 7.939093469599181e-05, "loss": 0.0559, "step": 26800 }, { "grad_norm": 0.46693935990333557, "learning_rate": 7.93742116228336e-05, "loss": 0.0532, "step": 26810 }, { "grad_norm": 0.5049833655357361, "learning_rate": 7.935748353042602e-05, "loss": 0.0457, "step": 26820 }, { "grad_norm": 0.7017560005187988, "learning_rate": 7.934075042162744e-05, "loss": 0.06, "step": 26830 }, { "grad_norm": 0.48097312450408936, "learning_rate": 7.932401229929705e-05, "loss": 0.0477, "step": 26840 }, { "grad_norm": 0.49292898178100586, "learning_rate": 7.9307269166295e-05, "loss": 0.0518, "step": 26850 }, { "grad_norm": 0.5302881002426147, "learning_rate": 7.92905210254822e-05, "loss": 0.0451, "step": 26860 }, { "grad_norm": 0.47636011242866516, "learning_rate": 7.927376787972045e-05, "loss": 0.0525, "step": 26870 }, { "grad_norm": 0.9919465184211731, "learning_rate": 7.92570097318724e-05, "loss": 0.0525, "step": 26880 }, { "grad_norm": 0.5977374315261841, "learning_rate": 7.924024658480158e-05, "loss": 0.0563, "step": 26890 }, { "grad_norm": 0.7135889530181885, "learning_rate": 7.922347844137233e-05, "loss": 0.0706, "step": 26900 }, { "grad_norm": 0.45466217398643494, "learning_rate": 7.92067053044499e-05, "loss": 0.0787, "step": 26910 }, { "grad_norm": 0.34748026728630066, "learning_rate": 7.918992717690031e-05, "loss": 0.0453, "step": 26920 }, { "grad_norm": 0.7672750353813171, "learning_rate": 7.917314406159053e-05, "loss": 0.0729, "step": 26930 }, { "grad_norm": 0.5892944931983948, "learning_rate": 7.915635596138832e-05, "loss": 0.0646, "step": 26940 }, { "grad_norm": 0.6780880689620972, "learning_rate": 7.913956287916228e-05, "loss": 0.0491, "step": 26950 }, { "grad_norm": 0.32714423537254333, "learning_rate": 7.912276481778193e-05, "loss": 0.0637, "step": 26960 }, { "grad_norm": 0.829246997833252, "learning_rate": 7.910596178011759e-05, "loss": 0.0574, "step": 26970 }, { "grad_norm": 0.3987463414669037, "learning_rate": 7.908915376904043e-05, "loss": 0.0456, "step": 26980 }, { "grad_norm": 0.4069289565086365, "learning_rate": 7.907234078742247e-05, "loss": 0.032, "step": 26990 }, { "grad_norm": 0.49404066801071167, "learning_rate": 7.90555228381366e-05, "loss": 0.0391, "step": 27000 }, { "grad_norm": 1.064391851425171, "learning_rate": 7.903869992405656e-05, "loss": 0.0571, "step": 27010 }, { "grad_norm": 0.5061320662498474, "learning_rate": 7.902187204805691e-05, "loss": 0.0522, "step": 27020 }, { "grad_norm": 0.4349416196346283, "learning_rate": 7.900503921301308e-05, "loss": 0.0581, "step": 27030 }, { "grad_norm": 1.0051556825637817, "learning_rate": 7.898820142180133e-05, "loss": 0.0549, "step": 27040 }, { "grad_norm": 0.43344640731811523, "learning_rate": 7.897135867729879e-05, "loss": 0.0738, "step": 27050 }, { "grad_norm": 0.6694864630699158, "learning_rate": 7.89545109823834e-05, "loss": 0.0568, "step": 27060 }, { "grad_norm": 0.46002262830734253, "learning_rate": 7.8937658339934e-05, "loss": 0.0792, "step": 27070 }, { "grad_norm": 0.988351583480835, "learning_rate": 7.892080075283026e-05, "loss": 0.0516, "step": 27080 }, { "grad_norm": 0.49613815546035767, "learning_rate": 7.890393822395263e-05, "loss": 0.0553, "step": 27090 }, { "grad_norm": 0.7847883105278015, "learning_rate": 7.88870707561825e-05, "loss": 0.0524, "step": 27100 }, { "grad_norm": 0.7919647097587585, "learning_rate": 7.887019835240203e-05, "loss": 0.0639, "step": 27110 }, { "grad_norm": 0.6271344423294067, "learning_rate": 7.885332101549427e-05, "loss": 0.0718, "step": 27120 }, { "grad_norm": 0.42065200209617615, "learning_rate": 7.883643874834308e-05, "loss": 0.0547, "step": 27130 }, { "grad_norm": 0.5417182445526123, "learning_rate": 7.881955155383321e-05, "loss": 0.0462, "step": 27140 }, { "grad_norm": 0.6792735457420349, "learning_rate": 7.880265943485017e-05, "loss": 0.0477, "step": 27150 }, { "grad_norm": 0.840930163860321, "learning_rate": 7.878576239428038e-05, "loss": 0.0622, "step": 27160 }, { "grad_norm": 0.7110920548439026, "learning_rate": 7.87688604350111e-05, "loss": 0.0844, "step": 27170 }, { "grad_norm": 0.33324024081230164, "learning_rate": 7.875195355993042e-05, "loss": 0.0468, "step": 27180 }, { "grad_norm": 0.45402640104293823, "learning_rate": 7.873504177192724e-05, "loss": 0.0872, "step": 27190 }, { "grad_norm": 0.6086936593055725, "learning_rate": 7.87181250738913e-05, "loss": 0.0608, "step": 27200 }, { "grad_norm": 0.3899020850658417, "learning_rate": 7.870120346871324e-05, "loss": 0.0474, "step": 27210 }, { "grad_norm": 0.5275940299034119, "learning_rate": 7.86842769592845e-05, "loss": 0.0425, "step": 27220 }, { "grad_norm": 0.3959120512008667, "learning_rate": 7.866734554849732e-05, "loss": 0.05, "step": 27230 }, { "grad_norm": 0.546501874923706, "learning_rate": 7.865040923924486e-05, "loss": 0.0799, "step": 27240 }, { "grad_norm": 0.5127246975898743, "learning_rate": 7.863346803442104e-05, "loss": 0.0518, "step": 27250 }, { "grad_norm": 0.48298728466033936, "learning_rate": 7.861652193692067e-05, "loss": 0.0417, "step": 27260 }, { "grad_norm": 0.41944095492362976, "learning_rate": 7.859957094963937e-05, "loss": 0.0407, "step": 27270 }, { "grad_norm": 0.6937837600708008, "learning_rate": 7.858261507547357e-05, "loss": 0.056, "step": 27280 }, { "grad_norm": 0.6005281805992126, "learning_rate": 7.856565431732061e-05, "loss": 0.06, "step": 27290 }, { "grad_norm": 0.4638076424598694, "learning_rate": 7.854868867807859e-05, "loss": 0.0638, "step": 27300 }, { "grad_norm": 0.6504153609275818, "learning_rate": 7.85317181606465e-05, "loss": 0.0601, "step": 27310 }, { "grad_norm": 0.7365071773529053, "learning_rate": 7.85147427679241e-05, "loss": 0.0526, "step": 27320 }, { "grad_norm": 0.45433735847473145, "learning_rate": 7.849776250281205e-05, "loss": 0.0407, "step": 27330 }, { "grad_norm": 0.5206131935119629, "learning_rate": 7.84807773682118e-05, "loss": 0.0538, "step": 27340 }, { "grad_norm": 1.1223019361495972, "learning_rate": 7.846378736702565e-05, "loss": 0.0586, "step": 27350 }, { "grad_norm": 0.413453072309494, "learning_rate": 7.844679250215671e-05, "loss": 0.0455, "step": 27360 }, { "grad_norm": 0.6105486154556274, "learning_rate": 7.842979277650898e-05, "loss": 0.0518, "step": 27370 }, { "grad_norm": 0.6623896956443787, "learning_rate": 7.84127881929872e-05, "loss": 0.0494, "step": 27380 }, { "grad_norm": 0.6227537989616394, "learning_rate": 7.839577875449704e-05, "loss": 0.0589, "step": 27390 }, { "grad_norm": 0.7632622122764587, "learning_rate": 7.837876446394489e-05, "loss": 0.047, "step": 27400 }, { "grad_norm": 0.7302005887031555, "learning_rate": 7.836174532423805e-05, "loss": 0.0522, "step": 27410 }, { "grad_norm": 0.7307083010673523, "learning_rate": 7.834472133828466e-05, "loss": 0.0484, "step": 27420 }, { "grad_norm": 0.5627347230911255, "learning_rate": 7.832769250899359e-05, "loss": 0.0475, "step": 27430 }, { "grad_norm": 0.49986517429351807, "learning_rate": 7.831065883927464e-05, "loss": 0.0556, "step": 27440 }, { "grad_norm": 0.4887232184410095, "learning_rate": 7.829362033203841e-05, "loss": 0.0401, "step": 27450 }, { "grad_norm": 0.4488120675086975, "learning_rate": 7.827657699019628e-05, "loss": 0.043, "step": 27460 }, { "grad_norm": 0.5648229122161865, "learning_rate": 7.825952881666052e-05, "loss": 0.0366, "step": 27470 }, { "grad_norm": 0.6803785562515259, "learning_rate": 7.824247581434418e-05, "loss": 0.0427, "step": 27480 }, { "grad_norm": 0.6612744927406311, "learning_rate": 7.822541798616116e-05, "loss": 0.055, "step": 27490 }, { "grad_norm": 0.3247365355491638, "learning_rate": 7.820835533502617e-05, "loss": 0.0414, "step": 27500 }, { "grad_norm": 0.5292696952819824, "learning_rate": 7.819128786385475e-05, "loss": 0.0595, "step": 27510 }, { "grad_norm": 0.5154598951339722, "learning_rate": 7.817421557556329e-05, "loss": 0.0835, "step": 27520 }, { "grad_norm": 0.4314756393432617, "learning_rate": 7.815713847306893e-05, "loss": 0.0491, "step": 27530 }, { "grad_norm": 0.7598555088043213, "learning_rate": 7.81400565592897e-05, "loss": 0.0481, "step": 27540 }, { "grad_norm": 0.2969847023487091, "learning_rate": 7.812296983714444e-05, "loss": 0.045, "step": 27550 }, { "grad_norm": 0.41926392912864685, "learning_rate": 7.810587830955281e-05, "loss": 0.0465, "step": 27560 }, { "grad_norm": 0.5091694593429565, "learning_rate": 7.808878197943528e-05, "loss": 0.0533, "step": 27570 }, { "grad_norm": 1.0621916055679321, "learning_rate": 7.807168084971312e-05, "loss": 0.0514, "step": 27580 }, { "grad_norm": 0.455098420381546, "learning_rate": 7.805457492330849e-05, "loss": 0.0434, "step": 27590 }, { "grad_norm": 0.4613129198551178, "learning_rate": 7.803746420314428e-05, "loss": 0.0458, "step": 27600 }, { "grad_norm": 0.5103310346603394, "learning_rate": 7.802034869214428e-05, "loss": 0.067, "step": 27610 }, { "grad_norm": 0.5069372653961182, "learning_rate": 7.800322839323303e-05, "loss": 0.0492, "step": 27620 }, { "grad_norm": 0.5704649090766907, "learning_rate": 7.798610330933593e-05, "loss": 0.0548, "step": 27630 }, { "grad_norm": 0.6548474431037903, "learning_rate": 7.796897344337922e-05, "loss": 0.061, "step": 27640 }, { "grad_norm": 0.5093953609466553, "learning_rate": 7.795183879828989e-05, "loss": 0.054, "step": 27650 }, { "grad_norm": 0.7256788611412048, "learning_rate": 7.793469937699579e-05, "loss": 0.0535, "step": 27660 }, { "grad_norm": 0.4419597387313843, "learning_rate": 7.791755518242558e-05, "loss": 0.0635, "step": 27670 }, { "grad_norm": 0.501137912273407, "learning_rate": 7.790040621750876e-05, "loss": 0.0534, "step": 27680 }, { "grad_norm": 0.3429276645183563, "learning_rate": 7.788325248517558e-05, "loss": 0.0355, "step": 27690 }, { "grad_norm": 0.675089418888092, "learning_rate": 7.786609398835715e-05, "loss": 0.0478, "step": 27700 }, { "grad_norm": 0.5577085614204407, "learning_rate": 7.784893072998541e-05, "loss": 0.0548, "step": 27710 }, { "grad_norm": 0.5348531007766724, "learning_rate": 7.783176271299306e-05, "loss": 0.0517, "step": 27720 }, { "grad_norm": 0.7074664831161499, "learning_rate": 7.781458994031368e-05, "loss": 0.0619, "step": 27730 }, { "grad_norm": 0.6499717235565186, "learning_rate": 7.779741241488161e-05, "loss": 0.0532, "step": 27740 }, { "grad_norm": 1.553612470626831, "learning_rate": 7.7780230139632e-05, "loss": 0.0629, "step": 27750 }, { "grad_norm": 1.133916974067688, "learning_rate": 7.776304311750087e-05, "loss": 0.0453, "step": 27760 }, { "grad_norm": 0.550697922706604, "learning_rate": 7.7745851351425e-05, "loss": 0.0528, "step": 27770 }, { "grad_norm": 0.8931840062141418, "learning_rate": 7.772865484434197e-05, "loss": 0.0642, "step": 27780 }, { "grad_norm": 0.6353334784507751, "learning_rate": 7.77114535991902e-05, "loss": 0.0559, "step": 27790 }, { "grad_norm": 0.28143373131752014, "learning_rate": 7.769424761890893e-05, "loss": 0.0398, "step": 27800 }, { "grad_norm": 0.9903985261917114, "learning_rate": 7.767703690643817e-05, "loss": 0.0548, "step": 27810 }, { "grad_norm": 0.41309744119644165, "learning_rate": 7.76598214647188e-05, "loss": 0.0421, "step": 27820 }, { "grad_norm": 0.4499862492084503, "learning_rate": 7.764260129669241e-05, "loss": 0.0626, "step": 27830 }, { "grad_norm": 0.5691275596618652, "learning_rate": 7.76253764053015e-05, "loss": 0.0508, "step": 27840 }, { "grad_norm": 0.7142775058746338, "learning_rate": 7.760814679348932e-05, "loss": 0.0538, "step": 27850 }, { "grad_norm": 0.5904514789581299, "learning_rate": 7.759091246419992e-05, "loss": 0.0606, "step": 27860 }, { "grad_norm": 0.621126115322113, "learning_rate": 7.75736734203782e-05, "loss": 0.0573, "step": 27870 }, { "grad_norm": 0.5139341950416565, "learning_rate": 7.755642966496985e-05, "loss": 0.0531, "step": 27880 }, { "grad_norm": 0.36548036336898804, "learning_rate": 7.753918120092132e-05, "loss": 0.0441, "step": 27890 }, { "grad_norm": 0.6116846799850464, "learning_rate": 7.752192803117993e-05, "loss": 0.0501, "step": 27900 }, { "grad_norm": 0.618453860282898, "learning_rate": 7.750467015869377e-05, "loss": 0.0518, "step": 27910 }, { "grad_norm": 0.5944697260856628, "learning_rate": 7.748740758641174e-05, "loss": 0.0538, "step": 27920 }, { "grad_norm": 0.43846815824508667, "learning_rate": 7.74701403172835e-05, "loss": 0.0612, "step": 27930 }, { "grad_norm": 0.5114140510559082, "learning_rate": 7.745286835425962e-05, "loss": 0.0797, "step": 27940 }, { "grad_norm": 0.3380921483039856, "learning_rate": 7.743559170029138e-05, "loss": 0.044, "step": 27950 }, { "grad_norm": 0.684750497341156, "learning_rate": 7.741831035833087e-05, "loss": 0.067, "step": 27960 }, { "grad_norm": 0.6006554961204529, "learning_rate": 7.740102433133102e-05, "loss": 0.0582, "step": 27970 }, { "grad_norm": 0.5456628203392029, "learning_rate": 7.738373362224553e-05, "loss": 0.0424, "step": 27980 }, { "grad_norm": 0.614881157875061, "learning_rate": 7.73664382340289e-05, "loss": 0.0603, "step": 27990 }, { "grad_norm": 0.7252930998802185, "learning_rate": 7.734913816963647e-05, "loss": 0.0511, "step": 28000 }, { "grad_norm": 0.6989389657974243, "learning_rate": 7.73318334320243e-05, "loss": 0.0549, "step": 28010 }, { "grad_norm": 0.518253743648529, "learning_rate": 7.731452402414934e-05, "loss": 0.0506, "step": 28020 }, { "grad_norm": 0.8826067447662354, "learning_rate": 7.729720994896928e-05, "loss": 0.0709, "step": 28030 }, { "grad_norm": 0.4861564040184021, "learning_rate": 7.727989120944262e-05, "loss": 0.0659, "step": 28040 }, { "grad_norm": 0.5560985207557678, "learning_rate": 7.726256780852865e-05, "loss": 0.0501, "step": 28050 }, { "grad_norm": 0.48230457305908203, "learning_rate": 7.724523974918749e-05, "loss": 0.0523, "step": 28060 }, { "grad_norm": 0.5012898445129395, "learning_rate": 7.722790703438002e-05, "loss": 0.0424, "step": 28070 }, { "grad_norm": 0.62242591381073, "learning_rate": 7.72105696670679e-05, "loss": 0.0467, "step": 28080 }, { "grad_norm": 0.6107966899871826, "learning_rate": 7.719322765021364e-05, "loss": 0.0483, "step": 28090 }, { "grad_norm": 0.4350246787071228, "learning_rate": 7.717588098678051e-05, "loss": 0.0452, "step": 28100 }, { "grad_norm": 0.6823537349700928, "learning_rate": 7.715852967973258e-05, "loss": 0.0564, "step": 28110 }, { "grad_norm": 0.5277339816093445, "learning_rate": 7.714117373203474e-05, "loss": 0.0467, "step": 28120 }, { "grad_norm": 0.4742606282234192, "learning_rate": 7.712381314665259e-05, "loss": 0.0528, "step": 28130 }, { "grad_norm": 0.7653809189796448, "learning_rate": 7.710644792655261e-05, "loss": 0.0538, "step": 28140 }, { "grad_norm": 0.4289858937263489, "learning_rate": 7.708907807470207e-05, "loss": 0.0475, "step": 28150 }, { "grad_norm": 0.40293657779693604, "learning_rate": 7.707170359406896e-05, "loss": 0.0453, "step": 28160 }, { "grad_norm": 0.46442192792892456, "learning_rate": 7.705432448762213e-05, "loss": 0.0582, "step": 28170 }, { "grad_norm": 0.5009303689002991, "learning_rate": 7.703694075833117e-05, "loss": 0.0372, "step": 28180 }, { "grad_norm": 0.7729337215423584, "learning_rate": 7.70195524091665e-05, "loss": 0.0523, "step": 28190 }, { "grad_norm": 0.43972423672676086, "learning_rate": 7.70021594430993e-05, "loss": 0.0483, "step": 28200 }, { "grad_norm": 0.3620920777320862, "learning_rate": 7.698476186310157e-05, "loss": 0.0418, "step": 28210 }, { "grad_norm": 0.44991961121559143, "learning_rate": 7.696735967214608e-05, "loss": 0.0643, "step": 28220 }, { "grad_norm": 0.4404211640357971, "learning_rate": 7.694995287320636e-05, "loss": 0.0442, "step": 28230 }, { "grad_norm": 0.5283715128898621, "learning_rate": 7.693254146925679e-05, "loss": 0.0491, "step": 28240 }, { "grad_norm": 0.4086558520793915, "learning_rate": 7.691512546327251e-05, "loss": 0.0373, "step": 28250 }, { "grad_norm": 0.5969226956367493, "learning_rate": 7.689770485822939e-05, "loss": 0.063, "step": 28260 }, { "grad_norm": 0.5722069144248962, "learning_rate": 7.688027965710416e-05, "loss": 0.059, "step": 28270 }, { "grad_norm": 0.6610327959060669, "learning_rate": 7.686284986287433e-05, "loss": 0.0527, "step": 28280 }, { "grad_norm": 0.36380335688591003, "learning_rate": 7.684541547851817e-05, "loss": 0.0397, "step": 28290 }, { "grad_norm": 0.40925660729408264, "learning_rate": 7.68279765070147e-05, "loss": 0.0556, "step": 28300 }, { "grad_norm": 0.917007565498352, "learning_rate": 7.68105329513438e-05, "loss": 0.0786, "step": 28310 }, { "grad_norm": 0.8017596006393433, "learning_rate": 7.67930848144861e-05, "loss": 0.0615, "step": 28320 }, { "grad_norm": 0.41045135259628296, "learning_rate": 7.6775632099423e-05, "loss": 0.0479, "step": 28330 }, { "grad_norm": 0.6916531920433044, "learning_rate": 7.675817480913667e-05, "loss": 0.06, "step": 28340 }, { "grad_norm": 0.6426004767417908, "learning_rate": 7.674071294661011e-05, "loss": 0.0465, "step": 28350 }, { "grad_norm": 0.6944481730461121, "learning_rate": 7.672324651482707e-05, "loss": 0.0599, "step": 28360 }, { "grad_norm": 0.6197952032089233, "learning_rate": 7.670577551677209e-05, "loss": 0.0658, "step": 28370 }, { "grad_norm": 0.8383909463882446, "learning_rate": 7.668829995543047e-05, "loss": 0.0632, "step": 28380 }, { "grad_norm": 0.9374382495880127, "learning_rate": 7.667081983378832e-05, "loss": 0.0596, "step": 28390 }, { "grad_norm": 0.7027829885482788, "learning_rate": 7.66533351548325e-05, "loss": 0.0563, "step": 28400 }, { "grad_norm": 0.5871483087539673, "learning_rate": 7.663584592155069e-05, "loss": 0.048, "step": 28410 }, { "grad_norm": 0.4952508211135864, "learning_rate": 7.661835213693129e-05, "loss": 0.058, "step": 28420 }, { "grad_norm": 0.550775408744812, "learning_rate": 7.660085380396353e-05, "loss": 0.0761, "step": 28430 }, { "grad_norm": 0.5748220086097717, "learning_rate": 7.658335092563738e-05, "loss": 0.0572, "step": 28440 }, { "grad_norm": 0.4886002540588379, "learning_rate": 7.656584350494362e-05, "loss": 0.0518, "step": 28450 }, { "grad_norm": 0.4463467001914978, "learning_rate": 7.654833154487378e-05, "loss": 0.0513, "step": 28460 }, { "grad_norm": 0.6972456574440002, "learning_rate": 7.653081504842017e-05, "loss": 0.0486, "step": 28470 }, { "grad_norm": 0.6276549696922302, "learning_rate": 7.65132940185759e-05, "loss": 0.049, "step": 28480 }, { "grad_norm": 0.5085862874984741, "learning_rate": 7.649576845833481e-05, "loss": 0.0627, "step": 28490 }, { "grad_norm": 0.5391444563865662, "learning_rate": 7.647823837069156e-05, "loss": 0.0556, "step": 28500 }, { "grad_norm": 0.5768287181854248, "learning_rate": 7.646070375864156e-05, "loss": 0.0644, "step": 28510 }, { "grad_norm": 0.4457032084465027, "learning_rate": 7.644316462518099e-05, "loss": 0.0546, "step": 28520 }, { "grad_norm": 0.44907453656196594, "learning_rate": 7.642562097330679e-05, "loss": 0.0599, "step": 28530 }, { "grad_norm": 0.621300220489502, "learning_rate": 7.640807280601671e-05, "loss": 0.0575, "step": 28540 }, { "grad_norm": 0.5470637679100037, "learning_rate": 7.639052012630927e-05, "loss": 0.0444, "step": 28550 }, { "grad_norm": 0.8119970560073853, "learning_rate": 7.63729629371837e-05, "loss": 0.0515, "step": 28560 }, { "grad_norm": 0.5441293716430664, "learning_rate": 7.635540124164009e-05, "loss": 0.0666, "step": 28570 }, { "grad_norm": 0.9075247049331665, "learning_rate": 7.633783504267922e-05, "loss": 0.0439, "step": 28580 }, { "grad_norm": 0.5640497207641602, "learning_rate": 7.632026434330269e-05, "loss": 0.0507, "step": 28590 }, { "grad_norm": 0.6613451838493347, "learning_rate": 7.630268914651282e-05, "loss": 0.0548, "step": 28600 }, { "grad_norm": 0.39858049154281616, "learning_rate": 7.628510945531278e-05, "loss": 0.0427, "step": 28610 }, { "grad_norm": 0.9941569566726685, "learning_rate": 7.626752527270641e-05, "loss": 0.0519, "step": 28620 }, { "grad_norm": 0.37268999218940735, "learning_rate": 7.62499366016984e-05, "loss": 0.0704, "step": 28630 }, { "grad_norm": 0.4467085301876068, "learning_rate": 7.623234344529416e-05, "loss": 0.0384, "step": 28640 }, { "grad_norm": 0.47145456075668335, "learning_rate": 7.62147458064999e-05, "loss": 0.0541, "step": 28650 }, { "grad_norm": 0.5894222855567932, "learning_rate": 7.619714368832254e-05, "loss": 0.0397, "step": 28660 }, { "grad_norm": 0.3709448575973511, "learning_rate": 7.61795370937698e-05, "loss": 0.0425, "step": 28670 }, { "grad_norm": 0.45807158946990967, "learning_rate": 7.61619260258502e-05, "loss": 0.0553, "step": 28680 }, { "grad_norm": 0.4886578619480133, "learning_rate": 7.614431048757298e-05, "loss": 0.0445, "step": 28690 }, { "grad_norm": 0.6714595556259155, "learning_rate": 7.612669048194814e-05, "loss": 0.0581, "step": 28700 }, { "grad_norm": 0.6732911467552185, "learning_rate": 7.610906601198646e-05, "loss": 0.042, "step": 28710 }, { "grad_norm": 0.9698162078857422, "learning_rate": 7.60914370806995e-05, "loss": 0.0466, "step": 28720 }, { "grad_norm": 0.4171202778816223, "learning_rate": 7.607380369109953e-05, "loss": 0.0508, "step": 28730 }, { "grad_norm": 0.5246805548667908, "learning_rate": 7.605616584619961e-05, "loss": 0.0533, "step": 28740 }, { "grad_norm": 0.5258064866065979, "learning_rate": 7.603852354901362e-05, "loss": 0.0469, "step": 28750 }, { "grad_norm": 0.5535130500793457, "learning_rate": 7.602087680255609e-05, "loss": 0.0484, "step": 28760 }, { "grad_norm": 0.6090837121009827, "learning_rate": 7.600322560984238e-05, "loss": 0.0542, "step": 28770 }, { "grad_norm": 0.6525554060935974, "learning_rate": 7.598556997388863e-05, "loss": 0.0658, "step": 28780 }, { "grad_norm": 0.43566495180130005, "learning_rate": 7.596790989771166e-05, "loss": 0.0444, "step": 28790 }, { "grad_norm": 0.45797282457351685, "learning_rate": 7.595024538432914e-05, "loss": 0.0399, "step": 28800 }, { "grad_norm": 0.3387131989002228, "learning_rate": 7.59325764367594e-05, "loss": 0.0491, "step": 28810 }, { "grad_norm": 0.5184364914894104, "learning_rate": 7.59149030580216e-05, "loss": 0.0552, "step": 28820 }, { "grad_norm": 0.5667352080345154, "learning_rate": 7.589722525113562e-05, "loss": 0.0479, "step": 28830 }, { "grad_norm": 0.5596851706504822, "learning_rate": 7.587954301912216e-05, "loss": 0.045, "step": 28840 }, { "grad_norm": 0.37759271264076233, "learning_rate": 7.586185636500263e-05, "loss": 0.0461, "step": 28850 }, { "grad_norm": 0.6150595545768738, "learning_rate": 7.584416529179914e-05, "loss": 0.0566, "step": 28860 }, { "grad_norm": 0.42267388105392456, "learning_rate": 7.582646980253465e-05, "loss": 0.0553, "step": 28870 }, { "grad_norm": 0.49458080530166626, "learning_rate": 7.580876990023282e-05, "loss": 0.0418, "step": 28880 }, { "grad_norm": 0.4788137376308441, "learning_rate": 7.579106558791809e-05, "loss": 0.059, "step": 28890 }, { "grad_norm": 0.7930386066436768, "learning_rate": 7.577335686861565e-05, "loss": 0.0476, "step": 28900 }, { "grad_norm": 0.5344968438148499, "learning_rate": 7.575564374535141e-05, "loss": 0.0455, "step": 28910 }, { "grad_norm": 0.8333723545074463, "learning_rate": 7.573792622115207e-05, "loss": 0.0417, "step": 28920 }, { "grad_norm": 0.5613278746604919, "learning_rate": 7.572020429904507e-05, "loss": 0.0455, "step": 28930 }, { "grad_norm": 0.5926421284675598, "learning_rate": 7.570247798205861e-05, "loss": 0.0723, "step": 28940 }, { "grad_norm": 1.2642624378204346, "learning_rate": 7.568474727322164e-05, "loss": 0.0535, "step": 28950 }, { "grad_norm": 0.5009538531303406, "learning_rate": 7.566701217556384e-05, "loss": 0.0538, "step": 28960 }, { "grad_norm": 0.4547271430492401, "learning_rate": 7.564927269211564e-05, "loss": 0.0528, "step": 28970 }, { "grad_norm": 0.539827823638916, "learning_rate": 7.563152882590824e-05, "loss": 0.0677, "step": 28980 }, { "grad_norm": 0.6827113628387451, "learning_rate": 7.56137805799736e-05, "loss": 0.0646, "step": 28990 }, { "grad_norm": 0.46956732869148254, "learning_rate": 7.559602795734439e-05, "loss": 0.054, "step": 29000 }, { "grad_norm": 0.4654534161090851, "learning_rate": 7.557827096105408e-05, "loss": 0.0513, "step": 29010 }, { "grad_norm": 0.4020843505859375, "learning_rate": 7.55605095941368e-05, "loss": 0.0431, "step": 29020 }, { "grad_norm": 0.5134803056716919, "learning_rate": 7.554274385962752e-05, "loss": 0.0487, "step": 29030 }, { "grad_norm": 0.4028429090976715, "learning_rate": 7.552497376056191e-05, "loss": 0.0435, "step": 29040 }, { "grad_norm": 0.6144540905952454, "learning_rate": 7.550719929997639e-05, "loss": 0.082, "step": 29050 }, { "grad_norm": 0.5812252759933472, "learning_rate": 7.548942048090813e-05, "loss": 0.0445, "step": 29060 }, { "grad_norm": 0.4412754774093628, "learning_rate": 7.547163730639506e-05, "loss": 0.0335, "step": 29070 }, { "grad_norm": 0.7579032778739929, "learning_rate": 7.545384977947583e-05, "loss": 0.0539, "step": 29080 }, { "grad_norm": 0.6139859557151794, "learning_rate": 7.543605790318981e-05, "loss": 0.0496, "step": 29090 }, { "grad_norm": 0.4834129214286804, "learning_rate": 7.54182616805772e-05, "loss": 0.048, "step": 29100 }, { "grad_norm": 0.49355700612068176, "learning_rate": 7.540046111467885e-05, "loss": 0.0486, "step": 29110 }, { "grad_norm": 0.508355975151062, "learning_rate": 7.53826562085364e-05, "loss": 0.0517, "step": 29120 }, { "grad_norm": 0.6478560566902161, "learning_rate": 7.536484696519221e-05, "loss": 0.0405, "step": 29130 }, { "grad_norm": 0.46720513701438904, "learning_rate": 7.534703338768942e-05, "loss": 0.066, "step": 29140 }, { "grad_norm": 0.5754474401473999, "learning_rate": 7.532921547907185e-05, "loss": 0.0518, "step": 29150 }, { "grad_norm": 0.7603769302368164, "learning_rate": 7.531139324238412e-05, "loss": 0.0586, "step": 29160 }, { "grad_norm": 0.40237414836883545, "learning_rate": 7.529356668067157e-05, "loss": 0.0624, "step": 29170 }, { "grad_norm": 1.0014598369598389, "learning_rate": 7.527573579698023e-05, "loss": 0.0685, "step": 29180 }, { "grad_norm": 0.5834318399429321, "learning_rate": 7.525790059435693e-05, "loss": 0.0488, "step": 29190 }, { "grad_norm": 0.34126728773117065, "learning_rate": 7.524006107584926e-05, "loss": 0.0614, "step": 29200 }, { "grad_norm": 0.5889216065406799, "learning_rate": 7.522221724450544e-05, "loss": 0.0333, "step": 29210 }, { "grad_norm": 0.6286783218383789, "learning_rate": 7.520436910337451e-05, "loss": 0.0483, "step": 29220 }, { "grad_norm": 0.46742764115333557, "learning_rate": 7.518651665550627e-05, "loss": 0.0543, "step": 29230 }, { "grad_norm": 0.5008029937744141, "learning_rate": 7.516865990395117e-05, "loss": 0.0431, "step": 29240 }, { "grad_norm": 0.29850298166275024, "learning_rate": 7.515079885176047e-05, "loss": 0.0276, "step": 29250 }, { "grad_norm": 0.8040362000465393, "learning_rate": 7.513293350198612e-05, "loss": 0.053, "step": 29260 }, { "grad_norm": 0.4811626672744751, "learning_rate": 7.511506385768081e-05, "loss": 0.0554, "step": 29270 }, { "grad_norm": 0.598279595375061, "learning_rate": 7.509718992189801e-05, "loss": 0.0393, "step": 29280 }, { "grad_norm": 0.5061681866645813, "learning_rate": 7.507931169769182e-05, "loss": 0.045, "step": 29290 }, { "grad_norm": 0.6976521015167236, "learning_rate": 7.506142918811722e-05, "loss": 0.0691, "step": 29300 }, { "grad_norm": 0.7630055546760559, "learning_rate": 7.504354239622978e-05, "loss": 0.0502, "step": 29310 }, { "grad_norm": 0.5295698046684265, "learning_rate": 7.50256513250859e-05, "loss": 0.0425, "step": 29320 }, { "grad_norm": 0.5401638746261597, "learning_rate": 7.500775597774265e-05, "loss": 0.0673, "step": 29330 }, { "grad_norm": 0.5173722505569458, "learning_rate": 7.498985635725788e-05, "loss": 0.0571, "step": 29340 }, { "grad_norm": 0.434440016746521, "learning_rate": 7.497195246669012e-05, "loss": 0.0378, "step": 29350 }, { "grad_norm": 0.5100142359733582, "learning_rate": 7.495404430909868e-05, "loss": 0.0488, "step": 29360 }, { "grad_norm": 0.43109747767448425, "learning_rate": 7.493613188754356e-05, "loss": 0.0532, "step": 29370 }, { "grad_norm": 0.8768615126609802, "learning_rate": 7.49182152050855e-05, "loss": 0.054, "step": 29380 }, { "grad_norm": 0.8194825053215027, "learning_rate": 7.490029426478598e-05, "loss": 0.0549, "step": 29390 }, { "grad_norm": 0.7689387202262878, "learning_rate": 7.488236906970719e-05, "loss": 0.0581, "step": 29400 }, { "grad_norm": 0.6812548041343689, "learning_rate": 7.486443962291207e-05, "loss": 0.0512, "step": 29410 }, { "grad_norm": 0.8873496651649475, "learning_rate": 7.484650592746424e-05, "loss": 0.0662, "step": 29420 }, { "grad_norm": 0.6481053233146667, "learning_rate": 7.482856798642811e-05, "loss": 0.0583, "step": 29430 }, { "grad_norm": 0.6275721788406372, "learning_rate": 7.481062580286878e-05, "loss": 0.056, "step": 29440 }, { "grad_norm": 0.5249293446540833, "learning_rate": 7.479267937985208e-05, "loss": 0.0526, "step": 29450 }, { "grad_norm": 0.3698330223560333, "learning_rate": 7.477472872044456e-05, "loss": 0.0606, "step": 29460 }, { "grad_norm": 0.4359968900680542, "learning_rate": 7.475677382771347e-05, "loss": 0.0579, "step": 29470 }, { "grad_norm": 0.44147029519081116, "learning_rate": 7.473881470472683e-05, "loss": 0.0401, "step": 29480 }, { "grad_norm": 0.7021834850311279, "learning_rate": 7.472085135455337e-05, "loss": 0.0461, "step": 29490 }, { "grad_norm": 0.5031159520149231, "learning_rate": 7.470288378026256e-05, "loss": 0.0643, "step": 29500 }, { "grad_norm": 0.4544684886932373, "learning_rate": 7.468491198492451e-05, "loss": 0.0505, "step": 29510 }, { "grad_norm": 0.6688838005065918, "learning_rate": 7.466693597161013e-05, "loss": 0.0508, "step": 29520 }, { "grad_norm": 0.6799508929252625, "learning_rate": 7.464895574339104e-05, "loss": 0.0419, "step": 29530 }, { "grad_norm": 0.5489168763160706, "learning_rate": 7.463097130333958e-05, "loss": 0.0539, "step": 29540 }, { "grad_norm": 0.3034588694572449, "learning_rate": 7.461298265452876e-05, "loss": 0.0467, "step": 29550 }, { "grad_norm": 0.7826828956604004, "learning_rate": 7.459498980003239e-05, "loss": 0.0421, "step": 29560 }, { "grad_norm": 0.449206680059433, "learning_rate": 7.457699274292493e-05, "loss": 0.0424, "step": 29570 }, { "grad_norm": 0.46932080388069153, "learning_rate": 7.455899148628159e-05, "loss": 0.0335, "step": 29580 }, { "grad_norm": 0.808917760848999, "learning_rate": 7.45409860331783e-05, "loss": 0.0543, "step": 29590 }, { "grad_norm": 0.8187258839607239, "learning_rate": 7.452297638669169e-05, "loss": 0.0614, "step": 29600 }, { "grad_norm": 0.6358350515365601, "learning_rate": 7.450496254989911e-05, "loss": 0.0474, "step": 29610 }, { "grad_norm": 0.4203015863895416, "learning_rate": 7.448694452587866e-05, "loss": 0.0541, "step": 29620 }, { "grad_norm": 0.48863276839256287, "learning_rate": 7.44689223177091e-05, "loss": 0.0674, "step": 29630 }, { "grad_norm": 0.5759286880493164, "learning_rate": 7.445089592846994e-05, "loss": 0.045, "step": 29640 }, { "grad_norm": 0.5333480834960938, "learning_rate": 7.443286536124141e-05, "loss": 0.0741, "step": 29650 }, { "grad_norm": 0.4368225336074829, "learning_rate": 7.441483061910443e-05, "loss": 0.0479, "step": 29660 }, { "grad_norm": 0.48699817061424255, "learning_rate": 7.439679170514064e-05, "loss": 0.0927, "step": 29670 }, { "grad_norm": 0.40153199434280396, "learning_rate": 7.43787486224324e-05, "loss": 0.0459, "step": 29680 }, { "grad_norm": 0.49731409549713135, "learning_rate": 7.436070137406276e-05, "loss": 0.0476, "step": 29690 }, { "grad_norm": 0.49559247493743896, "learning_rate": 7.434264996311556e-05, "loss": 0.0937, "step": 29700 }, { "grad_norm": 0.9458898305892944, "learning_rate": 7.432459439267525e-05, "loss": 0.0494, "step": 29710 }, { "grad_norm": 0.4798068106174469, "learning_rate": 7.430653466582701e-05, "loss": 0.0565, "step": 29720 }, { "grad_norm": 0.8720967173576355, "learning_rate": 7.42884707856568e-05, "loss": 0.0645, "step": 29730 }, { "grad_norm": 0.5179358124732971, "learning_rate": 7.427040275525122e-05, "loss": 0.0546, "step": 29740 }, { "grad_norm": 0.48572754859924316, "learning_rate": 7.42523305776976e-05, "loss": 0.0431, "step": 29750 }, { "grad_norm": 0.8022412061691284, "learning_rate": 7.4234254256084e-05, "loss": 0.0626, "step": 29760 }, { "grad_norm": 0.6944053173065186, "learning_rate": 7.421617379349915e-05, "loss": 0.0528, "step": 29770 }, { "grad_norm": 0.5609192252159119, "learning_rate": 7.41980891930325e-05, "loss": 0.0429, "step": 29780 }, { "grad_norm": 0.7620614767074585, "learning_rate": 7.418000045777425e-05, "loss": 0.052, "step": 29790 }, { "grad_norm": 0.6180248260498047, "learning_rate": 7.416190759081523e-05, "loss": 0.0444, "step": 29800 }, { "grad_norm": 0.6599485278129578, "learning_rate": 7.414381059524704e-05, "loss": 0.0442, "step": 29810 }, { "grad_norm": 0.7511146664619446, "learning_rate": 7.412570947416195e-05, "loss": 0.0529, "step": 29820 }, { "grad_norm": 0.4608929753303528, "learning_rate": 7.410760423065295e-05, "loss": 0.0397, "step": 29830 }, { "grad_norm": 0.4491628408432007, "learning_rate": 7.408949486781372e-05, "loss": 0.0452, "step": 29840 }, { "grad_norm": 0.353985458612442, "learning_rate": 7.407138138873868e-05, "loss": 0.0484, "step": 29850 }, { "grad_norm": 0.5129308700561523, "learning_rate": 7.405326379652292e-05, "loss": 0.0428, "step": 29860 }, { "grad_norm": 0.5684278011322021, "learning_rate": 7.403514209426222e-05, "loss": 0.0606, "step": 29870 }, { "grad_norm": 0.7607554793357849, "learning_rate": 7.40170162850531e-05, "loss": 0.0501, "step": 29880 }, { "grad_norm": 0.341021865606308, "learning_rate": 7.399888637199278e-05, "loss": 0.0318, "step": 29890 }, { "grad_norm": 0.6129428148269653, "learning_rate": 7.398075235817914e-05, "loss": 0.0782, "step": 29900 }, { "grad_norm": 0.4103606045246124, "learning_rate": 7.39626142467108e-05, "loss": 0.0403, "step": 29910 }, { "grad_norm": 0.7676213979721069, "learning_rate": 7.394447204068706e-05, "loss": 0.0645, "step": 29920 }, { "grad_norm": 0.4764541685581207, "learning_rate": 7.392632574320793e-05, "loss": 0.0349, "step": 29930 }, { "grad_norm": 0.7542891502380371, "learning_rate": 7.390817535737411e-05, "loss": 0.0597, "step": 29940 }, { "grad_norm": 0.5989865660667419, "learning_rate": 7.389002088628703e-05, "loss": 0.0578, "step": 29950 }, { "grad_norm": 0.5178682208061218, "learning_rate": 7.387186233304877e-05, "loss": 0.049, "step": 29960 }, { "grad_norm": 0.6574634313583374, "learning_rate": 7.385369970076212e-05, "loss": 0.0568, "step": 29970 }, { "grad_norm": 0.6095360517501831, "learning_rate": 7.38355329925306e-05, "loss": 0.0435, "step": 29980 }, { "grad_norm": 0.40922221541404724, "learning_rate": 7.381736221145838e-05, "loss": 0.0458, "step": 29990 }, { "grad_norm": 0.5613283514976501, "learning_rate": 7.37991873606504e-05, "loss": 0.0635, "step": 30000 }, { "grad_norm": 0.8491092920303345, "learning_rate": 7.378100844321218e-05, "loss": 0.0419, "step": 30010 }, { "grad_norm": 0.5961572527885437, "learning_rate": 7.376282546225004e-05, "loss": 0.0432, "step": 30020 }, { "grad_norm": 0.6548417210578918, "learning_rate": 7.374463842087094e-05, "loss": 0.0439, "step": 30030 }, { "grad_norm": 0.5028135776519775, "learning_rate": 7.372644732218254e-05, "loss": 0.0601, "step": 30040 }, { "grad_norm": 0.46019333600997925, "learning_rate": 7.370825216929322e-05, "loss": 0.0489, "step": 30050 }, { "grad_norm": 0.7294155955314636, "learning_rate": 7.369005296531205e-05, "loss": 0.0446, "step": 30060 }, { "grad_norm": 0.47723332047462463, "learning_rate": 7.367184971334873e-05, "loss": 0.0844, "step": 30070 }, { "grad_norm": 0.6768747568130493, "learning_rate": 7.365364241651371e-05, "loss": 0.0419, "step": 30080 }, { "grad_norm": 0.6392226815223694, "learning_rate": 7.363543107791815e-05, "loss": 0.047, "step": 30090 }, { "grad_norm": 0.6337289214134216, "learning_rate": 7.361721570067384e-05, "loss": 0.0703, "step": 30100 }, { "grad_norm": 0.46347835659980774, "learning_rate": 7.359899628789331e-05, "loss": 0.055, "step": 30110 }, { "grad_norm": 0.47631070017814636, "learning_rate": 7.358077284268974e-05, "loss": 0.0594, "step": 30120 }, { "grad_norm": 0.7133397459983826, "learning_rate": 7.356254536817702e-05, "loss": 0.0487, "step": 30130 }, { "grad_norm": 0.6287223100662231, "learning_rate": 7.354431386746973e-05, "loss": 0.0621, "step": 30140 }, { "grad_norm": 0.774346113204956, "learning_rate": 7.352607834368316e-05, "loss": 0.0476, "step": 30150 }, { "grad_norm": 0.4736230969429016, "learning_rate": 7.350783879993324e-05, "loss": 0.0487, "step": 30160 }, { "grad_norm": 0.667721152305603, "learning_rate": 7.348959523933658e-05, "loss": 0.0608, "step": 30170 }, { "grad_norm": 0.4671318829059601, "learning_rate": 7.347134766501057e-05, "loss": 0.0474, "step": 30180 }, { "grad_norm": 0.48762473464012146, "learning_rate": 7.345309608007315e-05, "loss": 0.0592, "step": 30190 }, { "grad_norm": 0.39188119769096375, "learning_rate": 7.343484048764308e-05, "loss": 0.0477, "step": 30200 }, { "grad_norm": 0.4662507474422455, "learning_rate": 7.341658089083972e-05, "loss": 0.0412, "step": 30210 }, { "grad_norm": 0.4879719913005829, "learning_rate": 7.339831729278313e-05, "loss": 0.0517, "step": 30220 }, { "grad_norm": 0.41564103960990906, "learning_rate": 7.338004969659404e-05, "loss": 0.0484, "step": 30230 }, { "grad_norm": 0.8351552486419678, "learning_rate": 7.336177810539391e-05, "loss": 0.0434, "step": 30240 }, { "grad_norm": 0.5267304182052612, "learning_rate": 7.334350252230485e-05, "loss": 0.0452, "step": 30250 }, { "grad_norm": 0.40121838450431824, "learning_rate": 7.332522295044965e-05, "loss": 0.0486, "step": 30260 }, { "grad_norm": 0.6123973727226257, "learning_rate": 7.33069393929518e-05, "loss": 0.0386, "step": 30270 }, { "grad_norm": 0.3634015619754791, "learning_rate": 7.328865185293545e-05, "loss": 0.0446, "step": 30280 }, { "grad_norm": 0.6156525611877441, "learning_rate": 7.327036033352546e-05, "loss": 0.0506, "step": 30290 }, { "grad_norm": 0.39773598313331604, "learning_rate": 7.325206483784733e-05, "loss": 0.0558, "step": 30300 }, { "grad_norm": 0.620606005191803, "learning_rate": 7.323376536902724e-05, "loss": 0.0645, "step": 30310 }, { "grad_norm": 0.634227454662323, "learning_rate": 7.321546193019213e-05, "loss": 0.0559, "step": 30320 }, { "grad_norm": 0.563107430934906, "learning_rate": 7.31971545244695e-05, "loss": 0.0387, "step": 30330 }, { "grad_norm": 0.5537030100822449, "learning_rate": 7.31788431549876e-05, "loss": 0.0688, "step": 30340 }, { "grad_norm": 0.562932014465332, "learning_rate": 7.316052782487534e-05, "loss": 0.0488, "step": 30350 }, { "grad_norm": 1.5962315797805786, "learning_rate": 7.314220853726234e-05, "loss": 0.0474, "step": 30360 }, { "grad_norm": 0.6709057688713074, "learning_rate": 7.312388529527884e-05, "loss": 0.0572, "step": 30370 }, { "grad_norm": 0.4266093075275421, "learning_rate": 7.310555810205577e-05, "loss": 0.0331, "step": 30380 }, { "grad_norm": 0.5966880917549133, "learning_rate": 7.308722696072476e-05, "loss": 0.0568, "step": 30390 }, { "grad_norm": 0.5820857286453247, "learning_rate": 7.306889187441811e-05, "loss": 0.0473, "step": 30400 }, { "grad_norm": 0.5367769002914429, "learning_rate": 7.305055284626876e-05, "loss": 0.044, "step": 30410 }, { "grad_norm": 0.31628894805908203, "learning_rate": 7.303220987941037e-05, "loss": 0.0349, "step": 30420 }, { "grad_norm": 0.4932696521282196, "learning_rate": 7.301386297697726e-05, "loss": 0.0338, "step": 30430 }, { "grad_norm": 0.7741850018501282, "learning_rate": 7.299551214210438e-05, "loss": 0.0521, "step": 30440 }, { "grad_norm": 0.6397730112075806, "learning_rate": 7.297715737792738e-05, "loss": 0.0398, "step": 30450 }, { "grad_norm": 0.5232501029968262, "learning_rate": 7.295879868758265e-05, "loss": 0.0402, "step": 30460 }, { "grad_norm": 0.39042800664901733, "learning_rate": 7.294043607420713e-05, "loss": 0.0392, "step": 30470 }, { "grad_norm": 0.5111897587776184, "learning_rate": 7.292206954093852e-05, "loss": 0.0478, "step": 30480 }, { "grad_norm": 0.5269081592559814, "learning_rate": 7.290369909091515e-05, "loss": 0.0491, "step": 30490 }, { "grad_norm": 0.48018237948417664, "learning_rate": 7.2885324727276e-05, "loss": 0.0721, "step": 30500 }, { "grad_norm": 0.7778648138046265, "learning_rate": 7.286694645316076e-05, "loss": 0.0686, "step": 30510 }, { "grad_norm": 0.7143461108207703, "learning_rate": 7.284856427170982e-05, "loss": 0.0565, "step": 30520 }, { "grad_norm": 0.554340660572052, "learning_rate": 7.283017818606414e-05, "loss": 0.0374, "step": 30530 }, { "grad_norm": 0.4654935896396637, "learning_rate": 7.28117881993654e-05, "loss": 0.0446, "step": 30540 }, { "grad_norm": 0.7052475810050964, "learning_rate": 7.279339431475598e-05, "loss": 0.0537, "step": 30550 }, { "grad_norm": 0.440878689289093, "learning_rate": 7.277499653537887e-05, "loss": 0.0325, "step": 30560 }, { "grad_norm": 0.6372168064117432, "learning_rate": 7.275659486437776e-05, "loss": 0.0606, "step": 30570 }, { "grad_norm": 0.6577532887458801, "learning_rate": 7.273818930489695e-05, "loss": 0.0568, "step": 30580 }, { "grad_norm": 0.7039077877998352, "learning_rate": 7.271977986008151e-05, "loss": 0.0482, "step": 30590 }, { "grad_norm": 0.5193545818328857, "learning_rate": 7.270136653307705e-05, "loss": 0.0493, "step": 30600 }, { "grad_norm": 0.5012040734291077, "learning_rate": 7.268294932702994e-05, "loss": 0.0725, "step": 30610 }, { "grad_norm": 0.49040669202804565, "learning_rate": 7.266452824508719e-05, "loss": 0.0343, "step": 30620 }, { "grad_norm": 0.5114156007766724, "learning_rate": 7.264610329039643e-05, "loss": 0.0549, "step": 30630 }, { "grad_norm": 0.47526252269744873, "learning_rate": 7.262767446610599e-05, "loss": 0.0398, "step": 30640 }, { "grad_norm": 0.4735732078552246, "learning_rate": 7.260924177536485e-05, "loss": 0.0374, "step": 30650 }, { "grad_norm": 0.6109406352043152, "learning_rate": 7.259080522132265e-05, "loss": 0.0425, "step": 30660 }, { "grad_norm": 0.46685972809791565, "learning_rate": 7.257236480712972e-05, "loss": 0.0418, "step": 30670 }, { "grad_norm": 0.495997816324234, "learning_rate": 7.255392053593697e-05, "loss": 0.0568, "step": 30680 }, { "grad_norm": 0.609550952911377, "learning_rate": 7.253547241089607e-05, "loss": 0.0528, "step": 30690 }, { "grad_norm": 0.4556388854980469, "learning_rate": 7.251702043515927e-05, "loss": 0.0594, "step": 30700 }, { "grad_norm": 0.6877837181091309, "learning_rate": 7.249856461187952e-05, "loss": 0.0496, "step": 30710 }, { "grad_norm": 0.6435277462005615, "learning_rate": 7.248010494421042e-05, "loss": 0.0459, "step": 30720 }, { "grad_norm": 0.5295370817184448, "learning_rate": 7.246164143530622e-05, "loss": 0.0626, "step": 30730 }, { "grad_norm": 0.6945603489875793, "learning_rate": 7.244317408832181e-05, "loss": 0.0399, "step": 30740 }, { "grad_norm": 0.6008121967315674, "learning_rate": 7.242470290641279e-05, "loss": 0.056, "step": 30750 }, { "grad_norm": 0.39217397570610046, "learning_rate": 7.240622789273536e-05, "loss": 0.0315, "step": 30760 }, { "grad_norm": 0.32659998536109924, "learning_rate": 7.238774905044638e-05, "loss": 0.0344, "step": 30770 }, { "grad_norm": 0.5448139905929565, "learning_rate": 7.236926638270341e-05, "loss": 0.0478, "step": 30780 }, { "grad_norm": 0.4994637668132782, "learning_rate": 7.23507798926646e-05, "loss": 0.0408, "step": 30790 }, { "grad_norm": 0.4493841826915741, "learning_rate": 7.23322895834888e-05, "loss": 0.0501, "step": 30800 }, { "grad_norm": 0.949108362197876, "learning_rate": 7.231379545833552e-05, "loss": 0.0504, "step": 30810 }, { "grad_norm": 1.162624478340149, "learning_rate": 7.229529752036487e-05, "loss": 0.0456, "step": 30820 }, { "grad_norm": 0.43669232726097107, "learning_rate": 7.227679577273765e-05, "loss": 0.0462, "step": 30830 }, { "grad_norm": 0.35868728160858154, "learning_rate": 7.225829021861529e-05, "loss": 0.0427, "step": 30840 }, { "grad_norm": 0.5031267404556274, "learning_rate": 7.223978086115992e-05, "loss": 0.0588, "step": 30850 }, { "grad_norm": 0.4564778506755829, "learning_rate": 7.222126770353425e-05, "loss": 0.0828, "step": 30860 }, { "grad_norm": 0.4266791343688965, "learning_rate": 7.22027507489017e-05, "loss": 0.0518, "step": 30870 }, { "grad_norm": 0.5162060260772705, "learning_rate": 7.218423000042627e-05, "loss": 0.0479, "step": 30880 }, { "grad_norm": 0.7068420648574829, "learning_rate": 7.216570546127268e-05, "loss": 0.0651, "step": 30890 }, { "grad_norm": 0.5200349688529968, "learning_rate": 7.214717713460626e-05, "loss": 0.0509, "step": 30900 }, { "grad_norm": 0.48846328258514404, "learning_rate": 7.2128645023593e-05, "loss": 0.0418, "step": 30910 }, { "grad_norm": 0.6039478182792664, "learning_rate": 7.211010913139951e-05, "loss": 0.0447, "step": 30920 }, { "grad_norm": 0.5335737466812134, "learning_rate": 7.209156946119308e-05, "loss": 0.0463, "step": 30930 }, { "grad_norm": 0.5466623902320862, "learning_rate": 7.207302601614166e-05, "loss": 0.0424, "step": 30940 }, { "grad_norm": 0.3572668135166168, "learning_rate": 7.205447879941378e-05, "loss": 0.0348, "step": 30950 }, { "grad_norm": 0.871961236000061, "learning_rate": 7.203592781417866e-05, "loss": 0.0683, "step": 30960 }, { "grad_norm": 0.38032516837120056, "learning_rate": 7.201737306360617e-05, "loss": 0.053, "step": 30970 }, { "grad_norm": 0.7316321730613708, "learning_rate": 7.19988145508668e-05, "loss": 0.0465, "step": 30980 }, { "grad_norm": 0.42939451336860657, "learning_rate": 7.198025227913168e-05, "loss": 0.0338, "step": 30990 }, { "grad_norm": 0.5993890166282654, "learning_rate": 7.196168625157261e-05, "loss": 0.0511, "step": 31000 }, { "grad_norm": 0.7117258310317993, "learning_rate": 7.194311647136201e-05, "loss": 0.0546, "step": 31010 }, { "grad_norm": 0.4381009340286255, "learning_rate": 7.192454294167297e-05, "loss": 0.0501, "step": 31020 }, { "grad_norm": 0.4730047583580017, "learning_rate": 7.190596566567917e-05, "loss": 0.039, "step": 31030 }, { "grad_norm": 0.45872727036476135, "learning_rate": 7.188738464655496e-05, "loss": 0.0361, "step": 31040 }, { "grad_norm": 0.48812687397003174, "learning_rate": 7.186879988747533e-05, "loss": 0.0476, "step": 31050 }, { "grad_norm": 0.4539312422275543, "learning_rate": 7.185021139161592e-05, "loss": 0.0381, "step": 31060 }, { "grad_norm": 0.38023293018341064, "learning_rate": 7.1831619162153e-05, "loss": 0.0471, "step": 31070 }, { "grad_norm": 0.2676549553871155, "learning_rate": 7.181302320226345e-05, "loss": 0.0323, "step": 31080 }, { "grad_norm": 0.5228103995323181, "learning_rate": 7.179442351512482e-05, "loss": 0.0427, "step": 31090 }, { "grad_norm": 0.49347206950187683, "learning_rate": 7.177582010391528e-05, "loss": 0.0446, "step": 31100 }, { "grad_norm": 0.5090417861938477, "learning_rate": 7.175721297181366e-05, "loss": 0.0544, "step": 31110 }, { "grad_norm": 0.596822202205658, "learning_rate": 7.173860212199942e-05, "loss": 0.0425, "step": 31120 }, { "grad_norm": 0.5817731022834778, "learning_rate": 7.171998755765263e-05, "loss": 0.06, "step": 31130 }, { "grad_norm": 0.7470046281814575, "learning_rate": 7.170136928195398e-05, "loss": 0.0502, "step": 31140 }, { "grad_norm": 0.5777588486671448, "learning_rate": 7.168274729808489e-05, "loss": 0.058, "step": 31150 }, { "grad_norm": 0.6278417110443115, "learning_rate": 7.166412160922728e-05, "loss": 0.0568, "step": 31160 }, { "grad_norm": 0.4542405307292938, "learning_rate": 7.164549221856382e-05, "loss": 0.0439, "step": 31170 }, { "grad_norm": 3.4782357215881348, "learning_rate": 7.162685912927775e-05, "loss": 0.0531, "step": 31180 }, { "grad_norm": 0.7300837635993958, "learning_rate": 7.160822234455294e-05, "loss": 0.0544, "step": 31190 }, { "grad_norm": 0.44775381684303284, "learning_rate": 7.158958186757391e-05, "loss": 0.045, "step": 31200 }, { "grad_norm": 0.5931660532951355, "learning_rate": 7.157093770152582e-05, "loss": 0.0563, "step": 31210 }, { "grad_norm": 0.46250206232070923, "learning_rate": 7.155228984959446e-05, "loss": 0.057, "step": 31220 }, { "grad_norm": 0.4872368574142456, "learning_rate": 7.153363831496621e-05, "loss": 0.0418, "step": 31230 }, { "grad_norm": 0.4920722544193268, "learning_rate": 7.151498310082811e-05, "loss": 0.0431, "step": 31240 }, { "grad_norm": 0.48163357377052307, "learning_rate": 7.149632421036784e-05, "loss": 0.0459, "step": 31250 }, { "grad_norm": 0.6887741088867188, "learning_rate": 7.147766164677369e-05, "loss": 0.0557, "step": 31260 }, { "grad_norm": 0.5322884917259216, "learning_rate": 7.145899541323459e-05, "loss": 0.0429, "step": 31270 }, { "grad_norm": 0.5660802721977234, "learning_rate": 7.144032551294007e-05, "loss": 0.0417, "step": 31280 }, { "grad_norm": 0.5015053749084473, "learning_rate": 7.14216519490803e-05, "loss": 0.0483, "step": 31290 }, { "grad_norm": 0.6775413751602173, "learning_rate": 7.140297472484609e-05, "loss": 0.0422, "step": 31300 }, { "grad_norm": 0.5622888803482056, "learning_rate": 7.138429384342891e-05, "loss": 0.0423, "step": 31310 }, { "grad_norm": 0.7602048516273499, "learning_rate": 7.136560930802074e-05, "loss": 0.0598, "step": 31320 }, { "grad_norm": 0.519782304763794, "learning_rate": 7.134692112181431e-05, "loss": 0.0408, "step": 31330 }, { "grad_norm": 0.6710020303726196, "learning_rate": 7.13282292880029e-05, "loss": 0.0539, "step": 31340 }, { "grad_norm": 0.35154446959495544, "learning_rate": 7.130953380978043e-05, "loss": 0.0523, "step": 31350 }, { "grad_norm": 0.32187679409980774, "learning_rate": 7.129083469034144e-05, "loss": 0.0414, "step": 31360 }, { "grad_norm": 0.32798823714256287, "learning_rate": 7.127213193288112e-05, "loss": 0.0445, "step": 31370 }, { "grad_norm": 0.5224504470825195, "learning_rate": 7.125342554059522e-05, "loss": 0.0501, "step": 31380 }, { "grad_norm": 0.501738965511322, "learning_rate": 7.12347155166802e-05, "loss": 0.0468, "step": 31390 }, { "grad_norm": 0.6729851365089417, "learning_rate": 7.121600186433306e-05, "loss": 0.053, "step": 31400 }, { "grad_norm": 0.5105595588684082, "learning_rate": 7.119728458675148e-05, "loss": 0.0555, "step": 31410 }, { "grad_norm": 0.6437304019927979, "learning_rate": 7.117856368713369e-05, "loss": 0.0542, "step": 31420 }, { "grad_norm": 0.39922353625297546, "learning_rate": 7.115983916867861e-05, "loss": 0.0465, "step": 31430 }, { "grad_norm": 0.6100232005119324, "learning_rate": 7.114111103458574e-05, "loss": 0.0363, "step": 31440 }, { "grad_norm": 0.6828243732452393, "learning_rate": 7.11223792880552e-05, "loss": 0.0443, "step": 31450 }, { "grad_norm": 0.6078697443008423, "learning_rate": 7.110364393228773e-05, "loss": 0.0522, "step": 31460 }, { "grad_norm": 0.5382857918739319, "learning_rate": 7.108490497048471e-05, "loss": 0.0437, "step": 31470 }, { "grad_norm": 0.42219972610473633, "learning_rate": 7.10661624058481e-05, "loss": 0.0415, "step": 31480 }, { "grad_norm": 0.8730614185333252, "learning_rate": 7.10474162415805e-05, "loss": 0.0597, "step": 31490 }, { "grad_norm": 0.6396957635879517, "learning_rate": 7.102866648088511e-05, "loss": 0.0447, "step": 31500 }, { "grad_norm": 0.43187323212623596, "learning_rate": 7.100991312696576e-05, "loss": 0.0432, "step": 31510 }, { "grad_norm": 0.5123885869979858, "learning_rate": 7.099115618302686e-05, "loss": 0.0481, "step": 31520 }, { "grad_norm": 0.6744593977928162, "learning_rate": 7.097239565227349e-05, "loss": 0.044, "step": 31530 }, { "grad_norm": 0.6310487389564514, "learning_rate": 7.09536315379113e-05, "loss": 0.0532, "step": 31540 }, { "grad_norm": 0.5698903799057007, "learning_rate": 7.093486384314656e-05, "loss": 0.0533, "step": 31550 }, { "grad_norm": 0.8446791768074036, "learning_rate": 7.091609257118616e-05, "loss": 0.0465, "step": 31560 }, { "grad_norm": 0.5153095126152039, "learning_rate": 7.08973177252376e-05, "loss": 0.0491, "step": 31570 }, { "grad_norm": 0.4752649962902069, "learning_rate": 7.087853930850898e-05, "loss": 0.053, "step": 31580 }, { "grad_norm": 0.687142550945282, "learning_rate": 7.085975732420903e-05, "loss": 0.0605, "step": 31590 }, { "grad_norm": 0.48356083035469055, "learning_rate": 7.084097177554706e-05, "loss": 0.0457, "step": 31600 }, { "grad_norm": 0.6020156145095825, "learning_rate": 7.082218266573301e-05, "loss": 0.0485, "step": 31610 }, { "grad_norm": 0.42620354890823364, "learning_rate": 7.080338999797743e-05, "loss": 0.0493, "step": 31620 }, { "grad_norm": 0.44415411353111267, "learning_rate": 7.07845937754915e-05, "loss": 0.0442, "step": 31630 }, { "grad_norm": 0.6563808917999268, "learning_rate": 7.076579400148693e-05, "loss": 0.0494, "step": 31640 }, { "grad_norm": 0.44954046607017517, "learning_rate": 7.074699067917611e-05, "loss": 0.0547, "step": 31650 }, { "grad_norm": 0.5375992059707642, "learning_rate": 7.072818381177201e-05, "loss": 0.0384, "step": 31660 }, { "grad_norm": 1.1781549453735352, "learning_rate": 7.070937340248823e-05, "loss": 0.0492, "step": 31670 }, { "grad_norm": 0.5899049043655396, "learning_rate": 7.069055945453893e-05, "loss": 0.0421, "step": 31680 }, { "grad_norm": 0.5309367179870605, "learning_rate": 7.067174197113892e-05, "loss": 0.0443, "step": 31690 }, { "grad_norm": 0.7489214539527893, "learning_rate": 7.065292095550355e-05, "loss": 0.0517, "step": 31700 }, { "grad_norm": 0.4341561496257782, "learning_rate": 7.063409641084887e-05, "loss": 0.0454, "step": 31710 }, { "grad_norm": 0.5190590620040894, "learning_rate": 7.061526834039145e-05, "loss": 0.0425, "step": 31720 }, { "grad_norm": 0.7082948088645935, "learning_rate": 7.05964367473485e-05, "loss": 0.0577, "step": 31730 }, { "grad_norm": 0.4200780987739563, "learning_rate": 7.057760163493783e-05, "loss": 0.0454, "step": 31740 }, { "grad_norm": 0.6458917856216431, "learning_rate": 7.055876300637783e-05, "loss": 0.0564, "step": 31750 }, { "grad_norm": 0.5127013921737671, "learning_rate": 7.053992086488753e-05, "loss": 0.0486, "step": 31760 }, { "grad_norm": 0.6523547768592834, "learning_rate": 7.052107521368651e-05, "loss": 0.0574, "step": 31770 }, { "grad_norm": 0.6262826919555664, "learning_rate": 7.0502226055995e-05, "loss": 0.0478, "step": 31780 }, { "grad_norm": 0.4694984257221222, "learning_rate": 7.048337339503379e-05, "loss": 0.0357, "step": 31790 }, { "grad_norm": 0.985752284526825, "learning_rate": 7.046451723402427e-05, "loss": 0.0445, "step": 31800 }, { "grad_norm": 0.4965660274028778, "learning_rate": 7.044565757618848e-05, "loss": 0.0424, "step": 31810 }, { "grad_norm": 0.5481279492378235, "learning_rate": 7.042679442474899e-05, "loss": 0.0411, "step": 31820 }, { "grad_norm": 0.4251774847507477, "learning_rate": 7.040792778292902e-05, "loss": 0.0406, "step": 31830 }, { "grad_norm": 0.4486905038356781, "learning_rate": 7.038905765395234e-05, "loss": 0.0457, "step": 31840 }, { "grad_norm": 0.5462509393692017, "learning_rate": 7.037018404104334e-05, "loss": 0.0402, "step": 31850 }, { "grad_norm": 0.6094315052032471, "learning_rate": 7.035130694742702e-05, "loss": 0.0447, "step": 31860 }, { "grad_norm": 0.3897845149040222, "learning_rate": 7.033242637632897e-05, "loss": 0.0336, "step": 31870 }, { "grad_norm": 1.2026091814041138, "learning_rate": 7.031354233097534e-05, "loss": 0.0357, "step": 31880 }, { "grad_norm": 0.42873114347457886, "learning_rate": 7.029465481459289e-05, "loss": 0.0422, "step": 31890 }, { "grad_norm": 0.6039527654647827, "learning_rate": 7.027576383040898e-05, "loss": 0.0388, "step": 31900 }, { "grad_norm": 0.43061432242393494, "learning_rate": 7.025686938165159e-05, "loss": 0.0405, "step": 31910 }, { "grad_norm": 0.55738765001297, "learning_rate": 7.023797147154924e-05, "loss": 0.0473, "step": 31920 }, { "grad_norm": 0.6616368293762207, "learning_rate": 7.021907010333111e-05, "loss": 0.0617, "step": 31930 }, { "grad_norm": 0.5103512406349182, "learning_rate": 7.020016528022685e-05, "loss": 0.0553, "step": 31940 }, { "grad_norm": 0.6229791641235352, "learning_rate": 7.018125700546683e-05, "loss": 0.0379, "step": 31950 }, { "grad_norm": 0.6929023265838623, "learning_rate": 7.016234528228196e-05, "loss": 0.0378, "step": 31960 }, { "grad_norm": 0.42539680004119873, "learning_rate": 7.014343011390372e-05, "loss": 0.052, "step": 31970 }, { "grad_norm": 0.43983379006385803, "learning_rate": 7.01245115035642e-05, "loss": 0.0564, "step": 31980 }, { "grad_norm": 0.4863318204879761, "learning_rate": 7.010558945449606e-05, "loss": 0.0455, "step": 31990 }, { "grad_norm": 0.5113188624382019, "learning_rate": 7.008666396993258e-05, "loss": 0.0391, "step": 32000 }, { "grad_norm": 0.5232723355293274, "learning_rate": 7.006773505310759e-05, "loss": 0.054, "step": 32010 }, { "grad_norm": 0.6583823561668396, "learning_rate": 7.004880270725553e-05, "loss": 0.048, "step": 32020 }, { "grad_norm": 0.5527577996253967, "learning_rate": 7.002986693561144e-05, "loss": 0.0526, "step": 32030 }, { "grad_norm": 0.6345615983009338, "learning_rate": 7.001092774141089e-05, "loss": 0.0438, "step": 32040 }, { "grad_norm": 0.5735673904418945, "learning_rate": 6.999198512789009e-05, "loss": 0.0534, "step": 32050 }, { "grad_norm": 0.47921866178512573, "learning_rate": 6.997303909828584e-05, "loss": 0.0488, "step": 32060 }, { "grad_norm": 0.8625393509864807, "learning_rate": 6.995408965583544e-05, "loss": 0.0475, "step": 32070 }, { "grad_norm": 0.39821815490722656, "learning_rate": 6.993513680377688e-05, "loss": 0.0484, "step": 32080 }, { "grad_norm": 0.8898906111717224, "learning_rate": 6.991618054534868e-05, "loss": 0.0442, "step": 32090 }, { "grad_norm": 0.3366783559322357, "learning_rate": 6.989722088378991e-05, "loss": 0.0405, "step": 32100 }, { "grad_norm": 0.549835205078125, "learning_rate": 6.987825782234027e-05, "loss": 0.0409, "step": 32110 }, { "grad_norm": 0.9733661413192749, "learning_rate": 6.985929136424006e-05, "loss": 0.0466, "step": 32120 }, { "grad_norm": 0.5969974398612976, "learning_rate": 6.984032151273012e-05, "loss": 0.0632, "step": 32130 }, { "grad_norm": 1.2620141506195068, "learning_rate": 6.982134827105186e-05, "loss": 0.0552, "step": 32140 }, { "grad_norm": 0.6178591251373291, "learning_rate": 6.980237164244729e-05, "loss": 0.0496, "step": 32150 }, { "grad_norm": 0.6174449920654297, "learning_rate": 6.9783391630159e-05, "loss": 0.0486, "step": 32160 }, { "grad_norm": 0.6586732268333435, "learning_rate": 6.976440823743015e-05, "loss": 0.0469, "step": 32170 }, { "grad_norm": 0.3841414153575897, "learning_rate": 6.974542146750451e-05, "loss": 0.0511, "step": 32180 }, { "grad_norm": 0.7473002672195435, "learning_rate": 6.972643132362637e-05, "loss": 0.0597, "step": 32190 }, { "grad_norm": 0.5361340641975403, "learning_rate": 6.970743780904064e-05, "loss": 0.0517, "step": 32200 }, { "grad_norm": 0.6139820218086243, "learning_rate": 6.968844092699277e-05, "loss": 0.0589, "step": 32210 }, { "grad_norm": 0.6009312272071838, "learning_rate": 6.966944068072883e-05, "loss": 0.0484, "step": 32220 }, { "grad_norm": 0.42381206154823303, "learning_rate": 6.965043707349545e-05, "loss": 0.0589, "step": 32230 }, { "grad_norm": 0.5959138870239258, "learning_rate": 6.963143010853982e-05, "loss": 0.0536, "step": 32240 }, { "grad_norm": 0.5136887431144714, "learning_rate": 6.961241978910971e-05, "loss": 0.0406, "step": 32250 }, { "grad_norm": 0.6322599649429321, "learning_rate": 6.959340611845344e-05, "loss": 0.0506, "step": 32260 }, { "grad_norm": 0.6440425515174866, "learning_rate": 6.957438909981995e-05, "loss": 0.0394, "step": 32270 }, { "grad_norm": 0.5925201177597046, "learning_rate": 6.955536873645872e-05, "loss": 0.0553, "step": 32280 }, { "grad_norm": 0.47919216752052307, "learning_rate": 6.953634503161982e-05, "loss": 0.0464, "step": 32290 }, { "grad_norm": 0.5085619688034058, "learning_rate": 6.951731798855387e-05, "loss": 0.0457, "step": 32300 }, { "grad_norm": 0.6383436322212219, "learning_rate": 6.949828761051208e-05, "loss": 0.0618, "step": 32310 }, { "grad_norm": 0.48543208837509155, "learning_rate": 6.947925390074622e-05, "loss": 0.0402, "step": 32320 }, { "grad_norm": 0.34783586859703064, "learning_rate": 6.946021686250863e-05, "loss": 0.0433, "step": 32330 }, { "grad_norm": 0.3372413218021393, "learning_rate": 6.94411764990522e-05, "loss": 0.0353, "step": 32340 }, { "grad_norm": 0.5778287053108215, "learning_rate": 6.942213281363044e-05, "loss": 0.0436, "step": 32350 }, { "grad_norm": 0.48784002661705017, "learning_rate": 6.940308580949737e-05, "loss": 0.04, "step": 32360 }, { "grad_norm": 0.549975574016571, "learning_rate": 6.93840354899076e-05, "loss": 0.0372, "step": 32370 }, { "grad_norm": 0.7051849365234375, "learning_rate": 6.936498185811633e-05, "loss": 0.0389, "step": 32380 }, { "grad_norm": 0.416452020406723, "learning_rate": 6.93459249173793e-05, "loss": 0.0465, "step": 32390 }, { "grad_norm": 0.44418689608573914, "learning_rate": 6.932686467095279e-05, "loss": 0.0379, "step": 32400 }, { "grad_norm": 0.45183759927749634, "learning_rate": 6.930780112209373e-05, "loss": 0.0694, "step": 32410 }, { "grad_norm": 0.44034111499786377, "learning_rate": 6.92887342740595e-05, "loss": 0.0392, "step": 32420 }, { "grad_norm": 1.0614128112792969, "learning_rate": 6.926966413010816e-05, "loss": 0.0486, "step": 32430 }, { "grad_norm": 0.36792609095573425, "learning_rate": 6.925059069349824e-05, "loss": 0.0326, "step": 32440 }, { "grad_norm": 0.5467274785041809, "learning_rate": 6.923151396748886e-05, "loss": 0.0471, "step": 32450 }, { "grad_norm": 0.30920788645744324, "learning_rate": 6.921243395533974e-05, "loss": 0.0549, "step": 32460 }, { "grad_norm": 0.4985678493976593, "learning_rate": 6.919335066031109e-05, "loss": 0.0521, "step": 32470 }, { "grad_norm": 0.4218123257160187, "learning_rate": 6.917426408566379e-05, "loss": 0.0429, "step": 32480 }, { "grad_norm": 0.29808998107910156, "learning_rate": 6.915517423465916e-05, "loss": 0.0475, "step": 32490 }, { "grad_norm": 0.48718327283859253, "learning_rate": 6.913608111055914e-05, "loss": 0.049, "step": 32500 }, { "grad_norm": 0.5678977370262146, "learning_rate": 6.911698471662623e-05, "loss": 0.0474, "step": 32510 }, { "grad_norm": 0.6086316108703613, "learning_rate": 6.90978850561235e-05, "loss": 0.056, "step": 32520 }, { "grad_norm": 0.4292129874229431, "learning_rate": 6.907878213231454e-05, "loss": 0.0377, "step": 32530 }, { "grad_norm": 0.3964032232761383, "learning_rate": 6.90596759484635e-05, "loss": 0.055, "step": 32540 }, { "grad_norm": 0.44263026118278503, "learning_rate": 6.904056650783514e-05, "loss": 0.0632, "step": 32550 }, { "grad_norm": 0.47635418176651, "learning_rate": 6.902145381369471e-05, "loss": 0.0395, "step": 32560 }, { "grad_norm": 0.5356208682060242, "learning_rate": 6.900233786930808e-05, "loss": 0.0498, "step": 32570 }, { "grad_norm": 0.3565826714038849, "learning_rate": 6.898321867794161e-05, "loss": 0.0411, "step": 32580 }, { "grad_norm": 0.4702075123786926, "learning_rate": 6.896409624286226e-05, "loss": 0.0313, "step": 32590 }, { "grad_norm": 0.37954282760620117, "learning_rate": 6.894497056733754e-05, "loss": 0.0411, "step": 32600 }, { "grad_norm": 0.5787203907966614, "learning_rate": 6.89258416546355e-05, "loss": 0.0572, "step": 32610 }, { "grad_norm": 0.42885035276412964, "learning_rate": 6.890670950802474e-05, "loss": 0.0542, "step": 32620 }, { "grad_norm": 0.4583437740802765, "learning_rate": 6.88875741307744e-05, "loss": 0.0344, "step": 32630 }, { "grad_norm": 0.4073426425457001, "learning_rate": 6.886843552615425e-05, "loss": 0.0447, "step": 32640 }, { "grad_norm": 0.44951605796813965, "learning_rate": 6.884929369743451e-05, "loss": 0.0583, "step": 32650 }, { "grad_norm": 0.6357453465461731, "learning_rate": 6.8830148647886e-05, "loss": 0.0457, "step": 32660 }, { "grad_norm": 0.4980791509151459, "learning_rate": 6.88110003807801e-05, "loss": 0.049, "step": 32670 }, { "grad_norm": 0.5972156524658203, "learning_rate": 6.87918488993887e-05, "loss": 0.0406, "step": 32680 }, { "grad_norm": 0.8119646906852722, "learning_rate": 6.877269420698431e-05, "loss": 0.0431, "step": 32690 }, { "grad_norm": 0.44922444224357605, "learning_rate": 6.875353630683989e-05, "loss": 0.0434, "step": 32700 }, { "grad_norm": 0.5342782139778137, "learning_rate": 6.873437520222905e-05, "loss": 0.047, "step": 32710 }, { "grad_norm": 0.4981009364128113, "learning_rate": 6.871521089642585e-05, "loss": 0.0394, "step": 32720 }, { "grad_norm": 0.6831677556037903, "learning_rate": 6.869604339270498e-05, "loss": 0.0436, "step": 32730 }, { "grad_norm": 0.7114524841308594, "learning_rate": 6.867687269434164e-05, "loss": 0.0571, "step": 32740 }, { "grad_norm": 0.7416750192642212, "learning_rate": 6.865769880461156e-05, "loss": 0.0508, "step": 32750 }, { "grad_norm": 0.5762982964515686, "learning_rate": 6.863852172679104e-05, "loss": 0.0348, "step": 32760 }, { "grad_norm": 0.4358603358268738, "learning_rate": 6.861934146415693e-05, "loss": 0.0491, "step": 32770 }, { "grad_norm": 0.45758193731307983, "learning_rate": 6.86001580199866e-05, "loss": 0.0466, "step": 32780 }, { "grad_norm": 0.6336572170257568, "learning_rate": 6.858097139755798e-05, "loss": 0.0477, "step": 32790 }, { "grad_norm": 0.6015558242797852, "learning_rate": 6.856178160014955e-05, "loss": 0.0439, "step": 32800 }, { "grad_norm": 0.5854575634002686, "learning_rate": 6.85425886310403e-05, "loss": 0.0463, "step": 32810 }, { "grad_norm": 0.6651691198348999, "learning_rate": 6.852339249350979e-05, "loss": 0.0453, "step": 32820 }, { "grad_norm": 0.4989462196826935, "learning_rate": 6.850419319083812e-05, "loss": 0.0428, "step": 32830 }, { "grad_norm": 0.5125022530555725, "learning_rate": 6.848499072630592e-05, "loss": 0.0465, "step": 32840 }, { "grad_norm": 0.5156419277191162, "learning_rate": 6.846578510319439e-05, "loss": 0.0528, "step": 32850 }, { "grad_norm": 0.6997029781341553, "learning_rate": 6.844657632478519e-05, "loss": 0.0572, "step": 32860 }, { "grad_norm": 0.5934537053108215, "learning_rate": 6.842736439436063e-05, "loss": 0.0399, "step": 32870 }, { "grad_norm": 0.5580911040306091, "learning_rate": 6.84081493152035e-05, "loss": 0.0416, "step": 32880 }, { "grad_norm": 0.4605739712715149, "learning_rate": 6.83889310905971e-05, "loss": 0.041, "step": 32890 }, { "grad_norm": 0.9710685014724731, "learning_rate": 6.836970972382533e-05, "loss": 0.0448, "step": 32900 }, { "grad_norm": 0.4770044982433319, "learning_rate": 6.835048521817257e-05, "loss": 0.0415, "step": 32910 }, { "grad_norm": 0.38299307227134705, "learning_rate": 6.833125757692379e-05, "loss": 0.056, "step": 32920 }, { "grad_norm": 0.4884554147720337, "learning_rate": 6.831202680336441e-05, "loss": 0.0442, "step": 32930 }, { "grad_norm": 0.5136785507202148, "learning_rate": 6.829279290078052e-05, "loss": 0.045, "step": 32940 }, { "grad_norm": 0.4947223365306854, "learning_rate": 6.827355587245863e-05, "loss": 0.0383, "step": 32950 }, { "grad_norm": 0.5507800579071045, "learning_rate": 6.82543157216858e-05, "loss": 0.0581, "step": 32960 }, { "grad_norm": 0.369097501039505, "learning_rate": 6.823507245174969e-05, "loss": 0.0322, "step": 32970 }, { "grad_norm": 0.7035688161849976, "learning_rate": 6.821582606593841e-05, "loss": 0.0495, "step": 32980 }, { "grad_norm": 0.4811916947364807, "learning_rate": 6.81965765675407e-05, "loss": 0.0413, "step": 32990 }, { "grad_norm": 0.47057145833969116, "learning_rate": 6.81773239598457e-05, "loss": 0.0373, "step": 33000 }, { "grad_norm": 0.6445510983467102, "learning_rate": 6.815806824614319e-05, "loss": 0.0415, "step": 33010 }, { "grad_norm": 0.9276332259178162, "learning_rate": 6.813880942972343e-05, "loss": 0.0462, "step": 33020 }, { "grad_norm": 0.4664739668369293, "learning_rate": 6.811954751387726e-05, "loss": 0.0431, "step": 33030 }, { "grad_norm": 0.5559079051017761, "learning_rate": 6.810028250189598e-05, "loss": 0.051, "step": 33040 }, { "grad_norm": 0.5706468224525452, "learning_rate": 6.808101439707147e-05, "loss": 0.0449, "step": 33050 }, { "grad_norm": 0.6309397220611572, "learning_rate": 6.806174320269609e-05, "loss": 0.0486, "step": 33060 }, { "grad_norm": 0.34772592782974243, "learning_rate": 6.804246892206281e-05, "loss": 0.0324, "step": 33070 }, { "grad_norm": 1.1387739181518555, "learning_rate": 6.802319155846506e-05, "loss": 0.0523, "step": 33080 }, { "grad_norm": 0.41563379764556885, "learning_rate": 6.800391111519679e-05, "loss": 0.0459, "step": 33090 }, { "grad_norm": 0.5114373564720154, "learning_rate": 6.798462759555253e-05, "loss": 0.0446, "step": 33100 }, { "grad_norm": 0.5748064517974854, "learning_rate": 6.79653410028273e-05, "loss": 0.0559, "step": 33110 }, { "grad_norm": 0.6081279516220093, "learning_rate": 6.794605134031663e-05, "loss": 0.0549, "step": 33120 }, { "grad_norm": 0.5078449845314026, "learning_rate": 6.792675861131661e-05, "loss": 0.0427, "step": 33130 }, { "grad_norm": 0.3353295922279358, "learning_rate": 6.790746281912386e-05, "loss": 0.0424, "step": 33140 }, { "grad_norm": 0.4710392355918884, "learning_rate": 6.788816396703546e-05, "loss": 0.0463, "step": 33150 }, { "grad_norm": 0.43697574734687805, "learning_rate": 6.78688620583491e-05, "loss": 0.0396, "step": 33160 }, { "grad_norm": 0.5664297938346863, "learning_rate": 6.784955709636292e-05, "loss": 0.0326, "step": 33170 }, { "grad_norm": 0.47976142168045044, "learning_rate": 6.783024908437564e-05, "loss": 0.035, "step": 33180 }, { "grad_norm": 0.32035019993782043, "learning_rate": 6.781093802568641e-05, "loss": 0.0378, "step": 33190 }, { "grad_norm": 0.5379282236099243, "learning_rate": 6.779162392359504e-05, "loss": 0.0441, "step": 33200 }, { "grad_norm": 0.7329285144805908, "learning_rate": 6.777230678140172e-05, "loss": 0.0382, "step": 33210 }, { "grad_norm": 0.4626694619655609, "learning_rate": 6.775298660240726e-05, "loss": 0.0474, "step": 33220 }, { "grad_norm": 0.39914897084236145, "learning_rate": 6.773366338991292e-05, "loss": 0.0464, "step": 33230 }, { "grad_norm": 0.35344335436820984, "learning_rate": 6.771433714722052e-05, "loss": 0.0393, "step": 33240 }, { "grad_norm": 0.4324227273464203, "learning_rate": 6.769500787763239e-05, "loss": 0.0342, "step": 33250 }, { "grad_norm": 0.33053386211395264, "learning_rate": 6.76756755844514e-05, "loss": 0.0331, "step": 33260 }, { "grad_norm": 0.4795700013637543, "learning_rate": 6.765634027098087e-05, "loss": 0.0412, "step": 33270 }, { "grad_norm": 0.5042476058006287, "learning_rate": 6.763700194052468e-05, "loss": 0.0618, "step": 33280 }, { "grad_norm": 0.6773169040679932, "learning_rate": 6.761766059638723e-05, "loss": 0.0495, "step": 33290 }, { "grad_norm": 0.40264028310775757, "learning_rate": 6.759831624187345e-05, "loss": 0.0395, "step": 33300 }, { "grad_norm": 0.634112536907196, "learning_rate": 6.757896888028871e-05, "loss": 0.043, "step": 33310 }, { "grad_norm": 0.6438288688659668, "learning_rate": 6.7559618514939e-05, "loss": 0.0415, "step": 33320 }, { "grad_norm": 0.46204873919487, "learning_rate": 6.754026514913073e-05, "loss": 0.032, "step": 33330 }, { "grad_norm": 0.5916316509246826, "learning_rate": 6.752090878617087e-05, "loss": 0.0795, "step": 33340 }, { "grad_norm": 0.2867026925086975, "learning_rate": 6.75015494293669e-05, "loss": 0.0285, "step": 33350 }, { "grad_norm": 0.45954230427742004, "learning_rate": 6.74821870820268e-05, "loss": 0.0444, "step": 33360 }, { "grad_norm": 0.3619239032268524, "learning_rate": 6.746282174745907e-05, "loss": 0.0359, "step": 33370 }, { "grad_norm": 0.4868829846382141, "learning_rate": 6.744345342897271e-05, "loss": 0.0475, "step": 33380 }, { "grad_norm": 0.454057902097702, "learning_rate": 6.742408212987724e-05, "loss": 0.0405, "step": 33390 }, { "grad_norm": 0.38673949241638184, "learning_rate": 6.740470785348269e-05, "loss": 0.0364, "step": 33400 }, { "grad_norm": 0.5420302748680115, "learning_rate": 6.738533060309958e-05, "loss": 0.0484, "step": 33410 }, { "grad_norm": 0.6648820638656616, "learning_rate": 6.736595038203894e-05, "loss": 0.0417, "step": 33420 }, { "grad_norm": 0.7704416513442993, "learning_rate": 6.734656719361236e-05, "loss": 0.0595, "step": 33430 }, { "grad_norm": 0.5572773814201355, "learning_rate": 6.732718104113189e-05, "loss": 0.0507, "step": 33440 }, { "grad_norm": 0.5962262153625488, "learning_rate": 6.730779192791006e-05, "loss": 0.0398, "step": 33450 }, { "grad_norm": 0.417297899723053, "learning_rate": 6.728839985725997e-05, "loss": 0.0631, "step": 33460 }, { "grad_norm": 0.39306414127349854, "learning_rate": 6.726900483249517e-05, "loss": 0.0509, "step": 33470 }, { "grad_norm": 0.5219950079917908, "learning_rate": 6.724960685692976e-05, "loss": 0.0523, "step": 33480 }, { "grad_norm": 0.42649197578430176, "learning_rate": 6.723020593387833e-05, "loss": 0.0407, "step": 33490 }, { "grad_norm": 0.7330330610275269, "learning_rate": 6.721080206665593e-05, "loss": 0.0582, "step": 33500 }, { "grad_norm": 0.27937865257263184, "learning_rate": 6.719139525857819e-05, "loss": 0.0334, "step": 33510 }, { "grad_norm": 0.562470555305481, "learning_rate": 6.717198551296117e-05, "loss": 0.0408, "step": 33520 }, { "grad_norm": 0.4942508041858673, "learning_rate": 6.715257283312148e-05, "loss": 0.0501, "step": 33530 }, { "grad_norm": 0.5617299675941467, "learning_rate": 6.713315722237623e-05, "loss": 0.0414, "step": 33540 }, { "grad_norm": 0.4485554099082947, "learning_rate": 6.7113738684043e-05, "loss": 0.042, "step": 33550 }, { "grad_norm": 0.6629064679145813, "learning_rate": 6.709431722143989e-05, "loss": 0.0565, "step": 33560 }, { "grad_norm": 0.44343411922454834, "learning_rate": 6.70748928378855e-05, "loss": 0.0506, "step": 33570 }, { "grad_norm": 0.33452925086021423, "learning_rate": 6.705546553669891e-05, "loss": 0.033, "step": 33580 }, { "grad_norm": 0.6381985545158386, "learning_rate": 6.703603532119974e-05, "loss": 0.0457, "step": 33590 }, { "grad_norm": 0.6767544746398926, "learning_rate": 6.701660219470808e-05, "loss": 0.0431, "step": 33600 }, { "grad_norm": 0.8863798379898071, "learning_rate": 6.69971661605445e-05, "loss": 0.0439, "step": 33610 }, { "grad_norm": 0.5558401942253113, "learning_rate": 6.697772722203008e-05, "loss": 0.0442, "step": 33620 }, { "grad_norm": 0.5499553680419922, "learning_rate": 6.695828538248643e-05, "loss": 0.043, "step": 33630 }, { "grad_norm": 0.3292674124240875, "learning_rate": 6.693884064523563e-05, "loss": 0.0365, "step": 33640 }, { "grad_norm": 0.4829912483692169, "learning_rate": 6.691939301360023e-05, "loss": 0.0461, "step": 33650 }, { "grad_norm": 0.5207998156547546, "learning_rate": 6.689994249090333e-05, "loss": 0.036, "step": 33660 }, { "grad_norm": 0.4582897126674652, "learning_rate": 6.688048908046845e-05, "loss": 0.0376, "step": 33670 }, { "grad_norm": 0.5684102177619934, "learning_rate": 6.686103278561969e-05, "loss": 0.0414, "step": 33680 }, { "grad_norm": 0.7413336038589478, "learning_rate": 6.684157360968156e-05, "loss": 0.0545, "step": 33690 }, { "grad_norm": 0.8067909479141235, "learning_rate": 6.682211155597911e-05, "loss": 0.0459, "step": 33700 }, { "grad_norm": 0.5013560056686401, "learning_rate": 6.680264662783789e-05, "loss": 0.0434, "step": 33710 }, { "grad_norm": 0.6267509460449219, "learning_rate": 6.678317882858391e-05, "loss": 0.0522, "step": 33720 }, { "grad_norm": 0.5279382467269897, "learning_rate": 6.67637081615437e-05, "loss": 0.0436, "step": 33730 }, { "grad_norm": 0.5299199819564819, "learning_rate": 6.674423463004427e-05, "loss": 0.039, "step": 33740 }, { "grad_norm": 0.47330692410469055, "learning_rate": 6.672475823741308e-05, "loss": 0.0357, "step": 33750 }, { "grad_norm": 0.4944644868373871, "learning_rate": 6.670527898697811e-05, "loss": 0.0422, "step": 33760 }, { "grad_norm": 0.45693299174308777, "learning_rate": 6.668579688206788e-05, "loss": 0.0491, "step": 33770 }, { "grad_norm": 0.6432951092720032, "learning_rate": 6.666631192601131e-05, "loss": 0.0401, "step": 33780 }, { "grad_norm": 1.007185935974121, "learning_rate": 6.664682412213785e-05, "loss": 0.0394, "step": 33790 }, { "grad_norm": 0.42167171835899353, "learning_rate": 6.662733347377745e-05, "loss": 0.0429, "step": 33800 }, { "grad_norm": 0.5172095894813538, "learning_rate": 6.660783998426051e-05, "loss": 0.04, "step": 33810 }, { "grad_norm": 0.5901033878326416, "learning_rate": 6.658834365691794e-05, "loss": 0.0509, "step": 33820 }, { "grad_norm": 0.6591936349868774, "learning_rate": 6.656884449508115e-05, "loss": 0.0426, "step": 33830 }, { "grad_norm": 0.5527634024620056, "learning_rate": 6.654934250208198e-05, "loss": 0.0401, "step": 33840 }, { "grad_norm": 0.3631369471549988, "learning_rate": 6.65298376812528e-05, "loss": 0.0438, "step": 33850 }, { "grad_norm": 0.7092177867889404, "learning_rate": 6.651033003592646e-05, "loss": 0.0456, "step": 33860 }, { "grad_norm": 0.549792468547821, "learning_rate": 6.649081956943626e-05, "loss": 0.0548, "step": 33870 }, { "grad_norm": 0.5583163499832153, "learning_rate": 6.647130628511604e-05, "loss": 0.0373, "step": 33880 }, { "grad_norm": 0.36360400915145874, "learning_rate": 6.645179018630005e-05, "loss": 0.0464, "step": 33890 }, { "grad_norm": 0.43291935324668884, "learning_rate": 6.643227127632309e-05, "loss": 0.0512, "step": 33900 }, { "grad_norm": 0.5497254729270935, "learning_rate": 6.641274955852038e-05, "loss": 0.047, "step": 33910 }, { "grad_norm": 0.5820241570472717, "learning_rate": 6.639322503622768e-05, "loss": 0.0469, "step": 33920 }, { "grad_norm": 0.6278632879257202, "learning_rate": 6.637369771278116e-05, "loss": 0.0561, "step": 33930 }, { "grad_norm": 0.6340512037277222, "learning_rate": 6.635416759151751e-05, "loss": 0.0522, "step": 33940 }, { "grad_norm": 0.540628969669342, "learning_rate": 6.633463467577394e-05, "loss": 0.0439, "step": 33950 }, { "grad_norm": 0.6136651635169983, "learning_rate": 6.631509896888803e-05, "loss": 0.0525, "step": 33960 }, { "grad_norm": 0.6524490714073181, "learning_rate": 6.629556047419794e-05, "loss": 0.0518, "step": 33970 }, { "grad_norm": 0.5239415168762207, "learning_rate": 6.627601919504223e-05, "loss": 0.0545, "step": 33980 }, { "grad_norm": 0.46061959862709045, "learning_rate": 6.625647513476001e-05, "loss": 0.0491, "step": 33990 }, { "grad_norm": 0.8369355797767639, "learning_rate": 6.62369282966908e-05, "loss": 0.0535, "step": 34000 }, { "grad_norm": 0.4400959014892578, "learning_rate": 6.621737868417464e-05, "loss": 0.0419, "step": 34010 }, { "grad_norm": 0.6892714500427246, "learning_rate": 6.619782630055198e-05, "loss": 0.056, "step": 34020 }, { "grad_norm": 0.5152133107185364, "learning_rate": 6.617827114916382e-05, "loss": 0.0541, "step": 34030 }, { "grad_norm": 0.5648447871208191, "learning_rate": 6.615871323335161e-05, "loss": 0.0529, "step": 34040 }, { "grad_norm": 0.4707593619823456, "learning_rate": 6.613915255645725e-05, "loss": 0.039, "step": 34050 }, { "grad_norm": 0.49616649746894836, "learning_rate": 6.611958912182312e-05, "loss": 0.0407, "step": 34060 }, { "grad_norm": 0.7273127436637878, "learning_rate": 6.610002293279207e-05, "loss": 0.0479, "step": 34070 }, { "grad_norm": 0.917661726474762, "learning_rate": 6.608045399270746e-05, "loss": 0.0382, "step": 34080 }, { "grad_norm": 0.3672366738319397, "learning_rate": 6.606088230491304e-05, "loss": 0.0435, "step": 34090 }, { "grad_norm": 0.6331964731216431, "learning_rate": 6.604130787275312e-05, "loss": 0.0451, "step": 34100 }, { "grad_norm": 0.9835085868835449, "learning_rate": 6.602173069957242e-05, "loss": 0.0422, "step": 34110 }, { "grad_norm": 0.8350191116333008, "learning_rate": 6.600215078871612e-05, "loss": 0.0741, "step": 34120 }, { "grad_norm": 0.5596519112586975, "learning_rate": 6.598256814352992e-05, "loss": 0.0388, "step": 34130 }, { "grad_norm": 0.40463000535964966, "learning_rate": 6.596298276735995e-05, "loss": 0.0455, "step": 34140 }, { "grad_norm": 0.44835928082466125, "learning_rate": 6.594339466355282e-05, "loss": 0.0418, "step": 34150 }, { "grad_norm": 0.5969957113265991, "learning_rate": 6.592380383545558e-05, "loss": 0.0348, "step": 34160 }, { "grad_norm": 0.38324880599975586, "learning_rate": 6.590421028641577e-05, "loss": 0.0403, "step": 34170 }, { "grad_norm": 0.55169278383255, "learning_rate": 6.588461401978143e-05, "loss": 0.0431, "step": 34180 }, { "grad_norm": 0.5275944471359253, "learning_rate": 6.586501503890099e-05, "loss": 0.0438, "step": 34190 }, { "grad_norm": 0.4180741608142853, "learning_rate": 6.584541334712338e-05, "loss": 0.0437, "step": 34200 }, { "grad_norm": 0.7628477215766907, "learning_rate": 6.582580894779802e-05, "loss": 0.0371, "step": 34210 }, { "grad_norm": 0.541207492351532, "learning_rate": 6.580620184427473e-05, "loss": 0.0612, "step": 34220 }, { "grad_norm": 0.8186712265014648, "learning_rate": 6.578659203990385e-05, "loss": 0.0435, "step": 34230 }, { "grad_norm": 0.5789433717727661, "learning_rate": 6.576697953803615e-05, "loss": 0.0436, "step": 34240 }, { "grad_norm": 0.407461017370224, "learning_rate": 6.57473643420229e-05, "loss": 0.0388, "step": 34250 }, { "grad_norm": 0.8085768222808838, "learning_rate": 6.572774645521574e-05, "loss": 0.0445, "step": 34260 }, { "grad_norm": 0.49283379316329956, "learning_rate": 6.570812588096688e-05, "loss": 0.0327, "step": 34270 }, { "grad_norm": 0.4754540026187897, "learning_rate": 6.568850262262893e-05, "loss": 0.0412, "step": 34280 }, { "grad_norm": 0.748955488204956, "learning_rate": 6.566887668355497e-05, "loss": 0.035, "step": 34290 }, { "grad_norm": 0.5171302556991577, "learning_rate": 6.564924806709851e-05, "loss": 0.0421, "step": 34300 }, { "grad_norm": 0.42636725306510925, "learning_rate": 6.562961677661359e-05, "loss": 0.0388, "step": 34310 }, { "grad_norm": 0.46966588497161865, "learning_rate": 6.56099828154546e-05, "loss": 0.0495, "step": 34320 }, { "grad_norm": 0.5967839956283569, "learning_rate": 6.55903461869765e-05, "loss": 0.0564, "step": 34330 }, { "grad_norm": 0.48365890979766846, "learning_rate": 6.557070689453465e-05, "loss": 0.0401, "step": 34340 }, { "grad_norm": 0.3763231337070465, "learning_rate": 6.555106494148482e-05, "loss": 0.0382, "step": 34350 }, { "grad_norm": 0.5448890328407288, "learning_rate": 6.553142033118333e-05, "loss": 0.0525, "step": 34360 }, { "grad_norm": 0.5141479969024658, "learning_rate": 6.551177306698688e-05, "loss": 0.0407, "step": 34370 }, { "grad_norm": 0.6786214709281921, "learning_rate": 6.549212315225267e-05, "loss": 0.0427, "step": 34380 }, { "grad_norm": 0.3660305440425873, "learning_rate": 6.547247059033833e-05, "loss": 0.0502, "step": 34390 }, { "grad_norm": 0.5028467178344727, "learning_rate": 6.545281538460193e-05, "loss": 0.0477, "step": 34400 }, { "grad_norm": 0.46729475259780884, "learning_rate": 6.543315753840202e-05, "loss": 0.0391, "step": 34410 }, { "grad_norm": 0.5968579053878784, "learning_rate": 6.541349705509758e-05, "loss": 0.0476, "step": 34420 }, { "grad_norm": 0.46243298053741455, "learning_rate": 6.539383393804805e-05, "loss": 0.0508, "step": 34430 }, { "grad_norm": 0.5248540043830872, "learning_rate": 6.537416819061333e-05, "loss": 0.0431, "step": 34440 }, { "grad_norm": 0.5725917220115662, "learning_rate": 6.535449981615375e-05, "loss": 0.0368, "step": 34450 }, { "grad_norm": 0.4508301019668579, "learning_rate": 6.53348288180301e-05, "loss": 0.037, "step": 34460 }, { "grad_norm": 0.4502609074115753, "learning_rate": 6.531515519960361e-05, "loss": 0.0453, "step": 34470 }, { "grad_norm": 0.5687374472618103, "learning_rate": 6.529547896423597e-05, "loss": 0.0456, "step": 34480 }, { "grad_norm": 0.45119979977607727, "learning_rate": 6.52758001152893e-05, "loss": 0.0749, "step": 34490 }, { "grad_norm": 0.555770218372345, "learning_rate": 6.525611865612618e-05, "loss": 0.0527, "step": 34500 }, { "grad_norm": 0.4815913140773773, "learning_rate": 6.523643459010966e-05, "loss": 0.0454, "step": 34510 }, { "grad_norm": 0.536443293094635, "learning_rate": 6.521674792060317e-05, "loss": 0.0451, "step": 34520 }, { "grad_norm": 0.5206589102745056, "learning_rate": 6.519705865097063e-05, "loss": 0.0481, "step": 34530 }, { "grad_norm": 0.42098328471183777, "learning_rate": 6.517736678457641e-05, "loss": 0.0356, "step": 34540 }, { "grad_norm": 0.5147231817245483, "learning_rate": 6.515767232478534e-05, "loss": 0.045, "step": 34550 }, { "grad_norm": 0.5158394575119019, "learning_rate": 6.51379752749626e-05, "loss": 0.0422, "step": 34560 }, { "grad_norm": 0.40769970417022705, "learning_rate": 6.511827563847393e-05, "loss": 0.0448, "step": 34570 }, { "grad_norm": 0.40663865208625793, "learning_rate": 6.509857341868542e-05, "loss": 0.0512, "step": 34580 }, { "grad_norm": 0.539756178855896, "learning_rate": 6.507886861896367e-05, "loss": 0.0419, "step": 34590 }, { "grad_norm": 0.5058627724647522, "learning_rate": 6.505916124267567e-05, "loss": 0.0371, "step": 34600 }, { "grad_norm": 0.9626567363739014, "learning_rate": 6.503945129318891e-05, "loss": 0.0441, "step": 34610 }, { "grad_norm": 0.50586998462677, "learning_rate": 6.501973877387122e-05, "loss": 0.0347, "step": 34620 }, { "grad_norm": 0.3532627820968628, "learning_rate": 6.500002368809098e-05, "loss": 0.0401, "step": 34630 }, { "grad_norm": 0.7666040062904358, "learning_rate": 6.498030603921694e-05, "loss": 0.0524, "step": 34640 }, { "grad_norm": 0.7188940048217773, "learning_rate": 6.496058583061832e-05, "loss": 0.0435, "step": 34650 }, { "grad_norm": 0.42058247327804565, "learning_rate": 6.494086306566475e-05, "loss": 0.0378, "step": 34660 }, { "grad_norm": 0.5455861687660217, "learning_rate": 6.492113774772632e-05, "loss": 0.0529, "step": 34670 }, { "grad_norm": 0.5128722786903381, "learning_rate": 6.490140988017354e-05, "loss": 0.046, "step": 34680 }, { "grad_norm": 0.6171111464500427, "learning_rate": 6.488167946637736e-05, "loss": 0.045, "step": 34690 }, { "grad_norm": 0.6762277483940125, "learning_rate": 6.486194650970915e-05, "loss": 0.0393, "step": 34700 }, { "grad_norm": 0.6963486671447754, "learning_rate": 6.48422110135408e-05, "loss": 0.0445, "step": 34710 }, { "grad_norm": 1.3440849781036377, "learning_rate": 6.482247298124451e-05, "loss": 0.0481, "step": 34720 }, { "grad_norm": 0.5808196663856506, "learning_rate": 6.480273241619297e-05, "loss": 0.0366, "step": 34730 }, { "grad_norm": 0.6330509781837463, "learning_rate": 6.478298932175933e-05, "loss": 0.0558, "step": 34740 }, { "grad_norm": 0.7198410034179688, "learning_rate": 6.476324370131712e-05, "loss": 0.0465, "step": 34750 }, { "grad_norm": 0.5026500821113586, "learning_rate": 6.474349555824036e-05, "loss": 0.0352, "step": 34760 }, { "grad_norm": 0.5994824171066284, "learning_rate": 6.472374489590342e-05, "loss": 0.0447, "step": 34770 }, { "grad_norm": 0.48617812991142273, "learning_rate": 6.470399171768118e-05, "loss": 0.0459, "step": 34780 }, { "grad_norm": 0.485044926404953, "learning_rate": 6.468423602694891e-05, "loss": 0.0353, "step": 34790 }, { "grad_norm": 0.49328094720840454, "learning_rate": 6.466447782708232e-05, "loss": 0.047, "step": 34800 }, { "grad_norm": 0.4245384633541107, "learning_rate": 6.464471712145754e-05, "loss": 0.0442, "step": 34810 }, { "grad_norm": 0.3368949294090271, "learning_rate": 6.462495391345114e-05, "loss": 0.0368, "step": 34820 }, { "grad_norm": 0.2650972902774811, "learning_rate": 6.46051882064401e-05, "loss": 0.0336, "step": 34830 }, { "grad_norm": 0.6985117197036743, "learning_rate": 6.458542000380186e-05, "loss": 0.0511, "step": 34840 }, { "grad_norm": 0.39008453488349915, "learning_rate": 6.456564930891424e-05, "loss": 0.037, "step": 34850 }, { "grad_norm": 0.5612831115722656, "learning_rate": 6.454587612515555e-05, "loss": 0.0416, "step": 34860 }, { "grad_norm": 0.41617193818092346, "learning_rate": 6.452610045590444e-05, "loss": 0.0312, "step": 34870 }, { "grad_norm": 0.508887529373169, "learning_rate": 6.450632230454005e-05, "loss": 0.0607, "step": 34880 }, { "grad_norm": 0.6312171816825867, "learning_rate": 6.448654167444195e-05, "loss": 0.0428, "step": 34890 }, { "grad_norm": 0.711881160736084, "learning_rate": 6.446675856899005e-05, "loss": 0.0618, "step": 34900 }, { "grad_norm": 0.5419835448265076, "learning_rate": 6.444697299156481e-05, "loss": 0.0393, "step": 34910 }, { "grad_norm": 0.35306641459465027, "learning_rate": 6.442718494554701e-05, "loss": 0.04, "step": 34920 }, { "grad_norm": 0.2777009606361389, "learning_rate": 6.440739443431787e-05, "loss": 0.0421, "step": 34930 }, { "grad_norm": 0.7006553411483765, "learning_rate": 6.438760146125906e-05, "loss": 0.036, "step": 34940 }, { "grad_norm": 0.557004988193512, "learning_rate": 6.436780602975267e-05, "loss": 0.0518, "step": 34950 }, { "grad_norm": 0.4230875074863434, "learning_rate": 6.43480081431812e-05, "loss": 0.048, "step": 34960 }, { "grad_norm": 0.7667468786239624, "learning_rate": 6.432820780492756e-05, "loss": 0.044, "step": 34970 }, { "grad_norm": 0.5144348740577698, "learning_rate": 6.430840501837506e-05, "loss": 0.0302, "step": 34980 }, { "grad_norm": 0.37813547253608704, "learning_rate": 6.428859978690748e-05, "loss": 0.0405, "step": 34990 }, { "grad_norm": 0.6108113527297974, "learning_rate": 6.426879211390901e-05, "loss": 0.0381, "step": 35000 }, { "grad_norm": 0.5803585648536682, "learning_rate": 6.424898200276422e-05, "loss": 0.0417, "step": 35010 }, { "grad_norm": 0.55318284034729, "learning_rate": 6.42291694568581e-05, "loss": 0.0473, "step": 35020 }, { "grad_norm": 0.6846518516540527, "learning_rate": 6.42093544795761e-05, "loss": 0.043, "step": 35030 }, { "grad_norm": 0.7233472466468811, "learning_rate": 6.418953707430403e-05, "loss": 0.0526, "step": 35040 }, { "grad_norm": 0.36903998255729675, "learning_rate": 6.416971724442819e-05, "loss": 0.0429, "step": 35050 }, { "grad_norm": 0.42574796080589294, "learning_rate": 6.414989499333519e-05, "loss": 0.0351, "step": 35060 }, { "grad_norm": 1.1712228059768677, "learning_rate": 6.413007032441214e-05, "loss": 0.0657, "step": 35070 }, { "grad_norm": 0.40219494700431824, "learning_rate": 6.411024324104653e-05, "loss": 0.038, "step": 35080 }, { "grad_norm": 0.4304877519607544, "learning_rate": 6.409041374662628e-05, "loss": 0.0319, "step": 35090 }, { "grad_norm": 0.31984278559684753, "learning_rate": 6.407058184453967e-05, "loss": 0.0378, "step": 35100 }, { "grad_norm": 0.47279489040374756, "learning_rate": 6.405074753817548e-05, "loss": 0.0561, "step": 35110 }, { "grad_norm": 0.7277394533157349, "learning_rate": 6.40309108309228e-05, "loss": 0.0606, "step": 35120 }, { "grad_norm": 0.39609450101852417, "learning_rate": 6.401107172617122e-05, "loss": 0.0439, "step": 35130 }, { "grad_norm": 0.7583276033401489, "learning_rate": 6.399123022731068e-05, "loss": 0.0584, "step": 35140 }, { "grad_norm": 0.5364418625831604, "learning_rate": 6.397138633773157e-05, "loss": 0.0419, "step": 35150 }, { "grad_norm": 0.43263375759124756, "learning_rate": 6.395154006082463e-05, "loss": 0.0373, "step": 35160 }, { "grad_norm": 0.6521447896957397, "learning_rate": 6.393169139998109e-05, "loss": 0.049, "step": 35170 }, { "grad_norm": 0.46769219636917114, "learning_rate": 6.39118403585925e-05, "loss": 0.0443, "step": 35180 }, { "grad_norm": 0.4257422685623169, "learning_rate": 6.38919869400509e-05, "loss": 0.0358, "step": 35190 }, { "grad_norm": 0.38661640882492065, "learning_rate": 6.387213114774865e-05, "loss": 0.0674, "step": 35200 }, { "grad_norm": 0.6508235931396484, "learning_rate": 6.385227298507863e-05, "loss": 0.0613, "step": 35210 }, { "grad_norm": 0.8158608078956604, "learning_rate": 6.3832412455434e-05, "loss": 0.0491, "step": 35220 }, { "grad_norm": 0.4282073378562927, "learning_rate": 6.381254956220841e-05, "loss": 0.0406, "step": 35230 }, { "grad_norm": 0.7532023787498474, "learning_rate": 6.379268430879586e-05, "loss": 0.0466, "step": 35240 }, { "grad_norm": 0.6728110313415527, "learning_rate": 6.37728166985908e-05, "loss": 0.0465, "step": 35250 }, { "grad_norm": 0.4360654056072235, "learning_rate": 6.375294673498804e-05, "loss": 0.0398, "step": 35260 }, { "grad_norm": 0.6170092225074768, "learning_rate": 6.373307442138284e-05, "loss": 0.0499, "step": 35270 }, { "grad_norm": 0.5670046806335449, "learning_rate": 6.371319976117081e-05, "loss": 0.0448, "step": 35280 }, { "grad_norm": 0.5151122808456421, "learning_rate": 6.3693322757748e-05, "loss": 0.045, "step": 35290 }, { "grad_norm": 0.6739092469215393, "learning_rate": 6.367344341451086e-05, "loss": 0.0385, "step": 35300 }, { "grad_norm": 0.3971610367298126, "learning_rate": 6.36535617348562e-05, "loss": 0.0339, "step": 35310 }, { "grad_norm": 0.37140193581581116, "learning_rate": 6.363367772218128e-05, "loss": 0.0596, "step": 35320 }, { "grad_norm": 0.41243207454681396, "learning_rate": 6.36137913798837e-05, "loss": 0.0368, "step": 35330 }, { "grad_norm": 0.49662765860557556, "learning_rate": 6.359390271136151e-05, "loss": 0.0513, "step": 35340 }, { "grad_norm": 0.7939421534538269, "learning_rate": 6.357401172001314e-05, "loss": 0.0491, "step": 35350 }, { "grad_norm": 0.6902651786804199, "learning_rate": 6.355411840923742e-05, "loss": 0.047, "step": 35360 }, { "grad_norm": 0.3263547122478485, "learning_rate": 6.353422278243358e-05, "loss": 0.0473, "step": 35370 }, { "grad_norm": 0.47542518377304077, "learning_rate": 6.351432484300121e-05, "loss": 0.0362, "step": 35380 }, { "grad_norm": 0.59917813539505, "learning_rate": 6.349442459434036e-05, "loss": 0.0565, "step": 35390 }, { "grad_norm": 0.45552870631217957, "learning_rate": 6.34745220398514e-05, "loss": 0.0325, "step": 35400 }, { "grad_norm": 0.6151089668273926, "learning_rate": 6.345461718293518e-05, "loss": 0.0412, "step": 35410 }, { "grad_norm": 0.5125537514686584, "learning_rate": 6.343471002699286e-05, "loss": 0.0441, "step": 35420 }, { "grad_norm": 0.5112981200218201, "learning_rate": 6.341480057542602e-05, "loss": 0.0366, "step": 35430 }, { "grad_norm": 0.5892192125320435, "learning_rate": 6.339488883163667e-05, "loss": 0.0549, "step": 35440 }, { "grad_norm": 0.5870867967605591, "learning_rate": 6.337497479902716e-05, "loss": 0.0372, "step": 35450 }, { "grad_norm": 0.47580140829086304, "learning_rate": 6.335505848100027e-05, "loss": 0.0358, "step": 35460 }, { "grad_norm": 0.5487498641014099, "learning_rate": 6.333513988095915e-05, "loss": 0.0421, "step": 35470 }, { "grad_norm": 0.382322758436203, "learning_rate": 6.331521900230735e-05, "loss": 0.0598, "step": 35480 }, { "grad_norm": 0.7201549410820007, "learning_rate": 6.329529584844878e-05, "loss": 0.0427, "step": 35490 }, { "grad_norm": 0.555072546005249, "learning_rate": 6.327537042278777e-05, "loss": 0.0705, "step": 35500 }, { "grad_norm": 0.5006661415100098, "learning_rate": 6.325544272872905e-05, "loss": 0.0429, "step": 35510 }, { "grad_norm": 0.45348846912384033, "learning_rate": 6.323551276967771e-05, "loss": 0.0584, "step": 35520 }, { "grad_norm": 0.6904719471931458, "learning_rate": 6.321558054903922e-05, "loss": 0.0406, "step": 35530 }, { "grad_norm": 0.42497628927230835, "learning_rate": 6.319564607021947e-05, "loss": 0.0347, "step": 35540 }, { "grad_norm": 0.524264931678772, "learning_rate": 6.31757093366247e-05, "loss": 0.0361, "step": 35550 }, { "grad_norm": 0.3301681876182556, "learning_rate": 6.315577035166154e-05, "loss": 0.0598, "step": 35560 }, { "grad_norm": 0.42894795536994934, "learning_rate": 6.313582911873708e-05, "loss": 0.0277, "step": 35570 }, { "grad_norm": 0.7122271060943604, "learning_rate": 6.311588564125865e-05, "loss": 0.0375, "step": 35580 }, { "grad_norm": 0.6836193203926086, "learning_rate": 6.30959399226341e-05, "loss": 0.0403, "step": 35590 }, { "grad_norm": 0.5783836841583252, "learning_rate": 6.30759919662716e-05, "loss": 0.0352, "step": 35600 }, { "grad_norm": 0.772138774394989, "learning_rate": 6.30560417755797e-05, "loss": 0.0382, "step": 35610 }, { "grad_norm": 0.4684266746044159, "learning_rate": 6.303608935396735e-05, "loss": 0.0499, "step": 35620 }, { "grad_norm": 0.769638180732727, "learning_rate": 6.301613470484386e-05, "loss": 0.0467, "step": 35630 }, { "grad_norm": 0.6350790858268738, "learning_rate": 6.299617783161893e-05, "loss": 0.0586, "step": 35640 }, { "grad_norm": 0.3720615804195404, "learning_rate": 6.297621873770266e-05, "loss": 0.0357, "step": 35650 }, { "grad_norm": 0.5884567499160767, "learning_rate": 6.29562574265055e-05, "loss": 0.053, "step": 35660 }, { "grad_norm": 0.4479207396507263, "learning_rate": 6.293629390143834e-05, "loss": 0.0373, "step": 35670 }, { "grad_norm": 0.6235770583152771, "learning_rate": 6.291632816591232e-05, "loss": 0.0505, "step": 35680 }, { "grad_norm": 0.4418685734272003, "learning_rate": 6.28963602233391e-05, "loss": 0.0385, "step": 35690 }, { "grad_norm": 0.4573710560798645, "learning_rate": 6.287639007713062e-05, "loss": 0.0525, "step": 35700 }, { "grad_norm": 0.822010338306427, "learning_rate": 6.285641773069926e-05, "loss": 0.0405, "step": 35710 }, { "grad_norm": 0.445413738489151, "learning_rate": 6.283644318745773e-05, "loss": 0.0382, "step": 35720 }, { "grad_norm": 0.6202298402786255, "learning_rate": 6.281646645081912e-05, "loss": 0.0385, "step": 35730 }, { "grad_norm": 0.5733619332313538, "learning_rate": 6.279648752419693e-05, "loss": 0.0454, "step": 35740 }, { "grad_norm": 0.358778715133667, "learning_rate": 6.2776506411005e-05, "loss": 0.0365, "step": 35750 }, { "grad_norm": 0.776419997215271, "learning_rate": 6.275652311465758e-05, "loss": 0.047, "step": 35760 }, { "grad_norm": 0.4275176525115967, "learning_rate": 6.273653763856926e-05, "loss": 0.0391, "step": 35770 }, { "grad_norm": 0.3316880762577057, "learning_rate": 6.271654998615501e-05, "loss": 0.0427, "step": 35780 }, { "grad_norm": 1.1218600273132324, "learning_rate": 6.269656016083013e-05, "loss": 0.0365, "step": 35790 }, { "grad_norm": 0.5879308581352234, "learning_rate": 6.267656816601038e-05, "loss": 0.0374, "step": 35800 }, { "grad_norm": 0.37412384152412415, "learning_rate": 6.265657400511185e-05, "loss": 0.03, "step": 35810 }, { "grad_norm": 0.5889235734939575, "learning_rate": 6.263657768155098e-05, "loss": 0.044, "step": 35820 }, { "grad_norm": 0.5195313692092896, "learning_rate": 6.261657919874457e-05, "loss": 0.0394, "step": 35830 }, { "grad_norm": 0.5757574439048767, "learning_rate": 6.259657856010986e-05, "loss": 0.0337, "step": 35840 }, { "grad_norm": 0.46971628069877625, "learning_rate": 6.257657576906439e-05, "loss": 0.0494, "step": 35850 }, { "grad_norm": 0.31778210401535034, "learning_rate": 6.255657082902609e-05, "loss": 0.0324, "step": 35860 }, { "grad_norm": 0.35150429606437683, "learning_rate": 6.253656374341325e-05, "loss": 0.0403, "step": 35870 }, { "grad_norm": 0.5700278878211975, "learning_rate": 6.251655451564457e-05, "loss": 0.0433, "step": 35880 }, { "grad_norm": 0.6927329301834106, "learning_rate": 6.249654314913902e-05, "loss": 0.0619, "step": 35890 }, { "grad_norm": 0.3713643550872803, "learning_rate": 6.247652964731604e-05, "loss": 0.0305, "step": 35900 }, { "grad_norm": 0.6155501008033752, "learning_rate": 6.245651401359537e-05, "loss": 0.0453, "step": 35910 }, { "grad_norm": 0.40192583203315735, "learning_rate": 6.243649625139715e-05, "loss": 0.0449, "step": 35920 }, { "grad_norm": 0.6690454483032227, "learning_rate": 6.241647636414185e-05, "loss": 0.0415, "step": 35930 }, { "grad_norm": 0.3027917742729187, "learning_rate": 6.239645435525034e-05, "loss": 0.0336, "step": 35940 }, { "grad_norm": 0.6140342950820923, "learning_rate": 6.237643022814381e-05, "loss": 0.062, "step": 35950 }, { "grad_norm": 0.5857582092285156, "learning_rate": 6.235640398624386e-05, "loss": 0.0481, "step": 35960 }, { "grad_norm": 0.5601363182067871, "learning_rate": 6.233637563297243e-05, "loss": 0.0399, "step": 35970 }, { "grad_norm": 0.4785367250442505, "learning_rate": 6.23163451717518e-05, "loss": 0.0455, "step": 35980 }, { "grad_norm": 0.5225421786308289, "learning_rate": 6.229631260600463e-05, "loss": 0.0589, "step": 35990 }, { "grad_norm": 0.7270310521125793, "learning_rate": 6.227627793915392e-05, "loss": 0.0337, "step": 36000 }, { "grad_norm": 0.44001564383506775, "learning_rate": 6.225624117462309e-05, "loss": 0.0434, "step": 36010 }, { "grad_norm": 0.5616369247436523, "learning_rate": 6.223620231583586e-05, "loss": 0.0865, "step": 36020 }, { "grad_norm": 0.46859440207481384, "learning_rate": 6.221616136621629e-05, "loss": 0.0483, "step": 36030 }, { "grad_norm": 0.31089910864830017, "learning_rate": 6.219611832918887e-05, "loss": 0.0366, "step": 36040 }, { "grad_norm": 0.5750100612640381, "learning_rate": 6.217607320817838e-05, "loss": 0.0565, "step": 36050 }, { "grad_norm": 0.4733801782131195, "learning_rate": 6.215602600661001e-05, "loss": 0.0353, "step": 36060 }, { "grad_norm": 0.5138092041015625, "learning_rate": 6.213597672790925e-05, "loss": 0.0394, "step": 36070 }, { "grad_norm": 0.6167211532592773, "learning_rate": 6.2115925375502e-05, "loss": 0.0407, "step": 36080 }, { "grad_norm": 0.6091152429580688, "learning_rate": 6.209587195281447e-05, "loss": 0.0486, "step": 36090 }, { "grad_norm": 0.46796324849128723, "learning_rate": 6.207581646327324e-05, "loss": 0.0368, "step": 36100 }, { "grad_norm": 0.438907653093338, "learning_rate": 6.205575891030526e-05, "loss": 0.0416, "step": 36110 }, { "grad_norm": 0.5153229236602783, "learning_rate": 6.203569929733781e-05, "loss": 0.0415, "step": 36120 }, { "grad_norm": 0.49655789136886597, "learning_rate": 6.201563762779852e-05, "loss": 0.0562, "step": 36130 }, { "grad_norm": 0.5652534365653992, "learning_rate": 6.199557390511538e-05, "loss": 0.0325, "step": 36140 }, { "grad_norm": 0.6672472357749939, "learning_rate": 6.197550813271675e-05, "loss": 0.0405, "step": 36150 }, { "grad_norm": 0.38941478729248047, "learning_rate": 6.195544031403131e-05, "loss": 0.0344, "step": 36160 }, { "grad_norm": 0.7124948501586914, "learning_rate": 6.19353704524881e-05, "loss": 0.0566, "step": 36170 }, { "grad_norm": 0.51595538854599, "learning_rate": 6.191529855151652e-05, "loss": 0.0366, "step": 36180 }, { "grad_norm": 0.4969753921031952, "learning_rate": 6.189522461454629e-05, "loss": 0.0421, "step": 36190 }, { "grad_norm": 0.4962790310382843, "learning_rate": 6.187514864500752e-05, "loss": 0.0546, "step": 36200 }, { "grad_norm": 0.3354364335536957, "learning_rate": 6.185507064633062e-05, "loss": 0.0399, "step": 36210 }, { "grad_norm": 0.7621604204177856, "learning_rate": 6.18349906219464e-05, "loss": 0.0451, "step": 36220 }, { "grad_norm": 0.4827462434768677, "learning_rate": 6.181490857528596e-05, "loss": 0.0532, "step": 36230 }, { "grad_norm": 0.6218323111534119, "learning_rate": 6.179482450978077e-05, "loss": 0.0417, "step": 36240 }, { "grad_norm": 0.43094533681869507, "learning_rate": 6.177473842886269e-05, "loss": 0.0401, "step": 36250 }, { "grad_norm": 0.7127171158790588, "learning_rate": 6.175465033596382e-05, "loss": 0.0449, "step": 36260 }, { "grad_norm": 0.4090985059738159, "learning_rate": 6.173456023451671e-05, "loss": 0.0467, "step": 36270 }, { "grad_norm": 2.0969650745391846, "learning_rate": 6.171446812795422e-05, "loss": 0.0463, "step": 36280 }, { "grad_norm": 0.4808281660079956, "learning_rate": 6.169437401970949e-05, "loss": 0.0377, "step": 36290 }, { "grad_norm": 0.6028499007225037, "learning_rate": 6.16742779132161e-05, "loss": 0.0363, "step": 36300 }, { "grad_norm": 0.7265965342521667, "learning_rate": 6.165417981190789e-05, "loss": 0.0425, "step": 36310 }, { "grad_norm": 0.4536028802394867, "learning_rate": 6.16340797192191e-05, "loss": 0.037, "step": 36320 }, { "grad_norm": 0.5703835487365723, "learning_rate": 6.161397763858427e-05, "loss": 0.0418, "step": 36330 }, { "grad_norm": 0.7802620530128479, "learning_rate": 6.159387357343834e-05, "loss": 0.0441, "step": 36340 }, { "grad_norm": 0.7764385342597961, "learning_rate": 6.157376752721648e-05, "loss": 0.0383, "step": 36350 }, { "grad_norm": 0.8605735898017883, "learning_rate": 6.155365950335428e-05, "loss": 0.0381, "step": 36360 }, { "grad_norm": 0.46584075689315796, "learning_rate": 6.153354950528768e-05, "loss": 0.04, "step": 36370 }, { "grad_norm": 0.2853653132915497, "learning_rate": 6.151343753645293e-05, "loss": 0.0434, "step": 36380 }, { "grad_norm": 0.40615159273147583, "learning_rate": 6.149332360028657e-05, "loss": 0.0365, "step": 36390 }, { "grad_norm": 0.4255734384059906, "learning_rate": 6.147320770022555e-05, "loss": 0.0306, "step": 36400 }, { "grad_norm": 0.44314178824424744, "learning_rate": 6.145308983970715e-05, "loss": 0.0449, "step": 36410 }, { "grad_norm": 0.42352059483528137, "learning_rate": 6.143297002216892e-05, "loss": 0.0259, "step": 36420 }, { "grad_norm": 0.4995476305484772, "learning_rate": 6.141284825104882e-05, "loss": 0.0368, "step": 36430 }, { "grad_norm": 0.694641649723053, "learning_rate": 6.13927245297851e-05, "loss": 0.0441, "step": 36440 }, { "grad_norm": 0.48472708463668823, "learning_rate": 6.137259886181633e-05, "loss": 0.034, "step": 36450 }, { "grad_norm": 0.4721730351448059, "learning_rate": 6.135247125058145e-05, "loss": 0.0369, "step": 36460 }, { "grad_norm": 0.5337848663330078, "learning_rate": 6.133234169951974e-05, "loss": 0.0393, "step": 36470 }, { "grad_norm": 0.6607767343521118, "learning_rate": 6.131221021207078e-05, "loss": 0.0334, "step": 36480 }, { "grad_norm": 0.47055667638778687, "learning_rate": 6.129207679167448e-05, "loss": 0.0518, "step": 36490 }, { "grad_norm": 0.9894503355026245, "learning_rate": 6.127194144177109e-05, "loss": 0.048, "step": 36500 }, { "grad_norm": 0.3491070866584778, "learning_rate": 6.125180416580118e-05, "loss": 0.0349, "step": 36510 }, { "grad_norm": 0.641945481300354, "learning_rate": 6.123166496720571e-05, "loss": 0.0485, "step": 36520 }, { "grad_norm": 0.4093146324157715, "learning_rate": 6.121152384942588e-05, "loss": 0.0419, "step": 36530 }, { "grad_norm": 0.5529224872589111, "learning_rate": 6.119138081590324e-05, "loss": 0.0349, "step": 36540 }, { "grad_norm": 0.4619150459766388, "learning_rate": 6.117123587007971e-05, "loss": 0.0334, "step": 36550 }, { "grad_norm": 0.5562582015991211, "learning_rate": 6.11510890153975e-05, "loss": 0.0314, "step": 36560 }, { "grad_norm": 0.6239798069000244, "learning_rate": 6.113094025529916e-05, "loss": 0.0373, "step": 36570 }, { "grad_norm": 0.7258429527282715, "learning_rate": 6.111078959322757e-05, "loss": 0.0413, "step": 36580 }, { "grad_norm": 0.4288107752799988, "learning_rate": 6.109063703262592e-05, "loss": 0.0325, "step": 36590 }, { "grad_norm": 0.594370424747467, "learning_rate": 6.107048257693772e-05, "loss": 0.0399, "step": 36600 }, { "grad_norm": 0.3974175751209259, "learning_rate": 6.105032622960683e-05, "loss": 0.0351, "step": 36610 }, { "grad_norm": 0.7432864904403687, "learning_rate": 6.103016799407743e-05, "loss": 0.0469, "step": 36620 }, { "grad_norm": 0.8531548380851746, "learning_rate": 6.1010007873793984e-05, "loss": 0.0406, "step": 36630 }, { "grad_norm": 0.4864896833896637, "learning_rate": 6.098984587220131e-05, "loss": 0.0484, "step": 36640 }, { "grad_norm": 0.2939881384372711, "learning_rate": 6.096968199274456e-05, "loss": 0.0398, "step": 36650 }, { "grad_norm": 0.6392294764518738, "learning_rate": 6.0949516238869166e-05, "loss": 0.0365, "step": 36660 }, { "grad_norm": 0.6384201049804688, "learning_rate": 6.092934861402092e-05, "loss": 0.042, "step": 36670 }, { "grad_norm": 0.5671483874320984, "learning_rate": 6.0909179121645924e-05, "loss": 0.0362, "step": 36680 }, { "grad_norm": 0.6531751155853271, "learning_rate": 6.0889007765190576e-05, "loss": 0.0373, "step": 36690 }, { "grad_norm": 0.47739407420158386, "learning_rate": 6.086883454810162e-05, "loss": 0.037, "step": 36700 }, { "grad_norm": 0.6382926106452942, "learning_rate": 6.0848659473826084e-05, "loss": 0.0348, "step": 36710 }, { "grad_norm": 0.6425788998603821, "learning_rate": 6.082848254581138e-05, "loss": 0.045, "step": 36720 }, { "grad_norm": 0.5296198129653931, "learning_rate": 6.080830376750517e-05, "loss": 0.0335, "step": 36730 }, { "grad_norm": 0.6566953063011169, "learning_rate": 6.0788123142355445e-05, "loss": 0.0293, "step": 36740 }, { "grad_norm": 0.425324022769928, "learning_rate": 6.076794067381052e-05, "loss": 0.033, "step": 36750 }, { "grad_norm": 0.6389082670211792, "learning_rate": 6.074775636531905e-05, "loss": 0.0395, "step": 36760 }, { "grad_norm": 0.4928758442401886, "learning_rate": 6.072757022032997e-05, "loss": 0.0371, "step": 36770 }, { "grad_norm": 0.43515244126319885, "learning_rate": 6.070738224229253e-05, "loss": 0.0336, "step": 36780 }, { "grad_norm": 0.535261332988739, "learning_rate": 6.0687192434656314e-05, "loss": 0.0423, "step": 36790 }, { "grad_norm": 0.39148426055908203, "learning_rate": 6.066700080087121e-05, "loss": 0.0536, "step": 36800 }, { "grad_norm": 0.5822014808654785, "learning_rate": 6.0646807344387424e-05, "loss": 0.0499, "step": 36810 }, { "grad_norm": 0.5369954705238342, "learning_rate": 6.062661206865543e-05, "loss": 0.0468, "step": 36820 }, { "grad_norm": 0.34286585450172424, "learning_rate": 6.06064149771261e-05, "loss": 0.0394, "step": 36830 }, { "grad_norm": 0.41808491945266724, "learning_rate": 6.058621607325051e-05, "loss": 0.0515, "step": 36840 }, { "grad_norm": 0.3248003423213959, "learning_rate": 6.056601536048014e-05, "loss": 0.0397, "step": 36850 }, { "grad_norm": 0.4262832999229431, "learning_rate": 6.0545812842266725e-05, "loss": 0.0327, "step": 36860 }, { "grad_norm": 0.5371469259262085, "learning_rate": 6.052560852206232e-05, "loss": 0.0311, "step": 36870 }, { "grad_norm": 0.5407285094261169, "learning_rate": 6.05054024033193e-05, "loss": 0.0263, "step": 36880 }, { "grad_norm": 0.4173141419887543, "learning_rate": 6.048519448949032e-05, "loss": 0.0416, "step": 36890 }, { "grad_norm": 0.4303615987300873, "learning_rate": 6.046498478402839e-05, "loss": 0.0573, "step": 36900 }, { "grad_norm": 0.5935033559799194, "learning_rate": 6.044477329038677e-05, "loss": 0.0441, "step": 36910 }, { "grad_norm": 0.6323651075363159, "learning_rate": 6.042456001201906e-05, "loss": 0.051, "step": 36920 }, { "grad_norm": 0.36802947521209717, "learning_rate": 6.040434495237917e-05, "loss": 0.0415, "step": 36930 }, { "grad_norm": 0.5063944458961487, "learning_rate": 6.0384128114921256e-05, "loss": 0.0557, "step": 36940 }, { "grad_norm": 0.4609330892562866, "learning_rate": 6.036390950309987e-05, "loss": 0.0417, "step": 36950 }, { "grad_norm": 0.47407156229019165, "learning_rate": 6.0343689120369805e-05, "loss": 0.0362, "step": 36960 }, { "grad_norm": 0.5366179347038269, "learning_rate": 6.032346697018616e-05, "loss": 0.0436, "step": 36970 }, { "grad_norm": 0.4142376780509949, "learning_rate": 6.0303243056004375e-05, "loss": 0.036, "step": 36980 }, { "grad_norm": 0.29362428188323975, "learning_rate": 6.0283017381280136e-05, "loss": 0.0625, "step": 36990 }, { "grad_norm": 0.6010977029800415, "learning_rate": 6.026278994946948e-05, "loss": 0.0389, "step": 37000 }, { "grad_norm": 0.5191525816917419, "learning_rate": 6.02425607640287e-05, "loss": 0.0455, "step": 37010 }, { "grad_norm": 0.6166524291038513, "learning_rate": 6.022232982841441e-05, "loss": 0.0401, "step": 37020 }, { "grad_norm": 0.529723048210144, "learning_rate": 6.020209714608355e-05, "loss": 0.0434, "step": 37030 }, { "grad_norm": 0.6298866868019104, "learning_rate": 6.018186272049331e-05, "loss": 0.0413, "step": 37040 }, { "grad_norm": 0.3972320854663849, "learning_rate": 6.01616265551012e-05, "loss": 0.0415, "step": 37050 }, { "grad_norm": 0.7665921449661255, "learning_rate": 6.014138865336503e-05, "loss": 0.0317, "step": 37060 }, { "grad_norm": 0.38375306129455566, "learning_rate": 6.0121149018742905e-05, "loss": 0.0324, "step": 37070 }, { "grad_norm": 0.298441082239151, "learning_rate": 6.010090765469325e-05, "loss": 0.0306, "step": 37080 }, { "grad_norm": 0.31577539443969727, "learning_rate": 6.008066456467473e-05, "loss": 0.0368, "step": 37090 }, { "grad_norm": 0.360240638256073, "learning_rate": 6.0060419752146335e-05, "loss": 0.0297, "step": 37100 }, { "grad_norm": 0.4160171151161194, "learning_rate": 6.0040173220567353e-05, "loss": 0.0422, "step": 37110 }, { "grad_norm": 0.42382848262786865, "learning_rate": 6.001992497339737e-05, "loss": 0.0643, "step": 37120 }, { "grad_norm": 0.5220874547958374, "learning_rate": 5.999967501409626e-05, "loss": 0.0325, "step": 37130 }, { "grad_norm": 0.5654336214065552, "learning_rate": 5.997942334612418e-05, "loss": 0.0482, "step": 37140 }, { "grad_norm": 0.41280612349510193, "learning_rate": 5.995916997294158e-05, "loss": 0.0438, "step": 37150 }, { "grad_norm": 0.6388242244720459, "learning_rate": 5.9938914898009214e-05, "loss": 0.0499, "step": 37160 }, { "grad_norm": 0.48248279094696045, "learning_rate": 5.991865812478813e-05, "loss": 0.0433, "step": 37170 }, { "grad_norm": 0.6360825300216675, "learning_rate": 5.989839965673964e-05, "loss": 0.0517, "step": 37180 }, { "grad_norm": 0.9023285508155823, "learning_rate": 5.987813949732539e-05, "loss": 0.05, "step": 37190 }, { "grad_norm": 0.5879956483840942, "learning_rate": 5.9857877650007255e-05, "loss": 0.0585, "step": 37200 }, { "grad_norm": 0.5296633839607239, "learning_rate": 5.983761411824744e-05, "loss": 0.0432, "step": 37210 }, { "grad_norm": 0.41472724080085754, "learning_rate": 5.981734890550844e-05, "loss": 0.0302, "step": 37220 }, { "grad_norm": 0.3828642964363098, "learning_rate": 5.979708201525301e-05, "loss": 0.0352, "step": 37230 }, { "grad_norm": 0.888331413269043, "learning_rate": 5.977681345094422e-05, "loss": 0.0601, "step": 37240 }, { "grad_norm": 0.5767579078674316, "learning_rate": 5.97565432160454e-05, "loss": 0.056, "step": 37250 }, { "grad_norm": 0.494056761264801, "learning_rate": 5.9736271314020186e-05, "loss": 0.0557, "step": 37260 }, { "grad_norm": 0.35792478919029236, "learning_rate": 5.971599774833251e-05, "loss": 0.0356, "step": 37270 }, { "grad_norm": 0.3091747760772705, "learning_rate": 5.9695722522446525e-05, "loss": 0.0433, "step": 37280 }, { "grad_norm": 0.4699556827545166, "learning_rate": 5.9675445639826765e-05, "loss": 0.042, "step": 37290 }, { "grad_norm": 0.5661477446556091, "learning_rate": 5.965516710393796e-05, "loss": 0.0367, "step": 37300 }, { "grad_norm": 0.6587666869163513, "learning_rate": 5.963488691824516e-05, "loss": 0.0367, "step": 37310 }, { "grad_norm": 0.41477224230766296, "learning_rate": 5.96146050862137e-05, "loss": 0.0439, "step": 37320 }, { "grad_norm": 0.5405175089836121, "learning_rate": 5.959432161130919e-05, "loss": 0.0501, "step": 37330 }, { "grad_norm": 0.49933087825775146, "learning_rate": 5.9574036496997545e-05, "loss": 0.0419, "step": 37340 }, { "grad_norm": 0.4039381146430969, "learning_rate": 5.955374974674488e-05, "loss": 0.0329, "step": 37350 }, { "grad_norm": 0.38158994913101196, "learning_rate": 5.9533461364017696e-05, "loss": 0.0444, "step": 37360 }, { "grad_norm": 0.38293755054473877, "learning_rate": 5.9513171352282716e-05, "loss": 0.0407, "step": 37370 }, { "grad_norm": 0.5359558463096619, "learning_rate": 5.949287971500692e-05, "loss": 0.0311, "step": 37380 }, { "grad_norm": 1.0833985805511475, "learning_rate": 5.947258645565762e-05, "loss": 0.0345, "step": 37390 }, { "grad_norm": 0.5703415274620056, "learning_rate": 5.945229157770237e-05, "loss": 0.0389, "step": 37400 }, { "grad_norm": 0.6317129135131836, "learning_rate": 5.9431995084609006e-05, "loss": 0.0342, "step": 37410 }, { "grad_norm": 0.30803239345550537, "learning_rate": 5.941169697984564e-05, "loss": 0.0325, "step": 37420 }, { "grad_norm": 0.5370569229125977, "learning_rate": 5.9391397266880675e-05, "loss": 0.0349, "step": 37430 }, { "grad_norm": 0.5153883695602417, "learning_rate": 5.937109594918279e-05, "loss": 0.0299, "step": 37440 }, { "grad_norm": 0.5045658946037292, "learning_rate": 5.9350793030220884e-05, "loss": 0.0712, "step": 37450 }, { "grad_norm": 0.2542881965637207, "learning_rate": 5.933048851346421e-05, "loss": 0.0597, "step": 37460 }, { "grad_norm": 0.4220263361930847, "learning_rate": 5.931018240238224e-05, "loss": 0.0333, "step": 37470 }, { "grad_norm": 0.5718974471092224, "learning_rate": 5.928987470044471e-05, "loss": 0.0348, "step": 37480 }, { "grad_norm": 0.5190686583518982, "learning_rate": 5.9269565411121695e-05, "loss": 0.0382, "step": 37490 }, { "grad_norm": 0.6513361930847168, "learning_rate": 5.924925453788347e-05, "loss": 0.0353, "step": 37500 }, { "grad_norm": 0.48873457312583923, "learning_rate": 5.92289420842006e-05, "loss": 0.032, "step": 37510 }, { "grad_norm": 0.31036704778671265, "learning_rate": 5.9208628053543945e-05, "loss": 0.0417, "step": 37520 }, { "grad_norm": 0.6033796668052673, "learning_rate": 5.918831244938462e-05, "loss": 0.0546, "step": 37530 }, { "grad_norm": 0.40718382596969604, "learning_rate": 5.9167995275194e-05, "loss": 0.0569, "step": 37540 }, { "grad_norm": 0.6641586422920227, "learning_rate": 5.914767653444373e-05, "loss": 0.0397, "step": 37550 }, { "grad_norm": 0.23521243035793304, "learning_rate": 5.912735623060572e-05, "loss": 0.0606, "step": 37560 }, { "grad_norm": 0.3366628885269165, "learning_rate": 5.910703436715217e-05, "loss": 0.0325, "step": 37570 }, { "grad_norm": 0.46491238474845886, "learning_rate": 5.908671094755552e-05, "loss": 0.0403, "step": 37580 }, { "grad_norm": 0.6394190192222595, "learning_rate": 5.906638597528851e-05, "loss": 0.0497, "step": 37590 }, { "grad_norm": 0.8602181673049927, "learning_rate": 5.9046059453824076e-05, "loss": 0.0417, "step": 37600 }, { "grad_norm": 0.4102073311805725, "learning_rate": 5.9025731386635505e-05, "loss": 0.0461, "step": 37610 }, { "grad_norm": 0.4427904188632965, "learning_rate": 5.900540177719629e-05, "loss": 0.0288, "step": 37620 }, { "grad_norm": 0.6070585250854492, "learning_rate": 5.898507062898021e-05, "loss": 0.0505, "step": 37630 }, { "grad_norm": 0.4842318892478943, "learning_rate": 5.8964737945461316e-05, "loss": 0.0396, "step": 37640 }, { "grad_norm": 1.09037184715271, "learning_rate": 5.8944403730113885e-05, "loss": 0.0399, "step": 37650 }, { "grad_norm": 0.41390082240104675, "learning_rate": 5.892406798641248e-05, "loss": 0.0425, "step": 37660 }, { "grad_norm": 0.3557102680206299, "learning_rate": 5.890373071783193e-05, "loss": 0.0367, "step": 37670 }, { "grad_norm": 0.4608297348022461, "learning_rate": 5.888339192784732e-05, "loss": 0.0359, "step": 37680 }, { "grad_norm": 0.5036818385124207, "learning_rate": 5.8863051619934003e-05, "loss": 0.0484, "step": 37690 }, { "grad_norm": 0.4885801374912262, "learning_rate": 5.8842709797567554e-05, "loss": 0.0378, "step": 37700 }, { "grad_norm": 0.46139559149742126, "learning_rate": 5.8822366464223855e-05, "loss": 0.0515, "step": 37710 }, { "grad_norm": 0.4032752215862274, "learning_rate": 5.880202162337901e-05, "loss": 0.0467, "step": 37720 }, { "grad_norm": 0.45074230432510376, "learning_rate": 5.8781675278509405e-05, "loss": 0.0369, "step": 37730 }, { "grad_norm": 0.5447295904159546, "learning_rate": 5.8761327433091696e-05, "loss": 0.0414, "step": 37740 }, { "grad_norm": 0.5869011878967285, "learning_rate": 5.874097809060275e-05, "loss": 0.0436, "step": 37750 }, { "grad_norm": 0.5624155402183533, "learning_rate": 5.87206272545197e-05, "loss": 0.0375, "step": 37760 }, { "grad_norm": 0.32673564553260803, "learning_rate": 5.8700274928319955e-05, "loss": 0.0415, "step": 37770 }, { "grad_norm": 0.2951319217681885, "learning_rate": 5.867992111548118e-05, "loss": 0.0498, "step": 37780 }, { "grad_norm": 0.3393334448337555, "learning_rate": 5.865956581948131e-05, "loss": 0.0419, "step": 37790 }, { "grad_norm": 0.402881920337677, "learning_rate": 5.863920904379845e-05, "loss": 0.0577, "step": 37800 }, { "grad_norm": 0.5591747760772705, "learning_rate": 5.8618850791911064e-05, "loss": 0.0331, "step": 37810 }, { "grad_norm": 0.38371384143829346, "learning_rate": 5.859849106729779e-05, "loss": 0.0379, "step": 37820 }, { "grad_norm": 0.776947557926178, "learning_rate": 5.857812987343758e-05, "loss": 0.0366, "step": 37830 }, { "grad_norm": 0.7904929518699646, "learning_rate": 5.855776721380957e-05, "loss": 0.0378, "step": 37840 }, { "grad_norm": 0.7575709819793701, "learning_rate": 5.8537403091893217e-05, "loss": 0.0424, "step": 37850 }, { "grad_norm": 0.5666773915290833, "learning_rate": 5.851703751116816e-05, "loss": 0.0474, "step": 37860 }, { "grad_norm": 0.6036673784255981, "learning_rate": 5.8496670475114336e-05, "loss": 0.0354, "step": 37870 }, { "grad_norm": 0.5022546648979187, "learning_rate": 5.84763019872119e-05, "loss": 0.0314, "step": 37880 }, { "grad_norm": 0.39514487981796265, "learning_rate": 5.845593205094131e-05, "loss": 0.0303, "step": 37890 }, { "grad_norm": 0.2727509140968323, "learning_rate": 5.843556066978318e-05, "loss": 0.0306, "step": 37900 }, { "grad_norm": 0.447710782289505, "learning_rate": 5.8415187847218455e-05, "loss": 0.0341, "step": 37910 }, { "grad_norm": 0.4459982216358185, "learning_rate": 5.839481358672827e-05, "loss": 0.0431, "step": 37920 }, { "grad_norm": 0.5764710903167725, "learning_rate": 5.837443789179407e-05, "loss": 0.0354, "step": 37930 }, { "grad_norm": 0.6609992384910583, "learning_rate": 5.8354060765897445e-05, "loss": 0.0405, "step": 37940 }, { "grad_norm": 0.42032432556152344, "learning_rate": 5.8333682212520334e-05, "loss": 0.0353, "step": 37950 }, { "grad_norm": 0.4797316789627075, "learning_rate": 5.831330223514486e-05, "loss": 0.0401, "step": 37960 }, { "grad_norm": 0.2552601993083954, "learning_rate": 5.8292920837253396e-05, "loss": 0.0338, "step": 37970 }, { "grad_norm": 0.46131330728530884, "learning_rate": 5.827253802232857e-05, "loss": 0.0417, "step": 37980 }, { "grad_norm": 0.630090057849884, "learning_rate": 5.825215379385327e-05, "loss": 0.0328, "step": 37990 }, { "grad_norm": 0.49818864464759827, "learning_rate": 5.823176815531057e-05, "loss": 0.0461, "step": 38000 }, { "grad_norm": 0.4271346926689148, "learning_rate": 5.8211381110183826e-05, "loss": 0.0451, "step": 38010 }, { "grad_norm": 0.3485530912876129, "learning_rate": 5.8190992661956645e-05, "loss": 0.039, "step": 38020 }, { "grad_norm": 0.5544759631156921, "learning_rate": 5.817060281411284e-05, "loss": 0.0408, "step": 38030 }, { "grad_norm": 0.6603007912635803, "learning_rate": 5.815021157013647e-05, "loss": 0.0345, "step": 38040 }, { "grad_norm": 0.6382023096084595, "learning_rate": 5.8129818933511856e-05, "loss": 0.0375, "step": 38050 }, { "grad_norm": 0.5563545823097229, "learning_rate": 5.8109424907723544e-05, "loss": 0.045, "step": 38060 }, { "grad_norm": 0.3490310609340668, "learning_rate": 5.80890294962563e-05, "loss": 0.0491, "step": 38070 }, { "grad_norm": 0.509425938129425, "learning_rate": 5.806863270259515e-05, "loss": 0.0347, "step": 38080 }, { "grad_norm": 0.5772629380226135, "learning_rate": 5.804823453022536e-05, "loss": 0.0431, "step": 38090 }, { "grad_norm": 0.7972419261932373, "learning_rate": 5.80278349826324e-05, "loss": 0.05, "step": 38100 }, { "grad_norm": 0.25915610790252686, "learning_rate": 5.8007434063302014e-05, "loss": 0.029, "step": 38110 }, { "grad_norm": 0.43584370613098145, "learning_rate": 5.7987031775720136e-05, "loss": 0.0404, "step": 38120 }, { "grad_norm": 1.0762969255447388, "learning_rate": 5.7966628123372976e-05, "loss": 0.0445, "step": 38130 }, { "grad_norm": 0.6491423845291138, "learning_rate": 5.7946223109746956e-05, "loss": 0.0432, "step": 38140 }, { "grad_norm": 0.4112643897533417, "learning_rate": 5.7925816738328754e-05, "loss": 0.0362, "step": 38150 }, { "grad_norm": 0.5898676514625549, "learning_rate": 5.790540901260521e-05, "loss": 0.0408, "step": 38160 }, { "grad_norm": 0.47686100006103516, "learning_rate": 5.788499993606351e-05, "loss": 0.0538, "step": 38170 }, { "grad_norm": 0.6496869325637817, "learning_rate": 5.786458951219096e-05, "loss": 0.0422, "step": 38180 }, { "grad_norm": 0.42020243406295776, "learning_rate": 5.784417774447517e-05, "loss": 0.0549, "step": 38190 }, { "grad_norm": 0.595544159412384, "learning_rate": 5.782376463640393e-05, "loss": 0.0316, "step": 38200 }, { "grad_norm": 0.8954882025718689, "learning_rate": 5.780335019146531e-05, "loss": 0.0363, "step": 38210 }, { "grad_norm": 0.7919787764549255, "learning_rate": 5.778293441314755e-05, "loss": 0.0401, "step": 38220 }, { "grad_norm": 0.6433899402618408, "learning_rate": 5.776251730493917e-05, "loss": 0.0345, "step": 38230 }, { "grad_norm": 0.34360066056251526, "learning_rate": 5.774209887032887e-05, "loss": 0.0289, "step": 38240 }, { "grad_norm": 0.3074503242969513, "learning_rate": 5.772167911280565e-05, "loss": 0.0378, "step": 38250 }, { "grad_norm": 0.4242721498012543, "learning_rate": 5.770125803585864e-05, "loss": 0.0549, "step": 38260 }, { "grad_norm": 0.6130184531211853, "learning_rate": 5.768083564297726e-05, "loss": 0.0296, "step": 38270 }, { "grad_norm": 0.4837193191051483, "learning_rate": 5.766041193765114e-05, "loss": 0.0384, "step": 38280 }, { "grad_norm": 0.4554039537906647, "learning_rate": 5.763998692337015e-05, "loss": 0.0294, "step": 38290 }, { "grad_norm": 0.2595798969268799, "learning_rate": 5.761956060362433e-05, "loss": 0.0236, "step": 38300 }, { "grad_norm": 0.4818957448005676, "learning_rate": 5.7599132981904e-05, "loss": 0.0385, "step": 38310 }, { "grad_norm": 0.44661787152290344, "learning_rate": 5.75787040616997e-05, "loss": 0.0439, "step": 38320 }, { "grad_norm": 0.39921942353248596, "learning_rate": 5.755827384650212e-05, "loss": 0.0296, "step": 38330 }, { "grad_norm": 1.0867724418640137, "learning_rate": 5.753784233980228e-05, "loss": 0.0556, "step": 38340 }, { "grad_norm": 0.4229718744754791, "learning_rate": 5.751740954509135e-05, "loss": 0.0448, "step": 38350 }, { "grad_norm": 0.5228754281997681, "learning_rate": 5.7496975465860715e-05, "loss": 0.0355, "step": 38360 }, { "grad_norm": 0.6487453579902649, "learning_rate": 5.747654010560202e-05, "loss": 0.0734, "step": 38370 }, { "grad_norm": 0.6026005744934082, "learning_rate": 5.7456103467807097e-05, "loss": 0.0376, "step": 38380 }, { "grad_norm": 0.5481817722320557, "learning_rate": 5.7435665555968046e-05, "loss": 0.0493, "step": 38390 }, { "grad_norm": 0.7163761258125305, "learning_rate": 5.74152263735771e-05, "loss": 0.0549, "step": 38400 }, { "grad_norm": 0.3433927595615387, "learning_rate": 5.739478592412677e-05, "loss": 0.0422, "step": 38410 }, { "grad_norm": 0.6268470883369446, "learning_rate": 5.7374344211109766e-05, "loss": 0.0286, "step": 38420 }, { "grad_norm": 0.6031409502029419, "learning_rate": 5.735390123801904e-05, "loss": 0.0334, "step": 38430 }, { "grad_norm": 0.4546515643596649, "learning_rate": 5.7333457008347704e-05, "loss": 0.0406, "step": 38440 }, { "grad_norm": 0.5583583116531372, "learning_rate": 5.7313011525589156e-05, "loss": 0.0396, "step": 38450 }, { "grad_norm": 0.8383157849311829, "learning_rate": 5.729256479323694e-05, "loss": 0.0508, "step": 38460 }, { "grad_norm": 0.39711013436317444, "learning_rate": 5.727211681478485e-05, "loss": 0.0378, "step": 38470 }, { "grad_norm": 0.3943306505680084, "learning_rate": 5.7251667593726886e-05, "loss": 0.0443, "step": 38480 }, { "grad_norm": 0.6081166863441467, "learning_rate": 5.723121713355728e-05, "loss": 0.0454, "step": 38490 }, { "grad_norm": 0.4572767913341522, "learning_rate": 5.721076543777044e-05, "loss": 0.0387, "step": 38500 }, { "grad_norm": 0.5936862230300903, "learning_rate": 5.7190312509860986e-05, "loss": 0.0316, "step": 38510 }, { "grad_norm": 1.0514535903930664, "learning_rate": 5.716985835332379e-05, "loss": 0.0371, "step": 38520 }, { "grad_norm": 0.42949244379997253, "learning_rate": 5.714940297165389e-05, "loss": 0.0301, "step": 38530 }, { "grad_norm": 0.6072158217430115, "learning_rate": 5.712894636834656e-05, "loss": 0.0335, "step": 38540 }, { "grad_norm": 4.244582176208496, "learning_rate": 5.7108488546897276e-05, "loss": 0.046, "step": 38550 }, { "grad_norm": 0.34254834055900574, "learning_rate": 5.708802951080172e-05, "loss": 0.0342, "step": 38560 }, { "grad_norm": 0.40202003717422485, "learning_rate": 5.706756926355576e-05, "loss": 0.0535, "step": 38570 }, { "grad_norm": 0.5566496253013611, "learning_rate": 5.704710780865554e-05, "loss": 0.0512, "step": 38580 }, { "grad_norm": 0.5986654162406921, "learning_rate": 5.7026645149597325e-05, "loss": 0.0381, "step": 38590 }, { "grad_norm": 0.6119598746299744, "learning_rate": 5.700618128987764e-05, "loss": 0.0525, "step": 38600 }, { "grad_norm": 0.4284651577472687, "learning_rate": 5.698571623299317e-05, "loss": 0.0372, "step": 38610 }, { "grad_norm": 0.4149814248085022, "learning_rate": 5.696524998244086e-05, "loss": 0.0349, "step": 38620 }, { "grad_norm": 1.3922942876815796, "learning_rate": 5.6944782541717836e-05, "loss": 0.0486, "step": 38630 }, { "grad_norm": 0.5300973653793335, "learning_rate": 5.69243139143214e-05, "loss": 0.0589, "step": 38640 }, { "grad_norm": 0.4517592787742615, "learning_rate": 5.6903844103749125e-05, "loss": 0.0379, "step": 38650 }, { "grad_norm": 0.5515922904014587, "learning_rate": 5.688337311349869e-05, "loss": 0.0667, "step": 38660 }, { "grad_norm": 0.5980228185653687, "learning_rate": 5.6862900947068074e-05, "loss": 0.0381, "step": 38670 }, { "grad_norm": 0.746497392654419, "learning_rate": 5.6842427607955374e-05, "loss": 0.0773, "step": 38680 }, { "grad_norm": 0.46709275245666504, "learning_rate": 5.682195309965893e-05, "loss": 0.0448, "step": 38690 }, { "grad_norm": 0.2721235454082489, "learning_rate": 5.6801477425677294e-05, "loss": 0.0425, "step": 38700 }, { "grad_norm": 0.4030846357345581, "learning_rate": 5.678100058950917e-05, "loss": 0.0324, "step": 38710 }, { "grad_norm": 0.6385922431945801, "learning_rate": 5.676052259465352e-05, "loss": 0.0354, "step": 38720 }, { "grad_norm": 0.380815327167511, "learning_rate": 5.674004344460945e-05, "loss": 0.0411, "step": 38730 }, { "grad_norm": 0.386834055185318, "learning_rate": 5.6719563142876295e-05, "loss": 0.0374, "step": 38740 }, { "grad_norm": 0.4339199662208557, "learning_rate": 5.669908169295359e-05, "loss": 0.0363, "step": 38750 }, { "grad_norm": 0.6994999647140503, "learning_rate": 5.667859909834105e-05, "loss": 0.0701, "step": 38760 }, { "grad_norm": 0.5443011522293091, "learning_rate": 5.6658115362538565e-05, "loss": 0.0474, "step": 38770 }, { "grad_norm": 0.4011799991130829, "learning_rate": 5.663763048904628e-05, "loss": 0.0381, "step": 38780 }, { "grad_norm": 0.4544498920440674, "learning_rate": 5.661714448136447e-05, "loss": 0.0409, "step": 38790 }, { "grad_norm": 0.5117974281311035, "learning_rate": 5.659665734299366e-05, "loss": 0.045, "step": 38800 }, { "grad_norm": 0.593938946723938, "learning_rate": 5.6576169077434516e-05, "loss": 0.0332, "step": 38810 }, { "grad_norm": 0.354440301656723, "learning_rate": 5.6555679688187944e-05, "loss": 0.028, "step": 38820 }, { "grad_norm": 0.3022119104862213, "learning_rate": 5.6535189178755e-05, "loss": 0.0297, "step": 38830 }, { "grad_norm": 0.4636264443397522, "learning_rate": 5.651469755263698e-05, "loss": 0.0394, "step": 38840 }, { "grad_norm": 0.39300408959388733, "learning_rate": 5.6494204813335316e-05, "loss": 0.0447, "step": 38850 }, { "grad_norm": 0.6312022805213928, "learning_rate": 5.647371096435168e-05, "loss": 0.0622, "step": 38860 }, { "grad_norm": 0.5068891644477844, "learning_rate": 5.645321600918788e-05, "loss": 0.0304, "step": 38870 }, { "grad_norm": 0.35621485114097595, "learning_rate": 5.643271995134597e-05, "loss": 0.0395, "step": 38880 }, { "grad_norm": 0.5859026312828064, "learning_rate": 5.641222279432814e-05, "loss": 0.0409, "step": 38890 }, { "grad_norm": 0.3668367266654968, "learning_rate": 5.6391724541636834e-05, "loss": 0.0345, "step": 38900 }, { "grad_norm": 0.6026226878166199, "learning_rate": 5.6371225196774605e-05, "loss": 0.0537, "step": 38910 }, { "grad_norm": 0.43187493085861206, "learning_rate": 5.635072476324423e-05, "loss": 0.0408, "step": 38920 }, { "grad_norm": 0.46686866879463196, "learning_rate": 5.63302232445487e-05, "loss": 0.04, "step": 38930 }, { "grad_norm": 0.651386559009552, "learning_rate": 5.6309720644191144e-05, "loss": 0.048, "step": 38940 }, { "grad_norm": 0.3877556622028351, "learning_rate": 5.628921696567491e-05, "loss": 0.0404, "step": 38950 }, { "grad_norm": 0.476874440908432, "learning_rate": 5.62687122125035e-05, "loss": 0.0335, "step": 38960 }, { "grad_norm": 0.48478221893310547, "learning_rate": 5.624820638818062e-05, "loss": 0.0445, "step": 38970 }, { "grad_norm": 0.3685701787471771, "learning_rate": 5.6227699496210164e-05, "loss": 0.0608, "step": 38980 }, { "grad_norm": 0.5001015663146973, "learning_rate": 5.6207191540096195e-05, "loss": 0.0354, "step": 38990 }, { "grad_norm": 0.44284287095069885, "learning_rate": 5.618668252334296e-05, "loss": 0.033, "step": 39000 }, { "grad_norm": 0.3745025396347046, "learning_rate": 5.616617244945488e-05, "loss": 0.0428, "step": 39010 }, { "grad_norm": 0.5319738388061523, "learning_rate": 5.614566132193656e-05, "loss": 0.0443, "step": 39020 }, { "grad_norm": 0.24255669116973877, "learning_rate": 5.612514914429282e-05, "loss": 0.0441, "step": 39030 }, { "grad_norm": 0.6999022364616394, "learning_rate": 5.610463592002863e-05, "loss": 0.0391, "step": 39040 }, { "grad_norm": 0.5413198471069336, "learning_rate": 5.608412165264909e-05, "loss": 0.0254, "step": 39050 }, { "grad_norm": 0.64799565076828, "learning_rate": 5.606360634565959e-05, "loss": 0.0339, "step": 39060 }, { "grad_norm": 0.41316163539886475, "learning_rate": 5.604309000256559e-05, "loss": 0.0444, "step": 39070 }, { "grad_norm": 0.7703383564949036, "learning_rate": 5.6022572626872785e-05, "loss": 0.035, "step": 39080 }, { "grad_norm": 0.5955260992050171, "learning_rate": 5.600205422208704e-05, "loss": 0.0411, "step": 39090 }, { "grad_norm": 0.5367544889450073, "learning_rate": 5.59815347917144e-05, "loss": 0.0431, "step": 39100 }, { "grad_norm": 0.6093878149986267, "learning_rate": 5.596101433926103e-05, "loss": 0.0523, "step": 39110 }, { "grad_norm": 0.4470142722129822, "learning_rate": 5.5940492868233364e-05, "loss": 0.049, "step": 39120 }, { "grad_norm": 0.6159268021583557, "learning_rate": 5.591997038213793e-05, "loss": 0.0435, "step": 39130 }, { "grad_norm": 0.3808121383190155, "learning_rate": 5.5899446884481475e-05, "loss": 0.0286, "step": 39140 }, { "grad_norm": 0.36493298411369324, "learning_rate": 5.5878922378770906e-05, "loss": 0.0428, "step": 39150 }, { "grad_norm": 0.5121458172798157, "learning_rate": 5.5858396868513285e-05, "loss": 0.0401, "step": 39160 }, { "grad_norm": 0.9292040467262268, "learning_rate": 5.583787035721586e-05, "loss": 0.0462, "step": 39170 }, { "grad_norm": 0.33683040738105774, "learning_rate": 5.581734284838606e-05, "loss": 0.0361, "step": 39180 }, { "grad_norm": 1.369724154472351, "learning_rate": 5.579681434553147e-05, "loss": 0.0375, "step": 39190 }, { "grad_norm": 0.6864435076713562, "learning_rate": 5.5776284852159854e-05, "loss": 0.0388, "step": 39200 }, { "grad_norm": 0.3415317237377167, "learning_rate": 5.575575437177913e-05, "loss": 0.0361, "step": 39210 }, { "grad_norm": 0.4388715922832489, "learning_rate": 5.573522290789742e-05, "loss": 0.0356, "step": 39220 }, { "grad_norm": 0.378081351518631, "learning_rate": 5.571469046402298e-05, "loss": 0.0419, "step": 39230 }, { "grad_norm": 0.4847550392150879, "learning_rate": 5.5694157043664205e-05, "loss": 0.0333, "step": 39240 }, { "grad_norm": 0.5976060628890991, "learning_rate": 5.567362265032975e-05, "loss": 0.0502, "step": 39250 }, { "grad_norm": 0.6505042910575867, "learning_rate": 5.565308728752836e-05, "loss": 0.0295, "step": 39260 }, { "grad_norm": 1.0863019227981567, "learning_rate": 5.5632550958768956e-05, "loss": 0.0405, "step": 39270 }, { "grad_norm": 0.388784795999527, "learning_rate": 5.5612013667560636e-05, "loss": 0.0433, "step": 39280 }, { "grad_norm": 0.4914911687374115, "learning_rate": 5.5591475417412676e-05, "loss": 0.0399, "step": 39290 }, { "grad_norm": 0.5727514624595642, "learning_rate": 5.557093621183451e-05, "loss": 0.0429, "step": 39300 }, { "grad_norm": 0.40594902634620667, "learning_rate": 5.55503960543357e-05, "loss": 0.0367, "step": 39310 }, { "grad_norm": 0.4642051160335541, "learning_rate": 5.552985494842601e-05, "loss": 0.0371, "step": 39320 }, { "grad_norm": 0.5891808867454529, "learning_rate": 5.550931289761534e-05, "loss": 0.0392, "step": 39330 }, { "grad_norm": 0.4313172698020935, "learning_rate": 5.548876990541378e-05, "loss": 0.0255, "step": 39340 }, { "grad_norm": 0.6948270201683044, "learning_rate": 5.5468225975331564e-05, "loss": 0.0347, "step": 39350 }, { "grad_norm": 0.3581996560096741, "learning_rate": 5.544768111087909e-05, "loss": 0.0358, "step": 39360 }, { "grad_norm": 0.4241708815097809, "learning_rate": 5.54271353155669e-05, "loss": 0.042, "step": 39370 }, { "grad_norm": 0.6884310841560364, "learning_rate": 5.5406588592905715e-05, "loss": 0.0382, "step": 39380 }, { "grad_norm": 0.5292034149169922, "learning_rate": 5.5386040946406416e-05, "loss": 0.0344, "step": 39390 }, { "grad_norm": 0.4180547893047333, "learning_rate": 5.536549237958004e-05, "loss": 0.0323, "step": 39400 }, { "grad_norm": 0.6013640761375427, "learning_rate": 5.5344942895937744e-05, "loss": 0.0417, "step": 39410 }, { "grad_norm": 0.39644598960876465, "learning_rate": 5.5324392498990904e-05, "loss": 0.0383, "step": 39420 }, { "grad_norm": 0.6364018321037292, "learning_rate": 5.5303841192251e-05, "loss": 0.0346, "step": 39430 }, { "grad_norm": 0.6868157386779785, "learning_rate": 5.52832889792297e-05, "loss": 0.0471, "step": 39440 }, { "grad_norm": 0.7792892456054688, "learning_rate": 5.526273586343881e-05, "loss": 0.0391, "step": 39450 }, { "grad_norm": 0.5183596611022949, "learning_rate": 5.5242181848390306e-05, "loss": 0.0432, "step": 39460 }, { "grad_norm": 0.5030085444450378, "learning_rate": 5.5221626937596285e-05, "loss": 0.0515, "step": 39470 }, { "grad_norm": 0.5025416016578674, "learning_rate": 5.520107113456903e-05, "loss": 0.0327, "step": 39480 }, { "grad_norm": 0.6814541816711426, "learning_rate": 5.5180514442820974e-05, "loss": 0.0315, "step": 39490 }, { "grad_norm": 0.5605167150497437, "learning_rate": 5.515995686586469e-05, "loss": 0.0386, "step": 39500 }, { "grad_norm": 0.37320777773857117, "learning_rate": 5.5139398407212916e-05, "loss": 0.0393, "step": 39510 }, { "grad_norm": 0.492496520280838, "learning_rate": 5.511883907037849e-05, "loss": 0.0418, "step": 39520 }, { "grad_norm": 0.4807118773460388, "learning_rate": 5.509827885887449e-05, "loss": 0.0318, "step": 39530 }, { "grad_norm": 0.46328204870224, "learning_rate": 5.507771777621406e-05, "loss": 0.0277, "step": 39540 }, { "grad_norm": 0.42469117045402527, "learning_rate": 5.505715582591052e-05, "loss": 0.0426, "step": 39550 }, { "grad_norm": 0.5355798602104187, "learning_rate": 5.50365930114774e-05, "loss": 0.032, "step": 39560 }, { "grad_norm": 0.4709698557853699, "learning_rate": 5.5016029336428255e-05, "loss": 0.0444, "step": 39570 }, { "grad_norm": 0.41128504276275635, "learning_rate": 5.49954648042769e-05, "loss": 0.0416, "step": 39580 }, { "grad_norm": 0.4005856215953827, "learning_rate": 5.4974899418537226e-05, "loss": 0.0354, "step": 39590 }, { "grad_norm": 0.7194512486457825, "learning_rate": 5.4954333182723316e-05, "loss": 0.0379, "step": 39600 }, { "grad_norm": 0.5494620203971863, "learning_rate": 5.493376610034937e-05, "loss": 0.0317, "step": 39610 }, { "grad_norm": 0.4496108293533325, "learning_rate": 5.4913198174929735e-05, "loss": 0.0424, "step": 39620 }, { "grad_norm": 0.5359855890274048, "learning_rate": 5.489262940997889e-05, "loss": 0.0444, "step": 39630 }, { "grad_norm": 0.39059457182884216, "learning_rate": 5.487205980901151e-05, "loss": 0.0306, "step": 39640 }, { "grad_norm": 0.43944332003593445, "learning_rate": 5.485148937554234e-05, "loss": 0.035, "step": 39650 }, { "grad_norm": 0.44467297196388245, "learning_rate": 5.483091811308635e-05, "loss": 0.0333, "step": 39660 }, { "grad_norm": 0.406043142080307, "learning_rate": 5.4810346025158564e-05, "loss": 0.0393, "step": 39670 }, { "grad_norm": 0.33779647946357727, "learning_rate": 5.478977311527421e-05, "loss": 0.0393, "step": 39680 }, { "grad_norm": 0.5461012125015259, "learning_rate": 5.476919938694863e-05, "loss": 0.0463, "step": 39690 }, { "grad_norm": 0.5866457223892212, "learning_rate": 5.474862484369733e-05, "loss": 0.0516, "step": 39700 }, { "grad_norm": 0.389946848154068, "learning_rate": 5.472804948903589e-05, "loss": 0.0312, "step": 39710 }, { "grad_norm": 0.6647882461547852, "learning_rate": 5.470747332648013e-05, "loss": 0.0645, "step": 39720 }, { "grad_norm": 0.3583804666996002, "learning_rate": 5.468689635954591e-05, "loss": 0.0282, "step": 39730 }, { "grad_norm": 0.5101736783981323, "learning_rate": 5.46663185917493e-05, "loss": 0.047, "step": 39740 }, { "grad_norm": 0.47873806953430176, "learning_rate": 5.464574002660645e-05, "loss": 0.0458, "step": 39750 }, { "grad_norm": 0.5673699378967285, "learning_rate": 5.4625160667633724e-05, "loss": 0.0406, "step": 39760 }, { "grad_norm": 0.6113342642784119, "learning_rate": 5.4604580518347505e-05, "loss": 0.0456, "step": 39770 }, { "grad_norm": 0.4671206474304199, "learning_rate": 5.458399958226442e-05, "loss": 0.0337, "step": 39780 }, { "grad_norm": 0.3783515989780426, "learning_rate": 5.456341786290119e-05, "loss": 0.0478, "step": 39790 }, { "grad_norm": 0.45115211606025696, "learning_rate": 5.454283536377465e-05, "loss": 0.0688, "step": 39800 }, { "grad_norm": 0.6229239702224731, "learning_rate": 5.452225208840179e-05, "loss": 0.0418, "step": 39810 }, { "grad_norm": 0.5436581969261169, "learning_rate": 5.450166804029975e-05, "loss": 0.0381, "step": 39820 }, { "grad_norm": 0.4034520089626312, "learning_rate": 5.448108322298574e-05, "loss": 0.0353, "step": 39830 }, { "grad_norm": 0.4123004376888275, "learning_rate": 5.446049763997717e-05, "loss": 0.0447, "step": 39840 }, { "grad_norm": 0.545071542263031, "learning_rate": 5.4439911294791546e-05, "loss": 0.0356, "step": 39850 }, { "grad_norm": 0.5853373408317566, "learning_rate": 5.441932419094652e-05, "loss": 0.0403, "step": 39860 }, { "grad_norm": 0.4083392024040222, "learning_rate": 5.439873633195985e-05, "loss": 0.0354, "step": 39870 }, { "grad_norm": 0.6871432065963745, "learning_rate": 5.437814772134947e-05, "loss": 0.0364, "step": 39880 }, { "grad_norm": 0.5217333436012268, "learning_rate": 5.4357558362633366e-05, "loss": 0.0325, "step": 39890 }, { "grad_norm": 0.549468457698822, "learning_rate": 5.4336968259329726e-05, "loss": 0.0445, "step": 39900 }, { "grad_norm": 0.6518493294715881, "learning_rate": 5.431637741495681e-05, "loss": 0.0442, "step": 39910 }, { "grad_norm": 0.5477796196937561, "learning_rate": 5.429578583303307e-05, "loss": 0.042, "step": 39920 }, { "grad_norm": 1.5377566814422607, "learning_rate": 5.427519351707701e-05, "loss": 0.0476, "step": 39930 }, { "grad_norm": 0.6596348285675049, "learning_rate": 5.42546004706073e-05, "loss": 0.0332, "step": 39940 }, { "grad_norm": 0.5663982033729553, "learning_rate": 5.4234006697142735e-05, "loss": 0.0424, "step": 39950 }, { "grad_norm": 0.5748822093009949, "learning_rate": 5.421341220020224e-05, "loss": 0.0478, "step": 39960 }, { "grad_norm": 1.3928289413452148, "learning_rate": 5.419281698330482e-05, "loss": 0.0496, "step": 39970 }, { "grad_norm": 0.5865259170532227, "learning_rate": 5.4172221049969665e-05, "loss": 0.0437, "step": 39980 }, { "grad_norm": 0.622166097164154, "learning_rate": 5.415162440371604e-05, "loss": 0.0449, "step": 39990 }, { "grad_norm": 0.9343371391296387, "learning_rate": 5.413102704806334e-05, "loss": 0.0468, "step": 40000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 40000, "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": 4, "trial_name": null, "trial_params": null }