{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.75, "eval_steps": 500, "global_step": 150000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.1597956418991089, "learning_rate": 9e-08, "loss": 1.5266, "step": 10 }, { "grad_norm": 0.9323684573173523, "learning_rate": 1.9e-07, "loss": 1.5197, "step": 20 }, { "grad_norm": 0.935326337814331, "learning_rate": 2.9e-07, "loss": 1.4696, "step": 30 }, { "grad_norm": 1.0758224725723267, "learning_rate": 3.9e-07, "loss": 1.5114, "step": 40 }, { "grad_norm": 0.9333432912826538, "learning_rate": 4.9e-07, "loss": 1.4916, "step": 50 }, { "grad_norm": 0.9084240198135376, "learning_rate": 5.9e-07, "loss": 1.4876, "step": 60 }, { "grad_norm": 0.9874604344367981, "learning_rate": 6.900000000000001e-07, "loss": 1.4969, "step": 70 }, { "grad_norm": 1.0728893280029297, "learning_rate": 7.900000000000002e-07, "loss": 1.4949, "step": 80 }, { "grad_norm": 1.3765158653259277, "learning_rate": 8.900000000000001e-07, "loss": 1.4524, "step": 90 }, { "grad_norm": 1.7702677249908447, "learning_rate": 9.9e-07, "loss": 1.3651, "step": 100 }, { "grad_norm": 9.079279899597168, "learning_rate": 1.0900000000000002e-06, "loss": 1.3656, "step": 110 }, { "grad_norm": 1.997900128364563, "learning_rate": 1.19e-06, "loss": 1.3184, "step": 120 }, { "grad_norm": 1.8172906637191772, "learning_rate": 1.2900000000000001e-06, "loss": 1.2682, "step": 130 }, { "grad_norm": 1.989872694015503, "learning_rate": 1.39e-06, "loss": 1.2996, "step": 140 }, { "grad_norm": 2.068657636642456, "learning_rate": 1.4900000000000001e-06, "loss": 1.312, "step": 150 }, { "grad_norm": 1.7207456827163696, "learning_rate": 1.5900000000000002e-06, "loss": 1.2867, "step": 160 }, { "grad_norm": 1.7290955781936646, "learning_rate": 1.69e-06, "loss": 1.299, "step": 170 }, { "grad_norm": 2.2355523109436035, "learning_rate": 1.79e-06, "loss": 1.28, "step": 180 }, { "grad_norm": 1.8418817520141602, "learning_rate": 1.8900000000000001e-06, "loss": 1.2862, "step": 190 }, { "grad_norm": 1.8650226593017578, "learning_rate": 1.99e-06, "loss": 1.2708, "step": 200 }, { "grad_norm": 2.1259586811065674, "learning_rate": 2.09e-06, "loss": 1.2487, "step": 210 }, { "grad_norm": 2.045229911804199, "learning_rate": 2.19e-06, "loss": 1.2312, "step": 220 }, { "grad_norm": 2.871631145477295, "learning_rate": 2.29e-06, "loss": 1.2091, "step": 230 }, { "grad_norm": 2.6972479820251465, "learning_rate": 2.3900000000000004e-06, "loss": 1.2191, "step": 240 }, { "grad_norm": 1.9211273193359375, "learning_rate": 2.49e-06, "loss": 1.1862, "step": 250 }, { "grad_norm": 4.512012958526611, "learning_rate": 2.59e-06, "loss": 1.1929, "step": 260 }, { "grad_norm": 2.6934359073638916, "learning_rate": 2.69e-06, "loss": 1.2237, "step": 270 }, { "grad_norm": 2.235018491744995, "learning_rate": 2.7900000000000004e-06, "loss": 1.1878, "step": 280 }, { "grad_norm": 2.525601863861084, "learning_rate": 2.89e-06, "loss": 1.1629, "step": 290 }, { "grad_norm": 2.9258246421813965, "learning_rate": 2.99e-06, "loss": 1.1508, "step": 300 }, { "grad_norm": 2.36173415184021, "learning_rate": 3.09e-06, "loss": 1.1731, "step": 310 }, { "grad_norm": 1.6708860397338867, "learning_rate": 3.19e-06, "loss": 1.1855, "step": 320 }, { "grad_norm": 2.7565224170684814, "learning_rate": 3.29e-06, "loss": 1.1462, "step": 330 }, { "grad_norm": 1.7039204835891724, "learning_rate": 3.39e-06, "loss": 1.1465, "step": 340 }, { "grad_norm": 2.832768440246582, "learning_rate": 3.49e-06, "loss": 1.1464, "step": 350 }, { "grad_norm": 6.346458435058594, "learning_rate": 3.5900000000000004e-06, "loss": 1.1277, "step": 360 }, { "grad_norm": 1.7389123439788818, "learning_rate": 3.6900000000000002e-06, "loss": 1.1163, "step": 370 }, { "grad_norm": 2.0680041313171387, "learning_rate": 3.7900000000000006e-06, "loss": 1.1262, "step": 380 }, { "grad_norm": 1.8653638362884521, "learning_rate": 3.89e-06, "loss": 1.1241, "step": 390 }, { "grad_norm": 2.541369915008545, "learning_rate": 3.99e-06, "loss": 1.1164, "step": 400 }, { "grad_norm": 2.1084954738616943, "learning_rate": 4.09e-06, "loss": 1.1312, "step": 410 }, { "grad_norm": 2.24845290184021, "learning_rate": 4.1900000000000005e-06, "loss": 1.088, "step": 420 }, { "grad_norm": 2.055889368057251, "learning_rate": 4.2900000000000004e-06, "loss": 1.1235, "step": 430 }, { "grad_norm": 1.8453360795974731, "learning_rate": 4.39e-06, "loss": 1.1338, "step": 440 }, { "grad_norm": 2.3952550888061523, "learning_rate": 4.49e-06, "loss": 1.1139, "step": 450 }, { "grad_norm": 2.2730326652526855, "learning_rate": 4.590000000000001e-06, "loss": 1.1419, "step": 460 }, { "grad_norm": 2.303610324859619, "learning_rate": 4.69e-06, "loss": 1.1168, "step": 470 }, { "grad_norm": 2.418318271636963, "learning_rate": 4.79e-06, "loss": 1.1082, "step": 480 }, { "grad_norm": 1.7430905103683472, "learning_rate": 4.89e-06, "loss": 1.0744, "step": 490 }, { "grad_norm": 3.2563443183898926, "learning_rate": 4.9900000000000005e-06, "loss": 1.0882, "step": 500 }, { "grad_norm": 1.9141615629196167, "learning_rate": 5.09e-06, "loss": 1.0942, "step": 510 }, { "grad_norm": 2.2211759090423584, "learning_rate": 5.19e-06, "loss": 1.0783, "step": 520 }, { "grad_norm": 2.10693359375, "learning_rate": 5.29e-06, "loss": 1.1192, "step": 530 }, { "grad_norm": 1.7789429426193237, "learning_rate": 5.390000000000001e-06, "loss": 1.0927, "step": 540 }, { "grad_norm": 1.9179975986480713, "learning_rate": 5.49e-06, "loss": 1.128, "step": 550 }, { "grad_norm": 1.5676915645599365, "learning_rate": 5.59e-06, "loss": 1.0895, "step": 560 }, { "grad_norm": 1.5406525135040283, "learning_rate": 5.690000000000001e-06, "loss": 1.0966, "step": 570 }, { "grad_norm": 1.926806926727295, "learning_rate": 5.7900000000000005e-06, "loss": 1.0906, "step": 580 }, { "grad_norm": 1.9671385288238525, "learning_rate": 5.89e-06, "loss": 1.0705, "step": 590 }, { "grad_norm": 2.451265335083008, "learning_rate": 5.99e-06, "loss": 1.0675, "step": 600 }, { "grad_norm": 1.86956787109375, "learning_rate": 6.090000000000001e-06, "loss": 1.0713, "step": 610 }, { "grad_norm": 1.954190731048584, "learning_rate": 6.19e-06, "loss": 1.0708, "step": 620 }, { "grad_norm": 9.347692489624023, "learning_rate": 6.29e-06, "loss": 1.0945, "step": 630 }, { "grad_norm": 2.3935024738311768, "learning_rate": 6.39e-06, "loss": 1.0856, "step": 640 }, { "grad_norm": 2.385981321334839, "learning_rate": 6.4900000000000005e-06, "loss": 1.1162, "step": 650 }, { "grad_norm": 1.5512551069259644, "learning_rate": 6.5900000000000004e-06, "loss": 1.0853, "step": 660 }, { "grad_norm": 1.7993474006652832, "learning_rate": 6.69e-06, "loss": 1.0757, "step": 670 }, { "grad_norm": 1.8108251094818115, "learning_rate": 6.79e-06, "loss": 1.1079, "step": 680 }, { "grad_norm": 1.7266809940338135, "learning_rate": 6.890000000000001e-06, "loss": 1.1029, "step": 690 }, { "grad_norm": 1.7809972763061523, "learning_rate": 6.990000000000001e-06, "loss": 1.0681, "step": 700 }, { "grad_norm": 1.5047270059585571, "learning_rate": 7.090000000000001e-06, "loss": 1.0899, "step": 710 }, { "grad_norm": 1.860785961151123, "learning_rate": 7.190000000000001e-06, "loss": 1.0843, "step": 720 }, { "grad_norm": 3.5653421878814697, "learning_rate": 7.290000000000001e-06, "loss": 1.096, "step": 730 }, { "grad_norm": 1.8767629861831665, "learning_rate": 7.3899999999999995e-06, "loss": 1.0898, "step": 740 }, { "grad_norm": 1.488114595413208, "learning_rate": 7.4899999999999994e-06, "loss": 1.0303, "step": 750 }, { "grad_norm": 2.186192750930786, "learning_rate": 7.59e-06, "loss": 1.0879, "step": 760 }, { "grad_norm": 1.6138622760772705, "learning_rate": 7.69e-06, "loss": 1.074, "step": 770 }, { "grad_norm": 1.8701101541519165, "learning_rate": 7.79e-06, "loss": 1.091, "step": 780 }, { "grad_norm": 1.713696002960205, "learning_rate": 7.89e-06, "loss": 1.0646, "step": 790 }, { "grad_norm": 1.8958468437194824, "learning_rate": 7.99e-06, "loss": 1.0839, "step": 800 }, { "grad_norm": 1.8440510034561157, "learning_rate": 8.09e-06, "loss": 1.0614, "step": 810 }, { "grad_norm": 1.6722640991210938, "learning_rate": 8.190000000000001e-06, "loss": 1.0642, "step": 820 }, { "grad_norm": 2.3286936283111572, "learning_rate": 8.29e-06, "loss": 1.0898, "step": 830 }, { "grad_norm": 2.0758705139160156, "learning_rate": 8.390000000000001e-06, "loss": 1.0883, "step": 840 }, { "grad_norm": 2.4472177028656006, "learning_rate": 8.49e-06, "loss": 1.0641, "step": 850 }, { "grad_norm": 1.6009747982025146, "learning_rate": 8.59e-06, "loss": 1.0343, "step": 860 }, { "grad_norm": 1.6581579446792603, "learning_rate": 8.690000000000002e-06, "loss": 1.0122, "step": 870 }, { "grad_norm": 2.036007881164551, "learning_rate": 8.79e-06, "loss": 1.008, "step": 880 }, { "grad_norm": 2.1194915771484375, "learning_rate": 8.890000000000001e-06, "loss": 1.0149, "step": 890 }, { "grad_norm": 7.445390224456787, "learning_rate": 8.99e-06, "loss": 1.0153, "step": 900 }, { "grad_norm": 2.8214237689971924, "learning_rate": 9.09e-06, "loss": 1.0221, "step": 910 }, { "grad_norm": 2.342575788497925, "learning_rate": 9.19e-06, "loss": 0.9742, "step": 920 }, { "grad_norm": 4.103582382202148, "learning_rate": 9.29e-06, "loss": 0.9703, "step": 930 }, { "grad_norm": 2.993591070175171, "learning_rate": 9.39e-06, "loss": 0.9391, "step": 940 }, { "grad_norm": 2.8020408153533936, "learning_rate": 9.49e-06, "loss": 0.9273, "step": 950 }, { "grad_norm": 2.4974875450134277, "learning_rate": 9.59e-06, "loss": 0.9308, "step": 960 }, { "grad_norm": 2.3020992279052734, "learning_rate": 9.69e-06, "loss": 0.8998, "step": 970 }, { "grad_norm": 2.4740002155303955, "learning_rate": 9.790000000000001e-06, "loss": 0.8667, "step": 980 }, { "grad_norm": 2.322462558746338, "learning_rate": 9.89e-06, "loss": 0.8353, "step": 990 }, { "grad_norm": 3.3505725860595703, "learning_rate": 9.990000000000001e-06, "loss": 0.8634, "step": 1000 }, { "grad_norm": 2.396606206893921, "learning_rate": 1.0090000000000002e-05, "loss": 0.8177, "step": 1010 }, { "grad_norm": 3.4732909202575684, "learning_rate": 1.019e-05, "loss": 0.82, "step": 1020 }, { "grad_norm": 3.048571825027466, "learning_rate": 1.0290000000000001e-05, "loss": 0.8099, "step": 1030 }, { "grad_norm": 2.4829654693603516, "learning_rate": 1.039e-05, "loss": 0.7641, "step": 1040 }, { "grad_norm": 2.6477584838867188, "learning_rate": 1.049e-05, "loss": 0.7766, "step": 1050 }, { "grad_norm": 3.0108087062835693, "learning_rate": 1.059e-05, "loss": 0.7264, "step": 1060 }, { "grad_norm": 3.090052604675293, "learning_rate": 1.069e-05, "loss": 0.7226, "step": 1070 }, { "grad_norm": 2.6824381351470947, "learning_rate": 1.079e-05, "loss": 0.6711, "step": 1080 }, { "grad_norm": 2.9290597438812256, "learning_rate": 1.089e-05, "loss": 0.7219, "step": 1090 }, { "grad_norm": 3.0847206115722656, "learning_rate": 1.099e-05, "loss": 0.6894, "step": 1100 }, { "grad_norm": 3.224210739135742, "learning_rate": 1.109e-05, "loss": 0.6745, "step": 1110 }, { "grad_norm": 3.4148528575897217, "learning_rate": 1.1190000000000001e-05, "loss": 0.6683, "step": 1120 }, { "grad_norm": 2.495229959487915, "learning_rate": 1.129e-05, "loss": 0.6294, "step": 1130 }, { "grad_norm": 3.4450790882110596, "learning_rate": 1.1390000000000001e-05, "loss": 0.6325, "step": 1140 }, { "grad_norm": 3.072985887527466, "learning_rate": 1.149e-05, "loss": 0.6192, "step": 1150 }, { "grad_norm": 3.072622299194336, "learning_rate": 1.159e-05, "loss": 0.5805, "step": 1160 }, { "grad_norm": 2.376331090927124, "learning_rate": 1.1690000000000002e-05, "loss": 0.6053, "step": 1170 }, { "grad_norm": 3.044680118560791, "learning_rate": 1.179e-05, "loss": 0.5822, "step": 1180 }, { "grad_norm": 3.175612688064575, "learning_rate": 1.1890000000000001e-05, "loss": 0.5375, "step": 1190 }, { "grad_norm": 2.816089391708374, "learning_rate": 1.199e-05, "loss": 0.5672, "step": 1200 }, { "grad_norm": 2.8884665966033936, "learning_rate": 1.209e-05, "loss": 0.5392, "step": 1210 }, { "grad_norm": 3.3984720706939697, "learning_rate": 1.219e-05, "loss": 0.5432, "step": 1220 }, { "grad_norm": 3.7283458709716797, "learning_rate": 1.2290000000000001e-05, "loss": 0.5269, "step": 1230 }, { "grad_norm": 2.5753977298736572, "learning_rate": 1.239e-05, "loss": 0.4775, "step": 1240 }, { "grad_norm": 3.382607936859131, "learning_rate": 1.249e-05, "loss": 0.4958, "step": 1250 }, { "grad_norm": 2.8088152408599854, "learning_rate": 1.2590000000000001e-05, "loss": 0.4821, "step": 1260 }, { "grad_norm": 3.692821502685547, "learning_rate": 1.2690000000000002e-05, "loss": 0.4661, "step": 1270 }, { "grad_norm": 4.548642635345459, "learning_rate": 1.2790000000000001e-05, "loss": 0.4713, "step": 1280 }, { "grad_norm": 4.1241655349731445, "learning_rate": 1.2889999999999999e-05, "loss": 0.4049, "step": 1290 }, { "grad_norm": 3.4829118251800537, "learning_rate": 1.299e-05, "loss": 0.4511, "step": 1300 }, { "grad_norm": 2.7466044425964355, "learning_rate": 1.309e-05, "loss": 0.4101, "step": 1310 }, { "grad_norm": 2.9366748332977295, "learning_rate": 1.3189999999999999e-05, "loss": 0.4287, "step": 1320 }, { "grad_norm": 3.8625328540802, "learning_rate": 1.329e-05, "loss": 0.3786, "step": 1330 }, { "grad_norm": 2.9389538764953613, "learning_rate": 1.339e-05, "loss": 0.3478, "step": 1340 }, { "grad_norm": 2.8211097717285156, "learning_rate": 1.349e-05, "loss": 0.3314, "step": 1350 }, { "grad_norm": 4.023573398590088, "learning_rate": 1.359e-05, "loss": 0.3855, "step": 1360 }, { "grad_norm": 3.225324869155884, "learning_rate": 1.369e-05, "loss": 0.3727, "step": 1370 }, { "grad_norm": 3.624112367630005, "learning_rate": 1.379e-05, "loss": 0.3545, "step": 1380 }, { "grad_norm": 3.676208257675171, "learning_rate": 1.389e-05, "loss": 0.3505, "step": 1390 }, { "grad_norm": 3.94494366645813, "learning_rate": 1.399e-05, "loss": 0.355, "step": 1400 }, { "grad_norm": 3.3171660900115967, "learning_rate": 1.409e-05, "loss": 0.2782, "step": 1410 }, { "grad_norm": 4.111855983734131, "learning_rate": 1.4190000000000001e-05, "loss": 0.3347, "step": 1420 }, { "grad_norm": 4.388930320739746, "learning_rate": 1.429e-05, "loss": 0.3229, "step": 1430 }, { "grad_norm": 3.2074239253997803, "learning_rate": 1.4390000000000001e-05, "loss": 0.3341, "step": 1440 }, { "grad_norm": 2.983560800552368, "learning_rate": 1.449e-05, "loss": 0.2559, "step": 1450 }, { "grad_norm": 2.797562599182129, "learning_rate": 1.4590000000000001e-05, "loss": 0.3026, "step": 1460 }, { "grad_norm": 3.633791446685791, "learning_rate": 1.4690000000000002e-05, "loss": 0.2529, "step": 1470 }, { "grad_norm": 3.473268508911133, "learning_rate": 1.479e-05, "loss": 0.2561, "step": 1480 }, { "grad_norm": 3.124117612838745, "learning_rate": 1.4890000000000001e-05, "loss": 0.3128, "step": 1490 }, { "grad_norm": 3.148956298828125, "learning_rate": 1.499e-05, "loss": 0.2772, "step": 1500 }, { "grad_norm": 3.9573323726654053, "learning_rate": 1.5090000000000001e-05, "loss": 0.2795, "step": 1510 }, { "grad_norm": 3.009986400604248, "learning_rate": 1.5190000000000002e-05, "loss": 0.2545, "step": 1520 }, { "grad_norm": 2.90029239654541, "learning_rate": 1.529e-05, "loss": 0.2252, "step": 1530 }, { "grad_norm": 3.043224811553955, "learning_rate": 1.539e-05, "loss": 0.2979, "step": 1540 }, { "grad_norm": 3.145740032196045, "learning_rate": 1.5490000000000002e-05, "loss": 0.3157, "step": 1550 }, { "grad_norm": 3.390120506286621, "learning_rate": 1.559e-05, "loss": 0.2936, "step": 1560 }, { "grad_norm": 2.8187272548675537, "learning_rate": 1.569e-05, "loss": 0.219, "step": 1570 }, { "grad_norm": 3.1568965911865234, "learning_rate": 1.5790000000000003e-05, "loss": 0.2268, "step": 1580 }, { "grad_norm": 3.053311586380005, "learning_rate": 1.5890000000000002e-05, "loss": 0.2329, "step": 1590 }, { "grad_norm": 2.989741325378418, "learning_rate": 1.599e-05, "loss": 0.2219, "step": 1600 }, { "grad_norm": 3.5009777545928955, "learning_rate": 1.609e-05, "loss": 0.2342, "step": 1610 }, { "grad_norm": 3.559318780899048, "learning_rate": 1.619e-05, "loss": 0.2515, "step": 1620 }, { "grad_norm": 2.412703514099121, "learning_rate": 1.6289999999999998e-05, "loss": 0.2276, "step": 1630 }, { "grad_norm": 3.7385940551757812, "learning_rate": 1.639e-05, "loss": 0.2511, "step": 1640 }, { "grad_norm": 4.001143932342529, "learning_rate": 1.649e-05, "loss": 0.2742, "step": 1650 }, { "grad_norm": 2.4864678382873535, "learning_rate": 1.659e-05, "loss": 0.2402, "step": 1660 }, { "grad_norm": 2.9319329261779785, "learning_rate": 1.669e-05, "loss": 0.2328, "step": 1670 }, { "grad_norm": 2.970522403717041, "learning_rate": 1.679e-05, "loss": 0.2587, "step": 1680 }, { "grad_norm": 4.011208534240723, "learning_rate": 1.689e-05, "loss": 0.2427, "step": 1690 }, { "grad_norm": 2.873157501220703, "learning_rate": 1.699e-05, "loss": 0.2559, "step": 1700 }, { "grad_norm": 2.1258091926574707, "learning_rate": 1.709e-05, "loss": 0.2098, "step": 1710 }, { "grad_norm": 2.5631399154663086, "learning_rate": 1.719e-05, "loss": 0.2263, "step": 1720 }, { "grad_norm": 3.056694984436035, "learning_rate": 1.7290000000000002e-05, "loss": 0.2312, "step": 1730 }, { "grad_norm": 3.1104767322540283, "learning_rate": 1.739e-05, "loss": 0.2228, "step": 1740 }, { "grad_norm": 2.637129306793213, "learning_rate": 1.749e-05, "loss": 0.2094, "step": 1750 }, { "grad_norm": 3.256718397140503, "learning_rate": 1.759e-05, "loss": 0.2144, "step": 1760 }, { "grad_norm": 3.787895679473877, "learning_rate": 1.7690000000000002e-05, "loss": 0.2323, "step": 1770 }, { "grad_norm": 2.4777700901031494, "learning_rate": 1.779e-05, "loss": 0.2116, "step": 1780 }, { "grad_norm": 2.651296615600586, "learning_rate": 1.789e-05, "loss": 0.201, "step": 1790 }, { "grad_norm": 2.17648983001709, "learning_rate": 1.7990000000000002e-05, "loss": 0.2242, "step": 1800 }, { "grad_norm": 2.8361196517944336, "learning_rate": 1.809e-05, "loss": 0.2135, "step": 1810 }, { "grad_norm": 2.4882349967956543, "learning_rate": 1.819e-05, "loss": 0.2271, "step": 1820 }, { "grad_norm": 3.376070737838745, "learning_rate": 1.8290000000000003e-05, "loss": 0.197, "step": 1830 }, { "grad_norm": 2.627925157546997, "learning_rate": 1.8390000000000002e-05, "loss": 0.1971, "step": 1840 }, { "grad_norm": 2.6930716037750244, "learning_rate": 1.849e-05, "loss": 0.1925, "step": 1850 }, { "grad_norm": 2.138472318649292, "learning_rate": 1.8590000000000003e-05, "loss": 0.2166, "step": 1860 }, { "grad_norm": 2.201000452041626, "learning_rate": 1.8690000000000002e-05, "loss": 0.2175, "step": 1870 }, { "grad_norm": 2.9070539474487305, "learning_rate": 1.879e-05, "loss": 0.1924, "step": 1880 }, { "grad_norm": 3.435210704803467, "learning_rate": 1.8890000000000004e-05, "loss": 0.1913, "step": 1890 }, { "grad_norm": 2.802631139755249, "learning_rate": 1.8990000000000003e-05, "loss": 0.2033, "step": 1900 }, { "grad_norm": 2.768099784851074, "learning_rate": 1.909e-05, "loss": 0.2308, "step": 1910 }, { "grad_norm": 2.432309150695801, "learning_rate": 1.919e-05, "loss": 0.1917, "step": 1920 }, { "grad_norm": 2.13704776763916, "learning_rate": 1.929e-05, "loss": 0.2442, "step": 1930 }, { "grad_norm": 2.7363295555114746, "learning_rate": 1.939e-05, "loss": 0.2098, "step": 1940 }, { "grad_norm": 3.04187273979187, "learning_rate": 1.949e-05, "loss": 0.1851, "step": 1950 }, { "grad_norm": 2.552834987640381, "learning_rate": 1.959e-05, "loss": 0.1894, "step": 1960 }, { "grad_norm": 2.548708438873291, "learning_rate": 1.969e-05, "loss": 0.1732, "step": 1970 }, { "grad_norm": 2.92378568649292, "learning_rate": 1.979e-05, "loss": 0.1609, "step": 1980 }, { "grad_norm": 2.1479456424713135, "learning_rate": 1.989e-05, "loss": 0.1786, "step": 1990 }, { "grad_norm": 3.0328502655029297, "learning_rate": 1.999e-05, "loss": 0.1974, "step": 2000 }, { "grad_norm": 2.2464962005615234, "learning_rate": 2.009e-05, "loss": 0.1647, "step": 2010 }, { "grad_norm": 2.3053643703460693, "learning_rate": 2.019e-05, "loss": 0.1842, "step": 2020 }, { "grad_norm": 2.3950035572052, "learning_rate": 2.029e-05, "loss": 0.2065, "step": 2030 }, { "grad_norm": 2.746720552444458, "learning_rate": 2.039e-05, "loss": 0.1887, "step": 2040 }, { "grad_norm": 2.634298324584961, "learning_rate": 2.0490000000000002e-05, "loss": 0.1648, "step": 2050 }, { "grad_norm": 2.191009998321533, "learning_rate": 2.059e-05, "loss": 0.1876, "step": 2060 }, { "grad_norm": 2.596604347229004, "learning_rate": 2.069e-05, "loss": 0.1769, "step": 2070 }, { "grad_norm": 2.6986732482910156, "learning_rate": 2.0790000000000003e-05, "loss": 0.1852, "step": 2080 }, { "grad_norm": 2.4511022567749023, "learning_rate": 2.089e-05, "loss": 0.1905, "step": 2090 }, { "grad_norm": 2.3509862422943115, "learning_rate": 2.099e-05, "loss": 0.1808, "step": 2100 }, { "grad_norm": 2.9866979122161865, "learning_rate": 2.1090000000000003e-05, "loss": 0.1966, "step": 2110 }, { "grad_norm": 2.0380055904388428, "learning_rate": 2.1190000000000002e-05, "loss": 0.2204, "step": 2120 }, { "grad_norm": 2.6596436500549316, "learning_rate": 2.129e-05, "loss": 0.2175, "step": 2130 }, { "grad_norm": 2.414684772491455, "learning_rate": 2.139e-05, "loss": 0.1827, "step": 2140 }, { "grad_norm": 2.6058809757232666, "learning_rate": 2.1490000000000003e-05, "loss": 0.2101, "step": 2150 }, { "grad_norm": 2.433520555496216, "learning_rate": 2.159e-05, "loss": 0.2138, "step": 2160 }, { "grad_norm": 3.0925207138061523, "learning_rate": 2.169e-05, "loss": 0.1785, "step": 2170 }, { "grad_norm": 2.208778142929077, "learning_rate": 2.1790000000000003e-05, "loss": 0.1924, "step": 2180 }, { "grad_norm": 2.3629233837127686, "learning_rate": 2.1890000000000002e-05, "loss": 0.1659, "step": 2190 }, { "grad_norm": 2.701658248901367, "learning_rate": 2.199e-05, "loss": 0.1489, "step": 2200 }, { "grad_norm": 2.5727269649505615, "learning_rate": 2.2090000000000004e-05, "loss": 0.2207, "step": 2210 }, { "grad_norm": 3.1485540866851807, "learning_rate": 2.219e-05, "loss": 0.2461, "step": 2220 }, { "grad_norm": 2.4928252696990967, "learning_rate": 2.229e-05, "loss": 0.1768, "step": 2230 }, { "grad_norm": 2.1574883460998535, "learning_rate": 2.239e-05, "loss": 0.16, "step": 2240 }, { "grad_norm": 2.549062490463257, "learning_rate": 2.249e-05, "loss": 0.1948, "step": 2250 }, { "grad_norm": 3.008831262588501, "learning_rate": 2.259e-05, "loss": 0.2216, "step": 2260 }, { "grad_norm": 2.843311071395874, "learning_rate": 2.269e-05, "loss": 0.2394, "step": 2270 }, { "grad_norm": 2.3166873455047607, "learning_rate": 2.279e-05, "loss": 0.2341, "step": 2280 }, { "grad_norm": 2.664923906326294, "learning_rate": 2.289e-05, "loss": 0.2004, "step": 2290 }, { "grad_norm": 4.021754741668701, "learning_rate": 2.2990000000000002e-05, "loss": 0.2176, "step": 2300 }, { "grad_norm": 2.923759698867798, "learning_rate": 2.309e-05, "loss": 0.1674, "step": 2310 }, { "grad_norm": 2.486494302749634, "learning_rate": 2.319e-05, "loss": 0.209, "step": 2320 }, { "grad_norm": 3.193664312362671, "learning_rate": 2.3290000000000002e-05, "loss": 0.175, "step": 2330 }, { "grad_norm": 2.339693307876587, "learning_rate": 2.339e-05, "loss": 0.2128, "step": 2340 }, { "grad_norm": 2.012749671936035, "learning_rate": 2.349e-05, "loss": 0.1704, "step": 2350 }, { "grad_norm": 2.650614023208618, "learning_rate": 2.359e-05, "loss": 0.145, "step": 2360 }, { "grad_norm": 2.5421769618988037, "learning_rate": 2.3690000000000002e-05, "loss": 0.1827, "step": 2370 }, { "grad_norm": 2.6003363132476807, "learning_rate": 2.379e-05, "loss": 0.1598, "step": 2380 }, { "grad_norm": 1.8409188985824585, "learning_rate": 2.389e-05, "loss": 0.1667, "step": 2390 }, { "grad_norm": 2.781139850616455, "learning_rate": 2.3990000000000002e-05, "loss": 0.2122, "step": 2400 }, { "grad_norm": 2.572331190109253, "learning_rate": 2.409e-05, "loss": 0.254, "step": 2410 }, { "grad_norm": 3.6820173263549805, "learning_rate": 2.419e-05, "loss": 0.2077, "step": 2420 }, { "grad_norm": 2.9247243404388428, "learning_rate": 2.4290000000000003e-05, "loss": 0.1715, "step": 2430 }, { "grad_norm": 2.673366069793701, "learning_rate": 2.4390000000000002e-05, "loss": 0.1733, "step": 2440 }, { "grad_norm": 2.2434511184692383, "learning_rate": 2.449e-05, "loss": 0.2225, "step": 2450 }, { "grad_norm": 2.1807749271392822, "learning_rate": 2.4590000000000003e-05, "loss": 0.1515, "step": 2460 }, { "grad_norm": 2.0708940029144287, "learning_rate": 2.4690000000000002e-05, "loss": 0.1978, "step": 2470 }, { "grad_norm": 2.3635849952697754, "learning_rate": 2.479e-05, "loss": 0.1394, "step": 2480 }, { "grad_norm": 2.1143858432769775, "learning_rate": 2.489e-05, "loss": 0.1761, "step": 2490 }, { "grad_norm": 2.7531254291534424, "learning_rate": 2.4990000000000003e-05, "loss": 0.1775, "step": 2500 }, { "grad_norm": 2.1499459743499756, "learning_rate": 2.5090000000000002e-05, "loss": 0.186, "step": 2510 }, { "grad_norm": 2.141038417816162, "learning_rate": 2.519e-05, "loss": 0.1749, "step": 2520 }, { "grad_norm": 1.8459936380386353, "learning_rate": 2.5290000000000004e-05, "loss": 0.1509, "step": 2530 }, { "grad_norm": 2.091891288757324, "learning_rate": 2.5390000000000003e-05, "loss": 0.1465, "step": 2540 }, { "grad_norm": 2.407561779022217, "learning_rate": 2.549e-05, "loss": 0.191, "step": 2550 }, { "grad_norm": 2.2205355167388916, "learning_rate": 2.5590000000000004e-05, "loss": 0.1835, "step": 2560 }, { "grad_norm": 2.3623480796813965, "learning_rate": 2.5690000000000003e-05, "loss": 0.2072, "step": 2570 }, { "grad_norm": 2.6856870651245117, "learning_rate": 2.5790000000000002e-05, "loss": 0.1521, "step": 2580 }, { "grad_norm": 2.532717227935791, "learning_rate": 2.5890000000000005e-05, "loss": 0.18, "step": 2590 }, { "grad_norm": 1.6929476261138916, "learning_rate": 2.5990000000000004e-05, "loss": 0.147, "step": 2600 }, { "grad_norm": 2.445797920227051, "learning_rate": 2.6090000000000003e-05, "loss": 0.1626, "step": 2610 }, { "grad_norm": 2.023866653442383, "learning_rate": 2.6190000000000005e-05, "loss": 0.1621, "step": 2620 }, { "grad_norm": 2.2956254482269287, "learning_rate": 2.6290000000000004e-05, "loss": 0.1502, "step": 2630 }, { "grad_norm": 1.8723139762878418, "learning_rate": 2.6390000000000003e-05, "loss": 0.1858, "step": 2640 }, { "grad_norm": 2.1642444133758545, "learning_rate": 2.6490000000000002e-05, "loss": 0.1428, "step": 2650 }, { "grad_norm": 1.608914852142334, "learning_rate": 2.6590000000000005e-05, "loss": 0.1522, "step": 2660 }, { "grad_norm": 1.7523161172866821, "learning_rate": 2.6690000000000004e-05, "loss": 0.1331, "step": 2670 }, { "grad_norm": 2.0015909671783447, "learning_rate": 2.6790000000000003e-05, "loss": 0.1662, "step": 2680 }, { "grad_norm": 1.5250941514968872, "learning_rate": 2.689e-05, "loss": 0.1746, "step": 2690 }, { "grad_norm": 1.7592893838882446, "learning_rate": 2.6989999999999997e-05, "loss": 0.203, "step": 2700 }, { "grad_norm": 2.15091609954834, "learning_rate": 2.709e-05, "loss": 0.1804, "step": 2710 }, { "grad_norm": 2.9020988941192627, "learning_rate": 2.719e-05, "loss": 0.1626, "step": 2720 }, { "grad_norm": 1.7552480697631836, "learning_rate": 2.7289999999999998e-05, "loss": 0.1534, "step": 2730 }, { "grad_norm": 2.1356029510498047, "learning_rate": 2.739e-05, "loss": 0.1562, "step": 2740 }, { "grad_norm": 2.206448554992676, "learning_rate": 2.749e-05, "loss": 0.1432, "step": 2750 }, { "grad_norm": 1.628509759902954, "learning_rate": 2.759e-05, "loss": 0.1611, "step": 2760 }, { "grad_norm": 2.1814422607421875, "learning_rate": 2.769e-05, "loss": 0.1656, "step": 2770 }, { "grad_norm": 1.7282811403274536, "learning_rate": 2.779e-05, "loss": 0.1769, "step": 2780 }, { "grad_norm": 1.4807851314544678, "learning_rate": 2.789e-05, "loss": 0.1341, "step": 2790 }, { "grad_norm": 1.4131330251693726, "learning_rate": 2.7989999999999998e-05, "loss": 0.1899, "step": 2800 }, { "grad_norm": 2.045030355453491, "learning_rate": 2.809e-05, "loss": 0.1648, "step": 2810 }, { "grad_norm": 1.5825222730636597, "learning_rate": 2.819e-05, "loss": 0.1615, "step": 2820 }, { "grad_norm": 1.6193199157714844, "learning_rate": 2.829e-05, "loss": 0.142, "step": 2830 }, { "grad_norm": 2.8306400775909424, "learning_rate": 2.839e-05, "loss": 0.179, "step": 2840 }, { "grad_norm": 3.215482711791992, "learning_rate": 2.849e-05, "loss": 0.1484, "step": 2850 }, { "grad_norm": 2.1931424140930176, "learning_rate": 2.859e-05, "loss": 0.1384, "step": 2860 }, { "grad_norm": 2.4697105884552, "learning_rate": 2.869e-05, "loss": 0.1705, "step": 2870 }, { "grad_norm": 1.9311448335647583, "learning_rate": 2.879e-05, "loss": 0.1616, "step": 2880 }, { "grad_norm": 1.6139051914215088, "learning_rate": 2.889e-05, "loss": 0.2034, "step": 2890 }, { "grad_norm": 1.7346769571304321, "learning_rate": 2.8990000000000002e-05, "loss": 0.1494, "step": 2900 }, { "grad_norm": 1.4380769729614258, "learning_rate": 2.909e-05, "loss": 0.1979, "step": 2910 }, { "grad_norm": 1.5383111238479614, "learning_rate": 2.919e-05, "loss": 0.133, "step": 2920 }, { "grad_norm": 1.6938873529434204, "learning_rate": 2.929e-05, "loss": 0.1689, "step": 2930 }, { "grad_norm": 1.4068087339401245, "learning_rate": 2.939e-05, "loss": 0.1577, "step": 2940 }, { "grad_norm": 1.66395902633667, "learning_rate": 2.949e-05, "loss": 0.1695, "step": 2950 }, { "grad_norm": 1.7688661813735962, "learning_rate": 2.959e-05, "loss": 0.1352, "step": 2960 }, { "grad_norm": 1.9321247339248657, "learning_rate": 2.9690000000000002e-05, "loss": 0.1522, "step": 2970 }, { "grad_norm": 1.8807789087295532, "learning_rate": 2.979e-05, "loss": 0.164, "step": 2980 }, { "grad_norm": 1.7379133701324463, "learning_rate": 2.989e-05, "loss": 0.1572, "step": 2990 }, { "grad_norm": 1.5475672483444214, "learning_rate": 2.9990000000000003e-05, "loss": 0.1512, "step": 3000 }, { "grad_norm": 2.837245225906372, "learning_rate": 3.009e-05, "loss": 0.1476, "step": 3010 }, { "grad_norm": 2.212566614151001, "learning_rate": 3.019e-05, "loss": 0.1441, "step": 3020 }, { "grad_norm": 1.400729775428772, "learning_rate": 3.0290000000000003e-05, "loss": 0.1493, "step": 3030 }, { "grad_norm": 2.399488687515259, "learning_rate": 3.0390000000000002e-05, "loss": 0.1725, "step": 3040 }, { "grad_norm": 1.6538830995559692, "learning_rate": 3.049e-05, "loss": 0.1286, "step": 3050 }, { "grad_norm": 1.3834540843963623, "learning_rate": 3.0590000000000004e-05, "loss": 0.179, "step": 3060 }, { "grad_norm": 2.004058361053467, "learning_rate": 3.069e-05, "loss": 0.158, "step": 3070 }, { "grad_norm": 1.7189937829971313, "learning_rate": 3.079e-05, "loss": 0.1462, "step": 3080 }, { "grad_norm": 2.086301326751709, "learning_rate": 3.0890000000000004e-05, "loss": 0.1446, "step": 3090 }, { "grad_norm": 1.642957329750061, "learning_rate": 3.099e-05, "loss": 0.1707, "step": 3100 }, { "grad_norm": 2.982144355773926, "learning_rate": 3.109e-05, "loss": 0.1933, "step": 3110 }, { "grad_norm": 2.094086170196533, "learning_rate": 3.1190000000000005e-05, "loss": 0.1641, "step": 3120 }, { "grad_norm": 1.5236698389053345, "learning_rate": 3.129e-05, "loss": 0.1436, "step": 3130 }, { "grad_norm": 2.070568084716797, "learning_rate": 3.139e-05, "loss": 0.1421, "step": 3140 }, { "grad_norm": 1.6263670921325684, "learning_rate": 3.1490000000000005e-05, "loss": 0.1386, "step": 3150 }, { "grad_norm": 1.6775598526000977, "learning_rate": 3.159e-05, "loss": 0.167, "step": 3160 }, { "grad_norm": 1.9060088396072388, "learning_rate": 3.169e-05, "loss": 0.1653, "step": 3170 }, { "grad_norm": 1.878879427909851, "learning_rate": 3.1790000000000006e-05, "loss": 0.1826, "step": 3180 }, { "grad_norm": 1.6158952713012695, "learning_rate": 3.189e-05, "loss": 0.1342, "step": 3190 }, { "grad_norm": 1.9469432830810547, "learning_rate": 3.1990000000000004e-05, "loss": 0.1652, "step": 3200 }, { "grad_norm": 1.8993737697601318, "learning_rate": 3.2090000000000006e-05, "loss": 0.1283, "step": 3210 }, { "grad_norm": 1.4363317489624023, "learning_rate": 3.219e-05, "loss": 0.1434, "step": 3220 }, { "grad_norm": 2.6434590816497803, "learning_rate": 3.2290000000000004e-05, "loss": 0.1603, "step": 3230 }, { "grad_norm": 1.2893071174621582, "learning_rate": 3.239000000000001e-05, "loss": 0.1596, "step": 3240 }, { "grad_norm": 1.9149582386016846, "learning_rate": 3.249e-05, "loss": 0.1602, "step": 3250 }, { "grad_norm": 1.5996023416519165, "learning_rate": 3.2590000000000005e-05, "loss": 0.1917, "step": 3260 }, { "grad_norm": 1.6979262828826904, "learning_rate": 3.269000000000001e-05, "loss": 0.1468, "step": 3270 }, { "grad_norm": 1.5993791818618774, "learning_rate": 3.279e-05, "loss": 0.1748, "step": 3280 }, { "grad_norm": 1.941141963005066, "learning_rate": 3.2890000000000005e-05, "loss": 0.1656, "step": 3290 }, { "grad_norm": 1.7107009887695312, "learning_rate": 3.299e-05, "loss": 0.1558, "step": 3300 }, { "grad_norm": 2.060408353805542, "learning_rate": 3.309e-05, "loss": 0.14, "step": 3310 }, { "grad_norm": 1.5489826202392578, "learning_rate": 3.319e-05, "loss": 0.1886, "step": 3320 }, { "grad_norm": 2.0892069339752197, "learning_rate": 3.329e-05, "loss": 0.1381, "step": 3330 }, { "grad_norm": 1.6879154443740845, "learning_rate": 3.339e-05, "loss": 0.1414, "step": 3340 }, { "grad_norm": 1.4951304197311401, "learning_rate": 3.349e-05, "loss": 0.1576, "step": 3350 }, { "grad_norm": 1.8529709577560425, "learning_rate": 3.359e-05, "loss": 0.1399, "step": 3360 }, { "grad_norm": 1.3347580432891846, "learning_rate": 3.369e-05, "loss": 0.1417, "step": 3370 }, { "grad_norm": 1.3337998390197754, "learning_rate": 3.379e-05, "loss": 0.1392, "step": 3380 }, { "grad_norm": 1.4121431112289429, "learning_rate": 3.389e-05, "loss": 0.162, "step": 3390 }, { "grad_norm": 1.5470808744430542, "learning_rate": 3.399e-05, "loss": 0.1476, "step": 3400 }, { "grad_norm": 1.5805673599243164, "learning_rate": 3.409e-05, "loss": 0.1495, "step": 3410 }, { "grad_norm": 2.0562498569488525, "learning_rate": 3.419e-05, "loss": 0.2066, "step": 3420 }, { "grad_norm": 1.3616365194320679, "learning_rate": 3.429e-05, "loss": 0.1449, "step": 3430 }, { "grad_norm": 1.8215268850326538, "learning_rate": 3.439e-05, "loss": 0.1679, "step": 3440 }, { "grad_norm": 1.4218429327011108, "learning_rate": 3.449e-05, "loss": 0.1601, "step": 3450 }, { "grad_norm": 1.759024977684021, "learning_rate": 3.459e-05, "loss": 0.1352, "step": 3460 }, { "grad_norm": 1.6747007369995117, "learning_rate": 3.469e-05, "loss": 0.1745, "step": 3470 }, { "grad_norm": 1.3088247776031494, "learning_rate": 3.479e-05, "loss": 0.1222, "step": 3480 }, { "grad_norm": 2.563368558883667, "learning_rate": 3.489e-05, "loss": 0.1651, "step": 3490 }, { "grad_norm": 2.0220303535461426, "learning_rate": 3.499e-05, "loss": 0.1682, "step": 3500 }, { "grad_norm": 2.7669899463653564, "learning_rate": 3.509e-05, "loss": 0.155, "step": 3510 }, { "grad_norm": 2.782804012298584, "learning_rate": 3.519e-05, "loss": 0.1479, "step": 3520 }, { "grad_norm": 1.615435004234314, "learning_rate": 3.529e-05, "loss": 0.1403, "step": 3530 }, { "grad_norm": 2.3901076316833496, "learning_rate": 3.539e-05, "loss": 0.1674, "step": 3540 }, { "grad_norm": 2.121290922164917, "learning_rate": 3.549e-05, "loss": 0.1877, "step": 3550 }, { "grad_norm": 1.5601614713668823, "learning_rate": 3.559e-05, "loss": 0.1617, "step": 3560 }, { "grad_norm": 1.690065622329712, "learning_rate": 3.569e-05, "loss": 0.1288, "step": 3570 }, { "grad_norm": 1.9170125722885132, "learning_rate": 3.579e-05, "loss": 0.1479, "step": 3580 }, { "grad_norm": 2.2584986686706543, "learning_rate": 3.5890000000000004e-05, "loss": 0.1472, "step": 3590 }, { "grad_norm": 0.9823827743530273, "learning_rate": 3.599e-05, "loss": 0.135, "step": 3600 }, { "grad_norm": 1.8295128345489502, "learning_rate": 3.609e-05, "loss": 0.1556, "step": 3610 }, { "grad_norm": 1.0946015119552612, "learning_rate": 3.6190000000000004e-05, "loss": 0.121, "step": 3620 }, { "grad_norm": 1.239712119102478, "learning_rate": 3.629e-05, "loss": 0.1294, "step": 3630 }, { "grad_norm": 2.567078113555908, "learning_rate": 3.639e-05, "loss": 0.1879, "step": 3640 }, { "grad_norm": 1.51886785030365, "learning_rate": 3.6490000000000005e-05, "loss": 0.1449, "step": 3650 }, { "grad_norm": 2.0216121673583984, "learning_rate": 3.659e-05, "loss": 0.1773, "step": 3660 }, { "grad_norm": 1.7771422863006592, "learning_rate": 3.669e-05, "loss": 0.1967, "step": 3670 }, { "grad_norm": 1.6641573905944824, "learning_rate": 3.6790000000000005e-05, "loss": 0.153, "step": 3680 }, { "grad_norm": 2.1637840270996094, "learning_rate": 3.689e-05, "loss": 0.1961, "step": 3690 }, { "grad_norm": 1.3956027030944824, "learning_rate": 3.699e-05, "loss": 0.1547, "step": 3700 }, { "grad_norm": 1.8394402265548706, "learning_rate": 3.7090000000000006e-05, "loss": 0.1542, "step": 3710 }, { "grad_norm": 1.459605097770691, "learning_rate": 3.719e-05, "loss": 0.1948, "step": 3720 }, { "grad_norm": 1.1745357513427734, "learning_rate": 3.7290000000000004e-05, "loss": 0.1354, "step": 3730 }, { "grad_norm": 1.234100103378296, "learning_rate": 3.739e-05, "loss": 0.1505, "step": 3740 }, { "grad_norm": 1.0653812885284424, "learning_rate": 3.749e-05, "loss": 0.1155, "step": 3750 }, { "grad_norm": 1.4260514974594116, "learning_rate": 3.7590000000000004e-05, "loss": 0.1357, "step": 3760 }, { "grad_norm": 1.567888617515564, "learning_rate": 3.769e-05, "loss": 0.1839, "step": 3770 }, { "grad_norm": 1.5779757499694824, "learning_rate": 3.779e-05, "loss": 0.1299, "step": 3780 }, { "grad_norm": 1.9295934438705444, "learning_rate": 3.7890000000000005e-05, "loss": 0.1517, "step": 3790 }, { "grad_norm": 1.108162522315979, "learning_rate": 3.799e-05, "loss": 0.1349, "step": 3800 }, { "grad_norm": 1.0407048463821411, "learning_rate": 3.809e-05, "loss": 0.1471, "step": 3810 }, { "grad_norm": 1.5243566036224365, "learning_rate": 3.8190000000000005e-05, "loss": 0.182, "step": 3820 }, { "grad_norm": 2.3156135082244873, "learning_rate": 3.829e-05, "loss": 0.167, "step": 3830 }, { "grad_norm": 1.2486193180084229, "learning_rate": 3.8390000000000003e-05, "loss": 0.1599, "step": 3840 }, { "grad_norm": 1.2965894937515259, "learning_rate": 3.8490000000000006e-05, "loss": 0.147, "step": 3850 }, { "grad_norm": 1.4468119144439697, "learning_rate": 3.859e-05, "loss": 0.1357, "step": 3860 }, { "grad_norm": 1.4265955686569214, "learning_rate": 3.8690000000000004e-05, "loss": 0.1441, "step": 3870 }, { "grad_norm": 2.2762644290924072, "learning_rate": 3.8790000000000006e-05, "loss": 0.1398, "step": 3880 }, { "grad_norm": 1.174649715423584, "learning_rate": 3.889e-05, "loss": 0.1357, "step": 3890 }, { "grad_norm": 1.8042694330215454, "learning_rate": 3.8990000000000004e-05, "loss": 0.125, "step": 3900 }, { "grad_norm": 1.3120208978652954, "learning_rate": 3.909000000000001e-05, "loss": 0.1585, "step": 3910 }, { "grad_norm": 1.039735198020935, "learning_rate": 3.919e-05, "loss": 0.159, "step": 3920 }, { "grad_norm": 1.245383381843567, "learning_rate": 3.9290000000000005e-05, "loss": 0.1344, "step": 3930 }, { "grad_norm": 1.5623055696487427, "learning_rate": 3.939e-05, "loss": 0.135, "step": 3940 }, { "grad_norm": 1.2958600521087646, "learning_rate": 3.9489999999999996e-05, "loss": 0.1463, "step": 3950 }, { "grad_norm": 1.2481951713562012, "learning_rate": 3.959e-05, "loss": 0.1626, "step": 3960 }, { "grad_norm": 1.353769063949585, "learning_rate": 3.969e-05, "loss": 0.14, "step": 3970 }, { "grad_norm": 1.471452236175537, "learning_rate": 3.979e-05, "loss": 0.1213, "step": 3980 }, { "grad_norm": 1.7451566457748413, "learning_rate": 3.989e-05, "loss": 0.1793, "step": 3990 }, { "grad_norm": 1.9809147119522095, "learning_rate": 3.999e-05, "loss": 0.1151, "step": 4000 }, { "grad_norm": 1.4987542629241943, "learning_rate": 4.009e-05, "loss": 0.1814, "step": 4010 }, { "grad_norm": 1.3449004888534546, "learning_rate": 4.019e-05, "loss": 0.1942, "step": 4020 }, { "grad_norm": 1.361050009727478, "learning_rate": 4.029e-05, "loss": 0.1308, "step": 4030 }, { "grad_norm": 1.4531471729278564, "learning_rate": 4.039e-05, "loss": 0.1626, "step": 4040 }, { "grad_norm": 1.637951374053955, "learning_rate": 4.049e-05, "loss": 0.1238, "step": 4050 }, { "grad_norm": 1.0609489679336548, "learning_rate": 4.059e-05, "loss": 0.1302, "step": 4060 }, { "grad_norm": 1.3431137800216675, "learning_rate": 4.069e-05, "loss": 0.122, "step": 4070 }, { "grad_norm": 1.3594876527786255, "learning_rate": 4.079e-05, "loss": 0.1602, "step": 4080 }, { "grad_norm": 1.7403985261917114, "learning_rate": 4.089e-05, "loss": 0.1525, "step": 4090 }, { "grad_norm": 1.7292977571487427, "learning_rate": 4.099e-05, "loss": 0.1358, "step": 4100 }, { "grad_norm": 1.2469737529754639, "learning_rate": 4.109e-05, "loss": 0.116, "step": 4110 }, { "grad_norm": 1.0748369693756104, "learning_rate": 4.1190000000000004e-05, "loss": 0.1425, "step": 4120 }, { "grad_norm": 1.2634974718093872, "learning_rate": 4.129e-05, "loss": 0.1234, "step": 4130 }, { "grad_norm": 1.7058881521224976, "learning_rate": 4.139e-05, "loss": 0.1294, "step": 4140 }, { "grad_norm": 1.10245943069458, "learning_rate": 4.1490000000000004e-05, "loss": 0.1299, "step": 4150 }, { "grad_norm": 1.2199500799179077, "learning_rate": 4.159e-05, "loss": 0.1426, "step": 4160 }, { "grad_norm": 1.5116628408432007, "learning_rate": 4.169e-05, "loss": 0.109, "step": 4170 }, { "grad_norm": 1.0903570652008057, "learning_rate": 4.179e-05, "loss": 0.1739, "step": 4180 }, { "grad_norm": 1.6142126321792603, "learning_rate": 4.189e-05, "loss": 0.1295, "step": 4190 }, { "grad_norm": 1.5894975662231445, "learning_rate": 4.199e-05, "loss": 0.1564, "step": 4200 }, { "grad_norm": 1.117369532585144, "learning_rate": 4.209e-05, "loss": 0.1519, "step": 4210 }, { "grad_norm": 1.3390915393829346, "learning_rate": 4.219e-05, "loss": 0.1648, "step": 4220 }, { "grad_norm": 2.012094259262085, "learning_rate": 4.229e-05, "loss": 0.1495, "step": 4230 }, { "grad_norm": 1.569614291191101, "learning_rate": 4.239e-05, "loss": 0.1317, "step": 4240 }, { "grad_norm": 1.1099754571914673, "learning_rate": 4.249e-05, "loss": 0.1402, "step": 4250 }, { "grad_norm": 1.3096421957015991, "learning_rate": 4.2590000000000004e-05, "loss": 0.1276, "step": 4260 }, { "grad_norm": 1.395736575126648, "learning_rate": 4.269e-05, "loss": 0.1161, "step": 4270 }, { "grad_norm": 1.9799820184707642, "learning_rate": 4.279e-05, "loss": 0.1204, "step": 4280 }, { "grad_norm": 1.9513689279556274, "learning_rate": 4.2890000000000004e-05, "loss": 0.1745, "step": 4290 }, { "grad_norm": 1.3805378675460815, "learning_rate": 4.299e-05, "loss": 0.1469, "step": 4300 }, { "grad_norm": 1.6310800313949585, "learning_rate": 4.309e-05, "loss": 0.1329, "step": 4310 }, { "grad_norm": 1.1723161935806274, "learning_rate": 4.3190000000000005e-05, "loss": 0.1266, "step": 4320 }, { "grad_norm": 1.1655337810516357, "learning_rate": 4.329e-05, "loss": 0.1455, "step": 4330 }, { "grad_norm": 1.1848574876785278, "learning_rate": 4.339e-05, "loss": 0.1157, "step": 4340 }, { "grad_norm": 1.5716606378555298, "learning_rate": 4.3490000000000005e-05, "loss": 0.1422, "step": 4350 }, { "grad_norm": 1.0073379278182983, "learning_rate": 4.359e-05, "loss": 0.1518, "step": 4360 }, { "grad_norm": 1.1269174814224243, "learning_rate": 4.3690000000000004e-05, "loss": 0.1364, "step": 4370 }, { "grad_norm": 1.308112621307373, "learning_rate": 4.3790000000000006e-05, "loss": 0.1275, "step": 4380 }, { "grad_norm": 1.1083263158798218, "learning_rate": 4.389e-05, "loss": 0.1433, "step": 4390 }, { "grad_norm": 0.9715483784675598, "learning_rate": 4.3990000000000004e-05, "loss": 0.1065, "step": 4400 }, { "grad_norm": 1.086820363998413, "learning_rate": 4.4090000000000006e-05, "loss": 0.1488, "step": 4410 }, { "grad_norm": 1.0471786260604858, "learning_rate": 4.419e-05, "loss": 0.1279, "step": 4420 }, { "grad_norm": 1.5245816707611084, "learning_rate": 4.4290000000000005e-05, "loss": 0.1458, "step": 4430 }, { "grad_norm": 1.8204214572906494, "learning_rate": 4.439000000000001e-05, "loss": 0.1277, "step": 4440 }, { "grad_norm": 1.1346112489700317, "learning_rate": 4.449e-05, "loss": 0.1371, "step": 4450 }, { "grad_norm": 1.398836374282837, "learning_rate": 4.4590000000000005e-05, "loss": 0.1609, "step": 4460 }, { "grad_norm": 1.731197714805603, "learning_rate": 4.469e-05, "loss": 0.1405, "step": 4470 }, { "grad_norm": 1.3361846208572388, "learning_rate": 4.479e-05, "loss": 0.1375, "step": 4480 }, { "grad_norm": 1.308600902557373, "learning_rate": 4.4890000000000006e-05, "loss": 0.1449, "step": 4490 }, { "grad_norm": 1.0327227115631104, "learning_rate": 4.499e-05, "loss": 0.1613, "step": 4500 }, { "grad_norm": 1.0392991304397583, "learning_rate": 4.5090000000000004e-05, "loss": 0.1288, "step": 4510 }, { "grad_norm": 0.9417699575424194, "learning_rate": 4.5190000000000006e-05, "loss": 0.1158, "step": 4520 }, { "grad_norm": 1.317484974861145, "learning_rate": 4.529e-05, "loss": 0.1283, "step": 4530 }, { "grad_norm": 1.1369163990020752, "learning_rate": 4.5390000000000004e-05, "loss": 0.1143, "step": 4540 }, { "grad_norm": 0.9834320545196533, "learning_rate": 4.549000000000001e-05, "loss": 0.1198, "step": 4550 }, { "grad_norm": 1.1130162477493286, "learning_rate": 4.559e-05, "loss": 0.1432, "step": 4560 }, { "grad_norm": 1.5538477897644043, "learning_rate": 4.569e-05, "loss": 0.1736, "step": 4570 }, { "grad_norm": 1.8129404783248901, "learning_rate": 4.579e-05, "loss": 0.1655, "step": 4580 }, { "grad_norm": 1.2421003580093384, "learning_rate": 4.589e-05, "loss": 0.1156, "step": 4590 }, { "grad_norm": 1.1745284795761108, "learning_rate": 4.599e-05, "loss": 0.1187, "step": 4600 }, { "grad_norm": 1.6860003471374512, "learning_rate": 4.609e-05, "loss": 0.1353, "step": 4610 }, { "grad_norm": 1.161636233329773, "learning_rate": 4.619e-05, "loss": 0.1249, "step": 4620 }, { "grad_norm": 1.0640716552734375, "learning_rate": 4.629e-05, "loss": 0.1129, "step": 4630 }, { "grad_norm": 1.044641137123108, "learning_rate": 4.639e-05, "loss": 0.1201, "step": 4640 }, { "grad_norm": 1.8544557094573975, "learning_rate": 4.649e-05, "loss": 0.1578, "step": 4650 }, { "grad_norm": 1.1540188789367676, "learning_rate": 4.659e-05, "loss": 0.13, "step": 4660 }, { "grad_norm": 1.3568944931030273, "learning_rate": 4.669e-05, "loss": 0.1348, "step": 4670 }, { "grad_norm": 1.3298848867416382, "learning_rate": 4.679e-05, "loss": 0.12, "step": 4680 }, { "grad_norm": 1.1511683464050293, "learning_rate": 4.689e-05, "loss": 0.1322, "step": 4690 }, { "grad_norm": 1.005913496017456, "learning_rate": 4.699e-05, "loss": 0.1329, "step": 4700 }, { "grad_norm": 1.0713367462158203, "learning_rate": 4.709e-05, "loss": 0.1185, "step": 4710 }, { "grad_norm": 1.1374387741088867, "learning_rate": 4.719e-05, "loss": 0.1577, "step": 4720 }, { "grad_norm": 1.2204569578170776, "learning_rate": 4.729e-05, "loss": 0.1435, "step": 4730 }, { "grad_norm": 0.9484572410583496, "learning_rate": 4.739e-05, "loss": 0.1112, "step": 4740 }, { "grad_norm": 1.2621692419052124, "learning_rate": 4.749e-05, "loss": 0.138, "step": 4750 }, { "grad_norm": 1.289969801902771, "learning_rate": 4.7590000000000003e-05, "loss": 0.1283, "step": 4760 }, { "grad_norm": 0.9329190254211426, "learning_rate": 4.769e-05, "loss": 0.1436, "step": 4770 }, { "grad_norm": 1.21550714969635, "learning_rate": 4.779e-05, "loss": 0.1593, "step": 4780 }, { "grad_norm": 1.5632598400115967, "learning_rate": 4.7890000000000004e-05, "loss": 0.1485, "step": 4790 }, { "grad_norm": 0.9465659856796265, "learning_rate": 4.799e-05, "loss": 0.1227, "step": 4800 }, { "grad_norm": 0.9933187365531921, "learning_rate": 4.809e-05, "loss": 0.1278, "step": 4810 }, { "grad_norm": 1.2423193454742432, "learning_rate": 4.8190000000000004e-05, "loss": 0.1316, "step": 4820 }, { "grad_norm": 1.1870195865631104, "learning_rate": 4.829e-05, "loss": 0.186, "step": 4830 }, { "grad_norm": 1.2630159854888916, "learning_rate": 4.839e-05, "loss": 0.1458, "step": 4840 }, { "grad_norm": 1.984775424003601, "learning_rate": 4.8490000000000005e-05, "loss": 0.1288, "step": 4850 }, { "grad_norm": 1.2594784498214722, "learning_rate": 4.859e-05, "loss": 0.1307, "step": 4860 }, { "grad_norm": 1.2532382011413574, "learning_rate": 4.869e-05, "loss": 0.1389, "step": 4870 }, { "grad_norm": 1.1481976509094238, "learning_rate": 4.8790000000000006e-05, "loss": 0.1211, "step": 4880 }, { "grad_norm": 0.9827946424484253, "learning_rate": 4.889e-05, "loss": 0.134, "step": 4890 }, { "grad_norm": 1.695766806602478, "learning_rate": 4.8990000000000004e-05, "loss": 0.1468, "step": 4900 }, { "grad_norm": 1.1141602993011475, "learning_rate": 4.9090000000000006e-05, "loss": 0.1459, "step": 4910 }, { "grad_norm": 1.0967342853546143, "learning_rate": 4.919e-05, "loss": 0.1003, "step": 4920 }, { "grad_norm": 1.0157840251922607, "learning_rate": 4.9290000000000004e-05, "loss": 0.1303, "step": 4930 }, { "grad_norm": 1.1933547258377075, "learning_rate": 4.939e-05, "loss": 0.1093, "step": 4940 }, { "grad_norm": 1.2433017492294312, "learning_rate": 4.949e-05, "loss": 0.1449, "step": 4950 }, { "grad_norm": 1.4360129833221436, "learning_rate": 4.9590000000000005e-05, "loss": 0.1642, "step": 4960 }, { "grad_norm": 1.2117246389389038, "learning_rate": 4.969e-05, "loss": 0.1199, "step": 4970 }, { "grad_norm": 0.9769688248634338, "learning_rate": 4.979e-05, "loss": 0.1241, "step": 4980 }, { "grad_norm": 1.1622202396392822, "learning_rate": 4.9890000000000005e-05, "loss": 0.1389, "step": 4990 }, { "grad_norm": 1.0423609018325806, "learning_rate": 4.999e-05, "loss": 0.1155, "step": 5000 }, { "grad_norm": 1.0224153995513916, "learning_rate": 5.009e-05, "loss": 0.1236, "step": 5010 }, { "grad_norm": 1.036453127861023, "learning_rate": 5.0190000000000006e-05, "loss": 0.1036, "step": 5020 }, { "grad_norm": 1.604305624961853, "learning_rate": 5.029e-05, "loss": 0.1462, "step": 5030 }, { "grad_norm": 0.9695850610733032, "learning_rate": 5.0390000000000004e-05, "loss": 0.1227, "step": 5040 }, { "grad_norm": 0.9313568472862244, "learning_rate": 5.0490000000000006e-05, "loss": 0.1222, "step": 5050 }, { "grad_norm": 1.4889432191848755, "learning_rate": 5.059e-05, "loss": 0.1111, "step": 5060 }, { "grad_norm": 1.4255702495574951, "learning_rate": 5.0690000000000004e-05, "loss": 0.1382, "step": 5070 }, { "grad_norm": 2.1976699829101562, "learning_rate": 5.079000000000001e-05, "loss": 0.1101, "step": 5080 }, { "grad_norm": 1.1300692558288574, "learning_rate": 5.089e-05, "loss": 0.1686, "step": 5090 }, { "grad_norm": 1.2727192640304565, "learning_rate": 5.0990000000000005e-05, "loss": 0.1065, "step": 5100 }, { "grad_norm": 0.8682419061660767, "learning_rate": 5.109000000000001e-05, "loss": 0.1306, "step": 5110 }, { "grad_norm": 0.9997276663780212, "learning_rate": 5.119e-05, "loss": 0.1159, "step": 5120 }, { "grad_norm": 1.6627448797225952, "learning_rate": 5.1290000000000005e-05, "loss": 0.1272, "step": 5130 }, { "grad_norm": 0.9265075922012329, "learning_rate": 5.139000000000001e-05, "loss": 0.1305, "step": 5140 }, { "grad_norm": 0.8790120482444763, "learning_rate": 5.149e-05, "loss": 0.1419, "step": 5150 }, { "grad_norm": 1.1092476844787598, "learning_rate": 5.1590000000000006e-05, "loss": 0.1256, "step": 5160 }, { "grad_norm": 1.4536547660827637, "learning_rate": 5.169000000000001e-05, "loss": 0.1481, "step": 5170 }, { "grad_norm": 1.113873839378357, "learning_rate": 5.1790000000000004e-05, "loss": 0.1014, "step": 5180 }, { "grad_norm": 1.3791991472244263, "learning_rate": 5.1890000000000006e-05, "loss": 0.1413, "step": 5190 }, { "grad_norm": 0.8126775622367859, "learning_rate": 5.199000000000001e-05, "loss": 0.1155, "step": 5200 }, { "grad_norm": 1.0822489261627197, "learning_rate": 5.2090000000000004e-05, "loss": 0.1097, "step": 5210 }, { "grad_norm": 1.1459898948669434, "learning_rate": 5.219000000000001e-05, "loss": 0.1309, "step": 5220 }, { "grad_norm": 1.4461294412612915, "learning_rate": 5.229e-05, "loss": 0.1103, "step": 5230 }, { "grad_norm": 1.065078854560852, "learning_rate": 5.2390000000000005e-05, "loss": 0.1227, "step": 5240 }, { "grad_norm": 1.330037236213684, "learning_rate": 5.249000000000001e-05, "loss": 0.1143, "step": 5250 }, { "grad_norm": 0.9446003437042236, "learning_rate": 5.259e-05, "loss": 0.1353, "step": 5260 }, { "grad_norm": 1.4848238229751587, "learning_rate": 5.2690000000000005e-05, "loss": 0.1302, "step": 5270 }, { "grad_norm": 1.2032711505889893, "learning_rate": 5.279000000000001e-05, "loss": 0.1511, "step": 5280 }, { "grad_norm": 1.5594899654388428, "learning_rate": 5.2890000000000004e-05, "loss": 0.1401, "step": 5290 }, { "grad_norm": 0.8914260864257812, "learning_rate": 5.2990000000000006e-05, "loss": 0.1282, "step": 5300 }, { "grad_norm": 0.9613694548606873, "learning_rate": 5.309000000000001e-05, "loss": 0.1257, "step": 5310 }, { "grad_norm": 1.2240673303604126, "learning_rate": 5.3190000000000004e-05, "loss": 0.1335, "step": 5320 }, { "grad_norm": 1.2454602718353271, "learning_rate": 5.3290000000000006e-05, "loss": 0.1363, "step": 5330 }, { "grad_norm": 1.0408670902252197, "learning_rate": 5.339000000000001e-05, "loss": 0.135, "step": 5340 }, { "grad_norm": 1.2548173666000366, "learning_rate": 5.3490000000000005e-05, "loss": 0.1431, "step": 5350 }, { "grad_norm": 0.7491536736488342, "learning_rate": 5.359000000000001e-05, "loss": 0.1572, "step": 5360 }, { "grad_norm": 1.1726189851760864, "learning_rate": 5.369000000000001e-05, "loss": 0.1542, "step": 5370 }, { "grad_norm": 1.0420390367507935, "learning_rate": 5.3790000000000005e-05, "loss": 0.1309, "step": 5380 }, { "grad_norm": 1.1252254247665405, "learning_rate": 5.389000000000001e-05, "loss": 0.1247, "step": 5390 }, { "grad_norm": 0.9938071370124817, "learning_rate": 5.399000000000001e-05, "loss": 0.1066, "step": 5400 }, { "grad_norm": 1.1762056350708008, "learning_rate": 5.4090000000000006e-05, "loss": 0.1165, "step": 5410 }, { "grad_norm": 1.1115885972976685, "learning_rate": 5.419000000000001e-05, "loss": 0.1268, "step": 5420 }, { "grad_norm": 1.0289325714111328, "learning_rate": 5.429000000000001e-05, "loss": 0.1217, "step": 5430 }, { "grad_norm": 1.1272711753845215, "learning_rate": 5.4390000000000006e-05, "loss": 0.1673, "step": 5440 }, { "grad_norm": 0.9461608529090881, "learning_rate": 5.449000000000001e-05, "loss": 0.1346, "step": 5450 }, { "grad_norm": 1.1741907596588135, "learning_rate": 5.459000000000001e-05, "loss": 0.1282, "step": 5460 }, { "grad_norm": 1.0948703289031982, "learning_rate": 5.469000000000001e-05, "loss": 0.108, "step": 5470 }, { "grad_norm": 0.8320397138595581, "learning_rate": 5.479000000000001e-05, "loss": 0.1284, "step": 5480 }, { "grad_norm": 1.2016342878341675, "learning_rate": 5.4890000000000005e-05, "loss": 0.1153, "step": 5490 }, { "grad_norm": 0.8514189124107361, "learning_rate": 5.499000000000001e-05, "loss": 0.1356, "step": 5500 }, { "grad_norm": 0.7469514608383179, "learning_rate": 5.5089999999999996e-05, "loss": 0.1134, "step": 5510 }, { "grad_norm": 0.9329110980033875, "learning_rate": 5.519e-05, "loss": 0.1229, "step": 5520 }, { "grad_norm": 0.9764323234558105, "learning_rate": 5.5289999999999994e-05, "loss": 0.0935, "step": 5530 }, { "grad_norm": 1.2447097301483154, "learning_rate": 5.5389999999999997e-05, "loss": 0.135, "step": 5540 }, { "grad_norm": 1.0942330360412598, "learning_rate": 5.549e-05, "loss": 0.13, "step": 5550 }, { "grad_norm": 1.1042649745941162, "learning_rate": 5.5589999999999995e-05, "loss": 0.141, "step": 5560 }, { "grad_norm": 0.9237393140792847, "learning_rate": 5.569e-05, "loss": 0.1316, "step": 5570 }, { "grad_norm": 1.136814832687378, "learning_rate": 5.579e-05, "loss": 0.1206, "step": 5580 }, { "grad_norm": 0.9772099852561951, "learning_rate": 5.5889999999999995e-05, "loss": 0.1173, "step": 5590 }, { "grad_norm": 0.8790267705917358, "learning_rate": 5.599e-05, "loss": 0.1327, "step": 5600 }, { "grad_norm": 1.4419783353805542, "learning_rate": 5.609e-05, "loss": 0.1354, "step": 5610 }, { "grad_norm": 1.1133975982666016, "learning_rate": 5.6189999999999996e-05, "loss": 0.1973, "step": 5620 }, { "grad_norm": 1.0263473987579346, "learning_rate": 5.629e-05, "loss": 0.1447, "step": 5630 }, { "grad_norm": 1.4077773094177246, "learning_rate": 5.639e-05, "loss": 0.1422, "step": 5640 }, { "grad_norm": 1.1721744537353516, "learning_rate": 5.6489999999999996e-05, "loss": 0.1214, "step": 5650 }, { "grad_norm": 0.8105905055999756, "learning_rate": 5.659e-05, "loss": 0.1275, "step": 5660 }, { "grad_norm": 1.2203807830810547, "learning_rate": 5.669e-05, "loss": 0.1345, "step": 5670 }, { "grad_norm": 0.9856700301170349, "learning_rate": 5.679e-05, "loss": 0.1479, "step": 5680 }, { "grad_norm": 0.8962683081626892, "learning_rate": 5.689e-05, "loss": 0.1373, "step": 5690 }, { "grad_norm": 1.0705022811889648, "learning_rate": 5.699e-05, "loss": 0.1334, "step": 5700 }, { "grad_norm": 1.4442496299743652, "learning_rate": 5.709e-05, "loss": 0.1122, "step": 5710 }, { "grad_norm": 1.3012189865112305, "learning_rate": 5.719e-05, "loss": 0.1333, "step": 5720 }, { "grad_norm": 0.8447312116622925, "learning_rate": 5.729e-05, "loss": 0.1195, "step": 5730 }, { "grad_norm": 1.7831627130508423, "learning_rate": 5.739e-05, "loss": 0.1204, "step": 5740 }, { "grad_norm": 1.1296782493591309, "learning_rate": 5.749e-05, "loss": 0.1017, "step": 5750 }, { "grad_norm": 1.0778084993362427, "learning_rate": 5.759e-05, "loss": 0.1088, "step": 5760 }, { "grad_norm": 1.2314789295196533, "learning_rate": 5.769e-05, "loss": 0.1357, "step": 5770 }, { "grad_norm": 0.8455005884170532, "learning_rate": 5.779e-05, "loss": 0.1202, "step": 5780 }, { "grad_norm": 0.9773204922676086, "learning_rate": 5.789e-05, "loss": 0.12, "step": 5790 }, { "grad_norm": 0.9113194346427917, "learning_rate": 5.799e-05, "loss": 0.1422, "step": 5800 }, { "grad_norm": 0.9764294624328613, "learning_rate": 5.809e-05, "loss": 0.116, "step": 5810 }, { "grad_norm": 1.2833714485168457, "learning_rate": 5.819e-05, "loss": 0.1163, "step": 5820 }, { "grad_norm": 1.1351208686828613, "learning_rate": 5.829e-05, "loss": 0.1239, "step": 5830 }, { "grad_norm": 0.8059297800064087, "learning_rate": 5.839e-05, "loss": 0.1222, "step": 5840 }, { "grad_norm": 0.9987894892692566, "learning_rate": 5.849e-05, "loss": 0.1014, "step": 5850 }, { "grad_norm": 0.9601706862449646, "learning_rate": 5.859e-05, "loss": 0.096, "step": 5860 }, { "grad_norm": 1.1250666379928589, "learning_rate": 5.869e-05, "loss": 0.1182, "step": 5870 }, { "grad_norm": 1.0579379796981812, "learning_rate": 5.879e-05, "loss": 0.0961, "step": 5880 }, { "grad_norm": 1.0324695110321045, "learning_rate": 5.889e-05, "loss": 0.1187, "step": 5890 }, { "grad_norm": 0.8730902671813965, "learning_rate": 5.899e-05, "loss": 0.1068, "step": 5900 }, { "grad_norm": 0.7905884385108948, "learning_rate": 5.909e-05, "loss": 0.1315, "step": 5910 }, { "grad_norm": 0.7306860685348511, "learning_rate": 5.919e-05, "loss": 0.1055, "step": 5920 }, { "grad_norm": 1.1042479276657104, "learning_rate": 5.929e-05, "loss": 0.1056, "step": 5930 }, { "grad_norm": 0.7627650499343872, "learning_rate": 5.939e-05, "loss": 0.1223, "step": 5940 }, { "grad_norm": 0.8679577112197876, "learning_rate": 5.949e-05, "loss": 0.1093, "step": 5950 }, { "grad_norm": 1.0768318176269531, "learning_rate": 5.9590000000000004e-05, "loss": 0.1238, "step": 5960 }, { "grad_norm": 1.313185214996338, "learning_rate": 5.969e-05, "loss": 0.1211, "step": 5970 }, { "grad_norm": 1.0541478395462036, "learning_rate": 5.979e-05, "loss": 0.1503, "step": 5980 }, { "grad_norm": 0.9547091722488403, "learning_rate": 5.9890000000000004e-05, "loss": 0.1397, "step": 5990 }, { "grad_norm": 0.972357988357544, "learning_rate": 5.999e-05, "loss": 0.1091, "step": 6000 }, { "grad_norm": 1.0835373401641846, "learning_rate": 6.009e-05, "loss": 0.1084, "step": 6010 }, { "grad_norm": 0.9856173992156982, "learning_rate": 6.0190000000000005e-05, "loss": 0.11, "step": 6020 }, { "grad_norm": 0.9679187536239624, "learning_rate": 6.029e-05, "loss": 0.1112, "step": 6030 }, { "grad_norm": 0.9462597966194153, "learning_rate": 6.039e-05, "loss": 0.1141, "step": 6040 }, { "grad_norm": 0.9209213256835938, "learning_rate": 6.0490000000000005e-05, "loss": 0.1321, "step": 6050 }, { "grad_norm": 0.7949337363243103, "learning_rate": 6.059e-05, "loss": 0.1013, "step": 6060 }, { "grad_norm": 0.8888867497444153, "learning_rate": 6.069e-05, "loss": 0.1103, "step": 6070 }, { "grad_norm": 1.0324453115463257, "learning_rate": 6.0790000000000006e-05, "loss": 0.1207, "step": 6080 }, { "grad_norm": 0.8693861961364746, "learning_rate": 6.089e-05, "loss": 0.1315, "step": 6090 }, { "grad_norm": 0.6550295948982239, "learning_rate": 6.0990000000000004e-05, "loss": 0.1176, "step": 6100 }, { "grad_norm": 0.935336709022522, "learning_rate": 6.109e-05, "loss": 0.1328, "step": 6110 }, { "grad_norm": 0.7859647274017334, "learning_rate": 6.119e-05, "loss": 0.1222, "step": 6120 }, { "grad_norm": 0.840678870677948, "learning_rate": 6.129e-05, "loss": 0.1058, "step": 6130 }, { "grad_norm": 1.0316786766052246, "learning_rate": 6.139000000000001e-05, "loss": 0.1023, "step": 6140 }, { "grad_norm": 1.3094611167907715, "learning_rate": 6.149000000000001e-05, "loss": 0.1086, "step": 6150 }, { "grad_norm": 1.0112197399139404, "learning_rate": 6.159e-05, "loss": 0.142, "step": 6160 }, { "grad_norm": 1.0255279541015625, "learning_rate": 6.169e-05, "loss": 0.1246, "step": 6170 }, { "grad_norm": 0.8183356523513794, "learning_rate": 6.179e-05, "loss": 0.1552, "step": 6180 }, { "grad_norm": 0.8438102006912231, "learning_rate": 6.189e-05, "loss": 0.113, "step": 6190 }, { "grad_norm": 1.2710213661193848, "learning_rate": 6.199000000000001e-05, "loss": 0.1113, "step": 6200 }, { "grad_norm": 0.8332725763320923, "learning_rate": 6.209000000000001e-05, "loss": 0.112, "step": 6210 }, { "grad_norm": 0.6341045498847961, "learning_rate": 6.219e-05, "loss": 0.1134, "step": 6220 }, { "grad_norm": 0.8169305324554443, "learning_rate": 6.229e-05, "loss": 0.1048, "step": 6230 }, { "grad_norm": 0.6750615239143372, "learning_rate": 6.239e-05, "loss": 0.0829, "step": 6240 }, { "grad_norm": 0.8523585796356201, "learning_rate": 6.249e-05, "loss": 0.0907, "step": 6250 }, { "grad_norm": 0.8387538194656372, "learning_rate": 6.259000000000001e-05, "loss": 0.1046, "step": 6260 }, { "grad_norm": 1.1740690469741821, "learning_rate": 6.269e-05, "loss": 0.1036, "step": 6270 }, { "grad_norm": 0.7315089702606201, "learning_rate": 6.279e-05, "loss": 0.1245, "step": 6280 }, { "grad_norm": 0.8466883897781372, "learning_rate": 6.289e-05, "loss": 0.1063, "step": 6290 }, { "grad_norm": 0.7435682415962219, "learning_rate": 6.299e-05, "loss": 0.1014, "step": 6300 }, { "grad_norm": 0.82485431432724, "learning_rate": 6.309000000000001e-05, "loss": 0.0967, "step": 6310 }, { "grad_norm": 1.01123046875, "learning_rate": 6.319000000000001e-05, "loss": 0.1301, "step": 6320 }, { "grad_norm": 1.0256685018539429, "learning_rate": 6.329e-05, "loss": 0.1037, "step": 6330 }, { "grad_norm": 0.7276252508163452, "learning_rate": 6.339e-05, "loss": 0.1012, "step": 6340 }, { "grad_norm": 0.8903923630714417, "learning_rate": 6.349e-05, "loss": 0.1203, "step": 6350 }, { "grad_norm": 0.8870490789413452, "learning_rate": 6.359e-05, "loss": 0.1277, "step": 6360 }, { "grad_norm": 0.9292553067207336, "learning_rate": 6.369000000000001e-05, "loss": 0.1247, "step": 6370 }, { "grad_norm": 0.8701168298721313, "learning_rate": 6.379000000000001e-05, "loss": 0.1287, "step": 6380 }, { "grad_norm": 0.9563761949539185, "learning_rate": 6.389e-05, "loss": 0.1053, "step": 6390 }, { "grad_norm": 0.7752581238746643, "learning_rate": 6.399e-05, "loss": 0.1297, "step": 6400 }, { "grad_norm": 0.893877387046814, "learning_rate": 6.409e-05, "loss": 0.1004, "step": 6410 }, { "grad_norm": 1.1340187788009644, "learning_rate": 6.419000000000001e-05, "loss": 0.0984, "step": 6420 }, { "grad_norm": 1.3698877096176147, "learning_rate": 6.429000000000001e-05, "loss": 0.1206, "step": 6430 }, { "grad_norm": 0.7787458896636963, "learning_rate": 6.439000000000001e-05, "loss": 0.1444, "step": 6440 }, { "grad_norm": 1.007850170135498, "learning_rate": 6.449e-05, "loss": 0.1466, "step": 6450 }, { "grad_norm": 0.731299102306366, "learning_rate": 6.459e-05, "loss": 0.1059, "step": 6460 }, { "grad_norm": 1.2477169036865234, "learning_rate": 6.469e-05, "loss": 0.1495, "step": 6470 }, { "grad_norm": 1.2738208770751953, "learning_rate": 6.479000000000001e-05, "loss": 0.1033, "step": 6480 }, { "grad_norm": 0.9361453056335449, "learning_rate": 6.489000000000001e-05, "loss": 0.1032, "step": 6490 }, { "grad_norm": 0.7055419087409973, "learning_rate": 6.499000000000001e-05, "loss": 0.149, "step": 6500 }, { "grad_norm": 0.9577810168266296, "learning_rate": 6.509e-05, "loss": 0.1137, "step": 6510 }, { "grad_norm": 1.859988808631897, "learning_rate": 6.519e-05, "loss": 0.1088, "step": 6520 }, { "grad_norm": 0.8579108715057373, "learning_rate": 6.529e-05, "loss": 0.1149, "step": 6530 }, { "grad_norm": 1.088701844215393, "learning_rate": 6.539000000000001e-05, "loss": 0.1302, "step": 6540 }, { "grad_norm": 1.4168336391448975, "learning_rate": 6.549000000000001e-05, "loss": 0.1206, "step": 6550 }, { "grad_norm": 0.8893941044807434, "learning_rate": 6.559e-05, "loss": 0.0938, "step": 6560 }, { "grad_norm": 1.4061110019683838, "learning_rate": 6.569e-05, "loss": 0.1126, "step": 6570 }, { "grad_norm": 0.7989781498908997, "learning_rate": 6.579e-05, "loss": 0.0974, "step": 6580 }, { "grad_norm": 0.7768872976303101, "learning_rate": 6.589000000000001e-05, "loss": 0.1064, "step": 6590 }, { "grad_norm": 0.9062947034835815, "learning_rate": 6.599000000000001e-05, "loss": 0.0895, "step": 6600 }, { "grad_norm": 0.9719805717468262, "learning_rate": 6.609000000000001e-05, "loss": 0.1398, "step": 6610 }, { "grad_norm": 0.7126299738883972, "learning_rate": 6.619e-05, "loss": 0.1029, "step": 6620 }, { "grad_norm": 1.0261019468307495, "learning_rate": 6.629e-05, "loss": 0.1323, "step": 6630 }, { "grad_norm": 0.6901127696037292, "learning_rate": 6.639e-05, "loss": 0.1502, "step": 6640 }, { "grad_norm": 1.1253154277801514, "learning_rate": 6.649000000000001e-05, "loss": 0.1167, "step": 6650 }, { "grad_norm": 0.8279752135276794, "learning_rate": 6.659000000000001e-05, "loss": 0.1082, "step": 6660 }, { "grad_norm": 0.7361634969711304, "learning_rate": 6.669000000000001e-05, "loss": 0.1248, "step": 6670 }, { "grad_norm": 2.4363372325897217, "learning_rate": 6.679e-05, "loss": 0.1157, "step": 6680 }, { "grad_norm": 0.7383310198783875, "learning_rate": 6.689e-05, "loss": 0.1175, "step": 6690 }, { "grad_norm": 1.0392268896102905, "learning_rate": 6.699000000000001e-05, "loss": 0.1119, "step": 6700 }, { "grad_norm": 0.8779187798500061, "learning_rate": 6.709000000000001e-05, "loss": 0.1003, "step": 6710 }, { "grad_norm": 0.8005078434944153, "learning_rate": 6.719000000000001e-05, "loss": 0.1246, "step": 6720 }, { "grad_norm": 0.8823389410972595, "learning_rate": 6.729000000000001e-05, "loss": 0.0946, "step": 6730 }, { "grad_norm": 0.8538469076156616, "learning_rate": 6.739e-05, "loss": 0.1093, "step": 6740 }, { "grad_norm": 1.107082486152649, "learning_rate": 6.749e-05, "loss": 0.1464, "step": 6750 }, { "grad_norm": 0.6935389637947083, "learning_rate": 6.759e-05, "loss": 0.1687, "step": 6760 }, { "grad_norm": 0.8310034871101379, "learning_rate": 6.769e-05, "loss": 0.1044, "step": 6770 }, { "grad_norm": 0.5981224775314331, "learning_rate": 6.779e-05, "loss": 0.097, "step": 6780 }, { "grad_norm": 0.8812510967254639, "learning_rate": 6.789e-05, "loss": 0.1346, "step": 6790 }, { "grad_norm": 0.8639986515045166, "learning_rate": 6.799e-05, "loss": 0.1161, "step": 6800 }, { "grad_norm": 1.0980374813079834, "learning_rate": 6.809e-05, "loss": 0.124, "step": 6810 }, { "grad_norm": 0.9929438233375549, "learning_rate": 6.819e-05, "loss": 0.1055, "step": 6820 }, { "grad_norm": 0.6213039755821228, "learning_rate": 6.829e-05, "loss": 0.1078, "step": 6830 }, { "grad_norm": 0.9570059180259705, "learning_rate": 6.839e-05, "loss": 0.1094, "step": 6840 }, { "grad_norm": 0.7508115172386169, "learning_rate": 6.849e-05, "loss": 0.1046, "step": 6850 }, { "grad_norm": 1.0042669773101807, "learning_rate": 6.858999999999999e-05, "loss": 0.1194, "step": 6860 }, { "grad_norm": 1.2766649723052979, "learning_rate": 6.869e-05, "loss": 0.1125, "step": 6870 }, { "grad_norm": 0.9548705816268921, "learning_rate": 6.879e-05, "loss": 0.1315, "step": 6880 }, { "grad_norm": 0.8312375545501709, "learning_rate": 6.889e-05, "loss": 0.1111, "step": 6890 }, { "grad_norm": 1.0502053499221802, "learning_rate": 6.899e-05, "loss": 0.1261, "step": 6900 }, { "grad_norm": 0.9631084203720093, "learning_rate": 6.909e-05, "loss": 0.0999, "step": 6910 }, { "grad_norm": 0.9157575368881226, "learning_rate": 6.918999999999999e-05, "loss": 0.0971, "step": 6920 }, { "grad_norm": 1.0381141901016235, "learning_rate": 6.929e-05, "loss": 0.1302, "step": 6930 }, { "grad_norm": 0.6944761276245117, "learning_rate": 6.939e-05, "loss": 0.1196, "step": 6940 }, { "grad_norm": 0.955314040184021, "learning_rate": 6.949e-05, "loss": 0.1088, "step": 6950 }, { "grad_norm": 0.8364238739013672, "learning_rate": 6.959e-05, "loss": 0.1075, "step": 6960 }, { "grad_norm": 0.8688331246376038, "learning_rate": 6.969e-05, "loss": 0.1074, "step": 6970 }, { "grad_norm": 0.8748021125793457, "learning_rate": 6.979e-05, "loss": 0.1194, "step": 6980 }, { "grad_norm": 0.9420182704925537, "learning_rate": 6.989e-05, "loss": 0.1123, "step": 6990 }, { "grad_norm": 1.0460361242294312, "learning_rate": 6.999e-05, "loss": 0.096, "step": 7000 }, { "grad_norm": 0.8880594372749329, "learning_rate": 7.009e-05, "loss": 0.1207, "step": 7010 }, { "grad_norm": 0.810185968875885, "learning_rate": 7.019e-05, "loss": 0.1119, "step": 7020 }, { "grad_norm": 0.8149952292442322, "learning_rate": 7.029e-05, "loss": 0.1005, "step": 7030 }, { "grad_norm": 0.8606831431388855, "learning_rate": 7.039e-05, "loss": 0.1272, "step": 7040 }, { "grad_norm": 0.9269838929176331, "learning_rate": 7.049e-05, "loss": 0.1105, "step": 7050 }, { "grad_norm": 1.0459978580474854, "learning_rate": 7.059e-05, "loss": 0.1159, "step": 7060 }, { "grad_norm": 0.8293987512588501, "learning_rate": 7.069e-05, "loss": 0.1072, "step": 7070 }, { "grad_norm": 0.7235618233680725, "learning_rate": 7.079e-05, "loss": 0.1144, "step": 7080 }, { "grad_norm": 0.8999929428100586, "learning_rate": 7.089000000000001e-05, "loss": 0.1128, "step": 7090 }, { "grad_norm": 0.8897308707237244, "learning_rate": 7.099e-05, "loss": 0.1203, "step": 7100 }, { "grad_norm": 0.6824089288711548, "learning_rate": 7.109e-05, "loss": 0.092, "step": 7110 }, { "grad_norm": 0.7877505421638489, "learning_rate": 7.119e-05, "loss": 0.0844, "step": 7120 }, { "grad_norm": 0.8141728043556213, "learning_rate": 7.129e-05, "loss": 0.1092, "step": 7130 }, { "grad_norm": 0.7298322319984436, "learning_rate": 7.139e-05, "loss": 0.1167, "step": 7140 }, { "grad_norm": 0.8018761277198792, "learning_rate": 7.149e-05, "loss": 0.1361, "step": 7150 }, { "grad_norm": 0.856553852558136, "learning_rate": 7.159e-05, "loss": 0.1186, "step": 7160 }, { "grad_norm": 0.9508381485939026, "learning_rate": 7.169e-05, "loss": 0.1354, "step": 7170 }, { "grad_norm": 0.8731581568717957, "learning_rate": 7.179e-05, "loss": 0.1099, "step": 7180 }, { "grad_norm": 0.8028404116630554, "learning_rate": 7.189e-05, "loss": 0.1055, "step": 7190 }, { "grad_norm": 0.6257680058479309, "learning_rate": 7.199000000000001e-05, "loss": 0.1032, "step": 7200 }, { "grad_norm": 0.9979420304298401, "learning_rate": 7.209e-05, "loss": 0.12, "step": 7210 }, { "grad_norm": 0.9263043999671936, "learning_rate": 7.219e-05, "loss": 0.1048, "step": 7220 }, { "grad_norm": 0.9590533375740051, "learning_rate": 7.229e-05, "loss": 0.104, "step": 7230 }, { "grad_norm": 0.7663519382476807, "learning_rate": 7.239e-05, "loss": 0.1137, "step": 7240 }, { "grad_norm": 0.7693989276885986, "learning_rate": 7.249e-05, "loss": 0.0951, "step": 7250 }, { "grad_norm": 0.6983858942985535, "learning_rate": 7.259000000000001e-05, "loss": 0.0915, "step": 7260 }, { "grad_norm": 0.9105063676834106, "learning_rate": 7.269e-05, "loss": 0.1088, "step": 7270 }, { "grad_norm": 1.118403673171997, "learning_rate": 7.279e-05, "loss": 0.1077, "step": 7280 }, { "grad_norm": 1.2539007663726807, "learning_rate": 7.289e-05, "loss": 0.1067, "step": 7290 }, { "grad_norm": 0.9875116348266602, "learning_rate": 7.299e-05, "loss": 0.1141, "step": 7300 }, { "grad_norm": 1.2972279787063599, "learning_rate": 7.309e-05, "loss": 0.1302, "step": 7310 }, { "grad_norm": 1.1812101602554321, "learning_rate": 7.319000000000001e-05, "loss": 0.0969, "step": 7320 }, { "grad_norm": 0.7690911889076233, "learning_rate": 7.329e-05, "loss": 0.1228, "step": 7330 }, { "grad_norm": 1.0311437845230103, "learning_rate": 7.339e-05, "loss": 0.1207, "step": 7340 }, { "grad_norm": 1.0403558015823364, "learning_rate": 7.349e-05, "loss": 0.1208, "step": 7350 }, { "grad_norm": 0.7895894050598145, "learning_rate": 7.359e-05, "loss": 0.1263, "step": 7360 }, { "grad_norm": 0.9845911860466003, "learning_rate": 7.369000000000001e-05, "loss": 0.114, "step": 7370 }, { "grad_norm": 0.721468985080719, "learning_rate": 7.379000000000001e-05, "loss": 0.1083, "step": 7380 }, { "grad_norm": 1.0693552494049072, "learning_rate": 7.389e-05, "loss": 0.1278, "step": 7390 }, { "grad_norm": 0.5564489960670471, "learning_rate": 7.399e-05, "loss": 0.1008, "step": 7400 }, { "grad_norm": 0.7150811553001404, "learning_rate": 7.409e-05, "loss": 0.1115, "step": 7410 }, { "grad_norm": 1.1781234741210938, "learning_rate": 7.419e-05, "loss": 0.1183, "step": 7420 }, { "grad_norm": 0.7744550704956055, "learning_rate": 7.429000000000001e-05, "loss": 0.1118, "step": 7430 }, { "grad_norm": 0.6486948132514954, "learning_rate": 7.439e-05, "loss": 0.1142, "step": 7440 }, { "grad_norm": 0.7652844786643982, "learning_rate": 7.449e-05, "loss": 0.1029, "step": 7450 }, { "grad_norm": 0.6150469779968262, "learning_rate": 7.459e-05, "loss": 0.088, "step": 7460 }, { "grad_norm": 0.8029166460037231, "learning_rate": 7.469e-05, "loss": 0.1085, "step": 7470 }, { "grad_norm": 0.7685571908950806, "learning_rate": 7.479000000000001e-05, "loss": 0.0939, "step": 7480 }, { "grad_norm": 0.9851720929145813, "learning_rate": 7.489000000000001e-05, "loss": 0.1323, "step": 7490 }, { "grad_norm": 0.8039880990982056, "learning_rate": 7.499e-05, "loss": 0.1361, "step": 7500 }, { "grad_norm": 1.1744712591171265, "learning_rate": 7.509e-05, "loss": 0.1152, "step": 7510 }, { "grad_norm": 0.7799496054649353, "learning_rate": 7.519e-05, "loss": 0.097, "step": 7520 }, { "grad_norm": 0.6255172491073608, "learning_rate": 7.529e-05, "loss": 0.1152, "step": 7530 }, { "grad_norm": 1.1555200815200806, "learning_rate": 7.539000000000001e-05, "loss": 0.1377, "step": 7540 }, { "grad_norm": 0.8543136715888977, "learning_rate": 7.549000000000001e-05, "loss": 0.1228, "step": 7550 }, { "grad_norm": 0.7149409055709839, "learning_rate": 7.559e-05, "loss": 0.1119, "step": 7560 }, { "grad_norm": 0.9120664596557617, "learning_rate": 7.569e-05, "loss": 0.1038, "step": 7570 }, { "grad_norm": 1.3391811847686768, "learning_rate": 7.579e-05, "loss": 0.0952, "step": 7580 }, { "grad_norm": 0.7847825288772583, "learning_rate": 7.589000000000001e-05, "loss": 0.1336, "step": 7590 }, { "grad_norm": 0.7908979654312134, "learning_rate": 7.599000000000001e-05, "loss": 0.1317, "step": 7600 }, { "grad_norm": 0.7850469350814819, "learning_rate": 7.609000000000001e-05, "loss": 0.1261, "step": 7610 }, { "grad_norm": 0.9408200979232788, "learning_rate": 7.619e-05, "loss": 0.0971, "step": 7620 }, { "grad_norm": 1.0297203063964844, "learning_rate": 7.629e-05, "loss": 0.1083, "step": 7630 }, { "grad_norm": 0.659436047077179, "learning_rate": 7.639e-05, "loss": 0.1147, "step": 7640 }, { "grad_norm": 0.7313481569290161, "learning_rate": 7.649000000000001e-05, "loss": 0.1122, "step": 7650 }, { "grad_norm": 0.7862698435783386, "learning_rate": 7.659000000000001e-05, "loss": 0.107, "step": 7660 }, { "grad_norm": 0.8498318195343018, "learning_rate": 7.669000000000001e-05, "loss": 0.0895, "step": 7670 }, { "grad_norm": 0.6571590304374695, "learning_rate": 7.679e-05, "loss": 0.1106, "step": 7680 }, { "grad_norm": 0.7642963528633118, "learning_rate": 7.689e-05, "loss": 0.1288, "step": 7690 }, { "grad_norm": 1.0183095932006836, "learning_rate": 7.699e-05, "loss": 0.1195, "step": 7700 }, { "grad_norm": 0.9381212592124939, "learning_rate": 7.709000000000001e-05, "loss": 0.0974, "step": 7710 }, { "grad_norm": 0.8611446022987366, "learning_rate": 7.719000000000001e-05, "loss": 0.1159, "step": 7720 }, { "grad_norm": 0.7527132034301758, "learning_rate": 7.729e-05, "loss": 0.118, "step": 7730 }, { "grad_norm": 0.7212865948677063, "learning_rate": 7.739e-05, "loss": 0.0979, "step": 7740 }, { "grad_norm": 0.7442691326141357, "learning_rate": 7.749e-05, "loss": 0.1125, "step": 7750 }, { "grad_norm": 0.8924878835678101, "learning_rate": 7.759000000000001e-05, "loss": 0.0947, "step": 7760 }, { "grad_norm": 0.7761420011520386, "learning_rate": 7.769000000000001e-05, "loss": 0.1297, "step": 7770 }, { "grad_norm": 0.6044732928276062, "learning_rate": 7.779000000000001e-05, "loss": 0.098, "step": 7780 }, { "grad_norm": 0.7506924867630005, "learning_rate": 7.789e-05, "loss": 0.108, "step": 7790 }, { "grad_norm": 0.7867915034294128, "learning_rate": 7.799e-05, "loss": 0.1068, "step": 7800 }, { "grad_norm": 0.7856225967407227, "learning_rate": 7.809e-05, "loss": 0.1065, "step": 7810 }, { "grad_norm": 0.9217194318771362, "learning_rate": 7.819000000000001e-05, "loss": 0.1034, "step": 7820 }, { "grad_norm": 0.7862571477890015, "learning_rate": 7.829000000000001e-05, "loss": 0.109, "step": 7830 }, { "grad_norm": 1.0002670288085938, "learning_rate": 7.839000000000001e-05, "loss": 0.1205, "step": 7840 }, { "grad_norm": 0.8491230010986328, "learning_rate": 7.849e-05, "loss": 0.1198, "step": 7850 }, { "grad_norm": 0.6717486381530762, "learning_rate": 7.859e-05, "loss": 0.0916, "step": 7860 }, { "grad_norm": 1.3053429126739502, "learning_rate": 7.869000000000001e-05, "loss": 0.1337, "step": 7870 }, { "grad_norm": 1.2145497798919678, "learning_rate": 7.879000000000001e-05, "loss": 0.1334, "step": 7880 }, { "grad_norm": 0.8512216806411743, "learning_rate": 7.889000000000001e-05, "loss": 0.1137, "step": 7890 }, { "grad_norm": 0.896957516670227, "learning_rate": 7.899000000000001e-05, "loss": 0.0993, "step": 7900 }, { "grad_norm": 0.9095954298973083, "learning_rate": 7.909e-05, "loss": 0.1272, "step": 7910 }, { "grad_norm": 1.0546730756759644, "learning_rate": 7.919e-05, "loss": 0.1055, "step": 7920 }, { "grad_norm": 0.8766406774520874, "learning_rate": 7.929000000000001e-05, "loss": 0.0886, "step": 7930 }, { "grad_norm": 0.8158206343650818, "learning_rate": 7.939000000000001e-05, "loss": 0.1123, "step": 7940 }, { "grad_norm": 0.6817150712013245, "learning_rate": 7.949000000000001e-05, "loss": 0.1315, "step": 7950 }, { "grad_norm": 0.6552171111106873, "learning_rate": 7.959000000000002e-05, "loss": 0.0912, "step": 7960 }, { "grad_norm": 0.8243007659912109, "learning_rate": 7.969e-05, "loss": 0.1207, "step": 7970 }, { "grad_norm": 1.1980174779891968, "learning_rate": 7.979000000000001e-05, "loss": 0.1072, "step": 7980 }, { "grad_norm": 0.8345234990119934, "learning_rate": 7.989000000000001e-05, "loss": 0.1248, "step": 7990 }, { "grad_norm": 1.2211310863494873, "learning_rate": 7.999000000000001e-05, "loss": 0.1181, "step": 8000 }, { "grad_norm": 0.6381802558898926, "learning_rate": 8.009e-05, "loss": 0.0871, "step": 8010 }, { "grad_norm": 0.8846755623817444, "learning_rate": 8.019e-05, "loss": 0.1156, "step": 8020 }, { "grad_norm": 1.4698905944824219, "learning_rate": 8.028999999999999e-05, "loss": 0.1242, "step": 8030 }, { "grad_norm": 1.040024757385254, "learning_rate": 8.039e-05, "loss": 0.1072, "step": 8040 }, { "grad_norm": 0.8775899410247803, "learning_rate": 8.049e-05, "loss": 0.1026, "step": 8050 }, { "grad_norm": 0.934634268283844, "learning_rate": 8.059e-05, "loss": 0.134, "step": 8060 }, { "grad_norm": 0.8075314164161682, "learning_rate": 8.069e-05, "loss": 0.0944, "step": 8070 }, { "grad_norm": 0.6065346002578735, "learning_rate": 8.079e-05, "loss": 0.1098, "step": 8080 }, { "grad_norm": 0.6221248507499695, "learning_rate": 8.088999999999999e-05, "loss": 0.1264, "step": 8090 }, { "grad_norm": 0.9363215565681458, "learning_rate": 8.099e-05, "loss": 0.1154, "step": 8100 }, { "grad_norm": 1.238507628440857, "learning_rate": 8.109e-05, "loss": 0.1073, "step": 8110 }, { "grad_norm": 0.8735936880111694, "learning_rate": 8.119e-05, "loss": 0.1122, "step": 8120 }, { "grad_norm": 0.7114894390106201, "learning_rate": 8.129e-05, "loss": 0.1078, "step": 8130 }, { "grad_norm": 0.8553372025489807, "learning_rate": 8.139e-05, "loss": 0.0944, "step": 8140 }, { "grad_norm": 0.8654554486274719, "learning_rate": 8.149e-05, "loss": 0.1211, "step": 8150 }, { "grad_norm": 1.0221205949783325, "learning_rate": 8.159e-05, "loss": 0.1044, "step": 8160 }, { "grad_norm": 0.9488891363143921, "learning_rate": 8.169e-05, "loss": 0.1, "step": 8170 }, { "grad_norm": 0.9532612562179565, "learning_rate": 8.179e-05, "loss": 0.1151, "step": 8180 }, { "grad_norm": 0.9754955172538757, "learning_rate": 8.189e-05, "loss": 0.0953, "step": 8190 }, { "grad_norm": 1.4063307046890259, "learning_rate": 8.199e-05, "loss": 0.1022, "step": 8200 }, { "grad_norm": 0.8412924408912659, "learning_rate": 8.209e-05, "loss": 0.1016, "step": 8210 }, { "grad_norm": 0.9507865905761719, "learning_rate": 8.219e-05, "loss": 0.1099, "step": 8220 }, { "grad_norm": 0.9438803195953369, "learning_rate": 8.229e-05, "loss": 0.1321, "step": 8230 }, { "grad_norm": 1.715269923210144, "learning_rate": 8.239e-05, "loss": 0.1457, "step": 8240 }, { "grad_norm": 0.6977145671844482, "learning_rate": 8.249e-05, "loss": 0.1084, "step": 8250 }, { "grad_norm": 0.6366567015647888, "learning_rate": 8.259000000000001e-05, "loss": 0.1041, "step": 8260 }, { "grad_norm": 1.2724367380142212, "learning_rate": 8.269e-05, "loss": 0.1188, "step": 8270 }, { "grad_norm": 0.8958494663238525, "learning_rate": 8.279e-05, "loss": 0.1052, "step": 8280 }, { "grad_norm": 0.8358458876609802, "learning_rate": 8.289e-05, "loss": 0.1223, "step": 8290 }, { "grad_norm": 0.8329949975013733, "learning_rate": 8.299e-05, "loss": 0.1354, "step": 8300 }, { "grad_norm": 1.1357029676437378, "learning_rate": 8.309e-05, "loss": 0.1153, "step": 8310 }, { "grad_norm": 0.9239726662635803, "learning_rate": 8.319e-05, "loss": 0.1047, "step": 8320 }, { "grad_norm": 1.6738183498382568, "learning_rate": 8.329e-05, "loss": 0.1038, "step": 8330 }, { "grad_norm": 0.7412547469139099, "learning_rate": 8.339e-05, "loss": 0.0969, "step": 8340 }, { "grad_norm": 0.8286996483802795, "learning_rate": 8.349e-05, "loss": 0.1065, "step": 8350 }, { "grad_norm": 0.7647151350975037, "learning_rate": 8.359e-05, "loss": 0.1021, "step": 8360 }, { "grad_norm": 0.6403796672821045, "learning_rate": 8.369000000000001e-05, "loss": 0.1148, "step": 8370 }, { "grad_norm": 1.0744503736495972, "learning_rate": 8.379e-05, "loss": 0.1007, "step": 8380 }, { "grad_norm": 0.8300666809082031, "learning_rate": 8.389e-05, "loss": 0.0952, "step": 8390 }, { "grad_norm": 0.6405826807022095, "learning_rate": 8.399e-05, "loss": 0.1185, "step": 8400 }, { "grad_norm": 0.9524831175804138, "learning_rate": 8.409e-05, "loss": 0.1109, "step": 8410 }, { "grad_norm": 0.755047082901001, "learning_rate": 8.419e-05, "loss": 0.1242, "step": 8420 }, { "grad_norm": 0.8794195652008057, "learning_rate": 8.429000000000001e-05, "loss": 0.1148, "step": 8430 }, { "grad_norm": 0.6230852603912354, "learning_rate": 8.439e-05, "loss": 0.1287, "step": 8440 }, { "grad_norm": 0.8849795460700989, "learning_rate": 8.449e-05, "loss": 0.1282, "step": 8450 }, { "grad_norm": 0.6673727035522461, "learning_rate": 8.459e-05, "loss": 0.1321, "step": 8460 }, { "grad_norm": 0.7113852500915527, "learning_rate": 8.469e-05, "loss": 0.1256, "step": 8470 }, { "grad_norm": 0.6869152188301086, "learning_rate": 8.479e-05, "loss": 0.1117, "step": 8480 }, { "grad_norm": 1.0796656608581543, "learning_rate": 8.489000000000001e-05, "loss": 0.1424, "step": 8490 }, { "grad_norm": 0.6411420106887817, "learning_rate": 8.499e-05, "loss": 0.1202, "step": 8500 }, { "grad_norm": 0.9099472165107727, "learning_rate": 8.509e-05, "loss": 0.1151, "step": 8510 }, { "grad_norm": 0.8348897695541382, "learning_rate": 8.519e-05, "loss": 0.1186, "step": 8520 }, { "grad_norm": 1.8735721111297607, "learning_rate": 8.529e-05, "loss": 0.1218, "step": 8530 }, { "grad_norm": 0.6638465523719788, "learning_rate": 8.539000000000001e-05, "loss": 0.1153, "step": 8540 }, { "grad_norm": 0.6044209599494934, "learning_rate": 8.549000000000001e-05, "loss": 0.1071, "step": 8550 }, { "grad_norm": 1.0116535425186157, "learning_rate": 8.559e-05, "loss": 0.1113, "step": 8560 }, { "grad_norm": 0.9716674089431763, "learning_rate": 8.569e-05, "loss": 0.1083, "step": 8570 }, { "grad_norm": 0.7499246001243591, "learning_rate": 8.579e-05, "loss": 0.1075, "step": 8580 }, { "grad_norm": 0.8012143969535828, "learning_rate": 8.589e-05, "loss": 0.1031, "step": 8590 }, { "grad_norm": 1.0194240808486938, "learning_rate": 8.599000000000001e-05, "loss": 0.0906, "step": 8600 }, { "grad_norm": 0.8011866211891174, "learning_rate": 8.609e-05, "loss": 0.1066, "step": 8610 }, { "grad_norm": 0.722503662109375, "learning_rate": 8.619e-05, "loss": 0.1037, "step": 8620 }, { "grad_norm": 0.7976551055908203, "learning_rate": 8.629e-05, "loss": 0.1784, "step": 8630 }, { "grad_norm": 0.8855199813842773, "learning_rate": 8.639e-05, "loss": 0.0886, "step": 8640 }, { "grad_norm": 0.5246119499206543, "learning_rate": 8.649000000000001e-05, "loss": 0.0914, "step": 8650 }, { "grad_norm": 0.7649372816085815, "learning_rate": 8.659000000000001e-05, "loss": 0.113, "step": 8660 }, { "grad_norm": 0.829483687877655, "learning_rate": 8.669e-05, "loss": 0.0977, "step": 8670 }, { "grad_norm": 0.6346505284309387, "learning_rate": 8.679e-05, "loss": 0.0893, "step": 8680 }, { "grad_norm": 0.49878618121147156, "learning_rate": 8.689e-05, "loss": 0.1017, "step": 8690 }, { "grad_norm": 1.5691807270050049, "learning_rate": 8.699e-05, "loss": 0.092, "step": 8700 }, { "grad_norm": 0.7634190320968628, "learning_rate": 8.709000000000001e-05, "loss": 0.1022, "step": 8710 }, { "grad_norm": 0.9104234576225281, "learning_rate": 8.719000000000001e-05, "loss": 0.0994, "step": 8720 }, { "grad_norm": 0.7287840843200684, "learning_rate": 8.729e-05, "loss": 0.1164, "step": 8730 }, { "grad_norm": 1.7977399826049805, "learning_rate": 8.739e-05, "loss": 0.1262, "step": 8740 }, { "grad_norm": 1.5058479309082031, "learning_rate": 8.749e-05, "loss": 0.1172, "step": 8750 }, { "grad_norm": 0.8767040967941284, "learning_rate": 8.759e-05, "loss": 0.1174, "step": 8760 }, { "grad_norm": 0.7044188380241394, "learning_rate": 8.769000000000001e-05, "loss": 0.1046, "step": 8770 }, { "grad_norm": 1.2278473377227783, "learning_rate": 8.779000000000001e-05, "loss": 0.1201, "step": 8780 }, { "grad_norm": 0.8969400525093079, "learning_rate": 8.789e-05, "loss": 0.1554, "step": 8790 }, { "grad_norm": 0.9754295349121094, "learning_rate": 8.799e-05, "loss": 0.1202, "step": 8800 }, { "grad_norm": 0.8776219487190247, "learning_rate": 8.809e-05, "loss": 0.125, "step": 8810 }, { "grad_norm": 0.8514942526817322, "learning_rate": 8.819000000000001e-05, "loss": 0.1393, "step": 8820 }, { "grad_norm": 0.8918728232383728, "learning_rate": 8.829000000000001e-05, "loss": 0.1352, "step": 8830 }, { "grad_norm": 0.6506112217903137, "learning_rate": 8.839000000000001e-05, "loss": 0.1192, "step": 8840 }, { "grad_norm": 0.9918071627616882, "learning_rate": 8.849e-05, "loss": 0.1219, "step": 8850 }, { "grad_norm": 0.749983549118042, "learning_rate": 8.859e-05, "loss": 0.095, "step": 8860 }, { "grad_norm": 0.8175488114356995, "learning_rate": 8.869e-05, "loss": 0.1161, "step": 8870 }, { "grad_norm": 0.884056031703949, "learning_rate": 8.879000000000001e-05, "loss": 0.1096, "step": 8880 }, { "grad_norm": 0.6860130429267883, "learning_rate": 8.889000000000001e-05, "loss": 0.1149, "step": 8890 }, { "grad_norm": 0.9091658592224121, "learning_rate": 8.899e-05, "loss": 0.123, "step": 8900 }, { "grad_norm": 0.7469496726989746, "learning_rate": 8.909e-05, "loss": 0.1168, "step": 8910 }, { "grad_norm": 0.690420389175415, "learning_rate": 8.919e-05, "loss": 0.0973, "step": 8920 }, { "grad_norm": 0.953324556350708, "learning_rate": 8.929000000000001e-05, "loss": 0.134, "step": 8930 }, { "grad_norm": 0.9958174228668213, "learning_rate": 8.939000000000001e-05, "loss": 0.0951, "step": 8940 }, { "grad_norm": 0.9288701415061951, "learning_rate": 8.949000000000001e-05, "loss": 0.1139, "step": 8950 }, { "grad_norm": 0.9118992686271667, "learning_rate": 8.959e-05, "loss": 0.1552, "step": 8960 }, { "grad_norm": 0.9384323954582214, "learning_rate": 8.969e-05, "loss": 0.1227, "step": 8970 }, { "grad_norm": 0.9010762572288513, "learning_rate": 8.979e-05, "loss": 0.106, "step": 8980 }, { "grad_norm": 0.8667104840278625, "learning_rate": 8.989000000000001e-05, "loss": 0.1209, "step": 8990 }, { "grad_norm": 0.9683027267456055, "learning_rate": 8.999000000000001e-05, "loss": 0.106, "step": 9000 }, { "grad_norm": 0.8360679149627686, "learning_rate": 9.009000000000001e-05, "loss": 0.1292, "step": 9010 }, { "grad_norm": 0.9607980847358704, "learning_rate": 9.019e-05, "loss": 0.1402, "step": 9020 }, { "grad_norm": 0.7085020542144775, "learning_rate": 9.029e-05, "loss": 0.1083, "step": 9030 }, { "grad_norm": 0.7802607417106628, "learning_rate": 9.039000000000001e-05, "loss": 0.1235, "step": 9040 }, { "grad_norm": 0.9685109853744507, "learning_rate": 9.049000000000001e-05, "loss": 0.1058, "step": 9050 }, { "grad_norm": 0.47269928455352783, "learning_rate": 9.059000000000001e-05, "loss": 0.1117, "step": 9060 }, { "grad_norm": 0.8312849402427673, "learning_rate": 9.069000000000001e-05, "loss": 0.1312, "step": 9070 }, { "grad_norm": 1.1591686010360718, "learning_rate": 9.079e-05, "loss": 0.1252, "step": 9080 }, { "grad_norm": 0.8206474781036377, "learning_rate": 9.089e-05, "loss": 0.1542, "step": 9090 }, { "grad_norm": 0.6397320628166199, "learning_rate": 9.099000000000001e-05, "loss": 0.0974, "step": 9100 }, { "grad_norm": 0.7310614585876465, "learning_rate": 9.109000000000001e-05, "loss": 0.1011, "step": 9110 }, { "grad_norm": 1.009570598602295, "learning_rate": 9.119000000000001e-05, "loss": 0.1221, "step": 9120 }, { "grad_norm": 0.8295372128486633, "learning_rate": 9.129000000000002e-05, "loss": 0.1101, "step": 9130 }, { "grad_norm": 0.7330739498138428, "learning_rate": 9.139e-05, "loss": 0.1248, "step": 9140 }, { "grad_norm": 0.8507656455039978, "learning_rate": 9.149e-05, "loss": 0.12, "step": 9150 }, { "grad_norm": 0.7655848860740662, "learning_rate": 9.159000000000001e-05, "loss": 0.112, "step": 9160 }, { "grad_norm": 0.6923570036888123, "learning_rate": 9.169000000000001e-05, "loss": 0.1138, "step": 9170 }, { "grad_norm": 0.74094557762146, "learning_rate": 9.179000000000001e-05, "loss": 0.1037, "step": 9180 }, { "grad_norm": 0.990933358669281, "learning_rate": 9.189e-05, "loss": 0.1212, "step": 9190 }, { "grad_norm": 0.8186109662055969, "learning_rate": 9.199e-05, "loss": 0.0953, "step": 9200 }, { "grad_norm": 1.046505331993103, "learning_rate": 9.209000000000001e-05, "loss": 0.1109, "step": 9210 }, { "grad_norm": 0.7732158899307251, "learning_rate": 9.219000000000001e-05, "loss": 0.1108, "step": 9220 }, { "grad_norm": 1.1698670387268066, "learning_rate": 9.229000000000001e-05, "loss": 0.0843, "step": 9230 }, { "grad_norm": 0.6777998208999634, "learning_rate": 9.239000000000001e-05, "loss": 0.1284, "step": 9240 }, { "grad_norm": 1.0150254964828491, "learning_rate": 9.249e-05, "loss": 0.0986, "step": 9250 }, { "grad_norm": 0.9639458060264587, "learning_rate": 9.258999999999999e-05, "loss": 0.1386, "step": 9260 }, { "grad_norm": 0.7445363402366638, "learning_rate": 9.269e-05, "loss": 0.0936, "step": 9270 }, { "grad_norm": 0.8966082334518433, "learning_rate": 9.279e-05, "loss": 0.1153, "step": 9280 }, { "grad_norm": 0.7503647208213806, "learning_rate": 9.289e-05, "loss": 0.11, "step": 9290 }, { "grad_norm": 0.6890041828155518, "learning_rate": 9.299e-05, "loss": 0.1085, "step": 9300 }, { "grad_norm": 0.7722306251525879, "learning_rate": 9.309e-05, "loss": 0.1338, "step": 9310 }, { "grad_norm": 0.8761326670646667, "learning_rate": 9.319e-05, "loss": 0.0996, "step": 9320 }, { "grad_norm": 1.1794902086257935, "learning_rate": 9.329e-05, "loss": 0.1062, "step": 9330 }, { "grad_norm": 1.0388062000274658, "learning_rate": 9.339e-05, "loss": 0.1079, "step": 9340 }, { "grad_norm": 0.8443741202354431, "learning_rate": 9.349e-05, "loss": 0.1268, "step": 9350 }, { "grad_norm": 0.7427790760993958, "learning_rate": 9.359e-05, "loss": 0.1312, "step": 9360 }, { "grad_norm": 0.8625345826148987, "learning_rate": 9.369e-05, "loss": 0.1134, "step": 9370 }, { "grad_norm": 0.7200597524642944, "learning_rate": 9.379e-05, "loss": 0.1428, "step": 9380 }, { "grad_norm": 0.6420333385467529, "learning_rate": 9.389e-05, "loss": 0.1138, "step": 9390 }, { "grad_norm": 0.7944412231445312, "learning_rate": 9.399e-05, "loss": 0.1007, "step": 9400 }, { "grad_norm": 0.7559469938278198, "learning_rate": 9.409e-05, "loss": 0.0984, "step": 9410 }, { "grad_norm": 0.6862945556640625, "learning_rate": 9.419e-05, "loss": 0.1168, "step": 9420 }, { "grad_norm": 0.8748782277107239, "learning_rate": 9.429000000000001e-05, "loss": 0.0949, "step": 9430 }, { "grad_norm": 0.8607191443443298, "learning_rate": 9.439e-05, "loss": 0.1022, "step": 9440 }, { "grad_norm": 0.6924458742141724, "learning_rate": 9.449e-05, "loss": 0.1054, "step": 9450 }, { "grad_norm": 0.633037269115448, "learning_rate": 9.459e-05, "loss": 0.1148, "step": 9460 }, { "grad_norm": 0.927825927734375, "learning_rate": 9.469e-05, "loss": 0.1246, "step": 9470 }, { "grad_norm": 1.2544233798980713, "learning_rate": 9.479e-05, "loss": 0.1014, "step": 9480 }, { "grad_norm": 0.655571460723877, "learning_rate": 9.489e-05, "loss": 0.1061, "step": 9490 }, { "grad_norm": 0.5895609855651855, "learning_rate": 9.499e-05, "loss": 0.0861, "step": 9500 }, { "grad_norm": 0.6156713962554932, "learning_rate": 9.509e-05, "loss": 0.1008, "step": 9510 }, { "grad_norm": 0.8699464201927185, "learning_rate": 9.519e-05, "loss": 0.1098, "step": 9520 }, { "grad_norm": 1.0816878080368042, "learning_rate": 9.529e-05, "loss": 0.135, "step": 9530 }, { "grad_norm": 1.3330824375152588, "learning_rate": 9.539e-05, "loss": 0.1266, "step": 9540 }, { "grad_norm": 0.9656386375427246, "learning_rate": 9.549e-05, "loss": 0.136, "step": 9550 }, { "grad_norm": 1.1430418491363525, "learning_rate": 9.559e-05, "loss": 0.1222, "step": 9560 }, { "grad_norm": 0.7832177877426147, "learning_rate": 9.569e-05, "loss": 0.1029, "step": 9570 }, { "grad_norm": 0.6447238326072693, "learning_rate": 9.579e-05, "loss": 0.1039, "step": 9580 }, { "grad_norm": 0.5840773582458496, "learning_rate": 9.589e-05, "loss": 0.1112, "step": 9590 }, { "grad_norm": 0.6518750786781311, "learning_rate": 9.599000000000001e-05, "loss": 0.0985, "step": 9600 }, { "grad_norm": 0.6659868955612183, "learning_rate": 9.609e-05, "loss": 0.0945, "step": 9610 }, { "grad_norm": 0.8236714005470276, "learning_rate": 9.619e-05, "loss": 0.1122, "step": 9620 }, { "grad_norm": 1.1134438514709473, "learning_rate": 9.629e-05, "loss": 0.1036, "step": 9630 }, { "grad_norm": 0.8966442346572876, "learning_rate": 9.639e-05, "loss": 0.1211, "step": 9640 }, { "grad_norm": 0.8044868111610413, "learning_rate": 9.649e-05, "loss": 0.1087, "step": 9650 }, { "grad_norm": 0.7582221627235413, "learning_rate": 9.659000000000001e-05, "loss": 0.098, "step": 9660 }, { "grad_norm": 0.8490458130836487, "learning_rate": 9.669e-05, "loss": 0.113, "step": 9670 }, { "grad_norm": 0.7920427322387695, "learning_rate": 9.679e-05, "loss": 0.1061, "step": 9680 }, { "grad_norm": 0.8877688050270081, "learning_rate": 9.689e-05, "loss": 0.1184, "step": 9690 }, { "grad_norm": 1.0008623600006104, "learning_rate": 9.699e-05, "loss": 0.1071, "step": 9700 }, { "grad_norm": 0.8438384532928467, "learning_rate": 9.709000000000001e-05, "loss": 0.0968, "step": 9710 }, { "grad_norm": 0.7979294061660767, "learning_rate": 9.719000000000001e-05, "loss": 0.097, "step": 9720 }, { "grad_norm": 0.6680224537849426, "learning_rate": 9.729e-05, "loss": 0.1116, "step": 9730 }, { "grad_norm": 1.0865802764892578, "learning_rate": 9.739e-05, "loss": 0.0969, "step": 9740 }, { "grad_norm": 0.8223173022270203, "learning_rate": 9.749e-05, "loss": 0.1048, "step": 9750 }, { "grad_norm": 0.7693291902542114, "learning_rate": 9.759e-05, "loss": 0.0988, "step": 9760 }, { "grad_norm": 0.8022322654724121, "learning_rate": 9.769000000000001e-05, "loss": 0.1309, "step": 9770 }, { "grad_norm": 0.7234243154525757, "learning_rate": 9.779e-05, "loss": 0.1247, "step": 9780 }, { "grad_norm": 1.034021019935608, "learning_rate": 9.789e-05, "loss": 0.0983, "step": 9790 }, { "grad_norm": 0.9601691961288452, "learning_rate": 9.799e-05, "loss": 0.1013, "step": 9800 }, { "grad_norm": 0.9926884770393372, "learning_rate": 9.809e-05, "loss": 0.0916, "step": 9810 }, { "grad_norm": 0.7213889360427856, "learning_rate": 9.819000000000001e-05, "loss": 0.115, "step": 9820 }, { "grad_norm": 0.8864683508872986, "learning_rate": 9.829000000000001e-05, "loss": 0.1339, "step": 9830 }, { "grad_norm": 0.700395941734314, "learning_rate": 9.839e-05, "loss": 0.1112, "step": 9840 }, { "grad_norm": 0.7275599241256714, "learning_rate": 9.849e-05, "loss": 0.1317, "step": 9850 }, { "grad_norm": 0.8605673909187317, "learning_rate": 9.859e-05, "loss": 0.1148, "step": 9860 }, { "grad_norm": 1.0375373363494873, "learning_rate": 9.869e-05, "loss": 0.1108, "step": 9870 }, { "grad_norm": 0.9773157835006714, "learning_rate": 9.879000000000001e-05, "loss": 0.1146, "step": 9880 }, { "grad_norm": 1.224466323852539, "learning_rate": 9.889000000000001e-05, "loss": 0.142, "step": 9890 }, { "grad_norm": 0.5765426754951477, "learning_rate": 9.899e-05, "loss": 0.1139, "step": 9900 }, { "grad_norm": 0.613315224647522, "learning_rate": 9.909e-05, "loss": 0.1051, "step": 9910 }, { "grad_norm": 0.626232385635376, "learning_rate": 9.919e-05, "loss": 0.1085, "step": 9920 }, { "grad_norm": 1.1613727807998657, "learning_rate": 9.929e-05, "loss": 0.1309, "step": 9930 }, { "grad_norm": 0.8042129874229431, "learning_rate": 9.939000000000001e-05, "loss": 0.0976, "step": 9940 }, { "grad_norm": 0.8563263416290283, "learning_rate": 9.949000000000001e-05, "loss": 0.1035, "step": 9950 }, { "grad_norm": 0.8037523627281189, "learning_rate": 9.959e-05, "loss": 0.1253, "step": 9960 }, { "grad_norm": 0.7394858598709106, "learning_rate": 9.969e-05, "loss": 0.1074, "step": 9970 }, { "grad_norm": 0.7797062397003174, "learning_rate": 9.979e-05, "loss": 0.1256, "step": 9980 }, { "grad_norm": 0.8144835233688354, "learning_rate": 9.989000000000001e-05, "loss": 0.1021, "step": 9990 }, { "grad_norm": 0.8877930641174316, "learning_rate": 9.999000000000001e-05, "loss": 0.1315, "step": 10000 }, { "grad_norm": 0.8264940977096558, "learning_rate": 9.999999944637262e-05, "loss": 0.1278, "step": 10010 }, { "grad_norm": 0.8573178648948669, "learning_rate": 9.999999753259892e-05, "loss": 0.1169, "step": 10020 }, { "grad_norm": 0.6292046308517456, "learning_rate": 9.999999425184409e-05, "loss": 0.1069, "step": 10030 }, { "grad_norm": 0.9978053569793701, "learning_rate": 9.999998960410819e-05, "loss": 0.1165, "step": 10040 }, { "grad_norm": 0.7574123740196228, "learning_rate": 9.999998358939136e-05, "loss": 0.1195, "step": 10050 }, { "grad_norm": 0.678922712802887, "learning_rate": 9.999997620769377e-05, "loss": 0.103, "step": 10060 }, { "grad_norm": 0.8895981311798096, "learning_rate": 9.999996745901561e-05, "loss": 0.0997, "step": 10070 }, { "grad_norm": 0.8450349569320679, "learning_rate": 9.999995734335712e-05, "loss": 0.1167, "step": 10080 }, { "grad_norm": 0.6437389850616455, "learning_rate": 9.99999458607186e-05, "loss": 0.1302, "step": 10090 }, { "grad_norm": 0.7665029764175415, "learning_rate": 9.999993301110033e-05, "loss": 0.1042, "step": 10100 }, { "grad_norm": 0.9622400999069214, "learning_rate": 9.999991879450268e-05, "loss": 0.0884, "step": 10110 }, { "grad_norm": 0.7476296424865723, "learning_rate": 9.999990321092603e-05, "loss": 0.0831, "step": 10120 }, { "grad_norm": 0.7989774346351624, "learning_rate": 9.99998862603708e-05, "loss": 0.1433, "step": 10130 }, { "grad_norm": 0.8561466932296753, "learning_rate": 9.999986794283748e-05, "loss": 0.1302, "step": 10140 }, { "grad_norm": 0.6702423691749573, "learning_rate": 9.999984825832654e-05, "loss": 0.1353, "step": 10150 }, { "grad_norm": 0.7844013571739197, "learning_rate": 9.999982720683854e-05, "loss": 0.1023, "step": 10160 }, { "grad_norm": 0.8346202373504639, "learning_rate": 9.999980478837405e-05, "loss": 0.0987, "step": 10170 }, { "grad_norm": 0.7990568280220032, "learning_rate": 9.999978100293367e-05, "loss": 0.1036, "step": 10180 }, { "grad_norm": 0.6999818086624146, "learning_rate": 9.999975585051808e-05, "loss": 0.1062, "step": 10190 }, { "grad_norm": 0.8309059143066406, "learning_rate": 9.999972933112795e-05, "loss": 0.1189, "step": 10200 }, { "grad_norm": 0.685377836227417, "learning_rate": 9.999970144476398e-05, "loss": 0.0924, "step": 10210 }, { "grad_norm": 0.8849321007728577, "learning_rate": 9.999967219142698e-05, "loss": 0.1285, "step": 10220 }, { "grad_norm": 0.779046893119812, "learning_rate": 9.999964157111771e-05, "loss": 0.1246, "step": 10230 }, { "grad_norm": 1.0976123809814453, "learning_rate": 9.999960958383704e-05, "loss": 0.1336, "step": 10240 }, { "grad_norm": 1.111795425415039, "learning_rate": 9.999957622958582e-05, "loss": 0.1158, "step": 10250 }, { "grad_norm": 0.8413555026054382, "learning_rate": 9.999954150836497e-05, "loss": 0.1213, "step": 10260 }, { "grad_norm": 0.9158645868301392, "learning_rate": 9.999950542017544e-05, "loss": 0.1127, "step": 10270 }, { "grad_norm": 0.9392321109771729, "learning_rate": 9.99994679650182e-05, "loss": 0.1038, "step": 10280 }, { "grad_norm": 1.0374712944030762, "learning_rate": 9.99994291428943e-05, "loss": 0.1127, "step": 10290 }, { "grad_norm": 0.796461284160614, "learning_rate": 9.999938895380481e-05, "loss": 0.0968, "step": 10300 }, { "grad_norm": 0.5726178288459778, "learning_rate": 9.999934739775079e-05, "loss": 0.1303, "step": 10310 }, { "grad_norm": 0.6630314588546753, "learning_rate": 9.99993044747334e-05, "loss": 0.1213, "step": 10320 }, { "grad_norm": 0.8380483984947205, "learning_rate": 9.99992601847538e-05, "loss": 0.1207, "step": 10330 }, { "grad_norm": 0.708931028842926, "learning_rate": 9.999921452781324e-05, "loss": 0.0948, "step": 10340 }, { "grad_norm": 1.0425593852996826, "learning_rate": 9.999916750391291e-05, "loss": 0.0988, "step": 10350 }, { "grad_norm": 0.8419025540351868, "learning_rate": 9.999911911305413e-05, "loss": 0.0948, "step": 10360 }, { "grad_norm": 1.1510485410690308, "learning_rate": 9.999906935523822e-05, "loss": 0.1171, "step": 10370 }, { "grad_norm": 0.9366475343704224, "learning_rate": 9.999901823046655e-05, "loss": 0.1357, "step": 10380 }, { "grad_norm": 0.8005967736244202, "learning_rate": 9.999896573874048e-05, "loss": 0.1046, "step": 10390 }, { "grad_norm": 0.9521505236625671, "learning_rate": 9.999891188006149e-05, "loss": 0.1042, "step": 10400 }, { "grad_norm": 0.5906554460525513, "learning_rate": 9.999885665443103e-05, "loss": 0.1033, "step": 10410 }, { "grad_norm": 0.8411032557487488, "learning_rate": 9.99988000618506e-05, "loss": 0.1221, "step": 10420 }, { "grad_norm": 0.7045748233795166, "learning_rate": 9.999874210232179e-05, "loss": 0.1038, "step": 10430 }, { "grad_norm": 0.781524658203125, "learning_rate": 9.999868277584612e-05, "loss": 0.0967, "step": 10440 }, { "grad_norm": 0.7099361419677734, "learning_rate": 9.999862208242526e-05, "loss": 0.0901, "step": 10450 }, { "grad_norm": 0.85304194688797, "learning_rate": 9.999856002206086e-05, "loss": 0.1084, "step": 10460 }, { "grad_norm": 0.5998881459236145, "learning_rate": 9.99984965947546e-05, "loss": 0.0946, "step": 10470 }, { "grad_norm": 0.4497559666633606, "learning_rate": 9.999843180050825e-05, "loss": 0.0823, "step": 10480 }, { "grad_norm": 0.7184239625930786, "learning_rate": 9.999836563932355e-05, "loss": 0.1259, "step": 10490 }, { "grad_norm": 0.6859263181686401, "learning_rate": 9.99982981112023e-05, "loss": 0.1174, "step": 10500 }, { "grad_norm": 0.8468312621116638, "learning_rate": 9.999822921614636e-05, "loss": 0.0974, "step": 10510 }, { "grad_norm": 0.7241590619087219, "learning_rate": 9.999815895415764e-05, "loss": 0.1199, "step": 10520 }, { "grad_norm": 0.6786900758743286, "learning_rate": 9.999808732523804e-05, "loss": 0.1273, "step": 10530 }, { "grad_norm": 0.7496358156204224, "learning_rate": 9.99980143293895e-05, "loss": 0.1137, "step": 10540 }, { "grad_norm": 1.0541305541992188, "learning_rate": 9.999793996661402e-05, "loss": 0.1245, "step": 10550 }, { "grad_norm": 0.7448856234550476, "learning_rate": 9.999786423691368e-05, "loss": 0.1162, "step": 10560 }, { "grad_norm": 0.7011431455612183, "learning_rate": 9.999778714029049e-05, "loss": 0.1409, "step": 10570 }, { "grad_norm": 0.872665524482727, "learning_rate": 9.999770867674658e-05, "loss": 0.0898, "step": 10580 }, { "grad_norm": 0.7696765065193176, "learning_rate": 9.999762884628413e-05, "loss": 0.1043, "step": 10590 }, { "grad_norm": 0.8603659868240356, "learning_rate": 9.999754764890525e-05, "loss": 0.0844, "step": 10600 }, { "grad_norm": 0.6151083707809448, "learning_rate": 9.999746508461222e-05, "loss": 0.0779, "step": 10610 }, { "grad_norm": 0.8696779012680054, "learning_rate": 9.999738115340727e-05, "loss": 0.0961, "step": 10620 }, { "grad_norm": 1.1314290761947632, "learning_rate": 9.999729585529271e-05, "loss": 0.121, "step": 10630 }, { "grad_norm": 0.7281700968742371, "learning_rate": 9.999720919027087e-05, "loss": 0.0976, "step": 10640 }, { "grad_norm": 0.6620031595230103, "learning_rate": 9.99971211583441e-05, "loss": 0.1279, "step": 10650 }, { "grad_norm": 0.8465137481689453, "learning_rate": 9.999703175951483e-05, "loss": 0.1002, "step": 10660 }, { "grad_norm": 1.1588407754898071, "learning_rate": 9.999694099378549e-05, "loss": 0.102, "step": 10670 }, { "grad_norm": 0.7457956671714783, "learning_rate": 9.999684886115857e-05, "loss": 0.1065, "step": 10680 }, { "grad_norm": 0.764904797077179, "learning_rate": 9.999675536163657e-05, "loss": 0.0908, "step": 10690 }, { "grad_norm": 0.7011914253234863, "learning_rate": 9.999666049522208e-05, "loss": 0.0965, "step": 10700 }, { "grad_norm": 0.9033677577972412, "learning_rate": 9.999656426191766e-05, "loss": 0.101, "step": 10710 }, { "grad_norm": 0.5427287817001343, "learning_rate": 9.999646666172597e-05, "loss": 0.1015, "step": 10720 }, { "grad_norm": 0.7387320399284363, "learning_rate": 9.999636769464967e-05, "loss": 0.09, "step": 10730 }, { "grad_norm": 0.5283268094062805, "learning_rate": 9.999626736069144e-05, "loss": 0.0972, "step": 10740 }, { "grad_norm": 0.6309165358543396, "learning_rate": 9.999616565985406e-05, "loss": 0.1011, "step": 10750 }, { "grad_norm": 1.2745935916900635, "learning_rate": 9.999606259214029e-05, "loss": 0.1048, "step": 10760 }, { "grad_norm": 0.5904372334480286, "learning_rate": 9.999595815755294e-05, "loss": 0.0865, "step": 10770 }, { "grad_norm": 0.7146846652030945, "learning_rate": 9.999585235609488e-05, "loss": 0.0919, "step": 10780 }, { "grad_norm": 0.7796810269355774, "learning_rate": 9.9995745187769e-05, "loss": 0.1189, "step": 10790 }, { "grad_norm": 0.9533534646034241, "learning_rate": 9.999563665257823e-05, "loss": 0.0839, "step": 10800 }, { "grad_norm": 0.9773382544517517, "learning_rate": 9.999552675052554e-05, "loss": 0.0949, "step": 10810 }, { "grad_norm": 0.6232694387435913, "learning_rate": 9.999541548161392e-05, "loss": 0.0922, "step": 10820 }, { "grad_norm": 0.6737024784088135, "learning_rate": 9.999530284584644e-05, "loss": 0.0986, "step": 10830 }, { "grad_norm": 0.5846999883651733, "learning_rate": 9.999518884322617e-05, "loss": 0.0798, "step": 10840 }, { "grad_norm": 0.6867079734802246, "learning_rate": 9.99950734737562e-05, "loss": 0.0994, "step": 10850 }, { "grad_norm": 0.7344138622283936, "learning_rate": 9.99949567374397e-05, "loss": 0.0925, "step": 10860 }, { "grad_norm": 0.6640591621398926, "learning_rate": 9.999483863427988e-05, "loss": 0.1223, "step": 10870 }, { "grad_norm": 0.756860077381134, "learning_rate": 9.999471916427993e-05, "loss": 0.1074, "step": 10880 }, { "grad_norm": 0.7165181040763855, "learning_rate": 9.999459832744318e-05, "loss": 0.0833, "step": 10890 }, { "grad_norm": 0.91389399766922, "learning_rate": 9.999447612377286e-05, "loss": 0.0994, "step": 10900 }, { "grad_norm": 0.7869278192520142, "learning_rate": 9.999435255327236e-05, "loss": 0.103, "step": 10910 }, { "grad_norm": 0.7266919612884521, "learning_rate": 9.999422761594503e-05, "loss": 0.0836, "step": 10920 }, { "grad_norm": 0.5481669306755066, "learning_rate": 9.999410131179431e-05, "loss": 0.0966, "step": 10930 }, { "grad_norm": 0.946270227432251, "learning_rate": 9.999397364082365e-05, "loss": 0.0913, "step": 10940 }, { "grad_norm": 0.6033931970596313, "learning_rate": 9.99938446030365e-05, "loss": 0.0814, "step": 10950 }, { "grad_norm": 0.6585834622383118, "learning_rate": 9.999371419843646e-05, "loss": 0.0957, "step": 10960 }, { "grad_norm": 0.7611512541770935, "learning_rate": 9.999358242702703e-05, "loss": 0.089, "step": 10970 }, { "grad_norm": 0.7189891934394836, "learning_rate": 9.999344928881185e-05, "loss": 0.0883, "step": 10980 }, { "grad_norm": 0.9125726222991943, "learning_rate": 9.999331478379455e-05, "loss": 0.1017, "step": 10990 }, { "grad_norm": 0.8826975226402283, "learning_rate": 9.999317891197881e-05, "loss": 0.0915, "step": 11000 }, { "grad_norm": 0.6583507657051086, "learning_rate": 9.999304167336832e-05, "loss": 0.1115, "step": 11010 }, { "grad_norm": 0.6950305700302124, "learning_rate": 9.999290306796687e-05, "loss": 0.1056, "step": 11020 }, { "grad_norm": 0.8852323293685913, "learning_rate": 9.99927630957782e-05, "loss": 0.097, "step": 11030 }, { "grad_norm": 0.6240564584732056, "learning_rate": 9.999262175680621e-05, "loss": 0.1004, "step": 11040 }, { "grad_norm": 0.6847505569458008, "learning_rate": 9.999247905105471e-05, "loss": 0.1142, "step": 11050 }, { "grad_norm": 0.9225022792816162, "learning_rate": 9.99923349785276e-05, "loss": 0.0993, "step": 11060 }, { "grad_norm": 0.7146589159965515, "learning_rate": 9.999218953922884e-05, "loss": 0.0926, "step": 11070 }, { "grad_norm": 0.6172218322753906, "learning_rate": 9.99920427331624e-05, "loss": 0.0828, "step": 11080 }, { "grad_norm": 0.6340732574462891, "learning_rate": 9.99918945603323e-05, "loss": 0.0931, "step": 11090 }, { "grad_norm": 0.47462403774261475, "learning_rate": 9.999174502074258e-05, "loss": 0.1036, "step": 11100 }, { "grad_norm": 0.8226956129074097, "learning_rate": 9.999159411439733e-05, "loss": 0.1078, "step": 11110 }, { "grad_norm": 0.7503790855407715, "learning_rate": 9.999144184130067e-05, "loss": 0.1038, "step": 11120 }, { "grad_norm": 0.8876129388809204, "learning_rate": 9.999128820145678e-05, "loss": 0.1087, "step": 11130 }, { "grad_norm": 0.7311369180679321, "learning_rate": 9.999113319486986e-05, "loss": 0.0907, "step": 11140 }, { "grad_norm": 0.6988862752914429, "learning_rate": 9.999097682154413e-05, "loss": 0.1153, "step": 11150 }, { "grad_norm": 0.9635097980499268, "learning_rate": 9.999081908148387e-05, "loss": 0.1104, "step": 11160 }, { "grad_norm": 1.0528329610824585, "learning_rate": 9.99906599746934e-05, "loss": 0.1035, "step": 11170 }, { "grad_norm": 0.7078378796577454, "learning_rate": 9.999049950117707e-05, "loss": 0.0942, "step": 11180 }, { "grad_norm": 0.7131977677345276, "learning_rate": 9.999033766093926e-05, "loss": 0.1241, "step": 11190 }, { "grad_norm": 0.5463507175445557, "learning_rate": 9.99901744539844e-05, "loss": 0.1329, "step": 11200 }, { "grad_norm": 1.0129623413085938, "learning_rate": 9.999000988031695e-05, "loss": 0.1145, "step": 11210 }, { "grad_norm": 0.7841231822967529, "learning_rate": 9.99898439399414e-05, "loss": 0.1051, "step": 11220 }, { "grad_norm": 0.5254455804824829, "learning_rate": 9.998967663286232e-05, "loss": 0.1005, "step": 11230 }, { "grad_norm": 0.7099396586418152, "learning_rate": 9.998950795908425e-05, "loss": 0.1037, "step": 11240 }, { "grad_norm": 0.611353874206543, "learning_rate": 9.99893379186118e-05, "loss": 0.0878, "step": 11250 }, { "grad_norm": 1.7543177604675293, "learning_rate": 9.998916651144966e-05, "loss": 0.0931, "step": 11260 }, { "grad_norm": 1.018093228340149, "learning_rate": 9.998899373760247e-05, "loss": 0.1179, "step": 11270 }, { "grad_norm": 0.6986900568008423, "learning_rate": 9.998881959707497e-05, "loss": 0.1001, "step": 11280 }, { "grad_norm": 0.9406096935272217, "learning_rate": 9.998864408987192e-05, "loss": 0.1046, "step": 11290 }, { "grad_norm": 0.6478930115699768, "learning_rate": 9.998846721599813e-05, "loss": 0.1267, "step": 11300 }, { "grad_norm": 1.1545261144638062, "learning_rate": 9.998828897545842e-05, "loss": 0.1018, "step": 11310 }, { "grad_norm": 0.8778548836708069, "learning_rate": 9.998810936825766e-05, "loss": 0.0989, "step": 11320 }, { "grad_norm": 0.5567822456359863, "learning_rate": 9.998792839440079e-05, "loss": 0.0917, "step": 11330 }, { "grad_norm": 1.2127128839492798, "learning_rate": 9.998774605389273e-05, "loss": 0.1218, "step": 11340 }, { "grad_norm": 0.7253042459487915, "learning_rate": 9.998756234673847e-05, "loss": 0.1114, "step": 11350 }, { "grad_norm": 0.6251216530799866, "learning_rate": 9.998737727294303e-05, "loss": 0.1157, "step": 11360 }, { "grad_norm": 0.7864600419998169, "learning_rate": 9.998719083251149e-05, "loss": 0.0894, "step": 11370 }, { "grad_norm": 0.836898148059845, "learning_rate": 9.998700302544893e-05, "loss": 0.1249, "step": 11380 }, { "grad_norm": 0.6214370727539062, "learning_rate": 9.998681385176047e-05, "loss": 0.0877, "step": 11390 }, { "grad_norm": 0.707731306552887, "learning_rate": 9.99866233114513e-05, "loss": 0.1208, "step": 11400 }, { "grad_norm": 0.7194560170173645, "learning_rate": 9.998643140452663e-05, "loss": 0.1062, "step": 11410 }, { "grad_norm": 0.7399633526802063, "learning_rate": 9.99862381309917e-05, "loss": 0.1183, "step": 11420 }, { "grad_norm": 0.9931272268295288, "learning_rate": 9.998604349085182e-05, "loss": 0.1097, "step": 11430 }, { "grad_norm": 1.1211464405059814, "learning_rate": 9.998584748411228e-05, "loss": 0.1211, "step": 11440 }, { "grad_norm": 0.7732235789299011, "learning_rate": 9.998565011077844e-05, "loss": 0.1131, "step": 11450 }, { "grad_norm": 0.8043196201324463, "learning_rate": 9.99854513708557e-05, "loss": 0.1065, "step": 11460 }, { "grad_norm": 0.916357696056366, "learning_rate": 9.998525126434952e-05, "loss": 0.1282, "step": 11470 }, { "grad_norm": 1.079914927482605, "learning_rate": 9.998504979126533e-05, "loss": 0.1172, "step": 11480 }, { "grad_norm": 0.6734189391136169, "learning_rate": 9.998484695160868e-05, "loss": 0.1046, "step": 11490 }, { "grad_norm": 0.475434273481369, "learning_rate": 9.998464274538507e-05, "loss": 0.0759, "step": 11500 }, { "grad_norm": 0.6184627413749695, "learning_rate": 9.99844371726001e-05, "loss": 0.107, "step": 11510 }, { "grad_norm": 0.7274813652038574, "learning_rate": 9.998423023325943e-05, "loss": 0.1012, "step": 11520 }, { "grad_norm": 0.9372004270553589, "learning_rate": 9.998402192736865e-05, "loss": 0.0876, "step": 11530 }, { "grad_norm": 0.729773998260498, "learning_rate": 9.99838122549335e-05, "loss": 0.1116, "step": 11540 }, { "grad_norm": 0.5963252782821655, "learning_rate": 9.998360121595972e-05, "loss": 0.1326, "step": 11550 }, { "grad_norm": 0.6125643849372864, "learning_rate": 9.998338881045304e-05, "loss": 0.0929, "step": 11560 }, { "grad_norm": 0.7562163472175598, "learning_rate": 9.998317503841927e-05, "loss": 0.0898, "step": 11570 }, { "grad_norm": 0.7680157423019409, "learning_rate": 9.99829598998643e-05, "loss": 0.0796, "step": 11580 }, { "grad_norm": 0.7987198233604431, "learning_rate": 9.998274339479398e-05, "loss": 0.1139, "step": 11590 }, { "grad_norm": 0.9012909531593323, "learning_rate": 9.998252552321423e-05, "loss": 0.0751, "step": 11600 }, { "grad_norm": 0.5387470126152039, "learning_rate": 9.998230628513101e-05, "loss": 0.098, "step": 11610 }, { "grad_norm": 0.8539097905158997, "learning_rate": 9.998208568055032e-05, "loss": 0.0984, "step": 11620 }, { "grad_norm": 0.8525514602661133, "learning_rate": 9.998186370947817e-05, "loss": 0.0856, "step": 11630 }, { "grad_norm": 0.8470929861068726, "learning_rate": 9.998164037192066e-05, "loss": 0.107, "step": 11640 }, { "grad_norm": 0.6210410594940186, "learning_rate": 9.998141566788387e-05, "loss": 0.1184, "step": 11650 }, { "grad_norm": 0.7087632417678833, "learning_rate": 9.998118959737395e-05, "loss": 0.1157, "step": 11660 }, { "grad_norm": 0.9937011003494263, "learning_rate": 9.998096216039709e-05, "loss": 0.1198, "step": 11670 }, { "grad_norm": 0.8083012104034424, "learning_rate": 9.99807333569595e-05, "loss": 0.0875, "step": 11680 }, { "grad_norm": 0.9571178555488586, "learning_rate": 9.998050318706744e-05, "loss": 0.0897, "step": 11690 }, { "grad_norm": 0.7533624768257141, "learning_rate": 9.998027165072718e-05, "loss": 0.0792, "step": 11700 }, { "grad_norm": 1.0104731321334839, "learning_rate": 9.998003874794511e-05, "loss": 0.1207, "step": 11710 }, { "grad_norm": 0.6219673752784729, "learning_rate": 9.997980447872753e-05, "loss": 0.0971, "step": 11720 }, { "grad_norm": 0.9560611844062805, "learning_rate": 9.997956884308086e-05, "loss": 0.1098, "step": 11730 }, { "grad_norm": 0.7190254926681519, "learning_rate": 9.997933184101158e-05, "loss": 0.0956, "step": 11740 }, { "grad_norm": 0.7112658619880676, "learning_rate": 9.997909347252613e-05, "loss": 0.1123, "step": 11750 }, { "grad_norm": 1.4141978025436401, "learning_rate": 9.997885373763106e-05, "loss": 0.0884, "step": 11760 }, { "grad_norm": 0.7546306848526001, "learning_rate": 9.997861263633288e-05, "loss": 0.1073, "step": 11770 }, { "grad_norm": 1.033457636833191, "learning_rate": 9.99783701686382e-05, "loss": 0.0951, "step": 11780 }, { "grad_norm": 0.8136588931083679, "learning_rate": 9.997812633455368e-05, "loss": 0.1029, "step": 11790 }, { "grad_norm": 0.7745634913444519, "learning_rate": 9.997788113408595e-05, "loss": 0.1565, "step": 11800 }, { "grad_norm": 0.8335391283035278, "learning_rate": 9.997763456724172e-05, "loss": 0.1234, "step": 11810 }, { "grad_norm": 0.7657585144042969, "learning_rate": 9.997738663402775e-05, "loss": 0.1192, "step": 11820 }, { "grad_norm": 0.757129967212677, "learning_rate": 9.99771373344508e-05, "loss": 0.0911, "step": 11830 }, { "grad_norm": 0.6248199939727783, "learning_rate": 9.99768866685177e-05, "loss": 0.1024, "step": 11840 }, { "grad_norm": 0.5315776467323303, "learning_rate": 9.997663463623528e-05, "loss": 0.0856, "step": 11850 }, { "grad_norm": 0.6881502270698547, "learning_rate": 9.997638123761046e-05, "loss": 0.0858, "step": 11860 }, { "grad_norm": 0.8027727007865906, "learning_rate": 9.997612647265014e-05, "loss": 0.1018, "step": 11870 }, { "grad_norm": 0.5783436298370361, "learning_rate": 9.99758703413613e-05, "loss": 0.1122, "step": 11880 }, { "grad_norm": 0.6867564916610718, "learning_rate": 9.997561284375093e-05, "loss": 0.0967, "step": 11890 }, { "grad_norm": 0.9013704061508179, "learning_rate": 9.997535397982609e-05, "loss": 0.0946, "step": 11900 }, { "grad_norm": 0.7165672779083252, "learning_rate": 9.997509374959384e-05, "loss": 0.0893, "step": 11910 }, { "grad_norm": 0.7323184609413147, "learning_rate": 9.997483215306132e-05, "loss": 0.0901, "step": 11920 }, { "grad_norm": 0.8799582719802856, "learning_rate": 9.997456919023565e-05, "loss": 0.083, "step": 11930 }, { "grad_norm": 0.8063368797302246, "learning_rate": 9.997430486112402e-05, "loss": 0.1134, "step": 11940 }, { "grad_norm": 0.7674893736839294, "learning_rate": 9.997403916573369e-05, "loss": 0.0934, "step": 11950 }, { "grad_norm": 0.661908745765686, "learning_rate": 9.997377210407191e-05, "loss": 0.0924, "step": 11960 }, { "grad_norm": 0.8022158145904541, "learning_rate": 9.997350367614596e-05, "loss": 0.1051, "step": 11970 }, { "grad_norm": 0.8835400342941284, "learning_rate": 9.99732338819632e-05, "loss": 0.1356, "step": 11980 }, { "grad_norm": 0.8130635619163513, "learning_rate": 9.997296272153098e-05, "loss": 0.1109, "step": 11990 }, { "grad_norm": 0.7604323625564575, "learning_rate": 9.997269019485675e-05, "loss": 0.155, "step": 12000 }, { "grad_norm": 0.5710409879684448, "learning_rate": 9.997241630194795e-05, "loss": 0.1067, "step": 12010 }, { "grad_norm": 0.7927724719047546, "learning_rate": 9.997214104281206e-05, "loss": 0.1004, "step": 12020 }, { "grad_norm": 0.7679644823074341, "learning_rate": 9.99718644174566e-05, "loss": 0.102, "step": 12030 }, { "grad_norm": 0.7078688144683838, "learning_rate": 9.997158642588913e-05, "loss": 0.1021, "step": 12040 }, { "grad_norm": 0.5869808197021484, "learning_rate": 9.997130706811728e-05, "loss": 0.0988, "step": 12050 }, { "grad_norm": 0.6515606045722961, "learning_rate": 9.997102634414866e-05, "loss": 0.1034, "step": 12060 }, { "grad_norm": 0.7984655499458313, "learning_rate": 9.997074425399095e-05, "loss": 0.0987, "step": 12070 }, { "grad_norm": 0.5600146651268005, "learning_rate": 9.997046079765187e-05, "loss": 0.1044, "step": 12080 }, { "grad_norm": 0.6712225675582886, "learning_rate": 9.997017597513917e-05, "loss": 0.1071, "step": 12090 }, { "grad_norm": 0.6227784156799316, "learning_rate": 9.996988978646062e-05, "loss": 0.1022, "step": 12100 }, { "grad_norm": 1.0928469896316528, "learning_rate": 9.996960223162406e-05, "loss": 0.1038, "step": 12110 }, { "grad_norm": 0.8702789545059204, "learning_rate": 9.996931331063735e-05, "loss": 0.0825, "step": 12120 }, { "grad_norm": 0.6770996451377869, "learning_rate": 9.996902302350838e-05, "loss": 0.116, "step": 12130 }, { "grad_norm": 0.7365726232528687, "learning_rate": 9.996873137024512e-05, "loss": 0.1004, "step": 12140 }, { "grad_norm": 0.6511625051498413, "learning_rate": 9.99684383508555e-05, "loss": 0.1033, "step": 12150 }, { "grad_norm": 0.7565473914146423, "learning_rate": 9.996814396534753e-05, "loss": 0.105, "step": 12160 }, { "grad_norm": 0.9984499216079712, "learning_rate": 9.99678482137293e-05, "loss": 0.1027, "step": 12170 }, { "grad_norm": 0.89345383644104, "learning_rate": 9.996755109600886e-05, "loss": 0.101, "step": 12180 }, { "grad_norm": 0.7502322793006897, "learning_rate": 9.996725261219435e-05, "loss": 0.1175, "step": 12190 }, { "grad_norm": 0.8098210096359253, "learning_rate": 9.996695276229392e-05, "loss": 0.1141, "step": 12200 }, { "grad_norm": 1.046172022819519, "learning_rate": 9.996665154631577e-05, "loss": 0.0975, "step": 12210 }, { "grad_norm": 0.7240563035011292, "learning_rate": 9.996634896426815e-05, "loss": 0.1018, "step": 12220 }, { "grad_norm": 0.866736650466919, "learning_rate": 9.996604501615931e-05, "loss": 0.0972, "step": 12230 }, { "grad_norm": 0.731220006942749, "learning_rate": 9.996573970199758e-05, "loss": 0.1236, "step": 12240 }, { "grad_norm": 0.6175957322120667, "learning_rate": 9.996543302179128e-05, "loss": 0.0856, "step": 12250 }, { "grad_norm": 0.5708375573158264, "learning_rate": 9.996512497554884e-05, "loss": 0.0793, "step": 12260 }, { "grad_norm": 0.8434088230133057, "learning_rate": 9.996481556327863e-05, "loss": 0.0767, "step": 12270 }, { "grad_norm": 0.6445064544677734, "learning_rate": 9.996450478498915e-05, "loss": 0.1074, "step": 12280 }, { "grad_norm": 0.9098164439201355, "learning_rate": 9.996419264068888e-05, "loss": 0.1265, "step": 12290 }, { "grad_norm": 0.7487826347351074, "learning_rate": 9.996387913038635e-05, "loss": 0.1226, "step": 12300 }, { "grad_norm": 0.8322883248329163, "learning_rate": 9.996356425409013e-05, "loss": 0.1055, "step": 12310 }, { "grad_norm": 0.7333357930183411, "learning_rate": 9.996324801180884e-05, "loss": 0.1476, "step": 12320 }, { "grad_norm": 0.7105318903923035, "learning_rate": 9.996293040355111e-05, "loss": 0.1164, "step": 12330 }, { "grad_norm": 0.9432327747344971, "learning_rate": 9.996261142932564e-05, "loss": 0.1, "step": 12340 }, { "grad_norm": 0.9389694929122925, "learning_rate": 9.996229108914113e-05, "loss": 0.0957, "step": 12350 }, { "grad_norm": 0.7405842542648315, "learning_rate": 9.996196938300638e-05, "loss": 0.0898, "step": 12360 }, { "grad_norm": 0.7367737889289856, "learning_rate": 9.996164631093012e-05, "loss": 0.1222, "step": 12370 }, { "grad_norm": 0.7641055583953857, "learning_rate": 9.996132187292125e-05, "loss": 0.1048, "step": 12380 }, { "grad_norm": 0.8103823661804199, "learning_rate": 9.99609960689886e-05, "loss": 0.1175, "step": 12390 }, { "grad_norm": 1.1075068712234497, "learning_rate": 9.996066889914109e-05, "loss": 0.0952, "step": 12400 }, { "grad_norm": 0.6826938986778259, "learning_rate": 9.996034036338766e-05, "loss": 0.0823, "step": 12410 }, { "grad_norm": 3.3124265670776367, "learning_rate": 9.996001046173728e-05, "loss": 0.1277, "step": 12420 }, { "grad_norm": 0.8083635568618774, "learning_rate": 9.9959679194199e-05, "loss": 0.1074, "step": 12430 }, { "grad_norm": 0.7762225270271301, "learning_rate": 9.995934656078186e-05, "loss": 0.0856, "step": 12440 }, { "grad_norm": 0.7045912742614746, "learning_rate": 9.995901256149494e-05, "loss": 0.1237, "step": 12450 }, { "grad_norm": 1.3325798511505127, "learning_rate": 9.99586771963474e-05, "loss": 0.0948, "step": 12460 }, { "grad_norm": 0.6620259284973145, "learning_rate": 9.995834046534839e-05, "loss": 0.1221, "step": 12470 }, { "grad_norm": 0.6798691153526306, "learning_rate": 9.995800236850712e-05, "loss": 0.0902, "step": 12480 }, { "grad_norm": 0.8909863829612732, "learning_rate": 9.995766290583283e-05, "loss": 0.0975, "step": 12490 }, { "grad_norm": 0.6984124183654785, "learning_rate": 9.99573220773348e-05, "loss": 0.117, "step": 12500 }, { "grad_norm": 0.7629874348640442, "learning_rate": 9.995697988302237e-05, "loss": 0.0831, "step": 12510 }, { "grad_norm": 0.8633049130439758, "learning_rate": 9.995663632290487e-05, "loss": 0.0917, "step": 12520 }, { "grad_norm": 1.3260425329208374, "learning_rate": 9.995629139699169e-05, "loss": 0.0884, "step": 12530 }, { "grad_norm": 0.8199912905693054, "learning_rate": 9.995594510529227e-05, "loss": 0.1098, "step": 12540 }, { "grad_norm": 0.6539613008499146, "learning_rate": 9.99555974478161e-05, "loss": 0.0861, "step": 12550 }, { "grad_norm": 0.5864291191101074, "learning_rate": 9.995524842457265e-05, "loss": 0.0952, "step": 12560 }, { "grad_norm": 0.8646998405456543, "learning_rate": 9.995489803557148e-05, "loss": 0.0902, "step": 12570 }, { "grad_norm": 0.7030901312828064, "learning_rate": 9.995454628082214e-05, "loss": 0.1107, "step": 12580 }, { "grad_norm": 0.5926002264022827, "learning_rate": 9.99541931603343e-05, "loss": 0.0906, "step": 12590 }, { "grad_norm": 0.7237545251846313, "learning_rate": 9.995383867411758e-05, "loss": 0.0866, "step": 12600 }, { "grad_norm": 0.5828007459640503, "learning_rate": 9.995348282218167e-05, "loss": 0.0919, "step": 12610 }, { "grad_norm": 0.5530545115470886, "learning_rate": 9.995312560453632e-05, "loss": 0.0765, "step": 12620 }, { "grad_norm": 0.9474378824234009, "learning_rate": 9.995276702119128e-05, "loss": 0.0917, "step": 12630 }, { "grad_norm": 0.83701491355896, "learning_rate": 9.995240707215635e-05, "loss": 0.0785, "step": 12640 }, { "grad_norm": 1.2333210706710815, "learning_rate": 9.995204575744137e-05, "loss": 0.1074, "step": 12650 }, { "grad_norm": 0.6286621689796448, "learning_rate": 9.995168307705624e-05, "loss": 0.0944, "step": 12660 }, { "grad_norm": 0.5188577175140381, "learning_rate": 9.995131903101084e-05, "loss": 0.1186, "step": 12670 }, { "grad_norm": 0.8740542531013489, "learning_rate": 9.995095361931516e-05, "loss": 0.092, "step": 12680 }, { "grad_norm": 0.6238161325454712, "learning_rate": 9.995058684197918e-05, "loss": 0.093, "step": 12690 }, { "grad_norm": 0.6366745233535767, "learning_rate": 9.995021869901291e-05, "loss": 0.0886, "step": 12700 }, { "grad_norm": 0.7961657643318176, "learning_rate": 9.994984919042642e-05, "loss": 0.1146, "step": 12710 }, { "grad_norm": 0.7434399127960205, "learning_rate": 9.994947831622983e-05, "loss": 0.0822, "step": 12720 }, { "grad_norm": 0.7943741679191589, "learning_rate": 9.994910607643327e-05, "loss": 0.0741, "step": 12730 }, { "grad_norm": 0.5922917723655701, "learning_rate": 9.99487324710469e-05, "loss": 0.1037, "step": 12740 }, { "grad_norm": 0.9335346221923828, "learning_rate": 9.994835750008096e-05, "loss": 0.0937, "step": 12750 }, { "grad_norm": 0.6454089283943176, "learning_rate": 9.994798116354568e-05, "loss": 0.0868, "step": 12760 }, { "grad_norm": 0.696412205696106, "learning_rate": 9.994760346145138e-05, "loss": 0.0889, "step": 12770 }, { "grad_norm": 1.4159343242645264, "learning_rate": 9.994722439380834e-05, "loss": 0.1149, "step": 12780 }, { "grad_norm": 0.898549497127533, "learning_rate": 9.994684396062696e-05, "loss": 0.0868, "step": 12790 }, { "grad_norm": 0.8018776178359985, "learning_rate": 9.994646216191766e-05, "loss": 0.1404, "step": 12800 }, { "grad_norm": 0.5164359211921692, "learning_rate": 9.994607899769082e-05, "loss": 0.0841, "step": 12810 }, { "grad_norm": 0.7319840788841248, "learning_rate": 9.994569446795694e-05, "loss": 0.0819, "step": 12820 }, { "grad_norm": 0.7175515294075012, "learning_rate": 9.994530857272655e-05, "loss": 0.1065, "step": 12830 }, { "grad_norm": 0.8665336966514587, "learning_rate": 9.994492131201017e-05, "loss": 0.0799, "step": 12840 }, { "grad_norm": 0.8879412412643433, "learning_rate": 9.994453268581842e-05, "loss": 0.0968, "step": 12850 }, { "grad_norm": 0.6156983971595764, "learning_rate": 9.99441426941619e-05, "loss": 0.0859, "step": 12860 }, { "grad_norm": 0.7495829463005066, "learning_rate": 9.994375133705129e-05, "loss": 0.0934, "step": 12870 }, { "grad_norm": 0.6328160762786865, "learning_rate": 9.994335861449728e-05, "loss": 0.0907, "step": 12880 }, { "grad_norm": 0.6421126127243042, "learning_rate": 9.99429645265106e-05, "loss": 0.0966, "step": 12890 }, { "grad_norm": 0.827418863773346, "learning_rate": 9.994256907310205e-05, "loss": 0.1004, "step": 12900 }, { "grad_norm": 0.7730570435523987, "learning_rate": 9.994217225428241e-05, "loss": 0.0936, "step": 12910 }, { "grad_norm": 0.5620275139808655, "learning_rate": 9.994177407006255e-05, "loss": 0.0982, "step": 12920 }, { "grad_norm": 0.6221848130226135, "learning_rate": 9.994137452045333e-05, "loss": 0.0861, "step": 12930 }, { "grad_norm": 0.5613845586776733, "learning_rate": 9.994097360546572e-05, "loss": 0.0959, "step": 12940 }, { "grad_norm": 0.6068636178970337, "learning_rate": 9.994057132511063e-05, "loss": 0.0804, "step": 12950 }, { "grad_norm": 0.6784394979476929, "learning_rate": 9.99401676793991e-05, "loss": 0.1083, "step": 12960 }, { "grad_norm": 0.732871949672699, "learning_rate": 9.993976266834214e-05, "loss": 0.0863, "step": 12970 }, { "grad_norm": 0.936134397983551, "learning_rate": 9.993935629195084e-05, "loss": 0.1224, "step": 12980 }, { "grad_norm": 0.5890068411827087, "learning_rate": 9.993894855023629e-05, "loss": 0.092, "step": 12990 }, { "grad_norm": 0.6284220814704895, "learning_rate": 9.993853944320964e-05, "loss": 0.0776, "step": 13000 }, { "grad_norm": 0.7406565546989441, "learning_rate": 9.99381289708821e-05, "loss": 0.0902, "step": 13010 }, { "grad_norm": 0.8287525773048401, "learning_rate": 9.993771713326487e-05, "loss": 0.0954, "step": 13020 }, { "grad_norm": 0.647223174571991, "learning_rate": 9.99373039303692e-05, "loss": 0.1027, "step": 13030 }, { "grad_norm": 0.8440384268760681, "learning_rate": 9.993688936220642e-05, "loss": 0.0964, "step": 13040 }, { "grad_norm": 0.6080214977264404, "learning_rate": 9.993647342878783e-05, "loss": 0.0977, "step": 13050 }, { "grad_norm": 0.7667699456214905, "learning_rate": 9.993605613012483e-05, "loss": 0.0821, "step": 13060 }, { "grad_norm": 0.9202483892440796, "learning_rate": 9.993563746622881e-05, "loss": 0.1017, "step": 13070 }, { "grad_norm": 0.6018708944320679, "learning_rate": 9.993521743711121e-05, "loss": 0.095, "step": 13080 }, { "grad_norm": 0.791617214679718, "learning_rate": 9.993479604278354e-05, "loss": 0.0971, "step": 13090 }, { "grad_norm": 0.7084881663322449, "learning_rate": 9.99343732832573e-05, "loss": 0.0908, "step": 13100 }, { "grad_norm": 0.8382473587989807, "learning_rate": 9.993394915854406e-05, "loss": 0.0898, "step": 13110 }, { "grad_norm": 0.8088220357894897, "learning_rate": 9.99335236686554e-05, "loss": 0.097, "step": 13120 }, { "grad_norm": 0.746724545955658, "learning_rate": 9.993309681360297e-05, "loss": 0.0819, "step": 13130 }, { "grad_norm": 0.790506899356842, "learning_rate": 9.993266859339842e-05, "loss": 0.1148, "step": 13140 }, { "grad_norm": 0.6869680285453796, "learning_rate": 9.993223900805348e-05, "loss": 0.0774, "step": 13150 }, { "grad_norm": 0.500083863735199, "learning_rate": 9.993180805757989e-05, "loss": 0.0868, "step": 13160 }, { "grad_norm": 1.0423212051391602, "learning_rate": 9.993137574198943e-05, "loss": 0.1194, "step": 13170 }, { "grad_norm": 0.679638147354126, "learning_rate": 9.99309420612939e-05, "loss": 0.1159, "step": 13180 }, { "grad_norm": 1.1395503282546997, "learning_rate": 9.993050701550517e-05, "loss": 0.1258, "step": 13190 }, { "grad_norm": 0.6520406007766724, "learning_rate": 9.993007060463515e-05, "loss": 0.1166, "step": 13200 }, { "grad_norm": 0.5411896705627441, "learning_rate": 9.992963282869575e-05, "loss": 0.0781, "step": 13210 }, { "grad_norm": 0.5592043995857239, "learning_rate": 9.992919368769895e-05, "loss": 0.0837, "step": 13220 }, { "grad_norm": 0.5887243747711182, "learning_rate": 9.992875318165676e-05, "loss": 0.0917, "step": 13230 }, { "grad_norm": 0.6873167753219604, "learning_rate": 9.992831131058119e-05, "loss": 0.0805, "step": 13240 }, { "grad_norm": 0.582960844039917, "learning_rate": 9.992786807448436e-05, "loss": 0.0796, "step": 13250 }, { "grad_norm": 0.8359052538871765, "learning_rate": 9.992742347337837e-05, "loss": 0.0789, "step": 13260 }, { "grad_norm": 0.7561572194099426, "learning_rate": 9.992697750727539e-05, "loss": 0.0908, "step": 13270 }, { "grad_norm": 0.5720493197441101, "learning_rate": 9.99265301761876e-05, "loss": 0.0788, "step": 13280 }, { "grad_norm": 0.741668701171875, "learning_rate": 9.992608148012722e-05, "loss": 0.0854, "step": 13290 }, { "grad_norm": 0.6915262937545776, "learning_rate": 9.992563141910654e-05, "loss": 0.1196, "step": 13300 }, { "grad_norm": 0.6977049112319946, "learning_rate": 9.992517999313784e-05, "loss": 0.0934, "step": 13310 }, { "grad_norm": 0.6832305192947388, "learning_rate": 9.992472720223349e-05, "loss": 0.0727, "step": 13320 }, { "grad_norm": 0.8576653599739075, "learning_rate": 9.992427304640584e-05, "loss": 0.0834, "step": 13330 }, { "grad_norm": 0.6862027049064636, "learning_rate": 9.992381752566733e-05, "loss": 0.0892, "step": 13340 }, { "grad_norm": 0.6628944277763367, "learning_rate": 9.992336064003041e-05, "loss": 0.1006, "step": 13350 }, { "grad_norm": 0.6249464154243469, "learning_rate": 9.992290238950754e-05, "loss": 0.0912, "step": 13360 }, { "grad_norm": 0.6408608555793762, "learning_rate": 9.99224427741113e-05, "loss": 0.1054, "step": 13370 }, { "grad_norm": 0.588215708732605, "learning_rate": 9.992198179385423e-05, "loss": 0.0942, "step": 13380 }, { "grad_norm": 0.6527678370475769, "learning_rate": 9.992151944874893e-05, "loss": 0.0912, "step": 13390 }, { "grad_norm": 0.6583364605903625, "learning_rate": 9.992105573880804e-05, "loss": 0.1087, "step": 13400 }, { "grad_norm": 0.6184884309768677, "learning_rate": 9.992059066404422e-05, "loss": 0.0929, "step": 13410 }, { "grad_norm": 0.7202732563018799, "learning_rate": 9.992012422447022e-05, "loss": 0.0965, "step": 13420 }, { "grad_norm": 0.7225697636604309, "learning_rate": 9.99196564200988e-05, "loss": 0.131, "step": 13430 }, { "grad_norm": 0.625339150428772, "learning_rate": 9.991918725094269e-05, "loss": 0.0941, "step": 13440 }, { "grad_norm": 0.760151207447052, "learning_rate": 9.991871671701479e-05, "loss": 0.0987, "step": 13450 }, { "grad_norm": 0.7753239870071411, "learning_rate": 9.99182448183279e-05, "loss": 0.1034, "step": 13460 }, { "grad_norm": 1.1804035902023315, "learning_rate": 9.991777155489496e-05, "loss": 0.1387, "step": 13470 }, { "grad_norm": 0.8524736762046814, "learning_rate": 9.99172969267289e-05, "loss": 0.1016, "step": 13480 }, { "grad_norm": 0.8856166005134583, "learning_rate": 9.99168209338427e-05, "loss": 0.0937, "step": 13490 }, { "grad_norm": 0.684822142124176, "learning_rate": 9.991634357624936e-05, "loss": 0.0797, "step": 13500 }, { "grad_norm": 0.8485726714134216, "learning_rate": 9.991586485396193e-05, "loss": 0.0951, "step": 13510 }, { "grad_norm": 0.5633108615875244, "learning_rate": 9.991538476699353e-05, "loss": 0.0938, "step": 13520 }, { "grad_norm": 0.5623761415481567, "learning_rate": 9.991490331535725e-05, "loss": 0.0906, "step": 13530 }, { "grad_norm": 0.6385155320167542, "learning_rate": 9.991442049906628e-05, "loss": 0.0812, "step": 13540 }, { "grad_norm": 0.5809624195098877, "learning_rate": 9.991393631813379e-05, "loss": 0.0871, "step": 13550 }, { "grad_norm": 0.6629335880279541, "learning_rate": 9.991345077257304e-05, "loss": 0.1026, "step": 13560 }, { "grad_norm": 0.8734977841377258, "learning_rate": 9.991296386239729e-05, "loss": 0.0908, "step": 13570 }, { "grad_norm": 1.0381520986557007, "learning_rate": 9.991247558761987e-05, "loss": 0.0783, "step": 13580 }, { "grad_norm": 0.9034871459007263, "learning_rate": 9.991198594825411e-05, "loss": 0.0908, "step": 13590 }, { "grad_norm": 0.8392394781112671, "learning_rate": 9.99114949443134e-05, "loss": 0.121, "step": 13600 }, { "grad_norm": 0.6381133794784546, "learning_rate": 9.991100257581119e-05, "loss": 0.0975, "step": 13610 }, { "grad_norm": 0.6767735481262207, "learning_rate": 9.991050884276089e-05, "loss": 0.0882, "step": 13620 }, { "grad_norm": 0.6374661922454834, "learning_rate": 9.991001374517606e-05, "loss": 0.0986, "step": 13630 }, { "grad_norm": 0.6501266956329346, "learning_rate": 9.990951728307019e-05, "loss": 0.1131, "step": 13640 }, { "grad_norm": 0.5923923254013062, "learning_rate": 9.990901945645687e-05, "loss": 0.0865, "step": 13650 }, { "grad_norm": 0.8167465329170227, "learning_rate": 9.990852026534969e-05, "loss": 0.1071, "step": 13660 }, { "grad_norm": 0.7444211840629578, "learning_rate": 9.990801970976234e-05, "loss": 0.1035, "step": 13670 }, { "grad_norm": 0.6168451309204102, "learning_rate": 9.990751778970848e-05, "loss": 0.0882, "step": 13680 }, { "grad_norm": 0.982017993927002, "learning_rate": 9.990701450520181e-05, "loss": 0.0972, "step": 13690 }, { "grad_norm": 0.694161057472229, "learning_rate": 9.990650985625613e-05, "loss": 0.0955, "step": 13700 }, { "grad_norm": 0.9315837621688843, "learning_rate": 9.990600384288523e-05, "loss": 0.1046, "step": 13710 }, { "grad_norm": 0.6427130103111267, "learning_rate": 9.990549646510291e-05, "loss": 0.0734, "step": 13720 }, { "grad_norm": 0.6334851384162903, "learning_rate": 9.990498772292309e-05, "loss": 0.1199, "step": 13730 }, { "grad_norm": 0.8884596824645996, "learning_rate": 9.990447761635964e-05, "loss": 0.1182, "step": 13740 }, { "grad_norm": 0.7000996470451355, "learning_rate": 9.990396614542652e-05, "loss": 0.0887, "step": 13750 }, { "grad_norm": 0.499691903591156, "learning_rate": 9.990345331013771e-05, "loss": 0.1192, "step": 13760 }, { "grad_norm": 0.6623421907424927, "learning_rate": 9.990293911050725e-05, "loss": 0.0896, "step": 13770 }, { "grad_norm": 0.7500148415565491, "learning_rate": 9.990242354654916e-05, "loss": 0.0967, "step": 13780 }, { "grad_norm": 0.5975396037101746, "learning_rate": 9.990190661827757e-05, "loss": 0.0822, "step": 13790 }, { "grad_norm": 0.7485140562057495, "learning_rate": 9.99013883257066e-05, "loss": 0.1119, "step": 13800 }, { "grad_norm": 0.6737509369850159, "learning_rate": 9.990086866885042e-05, "loss": 0.0846, "step": 13810 }, { "grad_norm": 0.7580047845840454, "learning_rate": 9.990034764772323e-05, "loss": 0.0887, "step": 13820 }, { "grad_norm": 0.8202012777328491, "learning_rate": 9.98998252623393e-05, "loss": 0.0972, "step": 13830 }, { "grad_norm": 0.8274821043014526, "learning_rate": 9.989930151271287e-05, "loss": 0.1165, "step": 13840 }, { "grad_norm": 0.8574158549308777, "learning_rate": 9.98987763988583e-05, "loss": 0.1018, "step": 13850 }, { "grad_norm": 0.670342743396759, "learning_rate": 9.989824992078993e-05, "loss": 0.0909, "step": 13860 }, { "grad_norm": 0.6151131391525269, "learning_rate": 9.989772207852215e-05, "loss": 0.0879, "step": 13870 }, { "grad_norm": 0.6992025375366211, "learning_rate": 9.989719287206939e-05, "loss": 0.0757, "step": 13880 }, { "grad_norm": 0.4888639450073242, "learning_rate": 9.989666230144612e-05, "loss": 0.0767, "step": 13890 }, { "grad_norm": 0.7336845397949219, "learning_rate": 9.989613036666686e-05, "loss": 0.085, "step": 13900 }, { "grad_norm": 0.8461738228797913, "learning_rate": 9.989559706774613e-05, "loss": 0.0984, "step": 13910 }, { "grad_norm": 0.7186420559883118, "learning_rate": 9.989506240469852e-05, "loss": 0.0801, "step": 13920 }, { "grad_norm": 0.667027473449707, "learning_rate": 9.989452637753867e-05, "loss": 0.0879, "step": 13930 }, { "grad_norm": 0.5247011184692383, "learning_rate": 9.989398898628119e-05, "loss": 0.0864, "step": 13940 }, { "grad_norm": 0.6410070061683655, "learning_rate": 9.98934502309408e-05, "loss": 0.0787, "step": 13950 }, { "grad_norm": 0.6736259460449219, "learning_rate": 9.989291011153224e-05, "loss": 0.1065, "step": 13960 }, { "grad_norm": 0.737144947052002, "learning_rate": 9.989236862807023e-05, "loss": 0.0713, "step": 13970 }, { "grad_norm": 0.9162158966064453, "learning_rate": 9.989182578056964e-05, "loss": 0.0883, "step": 13980 }, { "grad_norm": 0.7525148987770081, "learning_rate": 9.989128156904524e-05, "loss": 0.1009, "step": 13990 }, { "grad_norm": 0.9159949421882629, "learning_rate": 9.989073599351198e-05, "loss": 0.0845, "step": 14000 }, { "grad_norm": 0.7017428874969482, "learning_rate": 9.989018905398473e-05, "loss": 0.1095, "step": 14010 }, { "grad_norm": 0.6340438723564148, "learning_rate": 9.988964075047844e-05, "loss": 0.1058, "step": 14020 }, { "grad_norm": 0.7520162463188171, "learning_rate": 9.988909108300813e-05, "loss": 0.0863, "step": 14030 }, { "grad_norm": 0.5861961841583252, "learning_rate": 9.98885400515888e-05, "loss": 0.0965, "step": 14040 }, { "grad_norm": 0.7373747825622559, "learning_rate": 9.988798765623555e-05, "loss": 0.1164, "step": 14050 }, { "grad_norm": 0.7807149291038513, "learning_rate": 9.988743389696343e-05, "loss": 0.1042, "step": 14060 }, { "grad_norm": 0.6839778423309326, "learning_rate": 9.988687877378762e-05, "loss": 0.0894, "step": 14070 }, { "grad_norm": 0.6936168670654297, "learning_rate": 9.98863222867233e-05, "loss": 0.0907, "step": 14080 }, { "grad_norm": 0.6499001383781433, "learning_rate": 9.988576443578565e-05, "loss": 0.0778, "step": 14090 }, { "grad_norm": 0.6759138703346252, "learning_rate": 9.988520522098996e-05, "loss": 0.079, "step": 14100 }, { "grad_norm": 0.8812659382820129, "learning_rate": 9.988464464235148e-05, "loss": 0.0839, "step": 14110 }, { "grad_norm": 0.6467119455337524, "learning_rate": 9.988408269988559e-05, "loss": 0.0906, "step": 14120 }, { "grad_norm": 0.6002629399299622, "learning_rate": 9.988351939360758e-05, "loss": 0.0973, "step": 14130 }, { "grad_norm": 0.6987906694412231, "learning_rate": 9.98829547235329e-05, "loss": 0.0944, "step": 14140 }, { "grad_norm": 0.8726905584335327, "learning_rate": 9.988238868967699e-05, "loss": 0.1071, "step": 14150 }, { "grad_norm": 0.6979994177818298, "learning_rate": 9.988182129205532e-05, "loss": 0.1071, "step": 14160 }, { "grad_norm": 0.8963242173194885, "learning_rate": 9.988125253068338e-05, "loss": 0.0949, "step": 14170 }, { "grad_norm": 0.7985817790031433, "learning_rate": 9.988068240557673e-05, "loss": 0.1092, "step": 14180 }, { "grad_norm": 0.9073081612586975, "learning_rate": 9.988011091675098e-05, "loss": 0.069, "step": 14190 }, { "grad_norm": 0.8797186017036438, "learning_rate": 9.987953806422173e-05, "loss": 0.1043, "step": 14200 }, { "grad_norm": 0.5675152540206909, "learning_rate": 9.987896384800465e-05, "loss": 0.1157, "step": 14210 }, { "grad_norm": 0.7906568646430969, "learning_rate": 9.987838826811543e-05, "loss": 0.0975, "step": 14220 }, { "grad_norm": 0.5882888436317444, "learning_rate": 9.987781132456982e-05, "loss": 0.089, "step": 14230 }, { "grad_norm": 0.5259380340576172, "learning_rate": 9.987723301738358e-05, "loss": 0.0832, "step": 14240 }, { "grad_norm": 0.9866291880607605, "learning_rate": 9.987665334657253e-05, "loss": 0.0903, "step": 14250 }, { "grad_norm": 0.5202823877334595, "learning_rate": 9.987607231215251e-05, "loss": 0.0928, "step": 14260 }, { "grad_norm": 0.7495569586753845, "learning_rate": 9.987548991413942e-05, "loss": 0.1059, "step": 14270 }, { "grad_norm": 0.9348839521408081, "learning_rate": 9.987490615254917e-05, "loss": 0.1514, "step": 14280 }, { "grad_norm": 0.8281311988830566, "learning_rate": 9.987432102739773e-05, "loss": 0.1214, "step": 14290 }, { "grad_norm": 1.0045804977416992, "learning_rate": 9.987373453870107e-05, "loss": 0.1086, "step": 14300 }, { "grad_norm": 1.0343961715698242, "learning_rate": 9.987314668647526e-05, "loss": 0.0944, "step": 14310 }, { "grad_norm": 1.165006160736084, "learning_rate": 9.987255747073636e-05, "loss": 0.1119, "step": 14320 }, { "grad_norm": 0.9470888376235962, "learning_rate": 9.987196689150046e-05, "loss": 0.1159, "step": 14330 }, { "grad_norm": 0.8880698084831238, "learning_rate": 9.987137494878374e-05, "loss": 0.0947, "step": 14340 }, { "grad_norm": 0.7237434983253479, "learning_rate": 9.987078164260235e-05, "loss": 0.1053, "step": 14350 }, { "grad_norm": 0.7683876752853394, "learning_rate": 9.987018697297252e-05, "loss": 0.1007, "step": 14360 }, { "grad_norm": 0.7263169288635254, "learning_rate": 9.986959093991053e-05, "loss": 0.0957, "step": 14370 }, { "grad_norm": 0.8183110952377319, "learning_rate": 9.986899354343265e-05, "loss": 0.1002, "step": 14380 }, { "grad_norm": 0.5197404623031616, "learning_rate": 9.986839478355523e-05, "loss": 0.0888, "step": 14390 }, { "grad_norm": 1.0278186798095703, "learning_rate": 9.986779466029461e-05, "loss": 0.0939, "step": 14400 }, { "grad_norm": 0.7552916407585144, "learning_rate": 9.986719317366725e-05, "loss": 0.0794, "step": 14410 }, { "grad_norm": 0.7382943034172058, "learning_rate": 9.986659032368953e-05, "loss": 0.1119, "step": 14420 }, { "grad_norm": 0.7975528836250305, "learning_rate": 9.986598611037799e-05, "loss": 0.1028, "step": 14430 }, { "grad_norm": 0.8079559803009033, "learning_rate": 9.986538053374912e-05, "loss": 0.0974, "step": 14440 }, { "grad_norm": 0.6318050622940063, "learning_rate": 9.986477359381947e-05, "loss": 0.1017, "step": 14450 }, { "grad_norm": 0.5337705612182617, "learning_rate": 9.986416529060564e-05, "loss": 0.0981, "step": 14460 }, { "grad_norm": 1.0388864278793335, "learning_rate": 9.986355562412427e-05, "loss": 0.1283, "step": 14470 }, { "grad_norm": 0.8171789646148682, "learning_rate": 9.986294459439203e-05, "loss": 0.1082, "step": 14480 }, { "grad_norm": 0.6974749565124512, "learning_rate": 9.98623322014256e-05, "loss": 0.1013, "step": 14490 }, { "grad_norm": 0.5554207563400269, "learning_rate": 9.986171844524176e-05, "loss": 0.0842, "step": 14500 }, { "grad_norm": 0.8806113004684448, "learning_rate": 9.986110332585725e-05, "loss": 0.1021, "step": 14510 }, { "grad_norm": 0.6808772683143616, "learning_rate": 9.986048684328893e-05, "loss": 0.0863, "step": 14520 }, { "grad_norm": 0.9929015040397644, "learning_rate": 9.985986899755362e-05, "loss": 0.0997, "step": 14530 }, { "grad_norm": 0.463992714881897, "learning_rate": 9.985924978866822e-05, "loss": 0.0642, "step": 14540 }, { "grad_norm": 0.680844247341156, "learning_rate": 9.985862921664965e-05, "loss": 0.0714, "step": 14550 }, { "grad_norm": 0.5898401737213135, "learning_rate": 9.985800728151491e-05, "loss": 0.0904, "step": 14560 }, { "grad_norm": 0.6963229179382324, "learning_rate": 9.985738398328097e-05, "loss": 0.0854, "step": 14570 }, { "grad_norm": 0.8920849561691284, "learning_rate": 9.98567593219649e-05, "loss": 0.1043, "step": 14580 }, { "grad_norm": 0.567020833492279, "learning_rate": 9.985613329758375e-05, "loss": 0.0764, "step": 14590 }, { "grad_norm": 0.5076447129249573, "learning_rate": 9.985550591015463e-05, "loss": 0.0967, "step": 14600 }, { "grad_norm": 0.755061686038971, "learning_rate": 9.985487715969473e-05, "loss": 0.1005, "step": 14610 }, { "grad_norm": 0.6248704791069031, "learning_rate": 9.985424704622119e-05, "loss": 0.0804, "step": 14620 }, { "grad_norm": 0.627886950969696, "learning_rate": 9.985361556975131e-05, "loss": 0.0686, "step": 14630 }, { "grad_norm": 0.6368727684020996, "learning_rate": 9.985298273030226e-05, "loss": 0.0823, "step": 14640 }, { "grad_norm": 0.8077681660652161, "learning_rate": 9.985234852789142e-05, "loss": 0.0987, "step": 14650 }, { "grad_norm": 0.8428466320037842, "learning_rate": 9.98517129625361e-05, "loss": 0.0933, "step": 14660 }, { "grad_norm": 0.8216338753700256, "learning_rate": 9.985107603425367e-05, "loss": 0.0967, "step": 14670 }, { "grad_norm": 0.8728350400924683, "learning_rate": 9.985043774306155e-05, "loss": 0.0997, "step": 14680 }, { "grad_norm": 0.7882518172264099, "learning_rate": 9.984979808897719e-05, "loss": 0.0828, "step": 14690 }, { "grad_norm": 0.8808624744415283, "learning_rate": 9.984915707201807e-05, "loss": 0.0856, "step": 14700 }, { "grad_norm": 0.7228912115097046, "learning_rate": 9.984851469220174e-05, "loss": 0.0909, "step": 14710 }, { "grad_norm": 0.7677212953567505, "learning_rate": 9.984787094954573e-05, "loss": 0.0824, "step": 14720 }, { "grad_norm": 0.4740472435951233, "learning_rate": 9.984722584406767e-05, "loss": 0.0764, "step": 14730 }, { "grad_norm": 0.7594956755638123, "learning_rate": 9.984657937578516e-05, "loss": 0.0907, "step": 14740 }, { "grad_norm": 0.7848475575447083, "learning_rate": 9.984593154471592e-05, "loss": 0.0814, "step": 14750 }, { "grad_norm": 0.8399959802627563, "learning_rate": 9.984528235087763e-05, "loss": 0.0874, "step": 14760 }, { "grad_norm": 0.5924437046051025, "learning_rate": 9.984463179428804e-05, "loss": 0.0872, "step": 14770 }, { "grad_norm": 0.6154962182044983, "learning_rate": 9.984397987496494e-05, "loss": 0.0911, "step": 14780 }, { "grad_norm": 1.0758482217788696, "learning_rate": 9.984332659292616e-05, "loss": 0.099, "step": 14790 }, { "grad_norm": 0.8979820013046265, "learning_rate": 9.984267194818956e-05, "loss": 0.0857, "step": 14800 }, { "grad_norm": 0.7430589199066162, "learning_rate": 9.984201594077304e-05, "loss": 0.0854, "step": 14810 }, { "grad_norm": 0.6606823801994324, "learning_rate": 9.984135857069452e-05, "loss": 0.1049, "step": 14820 }, { "grad_norm": 0.6499386429786682, "learning_rate": 9.984069983797199e-05, "loss": 0.0823, "step": 14830 }, { "grad_norm": 0.8887962102890015, "learning_rate": 9.984003974262344e-05, "loss": 0.119, "step": 14840 }, { "grad_norm": 0.6577650904655457, "learning_rate": 9.983937828466694e-05, "loss": 0.1063, "step": 14850 }, { "grad_norm": 0.467933714389801, "learning_rate": 9.983871546412054e-05, "loss": 0.1008, "step": 14860 }, { "grad_norm": 0.635852038860321, "learning_rate": 9.98380512810024e-05, "loss": 0.1006, "step": 14870 }, { "grad_norm": 0.7254961133003235, "learning_rate": 9.983738573533065e-05, "loss": 0.0866, "step": 14880 }, { "grad_norm": 0.6622424721717834, "learning_rate": 9.983671882712351e-05, "loss": 0.0921, "step": 14890 }, { "grad_norm": 0.5705195665359497, "learning_rate": 9.983605055639919e-05, "loss": 0.0865, "step": 14900 }, { "grad_norm": 0.837651252746582, "learning_rate": 9.983538092317599e-05, "loss": 0.0959, "step": 14910 }, { "grad_norm": 0.48562636971473694, "learning_rate": 9.983470992747218e-05, "loss": 0.0844, "step": 14920 }, { "grad_norm": 0.745760440826416, "learning_rate": 9.983403756930611e-05, "loss": 0.0981, "step": 14930 }, { "grad_norm": 0.7505891919136047, "learning_rate": 9.98333638486962e-05, "loss": 0.1013, "step": 14940 }, { "grad_norm": 1.1507312059402466, "learning_rate": 9.983268876566083e-05, "loss": 0.1026, "step": 14950 }, { "grad_norm": 0.7467960715293884, "learning_rate": 9.983201232021847e-05, "loss": 0.107, "step": 14960 }, { "grad_norm": 0.875785231590271, "learning_rate": 9.983133451238762e-05, "loss": 0.1096, "step": 14970 }, { "grad_norm": 0.5358511805534363, "learning_rate": 9.98306553421868e-05, "loss": 0.1025, "step": 14980 }, { "grad_norm": 0.4432365596294403, "learning_rate": 9.982997480963458e-05, "loss": 0.0865, "step": 14990 }, { "grad_norm": 0.6598250865936279, "learning_rate": 9.982929291474958e-05, "loss": 0.094, "step": 15000 }, { "grad_norm": 0.5684540271759033, "learning_rate": 9.982860965755043e-05, "loss": 0.0795, "step": 15010 }, { "grad_norm": 0.9139082431793213, "learning_rate": 9.98279250380558e-05, "loss": 0.0854, "step": 15020 }, { "grad_norm": 0.4839598536491394, "learning_rate": 9.982723905628444e-05, "loss": 0.0915, "step": 15030 }, { "grad_norm": 0.5063598155975342, "learning_rate": 9.982655171225507e-05, "loss": 0.0822, "step": 15040 }, { "grad_norm": 0.5641692876815796, "learning_rate": 9.982586300598648e-05, "loss": 0.0686, "step": 15050 }, { "grad_norm": 0.6005620956420898, "learning_rate": 9.982517293749755e-05, "loss": 0.0692, "step": 15060 }, { "grad_norm": 0.894536554813385, "learning_rate": 9.98244815068071e-05, "loss": 0.0727, "step": 15070 }, { "grad_norm": 0.6182859539985657, "learning_rate": 9.982378871393402e-05, "loss": 0.0924, "step": 15080 }, { "grad_norm": 0.434815913438797, "learning_rate": 9.982309455889729e-05, "loss": 0.0709, "step": 15090 }, { "grad_norm": 0.7723525762557983, "learning_rate": 9.982239904171587e-05, "loss": 0.0998, "step": 15100 }, { "grad_norm": 0.5809647440910339, "learning_rate": 9.982170216240877e-05, "loss": 0.0795, "step": 15110 }, { "grad_norm": 0.7315045595169067, "learning_rate": 9.982100392099507e-05, "loss": 0.0849, "step": 15120 }, { "grad_norm": 1.0100011825561523, "learning_rate": 9.982030431749383e-05, "loss": 0.1084, "step": 15130 }, { "grad_norm": 0.9726934432983398, "learning_rate": 9.981960335192418e-05, "loss": 0.0893, "step": 15140 }, { "grad_norm": 0.5797505378723145, "learning_rate": 9.98189010243053e-05, "loss": 0.0809, "step": 15150 }, { "grad_norm": 0.7685850262641907, "learning_rate": 9.981819733465639e-05, "loss": 0.1092, "step": 15160 }, { "grad_norm": 0.5469709038734436, "learning_rate": 9.981749228299666e-05, "loss": 0.1019, "step": 15170 }, { "grad_norm": 0.5273554921150208, "learning_rate": 9.981678586934541e-05, "loss": 0.0927, "step": 15180 }, { "grad_norm": 1.1168227195739746, "learning_rate": 9.981607809372195e-05, "loss": 0.123, "step": 15190 }, { "grad_norm": 0.5026647448539734, "learning_rate": 9.981536895614564e-05, "loss": 0.0959, "step": 15200 }, { "grad_norm": 0.9137570858001709, "learning_rate": 9.981465845663584e-05, "loss": 0.1267, "step": 15210 }, { "grad_norm": 0.9388868808746338, "learning_rate": 9.981394659521202e-05, "loss": 0.0838, "step": 15220 }, { "grad_norm": 0.6480067372322083, "learning_rate": 9.98132333718936e-05, "loss": 0.1082, "step": 15230 }, { "grad_norm": 0.7978304624557495, "learning_rate": 9.981251878670008e-05, "loss": 0.0801, "step": 15240 }, { "grad_norm": 0.705848217010498, "learning_rate": 9.981180283965104e-05, "loss": 0.1044, "step": 15250 }, { "grad_norm": 0.642998993396759, "learning_rate": 9.981108553076599e-05, "loss": 0.0978, "step": 15260 }, { "grad_norm": 0.6401437520980835, "learning_rate": 9.981036686006459e-05, "loss": 0.076, "step": 15270 }, { "grad_norm": 0.6914892196655273, "learning_rate": 9.980964682756648e-05, "loss": 0.0973, "step": 15280 }, { "grad_norm": 0.5615313649177551, "learning_rate": 9.980892543329135e-05, "loss": 0.0777, "step": 15290 }, { "grad_norm": 0.8008688688278198, "learning_rate": 9.980820267725888e-05, "loss": 0.09, "step": 15300 }, { "grad_norm": 0.6160742044448853, "learning_rate": 9.980747855948887e-05, "loss": 0.0861, "step": 15310 }, { "grad_norm": 0.6481934785842896, "learning_rate": 9.980675308000111e-05, "loss": 0.0862, "step": 15320 }, { "grad_norm": 0.5813071131706238, "learning_rate": 9.980602623881544e-05, "loss": 0.0862, "step": 15330 }, { "grad_norm": 1.0799304246902466, "learning_rate": 9.980529803595172e-05, "loss": 0.095, "step": 15340 }, { "grad_norm": 0.6144507527351379, "learning_rate": 9.980456847142986e-05, "loss": 0.0864, "step": 15350 }, { "grad_norm": 0.9120813608169556, "learning_rate": 9.980383754526981e-05, "loss": 0.1057, "step": 15360 }, { "grad_norm": 0.7949371933937073, "learning_rate": 9.980310525749154e-05, "loss": 0.0996, "step": 15370 }, { "grad_norm": 0.6709441542625427, "learning_rate": 9.98023716081151e-05, "loss": 0.0792, "step": 15380 }, { "grad_norm": 0.6637035012245178, "learning_rate": 9.980163659716053e-05, "loss": 0.0923, "step": 15390 }, { "grad_norm": 0.6378656029701233, "learning_rate": 9.980090022464792e-05, "loss": 0.0759, "step": 15400 }, { "grad_norm": 0.8213024735450745, "learning_rate": 9.980016249059741e-05, "loss": 0.0828, "step": 15410 }, { "grad_norm": 0.8451032638549805, "learning_rate": 9.979942339502917e-05, "loss": 0.1236, "step": 15420 }, { "grad_norm": 0.731095552444458, "learning_rate": 9.979868293796339e-05, "loss": 0.0963, "step": 15430 }, { "grad_norm": 0.5021370053291321, "learning_rate": 9.979794111942033e-05, "loss": 0.0907, "step": 15440 }, { "grad_norm": 0.7625073194503784, "learning_rate": 9.979719793942027e-05, "loss": 0.0971, "step": 15450 }, { "grad_norm": 0.5467134714126587, "learning_rate": 9.979645339798354e-05, "loss": 0.0899, "step": 15460 }, { "grad_norm": 0.6784570217132568, "learning_rate": 9.979570749513045e-05, "loss": 0.0827, "step": 15470 }, { "grad_norm": 1.4714593887329102, "learning_rate": 9.979496023088145e-05, "loss": 0.1037, "step": 15480 }, { "grad_norm": 0.7312665581703186, "learning_rate": 9.979421160525694e-05, "loss": 0.0681, "step": 15490 }, { "grad_norm": 0.6686477065086365, "learning_rate": 9.979346161827739e-05, "loss": 0.0765, "step": 15500 }, { "grad_norm": 0.620975911617279, "learning_rate": 9.979271026996331e-05, "loss": 0.0835, "step": 15510 }, { "grad_norm": 0.5867982506752014, "learning_rate": 9.979195756033523e-05, "loss": 0.0911, "step": 15520 }, { "grad_norm": 0.7147915363311768, "learning_rate": 9.979120348941374e-05, "loss": 0.0854, "step": 15530 }, { "grad_norm": 0.6514164805412292, "learning_rate": 9.979044805721944e-05, "loss": 0.0847, "step": 15540 }, { "grad_norm": 0.5547724962234497, "learning_rate": 9.978969126377302e-05, "loss": 0.0811, "step": 15550 }, { "grad_norm": 0.9799763560295105, "learning_rate": 9.978893310909513e-05, "loss": 0.0835, "step": 15560 }, { "grad_norm": 0.6053034663200378, "learning_rate": 9.978817359320651e-05, "loss": 0.0885, "step": 15570 }, { "grad_norm": 0.7175306081771851, "learning_rate": 9.978741271612795e-05, "loss": 0.0918, "step": 15580 }, { "grad_norm": 0.8275865316390991, "learning_rate": 9.97866504778802e-05, "loss": 0.0836, "step": 15590 }, { "grad_norm": 0.6729199290275574, "learning_rate": 9.978588687848415e-05, "loss": 0.087, "step": 15600 }, { "grad_norm": 0.8162228465080261, "learning_rate": 9.978512191796065e-05, "loss": 0.0845, "step": 15610 }, { "grad_norm": 0.6448790431022644, "learning_rate": 9.978435559633064e-05, "loss": 0.0805, "step": 15620 }, { "grad_norm": 0.6115692257881165, "learning_rate": 9.978358791361501e-05, "loss": 0.077, "step": 15630 }, { "grad_norm": 0.8174669742584229, "learning_rate": 9.978281886983483e-05, "loss": 0.0745, "step": 15640 }, { "grad_norm": 0.5487907528877258, "learning_rate": 9.978204846501107e-05, "loss": 0.0833, "step": 15650 }, { "grad_norm": 0.8623830676078796, "learning_rate": 9.97812766991648e-05, "loss": 0.0694, "step": 15660 }, { "grad_norm": 0.9039743542671204, "learning_rate": 9.978050357231713e-05, "loss": 0.0813, "step": 15670 }, { "grad_norm": 0.6034120917320251, "learning_rate": 9.977972908448919e-05, "loss": 0.0697, "step": 15680 }, { "grad_norm": 0.9835641384124756, "learning_rate": 9.977895323570217e-05, "loss": 0.0951, "step": 15690 }, { "grad_norm": 0.6641436219215393, "learning_rate": 9.977817602597726e-05, "loss": 0.0726, "step": 15700 }, { "grad_norm": 0.7602996230125427, "learning_rate": 9.977739745533572e-05, "loss": 0.1048, "step": 15710 }, { "grad_norm": 0.7707087993621826, "learning_rate": 9.977661752379883e-05, "loss": 0.0854, "step": 15720 }, { "grad_norm": 0.6534422636032104, "learning_rate": 9.977583623138793e-05, "loss": 0.0899, "step": 15730 }, { "grad_norm": 0.9226329326629639, "learning_rate": 9.977505357812436e-05, "loss": 0.1064, "step": 15740 }, { "grad_norm": 0.4796634912490845, "learning_rate": 9.977426956402953e-05, "loss": 0.0976, "step": 15750 }, { "grad_norm": 0.786206841468811, "learning_rate": 9.977348418912486e-05, "loss": 0.0886, "step": 15760 }, { "grad_norm": 1.1541612148284912, "learning_rate": 9.977269745343183e-05, "loss": 0.0849, "step": 15770 }, { "grad_norm": 0.5915826559066772, "learning_rate": 9.977190935697196e-05, "loss": 0.089, "step": 15780 }, { "grad_norm": 1.269350528717041, "learning_rate": 9.977111989976678e-05, "loss": 0.1003, "step": 15790 }, { "grad_norm": 0.7163456082344055, "learning_rate": 9.977032908183789e-05, "loss": 0.1028, "step": 15800 }, { "grad_norm": 0.6705422401428223, "learning_rate": 9.97695369032069e-05, "loss": 0.1098, "step": 15810 }, { "grad_norm": 0.6661913394927979, "learning_rate": 9.976874336389546e-05, "loss": 0.0839, "step": 15820 }, { "grad_norm": 0.6790289282798767, "learning_rate": 9.976794846392527e-05, "loss": 0.1069, "step": 15830 }, { "grad_norm": 0.55353182554245, "learning_rate": 9.976715220331807e-05, "loss": 0.1117, "step": 15840 }, { "grad_norm": 0.5965957045555115, "learning_rate": 9.976635458209563e-05, "loss": 0.08, "step": 15850 }, { "grad_norm": 1.072232961654663, "learning_rate": 9.976555560027975e-05, "loss": 0.0809, "step": 15860 }, { "grad_norm": 0.4489087164402008, "learning_rate": 9.976475525789227e-05, "loss": 0.077, "step": 15870 }, { "grad_norm": 0.7485802173614502, "learning_rate": 9.976395355495509e-05, "loss": 0.1029, "step": 15880 }, { "grad_norm": 0.38592275977134705, "learning_rate": 9.976315049149009e-05, "loss": 0.0922, "step": 15890 }, { "grad_norm": 0.9122754335403442, "learning_rate": 9.976234606751928e-05, "loss": 0.0827, "step": 15900 }, { "grad_norm": 0.9412005543708801, "learning_rate": 9.976154028306461e-05, "loss": 0.0872, "step": 15910 }, { "grad_norm": 0.9074420928955078, "learning_rate": 9.976073313814814e-05, "loss": 0.1049, "step": 15920 }, { "grad_norm": 0.7841970920562744, "learning_rate": 9.975992463279189e-05, "loss": 0.0787, "step": 15930 }, { "grad_norm": 0.8374057412147522, "learning_rate": 9.975911476701801e-05, "loss": 0.1034, "step": 15940 }, { "grad_norm": 0.6384365558624268, "learning_rate": 9.975830354084863e-05, "loss": 0.0994, "step": 15950 }, { "grad_norm": 0.7978820204734802, "learning_rate": 9.975749095430591e-05, "loss": 0.0877, "step": 15960 }, { "grad_norm": 0.7582694292068481, "learning_rate": 9.97566770074121e-05, "loss": 0.092, "step": 15970 }, { "grad_norm": 0.5466494560241699, "learning_rate": 9.975586170018943e-05, "loss": 0.0922, "step": 15980 }, { "grad_norm": 0.5786107778549194, "learning_rate": 9.975504503266017e-05, "loss": 0.0813, "step": 15990 }, { "grad_norm": 0.6245849132537842, "learning_rate": 9.975422700484669e-05, "loss": 0.0792, "step": 16000 }, { "grad_norm": 0.7849588394165039, "learning_rate": 9.975340761677134e-05, "loss": 0.0869, "step": 16010 }, { "grad_norm": 0.9874016046524048, "learning_rate": 9.97525868684565e-05, "loss": 0.0835, "step": 16020 }, { "grad_norm": 0.7569478154182434, "learning_rate": 9.975176475992465e-05, "loss": 0.1029, "step": 16030 }, { "grad_norm": 0.7636578679084778, "learning_rate": 9.975094129119822e-05, "loss": 0.0737, "step": 16040 }, { "grad_norm": 0.9679352641105652, "learning_rate": 9.975011646229976e-05, "loss": 0.0816, "step": 16050 }, { "grad_norm": 0.4429354965686798, "learning_rate": 9.974929027325179e-05, "loss": 0.0669, "step": 16060 }, { "grad_norm": 0.5721522569656372, "learning_rate": 9.974846272407694e-05, "loss": 0.0708, "step": 16070 }, { "grad_norm": 0.7356047630310059, "learning_rate": 9.97476338147978e-05, "loss": 0.0709, "step": 16080 }, { "grad_norm": 0.6719717979431152, "learning_rate": 9.974680354543702e-05, "loss": 0.0894, "step": 16090 }, { "grad_norm": 0.7901291847229004, "learning_rate": 9.974597191601733e-05, "loss": 0.0958, "step": 16100 }, { "grad_norm": 0.7909145951271057, "learning_rate": 9.974513892656147e-05, "loss": 0.0914, "step": 16110 }, { "grad_norm": 0.5764238834381104, "learning_rate": 9.974430457709218e-05, "loss": 0.0874, "step": 16120 }, { "grad_norm": 0.738592803478241, "learning_rate": 9.974346886763231e-05, "loss": 0.0977, "step": 16130 }, { "grad_norm": 0.6818479895591736, "learning_rate": 9.974263179820469e-05, "loss": 0.0773, "step": 16140 }, { "grad_norm": 0.515086829662323, "learning_rate": 9.97417933688322e-05, "loss": 0.0923, "step": 16150 }, { "grad_norm": 0.6285178661346436, "learning_rate": 9.974095357953774e-05, "loss": 0.0948, "step": 16160 }, { "grad_norm": 1.1739271879196167, "learning_rate": 9.974011243034434e-05, "loss": 0.0979, "step": 16170 }, { "grad_norm": 0.6989249587059021, "learning_rate": 9.973926992127491e-05, "loss": 0.0845, "step": 16180 }, { "grad_norm": 0.6702224016189575, "learning_rate": 9.973842605235255e-05, "loss": 0.0948, "step": 16190 }, { "grad_norm": 0.6438866853713989, "learning_rate": 9.973758082360031e-05, "loss": 0.0767, "step": 16200 }, { "grad_norm": 0.6642608642578125, "learning_rate": 9.973673423504128e-05, "loss": 0.0957, "step": 16210 }, { "grad_norm": 0.6129983067512512, "learning_rate": 9.973588628669865e-05, "loss": 0.0791, "step": 16220 }, { "grad_norm": 0.6945782899856567, "learning_rate": 9.973503697859554e-05, "loss": 0.0918, "step": 16230 }, { "grad_norm": 0.6553422808647156, "learning_rate": 9.973418631075521e-05, "loss": 0.0883, "step": 16240 }, { "grad_norm": 0.6231864094734192, "learning_rate": 9.973333428320094e-05, "loss": 0.0822, "step": 16250 }, { "grad_norm": 0.6414456963539124, "learning_rate": 9.973248089595598e-05, "loss": 0.1185, "step": 16260 }, { "grad_norm": 0.7424886226654053, "learning_rate": 9.973162614904365e-05, "loss": 0.1093, "step": 16270 }, { "grad_norm": 0.49330848455429077, "learning_rate": 9.973077004248737e-05, "loss": 0.0939, "step": 16280 }, { "grad_norm": 0.6955891251564026, "learning_rate": 9.972991257631052e-05, "loss": 0.0747, "step": 16290 }, { "grad_norm": 0.9262967109680176, "learning_rate": 9.972905375053654e-05, "loss": 0.0912, "step": 16300 }, { "grad_norm": 0.8019174337387085, "learning_rate": 9.972819356518889e-05, "loss": 0.102, "step": 16310 }, { "grad_norm": 0.7202942967414856, "learning_rate": 9.972733202029115e-05, "loss": 0.077, "step": 16320 }, { "grad_norm": 0.7383132576942444, "learning_rate": 9.97264691158668e-05, "loss": 0.0984, "step": 16330 }, { "grad_norm": 0.8689553737640381, "learning_rate": 9.972560485193949e-05, "loss": 0.0829, "step": 16340 }, { "grad_norm": 0.6058828830718994, "learning_rate": 9.972473922853281e-05, "loss": 0.0807, "step": 16350 }, { "grad_norm": 0.5576794743537903, "learning_rate": 9.972387224567045e-05, "loss": 0.0819, "step": 16360 }, { "grad_norm": 0.6333582997322083, "learning_rate": 9.97230039033761e-05, "loss": 0.0726, "step": 16370 }, { "grad_norm": 0.945807695388794, "learning_rate": 9.972213420167349e-05, "loss": 0.0892, "step": 16380 }, { "grad_norm": 0.5635355114936829, "learning_rate": 9.972126314058643e-05, "loss": 0.0739, "step": 16390 }, { "grad_norm": 0.8903405070304871, "learning_rate": 9.97203907201387e-05, "loss": 0.0811, "step": 16400 }, { "grad_norm": 0.5757383108139038, "learning_rate": 9.971951694035417e-05, "loss": 0.0628, "step": 16410 }, { "grad_norm": 0.6837708950042725, "learning_rate": 9.971864180125674e-05, "loss": 0.0737, "step": 16420 }, { "grad_norm": 0.6155446171760559, "learning_rate": 9.97177653028703e-05, "loss": 0.0853, "step": 16430 }, { "grad_norm": 0.520035982131958, "learning_rate": 9.971688744521885e-05, "loss": 0.075, "step": 16440 }, { "grad_norm": 0.6030029058456421, "learning_rate": 9.971600822832636e-05, "loss": 0.0871, "step": 16450 }, { "grad_norm": 0.6059169173240662, "learning_rate": 9.971512765221691e-05, "loss": 0.0737, "step": 16460 }, { "grad_norm": 0.8220247626304626, "learning_rate": 9.971424571691452e-05, "loss": 0.0891, "step": 16470 }, { "grad_norm": 0.6888749003410339, "learning_rate": 9.971336242244333e-05, "loss": 0.0812, "step": 16480 }, { "grad_norm": 0.8521181344985962, "learning_rate": 9.971247776882751e-05, "loss": 0.0859, "step": 16490 }, { "grad_norm": 0.5192955732345581, "learning_rate": 9.971159175609121e-05, "loss": 0.083, "step": 16500 }, { "grad_norm": 0.47731390595436096, "learning_rate": 9.971070438425866e-05, "loss": 0.0799, "step": 16510 }, { "grad_norm": 0.6317752003669739, "learning_rate": 9.970981565335415e-05, "loss": 0.0876, "step": 16520 }, { "grad_norm": 0.9341726303100586, "learning_rate": 9.970892556340194e-05, "loss": 0.0785, "step": 16530 }, { "grad_norm": 0.8299347758293152, "learning_rate": 9.970803411442637e-05, "loss": 0.0843, "step": 16540 }, { "grad_norm": 0.6145433783531189, "learning_rate": 9.970714130645184e-05, "loss": 0.0881, "step": 16550 }, { "grad_norm": 0.7003065347671509, "learning_rate": 9.970624713950273e-05, "loss": 0.0968, "step": 16560 }, { "grad_norm": 0.7769551873207092, "learning_rate": 9.97053516136035e-05, "loss": 0.082, "step": 16570 }, { "grad_norm": 0.6988995671272278, "learning_rate": 9.970445472877863e-05, "loss": 0.0787, "step": 16580 }, { "grad_norm": 0.7976076602935791, "learning_rate": 9.970355648505264e-05, "loss": 0.091, "step": 16590 }, { "grad_norm": 0.7286537289619446, "learning_rate": 9.970265688245009e-05, "loss": 0.0943, "step": 16600 }, { "grad_norm": 1.0062934160232544, "learning_rate": 9.970175592099557e-05, "loss": 0.0917, "step": 16610 }, { "grad_norm": 0.6393295526504517, "learning_rate": 9.970085360071371e-05, "loss": 0.0825, "step": 16620 }, { "grad_norm": 0.7317168116569519, "learning_rate": 9.969994992162921e-05, "loss": 0.0716, "step": 16630 }, { "grad_norm": 0.6048678755760193, "learning_rate": 9.969904488376673e-05, "loss": 0.0708, "step": 16640 }, { "grad_norm": 0.6532891392707825, "learning_rate": 9.969813848715103e-05, "loss": 0.0769, "step": 16650 }, { "grad_norm": 0.7148872017860413, "learning_rate": 9.96972307318069e-05, "loss": 0.0922, "step": 16660 }, { "grad_norm": 0.9050301909446716, "learning_rate": 9.969632161775916e-05, "loss": 0.0711, "step": 16670 }, { "grad_norm": 0.6361347436904907, "learning_rate": 9.969541114503265e-05, "loss": 0.0804, "step": 16680 }, { "grad_norm": 0.4706277847290039, "learning_rate": 9.969449931365227e-05, "loss": 0.0999, "step": 16690 }, { "grad_norm": 0.5036236047744751, "learning_rate": 9.969358612364294e-05, "loss": 0.0687, "step": 16700 }, { "grad_norm": 0.5255247950553894, "learning_rate": 9.969267157502964e-05, "loss": 0.0784, "step": 16710 }, { "grad_norm": 1.0990986824035645, "learning_rate": 9.969175566783738e-05, "loss": 0.0794, "step": 16720 }, { "grad_norm": 0.8601037859916687, "learning_rate": 9.969083840209116e-05, "loss": 0.0745, "step": 16730 }, { "grad_norm": 0.5070042610168457, "learning_rate": 9.968991977781613e-05, "loss": 0.0784, "step": 16740 }, { "grad_norm": 0.7283571362495422, "learning_rate": 9.968899979503731e-05, "loss": 0.0917, "step": 16750 }, { "grad_norm": 0.6976105570793152, "learning_rate": 9.968807845377994e-05, "loss": 0.0933, "step": 16760 }, { "grad_norm": 1.5400121212005615, "learning_rate": 9.968715575406916e-05, "loss": 0.0823, "step": 16770 }, { "grad_norm": 0.6322422623634338, "learning_rate": 9.96862316959302e-05, "loss": 0.0881, "step": 16780 }, { "grad_norm": 0.8966118097305298, "learning_rate": 9.968530627938834e-05, "loss": 0.0734, "step": 16790 }, { "grad_norm": 0.7664239406585693, "learning_rate": 9.968437950446887e-05, "loss": 0.076, "step": 16800 }, { "grad_norm": 0.7740342617034912, "learning_rate": 9.968345137119715e-05, "loss": 0.0895, "step": 16810 }, { "grad_norm": 1.5450701713562012, "learning_rate": 9.96825218795985e-05, "loss": 0.077, "step": 16820 }, { "grad_norm": 1.1003059148788452, "learning_rate": 9.968159102969839e-05, "loss": 0.0767, "step": 16830 }, { "grad_norm": 0.5787093043327332, "learning_rate": 9.968065882152225e-05, "loss": 0.104, "step": 16840 }, { "grad_norm": 0.6349239945411682, "learning_rate": 9.967972525509554e-05, "loss": 0.068, "step": 16850 }, { "grad_norm": 0.5566478371620178, "learning_rate": 9.967879033044382e-05, "loss": 0.0891, "step": 16860 }, { "grad_norm": 0.7094630002975464, "learning_rate": 9.967785404759263e-05, "loss": 0.0913, "step": 16870 }, { "grad_norm": 0.5170134902000427, "learning_rate": 9.967691640656757e-05, "loss": 0.08, "step": 16880 }, { "grad_norm": 0.780204176902771, "learning_rate": 9.967597740739429e-05, "loss": 0.0833, "step": 16890 }, { "grad_norm": 0.8095187544822693, "learning_rate": 9.967503705009845e-05, "loss": 0.0816, "step": 16900 }, { "grad_norm": 0.7822887897491455, "learning_rate": 9.967409533470575e-05, "loss": 0.0891, "step": 16910 }, { "grad_norm": 0.6480308771133423, "learning_rate": 9.967315226124194e-05, "loss": 0.078, "step": 16920 }, { "grad_norm": 0.5769035816192627, "learning_rate": 9.967220782973282e-05, "loss": 0.0727, "step": 16930 }, { "grad_norm": 0.6693897247314453, "learning_rate": 9.967126204020421e-05, "loss": 0.0852, "step": 16940 }, { "grad_norm": 0.5876063704490662, "learning_rate": 9.967031489268193e-05, "loss": 0.069, "step": 16950 }, { "grad_norm": 0.6530678272247314, "learning_rate": 9.96693663871919e-05, "loss": 0.0724, "step": 16960 }, { "grad_norm": 0.6864531636238098, "learning_rate": 9.966841652376009e-05, "loss": 0.0769, "step": 16970 }, { "grad_norm": 0.7381351590156555, "learning_rate": 9.96674653024124e-05, "loss": 0.0832, "step": 16980 }, { "grad_norm": 0.7733074426651001, "learning_rate": 9.966651272317487e-05, "loss": 0.0848, "step": 16990 }, { "grad_norm": 0.9478386044502258, "learning_rate": 9.966555878607355e-05, "loss": 0.0995, "step": 17000 }, { "grad_norm": 0.5651631355285645, "learning_rate": 9.96646034911345e-05, "loss": 0.0869, "step": 17010 }, { "grad_norm": 0.5441485047340393, "learning_rate": 9.966364683838388e-05, "loss": 0.0792, "step": 17020 }, { "grad_norm": 0.5794316530227661, "learning_rate": 9.966268882784777e-05, "loss": 0.0784, "step": 17030 }, { "grad_norm": 0.7673960328102112, "learning_rate": 9.966172945955243e-05, "loss": 0.072, "step": 17040 }, { "grad_norm": 0.7322865724563599, "learning_rate": 9.966076873352405e-05, "loss": 0.076, "step": 17050 }, { "grad_norm": 0.7641483545303345, "learning_rate": 9.965980664978893e-05, "loss": 0.0915, "step": 17060 }, { "grad_norm": 0.5901033282279968, "learning_rate": 9.965884320837334e-05, "loss": 0.0911, "step": 17070 }, { "grad_norm": 0.5155192613601685, "learning_rate": 9.965787840930363e-05, "loss": 0.0922, "step": 17080 }, { "grad_norm": 0.5234880447387695, "learning_rate": 9.965691225260619e-05, "loss": 0.0866, "step": 17090 }, { "grad_norm": 0.5877949595451355, "learning_rate": 9.965594473830742e-05, "loss": 0.0658, "step": 17100 }, { "grad_norm": 0.5594019889831543, "learning_rate": 9.965497586643377e-05, "loss": 0.0714, "step": 17110 }, { "grad_norm": 0.5964386463165283, "learning_rate": 9.965400563701172e-05, "loss": 0.0765, "step": 17120 }, { "grad_norm": 0.8139293193817139, "learning_rate": 9.965303405006784e-05, "loss": 0.1513, "step": 17130 }, { "grad_norm": 0.6801536679267883, "learning_rate": 9.965206110562864e-05, "loss": 0.083, "step": 17140 }, { "grad_norm": 0.7077668905258179, "learning_rate": 9.965108680372076e-05, "loss": 0.0744, "step": 17150 }, { "grad_norm": 0.778688907623291, "learning_rate": 9.965011114437082e-05, "loss": 0.093, "step": 17160 }, { "grad_norm": 0.8326533436775208, "learning_rate": 9.964913412760549e-05, "loss": 0.0896, "step": 17170 }, { "grad_norm": 0.6085628271102905, "learning_rate": 9.964815575345148e-05, "loss": 0.0649, "step": 17180 }, { "grad_norm": 0.6609655618667603, "learning_rate": 9.964717602193554e-05, "loss": 0.0661, "step": 17190 }, { "grad_norm": 0.5500608086585999, "learning_rate": 9.964619493308447e-05, "loss": 0.0641, "step": 17200 }, { "grad_norm": 0.6532570123672485, "learning_rate": 9.964521248692507e-05, "loss": 0.0722, "step": 17210 }, { "grad_norm": 0.8695315718650818, "learning_rate": 9.964422868348421e-05, "loss": 0.0716, "step": 17220 }, { "grad_norm": 0.6510111689567566, "learning_rate": 9.96432435227888e-05, "loss": 0.0759, "step": 17230 }, { "grad_norm": 0.7046926617622375, "learning_rate": 9.964225700486576e-05, "loss": 0.0939, "step": 17240 }, { "grad_norm": 0.5189973711967468, "learning_rate": 9.964126912974207e-05, "loss": 0.0669, "step": 17250 }, { "grad_norm": 0.5210225582122803, "learning_rate": 9.964027989744471e-05, "loss": 0.0713, "step": 17260 }, { "grad_norm": 0.5487545728683472, "learning_rate": 9.963928930800077e-05, "loss": 0.0862, "step": 17270 }, { "grad_norm": 0.5787560343742371, "learning_rate": 9.96382973614373e-05, "loss": 0.0882, "step": 17280 }, { "grad_norm": 0.6499714255332947, "learning_rate": 9.963730405778142e-05, "loss": 0.0964, "step": 17290 }, { "grad_norm": 0.5263742208480835, "learning_rate": 9.96363093970603e-05, "loss": 0.0961, "step": 17300 }, { "grad_norm": 0.854817807674408, "learning_rate": 9.963531337930114e-05, "loss": 0.0858, "step": 17310 }, { "grad_norm": 0.5077136754989624, "learning_rate": 9.963431600453115e-05, "loss": 0.0846, "step": 17320 }, { "grad_norm": 0.6876624822616577, "learning_rate": 9.96333172727776e-05, "loss": 0.0902, "step": 17330 }, { "grad_norm": 0.44558948278427124, "learning_rate": 9.96323171840678e-05, "loss": 0.0689, "step": 17340 }, { "grad_norm": 0.8476610779762268, "learning_rate": 9.963131573842912e-05, "loss": 0.0659, "step": 17350 }, { "grad_norm": 0.8255099058151245, "learning_rate": 9.963031293588889e-05, "loss": 0.0704, "step": 17360 }, { "grad_norm": 0.5092073082923889, "learning_rate": 9.962930877647457e-05, "loss": 0.0752, "step": 17370 }, { "grad_norm": 0.4737416207790375, "learning_rate": 9.962830326021357e-05, "loss": 0.084, "step": 17380 }, { "grad_norm": 0.6522145867347717, "learning_rate": 9.962729638713341e-05, "loss": 0.0794, "step": 17390 }, { "grad_norm": 0.5471556782722473, "learning_rate": 9.96262881572616e-05, "loss": 0.085, "step": 17400 }, { "grad_norm": 0.5901970267295837, "learning_rate": 9.962527857062574e-05, "loss": 0.0889, "step": 17410 }, { "grad_norm": 0.6197850108146667, "learning_rate": 9.96242676272534e-05, "loss": 0.0895, "step": 17420 }, { "grad_norm": 0.7185725569725037, "learning_rate": 9.962325532717222e-05, "loss": 0.0781, "step": 17430 }, { "grad_norm": 0.7318128347396851, "learning_rate": 9.962224167040988e-05, "loss": 0.1332, "step": 17440 }, { "grad_norm": 0.8559635281562805, "learning_rate": 9.96212266569941e-05, "loss": 0.0917, "step": 17450 }, { "grad_norm": 0.6299645304679871, "learning_rate": 9.962021028695263e-05, "loss": 0.0847, "step": 17460 }, { "grad_norm": 0.6116491556167603, "learning_rate": 9.961919256031324e-05, "loss": 0.0859, "step": 17470 }, { "grad_norm": 0.7569529414176941, "learning_rate": 9.961817347710378e-05, "loss": 0.0794, "step": 17480 }, { "grad_norm": 0.6903420090675354, "learning_rate": 9.961715303735208e-05, "loss": 0.0774, "step": 17490 }, { "grad_norm": 0.5763995051383972, "learning_rate": 9.961613124108606e-05, "loss": 0.1095, "step": 17500 }, { "grad_norm": 0.5898011326789856, "learning_rate": 9.961510808833368e-05, "loss": 0.077, "step": 17510 }, { "grad_norm": 0.8138640522956848, "learning_rate": 9.961408357912286e-05, "loss": 0.0838, "step": 17520 }, { "grad_norm": 0.5948254466056824, "learning_rate": 9.961305771348163e-05, "loss": 0.0868, "step": 17530 }, { "grad_norm": 0.7057052850723267, "learning_rate": 9.961203049143805e-05, "loss": 0.0881, "step": 17540 }, { "grad_norm": 0.5610361695289612, "learning_rate": 9.961100191302019e-05, "loss": 0.0689, "step": 17550 }, { "grad_norm": 0.5273645520210266, "learning_rate": 9.960997197825618e-05, "loss": 0.1035, "step": 17560 }, { "grad_norm": 0.59166419506073, "learning_rate": 9.960894068717416e-05, "loss": 0.0973, "step": 17570 }, { "grad_norm": 0.6095185279846191, "learning_rate": 9.960790803980235e-05, "loss": 0.0762, "step": 17580 }, { "grad_norm": 0.5836493968963623, "learning_rate": 9.960687403616898e-05, "loss": 0.0757, "step": 17590 }, { "grad_norm": 0.43111664056777954, "learning_rate": 9.960583867630231e-05, "loss": 0.0647, "step": 17600 }, { "grad_norm": 0.5172200202941895, "learning_rate": 9.960480196023064e-05, "loss": 0.0803, "step": 17610 }, { "grad_norm": 0.5610156655311584, "learning_rate": 9.960376388798232e-05, "loss": 0.0731, "step": 17620 }, { "grad_norm": 0.42211127281188965, "learning_rate": 9.960272445958572e-05, "loss": 0.0682, "step": 17630 }, { "grad_norm": 1.0420233011245728, "learning_rate": 9.960168367506927e-05, "loss": 0.0809, "step": 17640 }, { "grad_norm": 0.5860196352005005, "learning_rate": 9.960064153446145e-05, "loss": 0.0892, "step": 17650 }, { "grad_norm": 0.6877866983413696, "learning_rate": 9.95995980377907e-05, "loss": 0.0882, "step": 17660 }, { "grad_norm": 1.0181466341018677, "learning_rate": 9.959855318508558e-05, "loss": 0.0839, "step": 17670 }, { "grad_norm": 0.7537293434143066, "learning_rate": 9.959750697637465e-05, "loss": 0.0765, "step": 17680 }, { "grad_norm": 0.8786177635192871, "learning_rate": 9.95964594116865e-05, "loss": 0.0616, "step": 17690 }, { "grad_norm": 0.5048705339431763, "learning_rate": 9.959541049104979e-05, "loss": 0.0773, "step": 17700 }, { "grad_norm": 0.6845552325248718, "learning_rate": 9.95943602144932e-05, "loss": 0.0736, "step": 17710 }, { "grad_norm": 0.633685290813446, "learning_rate": 9.959330858204542e-05, "loss": 0.0775, "step": 17720 }, { "grad_norm": 0.5676806569099426, "learning_rate": 9.959225559373521e-05, "loss": 0.0716, "step": 17730 }, { "grad_norm": 0.8417448401451111, "learning_rate": 9.959120124959138e-05, "loss": 0.0873, "step": 17740 }, { "grad_norm": 0.4200696349143982, "learning_rate": 9.959014554964273e-05, "loss": 0.0868, "step": 17750 }, { "grad_norm": 0.4789047837257385, "learning_rate": 9.958908849391812e-05, "loss": 0.0732, "step": 17760 }, { "grad_norm": 0.6685014367103577, "learning_rate": 9.958803008244649e-05, "loss": 0.1069, "step": 17770 }, { "grad_norm": 0.5434216856956482, "learning_rate": 9.958697031525672e-05, "loss": 0.0754, "step": 17780 }, { "grad_norm": 0.7152430415153503, "learning_rate": 9.958590919237783e-05, "loss": 0.0849, "step": 17790 }, { "grad_norm": 0.7335498332977295, "learning_rate": 9.958484671383879e-05, "loss": 0.0776, "step": 17800 }, { "grad_norm": 0.7635074257850647, "learning_rate": 9.958378287966868e-05, "loss": 0.0746, "step": 17810 }, { "grad_norm": 0.5453879833221436, "learning_rate": 9.958271768989658e-05, "loss": 0.0797, "step": 17820 }, { "grad_norm": 0.8937122821807861, "learning_rate": 9.95816511445516e-05, "loss": 0.1023, "step": 17830 }, { "grad_norm": 0.5456668138504028, "learning_rate": 9.958058324366291e-05, "loss": 0.0859, "step": 17840 }, { "grad_norm": 0.5963376760482788, "learning_rate": 9.957951398725969e-05, "loss": 0.0628, "step": 17850 }, { "grad_norm": 0.7943408489227295, "learning_rate": 9.95784433753712e-05, "loss": 0.0816, "step": 17860 }, { "grad_norm": 0.9351708292961121, "learning_rate": 9.957737140802666e-05, "loss": 0.0803, "step": 17870 }, { "grad_norm": 0.5199240446090698, "learning_rate": 9.957629808525544e-05, "loss": 0.09, "step": 17880 }, { "grad_norm": 0.7608328461647034, "learning_rate": 9.957522340708685e-05, "loss": 0.0704, "step": 17890 }, { "grad_norm": 0.6080573201179504, "learning_rate": 9.957414737355027e-05, "loss": 0.0587, "step": 17900 }, { "grad_norm": 0.5568516850471497, "learning_rate": 9.957306998467512e-05, "loss": 0.0671, "step": 17910 }, { "grad_norm": 0.8886057138442993, "learning_rate": 9.957199124049088e-05, "loss": 0.0809, "step": 17920 }, { "grad_norm": 0.6160197854042053, "learning_rate": 9.957091114102702e-05, "loss": 0.0766, "step": 17930 }, { "grad_norm": 0.5028795003890991, "learning_rate": 9.956982968631306e-05, "loss": 0.0654, "step": 17940 }, { "grad_norm": 0.6068440675735474, "learning_rate": 9.956874687637858e-05, "loss": 0.0934, "step": 17950 }, { "grad_norm": 0.6217051148414612, "learning_rate": 9.95676627112532e-05, "loss": 0.0664, "step": 17960 }, { "grad_norm": 0.5945988893508911, "learning_rate": 9.956657719096652e-05, "loss": 0.0969, "step": 17970 }, { "grad_norm": 0.5854355096817017, "learning_rate": 9.956549031554825e-05, "loss": 0.0978, "step": 17980 }, { "grad_norm": 0.5432735681533813, "learning_rate": 9.956440208502811e-05, "loss": 0.0907, "step": 17990 }, { "grad_norm": 0.6911395788192749, "learning_rate": 9.956331249943583e-05, "loss": 0.0649, "step": 18000 }, { "grad_norm": 0.6937709450721741, "learning_rate": 9.956222155880118e-05, "loss": 0.0711, "step": 18010 }, { "grad_norm": 0.4131618142127991, "learning_rate": 9.956112926315405e-05, "loss": 0.0628, "step": 18020 }, { "grad_norm": 0.6871560215950012, "learning_rate": 9.956003561252424e-05, "loss": 0.0734, "step": 18030 }, { "grad_norm": 1.1329628229141235, "learning_rate": 9.955894060694168e-05, "loss": 0.0933, "step": 18040 }, { "grad_norm": 0.6874530911445618, "learning_rate": 9.95578442464363e-05, "loss": 0.0991, "step": 18050 }, { "grad_norm": 0.5471956133842468, "learning_rate": 9.955674653103808e-05, "loss": 0.0726, "step": 18060 }, { "grad_norm": 0.5372629165649414, "learning_rate": 9.955564746077704e-05, "loss": 0.0786, "step": 18070 }, { "grad_norm": 0.6435133218765259, "learning_rate": 9.955454703568319e-05, "loss": 0.0828, "step": 18080 }, { "grad_norm": 1.053497552871704, "learning_rate": 9.955344525578667e-05, "loss": 0.0889, "step": 18090 }, { "grad_norm": 0.4682547152042389, "learning_rate": 9.955234212111755e-05, "loss": 0.0764, "step": 18100 }, { "grad_norm": 0.6612235903739929, "learning_rate": 9.955123763170602e-05, "loss": 0.1057, "step": 18110 }, { "grad_norm": 0.7940558791160583, "learning_rate": 9.955013178758228e-05, "loss": 0.0947, "step": 18120 }, { "grad_norm": 0.5672181248664856, "learning_rate": 9.954902458877654e-05, "loss": 0.0741, "step": 18130 }, { "grad_norm": 0.8739891648292542, "learning_rate": 9.954791603531908e-05, "loss": 0.0733, "step": 18140 }, { "grad_norm": 0.6261935830116272, "learning_rate": 9.954680612724021e-05, "loss": 0.0762, "step": 18150 }, { "grad_norm": 0.5794705152511597, "learning_rate": 9.954569486457028e-05, "loss": 0.0825, "step": 18160 }, { "grad_norm": 0.5946964025497437, "learning_rate": 9.954458224733966e-05, "loss": 0.1022, "step": 18170 }, { "grad_norm": 0.5497786402702332, "learning_rate": 9.954346827557879e-05, "loss": 0.0736, "step": 18180 }, { "grad_norm": 0.7041869759559631, "learning_rate": 9.954235294931809e-05, "loss": 0.0922, "step": 18190 }, { "grad_norm": 0.5720638632774353, "learning_rate": 9.954123626858808e-05, "loss": 0.0822, "step": 18200 }, { "grad_norm": 0.5233850479125977, "learning_rate": 9.954011823341928e-05, "loss": 0.0738, "step": 18210 }, { "grad_norm": 0.8568909764289856, "learning_rate": 9.953899884384228e-05, "loss": 0.0786, "step": 18220 }, { "grad_norm": 0.6614159345626831, "learning_rate": 9.953787809988764e-05, "loss": 0.0843, "step": 18230 }, { "grad_norm": 0.8840672969818115, "learning_rate": 9.953675600158603e-05, "loss": 0.0943, "step": 18240 }, { "grad_norm": 0.6388808488845825, "learning_rate": 9.953563254896812e-05, "loss": 0.1017, "step": 18250 }, { "grad_norm": 0.6756848692893982, "learning_rate": 9.953450774206462e-05, "loss": 0.0645, "step": 18260 }, { "grad_norm": 0.8149328231811523, "learning_rate": 9.953338158090629e-05, "loss": 0.0738, "step": 18270 }, { "grad_norm": 0.5886061191558838, "learning_rate": 9.953225406552392e-05, "loss": 0.0788, "step": 18280 }, { "grad_norm": 0.653188169002533, "learning_rate": 9.953112519594834e-05, "loss": 0.1121, "step": 18290 }, { "grad_norm": 0.6109516024589539, "learning_rate": 9.952999497221038e-05, "loss": 0.079, "step": 18300 }, { "grad_norm": 0.7842172384262085, "learning_rate": 9.9528863394341e-05, "loss": 0.0774, "step": 18310 }, { "grad_norm": 0.7289329171180725, "learning_rate": 9.952773046237108e-05, "loss": 0.0755, "step": 18320 }, { "grad_norm": 0.7689876556396484, "learning_rate": 9.95265961763316e-05, "loss": 0.0822, "step": 18330 }, { "grad_norm": 0.6075828075408936, "learning_rate": 9.952546053625362e-05, "loss": 0.0764, "step": 18340 }, { "grad_norm": 0.6479706764221191, "learning_rate": 9.952432354216815e-05, "loss": 0.0746, "step": 18350 }, { "grad_norm": 0.7659331560134888, "learning_rate": 9.952318519410625e-05, "loss": 0.0679, "step": 18360 }, { "grad_norm": 0.5836053490638733, "learning_rate": 9.952204549209909e-05, "loss": 0.0652, "step": 18370 }, { "grad_norm": 0.4501657485961914, "learning_rate": 9.952090443617783e-05, "loss": 0.0617, "step": 18380 }, { "grad_norm": 0.650675892829895, "learning_rate": 9.951976202637364e-05, "loss": 0.0879, "step": 18390 }, { "grad_norm": 0.7215585708618164, "learning_rate": 9.951861826271773e-05, "loss": 0.088, "step": 18400 }, { "grad_norm": 0.7586039304733276, "learning_rate": 9.951747314524143e-05, "loss": 0.0763, "step": 18410 }, { "grad_norm": 0.7561293840408325, "learning_rate": 9.9516326673976e-05, "loss": 0.0959, "step": 18420 }, { "grad_norm": 0.6351004242897034, "learning_rate": 9.95151788489528e-05, "loss": 0.1071, "step": 18430 }, { "grad_norm": 0.5647758841514587, "learning_rate": 9.951402967020322e-05, "loss": 0.0831, "step": 18440 }, { "grad_norm": 0.6836926341056824, "learning_rate": 9.951287913775869e-05, "loss": 0.0826, "step": 18450 }, { "grad_norm": 0.5780847668647766, "learning_rate": 9.951172725165061e-05, "loss": 0.0785, "step": 18460 }, { "grad_norm": 0.7386578917503357, "learning_rate": 9.951057401191053e-05, "loss": 0.07, "step": 18470 }, { "grad_norm": 0.5990845561027527, "learning_rate": 9.950941941856996e-05, "loss": 0.1052, "step": 18480 }, { "grad_norm": 0.5833959579467773, "learning_rate": 9.950826347166045e-05, "loss": 0.0697, "step": 18490 }, { "grad_norm": 0.4852882921695709, "learning_rate": 9.950710617121362e-05, "loss": 0.0787, "step": 18500 }, { "grad_norm": 0.4574964642524719, "learning_rate": 9.95059475172611e-05, "loss": 0.0789, "step": 18510 }, { "grad_norm": 0.7288281917572021, "learning_rate": 9.950478750983458e-05, "loss": 0.0745, "step": 18520 }, { "grad_norm": 0.9576372504234314, "learning_rate": 9.950362614896575e-05, "loss": 0.0683, "step": 18530 }, { "grad_norm": 0.8894375562667847, "learning_rate": 9.950246343468642e-05, "loss": 0.0887, "step": 18540 }, { "grad_norm": 0.587568998336792, "learning_rate": 9.95012993670283e-05, "loss": 0.0711, "step": 18550 }, { "grad_norm": 0.7948519587516785, "learning_rate": 9.950013394602327e-05, "loss": 0.0821, "step": 18560 }, { "grad_norm": 0.5172438621520996, "learning_rate": 9.949896717170316e-05, "loss": 0.0853, "step": 18570 }, { "grad_norm": 1.0113276243209839, "learning_rate": 9.949779904409989e-05, "loss": 0.087, "step": 18580 }, { "grad_norm": 0.7226466536521912, "learning_rate": 9.949662956324539e-05, "loss": 0.0728, "step": 18590 }, { "grad_norm": 0.5805397033691406, "learning_rate": 9.949545872917164e-05, "loss": 0.0938, "step": 18600 }, { "grad_norm": 0.6570537090301514, "learning_rate": 9.949428654191065e-05, "loss": 0.0979, "step": 18610 }, { "grad_norm": 0.9276039600372314, "learning_rate": 9.949311300149444e-05, "loss": 0.0788, "step": 18620 }, { "grad_norm": 0.4829694330692291, "learning_rate": 9.949193810795511e-05, "loss": 0.1125, "step": 18630 }, { "grad_norm": 0.5234202146530151, "learning_rate": 9.94907618613248e-05, "loss": 0.094, "step": 18640 }, { "grad_norm": 0.6796770691871643, "learning_rate": 9.948958426163566e-05, "loss": 0.0982, "step": 18650 }, { "grad_norm": 0.7556085586547852, "learning_rate": 9.948840530891987e-05, "loss": 0.068, "step": 18660 }, { "grad_norm": 1.2050825357437134, "learning_rate": 9.948722500320966e-05, "loss": 0.112, "step": 18670 }, { "grad_norm": 0.5647839903831482, "learning_rate": 9.948604334453733e-05, "loss": 0.0921, "step": 18680 }, { "grad_norm": 0.5488487482070923, "learning_rate": 9.948486033293515e-05, "loss": 0.0823, "step": 18690 }, { "grad_norm": 0.5395099520683289, "learning_rate": 9.948367596843549e-05, "loss": 0.0973, "step": 18700 }, { "grad_norm": 0.8873552083969116, "learning_rate": 9.94824902510707e-05, "loss": 0.0751, "step": 18710 }, { "grad_norm": 0.5866613388061523, "learning_rate": 9.948130318087325e-05, "loss": 0.0763, "step": 18720 }, { "grad_norm": 1.0089519023895264, "learning_rate": 9.948011475787553e-05, "loss": 0.0844, "step": 18730 }, { "grad_norm": 0.5705054998397827, "learning_rate": 9.947892498211007e-05, "loss": 0.1139, "step": 18740 }, { "grad_norm": 0.5532819628715515, "learning_rate": 9.947773385360939e-05, "loss": 0.0809, "step": 18750 }, { "grad_norm": 0.5920979976654053, "learning_rate": 9.947654137240606e-05, "loss": 0.0774, "step": 18760 }, { "grad_norm": 0.49217405915260315, "learning_rate": 9.947534753853267e-05, "loss": 0.0719, "step": 18770 }, { "grad_norm": 0.680561363697052, "learning_rate": 9.947415235202187e-05, "loss": 0.0745, "step": 18780 }, { "grad_norm": 0.44595927000045776, "learning_rate": 9.947295581290632e-05, "loss": 0.0879, "step": 18790 }, { "grad_norm": 0.6850140690803528, "learning_rate": 9.947175792121877e-05, "loss": 0.0943, "step": 18800 }, { "grad_norm": 0.7387519478797913, "learning_rate": 9.947055867699192e-05, "loss": 0.0934, "step": 18810 }, { "grad_norm": 1.005111575126648, "learning_rate": 9.946935808025859e-05, "loss": 0.0786, "step": 18820 }, { "grad_norm": 0.7907278537750244, "learning_rate": 9.946815613105159e-05, "loss": 0.0843, "step": 18830 }, { "grad_norm": 0.8985633850097656, "learning_rate": 9.94669528294038e-05, "loss": 0.0999, "step": 18840 }, { "grad_norm": 0.565900444984436, "learning_rate": 9.946574817534809e-05, "loss": 0.0754, "step": 18850 }, { "grad_norm": 0.8474990129470825, "learning_rate": 9.946454216891741e-05, "loss": 0.0735, "step": 18860 }, { "grad_norm": 0.6021759510040283, "learning_rate": 9.946333481014472e-05, "loss": 0.067, "step": 18870 }, { "grad_norm": 0.7359085083007812, "learning_rate": 9.946212609906305e-05, "loss": 0.072, "step": 18880 }, { "grad_norm": 0.8510693907737732, "learning_rate": 9.946091603570545e-05, "loss": 0.0718, "step": 18890 }, { "grad_norm": 0.6274131536483765, "learning_rate": 9.945970462010495e-05, "loss": 0.0489, "step": 18900 }, { "grad_norm": 0.8085114359855652, "learning_rate": 9.945849185229473e-05, "loss": 0.0707, "step": 18910 }, { "grad_norm": 0.5841957926750183, "learning_rate": 9.945727773230793e-05, "loss": 0.0606, "step": 18920 }, { "grad_norm": 0.5888252854347229, "learning_rate": 9.945606226017772e-05, "loss": 0.0936, "step": 18930 }, { "grad_norm": 0.5016506910324097, "learning_rate": 9.945484543593736e-05, "loss": 0.1135, "step": 18940 }, { "grad_norm": 0.6373016834259033, "learning_rate": 9.945362725962011e-05, "loss": 0.0714, "step": 18950 }, { "grad_norm": 0.6382120251655579, "learning_rate": 9.945240773125926e-05, "loss": 0.0691, "step": 18960 }, { "grad_norm": 0.6503155827522278, "learning_rate": 9.945118685088815e-05, "loss": 0.0669, "step": 18970 }, { "grad_norm": 0.5858096480369568, "learning_rate": 9.944996461854019e-05, "loss": 0.0811, "step": 18980 }, { "grad_norm": 0.573168933391571, "learning_rate": 9.944874103424876e-05, "loss": 0.074, "step": 18990 }, { "grad_norm": 0.7404623627662659, "learning_rate": 9.944751609804733e-05, "loss": 0.0743, "step": 19000 }, { "grad_norm": 0.5376530289649963, "learning_rate": 9.94462898099694e-05, "loss": 0.0714, "step": 19010 }, { "grad_norm": 2.054637908935547, "learning_rate": 9.944506217004847e-05, "loss": 0.0907, "step": 19020 }, { "grad_norm": 0.7149777412414551, "learning_rate": 9.944383317831812e-05, "loss": 0.0686, "step": 19030 }, { "grad_norm": 0.8094817399978638, "learning_rate": 9.944260283481194e-05, "loss": 0.0928, "step": 19040 }, { "grad_norm": 0.5898641347885132, "learning_rate": 9.944137113956358e-05, "loss": 0.089, "step": 19050 }, { "grad_norm": 1.2313085794448853, "learning_rate": 9.944013809260671e-05, "loss": 0.0936, "step": 19060 }, { "grad_norm": 0.874953031539917, "learning_rate": 9.943890369397502e-05, "loss": 0.0724, "step": 19070 }, { "grad_norm": 0.8216850757598877, "learning_rate": 9.943766794370228e-05, "loss": 0.1101, "step": 19080 }, { "grad_norm": 0.7446150779724121, "learning_rate": 9.943643084182229e-05, "loss": 0.0876, "step": 19090 }, { "grad_norm": 0.9350096583366394, "learning_rate": 9.943519238836885e-05, "loss": 0.0675, "step": 19100 }, { "grad_norm": 0.7413834929466248, "learning_rate": 9.943395258337581e-05, "loss": 0.0945, "step": 19110 }, { "grad_norm": 1.0092371702194214, "learning_rate": 9.943271142687709e-05, "loss": 0.0933, "step": 19120 }, { "grad_norm": 1.0332201719284058, "learning_rate": 9.94314689189066e-05, "loss": 0.0916, "step": 19130 }, { "grad_norm": 0.5292371511459351, "learning_rate": 9.943022505949833e-05, "loss": 0.1072, "step": 19140 }, { "grad_norm": 0.4833327531814575, "learning_rate": 9.942897984868626e-05, "loss": 0.0853, "step": 19150 }, { "grad_norm": 0.834291398525238, "learning_rate": 9.942773328650447e-05, "loss": 0.0964, "step": 19160 }, { "grad_norm": 0.6420473456382751, "learning_rate": 9.942648537298701e-05, "loss": 0.0782, "step": 19170 }, { "grad_norm": 0.7065240144729614, "learning_rate": 9.9425236108168e-05, "loss": 0.1077, "step": 19180 }, { "grad_norm": 0.8713886141777039, "learning_rate": 9.942398549208163e-05, "loss": 0.0839, "step": 19190 }, { "grad_norm": 0.6155529022216797, "learning_rate": 9.942273352476205e-05, "loss": 0.0786, "step": 19200 }, { "grad_norm": 0.6299723982810974, "learning_rate": 9.94214802062435e-05, "loss": 0.0747, "step": 19210 }, { "grad_norm": 0.6229888796806335, "learning_rate": 9.942022553656025e-05, "loss": 0.0769, "step": 19220 }, { "grad_norm": 0.522466778755188, "learning_rate": 9.941896951574659e-05, "loss": 0.0817, "step": 19230 }, { "grad_norm": 0.5513133406639099, "learning_rate": 9.941771214383689e-05, "loss": 0.0897, "step": 19240 }, { "grad_norm": 0.5818307995796204, "learning_rate": 9.94164534208655e-05, "loss": 0.0862, "step": 19250 }, { "grad_norm": 0.6078224182128906, "learning_rate": 9.941519334686682e-05, "loss": 0.0684, "step": 19260 }, { "grad_norm": 0.5784385800361633, "learning_rate": 9.941393192187534e-05, "loss": 0.0544, "step": 19270 }, { "grad_norm": 0.6301624774932861, "learning_rate": 9.941266914592551e-05, "loss": 0.0756, "step": 19280 }, { "grad_norm": 0.5835586190223694, "learning_rate": 9.941140501905188e-05, "loss": 0.0905, "step": 19290 }, { "grad_norm": 0.4279286861419678, "learning_rate": 9.941013954128899e-05, "loss": 0.0553, "step": 19300 }, { "grad_norm": 0.5617167353630066, "learning_rate": 9.940887271267145e-05, "loss": 0.0865, "step": 19310 }, { "grad_norm": 0.5577772259712219, "learning_rate": 9.94076045332339e-05, "loss": 0.0608, "step": 19320 }, { "grad_norm": 0.6222617030143738, "learning_rate": 9.9406335003011e-05, "loss": 0.0822, "step": 19330 }, { "grad_norm": 0.5529706478118896, "learning_rate": 9.940506412203746e-05, "loss": 0.0938, "step": 19340 }, { "grad_norm": 0.5531240105628967, "learning_rate": 9.940379189034802e-05, "loss": 0.0841, "step": 19350 }, { "grad_norm": 0.740673303604126, "learning_rate": 9.940251830797749e-05, "loss": 0.0669, "step": 19360 }, { "grad_norm": 1.0696684122085571, "learning_rate": 9.940124337496065e-05, "loss": 0.0899, "step": 19370 }, { "grad_norm": 0.7814590334892273, "learning_rate": 9.939996709133239e-05, "loss": 0.0824, "step": 19380 }, { "grad_norm": 0.7043203115463257, "learning_rate": 9.939868945712758e-05, "loss": 0.0737, "step": 19390 }, { "grad_norm": 0.6044790744781494, "learning_rate": 9.939741047238115e-05, "loss": 0.0732, "step": 19400 }, { "grad_norm": 1.0345704555511475, "learning_rate": 9.939613013712809e-05, "loss": 0.0887, "step": 19410 }, { "grad_norm": 0.5613470673561096, "learning_rate": 9.939484845140339e-05, "loss": 0.0675, "step": 19420 }, { "grad_norm": 0.4791855216026306, "learning_rate": 9.939356541524207e-05, "loss": 0.0701, "step": 19430 }, { "grad_norm": 0.598048985004425, "learning_rate": 9.939228102867926e-05, "loss": 0.0852, "step": 19440 }, { "grad_norm": 0.7389993667602539, "learning_rate": 9.939099529175003e-05, "loss": 0.0992, "step": 19450 }, { "grad_norm": 0.8387928009033203, "learning_rate": 9.938970820448951e-05, "loss": 0.0961, "step": 19460 }, { "grad_norm": 1.6681225299835205, "learning_rate": 9.938841976693296e-05, "loss": 0.078, "step": 19470 }, { "grad_norm": 0.7161552309989929, "learning_rate": 9.938712997911557e-05, "loss": 0.0803, "step": 19480 }, { "grad_norm": 0.83061683177948, "learning_rate": 9.938583884107257e-05, "loss": 0.0839, "step": 19490 }, { "grad_norm": 1.0533649921417236, "learning_rate": 9.938454635283932e-05, "loss": 0.0771, "step": 19500 }, { "grad_norm": 0.7864642143249512, "learning_rate": 9.938325251445112e-05, "loss": 0.0768, "step": 19510 }, { "grad_norm": 0.8131436705589294, "learning_rate": 9.938195732594335e-05, "loss": 0.0731, "step": 19520 }, { "grad_norm": 0.7060471177101135, "learning_rate": 9.93806607873514e-05, "loss": 0.0883, "step": 19530 }, { "grad_norm": 0.6929425597190857, "learning_rate": 9.937936289871075e-05, "loss": 0.0749, "step": 19540 }, { "grad_norm": 0.9010580778121948, "learning_rate": 9.937806366005686e-05, "loss": 0.0913, "step": 19550 }, { "grad_norm": 0.9015332460403442, "learning_rate": 9.937676307142527e-05, "loss": 0.0823, "step": 19560 }, { "grad_norm": 0.5615094900131226, "learning_rate": 9.937546113285152e-05, "loss": 0.08, "step": 19570 }, { "grad_norm": 1.0284972190856934, "learning_rate": 9.937415784437123e-05, "loss": 0.0872, "step": 19580 }, { "grad_norm": 0.6816398501396179, "learning_rate": 9.937285320601999e-05, "loss": 0.0828, "step": 19590 }, { "grad_norm": 0.7342271208763123, "learning_rate": 9.93715472178335e-05, "loss": 0.0719, "step": 19600 }, { "grad_norm": 0.8827897906303406, "learning_rate": 9.937023987984746e-05, "loss": 0.0829, "step": 19610 }, { "grad_norm": 0.6880467534065247, "learning_rate": 9.936893119209761e-05, "loss": 0.0793, "step": 19620 }, { "grad_norm": 0.5945860147476196, "learning_rate": 9.936762115461972e-05, "loss": 0.0978, "step": 19630 }, { "grad_norm": 0.6721397042274475, "learning_rate": 9.936630976744963e-05, "loss": 0.0672, "step": 19640 }, { "grad_norm": 0.6808332800865173, "learning_rate": 9.936499703062316e-05, "loss": 0.0856, "step": 19650 }, { "grad_norm": 0.5734970569610596, "learning_rate": 9.936368294417623e-05, "loss": 0.0819, "step": 19660 }, { "grad_norm": 0.9022446870803833, "learning_rate": 9.936236750814475e-05, "loss": 0.0976, "step": 19670 }, { "grad_norm": 0.5904114842414856, "learning_rate": 9.936105072256467e-05, "loss": 0.0952, "step": 19680 }, { "grad_norm": 0.5235656499862671, "learning_rate": 9.935973258747203e-05, "loss": 0.0936, "step": 19690 }, { "grad_norm": 0.686568021774292, "learning_rate": 9.935841310290283e-05, "loss": 0.0757, "step": 19700 }, { "grad_norm": 0.5609014630317688, "learning_rate": 9.935709226889319e-05, "loss": 0.0709, "step": 19710 }, { "grad_norm": 0.5957849621772766, "learning_rate": 9.935577008547915e-05, "loss": 0.0701, "step": 19720 }, { "grad_norm": 0.5945463180541992, "learning_rate": 9.935444655269692e-05, "loss": 0.0783, "step": 19730 }, { "grad_norm": 0.6415460109710693, "learning_rate": 9.935312167058265e-05, "loss": 0.0642, "step": 19740 }, { "grad_norm": 0.9696460366249084, "learning_rate": 9.935179543917258e-05, "loss": 0.0874, "step": 19750 }, { "grad_norm": 0.6641755700111389, "learning_rate": 9.935046785850297e-05, "loss": 0.0944, "step": 19760 }, { "grad_norm": 0.5902247428894043, "learning_rate": 9.93491389286101e-05, "loss": 0.0623, "step": 19770 }, { "grad_norm": 0.7113121747970581, "learning_rate": 9.934780864953031e-05, "loss": 0.0701, "step": 19780 }, { "grad_norm": 0.6962265968322754, "learning_rate": 9.934647702129998e-05, "loss": 0.0795, "step": 19790 }, { "grad_norm": 1.0143132209777832, "learning_rate": 9.934514404395549e-05, "loss": 0.0798, "step": 19800 }, { "grad_norm": 0.9334388971328735, "learning_rate": 9.934380971753333e-05, "loss": 0.0786, "step": 19810 }, { "grad_norm": 0.6635790467262268, "learning_rate": 9.934247404206992e-05, "loss": 0.0706, "step": 19820 }, { "grad_norm": 0.6803110241889954, "learning_rate": 9.934113701760181e-05, "loss": 0.0763, "step": 19830 }, { "grad_norm": 0.6204708218574524, "learning_rate": 9.933979864416558e-05, "loss": 0.0945, "step": 19840 }, { "grad_norm": 0.5740191340446472, "learning_rate": 9.933845892179776e-05, "loss": 0.0821, "step": 19850 }, { "grad_norm": 0.6324191689491272, "learning_rate": 9.933711785053502e-05, "loss": 0.0672, "step": 19860 }, { "grad_norm": 0.5102204084396362, "learning_rate": 9.933577543041401e-05, "loss": 0.0715, "step": 19870 }, { "grad_norm": 0.6801705956459045, "learning_rate": 9.933443166147144e-05, "loss": 0.0832, "step": 19880 }, { "grad_norm": 0.7128885984420776, "learning_rate": 9.933308654374403e-05, "loss": 0.0643, "step": 19890 }, { "grad_norm": 0.7425634264945984, "learning_rate": 9.933174007726858e-05, "loss": 0.0662, "step": 19900 }, { "grad_norm": 0.6978074908256531, "learning_rate": 9.933039226208189e-05, "loss": 0.0759, "step": 19910 }, { "grad_norm": 0.4897766411304474, "learning_rate": 9.93290430982208e-05, "loss": 0.0722, "step": 19920 }, { "grad_norm": 0.8518407940864563, "learning_rate": 9.93276925857222e-05, "loss": 0.079, "step": 19930 }, { "grad_norm": 0.6803897619247437, "learning_rate": 9.932634072462304e-05, "loss": 0.1022, "step": 19940 }, { "grad_norm": 0.9022562503814697, "learning_rate": 9.932498751496024e-05, "loss": 0.0787, "step": 19950 }, { "grad_norm": 0.46578237414360046, "learning_rate": 9.93236329567708e-05, "loss": 0.0839, "step": 19960 }, { "grad_norm": 0.7767977118492126, "learning_rate": 9.932227705009178e-05, "loss": 0.0937, "step": 19970 }, { "grad_norm": 0.7968692779541016, "learning_rate": 9.932091979496023e-05, "loss": 0.0748, "step": 19980 }, { "grad_norm": 0.6254744529724121, "learning_rate": 9.931956119141326e-05, "loss": 0.0811, "step": 19990 }, { "grad_norm": 0.912772536277771, "learning_rate": 9.931820123948803e-05, "loss": 0.1022, "step": 20000 }, { "grad_norm": 0.6012299060821533, "learning_rate": 9.931683993922169e-05, "loss": 0.0976, "step": 20010 }, { "grad_norm": 0.6716795563697815, "learning_rate": 9.931547729065147e-05, "loss": 0.0787, "step": 20020 }, { "grad_norm": 0.5236432552337646, "learning_rate": 9.931411329381464e-05, "loss": 0.0757, "step": 20030 }, { "grad_norm": 0.7657187581062317, "learning_rate": 9.931274794874847e-05, "loss": 0.0946, "step": 20040 }, { "grad_norm": 0.6216099262237549, "learning_rate": 9.931138125549031e-05, "loss": 0.0642, "step": 20050 }, { "grad_norm": 0.5644552707672119, "learning_rate": 9.93100132140775e-05, "loss": 0.0735, "step": 20060 }, { "grad_norm": 0.6562463641166687, "learning_rate": 9.930864382454746e-05, "loss": 0.0986, "step": 20070 }, { "grad_norm": 0.7289358973503113, "learning_rate": 9.930727308693762e-05, "loss": 0.0827, "step": 20080 }, { "grad_norm": 0.6872687339782715, "learning_rate": 9.930590100128545e-05, "loss": 0.0844, "step": 20090 }, { "grad_norm": 0.4844474494457245, "learning_rate": 9.930452756762848e-05, "loss": 0.0822, "step": 20100 }, { "grad_norm": 0.755044162273407, "learning_rate": 9.930315278600425e-05, "loss": 0.0908, "step": 20110 }, { "grad_norm": 0.6221271753311157, "learning_rate": 9.930177665645034e-05, "loss": 0.0726, "step": 20120 }, { "grad_norm": 0.5764560699462891, "learning_rate": 9.930039917900437e-05, "loss": 0.0659, "step": 20130 }, { "grad_norm": 0.857058584690094, "learning_rate": 9.929902035370403e-05, "loss": 0.0652, "step": 20140 }, { "grad_norm": 0.5327761173248291, "learning_rate": 9.929764018058698e-05, "loss": 0.0799, "step": 20150 }, { "grad_norm": 0.5390449166297913, "learning_rate": 9.929625865969096e-05, "loss": 0.0677, "step": 20160 }, { "grad_norm": 0.9739412665367126, "learning_rate": 9.929487579105376e-05, "loss": 0.0748, "step": 20170 }, { "grad_norm": 0.494037926197052, "learning_rate": 9.929349157471319e-05, "loss": 0.0742, "step": 20180 }, { "grad_norm": 0.5320767760276794, "learning_rate": 9.929210601070706e-05, "loss": 0.0553, "step": 20190 }, { "grad_norm": 0.6014220714569092, "learning_rate": 9.929071909907326e-05, "loss": 0.0859, "step": 20200 }, { "grad_norm": 0.9646050930023193, "learning_rate": 9.928933083984973e-05, "loss": 0.0613, "step": 20210 }, { "grad_norm": 0.53474360704422, "learning_rate": 9.92879412330744e-05, "loss": 0.0716, "step": 20220 }, { "grad_norm": 0.9199774265289307, "learning_rate": 9.928655027878529e-05, "loss": 0.0665, "step": 20230 }, { "grad_norm": 0.44643551111221313, "learning_rate": 9.92851579770204e-05, "loss": 0.0724, "step": 20240 }, { "grad_norm": 0.4572179317474365, "learning_rate": 9.928376432781783e-05, "loss": 0.0598, "step": 20250 }, { "grad_norm": 0.7902498841285706, "learning_rate": 9.928236933121564e-05, "loss": 0.0925, "step": 20260 }, { "grad_norm": 0.6501169800758362, "learning_rate": 9.928097298725199e-05, "loss": 0.0799, "step": 20270 }, { "grad_norm": 0.7219334244728088, "learning_rate": 9.927957529596505e-05, "loss": 0.07, "step": 20280 }, { "grad_norm": 0.8025805950164795, "learning_rate": 9.927817625739303e-05, "loss": 0.0792, "step": 20290 }, { "grad_norm": 0.5092989206314087, "learning_rate": 9.927677587157418e-05, "loss": 0.0799, "step": 20300 }, { "grad_norm": 0.5369877815246582, "learning_rate": 9.92753741385468e-05, "loss": 0.074, "step": 20310 }, { "grad_norm": 0.7053918242454529, "learning_rate": 9.927397105834922e-05, "loss": 0.0743, "step": 20320 }, { "grad_norm": 0.5175106525421143, "learning_rate": 9.927256663101975e-05, "loss": 0.0728, "step": 20330 }, { "grad_norm": 0.647278368473053, "learning_rate": 9.927116085659683e-05, "loss": 0.0814, "step": 20340 }, { "grad_norm": 0.6253474354743958, "learning_rate": 9.926975373511889e-05, "loss": 0.0922, "step": 20350 }, { "grad_norm": 0.566596508026123, "learning_rate": 9.926834526662439e-05, "loss": 0.0846, "step": 20360 }, { "grad_norm": 0.5539163947105408, "learning_rate": 9.926693545115184e-05, "loss": 0.0798, "step": 20370 }, { "grad_norm": 0.8610737323760986, "learning_rate": 9.926552428873979e-05, "loss": 0.1121, "step": 20380 }, { "grad_norm": 0.45455411076545715, "learning_rate": 9.926411177942681e-05, "loss": 0.0846, "step": 20390 }, { "grad_norm": 0.7093355059623718, "learning_rate": 9.926269792325151e-05, "loss": 0.0874, "step": 20400 }, { "grad_norm": 0.6704656481742859, "learning_rate": 9.926128272025257e-05, "loss": 0.0829, "step": 20410 }, { "grad_norm": 0.611197829246521, "learning_rate": 9.925986617046865e-05, "loss": 0.0789, "step": 20420 }, { "grad_norm": 0.5565042495727539, "learning_rate": 9.925844827393851e-05, "loss": 0.0829, "step": 20430 }, { "grad_norm": 0.6654995083808899, "learning_rate": 9.925702903070089e-05, "loss": 0.0579, "step": 20440 }, { "grad_norm": 0.5513204336166382, "learning_rate": 9.925560844079461e-05, "loss": 0.0783, "step": 20450 }, { "grad_norm": 0.702998161315918, "learning_rate": 9.92541865042585e-05, "loss": 0.0738, "step": 20460 }, { "grad_norm": 0.6631424427032471, "learning_rate": 9.925276322113143e-05, "loss": 0.0787, "step": 20470 }, { "grad_norm": 0.8645622730255127, "learning_rate": 9.925133859145232e-05, "loss": 0.0634, "step": 20480 }, { "grad_norm": 0.9498077630996704, "learning_rate": 9.924991261526011e-05, "loss": 0.094, "step": 20490 }, { "grad_norm": 0.593241810798645, "learning_rate": 9.92484852925938e-05, "loss": 0.0811, "step": 20500 }, { "grad_norm": 0.8796772360801697, "learning_rate": 9.92470566234924e-05, "loss": 0.0657, "step": 20510 }, { "grad_norm": 0.6845420002937317, "learning_rate": 9.924562660799497e-05, "loss": 0.0693, "step": 20520 }, { "grad_norm": 0.6233814358711243, "learning_rate": 9.924419524614061e-05, "loss": 0.0762, "step": 20530 }, { "grad_norm": 0.7937782406806946, "learning_rate": 9.924276253796845e-05, "loss": 0.0765, "step": 20540 }, { "grad_norm": 0.5315874814987183, "learning_rate": 9.924132848351767e-05, "loss": 0.0855, "step": 20550 }, { "grad_norm": 0.5541489720344543, "learning_rate": 9.923989308282746e-05, "loss": 0.0847, "step": 20560 }, { "grad_norm": 0.3925674855709076, "learning_rate": 9.923845633593708e-05, "loss": 0.0963, "step": 20570 }, { "grad_norm": 0.8070347309112549, "learning_rate": 9.92370182428858e-05, "loss": 0.0623, "step": 20580 }, { "grad_norm": 0.5540624260902405, "learning_rate": 9.923557880371293e-05, "loss": 0.0552, "step": 20590 }, { "grad_norm": 0.7651801705360413, "learning_rate": 9.923413801845785e-05, "loss": 0.0777, "step": 20600 }, { "grad_norm": 0.44970545172691345, "learning_rate": 9.923269588715991e-05, "loss": 0.0636, "step": 20610 }, { "grad_norm": 0.6746005415916443, "learning_rate": 9.923125240985857e-05, "loss": 0.0764, "step": 20620 }, { "grad_norm": 0.6592934131622314, "learning_rate": 9.922980758659331e-05, "loss": 0.0832, "step": 20630 }, { "grad_norm": 0.7059399485588074, "learning_rate": 9.922836141740357e-05, "loss": 0.0754, "step": 20640 }, { "grad_norm": 0.9892763495445251, "learning_rate": 9.922691390232894e-05, "loss": 0.0697, "step": 20650 }, { "grad_norm": 0.6928750872612, "learning_rate": 9.922546504140897e-05, "loss": 0.0761, "step": 20660 }, { "grad_norm": 0.7608811855316162, "learning_rate": 9.922401483468327e-05, "loss": 0.0858, "step": 20670 }, { "grad_norm": 0.880558967590332, "learning_rate": 9.922256328219151e-05, "loss": 0.0889, "step": 20680 }, { "grad_norm": 0.6218160390853882, "learning_rate": 9.922111038397337e-05, "loss": 0.0796, "step": 20690 }, { "grad_norm": 0.7276009321212769, "learning_rate": 9.921965614006855e-05, "loss": 0.0939, "step": 20700 }, { "grad_norm": 1.2338963747024536, "learning_rate": 9.921820055051682e-05, "loss": 0.0753, "step": 20710 }, { "grad_norm": 0.6375212073326111, "learning_rate": 9.9216743615358e-05, "loss": 0.0627, "step": 20720 }, { "grad_norm": 0.5063421130180359, "learning_rate": 9.921528533463186e-05, "loss": 0.084, "step": 20730 }, { "grad_norm": 0.5828374624252319, "learning_rate": 9.921382570837835e-05, "loss": 0.0559, "step": 20740 }, { "grad_norm": 0.6085830330848694, "learning_rate": 9.921236473663731e-05, "loss": 0.0658, "step": 20750 }, { "grad_norm": 0.5939198136329651, "learning_rate": 9.921090241944872e-05, "loss": 0.0871, "step": 20760 }, { "grad_norm": 0.6595045924186707, "learning_rate": 9.920943875685254e-05, "loss": 0.0661, "step": 20770 }, { "grad_norm": 0.9304861426353455, "learning_rate": 9.920797374888879e-05, "loss": 0.0714, "step": 20780 }, { "grad_norm": 0.5672391653060913, "learning_rate": 9.920650739559751e-05, "loss": 0.0858, "step": 20790 }, { "grad_norm": 0.4803975224494934, "learning_rate": 9.920503969701882e-05, "loss": 0.0758, "step": 20800 }, { "grad_norm": 0.6990278363227844, "learning_rate": 9.920357065319284e-05, "loss": 0.0825, "step": 20810 }, { "grad_norm": 0.7171846628189087, "learning_rate": 9.92021002641597e-05, "loss": 0.0819, "step": 20820 }, { "grad_norm": 0.7137464880943298, "learning_rate": 9.920062852995964e-05, "loss": 0.0667, "step": 20830 }, { "grad_norm": 0.6640031337738037, "learning_rate": 9.919915545063286e-05, "loss": 0.0819, "step": 20840 }, { "grad_norm": 0.5720766186714172, "learning_rate": 9.919768102621966e-05, "loss": 0.0663, "step": 20850 }, { "grad_norm": 1.0875829458236694, "learning_rate": 9.919620525676036e-05, "loss": 0.0686, "step": 20860 }, { "grad_norm": 0.6833270192146301, "learning_rate": 9.919472814229528e-05, "loss": 0.0532, "step": 20870 }, { "grad_norm": 0.9054814577102661, "learning_rate": 9.919324968286481e-05, "loss": 0.0849, "step": 20880 }, { "grad_norm": 0.8409886360168457, "learning_rate": 9.919176987850938e-05, "loss": 0.0867, "step": 20890 }, { "grad_norm": 0.5549073815345764, "learning_rate": 9.919028872926946e-05, "loss": 0.0907, "step": 20900 }, { "grad_norm": 0.7213913202285767, "learning_rate": 9.918880623518549e-05, "loss": 0.0792, "step": 20910 }, { "grad_norm": 1.1530420780181885, "learning_rate": 9.918732239629806e-05, "loss": 0.0821, "step": 20920 }, { "grad_norm": 0.5931182503700256, "learning_rate": 9.918583721264771e-05, "loss": 0.0887, "step": 20930 }, { "grad_norm": 0.6034256219863892, "learning_rate": 9.918435068427505e-05, "loss": 0.0663, "step": 20940 }, { "grad_norm": 0.5017592906951904, "learning_rate": 9.918286281122072e-05, "loss": 0.0751, "step": 20950 }, { "grad_norm": 0.9911549091339111, "learning_rate": 9.918137359352539e-05, "loss": 0.0698, "step": 20960 }, { "grad_norm": 0.6378101706504822, "learning_rate": 9.917988303122978e-05, "loss": 0.0674, "step": 20970 }, { "grad_norm": 0.4204789996147156, "learning_rate": 9.917839112437466e-05, "loss": 0.0601, "step": 20980 }, { "grad_norm": 0.6890564560890198, "learning_rate": 9.917689787300079e-05, "loss": 0.0611, "step": 20990 }, { "grad_norm": 0.46391019225120544, "learning_rate": 9.917540327714902e-05, "loss": 0.0657, "step": 21000 }, { "grad_norm": 0.5652409195899963, "learning_rate": 9.917390733686018e-05, "loss": 0.065, "step": 21010 }, { "grad_norm": 0.6425219774246216, "learning_rate": 9.91724100521752e-05, "loss": 0.0941, "step": 21020 }, { "grad_norm": 0.7409473061561584, "learning_rate": 9.917091142313499e-05, "loss": 0.0636, "step": 21030 }, { "grad_norm": 0.7623005509376526, "learning_rate": 9.916941144978056e-05, "loss": 0.0673, "step": 21040 }, { "grad_norm": 0.49288854002952576, "learning_rate": 9.916791013215288e-05, "loss": 0.0782, "step": 21050 }, { "grad_norm": 0.5485000610351562, "learning_rate": 9.9166407470293e-05, "loss": 0.0751, "step": 21060 }, { "grad_norm": 2.0505237579345703, "learning_rate": 9.9164903464242e-05, "loss": 0.0864, "step": 21070 }, { "grad_norm": 0.7012588977813721, "learning_rate": 9.916339811404104e-05, "loss": 0.0734, "step": 21080 }, { "grad_norm": 0.6504430174827576, "learning_rate": 9.916189141973122e-05, "loss": 0.0604, "step": 21090 }, { "grad_norm": 0.7268528342247009, "learning_rate": 9.916038338135378e-05, "loss": 0.0717, "step": 21100 }, { "grad_norm": 0.5579605102539062, "learning_rate": 9.915887399894992e-05, "loss": 0.0713, "step": 21110 }, { "grad_norm": 0.9375287890434265, "learning_rate": 9.91573632725609e-05, "loss": 0.0799, "step": 21120 }, { "grad_norm": 0.6937346458435059, "learning_rate": 9.915585120222806e-05, "loss": 0.0727, "step": 21130 }, { "grad_norm": 0.5330371260643005, "learning_rate": 9.915433778799269e-05, "loss": 0.0646, "step": 21140 }, { "grad_norm": 1.0943412780761719, "learning_rate": 9.915282302989619e-05, "loss": 0.0883, "step": 21150 }, { "grad_norm": 0.6259918212890625, "learning_rate": 9.915130692798e-05, "loss": 0.0937, "step": 21160 }, { "grad_norm": 0.8803867101669312, "learning_rate": 9.914978948228553e-05, "loss": 0.0971, "step": 21170 }, { "grad_norm": 0.7395297884941101, "learning_rate": 9.914827069285427e-05, "loss": 0.086, "step": 21180 }, { "grad_norm": 0.6466926336288452, "learning_rate": 9.914675055972777e-05, "loss": 0.0831, "step": 21190 }, { "grad_norm": 0.6298440098762512, "learning_rate": 9.914522908294757e-05, "loss": 0.0846, "step": 21200 }, { "grad_norm": 0.568876326084137, "learning_rate": 9.914370626255526e-05, "loss": 0.1023, "step": 21210 }, { "grad_norm": 0.4519944489002228, "learning_rate": 9.914218209859249e-05, "loss": 0.0861, "step": 21220 }, { "grad_norm": 1.0553175210952759, "learning_rate": 9.914065659110092e-05, "loss": 0.1061, "step": 21230 }, { "grad_norm": 1.073925495147705, "learning_rate": 9.913912974012228e-05, "loss": 0.0849, "step": 21240 }, { "grad_norm": 0.903737485408783, "learning_rate": 9.913760154569827e-05, "loss": 0.0766, "step": 21250 }, { "grad_norm": 0.6134679913520813, "learning_rate": 9.913607200787069e-05, "loss": 0.0761, "step": 21260 }, { "grad_norm": 1.5311802625656128, "learning_rate": 9.913454112668136e-05, "loss": 0.0911, "step": 21270 }, { "grad_norm": 2.3193297386169434, "learning_rate": 9.913300890217216e-05, "loss": 0.0878, "step": 21280 }, { "grad_norm": 0.6593215465545654, "learning_rate": 9.913147533438494e-05, "loss": 0.0954, "step": 21290 }, { "grad_norm": 0.6412304639816284, "learning_rate": 9.912994042336164e-05, "loss": 0.0765, "step": 21300 }, { "grad_norm": 0.8071780800819397, "learning_rate": 9.912840416914423e-05, "loss": 0.085, "step": 21310 }, { "grad_norm": 0.5547884702682495, "learning_rate": 9.91268665717747e-05, "loss": 0.069, "step": 21320 }, { "grad_norm": 0.9333174228668213, "learning_rate": 9.912532763129511e-05, "loss": 0.088, "step": 21330 }, { "grad_norm": 0.7414964437484741, "learning_rate": 9.91237873477475e-05, "loss": 0.0789, "step": 21340 }, { "grad_norm": 0.6653682589530945, "learning_rate": 9.912224572117402e-05, "loss": 0.0645, "step": 21350 }, { "grad_norm": 0.5332919955253601, "learning_rate": 9.912070275161677e-05, "loss": 0.0745, "step": 21360 }, { "grad_norm": 0.6270144581794739, "learning_rate": 9.911915843911797e-05, "loss": 0.0719, "step": 21370 }, { "grad_norm": 0.6532980799674988, "learning_rate": 9.911761278371985e-05, "loss": 0.0655, "step": 21380 }, { "grad_norm": 0.7794709801673889, "learning_rate": 9.911606578546462e-05, "loss": 0.0714, "step": 21390 }, { "grad_norm": 0.8336079120635986, "learning_rate": 9.911451744439461e-05, "loss": 0.1229, "step": 21400 }, { "grad_norm": 0.6057979464530945, "learning_rate": 9.911296776055217e-05, "loss": 0.0992, "step": 21410 }, { "grad_norm": 0.6999597549438477, "learning_rate": 9.911141673397961e-05, "loss": 0.0739, "step": 21420 }, { "grad_norm": 0.4732632040977478, "learning_rate": 9.910986436471938e-05, "loss": 0.0622, "step": 21430 }, { "grad_norm": 0.5589165687561035, "learning_rate": 9.910831065281391e-05, "loss": 0.0817, "step": 21440 }, { "grad_norm": 0.6074422001838684, "learning_rate": 9.91067555983057e-05, "loss": 0.0772, "step": 21450 }, { "grad_norm": 0.6262595057487488, "learning_rate": 9.910519920123719e-05, "loss": 0.0668, "step": 21460 }, { "grad_norm": 0.6680309772491455, "learning_rate": 9.910364146165102e-05, "loss": 0.0666, "step": 21470 }, { "grad_norm": 0.7086575031280518, "learning_rate": 9.910208237958974e-05, "loss": 0.0634, "step": 21480 }, { "grad_norm": 0.7655069828033447, "learning_rate": 9.910052195509595e-05, "loss": 0.0774, "step": 21490 }, { "grad_norm": 0.7654222846031189, "learning_rate": 9.909896018821234e-05, "loss": 0.092, "step": 21500 }, { "grad_norm": 0.6973251700401306, "learning_rate": 9.909739707898162e-05, "loss": 0.0928, "step": 21510 }, { "grad_norm": 0.6227676272392273, "learning_rate": 9.909583262744651e-05, "loss": 0.0674, "step": 21520 }, { "grad_norm": 0.5391992330551147, "learning_rate": 9.909426683364975e-05, "loss": 0.065, "step": 21530 }, { "grad_norm": 0.8288220763206482, "learning_rate": 9.909269969763422e-05, "loss": 0.0864, "step": 21540 }, { "grad_norm": 0.6434508562088013, "learning_rate": 9.909113121944269e-05, "loss": 0.0764, "step": 21550 }, { "grad_norm": 0.7599574327468872, "learning_rate": 9.90895613991181e-05, "loss": 0.0861, "step": 21560 }, { "grad_norm": 0.6515893936157227, "learning_rate": 9.908799023670333e-05, "loss": 0.0657, "step": 21570 }, { "grad_norm": 0.5859254002571106, "learning_rate": 9.908641773224134e-05, "loss": 0.0837, "step": 21580 }, { "grad_norm": 0.6964009404182434, "learning_rate": 9.908484388577514e-05, "loss": 0.096, "step": 21590 }, { "grad_norm": 0.6893118023872375, "learning_rate": 9.908326869734775e-05, "loss": 0.0893, "step": 21600 }, { "grad_norm": 0.6535158753395081, "learning_rate": 9.908169216700223e-05, "loss": 0.0637, "step": 21610 }, { "grad_norm": 0.6627980470657349, "learning_rate": 9.90801142947817e-05, "loss": 0.0756, "step": 21620 }, { "grad_norm": 0.55002361536026, "learning_rate": 9.907853508072927e-05, "loss": 0.0829, "step": 21630 }, { "grad_norm": 0.6559059023857117, "learning_rate": 9.907695452488812e-05, "loss": 0.0762, "step": 21640 }, { "grad_norm": 0.5466291308403015, "learning_rate": 9.907537262730149e-05, "loss": 0.073, "step": 21650 }, { "grad_norm": 0.7469327449798584, "learning_rate": 9.907378938801259e-05, "loss": 0.0776, "step": 21660 }, { "grad_norm": 0.7310678958892822, "learning_rate": 9.907220480706474e-05, "loss": 0.0994, "step": 21670 }, { "grad_norm": 0.5527225136756897, "learning_rate": 9.907061888450125e-05, "loss": 0.0605, "step": 21680 }, { "grad_norm": 0.5020173192024231, "learning_rate": 9.906903162036545e-05, "loss": 0.0694, "step": 21690 }, { "grad_norm": 0.9658920764923096, "learning_rate": 9.906744301470079e-05, "loss": 0.0898, "step": 21700 }, { "grad_norm": 0.5032933950424194, "learning_rate": 9.906585306755066e-05, "loss": 0.0736, "step": 21710 }, { "grad_norm": 0.6539530754089355, "learning_rate": 9.906426177895852e-05, "loss": 0.0744, "step": 21720 }, { "grad_norm": 0.693108320236206, "learning_rate": 9.90626691489679e-05, "loss": 0.0839, "step": 21730 }, { "grad_norm": 0.45658108592033386, "learning_rate": 9.906107517762238e-05, "loss": 0.0777, "step": 21740 }, { "grad_norm": 0.48961564898490906, "learning_rate": 9.905947986496546e-05, "loss": 0.0856, "step": 21750 }, { "grad_norm": 0.7029410600662231, "learning_rate": 9.905788321104079e-05, "loss": 0.0706, "step": 21760 }, { "grad_norm": 0.7790977954864502, "learning_rate": 9.905628521589204e-05, "loss": 0.0637, "step": 21770 }, { "grad_norm": 0.5555375218391418, "learning_rate": 9.905468587956285e-05, "loss": 0.0803, "step": 21780 }, { "grad_norm": 0.41178739070892334, "learning_rate": 9.905308520209702e-05, "loss": 0.0715, "step": 21790 }, { "grad_norm": 0.5468810200691223, "learning_rate": 9.905148318353824e-05, "loss": 0.0661, "step": 21800 }, { "grad_norm": 1.0380253791809082, "learning_rate": 9.904987982393035e-05, "loss": 0.0786, "step": 21810 }, { "grad_norm": 0.5023011565208435, "learning_rate": 9.904827512331717e-05, "loss": 0.0594, "step": 21820 }, { "grad_norm": 0.46022993326187134, "learning_rate": 9.904666908174257e-05, "loss": 0.071, "step": 21830 }, { "grad_norm": 0.85048907995224, "learning_rate": 9.904506169925047e-05, "loss": 0.0793, "step": 21840 }, { "grad_norm": 0.5784841775894165, "learning_rate": 9.90434529758848e-05, "loss": 0.0933, "step": 21850 }, { "grad_norm": 1.6930389404296875, "learning_rate": 9.904184291168956e-05, "loss": 0.0841, "step": 21860 }, { "grad_norm": 0.4674481153488159, "learning_rate": 9.904023150670876e-05, "loss": 0.0823, "step": 21870 }, { "grad_norm": 0.6654800772666931, "learning_rate": 9.903861876098644e-05, "loss": 0.086, "step": 21880 }, { "grad_norm": 0.5353226661682129, "learning_rate": 9.903700467456672e-05, "loss": 0.062, "step": 21890 }, { "grad_norm": 0.6680715084075928, "learning_rate": 9.903538924749371e-05, "loss": 0.0645, "step": 21900 }, { "grad_norm": 0.6927723288536072, "learning_rate": 9.903377247981159e-05, "loss": 0.0652, "step": 21910 }, { "grad_norm": 0.5352277159690857, "learning_rate": 9.903215437156452e-05, "loss": 0.057, "step": 21920 }, { "grad_norm": 0.4942409098148346, "learning_rate": 9.903053492279679e-05, "loss": 0.0704, "step": 21930 }, { "grad_norm": 0.5830907821655273, "learning_rate": 9.902891413355267e-05, "loss": 0.0689, "step": 21940 }, { "grad_norm": 0.6310718059539795, "learning_rate": 9.902729200387644e-05, "loss": 0.0745, "step": 21950 }, { "grad_norm": 0.7390317916870117, "learning_rate": 9.902566853381246e-05, "loss": 0.0595, "step": 21960 }, { "grad_norm": 0.4207310378551483, "learning_rate": 9.902404372340514e-05, "loss": 0.0558, "step": 21970 }, { "grad_norm": 0.5491604804992676, "learning_rate": 9.902241757269886e-05, "loss": 0.059, "step": 21980 }, { "grad_norm": 0.7438673377037048, "learning_rate": 9.902079008173811e-05, "loss": 0.0637, "step": 21990 }, { "grad_norm": 0.6679186820983887, "learning_rate": 9.901916125056738e-05, "loss": 0.0636, "step": 22000 }, { "grad_norm": 0.8742456436157227, "learning_rate": 9.90175310792312e-05, "loss": 0.0662, "step": 22010 }, { "grad_norm": 1.2557514905929565, "learning_rate": 9.901589956777413e-05, "loss": 0.0675, "step": 22020 }, { "grad_norm": 0.5898954272270203, "learning_rate": 9.901426671624077e-05, "loss": 0.07, "step": 22030 }, { "grad_norm": 0.5910346508026123, "learning_rate": 9.901263252467579e-05, "loss": 0.0696, "step": 22040 }, { "grad_norm": 0.6996574997901917, "learning_rate": 9.901099699312384e-05, "loss": 0.0668, "step": 22050 }, { "grad_norm": 1.1650335788726807, "learning_rate": 9.900936012162965e-05, "loss": 0.0836, "step": 22060 }, { "grad_norm": 0.5775374174118042, "learning_rate": 9.900772191023795e-05, "loss": 0.0599, "step": 22070 }, { "grad_norm": 0.6489575505256653, "learning_rate": 9.900608235899356e-05, "loss": 0.0715, "step": 22080 }, { "grad_norm": 0.5439215302467346, "learning_rate": 9.900444146794128e-05, "loss": 0.0683, "step": 22090 }, { "grad_norm": 0.7547674775123596, "learning_rate": 9.900279923712598e-05, "loss": 0.0584, "step": 22100 }, { "grad_norm": 0.7714028358459473, "learning_rate": 9.900115566659257e-05, "loss": 0.0635, "step": 22110 }, { "grad_norm": 0.6648194789886475, "learning_rate": 9.899951075638596e-05, "loss": 0.0794, "step": 22120 }, { "grad_norm": 0.5451257228851318, "learning_rate": 9.899786450655116e-05, "loss": 0.0686, "step": 22130 }, { "grad_norm": 0.43146419525146484, "learning_rate": 9.899621691713313e-05, "loss": 0.0615, "step": 22140 }, { "grad_norm": 0.519637942314148, "learning_rate": 9.899456798817693e-05, "loss": 0.0602, "step": 22150 }, { "grad_norm": 0.5895187854766846, "learning_rate": 9.899291771972765e-05, "loss": 0.0713, "step": 22160 }, { "grad_norm": 0.6728572249412537, "learning_rate": 9.899126611183042e-05, "loss": 0.0649, "step": 22170 }, { "grad_norm": 0.6260247230529785, "learning_rate": 9.898961316453036e-05, "loss": 0.075, "step": 22180 }, { "grad_norm": 0.5529462099075317, "learning_rate": 9.898795887787271e-05, "loss": 0.06, "step": 22190 }, { "grad_norm": 0.6379142999649048, "learning_rate": 9.898630325190265e-05, "loss": 0.0847, "step": 22200 }, { "grad_norm": 0.4602504372596741, "learning_rate": 9.898464628666546e-05, "loss": 0.0626, "step": 22210 }, { "grad_norm": 0.629386842250824, "learning_rate": 9.898298798220644e-05, "loss": 0.0946, "step": 22220 }, { "grad_norm": 0.5281068682670593, "learning_rate": 9.898132833857094e-05, "loss": 0.0568, "step": 22230 }, { "grad_norm": 0.8579805493354797, "learning_rate": 9.897966735580433e-05, "loss": 0.0881, "step": 22240 }, { "grad_norm": 0.5904453992843628, "learning_rate": 9.8978005033952e-05, "loss": 0.094, "step": 22250 }, { "grad_norm": 0.40490201115608215, "learning_rate": 9.897634137305943e-05, "loss": 0.0743, "step": 22260 }, { "grad_norm": 0.5274161100387573, "learning_rate": 9.897467637317207e-05, "loss": 0.0686, "step": 22270 }, { "grad_norm": 0.5208621025085449, "learning_rate": 9.897301003433545e-05, "loss": 0.0673, "step": 22280 }, { "grad_norm": 0.4981524646282196, "learning_rate": 9.897134235659515e-05, "loss": 0.051, "step": 22290 }, { "grad_norm": 0.8647127747535706, "learning_rate": 9.896967333999676e-05, "loss": 0.0677, "step": 22300 }, { "grad_norm": 0.7453901767730713, "learning_rate": 9.896800298458588e-05, "loss": 0.0693, "step": 22310 }, { "grad_norm": 0.9313946962356567, "learning_rate": 9.896633129040819e-05, "loss": 0.1001, "step": 22320 }, { "grad_norm": 0.6983513832092285, "learning_rate": 9.89646582575094e-05, "loss": 0.0793, "step": 22330 }, { "grad_norm": 0.6909811496734619, "learning_rate": 9.896298388593525e-05, "loss": 0.0739, "step": 22340 }, { "grad_norm": 0.5666958689689636, "learning_rate": 9.896130817573153e-05, "loss": 0.0824, "step": 22350 }, { "grad_norm": 0.8157311081886292, "learning_rate": 9.895963112694403e-05, "loss": 0.0788, "step": 22360 }, { "grad_norm": 0.6572942733764648, "learning_rate": 9.895795273961861e-05, "loss": 0.0808, "step": 22370 }, { "grad_norm": 1.3796948194503784, "learning_rate": 9.895627301380115e-05, "loss": 0.084, "step": 22380 }, { "grad_norm": 0.7535858154296875, "learning_rate": 9.895459194953758e-05, "loss": 0.0846, "step": 22390 }, { "grad_norm": 0.5913856625556946, "learning_rate": 9.895290954687385e-05, "loss": 0.0808, "step": 22400 }, { "grad_norm": 0.5149309039115906, "learning_rate": 9.895122580585597e-05, "loss": 0.0668, "step": 22410 }, { "grad_norm": 0.8555492758750916, "learning_rate": 9.894954072652996e-05, "loss": 0.0791, "step": 22420 }, { "grad_norm": 0.6215652823448181, "learning_rate": 9.894785430894191e-05, "loss": 0.0674, "step": 22430 }, { "grad_norm": 0.558629035949707, "learning_rate": 9.894616655313791e-05, "loss": 0.082, "step": 22440 }, { "grad_norm": 0.7194496989250183, "learning_rate": 9.894447745916409e-05, "loss": 0.1166, "step": 22450 }, { "grad_norm": 1.1513625383377075, "learning_rate": 9.894278702706665e-05, "loss": 0.0721, "step": 22460 }, { "grad_norm": 1.0960716009140015, "learning_rate": 9.89410952568918e-05, "loss": 0.1021, "step": 22470 }, { "grad_norm": 0.8474432826042175, "learning_rate": 9.89394021486858e-05, "loss": 0.0721, "step": 22480 }, { "grad_norm": 0.749993622303009, "learning_rate": 9.893770770249493e-05, "loss": 0.0601, "step": 22490 }, { "grad_norm": 0.5810113549232483, "learning_rate": 9.89360119183655e-05, "loss": 0.0822, "step": 22500 }, { "grad_norm": 0.6906539797782898, "learning_rate": 9.893431479634391e-05, "loss": 0.077, "step": 22510 }, { "grad_norm": 0.8280851244926453, "learning_rate": 9.893261633647652e-05, "loss": 0.0835, "step": 22520 }, { "grad_norm": 0.553741455078125, "learning_rate": 9.893091653880979e-05, "loss": 0.0689, "step": 22530 }, { "grad_norm": 0.6001688838005066, "learning_rate": 9.89292154033902e-05, "loss": 0.0708, "step": 22540 }, { "grad_norm": 0.7928604483604431, "learning_rate": 9.89275129302642e-05, "loss": 0.0676, "step": 22550 }, { "grad_norm": 0.7464116215705872, "learning_rate": 9.892580911947841e-05, "loss": 0.1012, "step": 22560 }, { "grad_norm": 0.7381909489631653, "learning_rate": 9.892410397107939e-05, "loss": 0.0871, "step": 22570 }, { "grad_norm": 0.8065599203109741, "learning_rate": 9.892239748511372e-05, "loss": 0.077, "step": 22580 }, { "grad_norm": 0.62738037109375, "learning_rate": 9.89206896616281e-05, "loss": 0.0726, "step": 22590 }, { "grad_norm": 0.5696205496788025, "learning_rate": 9.891898050066919e-05, "loss": 0.0745, "step": 22600 }, { "grad_norm": 0.7454226613044739, "learning_rate": 9.891727000228374e-05, "loss": 0.0648, "step": 22610 }, { "grad_norm": 0.7307091355323792, "learning_rate": 9.891555816651851e-05, "loss": 0.0876, "step": 22620 }, { "grad_norm": 0.7896071076393127, "learning_rate": 9.89138449934203e-05, "loss": 0.0729, "step": 22630 }, { "grad_norm": 0.5193443298339844, "learning_rate": 9.891213048303592e-05, "loss": 0.0713, "step": 22640 }, { "grad_norm": 0.7968478202819824, "learning_rate": 9.891041463541228e-05, "loss": 0.0605, "step": 22650 }, { "grad_norm": 0.5369075536727905, "learning_rate": 9.890869745059628e-05, "loss": 0.0626, "step": 22660 }, { "grad_norm": 0.8410562872886658, "learning_rate": 9.890697892863487e-05, "loss": 0.0601, "step": 22670 }, { "grad_norm": 0.8828100562095642, "learning_rate": 9.890525906957501e-05, "loss": 0.0709, "step": 22680 }, { "grad_norm": 0.5819839239120483, "learning_rate": 9.890353787346377e-05, "loss": 0.0976, "step": 22690 }, { "grad_norm": 0.9633632898330688, "learning_rate": 9.890181534034817e-05, "loss": 0.0752, "step": 22700 }, { "grad_norm": 0.7510591149330139, "learning_rate": 9.89000914702753e-05, "loss": 0.0683, "step": 22710 }, { "grad_norm": 0.568821907043457, "learning_rate": 9.889836626329229e-05, "loss": 0.0739, "step": 22720 }, { "grad_norm": 0.7952963709831238, "learning_rate": 9.889663971944634e-05, "loss": 0.0529, "step": 22730 }, { "grad_norm": 0.6263118386268616, "learning_rate": 9.88949118387846e-05, "loss": 0.0642, "step": 22740 }, { "grad_norm": 0.766901433467865, "learning_rate": 9.889318262135438e-05, "loss": 0.0743, "step": 22750 }, { "grad_norm": 0.45587649941444397, "learning_rate": 9.88914520672029e-05, "loss": 0.0811, "step": 22760 }, { "grad_norm": 0.5882430672645569, "learning_rate": 9.888972017637747e-05, "loss": 0.0941, "step": 22770 }, { "grad_norm": 0.6139439940452576, "learning_rate": 9.888798694892548e-05, "loss": 0.067, "step": 22780 }, { "grad_norm": 0.7149098515510559, "learning_rate": 9.888625238489428e-05, "loss": 0.0695, "step": 22790 }, { "grad_norm": 0.6348226070404053, "learning_rate": 9.888451648433131e-05, "loss": 0.0824, "step": 22800 }, { "grad_norm": 0.7234693765640259, "learning_rate": 9.888277924728402e-05, "loss": 0.0541, "step": 22810 }, { "grad_norm": 0.6377395391464233, "learning_rate": 9.88810406737999e-05, "loss": 0.0685, "step": 22820 }, { "grad_norm": 0.6783674359321594, "learning_rate": 9.887930076392651e-05, "loss": 0.0975, "step": 22830 }, { "grad_norm": 0.9546112418174744, "learning_rate": 9.887755951771139e-05, "loss": 0.0727, "step": 22840 }, { "grad_norm": 0.6618374586105347, "learning_rate": 9.887581693520216e-05, "loss": 0.0717, "step": 22850 }, { "grad_norm": 0.6746759414672852, "learning_rate": 9.887407301644646e-05, "loss": 0.0718, "step": 22860 }, { "grad_norm": 1.0000317096710205, "learning_rate": 9.887232776149195e-05, "loss": 0.0739, "step": 22870 }, { "grad_norm": 0.672568678855896, "learning_rate": 9.887058117038637e-05, "loss": 0.0758, "step": 22880 }, { "grad_norm": 0.6200418472290039, "learning_rate": 9.886883324317748e-05, "loss": 0.0648, "step": 22890 }, { "grad_norm": 0.7339642643928528, "learning_rate": 9.886708397991302e-05, "loss": 0.0783, "step": 22900 }, { "grad_norm": 0.6186084151268005, "learning_rate": 9.886533338064085e-05, "loss": 0.092, "step": 22910 }, { "grad_norm": 0.7336443662643433, "learning_rate": 9.886358144540882e-05, "loss": 0.0624, "step": 22920 }, { "grad_norm": 0.5330392122268677, "learning_rate": 9.886182817426483e-05, "loss": 0.0543, "step": 22930 }, { "grad_norm": 0.7779034972190857, "learning_rate": 9.886007356725685e-05, "loss": 0.0676, "step": 22940 }, { "grad_norm": 0.8071516156196594, "learning_rate": 9.885831762443276e-05, "loss": 0.0757, "step": 22950 }, { "grad_norm": 0.6127995848655701, "learning_rate": 9.885656034584065e-05, "loss": 0.0716, "step": 22960 }, { "grad_norm": 0.5958728790283203, "learning_rate": 9.885480173152854e-05, "loss": 0.0787, "step": 22970 }, { "grad_norm": 0.8493669033050537, "learning_rate": 9.885304178154449e-05, "loss": 0.0798, "step": 22980 }, { "grad_norm": 0.5803672671318054, "learning_rate": 9.885128049593664e-05, "loss": 0.0693, "step": 22990 }, { "grad_norm": 0.8153852224349976, "learning_rate": 9.884951787475313e-05, "loss": 0.0821, "step": 23000 }, { "grad_norm": 0.4158003330230713, "learning_rate": 9.884775391804216e-05, "loss": 0.0563, "step": 23010 }, { "grad_norm": 0.5639269351959229, "learning_rate": 9.884598862585193e-05, "loss": 0.0566, "step": 23020 }, { "grad_norm": 0.6796708106994629, "learning_rate": 9.884422199823073e-05, "loss": 0.0636, "step": 23030 }, { "grad_norm": 0.6226152181625366, "learning_rate": 9.884245403522686e-05, "loss": 0.0539, "step": 23040 }, { "grad_norm": 0.8246045112609863, "learning_rate": 9.884068473688863e-05, "loss": 0.0652, "step": 23050 }, { "grad_norm": 0.5041261315345764, "learning_rate": 9.883891410326444e-05, "loss": 0.063, "step": 23060 }, { "grad_norm": 0.5219712853431702, "learning_rate": 9.883714213440269e-05, "loss": 0.0587, "step": 23070 }, { "grad_norm": 0.5976848602294922, "learning_rate": 9.883536883035181e-05, "loss": 0.0621, "step": 23080 }, { "grad_norm": 0.7308478951454163, "learning_rate": 9.883359419116031e-05, "loss": 0.0757, "step": 23090 }, { "grad_norm": 0.6313096284866333, "learning_rate": 9.883181821687669e-05, "loss": 0.0629, "step": 23100 }, { "grad_norm": 0.6866416931152344, "learning_rate": 9.883004090754948e-05, "loss": 0.0769, "step": 23110 }, { "grad_norm": 0.5357614159584045, "learning_rate": 9.882826226322731e-05, "loss": 0.0704, "step": 23120 }, { "grad_norm": 0.6118890047073364, "learning_rate": 9.882648228395879e-05, "loss": 0.0648, "step": 23130 }, { "grad_norm": 0.825923502445221, "learning_rate": 9.882470096979259e-05, "loss": 0.0886, "step": 23140 }, { "grad_norm": 0.5600388646125793, "learning_rate": 9.882291832077741e-05, "loss": 0.0735, "step": 23150 }, { "grad_norm": 0.6524426341056824, "learning_rate": 9.882113433696199e-05, "loss": 0.0833, "step": 23160 }, { "grad_norm": 0.5376080274581909, "learning_rate": 9.88193490183951e-05, "loss": 0.0809, "step": 23170 }, { "grad_norm": 0.8198995590209961, "learning_rate": 9.881756236512554e-05, "loss": 0.09, "step": 23180 }, { "grad_norm": 0.7852233052253723, "learning_rate": 9.881577437720217e-05, "loss": 0.0662, "step": 23190 }, { "grad_norm": 0.8142662644386292, "learning_rate": 9.881398505467385e-05, "loss": 0.0687, "step": 23200 }, { "grad_norm": 0.6841519474983215, "learning_rate": 9.881219439758954e-05, "loss": 0.1049, "step": 23210 }, { "grad_norm": 0.7606642842292786, "learning_rate": 9.881040240599817e-05, "loss": 0.0832, "step": 23220 }, { "grad_norm": 0.6031937599182129, "learning_rate": 9.880860907994873e-05, "loss": 0.0924, "step": 23230 }, { "grad_norm": 0.9012818336486816, "learning_rate": 9.880681441949025e-05, "loss": 0.0914, "step": 23240 }, { "grad_norm": 0.4507986009120941, "learning_rate": 9.880501842467181e-05, "loss": 0.0839, "step": 23250 }, { "grad_norm": 1.174417495727539, "learning_rate": 9.880322109554249e-05, "loss": 0.1157, "step": 23260 }, { "grad_norm": 0.45794054865837097, "learning_rate": 9.880142243215145e-05, "loss": 0.071, "step": 23270 }, { "grad_norm": 0.7086172103881836, "learning_rate": 9.879962243454783e-05, "loss": 0.0783, "step": 23280 }, { "grad_norm": 0.9832475185394287, "learning_rate": 9.87978211027809e-05, "loss": 0.0841, "step": 23290 }, { "grad_norm": 0.7294554710388184, "learning_rate": 9.879601843689986e-05, "loss": 0.0786, "step": 23300 }, { "grad_norm": 0.8837796449661255, "learning_rate": 9.8794214436954e-05, "loss": 0.075, "step": 23310 }, { "grad_norm": 0.5914982557296753, "learning_rate": 9.879240910299265e-05, "loss": 0.067, "step": 23320 }, { "grad_norm": 0.663658082485199, "learning_rate": 9.879060243506518e-05, "loss": 0.0641, "step": 23330 }, { "grad_norm": 0.3747791349887848, "learning_rate": 9.878879443322095e-05, "loss": 0.0569, "step": 23340 }, { "grad_norm": 0.567226231098175, "learning_rate": 9.878698509750944e-05, "loss": 0.0788, "step": 23350 }, { "grad_norm": 0.5251399278640747, "learning_rate": 9.878517442798006e-05, "loss": 0.0674, "step": 23360 }, { "grad_norm": 0.6384533047676086, "learning_rate": 9.878336242468234e-05, "loss": 0.0691, "step": 23370 }, { "grad_norm": 0.660938560962677, "learning_rate": 9.878154908766581e-05, "loss": 0.0763, "step": 23380 }, { "grad_norm": 0.8496438264846802, "learning_rate": 9.877973441698007e-05, "loss": 0.0756, "step": 23390 }, { "grad_norm": 0.6655474305152893, "learning_rate": 9.877791841267472e-05, "loss": 0.0774, "step": 23400 }, { "grad_norm": 0.5714619159698486, "learning_rate": 9.87761010747994e-05, "loss": 0.0681, "step": 23410 }, { "grad_norm": 0.5589173436164856, "learning_rate": 9.87742824034038e-05, "loss": 0.0758, "step": 23420 }, { "grad_norm": 0.675061821937561, "learning_rate": 9.877246239853764e-05, "loss": 0.0968, "step": 23430 }, { "grad_norm": 1.1793268918991089, "learning_rate": 9.877064106025069e-05, "loss": 0.0679, "step": 23440 }, { "grad_norm": 0.795544445514679, "learning_rate": 9.876881838859273e-05, "loss": 0.0564, "step": 23450 }, { "grad_norm": 0.5508422255516052, "learning_rate": 9.876699438361359e-05, "loss": 0.0757, "step": 23460 }, { "grad_norm": 0.4496157169342041, "learning_rate": 9.876516904536315e-05, "loss": 0.0533, "step": 23470 }, { "grad_norm": 0.7584537863731384, "learning_rate": 9.876334237389131e-05, "loss": 0.0845, "step": 23480 }, { "grad_norm": 0.533897876739502, "learning_rate": 9.876151436924801e-05, "loss": 0.0541, "step": 23490 }, { "grad_norm": 0.6122130751609802, "learning_rate": 9.875968503148323e-05, "loss": 0.1005, "step": 23500 }, { "grad_norm": 0.8201525807380676, "learning_rate": 9.875785436064697e-05, "loss": 0.0744, "step": 23510 }, { "grad_norm": 0.529263973236084, "learning_rate": 9.875602235678929e-05, "loss": 0.057, "step": 23520 }, { "grad_norm": 0.5875462889671326, "learning_rate": 9.875418901996027e-05, "loss": 0.0879, "step": 23530 }, { "grad_norm": 0.7264320254325867, "learning_rate": 9.875235435021007e-05, "loss": 0.0695, "step": 23540 }, { "grad_norm": 0.42825260758399963, "learning_rate": 9.875051834758878e-05, "loss": 0.0565, "step": 23550 }, { "grad_norm": 0.6706088781356812, "learning_rate": 9.874868101214664e-05, "loss": 0.0807, "step": 23560 }, { "grad_norm": 0.4818483591079712, "learning_rate": 9.87468423439339e-05, "loss": 0.063, "step": 23570 }, { "grad_norm": 1.222352147102356, "learning_rate": 9.874500234300078e-05, "loss": 0.0686, "step": 23580 }, { "grad_norm": 1.1781673431396484, "learning_rate": 9.874316100939763e-05, "loss": 0.0738, "step": 23590 }, { "grad_norm": 0.6353667378425598, "learning_rate": 9.874131834317473e-05, "loss": 0.0687, "step": 23600 }, { "grad_norm": 0.6143390536308289, "learning_rate": 9.873947434438253e-05, "loss": 0.0976, "step": 23610 }, { "grad_norm": 0.5148978233337402, "learning_rate": 9.87376290130714e-05, "loss": 0.0846, "step": 23620 }, { "grad_norm": 0.7696484327316284, "learning_rate": 9.873578234929182e-05, "loss": 0.0738, "step": 23630 }, { "grad_norm": 0.5934996008872986, "learning_rate": 9.873393435309424e-05, "loss": 0.0699, "step": 23640 }, { "grad_norm": 0.7500200867652893, "learning_rate": 9.873208502452921e-05, "loss": 0.0774, "step": 23650 }, { "grad_norm": 0.6500495672225952, "learning_rate": 9.873023436364728e-05, "loss": 0.1136, "step": 23660 }, { "grad_norm": 0.5702710747718811, "learning_rate": 9.872838237049907e-05, "loss": 0.0611, "step": 23670 }, { "grad_norm": 0.6924976706504822, "learning_rate": 9.872652904513516e-05, "loss": 0.0804, "step": 23680 }, { "grad_norm": 1.5134546756744385, "learning_rate": 9.872467438760626e-05, "loss": 0.0708, "step": 23690 }, { "grad_norm": 0.5956273674964905, "learning_rate": 9.87228183979631e-05, "loss": 0.0723, "step": 23700 }, { "grad_norm": 1.182530403137207, "learning_rate": 9.872096107625634e-05, "loss": 0.0927, "step": 23710 }, { "grad_norm": 1.2074503898620605, "learning_rate": 9.871910242253683e-05, "loss": 0.0707, "step": 23720 }, { "grad_norm": 0.4731355905532837, "learning_rate": 9.871724243685538e-05, "loss": 0.0685, "step": 23730 }, { "grad_norm": 0.6125494241714478, "learning_rate": 9.87153811192628e-05, "loss": 0.0659, "step": 23740 }, { "grad_norm": 0.5584636926651001, "learning_rate": 9.871351846981e-05, "loss": 0.0966, "step": 23750 }, { "grad_norm": 0.514604926109314, "learning_rate": 9.871165448854794e-05, "loss": 0.0714, "step": 23760 }, { "grad_norm": 0.5023837685585022, "learning_rate": 9.87097891755275e-05, "loss": 0.0552, "step": 23770 }, { "grad_norm": 0.517258882522583, "learning_rate": 9.870792253079976e-05, "loss": 0.0664, "step": 23780 }, { "grad_norm": 0.5759910345077515, "learning_rate": 9.87060545544157e-05, "loss": 0.062, "step": 23790 }, { "grad_norm": 0.4849579930305481, "learning_rate": 9.87041852464264e-05, "loss": 0.0719, "step": 23800 }, { "grad_norm": 0.7575476765632629, "learning_rate": 9.870231460688299e-05, "loss": 0.0751, "step": 23810 }, { "grad_norm": 0.5008485317230225, "learning_rate": 9.870044263583658e-05, "loss": 0.0672, "step": 23820 }, { "grad_norm": 0.45230576395988464, "learning_rate": 9.869856933333838e-05, "loss": 0.0708, "step": 23830 }, { "grad_norm": 0.5730262398719788, "learning_rate": 9.869669469943957e-05, "loss": 0.0654, "step": 23840 }, { "grad_norm": 0.7663208842277527, "learning_rate": 9.869481873419145e-05, "loss": 0.0828, "step": 23850 }, { "grad_norm": 0.49249067902565, "learning_rate": 9.869294143764524e-05, "loss": 0.0759, "step": 23860 }, { "grad_norm": 0.7319296598434448, "learning_rate": 9.869106280985234e-05, "loss": 0.0616, "step": 23870 }, { "grad_norm": 0.5789247155189514, "learning_rate": 9.868918285086406e-05, "loss": 0.0583, "step": 23880 }, { "grad_norm": 0.5480086207389832, "learning_rate": 9.86873015607318e-05, "loss": 0.0596, "step": 23890 }, { "grad_norm": 0.557319164276123, "learning_rate": 9.868541893950702e-05, "loss": 0.0602, "step": 23900 }, { "grad_norm": 0.6591922044754028, "learning_rate": 9.868353498724118e-05, "loss": 0.0576, "step": 23910 }, { "grad_norm": 0.7186705470085144, "learning_rate": 9.868164970398579e-05, "loss": 0.064, "step": 23920 }, { "grad_norm": 0.8587645292282104, "learning_rate": 9.867976308979236e-05, "loss": 0.0697, "step": 23930 }, { "grad_norm": 0.5040963292121887, "learning_rate": 9.867787514471252e-05, "loss": 0.0833, "step": 23940 }, { "grad_norm": 0.7357170581817627, "learning_rate": 9.867598586879785e-05, "loss": 0.07, "step": 23950 }, { "grad_norm": 0.7633041739463806, "learning_rate": 9.86740952621e-05, "loss": 0.062, "step": 23960 }, { "grad_norm": 0.9302135109901428, "learning_rate": 9.867220332467069e-05, "loss": 0.0642, "step": 23970 }, { "grad_norm": 0.6086674928665161, "learning_rate": 9.867031005656163e-05, "loss": 0.0741, "step": 23980 }, { "grad_norm": 1.1836938858032227, "learning_rate": 9.866841545782457e-05, "loss": 0.0882, "step": 23990 }, { "grad_norm": 0.5689504146575928, "learning_rate": 9.866651952851131e-05, "loss": 0.0697, "step": 24000 }, { "grad_norm": 0.769990086555481, "learning_rate": 9.86646222686737e-05, "loss": 0.0761, "step": 24010 }, { "grad_norm": 0.7354530692100525, "learning_rate": 9.86627236783636e-05, "loss": 0.0695, "step": 24020 }, { "grad_norm": 0.6129624247550964, "learning_rate": 9.86608237576329e-05, "loss": 0.0668, "step": 24030 }, { "grad_norm": 0.7943939566612244, "learning_rate": 9.865892250653357e-05, "loss": 0.0655, "step": 24040 }, { "grad_norm": 0.6499873995780945, "learning_rate": 9.865701992511757e-05, "loss": 0.0645, "step": 24050 }, { "grad_norm": 0.8359246253967285, "learning_rate": 9.865511601343696e-05, "loss": 0.074, "step": 24060 }, { "grad_norm": 0.7536929249763489, "learning_rate": 9.865321077154372e-05, "loss": 0.0817, "step": 24070 }, { "grad_norm": 0.5834808349609375, "learning_rate": 9.865130419948998e-05, "loss": 0.0721, "step": 24080 }, { "grad_norm": 0.39033278822898865, "learning_rate": 9.864939629732788e-05, "loss": 0.0546, "step": 24090 }, { "grad_norm": 1.3034785985946655, "learning_rate": 9.864748706510954e-05, "loss": 0.087, "step": 24100 }, { "grad_norm": 0.6000214219093323, "learning_rate": 9.864557650288719e-05, "loss": 0.0672, "step": 24110 }, { "grad_norm": 0.5924542546272278, "learning_rate": 9.864366461071306e-05, "loss": 0.0673, "step": 24120 }, { "grad_norm": 0.5891140699386597, "learning_rate": 9.86417513886394e-05, "loss": 0.0843, "step": 24130 }, { "grad_norm": 0.7828443050384521, "learning_rate": 9.863983683671853e-05, "loss": 0.1, "step": 24140 }, { "grad_norm": 0.44507327675819397, "learning_rate": 9.863792095500282e-05, "loss": 0.0555, "step": 24150 }, { "grad_norm": 0.5723536610603333, "learning_rate": 9.863600374354459e-05, "loss": 0.0923, "step": 24160 }, { "grad_norm": 0.5375495553016663, "learning_rate": 9.86340852023963e-05, "loss": 0.0593, "step": 24170 }, { "grad_norm": 0.556950032711029, "learning_rate": 9.863216533161041e-05, "loss": 0.067, "step": 24180 }, { "grad_norm": 0.6935908198356628, "learning_rate": 9.863024413123936e-05, "loss": 0.0764, "step": 24190 }, { "grad_norm": 0.49815425276756287, "learning_rate": 9.862832160133572e-05, "loss": 0.0809, "step": 24200 }, { "grad_norm": 0.9589572548866272, "learning_rate": 9.862639774195205e-05, "loss": 0.0771, "step": 24210 }, { "grad_norm": 0.678415060043335, "learning_rate": 9.862447255314091e-05, "loss": 0.0653, "step": 24220 }, { "grad_norm": 0.541618824005127, "learning_rate": 9.862254603495498e-05, "loss": 0.0642, "step": 24230 }, { "grad_norm": 0.6251179575920105, "learning_rate": 9.86206181874469e-05, "loss": 0.0738, "step": 24240 }, { "grad_norm": 0.745089054107666, "learning_rate": 9.861868901066937e-05, "loss": 0.0818, "step": 24250 }, { "grad_norm": 0.6659358143806458, "learning_rate": 9.861675850467517e-05, "loss": 0.075, "step": 24260 }, { "grad_norm": 0.7869814038276672, "learning_rate": 9.861482666951705e-05, "loss": 0.0703, "step": 24270 }, { "grad_norm": 0.6649611592292786, "learning_rate": 9.861289350524781e-05, "loss": 0.0546, "step": 24280 }, { "grad_norm": 0.5137720108032227, "learning_rate": 9.861095901192035e-05, "loss": 0.0671, "step": 24290 }, { "grad_norm": 0.9837621450424194, "learning_rate": 9.860902318958753e-05, "loss": 0.0706, "step": 24300 }, { "grad_norm": 0.7990312576293945, "learning_rate": 9.860708603830227e-05, "loss": 0.0667, "step": 24310 }, { "grad_norm": 0.5898919105529785, "learning_rate": 9.860514755811753e-05, "loss": 0.0706, "step": 24320 }, { "grad_norm": 0.66888028383255, "learning_rate": 9.860320774908633e-05, "loss": 0.0654, "step": 24330 }, { "grad_norm": 0.5585422515869141, "learning_rate": 9.86012666112617e-05, "loss": 0.0627, "step": 24340 }, { "grad_norm": 0.5272592902183533, "learning_rate": 9.859932414469668e-05, "loss": 0.0622, "step": 24350 }, { "grad_norm": 0.7832601070404053, "learning_rate": 9.859738034944441e-05, "loss": 0.0975, "step": 24360 }, { "grad_norm": 0.5999487638473511, "learning_rate": 9.859543522555801e-05, "loss": 0.0589, "step": 24370 }, { "grad_norm": 0.9748600721359253, "learning_rate": 9.859348877309067e-05, "loss": 0.0764, "step": 24380 }, { "grad_norm": 0.6127532124519348, "learning_rate": 9.85915409920956e-05, "loss": 0.0866, "step": 24390 }, { "grad_norm": 0.42537403106689453, "learning_rate": 9.858959188262607e-05, "loss": 0.0685, "step": 24400 }, { "grad_norm": 1.0480279922485352, "learning_rate": 9.858764144473532e-05, "loss": 0.0946, "step": 24410 }, { "grad_norm": 0.5566164255142212, "learning_rate": 9.858568967847673e-05, "loss": 0.0734, "step": 24420 }, { "grad_norm": 0.7451068162918091, "learning_rate": 9.858373658390363e-05, "loss": 0.0698, "step": 24430 }, { "grad_norm": 0.780944287776947, "learning_rate": 9.858178216106943e-05, "loss": 0.073, "step": 24440 }, { "grad_norm": 0.6216530799865723, "learning_rate": 9.857982641002755e-05, "loss": 0.0755, "step": 24450 }, { "grad_norm": 0.4217599034309387, "learning_rate": 9.857786933083146e-05, "loss": 0.0707, "step": 24460 }, { "grad_norm": 0.5128370523452759, "learning_rate": 9.857591092353469e-05, "loss": 0.0671, "step": 24470 }, { "grad_norm": 0.586203396320343, "learning_rate": 9.857395118819075e-05, "loss": 0.0606, "step": 24480 }, { "grad_norm": 0.7282930612564087, "learning_rate": 9.857199012485324e-05, "loss": 0.0955, "step": 24490 }, { "grad_norm": 0.3385649621486664, "learning_rate": 9.857002773357577e-05, "loss": 0.0627, "step": 24500 }, { "grad_norm": 0.7272547483444214, "learning_rate": 9.856806401441199e-05, "loss": 0.0863, "step": 24510 }, { "grad_norm": 0.6961891055107117, "learning_rate": 9.856609896741558e-05, "loss": 0.0713, "step": 24520 }, { "grad_norm": 0.9954394102096558, "learning_rate": 9.856413259264025e-05, "loss": 0.0692, "step": 24530 }, { "grad_norm": 0.501248300075531, "learning_rate": 9.856216489013981e-05, "loss": 0.0649, "step": 24540 }, { "grad_norm": 0.8241854310035706, "learning_rate": 9.856019585996802e-05, "loss": 0.0712, "step": 24550 }, { "grad_norm": 0.5129530429840088, "learning_rate": 9.855822550217871e-05, "loss": 0.0638, "step": 24560 }, { "grad_norm": 0.6029219627380371, "learning_rate": 9.855625381682576e-05, "loss": 0.0782, "step": 24570 }, { "grad_norm": 0.6024878025054932, "learning_rate": 9.855428080396307e-05, "loss": 0.059, "step": 24580 }, { "grad_norm": 0.6714084148406982, "learning_rate": 9.85523064636446e-05, "loss": 0.0583, "step": 24590 }, { "grad_norm": 0.5772313475608826, "learning_rate": 9.85503307959243e-05, "loss": 0.0621, "step": 24600 }, { "grad_norm": 0.7054482102394104, "learning_rate": 9.854835380085619e-05, "loss": 0.0652, "step": 24610 }, { "grad_norm": 0.6638669371604919, "learning_rate": 9.854637547849434e-05, "loss": 0.0602, "step": 24620 }, { "grad_norm": 0.42519068717956543, "learning_rate": 9.854439582889281e-05, "loss": 0.0698, "step": 24630 }, { "grad_norm": 0.5886374711990356, "learning_rate": 9.854241485210573e-05, "loss": 0.0714, "step": 24640 }, { "grad_norm": 0.5609176754951477, "learning_rate": 9.854043254818728e-05, "loss": 0.0632, "step": 24650 }, { "grad_norm": 0.6970614790916443, "learning_rate": 9.853844891719163e-05, "loss": 0.0704, "step": 24660 }, { "grad_norm": 0.7265905141830444, "learning_rate": 9.853646395917302e-05, "loss": 0.0715, "step": 24670 }, { "grad_norm": 0.6347742676734924, "learning_rate": 9.853447767418573e-05, "loss": 0.0788, "step": 24680 }, { "grad_norm": 0.5445806384086609, "learning_rate": 9.853249006228405e-05, "loss": 0.0757, "step": 24690 }, { "grad_norm": 0.43418577313423157, "learning_rate": 9.853050112352231e-05, "loss": 0.0647, "step": 24700 }, { "grad_norm": 0.6576390862464905, "learning_rate": 9.852851085795493e-05, "loss": 0.0918, "step": 24710 }, { "grad_norm": 0.5500969290733337, "learning_rate": 9.852651926563628e-05, "loss": 0.0557, "step": 24720 }, { "grad_norm": 0.5264521837234497, "learning_rate": 9.852452634662081e-05, "loss": 0.0588, "step": 24730 }, { "grad_norm": 0.5328694581985474, "learning_rate": 9.852253210096304e-05, "loss": 0.0616, "step": 24740 }, { "grad_norm": 0.5070608258247375, "learning_rate": 9.852053652871746e-05, "loss": 0.0547, "step": 24750 }, { "grad_norm": 0.6979307532310486, "learning_rate": 9.851853962993864e-05, "loss": 0.0677, "step": 24760 }, { "grad_norm": 0.5326484441757202, "learning_rate": 9.851654140468118e-05, "loss": 0.0639, "step": 24770 }, { "grad_norm": 0.6280975341796875, "learning_rate": 9.851454185299971e-05, "loss": 0.0727, "step": 24780 }, { "grad_norm": 0.7610827088356018, "learning_rate": 9.851254097494887e-05, "loss": 0.0715, "step": 24790 }, { "grad_norm": 0.5090147852897644, "learning_rate": 9.851053877058341e-05, "loss": 0.0743, "step": 24800 }, { "grad_norm": 0.506554365158081, "learning_rate": 9.850853523995803e-05, "loss": 0.0699, "step": 24810 }, { "grad_norm": 0.6617531180381775, "learning_rate": 9.850653038312752e-05, "loss": 0.0774, "step": 24820 }, { "grad_norm": 0.3548245429992676, "learning_rate": 9.85045242001467e-05, "loss": 0.0571, "step": 24830 }, { "grad_norm": 0.654834508895874, "learning_rate": 9.85025166910704e-05, "loss": 0.0622, "step": 24840 }, { "grad_norm": 0.6361293792724609, "learning_rate": 9.850050785595352e-05, "loss": 0.06, "step": 24850 }, { "grad_norm": 0.8343828320503235, "learning_rate": 9.849849769485098e-05, "loss": 0.0903, "step": 24860 }, { "grad_norm": 0.524261474609375, "learning_rate": 9.849648620781773e-05, "loss": 0.0815, "step": 24870 }, { "grad_norm": 1.0051782131195068, "learning_rate": 9.849447339490877e-05, "loss": 0.0633, "step": 24880 }, { "grad_norm": 0.7950232028961182, "learning_rate": 9.849245925617911e-05, "loss": 0.069, "step": 24890 }, { "grad_norm": 0.5401838421821594, "learning_rate": 9.849044379168386e-05, "loss": 0.0586, "step": 24900 }, { "grad_norm": 0.4877359867095947, "learning_rate": 9.848842700147808e-05, "loss": 0.065, "step": 24910 }, { "grad_norm": 0.4868280589580536, "learning_rate": 9.848640888561691e-05, "loss": 0.0946, "step": 24920 }, { "grad_norm": 0.669326901435852, "learning_rate": 9.848438944415555e-05, "loss": 0.0813, "step": 24930 }, { "grad_norm": 0.41848284006118774, "learning_rate": 9.848236867714919e-05, "loss": 0.0629, "step": 24940 }, { "grad_norm": 0.6193104982376099, "learning_rate": 9.84803465846531e-05, "loss": 0.06, "step": 24950 }, { "grad_norm": 0.6330947279930115, "learning_rate": 9.847832316672254e-05, "loss": 0.0521, "step": 24960 }, { "grad_norm": 0.6016795039176941, "learning_rate": 9.847629842341282e-05, "loss": 0.0683, "step": 24970 }, { "grad_norm": 0.49457088112831116, "learning_rate": 9.847427235477933e-05, "loss": 0.0647, "step": 24980 }, { "grad_norm": 0.45297971367836, "learning_rate": 9.847224496087744e-05, "loss": 0.0823, "step": 24990 }, { "grad_norm": 0.5270340442657471, "learning_rate": 9.84702162417626e-05, "loss": 0.0591, "step": 25000 }, { "grad_norm": 0.7884156107902527, "learning_rate": 9.846818619749023e-05, "loss": 0.0588, "step": 25010 }, { "grad_norm": 0.7980835437774658, "learning_rate": 9.846615482811588e-05, "loss": 0.095, "step": 25020 }, { "grad_norm": 0.46511051058769226, "learning_rate": 9.846412213369504e-05, "loss": 0.0719, "step": 25030 }, { "grad_norm": 0.44193077087402344, "learning_rate": 9.846208811428332e-05, "loss": 0.0593, "step": 25040 }, { "grad_norm": 0.7146030068397522, "learning_rate": 9.846005276993633e-05, "loss": 0.0736, "step": 25050 }, { "grad_norm": 0.7115620374679565, "learning_rate": 9.845801610070969e-05, "loss": 0.0611, "step": 25060 }, { "grad_norm": 0.45223262906074524, "learning_rate": 9.84559781066591e-05, "loss": 0.0639, "step": 25070 }, { "grad_norm": 0.4760929346084595, "learning_rate": 9.845393878784026e-05, "loss": 0.0759, "step": 25080 }, { "grad_norm": 1.2446197271347046, "learning_rate": 9.845189814430895e-05, "loss": 0.0833, "step": 25090 }, { "grad_norm": 0.5716724991798401, "learning_rate": 9.844985617612095e-05, "loss": 0.0617, "step": 25100 }, { "grad_norm": 0.6547936797142029, "learning_rate": 9.844781288333207e-05, "loss": 0.0664, "step": 25110 }, { "grad_norm": 0.7294254302978516, "learning_rate": 9.84457682659982e-05, "loss": 0.0598, "step": 25120 }, { "grad_norm": 0.4895994961261749, "learning_rate": 9.844372232417522e-05, "loss": 0.0696, "step": 25130 }, { "grad_norm": 0.7128457427024841, "learning_rate": 9.844167505791907e-05, "loss": 0.0756, "step": 25140 }, { "grad_norm": 0.627737820148468, "learning_rate": 9.843962646728573e-05, "loss": 0.0716, "step": 25150 }, { "grad_norm": 0.6484552621841431, "learning_rate": 9.843757655233119e-05, "loss": 0.0707, "step": 25160 }, { "grad_norm": 0.606374204158783, "learning_rate": 9.843552531311151e-05, "loss": 0.0864, "step": 25170 }, { "grad_norm": 0.5407175421714783, "learning_rate": 9.843347274968275e-05, "loss": 0.0563, "step": 25180 }, { "grad_norm": 0.6796756982803345, "learning_rate": 9.843141886210107e-05, "loss": 0.0642, "step": 25190 }, { "grad_norm": 0.6843665242195129, "learning_rate": 9.842936365042256e-05, "loss": 0.0704, "step": 25200 }, { "grad_norm": 0.6006065607070923, "learning_rate": 9.842730711470347e-05, "loss": 0.0598, "step": 25210 }, { "grad_norm": 0.5541993379592896, "learning_rate": 9.842524925499999e-05, "loss": 0.0552, "step": 25220 }, { "grad_norm": 0.5808992385864258, "learning_rate": 9.842319007136838e-05, "loss": 0.0583, "step": 25230 }, { "grad_norm": 0.5919936299324036, "learning_rate": 9.842112956386494e-05, "loss": 0.0621, "step": 25240 }, { "grad_norm": 0.5720313787460327, "learning_rate": 9.841906773254602e-05, "loss": 0.0543, "step": 25250 }, { "grad_norm": 0.7334209084510803, "learning_rate": 9.841700457746799e-05, "loss": 0.0593, "step": 25260 }, { "grad_norm": 0.6753699779510498, "learning_rate": 9.841494009868722e-05, "loss": 0.0783, "step": 25270 }, { "grad_norm": 0.894197404384613, "learning_rate": 9.841287429626018e-05, "loss": 0.0894, "step": 25280 }, { "grad_norm": 0.7000298500061035, "learning_rate": 9.841080717024335e-05, "loss": 0.0629, "step": 25290 }, { "grad_norm": 0.7251766920089722, "learning_rate": 9.840873872069326e-05, "loss": 0.0693, "step": 25300 }, { "grad_norm": 0.3905307650566101, "learning_rate": 9.840666894766642e-05, "loss": 0.0562, "step": 25310 }, { "grad_norm": 0.7509567737579346, "learning_rate": 9.840459785121943e-05, "loss": 0.1028, "step": 25320 }, { "grad_norm": 0.46580299735069275, "learning_rate": 9.840252543140891e-05, "loss": 0.0677, "step": 25330 }, { "grad_norm": 0.6407303214073181, "learning_rate": 9.840045168829155e-05, "loss": 0.0675, "step": 25340 }, { "grad_norm": 0.581195592880249, "learning_rate": 9.839837662192401e-05, "loss": 0.0628, "step": 25350 }, { "grad_norm": 0.47064247727394104, "learning_rate": 9.839630023236306e-05, "loss": 0.0845, "step": 25360 }, { "grad_norm": 0.5329262614250183, "learning_rate": 9.839422251966541e-05, "loss": 0.068, "step": 25370 }, { "grad_norm": 0.5560061931610107, "learning_rate": 9.839214348388793e-05, "loss": 0.0672, "step": 25380 }, { "grad_norm": 0.5878323912620544, "learning_rate": 9.83900631250874e-05, "loss": 0.073, "step": 25390 }, { "grad_norm": 0.6451268792152405, "learning_rate": 9.838798144332074e-05, "loss": 0.0655, "step": 25400 }, { "grad_norm": 0.6696105599403381, "learning_rate": 9.838589843864484e-05, "loss": 0.071, "step": 25410 }, { "grad_norm": 0.5111460089683533, "learning_rate": 9.838381411111665e-05, "loss": 0.0619, "step": 25420 }, { "grad_norm": 0.6823230385780334, "learning_rate": 9.838172846079317e-05, "loss": 0.0741, "step": 25430 }, { "grad_norm": 0.829815685749054, "learning_rate": 9.83796414877314e-05, "loss": 0.06, "step": 25440 }, { "grad_norm": 0.5862039923667908, "learning_rate": 9.837755319198843e-05, "loss": 0.0558, "step": 25450 }, { "grad_norm": 0.6925121545791626, "learning_rate": 9.837546357362131e-05, "loss": 0.0689, "step": 25460 }, { "grad_norm": 0.6064973473548889, "learning_rate": 9.837337263268717e-05, "loss": 0.0688, "step": 25470 }, { "grad_norm": 0.6170687079429626, "learning_rate": 9.837128036924324e-05, "loss": 0.0699, "step": 25480 }, { "grad_norm": 1.5742545127868652, "learning_rate": 9.836918678334665e-05, "loss": 0.0837, "step": 25490 }, { "grad_norm": 1.8484083414077759, "learning_rate": 9.836709187505466e-05, "loss": 0.0709, "step": 25500 }, { "grad_norm": 0.703930675983429, "learning_rate": 9.836499564442457e-05, "loss": 0.0825, "step": 25510 }, { "grad_norm": 0.7402658462524414, "learning_rate": 9.836289809151365e-05, "loss": 0.0829, "step": 25520 }, { "grad_norm": 0.905866801738739, "learning_rate": 9.836079921637926e-05, "loss": 0.0983, "step": 25530 }, { "grad_norm": 0.6078210473060608, "learning_rate": 9.835869901907881e-05, "loss": 0.0803, "step": 25540 }, { "grad_norm": 1.094070553779602, "learning_rate": 9.835659749966967e-05, "loss": 0.0806, "step": 25550 }, { "grad_norm": 0.6547967791557312, "learning_rate": 9.835449465820935e-05, "loss": 0.0763, "step": 25560 }, { "grad_norm": 0.6703817248344421, "learning_rate": 9.83523904947553e-05, "loss": 0.0776, "step": 25570 }, { "grad_norm": 0.6889826655387878, "learning_rate": 9.835028500936505e-05, "loss": 0.0597, "step": 25580 }, { "grad_norm": 0.7232069373130798, "learning_rate": 9.834817820209619e-05, "loss": 0.0665, "step": 25590 }, { "grad_norm": 0.6482262015342712, "learning_rate": 9.834607007300629e-05, "loss": 0.0803, "step": 25600 }, { "grad_norm": 0.4890271723270416, "learning_rate": 9.8343960622153e-05, "loss": 0.0654, "step": 25610 }, { "grad_norm": 0.7151733040809631, "learning_rate": 9.834184984959398e-05, "loss": 0.0752, "step": 25620 }, { "grad_norm": 0.5510543584823608, "learning_rate": 9.833973775538697e-05, "loss": 0.076, "step": 25630 }, { "grad_norm": 0.6175277233123779, "learning_rate": 9.833762433958966e-05, "loss": 0.0725, "step": 25640 }, { "grad_norm": 0.6881520748138428, "learning_rate": 9.833550960225989e-05, "loss": 0.065, "step": 25650 }, { "grad_norm": 0.7378579378128052, "learning_rate": 9.833339354345541e-05, "loss": 0.0767, "step": 25660 }, { "grad_norm": 0.4895937740802765, "learning_rate": 9.833127616323413e-05, "loss": 0.0619, "step": 25670 }, { "grad_norm": 0.5132817029953003, "learning_rate": 9.832915746165392e-05, "loss": 0.0635, "step": 25680 }, { "grad_norm": 0.5938856601715088, "learning_rate": 9.83270374387727e-05, "loss": 0.0784, "step": 25690 }, { "grad_norm": 0.6020486354827881, "learning_rate": 9.832491609464843e-05, "loss": 0.0595, "step": 25700 }, { "grad_norm": 0.7867736220359802, "learning_rate": 9.83227934293391e-05, "loss": 0.0896, "step": 25710 }, { "grad_norm": 0.5145056843757629, "learning_rate": 9.832066944290277e-05, "loss": 0.0703, "step": 25720 }, { "grad_norm": 0.3839071989059448, "learning_rate": 9.831854413539747e-05, "loss": 0.0573, "step": 25730 }, { "grad_norm": 0.9172230362892151, "learning_rate": 9.831641750688135e-05, "loss": 0.0649, "step": 25740 }, { "grad_norm": 0.44743606448173523, "learning_rate": 9.83142895574125e-05, "loss": 0.0636, "step": 25750 }, { "grad_norm": 0.850589394569397, "learning_rate": 9.831216028704914e-05, "loss": 0.0624, "step": 25760 }, { "grad_norm": 0.6665005087852478, "learning_rate": 9.831002969584946e-05, "loss": 0.0661, "step": 25770 }, { "grad_norm": 0.7964568138122559, "learning_rate": 9.830789778387173e-05, "loss": 0.0824, "step": 25780 }, { "grad_norm": 0.714568555355072, "learning_rate": 9.830576455117422e-05, "loss": 0.0817, "step": 25790 }, { "grad_norm": 0.6462125182151794, "learning_rate": 9.830362999781524e-05, "loss": 0.0771, "step": 25800 }, { "grad_norm": 0.7728694677352905, "learning_rate": 9.830149412385317e-05, "loss": 0.0719, "step": 25810 }, { "grad_norm": 0.8802642226219177, "learning_rate": 9.829935692934639e-05, "loss": 0.0574, "step": 25820 }, { "grad_norm": 0.5136582255363464, "learning_rate": 9.829721841435336e-05, "loss": 0.055, "step": 25830 }, { "grad_norm": 0.6986721158027649, "learning_rate": 9.829507857893251e-05, "loss": 0.0549, "step": 25840 }, { "grad_norm": 0.8222113847732544, "learning_rate": 9.829293742314237e-05, "loss": 0.0565, "step": 25850 }, { "grad_norm": 0.4883970618247986, "learning_rate": 9.829079494704143e-05, "loss": 0.068, "step": 25860 }, { "grad_norm": 0.4850115478038788, "learning_rate": 9.828865115068833e-05, "loss": 0.0494, "step": 25870 }, { "grad_norm": 0.5193180441856384, "learning_rate": 9.828650603414162e-05, "loss": 0.0701, "step": 25880 }, { "grad_norm": 0.8302727341651917, "learning_rate": 9.828435959746e-05, "loss": 0.0668, "step": 25890 }, { "grad_norm": 0.4992939829826355, "learning_rate": 9.828221184070212e-05, "loss": 0.0524, "step": 25900 }, { "grad_norm": 2.4851412773132324, "learning_rate": 9.828006276392669e-05, "loss": 0.0833, "step": 25910 }, { "grad_norm": 0.8710256218910217, "learning_rate": 9.82779123671925e-05, "loss": 0.0627, "step": 25920 }, { "grad_norm": 0.6536652445793152, "learning_rate": 9.827576065055832e-05, "loss": 0.0614, "step": 25930 }, { "grad_norm": 0.7338285446166992, "learning_rate": 9.8273607614083e-05, "loss": 0.0581, "step": 25940 }, { "grad_norm": 0.5447090864181519, "learning_rate": 9.827145325782535e-05, "loss": 0.0597, "step": 25950 }, { "grad_norm": 0.5340520143508911, "learning_rate": 9.826929758184432e-05, "loss": 0.0581, "step": 25960 }, { "grad_norm": 0.7173455357551575, "learning_rate": 9.826714058619882e-05, "loss": 0.077, "step": 25970 }, { "grad_norm": 0.5671654939651489, "learning_rate": 9.826498227094784e-05, "loss": 0.0616, "step": 25980 }, { "grad_norm": 0.6803837418556213, "learning_rate": 9.826282263615036e-05, "loss": 0.0661, "step": 25990 }, { "grad_norm": 0.4307873249053955, "learning_rate": 9.826066168186547e-05, "loss": 0.0535, "step": 26000 }, { "grad_norm": 0.8116239309310913, "learning_rate": 9.82584994081522e-05, "loss": 0.0742, "step": 26010 }, { "grad_norm": 0.7441844940185547, "learning_rate": 9.825633581506968e-05, "loss": 0.0977, "step": 26020 }, { "grad_norm": 0.8213526010513306, "learning_rate": 9.825417090267707e-05, "loss": 0.0771, "step": 26030 }, { "grad_norm": 0.8771155476570129, "learning_rate": 9.825200467103357e-05, "loss": 0.0765, "step": 26040 }, { "grad_norm": 0.7774634957313538, "learning_rate": 9.824983712019838e-05, "loss": 0.0949, "step": 26050 }, { "grad_norm": 0.6347730159759521, "learning_rate": 9.824766825023077e-05, "loss": 0.0758, "step": 26060 }, { "grad_norm": 0.7779666781425476, "learning_rate": 9.824549806119005e-05, "loss": 0.0623, "step": 26070 }, { "grad_norm": 0.8380223512649536, "learning_rate": 9.824332655313553e-05, "loss": 0.0796, "step": 26080 }, { "grad_norm": 1.284346580505371, "learning_rate": 9.824115372612657e-05, "loss": 0.0707, "step": 26090 }, { "grad_norm": 0.5603412389755249, "learning_rate": 9.82389795802226e-05, "loss": 0.0575, "step": 26100 }, { "grad_norm": 0.5380796790122986, "learning_rate": 9.823680411548306e-05, "loss": 0.0615, "step": 26110 }, { "grad_norm": 0.628271222114563, "learning_rate": 9.823462733196741e-05, "loss": 0.0692, "step": 26120 }, { "grad_norm": 0.8022698163986206, "learning_rate": 9.823244922973518e-05, "loss": 0.0695, "step": 26130 }, { "grad_norm": 0.47927457094192505, "learning_rate": 9.823026980884589e-05, "loss": 0.0965, "step": 26140 }, { "grad_norm": 0.762226939201355, "learning_rate": 9.822808906935914e-05, "loss": 0.0741, "step": 26150 }, { "grad_norm": 0.9501785635948181, "learning_rate": 9.822590701133457e-05, "loss": 0.0648, "step": 26160 }, { "grad_norm": 0.6222162246704102, "learning_rate": 9.822372363483181e-05, "loss": 0.0529, "step": 26170 }, { "grad_norm": 0.782080352306366, "learning_rate": 9.822153893991057e-05, "loss": 0.0817, "step": 26180 }, { "grad_norm": 0.63153076171875, "learning_rate": 9.821935292663056e-05, "loss": 0.0671, "step": 26190 }, { "grad_norm": 0.8406867384910583, "learning_rate": 9.821716559505157e-05, "loss": 0.0695, "step": 26200 }, { "grad_norm": 0.8101731538772583, "learning_rate": 9.821497694523339e-05, "loss": 0.071, "step": 26210 }, { "grad_norm": 0.6517077684402466, "learning_rate": 9.821278697723583e-05, "loss": 0.0702, "step": 26220 }, { "grad_norm": 0.9574554562568665, "learning_rate": 9.821059569111881e-05, "loss": 0.071, "step": 26230 }, { "grad_norm": 0.8279696106910706, "learning_rate": 9.820840308694221e-05, "loss": 0.0568, "step": 26240 }, { "grad_norm": 0.78078293800354, "learning_rate": 9.820620916476597e-05, "loss": 0.0761, "step": 26250 }, { "grad_norm": 0.48491808772087097, "learning_rate": 9.820401392465011e-05, "loss": 0.0712, "step": 26260 }, { "grad_norm": 1.038433313369751, "learning_rate": 9.820181736665459e-05, "loss": 0.0792, "step": 26270 }, { "grad_norm": 0.6707736253738403, "learning_rate": 9.81996194908395e-05, "loss": 0.0741, "step": 26280 }, { "grad_norm": 0.71439528465271, "learning_rate": 9.819742029726492e-05, "loss": 0.0688, "step": 26290 }, { "grad_norm": 0.655556321144104, "learning_rate": 9.819521978599098e-05, "loss": 0.0663, "step": 26300 }, { "grad_norm": 0.5656362771987915, "learning_rate": 9.819301795707783e-05, "loss": 0.0538, "step": 26310 }, { "grad_norm": 0.5198747515678406, "learning_rate": 9.819081481058567e-05, "loss": 0.0461, "step": 26320 }, { "grad_norm": 0.7177080512046814, "learning_rate": 9.818861034657476e-05, "loss": 0.0682, "step": 26330 }, { "grad_norm": 0.640486478805542, "learning_rate": 9.818640456510533e-05, "loss": 0.0541, "step": 26340 }, { "grad_norm": 0.8291186094284058, "learning_rate": 9.81841974662377e-05, "loss": 0.0765, "step": 26350 }, { "grad_norm": 0.609794020652771, "learning_rate": 9.818198905003222e-05, "loss": 0.0579, "step": 26360 }, { "grad_norm": 0.4941074550151825, "learning_rate": 9.817977931654926e-05, "loss": 0.0557, "step": 26370 }, { "grad_norm": 0.5915853381156921, "learning_rate": 9.817756826584923e-05, "loss": 0.0614, "step": 26380 }, { "grad_norm": 0.6251067519187927, "learning_rate": 9.817535589799258e-05, "loss": 0.0568, "step": 26390 }, { "grad_norm": 0.7035349607467651, "learning_rate": 9.817314221303982e-05, "loss": 0.0781, "step": 26400 }, { "grad_norm": 0.5064131617546082, "learning_rate": 9.817092721105143e-05, "loss": 0.052, "step": 26410 }, { "grad_norm": 0.527330219745636, "learning_rate": 9.8168710892088e-05, "loss": 0.0482, "step": 26420 }, { "grad_norm": 0.6512566804885864, "learning_rate": 9.81664932562101e-05, "loss": 0.0706, "step": 26430 }, { "grad_norm": 0.56855708360672, "learning_rate": 9.816427430347836e-05, "loss": 0.0682, "step": 26440 }, { "grad_norm": 0.80031418800354, "learning_rate": 9.816205403395347e-05, "loss": 0.0564, "step": 26450 }, { "grad_norm": 0.7984063625335693, "learning_rate": 9.815983244769611e-05, "loss": 0.0701, "step": 26460 }, { "grad_norm": 0.725740373134613, "learning_rate": 9.815760954476703e-05, "loss": 0.0917, "step": 26470 }, { "grad_norm": 1.006803274154663, "learning_rate": 9.815538532522699e-05, "loss": 0.0831, "step": 26480 }, { "grad_norm": 0.6697885990142822, "learning_rate": 9.81531597891368e-05, "loss": 0.0677, "step": 26490 }, { "grad_norm": 0.682009756565094, "learning_rate": 9.815093293655733e-05, "loss": 0.0568, "step": 26500 }, { "grad_norm": 0.543407678604126, "learning_rate": 9.814870476754942e-05, "loss": 0.0646, "step": 26510 }, { "grad_norm": 0.5830398201942444, "learning_rate": 9.814647528217403e-05, "loss": 0.0731, "step": 26520 }, { "grad_norm": 0.7545248866081238, "learning_rate": 9.814424448049209e-05, "loss": 0.0803, "step": 26530 }, { "grad_norm": 0.8270124197006226, "learning_rate": 9.81420123625646e-05, "loss": 0.0681, "step": 26540 }, { "grad_norm": 0.5926297307014465, "learning_rate": 9.813977892845256e-05, "loss": 0.0706, "step": 26550 }, { "grad_norm": 0.5436320900917053, "learning_rate": 9.813754417821705e-05, "loss": 0.0702, "step": 26560 }, { "grad_norm": 0.5503366589546204, "learning_rate": 9.813530811191918e-05, "loss": 0.0495, "step": 26570 }, { "grad_norm": 0.8661223649978638, "learning_rate": 9.813307072962007e-05, "loss": 0.0783, "step": 26580 }, { "grad_norm": 0.6451756358146667, "learning_rate": 9.813083203138088e-05, "loss": 0.0631, "step": 26590 }, { "grad_norm": 0.5500257611274719, "learning_rate": 9.812859201726283e-05, "loss": 0.0829, "step": 26600 }, { "grad_norm": 0.8076847791671753, "learning_rate": 9.812635068732715e-05, "loss": 0.0677, "step": 26610 }, { "grad_norm": 0.47732922434806824, "learning_rate": 9.812410804163513e-05, "loss": 0.0747, "step": 26620 }, { "grad_norm": 0.623810350894928, "learning_rate": 9.812186408024806e-05, "loss": 0.077, "step": 26630 }, { "grad_norm": 0.8501935601234436, "learning_rate": 9.811961880322731e-05, "loss": 0.0911, "step": 26640 }, { "grad_norm": 0.609786868095398, "learning_rate": 9.811737221063429e-05, "loss": 0.0651, "step": 26650 }, { "grad_norm": 0.5386472940444946, "learning_rate": 9.811512430253036e-05, "loss": 0.0622, "step": 26660 }, { "grad_norm": 0.6340077519416809, "learning_rate": 9.811287507897701e-05, "loss": 0.0706, "step": 26670 }, { "grad_norm": 0.7504527568817139, "learning_rate": 9.811062454003572e-05, "loss": 0.0618, "step": 26680 }, { "grad_norm": 0.6821426153182983, "learning_rate": 9.810837268576805e-05, "loss": 0.0596, "step": 26690 }, { "grad_norm": 0.45112305879592896, "learning_rate": 9.810611951623554e-05, "loss": 0.056, "step": 26700 }, { "grad_norm": 0.549052357673645, "learning_rate": 9.810386503149979e-05, "loss": 0.0605, "step": 26710 }, { "grad_norm": 0.5707014203071594, "learning_rate": 9.810160923162245e-05, "loss": 0.0583, "step": 26720 }, { "grad_norm": 0.5847043991088867, "learning_rate": 9.809935211666518e-05, "loss": 0.0709, "step": 26730 }, { "grad_norm": 0.7067632675170898, "learning_rate": 9.809709368668969e-05, "loss": 0.0825, "step": 26740 }, { "grad_norm": 0.6148846745491028, "learning_rate": 9.809483394175773e-05, "loss": 0.0613, "step": 26750 }, { "grad_norm": 0.5450513958930969, "learning_rate": 9.809257288193107e-05, "loss": 0.0796, "step": 26760 }, { "grad_norm": 0.6110957860946655, "learning_rate": 9.809031050727155e-05, "loss": 0.069, "step": 26770 }, { "grad_norm": 0.676399290561676, "learning_rate": 9.8088046817841e-05, "loss": 0.0703, "step": 26780 }, { "grad_norm": 0.9481542706489563, "learning_rate": 9.808578181370131e-05, "loss": 0.0824, "step": 26790 }, { "grad_norm": 0.8340607285499573, "learning_rate": 9.808351549491443e-05, "loss": 0.0783, "step": 26800 }, { "grad_norm": 0.6085879802703857, "learning_rate": 9.808124786154228e-05, "loss": 0.0747, "step": 26810 }, { "grad_norm": 0.5861801505088806, "learning_rate": 9.807897891364688e-05, "loss": 0.0638, "step": 26820 }, { "grad_norm": 0.4359954595565796, "learning_rate": 9.807670865129026e-05, "loss": 0.0715, "step": 26830 }, { "grad_norm": 0.4732283055782318, "learning_rate": 9.80744370745345e-05, "loss": 0.0588, "step": 26840 }, { "grad_norm": 0.45384910702705383, "learning_rate": 9.807216418344167e-05, "loss": 0.056, "step": 26850 }, { "grad_norm": 0.7171075344085693, "learning_rate": 9.806988997807396e-05, "loss": 0.066, "step": 26860 }, { "grad_norm": 0.4267798066139221, "learning_rate": 9.806761445849348e-05, "loss": 0.0664, "step": 26870 }, { "grad_norm": 0.8302441835403442, "learning_rate": 9.806533762476252e-05, "loss": 0.0737, "step": 26880 }, { "grad_norm": 0.5992254614830017, "learning_rate": 9.806305947694325e-05, "loss": 0.0662, "step": 26890 }, { "grad_norm": 1.3917940855026245, "learning_rate": 9.806078001509801e-05, "loss": 0.0979, "step": 26900 }, { "grad_norm": 0.5952434539794922, "learning_rate": 9.80584992392891e-05, "loss": 0.0896, "step": 26910 }, { "grad_norm": 0.6734265685081482, "learning_rate": 9.805621714957886e-05, "loss": 0.0693, "step": 26920 }, { "grad_norm": 0.6512656211853027, "learning_rate": 9.805393374602972e-05, "loss": 0.0882, "step": 26930 }, { "grad_norm": 0.5492473840713501, "learning_rate": 9.805164902870406e-05, "loss": 0.0745, "step": 26940 }, { "grad_norm": 0.7878341674804688, "learning_rate": 9.804936299766438e-05, "loss": 0.0767, "step": 26950 }, { "grad_norm": 0.5461976528167725, "learning_rate": 9.804707565297318e-05, "loss": 0.0836, "step": 26960 }, { "grad_norm": 0.7069393396377563, "learning_rate": 9.804478699469296e-05, "loss": 0.0818, "step": 26970 }, { "grad_norm": 0.5291988253593445, "learning_rate": 9.804249702288633e-05, "loss": 0.0596, "step": 26980 }, { "grad_norm": 0.6200866103172302, "learning_rate": 9.804020573761587e-05, "loss": 0.0514, "step": 26990 }, { "grad_norm": 0.830742597579956, "learning_rate": 9.803791313894423e-05, "loss": 0.067, "step": 27000 }, { "grad_norm": 1.0527291297912598, "learning_rate": 9.80356192269341e-05, "loss": 0.0804, "step": 27010 }, { "grad_norm": 0.6889663338661194, "learning_rate": 9.803332400164817e-05, "loss": 0.0628, "step": 27020 }, { "grad_norm": 0.6692953705787659, "learning_rate": 9.803102746314922e-05, "loss": 0.0769, "step": 27030 }, { "grad_norm": 0.6279940605163574, "learning_rate": 9.802872961150003e-05, "loss": 0.0698, "step": 27040 }, { "grad_norm": 1.457688570022583, "learning_rate": 9.802643044676341e-05, "loss": 0.0873, "step": 27050 }, { "grad_norm": 0.7249099016189575, "learning_rate": 9.802412996900223e-05, "loss": 0.0677, "step": 27060 }, { "grad_norm": 0.6681095361709595, "learning_rate": 9.802182817827936e-05, "loss": 0.0743, "step": 27070 }, { "grad_norm": 0.7110199332237244, "learning_rate": 9.801952507465777e-05, "loss": 0.0536, "step": 27080 }, { "grad_norm": 0.5387560725212097, "learning_rate": 9.80172206582004e-05, "loss": 0.0592, "step": 27090 }, { "grad_norm": 0.47336530685424805, "learning_rate": 9.801491492897025e-05, "loss": 0.0619, "step": 27100 }, { "grad_norm": 0.4668413996696472, "learning_rate": 9.801260788703037e-05, "loss": 0.0688, "step": 27110 }, { "grad_norm": 0.6619738936424255, "learning_rate": 9.801029953244383e-05, "loss": 0.0722, "step": 27120 }, { "grad_norm": 0.3586488664150238, "learning_rate": 9.800798986527372e-05, "loss": 0.0564, "step": 27130 }, { "grad_norm": 0.4850705564022064, "learning_rate": 9.800567888558321e-05, "loss": 0.0489, "step": 27140 }, { "grad_norm": 0.758885383605957, "learning_rate": 9.800336659343549e-05, "loss": 0.0616, "step": 27150 }, { "grad_norm": 0.6216140389442444, "learning_rate": 9.800105298889375e-05, "loss": 0.0789, "step": 27160 }, { "grad_norm": 0.8103584051132202, "learning_rate": 9.799873807202124e-05, "loss": 0.0823, "step": 27170 }, { "grad_norm": 0.5489770770072937, "learning_rate": 9.799642184288129e-05, "loss": 0.0549, "step": 27180 }, { "grad_norm": 0.7181708812713623, "learning_rate": 9.799410430153719e-05, "loss": 0.105, "step": 27190 }, { "grad_norm": 0.588316023349762, "learning_rate": 9.799178544805229e-05, "loss": 0.0735, "step": 27200 }, { "grad_norm": 0.8649327754974365, "learning_rate": 9.798946528249003e-05, "loss": 0.0626, "step": 27210 }, { "grad_norm": 0.6690341234207153, "learning_rate": 9.79871438049138e-05, "loss": 0.0699, "step": 27220 }, { "grad_norm": 0.6309646964073181, "learning_rate": 9.798482101538711e-05, "loss": 0.0585, "step": 27230 }, { "grad_norm": 1.269525170326233, "learning_rate": 9.798249691397343e-05, "loss": 0.094, "step": 27240 }, { "grad_norm": 0.6586159467697144, "learning_rate": 9.79801715007363e-05, "loss": 0.0799, "step": 27250 }, { "grad_norm": 0.5516170859336853, "learning_rate": 9.79778447757393e-05, "loss": 0.0668, "step": 27260 }, { "grad_norm": 0.8529228568077087, "learning_rate": 9.797551673904607e-05, "loss": 0.0725, "step": 27270 }, { "grad_norm": 0.7052112221717834, "learning_rate": 9.797318739072023e-05, "loss": 0.0817, "step": 27280 }, { "grad_norm": 0.6021967530250549, "learning_rate": 9.797085673082547e-05, "loss": 0.0654, "step": 27290 }, { "grad_norm": 0.8305314183235168, "learning_rate": 9.79685247594255e-05, "loss": 0.0778, "step": 27300 }, { "grad_norm": 1.3224678039550781, "learning_rate": 9.796619147658408e-05, "loss": 0.079, "step": 27310 }, { "grad_norm": 0.8748931288719177, "learning_rate": 9.796385688236503e-05, "loss": 0.0716, "step": 27320 }, { "grad_norm": 0.44729480147361755, "learning_rate": 9.796152097683213e-05, "loss": 0.0606, "step": 27330 }, { "grad_norm": 0.8657302260398865, "learning_rate": 9.795918376004927e-05, "loss": 0.0726, "step": 27340 }, { "grad_norm": 0.6071323752403259, "learning_rate": 9.795684523208034e-05, "loss": 0.0667, "step": 27350 }, { "grad_norm": 0.4409150183200836, "learning_rate": 9.795450539298928e-05, "loss": 0.0581, "step": 27360 }, { "grad_norm": 0.4853850305080414, "learning_rate": 9.795216424284006e-05, "loss": 0.0559, "step": 27370 }, { "grad_norm": 0.8981728553771973, "learning_rate": 9.79498217816967e-05, "loss": 0.079, "step": 27380 }, { "grad_norm": 0.682630717754364, "learning_rate": 9.79474780096232e-05, "loss": 0.0613, "step": 27390 }, { "grad_norm": 0.853415846824646, "learning_rate": 9.794513292668368e-05, "loss": 0.0674, "step": 27400 }, { "grad_norm": 0.8687750101089478, "learning_rate": 9.794278653294222e-05, "loss": 0.0644, "step": 27410 }, { "grad_norm": 0.7434802055358887, "learning_rate": 9.794043882846302e-05, "loss": 0.0676, "step": 27420 }, { "grad_norm": 0.44696423411369324, "learning_rate": 9.793808981331021e-05, "loss": 0.0574, "step": 27430 }, { "grad_norm": 0.6519970893859863, "learning_rate": 9.793573948754802e-05, "loss": 0.0576, "step": 27440 }, { "grad_norm": 0.5521117448806763, "learning_rate": 9.793338785124076e-05, "loss": 0.0531, "step": 27450 }, { "grad_norm": 0.553585946559906, "learning_rate": 9.793103490445265e-05, "loss": 0.0523, "step": 27460 }, { "grad_norm": 0.7996609210968018, "learning_rate": 9.792868064724808e-05, "loss": 0.0556, "step": 27470 }, { "grad_norm": 0.5364789962768555, "learning_rate": 9.792632507969135e-05, "loss": 0.057, "step": 27480 }, { "grad_norm": 0.5274121761322021, "learning_rate": 9.792396820184693e-05, "loss": 0.0623, "step": 27490 }, { "grad_norm": 0.5079663991928101, "learning_rate": 9.792161001377921e-05, "loss": 0.0717, "step": 27500 }, { "grad_norm": 0.7173026204109192, "learning_rate": 9.791925051555269e-05, "loss": 0.0739, "step": 27510 }, { "grad_norm": 0.7608450651168823, "learning_rate": 9.791688970723184e-05, "loss": 0.1061, "step": 27520 }, { "grad_norm": 0.6757674217224121, "learning_rate": 9.791452758888125e-05, "loss": 0.0633, "step": 27530 }, { "grad_norm": 0.5977092981338501, "learning_rate": 9.791216416056545e-05, "loss": 0.0636, "step": 27540 }, { "grad_norm": 1.0900638103485107, "learning_rate": 9.79097994223491e-05, "loss": 0.072, "step": 27550 }, { "grad_norm": 0.9501396417617798, "learning_rate": 9.790743337429683e-05, "loss": 0.0582, "step": 27560 }, { "grad_norm": 0.8310208916664124, "learning_rate": 9.790506601647334e-05, "loss": 0.0701, "step": 27570 }, { "grad_norm": 1.0220383405685425, "learning_rate": 9.790269734894333e-05, "loss": 0.0658, "step": 27580 }, { "grad_norm": 0.43096277117729187, "learning_rate": 9.790032737177157e-05, "loss": 0.0574, "step": 27590 }, { "grad_norm": 0.6139349341392517, "learning_rate": 9.789795608502285e-05, "loss": 0.0683, "step": 27600 }, { "grad_norm": 0.8983789682388306, "learning_rate": 9.789558348876201e-05, "loss": 0.078, "step": 27610 }, { "grad_norm": 0.6945322751998901, "learning_rate": 9.789320958305392e-05, "loss": 0.0614, "step": 27620 }, { "grad_norm": 0.9368953108787537, "learning_rate": 9.789083436796347e-05, "loss": 0.0697, "step": 27630 }, { "grad_norm": 0.5841855406761169, "learning_rate": 9.788845784355558e-05, "loss": 0.0812, "step": 27640 }, { "grad_norm": 0.9682000875473022, "learning_rate": 9.788608000989528e-05, "loss": 0.0868, "step": 27650 }, { "grad_norm": 1.2946335077285767, "learning_rate": 9.78837008670475e-05, "loss": 0.0835, "step": 27660 }, { "grad_norm": 0.5537988543510437, "learning_rate": 9.788132041507736e-05, "loss": 0.086, "step": 27670 }, { "grad_norm": 0.561691403388977, "learning_rate": 9.787893865404989e-05, "loss": 0.0642, "step": 27680 }, { "grad_norm": 0.6729341149330139, "learning_rate": 9.787655558403023e-05, "loss": 0.0515, "step": 27690 }, { "grad_norm": 1.2765201330184937, "learning_rate": 9.787417120508354e-05, "loss": 0.0683, "step": 27700 }, { "grad_norm": 0.803502082824707, "learning_rate": 9.787178551727498e-05, "loss": 0.0791, "step": 27710 }, { "grad_norm": 1.064630150794983, "learning_rate": 9.786939852066979e-05, "loss": 0.0632, "step": 27720 }, { "grad_norm": 0.5869125127792358, "learning_rate": 9.786701021533322e-05, "loss": 0.0915, "step": 27730 }, { "grad_norm": 0.8703432083129883, "learning_rate": 9.786462060133059e-05, "loss": 0.0711, "step": 27740 }, { "grad_norm": 0.545014500617981, "learning_rate": 9.78622296787272e-05, "loss": 0.0642, "step": 27750 }, { "grad_norm": 0.4790741205215454, "learning_rate": 9.785983744758845e-05, "loss": 0.0539, "step": 27760 }, { "grad_norm": 0.7940358519554138, "learning_rate": 9.78574439079797e-05, "loss": 0.0628, "step": 27770 }, { "grad_norm": 0.9383813738822937, "learning_rate": 9.785504905996645e-05, "loss": 0.1027, "step": 27780 }, { "grad_norm": 0.852942943572998, "learning_rate": 9.78526529036141e-05, "loss": 0.0582, "step": 27790 }, { "grad_norm": 0.3946857154369354, "learning_rate": 9.785025543898822e-05, "loss": 0.0645, "step": 27800 }, { "grad_norm": 0.8496629595756531, "learning_rate": 9.784785666615433e-05, "loss": 0.0796, "step": 27810 }, { "grad_norm": 0.5833702087402344, "learning_rate": 9.7845456585178e-05, "loss": 0.0598, "step": 27820 }, { "grad_norm": 0.4591907262802124, "learning_rate": 9.784305519612487e-05, "loss": 0.067, "step": 27830 }, { "grad_norm": 0.6422338485717773, "learning_rate": 9.784065249906059e-05, "loss": 0.0681, "step": 27840 }, { "grad_norm": 0.5957201719284058, "learning_rate": 9.783824849405084e-05, "loss": 0.0515, "step": 27850 }, { "grad_norm": 0.6599706411361694, "learning_rate": 9.783584318116133e-05, "loss": 0.0681, "step": 27860 }, { "grad_norm": 0.5322542786598206, "learning_rate": 9.783343656045786e-05, "loss": 0.0811, "step": 27870 }, { "grad_norm": 0.7285740971565247, "learning_rate": 9.78310286320062e-05, "loss": 0.0568, "step": 27880 }, { "grad_norm": 0.6641424894332886, "learning_rate": 9.782861939587218e-05, "loss": 0.0626, "step": 27890 }, { "grad_norm": 0.6790198683738708, "learning_rate": 9.782620885212168e-05, "loss": 0.0785, "step": 27900 }, { "grad_norm": 0.6187946796417236, "learning_rate": 9.782379700082061e-05, "loss": 0.0679, "step": 27910 }, { "grad_norm": 0.5127753615379333, "learning_rate": 9.782138384203487e-05, "loss": 0.0709, "step": 27920 }, { "grad_norm": 0.6156544089317322, "learning_rate": 9.781896937583048e-05, "loss": 0.0881, "step": 27930 }, { "grad_norm": 0.49068737030029297, "learning_rate": 9.781655360227343e-05, "loss": 0.0902, "step": 27940 }, { "grad_norm": 0.7742199301719666, "learning_rate": 9.781413652142976e-05, "loss": 0.0601, "step": 27950 }, { "grad_norm": 0.508556604385376, "learning_rate": 9.781171813336556e-05, "loss": 0.0872, "step": 27960 }, { "grad_norm": 0.7320533990859985, "learning_rate": 9.780929843814696e-05, "loss": 0.0706, "step": 27970 }, { "grad_norm": 0.6384475827217102, "learning_rate": 9.78068774358401e-05, "loss": 0.0637, "step": 27980 }, { "grad_norm": 0.7296402454376221, "learning_rate": 9.780445512651116e-05, "loss": 0.0686, "step": 27990 }, { "grad_norm": 0.6880320310592651, "learning_rate": 9.780203151022639e-05, "loss": 0.0643, "step": 28000 }, { "grad_norm": 0.7251341342926025, "learning_rate": 9.779960658705203e-05, "loss": 0.0818, "step": 28010 }, { "grad_norm": 0.5954839587211609, "learning_rate": 9.779718035705439e-05, "loss": 0.0673, "step": 28020 }, { "grad_norm": 0.9535611867904663, "learning_rate": 9.779475282029977e-05, "loss": 0.0835, "step": 28030 }, { "grad_norm": 0.48707422614097595, "learning_rate": 9.779232397685458e-05, "loss": 0.0773, "step": 28040 }, { "grad_norm": 0.7103029489517212, "learning_rate": 9.778989382678523e-05, "loss": 0.0679, "step": 28050 }, { "grad_norm": 0.8660082221031189, "learning_rate": 9.77874623701581e-05, "loss": 0.0667, "step": 28060 }, { "grad_norm": 0.7042586207389832, "learning_rate": 9.778502960703972e-05, "loss": 0.0604, "step": 28070 }, { "grad_norm": 0.7822901606559753, "learning_rate": 9.778259553749658e-05, "loss": 0.0597, "step": 28080 }, { "grad_norm": 0.41029906272888184, "learning_rate": 9.778016016159524e-05, "loss": 0.0636, "step": 28090 }, { "grad_norm": 0.49983879923820496, "learning_rate": 9.777772347940226e-05, "loss": 0.0619, "step": 28100 }, { "grad_norm": 1.0183897018432617, "learning_rate": 9.777528549098427e-05, "loss": 0.082, "step": 28110 }, { "grad_norm": 0.5521094799041748, "learning_rate": 9.777284619640792e-05, "loss": 0.056, "step": 28120 }, { "grad_norm": 0.4511377215385437, "learning_rate": 9.777040559573992e-05, "loss": 0.0679, "step": 28130 }, { "grad_norm": 0.6586098074913025, "learning_rate": 9.776796368904695e-05, "loss": 0.0622, "step": 28140 }, { "grad_norm": 0.5920986533164978, "learning_rate": 9.776552047639581e-05, "loss": 0.0697, "step": 28150 }, { "grad_norm": 0.5323513150215149, "learning_rate": 9.77630759578533e-05, "loss": 0.0523, "step": 28160 }, { "grad_norm": 0.5329258441925049, "learning_rate": 9.776063013348622e-05, "loss": 0.0582, "step": 28170 }, { "grad_norm": 0.7325919270515442, "learning_rate": 9.775818300336146e-05, "loss": 0.0546, "step": 28180 }, { "grad_norm": 0.8670511841773987, "learning_rate": 9.775573456754591e-05, "loss": 0.0679, "step": 28190 }, { "grad_norm": 0.5505257248878479, "learning_rate": 9.775328482610653e-05, "loss": 0.0656, "step": 28200 }, { "grad_norm": 0.5633217096328735, "learning_rate": 9.775083377911028e-05, "loss": 0.0517, "step": 28210 }, { "grad_norm": 0.3789801001548767, "learning_rate": 9.774838142662417e-05, "loss": 0.0715, "step": 28220 }, { "grad_norm": 1.5502926111221313, "learning_rate": 9.774592776871525e-05, "loss": 0.0573, "step": 28230 }, { "grad_norm": 0.6957900524139404, "learning_rate": 9.77434728054506e-05, "loss": 0.0754, "step": 28240 }, { "grad_norm": 0.5296965837478638, "learning_rate": 9.774101653689735e-05, "loss": 0.0614, "step": 28250 }, { "grad_norm": 0.8053464889526367, "learning_rate": 9.773855896312263e-05, "loss": 0.0883, "step": 28260 }, { "grad_norm": 0.7344145178794861, "learning_rate": 9.773610008419365e-05, "loss": 0.0736, "step": 28270 }, { "grad_norm": 0.6034681797027588, "learning_rate": 9.773363990017763e-05, "loss": 0.0654, "step": 28280 }, { "grad_norm": 0.46838805079460144, "learning_rate": 9.773117841114182e-05, "loss": 0.0457, "step": 28290 }, { "grad_norm": 0.48713448643684387, "learning_rate": 9.772871561715353e-05, "loss": 0.067, "step": 28300 }, { "grad_norm": 0.9832140207290649, "learning_rate": 9.772625151828007e-05, "loss": 0.0865, "step": 28310 }, { "grad_norm": 0.9564993381500244, "learning_rate": 9.772378611458886e-05, "loss": 0.0736, "step": 28320 }, { "grad_norm": 0.690302848815918, "learning_rate": 9.772131940614724e-05, "loss": 0.0693, "step": 28330 }, { "grad_norm": 0.5499237775802612, "learning_rate": 9.771885139302268e-05, "loss": 0.0654, "step": 28340 }, { "grad_norm": 0.5945140719413757, "learning_rate": 9.771638207528264e-05, "loss": 0.0636, "step": 28350 }, { "grad_norm": 0.4104470908641815, "learning_rate": 9.771391145299465e-05, "loss": 0.0783, "step": 28360 }, { "grad_norm": 0.7222959399223328, "learning_rate": 9.771143952622626e-05, "loss": 0.0748, "step": 28370 }, { "grad_norm": 0.6542426943778992, "learning_rate": 9.770896629504502e-05, "loss": 0.0686, "step": 28380 }, { "grad_norm": 0.7149912714958191, "learning_rate": 9.770649175951857e-05, "loss": 0.0668, "step": 28390 }, { "grad_norm": 0.7353515028953552, "learning_rate": 9.770401591971455e-05, "loss": 0.0631, "step": 28400 }, { "grad_norm": 0.48507630825042725, "learning_rate": 9.770153877570067e-05, "loss": 0.0669, "step": 28410 }, { "grad_norm": 0.6020137667655945, "learning_rate": 9.769906032754464e-05, "loss": 0.0612, "step": 28420 }, { "grad_norm": 0.9090336561203003, "learning_rate": 9.769658057531422e-05, "loss": 0.0943, "step": 28430 }, { "grad_norm": 0.5792884230613708, "learning_rate": 9.769409951907719e-05, "loss": 0.0631, "step": 28440 }, { "grad_norm": 0.7606351375579834, "learning_rate": 9.76916171589014e-05, "loss": 0.0743, "step": 28450 }, { "grad_norm": 0.6127856969833374, "learning_rate": 9.768913349485472e-05, "loss": 0.0744, "step": 28460 }, { "grad_norm": 0.682151198387146, "learning_rate": 9.768664852700505e-05, "loss": 0.0674, "step": 28470 }, { "grad_norm": 0.6663133502006531, "learning_rate": 9.768416225542032e-05, "loss": 0.0613, "step": 28480 }, { "grad_norm": 0.5958750247955322, "learning_rate": 9.76816746801685e-05, "loss": 0.0837, "step": 28490 }, { "grad_norm": 0.5584161877632141, "learning_rate": 9.767918580131763e-05, "loss": 0.0662, "step": 28500 }, { "grad_norm": 0.5521354675292969, "learning_rate": 9.767669561893572e-05, "loss": 0.0694, "step": 28510 }, { "grad_norm": 0.6269077062606812, "learning_rate": 9.767420413309086e-05, "loss": 0.0605, "step": 28520 }, { "grad_norm": 0.5928886532783508, "learning_rate": 9.767171134385118e-05, "loss": 0.0723, "step": 28530 }, { "grad_norm": 0.553869903087616, "learning_rate": 9.766921725128481e-05, "loss": 0.0769, "step": 28540 }, { "grad_norm": 0.5729073882102966, "learning_rate": 9.766672185545995e-05, "loss": 0.0678, "step": 28550 }, { "grad_norm": 0.7680277824401855, "learning_rate": 9.766422515644482e-05, "loss": 0.0747, "step": 28560 }, { "grad_norm": 0.5615447163581848, "learning_rate": 9.766172715430769e-05, "loss": 0.0585, "step": 28570 }, { "grad_norm": 0.5780113935470581, "learning_rate": 9.765922784911683e-05, "loss": 0.0554, "step": 28580 }, { "grad_norm": 0.48996829986572266, "learning_rate": 9.76567272409406e-05, "loss": 0.066, "step": 28590 }, { "grad_norm": 0.7391840815544128, "learning_rate": 9.765422532984735e-05, "loss": 0.09, "step": 28600 }, { "grad_norm": 0.47889187932014465, "learning_rate": 9.765172211590547e-05, "loss": 0.0619, "step": 28610 }, { "grad_norm": 0.6479946374893188, "learning_rate": 9.764921759918342e-05, "loss": 0.0625, "step": 28620 }, { "grad_norm": 0.567376434803009, "learning_rate": 9.764671177974966e-05, "loss": 0.0795, "step": 28630 }, { "grad_norm": 0.4381171762943268, "learning_rate": 9.76442046576727e-05, "loss": 0.0551, "step": 28640 }, { "grad_norm": 0.618587076663971, "learning_rate": 9.764169623302109e-05, "loss": 0.0743, "step": 28650 }, { "grad_norm": 0.7096151113510132, "learning_rate": 9.763918650586338e-05, "loss": 0.0542, "step": 28660 }, { "grad_norm": 0.44278788566589355, "learning_rate": 9.763667547626825e-05, "loss": 0.055, "step": 28670 }, { "grad_norm": 0.6046476364135742, "learning_rate": 9.763416314430427e-05, "loss": 0.0632, "step": 28680 }, { "grad_norm": 0.43991148471832275, "learning_rate": 9.763164951004018e-05, "loss": 0.0612, "step": 28690 }, { "grad_norm": 0.8350784182548523, "learning_rate": 9.762913457354469e-05, "loss": 0.0661, "step": 28700 }, { "grad_norm": 0.6246744990348816, "learning_rate": 9.762661833488654e-05, "loss": 0.067, "step": 28710 }, { "grad_norm": 0.684108316898346, "learning_rate": 9.762410079413455e-05, "loss": 0.0592, "step": 28720 }, { "grad_norm": 0.6562833786010742, "learning_rate": 9.762158195135754e-05, "loss": 0.0659, "step": 28730 }, { "grad_norm": 0.6771926283836365, "learning_rate": 9.761906180662435e-05, "loss": 0.0689, "step": 28740 }, { "grad_norm": 0.6077961325645447, "learning_rate": 9.761654036000392e-05, "loss": 0.0601, "step": 28750 }, { "grad_norm": 0.5287161469459534, "learning_rate": 9.761401761156516e-05, "loss": 0.0706, "step": 28760 }, { "grad_norm": 0.5273951888084412, "learning_rate": 9.761149356137704e-05, "loss": 0.0629, "step": 28770 }, { "grad_norm": 0.5403570532798767, "learning_rate": 9.760896820950858e-05, "loss": 0.0815, "step": 28780 }, { "grad_norm": 0.5492487549781799, "learning_rate": 9.76064415560288e-05, "loss": 0.0721, "step": 28790 }, { "grad_norm": 0.7489679455757141, "learning_rate": 9.760391360100681e-05, "loss": 0.056, "step": 28800 }, { "grad_norm": 0.5127142071723938, "learning_rate": 9.76013843445117e-05, "loss": 0.0553, "step": 28810 }, { "grad_norm": 0.5661592483520508, "learning_rate": 9.759885378661263e-05, "loss": 0.0612, "step": 28820 }, { "grad_norm": 1.5588184595108032, "learning_rate": 9.759632192737878e-05, "loss": 0.0591, "step": 28830 }, { "grad_norm": 0.45544490218162537, "learning_rate": 9.759378876687937e-05, "loss": 0.0559, "step": 28840 }, { "grad_norm": 1.0041667222976685, "learning_rate": 9.759125430518365e-05, "loss": 0.0573, "step": 28850 }, { "grad_norm": 0.9767372608184814, "learning_rate": 9.758871854236091e-05, "loss": 0.0662, "step": 28860 }, { "grad_norm": 0.41621625423431396, "learning_rate": 9.758618147848049e-05, "loss": 0.0524, "step": 28870 }, { "grad_norm": 0.5460148453712463, "learning_rate": 9.758364311361174e-05, "loss": 0.0654, "step": 28880 }, { "grad_norm": 0.8844191431999207, "learning_rate": 9.758110344782408e-05, "loss": 0.0645, "step": 28890 }, { "grad_norm": 0.47078263759613037, "learning_rate": 9.75785624811869e-05, "loss": 0.059, "step": 28900 }, { "grad_norm": 0.5698981881141663, "learning_rate": 9.757602021376972e-05, "loss": 0.0694, "step": 28910 }, { "grad_norm": 0.5555749535560608, "learning_rate": 9.757347664564202e-05, "loss": 0.054, "step": 28920 }, { "grad_norm": 0.47207239270210266, "learning_rate": 9.757093177687333e-05, "loss": 0.0503, "step": 28930 }, { "grad_norm": 0.5743655562400818, "learning_rate": 9.756838560753325e-05, "loss": 0.0718, "step": 28940 }, { "grad_norm": 0.699832022190094, "learning_rate": 9.756583813769136e-05, "loss": 0.0598, "step": 28950 }, { "grad_norm": 0.6924501061439514, "learning_rate": 9.756328936741734e-05, "loss": 0.06, "step": 28960 }, { "grad_norm": 0.9670860171318054, "learning_rate": 9.756073929678085e-05, "loss": 0.0687, "step": 28970 }, { "grad_norm": 0.5661240220069885, "learning_rate": 9.75581879258516e-05, "loss": 0.061, "step": 28980 }, { "grad_norm": 0.5439299941062927, "learning_rate": 9.75556352546994e-05, "loss": 0.0768, "step": 28990 }, { "grad_norm": 3.1569411754608154, "learning_rate": 9.755308128339396e-05, "loss": 0.056, "step": 29000 }, { "grad_norm": 0.7123483419418335, "learning_rate": 9.755052601200517e-05, "loss": 0.062, "step": 29010 }, { "grad_norm": 0.4274228811264038, "learning_rate": 9.754796944060285e-05, "loss": 0.0693, "step": 29020 }, { "grad_norm": 0.5488526225090027, "learning_rate": 9.75454115692569e-05, "loss": 0.0563, "step": 29030 }, { "grad_norm": 0.6225519180297852, "learning_rate": 9.754285239803727e-05, "loss": 0.0636, "step": 29040 }, { "grad_norm": 0.597308337688446, "learning_rate": 9.75402919270139e-05, "loss": 0.1073, "step": 29050 }, { "grad_norm": 0.7548904418945312, "learning_rate": 9.753773015625682e-05, "loss": 0.0658, "step": 29060 }, { "grad_norm": 1.3418952226638794, "learning_rate": 9.753516708583607e-05, "loss": 0.0757, "step": 29070 }, { "grad_norm": 0.5721161365509033, "learning_rate": 9.75326027158217e-05, "loss": 0.0575, "step": 29080 }, { "grad_norm": 0.5448006391525269, "learning_rate": 9.753003704628383e-05, "loss": 0.0691, "step": 29090 }, { "grad_norm": 1.0835967063903809, "learning_rate": 9.752747007729259e-05, "loss": 0.0688, "step": 29100 }, { "grad_norm": 0.42788955569267273, "learning_rate": 9.75249018089182e-05, "loss": 0.069, "step": 29110 }, { "grad_norm": 0.4964495897293091, "learning_rate": 9.752233224123083e-05, "loss": 0.0629, "step": 29120 }, { "grad_norm": 0.8136771321296692, "learning_rate": 9.751976137430075e-05, "loss": 0.0658, "step": 29130 }, { "grad_norm": 0.6609320640563965, "learning_rate": 9.751718920819826e-05, "loss": 0.0756, "step": 29140 }, { "grad_norm": 0.4665728509426117, "learning_rate": 9.751461574299365e-05, "loss": 0.0685, "step": 29150 }, { "grad_norm": 1.004086971282959, "learning_rate": 9.75120409787573e-05, "loss": 0.0749, "step": 29160 }, { "grad_norm": 0.8448167443275452, "learning_rate": 9.750946491555961e-05, "loss": 0.076, "step": 29170 }, { "grad_norm": 0.4904869496822357, "learning_rate": 9.7506887553471e-05, "loss": 0.0816, "step": 29180 }, { "grad_norm": 0.5928230881690979, "learning_rate": 9.75043088925619e-05, "loss": 0.0586, "step": 29190 }, { "grad_norm": 0.7176315188407898, "learning_rate": 9.750172893290288e-05, "loss": 0.0752, "step": 29200 }, { "grad_norm": 1.006724238395691, "learning_rate": 9.749914767456441e-05, "loss": 0.0507, "step": 29210 }, { "grad_norm": 0.5646032094955444, "learning_rate": 9.74965651176171e-05, "loss": 0.0674, "step": 29220 }, { "grad_norm": 0.7447491884231567, "learning_rate": 9.749398126213154e-05, "loss": 0.0679, "step": 29230 }, { "grad_norm": 0.590513288974762, "learning_rate": 9.749139610817838e-05, "loss": 0.0506, "step": 29240 }, { "grad_norm": 0.6904083490371704, "learning_rate": 9.748880965582829e-05, "loss": 0.0523, "step": 29250 }, { "grad_norm": 0.8009453415870667, "learning_rate": 9.748622190515198e-05, "loss": 0.0677, "step": 29260 }, { "grad_norm": 0.8435172438621521, "learning_rate": 9.74836328562202e-05, "loss": 0.0613, "step": 29270 }, { "grad_norm": 0.6074042916297913, "learning_rate": 9.748104250910375e-05, "loss": 0.054, "step": 29280 }, { "grad_norm": 0.4717600643634796, "learning_rate": 9.747845086387342e-05, "loss": 0.0524, "step": 29290 }, { "grad_norm": 0.7282437086105347, "learning_rate": 9.74758579206001e-05, "loss": 0.0639, "step": 29300 }, { "grad_norm": 0.6202677488327026, "learning_rate": 9.747326367935464e-05, "loss": 0.0544, "step": 29310 }, { "grad_norm": 0.5409952402114868, "learning_rate": 9.7470668140208e-05, "loss": 0.0654, "step": 29320 }, { "grad_norm": 0.556053638458252, "learning_rate": 9.746807130323112e-05, "loss": 0.0716, "step": 29330 }, { "grad_norm": 0.6297544240951538, "learning_rate": 9.746547316849502e-05, "loss": 0.0764, "step": 29340 }, { "grad_norm": 0.6169591546058655, "learning_rate": 9.746287373607069e-05, "loss": 0.0663, "step": 29350 }, { "grad_norm": 0.5436936020851135, "learning_rate": 9.746027300602922e-05, "loss": 0.0557, "step": 29360 }, { "grad_norm": 0.6172446608543396, "learning_rate": 9.745767097844174e-05, "loss": 0.0515, "step": 29370 }, { "grad_norm": 0.5308115482330322, "learning_rate": 9.745506765337935e-05, "loss": 0.0506, "step": 29380 }, { "grad_norm": 0.4559507966041565, "learning_rate": 9.745246303091324e-05, "loss": 0.0648, "step": 29390 }, { "grad_norm": 0.538722574710846, "learning_rate": 9.744985711111464e-05, "loss": 0.0839, "step": 29400 }, { "grad_norm": 0.6507173180580139, "learning_rate": 9.744724989405474e-05, "loss": 0.0587, "step": 29410 }, { "grad_norm": 0.8053845167160034, "learning_rate": 9.744464137980487e-05, "loss": 0.0848, "step": 29420 }, { "grad_norm": 0.6605343818664551, "learning_rate": 9.744203156843632e-05, "loss": 0.0701, "step": 29430 }, { "grad_norm": 0.6276118755340576, "learning_rate": 9.743942046002047e-05, "loss": 0.0799, "step": 29440 }, { "grad_norm": 0.6511526703834534, "learning_rate": 9.743680805462864e-05, "loss": 0.0735, "step": 29450 }, { "grad_norm": 0.6873164176940918, "learning_rate": 9.743419435233234e-05, "loss": 0.0785, "step": 29460 }, { "grad_norm": 0.6523257493972778, "learning_rate": 9.743157935320297e-05, "loss": 0.0794, "step": 29470 }, { "grad_norm": 0.5884868502616882, "learning_rate": 9.742896305731206e-05, "loss": 0.0788, "step": 29480 }, { "grad_norm": 0.5552331209182739, "learning_rate": 9.742634546473111e-05, "loss": 0.0679, "step": 29490 }, { "grad_norm": 0.4559330940246582, "learning_rate": 9.742372657553167e-05, "loss": 0.0882, "step": 29500 }, { "grad_norm": 0.45356395840644836, "learning_rate": 9.742110638978538e-05, "loss": 0.0574, "step": 29510 }, { "grad_norm": 0.6627737283706665, "learning_rate": 9.741848490756387e-05, "loss": 0.0659, "step": 29520 }, { "grad_norm": 1.054279088973999, "learning_rate": 9.741586212893877e-05, "loss": 0.0617, "step": 29530 }, { "grad_norm": 0.5683802366256714, "learning_rate": 9.741323805398184e-05, "loss": 0.0557, "step": 29540 }, { "grad_norm": 0.7197126150131226, "learning_rate": 9.741061268276477e-05, "loss": 0.0667, "step": 29550 }, { "grad_norm": 0.6352335214614868, "learning_rate": 9.740798601535937e-05, "loss": 0.069, "step": 29560 }, { "grad_norm": 0.7665295004844666, "learning_rate": 9.740535805183744e-05, "loss": 0.0619, "step": 29570 }, { "grad_norm": 0.5368948578834534, "learning_rate": 9.740272879227084e-05, "loss": 0.0495, "step": 29580 }, { "grad_norm": 0.5220017433166504, "learning_rate": 9.740009823673144e-05, "loss": 0.0634, "step": 29590 }, { "grad_norm": 0.7098066806793213, "learning_rate": 9.739746638529115e-05, "loss": 0.0607, "step": 29600 }, { "grad_norm": 0.6051868200302124, "learning_rate": 9.739483323802194e-05, "loss": 0.0488, "step": 29610 }, { "grad_norm": 0.4638735055923462, "learning_rate": 9.73921987949958e-05, "loss": 0.0486, "step": 29620 }, { "grad_norm": 0.7599851489067078, "learning_rate": 9.738956305628476e-05, "loss": 0.0732, "step": 29630 }, { "grad_norm": 0.5819861888885498, "learning_rate": 9.738692602196085e-05, "loss": 0.0536, "step": 29640 }, { "grad_norm": 0.5934111475944519, "learning_rate": 9.738428769209618e-05, "loss": 0.1047, "step": 29650 }, { "grad_norm": 0.6109880208969116, "learning_rate": 9.73816480667629e-05, "loss": 0.0551, "step": 29660 }, { "grad_norm": 0.6247812509536743, "learning_rate": 9.737900714603316e-05, "loss": 0.11, "step": 29670 }, { "grad_norm": 0.4067780077457428, "learning_rate": 9.737636492997916e-05, "loss": 0.052, "step": 29680 }, { "grad_norm": 0.5359166264533997, "learning_rate": 9.737372141867315e-05, "loss": 0.0554, "step": 29690 }, { "grad_norm": 0.6640265583992004, "learning_rate": 9.737107661218738e-05, "loss": 0.1098, "step": 29700 }, { "grad_norm": 0.6397375464439392, "learning_rate": 9.73684305105942e-05, "loss": 0.0602, "step": 29710 }, { "grad_norm": 0.42622509598731995, "learning_rate": 9.736578311396589e-05, "loss": 0.0772, "step": 29720 }, { "grad_norm": 0.57331782579422, "learning_rate": 9.736313442237488e-05, "loss": 0.0572, "step": 29730 }, { "grad_norm": 0.4887990951538086, "learning_rate": 9.736048443589356e-05, "loss": 0.0576, "step": 29740 }, { "grad_norm": 0.6005712747573853, "learning_rate": 9.735783315459441e-05, "loss": 0.0547, "step": 29750 }, { "grad_norm": 1.2594014406204224, "learning_rate": 9.735518057854988e-05, "loss": 0.0875, "step": 29760 }, { "grad_norm": 0.5119661092758179, "learning_rate": 9.73525267078325e-05, "loss": 0.067, "step": 29770 }, { "grad_norm": 0.9393590688705444, "learning_rate": 9.734987154251483e-05, "loss": 0.0667, "step": 29780 }, { "grad_norm": 0.8647048473358154, "learning_rate": 9.734721508266946e-05, "loss": 0.057, "step": 29790 }, { "grad_norm": 0.7464419007301331, "learning_rate": 9.734455732836902e-05, "loss": 0.0523, "step": 29800 }, { "grad_norm": 0.40543532371520996, "learning_rate": 9.734189827968618e-05, "loss": 0.0569, "step": 29810 }, { "grad_norm": 0.8211783766746521, "learning_rate": 9.733923793669362e-05, "loss": 0.0794, "step": 29820 }, { "grad_norm": 0.6544003486633301, "learning_rate": 9.733657629946406e-05, "loss": 0.0517, "step": 29830 }, { "grad_norm": 0.5735533833503723, "learning_rate": 9.73339133680703e-05, "loss": 0.0549, "step": 29840 }, { "grad_norm": 0.630376935005188, "learning_rate": 9.733124914258514e-05, "loss": 0.0702, "step": 29850 }, { "grad_norm": 0.6640494465827942, "learning_rate": 9.732858362308141e-05, "loss": 0.0523, "step": 29860 }, { "grad_norm": 0.6844144463539124, "learning_rate": 9.732591680963196e-05, "loss": 0.0781, "step": 29870 }, { "grad_norm": 0.4698314964771271, "learning_rate": 9.732324870230975e-05, "loss": 0.0607, "step": 29880 }, { "grad_norm": 0.7172766923904419, "learning_rate": 9.732057930118769e-05, "loss": 0.0667, "step": 29890 }, { "grad_norm": 0.8350285291671753, "learning_rate": 9.731790860633878e-05, "loss": 0.0852, "step": 29900 }, { "grad_norm": 0.43749406933784485, "learning_rate": 9.731523661783601e-05, "loss": 0.0626, "step": 29910 }, { "grad_norm": 0.8101174831390381, "learning_rate": 9.731256333575246e-05, "loss": 0.0757, "step": 29920 }, { "grad_norm": 0.6004309058189392, "learning_rate": 9.730988876016118e-05, "loss": 0.054, "step": 29930 }, { "grad_norm": 0.7820136547088623, "learning_rate": 9.730721289113534e-05, "loss": 0.0801, "step": 29940 }, { "grad_norm": 0.5798683166503906, "learning_rate": 9.730453572874805e-05, "loss": 0.0717, "step": 29950 }, { "grad_norm": 0.5763264894485474, "learning_rate": 9.730185727307255e-05, "loss": 0.0604, "step": 29960 }, { "grad_norm": 0.5800615549087524, "learning_rate": 9.729917752418203e-05, "loss": 0.0563, "step": 29970 }, { "grad_norm": 0.5995568633079529, "learning_rate": 9.729649648214976e-05, "loss": 0.0563, "step": 29980 }, { "grad_norm": 0.5944081544876099, "learning_rate": 9.729381414704906e-05, "loss": 0.0585, "step": 29990 }, { "grad_norm": 0.5077909827232361, "learning_rate": 9.729113051895322e-05, "loss": 0.0684, "step": 30000 }, { "grad_norm": 0.6885697841644287, "learning_rate": 9.728844559793566e-05, "loss": 0.0598, "step": 30010 }, { "grad_norm": 0.5164771676063538, "learning_rate": 9.728575938406976e-05, "loss": 0.0598, "step": 30020 }, { "grad_norm": 0.6968998312950134, "learning_rate": 9.728307187742897e-05, "loss": 0.0619, "step": 30030 }, { "grad_norm": 0.5786165595054626, "learning_rate": 9.728038307808674e-05, "loss": 0.0647, "step": 30040 }, { "grad_norm": 0.49726352095603943, "learning_rate": 9.727769298611662e-05, "loss": 0.0554, "step": 30050 }, { "grad_norm": 0.7768471240997314, "learning_rate": 9.727500160159212e-05, "loss": 0.0588, "step": 30060 }, { "grad_norm": 0.5519285202026367, "learning_rate": 9.727230892458684e-05, "loss": 0.0972, "step": 30070 }, { "grad_norm": 0.8470852375030518, "learning_rate": 9.72696149551744e-05, "loss": 0.0751, "step": 30080 }, { "grad_norm": 0.5970964431762695, "learning_rate": 9.726691969342844e-05, "loss": 0.0864, "step": 30090 }, { "grad_norm": 0.5386285781860352, "learning_rate": 9.726422313942265e-05, "loss": 0.0922, "step": 30100 }, { "grad_norm": 0.44808727502822876, "learning_rate": 9.726152529323076e-05, "loss": 0.0726, "step": 30110 }, { "grad_norm": 0.6076274514198303, "learning_rate": 9.725882615492652e-05, "loss": 0.0701, "step": 30120 }, { "grad_norm": 0.5253189206123352, "learning_rate": 9.725612572458373e-05, "loss": 0.046, "step": 30130 }, { "grad_norm": 0.5088297128677368, "learning_rate": 9.725342400227621e-05, "loss": 0.0637, "step": 30140 }, { "grad_norm": 0.5558748841285706, "learning_rate": 9.725072098807784e-05, "loss": 0.0603, "step": 30150 }, { "grad_norm": 0.7782633900642395, "learning_rate": 9.72480166820625e-05, "loss": 0.0734, "step": 30160 }, { "grad_norm": 0.673608124256134, "learning_rate": 9.724531108430414e-05, "loss": 0.0661, "step": 30170 }, { "grad_norm": 0.55904620885849, "learning_rate": 9.724260419487674e-05, "loss": 0.0654, "step": 30180 }, { "grad_norm": 0.5674847364425659, "learning_rate": 9.723989601385425e-05, "loss": 0.0674, "step": 30190 }, { "grad_norm": 0.6669536828994751, "learning_rate": 9.723718654131078e-05, "loss": 0.0584, "step": 30200 }, { "grad_norm": 0.6499739289283752, "learning_rate": 9.723447577732037e-05, "loss": 0.0639, "step": 30210 }, { "grad_norm": 1.2924381494522095, "learning_rate": 9.723176372195715e-05, "loss": 0.0704, "step": 30220 }, { "grad_norm": 0.591788113117218, "learning_rate": 9.722905037529524e-05, "loss": 0.0571, "step": 30230 }, { "grad_norm": 0.6113234162330627, "learning_rate": 9.722633573740883e-05, "loss": 0.0549, "step": 30240 }, { "grad_norm": 0.4835471510887146, "learning_rate": 9.722361980837215e-05, "loss": 0.0575, "step": 30250 }, { "grad_norm": 0.5841102600097656, "learning_rate": 9.722090258825944e-05, "loss": 0.0525, "step": 30260 }, { "grad_norm": 0.5537900924682617, "learning_rate": 9.7218184077145e-05, "loss": 0.0433, "step": 30270 }, { "grad_norm": 0.7715656757354736, "learning_rate": 9.721546427510315e-05, "loss": 0.0478, "step": 30280 }, { "grad_norm": 0.8490579724311829, "learning_rate": 9.721274318220823e-05, "loss": 0.0655, "step": 30290 }, { "grad_norm": 0.5312951803207397, "learning_rate": 9.721002079853467e-05, "loss": 0.0694, "step": 30300 }, { "grad_norm": 0.7357950210571289, "learning_rate": 9.720729712415685e-05, "loss": 0.0746, "step": 30310 }, { "grad_norm": 0.6840144991874695, "learning_rate": 9.720457215914928e-05, "loss": 0.0722, "step": 30320 }, { "grad_norm": 0.3908993601799011, "learning_rate": 9.720184590358646e-05, "loss": 0.0571, "step": 30330 }, { "grad_norm": 0.7382761240005493, "learning_rate": 9.719911835754287e-05, "loss": 0.0799, "step": 30340 }, { "grad_norm": 0.6115530133247375, "learning_rate": 9.719638952109314e-05, "loss": 0.0581, "step": 30350 }, { "grad_norm": 0.4470806419849396, "learning_rate": 9.719365939431183e-05, "loss": 0.0575, "step": 30360 }, { "grad_norm": 0.5453231930732727, "learning_rate": 9.719092797727362e-05, "loss": 0.0598, "step": 30370 }, { "grad_norm": 0.5208419561386108, "learning_rate": 9.718819527005316e-05, "loss": 0.0492, "step": 30380 }, { "grad_norm": 0.6467622518539429, "learning_rate": 9.718546127272516e-05, "loss": 0.0846, "step": 30390 }, { "grad_norm": 0.5624409317970276, "learning_rate": 9.71827259853644e-05, "loss": 0.0594, "step": 30400 }, { "grad_norm": 0.5478466749191284, "learning_rate": 9.717998940804562e-05, "loss": 0.0639, "step": 30410 }, { "grad_norm": 0.344237357378006, "learning_rate": 9.717725154084365e-05, "loss": 0.0427, "step": 30420 }, { "grad_norm": 0.6735177636146545, "learning_rate": 9.717451238383335e-05, "loss": 0.0539, "step": 30430 }, { "grad_norm": 0.7208635210990906, "learning_rate": 9.717177193708962e-05, "loss": 0.0639, "step": 30440 }, { "grad_norm": 0.585046112537384, "learning_rate": 9.716903020068735e-05, "loss": 0.0537, "step": 30450 }, { "grad_norm": 0.5326189994812012, "learning_rate": 9.716628717470152e-05, "loss": 0.0514, "step": 30460 }, { "grad_norm": 0.9784504771232605, "learning_rate": 9.716354285920712e-05, "loss": 0.053, "step": 30470 }, { "grad_norm": 0.47731322050094604, "learning_rate": 9.716079725427918e-05, "loss": 0.0517, "step": 30480 }, { "grad_norm": 0.7803354859352112, "learning_rate": 9.715805035999275e-05, "loss": 0.0482, "step": 30490 }, { "grad_norm": 0.6669082641601562, "learning_rate": 9.715530217642295e-05, "loss": 0.0837, "step": 30500 }, { "grad_norm": 0.8485336303710938, "learning_rate": 9.715255270364489e-05, "loss": 0.0699, "step": 30510 }, { "grad_norm": 0.5990592837333679, "learning_rate": 9.714980194173376e-05, "loss": 0.0569, "step": 30520 }, { "grad_norm": 0.6033564209938049, "learning_rate": 9.714704989076477e-05, "loss": 0.0571, "step": 30530 }, { "grad_norm": 0.4015948474407196, "learning_rate": 9.714429655081314e-05, "loss": 0.046, "step": 30540 }, { "grad_norm": 0.6693099737167358, "learning_rate": 9.714154192195415e-05, "loss": 0.0555, "step": 30550 }, { "grad_norm": 0.5639888644218445, "learning_rate": 9.713878600426312e-05, "loss": 0.0482, "step": 30560 }, { "grad_norm": 0.635382354259491, "learning_rate": 9.713602879781538e-05, "loss": 0.0836, "step": 30570 }, { "grad_norm": 0.5435240864753723, "learning_rate": 9.713327030268633e-05, "loss": 0.0677, "step": 30580 }, { "grad_norm": 0.8006575107574463, "learning_rate": 9.713051051895137e-05, "loss": 0.0679, "step": 30590 }, { "grad_norm": 0.4714156687259674, "learning_rate": 9.712774944668596e-05, "loss": 0.0629, "step": 30600 }, { "grad_norm": 0.5976464152336121, "learning_rate": 9.712498708596559e-05, "loss": 0.0816, "step": 30610 }, { "grad_norm": 0.5813767313957214, "learning_rate": 9.712222343686578e-05, "loss": 0.0625, "step": 30620 }, { "grad_norm": 0.388942152261734, "learning_rate": 9.711945849946207e-05, "loss": 0.0521, "step": 30630 }, { "grad_norm": 1.0168370008468628, "learning_rate": 9.71166922738301e-05, "loss": 0.0455, "step": 30640 }, { "grad_norm": 0.3737016022205353, "learning_rate": 9.711392476004543e-05, "loss": 0.0462, "step": 30650 }, { "grad_norm": 0.6585996747016907, "learning_rate": 9.711115595818376e-05, "loss": 0.065, "step": 30660 }, { "grad_norm": 0.5069511532783508, "learning_rate": 9.71083858683208e-05, "loss": 0.0604, "step": 30670 }, { "grad_norm": 0.43752771615982056, "learning_rate": 9.710561449053226e-05, "loss": 0.0558, "step": 30680 }, { "grad_norm": 1.013378381729126, "learning_rate": 9.710284182489391e-05, "loss": 0.0666, "step": 30690 }, { "grad_norm": 0.6710794568061829, "learning_rate": 9.710006787148158e-05, "loss": 0.0793, "step": 30700 }, { "grad_norm": 0.7791982889175415, "learning_rate": 9.709729263037106e-05, "loss": 0.0511, "step": 30710 }, { "grad_norm": 0.5686426162719727, "learning_rate": 9.709451610163826e-05, "loss": 0.0593, "step": 30720 }, { "grad_norm": 0.6202645301818848, "learning_rate": 9.709173828535911e-05, "loss": 0.0683, "step": 30730 }, { "grad_norm": 0.5372584462165833, "learning_rate": 9.708895918160951e-05, "loss": 0.0574, "step": 30740 }, { "grad_norm": 0.704067587852478, "learning_rate": 9.708617879046544e-05, "loss": 0.0728, "step": 30750 }, { "grad_norm": 0.40465274453163147, "learning_rate": 9.708339711200295e-05, "loss": 0.0598, "step": 30760 }, { "grad_norm": 0.6180625557899475, "learning_rate": 9.708061414629808e-05, "loss": 0.0777, "step": 30770 }, { "grad_norm": 0.7292197942733765, "learning_rate": 9.707782989342688e-05, "loss": 0.0573, "step": 30780 }, { "grad_norm": 0.5845041275024414, "learning_rate": 9.707504435346552e-05, "loss": 0.0709, "step": 30790 }, { "grad_norm": 0.731235146522522, "learning_rate": 9.707225752649012e-05, "loss": 0.0575, "step": 30800 }, { "grad_norm": 0.6368486881256104, "learning_rate": 9.706946941257689e-05, "loss": 0.0514, "step": 30810 }, { "grad_norm": 0.5738276243209839, "learning_rate": 9.706668001180204e-05, "loss": 0.0457, "step": 30820 }, { "grad_norm": 0.6140639781951904, "learning_rate": 9.706388932424185e-05, "loss": 0.075, "step": 30830 }, { "grad_norm": 0.4602874517440796, "learning_rate": 9.70610973499726e-05, "loss": 0.0637, "step": 30840 }, { "grad_norm": 0.6843332648277283, "learning_rate": 9.705830408907063e-05, "loss": 0.0657, "step": 30850 }, { "grad_norm": 0.7429358959197998, "learning_rate": 9.705550954161233e-05, "loss": 0.0844, "step": 30860 }, { "grad_norm": 0.5726335048675537, "learning_rate": 9.705271370767403e-05, "loss": 0.0545, "step": 30870 }, { "grad_norm": 0.6664488911628723, "learning_rate": 9.704991658733225e-05, "loss": 0.059, "step": 30880 }, { "grad_norm": 0.4555465579032898, "learning_rate": 9.704711818066342e-05, "loss": 0.0723, "step": 30890 }, { "grad_norm": 0.4813016653060913, "learning_rate": 9.704431848774405e-05, "loss": 0.0552, "step": 30900 }, { "grad_norm": 0.5940849184989929, "learning_rate": 9.704151750865069e-05, "loss": 0.055, "step": 30910 }, { "grad_norm": 0.9176604151725769, "learning_rate": 9.70387152434599e-05, "loss": 0.0758, "step": 30920 }, { "grad_norm": 0.5355153679847717, "learning_rate": 9.703591169224833e-05, "loss": 0.0617, "step": 30930 }, { "grad_norm": 0.542524516582489, "learning_rate": 9.703310685509258e-05, "loss": 0.075, "step": 30940 }, { "grad_norm": 0.4325898587703705, "learning_rate": 9.703030073206939e-05, "loss": 0.0571, "step": 30950 }, { "grad_norm": 0.6710599660873413, "learning_rate": 9.70274933232554e-05, "loss": 0.0819, "step": 30960 }, { "grad_norm": 0.6687977313995361, "learning_rate": 9.702468462872745e-05, "loss": 0.0835, "step": 30970 }, { "grad_norm": 0.7228779196739197, "learning_rate": 9.702187464856227e-05, "loss": 0.0691, "step": 30980 }, { "grad_norm": 0.47927340865135193, "learning_rate": 9.70190633828367e-05, "loss": 0.0492, "step": 30990 }, { "grad_norm": 0.7247996926307678, "learning_rate": 9.701625083162763e-05, "loss": 0.0769, "step": 31000 }, { "grad_norm": 0.7771440744400024, "learning_rate": 9.701343699501189e-05, "loss": 0.0786, "step": 31010 }, { "grad_norm": 0.553693413734436, "learning_rate": 9.701062187306645e-05, "loss": 0.0648, "step": 31020 }, { "grad_norm": 0.45224812626838684, "learning_rate": 9.700780546586827e-05, "loss": 0.0452, "step": 31030 }, { "grad_norm": 0.7465060353279114, "learning_rate": 9.700498777349435e-05, "loss": 0.0553, "step": 31040 }, { "grad_norm": 0.5348809361457825, "learning_rate": 9.700216879602175e-05, "loss": 0.0405, "step": 31050 }, { "grad_norm": 0.7034539580345154, "learning_rate": 9.699934853352749e-05, "loss": 0.0526, "step": 31060 }, { "grad_norm": 0.5130443572998047, "learning_rate": 9.69965269860887e-05, "loss": 0.0703, "step": 31070 }, { "grad_norm": 0.5673633813858032, "learning_rate": 9.699370415378252e-05, "loss": 0.0535, "step": 31080 }, { "grad_norm": 0.7226769328117371, "learning_rate": 9.699088003668612e-05, "loss": 0.0532, "step": 31090 }, { "grad_norm": 0.6405554413795471, "learning_rate": 9.698805463487671e-05, "loss": 0.0474, "step": 31100 }, { "grad_norm": 0.6643682718276978, "learning_rate": 9.698522794843154e-05, "loss": 0.0682, "step": 31110 }, { "grad_norm": 0.6926302313804626, "learning_rate": 9.69823999774279e-05, "loss": 0.0491, "step": 31120 }, { "grad_norm": 0.5990027189254761, "learning_rate": 9.697957072194307e-05, "loss": 0.0631, "step": 31130 }, { "grad_norm": 0.5702651143074036, "learning_rate": 9.697674018205446e-05, "loss": 0.0649, "step": 31140 }, { "grad_norm": 0.6075853109359741, "learning_rate": 9.697390835783941e-05, "loss": 0.067, "step": 31150 }, { "grad_norm": 0.4558906853199005, "learning_rate": 9.697107524937534e-05, "loss": 0.0738, "step": 31160 }, { "grad_norm": 0.36887213587760925, "learning_rate": 9.696824085673973e-05, "loss": 0.052, "step": 31170 }, { "grad_norm": 0.5133751034736633, "learning_rate": 9.696540518001006e-05, "loss": 0.0548, "step": 31180 }, { "grad_norm": 0.511691689491272, "learning_rate": 9.696256821926386e-05, "loss": 0.0557, "step": 31190 }, { "grad_norm": 0.6217912435531616, "learning_rate": 9.695972997457867e-05, "loss": 0.0517, "step": 31200 }, { "grad_norm": 0.41904881596565247, "learning_rate": 9.695689044603212e-05, "loss": 0.0632, "step": 31210 }, { "grad_norm": 0.549009382724762, "learning_rate": 9.695404963370183e-05, "loss": 0.0661, "step": 31220 }, { "grad_norm": 0.4803968369960785, "learning_rate": 9.695120753766544e-05, "loss": 0.0574, "step": 31230 }, { "grad_norm": 0.49782800674438477, "learning_rate": 9.69483641580007e-05, "loss": 0.048, "step": 31240 }, { "grad_norm": 0.5072928071022034, "learning_rate": 9.694551949478533e-05, "loss": 0.0579, "step": 31250 }, { "grad_norm": 0.5611820220947266, "learning_rate": 9.694267354809705e-05, "loss": 0.0608, "step": 31260 }, { "grad_norm": 0.6045109033584595, "learning_rate": 9.693982631801374e-05, "loss": 0.0587, "step": 31270 }, { "grad_norm": 0.5532360076904297, "learning_rate": 9.693697780461322e-05, "loss": 0.0603, "step": 31280 }, { "grad_norm": 0.4445231556892395, "learning_rate": 9.693412800797335e-05, "loss": 0.0543, "step": 31290 }, { "grad_norm": 0.7203876972198486, "learning_rate": 9.693127692817204e-05, "loss": 0.0522, "step": 31300 }, { "grad_norm": 0.7461557984352112, "learning_rate": 9.692842456528726e-05, "loss": 0.0568, "step": 31310 }, { "grad_norm": 0.680303692817688, "learning_rate": 9.692557091939697e-05, "loss": 0.0732, "step": 31320 }, { "grad_norm": 0.621016800403595, "learning_rate": 9.692271599057922e-05, "loss": 0.0585, "step": 31330 }, { "grad_norm": 0.8061418533325195, "learning_rate": 9.691985977891205e-05, "loss": 0.064, "step": 31340 }, { "grad_norm": 0.6403115391731262, "learning_rate": 9.691700228447353e-05, "loss": 0.0651, "step": 31350 }, { "grad_norm": 0.7993391156196594, "learning_rate": 9.691414350734178e-05, "loss": 0.0577, "step": 31360 }, { "grad_norm": 1.227364420890808, "learning_rate": 9.691128344759499e-05, "loss": 0.0729, "step": 31370 }, { "grad_norm": 0.5343287587165833, "learning_rate": 9.690842210531133e-05, "loss": 0.0621, "step": 31380 }, { "grad_norm": 0.5796718597412109, "learning_rate": 9.690555948056903e-05, "loss": 0.0664, "step": 31390 }, { "grad_norm": 0.6025047302246094, "learning_rate": 9.690269557344635e-05, "loss": 0.0597, "step": 31400 }, { "grad_norm": 0.431545615196228, "learning_rate": 9.689983038402159e-05, "loss": 0.0578, "step": 31410 }, { "grad_norm": 0.5081874132156372, "learning_rate": 9.689696391237311e-05, "loss": 0.0631, "step": 31420 }, { "grad_norm": 0.34758251905441284, "learning_rate": 9.689409615857923e-05, "loss": 0.0442, "step": 31430 }, { "grad_norm": 0.8417641520500183, "learning_rate": 9.689122712271838e-05, "loss": 0.0557, "step": 31440 }, { "grad_norm": 0.6289563775062561, "learning_rate": 9.688835680486901e-05, "loss": 0.0696, "step": 31450 }, { "grad_norm": 0.6066879630088806, "learning_rate": 9.688548520510959e-05, "loss": 0.0683, "step": 31460 }, { "grad_norm": 0.7236223816871643, "learning_rate": 9.68826123235186e-05, "loss": 0.0562, "step": 31470 }, { "grad_norm": 0.35613200068473816, "learning_rate": 9.68797381601746e-05, "loss": 0.0517, "step": 31480 }, { "grad_norm": 0.5523813962936401, "learning_rate": 9.687686271515618e-05, "loss": 0.0734, "step": 31490 }, { "grad_norm": 0.7033831477165222, "learning_rate": 9.687398598854194e-05, "loss": 0.054, "step": 31500 }, { "grad_norm": 0.5269955396652222, "learning_rate": 9.687110798041053e-05, "loss": 0.0536, "step": 31510 }, { "grad_norm": 0.7842751145362854, "learning_rate": 9.686822869084065e-05, "loss": 0.0605, "step": 31520 }, { "grad_norm": 0.515598714351654, "learning_rate": 9.686534811991099e-05, "loss": 0.05, "step": 31530 }, { "grad_norm": 0.6486472487449646, "learning_rate": 9.686246626770032e-05, "loss": 0.0683, "step": 31540 }, { "grad_norm": 0.617555558681488, "learning_rate": 9.685958313428744e-05, "loss": 0.0665, "step": 31550 }, { "grad_norm": 0.451647013425827, "learning_rate": 9.685669871975113e-05, "loss": 0.061, "step": 31560 }, { "grad_norm": 0.6256020069122314, "learning_rate": 9.685381302417032e-05, "loss": 0.0823, "step": 31570 }, { "grad_norm": 0.7462055683135986, "learning_rate": 9.685092604762385e-05, "loss": 0.0622, "step": 31580 }, { "grad_norm": 0.7484483122825623, "learning_rate": 9.684803779019065e-05, "loss": 0.0666, "step": 31590 }, { "grad_norm": 0.4811736047267914, "learning_rate": 9.684514825194971e-05, "loss": 0.0551, "step": 31600 }, { "grad_norm": 0.4578191041946411, "learning_rate": 9.684225743298001e-05, "loss": 0.0732, "step": 31610 }, { "grad_norm": 0.3998435437679291, "learning_rate": 9.68393653333606e-05, "loss": 0.0634, "step": 31620 }, { "grad_norm": 0.7692849040031433, "learning_rate": 9.683647195317053e-05, "loss": 0.0616, "step": 31630 }, { "grad_norm": 0.7075166702270508, "learning_rate": 9.683357729248892e-05, "loss": 0.0614, "step": 31640 }, { "grad_norm": 0.6919487118721008, "learning_rate": 9.683068135139488e-05, "loss": 0.0617, "step": 31650 }, { "grad_norm": 0.5966624617576599, "learning_rate": 9.682778412996762e-05, "loss": 0.0599, "step": 31660 }, { "grad_norm": 0.4741901755332947, "learning_rate": 9.682488562828634e-05, "loss": 0.0557, "step": 31670 }, { "grad_norm": 0.7790765762329102, "learning_rate": 9.682198584643027e-05, "loss": 0.0613, "step": 31680 }, { "grad_norm": 1.1309504508972168, "learning_rate": 9.68190847844787e-05, "loss": 0.0609, "step": 31690 }, { "grad_norm": 0.7971528768539429, "learning_rate": 9.681618244251094e-05, "loss": 0.0535, "step": 31700 }, { "grad_norm": 0.5436838865280151, "learning_rate": 9.681327882060634e-05, "loss": 0.0533, "step": 31710 }, { "grad_norm": 0.5982955694198608, "learning_rate": 9.681037391884428e-05, "loss": 0.064, "step": 31720 }, { "grad_norm": 0.7151640057563782, "learning_rate": 9.680746773730421e-05, "loss": 0.0701, "step": 31730 }, { "grad_norm": 0.6729803085327148, "learning_rate": 9.680456027606552e-05, "loss": 0.0562, "step": 31740 }, { "grad_norm": 0.815657913684845, "learning_rate": 9.680165153520775e-05, "loss": 0.0827, "step": 31750 }, { "grad_norm": 0.5651172995567322, "learning_rate": 9.679874151481041e-05, "loss": 0.0533, "step": 31760 }, { "grad_norm": 0.6504471898078918, "learning_rate": 9.679583021495307e-05, "loss": 0.0722, "step": 31770 }, { "grad_norm": 0.4573388695716858, "learning_rate": 9.67929176357153e-05, "loss": 0.0554, "step": 31780 }, { "grad_norm": 0.38463377952575684, "learning_rate": 9.679000377717673e-05, "loss": 0.0507, "step": 31790 }, { "grad_norm": 0.5532046556472778, "learning_rate": 9.678708863941705e-05, "loss": 0.0621, "step": 31800 }, { "grad_norm": 0.5487226843833923, "learning_rate": 9.678417222251594e-05, "loss": 0.0537, "step": 31810 }, { "grad_norm": 0.7988343834877014, "learning_rate": 9.678125452655313e-05, "loss": 0.0592, "step": 31820 }, { "grad_norm": 0.5935947895050049, "learning_rate": 9.67783355516084e-05, "loss": 0.0608, "step": 31830 }, { "grad_norm": 0.48707902431488037, "learning_rate": 9.677541529776156e-05, "loss": 0.0584, "step": 31840 }, { "grad_norm": 0.6705939769744873, "learning_rate": 9.677249376509243e-05, "loss": 0.0736, "step": 31850 }, { "grad_norm": 0.5147388577461243, "learning_rate": 9.67695709536809e-05, "loss": 0.0719, "step": 31860 }, { "grad_norm": 0.6885640025138855, "learning_rate": 9.676664686360686e-05, "loss": 0.059, "step": 31870 }, { "grad_norm": 0.5097650289535522, "learning_rate": 9.676372149495027e-05, "loss": 0.049, "step": 31880 }, { "grad_norm": 0.5028774738311768, "learning_rate": 9.67607948477911e-05, "loss": 0.0549, "step": 31890 }, { "grad_norm": 0.5359739065170288, "learning_rate": 9.675786692220938e-05, "loss": 0.0521, "step": 31900 }, { "grad_norm": 0.7785214185714722, "learning_rate": 9.675493771828512e-05, "loss": 0.0576, "step": 31910 }, { "grad_norm": 0.5296103954315186, "learning_rate": 9.675200723609845e-05, "loss": 0.0502, "step": 31920 }, { "grad_norm": 0.5432544350624084, "learning_rate": 9.674907547572944e-05, "loss": 0.0495, "step": 31930 }, { "grad_norm": 0.6766061186790466, "learning_rate": 9.67461424372583e-05, "loss": 0.0693, "step": 31940 }, { "grad_norm": 0.7558256387710571, "learning_rate": 9.674320812076516e-05, "loss": 0.0581, "step": 31950 }, { "grad_norm": 0.7908263802528381, "learning_rate": 9.674027252633029e-05, "loss": 0.0591, "step": 31960 }, { "grad_norm": 1.1072180271148682, "learning_rate": 9.673733565403392e-05, "loss": 0.069, "step": 31970 }, { "grad_norm": 0.7104498744010925, "learning_rate": 9.673439750395636e-05, "loss": 0.0728, "step": 31980 }, { "grad_norm": 0.5944513082504272, "learning_rate": 9.673145807617792e-05, "loss": 0.063, "step": 31990 }, { "grad_norm": 0.5537567138671875, "learning_rate": 9.672851737077898e-05, "loss": 0.0605, "step": 32000 }, { "grad_norm": 0.7480613589286804, "learning_rate": 9.672557538783993e-05, "loss": 0.0732, "step": 32010 }, { "grad_norm": 0.6037663221359253, "learning_rate": 9.67226321274412e-05, "loss": 0.0573, "step": 32020 }, { "grad_norm": 0.5072296261787415, "learning_rate": 9.671968758966326e-05, "loss": 0.0613, "step": 32030 }, { "grad_norm": 0.7922964096069336, "learning_rate": 9.67167417745866e-05, "loss": 0.0593, "step": 32040 }, { "grad_norm": 0.44733256101608276, "learning_rate": 9.671379468229178e-05, "loss": 0.0591, "step": 32050 }, { "grad_norm": 0.43282923102378845, "learning_rate": 9.671084631285937e-05, "loss": 0.0491, "step": 32060 }, { "grad_norm": 0.6526686549186707, "learning_rate": 9.670789666636997e-05, "loss": 0.0724, "step": 32070 }, { "grad_norm": 0.3831295967102051, "learning_rate": 9.670494574290421e-05, "loss": 0.054, "step": 32080 }, { "grad_norm": 0.6117400527000427, "learning_rate": 9.670199354254279e-05, "loss": 0.0613, "step": 32090 }, { "grad_norm": 0.6059635877609253, "learning_rate": 9.66990400653664e-05, "loss": 0.0552, "step": 32100 }, { "grad_norm": 0.6449746489524841, "learning_rate": 9.669608531145581e-05, "loss": 0.0517, "step": 32110 }, { "grad_norm": 0.7717724442481995, "learning_rate": 9.669312928089181e-05, "loss": 0.0572, "step": 32120 }, { "grad_norm": 0.5167685747146606, "learning_rate": 9.669017197375517e-05, "loss": 0.0549, "step": 32130 }, { "grad_norm": 0.6130321025848389, "learning_rate": 9.668721339012677e-05, "loss": 0.0508, "step": 32140 }, { "grad_norm": 0.5946348905563354, "learning_rate": 9.668425353008749e-05, "loss": 0.0606, "step": 32150 }, { "grad_norm": 0.6323691010475159, "learning_rate": 9.668129239371828e-05, "loss": 0.0571, "step": 32160 }, { "grad_norm": 0.46538710594177246, "learning_rate": 9.667832998110006e-05, "loss": 0.0485, "step": 32170 }, { "grad_norm": 0.5529741644859314, "learning_rate": 9.667536629231384e-05, "loss": 0.0653, "step": 32180 }, { "grad_norm": 0.7351182699203491, "learning_rate": 9.667240132744063e-05, "loss": 0.0585, "step": 32190 }, { "grad_norm": 0.5353500247001648, "learning_rate": 9.66694350865615e-05, "loss": 0.0578, "step": 32200 }, { "grad_norm": 0.4712783694267273, "learning_rate": 9.666646756975757e-05, "loss": 0.0599, "step": 32210 }, { "grad_norm": 1.0046721696853638, "learning_rate": 9.666349877710991e-05, "loss": 0.0626, "step": 32220 }, { "grad_norm": 0.7288378477096558, "learning_rate": 9.666052870869977e-05, "loss": 0.0815, "step": 32230 }, { "grad_norm": 0.9350880980491638, "learning_rate": 9.665755736460828e-05, "loss": 0.081, "step": 32240 }, { "grad_norm": 0.9230968356132507, "learning_rate": 9.66545847449167e-05, "loss": 0.0549, "step": 32250 }, { "grad_norm": 0.6130694150924683, "learning_rate": 9.665161084970631e-05, "loss": 0.0748, "step": 32260 }, { "grad_norm": 0.648674726486206, "learning_rate": 9.66486356790584e-05, "loss": 0.0524, "step": 32270 }, { "grad_norm": 0.5595231652259827, "learning_rate": 9.664565923305432e-05, "loss": 0.0717, "step": 32280 }, { "grad_norm": 0.860980749130249, "learning_rate": 9.664268151177543e-05, "loss": 0.0705, "step": 32290 }, { "grad_norm": 0.5609794855117798, "learning_rate": 9.663970251530314e-05, "loss": 0.0602, "step": 32300 }, { "grad_norm": 0.7160289287567139, "learning_rate": 9.663672224371892e-05, "loss": 0.0692, "step": 32310 }, { "grad_norm": 0.5390477180480957, "learning_rate": 9.663374069710423e-05, "loss": 0.0468, "step": 32320 }, { "grad_norm": 0.5024919509887695, "learning_rate": 9.66307578755406e-05, "loss": 0.0545, "step": 32330 }, { "grad_norm": 0.5793772339820862, "learning_rate": 9.662777377910956e-05, "loss": 0.0645, "step": 32340 }, { "grad_norm": 0.6963393688201904, "learning_rate": 9.662478840789271e-05, "loss": 0.0514, "step": 32350 }, { "grad_norm": 0.618340253829956, "learning_rate": 9.662180176197164e-05, "loss": 0.0509, "step": 32360 }, { "grad_norm": 0.661268413066864, "learning_rate": 9.661881384142805e-05, "loss": 0.0473, "step": 32370 }, { "grad_norm": 0.8731707334518433, "learning_rate": 9.66158246463436e-05, "loss": 0.06, "step": 32380 }, { "grad_norm": 0.5727486610412598, "learning_rate": 9.66128341768e-05, "loss": 0.0582, "step": 32390 }, { "grad_norm": 0.544973611831665, "learning_rate": 9.660984243287903e-05, "loss": 0.0577, "step": 32400 }, { "grad_norm": 0.5378567576408386, "learning_rate": 9.66068494146625e-05, "loss": 0.084, "step": 32410 }, { "grad_norm": 0.37437763810157776, "learning_rate": 9.660385512223219e-05, "loss": 0.0507, "step": 32420 }, { "grad_norm": 0.6646509766578674, "learning_rate": 9.660085955567e-05, "loss": 0.0679, "step": 32430 }, { "grad_norm": 0.5744009613990784, "learning_rate": 9.659786271505782e-05, "loss": 0.0496, "step": 32440 }, { "grad_norm": 0.45891669392585754, "learning_rate": 9.659486460047757e-05, "loss": 0.0686, "step": 32450 }, { "grad_norm": 0.5454237461090088, "learning_rate": 9.659186521201122e-05, "loss": 0.0706, "step": 32460 }, { "grad_norm": 0.5514001250267029, "learning_rate": 9.65888645497408e-05, "loss": 0.0653, "step": 32470 }, { "grad_norm": 0.5486423969268799, "learning_rate": 9.658586261374832e-05, "loss": 0.0596, "step": 32480 }, { "grad_norm": 0.6906638145446777, "learning_rate": 9.658285940411584e-05, "loss": 0.0617, "step": 32490 }, { "grad_norm": 0.8585004806518555, "learning_rate": 9.657985492092551e-05, "loss": 0.0599, "step": 32500 }, { "grad_norm": 0.6520304679870605, "learning_rate": 9.657684916425944e-05, "loss": 0.0585, "step": 32510 }, { "grad_norm": 0.6645215153694153, "learning_rate": 9.65738421341998e-05, "loss": 0.0617, "step": 32520 }, { "grad_norm": 0.7293840646743774, "learning_rate": 9.657083383082881e-05, "loss": 0.0655, "step": 32530 }, { "grad_norm": 1.0717068910598755, "learning_rate": 9.656782425422874e-05, "loss": 0.0667, "step": 32540 }, { "grad_norm": 0.4978841245174408, "learning_rate": 9.656481340448184e-05, "loss": 0.071, "step": 32550 }, { "grad_norm": 0.7539440393447876, "learning_rate": 9.656180128167042e-05, "loss": 0.0532, "step": 32560 }, { "grad_norm": 0.5172296762466431, "learning_rate": 9.655878788587687e-05, "loss": 0.0535, "step": 32570 }, { "grad_norm": 0.48542338609695435, "learning_rate": 9.655577321718353e-05, "loss": 0.0544, "step": 32580 }, { "grad_norm": 0.6455007195472717, "learning_rate": 9.655275727567285e-05, "loss": 0.0525, "step": 32590 }, { "grad_norm": 0.4736669361591339, "learning_rate": 9.654974006142727e-05, "loss": 0.05, "step": 32600 }, { "grad_norm": 0.7491263747215271, "learning_rate": 9.654672157452929e-05, "loss": 0.0558, "step": 32610 }, { "grad_norm": 0.49263593554496765, "learning_rate": 9.654370181506143e-05, "loss": 0.0661, "step": 32620 }, { "grad_norm": 0.5706984996795654, "learning_rate": 9.654068078310625e-05, "loss": 0.0596, "step": 32630 }, { "grad_norm": 0.5678548216819763, "learning_rate": 9.653765847874634e-05, "loss": 0.0528, "step": 32640 }, { "grad_norm": 0.7864235639572144, "learning_rate": 9.653463490206432e-05, "loss": 0.0846, "step": 32650 }, { "grad_norm": 1.215809941291809, "learning_rate": 9.653161005314288e-05, "loss": 0.0598, "step": 32660 }, { "grad_norm": 0.561154305934906, "learning_rate": 9.65285839320647e-05, "loss": 0.0565, "step": 32670 }, { "grad_norm": 0.4779711961746216, "learning_rate": 9.65255565389125e-05, "loss": 0.0531, "step": 32680 }, { "grad_norm": 0.599826991558075, "learning_rate": 9.652252787376908e-05, "loss": 0.0537, "step": 32690 }, { "grad_norm": 0.47857680916786194, "learning_rate": 9.651949793671722e-05, "loss": 0.0479, "step": 32700 }, { "grad_norm": 0.5873757600784302, "learning_rate": 9.651646672783976e-05, "loss": 0.0649, "step": 32710 }, { "grad_norm": 0.6962327361106873, "learning_rate": 9.651343424721957e-05, "loss": 0.0559, "step": 32720 }, { "grad_norm": 0.6234087944030762, "learning_rate": 9.651040049493956e-05, "loss": 0.0569, "step": 32730 }, { "grad_norm": 0.679387092590332, "learning_rate": 9.65073654710827e-05, "loss": 0.0737, "step": 32740 }, { "grad_norm": 0.6478902697563171, "learning_rate": 9.650432917573192e-05, "loss": 0.0661, "step": 32750 }, { "grad_norm": 1.3312478065490723, "learning_rate": 9.650129160897025e-05, "loss": 0.0573, "step": 32760 }, { "grad_norm": 0.6469922065734863, "learning_rate": 9.649825277088073e-05, "loss": 0.0667, "step": 32770 }, { "grad_norm": 0.6160388588905334, "learning_rate": 9.649521266154646e-05, "loss": 0.0756, "step": 32780 }, { "grad_norm": 0.7517526149749756, "learning_rate": 9.649217128105055e-05, "loss": 0.0854, "step": 32790 }, { "grad_norm": 0.8401466012001038, "learning_rate": 9.648912862947613e-05, "loss": 0.0718, "step": 32800 }, { "grad_norm": 0.6613765358924866, "learning_rate": 9.64860847069064e-05, "loss": 0.0707, "step": 32810 }, { "grad_norm": 0.7209888696670532, "learning_rate": 9.648303951342459e-05, "loss": 0.0689, "step": 32820 }, { "grad_norm": 0.6697280406951904, "learning_rate": 9.647999304911392e-05, "loss": 0.0604, "step": 32830 }, { "grad_norm": 0.675656259059906, "learning_rate": 9.647694531405773e-05, "loss": 0.0608, "step": 32840 }, { "grad_norm": 0.5840195417404175, "learning_rate": 9.64738963083393e-05, "loss": 0.0628, "step": 32850 }, { "grad_norm": 0.6436601281166077, "learning_rate": 9.6470846032042e-05, "loss": 0.0607, "step": 32860 }, { "grad_norm": 0.4881269931793213, "learning_rate": 9.646779448524923e-05, "loss": 0.0594, "step": 32870 }, { "grad_norm": 0.6576771140098572, "learning_rate": 9.646474166804442e-05, "loss": 0.0525, "step": 32880 }, { "grad_norm": 0.5173324346542358, "learning_rate": 9.646168758051104e-05, "loss": 0.0512, "step": 32890 }, { "grad_norm": 0.7414448261260986, "learning_rate": 9.645863222273256e-05, "loss": 0.0552, "step": 32900 }, { "grad_norm": 0.489737331867218, "learning_rate": 9.645557559479253e-05, "loss": 0.0461, "step": 32910 }, { "grad_norm": 0.3885616362094879, "learning_rate": 9.645251769677452e-05, "loss": 0.0804, "step": 32920 }, { "grad_norm": 0.4138997197151184, "learning_rate": 9.644945852876213e-05, "loss": 0.0528, "step": 32930 }, { "grad_norm": 0.5025352239608765, "learning_rate": 9.6446398090839e-05, "loss": 0.0587, "step": 32940 }, { "grad_norm": 0.8986068964004517, "learning_rate": 9.644333638308878e-05, "loss": 0.0617, "step": 32950 }, { "grad_norm": 0.6639841794967651, "learning_rate": 9.644027340559519e-05, "loss": 0.0827, "step": 32960 }, { "grad_norm": 0.7110483050346375, "learning_rate": 9.643720915844199e-05, "loss": 0.0546, "step": 32970 }, { "grad_norm": 0.5204966068267822, "learning_rate": 9.643414364171292e-05, "loss": 0.0711, "step": 32980 }, { "grad_norm": 0.46540147066116333, "learning_rate": 9.643107685549182e-05, "loss": 0.0445, "step": 32990 }, { "grad_norm": 0.7943377494812012, "learning_rate": 9.642800879986251e-05, "loss": 0.0594, "step": 33000 }, { "grad_norm": 0.7324014902114868, "learning_rate": 9.642493947490889e-05, "loss": 0.0632, "step": 33010 }, { "grad_norm": 0.7548841834068298, "learning_rate": 9.642186888071486e-05, "loss": 0.0541, "step": 33020 }, { "grad_norm": 0.6519865989685059, "learning_rate": 9.641879701736438e-05, "loss": 0.0621, "step": 33030 }, { "grad_norm": 0.5870213508605957, "learning_rate": 9.641572388494143e-05, "loss": 0.0765, "step": 33040 }, { "grad_norm": 0.7550565004348755, "learning_rate": 9.641264948353001e-05, "loss": 0.0519, "step": 33050 }, { "grad_norm": 0.6224260926246643, "learning_rate": 9.640957381321422e-05, "loss": 0.0768, "step": 33060 }, { "grad_norm": 0.5816712975502014, "learning_rate": 9.64064968740781e-05, "loss": 0.0477, "step": 33070 }, { "grad_norm": 0.7936069369316101, "learning_rate": 9.64034186662058e-05, "loss": 0.0659, "step": 33080 }, { "grad_norm": 0.6855706572532654, "learning_rate": 9.640033918968143e-05, "loss": 0.0546, "step": 33090 }, { "grad_norm": 0.9404505491256714, "learning_rate": 9.639725844458926e-05, "loss": 0.0581, "step": 33100 }, { "grad_norm": 0.5699393153190613, "learning_rate": 9.639417643101345e-05, "loss": 0.0666, "step": 33110 }, { "grad_norm": 0.7635308504104614, "learning_rate": 9.63910931490383e-05, "loss": 0.066, "step": 33120 }, { "grad_norm": 0.6904157400131226, "learning_rate": 9.638800859874808e-05, "loss": 0.0737, "step": 33130 }, { "grad_norm": 0.5325438380241394, "learning_rate": 9.638492278022715e-05, "loss": 0.0701, "step": 33140 }, { "grad_norm": 0.7199145555496216, "learning_rate": 9.638183569355984e-05, "loss": 0.0886, "step": 33150 }, { "grad_norm": 0.7555208802223206, "learning_rate": 9.637874733883057e-05, "loss": 0.0672, "step": 33160 }, { "grad_norm": 0.6273117065429688, "learning_rate": 9.637565771612378e-05, "loss": 0.0502, "step": 33170 }, { "grad_norm": 0.5077223777770996, "learning_rate": 9.637256682552391e-05, "loss": 0.0477, "step": 33180 }, { "grad_norm": 0.6823986172676086, "learning_rate": 9.636947466711551e-05, "loss": 0.0685, "step": 33190 }, { "grad_norm": 0.41138002276420593, "learning_rate": 9.636638124098308e-05, "loss": 0.0491, "step": 33200 }, { "grad_norm": 0.5926910638809204, "learning_rate": 9.636328654721121e-05, "loss": 0.0493, "step": 33210 }, { "grad_norm": 0.6224141716957092, "learning_rate": 9.63601905858845e-05, "loss": 0.0507, "step": 33220 }, { "grad_norm": 0.5816009640693665, "learning_rate": 9.63570933570876e-05, "loss": 0.0596, "step": 33230 }, { "grad_norm": 0.8827009797096252, "learning_rate": 9.635399486090518e-05, "loss": 0.0517, "step": 33240 }, { "grad_norm": 0.438606858253479, "learning_rate": 9.635089509742196e-05, "loss": 0.0425, "step": 33250 }, { "grad_norm": 0.5901252627372742, "learning_rate": 9.634779406672267e-05, "loss": 0.0483, "step": 33260 }, { "grad_norm": 0.6360024213790894, "learning_rate": 9.634469176889211e-05, "loss": 0.063, "step": 33270 }, { "grad_norm": 1.0000663995742798, "learning_rate": 9.634158820401508e-05, "loss": 0.0617, "step": 33280 }, { "grad_norm": 0.5043017864227295, "learning_rate": 9.633848337217646e-05, "loss": 0.0577, "step": 33290 }, { "grad_norm": 0.3866040110588074, "learning_rate": 9.63353772734611e-05, "loss": 0.0529, "step": 33300 }, { "grad_norm": 0.5430019497871399, "learning_rate": 9.633226990795392e-05, "loss": 0.0599, "step": 33310 }, { "grad_norm": 0.41941773891448975, "learning_rate": 9.632916127573989e-05, "loss": 0.0597, "step": 33320 }, { "grad_norm": 0.5719395279884338, "learning_rate": 9.632605137690402e-05, "loss": 0.0569, "step": 33330 }, { "grad_norm": 1.1165251731872559, "learning_rate": 9.63229402115313e-05, "loss": 0.0971, "step": 33340 }, { "grad_norm": 1.3097975254058838, "learning_rate": 9.631982777970678e-05, "loss": 0.0533, "step": 33350 }, { "grad_norm": 0.563979983329773, "learning_rate": 9.631671408151559e-05, "loss": 0.0629, "step": 33360 }, { "grad_norm": 0.5581647157669067, "learning_rate": 9.631359911704281e-05, "loss": 0.066, "step": 33370 }, { "grad_norm": 0.8551870584487915, "learning_rate": 9.631048288637366e-05, "loss": 0.0726, "step": 33380 }, { "grad_norm": 0.7366802096366882, "learning_rate": 9.630736538959328e-05, "loss": 0.0576, "step": 33390 }, { "grad_norm": 0.6200405359268188, "learning_rate": 9.630424662678695e-05, "loss": 0.0664, "step": 33400 }, { "grad_norm": 0.777004063129425, "learning_rate": 9.630112659803991e-05, "loss": 0.0736, "step": 33410 }, { "grad_norm": 0.4099249839782715, "learning_rate": 9.629800530343745e-05, "loss": 0.0551, "step": 33420 }, { "grad_norm": 0.7173053026199341, "learning_rate": 9.629488274306493e-05, "loss": 0.0704, "step": 33430 }, { "grad_norm": 0.8132864236831665, "learning_rate": 9.62917589170077e-05, "loss": 0.0675, "step": 33440 }, { "grad_norm": 0.39239585399627686, "learning_rate": 9.628863382535118e-05, "loss": 0.0579, "step": 33450 }, { "grad_norm": 0.5682808756828308, "learning_rate": 9.628550746818079e-05, "loss": 0.0856, "step": 33460 }, { "grad_norm": 0.5435225367546082, "learning_rate": 9.628237984558203e-05, "loss": 0.0631, "step": 33470 }, { "grad_norm": 0.595609724521637, "learning_rate": 9.627925095764037e-05, "loss": 0.066, "step": 33480 }, { "grad_norm": 0.4474310874938965, "learning_rate": 9.62761208044414e-05, "loss": 0.0593, "step": 33490 }, { "grad_norm": 0.6288859844207764, "learning_rate": 9.627298938607066e-05, "loss": 0.0704, "step": 33500 }, { "grad_norm": 0.9539932012557983, "learning_rate": 9.626985670261377e-05, "loss": 0.0511, "step": 33510 }, { "grad_norm": 0.4070228338241577, "learning_rate": 9.626672275415639e-05, "loss": 0.066, "step": 33520 }, { "grad_norm": 0.7092669010162354, "learning_rate": 9.626358754078418e-05, "loss": 0.07, "step": 33530 }, { "grad_norm": 0.5991176962852478, "learning_rate": 9.626045106258288e-05, "loss": 0.064, "step": 33540 }, { "grad_norm": 0.5514965653419495, "learning_rate": 9.625731331963822e-05, "loss": 0.0545, "step": 33550 }, { "grad_norm": 0.5577341914176941, "learning_rate": 9.6254174312036e-05, "loss": 0.0684, "step": 33560 }, { "grad_norm": 0.5878657102584839, "learning_rate": 9.625103403986202e-05, "loss": 0.0547, "step": 33570 }, { "grad_norm": 0.6168043613433838, "learning_rate": 9.624789250320214e-05, "loss": 0.0563, "step": 33580 }, { "grad_norm": 0.46943771839141846, "learning_rate": 9.624474970214227e-05, "loss": 0.0743, "step": 33590 }, { "grad_norm": 0.47030147910118103, "learning_rate": 9.62416056367683e-05, "loss": 0.0577, "step": 33600 }, { "grad_norm": 0.37931618094444275, "learning_rate": 9.623846030716623e-05, "loss": 0.0475, "step": 33610 }, { "grad_norm": 1.3953425884246826, "learning_rate": 9.623531371342201e-05, "loss": 0.0654, "step": 33620 }, { "grad_norm": 0.4482352137565613, "learning_rate": 9.623216585562167e-05, "loss": 0.0519, "step": 33630 }, { "grad_norm": 0.4899759590625763, "learning_rate": 9.62290167338513e-05, "loss": 0.0473, "step": 33640 }, { "grad_norm": 0.8276827335357666, "learning_rate": 9.622586634819697e-05, "loss": 0.0457, "step": 33650 }, { "grad_norm": 0.41327449679374695, "learning_rate": 9.622271469874484e-05, "loss": 0.0588, "step": 33660 }, { "grad_norm": 0.5098702907562256, "learning_rate": 9.621956178558104e-05, "loss": 0.053, "step": 33670 }, { "grad_norm": 0.6207369565963745, "learning_rate": 9.621640760879179e-05, "loss": 0.074, "step": 33680 }, { "grad_norm": 0.7372641563415527, "learning_rate": 9.621325216846331e-05, "loss": 0.0759, "step": 33690 }, { "grad_norm": 0.5000006556510925, "learning_rate": 9.621009546468188e-05, "loss": 0.0568, "step": 33700 }, { "grad_norm": 0.5604759454727173, "learning_rate": 9.62069374975338e-05, "loss": 0.0547, "step": 33710 }, { "grad_norm": 0.45877358317375183, "learning_rate": 9.620377826710541e-05, "loss": 0.0895, "step": 33720 }, { "grad_norm": 0.46399417519569397, "learning_rate": 9.620061777348308e-05, "loss": 0.0557, "step": 33730 }, { "grad_norm": 0.5872248411178589, "learning_rate": 9.619745601675324e-05, "loss": 0.0516, "step": 33740 }, { "grad_norm": 1.4847171306610107, "learning_rate": 9.619429299700226e-05, "loss": 0.0466, "step": 33750 }, { "grad_norm": 0.7140119671821594, "learning_rate": 9.61911287143167e-05, "loss": 0.0607, "step": 33760 }, { "grad_norm": 0.70365309715271, "learning_rate": 9.618796316878302e-05, "loss": 0.0556, "step": 33770 }, { "grad_norm": 0.4713391661643982, "learning_rate": 9.618479636048778e-05, "loss": 0.0608, "step": 33780 }, { "grad_norm": 0.6295412182807922, "learning_rate": 9.618162828951757e-05, "loss": 0.0582, "step": 33790 }, { "grad_norm": 0.5129905939102173, "learning_rate": 9.617845895595898e-05, "loss": 0.0543, "step": 33800 }, { "grad_norm": 0.7431911826133728, "learning_rate": 9.617528835989867e-05, "loss": 0.0564, "step": 33810 }, { "grad_norm": 0.7487112283706665, "learning_rate": 9.617211650142332e-05, "loss": 0.0573, "step": 33820 }, { "grad_norm": 0.7072293162345886, "learning_rate": 9.616894338061967e-05, "loss": 0.0672, "step": 33830 }, { "grad_norm": 1.5898751020431519, "learning_rate": 9.616576899757444e-05, "loss": 0.0528, "step": 33840 }, { "grad_norm": 0.6831214427947998, "learning_rate": 9.616259335237443e-05, "loss": 0.0626, "step": 33850 }, { "grad_norm": 0.6879709362983704, "learning_rate": 9.615941644510647e-05, "loss": 0.063, "step": 33860 }, { "grad_norm": 0.7727422118186951, "learning_rate": 9.615623827585738e-05, "loss": 0.0568, "step": 33870 }, { "grad_norm": 0.643841028213501, "learning_rate": 9.615305884471411e-05, "loss": 0.0601, "step": 33880 }, { "grad_norm": 0.8230177164077759, "learning_rate": 9.614987815176352e-05, "loss": 0.0691, "step": 33890 }, { "grad_norm": 1.4165444374084473, "learning_rate": 9.614669619709262e-05, "loss": 0.0853, "step": 33900 }, { "grad_norm": 0.8015367984771729, "learning_rate": 9.614351298078836e-05, "loss": 0.0555, "step": 33910 }, { "grad_norm": 0.7744972109794617, "learning_rate": 9.614032850293782e-05, "loss": 0.0651, "step": 33920 }, { "grad_norm": 0.7192060351371765, "learning_rate": 9.613714276362802e-05, "loss": 0.0631, "step": 33930 }, { "grad_norm": 0.7430323958396912, "learning_rate": 9.613395576294606e-05, "loss": 0.0576, "step": 33940 }, { "grad_norm": 1.3279244899749756, "learning_rate": 9.613076750097909e-05, "loss": 0.0498, "step": 33950 }, { "grad_norm": 0.5386964082717896, "learning_rate": 9.612757797781426e-05, "loss": 0.0527, "step": 33960 }, { "grad_norm": 0.94545978307724, "learning_rate": 9.612438719353878e-05, "loss": 0.0655, "step": 33970 }, { "grad_norm": 0.7548125982284546, "learning_rate": 9.612119514823989e-05, "loss": 0.0718, "step": 33980 }, { "grad_norm": 0.47461825609207153, "learning_rate": 9.611800184200484e-05, "loss": 0.0642, "step": 33990 }, { "grad_norm": 1.1475332975387573, "learning_rate": 9.611480727492095e-05, "loss": 0.0716, "step": 34000 }, { "grad_norm": 0.6253029108047485, "learning_rate": 9.611161144707555e-05, "loss": 0.0573, "step": 34010 }, { "grad_norm": 0.7598142027854919, "learning_rate": 9.610841435855603e-05, "loss": 0.0773, "step": 34020 }, { "grad_norm": 0.631420910358429, "learning_rate": 9.610521600944978e-05, "loss": 0.0717, "step": 34030 }, { "grad_norm": 0.68561851978302, "learning_rate": 9.610201639984425e-05, "loss": 0.06, "step": 34040 }, { "grad_norm": 0.6007115840911865, "learning_rate": 9.60988155298269e-05, "loss": 0.0586, "step": 34050 }, { "grad_norm": 0.5899649858474731, "learning_rate": 9.609561339948524e-05, "loss": 0.0514, "step": 34060 }, { "grad_norm": 0.8096126317977905, "learning_rate": 9.609241000890685e-05, "loss": 0.0641, "step": 34070 }, { "grad_norm": 0.5200161337852478, "learning_rate": 9.608920535817927e-05, "loss": 0.0615, "step": 34080 }, { "grad_norm": 0.8158360123634338, "learning_rate": 9.608599944739014e-05, "loss": 0.0477, "step": 34090 }, { "grad_norm": 0.6954914927482605, "learning_rate": 9.608279227662708e-05, "loss": 0.0596, "step": 34100 }, { "grad_norm": 0.6320375204086304, "learning_rate": 9.607958384597782e-05, "loss": 0.053, "step": 34110 }, { "grad_norm": 0.7845332622528076, "learning_rate": 9.607637415553004e-05, "loss": 0.0855, "step": 34120 }, { "grad_norm": 0.5412559509277344, "learning_rate": 9.60731632053715e-05, "loss": 0.0462, "step": 34130 }, { "grad_norm": 0.43659669160842896, "learning_rate": 9.606995099558998e-05, "loss": 0.0565, "step": 34140 }, { "grad_norm": 0.5248022675514221, "learning_rate": 9.606673752627331e-05, "loss": 0.0646, "step": 34150 }, { "grad_norm": 0.6720101237297058, "learning_rate": 9.606352279750934e-05, "loss": 0.0653, "step": 34160 }, { "grad_norm": 0.5740312337875366, "learning_rate": 9.606030680938597e-05, "loss": 0.0566, "step": 34170 }, { "grad_norm": 0.4905911684036255, "learning_rate": 9.60570895619911e-05, "loss": 0.0635, "step": 34180 }, { "grad_norm": 0.5614591836929321, "learning_rate": 9.605387105541272e-05, "loss": 0.0557, "step": 34190 }, { "grad_norm": 0.538003146648407, "learning_rate": 9.605065128973882e-05, "loss": 0.0533, "step": 34200 }, { "grad_norm": 0.5979767441749573, "learning_rate": 9.60474302650574e-05, "loss": 0.0472, "step": 34210 }, { "grad_norm": 0.8557488918304443, "learning_rate": 9.604420798145652e-05, "loss": 0.0692, "step": 34220 }, { "grad_norm": 0.9228284358978271, "learning_rate": 9.604098443902431e-05, "loss": 0.0774, "step": 34230 }, { "grad_norm": 0.6311423778533936, "learning_rate": 9.603775963784888e-05, "loss": 0.0628, "step": 34240 }, { "grad_norm": 0.44636523723602295, "learning_rate": 9.60345335780184e-05, "loss": 0.0535, "step": 34250 }, { "grad_norm": 0.7334602475166321, "learning_rate": 9.603130625962105e-05, "loss": 0.0655, "step": 34260 }, { "grad_norm": 0.5957170128822327, "learning_rate": 9.602807768274509e-05, "loss": 0.056, "step": 34270 }, { "grad_norm": 0.526159405708313, "learning_rate": 9.602484784747879e-05, "loss": 0.0516, "step": 34280 }, { "grad_norm": 0.5958966016769409, "learning_rate": 9.602161675391042e-05, "loss": 0.0621, "step": 34290 }, { "grad_norm": 0.7561966776847839, "learning_rate": 9.601838440212835e-05, "loss": 0.0625, "step": 34300 }, { "grad_norm": 0.4861702620983124, "learning_rate": 9.601515079222094e-05, "loss": 0.0543, "step": 34310 }, { "grad_norm": 0.7962378263473511, "learning_rate": 9.601191592427658e-05, "loss": 0.0651, "step": 34320 }, { "grad_norm": 0.7198780179023743, "learning_rate": 9.600867979838374e-05, "loss": 0.075, "step": 34330 }, { "grad_norm": 0.6613696813583374, "learning_rate": 9.600544241463088e-05, "loss": 0.0693, "step": 34340 }, { "grad_norm": 0.742957592010498, "learning_rate": 9.60022037731065e-05, "loss": 0.0531, "step": 34350 }, { "grad_norm": 1.0293623208999634, "learning_rate": 9.599896387389915e-05, "loss": 0.0621, "step": 34360 }, { "grad_norm": 0.7814838886260986, "learning_rate": 9.599572271709741e-05, "loss": 0.0603, "step": 34370 }, { "grad_norm": 0.7239816188812256, "learning_rate": 9.599248030278989e-05, "loss": 0.055, "step": 34380 }, { "grad_norm": 0.5850377082824707, "learning_rate": 9.598923663106524e-05, "loss": 0.0621, "step": 34390 }, { "grad_norm": 0.5794128775596619, "learning_rate": 9.598599170201213e-05, "loss": 0.0596, "step": 34400 }, { "grad_norm": 0.44510626792907715, "learning_rate": 9.59827455157193e-05, "loss": 0.0501, "step": 34410 }, { "grad_norm": 0.5527411103248596, "learning_rate": 9.597949807227546e-05, "loss": 0.0588, "step": 34420 }, { "grad_norm": 0.5390773415565491, "learning_rate": 9.597624937176943e-05, "loss": 0.0636, "step": 34430 }, { "grad_norm": 0.4777369201183319, "learning_rate": 9.597299941429002e-05, "loss": 0.0451, "step": 34440 }, { "grad_norm": 0.41554245352745056, "learning_rate": 9.596974819992606e-05, "loss": 0.0519, "step": 34450 }, { "grad_norm": 0.566178023815155, "learning_rate": 9.596649572876648e-05, "loss": 0.0532, "step": 34460 }, { "grad_norm": 0.40229758620262146, "learning_rate": 9.596324200090016e-05, "loss": 0.0393, "step": 34470 }, { "grad_norm": 0.6423733830451965, "learning_rate": 9.595998701641606e-05, "loss": 0.0644, "step": 34480 }, { "grad_norm": 0.5026076436042786, "learning_rate": 9.595673077540319e-05, "loss": 0.0903, "step": 34490 }, { "grad_norm": 0.6153425574302673, "learning_rate": 9.595347327795058e-05, "loss": 0.0644, "step": 34500 }, { "grad_norm": 0.5175907015800476, "learning_rate": 9.595021452414727e-05, "loss": 0.0545, "step": 34510 }, { "grad_norm": 0.6515135765075684, "learning_rate": 9.594695451408235e-05, "loss": 0.0565, "step": 34520 }, { "grad_norm": 0.47194474935531616, "learning_rate": 9.594369324784495e-05, "loss": 0.0568, "step": 34530 }, { "grad_norm": 0.6143068671226501, "learning_rate": 9.594043072552425e-05, "loss": 0.0576, "step": 34540 }, { "grad_norm": 0.9912796020507812, "learning_rate": 9.593716694720942e-05, "loss": 0.0603, "step": 34550 }, { "grad_norm": 0.5472392439842224, "learning_rate": 9.59339019129897e-05, "loss": 0.0491, "step": 34560 }, { "grad_norm": 0.4569350779056549, "learning_rate": 9.593063562295437e-05, "loss": 0.0467, "step": 34570 }, { "grad_norm": 0.7115595936775208, "learning_rate": 9.592736807719271e-05, "loss": 0.0788, "step": 34580 }, { "grad_norm": 0.5582985281944275, "learning_rate": 9.592409927579407e-05, "loss": 0.0591, "step": 34590 }, { "grad_norm": 0.5097895264625549, "learning_rate": 9.592082921884782e-05, "loss": 0.0527, "step": 34600 }, { "grad_norm": 0.617218017578125, "learning_rate": 9.591755790644332e-05, "loss": 0.0649, "step": 34610 }, { "grad_norm": 0.45066431164741516, "learning_rate": 9.591428533867007e-05, "loss": 0.0552, "step": 34620 }, { "grad_norm": 0.5496455430984497, "learning_rate": 9.591101151561748e-05, "loss": 0.0397, "step": 34630 }, { "grad_norm": 0.5856770277023315, "learning_rate": 9.59077364373751e-05, "loss": 0.0502, "step": 34640 }, { "grad_norm": 0.7728607654571533, "learning_rate": 9.590446010403245e-05, "loss": 0.0498, "step": 34650 }, { "grad_norm": 0.4661693871021271, "learning_rate": 9.59011825156791e-05, "loss": 0.0473, "step": 34660 }, { "grad_norm": 0.669337809085846, "learning_rate": 9.589790367240465e-05, "loss": 0.062, "step": 34670 }, { "grad_norm": 0.7149866223335266, "learning_rate": 9.589462357429878e-05, "loss": 0.052, "step": 34680 }, { "grad_norm": 0.5191816091537476, "learning_rate": 9.589134222145114e-05, "loss": 0.0582, "step": 34690 }, { "grad_norm": 0.5471221208572388, "learning_rate": 9.588805961395144e-05, "loss": 0.0521, "step": 34700 }, { "grad_norm": 0.6187089085578918, "learning_rate": 9.588477575188943e-05, "loss": 0.0526, "step": 34710 }, { "grad_norm": 0.6353569030761719, "learning_rate": 9.58814906353549e-05, "loss": 0.0659, "step": 34720 }, { "grad_norm": 0.7723108530044556, "learning_rate": 9.587820426443763e-05, "loss": 0.0586, "step": 34730 }, { "grad_norm": 0.7946750521659851, "learning_rate": 9.58749166392275e-05, "loss": 0.0646, "step": 34740 }, { "grad_norm": 0.47253772616386414, "learning_rate": 9.587162775981439e-05, "loss": 0.0615, "step": 34750 }, { "grad_norm": 0.5871708393096924, "learning_rate": 9.586833762628821e-05, "loss": 0.0488, "step": 34760 }, { "grad_norm": 0.6637250185012817, "learning_rate": 9.58650462387389e-05, "loss": 0.0678, "step": 34770 }, { "grad_norm": 0.36017414927482605, "learning_rate": 9.586175359725648e-05, "loss": 0.0537, "step": 34780 }, { "grad_norm": 0.6303892135620117, "learning_rate": 9.585845970193093e-05, "loss": 0.0689, "step": 34790 }, { "grad_norm": 0.7828267812728882, "learning_rate": 9.585516455285231e-05, "loss": 0.0705, "step": 34800 }, { "grad_norm": 0.551641583442688, "learning_rate": 9.585186815011073e-05, "loss": 0.0567, "step": 34810 }, { "grad_norm": 0.5641936659812927, "learning_rate": 9.584857049379631e-05, "loss": 0.0487, "step": 34820 }, { "grad_norm": 0.5180937647819519, "learning_rate": 9.584527158399919e-05, "loss": 0.0609, "step": 34830 }, { "grad_norm": 0.6456660628318787, "learning_rate": 9.584197142080955e-05, "loss": 0.066, "step": 34840 }, { "grad_norm": 0.7237144112586975, "learning_rate": 9.583867000431765e-05, "loss": 0.0561, "step": 34850 }, { "grad_norm": 0.6417388916015625, "learning_rate": 9.583536733461373e-05, "loss": 0.0579, "step": 34860 }, { "grad_norm": 0.5884639620780945, "learning_rate": 9.58320634117881e-05, "loss": 0.0564, "step": 34870 }, { "grad_norm": 0.6090008616447449, "learning_rate": 9.582875823593106e-05, "loss": 0.0719, "step": 34880 }, { "grad_norm": 0.6372964382171631, "learning_rate": 9.582545180713298e-05, "loss": 0.0622, "step": 34890 }, { "grad_norm": 0.5928954482078552, "learning_rate": 9.582214412548428e-05, "loss": 0.0663, "step": 34900 }, { "grad_norm": 0.7613247632980347, "learning_rate": 9.581883519107538e-05, "loss": 0.0498, "step": 34910 }, { "grad_norm": 0.5658698081970215, "learning_rate": 9.581552500399671e-05, "loss": 0.0603, "step": 34920 }, { "grad_norm": 0.5641582608222961, "learning_rate": 9.581221356433882e-05, "loss": 0.0617, "step": 34930 }, { "grad_norm": 0.5806121826171875, "learning_rate": 9.580890087219221e-05, "loss": 0.0475, "step": 34940 }, { "grad_norm": 0.7180017828941345, "learning_rate": 9.580558692764747e-05, "loss": 0.0756, "step": 34950 }, { "grad_norm": 0.6661330461502075, "learning_rate": 9.580227173079519e-05, "loss": 0.0611, "step": 34960 }, { "grad_norm": 0.5046491026878357, "learning_rate": 9.5798955281726e-05, "loss": 0.0594, "step": 34970 }, { "grad_norm": 0.5065561532974243, "learning_rate": 9.579563758053058e-05, "loss": 0.0541, "step": 34980 }, { "grad_norm": 0.5238633155822754, "learning_rate": 9.579231862729965e-05, "loss": 0.0572, "step": 34990 }, { "grad_norm": 0.5719344615936279, "learning_rate": 9.578899842212392e-05, "loss": 0.0504, "step": 35000 }, { "grad_norm": 0.5957613587379456, "learning_rate": 9.578567696509418e-05, "loss": 0.0589, "step": 35010 }, { "grad_norm": 0.5978618860244751, "learning_rate": 9.578235425630123e-05, "loss": 0.0589, "step": 35020 }, { "grad_norm": 0.5111650228500366, "learning_rate": 9.57790302958359e-05, "loss": 0.059, "step": 35030 }, { "grad_norm": 0.4753875434398651, "learning_rate": 9.577570508378911e-05, "loss": 0.068, "step": 35040 }, { "grad_norm": 0.47226375341415405, "learning_rate": 9.57723786202517e-05, "loss": 0.0617, "step": 35050 }, { "grad_norm": 0.6676758527755737, "learning_rate": 9.576905090531468e-05, "loss": 0.0545, "step": 35060 }, { "grad_norm": 0.6088489294052124, "learning_rate": 9.576572193906901e-05, "loss": 0.0915, "step": 35070 }, { "grad_norm": 0.5754150748252869, "learning_rate": 9.576239172160568e-05, "loss": 0.0537, "step": 35080 }, { "grad_norm": 0.5704752206802368, "learning_rate": 9.575906025301575e-05, "loss": 0.0543, "step": 35090 }, { "grad_norm": 0.4135075807571411, "learning_rate": 9.575572753339032e-05, "loss": 0.0638, "step": 35100 }, { "grad_norm": 0.621456503868103, "learning_rate": 9.575239356282049e-05, "loss": 0.0746, "step": 35110 }, { "grad_norm": 0.511509358882904, "learning_rate": 9.57490583413974e-05, "loss": 0.0693, "step": 35120 }, { "grad_norm": 0.7872906923294067, "learning_rate": 9.574572186921222e-05, "loss": 0.0673, "step": 35130 }, { "grad_norm": 0.6182281374931335, "learning_rate": 9.574238414635621e-05, "loss": 0.075, "step": 35140 }, { "grad_norm": 0.3915617763996124, "learning_rate": 9.57390451729206e-05, "loss": 0.0627, "step": 35150 }, { "grad_norm": 0.6185944080352783, "learning_rate": 9.573570494899668e-05, "loss": 0.0661, "step": 35160 }, { "grad_norm": 1.0114463567733765, "learning_rate": 9.573236347467578e-05, "loss": 0.0635, "step": 35170 }, { "grad_norm": 0.7651988863945007, "learning_rate": 9.572902075004924e-05, "loss": 0.0624, "step": 35180 }, { "grad_norm": 0.5606789588928223, "learning_rate": 9.572567677520843e-05, "loss": 0.056, "step": 35190 }, { "grad_norm": 0.7031106948852539, "learning_rate": 9.572233155024482e-05, "loss": 0.0884, "step": 35200 }, { "grad_norm": 0.7114612460136414, "learning_rate": 9.571898507524984e-05, "loss": 0.0808, "step": 35210 }, { "grad_norm": 0.547363817691803, "learning_rate": 9.571563735031497e-05, "loss": 0.0606, "step": 35220 }, { "grad_norm": 0.7761225700378418, "learning_rate": 9.571228837553177e-05, "loss": 0.0575, "step": 35230 }, { "grad_norm": 0.6200740337371826, "learning_rate": 9.570893815099176e-05, "loss": 0.0695, "step": 35240 }, { "grad_norm": 0.5202545523643494, "learning_rate": 9.570558667678656e-05, "loss": 0.0545, "step": 35250 }, { "grad_norm": 0.6876357793807983, "learning_rate": 9.570223395300779e-05, "loss": 0.0623, "step": 35260 }, { "grad_norm": 0.6751470565795898, "learning_rate": 9.569887997974712e-05, "loss": 0.0644, "step": 35270 }, { "grad_norm": 0.48971542716026306, "learning_rate": 9.569552475709623e-05, "loss": 0.0564, "step": 35280 }, { "grad_norm": 0.4747253656387329, "learning_rate": 9.569216828514685e-05, "loss": 0.0487, "step": 35290 }, { "grad_norm": 0.49158668518066406, "learning_rate": 9.568881056399077e-05, "loss": 0.05, "step": 35300 }, { "grad_norm": 0.5603686571121216, "learning_rate": 9.568545159371978e-05, "loss": 0.0453, "step": 35310 }, { "grad_norm": 0.4552099406719208, "learning_rate": 9.56820913744257e-05, "loss": 0.0923, "step": 35320 }, { "grad_norm": 0.48779556155204773, "learning_rate": 9.56787299062004e-05, "loss": 0.0608, "step": 35330 }, { "grad_norm": 0.6944969892501831, "learning_rate": 9.567536718913578e-05, "loss": 0.0758, "step": 35340 }, { "grad_norm": 0.4469917118549347, "learning_rate": 9.567200322332379e-05, "loss": 0.0785, "step": 35350 }, { "grad_norm": 0.7723975777626038, "learning_rate": 9.566863800885638e-05, "loss": 0.0607, "step": 35360 }, { "grad_norm": 0.3879420757293701, "learning_rate": 9.566527154582558e-05, "loss": 0.0657, "step": 35370 }, { "grad_norm": 0.5794047117233276, "learning_rate": 9.56619038343234e-05, "loss": 0.0626, "step": 35380 }, { "grad_norm": 0.7735660076141357, "learning_rate": 9.565853487444192e-05, "loss": 0.0744, "step": 35390 }, { "grad_norm": 0.703999400138855, "learning_rate": 9.565516466627327e-05, "loss": 0.0518, "step": 35400 }, { "grad_norm": 0.8085862398147583, "learning_rate": 9.565179320990955e-05, "loss": 0.0635, "step": 35410 }, { "grad_norm": 0.7662864327430725, "learning_rate": 9.564842050544296e-05, "loss": 0.0686, "step": 35420 }, { "grad_norm": 0.6232629418373108, "learning_rate": 9.564504655296569e-05, "loss": 0.0577, "step": 35430 }, { "grad_norm": 0.523350179195404, "learning_rate": 9.564167135257001e-05, "loss": 0.0645, "step": 35440 }, { "grad_norm": 0.5771271586418152, "learning_rate": 9.56382949043482e-05, "loss": 0.0524, "step": 35450 }, { "grad_norm": 0.48224544525146484, "learning_rate": 9.563491720839253e-05, "loss": 0.0516, "step": 35460 }, { "grad_norm": 0.5241501331329346, "learning_rate": 9.563153826479536e-05, "loss": 0.0562, "step": 35470 }, { "grad_norm": 0.5425935387611389, "learning_rate": 9.562815807364909e-05, "loss": 0.0847, "step": 35480 }, { "grad_norm": 0.7321894764900208, "learning_rate": 9.562477663504611e-05, "loss": 0.0764, "step": 35490 }, { "grad_norm": 0.5121670365333557, "learning_rate": 9.562139394907888e-05, "loss": 0.0908, "step": 35500 }, { "grad_norm": 0.9676214456558228, "learning_rate": 9.56180100158399e-05, "loss": 0.0635, "step": 35510 }, { "grad_norm": 0.7217220067977905, "learning_rate": 9.561462483542164e-05, "loss": 0.075, "step": 35520 }, { "grad_norm": 0.7224625945091248, "learning_rate": 9.561123840791668e-05, "loss": 0.0678, "step": 35530 }, { "grad_norm": 1.8555511236190796, "learning_rate": 9.56078507334176e-05, "loss": 0.0623, "step": 35540 }, { "grad_norm": 0.5543866753578186, "learning_rate": 9.560446181201701e-05, "loss": 0.0576, "step": 35550 }, { "grad_norm": 0.7701137065887451, "learning_rate": 9.560107164380757e-05, "loss": 0.0889, "step": 35560 }, { "grad_norm": 0.4727461636066437, "learning_rate": 9.559768022888196e-05, "loss": 0.0491, "step": 35570 }, { "grad_norm": 0.5501593947410583, "learning_rate": 9.559428756733291e-05, "loss": 0.0591, "step": 35580 }, { "grad_norm": 0.6300615072250366, "learning_rate": 9.559089365925315e-05, "loss": 0.0525, "step": 35590 }, { "grad_norm": 1.1746891736984253, "learning_rate": 9.55874985047355e-05, "loss": 0.0624, "step": 35600 }, { "grad_norm": 0.4518706500530243, "learning_rate": 9.558410210387277e-05, "loss": 0.0547, "step": 35610 }, { "grad_norm": 0.7208671569824219, "learning_rate": 9.558070445675782e-05, "loss": 0.0732, "step": 35620 }, { "grad_norm": 0.43586719036102295, "learning_rate": 9.557730556348352e-05, "loss": 0.0465, "step": 35630 }, { "grad_norm": 0.9925081133842468, "learning_rate": 9.557390542414281e-05, "loss": 0.0681, "step": 35640 }, { "grad_norm": 1.6742511987686157, "learning_rate": 9.557050403882864e-05, "loss": 0.0705, "step": 35650 }, { "grad_norm": 0.7457919716835022, "learning_rate": 9.556710140763401e-05, "loss": 0.061, "step": 35660 }, { "grad_norm": 0.7037273645401001, "learning_rate": 9.556369753065197e-05, "loss": 0.0606, "step": 35670 }, { "grad_norm": 0.5833866000175476, "learning_rate": 9.556029240797553e-05, "loss": 0.0565, "step": 35680 }, { "grad_norm": 0.49441996216773987, "learning_rate": 9.555688603969783e-05, "loss": 0.0591, "step": 35690 }, { "grad_norm": 0.7227568626403809, "learning_rate": 9.555347842591198e-05, "loss": 0.0614, "step": 35700 }, { "grad_norm": 0.8251500725746155, "learning_rate": 9.555006956671113e-05, "loss": 0.0611, "step": 35710 }, { "grad_norm": 0.6895546913146973, "learning_rate": 9.55466594621885e-05, "loss": 0.0617, "step": 35720 }, { "grad_norm": 0.6439470648765564, "learning_rate": 9.554324811243731e-05, "loss": 0.0584, "step": 35730 }, { "grad_norm": 0.7889333963394165, "learning_rate": 9.553983551755081e-05, "loss": 0.0764, "step": 35740 }, { "grad_norm": 0.5258845090866089, "learning_rate": 9.553642167762233e-05, "loss": 0.0524, "step": 35750 }, { "grad_norm": 0.7384043335914612, "learning_rate": 9.55330065927452e-05, "loss": 0.0672, "step": 35760 }, { "grad_norm": 0.6059176325798035, "learning_rate": 9.552959026301276e-05, "loss": 0.0464, "step": 35770 }, { "grad_norm": 0.7376211285591125, "learning_rate": 9.552617268851844e-05, "loss": 0.0565, "step": 35780 }, { "grad_norm": 0.735230028629303, "learning_rate": 9.552275386935565e-05, "loss": 0.0545, "step": 35790 }, { "grad_norm": 0.9456732869148254, "learning_rate": 9.551933380561788e-05, "loss": 0.0526, "step": 35800 }, { "grad_norm": 0.9371166825294495, "learning_rate": 9.551591249739862e-05, "loss": 0.0588, "step": 35810 }, { "grad_norm": 0.8507806658744812, "learning_rate": 9.55124899447914e-05, "loss": 0.071, "step": 35820 }, { "grad_norm": 0.5291924476623535, "learning_rate": 9.550906614788982e-05, "loss": 0.0585, "step": 35830 }, { "grad_norm": 0.5582579970359802, "learning_rate": 9.550564110678746e-05, "loss": 0.0601, "step": 35840 }, { "grad_norm": 0.4796678423881531, "learning_rate": 9.550221482157797e-05, "loss": 0.0536, "step": 35850 }, { "grad_norm": 0.5581106543540955, "learning_rate": 9.549878729235502e-05, "loss": 0.0596, "step": 35860 }, { "grad_norm": 0.540395200252533, "learning_rate": 9.549535851921232e-05, "loss": 0.0496, "step": 35870 }, { "grad_norm": 0.43099936842918396, "learning_rate": 9.549192850224362e-05, "loss": 0.0455, "step": 35880 }, { "grad_norm": 0.6063694953918457, "learning_rate": 9.548849724154268e-05, "loss": 0.0837, "step": 35890 }, { "grad_norm": 0.6887842416763306, "learning_rate": 9.54850647372033e-05, "loss": 0.0472, "step": 35900 }, { "grad_norm": 0.6853367686271667, "learning_rate": 9.548163098931935e-05, "loss": 0.0609, "step": 35910 }, { "grad_norm": 0.5263974666595459, "learning_rate": 9.547819599798468e-05, "loss": 0.0619, "step": 35920 }, { "grad_norm": 2.1797547340393066, "learning_rate": 9.547475976329325e-05, "loss": 0.0631, "step": 35930 }, { "grad_norm": 0.9396882057189941, "learning_rate": 9.547132228533895e-05, "loss": 0.0726, "step": 35940 }, { "grad_norm": 0.6289934515953064, "learning_rate": 9.546788356421578e-05, "loss": 0.0739, "step": 35950 }, { "grad_norm": 0.6170966625213623, "learning_rate": 9.546444360001775e-05, "loss": 0.0641, "step": 35960 }, { "grad_norm": 0.7872222661972046, "learning_rate": 9.546100239283892e-05, "loss": 0.0517, "step": 35970 }, { "grad_norm": 0.76799476146698, "learning_rate": 9.545755994277337e-05, "loss": 0.0582, "step": 35980 }, { "grad_norm": 0.4540897309780121, "learning_rate": 9.545411624991521e-05, "loss": 0.0723, "step": 35990 }, { "grad_norm": 0.7192868590354919, "learning_rate": 9.545067131435858e-05, "loss": 0.0469, "step": 36000 }, { "grad_norm": 0.8165702819824219, "learning_rate": 9.544722513619769e-05, "loss": 0.0698, "step": 36010 }, { "grad_norm": 0.4133433997631073, "learning_rate": 9.54437777155267e-05, "loss": 0.0889, "step": 36020 }, { "grad_norm": 0.3982866108417511, "learning_rate": 9.544032905243994e-05, "loss": 0.0521, "step": 36030 }, { "grad_norm": 0.7274662256240845, "learning_rate": 9.543687914703162e-05, "loss": 0.055, "step": 36040 }, { "grad_norm": 0.6162336468696594, "learning_rate": 9.543342799939612e-05, "loss": 0.0743, "step": 36050 }, { "grad_norm": 0.5587760806083679, "learning_rate": 9.542997560962775e-05, "loss": 0.0477, "step": 36060 }, { "grad_norm": 0.5024696588516235, "learning_rate": 9.542652197782092e-05, "loss": 0.0697, "step": 36070 }, { "grad_norm": 0.645382285118103, "learning_rate": 9.542306710407005e-05, "loss": 0.0605, "step": 36080 }, { "grad_norm": 0.4964522123336792, "learning_rate": 9.541961098846959e-05, "loss": 0.0619, "step": 36090 }, { "grad_norm": 0.6891692876815796, "learning_rate": 9.541615363111402e-05, "loss": 0.0567, "step": 36100 }, { "grad_norm": 0.4780958592891693, "learning_rate": 9.541269503209787e-05, "loss": 0.0583, "step": 36110 }, { "grad_norm": 0.6397230625152588, "learning_rate": 9.540923519151571e-05, "loss": 0.0555, "step": 36120 }, { "grad_norm": 0.6355982422828674, "learning_rate": 9.540577410946212e-05, "loss": 0.0546, "step": 36130 }, { "grad_norm": 0.5540558695793152, "learning_rate": 9.540231178603172e-05, "loss": 0.0547, "step": 36140 }, { "grad_norm": 0.7409144639968872, "learning_rate": 9.539884822131918e-05, "loss": 0.062, "step": 36150 }, { "grad_norm": 0.7194812297821045, "learning_rate": 9.539538341541916e-05, "loss": 0.0569, "step": 36160 }, { "grad_norm": 0.4614640176296234, "learning_rate": 9.539191736842643e-05, "loss": 0.0695, "step": 36170 }, { "grad_norm": 0.639335036277771, "learning_rate": 9.538845008043573e-05, "loss": 0.0768, "step": 36180 }, { "grad_norm": 0.8373674750328064, "learning_rate": 9.538498155154185e-05, "loss": 0.0726, "step": 36190 }, { "grad_norm": 0.49775201082229614, "learning_rate": 9.538151178183965e-05, "loss": 0.0738, "step": 36200 }, { "grad_norm": 0.48877570033073425, "learning_rate": 9.537804077142394e-05, "loss": 0.0593, "step": 36210 }, { "grad_norm": 0.9701409935951233, "learning_rate": 9.537456852038965e-05, "loss": 0.0675, "step": 36220 }, { "grad_norm": 0.5745044946670532, "learning_rate": 9.53710950288317e-05, "loss": 0.0536, "step": 36230 }, { "grad_norm": 0.5877317190170288, "learning_rate": 9.536762029684507e-05, "loss": 0.0575, "step": 36240 }, { "grad_norm": 0.5055842399597168, "learning_rate": 9.536414432452473e-05, "loss": 0.0522, "step": 36250 }, { "grad_norm": 0.9660356640815735, "learning_rate": 9.536066711196573e-05, "loss": 0.0562, "step": 36260 }, { "grad_norm": 0.5875626802444458, "learning_rate": 9.535718865926314e-05, "loss": 0.0517, "step": 36270 }, { "grad_norm": 0.5082061290740967, "learning_rate": 9.535370896651205e-05, "loss": 0.0514, "step": 36280 }, { "grad_norm": 0.5363762974739075, "learning_rate": 9.535022803380759e-05, "loss": 0.0547, "step": 36290 }, { "grad_norm": 0.770366907119751, "learning_rate": 9.534674586124494e-05, "loss": 0.0673, "step": 36300 }, { "grad_norm": 0.9021984338760376, "learning_rate": 9.534326244891928e-05, "loss": 0.0596, "step": 36310 }, { "grad_norm": 0.7615953087806702, "learning_rate": 9.533977779692588e-05, "loss": 0.051, "step": 36320 }, { "grad_norm": 0.7518920302391052, "learning_rate": 9.533629190535997e-05, "loss": 0.0591, "step": 36330 }, { "grad_norm": 0.5663995742797852, "learning_rate": 9.533280477431688e-05, "loss": 0.0495, "step": 36340 }, { "grad_norm": 0.6978169679641724, "learning_rate": 9.532931640389193e-05, "loss": 0.0592, "step": 36350 }, { "grad_norm": 0.6534620523452759, "learning_rate": 9.532582679418052e-05, "loss": 0.0538, "step": 36360 }, { "grad_norm": 0.7007088661193848, "learning_rate": 9.532233594527802e-05, "loss": 0.055, "step": 36370 }, { "grad_norm": 0.583395779132843, "learning_rate": 9.531884385727987e-05, "loss": 0.0659, "step": 36380 }, { "grad_norm": 0.5092915296554565, "learning_rate": 9.531535053028158e-05, "loss": 0.0492, "step": 36390 }, { "grad_norm": 0.5218465328216553, "learning_rate": 9.531185596437861e-05, "loss": 0.0441, "step": 36400 }, { "grad_norm": 0.8620935678482056, "learning_rate": 9.530836015966654e-05, "loss": 0.0575, "step": 36410 }, { "grad_norm": 0.5291378498077393, "learning_rate": 9.53048631162409e-05, "loss": 0.0545, "step": 36420 }, { "grad_norm": 0.7601668238639832, "learning_rate": 9.530136483419733e-05, "loss": 0.0485, "step": 36430 }, { "grad_norm": 0.4559272527694702, "learning_rate": 9.529786531363148e-05, "loss": 0.0439, "step": 36440 }, { "grad_norm": 0.5070050358772278, "learning_rate": 9.5294364554639e-05, "loss": 0.0589, "step": 36450 }, { "grad_norm": 0.6178658604621887, "learning_rate": 9.52908625573156e-05, "loss": 0.0495, "step": 36460 }, { "grad_norm": 0.6992598176002502, "learning_rate": 9.528735932175704e-05, "loss": 0.0591, "step": 36470 }, { "grad_norm": 0.6441715955734253, "learning_rate": 9.52838548480591e-05, "loss": 0.0508, "step": 36480 }, { "grad_norm": 0.5467044711112976, "learning_rate": 9.528034913631756e-05, "loss": 0.0671, "step": 36490 }, { "grad_norm": 0.8629640936851501, "learning_rate": 9.527684218662831e-05, "loss": 0.0615, "step": 36500 }, { "grad_norm": 0.5836175680160522, "learning_rate": 9.527333399908718e-05, "loss": 0.0479, "step": 36510 }, { "grad_norm": 0.3558776378631592, "learning_rate": 9.526982457379012e-05, "loss": 0.0383, "step": 36520 }, { "grad_norm": 0.5004429817199707, "learning_rate": 9.526631391083307e-05, "loss": 0.062, "step": 36530 }, { "grad_norm": 0.7711041569709778, "learning_rate": 9.526280201031199e-05, "loss": 0.0452, "step": 36540 }, { "grad_norm": 0.5966451168060303, "learning_rate": 9.525928887232292e-05, "loss": 0.0481, "step": 36550 }, { "grad_norm": 0.7326532602310181, "learning_rate": 9.525577449696188e-05, "loss": 0.0495, "step": 36560 }, { "grad_norm": 0.38659465312957764, "learning_rate": 9.525225888432499e-05, "loss": 0.0538, "step": 36570 }, { "grad_norm": 0.6453984975814819, "learning_rate": 9.524874203450832e-05, "loss": 0.0513, "step": 36580 }, { "grad_norm": 0.5612689852714539, "learning_rate": 9.524522394760804e-05, "loss": 0.0464, "step": 36590 }, { "grad_norm": 0.601192831993103, "learning_rate": 9.524170462372035e-05, "loss": 0.0579, "step": 36600 }, { "grad_norm": 0.9297481775283813, "learning_rate": 9.523818406294145e-05, "loss": 0.0434, "step": 36610 }, { "grad_norm": 0.4310872554779053, "learning_rate": 9.523466226536759e-05, "loss": 0.0598, "step": 36620 }, { "grad_norm": 0.4843086898326874, "learning_rate": 9.523113923109506e-05, "loss": 0.0463, "step": 36630 }, { "grad_norm": 0.8166423439979553, "learning_rate": 9.522761496022017e-05, "loss": 0.0676, "step": 36640 }, { "grad_norm": 0.5627711415290833, "learning_rate": 9.52240894528393e-05, "loss": 0.0542, "step": 36650 }, { "grad_norm": 0.795179545879364, "learning_rate": 9.522056270904878e-05, "loss": 0.0563, "step": 36660 }, { "grad_norm": 0.8804706335067749, "learning_rate": 9.521703472894508e-05, "loss": 0.0563, "step": 36670 }, { "grad_norm": 0.6446207165718079, "learning_rate": 9.521350551262465e-05, "loss": 0.0633, "step": 36680 }, { "grad_norm": 0.7416960000991821, "learning_rate": 9.520997506018394e-05, "loss": 0.0581, "step": 36690 }, { "grad_norm": 0.5611290335655212, "learning_rate": 9.520644337171952e-05, "loss": 0.0503, "step": 36700 }, { "grad_norm": 0.4098645746707916, "learning_rate": 9.52029104473279e-05, "loss": 0.0625, "step": 36710 }, { "grad_norm": 0.7586246132850647, "learning_rate": 9.519937628710572e-05, "loss": 0.0483, "step": 36720 }, { "grad_norm": 0.3987685739994049, "learning_rate": 9.519584089114954e-05, "loss": 0.0497, "step": 36730 }, { "grad_norm": 0.5320476293563843, "learning_rate": 9.519230425955606e-05, "loss": 0.0469, "step": 36740 }, { "grad_norm": 0.5232864618301392, "learning_rate": 9.518876639242197e-05, "loss": 0.0563, "step": 36750 }, { "grad_norm": 0.56733638048172, "learning_rate": 9.518522728984399e-05, "loss": 0.0463, "step": 36760 }, { "grad_norm": 0.42975157499313354, "learning_rate": 9.518168695191885e-05, "loss": 0.0574, "step": 36770 }, { "grad_norm": 0.736921489238739, "learning_rate": 9.517814537874336e-05, "loss": 0.0496, "step": 36780 }, { "grad_norm": 0.7560990452766418, "learning_rate": 9.517460257041436e-05, "loss": 0.0549, "step": 36790 }, { "grad_norm": 0.5681348443031311, "learning_rate": 9.51710585270287e-05, "loss": 0.0702, "step": 36800 }, { "grad_norm": 0.7463192939758301, "learning_rate": 9.516751324868326e-05, "loss": 0.0658, "step": 36810 }, { "grad_norm": 0.6103019118309021, "learning_rate": 9.516396673547498e-05, "loss": 0.0783, "step": 36820 }, { "grad_norm": 0.7839944958686829, "learning_rate": 9.516041898750083e-05, "loss": 0.0686, "step": 36830 }, { "grad_norm": 0.5150975584983826, "learning_rate": 9.515687000485778e-05, "loss": 0.0682, "step": 36840 }, { "grad_norm": 0.5802522897720337, "learning_rate": 9.515331978764286e-05, "loss": 0.0625, "step": 36850 }, { "grad_norm": 0.6339281797409058, "learning_rate": 9.514976833595315e-05, "loss": 0.0605, "step": 36860 }, { "grad_norm": 0.6255648732185364, "learning_rate": 9.514621564988573e-05, "loss": 0.0541, "step": 36870 }, { "grad_norm": 0.6376101970672607, "learning_rate": 9.514266172953775e-05, "loss": 0.0572, "step": 36880 }, { "grad_norm": 0.6679726243019104, "learning_rate": 9.513910657500635e-05, "loss": 0.0444, "step": 36890 }, { "grad_norm": 0.43909451365470886, "learning_rate": 9.513555018638872e-05, "loss": 0.0689, "step": 36900 }, { "grad_norm": 0.48520803451538086, "learning_rate": 9.513199256378211e-05, "loss": 0.0422, "step": 36910 }, { "grad_norm": 0.5984159111976624, "learning_rate": 9.512843370728379e-05, "loss": 0.0654, "step": 36920 }, { "grad_norm": 0.7208818793296814, "learning_rate": 9.512487361699104e-05, "loss": 0.0612, "step": 36930 }, { "grad_norm": 0.4166052043437958, "learning_rate": 9.51213122930012e-05, "loss": 0.0652, "step": 36940 }, { "grad_norm": 0.5830631256103516, "learning_rate": 9.511774973541163e-05, "loss": 0.0597, "step": 36950 }, { "grad_norm": 0.5853204131126404, "learning_rate": 9.511418594431974e-05, "loss": 0.0442, "step": 36960 }, { "grad_norm": 0.6588292121887207, "learning_rate": 9.511062091982295e-05, "loss": 0.0451, "step": 36970 }, { "grad_norm": 0.9272167086601257, "learning_rate": 9.510705466201872e-05, "loss": 0.047, "step": 36980 }, { "grad_norm": 1.0275229215621948, "learning_rate": 9.510348717100457e-05, "loss": 0.067, "step": 36990 }, { "grad_norm": 0.48330816626548767, "learning_rate": 9.509991844687803e-05, "loss": 0.0511, "step": 37000 }, { "grad_norm": 0.5097610354423523, "learning_rate": 9.509634848973666e-05, "loss": 0.0583, "step": 37010 }, { "grad_norm": 0.6685498356819153, "learning_rate": 9.509277729967808e-05, "loss": 0.0474, "step": 37020 }, { "grad_norm": 0.5174744129180908, "learning_rate": 9.508920487679989e-05, "loss": 0.0544, "step": 37030 }, { "grad_norm": 0.5607062578201294, "learning_rate": 9.508563122119978e-05, "loss": 0.0681, "step": 37040 }, { "grad_norm": 0.6219741702079773, "learning_rate": 9.508205633297544e-05, "loss": 0.0474, "step": 37050 }, { "grad_norm": 0.5699092149734497, "learning_rate": 9.507848021222463e-05, "loss": 0.0629, "step": 37060 }, { "grad_norm": 0.4215136468410492, "learning_rate": 9.50749028590451e-05, "loss": 0.0492, "step": 37070 }, { "grad_norm": 0.7292858362197876, "learning_rate": 9.507132427353467e-05, "loss": 0.0517, "step": 37080 }, { "grad_norm": 0.7326087355613708, "learning_rate": 9.506774445579116e-05, "loss": 0.0545, "step": 37090 }, { "grad_norm": 0.5252734422683716, "learning_rate": 9.506416340591244e-05, "loss": 0.0507, "step": 37100 }, { "grad_norm": 0.5076687335968018, "learning_rate": 9.506058112399643e-05, "loss": 0.0623, "step": 37110 }, { "grad_norm": 0.45294472575187683, "learning_rate": 9.505699761014107e-05, "loss": 0.0762, "step": 37120 }, { "grad_norm": 0.504206657409668, "learning_rate": 9.505341286444432e-05, "loss": 0.0424, "step": 37130 }, { "grad_norm": 0.701120138168335, "learning_rate": 9.504982688700417e-05, "loss": 0.0546, "step": 37140 }, { "grad_norm": 0.5259209275245667, "learning_rate": 9.504623967791869e-05, "loss": 0.0447, "step": 37150 }, { "grad_norm": 0.5923762917518616, "learning_rate": 9.504265123728594e-05, "loss": 0.0531, "step": 37160 }, { "grad_norm": 0.6228054165840149, "learning_rate": 9.503906156520403e-05, "loss": 0.0497, "step": 37170 }, { "grad_norm": 0.6041321158409119, "learning_rate": 9.503547066177109e-05, "loss": 0.0526, "step": 37180 }, { "grad_norm": 0.9562049508094788, "learning_rate": 9.50318785270853e-05, "loss": 0.0797, "step": 37190 }, { "grad_norm": 0.6495671272277832, "learning_rate": 9.502828516124488e-05, "loss": 0.066, "step": 37200 }, { "grad_norm": 0.9276990294456482, "learning_rate": 9.502469056434805e-05, "loss": 0.0648, "step": 37210 }, { "grad_norm": 0.5979134440422058, "learning_rate": 9.50210947364931e-05, "loss": 0.0478, "step": 37220 }, { "grad_norm": 0.8627495765686035, "learning_rate": 9.501749767777832e-05, "loss": 0.0684, "step": 37230 }, { "grad_norm": 0.6337118148803711, "learning_rate": 9.50138993883021e-05, "loss": 0.0964, "step": 37240 }, { "grad_norm": 0.9432095289230347, "learning_rate": 9.501029986816274e-05, "loss": 0.0926, "step": 37250 }, { "grad_norm": 0.7604528069496155, "learning_rate": 9.500669911745871e-05, "loss": 0.0924, "step": 37260 }, { "grad_norm": 0.7645148038864136, "learning_rate": 9.500309713628845e-05, "loss": 0.0649, "step": 37270 }, { "grad_norm": 0.5761947631835938, "learning_rate": 9.49994939247504e-05, "loss": 0.0625, "step": 37280 }, { "grad_norm": 0.4052508473396301, "learning_rate": 9.499588948294309e-05, "loss": 0.0641, "step": 37290 }, { "grad_norm": 0.5676095485687256, "learning_rate": 9.499228381096507e-05, "loss": 0.0464, "step": 37300 }, { "grad_norm": 0.4165574908256531, "learning_rate": 9.498867690891493e-05, "loss": 0.0638, "step": 37310 }, { "grad_norm": 0.5035921931266785, "learning_rate": 9.498506877689125e-05, "loss": 0.0539, "step": 37320 }, { "grad_norm": 0.7567288875579834, "learning_rate": 9.498145941499269e-05, "loss": 0.0617, "step": 37330 }, { "grad_norm": 0.5520038604736328, "learning_rate": 9.497784882331793e-05, "loss": 0.0609, "step": 37340 }, { "grad_norm": 0.6614239811897278, "learning_rate": 9.497423700196568e-05, "loss": 0.0482, "step": 37350 }, { "grad_norm": 0.7171956896781921, "learning_rate": 9.497062395103469e-05, "loss": 0.0624, "step": 37360 }, { "grad_norm": 0.7163218855857849, "learning_rate": 9.496700967062373e-05, "loss": 0.0697, "step": 37370 }, { "grad_norm": 0.6720759868621826, "learning_rate": 9.496339416083162e-05, "loss": 0.0526, "step": 37380 }, { "grad_norm": 0.38605502247810364, "learning_rate": 9.495977742175722e-05, "loss": 0.0483, "step": 37390 }, { "grad_norm": 1.8855400085449219, "learning_rate": 9.495615945349937e-05, "loss": 0.0565, "step": 37400 }, { "grad_norm": 0.5838456749916077, "learning_rate": 9.495254025615701e-05, "loss": 0.055, "step": 37410 }, { "grad_norm": 0.4566616415977478, "learning_rate": 9.494891982982911e-05, "loss": 0.0423, "step": 37420 }, { "grad_norm": 0.8501203656196594, "learning_rate": 9.494529817461462e-05, "loss": 0.0617, "step": 37430 }, { "grad_norm": 0.5358686447143555, "learning_rate": 9.494167529061256e-05, "loss": 0.047, "step": 37440 }, { "grad_norm": 0.6337095499038696, "learning_rate": 9.493805117792199e-05, "loss": 0.0809, "step": 37450 }, { "grad_norm": 0.6733837723731995, "learning_rate": 9.493442583664196e-05, "loss": 0.0756, "step": 37460 }, { "grad_norm": 0.5568639636039734, "learning_rate": 9.493079926687163e-05, "loss": 0.0522, "step": 37470 }, { "grad_norm": 0.538271427154541, "learning_rate": 9.492717146871013e-05, "loss": 0.043, "step": 37480 }, { "grad_norm": 0.6207449436187744, "learning_rate": 9.492354244225663e-05, "loss": 0.0452, "step": 37490 }, { "grad_norm": 0.6612337231636047, "learning_rate": 9.491991218761036e-05, "loss": 0.0423, "step": 37500 }, { "grad_norm": 0.8426535129547119, "learning_rate": 9.491628070487056e-05, "loss": 0.0585, "step": 37510 }, { "grad_norm": 0.7346916198730469, "learning_rate": 9.491264799413653e-05, "loss": 0.0609, "step": 37520 }, { "grad_norm": 0.5172846913337708, "learning_rate": 9.490901405550757e-05, "loss": 0.067, "step": 37530 }, { "grad_norm": 0.48384737968444824, "learning_rate": 9.490537888908305e-05, "loss": 0.0798, "step": 37540 }, { "grad_norm": 0.527999222278595, "learning_rate": 9.490174249496231e-05, "loss": 0.0429, "step": 37550 }, { "grad_norm": 0.8840999603271484, "learning_rate": 9.489810487324483e-05, "loss": 0.0921, "step": 37560 }, { "grad_norm": 0.8559495210647583, "learning_rate": 9.489446602403002e-05, "loss": 0.0574, "step": 37570 }, { "grad_norm": 0.5144427418708801, "learning_rate": 9.489082594741739e-05, "loss": 0.0496, "step": 37580 }, { "grad_norm": 0.7432140111923218, "learning_rate": 9.48871846435064e-05, "loss": 0.056, "step": 37590 }, { "grad_norm": 0.757522702217102, "learning_rate": 9.48835421123967e-05, "loss": 0.0669, "step": 37600 }, { "grad_norm": 0.513016939163208, "learning_rate": 9.487989835418778e-05, "loss": 0.0721, "step": 37610 }, { "grad_norm": 0.6559446454048157, "learning_rate": 9.487625336897931e-05, "loss": 0.0501, "step": 37620 }, { "grad_norm": 0.7164962291717529, "learning_rate": 9.487260715687095e-05, "loss": 0.0577, "step": 37630 }, { "grad_norm": 0.566735029220581, "learning_rate": 9.486895971796234e-05, "loss": 0.0515, "step": 37640 }, { "grad_norm": 0.5520883798599243, "learning_rate": 9.486531105235326e-05, "loss": 0.0553, "step": 37650 }, { "grad_norm": 0.41785842180252075, "learning_rate": 9.48616611601434e-05, "loss": 0.0509, "step": 37660 }, { "grad_norm": 0.5307128429412842, "learning_rate": 9.48580100414326e-05, "loss": 0.0592, "step": 37670 }, { "grad_norm": 0.8062554597854614, "learning_rate": 9.485435769632064e-05, "loss": 0.049, "step": 37680 }, { "grad_norm": 0.6887593269348145, "learning_rate": 9.48507041249074e-05, "loss": 0.0644, "step": 37690 }, { "grad_norm": 0.7154922485351562, "learning_rate": 9.484704932729276e-05, "loss": 0.0626, "step": 37700 }, { "grad_norm": 0.6453598141670227, "learning_rate": 9.484339330357664e-05, "loss": 0.0639, "step": 37710 }, { "grad_norm": 0.5232036113739014, "learning_rate": 9.483973605385898e-05, "loss": 0.0605, "step": 37720 }, { "grad_norm": 0.43464604020118713, "learning_rate": 9.48360775782398e-05, "loss": 0.0501, "step": 37730 }, { "grad_norm": 0.6664551496505737, "learning_rate": 9.48324178768191e-05, "loss": 0.0571, "step": 37740 }, { "grad_norm": 0.617153525352478, "learning_rate": 9.482875694969693e-05, "loss": 0.0598, "step": 37750 }, { "grad_norm": 0.9737809896469116, "learning_rate": 9.482509479697339e-05, "loss": 0.0555, "step": 37760 }, { "grad_norm": 0.45216917991638184, "learning_rate": 9.48214314187486e-05, "loss": 0.0485, "step": 37770 }, { "grad_norm": 0.5434173345565796, "learning_rate": 9.481776681512272e-05, "loss": 0.0581, "step": 37780 }, { "grad_norm": 0.49230843782424927, "learning_rate": 9.481410098619592e-05, "loss": 0.0707, "step": 37790 }, { "grad_norm": 0.6577203869819641, "learning_rate": 9.481043393206845e-05, "loss": 0.0792, "step": 37800 }, { "grad_norm": 0.7790549397468567, "learning_rate": 9.480676565284054e-05, "loss": 0.0602, "step": 37810 }, { "grad_norm": 0.7063210606575012, "learning_rate": 9.480309614861249e-05, "loss": 0.0615, "step": 37820 }, { "grad_norm": 0.5726139545440674, "learning_rate": 9.479942541948463e-05, "loss": 0.0563, "step": 37830 }, { "grad_norm": 0.5640064477920532, "learning_rate": 9.47957534655573e-05, "loss": 0.0517, "step": 37840 }, { "grad_norm": 0.5324392318725586, "learning_rate": 9.47920802869309e-05, "loss": 0.0599, "step": 37850 }, { "grad_norm": 0.5061910152435303, "learning_rate": 9.478840588370585e-05, "loss": 0.0527, "step": 37860 }, { "grad_norm": 0.632411777973175, "learning_rate": 9.478473025598263e-05, "loss": 0.0514, "step": 37870 }, { "grad_norm": 0.6975783109664917, "learning_rate": 9.47810534038617e-05, "loss": 0.0567, "step": 37880 }, { "grad_norm": 0.4845316410064697, "learning_rate": 9.477737532744357e-05, "loss": 0.0487, "step": 37890 }, { "grad_norm": 0.4002242088317871, "learning_rate": 9.477369602682885e-05, "loss": 0.0482, "step": 37900 }, { "grad_norm": 0.5365331172943115, "learning_rate": 9.477001550211808e-05, "loss": 0.0577, "step": 37910 }, { "grad_norm": 0.9568166732788086, "learning_rate": 9.476633375341193e-05, "loss": 0.0715, "step": 37920 }, { "grad_norm": 0.7784495949745178, "learning_rate": 9.4762650780811e-05, "loss": 0.056, "step": 37930 }, { "grad_norm": 0.5441148281097412, "learning_rate": 9.475896658441605e-05, "loss": 0.0539, "step": 37940 }, { "grad_norm": 0.6450640559196472, "learning_rate": 9.475528116432775e-05, "loss": 0.0535, "step": 37950 }, { "grad_norm": 0.4829612374305725, "learning_rate": 9.475159452064688e-05, "loss": 0.0535, "step": 37960 }, { "grad_norm": 0.7897593975067139, "learning_rate": 9.474790665347423e-05, "loss": 0.0513, "step": 37970 }, { "grad_norm": 0.4505571126937866, "learning_rate": 9.474421756291061e-05, "loss": 0.05, "step": 37980 }, { "grad_norm": 0.46600207686424255, "learning_rate": 9.47405272490569e-05, "loss": 0.0491, "step": 37990 }, { "grad_norm": 0.5745378732681274, "learning_rate": 9.473683571201399e-05, "loss": 0.0718, "step": 38000 }, { "grad_norm": 0.3938022255897522, "learning_rate": 9.47331429518828e-05, "loss": 0.0535, "step": 38010 }, { "grad_norm": 1.061368703842163, "learning_rate": 9.472944896876427e-05, "loss": 0.0621, "step": 38020 }, { "grad_norm": 0.3883627653121948, "learning_rate": 9.472575376275943e-05, "loss": 0.0553, "step": 38030 }, { "grad_norm": 0.512153148651123, "learning_rate": 9.472205733396926e-05, "loss": 0.0533, "step": 38040 }, { "grad_norm": 0.5642637014389038, "learning_rate": 9.471835968249487e-05, "loss": 0.0492, "step": 38050 }, { "grad_norm": 0.7284396886825562, "learning_rate": 9.471466080843732e-05, "loss": 0.0602, "step": 38060 }, { "grad_norm": 0.5756042003631592, "learning_rate": 9.471096071189774e-05, "loss": 0.0716, "step": 38070 }, { "grad_norm": 0.4820091128349304, "learning_rate": 9.470725939297727e-05, "loss": 0.0493, "step": 38080 }, { "grad_norm": 0.5924922227859497, "learning_rate": 9.470355685177714e-05, "loss": 0.0546, "step": 38090 }, { "grad_norm": 0.5168845057487488, "learning_rate": 9.469985308839857e-05, "loss": 0.0604, "step": 38100 }, { "grad_norm": 0.5319409370422363, "learning_rate": 9.469614810294281e-05, "loss": 0.0472, "step": 38110 }, { "grad_norm": 0.6456055641174316, "learning_rate": 9.469244189551114e-05, "loss": 0.0488, "step": 38120 }, { "grad_norm": 0.6127622723579407, "learning_rate": 9.46887344662049e-05, "loss": 0.0532, "step": 38130 }, { "grad_norm": 0.5059843063354492, "learning_rate": 9.468502581512543e-05, "loss": 0.0501, "step": 38140 }, { "grad_norm": 0.5405874848365784, "learning_rate": 9.468131594237419e-05, "loss": 0.048, "step": 38150 }, { "grad_norm": 0.8069409728050232, "learning_rate": 9.467760484805253e-05, "loss": 0.0563, "step": 38160 }, { "grad_norm": 0.38162562251091003, "learning_rate": 9.467389253226193e-05, "loss": 0.0742, "step": 38170 }, { "grad_norm": 0.5935239195823669, "learning_rate": 9.46701789951039e-05, "loss": 0.0538, "step": 38180 }, { "grad_norm": 0.5422411561012268, "learning_rate": 9.466646423667996e-05, "loss": 0.0766, "step": 38190 }, { "grad_norm": 0.7644691467285156, "learning_rate": 9.466274825709166e-05, "loss": 0.0554, "step": 38200 }, { "grad_norm": 0.6284500956535339, "learning_rate": 9.465903105644061e-05, "loss": 0.0526, "step": 38210 }, { "grad_norm": 0.5803713202476501, "learning_rate": 9.465531263482843e-05, "loss": 0.0581, "step": 38220 }, { "grad_norm": 0.7365002632141113, "learning_rate": 9.465159299235677e-05, "loss": 0.0592, "step": 38230 }, { "grad_norm": 0.3992116451263428, "learning_rate": 9.464787212912733e-05, "loss": 0.0459, "step": 38240 }, { "grad_norm": 0.4218309819698334, "learning_rate": 9.464415004524185e-05, "loss": 0.0749, "step": 38250 }, { "grad_norm": 0.4749126434326172, "learning_rate": 9.46404267408021e-05, "loss": 0.0821, "step": 38260 }, { "grad_norm": 0.6375991106033325, "learning_rate": 9.463670221590982e-05, "loss": 0.0536, "step": 38270 }, { "grad_norm": 0.515113115310669, "learning_rate": 9.463297647066687e-05, "loss": 0.0501, "step": 38280 }, { "grad_norm": 0.6036314964294434, "learning_rate": 9.462924950517514e-05, "loss": 0.048, "step": 38290 }, { "grad_norm": 0.6186880469322205, "learning_rate": 9.462552131953646e-05, "loss": 0.0428, "step": 38300 }, { "grad_norm": 0.6101677417755127, "learning_rate": 9.462179191385281e-05, "loss": 0.0646, "step": 38310 }, { "grad_norm": 0.3987244665622711, "learning_rate": 9.461806128822614e-05, "loss": 0.0594, "step": 38320 }, { "grad_norm": 1.3485904932022095, "learning_rate": 9.461432944275843e-05, "loss": 0.0475, "step": 38330 }, { "grad_norm": 0.6861851215362549, "learning_rate": 9.46105963775517e-05, "loss": 0.0688, "step": 38340 }, { "grad_norm": 0.6529963612556458, "learning_rate": 9.460686209270803e-05, "loss": 0.0621, "step": 38350 }, { "grad_norm": 0.6768190860748291, "learning_rate": 9.460312658832953e-05, "loss": 0.0545, "step": 38360 }, { "grad_norm": 0.7455906867980957, "learning_rate": 9.459938986451828e-05, "loss": 0.0808, "step": 38370 }, { "grad_norm": 0.6228712797164917, "learning_rate": 9.45956519213765e-05, "loss": 0.0593, "step": 38380 }, { "grad_norm": 0.54902583360672, "learning_rate": 9.459191275900632e-05, "loss": 0.0514, "step": 38390 }, { "grad_norm": 0.7834385633468628, "learning_rate": 9.458817237751e-05, "loss": 0.0603, "step": 38400 }, { "grad_norm": 0.5582574605941772, "learning_rate": 9.458443077698979e-05, "loss": 0.0643, "step": 38410 }, { "grad_norm": 0.7417064309120178, "learning_rate": 9.458068795754802e-05, "loss": 0.0451, "step": 38420 }, { "grad_norm": 0.7128990888595581, "learning_rate": 9.457694391928698e-05, "loss": 0.0429, "step": 38430 }, { "grad_norm": 0.4733484983444214, "learning_rate": 9.457319866230902e-05, "loss": 0.0572, "step": 38440 }, { "grad_norm": 0.6563340425491333, "learning_rate": 9.456945218671657e-05, "loss": 0.0575, "step": 38450 }, { "grad_norm": 0.4664495885372162, "learning_rate": 9.456570449261203e-05, "loss": 0.0665, "step": 38460 }, { "grad_norm": 0.4253568649291992, "learning_rate": 9.456195558009791e-05, "loss": 0.0557, "step": 38470 }, { "grad_norm": 0.5116690397262573, "learning_rate": 9.455820544927662e-05, "loss": 0.053, "step": 38480 }, { "grad_norm": 0.782265841960907, "learning_rate": 9.455445410025076e-05, "loss": 0.0591, "step": 38490 }, { "grad_norm": 0.6348294615745544, "learning_rate": 9.455070153312285e-05, "loss": 0.0516, "step": 38500 }, { "grad_norm": 0.6077458262443542, "learning_rate": 9.45469477479955e-05, "loss": 0.0571, "step": 38510 }, { "grad_norm": 0.6723889112472534, "learning_rate": 9.454319274497134e-05, "loss": 0.0502, "step": 38520 }, { "grad_norm": 0.4661858379840851, "learning_rate": 9.453943652415302e-05, "loss": 0.0542, "step": 38530 }, { "grad_norm": 0.5240092873573303, "learning_rate": 9.453567908564323e-05, "loss": 0.0483, "step": 38540 }, { "grad_norm": 0.4746851325035095, "learning_rate": 9.453192042954471e-05, "loss": 0.0606, "step": 38550 }, { "grad_norm": 0.43724530935287476, "learning_rate": 9.452816055596022e-05, "loss": 0.0429, "step": 38560 }, { "grad_norm": 0.45979154109954834, "learning_rate": 9.452439946499254e-05, "loss": 0.0569, "step": 38570 }, { "grad_norm": 0.627076268196106, "learning_rate": 9.452063715674453e-05, "loss": 0.0625, "step": 38580 }, { "grad_norm": 0.7734909653663635, "learning_rate": 9.451687363131901e-05, "loss": 0.0458, "step": 38590 }, { "grad_norm": 0.5821864008903503, "learning_rate": 9.451310888881888e-05, "loss": 0.066, "step": 38600 }, { "grad_norm": 0.5544283986091614, "learning_rate": 9.450934292934708e-05, "loss": 0.0545, "step": 38610 }, { "grad_norm": 0.39072126150131226, "learning_rate": 9.450557575300655e-05, "loss": 0.0534, "step": 38620 }, { "grad_norm": 0.3305951654911041, "learning_rate": 9.450180735990032e-05, "loss": 0.0526, "step": 38630 }, { "grad_norm": 0.5864337682723999, "learning_rate": 9.44980377501314e-05, "loss": 0.0659, "step": 38640 }, { "grad_norm": 0.8364475965499878, "learning_rate": 9.449426692380284e-05, "loss": 0.0572, "step": 38650 }, { "grad_norm": 0.4416173994541168, "learning_rate": 9.449049488101773e-05, "loss": 0.0756, "step": 38660 }, { "grad_norm": 0.4203472435474396, "learning_rate": 9.448672162187921e-05, "loss": 0.0546, "step": 38670 }, { "grad_norm": 1.064218282699585, "learning_rate": 9.448294714649044e-05, "loss": 0.0926, "step": 38680 }, { "grad_norm": 0.7073491811752319, "learning_rate": 9.447917145495459e-05, "loss": 0.0686, "step": 38690 }, { "grad_norm": 0.41609832644462585, "learning_rate": 9.447539454737493e-05, "loss": 0.0653, "step": 38700 }, { "grad_norm": 0.41844573616981506, "learning_rate": 9.447161642385467e-05, "loss": 0.04, "step": 38710 }, { "grad_norm": 0.5649365782737732, "learning_rate": 9.446783708449714e-05, "loss": 0.0602, "step": 38720 }, { "grad_norm": 0.43155837059020996, "learning_rate": 9.446405652940562e-05, "loss": 0.0498, "step": 38730 }, { "grad_norm": 0.5011670589447021, "learning_rate": 9.446027475868352e-05, "loss": 0.0497, "step": 38740 }, { "grad_norm": 0.5371214747428894, "learning_rate": 9.445649177243421e-05, "loss": 0.0509, "step": 38750 }, { "grad_norm": 0.6833273768424988, "learning_rate": 9.445270757076112e-05, "loss": 0.0707, "step": 38760 }, { "grad_norm": 0.6853276491165161, "learning_rate": 9.44489221537677e-05, "loss": 0.0579, "step": 38770 }, { "grad_norm": 0.5821143984794617, "learning_rate": 9.444513552155744e-05, "loss": 0.053, "step": 38780 }, { "grad_norm": 0.5002533793449402, "learning_rate": 9.444134767423388e-05, "loss": 0.0436, "step": 38790 }, { "grad_norm": 0.38297125697135925, "learning_rate": 9.443755861190058e-05, "loss": 0.0548, "step": 38800 }, { "grad_norm": 0.773784339427948, "learning_rate": 9.44337683346611e-05, "loss": 0.0536, "step": 38810 }, { "grad_norm": 0.5072586536407471, "learning_rate": 9.44299768426191e-05, "loss": 0.0394, "step": 38820 }, { "grad_norm": 0.44834306836128235, "learning_rate": 9.442618413587822e-05, "loss": 0.0468, "step": 38830 }, { "grad_norm": 0.4628048837184906, "learning_rate": 9.442239021454215e-05, "loss": 0.0507, "step": 38840 }, { "grad_norm": 0.6916266679763794, "learning_rate": 9.441859507871465e-05, "loss": 0.0465, "step": 38850 }, { "grad_norm": 0.8468281626701355, "learning_rate": 9.44147987284994e-05, "loss": 0.0877, "step": 38860 }, { "grad_norm": 0.39569488167762756, "learning_rate": 9.441100116400028e-05, "loss": 0.0435, "step": 38870 }, { "grad_norm": 0.39846116304397583, "learning_rate": 9.440720238532106e-05, "loss": 0.058, "step": 38880 }, { "grad_norm": 0.5671919584274292, "learning_rate": 9.440340239256561e-05, "loss": 0.055, "step": 38890 }, { "grad_norm": 0.5348682403564453, "learning_rate": 9.439960118583782e-05, "loss": 0.0533, "step": 38900 }, { "grad_norm": 0.5548784136772156, "learning_rate": 9.43957987652416e-05, "loss": 0.0447, "step": 38910 }, { "grad_norm": 0.5689613223075867, "learning_rate": 9.439199513088094e-05, "loss": 0.0517, "step": 38920 }, { "grad_norm": 0.6938624978065491, "learning_rate": 9.43881902828598e-05, "loss": 0.0639, "step": 38930 }, { "grad_norm": 0.3545879125595093, "learning_rate": 9.438438422128221e-05, "loss": 0.0553, "step": 38940 }, { "grad_norm": 0.8119248151779175, "learning_rate": 9.438057694625224e-05, "loss": 0.0527, "step": 38950 }, { "grad_norm": 0.6444816589355469, "learning_rate": 9.437676845787397e-05, "loss": 0.0519, "step": 38960 }, { "grad_norm": 0.635748028755188, "learning_rate": 9.437295875625151e-05, "loss": 0.0726, "step": 38970 }, { "grad_norm": 0.8320780396461487, "learning_rate": 9.436914784148904e-05, "loss": 0.0668, "step": 38980 }, { "grad_norm": 0.689856767654419, "learning_rate": 9.436533571369073e-05, "loss": 0.0548, "step": 38990 }, { "grad_norm": 0.46177083253860474, "learning_rate": 9.43615223729608e-05, "loss": 0.0497, "step": 39000 }, { "grad_norm": 1.454907774925232, "learning_rate": 9.435770781940354e-05, "loss": 0.0684, "step": 39010 }, { "grad_norm": 0.6649978160858154, "learning_rate": 9.435389205312319e-05, "loss": 0.0766, "step": 39020 }, { "grad_norm": 0.4575692415237427, "learning_rate": 9.43500750742241e-05, "loss": 0.0508, "step": 39030 }, { "grad_norm": 0.5353708863258362, "learning_rate": 9.434625688281063e-05, "loss": 0.0481, "step": 39040 }, { "grad_norm": 0.6060479879379272, "learning_rate": 9.434243747898714e-05, "loss": 0.0473, "step": 39050 }, { "grad_norm": 0.7729438543319702, "learning_rate": 9.433861686285808e-05, "loss": 0.0657, "step": 39060 }, { "grad_norm": 0.736286461353302, "learning_rate": 9.433479503452789e-05, "loss": 0.0591, "step": 39070 }, { "grad_norm": 0.446487694978714, "learning_rate": 9.433097199410106e-05, "loss": 0.0526, "step": 39080 }, { "grad_norm": 0.44137492775917053, "learning_rate": 9.43271477416821e-05, "loss": 0.0616, "step": 39090 }, { "grad_norm": 0.6450083255767822, "learning_rate": 9.432332227737559e-05, "loss": 0.0573, "step": 39100 }, { "grad_norm": 0.556188702583313, "learning_rate": 9.431949560128607e-05, "loss": 0.0663, "step": 39110 }, { "grad_norm": 0.43591830134391785, "learning_rate": 9.431566771351823e-05, "loss": 0.0716, "step": 39120 }, { "grad_norm": 0.7986032962799072, "learning_rate": 9.431183861417666e-05, "loss": 0.0605, "step": 39130 }, { "grad_norm": 0.47470566630363464, "learning_rate": 9.430800830336608e-05, "loss": 0.0464, "step": 39140 }, { "grad_norm": 0.5974013805389404, "learning_rate": 9.430417678119119e-05, "loss": 0.0699, "step": 39150 }, { "grad_norm": 0.5917259454727173, "learning_rate": 9.430034404775676e-05, "loss": 0.045, "step": 39160 }, { "grad_norm": 0.7656868100166321, "learning_rate": 9.429651010316756e-05, "loss": 0.0539, "step": 39170 }, { "grad_norm": 0.49329477548599243, "learning_rate": 9.429267494752842e-05, "loss": 0.0444, "step": 39180 }, { "grad_norm": 0.447177529335022, "learning_rate": 9.428883858094418e-05, "loss": 0.0602, "step": 39190 }, { "grad_norm": 0.8612958192825317, "learning_rate": 9.428500100351973e-05, "loss": 0.0715, "step": 39200 }, { "grad_norm": 0.49064722657203674, "learning_rate": 9.428116221536e-05, "loss": 0.0637, "step": 39210 }, { "grad_norm": 0.6008786559104919, "learning_rate": 9.427732221656992e-05, "loss": 0.0771, "step": 39220 }, { "grad_norm": 0.5996813774108887, "learning_rate": 9.427348100725448e-05, "loss": 0.062, "step": 39230 }, { "grad_norm": 0.9614964723587036, "learning_rate": 9.426963858751872e-05, "loss": 0.0537, "step": 39240 }, { "grad_norm": 0.5676117539405823, "learning_rate": 9.426579495746767e-05, "loss": 0.0717, "step": 39250 }, { "grad_norm": 0.6750494241714478, "learning_rate": 9.42619501172064e-05, "loss": 0.0552, "step": 39260 }, { "grad_norm": 0.3603980839252472, "learning_rate": 9.425810406684006e-05, "loss": 0.0484, "step": 39270 }, { "grad_norm": 0.4633558988571167, "learning_rate": 9.425425680647377e-05, "loss": 0.0548, "step": 39280 }, { "grad_norm": 0.5188769698143005, "learning_rate": 9.425040833621273e-05, "loss": 0.0474, "step": 39290 }, { "grad_norm": 0.46420949697494507, "learning_rate": 9.424655865616215e-05, "loss": 0.0586, "step": 39300 }, { "grad_norm": 0.4513736069202423, "learning_rate": 9.424270776642728e-05, "loss": 0.0544, "step": 39310 }, { "grad_norm": 0.4651106595993042, "learning_rate": 9.423885566711339e-05, "loss": 0.048, "step": 39320 }, { "grad_norm": 0.582843005657196, "learning_rate": 9.423500235832584e-05, "loss": 0.0532, "step": 39330 }, { "grad_norm": 0.4097098410129547, "learning_rate": 9.42311478401699e-05, "loss": 0.0445, "step": 39340 }, { "grad_norm": 0.6311280727386475, "learning_rate": 9.422729211275101e-05, "loss": 0.0481, "step": 39350 }, { "grad_norm": 0.4525472819805145, "learning_rate": 9.422343517617458e-05, "loss": 0.0502, "step": 39360 }, { "grad_norm": 0.4662444591522217, "learning_rate": 9.421957703054604e-05, "loss": 0.0722, "step": 39370 }, { "grad_norm": 0.3647778630256653, "learning_rate": 9.421571767597086e-05, "loss": 0.0472, "step": 39380 }, { "grad_norm": 0.6018369793891907, "learning_rate": 9.42118571125546e-05, "loss": 0.0506, "step": 39390 }, { "grad_norm": 0.47158282995224, "learning_rate": 9.420799534040275e-05, "loss": 0.0519, "step": 39400 }, { "grad_norm": 0.8252403140068054, "learning_rate": 9.420413235962094e-05, "loss": 0.0614, "step": 39410 }, { "grad_norm": 0.672451913356781, "learning_rate": 9.420026817031473e-05, "loss": 0.0508, "step": 39420 }, { "grad_norm": 0.6505535244941711, "learning_rate": 9.419640277258981e-05, "loss": 0.0593, "step": 39430 }, { "grad_norm": 0.971851646900177, "learning_rate": 9.419253616655185e-05, "loss": 0.0621, "step": 39440 }, { "grad_norm": 0.5506219267845154, "learning_rate": 9.418866835230654e-05, "loss": 0.0482, "step": 39450 }, { "grad_norm": 0.5971584320068359, "learning_rate": 9.418479932995965e-05, "loss": 0.0553, "step": 39460 }, { "grad_norm": 0.46934112906455994, "learning_rate": 9.418092909961693e-05, "loss": 0.0542, "step": 39470 }, { "grad_norm": 0.5455076098442078, "learning_rate": 9.417705766138422e-05, "loss": 0.0414, "step": 39480 }, { "grad_norm": 0.43488427996635437, "learning_rate": 9.417318501536733e-05, "loss": 0.0342, "step": 39490 }, { "grad_norm": 1.8812072277069092, "learning_rate": 9.416931116167218e-05, "loss": 0.0554, "step": 39500 }, { "grad_norm": 0.6768897771835327, "learning_rate": 9.416543610040465e-05, "loss": 0.0528, "step": 39510 }, { "grad_norm": 0.47714078426361084, "learning_rate": 9.416155983167068e-05, "loss": 0.0673, "step": 39520 }, { "grad_norm": 1.0222766399383545, "learning_rate": 9.415768235557627e-05, "loss": 0.0602, "step": 39530 }, { "grad_norm": 0.5711783766746521, "learning_rate": 9.415380367222739e-05, "loss": 0.0553, "step": 39540 }, { "grad_norm": 0.676678478717804, "learning_rate": 9.414992378173013e-05, "loss": 0.0732, "step": 39550 }, { "grad_norm": 0.38917112350463867, "learning_rate": 9.414604268419054e-05, "loss": 0.0467, "step": 39560 }, { "grad_norm": 0.5268548130989075, "learning_rate": 9.414216037971471e-05, "loss": 0.058, "step": 39570 }, { "grad_norm": 0.7261295318603516, "learning_rate": 9.413827686840881e-05, "loss": 0.0568, "step": 39580 }, { "grad_norm": 0.54839688539505, "learning_rate": 9.4134392150379e-05, "loss": 0.0466, "step": 39590 }, { "grad_norm": 0.598871111869812, "learning_rate": 9.413050622573148e-05, "loss": 0.0576, "step": 39600 }, { "grad_norm": 0.7328147292137146, "learning_rate": 9.412661909457252e-05, "loss": 0.0551, "step": 39610 }, { "grad_norm": 0.5749231576919556, "learning_rate": 9.412273075700835e-05, "loss": 0.0488, "step": 39620 }, { "grad_norm": 0.6395809054374695, "learning_rate": 9.411884121314532e-05, "loss": 0.0627, "step": 39630 }, { "grad_norm": 0.5811132788658142, "learning_rate": 9.411495046308973e-05, "loss": 0.0463, "step": 39640 }, { "grad_norm": 0.8581499457359314, "learning_rate": 9.411105850694796e-05, "loss": 0.0471, "step": 39650 }, { "grad_norm": 0.5071620941162109, "learning_rate": 9.410716534482644e-05, "loss": 0.0519, "step": 39660 }, { "grad_norm": 1.543521761894226, "learning_rate": 9.410327097683158e-05, "loss": 0.0662, "step": 39670 }, { "grad_norm": 0.7431442141532898, "learning_rate": 9.409937540306986e-05, "loss": 0.0454, "step": 39680 }, { "grad_norm": 0.6767323613166809, "learning_rate": 9.409547862364778e-05, "loss": 0.0482, "step": 39690 }, { "grad_norm": 1.3396743535995483, "learning_rate": 9.40915806386719e-05, "loss": 0.0794, "step": 39700 }, { "grad_norm": 0.6136295795440674, "learning_rate": 9.408768144824875e-05, "loss": 0.0428, "step": 39710 }, { "grad_norm": 0.7352246046066284, "learning_rate": 9.408378105248496e-05, "loss": 0.0771, "step": 39720 }, { "grad_norm": 0.7608438730239868, "learning_rate": 9.407987945148716e-05, "loss": 0.0557, "step": 39730 }, { "grad_norm": 0.454725444316864, "learning_rate": 9.4075976645362e-05, "loss": 0.0498, "step": 39740 }, { "grad_norm": 0.5648193359375, "learning_rate": 9.407207263421622e-05, "loss": 0.0614, "step": 39750 }, { "grad_norm": 0.5904868245124817, "learning_rate": 9.406816741815652e-05, "loss": 0.0477, "step": 39760 }, { "grad_norm": 0.8889037370681763, "learning_rate": 9.406426099728967e-05, "loss": 0.0717, "step": 39770 }, { "grad_norm": 0.7656559944152832, "learning_rate": 9.40603533717225e-05, "loss": 0.0502, "step": 39780 }, { "grad_norm": 1.2550270557403564, "learning_rate": 9.40564445415618e-05, "loss": 0.0615, "step": 39790 }, { "grad_norm": 0.8840770125389099, "learning_rate": 9.405253450691446e-05, "loss": 0.0925, "step": 39800 }, { "grad_norm": 0.681941032409668, "learning_rate": 9.404862326788739e-05, "loss": 0.066, "step": 39810 }, { "grad_norm": 0.656845211982727, "learning_rate": 9.40447108245875e-05, "loss": 0.049, "step": 39820 }, { "grad_norm": 1.0447509288787842, "learning_rate": 9.404079717712176e-05, "loss": 0.0543, "step": 39830 }, { "grad_norm": 0.6145777702331543, "learning_rate": 9.403688232559718e-05, "loss": 0.0473, "step": 39840 }, { "grad_norm": 0.7541753053665161, "learning_rate": 9.403296627012077e-05, "loss": 0.0576, "step": 39850 }, { "grad_norm": 0.6986128687858582, "learning_rate": 9.402904901079961e-05, "loss": 0.0594, "step": 39860 }, { "grad_norm": 0.5697027444839478, "learning_rate": 9.40251305477408e-05, "loss": 0.0464, "step": 39870 }, { "grad_norm": 0.7457327842712402, "learning_rate": 9.402121088105144e-05, "loss": 0.0581, "step": 39880 }, { "grad_norm": 0.5150832533836365, "learning_rate": 9.401729001083873e-05, "loss": 0.0405, "step": 39890 }, { "grad_norm": 0.7487455606460571, "learning_rate": 9.401336793720983e-05, "loss": 0.0815, "step": 39900 }, { "grad_norm": 0.6710911393165588, "learning_rate": 9.400944466027201e-05, "loss": 0.0557, "step": 39910 }, { "grad_norm": 0.6803922653198242, "learning_rate": 9.40055201801325e-05, "loss": 0.0573, "step": 39920 }, { "grad_norm": 0.8422815203666687, "learning_rate": 9.400159449689858e-05, "loss": 0.0577, "step": 39930 }, { "grad_norm": 0.788274884223938, "learning_rate": 9.399766761067761e-05, "loss": 0.0578, "step": 39940 }, { "grad_norm": 0.7037048935890198, "learning_rate": 9.399373952157694e-05, "loss": 0.0616, "step": 39950 }, { "grad_norm": 1.6681535243988037, "learning_rate": 9.398981022970395e-05, "loss": 0.0663, "step": 39960 }, { "grad_norm": 0.5308334827423096, "learning_rate": 9.398587973516608e-05, "loss": 0.0603, "step": 39970 }, { "grad_norm": 0.5306441187858582, "learning_rate": 9.398194803807078e-05, "loss": 0.0509, "step": 39980 }, { "grad_norm": 0.46514439582824707, "learning_rate": 9.397801513852555e-05, "loss": 0.0441, "step": 39990 }, { "grad_norm": 0.44372448325157166, "learning_rate": 9.397408103663792e-05, "loss": 0.0655, "step": 40000 }, { "grad_norm": 0.7412471771240234, "learning_rate": 9.397014573251541e-05, "loss": 0.0527, "step": 40010 }, { "grad_norm": 0.39805853366851807, "learning_rate": 9.396620922626564e-05, "loss": 0.0502, "step": 40020 }, { "grad_norm": 0.5463377237319946, "learning_rate": 9.396227151799623e-05, "loss": 0.0582, "step": 40030 }, { "grad_norm": 0.5421064496040344, "learning_rate": 9.395833260781485e-05, "loss": 0.0542, "step": 40040 }, { "grad_norm": 0.9056123495101929, "learning_rate": 9.395439249582916e-05, "loss": 0.0712, "step": 40050 }, { "grad_norm": 1.1315022706985474, "learning_rate": 9.395045118214688e-05, "loss": 0.0569, "step": 40060 }, { "grad_norm": 0.44518572092056274, "learning_rate": 9.394650866687578e-05, "loss": 0.0942, "step": 40070 }, { "grad_norm": 0.7256097197532654, "learning_rate": 9.394256495012366e-05, "loss": 0.0528, "step": 40080 }, { "grad_norm": 0.9244576096534729, "learning_rate": 9.393862003199829e-05, "loss": 0.0633, "step": 40090 }, { "grad_norm": 0.5040315389633179, "learning_rate": 9.393467391260759e-05, "loss": 0.0652, "step": 40100 }, { "grad_norm": 0.3381940424442291, "learning_rate": 9.393072659205939e-05, "loss": 0.0422, "step": 40110 }, { "grad_norm": 1.0552481412887573, "learning_rate": 9.392677807046164e-05, "loss": 0.0603, "step": 40120 }, { "grad_norm": 0.36204513907432556, "learning_rate": 9.392282834792228e-05, "loss": 0.0468, "step": 40130 }, { "grad_norm": 0.9085021615028381, "learning_rate": 9.391887742454928e-05, "loss": 0.095, "step": 40140 }, { "grad_norm": 0.8438031077384949, "learning_rate": 9.391492530045068e-05, "loss": 0.0444, "step": 40150 }, { "grad_norm": 0.650852620601654, "learning_rate": 9.391097197573452e-05, "loss": 0.0571, "step": 40160 }, { "grad_norm": 0.6306065917015076, "learning_rate": 9.390701745050889e-05, "loss": 0.0769, "step": 40170 }, { "grad_norm": 0.5263387560844421, "learning_rate": 9.390306172488188e-05, "loss": 0.0493, "step": 40180 }, { "grad_norm": 0.564939022064209, "learning_rate": 9.389910479896167e-05, "loss": 0.0556, "step": 40190 }, { "grad_norm": 0.5886586308479309, "learning_rate": 9.389514667285642e-05, "loss": 0.0529, "step": 40200 }, { "grad_norm": 0.8666812777519226, "learning_rate": 9.389118734667435e-05, "loss": 0.0574, "step": 40210 }, { "grad_norm": 0.5539079904556274, "learning_rate": 9.388722682052371e-05, "loss": 0.0508, "step": 40220 }, { "grad_norm": 0.46653786301612854, "learning_rate": 9.388326509451278e-05, "loss": 0.0602, "step": 40230 }, { "grad_norm": 0.9844915270805359, "learning_rate": 9.387930216874987e-05, "loss": 0.0586, "step": 40240 }, { "grad_norm": 0.5750027894973755, "learning_rate": 9.387533804334332e-05, "loss": 0.0537, "step": 40250 }, { "grad_norm": 0.6519283652305603, "learning_rate": 9.387137271840151e-05, "loss": 0.056, "step": 40260 }, { "grad_norm": 0.482225626707077, "learning_rate": 9.386740619403285e-05, "loss": 0.0501, "step": 40270 }, { "grad_norm": 0.7927689552307129, "learning_rate": 9.386343847034578e-05, "loss": 0.052, "step": 40280 }, { "grad_norm": 0.7195131778717041, "learning_rate": 9.38594695474488e-05, "loss": 0.078, "step": 40290 }, { "grad_norm": 0.5508983135223389, "learning_rate": 9.385549942545038e-05, "loss": 0.0458, "step": 40300 }, { "grad_norm": 0.6527243256568909, "learning_rate": 9.38515281044591e-05, "loss": 0.0656, "step": 40310 }, { "grad_norm": 0.6618279218673706, "learning_rate": 9.384755558458349e-05, "loss": 0.0548, "step": 40320 }, { "grad_norm": 0.6581013798713684, "learning_rate": 9.38435818659322e-05, "loss": 0.0726, "step": 40330 }, { "grad_norm": 0.7268432378768921, "learning_rate": 9.383960694861385e-05, "loss": 0.0657, "step": 40340 }, { "grad_norm": 0.6222301721572876, "learning_rate": 9.383563083273712e-05, "loss": 0.0601, "step": 40350 }, { "grad_norm": 1.2311949729919434, "learning_rate": 9.383165351841071e-05, "loss": 0.0825, "step": 40360 }, { "grad_norm": 1.1235592365264893, "learning_rate": 9.382767500574337e-05, "loss": 0.0629, "step": 40370 }, { "grad_norm": 0.42400678992271423, "learning_rate": 9.382369529484383e-05, "loss": 0.0524, "step": 40380 }, { "grad_norm": 0.5675543546676636, "learning_rate": 9.381971438582096e-05, "loss": 0.047, "step": 40390 }, { "grad_norm": 0.798138439655304, "learning_rate": 9.381573227878353e-05, "loss": 0.0538, "step": 40400 }, { "grad_norm": 0.4509347081184387, "learning_rate": 9.381174897384047e-05, "loss": 0.0521, "step": 40410 }, { "grad_norm": 0.5592736601829529, "learning_rate": 9.380776447110062e-05, "loss": 0.0454, "step": 40420 }, { "grad_norm": 0.6470346450805664, "learning_rate": 9.380377877067297e-05, "loss": 0.068, "step": 40430 }, { "grad_norm": 0.9256389141082764, "learning_rate": 9.379979187266645e-05, "loss": 0.0537, "step": 40440 }, { "grad_norm": 0.49652865529060364, "learning_rate": 9.379580377719009e-05, "loss": 0.0444, "step": 40450 }, { "grad_norm": 1.0576714277267456, "learning_rate": 9.37918144843529e-05, "loss": 0.0829, "step": 40460 }, { "grad_norm": 0.9002124071121216, "learning_rate": 9.378782399426396e-05, "loss": 0.0464, "step": 40470 }, { "grad_norm": 0.5864037275314331, "learning_rate": 9.378383230703236e-05, "loss": 0.0633, "step": 40480 }, { "grad_norm": 0.459612637758255, "learning_rate": 9.377983942276723e-05, "loss": 0.0499, "step": 40490 }, { "grad_norm": 0.640424907207489, "learning_rate": 9.377584534157773e-05, "loss": 0.0506, "step": 40500 }, { "grad_norm": 0.6698077321052551, "learning_rate": 9.377185006357307e-05, "loss": 0.0475, "step": 40510 }, { "grad_norm": 0.5443724393844604, "learning_rate": 9.376785358886249e-05, "loss": 0.0604, "step": 40520 }, { "grad_norm": 0.8588810563087463, "learning_rate": 9.37638559175552e-05, "loss": 0.0515, "step": 40530 }, { "grad_norm": 0.6161930561065674, "learning_rate": 9.375985704976057e-05, "loss": 0.0486, "step": 40540 }, { "grad_norm": 0.5528659224510193, "learning_rate": 9.375585698558787e-05, "loss": 0.0531, "step": 40550 }, { "grad_norm": 1.2958667278289795, "learning_rate": 9.375185572514648e-05, "loss": 0.0673, "step": 40560 }, { "grad_norm": 0.6010765433311462, "learning_rate": 9.374785326854578e-05, "loss": 0.0695, "step": 40570 }, { "grad_norm": 0.47030243277549744, "learning_rate": 9.374384961589523e-05, "loss": 0.0618, "step": 40580 }, { "grad_norm": 0.5781823992729187, "learning_rate": 9.373984476730424e-05, "loss": 0.0479, "step": 40590 }, { "grad_norm": 0.6638754606246948, "learning_rate": 9.373583872288235e-05, "loss": 0.0526, "step": 40600 }, { "grad_norm": 0.853328287601471, "learning_rate": 9.373183148273905e-05, "loss": 0.0686, "step": 40610 }, { "grad_norm": 0.546632707118988, "learning_rate": 9.372782304698391e-05, "loss": 0.0496, "step": 40620 }, { "grad_norm": 0.6053578853607178, "learning_rate": 9.372381341572651e-05, "loss": 0.0505, "step": 40630 }, { "grad_norm": 0.5403996109962463, "learning_rate": 9.371980258907647e-05, "loss": 0.0543, "step": 40640 }, { "grad_norm": 0.5969939231872559, "learning_rate": 9.371579056714345e-05, "loss": 0.0581, "step": 40650 }, { "grad_norm": 0.6182414889335632, "learning_rate": 9.371177735003716e-05, "loss": 0.0598, "step": 40660 }, { "grad_norm": 0.7411803007125854, "learning_rate": 9.370776293786729e-05, "loss": 0.061, "step": 40670 }, { "grad_norm": 0.6551311016082764, "learning_rate": 9.37037473307436e-05, "loss": 0.0496, "step": 40680 }, { "grad_norm": 0.9291888475418091, "learning_rate": 9.369973052877587e-05, "loss": 0.0639, "step": 40690 }, { "grad_norm": 0.690711259841919, "learning_rate": 9.369571253207394e-05, "loss": 0.0569, "step": 40700 }, { "grad_norm": 0.7216624021530151, "learning_rate": 9.369169334074763e-05, "loss": 0.0612, "step": 40710 }, { "grad_norm": 0.6760673522949219, "learning_rate": 9.368767295490686e-05, "loss": 0.0549, "step": 40720 }, { "grad_norm": 0.5648287534713745, "learning_rate": 9.36836513746615e-05, "loss": 0.0537, "step": 40730 }, { "grad_norm": 0.6468533873558044, "learning_rate": 9.367962860012154e-05, "loss": 0.0733, "step": 40740 }, { "grad_norm": 0.7728064060211182, "learning_rate": 9.367560463139692e-05, "loss": 0.0512, "step": 40750 }, { "grad_norm": 0.6865125298500061, "learning_rate": 9.36715794685977e-05, "loss": 0.0809, "step": 40760 }, { "grad_norm": 0.5269246101379395, "learning_rate": 9.36675531118339e-05, "loss": 0.0564, "step": 40770 }, { "grad_norm": 0.6853713989257812, "learning_rate": 9.36635255612156e-05, "loss": 0.0642, "step": 40780 }, { "grad_norm": 0.44678565859794617, "learning_rate": 9.365949681685291e-05, "loss": 0.0622, "step": 40790 }, { "grad_norm": 0.9661129117012024, "learning_rate": 9.3655466878856e-05, "loss": 0.0514, "step": 40800 }, { "grad_norm": 0.5320448279380798, "learning_rate": 9.365143574733499e-05, "loss": 0.0445, "step": 40810 }, { "grad_norm": 0.6205260753631592, "learning_rate": 9.364740342240016e-05, "loss": 0.0565, "step": 40820 }, { "grad_norm": 0.5811701416969299, "learning_rate": 9.36433699041617e-05, "loss": 0.0493, "step": 40830 }, { "grad_norm": 0.6981018781661987, "learning_rate": 9.36393351927299e-05, "loss": 0.0505, "step": 40840 }, { "grad_norm": 0.7672936320304871, "learning_rate": 9.363529928821508e-05, "loss": 0.0436, "step": 40850 }, { "grad_norm": 1.1796679496765137, "learning_rate": 9.363126219072757e-05, "loss": 0.0606, "step": 40860 }, { "grad_norm": 0.4600078761577606, "learning_rate": 9.362722390037775e-05, "loss": 0.0577, "step": 40870 }, { "grad_norm": 0.8522085547447205, "learning_rate": 9.3623184417276e-05, "loss": 0.0547, "step": 40880 }, { "grad_norm": 0.5350392460823059, "learning_rate": 9.36191437415328e-05, "loss": 0.0632, "step": 40890 }, { "grad_norm": 0.5324291586875916, "learning_rate": 9.361510187325857e-05, "loss": 0.0505, "step": 40900 }, { "grad_norm": 0.8670983910560608, "learning_rate": 9.361105881256385e-05, "loss": 0.0937, "step": 40910 }, { "grad_norm": 0.49690085649490356, "learning_rate": 9.360701455955918e-05, "loss": 0.0652, "step": 40920 }, { "grad_norm": 0.8731522560119629, "learning_rate": 9.360296911435509e-05, "loss": 0.0637, "step": 40930 }, { "grad_norm": 0.9212355613708496, "learning_rate": 9.359892247706222e-05, "loss": 0.0611, "step": 40940 }, { "grad_norm": 0.6214748024940491, "learning_rate": 9.35948746477912e-05, "loss": 0.0503, "step": 40950 }, { "grad_norm": 0.5276445150375366, "learning_rate": 9.359082562665266e-05, "loss": 0.0555, "step": 40960 }, { "grad_norm": 0.40660175681114197, "learning_rate": 9.358677541375734e-05, "loss": 0.0429, "step": 40970 }, { "grad_norm": 0.3892348110675812, "learning_rate": 9.358272400921596e-05, "loss": 0.0582, "step": 40980 }, { "grad_norm": 0.5519060492515564, "learning_rate": 9.357867141313927e-05, "loss": 0.0417, "step": 40990 }, { "grad_norm": 0.4018663465976715, "learning_rate": 9.357461762563807e-05, "loss": 0.057, "step": 41000 }, { "grad_norm": 0.7870166897773743, "learning_rate": 9.35705626468232e-05, "loss": 0.0448, "step": 41010 }, { "grad_norm": 0.7232425808906555, "learning_rate": 9.356650647680551e-05, "loss": 0.0528, "step": 41020 }, { "grad_norm": 0.6450142860412598, "learning_rate": 9.356244911569591e-05, "loss": 0.056, "step": 41030 }, { "grad_norm": 0.7014408111572266, "learning_rate": 9.355839056360532e-05, "loss": 0.0555, "step": 41040 }, { "grad_norm": 0.5372626781463623, "learning_rate": 9.355433082064469e-05, "loss": 0.0365, "step": 41050 }, { "grad_norm": 0.7442078590393066, "learning_rate": 9.355026988692501e-05, "loss": 0.072, "step": 41060 }, { "grad_norm": 0.3816107511520386, "learning_rate": 9.354620776255731e-05, "loss": 0.0488, "step": 41070 }, { "grad_norm": 0.4577125012874603, "learning_rate": 9.354214444765267e-05, "loss": 0.0483, "step": 41080 }, { "grad_norm": 0.7131659984588623, "learning_rate": 9.353807994232214e-05, "loss": 0.0519, "step": 41090 }, { "grad_norm": 0.5259571075439453, "learning_rate": 9.353401424667687e-05, "loss": 0.0546, "step": 41100 }, { "grad_norm": 1.0818781852722168, "learning_rate": 9.3529947360828e-05, "loss": 0.0519, "step": 41110 }, { "grad_norm": 0.8467090129852295, "learning_rate": 9.352587928488673e-05, "loss": 0.0527, "step": 41120 }, { "grad_norm": 0.5099345445632935, "learning_rate": 9.352181001896427e-05, "loss": 0.0479, "step": 41130 }, { "grad_norm": 0.6231366991996765, "learning_rate": 9.351773956317187e-05, "loss": 0.0433, "step": 41140 }, { "grad_norm": 0.5646460652351379, "learning_rate": 9.351366791762081e-05, "loss": 0.0518, "step": 41150 }, { "grad_norm": 0.5827847123146057, "learning_rate": 9.350959508242244e-05, "loss": 0.0425, "step": 41160 }, { "grad_norm": 0.8276552557945251, "learning_rate": 9.350552105768808e-05, "loss": 0.0625, "step": 41170 }, { "grad_norm": 0.705795407295227, "learning_rate": 9.350144584352911e-05, "loss": 0.0514, "step": 41180 }, { "grad_norm": 1.1366366147994995, "learning_rate": 9.349736944005696e-05, "loss": 0.0634, "step": 41190 }, { "grad_norm": 0.6188697814941406, "learning_rate": 9.349329184738307e-05, "loss": 0.0574, "step": 41200 }, { "grad_norm": 0.5792099237442017, "learning_rate": 9.348921306561891e-05, "loss": 0.0494, "step": 41210 }, { "grad_norm": 0.874944806098938, "learning_rate": 9.348513309487601e-05, "loss": 0.0689, "step": 41220 }, { "grad_norm": 0.8223043084144592, "learning_rate": 9.348105193526591e-05, "loss": 0.0449, "step": 41230 }, { "grad_norm": 0.46923962235450745, "learning_rate": 9.347696958690019e-05, "loss": 0.0687, "step": 41240 }, { "grad_norm": 0.6774945259094238, "learning_rate": 9.347288604989044e-05, "loss": 0.063, "step": 41250 }, { "grad_norm": 0.5560785531997681, "learning_rate": 9.346880132434833e-05, "loss": 0.0511, "step": 41260 }, { "grad_norm": 0.6469944715499878, "learning_rate": 9.346471541038553e-05, "loss": 0.0527, "step": 41270 }, { "grad_norm": 0.7243626117706299, "learning_rate": 9.346062830811372e-05, "loss": 0.0511, "step": 41280 }, { "grad_norm": 0.6362908482551575, "learning_rate": 9.345654001764466e-05, "loss": 0.0601, "step": 41290 }, { "grad_norm": 0.45921647548675537, "learning_rate": 9.345245053909012e-05, "loss": 0.0431, "step": 41300 }, { "grad_norm": 0.5258449912071228, "learning_rate": 9.344835987256193e-05, "loss": 0.0343, "step": 41310 }, { "grad_norm": 0.48867085576057434, "learning_rate": 9.34442680181719e-05, "loss": 0.0486, "step": 41320 }, { "grad_norm": 0.6647264957427979, "learning_rate": 9.344017497603189e-05, "loss": 0.0493, "step": 41330 }, { "grad_norm": 0.5296662449836731, "learning_rate": 9.34360807462538e-05, "loss": 0.0386, "step": 41340 }, { "grad_norm": 0.6143789291381836, "learning_rate": 9.343198532894961e-05, "loss": 0.0711, "step": 41350 }, { "grad_norm": 0.6236795783042908, "learning_rate": 9.342788872423125e-05, "loss": 0.0539, "step": 41360 }, { "grad_norm": 0.6570321917533875, "learning_rate": 9.342379093221072e-05, "loss": 0.0445, "step": 41370 }, { "grad_norm": 0.8532869815826416, "learning_rate": 9.341969195300006e-05, "loss": 0.0545, "step": 41380 }, { "grad_norm": 0.5829930901527405, "learning_rate": 9.341559178671135e-05, "loss": 0.0467, "step": 41390 }, { "grad_norm": 0.3786059617996216, "learning_rate": 9.341149043345667e-05, "loss": 0.0664, "step": 41400 }, { "grad_norm": 0.5804378986358643, "learning_rate": 9.340738789334813e-05, "loss": 0.0566, "step": 41410 }, { "grad_norm": 0.633554995059967, "learning_rate": 9.340328416649792e-05, "loss": 0.0542, "step": 41420 }, { "grad_norm": 0.41899368166923523, "learning_rate": 9.339917925301824e-05, "loss": 0.0662, "step": 41430 }, { "grad_norm": 0.5842843055725098, "learning_rate": 9.339507315302129e-05, "loss": 0.0592, "step": 41440 }, { "grad_norm": 0.4896671175956726, "learning_rate": 9.339096586661934e-05, "loss": 0.05, "step": 41450 }, { "grad_norm": 0.6282545328140259, "learning_rate": 9.338685739392469e-05, "loss": 0.0838, "step": 41460 }, { "grad_norm": 0.5833984613418579, "learning_rate": 9.338274773504965e-05, "loss": 0.051, "step": 41470 }, { "grad_norm": 0.7247313261032104, "learning_rate": 9.33786368901066e-05, "loss": 0.0474, "step": 41480 }, { "grad_norm": 0.6240229606628418, "learning_rate": 9.33745248592079e-05, "loss": 0.0671, "step": 41490 }, { "grad_norm": 0.7187198996543884, "learning_rate": 9.3370411642466e-05, "loss": 0.0621, "step": 41500 }, { "grad_norm": 0.81465744972229, "learning_rate": 9.336629723999333e-05, "loss": 0.0741, "step": 41510 }, { "grad_norm": 0.7978882789611816, "learning_rate": 9.336218165190239e-05, "loss": 0.0648, "step": 41520 }, { "grad_norm": 0.7405453324317932, "learning_rate": 9.33580648783057e-05, "loss": 0.0643, "step": 41530 }, { "grad_norm": 0.47729673981666565, "learning_rate": 9.33539469193158e-05, "loss": 0.0752, "step": 41540 }, { "grad_norm": 0.5027186274528503, "learning_rate": 9.334982777504526e-05, "loss": 0.0523, "step": 41550 }, { "grad_norm": 0.5209764242172241, "learning_rate": 9.334570744560673e-05, "loss": 0.0808, "step": 41560 }, { "grad_norm": 0.6148762106895447, "learning_rate": 9.334158593111285e-05, "loss": 0.0571, "step": 41570 }, { "grad_norm": 0.9820826053619385, "learning_rate": 9.333746323167629e-05, "loss": 0.0516, "step": 41580 }, { "grad_norm": 0.5389305353164673, "learning_rate": 9.333333934740976e-05, "loss": 0.0754, "step": 41590 }, { "grad_norm": 0.7505671977996826, "learning_rate": 9.3329214278426e-05, "loss": 0.0594, "step": 41600 }, { "grad_norm": 0.856012225151062, "learning_rate": 9.332508802483782e-05, "loss": 0.0655, "step": 41610 }, { "grad_norm": 0.6468663215637207, "learning_rate": 9.332096058675801e-05, "loss": 0.05, "step": 41620 }, { "grad_norm": 0.7207782864570618, "learning_rate": 9.33168319642994e-05, "loss": 0.0614, "step": 41630 }, { "grad_norm": 0.49943751096725464, "learning_rate": 9.331270215757488e-05, "loss": 0.076, "step": 41640 }, { "grad_norm": 0.7069061994552612, "learning_rate": 9.330857116669736e-05, "loss": 0.0525, "step": 41650 }, { "grad_norm": 0.6374081373214722, "learning_rate": 9.330443899177978e-05, "loss": 0.0493, "step": 41660 }, { "grad_norm": 0.688879668712616, "learning_rate": 9.330030563293509e-05, "loss": 0.0591, "step": 41670 }, { "grad_norm": 0.5670687556266785, "learning_rate": 9.329617109027633e-05, "loss": 0.0511, "step": 41680 }, { "grad_norm": 0.7544358968734741, "learning_rate": 9.329203536391651e-05, "loss": 0.0567, "step": 41690 }, { "grad_norm": 0.8696749806404114, "learning_rate": 9.32878984539687e-05, "loss": 0.0555, "step": 41700 }, { "grad_norm": 0.4712385833263397, "learning_rate": 9.328376036054602e-05, "loss": 0.0735, "step": 41710 }, { "grad_norm": 0.6915056109428406, "learning_rate": 9.327962108376158e-05, "loss": 0.0575, "step": 41720 }, { "grad_norm": 0.4861137270927429, "learning_rate": 9.327548062372858e-05, "loss": 0.0517, "step": 41730 }, { "grad_norm": 0.702070415019989, "learning_rate": 9.327133898056017e-05, "loss": 0.0492, "step": 41740 }, { "grad_norm": 0.4695308208465576, "learning_rate": 9.326719615436963e-05, "loss": 0.0734, "step": 41750 }, { "grad_norm": 0.7464945316314697, "learning_rate": 9.326305214527019e-05, "loss": 0.0471, "step": 41760 }, { "grad_norm": 0.6828586459159851, "learning_rate": 9.325890695337516e-05, "loss": 0.0596, "step": 41770 }, { "grad_norm": 0.46143385767936707, "learning_rate": 9.325476057879787e-05, "loss": 0.0597, "step": 41780 }, { "grad_norm": 0.8865336179733276, "learning_rate": 9.325061302165167e-05, "loss": 0.0494, "step": 41790 }, { "grad_norm": 0.7477893233299255, "learning_rate": 9.324646428204997e-05, "loss": 0.0571, "step": 41800 }, { "grad_norm": 0.8752841949462891, "learning_rate": 9.324231436010617e-05, "loss": 0.0484, "step": 41810 }, { "grad_norm": 1.1193246841430664, "learning_rate": 9.323816325593375e-05, "loss": 0.0493, "step": 41820 }, { "grad_norm": 0.6734915375709534, "learning_rate": 9.323401096964618e-05, "loss": 0.047, "step": 41830 }, { "grad_norm": 0.5134198665618896, "learning_rate": 9.3229857501357e-05, "loss": 0.0488, "step": 41840 }, { "grad_norm": 0.5930250287055969, "learning_rate": 9.322570285117975e-05, "loss": 0.049, "step": 41850 }, { "grad_norm": 0.7029730081558228, "learning_rate": 9.322154701922803e-05, "loss": 0.0598, "step": 41860 }, { "grad_norm": 0.6512366533279419, "learning_rate": 9.321739000561545e-05, "loss": 0.0631, "step": 41870 }, { "grad_norm": 0.4626116454601288, "learning_rate": 9.321323181045567e-05, "loss": 0.0634, "step": 41880 }, { "grad_norm": 0.5308002233505249, "learning_rate": 9.320907243386237e-05, "loss": 0.0487, "step": 41890 }, { "grad_norm": 0.5349618196487427, "learning_rate": 9.320491187594925e-05, "loss": 0.0564, "step": 41900 }, { "grad_norm": 1.5585492849349976, "learning_rate": 9.320075013683007e-05, "loss": 0.0603, "step": 41910 }, { "grad_norm": 0.32236629724502563, "learning_rate": 9.319658721661861e-05, "loss": 0.0406, "step": 41920 }, { "grad_norm": 0.5544556975364685, "learning_rate": 9.319242311542868e-05, "loss": 0.0459, "step": 41930 }, { "grad_norm": 0.620168149471283, "learning_rate": 9.318825783337415e-05, "loss": 0.0573, "step": 41940 }, { "grad_norm": 0.45497703552246094, "learning_rate": 9.318409137056886e-05, "loss": 0.043, "step": 41950 }, { "grad_norm": 0.4064381718635559, "learning_rate": 9.317992372712674e-05, "loss": 0.0462, "step": 41960 }, { "grad_norm": 1.1811871528625488, "learning_rate": 9.317575490316173e-05, "loss": 0.0496, "step": 41970 }, { "grad_norm": 0.5775932669639587, "learning_rate": 9.317158489878779e-05, "loss": 0.0451, "step": 41980 }, { "grad_norm": 0.38494235277175903, "learning_rate": 9.316741371411895e-05, "loss": 0.0508, "step": 41990 }, { "grad_norm": 0.32763129472732544, "learning_rate": 9.316324134926923e-05, "loss": 0.0466, "step": 42000 }, { "grad_norm": 1.0575432777404785, "learning_rate": 9.31590678043527e-05, "loss": 0.0631, "step": 42010 }, { "grad_norm": 0.4824555218219757, "learning_rate": 9.315489307948349e-05, "loss": 0.0602, "step": 42020 }, { "grad_norm": 0.6503790616989136, "learning_rate": 9.31507171747757e-05, "loss": 0.0601, "step": 42030 }, { "grad_norm": 0.5201740860939026, "learning_rate": 9.314654009034352e-05, "loss": 0.0488, "step": 42040 }, { "grad_norm": 0.5292661786079407, "learning_rate": 9.314236182630114e-05, "loss": 0.0645, "step": 42050 }, { "grad_norm": 0.6237610578536987, "learning_rate": 9.31381823827628e-05, "loss": 0.0727, "step": 42060 }, { "grad_norm": 0.7298041582107544, "learning_rate": 9.313400175984275e-05, "loss": 0.0572, "step": 42070 }, { "grad_norm": 0.4824245870113373, "learning_rate": 9.31298199576553e-05, "loss": 0.0516, "step": 42080 }, { "grad_norm": 0.8254294395446777, "learning_rate": 9.312563697631478e-05, "loss": 0.065, "step": 42090 }, { "grad_norm": 0.6277993321418762, "learning_rate": 9.312145281593554e-05, "loss": 0.0482, "step": 42100 }, { "grad_norm": 0.5611412525177002, "learning_rate": 9.311726747663199e-05, "loss": 0.0526, "step": 42110 }, { "grad_norm": 0.7374159097671509, "learning_rate": 9.311308095851854e-05, "loss": 0.0816, "step": 42120 }, { "grad_norm": 0.5809485912322998, "learning_rate": 9.310889326170967e-05, "loss": 0.0562, "step": 42130 }, { "grad_norm": 0.6178372502326965, "learning_rate": 9.310470438631982e-05, "loss": 0.0559, "step": 42140 }, { "grad_norm": 0.7877609133720398, "learning_rate": 9.310051433246357e-05, "loss": 0.0528, "step": 42150 }, { "grad_norm": 0.37523019313812256, "learning_rate": 9.309632310025544e-05, "loss": 0.0556, "step": 42160 }, { "grad_norm": 0.5129857659339905, "learning_rate": 9.309213068981003e-05, "loss": 0.0568, "step": 42170 }, { "grad_norm": 0.66194748878479, "learning_rate": 9.308793710124195e-05, "loss": 0.067, "step": 42180 }, { "grad_norm": 0.7238625884056091, "learning_rate": 9.308374233466587e-05, "loss": 0.0691, "step": 42190 }, { "grad_norm": 0.7074739336967468, "learning_rate": 9.307954639019644e-05, "loss": 0.0608, "step": 42200 }, { "grad_norm": 0.5496061444282532, "learning_rate": 9.307534926794842e-05, "loss": 0.0495, "step": 42210 }, { "grad_norm": 0.5282689332962036, "learning_rate": 9.307115096803654e-05, "loss": 0.0448, "step": 42220 }, { "grad_norm": 0.637415885925293, "learning_rate": 9.306695149057554e-05, "loss": 0.041, "step": 42230 }, { "grad_norm": 0.43532466888427734, "learning_rate": 9.306275083568029e-05, "loss": 0.0439, "step": 42240 }, { "grad_norm": 0.552408754825592, "learning_rate": 9.30585490034656e-05, "loss": 0.0567, "step": 42250 }, { "grad_norm": 0.4456634819507599, "learning_rate": 9.305434599404637e-05, "loss": 0.0547, "step": 42260 }, { "grad_norm": 0.6483244299888611, "learning_rate": 9.305014180753749e-05, "loss": 0.0551, "step": 42270 }, { "grad_norm": 0.6093001961708069, "learning_rate": 9.304593644405388e-05, "loss": 0.0522, "step": 42280 }, { "grad_norm": 1.3443831205368042, "learning_rate": 9.304172990371056e-05, "loss": 0.0637, "step": 42290 }, { "grad_norm": 0.5721998810768127, "learning_rate": 9.303752218662251e-05, "loss": 0.0575, "step": 42300 }, { "grad_norm": 0.4952378571033478, "learning_rate": 9.303331329290479e-05, "loss": 0.0518, "step": 42310 }, { "grad_norm": 0.46710383892059326, "learning_rate": 9.302910322267244e-05, "loss": 0.0524, "step": 42320 }, { "grad_norm": 0.464115172624588, "learning_rate": 9.302489197604056e-05, "loss": 0.0495, "step": 42330 }, { "grad_norm": 0.6167890429496765, "learning_rate": 9.302067955312432e-05, "loss": 0.0509, "step": 42340 }, { "grad_norm": 0.746770441532135, "learning_rate": 9.301646595403884e-05, "loss": 0.0515, "step": 42350 }, { "grad_norm": 0.8115832209587097, "learning_rate": 9.301225117889936e-05, "loss": 0.0488, "step": 42360 }, { "grad_norm": 0.49287959933280945, "learning_rate": 9.300803522782108e-05, "loss": 0.0658, "step": 42370 }, { "grad_norm": 0.6704860329627991, "learning_rate": 9.300381810091929e-05, "loss": 0.0519, "step": 42380 }, { "grad_norm": 0.39795875549316406, "learning_rate": 9.299959979830925e-05, "loss": 0.0578, "step": 42390 }, { "grad_norm": 0.7114341855049133, "learning_rate": 9.299538032010631e-05, "loss": 0.0579, "step": 42400 }, { "grad_norm": 0.8155449628829956, "learning_rate": 9.299115966642584e-05, "loss": 0.0552, "step": 42410 }, { "grad_norm": 0.984279990196228, "learning_rate": 9.298693783738319e-05, "loss": 0.0504, "step": 42420 }, { "grad_norm": 0.6996873021125793, "learning_rate": 9.298271483309383e-05, "loss": 0.0631, "step": 42430 }, { "grad_norm": 0.6011309623718262, "learning_rate": 9.297849065367318e-05, "loss": 0.0505, "step": 42440 }, { "grad_norm": 0.6370531320571899, "learning_rate": 9.297426529923676e-05, "loss": 0.0643, "step": 42450 }, { "grad_norm": 0.5469176769256592, "learning_rate": 9.297003876990006e-05, "loss": 0.0501, "step": 42460 }, { "grad_norm": 0.46699219942092896, "learning_rate": 9.296581106577864e-05, "loss": 0.0496, "step": 42470 }, { "grad_norm": 0.549514889717102, "learning_rate": 9.296158218698809e-05, "loss": 0.0599, "step": 42480 }, { "grad_norm": 0.9174487590789795, "learning_rate": 9.295735213364403e-05, "loss": 0.0641, "step": 42490 }, { "grad_norm": 0.6269227266311646, "learning_rate": 9.295312090586208e-05, "loss": 0.0457, "step": 42500 }, { "grad_norm": 1.261903166770935, "learning_rate": 9.294888850375796e-05, "loss": 0.0761, "step": 42510 }, { "grad_norm": 0.6801533699035645, "learning_rate": 9.294465492744737e-05, "loss": 0.0581, "step": 42520 }, { "grad_norm": 0.6506511569023132, "learning_rate": 9.294042017704603e-05, "loss": 0.0639, "step": 42530 }, { "grad_norm": 0.5401962995529175, "learning_rate": 9.293618425266974e-05, "loss": 0.0475, "step": 42540 }, { "grad_norm": 0.579524576663971, "learning_rate": 9.293194715443429e-05, "loss": 0.0524, "step": 42550 }, { "grad_norm": 0.7067296504974365, "learning_rate": 9.292770888245557e-05, "loss": 0.063, "step": 42560 }, { "grad_norm": 0.4653911590576172, "learning_rate": 9.292346943684937e-05, "loss": 0.0562, "step": 42570 }, { "grad_norm": 0.3627658486366272, "learning_rate": 9.291922881773167e-05, "loss": 0.049, "step": 42580 }, { "grad_norm": 0.9543336629867554, "learning_rate": 9.291498702521838e-05, "loss": 0.0482, "step": 42590 }, { "grad_norm": 0.44462165236473083, "learning_rate": 9.291074405942546e-05, "loss": 0.0511, "step": 42600 }, { "grad_norm": 0.5203724503517151, "learning_rate": 9.290649992046892e-05, "loss": 0.0481, "step": 42610 }, { "grad_norm": 0.710424542427063, "learning_rate": 9.29022546084648e-05, "loss": 0.0425, "step": 42620 }, { "grad_norm": 0.8641668558120728, "learning_rate": 9.289800812352913e-05, "loss": 0.0503, "step": 42630 }, { "grad_norm": 0.465859055519104, "learning_rate": 9.289376046577806e-05, "loss": 0.0492, "step": 42640 }, { "grad_norm": 0.507477343082428, "learning_rate": 9.288951163532769e-05, "loss": 0.0527, "step": 42650 }, { "grad_norm": 0.33012303709983826, "learning_rate": 9.288526163229416e-05, "loss": 0.057, "step": 42660 }, { "grad_norm": 0.7720236778259277, "learning_rate": 9.288101045679371e-05, "loss": 0.0614, "step": 42670 }, { "grad_norm": 0.637692391872406, "learning_rate": 9.287675810894255e-05, "loss": 0.0513, "step": 42680 }, { "grad_norm": 0.4240078926086426, "learning_rate": 9.287250458885692e-05, "loss": 0.0637, "step": 42690 }, { "grad_norm": 0.35822993516921997, "learning_rate": 9.286824989665313e-05, "loss": 0.0427, "step": 42700 }, { "grad_norm": 0.5273385643959045, "learning_rate": 9.286399403244748e-05, "loss": 0.0432, "step": 42710 }, { "grad_norm": 0.9022397994995117, "learning_rate": 9.285973699635634e-05, "loss": 0.0553, "step": 42720 }, { "grad_norm": 0.789042055606842, "learning_rate": 9.285547878849609e-05, "loss": 0.056, "step": 42730 }, { "grad_norm": 0.47655874490737915, "learning_rate": 9.285121940898317e-05, "loss": 0.0411, "step": 42740 }, { "grad_norm": 0.8393275141716003, "learning_rate": 9.284695885793398e-05, "loss": 0.053, "step": 42750 }, { "grad_norm": 0.49927404522895813, "learning_rate": 9.284269713546505e-05, "loss": 0.0589, "step": 42760 }, { "grad_norm": 0.5827969908714294, "learning_rate": 9.283843424169288e-05, "loss": 0.0383, "step": 42770 }, { "grad_norm": 0.554665207862854, "learning_rate": 9.283417017673399e-05, "loss": 0.0462, "step": 42780 }, { "grad_norm": 0.3729223608970642, "learning_rate": 9.282990494070499e-05, "loss": 0.0646, "step": 42790 }, { "grad_norm": 0.4150024354457855, "learning_rate": 9.282563853372249e-05, "loss": 0.0511, "step": 42800 }, { "grad_norm": 0.4324381351470947, "learning_rate": 9.28213709559031e-05, "loss": 0.0545, "step": 42810 }, { "grad_norm": 0.4510764479637146, "learning_rate": 9.281710220736353e-05, "loss": 0.0766, "step": 42820 }, { "grad_norm": 0.7614931464195251, "learning_rate": 9.281283228822047e-05, "loss": 0.0496, "step": 42830 }, { "grad_norm": 0.5671886801719666, "learning_rate": 9.280856119859067e-05, "loss": 0.0652, "step": 42840 }, { "grad_norm": 0.49715879559516907, "learning_rate": 9.280428893859088e-05, "loss": 0.0447, "step": 42850 }, { "grad_norm": 0.4835386574268341, "learning_rate": 9.28000155083379e-05, "loss": 0.0437, "step": 42860 }, { "grad_norm": 0.5399529337882996, "learning_rate": 9.279574090794858e-05, "loss": 0.0507, "step": 42870 }, { "grad_norm": 0.42038920521736145, "learning_rate": 9.279146513753979e-05, "loss": 0.0397, "step": 42880 }, { "grad_norm": 0.7300376296043396, "learning_rate": 9.278718819722841e-05, "loss": 0.0535, "step": 42890 }, { "grad_norm": 0.4520849287509918, "learning_rate": 9.27829100871314e-05, "loss": 0.0669, "step": 42900 }, { "grad_norm": 0.5003454685211182, "learning_rate": 9.277863080736569e-05, "loss": 0.0522, "step": 42910 }, { "grad_norm": 0.6944496035575867, "learning_rate": 9.277435035804827e-05, "loss": 0.0451, "step": 42920 }, { "grad_norm": 1.4071259498596191, "learning_rate": 9.277006873929619e-05, "loss": 0.0753, "step": 42930 }, { "grad_norm": 0.766538679599762, "learning_rate": 9.276578595122649e-05, "loss": 0.0636, "step": 42940 }, { "grad_norm": 0.3722619116306305, "learning_rate": 9.276150199395628e-05, "loss": 0.0319, "step": 42950 }, { "grad_norm": 0.4204968214035034, "learning_rate": 9.275721686760267e-05, "loss": 0.0569, "step": 42960 }, { "grad_norm": 0.507361888885498, "learning_rate": 9.275293057228281e-05, "loss": 0.0483, "step": 42970 }, { "grad_norm": 0.5598690509796143, "learning_rate": 9.274864310811388e-05, "loss": 0.056, "step": 42980 }, { "grad_norm": 0.5937574505805969, "learning_rate": 9.274435447521313e-05, "loss": 0.0515, "step": 42990 }, { "grad_norm": 0.46899139881134033, "learning_rate": 9.274006467369778e-05, "loss": 0.0562, "step": 43000 }, { "grad_norm": 0.6031330227851868, "learning_rate": 9.273577370368512e-05, "loss": 0.0712, "step": 43010 }, { "grad_norm": 0.43848586082458496, "learning_rate": 9.273148156529246e-05, "loss": 0.047, "step": 43020 }, { "grad_norm": 0.43481871485710144, "learning_rate": 9.272718825863714e-05, "loss": 0.045, "step": 43030 }, { "grad_norm": 0.49929213523864746, "learning_rate": 9.272289378383654e-05, "loss": 0.0427, "step": 43040 }, { "grad_norm": 0.6656663417816162, "learning_rate": 9.271859814100808e-05, "loss": 0.0504, "step": 43050 }, { "grad_norm": 0.28778210282325745, "learning_rate": 9.27143013302692e-05, "loss": 0.0375, "step": 43060 }, { "grad_norm": 0.9228733777999878, "learning_rate": 9.271000335173736e-05, "loss": 0.0444, "step": 43070 }, { "grad_norm": 0.3495620787143707, "learning_rate": 9.270570420553008e-05, "loss": 0.0526, "step": 43080 }, { "grad_norm": 0.4487941861152649, "learning_rate": 9.27014038917649e-05, "loss": 0.0488, "step": 43090 }, { "grad_norm": 0.5736156702041626, "learning_rate": 9.269710241055937e-05, "loss": 0.0595, "step": 43100 }, { "grad_norm": 0.5008202791213989, "learning_rate": 9.26927997620311e-05, "loss": 0.0545, "step": 43110 }, { "grad_norm": 0.4322645962238312, "learning_rate": 9.268849594629773e-05, "loss": 0.0447, "step": 43120 }, { "grad_norm": 0.5798522233963013, "learning_rate": 9.268419096347692e-05, "loss": 0.0451, "step": 43130 }, { "grad_norm": 0.7008825540542603, "learning_rate": 9.267988481368635e-05, "loss": 0.0451, "step": 43140 }, { "grad_norm": 0.6959802508354187, "learning_rate": 9.267557749704377e-05, "loss": 0.0435, "step": 43150 }, { "grad_norm": 0.5525510907173157, "learning_rate": 9.267126901366694e-05, "loss": 0.0392, "step": 43160 }, { "grad_norm": 0.563388466835022, "learning_rate": 9.266695936367365e-05, "loss": 0.0483, "step": 43170 }, { "grad_norm": 0.5107530951499939, "learning_rate": 9.26626485471817e-05, "loss": 0.0506, "step": 43180 }, { "grad_norm": 0.5974330306053162, "learning_rate": 9.265833656430898e-05, "loss": 0.0487, "step": 43190 }, { "grad_norm": 0.3394741415977478, "learning_rate": 9.265402341517336e-05, "loss": 0.0464, "step": 43200 }, { "grad_norm": 0.7584654092788696, "learning_rate": 9.264970909989276e-05, "loss": 0.0503, "step": 43210 }, { "grad_norm": 0.542828381061554, "learning_rate": 9.264539361858515e-05, "loss": 0.0682, "step": 43220 }, { "grad_norm": 0.5024333000183105, "learning_rate": 9.26410769713685e-05, "loss": 0.0493, "step": 43230 }, { "grad_norm": 0.3786011338233948, "learning_rate": 9.263675915836081e-05, "loss": 0.0418, "step": 43240 }, { "grad_norm": 0.3871783912181854, "learning_rate": 9.263244017968014e-05, "loss": 0.0587, "step": 43250 }, { "grad_norm": 0.402093768119812, "learning_rate": 9.262812003544459e-05, "loss": 0.0415, "step": 43260 }, { "grad_norm": 0.4638914167881012, "learning_rate": 9.262379872577225e-05, "loss": 0.048, "step": 43270 }, { "grad_norm": 0.5966187119483948, "learning_rate": 9.261947625078126e-05, "loss": 0.0453, "step": 43280 }, { "grad_norm": 0.5305048823356628, "learning_rate": 9.26151526105898e-05, "loss": 0.0562, "step": 43290 }, { "grad_norm": 0.605937659740448, "learning_rate": 9.261082780531607e-05, "loss": 0.039, "step": 43300 }, { "grad_norm": 0.41368964314460754, "learning_rate": 9.260650183507834e-05, "loss": 0.0573, "step": 43310 }, { "grad_norm": 0.40455251932144165, "learning_rate": 9.260217469999483e-05, "loss": 0.0462, "step": 43320 }, { "grad_norm": 0.628331184387207, "learning_rate": 9.259784640018388e-05, "loss": 0.0766, "step": 43330 }, { "grad_norm": 0.6310299634933472, "learning_rate": 9.25935169357638e-05, "loss": 0.0465, "step": 43340 }, { "grad_norm": 0.5907789468765259, "learning_rate": 9.258918630685299e-05, "loss": 0.0468, "step": 43350 }, { "grad_norm": 0.6786477565765381, "learning_rate": 9.258485451356981e-05, "loss": 0.0511, "step": 43360 }, { "grad_norm": 0.7888342142105103, "learning_rate": 9.258052155603272e-05, "loss": 0.0509, "step": 43370 }, { "grad_norm": 0.7441941499710083, "learning_rate": 9.257618743436016e-05, "loss": 0.0544, "step": 43380 }, { "grad_norm": 0.5195417404174805, "learning_rate": 9.257185214867063e-05, "loss": 0.0548, "step": 43390 }, { "grad_norm": 0.7500154972076416, "learning_rate": 9.256751569908266e-05, "loss": 0.0523, "step": 43400 }, { "grad_norm": 0.9338284730911255, "learning_rate": 9.25631780857148e-05, "loss": 0.0529, "step": 43410 }, { "grad_norm": 0.6561309099197388, "learning_rate": 9.255883930868563e-05, "loss": 0.054, "step": 43420 }, { "grad_norm": 0.5465420484542847, "learning_rate": 9.25544993681138e-05, "loss": 0.0628, "step": 43430 }, { "grad_norm": 0.4980715811252594, "learning_rate": 9.255015826411794e-05, "loss": 0.0651, "step": 43440 }, { "grad_norm": 0.6431137919425964, "learning_rate": 9.254581599681674e-05, "loss": 0.0367, "step": 43450 }, { "grad_norm": 0.5439369082450867, "learning_rate": 9.25414725663289e-05, "loss": 0.0486, "step": 43460 }, { "grad_norm": 0.7930107116699219, "learning_rate": 9.253712797277319e-05, "loss": 0.0451, "step": 43470 }, { "grad_norm": 0.64630526304245, "learning_rate": 9.253278221626839e-05, "loss": 0.0564, "step": 43480 }, { "grad_norm": 0.48800408840179443, "learning_rate": 9.25284352969333e-05, "loss": 0.0492, "step": 43490 }, { "grad_norm": 0.5146664977073669, "learning_rate": 9.252408721488676e-05, "loss": 0.056, "step": 43500 }, { "grad_norm": 0.7283796072006226, "learning_rate": 9.251973797024764e-05, "loss": 0.0601, "step": 43510 }, { "grad_norm": 0.6324838995933533, "learning_rate": 9.251538756313488e-05, "loss": 0.0831, "step": 43520 }, { "grad_norm": 0.8210378289222717, "learning_rate": 9.251103599366739e-05, "loss": 0.0624, "step": 43530 }, { "grad_norm": 0.5703988671302795, "learning_rate": 9.250668326196415e-05, "loss": 0.0479, "step": 43540 }, { "grad_norm": 0.9770529866218567, "learning_rate": 9.250232936814413e-05, "loss": 0.0542, "step": 43550 }, { "grad_norm": 1.158162236213684, "learning_rate": 9.249797431232643e-05, "loss": 0.087, "step": 43560 }, { "grad_norm": 0.3940739631652832, "learning_rate": 9.249361809463007e-05, "loss": 0.038, "step": 43570 }, { "grad_norm": 0.6699449419975281, "learning_rate": 9.248926071517414e-05, "loss": 0.0475, "step": 43580 }, { "grad_norm": 0.38709819316864014, "learning_rate": 9.24849021740778e-05, "loss": 0.0426, "step": 43590 }, { "grad_norm": 0.5384904146194458, "learning_rate": 9.248054247146019e-05, "loss": 0.0568, "step": 43600 }, { "grad_norm": 0.44255781173706055, "learning_rate": 9.24761816074405e-05, "loss": 0.0483, "step": 43610 }, { "grad_norm": 0.31454339623451233, "learning_rate": 9.247181958213797e-05, "loss": 0.0354, "step": 43620 }, { "grad_norm": 0.5080539584159851, "learning_rate": 9.246745639567185e-05, "loss": 0.0503, "step": 43630 }, { "grad_norm": 0.4924686849117279, "learning_rate": 9.246309204816142e-05, "loss": 0.0535, "step": 43640 }, { "grad_norm": 0.7865267395973206, "learning_rate": 9.245872653972602e-05, "loss": 0.0529, "step": 43650 }, { "grad_norm": 0.9364225268363953, "learning_rate": 9.245435987048498e-05, "loss": 0.0562, "step": 43660 }, { "grad_norm": 0.5856350660324097, "learning_rate": 9.244999204055769e-05, "loss": 0.0477, "step": 43670 }, { "grad_norm": 0.3982318341732025, "learning_rate": 9.244562305006354e-05, "loss": 0.04, "step": 43680 }, { "grad_norm": 0.6455264687538147, "learning_rate": 9.244125289912205e-05, "loss": 0.0475, "step": 43690 }, { "grad_norm": 0.7299463152885437, "learning_rate": 9.243688158785263e-05, "loss": 0.0487, "step": 43700 }, { "grad_norm": 0.5987322926521301, "learning_rate": 9.243250911637479e-05, "loss": 0.0648, "step": 43710 }, { "grad_norm": 0.4317362606525421, "learning_rate": 9.242813548480811e-05, "loss": 0.0445, "step": 43720 }, { "grad_norm": 0.5198013186454773, "learning_rate": 9.242376069327215e-05, "loss": 0.0564, "step": 43730 }, { "grad_norm": 1.491249442100525, "learning_rate": 9.24193847418865e-05, "loss": 0.0467, "step": 43740 }, { "grad_norm": 0.7302250862121582, "learning_rate": 9.24150076307708e-05, "loss": 0.0496, "step": 43750 }, { "grad_norm": 0.5286884903907776, "learning_rate": 9.241062936004474e-05, "loss": 0.0408, "step": 43760 }, { "grad_norm": 0.5764294266700745, "learning_rate": 9.240624992982801e-05, "loss": 0.0453, "step": 43770 }, { "grad_norm": 0.37911707162857056, "learning_rate": 9.240186934024034e-05, "loss": 0.0434, "step": 43780 }, { "grad_norm": 0.7078136801719666, "learning_rate": 9.239748759140148e-05, "loss": 0.0402, "step": 43790 }, { "grad_norm": 0.5937661528587341, "learning_rate": 9.239310468343123e-05, "loss": 0.059, "step": 43800 }, { "grad_norm": 0.5157787203788757, "learning_rate": 9.238872061644943e-05, "loss": 0.0474, "step": 43810 }, { "grad_norm": 0.9252826571464539, "learning_rate": 9.238433539057594e-05, "loss": 0.0646, "step": 43820 }, { "grad_norm": 1.514417052268982, "learning_rate": 9.237994900593065e-05, "loss": 0.0411, "step": 43830 }, { "grad_norm": 0.6011796593666077, "learning_rate": 9.237556146263346e-05, "loss": 0.0579, "step": 43840 }, { "grad_norm": 1.3666425943374634, "learning_rate": 9.237117276080435e-05, "loss": 0.0528, "step": 43850 }, { "grad_norm": 0.4733925759792328, "learning_rate": 9.236678290056328e-05, "loss": 0.0424, "step": 43860 }, { "grad_norm": 0.7257065176963806, "learning_rate": 9.23623918820303e-05, "loss": 0.0465, "step": 43870 }, { "grad_norm": 0.6989455223083496, "learning_rate": 9.235799970532544e-05, "loss": 0.0664, "step": 43880 }, { "grad_norm": 0.5806035995483398, "learning_rate": 9.235360637056878e-05, "loss": 0.0434, "step": 43890 }, { "grad_norm": 0.4293154180049896, "learning_rate": 9.234921187788041e-05, "loss": 0.0445, "step": 43900 }, { "grad_norm": 0.5902295112609863, "learning_rate": 9.234481622738052e-05, "loss": 0.0494, "step": 43910 }, { "grad_norm": 0.6484107971191406, "learning_rate": 9.234041941918927e-05, "loss": 0.0458, "step": 43920 }, { "grad_norm": 1.0706146955490112, "learning_rate": 9.233602145342684e-05, "loss": 0.0542, "step": 43930 }, { "grad_norm": 0.5166855454444885, "learning_rate": 9.233162233021349e-05, "loss": 0.046, "step": 43940 }, { "grad_norm": 0.5501853227615356, "learning_rate": 9.23272220496695e-05, "loss": 0.0461, "step": 43950 }, { "grad_norm": 0.4848954677581787, "learning_rate": 9.232282061191515e-05, "loss": 0.0576, "step": 43960 }, { "grad_norm": 0.5312901735305786, "learning_rate": 9.231841801707079e-05, "loss": 0.0464, "step": 43970 }, { "grad_norm": 0.49802395701408386, "learning_rate": 9.231401426525679e-05, "loss": 0.0616, "step": 43980 }, { "grad_norm": 0.45840370655059814, "learning_rate": 9.230960935659352e-05, "loss": 0.0464, "step": 43990 }, { "grad_norm": 0.5931670069694519, "learning_rate": 9.230520329120142e-05, "loss": 0.052, "step": 44000 }, { "grad_norm": 0.3842460811138153, "learning_rate": 9.230079606920098e-05, "loss": 0.0399, "step": 44010 }, { "grad_norm": 0.6650837063789368, "learning_rate": 9.229638769071265e-05, "loss": 0.0457, "step": 44020 }, { "grad_norm": 0.5035260915756226, "learning_rate": 9.229197815585697e-05, "loss": 0.0642, "step": 44030 }, { "grad_norm": 0.6481939554214478, "learning_rate": 9.228756746475451e-05, "loss": 0.0549, "step": 44040 }, { "grad_norm": 0.6789657473564148, "learning_rate": 9.228315561752583e-05, "loss": 0.0657, "step": 44050 }, { "grad_norm": 1.903447151184082, "learning_rate": 9.227874261429157e-05, "loss": 0.0564, "step": 44060 }, { "grad_norm": 1.057673454284668, "learning_rate": 9.227432845517235e-05, "loss": 0.0595, "step": 44070 }, { "grad_norm": 0.47142356634140015, "learning_rate": 9.22699131402889e-05, "loss": 0.0636, "step": 44080 }, { "grad_norm": 0.3982986807823181, "learning_rate": 9.226549666976188e-05, "loss": 0.0494, "step": 44090 }, { "grad_norm": 0.4143809378147125, "learning_rate": 9.226107904371208e-05, "loss": 0.0558, "step": 44100 }, { "grad_norm": 0.9524703621864319, "learning_rate": 9.225666026226024e-05, "loss": 0.0452, "step": 44110 }, { "grad_norm": 0.5810269117355347, "learning_rate": 9.225224032552719e-05, "loss": 0.0458, "step": 44120 }, { "grad_norm": 0.7444759011268616, "learning_rate": 9.224781923363374e-05, "loss": 0.0564, "step": 44130 }, { "grad_norm": 0.3346802294254303, "learning_rate": 9.224339698670081e-05, "loss": 0.0395, "step": 44140 }, { "grad_norm": 0.5817285776138306, "learning_rate": 9.223897358484929e-05, "loss": 0.0436, "step": 44150 }, { "grad_norm": 0.7485262751579285, "learning_rate": 9.223454902820008e-05, "loss": 0.0535, "step": 44160 }, { "grad_norm": 1.2371701002120972, "learning_rate": 9.223012331687416e-05, "loss": 0.0753, "step": 44170 }, { "grad_norm": 0.828220784664154, "learning_rate": 9.222569645099254e-05, "loss": 0.0612, "step": 44180 }, { "grad_norm": 0.48915672302246094, "learning_rate": 9.222126843067624e-05, "loss": 0.0617, "step": 44190 }, { "grad_norm": 0.4502393305301666, "learning_rate": 9.221683925604634e-05, "loss": 0.0415, "step": 44200 }, { "grad_norm": 0.37716609239578247, "learning_rate": 9.221240892722389e-05, "loss": 0.064, "step": 44210 }, { "grad_norm": 0.5603666305541992, "learning_rate": 9.220797744433006e-05, "loss": 0.0815, "step": 44220 }, { "grad_norm": 0.7041752934455872, "learning_rate": 9.220354480748597e-05, "loss": 0.0767, "step": 44230 }, { "grad_norm": 0.7977063655853271, "learning_rate": 9.219911101681282e-05, "loss": 0.0663, "step": 44240 }, { "grad_norm": 0.7283099889755249, "learning_rate": 9.219467607243183e-05, "loss": 0.0683, "step": 44250 }, { "grad_norm": 0.6410419940948486, "learning_rate": 9.219023997446426e-05, "loss": 0.0508, "step": 44260 }, { "grad_norm": 0.7628041505813599, "learning_rate": 9.218580272303137e-05, "loss": 0.0511, "step": 44270 }, { "grad_norm": 0.47449904680252075, "learning_rate": 9.218136431825448e-05, "loss": 0.0613, "step": 44280 }, { "grad_norm": 0.5495802760124207, "learning_rate": 9.217692476025495e-05, "loss": 0.0503, "step": 44290 }, { "grad_norm": 0.5144543051719666, "learning_rate": 9.217248404915412e-05, "loss": 0.0657, "step": 44300 }, { "grad_norm": 0.4325868785381317, "learning_rate": 9.216804218507343e-05, "loss": 0.0501, "step": 44310 }, { "grad_norm": 0.41187235713005066, "learning_rate": 9.21635991681343e-05, "loss": 0.0567, "step": 44320 }, { "grad_norm": 0.4179842174053192, "learning_rate": 9.215915499845822e-05, "loss": 0.0478, "step": 44330 }, { "grad_norm": 0.937664806842804, "learning_rate": 9.215470967616668e-05, "loss": 0.051, "step": 44340 }, { "grad_norm": 0.5241880416870117, "learning_rate": 9.215026320138123e-05, "loss": 0.0455, "step": 44350 }, { "grad_norm": 0.48131582140922546, "learning_rate": 9.21458155742234e-05, "loss": 0.0766, "step": 44360 }, { "grad_norm": 0.569781482219696, "learning_rate": 9.21413667948148e-05, "loss": 0.0476, "step": 44370 }, { "grad_norm": 0.6391165256500244, "learning_rate": 9.213691686327707e-05, "loss": 0.0565, "step": 44380 }, { "grad_norm": 0.6651092171669006, "learning_rate": 9.213246577973187e-05, "loss": 0.0592, "step": 44390 }, { "grad_norm": 0.49528613686561584, "learning_rate": 9.212801354430088e-05, "loss": 0.0588, "step": 44400 }, { "grad_norm": 0.5605286359786987, "learning_rate": 9.212356015710581e-05, "loss": 0.0479, "step": 44410 }, { "grad_norm": 0.7258509993553162, "learning_rate": 9.211910561826845e-05, "loss": 0.054, "step": 44420 }, { "grad_norm": 0.7952747941017151, "learning_rate": 9.211464992791057e-05, "loss": 0.063, "step": 44430 }, { "grad_norm": 1.1260184049606323, "learning_rate": 9.211019308615396e-05, "loss": 0.0619, "step": 44440 }, { "grad_norm": 0.6040921211242676, "learning_rate": 9.21057350931205e-05, "loss": 0.0436, "step": 44450 }, { "grad_norm": 0.5775332450866699, "learning_rate": 9.210127594893207e-05, "loss": 0.0444, "step": 44460 }, { "grad_norm": 0.7367773056030273, "learning_rate": 9.209681565371058e-05, "loss": 0.0481, "step": 44470 }, { "grad_norm": 0.9045017957687378, "learning_rate": 9.209235420757794e-05, "loss": 0.0582, "step": 44480 }, { "grad_norm": 0.490456223487854, "learning_rate": 9.208789161065616e-05, "loss": 0.0522, "step": 44490 }, { "grad_norm": 0.5939602851867676, "learning_rate": 9.208342786306725e-05, "loss": 0.0464, "step": 44500 }, { "grad_norm": 0.5622963309288025, "learning_rate": 9.207896296493321e-05, "loss": 0.053, "step": 44510 }, { "grad_norm": 0.7435336709022522, "learning_rate": 9.207449691637616e-05, "loss": 0.0674, "step": 44520 }, { "grad_norm": 0.718758225440979, "learning_rate": 9.207002971751814e-05, "loss": 0.0588, "step": 44530 }, { "grad_norm": 1.0167946815490723, "learning_rate": 9.206556136848133e-05, "loss": 0.053, "step": 44540 }, { "grad_norm": 0.5385720133781433, "learning_rate": 9.206109186938785e-05, "loss": 0.0483, "step": 44550 }, { "grad_norm": 0.8349577784538269, "learning_rate": 9.205662122035994e-05, "loss": 0.0631, "step": 44560 }, { "grad_norm": 0.8797184824943542, "learning_rate": 9.20521494215198e-05, "loss": 0.0628, "step": 44570 }, { "grad_norm": 0.556582510471344, "learning_rate": 9.204767647298971e-05, "loss": 0.0449, "step": 44580 }, { "grad_norm": 0.37899819016456604, "learning_rate": 9.204320237489192e-05, "loss": 0.0467, "step": 44590 }, { "grad_norm": 1.3801584243774414, "learning_rate": 9.203872712734879e-05, "loss": 0.0687, "step": 44600 }, { "grad_norm": 0.9484851360321045, "learning_rate": 9.203425073048263e-05, "loss": 0.0667, "step": 44610 }, { "grad_norm": 1.076108455657959, "learning_rate": 9.202977318441587e-05, "loss": 0.0632, "step": 44620 }, { "grad_norm": 0.4471173584461212, "learning_rate": 9.202529448927091e-05, "loss": 0.0758, "step": 44630 }, { "grad_norm": 0.7664887309074402, "learning_rate": 9.202081464517016e-05, "loss": 0.0606, "step": 44640 }, { "grad_norm": 0.6930730938911438, "learning_rate": 9.201633365223614e-05, "loss": 0.0424, "step": 44650 }, { "grad_norm": 0.601469099521637, "learning_rate": 9.201185151059132e-05, "loss": 0.0514, "step": 44660 }, { "grad_norm": 0.5380877256393433, "learning_rate": 9.200736822035829e-05, "loss": 0.0673, "step": 44670 }, { "grad_norm": 1.4171620607376099, "learning_rate": 9.200288378165958e-05, "loss": 0.0543, "step": 44680 }, { "grad_norm": 0.6697529554367065, "learning_rate": 9.199839819461782e-05, "loss": 0.0487, "step": 44690 }, { "grad_norm": 0.6726697087287903, "learning_rate": 9.199391145935561e-05, "loss": 0.0516, "step": 44700 }, { "grad_norm": 0.5203090906143188, "learning_rate": 9.198942357599565e-05, "loss": 0.0426, "step": 44710 }, { "grad_norm": 0.7519232630729675, "learning_rate": 9.198493454466063e-05, "loss": 0.0493, "step": 44720 }, { "grad_norm": 0.6877142190933228, "learning_rate": 9.198044436547326e-05, "loss": 0.062, "step": 44730 }, { "grad_norm": 0.4596562087535858, "learning_rate": 9.197595303855633e-05, "loss": 0.0739, "step": 44740 }, { "grad_norm": 0.4284224808216095, "learning_rate": 9.19714605640326e-05, "loss": 0.0421, "step": 44750 }, { "grad_norm": 0.4204980731010437, "learning_rate": 9.19669669420249e-05, "loss": 0.0471, "step": 44760 }, { "grad_norm": 0.43914780020713806, "learning_rate": 9.19624721726561e-05, "loss": 0.0401, "step": 44770 }, { "grad_norm": 0.5688958168029785, "learning_rate": 9.195797625604908e-05, "loss": 0.0474, "step": 44780 }, { "grad_norm": 0.5475608706474304, "learning_rate": 9.195347919232673e-05, "loss": 0.0518, "step": 44790 }, { "grad_norm": 0.806693971157074, "learning_rate": 9.194898098161204e-05, "loss": 0.0578, "step": 44800 }, { "grad_norm": 0.9605935215950012, "learning_rate": 9.194448162402797e-05, "loss": 0.0775, "step": 44810 }, { "grad_norm": 0.504608154296875, "learning_rate": 9.193998111969753e-05, "loss": 0.0622, "step": 44820 }, { "grad_norm": 0.7195112705230713, "learning_rate": 9.193547946874374e-05, "loss": 0.0592, "step": 44830 }, { "grad_norm": 1.0793628692626953, "learning_rate": 9.193097667128972e-05, "loss": 0.0536, "step": 44840 }, { "grad_norm": 0.5537346005439758, "learning_rate": 9.192647272745855e-05, "loss": 0.0496, "step": 44850 }, { "grad_norm": 0.6824357509613037, "learning_rate": 9.192196763737335e-05, "loss": 0.0527, "step": 44860 }, { "grad_norm": 0.8101964592933655, "learning_rate": 9.191746140115732e-05, "loss": 0.0496, "step": 44870 }, { "grad_norm": 0.7782663106918335, "learning_rate": 9.191295401893365e-05, "loss": 0.047, "step": 44880 }, { "grad_norm": 0.7050342559814453, "learning_rate": 9.190844549082553e-05, "loss": 0.0608, "step": 44890 }, { "grad_norm": 0.7168493866920471, "learning_rate": 9.190393581695629e-05, "loss": 0.0493, "step": 44900 }, { "grad_norm": 0.41786089539527893, "learning_rate": 9.189942499744916e-05, "loss": 0.051, "step": 44910 }, { "grad_norm": 0.6010271310806274, "learning_rate": 9.189491303242752e-05, "loss": 0.0616, "step": 44920 }, { "grad_norm": 0.7062689661979675, "learning_rate": 9.189039992201467e-05, "loss": 0.0522, "step": 44930 }, { "grad_norm": 0.5120160579681396, "learning_rate": 9.188588566633403e-05, "loss": 0.0624, "step": 44940 }, { "grad_norm": 0.7742273807525635, "learning_rate": 9.188137026550903e-05, "loss": 0.0578, "step": 44950 }, { "grad_norm": 0.9392011165618896, "learning_rate": 9.187685371966308e-05, "loss": 0.0535, "step": 44960 }, { "grad_norm": 0.5576140880584717, "learning_rate": 9.187233602891969e-05, "loss": 0.0492, "step": 44970 }, { "grad_norm": 0.6307185888290405, "learning_rate": 9.186781719340236e-05, "loss": 0.0528, "step": 44980 }, { "grad_norm": 0.5296696424484253, "learning_rate": 9.186329721323464e-05, "loss": 0.0447, "step": 44990 }, { "grad_norm": 0.725208580493927, "learning_rate": 9.18587760885401e-05, "loss": 0.065, "step": 45000 }, { "grad_norm": 0.3395550549030304, "learning_rate": 9.185425381944234e-05, "loss": 0.0444, "step": 45010 }, { "grad_norm": 0.5359270572662354, "learning_rate": 9.184973040606502e-05, "loss": 0.0378, "step": 45020 }, { "grad_norm": 0.8808684945106506, "learning_rate": 9.18452058485318e-05, "loss": 0.0507, "step": 45030 }, { "grad_norm": 0.505407452583313, "learning_rate": 9.184068014696635e-05, "loss": 0.0709, "step": 45040 }, { "grad_norm": 0.8487017750740051, "learning_rate": 9.183615330149242e-05, "loss": 0.0855, "step": 45050 }, { "grad_norm": 0.513876736164093, "learning_rate": 9.183162531223379e-05, "loss": 0.0561, "step": 45060 }, { "grad_norm": 0.4306880235671997, "learning_rate": 9.182709617931424e-05, "loss": 0.0509, "step": 45070 }, { "grad_norm": 0.6956640481948853, "learning_rate": 9.18225659028576e-05, "loss": 0.0429, "step": 45080 }, { "grad_norm": 0.5231378078460693, "learning_rate": 9.181803448298771e-05, "loss": 0.0519, "step": 45090 }, { "grad_norm": 0.4678207039833069, "learning_rate": 9.181350191982848e-05, "loss": 0.0602, "step": 45100 }, { "grad_norm": 0.7185661792755127, "learning_rate": 9.18089682135038e-05, "loss": 0.0692, "step": 45110 }, { "grad_norm": 0.6786701679229736, "learning_rate": 9.180443336413764e-05, "loss": 0.0646, "step": 45120 }, { "grad_norm": 0.5906214118003845, "learning_rate": 9.179989737185397e-05, "loss": 0.0462, "step": 45130 }, { "grad_norm": 0.6238858699798584, "learning_rate": 9.179536023677682e-05, "loss": 0.058, "step": 45140 }, { "grad_norm": 0.5666710734367371, "learning_rate": 9.179082195903022e-05, "loss": 0.0474, "step": 45150 }, { "grad_norm": 0.4313807785511017, "learning_rate": 9.178628253873825e-05, "loss": 0.0488, "step": 45160 }, { "grad_norm": 0.5568516850471497, "learning_rate": 9.178174197602501e-05, "loss": 0.043, "step": 45170 }, { "grad_norm": 0.5107523798942566, "learning_rate": 9.177720027101464e-05, "loss": 0.0329, "step": 45180 }, { "grad_norm": 0.6064403057098389, "learning_rate": 9.177265742383132e-05, "loss": 0.0572, "step": 45190 }, { "grad_norm": 0.4429289400577545, "learning_rate": 9.176811343459923e-05, "loss": 0.0506, "step": 45200 }, { "grad_norm": 0.556701123714447, "learning_rate": 9.176356830344262e-05, "loss": 0.0421, "step": 45210 }, { "grad_norm": 0.7038702368736267, "learning_rate": 9.175902203048573e-05, "loss": 0.0508, "step": 45220 }, { "grad_norm": 0.5544828772544861, "learning_rate": 9.175447461585287e-05, "loss": 0.053, "step": 45230 }, { "grad_norm": 0.8918848037719727, "learning_rate": 9.174992605966835e-05, "loss": 0.0425, "step": 45240 }, { "grad_norm": 0.37321922183036804, "learning_rate": 9.174537636205656e-05, "loss": 0.048, "step": 45250 }, { "grad_norm": 0.5333330631256104, "learning_rate": 9.174082552314186e-05, "loss": 0.0482, "step": 45260 }, { "grad_norm": 0.6864514946937561, "learning_rate": 9.173627354304865e-05, "loss": 0.0711, "step": 45270 }, { "grad_norm": 0.692173957824707, "learning_rate": 9.173172042190141e-05, "loss": 0.0542, "step": 45280 }, { "grad_norm": 0.9624427556991577, "learning_rate": 9.172716615982461e-05, "loss": 0.0579, "step": 45290 }, { "grad_norm": 1.9537006616592407, "learning_rate": 9.172261075694277e-05, "loss": 0.0726, "step": 45300 }, { "grad_norm": 0.8682795763015747, "learning_rate": 9.171805421338043e-05, "loss": 0.0561, "step": 45310 }, { "grad_norm": 0.8231619000434875, "learning_rate": 9.171349652926216e-05, "loss": 0.0649, "step": 45320 }, { "grad_norm": 0.581597626209259, "learning_rate": 9.170893770471255e-05, "loss": 0.0453, "step": 45330 }, { "grad_norm": 0.7057535648345947, "learning_rate": 9.170437773985627e-05, "loss": 0.0592, "step": 45340 }, { "grad_norm": 0.5377574563026428, "learning_rate": 9.169981663481795e-05, "loss": 0.0583, "step": 45350 }, { "grad_norm": 0.6808138489723206, "learning_rate": 9.169525438972234e-05, "loss": 0.052, "step": 45360 }, { "grad_norm": 0.860271692276001, "learning_rate": 9.169069100469411e-05, "loss": 0.0696, "step": 45370 }, { "grad_norm": 0.44931578636169434, "learning_rate": 9.168612647985806e-05, "loss": 0.0344, "step": 45380 }, { "grad_norm": 0.48446598649024963, "learning_rate": 9.168156081533898e-05, "loss": 0.0485, "step": 45390 }, { "grad_norm": 0.501255989074707, "learning_rate": 9.167699401126167e-05, "loss": 0.0509, "step": 45400 }, { "grad_norm": 0.524485170841217, "learning_rate": 9.167242606775101e-05, "loss": 0.0432, "step": 45410 }, { "grad_norm": 0.7281507253646851, "learning_rate": 9.166785698493189e-05, "loss": 0.0743, "step": 45420 }, { "grad_norm": 0.4395451545715332, "learning_rate": 9.166328676292919e-05, "loss": 0.04, "step": 45430 }, { "grad_norm": 0.5900741219520569, "learning_rate": 9.16587154018679e-05, "loss": 0.0446, "step": 45440 }, { "grad_norm": 0.48957136273384094, "learning_rate": 9.165414290187298e-05, "loss": 0.0423, "step": 45450 }, { "grad_norm": 0.6250960826873779, "learning_rate": 9.164956926306944e-05, "loss": 0.0531, "step": 45460 }, { "grad_norm": 0.6358281970024109, "learning_rate": 9.164499448558232e-05, "loss": 0.0536, "step": 45470 }, { "grad_norm": 0.398351788520813, "learning_rate": 9.16404185695367e-05, "loss": 0.0446, "step": 45480 }, { "grad_norm": 0.6053804755210876, "learning_rate": 9.163584151505767e-05, "loss": 0.0768, "step": 45490 }, { "grad_norm": 0.8267833590507507, "learning_rate": 9.163126332227039e-05, "loss": 0.0672, "step": 45500 }, { "grad_norm": 0.6911640763282776, "learning_rate": 9.16266839913e-05, "loss": 0.07, "step": 45510 }, { "grad_norm": 0.5125473141670227, "learning_rate": 9.162210352227171e-05, "loss": 0.0623, "step": 45520 }, { "grad_norm": 0.6776713728904724, "learning_rate": 9.161752191531075e-05, "loss": 0.0617, "step": 45530 }, { "grad_norm": 0.5171457529067993, "learning_rate": 9.161293917054236e-05, "loss": 0.0521, "step": 45540 }, { "grad_norm": 0.5726033449172974, "learning_rate": 9.160835528809187e-05, "loss": 0.0479, "step": 45550 }, { "grad_norm": 0.6237253546714783, "learning_rate": 9.160377026808457e-05, "loss": 0.042, "step": 45560 }, { "grad_norm": 0.4225911796092987, "learning_rate": 9.159918411064583e-05, "loss": 0.053, "step": 45570 }, { "grad_norm": 0.49369776248931885, "learning_rate": 9.1594596815901e-05, "loss": 0.0485, "step": 45580 }, { "grad_norm": 0.7330127954483032, "learning_rate": 9.159000838397553e-05, "loss": 0.0555, "step": 45590 }, { "grad_norm": 0.5247395038604736, "learning_rate": 9.158541881499486e-05, "loss": 0.0393, "step": 45600 }, { "grad_norm": 0.5499431490898132, "learning_rate": 9.158082810908447e-05, "loss": 0.0432, "step": 45610 }, { "grad_norm": 0.6358917355537415, "learning_rate": 9.157623626636984e-05, "loss": 0.0654, "step": 45620 }, { "grad_norm": 0.6546777486801147, "learning_rate": 9.157164328697654e-05, "loss": 0.0574, "step": 45630 }, { "grad_norm": 0.4502878189086914, "learning_rate": 9.156704917103013e-05, "loss": 0.0594, "step": 45640 }, { "grad_norm": 0.3899550139904022, "learning_rate": 9.15624539186562e-05, "loss": 0.047, "step": 45650 }, { "grad_norm": 0.613751232624054, "learning_rate": 9.155785752998043e-05, "loss": 0.062, "step": 45660 }, { "grad_norm": 1.7088334560394287, "learning_rate": 9.155326000512841e-05, "loss": 0.0583, "step": 45670 }, { "grad_norm": 0.9712700843811035, "learning_rate": 9.15486613442259e-05, "loss": 0.0635, "step": 45680 }, { "grad_norm": 0.6287927627563477, "learning_rate": 9.154406154739858e-05, "loss": 0.0685, "step": 45690 }, { "grad_norm": 0.7424967885017395, "learning_rate": 9.153946061477223e-05, "loss": 0.0658, "step": 45700 }, { "grad_norm": 1.182976484298706, "learning_rate": 9.153485854647264e-05, "loss": 0.0527, "step": 45710 }, { "grad_norm": 0.568824291229248, "learning_rate": 9.153025534262561e-05, "loss": 0.052, "step": 45720 }, { "grad_norm": 0.6702147126197815, "learning_rate": 9.152565100335699e-05, "loss": 0.0519, "step": 45730 }, { "grad_norm": 0.48582208156585693, "learning_rate": 9.15210455287927e-05, "loss": 0.0422, "step": 45740 }, { "grad_norm": 0.512330949306488, "learning_rate": 9.15164389190586e-05, "loss": 0.049, "step": 45750 }, { "grad_norm": 0.8373093008995056, "learning_rate": 9.151183117428067e-05, "loss": 0.0666, "step": 45760 }, { "grad_norm": 0.6004441380500793, "learning_rate": 9.150722229458487e-05, "loss": 0.0455, "step": 45770 }, { "grad_norm": 0.6431151032447815, "learning_rate": 9.15026122800972e-05, "loss": 0.0411, "step": 45780 }, { "grad_norm": 0.4133138060569763, "learning_rate": 9.149800113094371e-05, "loss": 0.0663, "step": 45790 }, { "grad_norm": 0.9619796276092529, "learning_rate": 9.149338884725045e-05, "loss": 0.0671, "step": 45800 }, { "grad_norm": 1.3702857494354248, "learning_rate": 9.148877542914353e-05, "loss": 0.0645, "step": 45810 }, { "grad_norm": 0.5431200265884399, "learning_rate": 9.148416087674909e-05, "loss": 0.0496, "step": 45820 }, { "grad_norm": 1.0247610807418823, "learning_rate": 9.147954519019327e-05, "loss": 0.0694, "step": 45830 }, { "grad_norm": 0.600262463092804, "learning_rate": 9.147492836960225e-05, "loss": 0.0488, "step": 45840 }, { "grad_norm": 1.1370810270309448, "learning_rate": 9.147031041510229e-05, "loss": 0.0496, "step": 45850 }, { "grad_norm": 0.7989307045936584, "learning_rate": 9.146569132681961e-05, "loss": 0.0581, "step": 45860 }, { "grad_norm": 0.5140159130096436, "learning_rate": 9.146107110488051e-05, "loss": 0.0455, "step": 45870 }, { "grad_norm": 0.8620862364768982, "learning_rate": 9.145644974941132e-05, "loss": 0.0493, "step": 45880 }, { "grad_norm": 0.3517261743545532, "learning_rate": 9.145182726053835e-05, "loss": 0.0493, "step": 45890 }, { "grad_norm": 0.5925669074058533, "learning_rate": 9.144720363838799e-05, "loss": 0.0746, "step": 45900 }, { "grad_norm": 0.40957677364349365, "learning_rate": 9.144257888308667e-05, "loss": 0.0425, "step": 45910 }, { "grad_norm": 1.0607457160949707, "learning_rate": 9.143795299476081e-05, "loss": 0.0557, "step": 45920 }, { "grad_norm": 0.62669837474823, "learning_rate": 9.143332597353686e-05, "loss": 0.046, "step": 45930 }, { "grad_norm": 0.451094388961792, "learning_rate": 9.142869781954137e-05, "loss": 0.038, "step": 45940 }, { "grad_norm": 0.5024151802062988, "learning_rate": 9.142406853290084e-05, "loss": 0.044, "step": 45950 }, { "grad_norm": 0.47696149349212646, "learning_rate": 9.141943811374183e-05, "loss": 0.0608, "step": 45960 }, { "grad_norm": 0.35128313302993774, "learning_rate": 9.141480656219094e-05, "loss": 0.0474, "step": 45970 }, { "grad_norm": 0.45041006803512573, "learning_rate": 9.141017387837483e-05, "loss": 0.0579, "step": 45980 }, { "grad_norm": 0.6997491121292114, "learning_rate": 9.140554006242009e-05, "loss": 0.0495, "step": 45990 }, { "grad_norm": 0.5921828150749207, "learning_rate": 9.140090511445345e-05, "loss": 0.0423, "step": 46000 }, { "grad_norm": 0.6860007047653198, "learning_rate": 9.139626903460161e-05, "loss": 0.0565, "step": 46010 }, { "grad_norm": 0.3957616686820984, "learning_rate": 9.139163182299133e-05, "loss": 0.0403, "step": 46020 }, { "grad_norm": 0.5595121383666992, "learning_rate": 9.138699347974939e-05, "loss": 0.0469, "step": 46030 }, { "grad_norm": 0.5185771584510803, "learning_rate": 9.13823540050026e-05, "loss": 0.0436, "step": 46040 }, { "grad_norm": 0.467061847448349, "learning_rate": 9.137771339887779e-05, "loss": 0.0537, "step": 46050 }, { "grad_norm": 0.5686469078063965, "learning_rate": 9.137307166150185e-05, "loss": 0.0537, "step": 46060 }, { "grad_norm": 0.614370584487915, "learning_rate": 9.136842879300167e-05, "loss": 0.0507, "step": 46070 }, { "grad_norm": 0.449838250875473, "learning_rate": 9.136378479350418e-05, "loss": 0.0355, "step": 46080 }, { "grad_norm": 0.6357284784317017, "learning_rate": 9.135913966313637e-05, "loss": 0.0692, "step": 46090 }, { "grad_norm": 0.8380953669548035, "learning_rate": 9.13544934020252e-05, "loss": 0.0525, "step": 46100 }, { "grad_norm": 0.7198380827903748, "learning_rate": 9.134984601029773e-05, "loss": 0.0455, "step": 46110 }, { "grad_norm": 1.1042231321334839, "learning_rate": 9.1345197488081e-05, "loss": 0.0552, "step": 46120 }, { "grad_norm": 0.580158531665802, "learning_rate": 9.13405478355021e-05, "loss": 0.0471, "step": 46130 }, { "grad_norm": 0.754209041595459, "learning_rate": 9.133589705268817e-05, "loss": 0.0641, "step": 46140 }, { "grad_norm": 0.5280242562294006, "learning_rate": 9.133124513976633e-05, "loss": 0.0536, "step": 46150 }, { "grad_norm": 1.3732835054397583, "learning_rate": 9.132659209686377e-05, "loss": 0.0558, "step": 46160 }, { "grad_norm": 0.6465729475021362, "learning_rate": 9.132193792410771e-05, "loss": 0.0516, "step": 46170 }, { "grad_norm": 0.5633957982063293, "learning_rate": 9.13172826216254e-05, "loss": 0.0531, "step": 46180 }, { "grad_norm": 0.6495137214660645, "learning_rate": 9.13126261895441e-05, "loss": 0.0594, "step": 46190 }, { "grad_norm": 0.5736557245254517, "learning_rate": 9.130796862799111e-05, "loss": 0.0542, "step": 46200 }, { "grad_norm": 0.4875214397907257, "learning_rate": 9.13033099370938e-05, "loss": 0.0779, "step": 46210 }, { "grad_norm": 0.4573192894458771, "learning_rate": 9.12986501169795e-05, "loss": 0.0425, "step": 46220 }, { "grad_norm": 0.5581179857254028, "learning_rate": 9.129398916777561e-05, "loss": 0.0542, "step": 46230 }, { "grad_norm": 0.2615187168121338, "learning_rate": 9.128932708960959e-05, "loss": 0.0555, "step": 46240 }, { "grad_norm": 0.5850437879562378, "learning_rate": 9.128466388260888e-05, "loss": 0.05, "step": 46250 }, { "grad_norm": 0.8265694975852966, "learning_rate": 9.127999954690094e-05, "loss": 0.0473, "step": 46260 }, { "grad_norm": 0.37382060289382935, "learning_rate": 9.127533408261335e-05, "loss": 0.041, "step": 46270 }, { "grad_norm": 0.43341219425201416, "learning_rate": 9.127066748987363e-05, "loss": 0.0429, "step": 46280 }, { "grad_norm": 0.9419196248054504, "learning_rate": 9.126599976880934e-05, "loss": 0.0996, "step": 46290 }, { "grad_norm": 0.8142265677452087, "learning_rate": 9.126133091954814e-05, "loss": 0.0453, "step": 46300 }, { "grad_norm": 0.5004222393035889, "learning_rate": 9.125666094221766e-05, "loss": 0.0515, "step": 46310 }, { "grad_norm": 0.5928509831428528, "learning_rate": 9.125198983694556e-05, "loss": 0.0543, "step": 46320 }, { "grad_norm": 0.5918177366256714, "learning_rate": 9.124731760385956e-05, "loss": 0.0438, "step": 46330 }, { "grad_norm": 0.8915472626686096, "learning_rate": 9.124264424308738e-05, "loss": 0.0455, "step": 46340 }, { "grad_norm": 0.6283944845199585, "learning_rate": 9.12379697547568e-05, "loss": 0.0486, "step": 46350 }, { "grad_norm": 0.42950955033302307, "learning_rate": 9.123329413899564e-05, "loss": 0.0508, "step": 46360 }, { "grad_norm": 0.5849243998527527, "learning_rate": 9.122861739593169e-05, "loss": 0.0603, "step": 46370 }, { "grad_norm": 0.762880802154541, "learning_rate": 9.122393952569283e-05, "loss": 0.0452, "step": 46380 }, { "grad_norm": 0.43618613481521606, "learning_rate": 9.121926052840698e-05, "loss": 0.0324, "step": 46390 }, { "grad_norm": 0.37839967012405396, "learning_rate": 9.121458040420199e-05, "loss": 0.0566, "step": 46400 }, { "grad_norm": 0.6458873748779297, "learning_rate": 9.120989915320588e-05, "loss": 0.0516, "step": 46410 }, { "grad_norm": 0.6555989384651184, "learning_rate": 9.120521677554661e-05, "loss": 0.0475, "step": 46420 }, { "grad_norm": 0.5159918069839478, "learning_rate": 9.120053327135219e-05, "loss": 0.0431, "step": 46430 }, { "grad_norm": 0.5368824601173401, "learning_rate": 9.119584864075066e-05, "loss": 0.0468, "step": 46440 }, { "grad_norm": 0.7352043390274048, "learning_rate": 9.11911628838701e-05, "loss": 0.0515, "step": 46450 }, { "grad_norm": 0.5212388634681702, "learning_rate": 9.118647600083864e-05, "loss": 0.0541, "step": 46460 }, { "grad_norm": 1.0509746074676514, "learning_rate": 9.118178799178439e-05, "loss": 0.0496, "step": 46470 }, { "grad_norm": 0.3833107352256775, "learning_rate": 9.117709885683553e-05, "loss": 0.0446, "step": 46480 }, { "grad_norm": 0.663918137550354, "learning_rate": 9.117240859612026e-05, "loss": 0.0567, "step": 46490 }, { "grad_norm": 0.6904950141906738, "learning_rate": 9.11677172097668e-05, "loss": 0.0495, "step": 46500 }, { "grad_norm": 0.3342800438404083, "learning_rate": 9.116302469790341e-05, "loss": 0.0528, "step": 46510 }, { "grad_norm": 0.7126175165176392, "learning_rate": 9.11583310606584e-05, "loss": 0.04, "step": 46520 }, { "grad_norm": 0.522524356842041, "learning_rate": 9.115363629816009e-05, "loss": 0.053, "step": 46530 }, { "grad_norm": 0.8426415324211121, "learning_rate": 9.114894041053681e-05, "loss": 0.0606, "step": 46540 }, { "grad_norm": 0.5377634167671204, "learning_rate": 9.114424339791697e-05, "loss": 0.0384, "step": 46550 }, { "grad_norm": 0.5335493683815002, "learning_rate": 9.113954526042895e-05, "loss": 0.0496, "step": 46560 }, { "grad_norm": 0.7022128701210022, "learning_rate": 9.113484599820125e-05, "loss": 0.067, "step": 46570 }, { "grad_norm": 0.5656211972236633, "learning_rate": 9.113014561136229e-05, "loss": 0.0519, "step": 46580 }, { "grad_norm": 0.6518254280090332, "learning_rate": 9.11254441000406e-05, "loss": 0.0372, "step": 46590 }, { "grad_norm": 0.9485878944396973, "learning_rate": 9.112074146436473e-05, "loss": 0.0573, "step": 46600 }, { "grad_norm": 0.47687020897865295, "learning_rate": 9.111603770446324e-05, "loss": 0.0507, "step": 46610 }, { "grad_norm": 0.43687188625335693, "learning_rate": 9.111133282046471e-05, "loss": 0.0648, "step": 46620 }, { "grad_norm": 0.6728721857070923, "learning_rate": 9.11066268124978e-05, "loss": 0.05, "step": 46630 }, { "grad_norm": 0.7020560503005981, "learning_rate": 9.110191968069115e-05, "loss": 0.0616, "step": 46640 }, { "grad_norm": 0.45961570739746094, "learning_rate": 9.109721142517346e-05, "loss": 0.0479, "step": 46650 }, { "grad_norm": 0.652116060256958, "learning_rate": 9.109250204607344e-05, "loss": 0.0558, "step": 46660 }, { "grad_norm": 0.6372677087783813, "learning_rate": 9.108779154351986e-05, "loss": 0.0608, "step": 46670 }, { "grad_norm": 0.6866009831428528, "learning_rate": 9.108307991764148e-05, "loss": 0.0506, "step": 46680 }, { "grad_norm": 0.46059608459472656, "learning_rate": 9.107836716856715e-05, "loss": 0.0449, "step": 46690 }, { "grad_norm": 0.5820420980453491, "learning_rate": 9.107365329642569e-05, "loss": 0.0616, "step": 46700 }, { "grad_norm": 0.7304346561431885, "learning_rate": 9.106893830134598e-05, "loss": 0.0398, "step": 46710 }, { "grad_norm": 0.516380786895752, "learning_rate": 9.106422218345693e-05, "loss": 0.052, "step": 46720 }, { "grad_norm": 0.5816673636436462, "learning_rate": 9.105950494288747e-05, "loss": 0.0572, "step": 46730 }, { "grad_norm": 0.6032408475875854, "learning_rate": 9.105478657976656e-05, "loss": 0.0533, "step": 46740 }, { "grad_norm": 0.5617914795875549, "learning_rate": 9.105006709422321e-05, "loss": 0.0553, "step": 46750 }, { "grad_norm": 0.726952850818634, "learning_rate": 9.104534648638645e-05, "loss": 0.0714, "step": 46760 }, { "grad_norm": 0.39846867322921753, "learning_rate": 9.104062475638533e-05, "loss": 0.0365, "step": 46770 }, { "grad_norm": 0.8512837886810303, "learning_rate": 9.103590190434896e-05, "loss": 0.0704, "step": 46780 }, { "grad_norm": 0.3262530565261841, "learning_rate": 9.103117793040646e-05, "loss": 0.044, "step": 46790 }, { "grad_norm": 0.600463330745697, "learning_rate": 9.102645283468694e-05, "loss": 0.0666, "step": 46800 }, { "grad_norm": 0.5058342218399048, "learning_rate": 9.102172661731963e-05, "loss": 0.0677, "step": 46810 }, { "grad_norm": 0.6255159378051758, "learning_rate": 9.101699927843372e-05, "loss": 0.0551, "step": 46820 }, { "grad_norm": 0.48519188165664673, "learning_rate": 9.101227081815847e-05, "loss": 0.0448, "step": 46830 }, { "grad_norm": 0.9962022304534912, "learning_rate": 9.100754123662312e-05, "loss": 0.0515, "step": 46840 }, { "grad_norm": 0.5555611252784729, "learning_rate": 9.100281053395702e-05, "loss": 0.0535, "step": 46850 }, { "grad_norm": 0.5905733108520508, "learning_rate": 9.099807871028946e-05, "loss": 0.0516, "step": 46860 }, { "grad_norm": 0.547461986541748, "learning_rate": 9.099334576574985e-05, "loss": 0.0545, "step": 46870 }, { "grad_norm": 1.4923962354660034, "learning_rate": 9.098861170046757e-05, "loss": 0.0507, "step": 46880 }, { "grad_norm": 0.40702390670776367, "learning_rate": 9.098387651457204e-05, "loss": 0.0447, "step": 46890 }, { "grad_norm": 1.0911972522735596, "learning_rate": 9.097914020819272e-05, "loss": 0.0665, "step": 46900 }, { "grad_norm": 0.7250880599021912, "learning_rate": 9.09744027814591e-05, "loss": 0.0483, "step": 46910 }, { "grad_norm": 0.5302062630653381, "learning_rate": 9.096966423450071e-05, "loss": 0.0494, "step": 46920 }, { "grad_norm": 0.8475379347801208, "learning_rate": 9.096492456744709e-05, "loss": 0.0454, "step": 46930 }, { "grad_norm": 0.6022999882698059, "learning_rate": 9.096018378042782e-05, "loss": 0.0474, "step": 46940 }, { "grad_norm": 0.5491198301315308, "learning_rate": 9.095544187357252e-05, "loss": 0.0515, "step": 46950 }, { "grad_norm": 0.5184968113899231, "learning_rate": 9.095069884701082e-05, "loss": 0.0461, "step": 46960 }, { "grad_norm": 0.9138690233230591, "learning_rate": 9.09459547008724e-05, "loss": 0.0654, "step": 46970 }, { "grad_norm": 0.4773079752922058, "learning_rate": 9.094120943528694e-05, "loss": 0.0478, "step": 46980 }, { "grad_norm": 0.6907681226730347, "learning_rate": 9.093646305038421e-05, "loss": 0.0557, "step": 46990 }, { "grad_norm": 0.34435951709747314, "learning_rate": 9.093171554629396e-05, "loss": 0.0935, "step": 47000 }, { "grad_norm": 1.084743618965149, "learning_rate": 9.092696692314599e-05, "loss": 0.0624, "step": 47010 }, { "grad_norm": 0.8564534783363342, "learning_rate": 9.092221718107012e-05, "loss": 0.0497, "step": 47020 }, { "grad_norm": 0.7192468643188477, "learning_rate": 9.09174663201962e-05, "loss": 0.0851, "step": 47030 }, { "grad_norm": 0.7101294994354248, "learning_rate": 9.091271434065412e-05, "loss": 0.078, "step": 47040 }, { "grad_norm": 0.748819887638092, "learning_rate": 9.090796124257381e-05, "loss": 0.0544, "step": 47050 }, { "grad_norm": 0.5012989640235901, "learning_rate": 9.09032070260852e-05, "loss": 0.0422, "step": 47060 }, { "grad_norm": 1.5989402532577515, "learning_rate": 9.089845169131828e-05, "loss": 0.067, "step": 47070 }, { "grad_norm": 0.6970561742782593, "learning_rate": 9.089369523840305e-05, "loss": 0.0536, "step": 47080 }, { "grad_norm": 0.7380519509315491, "learning_rate": 9.088893766746956e-05, "loss": 0.0482, "step": 47090 }, { "grad_norm": 0.6683273315429688, "learning_rate": 9.088417897864787e-05, "loss": 0.045, "step": 47100 }, { "grad_norm": 1.583730697631836, "learning_rate": 9.08794191720681e-05, "loss": 0.0651, "step": 47110 }, { "grad_norm": 0.7831863760948181, "learning_rate": 9.087465824786035e-05, "loss": 0.0544, "step": 47120 }, { "grad_norm": 0.5472702980041504, "learning_rate": 9.086989620615482e-05, "loss": 0.0467, "step": 47130 }, { "grad_norm": 0.5171998143196106, "learning_rate": 9.086513304708166e-05, "loss": 0.0459, "step": 47140 }, { "grad_norm": 0.3458388149738312, "learning_rate": 9.086036877077113e-05, "loss": 0.0504, "step": 47150 }, { "grad_norm": 1.237705945968628, "learning_rate": 9.085560337735346e-05, "loss": 0.0827, "step": 47160 }, { "grad_norm": 0.8896883130073547, "learning_rate": 9.085083686695895e-05, "loss": 0.065, "step": 47170 }, { "grad_norm": 0.4774653911590576, "learning_rate": 9.08460692397179e-05, "loss": 0.0897, "step": 47180 }, { "grad_norm": 0.7971867918968201, "learning_rate": 9.084130049576067e-05, "loss": 0.0667, "step": 47190 }, { "grad_norm": 0.6631206274032593, "learning_rate": 9.083653063521762e-05, "loss": 0.0655, "step": 47200 }, { "grad_norm": 0.8329724073410034, "learning_rate": 9.083175965821917e-05, "loss": 0.0772, "step": 47210 }, { "grad_norm": 0.4697907269001007, "learning_rate": 9.082698756489573e-05, "loss": 0.0385, "step": 47220 }, { "grad_norm": 0.583356499671936, "learning_rate": 9.082221435537783e-05, "loss": 0.0481, "step": 47230 }, { "grad_norm": 0.4455253481864929, "learning_rate": 9.081744002979588e-05, "loss": 0.054, "step": 47240 }, { "grad_norm": 0.3689529299736023, "learning_rate": 9.081266458828048e-05, "loss": 0.045, "step": 47250 }, { "grad_norm": 0.5812131762504578, "learning_rate": 9.080788803096216e-05, "loss": 0.0527, "step": 47260 }, { "grad_norm": 0.7554506063461304, "learning_rate": 9.08031103579715e-05, "loss": 0.0464, "step": 47270 }, { "grad_norm": 0.6972821950912476, "learning_rate": 9.079833156943915e-05, "loss": 0.0604, "step": 47280 }, { "grad_norm": 0.805130660533905, "learning_rate": 9.079355166549575e-05, "loss": 0.047, "step": 47290 }, { "grad_norm": 0.4150356650352478, "learning_rate": 9.078877064627195e-05, "loss": 0.0577, "step": 47300 }, { "grad_norm": 0.6622716784477234, "learning_rate": 9.078398851189849e-05, "loss": 0.056, "step": 47310 }, { "grad_norm": 0.9606332182884216, "learning_rate": 9.077920526250611e-05, "loss": 0.061, "step": 47320 }, { "grad_norm": 0.5358687043190002, "learning_rate": 9.077442089822557e-05, "loss": 0.0584, "step": 47330 }, { "grad_norm": 0.7343239188194275, "learning_rate": 9.076963541918768e-05, "loss": 0.0599, "step": 47340 }, { "grad_norm": 0.7403658628463745, "learning_rate": 9.076484882552327e-05, "loss": 0.0645, "step": 47350 }, { "grad_norm": 0.4759751856327057, "learning_rate": 9.076006111736324e-05, "loss": 0.056, "step": 47360 }, { "grad_norm": 0.6199196577072144, "learning_rate": 9.075527229483842e-05, "loss": 0.0544, "step": 47370 }, { "grad_norm": 0.6173272132873535, "learning_rate": 9.075048235807977e-05, "loss": 0.048, "step": 47380 }, { "grad_norm": 0.629401683807373, "learning_rate": 9.074569130721826e-05, "loss": 0.0652, "step": 47390 }, { "grad_norm": 1.9132344722747803, "learning_rate": 9.074089914238484e-05, "loss": 0.0491, "step": 47400 }, { "grad_norm": 0.6173204183578491, "learning_rate": 9.073610586371055e-05, "loss": 0.042, "step": 47410 }, { "grad_norm": 0.5216284394264221, "learning_rate": 9.073131147132643e-05, "loss": 0.0488, "step": 47420 }, { "grad_norm": 0.7038136124610901, "learning_rate": 9.072651596536355e-05, "loss": 0.0545, "step": 47430 }, { "grad_norm": 0.5438323616981506, "learning_rate": 9.072171934595302e-05, "loss": 0.0448, "step": 47440 }, { "grad_norm": 0.5941234827041626, "learning_rate": 9.071692161322599e-05, "loss": 0.0639, "step": 47450 }, { "grad_norm": 0.8334625959396362, "learning_rate": 9.071212276731362e-05, "loss": 0.0536, "step": 47460 }, { "grad_norm": 0.3550044596195221, "learning_rate": 9.070732280834711e-05, "loss": 0.0368, "step": 47470 }, { "grad_norm": 0.42250365018844604, "learning_rate": 9.070252173645768e-05, "loss": 0.0482, "step": 47480 }, { "grad_norm": 0.5219340324401855, "learning_rate": 9.06977195517766e-05, "loss": 0.0504, "step": 47490 }, { "grad_norm": 0.7318758368492126, "learning_rate": 9.069291625443516e-05, "loss": 0.0462, "step": 47500 }, { "grad_norm": 0.5967262983322144, "learning_rate": 9.068811184456467e-05, "loss": 0.0543, "step": 47510 }, { "grad_norm": 0.5191282629966736, "learning_rate": 9.06833063222965e-05, "loss": 0.0472, "step": 47520 }, { "grad_norm": 0.553616464138031, "learning_rate": 9.067849968776201e-05, "loss": 0.0532, "step": 47530 }, { "grad_norm": 1.1301060914993286, "learning_rate": 9.067369194109263e-05, "loss": 0.056, "step": 47540 }, { "grad_norm": 0.816473662853241, "learning_rate": 9.066888308241979e-05, "loss": 0.0494, "step": 47550 }, { "grad_norm": 1.0704855918884277, "learning_rate": 9.066407311187495e-05, "loss": 0.0514, "step": 47560 }, { "grad_norm": 0.6291524767875671, "learning_rate": 9.065926202958964e-05, "loss": 0.0509, "step": 47570 }, { "grad_norm": 0.798058032989502, "learning_rate": 9.065444983569538e-05, "loss": 0.0549, "step": 47580 }, { "grad_norm": 0.7075178027153015, "learning_rate": 9.064963653032375e-05, "loss": 0.0664, "step": 47590 }, { "grad_norm": 0.42766624689102173, "learning_rate": 9.06448221136063e-05, "loss": 0.06, "step": 47600 }, { "grad_norm": 0.5539310574531555, "learning_rate": 9.06400065856747e-05, "loss": 0.0584, "step": 47610 }, { "grad_norm": 0.5514275431632996, "learning_rate": 9.063518994666058e-05, "loss": 0.0594, "step": 47620 }, { "grad_norm": 0.7954274415969849, "learning_rate": 9.063037219669563e-05, "loss": 0.0457, "step": 47630 }, { "grad_norm": 0.4057542085647583, "learning_rate": 9.062555333591159e-05, "loss": 0.0483, "step": 47640 }, { "grad_norm": 0.45202451944351196, "learning_rate": 9.062073336444016e-05, "loss": 0.0485, "step": 47650 }, { "grad_norm": 0.8211817741394043, "learning_rate": 9.061591228241314e-05, "loss": 0.057, "step": 47660 }, { "grad_norm": 0.809290885925293, "learning_rate": 9.061109008996236e-05, "loss": 0.0484, "step": 47670 }, { "grad_norm": 0.6668258309364319, "learning_rate": 9.060626678721961e-05, "loss": 0.0548, "step": 47680 }, { "grad_norm": 0.6451272964477539, "learning_rate": 9.06014423743168e-05, "loss": 0.0499, "step": 47690 }, { "grad_norm": 0.45477500557899475, "learning_rate": 9.05966168513858e-05, "loss": 0.057, "step": 47700 }, { "grad_norm": 1.0854840278625488, "learning_rate": 9.059179021855854e-05, "loss": 0.0615, "step": 47710 }, { "grad_norm": 0.4260334074497223, "learning_rate": 9.058696247596701e-05, "loss": 0.0537, "step": 47720 }, { "grad_norm": 0.7553231716156006, "learning_rate": 9.058213362374315e-05, "loss": 0.0707, "step": 47730 }, { "grad_norm": 0.5483633279800415, "learning_rate": 9.0577303662019e-05, "loss": 0.0488, "step": 47740 }, { "grad_norm": 0.522760272026062, "learning_rate": 9.057247259092662e-05, "loss": 0.0515, "step": 47750 }, { "grad_norm": 0.5381011962890625, "learning_rate": 9.056764041059809e-05, "loss": 0.0455, "step": 47760 }, { "grad_norm": 0.8709203004837036, "learning_rate": 9.056280712116551e-05, "loss": 0.042, "step": 47770 }, { "grad_norm": 0.5640631914138794, "learning_rate": 9.0557972722761e-05, "loss": 0.0627, "step": 47780 }, { "grad_norm": 0.7707463502883911, "learning_rate": 9.055313721551677e-05, "loss": 0.0551, "step": 47790 }, { "grad_norm": 0.7596168518066406, "learning_rate": 9.0548300599565e-05, "loss": 0.0464, "step": 47800 }, { "grad_norm": 0.6354097127914429, "learning_rate": 9.054346287503794e-05, "loss": 0.0455, "step": 47810 }, { "grad_norm": 0.835132896900177, "learning_rate": 9.053862404206782e-05, "loss": 0.0458, "step": 47820 }, { "grad_norm": 0.4726598262786865, "learning_rate": 9.053378410078694e-05, "loss": 0.0485, "step": 47830 }, { "grad_norm": 0.883826732635498, "learning_rate": 9.052894305132765e-05, "loss": 0.0635, "step": 47840 }, { "grad_norm": 0.6106964349746704, "learning_rate": 9.052410089382228e-05, "loss": 0.0529, "step": 47850 }, { "grad_norm": 0.46582266688346863, "learning_rate": 9.051925762840321e-05, "loss": 0.0471, "step": 47860 }, { "grad_norm": 0.6867358684539795, "learning_rate": 9.051441325520287e-05, "loss": 0.052, "step": 47870 }, { "grad_norm": 0.6920971274375916, "learning_rate": 9.050956777435367e-05, "loss": 0.0431, "step": 47880 }, { "grad_norm": 0.5031812787055969, "learning_rate": 9.050472118598811e-05, "loss": 0.0507, "step": 47890 }, { "grad_norm": 0.4526292681694031, "learning_rate": 9.04998734902387e-05, "loss": 0.0482, "step": 47900 }, { "grad_norm": 0.8931785821914673, "learning_rate": 9.049502468723796e-05, "loss": 0.0694, "step": 47910 }, { "grad_norm": 0.5041327476501465, "learning_rate": 9.049017477711847e-05, "loss": 0.083, "step": 47920 }, { "grad_norm": 0.6514850854873657, "learning_rate": 9.04853237600128e-05, "loss": 0.0582, "step": 47930 }, { "grad_norm": 0.6942211389541626, "learning_rate": 9.04804716360536e-05, "loss": 0.0636, "step": 47940 }, { "grad_norm": 0.7632321119308472, "learning_rate": 9.04756184053735e-05, "loss": 0.0602, "step": 47950 }, { "grad_norm": 0.6405106782913208, "learning_rate": 9.047076406810519e-05, "loss": 0.052, "step": 47960 }, { "grad_norm": 0.7411405444145203, "learning_rate": 9.046590862438142e-05, "loss": 0.0517, "step": 47970 }, { "grad_norm": 0.4864107370376587, "learning_rate": 9.046105207433489e-05, "loss": 0.0541, "step": 47980 }, { "grad_norm": 0.5443549156188965, "learning_rate": 9.04561944180984e-05, "loss": 0.0413, "step": 47990 }, { "grad_norm": 0.6594535708427429, "learning_rate": 9.045133565580476e-05, "loss": 0.0457, "step": 48000 }, { "grad_norm": 0.7314715385437012, "learning_rate": 9.04464757875868e-05, "loss": 0.0501, "step": 48010 }, { "grad_norm": 0.6822757720947266, "learning_rate": 9.04416148135774e-05, "loss": 0.0369, "step": 48020 }, { "grad_norm": 0.36561357975006104, "learning_rate": 9.043675273390942e-05, "loss": 0.0472, "step": 48030 }, { "grad_norm": 0.49974846839904785, "learning_rate": 9.043188954871582e-05, "loss": 0.0394, "step": 48040 }, { "grad_norm": 0.5215927958488464, "learning_rate": 9.042702525812957e-05, "loss": 0.0375, "step": 48050 }, { "grad_norm": 0.4140929579734802, "learning_rate": 9.042215986228359e-05, "loss": 0.0485, "step": 48060 }, { "grad_norm": 0.3995213806629181, "learning_rate": 9.041729336131099e-05, "loss": 0.0379, "step": 48070 }, { "grad_norm": 0.5911900401115417, "learning_rate": 9.041242575534474e-05, "loss": 0.0537, "step": 48080 }, { "grad_norm": 0.5383095145225525, "learning_rate": 9.040755704451798e-05, "loss": 0.0591, "step": 48090 }, { "grad_norm": 0.4487732946872711, "learning_rate": 9.040268722896378e-05, "loss": 0.0389, "step": 48100 }, { "grad_norm": 0.8848474621772766, "learning_rate": 9.039781630881528e-05, "loss": 0.0412, "step": 48110 }, { "grad_norm": 0.6309685707092285, "learning_rate": 9.039294428420568e-05, "loss": 0.0659, "step": 48120 }, { "grad_norm": 0.5778746008872986, "learning_rate": 9.038807115526815e-05, "loss": 0.0463, "step": 48130 }, { "grad_norm": 0.46797052025794983, "learning_rate": 9.038319692213591e-05, "loss": 0.0463, "step": 48140 }, { "grad_norm": 0.5757837295532227, "learning_rate": 9.037832158494226e-05, "loss": 0.0561, "step": 48150 }, { "grad_norm": 0.5324357151985168, "learning_rate": 9.037344514382046e-05, "loss": 0.0502, "step": 48160 }, { "grad_norm": 0.7878886461257935, "learning_rate": 9.036856759890383e-05, "loss": 0.0445, "step": 48170 }, { "grad_norm": 0.33385807275772095, "learning_rate": 9.036368895032575e-05, "loss": 0.0412, "step": 48180 }, { "grad_norm": 0.7259345054626465, "learning_rate": 9.035880919821956e-05, "loss": 0.0646, "step": 48190 }, { "grad_norm": 0.6119816899299622, "learning_rate": 9.035392834271868e-05, "loss": 0.064, "step": 48200 }, { "grad_norm": 0.5401572585105896, "learning_rate": 9.034904638395656e-05, "loss": 0.0481, "step": 48210 }, { "grad_norm": 1.0188559293746948, "learning_rate": 9.034416332206669e-05, "loss": 0.0458, "step": 48220 }, { "grad_norm": 0.5237802863121033, "learning_rate": 9.033927915718255e-05, "loss": 0.0385, "step": 48230 }, { "grad_norm": 0.4818427264690399, "learning_rate": 9.033439388943765e-05, "loss": 0.0513, "step": 48240 }, { "grad_norm": 0.968198299407959, "learning_rate": 9.032950751896558e-05, "loss": 0.0532, "step": 48250 }, { "grad_norm": 0.5814591646194458, "learning_rate": 9.032462004589994e-05, "loss": 0.0561, "step": 48260 }, { "grad_norm": 0.6805868148803711, "learning_rate": 9.031973147037433e-05, "loss": 0.0582, "step": 48270 }, { "grad_norm": 0.5734009146690369, "learning_rate": 9.031484179252241e-05, "loss": 0.0688, "step": 48280 }, { "grad_norm": 0.34779298305511475, "learning_rate": 9.030995101247785e-05, "loss": 0.0493, "step": 48290 }, { "grad_norm": 0.4965858459472656, "learning_rate": 9.03050591303744e-05, "loss": 0.047, "step": 48300 }, { "grad_norm": 0.545155942440033, "learning_rate": 9.030016614634576e-05, "loss": 0.0473, "step": 48310 }, { "grad_norm": 0.5008450150489807, "learning_rate": 9.029527206052571e-05, "loss": 0.05, "step": 48320 }, { "grad_norm": 0.4999806880950928, "learning_rate": 9.029037687304806e-05, "loss": 0.046, "step": 48330 }, { "grad_norm": 0.588161051273346, "learning_rate": 9.028548058404666e-05, "loss": 0.0504, "step": 48340 }, { "grad_norm": 0.5993198156356812, "learning_rate": 9.028058319365534e-05, "loss": 0.0538, "step": 48350 }, { "grad_norm": 0.5179465413093567, "learning_rate": 9.027568470200801e-05, "loss": 0.0508, "step": 48360 }, { "grad_norm": 0.5967801213264465, "learning_rate": 9.02707851092386e-05, "loss": 0.0439, "step": 48370 }, { "grad_norm": 0.7832048535346985, "learning_rate": 9.026588441548105e-05, "loss": 0.046, "step": 48380 }, { "grad_norm": 0.4902597963809967, "learning_rate": 9.026098262086934e-05, "loss": 0.0398, "step": 48390 }, { "grad_norm": 0.6817664504051208, "learning_rate": 9.025607972553749e-05, "loss": 0.0478, "step": 48400 }, { "grad_norm": 0.8840466737747192, "learning_rate": 9.025117572961956e-05, "loss": 0.0449, "step": 48410 }, { "grad_norm": 0.40859854221343994, "learning_rate": 9.024627063324959e-05, "loss": 0.0484, "step": 48420 }, { "grad_norm": 0.39960777759552, "learning_rate": 9.024136443656172e-05, "loss": 0.0443, "step": 48430 }, { "grad_norm": 0.5268636345863342, "learning_rate": 9.023645713969004e-05, "loss": 0.0599, "step": 48440 }, { "grad_norm": 0.6223659515380859, "learning_rate": 9.023154874276875e-05, "loss": 0.0437, "step": 48450 }, { "grad_norm": 0.46663764119148254, "learning_rate": 9.022663924593204e-05, "loss": 0.0504, "step": 48460 }, { "grad_norm": 0.7134353518486023, "learning_rate": 9.02217286493141e-05, "loss": 0.0457, "step": 48470 }, { "grad_norm": 0.38302624225616455, "learning_rate": 9.021681695304924e-05, "loss": 0.0358, "step": 48480 }, { "grad_norm": 0.583364725112915, "learning_rate": 9.02119041572717e-05, "loss": 0.0457, "step": 48490 }, { "grad_norm": 0.6830907464027405, "learning_rate": 9.020699026211581e-05, "loss": 0.0468, "step": 48500 }, { "grad_norm": 0.6658917665481567, "learning_rate": 9.02020752677159e-05, "loss": 0.0443, "step": 48510 }, { "grad_norm": 0.8464341759681702, "learning_rate": 9.019715917420637e-05, "loss": 0.0544, "step": 48520 }, { "grad_norm": 0.6367377638816833, "learning_rate": 9.01922419817216e-05, "loss": 0.0514, "step": 48530 }, { "grad_norm": 0.8984776139259338, "learning_rate": 9.018732369039604e-05, "loss": 0.049, "step": 48540 }, { "grad_norm": 0.5782573223114014, "learning_rate": 9.018240430036413e-05, "loss": 0.0507, "step": 48550 }, { "grad_norm": 0.5821977853775024, "learning_rate": 9.017748381176042e-05, "loss": 0.0569, "step": 48560 }, { "grad_norm": 0.9009358286857605, "learning_rate": 9.017256222471937e-05, "loss": 0.0485, "step": 48570 }, { "grad_norm": 0.514107346534729, "learning_rate": 9.016763953937555e-05, "loss": 0.0641, "step": 48580 }, { "grad_norm": 0.5643556118011475, "learning_rate": 9.016271575586356e-05, "loss": 0.0654, "step": 48590 }, { "grad_norm": 0.9543953537940979, "learning_rate": 9.015779087431803e-05, "loss": 0.0471, "step": 48600 }, { "grad_norm": 0.8943747878074646, "learning_rate": 9.015286489487358e-05, "loss": 0.0513, "step": 48610 }, { "grad_norm": 0.691168487071991, "learning_rate": 9.014793781766489e-05, "loss": 0.0584, "step": 48620 }, { "grad_norm": 0.3869909942150116, "learning_rate": 9.014300964282665e-05, "loss": 0.0429, "step": 48630 }, { "grad_norm": 0.6053623557090759, "learning_rate": 9.013808037049361e-05, "loss": 0.0592, "step": 48640 }, { "grad_norm": 0.5038873553276062, "learning_rate": 9.013315000080053e-05, "loss": 0.0461, "step": 48650 }, { "grad_norm": 0.6089090704917908, "learning_rate": 9.012821853388221e-05, "loss": 0.0437, "step": 48660 }, { "grad_norm": 0.5290356874465942, "learning_rate": 9.012328596987348e-05, "loss": 0.0468, "step": 48670 }, { "grad_norm": 0.6947416067123413, "learning_rate": 9.011835230890917e-05, "loss": 0.0506, "step": 48680 }, { "grad_norm": 0.7696360349655151, "learning_rate": 9.011341755112419e-05, "loss": 0.0467, "step": 48690 }, { "grad_norm": 0.6196701526641846, "learning_rate": 9.010848169665344e-05, "loss": 0.0517, "step": 48700 }, { "grad_norm": 0.5538312792778015, "learning_rate": 9.010354474563188e-05, "loss": 0.0473, "step": 48710 }, { "grad_norm": 0.6151639819145203, "learning_rate": 9.009860669819445e-05, "loss": 0.0342, "step": 48720 }, { "grad_norm": 0.3700309991836548, "learning_rate": 9.00936675544762e-05, "loss": 0.0646, "step": 48730 }, { "grad_norm": 0.564778208732605, "learning_rate": 9.008872731461211e-05, "loss": 0.0452, "step": 48740 }, { "grad_norm": 0.5400006175041199, "learning_rate": 9.00837859787373e-05, "loss": 0.0412, "step": 48750 }, { "grad_norm": 0.8262335658073425, "learning_rate": 9.007884354698683e-05, "loss": 0.0417, "step": 48760 }, { "grad_norm": 0.6302850246429443, "learning_rate": 9.007390001949585e-05, "loss": 0.0413, "step": 48770 }, { "grad_norm": 0.5029094815254211, "learning_rate": 9.006895539639948e-05, "loss": 0.0363, "step": 48780 }, { "grad_norm": 0.359756201505661, "learning_rate": 9.006400967783294e-05, "loss": 0.0439, "step": 48790 }, { "grad_norm": 0.5421027541160583, "learning_rate": 9.005906286393143e-05, "loss": 0.0556, "step": 48800 }, { "grad_norm": 0.6287646889686584, "learning_rate": 9.005411495483016e-05, "loss": 0.0522, "step": 48810 }, { "grad_norm": 0.7894074320793152, "learning_rate": 9.004916595066446e-05, "loss": 0.0572, "step": 48820 }, { "grad_norm": 0.7307039499282837, "learning_rate": 9.004421585156962e-05, "loss": 0.0461, "step": 48830 }, { "grad_norm": 0.8663018941879272, "learning_rate": 9.003926465768094e-05, "loss": 0.0559, "step": 48840 }, { "grad_norm": 0.5650827884674072, "learning_rate": 9.00343123691338e-05, "loss": 0.0499, "step": 48850 }, { "grad_norm": 0.6130111217498779, "learning_rate": 9.002935898606361e-05, "loss": 0.0556, "step": 48860 }, { "grad_norm": 0.5644935965538025, "learning_rate": 9.002440450860579e-05, "loss": 0.0409, "step": 48870 }, { "grad_norm": 0.3635275363922119, "learning_rate": 9.001944893689578e-05, "loss": 0.0411, "step": 48880 }, { "grad_norm": 0.5433704853057861, "learning_rate": 9.001449227106908e-05, "loss": 0.0513, "step": 48890 }, { "grad_norm": 0.5244060158729553, "learning_rate": 9.000953451126118e-05, "loss": 0.0533, "step": 48900 }, { "grad_norm": 0.5392687320709229, "learning_rate": 9.000457565760764e-05, "loss": 0.0457, "step": 48910 }, { "grad_norm": 0.5552855134010315, "learning_rate": 8.999961571024405e-05, "loss": 0.0616, "step": 48920 }, { "grad_norm": 0.5747401118278503, "learning_rate": 8.999465466930597e-05, "loss": 0.0484, "step": 48930 }, { "grad_norm": 0.564567506313324, "learning_rate": 8.998969253492906e-05, "loss": 0.0538, "step": 48940 }, { "grad_norm": 0.590727686882019, "learning_rate": 8.998472930724899e-05, "loss": 0.0504, "step": 48950 }, { "grad_norm": 0.6225653290748596, "learning_rate": 8.997976498640144e-05, "loss": 0.0561, "step": 48960 }, { "grad_norm": 2.0268733501434326, "learning_rate": 8.997479957252213e-05, "loss": 0.0638, "step": 48970 }, { "grad_norm": 0.55866938829422, "learning_rate": 8.996983306574682e-05, "loss": 0.0476, "step": 48980 }, { "grad_norm": 0.6366000771522522, "learning_rate": 8.996486546621129e-05, "loss": 0.0485, "step": 48990 }, { "grad_norm": 0.6165544986724854, "learning_rate": 8.995989677405135e-05, "loss": 0.0564, "step": 49000 }, { "grad_norm": 0.8068174123764038, "learning_rate": 8.995492698940284e-05, "loss": 0.0464, "step": 49010 }, { "grad_norm": 0.434734970331192, "learning_rate": 8.994995611240164e-05, "loss": 0.048, "step": 49020 }, { "grad_norm": 0.8893352150917053, "learning_rate": 8.994498414318366e-05, "loss": 0.0752, "step": 49030 }, { "grad_norm": 0.5206230282783508, "learning_rate": 8.994001108188481e-05, "loss": 0.0523, "step": 49040 }, { "grad_norm": 0.6267284750938416, "learning_rate": 8.993503692864107e-05, "loss": 0.038, "step": 49050 }, { "grad_norm": 0.8055266737937927, "learning_rate": 8.993006168358841e-05, "loss": 0.0588, "step": 49060 }, { "grad_norm": 0.5554595589637756, "learning_rate": 8.992508534686287e-05, "loss": 0.0555, "step": 49070 }, { "grad_norm": 0.798259973526001, "learning_rate": 8.992010791860051e-05, "loss": 0.047, "step": 49080 }, { "grad_norm": 0.7831912636756897, "learning_rate": 8.991512939893737e-05, "loss": 0.0489, "step": 49090 }, { "grad_norm": 0.6641559600830078, "learning_rate": 8.99101497880096e-05, "loss": 0.0476, "step": 49100 }, { "grad_norm": 0.48449578881263733, "learning_rate": 8.990516908595333e-05, "loss": 0.0455, "step": 49110 }, { "grad_norm": 0.30195972323417664, "learning_rate": 8.990018729290473e-05, "loss": 0.0369, "step": 49120 }, { "grad_norm": 0.38032156229019165, "learning_rate": 8.9895204409e-05, "loss": 0.0498, "step": 49130 }, { "grad_norm": 0.4533594846725464, "learning_rate": 8.989022043437536e-05, "loss": 0.0415, "step": 49140 }, { "grad_norm": 0.42653512954711914, "learning_rate": 8.988523536916709e-05, "loss": 0.0417, "step": 49150 }, { "grad_norm": 0.398877888917923, "learning_rate": 8.988024921351147e-05, "loss": 0.0483, "step": 49160 }, { "grad_norm": 0.5214009881019592, "learning_rate": 8.987526196754481e-05, "loss": 0.0424, "step": 49170 }, { "grad_norm": 0.4512822926044464, "learning_rate": 8.987027363140348e-05, "loss": 0.0404, "step": 49180 }, { "grad_norm": 0.4709712862968445, "learning_rate": 8.986528420522384e-05, "loss": 0.0427, "step": 49190 }, { "grad_norm": 0.49709925055503845, "learning_rate": 8.986029368914231e-05, "loss": 0.047, "step": 49200 }, { "grad_norm": 0.6028858423233032, "learning_rate": 8.985530208329531e-05, "loss": 0.0516, "step": 49210 }, { "grad_norm": 0.49899718165397644, "learning_rate": 8.985030938781935e-05, "loss": 0.048, "step": 49220 }, { "grad_norm": 0.9178043007850647, "learning_rate": 8.984531560285088e-05, "loss": 0.0443, "step": 49230 }, { "grad_norm": 0.6271170973777771, "learning_rate": 8.984032072852647e-05, "loss": 0.0336, "step": 49240 }, { "grad_norm": 0.6157333254814148, "learning_rate": 8.983532476498264e-05, "loss": 0.0589, "step": 49250 }, { "grad_norm": 1.0142216682434082, "learning_rate": 8.9830327712356e-05, "loss": 0.0639, "step": 49260 }, { "grad_norm": 0.44030657410621643, "learning_rate": 8.982532957078317e-05, "loss": 0.0509, "step": 49270 }, { "grad_norm": 0.6575967669487, "learning_rate": 8.982033034040078e-05, "loss": 0.0477, "step": 49280 }, { "grad_norm": 0.3965693712234497, "learning_rate": 8.981533002134553e-05, "loss": 0.0484, "step": 49290 }, { "grad_norm": 0.535734236240387, "learning_rate": 8.981032861375411e-05, "loss": 0.0472, "step": 49300 }, { "grad_norm": 0.49224305152893066, "learning_rate": 8.980532611776327e-05, "loss": 0.052, "step": 49310 }, { "grad_norm": 0.47045043110847473, "learning_rate": 8.980032253350976e-05, "loss": 0.0548, "step": 49320 }, { "grad_norm": 0.44549551606178284, "learning_rate": 8.979531786113039e-05, "loss": 0.0568, "step": 49330 }, { "grad_norm": 0.6533806324005127, "learning_rate": 8.979031210076199e-05, "loss": 0.0604, "step": 49340 }, { "grad_norm": 0.40419745445251465, "learning_rate": 8.978530525254138e-05, "loss": 0.0493, "step": 49350 }, { "grad_norm": 0.5673326253890991, "learning_rate": 8.978029731660548e-05, "loss": 0.042, "step": 49360 }, { "grad_norm": 0.41027435660362244, "learning_rate": 8.977528829309121e-05, "loss": 0.0408, "step": 49370 }, { "grad_norm": 0.4502699077129364, "learning_rate": 8.977027818213548e-05, "loss": 0.0381, "step": 49380 }, { "grad_norm": 0.6684470772743225, "learning_rate": 8.976526698387531e-05, "loss": 0.0437, "step": 49390 }, { "grad_norm": 0.7330896854400635, "learning_rate": 8.976025469844766e-05, "loss": 0.0604, "step": 49400 }, { "grad_norm": 1.06681489944458, "learning_rate": 8.975524132598958e-05, "loss": 0.0663, "step": 49410 }, { "grad_norm": 1.1041796207427979, "learning_rate": 8.975022686663816e-05, "loss": 0.0573, "step": 49420 }, { "grad_norm": 0.9883832335472107, "learning_rate": 8.974521132053045e-05, "loss": 0.0519, "step": 49430 }, { "grad_norm": 0.7496740221977234, "learning_rate": 8.97401946878036e-05, "loss": 0.0619, "step": 49440 }, { "grad_norm": 0.6344165802001953, "learning_rate": 8.973517696859476e-05, "loss": 0.0495, "step": 49450 }, { "grad_norm": 0.6653317809104919, "learning_rate": 8.97301581630411e-05, "loss": 0.0515, "step": 49460 }, { "grad_norm": 0.8324390053749084, "learning_rate": 8.972513827127987e-05, "loss": 0.0661, "step": 49470 }, { "grad_norm": 0.747373640537262, "learning_rate": 8.972011729344824e-05, "loss": 0.0503, "step": 49480 }, { "grad_norm": 0.5506784915924072, "learning_rate": 8.971509522968356e-05, "loss": 0.0611, "step": 49490 }, { "grad_norm": 0.5691434741020203, "learning_rate": 8.971007208012309e-05, "loss": 0.0692, "step": 49500 }, { "grad_norm": 0.6224528551101685, "learning_rate": 8.970504784490417e-05, "loss": 0.0451, "step": 49510 }, { "grad_norm": 0.6693350672721863, "learning_rate": 8.970002252416415e-05, "loss": 0.0604, "step": 49520 }, { "grad_norm": 0.8803062438964844, "learning_rate": 8.969499611804041e-05, "loss": 0.0598, "step": 49530 }, { "grad_norm": 0.5300438404083252, "learning_rate": 8.968996862667042e-05, "loss": 0.0515, "step": 49540 }, { "grad_norm": 0.4775875508785248, "learning_rate": 8.96849400501916e-05, "loss": 0.0443, "step": 49550 }, { "grad_norm": 1.051024317741394, "learning_rate": 8.967991038874142e-05, "loss": 0.0499, "step": 49560 }, { "grad_norm": 0.5674121975898743, "learning_rate": 8.967487964245738e-05, "loss": 0.0501, "step": 49570 }, { "grad_norm": 0.7666316628456116, "learning_rate": 8.966984781147705e-05, "loss": 0.072, "step": 49580 }, { "grad_norm": 0.46240299940109253, "learning_rate": 8.966481489593799e-05, "loss": 0.036, "step": 49590 }, { "grad_norm": 0.5191309452056885, "learning_rate": 8.965978089597778e-05, "loss": 0.0454, "step": 49600 }, { "grad_norm": 0.511976957321167, "learning_rate": 8.965474581173406e-05, "loss": 0.0543, "step": 49610 }, { "grad_norm": 0.5174558162689209, "learning_rate": 8.964970964334449e-05, "loss": 0.0514, "step": 49620 }, { "grad_norm": 0.9809896349906921, "learning_rate": 8.964467239094677e-05, "loss": 0.0553, "step": 49630 }, { "grad_norm": 0.7826922535896301, "learning_rate": 8.963963405467858e-05, "loss": 0.0547, "step": 49640 }, { "grad_norm": 1.0272942781448364, "learning_rate": 8.96345946346777e-05, "loss": 0.0634, "step": 49650 }, { "grad_norm": 0.42492806911468506, "learning_rate": 8.962955413108187e-05, "loss": 0.0364, "step": 49660 }, { "grad_norm": 0.4881943464279175, "learning_rate": 8.962451254402896e-05, "loss": 0.0412, "step": 49670 }, { "grad_norm": 1.2011654376983643, "learning_rate": 8.961946987365673e-05, "loss": 0.0484, "step": 49680 }, { "grad_norm": 0.4368753731250763, "learning_rate": 8.961442612010308e-05, "loss": 0.0482, "step": 49690 }, { "grad_norm": 0.4787503182888031, "learning_rate": 8.960938128350591e-05, "loss": 0.0415, "step": 49700 }, { "grad_norm": 1.2786734104156494, "learning_rate": 8.960433536400313e-05, "loss": 0.0526, "step": 49710 }, { "grad_norm": 0.490382581949234, "learning_rate": 8.95992883617327e-05, "loss": 0.0357, "step": 49720 }, { "grad_norm": 0.6624704003334045, "learning_rate": 8.95942402768326e-05, "loss": 0.0465, "step": 49730 }, { "grad_norm": 0.47221359610557556, "learning_rate": 8.958919110944087e-05, "loss": 0.079, "step": 49740 }, { "grad_norm": 0.4441322088241577, "learning_rate": 8.95841408596955e-05, "loss": 0.0457, "step": 49750 }, { "grad_norm": 0.40453052520751953, "learning_rate": 8.957908952773461e-05, "loss": 0.0358, "step": 49760 }, { "grad_norm": 0.6827235221862793, "learning_rate": 8.957403711369628e-05, "loss": 0.0556, "step": 49770 }, { "grad_norm": 0.5341347455978394, "learning_rate": 8.956898361771863e-05, "loss": 0.0454, "step": 49780 }, { "grad_norm": 0.4465799331665039, "learning_rate": 8.956392903993984e-05, "loss": 0.052, "step": 49790 }, { "grad_norm": 0.8135903477668762, "learning_rate": 8.955887338049809e-05, "loss": 0.0677, "step": 49800 }, { "grad_norm": 0.44220244884490967, "learning_rate": 8.955381663953159e-05, "loss": 0.0483, "step": 49810 }, { "grad_norm": 0.5949645638465881, "learning_rate": 8.954875881717864e-05, "loss": 0.043, "step": 49820 }, { "grad_norm": 0.3924080729484558, "learning_rate": 8.954369991357745e-05, "loss": 0.0489, "step": 49830 }, { "grad_norm": 0.3536205589771271, "learning_rate": 8.953863992886638e-05, "loss": 0.0415, "step": 49840 }, { "grad_norm": 0.9824429750442505, "learning_rate": 8.953357886318372e-05, "loss": 0.0439, "step": 49850 }, { "grad_norm": 0.42631417512893677, "learning_rate": 8.952851671666788e-05, "loss": 0.0414, "step": 49860 }, { "grad_norm": 0.5505945086479187, "learning_rate": 8.952345348945725e-05, "loss": 0.0441, "step": 49870 }, { "grad_norm": 0.6303761005401611, "learning_rate": 8.951838918169027e-05, "loss": 0.0587, "step": 49880 }, { "grad_norm": 0.3535996079444885, "learning_rate": 8.951332379350535e-05, "loss": 0.0422, "step": 49890 }, { "grad_norm": 0.3719363808631897, "learning_rate": 8.9508257325041e-05, "loss": 0.0613, "step": 49900 }, { "grad_norm": 0.8228573799133301, "learning_rate": 8.950318977643574e-05, "loss": 0.0434, "step": 49910 }, { "grad_norm": 1.1769800186157227, "learning_rate": 8.949812114782811e-05, "loss": 0.0562, "step": 49920 }, { "grad_norm": 0.6835915446281433, "learning_rate": 8.949305143935669e-05, "loss": 0.0536, "step": 49930 }, { "grad_norm": 0.8548206090927124, "learning_rate": 8.948798065116008e-05, "loss": 0.0526, "step": 49940 }, { "grad_norm": 0.6165660619735718, "learning_rate": 8.948290878337692e-05, "loss": 0.0413, "step": 49950 }, { "grad_norm": 0.4251493215560913, "learning_rate": 8.947783583614586e-05, "loss": 0.0503, "step": 49960 }, { "grad_norm": 0.7641896605491638, "learning_rate": 8.947276180960559e-05, "loss": 0.0426, "step": 49970 }, { "grad_norm": 0.48597145080566406, "learning_rate": 8.946768670389487e-05, "loss": 0.0641, "step": 49980 }, { "grad_norm": 0.7033815383911133, "learning_rate": 8.94626105191524e-05, "loss": 0.0384, "step": 49990 }, { "grad_norm": 0.5719709992408752, "learning_rate": 8.9457533255517e-05, "loss": 0.0509, "step": 50000 }, { "grad_norm": 0.914472222328186, "learning_rate": 8.945245491312744e-05, "loss": 0.0456, "step": 50010 }, { "grad_norm": 0.6189765930175781, "learning_rate": 8.944737549212261e-05, "loss": 0.0443, "step": 50020 }, { "grad_norm": 0.4910527467727661, "learning_rate": 8.944229499264135e-05, "loss": 0.0461, "step": 50030 }, { "grad_norm": 0.7193816900253296, "learning_rate": 8.943721341482253e-05, "loss": 0.0548, "step": 50040 }, { "grad_norm": 0.9121129512786865, "learning_rate": 8.943213075880514e-05, "loss": 0.0468, "step": 50050 }, { "grad_norm": 0.8441412448883057, "learning_rate": 8.942704702472812e-05, "loss": 0.0692, "step": 50060 }, { "grad_norm": 0.5036449432373047, "learning_rate": 8.942196221273043e-05, "loss": 0.0431, "step": 50070 }, { "grad_norm": 0.35901907086372375, "learning_rate": 8.94168763229511e-05, "loss": 0.0347, "step": 50080 }, { "grad_norm": 0.4704076051712036, "learning_rate": 8.941178935552918e-05, "loss": 0.0522, "step": 50090 }, { "grad_norm": 0.43674811720848083, "learning_rate": 8.940670131060375e-05, "loss": 0.0413, "step": 50100 }, { "grad_norm": 0.9303649663925171, "learning_rate": 8.940161218831391e-05, "loss": 0.0506, "step": 50110 }, { "grad_norm": 0.6524479985237122, "learning_rate": 8.939652198879879e-05, "loss": 0.0445, "step": 50120 }, { "grad_norm": 0.7617766857147217, "learning_rate": 8.939143071219758e-05, "loss": 0.0427, "step": 50130 }, { "grad_norm": 0.5124600529670715, "learning_rate": 8.938633835864944e-05, "loss": 0.0411, "step": 50140 }, { "grad_norm": 0.6810181140899658, "learning_rate": 8.93812449282936e-05, "loss": 0.038, "step": 50150 }, { "grad_norm": 0.7213748097419739, "learning_rate": 8.937615042126932e-05, "loss": 0.0476, "step": 50160 }, { "grad_norm": 0.6664255857467651, "learning_rate": 8.937105483771588e-05, "loss": 0.048, "step": 50170 }, { "grad_norm": 0.6796830296516418, "learning_rate": 8.936595817777261e-05, "loss": 0.0554, "step": 50180 }, { "grad_norm": 0.7034987211227417, "learning_rate": 8.936086044157881e-05, "loss": 0.0449, "step": 50190 }, { "grad_norm": 0.9291708469390869, "learning_rate": 8.935576162927388e-05, "loss": 0.0456, "step": 50200 }, { "grad_norm": 0.5608726143836975, "learning_rate": 8.935066174099721e-05, "loss": 0.0506, "step": 50210 }, { "grad_norm": 0.5928551554679871, "learning_rate": 8.934556077688823e-05, "loss": 0.0402, "step": 50220 }, { "grad_norm": 0.560652494430542, "learning_rate": 8.934045873708641e-05, "loss": 0.0548, "step": 50230 }, { "grad_norm": 0.6033722162246704, "learning_rate": 8.933535562173121e-05, "loss": 0.0547, "step": 50240 }, { "grad_norm": 0.7290727496147156, "learning_rate": 8.933025143096219e-05, "loss": 0.0745, "step": 50250 }, { "grad_norm": 0.6474990248680115, "learning_rate": 8.932514616491885e-05, "loss": 0.0614, "step": 50260 }, { "grad_norm": 0.6915316581726074, "learning_rate": 8.93200398237408e-05, "loss": 0.0593, "step": 50270 }, { "grad_norm": 0.8025569319725037, "learning_rate": 8.931493240756762e-05, "loss": 0.0674, "step": 50280 }, { "grad_norm": 0.459741473197937, "learning_rate": 8.930982391653897e-05, "loss": 0.0492, "step": 50290 }, { "grad_norm": 0.7773531079292297, "learning_rate": 8.93047143507945e-05, "loss": 0.0523, "step": 50300 }, { "grad_norm": 0.49940750002861023, "learning_rate": 8.92996037104739e-05, "loss": 0.0489, "step": 50310 }, { "grad_norm": 0.7442473769187927, "learning_rate": 8.929449199571692e-05, "loss": 0.0407, "step": 50320 }, { "grad_norm": 0.7105170488357544, "learning_rate": 8.928937920666325e-05, "loss": 0.0403, "step": 50330 }, { "grad_norm": 0.7410491704940796, "learning_rate": 8.928426534345274e-05, "loss": 0.0401, "step": 50340 }, { "grad_norm": 0.6643478274345398, "learning_rate": 8.927915040622518e-05, "loss": 0.045, "step": 50350 }, { "grad_norm": 0.5758014917373657, "learning_rate": 8.927403439512039e-05, "loss": 0.0494, "step": 50360 }, { "grad_norm": 0.740488588809967, "learning_rate": 8.926891731027826e-05, "loss": 0.068, "step": 50370 }, { "grad_norm": 0.46770206093788147, "learning_rate": 8.926379915183869e-05, "loss": 0.0479, "step": 50380 }, { "grad_norm": 0.5310547947883606, "learning_rate": 8.925867991994159e-05, "loss": 0.0563, "step": 50390 }, { "grad_norm": 0.43920883536338806, "learning_rate": 8.925355961472692e-05, "loss": 0.051, "step": 50400 }, { "grad_norm": 0.7709836363792419, "learning_rate": 8.924843823633471e-05, "loss": 0.0658, "step": 50410 }, { "grad_norm": 0.6025333404541016, "learning_rate": 8.924331578490492e-05, "loss": 0.0628, "step": 50420 }, { "grad_norm": 0.5859851241111755, "learning_rate": 8.923819226057762e-05, "loss": 0.066, "step": 50430 }, { "grad_norm": 0.44011610746383667, "learning_rate": 8.923306766349288e-05, "loss": 0.0456, "step": 50440 }, { "grad_norm": 0.6620529294013977, "learning_rate": 8.922794199379084e-05, "loss": 0.049, "step": 50450 }, { "grad_norm": 1.2061339616775513, "learning_rate": 8.922281525161157e-05, "loss": 0.0632, "step": 50460 }, { "grad_norm": 1.6312291622161865, "learning_rate": 8.921768743709528e-05, "loss": 0.0417, "step": 50470 }, { "grad_norm": 0.745357096195221, "learning_rate": 8.921255855038215e-05, "loss": 0.0568, "step": 50480 }, { "grad_norm": 0.7222909331321716, "learning_rate": 8.92074285916124e-05, "loss": 0.0483, "step": 50490 }, { "grad_norm": 0.5163357257843018, "learning_rate": 8.920229756092626e-05, "loss": 0.0426, "step": 50500 }, { "grad_norm": 0.5537539720535278, "learning_rate": 8.919716545846406e-05, "loss": 0.0442, "step": 50510 }, { "grad_norm": 0.5063778758049011, "learning_rate": 8.919203228436608e-05, "loss": 0.0464, "step": 50520 }, { "grad_norm": 0.4833618104457855, "learning_rate": 8.918689803877265e-05, "loss": 0.0508, "step": 50530 }, { "grad_norm": 0.5304726362228394, "learning_rate": 8.918176272182415e-05, "loss": 0.0537, "step": 50540 }, { "grad_norm": 0.5637051463127136, "learning_rate": 8.917662633366099e-05, "loss": 0.0471, "step": 50550 }, { "grad_norm": 0.4109736979007721, "learning_rate": 8.917148887442357e-05, "loss": 0.0425, "step": 50560 }, { "grad_norm": 0.5308313965797424, "learning_rate": 8.916635034425235e-05, "loss": 0.0495, "step": 50570 }, { "grad_norm": 0.6136873960494995, "learning_rate": 8.916121074328786e-05, "loss": 0.0374, "step": 50580 }, { "grad_norm": 0.4133606255054474, "learning_rate": 8.915607007167055e-05, "loss": 0.0469, "step": 50590 }, { "grad_norm": 0.4524261951446533, "learning_rate": 8.915092832954102e-05, "loss": 0.0422, "step": 50600 }, { "grad_norm": 0.6002103686332703, "learning_rate": 8.91457855170398e-05, "loss": 0.0438, "step": 50610 }, { "grad_norm": 0.8570559620857239, "learning_rate": 8.914064163430751e-05, "loss": 0.0472, "step": 50620 }, { "grad_norm": 1.1181151866912842, "learning_rate": 8.913549668148479e-05, "loss": 0.0382, "step": 50630 }, { "grad_norm": 0.5610479116439819, "learning_rate": 8.91303506587123e-05, "loss": 0.0424, "step": 50640 }, { "grad_norm": 0.39693906903266907, "learning_rate": 8.91252035661307e-05, "loss": 0.0692, "step": 50650 }, { "grad_norm": 0.7247180342674255, "learning_rate": 8.912005540388076e-05, "loss": 0.0531, "step": 50660 }, { "grad_norm": 0.394420862197876, "learning_rate": 8.911490617210319e-05, "loss": 0.036, "step": 50670 }, { "grad_norm": 0.3644818961620331, "learning_rate": 8.910975587093877e-05, "loss": 0.0559, "step": 50680 }, { "grad_norm": 0.5526459217071533, "learning_rate": 8.910460450052832e-05, "loss": 0.0402, "step": 50690 }, { "grad_norm": 0.9389615654945374, "learning_rate": 8.909945206101268e-05, "loss": 0.0473, "step": 50700 }, { "grad_norm": 0.426825612783432, "learning_rate": 8.909429855253271e-05, "loss": 0.04, "step": 50710 }, { "grad_norm": 0.7271881103515625, "learning_rate": 8.90891439752293e-05, "loss": 0.0571, "step": 50720 }, { "grad_norm": 0.602515697479248, "learning_rate": 8.908398832924338e-05, "loss": 0.054, "step": 50730 }, { "grad_norm": 0.7576499581336975, "learning_rate": 8.907883161471591e-05, "loss": 0.0528, "step": 50740 }, { "grad_norm": 0.5043877959251404, "learning_rate": 8.907367383178786e-05, "loss": 0.0455, "step": 50750 }, { "grad_norm": 0.44497451186180115, "learning_rate": 8.906851498060025e-05, "loss": 0.0415, "step": 50760 }, { "grad_norm": 0.8857517838478088, "learning_rate": 8.906335506129411e-05, "loss": 0.0459, "step": 50770 }, { "grad_norm": 0.5477506518363953, "learning_rate": 8.905819407401052e-05, "loss": 0.0479, "step": 50780 }, { "grad_norm": 0.5607220530509949, "learning_rate": 8.905303201889058e-05, "loss": 0.0428, "step": 50790 }, { "grad_norm": 0.6267293691635132, "learning_rate": 8.904786889607541e-05, "loss": 0.0421, "step": 50800 }, { "grad_norm": 0.6477819085121155, "learning_rate": 8.904270470570619e-05, "loss": 0.0447, "step": 50810 }, { "grad_norm": 0.654279887676239, "learning_rate": 8.903753944792408e-05, "loss": 0.0435, "step": 50820 }, { "grad_norm": 0.5194525122642517, "learning_rate": 8.903237312287031e-05, "loss": 0.0295, "step": 50830 }, { "grad_norm": 0.7054762244224548, "learning_rate": 8.902720573068612e-05, "loss": 0.0458, "step": 50840 }, { "grad_norm": 0.5588445067405701, "learning_rate": 8.90220372715128e-05, "loss": 0.0458, "step": 50850 }, { "grad_norm": 0.6336686015129089, "learning_rate": 8.901686774549162e-05, "loss": 0.0686, "step": 50860 }, { "grad_norm": 0.9106007218360901, "learning_rate": 8.901169715276394e-05, "loss": 0.0531, "step": 50870 }, { "grad_norm": 0.918599545955658, "learning_rate": 8.900652549347113e-05, "loss": 0.0467, "step": 50880 }, { "grad_norm": 0.9194434285163879, "learning_rate": 8.900135276775454e-05, "loss": 0.0805, "step": 50890 }, { "grad_norm": 0.598770797252655, "learning_rate": 8.899617897575564e-05, "loss": 0.0435, "step": 50900 }, { "grad_norm": 0.7770384550094604, "learning_rate": 8.899100411761585e-05, "loss": 0.0541, "step": 50910 }, { "grad_norm": 0.7711860537528992, "learning_rate": 8.898582819347665e-05, "loss": 0.0458, "step": 50920 }, { "grad_norm": 0.9448152184486389, "learning_rate": 8.898065120347956e-05, "loss": 0.0688, "step": 50930 }, { "grad_norm": 0.7779993414878845, "learning_rate": 8.89754731477661e-05, "loss": 0.0471, "step": 50940 }, { "grad_norm": 0.5104960203170776, "learning_rate": 8.897029402647787e-05, "loss": 0.0417, "step": 50950 }, { "grad_norm": 0.48429766297340393, "learning_rate": 8.896511383975643e-05, "loss": 0.0442, "step": 50960 }, { "grad_norm": 1.1681723594665527, "learning_rate": 8.89599325877434e-05, "loss": 0.046, "step": 50970 }, { "grad_norm": 0.5420631766319275, "learning_rate": 8.895475027058046e-05, "loss": 0.0461, "step": 50980 }, { "grad_norm": 0.6764789819717407, "learning_rate": 8.89495668884093e-05, "loss": 0.0466, "step": 50990 }, { "grad_norm": 0.7017671465873718, "learning_rate": 8.89443824413716e-05, "loss": 0.061, "step": 51000 }, { "grad_norm": 0.5155240893363953, "learning_rate": 8.893919692960912e-05, "loss": 0.0578, "step": 51010 }, { "grad_norm": 0.4962780177593231, "learning_rate": 8.89340103532636e-05, "loss": 0.042, "step": 51020 }, { "grad_norm": 1.2273186445236206, "learning_rate": 8.892882271247689e-05, "loss": 0.045, "step": 51030 }, { "grad_norm": 0.5128408670425415, "learning_rate": 8.892363400739079e-05, "loss": 0.0533, "step": 51040 }, { "grad_norm": 0.6570837497711182, "learning_rate": 8.891844423814715e-05, "loss": 0.0523, "step": 51050 }, { "grad_norm": 0.4501485526561737, "learning_rate": 8.891325340488787e-05, "loss": 0.0556, "step": 51060 }, { "grad_norm": 0.5679407715797424, "learning_rate": 8.890806150775486e-05, "loss": 0.0515, "step": 51070 }, { "grad_norm": 0.7301923036575317, "learning_rate": 8.890286854689007e-05, "loss": 0.06, "step": 51080 }, { "grad_norm": 1.7905679941177368, "learning_rate": 8.889767452243546e-05, "loss": 0.0415, "step": 51090 }, { "grad_norm": 0.9084315896034241, "learning_rate": 8.889247943453305e-05, "loss": 0.054, "step": 51100 }, { "grad_norm": 0.6174174547195435, "learning_rate": 8.888728328332488e-05, "loss": 0.045, "step": 51110 }, { "grad_norm": 0.36329302191734314, "learning_rate": 8.888208606895296e-05, "loss": 0.0618, "step": 51120 }, { "grad_norm": 0.802941083908081, "learning_rate": 8.887688779155943e-05, "loss": 0.0512, "step": 51130 }, { "grad_norm": 0.5990851521492004, "learning_rate": 8.88716884512864e-05, "loss": 0.0654, "step": 51140 }, { "grad_norm": 0.8670472502708435, "learning_rate": 8.8866488048276e-05, "loss": 0.0571, "step": 51150 }, { "grad_norm": 0.710052490234375, "learning_rate": 8.886128658267041e-05, "loss": 0.0514, "step": 51160 }, { "grad_norm": 0.7300209999084473, "learning_rate": 8.885608405461187e-05, "loss": 0.0603, "step": 51170 }, { "grad_norm": 0.5280730128288269, "learning_rate": 8.885088046424258e-05, "loss": 0.0433, "step": 51180 }, { "grad_norm": 0.5936734676361084, "learning_rate": 8.88456758117048e-05, "loss": 0.0542, "step": 51190 }, { "grad_norm": 0.7227916717529297, "learning_rate": 8.884047009714084e-05, "loss": 0.0501, "step": 51200 }, { "grad_norm": 0.5368281006813049, "learning_rate": 8.883526332069304e-05, "loss": 0.0698, "step": 51210 }, { "grad_norm": 0.43241578340530396, "learning_rate": 8.88300554825037e-05, "loss": 0.053, "step": 51220 }, { "grad_norm": 0.5746235847473145, "learning_rate": 8.882484658271526e-05, "loss": 0.0666, "step": 51230 }, { "grad_norm": 0.5404095649719238, "learning_rate": 8.881963662147008e-05, "loss": 0.0472, "step": 51240 }, { "grad_norm": 0.5370345711708069, "learning_rate": 8.881442559891063e-05, "loss": 0.0469, "step": 51250 }, { "grad_norm": 0.6915647387504578, "learning_rate": 8.880921351517934e-05, "loss": 0.057, "step": 51260 }, { "grad_norm": 0.6679696440696716, "learning_rate": 8.880400037041876e-05, "loss": 0.0452, "step": 51270 }, { "grad_norm": 0.5452608466148376, "learning_rate": 8.87987861647714e-05, "loss": 0.0434, "step": 51280 }, { "grad_norm": 0.4493235647678375, "learning_rate": 8.879357089837977e-05, "loss": 0.0459, "step": 51290 }, { "grad_norm": 0.7463662624359131, "learning_rate": 8.87883545713865e-05, "loss": 0.06, "step": 51300 }, { "grad_norm": 0.5803684592247009, "learning_rate": 8.87831371839342e-05, "loss": 0.0506, "step": 51310 }, { "grad_norm": 0.3467475175857544, "learning_rate": 8.877791873616548e-05, "loss": 0.0452, "step": 51320 }, { "grad_norm": 0.9199070930480957, "learning_rate": 8.877269922822305e-05, "loss": 0.0659, "step": 51330 }, { "grad_norm": 0.6773914098739624, "learning_rate": 8.876747866024956e-05, "loss": 0.0578, "step": 51340 }, { "grad_norm": 0.9336717128753662, "learning_rate": 8.876225703238778e-05, "loss": 0.051, "step": 51350 }, { "grad_norm": 0.5756871700286865, "learning_rate": 8.875703434478047e-05, "loss": 0.0507, "step": 51360 }, { "grad_norm": 0.64304518699646, "learning_rate": 8.875181059757038e-05, "loss": 0.0512, "step": 51370 }, { "grad_norm": 0.5576359033584595, "learning_rate": 8.874658579090036e-05, "loss": 0.0493, "step": 51380 }, { "grad_norm": 0.5254706144332886, "learning_rate": 8.874135992491323e-05, "loss": 0.0452, "step": 51390 }, { "grad_norm": 0.6728185415267944, "learning_rate": 8.87361329997519e-05, "loss": 0.0586, "step": 51400 }, { "grad_norm": 0.5876482725143433, "learning_rate": 8.873090501555922e-05, "loss": 0.0429, "step": 51410 }, { "grad_norm": 0.6455147862434387, "learning_rate": 8.872567597247816e-05, "loss": 0.0505, "step": 51420 }, { "grad_norm": 0.954248309135437, "learning_rate": 8.872044587065167e-05, "loss": 0.0505, "step": 51430 }, { "grad_norm": 0.6406594514846802, "learning_rate": 8.871521471022273e-05, "loss": 0.0387, "step": 51440 }, { "grad_norm": 0.6359517574310303, "learning_rate": 8.870998249133435e-05, "loss": 0.0596, "step": 51450 }, { "grad_norm": 0.5189629793167114, "learning_rate": 8.870474921412962e-05, "loss": 0.0491, "step": 51460 }, { "grad_norm": 0.6406062245368958, "learning_rate": 8.869951487875157e-05, "loss": 0.0477, "step": 51470 }, { "grad_norm": 0.7081637978553772, "learning_rate": 8.869427948534332e-05, "loss": 0.0477, "step": 51480 }, { "grad_norm": 0.8761494159698486, "learning_rate": 8.868904303404801e-05, "loss": 0.0473, "step": 51490 }, { "grad_norm": 0.7037971019744873, "learning_rate": 8.86838055250088e-05, "loss": 0.0561, "step": 51500 }, { "grad_norm": 0.5876902937889099, "learning_rate": 8.867856695836888e-05, "loss": 0.054, "step": 51510 }, { "grad_norm": 1.167954921722412, "learning_rate": 8.867332733427147e-05, "loss": 0.0577, "step": 51520 }, { "grad_norm": 0.6523448824882507, "learning_rate": 8.866808665285981e-05, "loss": 0.0511, "step": 51530 }, { "grad_norm": 0.641671359539032, "learning_rate": 8.866284491427721e-05, "loss": 0.0607, "step": 51540 }, { "grad_norm": 0.5229414105415344, "learning_rate": 8.865760211866694e-05, "loss": 0.0527, "step": 51550 }, { "grad_norm": 0.478037029504776, "learning_rate": 8.865235826617235e-05, "loss": 0.0409, "step": 51560 }, { "grad_norm": 0.688055694103241, "learning_rate": 8.86471133569368e-05, "loss": 0.0666, "step": 51570 }, { "grad_norm": 1.4759095907211304, "learning_rate": 8.86418673911037e-05, "loss": 0.0576, "step": 51580 }, { "grad_norm": 0.572922945022583, "learning_rate": 8.863662036881647e-05, "loss": 0.0439, "step": 51590 }, { "grad_norm": 0.5138960480690002, "learning_rate": 8.863137229021854e-05, "loss": 0.0579, "step": 51600 }, { "grad_norm": 0.7752368450164795, "learning_rate": 8.862612315545341e-05, "loss": 0.0395, "step": 51610 }, { "grad_norm": 0.5427494049072266, "learning_rate": 8.862087296466459e-05, "loss": 0.0477, "step": 51620 }, { "grad_norm": 0.7922075986862183, "learning_rate": 8.86156217179956e-05, "loss": 0.0475, "step": 51630 }, { "grad_norm": 0.45541858673095703, "learning_rate": 8.861036941559002e-05, "loss": 0.0489, "step": 51640 }, { "grad_norm": 0.5695688724517822, "learning_rate": 8.860511605759146e-05, "loss": 0.063, "step": 51650 }, { "grad_norm": 0.5591080188751221, "learning_rate": 8.859986164414351e-05, "loss": 0.0416, "step": 51660 }, { "grad_norm": 0.5646474957466125, "learning_rate": 8.859460617538986e-05, "loss": 0.0612, "step": 51670 }, { "grad_norm": 0.528770387172699, "learning_rate": 8.858934965147417e-05, "loss": 0.0466, "step": 51680 }, { "grad_norm": 0.6465003490447998, "learning_rate": 8.858409207254018e-05, "loss": 0.0532, "step": 51690 }, { "grad_norm": 0.5401199460029602, "learning_rate": 8.857883343873157e-05, "loss": 0.0466, "step": 51700 }, { "grad_norm": 0.3903225362300873, "learning_rate": 8.857357375019218e-05, "loss": 0.0462, "step": 51710 }, { "grad_norm": 0.5307822227478027, "learning_rate": 8.856831300706575e-05, "loss": 0.0491, "step": 51720 }, { "grad_norm": 0.9509256482124329, "learning_rate": 8.856305120949614e-05, "loss": 0.0455, "step": 51730 }, { "grad_norm": 0.5002453327178955, "learning_rate": 8.85577883576272e-05, "loss": 0.0505, "step": 51740 }, { "grad_norm": 0.44399774074554443, "learning_rate": 8.855252445160281e-05, "loss": 0.0748, "step": 51750 }, { "grad_norm": 1.1402812004089355, "learning_rate": 8.854725949156688e-05, "loss": 0.0446, "step": 51760 }, { "grad_norm": 0.5753199458122253, "learning_rate": 8.854199347766337e-05, "loss": 0.0558, "step": 51770 }, { "grad_norm": 0.7202351689338684, "learning_rate": 8.853672641003623e-05, "loss": 0.0426, "step": 51780 }, { "grad_norm": 3.3341894149780273, "learning_rate": 8.853145828882948e-05, "loss": 0.0567, "step": 51790 }, { "grad_norm": 0.5741007924079895, "learning_rate": 8.85261891141871e-05, "loss": 0.0493, "step": 51800 }, { "grad_norm": 0.7041148543357849, "learning_rate": 8.852091888625322e-05, "loss": 0.0537, "step": 51810 }, { "grad_norm": 0.2755487263202667, "learning_rate": 8.851564760517188e-05, "loss": 0.0585, "step": 51820 }, { "grad_norm": 0.515811026096344, "learning_rate": 8.85103752710872e-05, "loss": 0.0451, "step": 51830 }, { "grad_norm": 1.1219818592071533, "learning_rate": 8.850510188414334e-05, "loss": 0.0501, "step": 51840 }, { "grad_norm": 0.930336058139801, "learning_rate": 8.849982744448444e-05, "loss": 0.052, "step": 51850 }, { "grad_norm": 0.5204325914382935, "learning_rate": 8.849455195225473e-05, "loss": 0.0519, "step": 51860 }, { "grad_norm": 0.5827280879020691, "learning_rate": 8.848927540759843e-05, "loss": 0.0516, "step": 51870 }, { "grad_norm": 0.41790440678596497, "learning_rate": 8.848399781065981e-05, "loss": 0.0407, "step": 51880 }, { "grad_norm": 0.9932302236557007, "learning_rate": 8.847871916158312e-05, "loss": 0.0552, "step": 51890 }, { "grad_norm": 0.5525088310241699, "learning_rate": 8.847343946051273e-05, "loss": 0.0469, "step": 51900 }, { "grad_norm": 0.6006070375442505, "learning_rate": 8.846815870759295e-05, "loss": 0.0636, "step": 51910 }, { "grad_norm": 0.4012075960636139, "learning_rate": 8.846287690296817e-05, "loss": 0.0519, "step": 51920 }, { "grad_norm": 0.7296411991119385, "learning_rate": 8.845759404678278e-05, "loss": 0.0491, "step": 51930 }, { "grad_norm": 0.57028728723526, "learning_rate": 8.845231013918122e-05, "loss": 0.0467, "step": 51940 }, { "grad_norm": 0.8612443804740906, "learning_rate": 8.844702518030794e-05, "loss": 0.0412, "step": 51950 }, { "grad_norm": 0.781353771686554, "learning_rate": 8.844173917030744e-05, "loss": 0.0577, "step": 51960 }, { "grad_norm": 0.5750876069068909, "learning_rate": 8.843645210932423e-05, "loss": 0.048, "step": 51970 }, { "grad_norm": 0.7489938139915466, "learning_rate": 8.843116399750287e-05, "loss": 0.044, "step": 51980 }, { "grad_norm": 0.41358211636543274, "learning_rate": 8.842587483498791e-05, "loss": 0.045, "step": 51990 }, { "grad_norm": 0.7070242166519165, "learning_rate": 8.842058462192397e-05, "loss": 0.0454, "step": 52000 }, { "grad_norm": 0.7109112739562988, "learning_rate": 8.841529335845569e-05, "loss": 0.0349, "step": 52010 }, { "grad_norm": 0.760938286781311, "learning_rate": 8.841000104472772e-05, "loss": 0.0686, "step": 52020 }, { "grad_norm": 0.5506317019462585, "learning_rate": 8.840470768088473e-05, "loss": 0.0535, "step": 52030 }, { "grad_norm": 0.49195796251296997, "learning_rate": 8.839941326707149e-05, "loss": 0.0471, "step": 52040 }, { "grad_norm": 0.4223825931549072, "learning_rate": 8.839411780343269e-05, "loss": 0.0443, "step": 52050 }, { "grad_norm": 0.4345225691795349, "learning_rate": 8.838882129011317e-05, "loss": 0.0607, "step": 52060 }, { "grad_norm": 0.6179653406143188, "learning_rate": 8.838352372725767e-05, "loss": 0.0513, "step": 52070 }, { "grad_norm": 0.6871504783630371, "learning_rate": 8.837822511501105e-05, "loss": 0.052, "step": 52080 }, { "grad_norm": 0.6050460934638977, "learning_rate": 8.837292545351818e-05, "loss": 0.0518, "step": 52090 }, { "grad_norm": 0.3676072359085083, "learning_rate": 8.836762474292395e-05, "loss": 0.0598, "step": 52100 }, { "grad_norm": 0.6563490629196167, "learning_rate": 8.836232298337326e-05, "loss": 0.0454, "step": 52110 }, { "grad_norm": 0.7392621636390686, "learning_rate": 8.835702017501108e-05, "loss": 0.0462, "step": 52120 }, { "grad_norm": 0.64401775598526, "learning_rate": 8.835171631798238e-05, "loss": 0.0639, "step": 52130 }, { "grad_norm": 0.4839050769805908, "learning_rate": 8.834641141243217e-05, "loss": 0.0465, "step": 52140 }, { "grad_norm": 0.9062079191207886, "learning_rate": 8.834110545850545e-05, "loss": 0.0508, "step": 52150 }, { "grad_norm": 0.5096215009689331, "learning_rate": 8.833579845634735e-05, "loss": 0.082, "step": 52160 }, { "grad_norm": 0.672787606716156, "learning_rate": 8.83304904061029e-05, "loss": 0.0562, "step": 52170 }, { "grad_norm": 0.634833812713623, "learning_rate": 8.832518130791724e-05, "loss": 0.0648, "step": 52180 }, { "grad_norm": 0.4936014413833618, "learning_rate": 8.831987116193552e-05, "loss": 0.0419, "step": 52190 }, { "grad_norm": 0.5546761155128479, "learning_rate": 8.831455996830294e-05, "loss": 0.0662, "step": 52200 }, { "grad_norm": 0.49300724267959595, "learning_rate": 8.830924772716467e-05, "loss": 0.043, "step": 52210 }, { "grad_norm": 1.0336682796478271, "learning_rate": 8.830393443866596e-05, "loss": 0.0519, "step": 52220 }, { "grad_norm": 0.47294560074806213, "learning_rate": 8.829862010295207e-05, "loss": 0.0563, "step": 52230 }, { "grad_norm": 0.5481534004211426, "learning_rate": 8.829330472016828e-05, "loss": 0.0709, "step": 52240 }, { "grad_norm": 1.0303348302841187, "learning_rate": 8.828798829045994e-05, "loss": 0.0632, "step": 52250 }, { "grad_norm": 0.5292097926139832, "learning_rate": 8.82826708139724e-05, "loss": 0.0386, "step": 52260 }, { "grad_norm": 0.7270067930221558, "learning_rate": 8.827735229085101e-05, "loss": 0.0463, "step": 52270 }, { "grad_norm": 0.4255836009979248, "learning_rate": 8.827203272124119e-05, "loss": 0.0539, "step": 52280 }, { "grad_norm": 0.4655202329158783, "learning_rate": 8.826671210528836e-05, "loss": 0.0694, "step": 52290 }, { "grad_norm": 0.5166445970535278, "learning_rate": 8.826139044313801e-05, "loss": 0.0382, "step": 52300 }, { "grad_norm": 0.8230412602424622, "learning_rate": 8.825606773493561e-05, "loss": 0.048, "step": 52310 }, { "grad_norm": 0.8382673859596252, "learning_rate": 8.82507439808267e-05, "loss": 0.0493, "step": 52320 }, { "grad_norm": 0.5501798987388611, "learning_rate": 8.824541918095681e-05, "loss": 0.044, "step": 52330 }, { "grad_norm": 0.4049500524997711, "learning_rate": 8.824009333547154e-05, "loss": 0.0556, "step": 52340 }, { "grad_norm": 0.4870924949645996, "learning_rate": 8.823476644451646e-05, "loss": 0.0473, "step": 52350 }, { "grad_norm": 0.9624547958374023, "learning_rate": 8.822943850823726e-05, "loss": 0.0562, "step": 52360 }, { "grad_norm": 0.5058255791664124, "learning_rate": 8.822410952677956e-05, "loss": 0.0639, "step": 52370 }, { "grad_norm": 0.6052766442298889, "learning_rate": 8.821877950028908e-05, "loss": 0.052, "step": 52380 }, { "grad_norm": 0.5216665863990784, "learning_rate": 8.821344842891151e-05, "loss": 0.0529, "step": 52390 }, { "grad_norm": 0.48779964447021484, "learning_rate": 8.820811631279262e-05, "loss": 0.0541, "step": 52400 }, { "grad_norm": 0.647711455821991, "learning_rate": 8.820278315207817e-05, "loss": 0.0438, "step": 52410 }, { "grad_norm": 0.5632126927375793, "learning_rate": 8.8197448946914e-05, "loss": 0.067, "step": 52420 }, { "grad_norm": 2.401052236557007, "learning_rate": 8.81921136974459e-05, "loss": 0.0571, "step": 52430 }, { "grad_norm": 0.5884045362472534, "learning_rate": 8.81867774038198e-05, "loss": 0.0571, "step": 52440 }, { "grad_norm": 0.5810978412628174, "learning_rate": 8.818144006618154e-05, "loss": 0.0426, "step": 52450 }, { "grad_norm": 0.6385902762413025, "learning_rate": 8.817610168467704e-05, "loss": 0.0502, "step": 52460 }, { "grad_norm": 0.5318139791488647, "learning_rate": 8.817076225945227e-05, "loss": 0.0462, "step": 52470 }, { "grad_norm": 0.6862472891807556, "learning_rate": 8.816542179065319e-05, "loss": 0.0435, "step": 52480 }, { "grad_norm": 0.47138577699661255, "learning_rate": 8.816008027842582e-05, "loss": 0.037, "step": 52490 }, { "grad_norm": 0.46040454506874084, "learning_rate": 8.815473772291619e-05, "loss": 0.0505, "step": 52500 }, { "grad_norm": 1.244592547416687, "learning_rate": 8.814939412427037e-05, "loss": 0.0621, "step": 52510 }, { "grad_norm": 0.7161184549331665, "learning_rate": 8.814404948263445e-05, "loss": 0.0625, "step": 52520 }, { "grad_norm": 0.6556305885314941, "learning_rate": 8.813870379815454e-05, "loss": 0.056, "step": 52530 }, { "grad_norm": 0.39526528120040894, "learning_rate": 8.813335707097681e-05, "loss": 0.0394, "step": 52540 }, { "grad_norm": 0.7463733553886414, "learning_rate": 8.81280093012474e-05, "loss": 0.0861, "step": 52550 }, { "grad_norm": 0.8668938279151917, "learning_rate": 8.812266048911255e-05, "loss": 0.0455, "step": 52560 }, { "grad_norm": 0.385332852602005, "learning_rate": 8.811731063471849e-05, "loss": 0.0528, "step": 52570 }, { "grad_norm": 0.3934938311576843, "learning_rate": 8.811195973821147e-05, "loss": 0.0541, "step": 52580 }, { "grad_norm": 0.6892812252044678, "learning_rate": 8.81066077997378e-05, "loss": 0.0414, "step": 52590 }, { "grad_norm": 0.55488520860672, "learning_rate": 8.810125481944377e-05, "loss": 0.0554, "step": 52600 }, { "grad_norm": 0.4457516074180603, "learning_rate": 8.809590079747576e-05, "loss": 0.0552, "step": 52610 }, { "grad_norm": 0.6583343744277954, "learning_rate": 8.809054573398015e-05, "loss": 0.0669, "step": 52620 }, { "grad_norm": 0.7500734329223633, "learning_rate": 8.808518962910331e-05, "loss": 0.0486, "step": 52630 }, { "grad_norm": 0.4627821147441864, "learning_rate": 8.807983248299169e-05, "loss": 0.0475, "step": 52640 }, { "grad_norm": 0.5721616148948669, "learning_rate": 8.807447429579177e-05, "loss": 0.0598, "step": 52650 }, { "grad_norm": 0.6265650391578674, "learning_rate": 8.806911506765002e-05, "loss": 0.0485, "step": 52660 }, { "grad_norm": 0.6266689896583557, "learning_rate": 8.806375479871295e-05, "loss": 0.0412, "step": 52670 }, { "grad_norm": 0.7008988857269287, "learning_rate": 8.805839348912715e-05, "loss": 0.0531, "step": 52680 }, { "grad_norm": 0.5168386101722717, "learning_rate": 8.805303113903915e-05, "loss": 0.0465, "step": 52690 }, { "grad_norm": 0.7754213809967041, "learning_rate": 8.804766774859557e-05, "loss": 0.0372, "step": 52700 }, { "grad_norm": 0.6187829971313477, "learning_rate": 8.804230331794306e-05, "loss": 0.0566, "step": 52710 }, { "grad_norm": 0.6592797636985779, "learning_rate": 8.803693784722827e-05, "loss": 0.0561, "step": 52720 }, { "grad_norm": 0.43936648964881897, "learning_rate": 8.803157133659789e-05, "loss": 0.044, "step": 52730 }, { "grad_norm": 0.4613785445690155, "learning_rate": 8.802620378619864e-05, "loss": 0.0481, "step": 52740 }, { "grad_norm": 0.7410581707954407, "learning_rate": 8.802083519617725e-05, "loss": 0.0457, "step": 52750 }, { "grad_norm": 0.6857479810714722, "learning_rate": 8.801546556668052e-05, "loss": 0.0477, "step": 52760 }, { "grad_norm": 1.1104867458343506, "learning_rate": 8.801009489785526e-05, "loss": 0.0645, "step": 52770 }, { "grad_norm": 0.792060136795044, "learning_rate": 8.800472318984825e-05, "loss": 0.0468, "step": 52780 }, { "grad_norm": 0.6460438966751099, "learning_rate": 8.79993504428064e-05, "loss": 0.0637, "step": 52790 }, { "grad_norm": 0.6467125415802002, "learning_rate": 8.79939766568766e-05, "loss": 0.0592, "step": 52800 }, { "grad_norm": 0.6569516658782959, "learning_rate": 8.798860183220575e-05, "loss": 0.0516, "step": 52810 }, { "grad_norm": 0.5304391980171204, "learning_rate": 8.798322596894077e-05, "loss": 0.0565, "step": 52820 }, { "grad_norm": 0.6787130236625671, "learning_rate": 8.797784906722868e-05, "loss": 0.0443, "step": 52830 }, { "grad_norm": 0.9608675837516785, "learning_rate": 8.797247112721647e-05, "loss": 0.0531, "step": 52840 }, { "grad_norm": 0.597247838973999, "learning_rate": 8.796709214905116e-05, "loss": 0.0468, "step": 52850 }, { "grad_norm": 1.1726047992706299, "learning_rate": 8.79617121328798e-05, "loss": 0.0506, "step": 52860 }, { "grad_norm": 0.5221933126449585, "learning_rate": 8.795633107884951e-05, "loss": 0.0526, "step": 52870 }, { "grad_norm": 1.1070889234542847, "learning_rate": 8.795094898710739e-05, "loss": 0.0415, "step": 52880 }, { "grad_norm": 0.7053378224372864, "learning_rate": 8.794556585780056e-05, "loss": 0.0569, "step": 52890 }, { "grad_norm": 0.5089372992515564, "learning_rate": 8.794018169107623e-05, "loss": 0.034, "step": 52900 }, { "grad_norm": 0.8080387115478516, "learning_rate": 8.793479648708159e-05, "loss": 0.0319, "step": 52910 }, { "grad_norm": 0.5573858022689819, "learning_rate": 8.792941024596386e-05, "loss": 0.0562, "step": 52920 }, { "grad_norm": 0.5943167805671692, "learning_rate": 8.79240229678703e-05, "loss": 0.0379, "step": 52930 }, { "grad_norm": 0.5650584697723389, "learning_rate": 8.79186346529482e-05, "loss": 0.0422, "step": 52940 }, { "grad_norm": 0.47055140137672424, "learning_rate": 8.791324530134488e-05, "loss": 0.0349, "step": 52950 }, { "grad_norm": 1.1333500146865845, "learning_rate": 8.790785491320767e-05, "loss": 0.0435, "step": 52960 }, { "grad_norm": 0.5789627432823181, "learning_rate": 8.790246348868396e-05, "loss": 0.0459, "step": 52970 }, { "grad_norm": 0.892648458480835, "learning_rate": 8.789707102792112e-05, "loss": 0.0407, "step": 52980 }, { "grad_norm": 0.37804627418518066, "learning_rate": 8.789167753106661e-05, "loss": 0.0457, "step": 52990 }, { "grad_norm": 0.7688239216804504, "learning_rate": 8.788628299826786e-05, "loss": 0.0434, "step": 53000 }, { "grad_norm": 1.002594232559204, "learning_rate": 8.788088742967238e-05, "loss": 0.0695, "step": 53010 }, { "grad_norm": 0.7322565317153931, "learning_rate": 8.787549082542766e-05, "loss": 0.0528, "step": 53020 }, { "grad_norm": 0.9831551909446716, "learning_rate": 8.787009318568126e-05, "loss": 0.0486, "step": 53030 }, { "grad_norm": 0.5241119861602783, "learning_rate": 8.786469451058073e-05, "loss": 0.0554, "step": 53040 }, { "grad_norm": 0.7829786539077759, "learning_rate": 8.785929480027367e-05, "loss": 0.0578, "step": 53050 }, { "grad_norm": 0.7121362090110779, "learning_rate": 8.785389405490773e-05, "loss": 0.0558, "step": 53060 }, { "grad_norm": 1.3597444295883179, "learning_rate": 8.784849227463054e-05, "loss": 0.0525, "step": 53070 }, { "grad_norm": 0.6201826930046082, "learning_rate": 8.78430894595898e-05, "loss": 0.0503, "step": 53080 }, { "grad_norm": 0.583865225315094, "learning_rate": 8.783768560993317e-05, "loss": 0.0419, "step": 53090 }, { "grad_norm": 0.5202257633209229, "learning_rate": 8.783228072580847e-05, "loss": 0.0463, "step": 53100 }, { "grad_norm": 0.9816011786460876, "learning_rate": 8.782687480736343e-05, "loss": 0.071, "step": 53110 }, { "grad_norm": 0.5905801057815552, "learning_rate": 8.782146785474581e-05, "loss": 0.0581, "step": 53120 }, { "grad_norm": 0.5808166265487671, "learning_rate": 8.78160598681035e-05, "loss": 0.0445, "step": 53130 }, { "grad_norm": 0.7414993643760681, "learning_rate": 8.781065084758431e-05, "loss": 0.0404, "step": 53140 }, { "grad_norm": 0.4966774880886078, "learning_rate": 8.780524079333613e-05, "loss": 0.0446, "step": 53150 }, { "grad_norm": 0.6813881397247314, "learning_rate": 8.779982970550687e-05, "loss": 0.0447, "step": 53160 }, { "grad_norm": 0.3424035608768463, "learning_rate": 8.779441758424447e-05, "loss": 0.053, "step": 53170 }, { "grad_norm": 0.6355531811714172, "learning_rate": 8.778900442969688e-05, "loss": 0.0504, "step": 53180 }, { "grad_norm": 0.6671353578567505, "learning_rate": 8.778359024201212e-05, "loss": 0.0528, "step": 53190 }, { "grad_norm": 0.4644264280796051, "learning_rate": 8.777817502133819e-05, "loss": 0.0486, "step": 53200 }, { "grad_norm": 0.6370773911476135, "learning_rate": 8.777275876782315e-05, "loss": 0.04, "step": 53210 }, { "grad_norm": 0.5480477213859558, "learning_rate": 8.776734148161507e-05, "loss": 0.0556, "step": 53220 }, { "grad_norm": 0.5203929543495178, "learning_rate": 8.776192316286207e-05, "loss": 0.0476, "step": 53230 }, { "grad_norm": 0.568962574005127, "learning_rate": 8.775650381171227e-05, "loss": 0.0669, "step": 53240 }, { "grad_norm": 0.4658046066761017, "learning_rate": 8.775108342831385e-05, "loss": 0.0546, "step": 53250 }, { "grad_norm": 1.3973407745361328, "learning_rate": 8.774566201281499e-05, "loss": 0.0555, "step": 53260 }, { "grad_norm": 0.45811647176742554, "learning_rate": 8.774023956536391e-05, "loss": 0.0426, "step": 53270 }, { "grad_norm": 1.0410847663879395, "learning_rate": 8.773481608610886e-05, "loss": 0.0557, "step": 53280 }, { "grad_norm": 0.7492621541023254, "learning_rate": 8.77293915751981e-05, "loss": 0.0452, "step": 53290 }, { "grad_norm": 1.0768641233444214, "learning_rate": 8.772396603277997e-05, "loss": 0.0361, "step": 53300 }, { "grad_norm": 0.586916446685791, "learning_rate": 8.771853945900277e-05, "loss": 0.0496, "step": 53310 }, { "grad_norm": 0.6140584945678711, "learning_rate": 8.771311185401487e-05, "loss": 0.079, "step": 53320 }, { "grad_norm": 0.8946307301521301, "learning_rate": 8.770768321796466e-05, "loss": 0.0398, "step": 53330 }, { "grad_norm": 0.5741287469863892, "learning_rate": 8.770225355100054e-05, "loss": 0.0503, "step": 53340 }, { "grad_norm": 0.5629469156265259, "learning_rate": 8.7696822853271e-05, "loss": 0.0495, "step": 53350 }, { "grad_norm": 0.7390425205230713, "learning_rate": 8.769139112492447e-05, "loss": 0.0526, "step": 53360 }, { "grad_norm": 0.7754611968994141, "learning_rate": 8.768595836610948e-05, "loss": 0.0482, "step": 53370 }, { "grad_norm": 0.8546565175056458, "learning_rate": 8.768052457697454e-05, "loss": 0.0614, "step": 53380 }, { "grad_norm": 0.7256115674972534, "learning_rate": 8.76750897576682e-05, "loss": 0.0723, "step": 53390 }, { "grad_norm": 0.8796412944793701, "learning_rate": 8.766965390833907e-05, "loss": 0.0668, "step": 53400 }, { "grad_norm": 1.0640742778778076, "learning_rate": 8.766421702913576e-05, "loss": 0.0629, "step": 53410 }, { "grad_norm": 0.8476904034614563, "learning_rate": 8.76587791202069e-05, "loss": 0.0557, "step": 53420 }, { "grad_norm": 1.610935091972351, "learning_rate": 8.765334018170118e-05, "loss": 0.0435, "step": 53430 }, { "grad_norm": 0.4634873569011688, "learning_rate": 8.764790021376728e-05, "loss": 0.0416, "step": 53440 }, { "grad_norm": 0.8236021399497986, "learning_rate": 8.764245921655391e-05, "loss": 0.0414, "step": 53450 }, { "grad_norm": 1.366607427597046, "learning_rate": 8.763701719020987e-05, "loss": 0.0452, "step": 53460 }, { "grad_norm": 0.41772544384002686, "learning_rate": 8.763157413488392e-05, "loss": 0.0515, "step": 53470 }, { "grad_norm": 0.47846150398254395, "learning_rate": 8.762613005072486e-05, "loss": 0.0536, "step": 53480 }, { "grad_norm": 0.4463541805744171, "learning_rate": 8.762068493788154e-05, "loss": 0.0484, "step": 53490 }, { "grad_norm": 0.7129429578781128, "learning_rate": 8.761523879650283e-05, "loss": 0.0588, "step": 53500 }, { "grad_norm": 0.6331129670143127, "learning_rate": 8.760979162673763e-05, "loss": 0.0469, "step": 53510 }, { "grad_norm": 0.5826104283332825, "learning_rate": 8.760434342873484e-05, "loss": 0.0599, "step": 53520 }, { "grad_norm": 0.5615106225013733, "learning_rate": 8.759889420264345e-05, "loss": 0.0442, "step": 53530 }, { "grad_norm": 0.6736682057380676, "learning_rate": 8.75934439486124e-05, "loss": 0.052, "step": 53540 }, { "grad_norm": 0.7066115140914917, "learning_rate": 8.758799266679071e-05, "loss": 0.0549, "step": 53550 }, { "grad_norm": 0.515363872051239, "learning_rate": 8.758254035732744e-05, "loss": 0.046, "step": 53560 }, { "grad_norm": 0.4525022804737091, "learning_rate": 8.757708702037161e-05, "loss": 0.0606, "step": 53570 }, { "grad_norm": 0.43306291103363037, "learning_rate": 8.757163265607237e-05, "loss": 0.0406, "step": 53580 }, { "grad_norm": 1.3523305654525757, "learning_rate": 8.756617726457878e-05, "loss": 0.0313, "step": 53590 }, { "grad_norm": 0.6130388975143433, "learning_rate": 8.756072084604004e-05, "loss": 0.0411, "step": 53600 }, { "grad_norm": 0.5319221019744873, "learning_rate": 8.75552634006053e-05, "loss": 0.0464, "step": 53610 }, { "grad_norm": 0.4323735237121582, "learning_rate": 8.754980492842374e-05, "loss": 0.0391, "step": 53620 }, { "grad_norm": 0.33805862069129944, "learning_rate": 8.754434542964465e-05, "loss": 0.0396, "step": 53630 }, { "grad_norm": 0.9590712785720825, "learning_rate": 8.753888490441725e-05, "loss": 0.0479, "step": 53640 }, { "grad_norm": 0.5266739130020142, "learning_rate": 8.753342335289085e-05, "loss": 0.0412, "step": 53650 }, { "grad_norm": 0.5607447028160095, "learning_rate": 8.752796077521475e-05, "loss": 0.0493, "step": 53660 }, { "grad_norm": 0.5386539697647095, "learning_rate": 8.752249717153829e-05, "loss": 0.0519, "step": 53670 }, { "grad_norm": 0.6375105381011963, "learning_rate": 8.751703254201087e-05, "loss": 0.0551, "step": 53680 }, { "grad_norm": 1.197211503982544, "learning_rate": 8.751156688678186e-05, "loss": 0.0497, "step": 53690 }, { "grad_norm": 1.056283712387085, "learning_rate": 8.750610020600072e-05, "loss": 0.0535, "step": 53700 }, { "grad_norm": 0.4740922749042511, "learning_rate": 8.750063249981687e-05, "loss": 0.0441, "step": 53710 }, { "grad_norm": 0.3705071806907654, "learning_rate": 8.749516376837983e-05, "loss": 0.0318, "step": 53720 }, { "grad_norm": 0.3683938682079315, "learning_rate": 8.74896940118391e-05, "loss": 0.0396, "step": 53730 }, { "grad_norm": 0.6323266625404358, "learning_rate": 8.748422323034421e-05, "loss": 0.0461, "step": 53740 }, { "grad_norm": 0.5401157140731812, "learning_rate": 8.747875142404475e-05, "loss": 0.0351, "step": 53750 }, { "grad_norm": 0.6954439282417297, "learning_rate": 8.74732785930903e-05, "loss": 0.0486, "step": 53760 }, { "grad_norm": 0.4278619885444641, "learning_rate": 8.746780473763049e-05, "loss": 0.0433, "step": 53770 }, { "grad_norm": 0.38382184505462646, "learning_rate": 8.746232985781497e-05, "loss": 0.0344, "step": 53780 }, { "grad_norm": 0.4570102095603943, "learning_rate": 8.745685395379345e-05, "loss": 0.0453, "step": 53790 }, { "grad_norm": 0.5346582531929016, "learning_rate": 8.745137702571558e-05, "loss": 0.0328, "step": 53800 }, { "grad_norm": 0.6370213031768799, "learning_rate": 8.744589907373115e-05, "loss": 0.051, "step": 53810 }, { "grad_norm": 0.3243315517902374, "learning_rate": 8.744042009798991e-05, "loss": 0.0382, "step": 53820 }, { "grad_norm": 0.4264233410358429, "learning_rate": 8.743494009864165e-05, "loss": 0.0429, "step": 53830 }, { "grad_norm": 1.06130051612854, "learning_rate": 8.742945907583619e-05, "loss": 0.0504, "step": 53840 }, { "grad_norm": 0.738857090473175, "learning_rate": 8.742397702972338e-05, "loss": 0.0394, "step": 53850 }, { "grad_norm": 0.6191310882568359, "learning_rate": 8.74184939604531e-05, "loss": 0.0527, "step": 53860 }, { "grad_norm": 0.8838266134262085, "learning_rate": 8.741300986817524e-05, "loss": 0.0557, "step": 53870 }, { "grad_norm": 1.0697320699691772, "learning_rate": 8.740752475303975e-05, "loss": 0.0667, "step": 53880 }, { "grad_norm": 0.5735713243484497, "learning_rate": 8.74020386151966e-05, "loss": 0.0581, "step": 53890 }, { "grad_norm": 0.37882494926452637, "learning_rate": 8.739655145479576e-05, "loss": 0.0432, "step": 53900 }, { "grad_norm": 0.6022922396659851, "learning_rate": 8.739106327198724e-05, "loss": 0.0491, "step": 53910 }, { "grad_norm": 0.5241372585296631, "learning_rate": 8.73855740669211e-05, "loss": 0.0501, "step": 53920 }, { "grad_norm": 0.5877297520637512, "learning_rate": 8.738008383974742e-05, "loss": 0.0622, "step": 53930 }, { "grad_norm": 0.48707762360572815, "learning_rate": 8.737459259061629e-05, "loss": 0.0498, "step": 53940 }, { "grad_norm": 0.8631933331489563, "learning_rate": 8.736910031967784e-05, "loss": 0.0417, "step": 53950 }, { "grad_norm": 1.0686366558074951, "learning_rate": 8.736360702708222e-05, "loss": 0.0388, "step": 53960 }, { "grad_norm": 0.3413383960723877, "learning_rate": 8.735811271297963e-05, "loss": 0.0463, "step": 53970 }, { "grad_norm": 0.6031984686851501, "learning_rate": 8.735261737752025e-05, "loss": 0.0443, "step": 53980 }, { "grad_norm": 0.5116688013076782, "learning_rate": 8.734712102085436e-05, "loss": 0.0371, "step": 53990 }, { "grad_norm": 0.5744500756263733, "learning_rate": 8.734162364313222e-05, "loss": 0.0429, "step": 54000 }, { "grad_norm": 0.6643089056015015, "learning_rate": 8.733612524450412e-05, "loss": 0.0606, "step": 54010 }, { "grad_norm": 0.5270866751670837, "learning_rate": 8.733062582512037e-05, "loss": 0.0396, "step": 54020 }, { "grad_norm": 0.5384116768836975, "learning_rate": 8.732512538513134e-05, "loss": 0.0489, "step": 54030 }, { "grad_norm": 0.49401503801345825, "learning_rate": 8.73196239246874e-05, "loss": 0.0551, "step": 54040 }, { "grad_norm": 0.6910138130187988, "learning_rate": 8.731412144393896e-05, "loss": 0.0415, "step": 54050 }, { "grad_norm": 0.6479716897010803, "learning_rate": 8.730861794303646e-05, "loss": 0.0328, "step": 54060 }, { "grad_norm": 0.8224765658378601, "learning_rate": 8.730311342213036e-05, "loss": 0.0445, "step": 54070 }, { "grad_norm": 0.48549512028694153, "learning_rate": 8.729760788137116e-05, "loss": 0.0562, "step": 54080 }, { "grad_norm": 2.791243314743042, "learning_rate": 8.729210132090938e-05, "loss": 0.0356, "step": 54090 }, { "grad_norm": 0.8380087614059448, "learning_rate": 8.728659374089554e-05, "loss": 0.0478, "step": 54100 }, { "grad_norm": 0.8707197308540344, "learning_rate": 8.728108514148023e-05, "loss": 0.0555, "step": 54110 }, { "grad_norm": 0.6427738070487976, "learning_rate": 8.727557552281408e-05, "loss": 0.0351, "step": 54120 }, { "grad_norm": 1.1059776544570923, "learning_rate": 8.72700648850477e-05, "loss": 0.0421, "step": 54130 }, { "grad_norm": 0.46960270404815674, "learning_rate": 8.726455322833174e-05, "loss": 0.0416, "step": 54140 }, { "grad_norm": 0.8313402533531189, "learning_rate": 8.725904055281688e-05, "loss": 0.0486, "step": 54150 }, { "grad_norm": 0.6260781288146973, "learning_rate": 8.725352685865387e-05, "loss": 0.055, "step": 54160 }, { "grad_norm": 0.7133686542510986, "learning_rate": 8.724801214599343e-05, "loss": 0.046, "step": 54170 }, { "grad_norm": 0.6974632143974304, "learning_rate": 8.724249641498632e-05, "loss": 0.0485, "step": 54180 }, { "grad_norm": 0.6677164435386658, "learning_rate": 8.723697966578334e-05, "loss": 0.0436, "step": 54190 }, { "grad_norm": 0.6302991509437561, "learning_rate": 8.723146189853535e-05, "loss": 0.0425, "step": 54200 }, { "grad_norm": 0.49773192405700684, "learning_rate": 8.722594311339317e-05, "loss": 0.0446, "step": 54210 }, { "grad_norm": 0.7285786271095276, "learning_rate": 8.722042331050768e-05, "loss": 0.0507, "step": 54220 }, { "grad_norm": 0.874164879322052, "learning_rate": 8.721490249002982e-05, "loss": 0.0463, "step": 54230 }, { "grad_norm": 0.6983322501182556, "learning_rate": 8.72093806521105e-05, "loss": 0.0532, "step": 54240 }, { "grad_norm": 0.6875331997871399, "learning_rate": 8.720385779690067e-05, "loss": 0.0353, "step": 54250 }, { "grad_norm": 0.403373122215271, "learning_rate": 8.719833392455137e-05, "loss": 0.0447, "step": 54260 }, { "grad_norm": 1.576151967048645, "learning_rate": 8.719280903521358e-05, "loss": 0.0612, "step": 54270 }, { "grad_norm": 0.8409641981124878, "learning_rate": 8.718728312903838e-05, "loss": 0.0452, "step": 54280 }, { "grad_norm": 0.5236402153968811, "learning_rate": 8.718175620617681e-05, "loss": 0.0443, "step": 54290 }, { "grad_norm": 0.7013545036315918, "learning_rate": 8.717622826678001e-05, "loss": 0.044, "step": 54300 }, { "grad_norm": 0.588485598564148, "learning_rate": 8.717069931099908e-05, "loss": 0.0425, "step": 54310 }, { "grad_norm": 0.47814032435417175, "learning_rate": 8.716516933898521e-05, "loss": 0.0428, "step": 54320 }, { "grad_norm": 1.0834574699401855, "learning_rate": 8.715963835088959e-05, "loss": 0.0497, "step": 54330 }, { "grad_norm": 0.4890303909778595, "learning_rate": 8.715410634686339e-05, "loss": 0.0546, "step": 54340 }, { "grad_norm": 0.5129042863845825, "learning_rate": 8.714857332705788e-05, "loss": 0.0387, "step": 54350 }, { "grad_norm": 0.5489365458488464, "learning_rate": 8.714303929162433e-05, "loss": 0.0589, "step": 54360 }, { "grad_norm": 0.6299996972084045, "learning_rate": 8.713750424071405e-05, "loss": 0.0457, "step": 54370 }, { "grad_norm": 0.7435247898101807, "learning_rate": 8.713196817447835e-05, "loss": 0.0511, "step": 54380 }, { "grad_norm": 1.0509029626846313, "learning_rate": 8.71264310930686e-05, "loss": 0.0623, "step": 54390 }, { "grad_norm": 1.3746931552886963, "learning_rate": 8.712089299663617e-05, "loss": 0.0571, "step": 54400 }, { "grad_norm": 0.6249468326568604, "learning_rate": 8.711535388533246e-05, "loss": 0.0613, "step": 54410 }, { "grad_norm": 0.5365161299705505, "learning_rate": 8.710981375930893e-05, "loss": 0.0458, "step": 54420 }, { "grad_norm": 0.7193859219551086, "learning_rate": 8.710427261871703e-05, "loss": 0.0367, "step": 54430 }, { "grad_norm": 0.3707351088523865, "learning_rate": 8.709873046370825e-05, "loss": 0.0397, "step": 54440 }, { "grad_norm": 0.7364063858985901, "learning_rate": 8.709318729443411e-05, "loss": 0.0349, "step": 54450 }, { "grad_norm": 0.38937073945999146, "learning_rate": 8.708764311104618e-05, "loss": 0.0452, "step": 54460 }, { "grad_norm": 0.3559318482875824, "learning_rate": 8.708209791369601e-05, "loss": 0.0385, "step": 54470 }, { "grad_norm": 0.33205169439315796, "learning_rate": 8.707655170253524e-05, "loss": 0.036, "step": 54480 }, { "grad_norm": 0.5917271375656128, "learning_rate": 8.707100447771545e-05, "loss": 0.0588, "step": 54490 }, { "grad_norm": 1.5855584144592285, "learning_rate": 8.706545623938832e-05, "loss": 0.0594, "step": 54500 }, { "grad_norm": 0.5393316149711609, "learning_rate": 8.705990698770556e-05, "loss": 0.0319, "step": 54510 }, { "grad_norm": 0.47472697496414185, "learning_rate": 8.705435672281886e-05, "loss": 0.0445, "step": 54520 }, { "grad_norm": 1.0867468118667603, "learning_rate": 8.704880544487995e-05, "loss": 0.0553, "step": 54530 }, { "grad_norm": 0.5188086032867432, "learning_rate": 8.704325315404063e-05, "loss": 0.0384, "step": 54540 }, { "grad_norm": 0.329187273979187, "learning_rate": 8.70376998504527e-05, "loss": 0.0396, "step": 54550 }, { "grad_norm": 0.5666934251785278, "learning_rate": 8.703214553426795e-05, "loss": 0.0478, "step": 54560 }, { "grad_norm": 0.5112659931182861, "learning_rate": 8.702659020563826e-05, "loss": 0.0412, "step": 54570 }, { "grad_norm": 0.49584606289863586, "learning_rate": 8.702103386471551e-05, "loss": 0.0359, "step": 54580 }, { "grad_norm": 0.3473212718963623, "learning_rate": 8.70154765116516e-05, "loss": 0.0394, "step": 54590 }, { "grad_norm": 0.3065521717071533, "learning_rate": 8.700991814659847e-05, "loss": 0.0359, "step": 54600 }, { "grad_norm": 0.48899826407432556, "learning_rate": 8.700435876970807e-05, "loss": 0.0386, "step": 54610 }, { "grad_norm": 0.6918405294418335, "learning_rate": 8.69987983811324e-05, "loss": 0.0487, "step": 54620 }, { "grad_norm": 0.3667159676551819, "learning_rate": 8.699323698102351e-05, "loss": 0.0434, "step": 54630 }, { "grad_norm": 0.8041987419128418, "learning_rate": 8.69876745695334e-05, "loss": 0.0388, "step": 54640 }, { "grad_norm": 0.4322129189968109, "learning_rate": 8.698211114681417e-05, "loss": 0.0317, "step": 54650 }, { "grad_norm": 0.33523282408714294, "learning_rate": 8.697654671301789e-05, "loss": 0.0331, "step": 54660 }, { "grad_norm": 0.80292809009552, "learning_rate": 8.697098126829675e-05, "loss": 0.0416, "step": 54670 }, { "grad_norm": 0.6595636010169983, "learning_rate": 8.696541481280285e-05, "loss": 0.0433, "step": 54680 }, { "grad_norm": 0.6840540766716003, "learning_rate": 8.69598473466884e-05, "loss": 0.063, "step": 54690 }, { "grad_norm": 0.6107103228569031, "learning_rate": 8.69542788701056e-05, "loss": 0.0381, "step": 54700 }, { "grad_norm": 2.2353999614715576, "learning_rate": 8.694870938320671e-05, "loss": 0.0578, "step": 54710 }, { "grad_norm": 0.8273041248321533, "learning_rate": 8.6943138886144e-05, "loss": 0.0421, "step": 54720 }, { "grad_norm": 0.40460842847824097, "learning_rate": 8.693756737906972e-05, "loss": 0.0471, "step": 54730 }, { "grad_norm": 0.7541987299919128, "learning_rate": 8.693199486213626e-05, "loss": 0.0364, "step": 54740 }, { "grad_norm": 0.5547622442245483, "learning_rate": 8.692642133549591e-05, "loss": 0.0375, "step": 54750 }, { "grad_norm": 1.2510095834732056, "learning_rate": 8.692084679930108e-05, "loss": 0.0623, "step": 54760 }, { "grad_norm": 0.5767205953598022, "learning_rate": 8.691527125370417e-05, "loss": 0.042, "step": 54770 }, { "grad_norm": 0.7436186671257019, "learning_rate": 8.690969469885761e-05, "loss": 0.0437, "step": 54780 }, { "grad_norm": 0.6261321902275085, "learning_rate": 8.690411713491386e-05, "loss": 0.0429, "step": 54790 }, { "grad_norm": 0.5177961587905884, "learning_rate": 8.689853856202541e-05, "loss": 0.0371, "step": 54800 }, { "grad_norm": 0.7626355886459351, "learning_rate": 8.68929589803448e-05, "loss": 0.0648, "step": 54810 }, { "grad_norm": 0.4720478951931, "learning_rate": 8.688737839002452e-05, "loss": 0.0577, "step": 54820 }, { "grad_norm": 0.5833441615104675, "learning_rate": 8.688179679121719e-05, "loss": 0.0436, "step": 54830 }, { "grad_norm": 0.49709242582321167, "learning_rate": 8.68762141840754e-05, "loss": 0.041, "step": 54840 }, { "grad_norm": 0.6761584281921387, "learning_rate": 8.687063056875175e-05, "loss": 0.0436, "step": 54850 }, { "grad_norm": 0.7964344620704651, "learning_rate": 8.686504594539892e-05, "loss": 0.0626, "step": 54860 }, { "grad_norm": 0.6721751093864441, "learning_rate": 8.68594603141696e-05, "loss": 0.0577, "step": 54870 }, { "grad_norm": 0.47937119007110596, "learning_rate": 8.685387367521646e-05, "loss": 0.0686, "step": 54880 }, { "grad_norm": 0.5798193216323853, "learning_rate": 8.684828602869226e-05, "loss": 0.0581, "step": 54890 }, { "grad_norm": 0.5495821237564087, "learning_rate": 8.684269737474976e-05, "loss": 0.0568, "step": 54900 }, { "grad_norm": 0.8164705634117126, "learning_rate": 8.683710771354178e-05, "loss": 0.0571, "step": 54910 }, { "grad_norm": 0.9755966067314148, "learning_rate": 8.683151704522109e-05, "loss": 0.0543, "step": 54920 }, { "grad_norm": 1.285697102546692, "learning_rate": 8.682592536994057e-05, "loss": 0.059, "step": 54930 }, { "grad_norm": 1.2776838541030884, "learning_rate": 8.682033268785308e-05, "loss": 0.084, "step": 54940 }, { "grad_norm": 0.6473231315612793, "learning_rate": 8.681473899911154e-05, "loss": 0.0581, "step": 54950 }, { "grad_norm": 1.0578978061676025, "learning_rate": 8.680914430386885e-05, "loss": 0.0507, "step": 54960 }, { "grad_norm": 0.9411389231681824, "learning_rate": 8.6803548602278e-05, "loss": 0.0415, "step": 54970 }, { "grad_norm": 0.43845611810684204, "learning_rate": 8.679795189449193e-05, "loss": 0.0534, "step": 54980 }, { "grad_norm": 0.7107380628585815, "learning_rate": 8.679235418066371e-05, "loss": 0.0449, "step": 54990 }, { "grad_norm": 0.5603547692298889, "learning_rate": 8.678675546094633e-05, "loss": 0.0778, "step": 55000 }, { "grad_norm": 0.44312405586242676, "learning_rate": 8.678115573549288e-05, "loss": 0.0611, "step": 55010 }, { "grad_norm": 0.9895073175430298, "learning_rate": 8.677555500445645e-05, "loss": 0.0604, "step": 55020 }, { "grad_norm": 0.5000806450843811, "learning_rate": 8.676995326799017e-05, "loss": 0.043, "step": 55030 }, { "grad_norm": 0.9455545544624329, "learning_rate": 8.676435052624719e-05, "loss": 0.081, "step": 55040 }, { "grad_norm": 0.6232199668884277, "learning_rate": 8.675874677938066e-05, "loss": 0.0631, "step": 55050 }, { "grad_norm": 0.6390416622161865, "learning_rate": 8.675314202754379e-05, "loss": 0.059, "step": 55060 }, { "grad_norm": 0.8766713738441467, "learning_rate": 8.674753627088985e-05, "loss": 0.0532, "step": 55070 }, { "grad_norm": 0.5858731269836426, "learning_rate": 8.674192950957207e-05, "loss": 0.0454, "step": 55080 }, { "grad_norm": 0.6106569170951843, "learning_rate": 8.673632174374373e-05, "loss": 0.0449, "step": 55090 }, { "grad_norm": 0.981959879398346, "learning_rate": 8.673071297355817e-05, "loss": 0.0681, "step": 55100 }, { "grad_norm": 0.6042239665985107, "learning_rate": 8.67251031991687e-05, "loss": 0.0478, "step": 55110 }, { "grad_norm": 0.9080408811569214, "learning_rate": 8.671949242072873e-05, "loss": 0.0551, "step": 55120 }, { "grad_norm": 0.7516672015190125, "learning_rate": 8.67138806383916e-05, "loss": 0.0545, "step": 55130 }, { "grad_norm": 0.8431469798088074, "learning_rate": 8.670826785231078e-05, "loss": 0.0574, "step": 55140 }, { "grad_norm": 1.5016013383865356, "learning_rate": 8.670265406263971e-05, "loss": 0.0614, "step": 55150 }, { "grad_norm": 0.6802201271057129, "learning_rate": 8.669703926953187e-05, "loss": 0.0552, "step": 55160 }, { "grad_norm": 0.7485611438751221, "learning_rate": 8.669142347314074e-05, "loss": 0.0513, "step": 55170 }, { "grad_norm": 1.6146494150161743, "learning_rate": 8.668580667361991e-05, "loss": 0.0559, "step": 55180 }, { "grad_norm": 0.7356903553009033, "learning_rate": 8.668018887112287e-05, "loss": 0.047, "step": 55190 }, { "grad_norm": 0.5895802974700928, "learning_rate": 8.667457006580328e-05, "loss": 0.0581, "step": 55200 }, { "grad_norm": 0.6864249110221863, "learning_rate": 8.666895025781471e-05, "loss": 0.0621, "step": 55210 }, { "grad_norm": 0.7694905996322632, "learning_rate": 8.666332944731083e-05, "loss": 0.0616, "step": 55220 }, { "grad_norm": 0.9557541012763977, "learning_rate": 8.665770763444527e-05, "loss": 0.0495, "step": 55230 }, { "grad_norm": 0.7095834612846375, "learning_rate": 8.665208481937176e-05, "loss": 0.0517, "step": 55240 }, { "grad_norm": 0.9152218699455261, "learning_rate": 8.664646100224404e-05, "loss": 0.0395, "step": 55250 }, { "grad_norm": 0.46522676944732666, "learning_rate": 8.664083618321583e-05, "loss": 0.0478, "step": 55260 }, { "grad_norm": 0.44562235474586487, "learning_rate": 8.663521036244093e-05, "loss": 0.0414, "step": 55270 }, { "grad_norm": 0.6401655077934265, "learning_rate": 8.662958354007313e-05, "loss": 0.0463, "step": 55280 }, { "grad_norm": 0.9318197965621948, "learning_rate": 8.662395571626629e-05, "loss": 0.0532, "step": 55290 }, { "grad_norm": 0.7490055561065674, "learning_rate": 8.661832689117426e-05, "loss": 0.079, "step": 55300 }, { "grad_norm": 0.4575341045856476, "learning_rate": 8.661269706495092e-05, "loss": 0.0367, "step": 55310 }, { "grad_norm": 0.46579378843307495, "learning_rate": 8.660706623775021e-05, "loss": 0.074, "step": 55320 }, { "grad_norm": 1.4328441619873047, "learning_rate": 8.660143440972604e-05, "loss": 0.0624, "step": 55330 }, { "grad_norm": 0.46471554040908813, "learning_rate": 8.659580158103242e-05, "loss": 0.0453, "step": 55340 }, { "grad_norm": 0.5592676997184753, "learning_rate": 8.659016775182333e-05, "loss": 0.0474, "step": 55350 }, { "grad_norm": 0.9156274795532227, "learning_rate": 8.658453292225282e-05, "loss": 0.0621, "step": 55360 }, { "grad_norm": 0.5555698275566101, "learning_rate": 8.657889709247489e-05, "loss": 0.0401, "step": 55370 }, { "grad_norm": 0.5531944632530212, "learning_rate": 8.657326026264367e-05, "loss": 0.0516, "step": 55380 }, { "grad_norm": 0.6125370264053345, "learning_rate": 8.656762243291325e-05, "loss": 0.0491, "step": 55390 }, { "grad_norm": 0.5246103405952454, "learning_rate": 8.656198360343778e-05, "loss": 0.052, "step": 55400 }, { "grad_norm": 0.5644418001174927, "learning_rate": 8.655634377437141e-05, "loss": 0.0553, "step": 55410 }, { "grad_norm": 0.6314824819564819, "learning_rate": 8.655070294586833e-05, "loss": 0.0411, "step": 55420 }, { "grad_norm": 0.7305594682693481, "learning_rate": 8.654506111808277e-05, "loss": 0.0409, "step": 55430 }, { "grad_norm": 0.37023115158081055, "learning_rate": 8.653941829116896e-05, "loss": 0.0462, "step": 55440 }, { "grad_norm": 0.8608653545379639, "learning_rate": 8.653377446528119e-05, "loss": 0.0472, "step": 55450 }, { "grad_norm": 0.604619026184082, "learning_rate": 8.652812964057376e-05, "loss": 0.0398, "step": 55460 }, { "grad_norm": 0.814802885055542, "learning_rate": 8.652248381720097e-05, "loss": 0.0479, "step": 55470 }, { "grad_norm": 0.8603594303131104, "learning_rate": 8.65168369953172e-05, "loss": 0.0496, "step": 55480 }, { "grad_norm": 1.7347509860992432, "learning_rate": 8.651118917507681e-05, "loss": 0.063, "step": 55490 }, { "grad_norm": 0.5781630873680115, "learning_rate": 8.650554035663425e-05, "loss": 0.0453, "step": 55500 }, { "grad_norm": 0.563738226890564, "learning_rate": 8.649989054014392e-05, "loss": 0.0433, "step": 55510 }, { "grad_norm": 0.6778936982154846, "learning_rate": 8.64942397257603e-05, "loss": 0.0499, "step": 55520 }, { "grad_norm": 0.7013758420944214, "learning_rate": 8.648858791363786e-05, "loss": 0.0577, "step": 55530 }, { "grad_norm": 1.1875078678131104, "learning_rate": 8.648293510393115e-05, "loss": 0.0525, "step": 55540 }, { "grad_norm": 0.5712730288505554, "learning_rate": 8.647728129679469e-05, "loss": 0.0492, "step": 55550 }, { "grad_norm": 0.8430174589157104, "learning_rate": 8.647162649238307e-05, "loss": 0.0566, "step": 55560 }, { "grad_norm": 0.5996316075325012, "learning_rate": 8.64659706908509e-05, "loss": 0.0443, "step": 55570 }, { "grad_norm": 0.9313849806785583, "learning_rate": 8.646031389235277e-05, "loss": 0.0648, "step": 55580 }, { "grad_norm": 0.9401659369468689, "learning_rate": 8.645465609704335e-05, "loss": 0.0463, "step": 55590 }, { "grad_norm": 0.6044143438339233, "learning_rate": 8.644899730507734e-05, "loss": 0.048, "step": 55600 }, { "grad_norm": 0.41551801562309265, "learning_rate": 8.644333751660944e-05, "loss": 0.0482, "step": 55610 }, { "grad_norm": 0.6764141321182251, "learning_rate": 8.643767673179437e-05, "loss": 0.0664, "step": 55620 }, { "grad_norm": 0.5822838544845581, "learning_rate": 8.643201495078692e-05, "loss": 0.0371, "step": 55630 }, { "grad_norm": 0.4102145731449127, "learning_rate": 8.642635217374186e-05, "loss": 0.0333, "step": 55640 }, { "grad_norm": 0.659755527973175, "learning_rate": 8.6420688400814e-05, "loss": 0.0418, "step": 55650 }, { "grad_norm": 0.6648732423782349, "learning_rate": 8.641502363215823e-05, "loss": 0.039, "step": 55660 }, { "grad_norm": 0.8265186548233032, "learning_rate": 8.64093578679294e-05, "loss": 0.0481, "step": 55670 }, { "grad_norm": 0.47425681352615356, "learning_rate": 8.640369110828238e-05, "loss": 0.0514, "step": 55680 }, { "grad_norm": 0.8062150478363037, "learning_rate": 8.639802335337214e-05, "loss": 0.0453, "step": 55690 }, { "grad_norm": 0.5435823202133179, "learning_rate": 8.63923546033536e-05, "loss": 0.0491, "step": 55700 }, { "grad_norm": 0.4855024814605713, "learning_rate": 8.638668485838177e-05, "loss": 0.0343, "step": 55710 }, { "grad_norm": 0.8401912450790405, "learning_rate": 8.638101411861163e-05, "loss": 0.0534, "step": 55720 }, { "grad_norm": 0.48382556438446045, "learning_rate": 8.637534238419825e-05, "loss": 0.037, "step": 55730 }, { "grad_norm": 0.54803466796875, "learning_rate": 8.636966965529665e-05, "loss": 0.0402, "step": 55740 }, { "grad_norm": 0.660481870174408, "learning_rate": 8.636399593206197e-05, "loss": 0.0475, "step": 55750 }, { "grad_norm": 0.4622255861759186, "learning_rate": 8.63583212146493e-05, "loss": 0.0761, "step": 55760 }, { "grad_norm": 0.48377105593681335, "learning_rate": 8.635264550321378e-05, "loss": 0.0489, "step": 55770 }, { "grad_norm": 0.5606991052627563, "learning_rate": 8.634696879791058e-05, "loss": 0.0481, "step": 55780 }, { "grad_norm": 0.4584466516971588, "learning_rate": 8.634129109889493e-05, "loss": 0.0488, "step": 55790 }, { "grad_norm": 0.32307642698287964, "learning_rate": 8.633561240632202e-05, "loss": 0.0536, "step": 55800 }, { "grad_norm": 0.4786387085914612, "learning_rate": 8.63299327203471e-05, "loss": 0.0459, "step": 55810 }, { "grad_norm": 0.5964422821998596, "learning_rate": 8.632425204112551e-05, "loss": 0.0446, "step": 55820 }, { "grad_norm": 0.5475896596908569, "learning_rate": 8.631857036881248e-05, "loss": 0.062, "step": 55830 }, { "grad_norm": 1.7685132026672363, "learning_rate": 8.631288770356339e-05, "loss": 0.0597, "step": 55840 }, { "grad_norm": 0.5756087899208069, "learning_rate": 8.63072040455336e-05, "loss": 0.0515, "step": 55850 }, { "grad_norm": 0.3341887295246124, "learning_rate": 8.630151939487849e-05, "loss": 0.0586, "step": 55860 }, { "grad_norm": 0.5931572318077087, "learning_rate": 8.629583375175347e-05, "loss": 0.0428, "step": 55870 }, { "grad_norm": 0.5166654586791992, "learning_rate": 8.6290147116314e-05, "loss": 0.0489, "step": 55880 }, { "grad_norm": 0.68031245470047, "learning_rate": 8.628445948871553e-05, "loss": 0.038, "step": 55890 }, { "grad_norm": 2.7212131023406982, "learning_rate": 8.627877086911358e-05, "loss": 0.0558, "step": 55900 }, { "grad_norm": 0.5671154260635376, "learning_rate": 8.627308125766366e-05, "loss": 0.0411, "step": 55910 }, { "grad_norm": 0.7412532567977905, "learning_rate": 8.626739065452132e-05, "loss": 0.0783, "step": 55920 }, { "grad_norm": 0.7070919871330261, "learning_rate": 8.626169905984216e-05, "loss": 0.0458, "step": 55930 }, { "grad_norm": 0.5099592804908752, "learning_rate": 8.625600647378175e-05, "loss": 0.038, "step": 55940 }, { "grad_norm": 0.5580710768699646, "learning_rate": 8.625031289649576e-05, "loss": 0.0426, "step": 55950 }, { "grad_norm": 0.5106713175773621, "learning_rate": 8.624461832813983e-05, "loss": 0.0427, "step": 55960 }, { "grad_norm": 0.6707406044006348, "learning_rate": 8.623892276886964e-05, "loss": 0.0487, "step": 55970 }, { "grad_norm": 0.564365565776825, "learning_rate": 8.623322621884094e-05, "loss": 0.0411, "step": 55980 }, { "grad_norm": 0.9904279112815857, "learning_rate": 8.622752867820942e-05, "loss": 0.0571, "step": 55990 }, { "grad_norm": 0.5110026001930237, "learning_rate": 8.622183014713089e-05, "loss": 0.0449, "step": 56000 }, { "grad_norm": 0.4497823417186737, "learning_rate": 8.621613062576113e-05, "loss": 0.0465, "step": 56010 }, { "grad_norm": 0.37882140278816223, "learning_rate": 8.621043011425597e-05, "loss": 0.0389, "step": 56020 }, { "grad_norm": 0.6869708299636841, "learning_rate": 8.620472861277125e-05, "loss": 0.0467, "step": 56030 }, { "grad_norm": 0.5263819694519043, "learning_rate": 8.619902612146284e-05, "loss": 0.0519, "step": 56040 }, { "grad_norm": 0.5423586964607239, "learning_rate": 8.619332264048667e-05, "loss": 0.0536, "step": 56050 }, { "grad_norm": 0.8623807430267334, "learning_rate": 8.618761816999865e-05, "loss": 0.0531, "step": 56060 }, { "grad_norm": 0.7184329628944397, "learning_rate": 8.618191271015475e-05, "loss": 0.0443, "step": 56070 }, { "grad_norm": 0.5815322995185852, "learning_rate": 8.617620626111094e-05, "loss": 0.0454, "step": 56080 }, { "grad_norm": 0.6569144129753113, "learning_rate": 8.617049882302326e-05, "loss": 0.0466, "step": 56090 }, { "grad_norm": 0.7491828799247742, "learning_rate": 8.616479039604771e-05, "loss": 0.0425, "step": 56100 }, { "grad_norm": 0.33274394273757935, "learning_rate": 8.615908098034038e-05, "loss": 0.0365, "step": 56110 }, { "grad_norm": 0.5796927213668823, "learning_rate": 8.615337057605736e-05, "loss": 0.0498, "step": 56120 }, { "grad_norm": 0.7296192049980164, "learning_rate": 8.614765918335479e-05, "loss": 0.0547, "step": 56130 }, { "grad_norm": 0.515975832939148, "learning_rate": 8.614194680238878e-05, "loss": 0.0444, "step": 56140 }, { "grad_norm": 1.5021076202392578, "learning_rate": 8.613623343331553e-05, "loss": 0.0467, "step": 56150 }, { "grad_norm": 0.6328263282775879, "learning_rate": 8.613051907629123e-05, "loss": 0.0357, "step": 56160 }, { "grad_norm": 0.5664197206497192, "learning_rate": 8.612480373147211e-05, "loss": 0.0444, "step": 56170 }, { "grad_norm": 0.5962666273117065, "learning_rate": 8.611908739901444e-05, "loss": 0.053, "step": 56180 }, { "grad_norm": 0.3726191520690918, "learning_rate": 8.611337007907448e-05, "loss": 0.0497, "step": 56190 }, { "grad_norm": 0.6569390892982483, "learning_rate": 8.610765177180853e-05, "loss": 0.0312, "step": 56200 }, { "grad_norm": 1.598939061164856, "learning_rate": 8.610193247737297e-05, "loss": 0.0412, "step": 56210 }, { "grad_norm": 1.4982149600982666, "learning_rate": 8.609621219592413e-05, "loss": 0.0406, "step": 56220 }, { "grad_norm": 0.5446597337722778, "learning_rate": 8.60904909276184e-05, "loss": 0.049, "step": 56230 }, { "grad_norm": 0.6415248513221741, "learning_rate": 8.608476867261224e-05, "loss": 0.0416, "step": 56240 }, { "grad_norm": 0.8591428399085999, "learning_rate": 8.607904543106202e-05, "loss": 0.0426, "step": 56250 }, { "grad_norm": 0.6908059120178223, "learning_rate": 8.607332120312428e-05, "loss": 0.0461, "step": 56260 }, { "grad_norm": 0.7497454285621643, "learning_rate": 8.606759598895548e-05, "loss": 0.0389, "step": 56270 }, { "grad_norm": 1.078325867652893, "learning_rate": 8.606186978871215e-05, "loss": 0.0504, "step": 56280 }, { "grad_norm": 0.5368288159370422, "learning_rate": 8.605614260255087e-05, "loss": 0.0427, "step": 56290 }, { "grad_norm": 0.7352173328399658, "learning_rate": 8.605041443062818e-05, "loss": 0.0734, "step": 56300 }, { "grad_norm": 0.4594075381755829, "learning_rate": 8.60446852731007e-05, "loss": 0.0421, "step": 56310 }, { "grad_norm": 0.5341855883598328, "learning_rate": 8.603895513012507e-05, "loss": 0.0538, "step": 56320 }, { "grad_norm": 0.5378053188323975, "learning_rate": 8.603322400185796e-05, "loss": 0.0415, "step": 56330 }, { "grad_norm": 0.615790843963623, "learning_rate": 8.602749188845603e-05, "loss": 0.0495, "step": 56340 }, { "grad_norm": 0.5340067148208618, "learning_rate": 8.6021758790076e-05, "loss": 0.0703, "step": 56350 }, { "grad_norm": 0.7905066609382629, "learning_rate": 8.601602470687462e-05, "loss": 0.0424, "step": 56360 }, { "grad_norm": 0.6289423704147339, "learning_rate": 8.601028963900867e-05, "loss": 0.0622, "step": 56370 }, { "grad_norm": 0.8754615187644958, "learning_rate": 8.60045535866349e-05, "loss": 0.059, "step": 56380 }, { "grad_norm": 1.0491081476211548, "learning_rate": 8.59988165499102e-05, "loss": 0.0599, "step": 56390 }, { "grad_norm": 0.7014260292053223, "learning_rate": 8.599307852899135e-05, "loss": 0.0492, "step": 56400 }, { "grad_norm": 0.5094467401504517, "learning_rate": 8.598733952403527e-05, "loss": 0.0481, "step": 56410 }, { "grad_norm": 0.6157690286636353, "learning_rate": 8.598159953519885e-05, "loss": 0.0391, "step": 56420 }, { "grad_norm": 0.796897828578949, "learning_rate": 8.597585856263901e-05, "loss": 0.045, "step": 56430 }, { "grad_norm": 0.6931184530258179, "learning_rate": 8.597011660651271e-05, "loss": 0.0447, "step": 56440 }, { "grad_norm": 0.5211248993873596, "learning_rate": 8.596437366697692e-05, "loss": 0.045, "step": 56450 }, { "grad_norm": 0.6672224402427673, "learning_rate": 8.595862974418869e-05, "loss": 0.0519, "step": 56460 }, { "grad_norm": 0.5110237002372742, "learning_rate": 8.595288483830502e-05, "loss": 0.0348, "step": 56470 }, { "grad_norm": 0.5412623882293701, "learning_rate": 8.594713894948299e-05, "loss": 0.0386, "step": 56480 }, { "grad_norm": 0.6839032769203186, "learning_rate": 8.594139207787968e-05, "loss": 0.0473, "step": 56490 }, { "grad_norm": 0.9482561349868774, "learning_rate": 8.593564422365222e-05, "loss": 0.0524, "step": 56500 }, { "grad_norm": 0.5707886219024658, "learning_rate": 8.592989538695774e-05, "loss": 0.0545, "step": 56510 }, { "grad_norm": 0.6942980289459229, "learning_rate": 8.592414556795343e-05, "loss": 0.0556, "step": 56520 }, { "grad_norm": 0.3302472233772278, "learning_rate": 8.591839476679647e-05, "loss": 0.0441, "step": 56530 }, { "grad_norm": 0.8787384033203125, "learning_rate": 8.59126429836441e-05, "loss": 0.0497, "step": 56540 }, { "grad_norm": 0.4034753143787384, "learning_rate": 8.590689021865354e-05, "loss": 0.0537, "step": 56550 }, { "grad_norm": 0.48224398493766785, "learning_rate": 8.590113647198211e-05, "loss": 0.0435, "step": 56560 }, { "grad_norm": 0.5936315059661865, "learning_rate": 8.589538174378708e-05, "loss": 0.0451, "step": 56570 }, { "grad_norm": 0.5539041757583618, "learning_rate": 8.58896260342258e-05, "loss": 0.0397, "step": 56580 }, { "grad_norm": 0.7273064851760864, "learning_rate": 8.588386934345565e-05, "loss": 0.0387, "step": 56590 }, { "grad_norm": 0.5521100163459778, "learning_rate": 8.587811167163398e-05, "loss": 0.0404, "step": 56600 }, { "grad_norm": 0.45514780282974243, "learning_rate": 8.587235301891822e-05, "loss": 0.0337, "step": 56610 }, { "grad_norm": 0.5117447376251221, "learning_rate": 8.58665933854658e-05, "loss": 0.0447, "step": 56620 }, { "grad_norm": 1.3321386575698853, "learning_rate": 8.586083277143419e-05, "loss": 0.0546, "step": 56630 }, { "grad_norm": 1.2326713800430298, "learning_rate": 8.585507117698088e-05, "loss": 0.042, "step": 56640 }, { "grad_norm": 0.565963864326477, "learning_rate": 8.584930860226341e-05, "loss": 0.0542, "step": 56650 }, { "grad_norm": 1.056490182876587, "learning_rate": 8.58435450474393e-05, "loss": 0.0557, "step": 56660 }, { "grad_norm": 0.5245736241340637, "learning_rate": 8.583778051266613e-05, "loss": 0.052, "step": 56670 }, { "grad_norm": 0.7613095045089722, "learning_rate": 8.583201499810152e-05, "loss": 0.0353, "step": 56680 }, { "grad_norm": 0.7542089223861694, "learning_rate": 8.582624850390306e-05, "loss": 0.0373, "step": 56690 }, { "grad_norm": 0.606062650680542, "learning_rate": 8.582048103022844e-05, "loss": 0.0334, "step": 56700 }, { "grad_norm": 0.3068065643310547, "learning_rate": 8.581471257723532e-05, "loss": 0.0345, "step": 56710 }, { "grad_norm": 0.4625268578529358, "learning_rate": 8.58089431450814e-05, "loss": 0.0424, "step": 56720 }, { "grad_norm": 0.5459258556365967, "learning_rate": 8.580317273392446e-05, "loss": 0.04, "step": 56730 }, { "grad_norm": 0.6062901616096497, "learning_rate": 8.57974013439222e-05, "loss": 0.0535, "step": 56740 }, { "grad_norm": 0.721543550491333, "learning_rate": 8.579162897523246e-05, "loss": 0.045, "step": 56750 }, { "grad_norm": 0.515057384967804, "learning_rate": 8.5785855628013e-05, "loss": 0.0407, "step": 56760 }, { "grad_norm": 0.5419279336929321, "learning_rate": 8.578008130242171e-05, "loss": 0.0453, "step": 56770 }, { "grad_norm": 1.1566805839538574, "learning_rate": 8.577430599861644e-05, "loss": 0.049, "step": 56780 }, { "grad_norm": 0.47578558325767517, "learning_rate": 8.576852971675509e-05, "loss": 0.0488, "step": 56790 }, { "grad_norm": 1.1140928268432617, "learning_rate": 8.576275245699555e-05, "loss": 0.0409, "step": 56800 }, { "grad_norm": 0.6112644672393799, "learning_rate": 8.575697421949582e-05, "loss": 0.0396, "step": 56810 }, { "grad_norm": 0.6857845783233643, "learning_rate": 8.575119500441383e-05, "loss": 0.0538, "step": 56820 }, { "grad_norm": 0.5336726307868958, "learning_rate": 8.574541481190761e-05, "loss": 0.0588, "step": 56830 }, { "grad_norm": 0.5776033997535706, "learning_rate": 8.573963364213518e-05, "loss": 0.0425, "step": 56840 }, { "grad_norm": 0.3828484117984772, "learning_rate": 8.573385149525458e-05, "loss": 0.0407, "step": 56850 }, { "grad_norm": 0.6172454953193665, "learning_rate": 8.572806837142391e-05, "loss": 0.0487, "step": 56860 }, { "grad_norm": 0.7191513180732727, "learning_rate": 8.572228427080127e-05, "loss": 0.0454, "step": 56870 }, { "grad_norm": 0.7438210844993591, "learning_rate": 8.571649919354481e-05, "loss": 0.0513, "step": 56880 }, { "grad_norm": 0.4846300184726715, "learning_rate": 8.571071313981267e-05, "loss": 0.0442, "step": 56890 }, { "grad_norm": 0.45539310574531555, "learning_rate": 8.570492610976305e-05, "loss": 0.0442, "step": 56900 }, { "grad_norm": 0.3331497609615326, "learning_rate": 8.569913810355417e-05, "loss": 0.046, "step": 56910 }, { "grad_norm": 0.5335228443145752, "learning_rate": 8.569334912134426e-05, "loss": 0.0546, "step": 56920 }, { "grad_norm": 0.2985856235027313, "learning_rate": 8.568755916329159e-05, "loss": 0.0477, "step": 56930 }, { "grad_norm": 0.5872313380241394, "learning_rate": 8.568176822955447e-05, "loss": 0.056, "step": 56940 }, { "grad_norm": 0.4460860788822174, "learning_rate": 8.567597632029121e-05, "loss": 0.0529, "step": 56950 }, { "grad_norm": 0.6938721537590027, "learning_rate": 8.567018343566016e-05, "loss": 0.0504, "step": 56960 }, { "grad_norm": 0.5241279006004333, "learning_rate": 8.566438957581968e-05, "loss": 0.041, "step": 56970 }, { "grad_norm": 0.6441833972930908, "learning_rate": 8.565859474092822e-05, "loss": 0.0468, "step": 56980 }, { "grad_norm": 1.0249943733215332, "learning_rate": 8.565279893114415e-05, "loss": 0.054, "step": 56990 }, { "grad_norm": 0.4852568209171295, "learning_rate": 8.564700214662595e-05, "loss": 0.0448, "step": 57000 }, { "grad_norm": 0.44154274463653564, "learning_rate": 8.564120438753212e-05, "loss": 0.0403, "step": 57010 }, { "grad_norm": 0.741675853729248, "learning_rate": 8.563540565402115e-05, "loss": 0.046, "step": 57020 }, { "grad_norm": 1.015417456626892, "learning_rate": 8.562960594625155e-05, "loss": 0.047, "step": 57030 }, { "grad_norm": 0.4400978088378906, "learning_rate": 8.562380526438194e-05, "loss": 0.0559, "step": 57040 }, { "grad_norm": 0.5517438054084778, "learning_rate": 8.561800360857086e-05, "loss": 0.0469, "step": 57050 }, { "grad_norm": 0.7193424701690674, "learning_rate": 8.561220097897696e-05, "loss": 0.0588, "step": 57060 }, { "grad_norm": 0.7478760480880737, "learning_rate": 8.560639737575884e-05, "loss": 0.05, "step": 57070 }, { "grad_norm": 0.3807200491428375, "learning_rate": 8.56005927990752e-05, "loss": 0.0424, "step": 57080 }, { "grad_norm": 0.766995906829834, "learning_rate": 8.559478724908474e-05, "loss": 0.0576, "step": 57090 }, { "grad_norm": 1.293255090713501, "learning_rate": 8.558898072594617e-05, "loss": 0.0528, "step": 57100 }, { "grad_norm": 0.5646984577178955, "learning_rate": 8.558317322981821e-05, "loss": 0.0473, "step": 57110 }, { "grad_norm": 0.4498215615749359, "learning_rate": 8.557736476085968e-05, "loss": 0.0489, "step": 57120 }, { "grad_norm": 0.6522635817527771, "learning_rate": 8.557155531922937e-05, "loss": 0.0686, "step": 57130 }, { "grad_norm": 0.5033310651779175, "learning_rate": 8.556574490508611e-05, "loss": 0.0439, "step": 57140 }, { "grad_norm": 0.7052866816520691, "learning_rate": 8.555993351858874e-05, "loss": 0.0667, "step": 57150 }, { "grad_norm": 0.538368821144104, "learning_rate": 8.555412115989612e-05, "loss": 0.0348, "step": 57160 }, { "grad_norm": 0.6608472466468811, "learning_rate": 8.55483078291672e-05, "loss": 0.0498, "step": 57170 }, { "grad_norm": 0.4757680594921112, "learning_rate": 8.55424935265609e-05, "loss": 0.0431, "step": 57180 }, { "grad_norm": 0.42964601516723633, "learning_rate": 8.55366782522362e-05, "loss": 0.0423, "step": 57190 }, { "grad_norm": 0.5764769911766052, "learning_rate": 8.553086200635203e-05, "loss": 0.0441, "step": 57200 }, { "grad_norm": 0.5927271246910095, "learning_rate": 8.552504478906747e-05, "loss": 0.0751, "step": 57210 }, { "grad_norm": 0.4558502733707428, "learning_rate": 8.551922660054152e-05, "loss": 0.0567, "step": 57220 }, { "grad_norm": 1.2913181781768799, "learning_rate": 8.551340744093325e-05, "loss": 0.0523, "step": 57230 }, { "grad_norm": 0.6480346322059631, "learning_rate": 8.550758731040177e-05, "loss": 0.0633, "step": 57240 }, { "grad_norm": 0.7477328777313232, "learning_rate": 8.550176620910619e-05, "loss": 0.0427, "step": 57250 }, { "grad_norm": 0.9756854176521301, "learning_rate": 8.549594413720568e-05, "loss": 0.0437, "step": 57260 }, { "grad_norm": 0.7626193761825562, "learning_rate": 8.549012109485937e-05, "loss": 0.0548, "step": 57270 }, { "grad_norm": 0.5830135941505432, "learning_rate": 8.548429708222648e-05, "loss": 0.0463, "step": 57280 }, { "grad_norm": 0.6601433157920837, "learning_rate": 8.547847209946625e-05, "loss": 0.0501, "step": 57290 }, { "grad_norm": 0.947878360748291, "learning_rate": 8.547264614673792e-05, "loss": 0.0552, "step": 57300 }, { "grad_norm": 0.5362295508384705, "learning_rate": 8.546681922420076e-05, "loss": 0.045, "step": 57310 }, { "grad_norm": 0.7806485891342163, "learning_rate": 8.54609913320141e-05, "loss": 0.0576, "step": 57320 }, { "grad_norm": 0.4635346233844757, "learning_rate": 8.545516247033726e-05, "loss": 0.0393, "step": 57330 }, { "grad_norm": 0.6033256649971008, "learning_rate": 8.544933263932959e-05, "loss": 0.0418, "step": 57340 }, { "grad_norm": 0.6361967921257019, "learning_rate": 8.544350183915049e-05, "loss": 0.0463, "step": 57350 }, { "grad_norm": 0.603338897228241, "learning_rate": 8.543767006995935e-05, "loss": 0.0495, "step": 57360 }, { "grad_norm": 0.5019541382789612, "learning_rate": 8.543183733191564e-05, "loss": 0.0462, "step": 57370 }, { "grad_norm": 0.73886638879776, "learning_rate": 8.54260036251788e-05, "loss": 0.0427, "step": 57380 }, { "grad_norm": 0.5400415062904358, "learning_rate": 8.542016894990834e-05, "loss": 0.0567, "step": 57390 }, { "grad_norm": 0.41304415464401245, "learning_rate": 8.541433330626377e-05, "loss": 0.0419, "step": 57400 }, { "grad_norm": 0.6779380440711975, "learning_rate": 8.54084966944046e-05, "loss": 0.0404, "step": 57410 }, { "grad_norm": 0.480716347694397, "learning_rate": 8.540265911449045e-05, "loss": 0.041, "step": 57420 }, { "grad_norm": 1.3605608940124512, "learning_rate": 8.539682056668092e-05, "loss": 0.0401, "step": 57430 }, { "grad_norm": 0.8376333117485046, "learning_rate": 8.53909810511356e-05, "loss": 0.0608, "step": 57440 }, { "grad_norm": 1.7589315176010132, "learning_rate": 8.538514056801415e-05, "loss": 0.0536, "step": 57450 }, { "grad_norm": 0.4389480948448181, "learning_rate": 8.537929911747625e-05, "loss": 0.079, "step": 57460 }, { "grad_norm": 0.48637375235557556, "learning_rate": 8.537345669968161e-05, "loss": 0.038, "step": 57470 }, { "grad_norm": 0.3792324960231781, "learning_rate": 8.536761331478996e-05, "loss": 0.0405, "step": 57480 }, { "grad_norm": 0.5595366954803467, "learning_rate": 8.536176896296103e-05, "loss": 0.0541, "step": 57490 }, { "grad_norm": 0.5754213333129883, "learning_rate": 8.535592364435463e-05, "loss": 0.0516, "step": 57500 }, { "grad_norm": 0.556962251663208, "learning_rate": 8.535007735913056e-05, "loss": 0.0784, "step": 57510 }, { "grad_norm": 0.48955661058425903, "learning_rate": 8.534423010744865e-05, "loss": 0.0436, "step": 57520 }, { "grad_norm": 0.6426209807395935, "learning_rate": 8.53383818894688e-05, "loss": 0.0449, "step": 57530 }, { "grad_norm": 0.6947110891342163, "learning_rate": 8.533253270535082e-05, "loss": 0.042, "step": 57540 }, { "grad_norm": 0.590838611125946, "learning_rate": 8.532668255525468e-05, "loss": 0.0382, "step": 57550 }, { "grad_norm": 0.64238041639328, "learning_rate": 8.532083143934033e-05, "loss": 0.0399, "step": 57560 }, { "grad_norm": 0.6550117135047913, "learning_rate": 8.53149793577677e-05, "loss": 0.0452, "step": 57570 }, { "grad_norm": 0.38980597257614136, "learning_rate": 8.530912631069681e-05, "loss": 0.0428, "step": 57580 }, { "grad_norm": 0.7590561509132385, "learning_rate": 8.530327229828767e-05, "loss": 0.0491, "step": 57590 }, { "grad_norm": 0.657791018486023, "learning_rate": 8.529741732070033e-05, "loss": 0.0455, "step": 57600 }, { "grad_norm": 0.4567977488040924, "learning_rate": 8.529156137809484e-05, "loss": 0.0272, "step": 57610 }, { "grad_norm": 0.42773258686065674, "learning_rate": 8.528570447063136e-05, "loss": 0.052, "step": 57620 }, { "grad_norm": 0.8402823209762573, "learning_rate": 8.527984659846995e-05, "loss": 0.0456, "step": 57630 }, { "grad_norm": 1.2556087970733643, "learning_rate": 8.527398776177078e-05, "loss": 0.0461, "step": 57640 }, { "grad_norm": 1.3836297988891602, "learning_rate": 8.526812796069405e-05, "loss": 0.0472, "step": 57650 }, { "grad_norm": 0.4351571202278137, "learning_rate": 8.526226719539996e-05, "loss": 0.0432, "step": 57660 }, { "grad_norm": 0.4761451184749603, "learning_rate": 8.525640546604871e-05, "loss": 0.0471, "step": 57670 }, { "grad_norm": 0.6573383212089539, "learning_rate": 8.525054277280061e-05, "loss": 0.0435, "step": 57680 }, { "grad_norm": 0.6194483041763306, "learning_rate": 8.524467911581588e-05, "loss": 0.0555, "step": 57690 }, { "grad_norm": 0.610023021697998, "learning_rate": 8.523881449525488e-05, "loss": 0.0583, "step": 57700 }, { "grad_norm": 0.6945698857307434, "learning_rate": 8.523294891127794e-05, "loss": 0.0518, "step": 57710 }, { "grad_norm": 0.6018849611282349, "learning_rate": 8.52270823640454e-05, "loss": 0.0524, "step": 57720 }, { "grad_norm": 0.8116734623908997, "learning_rate": 8.522121485371767e-05, "loss": 0.0546, "step": 57730 }, { "grad_norm": 0.72627854347229, "learning_rate": 8.521534638045514e-05, "loss": 0.0402, "step": 57740 }, { "grad_norm": 0.4748493432998657, "learning_rate": 8.52094769444183e-05, "loss": 0.0591, "step": 57750 }, { "grad_norm": 0.349618524312973, "learning_rate": 8.520360654576757e-05, "loss": 0.0471, "step": 57760 }, { "grad_norm": 0.443308562040329, "learning_rate": 8.519773518466346e-05, "loss": 0.042, "step": 57770 }, { "grad_norm": 1.1095017194747925, "learning_rate": 8.51918628612665e-05, "loss": 0.0397, "step": 57780 }, { "grad_norm": 0.39793339371681213, "learning_rate": 8.518598957573722e-05, "loss": 0.0554, "step": 57790 }, { "grad_norm": 0.4431682527065277, "learning_rate": 8.518011532823621e-05, "loss": 0.04, "step": 57800 }, { "grad_norm": 0.4405677020549774, "learning_rate": 8.517424011892407e-05, "loss": 0.0541, "step": 57810 }, { "grad_norm": 0.5073069334030151, "learning_rate": 8.516836394796142e-05, "loss": 0.0504, "step": 57820 }, { "grad_norm": 0.7734730243682861, "learning_rate": 8.51624868155089e-05, "loss": 0.0431, "step": 57830 }, { "grad_norm": 0.6521824598312378, "learning_rate": 8.515660872172722e-05, "loss": 0.0492, "step": 57840 }, { "grad_norm": 0.531443178653717, "learning_rate": 8.515072966677704e-05, "loss": 0.0403, "step": 57850 }, { "grad_norm": 1.009893774986267, "learning_rate": 8.514484965081913e-05, "loss": 0.0465, "step": 57860 }, { "grad_norm": 0.5789510011672974, "learning_rate": 8.513896867401423e-05, "loss": 0.0425, "step": 57870 }, { "grad_norm": 0.5324895977973938, "learning_rate": 8.513308673652313e-05, "loss": 0.0495, "step": 57880 }, { "grad_norm": 1.1075769662857056, "learning_rate": 8.512720383850663e-05, "loss": 0.0498, "step": 57890 }, { "grad_norm": 0.8270159363746643, "learning_rate": 8.512131998012558e-05, "loss": 0.0413, "step": 57900 }, { "grad_norm": 0.5172551870346069, "learning_rate": 8.511543516154084e-05, "loss": 0.0785, "step": 57910 }, { "grad_norm": 0.5657901763916016, "learning_rate": 8.51095493829133e-05, "loss": 0.0367, "step": 57920 }, { "grad_norm": 0.546258807182312, "learning_rate": 8.510366264440384e-05, "loss": 0.0495, "step": 57930 }, { "grad_norm": 0.7533573508262634, "learning_rate": 8.509777494617346e-05, "loss": 0.0494, "step": 57940 }, { "grad_norm": 0.6970549821853638, "learning_rate": 8.50918862883831e-05, "loss": 0.0415, "step": 57950 }, { "grad_norm": 0.5233452916145325, "learning_rate": 8.508599667119374e-05, "loss": 0.0464, "step": 57960 }, { "grad_norm": 0.4245389401912689, "learning_rate": 8.508010609476643e-05, "loss": 0.0321, "step": 57970 }, { "grad_norm": 0.7024398446083069, "learning_rate": 8.507421455926217e-05, "loss": 0.0448, "step": 57980 }, { "grad_norm": 0.6611009836196899, "learning_rate": 8.506832206484207e-05, "loss": 0.0403, "step": 57990 }, { "grad_norm": 0.5675743818283081, "learning_rate": 8.506242861166723e-05, "loss": 0.0437, "step": 58000 }, { "grad_norm": 0.5454922318458557, "learning_rate": 8.505653419989876e-05, "loss": 0.0371, "step": 58010 }, { "grad_norm": 0.7770876288414001, "learning_rate": 8.505063882969782e-05, "loss": 0.0429, "step": 58020 }, { "grad_norm": 0.45626088976860046, "learning_rate": 8.504474250122558e-05, "loss": 0.0456, "step": 58030 }, { "grad_norm": 0.4878878593444824, "learning_rate": 8.503884521464322e-05, "loss": 0.0379, "step": 58040 }, { "grad_norm": 0.48145994544029236, "learning_rate": 8.503294697011202e-05, "loss": 0.0467, "step": 58050 }, { "grad_norm": 0.3524177074432373, "learning_rate": 8.50270477677932e-05, "loss": 0.0486, "step": 58060 }, { "grad_norm": 0.432647705078125, "learning_rate": 8.502114760784808e-05, "loss": 0.0387, "step": 58070 }, { "grad_norm": 0.43535536527633667, "learning_rate": 8.501524649043792e-05, "loss": 0.0532, "step": 58080 }, { "grad_norm": 0.9346721172332764, "learning_rate": 8.500934441572408e-05, "loss": 0.0548, "step": 58090 }, { "grad_norm": 0.4725319743156433, "learning_rate": 8.50034413838679e-05, "loss": 0.0476, "step": 58100 }, { "grad_norm": 0.5665035843849182, "learning_rate": 8.49975373950308e-05, "loss": 0.0447, "step": 58110 }, { "grad_norm": 0.9639389514923096, "learning_rate": 8.499163244937418e-05, "loss": 0.0501, "step": 58120 }, { "grad_norm": 0.5671550631523132, "learning_rate": 8.498572654705947e-05, "loss": 0.0487, "step": 58130 }, { "grad_norm": 1.3717859983444214, "learning_rate": 8.497981968824813e-05, "loss": 0.0524, "step": 58140 }, { "grad_norm": 0.5234237313270569, "learning_rate": 8.497391187310166e-05, "loss": 0.0401, "step": 58150 }, { "grad_norm": 0.40518009662628174, "learning_rate": 8.496800310178158e-05, "loss": 0.0418, "step": 58160 }, { "grad_norm": 1.1566493511199951, "learning_rate": 8.496209337444945e-05, "loss": 0.0538, "step": 58170 }, { "grad_norm": 0.718980073928833, "learning_rate": 8.495618269126681e-05, "loss": 0.0365, "step": 58180 }, { "grad_norm": 0.7026581764221191, "learning_rate": 8.495027105239526e-05, "loss": 0.046, "step": 58190 }, { "grad_norm": 0.5739437341690063, "learning_rate": 8.494435845799645e-05, "loss": 0.0406, "step": 58200 }, { "grad_norm": 0.5919190645217896, "learning_rate": 8.493844490823199e-05, "loss": 0.0369, "step": 58210 }, { "grad_norm": 0.5758712291717529, "learning_rate": 8.493253040326358e-05, "loss": 0.0415, "step": 58220 }, { "grad_norm": 0.9962618947029114, "learning_rate": 8.492661494325291e-05, "loss": 0.0373, "step": 58230 }, { "grad_norm": 0.6153793334960938, "learning_rate": 8.49206985283617e-05, "loss": 0.0381, "step": 58240 }, { "grad_norm": 0.41088858246803284, "learning_rate": 8.491478115875172e-05, "loss": 0.0377, "step": 58250 }, { "grad_norm": 0.5678367614746094, "learning_rate": 8.490886283458474e-05, "loss": 0.0385, "step": 58260 }, { "grad_norm": 0.667844831943512, "learning_rate": 8.490294355602258e-05, "loss": 0.0413, "step": 58270 }, { "grad_norm": 0.5160004496574402, "learning_rate": 8.489702332322703e-05, "loss": 0.041, "step": 58280 }, { "grad_norm": 0.8378148674964905, "learning_rate": 8.489110213635998e-05, "loss": 0.0371, "step": 58290 }, { "grad_norm": 0.42395642399787903, "learning_rate": 8.488517999558331e-05, "loss": 0.0472, "step": 58300 }, { "grad_norm": 1.063399314880371, "learning_rate": 8.487925690105891e-05, "loss": 0.0621, "step": 58310 }, { "grad_norm": 0.5215184092521667, "learning_rate": 8.487333285294875e-05, "loss": 0.0505, "step": 58320 }, { "grad_norm": 0.40608471632003784, "learning_rate": 8.486740785141476e-05, "loss": 0.0505, "step": 58330 }, { "grad_norm": 0.54546058177948, "learning_rate": 8.486148189661895e-05, "loss": 0.0583, "step": 58340 }, { "grad_norm": 0.7050957083702087, "learning_rate": 8.485555498872331e-05, "loss": 0.0505, "step": 58350 }, { "grad_norm": 0.5010291337966919, "learning_rate": 8.48496271278899e-05, "loss": 0.0455, "step": 58360 }, { "grad_norm": 0.7502813339233398, "learning_rate": 8.484369831428076e-05, "loss": 0.0362, "step": 58370 }, { "grad_norm": 0.7051516771316528, "learning_rate": 8.483776854805801e-05, "loss": 0.0495, "step": 58380 }, { "grad_norm": 0.518481433391571, "learning_rate": 8.483183782938375e-05, "loss": 0.0401, "step": 58390 }, { "grad_norm": 0.8360773921012878, "learning_rate": 8.482590615842014e-05, "loss": 0.0468, "step": 58400 }, { "grad_norm": 0.5093555450439453, "learning_rate": 8.481997353532934e-05, "loss": 0.0439, "step": 58410 }, { "grad_norm": 0.7074112892150879, "learning_rate": 8.481403996027354e-05, "loss": 0.0454, "step": 58420 }, { "grad_norm": 0.32578805088996887, "learning_rate": 8.480810543341497e-05, "loss": 0.0501, "step": 58430 }, { "grad_norm": 0.39040276408195496, "learning_rate": 8.480216995491586e-05, "loss": 0.0498, "step": 58440 }, { "grad_norm": 0.5100039839744568, "learning_rate": 8.47962335249385e-05, "loss": 0.0356, "step": 58450 }, { "grad_norm": 0.509796142578125, "learning_rate": 8.47902961436452e-05, "loss": 0.0382, "step": 58460 }, { "grad_norm": 0.6576712727546692, "learning_rate": 8.478435781119828e-05, "loss": 0.0665, "step": 58470 }, { "grad_norm": 0.48308828473091125, "learning_rate": 8.477841852776008e-05, "loss": 0.0461, "step": 58480 }, { "grad_norm": 0.434447318315506, "learning_rate": 8.477247829349297e-05, "loss": 0.0393, "step": 58490 }, { "grad_norm": 0.6426504254341125, "learning_rate": 8.476653710855938e-05, "loss": 0.0512, "step": 58500 }, { "grad_norm": 0.40288376808166504, "learning_rate": 8.476059497312171e-05, "loss": 0.0467, "step": 58510 }, { "grad_norm": 1.3426016569137573, "learning_rate": 8.475465188734244e-05, "loss": 0.0525, "step": 58520 }, { "grad_norm": 0.5540390014648438, "learning_rate": 8.474870785138406e-05, "loss": 0.0427, "step": 58530 }, { "grad_norm": 0.5016782283782959, "learning_rate": 8.474276286540905e-05, "loss": 0.0485, "step": 58540 }, { "grad_norm": 0.4495926797389984, "learning_rate": 8.473681692957996e-05, "loss": 0.0486, "step": 58550 }, { "grad_norm": 0.8641772270202637, "learning_rate": 8.473087004405932e-05, "loss": 0.045, "step": 58560 }, { "grad_norm": 0.7490901947021484, "learning_rate": 8.472492220900976e-05, "loss": 0.0503, "step": 58570 }, { "grad_norm": 0.942339301109314, "learning_rate": 8.471897342459389e-05, "loss": 0.0397, "step": 58580 }, { "grad_norm": 0.615370512008667, "learning_rate": 8.471302369097431e-05, "loss": 0.05, "step": 58590 }, { "grad_norm": 0.6505486369132996, "learning_rate": 8.470707300831371e-05, "loss": 0.0458, "step": 58600 }, { "grad_norm": 0.47420138120651245, "learning_rate": 8.470112137677476e-05, "loss": 0.0559, "step": 58610 }, { "grad_norm": 0.6109735369682312, "learning_rate": 8.469516879652019e-05, "loss": 0.0607, "step": 58620 }, { "grad_norm": 0.4715736210346222, "learning_rate": 8.468921526771276e-05, "loss": 0.0603, "step": 58630 }, { "grad_norm": 0.42667341232299805, "learning_rate": 8.468326079051519e-05, "loss": 0.0403, "step": 58640 }, { "grad_norm": 0.41530799865722656, "learning_rate": 8.467730536509033e-05, "loss": 0.0455, "step": 58650 }, { "grad_norm": 0.9979819655418396, "learning_rate": 8.467134899160093e-05, "loss": 0.0561, "step": 58660 }, { "grad_norm": 0.5350427627563477, "learning_rate": 8.46653916702099e-05, "loss": 0.0462, "step": 58670 }, { "grad_norm": 1.7202277183532715, "learning_rate": 8.46594334010801e-05, "loss": 0.0381, "step": 58680 }, { "grad_norm": 0.8602548241615295, "learning_rate": 8.465347418437439e-05, "loss": 0.0418, "step": 58690 }, { "grad_norm": 0.6745072603225708, "learning_rate": 8.46475140202557e-05, "loss": 0.0536, "step": 58700 }, { "grad_norm": 0.5166704654693604, "learning_rate": 8.464155290888703e-05, "loss": 0.0478, "step": 58710 }, { "grad_norm": 0.5323510766029358, "learning_rate": 8.463559085043129e-05, "loss": 0.0544, "step": 58720 }, { "grad_norm": 0.5403494238853455, "learning_rate": 8.462962784505152e-05, "loss": 0.0544, "step": 58730 }, { "grad_norm": 0.42667990922927856, "learning_rate": 8.462366389291073e-05, "loss": 0.0492, "step": 58740 }, { "grad_norm": 0.48785334825515747, "learning_rate": 8.461769899417198e-05, "loss": 0.0374, "step": 58750 }, { "grad_norm": 0.6598209142684937, "learning_rate": 8.461173314899833e-05, "loss": 0.0551, "step": 58760 }, { "grad_norm": 0.48008206486701965, "learning_rate": 8.460576635755292e-05, "loss": 0.0358, "step": 58770 }, { "grad_norm": 0.4115549623966217, "learning_rate": 8.459979861999885e-05, "loss": 0.0389, "step": 58780 }, { "grad_norm": 0.6356528997421265, "learning_rate": 8.45938299364993e-05, "loss": 0.0577, "step": 58790 }, { "grad_norm": 1.5307426452636719, "learning_rate": 8.458786030721741e-05, "loss": 0.0568, "step": 58800 }, { "grad_norm": 0.5051775574684143, "learning_rate": 8.458188973231642e-05, "loss": 0.0493, "step": 58810 }, { "grad_norm": 0.9529134035110474, "learning_rate": 8.457591821195956e-05, "loss": 0.0586, "step": 58820 }, { "grad_norm": 0.5141188502311707, "learning_rate": 8.456994574631008e-05, "loss": 0.0384, "step": 58830 }, { "grad_norm": 0.3795774579048157, "learning_rate": 8.456397233553129e-05, "loss": 0.0432, "step": 58840 }, { "grad_norm": 0.6383575201034546, "learning_rate": 8.455799797978645e-05, "loss": 0.0364, "step": 58850 }, { "grad_norm": 0.42492181062698364, "learning_rate": 8.455202267923895e-05, "loss": 0.0451, "step": 58860 }, { "grad_norm": 0.6146179437637329, "learning_rate": 8.454604643405213e-05, "loss": 0.0334, "step": 58870 }, { "grad_norm": 0.5076762437820435, "learning_rate": 8.454006924438937e-05, "loss": 0.0576, "step": 58880 }, { "grad_norm": 0.5833339095115662, "learning_rate": 8.45340911104141e-05, "loss": 0.0535, "step": 58890 }, { "grad_norm": 0.6835834980010986, "learning_rate": 8.452811203228974e-05, "loss": 0.0418, "step": 58900 }, { "grad_norm": 0.5469604134559631, "learning_rate": 8.45221320101798e-05, "loss": 0.0378, "step": 58910 }, { "grad_norm": 0.47297731041908264, "learning_rate": 8.451615104424771e-05, "loss": 0.0446, "step": 58920 }, { "grad_norm": 0.5771899223327637, "learning_rate": 8.451016913465702e-05, "loss": 0.0454, "step": 58930 }, { "grad_norm": 0.5671261548995972, "learning_rate": 8.450418628157128e-05, "loss": 0.0514, "step": 58940 }, { "grad_norm": 0.5177995562553406, "learning_rate": 8.449820248515404e-05, "loss": 0.0581, "step": 58950 }, { "grad_norm": 0.43818700313568115, "learning_rate": 8.449221774556892e-05, "loss": 0.0371, "step": 58960 }, { "grad_norm": 0.5518098473548889, "learning_rate": 8.448623206297951e-05, "loss": 0.0401, "step": 58970 }, { "grad_norm": 0.48986053466796875, "learning_rate": 8.448024543754949e-05, "loss": 0.0486, "step": 58980 }, { "grad_norm": 1.3089040517807007, "learning_rate": 8.447425786944249e-05, "loss": 0.0542, "step": 58990 }, { "grad_norm": 0.3905339241027832, "learning_rate": 8.446826935882227e-05, "loss": 0.0468, "step": 59000 }, { "grad_norm": 1.0219392776489258, "learning_rate": 8.446227990585248e-05, "loss": 0.0524, "step": 59010 }, { "grad_norm": 0.5034418106079102, "learning_rate": 8.445628951069689e-05, "loss": 0.048, "step": 59020 }, { "grad_norm": 0.7734388709068298, "learning_rate": 8.445029817351933e-05, "loss": 0.0434, "step": 59030 }, { "grad_norm": 0.3468821048736572, "learning_rate": 8.444430589448354e-05, "loss": 0.0377, "step": 59040 }, { "grad_norm": 0.6347732543945312, "learning_rate": 8.443831267375336e-05, "loss": 0.0474, "step": 59050 }, { "grad_norm": 1.1260089874267578, "learning_rate": 8.443231851149266e-05, "loss": 0.0482, "step": 59060 }, { "grad_norm": 0.6836808919906616, "learning_rate": 8.442632340786529e-05, "loss": 0.044, "step": 59070 }, { "grad_norm": 0.49622291326522827, "learning_rate": 8.442032736303521e-05, "loss": 0.0665, "step": 59080 }, { "grad_norm": 0.35993286967277527, "learning_rate": 8.441433037716628e-05, "loss": 0.0474, "step": 59090 }, { "grad_norm": 0.8966407179832458, "learning_rate": 8.440833245042249e-05, "loss": 0.0533, "step": 59100 }, { "grad_norm": 0.5679399967193604, "learning_rate": 8.440233358296783e-05, "loss": 0.0405, "step": 59110 }, { "grad_norm": 0.35337623953819275, "learning_rate": 8.439633377496628e-05, "loss": 0.0445, "step": 59120 }, { "grad_norm": 0.6833202242851257, "learning_rate": 8.43903330265819e-05, "loss": 0.047, "step": 59130 }, { "grad_norm": 0.6606881022453308, "learning_rate": 8.438433133797871e-05, "loss": 0.0595, "step": 59140 }, { "grad_norm": 0.7789595127105713, "learning_rate": 8.437832870932084e-05, "loss": 0.0525, "step": 59150 }, { "grad_norm": 0.3528268039226532, "learning_rate": 8.437232514077238e-05, "loss": 0.0393, "step": 59160 }, { "grad_norm": 0.6561287641525269, "learning_rate": 8.436632063249745e-05, "loss": 0.04, "step": 59170 }, { "grad_norm": 0.5216057896614075, "learning_rate": 8.436031518466024e-05, "loss": 0.038, "step": 59180 }, { "grad_norm": 0.5047731399536133, "learning_rate": 8.435430879742491e-05, "loss": 0.0465, "step": 59190 }, { "grad_norm": 0.4583533704280853, "learning_rate": 8.434830147095569e-05, "loss": 0.033, "step": 59200 }, { "grad_norm": 0.3882066607475281, "learning_rate": 8.434229320541681e-05, "loss": 0.0363, "step": 59210 }, { "grad_norm": 0.43007683753967285, "learning_rate": 8.433628400097253e-05, "loss": 0.0345, "step": 59220 }, { "grad_norm": 0.6836838126182556, "learning_rate": 8.433027385778716e-05, "loss": 0.0446, "step": 59230 }, { "grad_norm": 0.5549392104148865, "learning_rate": 8.432426277602498e-05, "loss": 0.0408, "step": 59240 }, { "grad_norm": 0.5304375290870667, "learning_rate": 8.431825075585038e-05, "loss": 0.0476, "step": 59250 }, { "grad_norm": 0.5574551820755005, "learning_rate": 8.431223779742768e-05, "loss": 0.0396, "step": 59260 }, { "grad_norm": 0.6711778044700623, "learning_rate": 8.430622390092128e-05, "loss": 0.0386, "step": 59270 }, { "grad_norm": 0.7940248250961304, "learning_rate": 8.43002090664956e-05, "loss": 0.0496, "step": 59280 }, { "grad_norm": 0.4687122702598572, "learning_rate": 8.429419329431511e-05, "loss": 0.0479, "step": 59290 }, { "grad_norm": 0.6010881066322327, "learning_rate": 8.428817658454426e-05, "loss": 0.0422, "step": 59300 }, { "grad_norm": 0.6572991013526917, "learning_rate": 8.428215893734754e-05, "loss": 0.0541, "step": 59310 }, { "grad_norm": 0.4516046643257141, "learning_rate": 8.427614035288947e-05, "loss": 0.0594, "step": 59320 }, { "grad_norm": 0.7961480617523193, "learning_rate": 8.42701208313346e-05, "loss": 0.053, "step": 59330 }, { "grad_norm": 0.5944185853004456, "learning_rate": 8.42641003728475e-05, "loss": 0.045, "step": 59340 }, { "grad_norm": 0.37472003698349, "learning_rate": 8.425807897759277e-05, "loss": 0.0548, "step": 59350 }, { "grad_norm": 0.5862919092178345, "learning_rate": 8.425205664573504e-05, "loss": 0.0393, "step": 59360 }, { "grad_norm": 0.9350698590278625, "learning_rate": 8.424603337743892e-05, "loss": 0.0426, "step": 59370 }, { "grad_norm": 0.4331657588481903, "learning_rate": 8.424000917286914e-05, "loss": 0.0354, "step": 59380 }, { "grad_norm": 0.5868877172470093, "learning_rate": 8.423398403219036e-05, "loss": 0.0395, "step": 59390 }, { "grad_norm": 0.8504162430763245, "learning_rate": 8.422795795556732e-05, "loss": 0.0518, "step": 59400 }, { "grad_norm": 0.45122796297073364, "learning_rate": 8.422193094316478e-05, "loss": 0.0382, "step": 59410 }, { "grad_norm": 0.9392406940460205, "learning_rate": 8.42159029951475e-05, "loss": 0.0531, "step": 59420 }, { "grad_norm": 0.5183367729187012, "learning_rate": 8.420987411168029e-05, "loss": 0.0499, "step": 59430 }, { "grad_norm": 0.43120890855789185, "learning_rate": 8.420384429292796e-05, "loss": 0.04, "step": 59440 }, { "grad_norm": 0.7520502805709839, "learning_rate": 8.419781353905538e-05, "loss": 0.0499, "step": 59450 }, { "grad_norm": 0.44672366976737976, "learning_rate": 8.419178185022743e-05, "loss": 0.0565, "step": 59460 }, { "grad_norm": 0.5909003019332886, "learning_rate": 8.418574922660901e-05, "loss": 0.042, "step": 59470 }, { "grad_norm": 0.45990797877311707, "learning_rate": 8.417971566836505e-05, "loss": 0.0391, "step": 59480 }, { "grad_norm": 0.6877949833869934, "learning_rate": 8.417368117566051e-05, "loss": 0.0521, "step": 59490 }, { "grad_norm": 0.5953488349914551, "learning_rate": 8.416764574866035e-05, "loss": 0.0663, "step": 59500 }, { "grad_norm": 0.5820127129554749, "learning_rate": 8.416160938752961e-05, "loss": 0.0456, "step": 59510 }, { "grad_norm": 0.30570507049560547, "learning_rate": 8.415557209243328e-05, "loss": 0.0377, "step": 59520 }, { "grad_norm": 0.5282612442970276, "learning_rate": 8.414953386353645e-05, "loss": 0.0522, "step": 59530 }, { "grad_norm": 0.4853452444076538, "learning_rate": 8.41434947010042e-05, "loss": 0.0449, "step": 59540 }, { "grad_norm": 0.5147560238838196, "learning_rate": 8.413745460500163e-05, "loss": 0.0364, "step": 59550 }, { "grad_norm": 0.7688568830490112, "learning_rate": 8.413141357569387e-05, "loss": 0.0375, "step": 59560 }, { "grad_norm": 0.5035019516944885, "learning_rate": 8.41253716132461e-05, "loss": 0.0505, "step": 59570 }, { "grad_norm": 0.3911789357662201, "learning_rate": 8.411932871782347e-05, "loss": 0.0347, "step": 59580 }, { "grad_norm": 0.9121109843254089, "learning_rate": 8.411328488959123e-05, "loss": 0.0382, "step": 59590 }, { "grad_norm": 0.559390664100647, "learning_rate": 8.410724012871459e-05, "loss": 0.0386, "step": 59600 }, { "grad_norm": 0.5647523999214172, "learning_rate": 8.41011944353588e-05, "loss": 0.0489, "step": 59610 }, { "grad_norm": 0.7474955320358276, "learning_rate": 8.409514780968918e-05, "loss": 0.0441, "step": 59620 }, { "grad_norm": 0.9092351198196411, "learning_rate": 8.408910025187103e-05, "loss": 0.0467, "step": 59630 }, { "grad_norm": 0.5335087776184082, "learning_rate": 8.40830517620697e-05, "loss": 0.0327, "step": 59640 }, { "grad_norm": 0.6383049488067627, "learning_rate": 8.407700234045051e-05, "loss": 0.0664, "step": 59650 }, { "grad_norm": 1.0251859426498413, "learning_rate": 8.407095198717889e-05, "loss": 0.0425, "step": 59660 }, { "grad_norm": 0.6588297486305237, "learning_rate": 8.406490070242024e-05, "loss": 0.0629, "step": 59670 }, { "grad_norm": 0.3589506447315216, "learning_rate": 8.405884848634003e-05, "loss": 0.034, "step": 59680 }, { "grad_norm": 0.3628748655319214, "learning_rate": 8.405279533910366e-05, "loss": 0.0558, "step": 59690 }, { "grad_norm": 0.5019807815551758, "learning_rate": 8.404674126087668e-05, "loss": 0.0434, "step": 59700 }, { "grad_norm": 0.5616974830627441, "learning_rate": 8.404068625182457e-05, "loss": 0.0398, "step": 59710 }, { "grad_norm": 0.48263019323349, "learning_rate": 8.403463031211289e-05, "loss": 0.0409, "step": 59720 }, { "grad_norm": 0.6056039929389954, "learning_rate": 8.402857344190722e-05, "loss": 0.041, "step": 59730 }, { "grad_norm": 0.612606406211853, "learning_rate": 8.402251564137311e-05, "loss": 0.033, "step": 59740 }, { "grad_norm": 0.49792397022247314, "learning_rate": 8.401645691067624e-05, "loss": 0.0435, "step": 59750 }, { "grad_norm": 0.8373286128044128, "learning_rate": 8.401039724998218e-05, "loss": 0.0526, "step": 59760 }, { "grad_norm": 0.30968162417411804, "learning_rate": 8.400433665945665e-05, "loss": 0.0381, "step": 59770 }, { "grad_norm": 0.3885623514652252, "learning_rate": 8.399827513926534e-05, "loss": 0.0348, "step": 59780 }, { "grad_norm": 0.7919995188713074, "learning_rate": 8.399221268957395e-05, "loss": 0.0709, "step": 59790 }, { "grad_norm": 1.130245327949524, "learning_rate": 8.398614931054825e-05, "loss": 0.0552, "step": 59800 }, { "grad_norm": 0.4506682753562927, "learning_rate": 8.398008500235397e-05, "loss": 0.0356, "step": 59810 }, { "grad_norm": 0.7948461771011353, "learning_rate": 8.397401976515695e-05, "loss": 0.0568, "step": 59820 }, { "grad_norm": 1.0150195360183716, "learning_rate": 8.396795359912298e-05, "loss": 0.0632, "step": 59830 }, { "grad_norm": 0.4910876452922821, "learning_rate": 8.396188650441793e-05, "loss": 0.0408, "step": 59840 }, { "grad_norm": 0.8225197792053223, "learning_rate": 8.395581848120766e-05, "loss": 0.048, "step": 59850 }, { "grad_norm": 0.6456068158149719, "learning_rate": 8.394974952965806e-05, "loss": 0.0624, "step": 59860 }, { "grad_norm": 0.5453260540962219, "learning_rate": 8.394367964993508e-05, "loss": 0.0413, "step": 59870 }, { "grad_norm": 0.6645123958587646, "learning_rate": 8.393760884220465e-05, "loss": 0.0506, "step": 59880 }, { "grad_norm": 0.6091613173484802, "learning_rate": 8.393153710663274e-05, "loss": 0.0541, "step": 59890 }, { "grad_norm": 0.48689037561416626, "learning_rate": 8.392546444338534e-05, "loss": 0.0344, "step": 59900 }, { "grad_norm": 0.6058265566825867, "learning_rate": 8.39193908526285e-05, "loss": 0.057, "step": 59910 }, { "grad_norm": 0.8771238327026367, "learning_rate": 8.391331633452824e-05, "loss": 0.0502, "step": 59920 }, { "grad_norm": 0.5220599174499512, "learning_rate": 8.390724088925068e-05, "loss": 0.0537, "step": 59930 }, { "grad_norm": 1.0461446046829224, "learning_rate": 8.390116451696188e-05, "loss": 0.0501, "step": 59940 }, { "grad_norm": 0.8598346710205078, "learning_rate": 8.389508721782797e-05, "loss": 0.0415, "step": 59950 }, { "grad_norm": 0.9247898459434509, "learning_rate": 8.388900899201511e-05, "loss": 0.0538, "step": 59960 }, { "grad_norm": 0.8863919377326965, "learning_rate": 8.388292983968948e-05, "loss": 0.0517, "step": 59970 }, { "grad_norm": 0.4545600116252899, "learning_rate": 8.387684976101728e-05, "loss": 0.0472, "step": 59980 }, { "grad_norm": 0.8358693718910217, "learning_rate": 8.387076875616473e-05, "loss": 0.0397, "step": 59990 }, { "grad_norm": 1.113781213760376, "learning_rate": 8.38646868252981e-05, "loss": 0.0517, "step": 60000 }, { "grad_norm": 0.8984826803207397, "learning_rate": 8.385860396858365e-05, "loss": 0.0458, "step": 60010 }, { "grad_norm": 0.7611017823219299, "learning_rate": 8.385252018618767e-05, "loss": 0.0518, "step": 60020 }, { "grad_norm": 0.5796512961387634, "learning_rate": 8.384643547827652e-05, "loss": 0.0514, "step": 60030 }, { "grad_norm": 0.7309781312942505, "learning_rate": 8.384034984501655e-05, "loss": 0.0422, "step": 60040 }, { "grad_norm": 0.7642314434051514, "learning_rate": 8.383426328657411e-05, "loss": 0.0438, "step": 60050 }, { "grad_norm": 0.578455924987793, "learning_rate": 8.382817580311562e-05, "loss": 0.0415, "step": 60060 }, { "grad_norm": 0.3884449601173401, "learning_rate": 8.382208739480752e-05, "loss": 0.0372, "step": 60070 }, { "grad_norm": 0.6107209324836731, "learning_rate": 8.381599806181625e-05, "loss": 0.0493, "step": 60080 }, { "grad_norm": 1.1438571214675903, "learning_rate": 8.380990780430831e-05, "loss": 0.0637, "step": 60090 }, { "grad_norm": 0.5665203928947449, "learning_rate": 8.380381662245019e-05, "loss": 0.0595, "step": 60100 }, { "grad_norm": 1.7741564512252808, "learning_rate": 8.379772451640841e-05, "loss": 0.0567, "step": 60110 }, { "grad_norm": 0.5672916173934937, "learning_rate": 8.379163148634956e-05, "loss": 0.0518, "step": 60120 }, { "grad_norm": 0.42968180775642395, "learning_rate": 8.378553753244019e-05, "loss": 0.0419, "step": 60130 }, { "grad_norm": 0.5930196642875671, "learning_rate": 8.377944265484691e-05, "loss": 0.0448, "step": 60140 }, { "grad_norm": 0.9060187339782715, "learning_rate": 8.377334685373636e-05, "loss": 0.0629, "step": 60150 }, { "grad_norm": 0.3362343907356262, "learning_rate": 8.376725012927522e-05, "loss": 0.0381, "step": 60160 }, { "grad_norm": 0.6990355849266052, "learning_rate": 8.376115248163012e-05, "loss": 0.0483, "step": 60170 }, { "grad_norm": 0.4623414874076843, "learning_rate": 8.37550539109678e-05, "loss": 0.0372, "step": 60180 }, { "grad_norm": 0.377036988735199, "learning_rate": 8.374895441745498e-05, "loss": 0.0419, "step": 60190 }, { "grad_norm": 0.39585167169570923, "learning_rate": 8.374285400125844e-05, "loss": 0.0504, "step": 60200 }, { "grad_norm": 0.44664856791496277, "learning_rate": 8.373675266254495e-05, "loss": 0.0546, "step": 60210 }, { "grad_norm": 0.6577928066253662, "learning_rate": 8.373065040148129e-05, "loss": 0.0551, "step": 60220 }, { "grad_norm": 0.487028032541275, "learning_rate": 8.372454721823435e-05, "loss": 0.0549, "step": 60230 }, { "grad_norm": 1.9279080629348755, "learning_rate": 8.371844311297094e-05, "loss": 0.0378, "step": 60240 }, { "grad_norm": 0.43966910243034363, "learning_rate": 8.371233808585796e-05, "loss": 0.0427, "step": 60250 }, { "grad_norm": 0.7133761644363403, "learning_rate": 8.370623213706234e-05, "loss": 0.0386, "step": 60260 }, { "grad_norm": 0.6642792224884033, "learning_rate": 8.3700125266751e-05, "loss": 0.0467, "step": 60270 }, { "grad_norm": 0.33393236994743347, "learning_rate": 8.369401747509088e-05, "loss": 0.0466, "step": 60280 }, { "grad_norm": 0.5100602507591248, "learning_rate": 8.368790876224898e-05, "loss": 0.0425, "step": 60290 }, { "grad_norm": 3.178260564804077, "learning_rate": 8.368179912839231e-05, "loss": 0.0371, "step": 60300 }, { "grad_norm": 0.46832191944122314, "learning_rate": 8.367568857368791e-05, "loss": 0.0378, "step": 60310 }, { "grad_norm": 0.41773155331611633, "learning_rate": 8.366957709830284e-05, "loss": 0.0383, "step": 60320 }, { "grad_norm": 0.9572724103927612, "learning_rate": 8.366346470240418e-05, "loss": 0.0516, "step": 60330 }, { "grad_norm": 0.4745147228240967, "learning_rate": 8.365735138615904e-05, "loss": 0.0506, "step": 60340 }, { "grad_norm": 0.5686249136924744, "learning_rate": 8.365123714973454e-05, "loss": 0.0413, "step": 60350 }, { "grad_norm": 0.6086007356643677, "learning_rate": 8.364512199329789e-05, "loss": 0.0389, "step": 60360 }, { "grad_norm": 0.37948355078697205, "learning_rate": 8.363900591701622e-05, "loss": 0.0351, "step": 60370 }, { "grad_norm": 1.8495677709579468, "learning_rate": 8.363288892105679e-05, "loss": 0.0494, "step": 60380 }, { "grad_norm": 0.6505844593048096, "learning_rate": 8.362677100558678e-05, "loss": 0.05, "step": 60390 }, { "grad_norm": 0.5422407984733582, "learning_rate": 8.362065217077351e-05, "loss": 0.0519, "step": 60400 }, { "grad_norm": 0.6310132741928101, "learning_rate": 8.361453241678423e-05, "loss": 0.0381, "step": 60410 }, { "grad_norm": 0.5128873586654663, "learning_rate": 8.360841174378626e-05, "loss": 0.0489, "step": 60420 }, { "grad_norm": 0.4172913134098053, "learning_rate": 8.360229015194693e-05, "loss": 0.0481, "step": 60430 }, { "grad_norm": 0.5981811285018921, "learning_rate": 8.359616764143363e-05, "loss": 0.0465, "step": 60440 }, { "grad_norm": 0.5763323307037354, "learning_rate": 8.359004421241374e-05, "loss": 0.0391, "step": 60450 }, { "grad_norm": 0.522071123123169, "learning_rate": 8.358391986505463e-05, "loss": 0.0407, "step": 60460 }, { "grad_norm": 0.6567674875259399, "learning_rate": 8.357779459952377e-05, "loss": 0.0396, "step": 60470 }, { "grad_norm": 0.47574764490127563, "learning_rate": 8.357166841598863e-05, "loss": 0.0495, "step": 60480 }, { "grad_norm": 0.5055723190307617, "learning_rate": 8.35655413146167e-05, "loss": 0.0336, "step": 60490 }, { "grad_norm": 0.40633493661880493, "learning_rate": 8.355941329557546e-05, "loss": 0.0383, "step": 60500 }, { "grad_norm": 0.4977414906024933, "learning_rate": 8.355328435903247e-05, "loss": 0.0389, "step": 60510 }, { "grad_norm": 0.4614315330982208, "learning_rate": 8.354715450515531e-05, "loss": 0.0334, "step": 60520 }, { "grad_norm": 0.5312139987945557, "learning_rate": 8.354102373411155e-05, "loss": 0.0552, "step": 60530 }, { "grad_norm": 0.7751870155334473, "learning_rate": 8.353489204606877e-05, "loss": 0.0437, "step": 60540 }, { "grad_norm": 0.6452052593231201, "learning_rate": 8.352875944119465e-05, "loss": 0.0463, "step": 60550 }, { "grad_norm": 0.9058330059051514, "learning_rate": 8.352262591965686e-05, "loss": 0.0733, "step": 60560 }, { "grad_norm": 0.9059296250343323, "learning_rate": 8.351649148162306e-05, "loss": 0.0454, "step": 60570 }, { "grad_norm": 0.5047587752342224, "learning_rate": 8.351035612726099e-05, "loss": 0.0574, "step": 60580 }, { "grad_norm": 0.8055457472801208, "learning_rate": 8.350421985673836e-05, "loss": 0.0401, "step": 60590 }, { "grad_norm": 0.6194103956222534, "learning_rate": 8.349808267022293e-05, "loss": 0.0515, "step": 60600 }, { "grad_norm": 0.6180238723754883, "learning_rate": 8.349194456788252e-05, "loss": 0.0805, "step": 60610 }, { "grad_norm": 0.7524135112762451, "learning_rate": 8.348580554988493e-05, "loss": 0.036, "step": 60620 }, { "grad_norm": 0.6871169209480286, "learning_rate": 8.3479665616398e-05, "loss": 0.061, "step": 60630 }, { "grad_norm": 0.5635843873023987, "learning_rate": 8.347352476758959e-05, "loss": 0.0553, "step": 60640 }, { "grad_norm": 0.7168835997581482, "learning_rate": 8.346738300362757e-05, "loss": 0.0388, "step": 60650 }, { "grad_norm": 0.5801082253456116, "learning_rate": 8.346124032467988e-05, "loss": 0.0497, "step": 60660 }, { "grad_norm": 0.8434810042381287, "learning_rate": 8.345509673091446e-05, "loss": 0.0555, "step": 60670 }, { "grad_norm": 0.39912572503089905, "learning_rate": 8.344895222249926e-05, "loss": 0.0362, "step": 60680 }, { "grad_norm": 0.551259458065033, "learning_rate": 8.344280679960228e-05, "loss": 0.0369, "step": 60690 }, { "grad_norm": 0.5090423822402954, "learning_rate": 8.343666046239152e-05, "loss": 0.0439, "step": 60700 }, { "grad_norm": 0.6579388380050659, "learning_rate": 8.343051321103501e-05, "loss": 0.0408, "step": 60710 }, { "grad_norm": 0.43362659215927124, "learning_rate": 8.342436504570084e-05, "loss": 0.0549, "step": 60720 }, { "grad_norm": 0.6245046257972717, "learning_rate": 8.341821596655708e-05, "loss": 0.0564, "step": 60730 }, { "grad_norm": 0.48540356755256653, "learning_rate": 8.341206597377186e-05, "loss": 0.0454, "step": 60740 }, { "grad_norm": 0.6512428522109985, "learning_rate": 8.34059150675133e-05, "loss": 0.0423, "step": 60750 }, { "grad_norm": 0.7542429566383362, "learning_rate": 8.339976324794958e-05, "loss": 0.0492, "step": 60760 }, { "grad_norm": 0.7804288864135742, "learning_rate": 8.339361051524888e-05, "loss": 0.0481, "step": 60770 }, { "grad_norm": 0.581548273563385, "learning_rate": 8.338745686957942e-05, "loss": 0.043, "step": 60780 }, { "grad_norm": 0.41168567538261414, "learning_rate": 8.338130231110941e-05, "loss": 0.0488, "step": 60790 }, { "grad_norm": 0.6425190567970276, "learning_rate": 8.337514684000716e-05, "loss": 0.0541, "step": 60800 }, { "grad_norm": 0.4697631299495697, "learning_rate": 8.336899045644091e-05, "loss": 0.0401, "step": 60810 }, { "grad_norm": 0.4305722117424011, "learning_rate": 8.336283316057903e-05, "loss": 0.0379, "step": 60820 }, { "grad_norm": 0.5445729494094849, "learning_rate": 8.33566749525898e-05, "loss": 0.0396, "step": 60830 }, { "grad_norm": 0.3636666238307953, "learning_rate": 8.335051583264161e-05, "loss": 0.0455, "step": 60840 }, { "grad_norm": 1.0429840087890625, "learning_rate": 8.334435580090285e-05, "loss": 0.0688, "step": 60850 }, { "grad_norm": 3.8721392154693604, "learning_rate": 8.333819485754193e-05, "loss": 0.0562, "step": 60860 }, { "grad_norm": 0.41807040572166443, "learning_rate": 8.333203300272728e-05, "loss": 0.0397, "step": 60870 }, { "grad_norm": 0.893079936504364, "learning_rate": 8.332587023662738e-05, "loss": 0.0449, "step": 60880 }, { "grad_norm": 0.471301406621933, "learning_rate": 8.33197065594107e-05, "loss": 0.0485, "step": 60890 }, { "grad_norm": 0.5516102313995361, "learning_rate": 8.331354197124576e-05, "loss": 0.0389, "step": 60900 }, { "grad_norm": 0.7537034153938293, "learning_rate": 8.330737647230111e-05, "loss": 0.0399, "step": 60910 }, { "grad_norm": 0.5193332433700562, "learning_rate": 8.33012100627453e-05, "loss": 0.0434, "step": 60920 }, { "grad_norm": 0.6852230429649353, "learning_rate": 8.32950427427469e-05, "loss": 0.0457, "step": 60930 }, { "grad_norm": 0.7761569023132324, "learning_rate": 8.328887451247455e-05, "loss": 0.0515, "step": 60940 }, { "grad_norm": 0.3958706557750702, "learning_rate": 8.328270537209688e-05, "loss": 0.0316, "step": 60950 }, { "grad_norm": 0.6314911842346191, "learning_rate": 8.327653532178254e-05, "loss": 0.0471, "step": 60960 }, { "grad_norm": 0.7015140056610107, "learning_rate": 8.327036436170024e-05, "loss": 0.0439, "step": 60970 }, { "grad_norm": 0.7255399227142334, "learning_rate": 8.326419249201867e-05, "loss": 0.0385, "step": 60980 }, { "grad_norm": 0.8246590495109558, "learning_rate": 8.325801971290657e-05, "loss": 0.0563, "step": 60990 }, { "grad_norm": 0.5774441361427307, "learning_rate": 8.32518460245327e-05, "loss": 0.0516, "step": 61000 }, { "grad_norm": 0.6493551135063171, "learning_rate": 8.324567142706586e-05, "loss": 0.0381, "step": 61010 }, { "grad_norm": 0.7522991895675659, "learning_rate": 8.323949592067485e-05, "loss": 0.0559, "step": 61020 }, { "grad_norm": 0.5957107543945312, "learning_rate": 8.323331950552851e-05, "loss": 0.0381, "step": 61030 }, { "grad_norm": 0.5569236874580383, "learning_rate": 8.322714218179571e-05, "loss": 0.0341, "step": 61040 }, { "grad_norm": 0.5038490295410156, "learning_rate": 8.322096394964532e-05, "loss": 0.0441, "step": 61050 }, { "grad_norm": 0.6778421401977539, "learning_rate": 8.321478480924624e-05, "loss": 0.0439, "step": 61060 }, { "grad_norm": 0.406046062707901, "learning_rate": 8.320860476076745e-05, "loss": 0.0403, "step": 61070 }, { "grad_norm": 0.7134881019592285, "learning_rate": 8.320242380437786e-05, "loss": 0.0588, "step": 61080 }, { "grad_norm": 0.85173100233078, "learning_rate": 8.319624194024649e-05, "loss": 0.0399, "step": 61090 }, { "grad_norm": 0.5361495018005371, "learning_rate": 8.319005916854232e-05, "loss": 0.0421, "step": 61100 }, { "grad_norm": 0.5989623665809631, "learning_rate": 8.318387548943441e-05, "loss": 0.036, "step": 61110 }, { "grad_norm": 0.585288405418396, "learning_rate": 8.317769090309182e-05, "loss": 0.0387, "step": 61120 }, { "grad_norm": 0.512328565120697, "learning_rate": 8.317150540968362e-05, "loss": 0.0551, "step": 61130 }, { "grad_norm": 0.35454803705215454, "learning_rate": 8.316531900937891e-05, "loss": 0.0504, "step": 61140 }, { "grad_norm": 0.9038816094398499, "learning_rate": 8.315913170234686e-05, "loss": 0.0573, "step": 61150 }, { "grad_norm": 0.493990421295166, "learning_rate": 8.31529434887566e-05, "loss": 0.031, "step": 61160 }, { "grad_norm": 0.41510576009750366, "learning_rate": 8.31467543687773e-05, "loss": 0.0413, "step": 61170 }, { "grad_norm": 0.5211282968521118, "learning_rate": 8.314056434257822e-05, "loss": 0.045, "step": 61180 }, { "grad_norm": 0.7843460440635681, "learning_rate": 8.313437341032853e-05, "loss": 0.0581, "step": 61190 }, { "grad_norm": 0.31456366181373596, "learning_rate": 8.312818157219755e-05, "loss": 0.0406, "step": 61200 }, { "grad_norm": 0.32521700859069824, "learning_rate": 8.312198882835452e-05, "loss": 0.0393, "step": 61210 }, { "grad_norm": 0.7120251059532166, "learning_rate": 8.311579517896875e-05, "loss": 0.0398, "step": 61220 }, { "grad_norm": 1.095576524734497, "learning_rate": 8.310960062420958e-05, "loss": 0.048, "step": 61230 }, { "grad_norm": 0.4329740107059479, "learning_rate": 8.310340516424637e-05, "loss": 0.0468, "step": 61240 }, { "grad_norm": 0.5307934880256653, "learning_rate": 8.30972087992485e-05, "loss": 0.0355, "step": 61250 }, { "grad_norm": 0.8491231203079224, "learning_rate": 8.309101152938538e-05, "loss": 0.0337, "step": 61260 }, { "grad_norm": 0.5011705756187439, "learning_rate": 8.308481335482644e-05, "loss": 0.0344, "step": 61270 }, { "grad_norm": 0.4461180865764618, "learning_rate": 8.307861427574113e-05, "loss": 0.0389, "step": 61280 }, { "grad_norm": 0.5160120725631714, "learning_rate": 8.307241429229893e-05, "loss": 0.0499, "step": 61290 }, { "grad_norm": 0.7046559453010559, "learning_rate": 8.306621340466934e-05, "loss": 0.0381, "step": 61300 }, { "grad_norm": 0.45245999097824097, "learning_rate": 8.30600116130219e-05, "loss": 0.0467, "step": 61310 }, { "grad_norm": 0.6878654360771179, "learning_rate": 8.305380891752616e-05, "loss": 0.0366, "step": 61320 }, { "grad_norm": 0.7067344784736633, "learning_rate": 8.304760531835172e-05, "loss": 0.0407, "step": 61330 }, { "grad_norm": 0.4962356686592102, "learning_rate": 8.304140081566815e-05, "loss": 0.039, "step": 61340 }, { "grad_norm": 0.4568948447704315, "learning_rate": 8.303519540964511e-05, "loss": 0.0337, "step": 61350 }, { "grad_norm": 0.521121621131897, "learning_rate": 8.302898910045222e-05, "loss": 0.0446, "step": 61360 }, { "grad_norm": 0.3191993832588196, "learning_rate": 8.302278188825918e-05, "loss": 0.0418, "step": 61370 }, { "grad_norm": 0.43921709060668945, "learning_rate": 8.30165737732357e-05, "loss": 0.042, "step": 61380 }, { "grad_norm": 0.4894377887248993, "learning_rate": 8.30103647555515e-05, "loss": 0.0449, "step": 61390 }, { "grad_norm": 0.4952895939350128, "learning_rate": 8.300415483537631e-05, "loss": 0.0564, "step": 61400 }, { "grad_norm": 0.45794665813446045, "learning_rate": 8.299794401287994e-05, "loss": 0.0424, "step": 61410 }, { "grad_norm": 0.3870500922203064, "learning_rate": 8.299173228823219e-05, "loss": 0.0459, "step": 61420 }, { "grad_norm": 0.2864238917827606, "learning_rate": 8.298551966160284e-05, "loss": 0.0372, "step": 61430 }, { "grad_norm": 1.2042497396469116, "learning_rate": 8.297930613316178e-05, "loss": 0.0641, "step": 61440 }, { "grad_norm": 0.5248977541923523, "learning_rate": 8.29730917030789e-05, "loss": 0.0471, "step": 61450 }, { "grad_norm": 0.7634519934654236, "learning_rate": 8.296687637152407e-05, "loss": 0.0645, "step": 61460 }, { "grad_norm": 0.6540836691856384, "learning_rate": 8.296066013866723e-05, "loss": 0.0454, "step": 61470 }, { "grad_norm": 0.4713396430015564, "learning_rate": 8.295444300467832e-05, "loss": 0.0453, "step": 61480 }, { "grad_norm": 0.4973621368408203, "learning_rate": 8.294822496972732e-05, "loss": 0.0469, "step": 61490 }, { "grad_norm": 0.4210458993911743, "learning_rate": 8.294200603398423e-05, "loss": 0.042, "step": 61500 }, { "grad_norm": 0.6430892944335938, "learning_rate": 8.293578619761906e-05, "loss": 0.0411, "step": 61510 }, { "grad_norm": 0.6244375109672546, "learning_rate": 8.292956546080189e-05, "loss": 0.05, "step": 61520 }, { "grad_norm": 0.41766849160194397, "learning_rate": 8.292334382370275e-05, "loss": 0.0352, "step": 61530 }, { "grad_norm": 0.6614400148391724, "learning_rate": 8.291712128649177e-05, "loss": 0.0404, "step": 61540 }, { "grad_norm": 0.5374654531478882, "learning_rate": 8.291089784933906e-05, "loss": 0.0433, "step": 61550 }, { "grad_norm": 0.3872773051261902, "learning_rate": 8.290467351241475e-05, "loss": 0.0461, "step": 61560 }, { "grad_norm": 0.5138658285140991, "learning_rate": 8.289844827588906e-05, "loss": 0.0443, "step": 61570 }, { "grad_norm": 1.5798776149749756, "learning_rate": 8.289222213993213e-05, "loss": 0.0474, "step": 61580 }, { "grad_norm": 0.7323685884475708, "learning_rate": 8.288599510471421e-05, "loss": 0.0353, "step": 61590 }, { "grad_norm": 1.4589110612869263, "learning_rate": 8.287976717040553e-05, "loss": 0.0573, "step": 61600 }, { "grad_norm": 0.44389620423316956, "learning_rate": 8.28735383371764e-05, "loss": 0.0453, "step": 61610 }, { "grad_norm": 1.2127329111099243, "learning_rate": 8.286730860519704e-05, "loss": 0.0498, "step": 61620 }, { "grad_norm": 0.42202675342559814, "learning_rate": 8.286107797463782e-05, "loss": 0.0389, "step": 61630 }, { "grad_norm": 0.457028865814209, "learning_rate": 8.285484644566909e-05, "loss": 0.045, "step": 61640 }, { "grad_norm": 0.720447301864624, "learning_rate": 8.284861401846119e-05, "loss": 0.0337, "step": 61650 }, { "grad_norm": 0.6210343837738037, "learning_rate": 8.284238069318453e-05, "loss": 0.0723, "step": 61660 }, { "grad_norm": 0.48900726437568665, "learning_rate": 8.283614647000952e-05, "loss": 0.0458, "step": 61670 }, { "grad_norm": 0.8338252902030945, "learning_rate": 8.282991134910658e-05, "loss": 0.0477, "step": 61680 }, { "grad_norm": 0.40231746435165405, "learning_rate": 8.282367533064624e-05, "loss": 0.0357, "step": 61690 }, { "grad_norm": 0.8695432543754578, "learning_rate": 8.281743841479891e-05, "loss": 0.0422, "step": 61700 }, { "grad_norm": 0.7969686985015869, "learning_rate": 8.281120060173515e-05, "loss": 0.049, "step": 61710 }, { "grad_norm": 0.28518301248550415, "learning_rate": 8.28049618916255e-05, "loss": 0.034, "step": 61720 }, { "grad_norm": 0.31806856393814087, "learning_rate": 8.279872228464052e-05, "loss": 0.0373, "step": 61730 }, { "grad_norm": 0.5244483351707458, "learning_rate": 8.279248178095078e-05, "loss": 0.0412, "step": 61740 }, { "grad_norm": 0.346306174993515, "learning_rate": 8.278624038072692e-05, "loss": 0.0464, "step": 61750 }, { "grad_norm": 0.4130052924156189, "learning_rate": 8.277999808413956e-05, "loss": 0.0493, "step": 61760 }, { "grad_norm": 0.5622959136962891, "learning_rate": 8.277375489135937e-05, "loss": 0.051, "step": 61770 }, { "grad_norm": 0.7554733157157898, "learning_rate": 8.276751080255703e-05, "loss": 0.0396, "step": 61780 }, { "grad_norm": 0.5211220979690552, "learning_rate": 8.276126581790327e-05, "loss": 0.0712, "step": 61790 }, { "grad_norm": 1.0075563192367554, "learning_rate": 8.275501993756878e-05, "loss": 0.056, "step": 61800 }, { "grad_norm": 1.1025786399841309, "learning_rate": 8.274877316172438e-05, "loss": 0.0632, "step": 61810 }, { "grad_norm": 0.7072575092315674, "learning_rate": 8.27425254905408e-05, "loss": 0.0452, "step": 61820 }, { "grad_norm": 0.5418517589569092, "learning_rate": 8.273627692418889e-05, "loss": 0.0458, "step": 61830 }, { "grad_norm": 0.2745772898197174, "learning_rate": 8.273002746283946e-05, "loss": 0.0483, "step": 61840 }, { "grad_norm": 0.6064556837081909, "learning_rate": 8.272377710666339e-05, "loss": 0.0376, "step": 61850 }, { "grad_norm": 0.4027990400791168, "learning_rate": 8.271752585583155e-05, "loss": 0.0364, "step": 61860 }, { "grad_norm": 0.46078792214393616, "learning_rate": 8.271127371051481e-05, "loss": 0.0361, "step": 61870 }, { "grad_norm": 0.46136194467544556, "learning_rate": 8.270502067088417e-05, "loss": 0.0505, "step": 61880 }, { "grad_norm": 0.555871307849884, "learning_rate": 8.269876673711053e-05, "loss": 0.0492, "step": 61890 }, { "grad_norm": 0.5800353288650513, "learning_rate": 8.269251190936491e-05, "loss": 0.0339, "step": 61900 }, { "grad_norm": 0.45943236351013184, "learning_rate": 8.268625618781829e-05, "loss": 0.0358, "step": 61910 }, { "grad_norm": 0.4104691445827484, "learning_rate": 8.26799995726417e-05, "loss": 0.0387, "step": 61920 }, { "grad_norm": 0.3375558853149414, "learning_rate": 8.26737420640062e-05, "loss": 0.0405, "step": 61930 }, { "grad_norm": 0.9413309693336487, "learning_rate": 8.266748366208285e-05, "loss": 0.0461, "step": 61940 }, { "grad_norm": 0.45305439829826355, "learning_rate": 8.26612243670428e-05, "loss": 0.0415, "step": 61950 }, { "grad_norm": 0.7776730060577393, "learning_rate": 8.265496417905714e-05, "loss": 0.0409, "step": 61960 }, { "grad_norm": 0.5118582248687744, "learning_rate": 8.264870309829702e-05, "loss": 0.0568, "step": 61970 }, { "grad_norm": 0.7582716345787048, "learning_rate": 8.264244112493362e-05, "loss": 0.0563, "step": 61980 }, { "grad_norm": 0.6153548955917358, "learning_rate": 8.263617825913814e-05, "loss": 0.0473, "step": 61990 }, { "grad_norm": 0.5999431610107422, "learning_rate": 8.262991450108182e-05, "loss": 0.0479, "step": 62000 }, { "grad_norm": 0.4296496510505676, "learning_rate": 8.262364985093589e-05, "loss": 0.039, "step": 62010 }, { "grad_norm": 0.7862755060195923, "learning_rate": 8.261738430887163e-05, "loss": 0.0459, "step": 62020 }, { "grad_norm": 0.5977700352668762, "learning_rate": 8.261111787506032e-05, "loss": 0.0371, "step": 62030 }, { "grad_norm": 0.7243539690971375, "learning_rate": 8.260485054967333e-05, "loss": 0.047, "step": 62040 }, { "grad_norm": 0.7978638410568237, "learning_rate": 8.259858233288196e-05, "loss": 0.0386, "step": 62050 }, { "grad_norm": 0.37551116943359375, "learning_rate": 8.259231322485759e-05, "loss": 0.0587, "step": 62060 }, { "grad_norm": 0.5907440185546875, "learning_rate": 8.258604322577163e-05, "loss": 0.047, "step": 62070 }, { "grad_norm": 0.657371461391449, "learning_rate": 8.257977233579548e-05, "loss": 0.0455, "step": 62080 }, { "grad_norm": 0.48640674352645874, "learning_rate": 8.257350055510061e-05, "loss": 0.0375, "step": 62090 }, { "grad_norm": 0.5216614603996277, "learning_rate": 8.256722788385847e-05, "loss": 0.0492, "step": 62100 }, { "grad_norm": 0.3764597475528717, "learning_rate": 8.256095432224053e-05, "loss": 0.0407, "step": 62110 }, { "grad_norm": 0.4293656349182129, "learning_rate": 8.255467987041835e-05, "loss": 0.0395, "step": 62120 }, { "grad_norm": 0.7773244976997375, "learning_rate": 8.254840452856345e-05, "loss": 0.0387, "step": 62130 }, { "grad_norm": 0.7286482453346252, "learning_rate": 8.25421282968474e-05, "loss": 0.0331, "step": 62140 }, { "grad_norm": 0.5344473719596863, "learning_rate": 8.25358511754418e-05, "loss": 0.0338, "step": 62150 }, { "grad_norm": 0.7937968373298645, "learning_rate": 8.252957316451824e-05, "loss": 0.0421, "step": 62160 }, { "grad_norm": 0.4952155649662018, "learning_rate": 8.252329426424837e-05, "loss": 0.0685, "step": 62170 }, { "grad_norm": 0.6306443214416504, "learning_rate": 8.251701447480384e-05, "loss": 0.0325, "step": 62180 }, { "grad_norm": 0.4977025091648102, "learning_rate": 8.251073379635637e-05, "loss": 0.0437, "step": 62190 }, { "grad_norm": 0.5758088827133179, "learning_rate": 8.250445222907764e-05, "loss": 0.043, "step": 62200 }, { "grad_norm": 0.9691364169120789, "learning_rate": 8.24981697731394e-05, "loss": 0.0551, "step": 62210 }, { "grad_norm": 1.5419957637786865, "learning_rate": 8.249188642871341e-05, "loss": 0.0514, "step": 62220 }, { "grad_norm": 1.5737115144729614, "learning_rate": 8.248560219597145e-05, "loss": 0.0437, "step": 62230 }, { "grad_norm": 0.5937957167625427, "learning_rate": 8.247931707508534e-05, "loss": 0.056, "step": 62240 }, { "grad_norm": 0.38464513421058655, "learning_rate": 8.24730310662269e-05, "loss": 0.0385, "step": 62250 }, { "grad_norm": 1.2086693048477173, "learning_rate": 8.246674416956797e-05, "loss": 0.0496, "step": 62260 }, { "grad_norm": 0.6834169626235962, "learning_rate": 8.246045638528047e-05, "loss": 0.0439, "step": 62270 }, { "grad_norm": 0.644884467124939, "learning_rate": 8.245416771353629e-05, "loss": 0.042, "step": 62280 }, { "grad_norm": 0.42315223813056946, "learning_rate": 8.244787815450735e-05, "loss": 0.0411, "step": 62290 }, { "grad_norm": 0.41647234559059143, "learning_rate": 8.244158770836561e-05, "loss": 0.0491, "step": 62300 }, { "grad_norm": 0.6048988103866577, "learning_rate": 8.243529637528306e-05, "loss": 0.0429, "step": 62310 }, { "grad_norm": 0.5421926975250244, "learning_rate": 8.242900415543167e-05, "loss": 0.0385, "step": 62320 }, { "grad_norm": 0.6242005228996277, "learning_rate": 8.242271104898352e-05, "loss": 0.0409, "step": 62330 }, { "grad_norm": 0.5448681116104126, "learning_rate": 8.241641705611061e-05, "loss": 0.0454, "step": 62340 }, { "grad_norm": 0.41155877709388733, "learning_rate": 8.241012217698503e-05, "loss": 0.0557, "step": 62350 }, { "grad_norm": 0.42059794068336487, "learning_rate": 8.240382641177891e-05, "loss": 0.0498, "step": 62360 }, { "grad_norm": 0.4287986159324646, "learning_rate": 8.239752976066432e-05, "loss": 0.0305, "step": 62370 }, { "grad_norm": 1.984868049621582, "learning_rate": 8.239123222381347e-05, "loss": 0.0522, "step": 62380 }, { "grad_norm": 0.6184206604957581, "learning_rate": 8.238493380139847e-05, "loss": 0.0306, "step": 62390 }, { "grad_norm": 0.4121311902999878, "learning_rate": 8.237863449359157e-05, "loss": 0.039, "step": 62400 }, { "grad_norm": 0.6272717714309692, "learning_rate": 8.237233430056495e-05, "loss": 0.0414, "step": 62410 }, { "grad_norm": 0.5546238422393799, "learning_rate": 8.236603322249088e-05, "loss": 0.0442, "step": 62420 }, { "grad_norm": 0.8322944045066833, "learning_rate": 8.235973125954162e-05, "loss": 0.0447, "step": 62430 }, { "grad_norm": 0.82122403383255, "learning_rate": 8.235342841188946e-05, "loss": 0.0492, "step": 62440 }, { "grad_norm": 0.9428316354751587, "learning_rate": 8.234712467970673e-05, "loss": 0.0529, "step": 62450 }, { "grad_norm": 0.6039222478866577, "learning_rate": 8.234082006316576e-05, "loss": 0.0374, "step": 62460 }, { "grad_norm": 0.9829263091087341, "learning_rate": 8.233451456243891e-05, "loss": 0.0539, "step": 62470 }, { "grad_norm": 0.6265493035316467, "learning_rate": 8.232820817769859e-05, "loss": 0.0393, "step": 62480 }, { "grad_norm": 0.6578014492988586, "learning_rate": 8.23219009091172e-05, "loss": 0.0392, "step": 62490 }, { "grad_norm": 0.3446427583694458, "learning_rate": 8.231559275686719e-05, "loss": 0.0495, "step": 62500 }, { "grad_norm": 0.5830228924751282, "learning_rate": 8.2309283721121e-05, "loss": 0.0561, "step": 62510 }, { "grad_norm": 0.8909366726875305, "learning_rate": 8.230297380205115e-05, "loss": 0.0871, "step": 62520 }, { "grad_norm": 0.5687381625175476, "learning_rate": 8.229666299983011e-05, "loss": 0.0453, "step": 62530 }, { "grad_norm": 0.724830687046051, "learning_rate": 8.229035131463045e-05, "loss": 0.0439, "step": 62540 }, { "grad_norm": 0.6404663324356079, "learning_rate": 8.228403874662471e-05, "loss": 0.0433, "step": 62550 }, { "grad_norm": 0.47781506180763245, "learning_rate": 8.22777252959855e-05, "loss": 0.0782, "step": 62560 }, { "grad_norm": 0.6630685329437256, "learning_rate": 8.227141096288537e-05, "loss": 0.0427, "step": 62570 }, { "grad_norm": 1.907402515411377, "learning_rate": 8.226509574749701e-05, "loss": 0.0576, "step": 62580 }, { "grad_norm": 0.47336500883102417, "learning_rate": 8.225877964999304e-05, "loss": 0.0463, "step": 62590 }, { "grad_norm": 0.4219774305820465, "learning_rate": 8.225246267054619e-05, "loss": 0.0654, "step": 62600 }, { "grad_norm": 0.6018417477607727, "learning_rate": 8.224614480932909e-05, "loss": 0.0751, "step": 62610 }, { "grad_norm": 0.47508367896080017, "learning_rate": 8.223982606651452e-05, "loss": 0.0556, "step": 62620 }, { "grad_norm": 0.6810644268989563, "learning_rate": 8.223350644227521e-05, "loss": 0.0435, "step": 62630 }, { "grad_norm": 0.5021542310714722, "learning_rate": 8.222718593678395e-05, "loss": 0.0377, "step": 62640 }, { "grad_norm": 0.49785807728767395, "learning_rate": 8.222086455021353e-05, "loss": 0.0348, "step": 62650 }, { "grad_norm": 0.4394710958003998, "learning_rate": 8.221454228273678e-05, "loss": 0.036, "step": 62660 }, { "grad_norm": 2.7070298194885254, "learning_rate": 8.220821913452655e-05, "loss": 0.0406, "step": 62670 }, { "grad_norm": 0.4453306496143341, "learning_rate": 8.22018951057557e-05, "loss": 0.0382, "step": 62680 }, { "grad_norm": 0.40318772196769714, "learning_rate": 8.219557019659714e-05, "loss": 0.0444, "step": 62690 }, { "grad_norm": 0.7949725389480591, "learning_rate": 8.218924440722378e-05, "loss": 0.0507, "step": 62700 }, { "grad_norm": 0.5651993751525879, "learning_rate": 8.218291773780858e-05, "loss": 0.0358, "step": 62710 }, { "grad_norm": 1.0849452018737793, "learning_rate": 8.217659018852451e-05, "loss": 0.0631, "step": 62720 }, { "grad_norm": 0.47324344515800476, "learning_rate": 8.217026175954454e-05, "loss": 0.0381, "step": 62730 }, { "grad_norm": 0.5809100866317749, "learning_rate": 8.216393245104169e-05, "loss": 0.0546, "step": 62740 }, { "grad_norm": 0.3777327239513397, "learning_rate": 8.215760226318902e-05, "loss": 0.0488, "step": 62750 }, { "grad_norm": 0.7081065773963928, "learning_rate": 8.215127119615958e-05, "loss": 0.0413, "step": 62760 }, { "grad_norm": 0.6381102800369263, "learning_rate": 8.214493925012648e-05, "loss": 0.0414, "step": 62770 }, { "grad_norm": 0.7065283060073853, "learning_rate": 8.213860642526278e-05, "loss": 0.0433, "step": 62780 }, { "grad_norm": 0.3882334232330322, "learning_rate": 8.21322727217417e-05, "loss": 0.0419, "step": 62790 }, { "grad_norm": 0.5042099356651306, "learning_rate": 8.212593813973632e-05, "loss": 0.0412, "step": 62800 }, { "grad_norm": 0.6223212480545044, "learning_rate": 8.21196026794199e-05, "loss": 0.0442, "step": 62810 }, { "grad_norm": 3.8089599609375, "learning_rate": 8.211326634096556e-05, "loss": 0.0412, "step": 62820 }, { "grad_norm": 0.7921205163002014, "learning_rate": 8.21069291245466e-05, "loss": 0.0379, "step": 62830 }, { "grad_norm": 0.5140101313591003, "learning_rate": 8.210059103033629e-05, "loss": 0.0443, "step": 62840 }, { "grad_norm": 0.7725275158882141, "learning_rate": 8.209425205850785e-05, "loss": 0.0439, "step": 62850 }, { "grad_norm": 0.6351220607757568, "learning_rate": 8.208791220923462e-05, "loss": 0.0419, "step": 62860 }, { "grad_norm": 0.9192352890968323, "learning_rate": 8.208157148268994e-05, "loss": 0.0389, "step": 62870 }, { "grad_norm": 0.7747792601585388, "learning_rate": 8.207522987904714e-05, "loss": 0.046, "step": 62880 }, { "grad_norm": 0.6013110280036926, "learning_rate": 8.20688873984796e-05, "loss": 0.041, "step": 62890 }, { "grad_norm": 0.2627648413181305, "learning_rate": 8.206254404116072e-05, "loss": 0.042, "step": 62900 }, { "grad_norm": 0.7956007122993469, "learning_rate": 8.205619980726394e-05, "loss": 0.0432, "step": 62910 }, { "grad_norm": 0.655699610710144, "learning_rate": 8.204985469696271e-05, "loss": 0.045, "step": 62920 }, { "grad_norm": 0.4520421326160431, "learning_rate": 8.204350871043047e-05, "loss": 0.0358, "step": 62930 }, { "grad_norm": 0.4656843841075897, "learning_rate": 8.203716184784075e-05, "loss": 0.0412, "step": 62940 }, { "grad_norm": 0.4705468416213989, "learning_rate": 8.203081410936708e-05, "loss": 0.0484, "step": 62950 }, { "grad_norm": 0.4028678834438324, "learning_rate": 8.202446549518297e-05, "loss": 0.0597, "step": 62960 }, { "grad_norm": 0.43951696157455444, "learning_rate": 8.201811600546199e-05, "loss": 0.0542, "step": 62970 }, { "grad_norm": 0.3942675292491913, "learning_rate": 8.201176564037777e-05, "loss": 0.0423, "step": 62980 }, { "grad_norm": 0.7792965769767761, "learning_rate": 8.20054144001039e-05, "loss": 0.0627, "step": 62990 }, { "grad_norm": 0.3743533194065094, "learning_rate": 8.199906228481402e-05, "loss": 0.0375, "step": 63000 }, { "grad_norm": 1.339630126953125, "learning_rate": 8.19927092946818e-05, "loss": 0.0386, "step": 63010 }, { "grad_norm": 0.7586463689804077, "learning_rate": 8.198635542988092e-05, "loss": 0.0557, "step": 63020 }, { "grad_norm": 0.765599250793457, "learning_rate": 8.198000069058509e-05, "loss": 0.035, "step": 63030 }, { "grad_norm": 0.7185304760932922, "learning_rate": 8.197364507696808e-05, "loss": 0.043, "step": 63040 }, { "grad_norm": 0.8468648791313171, "learning_rate": 8.196728858920361e-05, "loss": 0.0538, "step": 63050 }, { "grad_norm": 0.713786780834198, "learning_rate": 8.196093122746548e-05, "loss": 0.0464, "step": 63060 }, { "grad_norm": 0.721365749835968, "learning_rate": 8.195457299192751e-05, "loss": 0.0448, "step": 63070 }, { "grad_norm": 0.7042579054832458, "learning_rate": 8.19482138827635e-05, "loss": 0.0408, "step": 63080 }, { "grad_norm": 0.500848114490509, "learning_rate": 8.194185390014735e-05, "loss": 0.0338, "step": 63090 }, { "grad_norm": 0.6049621105194092, "learning_rate": 8.19354930442529e-05, "loss": 0.0385, "step": 63100 }, { "grad_norm": 0.4950026273727417, "learning_rate": 8.192913131525406e-05, "loss": 0.0405, "step": 63110 }, { "grad_norm": 0.6938900947570801, "learning_rate": 8.192276871332479e-05, "loss": 0.0398, "step": 63120 }, { "grad_norm": 0.5774931907653809, "learning_rate": 8.1916405238639e-05, "loss": 0.0473, "step": 63130 }, { "grad_norm": 1.834922432899475, "learning_rate": 8.191004089137066e-05, "loss": 0.0468, "step": 63140 }, { "grad_norm": 0.6120239496231079, "learning_rate": 8.190367567169381e-05, "loss": 0.0325, "step": 63150 }, { "grad_norm": 2.3058102130889893, "learning_rate": 8.189730957978247e-05, "loss": 0.0409, "step": 63160 }, { "grad_norm": 0.4458787441253662, "learning_rate": 8.189094261581065e-05, "loss": 0.0332, "step": 63170 }, { "grad_norm": 0.6052412390708923, "learning_rate": 8.188457477995244e-05, "loss": 0.0429, "step": 63180 }, { "grad_norm": 0.7416747212409973, "learning_rate": 8.187820607238194e-05, "loss": 0.0302, "step": 63190 }, { "grad_norm": 0.40174224972724915, "learning_rate": 8.187183649327325e-05, "loss": 0.0336, "step": 63200 }, { "grad_norm": 0.5246707797050476, "learning_rate": 8.186546604280053e-05, "loss": 0.0407, "step": 63210 }, { "grad_norm": 0.7923892736434937, "learning_rate": 8.185909472113795e-05, "loss": 0.05, "step": 63220 }, { "grad_norm": 0.8400330543518066, "learning_rate": 8.185272252845969e-05, "loss": 0.0418, "step": 63230 }, { "grad_norm": 0.6354644298553467, "learning_rate": 8.184634946493994e-05, "loss": 0.044, "step": 63240 }, { "grad_norm": 0.6365333795547485, "learning_rate": 8.183997553075298e-05, "loss": 0.0471, "step": 63250 }, { "grad_norm": 0.6071587800979614, "learning_rate": 8.183360072607304e-05, "loss": 0.0331, "step": 63260 }, { "grad_norm": 0.6535688042640686, "learning_rate": 8.182722505107442e-05, "loss": 0.0464, "step": 63270 }, { "grad_norm": 0.5151100158691406, "learning_rate": 8.182084850593142e-05, "loss": 0.0608, "step": 63280 }, { "grad_norm": 0.5618229508399963, "learning_rate": 8.181447109081839e-05, "loss": 0.0416, "step": 63290 }, { "grad_norm": 0.4667903482913971, "learning_rate": 8.180809280590966e-05, "loss": 0.0354, "step": 63300 }, { "grad_norm": 0.9245591759681702, "learning_rate": 8.180171365137963e-05, "loss": 0.0458, "step": 63310 }, { "grad_norm": 0.3808915615081787, "learning_rate": 8.17953336274027e-05, "loss": 0.0293, "step": 63320 }, { "grad_norm": 0.5574115514755249, "learning_rate": 8.178895273415329e-05, "loss": 0.0394, "step": 63330 }, { "grad_norm": 0.5545651912689209, "learning_rate": 8.178257097180583e-05, "loss": 0.0401, "step": 63340 }, { "grad_norm": 0.6539777517318726, "learning_rate": 8.177618834053484e-05, "loss": 0.0421, "step": 63350 }, { "grad_norm": 0.5247210264205933, "learning_rate": 8.17698048405148e-05, "loss": 0.0358, "step": 63360 }, { "grad_norm": 0.48502683639526367, "learning_rate": 8.176342047192023e-05, "loss": 0.0358, "step": 63370 }, { "grad_norm": 0.5906846523284912, "learning_rate": 8.175703523492568e-05, "loss": 0.0406, "step": 63380 }, { "grad_norm": 0.40114325284957886, "learning_rate": 8.175064912970571e-05, "loss": 0.0449, "step": 63390 }, { "grad_norm": 0.7870582342147827, "learning_rate": 8.174426215643493e-05, "loss": 0.0376, "step": 63400 }, { "grad_norm": 0.8803765177726746, "learning_rate": 8.173787431528794e-05, "loss": 0.0581, "step": 63410 }, { "grad_norm": 0.4918592870235443, "learning_rate": 8.17314856064394e-05, "loss": 0.0417, "step": 63420 }, { "grad_norm": 0.48959413170814514, "learning_rate": 8.172509603006395e-05, "loss": 0.0431, "step": 63430 }, { "grad_norm": 0.6710435152053833, "learning_rate": 8.17187055863363e-05, "loss": 0.0655, "step": 63440 }, { "grad_norm": 0.4087885916233063, "learning_rate": 8.171231427543115e-05, "loss": 0.0396, "step": 63450 }, { "grad_norm": 0.6182399392127991, "learning_rate": 8.170592209752326e-05, "loss": 0.0414, "step": 63460 }, { "grad_norm": 0.45415616035461426, "learning_rate": 8.169952905278734e-05, "loss": 0.0595, "step": 63470 }, { "grad_norm": 0.47039517760276794, "learning_rate": 8.169313514139823e-05, "loss": 0.0462, "step": 63480 }, { "grad_norm": 0.7979927659034729, "learning_rate": 8.16867403635307e-05, "loss": 0.0367, "step": 63490 }, { "grad_norm": 0.6347291469573975, "learning_rate": 8.16803447193596e-05, "loss": 0.0508, "step": 63500 }, { "grad_norm": 0.4217807948589325, "learning_rate": 8.167394820905977e-05, "loss": 0.053, "step": 63510 }, { "grad_norm": 0.5441542267799377, "learning_rate": 8.16675508328061e-05, "loss": 0.0361, "step": 63520 }, { "grad_norm": 0.5880966782569885, "learning_rate": 8.166115259077349e-05, "loss": 0.0544, "step": 63530 }, { "grad_norm": 0.5144261121749878, "learning_rate": 8.165475348313687e-05, "loss": 0.0383, "step": 63540 }, { "grad_norm": 0.5144681334495544, "learning_rate": 8.164835351007117e-05, "loss": 0.0502, "step": 63550 }, { "grad_norm": 0.7045838236808777, "learning_rate": 8.164195267175138e-05, "loss": 0.0517, "step": 63560 }, { "grad_norm": 0.9083784818649292, "learning_rate": 8.16355509683525e-05, "loss": 0.0565, "step": 63570 }, { "grad_norm": 0.4496736228466034, "learning_rate": 8.162914840004952e-05, "loss": 0.0634, "step": 63580 }, { "grad_norm": 0.4533114433288574, "learning_rate": 8.162274496701749e-05, "loss": 0.0435, "step": 63590 }, { "grad_norm": 0.712327241897583, "learning_rate": 8.161634066943153e-05, "loss": 0.0411, "step": 63600 }, { "grad_norm": 0.3142562508583069, "learning_rate": 8.160993550746668e-05, "loss": 0.0674, "step": 63610 }, { "grad_norm": 0.770427405834198, "learning_rate": 8.160352948129808e-05, "loss": 0.0469, "step": 63620 }, { "grad_norm": 0.5162129998207092, "learning_rate": 8.159712259110084e-05, "loss": 0.0439, "step": 63630 }, { "grad_norm": 0.5056813359260559, "learning_rate": 8.159071483705015e-05, "loss": 0.0374, "step": 63640 }, { "grad_norm": 0.4230304956436157, "learning_rate": 8.158430621932118e-05, "loss": 0.0425, "step": 63650 }, { "grad_norm": 0.3994000554084778, "learning_rate": 8.157789673808913e-05, "loss": 0.0532, "step": 63660 }, { "grad_norm": 0.5162056684494019, "learning_rate": 8.157148639352926e-05, "loss": 0.0591, "step": 63670 }, { "grad_norm": 0.5771387219429016, "learning_rate": 8.156507518581682e-05, "loss": 0.0361, "step": 63680 }, { "grad_norm": 0.5778725147247314, "learning_rate": 8.155866311512708e-05, "loss": 0.0386, "step": 63690 }, { "grad_norm": 1.4116064310073853, "learning_rate": 8.155225018163533e-05, "loss": 0.0586, "step": 63700 }, { "grad_norm": 0.46826621890068054, "learning_rate": 8.154583638551694e-05, "loss": 0.0421, "step": 63710 }, { "grad_norm": 0.732846736907959, "learning_rate": 8.153942172694721e-05, "loss": 0.036, "step": 63720 }, { "grad_norm": 0.6016547083854675, "learning_rate": 8.153300620610154e-05, "loss": 0.0358, "step": 63730 }, { "grad_norm": 0.6016938090324402, "learning_rate": 8.152658982315536e-05, "loss": 0.0477, "step": 63740 }, { "grad_norm": 0.4350077211856842, "learning_rate": 8.152017257828401e-05, "loss": 0.041, "step": 63750 }, { "grad_norm": 0.3899514377117157, "learning_rate": 8.151375447166302e-05, "loss": 0.0371, "step": 63760 }, { "grad_norm": 0.4682185649871826, "learning_rate": 8.150733550346781e-05, "loss": 0.0425, "step": 63770 }, { "grad_norm": 0.6856747269630432, "learning_rate": 8.150091567387389e-05, "loss": 0.0683, "step": 63780 }, { "grad_norm": 0.5197413563728333, "learning_rate": 8.149449498305674e-05, "loss": 0.0357, "step": 63790 }, { "grad_norm": 0.700742781162262, "learning_rate": 8.148807343119196e-05, "loss": 0.0368, "step": 63800 }, { "grad_norm": 0.6700085997581482, "learning_rate": 8.148165101845509e-05, "loss": 0.0449, "step": 63810 }, { "grad_norm": 0.6116388440132141, "learning_rate": 8.147522774502168e-05, "loss": 0.0649, "step": 63820 }, { "grad_norm": 0.49518513679504395, "learning_rate": 8.146880361106739e-05, "loss": 0.0405, "step": 63830 }, { "grad_norm": 0.6902034282684326, "learning_rate": 8.146237861676782e-05, "loss": 0.0708, "step": 63840 }, { "grad_norm": 0.4054378569126129, "learning_rate": 8.145595276229862e-05, "loss": 0.0398, "step": 63850 }, { "grad_norm": 0.35107702016830444, "learning_rate": 8.144952604783553e-05, "loss": 0.043, "step": 63860 }, { "grad_norm": 0.31282705068588257, "learning_rate": 8.144309847355418e-05, "loss": 0.0331, "step": 63870 }, { "grad_norm": 0.5357232093811035, "learning_rate": 8.143667003963034e-05, "loss": 0.0411, "step": 63880 }, { "grad_norm": 0.6485603451728821, "learning_rate": 8.143024074623975e-05, "loss": 0.0562, "step": 63890 }, { "grad_norm": 0.7889344692230225, "learning_rate": 8.14238105935582e-05, "loss": 0.0465, "step": 63900 }, { "grad_norm": 0.5387492179870605, "learning_rate": 8.141737958176146e-05, "loss": 0.0445, "step": 63910 }, { "grad_norm": 0.8222871422767639, "learning_rate": 8.141094771102537e-05, "loss": 0.0606, "step": 63920 }, { "grad_norm": 0.6959372758865356, "learning_rate": 8.140451498152578e-05, "loss": 0.0457, "step": 63930 }, { "grad_norm": 0.590292751789093, "learning_rate": 8.139808139343853e-05, "loss": 0.0502, "step": 63940 }, { "grad_norm": 0.4872193932533264, "learning_rate": 8.139164694693954e-05, "loss": 0.0365, "step": 63950 }, { "grad_norm": 0.4390363395214081, "learning_rate": 8.138521164220472e-05, "loss": 0.0454, "step": 63960 }, { "grad_norm": 0.9294240474700928, "learning_rate": 8.137877547941e-05, "loss": 0.0638, "step": 63970 }, { "grad_norm": 0.7505605816841125, "learning_rate": 8.137233845873133e-05, "loss": 0.0619, "step": 63980 }, { "grad_norm": 0.5215197205543518, "learning_rate": 8.136590058034472e-05, "loss": 0.0504, "step": 63990 }, { "grad_norm": 0.5823967456817627, "learning_rate": 8.135946184442619e-05, "loss": 0.0468, "step": 64000 }, { "grad_norm": 0.39943450689315796, "learning_rate": 8.135302225115175e-05, "loss": 0.0521, "step": 64010 }, { "grad_norm": 0.9443767070770264, "learning_rate": 8.134658180069745e-05, "loss": 0.0502, "step": 64020 }, { "grad_norm": 0.6012125015258789, "learning_rate": 8.134014049323939e-05, "loss": 0.0552, "step": 64030 }, { "grad_norm": 0.8906664252281189, "learning_rate": 8.133369832895364e-05, "loss": 0.0464, "step": 64040 }, { "grad_norm": 1.1148186922073364, "learning_rate": 8.132725530801638e-05, "loss": 0.0395, "step": 64050 }, { "grad_norm": 0.6822746396064758, "learning_rate": 8.13208114306037e-05, "loss": 0.0391, "step": 64060 }, { "grad_norm": 0.5945626497268677, "learning_rate": 8.131436669689183e-05, "loss": 0.0474, "step": 64070 }, { "grad_norm": 0.9280967712402344, "learning_rate": 8.130792110705692e-05, "loss": 0.0459, "step": 64080 }, { "grad_norm": 0.5071222186088562, "learning_rate": 8.130147466127522e-05, "loss": 0.0534, "step": 64090 }, { "grad_norm": 0.8940619826316833, "learning_rate": 8.129502735972295e-05, "loss": 0.0502, "step": 64100 }, { "grad_norm": 0.5158315300941467, "learning_rate": 8.12885792025764e-05, "loss": 0.0426, "step": 64110 }, { "grad_norm": 0.6166995763778687, "learning_rate": 8.128213019001185e-05, "loss": 0.0571, "step": 64120 }, { "grad_norm": 1.6460530757904053, "learning_rate": 8.12756803222056e-05, "loss": 0.0455, "step": 64130 }, { "grad_norm": 0.6397200226783752, "learning_rate": 8.126922959933401e-05, "loss": 0.0547, "step": 64140 }, { "grad_norm": 0.7460299730300903, "learning_rate": 8.126277802157343e-05, "loss": 0.0337, "step": 64150 }, { "grad_norm": 1.5630238056182861, "learning_rate": 8.125632558910023e-05, "loss": 0.0382, "step": 64160 }, { "grad_norm": 0.4188821613788605, "learning_rate": 8.124987230209085e-05, "loss": 0.0653, "step": 64170 }, { "grad_norm": 0.42559748888015747, "learning_rate": 8.12434181607217e-05, "loss": 0.0445, "step": 64180 }, { "grad_norm": 0.47829562425613403, "learning_rate": 8.123696316516922e-05, "loss": 0.0372, "step": 64190 }, { "grad_norm": 0.7952967882156372, "learning_rate": 8.123050731560991e-05, "loss": 0.0462, "step": 64200 }, { "grad_norm": 0.6180934309959412, "learning_rate": 8.122405061222027e-05, "loss": 0.0479, "step": 64210 }, { "grad_norm": 0.5032266974449158, "learning_rate": 8.12175930551768e-05, "loss": 0.0602, "step": 64220 }, { "grad_norm": 0.7276184558868408, "learning_rate": 8.121113464465608e-05, "loss": 0.0462, "step": 64230 }, { "grad_norm": 0.4647922217845917, "learning_rate": 8.120467538083465e-05, "loss": 0.0296, "step": 64240 }, { "grad_norm": 0.6081795692443848, "learning_rate": 8.119821526388913e-05, "loss": 0.0421, "step": 64250 }, { "grad_norm": 0.51776123046875, "learning_rate": 8.119175429399612e-05, "loss": 0.0467, "step": 64260 }, { "grad_norm": 0.4342036843299866, "learning_rate": 8.118529247133227e-05, "loss": 0.0444, "step": 64270 }, { "grad_norm": 0.4840763807296753, "learning_rate": 8.117882979607425e-05, "loss": 0.0377, "step": 64280 }, { "grad_norm": 0.38598576188087463, "learning_rate": 8.117236626839871e-05, "loss": 0.0396, "step": 64290 }, { "grad_norm": 0.9386047720909119, "learning_rate": 8.116590188848241e-05, "loss": 0.0375, "step": 64300 }, { "grad_norm": 0.6241111755371094, "learning_rate": 8.115943665650205e-05, "loss": 0.0436, "step": 64310 }, { "grad_norm": 1.0770323276519775, "learning_rate": 8.11529705726344e-05, "loss": 0.0404, "step": 64320 }, { "grad_norm": 0.4059458374977112, "learning_rate": 8.114650363705624e-05, "loss": 0.0365, "step": 64330 }, { "grad_norm": 0.5333682894706726, "learning_rate": 8.114003584994436e-05, "loss": 0.0468, "step": 64340 }, { "grad_norm": 0.5648695230484009, "learning_rate": 8.113356721147561e-05, "loss": 0.0376, "step": 64350 }, { "grad_norm": 0.6447440981864929, "learning_rate": 8.112709772182683e-05, "loss": 0.0333, "step": 64360 }, { "grad_norm": 0.9436164498329163, "learning_rate": 8.11206273811749e-05, "loss": 0.0373, "step": 64370 }, { "grad_norm": 1.2313766479492188, "learning_rate": 8.111415618969668e-05, "loss": 0.0385, "step": 64380 }, { "grad_norm": 0.45525723695755005, "learning_rate": 8.110768414756914e-05, "loss": 0.039, "step": 64390 }, { "grad_norm": 0.6306934952735901, "learning_rate": 8.11012112549692e-05, "loss": 0.0523, "step": 64400 }, { "grad_norm": 0.6465997099876404, "learning_rate": 8.109473751207384e-05, "loss": 0.0486, "step": 64410 }, { "grad_norm": 1.2487208843231201, "learning_rate": 8.108826291906001e-05, "loss": 0.0498, "step": 64420 }, { "grad_norm": 0.5602562427520752, "learning_rate": 8.108178747610477e-05, "loss": 0.0446, "step": 64430 }, { "grad_norm": 0.5584398508071899, "learning_rate": 8.107531118338514e-05, "loss": 0.0515, "step": 64440 }, { "grad_norm": 0.398737370967865, "learning_rate": 8.106883404107816e-05, "loss": 0.0421, "step": 64450 }, { "grad_norm": 0.5005357265472412, "learning_rate": 8.106235604936094e-05, "loss": 0.0468, "step": 64460 }, { "grad_norm": 0.7067667841911316, "learning_rate": 8.10558772084106e-05, "loss": 0.0585, "step": 64470 }, { "grad_norm": 0.3292985260486603, "learning_rate": 8.10493975184042e-05, "loss": 0.0449, "step": 64480 }, { "grad_norm": 0.6362219452857971, "learning_rate": 8.104291697951896e-05, "loss": 0.0403, "step": 64490 }, { "grad_norm": 0.5565330982208252, "learning_rate": 8.103643559193202e-05, "loss": 0.0535, "step": 64500 }, { "grad_norm": 0.4769496023654938, "learning_rate": 8.10299533558206e-05, "loss": 0.0399, "step": 64510 }, { "grad_norm": 0.6628263592720032, "learning_rate": 8.102347027136192e-05, "loss": 0.0397, "step": 64520 }, { "grad_norm": 0.6426058411598206, "learning_rate": 8.101698633873321e-05, "loss": 0.0456, "step": 64530 }, { "grad_norm": 0.8060640692710876, "learning_rate": 8.101050155811175e-05, "loss": 0.0434, "step": 64540 }, { "grad_norm": 0.3368840217590332, "learning_rate": 8.100401592967481e-05, "loss": 0.0423, "step": 64550 }, { "grad_norm": 0.8970251083374023, "learning_rate": 8.099752945359973e-05, "loss": 0.0426, "step": 64560 }, { "grad_norm": 0.5988641381263733, "learning_rate": 8.099104213006385e-05, "loss": 0.0409, "step": 64570 }, { "grad_norm": 0.5160107016563416, "learning_rate": 8.098455395924452e-05, "loss": 0.0436, "step": 64580 }, { "grad_norm": 0.46131807565689087, "learning_rate": 8.097806494131912e-05, "loss": 0.0378, "step": 64590 }, { "grad_norm": 0.8860964179039001, "learning_rate": 8.097157507646507e-05, "loss": 0.0386, "step": 64600 }, { "grad_norm": 0.7492954730987549, "learning_rate": 8.096508436485978e-05, "loss": 0.0469, "step": 64610 }, { "grad_norm": 1.927040934562683, "learning_rate": 8.095859280668072e-05, "loss": 0.0545, "step": 64620 }, { "grad_norm": 0.21699553728103638, "learning_rate": 8.095210040210536e-05, "loss": 0.0304, "step": 64630 }, { "grad_norm": 0.6288894414901733, "learning_rate": 8.09456071513112e-05, "loss": 0.04, "step": 64640 }, { "grad_norm": 0.6596752405166626, "learning_rate": 8.09391130544758e-05, "loss": 0.0399, "step": 64650 }, { "grad_norm": 0.3190162777900696, "learning_rate": 8.093261811177664e-05, "loss": 0.0369, "step": 64660 }, { "grad_norm": 0.5809559226036072, "learning_rate": 8.092612232339134e-05, "loss": 0.042, "step": 64670 }, { "grad_norm": 0.730648934841156, "learning_rate": 8.091962568949745e-05, "loss": 0.043, "step": 64680 }, { "grad_norm": 0.445941299200058, "learning_rate": 8.091312821027262e-05, "loss": 0.0349, "step": 64690 }, { "grad_norm": 1.3206517696380615, "learning_rate": 8.090662988589447e-05, "loss": 0.0466, "step": 64700 }, { "grad_norm": 0.44020986557006836, "learning_rate": 8.09001307165407e-05, "loss": 0.0444, "step": 64710 }, { "grad_norm": 0.4606761932373047, "learning_rate": 8.089363070238894e-05, "loss": 0.0642, "step": 64720 }, { "grad_norm": 0.38461077213287354, "learning_rate": 8.088712984361694e-05, "loss": 0.0451, "step": 64730 }, { "grad_norm": 0.5714570879936218, "learning_rate": 8.088062814040239e-05, "loss": 0.0473, "step": 64740 }, { "grad_norm": 0.8337383270263672, "learning_rate": 8.087412559292309e-05, "loss": 0.0471, "step": 64750 }, { "grad_norm": 1.0251502990722656, "learning_rate": 8.086762220135679e-05, "loss": 0.0373, "step": 64760 }, { "grad_norm": 0.5818232297897339, "learning_rate": 8.086111796588128e-05, "loss": 0.0298, "step": 64770 }, { "grad_norm": 0.398579478263855, "learning_rate": 8.085461288667442e-05, "loss": 0.0397, "step": 64780 }, { "grad_norm": 0.43143561482429504, "learning_rate": 8.084810696391403e-05, "loss": 0.0364, "step": 64790 }, { "grad_norm": 0.6978592276573181, "learning_rate": 8.084160019777797e-05, "loss": 0.0507, "step": 64800 }, { "grad_norm": 0.6137887835502625, "learning_rate": 8.083509258844415e-05, "loss": 0.0435, "step": 64810 }, { "grad_norm": 1.462996244430542, "learning_rate": 8.08285841360905e-05, "loss": 0.0439, "step": 64820 }, { "grad_norm": 0.5132328271865845, "learning_rate": 8.082207484089492e-05, "loss": 0.033, "step": 64830 }, { "grad_norm": 0.5503728985786438, "learning_rate": 8.081556470303542e-05, "loss": 0.0416, "step": 64840 }, { "grad_norm": 0.6894395351409912, "learning_rate": 8.080905372268994e-05, "loss": 0.0605, "step": 64850 }, { "grad_norm": 0.6458397507667542, "learning_rate": 8.080254190003651e-05, "loss": 0.0451, "step": 64860 }, { "grad_norm": 0.984588623046875, "learning_rate": 8.079602923525315e-05, "loss": 0.0354, "step": 64870 }, { "grad_norm": 0.49282926321029663, "learning_rate": 8.078951572851792e-05, "loss": 0.0363, "step": 64880 }, { "grad_norm": 0.3863956928253174, "learning_rate": 8.07830013800089e-05, "loss": 0.041, "step": 64890 }, { "grad_norm": 0.6018878221511841, "learning_rate": 8.07764861899042e-05, "loss": 0.0419, "step": 64900 }, { "grad_norm": 0.7590779066085815, "learning_rate": 8.076997015838192e-05, "loss": 0.0379, "step": 64910 }, { "grad_norm": 0.5742306709289551, "learning_rate": 8.07634532856202e-05, "loss": 0.0397, "step": 64920 }, { "grad_norm": 1.0829652547836304, "learning_rate": 8.075693557179724e-05, "loss": 0.043, "step": 64930 }, { "grad_norm": 2.8058009147644043, "learning_rate": 8.075041701709123e-05, "loss": 0.0519, "step": 64940 }, { "grad_norm": 0.3805111348628998, "learning_rate": 8.074389762168035e-05, "loss": 0.046, "step": 64950 }, { "grad_norm": 0.6061598658561707, "learning_rate": 8.073737738574288e-05, "loss": 0.0428, "step": 64960 }, { "grad_norm": 0.6166994571685791, "learning_rate": 8.073085630945703e-05, "loss": 0.0398, "step": 64970 }, { "grad_norm": 0.5768761038780212, "learning_rate": 8.072433439300113e-05, "loss": 0.0469, "step": 64980 }, { "grad_norm": 0.6905219554901123, "learning_rate": 8.071781163655346e-05, "loss": 0.0396, "step": 64990 }, { "grad_norm": 0.5657969117164612, "learning_rate": 8.071128804029238e-05, "loss": 0.049, "step": 65000 }, { "grad_norm": 0.9900477528572083, "learning_rate": 8.07047636043962e-05, "loss": 0.0432, "step": 65010 }, { "grad_norm": 0.4984596371650696, "learning_rate": 8.069823832904334e-05, "loss": 0.041, "step": 65020 }, { "grad_norm": 0.40874359011650085, "learning_rate": 8.069171221441217e-05, "loss": 0.0304, "step": 65030 }, { "grad_norm": 0.47534942626953125, "learning_rate": 8.068518526068112e-05, "loss": 0.0493, "step": 65040 }, { "grad_norm": 1.2436702251434326, "learning_rate": 8.067865746802862e-05, "loss": 0.0319, "step": 65050 }, { "grad_norm": 0.36667993664741516, "learning_rate": 8.067212883663315e-05, "loss": 0.0378, "step": 65060 }, { "grad_norm": 1.1429387331008911, "learning_rate": 8.066559936667321e-05, "loss": 0.0394, "step": 65070 }, { "grad_norm": 0.5454895496368408, "learning_rate": 8.065906905832731e-05, "loss": 0.046, "step": 65080 }, { "grad_norm": 0.9614434838294983, "learning_rate": 8.065253791177396e-05, "loss": 0.0556, "step": 65090 }, { "grad_norm": 0.6677563786506653, "learning_rate": 8.064600592719176e-05, "loss": 0.0501, "step": 65100 }, { "grad_norm": 0.4587096869945526, "learning_rate": 8.063947310475927e-05, "loss": 0.04, "step": 65110 }, { "grad_norm": 0.6207669973373413, "learning_rate": 8.063293944465508e-05, "loss": 0.0465, "step": 65120 }, { "grad_norm": 0.6909268498420715, "learning_rate": 8.062640494705784e-05, "loss": 0.0435, "step": 65130 }, { "grad_norm": 0.5464390516281128, "learning_rate": 8.06198696121462e-05, "loss": 0.0543, "step": 65140 }, { "grad_norm": 0.45612502098083496, "learning_rate": 8.061333344009882e-05, "loss": 0.0415, "step": 65150 }, { "grad_norm": 0.6537762880325317, "learning_rate": 8.060679643109442e-05, "loss": 0.0435, "step": 65160 }, { "grad_norm": 0.6483654379844666, "learning_rate": 8.06002585853117e-05, "loss": 0.048, "step": 65170 }, { "grad_norm": 0.5794326663017273, "learning_rate": 8.059371990292939e-05, "loss": 0.0449, "step": 65180 }, { "grad_norm": 0.3676927089691162, "learning_rate": 8.058718038412629e-05, "loss": 0.0362, "step": 65190 }, { "grad_norm": 0.5085698962211609, "learning_rate": 8.058064002908116e-05, "loss": 0.0443, "step": 65200 }, { "grad_norm": 0.5203081965446472, "learning_rate": 8.057409883797282e-05, "loss": 0.0515, "step": 65210 }, { "grad_norm": 0.5915449857711792, "learning_rate": 8.05675568109801e-05, "loss": 0.0396, "step": 65220 }, { "grad_norm": 0.42642587423324585, "learning_rate": 8.056101394828188e-05, "loss": 0.0404, "step": 65230 }, { "grad_norm": 0.36630767583847046, "learning_rate": 8.0554470250057e-05, "loss": 0.0405, "step": 65240 }, { "grad_norm": 0.4850059151649475, "learning_rate": 8.054792571648438e-05, "loss": 0.0368, "step": 65250 }, { "grad_norm": 1.721971035003662, "learning_rate": 8.054138034774297e-05, "loss": 0.0345, "step": 65260 }, { "grad_norm": 0.6960790157318115, "learning_rate": 8.053483414401167e-05, "loss": 0.0495, "step": 65270 }, { "grad_norm": 0.446821928024292, "learning_rate": 8.05282871054695e-05, "loss": 0.0302, "step": 65280 }, { "grad_norm": 0.7320622205734253, "learning_rate": 8.05217392322954e-05, "loss": 0.0374, "step": 65290 }, { "grad_norm": 0.5290915966033936, "learning_rate": 8.051519052466843e-05, "loss": 0.0398, "step": 65300 }, { "grad_norm": 0.44553348422050476, "learning_rate": 8.050864098276762e-05, "loss": 0.0433, "step": 65310 }, { "grad_norm": 0.6978806257247925, "learning_rate": 8.050209060677203e-05, "loss": 0.0414, "step": 65320 }, { "grad_norm": 0.46480298042297363, "learning_rate": 8.049553939686074e-05, "loss": 0.0497, "step": 65330 }, { "grad_norm": 0.5365428924560547, "learning_rate": 8.048898735321284e-05, "loss": 0.0346, "step": 65340 }, { "grad_norm": 0.6940374374389648, "learning_rate": 8.048243447600749e-05, "loss": 0.0429, "step": 65350 }, { "grad_norm": 0.5562579035758972, "learning_rate": 8.047588076542386e-05, "loss": 0.0426, "step": 65360 }, { "grad_norm": 0.4537111818790436, "learning_rate": 8.046932622164107e-05, "loss": 0.0493, "step": 65370 }, { "grad_norm": 0.8085053563117981, "learning_rate": 8.046277084483835e-05, "loss": 0.0476, "step": 65380 }, { "grad_norm": 1.0648256540298462, "learning_rate": 8.045621463519493e-05, "loss": 0.0478, "step": 65390 }, { "grad_norm": 0.6642525792121887, "learning_rate": 8.044965759289003e-05, "loss": 0.0526, "step": 65400 }, { "grad_norm": 0.591045618057251, "learning_rate": 8.044309971810294e-05, "loss": 0.0356, "step": 65410 }, { "grad_norm": 1.018542766571045, "learning_rate": 8.043654101101294e-05, "loss": 0.0553, "step": 65420 }, { "grad_norm": 0.5546302199363708, "learning_rate": 8.042998147179934e-05, "loss": 0.041, "step": 65430 }, { "grad_norm": 0.5659980773925781, "learning_rate": 8.042342110064148e-05, "loss": 0.045, "step": 65440 }, { "grad_norm": 1.0707951784133911, "learning_rate": 8.041685989771873e-05, "loss": 0.0432, "step": 65450 }, { "grad_norm": 0.3505379855632782, "learning_rate": 8.041029786321044e-05, "loss": 0.0431, "step": 65460 }, { "grad_norm": 0.5873782634735107, "learning_rate": 8.040373499729604e-05, "loss": 0.0459, "step": 65470 }, { "grad_norm": 0.7843451499938965, "learning_rate": 8.039717130015494e-05, "loss": 0.0598, "step": 65480 }, { "grad_norm": 0.907462477684021, "learning_rate": 8.039060677196661e-05, "loss": 0.0402, "step": 65490 }, { "grad_norm": 0.588532030582428, "learning_rate": 8.038404141291049e-05, "loss": 0.0405, "step": 65500 }, { "grad_norm": 0.370239794254303, "learning_rate": 8.03774752231661e-05, "loss": 0.0417, "step": 65510 }, { "grad_norm": 0.9744997024536133, "learning_rate": 8.037090820291296e-05, "loss": 0.0355, "step": 65520 }, { "grad_norm": 0.4054998755455017, "learning_rate": 8.036434035233058e-05, "loss": 0.0389, "step": 65530 }, { "grad_norm": 0.47717538475990295, "learning_rate": 8.035777167159856e-05, "loss": 0.044, "step": 65540 }, { "grad_norm": 0.619411826133728, "learning_rate": 8.035120216089647e-05, "loss": 0.0487, "step": 65550 }, { "grad_norm": 1.269339919090271, "learning_rate": 8.034463182040389e-05, "loss": 0.0462, "step": 65560 }, { "grad_norm": 0.5085673332214355, "learning_rate": 8.03380606503005e-05, "loss": 0.0334, "step": 65570 }, { "grad_norm": 0.49344462156295776, "learning_rate": 8.033148865076592e-05, "loss": 0.0338, "step": 65580 }, { "grad_norm": 0.6243752241134644, "learning_rate": 8.032491582197983e-05, "loss": 0.052, "step": 65590 }, { "grad_norm": 0.5802271962165833, "learning_rate": 8.031834216412195e-05, "loss": 0.0391, "step": 65600 }, { "grad_norm": 0.3311014175415039, "learning_rate": 8.031176767737196e-05, "loss": 0.0375, "step": 65610 }, { "grad_norm": 0.42378920316696167, "learning_rate": 8.030519236190966e-05, "loss": 0.035, "step": 65620 }, { "grad_norm": 0.5744569897651672, "learning_rate": 8.029861621791476e-05, "loss": 0.036, "step": 65630 }, { "grad_norm": 0.45525074005126953, "learning_rate": 8.029203924556708e-05, "loss": 0.0324, "step": 65640 }, { "grad_norm": 0.23724353313446045, "learning_rate": 8.028546144504643e-05, "loss": 0.0297, "step": 65650 }, { "grad_norm": 0.4999489188194275, "learning_rate": 8.027888281653265e-05, "loss": 0.0623, "step": 65660 }, { "grad_norm": 0.6920046210289001, "learning_rate": 8.027230336020559e-05, "loss": 0.0332, "step": 65670 }, { "grad_norm": 0.6655298471450806, "learning_rate": 8.026572307624512e-05, "loss": 0.0444, "step": 65680 }, { "grad_norm": 0.6202447414398193, "learning_rate": 8.025914196483116e-05, "loss": 0.0512, "step": 65690 }, { "grad_norm": 0.5018525123596191, "learning_rate": 8.025256002614361e-05, "loss": 0.0569, "step": 65700 }, { "grad_norm": 0.5217541456222534, "learning_rate": 8.024597726036245e-05, "loss": 0.0358, "step": 65710 }, { "grad_norm": 0.6340828537940979, "learning_rate": 8.023939366766763e-05, "loss": 0.0492, "step": 65720 }, { "grad_norm": 0.7394629120826721, "learning_rate": 8.023280924823915e-05, "loss": 0.0393, "step": 65730 }, { "grad_norm": 0.5646908283233643, "learning_rate": 8.022622400225701e-05, "loss": 0.0414, "step": 65740 }, { "grad_norm": 0.9491260051727295, "learning_rate": 8.021963792990129e-05, "loss": 0.0518, "step": 65750 }, { "grad_norm": 0.47258999943733215, "learning_rate": 8.0213051031352e-05, "loss": 0.0379, "step": 65760 }, { "grad_norm": 0.3602401316165924, "learning_rate": 8.020646330678924e-05, "loss": 0.0318, "step": 65770 }, { "grad_norm": 0.6091618537902832, "learning_rate": 8.019987475639312e-05, "loss": 0.0415, "step": 65780 }, { "grad_norm": 0.6630339622497559, "learning_rate": 8.019328538034378e-05, "loss": 0.0642, "step": 65790 }, { "grad_norm": 0.9878313541412354, "learning_rate": 8.018669517882136e-05, "loss": 0.0431, "step": 65800 }, { "grad_norm": 0.34369346499443054, "learning_rate": 8.018010415200604e-05, "loss": 0.0383, "step": 65810 }, { "grad_norm": 1.419795274734497, "learning_rate": 8.017351230007802e-05, "loss": 0.0357, "step": 65820 }, { "grad_norm": 0.3853718340396881, "learning_rate": 8.016691962321748e-05, "loss": 0.031, "step": 65830 }, { "grad_norm": 0.5779238343238831, "learning_rate": 8.016032612160472e-05, "loss": 0.0443, "step": 65840 }, { "grad_norm": 0.3550890386104584, "learning_rate": 8.015373179541995e-05, "loss": 0.0409, "step": 65850 }, { "grad_norm": 0.8068091869354248, "learning_rate": 8.014713664484349e-05, "loss": 0.0377, "step": 65860 }, { "grad_norm": 0.45140907168388367, "learning_rate": 8.014054067005564e-05, "loss": 0.0395, "step": 65870 }, { "grad_norm": 0.4089895486831665, "learning_rate": 8.013394387123674e-05, "loss": 0.0399, "step": 65880 }, { "grad_norm": 0.4500645399093628, "learning_rate": 8.012734624856712e-05, "loss": 0.0461, "step": 65890 }, { "grad_norm": 0.7005497813224792, "learning_rate": 8.012074780222718e-05, "loss": 0.0429, "step": 65900 }, { "grad_norm": 0.6058674454689026, "learning_rate": 8.011414853239732e-05, "loss": 0.0349, "step": 65910 }, { "grad_norm": 0.36048850417137146, "learning_rate": 8.010754843925794e-05, "loss": 0.0367, "step": 65920 }, { "grad_norm": 0.5484007596969604, "learning_rate": 8.01009475229895e-05, "loss": 0.0402, "step": 65930 }, { "grad_norm": 1.8628994226455688, "learning_rate": 8.009434578377246e-05, "loss": 0.0406, "step": 65940 }, { "grad_norm": 0.788490891456604, "learning_rate": 8.00877432217873e-05, "loss": 0.0346, "step": 65950 }, { "grad_norm": 0.3583052456378937, "learning_rate": 8.008113983721457e-05, "loss": 0.0434, "step": 65960 }, { "grad_norm": 0.7035466432571411, "learning_rate": 8.007453563023477e-05, "loss": 0.0338, "step": 65970 }, { "grad_norm": 0.5094318985939026, "learning_rate": 8.006793060102845e-05, "loss": 0.0427, "step": 65980 }, { "grad_norm": 0.5783898830413818, "learning_rate": 8.006132474977622e-05, "loss": 0.0325, "step": 65990 }, { "grad_norm": 0.666164219379425, "learning_rate": 8.005471807665866e-05, "loss": 0.0419, "step": 66000 }, { "grad_norm": 0.6159561276435852, "learning_rate": 8.004811058185639e-05, "loss": 0.0432, "step": 66010 }, { "grad_norm": 0.9425373077392578, "learning_rate": 8.004150226555008e-05, "loss": 0.0601, "step": 66020 }, { "grad_norm": 0.370810866355896, "learning_rate": 8.003489312792038e-05, "loss": 0.0315, "step": 66030 }, { "grad_norm": 0.5560862421989441, "learning_rate": 8.002828316914797e-05, "loss": 0.0524, "step": 66040 }, { "grad_norm": 0.6231657862663269, "learning_rate": 8.00216723894136e-05, "loss": 0.0488, "step": 66050 }, { "grad_norm": 0.7979594469070435, "learning_rate": 8.001506078889798e-05, "loss": 0.045, "step": 66060 }, { "grad_norm": 0.85443514585495, "learning_rate": 8.000844836778188e-05, "loss": 0.0424, "step": 66070 }, { "grad_norm": 0.46353790163993835, "learning_rate": 8.000183512624604e-05, "loss": 0.0363, "step": 66080 }, { "grad_norm": 0.5845217704772949, "learning_rate": 7.999522106447133e-05, "loss": 0.0526, "step": 66090 }, { "grad_norm": 0.7340012192726135, "learning_rate": 7.998860618263854e-05, "loss": 0.0512, "step": 66100 }, { "grad_norm": 0.8243312239646912, "learning_rate": 7.998199048092854e-05, "loss": 0.052, "step": 66110 }, { "grad_norm": 0.5629289150238037, "learning_rate": 7.997537395952215e-05, "loss": 0.0472, "step": 66120 }, { "grad_norm": 0.5102395415306091, "learning_rate": 7.996875661860032e-05, "loss": 0.0472, "step": 66130 }, { "grad_norm": 0.7627865672111511, "learning_rate": 7.996213845834393e-05, "loss": 0.0428, "step": 66140 }, { "grad_norm": 0.522953450679779, "learning_rate": 7.995551947893394e-05, "loss": 0.0464, "step": 66150 }, { "grad_norm": 0.5472820997238159, "learning_rate": 7.99488996805513e-05, "loss": 0.0662, "step": 66160 }, { "grad_norm": 0.6629661917686462, "learning_rate": 7.9942279063377e-05, "loss": 0.0362, "step": 66170 }, { "grad_norm": 0.288639098405838, "learning_rate": 7.993565762759203e-05, "loss": 0.0372, "step": 66180 }, { "grad_norm": 0.7055755257606506, "learning_rate": 7.992903537337743e-05, "loss": 0.0572, "step": 66190 }, { "grad_norm": 0.535601794719696, "learning_rate": 7.992241230091425e-05, "loss": 0.0338, "step": 66200 }, { "grad_norm": 0.6006422638893127, "learning_rate": 7.991578841038355e-05, "loss": 0.0368, "step": 66210 }, { "grad_norm": 0.5713355541229248, "learning_rate": 7.990916370196644e-05, "loss": 0.0327, "step": 66220 }, { "grad_norm": 0.4368429183959961, "learning_rate": 7.990253817584403e-05, "loss": 0.0363, "step": 66230 }, { "grad_norm": 0.6029165983200073, "learning_rate": 7.989591183219747e-05, "loss": 0.0388, "step": 66240 }, { "grad_norm": 0.6789546608924866, "learning_rate": 7.988928467120789e-05, "loss": 0.0477, "step": 66250 }, { "grad_norm": 0.32097387313842773, "learning_rate": 7.98826566930565e-05, "loss": 0.0339, "step": 66260 }, { "grad_norm": 0.47010165452957153, "learning_rate": 7.987602789792452e-05, "loss": 0.0531, "step": 66270 }, { "grad_norm": 0.6926090717315674, "learning_rate": 7.986939828599315e-05, "loss": 0.0368, "step": 66280 }, { "grad_norm": 0.44575709104537964, "learning_rate": 7.986276785744365e-05, "loss": 0.033, "step": 66290 }, { "grad_norm": 0.43962374329566956, "learning_rate": 7.98561366124573e-05, "loss": 0.0283, "step": 66300 }, { "grad_norm": 0.53537517786026, "learning_rate": 7.98495045512154e-05, "loss": 0.037, "step": 66310 }, { "grad_norm": 0.7934747338294983, "learning_rate": 7.984287167389924e-05, "loss": 0.0398, "step": 66320 }, { "grad_norm": 0.6651767492294312, "learning_rate": 7.98362379806902e-05, "loss": 0.0413, "step": 66330 }, { "grad_norm": 0.8580460548400879, "learning_rate": 7.98296034717696e-05, "loss": 0.0449, "step": 66340 }, { "grad_norm": 0.4320301413536072, "learning_rate": 7.982296814731886e-05, "loss": 0.0416, "step": 66350 }, { "grad_norm": 3.6082472801208496, "learning_rate": 7.981633200751937e-05, "loss": 0.0442, "step": 66360 }, { "grad_norm": 0.5014459490776062, "learning_rate": 7.980969505255258e-05, "loss": 0.034, "step": 66370 }, { "grad_norm": 0.8518579006195068, "learning_rate": 7.980305728259991e-05, "loss": 0.0345, "step": 66380 }, { "grad_norm": 0.4621298313140869, "learning_rate": 7.979641869784286e-05, "loss": 0.0448, "step": 66390 }, { "grad_norm": 0.6591896414756775, "learning_rate": 7.978977929846291e-05, "loss": 0.0407, "step": 66400 }, { "grad_norm": 0.563655436038971, "learning_rate": 7.97831390846416e-05, "loss": 0.0352, "step": 66410 }, { "grad_norm": 0.9845471978187561, "learning_rate": 7.977649805656043e-05, "loss": 0.0466, "step": 66420 }, { "grad_norm": 0.9932176470756531, "learning_rate": 7.9769856214401e-05, "loss": 0.049, "step": 66430 }, { "grad_norm": 0.6238579154014587, "learning_rate": 7.97632135583449e-05, "loss": 0.0363, "step": 66440 }, { "grad_norm": 0.7051265835762024, "learning_rate": 7.97565700885737e-05, "loss": 0.0637, "step": 66450 }, { "grad_norm": 0.5268487334251404, "learning_rate": 7.974992580526907e-05, "loss": 0.0484, "step": 66460 }, { "grad_norm": 0.5363688468933105, "learning_rate": 7.974328070861264e-05, "loss": 0.0401, "step": 66470 }, { "grad_norm": 0.7136755585670471, "learning_rate": 7.97366347987861e-05, "loss": 0.0444, "step": 66480 }, { "grad_norm": 0.7550502419471741, "learning_rate": 7.972998807597112e-05, "loss": 0.0349, "step": 66490 }, { "grad_norm": 0.5259597897529602, "learning_rate": 7.972334054034944e-05, "loss": 0.0441, "step": 66500 }, { "grad_norm": 0.7069973945617676, "learning_rate": 7.971669219210281e-05, "loss": 0.0454, "step": 66510 }, { "grad_norm": 0.7337439060211182, "learning_rate": 7.971004303141295e-05, "loss": 0.0424, "step": 66520 }, { "grad_norm": 0.8417528867721558, "learning_rate": 7.970339305846169e-05, "loss": 0.0579, "step": 66530 }, { "grad_norm": 0.44401997327804565, "learning_rate": 7.969674227343082e-05, "loss": 0.0334, "step": 66540 }, { "grad_norm": 0.5706638693809509, "learning_rate": 7.969009067650219e-05, "loss": 0.0391, "step": 66550 }, { "grad_norm": 0.5435660481452942, "learning_rate": 7.968343826785762e-05, "loss": 0.0389, "step": 66560 }, { "grad_norm": 0.5846661329269409, "learning_rate": 7.967678504767899e-05, "loss": 0.033, "step": 66570 }, { "grad_norm": 1.1417579650878906, "learning_rate": 7.967013101614822e-05, "loss": 0.0453, "step": 66580 }, { "grad_norm": 0.767166793346405, "learning_rate": 7.96634761734472e-05, "loss": 0.049, "step": 66590 }, { "grad_norm": 0.7358008623123169, "learning_rate": 7.965682051975792e-05, "loss": 0.0433, "step": 66600 }, { "grad_norm": 0.6452050805091858, "learning_rate": 7.965016405526228e-05, "loss": 0.0421, "step": 66610 }, { "grad_norm": 0.409106969833374, "learning_rate": 7.96435067801423e-05, "loss": 0.0424, "step": 66620 }, { "grad_norm": 0.3976249694824219, "learning_rate": 7.963684869457998e-05, "loss": 0.0441, "step": 66630 }, { "grad_norm": 0.8150283694267273, "learning_rate": 7.963018979875735e-05, "loss": 0.0467, "step": 66640 }, { "grad_norm": 0.5552700757980347, "learning_rate": 7.962353009285646e-05, "loss": 0.0425, "step": 66650 }, { "grad_norm": 0.7457661628723145, "learning_rate": 7.961686957705938e-05, "loss": 0.049, "step": 66660 }, { "grad_norm": 0.8353304862976074, "learning_rate": 7.961020825154824e-05, "loss": 0.0484, "step": 66670 }, { "grad_norm": 0.5060407519340515, "learning_rate": 7.960354611650512e-05, "loss": 0.0498, "step": 66680 }, { "grad_norm": 0.5234826803207397, "learning_rate": 7.959688317211216e-05, "loss": 0.0372, "step": 66690 }, { "grad_norm": 0.783707857131958, "learning_rate": 7.959021941855154e-05, "loss": 0.0386, "step": 66700 }, { "grad_norm": 0.49867188930511475, "learning_rate": 7.958355485600542e-05, "loss": 0.037, "step": 66710 }, { "grad_norm": 0.5300672054290771, "learning_rate": 7.957688948465604e-05, "loss": 0.0359, "step": 66720 }, { "grad_norm": 0.6552871465682983, "learning_rate": 7.957022330468561e-05, "loss": 0.0556, "step": 66730 }, { "grad_norm": 0.6698415279388428, "learning_rate": 7.95635563162764e-05, "loss": 0.0477, "step": 66740 }, { "grad_norm": 1.1023155450820923, "learning_rate": 7.955688851961064e-05, "loss": 0.0486, "step": 66750 }, { "grad_norm": 0.6235032081604004, "learning_rate": 7.955021991487067e-05, "loss": 0.0506, "step": 66760 }, { "grad_norm": 0.37371936440467834, "learning_rate": 7.954355050223879e-05, "loss": 0.0449, "step": 66770 }, { "grad_norm": 0.7363479733467102, "learning_rate": 7.953688028189732e-05, "loss": 0.044, "step": 66780 }, { "grad_norm": 0.7066335082054138, "learning_rate": 7.953020925402865e-05, "loss": 0.0419, "step": 66790 }, { "grad_norm": 0.39410915970802307, "learning_rate": 7.952353741881516e-05, "loss": 0.0316, "step": 66800 }, { "grad_norm": 0.7836428880691528, "learning_rate": 7.951686477643922e-05, "loss": 0.0442, "step": 66810 }, { "grad_norm": 0.35852962732315063, "learning_rate": 7.951019132708333e-05, "loss": 0.0502, "step": 66820 }, { "grad_norm": 0.5000956654548645, "learning_rate": 7.950351707092987e-05, "loss": 0.0398, "step": 66830 }, { "grad_norm": 0.47967520356178284, "learning_rate": 7.949684200816133e-05, "loss": 0.029, "step": 66840 }, { "grad_norm": 0.7015573978424072, "learning_rate": 7.949016613896023e-05, "loss": 0.0434, "step": 66850 }, { "grad_norm": 0.529508113861084, "learning_rate": 7.948348946350906e-05, "loss": 0.0415, "step": 66860 }, { "grad_norm": 0.49345827102661133, "learning_rate": 7.947681198199036e-05, "loss": 0.0371, "step": 66870 }, { "grad_norm": 0.4385996460914612, "learning_rate": 7.94701336945867e-05, "loss": 0.0442, "step": 66880 }, { "grad_norm": 0.441524863243103, "learning_rate": 7.946345460148065e-05, "loss": 0.0286, "step": 66890 }, { "grad_norm": 0.8357723355293274, "learning_rate": 7.945677470285482e-05, "loss": 0.0398, "step": 66900 }, { "grad_norm": 0.4833243489265442, "learning_rate": 7.945009399889185e-05, "loss": 0.0371, "step": 66910 }, { "grad_norm": 0.603645920753479, "learning_rate": 7.944341248977436e-05, "loss": 0.043, "step": 66920 }, { "grad_norm": 0.8359541296958923, "learning_rate": 7.943673017568504e-05, "loss": 0.0407, "step": 66930 }, { "grad_norm": 0.48952966928482056, "learning_rate": 7.943004705680656e-05, "loss": 0.0408, "step": 66940 }, { "grad_norm": 0.7459771633148193, "learning_rate": 7.942336313332169e-05, "loss": 0.0443, "step": 66950 }, { "grad_norm": 0.5060184001922607, "learning_rate": 7.941667840541309e-05, "loss": 0.0404, "step": 66960 }, { "grad_norm": 0.9661815166473389, "learning_rate": 7.940999287326356e-05, "loss": 0.0338, "step": 66970 }, { "grad_norm": 0.5365052819252014, "learning_rate": 7.940330653705588e-05, "loss": 0.0372, "step": 66980 }, { "grad_norm": 0.6044355630874634, "learning_rate": 7.939661939697284e-05, "loss": 0.0502, "step": 66990 }, { "grad_norm": 0.6608447432518005, "learning_rate": 7.938993145319727e-05, "loss": 0.0383, "step": 67000 }, { "grad_norm": 0.9369909763336182, "learning_rate": 7.938324270591202e-05, "loss": 0.0494, "step": 67010 }, { "grad_norm": 0.526509165763855, "learning_rate": 7.937655315529995e-05, "loss": 0.0426, "step": 67020 }, { "grad_norm": 1.049087643623352, "learning_rate": 7.936986280154395e-05, "loss": 0.0403, "step": 67030 }, { "grad_norm": 0.7581720352172852, "learning_rate": 7.936317164482694e-05, "loss": 0.0321, "step": 67040 }, { "grad_norm": 0.46004822850227356, "learning_rate": 7.935647968533184e-05, "loss": 0.0338, "step": 67050 }, { "grad_norm": 0.673076331615448, "learning_rate": 7.934978692324162e-05, "loss": 0.0639, "step": 67060 }, { "grad_norm": 0.7962045073509216, "learning_rate": 7.934309335873927e-05, "loss": 0.0466, "step": 67070 }, { "grad_norm": 0.5569140911102295, "learning_rate": 7.933639899200774e-05, "loss": 0.0351, "step": 67080 }, { "grad_norm": 0.8901545405387878, "learning_rate": 7.932970382323011e-05, "loss": 0.0505, "step": 67090 }, { "grad_norm": 0.6484209299087524, "learning_rate": 7.932300785258937e-05, "loss": 0.0421, "step": 67100 }, { "grad_norm": 0.5373373627662659, "learning_rate": 7.931631108026862e-05, "loss": 0.0373, "step": 67110 }, { "grad_norm": 0.5342562198638916, "learning_rate": 7.930961350645095e-05, "loss": 0.0541, "step": 67120 }, { "grad_norm": 0.5665781497955322, "learning_rate": 7.930291513131944e-05, "loss": 0.0433, "step": 67130 }, { "grad_norm": 0.6538094282150269, "learning_rate": 7.929621595505726e-05, "loss": 0.0501, "step": 67140 }, { "grad_norm": 0.43927934765815735, "learning_rate": 7.928951597784753e-05, "loss": 0.0472, "step": 67150 }, { "grad_norm": 0.9520275592803955, "learning_rate": 7.928281519987343e-05, "loss": 0.0319, "step": 67160 }, { "grad_norm": 1.6763161420822144, "learning_rate": 7.927611362131817e-05, "loss": 0.0496, "step": 67170 }, { "grad_norm": 0.31825006008148193, "learning_rate": 7.926941124236497e-05, "loss": 0.038, "step": 67180 }, { "grad_norm": 0.5984510779380798, "learning_rate": 7.926270806319705e-05, "loss": 0.0491, "step": 67190 }, { "grad_norm": 0.5235229730606079, "learning_rate": 7.92560040839977e-05, "loss": 0.038, "step": 67200 }, { "grad_norm": 0.3280402719974518, "learning_rate": 7.924929930495018e-05, "loss": 0.0303, "step": 67210 }, { "grad_norm": 1.665173053741455, "learning_rate": 7.92425937262378e-05, "loss": 0.044, "step": 67220 }, { "grad_norm": 0.8790895938873291, "learning_rate": 7.92358873480439e-05, "loss": 0.0564, "step": 67230 }, { "grad_norm": 0.5327475070953369, "learning_rate": 7.922918017055183e-05, "loss": 0.0336, "step": 67240 }, { "grad_norm": 0.42302820086479187, "learning_rate": 7.922247219394494e-05, "loss": 0.0468, "step": 67250 }, { "grad_norm": 0.6617704033851624, "learning_rate": 7.921576341840664e-05, "loss": 0.044, "step": 67260 }, { "grad_norm": 0.27455058693885803, "learning_rate": 7.920905384412036e-05, "loss": 0.0382, "step": 67270 }, { "grad_norm": 1.070205569267273, "learning_rate": 7.920234347126951e-05, "loss": 0.0531, "step": 67280 }, { "grad_norm": 4.986139297485352, "learning_rate": 7.919563230003755e-05, "loss": 0.0374, "step": 67290 }, { "grad_norm": 0.534092366695404, "learning_rate": 7.918892033060799e-05, "loss": 0.0446, "step": 67300 }, { "grad_norm": 0.5642482042312622, "learning_rate": 7.918220756316432e-05, "loss": 0.0438, "step": 67310 }, { "grad_norm": 0.34237655997276306, "learning_rate": 7.917549399789004e-05, "loss": 0.0357, "step": 67320 }, { "grad_norm": 0.7169159650802612, "learning_rate": 7.916877963496871e-05, "loss": 0.0589, "step": 67330 }, { "grad_norm": 0.8692095875740051, "learning_rate": 7.916206447458391e-05, "loss": 0.0425, "step": 67340 }, { "grad_norm": 1.08415687084198, "learning_rate": 7.915534851691923e-05, "loss": 0.0373, "step": 67350 }, { "grad_norm": 0.8138710260391235, "learning_rate": 7.914863176215828e-05, "loss": 0.0343, "step": 67360 }, { "grad_norm": 0.8318634033203125, "learning_rate": 7.914191421048467e-05, "loss": 0.0495, "step": 67370 }, { "grad_norm": 0.6186801791191101, "learning_rate": 7.91351958620821e-05, "loss": 0.0434, "step": 67380 }, { "grad_norm": 1.2458667755126953, "learning_rate": 7.91284767171342e-05, "loss": 0.0371, "step": 67390 }, { "grad_norm": 0.609779417514801, "learning_rate": 7.91217567758247e-05, "loss": 0.0389, "step": 67400 }, { "grad_norm": 0.742646336555481, "learning_rate": 7.911503603833729e-05, "loss": 0.045, "step": 67410 }, { "grad_norm": 0.6406974196434021, "learning_rate": 7.910831450485574e-05, "loss": 0.0396, "step": 67420 }, { "grad_norm": 0.6514495611190796, "learning_rate": 7.910159217556382e-05, "loss": 0.0462, "step": 67430 }, { "grad_norm": 0.6564139127731323, "learning_rate": 7.909486905064529e-05, "loss": 0.0365, "step": 67440 }, { "grad_norm": 0.7121584415435791, "learning_rate": 7.908814513028398e-05, "loss": 0.035, "step": 67450 }, { "grad_norm": 0.5621985793113708, "learning_rate": 7.90814204146637e-05, "loss": 0.0452, "step": 67460 }, { "grad_norm": 0.4654776453971863, "learning_rate": 7.907469490396832e-05, "loss": 0.0335, "step": 67470 }, { "grad_norm": 0.7720234394073486, "learning_rate": 7.90679685983817e-05, "loss": 0.0351, "step": 67480 }, { "grad_norm": 0.5646557807922363, "learning_rate": 7.906124149808776e-05, "loss": 0.0532, "step": 67490 }, { "grad_norm": 1.1763277053833008, "learning_rate": 7.905451360327037e-05, "loss": 0.0585, "step": 67500 }, { "grad_norm": 0.7599460482597351, "learning_rate": 7.904778491411349e-05, "loss": 0.037, "step": 67510 }, { "grad_norm": 0.47725099325180054, "learning_rate": 7.90410554308011e-05, "loss": 0.051, "step": 67520 }, { "grad_norm": 0.4325290024280548, "learning_rate": 7.903432515351716e-05, "loss": 0.0428, "step": 67530 }, { "grad_norm": 0.5339792370796204, "learning_rate": 7.902759408244566e-05, "loss": 0.0387, "step": 67540 }, { "grad_norm": 0.7676340937614441, "learning_rate": 7.902086221777066e-05, "loss": 0.0352, "step": 67550 }, { "grad_norm": 0.401869535446167, "learning_rate": 7.901412955967619e-05, "loss": 0.0365, "step": 67560 }, { "grad_norm": 0.5152249932289124, "learning_rate": 7.900739610834632e-05, "loss": 0.0429, "step": 67570 }, { "grad_norm": 0.6898306012153625, "learning_rate": 7.900066186396513e-05, "loss": 0.0485, "step": 67580 }, { "grad_norm": 0.33419036865234375, "learning_rate": 7.899392682671675e-05, "loss": 0.0368, "step": 67590 }, { "grad_norm": 0.8683757781982422, "learning_rate": 7.898719099678529e-05, "loss": 0.0487, "step": 67600 }, { "grad_norm": 0.5071070194244385, "learning_rate": 7.898045437435493e-05, "loss": 0.0335, "step": 67610 }, { "grad_norm": 0.43473029136657715, "learning_rate": 7.897371695960982e-05, "loss": 0.0388, "step": 67620 }, { "grad_norm": 0.7089622020721436, "learning_rate": 7.89669787527342e-05, "loss": 0.0306, "step": 67630 }, { "grad_norm": 0.5138701796531677, "learning_rate": 7.896023975391224e-05, "loss": 0.036, "step": 67640 }, { "grad_norm": 0.5507164597511292, "learning_rate": 7.895349996332821e-05, "loss": 0.0398, "step": 67650 }, { "grad_norm": 1.1028658151626587, "learning_rate": 7.894675938116637e-05, "loss": 0.0434, "step": 67660 }, { "grad_norm": 1.5883641242980957, "learning_rate": 7.8940018007611e-05, "loss": 0.0418, "step": 67670 }, { "grad_norm": 0.4070195257663727, "learning_rate": 7.893327584284642e-05, "loss": 0.0373, "step": 67680 }, { "grad_norm": 0.5844566822052002, "learning_rate": 7.892653288705693e-05, "loss": 0.0388, "step": 67690 }, { "grad_norm": 0.4046311676502228, "learning_rate": 7.891978914042691e-05, "loss": 0.0395, "step": 67700 }, { "grad_norm": 0.5359278321266174, "learning_rate": 7.891304460314073e-05, "loss": 0.0504, "step": 67710 }, { "grad_norm": 0.5737521052360535, "learning_rate": 7.890629927538276e-05, "loss": 0.0459, "step": 67720 }, { "grad_norm": 0.5319198369979858, "learning_rate": 7.889955315733742e-05, "loss": 0.0294, "step": 67730 }, { "grad_norm": 0.5312128067016602, "learning_rate": 7.889280624918917e-05, "loss": 0.0422, "step": 67740 }, { "grad_norm": 0.5914210677146912, "learning_rate": 7.888605855112244e-05, "loss": 0.0396, "step": 67750 }, { "grad_norm": 0.789317786693573, "learning_rate": 7.887931006332173e-05, "loss": 0.0365, "step": 67760 }, { "grad_norm": 0.5023782253265381, "learning_rate": 7.887256078597152e-05, "loss": 0.0292, "step": 67770 }, { "grad_norm": 0.7524036765098572, "learning_rate": 7.886581071925636e-05, "loss": 0.0556, "step": 67780 }, { "grad_norm": 0.6773175001144409, "learning_rate": 7.885905986336077e-05, "loss": 0.049, "step": 67790 }, { "grad_norm": 0.37813642621040344, "learning_rate": 7.885230821846931e-05, "loss": 0.0497, "step": 67800 }, { "grad_norm": 0.9561022520065308, "learning_rate": 7.88455557847666e-05, "loss": 0.0398, "step": 67810 }, { "grad_norm": 1.3424735069274902, "learning_rate": 7.883880256243724e-05, "loss": 0.0547, "step": 67820 }, { "grad_norm": 0.939982533454895, "learning_rate": 7.883204855166584e-05, "loss": 0.0739, "step": 67830 }, { "grad_norm": 0.3814399540424347, "learning_rate": 7.882529375263707e-05, "loss": 0.0609, "step": 67840 }, { "grad_norm": 0.5937349200248718, "learning_rate": 7.88185381655356e-05, "loss": 0.0512, "step": 67850 }, { "grad_norm": 0.619197428226471, "learning_rate": 7.881178179054611e-05, "loss": 0.0373, "step": 67860 }, { "grad_norm": 0.6038548946380615, "learning_rate": 7.880502462785335e-05, "loss": 0.05, "step": 67870 }, { "grad_norm": 0.5392666459083557, "learning_rate": 7.879826667764203e-05, "loss": 0.0358, "step": 67880 }, { "grad_norm": 1.322037696838379, "learning_rate": 7.879150794009691e-05, "loss": 0.0811, "step": 67890 }, { "grad_norm": 0.9471047520637512, "learning_rate": 7.878474841540278e-05, "loss": 0.0494, "step": 67900 }, { "grad_norm": 0.6220576167106628, "learning_rate": 7.877798810374446e-05, "loss": 0.0411, "step": 67910 }, { "grad_norm": 0.6409814953804016, "learning_rate": 7.877122700530672e-05, "loss": 0.0344, "step": 67920 }, { "grad_norm": 0.49099859595298767, "learning_rate": 7.876446512027448e-05, "loss": 0.0334, "step": 67930 }, { "grad_norm": 0.3206713795661926, "learning_rate": 7.875770244883256e-05, "loss": 0.035, "step": 67940 }, { "grad_norm": 0.905722975730896, "learning_rate": 7.875093899116586e-05, "loss": 0.0403, "step": 67950 }, { "grad_norm": 1.1415566205978394, "learning_rate": 7.874417474745928e-05, "loss": 0.0433, "step": 67960 }, { "grad_norm": 0.581306517124176, "learning_rate": 7.873740971789775e-05, "loss": 0.0373, "step": 67970 }, { "grad_norm": 1.1746931076049805, "learning_rate": 7.873064390266625e-05, "loss": 0.043, "step": 67980 }, { "grad_norm": 0.5736600160598755, "learning_rate": 7.872387730194974e-05, "loss": 0.0436, "step": 67990 }, { "grad_norm": 0.4541034996509552, "learning_rate": 7.871710991593322e-05, "loss": 0.0487, "step": 68000 }, { "grad_norm": 0.4003932774066925, "learning_rate": 7.87103417448017e-05, "loss": 0.037, "step": 68010 }, { "grad_norm": 0.7810644507408142, "learning_rate": 7.870357278874021e-05, "loss": 0.0529, "step": 68020 }, { "grad_norm": 0.8032855987548828, "learning_rate": 7.869680304793385e-05, "loss": 0.0372, "step": 68030 }, { "grad_norm": 0.5123417377471924, "learning_rate": 7.869003252256765e-05, "loss": 0.0498, "step": 68040 }, { "grad_norm": 0.4887758493423462, "learning_rate": 7.868326121282674e-05, "loss": 0.0525, "step": 68050 }, { "grad_norm": 0.6513171792030334, "learning_rate": 7.867648911889625e-05, "loss": 0.0488, "step": 68060 }, { "grad_norm": 0.5505173206329346, "learning_rate": 7.866971624096133e-05, "loss": 0.0405, "step": 68070 }, { "grad_norm": 0.7130863070487976, "learning_rate": 7.866294257920715e-05, "loss": 0.0459, "step": 68080 }, { "grad_norm": 0.6844847202301025, "learning_rate": 7.865616813381887e-05, "loss": 0.0465, "step": 68090 }, { "grad_norm": 0.5687243938446045, "learning_rate": 7.864939290498171e-05, "loss": 0.0391, "step": 68100 }, { "grad_norm": 0.3835335969924927, "learning_rate": 7.86426168928809e-05, "loss": 0.0324, "step": 68110 }, { "grad_norm": 0.5852607488632202, "learning_rate": 7.863584009770173e-05, "loss": 0.0511, "step": 68120 }, { "grad_norm": 0.45282888412475586, "learning_rate": 7.862906251962944e-05, "loss": 0.0476, "step": 68130 }, { "grad_norm": 0.3947010636329651, "learning_rate": 7.862228415884936e-05, "loss": 0.0349, "step": 68140 }, { "grad_norm": 0.448569655418396, "learning_rate": 7.861550501554676e-05, "loss": 0.0308, "step": 68150 }, { "grad_norm": 0.592065691947937, "learning_rate": 7.860872508990701e-05, "loss": 0.0385, "step": 68160 }, { "grad_norm": 0.6478938460350037, "learning_rate": 7.860194438211545e-05, "loss": 0.0331, "step": 68170 }, { "grad_norm": 1.262166976928711, "learning_rate": 7.859516289235748e-05, "loss": 0.0372, "step": 68180 }, { "grad_norm": 0.8535346388816833, "learning_rate": 7.858838062081853e-05, "loss": 0.0464, "step": 68190 }, { "grad_norm": 0.4669411778450012, "learning_rate": 7.858159756768396e-05, "loss": 0.0412, "step": 68200 }, { "grad_norm": 0.5219050049781799, "learning_rate": 7.857481373313926e-05, "loss": 0.034, "step": 68210 }, { "grad_norm": 0.6166796684265137, "learning_rate": 7.856802911736989e-05, "loss": 0.0415, "step": 68220 }, { "grad_norm": 0.43012773990631104, "learning_rate": 7.856124372056133e-05, "loss": 0.0352, "step": 68230 }, { "grad_norm": 0.7579917311668396, "learning_rate": 7.855445754289911e-05, "loss": 0.0474, "step": 68240 }, { "grad_norm": 0.6302482485771179, "learning_rate": 7.854767058456873e-05, "loss": 0.0309, "step": 68250 }, { "grad_norm": 0.43731972575187683, "learning_rate": 7.854088284575579e-05, "loss": 0.0408, "step": 68260 }, { "grad_norm": 0.48344293236732483, "learning_rate": 7.85340943266458e-05, "loss": 0.0306, "step": 68270 }, { "grad_norm": 0.9376006126403809, "learning_rate": 7.852730502742441e-05, "loss": 0.0399, "step": 68280 }, { "grad_norm": 0.6795499324798584, "learning_rate": 7.852051494827722e-05, "loss": 0.0363, "step": 68290 }, { "grad_norm": 0.38696715235710144, "learning_rate": 7.851372408938985e-05, "loss": 0.0429, "step": 68300 }, { "grad_norm": 0.8216124773025513, "learning_rate": 7.8506932450948e-05, "loss": 0.0419, "step": 68310 }, { "grad_norm": 0.9558783173561096, "learning_rate": 7.85001400331373e-05, "loss": 0.0487, "step": 68320 }, { "grad_norm": 0.9866251945495605, "learning_rate": 7.849334683614349e-05, "loss": 0.0488, "step": 68330 }, { "grad_norm": 0.5721470713615417, "learning_rate": 7.848655286015229e-05, "loss": 0.0411, "step": 68340 }, { "grad_norm": 0.5928543210029602, "learning_rate": 7.847975810534943e-05, "loss": 0.0448, "step": 68350 }, { "grad_norm": 2.4707937240600586, "learning_rate": 7.847296257192069e-05, "loss": 0.052, "step": 68360 }, { "grad_norm": 1.5536571741104126, "learning_rate": 7.846616626005183e-05, "loss": 0.0454, "step": 68370 }, { "grad_norm": 0.6224589347839355, "learning_rate": 7.845936916992869e-05, "loss": 0.0462, "step": 68380 }, { "grad_norm": 0.5244936347007751, "learning_rate": 7.845257130173709e-05, "loss": 0.0374, "step": 68390 }, { "grad_norm": 0.5977446436882019, "learning_rate": 7.844577265566289e-05, "loss": 0.0451, "step": 68400 }, { "grad_norm": 0.5160555243492126, "learning_rate": 7.843897323189194e-05, "loss": 0.0509, "step": 68410 }, { "grad_norm": 0.779833972454071, "learning_rate": 7.843217303061014e-05, "loss": 0.0497, "step": 68420 }, { "grad_norm": 0.5048444867134094, "learning_rate": 7.842537205200342e-05, "loss": 0.0399, "step": 68430 }, { "grad_norm": 0.5553749799728394, "learning_rate": 7.841857029625771e-05, "loss": 0.0463, "step": 68440 }, { "grad_norm": 0.617489755153656, "learning_rate": 7.841176776355897e-05, "loss": 0.0512, "step": 68450 }, { "grad_norm": 0.420218825340271, "learning_rate": 7.840496445409316e-05, "loss": 0.0348, "step": 68460 }, { "grad_norm": 0.6513285636901855, "learning_rate": 7.839816036804628e-05, "loss": 0.0425, "step": 68470 }, { "grad_norm": 0.47968733310699463, "learning_rate": 7.83913555056044e-05, "loss": 0.0327, "step": 68480 }, { "grad_norm": 1.8218529224395752, "learning_rate": 7.83845498669535e-05, "loss": 0.0378, "step": 68490 }, { "grad_norm": 1.014994740486145, "learning_rate": 7.837774345227966e-05, "loss": 0.0415, "step": 68500 }, { "grad_norm": 0.45720571279525757, "learning_rate": 7.8370936261769e-05, "loss": 0.0586, "step": 68510 }, { "grad_norm": 0.3312954604625702, "learning_rate": 7.83641282956076e-05, "loss": 0.0452, "step": 68520 }, { "grad_norm": 1.1064616441726685, "learning_rate": 7.835731955398158e-05, "loss": 0.0381, "step": 68530 }, { "grad_norm": 1.8052163124084473, "learning_rate": 7.83505100370771e-05, "loss": 0.0605, "step": 68540 }, { "grad_norm": 1.9610251188278198, "learning_rate": 7.834369974508033e-05, "loss": 0.0468, "step": 68550 }, { "grad_norm": 0.5402439832687378, "learning_rate": 7.833688867817744e-05, "loss": 0.0345, "step": 68560 }, { "grad_norm": 0.4772694706916809, "learning_rate": 7.833007683655466e-05, "loss": 0.0462, "step": 68570 }, { "grad_norm": 1.0361279249191284, "learning_rate": 7.832326422039823e-05, "loss": 0.0371, "step": 68580 }, { "grad_norm": 0.54310542345047, "learning_rate": 7.831645082989441e-05, "loss": 0.0362, "step": 68590 }, { "grad_norm": 0.4116908013820648, "learning_rate": 7.830963666522946e-05, "loss": 0.0342, "step": 68600 }, { "grad_norm": 1.0263636112213135, "learning_rate": 7.830282172658967e-05, "loss": 0.0411, "step": 68610 }, { "grad_norm": 0.47921597957611084, "learning_rate": 7.829600601416136e-05, "loss": 0.0446, "step": 68620 }, { "grad_norm": 0.42395150661468506, "learning_rate": 7.82891895281309e-05, "loss": 0.035, "step": 68630 }, { "grad_norm": 0.9787567853927612, "learning_rate": 7.82823722686846e-05, "loss": 0.0409, "step": 68640 }, { "grad_norm": 0.6126865744590759, "learning_rate": 7.827555423600889e-05, "loss": 0.0392, "step": 68650 }, { "grad_norm": 0.6104598045349121, "learning_rate": 7.826873543029013e-05, "loss": 0.0353, "step": 68660 }, { "grad_norm": 0.5381773114204407, "learning_rate": 7.826191585171477e-05, "loss": 0.041, "step": 68670 }, { "grad_norm": 0.30715465545654297, "learning_rate": 7.825509550046926e-05, "loss": 0.0374, "step": 68680 }, { "grad_norm": 0.657897412776947, "learning_rate": 7.824827437674002e-05, "loss": 0.0466, "step": 68690 }, { "grad_norm": 0.4083249270915985, "learning_rate": 7.824145248071361e-05, "loss": 0.0514, "step": 68700 }, { "grad_norm": 0.37947434186935425, "learning_rate": 7.823462981257648e-05, "loss": 0.0354, "step": 68710 }, { "grad_norm": 0.6846369504928589, "learning_rate": 7.822780637251518e-05, "loss": 0.0613, "step": 68720 }, { "grad_norm": 0.8674473166465759, "learning_rate": 7.822098216071626e-05, "loss": 0.0513, "step": 68730 }, { "grad_norm": 0.6650852560997009, "learning_rate": 7.821415717736629e-05, "loss": 0.0506, "step": 68740 }, { "grad_norm": 0.5424944758415222, "learning_rate": 7.820733142265186e-05, "loss": 0.0308, "step": 68750 }, { "grad_norm": 0.4266608655452728, "learning_rate": 7.820050489675959e-05, "loss": 0.0489, "step": 68760 }, { "grad_norm": 0.6355850100517273, "learning_rate": 7.81936775998761e-05, "loss": 0.0477, "step": 68770 }, { "grad_norm": 0.6743457317352295, "learning_rate": 7.818684953218807e-05, "loss": 0.0573, "step": 68780 }, { "grad_norm": 0.4666983485221863, "learning_rate": 7.818002069388215e-05, "loss": 0.0373, "step": 68790 }, { "grad_norm": 0.5391673445701599, "learning_rate": 7.817319108514507e-05, "loss": 0.0386, "step": 68800 }, { "grad_norm": 0.41833657026290894, "learning_rate": 7.816636070616352e-05, "loss": 0.0295, "step": 68810 }, { "grad_norm": 0.4918920397758484, "learning_rate": 7.815952955712425e-05, "loss": 0.0453, "step": 68820 }, { "grad_norm": 0.5980530977249146, "learning_rate": 7.815269763821401e-05, "loss": 0.0408, "step": 68830 }, { "grad_norm": 0.476874977350235, "learning_rate": 7.814586494961961e-05, "loss": 0.0377, "step": 68840 }, { "grad_norm": 0.38501182198524475, "learning_rate": 7.813903149152783e-05, "loss": 0.0322, "step": 68850 }, { "grad_norm": 0.3242429792881012, "learning_rate": 7.813219726412549e-05, "loss": 0.0396, "step": 68860 }, { "grad_norm": 0.5289587378501892, "learning_rate": 7.812536226759945e-05, "loss": 0.0378, "step": 68870 }, { "grad_norm": 0.6155468821525574, "learning_rate": 7.811852650213658e-05, "loss": 0.0546, "step": 68880 }, { "grad_norm": 0.8607448935508728, "learning_rate": 7.811168996792376e-05, "loss": 0.0619, "step": 68890 }, { "grad_norm": 0.6108772158622742, "learning_rate": 7.810485266514789e-05, "loss": 0.042, "step": 68900 }, { "grad_norm": 0.5787919163703918, "learning_rate": 7.809801459399592e-05, "loss": 0.0338, "step": 68910 }, { "grad_norm": 0.3357429504394531, "learning_rate": 7.809117575465476e-05, "loss": 0.038, "step": 68920 }, { "grad_norm": 0.5586287975311279, "learning_rate": 7.808433614731144e-05, "loss": 0.0507, "step": 68930 }, { "grad_norm": 0.86362224817276, "learning_rate": 7.80774957721529e-05, "loss": 0.0441, "step": 68940 }, { "grad_norm": 0.4267280697822571, "learning_rate": 7.807065462936618e-05, "loss": 0.0306, "step": 68950 }, { "grad_norm": 2.5053322315216064, "learning_rate": 7.80638127191383e-05, "loss": 0.0395, "step": 68960 }, { "grad_norm": 0.49693697690963745, "learning_rate": 7.805697004165633e-05, "loss": 0.0448, "step": 68970 }, { "grad_norm": 2.3983609676361084, "learning_rate": 7.805012659710734e-05, "loss": 0.0308, "step": 68980 }, { "grad_norm": 0.6398633122444153, "learning_rate": 7.804328238567844e-05, "loss": 0.039, "step": 68990 }, { "grad_norm": 0.2528798282146454, "learning_rate": 7.803643740755672e-05, "loss": 0.0474, "step": 69000 }, { "grad_norm": 1.0164780616760254, "learning_rate": 7.802959166292932e-05, "loss": 0.0468, "step": 69010 }, { "grad_norm": 0.7475897669792175, "learning_rate": 7.802274515198343e-05, "loss": 0.0527, "step": 69020 }, { "grad_norm": 0.6824316382408142, "learning_rate": 7.801589787490622e-05, "loss": 0.0407, "step": 69030 }, { "grad_norm": 0.8659976720809937, "learning_rate": 7.800904983188488e-05, "loss": 0.0484, "step": 69040 }, { "grad_norm": 0.5513163208961487, "learning_rate": 7.800220102310665e-05, "loss": 0.0492, "step": 69050 }, { "grad_norm": 0.5152549743652344, "learning_rate": 7.799535144875875e-05, "loss": 0.0413, "step": 69060 }, { "grad_norm": 0.6702459454536438, "learning_rate": 7.798850110902847e-05, "loss": 0.0368, "step": 69070 }, { "grad_norm": 0.40035828948020935, "learning_rate": 7.798165000410308e-05, "loss": 0.0328, "step": 69080 }, { "grad_norm": 0.5397055745124817, "learning_rate": 7.797479813416989e-05, "loss": 0.0335, "step": 69090 }, { "grad_norm": 0.47726327180862427, "learning_rate": 7.796794549941622e-05, "loss": 0.0403, "step": 69100 }, { "grad_norm": 0.6484902501106262, "learning_rate": 7.796109210002945e-05, "loss": 0.0275, "step": 69110 }, { "grad_norm": 0.5657492876052856, "learning_rate": 7.795423793619692e-05, "loss": 0.0335, "step": 69120 }, { "grad_norm": 0.6733711957931519, "learning_rate": 7.7947383008106e-05, "loss": 0.0414, "step": 69130 }, { "grad_norm": 0.38625073432922363, "learning_rate": 7.794052731594415e-05, "loss": 0.0416, "step": 69140 }, { "grad_norm": 1.642647624015808, "learning_rate": 7.793367085989877e-05, "loss": 0.0459, "step": 69150 }, { "grad_norm": 0.7372954487800598, "learning_rate": 7.792681364015735e-05, "loss": 0.042, "step": 69160 }, { "grad_norm": 0.5923042297363281, "learning_rate": 7.791995565690731e-05, "loss": 0.0463, "step": 69170 }, { "grad_norm": 1.5571519136428833, "learning_rate": 7.791309691033619e-05, "loss": 0.0355, "step": 69180 }, { "grad_norm": 1.2408121824264526, "learning_rate": 7.790623740063147e-05, "loss": 0.0643, "step": 69190 }, { "grad_norm": 0.9808353185653687, "learning_rate": 7.78993771279807e-05, "loss": 0.0372, "step": 69200 }, { "grad_norm": 1.1412361860275269, "learning_rate": 7.789251609257145e-05, "loss": 0.0593, "step": 69210 }, { "grad_norm": 0.5215720534324646, "learning_rate": 7.788565429459129e-05, "loss": 0.0458, "step": 69220 }, { "grad_norm": 0.6215035915374756, "learning_rate": 7.787879173422782e-05, "loss": 0.0421, "step": 69230 }, { "grad_norm": 0.7689270377159119, "learning_rate": 7.787192841166866e-05, "loss": 0.0326, "step": 69240 }, { "grad_norm": 0.6245806813240051, "learning_rate": 7.786506432710144e-05, "loss": 0.0427, "step": 69250 }, { "grad_norm": 0.68670654296875, "learning_rate": 7.785819948071382e-05, "loss": 0.0388, "step": 69260 }, { "grad_norm": 0.5313308835029602, "learning_rate": 7.78513338726935e-05, "loss": 0.0417, "step": 69270 }, { "grad_norm": 0.9072607755661011, "learning_rate": 7.784446750322818e-05, "loss": 0.0476, "step": 69280 }, { "grad_norm": 0.8891870975494385, "learning_rate": 7.783760037250558e-05, "loss": 0.0529, "step": 69290 }, { "grad_norm": 0.6464327573776245, "learning_rate": 7.783073248071344e-05, "loss": 0.0424, "step": 69300 }, { "grad_norm": 0.767043948173523, "learning_rate": 7.782386382803954e-05, "loss": 0.0351, "step": 69310 }, { "grad_norm": 0.5970797538757324, "learning_rate": 7.781699441467163e-05, "loss": 0.0391, "step": 69320 }, { "grad_norm": 0.48440104722976685, "learning_rate": 7.781012424079757e-05, "loss": 0.0367, "step": 69330 }, { "grad_norm": 1.1120589971542358, "learning_rate": 7.780325330660516e-05, "loss": 0.0422, "step": 69340 }, { "grad_norm": 0.47951048612594604, "learning_rate": 7.779638161228225e-05, "loss": 0.043, "step": 69350 }, { "grad_norm": 0.5723100900650024, "learning_rate": 7.778950915801669e-05, "loss": 0.0361, "step": 69360 }, { "grad_norm": 0.48479804396629333, "learning_rate": 7.778263594399639e-05, "loss": 0.0392, "step": 69370 }, { "grad_norm": 0.5657750368118286, "learning_rate": 7.777576197040927e-05, "loss": 0.0472, "step": 69380 }, { "grad_norm": 0.6804317235946655, "learning_rate": 7.776888723744327e-05, "loss": 0.0401, "step": 69390 }, { "grad_norm": 0.61316978931427, "learning_rate": 7.77620117452863e-05, "loss": 0.0424, "step": 69400 }, { "grad_norm": 0.6389497518539429, "learning_rate": 7.775513549412637e-05, "loss": 0.0465, "step": 69410 }, { "grad_norm": 0.33550891280174255, "learning_rate": 7.774825848415146e-05, "loss": 0.0332, "step": 69420 }, { "grad_norm": 1.6541122198104858, "learning_rate": 7.774138071554958e-05, "loss": 0.0677, "step": 69430 }, { "grad_norm": 0.7734764218330383, "learning_rate": 7.773450218850878e-05, "loss": 0.0325, "step": 69440 }, { "grad_norm": 0.47267159819602966, "learning_rate": 7.772762290321711e-05, "loss": 0.0352, "step": 69450 }, { "grad_norm": 0.5070846080780029, "learning_rate": 7.772074285986263e-05, "loss": 0.0385, "step": 69460 }, { "grad_norm": 3.523963212966919, "learning_rate": 7.77138620586335e-05, "loss": 0.0376, "step": 69470 }, { "grad_norm": 0.5594981908798218, "learning_rate": 7.770698049971776e-05, "loss": 0.0434, "step": 69480 }, { "grad_norm": 0.5503819584846497, "learning_rate": 7.770009818330359e-05, "loss": 0.042, "step": 69490 }, { "grad_norm": 0.8246371746063232, "learning_rate": 7.769321510957913e-05, "loss": 0.0402, "step": 69500 }, { "grad_norm": 1.233848214149475, "learning_rate": 7.768633127873257e-05, "loss": 0.0661, "step": 69510 }, { "grad_norm": 0.40687093138694763, "learning_rate": 7.767944669095213e-05, "loss": 0.0308, "step": 69520 }, { "grad_norm": 0.5855278968811035, "learning_rate": 7.767256134642602e-05, "loss": 0.0318, "step": 69530 }, { "grad_norm": 0.4135980010032654, "learning_rate": 7.766567524534248e-05, "loss": 0.0284, "step": 69540 }, { "grad_norm": 0.809014618396759, "learning_rate": 7.765878838788976e-05, "loss": 0.0594, "step": 69550 }, { "grad_norm": 0.5789884328842163, "learning_rate": 7.765190077425615e-05, "loss": 0.0441, "step": 69560 }, { "grad_norm": 0.5818321108818054, "learning_rate": 7.764501240462998e-05, "loss": 0.0382, "step": 69570 }, { "grad_norm": 0.5403252840042114, "learning_rate": 7.763812327919953e-05, "loss": 0.0408, "step": 69580 }, { "grad_norm": 1.239856481552124, "learning_rate": 7.76312333981532e-05, "loss": 0.0391, "step": 69590 }, { "grad_norm": 0.5419759154319763, "learning_rate": 7.762434276167931e-05, "loss": 0.0471, "step": 69600 }, { "grad_norm": 0.5010837912559509, "learning_rate": 7.761745136996628e-05, "loss": 0.0313, "step": 69610 }, { "grad_norm": 0.4631800055503845, "learning_rate": 7.76105592232025e-05, "loss": 0.0316, "step": 69620 }, { "grad_norm": 0.32572513818740845, "learning_rate": 7.76036663215764e-05, "loss": 0.0401, "step": 69630 }, { "grad_norm": 0.6953436732292175, "learning_rate": 7.759677266527644e-05, "loss": 0.0381, "step": 69640 }, { "grad_norm": 1.2216509580612183, "learning_rate": 7.758987825449108e-05, "loss": 0.0397, "step": 69650 }, { "grad_norm": 0.4642048478126526, "learning_rate": 7.75829830894088e-05, "loss": 0.0419, "step": 69660 }, { "grad_norm": 0.5915316343307495, "learning_rate": 7.757608717021816e-05, "loss": 0.0325, "step": 69670 }, { "grad_norm": 0.6599184274673462, "learning_rate": 7.756919049710762e-05, "loss": 0.0391, "step": 69680 }, { "grad_norm": 0.49027055501937866, "learning_rate": 7.75622930702658e-05, "loss": 0.0492, "step": 69690 }, { "grad_norm": 0.5296524167060852, "learning_rate": 7.755539488988122e-05, "loss": 0.0367, "step": 69700 }, { "grad_norm": 0.6775345802307129, "learning_rate": 7.75484959561425e-05, "loss": 0.0396, "step": 69710 }, { "grad_norm": 0.5592857599258423, "learning_rate": 7.754159626923824e-05, "loss": 0.0359, "step": 69720 }, { "grad_norm": 1.040100336074829, "learning_rate": 7.753469582935711e-05, "loss": 0.0496, "step": 69730 }, { "grad_norm": 0.4440348744392395, "learning_rate": 7.752779463668773e-05, "loss": 0.0327, "step": 69740 }, { "grad_norm": 0.35592252016067505, "learning_rate": 7.752089269141878e-05, "loss": 0.0349, "step": 69750 }, { "grad_norm": 0.5541458129882812, "learning_rate": 7.751398999373898e-05, "loss": 0.0404, "step": 69760 }, { "grad_norm": 0.6278367042541504, "learning_rate": 7.7507086543837e-05, "loss": 0.0595, "step": 69770 }, { "grad_norm": 1.063450574874878, "learning_rate": 7.750018234190163e-05, "loss": 0.0458, "step": 69780 }, { "grad_norm": 0.45330002903938293, "learning_rate": 7.74932773881216e-05, "loss": 0.0376, "step": 69790 }, { "grad_norm": 0.5002448558807373, "learning_rate": 7.74863716826857e-05, "loss": 0.0404, "step": 69800 }, { "grad_norm": 0.7069982886314392, "learning_rate": 7.747946522578272e-05, "loss": 0.0459, "step": 69810 }, { "grad_norm": 0.45005834102630615, "learning_rate": 7.747255801760149e-05, "loss": 0.0378, "step": 69820 }, { "grad_norm": 0.5909280180931091, "learning_rate": 7.746565005833083e-05, "loss": 0.0438, "step": 69830 }, { "grad_norm": 0.4119099974632263, "learning_rate": 7.745874134815962e-05, "loss": 0.0411, "step": 69840 }, { "grad_norm": 1.7285014390945435, "learning_rate": 7.745183188727674e-05, "loss": 0.0347, "step": 69850 }, { "grad_norm": 0.7770113348960876, "learning_rate": 7.74449216758711e-05, "loss": 0.0285, "step": 69860 }, { "grad_norm": 0.4184425473213196, "learning_rate": 7.743801071413161e-05, "loss": 0.039, "step": 69870 }, { "grad_norm": 0.884398341178894, "learning_rate": 7.743109900224721e-05, "loss": 0.0393, "step": 69880 }, { "grad_norm": 0.5130338668823242, "learning_rate": 7.742418654040686e-05, "loss": 0.0523, "step": 69890 }, { "grad_norm": 0.6677442789077759, "learning_rate": 7.741727332879956e-05, "loss": 0.0385, "step": 69900 }, { "grad_norm": 0.6148785948753357, "learning_rate": 7.74103593676143e-05, "loss": 0.0425, "step": 69910 }, { "grad_norm": 0.6451635360717773, "learning_rate": 7.740344465704013e-05, "loss": 0.0522, "step": 69920 }, { "grad_norm": 0.4270787835121155, "learning_rate": 7.739652919726606e-05, "loss": 0.0431, "step": 69930 }, { "grad_norm": 0.5096467137336731, "learning_rate": 7.73896129884812e-05, "loss": 0.0364, "step": 69940 }, { "grad_norm": 0.4636082649230957, "learning_rate": 7.738269603087458e-05, "loss": 0.047, "step": 69950 }, { "grad_norm": 0.3337055444717407, "learning_rate": 7.737577832463534e-05, "loss": 0.0477, "step": 69960 }, { "grad_norm": 0.36960116028785706, "learning_rate": 7.736885986995262e-05, "loss": 0.0407, "step": 69970 }, { "grad_norm": 1.0661003589630127, "learning_rate": 7.736194066701555e-05, "loss": 0.0322, "step": 69980 }, { "grad_norm": 0.36400848627090454, "learning_rate": 7.73550207160133e-05, "loss": 0.0364, "step": 69990 }, { "grad_norm": 1.7495898008346558, "learning_rate": 7.734810001713507e-05, "loss": 0.0438, "step": 70000 }, { "grad_norm": 0.6301487684249878, "learning_rate": 7.734117857057005e-05, "loss": 0.0506, "step": 70010 }, { "grad_norm": 0.5289168953895569, "learning_rate": 7.733425637650747e-05, "loss": 0.0351, "step": 70020 }, { "grad_norm": 0.6439501047134399, "learning_rate": 7.732733343513661e-05, "loss": 0.0384, "step": 70030 }, { "grad_norm": 0.5550447702407837, "learning_rate": 7.732040974664671e-05, "loss": 0.0284, "step": 70040 }, { "grad_norm": 0.4664800763130188, "learning_rate": 7.731348531122706e-05, "loss": 0.0538, "step": 70050 }, { "grad_norm": 1.8585445880889893, "learning_rate": 7.730656012906701e-05, "loss": 0.0504, "step": 70060 }, { "grad_norm": 0.47613221406936646, "learning_rate": 7.729963420035584e-05, "loss": 0.0389, "step": 70070 }, { "grad_norm": 1.7017288208007812, "learning_rate": 7.729270752528295e-05, "loss": 0.0498, "step": 70080 }, { "grad_norm": 0.8896299004554749, "learning_rate": 7.728578010403766e-05, "loss": 0.0354, "step": 70090 }, { "grad_norm": 0.6703274250030518, "learning_rate": 7.727885193680942e-05, "loss": 0.0355, "step": 70100 }, { "grad_norm": 0.8260025978088379, "learning_rate": 7.727192302378761e-05, "loss": 0.0444, "step": 70110 }, { "grad_norm": 0.6006300449371338, "learning_rate": 7.726499336516165e-05, "loss": 0.0371, "step": 70120 }, { "grad_norm": 0.5734149217605591, "learning_rate": 7.725806296112104e-05, "loss": 0.0441, "step": 70130 }, { "grad_norm": 0.7611039280891418, "learning_rate": 7.725113181185521e-05, "loss": 0.0398, "step": 70140 }, { "grad_norm": 1.0383145809173584, "learning_rate": 7.724419991755367e-05, "loss": 0.058, "step": 70150 }, { "grad_norm": 0.6072383522987366, "learning_rate": 7.723726727840596e-05, "loss": 0.0399, "step": 70160 }, { "grad_norm": 0.6259773373603821, "learning_rate": 7.723033389460159e-05, "loss": 0.0434, "step": 70170 }, { "grad_norm": 0.7584181427955627, "learning_rate": 7.722339976633012e-05, "loss": 0.0399, "step": 70180 }, { "grad_norm": 0.6211164593696594, "learning_rate": 7.721646489378112e-05, "loss": 0.0444, "step": 70190 }, { "grad_norm": 0.6136632561683655, "learning_rate": 7.72095292771442e-05, "loss": 0.05, "step": 70200 }, { "grad_norm": 0.358521431684494, "learning_rate": 7.720259291660898e-05, "loss": 0.0302, "step": 70210 }, { "grad_norm": 0.3821512758731842, "learning_rate": 7.719565581236508e-05, "loss": 0.0444, "step": 70220 }, { "grad_norm": 0.521308422088623, "learning_rate": 7.718871796460217e-05, "loss": 0.0346, "step": 70230 }, { "grad_norm": 0.5138885974884033, "learning_rate": 7.718177937350992e-05, "loss": 0.0422, "step": 70240 }, { "grad_norm": 0.619976282119751, "learning_rate": 7.717484003927804e-05, "loss": 0.04, "step": 70250 }, { "grad_norm": 0.7575899958610535, "learning_rate": 7.716789996209623e-05, "loss": 0.0463, "step": 70260 }, { "grad_norm": 0.49892425537109375, "learning_rate": 7.716095914215426e-05, "loss": 0.0543, "step": 70270 }, { "grad_norm": 0.8492416143417358, "learning_rate": 7.715401757964184e-05, "loss": 0.0681, "step": 70280 }, { "grad_norm": 0.4480989873409271, "learning_rate": 7.714707527474882e-05, "loss": 0.0377, "step": 70290 }, { "grad_norm": 0.6164549589157104, "learning_rate": 7.714013222766494e-05, "loss": 0.0388, "step": 70300 }, { "grad_norm": 0.8687291145324707, "learning_rate": 7.713318843858006e-05, "loss": 0.036, "step": 70310 }, { "grad_norm": 0.8686472177505493, "learning_rate": 7.712624390768397e-05, "loss": 0.0459, "step": 70320 }, { "grad_norm": 0.6122174263000488, "learning_rate": 7.711929863516658e-05, "loss": 0.0461, "step": 70330 }, { "grad_norm": 0.548553466796875, "learning_rate": 7.711235262121774e-05, "loss": 0.0381, "step": 70340 }, { "grad_norm": 0.47075897455215454, "learning_rate": 7.710540586602739e-05, "loss": 0.0374, "step": 70350 }, { "grad_norm": 0.5036875009536743, "learning_rate": 7.709845836978542e-05, "loss": 0.0382, "step": 70360 }, { "grad_norm": 0.4094997048377991, "learning_rate": 7.709151013268177e-05, "loss": 0.0434, "step": 70370 }, { "grad_norm": 0.4752804636955261, "learning_rate": 7.70845611549064e-05, "loss": 0.0375, "step": 70380 }, { "grad_norm": 0.425473153591156, "learning_rate": 7.707761143664933e-05, "loss": 0.039, "step": 70390 }, { "grad_norm": 0.7096099257469177, "learning_rate": 7.707066097810052e-05, "loss": 0.0437, "step": 70400 }, { "grad_norm": 0.5050592422485352, "learning_rate": 7.706370977945e-05, "loss": 0.0808, "step": 70410 }, { "grad_norm": 0.49912357330322266, "learning_rate": 7.705675784088784e-05, "loss": 0.0625, "step": 70420 }, { "grad_norm": 0.49362897872924805, "learning_rate": 7.704980516260407e-05, "loss": 0.0357, "step": 70430 }, { "grad_norm": 1.0015586614608765, "learning_rate": 7.70428517447888e-05, "loss": 0.0357, "step": 70440 }, { "grad_norm": 0.5533744096755981, "learning_rate": 7.703589758763212e-05, "loss": 0.0461, "step": 70450 }, { "grad_norm": 0.5019795894622803, "learning_rate": 7.702894269132417e-05, "loss": 0.0367, "step": 70460 }, { "grad_norm": 0.8960719704627991, "learning_rate": 7.702198705605506e-05, "loss": 0.0404, "step": 70470 }, { "grad_norm": 0.4830658435821533, "learning_rate": 7.701503068201499e-05, "loss": 0.0312, "step": 70480 }, { "grad_norm": 0.3991551995277405, "learning_rate": 7.700807356939411e-05, "loss": 0.0421, "step": 70490 }, { "grad_norm": 0.9326977729797363, "learning_rate": 7.700111571838266e-05, "loss": 0.0488, "step": 70500 }, { "grad_norm": 0.8311342000961304, "learning_rate": 7.699415712917087e-05, "loss": 0.0456, "step": 70510 }, { "grad_norm": 1.8236260414123535, "learning_rate": 7.698719780194892e-05, "loss": 0.0381, "step": 70520 }, { "grad_norm": 0.5744238495826721, "learning_rate": 7.698023773690714e-05, "loss": 0.0522, "step": 70530 }, { "grad_norm": 0.42759156227111816, "learning_rate": 7.697327693423581e-05, "loss": 0.0413, "step": 70540 }, { "grad_norm": 0.9700294733047485, "learning_rate": 7.696631539412523e-05, "loss": 0.0576, "step": 70550 }, { "grad_norm": 1.392796516418457, "learning_rate": 7.695935311676568e-05, "loss": 0.0589, "step": 70560 }, { "grad_norm": 0.6745227575302124, "learning_rate": 7.695239010234758e-05, "loss": 0.0401, "step": 70570 }, { "grad_norm": 0.5584630370140076, "learning_rate": 7.694542635106123e-05, "loss": 0.0478, "step": 70580 }, { "grad_norm": 0.5860633254051208, "learning_rate": 7.693846186309707e-05, "loss": 0.0533, "step": 70590 }, { "grad_norm": 0.7202990055084229, "learning_rate": 7.693149663864548e-05, "loss": 0.0455, "step": 70600 }, { "grad_norm": 0.5651992559432983, "learning_rate": 7.692453067789687e-05, "loss": 0.0373, "step": 70610 }, { "grad_norm": 0.6315271258354187, "learning_rate": 7.691756398104173e-05, "loss": 0.0438, "step": 70620 }, { "grad_norm": 0.8366517424583435, "learning_rate": 7.691059654827049e-05, "loss": 0.0395, "step": 70630 }, { "grad_norm": 0.5095695853233337, "learning_rate": 7.690362837977365e-05, "loss": 0.0475, "step": 70640 }, { "grad_norm": 0.5329989194869995, "learning_rate": 7.689665947574173e-05, "loss": 0.0444, "step": 70650 }, { "grad_norm": 0.6476349234580994, "learning_rate": 7.688968983636522e-05, "loss": 0.0296, "step": 70660 }, { "grad_norm": 0.5771377086639404, "learning_rate": 7.688271946183471e-05, "loss": 0.0351, "step": 70670 }, { "grad_norm": 1.563317060470581, "learning_rate": 7.687574835234075e-05, "loss": 0.0396, "step": 70680 }, { "grad_norm": 0.5807003378868103, "learning_rate": 7.686877650807394e-05, "loss": 0.0401, "step": 70690 }, { "grad_norm": 0.46310579776763916, "learning_rate": 7.686180392922485e-05, "loss": 0.0396, "step": 70700 }, { "grad_norm": 0.3773670494556427, "learning_rate": 7.685483061598413e-05, "loss": 0.0496, "step": 70710 }, { "grad_norm": 4.268719673156738, "learning_rate": 7.684785656854246e-05, "loss": 0.0424, "step": 70720 }, { "grad_norm": 0.6560173034667969, "learning_rate": 7.684088178709045e-05, "loss": 0.0442, "step": 70730 }, { "grad_norm": 1.015876293182373, "learning_rate": 7.683390627181881e-05, "loss": 0.0458, "step": 70740 }, { "grad_norm": 1.0849640369415283, "learning_rate": 7.682693002291827e-05, "loss": 0.0386, "step": 70750 }, { "grad_norm": 0.5790315270423889, "learning_rate": 7.681995304057956e-05, "loss": 0.0482, "step": 70760 }, { "grad_norm": 0.41781991720199585, "learning_rate": 7.681297532499337e-05, "loss": 0.0597, "step": 70770 }, { "grad_norm": 0.6716398000717163, "learning_rate": 7.680599687635055e-05, "loss": 0.0489, "step": 70780 }, { "grad_norm": 0.8502123355865479, "learning_rate": 7.679901769484179e-05, "loss": 0.0413, "step": 70790 }, { "grad_norm": 0.5571815967559814, "learning_rate": 7.6792037780658e-05, "loss": 0.0504, "step": 70800 }, { "grad_norm": 0.7049975991249084, "learning_rate": 7.678505713398994e-05, "loss": 0.039, "step": 70810 }, { "grad_norm": 0.43629011511802673, "learning_rate": 7.67780757550285e-05, "loss": 0.0361, "step": 70820 }, { "grad_norm": 0.6109269261360168, "learning_rate": 7.67710936439645e-05, "loss": 0.0498, "step": 70830 }, { "grad_norm": 0.41304317116737366, "learning_rate": 7.67641108009889e-05, "loss": 0.0435, "step": 70840 }, { "grad_norm": 0.653266191482544, "learning_rate": 7.675712722629253e-05, "loss": 0.0442, "step": 70850 }, { "grad_norm": 0.6102824807167053, "learning_rate": 7.675014292006637e-05, "loss": 0.0375, "step": 70860 }, { "grad_norm": 0.3981010913848877, "learning_rate": 7.674315788250134e-05, "loss": 0.0356, "step": 70870 }, { "grad_norm": 0.48269227147102356, "learning_rate": 7.673617211378842e-05, "loss": 0.0425, "step": 70880 }, { "grad_norm": 0.8909006714820862, "learning_rate": 7.672918561411862e-05, "loss": 0.0407, "step": 70890 }, { "grad_norm": 0.4983772039413452, "learning_rate": 7.672219838368291e-05, "loss": 0.0393, "step": 70900 }, { "grad_norm": 0.5983959436416626, "learning_rate": 7.671521042267233e-05, "loss": 0.0467, "step": 70910 }, { "grad_norm": 0.6549118757247925, "learning_rate": 7.670822173127794e-05, "loss": 0.0374, "step": 70920 }, { "grad_norm": 0.543030321598053, "learning_rate": 7.670123230969081e-05, "loss": 0.0303, "step": 70930 }, { "grad_norm": 0.5146280527114868, "learning_rate": 7.669424215810201e-05, "loss": 0.0445, "step": 70940 }, { "grad_norm": 0.6209496855735779, "learning_rate": 7.668725127670267e-05, "loss": 0.0382, "step": 70950 }, { "grad_norm": 0.4884307384490967, "learning_rate": 7.668025966568388e-05, "loss": 0.0419, "step": 70960 }, { "grad_norm": 0.9182809591293335, "learning_rate": 7.667326732523685e-05, "loss": 0.045, "step": 70970 }, { "grad_norm": 11.968158721923828, "learning_rate": 7.666627425555269e-05, "loss": 0.0387, "step": 70980 }, { "grad_norm": 0.5070690512657166, "learning_rate": 7.66592804568226e-05, "loss": 0.0557, "step": 70990 }, { "grad_norm": 0.9823341965675354, "learning_rate": 7.665228592923782e-05, "loss": 0.0477, "step": 71000 }, { "grad_norm": 0.8511924147605896, "learning_rate": 7.664529067298954e-05, "loss": 0.0383, "step": 71010 }, { "grad_norm": 0.6154729127883911, "learning_rate": 7.663829468826905e-05, "loss": 0.0503, "step": 71020 }, { "grad_norm": 0.4117182195186615, "learning_rate": 7.663129797526755e-05, "loss": 0.0428, "step": 71030 }, { "grad_norm": 0.8691420555114746, "learning_rate": 7.662430053417638e-05, "loss": 0.0527, "step": 71040 }, { "grad_norm": 0.6620068550109863, "learning_rate": 7.661730236518685e-05, "loss": 0.0414, "step": 71050 }, { "grad_norm": 0.6874279975891113, "learning_rate": 7.661030346849027e-05, "loss": 0.0369, "step": 71060 }, { "grad_norm": 0.4095863699913025, "learning_rate": 7.6603303844278e-05, "loss": 0.0354, "step": 71070 }, { "grad_norm": 1.262594223022461, "learning_rate": 7.659630349274138e-05, "loss": 0.0495, "step": 71080 }, { "grad_norm": 0.5016098618507385, "learning_rate": 7.658930241407183e-05, "loss": 0.0381, "step": 71090 }, { "grad_norm": 0.5880805850028992, "learning_rate": 7.658230060846072e-05, "loss": 0.0346, "step": 71100 }, { "grad_norm": 1.2326929569244385, "learning_rate": 7.65752980760995e-05, "loss": 0.0398, "step": 71110 }, { "grad_norm": 0.8049737215042114, "learning_rate": 7.656829481717962e-05, "loss": 0.0412, "step": 71120 }, { "grad_norm": 1.282378911972046, "learning_rate": 7.656129083189255e-05, "loss": 0.0443, "step": 71130 }, { "grad_norm": 0.6398971676826477, "learning_rate": 7.655428612042976e-05, "loss": 0.0327, "step": 71140 }, { "grad_norm": 0.9341295957565308, "learning_rate": 7.654728068298277e-05, "loss": 0.033, "step": 71150 }, { "grad_norm": 1.421560525894165, "learning_rate": 7.654027451974309e-05, "loss": 0.0415, "step": 71160 }, { "grad_norm": 0.681393027305603, "learning_rate": 7.653326763090227e-05, "loss": 0.041, "step": 71170 }, { "grad_norm": 1.308119773864746, "learning_rate": 7.652626001665189e-05, "loss": 0.0535, "step": 71180 }, { "grad_norm": 1.5594022274017334, "learning_rate": 7.651925167718353e-05, "loss": 0.0484, "step": 71190 }, { "grad_norm": 1.3067800998687744, "learning_rate": 7.651224261268878e-05, "loss": 0.0599, "step": 71200 }, { "grad_norm": 0.47931113839149475, "learning_rate": 7.650523282335929e-05, "loss": 0.0478, "step": 71210 }, { "grad_norm": 0.4470977783203125, "learning_rate": 7.64982223093867e-05, "loss": 0.045, "step": 71220 }, { "grad_norm": 0.4513147473335266, "learning_rate": 7.649121107096264e-05, "loss": 0.0343, "step": 71230 }, { "grad_norm": 0.8849328756332397, "learning_rate": 7.648419910827884e-05, "loss": 0.0258, "step": 71240 }, { "grad_norm": 0.7592186331748962, "learning_rate": 7.647718642152698e-05, "loss": 0.031, "step": 71250 }, { "grad_norm": 0.5082873106002808, "learning_rate": 7.64701730108988e-05, "loss": 0.0418, "step": 71260 }, { "grad_norm": 0.9057531356811523, "learning_rate": 7.646315887658603e-05, "loss": 0.0373, "step": 71270 }, { "grad_norm": 0.8496782779693604, "learning_rate": 7.645614401878044e-05, "loss": 0.0499, "step": 71280 }, { "grad_norm": 0.9697331190109253, "learning_rate": 7.64491284376738e-05, "loss": 0.0453, "step": 71290 }, { "grad_norm": 0.8626399636268616, "learning_rate": 7.644211213345794e-05, "loss": 0.0529, "step": 71300 }, { "grad_norm": 1.0054346323013306, "learning_rate": 7.643509510632465e-05, "loss": 0.044, "step": 71310 }, { "grad_norm": 0.7645377516746521, "learning_rate": 7.64280773564658e-05, "loss": 0.0474, "step": 71320 }, { "grad_norm": 1.0648893117904663, "learning_rate": 7.642105888407326e-05, "loss": 0.0449, "step": 71330 }, { "grad_norm": 0.8255796432495117, "learning_rate": 7.641403968933887e-05, "loss": 0.0374, "step": 71340 }, { "grad_norm": 0.7640125751495361, "learning_rate": 7.640701977245456e-05, "loss": 0.0423, "step": 71350 }, { "grad_norm": 0.3299263119697571, "learning_rate": 7.639999913361227e-05, "loss": 0.0375, "step": 71360 }, { "grad_norm": 0.47295641899108887, "learning_rate": 7.639297777300389e-05, "loss": 0.0338, "step": 71370 }, { "grad_norm": 0.5699352622032166, "learning_rate": 7.638595569082145e-05, "loss": 0.0396, "step": 71380 }, { "grad_norm": 0.44075584411621094, "learning_rate": 7.637893288725688e-05, "loss": 0.036, "step": 71390 }, { "grad_norm": 0.4633902311325073, "learning_rate": 7.637190936250219e-05, "loss": 0.0403, "step": 71400 }, { "grad_norm": 0.6580774188041687, "learning_rate": 7.63648851167494e-05, "loss": 0.033, "step": 71410 }, { "grad_norm": 0.6841376423835754, "learning_rate": 7.635786015019058e-05, "loss": 0.0398, "step": 71420 }, { "grad_norm": 0.3679279685020447, "learning_rate": 7.635083446301774e-05, "loss": 0.0319, "step": 71430 }, { "grad_norm": 1.727504849433899, "learning_rate": 7.6343808055423e-05, "loss": 0.0364, "step": 71440 }, { "grad_norm": 0.29048341512680054, "learning_rate": 7.633678092759843e-05, "loss": 0.0357, "step": 71450 }, { "grad_norm": 0.5598674416542053, "learning_rate": 7.632975307973618e-05, "loss": 0.0504, "step": 71460 }, { "grad_norm": 0.6933189034461975, "learning_rate": 7.632272451202837e-05, "loss": 0.031, "step": 71470 }, { "grad_norm": 0.9750687479972839, "learning_rate": 7.631569522466715e-05, "loss": 0.0325, "step": 71480 }, { "grad_norm": 0.7476276755332947, "learning_rate": 7.63086652178447e-05, "loss": 0.0468, "step": 71490 }, { "grad_norm": 0.4677172005176544, "learning_rate": 7.630163449175325e-05, "loss": 0.035, "step": 71500 }, { "grad_norm": 0.7439064383506775, "learning_rate": 7.6294603046585e-05, "loss": 0.0377, "step": 71510 }, { "grad_norm": 1.8141502141952515, "learning_rate": 7.628757088253216e-05, "loss": 0.0349, "step": 71520 }, { "grad_norm": 1.16060471534729, "learning_rate": 7.628053799978702e-05, "loss": 0.0339, "step": 71530 }, { "grad_norm": 0.9823324084281921, "learning_rate": 7.627350439854183e-05, "loss": 0.0486, "step": 71540 }, { "grad_norm": 0.35218650102615356, "learning_rate": 7.626647007898891e-05, "loss": 0.0555, "step": 71550 }, { "grad_norm": 0.5992467403411865, "learning_rate": 7.625943504132056e-05, "loss": 0.0453, "step": 71560 }, { "grad_norm": 0.4103071093559265, "learning_rate": 7.625239928572914e-05, "loss": 0.0424, "step": 71570 }, { "grad_norm": 0.3316606283187866, "learning_rate": 7.624536281240698e-05, "loss": 0.0296, "step": 71580 }, { "grad_norm": 0.825143575668335, "learning_rate": 7.623832562154645e-05, "loss": 0.0511, "step": 71590 }, { "grad_norm": 0.42977744340896606, "learning_rate": 7.623128771333995e-05, "loss": 0.0457, "step": 71600 }, { "grad_norm": 0.7741227149963379, "learning_rate": 7.622424908797992e-05, "loss": 0.0627, "step": 71610 }, { "grad_norm": 0.64826500415802, "learning_rate": 7.621720974565874e-05, "loss": 0.04, "step": 71620 }, { "grad_norm": 0.46081045269966125, "learning_rate": 7.621016968656892e-05, "loss": 0.0474, "step": 71630 }, { "grad_norm": 0.3292713165283203, "learning_rate": 7.62031289109029e-05, "loss": 0.046, "step": 71640 }, { "grad_norm": 2.0905067920684814, "learning_rate": 7.619608741885318e-05, "loss": 0.0439, "step": 71650 }, { "grad_norm": 1.1260552406311035, "learning_rate": 7.618904521061227e-05, "loss": 0.0454, "step": 71660 }, { "grad_norm": 0.705406129360199, "learning_rate": 7.618200228637271e-05, "loss": 0.0467, "step": 71670 }, { "grad_norm": 0.7285149097442627, "learning_rate": 7.617495864632703e-05, "loss": 0.0373, "step": 71680 }, { "grad_norm": 0.55559241771698, "learning_rate": 7.616791429066782e-05, "loss": 0.0351, "step": 71690 }, { "grad_norm": 2.1732017993927, "learning_rate": 7.616086921958767e-05, "loss": 0.0396, "step": 71700 }, { "grad_norm": 0.39780449867248535, "learning_rate": 7.615382343327918e-05, "loss": 0.0364, "step": 71710 }, { "grad_norm": 0.5541238784790039, "learning_rate": 7.614677693193499e-05, "loss": 0.035, "step": 71720 }, { "grad_norm": 0.6680925488471985, "learning_rate": 7.613972971574773e-05, "loss": 0.0698, "step": 71730 }, { "grad_norm": 0.8117258548736572, "learning_rate": 7.613268178491009e-05, "loss": 0.0381, "step": 71740 }, { "grad_norm": 0.49486565589904785, "learning_rate": 7.612563313961473e-05, "loss": 0.0409, "step": 71750 }, { "grad_norm": 0.6069406867027283, "learning_rate": 7.61185837800544e-05, "loss": 0.0435, "step": 71760 }, { "grad_norm": 0.9283690452575684, "learning_rate": 7.611153370642178e-05, "loss": 0.034, "step": 71770 }, { "grad_norm": 0.5229734778404236, "learning_rate": 7.610448291890966e-05, "loss": 0.0495, "step": 71780 }, { "grad_norm": 0.8204011917114258, "learning_rate": 7.609743141771076e-05, "loss": 0.0458, "step": 71790 }, { "grad_norm": 0.5822261571884155, "learning_rate": 7.609037920301792e-05, "loss": 0.0426, "step": 71800 }, { "grad_norm": 0.559024453163147, "learning_rate": 7.608332627502387e-05, "loss": 0.0637, "step": 71810 }, { "grad_norm": 1.4836182594299316, "learning_rate": 7.607627263392151e-05, "loss": 0.0388, "step": 71820 }, { "grad_norm": 1.0620843172073364, "learning_rate": 7.606921827990364e-05, "loss": 0.0557, "step": 71830 }, { "grad_norm": 0.477146178483963, "learning_rate": 7.606216321316315e-05, "loss": 0.0354, "step": 71840 }, { "grad_norm": 0.6417636275291443, "learning_rate": 7.605510743389291e-05, "loss": 0.0416, "step": 71850 }, { "grad_norm": 0.7924044728279114, "learning_rate": 7.60480509422858e-05, "loss": 0.0436, "step": 71860 }, { "grad_norm": 0.906408429145813, "learning_rate": 7.604099373853478e-05, "loss": 0.0524, "step": 71870 }, { "grad_norm": 0.7454427480697632, "learning_rate": 7.603393582283275e-05, "loss": 0.0391, "step": 71880 }, { "grad_norm": 1.3651517629623413, "learning_rate": 7.602687719537273e-05, "loss": 0.0359, "step": 71890 }, { "grad_norm": 0.5736173391342163, "learning_rate": 7.601981785634765e-05, "loss": 0.0483, "step": 71900 }, { "grad_norm": 0.7153334021568298, "learning_rate": 7.601275780595052e-05, "loss": 0.0282, "step": 71910 }, { "grad_norm": 0.6689392924308777, "learning_rate": 7.600569704437437e-05, "loss": 0.0361, "step": 71920 }, { "grad_norm": 0.5199954509735107, "learning_rate": 7.599863557181222e-05, "loss": 0.0433, "step": 71930 }, { "grad_norm": 0.6010630130767822, "learning_rate": 7.599157338845717e-05, "loss": 0.0387, "step": 71940 }, { "grad_norm": 0.6015309691429138, "learning_rate": 7.598451049450223e-05, "loss": 0.0325, "step": 71950 }, { "grad_norm": 0.4579901099205017, "learning_rate": 7.597744689014057e-05, "loss": 0.0522, "step": 71960 }, { "grad_norm": 0.30488142371177673, "learning_rate": 7.597038257556525e-05, "loss": 0.0418, "step": 71970 }, { "grad_norm": 0.5813340544700623, "learning_rate": 7.596331755096944e-05, "loss": 0.0309, "step": 71980 }, { "grad_norm": 0.7670179605484009, "learning_rate": 7.595625181654628e-05, "loss": 0.03, "step": 71990 }, { "grad_norm": 0.4716630280017853, "learning_rate": 7.594918537248893e-05, "loss": 0.0581, "step": 72000 }, { "grad_norm": 0.7206284999847412, "learning_rate": 7.594211821899063e-05, "loss": 0.0296, "step": 72010 }, { "grad_norm": 0.377783864736557, "learning_rate": 7.593505035624453e-05, "loss": 0.0424, "step": 72020 }, { "grad_norm": 0.6845493316650391, "learning_rate": 7.592798178444392e-05, "loss": 0.0412, "step": 72030 }, { "grad_norm": 0.7319647073745728, "learning_rate": 7.592091250378201e-05, "loss": 0.0443, "step": 72040 }, { "grad_norm": 0.5815756320953369, "learning_rate": 7.591384251445213e-05, "loss": 0.0409, "step": 72050 }, { "grad_norm": 0.4882560074329376, "learning_rate": 7.590677181664749e-05, "loss": 0.0325, "step": 72060 }, { "grad_norm": 0.49549639225006104, "learning_rate": 7.589970041056144e-05, "loss": 0.0462, "step": 72070 }, { "grad_norm": 0.6009385585784912, "learning_rate": 7.589262829638734e-05, "loss": 0.0322, "step": 72080 }, { "grad_norm": 0.714307427406311, "learning_rate": 7.58855554743185e-05, "loss": 0.0436, "step": 72090 }, { "grad_norm": 0.7410626411437988, "learning_rate": 7.587848194454828e-05, "loss": 0.0366, "step": 72100 }, { "grad_norm": 0.7034116387367249, "learning_rate": 7.587140770727012e-05, "loss": 0.049, "step": 72110 }, { "grad_norm": 1.357900619506836, "learning_rate": 7.586433276267738e-05, "loss": 0.0347, "step": 72120 }, { "grad_norm": 1.0949629545211792, "learning_rate": 7.585725711096348e-05, "loss": 0.0355, "step": 72130 }, { "grad_norm": 0.6145704984664917, "learning_rate": 7.585018075232188e-05, "loss": 0.0467, "step": 72140 }, { "grad_norm": 0.4206896424293518, "learning_rate": 7.584310368694607e-05, "loss": 0.0373, "step": 72150 }, { "grad_norm": 0.7224979996681213, "learning_rate": 7.583602591502952e-05, "loss": 0.0572, "step": 72160 }, { "grad_norm": 0.9756585955619812, "learning_rate": 7.58289474367657e-05, "loss": 0.0382, "step": 72170 }, { "grad_norm": 0.5567701458930969, "learning_rate": 7.582186825234818e-05, "loss": 0.0399, "step": 72180 }, { "grad_norm": 0.6258876919746399, "learning_rate": 7.581478836197049e-05, "loss": 0.0372, "step": 72190 }, { "grad_norm": 1.2408157587051392, "learning_rate": 7.580770776582616e-05, "loss": 0.0413, "step": 72200 }, { "grad_norm": 1.021589994430542, "learning_rate": 7.580062646410882e-05, "loss": 0.0443, "step": 72210 }, { "grad_norm": 0.598712146282196, "learning_rate": 7.579354445701204e-05, "loss": 0.0415, "step": 72220 }, { "grad_norm": 0.5294188857078552, "learning_rate": 7.578646174472943e-05, "loss": 0.051, "step": 72230 }, { "grad_norm": 1.255583643913269, "learning_rate": 7.577937832745466e-05, "loss": 0.0522, "step": 72240 }, { "grad_norm": 0.7328605055809021, "learning_rate": 7.577229420538135e-05, "loss": 0.0481, "step": 72250 }, { "grad_norm": 0.5722854137420654, "learning_rate": 7.576520937870322e-05, "loss": 0.0414, "step": 72260 }, { "grad_norm": 1.0071403980255127, "learning_rate": 7.575812384761394e-05, "loss": 0.0553, "step": 72270 }, { "grad_norm": 0.7497466802597046, "learning_rate": 7.575103761230724e-05, "loss": 0.0457, "step": 72280 }, { "grad_norm": 0.5863698124885559, "learning_rate": 7.574395067297683e-05, "loss": 0.0376, "step": 72290 }, { "grad_norm": 1.5581268072128296, "learning_rate": 7.573686302981648e-05, "loss": 0.0384, "step": 72300 }, { "grad_norm": 0.758482038974762, "learning_rate": 7.572977468301997e-05, "loss": 0.0471, "step": 72310 }, { "grad_norm": 0.7858142852783203, "learning_rate": 7.572268563278109e-05, "loss": 0.0413, "step": 72320 }, { "grad_norm": 0.48763614892959595, "learning_rate": 7.571559587929364e-05, "loss": 0.0374, "step": 72330 }, { "grad_norm": 0.706265926361084, "learning_rate": 7.570850542275146e-05, "loss": 0.0421, "step": 72340 }, { "grad_norm": 0.3858041763305664, "learning_rate": 7.57014142633484e-05, "loss": 0.0362, "step": 72350 }, { "grad_norm": 0.4952319264411926, "learning_rate": 7.569432240127833e-05, "loss": 0.0313, "step": 72360 }, { "grad_norm": 0.7463792562484741, "learning_rate": 7.568722983673516e-05, "loss": 0.0514, "step": 72370 }, { "grad_norm": 0.3246825635433197, "learning_rate": 7.568013656991275e-05, "loss": 0.029, "step": 72380 }, { "grad_norm": 0.7118174433708191, "learning_rate": 7.567304260100505e-05, "loss": 0.0636, "step": 72390 }, { "grad_norm": 0.7742963433265686, "learning_rate": 7.566594793020602e-05, "loss": 0.0452, "step": 72400 }, { "grad_norm": 0.587663471698761, "learning_rate": 7.565885255770964e-05, "loss": 0.044, "step": 72410 }, { "grad_norm": 1.3484876155853271, "learning_rate": 7.565175648370983e-05, "loss": 0.0539, "step": 72420 }, { "grad_norm": 0.5847611427307129, "learning_rate": 7.564465970840067e-05, "loss": 0.0532, "step": 72430 }, { "grad_norm": 1.314242959022522, "learning_rate": 7.563756223197613e-05, "loss": 0.0391, "step": 72440 }, { "grad_norm": 0.506868839263916, "learning_rate": 7.563046405463026e-05, "loss": 0.034, "step": 72450 }, { "grad_norm": 1.0806987285614014, "learning_rate": 7.562336517655714e-05, "loss": 0.0475, "step": 72460 }, { "grad_norm": 0.9217605590820312, "learning_rate": 7.561626559795085e-05, "loss": 0.0336, "step": 72470 }, { "grad_norm": 1.0659550428390503, "learning_rate": 7.560916531900549e-05, "loss": 0.0359, "step": 72480 }, { "grad_norm": 0.5715556740760803, "learning_rate": 7.560206433991515e-05, "loss": 0.0424, "step": 72490 }, { "grad_norm": 0.47816330194473267, "learning_rate": 7.5594962660874e-05, "loss": 0.0374, "step": 72500 }, { "grad_norm": 0.4794807732105255, "learning_rate": 7.558786028207617e-05, "loss": 0.0525, "step": 72510 }, { "grad_norm": 0.3667425215244293, "learning_rate": 7.558075720371587e-05, "loss": 0.0305, "step": 72520 }, { "grad_norm": 0.441989541053772, "learning_rate": 7.557365342598728e-05, "loss": 0.041, "step": 72530 }, { "grad_norm": 0.5164614915847778, "learning_rate": 7.556654894908461e-05, "loss": 0.0336, "step": 72540 }, { "grad_norm": 0.40211182832717896, "learning_rate": 7.55594437732021e-05, "loss": 0.0474, "step": 72550 }, { "grad_norm": 0.6607540845870972, "learning_rate": 7.5552337898534e-05, "loss": 0.0484, "step": 72560 }, { "grad_norm": 1.2129720449447632, "learning_rate": 7.554523132527455e-05, "loss": 0.0628, "step": 72570 }, { "grad_norm": 0.5185006260871887, "learning_rate": 7.553812405361811e-05, "loss": 0.0335, "step": 72580 }, { "grad_norm": 0.5326705574989319, "learning_rate": 7.553101608375894e-05, "loss": 0.0423, "step": 72590 }, { "grad_norm": 0.3214706480503082, "learning_rate": 7.552390741589137e-05, "loss": 0.0358, "step": 72600 }, { "grad_norm": 0.4447595179080963, "learning_rate": 7.551679805020979e-05, "loss": 0.0276, "step": 72610 }, { "grad_norm": 0.7424947619438171, "learning_rate": 7.550968798690852e-05, "loss": 0.0466, "step": 72620 }, { "grad_norm": 0.45807579159736633, "learning_rate": 7.550257722618194e-05, "loss": 0.0352, "step": 72630 }, { "grad_norm": 0.5960859656333923, "learning_rate": 7.549546576822452e-05, "loss": 0.047, "step": 72640 }, { "grad_norm": 0.3867030143737793, "learning_rate": 7.548835361323064e-05, "loss": 0.0373, "step": 72650 }, { "grad_norm": 0.6739728450775146, "learning_rate": 7.548124076139471e-05, "loss": 0.0434, "step": 72660 }, { "grad_norm": 1.811612844467163, "learning_rate": 7.547412721291126e-05, "loss": 0.0275, "step": 72670 }, { "grad_norm": 0.4425847828388214, "learning_rate": 7.546701296797474e-05, "loss": 0.0342, "step": 72680 }, { "grad_norm": 0.3580229580402374, "learning_rate": 7.545989802677966e-05, "loss": 0.0298, "step": 72690 }, { "grad_norm": 0.4585861265659332, "learning_rate": 7.545278238952052e-05, "loss": 0.041, "step": 72700 }, { "grad_norm": 0.6482888460159302, "learning_rate": 7.544566605639188e-05, "loss": 0.048, "step": 72710 }, { "grad_norm": 0.5153226256370544, "learning_rate": 7.543854902758828e-05, "loss": 0.0305, "step": 72720 }, { "grad_norm": 0.32484543323516846, "learning_rate": 7.54314313033043e-05, "loss": 0.0288, "step": 72730 }, { "grad_norm": 1.8797662258148193, "learning_rate": 7.542431288373457e-05, "loss": 0.048, "step": 72740 }, { "grad_norm": 1.0790140628814697, "learning_rate": 7.541719376907366e-05, "loss": 0.0383, "step": 72750 }, { "grad_norm": 1.1638356447219849, "learning_rate": 7.541007395951622e-05, "loss": 0.0502, "step": 72760 }, { "grad_norm": 0.4671931564807892, "learning_rate": 7.540295345525689e-05, "loss": 0.0384, "step": 72770 }, { "grad_norm": 0.6448723673820496, "learning_rate": 7.539583225649039e-05, "loss": 0.0411, "step": 72780 }, { "grad_norm": 0.5871506929397583, "learning_rate": 7.538871036341135e-05, "loss": 0.0431, "step": 72790 }, { "grad_norm": 0.4623662531375885, "learning_rate": 7.538158777621452e-05, "loss": 0.0361, "step": 72800 }, { "grad_norm": 0.8860825300216675, "learning_rate": 7.537446449509461e-05, "loss": 0.044, "step": 72810 }, { "grad_norm": 0.7306115031242371, "learning_rate": 7.536734052024638e-05, "loss": 0.0325, "step": 72820 }, { "grad_norm": 0.45934566855430603, "learning_rate": 7.536021585186456e-05, "loss": 0.0397, "step": 72830 }, { "grad_norm": 1.369850754737854, "learning_rate": 7.5353090490144e-05, "loss": 0.035, "step": 72840 }, { "grad_norm": 0.8060241937637329, "learning_rate": 7.534596443527946e-05, "loss": 0.0351, "step": 72850 }, { "grad_norm": 0.44572338461875916, "learning_rate": 7.53388376874658e-05, "loss": 0.0564, "step": 72860 }, { "grad_norm": 0.8045380711555481, "learning_rate": 7.533171024689781e-05, "loss": 0.0557, "step": 72870 }, { "grad_norm": 0.45236504077911377, "learning_rate": 7.532458211377037e-05, "loss": 0.0302, "step": 72880 }, { "grad_norm": 0.5779260396957397, "learning_rate": 7.531745328827837e-05, "loss": 0.0479, "step": 72890 }, { "grad_norm": 0.5402565598487854, "learning_rate": 7.531032377061673e-05, "loss": 0.0347, "step": 72900 }, { "grad_norm": 0.548890233039856, "learning_rate": 7.530319356098033e-05, "loss": 0.0458, "step": 72910 }, { "grad_norm": 0.47909748554229736, "learning_rate": 7.529606265956412e-05, "loss": 0.0366, "step": 72920 }, { "grad_norm": 0.3961699903011322, "learning_rate": 7.528893106656308e-05, "loss": 0.0275, "step": 72930 }, { "grad_norm": 0.5343101620674133, "learning_rate": 7.528179878217216e-05, "loss": 0.0307, "step": 72940 }, { "grad_norm": 0.5514586567878723, "learning_rate": 7.527466580658635e-05, "loss": 0.0544, "step": 72950 }, { "grad_norm": 0.42694324254989624, "learning_rate": 7.526753214000068e-05, "loss": 0.0496, "step": 72960 }, { "grad_norm": 0.5042679905891418, "learning_rate": 7.526039778261018e-05, "loss": 0.0408, "step": 72970 }, { "grad_norm": 0.4956511855125427, "learning_rate": 7.525326273460986e-05, "loss": 0.0473, "step": 72980 }, { "grad_norm": 0.4554094076156616, "learning_rate": 7.524612699619487e-05, "loss": 0.0318, "step": 72990 }, { "grad_norm": 0.37902429699897766, "learning_rate": 7.523899056756023e-05, "loss": 0.042, "step": 73000 }, { "grad_norm": 0.5251757502555847, "learning_rate": 7.523185344890107e-05, "loss": 0.0339, "step": 73010 }, { "grad_norm": 0.372281938791275, "learning_rate": 7.522471564041251e-05, "loss": 0.0306, "step": 73020 }, { "grad_norm": 1.246297836303711, "learning_rate": 7.521757714228971e-05, "loss": 0.0351, "step": 73030 }, { "grad_norm": 0.3648114800453186, "learning_rate": 7.521043795472781e-05, "loss": 0.0369, "step": 73040 }, { "grad_norm": 0.35380834341049194, "learning_rate": 7.520329807792204e-05, "loss": 0.0374, "step": 73050 }, { "grad_norm": 1.1207246780395508, "learning_rate": 7.519615751206755e-05, "loss": 0.0513, "step": 73060 }, { "grad_norm": 0.6133107542991638, "learning_rate": 7.518901625735957e-05, "loss": 0.0359, "step": 73070 }, { "grad_norm": 0.516654908657074, "learning_rate": 7.518187431399335e-05, "loss": 0.0397, "step": 73080 }, { "grad_norm": 0.49891847372055054, "learning_rate": 7.517473168216415e-05, "loss": 0.0367, "step": 73090 }, { "grad_norm": 0.8596031069755554, "learning_rate": 7.516758836206726e-05, "loss": 0.0332, "step": 73100 }, { "grad_norm": 0.6899697184562683, "learning_rate": 7.516044435389795e-05, "loss": 0.0473, "step": 73110 }, { "grad_norm": 0.5604550242424011, "learning_rate": 7.515329965785154e-05, "loss": 0.0464, "step": 73120 }, { "grad_norm": 0.2979585826396942, "learning_rate": 7.514615427412336e-05, "loss": 0.0296, "step": 73130 }, { "grad_norm": 0.48267364501953125, "learning_rate": 7.513900820290877e-05, "loss": 0.0325, "step": 73140 }, { "grad_norm": 0.27176371216773987, "learning_rate": 7.513186144440315e-05, "loss": 0.031, "step": 73150 }, { "grad_norm": 0.43227800726890564, "learning_rate": 7.512471399880188e-05, "loss": 0.0706, "step": 73160 }, { "grad_norm": 0.5527541637420654, "learning_rate": 7.511756586630036e-05, "loss": 0.0466, "step": 73170 }, { "grad_norm": 1.0388052463531494, "learning_rate": 7.511041704709403e-05, "loss": 0.0511, "step": 73180 }, { "grad_norm": 0.9111856818199158, "learning_rate": 7.510326754137834e-05, "loss": 0.043, "step": 73190 }, { "grad_norm": 0.44354599714279175, "learning_rate": 7.509611734934873e-05, "loss": 0.0431, "step": 73200 }, { "grad_norm": 0.6649079918861389, "learning_rate": 7.50889664712007e-05, "loss": 0.0401, "step": 73210 }, { "grad_norm": 0.678076446056366, "learning_rate": 7.508181490712975e-05, "loss": 0.0333, "step": 73220 }, { "grad_norm": 0.44169631600379944, "learning_rate": 7.507466265733143e-05, "loss": 0.0403, "step": 73230 }, { "grad_norm": 0.5732449889183044, "learning_rate": 7.506750972200123e-05, "loss": 0.0403, "step": 73240 }, { "grad_norm": 0.7070214748382568, "learning_rate": 7.506035610133474e-05, "loss": 0.0505, "step": 73250 }, { "grad_norm": 0.38622182607650757, "learning_rate": 7.505320179552753e-05, "loss": 0.0383, "step": 73260 }, { "grad_norm": 0.5748118758201599, "learning_rate": 7.504604680477518e-05, "loss": 0.0306, "step": 73270 }, { "grad_norm": 0.7990926504135132, "learning_rate": 7.503889112927334e-05, "loss": 0.0456, "step": 73280 }, { "grad_norm": 0.5183009505271912, "learning_rate": 7.50317347692176e-05, "loss": 0.0434, "step": 73290 }, { "grad_norm": 0.6940284967422485, "learning_rate": 7.502457772480367e-05, "loss": 0.0305, "step": 73300 }, { "grad_norm": 0.5103718638420105, "learning_rate": 7.501741999622716e-05, "loss": 0.0331, "step": 73310 }, { "grad_norm": 0.3625085949897766, "learning_rate": 7.501026158368379e-05, "loss": 0.0327, "step": 73320 }, { "grad_norm": 0.944839596748352, "learning_rate": 7.500310248736926e-05, "loss": 0.0525, "step": 73330 }, { "grad_norm": 0.678626537322998, "learning_rate": 7.499594270747932e-05, "loss": 0.0497, "step": 73340 }, { "grad_norm": 0.5837948322296143, "learning_rate": 7.498878224420967e-05, "loss": 0.0385, "step": 73350 }, { "grad_norm": 1.1294382810592651, "learning_rate": 7.498162109775613e-05, "loss": 0.0529, "step": 73360 }, { "grad_norm": 0.32439759373664856, "learning_rate": 7.497445926831444e-05, "loss": 0.0496, "step": 73370 }, { "grad_norm": 0.5110437870025635, "learning_rate": 7.49672967560804e-05, "loss": 0.0466, "step": 73380 }, { "grad_norm": 0.5876230001449585, "learning_rate": 7.496013356124987e-05, "loss": 0.0327, "step": 73390 }, { "grad_norm": 0.484394907951355, "learning_rate": 7.495296968401865e-05, "loss": 0.0391, "step": 73400 }, { "grad_norm": 0.8068536520004272, "learning_rate": 7.494580512458263e-05, "loss": 0.0319, "step": 73410 }, { "grad_norm": 0.8728707432746887, "learning_rate": 7.493863988313765e-05, "loss": 0.0483, "step": 73420 }, { "grad_norm": 1.5363094806671143, "learning_rate": 7.493147395987964e-05, "loss": 0.0337, "step": 73430 }, { "grad_norm": 0.7831894159317017, "learning_rate": 7.492430735500449e-05, "loss": 0.0463, "step": 73440 }, { "grad_norm": 0.4585077166557312, "learning_rate": 7.491714006870815e-05, "loss": 0.0301, "step": 73450 }, { "grad_norm": 1.3402646780014038, "learning_rate": 7.490997210118655e-05, "loss": 0.0414, "step": 73460 }, { "grad_norm": 0.3144072890281677, "learning_rate": 7.490280345263568e-05, "loss": 0.0352, "step": 73470 }, { "grad_norm": 0.5809911489486694, "learning_rate": 7.489563412325151e-05, "loss": 0.0308, "step": 73480 }, { "grad_norm": 0.5840776562690735, "learning_rate": 7.488846411323007e-05, "loss": 0.0288, "step": 73490 }, { "grad_norm": 0.7748580574989319, "learning_rate": 7.488129342276736e-05, "loss": 0.0315, "step": 73500 }, { "grad_norm": 0.759649395942688, "learning_rate": 7.487412205205943e-05, "loss": 0.0487, "step": 73510 }, { "grad_norm": 0.37240660190582275, "learning_rate": 7.486695000130237e-05, "loss": 0.0334, "step": 73520 }, { "grad_norm": 0.6563425660133362, "learning_rate": 7.485977727069221e-05, "loss": 0.0387, "step": 73530 }, { "grad_norm": 0.756106972694397, "learning_rate": 7.485260386042509e-05, "loss": 0.0512, "step": 73540 }, { "grad_norm": 0.5195733308792114, "learning_rate": 7.484542977069713e-05, "loss": 0.0327, "step": 73550 }, { "grad_norm": 0.6142443418502808, "learning_rate": 7.483825500170445e-05, "loss": 0.0537, "step": 73560 }, { "grad_norm": 0.705574631690979, "learning_rate": 7.48310795536432e-05, "loss": 0.0369, "step": 73570 }, { "grad_norm": 0.4681175947189331, "learning_rate": 7.482390342670958e-05, "loss": 0.049, "step": 73580 }, { "grad_norm": 1.1240198612213135, "learning_rate": 7.481672662109974e-05, "loss": 0.038, "step": 73590 }, { "grad_norm": 0.6236377358436584, "learning_rate": 7.480954913700992e-05, "loss": 0.0337, "step": 73600 }, { "grad_norm": 0.6177738308906555, "learning_rate": 7.480237097463637e-05, "loss": 0.0397, "step": 73610 }, { "grad_norm": 1.0121135711669922, "learning_rate": 7.47951921341753e-05, "loss": 0.0383, "step": 73620 }, { "grad_norm": 0.6751689314842224, "learning_rate": 7.4788012615823e-05, "loss": 0.0357, "step": 73630 }, { "grad_norm": 0.5088222622871399, "learning_rate": 7.478083241977575e-05, "loss": 0.0476, "step": 73640 }, { "grad_norm": 0.7333922982215881, "learning_rate": 7.477365154622983e-05, "loss": 0.044, "step": 73650 }, { "grad_norm": 2.329035758972168, "learning_rate": 7.476646999538161e-05, "loss": 0.0477, "step": 73660 }, { "grad_norm": 0.3154505789279938, "learning_rate": 7.47592877674274e-05, "loss": 0.0297, "step": 73670 }, { "grad_norm": 0.3366260826587677, "learning_rate": 7.475210486256356e-05, "loss": 0.0429, "step": 73680 }, { "grad_norm": 0.5238178372383118, "learning_rate": 7.474492128098646e-05, "loss": 0.0433, "step": 73690 }, { "grad_norm": 0.4578739404678345, "learning_rate": 7.473773702289253e-05, "loss": 0.0395, "step": 73700 }, { "grad_norm": 0.9916865229606628, "learning_rate": 7.473055208847815e-05, "loss": 0.0519, "step": 73710 }, { "grad_norm": 1.2884619235992432, "learning_rate": 7.472336647793976e-05, "loss": 0.0446, "step": 73720 }, { "grad_norm": 0.37557151913642883, "learning_rate": 7.471618019147381e-05, "loss": 0.0353, "step": 73730 }, { "grad_norm": 0.4810997247695923, "learning_rate": 7.470899322927681e-05, "loss": 0.0434, "step": 73740 }, { "grad_norm": 0.4472280442714691, "learning_rate": 7.47018055915452e-05, "loss": 0.0422, "step": 73750 }, { "grad_norm": 0.6014328598976135, "learning_rate": 7.469461727847553e-05, "loss": 0.0553, "step": 73760 }, { "grad_norm": 0.3983776271343231, "learning_rate": 7.468742829026425e-05, "loss": 0.043, "step": 73770 }, { "grad_norm": 0.9012638330459595, "learning_rate": 7.4680238627108e-05, "loss": 0.0312, "step": 73780 }, { "grad_norm": 0.5409608483314514, "learning_rate": 7.467304828920328e-05, "loss": 0.0368, "step": 73790 }, { "grad_norm": 0.5873892307281494, "learning_rate": 7.466585727674669e-05, "loss": 0.0403, "step": 73800 }, { "grad_norm": 0.4693324863910675, "learning_rate": 7.465866558993483e-05, "loss": 0.048, "step": 73810 }, { "grad_norm": 0.6436606049537659, "learning_rate": 7.46514732289643e-05, "loss": 0.037, "step": 73820 }, { "grad_norm": 0.6770743131637573, "learning_rate": 7.46442801940318e-05, "loss": 0.0407, "step": 73830 }, { "grad_norm": 0.562417209148407, "learning_rate": 7.463708648533389e-05, "loss": 0.048, "step": 73840 }, { "grad_norm": 0.4800419807434082, "learning_rate": 7.462989210306731e-05, "loss": 0.0323, "step": 73850 }, { "grad_norm": 0.7901262640953064, "learning_rate": 7.462269704742872e-05, "loss": 0.0505, "step": 73860 }, { "grad_norm": 0.5079373717308044, "learning_rate": 7.461550131861488e-05, "loss": 0.0435, "step": 73870 }, { "grad_norm": 0.6144339442253113, "learning_rate": 7.460830491682245e-05, "loss": 0.0462, "step": 73880 }, { "grad_norm": 0.3105461895465851, "learning_rate": 7.460110784224822e-05, "loss": 0.034, "step": 73890 }, { "grad_norm": 0.42329901456832886, "learning_rate": 7.459391009508894e-05, "loss": 0.0295, "step": 73900 }, { "grad_norm": 0.4945531189441681, "learning_rate": 7.458671167554138e-05, "loss": 0.0391, "step": 73910 }, { "grad_norm": 0.5662288665771484, "learning_rate": 7.457951258380238e-05, "loss": 0.0499, "step": 73920 }, { "grad_norm": 0.3495911955833435, "learning_rate": 7.457231282006875e-05, "loss": 0.0352, "step": 73930 }, { "grad_norm": 0.432142972946167, "learning_rate": 7.456511238453729e-05, "loss": 0.0491, "step": 73940 }, { "grad_norm": 0.6313925981521606, "learning_rate": 7.45579112774049e-05, "loss": 0.0587, "step": 73950 }, { "grad_norm": 0.511920690536499, "learning_rate": 7.455070949886845e-05, "loss": 0.0508, "step": 73960 }, { "grad_norm": 0.55131596326828, "learning_rate": 7.454350704912482e-05, "loss": 0.0376, "step": 73970 }, { "grad_norm": 0.8751404285430908, "learning_rate": 7.453630392837092e-05, "loss": 0.0433, "step": 73980 }, { "grad_norm": 0.5150318145751953, "learning_rate": 7.452910013680369e-05, "loss": 0.0303, "step": 73990 }, { "grad_norm": 1.5257214307785034, "learning_rate": 7.452189567462008e-05, "loss": 0.0577, "step": 74000 }, { "grad_norm": 0.4183588922023773, "learning_rate": 7.451469054201708e-05, "loss": 0.0469, "step": 74010 }, { "grad_norm": 0.5217524170875549, "learning_rate": 7.450748473919161e-05, "loss": 0.0336, "step": 74020 }, { "grad_norm": 0.5133609771728516, "learning_rate": 7.450027826634074e-05, "loss": 0.0341, "step": 74030 }, { "grad_norm": 0.39998576045036316, "learning_rate": 7.449307112366145e-05, "loss": 0.0391, "step": 74040 }, { "grad_norm": 0.3720256984233856, "learning_rate": 7.448586331135081e-05, "loss": 0.0364, "step": 74050 }, { "grad_norm": 0.6582076549530029, "learning_rate": 7.447865482960585e-05, "loss": 0.0352, "step": 74060 }, { "grad_norm": 0.4131133556365967, "learning_rate": 7.447144567862367e-05, "loss": 0.04, "step": 74070 }, { "grad_norm": 0.6571555137634277, "learning_rate": 7.446423585860138e-05, "loss": 0.0416, "step": 74080 }, { "grad_norm": 0.7537857890129089, "learning_rate": 7.445702536973604e-05, "loss": 0.0435, "step": 74090 }, { "grad_norm": 0.6066510081291199, "learning_rate": 7.444981421222483e-05, "loss": 0.0366, "step": 74100 }, { "grad_norm": 0.5471091270446777, "learning_rate": 7.44426023862649e-05, "loss": 0.0347, "step": 74110 }, { "grad_norm": 0.3807761073112488, "learning_rate": 7.443538989205339e-05, "loss": 0.0397, "step": 74120 }, { "grad_norm": 0.6550023555755615, "learning_rate": 7.442817672978749e-05, "loss": 0.0462, "step": 74130 }, { "grad_norm": 0.48106515407562256, "learning_rate": 7.442096289966443e-05, "loss": 0.0333, "step": 74140 }, { "grad_norm": 0.7126415371894836, "learning_rate": 7.441374840188141e-05, "loss": 0.0384, "step": 74150 }, { "grad_norm": 0.6038522720336914, "learning_rate": 7.440653323663568e-05, "loss": 0.0327, "step": 74160 }, { "grad_norm": 0.3578394651412964, "learning_rate": 7.439931740412449e-05, "loss": 0.034, "step": 74170 }, { "grad_norm": 0.60794597864151, "learning_rate": 7.439210090454517e-05, "loss": 0.0409, "step": 74180 }, { "grad_norm": 0.5543286204338074, "learning_rate": 7.438488373809494e-05, "loss": 0.04, "step": 74190 }, { "grad_norm": 0.9529294371604919, "learning_rate": 7.437766590497115e-05, "loss": 0.0426, "step": 74200 }, { "grad_norm": 0.7612006068229675, "learning_rate": 7.437044740537114e-05, "loss": 0.0361, "step": 74210 }, { "grad_norm": 0.44735026359558105, "learning_rate": 7.436322823949225e-05, "loss": 0.0379, "step": 74220 }, { "grad_norm": 0.4214302599430084, "learning_rate": 7.435600840753185e-05, "loss": 0.0341, "step": 74230 }, { "grad_norm": 0.6213170886039734, "learning_rate": 7.434878790968735e-05, "loss": 0.0326, "step": 74240 }, { "grad_norm": 0.42124179005622864, "learning_rate": 7.434156674615611e-05, "loss": 0.0323, "step": 74250 }, { "grad_norm": 0.8245782256126404, "learning_rate": 7.433434491713561e-05, "loss": 0.0515, "step": 74260 }, { "grad_norm": 1.5627827644348145, "learning_rate": 7.432712242282323e-05, "loss": 0.0462, "step": 74270 }, { "grad_norm": 0.4022962152957916, "learning_rate": 7.431989926341648e-05, "loss": 0.0284, "step": 74280 }, { "grad_norm": 0.3315155804157257, "learning_rate": 7.43126754391128e-05, "loss": 0.033, "step": 74290 }, { "grad_norm": 0.33591029047966003, "learning_rate": 7.430545095010973e-05, "loss": 0.0299, "step": 74300 }, { "grad_norm": 1.49473237991333, "learning_rate": 7.429822579660476e-05, "loss": 0.0534, "step": 74310 }, { "grad_norm": 0.550077497959137, "learning_rate": 7.42909999787954e-05, "loss": 0.04, "step": 74320 }, { "grad_norm": 0.37924423813819885, "learning_rate": 7.428377349687927e-05, "loss": 0.0279, "step": 74330 }, { "grad_norm": 0.6236900091171265, "learning_rate": 7.427654635105387e-05, "loss": 0.0532, "step": 74340 }, { "grad_norm": 0.4767071306705475, "learning_rate": 7.42693185415168e-05, "loss": 0.0472, "step": 74350 }, { "grad_norm": 0.8438669443130493, "learning_rate": 7.42620900684657e-05, "loss": 0.0704, "step": 74360 }, { "grad_norm": 1.9409807920455933, "learning_rate": 7.425486093209817e-05, "loss": 0.0421, "step": 74370 }, { "grad_norm": 0.6573659777641296, "learning_rate": 7.424763113261185e-05, "loss": 0.0354, "step": 74380 }, { "grad_norm": 0.547638475894928, "learning_rate": 7.424040067020442e-05, "loss": 0.035, "step": 74390 }, { "grad_norm": 0.5915595889091492, "learning_rate": 7.423316954507352e-05, "loss": 0.0387, "step": 74400 }, { "grad_norm": 0.8736211657524109, "learning_rate": 7.422593775741689e-05, "loss": 0.0367, "step": 74410 }, { "grad_norm": 1.1150221824645996, "learning_rate": 7.421870530743218e-05, "loss": 0.0384, "step": 74420 }, { "grad_norm": 0.3845925033092499, "learning_rate": 7.421147219531721e-05, "loss": 0.0332, "step": 74430 }, { "grad_norm": 0.6859534382820129, "learning_rate": 7.420423842126966e-05, "loss": 0.0356, "step": 74440 }, { "grad_norm": 0.9598626494407654, "learning_rate": 7.419700398548735e-05, "loss": 0.0527, "step": 74450 }, { "grad_norm": 0.5725756883621216, "learning_rate": 7.418976888816802e-05, "loss": 0.0325, "step": 74460 }, { "grad_norm": 0.5746057033538818, "learning_rate": 7.418253312950951e-05, "loss": 0.0369, "step": 74470 }, { "grad_norm": 0.7174174189567566, "learning_rate": 7.41752967097096e-05, "loss": 0.0391, "step": 74480 }, { "grad_norm": 0.4898305833339691, "learning_rate": 7.41680596289662e-05, "loss": 0.0365, "step": 74490 }, { "grad_norm": 0.4227432608604431, "learning_rate": 7.416082188747711e-05, "loss": 0.0401, "step": 74500 }, { "grad_norm": 0.47891196608543396, "learning_rate": 7.415358348544023e-05, "loss": 0.0441, "step": 74510 }, { "grad_norm": 0.4593202769756317, "learning_rate": 7.414634442305345e-05, "loss": 0.0333, "step": 74520 }, { "grad_norm": 0.4410054087638855, "learning_rate": 7.413910470051469e-05, "loss": 0.0481, "step": 74530 }, { "grad_norm": 0.7022548913955688, "learning_rate": 7.413186431802186e-05, "loss": 0.0316, "step": 74540 }, { "grad_norm": 0.4597349166870117, "learning_rate": 7.412462327577294e-05, "loss": 0.0443, "step": 74550 }, { "grad_norm": 1.3973498344421387, "learning_rate": 7.411738157396587e-05, "loss": 0.0478, "step": 74560 }, { "grad_norm": 0.7696905136108398, "learning_rate": 7.411013921279866e-05, "loss": 0.0372, "step": 74570 }, { "grad_norm": 0.7305848598480225, "learning_rate": 7.41028961924693e-05, "loss": 0.0505, "step": 74580 }, { "grad_norm": 1.7857414484024048, "learning_rate": 7.40956525131758e-05, "loss": 0.0411, "step": 74590 }, { "grad_norm": 0.5530229806900024, "learning_rate": 7.408840817511624e-05, "loss": 0.0397, "step": 74600 }, { "grad_norm": 0.9027025103569031, "learning_rate": 7.408116317848862e-05, "loss": 0.0434, "step": 74610 }, { "grad_norm": 0.5746373534202576, "learning_rate": 7.407391752349107e-05, "loss": 0.0388, "step": 74620 }, { "grad_norm": 0.5393884181976318, "learning_rate": 7.406667121032165e-05, "loss": 0.0514, "step": 74630 }, { "grad_norm": 0.6667599081993103, "learning_rate": 7.405942423917849e-05, "loss": 0.0537, "step": 74640 }, { "grad_norm": 0.5175408720970154, "learning_rate": 7.405217661025969e-05, "loss": 0.0457, "step": 74650 }, { "grad_norm": 0.6066992878913879, "learning_rate": 7.404492832376343e-05, "loss": 0.0385, "step": 74660 }, { "grad_norm": 0.5028160810470581, "learning_rate": 7.403767937988785e-05, "loss": 0.044, "step": 74670 }, { "grad_norm": 0.606320321559906, "learning_rate": 7.403042977883117e-05, "loss": 0.0441, "step": 74680 }, { "grad_norm": 0.5430427193641663, "learning_rate": 7.402317952079154e-05, "loss": 0.0357, "step": 74690 }, { "grad_norm": 0.6637331247329712, "learning_rate": 7.401592860596723e-05, "loss": 0.0432, "step": 74700 }, { "grad_norm": 0.5949918627738953, "learning_rate": 7.400867703455644e-05, "loss": 0.0426, "step": 74710 }, { "grad_norm": 0.47529858350753784, "learning_rate": 7.400142480675746e-05, "loss": 0.0341, "step": 74720 }, { "grad_norm": 0.7939667701721191, "learning_rate": 7.399417192276851e-05, "loss": 0.0387, "step": 74730 }, { "grad_norm": 0.6050395965576172, "learning_rate": 7.398691838278794e-05, "loss": 0.0417, "step": 74740 }, { "grad_norm": 0.6626872420310974, "learning_rate": 7.397966418701401e-05, "loss": 0.0385, "step": 74750 }, { "grad_norm": 0.5819980502128601, "learning_rate": 7.39724093356451e-05, "loss": 0.0338, "step": 74760 }, { "grad_norm": 0.7807685732841492, "learning_rate": 7.396515382887951e-05, "loss": 0.0379, "step": 74770 }, { "grad_norm": 0.2899363338947296, "learning_rate": 7.395789766691561e-05, "loss": 0.0358, "step": 74780 }, { "grad_norm": 0.33937907218933105, "learning_rate": 7.395064084995181e-05, "loss": 0.0382, "step": 74790 }, { "grad_norm": 0.42432501912117004, "learning_rate": 7.394338337818646e-05, "loss": 0.0373, "step": 74800 }, { "grad_norm": 0.5746806263923645, "learning_rate": 7.393612525181801e-05, "loss": 0.0348, "step": 74810 }, { "grad_norm": 0.48884910345077515, "learning_rate": 7.392886647104488e-05, "loss": 0.0473, "step": 74820 }, { "grad_norm": 0.6662476658821106, "learning_rate": 7.392160703606556e-05, "loss": 0.0364, "step": 74830 }, { "grad_norm": 0.5639509558677673, "learning_rate": 7.391434694707847e-05, "loss": 0.0356, "step": 74840 }, { "grad_norm": 0.5039001107215881, "learning_rate": 7.390708620428213e-05, "loss": 0.0358, "step": 74850 }, { "grad_norm": 0.6464117765426636, "learning_rate": 7.389982480787501e-05, "loss": 0.0321, "step": 74860 }, { "grad_norm": 0.7290310859680176, "learning_rate": 7.38925627580557e-05, "loss": 0.0698, "step": 74870 }, { "grad_norm": 0.5946241021156311, "learning_rate": 7.388530005502266e-05, "loss": 0.0469, "step": 74880 }, { "grad_norm": 1.0962187051773071, "learning_rate": 7.387803669897452e-05, "loss": 0.0552, "step": 74890 }, { "grad_norm": 0.6707829833030701, "learning_rate": 7.387077269010981e-05, "loss": 0.0695, "step": 74900 }, { "grad_norm": 0.8485026955604553, "learning_rate": 7.386350802862715e-05, "loss": 0.0421, "step": 74910 }, { "grad_norm": 0.36526140570640564, "learning_rate": 7.385624271472513e-05, "loss": 0.0367, "step": 74920 }, { "grad_norm": 0.5754607915878296, "learning_rate": 7.384897674860241e-05, "loss": 0.0607, "step": 74930 }, { "grad_norm": 0.5084073543548584, "learning_rate": 7.384171013045762e-05, "loss": 0.0491, "step": 74940 }, { "grad_norm": 0.9829872250556946, "learning_rate": 7.383444286048944e-05, "loss": 0.0348, "step": 74950 }, { "grad_norm": 0.6675472855567932, "learning_rate": 7.382717493889654e-05, "loss": 0.0394, "step": 74960 }, { "grad_norm": 0.3443349599838257, "learning_rate": 7.381990636587763e-05, "loss": 0.0397, "step": 74970 }, { "grad_norm": 0.5948379039764404, "learning_rate": 7.381263714163144e-05, "loss": 0.0489, "step": 74980 }, { "grad_norm": 0.43995150923728943, "learning_rate": 7.380536726635668e-05, "loss": 0.0505, "step": 74990 }, { "grad_norm": 0.6828826069831848, "learning_rate": 7.379809674025212e-05, "loss": 0.0398, "step": 75000 }, { "grad_norm": 0.8168054223060608, "learning_rate": 7.379082556351655e-05, "loss": 0.0365, "step": 75010 }, { "grad_norm": 0.5815768837928772, "learning_rate": 7.378355373634873e-05, "loss": 0.0466, "step": 75020 }, { "grad_norm": 0.8066537976264954, "learning_rate": 7.37762812589475e-05, "loss": 0.0533, "step": 75030 }, { "grad_norm": 0.5766963362693787, "learning_rate": 7.376900813151168e-05, "loss": 0.036, "step": 75040 }, { "grad_norm": 0.25332576036453247, "learning_rate": 7.376173435424008e-05, "loss": 0.0287, "step": 75050 }, { "grad_norm": 0.7042981386184692, "learning_rate": 7.375445992733161e-05, "loss": 0.0398, "step": 75060 }, { "grad_norm": 0.7342949509620667, "learning_rate": 7.374718485098513e-05, "loss": 0.0413, "step": 75070 }, { "grad_norm": 0.5636876225471497, "learning_rate": 7.373990912539953e-05, "loss": 0.0322, "step": 75080 }, { "grad_norm": 0.4029461741447449, "learning_rate": 7.373263275077374e-05, "loss": 0.0374, "step": 75090 }, { "grad_norm": 0.36363354325294495, "learning_rate": 7.372535572730669e-05, "loss": 0.0482, "step": 75100 }, { "grad_norm": 0.5234417915344238, "learning_rate": 7.371807805519732e-05, "loss": 0.0404, "step": 75110 }, { "grad_norm": 1.2425618171691895, "learning_rate": 7.37107997346446e-05, "loss": 0.0554, "step": 75120 }, { "grad_norm": 0.6134839057922363, "learning_rate": 7.370352076584753e-05, "loss": 0.0365, "step": 75130 }, { "grad_norm": 0.7788882851600647, "learning_rate": 7.369624114900511e-05, "loss": 0.0375, "step": 75140 }, { "grad_norm": 0.45371195673942566, "learning_rate": 7.368896088431635e-05, "loss": 0.0518, "step": 75150 }, { "grad_norm": 0.5850005745887756, "learning_rate": 7.368167997198031e-05, "loss": 0.0373, "step": 75160 }, { "grad_norm": 0.5700773000717163, "learning_rate": 7.367439841219603e-05, "loss": 0.0354, "step": 75170 }, { "grad_norm": 0.44604024291038513, "learning_rate": 7.366711620516258e-05, "loss": 0.0452, "step": 75180 }, { "grad_norm": 0.40365126729011536, "learning_rate": 7.365983335107908e-05, "loss": 0.0364, "step": 75190 }, { "grad_norm": 0.743218183517456, "learning_rate": 7.365254985014464e-05, "loss": 0.0396, "step": 75200 }, { "grad_norm": 0.5702033638954163, "learning_rate": 7.364526570255835e-05, "loss": 0.0308, "step": 75210 }, { "grad_norm": 1.4033626317977905, "learning_rate": 7.36379809085194e-05, "loss": 0.0458, "step": 75220 }, { "grad_norm": 0.5415608286857605, "learning_rate": 7.36306954682269e-05, "loss": 0.0479, "step": 75230 }, { "grad_norm": 0.4955465495586395, "learning_rate": 7.362340938188008e-05, "loss": 0.028, "step": 75240 }, { "grad_norm": 0.3284366726875305, "learning_rate": 7.361612264967811e-05, "loss": 0.0383, "step": 75250 }, { "grad_norm": 0.4742860794067383, "learning_rate": 7.360883527182025e-05, "loss": 0.0416, "step": 75260 }, { "grad_norm": 0.48686450719833374, "learning_rate": 7.360154724850568e-05, "loss": 0.0375, "step": 75270 }, { "grad_norm": 0.5142777562141418, "learning_rate": 7.359425857993367e-05, "loss": 0.0403, "step": 75280 }, { "grad_norm": 0.8224822282791138, "learning_rate": 7.358696926630351e-05, "loss": 0.0427, "step": 75290 }, { "grad_norm": 0.5759555101394653, "learning_rate": 7.357967930781446e-05, "loss": 0.0318, "step": 75300 }, { "grad_norm": 0.55811607837677, "learning_rate": 7.357238870466585e-05, "loss": 0.0562, "step": 75310 }, { "grad_norm": 0.6408155560493469, "learning_rate": 7.356509745705697e-05, "loss": 0.0394, "step": 75320 }, { "grad_norm": 0.32460060715675354, "learning_rate": 7.355780556518719e-05, "loss": 0.0368, "step": 75330 }, { "grad_norm": 0.7179728150367737, "learning_rate": 7.355051302925585e-05, "loss": 0.0307, "step": 75340 }, { "grad_norm": 0.4821002781391144, "learning_rate": 7.354321984946234e-05, "loss": 0.0467, "step": 75350 }, { "grad_norm": 0.5406652688980103, "learning_rate": 7.353592602600605e-05, "loss": 0.0378, "step": 75360 }, { "grad_norm": 0.608720064163208, "learning_rate": 7.352863155908636e-05, "loss": 0.0361, "step": 75370 }, { "grad_norm": 0.7616427540779114, "learning_rate": 7.352133644890273e-05, "loss": 0.0313, "step": 75380 }, { "grad_norm": 0.5601134300231934, "learning_rate": 7.351404069565462e-05, "loss": 0.0315, "step": 75390 }, { "grad_norm": 0.5490321516990662, "learning_rate": 7.350674429954146e-05, "loss": 0.0367, "step": 75400 }, { "grad_norm": 0.27895691990852356, "learning_rate": 7.349944726076274e-05, "loss": 0.0363, "step": 75410 }, { "grad_norm": 0.6013776659965515, "learning_rate": 7.349214957951795e-05, "loss": 0.0323, "step": 75420 }, { "grad_norm": 0.4391825795173645, "learning_rate": 7.348485125600663e-05, "loss": 0.0543, "step": 75430 }, { "grad_norm": 0.817649781703949, "learning_rate": 7.347755229042828e-05, "loss": 0.0486, "step": 75440 }, { "grad_norm": 0.6749547719955444, "learning_rate": 7.347025268298249e-05, "loss": 0.0358, "step": 75450 }, { "grad_norm": 0.7747499942779541, "learning_rate": 7.34629524338688e-05, "loss": 0.0379, "step": 75460 }, { "grad_norm": 1.6175421476364136, "learning_rate": 7.345565154328682e-05, "loss": 0.0481, "step": 75470 }, { "grad_norm": 1.155390739440918, "learning_rate": 7.344835001143611e-05, "loss": 0.0332, "step": 75480 }, { "grad_norm": 0.8896484971046448, "learning_rate": 7.344104783851634e-05, "loss": 0.0434, "step": 75490 }, { "grad_norm": 0.7402262687683105, "learning_rate": 7.343374502472713e-05, "loss": 0.0442, "step": 75500 }, { "grad_norm": 2.917646646499634, "learning_rate": 7.342644157026813e-05, "loss": 0.0773, "step": 75510 }, { "grad_norm": 0.6244267821311951, "learning_rate": 7.341913747533901e-05, "loss": 0.0505, "step": 75520 }, { "grad_norm": 0.6368706226348877, "learning_rate": 7.341183274013948e-05, "loss": 0.0528, "step": 75530 }, { "grad_norm": 4.736050128936768, "learning_rate": 7.340452736486925e-05, "loss": 0.0553, "step": 75540 }, { "grad_norm": 0.7833214998245239, "learning_rate": 7.3397221349728e-05, "loss": 0.044, "step": 75550 }, { "grad_norm": 0.7642663717269897, "learning_rate": 7.338991469491554e-05, "loss": 0.0436, "step": 75560 }, { "grad_norm": 0.4839797616004944, "learning_rate": 7.338260740063159e-05, "loss": 0.036, "step": 75570 }, { "grad_norm": 0.6585306525230408, "learning_rate": 7.337529946707593e-05, "loss": 0.0405, "step": 75580 }, { "grad_norm": 0.4958226680755615, "learning_rate": 7.336799089444837e-05, "loss": 0.0435, "step": 75590 }, { "grad_norm": 0.3372119665145874, "learning_rate": 7.336068168294873e-05, "loss": 0.0309, "step": 75600 }, { "grad_norm": 0.9372813105583191, "learning_rate": 7.335337183277682e-05, "loss": 0.0454, "step": 75610 }, { "grad_norm": 0.7349286079406738, "learning_rate": 7.334606134413249e-05, "loss": 0.0316, "step": 75620 }, { "grad_norm": 0.4968860149383545, "learning_rate": 7.333875021721562e-05, "loss": 0.0383, "step": 75630 }, { "grad_norm": 0.42045843601226807, "learning_rate": 7.333143845222607e-05, "loss": 0.0399, "step": 75640 }, { "grad_norm": 0.7776540517807007, "learning_rate": 7.332412604936378e-05, "loss": 0.0452, "step": 75650 }, { "grad_norm": 1.8374912738800049, "learning_rate": 7.331681300882865e-05, "loss": 0.043, "step": 75660 }, { "grad_norm": 0.5910296440124512, "learning_rate": 7.33094993308206e-05, "loss": 0.0362, "step": 75670 }, { "grad_norm": 0.5452921986579895, "learning_rate": 7.33021850155396e-05, "loss": 0.0491, "step": 75680 }, { "grad_norm": 0.5736867785453796, "learning_rate": 7.329487006318561e-05, "loss": 0.0531, "step": 75690 }, { "grad_norm": 0.531924307346344, "learning_rate": 7.328755447395863e-05, "loss": 0.0367, "step": 75700 }, { "grad_norm": 1.1317262649536133, "learning_rate": 7.328023824805865e-05, "loss": 0.049, "step": 75710 }, { "grad_norm": 4.233193397521973, "learning_rate": 7.327292138568571e-05, "loss": 0.0554, "step": 75720 }, { "grad_norm": 0.5862090587615967, "learning_rate": 7.326560388703985e-05, "loss": 0.0488, "step": 75730 }, { "grad_norm": 1.0090621709823608, "learning_rate": 7.325828575232109e-05, "loss": 0.0436, "step": 75740 }, { "grad_norm": 2.6176087856292725, "learning_rate": 7.325096698172956e-05, "loss": 0.0536, "step": 75750 }, { "grad_norm": 0.32396358251571655, "learning_rate": 7.324364757546534e-05, "loss": 0.0387, "step": 75760 }, { "grad_norm": 0.5216636657714844, "learning_rate": 7.32363275337285e-05, "loss": 0.0292, "step": 75770 }, { "grad_norm": 0.44178083539009094, "learning_rate": 7.322900685671921e-05, "loss": 0.0407, "step": 75780 }, { "grad_norm": 0.5120337605476379, "learning_rate": 7.32216855446376e-05, "loss": 0.0338, "step": 75790 }, { "grad_norm": 0.44080185890197754, "learning_rate": 7.321436359768382e-05, "loss": 0.0337, "step": 75800 }, { "grad_norm": 0.36222270131111145, "learning_rate": 7.320704101605808e-05, "loss": 0.0421, "step": 75810 }, { "grad_norm": 1.4113672971725464, "learning_rate": 7.319971779996053e-05, "loss": 0.0406, "step": 75820 }, { "grad_norm": 0.7296552658081055, "learning_rate": 7.319239394959143e-05, "loss": 0.0421, "step": 75830 }, { "grad_norm": 0.4949243664741516, "learning_rate": 7.3185069465151e-05, "loss": 0.0377, "step": 75840 }, { "grad_norm": 0.7936655282974243, "learning_rate": 7.317774434683946e-05, "loss": 0.0398, "step": 75850 }, { "grad_norm": 0.7002934217453003, "learning_rate": 7.31704185948571e-05, "loss": 0.0447, "step": 75860 }, { "grad_norm": 0.6528456211090088, "learning_rate": 7.316309220940422e-05, "loss": 0.0409, "step": 75870 }, { "grad_norm": 0.7797419428825378, "learning_rate": 7.315576519068108e-05, "loss": 0.0436, "step": 75880 }, { "grad_norm": 0.6830799579620361, "learning_rate": 7.314843753888803e-05, "loss": 0.0533, "step": 75890 }, { "grad_norm": 0.4424498677253723, "learning_rate": 7.314110925422539e-05, "loss": 0.0403, "step": 75900 }, { "grad_norm": 0.5427954792976379, "learning_rate": 7.313378033689352e-05, "loss": 0.0441, "step": 75910 }, { "grad_norm": 0.5107039213180542, "learning_rate": 7.312645078709278e-05, "loss": 0.0513, "step": 75920 }, { "grad_norm": 0.8275754451751709, "learning_rate": 7.311912060502357e-05, "loss": 0.0402, "step": 75930 }, { "grad_norm": 0.38343876600265503, "learning_rate": 7.311178979088628e-05, "loss": 0.034, "step": 75940 }, { "grad_norm": 0.4338926374912262, "learning_rate": 7.310445834488136e-05, "loss": 0.0255, "step": 75950 }, { "grad_norm": 0.5128186345100403, "learning_rate": 7.30971262672092e-05, "loss": 0.028, "step": 75960 }, { "grad_norm": 0.8226336240768433, "learning_rate": 7.308979355807031e-05, "loss": 0.0341, "step": 75970 }, { "grad_norm": 1.0507320165634155, "learning_rate": 7.308246021766511e-05, "loss": 0.0339, "step": 75980 }, { "grad_norm": 0.6713811159133911, "learning_rate": 7.307512624619415e-05, "loss": 0.0639, "step": 75990 }, { "grad_norm": 0.6069821119308472, "learning_rate": 7.30677916438579e-05, "loss": 0.0443, "step": 76000 }, { "grad_norm": 0.5299432873725891, "learning_rate": 7.306045641085688e-05, "loss": 0.0379, "step": 76010 }, { "grad_norm": 0.26312747597694397, "learning_rate": 7.305312054739166e-05, "loss": 0.057, "step": 76020 }, { "grad_norm": 1.009695291519165, "learning_rate": 7.304578405366276e-05, "loss": 0.0288, "step": 76030 }, { "grad_norm": 0.5998988747596741, "learning_rate": 7.303844692987081e-05, "loss": 0.0404, "step": 76040 }, { "grad_norm": 0.420695960521698, "learning_rate": 7.303110917621637e-05, "loss": 0.0493, "step": 76050 }, { "grad_norm": 0.288958340883255, "learning_rate": 7.302377079290004e-05, "loss": 0.0348, "step": 76060 }, { "grad_norm": 0.5199857950210571, "learning_rate": 7.301643178012247e-05, "loss": 0.0281, "step": 76070 }, { "grad_norm": 0.41184523701667786, "learning_rate": 7.300909213808432e-05, "loss": 0.0436, "step": 76080 }, { "grad_norm": 0.5779412984848022, "learning_rate": 7.300175186698622e-05, "loss": 0.0599, "step": 76090 }, { "grad_norm": 0.5711252093315125, "learning_rate": 7.299441096702888e-05, "loss": 0.0434, "step": 76100 }, { "grad_norm": 0.646570086479187, "learning_rate": 7.298706943841298e-05, "loss": 0.0526, "step": 76110 }, { "grad_norm": 0.5000240802764893, "learning_rate": 7.297972728133924e-05, "loss": 0.0448, "step": 76120 }, { "grad_norm": 0.542554497718811, "learning_rate": 7.297238449600837e-05, "loss": 0.0307, "step": 76130 }, { "grad_norm": 0.42096173763275146, "learning_rate": 7.296504108262114e-05, "loss": 0.0327, "step": 76140 }, { "grad_norm": 0.5447748899459839, "learning_rate": 7.295769704137832e-05, "loss": 0.0479, "step": 76150 }, { "grad_norm": 0.4602576792240143, "learning_rate": 7.29503523724807e-05, "loss": 0.0464, "step": 76160 }, { "grad_norm": 0.3889011740684509, "learning_rate": 7.294300707612905e-05, "loss": 0.0351, "step": 76170 }, { "grad_norm": 0.5741944909095764, "learning_rate": 7.293566115252423e-05, "loss": 0.0362, "step": 76180 }, { "grad_norm": 0.7386742234230042, "learning_rate": 7.292831460186704e-05, "loss": 0.0453, "step": 76190 }, { "grad_norm": 0.488019198179245, "learning_rate": 7.292096742435833e-05, "loss": 0.0411, "step": 76200 }, { "grad_norm": 0.4175091087818146, "learning_rate": 7.291361962019898e-05, "loss": 0.0435, "step": 76210 }, { "grad_norm": 1.0306254625320435, "learning_rate": 7.29062711895899e-05, "loss": 0.0425, "step": 76220 }, { "grad_norm": 0.7322680354118347, "learning_rate": 7.289892213273195e-05, "loss": 0.0422, "step": 76230 }, { "grad_norm": 0.69195556640625, "learning_rate": 7.289157244982607e-05, "loss": 0.0363, "step": 76240 }, { "grad_norm": 0.5921034812927246, "learning_rate": 7.288422214107322e-05, "loss": 0.03, "step": 76250 }, { "grad_norm": 0.38237687945365906, "learning_rate": 7.287687120667432e-05, "loss": 0.0605, "step": 76260 }, { "grad_norm": 0.48746758699417114, "learning_rate": 7.286951964683036e-05, "loss": 0.0477, "step": 76270 }, { "grad_norm": 1.1147494316101074, "learning_rate": 7.286216746174233e-05, "loss": 0.0582, "step": 76280 }, { "grad_norm": 0.28865286707878113, "learning_rate": 7.285481465161123e-05, "loss": 0.0288, "step": 76290 }, { "grad_norm": 0.6088605523109436, "learning_rate": 7.284746121663807e-05, "loss": 0.0324, "step": 76300 }, { "grad_norm": 0.2536008059978485, "learning_rate": 7.284010715702392e-05, "loss": 0.04, "step": 76310 }, { "grad_norm": 0.4718429744243622, "learning_rate": 7.28327524729698e-05, "loss": 0.0373, "step": 76320 }, { "grad_norm": 0.3398876488208771, "learning_rate": 7.282539716467682e-05, "loss": 0.0363, "step": 76330 }, { "grad_norm": 0.9450147151947021, "learning_rate": 7.281804123234605e-05, "loss": 0.0434, "step": 76340 }, { "grad_norm": 0.8299919962882996, "learning_rate": 7.281068467617862e-05, "loss": 0.0518, "step": 76350 }, { "grad_norm": 0.5603523254394531, "learning_rate": 7.280332749637561e-05, "loss": 0.0444, "step": 76360 }, { "grad_norm": 0.6244940757751465, "learning_rate": 7.279596969313823e-05, "loss": 0.0502, "step": 76370 }, { "grad_norm": 0.6228275299072266, "learning_rate": 7.278861126666758e-05, "loss": 0.0395, "step": 76380 }, { "grad_norm": 0.6526498794555664, "learning_rate": 7.278125221716486e-05, "loss": 0.0368, "step": 76390 }, { "grad_norm": 0.4900742471218109, "learning_rate": 7.277389254483127e-05, "loss": 0.0575, "step": 76400 }, { "grad_norm": 0.875350832939148, "learning_rate": 7.276653224986801e-05, "loss": 0.0444, "step": 76410 }, { "grad_norm": 0.6543457508087158, "learning_rate": 7.275917133247632e-05, "loss": 0.0465, "step": 76420 }, { "grad_norm": 0.580007791519165, "learning_rate": 7.275180979285743e-05, "loss": 0.0356, "step": 76430 }, { "grad_norm": 0.4904078543186188, "learning_rate": 7.27444476312126e-05, "loss": 0.0379, "step": 76440 }, { "grad_norm": 0.6877630949020386, "learning_rate": 7.273708484774312e-05, "loss": 0.0313, "step": 76450 }, { "grad_norm": 0.5320123434066772, "learning_rate": 7.272972144265029e-05, "loss": 0.0664, "step": 76460 }, { "grad_norm": 0.6701797842979431, "learning_rate": 7.272235741613541e-05, "loss": 0.0401, "step": 76470 }, { "grad_norm": 0.6280249357223511, "learning_rate": 7.271499276839982e-05, "loss": 0.045, "step": 76480 }, { "grad_norm": 0.36141613125801086, "learning_rate": 7.270762749964485e-05, "loss": 0.0406, "step": 76490 }, { "grad_norm": 0.5675387382507324, "learning_rate": 7.270026161007188e-05, "loss": 0.037, "step": 76500 }, { "grad_norm": 0.2958509922027588, "learning_rate": 7.269289509988229e-05, "loss": 0.0348, "step": 76510 }, { "grad_norm": 0.5999882221221924, "learning_rate": 7.268552796927745e-05, "loss": 0.0433, "step": 76520 }, { "grad_norm": 0.6610996723175049, "learning_rate": 7.267816021845882e-05, "loss": 0.058, "step": 76530 }, { "grad_norm": 0.5885528922080994, "learning_rate": 7.267079184762781e-05, "loss": 0.0353, "step": 76540 }, { "grad_norm": 1.1570566892623901, "learning_rate": 7.266342285698587e-05, "loss": 0.0467, "step": 76550 }, { "grad_norm": 0.6117709279060364, "learning_rate": 7.265605324673447e-05, "loss": 0.0433, "step": 76560 }, { "grad_norm": 0.4825775921344757, "learning_rate": 7.264868301707507e-05, "loss": 0.0332, "step": 76570 }, { "grad_norm": 0.34463202953338623, "learning_rate": 7.264131216820918e-05, "loss": 0.0327, "step": 76580 }, { "grad_norm": 0.8928911685943604, "learning_rate": 7.263394070033834e-05, "loss": 0.0515, "step": 76590 }, { "grad_norm": 0.8215903639793396, "learning_rate": 7.262656861366405e-05, "loss": 0.0317, "step": 76600 }, { "grad_norm": 0.3617337644100189, "learning_rate": 7.261919590838788e-05, "loss": 0.0263, "step": 76610 }, { "grad_norm": 0.3179371953010559, "learning_rate": 7.261182258471139e-05, "loss": 0.0486, "step": 76620 }, { "grad_norm": 0.4461956024169922, "learning_rate": 7.260444864283616e-05, "loss": 0.0506, "step": 76630 }, { "grad_norm": 0.8159703016281128, "learning_rate": 7.259707408296381e-05, "loss": 0.0379, "step": 76640 }, { "grad_norm": 0.5007601976394653, "learning_rate": 7.258969890529593e-05, "loss": 0.0316, "step": 76650 }, { "grad_norm": 0.693423867225647, "learning_rate": 7.258232311003418e-05, "loss": 0.0398, "step": 76660 }, { "grad_norm": 0.3800026476383209, "learning_rate": 7.257494669738018e-05, "loss": 0.045, "step": 76670 }, { "grad_norm": 0.6841953992843628, "learning_rate": 7.256756966753565e-05, "loss": 0.0372, "step": 76680 }, { "grad_norm": 0.4263128638267517, "learning_rate": 7.256019202070222e-05, "loss": 0.0368, "step": 76690 }, { "grad_norm": 0.6716954112052917, "learning_rate": 7.255281375708162e-05, "loss": 0.0283, "step": 76700 }, { "grad_norm": 0.7762892246246338, "learning_rate": 7.254543487687558e-05, "loss": 0.029, "step": 76710 }, { "grad_norm": 0.5392054915428162, "learning_rate": 7.25380553802858e-05, "loss": 0.0299, "step": 76720 }, { "grad_norm": 0.6618196368217468, "learning_rate": 7.253067526751406e-05, "loss": 0.041, "step": 76730 }, { "grad_norm": 0.3742814064025879, "learning_rate": 7.252329453876212e-05, "loss": 0.0307, "step": 76740 }, { "grad_norm": 0.30481037497520447, "learning_rate": 7.251591319423177e-05, "loss": 0.0472, "step": 76750 }, { "grad_norm": 0.41681408882141113, "learning_rate": 7.250853123412482e-05, "loss": 0.0437, "step": 76760 }, { "grad_norm": 0.47940757870674133, "learning_rate": 7.250114865864307e-05, "loss": 0.0373, "step": 76770 }, { "grad_norm": 0.7437270283699036, "learning_rate": 7.249376546798838e-05, "loss": 0.043, "step": 76780 }, { "grad_norm": 0.5812098383903503, "learning_rate": 7.248638166236259e-05, "loss": 0.0537, "step": 76790 }, { "grad_norm": 0.6457610130310059, "learning_rate": 7.247899724196756e-05, "loss": 0.0392, "step": 76800 }, { "grad_norm": 0.5525604486465454, "learning_rate": 7.247161220700522e-05, "loss": 0.0407, "step": 76810 }, { "grad_norm": 0.6202613711357117, "learning_rate": 7.246422655767742e-05, "loss": 0.0421, "step": 76820 }, { "grad_norm": 0.5213881731033325, "learning_rate": 7.24568402941861e-05, "loss": 0.0545, "step": 76830 }, { "grad_norm": 0.4674619138240814, "learning_rate": 7.244945341673322e-05, "loss": 0.0601, "step": 76840 }, { "grad_norm": 0.32466521859169006, "learning_rate": 7.244206592552071e-05, "loss": 0.0326, "step": 76850 }, { "grad_norm": 0.530667781829834, "learning_rate": 7.243467782075054e-05, "loss": 0.0402, "step": 76860 }, { "grad_norm": 0.7360541224479675, "learning_rate": 7.242728910262472e-05, "loss": 0.029, "step": 76870 }, { "grad_norm": 1.2197933197021484, "learning_rate": 7.241989977134524e-05, "loss": 0.0583, "step": 76880 }, { "grad_norm": 1.3049581050872803, "learning_rate": 7.241250982711413e-05, "loss": 0.0416, "step": 76890 }, { "grad_norm": 0.656181812286377, "learning_rate": 7.24051192701334e-05, "loss": 0.0498, "step": 76900 }, { "grad_norm": 0.7329277396202087, "learning_rate": 7.239772810060515e-05, "loss": 0.0373, "step": 76910 }, { "grad_norm": 2.8733181953430176, "learning_rate": 7.23903363187314e-05, "loss": 0.0317, "step": 76920 }, { "grad_norm": 0.41077569127082825, "learning_rate": 7.23829439247143e-05, "loss": 0.0267, "step": 76930 }, { "grad_norm": 0.7303817272186279, "learning_rate": 7.23755509187559e-05, "loss": 0.0395, "step": 76940 }, { "grad_norm": 0.42949730157852173, "learning_rate": 7.236815730105835e-05, "loss": 0.0424, "step": 76950 }, { "grad_norm": 0.6542243361473083, "learning_rate": 7.236076307182378e-05, "loss": 0.0469, "step": 76960 }, { "grad_norm": 0.5390422940254211, "learning_rate": 7.235336823125434e-05, "loss": 0.071, "step": 76970 }, { "grad_norm": 0.472996324300766, "learning_rate": 7.234597277955222e-05, "loss": 0.0376, "step": 76980 }, { "grad_norm": 0.5703572034835815, "learning_rate": 7.23385767169196e-05, "loss": 0.0356, "step": 76990 }, { "grad_norm": 0.5808945298194885, "learning_rate": 7.23311800435587e-05, "loss": 0.0374, "step": 77000 }, { "grad_norm": 0.4887534976005554, "learning_rate": 7.232378275967172e-05, "loss": 0.0284, "step": 77010 }, { "grad_norm": 0.625855565071106, "learning_rate": 7.231638486546089e-05, "loss": 0.0372, "step": 77020 }, { "grad_norm": 0.5491633415222168, "learning_rate": 7.230898636112848e-05, "loss": 0.0413, "step": 77030 }, { "grad_norm": 0.5449973940849304, "learning_rate": 7.230158724687678e-05, "loss": 0.0353, "step": 77040 }, { "grad_norm": 0.7311238050460815, "learning_rate": 7.229418752290805e-05, "loss": 0.0373, "step": 77050 }, { "grad_norm": 0.49717286229133606, "learning_rate": 7.228678718942463e-05, "loss": 0.0326, "step": 77060 }, { "grad_norm": 0.3488488495349884, "learning_rate": 7.22793862466288e-05, "loss": 0.0356, "step": 77070 }, { "grad_norm": 0.39864203333854675, "learning_rate": 7.227198469472292e-05, "loss": 0.0336, "step": 77080 }, { "grad_norm": 0.484374076128006, "learning_rate": 7.226458253390935e-05, "loss": 0.0331, "step": 77090 }, { "grad_norm": 0.6026151180267334, "learning_rate": 7.225717976439046e-05, "loss": 0.0356, "step": 77100 }, { "grad_norm": 0.24383409321308136, "learning_rate": 7.224977638636863e-05, "loss": 0.0306, "step": 77110 }, { "grad_norm": 0.45797285437583923, "learning_rate": 7.224237240004628e-05, "loss": 0.0336, "step": 77120 }, { "grad_norm": 0.7733824849128723, "learning_rate": 7.223496780562583e-05, "loss": 0.0374, "step": 77130 }, { "grad_norm": 0.46092700958251953, "learning_rate": 7.222756260330972e-05, "loss": 0.0479, "step": 77140 }, { "grad_norm": 1.9204397201538086, "learning_rate": 7.222015679330037e-05, "loss": 0.0483, "step": 77150 }, { "grad_norm": 0.542095959186554, "learning_rate": 7.221275037580029e-05, "loss": 0.0446, "step": 77160 }, { "grad_norm": 0.7934530377388, "learning_rate": 7.220534335101197e-05, "loss": 0.0357, "step": 77170 }, { "grad_norm": 0.5967861413955688, "learning_rate": 7.219793571913792e-05, "loss": 0.0505, "step": 77180 }, { "grad_norm": 0.3113898038864136, "learning_rate": 7.219052748038062e-05, "loss": 0.0292, "step": 77190 }, { "grad_norm": 0.4975797235965729, "learning_rate": 7.218311863494263e-05, "loss": 0.0644, "step": 77200 }, { "grad_norm": 0.6588749885559082, "learning_rate": 7.217570918302655e-05, "loss": 0.0353, "step": 77210 }, { "grad_norm": 0.5309670567512512, "learning_rate": 7.216829912483486e-05, "loss": 0.0427, "step": 77220 }, { "grad_norm": 0.4634060859680176, "learning_rate": 7.216088846057022e-05, "loss": 0.0252, "step": 77230 }, { "grad_norm": 0.9336604475975037, "learning_rate": 7.215347719043524e-05, "loss": 0.0338, "step": 77240 }, { "grad_norm": 0.3043406903743744, "learning_rate": 7.21460653146325e-05, "loss": 0.0524, "step": 77250 }, { "grad_norm": 0.7762683629989624, "learning_rate": 7.213865283336464e-05, "loss": 0.0398, "step": 77260 }, { "grad_norm": 0.719767153263092, "learning_rate": 7.213123974683436e-05, "loss": 0.0331, "step": 77270 }, { "grad_norm": 1.0403141975402832, "learning_rate": 7.212382605524425e-05, "loss": 0.0457, "step": 77280 }, { "grad_norm": 0.48150211572647095, "learning_rate": 7.211641175879709e-05, "loss": 0.0417, "step": 77290 }, { "grad_norm": 0.6679140329360962, "learning_rate": 7.210899685769552e-05, "loss": 0.0333, "step": 77300 }, { "grad_norm": 0.41636791825294495, "learning_rate": 7.21015813521423e-05, "loss": 0.034, "step": 77310 }, { "grad_norm": 0.415401816368103, "learning_rate": 7.209416524234013e-05, "loss": 0.0291, "step": 77320 }, { "grad_norm": 0.4643856883049011, "learning_rate": 7.208674852849181e-05, "loss": 0.0436, "step": 77330 }, { "grad_norm": 0.6477247476577759, "learning_rate": 7.207933121080006e-05, "loss": 0.0287, "step": 77340 }, { "grad_norm": 0.34397387504577637, "learning_rate": 7.20719132894677e-05, "loss": 0.0326, "step": 77350 }, { "grad_norm": 0.46374452114105225, "learning_rate": 7.20644947646975e-05, "loss": 0.0405, "step": 77360 }, { "grad_norm": 0.5883309841156006, "learning_rate": 7.205707563669232e-05, "loss": 0.0362, "step": 77370 }, { "grad_norm": 1.1576271057128906, "learning_rate": 7.204965590565497e-05, "loss": 0.0336, "step": 77380 }, { "grad_norm": 0.4558315873146057, "learning_rate": 7.204223557178834e-05, "loss": 0.0399, "step": 77390 }, { "grad_norm": 0.6099178194999695, "learning_rate": 7.203481463529525e-05, "loss": 0.0386, "step": 77400 }, { "grad_norm": 0.5142402648925781, "learning_rate": 7.20273930963786e-05, "loss": 0.054, "step": 77410 }, { "grad_norm": 0.9479755163192749, "learning_rate": 7.20199709552413e-05, "loss": 0.041, "step": 77420 }, { "grad_norm": 0.5834778547286987, "learning_rate": 7.201254821208627e-05, "loss": 0.0329, "step": 77430 }, { "grad_norm": 0.6413375735282898, "learning_rate": 7.200512486711644e-05, "loss": 0.0429, "step": 77440 }, { "grad_norm": 0.27099937200546265, "learning_rate": 7.199770092053476e-05, "loss": 0.0314, "step": 77450 }, { "grad_norm": 0.7402042746543884, "learning_rate": 7.199027637254421e-05, "loss": 0.0375, "step": 77460 }, { "grad_norm": 0.403001070022583, "learning_rate": 7.198285122334775e-05, "loss": 0.0368, "step": 77470 }, { "grad_norm": 0.8253275752067566, "learning_rate": 7.197542547314841e-05, "loss": 0.0483, "step": 77480 }, { "grad_norm": 0.7873509526252747, "learning_rate": 7.19679991221492e-05, "loss": 0.0368, "step": 77490 }, { "grad_norm": 0.5378254055976868, "learning_rate": 7.196057217055315e-05, "loss": 0.0355, "step": 77500 }, { "grad_norm": 0.4757232069969177, "learning_rate": 7.195314461856328e-05, "loss": 0.0352, "step": 77510 }, { "grad_norm": 0.9636372923851013, "learning_rate": 7.194571646638271e-05, "loss": 0.0481, "step": 77520 }, { "grad_norm": 0.5659991502761841, "learning_rate": 7.193828771421447e-05, "loss": 0.043, "step": 77530 }, { "grad_norm": 0.504941999912262, "learning_rate": 7.19308583622617e-05, "loss": 0.0386, "step": 77540 }, { "grad_norm": 0.4474281072616577, "learning_rate": 7.19234284107275e-05, "loss": 0.0281, "step": 77550 }, { "grad_norm": 0.3273301124572754, "learning_rate": 7.191599785981502e-05, "loss": 0.0365, "step": 77560 }, { "grad_norm": 0.39751747250556946, "learning_rate": 7.190856670972737e-05, "loss": 0.0299, "step": 77570 }, { "grad_norm": 0.7564520835876465, "learning_rate": 7.190113496066775e-05, "loss": 0.0379, "step": 77580 }, { "grad_norm": 0.6194102168083191, "learning_rate": 7.189370261283933e-05, "loss": 0.0336, "step": 77590 }, { "grad_norm": 0.36687660217285156, "learning_rate": 7.188626966644528e-05, "loss": 0.0289, "step": 77600 }, { "grad_norm": 0.7187610864639282, "learning_rate": 7.187883612168886e-05, "loss": 0.0422, "step": 77610 }, { "grad_norm": 0.4278734624385834, "learning_rate": 7.187140197877328e-05, "loss": 0.0484, "step": 77620 }, { "grad_norm": 0.4603163003921509, "learning_rate": 7.186396723790179e-05, "loss": 0.0411, "step": 77630 }, { "grad_norm": 0.5760947465896606, "learning_rate": 7.185653189927763e-05, "loss": 0.0424, "step": 77640 }, { "grad_norm": 0.6657668352127075, "learning_rate": 7.184909596310412e-05, "loss": 0.0317, "step": 77650 }, { "grad_norm": 0.5527024269104004, "learning_rate": 7.184165942958452e-05, "loss": 0.0506, "step": 77660 }, { "grad_norm": 0.7254223823547363, "learning_rate": 7.183422229892217e-05, "loss": 0.0422, "step": 77670 }, { "grad_norm": 0.7347515821456909, "learning_rate": 7.182678457132039e-05, "loss": 0.0394, "step": 77680 }, { "grad_norm": 0.7327767014503479, "learning_rate": 7.18193462469825e-05, "loss": 0.0545, "step": 77690 }, { "grad_norm": 0.33487650752067566, "learning_rate": 7.181190732611189e-05, "loss": 0.034, "step": 77700 }, { "grad_norm": 0.4510541260242462, "learning_rate": 7.180446780891192e-05, "loss": 0.0255, "step": 77710 }, { "grad_norm": 0.4549007713794708, "learning_rate": 7.179702769558601e-05, "loss": 0.0305, "step": 77720 }, { "grad_norm": 0.4998437762260437, "learning_rate": 7.178958698633753e-05, "loss": 0.032, "step": 77730 }, { "grad_norm": 0.4787391126155853, "learning_rate": 7.178214568136993e-05, "loss": 0.0355, "step": 77740 }, { "grad_norm": 1.7092446088790894, "learning_rate": 7.177470378088667e-05, "loss": 0.0518, "step": 77750 }, { "grad_norm": 0.7474032044410706, "learning_rate": 7.176726128509117e-05, "loss": 0.0365, "step": 77760 }, { "grad_norm": 0.5800366401672363, "learning_rate": 7.175981819418695e-05, "loss": 0.0315, "step": 77770 }, { "grad_norm": 0.5954287052154541, "learning_rate": 7.175237450837743e-05, "loss": 0.0418, "step": 77780 }, { "grad_norm": 0.4301438331604004, "learning_rate": 7.174493022786621e-05, "loss": 0.0359, "step": 77790 }, { "grad_norm": 1.1551368236541748, "learning_rate": 7.173748535285674e-05, "loss": 0.0409, "step": 77800 }, { "grad_norm": 0.5610386729240417, "learning_rate": 7.17300398835526e-05, "loss": 0.0415, "step": 77810 }, { "grad_norm": 1.3277902603149414, "learning_rate": 7.172259382015733e-05, "loss": 0.0449, "step": 77820 }, { "grad_norm": 0.44984596967697144, "learning_rate": 7.17151471628745e-05, "loss": 0.0416, "step": 77830 }, { "grad_norm": 0.6330439448356628, "learning_rate": 7.170769991190771e-05, "loss": 0.0518, "step": 77840 }, { "grad_norm": 0.9556574821472168, "learning_rate": 7.170025206746055e-05, "loss": 0.0482, "step": 77850 }, { "grad_norm": 0.5180816650390625, "learning_rate": 7.169280362973665e-05, "loss": 0.0329, "step": 77860 }, { "grad_norm": 0.4328075051307678, "learning_rate": 7.168535459893966e-05, "loss": 0.0595, "step": 77870 }, { "grad_norm": 0.5219211578369141, "learning_rate": 7.167790497527323e-05, "loss": 0.0337, "step": 77880 }, { "grad_norm": 0.6166178584098816, "learning_rate": 7.1670454758941e-05, "loss": 0.0329, "step": 77890 }, { "grad_norm": 0.8545597791671753, "learning_rate": 7.16630039501467e-05, "loss": 0.0476, "step": 77900 }, { "grad_norm": 1.046680212020874, "learning_rate": 7.165555254909398e-05, "loss": 0.0325, "step": 77910 }, { "grad_norm": 0.2959180772304535, "learning_rate": 7.164810055598661e-05, "loss": 0.0264, "step": 77920 }, { "grad_norm": 0.8822286128997803, "learning_rate": 7.164064797102831e-05, "loss": 0.0317, "step": 77930 }, { "grad_norm": 0.739043653011322, "learning_rate": 7.16331947944228e-05, "loss": 0.0402, "step": 77940 }, { "grad_norm": 0.9129494428634644, "learning_rate": 7.16257410263739e-05, "loss": 0.0418, "step": 77950 }, { "grad_norm": 0.44696325063705444, "learning_rate": 7.161828666708535e-05, "loss": 0.051, "step": 77960 }, { "grad_norm": 0.30573761463165283, "learning_rate": 7.161083171676097e-05, "loss": 0.0236, "step": 77970 }, { "grad_norm": 0.4797230660915375, "learning_rate": 7.160337617560455e-05, "loss": 0.0473, "step": 77980 }, { "grad_norm": 0.37489938735961914, "learning_rate": 7.159592004381996e-05, "loss": 0.0336, "step": 77990 }, { "grad_norm": 0.5211006999015808, "learning_rate": 7.158846332161104e-05, "loss": 0.0314, "step": 78000 }, { "grad_norm": 0.2424902766942978, "learning_rate": 7.158100600918163e-05, "loss": 0.0306, "step": 78010 }, { "grad_norm": 0.4832924008369446, "learning_rate": 7.157354810673562e-05, "loss": 0.0282, "step": 78020 }, { "grad_norm": 1.1568502187728882, "learning_rate": 7.15660896144769e-05, "loss": 0.0368, "step": 78030 }, { "grad_norm": 2.2107317447662354, "learning_rate": 7.155863053260942e-05, "loss": 0.0405, "step": 78040 }, { "grad_norm": 0.6234841346740723, "learning_rate": 7.155117086133705e-05, "loss": 0.0442, "step": 78050 }, { "grad_norm": 0.45821690559387207, "learning_rate": 7.154371060086378e-05, "loss": 0.0439, "step": 78060 }, { "grad_norm": 0.4431573748588562, "learning_rate": 7.153624975139356e-05, "loss": 0.0352, "step": 78070 }, { "grad_norm": 0.6477583646774292, "learning_rate": 7.152878831313037e-05, "loss": 0.0391, "step": 78080 }, { "grad_norm": 0.5616345405578613, "learning_rate": 7.152132628627818e-05, "loss": 0.0315, "step": 78090 }, { "grad_norm": 0.6665463447570801, "learning_rate": 7.151386367104102e-05, "loss": 0.059, "step": 78100 }, { "grad_norm": 0.7411730885505676, "learning_rate": 7.150640046762289e-05, "loss": 0.0394, "step": 78110 }, { "grad_norm": 0.6421431303024292, "learning_rate": 7.149893667622789e-05, "loss": 0.0423, "step": 78120 }, { "grad_norm": 1.253194808959961, "learning_rate": 7.149147229706e-05, "loss": 0.0436, "step": 78130 }, { "grad_norm": 0.2905971109867096, "learning_rate": 7.148400733032337e-05, "loss": 0.0285, "step": 78140 }, { "grad_norm": 0.7192613482475281, "learning_rate": 7.147654177622202e-05, "loss": 0.0348, "step": 78150 }, { "grad_norm": 0.6237886548042297, "learning_rate": 7.146907563496008e-05, "loss": 0.0355, "step": 78160 }, { "grad_norm": 0.918590247631073, "learning_rate": 7.146160890674169e-05, "loss": 0.0349, "step": 78170 }, { "grad_norm": 0.49458274245262146, "learning_rate": 7.145414159177099e-05, "loss": 0.0421, "step": 78180 }, { "grad_norm": 0.6164936423301697, "learning_rate": 7.144667369025211e-05, "loss": 0.0457, "step": 78190 }, { "grad_norm": 0.9752443432807922, "learning_rate": 7.143920520238921e-05, "loss": 0.0282, "step": 78200 }, { "grad_norm": 0.3849298357963562, "learning_rate": 7.143173612838652e-05, "loss": 0.0319, "step": 78210 }, { "grad_norm": 0.7527434825897217, "learning_rate": 7.14242664684482e-05, "loss": 0.0537, "step": 78220 }, { "grad_norm": 0.46187299489974976, "learning_rate": 7.14167962227785e-05, "loss": 0.0271, "step": 78230 }, { "grad_norm": 0.43091440200805664, "learning_rate": 7.140932539158162e-05, "loss": 0.0267, "step": 78240 }, { "grad_norm": 0.4326973557472229, "learning_rate": 7.140185397506182e-05, "loss": 0.0326, "step": 78250 }, { "grad_norm": 0.6804349422454834, "learning_rate": 7.139438197342339e-05, "loss": 0.0295, "step": 78260 }, { "grad_norm": 0.33672621846199036, "learning_rate": 7.13869093868706e-05, "loss": 0.0249, "step": 78270 }, { "grad_norm": 0.4893495738506317, "learning_rate": 7.137943621560773e-05, "loss": 0.0302, "step": 78280 }, { "grad_norm": 0.2804599106311798, "learning_rate": 7.137196245983911e-05, "loss": 0.0244, "step": 78290 }, { "grad_norm": 0.4323281943798065, "learning_rate": 7.136448811976905e-05, "loss": 0.0683, "step": 78300 }, { "grad_norm": 0.5744103789329529, "learning_rate": 7.135701319560193e-05, "loss": 0.0405, "step": 78310 }, { "grad_norm": 0.5559276938438416, "learning_rate": 7.134953768754208e-05, "loss": 0.0299, "step": 78320 }, { "grad_norm": 0.4076835811138153, "learning_rate": 7.13420615957939e-05, "loss": 0.0263, "step": 78330 }, { "grad_norm": 0.3400672972202301, "learning_rate": 7.133458492056178e-05, "loss": 0.0312, "step": 78340 }, { "grad_norm": 0.48352929949760437, "learning_rate": 7.132710766205012e-05, "loss": 0.0295, "step": 78350 }, { "grad_norm": 0.35140112042427063, "learning_rate": 7.131962982046335e-05, "loss": 0.0275, "step": 78360 }, { "grad_norm": 0.3076847493648529, "learning_rate": 7.131215139600589e-05, "loss": 0.0316, "step": 78370 }, { "grad_norm": 0.3602690100669861, "learning_rate": 7.130467238888222e-05, "loss": 0.0442, "step": 78380 }, { "grad_norm": 1.321748971939087, "learning_rate": 7.129719279929682e-05, "loss": 0.0297, "step": 78390 }, { "grad_norm": 0.48104327917099, "learning_rate": 7.128971262745419e-05, "loss": 0.0338, "step": 78400 }, { "grad_norm": 0.36536723375320435, "learning_rate": 7.128223187355878e-05, "loss": 0.048, "step": 78410 }, { "grad_norm": 0.876115620136261, "learning_rate": 7.127475053781517e-05, "loss": 0.0333, "step": 78420 }, { "grad_norm": 0.7443119287490845, "learning_rate": 7.126726862042786e-05, "loss": 0.0355, "step": 78430 }, { "grad_norm": 0.36512327194213867, "learning_rate": 7.125978612160143e-05, "loss": 0.0325, "step": 78440 }, { "grad_norm": 0.5220986008644104, "learning_rate": 7.125230304154042e-05, "loss": 0.0337, "step": 78450 }, { "grad_norm": 0.3446093499660492, "learning_rate": 7.124481938044942e-05, "loss": 0.0478, "step": 78460 }, { "grad_norm": 2.863680124282837, "learning_rate": 7.123733513853305e-05, "loss": 0.0426, "step": 78470 }, { "grad_norm": 0.5581591725349426, "learning_rate": 7.122985031599594e-05, "loss": 0.0425, "step": 78480 }, { "grad_norm": 0.6027995944023132, "learning_rate": 7.122236491304266e-05, "loss": 0.0435, "step": 78490 }, { "grad_norm": 0.8222156167030334, "learning_rate": 7.121487892987792e-05, "loss": 0.038, "step": 78500 }, { "grad_norm": 0.40983375906944275, "learning_rate": 7.120739236670634e-05, "loss": 0.0374, "step": 78510 }, { "grad_norm": 0.7142005562782288, "learning_rate": 7.119990522373265e-05, "loss": 0.0352, "step": 78520 }, { "grad_norm": 0.8470039367675781, "learning_rate": 7.11924175011615e-05, "loss": 0.0422, "step": 78530 }, { "grad_norm": 0.6378800868988037, "learning_rate": 7.118492919919761e-05, "loss": 0.0335, "step": 78540 }, { "grad_norm": 0.5990773439407349, "learning_rate": 7.117744031804572e-05, "loss": 0.0416, "step": 78550 }, { "grad_norm": 0.7484697103500366, "learning_rate": 7.116995085791057e-05, "loss": 0.0446, "step": 78560 }, { "grad_norm": 0.7909823060035706, "learning_rate": 7.116246081899691e-05, "loss": 0.0689, "step": 78570 }, { "grad_norm": 0.8076735138893127, "learning_rate": 7.115497020150953e-05, "loss": 0.0471, "step": 78580 }, { "grad_norm": 0.5215859413146973, "learning_rate": 7.11474790056532e-05, "loss": 0.0367, "step": 78590 }, { "grad_norm": 1.069048523902893, "learning_rate": 7.113998723163277e-05, "loss": 0.045, "step": 78600 }, { "grad_norm": 0.7688201665878296, "learning_rate": 7.1132494879653e-05, "loss": 0.0484, "step": 78610 }, { "grad_norm": 0.5495613217353821, "learning_rate": 7.112500194991876e-05, "loss": 0.0346, "step": 78620 }, { "grad_norm": 0.6387389898300171, "learning_rate": 7.111750844263491e-05, "loss": 0.04, "step": 78630 }, { "grad_norm": 0.46446841955184937, "learning_rate": 7.111001435800632e-05, "loss": 0.0297, "step": 78640 }, { "grad_norm": 0.5889864563941956, "learning_rate": 7.110251969623785e-05, "loss": 0.048, "step": 78650 }, { "grad_norm": 0.5332767963409424, "learning_rate": 7.109502445753444e-05, "loss": 0.0476, "step": 78660 }, { "grad_norm": 0.3780938982963562, "learning_rate": 7.108752864210098e-05, "loss": 0.0356, "step": 78670 }, { "grad_norm": 0.685051441192627, "learning_rate": 7.10800322501424e-05, "loss": 0.0415, "step": 78680 }, { "grad_norm": 0.4589696228504181, "learning_rate": 7.107253528186366e-05, "loss": 0.0627, "step": 78690 }, { "grad_norm": 0.5098222494125366, "learning_rate": 7.106503773746972e-05, "loss": 0.0387, "step": 78700 }, { "grad_norm": 0.35036933422088623, "learning_rate": 7.105753961716557e-05, "loss": 0.0345, "step": 78710 }, { "grad_norm": 0.5890026688575745, "learning_rate": 7.105004092115619e-05, "loss": 0.0335, "step": 78720 }, { "grad_norm": 0.4512687623500824, "learning_rate": 7.104254164964662e-05, "loss": 0.0362, "step": 78730 }, { "grad_norm": 0.6100274324417114, "learning_rate": 7.103504180284185e-05, "loss": 0.0441, "step": 78740 }, { "grad_norm": 0.594754695892334, "learning_rate": 7.102754138094693e-05, "loss": 0.0335, "step": 78750 }, { "grad_norm": 0.28454676270484924, "learning_rate": 7.102004038416694e-05, "loss": 0.041, "step": 78760 }, { "grad_norm": 0.7767502069473267, "learning_rate": 7.101253881270695e-05, "loss": 0.0429, "step": 78770 }, { "grad_norm": 0.3359987735748291, "learning_rate": 7.100503666677202e-05, "loss": 0.0382, "step": 78780 }, { "grad_norm": 0.5962264537811279, "learning_rate": 7.099753394656732e-05, "loss": 0.0507, "step": 78790 }, { "grad_norm": 0.8228698968887329, "learning_rate": 7.099003065229789e-05, "loss": 0.0465, "step": 78800 }, { "grad_norm": 0.43485215306282043, "learning_rate": 7.098252678416892e-05, "loss": 0.0418, "step": 78810 }, { "grad_norm": 0.6166824102401733, "learning_rate": 7.097502234238555e-05, "loss": 0.0553, "step": 78820 }, { "grad_norm": 0.5906214118003845, "learning_rate": 7.096751732715297e-05, "loss": 0.0357, "step": 78830 }, { "grad_norm": 0.5623594522476196, "learning_rate": 7.096001173867631e-05, "loss": 0.0293, "step": 78840 }, { "grad_norm": 0.5524118542671204, "learning_rate": 7.095250557716082e-05, "loss": 0.0279, "step": 78850 }, { "grad_norm": 0.6144195199012756, "learning_rate": 7.094499884281169e-05, "loss": 0.0321, "step": 78860 }, { "grad_norm": 0.42318806052207947, "learning_rate": 7.093749153583417e-05, "loss": 0.0353, "step": 78870 }, { "grad_norm": 0.5100373029708862, "learning_rate": 7.09299836564335e-05, "loss": 0.0359, "step": 78880 }, { "grad_norm": 0.35142508149147034, "learning_rate": 7.092247520481495e-05, "loss": 0.0297, "step": 78890 }, { "grad_norm": 0.45917806029319763, "learning_rate": 7.091496618118376e-05, "loss": 0.0353, "step": 78900 }, { "grad_norm": 0.3775028586387634, "learning_rate": 7.090745658574526e-05, "loss": 0.0287, "step": 78910 }, { "grad_norm": 0.6915757656097412, "learning_rate": 7.089994641870477e-05, "loss": 0.0479, "step": 78920 }, { "grad_norm": 0.2869965732097626, "learning_rate": 7.089243568026756e-05, "loss": 0.0374, "step": 78930 }, { "grad_norm": 0.718154788017273, "learning_rate": 7.088492437063904e-05, "loss": 0.0329, "step": 78940 }, { "grad_norm": 0.5179542899131775, "learning_rate": 7.087741249002452e-05, "loss": 0.035, "step": 78950 }, { "grad_norm": 0.6843746900558472, "learning_rate": 7.086990003862939e-05, "loss": 0.0468, "step": 78960 }, { "grad_norm": 0.40488141775131226, "learning_rate": 7.086238701665904e-05, "loss": 0.0338, "step": 78970 }, { "grad_norm": 0.7209373712539673, "learning_rate": 7.085487342431886e-05, "loss": 0.0645, "step": 78980 }, { "grad_norm": 0.605069637298584, "learning_rate": 7.084735926181428e-05, "loss": 0.0247, "step": 78990 }, { "grad_norm": 0.3562522828578949, "learning_rate": 7.083984452935071e-05, "loss": 0.0368, "step": 79000 }, { "grad_norm": 0.6438388228416443, "learning_rate": 7.083232922713364e-05, "loss": 0.027, "step": 79010 }, { "grad_norm": 0.417865127325058, "learning_rate": 7.082481335536851e-05, "loss": 0.0369, "step": 79020 }, { "grad_norm": 0.6317320466041565, "learning_rate": 7.08172969142608e-05, "loss": 0.0355, "step": 79030 }, { "grad_norm": 0.4996958076953888, "learning_rate": 7.080977990401602e-05, "loss": 0.031, "step": 79040 }, { "grad_norm": 0.6619974970817566, "learning_rate": 7.080226232483968e-05, "loss": 0.0422, "step": 79050 }, { "grad_norm": 0.874061107635498, "learning_rate": 7.07947441769373e-05, "loss": 0.0436, "step": 79060 }, { "grad_norm": 0.3945363759994507, "learning_rate": 7.078722546051442e-05, "loss": 0.0335, "step": 79070 }, { "grad_norm": 0.3834901750087738, "learning_rate": 7.077970617577662e-05, "loss": 0.0365, "step": 79080 }, { "grad_norm": 0.5333412885665894, "learning_rate": 7.077218632292944e-05, "loss": 0.0404, "step": 79090 }, { "grad_norm": 0.7332721948623657, "learning_rate": 7.07646659021785e-05, "loss": 0.0643, "step": 79100 }, { "grad_norm": 0.4732929468154907, "learning_rate": 7.07571449137294e-05, "loss": 0.0406, "step": 79110 }, { "grad_norm": 0.7807289958000183, "learning_rate": 7.074962335778775e-05, "loss": 0.0508, "step": 79120 }, { "grad_norm": 0.5057656764984131, "learning_rate": 7.074210123455918e-05, "loss": 0.0375, "step": 79130 }, { "grad_norm": 0.49342381954193115, "learning_rate": 7.073457854424938e-05, "loss": 0.0296, "step": 79140 }, { "grad_norm": 0.3123627305030823, "learning_rate": 7.072705528706396e-05, "loss": 0.0307, "step": 79150 }, { "grad_norm": 0.9465705156326294, "learning_rate": 7.071953146320867e-05, "loss": 0.0502, "step": 79160 }, { "grad_norm": 0.5255388617515564, "learning_rate": 7.071200707288916e-05, "loss": 0.0473, "step": 79170 }, { "grad_norm": 0.4881448447704315, "learning_rate": 7.070448211631115e-05, "loss": 0.0454, "step": 79180 }, { "grad_norm": 0.32306158542633057, "learning_rate": 7.069695659368038e-05, "loss": 0.0348, "step": 79190 }, { "grad_norm": 4.903568744659424, "learning_rate": 7.06894305052026e-05, "loss": 0.0458, "step": 79200 }, { "grad_norm": 0.786691427230835, "learning_rate": 7.068190385108356e-05, "loss": 0.0343, "step": 79210 }, { "grad_norm": 0.8595239520072937, "learning_rate": 7.067437663152903e-05, "loss": 0.031, "step": 79220 }, { "grad_norm": 0.8807206749916077, "learning_rate": 7.066684884674482e-05, "loss": 0.0336, "step": 79230 }, { "grad_norm": 0.4423881769180298, "learning_rate": 7.065932049693673e-05, "loss": 0.0314, "step": 79240 }, { "grad_norm": 0.5779803395271301, "learning_rate": 7.065179158231058e-05, "loss": 0.0427, "step": 79250 }, { "grad_norm": 0.26098495721817017, "learning_rate": 7.06442621030722e-05, "loss": 0.0379, "step": 79260 }, { "grad_norm": 0.4822060763835907, "learning_rate": 7.063673205942745e-05, "loss": 0.0295, "step": 79270 }, { "grad_norm": 0.44227948784828186, "learning_rate": 7.062920145158219e-05, "loss": 0.0551, "step": 79280 }, { "grad_norm": 0.8915663957595825, "learning_rate": 7.062167027974233e-05, "loss": 0.0832, "step": 79290 }, { "grad_norm": 0.3845881521701813, "learning_rate": 7.061413854411373e-05, "loss": 0.0389, "step": 79300 }, { "grad_norm": 1.7951475381851196, "learning_rate": 7.060660624490234e-05, "loss": 0.0411, "step": 79310 }, { "grad_norm": 0.3864067494869232, "learning_rate": 7.059907338231407e-05, "loss": 0.0364, "step": 79320 }, { "grad_norm": 0.5974902510643005, "learning_rate": 7.059153995655488e-05, "loss": 0.0448, "step": 79330 }, { "grad_norm": 1.8632055521011353, "learning_rate": 7.058400596783072e-05, "loss": 0.046, "step": 79340 }, { "grad_norm": 0.6135521531105042, "learning_rate": 7.057647141634758e-05, "loss": 0.0336, "step": 79350 }, { "grad_norm": 0.729562520980835, "learning_rate": 7.056893630231143e-05, "loss": 0.0373, "step": 79360 }, { "grad_norm": 0.3665148615837097, "learning_rate": 7.056140062592828e-05, "loss": 0.0322, "step": 79370 }, { "grad_norm": 0.54080730676651, "learning_rate": 7.055386438740418e-05, "loss": 0.0393, "step": 79380 }, { "grad_norm": 0.5251129269599915, "learning_rate": 7.054632758694514e-05, "loss": 0.032, "step": 79390 }, { "grad_norm": 0.4833230674266815, "learning_rate": 7.053879022475722e-05, "loss": 0.0482, "step": 79400 }, { "grad_norm": 0.5075897574424744, "learning_rate": 7.05312523010465e-05, "loss": 0.0397, "step": 79410 }, { "grad_norm": 0.5345545411109924, "learning_rate": 7.052371381601906e-05, "loss": 0.0316, "step": 79420 }, { "grad_norm": 0.5585489273071289, "learning_rate": 7.051617476988098e-05, "loss": 0.0335, "step": 79430 }, { "grad_norm": 0.6654512882232666, "learning_rate": 7.05086351628384e-05, "loss": 0.0398, "step": 79440 }, { "grad_norm": 0.8065393567085266, "learning_rate": 7.050109499509742e-05, "loss": 0.0547, "step": 79450 }, { "grad_norm": 1.3376184701919556, "learning_rate": 7.049355426686423e-05, "loss": 0.0419, "step": 79460 }, { "grad_norm": 0.46151813864707947, "learning_rate": 7.048601297834495e-05, "loss": 0.0383, "step": 79470 }, { "grad_norm": 0.3184405565261841, "learning_rate": 7.047847112974578e-05, "loss": 0.0305, "step": 79480 }, { "grad_norm": 0.6130873560905457, "learning_rate": 7.04709287212729e-05, "loss": 0.0481, "step": 79490 }, { "grad_norm": 0.5028256773948669, "learning_rate": 7.046338575313253e-05, "loss": 0.0385, "step": 79500 }, { "grad_norm": 0.609732449054718, "learning_rate": 7.045584222553086e-05, "loss": 0.0448, "step": 79510 }, { "grad_norm": 0.757811427116394, "learning_rate": 7.044829813867416e-05, "loss": 0.037, "step": 79520 }, { "grad_norm": 1.0200860500335693, "learning_rate": 7.044075349276866e-05, "loss": 0.0377, "step": 79530 }, { "grad_norm": 0.4266391098499298, "learning_rate": 7.043320828802066e-05, "loss": 0.0484, "step": 79540 }, { "grad_norm": 0.8456494808197021, "learning_rate": 7.042566252463641e-05, "loss": 0.0499, "step": 79550 }, { "grad_norm": 0.6387339234352112, "learning_rate": 7.041811620282223e-05, "loss": 0.0337, "step": 79560 }, { "grad_norm": 1.015594482421875, "learning_rate": 7.04105693227844e-05, "loss": 0.0444, "step": 79570 }, { "grad_norm": 0.24731509387493134, "learning_rate": 7.04030218847293e-05, "loss": 0.0296, "step": 79580 }, { "grad_norm": 0.5876072645187378, "learning_rate": 7.039547388886322e-05, "loss": 0.0358, "step": 79590 }, { "grad_norm": 0.663205623626709, "learning_rate": 7.038792533539255e-05, "loss": 0.0477, "step": 79600 }, { "grad_norm": 0.46568453311920166, "learning_rate": 7.038037622452368e-05, "loss": 0.0313, "step": 79610 }, { "grad_norm": 0.2855457365512848, "learning_rate": 7.037282655646297e-05, "loss": 0.0417, "step": 79620 }, { "grad_norm": 0.26003485918045044, "learning_rate": 7.036527633141684e-05, "loss": 0.0259, "step": 79630 }, { "grad_norm": 0.3791815936565399, "learning_rate": 7.03577255495917e-05, "loss": 0.0478, "step": 79640 }, { "grad_norm": 0.5624582171440125, "learning_rate": 7.035017421119399e-05, "loss": 0.0402, "step": 79650 }, { "grad_norm": 1.5184695720672607, "learning_rate": 7.034262231643016e-05, "loss": 0.0328, "step": 79660 }, { "grad_norm": 0.8675004839897156, "learning_rate": 7.033506986550669e-05, "loss": 0.0354, "step": 79670 }, { "grad_norm": 2.0577611923217773, "learning_rate": 7.032751685863004e-05, "loss": 0.0461, "step": 79680 }, { "grad_norm": 0.6070743203163147, "learning_rate": 7.031996329600673e-05, "loss": 0.0347, "step": 79690 }, { "grad_norm": 0.4469676911830902, "learning_rate": 7.031240917784323e-05, "loss": 0.034, "step": 79700 }, { "grad_norm": 0.7308461666107178, "learning_rate": 7.030485450434612e-05, "loss": 0.0364, "step": 79710 }, { "grad_norm": 0.7638355493545532, "learning_rate": 7.02972992757219e-05, "loss": 0.0358, "step": 79720 }, { "grad_norm": 0.44201886653900146, "learning_rate": 7.028974349217716e-05, "loss": 0.0408, "step": 79730 }, { "grad_norm": 0.5482105016708374, "learning_rate": 7.028218715391845e-05, "loss": 0.0438, "step": 79740 }, { "grad_norm": 0.4168913960456848, "learning_rate": 7.027463026115235e-05, "loss": 0.0428, "step": 79750 }, { "grad_norm": 0.7130826711654663, "learning_rate": 7.026707281408549e-05, "loss": 0.0398, "step": 79760 }, { "grad_norm": 0.5688403248786926, "learning_rate": 7.025951481292446e-05, "loss": 0.0357, "step": 79770 }, { "grad_norm": 0.4176277816295624, "learning_rate": 7.02519562578759e-05, "loss": 0.0277, "step": 79780 }, { "grad_norm": 0.7600124478340149, "learning_rate": 7.02443971491465e-05, "loss": 0.0341, "step": 79790 }, { "grad_norm": 0.9392589330673218, "learning_rate": 7.023683748694286e-05, "loss": 0.0189, "step": 79800 }, { "grad_norm": 0.5986735820770264, "learning_rate": 7.022927727147169e-05, "loss": 0.0382, "step": 79810 }, { "grad_norm": 0.38384053111076355, "learning_rate": 7.022171650293968e-05, "loss": 0.0302, "step": 79820 }, { "grad_norm": 0.46255937218666077, "learning_rate": 7.021415518155354e-05, "loss": 0.0362, "step": 79830 }, { "grad_norm": 0.6545071005821228, "learning_rate": 7.020659330751997e-05, "loss": 0.0394, "step": 79840 }, { "grad_norm": 1.6210371255874634, "learning_rate": 7.019903088104576e-05, "loss": 0.0516, "step": 79850 }, { "grad_norm": 0.34888169169425964, "learning_rate": 7.019146790233762e-05, "loss": 0.0448, "step": 79860 }, { "grad_norm": 0.8035064339637756, "learning_rate": 7.018390437160233e-05, "loss": 0.0363, "step": 79870 }, { "grad_norm": 0.6182776689529419, "learning_rate": 7.01763402890467e-05, "loss": 0.0461, "step": 79880 }, { "grad_norm": 0.8391448259353638, "learning_rate": 7.016877565487747e-05, "loss": 0.0301, "step": 79890 }, { "grad_norm": 0.5922822952270508, "learning_rate": 7.016121046930151e-05, "loss": 0.0465, "step": 79900 }, { "grad_norm": 0.9384959936141968, "learning_rate": 7.015364473252563e-05, "loss": 0.0393, "step": 79910 }, { "grad_norm": 0.9572750329971313, "learning_rate": 7.014607844475669e-05, "loss": 0.0397, "step": 79920 }, { "grad_norm": 1.3037010431289673, "learning_rate": 7.013851160620152e-05, "loss": 0.0526, "step": 79930 }, { "grad_norm": 0.2821888327598572, "learning_rate": 7.013094421706701e-05, "loss": 0.0311, "step": 79940 }, { "grad_norm": 0.3996128737926483, "learning_rate": 7.012337627756005e-05, "loss": 0.038, "step": 79950 }, { "grad_norm": 0.44060513377189636, "learning_rate": 7.011580778788754e-05, "loss": 0.0394, "step": 79960 }, { "grad_norm": 1.017431378364563, "learning_rate": 7.010823874825638e-05, "loss": 0.0352, "step": 79970 }, { "grad_norm": 0.6563628315925598, "learning_rate": 7.010066915887356e-05, "loss": 0.0429, "step": 79980 }, { "grad_norm": 0.6635863184928894, "learning_rate": 7.009309901994598e-05, "loss": 0.0508, "step": 79990 }, { "grad_norm": 0.6298401951789856, "learning_rate": 7.008552833168063e-05, "loss": 0.0411, "step": 80000 }, { "grad_norm": 1.0866668224334717, "learning_rate": 7.007795709428447e-05, "loss": 0.0378, "step": 80010 }, { "grad_norm": 0.6450172066688538, "learning_rate": 7.007038530796451e-05, "loss": 0.0426, "step": 80020 }, { "grad_norm": 0.7142420411109924, "learning_rate": 7.006281297292775e-05, "loss": 0.0358, "step": 80030 }, { "grad_norm": 0.511346161365509, "learning_rate": 7.005524008938124e-05, "loss": 0.0317, "step": 80040 }, { "grad_norm": 0.3704622983932495, "learning_rate": 7.004766665753198e-05, "loss": 0.0313, "step": 80050 }, { "grad_norm": 0.7490355968475342, "learning_rate": 7.004009267758705e-05, "loss": 0.0535, "step": 80060 }, { "grad_norm": 1.3095020055770874, "learning_rate": 7.003251814975353e-05, "loss": 0.0367, "step": 80070 }, { "grad_norm": 0.4952647089958191, "learning_rate": 7.002494307423849e-05, "loss": 0.0432, "step": 80080 }, { "grad_norm": 0.49366357922554016, "learning_rate": 7.001736745124901e-05, "loss": 0.0426, "step": 80090 }, { "grad_norm": 0.9358307719230652, "learning_rate": 7.000979128099223e-05, "loss": 0.0418, "step": 80100 }, { "grad_norm": 0.5012035965919495, "learning_rate": 7.000221456367527e-05, "loss": 0.0376, "step": 80110 }, { "grad_norm": 0.7486971616744995, "learning_rate": 6.999463729950528e-05, "loss": 0.0348, "step": 80120 }, { "grad_norm": 0.6010900735855103, "learning_rate": 6.998705948868944e-05, "loss": 0.0356, "step": 80130 }, { "grad_norm": 0.5845351815223694, "learning_rate": 6.997948113143488e-05, "loss": 0.0518, "step": 80140 }, { "grad_norm": 0.5175660848617554, "learning_rate": 6.99719022279488e-05, "loss": 0.0345, "step": 80150 }, { "grad_norm": 0.6129593849182129, "learning_rate": 6.996432277843843e-05, "loss": 0.037, "step": 80160 }, { "grad_norm": 0.5488085746765137, "learning_rate": 6.9956742783111e-05, "loss": 0.0357, "step": 80170 }, { "grad_norm": 0.4919091463088989, "learning_rate": 6.994916224217368e-05, "loss": 0.0473, "step": 80180 }, { "grad_norm": 0.5038741230964661, "learning_rate": 6.99415811558338e-05, "loss": 0.0421, "step": 80190 }, { "grad_norm": 0.6202775835990906, "learning_rate": 6.993399952429854e-05, "loss": 0.0352, "step": 80200 }, { "grad_norm": 0.7581196427345276, "learning_rate": 6.992641734777524e-05, "loss": 0.037, "step": 80210 }, { "grad_norm": 0.4688354730606079, "learning_rate": 6.991883462647117e-05, "loss": 0.0337, "step": 80220 }, { "grad_norm": 0.4003241956233978, "learning_rate": 6.991125136059366e-05, "loss": 0.0338, "step": 80230 }, { "grad_norm": 0.3932352066040039, "learning_rate": 6.990366755035001e-05, "loss": 0.0396, "step": 80240 }, { "grad_norm": 0.7150887250900269, "learning_rate": 6.989608319594756e-05, "loss": 0.0365, "step": 80250 }, { "grad_norm": 0.6616036891937256, "learning_rate": 6.988849829759369e-05, "loss": 0.0373, "step": 80260 }, { "grad_norm": 0.9981945753097534, "learning_rate": 6.988091285549573e-05, "loss": 0.05, "step": 80270 }, { "grad_norm": 0.49898552894592285, "learning_rate": 6.987332686986108e-05, "loss": 0.0469, "step": 80280 }, { "grad_norm": 0.5553387403488159, "learning_rate": 6.986574034089714e-05, "loss": 0.0509, "step": 80290 }, { "grad_norm": 0.8469057679176331, "learning_rate": 6.985815326881133e-05, "loss": 0.0423, "step": 80300 }, { "grad_norm": 0.592304527759552, "learning_rate": 6.985056565381107e-05, "loss": 0.0399, "step": 80310 }, { "grad_norm": 0.7037029266357422, "learning_rate": 6.98429774961038e-05, "loss": 0.041, "step": 80320 }, { "grad_norm": 0.7858840823173523, "learning_rate": 6.983538879589698e-05, "loss": 0.0331, "step": 80330 }, { "grad_norm": 0.4096837043762207, "learning_rate": 6.982779955339808e-05, "loss": 0.029, "step": 80340 }, { "grad_norm": 2.2094130516052246, "learning_rate": 6.982020976881459e-05, "loss": 0.0315, "step": 80350 }, { "grad_norm": 0.6887569427490234, "learning_rate": 6.981261944235402e-05, "loss": 0.0431, "step": 80360 }, { "grad_norm": 1.2295050621032715, "learning_rate": 6.980502857422387e-05, "loss": 0.0489, "step": 80370 }, { "grad_norm": 0.4756251871585846, "learning_rate": 6.979743716463168e-05, "loss": 0.0301, "step": 80380 }, { "grad_norm": 0.4276006519794464, "learning_rate": 6.9789845213785e-05, "loss": 0.0257, "step": 80390 }, { "grad_norm": 0.49480852484703064, "learning_rate": 6.978225272189138e-05, "loss": 0.0253, "step": 80400 }, { "grad_norm": 1.8568782806396484, "learning_rate": 6.977465968915839e-05, "loss": 0.0426, "step": 80410 }, { "grad_norm": 0.3124328553676605, "learning_rate": 6.976706611579365e-05, "loss": 0.046, "step": 80420 }, { "grad_norm": 0.5046926140785217, "learning_rate": 6.975947200200475e-05, "loss": 0.0456, "step": 80430 }, { "grad_norm": 0.3844790458679199, "learning_rate": 6.975187734799931e-05, "loss": 0.0384, "step": 80440 }, { "grad_norm": 2.406102418899536, "learning_rate": 6.974428215398495e-05, "loss": 0.039, "step": 80450 }, { "grad_norm": 0.7463926076889038, "learning_rate": 6.973668642016935e-05, "loss": 0.0457, "step": 80460 }, { "grad_norm": 2.050140380859375, "learning_rate": 6.972909014676013e-05, "loss": 0.0618, "step": 80470 }, { "grad_norm": 0.4382683038711548, "learning_rate": 6.972149333396502e-05, "loss": 0.0317, "step": 80480 }, { "grad_norm": 0.861677885055542, "learning_rate": 6.97138959819917e-05, "loss": 0.0358, "step": 80490 }, { "grad_norm": 0.3368372619152069, "learning_rate": 6.970629809104784e-05, "loss": 0.0282, "step": 80500 }, { "grad_norm": 0.3227634131908417, "learning_rate": 6.969869966134123e-05, "loss": 0.0326, "step": 80510 }, { "grad_norm": 0.5354247093200684, "learning_rate": 6.969110069307955e-05, "loss": 0.0537, "step": 80520 }, { "grad_norm": 0.4439369738101959, "learning_rate": 6.968350118647057e-05, "loss": 0.0659, "step": 80530 }, { "grad_norm": 0.7188904285430908, "learning_rate": 6.967590114172207e-05, "loss": 0.0306, "step": 80540 }, { "grad_norm": 0.44986841082572937, "learning_rate": 6.966830055904183e-05, "loss": 0.0336, "step": 80550 }, { "grad_norm": 0.741797149181366, "learning_rate": 6.966069943863763e-05, "loss": 0.0357, "step": 80560 }, { "grad_norm": 0.4216362535953522, "learning_rate": 6.965309778071731e-05, "loss": 0.0421, "step": 80570 }, { "grad_norm": 0.6120709180831909, "learning_rate": 6.964549558548866e-05, "loss": 0.0338, "step": 80580 }, { "grad_norm": 3.3389666080474854, "learning_rate": 6.963789285315954e-05, "loss": 0.0448, "step": 80590 }, { "grad_norm": 0.924060046672821, "learning_rate": 6.963028958393783e-05, "loss": 0.0336, "step": 80600 }, { "grad_norm": 0.6390471458435059, "learning_rate": 6.962268577803137e-05, "loss": 0.0387, "step": 80610 }, { "grad_norm": 1.0190949440002441, "learning_rate": 6.961508143564805e-05, "loss": 0.0456, "step": 80620 }, { "grad_norm": 0.2951563894748688, "learning_rate": 6.960747655699578e-05, "loss": 0.0483, "step": 80630 }, { "grad_norm": 0.9370633363723755, "learning_rate": 6.959987114228245e-05, "loss": 0.0358, "step": 80640 }, { "grad_norm": 0.364092081785202, "learning_rate": 6.959226519171602e-05, "loss": 0.0439, "step": 80650 }, { "grad_norm": 0.4861561954021454, "learning_rate": 6.95846587055044e-05, "loss": 0.029, "step": 80660 }, { "grad_norm": 0.8591511249542236, "learning_rate": 6.957705168385559e-05, "loss": 0.0307, "step": 80670 }, { "grad_norm": 0.4476264417171478, "learning_rate": 6.956944412697753e-05, "loss": 0.032, "step": 80680 }, { "grad_norm": 0.37197569012641907, "learning_rate": 6.956183603507824e-05, "loss": 0.0437, "step": 80690 }, { "grad_norm": 0.5020182728767395, "learning_rate": 6.955422740836568e-05, "loss": 0.0352, "step": 80700 }, { "grad_norm": 0.4019205868244171, "learning_rate": 6.954661824704789e-05, "loss": 0.0331, "step": 80710 }, { "grad_norm": 0.7336605191230774, "learning_rate": 6.95390085513329e-05, "loss": 0.0368, "step": 80720 }, { "grad_norm": 0.7132397294044495, "learning_rate": 6.953139832142876e-05, "loss": 0.0397, "step": 80730 }, { "grad_norm": 0.42088690400123596, "learning_rate": 6.952378755754353e-05, "loss": 0.0365, "step": 80740 }, { "grad_norm": 0.5954570770263672, "learning_rate": 6.951617625988528e-05, "loss": 0.0345, "step": 80750 }, { "grad_norm": 0.6387289762496948, "learning_rate": 6.950856442866212e-05, "loss": 0.0446, "step": 80760 }, { "grad_norm": 0.4431745111942291, "learning_rate": 6.950095206408211e-05, "loss": 0.0313, "step": 80770 }, { "grad_norm": 0.5336154103279114, "learning_rate": 6.949333916635342e-05, "loss": 0.0407, "step": 80780 }, { "grad_norm": 0.5553488731384277, "learning_rate": 6.948572573568414e-05, "loss": 0.0288, "step": 80790 }, { "grad_norm": 0.5750706195831299, "learning_rate": 6.947811177228244e-05, "loss": 0.05, "step": 80800 }, { "grad_norm": 0.7036243081092834, "learning_rate": 6.947049727635651e-05, "loss": 0.0299, "step": 80810 }, { "grad_norm": 0.795318067073822, "learning_rate": 6.946288224811448e-05, "loss": 0.0516, "step": 80820 }, { "grad_norm": 0.6539400815963745, "learning_rate": 6.945526668776456e-05, "loss": 0.0346, "step": 80830 }, { "grad_norm": 0.47481656074523926, "learning_rate": 6.944765059551496e-05, "loss": 0.0488, "step": 80840 }, { "grad_norm": 0.8245769739151001, "learning_rate": 6.944003397157388e-05, "loss": 0.0481, "step": 80850 }, { "grad_norm": 0.5081828832626343, "learning_rate": 6.94324168161496e-05, "loss": 0.0386, "step": 80860 }, { "grad_norm": 1.3713362216949463, "learning_rate": 6.942479912945033e-05, "loss": 0.0282, "step": 80870 }, { "grad_norm": 0.4559114873409271, "learning_rate": 6.941718091168436e-05, "loss": 0.0331, "step": 80880 }, { "grad_norm": 0.4445834457874298, "learning_rate": 6.940956216305996e-05, "loss": 0.0361, "step": 80890 }, { "grad_norm": 0.515321671962738, "learning_rate": 6.940194288378543e-05, "loss": 0.0379, "step": 80900 }, { "grad_norm": 0.5021325349807739, "learning_rate": 6.939432307406904e-05, "loss": 0.0314, "step": 80910 }, { "grad_norm": 0.4800433814525604, "learning_rate": 6.938670273411916e-05, "loss": 0.0394, "step": 80920 }, { "grad_norm": 0.313764363527298, "learning_rate": 6.937908186414411e-05, "loss": 0.032, "step": 80930 }, { "grad_norm": 0.5842393040657043, "learning_rate": 6.937146046435226e-05, "loss": 0.0317, "step": 80940 }, { "grad_norm": 0.7134732604026794, "learning_rate": 6.936383853495194e-05, "loss": 0.0395, "step": 80950 }, { "grad_norm": 0.5559786558151245, "learning_rate": 6.935621607615155e-05, "loss": 0.0395, "step": 80960 }, { "grad_norm": 0.4076206386089325, "learning_rate": 6.934859308815949e-05, "loss": 0.0269, "step": 80970 }, { "grad_norm": 0.31502947211265564, "learning_rate": 6.934096957118416e-05, "loss": 0.0248, "step": 80980 }, { "grad_norm": 0.6021945476531982, "learning_rate": 6.933334552543398e-05, "loss": 0.0317, "step": 80990 }, { "grad_norm": 0.5998706221580505, "learning_rate": 6.93257209511174e-05, "loss": 0.032, "step": 81000 }, { "grad_norm": 0.9870142936706543, "learning_rate": 6.931809584844287e-05, "loss": 0.0405, "step": 81010 }, { "grad_norm": 0.43716129660606384, "learning_rate": 6.931047021761886e-05, "loss": 0.0321, "step": 81020 }, { "grad_norm": 1.719236969947815, "learning_rate": 6.930284405885385e-05, "loss": 0.0321, "step": 81030 }, { "grad_norm": 0.7581937313079834, "learning_rate": 6.929521737235634e-05, "loss": 0.0414, "step": 81040 }, { "grad_norm": 0.8530296087265015, "learning_rate": 6.928759015833483e-05, "loss": 0.0401, "step": 81050 }, { "grad_norm": 0.4816839098930359, "learning_rate": 6.927996241699785e-05, "loss": 0.0362, "step": 81060 }, { "grad_norm": 0.46681132912635803, "learning_rate": 6.927233414855394e-05, "loss": 0.0289, "step": 81070 }, { "grad_norm": 0.32287663221359253, "learning_rate": 6.926470535321166e-05, "loss": 0.0325, "step": 81080 }, { "grad_norm": 1.1158428192138672, "learning_rate": 6.925707603117958e-05, "loss": 0.0401, "step": 81090 }, { "grad_norm": 0.44396382570266724, "learning_rate": 6.924944618266624e-05, "loss": 0.0323, "step": 81100 }, { "grad_norm": 0.521956205368042, "learning_rate": 6.924181580788031e-05, "loss": 0.0317, "step": 81110 }, { "grad_norm": 0.5774103403091431, "learning_rate": 6.923418490703036e-05, "loss": 0.0433, "step": 81120 }, { "grad_norm": 0.8304595947265625, "learning_rate": 6.922655348032501e-05, "loss": 0.0408, "step": 81130 }, { "grad_norm": 0.9332305192947388, "learning_rate": 6.921892152797291e-05, "loss": 0.0384, "step": 81140 }, { "grad_norm": 0.737800657749176, "learning_rate": 6.921128905018274e-05, "loss": 0.0341, "step": 81150 }, { "grad_norm": 0.4312739372253418, "learning_rate": 6.920365604716311e-05, "loss": 0.0277, "step": 81160 }, { "grad_norm": 0.6503357887268066, "learning_rate": 6.919602251912276e-05, "loss": 0.0386, "step": 81170 }, { "grad_norm": 0.4363735318183899, "learning_rate": 6.918838846627034e-05, "loss": 0.0343, "step": 81180 }, { "grad_norm": 0.3363853394985199, "learning_rate": 6.918075388881461e-05, "loss": 0.0502, "step": 81190 }, { "grad_norm": 0.2562863230705261, "learning_rate": 6.917311878696428e-05, "loss": 0.0309, "step": 81200 }, { "grad_norm": 0.688110888004303, "learning_rate": 6.916548316092809e-05, "loss": 0.04, "step": 81210 }, { "grad_norm": 0.5738890767097473, "learning_rate": 6.915784701091476e-05, "loss": 0.0364, "step": 81220 }, { "grad_norm": 0.6669570207595825, "learning_rate": 6.915021033713311e-05, "loss": 0.0366, "step": 81230 }, { "grad_norm": 0.3878924250602722, "learning_rate": 6.914257313979192e-05, "loss": 0.0261, "step": 81240 }, { "grad_norm": 0.44071856141090393, "learning_rate": 6.913493541909995e-05, "loss": 0.0252, "step": 81250 }, { "grad_norm": 0.5297979712486267, "learning_rate": 6.912729717526602e-05, "loss": 0.0406, "step": 81260 }, { "grad_norm": 0.6166331171989441, "learning_rate": 6.9119658408499e-05, "loss": 0.0365, "step": 81270 }, { "grad_norm": 0.3433469533920288, "learning_rate": 6.91120191190077e-05, "loss": 0.0365, "step": 81280 }, { "grad_norm": 0.3917697072029114, "learning_rate": 6.910437930700095e-05, "loss": 0.0353, "step": 81290 }, { "grad_norm": 0.5788329243659973, "learning_rate": 6.909673897268767e-05, "loss": 0.0272, "step": 81300 }, { "grad_norm": 1.2623910903930664, "learning_rate": 6.90890981162767e-05, "loss": 0.0378, "step": 81310 }, { "grad_norm": 0.5929707884788513, "learning_rate": 6.908145673797699e-05, "loss": 0.0402, "step": 81320 }, { "grad_norm": 0.6214590668678284, "learning_rate": 6.90738148379974e-05, "loss": 0.0302, "step": 81330 }, { "grad_norm": 0.7661365270614624, "learning_rate": 6.906617241654689e-05, "loss": 0.0464, "step": 81340 }, { "grad_norm": 0.5028041005134583, "learning_rate": 6.90585294738344e-05, "loss": 0.0245, "step": 81350 }, { "grad_norm": 0.4849652349948883, "learning_rate": 6.905088601006884e-05, "loss": 0.0327, "step": 81360 }, { "grad_norm": 0.8253174424171448, "learning_rate": 6.904324202545923e-05, "loss": 0.0361, "step": 81370 }, { "grad_norm": 0.31447258591651917, "learning_rate": 6.903559752021455e-05, "loss": 0.0343, "step": 81380 }, { "grad_norm": 0.4444376826286316, "learning_rate": 6.902795249454377e-05, "loss": 0.0358, "step": 81390 }, { "grad_norm": 0.6078206300735474, "learning_rate": 6.902030694865592e-05, "loss": 0.0292, "step": 81400 }, { "grad_norm": 1.184991717338562, "learning_rate": 6.901266088276003e-05, "loss": 0.0336, "step": 81410 }, { "grad_norm": 0.45216304063796997, "learning_rate": 6.900501429706511e-05, "loss": 0.0372, "step": 81420 }, { "grad_norm": 1.2467305660247803, "learning_rate": 6.899736719178025e-05, "loss": 0.0404, "step": 81430 }, { "grad_norm": 0.4966106414794922, "learning_rate": 6.898971956711452e-05, "loss": 0.0347, "step": 81440 }, { "grad_norm": 0.4630013108253479, "learning_rate": 6.8982071423277e-05, "loss": 0.0399, "step": 81450 }, { "grad_norm": 0.3989631235599518, "learning_rate": 6.897442276047677e-05, "loss": 0.0346, "step": 81460 }, { "grad_norm": 0.6858770251274109, "learning_rate": 6.896677357892294e-05, "loss": 0.0323, "step": 81470 }, { "grad_norm": 0.5215345621109009, "learning_rate": 6.895912387882465e-05, "loss": 0.0366, "step": 81480 }, { "grad_norm": 0.5674731135368347, "learning_rate": 6.895147366039103e-05, "loss": 0.0374, "step": 81490 }, { "grad_norm": 0.5671048164367676, "learning_rate": 6.894382292383126e-05, "loss": 0.0453, "step": 81500 }, { "grad_norm": 0.38103199005126953, "learning_rate": 6.893617166935446e-05, "loss": 0.033, "step": 81510 }, { "grad_norm": 0.2849627733230591, "learning_rate": 6.892851989716986e-05, "loss": 0.0363, "step": 81520 }, { "grad_norm": 0.6261642575263977, "learning_rate": 6.892086760748664e-05, "loss": 0.0395, "step": 81530 }, { "grad_norm": 1.8807536363601685, "learning_rate": 6.891321480051399e-05, "loss": 0.0405, "step": 81540 }, { "grad_norm": 0.6531252264976501, "learning_rate": 6.890556147646117e-05, "loss": 0.0315, "step": 81550 }, { "grad_norm": 0.6148874759674072, "learning_rate": 6.889790763553738e-05, "loss": 0.0377, "step": 81560 }, { "grad_norm": 0.6181339621543884, "learning_rate": 6.889025327795191e-05, "loss": 0.0415, "step": 81570 }, { "grad_norm": 0.7230722308158875, "learning_rate": 6.888259840391402e-05, "loss": 0.0436, "step": 81580 }, { "grad_norm": 0.8425885438919067, "learning_rate": 6.887494301363299e-05, "loss": 0.0339, "step": 81590 }, { "grad_norm": 1.3286463022232056, "learning_rate": 6.88672871073181e-05, "loss": 0.0354, "step": 81600 }, { "grad_norm": 1.4714233875274658, "learning_rate": 6.885963068517864e-05, "loss": 0.0615, "step": 81610 }, { "grad_norm": 0.4975014328956604, "learning_rate": 6.885197374742398e-05, "loss": 0.0362, "step": 81620 }, { "grad_norm": 0.5992901921272278, "learning_rate": 6.884431629426345e-05, "loss": 0.0343, "step": 81630 }, { "grad_norm": 0.910018265247345, "learning_rate": 6.883665832590639e-05, "loss": 0.0423, "step": 81640 }, { "grad_norm": 0.9614725708961487, "learning_rate": 6.882899984256216e-05, "loss": 0.0282, "step": 81650 }, { "grad_norm": 0.6775650382041931, "learning_rate": 6.882134084444017e-05, "loss": 0.0311, "step": 81660 }, { "grad_norm": 3.1886489391326904, "learning_rate": 6.881368133174977e-05, "loss": 0.0521, "step": 81670 }, { "grad_norm": 0.6087259650230408, "learning_rate": 6.880602130470039e-05, "loss": 0.0398, "step": 81680 }, { "grad_norm": 0.5604485869407654, "learning_rate": 6.879836076350146e-05, "loss": 0.0305, "step": 81690 }, { "grad_norm": 2.5547516345977783, "learning_rate": 6.87906997083624e-05, "loss": 0.0294, "step": 81700 }, { "grad_norm": 0.5399351119995117, "learning_rate": 6.87830381394927e-05, "loss": 0.0288, "step": 81710 }, { "grad_norm": 0.6960867643356323, "learning_rate": 6.877537605710176e-05, "loss": 0.0428, "step": 81720 }, { "grad_norm": 0.5420109033584595, "learning_rate": 6.876771346139909e-05, "loss": 0.0446, "step": 81730 }, { "grad_norm": 1.5462183952331543, "learning_rate": 6.876005035259419e-05, "loss": 0.0299, "step": 81740 }, { "grad_norm": 0.5072069764137268, "learning_rate": 6.87523867308966e-05, "loss": 0.0328, "step": 81750 }, { "grad_norm": 0.6554868221282959, "learning_rate": 6.874472259651575e-05, "loss": 0.0478, "step": 81760 }, { "grad_norm": 0.4922180771827698, "learning_rate": 6.873705794966124e-05, "loss": 0.0488, "step": 81770 }, { "grad_norm": 0.421750009059906, "learning_rate": 6.872939279054263e-05, "loss": 0.0432, "step": 81780 }, { "grad_norm": 0.5439856052398682, "learning_rate": 6.872172711936943e-05, "loss": 0.0316, "step": 81790 }, { "grad_norm": 0.948265552520752, "learning_rate": 6.871406093635124e-05, "loss": 0.0499, "step": 81800 }, { "grad_norm": 0.9302405118942261, "learning_rate": 6.870639424169767e-05, "loss": 0.0409, "step": 81810 }, { "grad_norm": 0.5106658339500427, "learning_rate": 6.869872703561831e-05, "loss": 0.036, "step": 81820 }, { "grad_norm": 0.49893805384635925, "learning_rate": 6.869105931832277e-05, "loss": 0.0352, "step": 81830 }, { "grad_norm": 0.5774646997451782, "learning_rate": 6.86833910900207e-05, "loss": 0.0413, "step": 81840 }, { "grad_norm": 0.4451061487197876, "learning_rate": 6.867572235092173e-05, "loss": 0.0481, "step": 81850 }, { "grad_norm": 0.4336298406124115, "learning_rate": 6.866805310123554e-05, "loss": 0.0322, "step": 81860 }, { "grad_norm": 0.3950614929199219, "learning_rate": 6.866038334117176e-05, "loss": 0.0315, "step": 81870 }, { "grad_norm": 1.0395793914794922, "learning_rate": 6.865271307094014e-05, "loss": 0.027, "step": 81880 }, { "grad_norm": 0.4723246693611145, "learning_rate": 6.864504229075033e-05, "loss": 0.0387, "step": 81890 }, { "grad_norm": 0.7553706765174866, "learning_rate": 6.863737100081209e-05, "loss": 0.0426, "step": 81900 }, { "grad_norm": 0.3527469038963318, "learning_rate": 6.862969920133512e-05, "loss": 0.0263, "step": 81910 }, { "grad_norm": 0.3705296814441681, "learning_rate": 6.862202689252917e-05, "loss": 0.0347, "step": 81920 }, { "grad_norm": 0.5668182373046875, "learning_rate": 6.861435407460399e-05, "loss": 0.0354, "step": 81930 }, { "grad_norm": 0.5950717329978943, "learning_rate": 6.860668074776938e-05, "loss": 0.0411, "step": 81940 }, { "grad_norm": 1.0016428232192993, "learning_rate": 6.859900691223509e-05, "loss": 0.0447, "step": 81950 }, { "grad_norm": 0.4729713499546051, "learning_rate": 6.859133256821096e-05, "loss": 0.036, "step": 81960 }, { "grad_norm": 0.7405773401260376, "learning_rate": 6.858365771590678e-05, "loss": 0.0341, "step": 81970 }, { "grad_norm": 0.7726830244064331, "learning_rate": 6.857598235553237e-05, "loss": 0.0353, "step": 81980 }, { "grad_norm": 0.5634690523147583, "learning_rate": 6.856830648729758e-05, "loss": 0.0366, "step": 81990 }, { "grad_norm": 0.6081894040107727, "learning_rate": 6.856063011141227e-05, "loss": 0.0475, "step": 82000 }, { "grad_norm": 0.8529224395751953, "learning_rate": 6.85529532280863e-05, "loss": 0.0353, "step": 82010 }, { "grad_norm": 0.434195339679718, "learning_rate": 6.854527583752956e-05, "loss": 0.0375, "step": 82020 }, { "grad_norm": 0.43027496337890625, "learning_rate": 6.853759793995196e-05, "loss": 0.0351, "step": 82030 }, { "grad_norm": 0.5298347473144531, "learning_rate": 6.85299195355634e-05, "loss": 0.0349, "step": 82040 }, { "grad_norm": 0.39381539821624756, "learning_rate": 6.852224062457378e-05, "loss": 0.0309, "step": 82050 }, { "grad_norm": 0.9963536262512207, "learning_rate": 6.851456120719306e-05, "loss": 0.054, "step": 82060 }, { "grad_norm": 1.023220181465149, "learning_rate": 6.85068812836312e-05, "loss": 0.0522, "step": 82070 }, { "grad_norm": 0.6450381278991699, "learning_rate": 6.849920085409816e-05, "loss": 0.0534, "step": 82080 }, { "grad_norm": 0.8201969265937805, "learning_rate": 6.849151991880392e-05, "loss": 0.0635, "step": 82090 }, { "grad_norm": 0.539799690246582, "learning_rate": 6.848383847795847e-05, "loss": 0.0362, "step": 82100 }, { "grad_norm": 0.39902162551879883, "learning_rate": 6.847615653177182e-05, "loss": 0.0297, "step": 82110 }, { "grad_norm": 1.0288809537887573, "learning_rate": 6.846847408045398e-05, "loss": 0.0352, "step": 82120 }, { "grad_norm": 0.4954218566417694, "learning_rate": 6.846079112421501e-05, "loss": 0.038, "step": 82130 }, { "grad_norm": 0.42466092109680176, "learning_rate": 6.845310766326494e-05, "loss": 0.0425, "step": 82140 }, { "grad_norm": 0.3967413306236267, "learning_rate": 6.844542369781385e-05, "loss": 0.0334, "step": 82150 }, { "grad_norm": 0.7749739289283752, "learning_rate": 6.84377392280718e-05, "loss": 0.0293, "step": 82160 }, { "grad_norm": 0.590461790561676, "learning_rate": 6.84300542542489e-05, "loss": 0.0536, "step": 82170 }, { "grad_norm": 0.7616425156593323, "learning_rate": 6.842236877655521e-05, "loss": 0.0593, "step": 82180 }, { "grad_norm": 0.361219584941864, "learning_rate": 6.84146827952009e-05, "loss": 0.0381, "step": 82190 }, { "grad_norm": 0.34012550115585327, "learning_rate": 6.840699631039606e-05, "loss": 0.0384, "step": 82200 }, { "grad_norm": 0.6526980996131897, "learning_rate": 6.839930932235089e-05, "loss": 0.044, "step": 82210 }, { "grad_norm": 0.3769835829734802, "learning_rate": 6.83916218312755e-05, "loss": 0.0434, "step": 82220 }, { "grad_norm": 0.5047123432159424, "learning_rate": 6.838393383738006e-05, "loss": 0.0463, "step": 82230 }, { "grad_norm": 0.29743799567222595, "learning_rate": 6.83762453408748e-05, "loss": 0.0528, "step": 82240 }, { "grad_norm": 0.4859180152416229, "learning_rate": 6.836855634196988e-05, "loss": 0.0455, "step": 82250 }, { "grad_norm": 0.7284715175628662, "learning_rate": 6.836086684087553e-05, "loss": 0.033, "step": 82260 }, { "grad_norm": 0.43622833490371704, "learning_rate": 6.835317683780198e-05, "loss": 0.0351, "step": 82270 }, { "grad_norm": 0.9857273697853088, "learning_rate": 6.834548633295948e-05, "loss": 0.0457, "step": 82280 }, { "grad_norm": 0.936621904373169, "learning_rate": 6.833779532655828e-05, "loss": 0.0288, "step": 82290 }, { "grad_norm": 0.7022523283958435, "learning_rate": 6.833010381880864e-05, "loss": 0.0378, "step": 82300 }, { "grad_norm": 0.8266796469688416, "learning_rate": 6.832241180992083e-05, "loss": 0.0583, "step": 82310 }, { "grad_norm": 0.9401037693023682, "learning_rate": 6.831471930010518e-05, "loss": 0.0403, "step": 82320 }, { "grad_norm": 0.3126593232154846, "learning_rate": 6.830702628957198e-05, "loss": 0.0462, "step": 82330 }, { "grad_norm": 0.9532082080841064, "learning_rate": 6.829933277853158e-05, "loss": 0.0422, "step": 82340 }, { "grad_norm": 0.5401037931442261, "learning_rate": 6.829163876719426e-05, "loss": 0.0351, "step": 82350 }, { "grad_norm": 1.9920848608016968, "learning_rate": 6.828394425577044e-05, "loss": 0.0331, "step": 82360 }, { "grad_norm": 0.6144149899482727, "learning_rate": 6.827624924447044e-05, "loss": 0.0347, "step": 82370 }, { "grad_norm": 0.7073801159858704, "learning_rate": 6.826855373350465e-05, "loss": 0.0419, "step": 82380 }, { "grad_norm": 0.6145668029785156, "learning_rate": 6.826085772308346e-05, "loss": 0.0307, "step": 82390 }, { "grad_norm": 1.0978320837020874, "learning_rate": 6.825316121341729e-05, "loss": 0.036, "step": 82400 }, { "grad_norm": 0.27243199944496155, "learning_rate": 6.824546420471653e-05, "loss": 0.0294, "step": 82410 }, { "grad_norm": 0.49392107129096985, "learning_rate": 6.823776669719166e-05, "loss": 0.0302, "step": 82420 }, { "grad_norm": 0.7308969497680664, "learning_rate": 6.823006869105308e-05, "loss": 0.0273, "step": 82430 }, { "grad_norm": 0.5170980095863342, "learning_rate": 6.822237018651129e-05, "loss": 0.0335, "step": 82440 }, { "grad_norm": 0.7264529466629028, "learning_rate": 6.821467118377672e-05, "loss": 0.0461, "step": 82450 }, { "grad_norm": 0.6405812501907349, "learning_rate": 6.820697168305992e-05, "loss": 0.0472, "step": 82460 }, { "grad_norm": 0.4895496666431427, "learning_rate": 6.819927168457131e-05, "loss": 0.0331, "step": 82470 }, { "grad_norm": 1.0921109914779663, "learning_rate": 6.819157118852147e-05, "loss": 0.0357, "step": 82480 }, { "grad_norm": 0.47638392448425293, "learning_rate": 6.818387019512092e-05, "loss": 0.0258, "step": 82490 }, { "grad_norm": 0.49928247928619385, "learning_rate": 6.817616870458018e-05, "loss": 0.0443, "step": 82500 }, { "grad_norm": 1.604494571685791, "learning_rate": 6.816846671710981e-05, "loss": 0.0531, "step": 82510 }, { "grad_norm": 0.702415943145752, "learning_rate": 6.816076423292038e-05, "loss": 0.0366, "step": 82520 }, { "grad_norm": 0.3179170787334442, "learning_rate": 6.815306125222249e-05, "loss": 0.0311, "step": 82530 }, { "grad_norm": 0.5611768364906311, "learning_rate": 6.814535777522673e-05, "loss": 0.0628, "step": 82540 }, { "grad_norm": 0.8152701258659363, "learning_rate": 6.81376538021437e-05, "loss": 0.0373, "step": 82550 }, { "grad_norm": 0.9701687097549438, "learning_rate": 6.812994933318403e-05, "loss": 0.032, "step": 82560 }, { "grad_norm": 0.7362422943115234, "learning_rate": 6.812224436855836e-05, "loss": 0.0328, "step": 82570 }, { "grad_norm": 0.4590894281864166, "learning_rate": 6.811453890847732e-05, "loss": 0.034, "step": 82580 }, { "grad_norm": 0.9709847569465637, "learning_rate": 6.810683295315162e-05, "loss": 0.0329, "step": 82590 }, { "grad_norm": 0.4915940463542938, "learning_rate": 6.809912650279188e-05, "loss": 0.04, "step": 82600 }, { "grad_norm": 0.517127275466919, "learning_rate": 6.809141955760885e-05, "loss": 0.0359, "step": 82610 }, { "grad_norm": 0.3597553074359894, "learning_rate": 6.808371211781318e-05, "loss": 0.0316, "step": 82620 }, { "grad_norm": 0.4372170567512512, "learning_rate": 6.807600418361562e-05, "loss": 0.0378, "step": 82630 }, { "grad_norm": 0.6224778294563293, "learning_rate": 6.80682957552269e-05, "loss": 0.0334, "step": 82640 }, { "grad_norm": 0.41829055547714233, "learning_rate": 6.806058683285776e-05, "loss": 0.0271, "step": 82650 }, { "grad_norm": 0.611838161945343, "learning_rate": 6.805287741671896e-05, "loss": 0.0329, "step": 82660 }, { "grad_norm": 0.32785463333129883, "learning_rate": 6.804516750702129e-05, "loss": 0.0297, "step": 82670 }, { "grad_norm": 0.9170925617218018, "learning_rate": 6.803745710397549e-05, "loss": 0.0425, "step": 82680 }, { "grad_norm": 0.3775390684604645, "learning_rate": 6.802974620779241e-05, "loss": 0.0268, "step": 82690 }, { "grad_norm": 0.6827302575111389, "learning_rate": 6.802203481868283e-05, "loss": 0.0334, "step": 82700 }, { "grad_norm": 0.4009494483470917, "learning_rate": 6.801432293685761e-05, "loss": 0.0275, "step": 82710 }, { "grad_norm": 0.5328832864761353, "learning_rate": 6.800661056252755e-05, "loss": 0.0383, "step": 82720 }, { "grad_norm": 0.3713217079639435, "learning_rate": 6.799889769590353e-05, "loss": 0.0312, "step": 82730 }, { "grad_norm": 0.4022045135498047, "learning_rate": 6.79911843371964e-05, "loss": 0.0396, "step": 82740 }, { "grad_norm": 0.5269448161125183, "learning_rate": 6.798347048661706e-05, "loss": 0.0344, "step": 82750 }, { "grad_norm": 0.697291910648346, "learning_rate": 6.797575614437639e-05, "loss": 0.035, "step": 82760 }, { "grad_norm": 0.544689416885376, "learning_rate": 6.79680413106853e-05, "loss": 0.0391, "step": 82770 }, { "grad_norm": 0.354977548122406, "learning_rate": 6.796032598575471e-05, "loss": 0.027, "step": 82780 }, { "grad_norm": 0.45084336400032043, "learning_rate": 6.795261016979555e-05, "loss": 0.0323, "step": 82790 }, { "grad_norm": 0.7812487483024597, "learning_rate": 6.794489386301878e-05, "loss": 0.0369, "step": 82800 }, { "grad_norm": 0.7146138548851013, "learning_rate": 6.793717706563534e-05, "loss": 0.0391, "step": 82810 }, { "grad_norm": 1.4048678874969482, "learning_rate": 6.792945977785623e-05, "loss": 0.0295, "step": 82820 }, { "grad_norm": 0.31264984607696533, "learning_rate": 6.792174199989243e-05, "loss": 0.0339, "step": 82830 }, { "grad_norm": 1.0080498456954956, "learning_rate": 6.791402373195493e-05, "loss": 0.0288, "step": 82840 }, { "grad_norm": 0.49369263648986816, "learning_rate": 6.790630497425475e-05, "loss": 0.0333, "step": 82850 }, { "grad_norm": 0.6527290344238281, "learning_rate": 6.789858572700293e-05, "loss": 0.0431, "step": 82860 }, { "grad_norm": 0.43183279037475586, "learning_rate": 6.789086599041047e-05, "loss": 0.0284, "step": 82870 }, { "grad_norm": 0.45110592246055603, "learning_rate": 6.788314576468849e-05, "loss": 0.0319, "step": 82880 }, { "grad_norm": 1.3658896684646606, "learning_rate": 6.7875425050048e-05, "loss": 0.0453, "step": 82890 }, { "grad_norm": 1.3727350234985352, "learning_rate": 6.786770384670013e-05, "loss": 0.034, "step": 82900 }, { "grad_norm": 0.5798864364624023, "learning_rate": 6.785998215485591e-05, "loss": 0.0388, "step": 82910 }, { "grad_norm": 0.4460943639278412, "learning_rate": 6.785225997472653e-05, "loss": 0.0334, "step": 82920 }, { "grad_norm": 0.675331711769104, "learning_rate": 6.784453730652305e-05, "loss": 0.0289, "step": 82930 }, { "grad_norm": 0.6786914467811584, "learning_rate": 6.783681415045661e-05, "loss": 0.0328, "step": 82940 }, { "grad_norm": 0.3212434947490692, "learning_rate": 6.782909050673839e-05, "loss": 0.0357, "step": 82950 }, { "grad_norm": 0.335931658744812, "learning_rate": 6.782136637557953e-05, "loss": 0.0544, "step": 82960 }, { "grad_norm": 0.5624614357948303, "learning_rate": 6.781364175719121e-05, "loss": 0.0342, "step": 82970 }, { "grad_norm": 0.5853679180145264, "learning_rate": 6.780591665178461e-05, "loss": 0.0332, "step": 82980 }, { "grad_norm": 0.2465510368347168, "learning_rate": 6.779819105957095e-05, "loss": 0.034, "step": 82990 }, { "grad_norm": 0.3029837906360626, "learning_rate": 6.779046498076143e-05, "loss": 0.0324, "step": 83000 }, { "grad_norm": 0.46163105964660645, "learning_rate": 6.778273841556725e-05, "loss": 0.0346, "step": 83010 }, { "grad_norm": 0.8157814741134644, "learning_rate": 6.777501136419972e-05, "loss": 0.0483, "step": 83020 }, { "grad_norm": 0.6350067257881165, "learning_rate": 6.776728382687005e-05, "loss": 0.0387, "step": 83030 }, { "grad_norm": 1.0767947435379028, "learning_rate": 6.775955580378951e-05, "loss": 0.0403, "step": 83040 }, { "grad_norm": 0.45421329140663147, "learning_rate": 6.775182729516939e-05, "loss": 0.0493, "step": 83050 }, { "grad_norm": 0.48212987184524536, "learning_rate": 6.774409830122097e-05, "loss": 0.0332, "step": 83060 }, { "grad_norm": 0.7063401341438293, "learning_rate": 6.77363688221556e-05, "loss": 0.0322, "step": 83070 }, { "grad_norm": 0.3286513090133667, "learning_rate": 6.772863885818452e-05, "loss": 0.036, "step": 83080 }, { "grad_norm": 0.3255874216556549, "learning_rate": 6.772090840951915e-05, "loss": 0.0262, "step": 83090 }, { "grad_norm": 0.5078086256980896, "learning_rate": 6.771317747637078e-05, "loss": 0.0335, "step": 83100 }, { "grad_norm": 0.6055443286895752, "learning_rate": 6.770544605895082e-05, "loss": 0.0292, "step": 83110 }, { "grad_norm": 0.3419671356678009, "learning_rate": 6.769771415747059e-05, "loss": 0.0321, "step": 83120 }, { "grad_norm": 0.6856301426887512, "learning_rate": 6.768998177214153e-05, "loss": 0.0282, "step": 83130 }, { "grad_norm": 1.1011356115341187, "learning_rate": 6.768224890317499e-05, "loss": 0.0568, "step": 83140 }, { "grad_norm": 0.46416330337524414, "learning_rate": 6.767451555078242e-05, "loss": 0.0439, "step": 83150 }, { "grad_norm": 0.7327677607536316, "learning_rate": 6.766678171517522e-05, "loss": 0.066, "step": 83160 }, { "grad_norm": 0.4486088156700134, "learning_rate": 6.765904739656486e-05, "loss": 0.0322, "step": 83170 }, { "grad_norm": 0.48250508308410645, "learning_rate": 6.765131259516277e-05, "loss": 0.0294, "step": 83180 }, { "grad_norm": 1.5780725479125977, "learning_rate": 6.764357731118043e-05, "loss": 0.0284, "step": 83190 }, { "grad_norm": 0.3748950958251953, "learning_rate": 6.763584154482932e-05, "loss": 0.026, "step": 83200 }, { "grad_norm": 1.4279245138168335, "learning_rate": 6.762810529632092e-05, "loss": 0.0392, "step": 83210 }, { "grad_norm": 0.5557972192764282, "learning_rate": 6.762036856586674e-05, "loss": 0.0303, "step": 83220 }, { "grad_norm": 0.3853994905948639, "learning_rate": 6.76126313536783e-05, "loss": 0.0339, "step": 83230 }, { "grad_norm": 0.41649743914604187, "learning_rate": 6.760489365996717e-05, "loss": 0.0395, "step": 83240 }, { "grad_norm": 0.5182700753211975, "learning_rate": 6.759715548494482e-05, "loss": 0.0429, "step": 83250 }, { "grad_norm": 5.457329273223877, "learning_rate": 6.758941682882287e-05, "loss": 0.0273, "step": 83260 }, { "grad_norm": 0.4961279332637787, "learning_rate": 6.758167769181285e-05, "loss": 0.0312, "step": 83270 }, { "grad_norm": 0.42969074845314026, "learning_rate": 6.757393807412639e-05, "loss": 0.0263, "step": 83280 }, { "grad_norm": 0.6404895186424255, "learning_rate": 6.756619797597506e-05, "loss": 0.034, "step": 83290 }, { "grad_norm": 0.5918794870376587, "learning_rate": 6.755845739757048e-05, "loss": 0.0375, "step": 83300 }, { "grad_norm": 0.3590775728225708, "learning_rate": 6.755071633912426e-05, "loss": 0.0469, "step": 83310 }, { "grad_norm": 0.8365153074264526, "learning_rate": 6.754297480084807e-05, "loss": 0.0456, "step": 83320 }, { "grad_norm": 0.49109703302383423, "learning_rate": 6.75352327829535e-05, "loss": 0.0328, "step": 83330 }, { "grad_norm": 0.45287856459617615, "learning_rate": 6.752749028565229e-05, "loss": 0.0351, "step": 83340 }, { "grad_norm": 0.6641997694969177, "learning_rate": 6.751974730915607e-05, "loss": 0.0372, "step": 83350 }, { "grad_norm": 0.5437332391738892, "learning_rate": 6.751200385367655e-05, "loss": 0.0303, "step": 83360 }, { "grad_norm": 0.47450461983680725, "learning_rate": 6.75042599194254e-05, "loss": 0.045, "step": 83370 }, { "grad_norm": 0.576070249080658, "learning_rate": 6.749651550661438e-05, "loss": 0.0236, "step": 83380 }, { "grad_norm": 0.6321496367454529, "learning_rate": 6.74887706154552e-05, "loss": 0.0333, "step": 83390 }, { "grad_norm": 0.48129498958587646, "learning_rate": 6.748102524615959e-05, "loss": 0.0357, "step": 83400 }, { "grad_norm": 0.5917499661445618, "learning_rate": 6.747327939893931e-05, "loss": 0.0365, "step": 83410 }, { "grad_norm": 0.6556980013847351, "learning_rate": 6.746553307400617e-05, "loss": 0.0295, "step": 83420 }, { "grad_norm": 0.4203638732433319, "learning_rate": 6.745778627157187e-05, "loss": 0.0342, "step": 83430 }, { "grad_norm": 0.9400101900100708, "learning_rate": 6.745003899184827e-05, "loss": 0.0492, "step": 83440 }, { "grad_norm": 0.8663519620895386, "learning_rate": 6.744229123504718e-05, "loss": 0.0502, "step": 83450 }, { "grad_norm": 0.7527684569358826, "learning_rate": 6.74345430013804e-05, "loss": 0.0347, "step": 83460 }, { "grad_norm": 0.4229450821876526, "learning_rate": 6.742679429105974e-05, "loss": 0.0329, "step": 83470 }, { "grad_norm": 1.057205319404602, "learning_rate": 6.741904510429706e-05, "loss": 0.033, "step": 83480 }, { "grad_norm": 1.6510471105575562, "learning_rate": 6.741129544130427e-05, "loss": 0.0427, "step": 83490 }, { "grad_norm": 0.5448567271232605, "learning_rate": 6.740354530229318e-05, "loss": 0.0335, "step": 83500 }, { "grad_norm": 0.6167947053909302, "learning_rate": 6.739579468747571e-05, "loss": 0.0345, "step": 83510 }, { "grad_norm": 0.7881117463111877, "learning_rate": 6.738804359706373e-05, "loss": 0.0398, "step": 83520 }, { "grad_norm": 1.6181590557098389, "learning_rate": 6.738029203126918e-05, "loss": 0.0393, "step": 83530 }, { "grad_norm": 0.888859212398529, "learning_rate": 6.7372539990304e-05, "loss": 0.0349, "step": 83540 }, { "grad_norm": 0.7805138826370239, "learning_rate": 6.736478747438007e-05, "loss": 0.0481, "step": 83550 }, { "grad_norm": 0.9923081994056702, "learning_rate": 6.735703448370938e-05, "loss": 0.0457, "step": 83560 }, { "grad_norm": 0.5782198309898376, "learning_rate": 6.734928101850391e-05, "loss": 0.0383, "step": 83570 }, { "grad_norm": 0.7869483232498169, "learning_rate": 6.73415270789756e-05, "loss": 0.0362, "step": 83580 }, { "grad_norm": 0.5931791663169861, "learning_rate": 6.733377266533644e-05, "loss": 0.0361, "step": 83590 }, { "grad_norm": 0.5840591788291931, "learning_rate": 6.732601777779847e-05, "loss": 0.0405, "step": 83600 }, { "grad_norm": 0.5442534685134888, "learning_rate": 6.731826241657368e-05, "loss": 0.0314, "step": 83610 }, { "grad_norm": 0.5977967381477356, "learning_rate": 6.73105065818741e-05, "loss": 0.0323, "step": 83620 }, { "grad_norm": 0.8163633942604065, "learning_rate": 6.730275027391178e-05, "loss": 0.0244, "step": 83630 }, { "grad_norm": 0.8347161412239075, "learning_rate": 6.729499349289874e-05, "loss": 0.0315, "step": 83640 }, { "grad_norm": 0.6592842936515808, "learning_rate": 6.728723623904709e-05, "loss": 0.0311, "step": 83650 }, { "grad_norm": 0.5593166947364807, "learning_rate": 6.72794785125689e-05, "loss": 0.0353, "step": 83660 }, { "grad_norm": 1.0688692331314087, "learning_rate": 6.727172031367627e-05, "loss": 0.0373, "step": 83670 }, { "grad_norm": 0.586489200592041, "learning_rate": 6.726396164258128e-05, "loss": 0.0385, "step": 83680 }, { "grad_norm": 0.68483966588974, "learning_rate": 6.725620249949606e-05, "loss": 0.0468, "step": 83690 }, { "grad_norm": 0.6324185729026794, "learning_rate": 6.724844288463275e-05, "loss": 0.0366, "step": 83700 }, { "grad_norm": 2.1493091583251953, "learning_rate": 6.72406827982035e-05, "loss": 0.0356, "step": 83710 }, { "grad_norm": 0.3000001013278961, "learning_rate": 6.723292224042045e-05, "loss": 0.0542, "step": 83720 }, { "grad_norm": 0.6835293173789978, "learning_rate": 6.722516121149578e-05, "loss": 0.0344, "step": 83730 }, { "grad_norm": 0.6894592046737671, "learning_rate": 6.721739971164169e-05, "loss": 0.0326, "step": 83740 }, { "grad_norm": 0.669043242931366, "learning_rate": 6.720963774107035e-05, "loss": 0.0462, "step": 83750 }, { "grad_norm": 0.6277201771736145, "learning_rate": 6.720187529999398e-05, "loss": 0.0357, "step": 83760 }, { "grad_norm": 0.8193809390068054, "learning_rate": 6.71941123886248e-05, "loss": 0.0379, "step": 83770 }, { "grad_norm": 0.42731696367263794, "learning_rate": 6.718634900717504e-05, "loss": 0.0335, "step": 83780 }, { "grad_norm": 0.6420512199401855, "learning_rate": 6.717858515585696e-05, "loss": 0.0429, "step": 83790 }, { "grad_norm": 0.7657015919685364, "learning_rate": 6.717082083488282e-05, "loss": 0.0387, "step": 83800 }, { "grad_norm": 0.4605855345726013, "learning_rate": 6.716305604446488e-05, "loss": 0.0433, "step": 83810 }, { "grad_norm": 0.4975062608718872, "learning_rate": 6.715529078481545e-05, "loss": 0.032, "step": 83820 }, { "grad_norm": 0.5653146505355835, "learning_rate": 6.714752505614681e-05, "loss": 0.0257, "step": 83830 }, { "grad_norm": 0.582150399684906, "learning_rate": 6.713975885867126e-05, "loss": 0.033, "step": 83840 }, { "grad_norm": 0.29493439197540283, "learning_rate": 6.713199219260116e-05, "loss": 0.0424, "step": 83850 }, { "grad_norm": 0.5970385074615479, "learning_rate": 6.712422505814882e-05, "loss": 0.0379, "step": 83860 }, { "grad_norm": 0.7131367921829224, "learning_rate": 6.711645745552661e-05, "loss": 0.0388, "step": 83870 }, { "grad_norm": 0.6266258358955383, "learning_rate": 6.710868938494688e-05, "loss": 0.0256, "step": 83880 }, { "grad_norm": 0.6464564204216003, "learning_rate": 6.710092084662201e-05, "loss": 0.0295, "step": 83890 }, { "grad_norm": 0.5469743609428406, "learning_rate": 6.709315184076438e-05, "loss": 0.0417, "step": 83900 }, { "grad_norm": 0.5207571983337402, "learning_rate": 6.708538236758642e-05, "loss": 0.0343, "step": 83910 }, { "grad_norm": 0.4291350245475769, "learning_rate": 6.70776124273005e-05, "loss": 0.0387, "step": 83920 }, { "grad_norm": 1.78788423538208, "learning_rate": 6.70698420201191e-05, "loss": 0.0546, "step": 83930 }, { "grad_norm": 0.6299231648445129, "learning_rate": 6.706207114625461e-05, "loss": 0.0271, "step": 83940 }, { "grad_norm": 0.6544201374053955, "learning_rate": 6.705429980591951e-05, "loss": 0.0308, "step": 83950 }, { "grad_norm": 0.5704312324523926, "learning_rate": 6.704652799932625e-05, "loss": 0.0388, "step": 83960 }, { "grad_norm": 0.5886865854263306, "learning_rate": 6.703875572668733e-05, "loss": 0.0394, "step": 83970 }, { "grad_norm": 0.46944335103034973, "learning_rate": 6.703098298821523e-05, "loss": 0.0489, "step": 83980 }, { "grad_norm": 0.8631066083908081, "learning_rate": 6.702320978412246e-05, "loss": 0.0269, "step": 83990 }, { "grad_norm": 0.47572681307792664, "learning_rate": 6.701543611462152e-05, "loss": 0.0268, "step": 84000 }, { "grad_norm": 1.7023108005523682, "learning_rate": 6.700766197992495e-05, "loss": 0.0428, "step": 84010 }, { "grad_norm": 0.5530990958213806, "learning_rate": 6.69998873802453e-05, "loss": 0.0395, "step": 84020 }, { "grad_norm": 0.4231397807598114, "learning_rate": 6.699211231579509e-05, "loss": 0.0356, "step": 84030 }, { "grad_norm": 0.9401429891586304, "learning_rate": 6.698433678678693e-05, "loss": 0.0583, "step": 84040 }, { "grad_norm": 1.1307125091552734, "learning_rate": 6.69765607934334e-05, "loss": 0.0526, "step": 84050 }, { "grad_norm": 0.7674691081047058, "learning_rate": 6.696878433594706e-05, "loss": 0.0346, "step": 84060 }, { "grad_norm": 0.5056787133216858, "learning_rate": 6.696100741454054e-05, "loss": 0.0427, "step": 84070 }, { "grad_norm": 0.7155230045318604, "learning_rate": 6.695323002942642e-05, "loss": 0.0624, "step": 84080 }, { "grad_norm": 2.368305206298828, "learning_rate": 6.69454521808174e-05, "loss": 0.0426, "step": 84090 }, { "grad_norm": 0.5996204614639282, "learning_rate": 6.693767386892605e-05, "loss": 0.0586, "step": 84100 }, { "grad_norm": 0.4973355233669281, "learning_rate": 6.692989509396508e-05, "loss": 0.0331, "step": 84110 }, { "grad_norm": 0.6373593211174011, "learning_rate": 6.692211585614711e-05, "loss": 0.0366, "step": 84120 }, { "grad_norm": 0.6778777837753296, "learning_rate": 6.691433615568489e-05, "loss": 0.0433, "step": 84130 }, { "grad_norm": 0.48603692650794983, "learning_rate": 6.690655599279106e-05, "loss": 0.0351, "step": 84140 }, { "grad_norm": 0.8835902810096741, "learning_rate": 6.689877536767833e-05, "loss": 0.0347, "step": 84150 }, { "grad_norm": 0.4476558268070221, "learning_rate": 6.689099428055942e-05, "loss": 0.0346, "step": 84160 }, { "grad_norm": 0.6147987842559814, "learning_rate": 6.688321273164711e-05, "loss": 0.0353, "step": 84170 }, { "grad_norm": 1.3223702907562256, "learning_rate": 6.687543072115408e-05, "loss": 0.0524, "step": 84180 }, { "grad_norm": 0.6215734481811523, "learning_rate": 6.686764824929311e-05, "loss": 0.0386, "step": 84190 }, { "grad_norm": 2.733269453048706, "learning_rate": 6.6859865316277e-05, "loss": 0.0424, "step": 84200 }, { "grad_norm": 0.9288244247436523, "learning_rate": 6.685208192231846e-05, "loss": 0.0667, "step": 84210 }, { "grad_norm": 0.44167661666870117, "learning_rate": 6.684429806763038e-05, "loss": 0.0358, "step": 84220 }, { "grad_norm": 0.4329083561897278, "learning_rate": 6.683651375242548e-05, "loss": 0.126, "step": 84230 }, { "grad_norm": 0.6763886213302612, "learning_rate": 6.682872897691665e-05, "loss": 0.0427, "step": 84240 }, { "grad_norm": 0.4995121657848358, "learning_rate": 6.682094374131668e-05, "loss": 0.033, "step": 84250 }, { "grad_norm": 0.4579395353794098, "learning_rate": 6.681315804583843e-05, "loss": 0.0377, "step": 84260 }, { "grad_norm": 0.499058336019516, "learning_rate": 6.680537189069475e-05, "loss": 0.0408, "step": 84270 }, { "grad_norm": 0.6148247122764587, "learning_rate": 6.679758527609852e-05, "loss": 0.0357, "step": 84280 }, { "grad_norm": 0.8681789636611938, "learning_rate": 6.678979820226261e-05, "loss": 0.0367, "step": 84290 }, { "grad_norm": 0.2909725308418274, "learning_rate": 6.678201066939993e-05, "loss": 0.0386, "step": 84300 }, { "grad_norm": 0.6146725416183472, "learning_rate": 6.677422267772338e-05, "loss": 0.0285, "step": 84310 }, { "grad_norm": 0.5014064311981201, "learning_rate": 6.67664342274459e-05, "loss": 0.0363, "step": 84320 }, { "grad_norm": 1.181687831878662, "learning_rate": 6.67586453187804e-05, "loss": 0.0408, "step": 84330 }, { "grad_norm": 0.5218877792358398, "learning_rate": 6.675085595193984e-05, "loss": 0.0399, "step": 84340 }, { "grad_norm": 0.4011802077293396, "learning_rate": 6.674306612713714e-05, "loss": 0.0364, "step": 84350 }, { "grad_norm": 0.708263635635376, "learning_rate": 6.673527584458534e-05, "loss": 0.0376, "step": 84360 }, { "grad_norm": 0.8724576234817505, "learning_rate": 6.672748510449736e-05, "loss": 0.0519, "step": 84370 }, { "grad_norm": 0.8981980085372925, "learning_rate": 6.671969390708623e-05, "loss": 0.0377, "step": 84380 }, { "grad_norm": 0.5077463388442993, "learning_rate": 6.671190225256496e-05, "loss": 0.028, "step": 84390 }, { "grad_norm": 0.44562816619873047, "learning_rate": 6.670411014114653e-05, "loss": 0.0305, "step": 84400 }, { "grad_norm": 0.5471062064170837, "learning_rate": 6.669631757304403e-05, "loss": 0.0365, "step": 84410 }, { "grad_norm": 0.6805342435836792, "learning_rate": 6.668852454847048e-05, "loss": 0.0287, "step": 84420 }, { "grad_norm": 0.43169716000556946, "learning_rate": 6.668073106763893e-05, "loss": 0.0329, "step": 84430 }, { "grad_norm": 0.7750944495201111, "learning_rate": 6.667293713076246e-05, "loss": 0.0389, "step": 84440 }, { "grad_norm": 0.5237770676612854, "learning_rate": 6.666514273805416e-05, "loss": 0.0301, "step": 84450 }, { "grad_norm": 1.1132748126983643, "learning_rate": 6.665734788972711e-05, "loss": 0.0452, "step": 84460 }, { "grad_norm": 0.4751316010951996, "learning_rate": 6.664955258599444e-05, "loss": 0.0406, "step": 84470 }, { "grad_norm": 0.6630964875221252, "learning_rate": 6.664175682706924e-05, "loss": 0.0429, "step": 84480 }, { "grad_norm": 0.38246771693229675, "learning_rate": 6.663396061316467e-05, "loss": 0.0242, "step": 84490 }, { "grad_norm": 0.4226347506046295, "learning_rate": 6.662616394449385e-05, "loss": 0.0315, "step": 84500 }, { "grad_norm": 0.6369099617004395, "learning_rate": 6.661836682126998e-05, "loss": 0.0456, "step": 84510 }, { "grad_norm": 0.5916298031806946, "learning_rate": 6.661056924370618e-05, "loss": 0.0391, "step": 84520 }, { "grad_norm": 0.2606525123119354, "learning_rate": 6.660277121201567e-05, "loss": 0.0323, "step": 84530 }, { "grad_norm": 1.6827393770217896, "learning_rate": 6.659497272641163e-05, "loss": 0.033, "step": 84540 }, { "grad_norm": 0.4060286581516266, "learning_rate": 6.658717378710728e-05, "loss": 0.0375, "step": 84550 }, { "grad_norm": 0.42465099692344666, "learning_rate": 6.657937439431582e-05, "loss": 0.0345, "step": 84560 }, { "grad_norm": 0.34728386998176575, "learning_rate": 6.65715745482505e-05, "loss": 0.0285, "step": 84570 }, { "grad_norm": 0.5715715289115906, "learning_rate": 6.656377424912455e-05, "loss": 0.0452, "step": 84580 }, { "grad_norm": 0.2608732283115387, "learning_rate": 6.655597349715125e-05, "loss": 0.032, "step": 84590 }, { "grad_norm": 0.4104941487312317, "learning_rate": 6.654817229254383e-05, "loss": 0.0307, "step": 84600 }, { "grad_norm": 0.4072457551956177, "learning_rate": 6.65403706355156e-05, "loss": 0.0286, "step": 84610 }, { "grad_norm": 0.7460728287696838, "learning_rate": 6.653256852627987e-05, "loss": 0.0343, "step": 84620 }, { "grad_norm": 0.9807737469673157, "learning_rate": 6.652476596504991e-05, "loss": 0.0347, "step": 84630 }, { "grad_norm": 1.9998091459274292, "learning_rate": 6.651696295203907e-05, "loss": 0.0416, "step": 84640 }, { "grad_norm": 0.40016135573387146, "learning_rate": 6.650915948746064e-05, "loss": 0.0322, "step": 84650 }, { "grad_norm": 0.6323404312133789, "learning_rate": 6.650135557152802e-05, "loss": 0.0365, "step": 84660 }, { "grad_norm": 0.6745389103889465, "learning_rate": 6.649355120445453e-05, "loss": 0.0338, "step": 84670 }, { "grad_norm": 0.637881338596344, "learning_rate": 6.648574638645355e-05, "loss": 0.0594, "step": 84680 }, { "grad_norm": 1.0799942016601562, "learning_rate": 6.647794111773843e-05, "loss": 0.0387, "step": 84690 }, { "grad_norm": 0.45622381567955017, "learning_rate": 6.647013539852264e-05, "loss": 0.0339, "step": 84700 }, { "grad_norm": 0.3327871561050415, "learning_rate": 6.646232922901951e-05, "loss": 0.0279, "step": 84710 }, { "grad_norm": 0.31784364581108093, "learning_rate": 6.645452260944248e-05, "loss": 0.0374, "step": 84720 }, { "grad_norm": 0.5203825235366821, "learning_rate": 6.644671554000498e-05, "loss": 0.0404, "step": 84730 }, { "grad_norm": 0.7194780111312866, "learning_rate": 6.643890802092046e-05, "loss": 0.0323, "step": 84740 }, { "grad_norm": 0.5360890626907349, "learning_rate": 6.643110005240237e-05, "loss": 0.0353, "step": 84750 }, { "grad_norm": 0.5817570090293884, "learning_rate": 6.642329163466418e-05, "loss": 0.0251, "step": 84760 }, { "grad_norm": 1.2345300912857056, "learning_rate": 6.641548276791938e-05, "loss": 0.0428, "step": 84770 }, { "grad_norm": 0.5366148352622986, "learning_rate": 6.640767345238143e-05, "loss": 0.0445, "step": 84780 }, { "grad_norm": 0.9636908173561096, "learning_rate": 6.639986368826386e-05, "loss": 0.041, "step": 84790 }, { "grad_norm": 0.4379761815071106, "learning_rate": 6.639205347578017e-05, "loss": 0.0303, "step": 84800 }, { "grad_norm": 0.5187910199165344, "learning_rate": 6.638424281514391e-05, "loss": 0.0284, "step": 84810 }, { "grad_norm": 0.645892322063446, "learning_rate": 6.63764317065686e-05, "loss": 0.0333, "step": 84820 }, { "grad_norm": 0.8185587525367737, "learning_rate": 6.63686201502678e-05, "loss": 0.0414, "step": 84830 }, { "grad_norm": 0.7680517435073853, "learning_rate": 6.636080814645506e-05, "loss": 0.0404, "step": 84840 }, { "grad_norm": 0.6498132348060608, "learning_rate": 6.635299569534398e-05, "loss": 0.0452, "step": 84850 }, { "grad_norm": 0.8278177976608276, "learning_rate": 6.634518279714815e-05, "loss": 0.0347, "step": 84860 }, { "grad_norm": 1.4437814950942993, "learning_rate": 6.633736945208115e-05, "loss": 0.0522, "step": 84870 }, { "grad_norm": 0.5056778788566589, "learning_rate": 6.632955566035663e-05, "loss": 0.0281, "step": 84880 }, { "grad_norm": 0.498738557100296, "learning_rate": 6.632174142218816e-05, "loss": 0.0389, "step": 84890 }, { "grad_norm": 0.6632082462310791, "learning_rate": 6.631392673778943e-05, "loss": 0.0483, "step": 84900 }, { "grad_norm": 0.6639951467514038, "learning_rate": 6.630611160737408e-05, "loss": 0.042, "step": 84910 }, { "grad_norm": 0.4999467134475708, "learning_rate": 6.629829603115576e-05, "loss": 0.0361, "step": 84920 }, { "grad_norm": 0.7203319668769836, "learning_rate": 6.629048000934814e-05, "loss": 0.0404, "step": 84930 }, { "grad_norm": 0.932187020778656, "learning_rate": 6.628266354216491e-05, "loss": 0.0401, "step": 84940 }, { "grad_norm": 0.6527381539344788, "learning_rate": 6.62748466298198e-05, "loss": 0.0355, "step": 84950 }, { "grad_norm": 0.9163842797279358, "learning_rate": 6.626702927252648e-05, "loss": 0.0381, "step": 84960 }, { "grad_norm": 0.3554418087005615, "learning_rate": 6.625921147049871e-05, "loss": 0.0437, "step": 84970 }, { "grad_norm": 0.5558323860168457, "learning_rate": 6.625139322395019e-05, "loss": 0.033, "step": 84980 }, { "grad_norm": 0.6836515069007874, "learning_rate": 6.624357453309468e-05, "loss": 0.0359, "step": 84990 }, { "grad_norm": 1.1074670553207397, "learning_rate": 6.623575539814595e-05, "loss": 0.0291, "step": 85000 }, { "grad_norm": 0.5906737446784973, "learning_rate": 6.622793581931776e-05, "loss": 0.0361, "step": 85010 }, { "grad_norm": 0.6857844591140747, "learning_rate": 6.622011579682392e-05, "loss": 0.0319, "step": 85020 }, { "grad_norm": 0.45472314953804016, "learning_rate": 6.621229533087821e-05, "loss": 0.037, "step": 85030 }, { "grad_norm": 0.3679478168487549, "learning_rate": 6.620447442169441e-05, "loss": 0.039, "step": 85040 }, { "grad_norm": 0.369460791349411, "learning_rate": 6.619665306948638e-05, "loss": 0.0301, "step": 85050 }, { "grad_norm": 0.8097197413444519, "learning_rate": 6.618883127446793e-05, "loss": 0.0398, "step": 85060 }, { "grad_norm": 0.6306586265563965, "learning_rate": 6.618100903685294e-05, "loss": 0.0346, "step": 85070 }, { "grad_norm": 0.48957720398902893, "learning_rate": 6.617318635685522e-05, "loss": 0.0372, "step": 85080 }, { "grad_norm": 0.6777923107147217, "learning_rate": 6.616536323468868e-05, "loss": 0.0426, "step": 85090 }, { "grad_norm": 0.5545994639396667, "learning_rate": 6.615753967056717e-05, "loss": 0.0378, "step": 85100 }, { "grad_norm": 0.5139283537864685, "learning_rate": 6.614971566470459e-05, "loss": 0.0309, "step": 85110 }, { "grad_norm": 0.7775079607963562, "learning_rate": 6.614189121731486e-05, "loss": 0.0462, "step": 85120 }, { "grad_norm": 0.9737276434898376, "learning_rate": 6.613406632861189e-05, "loss": 0.0326, "step": 85130 }, { "grad_norm": 0.8286780118942261, "learning_rate": 6.61262409988096e-05, "loss": 0.0292, "step": 85140 }, { "grad_norm": 0.4173325002193451, "learning_rate": 6.611841522812194e-05, "loss": 0.0239, "step": 85150 }, { "grad_norm": 0.48652806878089905, "learning_rate": 6.611058901676287e-05, "loss": 0.034, "step": 85160 }, { "grad_norm": 0.6972877383232117, "learning_rate": 6.610276236494633e-05, "loss": 0.0295, "step": 85170 }, { "grad_norm": 0.5297415256500244, "learning_rate": 6.609493527288634e-05, "loss": 0.0556, "step": 85180 }, { "grad_norm": 0.6767376661300659, "learning_rate": 6.608710774079685e-05, "loss": 0.0282, "step": 85190 }, { "grad_norm": 0.6997869610786438, "learning_rate": 6.607927976889189e-05, "loss": 0.0293, "step": 85200 }, { "grad_norm": 0.31187188625335693, "learning_rate": 6.607145135738547e-05, "loss": 0.0363, "step": 85210 }, { "grad_norm": 0.636613667011261, "learning_rate": 6.606362250649158e-05, "loss": 0.0412, "step": 85220 }, { "grad_norm": 0.49596256017684937, "learning_rate": 6.60557932164243e-05, "loss": 0.0589, "step": 85230 }, { "grad_norm": 0.5256730914115906, "learning_rate": 6.604796348739767e-05, "loss": 0.0359, "step": 85240 }, { "grad_norm": 0.3971622884273529, "learning_rate": 6.604013331962574e-05, "loss": 0.0428, "step": 85250 }, { "grad_norm": 0.5708963871002197, "learning_rate": 6.60323027133226e-05, "loss": 0.0437, "step": 85260 }, { "grad_norm": 1.713489055633545, "learning_rate": 6.602447166870231e-05, "loss": 0.04, "step": 85270 }, { "grad_norm": 0.6517512202262878, "learning_rate": 6.6016640185979e-05, "loss": 0.0243, "step": 85280 }, { "grad_norm": 0.41094064712524414, "learning_rate": 6.600880826536677e-05, "loss": 0.0413, "step": 85290 }, { "grad_norm": 0.41627708077430725, "learning_rate": 6.600097590707972e-05, "loss": 0.0278, "step": 85300 }, { "grad_norm": 0.44668617844581604, "learning_rate": 6.5993143111332e-05, "loss": 0.0333, "step": 85310 }, { "grad_norm": 0.6174898743629456, "learning_rate": 6.598530987833776e-05, "loss": 0.0296, "step": 85320 }, { "grad_norm": 0.5437849760055542, "learning_rate": 6.597747620831116e-05, "loss": 0.0273, "step": 85330 }, { "grad_norm": 0.3603280186653137, "learning_rate": 6.596964210146636e-05, "loss": 0.0258, "step": 85340 }, { "grad_norm": 0.3056466579437256, "learning_rate": 6.596180755801754e-05, "loss": 0.0384, "step": 85350 }, { "grad_norm": 0.49324357509613037, "learning_rate": 6.59539725781789e-05, "loss": 0.0288, "step": 85360 }, { "grad_norm": 0.3744196593761444, "learning_rate": 6.594613716216464e-05, "loss": 0.044, "step": 85370 }, { "grad_norm": 0.5391978025436401, "learning_rate": 6.593830131018898e-05, "loss": 0.0419, "step": 85380 }, { "grad_norm": 1.8295782804489136, "learning_rate": 6.593046502246616e-05, "loss": 0.0327, "step": 85390 }, { "grad_norm": 0.5911775827407837, "learning_rate": 6.59226282992104e-05, "loss": 0.0362, "step": 85400 }, { "grad_norm": 0.6293957233428955, "learning_rate": 6.591479114063597e-05, "loss": 0.0416, "step": 85410 }, { "grad_norm": 0.4580369293689728, "learning_rate": 6.590695354695714e-05, "loss": 0.0355, "step": 85420 }, { "grad_norm": 0.3739114999771118, "learning_rate": 6.589911551838815e-05, "loss": 0.0567, "step": 85430 }, { "grad_norm": 0.5657283067703247, "learning_rate": 6.589127705514334e-05, "loss": 0.0354, "step": 85440 }, { "grad_norm": 0.434749960899353, "learning_rate": 6.588343815743697e-05, "loss": 0.0332, "step": 85450 }, { "grad_norm": 0.3998088240623474, "learning_rate": 6.587559882548337e-05, "loss": 0.0325, "step": 85460 }, { "grad_norm": 0.6076531410217285, "learning_rate": 6.586775905949687e-05, "loss": 0.0402, "step": 85470 }, { "grad_norm": 0.5221181511878967, "learning_rate": 6.585991885969179e-05, "loss": 0.0386, "step": 85480 }, { "grad_norm": 0.6308976411819458, "learning_rate": 6.58520782262825e-05, "loss": 0.0417, "step": 85490 }, { "grad_norm": 1.0566853284835815, "learning_rate": 6.584423715948332e-05, "loss": 0.0349, "step": 85500 }, { "grad_norm": 0.7865243554115295, "learning_rate": 6.583639565950868e-05, "loss": 0.0434, "step": 85510 }, { "grad_norm": 0.6290062069892883, "learning_rate": 6.58285537265729e-05, "loss": 0.039, "step": 85520 }, { "grad_norm": 0.5100104212760925, "learning_rate": 6.582071136089044e-05, "loss": 0.0282, "step": 85530 }, { "grad_norm": 0.5579366683959961, "learning_rate": 6.581286856267565e-05, "loss": 0.031, "step": 85540 }, { "grad_norm": 0.6235112547874451, "learning_rate": 6.5805025332143e-05, "loss": 0.0417, "step": 85550 }, { "grad_norm": 0.4831676781177521, "learning_rate": 6.579718166950686e-05, "loss": 0.0334, "step": 85560 }, { "grad_norm": 0.7951130270957947, "learning_rate": 6.578933757498174e-05, "loss": 0.042, "step": 85570 }, { "grad_norm": 0.6560613512992859, "learning_rate": 6.578149304878204e-05, "loss": 0.0414, "step": 85580 }, { "grad_norm": 0.42856565117836, "learning_rate": 6.577364809112227e-05, "loss": 0.0334, "step": 85590 }, { "grad_norm": 1.349197268486023, "learning_rate": 6.576580270221687e-05, "loss": 0.0344, "step": 85600 }, { "grad_norm": 0.5474837422370911, "learning_rate": 6.575795688228035e-05, "loss": 0.0331, "step": 85610 }, { "grad_norm": 0.7248299717903137, "learning_rate": 6.575011063152721e-05, "loss": 0.0422, "step": 85620 }, { "grad_norm": 0.3018192648887634, "learning_rate": 6.574226395017196e-05, "loss": 0.0389, "step": 85630 }, { "grad_norm": 0.5688248872756958, "learning_rate": 6.573441683842913e-05, "loss": 0.03, "step": 85640 }, { "grad_norm": 0.3118721544742584, "learning_rate": 6.572656929651327e-05, "loss": 0.0393, "step": 85650 }, { "grad_norm": 0.543168842792511, "learning_rate": 6.57187213246389e-05, "loss": 0.041, "step": 85660 }, { "grad_norm": 0.49100956320762634, "learning_rate": 6.571087292302059e-05, "loss": 0.0387, "step": 85670 }, { "grad_norm": 0.563819408416748, "learning_rate": 6.570302409187292e-05, "loss": 0.0674, "step": 85680 }, { "grad_norm": 0.7450470328330994, "learning_rate": 6.569517483141047e-05, "loss": 0.0455, "step": 85690 }, { "grad_norm": 0.6274452209472656, "learning_rate": 6.568732514184784e-05, "loss": 0.0321, "step": 85700 }, { "grad_norm": 0.44845739006996155, "learning_rate": 6.567947502339962e-05, "loss": 0.0281, "step": 85710 }, { "grad_norm": 0.7302260994911194, "learning_rate": 6.567162447628048e-05, "loss": 0.0433, "step": 85720 }, { "grad_norm": 0.6631133556365967, "learning_rate": 6.566377350070499e-05, "loss": 0.0304, "step": 85730 }, { "grad_norm": 0.9355072975158691, "learning_rate": 6.565592209688781e-05, "loss": 0.0511, "step": 85740 }, { "grad_norm": 0.4746742844581604, "learning_rate": 6.56480702650436e-05, "loss": 0.0378, "step": 85750 }, { "grad_norm": 0.5621349811553955, "learning_rate": 6.564021800538705e-05, "loss": 0.0344, "step": 85760 }, { "grad_norm": 0.37004947662353516, "learning_rate": 6.563236531813278e-05, "loss": 0.0275, "step": 85770 }, { "grad_norm": 0.8524054288864136, "learning_rate": 6.562451220349555e-05, "loss": 0.053, "step": 85780 }, { "grad_norm": 0.5313725471496582, "learning_rate": 6.561665866169001e-05, "loss": 0.0324, "step": 85790 }, { "grad_norm": 0.4229811429977417, "learning_rate": 6.560880469293089e-05, "loss": 0.0337, "step": 85800 }, { "grad_norm": 0.7309830188751221, "learning_rate": 6.56009502974329e-05, "loss": 0.0341, "step": 85810 }, { "grad_norm": 0.7244088053703308, "learning_rate": 6.559309547541083e-05, "loss": 0.0343, "step": 85820 }, { "grad_norm": 0.5059237480163574, "learning_rate": 6.558524022707935e-05, "loss": 0.0391, "step": 85830 }, { "grad_norm": 0.432303249835968, "learning_rate": 6.557738455265327e-05, "loss": 0.0415, "step": 85840 }, { "grad_norm": 0.6769042015075684, "learning_rate": 6.556952845234735e-05, "loss": 0.0525, "step": 85850 }, { "grad_norm": 0.6155324578285217, "learning_rate": 6.556167192637637e-05, "loss": 0.0335, "step": 85860 }, { "grad_norm": 0.731413722038269, "learning_rate": 6.555381497495514e-05, "loss": 0.0413, "step": 85870 }, { "grad_norm": 0.529789388179779, "learning_rate": 6.554595759829842e-05, "loss": 0.0444, "step": 85880 }, { "grad_norm": 0.49218928813934326, "learning_rate": 6.553809979662109e-05, "loss": 0.041, "step": 85890 }, { "grad_norm": 3.410583019256592, "learning_rate": 6.553024157013793e-05, "loss": 0.0345, "step": 85900 }, { "grad_norm": 0.45722150802612305, "learning_rate": 6.552238291906381e-05, "loss": 0.035, "step": 85910 }, { "grad_norm": 0.307632714509964, "learning_rate": 6.551452384361356e-05, "loss": 0.0392, "step": 85920 }, { "grad_norm": 1.2769629955291748, "learning_rate": 6.550666434400207e-05, "loss": 0.0362, "step": 85930 }, { "grad_norm": 0.761199951171875, "learning_rate": 6.54988044204442e-05, "loss": 0.0418, "step": 85940 }, { "grad_norm": 0.6917794346809387, "learning_rate": 6.549094407315484e-05, "loss": 0.0346, "step": 85950 }, { "grad_norm": 1.0220872163772583, "learning_rate": 6.548308330234887e-05, "loss": 0.0298, "step": 85960 }, { "grad_norm": 0.4069030284881592, "learning_rate": 6.547522210824126e-05, "loss": 0.0376, "step": 85970 }, { "grad_norm": 0.45676514506340027, "learning_rate": 6.546736049104686e-05, "loss": 0.0325, "step": 85980 }, { "grad_norm": 0.7112253308296204, "learning_rate": 6.545949845098064e-05, "loss": 0.0267, "step": 85990 }, { "grad_norm": 0.4669356048107147, "learning_rate": 6.545163598825755e-05, "loss": 0.0408, "step": 86000 }, { "grad_norm": 0.3984357714653015, "learning_rate": 6.544377310309252e-05, "loss": 0.0258, "step": 86010 }, { "grad_norm": 0.8050486445426941, "learning_rate": 6.543590979570053e-05, "loss": 0.0414, "step": 86020 }, { "grad_norm": 0.613030195236206, "learning_rate": 6.54280460662966e-05, "loss": 0.0232, "step": 86030 }, { "grad_norm": 0.51813805103302, "learning_rate": 6.542018191509565e-05, "loss": 0.0378, "step": 86040 }, { "grad_norm": 1.2126953601837158, "learning_rate": 6.541231734231275e-05, "loss": 0.0346, "step": 86050 }, { "grad_norm": 0.711219310760498, "learning_rate": 6.540445234816287e-05, "loss": 0.0484, "step": 86060 }, { "grad_norm": 1.2903156280517578, "learning_rate": 6.539658693286105e-05, "loss": 0.0505, "step": 86070 }, { "grad_norm": 0.5681939125061035, "learning_rate": 6.538872109662232e-05, "loss": 0.0378, "step": 86080 }, { "grad_norm": 0.5470255613327026, "learning_rate": 6.538085483966174e-05, "loss": 0.0408, "step": 86090 }, { "grad_norm": 0.6295542120933533, "learning_rate": 6.537298816219437e-05, "loss": 0.0341, "step": 86100 }, { "grad_norm": 0.4193183183670044, "learning_rate": 6.536512106443529e-05, "loss": 0.0305, "step": 86110 }, { "grad_norm": 0.6848756074905396, "learning_rate": 6.535725354659956e-05, "loss": 0.044, "step": 86120 }, { "grad_norm": 0.9145874977111816, "learning_rate": 6.534938560890229e-05, "loss": 0.0351, "step": 86130 }, { "grad_norm": 0.4568074345588684, "learning_rate": 6.534151725155858e-05, "loss": 0.0558, "step": 86140 }, { "grad_norm": 0.6653626561164856, "learning_rate": 6.533364847478355e-05, "loss": 0.0412, "step": 86150 }, { "grad_norm": 0.35387948155403137, "learning_rate": 6.532577927879235e-05, "loss": 0.0359, "step": 86160 }, { "grad_norm": 0.6994425654411316, "learning_rate": 6.531790966380008e-05, "loss": 0.0454, "step": 86170 }, { "grad_norm": 0.4319108724594116, "learning_rate": 6.531003963002193e-05, "loss": 0.0349, "step": 86180 }, { "grad_norm": 0.44324958324432373, "learning_rate": 6.530216917767304e-05, "loss": 0.0396, "step": 86190 }, { "grad_norm": 0.6805684566497803, "learning_rate": 6.529429830696859e-05, "loss": 0.0351, "step": 86200 }, { "grad_norm": 0.456434965133667, "learning_rate": 6.528642701812378e-05, "loss": 0.0305, "step": 86210 }, { "grad_norm": 0.4159567058086395, "learning_rate": 6.527855531135379e-05, "loss": 0.0395, "step": 86220 }, { "grad_norm": 0.6832863092422485, "learning_rate": 6.527068318687384e-05, "loss": 0.0369, "step": 86230 }, { "grad_norm": 0.49620962142944336, "learning_rate": 6.526281064489916e-05, "loss": 0.0279, "step": 86240 }, { "grad_norm": 0.877291202545166, "learning_rate": 6.525493768564496e-05, "loss": 0.0389, "step": 86250 }, { "grad_norm": 0.4582393169403076, "learning_rate": 6.52470643093265e-05, "loss": 0.0282, "step": 86260 }, { "grad_norm": 0.40151888132095337, "learning_rate": 6.5239190516159e-05, "loss": 0.0374, "step": 86270 }, { "grad_norm": 1.8845996856689453, "learning_rate": 6.523131630635779e-05, "loss": 0.036, "step": 86280 }, { "grad_norm": 0.631775438785553, "learning_rate": 6.522344168013811e-05, "loss": 0.0293, "step": 86290 }, { "grad_norm": 0.44136497378349304, "learning_rate": 6.521556663771524e-05, "loss": 0.0275, "step": 86300 }, { "grad_norm": 0.4868853688240051, "learning_rate": 6.520769117930452e-05, "loss": 0.0597, "step": 86310 }, { "grad_norm": 0.4478941857814789, "learning_rate": 6.519981530512121e-05, "loss": 0.0362, "step": 86320 }, { "grad_norm": 0.960692822933197, "learning_rate": 6.519193901538066e-05, "loss": 0.023, "step": 86330 }, { "grad_norm": 0.35334107279777527, "learning_rate": 6.518406231029822e-05, "loss": 0.0421, "step": 86340 }, { "grad_norm": 0.33613109588623047, "learning_rate": 6.517618519008922e-05, "loss": 0.0404, "step": 86350 }, { "grad_norm": 0.6001811027526855, "learning_rate": 6.516830765496902e-05, "loss": 0.0452, "step": 86360 }, { "grad_norm": 0.48928573727607727, "learning_rate": 6.516042970515298e-05, "loss": 0.0302, "step": 86370 }, { "grad_norm": 0.4642563462257385, "learning_rate": 6.515255134085649e-05, "loss": 0.0266, "step": 86380 }, { "grad_norm": 0.5423688292503357, "learning_rate": 6.514467256229493e-05, "loss": 0.0359, "step": 86390 }, { "grad_norm": 0.39771828055381775, "learning_rate": 6.513679336968372e-05, "loss": 0.0379, "step": 86400 }, { "grad_norm": 0.6693467497825623, "learning_rate": 6.512891376323827e-05, "loss": 0.0379, "step": 86410 }, { "grad_norm": 0.48813119530677795, "learning_rate": 6.5121033743174e-05, "loss": 0.0369, "step": 86420 }, { "grad_norm": 0.47261902689933777, "learning_rate": 6.511315330970635e-05, "loss": 0.0289, "step": 86430 }, { "grad_norm": 0.33637961745262146, "learning_rate": 6.510527246305075e-05, "loss": 0.0241, "step": 86440 }, { "grad_norm": 0.46068477630615234, "learning_rate": 6.509739120342269e-05, "loss": 0.0562, "step": 86450 }, { "grad_norm": 0.40217986702919006, "learning_rate": 6.50895095310376e-05, "loss": 0.033, "step": 86460 }, { "grad_norm": 0.5260971784591675, "learning_rate": 6.508162744611104e-05, "loss": 0.0359, "step": 86470 }, { "grad_norm": 0.36968865990638733, "learning_rate": 6.507374494885841e-05, "loss": 0.0335, "step": 86480 }, { "grad_norm": 0.7346203327178955, "learning_rate": 6.506586203949528e-05, "loss": 0.0266, "step": 86490 }, { "grad_norm": 0.7188349366188049, "learning_rate": 6.505797871823712e-05, "loss": 0.0523, "step": 86500 }, { "grad_norm": 0.7335404753684998, "learning_rate": 6.505009498529949e-05, "loss": 0.0427, "step": 86510 }, { "grad_norm": 0.5732551217079163, "learning_rate": 6.504221084089791e-05, "loss": 0.035, "step": 86520 }, { "grad_norm": 2.144299030303955, "learning_rate": 6.503432628524795e-05, "loss": 0.0326, "step": 86530 }, { "grad_norm": 0.3710188865661621, "learning_rate": 6.502644131856514e-05, "loss": 0.0292, "step": 86540 }, { "grad_norm": 0.5429688692092896, "learning_rate": 6.501855594106508e-05, "loss": 0.0336, "step": 86550 }, { "grad_norm": 0.6674046516418457, "learning_rate": 6.501067015296336e-05, "loss": 0.0362, "step": 86560 }, { "grad_norm": 0.34026646614074707, "learning_rate": 6.500278395447553e-05, "loss": 0.0278, "step": 86570 }, { "grad_norm": 0.7405988574028015, "learning_rate": 6.499489734581724e-05, "loss": 0.0464, "step": 86580 }, { "grad_norm": 0.5015499591827393, "learning_rate": 6.498701032720406e-05, "loss": 0.0285, "step": 86590 }, { "grad_norm": 0.5183930993080139, "learning_rate": 6.497912289885169e-05, "loss": 0.0312, "step": 86600 }, { "grad_norm": 0.530387818813324, "learning_rate": 6.497123506097571e-05, "loss": 0.0304, "step": 86610 }, { "grad_norm": 0.5927067995071411, "learning_rate": 6.496334681379179e-05, "loss": 0.0364, "step": 86620 }, { "grad_norm": 0.38767698407173157, "learning_rate": 6.495545815751558e-05, "loss": 0.0371, "step": 86630 }, { "grad_norm": 0.6151883602142334, "learning_rate": 6.494756909236277e-05, "loss": 0.0504, "step": 86640 }, { "grad_norm": 0.9921599626541138, "learning_rate": 6.493967961854903e-05, "loss": 0.0454, "step": 86650 }, { "grad_norm": 0.524695098400116, "learning_rate": 6.493178973629007e-05, "loss": 0.0369, "step": 86660 }, { "grad_norm": 0.6478843688964844, "learning_rate": 6.492389944580159e-05, "loss": 0.0377, "step": 86670 }, { "grad_norm": 0.42296141386032104, "learning_rate": 6.491600874729931e-05, "loss": 0.0306, "step": 86680 }, { "grad_norm": 0.6777343153953552, "learning_rate": 6.490811764099893e-05, "loss": 0.0387, "step": 86690 }, { "grad_norm": 0.5926294326782227, "learning_rate": 6.490022612711624e-05, "loss": 0.037, "step": 86700 }, { "grad_norm": 0.4713480472564697, "learning_rate": 6.489233420586695e-05, "loss": 0.0342, "step": 86710 }, { "grad_norm": 0.8146805167198181, "learning_rate": 6.488444187746684e-05, "loss": 0.0289, "step": 86720 }, { "grad_norm": 0.3555412292480469, "learning_rate": 6.487654914213168e-05, "loss": 0.046, "step": 86730 }, { "grad_norm": 0.7418999075889587, "learning_rate": 6.486865600007726e-05, "loss": 0.0273, "step": 86740 }, { "grad_norm": 1.6927798986434937, "learning_rate": 6.486076245151937e-05, "loss": 0.0506, "step": 86750 }, { "grad_norm": 0.3302925229072571, "learning_rate": 6.485286849667382e-05, "loss": 0.037, "step": 86760 }, { "grad_norm": 0.4522939920425415, "learning_rate": 6.484497413575643e-05, "loss": 0.0389, "step": 86770 }, { "grad_norm": 0.31615373492240906, "learning_rate": 6.483707936898301e-05, "loss": 0.0284, "step": 86780 }, { "grad_norm": 0.48395469784736633, "learning_rate": 6.482918419656941e-05, "loss": 0.0244, "step": 86790 }, { "grad_norm": 0.7398173809051514, "learning_rate": 6.48212886187315e-05, "loss": 0.0312, "step": 86800 }, { "grad_norm": 0.6143785715103149, "learning_rate": 6.481339263568513e-05, "loss": 0.0348, "step": 86810 }, { "grad_norm": 1.5083353519439697, "learning_rate": 6.480549624764615e-05, "loss": 0.0355, "step": 86820 }, { "grad_norm": 0.9247990846633911, "learning_rate": 6.479759945483048e-05, "loss": 0.037, "step": 86830 }, { "grad_norm": 0.4086726903915405, "learning_rate": 6.4789702257454e-05, "loss": 0.0252, "step": 86840 }, { "grad_norm": 0.4059419631958008, "learning_rate": 6.478180465573263e-05, "loss": 0.0368, "step": 86850 }, { "grad_norm": 0.5516572594642639, "learning_rate": 6.477390664988226e-05, "loss": 0.0383, "step": 86860 }, { "grad_norm": 2.2849087715148926, "learning_rate": 6.476600824011883e-05, "loss": 0.0455, "step": 86870 }, { "grad_norm": 0.62727290391922, "learning_rate": 6.475810942665828e-05, "loss": 0.0296, "step": 86880 }, { "grad_norm": 0.4689337909221649, "learning_rate": 6.475021020971657e-05, "loss": 0.036, "step": 86890 }, { "grad_norm": 0.2810176908969879, "learning_rate": 6.474231058950963e-05, "loss": 0.0399, "step": 86900 }, { "grad_norm": 0.6249993443489075, "learning_rate": 6.473441056625349e-05, "loss": 0.0249, "step": 86910 }, { "grad_norm": 0.4820728003978729, "learning_rate": 6.472651014016408e-05, "loss": 0.0373, "step": 86920 }, { "grad_norm": 0.4461962580680847, "learning_rate": 6.471860931145744e-05, "loss": 0.0376, "step": 86930 }, { "grad_norm": 0.4509619474411011, "learning_rate": 6.471070808034953e-05, "loss": 0.0271, "step": 86940 }, { "grad_norm": 0.6176578998565674, "learning_rate": 6.47028064470564e-05, "loss": 0.0343, "step": 86950 }, { "grad_norm": 0.28100982308387756, "learning_rate": 6.469490441179405e-05, "loss": 0.0311, "step": 86960 }, { "grad_norm": 0.5093041658401489, "learning_rate": 6.468700197477853e-05, "loss": 0.0394, "step": 86970 }, { "grad_norm": 0.7090531587600708, "learning_rate": 6.46790991362259e-05, "loss": 0.0344, "step": 86980 }, { "grad_norm": 0.5300183892250061, "learning_rate": 6.467119589635223e-05, "loss": 0.0333, "step": 86990 }, { "grad_norm": 0.7061729431152344, "learning_rate": 6.466329225537356e-05, "loss": 0.0535, "step": 87000 }, { "grad_norm": 0.43646037578582764, "learning_rate": 6.4655388213506e-05, "loss": 0.029, "step": 87010 }, { "grad_norm": 1.2288422584533691, "learning_rate": 6.464748377096562e-05, "loss": 0.0376, "step": 87020 }, { "grad_norm": 0.4507386088371277, "learning_rate": 6.463957892796854e-05, "loss": 0.0292, "step": 87030 }, { "grad_norm": 0.6275814175605774, "learning_rate": 6.463167368473087e-05, "loss": 0.036, "step": 87040 }, { "grad_norm": 0.5839857459068298, "learning_rate": 6.462376804146874e-05, "loss": 0.0282, "step": 87050 }, { "grad_norm": 0.33218270540237427, "learning_rate": 6.461586199839831e-05, "loss": 0.0331, "step": 87060 }, { "grad_norm": 0.7650495171546936, "learning_rate": 6.460795555573567e-05, "loss": 0.0371, "step": 87070 }, { "grad_norm": 1.0454504489898682, "learning_rate": 6.460004871369703e-05, "loss": 0.0278, "step": 87080 }, { "grad_norm": 0.4971683919429779, "learning_rate": 6.459214147249852e-05, "loss": 0.0299, "step": 87090 }, { "grad_norm": 0.8234139680862427, "learning_rate": 6.458423383235638e-05, "loss": 0.0421, "step": 87100 }, { "grad_norm": 0.6459274888038635, "learning_rate": 6.457632579348674e-05, "loss": 0.0334, "step": 87110 }, { "grad_norm": 0.47400814294815063, "learning_rate": 6.456841735610585e-05, "loss": 0.0355, "step": 87120 }, { "grad_norm": 0.5317636728286743, "learning_rate": 6.456050852042987e-05, "loss": 0.0397, "step": 87130 }, { "grad_norm": 0.9675278663635254, "learning_rate": 6.45525992866751e-05, "loss": 0.0454, "step": 87140 }, { "grad_norm": 0.6333765387535095, "learning_rate": 6.454468965505768e-05, "loss": 0.0378, "step": 87150 }, { "grad_norm": 0.6012119054794312, "learning_rate": 6.453677962579396e-05, "loss": 0.0449, "step": 87160 }, { "grad_norm": 0.4188756048679352, "learning_rate": 6.452886919910012e-05, "loss": 0.0325, "step": 87170 }, { "grad_norm": 0.3405313789844513, "learning_rate": 6.452095837519246e-05, "loss": 0.0277, "step": 87180 }, { "grad_norm": 0.3272692859172821, "learning_rate": 6.451304715428725e-05, "loss": 0.0287, "step": 87190 }, { "grad_norm": 1.2869482040405273, "learning_rate": 6.450513553660079e-05, "loss": 0.0379, "step": 87200 }, { "grad_norm": 0.4784961938858032, "learning_rate": 6.449722352234937e-05, "loss": 0.0357, "step": 87210 }, { "grad_norm": 0.5070918798446655, "learning_rate": 6.448931111174929e-05, "loss": 0.0279, "step": 87220 }, { "grad_norm": 0.46927568316459656, "learning_rate": 6.448139830501692e-05, "loss": 0.0362, "step": 87230 }, { "grad_norm": 0.45866528153419495, "learning_rate": 6.447348510236853e-05, "loss": 0.0362, "step": 87240 }, { "grad_norm": 0.5952214002609253, "learning_rate": 6.446557150402051e-05, "loss": 0.0402, "step": 87250 }, { "grad_norm": 0.41923680901527405, "learning_rate": 6.44576575101892e-05, "loss": 0.0327, "step": 87260 }, { "grad_norm": 0.6937089562416077, "learning_rate": 6.444974312109097e-05, "loss": 0.0349, "step": 87270 }, { "grad_norm": 0.9524398446083069, "learning_rate": 6.444182833694218e-05, "loss": 0.0285, "step": 87280 }, { "grad_norm": 0.7586380839347839, "learning_rate": 6.443391315795923e-05, "loss": 0.0382, "step": 87290 }, { "grad_norm": 0.538335919380188, "learning_rate": 6.442599758435851e-05, "loss": 0.0325, "step": 87300 }, { "grad_norm": 0.7606346011161804, "learning_rate": 6.441808161635645e-05, "loss": 0.021, "step": 87310 }, { "grad_norm": 0.5718783140182495, "learning_rate": 6.441016525416945e-05, "loss": 0.0422, "step": 87320 }, { "grad_norm": 0.49305459856987, "learning_rate": 6.440224849801395e-05, "loss": 0.0359, "step": 87330 }, { "grad_norm": 0.7214454412460327, "learning_rate": 6.439433134810638e-05, "loss": 0.0346, "step": 87340 }, { "grad_norm": 0.3894760012626648, "learning_rate": 6.43864138046632e-05, "loss": 0.0289, "step": 87350 }, { "grad_norm": 0.6212543845176697, "learning_rate": 6.437849586790087e-05, "loss": 0.0225, "step": 87360 }, { "grad_norm": 0.6967631578445435, "learning_rate": 6.437057753803588e-05, "loss": 0.0387, "step": 87370 }, { "grad_norm": 0.7063909769058228, "learning_rate": 6.436265881528468e-05, "loss": 0.0334, "step": 87380 }, { "grad_norm": 0.3906162977218628, "learning_rate": 6.43547396998638e-05, "loss": 0.0275, "step": 87390 }, { "grad_norm": 0.6532503962516785, "learning_rate": 6.434682019198971e-05, "loss": 0.0583, "step": 87400 }, { "grad_norm": 0.3105660378932953, "learning_rate": 6.433890029187896e-05, "loss": 0.0265, "step": 87410 }, { "grad_norm": 0.4956212341785431, "learning_rate": 6.433097999974805e-05, "loss": 0.0445, "step": 87420 }, { "grad_norm": 0.7650719285011292, "learning_rate": 6.432305931581355e-05, "loss": 0.0329, "step": 87430 }, { "grad_norm": 0.5109889507293701, "learning_rate": 6.431513824029197e-05, "loss": 0.024, "step": 87440 }, { "grad_norm": 0.4550449252128601, "learning_rate": 6.430721677339991e-05, "loss": 0.0334, "step": 87450 }, { "grad_norm": 0.37165507674217224, "learning_rate": 6.429929491535389e-05, "loss": 0.0302, "step": 87460 }, { "grad_norm": 0.582252562046051, "learning_rate": 6.429137266637053e-05, "loss": 0.0419, "step": 87470 }, { "grad_norm": 0.35949110984802246, "learning_rate": 6.428345002666641e-05, "loss": 0.0326, "step": 87480 }, { "grad_norm": 0.528541088104248, "learning_rate": 6.427552699645814e-05, "loss": 0.0362, "step": 87490 }, { "grad_norm": 0.6917791962623596, "learning_rate": 6.426760357596234e-05, "loss": 0.0324, "step": 87500 }, { "grad_norm": 0.5851161479949951, "learning_rate": 6.425967976539558e-05, "loss": 0.0323, "step": 87510 }, { "grad_norm": 0.7164677977561951, "learning_rate": 6.425175556497456e-05, "loss": 0.0393, "step": 87520 }, { "grad_norm": 0.5084128379821777, "learning_rate": 6.424383097491589e-05, "loss": 0.0434, "step": 87530 }, { "grad_norm": 0.5758817195892334, "learning_rate": 6.423590599543624e-05, "loss": 0.0399, "step": 87540 }, { "grad_norm": 0.5593875646591187, "learning_rate": 6.422798062675224e-05, "loss": 0.0422, "step": 87550 }, { "grad_norm": 0.778343677520752, "learning_rate": 6.422005486908062e-05, "loss": 0.0381, "step": 87560 }, { "grad_norm": 0.5093082189559937, "learning_rate": 6.421212872263805e-05, "loss": 0.0326, "step": 87570 }, { "grad_norm": 0.6878107190132141, "learning_rate": 6.42042021876412e-05, "loss": 0.0271, "step": 87580 }, { "grad_norm": 0.7198955416679382, "learning_rate": 6.41962752643068e-05, "loss": 0.035, "step": 87590 }, { "grad_norm": 0.7767063975334167, "learning_rate": 6.418834795285158e-05, "loss": 0.0429, "step": 87600 }, { "grad_norm": 0.9516448974609375, "learning_rate": 6.418042025349224e-05, "loss": 0.038, "step": 87610 }, { "grad_norm": 1.0839534997940063, "learning_rate": 6.417249216644555e-05, "loss": 0.0396, "step": 87620 }, { "grad_norm": 0.6478956341743469, "learning_rate": 6.416456369192825e-05, "loss": 0.0377, "step": 87630 }, { "grad_norm": 0.4791717827320099, "learning_rate": 6.41566348301571e-05, "loss": 0.0351, "step": 87640 }, { "grad_norm": 0.4055403470993042, "learning_rate": 6.414870558134886e-05, "loss": 0.0366, "step": 87650 }, { "grad_norm": 0.41210654377937317, "learning_rate": 6.414077594572033e-05, "loss": 0.029, "step": 87660 }, { "grad_norm": 4.156254768371582, "learning_rate": 6.413284592348831e-05, "loss": 0.0327, "step": 87670 }, { "grad_norm": 0.5049816966056824, "learning_rate": 6.412491551486959e-05, "loss": 0.0345, "step": 87680 }, { "grad_norm": 0.6947370767593384, "learning_rate": 6.411698472008097e-05, "loss": 0.0352, "step": 87690 }, { "grad_norm": 0.5768458843231201, "learning_rate": 6.410905353933931e-05, "loss": 0.0381, "step": 87700 }, { "grad_norm": 0.3332970440387726, "learning_rate": 6.41011219728614e-05, "loss": 0.0316, "step": 87710 }, { "grad_norm": 0.43638500571250916, "learning_rate": 6.409319002086414e-05, "loss": 0.0405, "step": 87720 }, { "grad_norm": 0.48673924803733826, "learning_rate": 6.408525768356435e-05, "loss": 0.029, "step": 87730 }, { "grad_norm": 0.49703073501586914, "learning_rate": 6.407732496117891e-05, "loss": 0.0249, "step": 87740 }, { "grad_norm": 0.7202176451683044, "learning_rate": 6.406939185392469e-05, "loss": 0.0276, "step": 87750 }, { "grad_norm": 0.3768273890018463, "learning_rate": 6.406145836201858e-05, "loss": 0.0278, "step": 87760 }, { "grad_norm": 0.652201771736145, "learning_rate": 6.405352448567748e-05, "loss": 0.04, "step": 87770 }, { "grad_norm": 0.46140652894973755, "learning_rate": 6.40455902251183e-05, "loss": 0.0258, "step": 87780 }, { "grad_norm": 0.3469311594963074, "learning_rate": 6.403765558055796e-05, "loss": 0.0292, "step": 87790 }, { "grad_norm": 0.33972111344337463, "learning_rate": 6.40297205522134e-05, "loss": 0.0427, "step": 87800 }, { "grad_norm": 2.3587605953216553, "learning_rate": 6.402178514030154e-05, "loss": 0.0349, "step": 87810 }, { "grad_norm": 0.6719914078712463, "learning_rate": 6.401384934503935e-05, "loss": 0.0311, "step": 87820 }, { "grad_norm": 0.3343023359775543, "learning_rate": 6.400591316664379e-05, "loss": 0.0232, "step": 87830 }, { "grad_norm": 0.816428542137146, "learning_rate": 6.39979766053318e-05, "loss": 0.0437, "step": 87840 }, { "grad_norm": 0.35433441400527954, "learning_rate": 6.399003966132043e-05, "loss": 0.0323, "step": 87850 }, { "grad_norm": 1.0016498565673828, "learning_rate": 6.398210233482659e-05, "loss": 0.0447, "step": 87860 }, { "grad_norm": 0.8625040650367737, "learning_rate": 6.397416462606734e-05, "loss": 0.0305, "step": 87870 }, { "grad_norm": 0.6548200845718384, "learning_rate": 6.396622653525967e-05, "loss": 0.0302, "step": 87880 }, { "grad_norm": 0.5186771750450134, "learning_rate": 6.395828806262065e-05, "loss": 0.0422, "step": 87890 }, { "grad_norm": 0.23087041079998016, "learning_rate": 6.395034920836724e-05, "loss": 0.0426, "step": 87900 }, { "grad_norm": 0.45114609599113464, "learning_rate": 6.394240997271655e-05, "loss": 0.0361, "step": 87910 }, { "grad_norm": 0.9611589908599854, "learning_rate": 6.393447035588558e-05, "loss": 0.0485, "step": 87920 }, { "grad_norm": 0.6955994963645935, "learning_rate": 6.392653035809144e-05, "loss": 0.0382, "step": 87930 }, { "grad_norm": 0.3149162828922272, "learning_rate": 6.39185899795512e-05, "loss": 0.0458, "step": 87940 }, { "grad_norm": 0.7860917448997498, "learning_rate": 6.391064922048193e-05, "loss": 0.0294, "step": 87950 }, { "grad_norm": 0.8809729814529419, "learning_rate": 6.390270808110074e-05, "loss": 0.0483, "step": 87960 }, { "grad_norm": 2.0433144569396973, "learning_rate": 6.389476656162472e-05, "loss": 0.04, "step": 87970 }, { "grad_norm": 0.6906751394271851, "learning_rate": 6.388682466227101e-05, "loss": 0.0515, "step": 87980 }, { "grad_norm": 0.4865773022174835, "learning_rate": 6.387888238325674e-05, "loss": 0.0299, "step": 87990 }, { "grad_norm": 0.4591600298881531, "learning_rate": 6.387093972479903e-05, "loss": 0.0344, "step": 88000 }, { "grad_norm": 1.9903204441070557, "learning_rate": 6.386299668711504e-05, "loss": 0.0273, "step": 88010 }, { "grad_norm": 0.7435190081596375, "learning_rate": 6.385505327042193e-05, "loss": 0.0397, "step": 88020 }, { "grad_norm": 1.2018969058990479, "learning_rate": 6.384710947493689e-05, "loss": 0.0423, "step": 88030 }, { "grad_norm": 0.375845730304718, "learning_rate": 6.383916530087704e-05, "loss": 0.0385, "step": 88040 }, { "grad_norm": 0.6576921343803406, "learning_rate": 6.383122074845962e-05, "loss": 0.0337, "step": 88050 }, { "grad_norm": 0.8798921704292297, "learning_rate": 6.382327581790185e-05, "loss": 0.0291, "step": 88060 }, { "grad_norm": 0.35659143328666687, "learning_rate": 6.38153305094209e-05, "loss": 0.0373, "step": 88070 }, { "grad_norm": 0.4549218714237213, "learning_rate": 6.380738482323398e-05, "loss": 0.0265, "step": 88080 }, { "grad_norm": 0.5614176988601685, "learning_rate": 6.379943875955837e-05, "loss": 0.0388, "step": 88090 }, { "grad_norm": 0.5908771753311157, "learning_rate": 6.379149231861129e-05, "loss": 0.0355, "step": 88100 }, { "grad_norm": 0.9351261854171753, "learning_rate": 6.378354550060997e-05, "loss": 0.0315, "step": 88110 }, { "grad_norm": 0.5417541861534119, "learning_rate": 6.377559830577171e-05, "loss": 0.0281, "step": 88120 }, { "grad_norm": 0.6433850526809692, "learning_rate": 6.376765073431377e-05, "loss": 0.0365, "step": 88130 }, { "grad_norm": 0.5635045170783997, "learning_rate": 6.375970278645342e-05, "loss": 0.0378, "step": 88140 }, { "grad_norm": 0.5830293297767639, "learning_rate": 6.375175446240797e-05, "loss": 0.0308, "step": 88150 }, { "grad_norm": 0.41350412368774414, "learning_rate": 6.374380576239473e-05, "loss": 0.0338, "step": 88160 }, { "grad_norm": 0.7004556655883789, "learning_rate": 6.373585668663098e-05, "loss": 0.0408, "step": 88170 }, { "grad_norm": 0.5459281802177429, "learning_rate": 6.372790723533408e-05, "loss": 0.053, "step": 88180 }, { "grad_norm": 0.6647343039512634, "learning_rate": 6.371995740872136e-05, "loss": 0.0426, "step": 88190 }, { "grad_norm": 0.5744580626487732, "learning_rate": 6.371200720701016e-05, "loss": 0.0345, "step": 88200 }, { "grad_norm": 0.5475499629974365, "learning_rate": 6.370405663041783e-05, "loss": 0.0361, "step": 88210 }, { "grad_norm": 0.48917296528816223, "learning_rate": 6.369610567916173e-05, "loss": 0.0308, "step": 88220 }, { "grad_norm": 0.4589247703552246, "learning_rate": 6.368815435345924e-05, "loss": 0.0395, "step": 88230 }, { "grad_norm": 0.8843109011650085, "learning_rate": 6.368020265352779e-05, "loss": 0.036, "step": 88240 }, { "grad_norm": 0.8640470504760742, "learning_rate": 6.36722505795847e-05, "loss": 0.0313, "step": 88250 }, { "grad_norm": 0.7724282741546631, "learning_rate": 6.366429813184744e-05, "loss": 0.0254, "step": 88260 }, { "grad_norm": 1.5412851572036743, "learning_rate": 6.36563453105334e-05, "loss": 0.0291, "step": 88270 }, { "grad_norm": 0.4467605948448181, "learning_rate": 6.364839211586e-05, "loss": 0.0304, "step": 88280 }, { "grad_norm": 0.2888883352279663, "learning_rate": 6.36404385480447e-05, "loss": 0.0338, "step": 88290 }, { "grad_norm": 2.2104334831237793, "learning_rate": 6.36324846073049e-05, "loss": 0.0471, "step": 88300 }, { "grad_norm": 0.27299195528030396, "learning_rate": 6.362453029385813e-05, "loss": 0.0326, "step": 88310 }, { "grad_norm": 1.012695074081421, "learning_rate": 6.36165756079218e-05, "loss": 0.0342, "step": 88320 }, { "grad_norm": 0.8145851492881775, "learning_rate": 6.360862054971344e-05, "loss": 0.0302, "step": 88330 }, { "grad_norm": 0.4594522714614868, "learning_rate": 6.360066511945047e-05, "loss": 0.029, "step": 88340 }, { "grad_norm": 0.6443594694137573, "learning_rate": 6.359270931735044e-05, "loss": 0.0343, "step": 88350 }, { "grad_norm": 1.250640630722046, "learning_rate": 6.358475314363084e-05, "loss": 0.0342, "step": 88360 }, { "grad_norm": 1.0992403030395508, "learning_rate": 6.357679659850919e-05, "loss": 0.0412, "step": 88370 }, { "grad_norm": 0.38306400179862976, "learning_rate": 6.356883968220303e-05, "loss": 0.0433, "step": 88380 }, { "grad_norm": 0.3798239827156067, "learning_rate": 6.35608823949299e-05, "loss": 0.0277, "step": 88390 }, { "grad_norm": 1.0919697284698486, "learning_rate": 6.355292473690732e-05, "loss": 0.0453, "step": 88400 }, { "grad_norm": 0.628645658493042, "learning_rate": 6.354496670835288e-05, "loss": 0.0359, "step": 88410 }, { "grad_norm": 0.5066344738006592, "learning_rate": 6.353700830948412e-05, "loss": 0.0342, "step": 88420 }, { "grad_norm": 0.8158690333366394, "learning_rate": 6.352904954051865e-05, "loss": 0.0341, "step": 88430 }, { "grad_norm": 0.9723860621452332, "learning_rate": 6.352109040167407e-05, "loss": 0.034, "step": 88440 }, { "grad_norm": 0.21686053276062012, "learning_rate": 6.351313089316794e-05, "loss": 0.0281, "step": 88450 }, { "grad_norm": 0.48726850748062134, "learning_rate": 6.350517101521788e-05, "loss": 0.0289, "step": 88460 }, { "grad_norm": 0.2640705406665802, "learning_rate": 6.349721076804152e-05, "loss": 0.032, "step": 88470 }, { "grad_norm": 1.671098232269287, "learning_rate": 6.34892501518565e-05, "loss": 0.0354, "step": 88480 }, { "grad_norm": 0.5834330916404724, "learning_rate": 6.348128916688045e-05, "loss": 0.0411, "step": 88490 }, { "grad_norm": 0.6150866746902466, "learning_rate": 6.3473327813331e-05, "loss": 0.0286, "step": 88500 }, { "grad_norm": 2.3381459712982178, "learning_rate": 6.346536609142584e-05, "loss": 0.0448, "step": 88510 }, { "grad_norm": 0.8308809995651245, "learning_rate": 6.345740400138264e-05, "loss": 0.0361, "step": 88520 }, { "grad_norm": 0.3593336045742035, "learning_rate": 6.344944154341905e-05, "loss": 0.0507, "step": 88530 }, { "grad_norm": 0.47139468789100647, "learning_rate": 6.34414787177528e-05, "loss": 0.0281, "step": 88540 }, { "grad_norm": 0.43036022782325745, "learning_rate": 6.343351552460155e-05, "loss": 0.039, "step": 88550 }, { "grad_norm": 0.3522017002105713, "learning_rate": 6.342555196418306e-05, "loss": 0.0392, "step": 88560 }, { "grad_norm": 0.8631135821342468, "learning_rate": 6.3417588036715e-05, "loss": 0.0467, "step": 88570 }, { "grad_norm": 0.5033754110336304, "learning_rate": 6.340962374241514e-05, "loss": 0.0365, "step": 88580 }, { "grad_norm": 1.0774874687194824, "learning_rate": 6.340165908150119e-05, "loss": 0.0308, "step": 88590 }, { "grad_norm": 0.5550287961959839, "learning_rate": 6.339369405419093e-05, "loss": 0.0371, "step": 88600 }, { "grad_norm": 0.5957057476043701, "learning_rate": 6.338572866070209e-05, "loss": 0.0348, "step": 88610 }, { "grad_norm": 0.35893702507019043, "learning_rate": 6.337776290125247e-05, "loss": 0.025, "step": 88620 }, { "grad_norm": 0.9594422578811646, "learning_rate": 6.336979677605983e-05, "loss": 0.0369, "step": 88630 }, { "grad_norm": 0.5449469685554504, "learning_rate": 6.336183028534198e-05, "loss": 0.0382, "step": 88640 }, { "grad_norm": 0.5571485757827759, "learning_rate": 6.33538634293167e-05, "loss": 0.03, "step": 88650 }, { "grad_norm": 0.5540116429328918, "learning_rate": 6.334589620820182e-05, "loss": 0.0376, "step": 88660 }, { "grad_norm": 0.4373927712440491, "learning_rate": 6.333792862221514e-05, "loss": 0.0524, "step": 88670 }, { "grad_norm": 0.6317875981330872, "learning_rate": 6.33299606715745e-05, "loss": 0.0368, "step": 88680 }, { "grad_norm": 0.2862307131290436, "learning_rate": 6.332199235649776e-05, "loss": 0.0357, "step": 88690 }, { "grad_norm": 0.8175702095031738, "learning_rate": 6.331402367720275e-05, "loss": 0.0565, "step": 88700 }, { "grad_norm": 0.5092836022377014, "learning_rate": 6.330605463390732e-05, "loss": 0.0419, "step": 88710 }, { "grad_norm": 0.5873498320579529, "learning_rate": 6.329808522682937e-05, "loss": 0.0348, "step": 88720 }, { "grad_norm": 0.4533498287200928, "learning_rate": 6.329011545618676e-05, "loss": 0.0312, "step": 88730 }, { "grad_norm": 0.7728625535964966, "learning_rate": 6.328214532219739e-05, "loss": 0.0253, "step": 88740 }, { "grad_norm": 0.5134332776069641, "learning_rate": 6.327417482507914e-05, "loss": 0.0371, "step": 88750 }, { "grad_norm": 0.5162253379821777, "learning_rate": 6.326620396504994e-05, "loss": 0.0367, "step": 88760 }, { "grad_norm": 0.6487773060798645, "learning_rate": 6.325823274232773e-05, "loss": 0.0332, "step": 88770 }, { "grad_norm": 0.7478353977203369, "learning_rate": 6.325026115713039e-05, "loss": 0.0372, "step": 88780 }, { "grad_norm": 2.205335855484009, "learning_rate": 6.324228920967592e-05, "loss": 0.0518, "step": 88790 }, { "grad_norm": 1.0517889261245728, "learning_rate": 6.323431690018223e-05, "loss": 0.0356, "step": 88800 }, { "grad_norm": 0.43775489926338196, "learning_rate": 6.322634422886726e-05, "loss": 0.0598, "step": 88810 }, { "grad_norm": 0.49832478165626526, "learning_rate": 6.321837119594902e-05, "loss": 0.0357, "step": 88820 }, { "grad_norm": 0.44745028018951416, "learning_rate": 6.32103978016455e-05, "loss": 0.0337, "step": 88830 }, { "grad_norm": 0.4731820523738861, "learning_rate": 6.320242404617465e-05, "loss": 0.039, "step": 88840 }, { "grad_norm": 0.5478082299232483, "learning_rate": 6.31944499297545e-05, "loss": 0.0324, "step": 88850 }, { "grad_norm": 0.32350608706474304, "learning_rate": 6.318647545260302e-05, "loss": 0.0282, "step": 88860 }, { "grad_norm": 0.45708122849464417, "learning_rate": 6.317850061493827e-05, "loss": 0.0254, "step": 88870 }, { "grad_norm": 0.5746208429336548, "learning_rate": 6.317052541697826e-05, "loss": 0.0306, "step": 88880 }, { "grad_norm": 0.3729532063007355, "learning_rate": 6.316254985894103e-05, "loss": 0.0317, "step": 88890 }, { "grad_norm": 0.6088263392448425, "learning_rate": 6.315457394104463e-05, "loss": 0.0351, "step": 88900 }, { "grad_norm": 0.4917856752872467, "learning_rate": 6.314659766350714e-05, "loss": 0.0301, "step": 88910 }, { "grad_norm": 0.5079635977745056, "learning_rate": 6.31386210265466e-05, "loss": 0.0416, "step": 88920 }, { "grad_norm": 0.44933679699897766, "learning_rate": 6.31306440303811e-05, "loss": 0.0264, "step": 88930 }, { "grad_norm": 0.6411623358726501, "learning_rate": 6.31226666752287e-05, "loss": 0.0434, "step": 88940 }, { "grad_norm": 0.4276975989341736, "learning_rate": 6.311468896130755e-05, "loss": 0.0269, "step": 88950 }, { "grad_norm": 0.6479074358940125, "learning_rate": 6.310671088883571e-05, "loss": 0.0355, "step": 88960 }, { "grad_norm": 0.7236341834068298, "learning_rate": 6.309873245803135e-05, "loss": 0.0327, "step": 88970 }, { "grad_norm": 0.27364274859428406, "learning_rate": 6.309075366911255e-05, "loss": 0.0219, "step": 88980 }, { "grad_norm": 0.7892254590988159, "learning_rate": 6.308277452229744e-05, "loss": 0.0375, "step": 88990 }, { "grad_norm": 0.9553754925727844, "learning_rate": 6.30747950178042e-05, "loss": 0.0372, "step": 89000 }, { "grad_norm": 0.6182077527046204, "learning_rate": 6.3066815155851e-05, "loss": 0.0356, "step": 89010 }, { "grad_norm": 0.3040671646595001, "learning_rate": 6.305883493665596e-05, "loss": 0.0307, "step": 89020 }, { "grad_norm": 0.4767138957977295, "learning_rate": 6.30508543604373e-05, "loss": 0.0413, "step": 89030 }, { "grad_norm": 0.31582093238830566, "learning_rate": 6.304287342741317e-05, "loss": 0.0318, "step": 89040 }, { "grad_norm": 0.45211392641067505, "learning_rate": 6.303489213780178e-05, "loss": 0.0282, "step": 89050 }, { "grad_norm": 0.38832888007164, "learning_rate": 6.302691049182133e-05, "loss": 0.0338, "step": 89060 }, { "grad_norm": 0.5845457911491394, "learning_rate": 6.301892848969004e-05, "loss": 0.0426, "step": 89070 }, { "grad_norm": 0.6163671016693115, "learning_rate": 6.301094613162616e-05, "loss": 0.0317, "step": 89080 }, { "grad_norm": 0.5301283597946167, "learning_rate": 6.300296341784788e-05, "loss": 0.055, "step": 89090 }, { "grad_norm": 1.1722412109375, "learning_rate": 6.299498034857348e-05, "loss": 0.0357, "step": 89100 }, { "grad_norm": 1.1565793752670288, "learning_rate": 6.298699692402118e-05, "loss": 0.0316, "step": 89110 }, { "grad_norm": 0.9442541599273682, "learning_rate": 6.297901314440929e-05, "loss": 0.0351, "step": 89120 }, { "grad_norm": 0.40894317626953125, "learning_rate": 6.297102900995602e-05, "loss": 0.0337, "step": 89130 }, { "grad_norm": 0.41305801272392273, "learning_rate": 6.296304452087971e-05, "loss": 0.0514, "step": 89140 }, { "grad_norm": 0.3104378581047058, "learning_rate": 6.295505967739863e-05, "loss": 0.0808, "step": 89150 }, { "grad_norm": 0.4560509920120239, "learning_rate": 6.294707447973109e-05, "loss": 0.0318, "step": 89160 }, { "grad_norm": 0.5248727202415466, "learning_rate": 6.29390889280954e-05, "loss": 0.0279, "step": 89170 }, { "grad_norm": 0.40851765871047974, "learning_rate": 6.293110302270987e-05, "loss": 0.048, "step": 89180 }, { "grad_norm": 0.44693121314048767, "learning_rate": 6.292311676379284e-05, "loss": 0.0338, "step": 89190 }, { "grad_norm": 0.5915687084197998, "learning_rate": 6.291513015156267e-05, "loss": 0.0322, "step": 89200 }, { "grad_norm": 0.588457465171814, "learning_rate": 6.290714318623767e-05, "loss": 0.0332, "step": 89210 }, { "grad_norm": 1.0254144668579102, "learning_rate": 6.289915586803625e-05, "loss": 0.0444, "step": 89220 }, { "grad_norm": 0.857882559299469, "learning_rate": 6.289116819717675e-05, "loss": 0.033, "step": 89230 }, { "grad_norm": 1.2761292457580566, "learning_rate": 6.288318017387753e-05, "loss": 0.051, "step": 89240 }, { "grad_norm": 0.2763124108314514, "learning_rate": 6.287519179835702e-05, "loss": 0.038, "step": 89250 }, { "grad_norm": 6.49591588973999, "learning_rate": 6.28672030708336e-05, "loss": 0.0464, "step": 89260 }, { "grad_norm": 0.5631810426712036, "learning_rate": 6.285921399152569e-05, "loss": 0.038, "step": 89270 }, { "grad_norm": 0.6731019020080566, "learning_rate": 6.285122456065169e-05, "loss": 0.049, "step": 89280 }, { "grad_norm": 0.7287037968635559, "learning_rate": 6.284323477843005e-05, "loss": 0.0295, "step": 89290 }, { "grad_norm": 0.4356590211391449, "learning_rate": 6.28352446450792e-05, "loss": 0.0419, "step": 89300 }, { "grad_norm": 0.9196372032165527, "learning_rate": 6.282725416081757e-05, "loss": 0.0318, "step": 89310 }, { "grad_norm": 0.4143083393573761, "learning_rate": 6.281926332586362e-05, "loss": 0.0388, "step": 89320 }, { "grad_norm": 0.5101024508476257, "learning_rate": 6.281127214043586e-05, "loss": 0.0309, "step": 89330 }, { "grad_norm": 0.8244770765304565, "learning_rate": 6.280328060475269e-05, "loss": 0.0348, "step": 89340 }, { "grad_norm": 0.5982266664505005, "learning_rate": 6.279528871903267e-05, "loss": 0.0337, "step": 89350 }, { "grad_norm": 0.9486157298088074, "learning_rate": 6.278729648349426e-05, "loss": 0.0419, "step": 89360 }, { "grad_norm": 0.5963096618652344, "learning_rate": 6.277930389835597e-05, "loss": 0.0437, "step": 89370 }, { "grad_norm": 0.3029172122478485, "learning_rate": 6.277131096383628e-05, "loss": 0.0376, "step": 89380 }, { "grad_norm": 0.49880099296569824, "learning_rate": 6.276331768015378e-05, "loss": 0.0321, "step": 89390 }, { "grad_norm": 0.5730984210968018, "learning_rate": 6.275532404752696e-05, "loss": 0.0469, "step": 89400 }, { "grad_norm": 0.9246512055397034, "learning_rate": 6.274733006617438e-05, "loss": 0.0323, "step": 89410 }, { "grad_norm": 0.9883178472518921, "learning_rate": 6.273933573631457e-05, "loss": 0.0402, "step": 89420 }, { "grad_norm": 0.5783231258392334, "learning_rate": 6.273134105816612e-05, "loss": 0.0258, "step": 89430 }, { "grad_norm": 0.4230211675167084, "learning_rate": 6.272334603194758e-05, "loss": 0.0288, "step": 89440 }, { "grad_norm": 0.5310205817222595, "learning_rate": 6.271535065787754e-05, "loss": 0.0317, "step": 89450 }, { "grad_norm": 0.48011520504951477, "learning_rate": 6.27073549361746e-05, "loss": 0.0405, "step": 89460 }, { "grad_norm": 0.8776212334632874, "learning_rate": 6.269935886705733e-05, "loss": 0.0487, "step": 89470 }, { "grad_norm": 0.38576871156692505, "learning_rate": 6.269136245074439e-05, "loss": 0.033, "step": 89480 }, { "grad_norm": 0.522709310054779, "learning_rate": 6.268336568745434e-05, "loss": 0.0336, "step": 89490 }, { "grad_norm": 0.3211763799190521, "learning_rate": 6.267536857740585e-05, "loss": 0.0592, "step": 89500 }, { "grad_norm": 0.5272477865219116, "learning_rate": 6.266737112081754e-05, "loss": 0.0402, "step": 89510 }, { "grad_norm": 0.5442400574684143, "learning_rate": 6.265937331790808e-05, "loss": 0.0371, "step": 89520 }, { "grad_norm": 1.5205917358398438, "learning_rate": 6.265137516889609e-05, "loss": 0.0418, "step": 89530 }, { "grad_norm": 0.5392431020736694, "learning_rate": 6.264337667400026e-05, "loss": 0.05, "step": 89540 }, { "grad_norm": 0.6845933794975281, "learning_rate": 6.263537783343927e-05, "loss": 0.0394, "step": 89550 }, { "grad_norm": 0.5034310221672058, "learning_rate": 6.262737864743179e-05, "loss": 0.0439, "step": 89560 }, { "grad_norm": 0.9540495276451111, "learning_rate": 6.26193791161965e-05, "loss": 0.0505, "step": 89570 }, { "grad_norm": 0.4657614231109619, "learning_rate": 6.261137923995216e-05, "loss": 0.0306, "step": 89580 }, { "grad_norm": 0.3850420415401459, "learning_rate": 6.260337901891743e-05, "loss": 0.0316, "step": 89590 }, { "grad_norm": 1.0187482833862305, "learning_rate": 6.259537845331107e-05, "loss": 0.0333, "step": 89600 }, { "grad_norm": 1.465152621269226, "learning_rate": 6.258737754335178e-05, "loss": 0.0372, "step": 89610 }, { "grad_norm": 1.0317268371582031, "learning_rate": 6.257937628925832e-05, "loss": 0.0439, "step": 89620 }, { "grad_norm": 0.38093146681785583, "learning_rate": 6.257137469124944e-05, "loss": 0.041, "step": 89630 }, { "grad_norm": 0.32416635751724243, "learning_rate": 6.25633727495439e-05, "loss": 0.0534, "step": 89640 }, { "grad_norm": 0.5166881680488586, "learning_rate": 6.255537046436047e-05, "loss": 0.0322, "step": 89650 }, { "grad_norm": 0.5483273863792419, "learning_rate": 6.254736783591793e-05, "loss": 0.0439, "step": 89660 }, { "grad_norm": 0.4464588761329651, "learning_rate": 6.253936486443504e-05, "loss": 0.0301, "step": 89670 }, { "grad_norm": 1.1359814405441284, "learning_rate": 6.253136155013066e-05, "loss": 0.0528, "step": 89680 }, { "grad_norm": 0.45461317896842957, "learning_rate": 6.252335789322355e-05, "loss": 0.0305, "step": 89690 }, { "grad_norm": 0.5407418608665466, "learning_rate": 6.251535389393254e-05, "loss": 0.0347, "step": 89700 }, { "grad_norm": 1.3269299268722534, "learning_rate": 6.250734955247644e-05, "loss": 0.0325, "step": 89710 }, { "grad_norm": 1.0725687742233276, "learning_rate": 6.249934486907413e-05, "loss": 0.0373, "step": 89720 }, { "grad_norm": 0.49355074763298035, "learning_rate": 6.249133984394442e-05, "loss": 0.0411, "step": 89730 }, { "grad_norm": 0.6079521179199219, "learning_rate": 6.248333447730617e-05, "loss": 0.0378, "step": 89740 }, { "grad_norm": 0.5981754064559937, "learning_rate": 6.247532876937825e-05, "loss": 0.0289, "step": 89750 }, { "grad_norm": 0.4912419021129608, "learning_rate": 6.24673227203795e-05, "loss": 0.0326, "step": 89760 }, { "grad_norm": 0.4976333975791931, "learning_rate": 6.245931633052885e-05, "loss": 0.0345, "step": 89770 }, { "grad_norm": 0.518481969833374, "learning_rate": 6.245130960004517e-05, "loss": 0.0392, "step": 89780 }, { "grad_norm": 0.4135996699333191, "learning_rate": 6.244330252914738e-05, "loss": 0.027, "step": 89790 }, { "grad_norm": 0.5333378911018372, "learning_rate": 6.243529511805435e-05, "loss": 0.0312, "step": 89800 }, { "grad_norm": 0.4676593840122223, "learning_rate": 6.242728736698504e-05, "loss": 0.0278, "step": 89810 }, { "grad_norm": 0.6624797582626343, "learning_rate": 6.241927927615836e-05, "loss": 0.0318, "step": 89820 }, { "grad_norm": 0.5807161331176758, "learning_rate": 6.241127084579323e-05, "loss": 0.0448, "step": 89830 }, { "grad_norm": 0.8341360092163086, "learning_rate": 6.240326207610864e-05, "loss": 0.058, "step": 89840 }, { "grad_norm": 0.5104627013206482, "learning_rate": 6.239525296732352e-05, "loss": 0.0469, "step": 89850 }, { "grad_norm": 0.32940614223480225, "learning_rate": 6.238724351965683e-05, "loss": 0.029, "step": 89860 }, { "grad_norm": 0.39433056116104126, "learning_rate": 6.237923373332758e-05, "loss": 0.0371, "step": 89870 }, { "grad_norm": 0.3931412994861603, "learning_rate": 6.237122360855472e-05, "loss": 0.0317, "step": 89880 }, { "grad_norm": 0.72956782579422, "learning_rate": 6.236321314555725e-05, "loss": 0.0345, "step": 89890 }, { "grad_norm": 0.4595405161380768, "learning_rate": 6.235520234455417e-05, "loss": 0.0268, "step": 89900 }, { "grad_norm": 0.48191019892692566, "learning_rate": 6.234719120576451e-05, "loss": 0.0387, "step": 89910 }, { "grad_norm": 0.708411693572998, "learning_rate": 6.233917972940727e-05, "loss": 0.0396, "step": 89920 }, { "grad_norm": 0.3829731047153473, "learning_rate": 6.23311679157015e-05, "loss": 0.0359, "step": 89930 }, { "grad_norm": 0.6112335324287415, "learning_rate": 6.232315576486622e-05, "loss": 0.055, "step": 89940 }, { "grad_norm": 0.818328857421875, "learning_rate": 6.231514327712052e-05, "loss": 0.036, "step": 89950 }, { "grad_norm": 0.36645665764808655, "learning_rate": 6.230713045268339e-05, "loss": 0.0476, "step": 89960 }, { "grad_norm": 0.5875741243362427, "learning_rate": 6.229911729177396e-05, "loss": 0.0274, "step": 89970 }, { "grad_norm": 0.45601141452789307, "learning_rate": 6.22911037946113e-05, "loss": 0.0322, "step": 89980 }, { "grad_norm": 0.3033609390258789, "learning_rate": 6.228308996141446e-05, "loss": 0.0296, "step": 89990 }, { "grad_norm": 0.4632725119590759, "learning_rate": 6.227507579240254e-05, "loss": 0.048, "step": 90000 }, { "grad_norm": 0.6112654805183411, "learning_rate": 6.226706128779468e-05, "loss": 0.0315, "step": 90010 }, { "grad_norm": 0.34575021266937256, "learning_rate": 6.225904644780997e-05, "loss": 0.0294, "step": 90020 }, { "grad_norm": 0.3487606942653656, "learning_rate": 6.225103127266753e-05, "loss": 0.0415, "step": 90030 }, { "grad_norm": 0.4227263927459717, "learning_rate": 6.224301576258651e-05, "loss": 0.0393, "step": 90040 }, { "grad_norm": 0.36975380778312683, "learning_rate": 6.223499991778604e-05, "loss": 0.0349, "step": 90050 }, { "grad_norm": 0.6688131093978882, "learning_rate": 6.222698373848525e-05, "loss": 0.031, "step": 90060 }, { "grad_norm": 0.3560798466205597, "learning_rate": 6.221896722490333e-05, "loss": 0.0374, "step": 90070 }, { "grad_norm": 0.7038931846618652, "learning_rate": 6.221095037725942e-05, "loss": 0.0292, "step": 90080 }, { "grad_norm": 0.43518704175949097, "learning_rate": 6.220293319577273e-05, "loss": 0.0335, "step": 90090 }, { "grad_norm": 1.135730504989624, "learning_rate": 6.219491568066243e-05, "loss": 0.0353, "step": 90100 }, { "grad_norm": 0.5343167185783386, "learning_rate": 6.218689783214772e-05, "loss": 0.0405, "step": 90110 }, { "grad_norm": 0.38975393772125244, "learning_rate": 6.21788796504478e-05, "loss": 0.0438, "step": 90120 }, { "grad_norm": 0.5439395904541016, "learning_rate": 6.217086113578187e-05, "loss": 0.0279, "step": 90130 }, { "grad_norm": 0.4537172019481659, "learning_rate": 6.21628422883692e-05, "loss": 0.0294, "step": 90140 }, { "grad_norm": 0.42764660716056824, "learning_rate": 6.215482310842896e-05, "loss": 0.0323, "step": 90150 }, { "grad_norm": 0.7398531436920166, "learning_rate": 6.214680359618044e-05, "loss": 0.0392, "step": 90160 }, { "grad_norm": 0.6860285997390747, "learning_rate": 6.213878375184287e-05, "loss": 0.0367, "step": 90170 }, { "grad_norm": 0.4080336093902588, "learning_rate": 6.21307635756355e-05, "loss": 0.0326, "step": 90180 }, { "grad_norm": 0.5283743143081665, "learning_rate": 6.212274306777764e-05, "loss": 0.0366, "step": 90190 }, { "grad_norm": 0.47745582461357117, "learning_rate": 6.21147222284885e-05, "loss": 0.0269, "step": 90200 }, { "grad_norm": 0.7548861503601074, "learning_rate": 6.210670105798744e-05, "loss": 0.0326, "step": 90210 }, { "grad_norm": 0.9491684436798096, "learning_rate": 6.20986795564937e-05, "loss": 0.033, "step": 90220 }, { "grad_norm": 0.4958583116531372, "learning_rate": 6.209065772422661e-05, "loss": 0.0336, "step": 90230 }, { "grad_norm": 3.1738641262054443, "learning_rate": 6.208263556140549e-05, "loss": 0.0272, "step": 90240 }, { "grad_norm": 0.37580034136772156, "learning_rate": 6.207461306824965e-05, "loss": 0.0323, "step": 90250 }, { "grad_norm": 0.4258180260658264, "learning_rate": 6.206659024497843e-05, "loss": 0.0403, "step": 90260 }, { "grad_norm": 0.5370066165924072, "learning_rate": 6.205856709181115e-05, "loss": 0.0352, "step": 90270 }, { "grad_norm": 0.28379425406455994, "learning_rate": 6.205054360896718e-05, "loss": 0.026, "step": 90280 }, { "grad_norm": 0.5622945427894592, "learning_rate": 6.204251979666589e-05, "loss": 0.0325, "step": 90290 }, { "grad_norm": 0.5404784679412842, "learning_rate": 6.203449565512661e-05, "loss": 0.0281, "step": 90300 }, { "grad_norm": 0.30348995327949524, "learning_rate": 6.202647118456877e-05, "loss": 0.0337, "step": 90310 }, { "grad_norm": 0.4094267785549164, "learning_rate": 6.20184463852117e-05, "loss": 0.0557, "step": 90320 }, { "grad_norm": 0.38721466064453125, "learning_rate": 6.201042125727483e-05, "loss": 0.0316, "step": 90330 }, { "grad_norm": 0.4229050278663635, "learning_rate": 6.200239580097754e-05, "loss": 0.0325, "step": 90340 }, { "grad_norm": 0.7621817588806152, "learning_rate": 6.199437001653929e-05, "loss": 0.0352, "step": 90350 }, { "grad_norm": 0.6578339338302612, "learning_rate": 6.198634390417944e-05, "loss": 0.032, "step": 90360 }, { "grad_norm": 1.358498215675354, "learning_rate": 6.197831746411748e-05, "loss": 0.0446, "step": 90370 }, { "grad_norm": 0.5774692296981812, "learning_rate": 6.197029069657279e-05, "loss": 0.0399, "step": 90380 }, { "grad_norm": 0.9184568524360657, "learning_rate": 6.196226360176486e-05, "loss": 0.0314, "step": 90390 }, { "grad_norm": 0.4069538414478302, "learning_rate": 6.195423617991313e-05, "loss": 0.0273, "step": 90400 }, { "grad_norm": 0.5194627642631531, "learning_rate": 6.194620843123709e-05, "loss": 0.0319, "step": 90410 }, { "grad_norm": 0.2877979576587677, "learning_rate": 6.19381803559562e-05, "loss": 0.0315, "step": 90420 }, { "grad_norm": 1.6341372728347778, "learning_rate": 6.193015195428993e-05, "loss": 0.0314, "step": 90430 }, { "grad_norm": 0.6718907952308655, "learning_rate": 6.19221232264578e-05, "loss": 0.0311, "step": 90440 }, { "grad_norm": 0.723222553730011, "learning_rate": 6.191409417267929e-05, "loss": 0.0404, "step": 90450 }, { "grad_norm": 0.33889997005462646, "learning_rate": 6.190606479317392e-05, "loss": 0.0337, "step": 90460 }, { "grad_norm": 0.48439717292785645, "learning_rate": 6.189803508816121e-05, "loss": 0.039, "step": 90470 }, { "grad_norm": 0.5931333303451538, "learning_rate": 6.189000505786071e-05, "loss": 0.044, "step": 90480 }, { "grad_norm": 0.5536326766014099, "learning_rate": 6.188197470249192e-05, "loss": 0.0355, "step": 90490 }, { "grad_norm": 0.3428066074848175, "learning_rate": 6.187394402227442e-05, "loss": 0.0275, "step": 90500 }, { "grad_norm": 0.715519368648529, "learning_rate": 6.186591301742774e-05, "loss": 0.0225, "step": 90510 }, { "grad_norm": 0.5404621958732605, "learning_rate": 6.185788168817146e-05, "loss": 0.0341, "step": 90520 }, { "grad_norm": 0.5696438550949097, "learning_rate": 6.184985003472513e-05, "loss": 0.0336, "step": 90530 }, { "grad_norm": 0.8302581310272217, "learning_rate": 6.184181805730838e-05, "loss": 0.0269, "step": 90540 }, { "grad_norm": 0.539604127407074, "learning_rate": 6.183378575614077e-05, "loss": 0.0295, "step": 90550 }, { "grad_norm": 0.5924202799797058, "learning_rate": 6.18257531314419e-05, "loss": 0.0352, "step": 90560 }, { "grad_norm": 0.4030943810939789, "learning_rate": 6.181772018343137e-05, "loss": 0.0282, "step": 90570 }, { "grad_norm": 1.0398503541946411, "learning_rate": 6.180968691232882e-05, "loss": 0.0421, "step": 90580 }, { "grad_norm": 0.4013279378414154, "learning_rate": 6.180165331835386e-05, "loss": 0.0528, "step": 90590 }, { "grad_norm": 0.6539302468299866, "learning_rate": 6.179361940172614e-05, "loss": 0.0366, "step": 90600 }, { "grad_norm": 0.765529990196228, "learning_rate": 6.178558516266529e-05, "loss": 0.0304, "step": 90610 }, { "grad_norm": 0.3693026900291443, "learning_rate": 6.177755060139098e-05, "loss": 0.0556, "step": 90620 }, { "grad_norm": 1.0352705717086792, "learning_rate": 6.176951571812285e-05, "loss": 0.039, "step": 90630 }, { "grad_norm": 0.4107111096382141, "learning_rate": 6.176148051308059e-05, "loss": 0.0289, "step": 90640 }, { "grad_norm": 0.5593661665916443, "learning_rate": 6.175344498648385e-05, "loss": 0.0342, "step": 90650 }, { "grad_norm": 0.5612488985061646, "learning_rate": 6.174540913855237e-05, "loss": 0.0246, "step": 90660 }, { "grad_norm": 0.43026575446128845, "learning_rate": 6.173737296950579e-05, "loss": 0.0332, "step": 90670 }, { "grad_norm": 0.43902093172073364, "learning_rate": 6.172933647956386e-05, "loss": 0.0326, "step": 90680 }, { "grad_norm": 0.37680694460868835, "learning_rate": 6.172129966894627e-05, "loss": 0.0356, "step": 90690 }, { "grad_norm": 0.4854342043399811, "learning_rate": 6.171326253787275e-05, "loss": 0.0314, "step": 90700 }, { "grad_norm": 0.74051433801651, "learning_rate": 6.170522508656304e-05, "loss": 0.0511, "step": 90710 }, { "grad_norm": 0.8124783039093018, "learning_rate": 6.169718731523686e-05, "loss": 0.0463, "step": 90720 }, { "grad_norm": 0.3900129199028015, "learning_rate": 6.1689149224114e-05, "loss": 0.0258, "step": 90730 }, { "grad_norm": 0.5628536939620972, "learning_rate": 6.168111081341417e-05, "loss": 0.0334, "step": 90740 }, { "grad_norm": 0.4941759705543518, "learning_rate": 6.167307208335716e-05, "loss": 0.0319, "step": 90750 }, { "grad_norm": 0.43123894929885864, "learning_rate": 6.166503303416274e-05, "loss": 0.0244, "step": 90760 }, { "grad_norm": 1.0486083030700684, "learning_rate": 6.165699366605072e-05, "loss": 0.0312, "step": 90770 }, { "grad_norm": 0.2881828546524048, "learning_rate": 6.164895397924085e-05, "loss": 0.0336, "step": 90780 }, { "grad_norm": 0.3475782573223114, "learning_rate": 6.164091397395296e-05, "loss": 0.0405, "step": 90790 }, { "grad_norm": 0.6767386198043823, "learning_rate": 6.163287365040687e-05, "loss": 0.0272, "step": 90800 }, { "grad_norm": 0.3609964847564697, "learning_rate": 6.162483300882237e-05, "loss": 0.0235, "step": 90810 }, { "grad_norm": 1.3196614980697632, "learning_rate": 6.16167920494193e-05, "loss": 0.0457, "step": 90820 }, { "grad_norm": 0.34466058015823364, "learning_rate": 6.160875077241753e-05, "loss": 0.0282, "step": 90830 }, { "grad_norm": 0.44132423400878906, "learning_rate": 6.160070917803684e-05, "loss": 0.023, "step": 90840 }, { "grad_norm": 0.5327534079551697, "learning_rate": 6.159266726649713e-05, "loss": 0.0276, "step": 90850 }, { "grad_norm": 1.10667085647583, "learning_rate": 6.158462503801824e-05, "loss": 0.0425, "step": 90860 }, { "grad_norm": 1.2645924091339111, "learning_rate": 6.157658249282007e-05, "loss": 0.0389, "step": 90870 }, { "grad_norm": 0.8745163679122925, "learning_rate": 6.156853963112248e-05, "loss": 0.0423, "step": 90880 }, { "grad_norm": 0.7665811777114868, "learning_rate": 6.156049645314537e-05, "loss": 0.0476, "step": 90890 }, { "grad_norm": 0.48141175508499146, "learning_rate": 6.155245295910863e-05, "loss": 0.0396, "step": 90900 }, { "grad_norm": 0.45311856269836426, "learning_rate": 6.154440914923215e-05, "loss": 0.0431, "step": 90910 }, { "grad_norm": 0.45199310779571533, "learning_rate": 6.153636502373586e-05, "loss": 0.0437, "step": 90920 }, { "grad_norm": 0.9153753519058228, "learning_rate": 6.152832058283969e-05, "loss": 0.0304, "step": 90930 }, { "grad_norm": 0.4610769748687744, "learning_rate": 6.152027582676359e-05, "loss": 0.0366, "step": 90940 }, { "grad_norm": 1.0144236087799072, "learning_rate": 6.151223075572744e-05, "loss": 0.0409, "step": 90950 }, { "grad_norm": 0.48302462697029114, "learning_rate": 6.150418536995125e-05, "loss": 0.0296, "step": 90960 }, { "grad_norm": 0.5468060374259949, "learning_rate": 6.149613966965493e-05, "loss": 0.0507, "step": 90970 }, { "grad_norm": 0.5792583227157593, "learning_rate": 6.14880936550585e-05, "loss": 0.0229, "step": 90980 }, { "grad_norm": 0.5115638375282288, "learning_rate": 6.148004732638188e-05, "loss": 0.0327, "step": 90990 }, { "grad_norm": 0.6889946460723877, "learning_rate": 6.147200068384511e-05, "loss": 0.038, "step": 91000 }, { "grad_norm": 0.5558481812477112, "learning_rate": 6.146395372766812e-05, "loss": 0.0369, "step": 91010 }, { "grad_norm": 0.7582060098648071, "learning_rate": 6.145590645807097e-05, "loss": 0.0339, "step": 91020 }, { "grad_norm": 0.4770192503929138, "learning_rate": 6.144785887527362e-05, "loss": 0.0313, "step": 91030 }, { "grad_norm": 0.4172283113002777, "learning_rate": 6.143981097949611e-05, "loss": 0.0358, "step": 91040 }, { "grad_norm": 0.5439754724502563, "learning_rate": 6.143176277095847e-05, "loss": 0.0255, "step": 91050 }, { "grad_norm": 0.26659756898880005, "learning_rate": 6.142371424988074e-05, "loss": 0.0279, "step": 91060 }, { "grad_norm": 0.6154503226280212, "learning_rate": 6.141566541648294e-05, "loss": 0.0318, "step": 91070 }, { "grad_norm": 0.5247637033462524, "learning_rate": 6.140761627098515e-05, "loss": 0.0301, "step": 91080 }, { "grad_norm": 0.4914216101169586, "learning_rate": 6.13995668136074e-05, "loss": 0.0438, "step": 91090 }, { "grad_norm": 1.3177376985549927, "learning_rate": 6.139151704456978e-05, "loss": 0.0425, "step": 91100 }, { "grad_norm": 1.2152421474456787, "learning_rate": 6.138346696409236e-05, "loss": 0.0415, "step": 91110 }, { "grad_norm": 0.7091630697250366, "learning_rate": 6.137541657239525e-05, "loss": 0.034, "step": 91120 }, { "grad_norm": 0.7263021469116211, "learning_rate": 6.136736586969851e-05, "loss": 0.0389, "step": 91130 }, { "grad_norm": 0.7260653376579285, "learning_rate": 6.135931485622225e-05, "loss": 0.0324, "step": 91140 }, { "grad_norm": 0.36137646436691284, "learning_rate": 6.135126353218661e-05, "loss": 0.0358, "step": 91150 }, { "grad_norm": 0.5541040301322937, "learning_rate": 6.134321189781166e-05, "loss": 0.0351, "step": 91160 }, { "grad_norm": 1.4251445531845093, "learning_rate": 6.13351599533176e-05, "loss": 0.0425, "step": 91170 }, { "grad_norm": 0.494143545627594, "learning_rate": 6.132710769892449e-05, "loss": 0.0284, "step": 91180 }, { "grad_norm": 0.4907046854496002, "learning_rate": 6.131905513485252e-05, "loss": 0.0363, "step": 91190 }, { "grad_norm": 0.5929480791091919, "learning_rate": 6.131100226132184e-05, "loss": 0.0483, "step": 91200 }, { "grad_norm": 0.8684423565864563, "learning_rate": 6.130294907855261e-05, "loss": 0.034, "step": 91210 }, { "grad_norm": 0.6036779880523682, "learning_rate": 6.129489558676499e-05, "loss": 0.0331, "step": 91220 }, { "grad_norm": 1.411902904510498, "learning_rate": 6.128684178617917e-05, "loss": 0.0415, "step": 91230 }, { "grad_norm": 0.708432674407959, "learning_rate": 6.127878767701533e-05, "loss": 0.0512, "step": 91240 }, { "grad_norm": 0.5944589972496033, "learning_rate": 6.127073325949368e-05, "loss": 0.0347, "step": 91250 }, { "grad_norm": 0.4932643175125122, "learning_rate": 6.126267853383442e-05, "loss": 0.0362, "step": 91260 }, { "grad_norm": 0.8203845620155334, "learning_rate": 6.125462350025776e-05, "loss": 0.0342, "step": 91270 }, { "grad_norm": 0.47055891156196594, "learning_rate": 6.124656815898392e-05, "loss": 0.0314, "step": 91280 }, { "grad_norm": 0.5345668196678162, "learning_rate": 6.123851251023313e-05, "loss": 0.0284, "step": 91290 }, { "grad_norm": 0.3557053804397583, "learning_rate": 6.123045655422563e-05, "loss": 0.0412, "step": 91300 }, { "grad_norm": 0.48372289538383484, "learning_rate": 6.122240029118168e-05, "loss": 0.0304, "step": 91310 }, { "grad_norm": 0.6168065071105957, "learning_rate": 6.121434372132149e-05, "loss": 0.0256, "step": 91320 }, { "grad_norm": 1.8189914226531982, "learning_rate": 6.120628684486538e-05, "loss": 0.0312, "step": 91330 }, { "grad_norm": 0.8680362105369568, "learning_rate": 6.119822966203359e-05, "loss": 0.0327, "step": 91340 }, { "grad_norm": 0.8466756939888, "learning_rate": 6.119017217304641e-05, "loss": 0.0413, "step": 91350 }, { "grad_norm": 0.34724026918411255, "learning_rate": 6.118211437812412e-05, "loss": 0.0377, "step": 91360 }, { "grad_norm": 0.48732420802116394, "learning_rate": 6.117405627748703e-05, "loss": 0.031, "step": 91370 }, { "grad_norm": 0.6091507077217102, "learning_rate": 6.116599787135545e-05, "loss": 0.0358, "step": 91380 }, { "grad_norm": 1.9550726413726807, "learning_rate": 6.115793915994968e-05, "loss": 0.0425, "step": 91390 }, { "grad_norm": 0.40587034821510315, "learning_rate": 6.114988014349004e-05, "loss": 0.0315, "step": 91400 }, { "grad_norm": 0.6484859585762024, "learning_rate": 6.114182082219686e-05, "loss": 0.0291, "step": 91410 }, { "grad_norm": 0.3189232051372528, "learning_rate": 6.113376119629049e-05, "loss": 0.0373, "step": 91420 }, { "grad_norm": 0.6011351943016052, "learning_rate": 6.112570126599125e-05, "loss": 0.041, "step": 91430 }, { "grad_norm": 0.728020191192627, "learning_rate": 6.111764103151955e-05, "loss": 0.0379, "step": 91440 }, { "grad_norm": 0.6397368907928467, "learning_rate": 6.11095804930957e-05, "loss": 0.0292, "step": 91450 }, { "grad_norm": 0.5480132699012756, "learning_rate": 6.110151965094011e-05, "loss": 0.0286, "step": 91460 }, { "grad_norm": 0.6756817698478699, "learning_rate": 6.109345850527314e-05, "loss": 0.0395, "step": 91470 }, { "grad_norm": 0.48283207416534424, "learning_rate": 6.108539705631517e-05, "loss": 0.0281, "step": 91480 }, { "grad_norm": 1.2573479413986206, "learning_rate": 6.107733530428661e-05, "loss": 0.0432, "step": 91490 }, { "grad_norm": 0.521149754524231, "learning_rate": 6.106927324940788e-05, "loss": 0.0346, "step": 91500 }, { "grad_norm": 0.5433168411254883, "learning_rate": 6.106121089189937e-05, "loss": 0.049, "step": 91510 }, { "grad_norm": 0.5478119254112244, "learning_rate": 6.10531482319815e-05, "loss": 0.0385, "step": 91520 }, { "grad_norm": 0.8027316927909851, "learning_rate": 6.104508526987472e-05, "loss": 0.0327, "step": 91530 }, { "grad_norm": 0.5544983148574829, "learning_rate": 6.103702200579946e-05, "loss": 0.0382, "step": 91540 }, { "grad_norm": 0.498522013425827, "learning_rate": 6.1028958439976146e-05, "loss": 0.0291, "step": 91550 }, { "grad_norm": 0.4846738576889038, "learning_rate": 6.1020894572625275e-05, "loss": 0.0336, "step": 91560 }, { "grad_norm": 0.5218921899795532, "learning_rate": 6.101283040396728e-05, "loss": 0.043, "step": 91570 }, { "grad_norm": 1.0029151439666748, "learning_rate": 6.100476593422263e-05, "loss": 0.048, "step": 91580 }, { "grad_norm": 0.5011561512947083, "learning_rate": 6.099670116361182e-05, "loss": 0.0323, "step": 91590 }, { "grad_norm": 0.2839893698692322, "learning_rate": 6.0988636092355335e-05, "loss": 0.0332, "step": 91600 }, { "grad_norm": 0.5890867710113525, "learning_rate": 6.098057072067367e-05, "loss": 0.0364, "step": 91610 }, { "grad_norm": 0.8951244354248047, "learning_rate": 6.097250504878731e-05, "loss": 0.0376, "step": 91620 }, { "grad_norm": 0.5237593650817871, "learning_rate": 6.09644390769168e-05, "loss": 0.0373, "step": 91630 }, { "grad_norm": 0.49274730682373047, "learning_rate": 6.0956372805282634e-05, "loss": 0.0296, "step": 91640 }, { "grad_norm": 0.5798510909080505, "learning_rate": 6.094830623410537e-05, "loss": 0.0568, "step": 91650 }, { "grad_norm": 0.38340020179748535, "learning_rate": 6.094023936360551e-05, "loss": 0.0374, "step": 91660 }, { "grad_norm": 0.9708678126335144, "learning_rate": 6.093217219400362e-05, "loss": 0.0406, "step": 91670 }, { "grad_norm": 0.41863158345222473, "learning_rate": 6.092410472552025e-05, "loss": 0.0554, "step": 91680 }, { "grad_norm": 0.8731489777565002, "learning_rate": 6.0916036958375966e-05, "loss": 0.0317, "step": 91690 }, { "grad_norm": 0.5204538106918335, "learning_rate": 6.090796889279132e-05, "loss": 0.056, "step": 91700 }, { "grad_norm": 0.6837800741195679, "learning_rate": 6.089990052898692e-05, "loss": 0.0431, "step": 91710 }, { "grad_norm": 0.3805252015590668, "learning_rate": 6.0891831867183316e-05, "loss": 0.0285, "step": 91720 }, { "grad_norm": 0.8694709539413452, "learning_rate": 6.088376290760114e-05, "loss": 0.0381, "step": 91730 }, { "grad_norm": 0.6873107552528381, "learning_rate": 6.087569365046095e-05, "loss": 0.0352, "step": 91740 }, { "grad_norm": 1.2898846864700317, "learning_rate": 6.08676240959834e-05, "loss": 0.0483, "step": 91750 }, { "grad_norm": 0.36539706587791443, "learning_rate": 6.085955424438908e-05, "loss": 0.0334, "step": 91760 }, { "grad_norm": 0.685393750667572, "learning_rate": 6.085148409589864e-05, "loss": 0.0474, "step": 91770 }, { "grad_norm": 0.3729842007160187, "learning_rate": 6.0843413650732685e-05, "loss": 0.0429, "step": 91780 }, { "grad_norm": 0.613339364528656, "learning_rate": 6.083534290911188e-05, "loss": 0.039, "step": 91790 }, { "grad_norm": 1.2623984813690186, "learning_rate": 6.082727187125688e-05, "loss": 0.0351, "step": 91800 }, { "grad_norm": 0.4317694902420044, "learning_rate": 6.081920053738832e-05, "loss": 0.0355, "step": 91810 }, { "grad_norm": 1.0432451963424683, "learning_rate": 6.081112890772689e-05, "loss": 0.0363, "step": 91820 }, { "grad_norm": 0.693260908126831, "learning_rate": 6.0803056982493255e-05, "loss": 0.0343, "step": 91830 }, { "grad_norm": 0.5184222459793091, "learning_rate": 6.0794984761908114e-05, "loss": 0.0269, "step": 91840 }, { "grad_norm": 0.5299289226531982, "learning_rate": 6.078691224619213e-05, "loss": 0.0429, "step": 91850 }, { "grad_norm": 0.4414665400981903, "learning_rate": 6.0778839435566014e-05, "loss": 0.0277, "step": 91860 }, { "grad_norm": 0.40544846653938293, "learning_rate": 6.077076633025049e-05, "loss": 0.0305, "step": 91870 }, { "grad_norm": 1.2486315965652466, "learning_rate": 6.076269293046626e-05, "loss": 0.0274, "step": 91880 }, { "grad_norm": 0.6941802501678467, "learning_rate": 6.075461923643403e-05, "loss": 0.0328, "step": 91890 }, { "grad_norm": 0.3018885850906372, "learning_rate": 6.074654524837458e-05, "loss": 0.0288, "step": 91900 }, { "grad_norm": 0.4950471520423889, "learning_rate": 6.07384709665086e-05, "loss": 0.0359, "step": 91910 }, { "grad_norm": 0.5128141641616821, "learning_rate": 6.073039639105686e-05, "loss": 0.0202, "step": 91920 }, { "grad_norm": 0.376715749502182, "learning_rate": 6.072232152224011e-05, "loss": 0.0298, "step": 91930 }, { "grad_norm": 0.4297865927219391, "learning_rate": 6.071424636027912e-05, "loss": 0.0265, "step": 91940 }, { "grad_norm": 0.4125898778438568, "learning_rate": 6.0706170905394665e-05, "loss": 0.0399, "step": 91950 }, { "grad_norm": 0.4553672671318054, "learning_rate": 6.069809515780753e-05, "loss": 0.0324, "step": 91960 }, { "grad_norm": 0.4506685733795166, "learning_rate": 6.069001911773847e-05, "loss": 0.0644, "step": 91970 }, { "grad_norm": 0.4686774015426636, "learning_rate": 6.068194278540833e-05, "loss": 0.037, "step": 91980 }, { "grad_norm": 0.49894484877586365, "learning_rate": 6.067386616103785e-05, "loss": 0.0355, "step": 91990 }, { "grad_norm": 0.49187299609184265, "learning_rate": 6.066578924484791e-05, "loss": 0.0418, "step": 92000 }, { "grad_norm": 0.7067137360572815, "learning_rate": 6.065771203705929e-05, "loss": 0.0399, "step": 92010 }, { "grad_norm": 1.0377413034439087, "learning_rate": 6.064963453789283e-05, "loss": 0.0288, "step": 92020 }, { "grad_norm": 0.4189603328704834, "learning_rate": 6.064155674756936e-05, "loss": 0.0313, "step": 92030 }, { "grad_norm": 0.5583803057670593, "learning_rate": 6.0633478666309714e-05, "loss": 0.0374, "step": 92040 }, { "grad_norm": 0.6291442513465881, "learning_rate": 6.0625400294334765e-05, "loss": 0.0248, "step": 92050 }, { "grad_norm": 0.6158702373504639, "learning_rate": 6.061732163186537e-05, "loss": 0.0464, "step": 92060 }, { "grad_norm": 1.0664819478988647, "learning_rate": 6.060924267912239e-05, "loss": 0.0442, "step": 92070 }, { "grad_norm": 2.316025495529175, "learning_rate": 6.060116343632669e-05, "loss": 0.0386, "step": 92080 }, { "grad_norm": 0.514423131942749, "learning_rate": 6.059308390369919e-05, "loss": 0.0265, "step": 92090 }, { "grad_norm": 0.45842665433883667, "learning_rate": 6.0585004081460725e-05, "loss": 0.0309, "step": 92100 }, { "grad_norm": 0.2775304615497589, "learning_rate": 6.057692396983224e-05, "loss": 0.0452, "step": 92110 }, { "grad_norm": 0.6695839762687683, "learning_rate": 6.056884356903462e-05, "loss": 0.033, "step": 92120 }, { "grad_norm": 0.28899604082107544, "learning_rate": 6.05607628792888e-05, "loss": 0.0266, "step": 92130 }, { "grad_norm": 0.728162407875061, "learning_rate": 6.055268190081568e-05, "loss": 0.0588, "step": 92140 }, { "grad_norm": 1.2528389692306519, "learning_rate": 6.0544600633836214e-05, "loss": 0.0472, "step": 92150 }, { "grad_norm": 0.36532407999038696, "learning_rate": 6.053651907857131e-05, "loss": 0.0247, "step": 92160 }, { "grad_norm": 0.49348679184913635, "learning_rate": 6.052843723524195e-05, "loss": 0.0319, "step": 92170 }, { "grad_norm": 0.4683084785938263, "learning_rate": 6.052035510406905e-05, "loss": 0.0518, "step": 92180 }, { "grad_norm": 0.4025871455669403, "learning_rate": 6.0512272685273605e-05, "loss": 0.0374, "step": 92190 }, { "grad_norm": 0.5622091889381409, "learning_rate": 6.0504189979076574e-05, "loss": 0.0297, "step": 92200 }, { "grad_norm": 0.35138973593711853, "learning_rate": 6.0496106985698944e-05, "loss": 0.033, "step": 92210 }, { "grad_norm": 0.47347912192344666, "learning_rate": 6.048802370536168e-05, "loss": 0.0607, "step": 92220 }, { "grad_norm": 1.9569759368896484, "learning_rate": 6.04799401382858e-05, "loss": 0.0298, "step": 92230 }, { "grad_norm": 0.6675440669059753, "learning_rate": 6.0471856284692286e-05, "loss": 0.0271, "step": 92240 }, { "grad_norm": 0.6405589580535889, "learning_rate": 6.0463772144802145e-05, "loss": 0.0325, "step": 92250 }, { "grad_norm": 0.5027788877487183, "learning_rate": 6.045568771883642e-05, "loss": 0.0324, "step": 92260 }, { "grad_norm": 0.38909289240837097, "learning_rate": 6.044760300701612e-05, "loss": 0.0307, "step": 92270 }, { "grad_norm": 0.735790491104126, "learning_rate": 6.043951800956227e-05, "loss": 0.0293, "step": 92280 }, { "grad_norm": 0.7723713517189026, "learning_rate": 6.043143272669591e-05, "loss": 0.0262, "step": 92290 }, { "grad_norm": 0.31460875272750854, "learning_rate": 6.0423347158638134e-05, "loss": 0.0275, "step": 92300 }, { "grad_norm": 0.36700695753097534, "learning_rate": 6.0415261305609925e-05, "loss": 0.0224, "step": 92310 }, { "grad_norm": 0.9302812218666077, "learning_rate": 6.040717516783241e-05, "loss": 0.0322, "step": 92320 }, { "grad_norm": 0.5384142398834229, "learning_rate": 6.039908874552661e-05, "loss": 0.0281, "step": 92330 }, { "grad_norm": 0.2834148108959198, "learning_rate": 6.039100203891366e-05, "loss": 0.0498, "step": 92340 }, { "grad_norm": 0.7006524205207825, "learning_rate": 6.038291504821461e-05, "loss": 0.0311, "step": 92350 }, { "grad_norm": 0.7633166313171387, "learning_rate": 6.037482777365056e-05, "loss": 0.0465, "step": 92360 }, { "grad_norm": 0.6777400970458984, "learning_rate": 6.0366740215442595e-05, "loss": 0.0299, "step": 92370 }, { "grad_norm": 0.5296147465705872, "learning_rate": 6.035865237381187e-05, "loss": 0.0287, "step": 92380 }, { "grad_norm": 0.4989129900932312, "learning_rate": 6.035056424897948e-05, "loss": 0.0262, "step": 92390 }, { "grad_norm": 0.647163987159729, "learning_rate": 6.034247584116656e-05, "loss": 0.0392, "step": 92400 }, { "grad_norm": 0.6150621175765991, "learning_rate": 6.033438715059422e-05, "loss": 0.0382, "step": 92410 }, { "grad_norm": 0.5852656960487366, "learning_rate": 6.032629817748363e-05, "loss": 0.0366, "step": 92420 }, { "grad_norm": 0.5620895028114319, "learning_rate": 6.0318208922055905e-05, "loss": 0.0361, "step": 92430 }, { "grad_norm": 0.662101149559021, "learning_rate": 6.031011938453224e-05, "loss": 0.036, "step": 92440 }, { "grad_norm": 0.4637710452079773, "learning_rate": 6.0302029565133775e-05, "loss": 0.0304, "step": 92450 }, { "grad_norm": 0.24932245910167694, "learning_rate": 6.029393946408172e-05, "loss": 0.0359, "step": 92460 }, { "grad_norm": 1.0199936628341675, "learning_rate": 6.02858490815972e-05, "loss": 0.0329, "step": 92470 }, { "grad_norm": 0.45294782519340515, "learning_rate": 6.0277758417901444e-05, "loss": 0.0367, "step": 92480 }, { "grad_norm": 0.46778443455696106, "learning_rate": 6.026966747321563e-05, "loss": 0.0235, "step": 92490 }, { "grad_norm": 0.35728880763053894, "learning_rate": 6.026157624776095e-05, "loss": 0.0342, "step": 92500 }, { "grad_norm": 0.433438241481781, "learning_rate": 6.025348474175865e-05, "loss": 0.0404, "step": 92510 }, { "grad_norm": 3.030949115753174, "learning_rate": 6.0245392955429924e-05, "loss": 0.0426, "step": 92520 }, { "grad_norm": 0.3395324945449829, "learning_rate": 6.023730088899601e-05, "loss": 0.0335, "step": 92530 }, { "grad_norm": 0.5503639578819275, "learning_rate": 6.022920854267812e-05, "loss": 0.0269, "step": 92540 }, { "grad_norm": 0.7464969754219055, "learning_rate": 6.0221115916697524e-05, "loss": 0.0351, "step": 92550 }, { "grad_norm": 0.752062201499939, "learning_rate": 6.021302301127545e-05, "loss": 0.03, "step": 92560 }, { "grad_norm": 0.9857586622238159, "learning_rate": 6.020492982663317e-05, "loss": 0.027, "step": 92570 }, { "grad_norm": 0.6398594975471497, "learning_rate": 6.0196836362991935e-05, "loss": 0.0301, "step": 92580 }, { "grad_norm": 0.6620666980743408, "learning_rate": 6.0188742620573034e-05, "loss": 0.0369, "step": 92590 }, { "grad_norm": 0.8422115445137024, "learning_rate": 6.018064859959773e-05, "loss": 0.0371, "step": 92600 }, { "grad_norm": 1.6175670623779297, "learning_rate": 6.017255430028733e-05, "loss": 0.0407, "step": 92610 }, { "grad_norm": 0.4097318947315216, "learning_rate": 6.016445972286311e-05, "loss": 0.0247, "step": 92620 }, { "grad_norm": 0.4526313245296478, "learning_rate": 6.015636486754638e-05, "loss": 0.0324, "step": 92630 }, { "grad_norm": 0.4680594801902771, "learning_rate": 6.014826973455844e-05, "loss": 0.0282, "step": 92640 }, { "grad_norm": 0.4088941216468811, "learning_rate": 6.014017432412063e-05, "loss": 0.0223, "step": 92650 }, { "grad_norm": 0.5087025165557861, "learning_rate": 6.013207863645426e-05, "loss": 0.0353, "step": 92660 }, { "grad_norm": 0.48466089367866516, "learning_rate": 6.012398267178068e-05, "loss": 0.0373, "step": 92670 }, { "grad_norm": 1.4951823949813843, "learning_rate": 6.01158864303212e-05, "loss": 0.0573, "step": 92680 }, { "grad_norm": 0.37289944291114807, "learning_rate": 6.0107789912297185e-05, "loss": 0.034, "step": 92690 }, { "grad_norm": 0.6223412156105042, "learning_rate": 6.009969311792999e-05, "loss": 0.0276, "step": 92700 }, { "grad_norm": 0.3930400609970093, "learning_rate": 6.009159604744099e-05, "loss": 0.0346, "step": 92710 }, { "grad_norm": 0.5653695464134216, "learning_rate": 6.0083498701051545e-05, "loss": 0.0263, "step": 92720 }, { "grad_norm": 0.6708284020423889, "learning_rate": 6.007540107898303e-05, "loss": 0.033, "step": 92730 }, { "grad_norm": 0.43721914291381836, "learning_rate": 6.0067303181456834e-05, "loss": 0.0349, "step": 92740 }, { "grad_norm": 0.5147668123245239, "learning_rate": 6.005920500869434e-05, "loss": 0.0249, "step": 92750 }, { "grad_norm": 1.143075704574585, "learning_rate": 6.005110656091697e-05, "loss": 0.0361, "step": 92760 }, { "grad_norm": 0.3726729154586792, "learning_rate": 6.004300783834613e-05, "loss": 0.0399, "step": 92770 }, { "grad_norm": 0.4395466446876526, "learning_rate": 6.003490884120322e-05, "loss": 0.0403, "step": 92780 }, { "grad_norm": 0.371237188577652, "learning_rate": 6.0026809569709675e-05, "loss": 0.0305, "step": 92790 }, { "grad_norm": 0.5203465819358826, "learning_rate": 6.001871002408692e-05, "loss": 0.0372, "step": 92800 }, { "grad_norm": 0.2781396210193634, "learning_rate": 6.001061020455641e-05, "loss": 0.0349, "step": 92810 }, { "grad_norm": 0.6647344827651978, "learning_rate": 6.000251011133956e-05, "loss": 0.0445, "step": 92820 }, { "grad_norm": 0.7313388586044312, "learning_rate": 5.999440974465784e-05, "loss": 0.0346, "step": 92830 }, { "grad_norm": 0.5260970592498779, "learning_rate": 5.998630910473273e-05, "loss": 0.0306, "step": 92840 }, { "grad_norm": 0.5124896168708801, "learning_rate": 5.997820819178567e-05, "loss": 0.0331, "step": 92850 }, { "grad_norm": 1.5488150119781494, "learning_rate": 5.9970107006038155e-05, "loss": 0.0315, "step": 92860 }, { "grad_norm": 0.4386819899082184, "learning_rate": 5.996200554771165e-05, "loss": 0.0453, "step": 92870 }, { "grad_norm": 0.4430462121963501, "learning_rate": 5.9953903817027656e-05, "loss": 0.0332, "step": 92880 }, { "grad_norm": 0.5648428201675415, "learning_rate": 5.994580181420767e-05, "loss": 0.0371, "step": 92890 }, { "grad_norm": 0.38011622428894043, "learning_rate": 5.9937699539473215e-05, "loss": 0.0418, "step": 92900 }, { "grad_norm": 0.5681129693984985, "learning_rate": 5.9929596993045766e-05, "loss": 0.0281, "step": 92910 }, { "grad_norm": 0.7416437268257141, "learning_rate": 5.992149417514689e-05, "loss": 0.0318, "step": 92920 }, { "grad_norm": 0.5233862400054932, "learning_rate": 5.991339108599807e-05, "loss": 0.0323, "step": 92930 }, { "grad_norm": 0.4990457594394684, "learning_rate": 5.990528772582087e-05, "loss": 0.0428, "step": 92940 }, { "grad_norm": 0.7394016981124878, "learning_rate": 5.989718409483681e-05, "loss": 0.0461, "step": 92950 }, { "grad_norm": 0.5981101989746094, "learning_rate": 5.9889080193267475e-05, "loss": 0.0368, "step": 92960 }, { "grad_norm": 0.5578451752662659, "learning_rate": 5.98809760213344e-05, "loss": 0.0358, "step": 92970 }, { "grad_norm": 0.7579760551452637, "learning_rate": 5.987287157925915e-05, "loss": 0.0309, "step": 92980 }, { "grad_norm": 0.3579777777194977, "learning_rate": 5.98647668672633e-05, "loss": 0.0469, "step": 92990 }, { "grad_norm": 0.39779767394065857, "learning_rate": 5.985666188556842e-05, "loss": 0.0256, "step": 93000 }, { "grad_norm": 0.5589675903320312, "learning_rate": 5.9848556634396115e-05, "loss": 0.0536, "step": 93010 }, { "grad_norm": 0.416983038187027, "learning_rate": 5.9840451113967975e-05, "loss": 0.028, "step": 93020 }, { "grad_norm": 1.0686320066452026, "learning_rate": 5.98323453245056e-05, "loss": 0.0318, "step": 93030 }, { "grad_norm": 0.351030170917511, "learning_rate": 5.982423926623059e-05, "loss": 0.024, "step": 93040 }, { "grad_norm": 0.536455512046814, "learning_rate": 5.981613293936458e-05, "loss": 0.0557, "step": 93050 }, { "grad_norm": 1.100224494934082, "learning_rate": 5.980802634412916e-05, "loss": 0.0523, "step": 93060 }, { "grad_norm": 0.6078958511352539, "learning_rate": 5.9799919480746006e-05, "loss": 0.0355, "step": 93070 }, { "grad_norm": 0.9380820393562317, "learning_rate": 5.9791812349436726e-05, "loss": 0.0341, "step": 93080 }, { "grad_norm": 0.4166471064090729, "learning_rate": 5.978370495042298e-05, "loss": 0.0239, "step": 93090 }, { "grad_norm": 2.0788004398345947, "learning_rate": 5.977559728392641e-05, "loss": 0.0297, "step": 93100 }, { "grad_norm": 2.1239748001098633, "learning_rate": 5.97674893501687e-05, "loss": 0.0308, "step": 93110 }, { "grad_norm": 0.7975782155990601, "learning_rate": 5.9759381149371486e-05, "loss": 0.0313, "step": 93120 }, { "grad_norm": 1.501741886138916, "learning_rate": 5.9751272681756465e-05, "loss": 0.0337, "step": 93130 }, { "grad_norm": 6.508529186248779, "learning_rate": 5.974316394754529e-05, "loss": 0.048, "step": 93140 }, { "grad_norm": 0.5000334978103638, "learning_rate": 5.97350549469597e-05, "loss": 0.0374, "step": 93150 }, { "grad_norm": 0.4058002531528473, "learning_rate": 5.9726945680221344e-05, "loss": 0.0454, "step": 93160 }, { "grad_norm": 0.5893673300743103, "learning_rate": 5.9718836147551974e-05, "loss": 0.0319, "step": 93170 }, { "grad_norm": 0.44416412711143494, "learning_rate": 5.9710726349173254e-05, "loss": 0.0282, "step": 93180 }, { "grad_norm": 0.6781075596809387, "learning_rate": 5.970261628530693e-05, "loss": 0.0362, "step": 93190 }, { "grad_norm": 0.5898976922035217, "learning_rate": 5.969450595617472e-05, "loss": 0.0313, "step": 93200 }, { "grad_norm": 0.44227126240730286, "learning_rate": 5.968639536199836e-05, "loss": 0.029, "step": 93210 }, { "grad_norm": 0.7369299530982971, "learning_rate": 5.9678284502999596e-05, "loss": 0.0364, "step": 93220 }, { "grad_norm": 0.8181619048118591, "learning_rate": 5.967017337940016e-05, "loss": 0.0373, "step": 93230 }, { "grad_norm": 0.6572352647781372, "learning_rate": 5.966206199142182e-05, "loss": 0.0355, "step": 93240 }, { "grad_norm": 0.5648240447044373, "learning_rate": 5.9653950339286337e-05, "loss": 0.0375, "step": 93250 }, { "grad_norm": 0.4441784918308258, "learning_rate": 5.964583842321547e-05, "loss": 0.0273, "step": 93260 }, { "grad_norm": 0.6097153425216675, "learning_rate": 5.963772624343101e-05, "loss": 0.0332, "step": 93270 }, { "grad_norm": 0.41567471623420715, "learning_rate": 5.962961380015474e-05, "loss": 0.0313, "step": 93280 }, { "grad_norm": 1.7019261121749878, "learning_rate": 5.9621501093608445e-05, "loss": 0.0411, "step": 93290 }, { "grad_norm": 0.442853718996048, "learning_rate": 5.9613388124013926e-05, "loss": 0.0372, "step": 93300 }, { "grad_norm": 0.7382641434669495, "learning_rate": 5.960527489159298e-05, "loss": 0.0271, "step": 93310 }, { "grad_norm": 0.4574754238128662, "learning_rate": 5.9597161396567435e-05, "loss": 0.0337, "step": 93320 }, { "grad_norm": 0.5025722980499268, "learning_rate": 5.958904763915909e-05, "loss": 0.0327, "step": 93330 }, { "grad_norm": 0.5094088315963745, "learning_rate": 5.9580933619589796e-05, "loss": 0.0393, "step": 93340 }, { "grad_norm": 0.509128749370575, "learning_rate": 5.9572819338081375e-05, "loss": 0.0306, "step": 93350 }, { "grad_norm": 1.151216983795166, "learning_rate": 5.956470479485568e-05, "loss": 0.0424, "step": 93360 }, { "grad_norm": 0.8706373572349548, "learning_rate": 5.9556589990134524e-05, "loss": 0.0408, "step": 93370 }, { "grad_norm": 0.5470892786979675, "learning_rate": 5.954847492413981e-05, "loss": 0.0349, "step": 93380 }, { "grad_norm": 2.0154056549072266, "learning_rate": 5.9540359597093356e-05, "loss": 0.0385, "step": 93390 }, { "grad_norm": 0.7709202170372009, "learning_rate": 5.953224400921707e-05, "loss": 0.0308, "step": 93400 }, { "grad_norm": 0.8215320110321045, "learning_rate": 5.9524128160732814e-05, "loss": 0.0403, "step": 93410 }, { "grad_norm": 0.8033151626586914, "learning_rate": 5.951601205186248e-05, "loss": 0.037, "step": 93420 }, { "grad_norm": 0.31681424379348755, "learning_rate": 5.950789568282794e-05, "loss": 0.028, "step": 93430 }, { "grad_norm": 0.47033292055130005, "learning_rate": 5.9499779053851114e-05, "loss": 0.031, "step": 93440 }, { "grad_norm": 0.49390244483947754, "learning_rate": 5.9491662165153896e-05, "loss": 0.0278, "step": 93450 }, { "grad_norm": 0.4814494550228119, "learning_rate": 5.9483545016958186e-05, "loss": 0.039, "step": 93460 }, { "grad_norm": 0.35457175970077515, "learning_rate": 5.947542760948592e-05, "loss": 0.031, "step": 93470 }, { "grad_norm": 0.5511881709098816, "learning_rate": 5.946730994295904e-05, "loss": 0.0258, "step": 93480 }, { "grad_norm": 0.33699214458465576, "learning_rate": 5.945919201759944e-05, "loss": 0.0287, "step": 93490 }, { "grad_norm": 0.5482852458953857, "learning_rate": 5.94510738336291e-05, "loss": 0.0294, "step": 93500 }, { "grad_norm": 0.6447913646697998, "learning_rate": 5.944295539126995e-05, "loss": 0.0332, "step": 93510 }, { "grad_norm": 4.572673797607422, "learning_rate": 5.9434836690743944e-05, "loss": 0.045, "step": 93520 }, { "grad_norm": 0.7456363439559937, "learning_rate": 5.942671773227305e-05, "loss": 0.0345, "step": 93530 }, { "grad_norm": 0.2952421307563782, "learning_rate": 5.9418598516079226e-05, "loss": 0.032, "step": 93540 }, { "grad_norm": 0.8195688128471375, "learning_rate": 5.941047904238447e-05, "loss": 0.0351, "step": 93550 }, { "grad_norm": 0.7536357045173645, "learning_rate": 5.940235931141075e-05, "loss": 0.0454, "step": 93560 }, { "grad_norm": 0.4377247393131256, "learning_rate": 5.9394239323380064e-05, "loss": 0.0309, "step": 93570 }, { "grad_norm": 0.9282560348510742, "learning_rate": 5.938611907851438e-05, "loss": 0.0393, "step": 93580 }, { "grad_norm": 0.5646946430206299, "learning_rate": 5.937799857703574e-05, "loss": 0.0265, "step": 93590 }, { "grad_norm": 0.5245201587677002, "learning_rate": 5.9369877819166145e-05, "loss": 0.0492, "step": 93600 }, { "grad_norm": 0.41694900393486023, "learning_rate": 5.936175680512762e-05, "loss": 0.054, "step": 93610 }, { "grad_norm": 0.6584864854812622, "learning_rate": 5.935363553514217e-05, "loss": 0.0467, "step": 93620 }, { "grad_norm": 1.0426883697509766, "learning_rate": 5.934551400943185e-05, "loss": 0.0386, "step": 93630 }, { "grad_norm": 0.6278759837150574, "learning_rate": 5.9337392228218666e-05, "loss": 0.0355, "step": 93640 }, { "grad_norm": 0.5572735667228699, "learning_rate": 5.9329270191724694e-05, "loss": 0.0355, "step": 93650 }, { "grad_norm": 0.7275996804237366, "learning_rate": 5.932114790017198e-05, "loss": 0.0338, "step": 93660 }, { "grad_norm": 0.9062877893447876, "learning_rate": 5.931302535378259e-05, "loss": 0.0381, "step": 93670 }, { "grad_norm": 0.7217949628829956, "learning_rate": 5.930490255277859e-05, "loss": 0.029, "step": 93680 }, { "grad_norm": 0.4589754343032837, "learning_rate": 5.929677949738205e-05, "loss": 0.0277, "step": 93690 }, { "grad_norm": 0.6624327301979065, "learning_rate": 5.928865618781504e-05, "loss": 0.0346, "step": 93700 }, { "grad_norm": 0.47799524664878845, "learning_rate": 5.9280532624299654e-05, "loss": 0.0198, "step": 93710 }, { "grad_norm": 0.43254968523979187, "learning_rate": 5.9272408807058e-05, "loss": 0.0378, "step": 93720 }, { "grad_norm": 0.6222155690193176, "learning_rate": 5.926428473631218e-05, "loss": 0.036, "step": 93730 }, { "grad_norm": 0.40919604897499084, "learning_rate": 5.925616041228429e-05, "loss": 0.0229, "step": 93740 }, { "grad_norm": 0.5426287651062012, "learning_rate": 5.924803583519645e-05, "loss": 0.028, "step": 93750 }, { "grad_norm": 0.32689741253852844, "learning_rate": 5.92399110052708e-05, "loss": 0.0307, "step": 93760 }, { "grad_norm": 0.4133654236793518, "learning_rate": 5.923178592272943e-05, "loss": 0.0354, "step": 93770 }, { "grad_norm": 0.3482517898082733, "learning_rate": 5.9223660587794506e-05, "loss": 0.0305, "step": 93780 }, { "grad_norm": 1.2788677215576172, "learning_rate": 5.921553500068817e-05, "loss": 0.0345, "step": 93790 }, { "grad_norm": 1.7341241836547852, "learning_rate": 5.9207409161632585e-05, "loss": 0.0329, "step": 93800 }, { "grad_norm": 0.5002075433731079, "learning_rate": 5.9199283070849875e-05, "loss": 0.0368, "step": 93810 }, { "grad_norm": 2.0153472423553467, "learning_rate": 5.919115672856223e-05, "loss": 0.0311, "step": 93820 }, { "grad_norm": 0.24493925273418427, "learning_rate": 5.918303013499181e-05, "loss": 0.0365, "step": 93830 }, { "grad_norm": 0.4600871503353119, "learning_rate": 5.91749032903608e-05, "loss": 0.0333, "step": 93840 }, { "grad_norm": 0.6084553599357605, "learning_rate": 5.916677619489136e-05, "loss": 0.0361, "step": 93850 }, { "grad_norm": 0.6425840854644775, "learning_rate": 5.915864884880573e-05, "loss": 0.037, "step": 93860 }, { "grad_norm": 0.5594040155410767, "learning_rate": 5.915052125232607e-05, "loss": 0.0366, "step": 93870 }, { "grad_norm": 0.910332441329956, "learning_rate": 5.91423934056746e-05, "loss": 0.0314, "step": 93880 }, { "grad_norm": 0.4145147204399109, "learning_rate": 5.9134265309073525e-05, "loss": 0.0335, "step": 93890 }, { "grad_norm": 0.477994829416275, "learning_rate": 5.912613696274507e-05, "loss": 0.044, "step": 93900 }, { "grad_norm": 0.47904127836227417, "learning_rate": 5.911800836691145e-05, "loss": 0.0259, "step": 93910 }, { "grad_norm": 0.4008333384990692, "learning_rate": 5.910987952179492e-05, "loss": 0.03, "step": 93920 }, { "grad_norm": 0.47377264499664307, "learning_rate": 5.91017504276177e-05, "loss": 0.0297, "step": 93930 }, { "grad_norm": 0.6778913140296936, "learning_rate": 5.909362108460205e-05, "loss": 0.0296, "step": 93940 }, { "grad_norm": 0.7129138708114624, "learning_rate": 5.9085491492970205e-05, "loss": 0.0309, "step": 93950 }, { "grad_norm": 0.6860406398773193, "learning_rate": 5.907736165294444e-05, "loss": 0.0314, "step": 93960 }, { "grad_norm": 0.7426550984382629, "learning_rate": 5.906923156474702e-05, "loss": 0.0271, "step": 93970 }, { "grad_norm": 0.8764668703079224, "learning_rate": 5.906110122860021e-05, "loss": 0.0358, "step": 93980 }, { "grad_norm": 0.8832930326461792, "learning_rate": 5.90529706447263e-05, "loss": 0.0491, "step": 93990 }, { "grad_norm": 0.5398958921432495, "learning_rate": 5.904483981334757e-05, "loss": 0.0389, "step": 94000 }, { "grad_norm": 0.659375786781311, "learning_rate": 5.903670873468633e-05, "loss": 0.0345, "step": 94010 }, { "grad_norm": 0.4906490743160248, "learning_rate": 5.902857740896486e-05, "loss": 0.0358, "step": 94020 }, { "grad_norm": 0.5292624831199646, "learning_rate": 5.902044583640547e-05, "loss": 0.0282, "step": 94030 }, { "grad_norm": 0.49968692660331726, "learning_rate": 5.901231401723049e-05, "loss": 0.0404, "step": 94040 }, { "grad_norm": 0.8483418822288513, "learning_rate": 5.900418195166223e-05, "loss": 0.0402, "step": 94050 }, { "grad_norm": 0.4585250914096832, "learning_rate": 5.8996049639923e-05, "loss": 0.0361, "step": 94060 }, { "grad_norm": 0.33395296335220337, "learning_rate": 5.898791708223518e-05, "loss": 0.0409, "step": 94070 }, { "grad_norm": 0.4956508278846741, "learning_rate": 5.897978427882107e-05, "loss": 0.0288, "step": 94080 }, { "grad_norm": 0.32462894916534424, "learning_rate": 5.897165122990303e-05, "loss": 0.0238, "step": 94090 }, { "grad_norm": 0.5599524974822998, "learning_rate": 5.8963517935703404e-05, "loss": 0.0297, "step": 94100 }, { "grad_norm": 0.4273490607738495, "learning_rate": 5.8955384396444593e-05, "loss": 0.0265, "step": 94110 }, { "grad_norm": 0.542536199092865, "learning_rate": 5.894725061234892e-05, "loss": 0.0292, "step": 94120 }, { "grad_norm": 0.3741946518421173, "learning_rate": 5.893911658363878e-05, "loss": 0.0348, "step": 94130 }, { "grad_norm": 0.5575107336044312, "learning_rate": 5.8930982310536556e-05, "loss": 0.0308, "step": 94140 }, { "grad_norm": 0.9031896591186523, "learning_rate": 5.8922847793264624e-05, "loss": 0.0395, "step": 94150 }, { "grad_norm": 0.37508296966552734, "learning_rate": 5.8914713032045366e-05, "loss": 0.0277, "step": 94160 }, { "grad_norm": 0.7202122211456299, "learning_rate": 5.8906578027101235e-05, "loss": 0.0268, "step": 94170 }, { "grad_norm": 0.5988215804100037, "learning_rate": 5.889844277865459e-05, "loss": 0.0326, "step": 94180 }, { "grad_norm": 0.3963147699832916, "learning_rate": 5.8890307286927884e-05, "loss": 0.034, "step": 94190 }, { "grad_norm": 0.4395485520362854, "learning_rate": 5.8882171552143486e-05, "loss": 0.0336, "step": 94200 }, { "grad_norm": 0.7050032615661621, "learning_rate": 5.887403557452387e-05, "loss": 0.0313, "step": 94210 }, { "grad_norm": 0.49726197123527527, "learning_rate": 5.8865899354291445e-05, "loss": 0.0289, "step": 94220 }, { "grad_norm": 0.7603719830513, "learning_rate": 5.885776289166868e-05, "loss": 0.0273, "step": 94230 }, { "grad_norm": 0.8998610377311707, "learning_rate": 5.884962618687799e-05, "loss": 0.0303, "step": 94240 }, { "grad_norm": 1.219303846359253, "learning_rate": 5.8841489240141856e-05, "loss": 0.0294, "step": 94250 }, { "grad_norm": 0.6772059798240662, "learning_rate": 5.8833352051682725e-05, "loss": 0.0393, "step": 94260 }, { "grad_norm": 0.4309449791908264, "learning_rate": 5.8825214621723056e-05, "loss": 0.0409, "step": 94270 }, { "grad_norm": 1.7988744974136353, "learning_rate": 5.881707695048535e-05, "loss": 0.0438, "step": 94280 }, { "grad_norm": 0.3555532395839691, "learning_rate": 5.880893903819206e-05, "loss": 0.0275, "step": 94290 }, { "grad_norm": 0.586198091506958, "learning_rate": 5.88008008850657e-05, "loss": 0.0293, "step": 94300 }, { "grad_norm": 0.43841391801834106, "learning_rate": 5.879266249132873e-05, "loss": 0.0353, "step": 94310 }, { "grad_norm": 1.1508238315582275, "learning_rate": 5.87845238572037e-05, "loss": 0.0304, "step": 94320 }, { "grad_norm": 0.4158710837364197, "learning_rate": 5.877638498291307e-05, "loss": 0.0354, "step": 94330 }, { "grad_norm": 0.5363155603408813, "learning_rate": 5.876824586867937e-05, "loss": 0.0353, "step": 94340 }, { "grad_norm": 0.6238853931427002, "learning_rate": 5.876010651472512e-05, "loss": 0.0376, "step": 94350 }, { "grad_norm": 0.9462804198265076, "learning_rate": 5.875196692127286e-05, "loss": 0.0375, "step": 94360 }, { "grad_norm": 0.4868607521057129, "learning_rate": 5.87438270885451e-05, "loss": 0.0304, "step": 94370 }, { "grad_norm": 0.8164985775947571, "learning_rate": 5.873568701676441e-05, "loss": 0.0264, "step": 94380 }, { "grad_norm": 0.4346328377723694, "learning_rate": 5.8727546706153305e-05, "loss": 0.0296, "step": 94390 }, { "grad_norm": 0.453347384929657, "learning_rate": 5.8719406156934357e-05, "loss": 0.0301, "step": 94400 }, { "grad_norm": 0.5684769749641418, "learning_rate": 5.87112653693301e-05, "loss": 0.0221, "step": 94410 }, { "grad_norm": 0.35685256123542786, "learning_rate": 5.870312434356315e-05, "loss": 0.0262, "step": 94420 }, { "grad_norm": 0.5048567056655884, "learning_rate": 5.869498307985604e-05, "loss": 0.0354, "step": 94430 }, { "grad_norm": 0.5484380722045898, "learning_rate": 5.868684157843136e-05, "loss": 0.0281, "step": 94440 }, { "grad_norm": 0.5628528594970703, "learning_rate": 5.867869983951168e-05, "loss": 0.033, "step": 94450 }, { "grad_norm": 0.45118528604507446, "learning_rate": 5.867055786331962e-05, "loss": 0.0328, "step": 94460 }, { "grad_norm": 0.9713805913925171, "learning_rate": 5.866241565007776e-05, "loss": 0.0336, "step": 94470 }, { "grad_norm": 0.5590383410453796, "learning_rate": 5.865427320000872e-05, "loss": 0.0284, "step": 94480 }, { "grad_norm": 0.5002375245094299, "learning_rate": 5.86461305133351e-05, "loss": 0.043, "step": 94490 }, { "grad_norm": 0.4486755430698395, "learning_rate": 5.863798759027951e-05, "loss": 0.0429, "step": 94500 }, { "grad_norm": 2.2812390327453613, "learning_rate": 5.862984443106461e-05, "loss": 0.0392, "step": 94510 }, { "grad_norm": 1.028064250946045, "learning_rate": 5.862170103591298e-05, "loss": 0.0514, "step": 94520 }, { "grad_norm": 1.3652572631835938, "learning_rate": 5.861355740504729e-05, "loss": 0.0405, "step": 94530 }, { "grad_norm": 0.6345760822296143, "learning_rate": 5.8605413538690154e-05, "loss": 0.0325, "step": 94540 }, { "grad_norm": 0.600156843662262, "learning_rate": 5.859726943706428e-05, "loss": 0.0284, "step": 94550 }, { "grad_norm": 0.4847635328769684, "learning_rate": 5.8589125100392276e-05, "loss": 0.0385, "step": 94560 }, { "grad_norm": 3.4538533687591553, "learning_rate": 5.8580980528896824e-05, "loss": 0.0363, "step": 94570 }, { "grad_norm": 0.564947783946991, "learning_rate": 5.857283572280058e-05, "loss": 0.0344, "step": 94580 }, { "grad_norm": 0.4027915894985199, "learning_rate": 5.8564690682326237e-05, "loss": 0.0305, "step": 94590 }, { "grad_norm": 0.5366203188896179, "learning_rate": 5.855654540769645e-05, "loss": 0.0287, "step": 94600 }, { "grad_norm": 0.6146447062492371, "learning_rate": 5.8548399899133944e-05, "loss": 0.039, "step": 94610 }, { "grad_norm": 0.4840897023677826, "learning_rate": 5.854025415686139e-05, "loss": 0.03, "step": 94620 }, { "grad_norm": 0.5011682510375977, "learning_rate": 5.8532108181101496e-05, "loss": 0.0282, "step": 94630 }, { "grad_norm": 0.5527607798576355, "learning_rate": 5.852396197207697e-05, "loss": 0.0226, "step": 94640 }, { "grad_norm": 0.8765010833740234, "learning_rate": 5.851581553001053e-05, "loss": 0.0324, "step": 94650 }, { "grad_norm": 0.4599286913871765, "learning_rate": 5.8507668855124885e-05, "loss": 0.0173, "step": 94660 }, { "grad_norm": 0.23935741186141968, "learning_rate": 5.8499521947642764e-05, "loss": 0.0256, "step": 94670 }, { "grad_norm": 0.4892892837524414, "learning_rate": 5.849137480778691e-05, "loss": 0.0333, "step": 94680 }, { "grad_norm": 0.6479471921920776, "learning_rate": 5.848322743578008e-05, "loss": 0.0329, "step": 94690 }, { "grad_norm": 1.3608194589614868, "learning_rate": 5.8475079831844984e-05, "loss": 0.0397, "step": 94700 }, { "grad_norm": 0.34517285227775574, "learning_rate": 5.846693199620438e-05, "loss": 0.0298, "step": 94710 }, { "grad_norm": 0.5231087803840637, "learning_rate": 5.845878392908105e-05, "loss": 0.0317, "step": 94720 }, { "grad_norm": 1.2401241064071655, "learning_rate": 5.845063563069774e-05, "loss": 0.0305, "step": 94730 }, { "grad_norm": 0.36291664838790894, "learning_rate": 5.844248710127722e-05, "loss": 0.0499, "step": 94740 }, { "grad_norm": 0.3909893333911896, "learning_rate": 5.843433834104227e-05, "loss": 0.0394, "step": 94750 }, { "grad_norm": 0.6977999210357666, "learning_rate": 5.8426189350215696e-05, "loss": 0.0331, "step": 94760 }, { "grad_norm": 0.5565579533576965, "learning_rate": 5.841804012902027e-05, "loss": 0.0369, "step": 94770 }, { "grad_norm": 0.5465147495269775, "learning_rate": 5.840989067767878e-05, "loss": 0.0229, "step": 94780 }, { "grad_norm": 0.6709083914756775, "learning_rate": 5.8401740996414034e-05, "loss": 0.0512, "step": 94790 }, { "grad_norm": 0.45223379135131836, "learning_rate": 5.8393591085448865e-05, "loss": 0.0184, "step": 94800 }, { "grad_norm": 0.6993906497955322, "learning_rate": 5.8385440945006044e-05, "loss": 0.031, "step": 94810 }, { "grad_norm": 0.438129186630249, "learning_rate": 5.8377290575308444e-05, "loss": 0.0353, "step": 94820 }, { "grad_norm": 0.5383040308952332, "learning_rate": 5.836913997657885e-05, "loss": 0.0397, "step": 94830 }, { "grad_norm": 0.3721461892127991, "learning_rate": 5.8360989149040125e-05, "loss": 0.0361, "step": 94840 }, { "grad_norm": 0.5317333340644836, "learning_rate": 5.835283809291509e-05, "loss": 0.0364, "step": 94850 }, { "grad_norm": 0.6898958086967468, "learning_rate": 5.834468680842659e-05, "loss": 0.0282, "step": 94860 }, { "grad_norm": 0.3576701581478119, "learning_rate": 5.833653529579749e-05, "loss": 0.0517, "step": 94870 }, { "grad_norm": 0.6345551609992981, "learning_rate": 5.832838355525065e-05, "loss": 0.0278, "step": 94880 }, { "grad_norm": 0.4950910210609436, "learning_rate": 5.8320231587008946e-05, "loss": 0.0285, "step": 94890 }, { "grad_norm": 0.49363332986831665, "learning_rate": 5.831207939129523e-05, "loss": 0.0269, "step": 94900 }, { "grad_norm": 1.3156089782714844, "learning_rate": 5.8303926968332375e-05, "loss": 0.0581, "step": 94910 }, { "grad_norm": 1.6803780794143677, "learning_rate": 5.829577431834328e-05, "loss": 0.0312, "step": 94920 }, { "grad_norm": 0.6866757869720459, "learning_rate": 5.828762144155083e-05, "loss": 0.0338, "step": 94930 }, { "grad_norm": 0.46985769271850586, "learning_rate": 5.827946833817795e-05, "loss": 0.0383, "step": 94940 }, { "grad_norm": 1.1448684930801392, "learning_rate": 5.827131500844748e-05, "loss": 0.0351, "step": 94950 }, { "grad_norm": 0.5230758786201477, "learning_rate": 5.8263161452582384e-05, "loss": 0.0415, "step": 94960 }, { "grad_norm": 0.5969500541687012, "learning_rate": 5.8255007670805564e-05, "loss": 0.024, "step": 94970 }, { "grad_norm": 0.5518593788146973, "learning_rate": 5.824685366333992e-05, "loss": 0.0315, "step": 94980 }, { "grad_norm": 0.46476367115974426, "learning_rate": 5.8238699430408394e-05, "loss": 0.0298, "step": 94990 }, { "grad_norm": 0.5356079339981079, "learning_rate": 5.823054497223392e-05, "loss": 0.0431, "step": 95000 }, { "grad_norm": 0.4476495087146759, "learning_rate": 5.8222390289039456e-05, "loss": 0.0329, "step": 95010 }, { "grad_norm": 0.43433934450149536, "learning_rate": 5.8214235381047914e-05, "loss": 0.0284, "step": 95020 }, { "grad_norm": 0.42702779173851013, "learning_rate": 5.8206080248482275e-05, "loss": 0.0276, "step": 95030 }, { "grad_norm": 0.44209691882133484, "learning_rate": 5.819792489156548e-05, "loss": 0.0378, "step": 95040 }, { "grad_norm": 0.47762250900268555, "learning_rate": 5.8189769310520495e-05, "loss": 0.0222, "step": 95050 }, { "grad_norm": 0.536874532699585, "learning_rate": 5.8181613505570284e-05, "loss": 0.0342, "step": 95060 }, { "grad_norm": 0.3211955428123474, "learning_rate": 5.8173457476937856e-05, "loss": 0.0266, "step": 95070 }, { "grad_norm": 1.6203471422195435, "learning_rate": 5.8165301224846156e-05, "loss": 0.0323, "step": 95080 }, { "grad_norm": 0.39970773458480835, "learning_rate": 5.81571447495182e-05, "loss": 0.0343, "step": 95090 }, { "grad_norm": 0.7331780791282654, "learning_rate": 5.8148988051176954e-05, "loss": 0.0291, "step": 95100 }, { "grad_norm": 0.40209531784057617, "learning_rate": 5.8140831130045445e-05, "loss": 0.0475, "step": 95110 }, { "grad_norm": 0.6806427240371704, "learning_rate": 5.813267398634668e-05, "loss": 0.0357, "step": 95120 }, { "grad_norm": 0.37350034713745117, "learning_rate": 5.812451662030366e-05, "loss": 0.0269, "step": 95130 }, { "grad_norm": 0.3727569580078125, "learning_rate": 5.811635903213939e-05, "loss": 0.0327, "step": 95140 }, { "grad_norm": 0.5645720362663269, "learning_rate": 5.8108201222076944e-05, "loss": 0.0345, "step": 95150 }, { "grad_norm": 0.7271094918251038, "learning_rate": 5.8100043190339295e-05, "loss": 0.0269, "step": 95160 }, { "grad_norm": 0.5604318976402283, "learning_rate": 5.809188493714952e-05, "loss": 0.0236, "step": 95170 }, { "grad_norm": 1.3187975883483887, "learning_rate": 5.8083726462730635e-05, "loss": 0.0502, "step": 95180 }, { "grad_norm": 1.0085830688476562, "learning_rate": 5.807556776730574e-05, "loss": 0.0296, "step": 95190 }, { "grad_norm": 0.8455968499183655, "learning_rate": 5.8067408851097825e-05, "loss": 0.0624, "step": 95200 }, { "grad_norm": 0.8370541334152222, "learning_rate": 5.805924971432999e-05, "loss": 0.0366, "step": 95210 }, { "grad_norm": 0.4027293622493744, "learning_rate": 5.8051090357225314e-05, "loss": 0.0235, "step": 95220 }, { "grad_norm": 0.6827797293663025, "learning_rate": 5.804293078000683e-05, "loss": 0.0254, "step": 95230 }, { "grad_norm": 0.6454222798347473, "learning_rate": 5.803477098289765e-05, "loss": 0.0377, "step": 95240 }, { "grad_norm": 0.6976858973503113, "learning_rate": 5.802661096612083e-05, "loss": 0.0458, "step": 95250 }, { "grad_norm": 0.4924894869327545, "learning_rate": 5.801845072989951e-05, "loss": 0.0275, "step": 95260 }, { "grad_norm": 0.43605223298072815, "learning_rate": 5.801029027445675e-05, "loss": 0.0338, "step": 95270 }, { "grad_norm": 0.29318153858184814, "learning_rate": 5.800212960001567e-05, "loss": 0.0213, "step": 95280 }, { "grad_norm": 0.8345481753349304, "learning_rate": 5.7993968706799365e-05, "loss": 0.0322, "step": 95290 }, { "grad_norm": 0.4617365896701813, "learning_rate": 5.798580759503095e-05, "loss": 0.0337, "step": 95300 }, { "grad_norm": 0.5821983814239502, "learning_rate": 5.797764626493356e-05, "loss": 0.0289, "step": 95310 }, { "grad_norm": 0.588367223739624, "learning_rate": 5.796948471673033e-05, "loss": 0.0308, "step": 95320 }, { "grad_norm": 0.666019856929779, "learning_rate": 5.796132295064437e-05, "loss": 0.0286, "step": 95330 }, { "grad_norm": 1.0196372270584106, "learning_rate": 5.7953160966898846e-05, "loss": 0.0307, "step": 95340 }, { "grad_norm": 0.5135505199432373, "learning_rate": 5.7944998765716866e-05, "loss": 0.0325, "step": 95350 }, { "grad_norm": 2.2002217769622803, "learning_rate": 5.7936836347321634e-05, "loss": 0.0348, "step": 95360 }, { "grad_norm": 0.38131165504455566, "learning_rate": 5.7928673711936247e-05, "loss": 0.0228, "step": 95370 }, { "grad_norm": 0.47177669405937195, "learning_rate": 5.792051085978391e-05, "loss": 0.0219, "step": 95380 }, { "grad_norm": 0.2367016077041626, "learning_rate": 5.791234779108778e-05, "loss": 0.029, "step": 95390 }, { "grad_norm": 0.5315960645675659, "learning_rate": 5.7904184506071046e-05, "loss": 0.0283, "step": 95400 }, { "grad_norm": 0.5164232850074768, "learning_rate": 5.789602100495687e-05, "loss": 0.0289, "step": 95410 }, { "grad_norm": 1.2052408456802368, "learning_rate": 5.788785728796844e-05, "loss": 0.0347, "step": 95420 }, { "grad_norm": 0.4562908709049225, "learning_rate": 5.7879693355328954e-05, "loss": 0.0358, "step": 95430 }, { "grad_norm": 0.5438938736915588, "learning_rate": 5.787152920726163e-05, "loss": 0.0301, "step": 95440 }, { "grad_norm": 0.4394547641277313, "learning_rate": 5.786336484398964e-05, "loss": 0.0312, "step": 95450 }, { "grad_norm": 1.0884668827056885, "learning_rate": 5.785520026573621e-05, "loss": 0.0364, "step": 95460 }, { "grad_norm": 1.295399785041809, "learning_rate": 5.784703547272457e-05, "loss": 0.0557, "step": 95470 }, { "grad_norm": 0.3889925181865692, "learning_rate": 5.783887046517791e-05, "loss": 0.0263, "step": 95480 }, { "grad_norm": 0.4498007595539093, "learning_rate": 5.7830705243319486e-05, "loss": 0.0277, "step": 95490 }, { "grad_norm": 0.3830857574939728, "learning_rate": 5.782253980737251e-05, "loss": 0.0283, "step": 95500 }, { "grad_norm": 1.020450472831726, "learning_rate": 5.781437415756026e-05, "loss": 0.0429, "step": 95510 }, { "grad_norm": 1.0571722984313965, "learning_rate": 5.780620829410594e-05, "loss": 0.0284, "step": 95520 }, { "grad_norm": 0.5953222513198853, "learning_rate": 5.779804221723282e-05, "loss": 0.0382, "step": 95530 }, { "grad_norm": 1.290290117263794, "learning_rate": 5.7789875927164147e-05, "loss": 0.0413, "step": 95540 }, { "grad_norm": 0.49840864539146423, "learning_rate": 5.778170942412321e-05, "loss": 0.0359, "step": 95550 }, { "grad_norm": 0.40447568893432617, "learning_rate": 5.777354270833324e-05, "loss": 0.0454, "step": 95560 }, { "grad_norm": 0.42005208134651184, "learning_rate": 5.776537578001755e-05, "loss": 0.0311, "step": 95570 }, { "grad_norm": 0.4624519944190979, "learning_rate": 5.775720863939939e-05, "loss": 0.0393, "step": 95580 }, { "grad_norm": 0.5293760299682617, "learning_rate": 5.774904128670206e-05, "loss": 0.0342, "step": 95590 }, { "grad_norm": 0.5097499489784241, "learning_rate": 5.774087372214887e-05, "loss": 0.0267, "step": 95600 }, { "grad_norm": 0.46184593439102173, "learning_rate": 5.773270594596309e-05, "loss": 0.0271, "step": 95610 }, { "grad_norm": 0.36566510796546936, "learning_rate": 5.772453795836803e-05, "loss": 0.0372, "step": 95620 }, { "grad_norm": 0.5341957807540894, "learning_rate": 5.771636975958701e-05, "loss": 0.0295, "step": 95630 }, { "grad_norm": 0.36893147230148315, "learning_rate": 5.770820134984334e-05, "loss": 0.0682, "step": 95640 }, { "grad_norm": 1.2866339683532715, "learning_rate": 5.7700032729360346e-05, "loss": 0.0449, "step": 95650 }, { "grad_norm": 0.6607701778411865, "learning_rate": 5.769186389836133e-05, "loss": 0.031, "step": 95660 }, { "grad_norm": 0.5631270408630371, "learning_rate": 5.768369485706966e-05, "loss": 0.0304, "step": 95670 }, { "grad_norm": 0.4886820614337921, "learning_rate": 5.767552560570865e-05, "loss": 0.0426, "step": 95680 }, { "grad_norm": 0.9988505840301514, "learning_rate": 5.7667356144501675e-05, "loss": 0.0689, "step": 95690 }, { "grad_norm": 0.9249886870384216, "learning_rate": 5.765918647367203e-05, "loss": 0.043, "step": 95700 }, { "grad_norm": 1.4041200876235962, "learning_rate": 5.765101659344313e-05, "loss": 0.044, "step": 95710 }, { "grad_norm": 0.4839459955692291, "learning_rate": 5.764284650403831e-05, "loss": 0.0347, "step": 95720 }, { "grad_norm": 0.6469258069992065, "learning_rate": 5.7634676205680924e-05, "loss": 0.03, "step": 95730 }, { "grad_norm": 0.38833698630332947, "learning_rate": 5.762650569859438e-05, "loss": 0.0324, "step": 95740 }, { "grad_norm": 0.5135953426361084, "learning_rate": 5.7618334983002006e-05, "loss": 0.0315, "step": 95750 }, { "grad_norm": 0.6559708118438721, "learning_rate": 5.761016405912724e-05, "loss": 0.0417, "step": 95760 }, { "grad_norm": 0.7851808667182922, "learning_rate": 5.7601992927193435e-05, "loss": 0.0339, "step": 95770 }, { "grad_norm": 0.5713172554969788, "learning_rate": 5.759382158742401e-05, "loss": 0.0363, "step": 95780 }, { "grad_norm": 0.9226561784744263, "learning_rate": 5.758565004004235e-05, "loss": 0.0369, "step": 95790 }, { "grad_norm": 0.4489760100841522, "learning_rate": 5.757747828527187e-05, "loss": 0.0245, "step": 95800 }, { "grad_norm": 0.5411319136619568, "learning_rate": 5.756930632333596e-05, "loss": 0.0261, "step": 95810 }, { "grad_norm": 0.8015331625938416, "learning_rate": 5.7561134154458084e-05, "loss": 0.0326, "step": 95820 }, { "grad_norm": 0.6182081699371338, "learning_rate": 5.7552961778861624e-05, "loss": 0.0343, "step": 95830 }, { "grad_norm": 0.5232490301132202, "learning_rate": 5.7544789196770045e-05, "loss": 0.0456, "step": 95840 }, { "grad_norm": 0.49339061975479126, "learning_rate": 5.753661640840675e-05, "loss": 0.0297, "step": 95850 }, { "grad_norm": 0.5880131125450134, "learning_rate": 5.75284434139952e-05, "loss": 0.0331, "step": 95860 }, { "grad_norm": 1.1774519681930542, "learning_rate": 5.752027021375883e-05, "loss": 0.0431, "step": 95870 }, { "grad_norm": 0.5257726907730103, "learning_rate": 5.75120968079211e-05, "loss": 0.03, "step": 95880 }, { "grad_norm": 0.36144837737083435, "learning_rate": 5.7503923196705465e-05, "loss": 0.0236, "step": 95890 }, { "grad_norm": 0.2636711597442627, "learning_rate": 5.74957493803354e-05, "loss": 0.0325, "step": 95900 }, { "grad_norm": 0.6951267123222351, "learning_rate": 5.748757535903435e-05, "loss": 0.0451, "step": 95910 }, { "grad_norm": 0.7024590373039246, "learning_rate": 5.74794011330258e-05, "loss": 0.0408, "step": 95920 }, { "grad_norm": 0.6986164450645447, "learning_rate": 5.7471226702533244e-05, "loss": 0.0292, "step": 95930 }, { "grad_norm": 1.0770725011825562, "learning_rate": 5.746305206778016e-05, "loss": 0.0372, "step": 95940 }, { "grad_norm": 0.7453944683074951, "learning_rate": 5.745487722899001e-05, "loss": 0.0317, "step": 95950 }, { "grad_norm": 0.5071442723274231, "learning_rate": 5.744670218638634e-05, "loss": 0.0296, "step": 95960 }, { "grad_norm": 0.5259467363357544, "learning_rate": 5.7438526940192626e-05, "loss": 0.0448, "step": 95970 }, { "grad_norm": 0.28479570150375366, "learning_rate": 5.7430351490632374e-05, "loss": 0.0414, "step": 95980 }, { "grad_norm": 0.4995824694633484, "learning_rate": 5.742217583792912e-05, "loss": 0.0365, "step": 95990 }, { "grad_norm": 0.6876243948936462, "learning_rate": 5.7413999982306345e-05, "loss": 0.0293, "step": 96000 }, { "grad_norm": 0.3699477016925812, "learning_rate": 5.740582392398761e-05, "loss": 0.0315, "step": 96010 }, { "grad_norm": 0.6052810549736023, "learning_rate": 5.7397647663196416e-05, "loss": 0.028, "step": 96020 }, { "grad_norm": 0.35227593779563904, "learning_rate": 5.738947120015634e-05, "loss": 0.0435, "step": 96030 }, { "grad_norm": 0.4931417405605316, "learning_rate": 5.738129453509088e-05, "loss": 0.0344, "step": 96040 }, { "grad_norm": 0.4418470859527588, "learning_rate": 5.737311766822361e-05, "loss": 0.0345, "step": 96050 }, { "grad_norm": 0.29308322072029114, "learning_rate": 5.736494059977806e-05, "loss": 0.0291, "step": 96060 }, { "grad_norm": 0.6654694676399231, "learning_rate": 5.7356763329977804e-05, "loss": 0.0471, "step": 96070 }, { "grad_norm": 0.47091221809387207, "learning_rate": 5.73485858590464e-05, "loss": 0.0329, "step": 96080 }, { "grad_norm": 0.575046956539154, "learning_rate": 5.7340408187207427e-05, "loss": 0.0227, "step": 96090 }, { "grad_norm": 0.4994950294494629, "learning_rate": 5.733223031468445e-05, "loss": 0.0299, "step": 96100 }, { "grad_norm": 0.7672926187515259, "learning_rate": 5.732405224170105e-05, "loss": 0.0352, "step": 96110 }, { "grad_norm": 0.41025906801223755, "learning_rate": 5.731587396848081e-05, "loss": 0.0333, "step": 96120 }, { "grad_norm": 0.5090916156768799, "learning_rate": 5.730769549524733e-05, "loss": 0.0413, "step": 96130 }, { "grad_norm": 0.8955886363983154, "learning_rate": 5.72995168222242e-05, "loss": 0.0527, "step": 96140 }, { "grad_norm": 0.4223652780056, "learning_rate": 5.729133794963502e-05, "loss": 0.026, "step": 96150 }, { "grad_norm": 0.8798855543136597, "learning_rate": 5.72831588777034e-05, "loss": 0.044, "step": 96160 }, { "grad_norm": 0.47302258014678955, "learning_rate": 5.7274979606652944e-05, "loss": 0.0252, "step": 96170 }, { "grad_norm": 0.6732344031333923, "learning_rate": 5.72668001367073e-05, "loss": 0.0325, "step": 96180 }, { "grad_norm": 0.5817117094993591, "learning_rate": 5.7258620468090053e-05, "loss": 0.0327, "step": 96190 }, { "grad_norm": 0.5745179653167725, "learning_rate": 5.7250440601024844e-05, "loss": 0.0345, "step": 96200 }, { "grad_norm": 0.43093663454055786, "learning_rate": 5.7242260535735325e-05, "loss": 0.0404, "step": 96210 }, { "grad_norm": 0.8568543791770935, "learning_rate": 5.723408027244513e-05, "loss": 0.0417, "step": 96220 }, { "grad_norm": 0.422098308801651, "learning_rate": 5.722589981137788e-05, "loss": 0.0407, "step": 96230 }, { "grad_norm": 0.5218200087547302, "learning_rate": 5.721771915275727e-05, "loss": 0.0379, "step": 96240 }, { "grad_norm": 0.4373098313808441, "learning_rate": 5.720953829680691e-05, "loss": 0.0393, "step": 96250 }, { "grad_norm": 0.32299718260765076, "learning_rate": 5.7201357243750475e-05, "loss": 0.0274, "step": 96260 }, { "grad_norm": 0.45211470127105713, "learning_rate": 5.7193175993811646e-05, "loss": 0.0321, "step": 96270 }, { "grad_norm": 0.4680432677268982, "learning_rate": 5.71849945472141e-05, "loss": 0.0267, "step": 96280 }, { "grad_norm": 0.39872124791145325, "learning_rate": 5.7176812904181485e-05, "loss": 0.0279, "step": 96290 }, { "grad_norm": 0.33962002396583557, "learning_rate": 5.7168631064937515e-05, "loss": 0.0356, "step": 96300 }, { "grad_norm": 0.3756754994392395, "learning_rate": 5.7160449029705855e-05, "loss": 0.0284, "step": 96310 }, { "grad_norm": 0.5249772667884827, "learning_rate": 5.715226679871021e-05, "loss": 0.0301, "step": 96320 }, { "grad_norm": 0.6299551129341125, "learning_rate": 5.714408437217428e-05, "loss": 0.0388, "step": 96330 }, { "grad_norm": 0.58436119556427, "learning_rate": 5.7135901750321774e-05, "loss": 0.0337, "step": 96340 }, { "grad_norm": 0.46726903319358826, "learning_rate": 5.7127718933376384e-05, "loss": 0.027, "step": 96350 }, { "grad_norm": 0.624725878238678, "learning_rate": 5.711953592156184e-05, "loss": 0.0363, "step": 96360 }, { "grad_norm": 0.4036841094493866, "learning_rate": 5.7111352715101866e-05, "loss": 0.0339, "step": 96370 }, { "grad_norm": 0.7538849115371704, "learning_rate": 5.710316931422016e-05, "loss": 0.032, "step": 96380 }, { "grad_norm": 0.6233367323875427, "learning_rate": 5.709498571914048e-05, "loss": 0.0317, "step": 96390 }, { "grad_norm": 0.5989446640014648, "learning_rate": 5.7086801930086574e-05, "loss": 0.0391, "step": 96400 }, { "grad_norm": 0.2726397216320038, "learning_rate": 5.7078617947282155e-05, "loss": 0.0323, "step": 96410 }, { "grad_norm": 0.5281200408935547, "learning_rate": 5.707043377095097e-05, "loss": 0.0333, "step": 96420 }, { "grad_norm": 0.36897334456443787, "learning_rate": 5.70622494013168e-05, "loss": 0.025, "step": 96430 }, { "grad_norm": 0.6797336339950562, "learning_rate": 5.705406483860337e-05, "loss": 0.0354, "step": 96440 }, { "grad_norm": 0.5506234765052795, "learning_rate": 5.7045880083034464e-05, "loss": 0.0313, "step": 96450 }, { "grad_norm": 0.5084333419799805, "learning_rate": 5.703769513483383e-05, "loss": 0.0333, "step": 96460 }, { "grad_norm": 0.4965425431728363, "learning_rate": 5.702950999422528e-05, "loss": 0.0306, "step": 96470 }, { "grad_norm": 0.996036946773529, "learning_rate": 5.702132466143255e-05, "loss": 0.0515, "step": 96480 }, { "grad_norm": 0.45447978377342224, "learning_rate": 5.7013139136679464e-05, "loss": 0.0388, "step": 96490 }, { "grad_norm": 0.4689699709415436, "learning_rate": 5.700495342018977e-05, "loss": 0.0307, "step": 96500 }, { "grad_norm": 0.34897929430007935, "learning_rate": 5.699676751218728e-05, "loss": 0.0318, "step": 96510 }, { "grad_norm": 0.38493847846984863, "learning_rate": 5.698858141289579e-05, "loss": 0.0301, "step": 96520 }, { "grad_norm": 1.0957993268966675, "learning_rate": 5.698039512253912e-05, "loss": 0.0298, "step": 96530 }, { "grad_norm": 0.312806636095047, "learning_rate": 5.6972208641341064e-05, "loss": 0.0245, "step": 96540 }, { "grad_norm": 0.34617576003074646, "learning_rate": 5.696402196952546e-05, "loss": 0.0269, "step": 96550 }, { "grad_norm": 0.6477895379066467, "learning_rate": 5.695583510731609e-05, "loss": 0.0299, "step": 96560 }, { "grad_norm": 0.5671438574790955, "learning_rate": 5.694764805493682e-05, "loss": 0.0243, "step": 96570 }, { "grad_norm": 0.8395051956176758, "learning_rate": 5.6939460812611436e-05, "loss": 0.0363, "step": 96580 }, { "grad_norm": 1.140723705291748, "learning_rate": 5.693127338056382e-05, "loss": 0.0322, "step": 96590 }, { "grad_norm": 0.26309439539909363, "learning_rate": 5.692308575901779e-05, "loss": 0.0283, "step": 96600 }, { "grad_norm": 0.627852737903595, "learning_rate": 5.69148979481972e-05, "loss": 0.0296, "step": 96610 }, { "grad_norm": 1.047719120979309, "learning_rate": 5.690670994832589e-05, "loss": 0.0346, "step": 96620 }, { "grad_norm": 0.9190741777420044, "learning_rate": 5.689852175962772e-05, "loss": 0.0344, "step": 96630 }, { "grad_norm": 0.8969599604606628, "learning_rate": 5.6890333382326566e-05, "loss": 0.0352, "step": 96640 }, { "grad_norm": 0.44503533840179443, "learning_rate": 5.6882144816646286e-05, "loss": 0.0221, "step": 96650 }, { "grad_norm": 0.6313518285751343, "learning_rate": 5.687395606281075e-05, "loss": 0.0259, "step": 96660 }, { "grad_norm": 0.427726149559021, "learning_rate": 5.686576712104383e-05, "loss": 0.0243, "step": 96670 }, { "grad_norm": 0.3421591520309448, "learning_rate": 5.685757799156944e-05, "loss": 0.0311, "step": 96680 }, { "grad_norm": 0.7570383548736572, "learning_rate": 5.684938867461143e-05, "loss": 0.036, "step": 96690 }, { "grad_norm": 0.992483913898468, "learning_rate": 5.68411991703937e-05, "loss": 0.0385, "step": 96700 }, { "grad_norm": 0.5796877145767212, "learning_rate": 5.683300947914015e-05, "loss": 0.0209, "step": 96710 }, { "grad_norm": 0.6478826403617859, "learning_rate": 5.682481960107471e-05, "loss": 0.0246, "step": 96720 }, { "grad_norm": 0.5497615933418274, "learning_rate": 5.681662953642125e-05, "loss": 0.0428, "step": 96730 }, { "grad_norm": 0.3798612952232361, "learning_rate": 5.6808439285403716e-05, "loss": 0.0461, "step": 96740 }, { "grad_norm": 0.38550302386283875, "learning_rate": 5.6800248848246e-05, "loss": 0.0271, "step": 96750 }, { "grad_norm": 0.34990718960762024, "learning_rate": 5.679205822517204e-05, "loss": 0.0444, "step": 96760 }, { "grad_norm": 0.7315101027488708, "learning_rate": 5.6783867416405734e-05, "loss": 0.0439, "step": 96770 }, { "grad_norm": 0.3917868733406067, "learning_rate": 5.677567642217107e-05, "loss": 0.0304, "step": 96780 }, { "grad_norm": 0.5642104148864746, "learning_rate": 5.676748524269194e-05, "loss": 0.0318, "step": 96790 }, { "grad_norm": 0.548087477684021, "learning_rate": 5.675929387819232e-05, "loss": 0.0334, "step": 96800 }, { "grad_norm": 0.5323836207389832, "learning_rate": 5.675110232889614e-05, "loss": 0.0264, "step": 96810 }, { "grad_norm": 0.7248335480690002, "learning_rate": 5.6742910595027366e-05, "loss": 0.0459, "step": 96820 }, { "grad_norm": 1.7780487537384033, "learning_rate": 5.6734718676809926e-05, "loss": 0.0403, "step": 96830 }, { "grad_norm": 0.8291771411895752, "learning_rate": 5.672652657446783e-05, "loss": 0.0319, "step": 96840 }, { "grad_norm": 1.1124643087387085, "learning_rate": 5.671833428822502e-05, "loss": 0.0273, "step": 96850 }, { "grad_norm": 0.47669854760169983, "learning_rate": 5.6710141818305476e-05, "loss": 0.0513, "step": 96860 }, { "grad_norm": 0.42456290125846863, "learning_rate": 5.6701949164933165e-05, "loss": 0.0385, "step": 96870 }, { "grad_norm": 0.4215793311595917, "learning_rate": 5.6693756328332084e-05, "loss": 0.0293, "step": 96880 }, { "grad_norm": 0.49282148480415344, "learning_rate": 5.668556330872622e-05, "loss": 0.0311, "step": 96890 }, { "grad_norm": 0.4701014757156372, "learning_rate": 5.667737010633958e-05, "loss": 0.036, "step": 96900 }, { "grad_norm": 1.55496084690094, "learning_rate": 5.666917672139613e-05, "loss": 0.0384, "step": 96910 }, { "grad_norm": 0.20843355357646942, "learning_rate": 5.666098315411991e-05, "loss": 0.0329, "step": 96920 }, { "grad_norm": 0.6907283663749695, "learning_rate": 5.6652789404734904e-05, "loss": 0.0255, "step": 96930 }, { "grad_norm": 0.6062840223312378, "learning_rate": 5.664459547346514e-05, "loss": 0.04, "step": 96940 }, { "grad_norm": 0.41386863589286804, "learning_rate": 5.663640136053464e-05, "loss": 0.0288, "step": 96950 }, { "grad_norm": 0.6265805959701538, "learning_rate": 5.6628207066167395e-05, "loss": 0.0368, "step": 96960 }, { "grad_norm": 0.5119907855987549, "learning_rate": 5.662001259058748e-05, "loss": 0.0294, "step": 96970 }, { "grad_norm": 0.5660344958305359, "learning_rate": 5.661181793401889e-05, "loss": 0.0274, "step": 96980 }, { "grad_norm": 0.24024926126003265, "learning_rate": 5.660362309668571e-05, "loss": 0.0361, "step": 96990 }, { "grad_norm": 0.3019554018974304, "learning_rate": 5.659542807881193e-05, "loss": 0.0293, "step": 97000 }, { "grad_norm": 0.5986632704734802, "learning_rate": 5.6587232880621646e-05, "loss": 0.0378, "step": 97010 }, { "grad_norm": 0.3881347179412842, "learning_rate": 5.657903750233886e-05, "loss": 0.0373, "step": 97020 }, { "grad_norm": 0.3062974214553833, "learning_rate": 5.657084194418768e-05, "loss": 0.0334, "step": 97030 }, { "grad_norm": 0.5921733379364014, "learning_rate": 5.6562646206392146e-05, "loss": 0.0351, "step": 97040 }, { "grad_norm": 0.6466537714004517, "learning_rate": 5.655445028917634e-05, "loss": 0.0286, "step": 97050 }, { "grad_norm": 0.36827653646469116, "learning_rate": 5.6546254192764314e-05, "loss": 0.0419, "step": 97060 }, { "grad_norm": 0.5463423728942871, "learning_rate": 5.6538057917380174e-05, "loss": 0.0321, "step": 97070 }, { "grad_norm": 0.469431608915329, "learning_rate": 5.6529861463247966e-05, "loss": 0.0249, "step": 97080 }, { "grad_norm": 0.45140013098716736, "learning_rate": 5.65216648305918e-05, "loss": 0.0406, "step": 97090 }, { "grad_norm": 0.6823347806930542, "learning_rate": 5.6513468019635775e-05, "loss": 0.0308, "step": 97100 }, { "grad_norm": 0.6114554405212402, "learning_rate": 5.650527103060398e-05, "loss": 0.038, "step": 97110 }, { "grad_norm": 0.41846325993537903, "learning_rate": 5.649707386372052e-05, "loss": 0.029, "step": 97120 }, { "grad_norm": 1.2818193435668945, "learning_rate": 5.648887651920949e-05, "loss": 0.0421, "step": 97130 }, { "grad_norm": 0.5631383657455444, "learning_rate": 5.6480678997295036e-05, "loss": 0.0269, "step": 97140 }, { "grad_norm": 0.876508355140686, "learning_rate": 5.647248129820122e-05, "loss": 0.0307, "step": 97150 }, { "grad_norm": 0.979421854019165, "learning_rate": 5.646428342215223e-05, "loss": 0.0398, "step": 97160 }, { "grad_norm": 0.5304279327392578, "learning_rate": 5.6456085369372144e-05, "loss": 0.0341, "step": 97170 }, { "grad_norm": 0.6715513467788696, "learning_rate": 5.644788714008512e-05, "loss": 0.0275, "step": 97180 }, { "grad_norm": 0.9585444331169128, "learning_rate": 5.6439688734515274e-05, "loss": 0.0308, "step": 97190 }, { "grad_norm": 0.9615318775177002, "learning_rate": 5.6431490152886766e-05, "loss": 0.0298, "step": 97200 }, { "grad_norm": 0.5630415678024292, "learning_rate": 5.642329139542372e-05, "loss": 0.0613, "step": 97210 }, { "grad_norm": 0.6155786514282227, "learning_rate": 5.6415092462350304e-05, "loss": 0.069, "step": 97220 }, { "grad_norm": 0.38147473335266113, "learning_rate": 5.6406893353890675e-05, "loss": 0.0303, "step": 97230 }, { "grad_norm": 0.5621039271354675, "learning_rate": 5.6398694070268996e-05, "loss": 0.0289, "step": 97240 }, { "grad_norm": 0.5073451995849609, "learning_rate": 5.639049461170941e-05, "loss": 0.0305, "step": 97250 }, { "grad_norm": 0.40113168954849243, "learning_rate": 5.6382294978436124e-05, "loss": 0.025, "step": 97260 }, { "grad_norm": 1.0294981002807617, "learning_rate": 5.637409517067327e-05, "loss": 0.0339, "step": 97270 }, { "grad_norm": 0.5633421540260315, "learning_rate": 5.636589518864505e-05, "loss": 0.0348, "step": 97280 }, { "grad_norm": 0.6977658271789551, "learning_rate": 5.6357695032575644e-05, "loss": 0.0342, "step": 97290 }, { "grad_norm": 0.8165563941001892, "learning_rate": 5.6349494702689267e-05, "loss": 0.0417, "step": 97300 }, { "grad_norm": 0.3538430631160736, "learning_rate": 5.634129419921006e-05, "loss": 0.0252, "step": 97310 }, { "grad_norm": 0.7569912075996399, "learning_rate": 5.6333093522362276e-05, "loss": 0.0258, "step": 97320 }, { "grad_norm": 0.7721326351165771, "learning_rate": 5.632489267237009e-05, "loss": 0.0323, "step": 97330 }, { "grad_norm": 0.9186615347862244, "learning_rate": 5.63166916494577e-05, "loss": 0.033, "step": 97340 }, { "grad_norm": 0.7639920115470886, "learning_rate": 5.630849045384934e-05, "loss": 0.0267, "step": 97350 }, { "grad_norm": 0.5564056038856506, "learning_rate": 5.630028908576922e-05, "loss": 0.0327, "step": 97360 }, { "grad_norm": 0.6427054405212402, "learning_rate": 5.6292087545441575e-05, "loss": 0.0337, "step": 97370 }, { "grad_norm": 0.4314996898174286, "learning_rate": 5.628388583309061e-05, "loss": 0.0393, "step": 97380 }, { "grad_norm": 0.4814470410346985, "learning_rate": 5.627568394894057e-05, "loss": 0.0357, "step": 97390 }, { "grad_norm": 0.7534058094024658, "learning_rate": 5.626748189321568e-05, "loss": 0.0254, "step": 97400 }, { "grad_norm": 0.2977355718612671, "learning_rate": 5.6259279666140206e-05, "loss": 0.0351, "step": 97410 }, { "grad_norm": 0.47892773151397705, "learning_rate": 5.6251077267938366e-05, "loss": 0.0446, "step": 97420 }, { "grad_norm": 0.36399319767951965, "learning_rate": 5.6242874698834435e-05, "loss": 0.0294, "step": 97430 }, { "grad_norm": 0.5157135725021362, "learning_rate": 5.623467195905264e-05, "loss": 0.0282, "step": 97440 }, { "grad_norm": 0.47865402698516846, "learning_rate": 5.622646904881726e-05, "loss": 0.0478, "step": 97450 }, { "grad_norm": 0.4636740982532501, "learning_rate": 5.6218265968352556e-05, "loss": 0.0295, "step": 97460 }, { "grad_norm": 0.5621232986450195, "learning_rate": 5.621006271788279e-05, "loss": 0.0303, "step": 97470 }, { "grad_norm": 0.23064902424812317, "learning_rate": 5.6201859297632256e-05, "loss": 0.0352, "step": 97480 }, { "grad_norm": 0.6606371998786926, "learning_rate": 5.6193655707825224e-05, "loss": 0.0299, "step": 97490 }, { "grad_norm": 0.5087209939956665, "learning_rate": 5.618545194868595e-05, "loss": 0.0261, "step": 97500 }, { "grad_norm": 0.5411959886550903, "learning_rate": 5.6177248020438755e-05, "loss": 0.0338, "step": 97510 }, { "grad_norm": 0.2501802146434784, "learning_rate": 5.616904392330792e-05, "loss": 0.0394, "step": 97520 }, { "grad_norm": 0.37192094326019287, "learning_rate": 5.616083965751774e-05, "loss": 0.0294, "step": 97530 }, { "grad_norm": 0.5732018947601318, "learning_rate": 5.6152635223292505e-05, "loss": 0.0327, "step": 97540 }, { "grad_norm": 0.4452190101146698, "learning_rate": 5.6144430620856556e-05, "loss": 0.0249, "step": 97550 }, { "grad_norm": 0.31643661856651306, "learning_rate": 5.613622585043418e-05, "loss": 0.0372, "step": 97560 }, { "grad_norm": 0.4849341809749603, "learning_rate": 5.6128020912249676e-05, "loss": 0.0271, "step": 97570 }, { "grad_norm": 0.43558210134506226, "learning_rate": 5.611981580652741e-05, "loss": 0.0276, "step": 97580 }, { "grad_norm": 0.20417919754981995, "learning_rate": 5.6111610533491654e-05, "loss": 0.0306, "step": 97590 }, { "grad_norm": 0.504372775554657, "learning_rate": 5.610340509336677e-05, "loss": 0.0282, "step": 97600 }, { "grad_norm": 0.3219762146472931, "learning_rate": 5.609519948637708e-05, "loss": 0.027, "step": 97610 }, { "grad_norm": 1.1436975002288818, "learning_rate": 5.608699371274694e-05, "loss": 0.0421, "step": 97620 }, { "grad_norm": 1.0195263624191284, "learning_rate": 5.607878777270067e-05, "loss": 0.0272, "step": 97630 }, { "grad_norm": 0.24015243351459503, "learning_rate": 5.607058166646263e-05, "loss": 0.0324, "step": 97640 }, { "grad_norm": 0.7441185712814331, "learning_rate": 5.606237539425716e-05, "loss": 0.0375, "step": 97650 }, { "grad_norm": 0.5439512133598328, "learning_rate": 5.605416895630863e-05, "loss": 0.0392, "step": 97660 }, { "grad_norm": 0.7624025940895081, "learning_rate": 5.604596235284139e-05, "loss": 0.0368, "step": 97670 }, { "grad_norm": 0.3532242774963379, "learning_rate": 5.603775558407982e-05, "loss": 0.05, "step": 97680 }, { "grad_norm": 0.743155300617218, "learning_rate": 5.602954865024826e-05, "loss": 0.033, "step": 97690 }, { "grad_norm": 0.24137143790721893, "learning_rate": 5.6021341551571135e-05, "loss": 0.0296, "step": 97700 }, { "grad_norm": 0.38752424716949463, "learning_rate": 5.601313428827277e-05, "loss": 0.0364, "step": 97710 }, { "grad_norm": 0.2992742359638214, "learning_rate": 5.600492686057759e-05, "loss": 0.0308, "step": 97720 }, { "grad_norm": 0.3173584043979645, "learning_rate": 5.599671926870995e-05, "loss": 0.022, "step": 97730 }, { "grad_norm": 0.7326532602310181, "learning_rate": 5.598851151289427e-05, "loss": 0.0309, "step": 97740 }, { "grad_norm": 0.3422268331050873, "learning_rate": 5.598030359335492e-05, "loss": 0.0435, "step": 97750 }, { "grad_norm": 0.47826841473579407, "learning_rate": 5.5972095510316336e-05, "loss": 0.0242, "step": 97760 }, { "grad_norm": 0.402252197265625, "learning_rate": 5.59638872640029e-05, "loss": 0.0328, "step": 97770 }, { "grad_norm": 0.7990672588348389, "learning_rate": 5.5955678854639015e-05, "loss": 0.0262, "step": 97780 }, { "grad_norm": 0.6808543801307678, "learning_rate": 5.594747028244911e-05, "loss": 0.0391, "step": 97790 }, { "grad_norm": 0.34693780541419983, "learning_rate": 5.5939261547657604e-05, "loss": 0.0269, "step": 97800 }, { "grad_norm": 0.4065653681755066, "learning_rate": 5.5931052650488926e-05, "loss": 0.0355, "step": 97810 }, { "grad_norm": 0.5645700693130493, "learning_rate": 5.5922843591167486e-05, "loss": 0.0533, "step": 97820 }, { "grad_norm": 0.5519161820411682, "learning_rate": 5.591463436991774e-05, "loss": 0.0326, "step": 97830 }, { "grad_norm": 0.5838869214057922, "learning_rate": 5.59064249869641e-05, "loss": 0.0312, "step": 97840 }, { "grad_norm": 0.4514476954936981, "learning_rate": 5.589821544253102e-05, "loss": 0.0307, "step": 97850 }, { "grad_norm": 0.26909422874450684, "learning_rate": 5.5890005736842955e-05, "loss": 0.0294, "step": 97860 }, { "grad_norm": 0.3713429868221283, "learning_rate": 5.588179587012434e-05, "loss": 0.0335, "step": 97870 }, { "grad_norm": 0.46927565336227417, "learning_rate": 5.587358584259964e-05, "loss": 0.0325, "step": 97880 }, { "grad_norm": 0.473599374294281, "learning_rate": 5.586537565449331e-05, "loss": 0.0208, "step": 97890 }, { "grad_norm": 0.4182407259941101, "learning_rate": 5.58571653060298e-05, "loss": 0.0406, "step": 97900 }, { "grad_norm": 0.651139497756958, "learning_rate": 5.5848954797433606e-05, "loss": 0.0333, "step": 97910 }, { "grad_norm": 0.5627975463867188, "learning_rate": 5.584074412892918e-05, "loss": 0.0369, "step": 97920 }, { "grad_norm": 0.38358014822006226, "learning_rate": 5.583253330074101e-05, "loss": 0.0359, "step": 97930 }, { "grad_norm": 0.5723459124565125, "learning_rate": 5.582432231309356e-05, "loss": 0.0287, "step": 97940 }, { "grad_norm": 0.4747147858142853, "learning_rate": 5.5816111166211336e-05, "loss": 0.0427, "step": 97950 }, { "grad_norm": 0.7406452894210815, "learning_rate": 5.5807899860318803e-05, "loss": 0.0363, "step": 97960 }, { "grad_norm": 0.42782285809516907, "learning_rate": 5.579968839564049e-05, "loss": 0.0229, "step": 97970 }, { "grad_norm": 0.7803516387939453, "learning_rate": 5.579147677240085e-05, "loss": 0.0385, "step": 97980 }, { "grad_norm": 0.5031641125679016, "learning_rate": 5.5783264990824434e-05, "loss": 0.026, "step": 97990 }, { "grad_norm": 0.48279574513435364, "learning_rate": 5.57750530511357e-05, "loss": 0.0353, "step": 98000 }, { "grad_norm": 0.5552628636360168, "learning_rate": 5.576684095355922e-05, "loss": 0.0266, "step": 98010 }, { "grad_norm": 0.5807422399520874, "learning_rate": 5.575862869831945e-05, "loss": 0.0453, "step": 98020 }, { "grad_norm": 0.5507750511169434, "learning_rate": 5.575041628564094e-05, "loss": 0.0284, "step": 98030 }, { "grad_norm": 0.496061235666275, "learning_rate": 5.574220371574821e-05, "loss": 0.0302, "step": 98040 }, { "grad_norm": 0.23731033504009247, "learning_rate": 5.573399098886579e-05, "loss": 0.0271, "step": 98050 }, { "grad_norm": 0.4654640555381775, "learning_rate": 5.572577810521821e-05, "loss": 0.0347, "step": 98060 }, { "grad_norm": 0.6437534093856812, "learning_rate": 5.571756506503e-05, "loss": 0.0284, "step": 98070 }, { "grad_norm": 0.555944561958313, "learning_rate": 5.570935186852573e-05, "loss": 0.0351, "step": 98080 }, { "grad_norm": 0.4182700514793396, "learning_rate": 5.570113851592991e-05, "loss": 0.0347, "step": 98090 }, { "grad_norm": 0.575872540473938, "learning_rate": 5.569292500746709e-05, "loss": 0.0306, "step": 98100 }, { "grad_norm": 0.43108952045440674, "learning_rate": 5.568471134336185e-05, "loss": 0.0322, "step": 98110 }, { "grad_norm": 0.2561056911945343, "learning_rate": 5.567649752383876e-05, "loss": 0.0284, "step": 98120 }, { "grad_norm": 0.2799573838710785, "learning_rate": 5.5668283549122336e-05, "loss": 0.0439, "step": 98130 }, { "grad_norm": 0.7223587036132812, "learning_rate": 5.566006941943718e-05, "loss": 0.031, "step": 98140 }, { "grad_norm": 0.3376460075378418, "learning_rate": 5.565185513500785e-05, "loss": 0.0256, "step": 98150 }, { "grad_norm": 0.4076686203479767, "learning_rate": 5.564364069605892e-05, "loss": 0.0544, "step": 98160 }, { "grad_norm": 0.36395037174224854, "learning_rate": 5.563542610281496e-05, "loss": 0.0209, "step": 98170 }, { "grad_norm": 0.3139670491218567, "learning_rate": 5.5627211355500594e-05, "loss": 0.0327, "step": 98180 }, { "grad_norm": 0.5566185712814331, "learning_rate": 5.561899645434037e-05, "loss": 0.0396, "step": 98190 }, { "grad_norm": 0.8665329217910767, "learning_rate": 5.56107813995589e-05, "loss": 0.0325, "step": 98200 }, { "grad_norm": 0.47682860493659973, "learning_rate": 5.560256619138077e-05, "loss": 0.0312, "step": 98210 }, { "grad_norm": 0.40892910957336426, "learning_rate": 5.55943508300306e-05, "loss": 0.0247, "step": 98220 }, { "grad_norm": 1.8944920301437378, "learning_rate": 5.558613531573294e-05, "loss": 0.0392, "step": 98230 }, { "grad_norm": 0.4812558591365814, "learning_rate": 5.5577919648712475e-05, "loss": 0.0349, "step": 98240 }, { "grad_norm": 0.6622310876846313, "learning_rate": 5.556970382919376e-05, "loss": 0.0281, "step": 98250 }, { "grad_norm": 0.3040202856063843, "learning_rate": 5.556148785740146e-05, "loss": 0.0282, "step": 98260 }, { "grad_norm": 0.40724340081214905, "learning_rate": 5.5553271733560154e-05, "loss": 0.0245, "step": 98270 }, { "grad_norm": 0.5924420952796936, "learning_rate": 5.554505545789449e-05, "loss": 0.0334, "step": 98280 }, { "grad_norm": 0.8824102878570557, "learning_rate": 5.5536839030629084e-05, "loss": 0.0295, "step": 98290 }, { "grad_norm": 0.8643279075622559, "learning_rate": 5.55286224519886e-05, "loss": 0.0298, "step": 98300 }, { "grad_norm": 0.6623115539550781, "learning_rate": 5.552040572219763e-05, "loss": 0.0274, "step": 98310 }, { "grad_norm": 0.4339623749256134, "learning_rate": 5.5512188841480864e-05, "loss": 0.0349, "step": 98320 }, { "grad_norm": 0.5422537326812744, "learning_rate": 5.550397181006292e-05, "loss": 0.0294, "step": 98330 }, { "grad_norm": 0.46970340609550476, "learning_rate": 5.549575462816844e-05, "loss": 0.0371, "step": 98340 }, { "grad_norm": 0.6197744607925415, "learning_rate": 5.5487537296022106e-05, "loss": 0.0263, "step": 98350 }, { "grad_norm": 0.6144700050354004, "learning_rate": 5.547931981384855e-05, "loss": 0.0269, "step": 98360 }, { "grad_norm": 0.49539896845817566, "learning_rate": 5.547110218187247e-05, "loss": 0.0343, "step": 98370 }, { "grad_norm": 0.7848888635635376, "learning_rate": 5.5462884400318494e-05, "loss": 0.0346, "step": 98380 }, { "grad_norm": 0.3779284358024597, "learning_rate": 5.545466646941133e-05, "loss": 0.0235, "step": 98390 }, { "grad_norm": 0.5648841261863708, "learning_rate": 5.5446448389375614e-05, "loss": 0.0303, "step": 98400 }, { "grad_norm": 0.7872648239135742, "learning_rate": 5.5438230160436066e-05, "loss": 0.0309, "step": 98410 }, { "grad_norm": 0.2609250247478485, "learning_rate": 5.5430011782817326e-05, "loss": 0.0269, "step": 98420 }, { "grad_norm": 0.45733126997947693, "learning_rate": 5.5421793256744115e-05, "loss": 0.0316, "step": 98430 }, { "grad_norm": 0.6008514165878296, "learning_rate": 5.541357458244111e-05, "loss": 0.0309, "step": 98440 }, { "grad_norm": 2.2206852436065674, "learning_rate": 5.540535576013302e-05, "loss": 0.0223, "step": 98450 }, { "grad_norm": 1.0771840810775757, "learning_rate": 5.539713679004453e-05, "loss": 0.0342, "step": 98460 }, { "grad_norm": 0.5460710525512695, "learning_rate": 5.538891767240034e-05, "loss": 0.0309, "step": 98470 }, { "grad_norm": 0.7297230362892151, "learning_rate": 5.538069840742517e-05, "loss": 0.0294, "step": 98480 }, { "grad_norm": 0.7403075695037842, "learning_rate": 5.5372478995343726e-05, "loss": 0.0445, "step": 98490 }, { "grad_norm": 0.3727034032344818, "learning_rate": 5.536425943638072e-05, "loss": 0.0261, "step": 98500 }, { "grad_norm": 0.4247840344905853, "learning_rate": 5.535603973076089e-05, "loss": 0.0281, "step": 98510 }, { "grad_norm": 0.3567926585674286, "learning_rate": 5.534781987870893e-05, "loss": 0.0254, "step": 98520 }, { "grad_norm": 1.9736695289611816, "learning_rate": 5.533959988044959e-05, "loss": 0.024, "step": 98530 }, { "grad_norm": 0.4050154685974121, "learning_rate": 5.53313797362076e-05, "loss": 0.0276, "step": 98540 }, { "grad_norm": 0.85350102186203, "learning_rate": 5.532315944620769e-05, "loss": 0.0333, "step": 98550 }, { "grad_norm": 0.3479944169521332, "learning_rate": 5.531493901067458e-05, "loss": 0.0259, "step": 98560 }, { "grad_norm": 0.40536758303642273, "learning_rate": 5.530671842983305e-05, "loss": 0.0284, "step": 98570 }, { "grad_norm": 0.7098545432090759, "learning_rate": 5.529849770390785e-05, "loss": 0.0363, "step": 98580 }, { "grad_norm": 0.4504464864730835, "learning_rate": 5.5290276833123686e-05, "loss": 0.0316, "step": 98590 }, { "grad_norm": 0.829785168170929, "learning_rate": 5.528205581770536e-05, "loss": 0.0197, "step": 98600 }, { "grad_norm": 0.5994556546211243, "learning_rate": 5.527383465787759e-05, "loss": 0.0268, "step": 98610 }, { "grad_norm": 0.26177042722702026, "learning_rate": 5.526561335386519e-05, "loss": 0.0264, "step": 98620 }, { "grad_norm": 0.49420586228370667, "learning_rate": 5.525739190589287e-05, "loss": 0.0353, "step": 98630 }, { "grad_norm": 0.4949207603931427, "learning_rate": 5.524917031418546e-05, "loss": 0.034, "step": 98640 }, { "grad_norm": 1.2176916599273682, "learning_rate": 5.524094857896769e-05, "loss": 0.0302, "step": 98650 }, { "grad_norm": 0.5316818952560425, "learning_rate": 5.523272670046435e-05, "loss": 0.0347, "step": 98660 }, { "grad_norm": 0.5056130886077881, "learning_rate": 5.522450467890024e-05, "loss": 0.034, "step": 98670 }, { "grad_norm": 0.4737711250782013, "learning_rate": 5.521628251450012e-05, "loss": 0.0291, "step": 98680 }, { "grad_norm": 0.805754542350769, "learning_rate": 5.52080602074888e-05, "loss": 0.0364, "step": 98690 }, { "grad_norm": 0.41569414734840393, "learning_rate": 5.519983775809108e-05, "loss": 0.0364, "step": 98700 }, { "grad_norm": 0.7887876629829407, "learning_rate": 5.519161516653174e-05, "loss": 0.0376, "step": 98710 }, { "grad_norm": 0.562975287437439, "learning_rate": 5.518339243303561e-05, "loss": 0.0247, "step": 98720 }, { "grad_norm": 0.9220061302185059, "learning_rate": 5.5175169557827454e-05, "loss": 0.0314, "step": 98730 }, { "grad_norm": 0.3858847916126251, "learning_rate": 5.5166946541132116e-05, "loss": 0.0271, "step": 98740 }, { "grad_norm": 0.7329385876655579, "learning_rate": 5.5158723383174396e-05, "loss": 0.0358, "step": 98750 }, { "grad_norm": 0.5569372177124023, "learning_rate": 5.5150500084179124e-05, "loss": 0.0352, "step": 98760 }, { "grad_norm": 0.5676229596138, "learning_rate": 5.5142276644371116e-05, "loss": 0.0274, "step": 98770 }, { "grad_norm": 0.4875623285770416, "learning_rate": 5.513405306397519e-05, "loss": 0.0465, "step": 98780 }, { "grad_norm": 0.4936395287513733, "learning_rate": 5.5125829343216184e-05, "loss": 0.0284, "step": 98790 }, { "grad_norm": 0.5939506888389587, "learning_rate": 5.5117605482318927e-05, "loss": 0.0324, "step": 98800 }, { "grad_norm": 0.4642618000507355, "learning_rate": 5.510938148150826e-05, "loss": 0.0221, "step": 98810 }, { "grad_norm": 0.7027327418327332, "learning_rate": 5.510115734100902e-05, "loss": 0.0252, "step": 98820 }, { "grad_norm": 0.6032017469406128, "learning_rate": 5.509293306104607e-05, "loss": 0.0272, "step": 98830 }, { "grad_norm": 0.5465788245201111, "learning_rate": 5.508470864184423e-05, "loss": 0.0245, "step": 98840 }, { "grad_norm": 0.8284013867378235, "learning_rate": 5.507648408362838e-05, "loss": 0.0355, "step": 98850 }, { "grad_norm": 0.7488915920257568, "learning_rate": 5.506825938662335e-05, "loss": 0.0298, "step": 98860 }, { "grad_norm": 0.355071485042572, "learning_rate": 5.5060034551054e-05, "loss": 0.0295, "step": 98870 }, { "grad_norm": 0.5456687808036804, "learning_rate": 5.5051809577145216e-05, "loss": 0.0288, "step": 98880 }, { "grad_norm": 0.5403125286102295, "learning_rate": 5.5043584465121863e-05, "loss": 0.0423, "step": 98890 }, { "grad_norm": 0.3306410312652588, "learning_rate": 5.5035359215208805e-05, "loss": 0.0245, "step": 98900 }, { "grad_norm": 0.6373453736305237, "learning_rate": 5.5027133827630916e-05, "loss": 0.0253, "step": 98910 }, { "grad_norm": 0.5469528436660767, "learning_rate": 5.501890830261307e-05, "loss": 0.0329, "step": 98920 }, { "grad_norm": 0.45177900791168213, "learning_rate": 5.501068264038016e-05, "loss": 0.0346, "step": 98930 }, { "grad_norm": 0.30723336338996887, "learning_rate": 5.5002456841157066e-05, "loss": 0.029, "step": 98940 }, { "grad_norm": 0.6057271361351013, "learning_rate": 5.49942309051687e-05, "loss": 0.0293, "step": 98950 }, { "grad_norm": 0.7285621166229248, "learning_rate": 5.4986004832639915e-05, "loss": 0.0294, "step": 98960 }, { "grad_norm": 0.47374066710472107, "learning_rate": 5.497777862379565e-05, "loss": 0.0331, "step": 98970 }, { "grad_norm": 0.4973074197769165, "learning_rate": 5.496955227886078e-05, "loss": 0.0252, "step": 98980 }, { "grad_norm": 0.3439911901950836, "learning_rate": 5.4961325798060206e-05, "loss": 0.0361, "step": 98990 }, { "grad_norm": 0.6697579026222229, "learning_rate": 5.495309918161885e-05, "loss": 0.0411, "step": 99000 }, { "grad_norm": 1.1195522546768188, "learning_rate": 5.494487242976164e-05, "loss": 0.0293, "step": 99010 }, { "grad_norm": 1.0196559429168701, "learning_rate": 5.493664554271347e-05, "loss": 0.037, "step": 99020 }, { "grad_norm": 0.8217356204986572, "learning_rate": 5.492841852069926e-05, "loss": 0.0327, "step": 99030 }, { "grad_norm": 0.41342902183532715, "learning_rate": 5.492019136394394e-05, "loss": 0.0348, "step": 99040 }, { "grad_norm": 0.413683146238327, "learning_rate": 5.491196407267244e-05, "loss": 0.0367, "step": 99050 }, { "grad_norm": 0.5646896362304688, "learning_rate": 5.4903736647109684e-05, "loss": 0.032, "step": 99060 }, { "grad_norm": 0.44807344675064087, "learning_rate": 5.48955090874806e-05, "loss": 0.0238, "step": 99070 }, { "grad_norm": 0.4989403486251831, "learning_rate": 5.488728139401016e-05, "loss": 0.0268, "step": 99080 }, { "grad_norm": 0.44538742303848267, "learning_rate": 5.4879053566923264e-05, "loss": 0.0207, "step": 99090 }, { "grad_norm": 0.34237396717071533, "learning_rate": 5.4870825606444885e-05, "loss": 0.0426, "step": 99100 }, { "grad_norm": 1.1371984481811523, "learning_rate": 5.4862597512799964e-05, "loss": 0.031, "step": 99110 }, { "grad_norm": 0.6841449737548828, "learning_rate": 5.4854369286213446e-05, "loss": 0.0275, "step": 99120 }, { "grad_norm": 0.8847923278808594, "learning_rate": 5.4846140926910295e-05, "loss": 0.0318, "step": 99130 }, { "grad_norm": 0.8726252913475037, "learning_rate": 5.483791243511549e-05, "loss": 0.0502, "step": 99140 }, { "grad_norm": 0.5615969300270081, "learning_rate": 5.482968381105396e-05, "loss": 0.0405, "step": 99150 }, { "grad_norm": 0.7514541745185852, "learning_rate": 5.48214550549507e-05, "loss": 0.0262, "step": 99160 }, { "grad_norm": 0.6283766031265259, "learning_rate": 5.4813226167030665e-05, "loss": 0.0422, "step": 99170 }, { "grad_norm": 0.5210540294647217, "learning_rate": 5.480499714751883e-05, "loss": 0.0418, "step": 99180 }, { "grad_norm": 0.40636390447616577, "learning_rate": 5.4796767996640175e-05, "loss": 0.0333, "step": 99190 }, { "grad_norm": 0.9820419549942017, "learning_rate": 5.47885387146197e-05, "loss": 0.0326, "step": 99200 }, { "grad_norm": 0.49363285303115845, "learning_rate": 5.4780309301682364e-05, "loss": 0.027, "step": 99210 }, { "grad_norm": 0.9255826473236084, "learning_rate": 5.477207975805317e-05, "loss": 0.04, "step": 99220 }, { "grad_norm": 0.40679433941841125, "learning_rate": 5.4763850083957115e-05, "loss": 0.0352, "step": 99230 }, { "grad_norm": 0.5113542079925537, "learning_rate": 5.475562027961918e-05, "loss": 0.025, "step": 99240 }, { "grad_norm": 0.34924978017807007, "learning_rate": 5.474739034526437e-05, "loss": 0.0314, "step": 99250 }, { "grad_norm": 0.5603371262550354, "learning_rate": 5.473916028111771e-05, "loss": 0.0377, "step": 99260 }, { "grad_norm": 0.7901656031608582, "learning_rate": 5.473093008740416e-05, "loss": 0.0263, "step": 99270 }, { "grad_norm": 0.769119143486023, "learning_rate": 5.472269976434877e-05, "loss": 0.0315, "step": 99280 }, { "grad_norm": 0.465717077255249, "learning_rate": 5.471446931217656e-05, "loss": 0.044, "step": 99290 }, { "grad_norm": 0.7616139054298401, "learning_rate": 5.4706238731112516e-05, "loss": 0.0255, "step": 99300 }, { "grad_norm": 0.551078736782074, "learning_rate": 5.4698008021381665e-05, "loss": 0.032, "step": 99310 }, { "grad_norm": 0.32453978061676025, "learning_rate": 5.468977718320905e-05, "loss": 0.0367, "step": 99320 }, { "grad_norm": 0.7093350291252136, "learning_rate": 5.468154621681969e-05, "loss": 0.0321, "step": 99330 }, { "grad_norm": 0.5233401656150818, "learning_rate": 5.4673315122438604e-05, "loss": 0.0405, "step": 99340 }, { "grad_norm": 1.0642000436782837, "learning_rate": 5.466508390029086e-05, "loss": 0.0296, "step": 99350 }, { "grad_norm": 0.3567197620868683, "learning_rate": 5.465685255060146e-05, "loss": 0.0306, "step": 99360 }, { "grad_norm": 0.6736899018287659, "learning_rate": 5.464862107359546e-05, "loss": 0.033, "step": 99370 }, { "grad_norm": 1.5298079252243042, "learning_rate": 5.46403894694979e-05, "loss": 0.0297, "step": 99380 }, { "grad_norm": 0.4329909682273865, "learning_rate": 5.463215773853385e-05, "loss": 0.0388, "step": 99390 }, { "grad_norm": 0.4481748342514038, "learning_rate": 5.462392588092834e-05, "loss": 0.0305, "step": 99400 }, { "grad_norm": 0.9970000982284546, "learning_rate": 5.461569389690644e-05, "loss": 0.028, "step": 99410 }, { "grad_norm": 1.0269917249679565, "learning_rate": 5.4607461786693195e-05, "loss": 0.026, "step": 99420 }, { "grad_norm": 0.6269853711128235, "learning_rate": 5.4599229550513685e-05, "loss": 0.0301, "step": 99430 }, { "grad_norm": 0.47505322098731995, "learning_rate": 5.4590997188592954e-05, "loss": 0.0451, "step": 99440 }, { "grad_norm": 0.9144106507301331, "learning_rate": 5.4582764701156096e-05, "loss": 0.0471, "step": 99450 }, { "grad_norm": 1.4314719438552856, "learning_rate": 5.4574532088428164e-05, "loss": 0.0214, "step": 99460 }, { "grad_norm": 1.1766020059585571, "learning_rate": 5.456629935063425e-05, "loss": 0.0451, "step": 99470 }, { "grad_norm": 0.3813842535018921, "learning_rate": 5.455806648799943e-05, "loss": 0.0297, "step": 99480 }, { "grad_norm": 1.0084044933319092, "learning_rate": 5.454983350074877e-05, "loss": 0.0397, "step": 99490 }, { "grad_norm": 0.5254133343696594, "learning_rate": 5.454160038910737e-05, "loss": 0.0463, "step": 99500 }, { "grad_norm": 0.6431543231010437, "learning_rate": 5.453336715330034e-05, "loss": 0.0324, "step": 99510 }, { "grad_norm": 0.7360857129096985, "learning_rate": 5.452513379355275e-05, "loss": 0.0326, "step": 99520 }, { "grad_norm": 0.4276816248893738, "learning_rate": 5.451690031008969e-05, "loss": 0.0241, "step": 99530 }, { "grad_norm": 0.5347890257835388, "learning_rate": 5.450866670313629e-05, "loss": 0.038, "step": 99540 }, { "grad_norm": 0.859916090965271, "learning_rate": 5.450043297291763e-05, "loss": 0.0253, "step": 99550 }, { "grad_norm": 0.33791613578796387, "learning_rate": 5.4492199119658824e-05, "loss": 0.0252, "step": 99560 }, { "grad_norm": 0.4538193643093109, "learning_rate": 5.448396514358497e-05, "loss": 0.0244, "step": 99570 }, { "grad_norm": 0.5247684121131897, "learning_rate": 5.447573104492121e-05, "loss": 0.0318, "step": 99580 }, { "grad_norm": 1.0923070907592773, "learning_rate": 5.4467496823892626e-05, "loss": 0.0381, "step": 99590 }, { "grad_norm": 0.3277921676635742, "learning_rate": 5.445926248072438e-05, "loss": 0.0274, "step": 99600 }, { "grad_norm": 0.6288064122200012, "learning_rate": 5.445102801564156e-05, "loss": 0.0487, "step": 99610 }, { "grad_norm": 0.5682016015052795, "learning_rate": 5.444279342886931e-05, "loss": 0.0284, "step": 99620 }, { "grad_norm": 0.5701287388801575, "learning_rate": 5.443455872063274e-05, "loss": 0.0291, "step": 99630 }, { "grad_norm": 0.46336784958839417, "learning_rate": 5.442632389115703e-05, "loss": 0.0309, "step": 99640 }, { "grad_norm": 0.44368621706962585, "learning_rate": 5.441808894066727e-05, "loss": 0.0256, "step": 99650 }, { "grad_norm": 0.5016841888427734, "learning_rate": 5.4409853869388614e-05, "loss": 0.0298, "step": 99660 }, { "grad_norm": 0.48958921432495117, "learning_rate": 5.4401618677546215e-05, "loss": 0.0228, "step": 99670 }, { "grad_norm": 0.3721274435520172, "learning_rate": 5.439338336536522e-05, "loss": 0.0333, "step": 99680 }, { "grad_norm": 0.5461316108703613, "learning_rate": 5.438514793307077e-05, "loss": 0.0446, "step": 99690 }, { "grad_norm": 0.4393020272254944, "learning_rate": 5.437691238088802e-05, "loss": 0.0295, "step": 99700 }, { "grad_norm": 0.5876752138137817, "learning_rate": 5.436867670904213e-05, "loss": 0.0294, "step": 99710 }, { "grad_norm": 0.38955235481262207, "learning_rate": 5.436044091775827e-05, "loss": 0.0311, "step": 99720 }, { "grad_norm": 0.4453369081020355, "learning_rate": 5.435220500726159e-05, "loss": 0.0326, "step": 99730 }, { "grad_norm": 0.5272635221481323, "learning_rate": 5.4343968977777246e-05, "loss": 0.0244, "step": 99740 }, { "grad_norm": 0.464298278093338, "learning_rate": 5.433573282953045e-05, "loss": 0.0256, "step": 99750 }, { "grad_norm": 0.7756121754646301, "learning_rate": 5.432749656274633e-05, "loss": 0.0338, "step": 99760 }, { "grad_norm": 1.0495953559875488, "learning_rate": 5.431926017765007e-05, "loss": 0.0507, "step": 99770 }, { "grad_norm": 0.38006892800331116, "learning_rate": 5.431102367446686e-05, "loss": 0.0421, "step": 99780 }, { "grad_norm": 0.38949164748191833, "learning_rate": 5.430278705342191e-05, "loss": 0.0328, "step": 99790 }, { "grad_norm": 0.7261487245559692, "learning_rate": 5.429455031474036e-05, "loss": 0.037, "step": 99800 }, { "grad_norm": 0.3270784020423889, "learning_rate": 5.428631345864742e-05, "loss": 0.0226, "step": 99810 }, { "grad_norm": 0.5695765614509583, "learning_rate": 5.4278076485368255e-05, "loss": 0.0261, "step": 99820 }, { "grad_norm": 0.7434101104736328, "learning_rate": 5.426983939512812e-05, "loss": 0.0355, "step": 99830 }, { "grad_norm": 0.3166317641735077, "learning_rate": 5.4261602188152175e-05, "loss": 0.0443, "step": 99840 }, { "grad_norm": 0.5688899755477905, "learning_rate": 5.4253364864665625e-05, "loss": 0.0318, "step": 99850 }, { "grad_norm": 1.0651979446411133, "learning_rate": 5.424512742489367e-05, "loss": 0.0256, "step": 99860 }, { "grad_norm": 0.4752435088157654, "learning_rate": 5.4236889869061536e-05, "loss": 0.0325, "step": 99870 }, { "grad_norm": 1.1882244348526, "learning_rate": 5.4228652197394415e-05, "loss": 0.0349, "step": 99880 }, { "grad_norm": 0.5061211585998535, "learning_rate": 5.4220414410117536e-05, "loss": 0.0453, "step": 99890 }, { "grad_norm": 0.6679757833480835, "learning_rate": 5.42121765074561e-05, "loss": 0.0482, "step": 99900 }, { "grad_norm": 0.5573995113372803, "learning_rate": 5.4203938489635366e-05, "loss": 0.0302, "step": 99910 }, { "grad_norm": 0.5463385581970215, "learning_rate": 5.419570035688052e-05, "loss": 0.0334, "step": 99920 }, { "grad_norm": 0.4597403407096863, "learning_rate": 5.418746210941681e-05, "loss": 0.0327, "step": 99930 }, { "grad_norm": 1.0360522270202637, "learning_rate": 5.417922374746944e-05, "loss": 0.0381, "step": 99940 }, { "grad_norm": 0.40096649527549744, "learning_rate": 5.4170985271263676e-05, "loss": 0.0382, "step": 99950 }, { "grad_norm": 0.6078590750694275, "learning_rate": 5.416274668102474e-05, "loss": 0.0264, "step": 99960 }, { "grad_norm": 1.1088635921478271, "learning_rate": 5.4154507976977875e-05, "loss": 0.0348, "step": 99970 }, { "grad_norm": 0.5336252450942993, "learning_rate": 5.414626915934832e-05, "loss": 0.0275, "step": 99980 }, { "grad_norm": 0.47380033135414124, "learning_rate": 5.413803022836133e-05, "loss": 0.029, "step": 99990 }, { "grad_norm": 0.4581809639930725, "learning_rate": 5.412979118424214e-05, "loss": 0.0461, "step": 100000 }, { "grad_norm": 0.4541448950767517, "learning_rate": 5.412155202721602e-05, "loss": 0.0267, "step": 100010 }, { "grad_norm": 0.8704745769500732, "learning_rate": 5.4113312757508194e-05, "loss": 0.0367, "step": 100020 }, { "grad_norm": 0.8133975863456726, "learning_rate": 5.410507337534396e-05, "loss": 0.023, "step": 100030 }, { "grad_norm": 0.8646888136863708, "learning_rate": 5.409683388094857e-05, "loss": 0.0219, "step": 100040 }, { "grad_norm": 0.4522603750228882, "learning_rate": 5.4088594274547264e-05, "loss": 0.0283, "step": 100050 }, { "grad_norm": 0.39180755615234375, "learning_rate": 5.4080354556365344e-05, "loss": 0.0309, "step": 100060 }, { "grad_norm": 0.6793635487556458, "learning_rate": 5.407211472662805e-05, "loss": 0.0348, "step": 100070 }, { "grad_norm": 0.6107927560806274, "learning_rate": 5.406387478556067e-05, "loss": 0.0367, "step": 100080 }, { "grad_norm": 0.6999256610870361, "learning_rate": 5.405563473338847e-05, "loss": 0.0475, "step": 100090 }, { "grad_norm": 0.6388632655143738, "learning_rate": 5.4047394570336764e-05, "loss": 0.0276, "step": 100100 }, { "grad_norm": 0.6271086931228638, "learning_rate": 5.40391542966308e-05, "loss": 0.0311, "step": 100110 }, { "grad_norm": 0.6079747676849365, "learning_rate": 5.403091391249588e-05, "loss": 0.0299, "step": 100120 }, { "grad_norm": 0.46736419200897217, "learning_rate": 5.4022673418157276e-05, "loss": 0.0216, "step": 100130 }, { "grad_norm": 0.7592297196388245, "learning_rate": 5.40144328138403e-05, "loss": 0.029, "step": 100140 }, { "grad_norm": 0.6810134649276733, "learning_rate": 5.4006192099770225e-05, "loss": 0.0277, "step": 100150 }, { "grad_norm": 0.45070865750312805, "learning_rate": 5.399795127617239e-05, "loss": 0.0337, "step": 100160 }, { "grad_norm": 0.5007559657096863, "learning_rate": 5.398971034327205e-05, "loss": 0.0263, "step": 100170 }, { "grad_norm": 0.48654904961586, "learning_rate": 5.398146930129454e-05, "loss": 0.036, "step": 100180 }, { "grad_norm": 0.5916439890861511, "learning_rate": 5.397322815046515e-05, "loss": 0.0244, "step": 100190 }, { "grad_norm": 0.6849799752235413, "learning_rate": 5.396498689100918e-05, "loss": 0.0287, "step": 100200 }, { "grad_norm": 0.41148266196250916, "learning_rate": 5.395674552315198e-05, "loss": 0.0284, "step": 100210 }, { "grad_norm": 0.3747873604297638, "learning_rate": 5.394850404711883e-05, "loss": 0.0246, "step": 100220 }, { "grad_norm": 0.35403573513031006, "learning_rate": 5.394026246313506e-05, "loss": 0.0452, "step": 100230 }, { "grad_norm": 0.3371303081512451, "learning_rate": 5.3932020771426006e-05, "loss": 0.0258, "step": 100240 }, { "grad_norm": 0.5353532433509827, "learning_rate": 5.392377897221699e-05, "loss": 0.0269, "step": 100250 }, { "grad_norm": 0.44032716751098633, "learning_rate": 5.391553706573331e-05, "loss": 0.0322, "step": 100260 }, { "grad_norm": 0.779090404510498, "learning_rate": 5.390729505220032e-05, "loss": 0.0282, "step": 100270 }, { "grad_norm": 0.48477891087532043, "learning_rate": 5.389905293184335e-05, "loss": 0.0266, "step": 100280 }, { "grad_norm": 0.6728436946868896, "learning_rate": 5.389081070488775e-05, "loss": 0.029, "step": 100290 }, { "grad_norm": 0.4779417812824249, "learning_rate": 5.388256837155884e-05, "loss": 0.0285, "step": 100300 }, { "grad_norm": 0.7063356041908264, "learning_rate": 5.387432593208198e-05, "loss": 0.0318, "step": 100310 }, { "grad_norm": 0.5566504001617432, "learning_rate": 5.3866083386682496e-05, "loss": 0.0315, "step": 100320 }, { "grad_norm": 0.5728977918624878, "learning_rate": 5.385784073558574e-05, "loss": 0.0309, "step": 100330 }, { "grad_norm": 0.43749579787254333, "learning_rate": 5.3849597979017064e-05, "loss": 0.0224, "step": 100340 }, { "grad_norm": 0.3780371844768524, "learning_rate": 5.384135511720184e-05, "loss": 0.0296, "step": 100350 }, { "grad_norm": 0.9571576714515686, "learning_rate": 5.383311215036541e-05, "loss": 0.0418, "step": 100360 }, { "grad_norm": 0.7447491884231567, "learning_rate": 5.382486907873313e-05, "loss": 0.0284, "step": 100370 }, { "grad_norm": 0.45570605993270874, "learning_rate": 5.381662590253036e-05, "loss": 0.0404, "step": 100380 }, { "grad_norm": 0.39141932129859924, "learning_rate": 5.3808382621982475e-05, "loss": 0.0514, "step": 100390 }, { "grad_norm": 1.0628316402435303, "learning_rate": 5.380013923731483e-05, "loss": 0.0266, "step": 100400 }, { "grad_norm": 0.6748120784759521, "learning_rate": 5.379189574875282e-05, "loss": 0.036, "step": 100410 }, { "grad_norm": 0.4386484622955322, "learning_rate": 5.37836521565218e-05, "loss": 0.0264, "step": 100420 }, { "grad_norm": 0.7098807692527771, "learning_rate": 5.377540846084716e-05, "loss": 0.0281, "step": 100430 }, { "grad_norm": 0.4437408447265625, "learning_rate": 5.376716466195426e-05, "loss": 0.0413, "step": 100440 }, { "grad_norm": 0.28031450510025024, "learning_rate": 5.37589207600685e-05, "loss": 0.0289, "step": 100450 }, { "grad_norm": 0.4615088701248169, "learning_rate": 5.375067675541524e-05, "loss": 0.0268, "step": 100460 }, { "grad_norm": 0.2764362096786499, "learning_rate": 5.374243264821991e-05, "loss": 0.0229, "step": 100470 }, { "grad_norm": 0.3637422025203705, "learning_rate": 5.3734188438707866e-05, "loss": 0.0274, "step": 100480 }, { "grad_norm": 0.9281716346740723, "learning_rate": 5.3725944127104524e-05, "loss": 0.0252, "step": 100490 }, { "grad_norm": 0.3893531858921051, "learning_rate": 5.371769971363526e-05, "loss": 0.0254, "step": 100500 }, { "grad_norm": 0.4767635464668274, "learning_rate": 5.370945519852548e-05, "loss": 0.0228, "step": 100510 }, { "grad_norm": 0.4589698612689972, "learning_rate": 5.3701210582000594e-05, "loss": 0.0278, "step": 100520 }, { "grad_norm": 0.4981442987918854, "learning_rate": 5.369296586428599e-05, "loss": 0.0199, "step": 100530 }, { "grad_norm": 0.44503912329673767, "learning_rate": 5.368472104560711e-05, "loss": 0.0404, "step": 100540 }, { "grad_norm": 0.8133143782615662, "learning_rate": 5.367647612618932e-05, "loss": 0.0398, "step": 100550 }, { "grad_norm": 0.42564108967781067, "learning_rate": 5.366823110625807e-05, "loss": 0.0354, "step": 100560 }, { "grad_norm": 0.7839870452880859, "learning_rate": 5.365998598603874e-05, "loss": 0.0278, "step": 100570 }, { "grad_norm": 0.4536752998828888, "learning_rate": 5.36517407657568e-05, "loss": 0.0338, "step": 100580 }, { "grad_norm": 0.4075669050216675, "learning_rate": 5.3643495445637595e-05, "loss": 0.0303, "step": 100590 }, { "grad_norm": 0.447527140378952, "learning_rate": 5.3635250025906626e-05, "loss": 0.032, "step": 100600 }, { "grad_norm": 0.6132793426513672, "learning_rate": 5.362700450678927e-05, "loss": 0.0318, "step": 100610 }, { "grad_norm": 0.517399251461029, "learning_rate": 5.3618758888510986e-05, "loss": 0.0281, "step": 100620 }, { "grad_norm": 0.35735395550727844, "learning_rate": 5.3610513171297194e-05, "loss": 0.0365, "step": 100630 }, { "grad_norm": 0.5340665578842163, "learning_rate": 5.360226735537333e-05, "loss": 0.0237, "step": 100640 }, { "grad_norm": 0.7106504440307617, "learning_rate": 5.35940214409648e-05, "loss": 0.0245, "step": 100650 }, { "grad_norm": 0.34332436323165894, "learning_rate": 5.35857754282971e-05, "loss": 0.0384, "step": 100660 }, { "grad_norm": 0.6340485215187073, "learning_rate": 5.357752931759565e-05, "loss": 0.0372, "step": 100670 }, { "grad_norm": 0.6511096954345703, "learning_rate": 5.3569283109085886e-05, "loss": 0.0359, "step": 100680 }, { "grad_norm": 0.7754327058792114, "learning_rate": 5.356103680299327e-05, "loss": 0.0244, "step": 100690 }, { "grad_norm": 0.3830372095108032, "learning_rate": 5.355279039954323e-05, "loss": 0.0296, "step": 100700 }, { "grad_norm": 0.507595419883728, "learning_rate": 5.3544543898961244e-05, "loss": 0.0303, "step": 100710 }, { "grad_norm": 0.2995907664299011, "learning_rate": 5.353629730147276e-05, "loss": 0.0257, "step": 100720 }, { "grad_norm": 0.42878395318984985, "learning_rate": 5.352805060730324e-05, "loss": 0.0303, "step": 100730 }, { "grad_norm": 0.9326339364051819, "learning_rate": 5.351980381667813e-05, "loss": 0.0203, "step": 100740 }, { "grad_norm": 1.591748833656311, "learning_rate": 5.3511556929822914e-05, "loss": 0.0334, "step": 100750 }, { "grad_norm": 0.4573248624801636, "learning_rate": 5.350330994696305e-05, "loss": 0.0287, "step": 100760 }, { "grad_norm": 2.8782386779785156, "learning_rate": 5.349506286832402e-05, "loss": 0.0384, "step": 100770 }, { "grad_norm": 0.9256593585014343, "learning_rate": 5.3486815694131254e-05, "loss": 0.0372, "step": 100780 }, { "grad_norm": 0.4240419268608093, "learning_rate": 5.347856842461028e-05, "loss": 0.0228, "step": 100790 }, { "grad_norm": 1.0014184713363647, "learning_rate": 5.347032105998654e-05, "loss": 0.0395, "step": 100800 }, { "grad_norm": 0.43814998865127563, "learning_rate": 5.346207360048554e-05, "loss": 0.0426, "step": 100810 }, { "grad_norm": 0.5694029331207275, "learning_rate": 5.345382604633275e-05, "loss": 0.0282, "step": 100820 }, { "grad_norm": 0.44970789551734924, "learning_rate": 5.344557839775365e-05, "loss": 0.0198, "step": 100830 }, { "grad_norm": 0.6777507066726685, "learning_rate": 5.343733065497371e-05, "loss": 0.0226, "step": 100840 }, { "grad_norm": 0.6091100573539734, "learning_rate": 5.3429082818218456e-05, "loss": 0.0252, "step": 100850 }, { "grad_norm": 0.7846294045448303, "learning_rate": 5.3420834887713366e-05, "loss": 0.0325, "step": 100860 }, { "grad_norm": 0.3541819155216217, "learning_rate": 5.341258686368395e-05, "loss": 0.0356, "step": 100870 }, { "grad_norm": 0.36946746706962585, "learning_rate": 5.340433874635568e-05, "loss": 0.0302, "step": 100880 }, { "grad_norm": 0.40066781640052795, "learning_rate": 5.339609053595407e-05, "loss": 0.038, "step": 100890 }, { "grad_norm": 0.327677845954895, "learning_rate": 5.338784223270461e-05, "loss": 0.0265, "step": 100900 }, { "grad_norm": 0.4732741117477417, "learning_rate": 5.337959383683282e-05, "loss": 0.0405, "step": 100910 }, { "grad_norm": 0.46563223004341125, "learning_rate": 5.337134534856419e-05, "loss": 0.0288, "step": 100920 }, { "grad_norm": 0.3571290075778961, "learning_rate": 5.336309676812428e-05, "loss": 0.0226, "step": 100930 }, { "grad_norm": 0.3460666239261627, "learning_rate": 5.335484809573854e-05, "loss": 0.0313, "step": 100940 }, { "grad_norm": 0.6086898446083069, "learning_rate": 5.334659933163252e-05, "loss": 0.0263, "step": 100950 }, { "grad_norm": 0.3894731402397156, "learning_rate": 5.333835047603174e-05, "loss": 0.0225, "step": 100960 }, { "grad_norm": 0.26544004678726196, "learning_rate": 5.3330101529161705e-05, "loss": 0.0222, "step": 100970 }, { "grad_norm": 0.4489273428916931, "learning_rate": 5.3321852491247936e-05, "loss": 0.0303, "step": 100980 }, { "grad_norm": 0.8043726682662964, "learning_rate": 5.331360336251597e-05, "loss": 0.0437, "step": 100990 }, { "grad_norm": 0.33117732405662537, "learning_rate": 5.3305354143191354e-05, "loss": 0.0276, "step": 101000 }, { "grad_norm": 0.7665808796882629, "learning_rate": 5.329710483349959e-05, "loss": 0.0309, "step": 101010 }, { "grad_norm": 0.7130267024040222, "learning_rate": 5.328885543366622e-05, "loss": 0.0366, "step": 101020 }, { "grad_norm": 0.4443819522857666, "learning_rate": 5.328060594391675e-05, "loss": 0.0417, "step": 101030 }, { "grad_norm": 0.9213380813598633, "learning_rate": 5.3272356364476774e-05, "loss": 0.0448, "step": 101040 }, { "grad_norm": 0.6960297226905823, "learning_rate": 5.3264106695571795e-05, "loss": 0.0274, "step": 101050 }, { "grad_norm": 0.36837393045425415, "learning_rate": 5.3255856937427374e-05, "loss": 0.0297, "step": 101060 }, { "grad_norm": 0.3782931864261627, "learning_rate": 5.3247607090269034e-05, "loss": 0.0317, "step": 101070 }, { "grad_norm": 1.8969430923461914, "learning_rate": 5.3239357154322355e-05, "loss": 0.0464, "step": 101080 }, { "grad_norm": 0.5783740878105164, "learning_rate": 5.323110712981285e-05, "loss": 0.0259, "step": 101090 }, { "grad_norm": 0.3367588222026825, "learning_rate": 5.32228570169661e-05, "loss": 0.0291, "step": 101100 }, { "grad_norm": 0.6569714546203613, "learning_rate": 5.321460681600764e-05, "loss": 0.0343, "step": 101110 }, { "grad_norm": 1.0833598375320435, "learning_rate": 5.320635652716305e-05, "loss": 0.0265, "step": 101120 }, { "grad_norm": 0.3659898042678833, "learning_rate": 5.3198106150657865e-05, "loss": 0.0285, "step": 101130 }, { "grad_norm": 0.3799862265586853, "learning_rate": 5.318985568671767e-05, "loss": 0.0351, "step": 101140 }, { "grad_norm": 0.4986409842967987, "learning_rate": 5.3181605135568016e-05, "loss": 0.0247, "step": 101150 }, { "grad_norm": 0.5077848434448242, "learning_rate": 5.3173354497434466e-05, "loss": 0.0279, "step": 101160 }, { "grad_norm": 0.5329809784889221, "learning_rate": 5.31651037725426e-05, "loss": 0.0392, "step": 101170 }, { "grad_norm": 0.6545279622077942, "learning_rate": 5.315685296111799e-05, "loss": 0.03, "step": 101180 }, { "grad_norm": 0.5566681623458862, "learning_rate": 5.31486020633862e-05, "loss": 0.029, "step": 101190 }, { "grad_norm": 0.37400248646736145, "learning_rate": 5.314035107957281e-05, "loss": 0.0275, "step": 101200 }, { "grad_norm": 0.18373142182826996, "learning_rate": 5.313210000990342e-05, "loss": 0.0234, "step": 101210 }, { "grad_norm": 0.3985446095466614, "learning_rate": 5.312384885460358e-05, "loss": 0.0215, "step": 101220 }, { "grad_norm": 0.5786285400390625, "learning_rate": 5.3115597613898874e-05, "loss": 0.0281, "step": 101230 }, { "grad_norm": 0.27760350704193115, "learning_rate": 5.3107346288014903e-05, "loss": 0.0337, "step": 101240 }, { "grad_norm": 0.3191697597503662, "learning_rate": 5.309909487717727e-05, "loss": 0.0223, "step": 101250 }, { "grad_norm": 0.45840150117874146, "learning_rate": 5.309084338161153e-05, "loss": 0.049, "step": 101260 }, { "grad_norm": 0.5580337047576904, "learning_rate": 5.3082591801543315e-05, "loss": 0.0372, "step": 101270 }, { "grad_norm": 3.256467580795288, "learning_rate": 5.307434013719817e-05, "loss": 0.0474, "step": 101280 }, { "grad_norm": 0.35972386598587036, "learning_rate": 5.306608838880174e-05, "loss": 0.0389, "step": 101290 }, { "grad_norm": 0.46605148911476135, "learning_rate": 5.30578365565796e-05, "loss": 0.021, "step": 101300 }, { "grad_norm": 1.602473258972168, "learning_rate": 5.3049584640757355e-05, "loss": 0.0288, "step": 101310 }, { "grad_norm": 0.5171403288841248, "learning_rate": 5.3041332641560614e-05, "loss": 0.0276, "step": 101320 }, { "grad_norm": 0.7135312557220459, "learning_rate": 5.3033080559214985e-05, "loss": 0.0329, "step": 101330 }, { "grad_norm": 0.4653990864753723, "learning_rate": 5.302482839394607e-05, "loss": 0.0325, "step": 101340 }, { "grad_norm": 0.40344497561454773, "learning_rate": 5.301657614597947e-05, "loss": 0.0296, "step": 101350 }, { "grad_norm": 0.4072522819042206, "learning_rate": 5.3008323815540826e-05, "loss": 0.0423, "step": 101360 }, { "grad_norm": 0.36320745944976807, "learning_rate": 5.300007140285574e-05, "loss": 0.0267, "step": 101370 }, { "grad_norm": 0.28813934326171875, "learning_rate": 5.299181890814983e-05, "loss": 0.0301, "step": 101380 }, { "grad_norm": 2.823155641555786, "learning_rate": 5.29835663316487e-05, "loss": 0.0265, "step": 101390 }, { "grad_norm": 0.6475907564163208, "learning_rate": 5.2975313673578e-05, "loss": 0.0363, "step": 101400 }, { "grad_norm": 0.4772455096244812, "learning_rate": 5.296706093416334e-05, "loss": 0.0249, "step": 101410 }, { "grad_norm": 0.34151363372802734, "learning_rate": 5.2958808113630344e-05, "loss": 0.0267, "step": 101420 }, { "grad_norm": 0.42349106073379517, "learning_rate": 5.2950555212204655e-05, "loss": 0.0236, "step": 101430 }, { "grad_norm": 0.4581141471862793, "learning_rate": 5.2942302230111884e-05, "loss": 0.0245, "step": 101440 }, { "grad_norm": 0.6610183119773865, "learning_rate": 5.293404916757768e-05, "loss": 0.0286, "step": 101450 }, { "grad_norm": 0.5196149349212646, "learning_rate": 5.292579602482768e-05, "loss": 0.0334, "step": 101460 }, { "grad_norm": 0.4375154674053192, "learning_rate": 5.2917542802087506e-05, "loss": 0.039, "step": 101470 }, { "grad_norm": 0.3865843713283539, "learning_rate": 5.290928949958281e-05, "loss": 0.0282, "step": 101480 }, { "grad_norm": 0.6173596978187561, "learning_rate": 5.290103611753924e-05, "loss": 0.0321, "step": 101490 }, { "grad_norm": 0.6059337258338928, "learning_rate": 5.2892782656182425e-05, "loss": 0.0351, "step": 101500 }, { "grad_norm": 0.5623438358306885, "learning_rate": 5.2884529115738024e-05, "loss": 0.0284, "step": 101510 }, { "grad_norm": 0.32666832208633423, "learning_rate": 5.2876275496431684e-05, "loss": 0.0328, "step": 101520 }, { "grad_norm": 0.43561625480651855, "learning_rate": 5.286802179848904e-05, "loss": 0.0422, "step": 101530 }, { "grad_norm": 0.5027660131454468, "learning_rate": 5.285976802213576e-05, "loss": 0.0221, "step": 101540 }, { "grad_norm": 0.581565797328949, "learning_rate": 5.28515141675975e-05, "loss": 0.0383, "step": 101550 }, { "grad_norm": 0.31592854857444763, "learning_rate": 5.2843260235099924e-05, "loss": 0.0275, "step": 101560 }, { "grad_norm": 4.4046831130981445, "learning_rate": 5.283500622486867e-05, "loss": 0.043, "step": 101570 }, { "grad_norm": 0.4831790328025818, "learning_rate": 5.282675213712942e-05, "loss": 0.0218, "step": 101580 }, { "grad_norm": 0.610232412815094, "learning_rate": 5.281849797210782e-05, "loss": 0.0426, "step": 101590 }, { "grad_norm": 0.71791011095047, "learning_rate": 5.281024373002954e-05, "loss": 0.0413, "step": 101600 }, { "grad_norm": 0.45832034945487976, "learning_rate": 5.2801989411120245e-05, "loss": 0.0231, "step": 101610 }, { "grad_norm": 0.4590020775794983, "learning_rate": 5.279373501560564e-05, "loss": 0.0349, "step": 101620 }, { "grad_norm": 0.4634493291378021, "learning_rate": 5.2785480543711353e-05, "loss": 0.0269, "step": 101630 }, { "grad_norm": 0.6676647663116455, "learning_rate": 5.2777225995663074e-05, "loss": 0.0337, "step": 101640 }, { "grad_norm": 0.742085874080658, "learning_rate": 5.276897137168648e-05, "loss": 0.0307, "step": 101650 }, { "grad_norm": 0.743109405040741, "learning_rate": 5.276071667200725e-05, "loss": 0.0365, "step": 101660 }, { "grad_norm": 0.40301966667175293, "learning_rate": 5.2752461896851055e-05, "loss": 0.0228, "step": 101670 }, { "grad_norm": 0.34572333097457886, "learning_rate": 5.2744207046443604e-05, "loss": 0.0296, "step": 101680 }, { "grad_norm": 0.5535408854484558, "learning_rate": 5.273595212101055e-05, "loss": 0.0346, "step": 101690 }, { "grad_norm": 0.36204981803894043, "learning_rate": 5.2727697120777586e-05, "loss": 0.0355, "step": 101700 }, { "grad_norm": 1.3300203084945679, "learning_rate": 5.271944204597043e-05, "loss": 0.0363, "step": 101710 }, { "grad_norm": 0.6213182210922241, "learning_rate": 5.271118689681473e-05, "loss": 0.0356, "step": 101720 }, { "grad_norm": 0.4075838029384613, "learning_rate": 5.27029316735362e-05, "loss": 0.0475, "step": 101730 }, { "grad_norm": 0.5233167409896851, "learning_rate": 5.269467637636054e-05, "loss": 0.0455, "step": 101740 }, { "grad_norm": 0.6458942294120789, "learning_rate": 5.2686421005513454e-05, "loss": 0.0466, "step": 101750 }, { "grad_norm": 0.5006949305534363, "learning_rate": 5.26781655612206e-05, "loss": 0.0324, "step": 101760 }, { "grad_norm": 0.30892032384872437, "learning_rate": 5.2669910043707736e-05, "loss": 0.0248, "step": 101770 }, { "grad_norm": 1.3307970762252808, "learning_rate": 5.266165445320052e-05, "loss": 0.0296, "step": 101780 }, { "grad_norm": 0.4436424970626831, "learning_rate": 5.2653398789924666e-05, "loss": 0.0301, "step": 101790 }, { "grad_norm": 0.4232842028141022, "learning_rate": 5.264514305410587e-05, "loss": 0.0358, "step": 101800 }, { "grad_norm": 0.4276881217956543, "learning_rate": 5.263688724596989e-05, "loss": 0.03, "step": 101810 }, { "grad_norm": 0.4658087193965912, "learning_rate": 5.262863136574239e-05, "loss": 0.0274, "step": 101820 }, { "grad_norm": 1.5582548379898071, "learning_rate": 5.2620375413649096e-05, "loss": 0.0397, "step": 101830 }, { "grad_norm": 0.49889397621154785, "learning_rate": 5.2612119389915716e-05, "loss": 0.0342, "step": 101840 }, { "grad_norm": 0.5760471820831299, "learning_rate": 5.2603863294767986e-05, "loss": 0.0265, "step": 101850 }, { "grad_norm": 0.6447591185569763, "learning_rate": 5.259560712843159e-05, "loss": 0.0262, "step": 101860 }, { "grad_norm": 0.6612941026687622, "learning_rate": 5.2587350891132295e-05, "loss": 0.0231, "step": 101870 }, { "grad_norm": 0.4297146201133728, "learning_rate": 5.2579094583095776e-05, "loss": 0.0252, "step": 101880 }, { "grad_norm": 0.34922412037849426, "learning_rate": 5.25708382045478e-05, "loss": 0.0359, "step": 101890 }, { "grad_norm": 0.7772977948188782, "learning_rate": 5.2562581755714066e-05, "loss": 0.0379, "step": 101900 }, { "grad_norm": 0.3640477657318115, "learning_rate": 5.25543252368203e-05, "loss": 0.0333, "step": 101910 }, { "grad_norm": 0.5770970582962036, "learning_rate": 5.254606864809225e-05, "loss": 0.0264, "step": 101920 }, { "grad_norm": 0.450250506401062, "learning_rate": 5.253781198975565e-05, "loss": 0.0205, "step": 101930 }, { "grad_norm": 0.527385413646698, "learning_rate": 5.252955526203621e-05, "loss": 0.0256, "step": 101940 }, { "grad_norm": 0.47520482540130615, "learning_rate": 5.252129846515967e-05, "loss": 0.0348, "step": 101950 }, { "grad_norm": 0.32911238074302673, "learning_rate": 5.251304159935181e-05, "loss": 0.024, "step": 101960 }, { "grad_norm": 0.7274307012557983, "learning_rate": 5.250478466483831e-05, "loss": 0.0274, "step": 101970 }, { "grad_norm": 0.566866934299469, "learning_rate": 5.249652766184495e-05, "loss": 0.0288, "step": 101980 }, { "grad_norm": 0.503952145576477, "learning_rate": 5.2488270590597444e-05, "loss": 0.0239, "step": 101990 }, { "grad_norm": 0.29777243733406067, "learning_rate": 5.248001345132157e-05, "loss": 0.0327, "step": 102000 }, { "grad_norm": 0.527303159236908, "learning_rate": 5.247175624424306e-05, "loss": 0.0312, "step": 102010 }, { "grad_norm": 0.5424371361732483, "learning_rate": 5.246349896958766e-05, "loss": 0.0306, "step": 102020 }, { "grad_norm": 0.5086869597434998, "learning_rate": 5.245524162758113e-05, "loss": 0.0262, "step": 102030 }, { "grad_norm": 0.39528974890708923, "learning_rate": 5.2446984218449225e-05, "loss": 0.0241, "step": 102040 }, { "grad_norm": 0.5190277695655823, "learning_rate": 5.243872674241766e-05, "loss": 0.0262, "step": 102050 }, { "grad_norm": 0.3305445611476898, "learning_rate": 5.243046919971224e-05, "loss": 0.0238, "step": 102060 }, { "grad_norm": 0.6372914910316467, "learning_rate": 5.24222115905587e-05, "loss": 0.0358, "step": 102070 }, { "grad_norm": 0.4443569481372833, "learning_rate": 5.241395391518282e-05, "loss": 0.0257, "step": 102080 }, { "grad_norm": 0.42221948504447937, "learning_rate": 5.240569617381033e-05, "loss": 0.0214, "step": 102090 }, { "grad_norm": 0.5035356879234314, "learning_rate": 5.2397438366667015e-05, "loss": 0.0392, "step": 102100 }, { "grad_norm": 0.6017425656318665, "learning_rate": 5.238918049397864e-05, "loss": 0.0335, "step": 102110 }, { "grad_norm": 0.3254360258579254, "learning_rate": 5.238092255597096e-05, "loss": 0.0248, "step": 102120 }, { "grad_norm": 0.34564849734306335, "learning_rate": 5.237266455286976e-05, "loss": 0.0254, "step": 102130 }, { "grad_norm": 0.48590198159217834, "learning_rate": 5.2364406484900795e-05, "loss": 0.0303, "step": 102140 }, { "grad_norm": 0.3508943021297455, "learning_rate": 5.235614835228985e-05, "loss": 0.0289, "step": 102150 }, { "grad_norm": 1.08981192111969, "learning_rate": 5.234789015526269e-05, "loss": 0.0252, "step": 102160 }, { "grad_norm": 0.640948474407196, "learning_rate": 5.2339631894045106e-05, "loss": 0.0383, "step": 102170 }, { "grad_norm": 0.4483078420162201, "learning_rate": 5.233137356886285e-05, "loss": 0.0339, "step": 102180 }, { "grad_norm": 0.5712357759475708, "learning_rate": 5.232311517994173e-05, "loss": 0.0477, "step": 102190 }, { "grad_norm": 1.1202386617660522, "learning_rate": 5.23148567275075e-05, "loss": 0.0614, "step": 102200 }, { "grad_norm": 0.4962970018386841, "learning_rate": 5.230659821178597e-05, "loss": 0.0225, "step": 102210 }, { "grad_norm": 0.30897149443626404, "learning_rate": 5.22983396330029e-05, "loss": 0.0337, "step": 102220 }, { "grad_norm": 0.7797108292579651, "learning_rate": 5.229008099138411e-05, "loss": 0.0263, "step": 102230 }, { "grad_norm": 0.8559194803237915, "learning_rate": 5.228182228715532e-05, "loss": 0.0289, "step": 102240 }, { "grad_norm": 0.4994807243347168, "learning_rate": 5.227356352054241e-05, "loss": 0.0323, "step": 102250 }, { "grad_norm": 0.5630291104316711, "learning_rate": 5.226530469177111e-05, "loss": 0.0341, "step": 102260 }, { "grad_norm": 0.36460158228874207, "learning_rate": 5.225704580106724e-05, "loss": 0.0418, "step": 102270 }, { "grad_norm": 0.7306073904037476, "learning_rate": 5.224878684865657e-05, "loss": 0.0328, "step": 102280 }, { "grad_norm": 0.42552629113197327, "learning_rate": 5.224052783476494e-05, "loss": 0.0298, "step": 102290 }, { "grad_norm": 0.7532626986503601, "learning_rate": 5.2232268759618094e-05, "loss": 0.0505, "step": 102300 }, { "grad_norm": 0.5450787544250488, "learning_rate": 5.222400962344186e-05, "loss": 0.0244, "step": 102310 }, { "grad_norm": 0.41640356183052063, "learning_rate": 5.221575042646205e-05, "loss": 0.0531, "step": 102320 }, { "grad_norm": 0.5565289258956909, "learning_rate": 5.2207491168904455e-05, "loss": 0.0291, "step": 102330 }, { "grad_norm": 0.45882225036621094, "learning_rate": 5.219923185099488e-05, "loss": 0.0527, "step": 102340 }, { "grad_norm": 0.3710138201713562, "learning_rate": 5.2190972472959134e-05, "loss": 0.0286, "step": 102350 }, { "grad_norm": 0.4116935133934021, "learning_rate": 5.218271303502301e-05, "loss": 0.0361, "step": 102360 }, { "grad_norm": 0.49577727913856506, "learning_rate": 5.217445353741234e-05, "loss": 0.0283, "step": 102370 }, { "grad_norm": 0.40455591678619385, "learning_rate": 5.216619398035293e-05, "loss": 0.0257, "step": 102380 }, { "grad_norm": 1.4438616037368774, "learning_rate": 5.2157934364070594e-05, "loss": 0.0389, "step": 102390 }, { "grad_norm": 0.4266554117202759, "learning_rate": 5.214967468879113e-05, "loss": 0.0269, "step": 102400 }, { "grad_norm": 0.5840001702308655, "learning_rate": 5.2141414954740376e-05, "loss": 0.0529, "step": 102410 }, { "grad_norm": 0.37510353326797485, "learning_rate": 5.2133155162144144e-05, "loss": 0.0353, "step": 102420 }, { "grad_norm": 0.46161627769470215, "learning_rate": 5.212489531122825e-05, "loss": 0.0307, "step": 102430 }, { "grad_norm": 0.7089904546737671, "learning_rate": 5.2116635402218505e-05, "loss": 0.027, "step": 102440 }, { "grad_norm": 0.5190269947052002, "learning_rate": 5.210837543534075e-05, "loss": 0.0305, "step": 102450 }, { "grad_norm": 0.5239853262901306, "learning_rate": 5.210011541082081e-05, "loss": 0.0304, "step": 102460 }, { "grad_norm": 0.4733051359653473, "learning_rate": 5.20918553288845e-05, "loss": 0.0262, "step": 102470 }, { "grad_norm": 0.4023882746696472, "learning_rate": 5.2083595189757664e-05, "loss": 0.0317, "step": 102480 }, { "grad_norm": 0.3274003863334656, "learning_rate": 5.207533499366609e-05, "loss": 0.0409, "step": 102490 }, { "grad_norm": 0.7520623207092285, "learning_rate": 5.206707474083565e-05, "loss": 0.0358, "step": 102500 }, { "grad_norm": 0.45037510991096497, "learning_rate": 5.2058814431492165e-05, "loss": 0.0329, "step": 102510 }, { "grad_norm": 0.38887155055999756, "learning_rate": 5.205055406586147e-05, "loss": 0.0284, "step": 102520 }, { "grad_norm": 0.7878366708755493, "learning_rate": 5.204229364416939e-05, "loss": 0.0362, "step": 102530 }, { "grad_norm": 0.433633416891098, "learning_rate": 5.203403316664178e-05, "loss": 0.0344, "step": 102540 }, { "grad_norm": 0.6019538044929504, "learning_rate": 5.2025772633504455e-05, "loss": 0.0268, "step": 102550 }, { "grad_norm": 0.6976237893104553, "learning_rate": 5.201751204498327e-05, "loss": 0.0338, "step": 102560 }, { "grad_norm": 0.7179502248764038, "learning_rate": 5.2009251401304074e-05, "loss": 0.0318, "step": 102570 }, { "grad_norm": 0.5629257559776306, "learning_rate": 5.2000990702692683e-05, "loss": 0.0263, "step": 102580 }, { "grad_norm": 0.4448089599609375, "learning_rate": 5.199272994937497e-05, "loss": 0.0276, "step": 102590 }, { "grad_norm": 0.5803705453872681, "learning_rate": 5.198446914157677e-05, "loss": 0.0277, "step": 102600 }, { "grad_norm": 0.4672357439994812, "learning_rate": 5.197620827952393e-05, "loss": 0.0292, "step": 102610 }, { "grad_norm": 0.6202487945556641, "learning_rate": 5.1967947363442284e-05, "loss": 0.0299, "step": 102620 }, { "grad_norm": 0.2930891811847687, "learning_rate": 5.19596863935577e-05, "loss": 0.0405, "step": 102630 }, { "grad_norm": 0.6048433184623718, "learning_rate": 5.1951425370096044e-05, "loss": 0.0356, "step": 102640 }, { "grad_norm": 0.6330950260162354, "learning_rate": 5.194316429328313e-05, "loss": 0.0342, "step": 102650 }, { "grad_norm": 0.44836920499801636, "learning_rate": 5.193490316334484e-05, "loss": 0.0343, "step": 102660 }, { "grad_norm": 0.5943353176116943, "learning_rate": 5.192664198050704e-05, "loss": 0.0307, "step": 102670 }, { "grad_norm": 0.344828337430954, "learning_rate": 5.1918380744995556e-05, "loss": 0.0252, "step": 102680 }, { "grad_norm": 0.6530680656433105, "learning_rate": 5.191011945703625e-05, "loss": 0.0379, "step": 102690 }, { "grad_norm": 0.6444007158279419, "learning_rate": 5.1901858116855016e-05, "loss": 0.0235, "step": 102700 }, { "grad_norm": 1.4366415739059448, "learning_rate": 5.1893596724677706e-05, "loss": 0.0267, "step": 102710 }, { "grad_norm": 0.29313963651657104, "learning_rate": 5.188533528073015e-05, "loss": 0.0293, "step": 102720 }, { "grad_norm": 0.5701538324356079, "learning_rate": 5.187707378523825e-05, "loss": 0.0246, "step": 102730 }, { "grad_norm": 0.42156413197517395, "learning_rate": 5.1868812238427855e-05, "loss": 0.0342, "step": 102740 }, { "grad_norm": 0.49630647897720337, "learning_rate": 5.186055064052483e-05, "loss": 0.0262, "step": 102750 }, { "grad_norm": 0.4949011504650116, "learning_rate": 5.185228899175505e-05, "loss": 0.0311, "step": 102760 }, { "grad_norm": 0.539818525314331, "learning_rate": 5.184402729234439e-05, "loss": 0.0328, "step": 102770 }, { "grad_norm": 0.5183004140853882, "learning_rate": 5.183576554251871e-05, "loss": 0.0299, "step": 102780 }, { "grad_norm": 0.5751222372055054, "learning_rate": 5.1827503742503904e-05, "loss": 0.0269, "step": 102790 }, { "grad_norm": 0.40856167674064636, "learning_rate": 5.181924189252582e-05, "loss": 0.0391, "step": 102800 }, { "grad_norm": 0.9876253604888916, "learning_rate": 5.181097999281035e-05, "loss": 0.0332, "step": 102810 }, { "grad_norm": 0.30267417430877686, "learning_rate": 5.1802718043583356e-05, "loss": 0.0245, "step": 102820 }, { "grad_norm": 0.7823708057403564, "learning_rate": 5.179445604507074e-05, "loss": 0.0426, "step": 102830 }, { "grad_norm": 0.6882529258728027, "learning_rate": 5.1786193997498365e-05, "loss": 0.0282, "step": 102840 }, { "grad_norm": 0.4960191547870636, "learning_rate": 5.1777931901092124e-05, "loss": 0.0337, "step": 102850 }, { "grad_norm": 0.29464876651763916, "learning_rate": 5.176966975607789e-05, "loss": 0.0262, "step": 102860 }, { "grad_norm": 0.36246755719184875, "learning_rate": 5.176140756268154e-05, "loss": 0.0481, "step": 102870 }, { "grad_norm": 0.7384449243545532, "learning_rate": 5.175314532112897e-05, "loss": 0.0325, "step": 102880 }, { "grad_norm": 0.26266753673553467, "learning_rate": 5.1744883031646086e-05, "loss": 0.0312, "step": 102890 }, { "grad_norm": 0.3715779185295105, "learning_rate": 5.1736620694458735e-05, "loss": 0.0305, "step": 102900 }, { "grad_norm": 0.45822837948799133, "learning_rate": 5.172835830979284e-05, "loss": 0.0301, "step": 102910 }, { "grad_norm": 0.4712439775466919, "learning_rate": 5.172009587787426e-05, "loss": 0.0335, "step": 102920 }, { "grad_norm": 0.8670405149459839, "learning_rate": 5.171183339892892e-05, "loss": 0.0349, "step": 102930 }, { "grad_norm": 0.3632388412952423, "learning_rate": 5.170357087318269e-05, "loss": 0.0289, "step": 102940 }, { "grad_norm": 0.5997061133384705, "learning_rate": 5.1695308300861476e-05, "loss": 0.0302, "step": 102950 }, { "grad_norm": 0.47520583868026733, "learning_rate": 5.168704568219118e-05, "loss": 0.036, "step": 102960 }, { "grad_norm": 0.4491530954837799, "learning_rate": 5.167878301739767e-05, "loss": 0.0236, "step": 102970 }, { "grad_norm": 0.45290544629096985, "learning_rate": 5.1670520306706884e-05, "loss": 0.027, "step": 102980 }, { "grad_norm": 0.44061288237571716, "learning_rate": 5.1662257550344686e-05, "loss": 0.0387, "step": 102990 }, { "grad_norm": 0.7088019847869873, "learning_rate": 5.1653994748536996e-05, "loss": 0.0295, "step": 103000 }, { "grad_norm": 0.4991184175014496, "learning_rate": 5.16457319015097e-05, "loss": 0.0298, "step": 103010 }, { "grad_norm": 0.7069200873374939, "learning_rate": 5.1637469009488724e-05, "loss": 0.0365, "step": 103020 }, { "grad_norm": 0.5078380107879639, "learning_rate": 5.162920607269996e-05, "loss": 0.0295, "step": 103030 }, { "grad_norm": 0.5307647585868835, "learning_rate": 5.162094309136931e-05, "loss": 0.0436, "step": 103040 }, { "grad_norm": 0.7849757075309753, "learning_rate": 5.161268006572268e-05, "loss": 0.0429, "step": 103050 }, { "grad_norm": 0.36533617973327637, "learning_rate": 5.1604416995986e-05, "loss": 0.0326, "step": 103060 }, { "grad_norm": 0.3726770877838135, "learning_rate": 5.1596153882385145e-05, "loss": 0.0245, "step": 103070 }, { "grad_norm": 3.069251298904419, "learning_rate": 5.158789072514605e-05, "loss": 0.0267, "step": 103080 }, { "grad_norm": 0.780836820602417, "learning_rate": 5.157962752449461e-05, "loss": 0.0366, "step": 103090 }, { "grad_norm": 0.5580904483795166, "learning_rate": 5.1571364280656765e-05, "loss": 0.0279, "step": 103100 }, { "grad_norm": 0.6624284982681274, "learning_rate": 5.1563100993858405e-05, "loss": 0.0283, "step": 103110 }, { "grad_norm": 0.30213287472724915, "learning_rate": 5.155483766432544e-05, "loss": 0.027, "step": 103120 }, { "grad_norm": 0.2604427635669708, "learning_rate": 5.1546574292283803e-05, "loss": 0.027, "step": 103130 }, { "grad_norm": 0.46583595871925354, "learning_rate": 5.1538310877959396e-05, "loss": 0.0368, "step": 103140 }, { "grad_norm": 0.4640198051929474, "learning_rate": 5.153004742157816e-05, "loss": 0.0333, "step": 103150 }, { "grad_norm": 0.29513615369796753, "learning_rate": 5.152178392336601e-05, "loss": 0.0199, "step": 103160 }, { "grad_norm": 1.0630671977996826, "learning_rate": 5.1513520383548853e-05, "loss": 0.03, "step": 103170 }, { "grad_norm": 0.6839833855628967, "learning_rate": 5.1505256802352606e-05, "loss": 0.0277, "step": 103180 }, { "grad_norm": 0.4564085602760315, "learning_rate": 5.149699318000323e-05, "loss": 0.0305, "step": 103190 }, { "grad_norm": 1.0063624382019043, "learning_rate": 5.148872951672659e-05, "loss": 0.0241, "step": 103200 }, { "grad_norm": 0.323201447725296, "learning_rate": 5.148046581274866e-05, "loss": 0.0299, "step": 103210 }, { "grad_norm": 0.6275807619094849, "learning_rate": 5.147220206829535e-05, "loss": 0.0356, "step": 103220 }, { "grad_norm": 0.5363758206367493, "learning_rate": 5.146393828359259e-05, "loss": 0.0334, "step": 103230 }, { "grad_norm": 0.38194772601127625, "learning_rate": 5.14556744588663e-05, "loss": 0.0394, "step": 103240 }, { "grad_norm": 0.2989959418773651, "learning_rate": 5.144741059434244e-05, "loss": 0.023, "step": 103250 }, { "grad_norm": 0.3607182204723358, "learning_rate": 5.143914669024688e-05, "loss": 0.0311, "step": 103260 }, { "grad_norm": 0.20386753976345062, "learning_rate": 5.143088274680562e-05, "loss": 0.0282, "step": 103270 }, { "grad_norm": 0.45137181878089905, "learning_rate": 5.1422618764244544e-05, "loss": 0.023, "step": 103280 }, { "grad_norm": 0.49407997727394104, "learning_rate": 5.141435474278963e-05, "loss": 0.0259, "step": 103290 }, { "grad_norm": 0.5241004228591919, "learning_rate": 5.1406090682666775e-05, "loss": 0.0265, "step": 103300 }, { "grad_norm": 0.7956961393356323, "learning_rate": 5.139782658410193e-05, "loss": 0.0306, "step": 103310 }, { "grad_norm": 0.43703824281692505, "learning_rate": 5.138956244732103e-05, "loss": 0.0337, "step": 103320 }, { "grad_norm": 1.1040911674499512, "learning_rate": 5.1381298272550005e-05, "loss": 0.0325, "step": 103330 }, { "grad_norm": 0.5423431992530823, "learning_rate": 5.1373034060014816e-05, "loss": 0.0316, "step": 103340 }, { "grad_norm": 1.1638221740722656, "learning_rate": 5.136476980994138e-05, "loss": 0.0449, "step": 103350 }, { "grad_norm": 0.5679879784584045, "learning_rate": 5.135650552255567e-05, "loss": 0.0329, "step": 103360 }, { "grad_norm": 0.24629929661750793, "learning_rate": 5.134824119808359e-05, "loss": 0.0302, "step": 103370 }, { "grad_norm": 0.5016788244247437, "learning_rate": 5.133997683675113e-05, "loss": 0.0255, "step": 103380 }, { "grad_norm": 0.42214158177375793, "learning_rate": 5.133171243878416e-05, "loss": 0.0288, "step": 103390 }, { "grad_norm": 0.5113905072212219, "learning_rate": 5.1323448004408714e-05, "loss": 0.0257, "step": 103400 }, { "grad_norm": 0.6863647699356079, "learning_rate": 5.131518353385067e-05, "loss": 0.0294, "step": 103410 }, { "grad_norm": 0.6447385549545288, "learning_rate": 5.130691902733603e-05, "loss": 0.0366, "step": 103420 }, { "grad_norm": 0.5830914378166199, "learning_rate": 5.129865448509069e-05, "loss": 0.037, "step": 103430 }, { "grad_norm": 0.5420961976051331, "learning_rate": 5.129038990734064e-05, "loss": 0.032, "step": 103440 }, { "grad_norm": 0.43218427896499634, "learning_rate": 5.12821252943118e-05, "loss": 0.0292, "step": 103450 }, { "grad_norm": 0.5323514938354492, "learning_rate": 5.127386064623014e-05, "loss": 0.0364, "step": 103460 }, { "grad_norm": 0.40736883878707886, "learning_rate": 5.1265595963321625e-05, "loss": 0.0294, "step": 103470 }, { "grad_norm": 1.5250561237335205, "learning_rate": 5.125733124581218e-05, "loss": 0.0251, "step": 103480 }, { "grad_norm": 0.3845289647579193, "learning_rate": 5.1249066493927765e-05, "loss": 0.0311, "step": 103490 }, { "grad_norm": 1.4986294507980347, "learning_rate": 5.124080170789436e-05, "loss": 0.0303, "step": 103500 }, { "grad_norm": 0.3939840793609619, "learning_rate": 5.1232536887937876e-05, "loss": 0.0291, "step": 103510 }, { "grad_norm": 0.38792529702186584, "learning_rate": 5.1224272034284316e-05, "loss": 0.0227, "step": 103520 }, { "grad_norm": 0.24856778979301453, "learning_rate": 5.1216007147159606e-05, "loss": 0.0339, "step": 103530 }, { "grad_norm": 0.7962961792945862, "learning_rate": 5.120774222678972e-05, "loss": 0.0275, "step": 103540 }, { "grad_norm": 0.4980488419532776, "learning_rate": 5.119947727340062e-05, "loss": 0.0271, "step": 103550 }, { "grad_norm": 0.5823023915290833, "learning_rate": 5.119121228721826e-05, "loss": 0.0301, "step": 103560 }, { "grad_norm": 0.6536034941673279, "learning_rate": 5.118294726846862e-05, "loss": 0.0323, "step": 103570 }, { "grad_norm": 0.4298621118068695, "learning_rate": 5.117468221737762e-05, "loss": 0.038, "step": 103580 }, { "grad_norm": 0.46531325578689575, "learning_rate": 5.116641713417125e-05, "loss": 0.0274, "step": 103590 }, { "grad_norm": 0.39911437034606934, "learning_rate": 5.115815201907549e-05, "loss": 0.0221, "step": 103600 }, { "grad_norm": 0.4326862096786499, "learning_rate": 5.114988687231629e-05, "loss": 0.0271, "step": 103610 }, { "grad_norm": 0.41182011365890503, "learning_rate": 5.11416216941196e-05, "loss": 0.0392, "step": 103620 }, { "grad_norm": 0.539916455745697, "learning_rate": 5.113335648471141e-05, "loss": 0.0284, "step": 103630 }, { "grad_norm": 0.39236804842948914, "learning_rate": 5.1125091244317666e-05, "loss": 0.0222, "step": 103640 }, { "grad_norm": 0.5089298486709595, "learning_rate": 5.111682597316437e-05, "loss": 0.0315, "step": 103650 }, { "grad_norm": 0.5671442151069641, "learning_rate": 5.1108560671477466e-05, "loss": 0.0249, "step": 103660 }, { "grad_norm": 0.534764289855957, "learning_rate": 5.110029533948294e-05, "loss": 0.0305, "step": 103670 }, { "grad_norm": 0.4643038213253021, "learning_rate": 5.109202997740673e-05, "loss": 0.0415, "step": 103680 }, { "grad_norm": 0.5078378319740295, "learning_rate": 5.108376458547486e-05, "loss": 0.0249, "step": 103690 }, { "grad_norm": 1.6073085069656372, "learning_rate": 5.107549916391324e-05, "loss": 0.036, "step": 103700 }, { "grad_norm": 0.38855209946632385, "learning_rate": 5.106723371294789e-05, "loss": 0.0324, "step": 103710 }, { "grad_norm": 0.44146689772605896, "learning_rate": 5.105896823280477e-05, "loss": 0.0274, "step": 103720 }, { "grad_norm": 0.4278614819049835, "learning_rate": 5.1050702723709864e-05, "loss": 0.0271, "step": 103730 }, { "grad_norm": 0.6762109398841858, "learning_rate": 5.104243718588914e-05, "loss": 0.0419, "step": 103740 }, { "grad_norm": 1.0032150745391846, "learning_rate": 5.103417161956857e-05, "loss": 0.0365, "step": 103750 }, { "grad_norm": 0.4512339234352112, "learning_rate": 5.1025906024974135e-05, "loss": 0.0306, "step": 103760 }, { "grad_norm": 0.7776314616203308, "learning_rate": 5.1017640402331814e-05, "loss": 0.0365, "step": 103770 }, { "grad_norm": 1.1047848463058472, "learning_rate": 5.100937475186759e-05, "loss": 0.0322, "step": 103780 }, { "grad_norm": 3.7050039768218994, "learning_rate": 5.100110907380744e-05, "loss": 0.0279, "step": 103790 }, { "grad_norm": 1.8103904724121094, "learning_rate": 5.099284336837734e-05, "loss": 0.0341, "step": 103800 }, { "grad_norm": 0.33210936188697815, "learning_rate": 5.0984577635803276e-05, "loss": 0.0379, "step": 103810 }, { "grad_norm": 0.37837615609169006, "learning_rate": 5.097631187631124e-05, "loss": 0.048, "step": 103820 }, { "grad_norm": 0.5947933197021484, "learning_rate": 5.0968046090127195e-05, "loss": 0.0301, "step": 103830 }, { "grad_norm": 0.45075634121894836, "learning_rate": 5.095978027747713e-05, "loss": 0.0256, "step": 103840 }, { "grad_norm": 1.506962537765503, "learning_rate": 5.095151443858703e-05, "loss": 0.0364, "step": 103850 }, { "grad_norm": 0.4359970986843109, "learning_rate": 5.09432485736829e-05, "loss": 0.0325, "step": 103860 }, { "grad_norm": 0.5231059789657593, "learning_rate": 5.0934982682990696e-05, "loss": 0.0271, "step": 103870 }, { "grad_norm": 1.240476131439209, "learning_rate": 5.0926716766736425e-05, "loss": 0.0329, "step": 103880 }, { "grad_norm": 1.2149676084518433, "learning_rate": 5.091845082514606e-05, "loss": 0.0336, "step": 103890 }, { "grad_norm": 0.6210073232650757, "learning_rate": 5.091018485844559e-05, "loss": 0.043, "step": 103900 }, { "grad_norm": 0.5704084634780884, "learning_rate": 5.090191886686101e-05, "loss": 0.0229, "step": 103910 }, { "grad_norm": 0.7097941637039185, "learning_rate": 5.0893652850618315e-05, "loss": 0.0322, "step": 103920 }, { "grad_norm": 0.4984464943408966, "learning_rate": 5.088538680994347e-05, "loss": 0.0246, "step": 103930 }, { "grad_norm": 0.4353054463863373, "learning_rate": 5.08771207450625e-05, "loss": 0.0394, "step": 103940 }, { "grad_norm": 2.2983486652374268, "learning_rate": 5.086885465620137e-05, "loss": 0.0297, "step": 103950 }, { "grad_norm": 0.5407820343971252, "learning_rate": 5.0860588543586083e-05, "loss": 0.0269, "step": 103960 }, { "grad_norm": 0.9199707508087158, "learning_rate": 5.085232240744262e-05, "loss": 0.0256, "step": 103970 }, { "grad_norm": 0.589459240436554, "learning_rate": 5.0844056247997e-05, "loss": 0.0399, "step": 103980 }, { "grad_norm": 0.9629659056663513, "learning_rate": 5.0835790065475184e-05, "loss": 0.0461, "step": 103990 }, { "grad_norm": 0.8407498002052307, "learning_rate": 5.08275238601032e-05, "loss": 0.0345, "step": 104000 }, { "grad_norm": 0.47006404399871826, "learning_rate": 5.0819257632106996e-05, "loss": 0.0399, "step": 104010 }, { "grad_norm": 0.414384126663208, "learning_rate": 5.081099138171261e-05, "loss": 0.05, "step": 104020 }, { "grad_norm": 0.3975330591201782, "learning_rate": 5.080272510914602e-05, "loss": 0.0345, "step": 104030 }, { "grad_norm": 0.3797082304954529, "learning_rate": 5.079445881463325e-05, "loss": 0.0247, "step": 104040 }, { "grad_norm": 0.33454397320747375, "learning_rate": 5.0786192498400254e-05, "loss": 0.0282, "step": 104050 }, { "grad_norm": 0.3523446321487427, "learning_rate": 5.077792616067304e-05, "loss": 0.0356, "step": 104060 }, { "grad_norm": 0.40783464908599854, "learning_rate": 5.076965980167764e-05, "loss": 0.0509, "step": 104070 }, { "grad_norm": 0.5108580589294434, "learning_rate": 5.0761393421640024e-05, "loss": 0.027, "step": 104080 }, { "grad_norm": 0.4694380760192871, "learning_rate": 5.075312702078618e-05, "loss": 0.0327, "step": 104090 }, { "grad_norm": 0.8096480965614319, "learning_rate": 5.074486059934215e-05, "loss": 0.0353, "step": 104100 }, { "grad_norm": 0.6234089136123657, "learning_rate": 5.07365941575339e-05, "loss": 0.0291, "step": 104110 }, { "grad_norm": 0.3222731053829193, "learning_rate": 5.072832769558744e-05, "loss": 0.0396, "step": 104120 }, { "grad_norm": 0.4456828832626343, "learning_rate": 5.0720061213728784e-05, "loss": 0.032, "step": 104130 }, { "grad_norm": 0.5654675960540771, "learning_rate": 5.071179471218392e-05, "loss": 0.0321, "step": 104140 }, { "grad_norm": 0.5993722677230835, "learning_rate": 5.070352819117885e-05, "loss": 0.0273, "step": 104150 }, { "grad_norm": 0.6318230628967285, "learning_rate": 5.0695261650939584e-05, "loss": 0.0345, "step": 104160 }, { "grad_norm": 0.8503479361534119, "learning_rate": 5.068699509169215e-05, "loss": 0.0369, "step": 104170 }, { "grad_norm": 0.7922025322914124, "learning_rate": 5.06787285136625e-05, "loss": 0.0238, "step": 104180 }, { "grad_norm": 0.8475326299667358, "learning_rate": 5.0670461917076686e-05, "loss": 0.0405, "step": 104190 }, { "grad_norm": 0.26191213726997375, "learning_rate": 5.066219530216068e-05, "loss": 0.0335, "step": 104200 }, { "grad_norm": 0.28184589743614197, "learning_rate": 5.0653928669140514e-05, "loss": 0.0237, "step": 104210 }, { "grad_norm": 0.36401495337486267, "learning_rate": 5.064566201824217e-05, "loss": 0.0308, "step": 104220 }, { "grad_norm": 1.2679369449615479, "learning_rate": 5.063739534969169e-05, "loss": 0.0262, "step": 104230 }, { "grad_norm": 0.6316787004470825, "learning_rate": 5.062912866371503e-05, "loss": 0.0227, "step": 104240 }, { "grad_norm": 0.8439885973930359, "learning_rate": 5.062086196053826e-05, "loss": 0.0296, "step": 104250 }, { "grad_norm": 0.7486910820007324, "learning_rate": 5.061259524038734e-05, "loss": 0.052, "step": 104260 }, { "grad_norm": 0.8062781691551208, "learning_rate": 5.06043285034883e-05, "loss": 0.0321, "step": 104270 }, { "grad_norm": 0.34501251578330994, "learning_rate": 5.059606175006714e-05, "loss": 0.0307, "step": 104280 }, { "grad_norm": 0.35686635971069336, "learning_rate": 5.058779498034989e-05, "loss": 0.0219, "step": 104290 }, { "grad_norm": 0.4919479191303253, "learning_rate": 5.057952819456253e-05, "loss": 0.0247, "step": 104300 }, { "grad_norm": 0.5348969101905823, "learning_rate": 5.057126139293109e-05, "loss": 0.0198, "step": 104310 }, { "grad_norm": 0.24959100782871246, "learning_rate": 5.05629945756816e-05, "loss": 0.0236, "step": 104320 }, { "grad_norm": 0.35326507687568665, "learning_rate": 5.055472774304002e-05, "loss": 0.02, "step": 104330 }, { "grad_norm": 1.7314648628234863, "learning_rate": 5.054646089523241e-05, "loss": 0.0234, "step": 104340 }, { "grad_norm": 0.33520451188087463, "learning_rate": 5.053819403248475e-05, "loss": 0.04, "step": 104350 }, { "grad_norm": 0.8324881792068481, "learning_rate": 5.052992715502308e-05, "loss": 0.0316, "step": 104360 }, { "grad_norm": 0.7252883911132812, "learning_rate": 5.05216602630734e-05, "loss": 0.0311, "step": 104370 }, { "grad_norm": 0.33000218868255615, "learning_rate": 5.051339335686173e-05, "loss": 0.0345, "step": 104380 }, { "grad_norm": 1.158879041671753, "learning_rate": 5.050512643661406e-05, "loss": 0.0389, "step": 104390 }, { "grad_norm": 0.4351375997066498, "learning_rate": 5.049685950255645e-05, "loss": 0.0285, "step": 104400 }, { "grad_norm": 0.5394601821899414, "learning_rate": 5.0488592554914846e-05, "loss": 0.0254, "step": 104410 }, { "grad_norm": 0.49109768867492676, "learning_rate": 5.0480325593915345e-05, "loss": 0.0258, "step": 104420 }, { "grad_norm": 0.3760494589805603, "learning_rate": 5.04720586197839e-05, "loss": 0.0283, "step": 104430 }, { "grad_norm": 1.2791920900344849, "learning_rate": 5.046379163274656e-05, "loss": 0.0414, "step": 104440 }, { "grad_norm": 1.1426808834075928, "learning_rate": 5.0455524633029325e-05, "loss": 0.042, "step": 104450 }, { "grad_norm": 0.3780374228954315, "learning_rate": 5.044725762085823e-05, "loss": 0.038, "step": 104460 }, { "grad_norm": 0.6401551961898804, "learning_rate": 5.043899059645926e-05, "loss": 0.0277, "step": 104470 }, { "grad_norm": 0.4346177577972412, "learning_rate": 5.043072356005847e-05, "loss": 0.0315, "step": 104480 }, { "grad_norm": 0.4547555148601532, "learning_rate": 5.0422456511881844e-05, "loss": 0.0236, "step": 104490 }, { "grad_norm": 0.41987717151641846, "learning_rate": 5.0414189452155436e-05, "loss": 0.0285, "step": 104500 }, { "grad_norm": 0.4913521409034729, "learning_rate": 5.0405922381105234e-05, "loss": 0.0277, "step": 104510 }, { "grad_norm": 0.7014657855033875, "learning_rate": 5.039765529895726e-05, "loss": 0.0338, "step": 104520 }, { "grad_norm": 0.3987468183040619, "learning_rate": 5.038938820593754e-05, "loss": 0.032, "step": 104530 }, { "grad_norm": 1.118714451789856, "learning_rate": 5.03811211022721e-05, "loss": 0.0394, "step": 104540 }, { "grad_norm": 0.8523377776145935, "learning_rate": 5.037285398818694e-05, "loss": 0.0234, "step": 104550 }, { "grad_norm": 0.5674129724502563, "learning_rate": 5.0364586863908104e-05, "loss": 0.0286, "step": 104560 }, { "grad_norm": 0.46509990096092224, "learning_rate": 5.035631972966161e-05, "loss": 0.0311, "step": 104570 }, { "grad_norm": 0.6433192491531372, "learning_rate": 5.0348052585673446e-05, "loss": 0.0383, "step": 104580 }, { "grad_norm": 0.7301297783851624, "learning_rate": 5.0339785432169675e-05, "loss": 0.0256, "step": 104590 }, { "grad_norm": 0.48123663663864136, "learning_rate": 5.0331518269376264e-05, "loss": 0.0324, "step": 104600 }, { "grad_norm": 2.981180429458618, "learning_rate": 5.032325109751931e-05, "loss": 0.0244, "step": 104610 }, { "grad_norm": 0.6322000622749329, "learning_rate": 5.0314983916824764e-05, "loss": 0.032, "step": 104620 }, { "grad_norm": 0.3874950706958771, "learning_rate": 5.03067167275187e-05, "loss": 0.0333, "step": 104630 }, { "grad_norm": 0.3648187518119812, "learning_rate": 5.02984495298271e-05, "loss": 0.0296, "step": 104640 }, { "grad_norm": 0.5068695545196533, "learning_rate": 5.029018232397601e-05, "loss": 0.0549, "step": 104650 }, { "grad_norm": 0.6697445511817932, "learning_rate": 5.028191511019142e-05, "loss": 0.0262, "step": 104660 }, { "grad_norm": 0.508019208908081, "learning_rate": 5.027364788869941e-05, "loss": 0.024, "step": 104670 }, { "grad_norm": 0.5104590058326721, "learning_rate": 5.026538065972595e-05, "loss": 0.0382, "step": 104680 }, { "grad_norm": 0.7397738695144653, "learning_rate": 5.025711342349709e-05, "loss": 0.0232, "step": 104690 }, { "grad_norm": 0.4038258194923401, "learning_rate": 5.024884618023884e-05, "loss": 0.0306, "step": 104700 }, { "grad_norm": 0.7059868574142456, "learning_rate": 5.024057893017724e-05, "loss": 0.0274, "step": 104710 }, { "grad_norm": 0.7238040566444397, "learning_rate": 5.023231167353829e-05, "loss": 0.0296, "step": 104720 }, { "grad_norm": 0.4321790635585785, "learning_rate": 5.022404441054802e-05, "loss": 0.0231, "step": 104730 }, { "grad_norm": 0.682414174079895, "learning_rate": 5.021577714143247e-05, "loss": 0.036, "step": 104740 }, { "grad_norm": 0.48245519399642944, "learning_rate": 5.020750986641766e-05, "loss": 0.0345, "step": 104750 }, { "grad_norm": 0.6446180939674377, "learning_rate": 5.01992425857296e-05, "loss": 0.0351, "step": 104760 }, { "grad_norm": 0.5454267263412476, "learning_rate": 5.019097529959432e-05, "loss": 0.0289, "step": 104770 }, { "grad_norm": 0.441854864358902, "learning_rate": 5.018270800823787e-05, "loss": 0.0225, "step": 104780 }, { "grad_norm": 0.9937233328819275, "learning_rate": 5.017444071188622e-05, "loss": 0.031, "step": 104790 }, { "grad_norm": 0.2567211091518402, "learning_rate": 5.0166173410765437e-05, "loss": 0.0373, "step": 104800 }, { "grad_norm": 0.6831810474395752, "learning_rate": 5.0157906105101535e-05, "loss": 0.0269, "step": 104810 }, { "grad_norm": 1.1218823194503784, "learning_rate": 5.014963879512056e-05, "loss": 0.0232, "step": 104820 }, { "grad_norm": 0.6974474787712097, "learning_rate": 5.014137148104849e-05, "loss": 0.0363, "step": 104830 }, { "grad_norm": 0.8107836842536926, "learning_rate": 5.013310416311139e-05, "loss": 0.0325, "step": 104840 }, { "grad_norm": 0.26583731174468994, "learning_rate": 5.012483684153526e-05, "loss": 0.0244, "step": 104850 }, { "grad_norm": 0.44884106516838074, "learning_rate": 5.011656951654615e-05, "loss": 0.0268, "step": 104860 }, { "grad_norm": 0.437888503074646, "learning_rate": 5.010830218837007e-05, "loss": 0.0296, "step": 104870 }, { "grad_norm": 0.6395495533943176, "learning_rate": 5.0100034857233044e-05, "loss": 0.0305, "step": 104880 }, { "grad_norm": 0.5063795447349548, "learning_rate": 5.009176752336111e-05, "loss": 0.0243, "step": 104890 }, { "grad_norm": 0.3238978087902069, "learning_rate": 5.008350018698029e-05, "loss": 0.0279, "step": 104900 }, { "grad_norm": 1.1748034954071045, "learning_rate": 5.00752328483166e-05, "loss": 0.0269, "step": 104910 }, { "grad_norm": 0.7963975071907043, "learning_rate": 5.0066965507596065e-05, "loss": 0.0463, "step": 104920 }, { "grad_norm": 0.7237992286682129, "learning_rate": 5.0058698165044726e-05, "loss": 0.0353, "step": 104930 }, { "grad_norm": 0.44562727212905884, "learning_rate": 5.0050430820888604e-05, "loss": 0.0271, "step": 104940 }, { "grad_norm": 1.2052663564682007, "learning_rate": 5.004216347535371e-05, "loss": 0.0369, "step": 104950 }, { "grad_norm": 0.48643505573272705, "learning_rate": 5.0033896128666105e-05, "loss": 0.0465, "step": 104960 }, { "grad_norm": 0.5792027711868286, "learning_rate": 5.002562878105177e-05, "loss": 0.0556, "step": 104970 }, { "grad_norm": 0.405976802110672, "learning_rate": 5.0017361432736766e-05, "loss": 0.0294, "step": 104980 }, { "grad_norm": 0.38718029856681824, "learning_rate": 5.0009094083947096e-05, "loss": 0.0182, "step": 104990 }, { "grad_norm": 0.9738873839378357, "learning_rate": 5.000082673490881e-05, "loss": 0.0278, "step": 105000 }, { "grad_norm": 0.4977869689464569, "learning_rate": 4.99925593858479e-05, "loss": 0.0343, "step": 105010 }, { "grad_norm": 0.473842978477478, "learning_rate": 4.998429203699044e-05, "loss": 0.0267, "step": 105020 }, { "grad_norm": 0.4563935101032257, "learning_rate": 4.997602468856241e-05, "loss": 0.0234, "step": 105030 }, { "grad_norm": 0.5909298658370972, "learning_rate": 4.996775734078988e-05, "loss": 0.0256, "step": 105040 }, { "grad_norm": 0.55263352394104, "learning_rate": 4.995948999389883e-05, "loss": 0.0244, "step": 105050 }, { "grad_norm": 0.34469011425971985, "learning_rate": 4.995122264811533e-05, "loss": 0.0292, "step": 105060 }, { "grad_norm": 2.647970676422119, "learning_rate": 4.994295530366536e-05, "loss": 0.0339, "step": 105070 }, { "grad_norm": 0.64296555519104, "learning_rate": 4.993468796077498e-05, "loss": 0.0273, "step": 105080 }, { "grad_norm": 0.3445640206336975, "learning_rate": 4.9926420619670214e-05, "loss": 0.0253, "step": 105090 }, { "grad_norm": 0.36675551533699036, "learning_rate": 4.991815328057706e-05, "loss": 0.0302, "step": 105100 }, { "grad_norm": 0.39466872811317444, "learning_rate": 4.9909885943721596e-05, "loss": 0.0336, "step": 105110 }, { "grad_norm": 0.4564167261123657, "learning_rate": 4.990161860932978e-05, "loss": 0.032, "step": 105120 }, { "grad_norm": 0.36580395698547363, "learning_rate": 4.989335127762771e-05, "loss": 0.0321, "step": 105130 }, { "grad_norm": 0.33776143193244934, "learning_rate": 4.988508394884135e-05, "loss": 0.0302, "step": 105140 }, { "grad_norm": 0.5293816328048706, "learning_rate": 4.987681662319675e-05, "loss": 0.0268, "step": 105150 }, { "grad_norm": 0.4733603000640869, "learning_rate": 4.986854930091996e-05, "loss": 0.0364, "step": 105160 }, { "grad_norm": 0.47252804040908813, "learning_rate": 4.9860281982236965e-05, "loss": 0.0248, "step": 105170 }, { "grad_norm": 1.567253828048706, "learning_rate": 4.9852014667373825e-05, "loss": 0.026, "step": 105180 }, { "grad_norm": 0.6105237603187561, "learning_rate": 4.9843747356556534e-05, "loss": 0.022, "step": 105190 }, { "grad_norm": 0.6417559385299683, "learning_rate": 4.983548005001112e-05, "loss": 0.0301, "step": 105200 }, { "grad_norm": 0.8059595227241516, "learning_rate": 4.982721274796365e-05, "loss": 0.0382, "step": 105210 }, { "grad_norm": 0.5353494882583618, "learning_rate": 4.981894545064009e-05, "loss": 0.0298, "step": 105220 }, { "grad_norm": 0.4170162081718445, "learning_rate": 4.9810678158266514e-05, "loss": 0.0243, "step": 105230 }, { "grad_norm": 0.41744112968444824, "learning_rate": 4.980241087106891e-05, "loss": 0.0245, "step": 105240 }, { "grad_norm": 0.4871212840080261, "learning_rate": 4.979414358927334e-05, "loss": 0.0298, "step": 105250 }, { "grad_norm": 0.53499835729599, "learning_rate": 4.978587631310579e-05, "loss": 0.0296, "step": 105260 }, { "grad_norm": 0.7162114381790161, "learning_rate": 4.97776090427923e-05, "loss": 0.0344, "step": 105270 }, { "grad_norm": 0.3201374113559723, "learning_rate": 4.976934177855891e-05, "loss": 0.0331, "step": 105280 }, { "grad_norm": 0.37174028158187866, "learning_rate": 4.9761074520631616e-05, "loss": 0.0353, "step": 105290 }, { "grad_norm": 0.29675108194351196, "learning_rate": 4.9752807269236464e-05, "loss": 0.0253, "step": 105300 }, { "grad_norm": 0.41410553455352783, "learning_rate": 4.974454002459945e-05, "loss": 0.0316, "step": 105310 }, { "grad_norm": 1.0791805982589722, "learning_rate": 4.973627278694665e-05, "loss": 0.0364, "step": 105320 }, { "grad_norm": 0.9423328638076782, "learning_rate": 4.972800555650402e-05, "loss": 0.0329, "step": 105330 }, { "grad_norm": 0.39038050174713135, "learning_rate": 4.9719738333497615e-05, "loss": 0.0371, "step": 105340 }, { "grad_norm": 0.35706058144569397, "learning_rate": 4.971147111815348e-05, "loss": 0.0279, "step": 105350 }, { "grad_norm": 0.4246433973312378, "learning_rate": 4.970320391069761e-05, "loss": 0.026, "step": 105360 }, { "grad_norm": 0.3022223114967346, "learning_rate": 4.969493671135603e-05, "loss": 0.0257, "step": 105370 }, { "grad_norm": 0.4990835189819336, "learning_rate": 4.9686669520354764e-05, "loss": 0.0213, "step": 105380 }, { "grad_norm": 2.2487800121307373, "learning_rate": 4.9678402337919824e-05, "loss": 0.038, "step": 105390 }, { "grad_norm": 0.9268823266029358, "learning_rate": 4.9670135164277276e-05, "loss": 0.0349, "step": 105400 }, { "grad_norm": 1.687260627746582, "learning_rate": 4.966186799965308e-05, "loss": 0.0294, "step": 105410 }, { "grad_norm": 0.4115316569805145, "learning_rate": 4.9653600844273316e-05, "loss": 0.03, "step": 105420 }, { "grad_norm": 0.5247445106506348, "learning_rate": 4.9645333698363934e-05, "loss": 0.0344, "step": 105430 }, { "grad_norm": 1.1301624774932861, "learning_rate": 4.963706656215103e-05, "loss": 0.0293, "step": 105440 }, { "grad_norm": 0.9461063146591187, "learning_rate": 4.962879943586057e-05, "loss": 0.0255, "step": 105450 }, { "grad_norm": 0.37825286388397217, "learning_rate": 4.9620532319718594e-05, "loss": 0.0281, "step": 105460 }, { "grad_norm": 1.3648139238357544, "learning_rate": 4.9612265213951134e-05, "loss": 0.0258, "step": 105470 }, { "grad_norm": 0.6861958503723145, "learning_rate": 4.960399811878419e-05, "loss": 0.0227, "step": 105480 }, { "grad_norm": 0.473786860704422, "learning_rate": 4.959573103444379e-05, "loss": 0.0309, "step": 105490 }, { "grad_norm": 0.5265836715698242, "learning_rate": 4.958746396115596e-05, "loss": 0.0282, "step": 105500 }, { "grad_norm": 0.6852861642837524, "learning_rate": 4.95791968991467e-05, "loss": 0.0262, "step": 105510 }, { "grad_norm": 0.5684663653373718, "learning_rate": 4.957092984864203e-05, "loss": 0.0252, "step": 105520 }, { "grad_norm": 0.5906791687011719, "learning_rate": 4.956266280986798e-05, "loss": 0.0337, "step": 105530 }, { "grad_norm": 0.5228275656700134, "learning_rate": 4.9554395783050586e-05, "loss": 0.0356, "step": 105540 }, { "grad_norm": 0.38992804288864136, "learning_rate": 4.954612876841582e-05, "loss": 0.0213, "step": 105550 }, { "grad_norm": 0.30284616351127625, "learning_rate": 4.953786176618974e-05, "loss": 0.0306, "step": 105560 }, { "grad_norm": 0.8759164214134216, "learning_rate": 4.9529594776598334e-05, "loss": 0.0454, "step": 105570 }, { "grad_norm": 0.3731100559234619, "learning_rate": 4.952132779986764e-05, "loss": 0.036, "step": 105580 }, { "grad_norm": 0.4575427174568176, "learning_rate": 4.951306083622366e-05, "loss": 0.0363, "step": 105590 }, { "grad_norm": 0.8458718061447144, "learning_rate": 4.9504793885892395e-05, "loss": 0.0288, "step": 105600 }, { "grad_norm": 0.5755820274353027, "learning_rate": 4.9496526949099917e-05, "loss": 0.0284, "step": 105610 }, { "grad_norm": 0.7787804007530212, "learning_rate": 4.9488260026072164e-05, "loss": 0.0249, "step": 105620 }, { "grad_norm": 0.2735286056995392, "learning_rate": 4.947999311703523e-05, "loss": 0.0186, "step": 105630 }, { "grad_norm": 0.4022873640060425, "learning_rate": 4.9471726222215056e-05, "loss": 0.0248, "step": 105640 }, { "grad_norm": 0.3027130365371704, "learning_rate": 4.946345934183769e-05, "loss": 0.0221, "step": 105650 }, { "grad_norm": 0.41719233989715576, "learning_rate": 4.9455192476129165e-05, "loss": 0.0316, "step": 105660 }, { "grad_norm": 0.6582747101783752, "learning_rate": 4.944692562531546e-05, "loss": 0.0353, "step": 105670 }, { "grad_norm": 1.561692714691162, "learning_rate": 4.94386587896226e-05, "loss": 0.0277, "step": 105680 }, { "grad_norm": 1.0156266689300537, "learning_rate": 4.94303919692766e-05, "loss": 0.0245, "step": 105690 }, { "grad_norm": 0.3452755808830261, "learning_rate": 4.942212516450348e-05, "loss": 0.0197, "step": 105700 }, { "grad_norm": 0.550184428691864, "learning_rate": 4.941385837552923e-05, "loss": 0.0387, "step": 105710 }, { "grad_norm": 0.5307139158248901, "learning_rate": 4.940559160257985e-05, "loss": 0.0329, "step": 105720 }, { "grad_norm": 1.142878770828247, "learning_rate": 4.9397324845881404e-05, "loss": 0.0343, "step": 105730 }, { "grad_norm": 0.6006519794464111, "learning_rate": 4.9389058105659855e-05, "loss": 0.04, "step": 105740 }, { "grad_norm": 0.507453441619873, "learning_rate": 4.938079138214125e-05, "loss": 0.0339, "step": 105750 }, { "grad_norm": 0.5413435697555542, "learning_rate": 4.937252467555155e-05, "loss": 0.0225, "step": 105760 }, { "grad_norm": 0.3102702796459198, "learning_rate": 4.936425798611681e-05, "loss": 0.029, "step": 105770 }, { "grad_norm": 0.8665104508399963, "learning_rate": 4.9355991314063e-05, "loss": 0.029, "step": 105780 }, { "grad_norm": 0.19627375900745392, "learning_rate": 4.9347724659616155e-05, "loss": 0.0218, "step": 105790 }, { "grad_norm": 0.373982697725296, "learning_rate": 4.933945802300227e-05, "loss": 0.0361, "step": 105800 }, { "grad_norm": 0.5557425022125244, "learning_rate": 4.933119140444734e-05, "loss": 0.0275, "step": 105810 }, { "grad_norm": 0.3037565052509308, "learning_rate": 4.9322924804177424e-05, "loss": 0.0248, "step": 105820 }, { "grad_norm": 0.3131803572177887, "learning_rate": 4.931465822241845e-05, "loss": 0.0268, "step": 105830 }, { "grad_norm": 0.6554725766181946, "learning_rate": 4.9306391659396476e-05, "loss": 0.0241, "step": 105840 }, { "grad_norm": 0.6020379066467285, "learning_rate": 4.92981251153375e-05, "loss": 0.0308, "step": 105850 }, { "grad_norm": 0.6630390882492065, "learning_rate": 4.928985859046752e-05, "loss": 0.0324, "step": 105860 }, { "grad_norm": 1.0282368659973145, "learning_rate": 4.928159208501254e-05, "loss": 0.0454, "step": 105870 }, { "grad_norm": 0.7772493362426758, "learning_rate": 4.927332559919855e-05, "loss": 0.037, "step": 105880 }, { "grad_norm": 0.4264490306377411, "learning_rate": 4.926505913325158e-05, "loss": 0.0449, "step": 105890 }, { "grad_norm": 0.22189779579639435, "learning_rate": 4.9256792687397605e-05, "loss": 0.0283, "step": 105900 }, { "grad_norm": 0.711218535900116, "learning_rate": 4.9248526261862633e-05, "loss": 0.0216, "step": 105910 }, { "grad_norm": 0.35351672768592834, "learning_rate": 4.92402598568727e-05, "loss": 0.0212, "step": 105920 }, { "grad_norm": 0.3411169946193695, "learning_rate": 4.923199347265374e-05, "loss": 0.0264, "step": 105930 }, { "grad_norm": 0.44877514243125916, "learning_rate": 4.922372710943182e-05, "loss": 0.0492, "step": 105940 }, { "grad_norm": 0.6319447159767151, "learning_rate": 4.92154607674329e-05, "loss": 0.0298, "step": 105950 }, { "grad_norm": 0.378495991230011, "learning_rate": 4.9207194446882984e-05, "loss": 0.0346, "step": 105960 }, { "grad_norm": 0.4926080107688904, "learning_rate": 4.9198928148008075e-05, "loss": 0.0275, "step": 105970 }, { "grad_norm": 0.23142457008361816, "learning_rate": 4.9190661871034164e-05, "loss": 0.0196, "step": 105980 }, { "grad_norm": 0.3739964962005615, "learning_rate": 4.918239561618727e-05, "loss": 0.0241, "step": 105990 }, { "grad_norm": 0.3592219352722168, "learning_rate": 4.9174129383693355e-05, "loss": 0.0182, "step": 106000 }, { "grad_norm": 0.6116336584091187, "learning_rate": 4.916586317377845e-05, "loss": 0.0267, "step": 106010 }, { "grad_norm": 0.7545398473739624, "learning_rate": 4.915759698666851e-05, "loss": 0.0302, "step": 106020 }, { "grad_norm": 0.41875961422920227, "learning_rate": 4.914933082258956e-05, "loss": 0.0244, "step": 106030 }, { "grad_norm": 0.7333944439888, "learning_rate": 4.9141064681767594e-05, "loss": 0.0525, "step": 106040 }, { "grad_norm": 0.21360336244106293, "learning_rate": 4.913279856442859e-05, "loss": 0.0168, "step": 106050 }, { "grad_norm": 0.597303569316864, "learning_rate": 4.912453247079856e-05, "loss": 0.0377, "step": 106060 }, { "grad_norm": 0.8332979679107666, "learning_rate": 4.911626640110347e-05, "loss": 0.0362, "step": 106070 }, { "grad_norm": 0.8792454600334167, "learning_rate": 4.910800035556933e-05, "loss": 0.0499, "step": 106080 }, { "grad_norm": 0.3147227466106415, "learning_rate": 4.909973433442212e-05, "loss": 0.0261, "step": 106090 }, { "grad_norm": 0.774227499961853, "learning_rate": 4.909146833788783e-05, "loss": 0.0221, "step": 106100 }, { "grad_norm": 0.8846451044082642, "learning_rate": 4.908320236619248e-05, "loss": 0.0251, "step": 106110 }, { "grad_norm": 0.6701631546020508, "learning_rate": 4.9074936419562e-05, "loss": 0.0214, "step": 106120 }, { "grad_norm": 0.7615146040916443, "learning_rate": 4.906667049822244e-05, "loss": 0.0346, "step": 106130 }, { "grad_norm": 0.5596343278884888, "learning_rate": 4.905840460239973e-05, "loss": 0.0239, "step": 106140 }, { "grad_norm": 0.4523199498653412, "learning_rate": 4.905013873231991e-05, "loss": 0.0446, "step": 106150 }, { "grad_norm": 0.8550455570220947, "learning_rate": 4.9041872888208925e-05, "loss": 0.0572, "step": 106160 }, { "grad_norm": 0.43008410930633545, "learning_rate": 4.9033607070292774e-05, "loss": 0.034, "step": 106170 }, { "grad_norm": 0.7107563018798828, "learning_rate": 4.902534127879746e-05, "loss": 0.03, "step": 106180 }, { "grad_norm": 0.6729216575622559, "learning_rate": 4.901707551394893e-05, "loss": 0.0308, "step": 106190 }, { "grad_norm": 1.0710309743881226, "learning_rate": 4.90088097759732e-05, "loss": 0.037, "step": 106200 }, { "grad_norm": 1.2619833946228027, "learning_rate": 4.900054406509622e-05, "loss": 0.0328, "step": 106210 }, { "grad_norm": 1.4528895616531372, "learning_rate": 4.8992278381543994e-05, "loss": 0.0299, "step": 106220 }, { "grad_norm": 0.7860298752784729, "learning_rate": 4.8984012725542514e-05, "loss": 0.0284, "step": 106230 }, { "grad_norm": 0.9961211681365967, "learning_rate": 4.897574709731773e-05, "loss": 0.0311, "step": 106240 }, { "grad_norm": 0.7360586524009705, "learning_rate": 4.8967481497095654e-05, "loss": 0.0291, "step": 106250 }, { "grad_norm": 0.3201325237751007, "learning_rate": 4.895921592510223e-05, "loss": 0.0287, "step": 106260 }, { "grad_norm": 0.2694544792175293, "learning_rate": 4.8950950381563456e-05, "loss": 0.0273, "step": 106270 }, { "grad_norm": 0.5843542814254761, "learning_rate": 4.8942684866705305e-05, "loss": 0.0204, "step": 106280 }, { "grad_norm": 0.5659225583076477, "learning_rate": 4.8934419380753747e-05, "loss": 0.0271, "step": 106290 }, { "grad_norm": 0.2232527881860733, "learning_rate": 4.8926153923934774e-05, "loss": 0.0268, "step": 106300 }, { "grad_norm": 1.662300705909729, "learning_rate": 4.891788849647433e-05, "loss": 0.0561, "step": 106310 }, { "grad_norm": 0.3528306484222412, "learning_rate": 4.8909623098598436e-05, "loss": 0.0391, "step": 106320 }, { "grad_norm": 0.7438812851905823, "learning_rate": 4.8901357730533006e-05, "loss": 0.0356, "step": 106330 }, { "grad_norm": 2.4512696266174316, "learning_rate": 4.889309239250408e-05, "loss": 0.0406, "step": 106340 }, { "grad_norm": 0.6046850681304932, "learning_rate": 4.888482708473755e-05, "loss": 0.0333, "step": 106350 }, { "grad_norm": 0.5251549482345581, "learning_rate": 4.8876561807459445e-05, "loss": 0.0384, "step": 106360 }, { "grad_norm": 1.6760867834091187, "learning_rate": 4.8868296560895734e-05, "loss": 0.0287, "step": 106370 }, { "grad_norm": 0.7335927486419678, "learning_rate": 4.8860031345272355e-05, "loss": 0.0305, "step": 106380 }, { "grad_norm": 0.7822898626327515, "learning_rate": 4.8851766160815296e-05, "loss": 0.0268, "step": 106390 }, { "grad_norm": 0.550775945186615, "learning_rate": 4.884350100775052e-05, "loss": 0.0338, "step": 106400 }, { "grad_norm": 0.3531721234321594, "learning_rate": 4.883523588630397e-05, "loss": 0.0315, "step": 106410 }, { "grad_norm": 0.6770938038825989, "learning_rate": 4.882697079670167e-05, "loss": 0.0342, "step": 106420 }, { "grad_norm": 0.6810230612754822, "learning_rate": 4.8818705739169516e-05, "loss": 0.0294, "step": 106430 }, { "grad_norm": 0.3457878530025482, "learning_rate": 4.881044071393352e-05, "loss": 0.0231, "step": 106440 }, { "grad_norm": 0.6082356572151184, "learning_rate": 4.880217572121962e-05, "loss": 0.0347, "step": 106450 }, { "grad_norm": 0.5604608058929443, "learning_rate": 4.87939107612538e-05, "loss": 0.0383, "step": 106460 }, { "grad_norm": 0.8910039663314819, "learning_rate": 4.878564583426199e-05, "loss": 0.026, "step": 106470 }, { "grad_norm": 0.8652605414390564, "learning_rate": 4.877738094047018e-05, "loss": 0.0241, "step": 106480 }, { "grad_norm": 0.30667170882225037, "learning_rate": 4.8769116080104317e-05, "loss": 0.0385, "step": 106490 }, { "grad_norm": 0.27629804611206055, "learning_rate": 4.876085125339034e-05, "loss": 0.0277, "step": 106500 }, { "grad_norm": 0.21317359805107117, "learning_rate": 4.875258646055424e-05, "loss": 0.0255, "step": 106510 }, { "grad_norm": 0.9806617498397827, "learning_rate": 4.8744321701821946e-05, "loss": 0.0281, "step": 106520 }, { "grad_norm": 0.2268470823764801, "learning_rate": 4.873605697741945e-05, "loss": 0.0228, "step": 106530 }, { "grad_norm": 0.3362351953983307, "learning_rate": 4.872779228757264e-05, "loss": 0.0223, "step": 106540 }, { "grad_norm": 0.4617019593715668, "learning_rate": 4.8719527632507536e-05, "loss": 0.0389, "step": 106550 }, { "grad_norm": 0.5478069186210632, "learning_rate": 4.871126301245007e-05, "loss": 0.0267, "step": 106560 }, { "grad_norm": 0.4751288890838623, "learning_rate": 4.870299842762618e-05, "loss": 0.0311, "step": 106570 }, { "grad_norm": 0.7189112901687622, "learning_rate": 4.8694733878261836e-05, "loss": 0.0332, "step": 106580 }, { "grad_norm": 0.5887150764465332, "learning_rate": 4.868646936458297e-05, "loss": 0.032, "step": 106590 }, { "grad_norm": 0.7840403914451599, "learning_rate": 4.867820488681551e-05, "loss": 0.0277, "step": 106600 }, { "grad_norm": 0.5614836812019348, "learning_rate": 4.866994044518548e-05, "loss": 0.0355, "step": 106610 }, { "grad_norm": 0.3497677743434906, "learning_rate": 4.866167603991875e-05, "loss": 0.0396, "step": 106620 }, { "grad_norm": 1.3449748754501343, "learning_rate": 4.8653411671241316e-05, "loss": 0.039, "step": 106630 }, { "grad_norm": 0.5935612320899963, "learning_rate": 4.864514733937907e-05, "loss": 0.0345, "step": 106640 }, { "grad_norm": 0.32030466198921204, "learning_rate": 4.8636883044558014e-05, "loss": 0.0301, "step": 106650 }, { "grad_norm": 0.5300431251525879, "learning_rate": 4.862861878700405e-05, "loss": 0.0265, "step": 106660 }, { "grad_norm": 0.3682636022567749, "learning_rate": 4.862035456694313e-05, "loss": 0.0296, "step": 106670 }, { "grad_norm": 1.140230417251587, "learning_rate": 4.861209038460122e-05, "loss": 0.0384, "step": 106680 }, { "grad_norm": 0.6419647932052612, "learning_rate": 4.8603826240204214e-05, "loss": 0.025, "step": 106690 }, { "grad_norm": 0.4386718273162842, "learning_rate": 4.859556213397809e-05, "loss": 0.0236, "step": 106700 }, { "grad_norm": 1.4955811500549316, "learning_rate": 4.8587298066148756e-05, "loss": 0.0359, "step": 106710 }, { "grad_norm": 0.4324389398097992, "learning_rate": 4.857903403694217e-05, "loss": 0.0293, "step": 106720 }, { "grad_norm": 0.33798539638519287, "learning_rate": 4.857077004658425e-05, "loss": 0.0223, "step": 106730 }, { "grad_norm": 0.5402151346206665, "learning_rate": 4.856250609530094e-05, "loss": 0.0248, "step": 106740 }, { "grad_norm": 0.5472843647003174, "learning_rate": 4.855424218331819e-05, "loss": 0.0282, "step": 106750 }, { "grad_norm": 0.30832335352897644, "learning_rate": 4.8545978310861904e-05, "loss": 0.0234, "step": 106760 }, { "grad_norm": 0.45074915885925293, "learning_rate": 4.853771447815803e-05, "loss": 0.0267, "step": 106770 }, { "grad_norm": 0.5135844945907593, "learning_rate": 4.8529450685432485e-05, "loss": 0.0185, "step": 106780 }, { "grad_norm": 0.45448699593544006, "learning_rate": 4.852118693291122e-05, "loss": 0.0212, "step": 106790 }, { "grad_norm": 0.8530377745628357, "learning_rate": 4.851292322082013e-05, "loss": 0.0222, "step": 106800 }, { "grad_norm": 0.4723137319087982, "learning_rate": 4.850465954938516e-05, "loss": 0.054, "step": 106810 }, { "grad_norm": 0.46205875277519226, "learning_rate": 4.849639591883226e-05, "loss": 0.0275, "step": 106820 }, { "grad_norm": 0.702517032623291, "learning_rate": 4.84881323293873e-05, "loss": 0.0258, "step": 106830 }, { "grad_norm": 0.440742552280426, "learning_rate": 4.847986878127627e-05, "loss": 0.0291, "step": 106840 }, { "grad_norm": 0.9558896422386169, "learning_rate": 4.847160527472503e-05, "loss": 0.0221, "step": 106850 }, { "grad_norm": 1.2127399444580078, "learning_rate": 4.846334180995953e-05, "loss": 0.0371, "step": 106860 }, { "grad_norm": 0.27006182074546814, "learning_rate": 4.84550783872057e-05, "loss": 0.0264, "step": 106870 }, { "grad_norm": 0.5938071012496948, "learning_rate": 4.844681500668943e-05, "loss": 0.0386, "step": 106880 }, { "grad_norm": 0.4288839101791382, "learning_rate": 4.8438551668636675e-05, "loss": 0.0324, "step": 106890 }, { "grad_norm": 0.27440592646598816, "learning_rate": 4.8430288373273315e-05, "loss": 0.0226, "step": 106900 }, { "grad_norm": 0.7205871939659119, "learning_rate": 4.8422025120825286e-05, "loss": 0.0406, "step": 106910 }, { "grad_norm": 1.6830275058746338, "learning_rate": 4.8413761911518494e-05, "loss": 0.0285, "step": 106920 }, { "grad_norm": 0.3775669038295746, "learning_rate": 4.840549874557885e-05, "loss": 0.0188, "step": 106930 }, { "grad_norm": 0.5091567635536194, "learning_rate": 4.839723562323228e-05, "loss": 0.022, "step": 106940 }, { "grad_norm": 0.2571712136268616, "learning_rate": 4.838897254470468e-05, "loss": 0.0215, "step": 106950 }, { "grad_norm": 0.4754674732685089, "learning_rate": 4.838070951022198e-05, "loss": 0.0322, "step": 106960 }, { "grad_norm": 0.4231500029563904, "learning_rate": 4.8372446520010064e-05, "loss": 0.0281, "step": 106970 }, { "grad_norm": 0.42809873819351196, "learning_rate": 4.836418357429485e-05, "loss": 0.0277, "step": 106980 }, { "grad_norm": 0.7560818791389465, "learning_rate": 4.835592067330224e-05, "loss": 0.0362, "step": 106990 }, { "grad_norm": 0.34824511408805847, "learning_rate": 4.834765781725814e-05, "loss": 0.0244, "step": 107000 }, { "grad_norm": 0.5278708338737488, "learning_rate": 4.833939500638846e-05, "loss": 0.0288, "step": 107010 }, { "grad_norm": 0.3520784378051758, "learning_rate": 4.833113224091908e-05, "loss": 0.0252, "step": 107020 }, { "grad_norm": 0.6444190144538879, "learning_rate": 4.832286952107595e-05, "loss": 0.0333, "step": 107030 }, { "grad_norm": 0.4176841974258423, "learning_rate": 4.8314606847084904e-05, "loss": 0.0233, "step": 107040 }, { "grad_norm": 0.9991766214370728, "learning_rate": 4.8306344219171885e-05, "loss": 0.0298, "step": 107050 }, { "grad_norm": 1.215772032737732, "learning_rate": 4.829808163756279e-05, "loss": 0.0459, "step": 107060 }, { "grad_norm": 0.4669964015483856, "learning_rate": 4.828981910248349e-05, "loss": 0.0276, "step": 107070 }, { "grad_norm": 0.6010569334030151, "learning_rate": 4.828155661415991e-05, "loss": 0.0247, "step": 107080 }, { "grad_norm": 0.4687210023403168, "learning_rate": 4.8273294172817904e-05, "loss": 0.0238, "step": 107090 }, { "grad_norm": 0.2528811991214752, "learning_rate": 4.82650317786834e-05, "loss": 0.0406, "step": 107100 }, { "grad_norm": 0.5260227918624878, "learning_rate": 4.825676943198228e-05, "loss": 0.0254, "step": 107110 }, { "grad_norm": 0.43744680285453796, "learning_rate": 4.824850713294041e-05, "loss": 0.0283, "step": 107120 }, { "grad_norm": 0.3335998058319092, "learning_rate": 4.824024488178372e-05, "loss": 0.0196, "step": 107130 }, { "grad_norm": 0.8219140768051147, "learning_rate": 4.823198267873805e-05, "loss": 0.0316, "step": 107140 }, { "grad_norm": 0.5930449962615967, "learning_rate": 4.822372052402933e-05, "loss": 0.0318, "step": 107150 }, { "grad_norm": 1.393778681755066, "learning_rate": 4.821545841788341e-05, "loss": 0.0257, "step": 107160 }, { "grad_norm": 0.4488127529621124, "learning_rate": 4.82071963605262e-05, "loss": 0.0298, "step": 107170 }, { "grad_norm": 0.37405502796173096, "learning_rate": 4.819893435218356e-05, "loss": 0.0192, "step": 107180 }, { "grad_norm": 1.0942167043685913, "learning_rate": 4.819067239308137e-05, "loss": 0.0363, "step": 107190 }, { "grad_norm": 0.6582757234573364, "learning_rate": 4.8182410483445534e-05, "loss": 0.026, "step": 107200 }, { "grad_norm": 0.4148084223270416, "learning_rate": 4.81741486235019e-05, "loss": 0.0264, "step": 107210 }, { "grad_norm": 0.7039453387260437, "learning_rate": 4.816588681347637e-05, "loss": 0.0272, "step": 107220 }, { "grad_norm": 0.47185179591178894, "learning_rate": 4.815762505359478e-05, "loss": 0.0385, "step": 107230 }, { "grad_norm": 0.6307738423347473, "learning_rate": 4.814936334408304e-05, "loss": 0.0364, "step": 107240 }, { "grad_norm": 0.34605735540390015, "learning_rate": 4.814110168516702e-05, "loss": 0.0213, "step": 107250 }, { "grad_norm": 0.25680071115493774, "learning_rate": 4.8132840077072575e-05, "loss": 0.026, "step": 107260 }, { "grad_norm": 0.2901225984096527, "learning_rate": 4.8124578520025595e-05, "loss": 0.0286, "step": 107270 }, { "grad_norm": 0.5057669878005981, "learning_rate": 4.811631701425192e-05, "loss": 0.0269, "step": 107280 }, { "grad_norm": 0.2516912519931793, "learning_rate": 4.810805555997744e-05, "loss": 0.0337, "step": 107290 }, { "grad_norm": 0.6651016473770142, "learning_rate": 4.8099794157428e-05, "loss": 0.0315, "step": 107300 }, { "grad_norm": 0.7204443216323853, "learning_rate": 4.8091532806829456e-05, "loss": 0.0303, "step": 107310 }, { "grad_norm": 1.3831534385681152, "learning_rate": 4.808327150840772e-05, "loss": 0.0427, "step": 107320 }, { "grad_norm": 0.328485906124115, "learning_rate": 4.80750102623886e-05, "loss": 0.0244, "step": 107330 }, { "grad_norm": 0.76142418384552, "learning_rate": 4.806674906899801e-05, "loss": 0.0297, "step": 107340 }, { "grad_norm": 0.3347899913787842, "learning_rate": 4.805848792846173e-05, "loss": 0.0229, "step": 107350 }, { "grad_norm": 0.480019748210907, "learning_rate": 4.8050226841005685e-05, "loss": 0.0342, "step": 107360 }, { "grad_norm": 0.5883878469467163, "learning_rate": 4.80419658068557e-05, "loss": 0.0302, "step": 107370 }, { "grad_norm": 0.529625654220581, "learning_rate": 4.803370482623763e-05, "loss": 0.0377, "step": 107380 }, { "grad_norm": 3.9445273876190186, "learning_rate": 4.802544389937734e-05, "loss": 0.0405, "step": 107390 }, { "grad_norm": 0.9243860840797424, "learning_rate": 4.8017183026500664e-05, "loss": 0.0353, "step": 107400 }, { "grad_norm": 0.32786914706230164, "learning_rate": 4.800892220783347e-05, "loss": 0.0294, "step": 107410 }, { "grad_norm": 0.2899012863636017, "learning_rate": 4.800066144360159e-05, "loss": 0.0237, "step": 107420 }, { "grad_norm": 0.6413326263427734, "learning_rate": 4.7992400734030856e-05, "loss": 0.039, "step": 107430 }, { "grad_norm": 0.7574646472930908, "learning_rate": 4.798414007934716e-05, "loss": 0.0311, "step": 107440 }, { "grad_norm": 0.6544246673583984, "learning_rate": 4.7975879479776306e-05, "loss": 0.0427, "step": 107450 }, { "grad_norm": 0.285429984331131, "learning_rate": 4.796761893554415e-05, "loss": 0.0311, "step": 107460 }, { "grad_norm": 0.4631941616535187, "learning_rate": 4.795935844687652e-05, "loss": 0.0307, "step": 107470 }, { "grad_norm": 0.616437554359436, "learning_rate": 4.7951098013999276e-05, "loss": 0.034, "step": 107480 }, { "grad_norm": 0.41114699840545654, "learning_rate": 4.794283763713823e-05, "loss": 0.0216, "step": 107490 }, { "grad_norm": 0.5579277873039246, "learning_rate": 4.7934577316519235e-05, "loss": 0.0349, "step": 107500 }, { "grad_norm": 0.8167315721511841, "learning_rate": 4.792631705236813e-05, "loss": 0.035, "step": 107510 }, { "grad_norm": 0.37683382630348206, "learning_rate": 4.7918056844910714e-05, "loss": 0.0454, "step": 107520 }, { "grad_norm": 0.43710029125213623, "learning_rate": 4.790979669437287e-05, "loss": 0.0245, "step": 107530 }, { "grad_norm": 2.8947906494140625, "learning_rate": 4.790153660098037e-05, "loss": 0.0426, "step": 107540 }, { "grad_norm": 0.5504262447357178, "learning_rate": 4.78932765649591e-05, "loss": 0.0311, "step": 107550 }, { "grad_norm": 0.426364928483963, "learning_rate": 4.788501658653484e-05, "loss": 0.0315, "step": 107560 }, { "grad_norm": 0.9329233169555664, "learning_rate": 4.787675666593343e-05, "loss": 0.0343, "step": 107570 }, { "grad_norm": 0.8712198734283447, "learning_rate": 4.786849680338071e-05, "loss": 0.0336, "step": 107580 }, { "grad_norm": 0.31251364946365356, "learning_rate": 4.786023699910247e-05, "loss": 0.0212, "step": 107590 }, { "grad_norm": 0.6141643524169922, "learning_rate": 4.785197725332456e-05, "loss": 0.0253, "step": 107600 }, { "grad_norm": 0.6032164692878723, "learning_rate": 4.7843717566272775e-05, "loss": 0.0334, "step": 107610 }, { "grad_norm": 0.4403866231441498, "learning_rate": 4.7835457938172936e-05, "loss": 0.034, "step": 107620 }, { "grad_norm": 0.3154582679271698, "learning_rate": 4.7827198369250884e-05, "loss": 0.0308, "step": 107630 }, { "grad_norm": 0.3042670488357544, "learning_rate": 4.7818938859732385e-05, "loss": 0.0239, "step": 107640 }, { "grad_norm": 0.47948309779167175, "learning_rate": 4.7810679409843304e-05, "loss": 0.0358, "step": 107650 }, { "grad_norm": 0.44121280312538147, "learning_rate": 4.7802420019809413e-05, "loss": 0.0199, "step": 107660 }, { "grad_norm": 1.1203186511993408, "learning_rate": 4.7794160689856546e-05, "loss": 0.0397, "step": 107670 }, { "grad_norm": 1.5600205659866333, "learning_rate": 4.7785901420210486e-05, "loss": 0.0248, "step": 107680 }, { "grad_norm": 0.3789680600166321, "learning_rate": 4.7777642211097046e-05, "loss": 0.031, "step": 107690 }, { "grad_norm": 2.688246011734009, "learning_rate": 4.7769383062742045e-05, "loss": 0.0273, "step": 107700 }, { "grad_norm": 0.28582921624183655, "learning_rate": 4.7761123975371255e-05, "loss": 0.0239, "step": 107710 }, { "grad_norm": 0.3268294930458069, "learning_rate": 4.775286494921052e-05, "loss": 0.0265, "step": 107720 }, { "grad_norm": 0.4398963749408722, "learning_rate": 4.774460598448558e-05, "loss": 0.0342, "step": 107730 }, { "grad_norm": 0.618939220905304, "learning_rate": 4.77363470814223e-05, "loss": 0.0468, "step": 107740 }, { "grad_norm": 0.21047353744506836, "learning_rate": 4.772808824024641e-05, "loss": 0.0267, "step": 107750 }, { "grad_norm": 0.6115229725837708, "learning_rate": 4.771982946118375e-05, "loss": 0.0249, "step": 107760 }, { "grad_norm": 1.1985220909118652, "learning_rate": 4.77115707444601e-05, "loss": 0.0307, "step": 107770 }, { "grad_norm": 0.2932494580745697, "learning_rate": 4.7703312090301234e-05, "loss": 0.0266, "step": 107780 }, { "grad_norm": 0.4961872100830078, "learning_rate": 4.769505349893296e-05, "loss": 0.0272, "step": 107790 }, { "grad_norm": 0.6195487976074219, "learning_rate": 4.7686794970581053e-05, "loss": 0.0269, "step": 107800 }, { "grad_norm": 0.4617358148097992, "learning_rate": 4.7678536505471295e-05, "loss": 0.0248, "step": 107810 }, { "grad_norm": 0.5728958249092102, "learning_rate": 4.7670278103829505e-05, "loss": 0.0333, "step": 107820 }, { "grad_norm": 0.6545266509056091, "learning_rate": 4.766201976588141e-05, "loss": 0.0266, "step": 107830 }, { "grad_norm": 0.3762604296207428, "learning_rate": 4.7653761491852836e-05, "loss": 0.0251, "step": 107840 }, { "grad_norm": 0.3125813901424408, "learning_rate": 4.7645503281969514e-05, "loss": 0.0321, "step": 107850 }, { "grad_norm": 0.705470621585846, "learning_rate": 4.763724513645727e-05, "loss": 0.0386, "step": 107860 }, { "grad_norm": 0.5757971405982971, "learning_rate": 4.7628987055541846e-05, "loss": 0.0482, "step": 107870 }, { "grad_norm": 0.41556280851364136, "learning_rate": 4.762072903944903e-05, "loss": 0.0276, "step": 107880 }, { "grad_norm": 0.8179870843887329, "learning_rate": 4.761247108840459e-05, "loss": 0.0315, "step": 107890 }, { "grad_norm": 1.0785152912139893, "learning_rate": 4.760421320263429e-05, "loss": 0.0413, "step": 107900 }, { "grad_norm": 0.47153323888778687, "learning_rate": 4.759595538236391e-05, "loss": 0.0345, "step": 107910 }, { "grad_norm": 0.5791379809379578, "learning_rate": 4.758769762781919e-05, "loss": 0.0426, "step": 107920 }, { "grad_norm": 0.6847044229507446, "learning_rate": 4.757943993922593e-05, "loss": 0.0428, "step": 107930 }, { "grad_norm": 0.379096120595932, "learning_rate": 4.7571182316809845e-05, "loss": 0.0247, "step": 107940 }, { "grad_norm": 0.5018504858016968, "learning_rate": 4.756292476079674e-05, "loss": 0.0275, "step": 107950 }, { "grad_norm": 0.48924267292022705, "learning_rate": 4.755466727141236e-05, "loss": 0.0281, "step": 107960 }, { "grad_norm": 0.31153589487075806, "learning_rate": 4.7546409848882454e-05, "loss": 0.038, "step": 107970 }, { "grad_norm": 0.26378563046455383, "learning_rate": 4.753815249343279e-05, "loss": 0.0299, "step": 107980 }, { "grad_norm": 0.5353710651397705, "learning_rate": 4.75298952052891e-05, "loss": 0.024, "step": 107990 }, { "grad_norm": 0.6055176854133606, "learning_rate": 4.752163798467715e-05, "loss": 0.0263, "step": 108000 }, { "grad_norm": 0.45161715149879456, "learning_rate": 4.751338083182269e-05, "loss": 0.0329, "step": 108010 }, { "grad_norm": 0.8685879111289978, "learning_rate": 4.750512374695145e-05, "loss": 0.0258, "step": 108020 }, { "grad_norm": 0.5016330480575562, "learning_rate": 4.749686673028921e-05, "loss": 0.0275, "step": 108030 }, { "grad_norm": 0.4780254065990448, "learning_rate": 4.7488609782061674e-05, "loss": 0.0276, "step": 108040 }, { "grad_norm": 0.6310915946960449, "learning_rate": 4.748035290249463e-05, "loss": 0.0501, "step": 108050 }, { "grad_norm": 0.32596564292907715, "learning_rate": 4.7472096091813755e-05, "loss": 0.0401, "step": 108060 }, { "grad_norm": 1.7842540740966797, "learning_rate": 4.746383935024484e-05, "loss": 0.0307, "step": 108070 }, { "grad_norm": 0.39183759689331055, "learning_rate": 4.745558267801362e-05, "loss": 0.0362, "step": 108080 }, { "grad_norm": 0.21127523481845856, "learning_rate": 4.7447326075345797e-05, "loss": 0.0259, "step": 108090 }, { "grad_norm": 0.45685014128685, "learning_rate": 4.7439069542467126e-05, "loss": 0.0204, "step": 108100 }, { "grad_norm": 0.34899961948394775, "learning_rate": 4.7430813079603325e-05, "loss": 0.0251, "step": 108110 }, { "grad_norm": 0.38044488430023193, "learning_rate": 4.742255668698014e-05, "loss": 0.0282, "step": 108120 }, { "grad_norm": 0.3985764682292938, "learning_rate": 4.7414300364823275e-05, "loss": 0.025, "step": 108130 }, { "grad_norm": 0.42470237612724304, "learning_rate": 4.740604411335845e-05, "loss": 0.02, "step": 108140 }, { "grad_norm": 1.2158983945846558, "learning_rate": 4.739778793281144e-05, "loss": 0.034, "step": 108150 }, { "grad_norm": 0.4238022267818451, "learning_rate": 4.738953182340791e-05, "loss": 0.0339, "step": 108160 }, { "grad_norm": 0.46974343061447144, "learning_rate": 4.7381275785373616e-05, "loss": 0.0352, "step": 108170 }, { "grad_norm": 0.45016393065452576, "learning_rate": 4.737301981893424e-05, "loss": 0.02, "step": 108180 }, { "grad_norm": 0.35429975390434265, "learning_rate": 4.736476392431553e-05, "loss": 0.0256, "step": 108190 }, { "grad_norm": 0.6006202697753906, "learning_rate": 4.735650810174317e-05, "loss": 0.0248, "step": 108200 }, { "grad_norm": 0.5022130608558655, "learning_rate": 4.73482523514429e-05, "loss": 0.0336, "step": 108210 }, { "grad_norm": 0.385897696018219, "learning_rate": 4.733999667364041e-05, "loss": 0.0274, "step": 108220 }, { "grad_norm": 0.41671252250671387, "learning_rate": 4.733174106856139e-05, "loss": 0.0247, "step": 108230 }, { "grad_norm": 0.48379290103912354, "learning_rate": 4.7323485536431604e-05, "loss": 0.0205, "step": 108240 }, { "grad_norm": 0.4334673285484314, "learning_rate": 4.731523007747669e-05, "loss": 0.0409, "step": 108250 }, { "grad_norm": 0.6202670335769653, "learning_rate": 4.730697469192238e-05, "loss": 0.0259, "step": 108260 }, { "grad_norm": 2.5293636322021484, "learning_rate": 4.729871937999438e-05, "loss": 0.0296, "step": 108270 }, { "grad_norm": 0.7746390700340271, "learning_rate": 4.729046414191837e-05, "loss": 0.0281, "step": 108280 }, { "grad_norm": 0.4316136837005615, "learning_rate": 4.7282208977920064e-05, "loss": 0.0334, "step": 108290 }, { "grad_norm": 0.5175660848617554, "learning_rate": 4.7273953888225135e-05, "loss": 0.0232, "step": 108300 }, { "grad_norm": 0.6085466146469116, "learning_rate": 4.7265698873059286e-05, "loss": 0.0588, "step": 108310 }, { "grad_norm": 0.6524634957313538, "learning_rate": 4.72574439326482e-05, "loss": 0.0345, "step": 108320 }, { "grad_norm": 0.5029075741767883, "learning_rate": 4.724918906721755e-05, "loss": 0.0307, "step": 108330 }, { "grad_norm": 0.40356898307800293, "learning_rate": 4.724093427699308e-05, "loss": 0.0257, "step": 108340 }, { "grad_norm": 0.4560963213443756, "learning_rate": 4.723267956220038e-05, "loss": 0.029, "step": 108350 }, { "grad_norm": 0.38588520884513855, "learning_rate": 4.722442492306522e-05, "loss": 0.0315, "step": 108360 }, { "grad_norm": 0.5363911986351013, "learning_rate": 4.7216170359813214e-05, "loss": 0.0311, "step": 108370 }, { "grad_norm": 0.4036805033683777, "learning_rate": 4.720791587267009e-05, "loss": 0.0252, "step": 108380 }, { "grad_norm": 0.47438758611679077, "learning_rate": 4.719966146186148e-05, "loss": 0.0251, "step": 108390 }, { "grad_norm": 0.32782208919525146, "learning_rate": 4.719140712761307e-05, "loss": 0.0375, "step": 108400 }, { "grad_norm": 0.46360793709754944, "learning_rate": 4.718315287015055e-05, "loss": 0.0295, "step": 108410 }, { "grad_norm": 1.49762761592865, "learning_rate": 4.717489868969956e-05, "loss": 0.0303, "step": 108420 }, { "grad_norm": 0.5667160153388977, "learning_rate": 4.7166644586485786e-05, "loss": 0.0318, "step": 108430 }, { "grad_norm": 0.45341622829437256, "learning_rate": 4.715839056073487e-05, "loss": 0.0292, "step": 108440 }, { "grad_norm": 0.6361327767372131, "learning_rate": 4.715013661267249e-05, "loss": 0.032, "step": 108450 }, { "grad_norm": 1.5680296421051025, "learning_rate": 4.7141882742524315e-05, "loss": 0.0347, "step": 108460 }, { "grad_norm": 0.3519858717918396, "learning_rate": 4.713362895051599e-05, "loss": 0.0208, "step": 108470 }, { "grad_norm": 2.9937517642974854, "learning_rate": 4.712537523687318e-05, "loss": 0.0561, "step": 108480 }, { "grad_norm": 0.19870249927043915, "learning_rate": 4.711712160182152e-05, "loss": 0.023, "step": 108490 }, { "grad_norm": 0.37011927366256714, "learning_rate": 4.710886804558669e-05, "loss": 0.0322, "step": 108500 }, { "grad_norm": 0.5969224572181702, "learning_rate": 4.7100614568394296e-05, "loss": 0.0191, "step": 108510 }, { "grad_norm": 0.4523763656616211, "learning_rate": 4.7092361170470005e-05, "loss": 0.0218, "step": 108520 }, { "grad_norm": 0.6936390995979309, "learning_rate": 4.7084107852039506e-05, "loss": 0.0308, "step": 108530 }, { "grad_norm": 0.2330576330423355, "learning_rate": 4.707585461332836e-05, "loss": 0.0233, "step": 108540 }, { "grad_norm": 0.547771155834198, "learning_rate": 4.706760145456229e-05, "loss": 0.03, "step": 108550 }, { "grad_norm": 0.29407331347465515, "learning_rate": 4.7059348375966866e-05, "loss": 0.0195, "step": 108560 }, { "grad_norm": 0.5276154279708862, "learning_rate": 4.705109537776777e-05, "loss": 0.0553, "step": 108570 }, { "grad_norm": 0.4245421588420868, "learning_rate": 4.7042842460190606e-05, "loss": 0.0216, "step": 108580 }, { "grad_norm": 0.7375426888465881, "learning_rate": 4.703458962346102e-05, "loss": 0.0242, "step": 108590 }, { "grad_norm": 0.25708264112472534, "learning_rate": 4.702633686780465e-05, "loss": 0.0312, "step": 108600 }, { "grad_norm": 0.4288790225982666, "learning_rate": 4.7018084193447105e-05, "loss": 0.0305, "step": 108610 }, { "grad_norm": 0.5356788635253906, "learning_rate": 4.7009831600614024e-05, "loss": 0.0335, "step": 108620 }, { "grad_norm": 0.6050592064857483, "learning_rate": 4.700157908953102e-05, "loss": 0.0267, "step": 108630 }, { "grad_norm": 0.5950007438659668, "learning_rate": 4.6993326660423695e-05, "loss": 0.029, "step": 108640 }, { "grad_norm": 0.8468236923217773, "learning_rate": 4.698507431351772e-05, "loss": 0.049, "step": 108650 }, { "grad_norm": 0.41159147024154663, "learning_rate": 4.697682204903867e-05, "loss": 0.0306, "step": 108660 }, { "grad_norm": 1.0892078876495361, "learning_rate": 4.696856986721217e-05, "loss": 0.0343, "step": 108670 }, { "grad_norm": 0.8124831914901733, "learning_rate": 4.696031776826383e-05, "loss": 0.0367, "step": 108680 }, { "grad_norm": 0.5402088761329651, "learning_rate": 4.6952065752419264e-05, "loss": 0.0374, "step": 108690 }, { "grad_norm": 0.25114428997039795, "learning_rate": 4.694381381990407e-05, "loss": 0.02, "step": 108700 }, { "grad_norm": 0.9127808809280396, "learning_rate": 4.693556197094386e-05, "loss": 0.0272, "step": 108710 }, { "grad_norm": 1.1838619709014893, "learning_rate": 4.692731020576424e-05, "loss": 0.0251, "step": 108720 }, { "grad_norm": 0.381521999835968, "learning_rate": 4.691905852459079e-05, "loss": 0.0238, "step": 108730 }, { "grad_norm": 0.44284093379974365, "learning_rate": 4.691080692764915e-05, "loss": 0.025, "step": 108740 }, { "grad_norm": 0.45732763409614563, "learning_rate": 4.6902555415164855e-05, "loss": 0.0499, "step": 108750 }, { "grad_norm": 0.3327353596687317, "learning_rate": 4.689430398736356e-05, "loss": 0.0287, "step": 108760 }, { "grad_norm": 0.2820563316345215, "learning_rate": 4.68860526444708e-05, "loss": 0.0342, "step": 108770 }, { "grad_norm": 0.5089445114135742, "learning_rate": 4.6877801386712204e-05, "loss": 0.029, "step": 108780 }, { "grad_norm": 0.43644142150878906, "learning_rate": 4.686955021431335e-05, "loss": 0.0375, "step": 108790 }, { "grad_norm": 0.33709633350372314, "learning_rate": 4.686129912749981e-05, "loss": 0.0198, "step": 108800 }, { "grad_norm": 0.26316067576408386, "learning_rate": 4.6853048126497186e-05, "loss": 0.0274, "step": 108810 }, { "grad_norm": 0.2856556475162506, "learning_rate": 4.684479721153103e-05, "loss": 0.0249, "step": 108820 }, { "grad_norm": 0.330665647983551, "learning_rate": 4.683654638282692e-05, "loss": 0.0244, "step": 108830 }, { "grad_norm": 0.4724010229110718, "learning_rate": 4.6828295640610476e-05, "loss": 0.0314, "step": 108840 }, { "grad_norm": 0.6098425984382629, "learning_rate": 4.682004498510721e-05, "loss": 0.0163, "step": 108850 }, { "grad_norm": 0.34938758611679077, "learning_rate": 4.681179441654274e-05, "loss": 0.038, "step": 108860 }, { "grad_norm": 0.42103642225265503, "learning_rate": 4.68035439351426e-05, "loss": 0.0525, "step": 108870 }, { "grad_norm": 0.370410680770874, "learning_rate": 4.679529354113238e-05, "loss": 0.0269, "step": 108880 }, { "grad_norm": 0.4729439616203308, "learning_rate": 4.678704323473761e-05, "loss": 0.0324, "step": 108890 }, { "grad_norm": 0.5236992239952087, "learning_rate": 4.677879301618389e-05, "loss": 0.0231, "step": 108900 }, { "grad_norm": 0.2828810214996338, "learning_rate": 4.677054288569675e-05, "loss": 0.0193, "step": 108910 }, { "grad_norm": 0.3361881673336029, "learning_rate": 4.676229284350176e-05, "loss": 0.0432, "step": 108920 }, { "grad_norm": 0.4584607183933258, "learning_rate": 4.675404288982447e-05, "loss": 0.024, "step": 108930 }, { "grad_norm": 0.4807671904563904, "learning_rate": 4.6745793024890405e-05, "loss": 0.0336, "step": 108940 }, { "grad_norm": 0.37463441491127014, "learning_rate": 4.673754324892517e-05, "loss": 0.029, "step": 108950 }, { "grad_norm": 0.6888691782951355, "learning_rate": 4.6729293562154247e-05, "loss": 0.027, "step": 108960 }, { "grad_norm": 0.49527883529663086, "learning_rate": 4.6721043964803215e-05, "loss": 0.0347, "step": 108970 }, { "grad_norm": 0.5192437171936035, "learning_rate": 4.6712794457097625e-05, "loss": 0.0303, "step": 108980 }, { "grad_norm": 0.4006785452365875, "learning_rate": 4.670454503926299e-05, "loss": 0.0265, "step": 108990 }, { "grad_norm": 0.3921319246292114, "learning_rate": 4.669629571152486e-05, "loss": 0.0269, "step": 109000 }, { "grad_norm": 0.3370904326438904, "learning_rate": 4.668804647410876e-05, "loss": 0.025, "step": 109010 }, { "grad_norm": 0.8160083293914795, "learning_rate": 4.667979732724021e-05, "loss": 0.0273, "step": 109020 }, { "grad_norm": 0.5835104584693909, "learning_rate": 4.667154827114479e-05, "loss": 0.0272, "step": 109030 }, { "grad_norm": 0.3873286843299866, "learning_rate": 4.6663299306047954e-05, "loss": 0.0328, "step": 109040 }, { "grad_norm": 0.3054172694683075, "learning_rate": 4.665505043217529e-05, "loss": 0.0244, "step": 109050 }, { "grad_norm": 0.3893195688724518, "learning_rate": 4.664680164975226e-05, "loss": 0.0289, "step": 109060 }, { "grad_norm": 0.5037330389022827, "learning_rate": 4.663855295900444e-05, "loss": 0.025, "step": 109070 }, { "grad_norm": 0.40018996596336365, "learning_rate": 4.663030436015731e-05, "loss": 0.0222, "step": 109080 }, { "grad_norm": 0.43566304445266724, "learning_rate": 4.6622055853436387e-05, "loss": 0.0414, "step": 109090 }, { "grad_norm": 0.32950371503829956, "learning_rate": 4.6613807439067196e-05, "loss": 0.0373, "step": 109100 }, { "grad_norm": 0.18945425748825073, "learning_rate": 4.660555911727523e-05, "loss": 0.0196, "step": 109110 }, { "grad_norm": 0.349918395280838, "learning_rate": 4.6597310888286016e-05, "loss": 0.0211, "step": 109120 }, { "grad_norm": 0.36119943857192993, "learning_rate": 4.658906275232502e-05, "loss": 0.0217, "step": 109130 }, { "grad_norm": 0.5626786947250366, "learning_rate": 4.65808147096178e-05, "loss": 0.0185, "step": 109140 }, { "grad_norm": 0.5189014077186584, "learning_rate": 4.6572566760389794e-05, "loss": 0.0228, "step": 109150 }, { "grad_norm": 0.2901208698749542, "learning_rate": 4.6564318904866534e-05, "loss": 0.036, "step": 109160 }, { "grad_norm": 0.3802047371864319, "learning_rate": 4.655607114327351e-05, "loss": 0.0293, "step": 109170 }, { "grad_norm": 0.4693841338157654, "learning_rate": 4.654782347583619e-05, "loss": 0.0282, "step": 109180 }, { "grad_norm": 0.30813002586364746, "learning_rate": 4.653957590278011e-05, "loss": 0.0221, "step": 109190 }, { "grad_norm": 0.4513208568096161, "learning_rate": 4.65313284243307e-05, "loss": 0.0253, "step": 109200 }, { "grad_norm": 1.178437352180481, "learning_rate": 4.6523081040713455e-05, "loss": 0.0404, "step": 109210 }, { "grad_norm": 0.5182855725288391, "learning_rate": 4.651483375215391e-05, "loss": 0.0357, "step": 109220 }, { "grad_norm": 0.605248749256134, "learning_rate": 4.650658655887746e-05, "loss": 0.0353, "step": 109230 }, { "grad_norm": 0.6263518929481506, "learning_rate": 4.649833946110966e-05, "loss": 0.0369, "step": 109240 }, { "grad_norm": 1.1237828731536865, "learning_rate": 4.649009245907591e-05, "loss": 0.0305, "step": 109250 }, { "grad_norm": 0.37274953722953796, "learning_rate": 4.648184555300174e-05, "loss": 0.0203, "step": 109260 }, { "grad_norm": 1.0324647426605225, "learning_rate": 4.6473598743112565e-05, "loss": 0.0306, "step": 109270 }, { "grad_norm": 0.3010449707508087, "learning_rate": 4.646535202963388e-05, "loss": 0.024, "step": 109280 }, { "grad_norm": 0.5248420834541321, "learning_rate": 4.645710541279116e-05, "loss": 0.0286, "step": 109290 }, { "grad_norm": 1.2345361709594727, "learning_rate": 4.6448858892809833e-05, "loss": 0.0307, "step": 109300 }, { "grad_norm": 0.5079322457313538, "learning_rate": 4.644061246991538e-05, "loss": 0.0253, "step": 109310 }, { "grad_norm": 3.2910373210906982, "learning_rate": 4.643236614433324e-05, "loss": 0.0262, "step": 109320 }, { "grad_norm": 0.4409549832344055, "learning_rate": 4.642411991628887e-05, "loss": 0.0212, "step": 109330 }, { "grad_norm": 0.3894885182380676, "learning_rate": 4.64158737860077e-05, "loss": 0.023, "step": 109340 }, { "grad_norm": 0.7153393030166626, "learning_rate": 4.640762775371521e-05, "loss": 0.0254, "step": 109350 }, { "grad_norm": 0.5295670032501221, "learning_rate": 4.639938181963684e-05, "loss": 0.0206, "step": 109360 }, { "grad_norm": 0.5885034203529358, "learning_rate": 4.6391135983998005e-05, "loss": 0.0389, "step": 109370 }, { "grad_norm": 0.7027223706245422, "learning_rate": 4.638289024702416e-05, "loss": 0.035, "step": 109380 }, { "grad_norm": 0.8424449563026428, "learning_rate": 4.637464460894073e-05, "loss": 0.0305, "step": 109390 }, { "grad_norm": 0.34003084897994995, "learning_rate": 4.636639906997316e-05, "loss": 0.0241, "step": 109400 }, { "grad_norm": 0.8388358354568481, "learning_rate": 4.635815363034688e-05, "loss": 0.0234, "step": 109410 }, { "grad_norm": 0.22793027758598328, "learning_rate": 4.634990829028729e-05, "loss": 0.0254, "step": 109420 }, { "grad_norm": 0.5191069841384888, "learning_rate": 4.6341663050019866e-05, "loss": 0.0329, "step": 109430 }, { "grad_norm": 0.781156599521637, "learning_rate": 4.633341790976997e-05, "loss": 0.0322, "step": 109440 }, { "grad_norm": 0.36418479681015015, "learning_rate": 4.632517286976308e-05, "loss": 0.0277, "step": 109450 }, { "grad_norm": 4.561463832855225, "learning_rate": 4.631692793022456e-05, "loss": 0.0333, "step": 109460 }, { "grad_norm": 0.5176601409912109, "learning_rate": 4.630868309137985e-05, "loss": 0.0357, "step": 109470 }, { "grad_norm": 0.584998607635498, "learning_rate": 4.630043835345437e-05, "loss": 0.0364, "step": 109480 }, { "grad_norm": 0.30980005860328674, "learning_rate": 4.6292193716673504e-05, "loss": 0.0236, "step": 109490 }, { "grad_norm": 0.38658106327056885, "learning_rate": 4.628394918126268e-05, "loss": 0.0235, "step": 109500 }, { "grad_norm": 0.7744444012641907, "learning_rate": 4.627570474744728e-05, "loss": 0.0244, "step": 109510 }, { "grad_norm": 0.4852181077003479, "learning_rate": 4.626746041545272e-05, "loss": 0.0325, "step": 109520 }, { "grad_norm": 1.2983163595199585, "learning_rate": 4.625921618550439e-05, "loss": 0.0346, "step": 109530 }, { "grad_norm": 0.7398869395256042, "learning_rate": 4.625097205782766e-05, "loss": 0.0361, "step": 109540 }, { "grad_norm": 0.4866495132446289, "learning_rate": 4.624272803264798e-05, "loss": 0.0215, "step": 109550 }, { "grad_norm": 0.9735373854637146, "learning_rate": 4.623448411019069e-05, "loss": 0.0349, "step": 109560 }, { "grad_norm": 0.3590456545352936, "learning_rate": 4.6226240290681186e-05, "loss": 0.0208, "step": 109570 }, { "grad_norm": 0.5663689374923706, "learning_rate": 4.621799657434486e-05, "loss": 0.0331, "step": 109580 }, { "grad_norm": 0.47087210416793823, "learning_rate": 4.620975296140709e-05, "loss": 0.0291, "step": 109590 }, { "grad_norm": 0.47100237011909485, "learning_rate": 4.620150945209324e-05, "loss": 0.0237, "step": 109600 }, { "grad_norm": 0.4319283664226532, "learning_rate": 4.61932660466287e-05, "loss": 0.0315, "step": 109610 }, { "grad_norm": 0.7623132467269897, "learning_rate": 4.6185022745238844e-05, "loss": 0.0245, "step": 109620 }, { "grad_norm": 0.7623129487037659, "learning_rate": 4.617677954814901e-05, "loss": 0.0286, "step": 109630 }, { "grad_norm": 0.8421018123626709, "learning_rate": 4.6168536455584615e-05, "loss": 0.0326, "step": 109640 }, { "grad_norm": 0.5793534517288208, "learning_rate": 4.6160293467770964e-05, "loss": 0.0358, "step": 109650 }, { "grad_norm": 0.7459642291069031, "learning_rate": 4.615205058493347e-05, "loss": 0.0411, "step": 109660 }, { "grad_norm": 1.0176347494125366, "learning_rate": 4.614380780729746e-05, "loss": 0.0288, "step": 109670 }, { "grad_norm": 0.5012000799179077, "learning_rate": 4.6135565135088294e-05, "loss": 0.0272, "step": 109680 }, { "grad_norm": 0.45801103115081787, "learning_rate": 4.6127322568531335e-05, "loss": 0.0215, "step": 109690 }, { "grad_norm": 0.4808562994003296, "learning_rate": 4.6119080107851915e-05, "loss": 0.0227, "step": 109700 }, { "grad_norm": 0.3827802836894989, "learning_rate": 4.611083775327539e-05, "loss": 0.0244, "step": 109710 }, { "grad_norm": 0.607575535774231, "learning_rate": 4.61025955050271e-05, "loss": 0.0186, "step": 109720 }, { "grad_norm": 0.48571398854255676, "learning_rate": 4.609435336333236e-05, "loss": 0.0265, "step": 109730 }, { "grad_norm": 0.5504584908485413, "learning_rate": 4.608611132841658e-05, "loss": 0.0281, "step": 109740 }, { "grad_norm": 0.6229567527770996, "learning_rate": 4.6077869400505e-05, "loss": 0.0234, "step": 109750 }, { "grad_norm": 0.545093297958374, "learning_rate": 4.606962757982302e-05, "loss": 0.0379, "step": 109760 }, { "grad_norm": 0.5166342258453369, "learning_rate": 4.6061385866595944e-05, "loss": 0.0274, "step": 109770 }, { "grad_norm": 0.5665566921234131, "learning_rate": 4.60531442610491e-05, "loss": 0.0386, "step": 109780 }, { "grad_norm": 1.4094305038452148, "learning_rate": 4.6044902763407806e-05, "loss": 0.031, "step": 109790 }, { "grad_norm": 0.27954477071762085, "learning_rate": 4.603666137389738e-05, "loss": 0.0295, "step": 109800 }, { "grad_norm": 0.4055500328540802, "learning_rate": 4.602842009274315e-05, "loss": 0.0271, "step": 109810 }, { "grad_norm": 0.5229887366294861, "learning_rate": 4.602017892017042e-05, "loss": 0.0222, "step": 109820 }, { "grad_norm": 0.61249840259552, "learning_rate": 4.6011937856404506e-05, "loss": 0.0309, "step": 109830 }, { "grad_norm": 1.0352801084518433, "learning_rate": 4.6003696901670696e-05, "loss": 0.0433, "step": 109840 }, { "grad_norm": 0.3690192699432373, "learning_rate": 4.599545605619432e-05, "loss": 0.0321, "step": 109850 }, { "grad_norm": 0.5126132965087891, "learning_rate": 4.598721532020068e-05, "loss": 0.0313, "step": 109860 }, { "grad_norm": 0.6945070028305054, "learning_rate": 4.597897469391506e-05, "loss": 0.0301, "step": 109870 }, { "grad_norm": 0.5067228078842163, "learning_rate": 4.5970734177562775e-05, "loss": 0.0277, "step": 109880 }, { "grad_norm": 0.3718453645706177, "learning_rate": 4.5962493771369086e-05, "loss": 0.0345, "step": 109890 }, { "grad_norm": 0.6519992351531982, "learning_rate": 4.5954253475559316e-05, "loss": 0.0327, "step": 109900 }, { "grad_norm": 1.0103914737701416, "learning_rate": 4.594601329035872e-05, "loss": 0.04, "step": 109910 }, { "grad_norm": 0.3801601231098175, "learning_rate": 4.5937773215992594e-05, "loss": 0.022, "step": 109920 }, { "grad_norm": 0.5963356494903564, "learning_rate": 4.5929533252686244e-05, "loss": 0.0481, "step": 109930 }, { "grad_norm": 0.5822266936302185, "learning_rate": 4.59212934006649e-05, "loss": 0.0275, "step": 109940 }, { "grad_norm": 0.9652071595191956, "learning_rate": 4.59130536601539e-05, "loss": 0.0284, "step": 109950 }, { "grad_norm": 0.4749806225299835, "learning_rate": 4.590481403137844e-05, "loss": 0.0227, "step": 109960 }, { "grad_norm": 0.7297407388687134, "learning_rate": 4.589657451456384e-05, "loss": 0.0315, "step": 109970 }, { "grad_norm": 1.458465576171875, "learning_rate": 4.588833510993535e-05, "loss": 0.0308, "step": 109980 }, { "grad_norm": 1.0053277015686035, "learning_rate": 4.588009581771823e-05, "loss": 0.0336, "step": 109990 }, { "grad_norm": 2.4237284660339355, "learning_rate": 4.587185663813775e-05, "loss": 0.0293, "step": 110000 }, { "grad_norm": 0.41509076952934265, "learning_rate": 4.586361757141915e-05, "loss": 0.0237, "step": 110010 }, { "grad_norm": 0.41595184803009033, "learning_rate": 4.585537861778771e-05, "loss": 0.0275, "step": 110020 }, { "grad_norm": 0.7740074396133423, "learning_rate": 4.584713977746864e-05, "loss": 0.0221, "step": 110030 }, { "grad_norm": 0.42181432247161865, "learning_rate": 4.5838901050687206e-05, "loss": 0.0269, "step": 110040 }, { "grad_norm": 0.5213019847869873, "learning_rate": 4.583066243766867e-05, "loss": 0.0205, "step": 110050 }, { "grad_norm": 1.0110963582992554, "learning_rate": 4.5822423938638246e-05, "loss": 0.0285, "step": 110060 }, { "grad_norm": 0.6401641964912415, "learning_rate": 4.58141855538212e-05, "loss": 0.0348, "step": 110070 }, { "grad_norm": 0.4505738317966461, "learning_rate": 4.5805947283442726e-05, "loss": 0.0267, "step": 110080 }, { "grad_norm": 0.4710044264793396, "learning_rate": 4.57977091277281e-05, "loss": 0.0298, "step": 110090 }, { "grad_norm": 0.40360137820243835, "learning_rate": 4.578947108690251e-05, "loss": 0.0252, "step": 110100 }, { "grad_norm": 0.43339622020721436, "learning_rate": 4.57812331611912e-05, "loss": 0.0268, "step": 110110 }, { "grad_norm": 0.3861437141895294, "learning_rate": 4.57729953508194e-05, "loss": 0.0211, "step": 110120 }, { "grad_norm": 0.5316158533096313, "learning_rate": 4.57647576560123e-05, "loss": 0.0385, "step": 110130 }, { "grad_norm": 0.559501588344574, "learning_rate": 4.575652007699517e-05, "loss": 0.0227, "step": 110140 }, { "grad_norm": 0.5542821884155273, "learning_rate": 4.574828261399315e-05, "loss": 0.0206, "step": 110150 }, { "grad_norm": 0.7075895071029663, "learning_rate": 4.574004526723152e-05, "loss": 0.0337, "step": 110160 }, { "grad_norm": 0.4417893588542938, "learning_rate": 4.573180803693542e-05, "loss": 0.023, "step": 110170 }, { "grad_norm": 0.44346335530281067, "learning_rate": 4.572357092333009e-05, "loss": 0.0348, "step": 110180 }, { "grad_norm": 0.6061563491821289, "learning_rate": 4.571533392664074e-05, "loss": 0.0297, "step": 110190 }, { "grad_norm": 0.7435261607170105, "learning_rate": 4.570709704709254e-05, "loss": 0.0265, "step": 110200 }, { "grad_norm": 1.2589579820632935, "learning_rate": 4.569886028491071e-05, "loss": 0.0411, "step": 110210 }, { "grad_norm": 0.6318553686141968, "learning_rate": 4.56906236403204e-05, "loss": 0.0226, "step": 110220 }, { "grad_norm": 0.6901724338531494, "learning_rate": 4.568238711354681e-05, "loss": 0.0253, "step": 110230 }, { "grad_norm": 0.6373656392097473, "learning_rate": 4.5674150704815174e-05, "loss": 0.0241, "step": 110240 }, { "grad_norm": 1.0486738681793213, "learning_rate": 4.5665914414350606e-05, "loss": 0.0318, "step": 110250 }, { "grad_norm": 0.7017996311187744, "learning_rate": 4.5657678242378323e-05, "loss": 0.0292, "step": 110260 }, { "grad_norm": 0.4722627103328705, "learning_rate": 4.564944218912348e-05, "loss": 0.0393, "step": 110270 }, { "grad_norm": 0.5934538841247559, "learning_rate": 4.564120625481126e-05, "loss": 0.0282, "step": 110280 }, { "grad_norm": 0.3411732017993927, "learning_rate": 4.563297043966681e-05, "loss": 0.0282, "step": 110290 }, { "grad_norm": 0.9254568219184875, "learning_rate": 4.5624734743915316e-05, "loss": 0.0297, "step": 110300 }, { "grad_norm": 0.7954387068748474, "learning_rate": 4.5616499167781936e-05, "loss": 0.0239, "step": 110310 }, { "grad_norm": 0.49055901169776917, "learning_rate": 4.560826371149181e-05, "loss": 0.0299, "step": 110320 }, { "grad_norm": 0.6198493838310242, "learning_rate": 4.5600028375270114e-05, "loss": 0.0241, "step": 110330 }, { "grad_norm": 0.6716625094413757, "learning_rate": 4.5591793159341975e-05, "loss": 0.0265, "step": 110340 }, { "grad_norm": 0.4038658142089844, "learning_rate": 4.558355806393258e-05, "loss": 0.0291, "step": 110350 }, { "grad_norm": 0.519307017326355, "learning_rate": 4.557532308926702e-05, "loss": 0.0332, "step": 110360 }, { "grad_norm": 0.2529359757900238, "learning_rate": 4.5567088235570474e-05, "loss": 0.0249, "step": 110370 }, { "grad_norm": 0.7392414808273315, "learning_rate": 4.555885350306808e-05, "loss": 0.0294, "step": 110380 }, { "grad_norm": 0.4071483910083771, "learning_rate": 4.555061889198495e-05, "loss": 0.0219, "step": 110390 }, { "grad_norm": 0.5612903833389282, "learning_rate": 4.5542384402546244e-05, "loss": 0.0499, "step": 110400 }, { "grad_norm": 0.5195646286010742, "learning_rate": 4.553415003497706e-05, "loss": 0.0323, "step": 110410 }, { "grad_norm": 0.7046236991882324, "learning_rate": 4.552591578950253e-05, "loss": 0.0284, "step": 110420 }, { "grad_norm": 0.7032086849212646, "learning_rate": 4.55176816663478e-05, "loss": 0.0387, "step": 110430 }, { "grad_norm": 0.45472827553749084, "learning_rate": 4.5509447665737956e-05, "loss": 0.0297, "step": 110440 }, { "grad_norm": 0.6015634536743164, "learning_rate": 4.550121378789815e-05, "loss": 0.0288, "step": 110450 }, { "grad_norm": 0.6251515746116638, "learning_rate": 4.5492980033053426e-05, "loss": 0.0537, "step": 110460 }, { "grad_norm": 0.3924119770526886, "learning_rate": 4.5484746401428966e-05, "loss": 0.0235, "step": 110470 }, { "grad_norm": 0.35433363914489746, "learning_rate": 4.547651289324983e-05, "loss": 0.0375, "step": 110480 }, { "grad_norm": 0.428821861743927, "learning_rate": 4.546827950874114e-05, "loss": 0.0257, "step": 110490 }, { "grad_norm": 0.512167751789093, "learning_rate": 4.546004624812799e-05, "loss": 0.0208, "step": 110500 }, { "grad_norm": 0.6047458648681641, "learning_rate": 4.545181311163546e-05, "loss": 0.022, "step": 110510 }, { "grad_norm": 0.5568103194236755, "learning_rate": 4.5443580099488667e-05, "loss": 0.0272, "step": 110520 }, { "grad_norm": 0.48206478357315063, "learning_rate": 4.543534721191266e-05, "loss": 0.032, "step": 110530 }, { "grad_norm": 0.6354101300239563, "learning_rate": 4.542711444913256e-05, "loss": 0.0265, "step": 110540 }, { "grad_norm": 0.6523491144180298, "learning_rate": 4.541888181137342e-05, "loss": 0.0312, "step": 110550 }, { "grad_norm": 0.45961451530456543, "learning_rate": 4.541064929886033e-05, "loss": 0.0253, "step": 110560 }, { "grad_norm": 0.4292573630809784, "learning_rate": 4.540241691181838e-05, "loss": 0.0502, "step": 110570 }, { "grad_norm": 0.4186875522136688, "learning_rate": 4.539418465047261e-05, "loss": 0.036, "step": 110580 }, { "grad_norm": 0.3319636881351471, "learning_rate": 4.538595251504812e-05, "loss": 0.0269, "step": 110590 }, { "grad_norm": 2.196664810180664, "learning_rate": 4.537772050576994e-05, "loss": 0.0313, "step": 110600 }, { "grad_norm": 0.4590787887573242, "learning_rate": 4.536948862286316e-05, "loss": 0.0281, "step": 110610 }, { "grad_norm": 0.3413632810115814, "learning_rate": 4.5361256866552796e-05, "loss": 0.0308, "step": 110620 }, { "grad_norm": 0.614103376865387, "learning_rate": 4.535302523706393e-05, "loss": 0.0247, "step": 110630 }, { "grad_norm": 0.6922479867935181, "learning_rate": 4.5344793734621634e-05, "loss": 0.0382, "step": 110640 }, { "grad_norm": 0.5897587537765503, "learning_rate": 4.53365623594509e-05, "loss": 0.0264, "step": 110650 }, { "grad_norm": 0.4156968593597412, "learning_rate": 4.532833111177683e-05, "loss": 0.0289, "step": 110660 }, { "grad_norm": 0.32499685883522034, "learning_rate": 4.53200999918244e-05, "loss": 0.0309, "step": 110670 }, { "grad_norm": 0.36742621660232544, "learning_rate": 4.5311868999818685e-05, "loss": 0.0284, "step": 110680 }, { "grad_norm": 0.46136221289634705, "learning_rate": 4.5303638135984724e-05, "loss": 0.0276, "step": 110690 }, { "grad_norm": 0.8681749701499939, "learning_rate": 4.529540740054752e-05, "loss": 0.0203, "step": 110700 }, { "grad_norm": 0.24020788073539734, "learning_rate": 4.528717679373212e-05, "loss": 0.0266, "step": 110710 }, { "grad_norm": 0.43695151805877686, "learning_rate": 4.527894631576353e-05, "loss": 0.0272, "step": 110720 }, { "grad_norm": 0.4444250762462616, "learning_rate": 4.5270715966866786e-05, "loss": 0.0262, "step": 110730 }, { "grad_norm": 0.6389496922492981, "learning_rate": 4.5262485747266874e-05, "loss": 0.0191, "step": 110740 }, { "grad_norm": 0.42719414830207825, "learning_rate": 4.5254255657188816e-05, "loss": 0.0362, "step": 110750 }, { "grad_norm": 0.2536395490169525, "learning_rate": 4.524602569685764e-05, "loss": 0.0233, "step": 110760 }, { "grad_norm": 0.7971138954162598, "learning_rate": 4.5237795866498336e-05, "loss": 0.0299, "step": 110770 }, { "grad_norm": 0.3681476414203644, "learning_rate": 4.5229566166335904e-05, "loss": 0.0268, "step": 110780 }, { "grad_norm": 0.39873242378234863, "learning_rate": 4.522133659659533e-05, "loss": 0.0195, "step": 110790 }, { "grad_norm": 0.40463685989379883, "learning_rate": 4.521310715750164e-05, "loss": 0.0356, "step": 110800 }, { "grad_norm": 0.272938996553421, "learning_rate": 4.520487784927978e-05, "loss": 0.0426, "step": 110810 }, { "grad_norm": 0.3475976586341858, "learning_rate": 4.519664867215476e-05, "loss": 0.0354, "step": 110820 }, { "grad_norm": 0.4751823842525482, "learning_rate": 4.518841962635158e-05, "loss": 0.0368, "step": 110830 }, { "grad_norm": 0.4731859266757965, "learning_rate": 4.518019071209517e-05, "loss": 0.0387, "step": 110840 }, { "grad_norm": 0.2458990514278412, "learning_rate": 4.517196192961057e-05, "loss": 0.0303, "step": 110850 }, { "grad_norm": 0.3455391526222229, "learning_rate": 4.5163733279122685e-05, "loss": 0.0215, "step": 110860 }, { "grad_norm": 1.0587818622589111, "learning_rate": 4.515550476085653e-05, "loss": 0.0251, "step": 110870 }, { "grad_norm": 0.309773325920105, "learning_rate": 4.514727637503706e-05, "loss": 0.0431, "step": 110880 }, { "grad_norm": 0.3807370960712433, "learning_rate": 4.5139048121889214e-05, "loss": 0.0366, "step": 110890 }, { "grad_norm": 0.6268560290336609, "learning_rate": 4.5130820001637985e-05, "loss": 0.0247, "step": 110900 }, { "grad_norm": 0.3214820623397827, "learning_rate": 4.51225920145083e-05, "loss": 0.0267, "step": 110910 }, { "grad_norm": 0.2216702699661255, "learning_rate": 4.511436416072512e-05, "loss": 0.0235, "step": 110920 }, { "grad_norm": 0.7479110956192017, "learning_rate": 4.5106136440513386e-05, "loss": 0.0323, "step": 110930 }, { "grad_norm": 0.4005277156829834, "learning_rate": 4.5097908854098025e-05, "loss": 0.0296, "step": 110940 }, { "grad_norm": 0.5016008615493774, "learning_rate": 4.508968140170403e-05, "loss": 0.0238, "step": 110950 }, { "grad_norm": 0.3059837222099304, "learning_rate": 4.508145408355627e-05, "loss": 0.0207, "step": 110960 }, { "grad_norm": 0.5506308674812317, "learning_rate": 4.507322689987973e-05, "loss": 0.0324, "step": 110970 }, { "grad_norm": 0.3326881229877472, "learning_rate": 4.50649998508993e-05, "loss": 0.0274, "step": 110980 }, { "grad_norm": 0.3424730598926544, "learning_rate": 4.505677293683993e-05, "loss": 0.033, "step": 110990 }, { "grad_norm": 0.5250004529953003, "learning_rate": 4.504854615792652e-05, "loss": 0.0233, "step": 111000 }, { "grad_norm": 0.34618231654167175, "learning_rate": 4.504031951438401e-05, "loss": 0.0193, "step": 111010 }, { "grad_norm": 0.6141607165336609, "learning_rate": 4.50320930064373e-05, "loss": 0.0404, "step": 111020 }, { "grad_norm": 0.826828122138977, "learning_rate": 4.502386663431129e-05, "loss": 0.0274, "step": 111030 }, { "grad_norm": 0.53383469581604, "learning_rate": 4.501564039823091e-05, "loss": 0.026, "step": 111040 }, { "grad_norm": 0.318651020526886, "learning_rate": 4.500741429842103e-05, "loss": 0.025, "step": 111050 }, { "grad_norm": 0.43712204694747925, "learning_rate": 4.499918833510657e-05, "loss": 0.0249, "step": 111060 }, { "grad_norm": 0.4673464894294739, "learning_rate": 4.499096250851244e-05, "loss": 0.0229, "step": 111070 }, { "grad_norm": 0.826732337474823, "learning_rate": 4.49827368188635e-05, "loss": 0.029, "step": 111080 }, { "grad_norm": 0.7335944175720215, "learning_rate": 4.497451126638467e-05, "loss": 0.0236, "step": 111090 }, { "grad_norm": 0.4042501747608185, "learning_rate": 4.4966285851300796e-05, "loss": 0.0236, "step": 111100 }, { "grad_norm": 0.49834179878234863, "learning_rate": 4.4958060573836785e-05, "loss": 0.0309, "step": 111110 }, { "grad_norm": 0.5546310544013977, "learning_rate": 4.494983543421749e-05, "loss": 0.0249, "step": 111120 }, { "grad_norm": 0.5547651052474976, "learning_rate": 4.494161043266779e-05, "loss": 0.0324, "step": 111130 }, { "grad_norm": 0.45850270986557007, "learning_rate": 4.49333855694126e-05, "loss": 0.0223, "step": 111140 }, { "grad_norm": 0.4431093633174896, "learning_rate": 4.4925160844676704e-05, "loss": 0.0312, "step": 111150 }, { "grad_norm": 0.44700226187705994, "learning_rate": 4.4916936258685036e-05, "loss": 0.0311, "step": 111160 }, { "grad_norm": 0.28289222717285156, "learning_rate": 4.490871181166239e-05, "loss": 0.0243, "step": 111170 }, { "grad_norm": 0.5625526905059814, "learning_rate": 4.4900487503833675e-05, "loss": 0.0337, "step": 111180 }, { "grad_norm": 0.42374545335769653, "learning_rate": 4.48922633354237e-05, "loss": 0.0231, "step": 111190 }, { "grad_norm": 0.5194409489631653, "learning_rate": 4.4884039306657334e-05, "loss": 0.0438, "step": 111200 }, { "grad_norm": 0.5062872767448425, "learning_rate": 4.487581541775942e-05, "loss": 0.0305, "step": 111210 }, { "grad_norm": 0.36972641944885254, "learning_rate": 4.486759166895477e-05, "loss": 0.0176, "step": 111220 }, { "grad_norm": 0.5651742219924927, "learning_rate": 4.485936806046825e-05, "loss": 0.0288, "step": 111230 }, { "grad_norm": 0.4036778509616852, "learning_rate": 4.485114459252467e-05, "loss": 0.0224, "step": 111240 }, { "grad_norm": 0.33251333236694336, "learning_rate": 4.484292126534885e-05, "loss": 0.0282, "step": 111250 }, { "grad_norm": 0.449309766292572, "learning_rate": 4.4834698079165645e-05, "loss": 0.0233, "step": 111260 }, { "grad_norm": 0.7700099945068359, "learning_rate": 4.4826475034199836e-05, "loss": 0.0268, "step": 111270 }, { "grad_norm": 0.27583613991737366, "learning_rate": 4.481825213067627e-05, "loss": 0.0265, "step": 111280 }, { "grad_norm": 0.4551248550415039, "learning_rate": 4.481002936881975e-05, "loss": 0.0268, "step": 111290 }, { "grad_norm": 0.5688585638999939, "learning_rate": 4.480180674885507e-05, "loss": 0.0411, "step": 111300 }, { "grad_norm": 0.4215888977050781, "learning_rate": 4.479358427100703e-05, "loss": 0.0272, "step": 111310 }, { "grad_norm": 0.42813217639923096, "learning_rate": 4.478536193550045e-05, "loss": 0.0261, "step": 111320 }, { "grad_norm": 0.6825065612792969, "learning_rate": 4.477713974256012e-05, "loss": 0.0293, "step": 111330 }, { "grad_norm": 0.37445732951164246, "learning_rate": 4.476891769241081e-05, "loss": 0.0244, "step": 111340 }, { "grad_norm": 0.42497944831848145, "learning_rate": 4.4760695785277353e-05, "loss": 0.0371, "step": 111350 }, { "grad_norm": 0.5290000438690186, "learning_rate": 4.475247402138447e-05, "loss": 0.0335, "step": 111360 }, { "grad_norm": 0.4786798059940338, "learning_rate": 4.474425240095702e-05, "loss": 0.026, "step": 111370 }, { "grad_norm": 0.3309546709060669, "learning_rate": 4.473603092421969e-05, "loss": 0.0292, "step": 111380 }, { "grad_norm": 0.4672406315803528, "learning_rate": 4.4727809591397306e-05, "loss": 0.0219, "step": 111390 }, { "grad_norm": 0.46791207790374756, "learning_rate": 4.4719588402714644e-05, "loss": 0.0266, "step": 111400 }, { "grad_norm": 0.5189849138259888, "learning_rate": 4.471136735839644e-05, "loss": 0.0282, "step": 111410 }, { "grad_norm": 0.5717992782592773, "learning_rate": 4.470314645866748e-05, "loss": 0.0278, "step": 111420 }, { "grad_norm": 0.4667828381061554, "learning_rate": 4.469492570375249e-05, "loss": 0.0197, "step": 111430 }, { "grad_norm": 0.2564706802368164, "learning_rate": 4.4686705093876235e-05, "loss": 0.0219, "step": 111440 }, { "grad_norm": 0.37817299365997314, "learning_rate": 4.467848462926349e-05, "loss": 0.0236, "step": 111450 }, { "grad_norm": 0.43007591366767883, "learning_rate": 4.467026431013895e-05, "loss": 0.021, "step": 111460 }, { "grad_norm": 0.3527888059616089, "learning_rate": 4.46620441367274e-05, "loss": 0.0236, "step": 111470 }, { "grad_norm": 0.3394640386104584, "learning_rate": 4.4653824109253543e-05, "loss": 0.05, "step": 111480 }, { "grad_norm": 0.859093189239502, "learning_rate": 4.4645604227942146e-05, "loss": 0.0321, "step": 111490 }, { "grad_norm": 0.5718004703521729, "learning_rate": 4.46373844930179e-05, "loss": 0.0289, "step": 111500 }, { "grad_norm": 0.45623019337654114, "learning_rate": 4.462916490470555e-05, "loss": 0.0258, "step": 111510 }, { "grad_norm": 1.0917483568191528, "learning_rate": 4.4620945463229825e-05, "loss": 0.0285, "step": 111520 }, { "grad_norm": 0.3635154366493225, "learning_rate": 4.461272616881542e-05, "loss": 0.0348, "step": 111530 }, { "grad_norm": 0.39728647470474243, "learning_rate": 4.460450702168707e-05, "loss": 0.0295, "step": 111540 }, { "grad_norm": 0.45314139127731323, "learning_rate": 4.459628802206944e-05, "loss": 0.0208, "step": 111550 }, { "grad_norm": 0.40382707118988037, "learning_rate": 4.4588069170187306e-05, "loss": 0.025, "step": 111560 }, { "grad_norm": 0.4757373332977295, "learning_rate": 4.45798504662653e-05, "loss": 0.0383, "step": 111570 }, { "grad_norm": 1.0535451173782349, "learning_rate": 4.457163191052815e-05, "loss": 0.0231, "step": 111580 }, { "grad_norm": 0.6927486062049866, "learning_rate": 4.456341350320056e-05, "loss": 0.0291, "step": 111590 }, { "grad_norm": 0.5612444877624512, "learning_rate": 4.455519524450718e-05, "loss": 0.023, "step": 111600 }, { "grad_norm": 0.7318933606147766, "learning_rate": 4.454697713467273e-05, "loss": 0.0214, "step": 111610 }, { "grad_norm": 0.7210356593132019, "learning_rate": 4.453875917392188e-05, "loss": 0.0262, "step": 111620 }, { "grad_norm": 0.3008352220058441, "learning_rate": 4.4530541362479275e-05, "loss": 0.0271, "step": 111630 }, { "grad_norm": 1.1131917238235474, "learning_rate": 4.4522323700569657e-05, "loss": 0.0271, "step": 111640 }, { "grad_norm": 0.5846145749092102, "learning_rate": 4.451410618841761e-05, "loss": 0.0286, "step": 111650 }, { "grad_norm": 0.4399712085723877, "learning_rate": 4.450588882624788e-05, "loss": 0.0279, "step": 111660 }, { "grad_norm": 0.3870159983634949, "learning_rate": 4.449767161428505e-05, "loss": 0.0357, "step": 111670 }, { "grad_norm": 0.32982486486434937, "learning_rate": 4.4489454552753836e-05, "loss": 0.0221, "step": 111680 }, { "grad_norm": 0.45745009183883667, "learning_rate": 4.448123764187885e-05, "loss": 0.0274, "step": 111690 }, { "grad_norm": 0.4195016622543335, "learning_rate": 4.4473020881884755e-05, "loss": 0.0257, "step": 111700 }, { "grad_norm": 0.40383100509643555, "learning_rate": 4.4464804272996204e-05, "loss": 0.0232, "step": 111710 }, { "grad_norm": 0.6686253547668457, "learning_rate": 4.4456587815437814e-05, "loss": 0.0265, "step": 111720 }, { "grad_norm": 0.5092052221298218, "learning_rate": 4.444837150943423e-05, "loss": 0.0282, "step": 111730 }, { "grad_norm": 0.790966808795929, "learning_rate": 4.444015535521009e-05, "loss": 0.0236, "step": 111740 }, { "grad_norm": 0.3242513835430145, "learning_rate": 4.4431939352990014e-05, "loss": 0.0311, "step": 111750 }, { "grad_norm": 1.75358247756958, "learning_rate": 4.4423723502998606e-05, "loss": 0.0302, "step": 111760 }, { "grad_norm": 0.5939341187477112, "learning_rate": 4.441550780546051e-05, "loss": 0.0323, "step": 111770 }, { "grad_norm": 0.9854108095169067, "learning_rate": 4.440729226060035e-05, "loss": 0.0264, "step": 111780 }, { "grad_norm": 0.2627100348472595, "learning_rate": 4.4399076868642697e-05, "loss": 0.0224, "step": 111790 }, { "grad_norm": 0.2918699085712433, "learning_rate": 4.43908616298122e-05, "loss": 0.0286, "step": 111800 }, { "grad_norm": 0.42827367782592773, "learning_rate": 4.438264654433342e-05, "loss": 0.028, "step": 111810 }, { "grad_norm": 0.5466486215591431, "learning_rate": 4.437443161243099e-05, "loss": 0.0283, "step": 111820 }, { "grad_norm": 0.4731854796409607, "learning_rate": 4.436621683432947e-05, "loss": 0.0264, "step": 111830 }, { "grad_norm": 0.5600225329399109, "learning_rate": 4.4358002210253456e-05, "loss": 0.0203, "step": 111840 }, { "grad_norm": 0.5715610980987549, "learning_rate": 4.434978774042758e-05, "loss": 0.0304, "step": 111850 }, { "grad_norm": 0.3867853283882141, "learning_rate": 4.434157342507635e-05, "loss": 0.0251, "step": 111860 }, { "grad_norm": 0.3928832709789276, "learning_rate": 4.433335926442441e-05, "loss": 0.0318, "step": 111870 }, { "grad_norm": 0.2759312391281128, "learning_rate": 4.432514525869626e-05, "loss": 0.0216, "step": 111880 }, { "grad_norm": 0.6317099928855896, "learning_rate": 4.431693140811652e-05, "loss": 0.0282, "step": 111890 }, { "grad_norm": 0.4952632188796997, "learning_rate": 4.430871771290975e-05, "loss": 0.026, "step": 111900 }, { "grad_norm": 0.49662095308303833, "learning_rate": 4.430050417330049e-05, "loss": 0.0296, "step": 111910 }, { "grad_norm": 0.2711983919143677, "learning_rate": 4.429229078951332e-05, "loss": 0.0178, "step": 111920 }, { "grad_norm": 1.038905143737793, "learning_rate": 4.4284077561772765e-05, "loss": 0.0189, "step": 111930 }, { "grad_norm": 0.4359682500362396, "learning_rate": 4.4275864490303384e-05, "loss": 0.018, "step": 111940 }, { "grad_norm": 0.36669254302978516, "learning_rate": 4.426765157532971e-05, "loss": 0.029, "step": 111950 }, { "grad_norm": 0.4321756958961487, "learning_rate": 4.425943881707628e-05, "loss": 0.0286, "step": 111960 }, { "grad_norm": 0.32002541422843933, "learning_rate": 4.425122621576766e-05, "loss": 0.0243, "step": 111970 }, { "grad_norm": 0.7020336389541626, "learning_rate": 4.424301377162834e-05, "loss": 0.0297, "step": 111980 }, { "grad_norm": 0.5817008018493652, "learning_rate": 4.4234801484882874e-05, "loss": 0.0196, "step": 111990 }, { "grad_norm": 0.4806397259235382, "learning_rate": 4.422658935575575e-05, "loss": 0.0164, "step": 112000 }, { "grad_norm": 0.514907956123352, "learning_rate": 4.4218377384471535e-05, "loss": 0.0364, "step": 112010 }, { "grad_norm": 0.4977349042892456, "learning_rate": 4.421016557125468e-05, "loss": 0.0228, "step": 112020 }, { "grad_norm": 0.7205511927604675, "learning_rate": 4.420195391632973e-05, "loss": 0.0352, "step": 112030 }, { "grad_norm": 0.4303639829158783, "learning_rate": 4.4193742419921196e-05, "loss": 0.0257, "step": 112040 }, { "grad_norm": 0.2826414108276367, "learning_rate": 4.418553108225354e-05, "loss": 0.023, "step": 112050 }, { "grad_norm": 1.9811253547668457, "learning_rate": 4.417731990355131e-05, "loss": 0.0386, "step": 112060 }, { "grad_norm": 0.33809444308280945, "learning_rate": 4.4169108884038935e-05, "loss": 0.0199, "step": 112070 }, { "grad_norm": 0.30612125992774963, "learning_rate": 4.416089802394094e-05, "loss": 0.0311, "step": 112080 }, { "grad_norm": 0.3969040811061859, "learning_rate": 4.4152687323481814e-05, "loss": 0.0215, "step": 112090 }, { "grad_norm": 0.8021352887153625, "learning_rate": 4.414447678288601e-05, "loss": 0.0293, "step": 112100 }, { "grad_norm": 0.29229116439819336, "learning_rate": 4.413626640237802e-05, "loss": 0.0312, "step": 112110 }, { "grad_norm": 0.5228915214538574, "learning_rate": 4.412805618218229e-05, "loss": 0.0285, "step": 112120 }, { "grad_norm": 0.3678150773048401, "learning_rate": 4.411984612252331e-05, "loss": 0.0302, "step": 112130 }, { "grad_norm": 0.7417858242988586, "learning_rate": 4.411163622362552e-05, "loss": 0.0323, "step": 112140 }, { "grad_norm": 0.3380153179168701, "learning_rate": 4.4103426485713364e-05, "loss": 0.0289, "step": 112150 }, { "grad_norm": 0.43912777304649353, "learning_rate": 4.409521690901136e-05, "loss": 0.0323, "step": 112160 }, { "grad_norm": 0.681583046913147, "learning_rate": 4.4087007493743866e-05, "loss": 0.0289, "step": 112170 }, { "grad_norm": 0.3653188943862915, "learning_rate": 4.407879824013539e-05, "loss": 0.0354, "step": 112180 }, { "grad_norm": 0.5172504186630249, "learning_rate": 4.407058914841033e-05, "loss": 0.0245, "step": 112190 }, { "grad_norm": 0.3134647607803345, "learning_rate": 4.406238021879315e-05, "loss": 0.0267, "step": 112200 }, { "grad_norm": 0.4670129418373108, "learning_rate": 4.405417145150825e-05, "loss": 0.0319, "step": 112210 }, { "grad_norm": 0.920454740524292, "learning_rate": 4.404596284678008e-05, "loss": 0.0201, "step": 112220 }, { "grad_norm": 0.777661144733429, "learning_rate": 4.403775440483304e-05, "loss": 0.0251, "step": 112230 }, { "grad_norm": 0.39763468503952026, "learning_rate": 4.4029546125891554e-05, "loss": 0.0355, "step": 112240 }, { "grad_norm": 0.8612525463104248, "learning_rate": 4.402133801018005e-05, "loss": 0.0289, "step": 112250 }, { "grad_norm": 0.6018059253692627, "learning_rate": 4.4013130057922894e-05, "loss": 0.0215, "step": 112260 }, { "grad_norm": 0.2772963047027588, "learning_rate": 4.400492226934452e-05, "loss": 0.0217, "step": 112270 }, { "grad_norm": 0.6998903155326843, "learning_rate": 4.399671464466933e-05, "loss": 0.0283, "step": 112280 }, { "grad_norm": 0.36955150961875916, "learning_rate": 4.39885071841217e-05, "loss": 0.018, "step": 112290 }, { "grad_norm": 0.4215644896030426, "learning_rate": 4.398029988792604e-05, "loss": 0.021, "step": 112300 }, { "grad_norm": 0.34716206789016724, "learning_rate": 4.3972092756306696e-05, "loss": 0.027, "step": 112310 }, { "grad_norm": 0.41917896270751953, "learning_rate": 4.396388578948809e-05, "loss": 0.0381, "step": 112320 }, { "grad_norm": 0.2347734421491623, "learning_rate": 4.395567898769457e-05, "loss": 0.0193, "step": 112330 }, { "grad_norm": 0.5967540144920349, "learning_rate": 4.39474723511505e-05, "loss": 0.0348, "step": 112340 }, { "grad_norm": 0.6065030097961426, "learning_rate": 4.39392658800803e-05, "loss": 0.0259, "step": 112350 }, { "grad_norm": 0.4219687581062317, "learning_rate": 4.3931059574708255e-05, "loss": 0.0198, "step": 112360 }, { "grad_norm": 0.6209249496459961, "learning_rate": 4.39228534352588e-05, "loss": 0.0291, "step": 112370 }, { "grad_norm": 0.43336257338523865, "learning_rate": 4.391464746195621e-05, "loss": 0.0279, "step": 112380 }, { "grad_norm": 0.8575271964073181, "learning_rate": 4.39064416550249e-05, "loss": 0.0225, "step": 112390 }, { "grad_norm": 0.6100770235061646, "learning_rate": 4.389823601468917e-05, "loss": 0.0262, "step": 112400 }, { "grad_norm": 0.24304135143756866, "learning_rate": 4.389003054117338e-05, "loss": 0.0223, "step": 112410 }, { "grad_norm": 1.0105787515640259, "learning_rate": 4.3881825234701865e-05, "loss": 0.0344, "step": 112420 }, { "grad_norm": 0.36258408427238464, "learning_rate": 4.387362009549894e-05, "loss": 0.0192, "step": 112430 }, { "grad_norm": 0.5929961800575256, "learning_rate": 4.386541512378895e-05, "loss": 0.0259, "step": 112440 }, { "grad_norm": 0.55535888671875, "learning_rate": 4.3857210319796196e-05, "loss": 0.0216, "step": 112450 }, { "grad_norm": 0.44170820713043213, "learning_rate": 4.3849005683745e-05, "loss": 0.026, "step": 112460 }, { "grad_norm": 0.6905834674835205, "learning_rate": 4.384080121585969e-05, "loss": 0.0249, "step": 112470 }, { "grad_norm": 0.6482568979263306, "learning_rate": 4.383259691636455e-05, "loss": 0.0326, "step": 112480 }, { "grad_norm": 0.25125056505203247, "learning_rate": 4.382439278548391e-05, "loss": 0.0192, "step": 112490 }, { "grad_norm": 0.5934547781944275, "learning_rate": 4.381618882344204e-05, "loss": 0.0244, "step": 112500 }, { "grad_norm": 0.4856257140636444, "learning_rate": 4.380798503046325e-05, "loss": 0.0304, "step": 112510 }, { "grad_norm": 0.7117539644241333, "learning_rate": 4.379978140677182e-05, "loss": 0.038, "step": 112520 }, { "grad_norm": 0.449780136346817, "learning_rate": 4.3791577952592036e-05, "loss": 0.0374, "step": 112530 }, { "grad_norm": 0.5149775147438049, "learning_rate": 4.378337466814818e-05, "loss": 0.0257, "step": 112540 }, { "grad_norm": 0.6860264539718628, "learning_rate": 4.377517155366451e-05, "loss": 0.0316, "step": 112550 }, { "grad_norm": 0.5016123652458191, "learning_rate": 4.3766968609365343e-05, "loss": 0.0298, "step": 112560 }, { "grad_norm": 0.4123367369174957, "learning_rate": 4.375876583547488e-05, "loss": 0.0379, "step": 112570 }, { "grad_norm": 0.5132781863212585, "learning_rate": 4.375056323221745e-05, "loss": 0.0209, "step": 112580 }, { "grad_norm": 0.777776300907135, "learning_rate": 4.374236079981724e-05, "loss": 0.0335, "step": 112590 }, { "grad_norm": 0.8752300143241882, "learning_rate": 4.373415853849854e-05, "loss": 0.0216, "step": 112600 }, { "grad_norm": 0.505150318145752, "learning_rate": 4.37259564484856e-05, "loss": 0.0338, "step": 112610 }, { "grad_norm": 0.8007685542106628, "learning_rate": 4.3717754530002655e-05, "loss": 0.0193, "step": 112620 }, { "grad_norm": 0.3835029900074005, "learning_rate": 4.3709552783273946e-05, "loss": 0.0222, "step": 112630 }, { "grad_norm": 0.49106308817863464, "learning_rate": 4.370135120852369e-05, "loss": 0.0247, "step": 112640 }, { "grad_norm": 0.4467894732952118, "learning_rate": 4.369314980597611e-05, "loss": 0.0242, "step": 112650 }, { "grad_norm": 0.4058634638786316, "learning_rate": 4.368494857585549e-05, "loss": 0.0254, "step": 112660 }, { "grad_norm": 0.340864360332489, "learning_rate": 4.3676747518385956e-05, "loss": 0.0278, "step": 112670 }, { "grad_norm": 0.5958224534988403, "learning_rate": 4.36685466337918e-05, "loss": 0.026, "step": 112680 }, { "grad_norm": 1.1681138277053833, "learning_rate": 4.366034592229719e-05, "loss": 0.0339, "step": 112690 }, { "grad_norm": 0.36772188544273376, "learning_rate": 4.365214538412635e-05, "loss": 0.0333, "step": 112700 }, { "grad_norm": 0.523041844367981, "learning_rate": 4.3643945019503456e-05, "loss": 0.0234, "step": 112710 }, { "grad_norm": 0.2875042259693146, "learning_rate": 4.363574482865272e-05, "loss": 0.0303, "step": 112720 }, { "grad_norm": 0.44526928663253784, "learning_rate": 4.362754481179834e-05, "loss": 0.022, "step": 112730 }, { "grad_norm": 1.8995029926300049, "learning_rate": 4.361934496916448e-05, "loss": 0.023, "step": 112740 }, { "grad_norm": 0.5013496279716492, "learning_rate": 4.361114530097534e-05, "loss": 0.0305, "step": 112750 }, { "grad_norm": 0.524562656879425, "learning_rate": 4.360294580745507e-05, "loss": 0.0252, "step": 112760 }, { "grad_norm": 1.239210844039917, "learning_rate": 4.3594746488827895e-05, "loss": 0.0322, "step": 112770 }, { "grad_norm": 0.5939286947250366, "learning_rate": 4.35865473453179e-05, "loss": 0.0304, "step": 112780 }, { "grad_norm": 0.8107110857963562, "learning_rate": 4.357834837714931e-05, "loss": 0.0351, "step": 112790 }, { "grad_norm": 0.7254095673561096, "learning_rate": 4.357014958454627e-05, "loss": 0.042, "step": 112800 }, { "grad_norm": 0.7458251118659973, "learning_rate": 4.356195096773292e-05, "loss": 0.0292, "step": 112810 }, { "grad_norm": 0.4487350881099701, "learning_rate": 4.3553752526933415e-05, "loss": 0.0289, "step": 112820 }, { "grad_norm": 0.49291765689849854, "learning_rate": 4.354555426237189e-05, "loss": 0.0261, "step": 112830 }, { "grad_norm": 0.5305177569389343, "learning_rate": 4.353735617427247e-05, "loss": 0.0239, "step": 112840 }, { "grad_norm": 0.36372560262680054, "learning_rate": 4.3529158262859336e-05, "loss": 0.0197, "step": 112850 }, { "grad_norm": 0.3462976813316345, "learning_rate": 4.3520960528356557e-05, "loss": 0.0212, "step": 112860 }, { "grad_norm": 0.8381602168083191, "learning_rate": 4.3512762970988315e-05, "loss": 0.0247, "step": 112870 }, { "grad_norm": 0.35251641273498535, "learning_rate": 4.350456559097866e-05, "loss": 0.0336, "step": 112880 }, { "grad_norm": 0.5108811855316162, "learning_rate": 4.3496368388551765e-05, "loss": 0.0197, "step": 112890 }, { "grad_norm": 0.5461430549621582, "learning_rate": 4.3488171363931706e-05, "loss": 0.0285, "step": 112900 }, { "grad_norm": 0.19623415172100067, "learning_rate": 4.347997451734259e-05, "loss": 0.0269, "step": 112910 }, { "grad_norm": 0.9916272163391113, "learning_rate": 4.347177784900854e-05, "loss": 0.0442, "step": 112920 }, { "grad_norm": 0.6388118267059326, "learning_rate": 4.3463581359153614e-05, "loss": 0.0272, "step": 112930 }, { "grad_norm": 0.5712391138076782, "learning_rate": 4.3455385048001936e-05, "loss": 0.0306, "step": 112940 }, { "grad_norm": 0.3954235911369324, "learning_rate": 4.344718891577756e-05, "loss": 0.0299, "step": 112950 }, { "grad_norm": 0.47852832078933716, "learning_rate": 4.3438992962704587e-05, "loss": 0.0216, "step": 112960 }, { "grad_norm": 0.757563591003418, "learning_rate": 4.343079718900705e-05, "loss": 0.0256, "step": 112970 }, { "grad_norm": 0.6507226824760437, "learning_rate": 4.342260159490907e-05, "loss": 0.0295, "step": 112980 }, { "grad_norm": 0.4558170437812805, "learning_rate": 4.34144061806347e-05, "loss": 0.0412, "step": 112990 }, { "grad_norm": 0.500918447971344, "learning_rate": 4.3406210946407976e-05, "loss": 0.0477, "step": 113000 }, { "grad_norm": 0.2552717924118042, "learning_rate": 4.339801589245299e-05, "loss": 0.0333, "step": 113010 }, { "grad_norm": 0.3225332796573639, "learning_rate": 4.3389821018993745e-05, "loss": 0.0216, "step": 113020 }, { "grad_norm": 0.5871056318283081, "learning_rate": 4.338162632625433e-05, "loss": 0.025, "step": 113030 }, { "grad_norm": 0.43291643261909485, "learning_rate": 4.3373431814458756e-05, "loss": 0.0282, "step": 113040 }, { "grad_norm": 0.6195876002311707, "learning_rate": 4.3365237483831044e-05, "loss": 0.0327, "step": 113050 }, { "grad_norm": 0.4539920389652252, "learning_rate": 4.3357043334595296e-05, "loss": 0.0261, "step": 113060 }, { "grad_norm": 0.6613990068435669, "learning_rate": 4.334884936697544e-05, "loss": 0.0204, "step": 113070 }, { "grad_norm": 0.6326974630355835, "learning_rate": 4.334065558119558e-05, "loss": 0.0588, "step": 113080 }, { "grad_norm": 0.38581570982933044, "learning_rate": 4.3332461977479666e-05, "loss": 0.0328, "step": 113090 }, { "grad_norm": 0.7154549956321716, "learning_rate": 4.3324268556051746e-05, "loss": 0.0361, "step": 113100 }, { "grad_norm": 0.39530110359191895, "learning_rate": 4.3316075317135825e-05, "loss": 0.0279, "step": 113110 }, { "grad_norm": 0.40110892057418823, "learning_rate": 4.330788226095588e-05, "loss": 0.027, "step": 113120 }, { "grad_norm": 0.5654674172401428, "learning_rate": 4.3299689387735936e-05, "loss": 0.0227, "step": 113130 }, { "grad_norm": 0.5664215683937073, "learning_rate": 4.3291496697699954e-05, "loss": 0.0275, "step": 113140 }, { "grad_norm": 0.4812769889831543, "learning_rate": 4.328330419107194e-05, "loss": 0.0282, "step": 113150 }, { "grad_norm": 0.5059934854507446, "learning_rate": 4.327511186807586e-05, "loss": 0.0272, "step": 113160 }, { "grad_norm": 0.3276047110557556, "learning_rate": 4.3266919728935686e-05, "loss": 0.0232, "step": 113170 }, { "grad_norm": 0.2648254334926605, "learning_rate": 4.325872777387541e-05, "loss": 0.0234, "step": 113180 }, { "grad_norm": 0.6676785945892334, "learning_rate": 4.325053600311898e-05, "loss": 0.0283, "step": 113190 }, { "grad_norm": 0.3712911605834961, "learning_rate": 4.324234441689038e-05, "loss": 0.024, "step": 113200 }, { "grad_norm": 0.3836847245693207, "learning_rate": 4.323415301541351e-05, "loss": 0.0457, "step": 113210 }, { "grad_norm": 0.410403311252594, "learning_rate": 4.3225961798912385e-05, "loss": 0.0316, "step": 113220 }, { "grad_norm": 0.23873573541641235, "learning_rate": 4.3217770767610906e-05, "loss": 0.0243, "step": 113230 }, { "grad_norm": 0.29954254627227783, "learning_rate": 4.320957992173302e-05, "loss": 0.0237, "step": 113240 }, { "grad_norm": 0.5045401453971863, "learning_rate": 4.3201389261502685e-05, "loss": 0.0226, "step": 113250 }, { "grad_norm": 0.3866026699542999, "learning_rate": 4.319319878714379e-05, "loss": 0.0222, "step": 113260 }, { "grad_norm": 0.6133397817611694, "learning_rate": 4.318500849888031e-05, "loss": 0.0253, "step": 113270 }, { "grad_norm": 0.3339456021785736, "learning_rate": 4.317681839693611e-05, "loss": 0.0318, "step": 113280 }, { "grad_norm": 0.2411494106054306, "learning_rate": 4.316862848153514e-05, "loss": 0.0249, "step": 113290 }, { "grad_norm": 0.3329440951347351, "learning_rate": 4.3160438752901313e-05, "loss": 0.0229, "step": 113300 }, { "grad_norm": 0.33248767256736755, "learning_rate": 4.315224921125851e-05, "loss": 0.0261, "step": 113310 }, { "grad_norm": 0.42341890931129456, "learning_rate": 4.314405985683065e-05, "loss": 0.0378, "step": 113320 }, { "grad_norm": 0.48942601680755615, "learning_rate": 4.313587068984161e-05, "loss": 0.0182, "step": 113330 }, { "grad_norm": 1.4127813577651978, "learning_rate": 4.31276817105153e-05, "loss": 0.0366, "step": 113340 }, { "grad_norm": 0.41610461473464966, "learning_rate": 4.311949291907558e-05, "loss": 0.0274, "step": 113350 }, { "grad_norm": 0.5046583414077759, "learning_rate": 4.311130431574632e-05, "loss": 0.0265, "step": 113360 }, { "grad_norm": 0.4143485426902771, "learning_rate": 4.310311590075145e-05, "loss": 0.033, "step": 113370 }, { "grad_norm": 0.4727434515953064, "learning_rate": 4.3094927674314764e-05, "loss": 0.0322, "step": 113380 }, { "grad_norm": 0.6032549142837524, "learning_rate": 4.308673963666019e-05, "loss": 0.0331, "step": 113390 }, { "grad_norm": 0.3919380009174347, "learning_rate": 4.307855178801154e-05, "loss": 0.0399, "step": 113400 }, { "grad_norm": 0.39401674270629883, "learning_rate": 4.3070364128592695e-05, "loss": 0.0291, "step": 113410 }, { "grad_norm": 0.460123211145401, "learning_rate": 4.306217665862749e-05, "loss": 0.0195, "step": 113420 }, { "grad_norm": 0.44568803906440735, "learning_rate": 4.3053989378339756e-05, "loss": 0.0351, "step": 113430 }, { "grad_norm": 0.6067590117454529, "learning_rate": 4.304580228795335e-05, "loss": 0.0261, "step": 113440 }, { "grad_norm": 0.3996342420578003, "learning_rate": 4.30376153876921e-05, "loss": 0.0311, "step": 113450 }, { "grad_norm": 0.618588924407959, "learning_rate": 4.302942867777983e-05, "loss": 0.0297, "step": 113460 }, { "grad_norm": 0.4867796301841736, "learning_rate": 4.302124215844035e-05, "loss": 0.0277, "step": 113470 }, { "grad_norm": 0.5280079245567322, "learning_rate": 4.301305582989749e-05, "loss": 0.0248, "step": 113480 }, { "grad_norm": 0.2617725133895874, "learning_rate": 4.300486969237507e-05, "loss": 0.0496, "step": 113490 }, { "grad_norm": 0.6286978125572205, "learning_rate": 4.299668374609688e-05, "loss": 0.0262, "step": 113500 }, { "grad_norm": 0.3503718972206116, "learning_rate": 4.298849799128673e-05, "loss": 0.0205, "step": 113510 }, { "grad_norm": 0.4578225016593933, "learning_rate": 4.2980312428168404e-05, "loss": 0.0207, "step": 113520 }, { "grad_norm": 0.2038208544254303, "learning_rate": 4.2972127056965705e-05, "loss": 0.028, "step": 113530 }, { "grad_norm": 0.4735337197780609, "learning_rate": 4.296394187790241e-05, "loss": 0.0256, "step": 113540 }, { "grad_norm": 0.4026277959346771, "learning_rate": 4.295575689120228e-05, "loss": 0.0278, "step": 113550 }, { "grad_norm": 0.4360736608505249, "learning_rate": 4.294757209708915e-05, "loss": 0.0235, "step": 113560 }, { "grad_norm": 0.4204466938972473, "learning_rate": 4.293938749578671e-05, "loss": 0.026, "step": 113570 }, { "grad_norm": 0.437103271484375, "learning_rate": 4.29312030875188e-05, "loss": 0.0286, "step": 113580 }, { "grad_norm": 0.5192056894302368, "learning_rate": 4.2923018872509115e-05, "loss": 0.0233, "step": 113590 }, { "grad_norm": 0.310135155916214, "learning_rate": 4.2914834850981446e-05, "loss": 0.0323, "step": 113600 }, { "grad_norm": 0.44803446531295776, "learning_rate": 4.290665102315953e-05, "loss": 0.0255, "step": 113610 }, { "grad_norm": 1.0886056423187256, "learning_rate": 4.28984673892671e-05, "loss": 0.0318, "step": 113620 }, { "grad_norm": 0.4980332851409912, "learning_rate": 4.2890283949527924e-05, "loss": 0.0227, "step": 113630 }, { "grad_norm": 0.38730329275131226, "learning_rate": 4.2882100704165696e-05, "loss": 0.0218, "step": 113640 }, { "grad_norm": 0.5039039254188538, "learning_rate": 4.287391765340417e-05, "loss": 0.0198, "step": 113650 }, { "grad_norm": 0.31798288226127625, "learning_rate": 4.286573479746706e-05, "loss": 0.0233, "step": 113660 }, { "grad_norm": 0.579687774181366, "learning_rate": 4.2857552136578055e-05, "loss": 0.0231, "step": 113670 }, { "grad_norm": 0.2950417995452881, "learning_rate": 4.284936967096091e-05, "loss": 0.0293, "step": 113680 }, { "grad_norm": 2.179705858230591, "learning_rate": 4.284118740083932e-05, "loss": 0.0247, "step": 113690 }, { "grad_norm": 0.5511249303817749, "learning_rate": 4.283300532643698e-05, "loss": 0.0231, "step": 113700 }, { "grad_norm": 0.9418268203735352, "learning_rate": 4.2824823447977565e-05, "loss": 0.041, "step": 113710 }, { "grad_norm": 0.46012160181999207, "learning_rate": 4.2816641765684796e-05, "loss": 0.0259, "step": 113720 }, { "grad_norm": 0.7252095937728882, "learning_rate": 4.2808460279782334e-05, "loss": 0.0288, "step": 113730 }, { "grad_norm": 0.758285403251648, "learning_rate": 4.280027899049386e-05, "loss": 0.0294, "step": 113740 }, { "grad_norm": 0.5937193632125854, "learning_rate": 4.279209789804307e-05, "loss": 0.0212, "step": 113750 }, { "grad_norm": 0.5723949670791626, "learning_rate": 4.278391700265359e-05, "loss": 0.0197, "step": 113760 }, { "grad_norm": 0.4802634119987488, "learning_rate": 4.277573630454914e-05, "loss": 0.0248, "step": 113770 }, { "grad_norm": 0.3919582962989807, "learning_rate": 4.2767555803953316e-05, "loss": 0.0294, "step": 113780 }, { "grad_norm": 0.3714566230773926, "learning_rate": 4.275937550108984e-05, "loss": 0.0205, "step": 113790 }, { "grad_norm": 0.572782576084137, "learning_rate": 4.275119539618227e-05, "loss": 0.0249, "step": 113800 }, { "grad_norm": 0.576651394367218, "learning_rate": 4.274301548945431e-05, "loss": 0.0361, "step": 113810 }, { "grad_norm": 0.2531784176826477, "learning_rate": 4.2734835781129595e-05, "loss": 0.027, "step": 113820 }, { "grad_norm": 0.5570675134658813, "learning_rate": 4.272665627143173e-05, "loss": 0.0262, "step": 113830 }, { "grad_norm": 0.46838757395744324, "learning_rate": 4.271847696058436e-05, "loss": 0.0199, "step": 113840 }, { "grad_norm": 0.5914351344108582, "learning_rate": 4.2710297848811075e-05, "loss": 0.0341, "step": 113850 }, { "grad_norm": 0.3563995361328125, "learning_rate": 4.270211893633551e-05, "loss": 0.0207, "step": 113860 }, { "grad_norm": 0.8729016780853271, "learning_rate": 4.2693940223381295e-05, "loss": 0.0218, "step": 113870 }, { "grad_norm": 0.27374792098999023, "learning_rate": 4.2685761710171976e-05, "loss": 0.0229, "step": 113880 }, { "grad_norm": 0.42701393365859985, "learning_rate": 4.267758339693121e-05, "loss": 0.0268, "step": 113890 }, { "grad_norm": 0.3969675898551941, "learning_rate": 4.266940528388256e-05, "loss": 0.0188, "step": 113900 }, { "grad_norm": 0.823479950428009, "learning_rate": 4.266122737124962e-05, "loss": 0.0223, "step": 113910 }, { "grad_norm": 0.5143578052520752, "learning_rate": 4.265304965925595e-05, "loss": 0.0238, "step": 113920 }, { "grad_norm": 1.789044737815857, "learning_rate": 4.264487214812515e-05, "loss": 0.0403, "step": 113930 }, { "grad_norm": 0.5255306363105774, "learning_rate": 4.2636694838080786e-05, "loss": 0.0319, "step": 113940 }, { "grad_norm": 0.8502874970436096, "learning_rate": 4.262851772934641e-05, "loss": 0.0318, "step": 113950 }, { "grad_norm": 0.7019668221473694, "learning_rate": 4.262034082214561e-05, "loss": 0.0252, "step": 113960 }, { "grad_norm": 0.3826539218425751, "learning_rate": 4.261216411670189e-05, "loss": 0.0192, "step": 113970 }, { "grad_norm": 0.3515615165233612, "learning_rate": 4.260398761323886e-05, "loss": 0.0243, "step": 113980 }, { "grad_norm": 0.276947557926178, "learning_rate": 4.2595811311979995e-05, "loss": 0.0289, "step": 113990 }, { "grad_norm": 0.28204867243766785, "learning_rate": 4.258763521314888e-05, "loss": 0.0197, "step": 114000 }, { "grad_norm": 0.464994341135025, "learning_rate": 4.257945931696904e-05, "loss": 0.0364, "step": 114010 }, { "grad_norm": 0.48000138998031616, "learning_rate": 4.257128362366399e-05, "loss": 0.0229, "step": 114020 }, { "grad_norm": 0.5167235136032104, "learning_rate": 4.256310813345727e-05, "loss": 0.0383, "step": 114030 }, { "grad_norm": 0.5038664937019348, "learning_rate": 4.255493284657237e-05, "loss": 0.0217, "step": 114040 }, { "grad_norm": 0.37736159563064575, "learning_rate": 4.2546757763232786e-05, "loss": 0.0205, "step": 114050 }, { "grad_norm": 2.890261173248291, "learning_rate": 4.2538582883662084e-05, "loss": 0.0243, "step": 114060 }, { "grad_norm": 0.36279386281967163, "learning_rate": 4.253040820808369e-05, "loss": 0.0301, "step": 114070 }, { "grad_norm": 0.446725457906723, "learning_rate": 4.2522233736721164e-05, "loss": 0.0261, "step": 114080 }, { "grad_norm": 1.038954496383667, "learning_rate": 4.2514059469797915e-05, "loss": 0.0344, "step": 114090 }, { "grad_norm": 0.5977054834365845, "learning_rate": 4.2505885407537503e-05, "loss": 0.0218, "step": 114100 }, { "grad_norm": 0.4993635416030884, "learning_rate": 4.249771155016335e-05, "loss": 0.0233, "step": 114110 }, { "grad_norm": 0.2780870497226715, "learning_rate": 4.2489537897898944e-05, "loss": 0.0209, "step": 114120 }, { "grad_norm": 1.0474445819854736, "learning_rate": 4.248136445096777e-05, "loss": 0.0224, "step": 114130 }, { "grad_norm": 0.8450568914413452, "learning_rate": 4.247319120959325e-05, "loss": 0.021, "step": 114140 }, { "grad_norm": 0.6926120519638062, "learning_rate": 4.246501817399887e-05, "loss": 0.026, "step": 114150 }, { "grad_norm": 1.516335129737854, "learning_rate": 4.245684534440804e-05, "loss": 0.02, "step": 114160 }, { "grad_norm": 0.80100017786026, "learning_rate": 4.2448672721044246e-05, "loss": 0.0263, "step": 114170 }, { "grad_norm": 0.3239858150482178, "learning_rate": 4.244050030413088e-05, "loss": 0.0265, "step": 114180 }, { "grad_norm": 1.3954596519470215, "learning_rate": 4.243232809389141e-05, "loss": 0.0192, "step": 114190 }, { "grad_norm": 3.985076904296875, "learning_rate": 4.2424156090549244e-05, "loss": 0.0328, "step": 114200 }, { "grad_norm": 0.569042444229126, "learning_rate": 4.2415984294327806e-05, "loss": 0.0352, "step": 114210 }, { "grad_norm": 0.4936358332633972, "learning_rate": 4.240781270545051e-05, "loss": 0.0283, "step": 114220 }, { "grad_norm": 1.4736199378967285, "learning_rate": 4.239964132414076e-05, "loss": 0.0289, "step": 114230 }, { "grad_norm": 0.3947829306125641, "learning_rate": 4.239147015062196e-05, "loss": 0.0273, "step": 114240 }, { "grad_norm": 0.38356971740722656, "learning_rate": 4.2383299185117514e-05, "loss": 0.0209, "step": 114250 }, { "grad_norm": 0.37016889452934265, "learning_rate": 4.2375128427850785e-05, "loss": 0.0231, "step": 114260 }, { "grad_norm": 0.3133607804775238, "learning_rate": 4.2366957879045214e-05, "loss": 0.0205, "step": 114270 }, { "grad_norm": 0.6760798692703247, "learning_rate": 4.235878753892412e-05, "loss": 0.0269, "step": 114280 }, { "grad_norm": 0.4810600280761719, "learning_rate": 4.235061740771093e-05, "loss": 0.0288, "step": 114290 }, { "grad_norm": 0.4371524155139923, "learning_rate": 4.234244748562896e-05, "loss": 0.0211, "step": 114300 }, { "grad_norm": 0.25970378518104553, "learning_rate": 4.2334277772901595e-05, "loss": 0.0246, "step": 114310 }, { "grad_norm": 0.33594876527786255, "learning_rate": 4.232610826975222e-05, "loss": 0.0254, "step": 114320 }, { "grad_norm": 0.3104735016822815, "learning_rate": 4.2317938976404146e-05, "loss": 0.0228, "step": 114330 }, { "grad_norm": 0.6328189969062805, "learning_rate": 4.2309769893080756e-05, "loss": 0.0209, "step": 114340 }, { "grad_norm": 0.23142819106578827, "learning_rate": 4.2301601020005344e-05, "loss": 0.029, "step": 114350 }, { "grad_norm": 0.3146643042564392, "learning_rate": 4.229343235740129e-05, "loss": 0.0226, "step": 114360 }, { "grad_norm": 0.3192763328552246, "learning_rate": 4.228526390549188e-05, "loss": 0.0222, "step": 114370 }, { "grad_norm": 0.599134087562561, "learning_rate": 4.2277095664500446e-05, "loss": 0.0289, "step": 114380 }, { "grad_norm": 0.3164803981781006, "learning_rate": 4.2268927634650344e-05, "loss": 0.0318, "step": 114390 }, { "grad_norm": 0.36375242471694946, "learning_rate": 4.226075981616484e-05, "loss": 0.0193, "step": 114400 }, { "grad_norm": 0.44253796339035034, "learning_rate": 4.225259220926726e-05, "loss": 0.0283, "step": 114410 }, { "grad_norm": 0.5164222121238708, "learning_rate": 4.2244424814180894e-05, "loss": 0.0287, "step": 114420 }, { "grad_norm": 0.3089110851287842, "learning_rate": 4.223625763112905e-05, "loss": 0.0244, "step": 114430 }, { "grad_norm": 0.4116911292076111, "learning_rate": 4.2228090660334996e-05, "loss": 0.0281, "step": 114440 }, { "grad_norm": 0.3865585923194885, "learning_rate": 4.221992390202202e-05, "loss": 0.0223, "step": 114450 }, { "grad_norm": 0.44405919313430786, "learning_rate": 4.221175735641342e-05, "loss": 0.0171, "step": 114460 }, { "grad_norm": 0.8166553974151611, "learning_rate": 4.2203591023732426e-05, "loss": 0.0235, "step": 114470 }, { "grad_norm": 0.4205241799354553, "learning_rate": 4.219542490420235e-05, "loss": 0.0261, "step": 114480 }, { "grad_norm": 1.214360237121582, "learning_rate": 4.21872589980464e-05, "loss": 0.0382, "step": 114490 }, { "grad_norm": 0.5799291729927063, "learning_rate": 4.217909330548785e-05, "loss": 0.0294, "step": 114500 }, { "grad_norm": 0.4590449631214142, "learning_rate": 4.217092782674998e-05, "loss": 0.0426, "step": 114510 }, { "grad_norm": 0.4675571024417877, "learning_rate": 4.216276256205598e-05, "loss": 0.0401, "step": 114520 }, { "grad_norm": 0.3546397089958191, "learning_rate": 4.2154597511629126e-05, "loss": 0.025, "step": 114530 }, { "grad_norm": 1.2868181467056274, "learning_rate": 4.214643267569261e-05, "loss": 0.0308, "step": 114540 }, { "grad_norm": 0.6959080100059509, "learning_rate": 4.213826805446969e-05, "loss": 0.0271, "step": 114550 }, { "grad_norm": 0.6143209338188171, "learning_rate": 4.213010364818355e-05, "loss": 0.0224, "step": 114560 }, { "grad_norm": 0.5726414918899536, "learning_rate": 4.2121939457057414e-05, "loss": 0.0291, "step": 114570 }, { "grad_norm": 0.3701585531234741, "learning_rate": 4.211377548131453e-05, "loss": 0.0328, "step": 114580 }, { "grad_norm": 0.36075863242149353, "learning_rate": 4.210561172117802e-05, "loss": 0.0305, "step": 114590 }, { "grad_norm": 0.3052506744861603, "learning_rate": 4.209744817687115e-05, "loss": 0.0308, "step": 114600 }, { "grad_norm": 0.4219551980495453, "learning_rate": 4.208928484861706e-05, "loss": 0.0284, "step": 114610 }, { "grad_norm": 0.42306986451148987, "learning_rate": 4.208112173663897e-05, "loss": 0.027, "step": 114620 }, { "grad_norm": 0.7398540377616882, "learning_rate": 4.207295884116002e-05, "loss": 0.0262, "step": 114630 }, { "grad_norm": 0.5269853472709656, "learning_rate": 4.2064796162403395e-05, "loss": 0.021, "step": 114640 }, { "grad_norm": 0.5253850817680359, "learning_rate": 4.205663370059228e-05, "loss": 0.0251, "step": 114650 }, { "grad_norm": 0.45179399847984314, "learning_rate": 4.20484714559498e-05, "loss": 0.0185, "step": 114660 }, { "grad_norm": 0.2403290867805481, "learning_rate": 4.204030942869913e-05, "loss": 0.0194, "step": 114670 }, { "grad_norm": 0.6244991421699524, "learning_rate": 4.2032147619063397e-05, "loss": 0.0279, "step": 114680 }, { "grad_norm": 0.5499383807182312, "learning_rate": 4.202398602726576e-05, "loss": 0.0277, "step": 114690 }, { "grad_norm": 0.2861849367618561, "learning_rate": 4.201582465352936e-05, "loss": 0.0226, "step": 114700 }, { "grad_norm": 0.9269684553146362, "learning_rate": 4.200766349807731e-05, "loss": 0.0374, "step": 114710 }, { "grad_norm": 0.7500070929527283, "learning_rate": 4.199950256113274e-05, "loss": 0.0485, "step": 114720 }, { "grad_norm": 0.6954566836357117, "learning_rate": 4.199134184291875e-05, "loss": 0.0331, "step": 114730 }, { "grad_norm": 0.2055927962064743, "learning_rate": 4.198318134365848e-05, "loss": 0.0272, "step": 114740 }, { "grad_norm": 0.39848247170448303, "learning_rate": 4.1975021063575015e-05, "loss": 0.0219, "step": 114750 }, { "grad_norm": 0.41171932220458984, "learning_rate": 4.1966861002891445e-05, "loss": 0.0267, "step": 114760 }, { "grad_norm": 0.36991822719573975, "learning_rate": 4.195870116183091e-05, "loss": 0.0219, "step": 114770 }, { "grad_norm": 0.4505825340747833, "learning_rate": 4.195054154061643e-05, "loss": 0.0283, "step": 114780 }, { "grad_norm": 0.29244545102119446, "learning_rate": 4.1942382139471155e-05, "loss": 0.0318, "step": 114790 }, { "grad_norm": 0.5348671078681946, "learning_rate": 4.193422295861809e-05, "loss": 0.0212, "step": 114800 }, { "grad_norm": 0.5185478925704956, "learning_rate": 4.192606399828036e-05, "loss": 0.0228, "step": 114810 }, { "grad_norm": 0.4091721177101135, "learning_rate": 4.1917905258681003e-05, "loss": 0.0258, "step": 114820 }, { "grad_norm": 0.5796024203300476, "learning_rate": 4.190974674004308e-05, "loss": 0.0267, "step": 114830 }, { "grad_norm": 0.7699453234672546, "learning_rate": 4.190158844258964e-05, "loss": 0.0329, "step": 114840 }, { "grad_norm": 0.33665215969085693, "learning_rate": 4.1893430366543726e-05, "loss": 0.0175, "step": 114850 }, { "grad_norm": 0.3567805886268616, "learning_rate": 4.18852725121284e-05, "loss": 0.0291, "step": 114860 }, { "grad_norm": 0.6422708034515381, "learning_rate": 4.1877114879566654e-05, "loss": 0.0321, "step": 114870 }, { "grad_norm": 0.532666802406311, "learning_rate": 4.1868957469081524e-05, "loss": 0.0389, "step": 114880 }, { "grad_norm": 0.448076456785202, "learning_rate": 4.1860800280896065e-05, "loss": 0.0407, "step": 114890 }, { "grad_norm": 0.4339693486690521, "learning_rate": 4.185264331523326e-05, "loss": 0.0274, "step": 114900 }, { "grad_norm": 0.4849637746810913, "learning_rate": 4.184448657231614e-05, "loss": 0.0187, "step": 114910 }, { "grad_norm": 0.31784266233444214, "learning_rate": 4.1836330052367676e-05, "loss": 0.0372, "step": 114920 }, { "grad_norm": 0.9247984290122986, "learning_rate": 4.18281737556109e-05, "loss": 0.0257, "step": 114930 }, { "grad_norm": 0.41575273871421814, "learning_rate": 4.1820017682268765e-05, "loss": 0.0237, "step": 114940 }, { "grad_norm": 0.6740588545799255, "learning_rate": 4.181186183256426e-05, "loss": 0.0301, "step": 114950 }, { "grad_norm": 0.34242239594459534, "learning_rate": 4.1803706206720414e-05, "loss": 0.0225, "step": 114960 }, { "grad_norm": 0.5208346247673035, "learning_rate": 4.179555080496013e-05, "loss": 0.0294, "step": 114970 }, { "grad_norm": 0.45153218507766724, "learning_rate": 4.178739562750644e-05, "loss": 0.0184, "step": 114980 }, { "grad_norm": 0.49998021125793457, "learning_rate": 4.1779240674582234e-05, "loss": 0.0337, "step": 114990 }, { "grad_norm": 0.4366393983364105, "learning_rate": 4.177108594641053e-05, "loss": 0.0207, "step": 115000 }, { "grad_norm": 0.3239173889160156, "learning_rate": 4.176293144321421e-05, "loss": 0.0309, "step": 115010 }, { "grad_norm": 0.5169153809547424, "learning_rate": 4.175477716521626e-05, "loss": 0.0218, "step": 115020 }, { "grad_norm": 1.432078242301941, "learning_rate": 4.174662311263962e-05, "loss": 0.0354, "step": 115030 }, { "grad_norm": 0.8750325441360474, "learning_rate": 4.17384692857072e-05, "loss": 0.0293, "step": 115040 }, { "grad_norm": 0.48201823234558105, "learning_rate": 4.1730315684641915e-05, "loss": 0.0274, "step": 115050 }, { "grad_norm": 0.6807191967964172, "learning_rate": 4.1722162309666695e-05, "loss": 0.0372, "step": 115060 }, { "grad_norm": 0.4496775269508362, "learning_rate": 4.1714009161004435e-05, "loss": 0.0287, "step": 115070 }, { "grad_norm": 0.6357054710388184, "learning_rate": 4.170585623887809e-05, "loss": 0.043, "step": 115080 }, { "grad_norm": 0.8094896674156189, "learning_rate": 4.169770354351048e-05, "loss": 0.0326, "step": 115090 }, { "grad_norm": 0.15016211569309235, "learning_rate": 4.168955107512456e-05, "loss": 0.0188, "step": 115100 }, { "grad_norm": 0.34737342596054077, "learning_rate": 4.168139883394318e-05, "loss": 0.0215, "step": 115110 }, { "grad_norm": 0.8448832035064697, "learning_rate": 4.1673246820189245e-05, "loss": 0.0222, "step": 115120 }, { "grad_norm": 0.5203756093978882, "learning_rate": 4.16650950340856e-05, "loss": 0.0235, "step": 115130 }, { "grad_norm": 0.409410297870636, "learning_rate": 4.165694347585512e-05, "loss": 0.0314, "step": 115140 }, { "grad_norm": 0.5277624130249023, "learning_rate": 4.164879214572069e-05, "loss": 0.024, "step": 115150 }, { "grad_norm": 0.34101951122283936, "learning_rate": 4.164064104390512e-05, "loss": 0.0197, "step": 115160 }, { "grad_norm": 0.5635504722595215, "learning_rate": 4.163249017063132e-05, "loss": 0.0277, "step": 115170 }, { "grad_norm": 0.5579472184181213, "learning_rate": 4.162433952612206e-05, "loss": 0.0224, "step": 115180 }, { "grad_norm": 0.36667943000793457, "learning_rate": 4.161618911060024e-05, "loss": 0.0314, "step": 115190 }, { "grad_norm": 0.7309268712997437, "learning_rate": 4.160803892428863e-05, "loss": 0.0343, "step": 115200 }, { "grad_norm": 0.5385037064552307, "learning_rate": 4.1599888967410093e-05, "loss": 0.0378, "step": 115210 }, { "grad_norm": 0.4791164696216583, "learning_rate": 4.159173924018745e-05, "loss": 0.0241, "step": 115220 }, { "grad_norm": 0.4017777442932129, "learning_rate": 4.1583589742843486e-05, "loss": 0.0319, "step": 115230 }, { "grad_norm": 0.4836861491203308, "learning_rate": 4.157544047560102e-05, "loss": 0.0322, "step": 115240 }, { "grad_norm": 0.7965713143348694, "learning_rate": 4.156729143868285e-05, "loss": 0.0493, "step": 115250 }, { "grad_norm": 0.2940865159034729, "learning_rate": 4.155914263231174e-05, "loss": 0.0299, "step": 115260 }, { "grad_norm": 0.6076982021331787, "learning_rate": 4.1550994056710536e-05, "loss": 0.0324, "step": 115270 }, { "grad_norm": 0.5261138677597046, "learning_rate": 4.1542845712101955e-05, "loss": 0.0255, "step": 115280 }, { "grad_norm": 0.6308240294456482, "learning_rate": 4.153469759870882e-05, "loss": 0.0403, "step": 115290 }, { "grad_norm": 0.8807550668716431, "learning_rate": 4.1526549716753836e-05, "loss": 0.0217, "step": 115300 }, { "grad_norm": 0.2510758936405182, "learning_rate": 4.151840206645983e-05, "loss": 0.0286, "step": 115310 }, { "grad_norm": 0.39385420083999634, "learning_rate": 4.151025464804951e-05, "loss": 0.0218, "step": 115320 }, { "grad_norm": 0.4549338221549988, "learning_rate": 4.1502107461745635e-05, "loss": 0.0264, "step": 115330 }, { "grad_norm": 0.40935224294662476, "learning_rate": 4.1493960507770966e-05, "loss": 0.0268, "step": 115340 }, { "grad_norm": 0.500917911529541, "learning_rate": 4.1485813786348204e-05, "loss": 0.0222, "step": 115350 }, { "grad_norm": 0.39531591534614563, "learning_rate": 4.147766729770011e-05, "loss": 0.0242, "step": 115360 }, { "grad_norm": 0.3917883038520813, "learning_rate": 4.146952104204937e-05, "loss": 0.0237, "step": 115370 }, { "grad_norm": 0.4828289747238159, "learning_rate": 4.1461375019618754e-05, "loss": 0.0265, "step": 115380 }, { "grad_norm": 0.41305360198020935, "learning_rate": 4.1453229230630894e-05, "loss": 0.028, "step": 115390 }, { "grad_norm": 0.8329823613166809, "learning_rate": 4.144508367530856e-05, "loss": 0.028, "step": 115400 }, { "grad_norm": 0.5682989954948425, "learning_rate": 4.143693835387443e-05, "loss": 0.0294, "step": 115410 }, { "grad_norm": 0.3928399682044983, "learning_rate": 4.142879326655117e-05, "loss": 0.0219, "step": 115420 }, { "grad_norm": 0.27811479568481445, "learning_rate": 4.142064841356151e-05, "loss": 0.023, "step": 115430 }, { "grad_norm": 0.41217270493507385, "learning_rate": 4.1412503795128075e-05, "loss": 0.0257, "step": 115440 }, { "grad_norm": 1.969862461090088, "learning_rate": 4.1404359411473556e-05, "loss": 0.0238, "step": 115450 }, { "grad_norm": 0.5222855806350708, "learning_rate": 4.1396215262820645e-05, "loss": 0.0184, "step": 115460 }, { "grad_norm": 0.5326257348060608, "learning_rate": 4.138807134939195e-05, "loss": 0.0316, "step": 115470 }, { "grad_norm": 0.5096906423568726, "learning_rate": 4.137992767141018e-05, "loss": 0.0513, "step": 115480 }, { "grad_norm": 0.39823034405708313, "learning_rate": 4.137178422909792e-05, "loss": 0.0267, "step": 115490 }, { "grad_norm": 0.36516618728637695, "learning_rate": 4.1363641022677854e-05, "loss": 0.0265, "step": 115500 }, { "grad_norm": 0.5412433743476868, "learning_rate": 4.135549805237259e-05, "loss": 0.0189, "step": 115510 }, { "grad_norm": 0.37337085604667664, "learning_rate": 4.134735531840476e-05, "loss": 0.0292, "step": 115520 }, { "grad_norm": 0.4177818298339844, "learning_rate": 4.1339212820996996e-05, "loss": 0.0252, "step": 115530 }, { "grad_norm": 0.7263476848602295, "learning_rate": 4.13310705603719e-05, "loss": 0.0363, "step": 115540 }, { "grad_norm": 0.5570985078811646, "learning_rate": 4.1322928536752073e-05, "loss": 0.0234, "step": 115550 }, { "grad_norm": 0.45408695936203003, "learning_rate": 4.1314786750360116e-05, "loss": 0.0217, "step": 115560 }, { "grad_norm": 0.7525935769081116, "learning_rate": 4.130664520141864e-05, "loss": 0.0217, "step": 115570 }, { "grad_norm": 0.34459927678108215, "learning_rate": 4.129850389015019e-05, "loss": 0.0276, "step": 115580 }, { "grad_norm": 0.9519197344779968, "learning_rate": 4.129036281677739e-05, "loss": 0.0586, "step": 115590 }, { "grad_norm": 0.26983657479286194, "learning_rate": 4.1282221981522806e-05, "loss": 0.0394, "step": 115600 }, { "grad_norm": 0.7704334855079651, "learning_rate": 4.1274081384608984e-05, "loss": 0.0331, "step": 115610 }, { "grad_norm": 0.641628623008728, "learning_rate": 4.126594102625852e-05, "loss": 0.0296, "step": 115620 }, { "grad_norm": 0.31437432765960693, "learning_rate": 4.125780090669393e-05, "loss": 0.0199, "step": 115630 }, { "grad_norm": 1.0926364660263062, "learning_rate": 4.124966102613779e-05, "loss": 0.0331, "step": 115640 }, { "grad_norm": 0.4939119219779968, "learning_rate": 4.124152138481262e-05, "loss": 0.0297, "step": 115650 }, { "grad_norm": 0.4752441346645355, "learning_rate": 4.1233381982940954e-05, "loss": 0.0272, "step": 115660 }, { "grad_norm": 0.35504966974258423, "learning_rate": 4.122524282074537e-05, "loss": 0.0206, "step": 115670 }, { "grad_norm": 1.1994816064834595, "learning_rate": 4.1217103898448305e-05, "loss": 0.045, "step": 115680 }, { "grad_norm": 1.2505521774291992, "learning_rate": 4.120896521627237e-05, "loss": 0.0256, "step": 115690 }, { "grad_norm": 0.39970284700393677, "learning_rate": 4.120082677443997e-05, "loss": 0.0515, "step": 115700 }, { "grad_norm": 0.36052122712135315, "learning_rate": 4.119268857317368e-05, "loss": 0.0269, "step": 115710 }, { "grad_norm": 0.9657973647117615, "learning_rate": 4.118455061269599e-05, "loss": 0.0308, "step": 115720 }, { "grad_norm": 0.4120177626609802, "learning_rate": 4.1176412893229366e-05, "loss": 0.025, "step": 115730 }, { "grad_norm": 0.7120218276977539, "learning_rate": 4.1168275414996307e-05, "loss": 0.0213, "step": 115740 }, { "grad_norm": 0.45893633365631104, "learning_rate": 4.1160138178219266e-05, "loss": 0.0335, "step": 115750 }, { "grad_norm": 1.0479353666305542, "learning_rate": 4.1152001183120735e-05, "loss": 0.0326, "step": 115760 }, { "grad_norm": 1.8379613161087036, "learning_rate": 4.1143864429923166e-05, "loss": 0.0413, "step": 115770 }, { "grad_norm": 0.2755104601383209, "learning_rate": 4.1135727918848995e-05, "loss": 0.0292, "step": 115780 }, { "grad_norm": 0.45092418789863586, "learning_rate": 4.1127591650120716e-05, "loss": 0.0352, "step": 115790 }, { "grad_norm": 0.4353630244731903, "learning_rate": 4.111945562396074e-05, "loss": 0.0254, "step": 115800 }, { "grad_norm": 0.37691599130630493, "learning_rate": 4.1111319840591525e-05, "loss": 0.0182, "step": 115810 }, { "grad_norm": 0.46848398447036743, "learning_rate": 4.110318430023547e-05, "loss": 0.022, "step": 115820 }, { "grad_norm": 0.5949046015739441, "learning_rate": 4.109504900311503e-05, "loss": 0.0598, "step": 115830 }, { "grad_norm": 1.0720576047897339, "learning_rate": 4.108691394945259e-05, "loss": 0.0362, "step": 115840 }, { "grad_norm": 0.6648176908493042, "learning_rate": 4.1078779139470575e-05, "loss": 0.0283, "step": 115850 }, { "grad_norm": 5.746581554412842, "learning_rate": 4.10706445733914e-05, "loss": 0.0388, "step": 115860 }, { "grad_norm": 0.509922206401825, "learning_rate": 4.1062510251437417e-05, "loss": 0.0203, "step": 115870 }, { "grad_norm": 0.5177599787712097, "learning_rate": 4.105437617383109e-05, "loss": 0.0352, "step": 115880 }, { "grad_norm": 0.38940873742103577, "learning_rate": 4.104624234079472e-05, "loss": 0.0248, "step": 115890 }, { "grad_norm": 0.3886270523071289, "learning_rate": 4.103810875255073e-05, "loss": 0.0251, "step": 115900 }, { "grad_norm": 0.4894479513168335, "learning_rate": 4.10299754093215e-05, "loss": 0.0257, "step": 115910 }, { "grad_norm": 0.5886867642402649, "learning_rate": 4.102184231132935e-05, "loss": 0.0334, "step": 115920 }, { "grad_norm": 0.8062565922737122, "learning_rate": 4.101370945879667e-05, "loss": 0.0286, "step": 115930 }, { "grad_norm": 0.4199325442314148, "learning_rate": 4.100557685194579e-05, "loss": 0.0198, "step": 115940 }, { "grad_norm": 0.6411290764808655, "learning_rate": 4.099744449099907e-05, "loss": 0.0272, "step": 115950 }, { "grad_norm": 0.5603621602058411, "learning_rate": 4.098931237617883e-05, "loss": 0.0324, "step": 115960 }, { "grad_norm": 0.46815282106399536, "learning_rate": 4.0981180507707386e-05, "loss": 0.0259, "step": 115970 }, { "grad_norm": 0.36187443137168884, "learning_rate": 4.097304888580712e-05, "loss": 0.0198, "step": 115980 }, { "grad_norm": 0.5487422347068787, "learning_rate": 4.0964917510700265e-05, "loss": 0.0305, "step": 115990 }, { "grad_norm": 0.4771730899810791, "learning_rate": 4.0956786382609196e-05, "loss": 0.0213, "step": 116000 }, { "grad_norm": 0.828246533870697, "learning_rate": 4.0948655501756175e-05, "loss": 0.026, "step": 116010 }, { "grad_norm": 1.4540598392486572, "learning_rate": 4.0940524868363525e-05, "loss": 0.0302, "step": 116020 }, { "grad_norm": 0.47412535548210144, "learning_rate": 4.0932394482653514e-05, "loss": 0.0315, "step": 116030 }, { "grad_norm": 0.22579839825630188, "learning_rate": 4.0924264344848436e-05, "loss": 0.0204, "step": 116040 }, { "grad_norm": 0.8452045321464539, "learning_rate": 4.091613445517056e-05, "loss": 0.0278, "step": 116050 }, { "grad_norm": 0.2447158396244049, "learning_rate": 4.090800481384215e-05, "loss": 0.0176, "step": 116060 }, { "grad_norm": 0.2925610840320587, "learning_rate": 4.0899875421085485e-05, "loss": 0.0378, "step": 116070 }, { "grad_norm": 0.263714462518692, "learning_rate": 4.0891746277122786e-05, "loss": 0.0262, "step": 116080 }, { "grad_norm": 0.5674587488174438, "learning_rate": 4.088361738217633e-05, "loss": 0.0257, "step": 116090 }, { "grad_norm": 3.2194464206695557, "learning_rate": 4.087548873646836e-05, "loss": 0.0285, "step": 116100 }, { "grad_norm": 0.3299110531806946, "learning_rate": 4.0867360340221095e-05, "loss": 0.0215, "step": 116110 }, { "grad_norm": 0.37232428789138794, "learning_rate": 4.0859232193656774e-05, "loss": 0.0266, "step": 116120 }, { "grad_norm": 0.47590476274490356, "learning_rate": 4.08511042969976e-05, "loss": 0.0266, "step": 116130 }, { "grad_norm": 0.301266074180603, "learning_rate": 4.084297665046581e-05, "loss": 0.022, "step": 116140 }, { "grad_norm": 0.6982625126838684, "learning_rate": 4.083484925428358e-05, "loss": 0.0254, "step": 116150 }, { "grad_norm": 0.6422826647758484, "learning_rate": 4.082672210867313e-05, "loss": 0.037, "step": 116160 }, { "grad_norm": 0.3276459872722626, "learning_rate": 4.081859521385668e-05, "loss": 0.0325, "step": 116170 }, { "grad_norm": 1.4082062244415283, "learning_rate": 4.081046857005635e-05, "loss": 0.0231, "step": 116180 }, { "grad_norm": 0.5038050413131714, "learning_rate": 4.080234217749439e-05, "loss": 0.0336, "step": 116190 }, { "grad_norm": 0.57932448387146, "learning_rate": 4.0794216036392906e-05, "loss": 0.0239, "step": 116200 }, { "grad_norm": 0.4557991921901703, "learning_rate": 4.078609014697412e-05, "loss": 0.0374, "step": 116210 }, { "grad_norm": 0.3474794626235962, "learning_rate": 4.077796450946016e-05, "loss": 0.0194, "step": 116220 }, { "grad_norm": 0.5424354672431946, "learning_rate": 4.076983912407319e-05, "loss": 0.0299, "step": 116230 }, { "grad_norm": 0.7912410497665405, "learning_rate": 4.076171399103535e-05, "loss": 0.0342, "step": 116240 }, { "grad_norm": 0.6320852041244507, "learning_rate": 4.075358911056877e-05, "loss": 0.0296, "step": 116250 }, { "grad_norm": 0.7257422804832458, "learning_rate": 4.0745464482895604e-05, "loss": 0.0327, "step": 116260 }, { "grad_norm": 0.49582332372665405, "learning_rate": 4.073734010823795e-05, "loss": 0.031, "step": 116270 }, { "grad_norm": 0.4161394238471985, "learning_rate": 4.072921598681794e-05, "loss": 0.0286, "step": 116280 }, { "grad_norm": 0.8477327227592468, "learning_rate": 4.0721092118857686e-05, "loss": 0.0308, "step": 116290 }, { "grad_norm": 0.3977977931499481, "learning_rate": 4.071296850457928e-05, "loss": 0.0317, "step": 116300 }, { "grad_norm": 0.48970460891723633, "learning_rate": 4.0704845144204845e-05, "loss": 0.0241, "step": 116310 }, { "grad_norm": 0.48346826434135437, "learning_rate": 4.0696722037956434e-05, "loss": 0.0296, "step": 116320 }, { "grad_norm": 0.46367835998535156, "learning_rate": 4.0688599186056167e-05, "loss": 0.0218, "step": 116330 }, { "grad_norm": 0.5150217413902283, "learning_rate": 4.068047658872608e-05, "loss": 0.0245, "step": 116340 }, { "grad_norm": 0.9777884483337402, "learning_rate": 4.067235424618826e-05, "loss": 0.0207, "step": 116350 }, { "grad_norm": 0.2976146340370178, "learning_rate": 4.066423215866479e-05, "loss": 0.028, "step": 116360 }, { "grad_norm": 0.6661417484283447, "learning_rate": 4.065611032637768e-05, "loss": 0.0335, "step": 116370 }, { "grad_norm": 0.5031684637069702, "learning_rate": 4.064798874954903e-05, "loss": 0.0269, "step": 116380 }, { "grad_norm": 0.48179754614830017, "learning_rate": 4.063986742840083e-05, "loss": 0.0593, "step": 116390 }, { "grad_norm": 0.4757014811038971, "learning_rate": 4.063174636315516e-05, "loss": 0.0282, "step": 116400 }, { "grad_norm": 0.2883806526660919, "learning_rate": 4.062362555403401e-05, "loss": 0.0234, "step": 116410 }, { "grad_norm": 0.3905320465564728, "learning_rate": 4.0615505001259414e-05, "loss": 0.0295, "step": 116420 }, { "grad_norm": 0.9497338533401489, "learning_rate": 4.0607384705053396e-05, "loss": 0.0258, "step": 116430 }, { "grad_norm": 0.38216158747673035, "learning_rate": 4.0599264665637946e-05, "loss": 0.0294, "step": 116440 }, { "grad_norm": 1.2741881608963013, "learning_rate": 4.059114488323508e-05, "loss": 0.0207, "step": 116450 }, { "grad_norm": 0.5071540474891663, "learning_rate": 4.058302535806676e-05, "loss": 0.0217, "step": 116460 }, { "grad_norm": 0.5254930257797241, "learning_rate": 4.057490609035498e-05, "loss": 0.0314, "step": 116470 }, { "grad_norm": 0.3082949221134186, "learning_rate": 4.056678708032176e-05, "loss": 0.0303, "step": 116480 }, { "grad_norm": 0.7863908410072327, "learning_rate": 4.055866832818901e-05, "loss": 0.0324, "step": 116490 }, { "grad_norm": 0.3300810158252716, "learning_rate": 4.055054983417874e-05, "loss": 0.0214, "step": 116500 }, { "grad_norm": 0.45667579770088196, "learning_rate": 4.054243159851287e-05, "loss": 0.0274, "step": 116510 }, { "grad_norm": 0.2966972887516022, "learning_rate": 4.053431362141338e-05, "loss": 0.0247, "step": 116520 }, { "grad_norm": 0.4744175672531128, "learning_rate": 4.0526195903102196e-05, "loss": 0.0316, "step": 116530 }, { "grad_norm": 0.4131495952606201, "learning_rate": 4.051807844380125e-05, "loss": 0.0225, "step": 116540 }, { "grad_norm": 0.37995532155036926, "learning_rate": 4.0509961243732494e-05, "loss": 0.0185, "step": 116550 }, { "grad_norm": 0.42810335755348206, "learning_rate": 4.050184430311782e-05, "loss": 0.0316, "step": 116560 }, { "grad_norm": 0.7476525902748108, "learning_rate": 4.049372762217916e-05, "loss": 0.0362, "step": 116570 }, { "grad_norm": 0.46708711981773376, "learning_rate": 4.0485611201138396e-05, "loss": 0.0293, "step": 116580 }, { "grad_norm": 0.32899153232574463, "learning_rate": 4.047749504021748e-05, "loss": 0.0265, "step": 116590 }, { "grad_norm": 0.6191230416297913, "learning_rate": 4.046937913963824e-05, "loss": 0.0219, "step": 116600 }, { "grad_norm": 1.842949628829956, "learning_rate": 4.046126349962261e-05, "loss": 0.0287, "step": 116610 }, { "grad_norm": 0.6346182823181152, "learning_rate": 4.045314812039246e-05, "loss": 0.0255, "step": 116620 }, { "grad_norm": 1.2531987428665161, "learning_rate": 4.044503300216964e-05, "loss": 0.0295, "step": 116630 }, { "grad_norm": 0.43883875012397766, "learning_rate": 4.043691814517604e-05, "loss": 0.0182, "step": 116640 }, { "grad_norm": 0.4740793704986572, "learning_rate": 4.0428803549633496e-05, "loss": 0.021, "step": 116650 }, { "grad_norm": 0.5876947641372681, "learning_rate": 4.042068921576385e-05, "loss": 0.0493, "step": 116660 }, { "grad_norm": 0.47976693511009216, "learning_rate": 4.0412575143789e-05, "loss": 0.019, "step": 116670 }, { "grad_norm": 0.661298394203186, "learning_rate": 4.040446133393071e-05, "loss": 0.0274, "step": 116680 }, { "grad_norm": 0.46699604392051697, "learning_rate": 4.039634778641088e-05, "loss": 0.0302, "step": 116690 }, { "grad_norm": 0.5140056610107422, "learning_rate": 4.038823450145126e-05, "loss": 0.0197, "step": 116700 }, { "grad_norm": 0.5421626567840576, "learning_rate": 4.038012147927371e-05, "loss": 0.0258, "step": 116710 }, { "grad_norm": 0.6712685227394104, "learning_rate": 4.0372008720100026e-05, "loss": 0.0317, "step": 116720 }, { "grad_norm": 0.49403804540634155, "learning_rate": 4.0363896224152e-05, "loss": 0.0317, "step": 116730 }, { "grad_norm": 0.643886148929596, "learning_rate": 4.035578399165144e-05, "loss": 0.0302, "step": 116740 }, { "grad_norm": 0.4959730803966522, "learning_rate": 4.0347672022820115e-05, "loss": 0.0261, "step": 116750 }, { "grad_norm": 0.2819235026836395, "learning_rate": 4.033956031787982e-05, "loss": 0.0235, "step": 116760 }, { "grad_norm": 0.4802570044994354, "learning_rate": 4.0331448877052306e-05, "loss": 0.0257, "step": 116770 }, { "grad_norm": 0.27430668473243713, "learning_rate": 4.0323337700559363e-05, "loss": 0.0291, "step": 116780 }, { "grad_norm": 0.768020749092102, "learning_rate": 4.03152267886227e-05, "loss": 0.0232, "step": 116790 }, { "grad_norm": 0.5364721417427063, "learning_rate": 4.030711614146412e-05, "loss": 0.0214, "step": 116800 }, { "grad_norm": 0.4185499846935272, "learning_rate": 4.029900575930535e-05, "loss": 0.0289, "step": 116810 }, { "grad_norm": 0.3442559540271759, "learning_rate": 4.029089564236811e-05, "loss": 0.0236, "step": 116820 }, { "grad_norm": 0.41208773851394653, "learning_rate": 4.028278579087413e-05, "loss": 0.0231, "step": 116830 }, { "grad_norm": 0.45667845010757446, "learning_rate": 4.0274676205045145e-05, "loss": 0.0286, "step": 116840 }, { "grad_norm": 0.5387314558029175, "learning_rate": 4.026656688510286e-05, "loss": 0.0333, "step": 116850 }, { "grad_norm": 0.46317824721336365, "learning_rate": 4.025845783126897e-05, "loss": 0.0341, "step": 116860 }, { "grad_norm": 0.36205318570137024, "learning_rate": 4.025034904376518e-05, "loss": 0.0238, "step": 116870 }, { "grad_norm": 0.4551897943019867, "learning_rate": 4.0242240522813206e-05, "loss": 0.0223, "step": 116880 }, { "grad_norm": 0.46671950817108154, "learning_rate": 4.0234132268634686e-05, "loss": 0.0264, "step": 116890 }, { "grad_norm": 0.37570706009864807, "learning_rate": 4.0226024281451345e-05, "loss": 0.0338, "step": 116900 }, { "grad_norm": 0.34298792481422424, "learning_rate": 4.02179165614848e-05, "loss": 0.0176, "step": 116910 }, { "grad_norm": 0.4418301284313202, "learning_rate": 4.0209809108956755e-05, "loss": 0.0282, "step": 116920 }, { "grad_norm": 0.4784832298755646, "learning_rate": 4.0201701924088864e-05, "loss": 0.0256, "step": 116930 }, { "grad_norm": 0.37192583084106445, "learning_rate": 4.0193595007102744e-05, "loss": 0.0249, "step": 116940 }, { "grad_norm": 0.9074841141700745, "learning_rate": 4.018548835822006e-05, "loss": 0.0235, "step": 116950 }, { "grad_norm": 0.49894979596138, "learning_rate": 4.017738197766244e-05, "loss": 0.0272, "step": 116960 }, { "grad_norm": 1.0497591495513916, "learning_rate": 4.0169275865651504e-05, "loss": 0.0399, "step": 116970 }, { "grad_norm": 0.3759394586086273, "learning_rate": 4.016117002240887e-05, "loss": 0.0449, "step": 116980 }, { "grad_norm": 0.6811161637306213, "learning_rate": 4.0153064448156135e-05, "loss": 0.0331, "step": 116990 }, { "grad_norm": 0.44714173674583435, "learning_rate": 4.014495914311494e-05, "loss": 0.0395, "step": 117000 }, { "grad_norm": 0.7148145437240601, "learning_rate": 4.013685410750686e-05, "loss": 0.0303, "step": 117010 }, { "grad_norm": 0.5263222455978394, "learning_rate": 4.012874934155348e-05, "loss": 0.0206, "step": 117020 }, { "grad_norm": 0.41182741522789, "learning_rate": 4.012064484547638e-05, "loss": 0.0307, "step": 117030 }, { "grad_norm": 0.3039131164550781, "learning_rate": 4.0112540619497145e-05, "loss": 0.024, "step": 117040 }, { "grad_norm": 0.3730034828186035, "learning_rate": 4.010443666383733e-05, "loss": 0.0222, "step": 117050 }, { "grad_norm": 0.495914101600647, "learning_rate": 4.0096332978718495e-05, "loss": 0.031, "step": 117060 }, { "grad_norm": 0.4997846782207489, "learning_rate": 4.008822956436221e-05, "loss": 0.0323, "step": 117070 }, { "grad_norm": 0.33919578790664673, "learning_rate": 4.0080126420989985e-05, "loss": 0.0196, "step": 117080 }, { "grad_norm": 0.4122598171234131, "learning_rate": 4.00720235488234e-05, "loss": 0.0257, "step": 117090 }, { "grad_norm": 0.29661139845848083, "learning_rate": 4.006392094808393e-05, "loss": 0.0298, "step": 117100 }, { "grad_norm": 0.6335052251815796, "learning_rate": 4.005581861899315e-05, "loss": 0.0241, "step": 117110 }, { "grad_norm": 0.7303975820541382, "learning_rate": 4.0047716561772556e-05, "loss": 0.0311, "step": 117120 }, { "grad_norm": 0.3881579041481018, "learning_rate": 4.003961477664365e-05, "loss": 0.0247, "step": 117130 }, { "grad_norm": 0.8767660856246948, "learning_rate": 4.0031513263827934e-05, "loss": 0.0271, "step": 117140 }, { "grad_norm": 0.4211919903755188, "learning_rate": 4.002341202354689e-05, "loss": 0.0381, "step": 117150 }, { "grad_norm": 0.44945240020751953, "learning_rate": 4.0015311056022034e-05, "loss": 0.0224, "step": 117160 }, { "grad_norm": 1.0554527044296265, "learning_rate": 4.0007210361474806e-05, "loss": 0.0254, "step": 117170 }, { "grad_norm": 0.6983800530433655, "learning_rate": 3.999910994012668e-05, "loss": 0.0288, "step": 117180 }, { "grad_norm": 0.3019547164440155, "learning_rate": 3.999100979219917e-05, "loss": 0.0202, "step": 117190 }, { "grad_norm": 0.6003718376159668, "learning_rate": 3.9982909917913644e-05, "loss": 0.0276, "step": 117200 }, { "grad_norm": 0.5993437170982361, "learning_rate": 3.9974810317491635e-05, "loss": 0.0364, "step": 117210 }, { "grad_norm": 0.41644778847694397, "learning_rate": 3.996671099115453e-05, "loss": 0.031, "step": 117220 }, { "grad_norm": 0.33332058787345886, "learning_rate": 3.995861193912379e-05, "loss": 0.026, "step": 117230 }, { "grad_norm": 0.9714221358299255, "learning_rate": 3.995051316162081e-05, "loss": 0.031, "step": 117240 }, { "grad_norm": 0.5955374240875244, "learning_rate": 3.9942414658867035e-05, "loss": 0.0291, "step": 117250 }, { "grad_norm": 0.7955783009529114, "learning_rate": 3.993431643108387e-05, "loss": 0.0303, "step": 117260 }, { "grad_norm": 0.5263422131538391, "learning_rate": 3.992621847849269e-05, "loss": 0.0331, "step": 117270 }, { "grad_norm": 0.3079717755317688, "learning_rate": 3.991812080131494e-05, "loss": 0.0276, "step": 117280 }, { "grad_norm": 0.5083722472190857, "learning_rate": 3.991002339977196e-05, "loss": 0.023, "step": 117290 }, { "grad_norm": 0.7651686668395996, "learning_rate": 3.9901926274085146e-05, "loss": 0.0343, "step": 117300 }, { "grad_norm": 0.4931994378566742, "learning_rate": 3.989382942447589e-05, "loss": 0.0269, "step": 117310 }, { "grad_norm": 0.33885809779167175, "learning_rate": 3.988573285116554e-05, "loss": 0.0223, "step": 117320 }, { "grad_norm": 0.2561999261379242, "learning_rate": 3.987763655437545e-05, "loss": 0.0297, "step": 117330 }, { "grad_norm": 0.5959411859512329, "learning_rate": 3.986954053432697e-05, "loss": 0.0276, "step": 117340 }, { "grad_norm": 0.6497109532356262, "learning_rate": 3.986144479124145e-05, "loss": 0.0321, "step": 117350 }, { "grad_norm": 0.2706262767314911, "learning_rate": 3.985334932534022e-05, "loss": 0.032, "step": 117360 }, { "grad_norm": 0.7988306283950806, "learning_rate": 3.9845254136844593e-05, "loss": 0.0303, "step": 117370 }, { "grad_norm": 0.38391444087028503, "learning_rate": 3.9837159225975935e-05, "loss": 0.0252, "step": 117380 }, { "grad_norm": 0.7649648189544678, "learning_rate": 3.98290645929555e-05, "loss": 0.0298, "step": 117390 }, { "grad_norm": 0.33222925662994385, "learning_rate": 3.982097023800464e-05, "loss": 0.0232, "step": 117400 }, { "grad_norm": 0.4012901186943054, "learning_rate": 3.98128761613446e-05, "loss": 0.0247, "step": 117410 }, { "grad_norm": 0.614817202091217, "learning_rate": 3.9804782363196716e-05, "loss": 0.0366, "step": 117420 }, { "grad_norm": 0.28917771577835083, "learning_rate": 3.979668884378225e-05, "loss": 0.0228, "step": 117430 }, { "grad_norm": 0.29679805040359497, "learning_rate": 3.978859560332247e-05, "loss": 0.02, "step": 117440 }, { "grad_norm": 0.8135848045349121, "learning_rate": 3.978050264203867e-05, "loss": 0.0301, "step": 117450 }, { "grad_norm": 0.3052690625190735, "learning_rate": 3.9772409960152066e-05, "loss": 0.0207, "step": 117460 }, { "grad_norm": 0.44181615114212036, "learning_rate": 3.9764317557883954e-05, "loss": 0.0242, "step": 117470 }, { "grad_norm": 1.3623303174972534, "learning_rate": 3.975622543545554e-05, "loss": 0.026, "step": 117480 }, { "grad_norm": 0.9351062774658203, "learning_rate": 3.974813359308806e-05, "loss": 0.0322, "step": 117490 }, { "grad_norm": 0.6202865242958069, "learning_rate": 3.974004203100278e-05, "loss": 0.023, "step": 117500 }, { "grad_norm": 0.37397101521492004, "learning_rate": 3.9731950749420886e-05, "loss": 0.0236, "step": 117510 }, { "grad_norm": 0.41201478242874146, "learning_rate": 3.972385974856361e-05, "loss": 0.0288, "step": 117520 }, { "grad_norm": 0.504863977432251, "learning_rate": 3.971576902865214e-05, "loss": 0.0361, "step": 117530 }, { "grad_norm": 0.8119965195655823, "learning_rate": 3.9707678589907686e-05, "loss": 0.0226, "step": 117540 }, { "grad_norm": 0.6046937704086304, "learning_rate": 3.9699588432551424e-05, "loss": 0.0379, "step": 117550 }, { "grad_norm": 0.7980978488922119, "learning_rate": 3.9691498556804554e-05, "loss": 0.0233, "step": 117560 }, { "grad_norm": 0.3971122205257416, "learning_rate": 3.968340896288824e-05, "loss": 0.0184, "step": 117570 }, { "grad_norm": 1.6710290908813477, "learning_rate": 3.9675319651023636e-05, "loss": 0.0235, "step": 117580 }, { "grad_norm": 0.6736565232276917, "learning_rate": 3.9667230621431936e-05, "loss": 0.0279, "step": 117590 }, { "grad_norm": 0.3602178990840912, "learning_rate": 3.965914187433424e-05, "loss": 0.0206, "step": 117600 }, { "grad_norm": 0.31846868991851807, "learning_rate": 3.965105340995176e-05, "loss": 0.0495, "step": 117610 }, { "grad_norm": 0.352364718914032, "learning_rate": 3.964296522850556e-05, "loss": 0.0526, "step": 117620 }, { "grad_norm": 0.26027554273605347, "learning_rate": 3.9634877330216803e-05, "loss": 0.0435, "step": 117630 }, { "grad_norm": 0.2594195306301117, "learning_rate": 3.962678971530662e-05, "loss": 0.0451, "step": 117640 }, { "grad_norm": 0.34593573212623596, "learning_rate": 3.961870238399609e-05, "loss": 0.0137, "step": 117650 }, { "grad_norm": 0.7106725573539734, "learning_rate": 3.961061533650636e-05, "loss": 0.0305, "step": 117660 }, { "grad_norm": 0.44701558351516724, "learning_rate": 3.960252857305848e-05, "loss": 0.049, "step": 117670 }, { "grad_norm": 0.5011246800422668, "learning_rate": 3.959444209387356e-05, "loss": 0.0288, "step": 117680 }, { "grad_norm": 0.28912317752838135, "learning_rate": 3.9586355899172726e-05, "loss": 0.025, "step": 117690 }, { "grad_norm": 0.31873512268066406, "learning_rate": 3.9578269989176963e-05, "loss": 0.0231, "step": 117700 }, { "grad_norm": 0.46097368001937866, "learning_rate": 3.9570184364107404e-05, "loss": 0.0283, "step": 117710 }, { "grad_norm": 0.3312443494796753, "learning_rate": 3.9562099024185084e-05, "loss": 0.0281, "step": 117720 }, { "grad_norm": 0.2887021005153656, "learning_rate": 3.955401396963106e-05, "loss": 0.0388, "step": 117730 }, { "grad_norm": 0.5214877724647522, "learning_rate": 3.954592920066636e-05, "loss": 0.0299, "step": 117740 }, { "grad_norm": 0.5496942400932312, "learning_rate": 3.953784471751203e-05, "loss": 0.0394, "step": 117750 }, { "grad_norm": 0.5177728533744812, "learning_rate": 3.9529760520389104e-05, "loss": 0.0323, "step": 117760 }, { "grad_norm": 0.3703344762325287, "learning_rate": 3.952167660951858e-05, "loss": 0.0229, "step": 117770 }, { "grad_norm": 0.48192059993743896, "learning_rate": 3.9513592985121495e-05, "loss": 0.02, "step": 117780 }, { "grad_norm": 0.5707635283470154, "learning_rate": 3.950550964741881e-05, "loss": 0.0242, "step": 117790 }, { "grad_norm": 0.4291606545448303, "learning_rate": 3.9497426596631594e-05, "loss": 0.0215, "step": 117800 }, { "grad_norm": 0.33540552854537964, "learning_rate": 3.948934383298075e-05, "loss": 0.0395, "step": 117810 }, { "grad_norm": 0.4044581949710846, "learning_rate": 3.9481261356687306e-05, "loss": 0.0277, "step": 117820 }, { "grad_norm": 0.6707655191421509, "learning_rate": 3.947317916797224e-05, "loss": 0.0321, "step": 117830 }, { "grad_norm": 0.3077675998210907, "learning_rate": 3.9465097267056486e-05, "loss": 0.032, "step": 117840 }, { "grad_norm": 0.5394989848136902, "learning_rate": 3.9457015654161024e-05, "loss": 0.0379, "step": 117850 }, { "grad_norm": 0.3129247725009918, "learning_rate": 3.944893432950679e-05, "loss": 0.0267, "step": 117860 }, { "grad_norm": 0.38753652572631836, "learning_rate": 3.9440853293314704e-05, "loss": 0.0261, "step": 117870 }, { "grad_norm": 0.4594953954219818, "learning_rate": 3.943277254580576e-05, "loss": 0.0264, "step": 117880 }, { "grad_norm": 1.247202754020691, "learning_rate": 3.942469208720082e-05, "loss": 0.0403, "step": 117890 }, { "grad_norm": 0.39366278052330017, "learning_rate": 3.941661191772085e-05, "loss": 0.0294, "step": 117900 }, { "grad_norm": 0.4448227286338806, "learning_rate": 3.940853203758671e-05, "loss": 0.023, "step": 117910 }, { "grad_norm": 0.5525439977645874, "learning_rate": 3.940045244701934e-05, "loss": 0.0311, "step": 117920 }, { "grad_norm": 0.3561766445636749, "learning_rate": 3.93923731462396e-05, "loss": 0.0189, "step": 117930 }, { "grad_norm": 0.2401907593011856, "learning_rate": 3.9384294135468404e-05, "loss": 0.0203, "step": 117940 }, { "grad_norm": 0.38828203082084656, "learning_rate": 3.937621541492662e-05, "loss": 0.0221, "step": 117950 }, { "grad_norm": 0.9687463641166687, "learning_rate": 3.9368136984835116e-05, "loss": 0.0224, "step": 117960 }, { "grad_norm": 0.33188989758491516, "learning_rate": 3.936005884541476e-05, "loss": 0.0287, "step": 117970 }, { "grad_norm": 0.24203509092330933, "learning_rate": 3.935198099688638e-05, "loss": 0.0225, "step": 117980 }, { "grad_norm": 0.3022899925708771, "learning_rate": 3.934390343947086e-05, "loss": 0.0269, "step": 117990 }, { "grad_norm": 0.21449732780456543, "learning_rate": 3.933582617338899e-05, "loss": 0.033, "step": 118000 }, { "grad_norm": 0.5688530802726746, "learning_rate": 3.932774919886164e-05, "loss": 0.0269, "step": 118010 }, { "grad_norm": 0.44820940494537354, "learning_rate": 3.931967251610962e-05, "loss": 0.0267, "step": 118020 }, { "grad_norm": 0.3202359676361084, "learning_rate": 3.931159612535374e-05, "loss": 0.0224, "step": 118030 }, { "grad_norm": 0.7510448098182678, "learning_rate": 3.930352002681481e-05, "loss": 0.036, "step": 118040 }, { "grad_norm": 0.4106687605381012, "learning_rate": 3.9295444220713615e-05, "loss": 0.0224, "step": 118050 }, { "grad_norm": 0.3156684339046478, "learning_rate": 3.9287368707270964e-05, "loss": 0.0204, "step": 118060 }, { "grad_norm": 0.29748594760894775, "learning_rate": 3.927929348670763e-05, "loss": 0.0308, "step": 118070 }, { "grad_norm": 0.3852074146270752, "learning_rate": 3.9271218559244356e-05, "loss": 0.0215, "step": 118080 }, { "grad_norm": 0.6856679916381836, "learning_rate": 3.926314392510197e-05, "loss": 0.0306, "step": 118090 }, { "grad_norm": 0.41449838876724243, "learning_rate": 3.925506958450117e-05, "loss": 0.0308, "step": 118100 }, { "grad_norm": 0.6851648092269897, "learning_rate": 3.924699553766276e-05, "loss": 0.0404, "step": 118110 }, { "grad_norm": 0.35640573501586914, "learning_rate": 3.9238921784807426e-05, "loss": 0.0218, "step": 118120 }, { "grad_norm": 0.2526971697807312, "learning_rate": 3.923084832615593e-05, "loss": 0.0235, "step": 118130 }, { "grad_norm": 0.41857409477233887, "learning_rate": 3.922277516192901e-05, "loss": 0.0247, "step": 118140 }, { "grad_norm": 0.45923981070518494, "learning_rate": 3.921470229234736e-05, "loss": 0.0303, "step": 118150 }, { "grad_norm": 0.29813316464424133, "learning_rate": 3.92066297176317e-05, "loss": 0.0361, "step": 118160 }, { "grad_norm": 0.6784026026725769, "learning_rate": 3.9198557438002726e-05, "loss": 0.0318, "step": 118170 }, { "grad_norm": 0.32816001772880554, "learning_rate": 3.9190485453681136e-05, "loss": 0.0218, "step": 118180 }, { "grad_norm": 0.46837785840034485, "learning_rate": 3.918241376488761e-05, "loss": 0.0251, "step": 118190 }, { "grad_norm": 0.47487038373947144, "learning_rate": 3.917434237184281e-05, "loss": 0.0295, "step": 118200 }, { "grad_norm": 0.508162260055542, "learning_rate": 3.916627127476745e-05, "loss": 0.02, "step": 118210 }, { "grad_norm": 0.3296204209327698, "learning_rate": 3.915820047388215e-05, "loss": 0.0161, "step": 118220 }, { "grad_norm": 0.4512065649032593, "learning_rate": 3.915012996940759e-05, "loss": 0.0262, "step": 118230 }, { "grad_norm": 0.3485565483570099, "learning_rate": 3.914205976156439e-05, "loss": 0.0255, "step": 118240 }, { "grad_norm": 0.797653079032898, "learning_rate": 3.9133989850573204e-05, "loss": 0.0308, "step": 118250 }, { "grad_norm": 0.6275238394737244, "learning_rate": 3.912592023665464e-05, "loss": 0.0236, "step": 118260 }, { "grad_norm": 0.5747489333152771, "learning_rate": 3.9117850920029335e-05, "loss": 0.0243, "step": 118270 }, { "grad_norm": 0.47493860125541687, "learning_rate": 3.910978190091791e-05, "loss": 0.0311, "step": 118280 }, { "grad_norm": 0.4544287323951721, "learning_rate": 3.910171317954093e-05, "loss": 0.0299, "step": 118290 }, { "grad_norm": 0.3073090612888336, "learning_rate": 3.909364475611906e-05, "loss": 0.0297, "step": 118300 }, { "grad_norm": 0.8784205317497253, "learning_rate": 3.90855766308728e-05, "loss": 0.0389, "step": 118310 }, { "grad_norm": 0.3651144206523895, "learning_rate": 3.907750880402279e-05, "loss": 0.0279, "step": 118320 }, { "grad_norm": 0.8652234077453613, "learning_rate": 3.90694412757896e-05, "loss": 0.022, "step": 118330 }, { "grad_norm": 0.5253205299377441, "learning_rate": 3.906137404639377e-05, "loss": 0.0211, "step": 118340 }, { "grad_norm": 0.3054819107055664, "learning_rate": 3.9053307116055874e-05, "loss": 0.0216, "step": 118350 }, { "grad_norm": 0.44269314408302307, "learning_rate": 3.904524048499643e-05, "loss": 0.0312, "step": 118360 }, { "grad_norm": 1.2888091802597046, "learning_rate": 3.9037174153436015e-05, "loss": 0.0268, "step": 118370 }, { "grad_norm": 0.4360364079475403, "learning_rate": 3.902910812159512e-05, "loss": 0.019, "step": 118380 }, { "grad_norm": 0.9552055597305298, "learning_rate": 3.9021042389694285e-05, "loss": 0.0166, "step": 118390 }, { "grad_norm": 0.4800926148891449, "learning_rate": 3.9012976957954055e-05, "loss": 0.0221, "step": 118400 }, { "grad_norm": 0.20590873062610626, "learning_rate": 3.9004911826594876e-05, "loss": 0.0232, "step": 118410 }, { "grad_norm": 0.5008033514022827, "learning_rate": 3.89968469958373e-05, "loss": 0.0247, "step": 118420 }, { "grad_norm": 0.522729218006134, "learning_rate": 3.898878246590178e-05, "loss": 0.0166, "step": 118430 }, { "grad_norm": 0.7048375606536865, "learning_rate": 3.8980718237008825e-05, "loss": 0.0266, "step": 118440 }, { "grad_norm": 0.46999943256378174, "learning_rate": 3.897265430937888e-05, "loss": 0.0218, "step": 118450 }, { "grad_norm": 0.4237174391746521, "learning_rate": 3.8964590683232424e-05, "loss": 0.0253, "step": 118460 }, { "grad_norm": 0.329531729221344, "learning_rate": 3.895652735878992e-05, "loss": 0.023, "step": 118470 }, { "grad_norm": 0.9171802997589111, "learning_rate": 3.8948464336271805e-05, "loss": 0.0291, "step": 118480 }, { "grad_norm": 1.0010491609573364, "learning_rate": 3.8940401615898534e-05, "loss": 0.0289, "step": 118490 }, { "grad_norm": 0.4801982045173645, "learning_rate": 3.89323391978905e-05, "loss": 0.0179, "step": 118500 }, { "grad_norm": 0.47274142503738403, "learning_rate": 3.892427708246818e-05, "loss": 0.0216, "step": 118510 }, { "grad_norm": 0.42070141434669495, "learning_rate": 3.891621526985195e-05, "loss": 0.015, "step": 118520 }, { "grad_norm": 1.2688525915145874, "learning_rate": 3.890815376026224e-05, "loss": 0.0207, "step": 118530 }, { "grad_norm": 0.20037078857421875, "learning_rate": 3.890009255391944e-05, "loss": 0.0192, "step": 118540 }, { "grad_norm": 0.4223974049091339, "learning_rate": 3.8892031651043934e-05, "loss": 0.0276, "step": 118550 }, { "grad_norm": 0.521273672580719, "learning_rate": 3.888397105185612e-05, "loss": 0.0297, "step": 118560 }, { "grad_norm": 0.7089625000953674, "learning_rate": 3.8875910756576346e-05, "loss": 0.0231, "step": 118570 }, { "grad_norm": 0.6949350833892822, "learning_rate": 3.8867850765424986e-05, "loss": 0.024, "step": 118580 }, { "grad_norm": 0.6385612487792969, "learning_rate": 3.885979107862243e-05, "loss": 0.0273, "step": 118590 }, { "grad_norm": 0.48354947566986084, "learning_rate": 3.8851731696388975e-05, "loss": 0.0198, "step": 118600 }, { "grad_norm": 0.4022628664970398, "learning_rate": 3.884367261894502e-05, "loss": 0.0323, "step": 118610 }, { "grad_norm": 0.2702462077140808, "learning_rate": 3.883561384651083e-05, "loss": 0.0288, "step": 118620 }, { "grad_norm": 0.2713114619255066, "learning_rate": 3.8827555379306786e-05, "loss": 0.0272, "step": 118630 }, { "grad_norm": 0.2913482189178467, "learning_rate": 3.881949721755317e-05, "loss": 0.0242, "step": 118640 }, { "grad_norm": 0.5191738605499268, "learning_rate": 3.881143936147029e-05, "loss": 0.0197, "step": 118650 }, { "grad_norm": 0.2038082480430603, "learning_rate": 3.880338181127848e-05, "loss": 0.0236, "step": 118660 }, { "grad_norm": 0.8450612425804138, "learning_rate": 3.879532456719799e-05, "loss": 0.024, "step": 118670 }, { "grad_norm": 0.26772165298461914, "learning_rate": 3.878726762944912e-05, "loss": 0.023, "step": 118680 }, { "grad_norm": 0.42910563945770264, "learning_rate": 3.877921099825214e-05, "loss": 0.0202, "step": 118690 }, { "grad_norm": 0.35253867506980896, "learning_rate": 3.87711546738273e-05, "loss": 0.0292, "step": 118700 }, { "grad_norm": 0.43223780393600464, "learning_rate": 3.876309865639489e-05, "loss": 0.0243, "step": 118710 }, { "grad_norm": 0.48515748977661133, "learning_rate": 3.875504294617514e-05, "loss": 0.0362, "step": 118720 }, { "grad_norm": 0.3390304148197174, "learning_rate": 3.874698754338829e-05, "loss": 0.0229, "step": 118730 }, { "grad_norm": 0.5188717842102051, "learning_rate": 3.873893244825457e-05, "loss": 0.0311, "step": 118740 }, { "grad_norm": 0.3613761365413666, "learning_rate": 3.873087766099422e-05, "loss": 0.0181, "step": 118750 }, { "grad_norm": 0.5036362409591675, "learning_rate": 3.872282318182743e-05, "loss": 0.0335, "step": 118760 }, { "grad_norm": 0.9739229679107666, "learning_rate": 3.871476901097441e-05, "loss": 0.0239, "step": 118770 }, { "grad_norm": 0.5189883708953857, "learning_rate": 3.870671514865538e-05, "loss": 0.0251, "step": 118780 }, { "grad_norm": 0.5696446299552917, "learning_rate": 3.8698661595090494e-05, "loss": 0.027, "step": 118790 }, { "grad_norm": 1.1966474056243896, "learning_rate": 3.869060835049999e-05, "loss": 0.029, "step": 118800 }, { "grad_norm": 0.45146793127059937, "learning_rate": 3.8682555415103964e-05, "loss": 0.0207, "step": 118810 }, { "grad_norm": 0.28689900040626526, "learning_rate": 3.867450278912266e-05, "loss": 0.0221, "step": 118820 }, { "grad_norm": 0.5068771839141846, "learning_rate": 3.866645047277616e-05, "loss": 0.0298, "step": 118830 }, { "grad_norm": 0.44596561789512634, "learning_rate": 3.865839846628466e-05, "loss": 0.025, "step": 118840 }, { "grad_norm": 0.29456159472465515, "learning_rate": 3.86503467698683e-05, "loss": 0.031, "step": 118850 }, { "grad_norm": 0.786732017993927, "learning_rate": 3.864229538374717e-05, "loss": 0.0349, "step": 118860 }, { "grad_norm": 0.3923402428627014, "learning_rate": 3.863424430814144e-05, "loss": 0.0282, "step": 118870 }, { "grad_norm": 0.3280521035194397, "learning_rate": 3.862619354327119e-05, "loss": 0.0232, "step": 118880 }, { "grad_norm": 0.5031744241714478, "learning_rate": 3.861814308935652e-05, "loss": 0.0292, "step": 118890 }, { "grad_norm": 0.4662975072860718, "learning_rate": 3.8610092946617584e-05, "loss": 0.0166, "step": 118900 }, { "grad_norm": 0.5467290878295898, "learning_rate": 3.8602043115274386e-05, "loss": 0.0354, "step": 118910 }, { "grad_norm": 0.4682461619377136, "learning_rate": 3.859399359554707e-05, "loss": 0.0172, "step": 118920 }, { "grad_norm": 0.3905239403247833, "learning_rate": 3.858594438765567e-05, "loss": 0.0269, "step": 118930 }, { "grad_norm": 0.43207719922065735, "learning_rate": 3.857789549182028e-05, "loss": 0.0207, "step": 118940 }, { "grad_norm": 0.7779498100280762, "learning_rate": 3.856984690826092e-05, "loss": 0.0316, "step": 118950 }, { "grad_norm": 0.5198182463645935, "learning_rate": 3.856179863719765e-05, "loss": 0.0333, "step": 118960 }, { "grad_norm": 0.5512000918388367, "learning_rate": 3.8553750678850524e-05, "loss": 0.0269, "step": 118970 }, { "grad_norm": 0.45929205417633057, "learning_rate": 3.854570303343955e-05, "loss": 0.0282, "step": 118980 }, { "grad_norm": 0.6296274065971375, "learning_rate": 3.853765570118475e-05, "loss": 0.0226, "step": 118990 }, { "grad_norm": 0.7999004125595093, "learning_rate": 3.852960868230612e-05, "loss": 0.0254, "step": 119000 }, { "grad_norm": 0.4359738528728485, "learning_rate": 3.852156197702371e-05, "loss": 0.027, "step": 119010 }, { "grad_norm": 0.27323272824287415, "learning_rate": 3.851351558555746e-05, "loss": 0.0165, "step": 119020 }, { "grad_norm": 1.0305432081222534, "learning_rate": 3.850546950812738e-05, "loss": 0.0276, "step": 119030 }, { "grad_norm": 0.37062105536460876, "learning_rate": 3.8497423744953454e-05, "loss": 0.0318, "step": 119040 }, { "grad_norm": 1.0814560651779175, "learning_rate": 3.848937829625564e-05, "loss": 0.025, "step": 119050 }, { "grad_norm": 0.2954884171485901, "learning_rate": 3.84813331622539e-05, "loss": 0.0243, "step": 119060 }, { "grad_norm": 0.572236955165863, "learning_rate": 3.847328834316818e-05, "loss": 0.0263, "step": 119070 }, { "grad_norm": 0.5152071118354797, "learning_rate": 3.8465243839218414e-05, "loss": 0.0253, "step": 119080 }, { "grad_norm": 0.560500979423523, "learning_rate": 3.845719965062458e-05, "loss": 0.0246, "step": 119090 }, { "grad_norm": 1.130928874015808, "learning_rate": 3.844915577760654e-05, "loss": 0.0461, "step": 119100 }, { "grad_norm": 0.24832728505134583, "learning_rate": 3.844111222038427e-05, "loss": 0.0217, "step": 119110 }, { "grad_norm": 0.3663535714149475, "learning_rate": 3.8433068979177624e-05, "loss": 0.0188, "step": 119120 }, { "grad_norm": 0.5329051613807678, "learning_rate": 3.842502605420654e-05, "loss": 0.0288, "step": 119130 }, { "grad_norm": 0.29155051708221436, "learning_rate": 3.841698344569089e-05, "loss": 0.0321, "step": 119140 }, { "grad_norm": 0.5202053785324097, "learning_rate": 3.840894115385056e-05, "loss": 0.0229, "step": 119150 }, { "grad_norm": 0.4778342545032501, "learning_rate": 3.840089917890544e-05, "loss": 0.0188, "step": 119160 }, { "grad_norm": 0.46475449204444885, "learning_rate": 3.839285752107536e-05, "loss": 0.0226, "step": 119170 }, { "grad_norm": 0.754275918006897, "learning_rate": 3.838481618058021e-05, "loss": 0.0208, "step": 119180 }, { "grad_norm": 0.3483048379421234, "learning_rate": 3.8376775157639804e-05, "loss": 0.0215, "step": 119190 }, { "grad_norm": 0.19966548681259155, "learning_rate": 3.836873445247402e-05, "loss": 0.0222, "step": 119200 }, { "grad_norm": 0.8107498288154602, "learning_rate": 3.836069406530264e-05, "loss": 0.0227, "step": 119210 }, { "grad_norm": 0.66635662317276, "learning_rate": 3.835265399634552e-05, "loss": 0.0156, "step": 119220 }, { "grad_norm": 0.4021442234516144, "learning_rate": 3.834461424582248e-05, "loss": 0.0327, "step": 119230 }, { "grad_norm": 0.8710829019546509, "learning_rate": 3.833657481395329e-05, "loss": 0.039, "step": 119240 }, { "grad_norm": 0.46794652938842773, "learning_rate": 3.832853570095778e-05, "loss": 0.0274, "step": 119250 }, { "grad_norm": 0.49439048767089844, "learning_rate": 3.8320496907055703e-05, "loss": 0.0228, "step": 119260 }, { "grad_norm": 0.4537401497364044, "learning_rate": 3.831245843246687e-05, "loss": 0.0172, "step": 119270 }, { "grad_norm": 0.5323722958564758, "learning_rate": 3.8304420277411015e-05, "loss": 0.0257, "step": 119280 }, { "grad_norm": 0.48599904775619507, "learning_rate": 3.829638244210791e-05, "loss": 0.029, "step": 119290 }, { "grad_norm": 0.5591962933540344, "learning_rate": 3.828834492677734e-05, "loss": 0.0292, "step": 119300 }, { "grad_norm": 0.4363604784011841, "learning_rate": 3.828030773163899e-05, "loss": 0.0352, "step": 119310 }, { "grad_norm": 0.3863033652305603, "learning_rate": 3.827227085691266e-05, "loss": 0.0247, "step": 119320 }, { "grad_norm": 0.4249763488769531, "learning_rate": 3.8264234302818004e-05, "loss": 0.0538, "step": 119330 }, { "grad_norm": 0.30103743076324463, "learning_rate": 3.825619806957478e-05, "loss": 0.0215, "step": 119340 }, { "grad_norm": 0.4995807707309723, "learning_rate": 3.82481621574027e-05, "loss": 0.025, "step": 119350 }, { "grad_norm": 0.3460880219936371, "learning_rate": 3.8240126566521444e-05, "loss": 0.0385, "step": 119360 }, { "grad_norm": 0.8902797102928162, "learning_rate": 3.8232091297150716e-05, "loss": 0.0277, "step": 119370 }, { "grad_norm": 0.26934248208999634, "learning_rate": 3.822405634951019e-05, "loss": 0.021, "step": 119380 }, { "grad_norm": 0.2753485143184662, "learning_rate": 3.821602172381954e-05, "loss": 0.0363, "step": 119390 }, { "grad_norm": 0.7324347496032715, "learning_rate": 3.820798742029842e-05, "loss": 0.0338, "step": 119400 }, { "grad_norm": 0.4350998103618622, "learning_rate": 3.819995343916649e-05, "loss": 0.0216, "step": 119410 }, { "grad_norm": 0.6489830017089844, "learning_rate": 3.819191978064342e-05, "loss": 0.0282, "step": 119420 }, { "grad_norm": 0.6828741431236267, "learning_rate": 3.818388644494881e-05, "loss": 0.018, "step": 119430 }, { "grad_norm": 0.4006561040878296, "learning_rate": 3.817585343230232e-05, "loss": 0.0165, "step": 119440 }, { "grad_norm": 0.5690727233886719, "learning_rate": 3.8167820742923544e-05, "loss": 0.027, "step": 119450 }, { "grad_norm": 0.37336596846580505, "learning_rate": 3.815978837703211e-05, "loss": 0.0245, "step": 119460 }, { "grad_norm": 0.47053262591362, "learning_rate": 3.815175633484761e-05, "loss": 0.0225, "step": 119470 }, { "grad_norm": 0.4579387903213501, "learning_rate": 3.814372461658964e-05, "loss": 0.0272, "step": 119480 }, { "grad_norm": 0.5370524525642395, "learning_rate": 3.81356932224778e-05, "loss": 0.0227, "step": 119490 }, { "grad_norm": 0.45079731941223145, "learning_rate": 3.8127662152731615e-05, "loss": 0.0219, "step": 119500 }, { "grad_norm": 0.3328249156475067, "learning_rate": 3.8119631407570724e-05, "loss": 0.0209, "step": 119510 }, { "grad_norm": 0.4278540015220642, "learning_rate": 3.811160098721462e-05, "loss": 0.0227, "step": 119520 }, { "grad_norm": 0.8134616613388062, "learning_rate": 3.810357089188289e-05, "loss": 0.0426, "step": 119530 }, { "grad_norm": 0.37297022342681885, "learning_rate": 3.8095541121795064e-05, "loss": 0.0259, "step": 119540 }, { "grad_norm": 0.44679272174835205, "learning_rate": 3.808751167717067e-05, "loss": 0.016, "step": 119550 }, { "grad_norm": 0.7335307598114014, "learning_rate": 3.8079482558229235e-05, "loss": 0.033, "step": 119560 }, { "grad_norm": 0.39415016770362854, "learning_rate": 3.807145376519026e-05, "loss": 0.0268, "step": 119570 }, { "grad_norm": 0.4296659231185913, "learning_rate": 3.806342529827327e-05, "loss": 0.0246, "step": 119580 }, { "grad_norm": 0.3991130590438843, "learning_rate": 3.8055397157697735e-05, "loss": 0.0185, "step": 119590 }, { "grad_norm": 1.0171154737472534, "learning_rate": 3.804736934368315e-05, "loss": 0.0256, "step": 119600 }, { "grad_norm": 0.3640376031398773, "learning_rate": 3.8039341856449025e-05, "loss": 0.0225, "step": 119610 }, { "grad_norm": 0.39921408891677856, "learning_rate": 3.803131469621477e-05, "loss": 0.0236, "step": 119620 }, { "grad_norm": 0.3336980640888214, "learning_rate": 3.802328786319988e-05, "loss": 0.025, "step": 119630 }, { "grad_norm": 0.543971836566925, "learning_rate": 3.801526135762381e-05, "loss": 0.0278, "step": 119640 }, { "grad_norm": 0.21978966891765594, "learning_rate": 3.8007235179705995e-05, "loss": 0.0292, "step": 119650 }, { "grad_norm": 0.5609816312789917, "learning_rate": 3.799920932966586e-05, "loss": 0.0212, "step": 119660 }, { "grad_norm": 0.4471111297607422, "learning_rate": 3.799118380772282e-05, "loss": 0.0251, "step": 119670 }, { "grad_norm": 0.37663841247558594, "learning_rate": 3.798315861409632e-05, "loss": 0.0291, "step": 119680 }, { "grad_norm": 0.44382280111312866, "learning_rate": 3.797513374900575e-05, "loss": 0.0215, "step": 119690 }, { "grad_norm": 0.4805351197719574, "learning_rate": 3.796710921267051e-05, "loss": 0.0268, "step": 119700 }, { "grad_norm": 0.6397907733917236, "learning_rate": 3.795908500530996e-05, "loss": 0.0208, "step": 119710 }, { "grad_norm": 0.7204588055610657, "learning_rate": 3.7951061127143504e-05, "loss": 0.0215, "step": 119720 }, { "grad_norm": 0.29476359486579895, "learning_rate": 3.794303757839053e-05, "loss": 0.0185, "step": 119730 }, { "grad_norm": 0.534108579158783, "learning_rate": 3.793501435927037e-05, "loss": 0.025, "step": 119740 }, { "grad_norm": 0.3982536494731903, "learning_rate": 3.7926991470002394e-05, "loss": 0.0244, "step": 119750 }, { "grad_norm": 0.3388439416885376, "learning_rate": 3.791896891080592e-05, "loss": 0.0179, "step": 119760 }, { "grad_norm": 0.4890243709087372, "learning_rate": 3.791094668190032e-05, "loss": 0.0251, "step": 119770 }, { "grad_norm": 0.259184330701828, "learning_rate": 3.790292478350487e-05, "loss": 0.0201, "step": 119780 }, { "grad_norm": 0.31984198093414307, "learning_rate": 3.789490321583891e-05, "loss": 0.0186, "step": 119790 }, { "grad_norm": 0.4403105676174164, "learning_rate": 3.788688197912178e-05, "loss": 0.0284, "step": 119800 }, { "grad_norm": 0.6990579962730408, "learning_rate": 3.7878861073572713e-05, "loss": 0.0266, "step": 119810 }, { "grad_norm": 0.5114367008209229, "learning_rate": 3.7870840499411066e-05, "loss": 0.0246, "step": 119820 }, { "grad_norm": 0.6864703893661499, "learning_rate": 3.786282025685605e-05, "loss": 0.0222, "step": 119830 }, { "grad_norm": 0.3442935347557068, "learning_rate": 3.7854800346126985e-05, "loss": 0.0217, "step": 119840 }, { "grad_norm": 0.6532548069953918, "learning_rate": 3.78467807674431e-05, "loss": 0.0372, "step": 119850 }, { "grad_norm": 0.23457446694374084, "learning_rate": 3.7838761521023675e-05, "loss": 0.0257, "step": 119860 }, { "grad_norm": 0.49184003472328186, "learning_rate": 3.7830742607087946e-05, "loss": 0.0279, "step": 119870 }, { "grad_norm": 0.5477392673492432, "learning_rate": 3.782272402585513e-05, "loss": 0.0211, "step": 119880 }, { "grad_norm": 0.4025501310825348, "learning_rate": 3.781470577754448e-05, "loss": 0.0368, "step": 119890 }, { "grad_norm": 1.1953402757644653, "learning_rate": 3.780668786237518e-05, "loss": 0.0249, "step": 119900 }, { "grad_norm": 0.42632514238357544, "learning_rate": 3.779867028056645e-05, "loss": 0.0222, "step": 119910 }, { "grad_norm": 0.36420702934265137, "learning_rate": 3.779065303233751e-05, "loss": 0.0223, "step": 119920 }, { "grad_norm": 0.25607430934906006, "learning_rate": 3.778263611790751e-05, "loss": 0.0241, "step": 119930 }, { "grad_norm": 0.4601842164993286, "learning_rate": 3.777461953749567e-05, "loss": 0.0323, "step": 119940 }, { "grad_norm": 0.2605118453502655, "learning_rate": 3.7766603291321126e-05, "loss": 0.0265, "step": 119950 }, { "grad_norm": 0.2552122175693512, "learning_rate": 3.775858737960306e-05, "loss": 0.0251, "step": 119960 }, { "grad_norm": 0.4688560664653778, "learning_rate": 3.7750571802560605e-05, "loss": 0.0297, "step": 119970 }, { "grad_norm": 0.5551551580429077, "learning_rate": 3.7742556560412915e-05, "loss": 0.0232, "step": 119980 }, { "grad_norm": 0.5980046987533569, "learning_rate": 3.773454165337914e-05, "loss": 0.0192, "step": 119990 }, { "grad_norm": 0.5253344178199768, "learning_rate": 3.772652708167836e-05, "loss": 0.0213, "step": 120000 }, { "grad_norm": 0.3355696499347687, "learning_rate": 3.7718512845529756e-05, "loss": 0.0237, "step": 120010 }, { "grad_norm": 0.39060384035110474, "learning_rate": 3.771049894515236e-05, "loss": 0.0237, "step": 120020 }, { "grad_norm": 0.4085048735141754, "learning_rate": 3.7702485380765344e-05, "loss": 0.0226, "step": 120030 }, { "grad_norm": 0.27569910883903503, "learning_rate": 3.769447215258772e-05, "loss": 0.0239, "step": 120040 }, { "grad_norm": 0.3515770137310028, "learning_rate": 3.768645926083862e-05, "loss": 0.0204, "step": 120050 }, { "grad_norm": 0.37007737159729004, "learning_rate": 3.7678446705737105e-05, "loss": 0.0267, "step": 120060 }, { "grad_norm": 0.5307497978210449, "learning_rate": 3.7670434487502214e-05, "loss": 0.0217, "step": 120070 }, { "grad_norm": 0.2479773908853531, "learning_rate": 3.766242260635303e-05, "loss": 0.0287, "step": 120080 }, { "grad_norm": 0.6904592514038086, "learning_rate": 3.7654411062508565e-05, "loss": 0.0278, "step": 120090 }, { "grad_norm": 0.5187686681747437, "learning_rate": 3.7646399856187845e-05, "loss": 0.0189, "step": 120100 }, { "grad_norm": 0.5941237211227417, "learning_rate": 3.763838898760995e-05, "loss": 0.0186, "step": 120110 }, { "grad_norm": 0.31269535422325134, "learning_rate": 3.763037845699382e-05, "loss": 0.0274, "step": 120120 }, { "grad_norm": 0.8310630321502686, "learning_rate": 3.762236826455853e-05, "loss": 0.022, "step": 120130 }, { "grad_norm": 0.42246419191360474, "learning_rate": 3.761435841052302e-05, "loss": 0.0277, "step": 120140 }, { "grad_norm": 0.35057657957077026, "learning_rate": 3.760634889510631e-05, "loss": 0.0267, "step": 120150 }, { "grad_norm": 0.8199125528335571, "learning_rate": 3.759833971852735e-05, "loss": 0.0246, "step": 120160 }, { "grad_norm": 0.24208413064479828, "learning_rate": 3.759033088100513e-05, "loss": 0.0193, "step": 120170 }, { "grad_norm": 0.5755939483642578, "learning_rate": 3.758232238275862e-05, "loss": 0.0229, "step": 120180 }, { "grad_norm": 0.47916504740715027, "learning_rate": 3.757431422400672e-05, "loss": 0.0205, "step": 120190 }, { "grad_norm": 0.4243760406970978, "learning_rate": 3.756630640496843e-05, "loss": 0.0429, "step": 120200 }, { "grad_norm": 0.3817509114742279, "learning_rate": 3.755829892586262e-05, "loss": 0.0217, "step": 120210 }, { "grad_norm": 0.604267418384552, "learning_rate": 3.755029178690828e-05, "loss": 0.0296, "step": 120220 }, { "grad_norm": 0.44604164361953735, "learning_rate": 3.754228498832425e-05, "loss": 0.0267, "step": 120230 }, { "grad_norm": 0.48428788781166077, "learning_rate": 3.753427853032948e-05, "loss": 0.0202, "step": 120240 }, { "grad_norm": 0.6118766069412231, "learning_rate": 3.7526272413142865e-05, "loss": 0.0194, "step": 120250 }, { "grad_norm": 0.3625860810279846, "learning_rate": 3.751826663698326e-05, "loss": 0.0285, "step": 120260 }, { "grad_norm": 0.7133293747901917, "learning_rate": 3.7510261202069575e-05, "loss": 0.0304, "step": 120270 }, { "grad_norm": 0.325564444065094, "learning_rate": 3.750225610862064e-05, "loss": 0.0323, "step": 120280 }, { "grad_norm": 0.678210973739624, "learning_rate": 3.7494251356855317e-05, "loss": 0.0302, "step": 120290 }, { "grad_norm": 0.7220448851585388, "learning_rate": 3.74862469469925e-05, "loss": 0.0247, "step": 120300 }, { "grad_norm": 0.2678223252296448, "learning_rate": 3.747824287925095e-05, "loss": 0.0367, "step": 120310 }, { "grad_norm": 0.558983325958252, "learning_rate": 3.7470239153849575e-05, "loss": 0.0246, "step": 120320 }, { "grad_norm": 0.7507902383804321, "learning_rate": 3.7462235771007116e-05, "loss": 0.0276, "step": 120330 }, { "grad_norm": 1.3701492547988892, "learning_rate": 3.745423273094245e-05, "loss": 0.026, "step": 120340 }, { "grad_norm": 0.3428945541381836, "learning_rate": 3.744623003387433e-05, "loss": 0.0321, "step": 120350 }, { "grad_norm": 0.4435540735721588, "learning_rate": 3.743822768002157e-05, "loss": 0.0365, "step": 120360 }, { "grad_norm": 0.44276005029678345, "learning_rate": 3.743022566960295e-05, "loss": 0.0261, "step": 120370 }, { "grad_norm": 0.34022659063339233, "learning_rate": 3.7422224002837236e-05, "loss": 0.0249, "step": 120380 }, { "grad_norm": 0.2545178532600403, "learning_rate": 3.74142226799432e-05, "loss": 0.0186, "step": 120390 }, { "grad_norm": 0.27910515666007996, "learning_rate": 3.740622170113958e-05, "loss": 0.0175, "step": 120400 }, { "grad_norm": 0.4291003346443176, "learning_rate": 3.739822106664513e-05, "loss": 0.0244, "step": 120410 }, { "grad_norm": 0.7851347923278809, "learning_rate": 3.739022077667856e-05, "loss": 0.0195, "step": 120420 }, { "grad_norm": 0.5378105640411377, "learning_rate": 3.738222083145864e-05, "loss": 0.02, "step": 120430 }, { "grad_norm": 0.385409951210022, "learning_rate": 3.737422123120406e-05, "loss": 0.0299, "step": 120440 }, { "grad_norm": 0.4172860383987427, "learning_rate": 3.736622197613352e-05, "loss": 0.0244, "step": 120450 }, { "grad_norm": 0.534366250038147, "learning_rate": 3.7358223066465734e-05, "loss": 0.0393, "step": 120460 }, { "grad_norm": 0.4069274067878723, "learning_rate": 3.735022450241938e-05, "loss": 0.0235, "step": 120470 }, { "grad_norm": 0.27960219979286194, "learning_rate": 3.734222628421313e-05, "loss": 0.0237, "step": 120480 }, { "grad_norm": 0.5850474834442139, "learning_rate": 3.733422841206566e-05, "loss": 0.0217, "step": 120490 }, { "grad_norm": 0.8482068777084351, "learning_rate": 3.732623088619562e-05, "loss": 0.0283, "step": 120500 }, { "grad_norm": 0.333793044090271, "learning_rate": 3.7318233706821694e-05, "loss": 0.0217, "step": 120510 }, { "grad_norm": 0.40291082859039307, "learning_rate": 3.731023687416245e-05, "loss": 0.0252, "step": 120520 }, { "grad_norm": 0.43015220761299133, "learning_rate": 3.730224038843661e-05, "loss": 0.0265, "step": 120530 }, { "grad_norm": 0.5811775326728821, "learning_rate": 3.729424424986271e-05, "loss": 0.0223, "step": 120540 }, { "grad_norm": 0.25087618827819824, "learning_rate": 3.728624845865941e-05, "loss": 0.0209, "step": 120550 }, { "grad_norm": 0.39409521222114563, "learning_rate": 3.727825301504531e-05, "loss": 0.0259, "step": 120560 }, { "grad_norm": 0.26105374097824097, "learning_rate": 3.727025791923898e-05, "loss": 0.0194, "step": 120570 }, { "grad_norm": 0.2304765284061432, "learning_rate": 3.7262263171459034e-05, "loss": 0.0244, "step": 120580 }, { "grad_norm": 0.4666646420955658, "learning_rate": 3.7254268771924014e-05, "loss": 0.0181, "step": 120590 }, { "grad_norm": 0.39095866680145264, "learning_rate": 3.72462747208525e-05, "loss": 0.0204, "step": 120600 }, { "grad_norm": 0.29102808237075806, "learning_rate": 3.723828101846304e-05, "loss": 0.0214, "step": 120610 }, { "grad_norm": 0.6381334662437439, "learning_rate": 3.723028766497417e-05, "loss": 0.0241, "step": 120620 }, { "grad_norm": 1.3112214803695679, "learning_rate": 3.7222294660604466e-05, "loss": 0.0281, "step": 120630 }, { "grad_norm": 0.4028396010398865, "learning_rate": 3.721430200557241e-05, "loss": 0.0385, "step": 120640 }, { "grad_norm": 0.4284485876560211, "learning_rate": 3.720630970009655e-05, "loss": 0.0163, "step": 120650 }, { "grad_norm": 0.5025331377983093, "learning_rate": 3.719831774439536e-05, "loss": 0.024, "step": 120660 }, { "grad_norm": 0.5279074907302856, "learning_rate": 3.719032613868736e-05, "loss": 0.0348, "step": 120670 }, { "grad_norm": 0.3291596472263336, "learning_rate": 3.7182334883191026e-05, "loss": 0.0219, "step": 120680 }, { "grad_norm": 0.47686830163002014, "learning_rate": 3.717434397812484e-05, "loss": 0.021, "step": 120690 }, { "grad_norm": 0.415764719247818, "learning_rate": 3.716635342370728e-05, "loss": 0.0305, "step": 120700 }, { "grad_norm": 0.20474857091903687, "learning_rate": 3.715836322015678e-05, "loss": 0.0199, "step": 120710 }, { "grad_norm": 0.31937241554260254, "learning_rate": 3.7150373367691834e-05, "loss": 0.0204, "step": 120720 }, { "grad_norm": 0.25129517912864685, "learning_rate": 3.714238386653082e-05, "loss": 0.0172, "step": 120730 }, { "grad_norm": 0.36838358640670776, "learning_rate": 3.7134394716892215e-05, "loss": 0.0258, "step": 120740 }, { "grad_norm": 1.1254969835281372, "learning_rate": 3.7126405918994425e-05, "loss": 0.0293, "step": 120750 }, { "grad_norm": 0.6371150612831116, "learning_rate": 3.711841747305586e-05, "loss": 0.0286, "step": 120760 }, { "grad_norm": 0.39336875081062317, "learning_rate": 3.711042937929493e-05, "loss": 0.0285, "step": 120770 }, { "grad_norm": 3.430398941040039, "learning_rate": 3.710244163793001e-05, "loss": 0.0242, "step": 120780 }, { "grad_norm": 0.5049809813499451, "learning_rate": 3.70944542491795e-05, "loss": 0.0242, "step": 120790 }, { "grad_norm": 0.4285701513290405, "learning_rate": 3.708646721326175e-05, "loss": 0.025, "step": 120800 }, { "grad_norm": 0.7970572113990784, "learning_rate": 3.7078480530395124e-05, "loss": 0.0274, "step": 120810 }, { "grad_norm": 0.24614952504634857, "learning_rate": 3.7070494200798024e-05, "loss": 0.0305, "step": 120820 }, { "grad_norm": 0.6129167079925537, "learning_rate": 3.706250822468871e-05, "loss": 0.0239, "step": 120830 }, { "grad_norm": 1.5386033058166504, "learning_rate": 3.705452260228559e-05, "loss": 0.0306, "step": 120840 }, { "grad_norm": 0.710759162902832, "learning_rate": 3.704653733380695e-05, "loss": 0.0372, "step": 120850 }, { "grad_norm": 0.4028733968734741, "learning_rate": 3.703855241947112e-05, "loss": 0.0396, "step": 120860 }, { "grad_norm": 0.3354036211967468, "learning_rate": 3.703056785949639e-05, "loss": 0.0294, "step": 120870 }, { "grad_norm": 0.7004533410072327, "learning_rate": 3.702258365410106e-05, "loss": 0.0257, "step": 120880 }, { "grad_norm": 0.461077481508255, "learning_rate": 3.7014599803503436e-05, "loss": 0.0391, "step": 120890 }, { "grad_norm": 0.6890461444854736, "learning_rate": 3.700661630792174e-05, "loss": 0.0292, "step": 120900 }, { "grad_norm": 0.8288335204124451, "learning_rate": 3.6998633167574314e-05, "loss": 0.0494, "step": 120910 }, { "grad_norm": 0.44702500104904175, "learning_rate": 3.6990650382679336e-05, "loss": 0.0272, "step": 120920 }, { "grad_norm": 0.4471547305583954, "learning_rate": 3.6982667953455106e-05, "loss": 0.0292, "step": 120930 }, { "grad_norm": 0.432772696018219, "learning_rate": 3.697468588011985e-05, "loss": 0.0217, "step": 120940 }, { "grad_norm": 0.4322251081466675, "learning_rate": 3.696670416289177e-05, "loss": 0.0187, "step": 120950 }, { "grad_norm": 0.2860827147960663, "learning_rate": 3.6958722801989134e-05, "loss": 0.0168, "step": 120960 }, { "grad_norm": 0.6929062604904175, "learning_rate": 3.695074179763009e-05, "loss": 0.0272, "step": 120970 }, { "grad_norm": 1.086115837097168, "learning_rate": 3.694276115003288e-05, "loss": 0.0453, "step": 120980 }, { "grad_norm": 0.36674603819847107, "learning_rate": 3.693478085941565e-05, "loss": 0.0201, "step": 120990 }, { "grad_norm": 0.5016570091247559, "learning_rate": 3.6926800925996605e-05, "loss": 0.0216, "step": 121000 }, { "grad_norm": 0.6162925362586975, "learning_rate": 3.6918821349993946e-05, "loss": 0.0228, "step": 121010 }, { "grad_norm": 0.3250351846218109, "learning_rate": 3.691084213162576e-05, "loss": 0.0229, "step": 121020 }, { "grad_norm": 0.41897064447402954, "learning_rate": 3.690286327111026e-05, "loss": 0.0225, "step": 121030 }, { "grad_norm": 0.4386850595474243, "learning_rate": 3.689488476866554e-05, "loss": 0.0271, "step": 121040 }, { "grad_norm": 0.5989627838134766, "learning_rate": 3.688690662450975e-05, "loss": 0.0212, "step": 121050 }, { "grad_norm": 0.6648712754249573, "learning_rate": 3.687892883886102e-05, "loss": 0.0306, "step": 121060 }, { "grad_norm": 0.37256404757499695, "learning_rate": 3.6870951411937424e-05, "loss": 0.0274, "step": 121070 }, { "grad_norm": 0.4795117974281311, "learning_rate": 3.68629743439571e-05, "loss": 0.0291, "step": 121080 }, { "grad_norm": 0.3944070041179657, "learning_rate": 3.685499763513812e-05, "loss": 0.0287, "step": 121090 }, { "grad_norm": 0.5894351601600647, "learning_rate": 3.684702128569857e-05, "loss": 0.026, "step": 121100 }, { "grad_norm": 0.41680052876472473, "learning_rate": 3.6839045295856486e-05, "loss": 0.0215, "step": 121110 }, { "grad_norm": 0.39609840512275696, "learning_rate": 3.683106966582999e-05, "loss": 0.0271, "step": 121120 }, { "grad_norm": 0.5562645196914673, "learning_rate": 3.68230943958371e-05, "loss": 0.028, "step": 121130 }, { "grad_norm": 0.35912084579467773, "learning_rate": 3.681511948609585e-05, "loss": 0.0313, "step": 121140 }, { "grad_norm": 0.3775586783885956, "learning_rate": 3.680714493682429e-05, "loss": 0.0229, "step": 121150 }, { "grad_norm": 0.3860818147659302, "learning_rate": 3.679917074824043e-05, "loss": 0.026, "step": 121160 }, { "grad_norm": 0.8056314587593079, "learning_rate": 3.6791196920562275e-05, "loss": 0.0233, "step": 121170 }, { "grad_norm": 0.47496461868286133, "learning_rate": 3.678322345400784e-05, "loss": 0.0249, "step": 121180 }, { "grad_norm": 0.36239171028137207, "learning_rate": 3.6775250348795086e-05, "loss": 0.0253, "step": 121190 }, { "grad_norm": 0.98625248670578, "learning_rate": 3.6767277605142055e-05, "loss": 0.0185, "step": 121200 }, { "grad_norm": 0.6025276184082031, "learning_rate": 3.675930522326666e-05, "loss": 0.02, "step": 121210 }, { "grad_norm": 0.3864990174770355, "learning_rate": 3.675133320338692e-05, "loss": 0.0316, "step": 121220 }, { "grad_norm": 0.5924642086029053, "learning_rate": 3.674336154572071e-05, "loss": 0.0176, "step": 121230 }, { "grad_norm": 0.4531230628490448, "learning_rate": 3.673539025048605e-05, "loss": 0.0287, "step": 121240 }, { "grad_norm": 0.3717027008533478, "learning_rate": 3.6727419317900805e-05, "loss": 0.0297, "step": 121250 }, { "grad_norm": 0.29034048318862915, "learning_rate": 3.671944874818294e-05, "loss": 0.0204, "step": 121260 }, { "grad_norm": 0.5726889371871948, "learning_rate": 3.671147854155037e-05, "loss": 0.0278, "step": 121270 }, { "grad_norm": 0.3761485517024994, "learning_rate": 3.670350869822097e-05, "loss": 0.0247, "step": 121280 }, { "grad_norm": 0.3315230906009674, "learning_rate": 3.669553921841266e-05, "loss": 0.0142, "step": 121290 }, { "grad_norm": 0.24125522375106812, "learning_rate": 3.668757010234329e-05, "loss": 0.0256, "step": 121300 }, { "grad_norm": 0.49334853887557983, "learning_rate": 3.667960135023075e-05, "loss": 0.0199, "step": 121310 }, { "grad_norm": 0.30888643860816956, "learning_rate": 3.667163296229293e-05, "loss": 0.0198, "step": 121320 }, { "grad_norm": 0.5457450151443481, "learning_rate": 3.666366493874764e-05, "loss": 0.0357, "step": 121330 }, { "grad_norm": 0.3453553318977356, "learning_rate": 3.665569727981276e-05, "loss": 0.0242, "step": 121340 }, { "grad_norm": 0.39682304859161377, "learning_rate": 3.664772998570608e-05, "loss": 0.0282, "step": 121350 }, { "grad_norm": 0.26042604446411133, "learning_rate": 3.663976305664547e-05, "loss": 0.0317, "step": 121360 }, { "grad_norm": 0.3546326756477356, "learning_rate": 3.66317964928487e-05, "loss": 0.0249, "step": 121370 }, { "grad_norm": 0.568272590637207, "learning_rate": 3.662383029453359e-05, "loss": 0.0288, "step": 121380 }, { "grad_norm": 0.43068552017211914, "learning_rate": 3.661586446191795e-05, "loss": 0.0179, "step": 121390 }, { "grad_norm": 0.41513100266456604, "learning_rate": 3.6607898995219525e-05, "loss": 0.0256, "step": 121400 }, { "grad_norm": 0.6585609316825867, "learning_rate": 3.659993389465614e-05, "loss": 0.0345, "step": 121410 }, { "grad_norm": 0.43716946244239807, "learning_rate": 3.65919691604455e-05, "loss": 0.0221, "step": 121420 }, { "grad_norm": 0.4083345830440521, "learning_rate": 3.6584004792805415e-05, "loss": 0.026, "step": 121430 }, { "grad_norm": 0.572738766670227, "learning_rate": 3.657604079195357e-05, "loss": 0.025, "step": 121440 }, { "grad_norm": 0.3314396142959595, "learning_rate": 3.656807715810773e-05, "loss": 0.019, "step": 121450 }, { "grad_norm": 0.7676694989204407, "learning_rate": 3.6560113891485634e-05, "loss": 0.0256, "step": 121460 }, { "grad_norm": 0.6911763548851013, "learning_rate": 3.6552150992304954e-05, "loss": 0.0268, "step": 121470 }, { "grad_norm": 0.4794078469276428, "learning_rate": 3.654418846078343e-05, "loss": 0.0275, "step": 121480 }, { "grad_norm": 0.4152323305606842, "learning_rate": 3.653622629713873e-05, "loss": 0.029, "step": 121490 }, { "grad_norm": 0.7447405457496643, "learning_rate": 3.652826450158854e-05, "loss": 0.0378, "step": 121500 }, { "grad_norm": 0.3224257528781891, "learning_rate": 3.652030307435056e-05, "loss": 0.0298, "step": 121510 }, { "grad_norm": 0.2447902113199234, "learning_rate": 3.6512342015642396e-05, "loss": 0.0234, "step": 121520 }, { "grad_norm": 0.3680250942707062, "learning_rate": 3.650438132568176e-05, "loss": 0.0191, "step": 121530 }, { "grad_norm": 0.44088125228881836, "learning_rate": 3.6496421004686256e-05, "loss": 0.0283, "step": 121540 }, { "grad_norm": 0.5281039476394653, "learning_rate": 3.648846105287354e-05, "loss": 0.0399, "step": 121550 }, { "grad_norm": 0.527347207069397, "learning_rate": 3.6480501470461215e-05, "loss": 0.0323, "step": 121560 }, { "grad_norm": 0.40025386214256287, "learning_rate": 3.6472542257666896e-05, "loss": 0.029, "step": 121570 }, { "grad_norm": 0.6084333062171936, "learning_rate": 3.64645834147082e-05, "loss": 0.0261, "step": 121580 }, { "grad_norm": 0.29471299052238464, "learning_rate": 3.645662494180271e-05, "loss": 0.023, "step": 121590 }, { "grad_norm": 0.4658629894256592, "learning_rate": 3.6448666839168e-05, "loss": 0.0216, "step": 121600 }, { "grad_norm": 0.4280264973640442, "learning_rate": 3.644070910702164e-05, "loss": 0.0276, "step": 121610 }, { "grad_norm": 0.41846218705177307, "learning_rate": 3.6432751745581216e-05, "loss": 0.0197, "step": 121620 }, { "grad_norm": 0.534117579460144, "learning_rate": 3.642479475506424e-05, "loss": 0.0353, "step": 121630 }, { "grad_norm": 0.45422375202178955, "learning_rate": 3.641683813568829e-05, "loss": 0.0214, "step": 121640 }, { "grad_norm": 0.36543136835098267, "learning_rate": 3.640888188767089e-05, "loss": 0.0171, "step": 121650 }, { "grad_norm": 0.37742629647254944, "learning_rate": 3.640092601122953e-05, "loss": 0.0209, "step": 121660 }, { "grad_norm": 0.45861539244651794, "learning_rate": 3.639297050658177e-05, "loss": 0.0207, "step": 121670 }, { "grad_norm": 0.41593801975250244, "learning_rate": 3.638501537394507e-05, "loss": 0.022, "step": 121680 }, { "grad_norm": 0.3691204786300659, "learning_rate": 3.637706061353692e-05, "loss": 0.0259, "step": 121690 }, { "grad_norm": 0.3559333086013794, "learning_rate": 3.6369106225574856e-05, "loss": 0.0231, "step": 121700 }, { "grad_norm": 0.3752104341983795, "learning_rate": 3.636115221027627e-05, "loss": 0.0219, "step": 121710 }, { "grad_norm": 0.39938127994537354, "learning_rate": 3.6353198567858685e-05, "loss": 0.0248, "step": 121720 }, { "grad_norm": 0.16771508753299713, "learning_rate": 3.63452452985395e-05, "loss": 0.0213, "step": 121730 }, { "grad_norm": 0.48709413409233093, "learning_rate": 3.633729240253619e-05, "loss": 0.0191, "step": 121740 }, { "grad_norm": 0.5499780178070068, "learning_rate": 3.632933988006617e-05, "loss": 0.0202, "step": 121750 }, { "grad_norm": 0.286086767911911, "learning_rate": 3.6321387731346866e-05, "loss": 0.0249, "step": 121760 }, { "grad_norm": 0.30436915159225464, "learning_rate": 3.631343595659569e-05, "loss": 0.019, "step": 121770 }, { "grad_norm": 0.4779207706451416, "learning_rate": 3.630548455603002e-05, "loss": 0.0219, "step": 121780 }, { "grad_norm": 0.44171082973480225, "learning_rate": 3.629753352986727e-05, "loss": 0.0168, "step": 121790 }, { "grad_norm": 0.2577836811542511, "learning_rate": 3.6289582878324794e-05, "loss": 0.0207, "step": 121800 }, { "grad_norm": 1.1571195125579834, "learning_rate": 3.628163260161997e-05, "loss": 0.0267, "step": 121810 }, { "grad_norm": 0.7072025537490845, "learning_rate": 3.6273682699970154e-05, "loss": 0.0349, "step": 121820 }, { "grad_norm": 0.31098997592926025, "learning_rate": 3.62657331735927e-05, "loss": 0.0207, "step": 121830 }, { "grad_norm": 0.3613261580467224, "learning_rate": 3.6257784022704955e-05, "loss": 0.0251, "step": 121840 }, { "grad_norm": 0.339562326669693, "learning_rate": 3.624983524752422e-05, "loss": 0.0329, "step": 121850 }, { "grad_norm": 0.48377883434295654, "learning_rate": 3.624188684826783e-05, "loss": 0.0294, "step": 121860 }, { "grad_norm": 0.7119192481040955, "learning_rate": 3.623393882515308e-05, "loss": 0.0338, "step": 121870 }, { "grad_norm": 0.3009941279888153, "learning_rate": 3.622599117839728e-05, "loss": 0.0194, "step": 121880 }, { "grad_norm": 0.4417182505130768, "learning_rate": 3.621804390821769e-05, "loss": 0.0311, "step": 121890 }, { "grad_norm": 0.44275423884391785, "learning_rate": 3.6210097014831594e-05, "loss": 0.0353, "step": 121900 }, { "grad_norm": 0.3615298569202423, "learning_rate": 3.62021504984563e-05, "loss": 0.026, "step": 121910 }, { "grad_norm": 0.3038904368877411, "learning_rate": 3.6194204359308995e-05, "loss": 0.0193, "step": 121920 }, { "grad_norm": 0.17492716014385223, "learning_rate": 3.618625859760699e-05, "loss": 0.0289, "step": 121930 }, { "grad_norm": 0.40923380851745605, "learning_rate": 3.617831321356745e-05, "loss": 0.0288, "step": 121940 }, { "grad_norm": 1.355301856994629, "learning_rate": 3.6170368207407646e-05, "loss": 0.0452, "step": 121950 }, { "grad_norm": 0.4743514955043793, "learning_rate": 3.616242357934479e-05, "loss": 0.0232, "step": 121960 }, { "grad_norm": 0.520317792892456, "learning_rate": 3.615447932959606e-05, "loss": 0.0173, "step": 121970 }, { "grad_norm": 0.7508777379989624, "learning_rate": 3.614653545837868e-05, "loss": 0.0246, "step": 121980 }, { "grad_norm": 0.20337098836898804, "learning_rate": 3.613859196590981e-05, "loss": 0.016, "step": 121990 }, { "grad_norm": 0.4848058819770813, "learning_rate": 3.6130648852406627e-05, "loss": 0.0281, "step": 122000 }, { "grad_norm": 0.5895212292671204, "learning_rate": 3.612270611808629e-05, "loss": 0.0222, "step": 122010 }, { "grad_norm": 0.2472146898508072, "learning_rate": 3.6114763763165946e-05, "loss": 0.0179, "step": 122020 }, { "grad_norm": 0.13522127270698547, "learning_rate": 3.610682178786276e-05, "loss": 0.0245, "step": 122030 }, { "grad_norm": 0.5336438417434692, "learning_rate": 3.609888019239384e-05, "loss": 0.0212, "step": 122040 }, { "grad_norm": 0.3064413070678711, "learning_rate": 3.609093897697632e-05, "loss": 0.0203, "step": 122050 }, { "grad_norm": 0.2508518397808075, "learning_rate": 3.608299814182729e-05, "loss": 0.0261, "step": 122060 }, { "grad_norm": 0.3431552052497864, "learning_rate": 3.607505768716388e-05, "loss": 0.0267, "step": 122070 }, { "grad_norm": 0.4488224983215332, "learning_rate": 3.606711761320315e-05, "loss": 0.0328, "step": 122080 }, { "grad_norm": 0.32256951928138733, "learning_rate": 3.605917792016219e-05, "loss": 0.0275, "step": 122090 }, { "grad_norm": 1.3040863275527954, "learning_rate": 3.6051238608258075e-05, "loss": 0.0289, "step": 122100 }, { "grad_norm": 0.39418715238571167, "learning_rate": 3.604329967770784e-05, "loss": 0.029, "step": 122110 }, { "grad_norm": 0.8381692171096802, "learning_rate": 3.603536112872858e-05, "loss": 0.0365, "step": 122120 }, { "grad_norm": 0.5559929013252258, "learning_rate": 3.602742296153726e-05, "loss": 0.0255, "step": 122130 }, { "grad_norm": 0.2682167589664459, "learning_rate": 3.601948517635097e-05, "loss": 0.0236, "step": 122140 }, { "grad_norm": 0.4419136643409729, "learning_rate": 3.6011547773386706e-05, "loss": 0.0209, "step": 122150 }, { "grad_norm": 0.4667212665081024, "learning_rate": 3.600361075286146e-05, "loss": 0.0215, "step": 122160 }, { "grad_norm": 0.3357292115688324, "learning_rate": 3.599567411499225e-05, "loss": 0.024, "step": 122170 }, { "grad_norm": 0.4192315638065338, "learning_rate": 3.5987737859996043e-05, "loss": 0.0263, "step": 122180 }, { "grad_norm": 0.5882171988487244, "learning_rate": 3.5979801988089825e-05, "loss": 0.0305, "step": 122190 }, { "grad_norm": 0.5204571485519409, "learning_rate": 3.597186649949054e-05, "loss": 0.0204, "step": 122200 }, { "grad_norm": 0.21345438063144684, "learning_rate": 3.596393139441515e-05, "loss": 0.019, "step": 122210 }, { "grad_norm": 0.36971256136894226, "learning_rate": 3.5955996673080625e-05, "loss": 0.0172, "step": 122220 }, { "grad_norm": 0.3542498052120209, "learning_rate": 3.594806233570385e-05, "loss": 0.028, "step": 122230 }, { "grad_norm": 0.34582844376564026, "learning_rate": 3.594012838250179e-05, "loss": 0.0188, "step": 122240 }, { "grad_norm": 0.5794613361358643, "learning_rate": 3.593219481369134e-05, "loss": 0.0249, "step": 122250 }, { "grad_norm": 0.3222830295562744, "learning_rate": 3.592426162948939e-05, "loss": 0.0216, "step": 122260 }, { "grad_norm": 0.2543100118637085, "learning_rate": 3.591632883011284e-05, "loss": 0.0233, "step": 122270 }, { "grad_norm": 0.3484925925731659, "learning_rate": 3.590839641577857e-05, "loss": 0.0415, "step": 122280 }, { "grad_norm": 0.4973733425140381, "learning_rate": 3.590046438670345e-05, "loss": 0.0221, "step": 122290 }, { "grad_norm": 0.24032782018184662, "learning_rate": 3.589253274310433e-05, "loss": 0.0294, "step": 122300 }, { "grad_norm": 0.630251944065094, "learning_rate": 3.588460148519808e-05, "loss": 0.0339, "step": 122310 }, { "grad_norm": 0.29730224609375, "learning_rate": 3.587667061320149e-05, "loss": 0.0253, "step": 122320 }, { "grad_norm": 0.36031854152679443, "learning_rate": 3.586874012733144e-05, "loss": 0.0286, "step": 122330 }, { "grad_norm": 0.817104160785675, "learning_rate": 3.586081002780474e-05, "loss": 0.0229, "step": 122340 }, { "grad_norm": 0.5807763338088989, "learning_rate": 3.585288031483816e-05, "loss": 0.0204, "step": 122350 }, { "grad_norm": 0.2811470031738281, "learning_rate": 3.584495098864853e-05, "loss": 0.0171, "step": 122360 }, { "grad_norm": 0.5044372081756592, "learning_rate": 3.583702204945262e-05, "loss": 0.0267, "step": 122370 }, { "grad_norm": 0.40934544801712036, "learning_rate": 3.582909349746721e-05, "loss": 0.0256, "step": 122380 }, { "grad_norm": 0.801335334777832, "learning_rate": 3.5821165332909054e-05, "loss": 0.0207, "step": 122390 }, { "grad_norm": 0.36327266693115234, "learning_rate": 3.581323755599489e-05, "loss": 0.0235, "step": 122400 }, { "grad_norm": 0.3149360716342926, "learning_rate": 3.580531016694153e-05, "loss": 0.0205, "step": 122410 }, { "grad_norm": 0.345337450504303, "learning_rate": 3.5797383165965614e-05, "loss": 0.0246, "step": 122420 }, { "grad_norm": 0.32419049739837646, "learning_rate": 3.578945655328394e-05, "loss": 0.0272, "step": 122430 }, { "grad_norm": 7.421003818511963, "learning_rate": 3.578153032911315e-05, "loss": 0.0351, "step": 122440 }, { "grad_norm": 0.8499472737312317, "learning_rate": 3.577360449367001e-05, "loss": 0.0343, "step": 122450 }, { "grad_norm": 0.4775969088077545, "learning_rate": 3.576567904717116e-05, "loss": 0.0306, "step": 122460 }, { "grad_norm": 0.27822959423065186, "learning_rate": 3.575775398983331e-05, "loss": 0.0286, "step": 122470 }, { "grad_norm": 0.5567876696586609, "learning_rate": 3.574982932187311e-05, "loss": 0.0256, "step": 122480 }, { "grad_norm": 0.4193726181983948, "learning_rate": 3.5741905043507215e-05, "loss": 0.0239, "step": 122490 }, { "grad_norm": 0.5860385894775391, "learning_rate": 3.573398115495229e-05, "loss": 0.0256, "step": 122500 }, { "grad_norm": 0.3822520971298218, "learning_rate": 3.572605765642496e-05, "loss": 0.0399, "step": 122510 }, { "grad_norm": 0.32844775915145874, "learning_rate": 3.571813454814183e-05, "loss": 0.021, "step": 122520 }, { "grad_norm": 0.5281503200531006, "learning_rate": 3.571021183031956e-05, "loss": 0.0233, "step": 122530 }, { "grad_norm": 0.4312293827533722, "learning_rate": 3.570228950317473e-05, "loss": 0.0261, "step": 122540 }, { "grad_norm": 0.7413033843040466, "learning_rate": 3.5694367566923934e-05, "loss": 0.0379, "step": 122550 }, { "grad_norm": 0.5784947872161865, "learning_rate": 3.568644602178374e-05, "loss": 0.0353, "step": 122560 }, { "grad_norm": 0.5181698799133301, "learning_rate": 3.5678524867970754e-05, "loss": 0.0298, "step": 122570 }, { "grad_norm": 0.5999718308448792, "learning_rate": 3.56706041057015e-05, "loss": 0.027, "step": 122580 }, { "grad_norm": 0.3506501317024231, "learning_rate": 3.5662683735192547e-05, "loss": 0.0222, "step": 122590 }, { "grad_norm": 2.8163013458251953, "learning_rate": 3.565476375666045e-05, "loss": 0.0231, "step": 122600 }, { "grad_norm": 0.34411099553108215, "learning_rate": 3.56468441703217e-05, "loss": 0.0234, "step": 122610 }, { "grad_norm": 0.202274352312088, "learning_rate": 3.563892497639287e-05, "loss": 0.0187, "step": 122620 }, { "grad_norm": 1.0132982730865479, "learning_rate": 3.5631006175090405e-05, "loss": 0.0229, "step": 122630 }, { "grad_norm": 0.3309754729270935, "learning_rate": 3.562308776663086e-05, "loss": 0.0187, "step": 122640 }, { "grad_norm": 0.37748631834983826, "learning_rate": 3.561516975123067e-05, "loss": 0.0325, "step": 122650 }, { "grad_norm": 0.3198741376399994, "learning_rate": 3.5607252129106336e-05, "loss": 0.0316, "step": 122660 }, { "grad_norm": 0.6440446376800537, "learning_rate": 3.559933490047433e-05, "loss": 0.0214, "step": 122670 }, { "grad_norm": 0.5530405044555664, "learning_rate": 3.55914180655511e-05, "loss": 0.0243, "step": 122680 }, { "grad_norm": 0.8218242526054382, "learning_rate": 3.5583501624553095e-05, "loss": 0.0265, "step": 122690 }, { "grad_norm": 0.637859046459198, "learning_rate": 3.557558557769672e-05, "loss": 0.022, "step": 122700 }, { "grad_norm": 0.618497908115387, "learning_rate": 3.556766992519842e-05, "loss": 0.0178, "step": 122710 }, { "grad_norm": 0.32013779878616333, "learning_rate": 3.5559754667274625e-05, "loss": 0.0341, "step": 122720 }, { "grad_norm": 0.26281172037124634, "learning_rate": 3.555183980414167e-05, "loss": 0.026, "step": 122730 }, { "grad_norm": 0.3703939914703369, "learning_rate": 3.5543925336016024e-05, "loss": 0.0161, "step": 122740 }, { "grad_norm": 0.2030765265226364, "learning_rate": 3.553601126311402e-05, "loss": 0.0249, "step": 122750 }, { "grad_norm": 0.6595079302787781, "learning_rate": 3.552809758565203e-05, "loss": 0.0268, "step": 122760 }, { "grad_norm": 0.4076036810874939, "learning_rate": 3.5520184303846414e-05, "loss": 0.0185, "step": 122770 }, { "grad_norm": 0.5911450982093811, "learning_rate": 3.551227141791352e-05, "loss": 0.0188, "step": 122780 }, { "grad_norm": 0.44196492433547974, "learning_rate": 3.550435892806969e-05, "loss": 0.0252, "step": 122790 }, { "grad_norm": 1.370405673980713, "learning_rate": 3.5496446834531236e-05, "loss": 0.0241, "step": 122800 }, { "grad_norm": 0.3864850103855133, "learning_rate": 3.548853513751448e-05, "loss": 0.022, "step": 122810 }, { "grad_norm": 0.4048444628715515, "learning_rate": 3.548062383723571e-05, "loss": 0.0268, "step": 122820 }, { "grad_norm": 0.6475307941436768, "learning_rate": 3.547271293391124e-05, "loss": 0.0226, "step": 122830 }, { "grad_norm": 0.49410662055015564, "learning_rate": 3.546480242775733e-05, "loss": 0.027, "step": 122840 }, { "grad_norm": 0.47262707352638245, "learning_rate": 3.5456892318990264e-05, "loss": 0.0268, "step": 122850 }, { "grad_norm": 0.2982935309410095, "learning_rate": 3.544898260782631e-05, "loss": 0.0173, "step": 122860 }, { "grad_norm": 0.5028015375137329, "learning_rate": 3.54410732944817e-05, "loss": 0.0342, "step": 122870 }, { "grad_norm": 0.34123045206069946, "learning_rate": 3.543316437917268e-05, "loss": 0.0182, "step": 122880 }, { "grad_norm": 0.22391515970230103, "learning_rate": 3.5425255862115465e-05, "loss": 0.0223, "step": 122890 }, { "grad_norm": 0.3856329023838043, "learning_rate": 3.541734774352627e-05, "loss": 0.0178, "step": 122900 }, { "grad_norm": 0.4127732813358307, "learning_rate": 3.540944002362134e-05, "loss": 0.0194, "step": 122910 }, { "grad_norm": 1.4390279054641724, "learning_rate": 3.54015327026168e-05, "loss": 0.0185, "step": 122920 }, { "grad_norm": 0.8596125245094299, "learning_rate": 3.539362578072891e-05, "loss": 0.0292, "step": 122930 }, { "grad_norm": 0.37423980236053467, "learning_rate": 3.538571925817378e-05, "loss": 0.0217, "step": 122940 }, { "grad_norm": 0.4116712808609009, "learning_rate": 3.53778131351676e-05, "loss": 0.0231, "step": 122950 }, { "grad_norm": 0.7183703184127808, "learning_rate": 3.5369907411926516e-05, "loss": 0.0305, "step": 122960 }, { "grad_norm": 0.15141235291957855, "learning_rate": 3.536200208866666e-05, "loss": 0.0203, "step": 122970 }, { "grad_norm": 0.44895097613334656, "learning_rate": 3.535409716560418e-05, "loss": 0.0191, "step": 122980 }, { "grad_norm": 0.5339488387107849, "learning_rate": 3.534619264295518e-05, "loss": 0.0306, "step": 122990 }, { "grad_norm": 0.6893230080604553, "learning_rate": 3.5338288520935766e-05, "loss": 0.0283, "step": 123000 }, { "grad_norm": 0.4026329815387726, "learning_rate": 3.533038479976203e-05, "loss": 0.0274, "step": 123010 }, { "grad_norm": 0.6350938677787781, "learning_rate": 3.532248147965007e-05, "loss": 0.0375, "step": 123020 }, { "grad_norm": 0.4783245623111725, "learning_rate": 3.5314578560815936e-05, "loss": 0.0213, "step": 123030 }, { "grad_norm": 0.6234650015830994, "learning_rate": 3.5306676043475714e-05, "loss": 0.0223, "step": 123040 }, { "grad_norm": 0.5317831635475159, "learning_rate": 3.529877392784546e-05, "loss": 0.0195, "step": 123050 }, { "grad_norm": 0.3465333878993988, "learning_rate": 3.52908722141412e-05, "loss": 0.0228, "step": 123060 }, { "grad_norm": 1.3798415660858154, "learning_rate": 3.528297090257897e-05, "loss": 0.0249, "step": 123070 }, { "grad_norm": 0.3927284777164459, "learning_rate": 3.527506999337479e-05, "loss": 0.0246, "step": 123080 }, { "grad_norm": 0.45995041728019714, "learning_rate": 3.526716948674465e-05, "loss": 0.027, "step": 123090 }, { "grad_norm": 0.29313933849334717, "learning_rate": 3.525926938290457e-05, "loss": 0.0304, "step": 123100 }, { "grad_norm": 0.4407462775707245, "learning_rate": 3.525136968207051e-05, "loss": 0.0448, "step": 123110 }, { "grad_norm": 0.47382721304893494, "learning_rate": 3.52434703844585e-05, "loss": 0.024, "step": 123120 }, { "grad_norm": 0.32173746824264526, "learning_rate": 3.523557149028443e-05, "loss": 0.0255, "step": 123130 }, { "grad_norm": 0.4819570779800415, "learning_rate": 3.5227672999764316e-05, "loss": 0.0272, "step": 123140 }, { "grad_norm": 0.7581573724746704, "learning_rate": 3.5219774913114046e-05, "loss": 0.0272, "step": 123150 }, { "grad_norm": 0.6635355353355408, "learning_rate": 3.5211877230549586e-05, "loss": 0.0222, "step": 123160 }, { "grad_norm": 0.5460001826286316, "learning_rate": 3.520397995228686e-05, "loss": 0.0251, "step": 123170 }, { "grad_norm": 0.3430164158344269, "learning_rate": 3.5196083078541754e-05, "loss": 0.0352, "step": 123180 }, { "grad_norm": 0.46951138973236084, "learning_rate": 3.5188186609530186e-05, "loss": 0.0351, "step": 123190 }, { "grad_norm": 0.9457117319107056, "learning_rate": 3.5180290545468014e-05, "loss": 0.0383, "step": 123200 }, { "grad_norm": 0.4020296633243561, "learning_rate": 3.517239488657115e-05, "loss": 0.0328, "step": 123210 }, { "grad_norm": 0.40516817569732666, "learning_rate": 3.516449963305541e-05, "loss": 0.0251, "step": 123220 }, { "grad_norm": 0.33003711700439453, "learning_rate": 3.5156604785136685e-05, "loss": 0.0212, "step": 123230 }, { "grad_norm": 0.265359103679657, "learning_rate": 3.5148710343030825e-05, "loss": 0.0302, "step": 123240 }, { "grad_norm": 0.5649838447570801, "learning_rate": 3.514081630695364e-05, "loss": 0.0267, "step": 123250 }, { "grad_norm": 1.2642849683761597, "learning_rate": 3.513292267712096e-05, "loss": 0.0217, "step": 123260 }, { "grad_norm": 0.24091069400310516, "learning_rate": 3.5125029453748576e-05, "loss": 0.0208, "step": 123270 }, { "grad_norm": 0.35114389657974243, "learning_rate": 3.511713663705231e-05, "loss": 0.0183, "step": 123280 }, { "grad_norm": 0.790515661239624, "learning_rate": 3.510924422724793e-05, "loss": 0.0339, "step": 123290 }, { "grad_norm": 0.867194414138794, "learning_rate": 3.510135222455121e-05, "loss": 0.0331, "step": 123300 }, { "grad_norm": 0.41326266527175903, "learning_rate": 3.5093460629177946e-05, "loss": 0.0413, "step": 123310 }, { "grad_norm": 0.4566009044647217, "learning_rate": 3.508556944134384e-05, "loss": 0.0277, "step": 123320 }, { "grad_norm": 0.2925874590873718, "learning_rate": 3.507767866126469e-05, "loss": 0.0258, "step": 123330 }, { "grad_norm": 0.46134206652641296, "learning_rate": 3.506978828915617e-05, "loss": 0.0313, "step": 123340 }, { "grad_norm": 0.5188058614730835, "learning_rate": 3.5061898325234046e-05, "loss": 0.0185, "step": 123350 }, { "grad_norm": 0.47965461015701294, "learning_rate": 3.505400876971402e-05, "loss": 0.0233, "step": 123360 }, { "grad_norm": 0.40040239691734314, "learning_rate": 3.5046119622811754e-05, "loss": 0.0189, "step": 123370 }, { "grad_norm": 0.33323079347610474, "learning_rate": 3.503823088474299e-05, "loss": 0.0206, "step": 123380 }, { "grad_norm": 0.3590206503868103, "learning_rate": 3.503034255572335e-05, "loss": 0.0254, "step": 123390 }, { "grad_norm": 0.2561171054840088, "learning_rate": 3.502245463596854e-05, "loss": 0.018, "step": 123400 }, { "grad_norm": 2.437694549560547, "learning_rate": 3.501456712569418e-05, "loss": 0.0216, "step": 123410 }, { "grad_norm": 0.2782474756240845, "learning_rate": 3.5006680025115906e-05, "loss": 0.0281, "step": 123420 }, { "grad_norm": 0.49243390560150146, "learning_rate": 3.499879333444941e-05, "loss": 0.0237, "step": 123430 }, { "grad_norm": 0.5429016947746277, "learning_rate": 3.499090705391023e-05, "loss": 0.022, "step": 123440 }, { "grad_norm": 0.7114633917808533, "learning_rate": 3.498302118371403e-05, "loss": 0.0205, "step": 123450 }, { "grad_norm": 0.383973628282547, "learning_rate": 3.497513572407639e-05, "loss": 0.0232, "step": 123460 }, { "grad_norm": 0.32748010754585266, "learning_rate": 3.49672506752129e-05, "loss": 0.0262, "step": 123470 }, { "grad_norm": 1.5029487609863281, "learning_rate": 3.495936603733911e-05, "loss": 0.0235, "step": 123480 }, { "grad_norm": 0.7899563312530518, "learning_rate": 3.4951481810670616e-05, "loss": 0.0215, "step": 123490 }, { "grad_norm": 0.7132924795150757, "learning_rate": 3.494359799542296e-05, "loss": 0.0232, "step": 123500 }, { "grad_norm": 0.6328359246253967, "learning_rate": 3.493571459181167e-05, "loss": 0.0206, "step": 123510 }, { "grad_norm": 0.4484531581401825, "learning_rate": 3.492783160005229e-05, "loss": 0.0308, "step": 123520 }, { "grad_norm": 0.6133346557617188, "learning_rate": 3.491994902036031e-05, "loss": 0.0429, "step": 123530 }, { "grad_norm": 0.27281254529953003, "learning_rate": 3.4912066852951276e-05, "loss": 0.0167, "step": 123540 }, { "grad_norm": 0.6929419040679932, "learning_rate": 3.490418509804068e-05, "loss": 0.037, "step": 123550 }, { "grad_norm": 0.379155695438385, "learning_rate": 3.489630375584397e-05, "loss": 0.031, "step": 123560 }, { "grad_norm": 0.4146222770214081, "learning_rate": 3.488842282657665e-05, "loss": 0.0182, "step": 123570 }, { "grad_norm": 0.3673659861087799, "learning_rate": 3.4880542310454164e-05, "loss": 0.0239, "step": 123580 }, { "grad_norm": 0.6439489722251892, "learning_rate": 3.487266220769199e-05, "loss": 0.0421, "step": 123590 }, { "grad_norm": 0.41918209195137024, "learning_rate": 3.4864782518505536e-05, "loss": 0.0173, "step": 123600 }, { "grad_norm": 0.4814837872982025, "learning_rate": 3.485690324311023e-05, "loss": 0.0338, "step": 123610 }, { "grad_norm": 0.33495384454727173, "learning_rate": 3.484902438172153e-05, "loss": 0.0457, "step": 123620 }, { "grad_norm": 0.41078031063079834, "learning_rate": 3.484114593455478e-05, "loss": 0.0202, "step": 123630 }, { "grad_norm": 0.5362094640731812, "learning_rate": 3.483326790182544e-05, "loss": 0.024, "step": 123640 }, { "grad_norm": 0.4107537865638733, "learning_rate": 3.482539028374882e-05, "loss": 0.0242, "step": 123650 }, { "grad_norm": 1.1020046472549438, "learning_rate": 3.481751308054036e-05, "loss": 0.0239, "step": 123660 }, { "grad_norm": 0.3715933561325073, "learning_rate": 3.480963629241537e-05, "loss": 0.0183, "step": 123670 }, { "grad_norm": 1.5370755195617676, "learning_rate": 3.480175991958922e-05, "loss": 0.0261, "step": 123680 }, { "grad_norm": 0.29932767152786255, "learning_rate": 3.479388396227725e-05, "loss": 0.0259, "step": 123690 }, { "grad_norm": 0.7962906956672668, "learning_rate": 3.4786008420694774e-05, "loss": 0.0244, "step": 123700 }, { "grad_norm": 0.5829337239265442, "learning_rate": 3.477813329505711e-05, "loss": 0.0256, "step": 123710 }, { "grad_norm": 0.517777681350708, "learning_rate": 3.477025858557957e-05, "loss": 0.0406, "step": 123720 }, { "grad_norm": 0.5087254047393799, "learning_rate": 3.476238429247742e-05, "loss": 0.0233, "step": 123730 }, { "grad_norm": 0.3619499206542969, "learning_rate": 3.475451041596598e-05, "loss": 0.0291, "step": 123740 }, { "grad_norm": 0.33060741424560547, "learning_rate": 3.474663695626049e-05, "loss": 0.0155, "step": 123750 }, { "grad_norm": 0.19090761244297028, "learning_rate": 3.473876391357623e-05, "loss": 0.0237, "step": 123760 }, { "grad_norm": 0.565621018409729, "learning_rate": 3.473089128812841e-05, "loss": 0.0274, "step": 123770 }, { "grad_norm": 0.3611014783382416, "learning_rate": 3.47230190801323e-05, "loss": 0.021, "step": 123780 }, { "grad_norm": 0.4392707049846649, "learning_rate": 3.4715147289803105e-05, "loss": 0.0205, "step": 123790 }, { "grad_norm": 0.37218329310417175, "learning_rate": 3.470727591735604e-05, "loss": 0.0243, "step": 123800 }, { "grad_norm": 0.46072888374328613, "learning_rate": 3.469940496300631e-05, "loss": 0.0186, "step": 123810 }, { "grad_norm": 0.4446244239807129, "learning_rate": 3.469153442696908e-05, "loss": 0.0201, "step": 123820 }, { "grad_norm": 0.3931237459182739, "learning_rate": 3.468366430945959e-05, "loss": 0.0196, "step": 123830 }, { "grad_norm": 0.3026581108570099, "learning_rate": 3.4675794610692914e-05, "loss": 0.0221, "step": 123840 }, { "grad_norm": 0.38424152135849, "learning_rate": 3.4667925330884296e-05, "loss": 0.0256, "step": 123850 }, { "grad_norm": 0.30920305848121643, "learning_rate": 3.4660056470248816e-05, "loss": 0.0193, "step": 123860 }, { "grad_norm": 1.0737581253051758, "learning_rate": 3.465218802900163e-05, "loss": 0.0275, "step": 123870 }, { "grad_norm": 0.5170957446098328, "learning_rate": 3.464432000735787e-05, "loss": 0.0398, "step": 123880 }, { "grad_norm": 0.3357340693473816, "learning_rate": 3.4636452405532624e-05, "loss": 0.0225, "step": 123890 }, { "grad_norm": 0.41762787103652954, "learning_rate": 3.4628585223741e-05, "loss": 0.0329, "step": 123900 }, { "grad_norm": 0.8351077437400818, "learning_rate": 3.4620718462198084e-05, "loss": 0.0241, "step": 123910 }, { "grad_norm": 0.8334488868713379, "learning_rate": 3.4612852121118926e-05, "loss": 0.0244, "step": 123920 }, { "grad_norm": 0.44288018345832825, "learning_rate": 3.4604986200718646e-05, "loss": 0.0203, "step": 123930 }, { "grad_norm": 0.6718307137489319, "learning_rate": 3.459712070121223e-05, "loss": 0.0247, "step": 123940 }, { "grad_norm": 0.43389031291007996, "learning_rate": 3.458925562281476e-05, "loss": 0.022, "step": 123950 }, { "grad_norm": 0.33128607273101807, "learning_rate": 3.4581390965741254e-05, "loss": 0.0261, "step": 123960 }, { "grad_norm": 0.6920304894447327, "learning_rate": 3.457352673020673e-05, "loss": 0.0248, "step": 123970 }, { "grad_norm": 0.46411561965942383, "learning_rate": 3.456566291642618e-05, "loss": 0.0184, "step": 123980 }, { "grad_norm": 0.4951499104499817, "learning_rate": 3.4557799524614606e-05, "loss": 0.0315, "step": 123990 }, { "grad_norm": 0.3755267560482025, "learning_rate": 3.4549936554987e-05, "loss": 0.0235, "step": 124000 }, { "grad_norm": 0.6906748414039612, "learning_rate": 3.454207400775832e-05, "loss": 0.0326, "step": 124010 }, { "grad_norm": 0.5258301496505737, "learning_rate": 3.453421188314354e-05, "loss": 0.0328, "step": 124020 }, { "grad_norm": 0.517210841178894, "learning_rate": 3.452635018135757e-05, "loss": 0.0231, "step": 124030 }, { "grad_norm": 1.1205624341964722, "learning_rate": 3.451848890261541e-05, "loss": 0.0227, "step": 124040 }, { "grad_norm": 0.32250165939331055, "learning_rate": 3.451062804713191e-05, "loss": 0.0228, "step": 124050 }, { "grad_norm": 0.5999962091445923, "learning_rate": 3.4502767615122036e-05, "loss": 0.0261, "step": 124060 }, { "grad_norm": 0.4525371789932251, "learning_rate": 3.449490760680068e-05, "loss": 0.0249, "step": 124070 }, { "grad_norm": 0.38541319966316223, "learning_rate": 3.448704802238272e-05, "loss": 0.0271, "step": 124080 }, { "grad_norm": 0.30837011337280273, "learning_rate": 3.447918886208304e-05, "loss": 0.0272, "step": 124090 }, { "grad_norm": 0.5751372575759888, "learning_rate": 3.44713301261165e-05, "loss": 0.0219, "step": 124100 }, { "grad_norm": 0.41550886631011963, "learning_rate": 3.446347181469796e-05, "loss": 0.0227, "step": 124110 }, { "grad_norm": 0.23263047635555267, "learning_rate": 3.445561392804228e-05, "loss": 0.0158, "step": 124120 }, { "grad_norm": 0.6458568572998047, "learning_rate": 3.444775646636425e-05, "loss": 0.0192, "step": 124130 }, { "grad_norm": 0.3535769581794739, "learning_rate": 3.443989942987875e-05, "loss": 0.0216, "step": 124140 }, { "grad_norm": 0.6465983986854553, "learning_rate": 3.443204281880051e-05, "loss": 0.0248, "step": 124150 }, { "grad_norm": 0.4629609286785126, "learning_rate": 3.442418663334441e-05, "loss": 0.0249, "step": 124160 }, { "grad_norm": 0.9919087886810303, "learning_rate": 3.441633087372518e-05, "loss": 0.0285, "step": 124170 }, { "grad_norm": 0.4544961452484131, "learning_rate": 3.4408475540157606e-05, "loss": 0.036, "step": 124180 }, { "grad_norm": 0.39641255140304565, "learning_rate": 3.4400620632856464e-05, "loss": 0.0426, "step": 124190 }, { "grad_norm": 0.39644771814346313, "learning_rate": 3.439276615203648e-05, "loss": 0.0274, "step": 124200 }, { "grad_norm": 0.7311402559280396, "learning_rate": 3.438491209791242e-05, "loss": 0.0266, "step": 124210 }, { "grad_norm": 0.4180125594139099, "learning_rate": 3.4377058470698985e-05, "loss": 0.0258, "step": 124220 }, { "grad_norm": 0.3246649205684662, "learning_rate": 3.436920527061091e-05, "loss": 0.0172, "step": 124230 }, { "grad_norm": 0.8685842752456665, "learning_rate": 3.4361352497862874e-05, "loss": 0.0261, "step": 124240 }, { "grad_norm": 0.8966320753097534, "learning_rate": 3.4353500152669593e-05, "loss": 0.0316, "step": 124250 }, { "grad_norm": 0.4040336012840271, "learning_rate": 3.434564823524575e-05, "loss": 0.0249, "step": 124260 }, { "grad_norm": 0.3619920015335083, "learning_rate": 3.4337796745806e-05, "loss": 0.0258, "step": 124270 }, { "grad_norm": 0.4241500496864319, "learning_rate": 3.432994568456501e-05, "loss": 0.0287, "step": 124280 }, { "grad_norm": 0.5800385475158691, "learning_rate": 3.4322095051737404e-05, "loss": 0.0357, "step": 124290 }, { "grad_norm": 0.6486972570419312, "learning_rate": 3.431424484753784e-05, "loss": 0.0186, "step": 124300 }, { "grad_norm": 0.35307568311691284, "learning_rate": 3.4306395072180916e-05, "loss": 0.0437, "step": 124310 }, { "grad_norm": 0.5013561844825745, "learning_rate": 3.4298545725881256e-05, "loss": 0.031, "step": 124320 }, { "grad_norm": 0.32002881169319153, "learning_rate": 3.429069680885348e-05, "loss": 0.0213, "step": 124330 }, { "grad_norm": 0.3519818186759949, "learning_rate": 3.4282848321312124e-05, "loss": 0.0225, "step": 124340 }, { "grad_norm": 1.5186070203781128, "learning_rate": 3.4275000263471814e-05, "loss": 0.0302, "step": 124350 }, { "grad_norm": 0.3624362647533417, "learning_rate": 3.426715263554706e-05, "loss": 0.0307, "step": 124360 }, { "grad_norm": 0.2734183371067047, "learning_rate": 3.425930543775246e-05, "loss": 0.0207, "step": 124370 }, { "grad_norm": 0.7797427177429199, "learning_rate": 3.425145867030254e-05, "loss": 0.0193, "step": 124380 }, { "grad_norm": 0.4076932668685913, "learning_rate": 3.424361233341182e-05, "loss": 0.022, "step": 124390 }, { "grad_norm": 0.26860421895980835, "learning_rate": 3.4235766427294825e-05, "loss": 0.0238, "step": 124400 }, { "grad_norm": 0.5807006359100342, "learning_rate": 3.422792095216604e-05, "loss": 0.027, "step": 124410 }, { "grad_norm": 0.33949464559555054, "learning_rate": 3.422007590823999e-05, "loss": 0.0251, "step": 124420 }, { "grad_norm": 1.0993961095809937, "learning_rate": 3.4212231295731114e-05, "loss": 0.0263, "step": 124430 }, { "grad_norm": 0.4026690423488617, "learning_rate": 3.42043871148539e-05, "loss": 0.0284, "step": 124440 }, { "grad_norm": 0.5063566565513611, "learning_rate": 3.419654336582283e-05, "loss": 0.0177, "step": 124450 }, { "grad_norm": 0.5440405607223511, "learning_rate": 3.4188700048852314e-05, "loss": 0.0259, "step": 124460 }, { "grad_norm": 0.35982805490493774, "learning_rate": 3.418085716415681e-05, "loss": 0.0262, "step": 124470 }, { "grad_norm": 0.9810547232627869, "learning_rate": 3.417301471195071e-05, "loss": 0.0231, "step": 124480 }, { "grad_norm": 0.5393125414848328, "learning_rate": 3.4165172692448465e-05, "loss": 0.0194, "step": 124490 }, { "grad_norm": 0.8654168844223022, "learning_rate": 3.415733110586443e-05, "loss": 0.0251, "step": 124500 }, { "grad_norm": 0.42603039741516113, "learning_rate": 3.414948995241301e-05, "loss": 0.0336, "step": 124510 }, { "grad_norm": 0.42855334281921387, "learning_rate": 3.4141649232308604e-05, "loss": 0.0164, "step": 124520 }, { "grad_norm": 0.42393842339515686, "learning_rate": 3.4133808945765524e-05, "loss": 0.0189, "step": 124530 }, { "grad_norm": 0.42545703053474426, "learning_rate": 3.412596909299817e-05, "loss": 0.0242, "step": 124540 }, { "grad_norm": 0.5330047607421875, "learning_rate": 3.411812967422083e-05, "loss": 0.0294, "step": 124550 }, { "grad_norm": 0.45778974890708923, "learning_rate": 3.411029068964787e-05, "loss": 0.0193, "step": 124560 }, { "grad_norm": 0.3252594470977783, "learning_rate": 3.410245213949361e-05, "loss": 0.0209, "step": 124570 }, { "grad_norm": 0.5815754532814026, "learning_rate": 3.4094614023972315e-05, "loss": 0.0265, "step": 124580 }, { "grad_norm": 0.689024031162262, "learning_rate": 3.408677634329831e-05, "loss": 0.0219, "step": 124590 }, { "grad_norm": 0.42837589979171753, "learning_rate": 3.407893909768586e-05, "loss": 0.0222, "step": 124600 }, { "grad_norm": 0.43931418657302856, "learning_rate": 3.407110228734923e-05, "loss": 0.0233, "step": 124610 }, { "grad_norm": 0.3724110722541809, "learning_rate": 3.406326591250267e-05, "loss": 0.0169, "step": 124620 }, { "grad_norm": 0.5215358734130859, "learning_rate": 3.405542997336043e-05, "loss": 0.0231, "step": 124630 }, { "grad_norm": 0.374509334564209, "learning_rate": 3.404759447013678e-05, "loss": 0.0299, "step": 124640 }, { "grad_norm": 0.40457120537757874, "learning_rate": 3.403975940304586e-05, "loss": 0.0181, "step": 124650 }, { "grad_norm": 0.5873835682868958, "learning_rate": 3.4031924772301955e-05, "loss": 0.0292, "step": 124660 }, { "grad_norm": 0.8028077483177185, "learning_rate": 3.402409057811921e-05, "loss": 0.0241, "step": 124670 }, { "grad_norm": 0.579888105392456, "learning_rate": 3.401625682071184e-05, "loss": 0.0234, "step": 124680 }, { "grad_norm": 0.27362966537475586, "learning_rate": 3.400842350029399e-05, "loss": 0.0269, "step": 124690 }, { "grad_norm": 0.9599738121032715, "learning_rate": 3.400059061707983e-05, "loss": 0.0271, "step": 124700 }, { "grad_norm": 0.5141898393630981, "learning_rate": 3.3992758171283536e-05, "loss": 0.0285, "step": 124710 }, { "grad_norm": 0.40607109665870667, "learning_rate": 3.3984926163119204e-05, "loss": 0.0191, "step": 124720 }, { "grad_norm": 0.8264524936676025, "learning_rate": 3.3977094592800976e-05, "loss": 0.0323, "step": 124730 }, { "grad_norm": 0.4579809606075287, "learning_rate": 3.396926346054295e-05, "loss": 0.0225, "step": 124740 }, { "grad_norm": 0.3900855481624603, "learning_rate": 3.3961432766559244e-05, "loss": 0.0184, "step": 124750 }, { "grad_norm": 0.9960973262786865, "learning_rate": 3.3953602511063954e-05, "loss": 0.0395, "step": 124760 }, { "grad_norm": 0.4729582369327545, "learning_rate": 3.394577269427113e-05, "loss": 0.0291, "step": 124770 }, { "grad_norm": 0.3804160952568054, "learning_rate": 3.393794331639486e-05, "loss": 0.0247, "step": 124780 }, { "grad_norm": 0.47437816858291626, "learning_rate": 3.393011437764918e-05, "loss": 0.0231, "step": 124790 }, { "grad_norm": 0.6529944539070129, "learning_rate": 3.392228587824814e-05, "loss": 0.0159, "step": 124800 }, { "grad_norm": 0.3587714433670044, "learning_rate": 3.3914457818405766e-05, "loss": 0.0214, "step": 124810 }, { "grad_norm": 0.4989132583141327, "learning_rate": 3.3906630198336045e-05, "loss": 0.0192, "step": 124820 }, { "grad_norm": 0.26104941964149475, "learning_rate": 3.389880301825305e-05, "loss": 0.0275, "step": 124830 }, { "grad_norm": 0.3564930856227875, "learning_rate": 3.38909762783707e-05, "loss": 0.0177, "step": 124840 }, { "grad_norm": 0.4793790876865387, "learning_rate": 3.3883149978903036e-05, "loss": 0.0194, "step": 124850 }, { "grad_norm": 0.4598221182823181, "learning_rate": 3.387532412006396e-05, "loss": 0.0318, "step": 124860 }, { "grad_norm": 1.1301039457321167, "learning_rate": 3.386749870206749e-05, "loss": 0.0301, "step": 124870 }, { "grad_norm": 0.23891210556030273, "learning_rate": 3.385967372512754e-05, "loss": 0.0237, "step": 124880 }, { "grad_norm": 0.3518703281879425, "learning_rate": 3.385184918945804e-05, "loss": 0.0254, "step": 124890 }, { "grad_norm": 0.220443457365036, "learning_rate": 3.384402509527292e-05, "loss": 0.0261, "step": 124900 }, { "grad_norm": 0.5820611119270325, "learning_rate": 3.383620144278609e-05, "loss": 0.0298, "step": 124910 }, { "grad_norm": 0.5178495645523071, "learning_rate": 3.382837823221144e-05, "loss": 0.0306, "step": 124920 }, { "grad_norm": 0.841031014919281, "learning_rate": 3.382055546376285e-05, "loss": 0.0297, "step": 124930 }, { "grad_norm": 0.4474771022796631, "learning_rate": 3.3812733137654176e-05, "loss": 0.0182, "step": 124940 }, { "grad_norm": 0.41798776388168335, "learning_rate": 3.380491125409932e-05, "loss": 0.0163, "step": 124950 }, { "grad_norm": 0.5473369359970093, "learning_rate": 3.3797089813312103e-05, "loss": 0.0263, "step": 124960 }, { "grad_norm": 0.40799930691719055, "learning_rate": 3.378926881550637e-05, "loss": 0.0199, "step": 124970 }, { "grad_norm": 0.4180872142314911, "learning_rate": 3.3781448260895925e-05, "loss": 0.0203, "step": 124980 }, { "grad_norm": 0.342034250497818, "learning_rate": 3.37736281496946e-05, "loss": 0.0182, "step": 124990 }, { "grad_norm": 0.4440673291683197, "learning_rate": 3.376580848211618e-05, "loss": 0.0246, "step": 125000 }, { "grad_norm": 0.34412524104118347, "learning_rate": 3.375798925837446e-05, "loss": 0.0217, "step": 125010 }, { "grad_norm": 0.5821793675422668, "learning_rate": 3.375017047868322e-05, "loss": 0.022, "step": 125020 }, { "grad_norm": 1.618884563446045, "learning_rate": 3.374235214325618e-05, "loss": 0.0285, "step": 125030 }, { "grad_norm": 1.1441929340362549, "learning_rate": 3.373453425230717e-05, "loss": 0.0287, "step": 125040 }, { "grad_norm": 0.352668434381485, "learning_rate": 3.3726716806049855e-05, "loss": 0.0175, "step": 125050 }, { "grad_norm": 0.26405325531959534, "learning_rate": 3.3718899804698014e-05, "loss": 0.0212, "step": 125060 }, { "grad_norm": 0.48697128891944885, "learning_rate": 3.37110832484653e-05, "loss": 0.0214, "step": 125070 }, { "grad_norm": 0.24711288511753082, "learning_rate": 3.3703267137565465e-05, "loss": 0.0195, "step": 125080 }, { "grad_norm": 0.4502123296260834, "learning_rate": 3.36954514722122e-05, "loss": 0.0256, "step": 125090 }, { "grad_norm": 0.6136308908462524, "learning_rate": 3.368763625261915e-05, "loss": 0.0154, "step": 125100 }, { "grad_norm": 0.3604443073272705, "learning_rate": 3.3679821479000005e-05, "loss": 0.0207, "step": 125110 }, { "grad_norm": 0.5704986453056335, "learning_rate": 3.36720071515684e-05, "loss": 0.0218, "step": 125120 }, { "grad_norm": 0.6685199737548828, "learning_rate": 3.366419327053798e-05, "loss": 0.0308, "step": 125130 }, { "grad_norm": 0.21630172431468964, "learning_rate": 3.365637983612241e-05, "loss": 0.0166, "step": 125140 }, { "grad_norm": 0.2696548104286194, "learning_rate": 3.364856684853523e-05, "loss": 0.0165, "step": 125150 }, { "grad_norm": 0.3942376971244812, "learning_rate": 3.364075430799013e-05, "loss": 0.05, "step": 125160 }, { "grad_norm": 0.393794983625412, "learning_rate": 3.363294221470064e-05, "loss": 0.0247, "step": 125170 }, { "grad_norm": 0.37490254640579224, "learning_rate": 3.362513056888038e-05, "loss": 0.0269, "step": 125180 }, { "grad_norm": 0.3438234031200409, "learning_rate": 3.361731937074288e-05, "loss": 0.0197, "step": 125190 }, { "grad_norm": 0.4611256718635559, "learning_rate": 3.3609508620501716e-05, "loss": 0.0186, "step": 125200 }, { "grad_norm": 0.38638246059417725, "learning_rate": 3.3601698318370433e-05, "loss": 0.02, "step": 125210 }, { "grad_norm": 0.5373756289482117, "learning_rate": 3.3593888464562556e-05, "loss": 0.0202, "step": 125220 }, { "grad_norm": 0.5526480078697205, "learning_rate": 3.358607905929161e-05, "loss": 0.0231, "step": 125230 }, { "grad_norm": 0.42865800857543945, "learning_rate": 3.357827010277108e-05, "loss": 0.0173, "step": 125240 }, { "grad_norm": 0.4506802558898926, "learning_rate": 3.3570461595214504e-05, "loss": 0.0219, "step": 125250 }, { "grad_norm": 0.3015439510345459, "learning_rate": 3.35626535368353e-05, "loss": 0.0192, "step": 125260 }, { "grad_norm": 0.4526967704296112, "learning_rate": 3.355484592784699e-05, "loss": 0.0468, "step": 125270 }, { "grad_norm": 0.21371617913246155, "learning_rate": 3.3547038768463025e-05, "loss": 0.0173, "step": 125280 }, { "grad_norm": 0.505095362663269, "learning_rate": 3.353923205889683e-05, "loss": 0.0221, "step": 125290 }, { "grad_norm": 0.3605586588382721, "learning_rate": 3.353142579936185e-05, "loss": 0.0225, "step": 125300 }, { "grad_norm": 0.5393685698509216, "learning_rate": 3.35236199900715e-05, "loss": 0.0188, "step": 125310 }, { "grad_norm": 0.46296006441116333, "learning_rate": 3.3515814631239175e-05, "loss": 0.0203, "step": 125320 }, { "grad_norm": 0.440330445766449, "learning_rate": 3.350800972307831e-05, "loss": 0.0207, "step": 125330 }, { "grad_norm": 0.44567155838012695, "learning_rate": 3.3500205265802234e-05, "loss": 0.023, "step": 125340 }, { "grad_norm": 0.6859493255615234, "learning_rate": 3.349240125962438e-05, "loss": 0.0276, "step": 125350 }, { "grad_norm": 0.5418765544891357, "learning_rate": 3.348459770475805e-05, "loss": 0.0232, "step": 125360 }, { "grad_norm": 0.8371528387069702, "learning_rate": 3.347679460141663e-05, "loss": 0.0282, "step": 125370 }, { "grad_norm": 0.23556485772132874, "learning_rate": 3.3468991949813426e-05, "loss": 0.0189, "step": 125380 }, { "grad_norm": 1.0562686920166016, "learning_rate": 3.3461189750161774e-05, "loss": 0.0195, "step": 125390 }, { "grad_norm": 0.48620688915252686, "learning_rate": 3.345338800267499e-05, "loss": 0.0182, "step": 125400 }, { "grad_norm": 0.37746065855026245, "learning_rate": 3.3445586707566356e-05, "loss": 0.0261, "step": 125410 }, { "grad_norm": 0.19639624655246735, "learning_rate": 3.343778586504917e-05, "loss": 0.0188, "step": 125420 }, { "grad_norm": 0.46755701303482056, "learning_rate": 3.342998547533668e-05, "loss": 0.0178, "step": 125430 }, { "grad_norm": 0.45140907168388367, "learning_rate": 3.3422185538642175e-05, "loss": 0.023, "step": 125440 }, { "grad_norm": 0.6129220724105835, "learning_rate": 3.341438605517887e-05, "loss": 0.0293, "step": 125450 }, { "grad_norm": 0.7052303552627563, "learning_rate": 3.340658702516003e-05, "loss": 0.0165, "step": 125460 }, { "grad_norm": 0.3311740458011627, "learning_rate": 3.339878844879888e-05, "loss": 0.0138, "step": 125470 }, { "grad_norm": 0.37194693088531494, "learning_rate": 3.33909903263086e-05, "loss": 0.0292, "step": 125480 }, { "grad_norm": 0.5961225032806396, "learning_rate": 3.3383192657902424e-05, "loss": 0.0252, "step": 125490 }, { "grad_norm": 0.29390478134155273, "learning_rate": 3.33753954437935e-05, "loss": 0.0247, "step": 125500 }, { "grad_norm": 0.39471516013145447, "learning_rate": 3.3367598684195034e-05, "loss": 0.033, "step": 125510 }, { "grad_norm": 0.4457206428050995, "learning_rate": 3.335980237932016e-05, "loss": 0.0186, "step": 125520 }, { "grad_norm": 0.6387996077537537, "learning_rate": 3.335200652938203e-05, "loss": 0.0248, "step": 125530 }, { "grad_norm": 0.35402363538742065, "learning_rate": 3.334421113459381e-05, "loss": 0.0162, "step": 125540 }, { "grad_norm": 0.2919311225414276, "learning_rate": 3.333641619516857e-05, "loss": 0.0227, "step": 125550 }, { "grad_norm": 0.2953771650791168, "learning_rate": 3.3328621711319494e-05, "loss": 0.0156, "step": 125560 }, { "grad_norm": 0.27659872174263, "learning_rate": 3.332082768325959e-05, "loss": 0.0191, "step": 125570 }, { "grad_norm": 0.34590160846710205, "learning_rate": 3.331303411120201e-05, "loss": 0.0181, "step": 125580 }, { "grad_norm": 0.45545265078544617, "learning_rate": 3.330524099535982e-05, "loss": 0.0231, "step": 125590 }, { "grad_norm": 0.6111665368080139, "learning_rate": 3.3297448335946056e-05, "loss": 0.0277, "step": 125600 }, { "grad_norm": 0.48582983016967773, "learning_rate": 3.328965613317379e-05, "loss": 0.0245, "step": 125610 }, { "grad_norm": 0.3848526179790497, "learning_rate": 3.328186438725604e-05, "loss": 0.038, "step": 125620 }, { "grad_norm": 1.27117919921875, "learning_rate": 3.327407309840585e-05, "loss": 0.0244, "step": 125630 }, { "grad_norm": 0.13869528472423553, "learning_rate": 3.32662822668362e-05, "loss": 0.0277, "step": 125640 }, { "grad_norm": 0.30794429779052734, "learning_rate": 3.3258491892760104e-05, "loss": 0.0161, "step": 125650 }, { "grad_norm": 0.21998822689056396, "learning_rate": 3.3250701976390565e-05, "loss": 0.0198, "step": 125660 }, { "grad_norm": 0.7359896898269653, "learning_rate": 3.324291251794053e-05, "loss": 0.0265, "step": 125670 }, { "grad_norm": 0.4087701439857483, "learning_rate": 3.3235123517622985e-05, "loss": 0.0213, "step": 125680 }, { "grad_norm": 0.27370643615722656, "learning_rate": 3.3227334975650854e-05, "loss": 0.0194, "step": 125690 }, { "grad_norm": 0.6938113570213318, "learning_rate": 3.32195468922371e-05, "loss": 0.0302, "step": 125700 }, { "grad_norm": 0.2613016366958618, "learning_rate": 3.321175926759461e-05, "loss": 0.0251, "step": 125710 }, { "grad_norm": 0.24868400394916534, "learning_rate": 3.320397210193631e-05, "loss": 0.0285, "step": 125720 }, { "grad_norm": 0.5003488063812256, "learning_rate": 3.319618539547512e-05, "loss": 0.0217, "step": 125730 }, { "grad_norm": 0.2035694420337677, "learning_rate": 3.318839914842389e-05, "loss": 0.0207, "step": 125740 }, { "grad_norm": 0.417823851108551, "learning_rate": 3.318061336099554e-05, "loss": 0.0207, "step": 125750 }, { "grad_norm": 1.2699798345565796, "learning_rate": 3.317282803340288e-05, "loss": 0.0219, "step": 125760 }, { "grad_norm": 0.8811763525009155, "learning_rate": 3.316504316585878e-05, "loss": 0.0228, "step": 125770 }, { "grad_norm": 0.2592046558856964, "learning_rate": 3.3157258758576084e-05, "loss": 0.026, "step": 125780 }, { "grad_norm": 0.325100839138031, "learning_rate": 3.31494748117676e-05, "loss": 0.0199, "step": 125790 }, { "grad_norm": 0.22508715093135834, "learning_rate": 3.314169132564615e-05, "loss": 0.0216, "step": 125800 }, { "grad_norm": 0.39353904128074646, "learning_rate": 3.313390830042452e-05, "loss": 0.0143, "step": 125810 }, { "grad_norm": 0.6733782887458801, "learning_rate": 3.312612573631551e-05, "loss": 0.0163, "step": 125820 }, { "grad_norm": 0.2236563265323639, "learning_rate": 3.311834363353187e-05, "loss": 0.0222, "step": 125830 }, { "grad_norm": 1.3714691400527954, "learning_rate": 3.311056199228637e-05, "loss": 0.0333, "step": 125840 }, { "grad_norm": 0.37506693601608276, "learning_rate": 3.310278081279179e-05, "loss": 0.026, "step": 125850 }, { "grad_norm": 0.9172245264053345, "learning_rate": 3.3095000095260795e-05, "loss": 0.0334, "step": 125860 }, { "grad_norm": 0.2992621064186096, "learning_rate": 3.308721983990618e-05, "loss": 0.0236, "step": 125870 }, { "grad_norm": 0.5703729391098022, "learning_rate": 3.30794400469406e-05, "loss": 0.0229, "step": 125880 }, { "grad_norm": 0.540387749671936, "learning_rate": 3.307166071657678e-05, "loss": 0.0215, "step": 125890 }, { "grad_norm": 0.6938477158546448, "learning_rate": 3.306388184902739e-05, "loss": 0.0272, "step": 125900 }, { "grad_norm": 0.43006962537765503, "learning_rate": 3.3056103444505115e-05, "loss": 0.0279, "step": 125910 }, { "grad_norm": 0.5674951076507568, "learning_rate": 3.304832550322261e-05, "loss": 0.0207, "step": 125920 }, { "grad_norm": 0.4688103199005127, "learning_rate": 3.3040548025392504e-05, "loss": 0.024, "step": 125930 }, { "grad_norm": 0.620414137840271, "learning_rate": 3.3032771011227456e-05, "loss": 0.0232, "step": 125940 }, { "grad_norm": 0.378485769033432, "learning_rate": 3.302499446094006e-05, "loss": 0.0278, "step": 125950 }, { "grad_norm": 0.6210736632347107, "learning_rate": 3.301721837474294e-05, "loss": 0.0221, "step": 125960 }, { "grad_norm": 1.237515926361084, "learning_rate": 3.300944275284871e-05, "loss": 0.024, "step": 125970 }, { "grad_norm": 0.46157997846603394, "learning_rate": 3.300166759546991e-05, "loss": 0.02, "step": 125980 }, { "grad_norm": 0.38741570711135864, "learning_rate": 3.2993892902819155e-05, "loss": 0.0145, "step": 125990 }, { "grad_norm": 0.5147388577461243, "learning_rate": 3.298611867510896e-05, "loss": 0.0308, "step": 126000 }, { "grad_norm": 0.33033862709999084, "learning_rate": 3.29783449125519e-05, "loss": 0.0301, "step": 126010 }, { "grad_norm": 0.2923024594783783, "learning_rate": 3.297057161536048e-05, "loss": 0.0302, "step": 126020 }, { "grad_norm": 0.3626921772956848, "learning_rate": 3.296279878374723e-05, "loss": 0.0202, "step": 126030 }, { "grad_norm": 0.33456915616989136, "learning_rate": 3.29550264179247e-05, "loss": 0.0226, "step": 126040 }, { "grad_norm": 0.4242924153804779, "learning_rate": 3.2947254518105296e-05, "loss": 0.0203, "step": 126050 }, { "grad_norm": 0.4098478853702545, "learning_rate": 3.2939483084501584e-05, "loss": 0.0261, "step": 126060 }, { "grad_norm": 0.431144118309021, "learning_rate": 3.293171211732598e-05, "loss": 0.0315, "step": 126070 }, { "grad_norm": 0.7663939595222473, "learning_rate": 3.292394161679095e-05, "loss": 0.0177, "step": 126080 }, { "grad_norm": 0.22839394211769104, "learning_rate": 3.291617158310895e-05, "loss": 0.0202, "step": 126090 }, { "grad_norm": 0.37742671370506287, "learning_rate": 3.290840201649239e-05, "loss": 0.0235, "step": 126100 }, { "grad_norm": 0.3506004214286804, "learning_rate": 3.2900632917153705e-05, "loss": 0.0354, "step": 126110 }, { "grad_norm": 0.6311632990837097, "learning_rate": 3.2892864285305294e-05, "loss": 0.0238, "step": 126120 }, { "grad_norm": 0.9902317523956299, "learning_rate": 3.2885096121159556e-05, "loss": 0.033, "step": 126130 }, { "grad_norm": 0.7011279463768005, "learning_rate": 3.287732842492884e-05, "loss": 0.0216, "step": 126140 }, { "grad_norm": 0.5548401474952698, "learning_rate": 3.286956119682553e-05, "loss": 0.0218, "step": 126150 }, { "grad_norm": 0.7478533983230591, "learning_rate": 3.2861794437061995e-05, "loss": 0.0385, "step": 126160 }, { "grad_norm": 0.41812387108802795, "learning_rate": 3.285402814585056e-05, "loss": 0.0197, "step": 126170 }, { "grad_norm": 0.38378435373306274, "learning_rate": 3.2846262323403556e-05, "loss": 0.0201, "step": 126180 }, { "grad_norm": 0.35225605964660645, "learning_rate": 3.2838496969933284e-05, "loss": 0.0249, "step": 126190 }, { "grad_norm": 0.5343494415283203, "learning_rate": 3.283073208565207e-05, "loss": 0.0376, "step": 126200 }, { "grad_norm": 0.49811556935310364, "learning_rate": 3.2822967670772185e-05, "loss": 0.0209, "step": 126210 }, { "grad_norm": 0.5033469796180725, "learning_rate": 3.281520372550592e-05, "loss": 0.0175, "step": 126220 }, { "grad_norm": 0.37180060148239136, "learning_rate": 3.280744025006552e-05, "loss": 0.0358, "step": 126230 }, { "grad_norm": 0.46288490295410156, "learning_rate": 3.279967724466323e-05, "loss": 0.0208, "step": 126240 }, { "grad_norm": 0.6164326667785645, "learning_rate": 3.279191470951134e-05, "loss": 0.0278, "step": 126250 }, { "grad_norm": 0.5574764609336853, "learning_rate": 3.2784152644822e-05, "loss": 0.0177, "step": 126260 }, { "grad_norm": 0.34873539209365845, "learning_rate": 3.277639105080749e-05, "loss": 0.0205, "step": 126270 }, { "grad_norm": 0.3255484700202942, "learning_rate": 3.276862992767995e-05, "loss": 0.0215, "step": 126280 }, { "grad_norm": 0.8850418329238892, "learning_rate": 3.276086927565161e-05, "loss": 0.0261, "step": 126290 }, { "grad_norm": 0.47890594601631165, "learning_rate": 3.275310909493462e-05, "loss": 0.0278, "step": 126300 }, { "grad_norm": 0.6338419914245605, "learning_rate": 3.2745349385741156e-05, "loss": 0.0512, "step": 126310 }, { "grad_norm": 0.3932248055934906, "learning_rate": 3.273759014828336e-05, "loss": 0.0184, "step": 126320 }, { "grad_norm": 0.3965696096420288, "learning_rate": 3.272983138277335e-05, "loss": 0.0233, "step": 126330 }, { "grad_norm": 0.42666441202163696, "learning_rate": 3.2722073089423256e-05, "loss": 0.0248, "step": 126340 }, { "grad_norm": 0.3709907829761505, "learning_rate": 3.2714315268445226e-05, "loss": 0.0234, "step": 126350 }, { "grad_norm": 0.43746405839920044, "learning_rate": 3.270655792005129e-05, "loss": 0.0196, "step": 126360 }, { "grad_norm": 0.7037531733512878, "learning_rate": 3.269880104445358e-05, "loss": 0.0207, "step": 126370 }, { "grad_norm": 0.7394352555274963, "learning_rate": 3.269104464186414e-05, "loss": 0.0268, "step": 126380 }, { "grad_norm": 0.8114873170852661, "learning_rate": 3.268328871249505e-05, "loss": 0.0434, "step": 126390 }, { "grad_norm": 0.5996654629707336, "learning_rate": 3.267553325655833e-05, "loss": 0.0275, "step": 126400 }, { "grad_norm": 0.9397968053817749, "learning_rate": 3.266777827426603e-05, "loss": 0.0229, "step": 126410 }, { "grad_norm": 0.3554595410823822, "learning_rate": 3.266002376583016e-05, "loss": 0.0212, "step": 126420 }, { "grad_norm": 0.6870992183685303, "learning_rate": 3.265226973146272e-05, "loss": 0.0365, "step": 126430 }, { "grad_norm": 0.7194455862045288, "learning_rate": 3.264451617137572e-05, "loss": 0.029, "step": 126440 }, { "grad_norm": 0.606481671333313, "learning_rate": 3.2636763085781105e-05, "loss": 0.0179, "step": 126450 }, { "grad_norm": 0.49510911107063293, "learning_rate": 3.2629010474890895e-05, "loss": 0.0204, "step": 126460 }, { "grad_norm": 0.3225805461406708, "learning_rate": 3.262125833891698e-05, "loss": 0.0241, "step": 126470 }, { "grad_norm": 0.5928884744644165, "learning_rate": 3.261350667807135e-05, "loss": 0.0261, "step": 126480 }, { "grad_norm": 0.25133517384529114, "learning_rate": 3.260575549256592e-05, "loss": 0.0228, "step": 126490 }, { "grad_norm": 0.5102538466453552, "learning_rate": 3.259800478261259e-05, "loss": 0.0393, "step": 126500 }, { "grad_norm": 0.8033786416053772, "learning_rate": 3.2590254548423285e-05, "loss": 0.0318, "step": 126510 }, { "grad_norm": 0.3114439845085144, "learning_rate": 3.258250479020987e-05, "loss": 0.0242, "step": 126520 }, { "grad_norm": 0.3375873565673828, "learning_rate": 3.2574755508184215e-05, "loss": 0.0251, "step": 126530 }, { "grad_norm": 0.27707311511039734, "learning_rate": 3.2567006702558235e-05, "loss": 0.0209, "step": 126540 }, { "grad_norm": 0.5560433864593506, "learning_rate": 3.25592583735437e-05, "loss": 0.0335, "step": 126550 }, { "grad_norm": 0.672116219997406, "learning_rate": 3.2551510521352535e-05, "loss": 0.0192, "step": 126560 }, { "grad_norm": 0.5023679733276367, "learning_rate": 3.254376314619648e-05, "loss": 0.0275, "step": 126570 }, { "grad_norm": 0.5119009613990784, "learning_rate": 3.2536016248287405e-05, "loss": 0.0278, "step": 126580 }, { "grad_norm": 0.6124194860458374, "learning_rate": 3.252826982783707e-05, "loss": 0.0285, "step": 126590 }, { "grad_norm": 0.3700803518295288, "learning_rate": 3.252052388505728e-05, "loss": 0.0242, "step": 126600 }, { "grad_norm": 0.29139065742492676, "learning_rate": 3.251277842015981e-05, "loss": 0.0178, "step": 126610 }, { "grad_norm": 0.6603186726570129, "learning_rate": 3.250503343335639e-05, "loss": 0.0264, "step": 126620 }, { "grad_norm": 1.4348446130752563, "learning_rate": 3.24972889248588e-05, "loss": 0.0306, "step": 126630 }, { "grad_norm": 0.4147458076477051, "learning_rate": 3.248954489487874e-05, "loss": 0.0265, "step": 126640 }, { "grad_norm": 0.4992983937263489, "learning_rate": 3.248180134362796e-05, "loss": 0.032, "step": 126650 }, { "grad_norm": 0.6202115416526794, "learning_rate": 3.247405827131813e-05, "loss": 0.0258, "step": 126660 }, { "grad_norm": 0.9163758158683777, "learning_rate": 3.246631567816097e-05, "loss": 0.0193, "step": 126670 }, { "grad_norm": 0.30778083205223083, "learning_rate": 3.245857356436817e-05, "loss": 0.0247, "step": 126680 }, { "grad_norm": 0.29758235812187195, "learning_rate": 3.245083193015135e-05, "loss": 0.0283, "step": 126690 }, { "grad_norm": 0.36686962842941284, "learning_rate": 3.244309077572222e-05, "loss": 0.0222, "step": 126700 }, { "grad_norm": 1.5178085565567017, "learning_rate": 3.2435350101292375e-05, "loss": 0.0232, "step": 126710 }, { "grad_norm": 0.32116714119911194, "learning_rate": 3.2427609907073454e-05, "loss": 0.0226, "step": 126720 }, { "grad_norm": 0.760321855545044, "learning_rate": 3.2419870193277105e-05, "loss": 0.0575, "step": 126730 }, { "grad_norm": 0.5268489122390747, "learning_rate": 3.2412130960114864e-05, "loss": 0.0231, "step": 126740 }, { "grad_norm": 0.5022284388542175, "learning_rate": 3.2404392207798394e-05, "loss": 0.0234, "step": 126750 }, { "grad_norm": 0.4300893545150757, "learning_rate": 3.239665393653919e-05, "loss": 0.0238, "step": 126760 }, { "grad_norm": 0.2950942814350128, "learning_rate": 3.23889161465489e-05, "loss": 0.0228, "step": 126770 }, { "grad_norm": 0.8499290347099304, "learning_rate": 3.238117883803898e-05, "loss": 0.0265, "step": 126780 }, { "grad_norm": 0.3312411308288574, "learning_rate": 3.237344201122103e-05, "loss": 0.0215, "step": 126790 }, { "grad_norm": 0.36091718077659607, "learning_rate": 3.236570566630656e-05, "loss": 0.0158, "step": 126800 }, { "grad_norm": 0.7867004871368408, "learning_rate": 3.235796980350707e-05, "loss": 0.0287, "step": 126810 }, { "grad_norm": 0.32949233055114746, "learning_rate": 3.235023442303406e-05, "loss": 0.0311, "step": 126820 }, { "grad_norm": 0.5121615529060364, "learning_rate": 3.2342499525099e-05, "loss": 0.0218, "step": 126830 }, { "grad_norm": 0.6203959584236145, "learning_rate": 3.233476510991338e-05, "loss": 0.0288, "step": 126840 }, { "grad_norm": 0.6425186395645142, "learning_rate": 3.2327031177688636e-05, "loss": 0.0228, "step": 126850 }, { "grad_norm": 0.44610220193862915, "learning_rate": 3.231929772863621e-05, "loss": 0.0239, "step": 126860 }, { "grad_norm": 0.8839399218559265, "learning_rate": 3.231156476296757e-05, "loss": 0.0279, "step": 126870 }, { "grad_norm": 0.5824087262153625, "learning_rate": 3.2303832280894084e-05, "loss": 0.0185, "step": 126880 }, { "grad_norm": 0.3065287172794342, "learning_rate": 3.2296100282627184e-05, "loss": 0.0195, "step": 126890 }, { "grad_norm": 0.4758397936820984, "learning_rate": 3.2288368768378254e-05, "loss": 0.0213, "step": 126900 }, { "grad_norm": 0.5473920106887817, "learning_rate": 3.2280637738358666e-05, "loss": 0.0216, "step": 126910 }, { "grad_norm": 0.5674634575843811, "learning_rate": 3.227290719277978e-05, "loss": 0.0354, "step": 126920 }, { "grad_norm": 0.32304492592811584, "learning_rate": 3.226517713185294e-05, "loss": 0.0237, "step": 126930 }, { "grad_norm": 0.5563019514083862, "learning_rate": 3.225744755578952e-05, "loss": 0.0172, "step": 126940 }, { "grad_norm": 0.3888576924800873, "learning_rate": 3.224971846480079e-05, "loss": 0.0259, "step": 126950 }, { "grad_norm": 0.35804253816604614, "learning_rate": 3.224198985909813e-05, "loss": 0.03, "step": 126960 }, { "grad_norm": 0.7021423578262329, "learning_rate": 3.2234261738892756e-05, "loss": 0.0304, "step": 126970 }, { "grad_norm": 1.4554017782211304, "learning_rate": 3.2226534104396e-05, "loss": 0.0338, "step": 126980 }, { "grad_norm": 0.5353716611862183, "learning_rate": 3.2218806955819144e-05, "loss": 0.0254, "step": 126990 }, { "grad_norm": 0.5479439496994019, "learning_rate": 3.221108029337341e-05, "loss": 0.0304, "step": 127000 }, { "grad_norm": 0.4013582170009613, "learning_rate": 3.220335411727007e-05, "loss": 0.0171, "step": 127010 }, { "grad_norm": 0.4180261194705963, "learning_rate": 3.219562842772033e-05, "loss": 0.0201, "step": 127020 }, { "grad_norm": 0.45263251662254333, "learning_rate": 3.2187903224935426e-05, "loss": 0.0398, "step": 127030 }, { "grad_norm": 0.5401597619056702, "learning_rate": 3.218017850912655e-05, "loss": 0.0282, "step": 127040 }, { "grad_norm": 0.3921225368976593, "learning_rate": 3.2172454280504886e-05, "loss": 0.0178, "step": 127050 }, { "grad_norm": 0.18781401216983795, "learning_rate": 3.216473053928166e-05, "loss": 0.0343, "step": 127060 }, { "grad_norm": 0.9218363165855408, "learning_rate": 3.215700728566797e-05, "loss": 0.0188, "step": 127070 }, { "grad_norm": 0.19263888895511627, "learning_rate": 3.214928451987501e-05, "loss": 0.0201, "step": 127080 }, { "grad_norm": 0.45900386571884155, "learning_rate": 3.2141562242113906e-05, "loss": 0.024, "step": 127090 }, { "grad_norm": 1.3978523015975952, "learning_rate": 3.213384045259579e-05, "loss": 0.0253, "step": 127100 }, { "grad_norm": 0.34323829412460327, "learning_rate": 3.2126119151531744e-05, "loss": 0.0324, "step": 127110 }, { "grad_norm": 0.3868299126625061, "learning_rate": 3.211839833913289e-05, "loss": 0.0242, "step": 127120 }, { "grad_norm": 0.9743503928184509, "learning_rate": 3.211067801561032e-05, "loss": 0.0358, "step": 127130 }, { "grad_norm": 0.5322613716125488, "learning_rate": 3.210295818117507e-05, "loss": 0.0294, "step": 127140 }, { "grad_norm": 0.30103787779808044, "learning_rate": 3.209523883603826e-05, "loss": 0.0135, "step": 127150 }, { "grad_norm": 0.6462944746017456, "learning_rate": 3.2087519980410854e-05, "loss": 0.0178, "step": 127160 }, { "grad_norm": 0.22768467664718628, "learning_rate": 3.2079801614503924e-05, "loss": 0.0378, "step": 127170 }, { "grad_norm": 0.26491573452949524, "learning_rate": 3.207208373852851e-05, "loss": 0.0216, "step": 127180 }, { "grad_norm": 0.21477510035037994, "learning_rate": 3.206436635269557e-05, "loss": 0.0188, "step": 127190 }, { "grad_norm": 0.4585874676704407, "learning_rate": 3.2056649457216135e-05, "loss": 0.0204, "step": 127200 }, { "grad_norm": 0.2609902322292328, "learning_rate": 3.204893305230115e-05, "loss": 0.0217, "step": 127210 }, { "grad_norm": 1.5435823202133179, "learning_rate": 3.2041217138161605e-05, "loss": 0.0356, "step": 127220 }, { "grad_norm": 0.6206128597259521, "learning_rate": 3.2033501715008416e-05, "loss": 0.0228, "step": 127230 }, { "grad_norm": 0.6271637678146362, "learning_rate": 3.202578678305254e-05, "loss": 0.0197, "step": 127240 }, { "grad_norm": 0.5056488513946533, "learning_rate": 3.201807234250493e-05, "loss": 0.0323, "step": 127250 }, { "grad_norm": 0.31822851300239563, "learning_rate": 3.201035839357643e-05, "loss": 0.0342, "step": 127260 }, { "grad_norm": 0.538180410861969, "learning_rate": 3.2002644936478e-05, "loss": 0.0196, "step": 127270 }, { "grad_norm": 0.28607261180877686, "learning_rate": 3.199493197142048e-05, "loss": 0.0247, "step": 127280 }, { "grad_norm": 0.29049187898635864, "learning_rate": 3.198721949861477e-05, "loss": 0.0276, "step": 127290 }, { "grad_norm": 0.35838624835014343, "learning_rate": 3.1979507518271705e-05, "loss": 0.0193, "step": 127300 }, { "grad_norm": 0.6059530377388, "learning_rate": 3.197179603060213e-05, "loss": 0.0266, "step": 127310 }, { "grad_norm": 0.5237556099891663, "learning_rate": 3.196408503581689e-05, "loss": 0.019, "step": 127320 }, { "grad_norm": 1.023625135421753, "learning_rate": 3.195637453412678e-05, "loss": 0.0232, "step": 127330 }, { "grad_norm": 0.305008202791214, "learning_rate": 3.194866452574262e-05, "loss": 0.0264, "step": 127340 }, { "grad_norm": 0.1808747798204422, "learning_rate": 3.194095501087517e-05, "loss": 0.0216, "step": 127350 }, { "grad_norm": 0.33781424164772034, "learning_rate": 3.193324598973524e-05, "loss": 0.0223, "step": 127360 }, { "grad_norm": 0.3955981135368347, "learning_rate": 3.192553746253358e-05, "loss": 0.0222, "step": 127370 }, { "grad_norm": 0.33935055136680603, "learning_rate": 3.191782942948093e-05, "loss": 0.048, "step": 127380 }, { "grad_norm": 0.40281301736831665, "learning_rate": 3.191012189078804e-05, "loss": 0.0362, "step": 127390 }, { "grad_norm": 0.4090995490550995, "learning_rate": 3.190241484666561e-05, "loss": 0.0356, "step": 127400 }, { "grad_norm": 0.45232725143432617, "learning_rate": 3.1894708297324373e-05, "loss": 0.0259, "step": 127410 }, { "grad_norm": 0.41870927810668945, "learning_rate": 3.188700224297499e-05, "loss": 0.0264, "step": 127420 }, { "grad_norm": 0.7286515235900879, "learning_rate": 3.1879296683828164e-05, "loss": 0.0289, "step": 127430 }, { "grad_norm": 0.43615925312042236, "learning_rate": 3.187159162009458e-05, "loss": 0.0185, "step": 127440 }, { "grad_norm": 0.25748851895332336, "learning_rate": 3.186388705198484e-05, "loss": 0.0193, "step": 127450 }, { "grad_norm": 0.7959884405136108, "learning_rate": 3.1856182979709646e-05, "loss": 0.0187, "step": 127460 }, { "grad_norm": 0.5197843313217163, "learning_rate": 3.184847940347957e-05, "loss": 0.021, "step": 127470 }, { "grad_norm": 0.4116049110889435, "learning_rate": 3.1840776323505264e-05, "loss": 0.018, "step": 127480 }, { "grad_norm": 0.36685866117477417, "learning_rate": 3.18330737399973e-05, "loss": 0.0232, "step": 127490 }, { "grad_norm": 0.6634200811386108, "learning_rate": 3.182537165316627e-05, "loss": 0.0209, "step": 127500 }, { "grad_norm": 0.7267566323280334, "learning_rate": 3.181767006322277e-05, "loss": 0.0286, "step": 127510 }, { "grad_norm": 0.7068060040473938, "learning_rate": 3.1809968970377324e-05, "loss": 0.0239, "step": 127520 }, { "grad_norm": 0.4021267890930176, "learning_rate": 3.1802268374840503e-05, "loss": 0.0133, "step": 127530 }, { "grad_norm": 0.4196840524673462, "learning_rate": 3.179456827682282e-05, "loss": 0.0254, "step": 127540 }, { "grad_norm": 0.41090962290763855, "learning_rate": 3.178686867653479e-05, "loss": 0.0249, "step": 127550 }, { "grad_norm": 0.28682956099510193, "learning_rate": 3.177916957418694e-05, "loss": 0.0339, "step": 127560 }, { "grad_norm": 0.7818017601966858, "learning_rate": 3.177147096998975e-05, "loss": 0.0208, "step": 127570 }, { "grad_norm": 0.47545966506004333, "learning_rate": 3.1763772864153705e-05, "loss": 0.0212, "step": 127580 }, { "grad_norm": 0.6903296113014221, "learning_rate": 3.175607525688925e-05, "loss": 0.0277, "step": 127590 }, { "grad_norm": 0.4438837468624115, "learning_rate": 3.174837814840684e-05, "loss": 0.0244, "step": 127600 }, { "grad_norm": 0.326126366853714, "learning_rate": 3.1740681538916914e-05, "loss": 0.026, "step": 127610 }, { "grad_norm": 0.7371900677680969, "learning_rate": 3.1732985428629905e-05, "loss": 0.0296, "step": 127620 }, { "grad_norm": 0.28644388914108276, "learning_rate": 3.172528981775621e-05, "loss": 0.0188, "step": 127630 }, { "grad_norm": 1.4507406949996948, "learning_rate": 3.17175947065062e-05, "loss": 0.033, "step": 127640 }, { "grad_norm": 0.3230453431606293, "learning_rate": 3.170990009509033e-05, "loss": 0.0318, "step": 127650 }, { "grad_norm": 0.40899860858917236, "learning_rate": 3.1702205983718874e-05, "loss": 0.0176, "step": 127660 }, { "grad_norm": 1.3286590576171875, "learning_rate": 3.169451237260227e-05, "loss": 0.024, "step": 127670 }, { "grad_norm": 0.8287908434867859, "learning_rate": 3.168681926195079e-05, "loss": 0.028, "step": 127680 }, { "grad_norm": 0.5213775634765625, "learning_rate": 3.16791266519748e-05, "loss": 0.0315, "step": 127690 }, { "grad_norm": 0.3912901282310486, "learning_rate": 3.1671434542884615e-05, "loss": 0.0212, "step": 127700 }, { "grad_norm": 0.44412732124328613, "learning_rate": 3.166374293489052e-05, "loss": 0.0186, "step": 127710 }, { "grad_norm": 0.9657526612281799, "learning_rate": 3.1656051828202804e-05, "loss": 0.0244, "step": 127720 }, { "grad_norm": 0.23886598646640778, "learning_rate": 3.164836122303173e-05, "loss": 0.0185, "step": 127730 }, { "grad_norm": 0.5697342753410339, "learning_rate": 3.1640671119587565e-05, "loss": 0.0285, "step": 127740 }, { "grad_norm": 0.6888399124145508, "learning_rate": 3.1632981518080575e-05, "loss": 0.0275, "step": 127750 }, { "grad_norm": 0.4320600628852844, "learning_rate": 3.162529241872094e-05, "loss": 0.0274, "step": 127760 }, { "grad_norm": 0.37259453535079956, "learning_rate": 3.1617603821718936e-05, "loss": 0.0251, "step": 127770 }, { "grad_norm": 0.7415457963943481, "learning_rate": 3.160991572728472e-05, "loss": 0.0236, "step": 127780 }, { "grad_norm": 0.3620489537715912, "learning_rate": 3.160222813562852e-05, "loss": 0.0173, "step": 127790 }, { "grad_norm": 0.4372328221797943, "learning_rate": 3.159454104696047e-05, "loss": 0.0157, "step": 127800 }, { "grad_norm": 0.3373115062713623, "learning_rate": 3.158685446149076e-05, "loss": 0.0343, "step": 127810 }, { "grad_norm": 0.9679718613624573, "learning_rate": 3.157916837942954e-05, "loss": 0.0234, "step": 127820 }, { "grad_norm": 0.44016188383102417, "learning_rate": 3.157148280098692e-05, "loss": 0.0248, "step": 127830 }, { "grad_norm": 0.39494428038597107, "learning_rate": 3.156379772637305e-05, "loss": 0.0321, "step": 127840 }, { "grad_norm": 0.46544384956359863, "learning_rate": 3.1556113155798006e-05, "loss": 0.0201, "step": 127850 }, { "grad_norm": 2.6699910163879395, "learning_rate": 3.1548429089471925e-05, "loss": 0.0304, "step": 127860 }, { "grad_norm": 0.4049639403820038, "learning_rate": 3.154074552760484e-05, "loss": 0.0179, "step": 127870 }, { "grad_norm": 0.38669079542160034, "learning_rate": 3.1533062470406836e-05, "loss": 0.0248, "step": 127880 }, { "grad_norm": 0.42945167422294617, "learning_rate": 3.1525379918087984e-05, "loss": 0.0218, "step": 127890 }, { "grad_norm": 0.6280928254127502, "learning_rate": 3.151769787085829e-05, "loss": 0.0258, "step": 127900 }, { "grad_norm": 0.30710461735725403, "learning_rate": 3.151001632892781e-05, "loss": 0.0242, "step": 127910 }, { "grad_norm": 0.4139852523803711, "learning_rate": 3.150233529250652e-05, "loss": 0.0208, "step": 127920 }, { "grad_norm": 0.3277983069419861, "learning_rate": 3.149465476180442e-05, "loss": 0.0242, "step": 127930 }, { "grad_norm": 0.7536188364028931, "learning_rate": 3.148697473703155e-05, "loss": 0.0154, "step": 127940 }, { "grad_norm": 0.9606106281280518, "learning_rate": 3.14792952183978e-05, "loss": 0.032, "step": 127950 }, { "grad_norm": 0.18691977858543396, "learning_rate": 3.1471616206113185e-05, "loss": 0.0217, "step": 127960 }, { "grad_norm": 0.2793366611003876, "learning_rate": 3.14639377003876e-05, "loss": 0.023, "step": 127970 }, { "grad_norm": 0.972171425819397, "learning_rate": 3.1456259701431006e-05, "loss": 0.0237, "step": 127980 }, { "grad_norm": 0.5749902725219727, "learning_rate": 3.1448582209453304e-05, "loss": 0.0185, "step": 127990 }, { "grad_norm": 0.4567693769931793, "learning_rate": 3.144090522466438e-05, "loss": 0.0182, "step": 128000 }, { "grad_norm": 0.7461203336715698, "learning_rate": 3.143322874727417e-05, "loss": 0.0191, "step": 128010 }, { "grad_norm": 0.518928050994873, "learning_rate": 3.1425552777492486e-05, "loss": 0.0223, "step": 128020 }, { "grad_norm": 0.7904971241950989, "learning_rate": 3.141787731552922e-05, "loss": 0.0229, "step": 128030 }, { "grad_norm": 0.4257368743419647, "learning_rate": 3.1410202361594196e-05, "loss": 0.0192, "step": 128040 }, { "grad_norm": 0.39156705141067505, "learning_rate": 3.140252791589727e-05, "loss": 0.0213, "step": 128050 }, { "grad_norm": 0.6522127985954285, "learning_rate": 3.139485397864822e-05, "loss": 0.023, "step": 128060 }, { "grad_norm": 0.4855394959449768, "learning_rate": 3.138718055005689e-05, "loss": 0.0192, "step": 128070 }, { "grad_norm": 0.4686092436313629, "learning_rate": 3.137950763033305e-05, "loss": 0.0211, "step": 128080 }, { "grad_norm": 0.8613561391830444, "learning_rate": 3.137183521968647e-05, "loss": 0.0164, "step": 128090 }, { "grad_norm": 0.6210594177246094, "learning_rate": 3.1364163318326934e-05, "loss": 0.0178, "step": 128100 }, { "grad_norm": 0.30715128779411316, "learning_rate": 3.135649192646416e-05, "loss": 0.0211, "step": 128110 }, { "grad_norm": 0.4880869388580322, "learning_rate": 3.13488210443079e-05, "loss": 0.0191, "step": 128120 }, { "grad_norm": 0.9950463175773621, "learning_rate": 3.134115067206786e-05, "loss": 0.0194, "step": 128130 }, { "grad_norm": 0.374211847782135, "learning_rate": 3.133348080995373e-05, "loss": 0.0208, "step": 128140 }, { "grad_norm": 0.45389917492866516, "learning_rate": 3.132581145817526e-05, "loss": 0.0233, "step": 128150 }, { "grad_norm": 0.3306991755962372, "learning_rate": 3.131814261694206e-05, "loss": 0.0309, "step": 128160 }, { "grad_norm": 0.45953667163848877, "learning_rate": 3.131047428646384e-05, "loss": 0.0228, "step": 128170 }, { "grad_norm": 0.3798445761203766, "learning_rate": 3.130280646695022e-05, "loss": 0.0183, "step": 128180 }, { "grad_norm": 0.3444133698940277, "learning_rate": 3.129513915861084e-05, "loss": 0.0192, "step": 128190 }, { "grad_norm": 0.2958352565765381, "learning_rate": 3.128747236165535e-05, "loss": 0.0213, "step": 128200 }, { "grad_norm": 0.46513333916664124, "learning_rate": 3.127980607629333e-05, "loss": 0.0192, "step": 128210 }, { "grad_norm": 0.5127432346343994, "learning_rate": 3.127214030273438e-05, "loss": 0.0163, "step": 128220 }, { "grad_norm": 0.3307650685310364, "learning_rate": 3.126447504118807e-05, "loss": 0.0251, "step": 128230 }, { "grad_norm": 0.8877321481704712, "learning_rate": 3.125681029186398e-05, "loss": 0.0251, "step": 128240 }, { "grad_norm": 0.7657520174980164, "learning_rate": 3.124914605497164e-05, "loss": 0.0343, "step": 128250 }, { "grad_norm": 0.3217928409576416, "learning_rate": 3.1241482330720606e-05, "loss": 0.0261, "step": 128260 }, { "grad_norm": 0.43575868010520935, "learning_rate": 3.123381911932042e-05, "loss": 0.0232, "step": 128270 }, { "grad_norm": 0.31824976205825806, "learning_rate": 3.122615642098055e-05, "loss": 0.0197, "step": 128280 }, { "grad_norm": 0.5552522540092468, "learning_rate": 3.1218494235910524e-05, "loss": 0.0262, "step": 128290 }, { "grad_norm": 0.5706861019134521, "learning_rate": 3.1210832564319794e-05, "loss": 0.0321, "step": 128300 }, { "grad_norm": 0.3361353576183319, "learning_rate": 3.120317140641786e-05, "loss": 0.0208, "step": 128310 }, { "grad_norm": 1.1533714532852173, "learning_rate": 3.119551076241415e-05, "loss": 0.021, "step": 128320 }, { "grad_norm": 0.9810850024223328, "learning_rate": 3.11878506325181e-05, "loss": 0.0275, "step": 128330 }, { "grad_norm": 0.40465810894966125, "learning_rate": 3.118019101693917e-05, "loss": 0.0312, "step": 128340 }, { "grad_norm": 0.3879747986793518, "learning_rate": 3.1172531915886716e-05, "loss": 0.0312, "step": 128350 }, { "grad_norm": 0.4071570634841919, "learning_rate": 3.1164873329570196e-05, "loss": 0.0264, "step": 128360 }, { "grad_norm": 0.6580514907836914, "learning_rate": 3.115721525819894e-05, "loss": 0.0201, "step": 128370 }, { "grad_norm": 3.120204448699951, "learning_rate": 3.114955770198234e-05, "loss": 0.025, "step": 128380 }, { "grad_norm": 0.2281816452741623, "learning_rate": 3.1141900661129775e-05, "loss": 0.0269, "step": 128390 }, { "grad_norm": 0.48010575771331787, "learning_rate": 3.1134244135850544e-05, "loss": 0.0217, "step": 128400 }, { "grad_norm": 0.43946483731269836, "learning_rate": 3.112658812635401e-05, "loss": 0.0276, "step": 128410 }, { "grad_norm": 0.2690913677215576, "learning_rate": 3.111893263284944e-05, "loss": 0.0236, "step": 128420 }, { "grad_norm": 0.5370691418647766, "learning_rate": 3.111127765554618e-05, "loss": 0.0155, "step": 128430 }, { "grad_norm": 0.4958869218826294, "learning_rate": 3.110362319465348e-05, "loss": 0.0224, "step": 128440 }, { "grad_norm": 0.3117777705192566, "learning_rate": 3.1095969250380605e-05, "loss": 0.0275, "step": 128450 }, { "grad_norm": 0.4155365824699402, "learning_rate": 3.1088315822936865e-05, "loss": 0.0233, "step": 128460 }, { "grad_norm": 0.6735824942588806, "learning_rate": 3.108066291253144e-05, "loss": 0.0239, "step": 128470 }, { "grad_norm": 0.3406924903392792, "learning_rate": 3.10730105193736e-05, "loss": 0.0165, "step": 128480 }, { "grad_norm": 0.39270398020744324, "learning_rate": 3.106535864367253e-05, "loss": 0.0218, "step": 128490 }, { "grad_norm": 0.5340511798858643, "learning_rate": 3.105770728563745e-05, "loss": 0.0222, "step": 128500 }, { "grad_norm": 0.32652992010116577, "learning_rate": 3.1050056445477526e-05, "loss": 0.0173, "step": 128510 }, { "grad_norm": 0.53807133436203, "learning_rate": 3.104240612340195e-05, "loss": 0.0168, "step": 128520 }, { "grad_norm": 0.46058276295661926, "learning_rate": 3.103475631961987e-05, "loss": 0.0262, "step": 128530 }, { "grad_norm": 0.42516028881073, "learning_rate": 3.102710703434042e-05, "loss": 0.0276, "step": 128540 }, { "grad_norm": 0.5730499029159546, "learning_rate": 3.101945826777274e-05, "loss": 0.0262, "step": 128550 }, { "grad_norm": 0.6663663387298584, "learning_rate": 3.1011810020125924e-05, "loss": 0.0431, "step": 128560 }, { "grad_norm": 0.8293857574462891, "learning_rate": 3.10041622916091e-05, "loss": 0.0295, "step": 128570 }, { "grad_norm": 1.4493507146835327, "learning_rate": 3.099651508243136e-05, "loss": 0.0245, "step": 128580 }, { "grad_norm": 0.449470192193985, "learning_rate": 3.098886839280173e-05, "loss": 0.0322, "step": 128590 }, { "grad_norm": 1.6462482213974, "learning_rate": 3.0981222222929315e-05, "loss": 0.0204, "step": 128600 }, { "grad_norm": 0.415825217962265, "learning_rate": 3.0973576573023134e-05, "loss": 0.0222, "step": 128610 }, { "grad_norm": 0.6703187823295593, "learning_rate": 3.0965931443292226e-05, "loss": 0.0238, "step": 128620 }, { "grad_norm": 0.376840204000473, "learning_rate": 3.095828683394559e-05, "loss": 0.0253, "step": 128630 }, { "grad_norm": 0.6310621500015259, "learning_rate": 3.0950642745192234e-05, "loss": 0.0376, "step": 128640 }, { "grad_norm": 0.3721957504749298, "learning_rate": 3.0942999177241186e-05, "loss": 0.034, "step": 128650 }, { "grad_norm": 1.6851054430007935, "learning_rate": 3.0935356130301336e-05, "loss": 0.0284, "step": 128660 }, { "grad_norm": 1.3113733530044556, "learning_rate": 3.092771360458172e-05, "loss": 0.0245, "step": 128670 }, { "grad_norm": 1.237754464149475, "learning_rate": 3.092007160029122e-05, "loss": 0.0314, "step": 128680 }, { "grad_norm": 0.24645587801933289, "learning_rate": 3.0912430117638824e-05, "loss": 0.0241, "step": 128690 }, { "grad_norm": 1.6076358556747437, "learning_rate": 3.0904789156833394e-05, "loss": 0.0187, "step": 128700 }, { "grad_norm": 0.5298351645469666, "learning_rate": 3.089714871808387e-05, "loss": 0.0214, "step": 128710 }, { "grad_norm": 0.329875648021698, "learning_rate": 3.088950880159912e-05, "loss": 0.0242, "step": 128720 }, { "grad_norm": 0.24260665476322174, "learning_rate": 3.0881869407588014e-05, "loss": 0.0224, "step": 128730 }, { "grad_norm": 0.27703359723091125, "learning_rate": 3.087423053625943e-05, "loss": 0.018, "step": 128740 }, { "grad_norm": 0.392395555973053, "learning_rate": 3.086659218782218e-05, "loss": 0.0228, "step": 128750 }, { "grad_norm": 1.4197808504104614, "learning_rate": 3.0858954362485105e-05, "loss": 0.0319, "step": 128760 }, { "grad_norm": 0.35854434967041016, "learning_rate": 3.0851317060457045e-05, "loss": 0.0208, "step": 128770 }, { "grad_norm": 0.3354211449623108, "learning_rate": 3.084368028194677e-05, "loss": 0.0199, "step": 128780 }, { "grad_norm": 0.8560556769371033, "learning_rate": 3.08360440271631e-05, "loss": 0.019, "step": 128790 }, { "grad_norm": 0.29411137104034424, "learning_rate": 3.082840829631477e-05, "loss": 0.0178, "step": 128800 }, { "grad_norm": 0.42042386531829834, "learning_rate": 3.0820773089610564e-05, "loss": 0.0265, "step": 128810 }, { "grad_norm": 3.2511959075927734, "learning_rate": 3.081313840725921e-05, "loss": 0.0213, "step": 128820 }, { "grad_norm": 6.021826267242432, "learning_rate": 3.0805504249469444e-05, "loss": 0.0579, "step": 128830 }, { "grad_norm": 0.32783591747283936, "learning_rate": 3.079787061644999e-05, "loss": 0.0161, "step": 128840 }, { "grad_norm": 0.5766112208366394, "learning_rate": 3.079023750840952e-05, "loss": 0.0215, "step": 128850 }, { "grad_norm": 0.5912038683891296, "learning_rate": 3.078260492555677e-05, "loss": 0.0279, "step": 128860 }, { "grad_norm": 0.3616774082183838, "learning_rate": 3.077497286810035e-05, "loss": 0.0246, "step": 128870 }, { "grad_norm": 0.6390633583068848, "learning_rate": 3.076734133624898e-05, "loss": 0.0214, "step": 128880 }, { "grad_norm": 0.39977824687957764, "learning_rate": 3.075971033021124e-05, "loss": 0.0245, "step": 128890 }, { "grad_norm": 0.43503573536872864, "learning_rate": 3.0752079850195805e-05, "loss": 0.0245, "step": 128900 }, { "grad_norm": 0.6332451701164246, "learning_rate": 3.074444989641128e-05, "loss": 0.0215, "step": 128910 }, { "grad_norm": 0.4920494258403778, "learning_rate": 3.0736820469066255e-05, "loss": 0.0214, "step": 128920 }, { "grad_norm": 0.4797636568546295, "learning_rate": 3.072919156836933e-05, "loss": 0.0238, "step": 128930 }, { "grad_norm": 0.4925305247306824, "learning_rate": 3.072156319452906e-05, "loss": 0.0242, "step": 128940 }, { "grad_norm": 0.26120874285697937, "learning_rate": 3.0713935347754e-05, "loss": 0.0265, "step": 128950 }, { "grad_norm": 0.848124623298645, "learning_rate": 3.0706308028252726e-05, "loss": 0.0198, "step": 128960 }, { "grad_norm": 2.6353671550750732, "learning_rate": 3.0698681236233714e-05, "loss": 0.0281, "step": 128970 }, { "grad_norm": 0.3484167754650116, "learning_rate": 3.0691054971905535e-05, "loss": 0.019, "step": 128980 }, { "grad_norm": 0.4257907271385193, "learning_rate": 3.068342923547665e-05, "loss": 0.0201, "step": 128990 }, { "grad_norm": 0.4555818736553192, "learning_rate": 3.067580402715555e-05, "loss": 0.0195, "step": 129000 }, { "grad_norm": 0.26045408844947815, "learning_rate": 3.06681793471507e-05, "loss": 0.0246, "step": 129010 }, { "grad_norm": 0.35378071665763855, "learning_rate": 3.066055519567057e-05, "loss": 0.0197, "step": 129020 }, { "grad_norm": 0.2605869174003601, "learning_rate": 3.06529315729236e-05, "loss": 0.0238, "step": 129030 }, { "grad_norm": 0.3811314105987549, "learning_rate": 3.064530847911821e-05, "loss": 0.0314, "step": 129040 }, { "grad_norm": 0.5855749249458313, "learning_rate": 3.063768591446282e-05, "loss": 0.0288, "step": 129050 }, { "grad_norm": 0.4783664345741272, "learning_rate": 3.06300638791658e-05, "loss": 0.0297, "step": 129060 }, { "grad_norm": 0.302859902381897, "learning_rate": 3.06224423734356e-05, "loss": 0.0244, "step": 129070 }, { "grad_norm": 0.3036821484565735, "learning_rate": 3.06148213974805e-05, "loss": 0.0294, "step": 129080 }, { "grad_norm": 0.46763354539871216, "learning_rate": 3.060720095150892e-05, "loss": 0.0225, "step": 129090 }, { "grad_norm": 0.6015490293502808, "learning_rate": 3.0599581035729184e-05, "loss": 0.0241, "step": 129100 }, { "grad_norm": 1.3076386451721191, "learning_rate": 3.059196165034961e-05, "loss": 0.0282, "step": 129110 }, { "grad_norm": 0.3161262273788452, "learning_rate": 3.058434279557852e-05, "loss": 0.0214, "step": 129120 }, { "grad_norm": 0.3519665598869324, "learning_rate": 3.057672447162419e-05, "loss": 0.0228, "step": 129130 }, { "grad_norm": 0.466778427362442, "learning_rate": 3.0569106678694915e-05, "loss": 0.0364, "step": 129140 }, { "grad_norm": 0.4231880307197571, "learning_rate": 3.056148941699899e-05, "loss": 0.0264, "step": 129150 }, { "grad_norm": 0.5619151592254639, "learning_rate": 3.055387268674461e-05, "loss": 0.0295, "step": 129160 }, { "grad_norm": 0.3324708044528961, "learning_rate": 3.054625648814007e-05, "loss": 0.0202, "step": 129170 }, { "grad_norm": 0.4399178922176361, "learning_rate": 3.053864082139355e-05, "loss": 0.0269, "step": 129180 }, { "grad_norm": 0.6956555247306824, "learning_rate": 3.053102568671329e-05, "loss": 0.0232, "step": 129190 }, { "grad_norm": 0.44392579793930054, "learning_rate": 3.052341108430746e-05, "loss": 0.0262, "step": 129200 }, { "grad_norm": 0.1661599725484848, "learning_rate": 3.051579701438426e-05, "loss": 0.0364, "step": 129210 }, { "grad_norm": 0.5448340177536011, "learning_rate": 3.0508183477151864e-05, "loss": 0.0271, "step": 129220 }, { "grad_norm": 0.4774165749549866, "learning_rate": 3.0500570472818398e-05, "loss": 0.0207, "step": 129230 }, { "grad_norm": 0.5566266179084778, "learning_rate": 3.049295800159202e-05, "loss": 0.035, "step": 129240 }, { "grad_norm": 0.5634211897850037, "learning_rate": 3.0485346063680825e-05, "loss": 0.027, "step": 129250 }, { "grad_norm": 0.34964069724082947, "learning_rate": 3.0477734659292957e-05, "loss": 0.0268, "step": 129260 }, { "grad_norm": 0.4626467227935791, "learning_rate": 3.0470123788636457e-05, "loss": 0.0251, "step": 129270 }, { "grad_norm": 0.35929885506629944, "learning_rate": 3.046251345191945e-05, "loss": 0.0252, "step": 129280 }, { "grad_norm": 0.5067981481552124, "learning_rate": 3.0454903649349996e-05, "loss": 0.0215, "step": 129290 }, { "grad_norm": 0.7785513401031494, "learning_rate": 3.0447294381136128e-05, "loss": 0.0272, "step": 129300 }, { "grad_norm": 0.21392056345939636, "learning_rate": 3.0439685647485895e-05, "loss": 0.0238, "step": 129310 }, { "grad_norm": 0.47558265924453735, "learning_rate": 3.0432077448607294e-05, "loss": 0.0205, "step": 129320 }, { "grad_norm": 0.2730603814125061, "learning_rate": 3.0424469784708353e-05, "loss": 0.0217, "step": 129330 }, { "grad_norm": 0.47636091709136963, "learning_rate": 3.0416862655997046e-05, "loss": 0.0204, "step": 129340 }, { "grad_norm": 0.7222682237625122, "learning_rate": 3.0409256062681353e-05, "loss": 0.0237, "step": 129350 }, { "grad_norm": 0.3948765993118286, "learning_rate": 3.040165000496926e-05, "loss": 0.0214, "step": 129360 }, { "grad_norm": 0.33509236574172974, "learning_rate": 3.0394044483068674e-05, "loss": 0.0213, "step": 129370 }, { "grad_norm": 0.49507102370262146, "learning_rate": 3.038643949718757e-05, "loss": 0.04, "step": 129380 }, { "grad_norm": 0.4702315032482147, "learning_rate": 3.0378835047533805e-05, "loss": 0.0331, "step": 129390 }, { "grad_norm": 0.19488419592380524, "learning_rate": 3.0371231134315335e-05, "loss": 0.0231, "step": 129400 }, { "grad_norm": 0.3620550036430359, "learning_rate": 3.0363627757740044e-05, "loss": 0.0222, "step": 129410 }, { "grad_norm": 0.24939027428627014, "learning_rate": 3.035602491801578e-05, "loss": 0.0185, "step": 129420 }, { "grad_norm": 0.5624231100082397, "learning_rate": 3.0348422615350425e-05, "loss": 0.0277, "step": 129430 }, { "grad_norm": 0.27522560954093933, "learning_rate": 3.03408208499518e-05, "loss": 0.026, "step": 129440 }, { "grad_norm": 0.4124748110771179, "learning_rate": 3.0333219622027764e-05, "loss": 0.0242, "step": 129450 }, { "grad_norm": 0.4675120711326599, "learning_rate": 3.0325618931786104e-05, "loss": 0.0229, "step": 129460 }, { "grad_norm": 0.5130429863929749, "learning_rate": 3.0318018779434615e-05, "loss": 0.0326, "step": 129470 }, { "grad_norm": 0.4949721395969391, "learning_rate": 3.0310419165181124e-05, "loss": 0.0286, "step": 129480 }, { "grad_norm": 0.4329153001308441, "learning_rate": 3.0302820089233364e-05, "loss": 0.0231, "step": 129490 }, { "grad_norm": 0.5440178513526917, "learning_rate": 3.0295221551799114e-05, "loss": 0.0198, "step": 129500 }, { "grad_norm": 0.26195207238197327, "learning_rate": 3.0287623553086088e-05, "loss": 0.0242, "step": 129510 }, { "grad_norm": 0.46187636256217957, "learning_rate": 3.0280026093302044e-05, "loss": 0.0239, "step": 129520 }, { "grad_norm": 0.3092185854911804, "learning_rate": 3.027242917265467e-05, "loss": 0.0195, "step": 129530 }, { "grad_norm": 0.40584030747413635, "learning_rate": 3.0264832791351667e-05, "loss": 0.0224, "step": 129540 }, { "grad_norm": 0.48538675904273987, "learning_rate": 3.025723694960073e-05, "loss": 0.017, "step": 129550 }, { "grad_norm": 0.7872238159179688, "learning_rate": 3.0249641647609507e-05, "loss": 0.028, "step": 129560 }, { "grad_norm": 0.36948174238204956, "learning_rate": 3.0242046885585684e-05, "loss": 0.0297, "step": 129570 }, { "grad_norm": 0.7379759550094604, "learning_rate": 3.0234452663736846e-05, "loss": 0.0411, "step": 129580 }, { "grad_norm": 0.27764585614204407, "learning_rate": 3.0226858982270657e-05, "loss": 0.0202, "step": 129590 }, { "grad_norm": 0.27652236819267273, "learning_rate": 3.0219265841394728e-05, "loss": 0.0183, "step": 129600 }, { "grad_norm": 0.30458077788352966, "learning_rate": 3.021167324131663e-05, "loss": 0.0226, "step": 129610 }, { "grad_norm": 0.4923683702945709, "learning_rate": 3.020408118224396e-05, "loss": 0.0278, "step": 129620 }, { "grad_norm": 0.9166706204414368, "learning_rate": 3.019648966438426e-05, "loss": 0.026, "step": 129630 }, { "grad_norm": 0.3942982256412506, "learning_rate": 3.0188898687945106e-05, "loss": 0.0151, "step": 129640 }, { "grad_norm": 0.3985273540019989, "learning_rate": 3.018130825313401e-05, "loss": 0.0206, "step": 129650 }, { "grad_norm": 1.2991989850997925, "learning_rate": 3.017371836015849e-05, "loss": 0.0302, "step": 129660 }, { "grad_norm": 0.23350633680820465, "learning_rate": 3.0166129009226085e-05, "loss": 0.0214, "step": 129670 }, { "grad_norm": 0.23047178983688354, "learning_rate": 3.0158540200544238e-05, "loss": 0.0193, "step": 129680 }, { "grad_norm": 0.3390565812587738, "learning_rate": 3.015095193432047e-05, "loss": 0.0235, "step": 129690 }, { "grad_norm": 0.24683013558387756, "learning_rate": 3.0143364210762202e-05, "loss": 0.0189, "step": 129700 }, { "grad_norm": 0.35241132974624634, "learning_rate": 3.0135777030076918e-05, "loss": 0.0215, "step": 129710 }, { "grad_norm": 0.4651627838611603, "learning_rate": 3.012819039247201e-05, "loss": 0.0237, "step": 129720 }, { "grad_norm": 0.8957154154777527, "learning_rate": 3.0120604298154908e-05, "loss": 0.0394, "step": 129730 }, { "grad_norm": 0.24272571504116058, "learning_rate": 3.0113018747333033e-05, "loss": 0.0275, "step": 129740 }, { "grad_norm": 0.5017423629760742, "learning_rate": 3.010543374021374e-05, "loss": 0.0231, "step": 129750 }, { "grad_norm": 0.33638158440589905, "learning_rate": 3.009784927700443e-05, "loss": 0.0228, "step": 129760 }, { "grad_norm": 0.3585851192474365, "learning_rate": 3.0090265357912416e-05, "loss": 0.0262, "step": 129770 }, { "grad_norm": 0.669248104095459, "learning_rate": 3.0082681983145084e-05, "loss": 0.0213, "step": 129780 }, { "grad_norm": 0.5771088600158691, "learning_rate": 3.007509915290975e-05, "loss": 0.0206, "step": 129790 }, { "grad_norm": 0.32396507263183594, "learning_rate": 3.0067516867413715e-05, "loss": 0.024, "step": 129800 }, { "grad_norm": 0.30631014704704285, "learning_rate": 3.005993512686428e-05, "loss": 0.0271, "step": 129810 }, { "grad_norm": 0.5053199529647827, "learning_rate": 3.0052353931468724e-05, "loss": 0.0317, "step": 129820 }, { "grad_norm": 0.35139280557632446, "learning_rate": 3.004477328143433e-05, "loss": 0.0271, "step": 129830 }, { "grad_norm": 0.4708685278892517, "learning_rate": 3.0037193176968326e-05, "loss": 0.0206, "step": 129840 }, { "grad_norm": 0.45476779341697693, "learning_rate": 3.002961361827795e-05, "loss": 0.0222, "step": 129850 }, { "grad_norm": 0.3817902207374573, "learning_rate": 3.0022034605570466e-05, "loss": 0.02, "step": 129860 }, { "grad_norm": 0.28697827458381653, "learning_rate": 3.0014456139053016e-05, "loss": 0.0168, "step": 129870 }, { "grad_norm": 0.590635359287262, "learning_rate": 3.0006878218932866e-05, "loss": 0.024, "step": 129880 }, { "grad_norm": 0.4817083179950714, "learning_rate": 2.9999300845417116e-05, "loss": 0.0219, "step": 129890 }, { "grad_norm": 0.8265025019645691, "learning_rate": 2.9991724018712993e-05, "loss": 0.023, "step": 129900 }, { "grad_norm": 0.5211712718009949, "learning_rate": 2.99841477390276e-05, "loss": 0.0311, "step": 129910 }, { "grad_norm": 0.24006035923957825, "learning_rate": 2.9976572006568094e-05, "loss": 0.0316, "step": 129920 }, { "grad_norm": 0.5609967708587646, "learning_rate": 2.9968996821541602e-05, "loss": 0.0301, "step": 129930 }, { "grad_norm": 0.38009151816368103, "learning_rate": 2.9961422184155196e-05, "loss": 0.0343, "step": 129940 }, { "grad_norm": 0.6987277865409851, "learning_rate": 2.9953848094615988e-05, "loss": 0.0254, "step": 129950 }, { "grad_norm": 0.7231366634368896, "learning_rate": 2.9946274553131038e-05, "loss": 0.0283, "step": 129960 }, { "grad_norm": 0.38310107588768005, "learning_rate": 2.993870155990739e-05, "loss": 0.0173, "step": 129970 }, { "grad_norm": 0.26004666090011597, "learning_rate": 2.9931129115152128e-05, "loss": 0.0231, "step": 129980 }, { "grad_norm": 0.43352439999580383, "learning_rate": 2.9923557219072246e-05, "loss": 0.0202, "step": 129990 }, { "grad_norm": 0.440700501203537, "learning_rate": 2.991598587187478e-05, "loss": 0.0237, "step": 130000 }, { "grad_norm": 0.48619306087493896, "learning_rate": 2.9908415073766704e-05, "loss": 0.0298, "step": 130010 }, { "grad_norm": 0.320678174495697, "learning_rate": 2.990084482495502e-05, "loss": 0.021, "step": 130020 }, { "grad_norm": 0.23034720122814178, "learning_rate": 2.989327512564668e-05, "loss": 0.0231, "step": 130030 }, { "grad_norm": 0.3296123445034027, "learning_rate": 2.988570597604864e-05, "loss": 0.0239, "step": 130040 }, { "grad_norm": 0.3906027674674988, "learning_rate": 2.9878137376367854e-05, "loss": 0.0225, "step": 130050 }, { "grad_norm": 0.6999081373214722, "learning_rate": 2.9870569326811217e-05, "loss": 0.0233, "step": 130060 }, { "grad_norm": 0.8860864043235779, "learning_rate": 2.986300182758568e-05, "loss": 0.0237, "step": 130070 }, { "grad_norm": 2.1527979373931885, "learning_rate": 2.985543487889807e-05, "loss": 0.0268, "step": 130080 }, { "grad_norm": 0.7009803056716919, "learning_rate": 2.9847868480955345e-05, "loss": 0.0265, "step": 130090 }, { "grad_norm": 0.2570870518684387, "learning_rate": 2.984030263396429e-05, "loss": 0.0181, "step": 130100 }, { "grad_norm": 0.43868762254714966, "learning_rate": 2.983273733813179e-05, "loss": 0.0285, "step": 130110 }, { "grad_norm": 0.3625811040401459, "learning_rate": 2.9825172593664695e-05, "loss": 0.0191, "step": 130120 }, { "grad_norm": 0.4923836290836334, "learning_rate": 2.9817608400769786e-05, "loss": 0.0302, "step": 130130 }, { "grad_norm": 0.183823361992836, "learning_rate": 2.9810044759653894e-05, "loss": 0.0179, "step": 130140 }, { "grad_norm": 0.8145902752876282, "learning_rate": 2.9802481670523786e-05, "loss": 0.0177, "step": 130150 }, { "grad_norm": 0.3208785355091095, "learning_rate": 2.9794919133586224e-05, "loss": 0.0138, "step": 130160 }, { "grad_norm": 0.6282291412353516, "learning_rate": 2.9787357149048018e-05, "loss": 0.0183, "step": 130170 }, { "grad_norm": 0.24659110605716705, "learning_rate": 2.9779795717115842e-05, "loss": 0.0224, "step": 130180 }, { "grad_norm": 0.24584808945655823, "learning_rate": 2.977223483799647e-05, "loss": 0.0168, "step": 130190 }, { "grad_norm": 3.396437168121338, "learning_rate": 2.9764674511896595e-05, "loss": 0.0242, "step": 130200 }, { "grad_norm": 0.5058797597885132, "learning_rate": 2.975711473902293e-05, "loss": 0.0155, "step": 130210 }, { "grad_norm": 0.2703879773616791, "learning_rate": 2.974955551958213e-05, "loss": 0.0198, "step": 130220 }, { "grad_norm": 0.34636756777763367, "learning_rate": 2.974199685378087e-05, "loss": 0.0167, "step": 130230 }, { "grad_norm": 0.4433784782886505, "learning_rate": 2.9734438741825816e-05, "loss": 0.038, "step": 130240 }, { "grad_norm": 0.7945601344108582, "learning_rate": 2.9726881183923582e-05, "loss": 0.0345, "step": 130250 }, { "grad_norm": 0.5421428084373474, "learning_rate": 2.9719324180280816e-05, "loss": 0.0245, "step": 130260 }, { "grad_norm": 0.4916900098323822, "learning_rate": 2.9711767731104083e-05, "loss": 0.0375, "step": 130270 }, { "grad_norm": 0.43459850549697876, "learning_rate": 2.970421183660003e-05, "loss": 0.0176, "step": 130280 }, { "grad_norm": 0.45801007747650146, "learning_rate": 2.969665649697517e-05, "loss": 0.0225, "step": 130290 }, { "grad_norm": 0.5169907808303833, "learning_rate": 2.9689101712436108e-05, "loss": 0.0224, "step": 130300 }, { "grad_norm": 0.4546600878238678, "learning_rate": 2.968154748318938e-05, "loss": 0.0172, "step": 130310 }, { "grad_norm": 0.384724497795105, "learning_rate": 2.96739938094415e-05, "loss": 0.0141, "step": 130320 }, { "grad_norm": 0.45007097721099854, "learning_rate": 2.9666440691399012e-05, "loss": 0.0243, "step": 130330 }, { "grad_norm": 0.5342477560043335, "learning_rate": 2.9658888129268373e-05, "loss": 0.0216, "step": 130340 }, { "grad_norm": 0.5248407125473022, "learning_rate": 2.965133612325609e-05, "loss": 0.0192, "step": 130350 }, { "grad_norm": 0.26466652750968933, "learning_rate": 2.9643784673568665e-05, "loss": 0.0213, "step": 130360 }, { "grad_norm": 0.40401938557624817, "learning_rate": 2.963623378041248e-05, "loss": 0.025, "step": 130370 }, { "grad_norm": 1.3091782331466675, "learning_rate": 2.9628683443994047e-05, "loss": 0.0382, "step": 130380 }, { "grad_norm": 0.3028952479362488, "learning_rate": 2.962113366451973e-05, "loss": 0.0203, "step": 130390 }, { "grad_norm": 0.52834552526474, "learning_rate": 2.9613584442195973e-05, "loss": 0.0221, "step": 130400 }, { "grad_norm": 0.7004006505012512, "learning_rate": 2.960603577722916e-05, "loss": 0.0211, "step": 130410 }, { "grad_norm": 0.5366135239601135, "learning_rate": 2.9598487669825658e-05, "loss": 0.0147, "step": 130420 }, { "grad_norm": 0.37477871775627136, "learning_rate": 2.9590940120191855e-05, "loss": 0.0285, "step": 130430 }, { "grad_norm": 0.28611642122268677, "learning_rate": 2.958339312853407e-05, "loss": 0.0326, "step": 130440 }, { "grad_norm": 0.9416021704673767, "learning_rate": 2.9575846695058657e-05, "loss": 0.0162, "step": 130450 }, { "grad_norm": 0.6332621574401855, "learning_rate": 2.9568300819971904e-05, "loss": 0.0219, "step": 130460 }, { "grad_norm": 0.33567753434181213, "learning_rate": 2.956075550348015e-05, "loss": 0.0215, "step": 130470 }, { "grad_norm": 0.6943308115005493, "learning_rate": 2.9553210745789638e-05, "loss": 0.0242, "step": 130480 }, { "grad_norm": 0.5505313873291016, "learning_rate": 2.9545666547106676e-05, "loss": 0.0255, "step": 130490 }, { "grad_norm": 0.22450478374958038, "learning_rate": 2.9538122907637522e-05, "loss": 0.0194, "step": 130500 }, { "grad_norm": 0.6002326607704163, "learning_rate": 2.9530579827588378e-05, "loss": 0.0287, "step": 130510 }, { "grad_norm": 0.46586740016937256, "learning_rate": 2.9523037307165507e-05, "loss": 0.0238, "step": 130520 }, { "grad_norm": 0.2363877296447754, "learning_rate": 2.9515495346575096e-05, "loss": 0.0215, "step": 130530 }, { "grad_norm": 0.3747173547744751, "learning_rate": 2.9507953946023354e-05, "loss": 0.0231, "step": 130540 }, { "grad_norm": 0.34312888979911804, "learning_rate": 2.950041310571644e-05, "loss": 0.0169, "step": 130550 }, { "grad_norm": 0.9704059958457947, "learning_rate": 2.9492872825860522e-05, "loss": 0.0175, "step": 130560 }, { "grad_norm": 0.5921598076820374, "learning_rate": 2.9485333106661788e-05, "loss": 0.0178, "step": 130570 }, { "grad_norm": 2.6472244262695312, "learning_rate": 2.947779394832631e-05, "loss": 0.0229, "step": 130580 }, { "grad_norm": 0.24742062389850616, "learning_rate": 2.947025535106026e-05, "loss": 0.0231, "step": 130590 }, { "grad_norm": 0.891701340675354, "learning_rate": 2.94627173150697e-05, "loss": 0.0172, "step": 130600 }, { "grad_norm": 0.31533679366111755, "learning_rate": 2.9455179840560733e-05, "loss": 0.0178, "step": 130610 }, { "grad_norm": 1.0602174997329712, "learning_rate": 2.944764292773944e-05, "loss": 0.0325, "step": 130620 }, { "grad_norm": 0.4387511909008026, "learning_rate": 2.9440106576811855e-05, "loss": 0.0224, "step": 130630 }, { "grad_norm": 0.8841016888618469, "learning_rate": 2.943257078798405e-05, "loss": 0.0279, "step": 130640 }, { "grad_norm": 0.5322188138961792, "learning_rate": 2.9425035561462023e-05, "loss": 0.0274, "step": 130650 }, { "grad_norm": 0.39011555910110474, "learning_rate": 2.9417500897451798e-05, "loss": 0.0265, "step": 130660 }, { "grad_norm": 0.3427865505218506, "learning_rate": 2.940996679615936e-05, "loss": 0.0242, "step": 130670 }, { "grad_norm": 0.5717509984970093, "learning_rate": 2.9402433257790685e-05, "loss": 0.0197, "step": 130680 }, { "grad_norm": 0.6728214025497437, "learning_rate": 2.9394900282551762e-05, "loss": 0.0176, "step": 130690 }, { "grad_norm": 0.5618331432342529, "learning_rate": 2.938736787064852e-05, "loss": 0.0154, "step": 130700 }, { "grad_norm": 0.3430277109146118, "learning_rate": 2.93798360222869e-05, "loss": 0.0202, "step": 130710 }, { "grad_norm": 0.44955629110336304, "learning_rate": 2.9372304737672812e-05, "loss": 0.0231, "step": 130720 }, { "grad_norm": 0.45483967661857605, "learning_rate": 2.9364774017012175e-05, "loss": 0.018, "step": 130730 }, { "grad_norm": 0.3303823471069336, "learning_rate": 2.9357243860510853e-05, "loss": 0.0171, "step": 130740 }, { "grad_norm": 0.2930273115634918, "learning_rate": 2.934971426837472e-05, "loss": 0.0205, "step": 130750 }, { "grad_norm": 0.2885183095932007, "learning_rate": 2.9342185240809662e-05, "loss": 0.0189, "step": 130760 }, { "grad_norm": 0.38818684220314026, "learning_rate": 2.9334656778021475e-05, "loss": 0.0325, "step": 130770 }, { "grad_norm": 0.5870450735092163, "learning_rate": 2.9327128880216038e-05, "loss": 0.0236, "step": 130780 }, { "grad_norm": 0.38983219861984253, "learning_rate": 2.9319601547599095e-05, "loss": 0.0223, "step": 130790 }, { "grad_norm": 0.3700959384441376, "learning_rate": 2.931207478037649e-05, "loss": 0.0209, "step": 130800 }, { "grad_norm": 0.29887115955352783, "learning_rate": 2.9304548578754e-05, "loss": 0.0239, "step": 130810 }, { "grad_norm": 0.5500478744506836, "learning_rate": 2.929702294293737e-05, "loss": 0.0316, "step": 130820 }, { "grad_norm": 0.47232377529144287, "learning_rate": 2.928949787313236e-05, "loss": 0.0218, "step": 130830 }, { "grad_norm": 0.4834631383419037, "learning_rate": 2.928197336954469e-05, "loss": 0.0144, "step": 130840 }, { "grad_norm": 0.32826992869377136, "learning_rate": 2.927444943238009e-05, "loss": 0.0237, "step": 130850 }, { "grad_norm": 0.2108919322490692, "learning_rate": 2.9266926061844258e-05, "loss": 0.0323, "step": 130860 }, { "grad_norm": 0.3186053931713104, "learning_rate": 2.925940325814286e-05, "loss": 0.025, "step": 130870 }, { "grad_norm": 0.42667609453201294, "learning_rate": 2.9251881021481614e-05, "loss": 0.0272, "step": 130880 }, { "grad_norm": 0.6445286273956299, "learning_rate": 2.924435935206612e-05, "loss": 0.0213, "step": 130890 }, { "grad_norm": 0.22970443964004517, "learning_rate": 2.923683825010206e-05, "loss": 0.0213, "step": 130900 }, { "grad_norm": 0.42571404576301575, "learning_rate": 2.9229317715795036e-05, "loss": 0.0169, "step": 130910 }, { "grad_norm": 0.36320269107818604, "learning_rate": 2.9221797749350665e-05, "loss": 0.0266, "step": 130920 }, { "grad_norm": 0.3688536584377289, "learning_rate": 2.9214278350974535e-05, "loss": 0.0199, "step": 130930 }, { "grad_norm": 0.5400039553642273, "learning_rate": 2.9206759520872222e-05, "loss": 0.0196, "step": 130940 }, { "grad_norm": 0.43155011534690857, "learning_rate": 2.919924125924931e-05, "loss": 0.0226, "step": 130950 }, { "grad_norm": 0.41203540563583374, "learning_rate": 2.9191723566311307e-05, "loss": 0.0248, "step": 130960 }, { "grad_norm": 0.2741439938545227, "learning_rate": 2.918420644226378e-05, "loss": 0.0167, "step": 130970 }, { "grad_norm": 0.3292451798915863, "learning_rate": 2.9176689887312204e-05, "loss": 0.0158, "step": 130980 }, { "grad_norm": 0.538326621055603, "learning_rate": 2.916917390166213e-05, "loss": 0.0179, "step": 130990 }, { "grad_norm": 0.3246479332447052, "learning_rate": 2.9161658485519012e-05, "loss": 0.0369, "step": 131000 }, { "grad_norm": 0.3601612448692322, "learning_rate": 2.9154143639088304e-05, "loss": 0.0266, "step": 131010 }, { "grad_norm": 4.371501445770264, "learning_rate": 2.9146629362575495e-05, "loss": 0.0188, "step": 131020 }, { "grad_norm": 0.5515604019165039, "learning_rate": 2.913911565618599e-05, "loss": 0.0482, "step": 131030 }, { "grad_norm": 0.39881432056427, "learning_rate": 2.9131602520125256e-05, "loss": 0.016, "step": 131040 }, { "grad_norm": 0.429901659488678, "learning_rate": 2.9124089954598645e-05, "loss": 0.0265, "step": 131050 }, { "grad_norm": 0.4361029267311096, "learning_rate": 2.9116577959811558e-05, "loss": 0.0219, "step": 131060 }, { "grad_norm": 0.24487845599651337, "learning_rate": 2.910906653596941e-05, "loss": 0.019, "step": 131070 }, { "grad_norm": 0.5827226638793945, "learning_rate": 2.910155568327753e-05, "loss": 0.0288, "step": 131080 }, { "grad_norm": 0.6184825301170349, "learning_rate": 2.9094045401941273e-05, "loss": 0.0215, "step": 131090 }, { "grad_norm": 0.46711528301239014, "learning_rate": 2.908653569216594e-05, "loss": 0.0136, "step": 131100 }, { "grad_norm": 0.3248073160648346, "learning_rate": 2.9079026554156874e-05, "loss": 0.0244, "step": 131110 }, { "grad_norm": 0.5682010054588318, "learning_rate": 2.9071517988119374e-05, "loss": 0.0181, "step": 131120 }, { "grad_norm": 0.7292381525039673, "learning_rate": 2.9064009994258684e-05, "loss": 0.0243, "step": 131130 }, { "grad_norm": 0.6092053651809692, "learning_rate": 2.905650257278012e-05, "loss": 0.0219, "step": 131140 }, { "grad_norm": 0.21115168929100037, "learning_rate": 2.9048995723888884e-05, "loss": 0.0293, "step": 131150 }, { "grad_norm": 0.3872910439968109, "learning_rate": 2.9041489447790256e-05, "loss": 0.023, "step": 131160 }, { "grad_norm": 0.2433592528104782, "learning_rate": 2.903398374468943e-05, "loss": 0.0198, "step": 131170 }, { "grad_norm": 0.49484530091285706, "learning_rate": 2.9026478614791598e-05, "loss": 0.0205, "step": 131180 }, { "grad_norm": 0.6570601463317871, "learning_rate": 2.9018974058301972e-05, "loss": 0.0203, "step": 131190 }, { "grad_norm": 0.3370015025138855, "learning_rate": 2.901147007542572e-05, "loss": 0.0158, "step": 131200 }, { "grad_norm": 0.4611160159111023, "learning_rate": 2.9003966666367988e-05, "loss": 0.0183, "step": 131210 }, { "grad_norm": 0.41219642758369446, "learning_rate": 2.8996463831333898e-05, "loss": 0.0215, "step": 131220 }, { "grad_norm": 0.3672814667224884, "learning_rate": 2.8988961570528616e-05, "loss": 0.0213, "step": 131230 }, { "grad_norm": 0.5204527378082275, "learning_rate": 2.8981459884157214e-05, "loss": 0.0275, "step": 131240 }, { "grad_norm": 0.3276183307170868, "learning_rate": 2.8973958772424824e-05, "loss": 0.0277, "step": 131250 }, { "grad_norm": 0.30249762535095215, "learning_rate": 2.8966458235536498e-05, "loss": 0.0256, "step": 131260 }, { "grad_norm": 0.22387687861919403, "learning_rate": 2.895895827369729e-05, "loss": 0.0228, "step": 131270 }, { "grad_norm": 0.44701504707336426, "learning_rate": 2.895145888711227e-05, "loss": 0.0218, "step": 131280 }, { "grad_norm": 0.34775426983833313, "learning_rate": 2.8943960075986458e-05, "loss": 0.0185, "step": 131290 }, { "grad_norm": 0.2984231412410736, "learning_rate": 2.8936461840524864e-05, "loss": 0.02, "step": 131300 }, { "grad_norm": 0.9293723702430725, "learning_rate": 2.8928964180932477e-05, "loss": 0.0228, "step": 131310 }, { "grad_norm": 0.45245984196662903, "learning_rate": 2.892146709741429e-05, "loss": 0.0205, "step": 131320 }, { "grad_norm": 0.6779378652572632, "learning_rate": 2.8913970590175315e-05, "loss": 0.0242, "step": 131330 }, { "grad_norm": 0.670417308807373, "learning_rate": 2.8906474659420425e-05, "loss": 0.024, "step": 131340 }, { "grad_norm": 0.44948089122772217, "learning_rate": 2.8898979305354607e-05, "loss": 0.0216, "step": 131350 }, { "grad_norm": 0.2583845257759094, "learning_rate": 2.8891484528182748e-05, "loss": 0.0222, "step": 131360 }, { "grad_norm": 0.27318084239959717, "learning_rate": 2.8883990328109795e-05, "loss": 0.0188, "step": 131370 }, { "grad_norm": 0.29618507623672485, "learning_rate": 2.8876496705340618e-05, "loss": 0.0171, "step": 131380 }, { "grad_norm": 0.32697394490242004, "learning_rate": 2.8869003660080052e-05, "loss": 0.0226, "step": 131390 }, { "grad_norm": 0.3531944155693054, "learning_rate": 2.886151119253302e-05, "loss": 0.0208, "step": 131400 }, { "grad_norm": 0.784650444984436, "learning_rate": 2.8854019302904322e-05, "loss": 0.0257, "step": 131410 }, { "grad_norm": 0.3861343264579773, "learning_rate": 2.8846527991398796e-05, "loss": 0.0221, "step": 131420 }, { "grad_norm": 3.0835351943969727, "learning_rate": 2.883903725822123e-05, "loss": 0.0204, "step": 131430 }, { "grad_norm": 0.4063613712787628, "learning_rate": 2.8831547103576427e-05, "loss": 0.0217, "step": 131440 }, { "grad_norm": 0.4617671072483063, "learning_rate": 2.8824057527669203e-05, "loss": 0.0174, "step": 131450 }, { "grad_norm": 0.36172789335250854, "learning_rate": 2.8816568530704292e-05, "loss": 0.0255, "step": 131460 }, { "grad_norm": 0.3400033116340637, "learning_rate": 2.8809080112886433e-05, "loss": 0.0178, "step": 131470 }, { "grad_norm": 0.2922501862049103, "learning_rate": 2.8801592274420342e-05, "loss": 0.0178, "step": 131480 }, { "grad_norm": 0.521759569644928, "learning_rate": 2.8794105015510776e-05, "loss": 0.0188, "step": 131490 }, { "grad_norm": 0.4927009642124176, "learning_rate": 2.8786618336362413e-05, "loss": 0.0211, "step": 131500 }, { "grad_norm": 0.2597176432609558, "learning_rate": 2.877913223717992e-05, "loss": 0.0172, "step": 131510 }, { "grad_norm": 0.5095505118370056, "learning_rate": 2.8771646718167987e-05, "loss": 0.0289, "step": 131520 }, { "grad_norm": 0.4377647042274475, "learning_rate": 2.876416177953125e-05, "loss": 0.0223, "step": 131530 }, { "grad_norm": 0.48074740171432495, "learning_rate": 2.8756677421474388e-05, "loss": 0.0202, "step": 131540 }, { "grad_norm": 0.2284223884344101, "learning_rate": 2.8749193644201943e-05, "loss": 0.024, "step": 131550 }, { "grad_norm": 0.4074338674545288, "learning_rate": 2.8741710447918558e-05, "loss": 0.023, "step": 131560 }, { "grad_norm": 0.24289214611053467, "learning_rate": 2.8734227832828846e-05, "loss": 0.0219, "step": 131570 }, { "grad_norm": 0.5070481896400452, "learning_rate": 2.8726745799137354e-05, "loss": 0.0154, "step": 131580 }, { "grad_norm": 0.4140041768550873, "learning_rate": 2.8719264347048646e-05, "loss": 0.0146, "step": 131590 }, { "grad_norm": 0.9278138875961304, "learning_rate": 2.8711783476767228e-05, "loss": 0.0216, "step": 131600 }, { "grad_norm": 0.4165962040424347, "learning_rate": 2.870430318849768e-05, "loss": 0.0172, "step": 131610 }, { "grad_norm": 0.3648320734500885, "learning_rate": 2.869682348244448e-05, "loss": 0.0162, "step": 131620 }, { "grad_norm": 0.36254701018333435, "learning_rate": 2.8689344358812098e-05, "loss": 0.0167, "step": 131630 }, { "grad_norm": 1.9381921291351318, "learning_rate": 2.8681865817805054e-05, "loss": 0.0285, "step": 131640 }, { "grad_norm": 0.49409499764442444, "learning_rate": 2.8674387859627777e-05, "loss": 0.0197, "step": 131650 }, { "grad_norm": 0.7037181854248047, "learning_rate": 2.8666910484484742e-05, "loss": 0.0254, "step": 131660 }, { "grad_norm": 0.35661059617996216, "learning_rate": 2.865943369258036e-05, "loss": 0.021, "step": 131670 }, { "grad_norm": 0.49498969316482544, "learning_rate": 2.865195748411904e-05, "loss": 0.0307, "step": 131680 }, { "grad_norm": 0.3252015709877014, "learning_rate": 2.8644481859305162e-05, "loss": 0.0187, "step": 131690 }, { "grad_norm": 0.3611985743045807, "learning_rate": 2.8637006818343155e-05, "loss": 0.0196, "step": 131700 }, { "grad_norm": 0.2606452703475952, "learning_rate": 2.862953236143735e-05, "loss": 0.015, "step": 131710 }, { "grad_norm": 0.40264463424682617, "learning_rate": 2.8622058488792082e-05, "loss": 0.0155, "step": 131720 }, { "grad_norm": 0.3693820834159851, "learning_rate": 2.861458520061172e-05, "loss": 0.0171, "step": 131730 }, { "grad_norm": 0.477174311876297, "learning_rate": 2.8607112497100552e-05, "loss": 0.0207, "step": 131740 }, { "grad_norm": 0.45471298694610596, "learning_rate": 2.8599640378462912e-05, "loss": 0.0306, "step": 131750 }, { "grad_norm": 0.33188390731811523, "learning_rate": 2.8592168844903055e-05, "loss": 0.02, "step": 131760 }, { "grad_norm": 0.1989911049604416, "learning_rate": 2.8584697896625245e-05, "loss": 0.0251, "step": 131770 }, { "grad_norm": 0.2791670262813568, "learning_rate": 2.8577227533833765e-05, "loss": 0.0216, "step": 131780 }, { "grad_norm": 0.35998132824897766, "learning_rate": 2.856975775673284e-05, "loss": 0.025, "step": 131790 }, { "grad_norm": 0.29337504506111145, "learning_rate": 2.8562288565526683e-05, "loss": 0.0226, "step": 131800 }, { "grad_norm": 0.26870518922805786, "learning_rate": 2.8554819960419493e-05, "loss": 0.0228, "step": 131810 }, { "grad_norm": 1.0418727397918701, "learning_rate": 2.8547351941615465e-05, "loss": 0.031, "step": 131820 }, { "grad_norm": 0.7090299129486084, "learning_rate": 2.853988450931881e-05, "loss": 0.0187, "step": 131830 }, { "grad_norm": 0.5158100128173828, "learning_rate": 2.8532417663733613e-05, "loss": 0.016, "step": 131840 }, { "grad_norm": 0.2970050275325775, "learning_rate": 2.8524951405064072e-05, "loss": 0.029, "step": 131850 }, { "grad_norm": 0.27086323499679565, "learning_rate": 2.851748573351427e-05, "loss": 0.0272, "step": 131860 }, { "grad_norm": 0.30783405900001526, "learning_rate": 2.851002064928836e-05, "loss": 0.0184, "step": 131870 }, { "grad_norm": 0.6839306950569153, "learning_rate": 2.850255615259041e-05, "loss": 0.0406, "step": 131880 }, { "grad_norm": 0.4122909605503082, "learning_rate": 2.8495092243624476e-05, "loss": 0.0248, "step": 131890 }, { "grad_norm": 0.4946550130844116, "learning_rate": 2.8487628922594667e-05, "loss": 0.0195, "step": 131900 }, { "grad_norm": 0.43581900000572205, "learning_rate": 2.8480166189705005e-05, "loss": 0.0232, "step": 131910 }, { "grad_norm": 0.3723241686820984, "learning_rate": 2.8472704045159504e-05, "loss": 0.0231, "step": 131920 }, { "grad_norm": 0.5034899115562439, "learning_rate": 2.8465242489162176e-05, "loss": 0.0276, "step": 131930 }, { "grad_norm": 0.5027072429656982, "learning_rate": 2.8457781521917042e-05, "loss": 0.0222, "step": 131940 }, { "grad_norm": 0.4231618344783783, "learning_rate": 2.845032114362805e-05, "loss": 0.0246, "step": 131950 }, { "grad_norm": 0.493873655796051, "learning_rate": 2.8442861354499208e-05, "loss": 0.0194, "step": 131960 }, { "grad_norm": 0.6781842112541199, "learning_rate": 2.8435402154734425e-05, "loss": 0.0296, "step": 131970 }, { "grad_norm": 0.3723393976688385, "learning_rate": 2.8427943544537638e-05, "loss": 0.0207, "step": 131980 }, { "grad_norm": 0.4461900293827057, "learning_rate": 2.8420485524112782e-05, "loss": 0.0261, "step": 131990 }, { "grad_norm": 0.45057934522628784, "learning_rate": 2.8413028093663752e-05, "loss": 0.0265, "step": 132000 }, { "grad_norm": 0.42821651697158813, "learning_rate": 2.8405571253394397e-05, "loss": 0.021, "step": 132010 }, { "grad_norm": 0.6537704467773438, "learning_rate": 2.8398115003508633e-05, "loss": 0.022, "step": 132020 }, { "grad_norm": 0.41768747568130493, "learning_rate": 2.839065934421027e-05, "loss": 0.0202, "step": 132030 }, { "grad_norm": 0.6987510919570923, "learning_rate": 2.83832042757032e-05, "loss": 0.018, "step": 132040 }, { "grad_norm": 0.18614478409290314, "learning_rate": 2.8375749798191175e-05, "loss": 0.0146, "step": 132050 }, { "grad_norm": 0.3073199689388275, "learning_rate": 2.8368295911878036e-05, "loss": 0.0237, "step": 132060 }, { "grad_norm": 0.4283261299133301, "learning_rate": 2.8360842616967543e-05, "loss": 0.0248, "step": 132070 }, { "grad_norm": 0.5313020348548889, "learning_rate": 2.83533899136635e-05, "loss": 0.0257, "step": 132080 }, { "grad_norm": 0.411780446767807, "learning_rate": 2.8345937802169654e-05, "loss": 0.0234, "step": 132090 }, { "grad_norm": 0.32880699634552, "learning_rate": 2.8338486282689702e-05, "loss": 0.0268, "step": 132100 }, { "grad_norm": 0.41478797793388367, "learning_rate": 2.8331035355427428e-05, "loss": 0.0201, "step": 132110 }, { "grad_norm": 0.3837585747241974, "learning_rate": 2.8323585020586495e-05, "loss": 0.021, "step": 132120 }, { "grad_norm": 0.30551081895828247, "learning_rate": 2.8316135278370616e-05, "loss": 0.021, "step": 132130 }, { "grad_norm": 0.25585609674453735, "learning_rate": 2.830868612898342e-05, "loss": 0.0177, "step": 132140 }, { "grad_norm": 1.2360914945602417, "learning_rate": 2.83012375726286e-05, "loss": 0.0306, "step": 132150 }, { "grad_norm": 0.4660358726978302, "learning_rate": 2.8293789609509825e-05, "loss": 0.026, "step": 132160 }, { "grad_norm": 0.5761998295783997, "learning_rate": 2.8286342239830676e-05, "loss": 0.0207, "step": 132170 }, { "grad_norm": 0.42505571246147156, "learning_rate": 2.8278895463794775e-05, "loss": 0.0177, "step": 132180 }, { "grad_norm": 0.21130309998989105, "learning_rate": 2.8271449281605687e-05, "loss": 0.029, "step": 132190 }, { "grad_norm": 0.3842610716819763, "learning_rate": 2.826400369346704e-05, "loss": 0.0168, "step": 132200 }, { "grad_norm": 0.3096408247947693, "learning_rate": 2.8256558699582358e-05, "loss": 0.0287, "step": 132210 }, { "grad_norm": 0.664115309715271, "learning_rate": 2.8249114300155177e-05, "loss": 0.0193, "step": 132220 }, { "grad_norm": 1.5942472219467163, "learning_rate": 2.824167049538906e-05, "loss": 0.022, "step": 132230 }, { "grad_norm": 0.34753888845443726, "learning_rate": 2.823422728548748e-05, "loss": 0.014, "step": 132240 }, { "grad_norm": 0.5138251781463623, "learning_rate": 2.8226784670653988e-05, "loss": 0.0168, "step": 132250 }, { "grad_norm": 0.4440738260746002, "learning_rate": 2.8219342651091984e-05, "loss": 0.016, "step": 132260 }, { "grad_norm": 0.20532286167144775, "learning_rate": 2.8211901227004965e-05, "loss": 0.0205, "step": 132270 }, { "grad_norm": 0.3262947201728821, "learning_rate": 2.8204460398596407e-05, "loss": 0.0296, "step": 132280 }, { "grad_norm": 0.48205190896987915, "learning_rate": 2.819702016606971e-05, "loss": 0.025, "step": 132290 }, { "grad_norm": 0.6647748351097107, "learning_rate": 2.818958052962829e-05, "loss": 0.0252, "step": 132300 }, { "grad_norm": 0.33476629853248596, "learning_rate": 2.8182141489475523e-05, "loss": 0.0211, "step": 132310 }, { "grad_norm": 0.5064487457275391, "learning_rate": 2.817470304581483e-05, "loss": 0.0205, "step": 132320 }, { "grad_norm": 0.35170599818229675, "learning_rate": 2.816726519884956e-05, "loss": 0.0173, "step": 132330 }, { "grad_norm": 0.6885849237442017, "learning_rate": 2.815982794878304e-05, "loss": 0.0349, "step": 132340 }, { "grad_norm": 1.3079217672348022, "learning_rate": 2.8152391295818637e-05, "loss": 0.0352, "step": 132350 }, { "grad_norm": 0.5016785860061646, "learning_rate": 2.8144955240159627e-05, "loss": 0.0203, "step": 132360 }, { "grad_norm": 0.740662157535553, "learning_rate": 2.8137519782009347e-05, "loss": 0.0249, "step": 132370 }, { "grad_norm": 0.4691775143146515, "learning_rate": 2.8130084921571064e-05, "loss": 0.0169, "step": 132380 }, { "grad_norm": 0.5407587885856628, "learning_rate": 2.812265065904803e-05, "loss": 0.0196, "step": 132390 }, { "grad_norm": 0.3597063422203064, "learning_rate": 2.8115216994643524e-05, "loss": 0.0211, "step": 132400 }, { "grad_norm": 1.5209779739379883, "learning_rate": 2.8107783928560762e-05, "loss": 0.0264, "step": 132410 }, { "grad_norm": 0.32942256331443787, "learning_rate": 2.810035146100296e-05, "loss": 0.033, "step": 132420 }, { "grad_norm": 0.5383554697036743, "learning_rate": 2.8092919592173305e-05, "loss": 0.0213, "step": 132430 }, { "grad_norm": 0.5869735479354858, "learning_rate": 2.808548832227502e-05, "loss": 0.023, "step": 132440 }, { "grad_norm": 0.4555942714214325, "learning_rate": 2.807805765151123e-05, "loss": 0.0232, "step": 132450 }, { "grad_norm": 0.49260586500167847, "learning_rate": 2.807062758008513e-05, "loss": 0.025, "step": 132460 }, { "grad_norm": 0.2328321784734726, "learning_rate": 2.806319810819983e-05, "loss": 0.015, "step": 132470 }, { "grad_norm": 0.5040455460548401, "learning_rate": 2.805576923605843e-05, "loss": 0.0269, "step": 132480 }, { "grad_norm": 0.2707942724227905, "learning_rate": 2.8048340963864083e-05, "loss": 0.0245, "step": 132490 }, { "grad_norm": 0.7240216732025146, "learning_rate": 2.8040913291819842e-05, "loss": 0.0286, "step": 132500 }, { "grad_norm": 0.3809971809387207, "learning_rate": 2.803348622012878e-05, "loss": 0.0267, "step": 132510 }, { "grad_norm": 2.1501989364624023, "learning_rate": 2.802605974899394e-05, "loss": 0.0246, "step": 132520 }, { "grad_norm": 10.161870002746582, "learning_rate": 2.8018633878618365e-05, "loss": 0.0192, "step": 132530 }, { "grad_norm": 0.635984480381012, "learning_rate": 2.8011208609205126e-05, "loss": 0.0202, "step": 132540 }, { "grad_norm": 0.24550920724868774, "learning_rate": 2.800378394095714e-05, "loss": 0.0222, "step": 132550 }, { "grad_norm": 0.2065657675266266, "learning_rate": 2.7996359874077455e-05, "loss": 0.0219, "step": 132560 }, { "grad_norm": 0.8783692717552185, "learning_rate": 2.7988936408769007e-05, "loss": 0.0286, "step": 132570 }, { "grad_norm": 0.2090909332036972, "learning_rate": 2.7981513545234788e-05, "loss": 0.0231, "step": 132580 }, { "grad_norm": 0.4590967297554016, "learning_rate": 2.7974091283677716e-05, "loss": 0.0298, "step": 132590 }, { "grad_norm": 0.3973250687122345, "learning_rate": 2.79666696243007e-05, "loss": 0.0192, "step": 132600 }, { "grad_norm": 1.4252036809921265, "learning_rate": 2.7959248567306672e-05, "loss": 0.0251, "step": 132610 }, { "grad_norm": 0.2972877323627472, "learning_rate": 2.795182811289851e-05, "loss": 0.0183, "step": 132620 }, { "grad_norm": 0.4468490183353424, "learning_rate": 2.7944408261279088e-05, "loss": 0.0203, "step": 132630 }, { "grad_norm": 0.807903528213501, "learning_rate": 2.7936989012651227e-05, "loss": 0.0308, "step": 132640 }, { "grad_norm": 0.5768556594848633, "learning_rate": 2.7929570367217804e-05, "loss": 0.0235, "step": 132650 }, { "grad_norm": 0.3371683657169342, "learning_rate": 2.792215232518166e-05, "loss": 0.0234, "step": 132660 }, { "grad_norm": 0.4019278585910797, "learning_rate": 2.7914734886745574e-05, "loss": 0.0244, "step": 132670 }, { "grad_norm": 0.5450245141983032, "learning_rate": 2.7907318052112342e-05, "loss": 0.0178, "step": 132680 }, { "grad_norm": 0.43558016419410706, "learning_rate": 2.789990182148471e-05, "loss": 0.0179, "step": 132690 }, { "grad_norm": 0.25503188371658325, "learning_rate": 2.789248619506548e-05, "loss": 0.0192, "step": 132700 }, { "grad_norm": 0.29407745599746704, "learning_rate": 2.7885071173057375e-05, "loss": 0.0169, "step": 132710 }, { "grad_norm": 0.2504351735115051, "learning_rate": 2.7877656755663095e-05, "loss": 0.0213, "step": 132720 }, { "grad_norm": 0.4951808452606201, "learning_rate": 2.787024294308539e-05, "loss": 0.0193, "step": 132730 }, { "grad_norm": 0.29218417406082153, "learning_rate": 2.78628297355269e-05, "loss": 0.0273, "step": 132740 }, { "grad_norm": 0.30044499039649963, "learning_rate": 2.7855417133190375e-05, "loss": 0.0195, "step": 132750 }, { "grad_norm": 0.30703383684158325, "learning_rate": 2.7848005136278388e-05, "loss": 0.0294, "step": 132760 }, { "grad_norm": 0.28536316752433777, "learning_rate": 2.7840593744993615e-05, "loss": 0.022, "step": 132770 }, { "grad_norm": 0.44653964042663574, "learning_rate": 2.7833182959538706e-05, "loss": 0.022, "step": 132780 }, { "grad_norm": 0.25279295444488525, "learning_rate": 2.782577278011625e-05, "loss": 0.0188, "step": 132790 }, { "grad_norm": 0.5763594508171082, "learning_rate": 2.781836320692883e-05, "loss": 0.0285, "step": 132800 }, { "grad_norm": 0.7300931215286255, "learning_rate": 2.781095424017901e-05, "loss": 0.0271, "step": 132810 }, { "grad_norm": 0.32764190435409546, "learning_rate": 2.7803545880069386e-05, "loss": 0.0211, "step": 132820 }, { "grad_norm": 0.5255294442176819, "learning_rate": 2.779613812680248e-05, "loss": 0.0311, "step": 132830 }, { "grad_norm": 0.45044246315956116, "learning_rate": 2.778873098058079e-05, "loss": 0.0225, "step": 132840 }, { "grad_norm": 0.6663727164268494, "learning_rate": 2.7781324441606875e-05, "loss": 0.0173, "step": 132850 }, { "grad_norm": 0.21658585965633392, "learning_rate": 2.7773918510083184e-05, "loss": 0.027, "step": 132860 }, { "grad_norm": 0.28707030415534973, "learning_rate": 2.776651318621223e-05, "loss": 0.0158, "step": 132870 }, { "grad_norm": 0.3898742198944092, "learning_rate": 2.7759108470196448e-05, "loss": 0.0295, "step": 132880 }, { "grad_norm": 0.4759385883808136, "learning_rate": 2.7751704362238283e-05, "loss": 0.0228, "step": 132890 }, { "grad_norm": 0.5252234935760498, "learning_rate": 2.774430086254015e-05, "loss": 0.0207, "step": 132900 }, { "grad_norm": 0.5356399416923523, "learning_rate": 2.773689797130449e-05, "loss": 0.0212, "step": 132910 }, { "grad_norm": 0.4636170566082001, "learning_rate": 2.7729495688733676e-05, "loss": 0.0293, "step": 132920 }, { "grad_norm": 0.5379051566123962, "learning_rate": 2.7722094015030058e-05, "loss": 0.0253, "step": 132930 }, { "grad_norm": 0.3463233709335327, "learning_rate": 2.7714692950396048e-05, "loss": 0.026, "step": 132940 }, { "grad_norm": 0.3830753266811371, "learning_rate": 2.770729249503393e-05, "loss": 0.0283, "step": 132950 }, { "grad_norm": 0.41891807317733765, "learning_rate": 2.7699892649146085e-05, "loss": 0.0175, "step": 132960 }, { "grad_norm": 0.16683724522590637, "learning_rate": 2.76924934129348e-05, "loss": 0.0228, "step": 132970 }, { "grad_norm": 0.8898942470550537, "learning_rate": 2.7685094786602344e-05, "loss": 0.0252, "step": 132980 }, { "grad_norm": 0.23328378796577454, "learning_rate": 2.767769677035103e-05, "loss": 0.0178, "step": 132990 }, { "grad_norm": 0.3940861225128174, "learning_rate": 2.76702993643831e-05, "loss": 0.0162, "step": 133000 }, { "grad_norm": 0.3524504005908966, "learning_rate": 2.7662902568900807e-05, "loss": 0.0239, "step": 133010 }, { "grad_norm": 0.4083004593849182, "learning_rate": 2.7655506384106334e-05, "loss": 0.0193, "step": 133020 }, { "grad_norm": 0.4115762710571289, "learning_rate": 2.7648110810201923e-05, "loss": 0.0161, "step": 133030 }, { "grad_norm": 0.21502803266048431, "learning_rate": 2.7640715847389808e-05, "loss": 0.0191, "step": 133040 }, { "grad_norm": 0.5997680425643921, "learning_rate": 2.7633321495872084e-05, "loss": 0.0316, "step": 133050 }, { "grad_norm": 0.41195908188819885, "learning_rate": 2.7625927755850965e-05, "loss": 0.0262, "step": 133060 }, { "grad_norm": 0.38012200593948364, "learning_rate": 2.761853462752856e-05, "loss": 0.0225, "step": 133070 }, { "grad_norm": 0.2840381860733032, "learning_rate": 2.7611142111107018e-05, "loss": 0.0177, "step": 133080 }, { "grad_norm": 0.38706064224243164, "learning_rate": 2.760375020678845e-05, "loss": 0.0228, "step": 133090 }, { "grad_norm": 0.19125184416770935, "learning_rate": 2.7596358914774913e-05, "loss": 0.0304, "step": 133100 }, { "grad_norm": 0.48339250683784485, "learning_rate": 2.758896823526853e-05, "loss": 0.0271, "step": 133110 }, { "grad_norm": 0.30199772119522095, "learning_rate": 2.7581578168471333e-05, "loss": 0.0191, "step": 133120 }, { "grad_norm": 0.2996939420700073, "learning_rate": 2.7574188714585368e-05, "loss": 0.0172, "step": 133130 }, { "grad_norm": 0.3844325840473175, "learning_rate": 2.756679987381264e-05, "loss": 0.0256, "step": 133140 }, { "grad_norm": 0.6453239321708679, "learning_rate": 2.7559411646355195e-05, "loss": 0.017, "step": 133150 }, { "grad_norm": 0.6982542872428894, "learning_rate": 2.755202403241498e-05, "loss": 0.0174, "step": 133160 }, { "grad_norm": 0.27308666706085205, "learning_rate": 2.754463703219402e-05, "loss": 0.022, "step": 133170 }, { "grad_norm": 0.36323827505111694, "learning_rate": 2.7537250645894252e-05, "loss": 0.0423, "step": 133180 }, { "grad_norm": 1.4086331129074097, "learning_rate": 2.7529864873717582e-05, "loss": 0.0237, "step": 133190 }, { "grad_norm": 0.3794777989387512, "learning_rate": 2.752247971586599e-05, "loss": 0.016, "step": 133200 }, { "grad_norm": 0.4156041741371155, "learning_rate": 2.751509517254136e-05, "loss": 0.0211, "step": 133210 }, { "grad_norm": 0.4113713204860687, "learning_rate": 2.7507711243945554e-05, "loss": 0.0276, "step": 133220 }, { "grad_norm": 0.31132879853248596, "learning_rate": 2.75003279302805e-05, "loss": 0.018, "step": 133230 }, { "grad_norm": 0.3240249752998352, "learning_rate": 2.7492945231748003e-05, "loss": 0.0279, "step": 133240 }, { "grad_norm": 0.3475547432899475, "learning_rate": 2.7485563148549974e-05, "loss": 0.0312, "step": 133250 }, { "grad_norm": 0.32769596576690674, "learning_rate": 2.7478181680888145e-05, "loss": 0.0269, "step": 133260 }, { "grad_norm": 0.315581738948822, "learning_rate": 2.7470800828964387e-05, "loss": 0.0177, "step": 133270 }, { "grad_norm": 0.40524834394454956, "learning_rate": 2.7463420592980455e-05, "loss": 0.0174, "step": 133280 }, { "grad_norm": 0.348628431558609, "learning_rate": 2.7456040973138154e-05, "loss": 0.0277, "step": 133290 }, { "grad_norm": 0.32353484630584717, "learning_rate": 2.7448661969639222e-05, "loss": 0.0239, "step": 133300 }, { "grad_norm": 0.3667260706424713, "learning_rate": 2.7441283582685384e-05, "loss": 0.0195, "step": 133310 }, { "grad_norm": 0.45311790704727173, "learning_rate": 2.7433905812478393e-05, "loss": 0.0203, "step": 133320 }, { "grad_norm": 0.30484911799430847, "learning_rate": 2.742652865921994e-05, "loss": 0.0213, "step": 133330 }, { "grad_norm": 0.5745337009429932, "learning_rate": 2.7419152123111718e-05, "loss": 0.024, "step": 133340 }, { "grad_norm": 0.3757386803627014, "learning_rate": 2.741177620435537e-05, "loss": 0.0213, "step": 133350 }, { "grad_norm": 0.3541036546230316, "learning_rate": 2.7404400903152578e-05, "loss": 0.0268, "step": 133360 }, { "grad_norm": 0.3689286410808563, "learning_rate": 2.739702621970499e-05, "loss": 0.02, "step": 133370 }, { "grad_norm": 0.42017942667007446, "learning_rate": 2.738965215421422e-05, "loss": 0.0178, "step": 133380 }, { "grad_norm": 1.584399938583374, "learning_rate": 2.738227870688186e-05, "loss": 0.0233, "step": 133390 }, { "grad_norm": 0.3904276490211487, "learning_rate": 2.7374905877909492e-05, "loss": 0.0234, "step": 133400 }, { "grad_norm": 0.3520102798938751, "learning_rate": 2.7367533667498712e-05, "loss": 0.0286, "step": 133410 }, { "grad_norm": 0.34472960233688354, "learning_rate": 2.736016207585106e-05, "loss": 0.022, "step": 133420 }, { "grad_norm": 0.5656941533088684, "learning_rate": 2.7352791103168046e-05, "loss": 0.022, "step": 133430 }, { "grad_norm": 0.3466922640800476, "learning_rate": 2.7345420749651242e-05, "loss": 0.0169, "step": 133440 }, { "grad_norm": 1.2571992874145508, "learning_rate": 2.7338051015502096e-05, "loss": 0.0278, "step": 133450 }, { "grad_norm": 0.36997082829475403, "learning_rate": 2.7330681900922166e-05, "loss": 0.0181, "step": 133460 }, { "grad_norm": 0.38305068016052246, "learning_rate": 2.732331340611284e-05, "loss": 0.0217, "step": 133470 }, { "grad_norm": 0.8734263181686401, "learning_rate": 2.73159455312756e-05, "loss": 0.0161, "step": 133480 }, { "grad_norm": 0.3386831283569336, "learning_rate": 2.7308578276611912e-05, "loss": 0.0297, "step": 133490 }, { "grad_norm": 0.6295902729034424, "learning_rate": 2.7301211642323165e-05, "loss": 0.0165, "step": 133500 }, { "grad_norm": 0.6462217569351196, "learning_rate": 2.729384562861077e-05, "loss": 0.023, "step": 133510 }, { "grad_norm": 0.4506627917289734, "learning_rate": 2.728648023567608e-05, "loss": 0.0314, "step": 133520 }, { "grad_norm": 0.3084905445575714, "learning_rate": 2.7279115463720505e-05, "loss": 0.015, "step": 133530 }, { "grad_norm": 0.5021024942398071, "learning_rate": 2.7271751312945387e-05, "loss": 0.0225, "step": 133540 }, { "grad_norm": 0.43373724818229675, "learning_rate": 2.7264387783552026e-05, "loss": 0.0164, "step": 133550 }, { "grad_norm": 0.28576183319091797, "learning_rate": 2.7257024875741776e-05, "loss": 0.0194, "step": 133560 }, { "grad_norm": 0.4645770490169525, "learning_rate": 2.724966258971591e-05, "loss": 0.0212, "step": 133570 }, { "grad_norm": 0.40908095240592957, "learning_rate": 2.7242300925675733e-05, "loss": 0.0186, "step": 133580 }, { "grad_norm": 0.42493847012519836, "learning_rate": 2.7234939883822508e-05, "loss": 0.0392, "step": 133590 }, { "grad_norm": 0.428400456905365, "learning_rate": 2.7227579464357444e-05, "loss": 0.0393, "step": 133600 }, { "grad_norm": 0.41999444365501404, "learning_rate": 2.7220219667481826e-05, "loss": 0.0235, "step": 133610 }, { "grad_norm": 0.3766058683395386, "learning_rate": 2.7212860493396853e-05, "loss": 0.0219, "step": 133620 }, { "grad_norm": 0.49009594321250916, "learning_rate": 2.72055019423037e-05, "loss": 0.0231, "step": 133630 }, { "grad_norm": 0.15976519882678986, "learning_rate": 2.7198144014403548e-05, "loss": 0.0194, "step": 133640 }, { "grad_norm": 0.4969623386859894, "learning_rate": 2.7190786709897594e-05, "loss": 0.0201, "step": 133650 }, { "grad_norm": 1.0100200176239014, "learning_rate": 2.7183430028986935e-05, "loss": 0.0165, "step": 133660 }, { "grad_norm": 0.29259395599365234, "learning_rate": 2.717607397187275e-05, "loss": 0.028, "step": 133670 }, { "grad_norm": 0.34886839985847473, "learning_rate": 2.716871853875613e-05, "loss": 0.0156, "step": 133680 }, { "grad_norm": 0.2904062271118164, "learning_rate": 2.7161363729838152e-05, "loss": 0.0135, "step": 133690 }, { "grad_norm": 0.31086084246635437, "learning_rate": 2.7154009545319926e-05, "loss": 0.0285, "step": 133700 }, { "grad_norm": 0.27969419956207275, "learning_rate": 2.714665598540249e-05, "loss": 0.0208, "step": 133710 }, { "grad_norm": 0.36833322048187256, "learning_rate": 2.71393030502869e-05, "loss": 0.0191, "step": 133720 }, { "grad_norm": 0.6360757946968079, "learning_rate": 2.713195074017416e-05, "loss": 0.0177, "step": 133730 }, { "grad_norm": 0.6084601879119873, "learning_rate": 2.7124599055265286e-05, "loss": 0.0506, "step": 133740 }, { "grad_norm": 0.5255717635154724, "learning_rate": 2.7117247995761332e-05, "loss": 0.0319, "step": 133750 }, { "grad_norm": 0.7667859792709351, "learning_rate": 2.7109897561863173e-05, "loss": 0.026, "step": 133760 }, { "grad_norm": 1.0709123611450195, "learning_rate": 2.7102547753771833e-05, "loss": 0.0163, "step": 133770 }, { "grad_norm": 0.3742901682853699, "learning_rate": 2.7095198571688217e-05, "loss": 0.0311, "step": 133780 }, { "grad_norm": 0.6630753874778748, "learning_rate": 2.7087850015813287e-05, "loss": 0.02, "step": 133790 }, { "grad_norm": 0.6898211240768433, "learning_rate": 2.7080502086347936e-05, "loss": 0.0229, "step": 133800 }, { "grad_norm": 0.5229199528694153, "learning_rate": 2.7073154783493023e-05, "loss": 0.0197, "step": 133810 }, { "grad_norm": 0.36800503730773926, "learning_rate": 2.7065808107449463e-05, "loss": 0.0243, "step": 133820 }, { "grad_norm": 0.4196782410144806, "learning_rate": 2.70584620584181e-05, "loss": 0.0265, "step": 133830 }, { "grad_norm": 0.30015408992767334, "learning_rate": 2.705111663659976e-05, "loss": 0.0162, "step": 133840 }, { "grad_norm": 0.2787371873855591, "learning_rate": 2.7043771842195248e-05, "loss": 0.0172, "step": 133850 }, { "grad_norm": 0.4167217016220093, "learning_rate": 2.7036427675405395e-05, "loss": 0.0158, "step": 133860 }, { "grad_norm": 0.5600649118423462, "learning_rate": 2.7029084136430992e-05, "loss": 0.0269, "step": 133870 }, { "grad_norm": 0.2952806055545807, "learning_rate": 2.7021741225472806e-05, "loss": 0.0211, "step": 133880 }, { "grad_norm": 0.4150060713291168, "learning_rate": 2.7014398942731588e-05, "loss": 0.0163, "step": 133890 }, { "grad_norm": 0.7010473012924194, "learning_rate": 2.7007057288408044e-05, "loss": 0.0246, "step": 133900 }, { "grad_norm": 0.7033064365386963, "learning_rate": 2.6999716262702934e-05, "loss": 0.0255, "step": 133910 }, { "grad_norm": 0.6712671518325806, "learning_rate": 2.6992375865816944e-05, "loss": 0.0174, "step": 133920 }, { "grad_norm": 0.3639466464519501, "learning_rate": 2.698503609795073e-05, "loss": 0.0192, "step": 133930 }, { "grad_norm": 0.6231164336204529, "learning_rate": 2.6977696959305e-05, "loss": 0.0247, "step": 133940 }, { "grad_norm": 0.22378213703632355, "learning_rate": 2.697035845008037e-05, "loss": 0.0165, "step": 133950 }, { "grad_norm": 0.36940500140190125, "learning_rate": 2.6963020570477527e-05, "loss": 0.0221, "step": 133960 }, { "grad_norm": 0.28653833270072937, "learning_rate": 2.695568332069701e-05, "loss": 0.0187, "step": 133970 }, { "grad_norm": 0.30973151326179504, "learning_rate": 2.6948346700939452e-05, "loss": 0.0168, "step": 133980 }, { "grad_norm": 0.5060839653015137, "learning_rate": 2.6941010711405458e-05, "loss": 0.0238, "step": 133990 }, { "grad_norm": 0.27609726786613464, "learning_rate": 2.6933675352295563e-05, "loss": 0.0216, "step": 134000 }, { "grad_norm": 0.39323827624320984, "learning_rate": 2.692634062381032e-05, "loss": 0.0196, "step": 134010 }, { "grad_norm": 0.28305673599243164, "learning_rate": 2.6919006526150237e-05, "loss": 0.0191, "step": 134020 }, { "grad_norm": 0.5201022028923035, "learning_rate": 2.6911673059515863e-05, "loss": 0.0265, "step": 134030 }, { "grad_norm": 0.5050598382949829, "learning_rate": 2.690434022410767e-05, "loss": 0.0222, "step": 134040 }, { "grad_norm": 0.7559916973114014, "learning_rate": 2.6897008020126125e-05, "loss": 0.0336, "step": 134050 }, { "grad_norm": 0.49984338879585266, "learning_rate": 2.6889676447771718e-05, "loss": 0.0209, "step": 134060 }, { "grad_norm": 0.2173244208097458, "learning_rate": 2.6882345507244855e-05, "loss": 0.0205, "step": 134070 }, { "grad_norm": 0.45204460620880127, "learning_rate": 2.6875015198746002e-05, "loss": 0.0185, "step": 134080 }, { "grad_norm": 0.546170711517334, "learning_rate": 2.686768552247555e-05, "loss": 0.0184, "step": 134090 }, { "grad_norm": 0.9339386224746704, "learning_rate": 2.686035647863388e-05, "loss": 0.0453, "step": 134100 }, { "grad_norm": 0.3762741982936859, "learning_rate": 2.6853028067421348e-05, "loss": 0.0222, "step": 134110 }, { "grad_norm": 0.6018154621124268, "learning_rate": 2.684570028903836e-05, "loss": 0.0276, "step": 134120 }, { "grad_norm": 0.5905981659889221, "learning_rate": 2.6838373143685224e-05, "loss": 0.0299, "step": 134130 }, { "grad_norm": 0.43679141998291016, "learning_rate": 2.6831046631562247e-05, "loss": 0.0227, "step": 134140 }, { "grad_norm": 1.4537994861602783, "learning_rate": 2.682372075286977e-05, "loss": 0.0175, "step": 134150 }, { "grad_norm": 0.36109238862991333, "learning_rate": 2.6816395507808035e-05, "loss": 0.0142, "step": 134160 }, { "grad_norm": 0.41031816601753235, "learning_rate": 2.6809070896577366e-05, "loss": 0.0209, "step": 134170 }, { "grad_norm": 0.49072808027267456, "learning_rate": 2.6801746919377977e-05, "loss": 0.0206, "step": 134180 }, { "grad_norm": 0.29733791947364807, "learning_rate": 2.67944235764101e-05, "loss": 0.0217, "step": 134190 }, { "grad_norm": 0.5234381556510925, "learning_rate": 2.6787100867873983e-05, "loss": 0.023, "step": 134200 }, { "grad_norm": 0.33398860692977905, "learning_rate": 2.67797787939698e-05, "loss": 0.0272, "step": 134210 }, { "grad_norm": 0.5555232167243958, "learning_rate": 2.6772457354897752e-05, "loss": 0.0229, "step": 134220 }, { "grad_norm": 0.2490633875131607, "learning_rate": 2.676513655085797e-05, "loss": 0.0236, "step": 134230 }, { "grad_norm": 0.6003566384315491, "learning_rate": 2.6757816382050627e-05, "loss": 0.0263, "step": 134240 }, { "grad_norm": 0.38505205512046814, "learning_rate": 2.675049684867589e-05, "loss": 0.0249, "step": 134250 }, { "grad_norm": 0.5120921730995178, "learning_rate": 2.6743177950933795e-05, "loss": 0.0231, "step": 134260 }, { "grad_norm": 0.20434722304344177, "learning_rate": 2.6735859689024494e-05, "loss": 0.0199, "step": 134270 }, { "grad_norm": 0.7518441081047058, "learning_rate": 2.6728542063148032e-05, "loss": 0.0325, "step": 134280 }, { "grad_norm": 0.28371578454971313, "learning_rate": 2.672122507350451e-05, "loss": 0.0221, "step": 134290 }, { "grad_norm": 0.28843238949775696, "learning_rate": 2.6713908720293945e-05, "loss": 0.0188, "step": 134300 }, { "grad_norm": 0.4057159721851349, "learning_rate": 2.670659300371635e-05, "loss": 0.0154, "step": 134310 }, { "grad_norm": 0.6224492192268372, "learning_rate": 2.6699277923971776e-05, "loss": 0.0198, "step": 134320 }, { "grad_norm": 0.11423088610172272, "learning_rate": 2.6691963481260184e-05, "loss": 0.015, "step": 134330 }, { "grad_norm": 0.3619573712348938, "learning_rate": 2.6684649675781553e-05, "loss": 0.0196, "step": 134340 }, { "grad_norm": 0.5089856386184692, "learning_rate": 2.6677336507735823e-05, "loss": 0.0167, "step": 134350 }, { "grad_norm": 0.40273433923721313, "learning_rate": 2.6670023977322972e-05, "loss": 0.0378, "step": 134360 }, { "grad_norm": 0.4160023033618927, "learning_rate": 2.666271208474288e-05, "loss": 0.0301, "step": 134370 }, { "grad_norm": 0.4315737187862396, "learning_rate": 2.6655400830195486e-05, "loss": 0.0273, "step": 134380 }, { "grad_norm": 0.2921612858772278, "learning_rate": 2.6648090213880674e-05, "loss": 0.0169, "step": 134390 }, { "grad_norm": 0.3098275065422058, "learning_rate": 2.6640780235998276e-05, "loss": 0.0147, "step": 134400 }, { "grad_norm": 0.4746428430080414, "learning_rate": 2.663347089674819e-05, "loss": 0.0292, "step": 134410 }, { "grad_norm": 0.3941071331501007, "learning_rate": 2.6626162196330235e-05, "loss": 0.0215, "step": 134420 }, { "grad_norm": 0.2698507606983185, "learning_rate": 2.6618854134944205e-05, "loss": 0.0326, "step": 134430 }, { "grad_norm": 0.790649950504303, "learning_rate": 2.6611546712789938e-05, "loss": 0.0237, "step": 134440 }, { "grad_norm": 0.3305165469646454, "learning_rate": 2.6604239930067176e-05, "loss": 0.0421, "step": 134450 }, { "grad_norm": 0.48017019033432007, "learning_rate": 2.6596933786975743e-05, "loss": 0.0247, "step": 134460 }, { "grad_norm": 0.21858865022659302, "learning_rate": 2.6589628283715305e-05, "loss": 0.021, "step": 134470 }, { "grad_norm": 0.475116103887558, "learning_rate": 2.6582323420485665e-05, "loss": 0.0267, "step": 134480 }, { "grad_norm": 0.3907869756221771, "learning_rate": 2.6575019197486478e-05, "loss": 0.0236, "step": 134490 }, { "grad_norm": 0.4893021881580353, "learning_rate": 2.6567715614917487e-05, "loss": 0.0242, "step": 134500 }, { "grad_norm": 0.4532909095287323, "learning_rate": 2.6560412672978353e-05, "loss": 0.0233, "step": 134510 }, { "grad_norm": 0.28983578085899353, "learning_rate": 2.6553110371868706e-05, "loss": 0.0205, "step": 134520 }, { "grad_norm": 0.22292977571487427, "learning_rate": 2.6545808711788234e-05, "loss": 0.022, "step": 134530 }, { "grad_norm": 0.4738655388355255, "learning_rate": 2.653850769293654e-05, "loss": 0.0231, "step": 134540 }, { "grad_norm": 0.4250527620315552, "learning_rate": 2.653120731551323e-05, "loss": 0.0189, "step": 134550 }, { "grad_norm": 0.438642293214798, "learning_rate": 2.6523907579717878e-05, "loss": 0.027, "step": 134560 }, { "grad_norm": 0.3820166289806366, "learning_rate": 2.6516608485750073e-05, "loss": 0.0355, "step": 134570 }, { "grad_norm": 0.7245982885360718, "learning_rate": 2.6509310033809386e-05, "loss": 0.023, "step": 134580 }, { "grad_norm": 0.26774537563323975, "learning_rate": 2.650201222409534e-05, "loss": 0.0291, "step": 134590 }, { "grad_norm": 0.6070482134819031, "learning_rate": 2.649471505680745e-05, "loss": 0.0291, "step": 134600 }, { "grad_norm": 0.8547091484069824, "learning_rate": 2.6487418532145202e-05, "loss": 0.0148, "step": 134610 }, { "grad_norm": 0.5043979287147522, "learning_rate": 2.648012265030812e-05, "loss": 0.0186, "step": 134620 }, { "grad_norm": 0.562998354434967, "learning_rate": 2.647282741149565e-05, "loss": 0.0257, "step": 134630 }, { "grad_norm": 0.3703847825527191, "learning_rate": 2.6465532815907214e-05, "loss": 0.0318, "step": 134640 }, { "grad_norm": 0.5158290266990662, "learning_rate": 2.645823886374229e-05, "loss": 0.0173, "step": 134650 }, { "grad_norm": 0.527772068977356, "learning_rate": 2.645094555520026e-05, "loss": 0.0187, "step": 134660 }, { "grad_norm": 0.4639836549758911, "learning_rate": 2.6443652890480558e-05, "loss": 0.0366, "step": 134670 }, { "grad_norm": 0.4431595504283905, "learning_rate": 2.6436360869782504e-05, "loss": 0.0284, "step": 134680 }, { "grad_norm": 0.15448567271232605, "learning_rate": 2.6429069493305493e-05, "loss": 0.0249, "step": 134690 }, { "grad_norm": 0.9648197889328003, "learning_rate": 2.6421778761248884e-05, "loss": 0.0178, "step": 134700 }, { "grad_norm": 0.49131330847740173, "learning_rate": 2.641448867381199e-05, "loss": 0.0196, "step": 134710 }, { "grad_norm": 0.31255534291267395, "learning_rate": 2.6407199231194103e-05, "loss": 0.0277, "step": 134720 }, { "grad_norm": 0.37596777081489563, "learning_rate": 2.6399910433594514e-05, "loss": 0.019, "step": 134730 }, { "grad_norm": 0.34314507246017456, "learning_rate": 2.6392622281212532e-05, "loss": 0.0161, "step": 134740 }, { "grad_norm": 0.5687873363494873, "learning_rate": 2.638533477424738e-05, "loss": 0.0219, "step": 134750 }, { "grad_norm": 0.4560721218585968, "learning_rate": 2.6378047912898286e-05, "loss": 0.0347, "step": 134760 }, { "grad_norm": 0.24272674322128296, "learning_rate": 2.637076169736451e-05, "loss": 0.0246, "step": 134770 }, { "grad_norm": 0.4003734290599823, "learning_rate": 2.636347612784521e-05, "loss": 0.0164, "step": 134780 }, { "grad_norm": 0.7997902035713196, "learning_rate": 2.6356191204539614e-05, "loss": 0.025, "step": 134790 }, { "grad_norm": 0.2152824103832245, "learning_rate": 2.6348906927646866e-05, "loss": 0.0207, "step": 134800 }, { "grad_norm": 0.321400910615921, "learning_rate": 2.6341623297366103e-05, "loss": 0.0215, "step": 134810 }, { "grad_norm": 0.39550358057022095, "learning_rate": 2.6334340313896484e-05, "loss": 0.0159, "step": 134820 }, { "grad_norm": 0.3972264230251312, "learning_rate": 2.6327057977437108e-05, "loss": 0.0284, "step": 134830 }, { "grad_norm": 0.32054755091667175, "learning_rate": 2.6319776288187083e-05, "loss": 0.0176, "step": 134840 }, { "grad_norm": 0.4030204117298126, "learning_rate": 2.6312495246345454e-05, "loss": 0.0215, "step": 134850 }, { "grad_norm": 0.7353387475013733, "learning_rate": 2.6305214852111325e-05, "loss": 0.0224, "step": 134860 }, { "grad_norm": 0.2886642813682556, "learning_rate": 2.62979351056837e-05, "loss": 0.0183, "step": 134870 }, { "grad_norm": 0.3319180905818939, "learning_rate": 2.6290656007261642e-05, "loss": 0.0213, "step": 134880 }, { "grad_norm": 0.30881571769714355, "learning_rate": 2.6283377557044144e-05, "loss": 0.0163, "step": 134890 }, { "grad_norm": 0.2849731147289276, "learning_rate": 2.6276099755230176e-05, "loss": 0.0211, "step": 134900 }, { "grad_norm": 0.5791530609130859, "learning_rate": 2.626882260201875e-05, "loss": 0.0164, "step": 134910 }, { "grad_norm": 0.5289086103439331, "learning_rate": 2.6261546097608792e-05, "loss": 0.0163, "step": 134920 }, { "grad_norm": 0.6465117931365967, "learning_rate": 2.6254270242199247e-05, "loss": 0.0218, "step": 134930 }, { "grad_norm": 0.49219074845314026, "learning_rate": 2.6246995035989018e-05, "loss": 0.0167, "step": 134940 }, { "grad_norm": 0.6284700632095337, "learning_rate": 2.6239720479177015e-05, "loss": 0.0248, "step": 134950 }, { "grad_norm": 0.32243019342422485, "learning_rate": 2.623244657196217e-05, "loss": 0.0195, "step": 134960 }, { "grad_norm": 0.32014593482017517, "learning_rate": 2.6225173314543266e-05, "loss": 0.0315, "step": 134970 }, { "grad_norm": 0.30294936895370483, "learning_rate": 2.6217900707119213e-05, "loss": 0.0239, "step": 134980 }, { "grad_norm": 0.5930161476135254, "learning_rate": 2.6210628749888798e-05, "loss": 0.0247, "step": 134990 }, { "grad_norm": 0.567188024520874, "learning_rate": 2.620335744305088e-05, "loss": 0.0208, "step": 135000 }, { "grad_norm": 0.3065260350704193, "learning_rate": 2.619608678680423e-05, "loss": 0.026, "step": 135010 }, { "grad_norm": 0.575390636920929, "learning_rate": 2.61888167813476e-05, "loss": 0.0236, "step": 135020 }, { "grad_norm": 0.2336803823709488, "learning_rate": 2.6181547426879793e-05, "loss": 0.0238, "step": 135030 }, { "grad_norm": 0.7665029168128967, "learning_rate": 2.6174278723599533e-05, "loss": 0.0218, "step": 135040 }, { "grad_norm": 0.320372074842453, "learning_rate": 2.6167010671705543e-05, "loss": 0.0164, "step": 135050 }, { "grad_norm": 0.5019031763076782, "learning_rate": 2.615974327139651e-05, "loss": 0.0335, "step": 135060 }, { "grad_norm": 0.3135424256324768, "learning_rate": 2.615247652287114e-05, "loss": 0.0196, "step": 135070 }, { "grad_norm": 0.795230507850647, "learning_rate": 2.6145210426328114e-05, "loss": 0.0384, "step": 135080 }, { "grad_norm": 0.38897693157196045, "learning_rate": 2.6137944981966078e-05, "loss": 0.0193, "step": 135090 }, { "grad_norm": 0.2796315550804138, "learning_rate": 2.6130680189983665e-05, "loss": 0.0304, "step": 135100 }, { "grad_norm": 1.4293501377105713, "learning_rate": 2.6123416050579463e-05, "loss": 0.0192, "step": 135110 }, { "grad_norm": 0.5251771211624146, "learning_rate": 2.6116152563952118e-05, "loss": 0.0319, "step": 135120 }, { "grad_norm": 0.6020410060882568, "learning_rate": 2.610888973030019e-05, "loss": 0.0277, "step": 135130 }, { "grad_norm": 0.45899146795272827, "learning_rate": 2.6101627549822223e-05, "loss": 0.0212, "step": 135140 }, { "grad_norm": 0.42649325728416443, "learning_rate": 2.60943660227168e-05, "loss": 0.0318, "step": 135150 }, { "grad_norm": 0.7382918000221252, "learning_rate": 2.6087105149182407e-05, "loss": 0.0259, "step": 135160 }, { "grad_norm": 0.35765933990478516, "learning_rate": 2.6079844929417612e-05, "loss": 0.0335, "step": 135170 }, { "grad_norm": 1.1385630369186401, "learning_rate": 2.607258536362084e-05, "loss": 0.024, "step": 135180 }, { "grad_norm": 0.3836027979850769, "learning_rate": 2.6065326451990584e-05, "loss": 0.0239, "step": 135190 }, { "grad_norm": 0.33111315965652466, "learning_rate": 2.605806819472534e-05, "loss": 0.0306, "step": 135200 }, { "grad_norm": 0.2893281579017639, "learning_rate": 2.6050810592023513e-05, "loss": 0.0186, "step": 135210 }, { "grad_norm": 0.5651339292526245, "learning_rate": 2.604355364408353e-05, "loss": 0.0232, "step": 135220 }, { "grad_norm": 0.4783390760421753, "learning_rate": 2.6036297351103768e-05, "loss": 0.0187, "step": 135230 }, { "grad_norm": 0.572471559047699, "learning_rate": 2.6029041713282653e-05, "loss": 0.03, "step": 135240 }, { "grad_norm": 0.583096981048584, "learning_rate": 2.602178673081853e-05, "loss": 0.0181, "step": 135250 }, { "grad_norm": 0.3988472521305084, "learning_rate": 2.601453240390973e-05, "loss": 0.0203, "step": 135260 }, { "grad_norm": 0.2889866828918457, "learning_rate": 2.6007278732754624e-05, "loss": 0.0233, "step": 135270 }, { "grad_norm": 0.7325084805488586, "learning_rate": 2.600002571755149e-05, "loss": 0.0193, "step": 135280 }, { "grad_norm": 0.34894779324531555, "learning_rate": 2.599277335849865e-05, "loss": 0.0183, "step": 135290 }, { "grad_norm": 0.24560430645942688, "learning_rate": 2.5985521655794366e-05, "loss": 0.0196, "step": 135300 }, { "grad_norm": 0.2964843213558197, "learning_rate": 2.59782706096369e-05, "loss": 0.0169, "step": 135310 }, { "grad_norm": 0.232564315199852, "learning_rate": 2.5971020220224474e-05, "loss": 0.0216, "step": 135320 }, { "grad_norm": 8.87320327758789, "learning_rate": 2.5963770487755346e-05, "loss": 0.0438, "step": 135330 }, { "grad_norm": 1.001957654953003, "learning_rate": 2.5956521412427703e-05, "loss": 0.0389, "step": 135340 }, { "grad_norm": 0.2892993986606598, "learning_rate": 2.5949272994439712e-05, "loss": 0.0143, "step": 135350 }, { "grad_norm": 0.2807340621948242, "learning_rate": 2.5942025233989587e-05, "loss": 0.015, "step": 135360 }, { "grad_norm": 0.2456967532634735, "learning_rate": 2.5934778131275427e-05, "loss": 0.0151, "step": 135370 }, { "grad_norm": 0.3575689196586609, "learning_rate": 2.592753168649541e-05, "loss": 0.021, "step": 135380 }, { "grad_norm": 0.7687050700187683, "learning_rate": 2.592028589984764e-05, "loss": 0.0224, "step": 135390 }, { "grad_norm": 0.4194760322570801, "learning_rate": 2.5913040771530182e-05, "loss": 0.0209, "step": 135400 }, { "grad_norm": 0.3957512676715851, "learning_rate": 2.590579630174116e-05, "loss": 0.0225, "step": 135410 }, { "grad_norm": 0.26583707332611084, "learning_rate": 2.589855249067862e-05, "loss": 0.0224, "step": 135420 }, { "grad_norm": 0.6325483322143555, "learning_rate": 2.58913093385406e-05, "loss": 0.0217, "step": 135430 }, { "grad_norm": 0.3547983169555664, "learning_rate": 2.5884066845525102e-05, "loss": 0.0263, "step": 135440 }, { "grad_norm": 0.6092463135719299, "learning_rate": 2.5876825011830158e-05, "loss": 0.0152, "step": 135450 }, { "grad_norm": 0.6295222043991089, "learning_rate": 2.5869583837653793e-05, "loss": 0.0166, "step": 135460 }, { "grad_norm": 0.4121575355529785, "learning_rate": 2.5862343323193904e-05, "loss": 0.0146, "step": 135470 }, { "grad_norm": 0.5131202340126038, "learning_rate": 2.58551034686485e-05, "loss": 0.0182, "step": 135480 }, { "grad_norm": 0.32186973094940186, "learning_rate": 2.5847864274215473e-05, "loss": 0.0235, "step": 135490 }, { "grad_norm": 0.6231058835983276, "learning_rate": 2.5840625740092788e-05, "loss": 0.0346, "step": 135500 }, { "grad_norm": 0.3587934672832489, "learning_rate": 2.5833387866478313e-05, "loss": 0.0211, "step": 135510 }, { "grad_norm": 0.39961835741996765, "learning_rate": 2.5826150653569924e-05, "loss": 0.0217, "step": 135520 }, { "grad_norm": 0.5754046440124512, "learning_rate": 2.5818914101565504e-05, "loss": 0.0244, "step": 135530 }, { "grad_norm": 0.27514126896858215, "learning_rate": 2.5811678210662893e-05, "loss": 0.0136, "step": 135540 }, { "grad_norm": 0.1891646683216095, "learning_rate": 2.5804442981059917e-05, "loss": 0.0121, "step": 135550 }, { "grad_norm": 0.32044896483421326, "learning_rate": 2.5797208412954355e-05, "loss": 0.0229, "step": 135560 }, { "grad_norm": 1.2558948993682861, "learning_rate": 2.5789974506544047e-05, "loss": 0.0314, "step": 135570 }, { "grad_norm": 0.4163420796394348, "learning_rate": 2.578274126202671e-05, "loss": 0.0224, "step": 135580 }, { "grad_norm": 0.4128462076187134, "learning_rate": 2.5775508679600157e-05, "loss": 0.0215, "step": 135590 }, { "grad_norm": 0.2632577121257782, "learning_rate": 2.5768276759462097e-05, "loss": 0.026, "step": 135600 }, { "grad_norm": 0.41107019782066345, "learning_rate": 2.5761045501810222e-05, "loss": 0.0241, "step": 135610 }, { "grad_norm": 0.6137978434562683, "learning_rate": 2.5753814906842278e-05, "loss": 0.0231, "step": 135620 }, { "grad_norm": 0.5451814532279968, "learning_rate": 2.574658497475593e-05, "loss": 0.0179, "step": 135630 }, { "grad_norm": 1.0117136240005493, "learning_rate": 2.5739355705748814e-05, "loss": 0.0427, "step": 135640 }, { "grad_norm": 0.4335381090641022, "learning_rate": 2.573212710001862e-05, "loss": 0.0275, "step": 135650 }, { "grad_norm": 0.22801361978054047, "learning_rate": 2.572489915776293e-05, "loss": 0.0214, "step": 135660 }, { "grad_norm": 0.4399343729019165, "learning_rate": 2.5717671879179423e-05, "loss": 0.0193, "step": 135670 }, { "grad_norm": 0.624625563621521, "learning_rate": 2.57104452644656e-05, "loss": 0.0163, "step": 135680 }, { "grad_norm": 0.47320523858070374, "learning_rate": 2.5703219313819104e-05, "loss": 0.0268, "step": 135690 }, { "grad_norm": 0.3648129105567932, "learning_rate": 2.5695994027437443e-05, "loss": 0.0226, "step": 135700 }, { "grad_norm": 0.36469385027885437, "learning_rate": 2.5688769405518188e-05, "loss": 0.0339, "step": 135710 }, { "grad_norm": 0.4197501540184021, "learning_rate": 2.568154544825885e-05, "loss": 0.0212, "step": 135720 }, { "grad_norm": 1.405770182609558, "learning_rate": 2.56743221558569e-05, "loss": 0.0232, "step": 135730 }, { "grad_norm": 0.4695717990398407, "learning_rate": 2.5667099528509864e-05, "loss": 0.0241, "step": 135740 }, { "grad_norm": 0.2436956763267517, "learning_rate": 2.5659877566415186e-05, "loss": 0.0176, "step": 135750 }, { "grad_norm": 0.365461528301239, "learning_rate": 2.5652656269770313e-05, "loss": 0.0212, "step": 135760 }, { "grad_norm": 0.5297716856002808, "learning_rate": 2.5645435638772647e-05, "loss": 0.0271, "step": 135770 }, { "grad_norm": 0.4254610240459442, "learning_rate": 2.5638215673619627e-05, "loss": 0.0184, "step": 135780 }, { "grad_norm": 0.34861642122268677, "learning_rate": 2.5630996374508653e-05, "loss": 0.0161, "step": 135790 }, { "grad_norm": 0.4735634922981262, "learning_rate": 2.5623777741637085e-05, "loss": 0.0175, "step": 135800 }, { "grad_norm": 0.2959606647491455, "learning_rate": 2.5616559775202276e-05, "loss": 0.0234, "step": 135810 }, { "grad_norm": 1.1751354932785034, "learning_rate": 2.560934247540154e-05, "loss": 0.0222, "step": 135820 }, { "grad_norm": 0.3658336400985718, "learning_rate": 2.5602125842432222e-05, "loss": 0.0152, "step": 135830 }, { "grad_norm": 0.5018882751464844, "learning_rate": 2.559490987649165e-05, "loss": 0.0226, "step": 135840 }, { "grad_norm": 0.45973220467567444, "learning_rate": 2.5587694577777043e-05, "loss": 0.0194, "step": 135850 }, { "grad_norm": 0.3591521084308624, "learning_rate": 2.5580479946485707e-05, "loss": 0.0213, "step": 135860 }, { "grad_norm": 0.32630655169487, "learning_rate": 2.557326598281486e-05, "loss": 0.0189, "step": 135870 }, { "grad_norm": 0.16060568392276764, "learning_rate": 2.5566052686961783e-05, "loss": 0.0188, "step": 135880 }, { "grad_norm": 0.2931232154369354, "learning_rate": 2.55588400591236e-05, "loss": 0.0152, "step": 135890 }, { "grad_norm": 0.3794303834438324, "learning_rate": 2.555162809949756e-05, "loss": 0.0387, "step": 135900 }, { "grad_norm": 0.45265963673591614, "learning_rate": 2.5544416808280837e-05, "loss": 0.0216, "step": 135910 }, { "grad_norm": 0.4282968044281006, "learning_rate": 2.553720618567057e-05, "loss": 0.0247, "step": 135920 }, { "grad_norm": 0.3431854248046875, "learning_rate": 2.5529996231863896e-05, "loss": 0.0195, "step": 135930 }, { "grad_norm": 0.3242965042591095, "learning_rate": 2.552278694705791e-05, "loss": 0.0183, "step": 135940 }, { "grad_norm": 0.34538501501083374, "learning_rate": 2.551557833144976e-05, "loss": 0.0242, "step": 135950 }, { "grad_norm": 0.2866760492324829, "learning_rate": 2.5508370385236476e-05, "loss": 0.0248, "step": 135960 }, { "grad_norm": 0.1814352571964264, "learning_rate": 2.550116310861516e-05, "loss": 0.0377, "step": 135970 }, { "grad_norm": 0.24236637353897095, "learning_rate": 2.5493956501782846e-05, "loss": 0.0199, "step": 135980 }, { "grad_norm": 0.28306737542152405, "learning_rate": 2.5486750564936533e-05, "loss": 0.0236, "step": 135990 }, { "grad_norm": 0.43570223450660706, "learning_rate": 2.5479545298273273e-05, "loss": 0.0332, "step": 136000 }, { "grad_norm": 0.6151817440986633, "learning_rate": 2.5472340701990027e-05, "loss": 0.0203, "step": 136010 }, { "grad_norm": 0.30556127429008484, "learning_rate": 2.546513677628376e-05, "loss": 0.0241, "step": 136020 }, { "grad_norm": 0.30857717990875244, "learning_rate": 2.545793352135145e-05, "loss": 0.0228, "step": 136030 }, { "grad_norm": 0.33341577649116516, "learning_rate": 2.545073093739e-05, "loss": 0.0182, "step": 136040 }, { "grad_norm": 0.754616379737854, "learning_rate": 2.544352902459638e-05, "loss": 0.0205, "step": 136050 }, { "grad_norm": 0.3461916446685791, "learning_rate": 2.543632778316741e-05, "loss": 0.0212, "step": 136060 }, { "grad_norm": 0.26314443349838257, "learning_rate": 2.542912721330004e-05, "loss": 0.022, "step": 136070 }, { "grad_norm": 0.2697409391403198, "learning_rate": 2.5421927315191075e-05, "loss": 0.0155, "step": 136080 }, { "grad_norm": 0.3008742332458496, "learning_rate": 2.54147280890374e-05, "loss": 0.0238, "step": 136090 }, { "grad_norm": 0.24029478430747986, "learning_rate": 2.5407529535035828e-05, "loss": 0.0306, "step": 136100 }, { "grad_norm": 1.022120714187622, "learning_rate": 2.5400331653383136e-05, "loss": 0.0244, "step": 136110 }, { "grad_norm": 0.26003479957580566, "learning_rate": 2.5393134444276156e-05, "loss": 0.0152, "step": 136120 }, { "grad_norm": 0.40608930587768555, "learning_rate": 2.538593790791164e-05, "loss": 0.0178, "step": 136130 }, { "grad_norm": 0.6138625144958496, "learning_rate": 2.5378742044486333e-05, "loss": 0.023, "step": 136140 }, { "grad_norm": 0.5263688564300537, "learning_rate": 2.5371546854196947e-05, "loss": 0.0167, "step": 136150 }, { "grad_norm": 0.34986692667007446, "learning_rate": 2.5364352337240214e-05, "loss": 0.0168, "step": 136160 }, { "grad_norm": 0.6616045832633972, "learning_rate": 2.535715849381286e-05, "loss": 0.024, "step": 136170 }, { "grad_norm": 0.5683114528656006, "learning_rate": 2.534996532411153e-05, "loss": 0.0236, "step": 136180 }, { "grad_norm": 0.5626118779182434, "learning_rate": 2.534277282833289e-05, "loss": 0.0207, "step": 136190 }, { "grad_norm": 0.4563431143760681, "learning_rate": 2.533558100667356e-05, "loss": 0.0243, "step": 136200 }, { "grad_norm": 0.2514645457267761, "learning_rate": 2.532838985933019e-05, "loss": 0.0143, "step": 136210 }, { "grad_norm": 0.7200549244880676, "learning_rate": 2.5321199386499377e-05, "loss": 0.0167, "step": 136220 }, { "grad_norm": 0.32243940234184265, "learning_rate": 2.5314009588377684e-05, "loss": 0.0207, "step": 136230 }, { "grad_norm": 0.44583502411842346, "learning_rate": 2.5306820465161714e-05, "loss": 0.0189, "step": 136240 }, { "grad_norm": 0.5714377760887146, "learning_rate": 2.5299632017047976e-05, "loss": 0.0206, "step": 136250 }, { "grad_norm": 0.4173212945461273, "learning_rate": 2.529244424423306e-05, "loss": 0.0143, "step": 136260 }, { "grad_norm": 0.1589955985546112, "learning_rate": 2.5285257146913388e-05, "loss": 0.0181, "step": 136270 }, { "grad_norm": 0.343440979719162, "learning_rate": 2.527807072528551e-05, "loss": 0.0255, "step": 136280 }, { "grad_norm": 0.2868259847164154, "learning_rate": 2.5270884979545907e-05, "loss": 0.0148, "step": 136290 }, { "grad_norm": 0.5828797817230225, "learning_rate": 2.526369990989102e-05, "loss": 0.026, "step": 136300 }, { "grad_norm": 0.31093519926071167, "learning_rate": 2.525651551651729e-05, "loss": 0.0217, "step": 136310 }, { "grad_norm": 0.6731470227241516, "learning_rate": 2.5249331799621102e-05, "loss": 0.0198, "step": 136320 }, { "grad_norm": 0.6182685494422913, "learning_rate": 2.52421487593989e-05, "loss": 0.0347, "step": 136330 }, { "grad_norm": 0.464964896440506, "learning_rate": 2.5234966396047055e-05, "loss": 0.0223, "step": 136340 }, { "grad_norm": 0.8525711297988892, "learning_rate": 2.522778470976191e-05, "loss": 0.0205, "step": 136350 }, { "grad_norm": 0.4134300947189331, "learning_rate": 2.5220603700739842e-05, "loss": 0.0236, "step": 136360 }, { "grad_norm": 0.7140554785728455, "learning_rate": 2.5213423369177136e-05, "loss": 0.0224, "step": 136370 }, { "grad_norm": 0.282000333070755, "learning_rate": 2.5206243715270146e-05, "loss": 0.0183, "step": 136380 }, { "grad_norm": 0.21670222282409668, "learning_rate": 2.5199064739215132e-05, "loss": 0.028, "step": 136390 }, { "grad_norm": 0.932401716709137, "learning_rate": 2.5191886441208356e-05, "loss": 0.0152, "step": 136400 }, { "grad_norm": 0.2821008265018463, "learning_rate": 2.5184708821446106e-05, "loss": 0.0236, "step": 136410 }, { "grad_norm": 0.4824688732624054, "learning_rate": 2.5177531880124596e-05, "loss": 0.0196, "step": 136420 }, { "grad_norm": 0.330484002828598, "learning_rate": 2.5170355617440034e-05, "loss": 0.0249, "step": 136430 }, { "grad_norm": 0.873049259185791, "learning_rate": 2.5163180033588606e-05, "loss": 0.0166, "step": 136440 }, { "grad_norm": 0.5631118416786194, "learning_rate": 2.5156005128766526e-05, "loss": 0.0266, "step": 136450 }, { "grad_norm": 0.2893696427345276, "learning_rate": 2.5148830903169908e-05, "loss": 0.0191, "step": 136460 }, { "grad_norm": 0.47138169407844543, "learning_rate": 2.514165735699493e-05, "loss": 0.0222, "step": 136470 }, { "grad_norm": 0.20844164490699768, "learning_rate": 2.5134484490437715e-05, "loss": 0.0185, "step": 136480 }, { "grad_norm": 0.30998164415359497, "learning_rate": 2.5127312303694324e-05, "loss": 0.0195, "step": 136490 }, { "grad_norm": 0.4857632517814636, "learning_rate": 2.512014079696089e-05, "loss": 0.0226, "step": 136500 }, { "grad_norm": 0.3800557255744934, "learning_rate": 2.5112969970433464e-05, "loss": 0.0131, "step": 136510 }, { "grad_norm": 0.2772051692008972, "learning_rate": 2.5105799824308083e-05, "loss": 0.0188, "step": 136520 }, { "grad_norm": 0.2940608561038971, "learning_rate": 2.5098630358780774e-05, "loss": 0.0251, "step": 136530 }, { "grad_norm": 0.2137804925441742, "learning_rate": 2.509146157404755e-05, "loss": 0.0267, "step": 136540 }, { "grad_norm": 0.3397390842437744, "learning_rate": 2.5084293470304443e-05, "loss": 0.0452, "step": 136550 }, { "grad_norm": 0.8911265134811401, "learning_rate": 2.5077126047747358e-05, "loss": 0.0231, "step": 136560 }, { "grad_norm": 0.29674455523490906, "learning_rate": 2.5069959306572306e-05, "loss": 0.0287, "step": 136570 }, { "grad_norm": 0.35906150937080383, "learning_rate": 2.506279324697518e-05, "loss": 0.0192, "step": 136580 }, { "grad_norm": 0.8260379433631897, "learning_rate": 2.5055627869151934e-05, "loss": 0.019, "step": 136590 }, { "grad_norm": 0.5134591460227966, "learning_rate": 2.5048463173298453e-05, "loss": 0.0298, "step": 136600 }, { "grad_norm": 0.4131214916706085, "learning_rate": 2.5041299159610593e-05, "loss": 0.0205, "step": 136610 }, { "grad_norm": 0.2732847332954407, "learning_rate": 2.5034135828284256e-05, "loss": 0.0258, "step": 136620 }, { "grad_norm": 0.5963036417961121, "learning_rate": 2.5026973179515273e-05, "loss": 0.0199, "step": 136630 }, { "grad_norm": 0.33507031202316284, "learning_rate": 2.5019811213499456e-05, "loss": 0.0178, "step": 136640 }, { "grad_norm": 0.28687357902526855, "learning_rate": 2.5012649930432596e-05, "loss": 0.0212, "step": 136650 }, { "grad_norm": 0.31485241651535034, "learning_rate": 2.50054893305105e-05, "loss": 0.0238, "step": 136660 }, { "grad_norm": 0.6383006572723389, "learning_rate": 2.4998329413928962e-05, "loss": 0.0214, "step": 136670 }, { "grad_norm": 0.6826855540275574, "learning_rate": 2.49911701808837e-05, "loss": 0.02, "step": 136680 }, { "grad_norm": 0.4662032127380371, "learning_rate": 2.498401163157045e-05, "loss": 0.0137, "step": 136690 }, { "grad_norm": 0.48760491609573364, "learning_rate": 2.4976853766184905e-05, "loss": 0.0189, "step": 136700 }, { "grad_norm": 0.39389804005622864, "learning_rate": 2.496969658492279e-05, "loss": 0.0212, "step": 136710 }, { "grad_norm": 0.37726661562919617, "learning_rate": 2.4962540087979774e-05, "loss": 0.0171, "step": 136720 }, { "grad_norm": 0.4008406698703766, "learning_rate": 2.4955384275551484e-05, "loss": 0.0138, "step": 136730 }, { "grad_norm": 0.33008626103401184, "learning_rate": 2.49482291478336e-05, "loss": 0.0288, "step": 136740 }, { "grad_norm": 0.42188379168510437, "learning_rate": 2.4941074705021707e-05, "loss": 0.0221, "step": 136750 }, { "grad_norm": 0.2790685296058655, "learning_rate": 2.4933920947311447e-05, "loss": 0.0318, "step": 136760 }, { "grad_norm": 0.3216865062713623, "learning_rate": 2.4926767874898332e-05, "loss": 0.0196, "step": 136770 }, { "grad_norm": 0.3868141174316406, "learning_rate": 2.4919615487977987e-05, "loss": 0.0141, "step": 136780 }, { "grad_norm": 0.4022132158279419, "learning_rate": 2.491246378674591e-05, "loss": 0.0158, "step": 136790 }, { "grad_norm": 0.656608521938324, "learning_rate": 2.4905312771397672e-05, "loss": 0.0228, "step": 136800 }, { "grad_norm": 0.44982999563217163, "learning_rate": 2.4898162442128753e-05, "loss": 0.0273, "step": 136810 }, { "grad_norm": 1.4182275533676147, "learning_rate": 2.4891012799134622e-05, "loss": 0.0235, "step": 136820 }, { "grad_norm": 0.5075586438179016, "learning_rate": 2.4883863842610788e-05, "loss": 0.0193, "step": 136830 }, { "grad_norm": 0.21156193315982819, "learning_rate": 2.487671557275268e-05, "loss": 0.0195, "step": 136840 }, { "grad_norm": 0.32516616582870483, "learning_rate": 2.486956798975571e-05, "loss": 0.0193, "step": 136850 }, { "grad_norm": 0.5406503677368164, "learning_rate": 2.486242109381533e-05, "loss": 0.0271, "step": 136860 }, { "grad_norm": 0.4746715724468231, "learning_rate": 2.4855274885126893e-05, "loss": 0.0201, "step": 136870 }, { "grad_norm": 0.300245076417923, "learning_rate": 2.4848129363885818e-05, "loss": 0.029, "step": 136880 }, { "grad_norm": 0.4780164062976837, "learning_rate": 2.4840984530287438e-05, "loss": 0.0247, "step": 136890 }, { "grad_norm": 0.41957247257232666, "learning_rate": 2.4833840384527085e-05, "loss": 0.019, "step": 136900 }, { "grad_norm": 0.2897111773490906, "learning_rate": 2.482669692680006e-05, "loss": 0.0199, "step": 136910 }, { "grad_norm": 1.4849025011062622, "learning_rate": 2.4819554157301705e-05, "loss": 0.0338, "step": 136920 }, { "grad_norm": 0.2500579357147217, "learning_rate": 2.4812412076227283e-05, "loss": 0.0169, "step": 136930 }, { "grad_norm": 0.4033868908882141, "learning_rate": 2.480527068377203e-05, "loss": 0.0222, "step": 136940 }, { "grad_norm": 0.2814151644706726, "learning_rate": 2.479812998013123e-05, "loss": 0.0175, "step": 136950 }, { "grad_norm": 0.32897552847862244, "learning_rate": 2.4790989965500073e-05, "loss": 0.0232, "step": 136960 }, { "grad_norm": 0.4604320526123047, "learning_rate": 2.4783850640073814e-05, "loss": 0.0167, "step": 136970 }, { "grad_norm": 0.26379817724227905, "learning_rate": 2.477671200404757e-05, "loss": 0.024, "step": 136980 }, { "grad_norm": 0.7210800647735596, "learning_rate": 2.476957405761654e-05, "loss": 0.0323, "step": 136990 }, { "grad_norm": 0.5789843201637268, "learning_rate": 2.4762436800975897e-05, "loss": 0.0235, "step": 137000 }, { "grad_norm": 0.7648786902427673, "learning_rate": 2.4755300234320748e-05, "loss": 0.0291, "step": 137010 }, { "grad_norm": 0.6941007971763611, "learning_rate": 2.4748164357846215e-05, "loss": 0.028, "step": 137020 }, { "grad_norm": 1.5685995817184448, "learning_rate": 2.4741029171747353e-05, "loss": 0.0294, "step": 137030 }, { "grad_norm": 0.5349649786949158, "learning_rate": 2.473389467621927e-05, "loss": 0.0244, "step": 137040 }, { "grad_norm": 0.6381176114082336, "learning_rate": 2.4726760871457057e-05, "loss": 0.0298, "step": 137050 }, { "grad_norm": 0.48337483406066895, "learning_rate": 2.471962775765566e-05, "loss": 0.0302, "step": 137060 }, { "grad_norm": 0.8900501728057861, "learning_rate": 2.4712495335010168e-05, "loss": 0.0368, "step": 137070 }, { "grad_norm": 0.28021374344825745, "learning_rate": 2.470536360371553e-05, "loss": 0.0177, "step": 137080 }, { "grad_norm": 0.4245445132255554, "learning_rate": 2.4698232563966765e-05, "loss": 0.0313, "step": 137090 }, { "grad_norm": 0.4160219728946686, "learning_rate": 2.4691102215958823e-05, "loss": 0.0143, "step": 137100 }, { "grad_norm": 0.36883190274238586, "learning_rate": 2.4683972559886616e-05, "loss": 0.0194, "step": 137110 }, { "grad_norm": 0.2915154993534088, "learning_rate": 2.467684359594511e-05, "loss": 0.0194, "step": 137120 }, { "grad_norm": 0.20702743530273438, "learning_rate": 2.4669715324329186e-05, "loss": 0.0154, "step": 137130 }, { "grad_norm": 0.19898997247219086, "learning_rate": 2.4662587745233723e-05, "loss": 0.0196, "step": 137140 }, { "grad_norm": 0.5297536253929138, "learning_rate": 2.465546085885358e-05, "loss": 0.0209, "step": 137150 }, { "grad_norm": 0.47843366861343384, "learning_rate": 2.4648334665383634e-05, "loss": 0.0177, "step": 137160 }, { "grad_norm": 0.23857617378234863, "learning_rate": 2.464120916501867e-05, "loss": 0.023, "step": 137170 }, { "grad_norm": 0.45864686369895935, "learning_rate": 2.4634084357953547e-05, "loss": 0.018, "step": 137180 }, { "grad_norm": 0.8874956369400024, "learning_rate": 2.4626960244383022e-05, "loss": 0.0295, "step": 137190 }, { "grad_norm": 0.3605823814868927, "learning_rate": 2.4619836824501853e-05, "loss": 0.0205, "step": 137200 }, { "grad_norm": 1.0783635377883911, "learning_rate": 2.461271409850483e-05, "loss": 0.0269, "step": 137210 }, { "grad_norm": 0.4134065806865692, "learning_rate": 2.4605592066586668e-05, "loss": 0.0185, "step": 137220 }, { "grad_norm": 0.2763848602771759, "learning_rate": 2.459847072894206e-05, "loss": 0.0249, "step": 137230 }, { "grad_norm": 0.3483864665031433, "learning_rate": 2.4591350085765735e-05, "loss": 0.0168, "step": 137240 }, { "grad_norm": 0.3113669753074646, "learning_rate": 2.458423013725234e-05, "loss": 0.0193, "step": 137250 }, { "grad_norm": 0.1892010122537613, "learning_rate": 2.4577110883596577e-05, "loss": 0.0153, "step": 137260 }, { "grad_norm": 0.38522958755493164, "learning_rate": 2.4569992324993015e-05, "loss": 0.0203, "step": 137270 }, { "grad_norm": 0.5534731149673462, "learning_rate": 2.456287446163633e-05, "loss": 0.0182, "step": 137280 }, { "grad_norm": 0.28090474009513855, "learning_rate": 2.4555757293721084e-05, "loss": 0.0176, "step": 137290 }, { "grad_norm": 0.3789772689342499, "learning_rate": 2.4548640821441887e-05, "loss": 0.0229, "step": 137300 }, { "grad_norm": 0.3842330873012543, "learning_rate": 2.454152504499329e-05, "loss": 0.0185, "step": 137310 }, { "grad_norm": 0.718132734298706, "learning_rate": 2.4534409964569815e-05, "loss": 0.0302, "step": 137320 }, { "grad_norm": 0.4114300608634949, "learning_rate": 2.4527295580366028e-05, "loss": 0.0243, "step": 137330 }, { "grad_norm": 0.39696091413497925, "learning_rate": 2.4520181892576405e-05, "loss": 0.0203, "step": 137340 }, { "grad_norm": 0.5409038066864014, "learning_rate": 2.451306890139544e-05, "loss": 0.0279, "step": 137350 }, { "grad_norm": 0.17126309871673584, "learning_rate": 2.4505956607017578e-05, "loss": 0.0179, "step": 137360 }, { "grad_norm": 0.29835036396980286, "learning_rate": 2.4498845009637283e-05, "loss": 0.0176, "step": 137370 }, { "grad_norm": 0.4364735186100006, "learning_rate": 2.4491734109449004e-05, "loss": 0.0279, "step": 137380 }, { "grad_norm": 1.2693164348602295, "learning_rate": 2.4484623906647132e-05, "loss": 0.0212, "step": 137390 }, { "grad_norm": 0.2511357367038727, "learning_rate": 2.4477514401426053e-05, "loss": 0.0306, "step": 137400 }, { "grad_norm": 0.38046208024024963, "learning_rate": 2.447040559398013e-05, "loss": 0.0197, "step": 137410 }, { "grad_norm": 0.8033462166786194, "learning_rate": 2.446329748450374e-05, "loss": 0.0333, "step": 137420 }, { "grad_norm": 0.42785897850990295, "learning_rate": 2.4456190073191203e-05, "loss": 0.0222, "step": 137430 }, { "grad_norm": 0.24809277057647705, "learning_rate": 2.4449083360236817e-05, "loss": 0.021, "step": 137440 }, { "grad_norm": 0.5619657635688782, "learning_rate": 2.4441977345834914e-05, "loss": 0.0209, "step": 137450 }, { "grad_norm": 0.41119876503944397, "learning_rate": 2.443487203017973e-05, "loss": 0.014, "step": 137460 }, { "grad_norm": 0.28197577595710754, "learning_rate": 2.442776741346557e-05, "loss": 0.0158, "step": 137470 }, { "grad_norm": 0.2521912455558777, "learning_rate": 2.442066349588661e-05, "loss": 0.0178, "step": 137480 }, { "grad_norm": 0.416657030582428, "learning_rate": 2.44135602776371e-05, "loss": 0.018, "step": 137490 }, { "grad_norm": 1.7122541666030884, "learning_rate": 2.4406457758911254e-05, "loss": 0.0282, "step": 137500 }, { "grad_norm": 0.24164029955863953, "learning_rate": 2.4399355939903245e-05, "loss": 0.0125, "step": 137510 }, { "grad_norm": 0.38817930221557617, "learning_rate": 2.4392254820807226e-05, "loss": 0.026, "step": 137520 }, { "grad_norm": 0.2851036489009857, "learning_rate": 2.4385154401817318e-05, "loss": 0.0209, "step": 137530 }, { "grad_norm": 0.36507660150527954, "learning_rate": 2.437805468312768e-05, "loss": 0.0211, "step": 137540 }, { "grad_norm": 0.5152348875999451, "learning_rate": 2.437095566493241e-05, "loss": 0.0245, "step": 137550 }, { "grad_norm": 0.4592231214046478, "learning_rate": 2.436385734742556e-05, "loss": 0.0184, "step": 137560 }, { "grad_norm": 0.6201115250587463, "learning_rate": 2.4356759730801245e-05, "loss": 0.0348, "step": 137570 }, { "grad_norm": 0.47728052735328674, "learning_rate": 2.4349662815253455e-05, "loss": 0.0215, "step": 137580 }, { "grad_norm": 0.33811652660369873, "learning_rate": 2.4342566600976273e-05, "loss": 0.0168, "step": 137590 }, { "grad_norm": 0.2906545102596283, "learning_rate": 2.433547108816368e-05, "loss": 0.0208, "step": 137600 }, { "grad_norm": 0.41091352701187134, "learning_rate": 2.4328376277009647e-05, "loss": 0.0385, "step": 137610 }, { "grad_norm": 0.3417311906814575, "learning_rate": 2.432128216770818e-05, "loss": 0.0195, "step": 137620 }, { "grad_norm": 0.2921736538410187, "learning_rate": 2.431418876045321e-05, "loss": 0.023, "step": 137630 }, { "grad_norm": 0.325236439704895, "learning_rate": 2.430709605543867e-05, "loss": 0.018, "step": 137640 }, { "grad_norm": 0.5058897733688354, "learning_rate": 2.430000405285846e-05, "loss": 0.027, "step": 137650 }, { "grad_norm": 0.5745236873626709, "learning_rate": 2.4292912752906494e-05, "loss": 0.02, "step": 137660 }, { "grad_norm": 0.21510517597198486, "learning_rate": 2.428582215577662e-05, "loss": 0.0224, "step": 137670 }, { "grad_norm": 1.1399221420288086, "learning_rate": 2.4278732261662727e-05, "loss": 0.0246, "step": 137680 }, { "grad_norm": 0.6356798410415649, "learning_rate": 2.4271643070758633e-05, "loss": 0.0212, "step": 137690 }, { "grad_norm": 1.1037397384643555, "learning_rate": 2.426455458325813e-05, "loss": 0.0235, "step": 137700 }, { "grad_norm": 0.5347818732261658, "learning_rate": 2.4257466799355055e-05, "loss": 0.0167, "step": 137710 }, { "grad_norm": 0.6057494878768921, "learning_rate": 2.4250379719243167e-05, "loss": 0.0259, "step": 137720 }, { "grad_norm": 0.3478292226791382, "learning_rate": 2.4243293343116224e-05, "loss": 0.0254, "step": 137730 }, { "grad_norm": 0.4640325903892517, "learning_rate": 2.4236207671167948e-05, "loss": 0.024, "step": 137740 }, { "grad_norm": 0.3771439790725708, "learning_rate": 2.4229122703592073e-05, "loss": 0.0213, "step": 137750 }, { "grad_norm": 0.2664586007595062, "learning_rate": 2.422203844058234e-05, "loss": 0.017, "step": 137760 }, { "grad_norm": 0.6128882169723511, "learning_rate": 2.4214954882332353e-05, "loss": 0.0263, "step": 137770 }, { "grad_norm": 0.7102706432342529, "learning_rate": 2.4207872029035835e-05, "loss": 0.0257, "step": 137780 }, { "grad_norm": 0.30529695749282837, "learning_rate": 2.420078988088638e-05, "loss": 0.019, "step": 137790 }, { "grad_norm": 0.46931833028793335, "learning_rate": 2.419370843807766e-05, "loss": 0.0151, "step": 137800 }, { "grad_norm": 0.26625123620033264, "learning_rate": 2.418662770080326e-05, "loss": 0.0206, "step": 137810 }, { "grad_norm": 0.27857911586761475, "learning_rate": 2.417954766925674e-05, "loss": 0.0164, "step": 137820 }, { "grad_norm": 0.4630124270915985, "learning_rate": 2.4172468343631704e-05, "loss": 0.0167, "step": 137830 }, { "grad_norm": 0.34651729464530945, "learning_rate": 2.4165389724121677e-05, "loss": 0.0212, "step": 137840 }, { "grad_norm": 0.35442274808883667, "learning_rate": 2.4158311810920198e-05, "loss": 0.0174, "step": 137850 }, { "grad_norm": 0.310587614774704, "learning_rate": 2.415123460422074e-05, "loss": 0.0206, "step": 137860 }, { "grad_norm": 0.3384197950363159, "learning_rate": 2.414415810421682e-05, "loss": 0.0176, "step": 137870 }, { "grad_norm": 0.23325979709625244, "learning_rate": 2.413708231110192e-05, "loss": 0.0151, "step": 137880 }, { "grad_norm": 0.2709861397743225, "learning_rate": 2.413000722506948e-05, "loss": 0.0206, "step": 137890 }, { "grad_norm": 0.6296613216400146, "learning_rate": 2.412293284631292e-05, "loss": 0.0269, "step": 137900 }, { "grad_norm": 1.0835789442062378, "learning_rate": 2.4115859175025628e-05, "loss": 0.017, "step": 137910 }, { "grad_norm": 0.547764241695404, "learning_rate": 2.4108786211401047e-05, "loss": 0.0223, "step": 137920 }, { "grad_norm": 0.12016919255256653, "learning_rate": 2.4101713955632528e-05, "loss": 0.0176, "step": 137930 }, { "grad_norm": 0.5132116675376892, "learning_rate": 2.4094642407913398e-05, "loss": 0.0576, "step": 137940 }, { "grad_norm": 0.491107314825058, "learning_rate": 2.4087571568437028e-05, "loss": 0.0193, "step": 137950 }, { "grad_norm": 0.5012580156326294, "learning_rate": 2.40805014373967e-05, "loss": 0.0239, "step": 137960 }, { "grad_norm": 0.9842888116836548, "learning_rate": 2.4073432014985764e-05, "loss": 0.0228, "step": 137970 }, { "grad_norm": 0.5835340023040771, "learning_rate": 2.4066363301397415e-05, "loss": 0.0147, "step": 137980 }, { "grad_norm": 0.30826374888420105, "learning_rate": 2.4059295296824958e-05, "loss": 0.0183, "step": 137990 }, { "grad_norm": 0.5392017364501953, "learning_rate": 2.4052228001461636e-05, "loss": 0.0191, "step": 138000 }, { "grad_norm": 0.31962770223617554, "learning_rate": 2.404516141550065e-05, "loss": 0.0212, "step": 138010 }, { "grad_norm": 0.41025176644325256, "learning_rate": 2.4038095539135203e-05, "loss": 0.0235, "step": 138020 }, { "grad_norm": 0.2841476500034332, "learning_rate": 2.403103037255845e-05, "loss": 0.0232, "step": 138030 }, { "grad_norm": 0.260186105966568, "learning_rate": 2.402396591596359e-05, "loss": 0.0253, "step": 138040 }, { "grad_norm": 0.305534690618515, "learning_rate": 2.4016902169543747e-05, "loss": 0.0155, "step": 138050 }, { "grad_norm": 0.5269971489906311, "learning_rate": 2.4009839133492018e-05, "loss": 0.0184, "step": 138060 }, { "grad_norm": 0.20049341022968292, "learning_rate": 2.4002776808001538e-05, "loss": 0.0284, "step": 138070 }, { "grad_norm": 0.4082963466644287, "learning_rate": 2.3995715193265355e-05, "loss": 0.0338, "step": 138080 }, { "grad_norm": 0.18470680713653564, "learning_rate": 2.398865428947657e-05, "loss": 0.02, "step": 138090 }, { "grad_norm": 0.6234345436096191, "learning_rate": 2.3981594096828203e-05, "loss": 0.0245, "step": 138100 }, { "grad_norm": 0.5114167928695679, "learning_rate": 2.3974534615513278e-05, "loss": 0.0186, "step": 138110 }, { "grad_norm": 0.4492151737213135, "learning_rate": 2.3967475845724785e-05, "loss": 0.0182, "step": 138120 }, { "grad_norm": 0.40852609276771545, "learning_rate": 2.3960417787655736e-05, "loss": 0.0186, "step": 138130 }, { "grad_norm": 0.11812612414360046, "learning_rate": 2.3953360441499085e-05, "loss": 0.0195, "step": 138140 }, { "grad_norm": 1.5165622234344482, "learning_rate": 2.394630380744775e-05, "loss": 0.0405, "step": 138150 }, { "grad_norm": 0.6803697347640991, "learning_rate": 2.393924788569471e-05, "loss": 0.0248, "step": 138160 }, { "grad_norm": 0.49981755018234253, "learning_rate": 2.3932192676432813e-05, "loss": 0.0176, "step": 138170 }, { "grad_norm": 0.44531872868537903, "learning_rate": 2.3925138179855022e-05, "loss": 0.0216, "step": 138180 }, { "grad_norm": 0.36286425590515137, "learning_rate": 2.391808439615411e-05, "loss": 0.0173, "step": 138190 }, { "grad_norm": 0.23984889686107635, "learning_rate": 2.391103132552298e-05, "loss": 0.0153, "step": 138200 }, { "grad_norm": 0.5361880660057068, "learning_rate": 2.3903978968154468e-05, "loss": 0.0169, "step": 138210 }, { "grad_norm": 0.6813249588012695, "learning_rate": 2.3896927324241368e-05, "loss": 0.0249, "step": 138220 }, { "grad_norm": 0.3780819773674011, "learning_rate": 2.3889876393976468e-05, "loss": 0.0375, "step": 138230 }, { "grad_norm": 0.4287271499633789, "learning_rate": 2.388282617755252e-05, "loss": 0.0267, "step": 138240 }, { "grad_norm": 0.3800694942474365, "learning_rate": 2.3875776675162292e-05, "loss": 0.0228, "step": 138250 }, { "grad_norm": 0.38290441036224365, "learning_rate": 2.3868727886998555e-05, "loss": 0.0227, "step": 138260 }, { "grad_norm": 0.49463531374931335, "learning_rate": 2.386167981325394e-05, "loss": 0.0283, "step": 138270 }, { "grad_norm": 0.3307836055755615, "learning_rate": 2.38546324541212e-05, "loss": 0.0246, "step": 138280 }, { "grad_norm": 0.5649271607398987, "learning_rate": 2.3847585809792965e-05, "loss": 0.0239, "step": 138290 }, { "grad_norm": 0.30998703837394714, "learning_rate": 2.3840539880461926e-05, "loss": 0.0188, "step": 138300 }, { "grad_norm": 0.3132469654083252, "learning_rate": 2.3833494666320706e-05, "loss": 0.0251, "step": 138310 }, { "grad_norm": 0.37695592641830444, "learning_rate": 2.3826450167561887e-05, "loss": 0.0176, "step": 138320 }, { "grad_norm": 0.1655128300189972, "learning_rate": 2.3819406384378107e-05, "loss": 0.0159, "step": 138330 }, { "grad_norm": 0.6438235640525818, "learning_rate": 2.3812363316961922e-05, "loss": 0.0349, "step": 138340 }, { "grad_norm": 0.5885739326477051, "learning_rate": 2.380532096550589e-05, "loss": 0.0228, "step": 138350 }, { "grad_norm": 0.3505946397781372, "learning_rate": 2.3798279330202518e-05, "loss": 0.026, "step": 138360 }, { "grad_norm": 0.39390331506729126, "learning_rate": 2.3791238411244366e-05, "loss": 0.0232, "step": 138370 }, { "grad_norm": 0.6050373911857605, "learning_rate": 2.3784198208823887e-05, "loss": 0.0212, "step": 138380 }, { "grad_norm": 0.3614589273929596, "learning_rate": 2.37771587231336e-05, "loss": 0.0223, "step": 138390 }, { "grad_norm": 1.1933577060699463, "learning_rate": 2.377011995436595e-05, "loss": 0.02, "step": 138400 }, { "grad_norm": 0.2607761025428772, "learning_rate": 2.376308190271334e-05, "loss": 0.0176, "step": 138410 }, { "grad_norm": 0.36992695927619934, "learning_rate": 2.3756044568368237e-05, "loss": 0.0206, "step": 138420 }, { "grad_norm": 0.41294899582862854, "learning_rate": 2.374900795152301e-05, "loss": 0.0247, "step": 138430 }, { "grad_norm": 0.4786524772644043, "learning_rate": 2.3741972052370025e-05, "loss": 0.0289, "step": 138440 }, { "grad_norm": 0.3804129660129547, "learning_rate": 2.3734936871101687e-05, "loss": 0.0253, "step": 138450 }, { "grad_norm": 0.2458886355161667, "learning_rate": 2.372790240791028e-05, "loss": 0.0159, "step": 138460 }, { "grad_norm": 0.46620315313339233, "learning_rate": 2.3720868662988193e-05, "loss": 0.0205, "step": 138470 }, { "grad_norm": 0.545607328414917, "learning_rate": 2.3713835636527647e-05, "loss": 0.0428, "step": 138480 }, { "grad_norm": 0.6172824501991272, "learning_rate": 2.3706803328720977e-05, "loss": 0.0246, "step": 138490 }, { "grad_norm": 1.184047818183899, "learning_rate": 2.3699771739760406e-05, "loss": 0.0218, "step": 138500 }, { "grad_norm": 0.2265242636203766, "learning_rate": 2.3692740869838215e-05, "loss": 0.017, "step": 138510 }, { "grad_norm": 0.5247446298599243, "learning_rate": 2.368571071914661e-05, "loss": 0.0149, "step": 138520 }, { "grad_norm": 1.001487374305725, "learning_rate": 2.3678681287877763e-05, "loss": 0.0311, "step": 138530 }, { "grad_norm": 0.2848363518714905, "learning_rate": 2.36716525762239e-05, "loss": 0.0255, "step": 138540 }, { "grad_norm": 0.45544418692588806, "learning_rate": 2.3664624584377164e-05, "loss": 0.017, "step": 138550 }, { "grad_norm": 0.33698633313179016, "learning_rate": 2.3657597312529706e-05, "loss": 0.0211, "step": 138560 }, { "grad_norm": 0.2997940182685852, "learning_rate": 2.3650570760873615e-05, "loss": 0.026, "step": 138570 }, { "grad_norm": 0.4209611415863037, "learning_rate": 2.364354492960102e-05, "loss": 0.0154, "step": 138580 }, { "grad_norm": 0.3792248070240021, "learning_rate": 2.363651981890403e-05, "loss": 0.0134, "step": 138590 }, { "grad_norm": 0.2629702091217041, "learning_rate": 2.3629495428974674e-05, "loss": 0.0324, "step": 138600 }, { "grad_norm": 0.4645388722419739, "learning_rate": 2.3622471760005015e-05, "loss": 0.0242, "step": 138610 }, { "grad_norm": 0.42759352922439575, "learning_rate": 2.3615448812187047e-05, "loss": 0.0264, "step": 138620 }, { "grad_norm": 0.8545157313346863, "learning_rate": 2.360842658571281e-05, "loss": 0.0284, "step": 138630 }, { "grad_norm": 0.6381808519363403, "learning_rate": 2.3601405080774286e-05, "loss": 0.0209, "step": 138640 }, { "grad_norm": 0.20956462621688843, "learning_rate": 2.3594384297563405e-05, "loss": 0.0364, "step": 138650 }, { "grad_norm": 0.5394247174263, "learning_rate": 2.358736423627215e-05, "loss": 0.0143, "step": 138660 }, { "grad_norm": 0.6796371340751648, "learning_rate": 2.358034489709242e-05, "loss": 0.0189, "step": 138670 }, { "grad_norm": 0.5321426391601562, "learning_rate": 2.357332628021618e-05, "loss": 0.0217, "step": 138680 }, { "grad_norm": 0.2722914218902588, "learning_rate": 2.3566308385835224e-05, "loss": 0.0275, "step": 138690 }, { "grad_norm": 0.3239484131336212, "learning_rate": 2.355929121414147e-05, "loss": 0.0185, "step": 138700 }, { "grad_norm": 0.6650248169898987, "learning_rate": 2.3552274765326777e-05, "loss": 0.0271, "step": 138710 }, { "grad_norm": 0.2980763614177704, "learning_rate": 2.3545259039582962e-05, "loss": 0.0319, "step": 138720 }, { "grad_norm": 0.9822301864624023, "learning_rate": 2.353824403710182e-05, "loss": 0.0417, "step": 138730 }, { "grad_norm": 0.43875470757484436, "learning_rate": 2.353122975807513e-05, "loss": 0.0232, "step": 138740 }, { "grad_norm": 0.6004987955093384, "learning_rate": 2.3524216202694687e-05, "loss": 0.0198, "step": 138750 }, { "grad_norm": 0.42017704248428345, "learning_rate": 2.3517203371152237e-05, "loss": 0.0224, "step": 138760 }, { "grad_norm": 0.6648881435394287, "learning_rate": 2.351019126363947e-05, "loss": 0.034, "step": 138770 }, { "grad_norm": 0.5120417475700378, "learning_rate": 2.3503179880348147e-05, "loss": 0.0161, "step": 138780 }, { "grad_norm": 0.405117928981781, "learning_rate": 2.349616922146991e-05, "loss": 0.0217, "step": 138790 }, { "grad_norm": 0.3824055790901184, "learning_rate": 2.3489159287196465e-05, "loss": 0.0261, "step": 138800 }, { "grad_norm": 0.18782082200050354, "learning_rate": 2.348215007771945e-05, "loss": 0.0334, "step": 138810 }, { "grad_norm": 0.3449844717979431, "learning_rate": 2.3475141593230472e-05, "loss": 0.0209, "step": 138820 }, { "grad_norm": 0.2908575236797333, "learning_rate": 2.346813383392117e-05, "loss": 0.0192, "step": 138830 }, { "grad_norm": 0.3743360638618469, "learning_rate": 2.3461126799983126e-05, "loss": 0.033, "step": 138840 }, { "grad_norm": 0.20849795639514923, "learning_rate": 2.3454120491607905e-05, "loss": 0.026, "step": 138850 }, { "grad_norm": 0.7581380605697632, "learning_rate": 2.344711490898704e-05, "loss": 0.0178, "step": 138860 }, { "grad_norm": 0.3251528739929199, "learning_rate": 2.344011005231209e-05, "loss": 0.0163, "step": 138870 }, { "grad_norm": 0.6015259623527527, "learning_rate": 2.343310592177454e-05, "loss": 0.021, "step": 138880 }, { "grad_norm": 0.5415794253349304, "learning_rate": 2.342610251756591e-05, "loss": 0.0176, "step": 138890 }, { "grad_norm": 0.3648890256881714, "learning_rate": 2.3419099839877655e-05, "loss": 0.0239, "step": 138900 }, { "grad_norm": 0.2645260989665985, "learning_rate": 2.3412097888901207e-05, "loss": 0.0244, "step": 138910 }, { "grad_norm": 0.26028379797935486, "learning_rate": 2.340509666482803e-05, "loss": 0.0198, "step": 138920 }, { "grad_norm": 0.4514007270336151, "learning_rate": 2.339809616784952e-05, "loss": 0.0172, "step": 138930 }, { "grad_norm": 0.33627039194107056, "learning_rate": 2.339109639815707e-05, "loss": 0.0207, "step": 138940 }, { "grad_norm": 0.2842005491256714, "learning_rate": 2.3384097355942025e-05, "loss": 0.0174, "step": 138950 }, { "grad_norm": 0.3637247085571289, "learning_rate": 2.337709904139576e-05, "loss": 0.023, "step": 138960 }, { "grad_norm": 0.30902615189552307, "learning_rate": 2.3370101454709642e-05, "loss": 0.0215, "step": 138970 }, { "grad_norm": 0.23094257712364197, "learning_rate": 2.3363104596074914e-05, "loss": 0.0156, "step": 138980 }, { "grad_norm": 0.17122632265090942, "learning_rate": 2.3356108465682912e-05, "loss": 0.021, "step": 138990 }, { "grad_norm": 0.21227559447288513, "learning_rate": 2.3349113063724876e-05, "loss": 0.0143, "step": 139000 }, { "grad_norm": 1.1572973728179932, "learning_rate": 2.33421183903921e-05, "loss": 0.0286, "step": 139010 }, { "grad_norm": 0.4948175847530365, "learning_rate": 2.333512444587579e-05, "loss": 0.0234, "step": 139020 }, { "grad_norm": 0.4178679883480072, "learning_rate": 2.332813123036714e-05, "loss": 0.0173, "step": 139030 }, { "grad_norm": 0.33378955721855164, "learning_rate": 2.332113874405738e-05, "loss": 0.0197, "step": 139040 }, { "grad_norm": 0.5243502259254456, "learning_rate": 2.331414698713767e-05, "loss": 0.0197, "step": 139050 }, { "grad_norm": 0.6584689021110535, "learning_rate": 2.3307155959799148e-05, "loss": 0.0203, "step": 139060 }, { "grad_norm": 0.4081437885761261, "learning_rate": 2.3300165662232942e-05, "loss": 0.0249, "step": 139070 }, { "grad_norm": 0.3331068158149719, "learning_rate": 2.329317609463017e-05, "loss": 0.0143, "step": 139080 }, { "grad_norm": 0.538702666759491, "learning_rate": 2.3286187257181953e-05, "loss": 0.0187, "step": 139090 }, { "grad_norm": 0.4170125722885132, "learning_rate": 2.327919915007934e-05, "loss": 0.0196, "step": 139100 }, { "grad_norm": 0.40887516736984253, "learning_rate": 2.3272211773513385e-05, "loss": 0.0173, "step": 139110 }, { "grad_norm": 0.3540034294128418, "learning_rate": 2.32652251276751e-05, "loss": 0.0182, "step": 139120 }, { "grad_norm": 0.5023772120475769, "learning_rate": 2.3258239212755527e-05, "loss": 0.017, "step": 139130 }, { "grad_norm": 0.5334282517433167, "learning_rate": 2.3251254028945657e-05, "loss": 0.0173, "step": 139140 }, { "grad_norm": 0.5065156817436218, "learning_rate": 2.3244269576436433e-05, "loss": 0.0182, "step": 139150 }, { "grad_norm": 0.4620285928249359, "learning_rate": 2.3237285855418845e-05, "loss": 0.0258, "step": 139160 }, { "grad_norm": 0.408724844455719, "learning_rate": 2.3230302866083788e-05, "loss": 0.0188, "step": 139170 }, { "grad_norm": 0.5042911767959595, "learning_rate": 2.322332060862223e-05, "loss": 0.0217, "step": 139180 }, { "grad_norm": 0.5655160546302795, "learning_rate": 2.3216339083224985e-05, "loss": 0.0168, "step": 139190 }, { "grad_norm": 0.3908800780773163, "learning_rate": 2.3209358290082977e-05, "loss": 0.0225, "step": 139200 }, { "grad_norm": 0.6352102160453796, "learning_rate": 2.3202378229387062e-05, "loss": 0.0206, "step": 139210 }, { "grad_norm": 3.822237730026245, "learning_rate": 2.3195398901328057e-05, "loss": 0.0198, "step": 139220 }, { "grad_norm": 0.51912921667099, "learning_rate": 2.318842030609678e-05, "loss": 0.0231, "step": 139230 }, { "grad_norm": 0.46476662158966064, "learning_rate": 2.3181442443884e-05, "loss": 0.017, "step": 139240 }, { "grad_norm": 0.4659985899925232, "learning_rate": 2.3174465314880528e-05, "loss": 0.0186, "step": 139250 }, { "grad_norm": 0.26878657937049866, "learning_rate": 2.316748891927709e-05, "loss": 0.0239, "step": 139260 }, { "grad_norm": 0.7400495409965515, "learning_rate": 2.3160513257264416e-05, "loss": 0.0169, "step": 139270 }, { "grad_norm": 0.40975895524024963, "learning_rate": 2.3153538329033237e-05, "loss": 0.0178, "step": 139280 }, { "grad_norm": 0.3708794116973877, "learning_rate": 2.314656413477422e-05, "loss": 0.0167, "step": 139290 }, { "grad_norm": 0.3817601501941681, "learning_rate": 2.313959067467807e-05, "loss": 0.0242, "step": 139300 }, { "grad_norm": 0.5127875208854675, "learning_rate": 2.3132617948935415e-05, "loss": 0.0204, "step": 139310 }, { "grad_norm": 0.38443613052368164, "learning_rate": 2.3125645957736898e-05, "loss": 0.0148, "step": 139320 }, { "grad_norm": 0.48191046714782715, "learning_rate": 2.3118674701273104e-05, "loss": 0.025, "step": 139330 }, { "grad_norm": 0.672265350818634, "learning_rate": 2.3111704179734657e-05, "loss": 0.0225, "step": 139340 }, { "grad_norm": 0.3407588005065918, "learning_rate": 2.310473439331212e-05, "loss": 0.0197, "step": 139350 }, { "grad_norm": 0.335071861743927, "learning_rate": 2.3097765342196027e-05, "loss": 0.0269, "step": 139360 }, { "grad_norm": 0.44744881987571716, "learning_rate": 2.3090797026576937e-05, "loss": 0.0194, "step": 139370 }, { "grad_norm": 0.22177192568778992, "learning_rate": 2.308382944664533e-05, "loss": 0.0309, "step": 139380 }, { "grad_norm": 0.4157773554325104, "learning_rate": 2.3076862602591746e-05, "loss": 0.021, "step": 139390 }, { "grad_norm": 0.537774920463562, "learning_rate": 2.306989649460659e-05, "loss": 0.0179, "step": 139400 }, { "grad_norm": 0.22748760879039764, "learning_rate": 2.3062931122880348e-05, "loss": 0.0218, "step": 139410 }, { "grad_norm": 0.532048761844635, "learning_rate": 2.305596648760347e-05, "loss": 0.017, "step": 139420 }, { "grad_norm": 0.3248264789581299, "learning_rate": 2.3049002588966346e-05, "loss": 0.0425, "step": 139430 }, { "grad_norm": 0.2268696129322052, "learning_rate": 2.3042039427159363e-05, "loss": 0.0158, "step": 139440 }, { "grad_norm": 0.5430521368980408, "learning_rate": 2.3035077002372874e-05, "loss": 0.0226, "step": 139450 }, { "grad_norm": 0.3282322287559509, "learning_rate": 2.302811531479725e-05, "loss": 0.0138, "step": 139460 }, { "grad_norm": 0.2897559404373169, "learning_rate": 2.302115436462286e-05, "loss": 0.0192, "step": 139470 }, { "grad_norm": 0.26976239681243896, "learning_rate": 2.3014194152039935e-05, "loss": 0.0204, "step": 139480 }, { "grad_norm": 0.5807519555091858, "learning_rate": 2.3007234677238825e-05, "loss": 0.0196, "step": 139490 }, { "grad_norm": 0.38781580328941345, "learning_rate": 2.300027594040976e-05, "loss": 0.0223, "step": 139500 }, { "grad_norm": 0.19259682297706604, "learning_rate": 2.2993317941743027e-05, "loss": 0.0158, "step": 139510 }, { "grad_norm": 0.8418973684310913, "learning_rate": 2.2986360681428832e-05, "loss": 0.0154, "step": 139520 }, { "grad_norm": 0.5277628898620605, "learning_rate": 2.297940415965737e-05, "loss": 0.0195, "step": 139530 }, { "grad_norm": 0.33560892939567566, "learning_rate": 2.297244837661886e-05, "loss": 0.0158, "step": 139540 }, { "grad_norm": 1.0240167379379272, "learning_rate": 2.2965493332503468e-05, "loss": 0.0218, "step": 139550 }, { "grad_norm": 0.7154567837715149, "learning_rate": 2.2958539027501324e-05, "loss": 0.0269, "step": 139560 }, { "grad_norm": 0.4199052155017853, "learning_rate": 2.295158546180255e-05, "loss": 0.0216, "step": 139570 }, { "grad_norm": 0.3900560736656189, "learning_rate": 2.2944632635597286e-05, "loss": 0.0192, "step": 139580 }, { "grad_norm": 0.40789952874183655, "learning_rate": 2.293768054907558e-05, "loss": 0.0269, "step": 139590 }, { "grad_norm": 0.41755279898643494, "learning_rate": 2.2930729202427542e-05, "loss": 0.0125, "step": 139600 }, { "grad_norm": 0.3272801637649536, "learning_rate": 2.2923778595843198e-05, "loss": 0.0177, "step": 139610 }, { "grad_norm": 0.25115013122558594, "learning_rate": 2.291682872951255e-05, "loss": 0.0149, "step": 139620 }, { "grad_norm": 0.6404113173484802, "learning_rate": 2.290987960362566e-05, "loss": 0.0139, "step": 139630 }, { "grad_norm": 0.7479946613311768, "learning_rate": 2.2902931218372475e-05, "loss": 0.0244, "step": 139640 }, { "grad_norm": 0.22056755423545837, "learning_rate": 2.2895983573942958e-05, "loss": 0.0168, "step": 139650 }, { "grad_norm": 0.819168210029602, "learning_rate": 2.2889036670527086e-05, "loss": 0.02, "step": 139660 }, { "grad_norm": 0.4004690647125244, "learning_rate": 2.2882090508314746e-05, "loss": 0.0297, "step": 139670 }, { "grad_norm": 1.0577999353408813, "learning_rate": 2.28751450874959e-05, "loss": 0.0188, "step": 139680 }, { "grad_norm": 0.32856056094169617, "learning_rate": 2.286820040826036e-05, "loss": 0.0145, "step": 139690 }, { "grad_norm": 0.24854373931884766, "learning_rate": 2.286125647079805e-05, "loss": 0.0181, "step": 139700 }, { "grad_norm": 0.27447575330734253, "learning_rate": 2.2854313275298767e-05, "loss": 0.0157, "step": 139710 }, { "grad_norm": 0.3028700351715088, "learning_rate": 2.2847370821952385e-05, "loss": 0.0205, "step": 139720 }, { "grad_norm": 0.36052632331848145, "learning_rate": 2.284042911094868e-05, "loss": 0.0195, "step": 139730 }, { "grad_norm": 0.3413001298904419, "learning_rate": 2.283348814247742e-05, "loss": 0.0176, "step": 139740 }, { "grad_norm": 0.5452853441238403, "learning_rate": 2.2826547916728415e-05, "loss": 0.0206, "step": 139750 }, { "grad_norm": 0.5099776387214661, "learning_rate": 2.2819608433891376e-05, "loss": 0.0226, "step": 139760 }, { "grad_norm": 0.5636218190193176, "learning_rate": 2.281266969415603e-05, "loss": 0.0158, "step": 139770 }, { "grad_norm": 0.484934538602829, "learning_rate": 2.2805731697712063e-05, "loss": 0.0165, "step": 139780 }, { "grad_norm": 0.501014232635498, "learning_rate": 2.2798794444749177e-05, "loss": 0.0217, "step": 139790 }, { "grad_norm": 0.8022445440292358, "learning_rate": 2.279185793545705e-05, "loss": 0.0256, "step": 139800 }, { "grad_norm": 0.4071425795555115, "learning_rate": 2.278492217002531e-05, "loss": 0.0144, "step": 139810 }, { "grad_norm": 0.5534414052963257, "learning_rate": 2.2777987148643564e-05, "loss": 0.0235, "step": 139820 }, { "grad_norm": 0.4746967852115631, "learning_rate": 2.277105287150141e-05, "loss": 0.0273, "step": 139830 }, { "grad_norm": 0.32688626646995544, "learning_rate": 2.276411933878846e-05, "loss": 0.0196, "step": 139840 }, { "grad_norm": 0.5668461322784424, "learning_rate": 2.2757186550694247e-05, "loss": 0.0235, "step": 139850 }, { "grad_norm": 0.5180675983428955, "learning_rate": 2.275025450740831e-05, "loss": 0.0265, "step": 139860 }, { "grad_norm": 0.42407137155532837, "learning_rate": 2.2743323209120193e-05, "loss": 0.0153, "step": 139870 }, { "grad_norm": 0.419310986995697, "learning_rate": 2.2736392656019363e-05, "loss": 0.0184, "step": 139880 }, { "grad_norm": 0.7100100517272949, "learning_rate": 2.2729462848295347e-05, "loss": 0.0224, "step": 139890 }, { "grad_norm": 0.31844496726989746, "learning_rate": 2.272253378613754e-05, "loss": 0.0163, "step": 139900 }, { "grad_norm": 0.2803754508495331, "learning_rate": 2.2715605469735406e-05, "loss": 0.0247, "step": 139910 }, { "grad_norm": 0.47844719886779785, "learning_rate": 2.2708677899278392e-05, "loss": 0.0158, "step": 139920 }, { "grad_norm": 0.46239328384399414, "learning_rate": 2.270175107495587e-05, "loss": 0.0239, "step": 139930 }, { "grad_norm": 0.40338876843452454, "learning_rate": 2.269482499695722e-05, "loss": 0.0168, "step": 139940 }, { "grad_norm": 0.4437134563922882, "learning_rate": 2.2687899665471775e-05, "loss": 0.0367, "step": 139950 }, { "grad_norm": 0.4013305902481079, "learning_rate": 2.268097508068891e-05, "loss": 0.0132, "step": 139960 }, { "grad_norm": 0.3091250956058502, "learning_rate": 2.267405124279792e-05, "loss": 0.023, "step": 139970 }, { "grad_norm": 0.24406908452510834, "learning_rate": 2.2667128151988094e-05, "loss": 0.0136, "step": 139980 }, { "grad_norm": 0.7336640954017639, "learning_rate": 2.266020580844873e-05, "loss": 0.0156, "step": 139990 }, { "grad_norm": 0.28446078300476074, "learning_rate": 2.2653284212369048e-05, "loss": 0.0165, "step": 140000 }, { "grad_norm": 0.23886729776859283, "learning_rate": 2.2646363363938324e-05, "loss": 0.0152, "step": 140010 }, { "grad_norm": 0.24781982600688934, "learning_rate": 2.2639443263345745e-05, "loss": 0.0207, "step": 140020 }, { "grad_norm": 0.5198742151260376, "learning_rate": 2.263252391078049e-05, "loss": 0.0212, "step": 140030 }, { "grad_norm": 0.5015899538993835, "learning_rate": 2.2625605306431773e-05, "loss": 0.0175, "step": 140040 }, { "grad_norm": 0.3974667489528656, "learning_rate": 2.261868745048872e-05, "loss": 0.0145, "step": 140050 }, { "grad_norm": 0.19391924142837524, "learning_rate": 2.2611770343140464e-05, "loss": 0.0164, "step": 140060 }, { "grad_norm": 0.31981006264686584, "learning_rate": 2.26048539845761e-05, "loss": 0.0148, "step": 140070 }, { "grad_norm": 0.35758712887763977, "learning_rate": 2.2597938374984755e-05, "loss": 0.0164, "step": 140080 }, { "grad_norm": 1.2500125169754028, "learning_rate": 2.2591023514555464e-05, "loss": 0.0194, "step": 140090 }, { "grad_norm": 0.9257835149765015, "learning_rate": 2.258410940347731e-05, "loss": 0.0351, "step": 140100 }, { "grad_norm": 0.47870415449142456, "learning_rate": 2.2577196041939302e-05, "loss": 0.0225, "step": 140110 }, { "grad_norm": 0.3066539764404297, "learning_rate": 2.257028343013044e-05, "loss": 0.0224, "step": 140120 }, { "grad_norm": 0.475314199924469, "learning_rate": 2.256337156823974e-05, "loss": 0.0263, "step": 140130 }, { "grad_norm": 0.4031587243080139, "learning_rate": 2.2556460456456152e-05, "loss": 0.0123, "step": 140140 }, { "grad_norm": 0.646021842956543, "learning_rate": 2.2549550094968624e-05, "loss": 0.0168, "step": 140150 }, { "grad_norm": 0.4510941505432129, "learning_rate": 2.2542640483966067e-05, "loss": 0.0192, "step": 140160 }, { "grad_norm": 0.8890885710716248, "learning_rate": 2.2535731623637403e-05, "loss": 0.0319, "step": 140170 }, { "grad_norm": 0.33697623014450073, "learning_rate": 2.252882351417156e-05, "loss": 0.0238, "step": 140180 }, { "grad_norm": 0.18547563254833221, "learning_rate": 2.2521916155757317e-05, "loss": 0.0181, "step": 140190 }, { "grad_norm": 0.4464944303035736, "learning_rate": 2.2515009548583587e-05, "loss": 0.0169, "step": 140200 }, { "grad_norm": 0.28298234939575195, "learning_rate": 2.250810369283915e-05, "loss": 0.018, "step": 140210 }, { "grad_norm": 0.31298378109931946, "learning_rate": 2.250119858871285e-05, "loss": 0.0195, "step": 140220 }, { "grad_norm": 0.42031949758529663, "learning_rate": 2.2494294236393447e-05, "loss": 0.0235, "step": 140230 }, { "grad_norm": 0.4039592146873474, "learning_rate": 2.2487390636069695e-05, "loss": 0.0256, "step": 140240 }, { "grad_norm": 0.46406570076942444, "learning_rate": 2.248048778793036e-05, "loss": 0.0294, "step": 140250 }, { "grad_norm": 0.9592124223709106, "learning_rate": 2.2473585692164144e-05, "loss": 0.0302, "step": 140260 }, { "grad_norm": 0.45132914185523987, "learning_rate": 2.2466684348959765e-05, "loss": 0.0228, "step": 140270 }, { "grad_norm": 0.39366820454597473, "learning_rate": 2.245978375850587e-05, "loss": 0.0388, "step": 140280 }, { "grad_norm": 0.41259902715682983, "learning_rate": 2.2452883920991134e-05, "loss": 0.0252, "step": 140290 }, { "grad_norm": 0.40166231989860535, "learning_rate": 2.2445984836604228e-05, "loss": 0.0227, "step": 140300 }, { "grad_norm": 0.45137476921081543, "learning_rate": 2.2439086505533747e-05, "loss": 0.0233, "step": 140310 }, { "grad_norm": 0.3125394880771637, "learning_rate": 2.2432188927968283e-05, "loss": 0.017, "step": 140320 }, { "grad_norm": 0.4604676067829132, "learning_rate": 2.2425292104096397e-05, "loss": 0.0228, "step": 140330 }, { "grad_norm": 0.24396178126335144, "learning_rate": 2.2418396034106687e-05, "loss": 0.0324, "step": 140340 }, { "grad_norm": 0.3349090814590454, "learning_rate": 2.2411500718187666e-05, "loss": 0.0173, "step": 140350 }, { "grad_norm": 0.2570822238922119, "learning_rate": 2.2404606156527835e-05, "loss": 0.0394, "step": 140360 }, { "grad_norm": 0.46852031350135803, "learning_rate": 2.239771234931572e-05, "loss": 0.0138, "step": 140370 }, { "grad_norm": 0.46170610189437866, "learning_rate": 2.2390819296739763e-05, "loss": 0.0189, "step": 140380 }, { "grad_norm": 0.748119592666626, "learning_rate": 2.2383926998988468e-05, "loss": 0.0201, "step": 140390 }, { "grad_norm": 0.2169032096862793, "learning_rate": 2.23770354562502e-05, "loss": 0.0174, "step": 140400 }, { "grad_norm": 0.9199990034103394, "learning_rate": 2.2370144668713404e-05, "loss": 0.0427, "step": 140410 }, { "grad_norm": 0.2749679684638977, "learning_rate": 2.2363254636566494e-05, "loss": 0.0159, "step": 140420 }, { "grad_norm": 0.27036136388778687, "learning_rate": 2.2356365359997818e-05, "loss": 0.0169, "step": 140430 }, { "grad_norm": 0.5263622403144836, "learning_rate": 2.2349476839195725e-05, "loss": 0.0177, "step": 140440 }, { "grad_norm": 0.4311840236186981, "learning_rate": 2.234258907434853e-05, "loss": 0.0224, "step": 140450 }, { "grad_norm": 0.4934714734554291, "learning_rate": 2.233570206564458e-05, "loss": 0.0203, "step": 140460 }, { "grad_norm": 0.40979087352752686, "learning_rate": 2.232881581327214e-05, "loss": 0.0204, "step": 140470 }, { "grad_norm": 0.7922412753105164, "learning_rate": 2.2321930317419464e-05, "loss": 0.024, "step": 140480 }, { "grad_norm": 0.3868381977081299, "learning_rate": 2.231504557827483e-05, "loss": 0.0185, "step": 140490 }, { "grad_norm": 0.42668208479881287, "learning_rate": 2.2308161596026434e-05, "loss": 0.0208, "step": 140500 }, { "grad_norm": 0.408716082572937, "learning_rate": 2.2301278370862516e-05, "loss": 0.0162, "step": 140510 }, { "grad_norm": 0.22872325778007507, "learning_rate": 2.2294395902971238e-05, "loss": 0.0252, "step": 140520 }, { "grad_norm": 0.5298582315444946, "learning_rate": 2.228751419254077e-05, "loss": 0.022, "step": 140530 }, { "grad_norm": 0.450878769159317, "learning_rate": 2.228063323975923e-05, "loss": 0.0204, "step": 140540 }, { "grad_norm": 0.3670327365398407, "learning_rate": 2.2273753044814787e-05, "loss": 0.0215, "step": 140550 }, { "grad_norm": 0.4312748312950134, "learning_rate": 2.2266873607895517e-05, "loss": 0.0141, "step": 140560 }, { "grad_norm": 0.20475904643535614, "learning_rate": 2.2259994929189486e-05, "loss": 0.0305, "step": 140570 }, { "grad_norm": 0.46670112013816833, "learning_rate": 2.2253117008884795e-05, "loss": 0.0194, "step": 140580 }, { "grad_norm": 0.3866865038871765, "learning_rate": 2.2246239847169435e-05, "loss": 0.0199, "step": 140590 }, { "grad_norm": 0.33285558223724365, "learning_rate": 2.2239363444231492e-05, "loss": 0.0176, "step": 140600 }, { "grad_norm": 0.32813093066215515, "learning_rate": 2.2232487800258882e-05, "loss": 0.011, "step": 140610 }, { "grad_norm": 0.5476661920547485, "learning_rate": 2.2225612915439632e-05, "loss": 0.0228, "step": 140620 }, { "grad_norm": 0.500019907951355, "learning_rate": 2.2218738789961702e-05, "loss": 0.0225, "step": 140630 }, { "grad_norm": 0.6813263893127441, "learning_rate": 2.221186542401302e-05, "loss": 0.0254, "step": 140640 }, { "grad_norm": 0.42964738607406616, "learning_rate": 2.2204992817781494e-05, "loss": 0.0174, "step": 140650 }, { "grad_norm": 0.8309909701347351, "learning_rate": 2.2198120971455004e-05, "loss": 0.0244, "step": 140660 }, { "grad_norm": 0.38839539885520935, "learning_rate": 2.219124988522145e-05, "loss": 0.0165, "step": 140670 }, { "grad_norm": 0.8283153772354126, "learning_rate": 2.218437955926871e-05, "loss": 0.0244, "step": 140680 }, { "grad_norm": 0.29631873965263367, "learning_rate": 2.217750999378454e-05, "loss": 0.0142, "step": 140690 }, { "grad_norm": 0.5391461253166199, "learning_rate": 2.217064118895682e-05, "loss": 0.0199, "step": 140700 }, { "grad_norm": 0.23802131414413452, "learning_rate": 2.2163773144973293e-05, "loss": 0.0142, "step": 140710 }, { "grad_norm": 0.4414826035499573, "learning_rate": 2.2156905862021766e-05, "loss": 0.0188, "step": 140720 }, { "grad_norm": 0.18705938756465912, "learning_rate": 2.215003934028998e-05, "loss": 0.0218, "step": 140730 }, { "grad_norm": 0.26937806606292725, "learning_rate": 2.2143173579965632e-05, "loss": 0.0155, "step": 140740 }, { "grad_norm": 0.6545014381408691, "learning_rate": 2.213630858123647e-05, "loss": 0.0202, "step": 140750 }, { "grad_norm": 0.36412301659584045, "learning_rate": 2.2129444344290167e-05, "loss": 0.0189, "step": 140760 }, { "grad_norm": 0.5234718322753906, "learning_rate": 2.2122580869314385e-05, "loss": 0.0224, "step": 140770 }, { "grad_norm": 0.6483821868896484, "learning_rate": 2.2115718156496752e-05, "loss": 0.0219, "step": 140780 }, { "grad_norm": 0.32526707649230957, "learning_rate": 2.2108856206024926e-05, "loss": 0.0166, "step": 140790 }, { "grad_norm": 0.9427875876426697, "learning_rate": 2.2101995018086473e-05, "loss": 0.0226, "step": 140800 }, { "grad_norm": 0.6596044898033142, "learning_rate": 2.2095134592869017e-05, "loss": 0.0202, "step": 140810 }, { "grad_norm": 0.538537859916687, "learning_rate": 2.2088274930560095e-05, "loss": 0.0172, "step": 140820 }, { "grad_norm": 0.387747198343277, "learning_rate": 2.208141603134723e-05, "loss": 0.0226, "step": 140830 }, { "grad_norm": 0.3140234351158142, "learning_rate": 2.2074557895417986e-05, "loss": 0.0187, "step": 140840 }, { "grad_norm": 0.40702563524246216, "learning_rate": 2.2067700522959834e-05, "loss": 0.0196, "step": 140850 }, { "grad_norm": 0.4663445055484772, "learning_rate": 2.206084391416024e-05, "loss": 0.0172, "step": 140860 }, { "grad_norm": 0.5394144058227539, "learning_rate": 2.2053988069206695e-05, "loss": 0.0417, "step": 140870 }, { "grad_norm": 0.337161660194397, "learning_rate": 2.2047132988286596e-05, "loss": 0.0175, "step": 140880 }, { "grad_norm": 0.33050575852394104, "learning_rate": 2.2040278671587422e-05, "loss": 0.0169, "step": 140890 }, { "grad_norm": 0.2601403295993805, "learning_rate": 2.203342511929648e-05, "loss": 0.0194, "step": 140900 }, { "grad_norm": 0.34321823716163635, "learning_rate": 2.2026572331601215e-05, "loss": 0.0215, "step": 140910 }, { "grad_norm": 0.4767376184463501, "learning_rate": 2.2019720308688933e-05, "loss": 0.0237, "step": 140920 }, { "grad_norm": 0.2608132064342499, "learning_rate": 2.2012869050747003e-05, "loss": 0.0238, "step": 140930 }, { "grad_norm": 0.36850136518478394, "learning_rate": 2.2006018557962722e-05, "loss": 0.0151, "step": 140940 }, { "grad_norm": 0.5772277116775513, "learning_rate": 2.1999168830523354e-05, "loss": 0.014, "step": 140950 }, { "grad_norm": 0.36870428919792175, "learning_rate": 2.1992319868616212e-05, "loss": 0.0187, "step": 140960 }, { "grad_norm": 0.36294594407081604, "learning_rate": 2.1985471672428522e-05, "loss": 0.0186, "step": 140970 }, { "grad_norm": 0.3098183572292328, "learning_rate": 2.197862424214751e-05, "loss": 0.0159, "step": 140980 }, { "grad_norm": 1.2147904634475708, "learning_rate": 2.1971777577960367e-05, "loss": 0.034, "step": 140990 }, { "grad_norm": 0.5327812433242798, "learning_rate": 2.1964931680054295e-05, "loss": 0.021, "step": 141000 }, { "grad_norm": 0.6520422101020813, "learning_rate": 2.195808654861648e-05, "loss": 0.0175, "step": 141010 }, { "grad_norm": 0.489189088344574, "learning_rate": 2.1951242183834043e-05, "loss": 0.0162, "step": 141020 }, { "grad_norm": 0.26971614360809326, "learning_rate": 2.1944398585894103e-05, "loss": 0.0139, "step": 141030 }, { "grad_norm": 0.3454742729663849, "learning_rate": 2.193755575498375e-05, "loss": 0.0276, "step": 141040 }, { "grad_norm": 0.24123597145080566, "learning_rate": 2.19307136912901e-05, "loss": 0.0203, "step": 141050 }, { "grad_norm": 0.477939635515213, "learning_rate": 2.1923872395000193e-05, "loss": 0.0232, "step": 141060 }, { "grad_norm": 0.6803086400032043, "learning_rate": 2.1917031866301048e-05, "loss": 0.0181, "step": 141070 }, { "grad_norm": 0.5641144514083862, "learning_rate": 2.191019210537972e-05, "loss": 0.0267, "step": 141080 }, { "grad_norm": 0.35092952847480774, "learning_rate": 2.1903353112423165e-05, "loss": 0.0211, "step": 141090 }, { "grad_norm": 0.48122286796569824, "learning_rate": 2.189651488761842e-05, "loss": 0.0268, "step": 141100 }, { "grad_norm": 0.9514409303665161, "learning_rate": 2.1889677431152362e-05, "loss": 0.0265, "step": 141110 }, { "grad_norm": 0.3012084364891052, "learning_rate": 2.1882840743211962e-05, "loss": 0.0212, "step": 141120 }, { "grad_norm": 0.2012384831905365, "learning_rate": 2.1876004823984154e-05, "loss": 0.0176, "step": 141130 }, { "grad_norm": 0.5041446089744568, "learning_rate": 2.186916967365581e-05, "loss": 0.0256, "step": 141140 }, { "grad_norm": 0.27753955125808716, "learning_rate": 2.1862335292413793e-05, "loss": 0.0175, "step": 141150 }, { "grad_norm": 0.3627091348171234, "learning_rate": 2.1855501680444945e-05, "loss": 0.0188, "step": 141160 }, { "grad_norm": 0.6013097167015076, "learning_rate": 2.1848668837936127e-05, "loss": 0.0236, "step": 141170 }, { "grad_norm": 0.3901308476924896, "learning_rate": 2.184183676507412e-05, "loss": 0.0164, "step": 141180 }, { "grad_norm": 0.5220308899879456, "learning_rate": 2.1835005462045706e-05, "loss": 0.0253, "step": 141190 }, { "grad_norm": 0.5435779094696045, "learning_rate": 2.1828174929037677e-05, "loss": 0.0151, "step": 141200 }, { "grad_norm": 0.4347541332244873, "learning_rate": 2.182134516623674e-05, "loss": 0.0219, "step": 141210 }, { "grad_norm": 0.4065355956554413, "learning_rate": 2.1814516173829663e-05, "loss": 0.0192, "step": 141220 }, { "grad_norm": 0.6096925735473633, "learning_rate": 2.1807687952003124e-05, "loss": 0.0296, "step": 141230 }, { "grad_norm": 0.21152733266353607, "learning_rate": 2.1800860500943786e-05, "loss": 0.0176, "step": 141240 }, { "grad_norm": 7.22463321685791, "learning_rate": 2.1794033820838344e-05, "loss": 0.0298, "step": 141250 }, { "grad_norm": 0.41767990589141846, "learning_rate": 2.1787207911873424e-05, "loss": 0.0202, "step": 141260 }, { "grad_norm": 0.24808673560619354, "learning_rate": 2.1780382774235642e-05, "loss": 0.0301, "step": 141270 }, { "grad_norm": 0.2510799467563629, "learning_rate": 2.1773558408111573e-05, "loss": 0.0174, "step": 141280 }, { "grad_norm": 0.44017237424850464, "learning_rate": 2.1766734813687827e-05, "loss": 0.0184, "step": 141290 }, { "grad_norm": 0.474965900182724, "learning_rate": 2.1759911991150933e-05, "loss": 0.0301, "step": 141300 }, { "grad_norm": 0.31682834029197693, "learning_rate": 2.1753089940687456e-05, "loss": 0.0263, "step": 141310 }, { "grad_norm": 0.3992713689804077, "learning_rate": 2.1746268662483883e-05, "loss": 0.027, "step": 141320 }, { "grad_norm": 0.7890425324440002, "learning_rate": 2.1739448156726688e-05, "loss": 0.0262, "step": 141330 }, { "grad_norm": 0.20145787298679352, "learning_rate": 2.1732628423602386e-05, "loss": 0.0202, "step": 141340 }, { "grad_norm": 0.22939839959144592, "learning_rate": 2.172580946329741e-05, "loss": 0.0126, "step": 141350 }, { "grad_norm": 0.2660191059112549, "learning_rate": 2.1718991275998173e-05, "loss": 0.0167, "step": 141360 }, { "grad_norm": 0.4183310866355896, "learning_rate": 2.1712173861891072e-05, "loss": 0.0155, "step": 141370 }, { "grad_norm": 0.24698041379451752, "learning_rate": 2.1705357221162515e-05, "loss": 0.0228, "step": 141380 }, { "grad_norm": 0.32317236065864563, "learning_rate": 2.1698541353998892e-05, "loss": 0.0166, "step": 141390 }, { "grad_norm": 0.35108450055122375, "learning_rate": 2.169172626058648e-05, "loss": 0.0169, "step": 141400 }, { "grad_norm": 0.6895926594734192, "learning_rate": 2.1684911941111663e-05, "loss": 0.0167, "step": 141410 }, { "grad_norm": 0.39996278285980225, "learning_rate": 2.1678098395760694e-05, "loss": 0.0316, "step": 141420 }, { "grad_norm": 0.36982789635658264, "learning_rate": 2.167128562471989e-05, "loss": 0.0204, "step": 141430 }, { "grad_norm": 0.28994515538215637, "learning_rate": 2.1664473628175495e-05, "loss": 0.0195, "step": 141440 }, { "grad_norm": 0.9545009136199951, "learning_rate": 2.165766240631373e-05, "loss": 0.0213, "step": 141450 }, { "grad_norm": 0.34068912267684937, "learning_rate": 2.1650851959320845e-05, "loss": 0.0186, "step": 141460 }, { "grad_norm": 0.330106645822525, "learning_rate": 2.1644042287383015e-05, "loss": 0.0226, "step": 141470 }, { "grad_norm": 0.4116394817829132, "learning_rate": 2.163723339068642e-05, "loss": 0.0218, "step": 141480 }, { "grad_norm": 0.5254673957824707, "learning_rate": 2.1630425269417186e-05, "loss": 0.0197, "step": 141490 }, { "grad_norm": 0.5860031843185425, "learning_rate": 2.1623617923761468e-05, "loss": 0.018, "step": 141500 }, { "grad_norm": 0.25857067108154297, "learning_rate": 2.1616811353905398e-05, "loss": 0.0133, "step": 141510 }, { "grad_norm": 0.46382004022598267, "learning_rate": 2.1610005560035035e-05, "loss": 0.0195, "step": 141520 }, { "grad_norm": 0.20800556242465973, "learning_rate": 2.160320054233646e-05, "loss": 0.0241, "step": 141530 }, { "grad_norm": 0.9456473588943481, "learning_rate": 2.1596396300995697e-05, "loss": 0.0315, "step": 141540 }, { "grad_norm": 0.39995822310447693, "learning_rate": 2.1589592836198802e-05, "loss": 0.0264, "step": 141550 }, { "grad_norm": 0.6065558791160583, "learning_rate": 2.1582790148131767e-05, "loss": 0.019, "step": 141560 }, { "grad_norm": 0.3406146168708801, "learning_rate": 2.1575988236980556e-05, "loss": 0.0471, "step": 141570 }, { "grad_norm": 0.25680434703826904, "learning_rate": 2.1569187102931164e-05, "loss": 0.023, "step": 141580 }, { "grad_norm": 0.2483721673488617, "learning_rate": 2.15623867461695e-05, "loss": 0.0221, "step": 141590 }, { "grad_norm": 0.5378167629241943, "learning_rate": 2.1555587166881535e-05, "loss": 0.0194, "step": 141600 }, { "grad_norm": 0.2632755935192108, "learning_rate": 2.15487883652531e-05, "loss": 0.0183, "step": 141610 }, { "grad_norm": 0.4744248688220978, "learning_rate": 2.154199034147009e-05, "loss": 0.0198, "step": 141620 }, { "grad_norm": 0.3908984065055847, "learning_rate": 2.1535193095718403e-05, "loss": 0.0289, "step": 141630 }, { "grad_norm": 0.48788926005363464, "learning_rate": 2.152839662818384e-05, "loss": 0.0246, "step": 141640 }, { "grad_norm": 0.3991146981716156, "learning_rate": 2.1521600939052212e-05, "loss": 0.0151, "step": 141650 }, { "grad_norm": 0.4791654050350189, "learning_rate": 2.15148060285093e-05, "loss": 0.0197, "step": 141660 }, { "grad_norm": 0.3302538990974426, "learning_rate": 2.1508011896740908e-05, "loss": 0.0256, "step": 141670 }, { "grad_norm": 0.25567302107810974, "learning_rate": 2.1501218543932767e-05, "loss": 0.02, "step": 141680 }, { "grad_norm": 0.6837798357009888, "learning_rate": 2.1494425970270587e-05, "loss": 0.0188, "step": 141690 }, { "grad_norm": 0.33846354484558105, "learning_rate": 2.1487634175940107e-05, "loss": 0.0143, "step": 141700 }, { "grad_norm": 0.31233111023902893, "learning_rate": 2.1480843161126985e-05, "loss": 0.0187, "step": 141710 }, { "grad_norm": 0.7165054678916931, "learning_rate": 2.1474052926016906e-05, "loss": 0.0293, "step": 141720 }, { "grad_norm": 0.6157118082046509, "learning_rate": 2.146726347079551e-05, "loss": 0.0256, "step": 141730 }, { "grad_norm": 0.2827639877796173, "learning_rate": 2.146047479564841e-05, "loss": 0.0209, "step": 141740 }, { "grad_norm": 0.23335587978363037, "learning_rate": 2.1453686900761188e-05, "loss": 0.0141, "step": 141750 }, { "grad_norm": 0.5627097487449646, "learning_rate": 2.1446899786319462e-05, "loss": 0.0342, "step": 141760 }, { "grad_norm": 0.3062632083892822, "learning_rate": 2.144011345250877e-05, "loss": 0.0162, "step": 141770 }, { "grad_norm": 0.3243979215621948, "learning_rate": 2.143332789951463e-05, "loss": 0.0183, "step": 141780 }, { "grad_norm": 2.5852065086364746, "learning_rate": 2.1426543127522585e-05, "loss": 0.0248, "step": 141790 }, { "grad_norm": 0.6378735303878784, "learning_rate": 2.14197591367181e-05, "loss": 0.0168, "step": 141800 }, { "grad_norm": 0.30106890201568604, "learning_rate": 2.1412975927286698e-05, "loss": 0.0231, "step": 141810 }, { "grad_norm": 0.38174325227737427, "learning_rate": 2.1406193499413762e-05, "loss": 0.0252, "step": 141820 }, { "grad_norm": 0.3767818510532379, "learning_rate": 2.1399411853284756e-05, "loss": 0.0336, "step": 141830 }, { "grad_norm": 0.39669081568717957, "learning_rate": 2.13926309890851e-05, "loss": 0.0233, "step": 141840 }, { "grad_norm": 1.118973970413208, "learning_rate": 2.1385850907000164e-05, "loss": 0.0289, "step": 141850 }, { "grad_norm": 0.46052926778793335, "learning_rate": 2.137907160721532e-05, "loss": 0.0234, "step": 141860 }, { "grad_norm": 0.34523966908454895, "learning_rate": 2.1372293089915877e-05, "loss": 0.0202, "step": 141870 }, { "grad_norm": 0.9086145162582397, "learning_rate": 2.1365515355287192e-05, "loss": 0.022, "step": 141880 }, { "grad_norm": 0.4112466275691986, "learning_rate": 2.13587384035146e-05, "loss": 0.0214, "step": 141890 }, { "grad_norm": 0.3313516676425934, "learning_rate": 2.1351962234783297e-05, "loss": 0.0143, "step": 141900 }, { "grad_norm": 0.16895060241222382, "learning_rate": 2.1345186849278603e-05, "loss": 0.034, "step": 141910 }, { "grad_norm": 0.4492649435997009, "learning_rate": 2.1338412247185724e-05, "loss": 0.019, "step": 141920 }, { "grad_norm": 0.4099378287792206, "learning_rate": 2.1331638428689898e-05, "loss": 0.0166, "step": 141930 }, { "grad_norm": 0.4416152536869049, "learning_rate": 2.1324865393976305e-05, "loss": 0.0272, "step": 141940 }, { "grad_norm": 0.49165406823158264, "learning_rate": 2.131809314323011e-05, "loss": 0.0196, "step": 141950 }, { "grad_norm": 0.5173073410987854, "learning_rate": 2.1311321676636486e-05, "loss": 0.0196, "step": 141960 }, { "grad_norm": 0.17244333028793335, "learning_rate": 2.130455099438055e-05, "loss": 0.0182, "step": 141970 }, { "grad_norm": 0.2300470471382141, "learning_rate": 2.129778109664741e-05, "loss": 0.0188, "step": 141980 }, { "grad_norm": 0.5076035261154175, "learning_rate": 2.1291011983622133e-05, "loss": 0.0137, "step": 141990 }, { "grad_norm": 0.38394415378570557, "learning_rate": 2.1284243655489822e-05, "loss": 0.0171, "step": 142000 }, { "grad_norm": 2.0786936283111572, "learning_rate": 2.1277476112435484e-05, "loss": 0.0177, "step": 142010 }, { "grad_norm": 0.25236162543296814, "learning_rate": 2.1270709354644175e-05, "loss": 0.014, "step": 142020 }, { "grad_norm": 0.34272119402885437, "learning_rate": 2.1263943382300876e-05, "loss": 0.0182, "step": 142030 }, { "grad_norm": 0.26653939485549927, "learning_rate": 2.1257178195590553e-05, "loss": 0.0156, "step": 142040 }, { "grad_norm": 0.52967369556427, "learning_rate": 2.1250413794698205e-05, "loss": 0.0253, "step": 142050 }, { "grad_norm": 1.5855698585510254, "learning_rate": 2.1243650179808738e-05, "loss": 0.026, "step": 142060 }, { "grad_norm": 0.3370126485824585, "learning_rate": 2.1236887351107053e-05, "loss": 0.0231, "step": 142070 }, { "grad_norm": 0.44752129912376404, "learning_rate": 2.1230125308778082e-05, "loss": 0.0315, "step": 142080 }, { "grad_norm": 0.40953612327575684, "learning_rate": 2.1223364053006657e-05, "loss": 0.0205, "step": 142090 }, { "grad_norm": 0.6490787863731384, "learning_rate": 2.1216603583977685e-05, "loss": 0.0411, "step": 142100 }, { "grad_norm": 0.740985631942749, "learning_rate": 2.120984390187592e-05, "loss": 0.0292, "step": 142110 }, { "grad_norm": 0.25602665543556213, "learning_rate": 2.1203085006886225e-05, "loss": 0.0114, "step": 142120 }, { "grad_norm": 0.2957034707069397, "learning_rate": 2.119632689919335e-05, "loss": 0.0209, "step": 142130 }, { "grad_norm": 0.34950947761535645, "learning_rate": 2.1189569578982093e-05, "loss": 0.0282, "step": 142140 }, { "grad_norm": 0.5251956582069397, "learning_rate": 2.1182813046437184e-05, "loss": 0.022, "step": 142150 }, { "grad_norm": 0.4068724513053894, "learning_rate": 2.117605730174332e-05, "loss": 0.0164, "step": 142160 }, { "grad_norm": 0.43983137607574463, "learning_rate": 2.116930234508524e-05, "loss": 0.014, "step": 142170 }, { "grad_norm": 0.17606571316719055, "learning_rate": 2.11625481766476e-05, "loss": 0.0143, "step": 142180 }, { "grad_norm": 0.6497757434844971, "learning_rate": 2.1155794796615058e-05, "loss": 0.0258, "step": 142190 }, { "grad_norm": 0.3497461676597595, "learning_rate": 2.1149042205172236e-05, "loss": 0.017, "step": 142200 }, { "grad_norm": 0.41725456714630127, "learning_rate": 2.1142290402503763e-05, "loss": 0.0296, "step": 142210 }, { "grad_norm": 0.2214018851518631, "learning_rate": 2.113553938879424e-05, "loss": 0.0299, "step": 142220 }, { "grad_norm": 0.35486724972724915, "learning_rate": 2.112878916422823e-05, "loss": 0.0198, "step": 142230 }, { "grad_norm": 0.42294442653656006, "learning_rate": 2.112203972899028e-05, "loss": 0.0152, "step": 142240 }, { "grad_norm": 0.32334306836128235, "learning_rate": 2.1115291083264893e-05, "loss": 0.0312, "step": 142250 }, { "grad_norm": 0.2796299159526825, "learning_rate": 2.1108543227236616e-05, "loss": 0.0259, "step": 142260 }, { "grad_norm": 1.8588711023330688, "learning_rate": 2.1101796161089916e-05, "loss": 0.0203, "step": 142270 }, { "grad_norm": 0.5239871144294739, "learning_rate": 2.1095049885009234e-05, "loss": 0.0322, "step": 142280 }, { "grad_norm": 0.23956671357154846, "learning_rate": 2.1088304399179042e-05, "loss": 0.0206, "step": 142290 }, { "grad_norm": 0.274252712726593, "learning_rate": 2.108155970378373e-05, "loss": 0.0142, "step": 142300 }, { "grad_norm": 0.3006769120693207, "learning_rate": 2.1074815799007753e-05, "loss": 0.0177, "step": 142310 }, { "grad_norm": 0.4035564959049225, "learning_rate": 2.10680726850354e-05, "loss": 0.0193, "step": 142320 }, { "grad_norm": 0.4456486999988556, "learning_rate": 2.106133036205108e-05, "loss": 0.021, "step": 142330 }, { "grad_norm": 0.5396602153778076, "learning_rate": 2.1054588830239124e-05, "loss": 0.0167, "step": 142340 }, { "grad_norm": 0.3592164218425751, "learning_rate": 2.1047848089783846e-05, "loss": 0.0178, "step": 142350 }, { "grad_norm": 0.25364232063293457, "learning_rate": 2.1041108140869514e-05, "loss": 0.0332, "step": 142360 }, { "grad_norm": 0.3157702386379242, "learning_rate": 2.1034368983680397e-05, "loss": 0.0252, "step": 142370 }, { "grad_norm": 1.4246201515197754, "learning_rate": 2.102763061840076e-05, "loss": 0.0279, "step": 142380 }, { "grad_norm": 6.16881799697876, "learning_rate": 2.1020893045214823e-05, "loss": 0.0314, "step": 142390 }, { "grad_norm": 0.3905235826969147, "learning_rate": 2.1014156264306767e-05, "loss": 0.0244, "step": 142400 }, { "grad_norm": 0.32293039560317993, "learning_rate": 2.1007420275860802e-05, "loss": 0.0271, "step": 142410 }, { "grad_norm": 0.4510710537433624, "learning_rate": 2.1000685080061062e-05, "loss": 0.0278, "step": 142420 }, { "grad_norm": 0.4093158543109894, "learning_rate": 2.0993950677091713e-05, "loss": 0.0199, "step": 142430 }, { "grad_norm": 0.36214032769203186, "learning_rate": 2.098721706713685e-05, "loss": 0.0176, "step": 142440 }, { "grad_norm": 0.5469124913215637, "learning_rate": 2.098048425038056e-05, "loss": 0.0326, "step": 142450 }, { "grad_norm": 0.2649679481983185, "learning_rate": 2.097375222700695e-05, "loss": 0.0117, "step": 142460 }, { "grad_norm": 0.580830991268158, "learning_rate": 2.0967020997200043e-05, "loss": 0.0201, "step": 142470 }, { "grad_norm": 0.323736310005188, "learning_rate": 2.0960290561143873e-05, "loss": 0.0236, "step": 142480 }, { "grad_norm": 0.45236101746559143, "learning_rate": 2.0953560919022436e-05, "loss": 0.024, "step": 142490 }, { "grad_norm": 0.33484765887260437, "learning_rate": 2.094683207101975e-05, "loss": 0.0213, "step": 142500 }, { "grad_norm": 0.3272281587123871, "learning_rate": 2.094010401731974e-05, "loss": 0.0171, "step": 142510 }, { "grad_norm": 0.24340137839317322, "learning_rate": 2.0933376758106386e-05, "loss": 0.0115, "step": 142520 }, { "grad_norm": 0.4089815020561218, "learning_rate": 2.09266502935636e-05, "loss": 0.0139, "step": 142530 }, { "grad_norm": 0.8283567428588867, "learning_rate": 2.0919924623875248e-05, "loss": 0.0208, "step": 142540 }, { "grad_norm": 0.29008376598358154, "learning_rate": 2.091319974922525e-05, "loss": 0.0194, "step": 142550 }, { "grad_norm": 0.4843222200870514, "learning_rate": 2.0906475669797444e-05, "loss": 0.0315, "step": 142560 }, { "grad_norm": 0.3754345178604126, "learning_rate": 2.0899752385775662e-05, "loss": 0.0166, "step": 142570 }, { "grad_norm": 0.5787218809127808, "learning_rate": 2.0893029897343696e-05, "loss": 0.016, "step": 142580 }, { "grad_norm": 0.3976743221282959, "learning_rate": 2.088630820468536e-05, "loss": 0.0281, "step": 142590 }, { "grad_norm": 0.28298547863960266, "learning_rate": 2.087958730798446e-05, "loss": 0.0171, "step": 142600 }, { "grad_norm": 0.929165780544281, "learning_rate": 2.0872867207424668e-05, "loss": 0.0221, "step": 142610 }, { "grad_norm": 0.645475447177887, "learning_rate": 2.086614790318976e-05, "loss": 0.0191, "step": 142620 }, { "grad_norm": 0.4136289358139038, "learning_rate": 2.0859429395463405e-05, "loss": 0.0203, "step": 142630 }, { "grad_norm": 0.34431201219558716, "learning_rate": 2.0852711684429314e-05, "loss": 0.023, "step": 142640 }, { "grad_norm": 0.2857010066509247, "learning_rate": 2.0845994770271142e-05, "loss": 0.0261, "step": 142650 }, { "grad_norm": 0.43341800570487976, "learning_rate": 2.0839278653172505e-05, "loss": 0.0331, "step": 142660 }, { "grad_norm": 0.23364786803722382, "learning_rate": 2.0832563333317046e-05, "loss": 0.0189, "step": 142670 }, { "grad_norm": 0.6457808017730713, "learning_rate": 2.0825848810888355e-05, "loss": 0.0276, "step": 142680 }, { "grad_norm": 0.44170311093330383, "learning_rate": 2.081913508606999e-05, "loss": 0.0232, "step": 142690 }, { "grad_norm": 0.42726433277130127, "learning_rate": 2.0812422159045497e-05, "loss": 0.0204, "step": 142700 }, { "grad_norm": 0.5164638757705688, "learning_rate": 2.080571002999841e-05, "loss": 0.0264, "step": 142710 }, { "grad_norm": 0.5654779076576233, "learning_rate": 2.0798998699112266e-05, "loss": 0.0233, "step": 142720 }, { "grad_norm": 1.2024967670440674, "learning_rate": 2.0792288166570528e-05, "loss": 0.034, "step": 142730 }, { "grad_norm": 0.36113324761390686, "learning_rate": 2.0785578432556656e-05, "loss": 0.0146, "step": 142740 }, { "grad_norm": 0.4951360523700714, "learning_rate": 2.0778869497254077e-05, "loss": 0.0289, "step": 142750 }, { "grad_norm": 0.3381354808807373, "learning_rate": 2.0772161360846248e-05, "loss": 0.0511, "step": 142760 }, { "grad_norm": 0.5530857443809509, "learning_rate": 2.0765454023516546e-05, "loss": 0.0236, "step": 142770 }, { "grad_norm": 0.42005544900894165, "learning_rate": 2.0758747485448328e-05, "loss": 0.0159, "step": 142780 }, { "grad_norm": 0.839381754398346, "learning_rate": 2.0752041746824985e-05, "loss": 0.0294, "step": 142790 }, { "grad_norm": 0.6482917666435242, "learning_rate": 2.0745336807829817e-05, "loss": 0.018, "step": 142800 }, { "grad_norm": 0.3153923749923706, "learning_rate": 2.073863266864619e-05, "loss": 0.0197, "step": 142810 }, { "grad_norm": 0.49829235672950745, "learning_rate": 2.073192932945731e-05, "loss": 0.0161, "step": 142820 }, { "grad_norm": 0.2461981177330017, "learning_rate": 2.0725226790446495e-05, "loss": 0.0183, "step": 142830 }, { "grad_norm": 0.16361242532730103, "learning_rate": 2.0718525051796993e-05, "loss": 0.0183, "step": 142840 }, { "grad_norm": 0.44121289253234863, "learning_rate": 2.0711824113692018e-05, "loss": 0.0221, "step": 142850 }, { "grad_norm": 0.15583810210227966, "learning_rate": 2.0705123976314765e-05, "loss": 0.0154, "step": 142860 }, { "grad_norm": 0.38748493790626526, "learning_rate": 2.06984246398484e-05, "loss": 0.0174, "step": 142870 }, { "grad_norm": 0.953416645526886, "learning_rate": 2.0691726104476122e-05, "loss": 0.0213, "step": 142880 }, { "grad_norm": 0.9969333410263062, "learning_rate": 2.068502837038104e-05, "loss": 0.0219, "step": 142890 }, { "grad_norm": 0.5377387404441833, "learning_rate": 2.0678331437746244e-05, "loss": 0.0109, "step": 142900 }, { "grad_norm": 0.2715563178062439, "learning_rate": 2.0671635306754872e-05, "loss": 0.0264, "step": 142910 }, { "grad_norm": 1.4491602182388306, "learning_rate": 2.0664939977589953e-05, "loss": 0.0257, "step": 142920 }, { "grad_norm": 0.3176664412021637, "learning_rate": 2.0658245450434576e-05, "loss": 0.0153, "step": 142930 }, { "grad_norm": 0.36308473348617554, "learning_rate": 2.0651551725471746e-05, "loss": 0.0134, "step": 142940 }, { "grad_norm": 0.6152035593986511, "learning_rate": 2.0644858802884463e-05, "loss": 0.0182, "step": 142950 }, { "grad_norm": 0.2539807856082916, "learning_rate": 2.0638166682855697e-05, "loss": 0.0156, "step": 142960 }, { "grad_norm": 0.6101025342941284, "learning_rate": 2.0631475365568438e-05, "loss": 0.0204, "step": 142970 }, { "grad_norm": 0.6565963625907898, "learning_rate": 2.062478485120561e-05, "loss": 0.0339, "step": 142980 }, { "grad_norm": 0.5240156650543213, "learning_rate": 2.0618095139950112e-05, "loss": 0.0232, "step": 142990 }, { "grad_norm": 0.540149986743927, "learning_rate": 2.061140623198487e-05, "loss": 0.0195, "step": 143000 }, { "grad_norm": 0.8090294003486633, "learning_rate": 2.0604718127492728e-05, "loss": 0.0268, "step": 143010 }, { "grad_norm": 0.42610689997673035, "learning_rate": 2.0598030826656577e-05, "loss": 0.0182, "step": 143020 }, { "grad_norm": 0.4247860610485077, "learning_rate": 2.059134432965919e-05, "loss": 0.0234, "step": 143030 }, { "grad_norm": 0.6364171504974365, "learning_rate": 2.0584658636683395e-05, "loss": 0.0283, "step": 143040 }, { "grad_norm": 0.44493094086647034, "learning_rate": 2.0577973747911998e-05, "loss": 0.0211, "step": 143050 }, { "grad_norm": 0.4355512857437134, "learning_rate": 2.0571289663527753e-05, "loss": 0.0235, "step": 143060 }, { "grad_norm": 0.3123239576816559, "learning_rate": 2.0564606383713392e-05, "loss": 0.0174, "step": 143070 }, { "grad_norm": 0.3221505284309387, "learning_rate": 2.055792390865161e-05, "loss": 0.0193, "step": 143080 }, { "grad_norm": 1.6912126541137695, "learning_rate": 2.055124223852513e-05, "loss": 0.0253, "step": 143090 }, { "grad_norm": 0.1409294158220291, "learning_rate": 2.0544561373516664e-05, "loss": 0.0144, "step": 143100 }, { "grad_norm": 0.26155316829681396, "learning_rate": 2.0537881313808783e-05, "loss": 0.0163, "step": 143110 }, { "grad_norm": 0.27560874819755554, "learning_rate": 2.0531202059584177e-05, "loss": 0.0245, "step": 143120 }, { "grad_norm": 0.19099533557891846, "learning_rate": 2.052452361102542e-05, "loss": 0.0219, "step": 143130 }, { "grad_norm": 0.962867259979248, "learning_rate": 2.051784596831513e-05, "loss": 0.0229, "step": 143140 }, { "grad_norm": 1.117475986480713, "learning_rate": 2.0511169131635855e-05, "loss": 0.0445, "step": 143150 }, { "grad_norm": 0.49348849058151245, "learning_rate": 2.0504493101170117e-05, "loss": 0.0215, "step": 143160 }, { "grad_norm": 0.27001965045928955, "learning_rate": 2.0497817877100484e-05, "loss": 0.0119, "step": 143170 }, { "grad_norm": 0.43815410137176514, "learning_rate": 2.049114345960942e-05, "loss": 0.0204, "step": 143180 }, { "grad_norm": 0.5425558090209961, "learning_rate": 2.0484469848879407e-05, "loss": 0.0155, "step": 143190 }, { "grad_norm": 0.4066406786441803, "learning_rate": 2.0477797045092884e-05, "loss": 0.0193, "step": 143200 }, { "grad_norm": 0.4943035840988159, "learning_rate": 2.0471125048432317e-05, "loss": 0.0215, "step": 143210 }, { "grad_norm": 0.32959720492362976, "learning_rate": 2.046445385908008e-05, "loss": 0.0155, "step": 143220 }, { "grad_norm": 0.5382245779037476, "learning_rate": 2.0457783477218596e-05, "loss": 0.0218, "step": 143230 }, { "grad_norm": 0.8114638924598694, "learning_rate": 2.045111390303021e-05, "loss": 0.0212, "step": 143240 }, { "grad_norm": 0.3000394403934479, "learning_rate": 2.0444445136697254e-05, "loss": 0.015, "step": 143250 }, { "grad_norm": 0.5090539455413818, "learning_rate": 2.0437777178402086e-05, "loss": 0.0228, "step": 143260 }, { "grad_norm": 0.4075483977794647, "learning_rate": 2.0431110028326976e-05, "loss": 0.0208, "step": 143270 }, { "grad_norm": 0.21135035157203674, "learning_rate": 2.04244436866542e-05, "loss": 0.0279, "step": 143280 }, { "grad_norm": 0.3021719455718994, "learning_rate": 2.041777815356603e-05, "loss": 0.02, "step": 143290 }, { "grad_norm": 0.34162092208862305, "learning_rate": 2.0411113429244684e-05, "loss": 0.0176, "step": 143300 }, { "grad_norm": 0.3086877763271332, "learning_rate": 2.040444951387241e-05, "loss": 0.0165, "step": 143310 }, { "grad_norm": 0.41221609711647034, "learning_rate": 2.0397786407631338e-05, "loss": 0.022, "step": 143320 }, { "grad_norm": 0.31906911730766296, "learning_rate": 2.0391124110703673e-05, "loss": 0.0514, "step": 143330 }, { "grad_norm": 0.4689393639564514, "learning_rate": 2.0384462623271543e-05, "loss": 0.0233, "step": 143340 }, { "grad_norm": 0.4110751450061798, "learning_rate": 2.0377801945517095e-05, "loss": 0.022, "step": 143350 }, { "grad_norm": 0.2765221893787384, "learning_rate": 2.0371142077622408e-05, "loss": 0.0151, "step": 143360 }, { "grad_norm": 0.2915133833885193, "learning_rate": 2.0364483019769554e-05, "loss": 0.0208, "step": 143370 }, { "grad_norm": 0.44979315996170044, "learning_rate": 2.0357824772140614e-05, "loss": 0.0254, "step": 143380 }, { "grad_norm": 0.4533785879611969, "learning_rate": 2.035116733491761e-05, "loss": 0.0201, "step": 143390 }, { "grad_norm": 0.9740491509437561, "learning_rate": 2.0344510708282556e-05, "loss": 0.0226, "step": 143400 }, { "grad_norm": 0.6504157781600952, "learning_rate": 2.0337854892417414e-05, "loss": 0.02, "step": 143410 }, { "grad_norm": 0.8435676097869873, "learning_rate": 2.033119988750418e-05, "loss": 0.0201, "step": 143420 }, { "grad_norm": 0.15425805747509003, "learning_rate": 2.032454569372481e-05, "loss": 0.0247, "step": 143430 }, { "grad_norm": 0.56208735704422, "learning_rate": 2.0317892311261215e-05, "loss": 0.0195, "step": 143440 }, { "grad_norm": 0.5715231895446777, "learning_rate": 2.031123974029529e-05, "loss": 0.0227, "step": 143450 }, { "grad_norm": 0.686856210231781, "learning_rate": 2.0304587981008904e-05, "loss": 0.0193, "step": 143460 }, { "grad_norm": 0.21530796587467194, "learning_rate": 2.029793703358394e-05, "loss": 0.0158, "step": 143470 }, { "grad_norm": 0.3497295677661896, "learning_rate": 2.0291286898202223e-05, "loss": 0.0216, "step": 143480 }, { "grad_norm": 0.41479170322418213, "learning_rate": 2.028463757504554e-05, "loss": 0.0209, "step": 143490 }, { "grad_norm": 0.4820599853992462, "learning_rate": 2.027798906429572e-05, "loss": 0.0228, "step": 143500 }, { "grad_norm": 0.4289897680282593, "learning_rate": 2.0271341366134493e-05, "loss": 0.0159, "step": 143510 }, { "grad_norm": 0.878868043422699, "learning_rate": 2.026469448074367e-05, "loss": 0.0255, "step": 143520 }, { "grad_norm": 0.3989374339580536, "learning_rate": 2.025804840830488e-05, "loss": 0.0217, "step": 143530 }, { "grad_norm": 0.608498215675354, "learning_rate": 2.025140314899988e-05, "loss": 0.0236, "step": 143540 }, { "grad_norm": 0.5743672847747803, "learning_rate": 2.0244758703010358e-05, "loss": 0.0255, "step": 143550 }, { "grad_norm": 0.8249819278717041, "learning_rate": 2.023811507051796e-05, "loss": 0.024, "step": 143560 }, { "grad_norm": 0.34686991572380066, "learning_rate": 2.023147225170431e-05, "loss": 0.0132, "step": 143570 }, { "grad_norm": 0.5257822871208191, "learning_rate": 2.022483024675101e-05, "loss": 0.0204, "step": 143580 }, { "grad_norm": 0.48209458589553833, "learning_rate": 2.0218189055839677e-05, "loss": 0.019, "step": 143590 }, { "grad_norm": 0.26137837767601013, "learning_rate": 2.021154867915187e-05, "loss": 0.0147, "step": 143600 }, { "grad_norm": 0.6792541742324829, "learning_rate": 2.0204909116869115e-05, "loss": 0.023, "step": 143610 }, { "grad_norm": 0.32103875279426575, "learning_rate": 2.0198270369172968e-05, "loss": 0.0258, "step": 143620 }, { "grad_norm": 0.1946636140346527, "learning_rate": 2.0191632436244894e-05, "loss": 0.0166, "step": 143630 }, { "grad_norm": 0.2727496325969696, "learning_rate": 2.018499531826641e-05, "loss": 0.0195, "step": 143640 }, { "grad_norm": 1.3229877948760986, "learning_rate": 2.017835901541895e-05, "loss": 0.0189, "step": 143650 }, { "grad_norm": 0.5344796776771545, "learning_rate": 2.017172352788394e-05, "loss": 0.0225, "step": 143660 }, { "grad_norm": 0.41904303431510925, "learning_rate": 2.0165088855842813e-05, "loss": 0.0197, "step": 143670 }, { "grad_norm": 0.5405056476593018, "learning_rate": 2.015845499947695e-05, "loss": 0.0348, "step": 143680 }, { "grad_norm": 0.36127403378486633, "learning_rate": 2.015182195896772e-05, "loss": 0.0283, "step": 143690 }, { "grad_norm": 0.3017818331718445, "learning_rate": 2.0145189734496446e-05, "loss": 0.0172, "step": 143700 }, { "grad_norm": 0.8995538949966431, "learning_rate": 2.0138558326244483e-05, "loss": 0.0193, "step": 143710 }, { "grad_norm": 0.7702877521514893, "learning_rate": 2.0131927734393103e-05, "loss": 0.0286, "step": 143720 }, { "grad_norm": 0.22330498695373535, "learning_rate": 2.0125297959123613e-05, "loss": 0.0311, "step": 143730 }, { "grad_norm": 0.7500833868980408, "learning_rate": 2.0118669000617253e-05, "loss": 0.0185, "step": 143740 }, { "grad_norm": 0.2773337960243225, "learning_rate": 2.011204085905524e-05, "loss": 0.0295, "step": 143750 }, { "grad_norm": 0.3598548471927643, "learning_rate": 2.0105413534618822e-05, "loss": 0.0176, "step": 143760 }, { "grad_norm": 0.4615829586982727, "learning_rate": 2.0098787027489163e-05, "loss": 0.0217, "step": 143770 }, { "grad_norm": 0.3262762129306793, "learning_rate": 2.0092161337847433e-05, "loss": 0.0155, "step": 143780 }, { "grad_norm": 0.4055663049221039, "learning_rate": 2.008553646587476e-05, "loss": 0.0137, "step": 143790 }, { "grad_norm": 3.1196370124816895, "learning_rate": 2.007891241175228e-05, "loss": 0.0223, "step": 143800 }, { "grad_norm": 0.29093825817108154, "learning_rate": 2.0072289175661136e-05, "loss": 0.0144, "step": 143810 }, { "grad_norm": 0.41365042328834534, "learning_rate": 2.006566675778232e-05, "loss": 0.0171, "step": 143820 }, { "grad_norm": 0.5105692148208618, "learning_rate": 2.0059045158296948e-05, "loss": 0.0226, "step": 143830 }, { "grad_norm": 0.4463516175746918, "learning_rate": 2.0052424377386013e-05, "loss": 0.0169, "step": 143840 }, { "grad_norm": 0.5433361530303955, "learning_rate": 2.004580441523057e-05, "loss": 0.0203, "step": 143850 }, { "grad_norm": 0.46988484263420105, "learning_rate": 2.0039185272011575e-05, "loss": 0.0199, "step": 143860 }, { "grad_norm": 0.31681790947914124, "learning_rate": 2.0032566947909986e-05, "loss": 0.0137, "step": 143870 }, { "grad_norm": 0.4108678996562958, "learning_rate": 2.0025949443106772e-05, "loss": 0.0257, "step": 143880 }, { "grad_norm": 0.5903523564338684, "learning_rate": 2.0019332757782838e-05, "loss": 0.0149, "step": 143890 }, { "grad_norm": 0.4418486952781677, "learning_rate": 2.0012716892119087e-05, "loss": 0.024, "step": 143900 }, { "grad_norm": 0.3250992000102997, "learning_rate": 2.000610184629637e-05, "loss": 0.0219, "step": 143910 }, { "grad_norm": 0.31661057472229004, "learning_rate": 1.9999487620495566e-05, "loss": 0.0156, "step": 143920 }, { "grad_norm": 0.4736306965351105, "learning_rate": 1.9992874214897512e-05, "loss": 0.0185, "step": 143930 }, { "grad_norm": 0.3389824628829956, "learning_rate": 1.9986261629683008e-05, "loss": 0.019, "step": 143940 }, { "grad_norm": 0.30323082208633423, "learning_rate": 1.997964986503283e-05, "loss": 0.0178, "step": 143950 }, { "grad_norm": 0.4086582660675049, "learning_rate": 1.9973038921127728e-05, "loss": 0.0217, "step": 143960 }, { "grad_norm": 0.475615918636322, "learning_rate": 1.9966428798148483e-05, "loss": 0.0154, "step": 143970 }, { "grad_norm": 0.27298977971076965, "learning_rate": 1.9959819496275794e-05, "loss": 0.0168, "step": 143980 }, { "grad_norm": 0.39841270446777344, "learning_rate": 1.995321101569033e-05, "loss": 0.0218, "step": 143990 }, { "grad_norm": 0.31404417753219604, "learning_rate": 1.994660335657281e-05, "loss": 0.0349, "step": 144000 }, { "grad_norm": 0.36679986119270325, "learning_rate": 1.9939996519103848e-05, "loss": 0.0212, "step": 144010 }, { "grad_norm": 0.38242459297180176, "learning_rate": 1.9933390503464123e-05, "loss": 0.0152, "step": 144020 }, { "grad_norm": 0.32454997301101685, "learning_rate": 1.992678530983417e-05, "loss": 0.0238, "step": 144030 }, { "grad_norm": 0.29328957200050354, "learning_rate": 1.9920180938394607e-05, "loss": 0.032, "step": 144040 }, { "grad_norm": 0.7809164524078369, "learning_rate": 1.9913577389326015e-05, "loss": 0.0192, "step": 144050 }, { "grad_norm": 0.3205682933330536, "learning_rate": 1.9906974662808912e-05, "loss": 0.0184, "step": 144060 }, { "grad_norm": 0.795133650302887, "learning_rate": 1.9900372759023822e-05, "loss": 0.0231, "step": 144070 }, { "grad_norm": 0.2896899878978729, "learning_rate": 1.9893771678151213e-05, "loss": 0.0175, "step": 144080 }, { "grad_norm": 0.3636912703514099, "learning_rate": 1.988717142037159e-05, "loss": 0.0439, "step": 144090 }, { "grad_norm": 0.37881016731262207, "learning_rate": 1.9880571985865393e-05, "loss": 0.0188, "step": 144100 }, { "grad_norm": 0.3901784121990204, "learning_rate": 1.987397337481302e-05, "loss": 0.0236, "step": 144110 }, { "grad_norm": 0.5074739456176758, "learning_rate": 1.986737558739492e-05, "loss": 0.021, "step": 144120 }, { "grad_norm": 0.3024890422821045, "learning_rate": 1.9860778623791432e-05, "loss": 0.0183, "step": 144130 }, { "grad_norm": 0.6992849111557007, "learning_rate": 1.985418248418295e-05, "loss": 0.0206, "step": 144140 }, { "grad_norm": 0.5226829648017883, "learning_rate": 1.9847587168749794e-05, "loss": 0.0171, "step": 144150 }, { "grad_norm": 0.5715170502662659, "learning_rate": 1.984099267767227e-05, "loss": 0.0193, "step": 144160 }, { "grad_norm": 0.2841614782810211, "learning_rate": 1.9834399011130665e-05, "loss": 0.0136, "step": 144170 }, { "grad_norm": 0.6247631311416626, "learning_rate": 1.982780616930528e-05, "loss": 0.0363, "step": 144180 }, { "grad_norm": 0.5990033149719238, "learning_rate": 1.9821214152376332e-05, "loss": 0.025, "step": 144190 }, { "grad_norm": 0.6108908653259277, "learning_rate": 1.9814622960524038e-05, "loss": 0.0228, "step": 144200 }, { "grad_norm": 0.3182103931903839, "learning_rate": 1.980803259392863e-05, "loss": 0.0233, "step": 144210 }, { "grad_norm": 0.31767508387565613, "learning_rate": 1.9801443052770245e-05, "loss": 0.0246, "step": 144220 }, { "grad_norm": 0.41450706124305725, "learning_rate": 1.9794854337229085e-05, "loss": 0.0226, "step": 144230 }, { "grad_norm": 0.4982900619506836, "learning_rate": 1.9788266447485258e-05, "loss": 0.0196, "step": 144240 }, { "grad_norm": 0.6171665191650391, "learning_rate": 1.978167938371886e-05, "loss": 0.0264, "step": 144250 }, { "grad_norm": 0.32409167289733887, "learning_rate": 1.9775093146110013e-05, "loss": 0.0215, "step": 144260 }, { "grad_norm": 0.6511872410774231, "learning_rate": 1.9768507734838766e-05, "loss": 0.0189, "step": 144270 }, { "grad_norm": 0.3743637502193451, "learning_rate": 1.9761923150085156e-05, "loss": 0.0294, "step": 144280 }, { "grad_norm": 0.21030882000923157, "learning_rate": 1.9755339392029186e-05, "loss": 0.0183, "step": 144290 }, { "grad_norm": 0.22663360834121704, "learning_rate": 1.974875646085088e-05, "loss": 0.0267, "step": 144300 }, { "grad_norm": 0.3689880967140198, "learning_rate": 1.9742174356730248e-05, "loss": 0.028, "step": 144310 }, { "grad_norm": 0.6195332407951355, "learning_rate": 1.973559307984716e-05, "loss": 0.0118, "step": 144320 }, { "grad_norm": 0.3002958297729492, "learning_rate": 1.9729012630381604e-05, "loss": 0.0165, "step": 144330 }, { "grad_norm": 0.8771830797195435, "learning_rate": 1.9722433008513457e-05, "loss": 0.0287, "step": 144340 }, { "grad_norm": 0.46822410821914673, "learning_rate": 1.971585421442263e-05, "loss": 0.0253, "step": 144350 }, { "grad_norm": 0.6467915773391724, "learning_rate": 1.9709276248288978e-05, "loss": 0.031, "step": 144360 }, { "grad_norm": 0.3175774812698364, "learning_rate": 1.9702699110292307e-05, "loss": 0.0217, "step": 144370 }, { "grad_norm": 0.9056268930435181, "learning_rate": 1.9696122800612482e-05, "loss": 0.0329, "step": 144380 }, { "grad_norm": 0.45004215836524963, "learning_rate": 1.9689547319429274e-05, "loss": 0.0195, "step": 144390 }, { "grad_norm": 0.2344980686903, "learning_rate": 1.968297266692246e-05, "loss": 0.0147, "step": 144400 }, { "grad_norm": 0.33607375621795654, "learning_rate": 1.9676398843271767e-05, "loss": 0.0216, "step": 144410 }, { "grad_norm": 0.8593150973320007, "learning_rate": 1.9669825848656948e-05, "loss": 0.0202, "step": 144420 }, { "grad_norm": 0.4644504487514496, "learning_rate": 1.9663253683257684e-05, "loss": 0.0178, "step": 144430 }, { "grad_norm": 0.2443803995847702, "learning_rate": 1.965668234725368e-05, "loss": 0.0347, "step": 144440 }, { "grad_norm": 0.43142926692962646, "learning_rate": 1.9650111840824582e-05, "loss": 0.0281, "step": 144450 }, { "grad_norm": 0.24175147712230682, "learning_rate": 1.964354216415001e-05, "loss": 0.0194, "step": 144460 }, { "grad_norm": 0.4888818562030792, "learning_rate": 1.9636973317409603e-05, "loss": 0.0204, "step": 144470 }, { "grad_norm": 0.23544763028621674, "learning_rate": 1.9630405300782946e-05, "loss": 0.0137, "step": 144480 }, { "grad_norm": 0.2854057848453522, "learning_rate": 1.962383811444957e-05, "loss": 0.018, "step": 144490 }, { "grad_norm": 0.45312950015068054, "learning_rate": 1.9617271758589077e-05, "loss": 0.0156, "step": 144500 }, { "grad_norm": 0.32592275738716125, "learning_rate": 1.9610706233380936e-05, "loss": 0.0221, "step": 144510 }, { "grad_norm": 0.3381722867488861, "learning_rate": 1.960414153900471e-05, "loss": 0.0212, "step": 144520 }, { "grad_norm": 0.3900507986545563, "learning_rate": 1.9597577675639795e-05, "loss": 0.0195, "step": 144530 }, { "grad_norm": 0.21844883263111115, "learning_rate": 1.9591014643465706e-05, "loss": 0.0192, "step": 144540 }, { "grad_norm": 0.34611812233924866, "learning_rate": 1.958445244266183e-05, "loss": 0.0227, "step": 144550 }, { "grad_norm": 0.405966192483902, "learning_rate": 1.9577891073407622e-05, "loss": 0.0179, "step": 144560 }, { "grad_norm": 0.223039910197258, "learning_rate": 1.9571330535882437e-05, "loss": 0.0171, "step": 144570 }, { "grad_norm": 0.3487691879272461, "learning_rate": 1.9564770830265633e-05, "loss": 0.0163, "step": 144580 }, { "grad_norm": 0.25650152564048767, "learning_rate": 1.955821195673657e-05, "loss": 0.024, "step": 144590 }, { "grad_norm": 0.21157948672771454, "learning_rate": 1.955165391547456e-05, "loss": 0.0164, "step": 144600 }, { "grad_norm": 0.49615180492401123, "learning_rate": 1.9545096706658893e-05, "loss": 0.0222, "step": 144610 }, { "grad_norm": 0.3578959107398987, "learning_rate": 1.9538540330468826e-05, "loss": 0.0229, "step": 144620 }, { "grad_norm": 0.33544638752937317, "learning_rate": 1.953198478708362e-05, "loss": 0.0167, "step": 144630 }, { "grad_norm": 0.322536438703537, "learning_rate": 1.9525430076682516e-05, "loss": 0.0216, "step": 144640 }, { "grad_norm": 0.4050963819026947, "learning_rate": 1.9518876199444707e-05, "loss": 0.0259, "step": 144650 }, { "grad_norm": 0.3681952655315399, "learning_rate": 1.9512323155549377e-05, "loss": 0.0168, "step": 144660 }, { "grad_norm": 0.2527701258659363, "learning_rate": 1.9505770945175656e-05, "loss": 0.0102, "step": 144670 }, { "grad_norm": 0.518076479434967, "learning_rate": 1.949921956850272e-05, "loss": 0.0204, "step": 144680 }, { "grad_norm": 0.3646335303783417, "learning_rate": 1.9492669025709665e-05, "loss": 0.0116, "step": 144690 }, { "grad_norm": 0.26168113946914673, "learning_rate": 1.9486119316975553e-05, "loss": 0.0166, "step": 144700 }, { "grad_norm": 0.4679000675678253, "learning_rate": 1.94795704424795e-05, "loss": 0.0171, "step": 144710 }, { "grad_norm": 0.5420842170715332, "learning_rate": 1.9473022402400503e-05, "loss": 0.0272, "step": 144720 }, { "grad_norm": 0.2566165328025818, "learning_rate": 1.9466475196917643e-05, "loss": 0.0382, "step": 144730 }, { "grad_norm": 0.20991675555706024, "learning_rate": 1.9459928826209845e-05, "loss": 0.0182, "step": 144740 }, { "grad_norm": 0.3720760941505432, "learning_rate": 1.945338329045612e-05, "loss": 0.023, "step": 144750 }, { "grad_norm": 0.3453637659549713, "learning_rate": 1.9446838589835437e-05, "loss": 0.0182, "step": 144760 }, { "grad_norm": 0.7578839659690857, "learning_rate": 1.944029472452671e-05, "loss": 0.0183, "step": 144770 }, { "grad_norm": 0.3970368504524231, "learning_rate": 1.943375169470884e-05, "loss": 0.0165, "step": 144780 }, { "grad_norm": 0.3436891436576843, "learning_rate": 1.94272095005607e-05, "loss": 0.0172, "step": 144790 }, { "grad_norm": 0.18644028902053833, "learning_rate": 1.942066814226119e-05, "loss": 0.0168, "step": 144800 }, { "grad_norm": 0.5571170449256897, "learning_rate": 1.9414127619989115e-05, "loss": 0.0286, "step": 144810 }, { "grad_norm": 0.1653672456741333, "learning_rate": 1.940758793392329e-05, "loss": 0.0251, "step": 144820 }, { "grad_norm": 0.39211252331733704, "learning_rate": 1.9401049084242533e-05, "loss": 0.0192, "step": 144830 }, { "grad_norm": 0.24165426194667816, "learning_rate": 1.9394511071125582e-05, "loss": 0.023, "step": 144840 }, { "grad_norm": 0.7436455488204956, "learning_rate": 1.9387973894751217e-05, "loss": 0.0278, "step": 144850 }, { "grad_norm": 0.183578759431839, "learning_rate": 1.9381437555298142e-05, "loss": 0.02, "step": 144860 }, { "grad_norm": 0.3772742748260498, "learning_rate": 1.937490205294505e-05, "loss": 0.0235, "step": 144870 }, { "grad_norm": 0.4221534729003906, "learning_rate": 1.936836738787064e-05, "loss": 0.0216, "step": 144880 }, { "grad_norm": 0.1888568103313446, "learning_rate": 1.9361833560253567e-05, "loss": 0.018, "step": 144890 }, { "grad_norm": 0.3719569742679596, "learning_rate": 1.935530057027245e-05, "loss": 0.0155, "step": 144900 }, { "grad_norm": 0.37689295411109924, "learning_rate": 1.9348768418105884e-05, "loss": 0.0164, "step": 144910 }, { "grad_norm": 0.43438246846199036, "learning_rate": 1.934223710393249e-05, "loss": 0.0229, "step": 144920 }, { "grad_norm": 0.44437822699546814, "learning_rate": 1.9335706627930804e-05, "loss": 0.0207, "step": 144930 }, { "grad_norm": 0.4170830547809601, "learning_rate": 1.932917699027939e-05, "loss": 0.0225, "step": 144940 }, { "grad_norm": 0.39970675110816956, "learning_rate": 1.932264819115676e-05, "loss": 0.0253, "step": 144950 }, { "grad_norm": 0.3262040615081787, "learning_rate": 1.9316120230741385e-05, "loss": 0.0188, "step": 144960 }, { "grad_norm": 0.276217520236969, "learning_rate": 1.9309593109211773e-05, "loss": 0.0232, "step": 144970 }, { "grad_norm": 0.5735220313072205, "learning_rate": 1.9303066826746358e-05, "loss": 0.0113, "step": 144980 }, { "grad_norm": 0.2824963331222534, "learning_rate": 1.929654138352356e-05, "loss": 0.0171, "step": 144990 }, { "grad_norm": 0.33909526467323303, "learning_rate": 1.9290016779721763e-05, "loss": 0.0165, "step": 145000 }, { "grad_norm": 0.30124834179878235, "learning_rate": 1.928349301551938e-05, "loss": 0.0268, "step": 145010 }, { "grad_norm": 1.862272024154663, "learning_rate": 1.927697009109479e-05, "loss": 0.0196, "step": 145020 }, { "grad_norm": 0.2776407301425934, "learning_rate": 1.9270448006626256e-05, "loss": 0.0165, "step": 145030 }, { "grad_norm": 0.5975817441940308, "learning_rate": 1.926392676229215e-05, "loss": 0.0177, "step": 145040 }, { "grad_norm": 0.2882000505924225, "learning_rate": 1.9257406358270724e-05, "loss": 0.0206, "step": 145050 }, { "grad_norm": 0.7650305032730103, "learning_rate": 1.9250886794740275e-05, "loss": 0.0271, "step": 145060 }, { "grad_norm": 0.5766596794128418, "learning_rate": 1.9244368071879027e-05, "loss": 0.0296, "step": 145070 }, { "grad_norm": 0.8145327568054199, "learning_rate": 1.9237850189865187e-05, "loss": 0.0272, "step": 145080 }, { "grad_norm": 0.36917704343795776, "learning_rate": 1.9231333148876984e-05, "loss": 0.0204, "step": 145090 }, { "grad_norm": 0.5448872447013855, "learning_rate": 1.9224816949092574e-05, "loss": 0.02, "step": 145100 }, { "grad_norm": 0.3892684578895569, "learning_rate": 1.9218301590690103e-05, "loss": 0.0332, "step": 145110 }, { "grad_norm": 0.27807796001434326, "learning_rate": 1.9211787073847682e-05, "loss": 0.0159, "step": 145120 }, { "grad_norm": 0.5572267770767212, "learning_rate": 1.920527339874344e-05, "loss": 0.0264, "step": 145130 }, { "grad_norm": 0.2564287781715393, "learning_rate": 1.9198760565555473e-05, "loss": 0.014, "step": 145140 }, { "grad_norm": 0.6202947497367859, "learning_rate": 1.9192248574461818e-05, "loss": 0.0244, "step": 145150 }, { "grad_norm": 0.1980433315038681, "learning_rate": 1.9185737425640517e-05, "loss": 0.0141, "step": 145160 }, { "grad_norm": 0.29020053148269653, "learning_rate": 1.9179227119269555e-05, "loss": 0.0197, "step": 145170 }, { "grad_norm": 0.33111661672592163, "learning_rate": 1.9172717655526967e-05, "loss": 0.0168, "step": 145180 }, { "grad_norm": 0.17185410857200623, "learning_rate": 1.91662090345907e-05, "loss": 0.0156, "step": 145190 }, { "grad_norm": 0.3729868531227112, "learning_rate": 1.915970125663867e-05, "loss": 0.0205, "step": 145200 }, { "grad_norm": 0.19966299831867218, "learning_rate": 1.915319432184884e-05, "loss": 0.0229, "step": 145210 }, { "grad_norm": 0.3303511142730713, "learning_rate": 1.9146688230399075e-05, "loss": 0.0251, "step": 145220 }, { "grad_norm": 0.41546809673309326, "learning_rate": 1.9140182982467292e-05, "loss": 0.019, "step": 145230 }, { "grad_norm": 0.20987653732299805, "learning_rate": 1.9133678578231286e-05, "loss": 0.0176, "step": 145240 }, { "grad_norm": 0.1905069202184677, "learning_rate": 1.9127175017868902e-05, "loss": 0.024, "step": 145250 }, { "grad_norm": 0.3022612929344177, "learning_rate": 1.9120672301557978e-05, "loss": 0.0218, "step": 145260 }, { "grad_norm": 0.4065278172492981, "learning_rate": 1.911417042947627e-05, "loss": 0.0276, "step": 145270 }, { "grad_norm": 0.2526227533817291, "learning_rate": 1.9107669401801543e-05, "loss": 0.0173, "step": 145280 }, { "grad_norm": 0.42762789130210876, "learning_rate": 1.9101169218711508e-05, "loss": 0.0158, "step": 145290 }, { "grad_norm": 0.290932834148407, "learning_rate": 1.9094669880383915e-05, "loss": 0.0179, "step": 145300 }, { "grad_norm": 0.3220556080341339, "learning_rate": 1.908817138699644e-05, "loss": 0.0121, "step": 145310 }, { "grad_norm": 0.3065010905265808, "learning_rate": 1.908167373872673e-05, "loss": 0.0181, "step": 145320 }, { "grad_norm": 0.3274383246898651, "learning_rate": 1.907517693575246e-05, "loss": 0.0259, "step": 145330 }, { "grad_norm": 0.7642289996147156, "learning_rate": 1.9068680978251213e-05, "loss": 0.0241, "step": 145340 }, { "grad_norm": 1.7477257251739502, "learning_rate": 1.9062185866400628e-05, "loss": 0.0223, "step": 145350 }, { "grad_norm": 0.48825785517692566, "learning_rate": 1.905569160037826e-05, "loss": 0.0181, "step": 145360 }, { "grad_norm": 0.39599233865737915, "learning_rate": 1.904919818036166e-05, "loss": 0.017, "step": 145370 }, { "grad_norm": 0.3262137770652771, "learning_rate": 1.9042705606528333e-05, "loss": 0.0231, "step": 145380 }, { "grad_norm": 0.27814361453056335, "learning_rate": 1.9036213879055825e-05, "loss": 0.0171, "step": 145390 }, { "grad_norm": 0.22804638743400574, "learning_rate": 1.902972299812159e-05, "loss": 0.0167, "step": 145400 }, { "grad_norm": 0.5321765542030334, "learning_rate": 1.9023232963903086e-05, "loss": 0.0281, "step": 145410 }, { "grad_norm": 0.6440600156784058, "learning_rate": 1.9016743776577768e-05, "loss": 0.0217, "step": 145420 }, { "grad_norm": 0.34315773844718933, "learning_rate": 1.9010255436323016e-05, "loss": 0.0254, "step": 145430 }, { "grad_norm": 0.48924386501312256, "learning_rate": 1.900376794331626e-05, "loss": 0.0202, "step": 145440 }, { "grad_norm": 0.349382609128952, "learning_rate": 1.8997281297734835e-05, "loss": 0.0151, "step": 145450 }, { "grad_norm": 0.3596048355102539, "learning_rate": 1.8990795499756087e-05, "loss": 0.0214, "step": 145460 }, { "grad_norm": 0.38219061493873596, "learning_rate": 1.898431054955735e-05, "loss": 0.0137, "step": 145470 }, { "grad_norm": 0.2373628169298172, "learning_rate": 1.897782644731591e-05, "loss": 0.0299, "step": 145480 }, { "grad_norm": 0.2747536599636078, "learning_rate": 1.8971343193209045e-05, "loss": 0.0221, "step": 145490 }, { "grad_norm": 1.1724754571914673, "learning_rate": 1.896486078741398e-05, "loss": 0.032, "step": 145500 }, { "grad_norm": 0.5452468991279602, "learning_rate": 1.895837923010796e-05, "loss": 0.0144, "step": 145510 }, { "grad_norm": 0.31970009207725525, "learning_rate": 1.895189852146823e-05, "loss": 0.0237, "step": 145520 }, { "grad_norm": 0.535266101360321, "learning_rate": 1.894541866167189e-05, "loss": 0.017, "step": 145530 }, { "grad_norm": 0.7851859331130981, "learning_rate": 1.893893965089616e-05, "loss": 0.0162, "step": 145540 }, { "grad_norm": 0.5566343069076538, "learning_rate": 1.8932461489318125e-05, "loss": 0.0189, "step": 145550 }, { "grad_norm": 0.3335373103618622, "learning_rate": 1.8925984177114946e-05, "loss": 0.0158, "step": 145560 }, { "grad_norm": 0.8262741565704346, "learning_rate": 1.8919507714463685e-05, "loss": 0.0212, "step": 145570 }, { "grad_norm": 0.37976202368736267, "learning_rate": 1.891303210154139e-05, "loss": 0.0206, "step": 145580 }, { "grad_norm": 0.5740960240364075, "learning_rate": 1.8906557338525132e-05, "loss": 0.0149, "step": 145590 }, { "grad_norm": 0.3650228679180145, "learning_rate": 1.8900083425591913e-05, "loss": 0.0209, "step": 145600 }, { "grad_norm": 0.19176332652568817, "learning_rate": 1.8893610362918736e-05, "loss": 0.0127, "step": 145610 }, { "grad_norm": 0.3547498881816864, "learning_rate": 1.8887138150682548e-05, "loss": 0.0151, "step": 145620 }, { "grad_norm": 0.3499632775783539, "learning_rate": 1.8880666789060326e-05, "loss": 0.0164, "step": 145630 }, { "grad_norm": 0.3075156509876251, "learning_rate": 1.8874196278228973e-05, "loss": 0.0161, "step": 145640 }, { "grad_norm": 0.3975301682949066, "learning_rate": 1.8867726618365412e-05, "loss": 0.0178, "step": 145650 }, { "grad_norm": 0.5286100506782532, "learning_rate": 1.886125780964651e-05, "loss": 0.0251, "step": 145660 }, { "grad_norm": 0.2560763657093048, "learning_rate": 1.8854789852249106e-05, "loss": 0.0172, "step": 145670 }, { "grad_norm": 0.319614440202713, "learning_rate": 1.884832274635006e-05, "loss": 0.0117, "step": 145680 }, { "grad_norm": 0.49147579073905945, "learning_rate": 1.884185649212617e-05, "loss": 0.0173, "step": 145690 }, { "grad_norm": 0.385210245847702, "learning_rate": 1.88353910897542e-05, "loss": 0.0234, "step": 145700 }, { "grad_norm": 0.6498017311096191, "learning_rate": 1.882892653941095e-05, "loss": 0.0204, "step": 145710 }, { "grad_norm": 0.4304610788822174, "learning_rate": 1.882246284127312e-05, "loss": 0.0192, "step": 145720 }, { "grad_norm": 0.6911530494689941, "learning_rate": 1.8815999995517475e-05, "loss": 0.0207, "step": 145730 }, { "grad_norm": 0.6003789901733398, "learning_rate": 1.8809538002320644e-05, "loss": 0.0161, "step": 145740 }, { "grad_norm": 0.3100363612174988, "learning_rate": 1.8803076861859338e-05, "loss": 0.0168, "step": 145750 }, { "grad_norm": 0.3736073076725006, "learning_rate": 1.879661657431017e-05, "loss": 0.0183, "step": 145760 }, { "grad_norm": 0.2974121868610382, "learning_rate": 1.8790157139849813e-05, "loss": 0.0173, "step": 145770 }, { "grad_norm": 0.34348762035369873, "learning_rate": 1.8783698558654823e-05, "loss": 0.0178, "step": 145780 }, { "grad_norm": 0.25950202345848083, "learning_rate": 1.8777240830901777e-05, "loss": 0.0176, "step": 145790 }, { "grad_norm": 0.2306136041879654, "learning_rate": 1.877078395676725e-05, "loss": 0.0186, "step": 145800 }, { "grad_norm": 0.31622761487960815, "learning_rate": 1.8764327936427757e-05, "loss": 0.0218, "step": 145810 }, { "grad_norm": 0.37383121252059937, "learning_rate": 1.8757872770059803e-05, "loss": 0.0196, "step": 145820 }, { "grad_norm": 0.544363260269165, "learning_rate": 1.875141845783986e-05, "loss": 0.0408, "step": 145830 }, { "grad_norm": 0.720500648021698, "learning_rate": 1.8744964999944392e-05, "loss": 0.0165, "step": 145840 }, { "grad_norm": 0.39389318227767944, "learning_rate": 1.873851239654985e-05, "loss": 0.0171, "step": 145850 }, { "grad_norm": 0.4030713140964508, "learning_rate": 1.8732060647832645e-05, "loss": 0.0223, "step": 145860 }, { "grad_norm": 0.5244157314300537, "learning_rate": 1.8725609753969146e-05, "loss": 0.0185, "step": 145870 }, { "grad_norm": 0.3985961377620697, "learning_rate": 1.871915971513572e-05, "loss": 0.0197, "step": 145880 }, { "grad_norm": 0.2616735100746155, "learning_rate": 1.8712710531508724e-05, "loss": 0.0207, "step": 145890 }, { "grad_norm": 0.2821221649646759, "learning_rate": 1.8706262203264475e-05, "loss": 0.0118, "step": 145900 }, { "grad_norm": 1.137148141860962, "learning_rate": 1.869981473057924e-05, "loss": 0.0161, "step": 145910 }, { "grad_norm": 0.4853840470314026, "learning_rate": 1.869336811362933e-05, "loss": 0.0242, "step": 145920 }, { "grad_norm": 0.5004569888114929, "learning_rate": 1.868692235259095e-05, "loss": 0.0258, "step": 145930 }, { "grad_norm": 0.41328951716423035, "learning_rate": 1.8680477447640394e-05, "loss": 0.0186, "step": 145940 }, { "grad_norm": 0.4598420560359955, "learning_rate": 1.8674033398953777e-05, "loss": 0.029, "step": 145950 }, { "grad_norm": 0.2701210081577301, "learning_rate": 1.8667590206707318e-05, "loss": 0.0139, "step": 145960 }, { "grad_norm": 0.3491255044937134, "learning_rate": 1.8661147871077196e-05, "loss": 0.0189, "step": 145970 }, { "grad_norm": 0.4107924699783325, "learning_rate": 1.8654706392239514e-05, "loss": 0.0173, "step": 145980 }, { "grad_norm": 0.5660713911056519, "learning_rate": 1.8648265770370384e-05, "loss": 0.0281, "step": 145990 }, { "grad_norm": 0.22749784588813782, "learning_rate": 1.8641826005645867e-05, "loss": 0.0197, "step": 146000 }, { "grad_norm": 0.2354765087366104, "learning_rate": 1.863538709824207e-05, "loss": 0.0161, "step": 146010 }, { "grad_norm": 0.3202509582042694, "learning_rate": 1.8628949048335005e-05, "loss": 0.0146, "step": 146020 }, { "grad_norm": 0.34078624844551086, "learning_rate": 1.8622511856100665e-05, "loss": 0.0214, "step": 146030 }, { "grad_norm": 0.39401593804359436, "learning_rate": 1.8616075521715088e-05, "loss": 0.0201, "step": 146040 }, { "grad_norm": 0.3856837749481201, "learning_rate": 1.8609640045354194e-05, "loss": 0.0154, "step": 146050 }, { "grad_norm": 0.3928774297237396, "learning_rate": 1.8603205427193964e-05, "loss": 0.0152, "step": 146060 }, { "grad_norm": 0.32048487663269043, "learning_rate": 1.8596771667410306e-05, "loss": 0.0175, "step": 146070 }, { "grad_norm": 0.4401865303516388, "learning_rate": 1.8590338766179093e-05, "loss": 0.0175, "step": 146080 }, { "grad_norm": 0.6877092123031616, "learning_rate": 1.8583906723676242e-05, "loss": 0.0232, "step": 146090 }, { "grad_norm": 0.6194220185279846, "learning_rate": 1.857747554007757e-05, "loss": 0.0188, "step": 146100 }, { "grad_norm": 0.2813850939273834, "learning_rate": 1.857104521555892e-05, "loss": 0.0175, "step": 146110 }, { "grad_norm": 0.39203324913978577, "learning_rate": 1.856461575029606e-05, "loss": 0.0167, "step": 146120 }, { "grad_norm": 0.27505576610565186, "learning_rate": 1.8558187144464822e-05, "loss": 0.0132, "step": 146130 }, { "grad_norm": 0.3641904592514038, "learning_rate": 1.8551759398240914e-05, "loss": 0.0175, "step": 146140 }, { "grad_norm": 0.3550736606121063, "learning_rate": 1.8545332511800106e-05, "loss": 0.016, "step": 146150 }, { "grad_norm": 0.4543795883655548, "learning_rate": 1.853890648531809e-05, "loss": 0.0218, "step": 146160 }, { "grad_norm": 0.3930942714214325, "learning_rate": 1.8532481318970535e-05, "loss": 0.0219, "step": 146170 }, { "grad_norm": 0.46756643056869507, "learning_rate": 1.8526057012933134e-05, "loss": 0.0218, "step": 146180 }, { "grad_norm": 0.6223098039627075, "learning_rate": 1.851963356738151e-05, "loss": 0.021, "step": 146190 }, { "grad_norm": 1.098658561706543, "learning_rate": 1.8513210982491284e-05, "loss": 0.026, "step": 146200 }, { "grad_norm": 0.39492225646972656, "learning_rate": 1.8506789258438016e-05, "loss": 0.0155, "step": 146210 }, { "grad_norm": 0.167291060090065, "learning_rate": 1.8500368395397306e-05, "loss": 0.0146, "step": 146220 }, { "grad_norm": 0.3035621643066406, "learning_rate": 1.8493948393544718e-05, "loss": 0.0111, "step": 146230 }, { "grad_norm": 0.5151153206825256, "learning_rate": 1.8487529253055712e-05, "loss": 0.0204, "step": 146240 }, { "grad_norm": 0.2675969898700714, "learning_rate": 1.8481110974105835e-05, "loss": 0.02, "step": 146250 }, { "grad_norm": 0.28036320209503174, "learning_rate": 1.8474693556870527e-05, "loss": 0.018, "step": 146260 }, { "grad_norm": 1.7787678241729736, "learning_rate": 1.846827700152527e-05, "loss": 0.0215, "step": 146270 }, { "grad_norm": 0.4968436360359192, "learning_rate": 1.846186130824547e-05, "loss": 0.0236, "step": 146280 }, { "grad_norm": 0.19451099634170532, "learning_rate": 1.8455446477206523e-05, "loss": 0.0236, "step": 146290 }, { "grad_norm": 0.30608201026916504, "learning_rate": 1.844903250858383e-05, "loss": 0.0193, "step": 146300 }, { "grad_norm": 0.19892540574073792, "learning_rate": 1.844261940255274e-05, "loss": 0.0189, "step": 146310 }, { "grad_norm": 0.35391756892204285, "learning_rate": 1.8436207159288578e-05, "loss": 0.0227, "step": 146320 }, { "grad_norm": 0.2792070806026459, "learning_rate": 1.8429795778966634e-05, "loss": 0.0221, "step": 146330 }, { "grad_norm": 1.4088853597640991, "learning_rate": 1.8423385261762216e-05, "loss": 0.0235, "step": 146340 }, { "grad_norm": 1.0607869625091553, "learning_rate": 1.84169756078506e-05, "loss": 0.022, "step": 146350 }, { "grad_norm": 0.29969340562820435, "learning_rate": 1.8410566817407004e-05, "loss": 0.0218, "step": 146360 }, { "grad_norm": 0.7736327648162842, "learning_rate": 1.8404158890606644e-05, "loss": 0.0192, "step": 146370 }, { "grad_norm": 0.40624311566352844, "learning_rate": 1.8397751827624693e-05, "loss": 0.0166, "step": 146380 }, { "grad_norm": 0.2787154018878937, "learning_rate": 1.839134562863635e-05, "loss": 0.0181, "step": 146390 }, { "grad_norm": 0.7096915245056152, "learning_rate": 1.838494029381675e-05, "loss": 0.0326, "step": 146400 }, { "grad_norm": 0.394487202167511, "learning_rate": 1.837853582334098e-05, "loss": 0.0185, "step": 146410 }, { "grad_norm": 1.5351048707962036, "learning_rate": 1.837213221738418e-05, "loss": 0.0236, "step": 146420 }, { "grad_norm": 0.3369779586791992, "learning_rate": 1.8365729476121386e-05, "loss": 0.0159, "step": 146430 }, { "grad_norm": 0.2824101150035858, "learning_rate": 1.8359327599727698e-05, "loss": 0.022, "step": 146440 }, { "grad_norm": 0.4665517210960388, "learning_rate": 1.8352926588378074e-05, "loss": 0.0212, "step": 146450 }, { "grad_norm": 0.27262791991233826, "learning_rate": 1.8346526442247546e-05, "loss": 0.0234, "step": 146460 }, { "grad_norm": 1.3324764966964722, "learning_rate": 1.8340127161511112e-05, "loss": 0.0421, "step": 146470 }, { "grad_norm": 0.5683314800262451, "learning_rate": 1.833372874634371e-05, "loss": 0.0163, "step": 146480 }, { "grad_norm": 0.5184077620506287, "learning_rate": 1.832733119692026e-05, "loss": 0.0186, "step": 146490 }, { "grad_norm": 0.3100760281085968, "learning_rate": 1.8320934513415667e-05, "loss": 0.0253, "step": 146500 }, { "grad_norm": 0.2925208508968353, "learning_rate": 1.8314538696004834e-05, "loss": 0.0126, "step": 146510 }, { "grad_norm": 0.5167959332466125, "learning_rate": 1.830814374486261e-05, "loss": 0.0139, "step": 146520 }, { "grad_norm": 0.538474977016449, "learning_rate": 1.830174966016382e-05, "loss": 0.0141, "step": 146530 }, { "grad_norm": 0.2594202756881714, "learning_rate": 1.8295356442083294e-05, "loss": 0.0143, "step": 146540 }, { "grad_norm": 0.23451228439807892, "learning_rate": 1.82889640907958e-05, "loss": 0.0162, "step": 146550 }, { "grad_norm": 0.2594491243362427, "learning_rate": 1.828257260647613e-05, "loss": 0.0164, "step": 146560 }, { "grad_norm": 0.19229373335838318, "learning_rate": 1.8276181989299006e-05, "loss": 0.0153, "step": 146570 }, { "grad_norm": 1.2162522077560425, "learning_rate": 1.8269792239439156e-05, "loss": 0.0181, "step": 146580 }, { "grad_norm": 0.4101507365703583, "learning_rate": 1.8263403357071242e-05, "loss": 0.0286, "step": 146590 }, { "grad_norm": 0.3581468462944031, "learning_rate": 1.8257015342369978e-05, "loss": 0.0155, "step": 146600 }, { "grad_norm": 0.2995465099811554, "learning_rate": 1.8250628195509988e-05, "loss": 0.0151, "step": 146610 }, { "grad_norm": 1.3689697980880737, "learning_rate": 1.824424191666588e-05, "loss": 0.0242, "step": 146620 }, { "grad_norm": 0.41724589467048645, "learning_rate": 1.823785650601228e-05, "loss": 0.0195, "step": 146630 }, { "grad_norm": 0.21262744069099426, "learning_rate": 1.823147196372374e-05, "loss": 0.0355, "step": 146640 }, { "grad_norm": 0.4248824417591095, "learning_rate": 1.8225088289974835e-05, "loss": 0.016, "step": 146650 }, { "grad_norm": 0.6408153176307678, "learning_rate": 1.821870548494008e-05, "loss": 0.0182, "step": 146660 }, { "grad_norm": 0.22080402076244354, "learning_rate": 1.821232354879396e-05, "loss": 0.022, "step": 146670 }, { "grad_norm": 0.47825655341148376, "learning_rate": 1.8205942481710987e-05, "loss": 0.0216, "step": 146680 }, { "grad_norm": 0.5998581051826477, "learning_rate": 1.8199562283865607e-05, "loss": 0.0125, "step": 146690 }, { "grad_norm": 0.3379390835762024, "learning_rate": 1.819318295543224e-05, "loss": 0.0241, "step": 146700 }, { "grad_norm": 0.37303426861763, "learning_rate": 1.818680449658529e-05, "loss": 0.0235, "step": 146710 }, { "grad_norm": 0.4999482333660126, "learning_rate": 1.8180426907499154e-05, "loss": 0.0211, "step": 146720 }, { "grad_norm": 0.3831866681575775, "learning_rate": 1.8174050188348224e-05, "loss": 0.0175, "step": 146730 }, { "grad_norm": 0.39284414052963257, "learning_rate": 1.8167674339306773e-05, "loss": 0.0225, "step": 146740 }, { "grad_norm": 0.40543094277381897, "learning_rate": 1.816129936054916e-05, "loss": 0.0162, "step": 146750 }, { "grad_norm": 0.32991984486579895, "learning_rate": 1.815492525224965e-05, "loss": 0.0188, "step": 146760 }, { "grad_norm": 0.38784393668174744, "learning_rate": 1.8148552014582532e-05, "loss": 0.0167, "step": 146770 }, { "grad_norm": 0.30563923716545105, "learning_rate": 1.8142179647722042e-05, "loss": 0.0212, "step": 146780 }, { "grad_norm": 0.648912787437439, "learning_rate": 1.8135808151842364e-05, "loss": 0.0184, "step": 146790 }, { "grad_norm": 0.17896023392677307, "learning_rate": 1.812943752711775e-05, "loss": 0.013, "step": 146800 }, { "grad_norm": 0.420775443315506, "learning_rate": 1.8123067773722336e-05, "loss": 0.0216, "step": 146810 }, { "grad_norm": 0.49563977122306824, "learning_rate": 1.811669889183027e-05, "loss": 0.018, "step": 146820 }, { "grad_norm": 0.3734883964061737, "learning_rate": 1.811033088161566e-05, "loss": 0.0135, "step": 146830 }, { "grad_norm": 0.23057971894741058, "learning_rate": 1.810396374325265e-05, "loss": 0.0243, "step": 146840 }, { "grad_norm": 0.17487485706806183, "learning_rate": 1.8097597476915263e-05, "loss": 0.0236, "step": 146850 }, { "grad_norm": 0.18646948039531708, "learning_rate": 1.8091232082777597e-05, "loss": 0.0274, "step": 146860 }, { "grad_norm": 0.48977676033973694, "learning_rate": 1.8084867561013658e-05, "loss": 0.014, "step": 146870 }, { "grad_norm": 3.323747158050537, "learning_rate": 1.807850391179743e-05, "loss": 0.0263, "step": 146880 }, { "grad_norm": 0.3377310037612915, "learning_rate": 1.8072141135302927e-05, "loss": 0.0172, "step": 146890 }, { "grad_norm": 0.31930893659591675, "learning_rate": 1.80657792317041e-05, "loss": 0.0164, "step": 146900 }, { "grad_norm": 0.3533879220485687, "learning_rate": 1.805941820117485e-05, "loss": 0.0226, "step": 146910 }, { "grad_norm": 0.4774247109889984, "learning_rate": 1.805305804388912e-05, "loss": 0.0157, "step": 146920 }, { "grad_norm": 0.5347882509231567, "learning_rate": 1.804669876002077e-05, "loss": 0.0183, "step": 146930 }, { "grad_norm": 0.307693213224411, "learning_rate": 1.8040340349743702e-05, "loss": 0.0195, "step": 146940 }, { "grad_norm": 0.36062756180763245, "learning_rate": 1.8033982813231685e-05, "loss": 0.0233, "step": 146950 }, { "grad_norm": 0.27094653248786926, "learning_rate": 1.8027626150658594e-05, "loss": 0.0409, "step": 146960 }, { "grad_norm": 0.27812185883522034, "learning_rate": 1.8021270362198174e-05, "loss": 0.0126, "step": 146970 }, { "grad_norm": 0.5139604806900024, "learning_rate": 1.8014915448024224e-05, "loss": 0.0199, "step": 146980 }, { "grad_norm": 0.7465311884880066, "learning_rate": 1.8008561408310466e-05, "loss": 0.023, "step": 146990 }, { "grad_norm": 0.28732675313949585, "learning_rate": 1.8002208243230607e-05, "loss": 0.0212, "step": 147000 }, { "grad_norm": 0.35832372307777405, "learning_rate": 1.799585595295837e-05, "loss": 0.0162, "step": 147010 }, { "grad_norm": 0.27477267384529114, "learning_rate": 1.7989504537667406e-05, "loss": 0.021, "step": 147020 }, { "grad_norm": 0.3884989023208618, "learning_rate": 1.7983153997531365e-05, "loss": 0.0212, "step": 147030 }, { "grad_norm": 0.4479949474334717, "learning_rate": 1.7976804332723847e-05, "loss": 0.0223, "step": 147040 }, { "grad_norm": 0.4004298746585846, "learning_rate": 1.7970455543418467e-05, "loss": 0.0185, "step": 147050 }, { "grad_norm": 0.3382367491722107, "learning_rate": 1.796410762978882e-05, "loss": 0.0189, "step": 147060 }, { "grad_norm": 0.6336799263954163, "learning_rate": 1.7957760592008434e-05, "loss": 0.0189, "step": 147070 }, { "grad_norm": 0.4957381784915924, "learning_rate": 1.7951414430250833e-05, "loss": 0.0216, "step": 147080 }, { "grad_norm": 0.16562238335609436, "learning_rate": 1.79450691446895e-05, "loss": 0.0178, "step": 147090 }, { "grad_norm": 0.4738374352455139, "learning_rate": 1.7938724735497963e-05, "loss": 0.0213, "step": 147100 }, { "grad_norm": 0.3122958540916443, "learning_rate": 1.793238120284964e-05, "loss": 0.0205, "step": 147110 }, { "grad_norm": 0.3030683994293213, "learning_rate": 1.792603854691795e-05, "loss": 0.0135, "step": 147120 }, { "grad_norm": 0.42928844690322876, "learning_rate": 1.7919696767876336e-05, "loss": 0.0194, "step": 147130 }, { "grad_norm": 0.31161001324653625, "learning_rate": 1.791335586589814e-05, "loss": 0.0221, "step": 147140 }, { "grad_norm": 0.26902323961257935, "learning_rate": 1.7907015841156782e-05, "loss": 0.0353, "step": 147150 }, { "grad_norm": 0.25091519951820374, "learning_rate": 1.7900676693825515e-05, "loss": 0.0224, "step": 147160 }, { "grad_norm": 0.2170565128326416, "learning_rate": 1.7894338424077683e-05, "loss": 0.0167, "step": 147170 }, { "grad_norm": 0.549664318561554, "learning_rate": 1.78880010320866e-05, "loss": 0.03, "step": 147180 }, { "grad_norm": 0.5138140916824341, "learning_rate": 1.7881664518025503e-05, "loss": 0.0147, "step": 147190 }, { "grad_norm": 0.47053566575050354, "learning_rate": 1.7875328882067634e-05, "loss": 0.0136, "step": 147200 }, { "grad_norm": 0.9850460290908813, "learning_rate": 1.7868994124386183e-05, "loss": 0.0505, "step": 147210 }, { "grad_norm": 0.37780970335006714, "learning_rate": 1.7862660245154383e-05, "loss": 0.0204, "step": 147220 }, { "grad_norm": 0.3062055706977844, "learning_rate": 1.785632724454538e-05, "loss": 0.015, "step": 147230 }, { "grad_norm": 0.4695720076560974, "learning_rate": 1.7849995122732292e-05, "loss": 0.0236, "step": 147240 }, { "grad_norm": 0.39931273460388184, "learning_rate": 1.784366387988828e-05, "loss": 0.0218, "step": 147250 }, { "grad_norm": 0.40137287974357605, "learning_rate": 1.78373335161864e-05, "loss": 0.0298, "step": 147260 }, { "grad_norm": 0.2382940948009491, "learning_rate": 1.783100403179975e-05, "loss": 0.0213, "step": 147270 }, { "grad_norm": 0.578987181186676, "learning_rate": 1.782467542690137e-05, "loss": 0.017, "step": 147280 }, { "grad_norm": 0.6096801161766052, "learning_rate": 1.7818347701664262e-05, "loss": 0.0192, "step": 147290 }, { "grad_norm": 0.3975507318973541, "learning_rate": 1.7812020856261445e-05, "loss": 0.0133, "step": 147300 }, { "grad_norm": 2.2031748294830322, "learning_rate": 1.7805694890865894e-05, "loss": 0.0268, "step": 147310 }, { "grad_norm": 0.3391657769680023, "learning_rate": 1.7799369805650546e-05, "loss": 0.0152, "step": 147320 }, { "grad_norm": 0.30824604630470276, "learning_rate": 1.7793045600788317e-05, "loss": 0.0238, "step": 147330 }, { "grad_norm": 0.27345141768455505, "learning_rate": 1.7786722276452134e-05, "loss": 0.0142, "step": 147340 }, { "grad_norm": 0.4585617780685425, "learning_rate": 1.778039983281485e-05, "loss": 0.0175, "step": 147350 }, { "grad_norm": 0.24456962943077087, "learning_rate": 1.777407827004934e-05, "loss": 0.0121, "step": 147360 }, { "grad_norm": 0.45729514956474304, "learning_rate": 1.7767757588328426e-05, "loss": 0.0197, "step": 147370 }, { "grad_norm": 0.22055646777153015, "learning_rate": 1.7761437787824897e-05, "loss": 0.0192, "step": 147380 }, { "grad_norm": 0.44987720251083374, "learning_rate": 1.775511886871156e-05, "loss": 0.0166, "step": 147390 }, { "grad_norm": 0.4387975037097931, "learning_rate": 1.774880083116116e-05, "loss": 0.0202, "step": 147400 }, { "grad_norm": 0.4149017930030823, "learning_rate": 1.774248367534643e-05, "loss": 0.0221, "step": 147410 }, { "grad_norm": 0.35711297392845154, "learning_rate": 1.773616740144007e-05, "loss": 0.0202, "step": 147420 }, { "grad_norm": 0.5841598510742188, "learning_rate": 1.7729852009614762e-05, "loss": 0.018, "step": 147430 }, { "grad_norm": 0.3361721336841583, "learning_rate": 1.772353750004321e-05, "loss": 0.0184, "step": 147440 }, { "grad_norm": 0.33954495191574097, "learning_rate": 1.7717223872897986e-05, "loss": 0.0176, "step": 147450 }, { "grad_norm": 0.4047622084617615, "learning_rate": 1.771091112835175e-05, "loss": 0.0181, "step": 147460 }, { "grad_norm": 0.5512911677360535, "learning_rate": 1.7704599266577054e-05, "loss": 0.0228, "step": 147470 }, { "grad_norm": 0.26828017830848694, "learning_rate": 1.76982882877465e-05, "loss": 0.0114, "step": 147480 }, { "grad_norm": 0.3778909742832184, "learning_rate": 1.769197819203261e-05, "loss": 0.0221, "step": 147490 }, { "grad_norm": 0.301018625497818, "learning_rate": 1.768566897960788e-05, "loss": 0.023, "step": 147500 }, { "grad_norm": 0.5061416625976562, "learning_rate": 1.7679360650644838e-05, "loss": 0.0203, "step": 147510 }, { "grad_norm": 0.7941271066665649, "learning_rate": 1.767305320531593e-05, "loss": 0.0345, "step": 147520 }, { "grad_norm": 0.3578961193561554, "learning_rate": 1.766674664379361e-05, "loss": 0.0252, "step": 147530 }, { "grad_norm": 0.5450003743171692, "learning_rate": 1.766044096625027e-05, "loss": 0.026, "step": 147540 }, { "grad_norm": 0.32580843567848206, "learning_rate": 1.7654136172858323e-05, "loss": 0.0172, "step": 147550 }, { "grad_norm": 0.37536540627479553, "learning_rate": 1.7647832263790158e-05, "loss": 0.0179, "step": 147560 }, { "grad_norm": 0.27826911211013794, "learning_rate": 1.7641529239218106e-05, "loss": 0.0223, "step": 147570 }, { "grad_norm": 0.21590854227542877, "learning_rate": 1.763522709931449e-05, "loss": 0.0166, "step": 147580 }, { "grad_norm": 0.6920477151870728, "learning_rate": 1.762892584425158e-05, "loss": 0.023, "step": 147590 }, { "grad_norm": 0.25947779417037964, "learning_rate": 1.7622625474201703e-05, "loss": 0.0155, "step": 147600 }, { "grad_norm": 0.3504115045070648, "learning_rate": 1.761632598933708e-05, "loss": 0.0218, "step": 147610 }, { "grad_norm": 0.2721155285835266, "learning_rate": 1.761002738982992e-05, "loss": 0.0126, "step": 147620 }, { "grad_norm": 0.8900970220565796, "learning_rate": 1.760372967585246e-05, "loss": 0.0302, "step": 147630 }, { "grad_norm": 0.26610371470451355, "learning_rate": 1.7597432847576844e-05, "loss": 0.0201, "step": 147640 }, { "grad_norm": 0.1846819370985031, "learning_rate": 1.7591136905175278e-05, "loss": 0.0163, "step": 147650 }, { "grad_norm": 0.3802908658981323, "learning_rate": 1.758484184881981e-05, "loss": 0.0265, "step": 147660 }, { "grad_norm": 1.8293763399124146, "learning_rate": 1.7578547678682594e-05, "loss": 0.0255, "step": 147670 }, { "grad_norm": 0.5657772421836853, "learning_rate": 1.757225439493572e-05, "loss": 0.0199, "step": 147680 }, { "grad_norm": 0.5065324306488037, "learning_rate": 1.756596199775123e-05, "loss": 0.0155, "step": 147690 }, { "grad_norm": 0.6168361902236938, "learning_rate": 1.7559670487301157e-05, "loss": 0.0204, "step": 147700 }, { "grad_norm": 0.5480488538742065, "learning_rate": 1.7553379863757487e-05, "loss": 0.0214, "step": 147710 }, { "grad_norm": 0.36615437269210815, "learning_rate": 1.7547090127292233e-05, "loss": 0.0244, "step": 147720 }, { "grad_norm": 0.5010129809379578, "learning_rate": 1.7540801278077345e-05, "loss": 0.0176, "step": 147730 }, { "grad_norm": 0.33350613713264465, "learning_rate": 1.753451331628474e-05, "loss": 0.0222, "step": 147740 }, { "grad_norm": 0.4967767596244812, "learning_rate": 1.7528226242086364e-05, "loss": 0.0194, "step": 147750 }, { "grad_norm": 0.37993553280830383, "learning_rate": 1.752194005565406e-05, "loss": 0.0145, "step": 147760 }, { "grad_norm": 0.1889292597770691, "learning_rate": 1.751565475715974e-05, "loss": 0.0142, "step": 147770 }, { "grad_norm": 0.36105623841285706, "learning_rate": 1.7509370346775207e-05, "loss": 0.0152, "step": 147780 }, { "grad_norm": 0.3877922594547272, "learning_rate": 1.7503086824672287e-05, "loss": 0.0154, "step": 147790 }, { "grad_norm": 0.1362587958574295, "learning_rate": 1.7496804191022747e-05, "loss": 0.016, "step": 147800 }, { "grad_norm": 0.3463985025882721, "learning_rate": 1.7490522445998363e-05, "loss": 0.0193, "step": 147810 }, { "grad_norm": 0.326826274394989, "learning_rate": 1.7484241589770928e-05, "loss": 0.0158, "step": 147820 }, { "grad_norm": 0.356085866689682, "learning_rate": 1.7477961622512066e-05, "loss": 0.0283, "step": 147830 }, { "grad_norm": 0.28567126393318176, "learning_rate": 1.7471682544393542e-05, "loss": 0.0327, "step": 147840 }, { "grad_norm": 0.5241082906723022, "learning_rate": 1.7465404355586974e-05, "loss": 0.0247, "step": 147850 }, { "grad_norm": 0.6736773252487183, "learning_rate": 1.7459127056264046e-05, "loss": 0.0215, "step": 147860 }, { "grad_norm": 0.7185114026069641, "learning_rate": 1.7452850646596363e-05, "loss": 0.0201, "step": 147870 }, { "grad_norm": 0.5174068212509155, "learning_rate": 1.7446575126755497e-05, "loss": 0.0249, "step": 147880 }, { "grad_norm": 0.20153440535068512, "learning_rate": 1.7440300496913054e-05, "loss": 0.018, "step": 147890 }, { "grad_norm": 0.3273630142211914, "learning_rate": 1.7434026757240564e-05, "loss": 0.0115, "step": 147900 }, { "grad_norm": 0.6340059041976929, "learning_rate": 1.7427753907909545e-05, "loss": 0.0244, "step": 147910 }, { "grad_norm": 0.4950276017189026, "learning_rate": 1.7421481949091483e-05, "loss": 0.0165, "step": 147920 }, { "grad_norm": 0.9913126230239868, "learning_rate": 1.7415210880957867e-05, "loss": 0.0223, "step": 147930 }, { "grad_norm": 0.5445325374603271, "learning_rate": 1.7408940703680154e-05, "loss": 0.017, "step": 147940 }, { "grad_norm": 0.2849355936050415, "learning_rate": 1.7402671417429765e-05, "loss": 0.0142, "step": 147950 }, { "grad_norm": 0.537901759147644, "learning_rate": 1.7396403022378095e-05, "loss": 0.0201, "step": 147960 }, { "grad_norm": 0.1940840482711792, "learning_rate": 1.7390135518696492e-05, "loss": 0.0308, "step": 147970 }, { "grad_norm": 0.21791045367717743, "learning_rate": 1.7383868906556355e-05, "loss": 0.0157, "step": 147980 }, { "grad_norm": 0.4133894741535187, "learning_rate": 1.7377603186128988e-05, "loss": 0.0138, "step": 147990 }, { "grad_norm": 0.13372646272182465, "learning_rate": 1.737133835758568e-05, "loss": 0.0359, "step": 148000 }, { "grad_norm": 0.3806176483631134, "learning_rate": 1.7365074421097737e-05, "loss": 0.0255, "step": 148010 }, { "grad_norm": 0.7413172125816345, "learning_rate": 1.7358811376836387e-05, "loss": 0.0224, "step": 148020 }, { "grad_norm": 0.7384961247444153, "learning_rate": 1.7352549224972897e-05, "loss": 0.0238, "step": 148030 }, { "grad_norm": 0.22846795618534088, "learning_rate": 1.734628796567842e-05, "loss": 0.0206, "step": 148040 }, { "grad_norm": 0.16863667964935303, "learning_rate": 1.7340027599124175e-05, "loss": 0.0248, "step": 148050 }, { "grad_norm": 0.3811183571815491, "learning_rate": 1.7333768125481293e-05, "loss": 0.0146, "step": 148060 }, { "grad_norm": 0.38555747270584106, "learning_rate": 1.7327509544920928e-05, "loss": 0.0242, "step": 148070 }, { "grad_norm": 0.5596863031387329, "learning_rate": 1.7321251857614183e-05, "loss": 0.0398, "step": 148080 }, { "grad_norm": 0.1470949947834015, "learning_rate": 1.7314995063732113e-05, "loss": 0.0214, "step": 148090 }, { "grad_norm": 0.5200960636138916, "learning_rate": 1.7308739163445825e-05, "loss": 0.0218, "step": 148100 }, { "grad_norm": 0.4613744020462036, "learning_rate": 1.730248415692632e-05, "loss": 0.0373, "step": 148110 }, { "grad_norm": 0.3616591989994049, "learning_rate": 1.7296230044344598e-05, "loss": 0.0187, "step": 148120 }, { "grad_norm": 0.4407232701778412, "learning_rate": 1.728997682587168e-05, "loss": 0.0165, "step": 148130 }, { "grad_norm": 0.17031268775463104, "learning_rate": 1.728372450167849e-05, "loss": 0.0194, "step": 148140 }, { "grad_norm": 0.13346216082572937, "learning_rate": 1.7277473071935996e-05, "loss": 0.0194, "step": 148150 }, { "grad_norm": 0.31381845474243164, "learning_rate": 1.727122253681509e-05, "loss": 0.0138, "step": 148160 }, { "grad_norm": 0.2812192440032959, "learning_rate": 1.726497289648667e-05, "loss": 0.0194, "step": 148170 }, { "grad_norm": 0.4137652814388275, "learning_rate": 1.725872415112158e-05, "loss": 0.0285, "step": 148180 }, { "grad_norm": 0.4122503399848938, "learning_rate": 1.725247630089068e-05, "loss": 0.0246, "step": 148190 }, { "grad_norm": 0.6511193513870239, "learning_rate": 1.724622934596478e-05, "loss": 0.0241, "step": 148200 }, { "grad_norm": 0.32367101311683655, "learning_rate": 1.723998328651465e-05, "loss": 0.0176, "step": 148210 }, { "grad_norm": 0.5226243138313293, "learning_rate": 1.723373812271108e-05, "loss": 0.0319, "step": 148220 }, { "grad_norm": 0.22039854526519775, "learning_rate": 1.7227493854724784e-05, "loss": 0.0285, "step": 148230 }, { "grad_norm": 0.6065847277641296, "learning_rate": 1.722125048272653e-05, "loss": 0.0242, "step": 148240 }, { "grad_norm": 0.7902690768241882, "learning_rate": 1.7215008006886946e-05, "loss": 0.0185, "step": 148250 }, { "grad_norm": 0.262703537940979, "learning_rate": 1.7208766427376715e-05, "loss": 0.016, "step": 148260 }, { "grad_norm": 0.42582741379737854, "learning_rate": 1.7202525744366515e-05, "loss": 0.0478, "step": 148270 }, { "grad_norm": 0.19258177280426025, "learning_rate": 1.7196285958026937e-05, "loss": 0.0161, "step": 148280 }, { "grad_norm": 0.4980139434337616, "learning_rate": 1.7190047068528582e-05, "loss": 0.0213, "step": 148290 }, { "grad_norm": 0.459272563457489, "learning_rate": 1.718380907604199e-05, "loss": 0.0219, "step": 148300 }, { "grad_norm": 0.9679397344589233, "learning_rate": 1.7177571980737732e-05, "loss": 0.0303, "step": 148310 }, { "grad_norm": 0.3170870542526245, "learning_rate": 1.7171335782786357e-05, "loss": 0.0196, "step": 148320 }, { "grad_norm": 0.3292924761772156, "learning_rate": 1.71651004823583e-05, "loss": 0.0236, "step": 148330 }, { "grad_norm": 0.3144012987613678, "learning_rate": 1.7158866079624075e-05, "loss": 0.0169, "step": 148340 }, { "grad_norm": 0.5830413103103638, "learning_rate": 1.7152632574754094e-05, "loss": 0.0172, "step": 148350 }, { "grad_norm": 0.19765400886535645, "learning_rate": 1.714639996791882e-05, "loss": 0.0228, "step": 148360 }, { "grad_norm": 0.42216309905052185, "learning_rate": 1.7140168259288625e-05, "loss": 0.0202, "step": 148370 }, { "grad_norm": 0.2266615331172943, "learning_rate": 1.7133937449033866e-05, "loss": 0.0153, "step": 148380 }, { "grad_norm": 0.2886061668395996, "learning_rate": 1.7127707537324926e-05, "loss": 0.0161, "step": 148390 }, { "grad_norm": 0.3002031743526459, "learning_rate": 1.712147852433212e-05, "loss": 0.0184, "step": 148400 }, { "grad_norm": 0.1889614760875702, "learning_rate": 1.711525041022573e-05, "loss": 0.0098, "step": 148410 }, { "grad_norm": 0.3136744499206543, "learning_rate": 1.7109023195176033e-05, "loss": 0.0389, "step": 148420 }, { "grad_norm": 0.26591989398002625, "learning_rate": 1.7102796879353296e-05, "loss": 0.026, "step": 148430 }, { "grad_norm": 0.970903217792511, "learning_rate": 1.7096571462927713e-05, "loss": 0.0165, "step": 148440 }, { "grad_norm": 0.4215907156467438, "learning_rate": 1.709034694606953e-05, "loss": 0.0187, "step": 148450 }, { "grad_norm": 0.3223123252391815, "learning_rate": 1.708412332894889e-05, "loss": 0.0234, "step": 148460 }, { "grad_norm": 0.30688154697418213, "learning_rate": 1.707790061173594e-05, "loss": 0.0234, "step": 148470 }, { "grad_norm": 0.26261106133461, "learning_rate": 1.7071678794600836e-05, "loss": 0.0123, "step": 148480 }, { "grad_norm": 0.32234013080596924, "learning_rate": 1.7065457877713663e-05, "loss": 0.024, "step": 148490 }, { "grad_norm": 0.30663254857063293, "learning_rate": 1.7059237861244482e-05, "loss": 0.0206, "step": 148500 }, { "grad_norm": 0.4428611993789673, "learning_rate": 1.705301874536338e-05, "loss": 0.0239, "step": 148510 }, { "grad_norm": 0.3143925964832306, "learning_rate": 1.704680053024035e-05, "loss": 0.0182, "step": 148520 }, { "grad_norm": 0.2857389450073242, "learning_rate": 1.7040583216045448e-05, "loss": 0.0183, "step": 148530 }, { "grad_norm": 0.41207414865493774, "learning_rate": 1.703436680294858e-05, "loss": 0.0209, "step": 148540 }, { "grad_norm": 0.9885429739952087, "learning_rate": 1.702815129111976e-05, "loss": 0.022, "step": 148550 }, { "grad_norm": 0.4668305218219757, "learning_rate": 1.702193668072888e-05, "loss": 0.0147, "step": 148560 }, { "grad_norm": 0.2997579276561737, "learning_rate": 1.701572297194588e-05, "loss": 0.0198, "step": 148570 }, { "grad_norm": 0.5826438069343567, "learning_rate": 1.7009510164940616e-05, "loss": 0.0202, "step": 148580 }, { "grad_norm": 0.24490423500537872, "learning_rate": 1.7003298259882944e-05, "loss": 0.0134, "step": 148590 }, { "grad_norm": 0.3031160235404968, "learning_rate": 1.6997087256942706e-05, "loss": 0.0146, "step": 148600 }, { "grad_norm": 0.43872353434562683, "learning_rate": 1.6990877156289712e-05, "loss": 0.0203, "step": 148610 }, { "grad_norm": 0.7244512438774109, "learning_rate": 1.6984667958093735e-05, "loss": 0.0276, "step": 148620 }, { "grad_norm": 0.5307899117469788, "learning_rate": 1.6978459662524514e-05, "loss": 0.0307, "step": 148630 }, { "grad_norm": 0.7541978359222412, "learning_rate": 1.69722522697518e-05, "loss": 0.0188, "step": 148640 }, { "grad_norm": 0.48822930455207825, "learning_rate": 1.6966045779945315e-05, "loss": 0.0198, "step": 148650 }, { "grad_norm": 0.3104594349861145, "learning_rate": 1.6959840193274734e-05, "loss": 0.0168, "step": 148660 }, { "grad_norm": 0.3543863594532013, "learning_rate": 1.695363550990971e-05, "loss": 0.0257, "step": 148670 }, { "grad_norm": 0.8012112975120544, "learning_rate": 1.6947431730019853e-05, "loss": 0.0385, "step": 148680 }, { "grad_norm": 0.193651482462883, "learning_rate": 1.6941228853774814e-05, "loss": 0.0204, "step": 148690 }, { "grad_norm": 0.5782427787780762, "learning_rate": 1.6935026881344162e-05, "loss": 0.0294, "step": 148700 }, { "grad_norm": 0.33707475662231445, "learning_rate": 1.6928825812897436e-05, "loss": 0.0207, "step": 148710 }, { "grad_norm": 0.5046456456184387, "learning_rate": 1.69226256486042e-05, "loss": 0.0239, "step": 148720 }, { "grad_norm": 0.29032689332962036, "learning_rate": 1.6916426388633943e-05, "loss": 0.0263, "step": 148730 }, { "grad_norm": 0.24400366842746735, "learning_rate": 1.691022803315619e-05, "loss": 0.021, "step": 148740 }, { "grad_norm": 0.4506492018699646, "learning_rate": 1.6904030582340337e-05, "loss": 0.0325, "step": 148750 }, { "grad_norm": 1.648492693901062, "learning_rate": 1.6897834036355846e-05, "loss": 0.0212, "step": 148760 }, { "grad_norm": 0.3460257053375244, "learning_rate": 1.6891638395372167e-05, "loss": 0.0192, "step": 148770 }, { "grad_norm": 0.5372751355171204, "learning_rate": 1.6885443659558647e-05, "loss": 0.0169, "step": 148780 }, { "grad_norm": 0.16990794241428375, "learning_rate": 1.6879249829084658e-05, "loss": 0.015, "step": 148790 }, { "grad_norm": 0.1803494542837143, "learning_rate": 1.687305690411952e-05, "loss": 0.0235, "step": 148800 }, { "grad_norm": 0.41898125410079956, "learning_rate": 1.6866864884832583e-05, "loss": 0.0239, "step": 148810 }, { "grad_norm": 0.39677390456199646, "learning_rate": 1.686067377139311e-05, "loss": 0.0159, "step": 148820 }, { "grad_norm": 0.23478847742080688, "learning_rate": 1.6854483563970348e-05, "loss": 0.0203, "step": 148830 }, { "grad_norm": 0.2667742669582367, "learning_rate": 1.6848294262733567e-05, "loss": 0.0236, "step": 148840 }, { "grad_norm": 0.3445814549922943, "learning_rate": 1.6842105867851953e-05, "loss": 0.0201, "step": 148850 }, { "grad_norm": 0.3115549385547638, "learning_rate": 1.6835918379494726e-05, "loss": 0.0202, "step": 148860 }, { "grad_norm": 0.31212377548217773, "learning_rate": 1.6829731797831027e-05, "loss": 0.017, "step": 148870 }, { "grad_norm": 0.2541424334049225, "learning_rate": 1.6823546123029982e-05, "loss": 0.0243, "step": 148880 }, { "grad_norm": 0.47537797689437866, "learning_rate": 1.6817361355260742e-05, "loss": 0.0192, "step": 148890 }, { "grad_norm": 0.35042503476142883, "learning_rate": 1.6811177494692377e-05, "loss": 0.0242, "step": 148900 }, { "grad_norm": 0.14484190940856934, "learning_rate": 1.6804994541493953e-05, "loss": 0.0176, "step": 148910 }, { "grad_norm": 0.12492643296718597, "learning_rate": 1.6798812495834482e-05, "loss": 0.0135, "step": 148920 }, { "grad_norm": 0.3441249430179596, "learning_rate": 1.6792631357883025e-05, "loss": 0.0189, "step": 148930 }, { "grad_norm": 0.3475227653980255, "learning_rate": 1.678645112780854e-05, "loss": 0.0146, "step": 148940 }, { "grad_norm": 0.3904566466808319, "learning_rate": 1.6780271805780013e-05, "loss": 0.0222, "step": 148950 }, { "grad_norm": 0.20114055275917053, "learning_rate": 1.6774093391966377e-05, "loss": 0.0216, "step": 148960 }, { "grad_norm": 0.4689573347568512, "learning_rate": 1.6767915886536535e-05, "loss": 0.0188, "step": 148970 }, { "grad_norm": 0.516485333442688, "learning_rate": 1.6761739289659395e-05, "loss": 0.0168, "step": 148980 }, { "grad_norm": 0.401703804731369, "learning_rate": 1.675556360150382e-05, "loss": 0.0181, "step": 148990 }, { "grad_norm": 1.1538147926330566, "learning_rate": 1.674938882223865e-05, "loss": 0.0314, "step": 149000 }, { "grad_norm": 0.28039857745170593, "learning_rate": 1.6743214952032677e-05, "loss": 0.0181, "step": 149010 }, { "grad_norm": 0.9044480919837952, "learning_rate": 1.6737041991054713e-05, "loss": 0.0269, "step": 149020 }, { "grad_norm": 0.2287079393863678, "learning_rate": 1.673086993947356e-05, "loss": 0.0144, "step": 149030 }, { "grad_norm": 0.3083380460739136, "learning_rate": 1.6724698797457888e-05, "loss": 0.0161, "step": 149040 }, { "grad_norm": 0.4143659770488739, "learning_rate": 1.6718528565176465e-05, "loss": 0.0144, "step": 149050 }, { "grad_norm": 0.35399290919303894, "learning_rate": 1.671235924279796e-05, "loss": 0.02, "step": 149060 }, { "grad_norm": 0.7235307693481445, "learning_rate": 1.6706190830491053e-05, "loss": 0.0166, "step": 149070 }, { "grad_norm": 0.3873840868473053, "learning_rate": 1.670002332842439e-05, "loss": 0.019, "step": 149080 }, { "grad_norm": 0.48564019799232483, "learning_rate": 1.669385673676656e-05, "loss": 0.0153, "step": 149090 }, { "grad_norm": 0.24577872455120087, "learning_rate": 1.6687691055686187e-05, "loss": 0.0179, "step": 149100 }, { "grad_norm": 0.3968072235584259, "learning_rate": 1.6681526285351828e-05, "loss": 0.017, "step": 149110 }, { "grad_norm": 0.29114869236946106, "learning_rate": 1.6675362425932028e-05, "loss": 0.0146, "step": 149120 }, { "grad_norm": 0.3057517111301422, "learning_rate": 1.6669199477595275e-05, "loss": 0.0208, "step": 149130 }, { "grad_norm": 0.7616528868675232, "learning_rate": 1.6663037440510093e-05, "loss": 0.0227, "step": 149140 }, { "grad_norm": 0.1758759766817093, "learning_rate": 1.6656876314844956e-05, "loss": 0.0221, "step": 149150 }, { "grad_norm": 0.4572036862373352, "learning_rate": 1.6650716100768298e-05, "loss": 0.0168, "step": 149160 }, { "grad_norm": 0.3828851580619812, "learning_rate": 1.664455679844853e-05, "loss": 0.0181, "step": 149170 }, { "grad_norm": 0.6343505382537842, "learning_rate": 1.663839840805403e-05, "loss": 0.0192, "step": 149180 }, { "grad_norm": 0.49820518493652344, "learning_rate": 1.66322409297532e-05, "loss": 0.0191, "step": 149190 }, { "grad_norm": 0.36761897802352905, "learning_rate": 1.662608436371437e-05, "loss": 0.017, "step": 149200 }, { "grad_norm": 0.38106462359428406, "learning_rate": 1.6619928710105843e-05, "loss": 0.0243, "step": 149210 }, { "grad_norm": 0.4365946352481842, "learning_rate": 1.6613773969095937e-05, "loss": 0.0194, "step": 149220 }, { "grad_norm": 0.343993216753006, "learning_rate": 1.6607620140852893e-05, "loss": 0.0133, "step": 149230 }, { "grad_norm": 0.4319615364074707, "learning_rate": 1.6601467225545e-05, "loss": 0.0272, "step": 149240 }, { "grad_norm": 0.326152503490448, "learning_rate": 1.6595315223340414e-05, "loss": 0.0153, "step": 149250 }, { "grad_norm": 0.39528268575668335, "learning_rate": 1.6589164134407376e-05, "loss": 0.0135, "step": 149260 }, { "grad_norm": 0.760906457901001, "learning_rate": 1.6583013958914022e-05, "loss": 0.02, "step": 149270 }, { "grad_norm": 0.5500805377960205, "learning_rate": 1.6576864697028534e-05, "loss": 0.0263, "step": 149280 }, { "grad_norm": 0.1931149810552597, "learning_rate": 1.6570716348919e-05, "loss": 0.014, "step": 149290 }, { "grad_norm": 0.3055424988269806, "learning_rate": 1.6564568914753516e-05, "loss": 0.0342, "step": 149300 }, { "grad_norm": 0.22980383038520813, "learning_rate": 1.6558422394700163e-05, "loss": 0.0196, "step": 149310 }, { "grad_norm": 0.3804989755153656, "learning_rate": 1.6552276788926984e-05, "loss": 0.0233, "step": 149320 }, { "grad_norm": 0.3471719026565552, "learning_rate": 1.6546132097601975e-05, "loss": 0.0153, "step": 149330 }, { "grad_norm": 0.21395398676395416, "learning_rate": 1.6539988320893158e-05, "loss": 0.0393, "step": 149340 }, { "grad_norm": 0.364499568939209, "learning_rate": 1.653384545896848e-05, "loss": 0.0392, "step": 149350 }, { "grad_norm": 0.43272215127944946, "learning_rate": 1.6527703511995908e-05, "loss": 0.0186, "step": 149360 }, { "grad_norm": 0.7821706533432007, "learning_rate": 1.6521562480143347e-05, "loss": 0.0138, "step": 149370 }, { "grad_norm": 0.17946630716323853, "learning_rate": 1.6515422363578692e-05, "loss": 0.0285, "step": 149380 }, { "grad_norm": 0.48557302355766296, "learning_rate": 1.650928316246979e-05, "loss": 0.0134, "step": 149390 }, { "grad_norm": 0.4825190007686615, "learning_rate": 1.6503144876984522e-05, "loss": 0.0358, "step": 149400 }, { "grad_norm": 0.974905788898468, "learning_rate": 1.649700750729069e-05, "loss": 0.0238, "step": 149410 }, { "grad_norm": 0.329833060503006, "learning_rate": 1.6490871053556073e-05, "loss": 0.0174, "step": 149420 }, { "grad_norm": 0.2896270155906677, "learning_rate": 1.6484735515948467e-05, "loss": 0.019, "step": 149430 }, { "grad_norm": 0.22377698123455048, "learning_rate": 1.647860089463558e-05, "loss": 0.0153, "step": 149440 }, { "grad_norm": 0.6214963793754578, "learning_rate": 1.647246718978519e-05, "loss": 0.0222, "step": 149450 }, { "grad_norm": 0.5925918817520142, "learning_rate": 1.6466334401564915e-05, "loss": 0.0227, "step": 149460 }, { "grad_norm": 0.44731882214546204, "learning_rate": 1.646020253014246e-05, "loss": 0.0174, "step": 149470 }, { "grad_norm": 1.1156185865402222, "learning_rate": 1.6454071575685488e-05, "loss": 0.0164, "step": 149480 }, { "grad_norm": 0.32532474398612976, "learning_rate": 1.6447941538361593e-05, "loss": 0.014, "step": 149490 }, { "grad_norm": 0.1648167371749878, "learning_rate": 1.644181241833837e-05, "loss": 0.0246, "step": 149500 }, { "grad_norm": 0.5680223703384399, "learning_rate": 1.6435684215783377e-05, "loss": 0.0292, "step": 149510 }, { "grad_norm": 0.4691929817199707, "learning_rate": 1.642955693086416e-05, "loss": 0.0176, "step": 149520 }, { "grad_norm": 0.2726694345474243, "learning_rate": 1.6423430563748282e-05, "loss": 0.0186, "step": 149530 }, { "grad_norm": 0.2634336054325104, "learning_rate": 1.6417305114603165e-05, "loss": 0.0131, "step": 149540 }, { "grad_norm": 0.4073016941547394, "learning_rate": 1.6411180583596324e-05, "loss": 0.0146, "step": 149550 }, { "grad_norm": 0.45379456877708435, "learning_rate": 1.6405056970895178e-05, "loss": 0.0198, "step": 149560 }, { "grad_norm": 0.3814542889595032, "learning_rate": 1.639893427666716e-05, "loss": 0.0354, "step": 149570 }, { "grad_norm": 0.5109282732009888, "learning_rate": 1.6392812501079663e-05, "loss": 0.0209, "step": 149580 }, { "grad_norm": 0.5679459571838379, "learning_rate": 1.638669164430003e-05, "loss": 0.0235, "step": 149590 }, { "grad_norm": 0.24955447018146515, "learning_rate": 1.638057170649563e-05, "loss": 0.0141, "step": 149600 }, { "grad_norm": 0.4577651917934418, "learning_rate": 1.6374452687833775e-05, "loss": 0.0296, "step": 149610 }, { "grad_norm": 0.6185445785522461, "learning_rate": 1.6368334588481748e-05, "loss": 0.0261, "step": 149620 }, { "grad_norm": 0.3239096999168396, "learning_rate": 1.63622174086068e-05, "loss": 0.0203, "step": 149630 }, { "grad_norm": 0.5273399949073792, "learning_rate": 1.6356101148376212e-05, "loss": 0.0199, "step": 149640 }, { "grad_norm": 0.3168490529060364, "learning_rate": 1.6349985807957163e-05, "loss": 0.015, "step": 149650 }, { "grad_norm": 0.48796114325523376, "learning_rate": 1.6343871387516874e-05, "loss": 0.0206, "step": 149660 }, { "grad_norm": 0.28402191400527954, "learning_rate": 1.63377578872225e-05, "loss": 0.0116, "step": 149670 }, { "grad_norm": 0.3736550807952881, "learning_rate": 1.6331645307241155e-05, "loss": 0.0194, "step": 149680 }, { "grad_norm": 0.2918042242527008, "learning_rate": 1.6325533647740005e-05, "loss": 0.0187, "step": 149690 }, { "grad_norm": 0.4849667251110077, "learning_rate": 1.6319422908886107e-05, "loss": 0.0195, "step": 149700 }, { "grad_norm": 0.4461943805217743, "learning_rate": 1.631331309084652e-05, "loss": 0.0186, "step": 149710 }, { "grad_norm": 0.4250064790248871, "learning_rate": 1.6307204193788317e-05, "loss": 0.0185, "step": 149720 }, { "grad_norm": 0.22609633207321167, "learning_rate": 1.630109621787847e-05, "loss": 0.0132, "step": 149730 }, { "grad_norm": 0.20615805685520172, "learning_rate": 1.629498916328403e-05, "loss": 0.0165, "step": 149740 }, { "grad_norm": 0.31757983565330505, "learning_rate": 1.6288883030171892e-05, "loss": 0.0185, "step": 149750 }, { "grad_norm": 0.572482168674469, "learning_rate": 1.628277781870904e-05, "loss": 0.0257, "step": 149760 }, { "grad_norm": 0.23367683589458466, "learning_rate": 1.6276673529062365e-05, "loss": 0.0167, "step": 149770 }, { "grad_norm": 1.4325065612792969, "learning_rate": 1.6270570161398775e-05, "loss": 0.0169, "step": 149780 }, { "grad_norm": 0.1977093517780304, "learning_rate": 1.6264467715885128e-05, "loss": 0.0131, "step": 149790 }, { "grad_norm": 0.3023398220539093, "learning_rate": 1.6258366192688246e-05, "loss": 0.02, "step": 149800 }, { "grad_norm": 0.39465850591659546, "learning_rate": 1.625226559197497e-05, "loss": 0.0164, "step": 149810 }, { "grad_norm": 0.6057740449905396, "learning_rate": 1.6246165913912075e-05, "loss": 0.0201, "step": 149820 }, { "grad_norm": 0.26067984104156494, "learning_rate": 1.6240067158666324e-05, "loss": 0.0192, "step": 149830 }, { "grad_norm": 0.3877856731414795, "learning_rate": 1.6233969326404437e-05, "loss": 0.0129, "step": 149840 }, { "grad_norm": 0.6199467182159424, "learning_rate": 1.6227872417293144e-05, "loss": 0.0187, "step": 149850 }, { "grad_norm": 0.7037124633789062, "learning_rate": 1.6221776431499148e-05, "loss": 0.0174, "step": 149860 }, { "grad_norm": 0.6929872035980225, "learning_rate": 1.6215681369189094e-05, "loss": 0.022, "step": 149870 }, { "grad_norm": 0.39508798718452454, "learning_rate": 1.6209587230529614e-05, "loss": 0.0246, "step": 149880 }, { "grad_norm": 0.33964577317237854, "learning_rate": 1.6203494015687316e-05, "loss": 0.0215, "step": 149890 }, { "grad_norm": 0.5886121988296509, "learning_rate": 1.6197401724828808e-05, "loss": 0.0185, "step": 149900 }, { "grad_norm": 0.4165380597114563, "learning_rate": 1.6191310358120637e-05, "loss": 0.0247, "step": 149910 }, { "grad_norm": 0.447143018245697, "learning_rate": 1.6185219915729326e-05, "loss": 0.0213, "step": 149920 }, { "grad_norm": 0.2787947654724121, "learning_rate": 1.6179130397821412e-05, "loss": 0.0153, "step": 149930 }, { "grad_norm": 0.23500216007232666, "learning_rate": 1.6173041804563354e-05, "loss": 0.0181, "step": 149940 }, { "grad_norm": 0.6271631121635437, "learning_rate": 1.6166954136121654e-05, "loss": 0.0184, "step": 149950 }, { "grad_norm": 0.8596580624580383, "learning_rate": 1.6160867392662688e-05, "loss": 0.0189, "step": 149960 }, { "grad_norm": 0.35023239254951477, "learning_rate": 1.6154781574352894e-05, "loss": 0.0165, "step": 149970 }, { "grad_norm": 0.497040718793869, "learning_rate": 1.6148696681358672e-05, "loss": 0.0215, "step": 149980 }, { "grad_norm": 0.8202377557754517, "learning_rate": 1.6142612713846367e-05, "loss": 0.0175, "step": 149990 }, { "grad_norm": 0.856261670589447, "learning_rate": 1.613652967198231e-05, "loss": 0.0214, "step": 150000 } ], "logging_steps": 10, "max_steps": 200000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 50000, "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 }