{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.3248094320297241, "learning_rate": 2.25e-07, "loss": 1.2955, "step": 10 }, { "grad_norm": 1.3657076358795166, "learning_rate": 4.75e-07, "loss": 1.2928, "step": 20 }, { "grad_norm": 1.274039387702942, "learning_rate": 7.25e-07, "loss": 1.2887, "step": 30 }, { "grad_norm": 1.1349459886550903, "learning_rate": 9.75e-07, "loss": 1.2902, "step": 40 }, { "grad_norm": 1.0769822597503662, "learning_rate": 1.2250000000000001e-06, "loss": 1.2711, "step": 50 }, { "grad_norm": 0.6751026511192322, "learning_rate": 1.475e-06, "loss": 1.2525, "step": 60 }, { "grad_norm": 0.48190054297447205, "learning_rate": 1.7250000000000002e-06, "loss": 1.2296, "step": 70 }, { "grad_norm": 0.30119672417640686, "learning_rate": 1.975e-06, "loss": 1.2196, "step": 80 }, { "grad_norm": 0.39081430435180664, "learning_rate": 2.225e-06, "loss": 1.2085, "step": 90 }, { "grad_norm": 0.5773189067840576, "learning_rate": 2.4750000000000004e-06, "loss": 1.2259, "step": 100 }, { "grad_norm": 0.35268476605415344, "learning_rate": 2.725e-06, "loss": 1.2057, "step": 110 }, { "grad_norm": 0.6626057028770447, "learning_rate": 2.975e-06, "loss": 1.2087, "step": 120 }, { "grad_norm": 0.5878528356552124, "learning_rate": 3.225e-06, "loss": 1.1747, "step": 130 }, { "grad_norm": 0.5314624309539795, "learning_rate": 3.4750000000000006e-06, "loss": 1.1396, "step": 140 }, { "grad_norm": 0.45975780487060547, "learning_rate": 3.725e-06, "loss": 1.1075, "step": 150 }, { "grad_norm": 0.39815983176231384, "learning_rate": 3.975e-06, "loss": 1.0944, "step": 160 }, { "grad_norm": 0.4668089747428894, "learning_rate": 4.225e-06, "loss": 1.0885, "step": 170 }, { "grad_norm": 0.5085218548774719, "learning_rate": 4.475e-06, "loss": 1.0879, "step": 180 }, { "grad_norm": 0.49683114886283875, "learning_rate": 4.7250000000000005e-06, "loss": 1.0784, "step": 190 }, { "grad_norm": 0.4529138505458832, "learning_rate": 4.975000000000001e-06, "loss": 1.0786, "step": 200 }, { "grad_norm": 0.5527579188346863, "learning_rate": 5.225e-06, "loss": 1.0725, "step": 210 }, { "grad_norm": 0.4353656470775604, "learning_rate": 5.475e-06, "loss": 1.0648, "step": 220 }, { "grad_norm": 0.6097522974014282, "learning_rate": 5.725e-06, "loss": 1.0569, "step": 230 }, { "grad_norm": 0.6625789403915405, "learning_rate": 5.975e-06, "loss": 1.0514, "step": 240 }, { "grad_norm": 0.6466538310050964, "learning_rate": 6.2250000000000005e-06, "loss": 1.0476, "step": 250 }, { "grad_norm": 0.5459293723106384, "learning_rate": 6.475000000000001e-06, "loss": 1.0516, "step": 260 }, { "grad_norm": 0.48585712909698486, "learning_rate": 6.725000000000001e-06, "loss": 1.0454, "step": 270 }, { "grad_norm": 0.49608978629112244, "learning_rate": 6.975000000000001e-06, "loss": 1.0439, "step": 280 }, { "grad_norm": 0.5594484806060791, "learning_rate": 7.2249999999999994e-06, "loss": 1.0366, "step": 290 }, { "grad_norm": 0.5457912087440491, "learning_rate": 7.4750000000000004e-06, "loss": 1.0383, "step": 300 }, { "grad_norm": 0.5215047597885132, "learning_rate": 7.725e-06, "loss": 1.0343, "step": 310 }, { "grad_norm": 0.624180793762207, "learning_rate": 7.975e-06, "loss": 1.0409, "step": 320 }, { "grad_norm": 0.47685298323631287, "learning_rate": 8.225e-06, "loss": 1.0332, "step": 330 }, { "grad_norm": 0.5333760976791382, "learning_rate": 8.475000000000001e-06, "loss": 1.0266, "step": 340 }, { "grad_norm": 0.44368305802345276, "learning_rate": 8.725e-06, "loss": 1.0349, "step": 350 }, { "grad_norm": 0.5649255514144897, "learning_rate": 8.975e-06, "loss": 1.0235, "step": 360 }, { "grad_norm": 0.8148795366287231, "learning_rate": 9.225e-06, "loss": 1.0108, "step": 370 }, { "grad_norm": 0.8726373314857483, "learning_rate": 9.475e-06, "loss": 0.9987, "step": 380 }, { "grad_norm": 0.9848608374595642, "learning_rate": 9.725000000000001e-06, "loss": 0.9884, "step": 390 }, { "grad_norm": 0.9186549186706543, "learning_rate": 9.975e-06, "loss": 0.9617, "step": 400 }, { "grad_norm": 0.8507822751998901, "learning_rate": 1.0225e-05, "loss": 0.9401, "step": 410 }, { "grad_norm": 1.2925907373428345, "learning_rate": 1.0475e-05, "loss": 0.9211, "step": 420 }, { "grad_norm": 1.152266025543213, "learning_rate": 1.0725e-05, "loss": 0.9049, "step": 430 }, { "grad_norm": 1.470323085784912, "learning_rate": 1.0975e-05, "loss": 0.8713, "step": 440 }, { "grad_norm": 1.2892720699310303, "learning_rate": 1.1225e-05, "loss": 0.8466, "step": 450 }, { "grad_norm": 1.4253071546554565, "learning_rate": 1.1475000000000001e-05, "loss": 0.8097, "step": 460 }, { "grad_norm": 1.3366661071777344, "learning_rate": 1.1725e-05, "loss": 0.7831, "step": 470 }, { "grad_norm": 1.697050929069519, "learning_rate": 1.1975e-05, "loss": 0.75, "step": 480 }, { "grad_norm": 1.5356013774871826, "learning_rate": 1.2225e-05, "loss": 0.708, "step": 490 }, { "grad_norm": 1.9125316143035889, "learning_rate": 1.2475e-05, "loss": 0.6837, "step": 500 }, { "grad_norm": 2.116225242614746, "learning_rate": 1.2725000000000001e-05, "loss": 0.6443, "step": 510 }, { "grad_norm": 1.851104974746704, "learning_rate": 1.2975e-05, "loss": 0.6182, "step": 520 }, { "grad_norm": 1.905001163482666, "learning_rate": 1.3225000000000001e-05, "loss": 0.5871, "step": 530 }, { "grad_norm": 1.896436333656311, "learning_rate": 1.3475000000000002e-05, "loss": 0.5557, "step": 540 }, { "grad_norm": 1.6398714780807495, "learning_rate": 1.3725000000000002e-05, "loss": 0.5148, "step": 550 }, { "grad_norm": 1.4757729768753052, "learning_rate": 1.3975000000000003e-05, "loss": 0.4921, "step": 560 }, { "grad_norm": 1.8107072114944458, "learning_rate": 1.4225e-05, "loss": 0.4692, "step": 570 }, { "grad_norm": 2.0016868114471436, "learning_rate": 1.4475e-05, "loss": 0.4303, "step": 580 }, { "grad_norm": 2.0596532821655273, "learning_rate": 1.4725e-05, "loss": 0.4023, "step": 590 }, { "grad_norm": 2.263878107070923, "learning_rate": 1.4975e-05, "loss": 0.3845, "step": 600 }, { "grad_norm": 1.6843310594558716, "learning_rate": 1.5225e-05, "loss": 0.3529, "step": 610 }, { "grad_norm": 1.8144149780273438, "learning_rate": 1.5475e-05, "loss": 0.338, "step": 620 }, { "grad_norm": 2.27608323097229, "learning_rate": 1.5725e-05, "loss": 0.3118, "step": 630 }, { "grad_norm": 1.7618740797042847, "learning_rate": 1.5975000000000002e-05, "loss": 0.2893, "step": 640 }, { "grad_norm": 2.1664535999298096, "learning_rate": 1.6225e-05, "loss": 0.2832, "step": 650 }, { "grad_norm": 1.4426337480545044, "learning_rate": 1.6475e-05, "loss": 0.27, "step": 660 }, { "grad_norm": 2.073101282119751, "learning_rate": 1.6725000000000003e-05, "loss": 0.2417, "step": 670 }, { "grad_norm": 2.791332244873047, "learning_rate": 1.6975000000000003e-05, "loss": 0.2286, "step": 680 }, { "grad_norm": 1.8503005504608154, "learning_rate": 1.7225e-05, "loss": 0.2235, "step": 690 }, { "grad_norm": 1.5837299823760986, "learning_rate": 1.7475e-05, "loss": 0.2084, "step": 700 }, { "grad_norm": 2.5951075553894043, "learning_rate": 1.7725e-05, "loss": 0.2215, "step": 710 }, { "grad_norm": 1.9050532579421997, "learning_rate": 1.7975e-05, "loss": 0.1993, "step": 720 }, { "grad_norm": 2.3153669834136963, "learning_rate": 1.8225e-05, "loss": 0.1742, "step": 730 }, { "grad_norm": 2.00223445892334, "learning_rate": 1.8475000000000002e-05, "loss": 0.1718, "step": 740 }, { "grad_norm": 2.2380738258361816, "learning_rate": 1.8725e-05, "loss": 0.1635, "step": 750 }, { "grad_norm": 1.6764836311340332, "learning_rate": 1.8975e-05, "loss": 0.1612, "step": 760 }, { "grad_norm": 1.7630505561828613, "learning_rate": 1.9225e-05, "loss": 0.1482, "step": 770 }, { "grad_norm": 1.9183069467544556, "learning_rate": 1.9475000000000002e-05, "loss": 0.1443, "step": 780 }, { "grad_norm": 1.5340920686721802, "learning_rate": 1.9725000000000002e-05, "loss": 0.1433, "step": 790 }, { "grad_norm": 2.148404836654663, "learning_rate": 1.9975e-05, "loss": 0.1394, "step": 800 }, { "grad_norm": 1.509039044380188, "learning_rate": 2.0225000000000004e-05, "loss": 0.1339, "step": 810 }, { "grad_norm": 1.9193981885910034, "learning_rate": 2.0475e-05, "loss": 0.1269, "step": 820 }, { "grad_norm": 1.7333346605300903, "learning_rate": 2.0725e-05, "loss": 0.1302, "step": 830 }, { "grad_norm": 2.1484999656677246, "learning_rate": 2.0975e-05, "loss": 0.1244, "step": 840 }, { "grad_norm": 1.6308374404907227, "learning_rate": 2.1225e-05, "loss": 0.1281, "step": 850 }, { "grad_norm": 2.0011863708496094, "learning_rate": 2.1475e-05, "loss": 0.132, "step": 860 }, { "grad_norm": 1.9148610830307007, "learning_rate": 2.1725e-05, "loss": 0.1205, "step": 870 }, { "grad_norm": 1.8098151683807373, "learning_rate": 2.1975000000000002e-05, "loss": 0.1225, "step": 880 }, { "grad_norm": 1.6718387603759766, "learning_rate": 2.2225e-05, "loss": 0.1134, "step": 890 }, { "grad_norm": 1.8246285915374756, "learning_rate": 2.2475e-05, "loss": 0.1097, "step": 900 }, { "grad_norm": 1.636834979057312, "learning_rate": 2.2725000000000003e-05, "loss": 0.107, "step": 910 }, { "grad_norm": 1.624720811843872, "learning_rate": 2.2975000000000003e-05, "loss": 0.107, "step": 920 }, { "grad_norm": 1.537421464920044, "learning_rate": 2.3225000000000002e-05, "loss": 0.1076, "step": 930 }, { "grad_norm": 1.9134464263916016, "learning_rate": 2.3475e-05, "loss": 0.1057, "step": 940 }, { "grad_norm": 1.350211501121521, "learning_rate": 2.3725e-05, "loss": 0.1081, "step": 950 }, { "grad_norm": 1.6219186782836914, "learning_rate": 2.3975e-05, "loss": 0.1089, "step": 960 }, { "grad_norm": 1.7075656652450562, "learning_rate": 2.4225e-05, "loss": 0.1054, "step": 970 }, { "grad_norm": 1.580180048942566, "learning_rate": 2.4475000000000002e-05, "loss": 0.1088, "step": 980 }, { "grad_norm": 1.7007418870925903, "learning_rate": 2.4725e-05, "loss": 0.1094, "step": 990 }, { "grad_norm": 1.3982478380203247, "learning_rate": 2.4975e-05, "loss": 0.1103, "step": 1000 }, { "grad_norm": 1.6127362251281738, "learning_rate": 2.5225e-05, "loss": 0.1031, "step": 1010 }, { "grad_norm": 1.3123494386672974, "learning_rate": 2.5475e-05, "loss": 0.109, "step": 1020 }, { "grad_norm": 1.5768927335739136, "learning_rate": 2.5725e-05, "loss": 0.1096, "step": 1030 }, { "grad_norm": 1.4229018688201904, "learning_rate": 2.5974999999999998e-05, "loss": 0.1019, "step": 1040 }, { "grad_norm": 1.8763872385025024, "learning_rate": 2.6225e-05, "loss": 0.0959, "step": 1050 }, { "grad_norm": 1.5930166244506836, "learning_rate": 2.6475e-05, "loss": 0.093, "step": 1060 }, { "grad_norm": 2.02498197555542, "learning_rate": 2.6725e-05, "loss": 0.0996, "step": 1070 }, { "grad_norm": 1.4271013736724854, "learning_rate": 2.6975000000000002e-05, "loss": 0.09, "step": 1080 }, { "grad_norm": 1.2051818370819092, "learning_rate": 2.7225e-05, "loss": 0.0921, "step": 1090 }, { "grad_norm": 1.3378522396087646, "learning_rate": 2.7475e-05, "loss": 0.1068, "step": 1100 }, { "grad_norm": 1.245944619178772, "learning_rate": 2.7725e-05, "loss": 0.0897, "step": 1110 }, { "grad_norm": 1.4278147220611572, "learning_rate": 2.7975000000000002e-05, "loss": 0.0991, "step": 1120 }, { "grad_norm": 1.3979853391647339, "learning_rate": 2.8225e-05, "loss": 0.0965, "step": 1130 }, { "grad_norm": 1.7562754154205322, "learning_rate": 2.8475e-05, "loss": 0.0952, "step": 1140 }, { "grad_norm": 1.2771732807159424, "learning_rate": 2.8725e-05, "loss": 0.0955, "step": 1150 }, { "grad_norm": 1.3724713325500488, "learning_rate": 2.8975000000000003e-05, "loss": 0.0894, "step": 1160 }, { "grad_norm": 1.6313719749450684, "learning_rate": 2.9225000000000002e-05, "loss": 0.0937, "step": 1170 }, { "grad_norm": 1.5779852867126465, "learning_rate": 2.9475e-05, "loss": 0.0974, "step": 1180 }, { "grad_norm": 1.3955227136611938, "learning_rate": 2.9725000000000004e-05, "loss": 0.0893, "step": 1190 }, { "grad_norm": 1.4362478256225586, "learning_rate": 2.9975000000000004e-05, "loss": 0.0949, "step": 1200 }, { "grad_norm": 1.1792192459106445, "learning_rate": 3.0225000000000003e-05, "loss": 0.0893, "step": 1210 }, { "grad_norm": 1.2383177280426025, "learning_rate": 3.0475000000000002e-05, "loss": 0.0815, "step": 1220 }, { "grad_norm": 1.1631431579589844, "learning_rate": 3.0725e-05, "loss": 0.0851, "step": 1230 }, { "grad_norm": 1.1800059080123901, "learning_rate": 3.0975e-05, "loss": 0.0924, "step": 1240 }, { "grad_norm": 1.291428565979004, "learning_rate": 3.122500000000001e-05, "loss": 0.0879, "step": 1250 }, { "grad_norm": 1.5706795454025269, "learning_rate": 3.1475e-05, "loss": 0.0772, "step": 1260 }, { "grad_norm": 1.444273591041565, "learning_rate": 3.1725e-05, "loss": 0.0974, "step": 1270 }, { "grad_norm": 1.6144702434539795, "learning_rate": 3.1975e-05, "loss": 0.0869, "step": 1280 }, { "grad_norm": 1.0006299018859863, "learning_rate": 3.2225e-05, "loss": 0.0867, "step": 1290 }, { "grad_norm": 1.5861685276031494, "learning_rate": 3.2474999999999997e-05, "loss": 0.0867, "step": 1300 }, { "grad_norm": 1.2715297937393188, "learning_rate": 3.2725e-05, "loss": 0.0801, "step": 1310 }, { "grad_norm": 1.3936710357666016, "learning_rate": 3.2975e-05, "loss": 0.0952, "step": 1320 }, { "grad_norm": 1.0783295631408691, "learning_rate": 3.3225e-05, "loss": 0.0776, "step": 1330 }, { "grad_norm": 1.2263141870498657, "learning_rate": 3.3475e-05, "loss": 0.0746, "step": 1340 }, { "grad_norm": 1.4308931827545166, "learning_rate": 3.3725e-05, "loss": 0.0844, "step": 1350 }, { "grad_norm": 1.4008827209472656, "learning_rate": 3.3975e-05, "loss": 0.0903, "step": 1360 }, { "grad_norm": 1.4701049327850342, "learning_rate": 3.4225e-05, "loss": 0.0814, "step": 1370 }, { "grad_norm": 1.1517783403396606, "learning_rate": 3.4475000000000005e-05, "loss": 0.0768, "step": 1380 }, { "grad_norm": 1.1980280876159668, "learning_rate": 3.4725000000000004e-05, "loss": 0.0707, "step": 1390 }, { "grad_norm": 1.1788345575332642, "learning_rate": 3.4975e-05, "loss": 0.0727, "step": 1400 }, { "grad_norm": 1.399664282798767, "learning_rate": 3.5225e-05, "loss": 0.0725, "step": 1410 }, { "grad_norm": 1.028367519378662, "learning_rate": 3.5475e-05, "loss": 0.079, "step": 1420 }, { "grad_norm": 1.2859041690826416, "learning_rate": 3.5725e-05, "loss": 0.0861, "step": 1430 }, { "grad_norm": 1.4616445302963257, "learning_rate": 3.5975e-05, "loss": 0.0765, "step": 1440 }, { "grad_norm": 1.2596211433410645, "learning_rate": 3.6225000000000006e-05, "loss": 0.0716, "step": 1450 }, { "grad_norm": 1.2865813970565796, "learning_rate": 3.6475000000000006e-05, "loss": 0.0784, "step": 1460 }, { "grad_norm": 1.413537621498108, "learning_rate": 3.6725000000000005e-05, "loss": 0.0734, "step": 1470 }, { "grad_norm": 1.0321540832519531, "learning_rate": 3.6975000000000004e-05, "loss": 0.0705, "step": 1480 }, { "grad_norm": 1.0979225635528564, "learning_rate": 3.7225000000000004e-05, "loss": 0.0782, "step": 1490 }, { "grad_norm": 1.077423334121704, "learning_rate": 3.7475e-05, "loss": 0.0734, "step": 1500 }, { "grad_norm": 1.1100037097930908, "learning_rate": 3.7725e-05, "loss": 0.0765, "step": 1510 }, { "grad_norm": 1.1875108480453491, "learning_rate": 3.7975e-05, "loss": 0.0747, "step": 1520 }, { "grad_norm": 1.3922284841537476, "learning_rate": 3.8225e-05, "loss": 0.0741, "step": 1530 }, { "grad_norm": 1.0001031160354614, "learning_rate": 3.8475e-05, "loss": 0.0759, "step": 1540 }, { "grad_norm": 1.1327162981033325, "learning_rate": 3.8725e-05, "loss": 0.0716, "step": 1550 }, { "grad_norm": 1.3340402841567993, "learning_rate": 3.8975e-05, "loss": 0.0747, "step": 1560 }, { "grad_norm": 1.425162434577942, "learning_rate": 3.9225e-05, "loss": 0.0765, "step": 1570 }, { "grad_norm": 1.1320658922195435, "learning_rate": 3.9475000000000004e-05, "loss": 0.0694, "step": 1580 }, { "grad_norm": 1.0466432571411133, "learning_rate": 3.9725e-05, "loss": 0.0683, "step": 1590 }, { "grad_norm": 1.0965310335159302, "learning_rate": 3.9975e-05, "loss": 0.076, "step": 1600 }, { "grad_norm": 1.0701987743377686, "learning_rate": 4.0225e-05, "loss": 0.0691, "step": 1610 }, { "grad_norm": 1.0443925857543945, "learning_rate": 4.0475e-05, "loss": 0.0685, "step": 1620 }, { "grad_norm": 1.0991504192352295, "learning_rate": 4.0725e-05, "loss": 0.0761, "step": 1630 }, { "grad_norm": 1.2984572649002075, "learning_rate": 4.0975e-05, "loss": 0.0727, "step": 1640 }, { "grad_norm": 1.0625362396240234, "learning_rate": 4.1225e-05, "loss": 0.0683, "step": 1650 }, { "grad_norm": 1.0158811807632446, "learning_rate": 4.1475000000000005e-05, "loss": 0.0667, "step": 1660 }, { "grad_norm": 0.9768041372299194, "learning_rate": 4.1725000000000005e-05, "loss": 0.0724, "step": 1670 }, { "grad_norm": 1.0482361316680908, "learning_rate": 4.1975000000000004e-05, "loss": 0.0647, "step": 1680 }, { "grad_norm": 1.018424391746521, "learning_rate": 4.2225e-05, "loss": 0.0638, "step": 1690 }, { "grad_norm": 0.9575387239456177, "learning_rate": 4.2475e-05, "loss": 0.0792, "step": 1700 }, { "grad_norm": 0.9824761152267456, "learning_rate": 4.2725e-05, "loss": 0.069, "step": 1710 }, { "grad_norm": 1.0223362445831299, "learning_rate": 4.2975e-05, "loss": 0.0709, "step": 1720 }, { "grad_norm": 1.0675512552261353, "learning_rate": 4.322500000000001e-05, "loss": 0.0705, "step": 1730 }, { "grad_norm": 0.9600633382797241, "learning_rate": 4.3475000000000006e-05, "loss": 0.0659, "step": 1740 }, { "grad_norm": 1.069966197013855, "learning_rate": 4.3725000000000006e-05, "loss": 0.0654, "step": 1750 }, { "grad_norm": 0.8766752481460571, "learning_rate": 4.3975e-05, "loss": 0.0619, "step": 1760 }, { "grad_norm": 1.2014628648757935, "learning_rate": 4.4225e-05, "loss": 0.0783, "step": 1770 }, { "grad_norm": 0.9762494564056396, "learning_rate": 4.4475e-05, "loss": 0.0711, "step": 1780 }, { "grad_norm": 0.9272044897079468, "learning_rate": 4.4725e-05, "loss": 0.0665, "step": 1790 }, { "grad_norm": 0.7298517823219299, "learning_rate": 4.4975e-05, "loss": 0.0741, "step": 1800 }, { "grad_norm": 0.9515014886856079, "learning_rate": 4.5225e-05, "loss": 0.0673, "step": 1810 }, { "grad_norm": 1.072855830192566, "learning_rate": 4.5475e-05, "loss": 0.0665, "step": 1820 }, { "grad_norm": 1.0082136392593384, "learning_rate": 4.5725e-05, "loss": 0.0634, "step": 1830 }, { "grad_norm": 1.0119860172271729, "learning_rate": 4.5975e-05, "loss": 0.0608, "step": 1840 }, { "grad_norm": 1.0505754947662354, "learning_rate": 4.6225e-05, "loss": 0.0645, "step": 1850 }, { "grad_norm": 0.831743061542511, "learning_rate": 4.6475000000000005e-05, "loss": 0.0635, "step": 1860 }, { "grad_norm": 0.8519420623779297, "learning_rate": 4.6725000000000004e-05, "loss": 0.0661, "step": 1870 }, { "grad_norm": 1.0972480773925781, "learning_rate": 4.6975000000000003e-05, "loss": 0.0621, "step": 1880 }, { "grad_norm": 1.034976840019226, "learning_rate": 4.7225e-05, "loss": 0.0603, "step": 1890 }, { "grad_norm": 0.8545805811882019, "learning_rate": 4.7475e-05, "loss": 0.0689, "step": 1900 }, { "grad_norm": 1.0038020610809326, "learning_rate": 4.7725e-05, "loss": 0.0674, "step": 1910 }, { "grad_norm": 0.8302486538887024, "learning_rate": 4.7975e-05, "loss": 0.0599, "step": 1920 }, { "grad_norm": 0.8866508603096008, "learning_rate": 4.822500000000001e-05, "loss": 0.0629, "step": 1930 }, { "grad_norm": 1.210331678390503, "learning_rate": 4.8475000000000006e-05, "loss": 0.0766, "step": 1940 }, { "grad_norm": 0.8972250819206238, "learning_rate": 4.8725000000000005e-05, "loss": 0.0659, "step": 1950 }, { "grad_norm": 0.8315176963806152, "learning_rate": 4.8975000000000005e-05, "loss": 0.0588, "step": 1960 }, { "grad_norm": 1.2388745546340942, "learning_rate": 4.9225000000000004e-05, "loss": 0.0626, "step": 1970 }, { "grad_norm": 0.8640565872192383, "learning_rate": 4.9475e-05, "loss": 0.0691, "step": 1980 }, { "grad_norm": 0.8440235257148743, "learning_rate": 4.9725e-05, "loss": 0.0601, "step": 1990 }, { "grad_norm": 0.8946271538734436, "learning_rate": 4.9975e-05, "loss": 0.0625, "step": 2000 }, { "grad_norm": 0.8231354355812073, "learning_rate": 5.0225e-05, "loss": 0.0625, "step": 2010 }, { "grad_norm": 1.028838872909546, "learning_rate": 5.047500000000001e-05, "loss": 0.0646, "step": 2020 }, { "grad_norm": 0.7520943284034729, "learning_rate": 5.0725e-05, "loss": 0.058, "step": 2030 }, { "grad_norm": 0.7868366837501526, "learning_rate": 5.0975000000000006e-05, "loss": 0.0668, "step": 2040 }, { "grad_norm": 1.051331877708435, "learning_rate": 5.1225e-05, "loss": 0.0696, "step": 2050 }, { "grad_norm": 1.1799931526184082, "learning_rate": 5.1475000000000004e-05, "loss": 0.0661, "step": 2060 }, { "grad_norm": 1.0628348588943481, "learning_rate": 5.1725000000000004e-05, "loss": 0.0638, "step": 2070 }, { "grad_norm": 0.9143269658088684, "learning_rate": 5.197500000000001e-05, "loss": 0.0643, "step": 2080 }, { "grad_norm": 1.0322067737579346, "learning_rate": 5.2225e-05, "loss": 0.0656, "step": 2090 }, { "grad_norm": 0.8795320391654968, "learning_rate": 5.247500000000001e-05, "loss": 0.0623, "step": 2100 }, { "grad_norm": 0.9428721070289612, "learning_rate": 5.2725e-05, "loss": 0.0606, "step": 2110 }, { "grad_norm": 1.1099660396575928, "learning_rate": 5.297500000000001e-05, "loss": 0.0617, "step": 2120 }, { "grad_norm": 0.7510033249855042, "learning_rate": 5.3225e-05, "loss": 0.0658, "step": 2130 }, { "grad_norm": 0.9254029393196106, "learning_rate": 5.3475e-05, "loss": 0.0612, "step": 2140 }, { "grad_norm": 0.751883864402771, "learning_rate": 5.3725000000000005e-05, "loss": 0.0594, "step": 2150 }, { "grad_norm": 0.7770131826400757, "learning_rate": 5.3975e-05, "loss": 0.0509, "step": 2160 }, { "grad_norm": 0.8493210077285767, "learning_rate": 5.4225000000000003e-05, "loss": 0.06, "step": 2170 }, { "grad_norm": 0.7980387210845947, "learning_rate": 5.4474999999999996e-05, "loss": 0.0602, "step": 2180 }, { "grad_norm": 0.6189719438552856, "learning_rate": 5.4725e-05, "loss": 0.0591, "step": 2190 }, { "grad_norm": 0.7455147504806519, "learning_rate": 5.4975e-05, "loss": 0.0551, "step": 2200 }, { "grad_norm": 0.7110676765441895, "learning_rate": 5.522500000000001e-05, "loss": 0.0529, "step": 2210 }, { "grad_norm": 0.7411757707595825, "learning_rate": 5.5475e-05, "loss": 0.055, "step": 2220 }, { "grad_norm": 0.7644950747489929, "learning_rate": 5.5725000000000006e-05, "loss": 0.0555, "step": 2230 }, { "grad_norm": 0.7163835763931274, "learning_rate": 5.5975e-05, "loss": 0.0593, "step": 2240 }, { "grad_norm": 0.6163251399993896, "learning_rate": 5.6225000000000005e-05, "loss": 0.0643, "step": 2250 }, { "grad_norm": 0.7418830394744873, "learning_rate": 5.6475e-05, "loss": 0.066, "step": 2260 }, { "grad_norm": 0.8261330127716064, "learning_rate": 5.6725e-05, "loss": 0.0637, "step": 2270 }, { "grad_norm": 0.6793302893638611, "learning_rate": 5.6975e-05, "loss": 0.0566, "step": 2280 }, { "grad_norm": 0.6515572667121887, "learning_rate": 5.722500000000001e-05, "loss": 0.0504, "step": 2290 }, { "grad_norm": 0.7960675954818726, "learning_rate": 5.7475e-05, "loss": 0.0555, "step": 2300 }, { "grad_norm": 0.5864563584327698, "learning_rate": 5.772500000000001e-05, "loss": 0.0571, "step": 2310 }, { "grad_norm": 0.637776255607605, "learning_rate": 5.7975e-05, "loss": 0.0575, "step": 2320 }, { "grad_norm": 0.873711884021759, "learning_rate": 5.8225000000000006e-05, "loss": 0.0638, "step": 2330 }, { "grad_norm": 0.6177480816841125, "learning_rate": 5.8475000000000005e-05, "loss": 0.0612, "step": 2340 }, { "grad_norm": 0.8055627346038818, "learning_rate": 5.8725000000000004e-05, "loss": 0.0564, "step": 2350 }, { "grad_norm": 0.8251268863677979, "learning_rate": 5.8975000000000004e-05, "loss": 0.0613, "step": 2360 }, { "grad_norm": 0.6766011714935303, "learning_rate": 5.922500000000001e-05, "loss": 0.0619, "step": 2370 }, { "grad_norm": 0.9299607276916504, "learning_rate": 5.9475e-05, "loss": 0.0604, "step": 2380 }, { "grad_norm": 0.7551127076148987, "learning_rate": 5.9724999999999995e-05, "loss": 0.0564, "step": 2390 }, { "grad_norm": 0.6799227595329285, "learning_rate": 5.9975e-05, "loss": 0.0592, "step": 2400 }, { "grad_norm": 0.5858661532402039, "learning_rate": 6.0225e-05, "loss": 0.0527, "step": 2410 }, { "grad_norm": 0.804330587387085, "learning_rate": 6.0475000000000006e-05, "loss": 0.058, "step": 2420 }, { "grad_norm": 0.7078824639320374, "learning_rate": 6.0725e-05, "loss": 0.0566, "step": 2430 }, { "grad_norm": 0.8924959897994995, "learning_rate": 6.0975000000000005e-05, "loss": 0.0578, "step": 2440 }, { "grad_norm": 0.6893863081932068, "learning_rate": 6.1225e-05, "loss": 0.0589, "step": 2450 }, { "grad_norm": 0.7928688526153564, "learning_rate": 6.1475e-05, "loss": 0.063, "step": 2460 }, { "grad_norm": 0.5326719880104065, "learning_rate": 6.1725e-05, "loss": 0.054, "step": 2470 }, { "grad_norm": 0.7189964652061462, "learning_rate": 6.1975e-05, "loss": 0.062, "step": 2480 }, { "grad_norm": 0.6169294714927673, "learning_rate": 6.2225e-05, "loss": 0.0534, "step": 2490 }, { "grad_norm": 0.6136389374732971, "learning_rate": 6.2475e-05, "loss": 0.0568, "step": 2500 }, { "grad_norm": 0.7750381827354431, "learning_rate": 6.2725e-05, "loss": 0.0576, "step": 2510 }, { "grad_norm": 0.6517468094825745, "learning_rate": 6.297500000000001e-05, "loss": 0.0544, "step": 2520 }, { "grad_norm": 0.6843379735946655, "learning_rate": 6.3225e-05, "loss": 0.05, "step": 2530 }, { "grad_norm": 0.5966778993606567, "learning_rate": 6.347500000000001e-05, "loss": 0.0532, "step": 2540 }, { "grad_norm": 0.8587526082992554, "learning_rate": 6.3725e-05, "loss": 0.0587, "step": 2550 }, { "grad_norm": 0.8257063031196594, "learning_rate": 6.397500000000001e-05, "loss": 0.0526, "step": 2560 }, { "grad_norm": 0.918013334274292, "learning_rate": 6.4225e-05, "loss": 0.0527, "step": 2570 }, { "grad_norm": 0.7775500416755676, "learning_rate": 6.447500000000001e-05, "loss": 0.0595, "step": 2580 }, { "grad_norm": 0.7274789810180664, "learning_rate": 6.4725e-05, "loss": 0.0572, "step": 2590 }, { "grad_norm": 0.5990054607391357, "learning_rate": 6.497500000000001e-05, "loss": 0.0571, "step": 2600 }, { "grad_norm": 0.7742208242416382, "learning_rate": 6.5225e-05, "loss": 0.0554, "step": 2610 }, { "grad_norm": 0.6569889187812805, "learning_rate": 6.5475e-05, "loss": 0.052, "step": 2620 }, { "grad_norm": 1.0096547603607178, "learning_rate": 6.5725e-05, "loss": 0.0608, "step": 2630 }, { "grad_norm": 0.72808438539505, "learning_rate": 6.5975e-05, "loss": 0.0589, "step": 2640 }, { "grad_norm": 0.4831610321998596, "learning_rate": 6.6225e-05, "loss": 0.0541, "step": 2650 }, { "grad_norm": 0.5575710535049438, "learning_rate": 6.6475e-05, "loss": 0.0504, "step": 2660 }, { "grad_norm": 0.7136884331703186, "learning_rate": 6.672500000000001e-05, "loss": 0.0524, "step": 2670 }, { "grad_norm": 0.7139206528663635, "learning_rate": 6.6975e-05, "loss": 0.05, "step": 2680 }, { "grad_norm": 0.6018664240837097, "learning_rate": 6.722500000000001e-05, "loss": 0.0556, "step": 2690 }, { "grad_norm": 0.6141428351402283, "learning_rate": 6.7475e-05, "loss": 0.0534, "step": 2700 }, { "grad_norm": 0.6442875862121582, "learning_rate": 6.7725e-05, "loss": 0.057, "step": 2710 }, { "grad_norm": 0.751080334186554, "learning_rate": 6.7975e-05, "loss": 0.053, "step": 2720 }, { "grad_norm": 0.8846927285194397, "learning_rate": 6.8225e-05, "loss": 0.0556, "step": 2730 }, { "grad_norm": 0.7933576107025146, "learning_rate": 6.8475e-05, "loss": 0.0575, "step": 2740 }, { "grad_norm": 0.478277325630188, "learning_rate": 6.8725e-05, "loss": 0.0535, "step": 2750 }, { "grad_norm": 0.6459885835647583, "learning_rate": 6.8975e-05, "loss": 0.0577, "step": 2760 }, { "grad_norm": 0.5077873468399048, "learning_rate": 6.9225e-05, "loss": 0.0476, "step": 2770 }, { "grad_norm": 0.5954045057296753, "learning_rate": 6.9475e-05, "loss": 0.0536, "step": 2780 }, { "grad_norm": 0.5597285628318787, "learning_rate": 6.9725e-05, "loss": 0.0578, "step": 2790 }, { "grad_norm": 0.7672089338302612, "learning_rate": 6.997500000000001e-05, "loss": 0.0516, "step": 2800 }, { "grad_norm": 0.6561864614486694, "learning_rate": 7.022500000000001e-05, "loss": 0.0504, "step": 2810 }, { "grad_norm": 0.7219908237457275, "learning_rate": 7.0475e-05, "loss": 0.0444, "step": 2820 }, { "grad_norm": 0.6153291463851929, "learning_rate": 7.072500000000001e-05, "loss": 0.0514, "step": 2830 }, { "grad_norm": 0.5270136594772339, "learning_rate": 7.0975e-05, "loss": 0.0541, "step": 2840 }, { "grad_norm": 0.4338686764240265, "learning_rate": 7.122500000000001e-05, "loss": 0.0506, "step": 2850 }, { "grad_norm": 0.6230470538139343, "learning_rate": 7.1475e-05, "loss": 0.0514, "step": 2860 }, { "grad_norm": 0.5985748171806335, "learning_rate": 7.172500000000001e-05, "loss": 0.0524, "step": 2870 }, { "grad_norm": 0.6185658574104309, "learning_rate": 7.1975e-05, "loss": 0.0503, "step": 2880 }, { "grad_norm": 0.5066505074501038, "learning_rate": 7.2225e-05, "loss": 0.0481, "step": 2890 }, { "grad_norm": 0.7561277747154236, "learning_rate": 7.2475e-05, "loss": 0.0504, "step": 2900 }, { "grad_norm": 0.7281303405761719, "learning_rate": 7.272499999999999e-05, "loss": 0.0502, "step": 2910 }, { "grad_norm": 0.5447114109992981, "learning_rate": 7.2975e-05, "loss": 0.0468, "step": 2920 }, { "grad_norm": 0.5884737968444824, "learning_rate": 7.3225e-05, "loss": 0.056, "step": 2930 }, { "grad_norm": 0.5766794085502625, "learning_rate": 7.347500000000001e-05, "loss": 0.0582, "step": 2940 }, { "grad_norm": 0.5330209136009216, "learning_rate": 7.3725e-05, "loss": 0.0554, "step": 2950 }, { "grad_norm": 0.6269991397857666, "learning_rate": 7.397500000000001e-05, "loss": 0.0555, "step": 2960 }, { "grad_norm": 0.47158142924308777, "learning_rate": 7.4225e-05, "loss": 0.0506, "step": 2970 }, { "grad_norm": 0.7803083658218384, "learning_rate": 7.447500000000001e-05, "loss": 0.0483, "step": 2980 }, { "grad_norm": 0.545549750328064, "learning_rate": 7.4725e-05, "loss": 0.049, "step": 2990 }, { "grad_norm": 0.5283257365226746, "learning_rate": 7.4975e-05, "loss": 0.0501, "step": 3000 }, { "grad_norm": 0.6652109026908875, "learning_rate": 7.5225e-05, "loss": 0.0554, "step": 3010 }, { "grad_norm": 0.6098585724830627, "learning_rate": 7.5475e-05, "loss": 0.0493, "step": 3020 }, { "grad_norm": 0.5460692048072815, "learning_rate": 7.5725e-05, "loss": 0.0474, "step": 3030 }, { "grad_norm": 0.666548490524292, "learning_rate": 7.5975e-05, "loss": 0.0475, "step": 3040 }, { "grad_norm": 0.5687774419784546, "learning_rate": 7.6225e-05, "loss": 0.0499, "step": 3050 }, { "grad_norm": 0.5317971110343933, "learning_rate": 7.6475e-05, "loss": 0.0487, "step": 3060 }, { "grad_norm": 0.6172863245010376, "learning_rate": 7.672500000000001e-05, "loss": 0.0499, "step": 3070 }, { "grad_norm": 0.5055314302444458, "learning_rate": 7.697500000000001e-05, "loss": 0.0489, "step": 3080 }, { "grad_norm": 0.6995224356651306, "learning_rate": 7.722500000000001e-05, "loss": 0.0489, "step": 3090 }, { "grad_norm": 0.8053944110870361, "learning_rate": 7.747500000000001e-05, "loss": 0.0553, "step": 3100 }, { "grad_norm": 0.8026812076568604, "learning_rate": 7.7725e-05, "loss": 0.0574, "step": 3110 }, { "grad_norm": 0.6838295459747314, "learning_rate": 7.797500000000001e-05, "loss": 0.0484, "step": 3120 }, { "grad_norm": 0.47922876477241516, "learning_rate": 7.8225e-05, "loss": 0.0585, "step": 3130 }, { "grad_norm": 0.6339093446731567, "learning_rate": 7.8475e-05, "loss": 0.0466, "step": 3140 }, { "grad_norm": 0.5474699139595032, "learning_rate": 7.8725e-05, "loss": 0.0528, "step": 3150 }, { "grad_norm": 0.601908266544342, "learning_rate": 7.8975e-05, "loss": 0.0447, "step": 3160 }, { "grad_norm": 0.4975273311138153, "learning_rate": 7.9225e-05, "loss": 0.0496, "step": 3170 }, { "grad_norm": 0.503985583782196, "learning_rate": 7.9475e-05, "loss": 0.0534, "step": 3180 }, { "grad_norm": 0.4808591306209564, "learning_rate": 7.9725e-05, "loss": 0.0522, "step": 3190 }, { "grad_norm": 0.5129116177558899, "learning_rate": 7.9975e-05, "loss": 0.0466, "step": 3200 }, { "grad_norm": 0.5424265265464783, "learning_rate": 8.022500000000001e-05, "loss": 0.0457, "step": 3210 }, { "grad_norm": 0.5987427830696106, "learning_rate": 8.0475e-05, "loss": 0.0483, "step": 3220 }, { "grad_norm": 0.46617454290390015, "learning_rate": 8.072500000000001e-05, "loss": 0.0564, "step": 3230 }, { "grad_norm": 0.6292984485626221, "learning_rate": 8.0975e-05, "loss": 0.05, "step": 3240 }, { "grad_norm": 0.7358474731445312, "learning_rate": 8.122500000000001e-05, "loss": 0.0515, "step": 3250 }, { "grad_norm": 0.5888814330101013, "learning_rate": 8.1475e-05, "loss": 0.0524, "step": 3260 }, { "grad_norm": 0.56321781873703, "learning_rate": 8.172500000000001e-05, "loss": 0.0531, "step": 3270 }, { "grad_norm": 0.5575167536735535, "learning_rate": 8.1975e-05, "loss": 0.0468, "step": 3280 }, { "grad_norm": 0.6296348571777344, "learning_rate": 8.2225e-05, "loss": 0.0518, "step": 3290 }, { "grad_norm": 0.5492444038391113, "learning_rate": 8.2475e-05, "loss": 0.053, "step": 3300 }, { "grad_norm": 0.5463634133338928, "learning_rate": 8.2725e-05, "loss": 0.0519, "step": 3310 }, { "grad_norm": 0.48049548268318176, "learning_rate": 8.2975e-05, "loss": 0.0529, "step": 3320 }, { "grad_norm": 0.4694688320159912, "learning_rate": 8.3225e-05, "loss": 0.0496, "step": 3330 }, { "grad_norm": 0.45655393600463867, "learning_rate": 8.347500000000001e-05, "loss": 0.0503, "step": 3340 }, { "grad_norm": 0.5109291076660156, "learning_rate": 8.3725e-05, "loss": 0.0503, "step": 3350 }, { "grad_norm": 0.4900495409965515, "learning_rate": 8.397500000000001e-05, "loss": 0.0471, "step": 3360 }, { "grad_norm": 0.6177843809127808, "learning_rate": 8.422500000000001e-05, "loss": 0.047, "step": 3370 }, { "grad_norm": 0.5055627226829529, "learning_rate": 8.447500000000001e-05, "loss": 0.0472, "step": 3380 }, { "grad_norm": 0.5365391969680786, "learning_rate": 8.4725e-05, "loss": 0.0492, "step": 3390 }, { "grad_norm": 0.5930346846580505, "learning_rate": 8.4975e-05, "loss": 0.0536, "step": 3400 }, { "grad_norm": 0.46174219250679016, "learning_rate": 8.5225e-05, "loss": 0.0541, "step": 3410 }, { "grad_norm": 0.5849230885505676, "learning_rate": 8.5475e-05, "loss": 0.0465, "step": 3420 }, { "grad_norm": 0.6723579168319702, "learning_rate": 8.5725e-05, "loss": 0.0459, "step": 3430 }, { "grad_norm": 0.6256141662597656, "learning_rate": 8.5975e-05, "loss": 0.052, "step": 3440 }, { "grad_norm": 0.5310952663421631, "learning_rate": 8.6225e-05, "loss": 0.0423, "step": 3450 }, { "grad_norm": 0.5102205872535706, "learning_rate": 8.6475e-05, "loss": 0.0472, "step": 3460 }, { "grad_norm": 0.4807256758213043, "learning_rate": 8.672500000000001e-05, "loss": 0.0522, "step": 3470 }, { "grad_norm": 0.5060203075408936, "learning_rate": 8.6975e-05, "loss": 0.0488, "step": 3480 }, { "grad_norm": 0.5258985757827759, "learning_rate": 8.7225e-05, "loss": 0.047, "step": 3490 }, { "grad_norm": 0.4901120066642761, "learning_rate": 8.747500000000001e-05, "loss": 0.0465, "step": 3500 }, { "grad_norm": 0.5898966193199158, "learning_rate": 8.7725e-05, "loss": 0.0544, "step": 3510 }, { "grad_norm": 0.5114065408706665, "learning_rate": 8.797500000000001e-05, "loss": 0.0488, "step": 3520 }, { "grad_norm": 0.5239070653915405, "learning_rate": 8.8225e-05, "loss": 0.0486, "step": 3530 }, { "grad_norm": 0.45670586824417114, "learning_rate": 8.847500000000001e-05, "loss": 0.0473, "step": 3540 }, { "grad_norm": 0.4664277732372284, "learning_rate": 8.8725e-05, "loss": 0.0483, "step": 3550 }, { "grad_norm": 0.6246154308319092, "learning_rate": 8.897500000000001e-05, "loss": 0.049, "step": 3560 }, { "grad_norm": 0.541932225227356, "learning_rate": 8.9225e-05, "loss": 0.0529, "step": 3570 }, { "grad_norm": 0.5327295660972595, "learning_rate": 8.9475e-05, "loss": 0.0536, "step": 3580 }, { "grad_norm": 0.6369325518608093, "learning_rate": 8.9725e-05, "loss": 0.0477, "step": 3590 }, { "grad_norm": 0.5344922542572021, "learning_rate": 8.9975e-05, "loss": 0.049, "step": 3600 }, { "grad_norm": 0.6004131436347961, "learning_rate": 9.0225e-05, "loss": 0.0473, "step": 3610 }, { "grad_norm": 0.5958373546600342, "learning_rate": 9.0475e-05, "loss": 0.0471, "step": 3620 }, { "grad_norm": 0.4797953963279724, "learning_rate": 9.072500000000001e-05, "loss": 0.0501, "step": 3630 }, { "grad_norm": 0.5953472852706909, "learning_rate": 9.0975e-05, "loss": 0.0446, "step": 3640 }, { "grad_norm": 0.431241512298584, "learning_rate": 9.122500000000001e-05, "loss": 0.0422, "step": 3650 }, { "grad_norm": 0.5011346936225891, "learning_rate": 9.1475e-05, "loss": 0.0451, "step": 3660 }, { "grad_norm": 0.4215828478336334, "learning_rate": 9.172500000000001e-05, "loss": 0.0439, "step": 3670 }, { "grad_norm": 0.5788990259170532, "learning_rate": 9.1975e-05, "loss": 0.044, "step": 3680 }, { "grad_norm": 0.47038429975509644, "learning_rate": 9.2225e-05, "loss": 0.0518, "step": 3690 }, { "grad_norm": 0.42414426803588867, "learning_rate": 9.2475e-05, "loss": 0.0394, "step": 3700 }, { "grad_norm": 0.5336163640022278, "learning_rate": 9.2725e-05, "loss": 0.044, "step": 3710 }, { "grad_norm": 0.4172574579715729, "learning_rate": 9.2975e-05, "loss": 0.0446, "step": 3720 }, { "grad_norm": 0.4402601420879364, "learning_rate": 9.3225e-05, "loss": 0.0482, "step": 3730 }, { "grad_norm": 0.4593372344970703, "learning_rate": 9.3475e-05, "loss": 0.047, "step": 3740 }, { "grad_norm": 0.39701226353645325, "learning_rate": 9.3725e-05, "loss": 0.0457, "step": 3750 }, { "grad_norm": 0.3079114556312561, "learning_rate": 9.397500000000001e-05, "loss": 0.0461, "step": 3760 }, { "grad_norm": 0.5168612599372864, "learning_rate": 9.422500000000001e-05, "loss": 0.0393, "step": 3770 }, { "grad_norm": 0.5133991241455078, "learning_rate": 9.4475e-05, "loss": 0.0459, "step": 3780 }, { "grad_norm": 0.5842492580413818, "learning_rate": 9.472500000000001e-05, "loss": 0.0444, "step": 3790 }, { "grad_norm": 0.553408682346344, "learning_rate": 9.4975e-05, "loss": 0.0482, "step": 3800 }, { "grad_norm": 0.4971154034137726, "learning_rate": 9.522500000000001e-05, "loss": 0.0502, "step": 3810 }, { "grad_norm": 0.5130455493927002, "learning_rate": 9.5475e-05, "loss": 0.0516, "step": 3820 }, { "grad_norm": 0.5035747289657593, "learning_rate": 9.572500000000001e-05, "loss": 0.045, "step": 3830 }, { "grad_norm": 0.4808500409126282, "learning_rate": 9.5975e-05, "loss": 0.0449, "step": 3840 }, { "grad_norm": 0.6116794943809509, "learning_rate": 9.622500000000001e-05, "loss": 0.0458, "step": 3850 }, { "grad_norm": 0.46652448177337646, "learning_rate": 9.6475e-05, "loss": 0.0477, "step": 3860 }, { "grad_norm": 0.4838227331638336, "learning_rate": 9.6725e-05, "loss": 0.0418, "step": 3870 }, { "grad_norm": 0.4763167202472687, "learning_rate": 9.6975e-05, "loss": 0.0427, "step": 3880 }, { "grad_norm": 0.5234687328338623, "learning_rate": 9.7225e-05, "loss": 0.0488, "step": 3890 }, { "grad_norm": 0.5023638010025024, "learning_rate": 9.747500000000001e-05, "loss": 0.0526, "step": 3900 }, { "grad_norm": 0.4059363901615143, "learning_rate": 9.7725e-05, "loss": 0.0472, "step": 3910 }, { "grad_norm": 0.4368871748447418, "learning_rate": 9.797500000000001e-05, "loss": 0.0526, "step": 3920 }, { "grad_norm": 0.426087886095047, "learning_rate": 9.8225e-05, "loss": 0.042, "step": 3930 }, { "grad_norm": 0.4688669443130493, "learning_rate": 9.847500000000001e-05, "loss": 0.0458, "step": 3940 }, { "grad_norm": 0.45402318239212036, "learning_rate": 9.8725e-05, "loss": 0.0485, "step": 3950 }, { "grad_norm": 0.427382230758667, "learning_rate": 9.897500000000001e-05, "loss": 0.0465, "step": 3960 }, { "grad_norm": 0.43280676007270813, "learning_rate": 9.9225e-05, "loss": 0.0438, "step": 3970 }, { "grad_norm": 0.4968304932117462, "learning_rate": 9.9475e-05, "loss": 0.0427, "step": 3980 }, { "grad_norm": 0.46578094363212585, "learning_rate": 9.9725e-05, "loss": 0.045, "step": 3990 }, { "grad_norm": 0.5831339359283447, "learning_rate": 9.9975e-05, "loss": 0.0479, "step": 4000 }, { "grad_norm": 0.5384007096290588, "learning_rate": 9.999999653982884e-05, "loss": 0.043, "step": 4010 }, { "grad_norm": 0.5871061086654663, "learning_rate": 9.999998457874392e-05, "loss": 0.0431, "step": 4020 }, { "grad_norm": 0.3908650875091553, "learning_rate": 9.999996407402913e-05, "loss": 0.0486, "step": 4030 }, { "grad_norm": 0.39945098757743835, "learning_rate": 9.999993502568801e-05, "loss": 0.0422, "step": 4040 }, { "grad_norm": 0.6500047445297241, "learning_rate": 9.999989743372548e-05, "loss": 0.0474, "step": 4050 }, { "grad_norm": 0.42788901925086975, "learning_rate": 9.999985129814798e-05, "loss": 0.0456, "step": 4060 }, { "grad_norm": 0.3542851507663727, "learning_rate": 9.99997966189634e-05, "loss": 0.0389, "step": 4070 }, { "grad_norm": 0.38482391834259033, "learning_rate": 9.999973339618107e-05, "loss": 0.0437, "step": 4080 }, { "grad_norm": 0.5474271774291992, "learning_rate": 9.999966162981179e-05, "loss": 0.0424, "step": 4090 }, { "grad_norm": 0.4295892119407654, "learning_rate": 9.999958131986784e-05, "loss": 0.0405, "step": 4100 }, { "grad_norm": 0.45272096991539, "learning_rate": 9.999949246636293e-05, "loss": 0.0483, "step": 4110 }, { "grad_norm": 0.48269394040107727, "learning_rate": 9.999939506931224e-05, "loss": 0.048, "step": 4120 }, { "grad_norm": 0.6036842465400696, "learning_rate": 9.999928912873243e-05, "loss": 0.0503, "step": 4130 }, { "grad_norm": 0.5376411080360413, "learning_rate": 9.999917464464159e-05, "loss": 0.0439, "step": 4140 }, { "grad_norm": 0.5487960577011108, "learning_rate": 9.999905161705929e-05, "loss": 0.0472, "step": 4150 }, { "grad_norm": 0.41787075996398926, "learning_rate": 9.999892004600653e-05, "loss": 0.0439, "step": 4160 }, { "grad_norm": 0.4618934690952301, "learning_rate": 9.999877993150581e-05, "loss": 0.0461, "step": 4170 }, { "grad_norm": 0.34073442220687866, "learning_rate": 9.999863127358108e-05, "loss": 0.0454, "step": 4180 }, { "grad_norm": 0.4129701256752014, "learning_rate": 9.999847407225773e-05, "loss": 0.0474, "step": 4190 }, { "grad_norm": 0.3401712477207184, "learning_rate": 9.999830832756262e-05, "loss": 0.04, "step": 4200 }, { "grad_norm": 0.35181915760040283, "learning_rate": 9.999813403952407e-05, "loss": 0.0421, "step": 4210 }, { "grad_norm": 0.5450085401535034, "learning_rate": 9.999795120817187e-05, "loss": 0.0426, "step": 4220 }, { "grad_norm": 0.3971133828163147, "learning_rate": 9.999775983353725e-05, "loss": 0.0423, "step": 4230 }, { "grad_norm": 0.4853639006614685, "learning_rate": 9.999755991565292e-05, "loss": 0.042, "step": 4240 }, { "grad_norm": 0.4363464117050171, "learning_rate": 9.999735145455303e-05, "loss": 0.0454, "step": 4250 }, { "grad_norm": 0.39583849906921387, "learning_rate": 9.99971344502732e-05, "loss": 0.0445, "step": 4260 }, { "grad_norm": 0.4322342872619629, "learning_rate": 9.999690890285053e-05, "loss": 0.0433, "step": 4270 }, { "grad_norm": 0.6804735660552979, "learning_rate": 9.999667481232356e-05, "loss": 0.0465, "step": 4280 }, { "grad_norm": 0.354281485080719, "learning_rate": 9.999643217873225e-05, "loss": 0.0397, "step": 4290 }, { "grad_norm": 0.36661094427108765, "learning_rate": 9.999618100211809e-05, "loss": 0.0382, "step": 4300 }, { "grad_norm": 0.5432329177856445, "learning_rate": 9.999592128252402e-05, "loss": 0.0402, "step": 4310 }, { "grad_norm": 0.3780505061149597, "learning_rate": 9.999565301999437e-05, "loss": 0.0444, "step": 4320 }, { "grad_norm": 0.4568091034889221, "learning_rate": 9.999537621457502e-05, "loss": 0.042, "step": 4330 }, { "grad_norm": 0.3890150487422943, "learning_rate": 9.999509086631323e-05, "loss": 0.0435, "step": 4340 }, { "grad_norm": 0.3866572082042694, "learning_rate": 9.99947969752578e-05, "loss": 0.0481, "step": 4350 }, { "grad_norm": 0.4701360762119293, "learning_rate": 9.999449454145891e-05, "loss": 0.0417, "step": 4360 }, { "grad_norm": 0.4596754312515259, "learning_rate": 9.999418356496827e-05, "loss": 0.0438, "step": 4370 }, { "grad_norm": 0.43898117542266846, "learning_rate": 9.999386404583899e-05, "loss": 0.0474, "step": 4380 }, { "grad_norm": 0.36950942873954773, "learning_rate": 9.999353598412568e-05, "loss": 0.0497, "step": 4390 }, { "grad_norm": 0.5559619665145874, "learning_rate": 9.999319937988442e-05, "loss": 0.0426, "step": 4400 }, { "grad_norm": 0.38906940817832947, "learning_rate": 9.999285423317268e-05, "loss": 0.0428, "step": 4410 }, { "grad_norm": 0.3974049985408783, "learning_rate": 9.999250054404947e-05, "loss": 0.0467, "step": 4420 }, { "grad_norm": 0.3877277374267578, "learning_rate": 9.99921383125752e-05, "loss": 0.0398, "step": 4430 }, { "grad_norm": 0.3927239179611206, "learning_rate": 9.99917675388118e-05, "loss": 0.0422, "step": 4440 }, { "grad_norm": 0.411536306142807, "learning_rate": 9.99913882228226e-05, "loss": 0.0439, "step": 4450 }, { "grad_norm": 0.3637375235557556, "learning_rate": 9.999100036467242e-05, "loss": 0.0429, "step": 4460 }, { "grad_norm": 0.4128304719924927, "learning_rate": 9.999060396442753e-05, "loss": 0.0413, "step": 4470 }, { "grad_norm": 0.38868555426597595, "learning_rate": 9.999019902215566e-05, "loss": 0.0431, "step": 4480 }, { "grad_norm": 0.44928178191185, "learning_rate": 9.998978553792602e-05, "loss": 0.047, "step": 4490 }, { "grad_norm": 0.2985336482524872, "learning_rate": 9.998936351180926e-05, "loss": 0.0399, "step": 4500 }, { "grad_norm": 0.4610495865345001, "learning_rate": 9.998893294387747e-05, "loss": 0.0381, "step": 4510 }, { "grad_norm": 0.4093915820121765, "learning_rate": 9.998849383420426e-05, "loss": 0.0409, "step": 4520 }, { "grad_norm": 0.4313364028930664, "learning_rate": 9.998804618286465e-05, "loss": 0.0438, "step": 4530 }, { "grad_norm": 0.5056940317153931, "learning_rate": 9.99875899899351e-05, "loss": 0.0429, "step": 4540 }, { "grad_norm": 0.4750808775424957, "learning_rate": 9.99871252554936e-05, "loss": 0.0385, "step": 4550 }, { "grad_norm": 0.43604958057403564, "learning_rate": 9.998665197961955e-05, "loss": 0.0411, "step": 4560 }, { "grad_norm": 0.31070080399513245, "learning_rate": 9.998617016239379e-05, "loss": 0.0393, "step": 4570 }, { "grad_norm": 0.36653444170951843, "learning_rate": 9.998567980389869e-05, "loss": 0.0452, "step": 4580 }, { "grad_norm": 0.43733397126197815, "learning_rate": 9.998518090421802e-05, "loss": 0.047, "step": 4590 }, { "grad_norm": 0.38917210698127747, "learning_rate": 9.998467346343703e-05, "loss": 0.0396, "step": 4600 }, { "grad_norm": 0.5385326743125916, "learning_rate": 9.998415748164243e-05, "loss": 0.0449, "step": 4610 }, { "grad_norm": 0.3539160192012787, "learning_rate": 9.998363295892238e-05, "loss": 0.0406, "step": 4620 }, { "grad_norm": 0.3558367192745209, "learning_rate": 9.998309989536652e-05, "loss": 0.0392, "step": 4630 }, { "grad_norm": 0.3908977508544922, "learning_rate": 9.998255829106593e-05, "loss": 0.044, "step": 4640 }, { "grad_norm": 0.3682451546192169, "learning_rate": 9.998200814611316e-05, "loss": 0.0378, "step": 4650 }, { "grad_norm": 0.3557581901550293, "learning_rate": 9.998144946060219e-05, "loss": 0.0395, "step": 4660 }, { "grad_norm": 0.4224003553390503, "learning_rate": 9.998088223462852e-05, "loss": 0.0373, "step": 4670 }, { "grad_norm": 0.45699042081832886, "learning_rate": 9.998030646828905e-05, "loss": 0.0375, "step": 4680 }, { "grad_norm": 0.49147486686706543, "learning_rate": 9.997972216168217e-05, "loss": 0.0412, "step": 4690 }, { "grad_norm": 0.4647693634033203, "learning_rate": 9.997912931490771e-05, "loss": 0.0488, "step": 4700 }, { "grad_norm": 0.381600558757782, "learning_rate": 9.9978527928067e-05, "loss": 0.0409, "step": 4710 }, { "grad_norm": 0.45850738883018494, "learning_rate": 9.997791800126277e-05, "loss": 0.0363, "step": 4720 }, { "grad_norm": 0.3741294741630554, "learning_rate": 9.997729953459927e-05, "loss": 0.0393, "step": 4730 }, { "grad_norm": 0.32280513644218445, "learning_rate": 9.997667252818214e-05, "loss": 0.0406, "step": 4740 }, { "grad_norm": 0.42173606157302856, "learning_rate": 9.997603698211855e-05, "loss": 0.0382, "step": 4750 }, { "grad_norm": 0.3607245981693268, "learning_rate": 9.99753928965171e-05, "loss": 0.0414, "step": 4760 }, { "grad_norm": 0.3502008020877838, "learning_rate": 9.997474027148781e-05, "loss": 0.0441, "step": 4770 }, { "grad_norm": 0.373586505651474, "learning_rate": 9.997407910714223e-05, "loss": 0.039, "step": 4780 }, { "grad_norm": 0.3898361325263977, "learning_rate": 9.997340940359332e-05, "loss": 0.042, "step": 4790 }, { "grad_norm": 0.4134279787540436, "learning_rate": 9.997273116095552e-05, "loss": 0.0394, "step": 4800 }, { "grad_norm": 0.4778191149234772, "learning_rate": 9.997204437934473e-05, "loss": 0.0371, "step": 4810 }, { "grad_norm": 0.43765687942504883, "learning_rate": 9.997134905887829e-05, "loss": 0.0397, "step": 4820 }, { "grad_norm": 0.2678273320198059, "learning_rate": 9.997064519967501e-05, "loss": 0.0378, "step": 4830 }, { "grad_norm": 0.36056452989578247, "learning_rate": 9.996993280185517e-05, "loss": 0.0392, "step": 4840 }, { "grad_norm": 0.3759630620479584, "learning_rate": 9.99692118655405e-05, "loss": 0.0394, "step": 4850 }, { "grad_norm": 0.4456116557121277, "learning_rate": 9.996848239085417e-05, "loss": 0.0351, "step": 4860 }, { "grad_norm": 0.35977768898010254, "learning_rate": 9.996774437792085e-05, "loss": 0.0426, "step": 4870 }, { "grad_norm": 0.4139716923236847, "learning_rate": 9.996699782686664e-05, "loss": 0.039, "step": 4880 }, { "grad_norm": 0.3568178117275238, "learning_rate": 9.996624273781909e-05, "loss": 0.0395, "step": 4890 }, { "grad_norm": 0.392555832862854, "learning_rate": 9.996547911090725e-05, "loss": 0.0393, "step": 4900 }, { "grad_norm": 0.2299443483352661, "learning_rate": 9.996470694626157e-05, "loss": 0.0361, "step": 4910 }, { "grad_norm": 0.33831363916397095, "learning_rate": 9.996392624401403e-05, "loss": 0.0403, "step": 4920 }, { "grad_norm": 0.33414116501808167, "learning_rate": 9.996313700429801e-05, "loss": 0.0397, "step": 4930 }, { "grad_norm": 0.42402130365371704, "learning_rate": 9.996233922724836e-05, "loss": 0.0336, "step": 4940 }, { "grad_norm": 0.43308570981025696, "learning_rate": 9.996153291300141e-05, "loss": 0.0385, "step": 4950 }, { "grad_norm": 0.2647453248500824, "learning_rate": 9.996071806169494e-05, "loss": 0.0409, "step": 4960 }, { "grad_norm": 0.33003202080726624, "learning_rate": 9.995989467346817e-05, "loss": 0.0328, "step": 4970 }, { "grad_norm": 0.334346741437912, "learning_rate": 9.995906274846183e-05, "loss": 0.0432, "step": 4980 }, { "grad_norm": 0.31773364543914795, "learning_rate": 9.995822228681803e-05, "loss": 0.0396, "step": 4990 }, { "grad_norm": 0.343639999628067, "learning_rate": 9.99573732886804e-05, "loss": 0.0384, "step": 5000 }, { "grad_norm": 0.4441509544849396, "learning_rate": 9.995651575419402e-05, "loss": 0.0403, "step": 5010 }, { "grad_norm": 0.42626649141311646, "learning_rate": 9.995564968350541e-05, "loss": 0.0427, "step": 5020 }, { "grad_norm": 0.3693671226501465, "learning_rate": 9.995477507676256e-05, "loss": 0.0351, "step": 5030 }, { "grad_norm": 0.3732265830039978, "learning_rate": 9.995389193411493e-05, "loss": 0.0337, "step": 5040 }, { "grad_norm": 0.293202668428421, "learning_rate": 9.995300025571339e-05, "loss": 0.0378, "step": 5050 }, { "grad_norm": 0.3907698690891266, "learning_rate": 9.995210004171034e-05, "loss": 0.0409, "step": 5060 }, { "grad_norm": 0.3382211923599243, "learning_rate": 9.995119129225956e-05, "loss": 0.0377, "step": 5070 }, { "grad_norm": 0.39819952845573425, "learning_rate": 9.995027400751637e-05, "loss": 0.0366, "step": 5080 }, { "grad_norm": 0.3779410123825073, "learning_rate": 9.994934818763751e-05, "loss": 0.0413, "step": 5090 }, { "grad_norm": 0.4228343069553375, "learning_rate": 9.994841383278115e-05, "loss": 0.0407, "step": 5100 }, { "grad_norm": 0.3468967378139496, "learning_rate": 9.994747094310695e-05, "loss": 0.04, "step": 5110 }, { "grad_norm": 0.34337660670280457, "learning_rate": 9.994651951877604e-05, "loss": 0.035, "step": 5120 }, { "grad_norm": 0.298493891954422, "learning_rate": 9.994555955995099e-05, "loss": 0.0351, "step": 5130 }, { "grad_norm": 0.4798490107059479, "learning_rate": 9.994459106679581e-05, "loss": 0.0376, "step": 5140 }, { "grad_norm": 0.3636314272880554, "learning_rate": 9.994361403947603e-05, "loss": 0.0385, "step": 5150 }, { "grad_norm": 0.3704986572265625, "learning_rate": 9.994262847815854e-05, "loss": 0.0389, "step": 5160 }, { "grad_norm": 0.3189834654331207, "learning_rate": 9.99416343830118e-05, "loss": 0.0351, "step": 5170 }, { "grad_norm": 0.3813282251358032, "learning_rate": 9.994063175420565e-05, "loss": 0.0339, "step": 5180 }, { "grad_norm": 0.43352627754211426, "learning_rate": 9.99396205919114e-05, "loss": 0.0395, "step": 5190 }, { "grad_norm": 0.3573223948478699, "learning_rate": 9.993860089630185e-05, "loss": 0.0361, "step": 5200 }, { "grad_norm": 0.4510124921798706, "learning_rate": 9.993757266755123e-05, "loss": 0.0443, "step": 5210 }, { "grad_norm": 0.5565505027770996, "learning_rate": 9.993653590583522e-05, "loss": 0.0463, "step": 5220 }, { "grad_norm": 0.33148685097694397, "learning_rate": 9.993549061133102e-05, "loss": 0.0397, "step": 5230 }, { "grad_norm": 0.37936991453170776, "learning_rate": 9.993443678421719e-05, "loss": 0.0351, "step": 5240 }, { "grad_norm": 0.37209123373031616, "learning_rate": 9.993337442467384e-05, "loss": 0.0389, "step": 5250 }, { "grad_norm": 0.5182049870491028, "learning_rate": 9.993230353288248e-05, "loss": 0.0395, "step": 5260 }, { "grad_norm": 0.4582339823246002, "learning_rate": 9.993122410902608e-05, "loss": 0.039, "step": 5270 }, { "grad_norm": 0.29979607462882996, "learning_rate": 9.993013615328912e-05, "loss": 0.0381, "step": 5280 }, { "grad_norm": 0.3988814949989319, "learning_rate": 9.992903966585747e-05, "loss": 0.0391, "step": 5290 }, { "grad_norm": 0.4986861050128937, "learning_rate": 9.992793464691852e-05, "loss": 0.0422, "step": 5300 }, { "grad_norm": 0.3286103904247284, "learning_rate": 9.992682109666105e-05, "loss": 0.0437, "step": 5310 }, { "grad_norm": 0.3437652289867401, "learning_rate": 9.992569901527538e-05, "loss": 0.0454, "step": 5320 }, { "grad_norm": 0.31820130348205566, "learning_rate": 9.99245684029532e-05, "loss": 0.0375, "step": 5330 }, { "grad_norm": 0.3070760667324066, "learning_rate": 9.992342925988774e-05, "loss": 0.0394, "step": 5340 }, { "grad_norm": 0.31057488918304443, "learning_rate": 9.992228158627361e-05, "loss": 0.0397, "step": 5350 }, { "grad_norm": 0.3440532088279724, "learning_rate": 9.992112538230693e-05, "loss": 0.0346, "step": 5360 }, { "grad_norm": 0.4492149353027344, "learning_rate": 9.991996064818527e-05, "loss": 0.0356, "step": 5370 }, { "grad_norm": 0.3624633252620697, "learning_rate": 9.991878738410768e-05, "loss": 0.0359, "step": 5380 }, { "grad_norm": 0.353118896484375, "learning_rate": 9.991760559027457e-05, "loss": 0.0358, "step": 5390 }, { "grad_norm": 0.33083733916282654, "learning_rate": 9.991641526688793e-05, "loss": 0.0334, "step": 5400 }, { "grad_norm": 0.352997750043869, "learning_rate": 9.991521641415113e-05, "loss": 0.0368, "step": 5410 }, { "grad_norm": 0.23542548716068268, "learning_rate": 9.991400903226904e-05, "loss": 0.0343, "step": 5420 }, { "grad_norm": 0.451647013425827, "learning_rate": 9.991279312144794e-05, "loss": 0.039, "step": 5430 }, { "grad_norm": 0.38292258977890015, "learning_rate": 9.991156868189564e-05, "loss": 0.0386, "step": 5440 }, { "grad_norm": 0.3513263463973999, "learning_rate": 9.991033571382131e-05, "loss": 0.0381, "step": 5450 }, { "grad_norm": 0.36475735902786255, "learning_rate": 9.990909421743569e-05, "loss": 0.0377, "step": 5460 }, { "grad_norm": 0.42746368050575256, "learning_rate": 9.990784419295085e-05, "loss": 0.0357, "step": 5470 }, { "grad_norm": 0.3543992340564728, "learning_rate": 9.990658564058044e-05, "loss": 0.0389, "step": 5480 }, { "grad_norm": 0.29717639088630676, "learning_rate": 9.990531856053948e-05, "loss": 0.0388, "step": 5490 }, { "grad_norm": 0.3424909710884094, "learning_rate": 9.99040429530445e-05, "loss": 0.0351, "step": 5500 }, { "grad_norm": 0.32598212361335754, "learning_rate": 9.990275881831346e-05, "loss": 0.0372, "step": 5510 }, { "grad_norm": 0.3438681364059448, "learning_rate": 9.990146615656577e-05, "loss": 0.0388, "step": 5520 }, { "grad_norm": 0.4006127417087555, "learning_rate": 9.990016496802233e-05, "loss": 0.0369, "step": 5530 }, { "grad_norm": 0.3567691743373871, "learning_rate": 9.989885525290548e-05, "loss": 0.0354, "step": 5540 }, { "grad_norm": 0.3159230649471283, "learning_rate": 9.989753701143897e-05, "loss": 0.0365, "step": 5550 }, { "grad_norm": 0.46280181407928467, "learning_rate": 9.989621024384812e-05, "loss": 0.0361, "step": 5560 }, { "grad_norm": 0.3141930103302002, "learning_rate": 9.989487495035959e-05, "loss": 0.0375, "step": 5570 }, { "grad_norm": 0.47234684228897095, "learning_rate": 9.989353113120156e-05, "loss": 0.0338, "step": 5580 }, { "grad_norm": 0.3736853301525116, "learning_rate": 9.989217878660366e-05, "loss": 0.038, "step": 5590 }, { "grad_norm": 0.33914294838905334, "learning_rate": 9.989081791679695e-05, "loss": 0.0351, "step": 5600 }, { "grad_norm": 0.27024826407432556, "learning_rate": 9.988944852201397e-05, "loss": 0.0345, "step": 5610 }, { "grad_norm": 0.3808518350124359, "learning_rate": 9.988807060248873e-05, "loss": 0.0355, "step": 5620 }, { "grad_norm": 0.3226263225078583, "learning_rate": 9.988668415845665e-05, "loss": 0.0319, "step": 5630 }, { "grad_norm": 0.2879692316055298, "learning_rate": 9.988528919015466e-05, "loss": 0.0331, "step": 5640 }, { "grad_norm": 0.3132660686969757, "learning_rate": 9.988388569782112e-05, "loss": 0.0342, "step": 5650 }, { "grad_norm": 0.34015023708343506, "learning_rate": 9.988247368169583e-05, "loss": 0.0432, "step": 5660 }, { "grad_norm": 0.3346807360649109, "learning_rate": 9.988105314202007e-05, "loss": 0.0349, "step": 5670 }, { "grad_norm": 0.38133320212364197, "learning_rate": 9.987962407903659e-05, "loss": 0.0381, "step": 5680 }, { "grad_norm": 0.3601982295513153, "learning_rate": 9.987818649298957e-05, "loss": 0.0319, "step": 5690 }, { "grad_norm": 0.37653425335884094, "learning_rate": 9.987674038412465e-05, "loss": 0.0368, "step": 5700 }, { "grad_norm": 0.43179813027381897, "learning_rate": 9.987528575268891e-05, "loss": 0.0332, "step": 5710 }, { "grad_norm": 0.3311007618904114, "learning_rate": 9.987382259893095e-05, "loss": 0.0354, "step": 5720 }, { "grad_norm": 0.3272186815738678, "learning_rate": 9.987235092310074e-05, "loss": 0.035, "step": 5730 }, { "grad_norm": 0.27053195238113403, "learning_rate": 9.987087072544978e-05, "loss": 0.0374, "step": 5740 }, { "grad_norm": 0.3311319351196289, "learning_rate": 9.9869382006231e-05, "loss": 0.0369, "step": 5750 }, { "grad_norm": 0.4038267135620117, "learning_rate": 9.986788476569875e-05, "loss": 0.033, "step": 5760 }, { "grad_norm": 0.39773479104042053, "learning_rate": 9.986637900410887e-05, "loss": 0.0389, "step": 5770 }, { "grad_norm": 0.36153969168663025, "learning_rate": 9.986486472171869e-05, "loss": 0.0352, "step": 5780 }, { "grad_norm": 0.384724885225296, "learning_rate": 9.986334191878692e-05, "loss": 0.0378, "step": 5790 }, { "grad_norm": 0.37608346343040466, "learning_rate": 9.986181059557378e-05, "loss": 0.0367, "step": 5800 }, { "grad_norm": 0.2913546562194824, "learning_rate": 9.986027075234094e-05, "loss": 0.0329, "step": 5810 }, { "grad_norm": 0.2533680200576782, "learning_rate": 9.985872238935152e-05, "loss": 0.0346, "step": 5820 }, { "grad_norm": 0.34604325890541077, "learning_rate": 9.985716550687008e-05, "loss": 0.0369, "step": 5830 }, { "grad_norm": 0.41734206676483154, "learning_rate": 9.985560010516264e-05, "loss": 0.0365, "step": 5840 }, { "grad_norm": 0.33128786087036133, "learning_rate": 9.985402618449668e-05, "loss": 0.0355, "step": 5850 }, { "grad_norm": 0.39783960580825806, "learning_rate": 9.985244374514118e-05, "loss": 0.0389, "step": 5860 }, { "grad_norm": 0.44943127036094666, "learning_rate": 9.985085278736651e-05, "loss": 0.0416, "step": 5870 }, { "grad_norm": 0.32332780957221985, "learning_rate": 9.984925331144452e-05, "loss": 0.0365, "step": 5880 }, { "grad_norm": 0.3049481213092804, "learning_rate": 9.984764531764851e-05, "loss": 0.0312, "step": 5890 }, { "grad_norm": 0.3635876178741455, "learning_rate": 9.984602880625326e-05, "loss": 0.0356, "step": 5900 }, { "grad_norm": 0.4217691123485565, "learning_rate": 9.9844403777535e-05, "loss": 0.037, "step": 5910 }, { "grad_norm": 0.43595367670059204, "learning_rate": 9.984277023177135e-05, "loss": 0.0353, "step": 5920 }, { "grad_norm": 0.282981276512146, "learning_rate": 9.984112816924148e-05, "loss": 0.0319, "step": 5930 }, { "grad_norm": 0.30289626121520996, "learning_rate": 9.983947759022596e-05, "loss": 0.0356, "step": 5940 }, { "grad_norm": 0.3619740605354309, "learning_rate": 9.983781849500682e-05, "loss": 0.0331, "step": 5950 }, { "grad_norm": 0.3639616072177887, "learning_rate": 9.98361508838676e-05, "loss": 0.0329, "step": 5960 }, { "grad_norm": 0.3376958668231964, "learning_rate": 9.98344747570932e-05, "loss": 0.0361, "step": 5970 }, { "grad_norm": 0.26860982179641724, "learning_rate": 9.983279011497004e-05, "loss": 0.0332, "step": 5980 }, { "grad_norm": 0.34877878427505493, "learning_rate": 9.983109695778596e-05, "loss": 0.0371, "step": 5990 }, { "grad_norm": 0.34162575006484985, "learning_rate": 9.982939528583032e-05, "loss": 0.0366, "step": 6000 }, { "grad_norm": 0.322662353515625, "learning_rate": 9.982768509939385e-05, "loss": 0.0321, "step": 6010 }, { "grad_norm": 0.2549271583557129, "learning_rate": 9.982596639876879e-05, "loss": 0.0322, "step": 6020 }, { "grad_norm": 0.3829749524593353, "learning_rate": 9.982423918424881e-05, "loss": 0.0293, "step": 6030 }, { "grad_norm": 0.3329724669456482, "learning_rate": 9.982250345612908e-05, "loss": 0.0356, "step": 6040 }, { "grad_norm": 0.34320077300071716, "learning_rate": 9.982075921470611e-05, "loss": 0.0336, "step": 6050 }, { "grad_norm": 0.34278684854507446, "learning_rate": 9.981900646027802e-05, "loss": 0.0311, "step": 6060 }, { "grad_norm": 0.3491262197494507, "learning_rate": 9.981724519314425e-05, "loss": 0.0349, "step": 6070 }, { "grad_norm": 0.29859310388565063, "learning_rate": 9.981547541360581e-05, "loss": 0.036, "step": 6080 }, { "grad_norm": 0.3209095597267151, "learning_rate": 9.981369712196508e-05, "loss": 0.0296, "step": 6090 }, { "grad_norm": 0.3610628843307495, "learning_rate": 9.981191031852592e-05, "loss": 0.0396, "step": 6100 }, { "grad_norm": 0.33447033166885376, "learning_rate": 9.981011500359362e-05, "loss": 0.0344, "step": 6110 }, { "grad_norm": 0.338122695684433, "learning_rate": 9.9808311177475e-05, "loss": 0.0405, "step": 6120 }, { "grad_norm": 0.35023677349090576, "learning_rate": 9.980649884047826e-05, "loss": 0.0336, "step": 6130 }, { "grad_norm": 0.3216910660266876, "learning_rate": 9.980467799291307e-05, "loss": 0.0337, "step": 6140 }, { "grad_norm": 0.3336784541606903, "learning_rate": 9.980284863509058e-05, "loss": 0.0393, "step": 6150 }, { "grad_norm": 0.2923177182674408, "learning_rate": 9.980101076732334e-05, "loss": 0.035, "step": 6160 }, { "grad_norm": 0.3564910292625427, "learning_rate": 9.979916438992544e-05, "loss": 0.0377, "step": 6170 }, { "grad_norm": 0.3874610364437103, "learning_rate": 9.979730950321237e-05, "loss": 0.0362, "step": 6180 }, { "grad_norm": 0.30740854144096375, "learning_rate": 9.979544610750104e-05, "loss": 0.0327, "step": 6190 }, { "grad_norm": 0.32619091868400574, "learning_rate": 9.97935742031099e-05, "loss": 0.0341, "step": 6200 }, { "grad_norm": 0.32695430517196655, "learning_rate": 9.979169379035878e-05, "loss": 0.0344, "step": 6210 }, { "grad_norm": 0.33780378103256226, "learning_rate": 9.978980486956899e-05, "loss": 0.0382, "step": 6220 }, { "grad_norm": 0.24305669963359833, "learning_rate": 9.978790744106332e-05, "loss": 0.0364, "step": 6230 }, { "grad_norm": 0.3300056457519531, "learning_rate": 9.978600150516594e-05, "loss": 0.0338, "step": 6240 }, { "grad_norm": 0.33591482043266296, "learning_rate": 9.978408706220259e-05, "loss": 0.0392, "step": 6250 }, { "grad_norm": 0.36712411046028137, "learning_rate": 9.978216411250032e-05, "loss": 0.0384, "step": 6260 }, { "grad_norm": 0.2993386387825012, "learning_rate": 9.978023265638778e-05, "loss": 0.0316, "step": 6270 }, { "grad_norm": 0.3726228177547455, "learning_rate": 9.977829269419495e-05, "loss": 0.0357, "step": 6280 }, { "grad_norm": 0.26372361183166504, "learning_rate": 9.977634422625335e-05, "loss": 0.0326, "step": 6290 }, { "grad_norm": 0.3961373567581177, "learning_rate": 9.97743872528959e-05, "loss": 0.0389, "step": 6300 }, { "grad_norm": 0.27624598145484924, "learning_rate": 9.9772421774457e-05, "loss": 0.0361, "step": 6310 }, { "grad_norm": 0.3622264564037323, "learning_rate": 9.977044779127252e-05, "loss": 0.0345, "step": 6320 }, { "grad_norm": 0.3493199348449707, "learning_rate": 9.976846530367971e-05, "loss": 0.035, "step": 6330 }, { "grad_norm": 0.3156735599040985, "learning_rate": 9.976647431201735e-05, "loss": 0.0358, "step": 6340 }, { "grad_norm": 0.3431600034236908, "learning_rate": 9.976447481662568e-05, "loss": 0.0347, "step": 6350 }, { "grad_norm": 0.29227200150489807, "learning_rate": 9.976246681784629e-05, "loss": 0.0369, "step": 6360 }, { "grad_norm": 0.4043862521648407, "learning_rate": 9.976045031602234e-05, "loss": 0.0366, "step": 6370 }, { "grad_norm": 0.30286914110183716, "learning_rate": 9.975842531149837e-05, "loss": 0.037, "step": 6380 }, { "grad_norm": 0.4056163728237152, "learning_rate": 9.975639180462043e-05, "loss": 0.0357, "step": 6390 }, { "grad_norm": 0.41694116592407227, "learning_rate": 9.975434979573596e-05, "loss": 0.0333, "step": 6400 }, { "grad_norm": 0.34024643898010254, "learning_rate": 9.97522992851939e-05, "loss": 0.0386, "step": 6410 }, { "grad_norm": 0.4286046028137207, "learning_rate": 9.975024027334461e-05, "loss": 0.0369, "step": 6420 }, { "grad_norm": 0.37317749857902527, "learning_rate": 9.974817276053993e-05, "loss": 0.0339, "step": 6430 }, { "grad_norm": 0.3410826623439789, "learning_rate": 9.974609674713315e-05, "loss": 0.0329, "step": 6440 }, { "grad_norm": 0.34463635087013245, "learning_rate": 9.9744012233479e-05, "loss": 0.0329, "step": 6450 }, { "grad_norm": 0.3897024393081665, "learning_rate": 9.974191921993366e-05, "loss": 0.0331, "step": 6460 }, { "grad_norm": 0.3651590943336487, "learning_rate": 9.973981770685474e-05, "loss": 0.0336, "step": 6470 }, { "grad_norm": 0.31610551476478577, "learning_rate": 9.97377076946014e-05, "loss": 0.0366, "step": 6480 }, { "grad_norm": 0.30003824830055237, "learning_rate": 9.973558918353412e-05, "loss": 0.0365, "step": 6490 }, { "grad_norm": 0.29989054799079895, "learning_rate": 9.973346217401494e-05, "loss": 0.0325, "step": 6500 }, { "grad_norm": 0.29656538367271423, "learning_rate": 9.973132666640726e-05, "loss": 0.0317, "step": 6510 }, { "grad_norm": 0.3443485498428345, "learning_rate": 9.972918266107602e-05, "loss": 0.0314, "step": 6520 }, { "grad_norm": 0.38496634364128113, "learning_rate": 9.972703015838756e-05, "loss": 0.0329, "step": 6530 }, { "grad_norm": 0.3501671254634857, "learning_rate": 9.97248691587097e-05, "loss": 0.0323, "step": 6540 }, { "grad_norm": 0.3363714814186096, "learning_rate": 9.972269966241166e-05, "loss": 0.0364, "step": 6550 }, { "grad_norm": 0.3181905448436737, "learning_rate": 9.972052166986417e-05, "loss": 0.0359, "step": 6560 }, { "grad_norm": 0.28684002161026, "learning_rate": 9.971833518143938e-05, "loss": 0.0356, "step": 6570 }, { "grad_norm": 0.29322120547294617, "learning_rate": 9.971614019751093e-05, "loss": 0.03, "step": 6580 }, { "grad_norm": 0.3607814610004425, "learning_rate": 9.971393671845383e-05, "loss": 0.0321, "step": 6590 }, { "grad_norm": 0.2465900480747223, "learning_rate": 9.971172474464464e-05, "loss": 0.037, "step": 6600 }, { "grad_norm": 0.23347820341587067, "learning_rate": 9.97095042764613e-05, "loss": 0.0383, "step": 6610 }, { "grad_norm": 0.3259611129760742, "learning_rate": 9.970727531428324e-05, "loss": 0.0276, "step": 6620 }, { "grad_norm": 0.26180151104927063, "learning_rate": 9.970503785849132e-05, "loss": 0.0395, "step": 6630 }, { "grad_norm": 0.2953731417655945, "learning_rate": 9.970279190946788e-05, "loss": 0.0346, "step": 6640 }, { "grad_norm": 0.31294429302215576, "learning_rate": 9.970053746759667e-05, "loss": 0.032, "step": 6650 }, { "grad_norm": 0.2886003851890564, "learning_rate": 9.969827453326292e-05, "loss": 0.0351, "step": 6660 }, { "grad_norm": 0.25583508610725403, "learning_rate": 9.969600310685332e-05, "loss": 0.0349, "step": 6670 }, { "grad_norm": 0.2873946726322174, "learning_rate": 9.969372318875596e-05, "loss": 0.0351, "step": 6680 }, { "grad_norm": 0.3436180353164673, "learning_rate": 9.969143477936043e-05, "loss": 0.0345, "step": 6690 }, { "grad_norm": 0.32176849246025085, "learning_rate": 9.968913787905775e-05, "loss": 0.0343, "step": 6700 }, { "grad_norm": 0.334466814994812, "learning_rate": 9.968683248824045e-05, "loss": 0.035, "step": 6710 }, { "grad_norm": 0.28519490361213684, "learning_rate": 9.968451860730238e-05, "loss": 0.0336, "step": 6720 }, { "grad_norm": 0.42979928851127625, "learning_rate": 9.968219623663896e-05, "loss": 0.0318, "step": 6730 }, { "grad_norm": 0.3103415369987488, "learning_rate": 9.967986537664702e-05, "loss": 0.0328, "step": 6740 }, { "grad_norm": 0.33410847187042236, "learning_rate": 9.967752602772483e-05, "loss": 0.031, "step": 6750 }, { "grad_norm": 0.4520779252052307, "learning_rate": 9.967517819027212e-05, "loss": 0.0311, "step": 6760 }, { "grad_norm": 0.3019709587097168, "learning_rate": 9.967282186469009e-05, "loss": 0.0298, "step": 6770 }, { "grad_norm": 0.32873353362083435, "learning_rate": 9.967045705138135e-05, "loss": 0.033, "step": 6780 }, { "grad_norm": 0.26213502883911133, "learning_rate": 9.966808375074998e-05, "loss": 0.0331, "step": 6790 }, { "grad_norm": 0.33900928497314453, "learning_rate": 9.966570196320154e-05, "loss": 0.0315, "step": 6800 }, { "grad_norm": 0.4208446443080902, "learning_rate": 9.966331168914299e-05, "loss": 0.0336, "step": 6810 }, { "grad_norm": 0.36674705147743225, "learning_rate": 9.966091292898277e-05, "loss": 0.0355, "step": 6820 }, { "grad_norm": 0.3064243495464325, "learning_rate": 9.965850568313076e-05, "loss": 0.0298, "step": 6830 }, { "grad_norm": 0.2821328938007355, "learning_rate": 9.965608995199827e-05, "loss": 0.0324, "step": 6840 }, { "grad_norm": 0.3195745646953583, "learning_rate": 9.965366573599812e-05, "loss": 0.0349, "step": 6850 }, { "grad_norm": 0.32105785608291626, "learning_rate": 9.965123303554453e-05, "loss": 0.0361, "step": 6860 }, { "grad_norm": 0.2701222598552704, "learning_rate": 9.964879185105317e-05, "loss": 0.0304, "step": 6870 }, { "grad_norm": 0.2757580280303955, "learning_rate": 9.964634218294119e-05, "loss": 0.0365, "step": 6880 }, { "grad_norm": 0.26895761489868164, "learning_rate": 9.964388403162714e-05, "loss": 0.0277, "step": 6890 }, { "grad_norm": 0.29643258452415466, "learning_rate": 9.96414173975311e-05, "loss": 0.031, "step": 6900 }, { "grad_norm": 0.2892759144306183, "learning_rate": 9.963894228107451e-05, "loss": 0.0284, "step": 6910 }, { "grad_norm": 0.36054250597953796, "learning_rate": 9.963645868268032e-05, "loss": 0.0325, "step": 6920 }, { "grad_norm": 0.28795769810676575, "learning_rate": 9.963396660277289e-05, "loss": 0.0334, "step": 6930 }, { "grad_norm": 0.3308853507041931, "learning_rate": 9.963146604177807e-05, "loss": 0.0313, "step": 6940 }, { "grad_norm": 0.31371352076530457, "learning_rate": 9.962895700012311e-05, "loss": 0.0323, "step": 6950 }, { "grad_norm": 0.43347030878067017, "learning_rate": 9.962643947823677e-05, "loss": 0.0341, "step": 6960 }, { "grad_norm": 0.2730543613433838, "learning_rate": 9.962391347654921e-05, "loss": 0.0319, "step": 6970 }, { "grad_norm": 0.3044643998146057, "learning_rate": 9.962137899549204e-05, "loss": 0.0353, "step": 6980 }, { "grad_norm": 0.26576516032218933, "learning_rate": 9.961883603549835e-05, "loss": 0.0313, "step": 6990 }, { "grad_norm": 0.2674054503440857, "learning_rate": 9.961628459700267e-05, "loss": 0.0319, "step": 7000 }, { "grad_norm": 0.32691922783851624, "learning_rate": 9.961372468044095e-05, "loss": 0.0365, "step": 7010 }, { "grad_norm": 0.27323243021965027, "learning_rate": 9.961115628625062e-05, "loss": 0.0357, "step": 7020 }, { "grad_norm": 0.3162055015563965, "learning_rate": 9.960857941487056e-05, "loss": 0.0302, "step": 7030 }, { "grad_norm": 0.36620739102363586, "learning_rate": 9.960599406674106e-05, "loss": 0.0303, "step": 7040 }, { "grad_norm": 0.28314870595932007, "learning_rate": 9.960340024230393e-05, "loss": 0.029, "step": 7050 }, { "grad_norm": 0.39889049530029297, "learning_rate": 9.960079794200232e-05, "loss": 0.032, "step": 7060 }, { "grad_norm": 0.3084174394607544, "learning_rate": 9.959818716628096e-05, "loss": 0.0298, "step": 7070 }, { "grad_norm": 0.32021984457969666, "learning_rate": 9.95955679155859e-05, "loss": 0.0301, "step": 7080 }, { "grad_norm": 0.32022449374198914, "learning_rate": 9.959294019036472e-05, "loss": 0.0329, "step": 7090 }, { "grad_norm": 0.36562439799308777, "learning_rate": 9.959030399106646e-05, "loss": 0.0287, "step": 7100 }, { "grad_norm": 0.2498139590024948, "learning_rate": 9.958765931814153e-05, "loss": 0.0336, "step": 7110 }, { "grad_norm": 0.2524566650390625, "learning_rate": 9.958500617204184e-05, "loss": 0.0323, "step": 7120 }, { "grad_norm": 0.37731635570526123, "learning_rate": 9.958234455322075e-05, "loss": 0.0282, "step": 7130 }, { "grad_norm": 0.3420332670211792, "learning_rate": 9.957967446213308e-05, "loss": 0.0337, "step": 7140 }, { "grad_norm": 0.3113129436969757, "learning_rate": 9.957699589923501e-05, "loss": 0.029, "step": 7150 }, { "grad_norm": 0.3556317687034607, "learning_rate": 9.957430886498431e-05, "loss": 0.0311, "step": 7160 }, { "grad_norm": 0.30398181080818176, "learning_rate": 9.957161335984008e-05, "loss": 0.028, "step": 7170 }, { "grad_norm": 0.29735246300697327, "learning_rate": 9.956890938426291e-05, "loss": 0.0314, "step": 7180 }, { "grad_norm": 0.3482179045677185, "learning_rate": 9.956619693871482e-05, "loss": 0.03, "step": 7190 }, { "grad_norm": 0.28565502166748047, "learning_rate": 9.956347602365934e-05, "loss": 0.0316, "step": 7200 }, { "grad_norm": 0.3121965229511261, "learning_rate": 9.956074663956135e-05, "loss": 0.0321, "step": 7210 }, { "grad_norm": 0.29667773842811584, "learning_rate": 9.955800878688726e-05, "loss": 0.031, "step": 7220 }, { "grad_norm": 0.30837303400039673, "learning_rate": 9.955526246610489e-05, "loss": 0.0341, "step": 7230 }, { "grad_norm": 0.27572861313819885, "learning_rate": 9.955250767768349e-05, "loss": 0.0286, "step": 7240 }, { "grad_norm": 0.4410313665866852, "learning_rate": 9.95497444220938e-05, "loss": 0.0282, "step": 7250 }, { "grad_norm": 0.2930759787559509, "learning_rate": 9.954697269980797e-05, "loss": 0.0303, "step": 7260 }, { "grad_norm": 0.33841434121131897, "learning_rate": 9.954419251129962e-05, "loss": 0.0302, "step": 7270 }, { "grad_norm": 0.3518082797527313, "learning_rate": 9.95414038570438e-05, "loss": 0.0273, "step": 7280 }, { "grad_norm": 0.22279265522956848, "learning_rate": 9.953860673751703e-05, "loss": 0.0324, "step": 7290 }, { "grad_norm": 0.29210400581359863, "learning_rate": 9.953580115319725e-05, "loss": 0.032, "step": 7300 }, { "grad_norm": 0.32945048809051514, "learning_rate": 9.953298710456387e-05, "loss": 0.0304, "step": 7310 }, { "grad_norm": 0.3086070120334625, "learning_rate": 9.953016459209771e-05, "loss": 0.0309, "step": 7320 }, { "grad_norm": 0.28039342164993286, "learning_rate": 9.952733361628108e-05, "loss": 0.0355, "step": 7330 }, { "grad_norm": 0.293395072221756, "learning_rate": 9.952449417759772e-05, "loss": 0.0309, "step": 7340 }, { "grad_norm": 0.47441554069519043, "learning_rate": 9.952164627653279e-05, "loss": 0.0324, "step": 7350 }, { "grad_norm": 0.33182457089424133, "learning_rate": 9.951878991357292e-05, "loss": 0.0291, "step": 7360 }, { "grad_norm": 0.3146910071372986, "learning_rate": 9.951592508920622e-05, "loss": 0.0312, "step": 7370 }, { "grad_norm": 0.3261874318122864, "learning_rate": 9.951305180392219e-05, "loss": 0.0321, "step": 7380 }, { "grad_norm": 0.2843770682811737, "learning_rate": 9.951017005821178e-05, "loss": 0.0311, "step": 7390 }, { "grad_norm": 0.3669048547744751, "learning_rate": 9.95072798525674e-05, "loss": 0.0299, "step": 7400 }, { "grad_norm": 0.30616647005081177, "learning_rate": 9.950438118748293e-05, "loss": 0.033, "step": 7410 }, { "grad_norm": 0.2812444865703583, "learning_rate": 9.950147406345366e-05, "loss": 0.0283, "step": 7420 }, { "grad_norm": 0.24672824144363403, "learning_rate": 9.949855848097635e-05, "loss": 0.0357, "step": 7430 }, { "grad_norm": 0.25833192467689514, "learning_rate": 9.949563444054916e-05, "loss": 0.0328, "step": 7440 }, { "grad_norm": 0.3660585880279541, "learning_rate": 9.949270194267178e-05, "loss": 0.0333, "step": 7450 }, { "grad_norm": 0.3565424680709839, "learning_rate": 9.948976098784526e-05, "loss": 0.0314, "step": 7460 }, { "grad_norm": 0.3308977484703064, "learning_rate": 9.948681157657213e-05, "loss": 0.036, "step": 7470 }, { "grad_norm": 0.2970162630081177, "learning_rate": 9.948385370935638e-05, "loss": 0.0284, "step": 7480 }, { "grad_norm": 0.3377671241760254, "learning_rate": 9.94808873867034e-05, "loss": 0.0321, "step": 7490 }, { "grad_norm": 0.3049936592578888, "learning_rate": 9.947791260912009e-05, "loss": 0.0314, "step": 7500 }, { "grad_norm": 0.30066829919815063, "learning_rate": 9.947492937711474e-05, "loss": 0.0299, "step": 7510 }, { "grad_norm": 0.3367052376270294, "learning_rate": 9.947193769119707e-05, "loss": 0.028, "step": 7520 }, { "grad_norm": 0.33062928915023804, "learning_rate": 9.946893755187834e-05, "loss": 0.0311, "step": 7530 }, { "grad_norm": 0.24922950565814972, "learning_rate": 9.946592895967115e-05, "loss": 0.029, "step": 7540 }, { "grad_norm": 0.2623707056045532, "learning_rate": 9.94629119150896e-05, "loss": 0.0294, "step": 7550 }, { "grad_norm": 0.34724870324134827, "learning_rate": 9.94598864186492e-05, "loss": 0.0315, "step": 7560 }, { "grad_norm": 0.2699770927429199, "learning_rate": 9.945685247086696e-05, "loss": 0.0284, "step": 7570 }, { "grad_norm": 0.24909426271915436, "learning_rate": 9.945381007226129e-05, "loss": 0.0328, "step": 7580 }, { "grad_norm": 0.3560238778591156, "learning_rate": 9.945075922335203e-05, "loss": 0.0252, "step": 7590 }, { "grad_norm": 0.26795119047164917, "learning_rate": 9.944769992466049e-05, "loss": 0.0293, "step": 7600 }, { "grad_norm": 0.27005648612976074, "learning_rate": 9.944463217670945e-05, "loss": 0.0297, "step": 7610 }, { "grad_norm": 0.34101948142051697, "learning_rate": 9.944155598002307e-05, "loss": 0.032, "step": 7620 }, { "grad_norm": 0.36848586797714233, "learning_rate": 9.943847133512701e-05, "loss": 0.029, "step": 7630 }, { "grad_norm": 0.30031442642211914, "learning_rate": 9.943537824254834e-05, "loss": 0.037, "step": 7640 }, { "grad_norm": 0.37116560339927673, "learning_rate": 9.943227670281559e-05, "loss": 0.0343, "step": 7650 }, { "grad_norm": 0.3808576464653015, "learning_rate": 9.942916671645873e-05, "loss": 0.0265, "step": 7660 }, { "grad_norm": 0.30678391456604004, "learning_rate": 9.942604828400916e-05, "loss": 0.0299, "step": 7670 }, { "grad_norm": 0.33694988489151, "learning_rate": 9.942292140599975e-05, "loss": 0.0306, "step": 7680 }, { "grad_norm": 0.2833749055862427, "learning_rate": 9.94197860829648e-05, "loss": 0.0322, "step": 7690 }, { "grad_norm": 0.34629881381988525, "learning_rate": 9.941664231544004e-05, "loss": 0.0313, "step": 7700 }, { "grad_norm": 0.29000788927078247, "learning_rate": 9.941349010396264e-05, "loss": 0.0307, "step": 7710 }, { "grad_norm": 0.3303966224193573, "learning_rate": 9.941032944907125e-05, "loss": 0.0267, "step": 7720 }, { "grad_norm": 0.3532915711402893, "learning_rate": 9.940716035130596e-05, "loss": 0.0319, "step": 7730 }, { "grad_norm": 0.25930994749069214, "learning_rate": 9.940398281120821e-05, "loss": 0.0295, "step": 7740 }, { "grad_norm": 0.29536816477775574, "learning_rate": 9.940079682932102e-05, "loss": 0.0299, "step": 7750 }, { "grad_norm": 0.335245817899704, "learning_rate": 9.939760240618877e-05, "loss": 0.0321, "step": 7760 }, { "grad_norm": 0.2695044279098511, "learning_rate": 9.939439954235729e-05, "loss": 0.0287, "step": 7770 }, { "grad_norm": 0.2595720887184143, "learning_rate": 9.939118823837387e-05, "loss": 0.0306, "step": 7780 }, { "grad_norm": 0.29214048385620117, "learning_rate": 9.938796849478725e-05, "loss": 0.0276, "step": 7790 }, { "grad_norm": 0.2787722647190094, "learning_rate": 9.938474031214755e-05, "loss": 0.0314, "step": 7800 }, { "grad_norm": 0.213326558470726, "learning_rate": 9.938150369100643e-05, "loss": 0.0335, "step": 7810 }, { "grad_norm": 0.28689250349998474, "learning_rate": 9.93782586319169e-05, "loss": 0.0292, "step": 7820 }, { "grad_norm": 0.3023125231266022, "learning_rate": 9.937500513543348e-05, "loss": 0.0275, "step": 7830 }, { "grad_norm": 0.406034916639328, "learning_rate": 9.937174320211207e-05, "loss": 0.0303, "step": 7840 }, { "grad_norm": 0.2663906216621399, "learning_rate": 9.936847283251009e-05, "loss": 0.0243, "step": 7850 }, { "grad_norm": 0.25763583183288574, "learning_rate": 9.936519402718632e-05, "loss": 0.0262, "step": 7860 }, { "grad_norm": 0.2624443471431732, "learning_rate": 9.936190678670102e-05, "loss": 0.0261, "step": 7870 }, { "grad_norm": 0.32760635018348694, "learning_rate": 9.935861111161593e-05, "loss": 0.0318, "step": 7880 }, { "grad_norm": 0.27221572399139404, "learning_rate": 9.935530700249416e-05, "loss": 0.0287, "step": 7890 }, { "grad_norm": 0.2860610783100128, "learning_rate": 9.935199445990028e-05, "loss": 0.0301, "step": 7900 }, { "grad_norm": 0.43569415807724, "learning_rate": 9.934867348440033e-05, "loss": 0.0343, "step": 7910 }, { "grad_norm": 0.2852897644042969, "learning_rate": 9.934534407656176e-05, "loss": 0.0298, "step": 7920 }, { "grad_norm": 0.29880890250205994, "learning_rate": 9.93420062369535e-05, "loss": 0.0319, "step": 7930 }, { "grad_norm": 0.2859969735145569, "learning_rate": 9.933865996614589e-05, "loss": 0.0315, "step": 7940 }, { "grad_norm": 0.24886801838874817, "learning_rate": 9.933530526471068e-05, "loss": 0.0262, "step": 7950 }, { "grad_norm": 0.29580792784690857, "learning_rate": 9.933194213322114e-05, "loss": 0.0281, "step": 7960 }, { "grad_norm": 0.2344650775194168, "learning_rate": 9.932857057225192e-05, "loss": 0.0283, "step": 7970 }, { "grad_norm": 0.30398857593536377, "learning_rate": 9.932519058237912e-05, "loss": 0.0297, "step": 7980 }, { "grad_norm": 0.3496411144733429, "learning_rate": 9.932180216418032e-05, "loss": 0.0281, "step": 7990 }, { "grad_norm": 0.23623143136501312, "learning_rate": 9.931840531823446e-05, "loss": 0.028, "step": 8000 }, { "grad_norm": 0.3095143437385559, "learning_rate": 9.9315000045122e-05, "loss": 0.0301, "step": 8010 }, { "grad_norm": 0.3622783422470093, "learning_rate": 9.931158634542481e-05, "loss": 0.0322, "step": 8020 }, { "grad_norm": 0.28730127215385437, "learning_rate": 9.930816421972617e-05, "loss": 0.0276, "step": 8030 }, { "grad_norm": 0.2649135887622833, "learning_rate": 9.930473366861086e-05, "loss": 0.0296, "step": 8040 }, { "grad_norm": 0.21076826751232147, "learning_rate": 9.930129469266505e-05, "loss": 0.0236, "step": 8050 }, { "grad_norm": 0.2395796775817871, "learning_rate": 9.929784729247638e-05, "loss": 0.0256, "step": 8060 }, { "grad_norm": 0.2573155462741852, "learning_rate": 9.929439146863389e-05, "loss": 0.0272, "step": 8070 }, { "grad_norm": 0.3180524408817291, "learning_rate": 9.92909272217281e-05, "loss": 0.0282, "step": 8080 }, { "grad_norm": 0.1964866816997528, "learning_rate": 9.928745455235097e-05, "loss": 0.0251, "step": 8090 }, { "grad_norm": 0.23427104949951172, "learning_rate": 9.928397346109588e-05, "loss": 0.0272, "step": 8100 }, { "grad_norm": 0.2629137933254242, "learning_rate": 9.928048394855762e-05, "loss": 0.0306, "step": 8110 }, { "grad_norm": 0.31017664074897766, "learning_rate": 9.92769860153325e-05, "loss": 0.0295, "step": 8120 }, { "grad_norm": 0.2915886342525482, "learning_rate": 9.927347966201819e-05, "loss": 0.0285, "step": 8130 }, { "grad_norm": 0.36048775911331177, "learning_rate": 9.926996488921383e-05, "loss": 0.0305, "step": 8140 }, { "grad_norm": 0.2941038906574249, "learning_rate": 9.926644169752001e-05, "loss": 0.0301, "step": 8150 }, { "grad_norm": 0.2892025411128998, "learning_rate": 9.926291008753875e-05, "loss": 0.031, "step": 8160 }, { "grad_norm": 0.2791893780231476, "learning_rate": 9.92593700598735e-05, "loss": 0.029, "step": 8170 }, { "grad_norm": 0.3169870674610138, "learning_rate": 9.925582161512915e-05, "loss": 0.0297, "step": 8180 }, { "grad_norm": 0.32276034355163574, "learning_rate": 9.925226475391205e-05, "loss": 0.0287, "step": 8190 }, { "grad_norm": 0.2249068170785904, "learning_rate": 9.924869947682993e-05, "loss": 0.0255, "step": 8200 }, { "grad_norm": 0.24731381237506866, "learning_rate": 9.924512578449204e-05, "loss": 0.0278, "step": 8210 }, { "grad_norm": 0.2790684998035431, "learning_rate": 9.924154367750901e-05, "loss": 0.0258, "step": 8220 }, { "grad_norm": 0.27847883105278015, "learning_rate": 9.923795315649293e-05, "loss": 0.0271, "step": 8230 }, { "grad_norm": 0.28081899881362915, "learning_rate": 9.92343542220573e-05, "loss": 0.0262, "step": 8240 }, { "grad_norm": 0.21870897710323334, "learning_rate": 9.92307468748171e-05, "loss": 0.0281, "step": 8250 }, { "grad_norm": 0.22965073585510254, "learning_rate": 9.922713111538873e-05, "loss": 0.0281, "step": 8260 }, { "grad_norm": 0.25257420539855957, "learning_rate": 9.922350694439003e-05, "loss": 0.0319, "step": 8270 }, { "grad_norm": 0.2984631061553955, "learning_rate": 9.921987436244024e-05, "loss": 0.0322, "step": 8280 }, { "grad_norm": 0.23995836079120636, "learning_rate": 9.921623337016008e-05, "loss": 0.0291, "step": 8290 }, { "grad_norm": 0.3045021891593933, "learning_rate": 9.921258396817172e-05, "loss": 0.0272, "step": 8300 }, { "grad_norm": 0.22881236672401428, "learning_rate": 9.920892615709874e-05, "loss": 0.0302, "step": 8310 }, { "grad_norm": 0.2488105744123459, "learning_rate": 9.920525993756612e-05, "loss": 0.0291, "step": 8320 }, { "grad_norm": 0.3306199908256531, "learning_rate": 9.920158531020036e-05, "loss": 0.0259, "step": 8330 }, { "grad_norm": 0.32168129086494446, "learning_rate": 9.919790227562933e-05, "loss": 0.0276, "step": 8340 }, { "grad_norm": 0.26897913217544556, "learning_rate": 9.919421083448237e-05, "loss": 0.0269, "step": 8350 }, { "grad_norm": 0.23823562264442444, "learning_rate": 9.919051098739022e-05, "loss": 0.0265, "step": 8360 }, { "grad_norm": 0.32039517164230347, "learning_rate": 9.918680273498514e-05, "loss": 0.0262, "step": 8370 }, { "grad_norm": 0.319292813539505, "learning_rate": 9.918308607790072e-05, "loss": 0.0283, "step": 8380 }, { "grad_norm": 0.26043254137039185, "learning_rate": 9.917936101677205e-05, "loss": 0.0265, "step": 8390 }, { "grad_norm": 0.2665093243122101, "learning_rate": 9.917562755223564e-05, "loss": 0.0272, "step": 8400 }, { "grad_norm": 0.27096131443977356, "learning_rate": 9.917188568492944e-05, "loss": 0.0259, "step": 8410 }, { "grad_norm": 0.26681190729141235, "learning_rate": 9.916813541549283e-05, "loss": 0.0268, "step": 8420 }, { "grad_norm": 0.2953859865665436, "learning_rate": 9.916437674456663e-05, "loss": 0.0262, "step": 8430 }, { "grad_norm": 0.22404322028160095, "learning_rate": 9.916060967279308e-05, "loss": 0.0297, "step": 8440 }, { "grad_norm": 0.32462626695632935, "learning_rate": 9.91568342008159e-05, "loss": 0.031, "step": 8450 }, { "grad_norm": 0.34318259358406067, "learning_rate": 9.915305032928019e-05, "loss": 0.0295, "step": 8460 }, { "grad_norm": 0.2582516670227051, "learning_rate": 9.914925805883253e-05, "loss": 0.029, "step": 8470 }, { "grad_norm": 0.29810529947280884, "learning_rate": 9.914545739012088e-05, "loss": 0.0267, "step": 8480 }, { "grad_norm": 0.2799208462238312, "learning_rate": 9.91416483237947e-05, "loss": 0.0261, "step": 8490 }, { "grad_norm": 0.3227693438529968, "learning_rate": 9.913783086050485e-05, "loss": 0.028, "step": 8500 }, { "grad_norm": 0.2845383882522583, "learning_rate": 9.913400500090364e-05, "loss": 0.028, "step": 8510 }, { "grad_norm": 0.29315465688705444, "learning_rate": 9.913017074564479e-05, "loss": 0.0305, "step": 8520 }, { "grad_norm": 0.28585702180862427, "learning_rate": 9.912632809538348e-05, "loss": 0.031, "step": 8530 }, { "grad_norm": 0.3429509401321411, "learning_rate": 9.912247705077629e-05, "loss": 0.0308, "step": 8540 }, { "grad_norm": 0.3061855733394623, "learning_rate": 9.911861761248127e-05, "loss": 0.0296, "step": 8550 }, { "grad_norm": 0.3516901731491089, "learning_rate": 9.91147497811579e-05, "loss": 0.0296, "step": 8560 }, { "grad_norm": 0.2861192226409912, "learning_rate": 9.911087355746709e-05, "loss": 0.0249, "step": 8570 }, { "grad_norm": 0.3234238922595978, "learning_rate": 9.910698894207117e-05, "loss": 0.0286, "step": 8580 }, { "grad_norm": 0.49911239743232727, "learning_rate": 9.910309593563392e-05, "loss": 0.0264, "step": 8590 }, { "grad_norm": 0.21117472648620605, "learning_rate": 9.909919453882057e-05, "loss": 0.0254, "step": 8600 }, { "grad_norm": 0.25785863399505615, "learning_rate": 9.90952847522977e-05, "loss": 0.0274, "step": 8610 }, { "grad_norm": 0.35120418667793274, "learning_rate": 9.909136657673346e-05, "loss": 0.0283, "step": 8620 }, { "grad_norm": 0.35102033615112305, "learning_rate": 9.908744001279731e-05, "loss": 0.0232, "step": 8630 }, { "grad_norm": 0.25661784410476685, "learning_rate": 9.90835050611602e-05, "loss": 0.0269, "step": 8640 }, { "grad_norm": 0.25306740403175354, "learning_rate": 9.90795617224945e-05, "loss": 0.0265, "step": 8650 }, { "grad_norm": 0.2748759686946869, "learning_rate": 9.907560999747405e-05, "loss": 0.0278, "step": 8660 }, { "grad_norm": 0.2409040778875351, "learning_rate": 9.907164988677408e-05, "loss": 0.0307, "step": 8670 }, { "grad_norm": 0.29659518599510193, "learning_rate": 9.906768139107124e-05, "loss": 0.0286, "step": 8680 }, { "grad_norm": 0.29645028710365295, "learning_rate": 9.906370451104367e-05, "loss": 0.0272, "step": 8690 }, { "grad_norm": 0.3434709906578064, "learning_rate": 9.905971924737088e-05, "loss": 0.024, "step": 8700 }, { "grad_norm": 0.3046552240848541, "learning_rate": 9.905572560073387e-05, "loss": 0.0265, "step": 8710 }, { "grad_norm": 0.2530934512615204, "learning_rate": 9.905172357181501e-05, "loss": 0.0227, "step": 8720 }, { "grad_norm": 0.30435165762901306, "learning_rate": 9.904771316129817e-05, "loss": 0.0282, "step": 8730 }, { "grad_norm": 0.20620696246623993, "learning_rate": 9.904369436986862e-05, "loss": 0.0252, "step": 8740 }, { "grad_norm": 0.27456456422805786, "learning_rate": 9.903966719821303e-05, "loss": 0.0299, "step": 8750 }, { "grad_norm": 0.2697552442550659, "learning_rate": 9.903563164701956e-05, "loss": 0.0255, "step": 8760 }, { "grad_norm": 0.2198791652917862, "learning_rate": 9.903158771697778e-05, "loss": 0.0233, "step": 8770 }, { "grad_norm": 0.3000161349773407, "learning_rate": 9.902753540877867e-05, "loss": 0.0249, "step": 8780 }, { "grad_norm": 0.3050804138183594, "learning_rate": 9.902347472311466e-05, "loss": 0.0253, "step": 8790 }, { "grad_norm": 0.2566889524459839, "learning_rate": 9.901940566067962e-05, "loss": 0.0305, "step": 8800 }, { "grad_norm": 0.23254813253879547, "learning_rate": 9.901532822216883e-05, "loss": 0.0261, "step": 8810 }, { "grad_norm": 0.29862529039382935, "learning_rate": 9.901124240827904e-05, "loss": 0.029, "step": 8820 }, { "grad_norm": 0.28771379590034485, "learning_rate": 9.900714821970835e-05, "loss": 0.0327, "step": 8830 }, { "grad_norm": 0.30176857113838196, "learning_rate": 9.900304565715641e-05, "loss": 0.03, "step": 8840 }, { "grad_norm": 0.2580115497112274, "learning_rate": 9.899893472132419e-05, "loss": 0.0322, "step": 8850 }, { "grad_norm": 0.3875986635684967, "learning_rate": 9.899481541291415e-05, "loss": 0.0294, "step": 8860 }, { "grad_norm": 0.283587247133255, "learning_rate": 9.899068773263016e-05, "loss": 0.0297, "step": 8870 }, { "grad_norm": 0.30284956097602844, "learning_rate": 9.898655168117754e-05, "loss": 0.0242, "step": 8880 }, { "grad_norm": 0.2892449200153351, "learning_rate": 9.898240725926302e-05, "loss": 0.0284, "step": 8890 }, { "grad_norm": 0.29546940326690674, "learning_rate": 9.897825446759478e-05, "loss": 0.0236, "step": 8900 }, { "grad_norm": 0.2503766119480133, "learning_rate": 9.897409330688241e-05, "loss": 0.0267, "step": 8910 }, { "grad_norm": 0.3384060561656952, "learning_rate": 9.896992377783692e-05, "loss": 0.0251, "step": 8920 }, { "grad_norm": 0.33530983328819275, "learning_rate": 9.89657458811708e-05, "loss": 0.0258, "step": 8930 }, { "grad_norm": 0.28765419125556946, "learning_rate": 9.896155961759792e-05, "loss": 0.0259, "step": 8940 }, { "grad_norm": 0.41243207454681396, "learning_rate": 9.895736498783361e-05, "loss": 0.0294, "step": 8950 }, { "grad_norm": 0.2948262095451355, "learning_rate": 9.895316199259462e-05, "loss": 0.0306, "step": 8960 }, { "grad_norm": 0.279682457447052, "learning_rate": 9.894895063259909e-05, "loss": 0.0282, "step": 8970 }, { "grad_norm": 0.2098907232284546, "learning_rate": 9.894473090856667e-05, "loss": 0.0297, "step": 8980 }, { "grad_norm": 0.37159550189971924, "learning_rate": 9.894050282121839e-05, "loss": 0.0275, "step": 8990 }, { "grad_norm": 0.2978190779685974, "learning_rate": 9.893626637127668e-05, "loss": 0.0316, "step": 9000 }, { "grad_norm": 0.2476150393486023, "learning_rate": 9.893202155946546e-05, "loss": 0.0307, "step": 9010 }, { "grad_norm": 0.26011380553245544, "learning_rate": 9.892776838651006e-05, "loss": 0.024, "step": 9020 }, { "grad_norm": 0.24072998762130737, "learning_rate": 9.892350685313722e-05, "loss": 0.0283, "step": 9030 }, { "grad_norm": 0.3597114384174347, "learning_rate": 9.891923696007513e-05, "loss": 0.029, "step": 9040 }, { "grad_norm": 0.2600955665111542, "learning_rate": 9.891495870805336e-05, "loss": 0.0244, "step": 9050 }, { "grad_norm": 0.2049400508403778, "learning_rate": 9.891067209780298e-05, "loss": 0.0262, "step": 9060 }, { "grad_norm": 0.28217488527297974, "learning_rate": 9.890637713005646e-05, "loss": 0.0274, "step": 9070 }, { "grad_norm": 0.28004205226898193, "learning_rate": 9.890207380554767e-05, "loss": 0.027, "step": 9080 }, { "grad_norm": 0.2600816488265991, "learning_rate": 9.889776212501196e-05, "loss": 0.0277, "step": 9090 }, { "grad_norm": 0.328202486038208, "learning_rate": 9.889344208918605e-05, "loss": 0.026, "step": 9100 }, { "grad_norm": 0.30145463347435, "learning_rate": 9.888911369880812e-05, "loss": 0.0272, "step": 9110 }, { "grad_norm": 0.24766550958156586, "learning_rate": 9.888477695461777e-05, "loss": 0.0266, "step": 9120 }, { "grad_norm": 0.3101094961166382, "learning_rate": 9.888043185735607e-05, "loss": 0.025, "step": 9130 }, { "grad_norm": 0.3316406011581421, "learning_rate": 9.887607840776542e-05, "loss": 0.028, "step": 9140 }, { "grad_norm": 0.27868255972862244, "learning_rate": 9.887171660658975e-05, "loss": 0.0231, "step": 9150 }, { "grad_norm": 0.25141122937202454, "learning_rate": 9.886734645457435e-05, "loss": 0.0251, "step": 9160 }, { "grad_norm": 0.235273540019989, "learning_rate": 9.886296795246597e-05, "loss": 0.0275, "step": 9170 }, { "grad_norm": 0.27024224400520325, "learning_rate": 9.885858110101276e-05, "loss": 0.024, "step": 9180 }, { "grad_norm": 0.2795313000679016, "learning_rate": 9.885418590096434e-05, "loss": 0.0246, "step": 9190 }, { "grad_norm": 0.2418869435787201, "learning_rate": 9.88497823530717e-05, "loss": 0.0261, "step": 9200 }, { "grad_norm": 0.2891831696033478, "learning_rate": 9.884537045808732e-05, "loss": 0.0265, "step": 9210 }, { "grad_norm": 0.22151601314544678, "learning_rate": 9.884095021676502e-05, "loss": 0.0272, "step": 9220 }, { "grad_norm": 0.22702912986278534, "learning_rate": 9.883652162986017e-05, "loss": 0.0253, "step": 9230 }, { "grad_norm": 0.29362452030181885, "learning_rate": 9.883208469812943e-05, "loss": 0.0273, "step": 9240 }, { "grad_norm": 0.21563395857810974, "learning_rate": 9.882763942233098e-05, "loss": 0.0234, "step": 9250 }, { "grad_norm": 0.26447415351867676, "learning_rate": 9.882318580322441e-05, "loss": 0.024, "step": 9260 }, { "grad_norm": 0.23151524364948273, "learning_rate": 9.881872384157067e-05, "loss": 0.0238, "step": 9270 }, { "grad_norm": 0.25850212574005127, "learning_rate": 9.881425353813225e-05, "loss": 0.0268, "step": 9280 }, { "grad_norm": 0.28777140378952026, "learning_rate": 9.880977489367296e-05, "loss": 0.0263, "step": 9290 }, { "grad_norm": 0.32817912101745605, "learning_rate": 9.88052879089581e-05, "loss": 0.0276, "step": 9300 }, { "grad_norm": 0.2871648967266083, "learning_rate": 9.880079258475434e-05, "loss": 0.024, "step": 9310 }, { "grad_norm": 0.26717129349708557, "learning_rate": 9.879628892182985e-05, "loss": 0.0227, "step": 9320 }, { "grad_norm": 0.2416200190782547, "learning_rate": 9.879177692095416e-05, "loss": 0.0236, "step": 9330 }, { "grad_norm": 0.33232495188713074, "learning_rate": 9.878725658289825e-05, "loss": 0.0233, "step": 9340 }, { "grad_norm": 0.33320462703704834, "learning_rate": 9.878272790843454e-05, "loss": 0.0265, "step": 9350 }, { "grad_norm": 0.2566198408603668, "learning_rate": 9.877819089833682e-05, "loss": 0.0252, "step": 9360 }, { "grad_norm": 0.3075855076313019, "learning_rate": 9.877364555338038e-05, "loss": 0.0226, "step": 9370 }, { "grad_norm": 0.231120303273201, "learning_rate": 9.876909187434186e-05, "loss": 0.0252, "step": 9380 }, { "grad_norm": 0.2447887808084488, "learning_rate": 9.876452986199939e-05, "loss": 0.0213, "step": 9390 }, { "grad_norm": 0.2527186870574951, "learning_rate": 9.875995951713248e-05, "loss": 0.024, "step": 9400 }, { "grad_norm": 0.38442882895469666, "learning_rate": 9.875538084052207e-05, "loss": 0.0208, "step": 9410 }, { "grad_norm": 0.3187733590602875, "learning_rate": 9.875079383295053e-05, "loss": 0.0259, "step": 9420 }, { "grad_norm": 0.22343404591083527, "learning_rate": 9.874619849520167e-05, "loss": 0.0274, "step": 9430 }, { "grad_norm": 0.24663791060447693, "learning_rate": 9.874159482806069e-05, "loss": 0.0255, "step": 9440 }, { "grad_norm": 0.23998089134693146, "learning_rate": 9.873698283231426e-05, "loss": 0.0274, "step": 9450 }, { "grad_norm": 0.20122860372066498, "learning_rate": 9.87323625087504e-05, "loss": 0.0235, "step": 9460 }, { "grad_norm": 0.23733201622962952, "learning_rate": 9.872773385815863e-05, "loss": 0.0247, "step": 9470 }, { "grad_norm": 0.261588990688324, "learning_rate": 9.872309688132986e-05, "loss": 0.0263, "step": 9480 }, { "grad_norm": 0.24089820683002472, "learning_rate": 9.871845157905639e-05, "loss": 0.0246, "step": 9490 }, { "grad_norm": 0.28446510434150696, "learning_rate": 9.871379795213201e-05, "loss": 0.0254, "step": 9500 }, { "grad_norm": 0.17613108456134796, "learning_rate": 9.87091360013519e-05, "loss": 0.0265, "step": 9510 }, { "grad_norm": 0.1985762119293213, "learning_rate": 9.870446572751262e-05, "loss": 0.0222, "step": 9520 }, { "grad_norm": 0.18427501618862152, "learning_rate": 9.869978713141224e-05, "loss": 0.02, "step": 9530 }, { "grad_norm": 0.32171618938446045, "learning_rate": 9.869510021385016e-05, "loss": 0.0255, "step": 9540 }, { "grad_norm": 0.2887208163738251, "learning_rate": 9.869040497562727e-05, "loss": 0.0294, "step": 9550 }, { "grad_norm": 0.3040117025375366, "learning_rate": 9.868570141754587e-05, "loss": 0.0263, "step": 9560 }, { "grad_norm": 0.25922444462776184, "learning_rate": 9.868098954040965e-05, "loss": 0.0239, "step": 9570 }, { "grad_norm": 0.2811874449253082, "learning_rate": 9.867626934502374e-05, "loss": 0.0293, "step": 9580 }, { "grad_norm": 0.25736549496650696, "learning_rate": 9.86715408321947e-05, "loss": 0.0276, "step": 9590 }, { "grad_norm": 0.21450766921043396, "learning_rate": 9.86668040027305e-05, "loss": 0.0218, "step": 9600 }, { "grad_norm": 0.30204737186431885, "learning_rate": 9.866205885744053e-05, "loss": 0.0266, "step": 9610 }, { "grad_norm": 0.2931254208087921, "learning_rate": 9.865730539713563e-05, "loss": 0.0222, "step": 9620 }, { "grad_norm": 0.24818548560142517, "learning_rate": 9.8652543622628e-05, "loss": 0.0224, "step": 9630 }, { "grad_norm": 0.20985734462738037, "learning_rate": 9.864777353473132e-05, "loss": 0.0272, "step": 9640 }, { "grad_norm": 0.2521359622478485, "learning_rate": 9.864299513426068e-05, "loss": 0.0284, "step": 9650 }, { "grad_norm": 0.2349608987569809, "learning_rate": 9.863820842203254e-05, "loss": 0.027, "step": 9660 }, { "grad_norm": 0.350739985704422, "learning_rate": 9.863341339886483e-05, "loss": 0.0216, "step": 9670 }, { "grad_norm": 0.30511465668678284, "learning_rate": 9.86286100655769e-05, "loss": 0.0214, "step": 9680 }, { "grad_norm": 0.23080308735370636, "learning_rate": 9.862379842298953e-05, "loss": 0.0255, "step": 9690 }, { "grad_norm": 0.28018489480018616, "learning_rate": 9.861897847192485e-05, "loss": 0.0253, "step": 9700 }, { "grad_norm": 0.2550153136253357, "learning_rate": 9.86141502132065e-05, "loss": 0.0248, "step": 9710 }, { "grad_norm": 0.2044994980096817, "learning_rate": 9.860931364765946e-05, "loss": 0.0245, "step": 9720 }, { "grad_norm": 0.21526765823364258, "learning_rate": 9.860446877611021e-05, "loss": 0.0255, "step": 9730 }, { "grad_norm": 0.30527254939079285, "learning_rate": 9.859961559938655e-05, "loss": 0.0266, "step": 9740 }, { "grad_norm": 0.2860325872898102, "learning_rate": 9.85947541183178e-05, "loss": 0.0274, "step": 9750 }, { "grad_norm": 0.22889266908168793, "learning_rate": 9.858988433373463e-05, "loss": 0.0247, "step": 9760 }, { "grad_norm": 0.20900379121303558, "learning_rate": 9.858500624646918e-05, "loss": 0.0238, "step": 9770 }, { "grad_norm": 0.24386528134346008, "learning_rate": 9.858011985735497e-05, "loss": 0.0207, "step": 9780 }, { "grad_norm": 0.3092852234840393, "learning_rate": 9.857522516722693e-05, "loss": 0.0225, "step": 9790 }, { "grad_norm": 0.22450950741767883, "learning_rate": 9.857032217692145e-05, "loss": 0.0238, "step": 9800 }, { "grad_norm": 0.2719589173793793, "learning_rate": 9.856541088727631e-05, "loss": 0.0316, "step": 9810 }, { "grad_norm": 0.31458044052124023, "learning_rate": 9.856049129913072e-05, "loss": 0.0275, "step": 9820 }, { "grad_norm": 0.2636988162994385, "learning_rate": 9.85555634133253e-05, "loss": 0.0276, "step": 9830 }, { "grad_norm": 0.22225773334503174, "learning_rate": 9.855062723070208e-05, "loss": 0.0258, "step": 9840 }, { "grad_norm": 0.2955557405948639, "learning_rate": 9.854568275210454e-05, "loss": 0.0275, "step": 9850 }, { "grad_norm": 0.2688165307044983, "learning_rate": 9.854072997837754e-05, "loss": 0.0257, "step": 9860 }, { "grad_norm": 0.21458207070827484, "learning_rate": 9.853576891036737e-05, "loss": 0.0211, "step": 9870 }, { "grad_norm": 0.30546480417251587, "learning_rate": 9.853079954892177e-05, "loss": 0.0212, "step": 9880 }, { "grad_norm": 0.24983032047748566, "learning_rate": 9.852582189488983e-05, "loss": 0.0231, "step": 9890 }, { "grad_norm": 0.25877729058265686, "learning_rate": 9.852083594912212e-05, "loss": 0.023, "step": 9900 }, { "grad_norm": 0.24235117435455322, "learning_rate": 9.851584171247058e-05, "loss": 0.0244, "step": 9910 }, { "grad_norm": 0.2512979507446289, "learning_rate": 9.851083918578863e-05, "loss": 0.0224, "step": 9920 }, { "grad_norm": 0.21787101030349731, "learning_rate": 9.850582836993103e-05, "loss": 0.0232, "step": 9930 }, { "grad_norm": 0.2501905858516693, "learning_rate": 9.850080926575397e-05, "loss": 0.0204, "step": 9940 }, { "grad_norm": 0.2442246973514557, "learning_rate": 9.849578187411515e-05, "loss": 0.028, "step": 9950 }, { "grad_norm": 0.29922616481781006, "learning_rate": 9.849074619587354e-05, "loss": 0.0251, "step": 9960 }, { "grad_norm": 0.21666263043880463, "learning_rate": 9.848570223188964e-05, "loss": 0.023, "step": 9970 }, { "grad_norm": 0.25425854325294495, "learning_rate": 9.848064998302531e-05, "loss": 0.0201, "step": 9980 }, { "grad_norm": 0.25501516461372375, "learning_rate": 9.847558945014386e-05, "loss": 0.0223, "step": 9990 }, { "grad_norm": 0.21783235669136047, "learning_rate": 9.847052063410996e-05, "loss": 0.0227, "step": 10000 }, { "grad_norm": 0.24500063061714172, "learning_rate": 9.846544353578977e-05, "loss": 0.0209, "step": 10010 }, { "grad_norm": 0.27603912353515625, "learning_rate": 9.846035815605081e-05, "loss": 0.0235, "step": 10020 }, { "grad_norm": 0.3227001130580902, "learning_rate": 9.845526449576204e-05, "loss": 0.0239, "step": 10030 }, { "grad_norm": 0.28614410758018494, "learning_rate": 9.845016255579383e-05, "loss": 0.022, "step": 10040 }, { "grad_norm": 0.2617984116077423, "learning_rate": 9.844505233701794e-05, "loss": 0.022, "step": 10050 }, { "grad_norm": 0.23960858583450317, "learning_rate": 9.843993384030757e-05, "loss": 0.0209, "step": 10060 }, { "grad_norm": 0.2163446694612503, "learning_rate": 9.843480706653737e-05, "loss": 0.0209, "step": 10070 }, { "grad_norm": 0.2481665462255478, "learning_rate": 9.84296720165833e-05, "loss": 0.0288, "step": 10080 }, { "grad_norm": 0.32253608107566833, "learning_rate": 9.842452869132286e-05, "loss": 0.0224, "step": 10090 }, { "grad_norm": 0.2658237814903259, "learning_rate": 9.841937709163489e-05, "loss": 0.0236, "step": 10100 }, { "grad_norm": 0.1907413750886917, "learning_rate": 9.841421721839962e-05, "loss": 0.0201, "step": 10110 }, { "grad_norm": 0.23303678631782532, "learning_rate": 9.840904907249879e-05, "loss": 0.0227, "step": 10120 }, { "grad_norm": 0.2713325023651123, "learning_rate": 9.840387265481545e-05, "loss": 0.0251, "step": 10130 }, { "grad_norm": 0.27836093306541443, "learning_rate": 9.839868796623411e-05, "loss": 0.0225, "step": 10140 }, { "grad_norm": 0.24270500242710114, "learning_rate": 9.839349500764072e-05, "loss": 0.0214, "step": 10150 }, { "grad_norm": 0.29587307572364807, "learning_rate": 9.83882937799226e-05, "loss": 0.0248, "step": 10160 }, { "grad_norm": 0.1698063760995865, "learning_rate": 9.838308428396849e-05, "loss": 0.0239, "step": 10170 }, { "grad_norm": 0.3109075725078583, "learning_rate": 9.837786652066854e-05, "loss": 0.0256, "step": 10180 }, { "grad_norm": 0.2277935892343521, "learning_rate": 9.837264049091437e-05, "loss": 0.0234, "step": 10190 }, { "grad_norm": 0.3449375629425049, "learning_rate": 9.836740619559893e-05, "loss": 0.0238, "step": 10200 }, { "grad_norm": 0.3596870005130768, "learning_rate": 9.836216363561659e-05, "loss": 0.0281, "step": 10210 }, { "grad_norm": 0.4000246822834015, "learning_rate": 9.835691281186322e-05, "loss": 0.0282, "step": 10220 }, { "grad_norm": 0.2865615785121918, "learning_rate": 9.8351653725236e-05, "loss": 0.024, "step": 10230 }, { "grad_norm": 0.20724831521511078, "learning_rate": 9.83463863766336e-05, "loss": 0.0224, "step": 10240 }, { "grad_norm": 0.2764706313610077, "learning_rate": 9.834111076695602e-05, "loss": 0.021, "step": 10250 }, { "grad_norm": 0.17200616002082825, "learning_rate": 9.833582689710477e-05, "loss": 0.0266, "step": 10260 }, { "grad_norm": 0.24495764076709747, "learning_rate": 9.833053476798268e-05, "loss": 0.0241, "step": 10270 }, { "grad_norm": 0.24808824062347412, "learning_rate": 9.832523438049404e-05, "loss": 0.0241, "step": 10280 }, { "grad_norm": 0.22137804329395294, "learning_rate": 9.831992573554454e-05, "loss": 0.0201, "step": 10290 }, { "grad_norm": 0.20455269515514374, "learning_rate": 9.831460883404128e-05, "loss": 0.0215, "step": 10300 }, { "grad_norm": 0.26773127913475037, "learning_rate": 9.830928367689278e-05, "loss": 0.0225, "step": 10310 }, { "grad_norm": 0.2562367022037506, "learning_rate": 9.830395026500896e-05, "loss": 0.0212, "step": 10320 }, { "grad_norm": 0.24280409514904022, "learning_rate": 9.829860859930115e-05, "loss": 0.0232, "step": 10330 }, { "grad_norm": 0.20335355401039124, "learning_rate": 9.829325868068212e-05, "loss": 0.0214, "step": 10340 }, { "grad_norm": 0.30372750759124756, "learning_rate": 9.8287900510066e-05, "loss": 0.0269, "step": 10350 }, { "grad_norm": 0.26752519607543945, "learning_rate": 9.828253408836834e-05, "loss": 0.0257, "step": 10360 }, { "grad_norm": 0.26025673747062683, "learning_rate": 9.827715941650615e-05, "loss": 0.0235, "step": 10370 }, { "grad_norm": 0.2509560286998749, "learning_rate": 9.82717764953978e-05, "loss": 0.0215, "step": 10380 }, { "grad_norm": 0.23601439595222473, "learning_rate": 9.826638532596308e-05, "loss": 0.0221, "step": 10390 }, { "grad_norm": 0.24748827517032623, "learning_rate": 9.82609859091232e-05, "loss": 0.0213, "step": 10400 }, { "grad_norm": 0.2148655503988266, "learning_rate": 9.825557824580076e-05, "loss": 0.0214, "step": 10410 }, { "grad_norm": 0.22486890852451324, "learning_rate": 9.82501623369198e-05, "loss": 0.0193, "step": 10420 }, { "grad_norm": 0.2562028467655182, "learning_rate": 9.824473818340574e-05, "loss": 0.0232, "step": 10430 }, { "grad_norm": 0.23531204462051392, "learning_rate": 9.823930578618541e-05, "loss": 0.022, "step": 10440 }, { "grad_norm": 0.2482888102531433, "learning_rate": 9.823386514618709e-05, "loss": 0.0201, "step": 10450 }, { "grad_norm": 0.2269417941570282, "learning_rate": 9.82284162643404e-05, "loss": 0.0225, "step": 10460 }, { "grad_norm": 0.25478675961494446, "learning_rate": 9.822295914157642e-05, "loss": 0.022, "step": 10470 }, { "grad_norm": 0.2134317308664322, "learning_rate": 9.821749377882763e-05, "loss": 0.0207, "step": 10480 }, { "grad_norm": 0.1966019719839096, "learning_rate": 9.821202017702791e-05, "loss": 0.0209, "step": 10490 }, { "grad_norm": 0.26036450266838074, "learning_rate": 9.820653833711253e-05, "loss": 0.0216, "step": 10500 }, { "grad_norm": 0.26643022894859314, "learning_rate": 9.820104826001822e-05, "loss": 0.0235, "step": 10510 }, { "grad_norm": 0.28584229946136475, "learning_rate": 9.819554994668305e-05, "loss": 0.0246, "step": 10520 }, { "grad_norm": 0.30917075276374817, "learning_rate": 9.819004339804654e-05, "loss": 0.023, "step": 10530 }, { "grad_norm": 0.2620486319065094, "learning_rate": 9.818452861504961e-05, "loss": 0.0236, "step": 10540 }, { "grad_norm": 0.31446725130081177, "learning_rate": 9.81790055986346e-05, "loss": 0.0259, "step": 10550 }, { "grad_norm": 0.3023761510848999, "learning_rate": 9.817347434974523e-05, "loss": 0.0203, "step": 10560 }, { "grad_norm": 0.31798455119132996, "learning_rate": 9.816793486932664e-05, "loss": 0.0232, "step": 10570 }, { "grad_norm": 0.2649776339530945, "learning_rate": 9.816238715832538e-05, "loss": 0.0222, "step": 10580 }, { "grad_norm": 0.19242195785045624, "learning_rate": 9.815683121768939e-05, "loss": 0.0252, "step": 10590 }, { "grad_norm": 0.22264128923416138, "learning_rate": 9.815126704836804e-05, "loss": 0.0247, "step": 10600 }, { "grad_norm": 0.23237144947052002, "learning_rate": 9.81456946513121e-05, "loss": 0.0264, "step": 10610 }, { "grad_norm": 0.293218731880188, "learning_rate": 9.814011402747373e-05, "loss": 0.0278, "step": 10620 }, { "grad_norm": 0.2137145698070526, "learning_rate": 9.813452517780651e-05, "loss": 0.022, "step": 10630 }, { "grad_norm": 0.2338603436946869, "learning_rate": 9.81289281032654e-05, "loss": 0.0246, "step": 10640 }, { "grad_norm": 0.24001209437847137, "learning_rate": 9.812332280480683e-05, "loss": 0.0219, "step": 10650 }, { "grad_norm": 0.2928566038608551, "learning_rate": 9.811770928338854e-05, "loss": 0.0236, "step": 10660 }, { "grad_norm": 0.23578087985515594, "learning_rate": 9.811208753996979e-05, "loss": 0.0193, "step": 10670 }, { "grad_norm": 0.2514597773551941, "learning_rate": 9.810645757551113e-05, "loss": 0.019, "step": 10680 }, { "grad_norm": 0.2798321545124054, "learning_rate": 9.810081939097459e-05, "loss": 0.0207, "step": 10690 }, { "grad_norm": 0.18075917661190033, "learning_rate": 9.809517298732356e-05, "loss": 0.0224, "step": 10700 }, { "grad_norm": 0.3092341423034668, "learning_rate": 9.80895183655229e-05, "loss": 0.0196, "step": 10710 }, { "grad_norm": 0.3007313013076782, "learning_rate": 9.808385552653877e-05, "loss": 0.0246, "step": 10720 }, { "grad_norm": 0.23829351365566254, "learning_rate": 9.807818447133886e-05, "loss": 0.0244, "step": 10730 }, { "grad_norm": 0.20461584627628326, "learning_rate": 9.807250520089215e-05, "loss": 0.021, "step": 10740 }, { "grad_norm": 0.19856837391853333, "learning_rate": 9.806681771616908e-05, "loss": 0.0201, "step": 10750 }, { "grad_norm": 0.38346758484840393, "learning_rate": 9.80611220181415e-05, "loss": 0.0236, "step": 10760 }, { "grad_norm": 0.284730464220047, "learning_rate": 9.805541810778264e-05, "loss": 0.0237, "step": 10770 }, { "grad_norm": 0.2834254801273346, "learning_rate": 9.804970598606716e-05, "loss": 0.0238, "step": 10780 }, { "grad_norm": 0.23701736330986023, "learning_rate": 9.804398565397106e-05, "loss": 0.0275, "step": 10790 }, { "grad_norm": 0.21856489777565002, "learning_rate": 9.803825711247183e-05, "loss": 0.0234, "step": 10800 }, { "grad_norm": 0.27854201197624207, "learning_rate": 9.803252036254831e-05, "loss": 0.0179, "step": 10810 }, { "grad_norm": 0.24467256665229797, "learning_rate": 9.802677540518076e-05, "loss": 0.0243, "step": 10820 }, { "grad_norm": 0.27034443616867065, "learning_rate": 9.802102224135081e-05, "loss": 0.0251, "step": 10830 }, { "grad_norm": 0.2677966058254242, "learning_rate": 9.801526087204155e-05, "loss": 0.0224, "step": 10840 }, { "grad_norm": 0.2957170009613037, "learning_rate": 9.800949129823743e-05, "loss": 0.0205, "step": 10850 }, { "grad_norm": 0.19793492555618286, "learning_rate": 9.80037135209243e-05, "loss": 0.0249, "step": 10860 }, { "grad_norm": 0.17372824251651764, "learning_rate": 9.799792754108946e-05, "loss": 0.0226, "step": 10870 }, { "grad_norm": 0.23676173388957977, "learning_rate": 9.799213335972152e-05, "loss": 0.0203, "step": 10880 }, { "grad_norm": 0.16660527884960175, "learning_rate": 9.798633097781058e-05, "loss": 0.0244, "step": 10890 }, { "grad_norm": 0.17037983238697052, "learning_rate": 9.79805203963481e-05, "loss": 0.0191, "step": 10900 }, { "grad_norm": 0.2789076268672943, "learning_rate": 9.797470161632697e-05, "loss": 0.0228, "step": 10910 }, { "grad_norm": 0.20417703688144684, "learning_rate": 9.796887463874145e-05, "loss": 0.0231, "step": 10920 }, { "grad_norm": 0.25550517439842224, "learning_rate": 9.796303946458718e-05, "loss": 0.0212, "step": 10930 }, { "grad_norm": 0.18050308525562286, "learning_rate": 9.795719609486127e-05, "loss": 0.0223, "step": 10940 }, { "grad_norm": 0.2530144453048706, "learning_rate": 9.795134453056219e-05, "loss": 0.0175, "step": 10950 }, { "grad_norm": 0.299027681350708, "learning_rate": 9.794548477268979e-05, "loss": 0.0189, "step": 10960 }, { "grad_norm": 0.16337402164936066, "learning_rate": 9.793961682224537e-05, "loss": 0.0196, "step": 10970 }, { "grad_norm": 0.22443214058876038, "learning_rate": 9.793374068023156e-05, "loss": 0.0223, "step": 10980 }, { "grad_norm": 0.281933456659317, "learning_rate": 9.792785634765247e-05, "loss": 0.0233, "step": 10990 }, { "grad_norm": 0.26232782006263733, "learning_rate": 9.792196382551357e-05, "loss": 0.0225, "step": 11000 }, { "grad_norm": 0.2414177507162094, "learning_rate": 9.791606311482171e-05, "loss": 0.0198, "step": 11010 }, { "grad_norm": 0.24049389362335205, "learning_rate": 9.791015421658518e-05, "loss": 0.0203, "step": 11020 }, { "grad_norm": 0.27038127183914185, "learning_rate": 9.790423713181362e-05, "loss": 0.0225, "step": 11030 }, { "grad_norm": 0.1856585294008255, "learning_rate": 9.789831186151814e-05, "loss": 0.0216, "step": 11040 }, { "grad_norm": 0.28956085443496704, "learning_rate": 9.789237840671118e-05, "loss": 0.0223, "step": 11050 }, { "grad_norm": 0.3246668577194214, "learning_rate": 9.78864367684066e-05, "loss": 0.0227, "step": 11060 }, { "grad_norm": 0.2586440443992615, "learning_rate": 9.788048694761968e-05, "loss": 0.022, "step": 11070 }, { "grad_norm": 0.3274093568325043, "learning_rate": 9.787452894536709e-05, "loss": 0.0229, "step": 11080 }, { "grad_norm": 0.19711408019065857, "learning_rate": 9.786856276266685e-05, "loss": 0.0197, "step": 11090 }, { "grad_norm": 0.22190113365650177, "learning_rate": 9.786258840053845e-05, "loss": 0.0218, "step": 11100 }, { "grad_norm": 0.20367281138896942, "learning_rate": 9.785660586000273e-05, "loss": 0.0205, "step": 11110 }, { "grad_norm": 0.217771515250206, "learning_rate": 9.785061514208196e-05, "loss": 0.0191, "step": 11120 }, { "grad_norm": 0.1872079223394394, "learning_rate": 9.784461624779977e-05, "loss": 0.0202, "step": 11130 }, { "grad_norm": 0.17307022213935852, "learning_rate": 9.783860917818123e-05, "loss": 0.0202, "step": 11140 }, { "grad_norm": 0.285285085439682, "learning_rate": 9.783259393425277e-05, "loss": 0.0224, "step": 11150 }, { "grad_norm": 0.2336975485086441, "learning_rate": 9.782657051704221e-05, "loss": 0.0177, "step": 11160 }, { "grad_norm": 0.22625291347503662, "learning_rate": 9.782053892757883e-05, "loss": 0.0199, "step": 11170 }, { "grad_norm": 0.20324388146400452, "learning_rate": 9.781449916689324e-05, "loss": 0.0236, "step": 11180 }, { "grad_norm": 0.2884662449359894, "learning_rate": 9.780845123601746e-05, "loss": 0.0214, "step": 11190 }, { "grad_norm": 0.23633801937103271, "learning_rate": 9.780239513598492e-05, "loss": 0.0216, "step": 11200 }, { "grad_norm": 0.20358680188655853, "learning_rate": 9.779633086783047e-05, "loss": 0.0238, "step": 11210 }, { "grad_norm": 0.28020980954170227, "learning_rate": 9.779025843259031e-05, "loss": 0.0197, "step": 11220 }, { "grad_norm": 0.21567083895206451, "learning_rate": 9.778417783130204e-05, "loss": 0.0208, "step": 11230 }, { "grad_norm": 0.21748997271060944, "learning_rate": 9.777808906500468e-05, "loss": 0.0189, "step": 11240 }, { "grad_norm": 0.21000462770462036, "learning_rate": 9.777199213473862e-05, "loss": 0.022, "step": 11250 }, { "grad_norm": 0.28776082396507263, "learning_rate": 9.77658870415457e-05, "loss": 0.0242, "step": 11260 }, { "grad_norm": 0.32829102873802185, "learning_rate": 9.775977378646906e-05, "loss": 0.0263, "step": 11270 }, { "grad_norm": 0.22459356486797333, "learning_rate": 9.775365237055331e-05, "loss": 0.0202, "step": 11280 }, { "grad_norm": 0.23483000695705414, "learning_rate": 9.774752279484445e-05, "loss": 0.0193, "step": 11290 }, { "grad_norm": 0.1971285045146942, "learning_rate": 9.774138506038984e-05, "loss": 0.0204, "step": 11300 }, { "grad_norm": 0.2575632929801941, "learning_rate": 9.773523916823826e-05, "loss": 0.0231, "step": 11310 }, { "grad_norm": 0.2067289799451828, "learning_rate": 9.772908511943986e-05, "loss": 0.0181, "step": 11320 }, { "grad_norm": 0.24803543090820312, "learning_rate": 9.77229229150462e-05, "loss": 0.0198, "step": 11330 }, { "grad_norm": 0.22251877188682556, "learning_rate": 9.771675255611024e-05, "loss": 0.0235, "step": 11340 }, { "grad_norm": 0.25758737325668335, "learning_rate": 9.771057404368632e-05, "loss": 0.0228, "step": 11350 }, { "grad_norm": 0.20297816395759583, "learning_rate": 9.770438737883018e-05, "loss": 0.0185, "step": 11360 }, { "grad_norm": 0.22432343661785126, "learning_rate": 9.769819256259898e-05, "loss": 0.0199, "step": 11370 }, { "grad_norm": 0.2995404601097107, "learning_rate": 9.769198959605119e-05, "loss": 0.0233, "step": 11380 }, { "grad_norm": 0.2603840231895447, "learning_rate": 9.768577848024678e-05, "loss": 0.0228, "step": 11390 }, { "grad_norm": 0.22371578216552734, "learning_rate": 9.767955921624702e-05, "loss": 0.0204, "step": 11400 }, { "grad_norm": 0.24255388975143433, "learning_rate": 9.767333180511465e-05, "loss": 0.023, "step": 11410 }, { "grad_norm": 0.22612029314041138, "learning_rate": 9.766709624791373e-05, "loss": 0.0269, "step": 11420 }, { "grad_norm": 0.2586194574832916, "learning_rate": 9.766085254570975e-05, "loss": 0.0236, "step": 11430 }, { "grad_norm": 0.3205134868621826, "learning_rate": 9.76546006995696e-05, "loss": 0.0226, "step": 11440 }, { "grad_norm": 0.28528672456741333, "learning_rate": 9.764834071056155e-05, "loss": 0.0222, "step": 11450 }, { "grad_norm": 0.21503591537475586, "learning_rate": 9.764207257975526e-05, "loss": 0.0223, "step": 11460 }, { "grad_norm": 0.1945355087518692, "learning_rate": 9.763579630822179e-05, "loss": 0.0219, "step": 11470 }, { "grad_norm": 0.2340921014547348, "learning_rate": 9.762951189703356e-05, "loss": 0.0217, "step": 11480 }, { "grad_norm": 0.2117173671722412, "learning_rate": 9.762321934726442e-05, "loss": 0.0191, "step": 11490 }, { "grad_norm": 0.27084508538246155, "learning_rate": 9.761691865998959e-05, "loss": 0.024, "step": 11500 }, { "grad_norm": 0.2395874410867691, "learning_rate": 9.76106098362857e-05, "loss": 0.0217, "step": 11510 }, { "grad_norm": 0.3085693418979645, "learning_rate": 9.760429287723072e-05, "loss": 0.0224, "step": 11520 }, { "grad_norm": 0.22012463212013245, "learning_rate": 9.759796778390406e-05, "loss": 0.0227, "step": 11530 }, { "grad_norm": 0.21555419266223907, "learning_rate": 9.759163455738653e-05, "loss": 0.0226, "step": 11540 }, { "grad_norm": 0.1964767426252365, "learning_rate": 9.75852931987603e-05, "loss": 0.0171, "step": 11550 }, { "grad_norm": 0.20163992047309875, "learning_rate": 9.757894370910891e-05, "loss": 0.0232, "step": 11560 }, { "grad_norm": 0.2494324892759323, "learning_rate": 9.757258608951733e-05, "loss": 0.0227, "step": 11570 }, { "grad_norm": 0.20367120206356049, "learning_rate": 9.75662203410719e-05, "loss": 0.0187, "step": 11580 }, { "grad_norm": 0.23401139676570892, "learning_rate": 9.755984646486034e-05, "loss": 0.0183, "step": 11590 }, { "grad_norm": 0.2253032922744751, "learning_rate": 9.75534644619718e-05, "loss": 0.0173, "step": 11600 }, { "grad_norm": 0.2313094139099121, "learning_rate": 9.754707433349676e-05, "loss": 0.0208, "step": 11610 }, { "grad_norm": 0.3299981653690338, "learning_rate": 9.754067608052715e-05, "loss": 0.0237, "step": 11620 }, { "grad_norm": 0.3725310266017914, "learning_rate": 9.753426970415622e-05, "loss": 0.0203, "step": 11630 }, { "grad_norm": 0.26184287667274475, "learning_rate": 9.752785520547868e-05, "loss": 0.0206, "step": 11640 }, { "grad_norm": 0.2565937042236328, "learning_rate": 9.752143258559056e-05, "loss": 0.0201, "step": 11650 }, { "grad_norm": 0.23180244863033295, "learning_rate": 9.751500184558933e-05, "loss": 0.02, "step": 11660 }, { "grad_norm": 0.18693843483924866, "learning_rate": 9.750856298657383e-05, "loss": 0.0174, "step": 11670 }, { "grad_norm": 0.22915230691432953, "learning_rate": 9.750211600964428e-05, "loss": 0.0199, "step": 11680 }, { "grad_norm": 0.33791911602020264, "learning_rate": 9.749566091590226e-05, "loss": 0.0235, "step": 11690 }, { "grad_norm": 0.2353965789079666, "learning_rate": 9.748919770645083e-05, "loss": 0.0224, "step": 11700 }, { "grad_norm": 0.2327210009098053, "learning_rate": 9.748272638239432e-05, "loss": 0.0212, "step": 11710 }, { "grad_norm": 0.23285925388336182, "learning_rate": 9.747624694483855e-05, "loss": 0.022, "step": 11720 }, { "grad_norm": 0.2113644927740097, "learning_rate": 9.746975939489065e-05, "loss": 0.018, "step": 11730 }, { "grad_norm": 0.25779280066490173, "learning_rate": 9.746326373365918e-05, "loss": 0.0184, "step": 11740 }, { "grad_norm": 0.3011162579059601, "learning_rate": 9.745675996225403e-05, "loss": 0.0198, "step": 11750 }, { "grad_norm": 0.22088566422462463, "learning_rate": 9.745024808178657e-05, "loss": 0.0201, "step": 11760 }, { "grad_norm": 0.22271592915058136, "learning_rate": 9.744372809336947e-05, "loss": 0.02, "step": 11770 }, { "grad_norm": 0.22347968816757202, "learning_rate": 9.743719999811682e-05, "loss": 0.0249, "step": 11780 }, { "grad_norm": 0.24478104710578918, "learning_rate": 9.743066379714412e-05, "loss": 0.018, "step": 11790 }, { "grad_norm": 0.22712597250938416, "learning_rate": 9.74241194915682e-05, "loss": 0.0256, "step": 11800 }, { "grad_norm": 0.19739416241645813, "learning_rate": 9.741756708250731e-05, "loss": 0.0193, "step": 11810 }, { "grad_norm": 0.22674211859703064, "learning_rate": 9.741100657108109e-05, "loss": 0.0234, "step": 11820 }, { "grad_norm": 0.23293668031692505, "learning_rate": 9.740443795841054e-05, "loss": 0.021, "step": 11830 }, { "grad_norm": 0.21996499598026276, "learning_rate": 9.739786124561805e-05, "loss": 0.0201, "step": 11840 }, { "grad_norm": 0.2687044143676758, "learning_rate": 9.73912764338274e-05, "loss": 0.0225, "step": 11850 }, { "grad_norm": 0.26589709520339966, "learning_rate": 9.738468352416377e-05, "loss": 0.0227, "step": 11860 }, { "grad_norm": 0.21743445098400116, "learning_rate": 9.737808251775369e-05, "loss": 0.0239, "step": 11870 }, { "grad_norm": 0.26396623253822327, "learning_rate": 9.737147341572512e-05, "loss": 0.0195, "step": 11880 }, { "grad_norm": 0.23206453025341034, "learning_rate": 9.736485621920735e-05, "loss": 0.0212, "step": 11890 }, { "grad_norm": 0.20866981148719788, "learning_rate": 9.735823092933108e-05, "loss": 0.0285, "step": 11900 }, { "grad_norm": 0.2133927047252655, "learning_rate": 9.735159754722838e-05, "loss": 0.02, "step": 11910 }, { "grad_norm": 0.29474127292633057, "learning_rate": 9.734495607403275e-05, "loss": 0.0217, "step": 11920 }, { "grad_norm": 0.2723054885864258, "learning_rate": 9.733830651087901e-05, "loss": 0.0218, "step": 11930 }, { "grad_norm": 0.2360614687204361, "learning_rate": 9.733164885890338e-05, "loss": 0.0183, "step": 11940 }, { "grad_norm": 0.20108850300312042, "learning_rate": 9.732498311924349e-05, "loss": 0.0235, "step": 11950 }, { "grad_norm": 0.24269340932369232, "learning_rate": 9.731830929303833e-05, "loss": 0.0219, "step": 11960 }, { "grad_norm": 0.26113152503967285, "learning_rate": 9.731162738142827e-05, "loss": 0.021, "step": 11970 }, { "grad_norm": 0.24774861335754395, "learning_rate": 9.730493738555506e-05, "loss": 0.0247, "step": 11980 }, { "grad_norm": 0.3034335970878601, "learning_rate": 9.729823930656186e-05, "loss": 0.0222, "step": 11990 }, { "grad_norm": 0.267693430185318, "learning_rate": 9.729153314559316e-05, "loss": 0.0227, "step": 12000 }, { "grad_norm": 0.22470679879188538, "learning_rate": 9.728481890379486e-05, "loss": 0.0242, "step": 12010 }, { "grad_norm": 0.28605735301971436, "learning_rate": 9.727809658231428e-05, "loss": 0.0226, "step": 12020 }, { "grad_norm": 0.22513772547245026, "learning_rate": 9.727136618230003e-05, "loss": 0.0235, "step": 12030 }, { "grad_norm": 0.22973431646823883, "learning_rate": 9.726462770490219e-05, "loss": 0.0233, "step": 12040 }, { "grad_norm": 0.25740838050842285, "learning_rate": 9.725788115127214e-05, "loss": 0.0191, "step": 12050 }, { "grad_norm": 0.23880887031555176, "learning_rate": 9.725112652256274e-05, "loss": 0.019, "step": 12060 }, { "grad_norm": 0.2571132481098175, "learning_rate": 9.724436381992812e-05, "loss": 0.0198, "step": 12070 }, { "grad_norm": 0.18910054862499237, "learning_rate": 9.723759304452387e-05, "loss": 0.0203, "step": 12080 }, { "grad_norm": 0.2384314388036728, "learning_rate": 9.72308141975069e-05, "loss": 0.018, "step": 12090 }, { "grad_norm": 0.22752459347248077, "learning_rate": 9.722402728003557e-05, "loss": 0.0214, "step": 12100 }, { "grad_norm": 0.19598062336444855, "learning_rate": 9.721723229326953e-05, "loss": 0.0214, "step": 12110 }, { "grad_norm": 0.23425628244876862, "learning_rate": 9.721042923836992e-05, "loss": 0.0213, "step": 12120 }, { "grad_norm": 0.22556550800800323, "learning_rate": 9.720361811649914e-05, "loss": 0.019, "step": 12130 }, { "grad_norm": 0.18114671111106873, "learning_rate": 9.719679892882106e-05, "loss": 0.0152, "step": 12140 }, { "grad_norm": 0.18827801942825317, "learning_rate": 9.718997167650085e-05, "loss": 0.0184, "step": 12150 }, { "grad_norm": 0.2264823019504547, "learning_rate": 9.718313636070515e-05, "loss": 0.023, "step": 12160 }, { "grad_norm": 0.2318115234375, "learning_rate": 9.717629298260192e-05, "loss": 0.0227, "step": 12170 }, { "grad_norm": 0.19769231975078583, "learning_rate": 9.716944154336047e-05, "loss": 0.0187, "step": 12180 }, { "grad_norm": 0.23592644929885864, "learning_rate": 9.716258204415157e-05, "loss": 0.0191, "step": 12190 }, { "grad_norm": 0.28944891691207886, "learning_rate": 9.715571448614728e-05, "loss": 0.023, "step": 12200 }, { "grad_norm": 0.20865662395954132, "learning_rate": 9.71488388705211e-05, "loss": 0.0188, "step": 12210 }, { "grad_norm": 0.24624048173427582, "learning_rate": 9.714195519844788e-05, "loss": 0.0202, "step": 12220 }, { "grad_norm": 0.18549981713294983, "learning_rate": 9.713506347110386e-05, "loss": 0.0184, "step": 12230 }, { "grad_norm": 0.29709866642951965, "learning_rate": 9.712816368966662e-05, "loss": 0.0196, "step": 12240 }, { "grad_norm": 0.2266109734773636, "learning_rate": 9.712125585531517e-05, "loss": 0.0221, "step": 12250 }, { "grad_norm": 0.22912755608558655, "learning_rate": 9.711433996922988e-05, "loss": 0.0196, "step": 12260 }, { "grad_norm": 0.28323227167129517, "learning_rate": 9.710741603259245e-05, "loss": 0.0238, "step": 12270 }, { "grad_norm": 0.2744521200656891, "learning_rate": 9.710048404658603e-05, "loss": 0.0205, "step": 12280 }, { "grad_norm": 0.21003654599189758, "learning_rate": 9.709354401239508e-05, "loss": 0.0196, "step": 12290 }, { "grad_norm": 0.21098195016384125, "learning_rate": 9.708659593120546e-05, "loss": 0.0214, "step": 12300 }, { "grad_norm": 0.21342600882053375, "learning_rate": 9.707963980420443e-05, "loss": 0.0223, "step": 12310 }, { "grad_norm": 0.21596969664096832, "learning_rate": 9.707267563258058e-05, "loss": 0.0176, "step": 12320 }, { "grad_norm": 0.19766972959041595, "learning_rate": 9.70657034175239e-05, "loss": 0.0199, "step": 12330 }, { "grad_norm": 0.2970128059387207, "learning_rate": 9.705872316022577e-05, "loss": 0.0253, "step": 12340 }, { "grad_norm": 0.17618519067764282, "learning_rate": 9.705173486187891e-05, "loss": 0.0198, "step": 12350 }, { "grad_norm": 0.2620709538459778, "learning_rate": 9.704473852367741e-05, "loss": 0.0254, "step": 12360 }, { "grad_norm": 0.2063242644071579, "learning_rate": 9.70377341468168e-05, "loss": 0.0208, "step": 12370 }, { "grad_norm": 0.2820776700973511, "learning_rate": 9.703072173249389e-05, "loss": 0.0207, "step": 12380 }, { "grad_norm": 0.2393064647912979, "learning_rate": 9.702370128190693e-05, "loss": 0.0191, "step": 12390 }, { "grad_norm": 0.19788607954978943, "learning_rate": 9.701667279625552e-05, "loss": 0.0228, "step": 12400 }, { "grad_norm": 0.16471737623214722, "learning_rate": 9.700963627674065e-05, "loss": 0.0195, "step": 12410 }, { "grad_norm": 0.19515444338321686, "learning_rate": 9.700259172456466e-05, "loss": 0.0176, "step": 12420 }, { "grad_norm": 0.19798420369625092, "learning_rate": 9.699553914093124e-05, "loss": 0.0165, "step": 12430 }, { "grad_norm": 0.24735590815544128, "learning_rate": 9.698847852704553e-05, "loss": 0.0206, "step": 12440 }, { "grad_norm": 0.22704102098941803, "learning_rate": 9.6981409884114e-05, "loss": 0.0277, "step": 12450 }, { "grad_norm": 0.22644616663455963, "learning_rate": 9.697433321334443e-05, "loss": 0.0195, "step": 12460 }, { "grad_norm": 0.3123067021369934, "learning_rate": 9.696724851594607e-05, "loss": 0.0232, "step": 12470 }, { "grad_norm": 0.18518102169036865, "learning_rate": 9.696015579312952e-05, "loss": 0.0174, "step": 12480 }, { "grad_norm": 0.2391195148229599, "learning_rate": 9.695305504610668e-05, "loss": 0.0182, "step": 12490 }, { "grad_norm": 0.2614697217941284, "learning_rate": 9.694594627609092e-05, "loss": 0.0176, "step": 12500 }, { "grad_norm": 0.20462125539779663, "learning_rate": 9.693882948429691e-05, "loss": 0.0197, "step": 12510 }, { "grad_norm": 0.2331008017063141, "learning_rate": 9.693170467194071e-05, "loss": 0.0201, "step": 12520 }, { "grad_norm": 0.25742819905281067, "learning_rate": 9.692457184023977e-05, "loss": 0.0214, "step": 12530 }, { "grad_norm": 0.26908794045448303, "learning_rate": 9.691743099041291e-05, "loss": 0.02, "step": 12540 }, { "grad_norm": 0.23548005521297455, "learning_rate": 9.691028212368027e-05, "loss": 0.0194, "step": 12550 }, { "grad_norm": 0.22155819833278656, "learning_rate": 9.690312524126342e-05, "loss": 0.0165, "step": 12560 }, { "grad_norm": 0.17966237664222717, "learning_rate": 9.689596034438527e-05, "loss": 0.0171, "step": 12570 }, { "grad_norm": 0.2010742425918579, "learning_rate": 9.688878743427012e-05, "loss": 0.0173, "step": 12580 }, { "grad_norm": 0.2549093961715698, "learning_rate": 9.688160651214359e-05, "loss": 0.0183, "step": 12590 }, { "grad_norm": 0.28901490569114685, "learning_rate": 9.687441757923273e-05, "loss": 0.02, "step": 12600 }, { "grad_norm": 0.24335475265979767, "learning_rate": 9.68672206367659e-05, "loss": 0.0197, "step": 12610 }, { "grad_norm": 0.17468833923339844, "learning_rate": 9.686001568597291e-05, "loss": 0.0206, "step": 12620 }, { "grad_norm": 0.24295170605182648, "learning_rate": 9.685280272808486e-05, "loss": 0.0182, "step": 12630 }, { "grad_norm": 0.3052903413772583, "learning_rate": 9.684558176433424e-05, "loss": 0.0197, "step": 12640 }, { "grad_norm": 0.19417797029018402, "learning_rate": 9.683835279595495e-05, "loss": 0.0182, "step": 12650 }, { "grad_norm": 0.22083096206188202, "learning_rate": 9.683111582418216e-05, "loss": 0.0153, "step": 12660 }, { "grad_norm": 0.2711217999458313, "learning_rate": 9.682387085025254e-05, "loss": 0.0165, "step": 12670 }, { "grad_norm": 0.26490098237991333, "learning_rate": 9.681661787540401e-05, "loss": 0.0198, "step": 12680 }, { "grad_norm": 0.24799922108650208, "learning_rate": 9.680935690087593e-05, "loss": 0.0209, "step": 12690 }, { "grad_norm": 0.2645299732685089, "learning_rate": 9.680208792790901e-05, "loss": 0.0174, "step": 12700 }, { "grad_norm": 0.18331444263458252, "learning_rate": 9.679481095774529e-05, "loss": 0.0162, "step": 12710 }, { "grad_norm": 0.2288442850112915, "learning_rate": 9.678752599162822e-05, "loss": 0.0193, "step": 12720 }, { "grad_norm": 0.2129957377910614, "learning_rate": 9.678023303080259e-05, "loss": 0.0176, "step": 12730 }, { "grad_norm": 0.2444288581609726, "learning_rate": 9.677293207651459e-05, "loss": 0.0191, "step": 12740 }, { "grad_norm": 0.20178069174289703, "learning_rate": 9.676562313001173e-05, "loss": 0.0186, "step": 12750 }, { "grad_norm": 0.2262328416109085, "learning_rate": 9.675830619254293e-05, "loss": 0.0189, "step": 12760 }, { "grad_norm": 0.2637610137462616, "learning_rate": 9.675098126535843e-05, "loss": 0.0205, "step": 12770 }, { "grad_norm": 0.1965716928243637, "learning_rate": 9.674364834970988e-05, "loss": 0.022, "step": 12780 }, { "grad_norm": 0.25320133566856384, "learning_rate": 9.673630744685028e-05, "loss": 0.0177, "step": 12790 }, { "grad_norm": 0.19425520300865173, "learning_rate": 9.672895855803397e-05, "loss": 0.0191, "step": 12800 }, { "grad_norm": 0.22571232914924622, "learning_rate": 9.672160168451667e-05, "loss": 0.019, "step": 12810 }, { "grad_norm": 0.2990100681781769, "learning_rate": 9.671423682755549e-05, "loss": 0.0202, "step": 12820 }, { "grad_norm": 0.252921462059021, "learning_rate": 9.670686398840888e-05, "loss": 0.0193, "step": 12830 }, { "grad_norm": 0.25662124156951904, "learning_rate": 9.669948316833664e-05, "loss": 0.0207, "step": 12840 }, { "grad_norm": 0.2220335155725479, "learning_rate": 9.669209436859997e-05, "loss": 0.0203, "step": 12850 }, { "grad_norm": 0.2849178910255432, "learning_rate": 9.66846975904614e-05, "loss": 0.0226, "step": 12860 }, { "grad_norm": 0.20663200318813324, "learning_rate": 9.667729283518483e-05, "loss": 0.0177, "step": 12870 }, { "grad_norm": 0.22346536815166473, "learning_rate": 9.666988010403557e-05, "loss": 0.0173, "step": 12880 }, { "grad_norm": 0.23831309378147125, "learning_rate": 9.66624593982802e-05, "loss": 0.0201, "step": 12890 }, { "grad_norm": 0.26248469948768616, "learning_rate": 9.665503071918675e-05, "loss": 0.0199, "step": 12900 }, { "grad_norm": 0.2517752945423126, "learning_rate": 9.664759406802456e-05, "loss": 0.0183, "step": 12910 }, { "grad_norm": 0.24715617299079895, "learning_rate": 9.664014944606437e-05, "loss": 0.0192, "step": 12920 }, { "grad_norm": 0.23433464765548706, "learning_rate": 9.663269685457822e-05, "loss": 0.0198, "step": 12930 }, { "grad_norm": 0.24293453991413116, "learning_rate": 9.662523629483962e-05, "loss": 0.0205, "step": 12940 }, { "grad_norm": 0.1740095317363739, "learning_rate": 9.661776776812333e-05, "loss": 0.0173, "step": 12950 }, { "grad_norm": 0.20230144262313843, "learning_rate": 9.661029127570553e-05, "loss": 0.019, "step": 12960 }, { "grad_norm": 0.21812960505485535, "learning_rate": 9.660280681886373e-05, "loss": 0.0175, "step": 12970 }, { "grad_norm": 0.22678661346435547, "learning_rate": 9.659531439887685e-05, "loss": 0.0192, "step": 12980 }, { "grad_norm": 0.24554947018623352, "learning_rate": 9.658781401702511e-05, "loss": 0.0241, "step": 12990 }, { "grad_norm": 0.19873136281967163, "learning_rate": 9.658030567459015e-05, "loss": 0.0172, "step": 13000 }, { "grad_norm": 0.20409783720970154, "learning_rate": 9.65727893728549e-05, "loss": 0.0197, "step": 13010 }, { "grad_norm": 0.26766324043273926, "learning_rate": 9.656526511310375e-05, "loss": 0.0184, "step": 13020 }, { "grad_norm": 0.18929125368595123, "learning_rate": 9.655773289662233e-05, "loss": 0.0166, "step": 13030 }, { "grad_norm": 0.1778789907693863, "learning_rate": 9.655019272469772e-05, "loss": 0.023, "step": 13040 }, { "grad_norm": 0.2896409034729004, "learning_rate": 9.654264459861832e-05, "loss": 0.0222, "step": 13050 }, { "grad_norm": 0.26118218898773193, "learning_rate": 9.653508851967391e-05, "loss": 0.0182, "step": 13060 }, { "grad_norm": 0.15834182500839233, "learning_rate": 9.65275244891556e-05, "loss": 0.0181, "step": 13070 }, { "grad_norm": 0.2623472213745117, "learning_rate": 9.651995250835591e-05, "loss": 0.0176, "step": 13080 }, { "grad_norm": 0.1983652114868164, "learning_rate": 9.651237257856862e-05, "loss": 0.0176, "step": 13090 }, { "grad_norm": 0.17963089048862457, "learning_rate": 9.6504784701089e-05, "loss": 0.0154, "step": 13100 }, { "grad_norm": 0.24453362822532654, "learning_rate": 9.649718887721357e-05, "loss": 0.02, "step": 13110 }, { "grad_norm": 0.2167995423078537, "learning_rate": 9.648958510824028e-05, "loss": 0.0175, "step": 13120 }, { "grad_norm": 0.20352332293987274, "learning_rate": 9.648197339546837e-05, "loss": 0.0203, "step": 13130 }, { "grad_norm": 0.15380501747131348, "learning_rate": 9.647435374019851e-05, "loss": 0.0203, "step": 13140 }, { "grad_norm": 0.2670317590236664, "learning_rate": 9.646672614373266e-05, "loss": 0.0218, "step": 13150 }, { "grad_norm": 0.2387571483850479, "learning_rate": 9.645909060737418e-05, "loss": 0.0189, "step": 13160 }, { "grad_norm": 0.22580556571483612, "learning_rate": 9.645144713242778e-05, "loss": 0.0219, "step": 13170 }, { "grad_norm": 0.285839319229126, "learning_rate": 9.64437957201995e-05, "loss": 0.0184, "step": 13180 }, { "grad_norm": 0.3285474181175232, "learning_rate": 9.643613637199678e-05, "loss": 0.0208, "step": 13190 }, { "grad_norm": 0.1518661081790924, "learning_rate": 9.642846908912839e-05, "loss": 0.0198, "step": 13200 }, { "grad_norm": 0.19372400641441345, "learning_rate": 9.642079387290444e-05, "loss": 0.0178, "step": 13210 }, { "grad_norm": 0.3763114809989929, "learning_rate": 9.641311072463644e-05, "loss": 0.0215, "step": 13220 }, { "grad_norm": 0.22381256520748138, "learning_rate": 9.640541964563722e-05, "loss": 0.0231, "step": 13230 }, { "grad_norm": 0.2480781525373459, "learning_rate": 9.639772063722096e-05, "loss": 0.0233, "step": 13240 }, { "grad_norm": 0.23770898580551147, "learning_rate": 9.639001370070324e-05, "loss": 0.0198, "step": 13250 }, { "grad_norm": 0.3088310658931732, "learning_rate": 9.638229883740095e-05, "loss": 0.0194, "step": 13260 }, { "grad_norm": 0.1621241420507431, "learning_rate": 9.637457604863233e-05, "loss": 0.0169, "step": 13270 }, { "grad_norm": 0.23033303022384644, "learning_rate": 9.636684533571703e-05, "loss": 0.0186, "step": 13280 }, { "grad_norm": 0.18699456751346588, "learning_rate": 9.635910669997599e-05, "loss": 0.0149, "step": 13290 }, { "grad_norm": 0.24491038918495178, "learning_rate": 9.635136014273154e-05, "loss": 0.0201, "step": 13300 }, { "grad_norm": 0.2503589689731598, "learning_rate": 9.634360566530735e-05, "loss": 0.0195, "step": 13310 }, { "grad_norm": 0.21669413149356842, "learning_rate": 9.633584326902845e-05, "loss": 0.0192, "step": 13320 }, { "grad_norm": 0.20877233147621155, "learning_rate": 9.632807295522124e-05, "loss": 0.0205, "step": 13330 }, { "grad_norm": 0.2800540626049042, "learning_rate": 9.632029472521342e-05, "loss": 0.0185, "step": 13340 }, { "grad_norm": 0.20940355956554413, "learning_rate": 9.631250858033409e-05, "loss": 0.0177, "step": 13350 }, { "grad_norm": 0.2826054096221924, "learning_rate": 9.630471452191371e-05, "loss": 0.0184, "step": 13360 }, { "grad_norm": 0.29505354166030884, "learning_rate": 9.629691255128405e-05, "loss": 0.0185, "step": 13370 }, { "grad_norm": 0.2878403961658478, "learning_rate": 9.628910266977825e-05, "loss": 0.0176, "step": 13380 }, { "grad_norm": 0.23909346759319305, "learning_rate": 9.628128487873083e-05, "loss": 0.0183, "step": 13390 }, { "grad_norm": 0.2278115153312683, "learning_rate": 9.627345917947761e-05, "loss": 0.0196, "step": 13400 }, { "grad_norm": 0.22357414662837982, "learning_rate": 9.626562557335579e-05, "loss": 0.016, "step": 13410 }, { "grad_norm": 0.1724267452955246, "learning_rate": 9.625778406170393e-05, "loss": 0.0162, "step": 13420 }, { "grad_norm": 0.15001016855239868, "learning_rate": 9.624993464586193e-05, "loss": 0.0187, "step": 13430 }, { "grad_norm": 0.2128591388463974, "learning_rate": 9.624207732717105e-05, "loss": 0.0206, "step": 13440 }, { "grad_norm": 0.24547003209590912, "learning_rate": 9.623421210697386e-05, "loss": 0.0193, "step": 13450 }, { "grad_norm": 0.2538815438747406, "learning_rate": 9.622633898661434e-05, "loss": 0.0201, "step": 13460 }, { "grad_norm": 0.2802901864051819, "learning_rate": 9.621845796743778e-05, "loss": 0.0222, "step": 13470 }, { "grad_norm": 0.2703281342983246, "learning_rate": 9.621056905079082e-05, "loss": 0.0185, "step": 13480 }, { "grad_norm": 0.20312055945396423, "learning_rate": 9.620267223802149e-05, "loss": 0.0169, "step": 13490 }, { "grad_norm": 0.2515767216682434, "learning_rate": 9.619476753047911e-05, "loss": 0.017, "step": 13500 }, { "grad_norm": 0.20367610454559326, "learning_rate": 9.618685492951438e-05, "loss": 0.0199, "step": 13510 }, { "grad_norm": 0.28263336420059204, "learning_rate": 9.617893443647938e-05, "loss": 0.0173, "step": 13520 }, { "grad_norm": 0.2662907540798187, "learning_rate": 9.617100605272746e-05, "loss": 0.0164, "step": 13530 }, { "grad_norm": 0.23603777587413788, "learning_rate": 9.616306977961338e-05, "loss": 0.0166, "step": 13540 }, { "grad_norm": 0.20641499757766724, "learning_rate": 9.615512561849326e-05, "loss": 0.0173, "step": 13550 }, { "grad_norm": 0.23307539522647858, "learning_rate": 9.61471735707245e-05, "loss": 0.023, "step": 13560 }, { "grad_norm": 0.3097657263278961, "learning_rate": 9.613921363766592e-05, "loss": 0.0191, "step": 13570 }, { "grad_norm": 0.2325686365365982, "learning_rate": 9.613124582067763e-05, "loss": 0.0179, "step": 13580 }, { "grad_norm": 0.1898065060377121, "learning_rate": 9.612327012112112e-05, "loss": 0.0189, "step": 13590 }, { "grad_norm": 0.2859397828578949, "learning_rate": 9.611528654035921e-05, "loss": 0.0166, "step": 13600 }, { "grad_norm": 0.18663649260997772, "learning_rate": 9.610729507975611e-05, "loss": 0.0184, "step": 13610 }, { "grad_norm": 0.21955572068691254, "learning_rate": 9.609929574067731e-05, "loss": 0.0171, "step": 13620 }, { "grad_norm": 0.2153111696243286, "learning_rate": 9.609128852448967e-05, "loss": 0.0173, "step": 13630 }, { "grad_norm": 0.24792656302452087, "learning_rate": 9.608327343256143e-05, "loss": 0.0183, "step": 13640 }, { "grad_norm": 0.2483137547969818, "learning_rate": 9.607525046626216e-05, "loss": 0.018, "step": 13650 }, { "grad_norm": 0.1641301065683365, "learning_rate": 9.606721962696272e-05, "loss": 0.015, "step": 13660 }, { "grad_norm": 0.18928930163383484, "learning_rate": 9.60591809160354e-05, "loss": 0.0177, "step": 13670 }, { "grad_norm": 0.2638288140296936, "learning_rate": 9.605113433485378e-05, "loss": 0.0177, "step": 13680 }, { "grad_norm": 0.3327926695346832, "learning_rate": 9.604307988479279e-05, "loss": 0.0192, "step": 13690 }, { "grad_norm": 0.29844462871551514, "learning_rate": 9.603501756722876e-05, "loss": 0.0188, "step": 13700 }, { "grad_norm": 0.1706239879131317, "learning_rate": 9.602694738353927e-05, "loss": 0.0214, "step": 13710 }, { "grad_norm": 0.2318049967288971, "learning_rate": 9.601886933510331e-05, "loss": 0.0213, "step": 13720 }, { "grad_norm": 0.23588179051876068, "learning_rate": 9.60107834233012e-05, "loss": 0.0175, "step": 13730 }, { "grad_norm": 0.31102654337882996, "learning_rate": 9.60026896495146e-05, "loss": 0.0219, "step": 13740 }, { "grad_norm": 0.22339271008968353, "learning_rate": 9.599458801512652e-05, "loss": 0.016, "step": 13750 }, { "grad_norm": 0.23928490281105042, "learning_rate": 9.598647852152129e-05, "loss": 0.0168, "step": 13760 }, { "grad_norm": 0.21425820887088776, "learning_rate": 9.597836117008462e-05, "loss": 0.0193, "step": 13770 }, { "grad_norm": 0.23441389203071594, "learning_rate": 9.597023596220356e-05, "loss": 0.0209, "step": 13780 }, { "grad_norm": 0.2686905264854431, "learning_rate": 9.596210289926643e-05, "loss": 0.0175, "step": 13790 }, { "grad_norm": 0.25438934564590454, "learning_rate": 9.5953961982663e-05, "loss": 0.018, "step": 13800 }, { "grad_norm": 0.22508274018764496, "learning_rate": 9.594581321378431e-05, "loss": 0.0185, "step": 13810 }, { "grad_norm": 0.18009498715400696, "learning_rate": 9.593765659402276e-05, "loss": 0.0189, "step": 13820 }, { "grad_norm": 0.18233662843704224, "learning_rate": 9.59294921247721e-05, "loss": 0.0165, "step": 13830 }, { "grad_norm": 0.27595940232276917, "learning_rate": 9.59213198074274e-05, "loss": 0.0168, "step": 13840 }, { "grad_norm": 0.15431390702724457, "learning_rate": 9.59131396433851e-05, "loss": 0.0179, "step": 13850 }, { "grad_norm": 0.18561847507953644, "learning_rate": 9.590495163404297e-05, "loss": 0.0186, "step": 13860 }, { "grad_norm": 0.21259263157844543, "learning_rate": 9.589675578080009e-05, "loss": 0.02, "step": 13870 }, { "grad_norm": 0.2545165717601776, "learning_rate": 9.588855208505694e-05, "loss": 0.0179, "step": 13880 }, { "grad_norm": 0.17123430967330933, "learning_rate": 9.588034054821529e-05, "loss": 0.0159, "step": 13890 }, { "grad_norm": 0.19636934995651245, "learning_rate": 9.587212117167826e-05, "loss": 0.0188, "step": 13900 }, { "grad_norm": 0.23446400463581085, "learning_rate": 9.586389395685033e-05, "loss": 0.0145, "step": 13910 }, { "grad_norm": 0.2727786600589752, "learning_rate": 9.585565890513733e-05, "loss": 0.0164, "step": 13920 }, { "grad_norm": 0.20927998423576355, "learning_rate": 9.584741601794636e-05, "loss": 0.017, "step": 13930 }, { "grad_norm": 0.20892561972141266, "learning_rate": 9.58391652966859e-05, "loss": 0.0161, "step": 13940 }, { "grad_norm": 0.22676165401935577, "learning_rate": 9.583090674276583e-05, "loss": 0.0187, "step": 13950 }, { "grad_norm": 0.2772424817085266, "learning_rate": 9.582264035759726e-05, "loss": 0.0185, "step": 13960 }, { "grad_norm": 0.24019134044647217, "learning_rate": 9.58143661425927e-05, "loss": 0.0186, "step": 13970 }, { "grad_norm": 0.2597690522670746, "learning_rate": 9.580608409916601e-05, "loss": 0.0197, "step": 13980 }, { "grad_norm": 0.1767072081565857, "learning_rate": 9.579779422873233e-05, "loss": 0.0211, "step": 13990 }, { "grad_norm": 0.201277494430542, "learning_rate": 9.578949653270819e-05, "loss": 0.0231, "step": 14000 }, { "grad_norm": 0.2666541337966919, "learning_rate": 9.578119101251144e-05, "loss": 0.0157, "step": 14010 }, { "grad_norm": 0.16999830305576324, "learning_rate": 9.577287766956127e-05, "loss": 0.0177, "step": 14020 }, { "grad_norm": 0.24336989223957062, "learning_rate": 9.57645565052782e-05, "loss": 0.0167, "step": 14030 }, { "grad_norm": 0.2469547539949417, "learning_rate": 9.575622752108407e-05, "loss": 0.0193, "step": 14040 }, { "grad_norm": 0.23341715335845947, "learning_rate": 9.57478907184021e-05, "loss": 0.0195, "step": 14050 }, { "grad_norm": 0.2818749248981476, "learning_rate": 9.573954609865681e-05, "loss": 0.0181, "step": 14060 }, { "grad_norm": 0.22201073169708252, "learning_rate": 9.573119366327408e-05, "loss": 0.0197, "step": 14070 }, { "grad_norm": 0.2753854990005493, "learning_rate": 9.57228334136811e-05, "loss": 0.0174, "step": 14080 }, { "grad_norm": 0.20094798505306244, "learning_rate": 9.571446535130641e-05, "loss": 0.0193, "step": 14090 }, { "grad_norm": 0.216798335313797, "learning_rate": 9.570608947757988e-05, "loss": 0.0181, "step": 14100 }, { "grad_norm": 0.2001463621854782, "learning_rate": 9.569770579393274e-05, "loss": 0.0168, "step": 14110 }, { "grad_norm": 0.17842450737953186, "learning_rate": 9.56893143017975e-05, "loss": 0.0179, "step": 14120 }, { "grad_norm": 0.17666879296302795, "learning_rate": 9.568091500260806e-05, "loss": 0.0172, "step": 14130 }, { "grad_norm": 0.2112829089164734, "learning_rate": 9.567250789779961e-05, "loss": 0.0169, "step": 14140 }, { "grad_norm": 0.20877088606357574, "learning_rate": 9.566409298880872e-05, "loss": 0.0183, "step": 14150 }, { "grad_norm": 0.20149096846580505, "learning_rate": 9.565567027707326e-05, "loss": 0.0206, "step": 14160 }, { "grad_norm": 0.22562596201896667, "learning_rate": 9.56472397640324e-05, "loss": 0.0193, "step": 14170 }, { "grad_norm": 0.2108926922082901, "learning_rate": 9.563880145112675e-05, "loss": 0.0221, "step": 14180 }, { "grad_norm": 0.21282260119915009, "learning_rate": 9.563035533979814e-05, "loss": 0.0177, "step": 14190 }, { "grad_norm": 0.1656942516565323, "learning_rate": 9.562190143148981e-05, "loss": 0.0206, "step": 14200 }, { "grad_norm": 0.21454671025276184, "learning_rate": 9.561343972764627e-05, "loss": 0.0133, "step": 14210 }, { "grad_norm": 0.1996602565050125, "learning_rate": 9.560497022971343e-05, "loss": 0.0171, "step": 14220 }, { "grad_norm": 0.46640482544898987, "learning_rate": 9.559649293913847e-05, "loss": 0.0169, "step": 14230 }, { "grad_norm": 0.19516108930110931, "learning_rate": 9.558800785736993e-05, "loss": 0.02, "step": 14240 }, { "grad_norm": 0.22699633240699768, "learning_rate": 9.557951498585767e-05, "loss": 0.0193, "step": 14250 }, { "grad_norm": 0.2022668868303299, "learning_rate": 9.557101432605293e-05, "loss": 0.017, "step": 14260 }, { "grad_norm": 0.2662869691848755, "learning_rate": 9.556250587940818e-05, "loss": 0.0184, "step": 14270 }, { "grad_norm": 0.24089279770851135, "learning_rate": 9.555398964737734e-05, "loss": 0.0189, "step": 14280 }, { "grad_norm": 0.19973328709602356, "learning_rate": 9.554546563141555e-05, "loss": 0.0178, "step": 14290 }, { "grad_norm": 0.22417576611042023, "learning_rate": 9.553693383297937e-05, "loss": 0.0205, "step": 14300 }, { "grad_norm": 0.24548542499542236, "learning_rate": 9.552839425352663e-05, "loss": 0.0168, "step": 14310 }, { "grad_norm": 0.308933824300766, "learning_rate": 9.551984689451652e-05, "loss": 0.0193, "step": 14320 }, { "grad_norm": 0.1953035593032837, "learning_rate": 9.551129175740953e-05, "loss": 0.0193, "step": 14330 }, { "grad_norm": 0.1330455243587494, "learning_rate": 9.550272884366754e-05, "loss": 0.0164, "step": 14340 }, { "grad_norm": 0.23597176373004913, "learning_rate": 9.549415815475369e-05, "loss": 0.0191, "step": 14350 }, { "grad_norm": 0.23321223258972168, "learning_rate": 9.548557969213247e-05, "loss": 0.0167, "step": 14360 }, { "grad_norm": 0.18431662023067474, "learning_rate": 9.547699345726972e-05, "loss": 0.0147, "step": 14370 }, { "grad_norm": 0.15540356934070587, "learning_rate": 9.546839945163257e-05, "loss": 0.0149, "step": 14380 }, { "grad_norm": 0.18043796718120575, "learning_rate": 9.545979767668953e-05, "loss": 0.0182, "step": 14390 }, { "grad_norm": 0.22333696484565735, "learning_rate": 9.54511881339104e-05, "loss": 0.0179, "step": 14400 }, { "grad_norm": 0.2825668454170227, "learning_rate": 9.54425708247663e-05, "loss": 0.021, "step": 14410 }, { "grad_norm": 0.21178191900253296, "learning_rate": 9.543394575072972e-05, "loss": 0.0175, "step": 14420 }, { "grad_norm": 0.17119905352592468, "learning_rate": 9.542531291327441e-05, "loss": 0.0163, "step": 14430 }, { "grad_norm": 0.19318461418151855, "learning_rate": 9.541667231387552e-05, "loss": 0.0159, "step": 14440 }, { "grad_norm": 0.23642507195472717, "learning_rate": 9.540802395400949e-05, "loss": 0.0173, "step": 14450 }, { "grad_norm": 0.21837939321994781, "learning_rate": 9.539936783515406e-05, "loss": 0.0177, "step": 14460 }, { "grad_norm": 0.19333326816558838, "learning_rate": 9.539070395878835e-05, "loss": 0.0186, "step": 14470 }, { "grad_norm": 0.18287822604179382, "learning_rate": 9.538203232639277e-05, "loss": 0.0159, "step": 14480 }, { "grad_norm": 0.13648758828639984, "learning_rate": 9.537335293944907e-05, "loss": 0.0174, "step": 14490 }, { "grad_norm": 0.21224673092365265, "learning_rate": 9.536466579944032e-05, "loss": 0.0163, "step": 14500 }, { "grad_norm": 0.2200491726398468, "learning_rate": 9.535597090785091e-05, "loss": 0.0186, "step": 14510 }, { "grad_norm": 0.222926065325737, "learning_rate": 9.534726826616656e-05, "loss": 0.0171, "step": 14520 }, { "grad_norm": 0.23199574649333954, "learning_rate": 9.53385578758743e-05, "loss": 0.0176, "step": 14530 }, { "grad_norm": 0.2046254724264145, "learning_rate": 9.532983973846252e-05, "loss": 0.0206, "step": 14540 }, { "grad_norm": 0.25222352147102356, "learning_rate": 9.53211138554209e-05, "loss": 0.0163, "step": 14550 }, { "grad_norm": 0.22240720689296722, "learning_rate": 9.531238022824047e-05, "loss": 0.0167, "step": 14560 }, { "grad_norm": 0.1880016028881073, "learning_rate": 9.530363885841355e-05, "loss": 0.0197, "step": 14570 }, { "grad_norm": 0.19763392210006714, "learning_rate": 9.52948897474338e-05, "loss": 0.0176, "step": 14580 }, { "grad_norm": 0.22174404561519623, "learning_rate": 9.528613289679622e-05, "loss": 0.0167, "step": 14590 }, { "grad_norm": 0.17914429306983948, "learning_rate": 9.52773683079971e-05, "loss": 0.0154, "step": 14600 }, { "grad_norm": 0.19894734025001526, "learning_rate": 9.526859598253407e-05, "loss": 0.0171, "step": 14610 }, { "grad_norm": 0.21400132775306702, "learning_rate": 9.525981592190609e-05, "loss": 0.0189, "step": 14620 }, { "grad_norm": 0.3182612359523773, "learning_rate": 9.525102812761342e-05, "loss": 0.0191, "step": 14630 }, { "grad_norm": 0.12119775265455246, "learning_rate": 9.524223260115768e-05, "loss": 0.0174, "step": 14640 }, { "grad_norm": 0.21156087517738342, "learning_rate": 9.523342934404175e-05, "loss": 0.0155, "step": 14650 }, { "grad_norm": 0.17518410086631775, "learning_rate": 9.522461835776989e-05, "loss": 0.021, "step": 14660 }, { "grad_norm": 0.26419615745544434, "learning_rate": 9.521579964384764e-05, "loss": 0.0161, "step": 14670 }, { "grad_norm": 0.19576025009155273, "learning_rate": 9.52069732037819e-05, "loss": 0.0185, "step": 14680 }, { "grad_norm": 0.25225329399108887, "learning_rate": 9.519813903908083e-05, "loss": 0.0185, "step": 14690 }, { "grad_norm": 0.2019978165626526, "learning_rate": 9.5189297151254e-05, "loss": 0.0183, "step": 14700 }, { "grad_norm": 0.23220981657505035, "learning_rate": 9.518044754181218e-05, "loss": 0.0148, "step": 14710 }, { "grad_norm": 0.2033349871635437, "learning_rate": 9.51715902122676e-05, "loss": 0.0186, "step": 14720 }, { "grad_norm": 0.23110073804855347, "learning_rate": 9.516272516413368e-05, "loss": 0.0177, "step": 14730 }, { "grad_norm": 0.20821796357631683, "learning_rate": 9.515385239892525e-05, "loss": 0.0213, "step": 14740 }, { "grad_norm": 0.2692054808139801, "learning_rate": 9.514497191815839e-05, "loss": 0.0193, "step": 14750 }, { "grad_norm": 0.181192547082901, "learning_rate": 9.513608372335055e-05, "loss": 0.0164, "step": 14760 }, { "grad_norm": 0.18257319927215576, "learning_rate": 9.512718781602045e-05, "loss": 0.0172, "step": 14770 }, { "grad_norm": 0.16215480864048004, "learning_rate": 9.511828419768823e-05, "loss": 0.0156, "step": 14780 }, { "grad_norm": 0.20430433750152588, "learning_rate": 9.510937286987521e-05, "loss": 0.0175, "step": 14790 }, { "grad_norm": 0.19648106396198273, "learning_rate": 9.510045383410408e-05, "loss": 0.0169, "step": 14800 }, { "grad_norm": 0.1758844256401062, "learning_rate": 9.509152709189892e-05, "loss": 0.0164, "step": 14810 }, { "grad_norm": 0.2364576756954193, "learning_rate": 9.508259264478504e-05, "loss": 0.0191, "step": 14820 }, { "grad_norm": 0.21707139909267426, "learning_rate": 9.507365049428909e-05, "loss": 0.0188, "step": 14830 }, { "grad_norm": 0.17586635053157806, "learning_rate": 9.506470064193902e-05, "loss": 0.0193, "step": 14840 }, { "grad_norm": 0.2182299941778183, "learning_rate": 9.505574308926414e-05, "loss": 0.0167, "step": 14850 }, { "grad_norm": 0.16961079835891724, "learning_rate": 9.504677783779505e-05, "loss": 0.014, "step": 14860 }, { "grad_norm": 0.18077220022678375, "learning_rate": 9.503780488906365e-05, "loss": 0.016, "step": 14870 }, { "grad_norm": 0.22547324001789093, "learning_rate": 9.502882424460319e-05, "loss": 0.0186, "step": 14880 }, { "grad_norm": 0.17867591977119446, "learning_rate": 9.501983590594821e-05, "loss": 0.0173, "step": 14890 }, { "grad_norm": 0.2028050720691681, "learning_rate": 9.501083987463455e-05, "loss": 0.0161, "step": 14900 }, { "grad_norm": 0.17167195677757263, "learning_rate": 9.500183615219942e-05, "loss": 0.0154, "step": 14910 }, { "grad_norm": 0.19220700860023499, "learning_rate": 9.49928247401813e-05, "loss": 0.0206, "step": 14920 }, { "grad_norm": 0.25050416588783264, "learning_rate": 9.498380564011997e-05, "loss": 0.0192, "step": 14930 }, { "grad_norm": 0.26076555252075195, "learning_rate": 9.497477885355656e-05, "loss": 0.0212, "step": 14940 }, { "grad_norm": 0.22352007031440735, "learning_rate": 9.496574438203353e-05, "loss": 0.0162, "step": 14950 }, { "grad_norm": 0.2665708065032959, "learning_rate": 9.495670222709459e-05, "loss": 0.0157, "step": 14960 }, { "grad_norm": 0.20749227702617645, "learning_rate": 9.494765239028483e-05, "loss": 0.023, "step": 14970 }, { "grad_norm": 0.21764115989208221, "learning_rate": 9.493859487315057e-05, "loss": 0.0173, "step": 14980 }, { "grad_norm": 0.2015552520751953, "learning_rate": 9.492952967723953e-05, "loss": 0.0165, "step": 14990 }, { "grad_norm": 0.27121618390083313, "learning_rate": 9.492045680410068e-05, "loss": 0.0188, "step": 15000 }, { "grad_norm": 0.2932228147983551, "learning_rate": 9.491137625528436e-05, "loss": 0.017, "step": 15010 }, { "grad_norm": 0.2394183725118637, "learning_rate": 9.490228803234215e-05, "loss": 0.0176, "step": 15020 }, { "grad_norm": 0.1776198446750641, "learning_rate": 9.489319213682701e-05, "loss": 0.0152, "step": 15030 }, { "grad_norm": 0.17298100888729095, "learning_rate": 9.488408857029316e-05, "loss": 0.0138, "step": 15040 }, { "grad_norm": 0.17423640191555023, "learning_rate": 9.487497733429616e-05, "loss": 0.0143, "step": 15050 }, { "grad_norm": 0.2674310505390167, "learning_rate": 9.486585843039286e-05, "loss": 0.014, "step": 15060 }, { "grad_norm": 0.18638771772384644, "learning_rate": 9.485673186014143e-05, "loss": 0.0151, "step": 15070 }, { "grad_norm": 0.21161139011383057, "learning_rate": 9.484759762510137e-05, "loss": 0.0169, "step": 15080 }, { "grad_norm": 0.17263533174991608, "learning_rate": 9.483845572683346e-05, "loss": 0.0152, "step": 15090 }, { "grad_norm": 0.16325461864471436, "learning_rate": 9.48293061668998e-05, "loss": 0.0162, "step": 15100 }, { "grad_norm": 0.2005191296339035, "learning_rate": 9.48201489468638e-05, "loss": 0.0168, "step": 15110 }, { "grad_norm": 0.14025577902793884, "learning_rate": 9.481098406829016e-05, "loss": 0.0147, "step": 15120 }, { "grad_norm": 0.18310874700546265, "learning_rate": 9.480181153274495e-05, "loss": 0.0166, "step": 15130 }, { "grad_norm": 0.24855627119541168, "learning_rate": 9.479263134179548e-05, "loss": 0.02, "step": 15140 }, { "grad_norm": 0.16985197365283966, "learning_rate": 9.478344349701039e-05, "loss": 0.0201, "step": 15150 }, { "grad_norm": 0.15032735466957092, "learning_rate": 9.477424799995964e-05, "loss": 0.0181, "step": 15160 }, { "grad_norm": 0.21628977358341217, "learning_rate": 9.476504485221448e-05, "loss": 0.0167, "step": 15170 }, { "grad_norm": 0.22928835451602936, "learning_rate": 9.475583405534748e-05, "loss": 0.0169, "step": 15180 }, { "grad_norm": 0.20214101672172546, "learning_rate": 9.474661561093251e-05, "loss": 0.0198, "step": 15190 }, { "grad_norm": 0.24087095260620117, "learning_rate": 9.473738952054478e-05, "loss": 0.0166, "step": 15200 }, { "grad_norm": 0.2216915339231491, "learning_rate": 9.472815578576073e-05, "loss": 0.0187, "step": 15210 }, { "grad_norm": 0.19907189905643463, "learning_rate": 9.471891440815817e-05, "loss": 0.0165, "step": 15220 }, { "grad_norm": 0.18786166608333588, "learning_rate": 9.470966538931621e-05, "loss": 0.0155, "step": 15230 }, { "grad_norm": 0.17045347392559052, "learning_rate": 9.470040873081525e-05, "loss": 0.0169, "step": 15240 }, { "grad_norm": 0.22977611422538757, "learning_rate": 9.469114443423698e-05, "loss": 0.0166, "step": 15250 }, { "grad_norm": 0.1917681246995926, "learning_rate": 9.468187250116445e-05, "loss": 0.0175, "step": 15260 }, { "grad_norm": 0.20041422545909882, "learning_rate": 9.467259293318197e-05, "loss": 0.0156, "step": 15270 }, { "grad_norm": 0.33648988604545593, "learning_rate": 9.466330573187514e-05, "loss": 0.0183, "step": 15280 }, { "grad_norm": 0.16208328306674957, "learning_rate": 9.46540108988309e-05, "loss": 0.0153, "step": 15290 }, { "grad_norm": 0.2382642924785614, "learning_rate": 9.46447084356375e-05, "loss": 0.0147, "step": 15300 }, { "grad_norm": 0.17117376625537872, "learning_rate": 9.463539834388447e-05, "loss": 0.0146, "step": 15310 }, { "grad_norm": 0.21283882856369019, "learning_rate": 9.462608062516263e-05, "loss": 0.0173, "step": 15320 }, { "grad_norm": 0.16136927902698517, "learning_rate": 9.461675528106413e-05, "loss": 0.0151, "step": 15330 }, { "grad_norm": 0.21558208763599396, "learning_rate": 9.460742231318244e-05, "loss": 0.0174, "step": 15340 }, { "grad_norm": 0.24186168611049652, "learning_rate": 9.459808172311229e-05, "loss": 0.018, "step": 15350 }, { "grad_norm": 0.17744502425193787, "learning_rate": 9.458873351244972e-05, "loss": 0.0169, "step": 15360 }, { "grad_norm": 0.2093506008386612, "learning_rate": 9.457937768279211e-05, "loss": 0.0191, "step": 15370 }, { "grad_norm": 0.22112402319908142, "learning_rate": 9.45700142357381e-05, "loss": 0.0189, "step": 15380 }, { "grad_norm": 0.18296417593955994, "learning_rate": 9.456064317288765e-05, "loss": 0.0173, "step": 15390 }, { "grad_norm": 0.24767854809761047, "learning_rate": 9.455126449584201e-05, "loss": 0.0194, "step": 15400 }, { "grad_norm": 0.19108565151691437, "learning_rate": 9.454187820620375e-05, "loss": 0.017, "step": 15410 }, { "grad_norm": 0.18035200238227844, "learning_rate": 9.453248430557673e-05, "loss": 0.0164, "step": 15420 }, { "grad_norm": 0.15933695435523987, "learning_rate": 9.452308279556611e-05, "loss": 0.0187, "step": 15430 }, { "grad_norm": 0.19889138638973236, "learning_rate": 9.451367367777835e-05, "loss": 0.0191, "step": 15440 }, { "grad_norm": 0.2053414136171341, "learning_rate": 9.450425695382122e-05, "loss": 0.0163, "step": 15450 }, { "grad_norm": 0.18876861035823822, "learning_rate": 9.449483262530375e-05, "loss": 0.017, "step": 15460 }, { "grad_norm": 0.20608380436897278, "learning_rate": 9.448540069383633e-05, "loss": 0.0191, "step": 15470 }, { "grad_norm": 0.2796020209789276, "learning_rate": 9.447596116103061e-05, "loss": 0.0212, "step": 15480 }, { "grad_norm": 0.19834858179092407, "learning_rate": 9.446651402849955e-05, "loss": 0.0264, "step": 15490 }, { "grad_norm": 0.2507774531841278, "learning_rate": 9.44570592978574e-05, "loss": 0.0222, "step": 15500 }, { "grad_norm": 0.22513042390346527, "learning_rate": 9.444759697071972e-05, "loss": 0.0188, "step": 15510 }, { "grad_norm": 0.17788012325763702, "learning_rate": 9.443812704870336e-05, "loss": 0.0183, "step": 15520 }, { "grad_norm": 0.17039208114147186, "learning_rate": 9.442864953342649e-05, "loss": 0.0167, "step": 15530 }, { "grad_norm": 0.15799374878406525, "learning_rate": 9.441916442650852e-05, "loss": 0.016, "step": 15540 }, { "grad_norm": 0.23018410801887512, "learning_rate": 9.440967172957023e-05, "loss": 0.0177, "step": 15550 }, { "grad_norm": 0.2660403251647949, "learning_rate": 9.440017144423364e-05, "loss": 0.0164, "step": 15560 }, { "grad_norm": 0.16290542483329773, "learning_rate": 9.439066357212209e-05, "loss": 0.0158, "step": 15570 }, { "grad_norm": 0.1743885576725006, "learning_rate": 9.438114811486022e-05, "loss": 0.0141, "step": 15580 }, { "grad_norm": 0.1824100911617279, "learning_rate": 9.4371625074074e-05, "loss": 0.0169, "step": 15590 }, { "grad_norm": 0.18937452137470245, "learning_rate": 9.436209445139059e-05, "loss": 0.017, "step": 15600 }, { "grad_norm": 0.198118656873703, "learning_rate": 9.435255624843855e-05, "loss": 0.0169, "step": 15610 }, { "grad_norm": 0.17378833889961243, "learning_rate": 9.43430104668477e-05, "loss": 0.0153, "step": 15620 }, { "grad_norm": 0.2162749469280243, "learning_rate": 9.433345710824914e-05, "loss": 0.0169, "step": 15630 }, { "grad_norm": 0.20189301669597626, "learning_rate": 9.432389617427529e-05, "loss": 0.0166, "step": 15640 }, { "grad_norm": 0.18564820289611816, "learning_rate": 9.431432766655984e-05, "loss": 0.0171, "step": 15650 }, { "grad_norm": 0.2708279490470886, "learning_rate": 9.430475158673778e-05, "loss": 0.0183, "step": 15660 }, { "grad_norm": 0.32463914155960083, "learning_rate": 9.429516793644542e-05, "loss": 0.0187, "step": 15670 }, { "grad_norm": 0.22075265645980835, "learning_rate": 9.428557671732034e-05, "loss": 0.0171, "step": 15680 }, { "grad_norm": 0.18305577337741852, "learning_rate": 9.42759779310014e-05, "loss": 0.0148, "step": 15690 }, { "grad_norm": 0.16864627599716187, "learning_rate": 9.426637157912879e-05, "loss": 0.0133, "step": 15700 }, { "grad_norm": 0.22514590620994568, "learning_rate": 9.425675766334397e-05, "loss": 0.0155, "step": 15710 }, { "grad_norm": 0.25346463918685913, "learning_rate": 9.424713618528968e-05, "loss": 0.0164, "step": 15720 }, { "grad_norm": 0.23329930007457733, "learning_rate": 9.423750714661e-05, "loss": 0.0184, "step": 15730 }, { "grad_norm": 0.28267213702201843, "learning_rate": 9.422787054895022e-05, "loss": 0.0173, "step": 15740 }, { "grad_norm": 0.20812273025512695, "learning_rate": 9.4218226393957e-05, "loss": 0.0176, "step": 15750 }, { "grad_norm": 0.2085316777229309, "learning_rate": 9.420857468327828e-05, "loss": 0.0183, "step": 15760 }, { "grad_norm": 0.18798404932022095, "learning_rate": 9.419891541856323e-05, "loss": 0.0177, "step": 15770 }, { "grad_norm": 0.2135792374610901, "learning_rate": 9.41892486014624e-05, "loss": 0.0179, "step": 15780 }, { "grad_norm": 0.21633292734622955, "learning_rate": 9.417957423362756e-05, "loss": 0.0163, "step": 15790 }, { "grad_norm": 0.20645570755004883, "learning_rate": 9.416989231671178e-05, "loss": 0.0161, "step": 15800 }, { "grad_norm": 0.19903422892093658, "learning_rate": 9.416020285236946e-05, "loss": 0.0143, "step": 15810 }, { "grad_norm": 0.1866179257631302, "learning_rate": 9.415050584225626e-05, "loss": 0.0206, "step": 15820 }, { "grad_norm": 0.18957048654556274, "learning_rate": 9.414080128802914e-05, "loss": 0.0136, "step": 15830 }, { "grad_norm": 0.1633259654045105, "learning_rate": 9.413108919134632e-05, "loss": 0.0201, "step": 15840 }, { "grad_norm": 0.1852162927389145, "learning_rate": 9.412136955386734e-05, "loss": 0.0176, "step": 15850 }, { "grad_norm": 0.2154422551393509, "learning_rate": 9.411164237725303e-05, "loss": 0.0181, "step": 15860 }, { "grad_norm": 0.20235981047153473, "learning_rate": 9.41019076631655e-05, "loss": 0.0152, "step": 15870 }, { "grad_norm": 0.21256425976753235, "learning_rate": 9.409216541326815e-05, "loss": 0.0181, "step": 15880 }, { "grad_norm": 0.1710626631975174, "learning_rate": 9.408241562922564e-05, "loss": 0.0175, "step": 15890 }, { "grad_norm": 0.22851169109344482, "learning_rate": 9.407265831270395e-05, "loss": 0.0217, "step": 15900 }, { "grad_norm": 0.2027467042207718, "learning_rate": 9.406289346537035e-05, "loss": 0.0182, "step": 15910 }, { "grad_norm": 0.16263660788536072, "learning_rate": 9.405312108889339e-05, "loss": 0.0224, "step": 15920 }, { "grad_norm": 0.20540837943553925, "learning_rate": 9.404334118494288e-05, "loss": 0.0158, "step": 15930 }, { "grad_norm": 0.2841065227985382, "learning_rate": 9.403355375518995e-05, "loss": 0.0177, "step": 15940 }, { "grad_norm": 0.19251461327075958, "learning_rate": 9.4023758801307e-05, "loss": 0.0195, "step": 15950 }, { "grad_norm": 0.20433549582958221, "learning_rate": 9.401395632496774e-05, "loss": 0.0166, "step": 15960 }, { "grad_norm": 0.19288328289985657, "learning_rate": 9.400414632784711e-05, "loss": 0.0168, "step": 15970 }, { "grad_norm": 0.2357155680656433, "learning_rate": 9.39943288116214e-05, "loss": 0.0156, "step": 15980 }, { "grad_norm": 0.20150449872016907, "learning_rate": 9.398450377796815e-05, "loss": 0.0153, "step": 15990 }, { "grad_norm": 0.21195761859416962, "learning_rate": 9.397467122856616e-05, "loss": 0.0199, "step": 16000 }, { "grad_norm": 0.22187116742134094, "learning_rate": 9.396483116509558e-05, "loss": 0.0177, "step": 16010 }, { "grad_norm": 0.1657458245754242, "learning_rate": 9.39549835892378e-05, "loss": 0.0181, "step": 16020 }, { "grad_norm": 0.15144328773021698, "learning_rate": 9.39451285026755e-05, "loss": 0.0166, "step": 16030 }, { "grad_norm": 0.1855260282754898, "learning_rate": 9.393526590709262e-05, "loss": 0.0157, "step": 16040 }, { "grad_norm": 0.17325830459594727, "learning_rate": 9.392539580417444e-05, "loss": 0.0172, "step": 16050 }, { "grad_norm": 0.1914627104997635, "learning_rate": 9.391551819560747e-05, "loss": 0.0174, "step": 16060 }, { "grad_norm": 0.21626824140548706, "learning_rate": 9.390563308307955e-05, "loss": 0.0149, "step": 16070 }, { "grad_norm": 0.19626294076442719, "learning_rate": 9.389574046827974e-05, "loss": 0.0179, "step": 16080 }, { "grad_norm": 0.184874027967453, "learning_rate": 9.388584035289845e-05, "loss": 0.0162, "step": 16090 }, { "grad_norm": 0.19669906795024872, "learning_rate": 9.387593273862732e-05, "loss": 0.0156, "step": 16100 }, { "grad_norm": 0.1762658655643463, "learning_rate": 9.386601762715929e-05, "loss": 0.0154, "step": 16110 }, { "grad_norm": 0.22227753698825836, "learning_rate": 9.38560950201886e-05, "loss": 0.0151, "step": 16120 }, { "grad_norm": 0.1862933486700058, "learning_rate": 9.384616491941071e-05, "loss": 0.0168, "step": 16130 }, { "grad_norm": 0.17803311347961426, "learning_rate": 9.383622732652245e-05, "loss": 0.0162, "step": 16140 }, { "grad_norm": 0.21574898064136505, "learning_rate": 9.382628224322187e-05, "loss": 0.0167, "step": 16150 }, { "grad_norm": 0.20011867582798004, "learning_rate": 9.381632967120829e-05, "loss": 0.0166, "step": 16160 }, { "grad_norm": 0.21673183143138885, "learning_rate": 9.380636961218235e-05, "loss": 0.0157, "step": 16170 }, { "grad_norm": 0.17446136474609375, "learning_rate": 9.379640206784597e-05, "loss": 0.0171, "step": 16180 }, { "grad_norm": 0.19700782001018524, "learning_rate": 9.378642703990229e-05, "loss": 0.0148, "step": 16190 }, { "grad_norm": 0.22514690458774567, "learning_rate": 9.37764445300558e-05, "loss": 0.0154, "step": 16200 }, { "grad_norm": 0.22747434675693512, "learning_rate": 9.376645454001222e-05, "loss": 0.0162, "step": 16210 }, { "grad_norm": 0.21013469994068146, "learning_rate": 9.375645707147858e-05, "loss": 0.0174, "step": 16220 }, { "grad_norm": 0.23189575970172882, "learning_rate": 9.374645212616316e-05, "loss": 0.0157, "step": 16230 }, { "grad_norm": 0.22253552079200745, "learning_rate": 9.373643970577555e-05, "loss": 0.0174, "step": 16240 }, { "grad_norm": 0.20317023992538452, "learning_rate": 9.372641981202659e-05, "loss": 0.0139, "step": 16250 }, { "grad_norm": 0.25173935294151306, "learning_rate": 9.37163924466284e-05, "loss": 0.0187, "step": 16260 }, { "grad_norm": 0.22008183598518372, "learning_rate": 9.370635761129438e-05, "loss": 0.0159, "step": 16270 }, { "grad_norm": 0.2198827713727951, "learning_rate": 9.36963153077392e-05, "loss": 0.0188, "step": 16280 }, { "grad_norm": 0.16604545712471008, "learning_rate": 9.368626553767888e-05, "loss": 0.0138, "step": 16290 }, { "grad_norm": 0.20477358996868134, "learning_rate": 9.367620830283057e-05, "loss": 0.0182, "step": 16300 }, { "grad_norm": 0.16951985657215118, "learning_rate": 9.366614360491281e-05, "loss": 0.0183, "step": 16310 }, { "grad_norm": 0.1850765347480774, "learning_rate": 9.365607144564539e-05, "loss": 0.0169, "step": 16320 }, { "grad_norm": 0.13354118168354034, "learning_rate": 9.364599182674934e-05, "loss": 0.014, "step": 16330 }, { "grad_norm": 0.24764461815357208, "learning_rate": 9.3635904749947e-05, "loss": 0.0144, "step": 16340 }, { "grad_norm": 0.28293555974960327, "learning_rate": 9.362581021696202e-05, "loss": 0.0185, "step": 16350 }, { "grad_norm": 0.22827427089214325, "learning_rate": 9.361570822951921e-05, "loss": 0.0206, "step": 16360 }, { "grad_norm": 0.18839265406131744, "learning_rate": 9.360559878934476e-05, "loss": 0.0158, "step": 16370 }, { "grad_norm": 0.17252494394779205, "learning_rate": 9.359548189816611e-05, "loss": 0.0141, "step": 16380 }, { "grad_norm": 0.20518650114536285, "learning_rate": 9.358535755771193e-05, "loss": 0.0159, "step": 16390 }, { "grad_norm": 0.17542128264904022, "learning_rate": 9.357522576971221e-05, "loss": 0.0189, "step": 16400 }, { "grad_norm": 0.17986515164375305, "learning_rate": 9.356508653589819e-05, "loss": 0.0164, "step": 16410 }, { "grad_norm": 0.22194993495941162, "learning_rate": 9.355493985800237e-05, "loss": 0.0178, "step": 16420 }, { "grad_norm": 0.30305662751197815, "learning_rate": 9.354478573775857e-05, "loss": 0.0153, "step": 16430 }, { "grad_norm": 0.1972101777791977, "learning_rate": 9.353462417690186e-05, "loss": 0.0152, "step": 16440 }, { "grad_norm": 0.18519209325313568, "learning_rate": 9.352445517716853e-05, "loss": 0.0167, "step": 16450 }, { "grad_norm": 0.21843321621418, "learning_rate": 9.351427874029621e-05, "loss": 0.0176, "step": 16460 }, { "grad_norm": 0.1926446557044983, "learning_rate": 9.350409486802379e-05, "loss": 0.0166, "step": 16470 }, { "grad_norm": 0.2131846398115158, "learning_rate": 9.349390356209138e-05, "loss": 0.0167, "step": 16480 }, { "grad_norm": 0.21725523471832275, "learning_rate": 9.348370482424042e-05, "loss": 0.0155, "step": 16490 }, { "grad_norm": 0.18800406157970428, "learning_rate": 9.347349865621357e-05, "loss": 0.0149, "step": 16500 }, { "grad_norm": 0.2481091320514679, "learning_rate": 9.346328505975481e-05, "loss": 0.0126, "step": 16510 }, { "grad_norm": 0.23423771560192108, "learning_rate": 9.345306403660936e-05, "loss": 0.0179, "step": 16520 }, { "grad_norm": 0.21575038135051727, "learning_rate": 9.344283558852371e-05, "loss": 0.0175, "step": 16530 }, { "grad_norm": 0.2003353089094162, "learning_rate": 9.343259971724563e-05, "loss": 0.0163, "step": 16540 }, { "grad_norm": 0.18481719493865967, "learning_rate": 9.342235642452413e-05, "loss": 0.0136, "step": 16550 }, { "grad_norm": 0.24412228167057037, "learning_rate": 9.341210571210954e-05, "loss": 0.0167, "step": 16560 }, { "grad_norm": 0.20882023870944977, "learning_rate": 9.340184758175338e-05, "loss": 0.018, "step": 16570 }, { "grad_norm": 0.16447320580482483, "learning_rate": 9.339158203520854e-05, "loss": 0.0145, "step": 16580 }, { "grad_norm": 0.15811686217784882, "learning_rate": 9.338130907422908e-05, "loss": 0.0155, "step": 16590 }, { "grad_norm": 0.20746149122714996, "learning_rate": 9.337102870057037e-05, "loss": 0.0156, "step": 16600 }, { "grad_norm": 0.15577206015586853, "learning_rate": 9.336074091598907e-05, "loss": 0.0134, "step": 16610 }, { "grad_norm": 0.2174607813358307, "learning_rate": 9.335044572224306e-05, "loss": 0.0147, "step": 16620 }, { "grad_norm": 0.21648705005645752, "learning_rate": 9.334014312109151e-05, "loss": 0.0176, "step": 16630 }, { "grad_norm": 0.2084072232246399, "learning_rate": 9.332983311429486e-05, "loss": 0.0153, "step": 16640 }, { "grad_norm": 0.2569568157196045, "learning_rate": 9.33195157036148e-05, "loss": 0.02, "step": 16650 }, { "grad_norm": 0.2307877540588379, "learning_rate": 9.330919089081432e-05, "loss": 0.0151, "step": 16660 }, { "grad_norm": 0.2151934802532196, "learning_rate": 9.32988586776576e-05, "loss": 0.0186, "step": 16670 }, { "grad_norm": 0.18874125182628632, "learning_rate": 9.328851906591016e-05, "loss": 0.0157, "step": 16680 }, { "grad_norm": 0.16613715887069702, "learning_rate": 9.327817205733875e-05, "loss": 0.0143, "step": 16690 }, { "grad_norm": 0.17120184004306793, "learning_rate": 9.326781765371142e-05, "loss": 0.0139, "step": 16700 }, { "grad_norm": 0.19419999420642853, "learning_rate": 9.325745585679741e-05, "loss": 0.017, "step": 16710 }, { "grad_norm": 0.1756373941898346, "learning_rate": 9.32470866683673e-05, "loss": 0.0155, "step": 16720 }, { "grad_norm": 0.33863914012908936, "learning_rate": 9.323671009019288e-05, "loss": 0.0156, "step": 16730 }, { "grad_norm": 0.2438642978668213, "learning_rate": 9.322632612404725e-05, "loss": 0.0193, "step": 16740 }, { "grad_norm": 0.20341092348098755, "learning_rate": 9.321593477170471e-05, "loss": 0.0164, "step": 16750 }, { "grad_norm": 0.25002750754356384, "learning_rate": 9.320553603494088e-05, "loss": 0.0167, "step": 16760 }, { "grad_norm": 0.2445005476474762, "learning_rate": 9.319512991553261e-05, "loss": 0.0171, "step": 16770 }, { "grad_norm": 0.23072217404842377, "learning_rate": 9.318471641525803e-05, "loss": 0.0161, "step": 16780 }, { "grad_norm": 0.22750265896320343, "learning_rate": 9.317429553589652e-05, "loss": 0.0142, "step": 16790 }, { "grad_norm": 0.17908407747745514, "learning_rate": 9.316386727922873e-05, "loss": 0.0161, "step": 16800 }, { "grad_norm": 0.20534121990203857, "learning_rate": 9.315343164703656e-05, "loss": 0.0123, "step": 16810 }, { "grad_norm": 0.29159414768218994, "learning_rate": 9.314298864110316e-05, "loss": 0.0154, "step": 16820 }, { "grad_norm": 0.18552787601947784, "learning_rate": 9.313253826321295e-05, "loss": 0.0152, "step": 16830 }, { "grad_norm": 0.21608412265777588, "learning_rate": 9.312208051515165e-05, "loss": 0.0142, "step": 16840 }, { "grad_norm": 0.2133447825908661, "learning_rate": 9.311161539870618e-05, "loss": 0.0166, "step": 16850 }, { "grad_norm": 0.1942083090543747, "learning_rate": 9.310114291566474e-05, "loss": 0.0145, "step": 16860 }, { "grad_norm": 0.17371827363967896, "learning_rate": 9.309066306781679e-05, "loss": 0.0145, "step": 16870 }, { "grad_norm": 0.20373061299324036, "learning_rate": 9.308017585695306e-05, "loss": 0.0148, "step": 16880 }, { "grad_norm": 0.24265919625759125, "learning_rate": 9.306968128486552e-05, "loss": 0.0162, "step": 16890 }, { "grad_norm": 0.22664374113082886, "learning_rate": 9.30591793533474e-05, "loss": 0.0172, "step": 16900 }, { "grad_norm": 0.23448695242404938, "learning_rate": 9.304867006419321e-05, "loss": 0.0147, "step": 16910 }, { "grad_norm": 0.19309574365615845, "learning_rate": 9.303815341919868e-05, "loss": 0.0183, "step": 16920 }, { "grad_norm": 0.14278322458267212, "learning_rate": 9.302762942016084e-05, "loss": 0.0135, "step": 16930 }, { "grad_norm": 0.1741003543138504, "learning_rate": 9.301709806887792e-05, "loss": 0.0147, "step": 16940 }, { "grad_norm": 0.21804402768611908, "learning_rate": 9.300655936714948e-05, "loss": 0.0171, "step": 16950 }, { "grad_norm": 0.16624535620212555, "learning_rate": 9.299601331677627e-05, "loss": 0.0183, "step": 16960 }, { "grad_norm": 0.23530995845794678, "learning_rate": 9.298545991956033e-05, "loss": 0.0184, "step": 16970 }, { "grad_norm": 0.1883377581834793, "learning_rate": 9.297489917730493e-05, "loss": 0.0152, "step": 16980 }, { "grad_norm": 0.1535339206457138, "learning_rate": 9.296433109181464e-05, "loss": 0.0138, "step": 16990 }, { "grad_norm": 0.22326801717281342, "learning_rate": 9.295375566489523e-05, "loss": 0.0157, "step": 17000 }, { "grad_norm": 0.22030118107795715, "learning_rate": 9.294317289835379e-05, "loss": 0.0199, "step": 17010 }, { "grad_norm": 0.14768822491168976, "learning_rate": 9.293258279399859e-05, "loss": 0.0172, "step": 17020 }, { "grad_norm": 0.2082662284374237, "learning_rate": 9.292198535363919e-05, "loss": 0.0142, "step": 17030 }, { "grad_norm": 0.18503613770008087, "learning_rate": 9.291138057908641e-05, "loss": 0.0164, "step": 17040 }, { "grad_norm": 0.22832205891609192, "learning_rate": 9.290076847215234e-05, "loss": 0.0175, "step": 17050 }, { "grad_norm": 0.17988304793834686, "learning_rate": 9.289014903465025e-05, "loss": 0.0171, "step": 17060 }, { "grad_norm": 0.19305795431137085, "learning_rate": 9.287952226839475e-05, "loss": 0.0166, "step": 17070 }, { "grad_norm": 0.1628657877445221, "learning_rate": 9.286888817520164e-05, "loss": 0.014, "step": 17080 }, { "grad_norm": 0.21273066103458405, "learning_rate": 9.285824675688803e-05, "loss": 0.0178, "step": 17090 }, { "grad_norm": 0.18423707783222198, "learning_rate": 9.28475980152722e-05, "loss": 0.0137, "step": 17100 }, { "grad_norm": 0.28487297892570496, "learning_rate": 9.283694195217379e-05, "loss": 0.0148, "step": 17110 }, { "grad_norm": 0.20248186588287354, "learning_rate": 9.282627856941356e-05, "loss": 0.0162, "step": 17120 }, { "grad_norm": 0.25443729758262634, "learning_rate": 9.281560786881363e-05, "loss": 0.0156, "step": 17130 }, { "grad_norm": 0.22365933656692505, "learning_rate": 9.280492985219733e-05, "loss": 0.0176, "step": 17140 }, { "grad_norm": 0.2245764136314392, "learning_rate": 9.279424452138924e-05, "loss": 0.0168, "step": 17150 }, { "grad_norm": 0.2116972655057907, "learning_rate": 9.278355187821517e-05, "loss": 0.0172, "step": 17160 }, { "grad_norm": 0.26784712076187134, "learning_rate": 9.277285192450224e-05, "loss": 0.0144, "step": 17170 }, { "grad_norm": 0.17824362218379974, "learning_rate": 9.276214466207875e-05, "loss": 0.0146, "step": 17180 }, { "grad_norm": 0.20076489448547363, "learning_rate": 9.275143009277427e-05, "loss": 0.019, "step": 17190 }, { "grad_norm": 0.17985263466835022, "learning_rate": 9.274070821841964e-05, "loss": 0.0151, "step": 17200 }, { "grad_norm": 0.2143898755311966, "learning_rate": 9.272997904084696e-05, "loss": 0.0141, "step": 17210 }, { "grad_norm": 0.20046204328536987, "learning_rate": 9.271924256188951e-05, "loss": 0.0165, "step": 17220 }, { "grad_norm": 0.22431564331054688, "learning_rate": 9.270849878338189e-05, "loss": 0.0193, "step": 17230 }, { "grad_norm": 0.16336245834827423, "learning_rate": 9.269774770715991e-05, "loss": 0.0178, "step": 17240 }, { "grad_norm": 0.19661618769168854, "learning_rate": 9.268698933506061e-05, "loss": 0.0145, "step": 17250 }, { "grad_norm": 0.1488388478755951, "learning_rate": 9.267622366892235e-05, "loss": 0.0163, "step": 17260 }, { "grad_norm": 0.1944105476140976, "learning_rate": 9.266545071058465e-05, "loss": 0.0151, "step": 17270 }, { "grad_norm": 0.15423598885536194, "learning_rate": 9.265467046188833e-05, "loss": 0.0147, "step": 17280 }, { "grad_norm": 0.1343943327665329, "learning_rate": 9.264388292467543e-05, "loss": 0.0139, "step": 17290 }, { "grad_norm": 0.19392095506191254, "learning_rate": 9.263308810078926e-05, "loss": 0.016, "step": 17300 }, { "grad_norm": 0.21537676453590393, "learning_rate": 9.262228599207434e-05, "loss": 0.0158, "step": 17310 }, { "grad_norm": 0.19642344117164612, "learning_rate": 9.261147660037647e-05, "loss": 0.016, "step": 17320 }, { "grad_norm": 0.19457611441612244, "learning_rate": 9.26006599275427e-05, "loss": 0.0183, "step": 17330 }, { "grad_norm": 0.2242508828639984, "learning_rate": 9.258983597542124e-05, "loss": 0.0153, "step": 17340 }, { "grad_norm": 0.2544739544391632, "learning_rate": 9.257900474586167e-05, "loss": 0.0181, "step": 17350 }, { "grad_norm": 0.22956262528896332, "learning_rate": 9.256816624071471e-05, "loss": 0.0147, "step": 17360 }, { "grad_norm": 0.17696543037891388, "learning_rate": 9.25573204618324e-05, "loss": 0.0137, "step": 17370 }, { "grad_norm": 0.23761248588562012, "learning_rate": 9.254646741106796e-05, "loss": 0.0153, "step": 17380 }, { "grad_norm": 0.18781046569347382, "learning_rate": 9.253560709027589e-05, "loss": 0.0156, "step": 17390 }, { "grad_norm": 0.15873503684997559, "learning_rate": 9.252473950131192e-05, "loss": 0.0165, "step": 17400 }, { "grad_norm": 0.18579566478729248, "learning_rate": 9.251386464603302e-05, "loss": 0.0159, "step": 17410 }, { "grad_norm": 0.18774931132793427, "learning_rate": 9.250298252629741e-05, "loss": 0.0153, "step": 17420 }, { "grad_norm": 0.23183225095272064, "learning_rate": 9.249209314396454e-05, "loss": 0.0157, "step": 17430 }, { "grad_norm": 0.1845218539237976, "learning_rate": 9.248119650089513e-05, "loss": 0.0162, "step": 17440 }, { "grad_norm": 0.20635589957237244, "learning_rate": 9.247029259895108e-05, "loss": 0.0163, "step": 17450 }, { "grad_norm": 0.16125896573066711, "learning_rate": 9.24593814399956e-05, "loss": 0.0137, "step": 17460 }, { "grad_norm": 0.156382218003273, "learning_rate": 9.244846302589309e-05, "loss": 0.0166, "step": 17470 }, { "grad_norm": 0.27641087770462036, "learning_rate": 9.243753735850923e-05, "loss": 0.0164, "step": 17480 }, { "grad_norm": 0.19632089138031006, "learning_rate": 9.24266044397109e-05, "loss": 0.0155, "step": 17490 }, { "grad_norm": 0.15830808877944946, "learning_rate": 9.241566427136624e-05, "loss": 0.0147, "step": 17500 }, { "grad_norm": 0.22146785259246826, "learning_rate": 9.240471685534463e-05, "loss": 0.0159, "step": 17510 }, { "grad_norm": 0.2167334258556366, "learning_rate": 9.239376219351667e-05, "loss": 0.0131, "step": 17520 }, { "grad_norm": 0.2280401587486267, "learning_rate": 9.238280028775425e-05, "loss": 0.0148, "step": 17530 }, { "grad_norm": 0.22130770981311798, "learning_rate": 9.237183113993041e-05, "loss": 0.0153, "step": 17540 }, { "grad_norm": 0.2019968032836914, "learning_rate": 9.236085475191952e-05, "loss": 0.0139, "step": 17550 }, { "grad_norm": 0.23655526340007782, "learning_rate": 9.234987112559709e-05, "loss": 0.0158, "step": 17560 }, { "grad_norm": 0.21480119228363037, "learning_rate": 9.233888026283999e-05, "loss": 0.0159, "step": 17570 }, { "grad_norm": 0.1433519572019577, "learning_rate": 9.232788216552619e-05, "loss": 0.0159, "step": 17580 }, { "grad_norm": 0.17478787899017334, "learning_rate": 9.231687683553502e-05, "loss": 0.0164, "step": 17590 }, { "grad_norm": 0.24310848116874695, "learning_rate": 9.230586427474698e-05, "loss": 0.0132, "step": 17600 }, { "grad_norm": 0.22778542339801788, "learning_rate": 9.229484448504379e-05, "loss": 0.0169, "step": 17610 }, { "grad_norm": 0.2090940773487091, "learning_rate": 9.228381746830843e-05, "loss": 0.0153, "step": 17620 }, { "grad_norm": 0.24433809518814087, "learning_rate": 9.227278322642514e-05, "loss": 0.0185, "step": 17630 }, { "grad_norm": 0.15867960453033447, "learning_rate": 9.226174176127937e-05, "loss": 0.0157, "step": 17640 }, { "grad_norm": 0.1322728395462036, "learning_rate": 9.22506930747578e-05, "loss": 0.0126, "step": 17650 }, { "grad_norm": 0.18855604529380798, "learning_rate": 9.223963716874831e-05, "loss": 0.0128, "step": 17660 }, { "grad_norm": 0.1878940612077713, "learning_rate": 9.222857404514012e-05, "loss": 0.015, "step": 17670 }, { "grad_norm": 0.2210339605808258, "learning_rate": 9.221750370582355e-05, "loss": 0.021, "step": 17680 }, { "grad_norm": 0.19558565318584442, "learning_rate": 9.220642615269028e-05, "loss": 0.0162, "step": 17690 }, { "grad_norm": 0.20437858998775482, "learning_rate": 9.219534138763311e-05, "loss": 0.0149, "step": 17700 }, { "grad_norm": 0.1471864879131317, "learning_rate": 9.218424941254613e-05, "loss": 0.0149, "step": 17710 }, { "grad_norm": 0.171221062541008, "learning_rate": 9.217315022932468e-05, "loss": 0.018, "step": 17720 }, { "grad_norm": 0.2091982066631317, "learning_rate": 9.216204383986528e-05, "loss": 0.0154, "step": 17730 }, { "grad_norm": 0.2384805828332901, "learning_rate": 9.215093024606574e-05, "loss": 0.0188, "step": 17740 }, { "grad_norm": 0.20179937779903412, "learning_rate": 9.213980944982506e-05, "loss": 0.0145, "step": 17750 }, { "grad_norm": 0.22114476561546326, "learning_rate": 9.212868145304346e-05, "loss": 0.0231, "step": 17760 }, { "grad_norm": 0.23668283224105835, "learning_rate": 9.211754625762241e-05, "loss": 0.0161, "step": 17770 }, { "grad_norm": 0.213945671916008, "learning_rate": 9.210640386546463e-05, "loss": 0.0158, "step": 17780 }, { "grad_norm": 0.1554774045944214, "learning_rate": 9.209525427847405e-05, "loss": 0.0169, "step": 17790 }, { "grad_norm": 0.21568749845027924, "learning_rate": 9.208409749855583e-05, "loss": 0.0145, "step": 17800 }, { "grad_norm": 0.22380153834819794, "learning_rate": 9.207293352761633e-05, "loss": 0.0189, "step": 17810 }, { "grad_norm": 0.1777782440185547, "learning_rate": 9.206176236756319e-05, "loss": 0.0164, "step": 17820 }, { "grad_norm": 0.18630225956439972, "learning_rate": 9.205058402030525e-05, "loss": 0.0139, "step": 17830 }, { "grad_norm": 0.20195023715496063, "learning_rate": 9.203939848775259e-05, "loss": 0.0172, "step": 17840 }, { "grad_norm": 0.2419603019952774, "learning_rate": 9.202820577181652e-05, "loss": 0.0158, "step": 17850 }, { "grad_norm": 0.22975051403045654, "learning_rate": 9.201700587440953e-05, "loss": 0.0167, "step": 17860 }, { "grad_norm": 0.23125441372394562, "learning_rate": 9.200579879744544e-05, "loss": 0.0194, "step": 17870 }, { "grad_norm": 0.17434144020080566, "learning_rate": 9.199458454283918e-05, "loss": 0.0156, "step": 17880 }, { "grad_norm": 0.28422611951828003, "learning_rate": 9.198336311250697e-05, "loss": 0.018, "step": 17890 }, { "grad_norm": 0.2194548398256302, "learning_rate": 9.197213450836626e-05, "loss": 0.0168, "step": 17900 }, { "grad_norm": 0.2738865315914154, "learning_rate": 9.19608987323357e-05, "loss": 0.0179, "step": 17910 }, { "grad_norm": 0.18120728433132172, "learning_rate": 9.194965578633517e-05, "loss": 0.0176, "step": 17920 }, { "grad_norm": 0.2072477489709854, "learning_rate": 9.193840567228582e-05, "loss": 0.0137, "step": 17930 }, { "grad_norm": 0.2094932496547699, "learning_rate": 9.192714839210994e-05, "loss": 0.0172, "step": 17940 }, { "grad_norm": 0.18869231641292572, "learning_rate": 9.19158839477311e-05, "loss": 0.013, "step": 17950 }, { "grad_norm": 0.2264290302991867, "learning_rate": 9.190461234107411e-05, "loss": 0.0187, "step": 17960 }, { "grad_norm": 0.2295251488685608, "learning_rate": 9.189333357406496e-05, "loss": 0.0154, "step": 17970 }, { "grad_norm": 0.19297988712787628, "learning_rate": 9.188204764863089e-05, "loss": 0.0148, "step": 17980 }, { "grad_norm": 0.15206381678581238, "learning_rate": 9.187075456670033e-05, "loss": 0.0175, "step": 17990 }, { "grad_norm": 0.15293127298355103, "learning_rate": 9.1859454330203e-05, "loss": 0.0146, "step": 18000 }, { "grad_norm": 0.2115863561630249, "learning_rate": 9.18481469410698e-05, "loss": 0.0162, "step": 18010 }, { "grad_norm": 0.2491154968738556, "learning_rate": 9.183683240123281e-05, "loss": 0.0155, "step": 18020 }, { "grad_norm": 0.24062032997608185, "learning_rate": 9.182551071262541e-05, "loss": 0.0151, "step": 18030 }, { "grad_norm": 0.21897129714488983, "learning_rate": 9.181418187718218e-05, "loss": 0.0156, "step": 18040 }, { "grad_norm": 0.18327610194683075, "learning_rate": 9.180284589683888e-05, "loss": 0.0174, "step": 18050 }, { "grad_norm": 0.2400822788476944, "learning_rate": 9.17915027735325e-05, "loss": 0.0179, "step": 18060 }, { "grad_norm": 0.18352659046649933, "learning_rate": 9.178015250920133e-05, "loss": 0.0176, "step": 18070 }, { "grad_norm": 0.23200705647468567, "learning_rate": 9.176879510578477e-05, "loss": 0.0148, "step": 18080 }, { "grad_norm": 0.21899163722991943, "learning_rate": 9.17574305652235e-05, "loss": 0.0158, "step": 18090 }, { "grad_norm": 0.2137572169303894, "learning_rate": 9.174605888945942e-05, "loss": 0.0155, "step": 18100 }, { "grad_norm": 0.15423010289669037, "learning_rate": 9.173468008043564e-05, "loss": 0.0139, "step": 18110 }, { "grad_norm": 0.22833023965358734, "learning_rate": 9.172329414009648e-05, "loss": 0.0208, "step": 18120 }, { "grad_norm": 0.2551521062850952, "learning_rate": 9.171190107038747e-05, "loss": 0.017, "step": 18130 }, { "grad_norm": 0.1856410801410675, "learning_rate": 9.170050087325541e-05, "loss": 0.0134, "step": 18140 }, { "grad_norm": 0.2181520313024521, "learning_rate": 9.168909355064824e-05, "loss": 0.0135, "step": 18150 }, { "grad_norm": 0.21698829531669617, "learning_rate": 9.167767910451519e-05, "loss": 0.0155, "step": 18160 }, { "grad_norm": 0.1473558247089386, "learning_rate": 9.166625753680669e-05, "loss": 0.0132, "step": 18170 }, { "grad_norm": 0.1963203102350235, "learning_rate": 9.165482884947431e-05, "loss": 0.0143, "step": 18180 }, { "grad_norm": 0.12935099005699158, "learning_rate": 9.164339304447098e-05, "loss": 0.0155, "step": 18190 }, { "grad_norm": 0.22011786699295044, "learning_rate": 9.163195012375072e-05, "loss": 0.0159, "step": 18200 }, { "grad_norm": 0.1822383552789688, "learning_rate": 9.16205000892688e-05, "loss": 0.0146, "step": 18210 }, { "grad_norm": 0.1515713632106781, "learning_rate": 9.160904294298175e-05, "loss": 0.0141, "step": 18220 }, { "grad_norm": 0.19422267377376556, "learning_rate": 9.159757868684727e-05, "loss": 0.0171, "step": 18230 }, { "grad_norm": 0.15991586446762085, "learning_rate": 9.15861073228243e-05, "loss": 0.0183, "step": 18240 }, { "grad_norm": 0.17331469058990479, "learning_rate": 9.157462885287296e-05, "loss": 0.0156, "step": 18250 }, { "grad_norm": 0.1882091611623764, "learning_rate": 9.156314327895461e-05, "loss": 0.0152, "step": 18260 }, { "grad_norm": 0.18795306980609894, "learning_rate": 9.155165060303185e-05, "loss": 0.0175, "step": 18270 }, { "grad_norm": 0.19392640888690948, "learning_rate": 9.154015082706841e-05, "loss": 0.0161, "step": 18280 }, { "grad_norm": 0.20710159838199615, "learning_rate": 9.152864395302936e-05, "loss": 0.0144, "step": 18290 }, { "grad_norm": 0.17990900576114655, "learning_rate": 9.151712998288085e-05, "loss": 0.0148, "step": 18300 }, { "grad_norm": 0.17926733195781708, "learning_rate": 9.150560891859031e-05, "loss": 0.0198, "step": 18310 }, { "grad_norm": 0.1835915446281433, "learning_rate": 9.14940807621264e-05, "loss": 0.0139, "step": 18320 }, { "grad_norm": 0.2007942795753479, "learning_rate": 9.148254551545894e-05, "loss": 0.0185, "step": 18330 }, { "grad_norm": 0.17094913125038147, "learning_rate": 9.147100318055901e-05, "loss": 0.0156, "step": 18340 }, { "grad_norm": 0.1744059920310974, "learning_rate": 9.145945375939888e-05, "loss": 0.0153, "step": 18350 }, { "grad_norm": 0.2515406906604767, "learning_rate": 9.144789725395203e-05, "loss": 0.0142, "step": 18360 }, { "grad_norm": 0.16089186072349548, "learning_rate": 9.14363336661931e-05, "loss": 0.014, "step": 18370 }, { "grad_norm": 0.23387756943702698, "learning_rate": 9.142476299809806e-05, "loss": 0.017, "step": 18380 }, { "grad_norm": 0.2014913558959961, "learning_rate": 9.1413185251644e-05, "loss": 0.0137, "step": 18390 }, { "grad_norm": 0.2008383423089981, "learning_rate": 9.140160042880923e-05, "loss": 0.0139, "step": 18400 }, { "grad_norm": 0.23514604568481445, "learning_rate": 9.139000853157327e-05, "loss": 0.0152, "step": 18410 }, { "grad_norm": 0.18037764728069305, "learning_rate": 9.137840956191688e-05, "loss": 0.0137, "step": 18420 }, { "grad_norm": 0.13571926951408386, "learning_rate": 9.136680352182199e-05, "loss": 0.0168, "step": 18430 }, { "grad_norm": 0.20956337451934814, "learning_rate": 9.135519041327177e-05, "loss": 0.0171, "step": 18440 }, { "grad_norm": 0.19641654193401337, "learning_rate": 9.134357023825058e-05, "loss": 0.0168, "step": 18450 }, { "grad_norm": 0.21104282140731812, "learning_rate": 9.133194299874398e-05, "loss": 0.0128, "step": 18460 }, { "grad_norm": 0.20396216213703156, "learning_rate": 9.132030869673876e-05, "loss": 0.0139, "step": 18470 }, { "grad_norm": 0.2029760479927063, "learning_rate": 9.130866733422288e-05, "loss": 0.0131, "step": 18480 }, { "grad_norm": 0.19328975677490234, "learning_rate": 9.129701891318556e-05, "loss": 0.0176, "step": 18490 }, { "grad_norm": 0.20780503749847412, "learning_rate": 9.128536343561718e-05, "loss": 0.0158, "step": 18500 }, { "grad_norm": 0.20701399445533752, "learning_rate": 9.127370090350934e-05, "loss": 0.0145, "step": 18510 }, { "grad_norm": 0.1652124673128128, "learning_rate": 9.126203131885487e-05, "loss": 0.0167, "step": 18520 }, { "grad_norm": 0.1976790726184845, "learning_rate": 9.125035468364775e-05, "loss": 0.014, "step": 18530 }, { "grad_norm": 0.20142331719398499, "learning_rate": 9.123867099988322e-05, "loss": 0.0154, "step": 18540 }, { "grad_norm": 0.1848927140235901, "learning_rate": 9.122698026955769e-05, "loss": 0.0169, "step": 18550 }, { "grad_norm": 0.215862438082695, "learning_rate": 9.12152824946688e-05, "loss": 0.0141, "step": 18560 }, { "grad_norm": 0.16914594173431396, "learning_rate": 9.120357767721538e-05, "loss": 0.0127, "step": 18570 }, { "grad_norm": 0.19146166741847992, "learning_rate": 9.119186581919745e-05, "loss": 0.0129, "step": 18580 }, { "grad_norm": 0.17417743802070618, "learning_rate": 9.118014692261624e-05, "loss": 0.0145, "step": 18590 }, { "grad_norm": 0.28459328413009644, "learning_rate": 9.116842098947422e-05, "loss": 0.0137, "step": 18600 }, { "grad_norm": 0.2087567299604416, "learning_rate": 9.115668802177499e-05, "loss": 0.0174, "step": 18610 }, { "grad_norm": 0.1817602515220642, "learning_rate": 9.114494802152342e-05, "loss": 0.0173, "step": 18620 }, { "grad_norm": 0.20222388207912445, "learning_rate": 9.113320099072555e-05, "loss": 0.0165, "step": 18630 }, { "grad_norm": 0.13394522666931152, "learning_rate": 9.112144693138864e-05, "loss": 0.0129, "step": 18640 }, { "grad_norm": 0.17945821583271027, "learning_rate": 9.110968584552111e-05, "loss": 0.0141, "step": 18650 }, { "grad_norm": 0.1790282130241394, "learning_rate": 9.109791773513264e-05, "loss": 0.0134, "step": 18660 }, { "grad_norm": 0.2836844325065613, "learning_rate": 9.108614260223403e-05, "loss": 0.0135, "step": 18670 }, { "grad_norm": 0.27849486470222473, "learning_rate": 9.107436044883738e-05, "loss": 0.0177, "step": 18680 }, { "grad_norm": 0.20256581902503967, "learning_rate": 9.10625712769559e-05, "loss": 0.0147, "step": 18690 }, { "grad_norm": 0.2059105783700943, "learning_rate": 9.105077508860406e-05, "loss": 0.0149, "step": 18700 }, { "grad_norm": 0.16628111898899078, "learning_rate": 9.103897188579751e-05, "loss": 0.0144, "step": 18710 }, { "grad_norm": 0.221982941031456, "learning_rate": 9.102716167055308e-05, "loss": 0.0141, "step": 18720 }, { "grad_norm": 0.2357497662305832, "learning_rate": 9.10153444448888e-05, "loss": 0.0171, "step": 18730 }, { "grad_norm": 0.1901453137397766, "learning_rate": 9.100352021082393e-05, "loss": 0.0152, "step": 18740 }, { "grad_norm": 0.1819319725036621, "learning_rate": 9.099168897037891e-05, "loss": 0.0139, "step": 18750 }, { "grad_norm": 0.12922757863998413, "learning_rate": 9.097985072557538e-05, "loss": 0.0116, "step": 18760 }, { "grad_norm": 0.20320315659046173, "learning_rate": 9.096800547843615e-05, "loss": 0.0122, "step": 18770 }, { "grad_norm": 0.2718007266521454, "learning_rate": 9.095615323098526e-05, "loss": 0.0139, "step": 18780 }, { "grad_norm": 0.14184245467185974, "learning_rate": 9.094429398524795e-05, "loss": 0.0127, "step": 18790 }, { "grad_norm": 0.23905710875988007, "learning_rate": 9.093242774325061e-05, "loss": 0.0149, "step": 18800 }, { "grad_norm": 0.17530550062656403, "learning_rate": 9.092055450702088e-05, "loss": 0.0147, "step": 18810 }, { "grad_norm": 0.19547902047634125, "learning_rate": 9.090867427858756e-05, "loss": 0.019, "step": 18820 }, { "grad_norm": 0.19305744767189026, "learning_rate": 9.089678705998066e-05, "loss": 0.0153, "step": 18830 }, { "grad_norm": 0.22772552073001862, "learning_rate": 9.088489285323139e-05, "loss": 0.0155, "step": 18840 }, { "grad_norm": 0.22432969510555267, "learning_rate": 9.087299166037212e-05, "loss": 0.0158, "step": 18850 }, { "grad_norm": 0.2519528567790985, "learning_rate": 9.086108348343647e-05, "loss": 0.0135, "step": 18860 }, { "grad_norm": 0.18351376056671143, "learning_rate": 9.08491683244592e-05, "loss": 0.0132, "step": 18870 }, { "grad_norm": 0.181998148560524, "learning_rate": 9.08372461854763e-05, "loss": 0.0148, "step": 18880 }, { "grad_norm": 0.1698172688484192, "learning_rate": 9.082531706852492e-05, "loss": 0.0182, "step": 18890 }, { "grad_norm": 0.1494116187095642, "learning_rate": 9.081338097564342e-05, "loss": 0.0133, "step": 18900 }, { "grad_norm": 0.18105441331863403, "learning_rate": 9.080143790887137e-05, "loss": 0.0143, "step": 18910 }, { "grad_norm": 0.20605793595314026, "learning_rate": 9.07894878702495e-05, "loss": 0.0147, "step": 18920 }, { "grad_norm": 0.2390253096818924, "learning_rate": 9.077753086181974e-05, "loss": 0.0157, "step": 18930 }, { "grad_norm": 0.2409971058368683, "learning_rate": 9.076556688562524e-05, "loss": 0.0139, "step": 18940 }, { "grad_norm": 0.17068344354629517, "learning_rate": 9.075359594371029e-05, "loss": 0.0153, "step": 18950 }, { "grad_norm": 0.15518593788146973, "learning_rate": 9.07416180381204e-05, "loss": 0.0158, "step": 18960 }, { "grad_norm": 0.22902202606201172, "learning_rate": 9.072963317090228e-05, "loss": 0.0145, "step": 18970 }, { "grad_norm": 0.2277349829673767, "learning_rate": 9.071764134410382e-05, "loss": 0.0146, "step": 18980 }, { "grad_norm": 0.1831721067428589, "learning_rate": 9.070564255977407e-05, "loss": 0.0155, "step": 18990 }, { "grad_norm": 0.2209821343421936, "learning_rate": 9.06936368199633e-05, "loss": 0.0164, "step": 19000 }, { "grad_norm": 0.1953006237745285, "learning_rate": 9.0681624126723e-05, "loss": 0.0148, "step": 19010 }, { "grad_norm": 0.15778358280658722, "learning_rate": 9.066960448210576e-05, "loss": 0.0142, "step": 19020 }, { "grad_norm": 0.2077333778142929, "learning_rate": 9.065757788816543e-05, "loss": 0.0159, "step": 19030 }, { "grad_norm": 0.1831001341342926, "learning_rate": 9.064554434695705e-05, "loss": 0.013, "step": 19040 }, { "grad_norm": 0.1974845826625824, "learning_rate": 9.063350386053677e-05, "loss": 0.0136, "step": 19050 }, { "grad_norm": 0.20603889226913452, "learning_rate": 9.062145643096202e-05, "loss": 0.0138, "step": 19060 }, { "grad_norm": 0.16549959778785706, "learning_rate": 9.060940206029136e-05, "loss": 0.0151, "step": 19070 }, { "grad_norm": 0.1897560954093933, "learning_rate": 9.059734075058457e-05, "loss": 0.0181, "step": 19080 }, { "grad_norm": 0.20726171135902405, "learning_rate": 9.058527250390257e-05, "loss": 0.0155, "step": 19090 }, { "grad_norm": 0.1684131771326065, "learning_rate": 9.057319732230752e-05, "loss": 0.0144, "step": 19100 }, { "grad_norm": 0.20382775366306305, "learning_rate": 9.056111520786273e-05, "loss": 0.0136, "step": 19110 }, { "grad_norm": 0.21407228708267212, "learning_rate": 9.054902616263268e-05, "loss": 0.0164, "step": 19120 }, { "grad_norm": 0.20468781888484955, "learning_rate": 9.05369301886831e-05, "loss": 0.0135, "step": 19130 }, { "grad_norm": 0.1997767984867096, "learning_rate": 9.052482728808083e-05, "loss": 0.0136, "step": 19140 }, { "grad_norm": 0.1425315886735916, "learning_rate": 9.051271746289391e-05, "loss": 0.0144, "step": 19150 }, { "grad_norm": 0.2224009931087494, "learning_rate": 9.050060071519162e-05, "loss": 0.0128, "step": 19160 }, { "grad_norm": 0.167444109916687, "learning_rate": 9.048847704704437e-05, "loss": 0.0124, "step": 19170 }, { "grad_norm": 0.19326317310333252, "learning_rate": 9.047634646052376e-05, "loss": 0.0159, "step": 19180 }, { "grad_norm": 0.14694605767726898, "learning_rate": 9.046420895770256e-05, "loss": 0.0139, "step": 19190 }, { "grad_norm": 0.20416004955768585, "learning_rate": 9.045206454065473e-05, "loss": 0.0143, "step": 19200 }, { "grad_norm": 0.1538913995027542, "learning_rate": 9.043991321145546e-05, "loss": 0.0119, "step": 19210 }, { "grad_norm": 0.13396893441677094, "learning_rate": 9.042775497218105e-05, "loss": 0.0108, "step": 19220 }, { "grad_norm": 0.24955067038536072, "learning_rate": 9.041558982490901e-05, "loss": 0.0142, "step": 19230 }, { "grad_norm": 0.23064608871936798, "learning_rate": 9.040341777171805e-05, "loss": 0.0143, "step": 19240 }, { "grad_norm": 0.1785823553800583, "learning_rate": 9.039123881468802e-05, "loss": 0.0172, "step": 19250 }, { "grad_norm": 0.18419894576072693, "learning_rate": 9.037905295589998e-05, "loss": 0.0118, "step": 19260 }, { "grad_norm": 0.14959625899791718, "learning_rate": 9.036686019743617e-05, "loss": 0.0128, "step": 19270 }, { "grad_norm": 0.17264819145202637, "learning_rate": 9.035466054137997e-05, "loss": 0.0151, "step": 19280 }, { "grad_norm": 0.14689727127552032, "learning_rate": 9.0342453989816e-05, "loss": 0.0148, "step": 19290 }, { "grad_norm": 0.2061976045370102, "learning_rate": 9.033024054483e-05, "loss": 0.0115, "step": 19300 }, { "grad_norm": 0.24607695639133453, "learning_rate": 9.031802020850894e-05, "loss": 0.0131, "step": 19310 }, { "grad_norm": 0.1970769315958023, "learning_rate": 9.030579298294092e-05, "loss": 0.0147, "step": 19320 }, { "grad_norm": 0.1360345482826233, "learning_rate": 9.029355887021524e-05, "loss": 0.0165, "step": 19330 }, { "grad_norm": 0.19522202014923096, "learning_rate": 9.028131787242238e-05, "loss": 0.0161, "step": 19340 }, { "grad_norm": 0.1535138636827469, "learning_rate": 9.026906999165399e-05, "loss": 0.013, "step": 19350 }, { "grad_norm": 0.17604191601276398, "learning_rate": 9.025681523000291e-05, "loss": 0.014, "step": 19360 }, { "grad_norm": 0.16911791265010834, "learning_rate": 9.024455358956315e-05, "loss": 0.0129, "step": 19370 }, { "grad_norm": 0.2301788628101349, "learning_rate": 9.023228507242984e-05, "loss": 0.0147, "step": 19380 }, { "grad_norm": 0.15779805183410645, "learning_rate": 9.022000968069937e-05, "loss": 0.0166, "step": 19390 }, { "grad_norm": 0.24756693840026855, "learning_rate": 9.020772741646928e-05, "loss": 0.0148, "step": 19400 }, { "grad_norm": 0.23536905646324158, "learning_rate": 9.019543828183826e-05, "loss": 0.0169, "step": 19410 }, { "grad_norm": 0.16221657395362854, "learning_rate": 9.018314227890616e-05, "loss": 0.0122, "step": 19420 }, { "grad_norm": 0.19824133813381195, "learning_rate": 9.017083940977408e-05, "loss": 0.0142, "step": 19430 }, { "grad_norm": 0.18140828609466553, "learning_rate": 9.015852967654422e-05, "loss": 0.0135, "step": 19440 }, { "grad_norm": 0.19082289934158325, "learning_rate": 9.014621308131996e-05, "loss": 0.0146, "step": 19450 }, { "grad_norm": 0.13355126976966858, "learning_rate": 9.01338896262059e-05, "loss": 0.0132, "step": 19460 }, { "grad_norm": 0.21163499355316162, "learning_rate": 9.012155931330777e-05, "loss": 0.016, "step": 19470 }, { "grad_norm": 0.19951733946800232, "learning_rate": 9.010922214473246e-05, "loss": 0.0157, "step": 19480 }, { "grad_norm": 0.16000686585903168, "learning_rate": 9.009687812258808e-05, "loss": 0.0132, "step": 19490 }, { "grad_norm": 0.19119521975517273, "learning_rate": 9.00845272489839e-05, "loss": 0.0122, "step": 19500 }, { "grad_norm": 0.2851684093475342, "learning_rate": 9.007216952603031e-05, "loss": 0.0171, "step": 19510 }, { "grad_norm": 0.21237757802009583, "learning_rate": 9.005980495583894e-05, "loss": 0.0133, "step": 19520 }, { "grad_norm": 0.17266280949115753, "learning_rate": 9.004743354052252e-05, "loss": 0.0144, "step": 19530 }, { "grad_norm": 0.13222190737724304, "learning_rate": 9.003505528219503e-05, "loss": 0.0126, "step": 19540 }, { "grad_norm": 0.2066868543624878, "learning_rate": 9.002267018297154e-05, "loss": 0.0158, "step": 19550 }, { "grad_norm": 0.22258178889751434, "learning_rate": 9.001027824496834e-05, "loss": 0.0138, "step": 19560 }, { "grad_norm": 0.1576356589794159, "learning_rate": 8.999787947030287e-05, "loss": 0.0148, "step": 19570 }, { "grad_norm": 0.19745096564292908, "learning_rate": 8.998547386109376e-05, "loss": 0.012, "step": 19580 }, { "grad_norm": 0.24102316796779633, "learning_rate": 8.997306141946073e-05, "loss": 0.0152, "step": 19590 }, { "grad_norm": 0.2350103259086609, "learning_rate": 8.996064214752481e-05, "loss": 0.0168, "step": 19600 }, { "grad_norm": 0.18151973187923431, "learning_rate": 8.994821604740806e-05, "loss": 0.0136, "step": 19610 }, { "grad_norm": 0.21060454845428467, "learning_rate": 8.993578312123377e-05, "loss": 0.0151, "step": 19620 }, { "grad_norm": 0.1440563201904297, "learning_rate": 8.992334337112639e-05, "loss": 0.0128, "step": 19630 }, { "grad_norm": 0.20390728116035461, "learning_rate": 8.991089679921154e-05, "loss": 0.0164, "step": 19640 }, { "grad_norm": 0.19067710638046265, "learning_rate": 8.989844340761599e-05, "loss": 0.0133, "step": 19650 }, { "grad_norm": 0.20547810196876526, "learning_rate": 8.988598319846768e-05, "loss": 0.0126, "step": 19660 }, { "grad_norm": 0.18603329360485077, "learning_rate": 8.987351617389574e-05, "loss": 0.014, "step": 19670 }, { "grad_norm": 0.18866066634655, "learning_rate": 8.98610423360304e-05, "loss": 0.0163, "step": 19680 }, { "grad_norm": 0.2070450335741043, "learning_rate": 8.984856168700317e-05, "loss": 0.0126, "step": 19690 }, { "grad_norm": 0.20723335444927216, "learning_rate": 8.983607422894658e-05, "loss": 0.0118, "step": 19700 }, { "grad_norm": 0.1939781904220581, "learning_rate": 8.982357996399442e-05, "loss": 0.0151, "step": 19710 }, { "grad_norm": 0.16633455455303192, "learning_rate": 8.981107889428164e-05, "loss": 0.0129, "step": 19720 }, { "grad_norm": 0.16360190510749817, "learning_rate": 8.979857102194428e-05, "loss": 0.0169, "step": 19730 }, { "grad_norm": 0.14949151873588562, "learning_rate": 8.978605634911968e-05, "loss": 0.014, "step": 19740 }, { "grad_norm": 0.22480864822864532, "learning_rate": 8.977353487794616e-05, "loss": 0.0142, "step": 19750 }, { "grad_norm": 0.20359280705451965, "learning_rate": 8.976100661056334e-05, "loss": 0.0124, "step": 19760 }, { "grad_norm": 0.25286415219306946, "learning_rate": 8.974847154911197e-05, "loss": 0.0163, "step": 19770 }, { "grad_norm": 0.24760891497135162, "learning_rate": 8.973592969573393e-05, "loss": 0.0141, "step": 19780 }, { "grad_norm": 0.23267097771167755, "learning_rate": 8.972338105257228e-05, "loss": 0.0144, "step": 19790 }, { "grad_norm": 0.14128327369689941, "learning_rate": 8.971082562177125e-05, "loss": 0.012, "step": 19800 }, { "grad_norm": 0.17237859964370728, "learning_rate": 8.96982634054762e-05, "loss": 0.0142, "step": 19810 }, { "grad_norm": 0.20075520873069763, "learning_rate": 8.96856944058337e-05, "loss": 0.0145, "step": 19820 }, { "grad_norm": 0.2532138228416443, "learning_rate": 8.967311862499144e-05, "loss": 0.016, "step": 19830 }, { "grad_norm": 0.19043655693531036, "learning_rate": 8.966053606509825e-05, "loss": 0.0153, "step": 19840 }, { "grad_norm": 0.2027634233236313, "learning_rate": 8.964794672830417e-05, "loss": 0.0128, "step": 19850 }, { "grad_norm": 0.31395742297172546, "learning_rate": 8.963535061676038e-05, "loss": 0.0158, "step": 19860 }, { "grad_norm": 0.1506483405828476, "learning_rate": 8.962274773261918e-05, "loss": 0.016, "step": 19870 }, { "grad_norm": 0.19397646188735962, "learning_rate": 8.961013807803409e-05, "loss": 0.0167, "step": 19880 }, { "grad_norm": 0.1272292584180832, "learning_rate": 8.959752165515973e-05, "loss": 0.0142, "step": 19890 }, { "grad_norm": 0.20496663451194763, "learning_rate": 8.958489846615193e-05, "loss": 0.0148, "step": 19900 }, { "grad_norm": 0.24663515388965607, "learning_rate": 8.957226851316762e-05, "loss": 0.0176, "step": 19910 }, { "grad_norm": 0.15316393971443176, "learning_rate": 8.955963179836493e-05, "loss": 0.0133, "step": 19920 }, { "grad_norm": 0.21034912765026093, "learning_rate": 8.954698832390312e-05, "loss": 0.0134, "step": 19930 }, { "grad_norm": 0.1830831915140152, "learning_rate": 8.953433809194263e-05, "loss": 0.0143, "step": 19940 }, { "grad_norm": 0.1403685361146927, "learning_rate": 8.9521681104645e-05, "loss": 0.0116, "step": 19950 }, { "grad_norm": 0.1982526183128357, "learning_rate": 8.9509017364173e-05, "loss": 0.014, "step": 19960 }, { "grad_norm": 0.23927564918994904, "learning_rate": 8.949634687269052e-05, "loss": 0.0153, "step": 19970 }, { "grad_norm": 0.22489672899246216, "learning_rate": 8.948366963236259e-05, "loss": 0.0131, "step": 19980 }, { "grad_norm": 0.15809576213359833, "learning_rate": 8.947098564535538e-05, "loss": 0.013, "step": 19990 }, { "grad_norm": 0.21632297337055206, "learning_rate": 8.945829491383627e-05, "loss": 0.0138, "step": 20000 }, { "grad_norm": 0.20399565994739532, "learning_rate": 8.944559743997374e-05, "loss": 0.0166, "step": 20010 }, { "grad_norm": 0.17033009231090546, "learning_rate": 8.943289322593746e-05, "loss": 0.0138, "step": 20020 }, { "grad_norm": 0.2633662223815918, "learning_rate": 8.942018227389821e-05, "loss": 0.015, "step": 20030 }, { "grad_norm": 0.16509316861629486, "learning_rate": 8.940746458602795e-05, "loss": 0.0139, "step": 20040 }, { "grad_norm": 0.1511375457048416, "learning_rate": 8.939474016449979e-05, "loss": 0.0129, "step": 20050 }, { "grad_norm": 0.1570492535829544, "learning_rate": 8.938200901148799e-05, "loss": 0.0142, "step": 20060 }, { "grad_norm": 0.17072664201259613, "learning_rate": 8.936927112916795e-05, "loss": 0.0151, "step": 20070 }, { "grad_norm": 0.17422153055667877, "learning_rate": 8.935652651971622e-05, "loss": 0.0131, "step": 20080 }, { "grad_norm": 0.16759337484836578, "learning_rate": 8.934377518531052e-05, "loss": 0.0163, "step": 20090 }, { "grad_norm": 0.3128379285335541, "learning_rate": 8.933101712812967e-05, "loss": 0.0184, "step": 20100 }, { "grad_norm": 0.422977089881897, "learning_rate": 8.931825235035374e-05, "loss": 0.0175, "step": 20110 }, { "grad_norm": 0.19754433631896973, "learning_rate": 8.930548085416382e-05, "loss": 0.0172, "step": 20120 }, { "grad_norm": 0.18259835243225098, "learning_rate": 8.92927026417422e-05, "loss": 0.0153, "step": 20130 }, { "grad_norm": 0.20019912719726562, "learning_rate": 8.92799177152724e-05, "loss": 0.0148, "step": 20140 }, { "grad_norm": 0.23379859328269958, "learning_rate": 8.926712607693895e-05, "loss": 0.0173, "step": 20150 }, { "grad_norm": 0.20662833750247955, "learning_rate": 8.925432772892762e-05, "loss": 0.013, "step": 20160 }, { "grad_norm": 0.18602882325649261, "learning_rate": 8.924152267342529e-05, "loss": 0.0125, "step": 20170 }, { "grad_norm": 0.18352799117565155, "learning_rate": 8.922871091261998e-05, "loss": 0.0128, "step": 20180 }, { "grad_norm": 0.13096316158771515, "learning_rate": 8.92158924487009e-05, "loss": 0.0133, "step": 20190 }, { "grad_norm": 0.16705837845802307, "learning_rate": 8.920306728385834e-05, "loss": 0.0122, "step": 20200 }, { "grad_norm": 0.16929498314857483, "learning_rate": 8.919023542028379e-05, "loss": 0.0133, "step": 20210 }, { "grad_norm": 0.2575472295284271, "learning_rate": 8.917739686016988e-05, "loss": 0.0153, "step": 20220 }, { "grad_norm": 0.13490058481693268, "learning_rate": 8.916455160571033e-05, "loss": 0.0117, "step": 20230 }, { "grad_norm": 0.21043115854263306, "learning_rate": 8.915169965910008e-05, "loss": 0.0127, "step": 20240 }, { "grad_norm": 0.22035078704357147, "learning_rate": 8.913884102253514e-05, "loss": 0.0168, "step": 20250 }, { "grad_norm": 0.2644292414188385, "learning_rate": 8.912597569821273e-05, "loss": 0.0172, "step": 20260 }, { "grad_norm": 0.21576066315174103, "learning_rate": 8.911310368833118e-05, "loss": 0.0142, "step": 20270 }, { "grad_norm": 0.19427929818630219, "learning_rate": 8.910022499508994e-05, "loss": 0.0149, "step": 20280 }, { "grad_norm": 0.18356621265411377, "learning_rate": 8.908733962068965e-05, "loss": 0.0128, "step": 20290 }, { "grad_norm": 0.1710035651922226, "learning_rate": 8.907444756733207e-05, "loss": 0.0129, "step": 20300 }, { "grad_norm": 0.16924183070659637, "learning_rate": 8.906154883722006e-05, "loss": 0.0167, "step": 20310 }, { "grad_norm": 0.20146918296813965, "learning_rate": 8.904864343255773e-05, "loss": 0.0129, "step": 20320 }, { "grad_norm": 0.18225741386413574, "learning_rate": 8.90357313555502e-05, "loss": 0.0136, "step": 20330 }, { "grad_norm": 0.17830154299736023, "learning_rate": 8.90228126084038e-05, "loss": 0.0135, "step": 20340 }, { "grad_norm": 0.19678881764411926, "learning_rate": 8.900988719332601e-05, "loss": 0.0149, "step": 20350 }, { "grad_norm": 0.1701298952102661, "learning_rate": 8.899695511252542e-05, "loss": 0.0138, "step": 20360 }, { "grad_norm": 0.19713902473449707, "learning_rate": 8.898401636821176e-05, "loss": 0.0144, "step": 20370 }, { "grad_norm": 0.1758105307817459, "learning_rate": 8.897107096259593e-05, "loss": 0.0157, "step": 20380 }, { "grad_norm": 0.21636728942394257, "learning_rate": 8.895811889788994e-05, "loss": 0.0159, "step": 20390 }, { "grad_norm": 0.16502691805362701, "learning_rate": 8.894516017630692e-05, "loss": 0.0151, "step": 20400 }, { "grad_norm": 0.2873676121234894, "learning_rate": 8.893219480006118e-05, "loss": 0.0148, "step": 20410 }, { "grad_norm": 0.14899688959121704, "learning_rate": 8.891922277136817e-05, "loss": 0.0097, "step": 20420 }, { "grad_norm": 0.20323479175567627, "learning_rate": 8.890624409244441e-05, "loss": 0.0116, "step": 20430 }, { "grad_norm": 0.15658889710903168, "learning_rate": 8.889325876550763e-05, "loss": 0.0161, "step": 20440 }, { "grad_norm": 0.18018978834152222, "learning_rate": 8.888026679277666e-05, "loss": 0.0133, "step": 20450 }, { "grad_norm": 0.15479981899261475, "learning_rate": 8.886726817647147e-05, "loss": 0.0142, "step": 20460 }, { "grad_norm": 0.21088925004005432, "learning_rate": 8.885426291881319e-05, "loss": 0.0137, "step": 20470 }, { "grad_norm": 0.16753162443637848, "learning_rate": 8.884125102202401e-05, "loss": 0.0147, "step": 20480 }, { "grad_norm": 0.19618766009807587, "learning_rate": 8.882823248832736e-05, "loss": 0.0169, "step": 20490 }, { "grad_norm": 0.1579122543334961, "learning_rate": 8.881520731994772e-05, "loss": 0.0148, "step": 20500 }, { "grad_norm": 0.1987248659133911, "learning_rate": 8.880217551911077e-05, "loss": 0.0126, "step": 20510 }, { "grad_norm": 0.18648721277713776, "learning_rate": 8.878913708804323e-05, "loss": 0.0181, "step": 20520 }, { "grad_norm": 0.14049747586250305, "learning_rate": 8.877609202897308e-05, "loss": 0.016, "step": 20530 }, { "grad_norm": 0.1629110425710678, "learning_rate": 8.876304034412933e-05, "loss": 0.0113, "step": 20540 }, { "grad_norm": 0.16397501528263092, "learning_rate": 8.874998203574214e-05, "loss": 0.014, "step": 20550 }, { "grad_norm": 0.17862199246883392, "learning_rate": 8.873691710604284e-05, "loss": 0.0159, "step": 20560 }, { "grad_norm": 0.2574719786643982, "learning_rate": 8.872384555726387e-05, "loss": 0.0141, "step": 20570 }, { "grad_norm": 0.20427173376083374, "learning_rate": 8.871076739163878e-05, "loss": 0.0155, "step": 20580 }, { "grad_norm": 0.15796862542629242, "learning_rate": 8.86976826114023e-05, "loss": 0.0141, "step": 20590 }, { "grad_norm": 0.14524494111537933, "learning_rate": 8.868459121879023e-05, "loss": 0.0117, "step": 20600 }, { "grad_norm": 0.16531553864479065, "learning_rate": 8.867149321603956e-05, "loss": 0.0135, "step": 20610 }, { "grad_norm": 0.16412927210330963, "learning_rate": 8.865838860538835e-05, "loss": 0.0111, "step": 20620 }, { "grad_norm": 0.20069542527198792, "learning_rate": 8.864527738907585e-05, "loss": 0.0154, "step": 20630 }, { "grad_norm": 0.1682129055261612, "learning_rate": 8.863215956934239e-05, "loss": 0.0137, "step": 20640 }, { "grad_norm": 0.1632651686668396, "learning_rate": 8.861903514842947e-05, "loss": 0.0164, "step": 20650 }, { "grad_norm": 0.12922824919223785, "learning_rate": 8.860590412857966e-05, "loss": 0.0167, "step": 20660 }, { "grad_norm": 0.17531393468379974, "learning_rate": 8.85927665120367e-05, "loss": 0.0158, "step": 20670 }, { "grad_norm": 0.1869802325963974, "learning_rate": 8.857962230104546e-05, "loss": 0.0142, "step": 20680 }, { "grad_norm": 0.18937397003173828, "learning_rate": 8.856647149785193e-05, "loss": 0.0142, "step": 20690 }, { "grad_norm": 0.177892804145813, "learning_rate": 8.855331410470322e-05, "loss": 0.0149, "step": 20700 }, { "grad_norm": 0.17705802619457245, "learning_rate": 8.854015012384756e-05, "loss": 0.0143, "step": 20710 }, { "grad_norm": 0.21989165246486664, "learning_rate": 8.852697955753433e-05, "loss": 0.0133, "step": 20720 }, { "grad_norm": 0.21125897765159607, "learning_rate": 8.851380240801399e-05, "loss": 0.0165, "step": 20730 }, { "grad_norm": 0.15069331228733063, "learning_rate": 8.850061867753818e-05, "loss": 0.0168, "step": 20740 }, { "grad_norm": 0.18950539827346802, "learning_rate": 8.848742836835963e-05, "loss": 0.0146, "step": 20750 }, { "grad_norm": 0.18876218795776367, "learning_rate": 8.847423148273221e-05, "loss": 0.0156, "step": 20760 }, { "grad_norm": 0.1936938315629959, "learning_rate": 8.846102802291092e-05, "loss": 0.0162, "step": 20770 }, { "grad_norm": 0.17662973701953888, "learning_rate": 8.844781799115183e-05, "loss": 0.0123, "step": 20780 }, { "grad_norm": 0.188248872756958, "learning_rate": 8.84346013897122e-05, "loss": 0.0127, "step": 20790 }, { "grad_norm": 0.18770219385623932, "learning_rate": 8.842137822085038e-05, "loss": 0.0136, "step": 20800 }, { "grad_norm": 0.22623059153556824, "learning_rate": 8.840814848682585e-05, "loss": 0.0119, "step": 20810 }, { "grad_norm": 0.16464875638484955, "learning_rate": 8.83949121898992e-05, "loss": 0.0136, "step": 20820 }, { "grad_norm": 0.19643379747867584, "learning_rate": 8.838166933233217e-05, "loss": 0.0154, "step": 20830 }, { "grad_norm": 0.22174789011478424, "learning_rate": 8.83684199163876e-05, "loss": 0.0133, "step": 20840 }, { "grad_norm": 0.19293873012065887, "learning_rate": 8.835516394432943e-05, "loss": 0.0138, "step": 20850 }, { "grad_norm": 0.20653818547725677, "learning_rate": 8.834190141842276e-05, "loss": 0.0147, "step": 20860 }, { "grad_norm": 0.12067646533250809, "learning_rate": 8.83286323409338e-05, "loss": 0.0125, "step": 20870 }, { "grad_norm": 0.1936495453119278, "learning_rate": 8.831535671412986e-05, "loss": 0.0137, "step": 20880 }, { "grad_norm": 0.19554246962070465, "learning_rate": 8.830207454027938e-05, "loss": 0.0135, "step": 20890 }, { "grad_norm": 0.16039274632930756, "learning_rate": 8.828878582165192e-05, "loss": 0.0117, "step": 20900 }, { "grad_norm": 0.17845940589904785, "learning_rate": 8.827549056051818e-05, "loss": 0.0142, "step": 20910 }, { "grad_norm": 0.23462650179862976, "learning_rate": 8.826218875914993e-05, "loss": 0.013, "step": 20920 }, { "grad_norm": 0.20699866116046906, "learning_rate": 8.82488804198201e-05, "loss": 0.0138, "step": 20930 }, { "grad_norm": 0.1921994984149933, "learning_rate": 8.82355655448027e-05, "loss": 0.0171, "step": 20940 }, { "grad_norm": 0.21071040630340576, "learning_rate": 8.822224413637293e-05, "loss": 0.012, "step": 20950 }, { "grad_norm": 0.18586339056491852, "learning_rate": 8.820891619680697e-05, "loss": 0.0127, "step": 20960 }, { "grad_norm": 0.12567588686943054, "learning_rate": 8.819558172838227e-05, "loss": 0.0136, "step": 20970 }, { "grad_norm": 0.15329620242118835, "learning_rate": 8.818224073337731e-05, "loss": 0.0145, "step": 20980 }, { "grad_norm": 0.18853460252285004, "learning_rate": 8.816889321407169e-05, "loss": 0.0113, "step": 20990 }, { "grad_norm": 0.20020808279514313, "learning_rate": 8.815553917274615e-05, "loss": 0.0167, "step": 21000 }, { "grad_norm": 0.18812070786952972, "learning_rate": 8.81421786116825e-05, "loss": 0.014, "step": 21010 }, { "grad_norm": 0.20136946439743042, "learning_rate": 8.812881153316373e-05, "loss": 0.015, "step": 21020 }, { "grad_norm": 0.19995544850826263, "learning_rate": 8.81154379394739e-05, "loss": 0.014, "step": 21030 }, { "grad_norm": 0.1922861635684967, "learning_rate": 8.810205783289818e-05, "loss": 0.0139, "step": 21040 }, { "grad_norm": 0.1567908525466919, "learning_rate": 8.808867121572286e-05, "loss": 0.0172, "step": 21050 }, { "grad_norm": 0.17735396325588226, "learning_rate": 8.807527809023537e-05, "loss": 0.0127, "step": 21060 }, { "grad_norm": 0.17014843225479126, "learning_rate": 8.80618784587242e-05, "loss": 0.0172, "step": 21070 }, { "grad_norm": 0.19468118250370026, "learning_rate": 8.804847232347902e-05, "loss": 0.0145, "step": 21080 }, { "grad_norm": 0.2301924228668213, "learning_rate": 8.803505968679054e-05, "loss": 0.015, "step": 21090 }, { "grad_norm": 0.20528210699558258, "learning_rate": 8.802164055095061e-05, "loss": 0.0139, "step": 21100 }, { "grad_norm": 0.20582729578018188, "learning_rate": 8.80082149182522e-05, "loss": 0.0166, "step": 21110 }, { "grad_norm": 0.1621813327074051, "learning_rate": 8.79947827909894e-05, "loss": 0.0123, "step": 21120 }, { "grad_norm": 0.22177107632160187, "learning_rate": 8.798134417145738e-05, "loss": 0.0151, "step": 21130 }, { "grad_norm": 0.2559068500995636, "learning_rate": 8.796789906195243e-05, "loss": 0.0155, "step": 21140 }, { "grad_norm": 0.1554238349199295, "learning_rate": 8.795444746477195e-05, "loss": 0.0166, "step": 21150 }, { "grad_norm": 0.2034590244293213, "learning_rate": 8.794098938221446e-05, "loss": 0.0142, "step": 21160 }, { "grad_norm": 0.18295128643512726, "learning_rate": 8.792752481657957e-05, "loss": 0.0139, "step": 21170 }, { "grad_norm": 0.2232232540845871, "learning_rate": 8.791405377016802e-05, "loss": 0.0164, "step": 21180 }, { "grad_norm": 0.18857412040233612, "learning_rate": 8.790057624528163e-05, "loss": 0.0161, "step": 21190 }, { "grad_norm": 0.18944978713989258, "learning_rate": 8.788709224422333e-05, "loss": 0.014, "step": 21200 }, { "grad_norm": 0.18817360699176788, "learning_rate": 8.787360176929717e-05, "loss": 0.0119, "step": 21210 }, { "grad_norm": 0.15551312267780304, "learning_rate": 8.786010482280834e-05, "loss": 0.0141, "step": 21220 }, { "grad_norm": 0.16087478399276733, "learning_rate": 8.784660140706306e-05, "loss": 0.0148, "step": 21230 }, { "grad_norm": 0.17199833691120148, "learning_rate": 8.783309152436872e-05, "loss": 0.0135, "step": 21240 }, { "grad_norm": 0.1458941102027893, "learning_rate": 8.781957517703375e-05, "loss": 0.0121, "step": 21250 }, { "grad_norm": 0.10452006757259369, "learning_rate": 8.780605236736776e-05, "loss": 0.0138, "step": 21260 }, { "grad_norm": 0.21070238947868347, "learning_rate": 8.779252309768142e-05, "loss": 0.0144, "step": 21270 }, { "grad_norm": 0.16083364188671112, "learning_rate": 8.777898737028652e-05, "loss": 0.0137, "step": 21280 }, { "grad_norm": 0.16639548540115356, "learning_rate": 8.776544518749591e-05, "loss": 0.0141, "step": 21290 }, { "grad_norm": 0.20902955532073975, "learning_rate": 8.775189655162364e-05, "loss": 0.0129, "step": 21300 }, { "grad_norm": 0.20885387063026428, "learning_rate": 8.773834146498474e-05, "loss": 0.0166, "step": 21310 }, { "grad_norm": 0.14471666514873505, "learning_rate": 8.772477992989545e-05, "loss": 0.0139, "step": 21320 }, { "grad_norm": 0.1302000880241394, "learning_rate": 8.771121194867304e-05, "loss": 0.0106, "step": 21330 }, { "grad_norm": 0.15136633813381195, "learning_rate": 8.769763752363589e-05, "loss": 0.0152, "step": 21340 }, { "grad_norm": 0.16978910565376282, "learning_rate": 8.768405665710352e-05, "loss": 0.0143, "step": 21350 }, { "grad_norm": 0.17716267704963684, "learning_rate": 8.767046935139655e-05, "loss": 0.0122, "step": 21360 }, { "grad_norm": 0.17317689955234528, "learning_rate": 8.765687560883666e-05, "loss": 0.0114, "step": 21370 }, { "grad_norm": 0.1610879749059677, "learning_rate": 8.764327543174664e-05, "loss": 0.0106, "step": 21380 }, { "grad_norm": 0.20215864479541779, "learning_rate": 8.762966882245038e-05, "loss": 0.0142, "step": 21390 }, { "grad_norm": 0.18537244200706482, "learning_rate": 8.761605578327291e-05, "loss": 0.0122, "step": 21400 }, { "grad_norm": 0.18361896276474, "learning_rate": 8.76024363165403e-05, "loss": 0.0136, "step": 21410 }, { "grad_norm": 0.16111691296100616, "learning_rate": 8.758881042457976e-05, "loss": 0.013, "step": 21420 }, { "grad_norm": 0.15778738260269165, "learning_rate": 8.757517810971957e-05, "loss": 0.0123, "step": 21430 }, { "grad_norm": 0.14876633882522583, "learning_rate": 8.756153937428913e-05, "loss": 0.0136, "step": 21440 }, { "grad_norm": 0.17574000358581543, "learning_rate": 8.754789422061889e-05, "loss": 0.0168, "step": 21450 }, { "grad_norm": 0.17153431475162506, "learning_rate": 8.753424265104052e-05, "loss": 0.0118, "step": 21460 }, { "grad_norm": 0.13795264065265656, "learning_rate": 8.752058466788659e-05, "loss": 0.0107, "step": 21470 }, { "grad_norm": 0.2080654352903366, "learning_rate": 8.750692027349097e-05, "loss": 0.0123, "step": 21480 }, { "grad_norm": 0.1610112339258194, "learning_rate": 8.749324947018847e-05, "loss": 0.0133, "step": 21490 }, { "grad_norm": 0.1934899240732193, "learning_rate": 8.747957226031507e-05, "loss": 0.0168, "step": 21500 }, { "grad_norm": 0.17061884701251984, "learning_rate": 8.746588864620787e-05, "loss": 0.0125, "step": 21510 }, { "grad_norm": 0.27462297677993774, "learning_rate": 8.745219863020498e-05, "loss": 0.0143, "step": 21520 }, { "grad_norm": 0.18011689186096191, "learning_rate": 8.743850221464564e-05, "loss": 0.0143, "step": 21530 }, { "grad_norm": 0.16215217113494873, "learning_rate": 8.742479940187026e-05, "loss": 0.0151, "step": 21540 }, { "grad_norm": 0.20739471912384033, "learning_rate": 8.74110901942202e-05, "loss": 0.0136, "step": 21550 }, { "grad_norm": 0.2599533498287201, "learning_rate": 8.739737459403803e-05, "loss": 0.0153, "step": 21560 }, { "grad_norm": 0.2018802911043167, "learning_rate": 8.738365260366737e-05, "loss": 0.0126, "step": 21570 }, { "grad_norm": 0.15600335597991943, "learning_rate": 8.736992422545292e-05, "loss": 0.016, "step": 21580 }, { "grad_norm": 0.19355903565883636, "learning_rate": 8.73561894617405e-05, "loss": 0.0153, "step": 21590 }, { "grad_norm": 0.19014789164066315, "learning_rate": 8.734244831487697e-05, "loss": 0.0157, "step": 21600 }, { "grad_norm": 0.16131789982318878, "learning_rate": 8.732870078721035e-05, "loss": 0.013, "step": 21610 }, { "grad_norm": 0.18706610798835754, "learning_rate": 8.731494688108972e-05, "loss": 0.0134, "step": 21620 }, { "grad_norm": 0.23094616830348969, "learning_rate": 8.730118659886523e-05, "loss": 0.0148, "step": 21630 }, { "grad_norm": 0.18942806124687195, "learning_rate": 8.728741994288814e-05, "loss": 0.0139, "step": 21640 }, { "grad_norm": 0.1722007840871811, "learning_rate": 8.727364691551079e-05, "loss": 0.012, "step": 21650 }, { "grad_norm": 0.15968765318393707, "learning_rate": 8.725986751908661e-05, "loss": 0.0127, "step": 21660 }, { "grad_norm": 0.18266509473323822, "learning_rate": 8.724608175597016e-05, "loss": 0.0132, "step": 21670 }, { "grad_norm": 0.21753950417041779, "learning_rate": 8.723228962851699e-05, "loss": 0.0157, "step": 21680 }, { "grad_norm": 0.19835413992404938, "learning_rate": 8.721849113908385e-05, "loss": 0.0154, "step": 21690 }, { "grad_norm": 0.16700005531311035, "learning_rate": 8.720468629002848e-05, "loss": 0.0144, "step": 21700 }, { "grad_norm": 0.2124863713979721, "learning_rate": 8.719087508370978e-05, "loss": 0.0146, "step": 21710 }, { "grad_norm": 0.18665318191051483, "learning_rate": 8.717705752248772e-05, "loss": 0.0205, "step": 21720 }, { "grad_norm": 0.17159441113471985, "learning_rate": 8.71632336087233e-05, "loss": 0.0171, "step": 21730 }, { "grad_norm": 0.1750337928533554, "learning_rate": 8.71494033447787e-05, "loss": 0.0135, "step": 21740 }, { "grad_norm": 0.23318582773208618, "learning_rate": 8.713556673301708e-05, "loss": 0.0136, "step": 21750 }, { "grad_norm": 0.21332304179668427, "learning_rate": 8.712172377580278e-05, "loss": 0.0155, "step": 21760 }, { "grad_norm": 0.21032211184501648, "learning_rate": 8.710787447550114e-05, "loss": 0.0167, "step": 21770 }, { "grad_norm": 0.1909770965576172, "learning_rate": 8.70940188344787e-05, "loss": 0.0133, "step": 21780 }, { "grad_norm": 0.14542225003242493, "learning_rate": 8.708015685510293e-05, "loss": 0.0131, "step": 21790 }, { "grad_norm": 0.19782818853855133, "learning_rate": 8.706628853974252e-05, "loss": 0.0121, "step": 21800 }, { "grad_norm": 0.17101067304611206, "learning_rate": 8.705241389076715e-05, "loss": 0.0132, "step": 21810 }, { "grad_norm": 0.13859626650810242, "learning_rate": 8.703853291054764e-05, "loss": 0.0108, "step": 21820 }, { "grad_norm": 0.16377654671669006, "learning_rate": 8.702464560145587e-05, "loss": 0.0134, "step": 21830 }, { "grad_norm": 0.18363332748413086, "learning_rate": 8.701075196586476e-05, "loss": 0.0128, "step": 21840 }, { "grad_norm": 0.20279936492443085, "learning_rate": 8.699685200614842e-05, "loss": 0.0151, "step": 21850 }, { "grad_norm": 0.11766459047794342, "learning_rate": 8.698294572468193e-05, "loss": 0.0129, "step": 21860 }, { "grad_norm": 0.19161957502365112, "learning_rate": 8.696903312384148e-05, "loss": 0.0146, "step": 21870 }, { "grad_norm": 0.17111535370349884, "learning_rate": 8.695511420600439e-05, "loss": 0.0129, "step": 21880 }, { "grad_norm": 0.16353276371955872, "learning_rate": 8.694118897354901e-05, "loss": 0.0141, "step": 21890 }, { "grad_norm": 0.2590918242931366, "learning_rate": 8.692725742885478e-05, "loss": 0.015, "step": 21900 }, { "grad_norm": 0.2253590077161789, "learning_rate": 8.691331957430221e-05, "loss": 0.0139, "step": 21910 }, { "grad_norm": 0.1932917982339859, "learning_rate": 8.68993754122729e-05, "loss": 0.0143, "step": 21920 }, { "grad_norm": 0.20362532138824463, "learning_rate": 8.688542494514955e-05, "loss": 0.0154, "step": 21930 }, { "grad_norm": 0.14792314171791077, "learning_rate": 8.68714681753159e-05, "loss": 0.0145, "step": 21940 }, { "grad_norm": 0.1944016069173813, "learning_rate": 8.685750510515676e-05, "loss": 0.0122, "step": 21950 }, { "grad_norm": 0.14430133998394012, "learning_rate": 8.684353573705805e-05, "loss": 0.0127, "step": 21960 }, { "grad_norm": 0.15934903919696808, "learning_rate": 8.682956007340677e-05, "loss": 0.0148, "step": 21970 }, { "grad_norm": 0.16230474412441254, "learning_rate": 8.681557811659095e-05, "loss": 0.0159, "step": 21980 }, { "grad_norm": 0.12213533371686935, "learning_rate": 8.680158986899974e-05, "loss": 0.013, "step": 21990 }, { "grad_norm": 0.15783832967281342, "learning_rate": 8.678759533302335e-05, "loss": 0.0118, "step": 22000 }, { "grad_norm": 0.17973759770393372, "learning_rate": 8.677359451105308e-05, "loss": 0.0136, "step": 22010 }, { "grad_norm": 0.1569179743528366, "learning_rate": 8.675958740548123e-05, "loss": 0.0128, "step": 22020 }, { "grad_norm": 0.2147749960422516, "learning_rate": 8.674557401870129e-05, "loss": 0.0123, "step": 22030 }, { "grad_norm": 0.15379056334495544, "learning_rate": 8.673155435310775e-05, "loss": 0.014, "step": 22040 }, { "grad_norm": 0.1785944104194641, "learning_rate": 8.671752841109617e-05, "loss": 0.012, "step": 22050 }, { "grad_norm": 0.17414793372154236, "learning_rate": 8.670349619506321e-05, "loss": 0.0109, "step": 22060 }, { "grad_norm": 0.1664130687713623, "learning_rate": 8.66894577074066e-05, "loss": 0.0139, "step": 22070 }, { "grad_norm": 0.15235206484794617, "learning_rate": 8.667541295052513e-05, "loss": 0.0115, "step": 22080 }, { "grad_norm": 0.199070543050766, "learning_rate": 8.666136192681865e-05, "loss": 0.0124, "step": 22090 }, { "grad_norm": 0.16819855570793152, "learning_rate": 8.664730463868811e-05, "loss": 0.0155, "step": 22100 }, { "grad_norm": 0.19031168520450592, "learning_rate": 8.663324108853552e-05, "loss": 0.0132, "step": 22110 }, { "grad_norm": 0.16668696701526642, "learning_rate": 8.661917127876395e-05, "loss": 0.0137, "step": 22120 }, { "grad_norm": 0.16584795713424683, "learning_rate": 8.660509521177754e-05, "loss": 0.0169, "step": 22130 }, { "grad_norm": 0.10837186127901077, "learning_rate": 8.65910128899815e-05, "loss": 0.0132, "step": 22140 }, { "grad_norm": 0.16999438405036926, "learning_rate": 8.657692431578214e-05, "loss": 0.0126, "step": 22150 }, { "grad_norm": 0.18497870862483978, "learning_rate": 8.656282949158679e-05, "loss": 0.0111, "step": 22160 }, { "grad_norm": 0.15840935707092285, "learning_rate": 8.654872841980388e-05, "loss": 0.0131, "step": 22170 }, { "grad_norm": 0.14369483292102814, "learning_rate": 8.653462110284289e-05, "loss": 0.0132, "step": 22180 }, { "grad_norm": 0.20614343881607056, "learning_rate": 8.652050754311437e-05, "loss": 0.0126, "step": 22190 }, { "grad_norm": 0.15380893647670746, "learning_rate": 8.650638774302995e-05, "loss": 0.0128, "step": 22200 }, { "grad_norm": 0.22914929687976837, "learning_rate": 8.649226170500233e-05, "loss": 0.0133, "step": 22210 }, { "grad_norm": 0.16008688509464264, "learning_rate": 8.647812943144524e-05, "loss": 0.0155, "step": 22220 }, { "grad_norm": 0.2054598033428192, "learning_rate": 8.646399092477351e-05, "loss": 0.0127, "step": 22230 }, { "grad_norm": 0.16207846999168396, "learning_rate": 8.644984618740301e-05, "loss": 0.0154, "step": 22240 }, { "grad_norm": 0.21952560544013977, "learning_rate": 8.643569522175073e-05, "loss": 0.0168, "step": 22250 }, { "grad_norm": 0.18455879390239716, "learning_rate": 8.642153803023463e-05, "loss": 0.0137, "step": 22260 }, { "grad_norm": 0.13467362523078918, "learning_rate": 8.640737461527383e-05, "loss": 0.0127, "step": 22270 }, { "grad_norm": 0.17443318665027618, "learning_rate": 8.639320497928845e-05, "loss": 0.0113, "step": 22280 }, { "grad_norm": 0.15202166140079498, "learning_rate": 8.637902912469969e-05, "loss": 0.0126, "step": 22290 }, { "grad_norm": 0.12851634621620178, "learning_rate": 8.636484705392982e-05, "loss": 0.0102, "step": 22300 }, { "grad_norm": 0.18062232434749603, "learning_rate": 8.635065876940216e-05, "loss": 0.0134, "step": 22310 }, { "grad_norm": 0.14850673079490662, "learning_rate": 8.633646427354112e-05, "loss": 0.0126, "step": 22320 }, { "grad_norm": 0.174448624253273, "learning_rate": 8.632226356877213e-05, "loss": 0.0156, "step": 22330 }, { "grad_norm": 0.16347377002239227, "learning_rate": 8.630805665752173e-05, "loss": 0.0141, "step": 22340 }, { "grad_norm": 0.21396392583847046, "learning_rate": 8.629384354221748e-05, "loss": 0.0129, "step": 22350 }, { "grad_norm": 0.14222098886966705, "learning_rate": 8.627962422528797e-05, "loss": 0.0129, "step": 22360 }, { "grad_norm": 0.13207657635211945, "learning_rate": 8.626539870916296e-05, "loss": 0.0128, "step": 22370 }, { "grad_norm": 0.18369723856449127, "learning_rate": 8.625116699627317e-05, "loss": 0.0115, "step": 22380 }, { "grad_norm": 0.20117615163326263, "learning_rate": 8.623692908905041e-05, "loss": 0.0166, "step": 22390 }, { "grad_norm": 0.18396256864070892, "learning_rate": 8.622268498992755e-05, "loss": 0.0148, "step": 22400 }, { "grad_norm": 0.18550808727741241, "learning_rate": 8.620843470133851e-05, "loss": 0.0133, "step": 22410 }, { "grad_norm": 0.13316725194454193, "learning_rate": 8.619417822571829e-05, "loss": 0.0152, "step": 22420 }, { "grad_norm": 0.1614820957183838, "learning_rate": 8.617991556550292e-05, "loss": 0.0154, "step": 22430 }, { "grad_norm": 0.23010309040546417, "learning_rate": 8.616564672312952e-05, "loss": 0.0157, "step": 22440 }, { "grad_norm": 0.23984253406524658, "learning_rate": 8.61513717010362e-05, "loss": 0.0129, "step": 22450 }, { "grad_norm": 0.1916763037443161, "learning_rate": 8.613709050166221e-05, "loss": 0.0128, "step": 22460 }, { "grad_norm": 0.15464994311332703, "learning_rate": 8.61228031274478e-05, "loss": 0.0119, "step": 22470 }, { "grad_norm": 0.21420593559741974, "learning_rate": 8.610850958083431e-05, "loss": 0.0115, "step": 22480 }, { "grad_norm": 0.15531612932682037, "learning_rate": 8.609420986426409e-05, "loss": 0.0115, "step": 22490 }, { "grad_norm": 0.1516834795475006, "learning_rate": 8.60799039801806e-05, "loss": 0.0127, "step": 22500 }, { "grad_norm": 0.1572219431400299, "learning_rate": 8.606559193102828e-05, "loss": 0.0132, "step": 22510 }, { "grad_norm": 0.11535004526376724, "learning_rate": 8.605127371925273e-05, "loss": 0.0121, "step": 22520 }, { "grad_norm": 0.1449630707502365, "learning_rate": 8.603694934730047e-05, "loss": 0.0134, "step": 22530 }, { "grad_norm": 0.2016325443983078, "learning_rate": 8.602261881761919e-05, "loss": 0.0146, "step": 22540 }, { "grad_norm": 0.15158230066299438, "learning_rate": 8.600828213265759e-05, "loss": 0.0119, "step": 22550 }, { "grad_norm": 0.1962612122297287, "learning_rate": 8.599393929486539e-05, "loss": 0.0137, "step": 22560 }, { "grad_norm": 0.20098350942134857, "learning_rate": 8.59795903066934e-05, "loss": 0.0123, "step": 22570 }, { "grad_norm": 0.14842820167541504, "learning_rate": 8.596523517059347e-05, "loss": 0.0102, "step": 22580 }, { "grad_norm": 0.15202875435352325, "learning_rate": 8.59508738890185e-05, "loss": 0.0122, "step": 22590 }, { "grad_norm": 0.16384001076221466, "learning_rate": 8.593650646442246e-05, "loss": 0.0131, "step": 22600 }, { "grad_norm": 0.20021259784698486, "learning_rate": 8.59221328992603e-05, "loss": 0.0122, "step": 22610 }, { "grad_norm": 0.17772312462329865, "learning_rate": 8.590775319598813e-05, "loss": 0.0107, "step": 22620 }, { "grad_norm": 0.14218638837337494, "learning_rate": 8.589336735706301e-05, "loss": 0.0128, "step": 22630 }, { "grad_norm": 0.20085354149341583, "learning_rate": 8.587897538494307e-05, "loss": 0.013, "step": 22640 }, { "grad_norm": 0.22800545394420624, "learning_rate": 8.586457728208756e-05, "loss": 0.0141, "step": 22650 }, { "grad_norm": 0.1399993896484375, "learning_rate": 8.585017305095667e-05, "loss": 0.0128, "step": 22660 }, { "grad_norm": 0.23741652071475983, "learning_rate": 8.583576269401173e-05, "loss": 0.0193, "step": 22670 }, { "grad_norm": 0.20040397346019745, "learning_rate": 8.582134621371504e-05, "loss": 0.0144, "step": 22680 }, { "grad_norm": 0.21653085947036743, "learning_rate": 8.580692361253e-05, "loss": 0.0133, "step": 22690 }, { "grad_norm": 0.24451686441898346, "learning_rate": 8.579249489292104e-05, "loss": 0.013, "step": 22700 }, { "grad_norm": 0.18504169583320618, "learning_rate": 8.577806005735363e-05, "loss": 0.0143, "step": 22710 }, { "grad_norm": 0.2351469248533249, "learning_rate": 8.576361910829429e-05, "loss": 0.0125, "step": 22720 }, { "grad_norm": 0.145010307431221, "learning_rate": 8.574917204821057e-05, "loss": 0.012, "step": 22730 }, { "grad_norm": 0.1685410588979721, "learning_rate": 8.57347188795711e-05, "loss": 0.0135, "step": 22740 }, { "grad_norm": 0.14433155953884125, "learning_rate": 8.572025960484551e-05, "loss": 0.0123, "step": 22750 }, { "grad_norm": 0.1496257334947586, "learning_rate": 8.57057942265045e-05, "loss": 0.0121, "step": 22760 }, { "grad_norm": 0.16906818747520447, "learning_rate": 8.569132274701984e-05, "loss": 0.0148, "step": 22770 }, { "grad_norm": 0.18685273826122284, "learning_rate": 8.567684516886427e-05, "loss": 0.0133, "step": 22780 }, { "grad_norm": 0.17691077291965485, "learning_rate": 8.56623614945116e-05, "loss": 0.0123, "step": 22790 }, { "grad_norm": 0.22294864058494568, "learning_rate": 8.564787172643675e-05, "loss": 0.0129, "step": 22800 }, { "grad_norm": 0.23564667999744415, "learning_rate": 8.563337586711559e-05, "loss": 0.0174, "step": 22810 }, { "grad_norm": 0.19081977009773254, "learning_rate": 8.561887391902506e-05, "loss": 0.0101, "step": 22820 }, { "grad_norm": 0.16347236931324005, "learning_rate": 8.560436588464316e-05, "loss": 0.0137, "step": 22830 }, { "grad_norm": 0.2819353938102722, "learning_rate": 8.55898517664489e-05, "loss": 0.0118, "step": 22840 }, { "grad_norm": 0.23593516647815704, "learning_rate": 8.557533156692236e-05, "loss": 0.0161, "step": 22850 }, { "grad_norm": 0.18704655766487122, "learning_rate": 8.556080528854467e-05, "loss": 0.0129, "step": 22860 }, { "grad_norm": 0.18208426237106323, "learning_rate": 8.554627293379791e-05, "loss": 0.0127, "step": 22870 }, { "grad_norm": 0.1552094668149948, "learning_rate": 8.553173450516531e-05, "loss": 0.0126, "step": 22880 }, { "grad_norm": 0.16670365631580353, "learning_rate": 8.551719000513108e-05, "loss": 0.0115, "step": 22890 }, { "grad_norm": 0.20955224335193634, "learning_rate": 8.550263943618049e-05, "loss": 0.0141, "step": 22900 }, { "grad_norm": 0.14464426040649414, "learning_rate": 8.54880828007998e-05, "loss": 0.0131, "step": 22910 }, { "grad_norm": 0.15392836928367615, "learning_rate": 8.547352010147637e-05, "loss": 0.0138, "step": 22920 }, { "grad_norm": 0.2394436150789261, "learning_rate": 8.545895134069855e-05, "loss": 0.0126, "step": 22930 }, { "grad_norm": 0.1913074105978012, "learning_rate": 8.544437652095576e-05, "loss": 0.0139, "step": 22940 }, { "grad_norm": 0.18053589761257172, "learning_rate": 8.542979564473843e-05, "loss": 0.014, "step": 22950 }, { "grad_norm": 0.23391567170619965, "learning_rate": 8.541520871453802e-05, "loss": 0.0149, "step": 22960 }, { "grad_norm": 0.19781403243541718, "learning_rate": 8.540061573284705e-05, "loss": 0.0152, "step": 22970 }, { "grad_norm": 0.14780987799167633, "learning_rate": 8.538601670215906e-05, "loss": 0.0116, "step": 22980 }, { "grad_norm": 0.13591064512729645, "learning_rate": 8.537141162496864e-05, "loss": 0.0113, "step": 22990 }, { "grad_norm": 0.12533800303936005, "learning_rate": 8.535680050377137e-05, "loss": 0.0137, "step": 23000 }, { "grad_norm": 0.14003333449363708, "learning_rate": 8.534218334106391e-05, "loss": 0.012, "step": 23010 }, { "grad_norm": 0.1917215883731842, "learning_rate": 8.532756013934393e-05, "loss": 0.0116, "step": 23020 }, { "grad_norm": 0.2022702991962433, "learning_rate": 8.531293090111012e-05, "loss": 0.0105, "step": 23030 }, { "grad_norm": 0.17856283485889435, "learning_rate": 8.529829562886225e-05, "loss": 0.0136, "step": 23040 }, { "grad_norm": 0.20259989798069, "learning_rate": 8.528365432510105e-05, "loss": 0.0121, "step": 23050 }, { "grad_norm": 0.19927220046520233, "learning_rate": 8.526900699232833e-05, "loss": 0.0134, "step": 23060 }, { "grad_norm": 0.18769092857837677, "learning_rate": 8.525435363304695e-05, "loss": 0.0132, "step": 23070 }, { "grad_norm": 0.22002840042114258, "learning_rate": 8.523969424976072e-05, "loss": 0.0143, "step": 23080 }, { "grad_norm": 0.1985943466424942, "learning_rate": 8.522502884497457e-05, "loss": 0.0137, "step": 23090 }, { "grad_norm": 0.21383976936340332, "learning_rate": 8.521035742119437e-05, "loss": 0.0146, "step": 23100 }, { "grad_norm": 0.14770837128162384, "learning_rate": 8.519567998092712e-05, "loss": 0.0124, "step": 23110 }, { "grad_norm": 0.17656166851520538, "learning_rate": 8.518099652668075e-05, "loss": 0.0111, "step": 23120 }, { "grad_norm": 0.1433616280555725, "learning_rate": 8.516630706096429e-05, "loss": 0.0118, "step": 23130 }, { "grad_norm": 0.14664946496486664, "learning_rate": 8.515161158628773e-05, "loss": 0.0105, "step": 23140 }, { "grad_norm": 0.16491222381591797, "learning_rate": 8.513691010516216e-05, "loss": 0.0123, "step": 23150 }, { "grad_norm": 0.13798275589942932, "learning_rate": 8.512220262009966e-05, "loss": 0.0114, "step": 23160 }, { "grad_norm": 0.17246633768081665, "learning_rate": 8.510748913361332e-05, "loss": 0.0112, "step": 23170 }, { "grad_norm": 0.17950594425201416, "learning_rate": 8.509276964821726e-05, "loss": 0.0118, "step": 23180 }, { "grad_norm": 0.18774567544460297, "learning_rate": 8.507804416642669e-05, "loss": 0.0147, "step": 23190 }, { "grad_norm": 0.1577351987361908, "learning_rate": 8.506331269075774e-05, "loss": 0.017, "step": 23200 }, { "grad_norm": 0.17213204503059387, "learning_rate": 8.504857522372765e-05, "loss": 0.0108, "step": 23210 }, { "grad_norm": 0.13126708567142487, "learning_rate": 8.503383176785461e-05, "loss": 0.0098, "step": 23220 }, { "grad_norm": 0.1818597912788391, "learning_rate": 8.501908232565792e-05, "loss": 0.0123, "step": 23230 }, { "grad_norm": 0.16360940039157867, "learning_rate": 8.50043268996578e-05, "loss": 0.0115, "step": 23240 }, { "grad_norm": 0.17393872141838074, "learning_rate": 8.498956549237562e-05, "loss": 0.0096, "step": 23250 }, { "grad_norm": 0.1303960084915161, "learning_rate": 8.497479810633366e-05, "loss": 0.0123, "step": 23260 }, { "grad_norm": 0.15944769978523254, "learning_rate": 8.496002474405525e-05, "loss": 0.0131, "step": 23270 }, { "grad_norm": 0.18428783118724823, "learning_rate": 8.494524540806478e-05, "loss": 0.0148, "step": 23280 }, { "grad_norm": 0.2064494490623474, "learning_rate": 8.493046010088761e-05, "loss": 0.0127, "step": 23290 }, { "grad_norm": 0.16298674046993256, "learning_rate": 8.491566882505018e-05, "loss": 0.0127, "step": 23300 }, { "grad_norm": 0.17456766963005066, "learning_rate": 8.490087158307988e-05, "loss": 0.0122, "step": 23310 }, { "grad_norm": 0.1595582664012909, "learning_rate": 8.488606837750518e-05, "loss": 0.0138, "step": 23320 }, { "grad_norm": 0.20436660945415497, "learning_rate": 8.487125921085552e-05, "loss": 0.0119, "step": 23330 }, { "grad_norm": 0.14774271845817566, "learning_rate": 8.485644408566141e-05, "loss": 0.0109, "step": 23340 }, { "grad_norm": 0.20743423700332642, "learning_rate": 8.484162300445431e-05, "loss": 0.0126, "step": 23350 }, { "grad_norm": 0.21242903172969818, "learning_rate": 8.482679596976676e-05, "loss": 0.014, "step": 23360 }, { "grad_norm": 0.1263130009174347, "learning_rate": 8.48119629841323e-05, "loss": 0.0141, "step": 23370 }, { "grad_norm": 0.17559507489204407, "learning_rate": 8.479712405008547e-05, "loss": 0.0158, "step": 23380 }, { "grad_norm": 0.2077982872724533, "learning_rate": 8.478227917016184e-05, "loss": 0.0128, "step": 23390 }, { "grad_norm": 0.20183001458644867, "learning_rate": 8.476742834689801e-05, "loss": 0.0136, "step": 23400 }, { "grad_norm": 0.13166436553001404, "learning_rate": 8.475257158283157e-05, "loss": 0.0132, "step": 23410 }, { "grad_norm": 0.1763761341571808, "learning_rate": 8.473770888050112e-05, "loss": 0.0156, "step": 23420 }, { "grad_norm": 0.1914791762828827, "learning_rate": 8.47228402424463e-05, "loss": 0.0159, "step": 23430 }, { "grad_norm": 0.1980430632829666, "learning_rate": 8.470796567120775e-05, "loss": 0.0123, "step": 23440 }, { "grad_norm": 0.1529274731874466, "learning_rate": 8.469308516932714e-05, "loss": 0.0148, "step": 23450 }, { "grad_norm": 0.16436666250228882, "learning_rate": 8.467819873934714e-05, "loss": 0.0139, "step": 23460 }, { "grad_norm": 0.16010263562202454, "learning_rate": 8.466330638381143e-05, "loss": 0.0142, "step": 23470 }, { "grad_norm": 0.18539147078990936, "learning_rate": 8.464840810526469e-05, "loss": 0.0122, "step": 23480 }, { "grad_norm": 0.17249543964862823, "learning_rate": 8.463350390625264e-05, "loss": 0.0097, "step": 23490 }, { "grad_norm": 0.16833414137363434, "learning_rate": 8.4618593789322e-05, "loss": 0.0134, "step": 23500 }, { "grad_norm": 0.19284215569496155, "learning_rate": 8.46036777570205e-05, "loss": 0.0141, "step": 23510 }, { "grad_norm": 0.2107212096452713, "learning_rate": 8.458875581189688e-05, "loss": 0.0146, "step": 23520 }, { "grad_norm": 0.16327619552612305, "learning_rate": 8.457382795650092e-05, "loss": 0.0142, "step": 23530 }, { "grad_norm": 0.15534082055091858, "learning_rate": 8.455889419338335e-05, "loss": 0.015, "step": 23540 }, { "grad_norm": 0.1987888664007187, "learning_rate": 8.454395452509593e-05, "loss": 0.0125, "step": 23550 }, { "grad_norm": 0.18365146219730377, "learning_rate": 8.452900895419146e-05, "loss": 0.0105, "step": 23560 }, { "grad_norm": 0.14206844568252563, "learning_rate": 8.451405748322376e-05, "loss": 0.0124, "step": 23570 }, { "grad_norm": 0.1839594542980194, "learning_rate": 8.449910011474759e-05, "loss": 0.0175, "step": 23580 }, { "grad_norm": 0.17566266655921936, "learning_rate": 8.448413685131876e-05, "loss": 0.0135, "step": 23590 }, { "grad_norm": 0.21478013694286346, "learning_rate": 8.446916769549407e-05, "loss": 0.0129, "step": 23600 }, { "grad_norm": 0.13358882069587708, "learning_rate": 8.445419264983136e-05, "loss": 0.0119, "step": 23610 }, { "grad_norm": 0.21689066290855408, "learning_rate": 8.443921171688947e-05, "loss": 0.0106, "step": 23620 }, { "grad_norm": 0.20499223470687866, "learning_rate": 8.442422489922819e-05, "loss": 0.013, "step": 23630 }, { "grad_norm": 0.20308536291122437, "learning_rate": 8.440923219940838e-05, "loss": 0.0131, "step": 23640 }, { "grad_norm": 0.17517851293087006, "learning_rate": 8.439423361999189e-05, "loss": 0.0112, "step": 23650 }, { "grad_norm": 0.19433894753456116, "learning_rate": 8.437922916354155e-05, "loss": 0.0112, "step": 23660 }, { "grad_norm": 0.187160462141037, "learning_rate": 8.436421883262123e-05, "loss": 0.0123, "step": 23670 }, { "grad_norm": 0.12598168849945068, "learning_rate": 8.434920262979577e-05, "loss": 0.0099, "step": 23680 }, { "grad_norm": 0.2022571861743927, "learning_rate": 8.433418055763104e-05, "loss": 0.0106, "step": 23690 }, { "grad_norm": 0.12840133905410767, "learning_rate": 8.431915261869389e-05, "loss": 0.0127, "step": 23700 }, { "grad_norm": 0.15141424536705017, "learning_rate": 8.43041188155522e-05, "loss": 0.0102, "step": 23710 }, { "grad_norm": 0.17377449572086334, "learning_rate": 8.428907915077481e-05, "loss": 0.0126, "step": 23720 }, { "grad_norm": 0.1479218453168869, "learning_rate": 8.42740336269316e-05, "loss": 0.0109, "step": 23730 }, { "grad_norm": 0.1814265251159668, "learning_rate": 8.425898224659345e-05, "loss": 0.0137, "step": 23740 }, { "grad_norm": 0.16510598361492157, "learning_rate": 8.42439250123322e-05, "loss": 0.013, "step": 23750 }, { "grad_norm": 0.1828743815422058, "learning_rate": 8.422886192672076e-05, "loss": 0.0134, "step": 23760 }, { "grad_norm": 0.20277613401412964, "learning_rate": 8.421379299233297e-05, "loss": 0.0129, "step": 23770 }, { "grad_norm": 0.1986384093761444, "learning_rate": 8.419871821174371e-05, "loss": 0.0135, "step": 23780 }, { "grad_norm": 0.1797793209552765, "learning_rate": 8.418363758752884e-05, "loss": 0.0129, "step": 23790 }, { "grad_norm": 0.16133950650691986, "learning_rate": 8.416855112226523e-05, "loss": 0.0135, "step": 23800 }, { "grad_norm": 0.16970977187156677, "learning_rate": 8.415345881853075e-05, "loss": 0.0123, "step": 23810 }, { "grad_norm": 0.19882509112358093, "learning_rate": 8.413836067890426e-05, "loss": 0.0139, "step": 23820 }, { "grad_norm": 0.1617104560136795, "learning_rate": 8.41232567059656e-05, "loss": 0.0104, "step": 23830 }, { "grad_norm": 0.14430148899555206, "learning_rate": 8.410814690229565e-05, "loss": 0.0125, "step": 23840 }, { "grad_norm": 0.12528741359710693, "learning_rate": 8.409303127047626e-05, "loss": 0.0118, "step": 23850 }, { "grad_norm": 0.21685923635959625, "learning_rate": 8.407790981309028e-05, "loss": 0.0113, "step": 23860 }, { "grad_norm": 0.1302449107170105, "learning_rate": 8.406278253272153e-05, "loss": 0.0109, "step": 23870 }, { "grad_norm": 0.16127188503742218, "learning_rate": 8.404764943195487e-05, "loss": 0.0112, "step": 23880 }, { "grad_norm": 0.16836552321910858, "learning_rate": 8.403251051337613e-05, "loss": 0.013, "step": 23890 }, { "grad_norm": 0.11499406397342682, "learning_rate": 8.401736577957214e-05, "loss": 0.0117, "step": 23900 }, { "grad_norm": 0.14725933969020844, "learning_rate": 8.40022152331307e-05, "loss": 0.0105, "step": 23910 }, { "grad_norm": 0.14342640340328217, "learning_rate": 8.398705887664064e-05, "loss": 0.011, "step": 23920 }, { "grad_norm": 0.2604374289512634, "learning_rate": 8.397189671269177e-05, "loss": 0.0107, "step": 23930 }, { "grad_norm": 0.16997209191322327, "learning_rate": 8.395672874387488e-05, "loss": 0.0124, "step": 23940 }, { "grad_norm": 0.20453055202960968, "learning_rate": 8.394155497278177e-05, "loss": 0.0122, "step": 23950 }, { "grad_norm": 0.15866397321224213, "learning_rate": 8.392637540200523e-05, "loss": 0.0139, "step": 23960 }, { "grad_norm": 0.1447560340166092, "learning_rate": 8.391119003413902e-05, "loss": 0.0101, "step": 23970 }, { "grad_norm": 0.1942092627286911, "learning_rate": 8.38959988717779e-05, "loss": 0.0102, "step": 23980 }, { "grad_norm": 0.1671309620141983, "learning_rate": 8.388080191751764e-05, "loss": 0.0121, "step": 23990 }, { "grad_norm": 0.148244246840477, "learning_rate": 8.386559917395496e-05, "loss": 0.0121, "step": 24000 }, { "grad_norm": 0.11933386325836182, "learning_rate": 8.385039064368761e-05, "loss": 0.0105, "step": 24010 }, { "grad_norm": 0.15464258193969727, "learning_rate": 8.383517632931431e-05, "loss": 0.0144, "step": 24020 }, { "grad_norm": 0.15319743752479553, "learning_rate": 8.381995623343477e-05, "loss": 0.0116, "step": 24030 }, { "grad_norm": 0.23701004683971405, "learning_rate": 8.380473035864968e-05, "loss": 0.0141, "step": 24040 }, { "grad_norm": 0.22908514738082886, "learning_rate": 8.378949870756076e-05, "loss": 0.0137, "step": 24050 }, { "grad_norm": 0.13687863945960999, "learning_rate": 8.377426128277063e-05, "loss": 0.0153, "step": 24060 }, { "grad_norm": 0.18062271177768707, "learning_rate": 8.375901808688298e-05, "loss": 0.0118, "step": 24070 }, { "grad_norm": 0.20212779939174652, "learning_rate": 8.374376912250246e-05, "loss": 0.0108, "step": 24080 }, { "grad_norm": 0.2158302366733551, "learning_rate": 8.372851439223468e-05, "loss": 0.0142, "step": 24090 }, { "grad_norm": 0.1729850471019745, "learning_rate": 8.371325389868627e-05, "loss": 0.0123, "step": 24100 }, { "grad_norm": 0.15804632008075714, "learning_rate": 8.369798764446482e-05, "loss": 0.0123, "step": 24110 }, { "grad_norm": 0.17898736894130707, "learning_rate": 8.368271563217893e-05, "loss": 0.0134, "step": 24120 }, { "grad_norm": 0.18372376263141632, "learning_rate": 8.366743786443817e-05, "loss": 0.0116, "step": 24130 }, { "grad_norm": 0.1343306303024292, "learning_rate": 8.365215434385309e-05, "loss": 0.0118, "step": 24140 }, { "grad_norm": 0.12328104674816132, "learning_rate": 8.36368650730352e-05, "loss": 0.0128, "step": 24150 }, { "grad_norm": 0.12953941524028778, "learning_rate": 8.362157005459705e-05, "loss": 0.0133, "step": 24160 }, { "grad_norm": 0.12952014803886414, "learning_rate": 8.360626929115213e-05, "loss": 0.0095, "step": 24170 }, { "grad_norm": 0.15160724520683289, "learning_rate": 8.359096278531492e-05, "loss": 0.0155, "step": 24180 }, { "grad_norm": 0.17030467092990875, "learning_rate": 8.357565053970088e-05, "loss": 0.0149, "step": 24190 }, { "grad_norm": 0.15617656707763672, "learning_rate": 8.356033255692647e-05, "loss": 0.0109, "step": 24200 }, { "grad_norm": 0.13231788575649261, "learning_rate": 8.354500883960911e-05, "loss": 0.0123, "step": 24210 }, { "grad_norm": 0.13698430359363556, "learning_rate": 8.352967939036717e-05, "loss": 0.0157, "step": 24220 }, { "grad_norm": 0.14784923195838928, "learning_rate": 8.35143442118201e-05, "loss": 0.0135, "step": 24230 }, { "grad_norm": 0.17312902212142944, "learning_rate": 8.349900330658819e-05, "loss": 0.0129, "step": 24240 }, { "grad_norm": 0.17051798105239868, "learning_rate": 8.348365667729284e-05, "loss": 0.0144, "step": 24250 }, { "grad_norm": 0.1936093419790268, "learning_rate": 8.346830432655633e-05, "loss": 0.0122, "step": 24260 }, { "grad_norm": 0.2946666181087494, "learning_rate": 8.345294625700195e-05, "loss": 0.0118, "step": 24270 }, { "grad_norm": 0.19242024421691895, "learning_rate": 8.343758247125402e-05, "loss": 0.0116, "step": 24280 }, { "grad_norm": 0.15991932153701782, "learning_rate": 8.342221297193776e-05, "loss": 0.015, "step": 24290 }, { "grad_norm": 0.14196684956550598, "learning_rate": 8.34068377616794e-05, "loss": 0.0124, "step": 24300 }, { "grad_norm": 0.18877555429935455, "learning_rate": 8.339145684310615e-05, "loss": 0.0136, "step": 24310 }, { "grad_norm": 0.20063038170337677, "learning_rate": 8.337607021884618e-05, "loss": 0.0135, "step": 24320 }, { "grad_norm": 0.16129347681999207, "learning_rate": 8.336067789152867e-05, "loss": 0.0128, "step": 24330 }, { "grad_norm": 0.19551518559455872, "learning_rate": 8.334527986378369e-05, "loss": 0.013, "step": 24340 }, { "grad_norm": 0.2262812703847885, "learning_rate": 8.332987613824239e-05, "loss": 0.0119, "step": 24350 }, { "grad_norm": 0.1500687599182129, "learning_rate": 8.331446671753685e-05, "loss": 0.0131, "step": 24360 }, { "grad_norm": 0.1433422565460205, "learning_rate": 8.329905160430007e-05, "loss": 0.0136, "step": 24370 }, { "grad_norm": 0.17094939947128296, "learning_rate": 8.328363080116611e-05, "loss": 0.011, "step": 24380 }, { "grad_norm": 0.14510920643806458, "learning_rate": 8.326820431076997e-05, "loss": 0.0128, "step": 24390 }, { "grad_norm": 0.14638616144657135, "learning_rate": 8.325277213574759e-05, "loss": 0.0107, "step": 24400 }, { "grad_norm": 0.148738294839859, "learning_rate": 8.32373342787359e-05, "loss": 0.0124, "step": 24410 }, { "grad_norm": 0.12009231001138687, "learning_rate": 8.322189074237285e-05, "loss": 0.0123, "step": 24420 }, { "grad_norm": 0.16682514548301697, "learning_rate": 8.32064415292973e-05, "loss": 0.0136, "step": 24430 }, { "grad_norm": 0.1853889375925064, "learning_rate": 8.319098664214907e-05, "loss": 0.0158, "step": 24440 }, { "grad_norm": 0.13446636497974396, "learning_rate": 8.3175526083569e-05, "loss": 0.0121, "step": 24450 }, { "grad_norm": 0.1439325511455536, "learning_rate": 8.316005985619889e-05, "loss": 0.0109, "step": 24460 }, { "grad_norm": 0.22113096714019775, "learning_rate": 8.314458796268147e-05, "loss": 0.0113, "step": 24470 }, { "grad_norm": 0.16084280610084534, "learning_rate": 8.312911040566047e-05, "loss": 0.0126, "step": 24480 }, { "grad_norm": 0.20063269138336182, "learning_rate": 8.31136271877806e-05, "loss": 0.0139, "step": 24490 }, { "grad_norm": 0.18246020376682281, "learning_rate": 8.309813831168748e-05, "loss": 0.0149, "step": 24500 }, { "grad_norm": 0.20669227838516235, "learning_rate": 8.308264378002777e-05, "loss": 0.0135, "step": 24510 }, { "grad_norm": 0.19855885207653046, "learning_rate": 8.306714359544906e-05, "loss": 0.0127, "step": 24520 }, { "grad_norm": 0.16027437150478363, "learning_rate": 8.30516377605999e-05, "loss": 0.0125, "step": 24530 }, { "grad_norm": 0.15327446162700653, "learning_rate": 8.30361262781298e-05, "loss": 0.0159, "step": 24540 }, { "grad_norm": 0.14397992193698883, "learning_rate": 8.302060915068924e-05, "loss": 0.0117, "step": 24550 }, { "grad_norm": 0.17003291845321655, "learning_rate": 8.300508638092972e-05, "loss": 0.0122, "step": 24560 }, { "grad_norm": 0.1402270793914795, "learning_rate": 8.298955797150361e-05, "loss": 0.0155, "step": 24570 }, { "grad_norm": 0.16587433218955994, "learning_rate": 8.297402392506433e-05, "loss": 0.0108, "step": 24580 }, { "grad_norm": 0.166418194770813, "learning_rate": 8.295848424426617e-05, "loss": 0.0119, "step": 24590 }, { "grad_norm": 0.16677768528461456, "learning_rate": 8.29429389317645e-05, "loss": 0.016, "step": 24600 }, { "grad_norm": 0.15351265668869019, "learning_rate": 8.292738799021556e-05, "loss": 0.0125, "step": 24610 }, { "grad_norm": 0.15703532099723816, "learning_rate": 8.291183142227656e-05, "loss": 0.014, "step": 24620 }, { "grad_norm": 0.21245183050632477, "learning_rate": 8.289626923060572e-05, "loss": 0.0114, "step": 24630 }, { "grad_norm": 0.1660783588886261, "learning_rate": 8.288070141786218e-05, "loss": 0.0105, "step": 24640 }, { "grad_norm": 0.13795675337314606, "learning_rate": 8.286512798670605e-05, "loss": 0.0113, "step": 24650 }, { "grad_norm": 0.16444694995880127, "learning_rate": 8.284954893979842e-05, "loss": 0.0125, "step": 24660 }, { "grad_norm": 0.16089820861816406, "learning_rate": 8.283396427980131e-05, "loss": 0.0122, "step": 24670 }, { "grad_norm": 0.14760486781597137, "learning_rate": 8.281837400937771e-05, "loss": 0.0126, "step": 24680 }, { "grad_norm": 0.14522840082645416, "learning_rate": 8.28027781311916e-05, "loss": 0.0128, "step": 24690 }, { "grad_norm": 0.15640680491924286, "learning_rate": 8.278717664790785e-05, "loss": 0.0108, "step": 24700 }, { "grad_norm": 0.210169717669487, "learning_rate": 8.277156956219234e-05, "loss": 0.012, "step": 24710 }, { "grad_norm": 0.17579245567321777, "learning_rate": 8.275595687671189e-05, "loss": 0.0146, "step": 24720 }, { "grad_norm": 0.14635735750198364, "learning_rate": 8.27403385941343e-05, "loss": 0.0123, "step": 24730 }, { "grad_norm": 0.17356741428375244, "learning_rate": 8.272471471712828e-05, "loss": 0.0119, "step": 24740 }, { "grad_norm": 0.1501069962978363, "learning_rate": 8.270908524836355e-05, "loss": 0.0114, "step": 24750 }, { "grad_norm": 0.21042156219482422, "learning_rate": 8.269345019051074e-05, "loss": 0.0153, "step": 24760 }, { "grad_norm": 0.1519034504890442, "learning_rate": 8.267780954624147e-05, "loss": 0.0116, "step": 24770 }, { "grad_norm": 0.17142824828624725, "learning_rate": 8.266216331822827e-05, "loss": 0.0121, "step": 24780 }, { "grad_norm": 0.1576225757598877, "learning_rate": 8.264651150914469e-05, "loss": 0.0132, "step": 24790 }, { "grad_norm": 0.18365788459777832, "learning_rate": 8.263085412166517e-05, "loss": 0.0112, "step": 24800 }, { "grad_norm": 0.2222716510295868, "learning_rate": 8.261519115846514e-05, "loss": 0.0131, "step": 24810 }, { "grad_norm": 0.17891418933868408, "learning_rate": 8.259952262222096e-05, "loss": 0.011, "step": 24820 }, { "grad_norm": 0.21015939116477966, "learning_rate": 8.258384851560997e-05, "loss": 0.0124, "step": 24830 }, { "grad_norm": 0.16603825986385345, "learning_rate": 8.256816884131044e-05, "loss": 0.0104, "step": 24840 }, { "grad_norm": 0.15462583303451538, "learning_rate": 8.255248360200159e-05, "loss": 0.0117, "step": 24850 }, { "grad_norm": 0.16558459401130676, "learning_rate": 8.253679280036359e-05, "loss": 0.0092, "step": 24860 }, { "grad_norm": 0.17369936406612396, "learning_rate": 8.252109643907762e-05, "loss": 0.01, "step": 24870 }, { "grad_norm": 0.14011317491531372, "learning_rate": 8.250539452082569e-05, "loss": 0.0109, "step": 24880 }, { "grad_norm": 0.11702771484851837, "learning_rate": 8.248968704829087e-05, "loss": 0.0098, "step": 24890 }, { "grad_norm": 0.15445390343666077, "learning_rate": 8.247397402415714e-05, "loss": 0.0104, "step": 24900 }, { "grad_norm": 0.11762432754039764, "learning_rate": 8.24582554511094e-05, "loss": 0.0115, "step": 24910 }, { "grad_norm": 0.13884237408638, "learning_rate": 8.244253133183355e-05, "loss": 0.013, "step": 24920 }, { "grad_norm": 0.19828028976917267, "learning_rate": 8.24268016690164e-05, "loss": 0.0119, "step": 24930 }, { "grad_norm": 0.18110764026641846, "learning_rate": 8.241106646534571e-05, "loss": 0.0112, "step": 24940 }, { "grad_norm": 0.13866257667541504, "learning_rate": 8.23953257235102e-05, "loss": 0.0116, "step": 24950 }, { "grad_norm": 0.17613668739795685, "learning_rate": 8.237957944619956e-05, "loss": 0.0122, "step": 24960 }, { "grad_norm": 0.2027829885482788, "learning_rate": 8.236382763610437e-05, "loss": 0.0115, "step": 24970 }, { "grad_norm": 0.18336468935012817, "learning_rate": 8.234807029591619e-05, "loss": 0.0141, "step": 24980 }, { "grad_norm": 0.15685084462165833, "learning_rate": 8.233230742832752e-05, "loss": 0.0127, "step": 24990 }, { "grad_norm": 0.1911257952451706, "learning_rate": 8.231653903603178e-05, "loss": 0.0123, "step": 25000 }, { "grad_norm": 0.15395990014076233, "learning_rate": 8.23007651217234e-05, "loss": 0.0116, "step": 25010 }, { "grad_norm": 0.15200772881507874, "learning_rate": 8.228498568809769e-05, "loss": 0.0113, "step": 25020 }, { "grad_norm": 0.14657136797904968, "learning_rate": 8.22692007378509e-05, "loss": 0.0124, "step": 25030 }, { "grad_norm": 0.1292056292295456, "learning_rate": 8.225341027368028e-05, "loss": 0.0116, "step": 25040 }, { "grad_norm": 0.14096994698047638, "learning_rate": 8.223761429828399e-05, "loss": 0.0121, "step": 25050 }, { "grad_norm": 0.16097058355808258, "learning_rate": 8.22218128143611e-05, "loss": 0.0119, "step": 25060 }, { "grad_norm": 0.20488053560256958, "learning_rate": 8.220600582461166e-05, "loss": 0.0154, "step": 25070 }, { "grad_norm": 0.1787697672843933, "learning_rate": 8.219019333173668e-05, "loss": 0.0112, "step": 25080 }, { "grad_norm": 0.18978777527809143, "learning_rate": 8.217437533843805e-05, "loss": 0.0122, "step": 25090 }, { "grad_norm": 0.1484893560409546, "learning_rate": 8.215855184741867e-05, "loss": 0.0128, "step": 25100 }, { "grad_norm": 0.1213177815079689, "learning_rate": 8.21427228613823e-05, "loss": 0.0115, "step": 25110 }, { "grad_norm": 0.13940489292144775, "learning_rate": 8.21268883830337e-05, "loss": 0.0132, "step": 25120 }, { "grad_norm": 0.12276280671358109, "learning_rate": 8.211104841507855e-05, "loss": 0.0118, "step": 25130 }, { "grad_norm": 0.13477061688899994, "learning_rate": 8.209520296022346e-05, "loss": 0.0128, "step": 25140 }, { "grad_norm": 0.21043981611728668, "learning_rate": 8.207935202117599e-05, "loss": 0.0121, "step": 25150 }, { "grad_norm": 0.1605938822031021, "learning_rate": 8.206349560064463e-05, "loss": 0.0121, "step": 25160 }, { "grad_norm": 0.15878278017044067, "learning_rate": 8.204763370133881e-05, "loss": 0.0115, "step": 25170 }, { "grad_norm": 0.1785695105791092, "learning_rate": 8.203176632596892e-05, "loss": 0.0135, "step": 25180 }, { "grad_norm": 0.16130895912647247, "learning_rate": 8.20158934772462e-05, "loss": 0.0119, "step": 25190 }, { "grad_norm": 0.1430085450410843, "learning_rate": 8.200001515788294e-05, "loss": 0.0097, "step": 25200 }, { "grad_norm": 0.14859522879123688, "learning_rate": 8.198413137059228e-05, "loss": 0.0095, "step": 25210 }, { "grad_norm": 0.1820550560951233, "learning_rate": 8.196824211808835e-05, "loss": 0.0144, "step": 25220 }, { "grad_norm": 0.1551804095506668, "learning_rate": 8.195234740308617e-05, "loss": 0.0125, "step": 25230 }, { "grad_norm": 0.17808042466640472, "learning_rate": 8.193644722830171e-05, "loss": 0.0113, "step": 25240 }, { "grad_norm": 0.1416066437959671, "learning_rate": 8.19205415964519e-05, "loss": 0.0098, "step": 25250 }, { "grad_norm": 0.15036724507808685, "learning_rate": 8.190463051025456e-05, "loss": 0.0103, "step": 25260 }, { "grad_norm": 0.13608461618423462, "learning_rate": 8.188871397242843e-05, "loss": 0.0127, "step": 25270 }, { "grad_norm": 0.2137366086244583, "learning_rate": 8.187279198569326e-05, "loss": 0.0109, "step": 25280 }, { "grad_norm": 0.16456164419651031, "learning_rate": 8.185686455276966e-05, "loss": 0.0139, "step": 25290 }, { "grad_norm": 0.14578355848789215, "learning_rate": 8.184093167637921e-05, "loss": 0.0121, "step": 25300 }, { "grad_norm": 0.15478193759918213, "learning_rate": 8.182499335924437e-05, "loss": 0.0133, "step": 25310 }, { "grad_norm": 0.18802273273468018, "learning_rate": 8.18090496040886e-05, "loss": 0.0143, "step": 25320 }, { "grad_norm": 0.12880651652812958, "learning_rate": 8.179310041363621e-05, "loss": 0.011, "step": 25330 }, { "grad_norm": 0.17994245886802673, "learning_rate": 8.17771457906125e-05, "loss": 0.0111, "step": 25340 }, { "grad_norm": 0.1492520421743393, "learning_rate": 8.176118573774371e-05, "loss": 0.0102, "step": 25350 }, { "grad_norm": 0.18474872410297394, "learning_rate": 8.174522025775692e-05, "loss": 0.0112, "step": 25360 }, { "grad_norm": 0.13493207097053528, "learning_rate": 8.172924935338022e-05, "loss": 0.0127, "step": 25370 }, { "grad_norm": 0.13011349737644196, "learning_rate": 8.171327302734262e-05, "loss": 0.0128, "step": 25380 }, { "grad_norm": 0.12018559128046036, "learning_rate": 8.169729128237401e-05, "loss": 0.013, "step": 25390 }, { "grad_norm": 0.151035875082016, "learning_rate": 8.168130412120525e-05, "loss": 0.0123, "step": 25400 }, { "grad_norm": 0.14216026663780212, "learning_rate": 8.16653115465681e-05, "loss": 0.0139, "step": 25410 }, { "grad_norm": 0.15012358129024506, "learning_rate": 8.164931356119526e-05, "loss": 0.0138, "step": 25420 }, { "grad_norm": 0.12410815805196762, "learning_rate": 8.163331016782032e-05, "loss": 0.0105, "step": 25430 }, { "grad_norm": 0.13634471595287323, "learning_rate": 8.161730136917785e-05, "loss": 0.0137, "step": 25440 }, { "grad_norm": 0.16572225093841553, "learning_rate": 8.160128716800333e-05, "loss": 0.0108, "step": 25450 }, { "grad_norm": 0.1545735001564026, "learning_rate": 8.158526756703313e-05, "loss": 0.0147, "step": 25460 }, { "grad_norm": 0.11455975472927094, "learning_rate": 8.156924256900455e-05, "loss": 0.0158, "step": 25470 }, { "grad_norm": 0.13410848379135132, "learning_rate": 8.155321217665584e-05, "loss": 0.0158, "step": 25480 }, { "grad_norm": 0.1706921011209488, "learning_rate": 8.153717639272614e-05, "loss": 0.0141, "step": 25490 }, { "grad_norm": 0.19279845058918, "learning_rate": 8.152113521995555e-05, "loss": 0.0135, "step": 25500 }, { "grad_norm": 0.141779363155365, "learning_rate": 8.150508866108505e-05, "loss": 0.0127, "step": 25510 }, { "grad_norm": 0.21391621232032776, "learning_rate": 8.148903671885657e-05, "loss": 0.0147, "step": 25520 }, { "grad_norm": 0.1712331771850586, "learning_rate": 8.147297939601292e-05, "loss": 0.0134, "step": 25530 }, { "grad_norm": 0.20690587162971497, "learning_rate": 8.145691669529792e-05, "loss": 0.0141, "step": 25540 }, { "grad_norm": 0.1499333381652832, "learning_rate": 8.144084861945618e-05, "loss": 0.0089, "step": 25550 }, { "grad_norm": 0.16908253729343414, "learning_rate": 8.142477517123333e-05, "loss": 0.0113, "step": 25560 }, { "grad_norm": 0.19470351934432983, "learning_rate": 8.140869635337586e-05, "loss": 0.0128, "step": 25570 }, { "grad_norm": 0.19605739414691925, "learning_rate": 8.139261216863123e-05, "loss": 0.0135, "step": 25580 }, { "grad_norm": 0.16003303229808807, "learning_rate": 8.137652261974776e-05, "loss": 0.0148, "step": 25590 }, { "grad_norm": 0.17201508581638336, "learning_rate": 8.136042770947472e-05, "loss": 0.0155, "step": 25600 }, { "grad_norm": 0.19203481078147888, "learning_rate": 8.134432744056228e-05, "loss": 0.0118, "step": 25610 }, { "grad_norm": 0.1629059761762619, "learning_rate": 8.132822181576158e-05, "loss": 0.0134, "step": 25620 }, { "grad_norm": 0.18905843794345856, "learning_rate": 8.131211083782459e-05, "loss": 0.0139, "step": 25630 }, { "grad_norm": 0.20675796270370483, "learning_rate": 8.129599450950424e-05, "loss": 0.0115, "step": 25640 }, { "grad_norm": 0.14765027165412903, "learning_rate": 8.127987283355438e-05, "loss": 0.0121, "step": 25650 }, { "grad_norm": 0.1575659066438675, "learning_rate": 8.126374581272976e-05, "loss": 0.0131, "step": 25660 }, { "grad_norm": 0.12256266176700592, "learning_rate": 8.124761344978605e-05, "loss": 0.0169, "step": 25670 }, { "grad_norm": 0.10626032948493958, "learning_rate": 8.12314757474798e-05, "loss": 0.0127, "step": 25680 }, { "grad_norm": 0.14535346627235413, "learning_rate": 8.121533270856856e-05, "loss": 0.0097, "step": 25690 }, { "grad_norm": 0.17059193551540375, "learning_rate": 8.119918433581069e-05, "loss": 0.0129, "step": 25700 }, { "grad_norm": 0.16574954986572266, "learning_rate": 8.118303063196551e-05, "loss": 0.0139, "step": 25710 }, { "grad_norm": 0.1569165140390396, "learning_rate": 8.116687159979326e-05, "loss": 0.0107, "step": 25720 }, { "grad_norm": 0.16323402523994446, "learning_rate": 8.115070724205508e-05, "loss": 0.0127, "step": 25730 }, { "grad_norm": 0.1664363294839859, "learning_rate": 8.113453756151296e-05, "loss": 0.0141, "step": 25740 }, { "grad_norm": 0.16838745772838593, "learning_rate": 8.111836256092995e-05, "loss": 0.0138, "step": 25750 }, { "grad_norm": 0.24740271270275116, "learning_rate": 8.110218224306985e-05, "loss": 0.0141, "step": 25760 }, { "grad_norm": 0.15080486238002777, "learning_rate": 8.108599661069745e-05, "loss": 0.0114, "step": 25770 }, { "grad_norm": 0.18075525760650635, "learning_rate": 8.106980566657845e-05, "loss": 0.0113, "step": 25780 }, { "grad_norm": 0.19306203722953796, "learning_rate": 8.10536094134794e-05, "loss": 0.0131, "step": 25790 }, { "grad_norm": 0.14547811448574066, "learning_rate": 8.103740785416783e-05, "loss": 0.0107, "step": 25800 }, { "grad_norm": 0.15591572225093842, "learning_rate": 8.102120099141212e-05, "loss": 0.0123, "step": 25810 }, { "grad_norm": 0.13745857775211334, "learning_rate": 8.100498882798163e-05, "loss": 0.013, "step": 25820 }, { "grad_norm": 0.14568108320236206, "learning_rate": 8.09887713666465e-05, "loss": 0.0141, "step": 25830 }, { "grad_norm": 0.15248246490955353, "learning_rate": 8.09725486101779e-05, "loss": 0.0126, "step": 25840 }, { "grad_norm": 0.07536418735980988, "learning_rate": 8.095632056134784e-05, "loss": 0.0112, "step": 25850 }, { "grad_norm": 0.1768156737089157, "learning_rate": 8.094008722292925e-05, "loss": 0.0149, "step": 25860 }, { "grad_norm": 0.12879188358783722, "learning_rate": 8.092384859769598e-05, "loss": 0.0103, "step": 25870 }, { "grad_norm": 0.1859772801399231, "learning_rate": 8.090760468842275e-05, "loss": 0.0157, "step": 25880 }, { "grad_norm": 0.17298583686351776, "learning_rate": 8.089135549788521e-05, "loss": 0.011, "step": 25890 }, { "grad_norm": 0.20719796419143677, "learning_rate": 8.087510102885987e-05, "loss": 0.0112, "step": 25900 }, { "grad_norm": 0.25626543164253235, "learning_rate": 8.085884128412422e-05, "loss": 0.014, "step": 25910 }, { "grad_norm": 0.15399578213691711, "learning_rate": 8.084257626645659e-05, "loss": 0.0132, "step": 25920 }, { "grad_norm": 0.2069472223520279, "learning_rate": 8.08263059786362e-05, "loss": 0.0105, "step": 25930 }, { "grad_norm": 0.16342885792255402, "learning_rate": 8.081003042344325e-05, "loss": 0.0097, "step": 25940 }, { "grad_norm": 0.1632104218006134, "learning_rate": 8.079374960365872e-05, "loss": 0.0102, "step": 25950 }, { "grad_norm": 0.15121148526668549, "learning_rate": 8.077746352206463e-05, "loss": 0.0101, "step": 25960 }, { "grad_norm": 0.14169415831565857, "learning_rate": 8.076117218144377e-05, "loss": 0.0093, "step": 25970 }, { "grad_norm": 0.14929921925067902, "learning_rate": 8.074487558457991e-05, "loss": 0.0109, "step": 25980 }, { "grad_norm": 0.2494577318429947, "learning_rate": 8.072857373425768e-05, "loss": 0.011, "step": 25990 }, { "grad_norm": 0.155247300863266, "learning_rate": 8.071226663326264e-05, "loss": 0.0098, "step": 26000 }, { "grad_norm": 0.15639151632785797, "learning_rate": 8.069595428438121e-05, "loss": 0.0154, "step": 26010 }, { "grad_norm": 0.11754364520311356, "learning_rate": 8.067963669040072e-05, "loss": 0.01, "step": 26020 }, { "grad_norm": 0.1843928098678589, "learning_rate": 8.066331385410942e-05, "loss": 0.0142, "step": 26030 }, { "grad_norm": 0.185310959815979, "learning_rate": 8.064698577829641e-05, "loss": 0.0118, "step": 26040 }, { "grad_norm": 0.208592489361763, "learning_rate": 8.063065246575175e-05, "loss": 0.011, "step": 26050 }, { "grad_norm": 0.1646731197834015, "learning_rate": 8.061431391926631e-05, "loss": 0.012, "step": 26060 }, { "grad_norm": 0.1735275238752365, "learning_rate": 8.059797014163195e-05, "loss": 0.0128, "step": 26070 }, { "grad_norm": 0.12640872597694397, "learning_rate": 8.058162113564133e-05, "loss": 0.0127, "step": 26080 }, { "grad_norm": 0.23463015258312225, "learning_rate": 8.056526690408806e-05, "loss": 0.013, "step": 26090 }, { "grad_norm": 0.16346020996570587, "learning_rate": 8.054890744976666e-05, "loss": 0.0121, "step": 26100 }, { "grad_norm": 0.12600953876972198, "learning_rate": 8.053254277547248e-05, "loss": 0.0115, "step": 26110 }, { "grad_norm": 0.14788056910037994, "learning_rate": 8.051617288400182e-05, "loss": 0.0099, "step": 26120 }, { "grad_norm": 0.17344902455806732, "learning_rate": 8.049979777815182e-05, "loss": 0.0119, "step": 26130 }, { "grad_norm": 0.14866231381893158, "learning_rate": 8.048341746072054e-05, "loss": 0.0116, "step": 26140 }, { "grad_norm": 0.15063704550266266, "learning_rate": 8.046703193450696e-05, "loss": 0.0135, "step": 26150 }, { "grad_norm": 0.11102072149515152, "learning_rate": 8.04506412023109e-05, "loss": 0.0114, "step": 26160 }, { "grad_norm": 0.1543712168931961, "learning_rate": 8.043424526693306e-05, "loss": 0.0131, "step": 26170 }, { "grad_norm": 0.13784180581569672, "learning_rate": 8.04178441311751e-05, "loss": 0.0101, "step": 26180 }, { "grad_norm": 0.17017705738544464, "learning_rate": 8.04014377978395e-05, "loss": 0.0107, "step": 26190 }, { "grad_norm": 0.17562197148799896, "learning_rate": 8.038502626972967e-05, "loss": 0.0125, "step": 26200 }, { "grad_norm": 0.1460225135087967, "learning_rate": 8.036860954964989e-05, "loss": 0.0146, "step": 26210 }, { "grad_norm": 0.16391752660274506, "learning_rate": 8.035218764040531e-05, "loss": 0.0116, "step": 26220 }, { "grad_norm": 0.22932566702365875, "learning_rate": 8.033576054480199e-05, "loss": 0.0152, "step": 26230 }, { "grad_norm": 0.14550255239009857, "learning_rate": 8.031932826564688e-05, "loss": 0.0115, "step": 26240 }, { "grad_norm": 0.185557559132576, "learning_rate": 8.030289080574782e-05, "loss": 0.0108, "step": 26250 }, { "grad_norm": 0.14722786843776703, "learning_rate": 8.028644816791349e-05, "loss": 0.0119, "step": 26260 }, { "grad_norm": 0.16142989695072174, "learning_rate": 8.027000035495351e-05, "loss": 0.0109, "step": 26270 }, { "grad_norm": 0.09813769161701202, "learning_rate": 8.025354736967836e-05, "loss": 0.0089, "step": 26280 }, { "grad_norm": 0.2099316269159317, "learning_rate": 8.023708921489941e-05, "loss": 0.0115, "step": 26290 }, { "grad_norm": 0.15031543374061584, "learning_rate": 8.022062589342887e-05, "loss": 0.0124, "step": 26300 }, { "grad_norm": 0.18674662709236145, "learning_rate": 8.020415740807993e-05, "loss": 0.0106, "step": 26310 }, { "grad_norm": 0.13057738542556763, "learning_rate": 8.018768376166656e-05, "loss": 0.012, "step": 26320 }, { "grad_norm": 0.16832596063613892, "learning_rate": 8.017120495700368e-05, "loss": 0.0134, "step": 26330 }, { "grad_norm": 0.16871751844882965, "learning_rate": 8.015472099690704e-05, "loss": 0.0123, "step": 26340 }, { "grad_norm": 0.15893687307834625, "learning_rate": 8.013823188419332e-05, "loss": 0.015, "step": 26350 }, { "grad_norm": 0.2056180238723755, "learning_rate": 8.012173762168006e-05, "loss": 0.0107, "step": 26360 }, { "grad_norm": 0.12794485688209534, "learning_rate": 8.010523821218567e-05, "loss": 0.0107, "step": 26370 }, { "grad_norm": 0.1581486463546753, "learning_rate": 8.008873365852945e-05, "loss": 0.0124, "step": 26380 }, { "grad_norm": 0.1501310169696808, "learning_rate": 8.007222396353157e-05, "loss": 0.0132, "step": 26390 }, { "grad_norm": 0.13273771107196808, "learning_rate": 8.00557091300131e-05, "loss": 0.0115, "step": 26400 }, { "grad_norm": 0.20065724849700928, "learning_rate": 8.003918916079597e-05, "loss": 0.012, "step": 26410 }, { "grad_norm": 0.15628136694431305, "learning_rate": 8.002266405870298e-05, "loss": 0.0118, "step": 26420 }, { "grad_norm": 0.18247726559638977, "learning_rate": 8.000613382655782e-05, "loss": 0.0116, "step": 26430 }, { "grad_norm": 0.15224792063236237, "learning_rate": 7.998959846718505e-05, "loss": 0.0132, "step": 26440 }, { "grad_norm": 0.13236474990844727, "learning_rate": 7.997305798341012e-05, "loss": 0.0112, "step": 26450 }, { "grad_norm": 0.20390202105045319, "learning_rate": 7.995651237805937e-05, "loss": 0.0113, "step": 26460 }, { "grad_norm": 0.16588576138019562, "learning_rate": 7.993996165395996e-05, "loss": 0.0111, "step": 26470 }, { "grad_norm": 0.17002952098846436, "learning_rate": 7.992340581393996e-05, "loss": 0.0121, "step": 26480 }, { "grad_norm": 0.1210218220949173, "learning_rate": 7.990684486082831e-05, "loss": 0.0126, "step": 26490 }, { "grad_norm": 0.12252987921237946, "learning_rate": 7.989027879745482e-05, "loss": 0.0101, "step": 26500 }, { "grad_norm": 0.18986709415912628, "learning_rate": 7.98737076266502e-05, "loss": 0.0115, "step": 26510 }, { "grad_norm": 0.18068265914916992, "learning_rate": 7.985713135124598e-05, "loss": 0.0111, "step": 26520 }, { "grad_norm": 0.19357721507549286, "learning_rate": 7.98405499740746e-05, "loss": 0.0118, "step": 26530 }, { "grad_norm": 0.2565384805202484, "learning_rate": 7.98239634979694e-05, "loss": 0.0137, "step": 26540 }, { "grad_norm": 0.18816809356212616, "learning_rate": 7.98073719257645e-05, "loss": 0.0125, "step": 26550 }, { "grad_norm": 0.1375349760055542, "learning_rate": 7.979077526029499e-05, "loss": 0.0117, "step": 26560 }, { "grad_norm": 0.16399036347866058, "learning_rate": 7.977417350439675e-05, "loss": 0.0145, "step": 26570 }, { "grad_norm": 0.1508973091840744, "learning_rate": 7.97575666609066e-05, "loss": 0.0128, "step": 26580 }, { "grad_norm": 0.15271009504795074, "learning_rate": 7.974095473266216e-05, "loss": 0.0105, "step": 26590 }, { "grad_norm": 0.12570326030254364, "learning_rate": 7.972433772250198e-05, "loss": 0.0114, "step": 26600 }, { "grad_norm": 0.1720360517501831, "learning_rate": 7.970771563326544e-05, "loss": 0.0139, "step": 26610 }, { "grad_norm": 0.13691431283950806, "learning_rate": 7.96910884677928e-05, "loss": 0.0126, "step": 26620 }, { "grad_norm": 0.2278885394334793, "learning_rate": 7.967445622892523e-05, "loss": 0.011, "step": 26630 }, { "grad_norm": 0.16774122416973114, "learning_rate": 7.965781891950465e-05, "loss": 0.0119, "step": 26640 }, { "grad_norm": 0.1371527910232544, "learning_rate": 7.964117654237397e-05, "loss": 0.0102, "step": 26650 }, { "grad_norm": 0.14594562351703644, "learning_rate": 7.962452910037692e-05, "loss": 0.0121, "step": 26660 }, { "grad_norm": 0.13582536578178406, "learning_rate": 7.96078765963581e-05, "loss": 0.0112, "step": 26670 }, { "grad_norm": 0.10741212964057922, "learning_rate": 7.95912190331629e-05, "loss": 0.0097, "step": 26680 }, { "grad_norm": 0.11986242979764938, "learning_rate": 7.957455641363772e-05, "loss": 0.0104, "step": 26690 }, { "grad_norm": 0.17212294042110443, "learning_rate": 7.955788874062968e-05, "loss": 0.011, "step": 26700 }, { "grad_norm": 0.1517718881368637, "learning_rate": 7.95412160169869e-05, "loss": 0.013, "step": 26710 }, { "grad_norm": 0.2138395756483078, "learning_rate": 7.952453824555824e-05, "loss": 0.011, "step": 26720 }, { "grad_norm": 0.15925529599189758, "learning_rate": 7.95078554291935e-05, "loss": 0.0108, "step": 26730 }, { "grad_norm": 0.15392178297042847, "learning_rate": 7.94911675707433e-05, "loss": 0.0092, "step": 26740 }, { "grad_norm": 0.1424666941165924, "learning_rate": 7.947447467305915e-05, "loss": 0.0111, "step": 26750 }, { "grad_norm": 0.1775674670934677, "learning_rate": 7.94577767389934e-05, "loss": 0.0113, "step": 26760 }, { "grad_norm": 0.18774162232875824, "learning_rate": 7.944107377139928e-05, "loss": 0.0104, "step": 26770 }, { "grad_norm": 0.16873939335346222, "learning_rate": 7.942436577313088e-05, "loss": 0.0134, "step": 26780 }, { "grad_norm": 0.13619457185268402, "learning_rate": 7.940765274704312e-05, "loss": 0.0087, "step": 26790 }, { "grad_norm": 0.24745164811611176, "learning_rate": 7.939093469599181e-05, "loss": 0.0125, "step": 26800 }, { "grad_norm": 0.11479558795690536, "learning_rate": 7.93742116228336e-05, "loss": 0.0107, "step": 26810 }, { "grad_norm": 0.17479853332042694, "learning_rate": 7.935748353042602e-05, "loss": 0.0117, "step": 26820 }, { "grad_norm": 0.1878834068775177, "learning_rate": 7.934075042162744e-05, "loss": 0.0132, "step": 26830 }, { "grad_norm": 0.18755744397640228, "learning_rate": 7.932401229929705e-05, "loss": 0.0108, "step": 26840 }, { "grad_norm": 0.12697067856788635, "learning_rate": 7.9307269166295e-05, "loss": 0.0107, "step": 26850 }, { "grad_norm": 0.14109089970588684, "learning_rate": 7.92905210254822e-05, "loss": 0.01, "step": 26860 }, { "grad_norm": 0.22203435003757477, "learning_rate": 7.927376787972045e-05, "loss": 0.014, "step": 26870 }, { "grad_norm": 0.19398342072963715, "learning_rate": 7.92570097318724e-05, "loss": 0.0126, "step": 26880 }, { "grad_norm": 0.17667175829410553, "learning_rate": 7.924024658480158e-05, "loss": 0.0135, "step": 26890 }, { "grad_norm": 0.16498951613903046, "learning_rate": 7.922347844137233e-05, "loss": 0.0122, "step": 26900 }, { "grad_norm": 0.15234173834323883, "learning_rate": 7.92067053044499e-05, "loss": 0.0109, "step": 26910 }, { "grad_norm": 0.1742376834154129, "learning_rate": 7.918992717690031e-05, "loss": 0.0114, "step": 26920 }, { "grad_norm": 0.19729828834533691, "learning_rate": 7.917314406159053e-05, "loss": 0.0113, "step": 26930 }, { "grad_norm": 0.18975450098514557, "learning_rate": 7.915635596138832e-05, "loss": 0.0121, "step": 26940 }, { "grad_norm": 0.16328950226306915, "learning_rate": 7.913956287916228e-05, "loss": 0.0152, "step": 26950 }, { "grad_norm": 0.18426239490509033, "learning_rate": 7.912276481778193e-05, "loss": 0.0119, "step": 26960 }, { "grad_norm": 0.1601221263408661, "learning_rate": 7.910596178011759e-05, "loss": 0.011, "step": 26970 }, { "grad_norm": 0.16613325476646423, "learning_rate": 7.908915376904043e-05, "loss": 0.0114, "step": 26980 }, { "grad_norm": 0.16806760430335999, "learning_rate": 7.907234078742247e-05, "loss": 0.0121, "step": 26990 }, { "grad_norm": 0.18018382787704468, "learning_rate": 7.90555228381366e-05, "loss": 0.0104, "step": 27000 }, { "grad_norm": 0.13873830437660217, "learning_rate": 7.903869992405656e-05, "loss": 0.0128, "step": 27010 }, { "grad_norm": 0.13020159304141998, "learning_rate": 7.902187204805691e-05, "loss": 0.0109, "step": 27020 }, { "grad_norm": 0.152888685464859, "learning_rate": 7.900503921301308e-05, "loss": 0.0107, "step": 27030 }, { "grad_norm": 0.12735894322395325, "learning_rate": 7.898820142180133e-05, "loss": 0.0098, "step": 27040 }, { "grad_norm": 0.14879879355430603, "learning_rate": 7.897135867729879e-05, "loss": 0.0094, "step": 27050 }, { "grad_norm": 0.15227095782756805, "learning_rate": 7.89545109823834e-05, "loss": 0.0122, "step": 27060 }, { "grad_norm": 0.17765343189239502, "learning_rate": 7.8937658339934e-05, "loss": 0.0107, "step": 27070 }, { "grad_norm": 0.1725626289844513, "learning_rate": 7.892080075283026e-05, "loss": 0.0117, "step": 27080 }, { "grad_norm": 0.13974811136722565, "learning_rate": 7.890393822395263e-05, "loss": 0.0096, "step": 27090 }, { "grad_norm": 0.1634090691804886, "learning_rate": 7.88870707561825e-05, "loss": 0.0099, "step": 27100 }, { "grad_norm": 0.13589464128017426, "learning_rate": 7.887019835240203e-05, "loss": 0.0102, "step": 27110 }, { "grad_norm": 0.13372035324573517, "learning_rate": 7.885332101549427e-05, "loss": 0.0105, "step": 27120 }, { "grad_norm": 0.21007198095321655, "learning_rate": 7.883643874834308e-05, "loss": 0.0108, "step": 27130 }, { "grad_norm": 0.12590208649635315, "learning_rate": 7.881955155383321e-05, "loss": 0.0089, "step": 27140 }, { "grad_norm": 0.12979045510292053, "learning_rate": 7.880265943485017e-05, "loss": 0.0117, "step": 27150 }, { "grad_norm": 0.15509071946144104, "learning_rate": 7.878576239428038e-05, "loss": 0.0142, "step": 27160 }, { "grad_norm": 0.20132188498973846, "learning_rate": 7.87688604350111e-05, "loss": 0.015, "step": 27170 }, { "grad_norm": 0.10876022279262543, "learning_rate": 7.875195355993042e-05, "loss": 0.0113, "step": 27180 }, { "grad_norm": 0.1475232094526291, "learning_rate": 7.873504177192724e-05, "loss": 0.0143, "step": 27190 }, { "grad_norm": 0.13021819293498993, "learning_rate": 7.87181250738913e-05, "loss": 0.0102, "step": 27200 }, { "grad_norm": 0.1490197628736496, "learning_rate": 7.870120346871324e-05, "loss": 0.0112, "step": 27210 }, { "grad_norm": 0.1449505090713501, "learning_rate": 7.86842769592845e-05, "loss": 0.0098, "step": 27220 }, { "grad_norm": 0.19084563851356506, "learning_rate": 7.866734554849732e-05, "loss": 0.0106, "step": 27230 }, { "grad_norm": 0.1594018191099167, "learning_rate": 7.865040923924486e-05, "loss": 0.0104, "step": 27240 }, { "grad_norm": 0.1627141684293747, "learning_rate": 7.863346803442104e-05, "loss": 0.0127, "step": 27250 }, { "grad_norm": 0.16617703437805176, "learning_rate": 7.861652193692067e-05, "loss": 0.0129, "step": 27260 }, { "grad_norm": 0.17088690400123596, "learning_rate": 7.859957094963937e-05, "loss": 0.0117, "step": 27270 }, { "grad_norm": 0.11604432016611099, "learning_rate": 7.858261507547357e-05, "loss": 0.0114, "step": 27280 }, { "grad_norm": 0.10807806998491287, "learning_rate": 7.856565431732061e-05, "loss": 0.0116, "step": 27290 }, { "grad_norm": 0.1811112016439438, "learning_rate": 7.854868867807859e-05, "loss": 0.0137, "step": 27300 }, { "grad_norm": 0.13965992629528046, "learning_rate": 7.85317181606465e-05, "loss": 0.0121, "step": 27310 }, { "grad_norm": 0.1698170304298401, "learning_rate": 7.85147427679241e-05, "loss": 0.0115, "step": 27320 }, { "grad_norm": 0.1432223916053772, "learning_rate": 7.849776250281205e-05, "loss": 0.0093, "step": 27330 }, { "grad_norm": 0.12268104404211044, "learning_rate": 7.84807773682118e-05, "loss": 0.0095, "step": 27340 }, { "grad_norm": 0.14617981016635895, "learning_rate": 7.846378736702565e-05, "loss": 0.0099, "step": 27350 }, { "grad_norm": 0.12168613076210022, "learning_rate": 7.844679250215671e-05, "loss": 0.0103, "step": 27360 }, { "grad_norm": 0.16508975625038147, "learning_rate": 7.842979277650898e-05, "loss": 0.0127, "step": 27370 }, { "grad_norm": 0.22557561099529266, "learning_rate": 7.84127881929872e-05, "loss": 0.0111, "step": 27380 }, { "grad_norm": 0.13268914818763733, "learning_rate": 7.839577875449704e-05, "loss": 0.0104, "step": 27390 }, { "grad_norm": 0.1901417374610901, "learning_rate": 7.837876446394489e-05, "loss": 0.0105, "step": 27400 }, { "grad_norm": 0.14102284610271454, "learning_rate": 7.836174532423805e-05, "loss": 0.0116, "step": 27410 }, { "grad_norm": 0.12556155025959015, "learning_rate": 7.834472133828466e-05, "loss": 0.0111, "step": 27420 }, { "grad_norm": 0.16066637635231018, "learning_rate": 7.832769250899359e-05, "loss": 0.0126, "step": 27430 }, { "grad_norm": 0.17660780251026154, "learning_rate": 7.831065883927464e-05, "loss": 0.0116, "step": 27440 }, { "grad_norm": 0.16871964931488037, "learning_rate": 7.829362033203841e-05, "loss": 0.0125, "step": 27450 }, { "grad_norm": 0.15161308646202087, "learning_rate": 7.827657699019628e-05, "loss": 0.0092, "step": 27460 }, { "grad_norm": 0.11619845777750015, "learning_rate": 7.825952881666052e-05, "loss": 0.0135, "step": 27470 }, { "grad_norm": 0.12464790791273117, "learning_rate": 7.824247581434418e-05, "loss": 0.0105, "step": 27480 }, { "grad_norm": 0.14634789526462555, "learning_rate": 7.822541798616116e-05, "loss": 0.0097, "step": 27490 }, { "grad_norm": 0.14423377811908722, "learning_rate": 7.820835533502617e-05, "loss": 0.0111, "step": 27500 }, { "grad_norm": 0.16490429639816284, "learning_rate": 7.819128786385475e-05, "loss": 0.0101, "step": 27510 }, { "grad_norm": 0.12880562245845795, "learning_rate": 7.817421557556329e-05, "loss": 0.0105, "step": 27520 }, { "grad_norm": 0.17587608098983765, "learning_rate": 7.815713847306893e-05, "loss": 0.0092, "step": 27530 }, { "grad_norm": 0.1745913177728653, "learning_rate": 7.81400565592897e-05, "loss": 0.012, "step": 27540 }, { "grad_norm": 0.16409403085708618, "learning_rate": 7.812296983714444e-05, "loss": 0.0111, "step": 27550 }, { "grad_norm": 0.13592642545700073, "learning_rate": 7.810587830955281e-05, "loss": 0.0104, "step": 27560 }, { "grad_norm": 0.13066108524799347, "learning_rate": 7.808878197943528e-05, "loss": 0.0113, "step": 27570 }, { "grad_norm": 0.15302598476409912, "learning_rate": 7.807168084971312e-05, "loss": 0.0115, "step": 27580 }, { "grad_norm": 0.14680072665214539, "learning_rate": 7.805457492330849e-05, "loss": 0.0129, "step": 27590 }, { "grad_norm": 0.10788451135158539, "learning_rate": 7.803746420314428e-05, "loss": 0.0098, "step": 27600 }, { "grad_norm": 0.12029585987329483, "learning_rate": 7.802034869214428e-05, "loss": 0.013, "step": 27610 }, { "grad_norm": 0.16825374960899353, "learning_rate": 7.800322839323303e-05, "loss": 0.0151, "step": 27620 }, { "grad_norm": 0.1683899611234665, "learning_rate": 7.798610330933593e-05, "loss": 0.0126, "step": 27630 }, { "grad_norm": 0.19792847335338593, "learning_rate": 7.796897344337922e-05, "loss": 0.0109, "step": 27640 }, { "grad_norm": 0.1417386680841446, "learning_rate": 7.795183879828989e-05, "loss": 0.0113, "step": 27650 }, { "grad_norm": 0.16939213871955872, "learning_rate": 7.793469937699579e-05, "loss": 0.0109, "step": 27660 }, { "grad_norm": 0.18531736731529236, "learning_rate": 7.791755518242558e-05, "loss": 0.0104, "step": 27670 }, { "grad_norm": 0.14621081948280334, "learning_rate": 7.790040621750876e-05, "loss": 0.009, "step": 27680 }, { "grad_norm": 0.14067621529102325, "learning_rate": 7.788325248517558e-05, "loss": 0.0112, "step": 27690 }, { "grad_norm": 0.20170247554779053, "learning_rate": 7.786609398835715e-05, "loss": 0.0131, "step": 27700 }, { "grad_norm": 0.16121985018253326, "learning_rate": 7.784893072998541e-05, "loss": 0.011, "step": 27710 }, { "grad_norm": 0.16764888167381287, "learning_rate": 7.783176271299306e-05, "loss": 0.0096, "step": 27720 }, { "grad_norm": 0.1457129418849945, "learning_rate": 7.781458994031368e-05, "loss": 0.0106, "step": 27730 }, { "grad_norm": 0.13039763271808624, "learning_rate": 7.779741241488161e-05, "loss": 0.0101, "step": 27740 }, { "grad_norm": 0.13382361829280853, "learning_rate": 7.7780230139632e-05, "loss": 0.0125, "step": 27750 }, { "grad_norm": 0.11116662621498108, "learning_rate": 7.776304311750087e-05, "loss": 0.011, "step": 27760 }, { "grad_norm": 0.16091075539588928, "learning_rate": 7.7745851351425e-05, "loss": 0.0131, "step": 27770 }, { "grad_norm": 0.18174050748348236, "learning_rate": 7.772865484434197e-05, "loss": 0.009, "step": 27780 }, { "grad_norm": 0.1573939323425293, "learning_rate": 7.77114535991902e-05, "loss": 0.0114, "step": 27790 }, { "grad_norm": 0.17984674870967865, "learning_rate": 7.769424761890893e-05, "loss": 0.0111, "step": 27800 }, { "grad_norm": 0.11755476891994476, "learning_rate": 7.767703690643817e-05, "loss": 0.0108, "step": 27810 }, { "grad_norm": 0.16693851351737976, "learning_rate": 7.76598214647188e-05, "loss": 0.013, "step": 27820 }, { "grad_norm": 0.12074416130781174, "learning_rate": 7.764260129669241e-05, "loss": 0.0107, "step": 27830 }, { "grad_norm": 0.12331463396549225, "learning_rate": 7.76253764053015e-05, "loss": 0.0098, "step": 27840 }, { "grad_norm": 0.1735437959432602, "learning_rate": 7.760814679348932e-05, "loss": 0.0108, "step": 27850 }, { "grad_norm": 0.15117444097995758, "learning_rate": 7.759091246419992e-05, "loss": 0.0116, "step": 27860 }, { "grad_norm": 0.15478628873825073, "learning_rate": 7.75736734203782e-05, "loss": 0.0132, "step": 27870 }, { "grad_norm": 0.17551539838314056, "learning_rate": 7.755642966496985e-05, "loss": 0.0126, "step": 27880 }, { "grad_norm": 0.1579257696866989, "learning_rate": 7.753918120092132e-05, "loss": 0.0131, "step": 27890 }, { "grad_norm": 0.17926335334777832, "learning_rate": 7.752192803117993e-05, "loss": 0.0102, "step": 27900 }, { "grad_norm": 0.15957532823085785, "learning_rate": 7.750467015869377e-05, "loss": 0.0137, "step": 27910 }, { "grad_norm": 0.16586631536483765, "learning_rate": 7.748740758641174e-05, "loss": 0.0105, "step": 27920 }, { "grad_norm": 0.15845799446105957, "learning_rate": 7.74701403172835e-05, "loss": 0.0133, "step": 27930 }, { "grad_norm": 0.19816724956035614, "learning_rate": 7.745286835425962e-05, "loss": 0.0109, "step": 27940 }, { "grad_norm": 0.17457184195518494, "learning_rate": 7.743559170029138e-05, "loss": 0.014, "step": 27950 }, { "grad_norm": 0.16953645646572113, "learning_rate": 7.741831035833087e-05, "loss": 0.0108, "step": 27960 }, { "grad_norm": 0.13702651858329773, "learning_rate": 7.740102433133102e-05, "loss": 0.0123, "step": 27970 }, { "grad_norm": 0.16514542698860168, "learning_rate": 7.738373362224553e-05, "loss": 0.0104, "step": 27980 }, { "grad_norm": 0.15640267729759216, "learning_rate": 7.73664382340289e-05, "loss": 0.0125, "step": 27990 }, { "grad_norm": 0.136526957154274, "learning_rate": 7.734913816963647e-05, "loss": 0.01, "step": 28000 }, { "grad_norm": 0.18432950973510742, "learning_rate": 7.73318334320243e-05, "loss": 0.012, "step": 28010 }, { "grad_norm": 0.16978998482227325, "learning_rate": 7.731452402414934e-05, "loss": 0.0119, "step": 28020 }, { "grad_norm": 0.23961509764194489, "learning_rate": 7.729720994896928e-05, "loss": 0.0114, "step": 28030 }, { "grad_norm": 0.10591244697570801, "learning_rate": 7.727989120944262e-05, "loss": 0.0112, "step": 28040 }, { "grad_norm": 0.1730319708585739, "learning_rate": 7.726256780852865e-05, "loss": 0.0132, "step": 28050 }, { "grad_norm": 0.20611095428466797, "learning_rate": 7.724523974918749e-05, "loss": 0.0123, "step": 28060 }, { "grad_norm": 0.1881512850522995, "learning_rate": 7.722790703438002e-05, "loss": 0.0102, "step": 28070 }, { "grad_norm": 0.1868513971567154, "learning_rate": 7.72105696670679e-05, "loss": 0.01, "step": 28080 }, { "grad_norm": 0.1376691311597824, "learning_rate": 7.719322765021364e-05, "loss": 0.0088, "step": 28090 }, { "grad_norm": 0.16366341710090637, "learning_rate": 7.717588098678051e-05, "loss": 0.0133, "step": 28100 }, { "grad_norm": 0.1650005728006363, "learning_rate": 7.715852967973258e-05, "loss": 0.0101, "step": 28110 }, { "grad_norm": 0.15672451257705688, "learning_rate": 7.714117373203474e-05, "loss": 0.0102, "step": 28120 }, { "grad_norm": 0.13281047344207764, "learning_rate": 7.712381314665259e-05, "loss": 0.0093, "step": 28130 }, { "grad_norm": 0.13001978397369385, "learning_rate": 7.710644792655261e-05, "loss": 0.0114, "step": 28140 }, { "grad_norm": 0.16325928270816803, "learning_rate": 7.708907807470207e-05, "loss": 0.0087, "step": 28150 }, { "grad_norm": 0.10252027213573456, "learning_rate": 7.707170359406896e-05, "loss": 0.0116, "step": 28160 }, { "grad_norm": 0.17680715024471283, "learning_rate": 7.705432448762213e-05, "loss": 0.0136, "step": 28170 }, { "grad_norm": 0.15465687215328217, "learning_rate": 7.703694075833117e-05, "loss": 0.0101, "step": 28180 }, { "grad_norm": 0.1410929560661316, "learning_rate": 7.70195524091665e-05, "loss": 0.0127, "step": 28190 }, { "grad_norm": 0.16019687056541443, "learning_rate": 7.70021594430993e-05, "loss": 0.0108, "step": 28200 }, { "grad_norm": 0.10341254621744156, "learning_rate": 7.698476186310157e-05, "loss": 0.01, "step": 28210 }, { "grad_norm": 0.20392575860023499, "learning_rate": 7.696735967214608e-05, "loss": 0.0115, "step": 28220 }, { "grad_norm": 0.1452796757221222, "learning_rate": 7.694995287320636e-05, "loss": 0.0099, "step": 28230 }, { "grad_norm": 0.17120520770549774, "learning_rate": 7.693254146925679e-05, "loss": 0.0111, "step": 28240 }, { "grad_norm": 0.19986483454704285, "learning_rate": 7.691512546327251e-05, "loss": 0.011, "step": 28250 }, { "grad_norm": 0.11908718198537827, "learning_rate": 7.689770485822939e-05, "loss": 0.0112, "step": 28260 }, { "grad_norm": 0.1270158588886261, "learning_rate": 7.688027965710416e-05, "loss": 0.0095, "step": 28270 }, { "grad_norm": 0.1622220128774643, "learning_rate": 7.686284986287433e-05, "loss": 0.0103, "step": 28280 }, { "grad_norm": 0.14751943945884705, "learning_rate": 7.684541547851817e-05, "loss": 0.0094, "step": 28290 }, { "grad_norm": 0.1326712816953659, "learning_rate": 7.68279765070147e-05, "loss": 0.0103, "step": 28300 }, { "grad_norm": 0.11839442700147629, "learning_rate": 7.68105329513438e-05, "loss": 0.0119, "step": 28310 }, { "grad_norm": 0.3558816909790039, "learning_rate": 7.67930848144861e-05, "loss": 0.0124, "step": 28320 }, { "grad_norm": 0.11014202237129211, "learning_rate": 7.6775632099423e-05, "loss": 0.0101, "step": 28330 }, { "grad_norm": 0.15765829384326935, "learning_rate": 7.675817480913667e-05, "loss": 0.0095, "step": 28340 }, { "grad_norm": 0.18288229405879974, "learning_rate": 7.674071294661011e-05, "loss": 0.0099, "step": 28350 }, { "grad_norm": 0.13849559426307678, "learning_rate": 7.672324651482707e-05, "loss": 0.009, "step": 28360 }, { "grad_norm": 0.19355662167072296, "learning_rate": 7.670577551677209e-05, "loss": 0.0087, "step": 28370 }, { "grad_norm": 0.17009365558624268, "learning_rate": 7.668829995543047e-05, "loss": 0.0106, "step": 28380 }, { "grad_norm": 0.1732129603624344, "learning_rate": 7.667081983378832e-05, "loss": 0.0112, "step": 28390 }, { "grad_norm": 0.12878815829753876, "learning_rate": 7.66533351548325e-05, "loss": 0.0087, "step": 28400 }, { "grad_norm": 0.13205823302268982, "learning_rate": 7.663584592155069e-05, "loss": 0.0145, "step": 28410 }, { "grad_norm": 0.16234441101551056, "learning_rate": 7.661835213693129e-05, "loss": 0.01, "step": 28420 }, { "grad_norm": 0.17726518213748932, "learning_rate": 7.660085380396353e-05, "loss": 0.0125, "step": 28430 }, { "grad_norm": 0.12196049839258194, "learning_rate": 7.658335092563738e-05, "loss": 0.0108, "step": 28440 }, { "grad_norm": 0.17009244859218597, "learning_rate": 7.656584350494362e-05, "loss": 0.0118, "step": 28450 }, { "grad_norm": 0.2554573714733124, "learning_rate": 7.654833154487378e-05, "loss": 0.0103, "step": 28460 }, { "grad_norm": 0.18397116661071777, "learning_rate": 7.653081504842017e-05, "loss": 0.013, "step": 28470 }, { "grad_norm": 0.2010524421930313, "learning_rate": 7.65132940185759e-05, "loss": 0.0129, "step": 28480 }, { "grad_norm": 0.1802714467048645, "learning_rate": 7.649576845833481e-05, "loss": 0.0102, "step": 28490 }, { "grad_norm": 0.15712298452854156, "learning_rate": 7.647823837069156e-05, "loss": 0.0114, "step": 28500 }, { "grad_norm": 0.14053697884082794, "learning_rate": 7.646070375864156e-05, "loss": 0.0089, "step": 28510 }, { "grad_norm": 0.1572110950946808, "learning_rate": 7.644316462518099e-05, "loss": 0.0094, "step": 28520 }, { "grad_norm": 0.15088234841823578, "learning_rate": 7.642562097330679e-05, "loss": 0.0112, "step": 28530 }, { "grad_norm": 0.14920207858085632, "learning_rate": 7.640807280601671e-05, "loss": 0.0103, "step": 28540 }, { "grad_norm": 0.17217811942100525, "learning_rate": 7.639052012630927e-05, "loss": 0.0111, "step": 28550 }, { "grad_norm": 0.14647777378559113, "learning_rate": 7.63729629371837e-05, "loss": 0.0114, "step": 28560 }, { "grad_norm": 0.13494804501533508, "learning_rate": 7.635540124164009e-05, "loss": 0.0098, "step": 28570 }, { "grad_norm": 0.1362437903881073, "learning_rate": 7.633783504267922e-05, "loss": 0.0095, "step": 28580 }, { "grad_norm": 0.19214902818202972, "learning_rate": 7.632026434330269e-05, "loss": 0.0106, "step": 28590 }, { "grad_norm": 0.1349097639322281, "learning_rate": 7.630268914651282e-05, "loss": 0.0095, "step": 28600 }, { "grad_norm": 0.22055162489414215, "learning_rate": 7.628510945531278e-05, "loss": 0.0102, "step": 28610 }, { "grad_norm": 0.09885627776384354, "learning_rate": 7.626752527270641e-05, "loss": 0.0103, "step": 28620 }, { "grad_norm": 0.17296786606311798, "learning_rate": 7.62499366016984e-05, "loss": 0.0123, "step": 28630 }, { "grad_norm": 0.1708535999059677, "learning_rate": 7.623234344529416e-05, "loss": 0.0094, "step": 28640 }, { "grad_norm": 0.18916194140911102, "learning_rate": 7.62147458064999e-05, "loss": 0.0117, "step": 28650 }, { "grad_norm": 0.15176749229431152, "learning_rate": 7.619714368832254e-05, "loss": 0.0087, "step": 28660 }, { "grad_norm": 0.14470799267292023, "learning_rate": 7.61795370937698e-05, "loss": 0.0087, "step": 28670 }, { "grad_norm": 0.1977493166923523, "learning_rate": 7.61619260258502e-05, "loss": 0.0121, "step": 28680 }, { "grad_norm": 0.17504513263702393, "learning_rate": 7.614431048757298e-05, "loss": 0.0116, "step": 28690 }, { "grad_norm": 0.16534166038036346, "learning_rate": 7.612669048194814e-05, "loss": 0.0121, "step": 28700 }, { "grad_norm": 0.20267824828624725, "learning_rate": 7.610906601198646e-05, "loss": 0.0111, "step": 28710 }, { "grad_norm": 0.11677301675081253, "learning_rate": 7.60914370806995e-05, "loss": 0.009, "step": 28720 }, { "grad_norm": 0.1342749148607254, "learning_rate": 7.607380369109953e-05, "loss": 0.0121, "step": 28730 }, { "grad_norm": 0.15270008146762848, "learning_rate": 7.605616584619961e-05, "loss": 0.0107, "step": 28740 }, { "grad_norm": 0.17320628464221954, "learning_rate": 7.603852354901362e-05, "loss": 0.0118, "step": 28750 }, { "grad_norm": 0.18512146174907684, "learning_rate": 7.602087680255609e-05, "loss": 0.0102, "step": 28760 }, { "grad_norm": 0.1272171437740326, "learning_rate": 7.600322560984238e-05, "loss": 0.0111, "step": 28770 }, { "grad_norm": 0.1411169171333313, "learning_rate": 7.598556997388863e-05, "loss": 0.0121, "step": 28780 }, { "grad_norm": 0.13254356384277344, "learning_rate": 7.596790989771166e-05, "loss": 0.0083, "step": 28790 }, { "grad_norm": 0.12499869614839554, "learning_rate": 7.595024538432914e-05, "loss": 0.0101, "step": 28800 }, { "grad_norm": 0.15044516324996948, "learning_rate": 7.59325764367594e-05, "loss": 0.0098, "step": 28810 }, { "grad_norm": 0.1405669003725052, "learning_rate": 7.59149030580216e-05, "loss": 0.0126, "step": 28820 }, { "grad_norm": 0.1588500291109085, "learning_rate": 7.589722525113562e-05, "loss": 0.0096, "step": 28830 }, { "grad_norm": 0.16035866737365723, "learning_rate": 7.587954301912216e-05, "loss": 0.0109, "step": 28840 }, { "grad_norm": 0.10898991674184799, "learning_rate": 7.586185636500263e-05, "loss": 0.0096, "step": 28850 }, { "grad_norm": 0.12259998172521591, "learning_rate": 7.584416529179914e-05, "loss": 0.0091, "step": 28860 }, { "grad_norm": 0.17540515959262848, "learning_rate": 7.582646980253465e-05, "loss": 0.0102, "step": 28870 }, { "grad_norm": 0.1619962453842163, "learning_rate": 7.580876990023282e-05, "loss": 0.0096, "step": 28880 }, { "grad_norm": 0.10979241877794266, "learning_rate": 7.579106558791809e-05, "loss": 0.0091, "step": 28890 }, { "grad_norm": 0.17702747881412506, "learning_rate": 7.577335686861565e-05, "loss": 0.0116, "step": 28900 }, { "grad_norm": 0.18343596160411835, "learning_rate": 7.575564374535141e-05, "loss": 0.0093, "step": 28910 }, { "grad_norm": 0.14628811180591583, "learning_rate": 7.573792622115207e-05, "loss": 0.0082, "step": 28920 }, { "grad_norm": 0.1272328644990921, "learning_rate": 7.572020429904507e-05, "loss": 0.0117, "step": 28930 }, { "grad_norm": 0.11523675173521042, "learning_rate": 7.570247798205861e-05, "loss": 0.0091, "step": 28940 }, { "grad_norm": 0.14304542541503906, "learning_rate": 7.568474727322164e-05, "loss": 0.0096, "step": 28950 }, { "grad_norm": 0.1584223359823227, "learning_rate": 7.566701217556384e-05, "loss": 0.0123, "step": 28960 }, { "grad_norm": 0.1295626312494278, "learning_rate": 7.564927269211564e-05, "loss": 0.011, "step": 28970 }, { "grad_norm": 0.14879296720027924, "learning_rate": 7.563152882590824e-05, "loss": 0.0111, "step": 28980 }, { "grad_norm": 0.14851494133472443, "learning_rate": 7.56137805799736e-05, "loss": 0.0116, "step": 28990 }, { "grad_norm": 0.17275261878967285, "learning_rate": 7.559602795734439e-05, "loss": 0.0115, "step": 29000 }, { "grad_norm": 0.12945134937763214, "learning_rate": 7.557827096105408e-05, "loss": 0.0104, "step": 29010 }, { "grad_norm": 0.11751731485128403, "learning_rate": 7.55605095941368e-05, "loss": 0.0099, "step": 29020 }, { "grad_norm": 0.16284359991550446, "learning_rate": 7.554274385962752e-05, "loss": 0.0097, "step": 29030 }, { "grad_norm": 0.17733153700828552, "learning_rate": 7.552497376056191e-05, "loss": 0.013, "step": 29040 }, { "grad_norm": 0.126778244972229, "learning_rate": 7.550719929997639e-05, "loss": 0.0073, "step": 29050 }, { "grad_norm": 0.08756773918867111, "learning_rate": 7.548942048090813e-05, "loss": 0.0102, "step": 29060 }, { "grad_norm": 0.14499050378799438, "learning_rate": 7.547163730639506e-05, "loss": 0.011, "step": 29070 }, { "grad_norm": 0.14725929498672485, "learning_rate": 7.545384977947583e-05, "loss": 0.0105, "step": 29080 }, { "grad_norm": 0.13037820160388947, "learning_rate": 7.543605790318981e-05, "loss": 0.0104, "step": 29090 }, { "grad_norm": 0.17268700897693634, "learning_rate": 7.54182616805772e-05, "loss": 0.011, "step": 29100 }, { "grad_norm": 0.1823243796825409, "learning_rate": 7.540046111467885e-05, "loss": 0.0106, "step": 29110 }, { "grad_norm": 0.18889674544334412, "learning_rate": 7.53826562085364e-05, "loss": 0.0106, "step": 29120 }, { "grad_norm": 0.1122208833694458, "learning_rate": 7.536484696519221e-05, "loss": 0.0114, "step": 29130 }, { "grad_norm": 0.10466967523097992, "learning_rate": 7.534703338768942e-05, "loss": 0.012, "step": 29140 }, { "grad_norm": 0.19452880322933197, "learning_rate": 7.532921547907185e-05, "loss": 0.012, "step": 29150 }, { "grad_norm": 0.16766582429409027, "learning_rate": 7.531139324238412e-05, "loss": 0.0117, "step": 29160 }, { "grad_norm": 0.11370737850666046, "learning_rate": 7.529356668067157e-05, "loss": 0.0115, "step": 29170 }, { "grad_norm": 0.17846465110778809, "learning_rate": 7.527573579698023e-05, "loss": 0.0106, "step": 29180 }, { "grad_norm": 0.17656078934669495, "learning_rate": 7.525790059435693e-05, "loss": 0.0104, "step": 29190 }, { "grad_norm": 0.14021344482898712, "learning_rate": 7.524006107584926e-05, "loss": 0.0085, "step": 29200 }, { "grad_norm": 0.1296965628862381, "learning_rate": 7.522221724450544e-05, "loss": 0.0099, "step": 29210 }, { "grad_norm": 0.14506015181541443, "learning_rate": 7.520436910337451e-05, "loss": 0.0101, "step": 29220 }, { "grad_norm": 0.23167282342910767, "learning_rate": 7.518651665550627e-05, "loss": 0.0115, "step": 29230 }, { "grad_norm": 0.17564813792705536, "learning_rate": 7.516865990395117e-05, "loss": 0.0133, "step": 29240 }, { "grad_norm": 0.1432042121887207, "learning_rate": 7.515079885176047e-05, "loss": 0.0111, "step": 29250 }, { "grad_norm": 0.11477933824062347, "learning_rate": 7.513293350198612e-05, "loss": 0.0093, "step": 29260 }, { "grad_norm": 0.1344379037618637, "learning_rate": 7.511506385768081e-05, "loss": 0.0098, "step": 29270 }, { "grad_norm": 0.1428169161081314, "learning_rate": 7.509718992189801e-05, "loss": 0.0101, "step": 29280 }, { "grad_norm": 0.15387295186519623, "learning_rate": 7.507931169769182e-05, "loss": 0.0121, "step": 29290 }, { "grad_norm": 0.14172933995723724, "learning_rate": 7.506142918811722e-05, "loss": 0.0116, "step": 29300 }, { "grad_norm": 0.16687387228012085, "learning_rate": 7.504354239622978e-05, "loss": 0.0105, "step": 29310 }, { "grad_norm": 0.16117511689662933, "learning_rate": 7.50256513250859e-05, "loss": 0.0104, "step": 29320 }, { "grad_norm": 0.14509141445159912, "learning_rate": 7.500775597774265e-05, "loss": 0.011, "step": 29330 }, { "grad_norm": 0.14113205671310425, "learning_rate": 7.498985635725788e-05, "loss": 0.009, "step": 29340 }, { "grad_norm": 0.2199842482805252, "learning_rate": 7.497195246669012e-05, "loss": 0.0113, "step": 29350 }, { "grad_norm": 0.20182906091213226, "learning_rate": 7.495404430909868e-05, "loss": 0.0109, "step": 29360 }, { "grad_norm": 0.14785239100456238, "learning_rate": 7.493613188754356e-05, "loss": 0.0105, "step": 29370 }, { "grad_norm": 0.1649293303489685, "learning_rate": 7.49182152050855e-05, "loss": 0.009, "step": 29380 }, { "grad_norm": 0.16730281710624695, "learning_rate": 7.490029426478598e-05, "loss": 0.0101, "step": 29390 }, { "grad_norm": 0.15834663808345795, "learning_rate": 7.488236906970719e-05, "loss": 0.01, "step": 29400 }, { "grad_norm": 0.1610899567604065, "learning_rate": 7.486443962291207e-05, "loss": 0.0088, "step": 29410 }, { "grad_norm": 0.14335542917251587, "learning_rate": 7.484650592746424e-05, "loss": 0.0093, "step": 29420 }, { "grad_norm": 0.13425076007843018, "learning_rate": 7.482856798642811e-05, "loss": 0.0113, "step": 29430 }, { "grad_norm": 0.13479384779930115, "learning_rate": 7.481062580286878e-05, "loss": 0.0095, "step": 29440 }, { "grad_norm": 0.14730823040008545, "learning_rate": 7.479267937985208e-05, "loss": 0.0101, "step": 29450 }, { "grad_norm": 0.22315920889377594, "learning_rate": 7.477472872044456e-05, "loss": 0.0115, "step": 29460 }, { "grad_norm": 0.15438759326934814, "learning_rate": 7.475677382771347e-05, "loss": 0.0089, "step": 29470 }, { "grad_norm": 0.1688223034143448, "learning_rate": 7.473881470472683e-05, "loss": 0.0106, "step": 29480 }, { "grad_norm": 0.14191922545433044, "learning_rate": 7.472085135455337e-05, "loss": 0.0089, "step": 29490 }, { "grad_norm": 0.1563543677330017, "learning_rate": 7.470288378026256e-05, "loss": 0.011, "step": 29500 }, { "grad_norm": 0.11685585230588913, "learning_rate": 7.468491198492451e-05, "loss": 0.0119, "step": 29510 }, { "grad_norm": 0.16225665807724, "learning_rate": 7.466693597161013e-05, "loss": 0.0102, "step": 29520 }, { "grad_norm": 0.14148592948913574, "learning_rate": 7.464895574339104e-05, "loss": 0.0085, "step": 29530 }, { "grad_norm": 0.14728038012981415, "learning_rate": 7.463097130333958e-05, "loss": 0.0129, "step": 29540 }, { "grad_norm": 0.14225471019744873, "learning_rate": 7.461298265452876e-05, "loss": 0.0095, "step": 29550 }, { "grad_norm": 0.19644801318645477, "learning_rate": 7.459498980003239e-05, "loss": 0.0115, "step": 29560 }, { "grad_norm": 0.18170085549354553, "learning_rate": 7.457699274292493e-05, "loss": 0.0099, "step": 29570 }, { "grad_norm": 0.12512698769569397, "learning_rate": 7.455899148628159e-05, "loss": 0.0097, "step": 29580 }, { "grad_norm": 0.14073824882507324, "learning_rate": 7.45409860331783e-05, "loss": 0.0101, "step": 29590 }, { "grad_norm": 0.16548515856266022, "learning_rate": 7.452297638669169e-05, "loss": 0.0113, "step": 29600 }, { "grad_norm": 0.1627236008644104, "learning_rate": 7.450496254989911e-05, "loss": 0.013, "step": 29610 }, { "grad_norm": 0.13127613067626953, "learning_rate": 7.448694452587866e-05, "loss": 0.0108, "step": 29620 }, { "grad_norm": 0.15223491191864014, "learning_rate": 7.44689223177091e-05, "loss": 0.0094, "step": 29630 }, { "grad_norm": 0.2135103940963745, "learning_rate": 7.445089592846994e-05, "loss": 0.0099, "step": 29640 }, { "grad_norm": 0.11572836339473724, "learning_rate": 7.443286536124141e-05, "loss": 0.009, "step": 29650 }, { "grad_norm": 0.1628730446100235, "learning_rate": 7.441483061910443e-05, "loss": 0.0089, "step": 29660 }, { "grad_norm": 0.11625181883573532, "learning_rate": 7.439679170514064e-05, "loss": 0.008, "step": 29670 }, { "grad_norm": 0.1613348126411438, "learning_rate": 7.43787486224324e-05, "loss": 0.009, "step": 29680 }, { "grad_norm": 0.1332802027463913, "learning_rate": 7.436070137406276e-05, "loss": 0.0097, "step": 29690 }, { "grad_norm": 0.20322133600711823, "learning_rate": 7.434264996311556e-05, "loss": 0.0116, "step": 29700 }, { "grad_norm": 0.17923122644424438, "learning_rate": 7.432459439267525e-05, "loss": 0.0101, "step": 29710 }, { "grad_norm": 0.13570669293403625, "learning_rate": 7.430653466582701e-05, "loss": 0.0095, "step": 29720 }, { "grad_norm": 0.15773287415504456, "learning_rate": 7.42884707856568e-05, "loss": 0.0131, "step": 29730 }, { "grad_norm": 0.1236308142542839, "learning_rate": 7.427040275525122e-05, "loss": 0.0082, "step": 29740 }, { "grad_norm": 0.1302952617406845, "learning_rate": 7.42523305776976e-05, "loss": 0.0115, "step": 29750 }, { "grad_norm": 0.12797918915748596, "learning_rate": 7.4234254256084e-05, "loss": 0.0104, "step": 29760 }, { "grad_norm": 0.17832361161708832, "learning_rate": 7.421617379349915e-05, "loss": 0.0121, "step": 29770 }, { "grad_norm": 0.1745089739561081, "learning_rate": 7.41980891930325e-05, "loss": 0.0106, "step": 29780 }, { "grad_norm": 0.19139212369918823, "learning_rate": 7.418000045777425e-05, "loss": 0.0106, "step": 29790 }, { "grad_norm": 0.18031847476959229, "learning_rate": 7.416190759081523e-05, "loss": 0.0139, "step": 29800 }, { "grad_norm": 0.14026859402656555, "learning_rate": 7.414381059524704e-05, "loss": 0.0087, "step": 29810 }, { "grad_norm": 0.14381520450115204, "learning_rate": 7.412570947416195e-05, "loss": 0.0092, "step": 29820 }, { "grad_norm": 0.15528574585914612, "learning_rate": 7.410760423065295e-05, "loss": 0.0107, "step": 29830 }, { "grad_norm": 0.19785402715206146, "learning_rate": 7.408949486781372e-05, "loss": 0.0096, "step": 29840 }, { "grad_norm": 0.15413157641887665, "learning_rate": 7.407138138873868e-05, "loss": 0.0117, "step": 29850 }, { "grad_norm": 0.17549645900726318, "learning_rate": 7.405326379652292e-05, "loss": 0.0117, "step": 29860 }, { "grad_norm": 0.201041117310524, "learning_rate": 7.403514209426222e-05, "loss": 0.0118, "step": 29870 }, { "grad_norm": 0.1503455638885498, "learning_rate": 7.40170162850531e-05, "loss": 0.0127, "step": 29880 }, { "grad_norm": 0.15775546431541443, "learning_rate": 7.399888637199278e-05, "loss": 0.0131, "step": 29890 }, { "grad_norm": 0.16897734999656677, "learning_rate": 7.398075235817914e-05, "loss": 0.0107, "step": 29900 }, { "grad_norm": 0.18893200159072876, "learning_rate": 7.39626142467108e-05, "loss": 0.0123, "step": 29910 }, { "grad_norm": 0.16647064685821533, "learning_rate": 7.394447204068706e-05, "loss": 0.0108, "step": 29920 }, { "grad_norm": 0.1866833120584488, "learning_rate": 7.392632574320793e-05, "loss": 0.0145, "step": 29930 }, { "grad_norm": 0.0901191383600235, "learning_rate": 7.390817535737411e-05, "loss": 0.0119, "step": 29940 }, { "grad_norm": 0.1717037558555603, "learning_rate": 7.389002088628703e-05, "loss": 0.0116, "step": 29950 }, { "grad_norm": 0.16254913806915283, "learning_rate": 7.387186233304877e-05, "loss": 0.0097, "step": 29960 }, { "grad_norm": 0.15537108480930328, "learning_rate": 7.385369970076212e-05, "loss": 0.0091, "step": 29970 }, { "grad_norm": 0.12358340620994568, "learning_rate": 7.38355329925306e-05, "loss": 0.0076, "step": 29980 }, { "grad_norm": 0.14623405039310455, "learning_rate": 7.381736221145838e-05, "loss": 0.0087, "step": 29990 }, { "grad_norm": 0.1277451068162918, "learning_rate": 7.37991873606504e-05, "loss": 0.0105, "step": 30000 }, { "grad_norm": 0.12565921247005463, "learning_rate": 7.378100844321218e-05, "loss": 0.0092, "step": 30010 }, { "grad_norm": 0.14537329971790314, "learning_rate": 7.376282546225004e-05, "loss": 0.0134, "step": 30020 }, { "grad_norm": 0.16036608815193176, "learning_rate": 7.374463842087094e-05, "loss": 0.0114, "step": 30030 }, { "grad_norm": 0.08866382390260696, "learning_rate": 7.372644732218254e-05, "loss": 0.0084, "step": 30040 }, { "grad_norm": 0.1371430903673172, "learning_rate": 7.370825216929322e-05, "loss": 0.0123, "step": 30050 }, { "grad_norm": 0.13088330626487732, "learning_rate": 7.369005296531205e-05, "loss": 0.0117, "step": 30060 }, { "grad_norm": 0.1441064178943634, "learning_rate": 7.367184971334873e-05, "loss": 0.0093, "step": 30070 }, { "grad_norm": 0.10943841934204102, "learning_rate": 7.365364241651371e-05, "loss": 0.0096, "step": 30080 }, { "grad_norm": 0.13393428921699524, "learning_rate": 7.363543107791815e-05, "loss": 0.0108, "step": 30090 }, { "grad_norm": 0.18410739302635193, "learning_rate": 7.361721570067384e-05, "loss": 0.0109, "step": 30100 }, { "grad_norm": 0.1364559531211853, "learning_rate": 7.359899628789331e-05, "loss": 0.011, "step": 30110 }, { "grad_norm": 0.11982589215040207, "learning_rate": 7.358077284268974e-05, "loss": 0.0107, "step": 30120 }, { "grad_norm": 0.16873802244663239, "learning_rate": 7.356254536817702e-05, "loss": 0.0108, "step": 30130 }, { "grad_norm": 0.16629016399383545, "learning_rate": 7.354431386746973e-05, "loss": 0.0135, "step": 30140 }, { "grad_norm": 0.1897154003381729, "learning_rate": 7.352607834368316e-05, "loss": 0.0116, "step": 30150 }, { "grad_norm": 0.1825110763311386, "learning_rate": 7.350783879993324e-05, "loss": 0.0114, "step": 30160 }, { "grad_norm": 0.1156729906797409, "learning_rate": 7.348959523933658e-05, "loss": 0.0126, "step": 30170 }, { "grad_norm": 0.13616818189620972, "learning_rate": 7.347134766501057e-05, "loss": 0.0105, "step": 30180 }, { "grad_norm": 0.14632733166217804, "learning_rate": 7.345309608007315e-05, "loss": 0.011, "step": 30190 }, { "grad_norm": 0.14156578481197357, "learning_rate": 7.343484048764308e-05, "loss": 0.0118, "step": 30200 }, { "grad_norm": 0.18443678319454193, "learning_rate": 7.341658089083972e-05, "loss": 0.0113, "step": 30210 }, { "grad_norm": 0.17565330862998962, "learning_rate": 7.339831729278313e-05, "loss": 0.0137, "step": 30220 }, { "grad_norm": 0.18699219822883606, "learning_rate": 7.338004969659404e-05, "loss": 0.0094, "step": 30230 }, { "grad_norm": 0.1402943730354309, "learning_rate": 7.336177810539391e-05, "loss": 0.0107, "step": 30240 }, { "grad_norm": 0.17186321318149567, "learning_rate": 7.334350252230485e-05, "loss": 0.014, "step": 30250 }, { "grad_norm": 0.12433134019374847, "learning_rate": 7.332522295044965e-05, "loss": 0.0091, "step": 30260 }, { "grad_norm": 0.15119251608848572, "learning_rate": 7.33069393929518e-05, "loss": 0.0117, "step": 30270 }, { "grad_norm": 0.18598537147045135, "learning_rate": 7.328865185293545e-05, "loss": 0.0128, "step": 30280 }, { "grad_norm": 0.18645772337913513, "learning_rate": 7.327036033352546e-05, "loss": 0.0112, "step": 30290 }, { "grad_norm": 0.1788434386253357, "learning_rate": 7.325206483784733e-05, "loss": 0.0104, "step": 30300 }, { "grad_norm": 0.14278492331504822, "learning_rate": 7.323376536902724e-05, "loss": 0.0092, "step": 30310 }, { "grad_norm": 0.17303550243377686, "learning_rate": 7.321546193019213e-05, "loss": 0.0117, "step": 30320 }, { "grad_norm": 0.1473180204629898, "learning_rate": 7.31971545244695e-05, "loss": 0.0106, "step": 30330 }, { "grad_norm": 0.1586506962776184, "learning_rate": 7.31788431549876e-05, "loss": 0.0118, "step": 30340 }, { "grad_norm": 0.14793168008327484, "learning_rate": 7.316052782487534e-05, "loss": 0.0113, "step": 30350 }, { "grad_norm": 0.21149399876594543, "learning_rate": 7.314220853726234e-05, "loss": 0.0116, "step": 30360 }, { "grad_norm": 0.1661619246006012, "learning_rate": 7.312388529527884e-05, "loss": 0.0115, "step": 30370 }, { "grad_norm": 0.1742565929889679, "learning_rate": 7.310555810205577e-05, "loss": 0.0095, "step": 30380 }, { "grad_norm": 0.1493779420852661, "learning_rate": 7.308722696072476e-05, "loss": 0.0106, "step": 30390 }, { "grad_norm": 0.13984136283397675, "learning_rate": 7.306889187441811e-05, "loss": 0.0091, "step": 30400 }, { "grad_norm": 0.11740963906049728, "learning_rate": 7.305055284626876e-05, "loss": 0.0099, "step": 30410 }, { "grad_norm": 0.10805866122245789, "learning_rate": 7.303220987941037e-05, "loss": 0.0095, "step": 30420 }, { "grad_norm": 0.10786322504281998, "learning_rate": 7.301386297697726e-05, "loss": 0.0074, "step": 30430 }, { "grad_norm": 0.133237823843956, "learning_rate": 7.299551214210438e-05, "loss": 0.0082, "step": 30440 }, { "grad_norm": 0.22217662632465363, "learning_rate": 7.297715737792738e-05, "loss": 0.009, "step": 30450 }, { "grad_norm": 0.11256197094917297, "learning_rate": 7.295879868758265e-05, "loss": 0.009, "step": 30460 }, { "grad_norm": 0.16630521416664124, "learning_rate": 7.294043607420713e-05, "loss": 0.0101, "step": 30470 }, { "grad_norm": 0.12334199249744415, "learning_rate": 7.292206954093852e-05, "loss": 0.0088, "step": 30480 }, { "grad_norm": 0.18702106177806854, "learning_rate": 7.290369909091515e-05, "loss": 0.0107, "step": 30490 }, { "grad_norm": 0.19171644747257233, "learning_rate": 7.2885324727276e-05, "loss": 0.0107, "step": 30500 }, { "grad_norm": 0.13670355081558228, "learning_rate": 7.286694645316076e-05, "loss": 0.0101, "step": 30510 }, { "grad_norm": 0.13328011333942413, "learning_rate": 7.284856427170982e-05, "loss": 0.0103, "step": 30520 }, { "grad_norm": 0.16725926101207733, "learning_rate": 7.283017818606414e-05, "loss": 0.0109, "step": 30530 }, { "grad_norm": 0.11337985843420029, "learning_rate": 7.28117881993654e-05, "loss": 0.0094, "step": 30540 }, { "grad_norm": 0.1551070362329483, "learning_rate": 7.279339431475598e-05, "loss": 0.0102, "step": 30550 }, { "grad_norm": 0.1322881430387497, "learning_rate": 7.277499653537887e-05, "loss": 0.0127, "step": 30560 }, { "grad_norm": 0.13647599518299103, "learning_rate": 7.275659486437776e-05, "loss": 0.0139, "step": 30570 }, { "grad_norm": 0.11713307350873947, "learning_rate": 7.273818930489695e-05, "loss": 0.0117, "step": 30580 }, { "grad_norm": 0.14263521134853363, "learning_rate": 7.271977986008151e-05, "loss": 0.0088, "step": 30590 }, { "grad_norm": 0.1725665181875229, "learning_rate": 7.270136653307705e-05, "loss": 0.01, "step": 30600 }, { "grad_norm": 0.13386334478855133, "learning_rate": 7.268294932702994e-05, "loss": 0.0108, "step": 30610 }, { "grad_norm": 0.14801770448684692, "learning_rate": 7.266452824508719e-05, "loss": 0.0078, "step": 30620 }, { "grad_norm": 0.16794046759605408, "learning_rate": 7.264610329039643e-05, "loss": 0.0114, "step": 30630 }, { "grad_norm": 0.14683954417705536, "learning_rate": 7.262767446610599e-05, "loss": 0.0113, "step": 30640 }, { "grad_norm": 0.1294189840555191, "learning_rate": 7.260924177536485e-05, "loss": 0.0093, "step": 30650 }, { "grad_norm": 0.14496926963329315, "learning_rate": 7.259080522132265e-05, "loss": 0.0118, "step": 30660 }, { "grad_norm": 0.14289434254169464, "learning_rate": 7.257236480712972e-05, "loss": 0.0091, "step": 30670 }, { "grad_norm": 0.14760743081569672, "learning_rate": 7.255392053593697e-05, "loss": 0.0093, "step": 30680 }, { "grad_norm": 0.16943782567977905, "learning_rate": 7.253547241089607e-05, "loss": 0.0091, "step": 30690 }, { "grad_norm": 0.18273775279521942, "learning_rate": 7.251702043515927e-05, "loss": 0.0112, "step": 30700 }, { "grad_norm": 0.13185912370681763, "learning_rate": 7.249856461187952e-05, "loss": 0.0102, "step": 30710 }, { "grad_norm": 0.10670153051614761, "learning_rate": 7.248010494421042e-05, "loss": 0.0104, "step": 30720 }, { "grad_norm": 0.1475757658481598, "learning_rate": 7.246164143530622e-05, "loss": 0.0121, "step": 30730 }, { "grad_norm": 0.1136159896850586, "learning_rate": 7.244317408832181e-05, "loss": 0.0105, "step": 30740 }, { "grad_norm": 0.1544303148984909, "learning_rate": 7.242470290641279e-05, "loss": 0.0086, "step": 30750 }, { "grad_norm": 0.1048705130815506, "learning_rate": 7.240622789273536e-05, "loss": 0.0104, "step": 30760 }, { "grad_norm": 0.1522100567817688, "learning_rate": 7.238774905044638e-05, "loss": 0.0102, "step": 30770 }, { "grad_norm": 0.1978926807641983, "learning_rate": 7.236926638270341e-05, "loss": 0.0125, "step": 30780 }, { "grad_norm": 0.22705651819705963, "learning_rate": 7.23507798926646e-05, "loss": 0.0092, "step": 30790 }, { "grad_norm": 0.17303578555583954, "learning_rate": 7.23322895834888e-05, "loss": 0.0117, "step": 30800 }, { "grad_norm": 0.144395112991333, "learning_rate": 7.231379545833552e-05, "loss": 0.011, "step": 30810 }, { "grad_norm": 0.17628978192806244, "learning_rate": 7.229529752036487e-05, "loss": 0.0127, "step": 30820 }, { "grad_norm": 0.1320701241493225, "learning_rate": 7.227679577273765e-05, "loss": 0.0109, "step": 30830 }, { "grad_norm": 0.14657846093177795, "learning_rate": 7.225829021861529e-05, "loss": 0.0081, "step": 30840 }, { "grad_norm": 0.18837076425552368, "learning_rate": 7.223978086115992e-05, "loss": 0.0099, "step": 30850 }, { "grad_norm": 0.13259832561016083, "learning_rate": 7.222126770353425e-05, "loss": 0.0115, "step": 30860 }, { "grad_norm": 0.1555519700050354, "learning_rate": 7.22027507489017e-05, "loss": 0.0134, "step": 30870 }, { "grad_norm": 0.16250848770141602, "learning_rate": 7.218423000042627e-05, "loss": 0.0111, "step": 30880 }, { "grad_norm": 0.19182941317558289, "learning_rate": 7.216570546127268e-05, "loss": 0.0091, "step": 30890 }, { "grad_norm": 0.15304163098335266, "learning_rate": 7.214717713460626e-05, "loss": 0.011, "step": 30900 }, { "grad_norm": 0.14144855737686157, "learning_rate": 7.2128645023593e-05, "loss": 0.0092, "step": 30910 }, { "grad_norm": 0.14043405652046204, "learning_rate": 7.211010913139951e-05, "loss": 0.0101, "step": 30920 }, { "grad_norm": 0.10314509272575378, "learning_rate": 7.209156946119308e-05, "loss": 0.0099, "step": 30930 }, { "grad_norm": 0.15788543224334717, "learning_rate": 7.207302601614166e-05, "loss": 0.0087, "step": 30940 }, { "grad_norm": 0.11211904138326645, "learning_rate": 7.205447879941378e-05, "loss": 0.0087, "step": 30950 }, { "grad_norm": 0.16649805009365082, "learning_rate": 7.203592781417866e-05, "loss": 0.0108, "step": 30960 }, { "grad_norm": 0.14092005789279938, "learning_rate": 7.201737306360617e-05, "loss": 0.0095, "step": 30970 }, { "grad_norm": 0.13089053332805634, "learning_rate": 7.19988145508668e-05, "loss": 0.0097, "step": 30980 }, { "grad_norm": 0.11707255989313126, "learning_rate": 7.198025227913168e-05, "loss": 0.0113, "step": 30990 }, { "grad_norm": 0.15138043463230133, "learning_rate": 7.196168625157261e-05, "loss": 0.0087, "step": 31000 }, { "grad_norm": 0.14747492969036102, "learning_rate": 7.194311647136201e-05, "loss": 0.0105, "step": 31010 }, { "grad_norm": 0.16362348198890686, "learning_rate": 7.192454294167297e-05, "loss": 0.0104, "step": 31020 }, { "grad_norm": 0.16385313868522644, "learning_rate": 7.190596566567917e-05, "loss": 0.0103, "step": 31030 }, { "grad_norm": 0.17285649478435516, "learning_rate": 7.188738464655496e-05, "loss": 0.0122, "step": 31040 }, { "grad_norm": 0.13051837682724, "learning_rate": 7.186879988747533e-05, "loss": 0.0098, "step": 31050 }, { "grad_norm": 0.15533824265003204, "learning_rate": 7.185021139161592e-05, "loss": 0.0091, "step": 31060 }, { "grad_norm": 0.11842969805002213, "learning_rate": 7.1831619162153e-05, "loss": 0.0096, "step": 31070 }, { "grad_norm": 0.13922452926635742, "learning_rate": 7.181302320226345e-05, "loss": 0.0117, "step": 31080 }, { "grad_norm": 0.12315239757299423, "learning_rate": 7.179442351512482e-05, "loss": 0.0095, "step": 31090 }, { "grad_norm": 0.1536729633808136, "learning_rate": 7.177582010391528e-05, "loss": 0.0108, "step": 31100 }, { "grad_norm": 0.18693913519382477, "learning_rate": 7.175721297181366e-05, "loss": 0.01, "step": 31110 }, { "grad_norm": 0.20030181109905243, "learning_rate": 7.173860212199942e-05, "loss": 0.0115, "step": 31120 }, { "grad_norm": 0.14858128130435944, "learning_rate": 7.171998755765263e-05, "loss": 0.0103, "step": 31130 }, { "grad_norm": 0.17263492941856384, "learning_rate": 7.170136928195398e-05, "loss": 0.0109, "step": 31140 }, { "grad_norm": 0.18125160038471222, "learning_rate": 7.168274729808489e-05, "loss": 0.0103, "step": 31150 }, { "grad_norm": 0.17417363822460175, "learning_rate": 7.166412160922728e-05, "loss": 0.0115, "step": 31160 }, { "grad_norm": 0.13820818066596985, "learning_rate": 7.164549221856382e-05, "loss": 0.0099, "step": 31170 }, { "grad_norm": 0.18450503051280975, "learning_rate": 7.162685912927775e-05, "loss": 0.0099, "step": 31180 }, { "grad_norm": 0.15058204531669617, "learning_rate": 7.160822234455294e-05, "loss": 0.0099, "step": 31190 }, { "grad_norm": 0.15802794694900513, "learning_rate": 7.158958186757391e-05, "loss": 0.0086, "step": 31200 }, { "grad_norm": 0.16500195860862732, "learning_rate": 7.157093770152582e-05, "loss": 0.0103, "step": 31210 }, { "grad_norm": 0.15006402134895325, "learning_rate": 7.155228984959446e-05, "loss": 0.0125, "step": 31220 }, { "grad_norm": 0.13940125703811646, "learning_rate": 7.153363831496621e-05, "loss": 0.0113, "step": 31230 }, { "grad_norm": 0.10453125834465027, "learning_rate": 7.151498310082811e-05, "loss": 0.0114, "step": 31240 }, { "grad_norm": 0.14427459239959717, "learning_rate": 7.149632421036784e-05, "loss": 0.0095, "step": 31250 }, { "grad_norm": 0.19118204712867737, "learning_rate": 7.147766164677369e-05, "loss": 0.0093, "step": 31260 }, { "grad_norm": 0.13059747219085693, "learning_rate": 7.145899541323459e-05, "loss": 0.0101, "step": 31270 }, { "grad_norm": 0.1407090276479721, "learning_rate": 7.144032551294007e-05, "loss": 0.0121, "step": 31280 }, { "grad_norm": 0.16883482038974762, "learning_rate": 7.14216519490803e-05, "loss": 0.0124, "step": 31290 }, { "grad_norm": 0.20829661190509796, "learning_rate": 7.140297472484609e-05, "loss": 0.0129, "step": 31300 }, { "grad_norm": 0.12047181278467178, "learning_rate": 7.138429384342891e-05, "loss": 0.0087, "step": 31310 }, { "grad_norm": 0.16294720768928528, "learning_rate": 7.136560930802074e-05, "loss": 0.0095, "step": 31320 }, { "grad_norm": 0.16270968317985535, "learning_rate": 7.134692112181431e-05, "loss": 0.0111, "step": 31330 }, { "grad_norm": 0.17890992760658264, "learning_rate": 7.13282292880029e-05, "loss": 0.0107, "step": 31340 }, { "grad_norm": 0.15168659389019012, "learning_rate": 7.130953380978043e-05, "loss": 0.0118, "step": 31350 }, { "grad_norm": 0.12374977767467499, "learning_rate": 7.129083469034144e-05, "loss": 0.0113, "step": 31360 }, { "grad_norm": 0.148330956697464, "learning_rate": 7.127213193288112e-05, "loss": 0.0115, "step": 31370 }, { "grad_norm": 0.13662318885326385, "learning_rate": 7.125342554059522e-05, "loss": 0.011, "step": 31380 }, { "grad_norm": 0.2154042273759842, "learning_rate": 7.12347155166802e-05, "loss": 0.0139, "step": 31390 }, { "grad_norm": 0.12005968391895294, "learning_rate": 7.121600186433306e-05, "loss": 0.0107, "step": 31400 }, { "grad_norm": 0.18649183213710785, "learning_rate": 7.119728458675148e-05, "loss": 0.0122, "step": 31410 }, { "grad_norm": 0.20223459601402283, "learning_rate": 7.117856368713369e-05, "loss": 0.013, "step": 31420 }, { "grad_norm": 0.12005577981472015, "learning_rate": 7.115983916867861e-05, "loss": 0.0094, "step": 31430 }, { "grad_norm": 0.13120050728321075, "learning_rate": 7.114111103458574e-05, "loss": 0.0101, "step": 31440 }, { "grad_norm": 0.15954740345478058, "learning_rate": 7.11223792880552e-05, "loss": 0.0094, "step": 31450 }, { "grad_norm": 0.1523853838443756, "learning_rate": 7.110364393228773e-05, "loss": 0.0111, "step": 31460 }, { "grad_norm": 0.14956651628017426, "learning_rate": 7.108490497048471e-05, "loss": 0.0093, "step": 31470 }, { "grad_norm": 0.13401873409748077, "learning_rate": 7.10661624058481e-05, "loss": 0.0096, "step": 31480 }, { "grad_norm": 0.14329127967357635, "learning_rate": 7.10474162415805e-05, "loss": 0.0106, "step": 31490 }, { "grad_norm": 0.13975751399993896, "learning_rate": 7.102866648088511e-05, "loss": 0.0097, "step": 31500 }, { "grad_norm": 0.11940133571624756, "learning_rate": 7.100991312696576e-05, "loss": 0.0092, "step": 31510 }, { "grad_norm": 0.13104674220085144, "learning_rate": 7.099115618302686e-05, "loss": 0.0084, "step": 31520 }, { "grad_norm": 0.145146444439888, "learning_rate": 7.097239565227349e-05, "loss": 0.0095, "step": 31530 }, { "grad_norm": 0.13807594776153564, "learning_rate": 7.09536315379113e-05, "loss": 0.0111, "step": 31540 }, { "grad_norm": 0.14506353437900543, "learning_rate": 7.093486384314656e-05, "loss": 0.014, "step": 31550 }, { "grad_norm": 0.12137903273105621, "learning_rate": 7.091609257118616e-05, "loss": 0.0097, "step": 31560 }, { "grad_norm": 0.12922970950603485, "learning_rate": 7.08973177252376e-05, "loss": 0.0081, "step": 31570 }, { "grad_norm": 0.15587538480758667, "learning_rate": 7.087853930850898e-05, "loss": 0.0081, "step": 31580 }, { "grad_norm": 0.17842227220535278, "learning_rate": 7.085975732420903e-05, "loss": 0.0093, "step": 31590 }, { "grad_norm": 0.1921517550945282, "learning_rate": 7.084097177554706e-05, "loss": 0.0097, "step": 31600 }, { "grad_norm": 0.18549636006355286, "learning_rate": 7.082218266573301e-05, "loss": 0.0086, "step": 31610 }, { "grad_norm": 0.17941877245903015, "learning_rate": 7.080338999797743e-05, "loss": 0.0123, "step": 31620 }, { "grad_norm": 0.14374296367168427, "learning_rate": 7.07845937754915e-05, "loss": 0.0071, "step": 31630 }, { "grad_norm": 0.17043209075927734, "learning_rate": 7.076579400148693e-05, "loss": 0.0107, "step": 31640 }, { "grad_norm": 0.14315038919448853, "learning_rate": 7.074699067917611e-05, "loss": 0.0103, "step": 31650 }, { "grad_norm": 0.1518392264842987, "learning_rate": 7.072818381177201e-05, "loss": 0.0114, "step": 31660 }, { "grad_norm": 0.14043983817100525, "learning_rate": 7.070937340248823e-05, "loss": 0.0112, "step": 31670 }, { "grad_norm": 0.15783827006816864, "learning_rate": 7.069055945453893e-05, "loss": 0.01, "step": 31680 }, { "grad_norm": 0.12383440881967545, "learning_rate": 7.067174197113892e-05, "loss": 0.0089, "step": 31690 }, { "grad_norm": 0.1574370563030243, "learning_rate": 7.065292095550355e-05, "loss": 0.0111, "step": 31700 }, { "grad_norm": 0.16034762561321259, "learning_rate": 7.063409641084887e-05, "loss": 0.0096, "step": 31710 }, { "grad_norm": 0.14964407682418823, "learning_rate": 7.061526834039145e-05, "loss": 0.0092, "step": 31720 }, { "grad_norm": 0.19344744086265564, "learning_rate": 7.05964367473485e-05, "loss": 0.0101, "step": 31730 }, { "grad_norm": 0.13556736707687378, "learning_rate": 7.057760163493783e-05, "loss": 0.0105, "step": 31740 }, { "grad_norm": 0.1183771938085556, "learning_rate": 7.055876300637783e-05, "loss": 0.0111, "step": 31750 }, { "grad_norm": 0.1740601807832718, "learning_rate": 7.053992086488753e-05, "loss": 0.0098, "step": 31760 }, { "grad_norm": 0.1634240597486496, "learning_rate": 7.052107521368651e-05, "loss": 0.0114, "step": 31770 }, { "grad_norm": 0.13164366781711578, "learning_rate": 7.0502226055995e-05, "loss": 0.0097, "step": 31780 }, { "grad_norm": 0.17171615362167358, "learning_rate": 7.048337339503379e-05, "loss": 0.0101, "step": 31790 }, { "grad_norm": 0.20668639242649078, "learning_rate": 7.046451723402427e-05, "loss": 0.0106, "step": 31800 }, { "grad_norm": 0.13364148139953613, "learning_rate": 7.044565757618848e-05, "loss": 0.0114, "step": 31810 }, { "grad_norm": 0.13817158341407776, "learning_rate": 7.042679442474899e-05, "loss": 0.0092, "step": 31820 }, { "grad_norm": 0.10813794285058975, "learning_rate": 7.040792778292902e-05, "loss": 0.01, "step": 31830 }, { "grad_norm": 0.3001505732536316, "learning_rate": 7.038905765395234e-05, "loss": 0.0097, "step": 31840 }, { "grad_norm": 0.23946642875671387, "learning_rate": 7.037018404104334e-05, "loss": 0.0096, "step": 31850 }, { "grad_norm": 0.23920533061027527, "learning_rate": 7.035130694742702e-05, "loss": 0.0106, "step": 31860 }, { "grad_norm": 0.17499913275241852, "learning_rate": 7.033242637632897e-05, "loss": 0.0107, "step": 31870 }, { "grad_norm": 0.1320905238389969, "learning_rate": 7.031354233097534e-05, "loss": 0.0093, "step": 31880 }, { "grad_norm": 0.11577470600605011, "learning_rate": 7.029465481459289e-05, "loss": 0.0081, "step": 31890 }, { "grad_norm": 0.1332392543554306, "learning_rate": 7.027576383040898e-05, "loss": 0.0094, "step": 31900 }, { "grad_norm": 0.15073280036449432, "learning_rate": 7.025686938165159e-05, "loss": 0.0107, "step": 31910 }, { "grad_norm": 0.20049282908439636, "learning_rate": 7.023797147154924e-05, "loss": 0.0098, "step": 31920 }, { "grad_norm": 0.14544354379177094, "learning_rate": 7.021907010333111e-05, "loss": 0.0091, "step": 31930 }, { "grad_norm": 0.13671119511127472, "learning_rate": 7.020016528022685e-05, "loss": 0.0125, "step": 31940 }, { "grad_norm": 0.1628483682870865, "learning_rate": 7.018125700546683e-05, "loss": 0.01, "step": 31950 }, { "grad_norm": 0.17767685651779175, "learning_rate": 7.016234528228196e-05, "loss": 0.0141, "step": 31960 }, { "grad_norm": 0.12076114118099213, "learning_rate": 7.014343011390372e-05, "loss": 0.009, "step": 31970 }, { "grad_norm": 0.12921661138534546, "learning_rate": 7.01245115035642e-05, "loss": 0.0118, "step": 31980 }, { "grad_norm": 0.11733341217041016, "learning_rate": 7.010558945449606e-05, "loss": 0.0084, "step": 31990 }, { "grad_norm": 0.1831863671541214, "learning_rate": 7.008666396993258e-05, "loss": 0.0091, "step": 32000 }, { "grad_norm": 0.12241371721029282, "learning_rate": 7.006773505310759e-05, "loss": 0.0122, "step": 32010 }, { "grad_norm": 0.12407717853784561, "learning_rate": 7.004880270725553e-05, "loss": 0.0092, "step": 32020 }, { "grad_norm": 0.11992668360471725, "learning_rate": 7.002986693561144e-05, "loss": 0.0092, "step": 32030 }, { "grad_norm": 0.14292971789836884, "learning_rate": 7.001092774141089e-05, "loss": 0.0085, "step": 32040 }, { "grad_norm": 0.13844351470470428, "learning_rate": 6.999198512789009e-05, "loss": 0.0087, "step": 32050 }, { "grad_norm": 0.19111111760139465, "learning_rate": 6.997303909828584e-05, "loss": 0.0108, "step": 32060 }, { "grad_norm": 0.13292518258094788, "learning_rate": 6.995408965583544e-05, "loss": 0.0078, "step": 32070 }, { "grad_norm": 0.1990731954574585, "learning_rate": 6.993513680377688e-05, "loss": 0.0115, "step": 32080 }, { "grad_norm": 0.14693614840507507, "learning_rate": 6.991618054534868e-05, "loss": 0.0113, "step": 32090 }, { "grad_norm": 0.15851950645446777, "learning_rate": 6.989722088378991e-05, "loss": 0.0088, "step": 32100 }, { "grad_norm": 0.09046100825071335, "learning_rate": 6.987825782234027e-05, "loss": 0.0088, "step": 32110 }, { "grad_norm": 0.14453928172588348, "learning_rate": 6.985929136424006e-05, "loss": 0.01, "step": 32120 }, { "grad_norm": 0.15067066252231598, "learning_rate": 6.984032151273012e-05, "loss": 0.0098, "step": 32130 }, { "grad_norm": 0.10048504918813705, "learning_rate": 6.982134827105186e-05, "loss": 0.0101, "step": 32140 }, { "grad_norm": 0.14178259670734406, "learning_rate": 6.980237164244729e-05, "loss": 0.0113, "step": 32150 }, { "grad_norm": 0.1944541186094284, "learning_rate": 6.9783391630159e-05, "loss": 0.0089, "step": 32160 }, { "grad_norm": 0.1962420493364334, "learning_rate": 6.976440823743015e-05, "loss": 0.0093, "step": 32170 }, { "grad_norm": 0.17284433543682098, "learning_rate": 6.974542146750451e-05, "loss": 0.0097, "step": 32180 }, { "grad_norm": 0.12876221537590027, "learning_rate": 6.972643132362637e-05, "loss": 0.0082, "step": 32190 }, { "grad_norm": 0.13045623898506165, "learning_rate": 6.970743780904064e-05, "loss": 0.0086, "step": 32200 }, { "grad_norm": 0.15806859731674194, "learning_rate": 6.968844092699277e-05, "loss": 0.0135, "step": 32210 }, { "grad_norm": 0.1732417494058609, "learning_rate": 6.966944068072883e-05, "loss": 0.0102, "step": 32220 }, { "grad_norm": 0.19997906684875488, "learning_rate": 6.965043707349545e-05, "loss": 0.0101, "step": 32230 }, { "grad_norm": 0.22364577651023865, "learning_rate": 6.963143010853982e-05, "loss": 0.0101, "step": 32240 }, { "grad_norm": 0.1477157026529312, "learning_rate": 6.961241978910971e-05, "loss": 0.0083, "step": 32250 }, { "grad_norm": 0.12908363342285156, "learning_rate": 6.959340611845344e-05, "loss": 0.0126, "step": 32260 }, { "grad_norm": 0.11618319898843765, "learning_rate": 6.957438909981995e-05, "loss": 0.0094, "step": 32270 }, { "grad_norm": 0.16328325867652893, "learning_rate": 6.955536873645872e-05, "loss": 0.0107, "step": 32280 }, { "grad_norm": 0.1531083881855011, "learning_rate": 6.953634503161982e-05, "loss": 0.0097, "step": 32290 }, { "grad_norm": 0.12341529875993729, "learning_rate": 6.951731798855387e-05, "loss": 0.01, "step": 32300 }, { "grad_norm": 0.16203755140304565, "learning_rate": 6.949828761051208e-05, "loss": 0.0083, "step": 32310 }, { "grad_norm": 0.09828832745552063, "learning_rate": 6.947925390074622e-05, "loss": 0.0066, "step": 32320 }, { "grad_norm": 0.13025003671646118, "learning_rate": 6.946021686250863e-05, "loss": 0.0096, "step": 32330 }, { "grad_norm": 0.13497038185596466, "learning_rate": 6.94411764990522e-05, "loss": 0.0095, "step": 32340 }, { "grad_norm": 0.08465612679719925, "learning_rate": 6.942213281363044e-05, "loss": 0.009, "step": 32350 }, { "grad_norm": 0.11702406406402588, "learning_rate": 6.940308580949737e-05, "loss": 0.009, "step": 32360 }, { "grad_norm": 0.15027378499507904, "learning_rate": 6.93840354899076e-05, "loss": 0.0105, "step": 32370 }, { "grad_norm": 0.13778311014175415, "learning_rate": 6.936498185811633e-05, "loss": 0.0084, "step": 32380 }, { "grad_norm": 0.16544118523597717, "learning_rate": 6.93459249173793e-05, "loss": 0.012, "step": 32390 }, { "grad_norm": 0.17703242599964142, "learning_rate": 6.932686467095279e-05, "loss": 0.0102, "step": 32400 }, { "grad_norm": 0.15937796235084534, "learning_rate": 6.930780112209373e-05, "loss": 0.0134, "step": 32410 }, { "grad_norm": 0.15185073018074036, "learning_rate": 6.92887342740595e-05, "loss": 0.0106, "step": 32420 }, { "grad_norm": 0.12279563397169113, "learning_rate": 6.926966413010816e-05, "loss": 0.0095, "step": 32430 }, { "grad_norm": 0.12348916381597519, "learning_rate": 6.925059069349824e-05, "loss": 0.009, "step": 32440 }, { "grad_norm": 0.1531372368335724, "learning_rate": 6.923151396748886e-05, "loss": 0.0099, "step": 32450 }, { "grad_norm": 0.12327033281326294, "learning_rate": 6.921243395533974e-05, "loss": 0.011, "step": 32460 }, { "grad_norm": 0.15393616259098053, "learning_rate": 6.919335066031109e-05, "loss": 0.0116, "step": 32470 }, { "grad_norm": 0.13543309271335602, "learning_rate": 6.917426408566379e-05, "loss": 0.0099, "step": 32480 }, { "grad_norm": 0.15369349718093872, "learning_rate": 6.915517423465916e-05, "loss": 0.0089, "step": 32490 }, { "grad_norm": 0.1141183152794838, "learning_rate": 6.913608111055914e-05, "loss": 0.0085, "step": 32500 }, { "grad_norm": 0.16748811304569244, "learning_rate": 6.911698471662623e-05, "loss": 0.0093, "step": 32510 }, { "grad_norm": 0.2044326364994049, "learning_rate": 6.90978850561235e-05, "loss": 0.0098, "step": 32520 }, { "grad_norm": 0.13485588133335114, "learning_rate": 6.907878213231454e-05, "loss": 0.0081, "step": 32530 }, { "grad_norm": 0.20138351619243622, "learning_rate": 6.90596759484635e-05, "loss": 0.0152, "step": 32540 }, { "grad_norm": 0.13151989877223969, "learning_rate": 6.904056650783514e-05, "loss": 0.0094, "step": 32550 }, { "grad_norm": 0.14155234396457672, "learning_rate": 6.902145381369471e-05, "loss": 0.0086, "step": 32560 }, { "grad_norm": 0.17302285134792328, "learning_rate": 6.900233786930808e-05, "loss": 0.0102, "step": 32570 }, { "grad_norm": 0.17169806361198425, "learning_rate": 6.898321867794161e-05, "loss": 0.0121, "step": 32580 }, { "grad_norm": 0.12262528389692307, "learning_rate": 6.896409624286226e-05, "loss": 0.012, "step": 32590 }, { "grad_norm": 0.10052776336669922, "learning_rate": 6.894497056733754e-05, "loss": 0.0128, "step": 32600 }, { "grad_norm": 0.14377711713314056, "learning_rate": 6.89258416546355e-05, "loss": 0.0083, "step": 32610 }, { "grad_norm": 0.13720054924488068, "learning_rate": 6.890670950802474e-05, "loss": 0.0076, "step": 32620 }, { "grad_norm": 0.12924307584762573, "learning_rate": 6.88875741307744e-05, "loss": 0.0106, "step": 32630 }, { "grad_norm": 0.16033674776554108, "learning_rate": 6.886843552615425e-05, "loss": 0.0098, "step": 32640 }, { "grad_norm": 0.18783804774284363, "learning_rate": 6.884929369743451e-05, "loss": 0.0081, "step": 32650 }, { "grad_norm": 0.14557848870754242, "learning_rate": 6.8830148647886e-05, "loss": 0.0097, "step": 32660 }, { "grad_norm": 0.12060637027025223, "learning_rate": 6.88110003807801e-05, "loss": 0.0096, "step": 32670 }, { "grad_norm": 0.13168705999851227, "learning_rate": 6.87918488993887e-05, "loss": 0.0118, "step": 32680 }, { "grad_norm": 0.12695381045341492, "learning_rate": 6.877269420698431e-05, "loss": 0.0095, "step": 32690 }, { "grad_norm": 0.13875871896743774, "learning_rate": 6.875353630683989e-05, "loss": 0.0108, "step": 32700 }, { "grad_norm": 0.14278444647789001, "learning_rate": 6.873437520222905e-05, "loss": 0.008, "step": 32710 }, { "grad_norm": 0.13917139172554016, "learning_rate": 6.871521089642585e-05, "loss": 0.0085, "step": 32720 }, { "grad_norm": 0.14683395624160767, "learning_rate": 6.869604339270498e-05, "loss": 0.0105, "step": 32730 }, { "grad_norm": 0.10903777182102203, "learning_rate": 6.867687269434164e-05, "loss": 0.0072, "step": 32740 }, { "grad_norm": 0.12193501740694046, "learning_rate": 6.865769880461156e-05, "loss": 0.0097, "step": 32750 }, { "grad_norm": 0.1870705783367157, "learning_rate": 6.863852172679104e-05, "loss": 0.0094, "step": 32760 }, { "grad_norm": 0.12752728164196014, "learning_rate": 6.861934146415693e-05, "loss": 0.0105, "step": 32770 }, { "grad_norm": 0.1551402509212494, "learning_rate": 6.86001580199866e-05, "loss": 0.0095, "step": 32780 }, { "grad_norm": 0.15493999421596527, "learning_rate": 6.858097139755798e-05, "loss": 0.0085, "step": 32790 }, { "grad_norm": 0.1440151184797287, "learning_rate": 6.856178160014955e-05, "loss": 0.0104, "step": 32800 }, { "grad_norm": 0.15504896640777588, "learning_rate": 6.85425886310403e-05, "loss": 0.0087, "step": 32810 }, { "grad_norm": 0.13931693136692047, "learning_rate": 6.852339249350979e-05, "loss": 0.0087, "step": 32820 }, { "grad_norm": 0.14661258459091187, "learning_rate": 6.850419319083812e-05, "loss": 0.0125, "step": 32830 }, { "grad_norm": 0.13804034888744354, "learning_rate": 6.848499072630592e-05, "loss": 0.0083, "step": 32840 }, { "grad_norm": 0.12534624338150024, "learning_rate": 6.846578510319439e-05, "loss": 0.009, "step": 32850 }, { "grad_norm": 0.15415270626544952, "learning_rate": 6.844657632478519e-05, "loss": 0.0087, "step": 32860 }, { "grad_norm": 0.11886785179376602, "learning_rate": 6.842736439436063e-05, "loss": 0.0083, "step": 32870 }, { "grad_norm": 0.1315975934267044, "learning_rate": 6.84081493152035e-05, "loss": 0.0116, "step": 32880 }, { "grad_norm": 0.12106169760227203, "learning_rate": 6.83889310905971e-05, "loss": 0.008, "step": 32890 }, { "grad_norm": 0.1443222612142563, "learning_rate": 6.836970972382533e-05, "loss": 0.0092, "step": 32900 }, { "grad_norm": 0.11998096853494644, "learning_rate": 6.835048521817257e-05, "loss": 0.0072, "step": 32910 }, { "grad_norm": 0.13046003878116608, "learning_rate": 6.833125757692379e-05, "loss": 0.0102, "step": 32920 }, { "grad_norm": 0.153686061501503, "learning_rate": 6.831202680336441e-05, "loss": 0.0092, "step": 32930 }, { "grad_norm": 0.12054458260536194, "learning_rate": 6.829279290078052e-05, "loss": 0.0133, "step": 32940 }, { "grad_norm": 0.1826963573694229, "learning_rate": 6.827355587245863e-05, "loss": 0.0105, "step": 32950 }, { "grad_norm": 0.11694594472646713, "learning_rate": 6.82543157216858e-05, "loss": 0.0081, "step": 32960 }, { "grad_norm": 0.13141265511512756, "learning_rate": 6.823507245174969e-05, "loss": 0.0086, "step": 32970 }, { "grad_norm": 0.18059182167053223, "learning_rate": 6.821582606593841e-05, "loss": 0.0091, "step": 32980 }, { "grad_norm": 0.15979671478271484, "learning_rate": 6.81965765675407e-05, "loss": 0.01, "step": 32990 }, { "grad_norm": 0.16664670407772064, "learning_rate": 6.81773239598457e-05, "loss": 0.0095, "step": 33000 }, { "grad_norm": 0.11795108020305634, "learning_rate": 6.815806824614319e-05, "loss": 0.0089, "step": 33010 }, { "grad_norm": 0.1230592131614685, "learning_rate": 6.813880942972343e-05, "loss": 0.0097, "step": 33020 }, { "grad_norm": 0.11540485173463821, "learning_rate": 6.811954751387726e-05, "loss": 0.0108, "step": 33030 }, { "grad_norm": 0.12112277001142502, "learning_rate": 6.810028250189598e-05, "loss": 0.0096, "step": 33040 }, { "grad_norm": 0.15725553035736084, "learning_rate": 6.808101439707147e-05, "loss": 0.0105, "step": 33050 }, { "grad_norm": 0.15847638249397278, "learning_rate": 6.806174320269609e-05, "loss": 0.0107, "step": 33060 }, { "grad_norm": 0.1750524640083313, "learning_rate": 6.804246892206281e-05, "loss": 0.0091, "step": 33070 }, { "grad_norm": 0.15640948712825775, "learning_rate": 6.802319155846506e-05, "loss": 0.0112, "step": 33080 }, { "grad_norm": 0.14239546656608582, "learning_rate": 6.800391111519679e-05, "loss": 0.0125, "step": 33090 }, { "grad_norm": 0.1815686821937561, "learning_rate": 6.798462759555253e-05, "loss": 0.0105, "step": 33100 }, { "grad_norm": 0.13058525323867798, "learning_rate": 6.79653410028273e-05, "loss": 0.0092, "step": 33110 }, { "grad_norm": 0.14230895042419434, "learning_rate": 6.794605134031663e-05, "loss": 0.0087, "step": 33120 }, { "grad_norm": 0.13709992170333862, "learning_rate": 6.792675861131661e-05, "loss": 0.0082, "step": 33130 }, { "grad_norm": 0.1858038455247879, "learning_rate": 6.790746281912386e-05, "loss": 0.0132, "step": 33140 }, { "grad_norm": 0.15200942754745483, "learning_rate": 6.788816396703546e-05, "loss": 0.0101, "step": 33150 }, { "grad_norm": 0.11864624917507172, "learning_rate": 6.78688620583491e-05, "loss": 0.0095, "step": 33160 }, { "grad_norm": 0.10487104207277298, "learning_rate": 6.784955709636292e-05, "loss": 0.0106, "step": 33170 }, { "grad_norm": 0.15528899431228638, "learning_rate": 6.783024908437564e-05, "loss": 0.0109, "step": 33180 }, { "grad_norm": 0.1577674299478531, "learning_rate": 6.781093802568641e-05, "loss": 0.0084, "step": 33190 }, { "grad_norm": 0.14581738412380219, "learning_rate": 6.779162392359504e-05, "loss": 0.0083, "step": 33200 }, { "grad_norm": 0.13934412598609924, "learning_rate": 6.777230678140172e-05, "loss": 0.0123, "step": 33210 }, { "grad_norm": 0.11044397205114365, "learning_rate": 6.775298660240726e-05, "loss": 0.0088, "step": 33220 }, { "grad_norm": 0.15619347989559174, "learning_rate": 6.773366338991292e-05, "loss": 0.0099, "step": 33230 }, { "grad_norm": 0.13701212406158447, "learning_rate": 6.771433714722052e-05, "loss": 0.009, "step": 33240 }, { "grad_norm": 0.08528436720371246, "learning_rate": 6.769500787763239e-05, "loss": 0.0089, "step": 33250 }, { "grad_norm": 0.11443664878606796, "learning_rate": 6.76756755844514e-05, "loss": 0.0132, "step": 33260 }, { "grad_norm": 0.11660309880971909, "learning_rate": 6.765634027098087e-05, "loss": 0.0108, "step": 33270 }, { "grad_norm": 0.11438263207674026, "learning_rate": 6.763700194052468e-05, "loss": 0.0099, "step": 33280 }, { "grad_norm": 0.13812802731990814, "learning_rate": 6.761766059638723e-05, "loss": 0.0101, "step": 33290 }, { "grad_norm": 0.1245337426662445, "learning_rate": 6.759831624187345e-05, "loss": 0.0095, "step": 33300 }, { "grad_norm": 0.14358673989772797, "learning_rate": 6.757896888028871e-05, "loss": 0.0099, "step": 33310 }, { "grad_norm": 0.12219132483005524, "learning_rate": 6.7559618514939e-05, "loss": 0.0082, "step": 33320 }, { "grad_norm": 0.15406028926372528, "learning_rate": 6.754026514913073e-05, "loss": 0.0134, "step": 33330 }, { "grad_norm": 0.15885083377361298, "learning_rate": 6.752090878617087e-05, "loss": 0.0095, "step": 33340 }, { "grad_norm": 0.16572533547878265, "learning_rate": 6.75015494293669e-05, "loss": 0.008, "step": 33350 }, { "grad_norm": 0.18530325591564178, "learning_rate": 6.74821870820268e-05, "loss": 0.0096, "step": 33360 }, { "grad_norm": 0.14066164195537567, "learning_rate": 6.746282174745907e-05, "loss": 0.0092, "step": 33370 }, { "grad_norm": 0.13364064693450928, "learning_rate": 6.744345342897271e-05, "loss": 0.0101, "step": 33380 }, { "grad_norm": 0.15324527025222778, "learning_rate": 6.742408212987724e-05, "loss": 0.0108, "step": 33390 }, { "grad_norm": 0.14059928059577942, "learning_rate": 6.740470785348269e-05, "loss": 0.0088, "step": 33400 }, { "grad_norm": 0.1677529662847519, "learning_rate": 6.738533060309958e-05, "loss": 0.0104, "step": 33410 }, { "grad_norm": 0.16453145444393158, "learning_rate": 6.736595038203894e-05, "loss": 0.0081, "step": 33420 }, { "grad_norm": 0.13108734786510468, "learning_rate": 6.734656719361236e-05, "loss": 0.0094, "step": 33430 }, { "grad_norm": 0.1736723929643631, "learning_rate": 6.732718104113189e-05, "loss": 0.0135, "step": 33440 }, { "grad_norm": 0.17612606287002563, "learning_rate": 6.730779192791006e-05, "loss": 0.0096, "step": 33450 }, { "grad_norm": 0.11802655458450317, "learning_rate": 6.728839985725997e-05, "loss": 0.0078, "step": 33460 }, { "grad_norm": 0.17555071413516998, "learning_rate": 6.726900483249517e-05, "loss": 0.0114, "step": 33470 }, { "grad_norm": 0.20743049681186676, "learning_rate": 6.724960685692976e-05, "loss": 0.0119, "step": 33480 }, { "grad_norm": 0.10778863728046417, "learning_rate": 6.723020593387833e-05, "loss": 0.0087, "step": 33490 }, { "grad_norm": 0.09693140536546707, "learning_rate": 6.721080206665593e-05, "loss": 0.0084, "step": 33500 }, { "grad_norm": 0.15232786536216736, "learning_rate": 6.719139525857819e-05, "loss": 0.0128, "step": 33510 }, { "grad_norm": 0.14346127212047577, "learning_rate": 6.717198551296117e-05, "loss": 0.0079, "step": 33520 }, { "grad_norm": 0.16123151779174805, "learning_rate": 6.715257283312148e-05, "loss": 0.0116, "step": 33530 }, { "grad_norm": 0.24047663807868958, "learning_rate": 6.713315722237623e-05, "loss": 0.0079, "step": 33540 }, { "grad_norm": 0.12039835005998611, "learning_rate": 6.7113738684043e-05, "loss": 0.01, "step": 33550 }, { "grad_norm": 0.11503557115793228, "learning_rate": 6.709431722143989e-05, "loss": 0.0085, "step": 33560 }, { "grad_norm": 0.13703246414661407, "learning_rate": 6.70748928378855e-05, "loss": 0.0083, "step": 33570 }, { "grad_norm": 0.15839456021785736, "learning_rate": 6.705546553669891e-05, "loss": 0.0102, "step": 33580 }, { "grad_norm": 0.15494908392429352, "learning_rate": 6.703603532119974e-05, "loss": 0.01, "step": 33590 }, { "grad_norm": 0.12341190129518509, "learning_rate": 6.701660219470808e-05, "loss": 0.008, "step": 33600 }, { "grad_norm": 0.1319519430398941, "learning_rate": 6.69971661605445e-05, "loss": 0.0102, "step": 33610 }, { "grad_norm": 0.10414155572652817, "learning_rate": 6.697772722203008e-05, "loss": 0.0084, "step": 33620 }, { "grad_norm": 0.12142716348171234, "learning_rate": 6.695828538248643e-05, "loss": 0.0097, "step": 33630 }, { "grad_norm": 0.15286268293857574, "learning_rate": 6.693884064523563e-05, "loss": 0.0087, "step": 33640 }, { "grad_norm": 0.1447957158088684, "learning_rate": 6.691939301360023e-05, "loss": 0.0099, "step": 33650 }, { "grad_norm": 0.16251367330551147, "learning_rate": 6.689994249090333e-05, "loss": 0.0091, "step": 33660 }, { "grad_norm": 0.15816453099250793, "learning_rate": 6.688048908046845e-05, "loss": 0.0101, "step": 33670 }, { "grad_norm": 0.10389126092195511, "learning_rate": 6.686103278561969e-05, "loss": 0.0076, "step": 33680 }, { "grad_norm": 0.13132260739803314, "learning_rate": 6.684157360968156e-05, "loss": 0.0107, "step": 33690 }, { "grad_norm": 0.10896164923906326, "learning_rate": 6.682211155597911e-05, "loss": 0.0082, "step": 33700 }, { "grad_norm": 0.1578698605298996, "learning_rate": 6.680264662783789e-05, "loss": 0.0108, "step": 33710 }, { "grad_norm": 0.12208637595176697, "learning_rate": 6.678317882858391e-05, "loss": 0.0099, "step": 33720 }, { "grad_norm": 0.13405650854110718, "learning_rate": 6.67637081615437e-05, "loss": 0.0089, "step": 33730 }, { "grad_norm": 0.12068218737840652, "learning_rate": 6.674423463004427e-05, "loss": 0.008, "step": 33740 }, { "grad_norm": 0.11782514303922653, "learning_rate": 6.672475823741308e-05, "loss": 0.0089, "step": 33750 }, { "grad_norm": 0.10488855838775635, "learning_rate": 6.670527898697811e-05, "loss": 0.0086, "step": 33760 }, { "grad_norm": 0.15446524322032928, "learning_rate": 6.668579688206788e-05, "loss": 0.0097, "step": 33770 }, { "grad_norm": 0.16497968137264252, "learning_rate": 6.666631192601131e-05, "loss": 0.0095, "step": 33780 }, { "grad_norm": 0.11582434922456741, "learning_rate": 6.664682412213785e-05, "loss": 0.0084, "step": 33790 }, { "grad_norm": 0.10219226777553558, "learning_rate": 6.662733347377745e-05, "loss": 0.0107, "step": 33800 }, { "grad_norm": 0.14063666760921478, "learning_rate": 6.660783998426051e-05, "loss": 0.0098, "step": 33810 }, { "grad_norm": 0.10835310071706772, "learning_rate": 6.658834365691794e-05, "loss": 0.0075, "step": 33820 }, { "grad_norm": 0.1480705440044403, "learning_rate": 6.656884449508115e-05, "loss": 0.0102, "step": 33830 }, { "grad_norm": 0.12912403047084808, "learning_rate": 6.654934250208198e-05, "loss": 0.0082, "step": 33840 }, { "grad_norm": 0.3990298807621002, "learning_rate": 6.65298376812528e-05, "loss": 0.0098, "step": 33850 }, { "grad_norm": 0.11444091796875, "learning_rate": 6.651033003592646e-05, "loss": 0.0069, "step": 33860 }, { "grad_norm": 0.09850923717021942, "learning_rate": 6.649081956943626e-05, "loss": 0.0077, "step": 33870 }, { "grad_norm": 0.13654054701328278, "learning_rate": 6.647130628511604e-05, "loss": 0.0095, "step": 33880 }, { "grad_norm": 0.15310104191303253, "learning_rate": 6.645179018630005e-05, "loss": 0.0095, "step": 33890 }, { "grad_norm": 0.16513346135616302, "learning_rate": 6.643227127632309e-05, "loss": 0.0112, "step": 33900 }, { "grad_norm": 0.10778626799583435, "learning_rate": 6.641274955852038e-05, "loss": 0.0076, "step": 33910 }, { "grad_norm": 0.109223872423172, "learning_rate": 6.639322503622768e-05, "loss": 0.0088, "step": 33920 }, { "grad_norm": 0.13841545581817627, "learning_rate": 6.637369771278116e-05, "loss": 0.0119, "step": 33930 }, { "grad_norm": 0.10008282214403152, "learning_rate": 6.635416759151751e-05, "loss": 0.0082, "step": 33940 }, { "grad_norm": 0.11965849995613098, "learning_rate": 6.633463467577394e-05, "loss": 0.0077, "step": 33950 }, { "grad_norm": 0.11608099937438965, "learning_rate": 6.631509896888803e-05, "loss": 0.0104, "step": 33960 }, { "grad_norm": 0.16473673284053802, "learning_rate": 6.629556047419794e-05, "loss": 0.0099, "step": 33970 }, { "grad_norm": 0.14226895570755005, "learning_rate": 6.627601919504223e-05, "loss": 0.0108, "step": 33980 }, { "grad_norm": 0.1683964878320694, "learning_rate": 6.625647513476001e-05, "loss": 0.0093, "step": 33990 }, { "grad_norm": 0.1240696907043457, "learning_rate": 6.62369282966908e-05, "loss": 0.0119, "step": 34000 }, { "grad_norm": 0.12517112493515015, "learning_rate": 6.621737868417464e-05, "loss": 0.0083, "step": 34010 }, { "grad_norm": 0.12097883224487305, "learning_rate": 6.619782630055198e-05, "loss": 0.0094, "step": 34020 }, { "grad_norm": 0.1439327448606491, "learning_rate": 6.617827114916382e-05, "loss": 0.0114, "step": 34030 }, { "grad_norm": 0.14830873906612396, "learning_rate": 6.615871323335161e-05, "loss": 0.0093, "step": 34040 }, { "grad_norm": 0.09883216768503189, "learning_rate": 6.613915255645725e-05, "loss": 0.0093, "step": 34050 }, { "grad_norm": 0.14915859699249268, "learning_rate": 6.611958912182312e-05, "loss": 0.0126, "step": 34060 }, { "grad_norm": 0.15039730072021484, "learning_rate": 6.610002293279207e-05, "loss": 0.0097, "step": 34070 }, { "grad_norm": 0.17050392925739288, "learning_rate": 6.608045399270746e-05, "loss": 0.011, "step": 34080 }, { "grad_norm": 0.11564358323812485, "learning_rate": 6.606088230491304e-05, "loss": 0.0109, "step": 34090 }, { "grad_norm": 0.19263216853141785, "learning_rate": 6.604130787275312e-05, "loss": 0.0087, "step": 34100 }, { "grad_norm": 0.16518568992614746, "learning_rate": 6.602173069957242e-05, "loss": 0.0118, "step": 34110 }, { "grad_norm": 0.1384253054857254, "learning_rate": 6.600215078871612e-05, "loss": 0.0105, "step": 34120 }, { "grad_norm": 0.1589103639125824, "learning_rate": 6.598256814352992e-05, "loss": 0.0107, "step": 34130 }, { "grad_norm": 0.16371005773544312, "learning_rate": 6.596298276735995e-05, "loss": 0.0086, "step": 34140 }, { "grad_norm": 0.14304976165294647, "learning_rate": 6.594339466355282e-05, "loss": 0.0103, "step": 34150 }, { "grad_norm": 0.15009716153144836, "learning_rate": 6.592380383545558e-05, "loss": 0.0082, "step": 34160 }, { "grad_norm": 0.12215034663677216, "learning_rate": 6.590421028641577e-05, "loss": 0.0084, "step": 34170 }, { "grad_norm": 0.11851100623607635, "learning_rate": 6.588461401978143e-05, "loss": 0.0099, "step": 34180 }, { "grad_norm": 0.1313459277153015, "learning_rate": 6.586501503890099e-05, "loss": 0.0082, "step": 34190 }, { "grad_norm": 0.18359345197677612, "learning_rate": 6.584541334712338e-05, "loss": 0.0082, "step": 34200 }, { "grad_norm": 0.14242587983608246, "learning_rate": 6.582580894779802e-05, "loss": 0.0081, "step": 34210 }, { "grad_norm": 0.13720731437206268, "learning_rate": 6.580620184427473e-05, "loss": 0.009, "step": 34220 }, { "grad_norm": 0.14609573781490326, "learning_rate": 6.578659203990385e-05, "loss": 0.009, "step": 34230 }, { "grad_norm": 0.14875806868076324, "learning_rate": 6.576697953803615e-05, "loss": 0.0097, "step": 34240 }, { "grad_norm": 0.16523851454257965, "learning_rate": 6.57473643420229e-05, "loss": 0.0092, "step": 34250 }, { "grad_norm": 0.17354631423950195, "learning_rate": 6.572774645521574e-05, "loss": 0.0111, "step": 34260 }, { "grad_norm": 0.14844325184822083, "learning_rate": 6.570812588096688e-05, "loss": 0.0094, "step": 34270 }, { "grad_norm": 0.1235414519906044, "learning_rate": 6.568850262262893e-05, "loss": 0.0107, "step": 34280 }, { "grad_norm": 0.13632281124591827, "learning_rate": 6.566887668355497e-05, "loss": 0.009, "step": 34290 }, { "grad_norm": 0.12943920493125916, "learning_rate": 6.564924806709851e-05, "loss": 0.0076, "step": 34300 }, { "grad_norm": 0.18183496594429016, "learning_rate": 6.562961677661359e-05, "loss": 0.0078, "step": 34310 }, { "grad_norm": 0.13028140366077423, "learning_rate": 6.56099828154546e-05, "loss": 0.0103, "step": 34320 }, { "grad_norm": 0.09949102252721786, "learning_rate": 6.55903461869765e-05, "loss": 0.0088, "step": 34330 }, { "grad_norm": 0.12123191356658936, "learning_rate": 6.557070689453465e-05, "loss": 0.0077, "step": 34340 }, { "grad_norm": 0.11898631602525711, "learning_rate": 6.555106494148482e-05, "loss": 0.0087, "step": 34350 }, { "grad_norm": 0.16053543984889984, "learning_rate": 6.553142033118333e-05, "loss": 0.0096, "step": 34360 }, { "grad_norm": 0.12412213534116745, "learning_rate": 6.551177306698688e-05, "loss": 0.0082, "step": 34370 }, { "grad_norm": 0.15461181104183197, "learning_rate": 6.549212315225267e-05, "loss": 0.0099, "step": 34380 }, { "grad_norm": 0.20197150111198425, "learning_rate": 6.547247059033833e-05, "loss": 0.0133, "step": 34390 }, { "grad_norm": 0.1233319565653801, "learning_rate": 6.545281538460193e-05, "loss": 0.0079, "step": 34400 }, { "grad_norm": 0.103717140853405, "learning_rate": 6.543315753840202e-05, "loss": 0.0088, "step": 34410 }, { "grad_norm": 0.10672254860401154, "learning_rate": 6.541349705509758e-05, "loss": 0.0084, "step": 34420 }, { "grad_norm": 0.1286822110414505, "learning_rate": 6.539383393804805e-05, "loss": 0.0095, "step": 34430 }, { "grad_norm": 0.1703711897134781, "learning_rate": 6.537416819061333e-05, "loss": 0.009, "step": 34440 }, { "grad_norm": 0.17575357854366302, "learning_rate": 6.535449981615375e-05, "loss": 0.0098, "step": 34450 }, { "grad_norm": 0.12715858221054077, "learning_rate": 6.53348288180301e-05, "loss": 0.009, "step": 34460 }, { "grad_norm": 0.13053195178508759, "learning_rate": 6.531515519960361e-05, "loss": 0.0086, "step": 34470 }, { "grad_norm": 0.1171117052435875, "learning_rate": 6.529547896423597e-05, "loss": 0.0093, "step": 34480 }, { "grad_norm": 0.18868568539619446, "learning_rate": 6.52758001152893e-05, "loss": 0.0087, "step": 34490 }, { "grad_norm": 0.12111896276473999, "learning_rate": 6.525611865612618e-05, "loss": 0.0092, "step": 34500 }, { "grad_norm": 0.13554807007312775, "learning_rate": 6.523643459010966e-05, "loss": 0.01, "step": 34510 }, { "grad_norm": 0.18192799389362335, "learning_rate": 6.521674792060317e-05, "loss": 0.0095, "step": 34520 }, { "grad_norm": 0.14750485122203827, "learning_rate": 6.519705865097063e-05, "loss": 0.0101, "step": 34530 }, { "grad_norm": 0.1496707797050476, "learning_rate": 6.517736678457641e-05, "loss": 0.0087, "step": 34540 }, { "grad_norm": 0.11552002280950546, "learning_rate": 6.515767232478534e-05, "loss": 0.0098, "step": 34550 }, { "grad_norm": 0.1362028270959854, "learning_rate": 6.51379752749626e-05, "loss": 0.008, "step": 34560 }, { "grad_norm": 0.1581670492887497, "learning_rate": 6.511827563847393e-05, "loss": 0.0099, "step": 34570 }, { "grad_norm": 0.14316552877426147, "learning_rate": 6.509857341868542e-05, "loss": 0.0105, "step": 34580 }, { "grad_norm": 0.13436287641525269, "learning_rate": 6.507886861896367e-05, "loss": 0.0078, "step": 34590 }, { "grad_norm": 0.17277659475803375, "learning_rate": 6.505916124267567e-05, "loss": 0.0083, "step": 34600 }, { "grad_norm": 0.11999691277742386, "learning_rate": 6.503945129318891e-05, "loss": 0.009, "step": 34610 }, { "grad_norm": 0.13797235488891602, "learning_rate": 6.501973877387122e-05, "loss": 0.0096, "step": 34620 }, { "grad_norm": 0.15466196835041046, "learning_rate": 6.500002368809098e-05, "loss": 0.01, "step": 34630 }, { "grad_norm": 0.1608894169330597, "learning_rate": 6.498030603921694e-05, "loss": 0.0116, "step": 34640 }, { "grad_norm": 0.1761058270931244, "learning_rate": 6.496058583061832e-05, "loss": 0.0116, "step": 34650 }, { "grad_norm": 0.10234169661998749, "learning_rate": 6.494086306566475e-05, "loss": 0.0077, "step": 34660 }, { "grad_norm": 0.17682601511478424, "learning_rate": 6.492113774772632e-05, "loss": 0.008, "step": 34670 }, { "grad_norm": 0.16009651124477386, "learning_rate": 6.490140988017354e-05, "loss": 0.0106, "step": 34680 }, { "grad_norm": 0.18917402625083923, "learning_rate": 6.488167946637736e-05, "loss": 0.0083, "step": 34690 }, { "grad_norm": 0.13012173771858215, "learning_rate": 6.486194650970915e-05, "loss": 0.0101, "step": 34700 }, { "grad_norm": 0.12209083139896393, "learning_rate": 6.48422110135408e-05, "loss": 0.0084, "step": 34710 }, { "grad_norm": 0.19162651896476746, "learning_rate": 6.482247298124451e-05, "loss": 0.0092, "step": 34720 }, { "grad_norm": 0.1674707531929016, "learning_rate": 6.480273241619297e-05, "loss": 0.0075, "step": 34730 }, { "grad_norm": 0.13753493130207062, "learning_rate": 6.478298932175933e-05, "loss": 0.0099, "step": 34740 }, { "grad_norm": 0.12536847591400146, "learning_rate": 6.476324370131712e-05, "loss": 0.0098, "step": 34750 }, { "grad_norm": 0.14549857378005981, "learning_rate": 6.474349555824036e-05, "loss": 0.0088, "step": 34760 }, { "grad_norm": 0.18047119677066803, "learning_rate": 6.472374489590342e-05, "loss": 0.0088, "step": 34770 }, { "grad_norm": 0.16272486746311188, "learning_rate": 6.470399171768118e-05, "loss": 0.0072, "step": 34780 }, { "grad_norm": 0.12330887466669083, "learning_rate": 6.468423602694891e-05, "loss": 0.0088, "step": 34790 }, { "grad_norm": 0.15167053043842316, "learning_rate": 6.466447782708232e-05, "loss": 0.0084, "step": 34800 }, { "grad_norm": 0.14543576538562775, "learning_rate": 6.464471712145754e-05, "loss": 0.0111, "step": 34810 }, { "grad_norm": 0.1368837058544159, "learning_rate": 6.462495391345114e-05, "loss": 0.0084, "step": 34820 }, { "grad_norm": 0.13053478300571442, "learning_rate": 6.46051882064401e-05, "loss": 0.0105, "step": 34830 }, { "grad_norm": 0.1839286983013153, "learning_rate": 6.458542000380186e-05, "loss": 0.0081, "step": 34840 }, { "grad_norm": 0.10118689388036728, "learning_rate": 6.456564930891424e-05, "loss": 0.0078, "step": 34850 }, { "grad_norm": 0.11833352595567703, "learning_rate": 6.454587612515555e-05, "loss": 0.0071, "step": 34860 }, { "grad_norm": 0.13847123086452484, "learning_rate": 6.452610045590444e-05, "loss": 0.0101, "step": 34870 }, { "grad_norm": 0.11205646395683289, "learning_rate": 6.450632230454005e-05, "loss": 0.0085, "step": 34880 }, { "grad_norm": 0.12259619683027267, "learning_rate": 6.448654167444195e-05, "loss": 0.0094, "step": 34890 }, { "grad_norm": 0.11165999621152878, "learning_rate": 6.446675856899005e-05, "loss": 0.0083, "step": 34900 }, { "grad_norm": 0.15433304011821747, "learning_rate": 6.444697299156481e-05, "loss": 0.0079, "step": 34910 }, { "grad_norm": 0.1754053384065628, "learning_rate": 6.442718494554701e-05, "loss": 0.0077, "step": 34920 }, { "grad_norm": 0.12593217194080353, "learning_rate": 6.440739443431787e-05, "loss": 0.0081, "step": 34930 }, { "grad_norm": 0.14041048288345337, "learning_rate": 6.438760146125906e-05, "loss": 0.0102, "step": 34940 }, { "grad_norm": 0.15306931734085083, "learning_rate": 6.436780602975267e-05, "loss": 0.0085, "step": 34950 }, { "grad_norm": 0.14535677433013916, "learning_rate": 6.43480081431812e-05, "loss": 0.0082, "step": 34960 }, { "grad_norm": 0.09070346504449844, "learning_rate": 6.432820780492756e-05, "loss": 0.0084, "step": 34970 }, { "grad_norm": 0.13262099027633667, "learning_rate": 6.430840501837506e-05, "loss": 0.0111, "step": 34980 }, { "grad_norm": 0.15804798901081085, "learning_rate": 6.428859978690748e-05, "loss": 0.0102, "step": 34990 }, { "grad_norm": 0.14006410539150238, "learning_rate": 6.426879211390901e-05, "loss": 0.0085, "step": 35000 }, { "grad_norm": 0.13749730587005615, "learning_rate": 6.424898200276422e-05, "loss": 0.0128, "step": 35010 }, { "grad_norm": 0.15041349828243256, "learning_rate": 6.42291694568581e-05, "loss": 0.009, "step": 35020 }, { "grad_norm": 0.13170897960662842, "learning_rate": 6.42093544795761e-05, "loss": 0.0112, "step": 35030 }, { "grad_norm": 0.12648391723632812, "learning_rate": 6.418953707430403e-05, "loss": 0.0091, "step": 35040 }, { "grad_norm": 0.0960114374756813, "learning_rate": 6.416971724442819e-05, "loss": 0.0077, "step": 35050 }, { "grad_norm": 0.09859751909971237, "learning_rate": 6.414989499333519e-05, "loss": 0.0077, "step": 35060 }, { "grad_norm": 0.14534585177898407, "learning_rate": 6.413007032441214e-05, "loss": 0.009, "step": 35070 }, { "grad_norm": 0.11856398731470108, "learning_rate": 6.411024324104653e-05, "loss": 0.0131, "step": 35080 }, { "grad_norm": 0.15307630598545074, "learning_rate": 6.409041374662628e-05, "loss": 0.0102, "step": 35090 }, { "grad_norm": 0.15785783529281616, "learning_rate": 6.407058184453967e-05, "loss": 0.0105, "step": 35100 }, { "grad_norm": 0.13607577979564667, "learning_rate": 6.405074753817548e-05, "loss": 0.0134, "step": 35110 }, { "grad_norm": 0.15685512125492096, "learning_rate": 6.40309108309228e-05, "loss": 0.0103, "step": 35120 }, { "grad_norm": 0.1252901554107666, "learning_rate": 6.401107172617122e-05, "loss": 0.009, "step": 35130 }, { "grad_norm": 0.1385154277086258, "learning_rate": 6.399123022731068e-05, "loss": 0.0084, "step": 35140 }, { "grad_norm": 0.10314377397298813, "learning_rate": 6.397138633773157e-05, "loss": 0.0089, "step": 35150 }, { "grad_norm": 0.1374790370464325, "learning_rate": 6.395154006082463e-05, "loss": 0.0083, "step": 35160 }, { "grad_norm": 0.11758182942867279, "learning_rate": 6.393169139998109e-05, "loss": 0.0077, "step": 35170 }, { "grad_norm": 0.12852634489536285, "learning_rate": 6.39118403585925e-05, "loss": 0.0116, "step": 35180 }, { "grad_norm": 0.19580820202827454, "learning_rate": 6.38919869400509e-05, "loss": 0.0105, "step": 35190 }, { "grad_norm": 0.1455964893102646, "learning_rate": 6.387213114774865e-05, "loss": 0.0124, "step": 35200 }, { "grad_norm": 0.13327349722385406, "learning_rate": 6.385227298507863e-05, "loss": 0.0086, "step": 35210 }, { "grad_norm": 0.15516161918640137, "learning_rate": 6.3832412455434e-05, "loss": 0.0082, "step": 35220 }, { "grad_norm": 0.14677536487579346, "learning_rate": 6.381254956220841e-05, "loss": 0.0082, "step": 35230 }, { "grad_norm": 0.12767057120800018, "learning_rate": 6.379268430879586e-05, "loss": 0.0089, "step": 35240 }, { "grad_norm": 0.11366351693868637, "learning_rate": 6.37728166985908e-05, "loss": 0.009, "step": 35250 }, { "grad_norm": 0.1553105264902115, "learning_rate": 6.375294673498804e-05, "loss": 0.0094, "step": 35260 }, { "grad_norm": 0.18124471604824066, "learning_rate": 6.373307442138284e-05, "loss": 0.0086, "step": 35270 }, { "grad_norm": 0.10474932193756104, "learning_rate": 6.371319976117081e-05, "loss": 0.0073, "step": 35280 }, { "grad_norm": 0.1376945823431015, "learning_rate": 6.3693322757748e-05, "loss": 0.0101, "step": 35290 }, { "grad_norm": 0.13285033404827118, "learning_rate": 6.367344341451086e-05, "loss": 0.0097, "step": 35300 }, { "grad_norm": 0.15947659313678741, "learning_rate": 6.36535617348562e-05, "loss": 0.0096, "step": 35310 }, { "grad_norm": 0.16154909133911133, "learning_rate": 6.363367772218128e-05, "loss": 0.01, "step": 35320 }, { "grad_norm": 0.18291375041007996, "learning_rate": 6.36137913798837e-05, "loss": 0.0111, "step": 35330 }, { "grad_norm": 0.12219500541687012, "learning_rate": 6.359390271136151e-05, "loss": 0.0098, "step": 35340 }, { "grad_norm": 0.18115688860416412, "learning_rate": 6.357401172001314e-05, "loss": 0.0092, "step": 35350 }, { "grad_norm": 0.12967641651630402, "learning_rate": 6.355411840923742e-05, "loss": 0.0081, "step": 35360 }, { "grad_norm": 0.15280702710151672, "learning_rate": 6.353422278243358e-05, "loss": 0.0132, "step": 35370 }, { "grad_norm": 0.11128456145524979, "learning_rate": 6.351432484300121e-05, "loss": 0.0072, "step": 35380 }, { "grad_norm": 0.13709813356399536, "learning_rate": 6.349442459434036e-05, "loss": 0.009, "step": 35390 }, { "grad_norm": 0.17246569693088531, "learning_rate": 6.34745220398514e-05, "loss": 0.0101, "step": 35400 }, { "grad_norm": 0.18103142082691193, "learning_rate": 6.345461718293518e-05, "loss": 0.0085, "step": 35410 }, { "grad_norm": 0.15212509036064148, "learning_rate": 6.343471002699286e-05, "loss": 0.0127, "step": 35420 }, { "grad_norm": 0.093815878033638, "learning_rate": 6.341480057542602e-05, "loss": 0.0105, "step": 35430 }, { "grad_norm": 0.14236973226070404, "learning_rate": 6.339488883163667e-05, "loss": 0.0132, "step": 35440 }, { "grad_norm": 0.1589837521314621, "learning_rate": 6.337497479902716e-05, "loss": 0.0102, "step": 35450 }, { "grad_norm": 0.1404905617237091, "learning_rate": 6.335505848100027e-05, "loss": 0.0098, "step": 35460 }, { "grad_norm": 0.1756250113248825, "learning_rate": 6.333513988095915e-05, "loss": 0.0109, "step": 35470 }, { "grad_norm": 0.1555190086364746, "learning_rate": 6.331521900230735e-05, "loss": 0.01, "step": 35480 }, { "grad_norm": 0.12065453827381134, "learning_rate": 6.329529584844878e-05, "loss": 0.0062, "step": 35490 }, { "grad_norm": 0.1461297869682312, "learning_rate": 6.327537042278777e-05, "loss": 0.0107, "step": 35500 }, { "grad_norm": 0.11911161243915558, "learning_rate": 6.325544272872905e-05, "loss": 0.0123, "step": 35510 }, { "grad_norm": 0.12856915593147278, "learning_rate": 6.323551276967771e-05, "loss": 0.0104, "step": 35520 }, { "grad_norm": 0.11253461986780167, "learning_rate": 6.321558054903922e-05, "loss": 0.0093, "step": 35530 }, { "grad_norm": 0.17054633796215057, "learning_rate": 6.319564607021947e-05, "loss": 0.0097, "step": 35540 }, { "grad_norm": 0.15594927966594696, "learning_rate": 6.31757093366247e-05, "loss": 0.0084, "step": 35550 }, { "grad_norm": 0.10598254203796387, "learning_rate": 6.315577035166154e-05, "loss": 0.0089, "step": 35560 }, { "grad_norm": 0.12360940128564835, "learning_rate": 6.313582911873708e-05, "loss": 0.009, "step": 35570 }, { "grad_norm": 0.1479555070400238, "learning_rate": 6.311588564125865e-05, "loss": 0.0088, "step": 35580 }, { "grad_norm": 0.18622492253780365, "learning_rate": 6.30959399226341e-05, "loss": 0.0096, "step": 35590 }, { "grad_norm": 0.14209146797657013, "learning_rate": 6.30759919662716e-05, "loss": 0.0088, "step": 35600 }, { "grad_norm": 0.1292075365781784, "learning_rate": 6.30560417755797e-05, "loss": 0.0091, "step": 35610 }, { "grad_norm": 0.15628068149089813, "learning_rate": 6.303608935396735e-05, "loss": 0.0101, "step": 35620 }, { "grad_norm": 0.13798293471336365, "learning_rate": 6.301613470484386e-05, "loss": 0.0108, "step": 35630 }, { "grad_norm": 0.1912301629781723, "learning_rate": 6.299617783161893e-05, "loss": 0.0115, "step": 35640 }, { "grad_norm": 0.16548138856887817, "learning_rate": 6.297621873770266e-05, "loss": 0.0089, "step": 35650 }, { "grad_norm": 0.1352074146270752, "learning_rate": 6.29562574265055e-05, "loss": 0.01, "step": 35660 }, { "grad_norm": 0.12748606503009796, "learning_rate": 6.293629390143834e-05, "loss": 0.0092, "step": 35670 }, { "grad_norm": 0.12047647684812546, "learning_rate": 6.291632816591232e-05, "loss": 0.0091, "step": 35680 }, { "grad_norm": 0.12442407011985779, "learning_rate": 6.28963602233391e-05, "loss": 0.0106, "step": 35690 }, { "grad_norm": 0.14335747063159943, "learning_rate": 6.287639007713062e-05, "loss": 0.0092, "step": 35700 }, { "grad_norm": 0.11454543471336365, "learning_rate": 6.285641773069926e-05, "loss": 0.0077, "step": 35710 }, { "grad_norm": 0.14557528495788574, "learning_rate": 6.283644318745773e-05, "loss": 0.0112, "step": 35720 }, { "grad_norm": 0.12514832615852356, "learning_rate": 6.281646645081912e-05, "loss": 0.011, "step": 35730 }, { "grad_norm": 0.1431799679994583, "learning_rate": 6.279648752419693e-05, "loss": 0.0091, "step": 35740 }, { "grad_norm": 0.14689406752586365, "learning_rate": 6.2776506411005e-05, "loss": 0.0086, "step": 35750 }, { "grad_norm": 0.1587585061788559, "learning_rate": 6.275652311465758e-05, "loss": 0.0086, "step": 35760 }, { "grad_norm": 0.1536289006471634, "learning_rate": 6.273653763856926e-05, "loss": 0.0107, "step": 35770 }, { "grad_norm": 0.1478891521692276, "learning_rate": 6.271654998615501e-05, "loss": 0.0071, "step": 35780 }, { "grad_norm": 0.15620066225528717, "learning_rate": 6.269656016083013e-05, "loss": 0.0082, "step": 35790 }, { "grad_norm": 0.15615950524806976, "learning_rate": 6.267656816601038e-05, "loss": 0.0105, "step": 35800 }, { "grad_norm": 0.1370677500963211, "learning_rate": 6.265657400511185e-05, "loss": 0.0108, "step": 35810 }, { "grad_norm": 0.15288332104682922, "learning_rate": 6.263657768155098e-05, "loss": 0.0084, "step": 35820 }, { "grad_norm": 0.12971670925617218, "learning_rate": 6.261657919874457e-05, "loss": 0.0077, "step": 35830 }, { "grad_norm": 0.12946583330631256, "learning_rate": 6.259657856010986e-05, "loss": 0.0083, "step": 35840 }, { "grad_norm": 0.11685734987258911, "learning_rate": 6.257657576906439e-05, "loss": 0.0086, "step": 35850 }, { "grad_norm": 0.1271001249551773, "learning_rate": 6.255657082902609e-05, "loss": 0.0075, "step": 35860 }, { "grad_norm": 0.11714619398117065, "learning_rate": 6.253656374341325e-05, "loss": 0.0118, "step": 35870 }, { "grad_norm": 0.11695195734500885, "learning_rate": 6.251655451564457e-05, "loss": 0.007, "step": 35880 }, { "grad_norm": 0.1389402151107788, "learning_rate": 6.249654314913902e-05, "loss": 0.0077, "step": 35890 }, { "grad_norm": 0.16051997244358063, "learning_rate": 6.247652964731604e-05, "loss": 0.0081, "step": 35900 }, { "grad_norm": 0.15353655815124512, "learning_rate": 6.245651401359537e-05, "loss": 0.0106, "step": 35910 }, { "grad_norm": 0.1808718889951706, "learning_rate": 6.243649625139715e-05, "loss": 0.0077, "step": 35920 }, { "grad_norm": 0.1420517861843109, "learning_rate": 6.241647636414185e-05, "loss": 0.0091, "step": 35930 }, { "grad_norm": 0.12462937831878662, "learning_rate": 6.239645435525034e-05, "loss": 0.0084, "step": 35940 }, { "grad_norm": 0.0936473086476326, "learning_rate": 6.237643022814381e-05, "loss": 0.0086, "step": 35950 }, { "grad_norm": 0.08954925835132599, "learning_rate": 6.235640398624386e-05, "loss": 0.0073, "step": 35960 }, { "grad_norm": 0.17925041913986206, "learning_rate": 6.233637563297243e-05, "loss": 0.0093, "step": 35970 }, { "grad_norm": 0.13860303163528442, "learning_rate": 6.23163451717518e-05, "loss": 0.0091, "step": 35980 }, { "grad_norm": 0.20323802530765533, "learning_rate": 6.229631260600463e-05, "loss": 0.0101, "step": 35990 }, { "grad_norm": 0.12351475656032562, "learning_rate": 6.227627793915392e-05, "loss": 0.0101, "step": 36000 }, { "grad_norm": 0.1414506882429123, "learning_rate": 6.225624117462309e-05, "loss": 0.0084, "step": 36010 }, { "grad_norm": 0.19040009379386902, "learning_rate": 6.223620231583586e-05, "loss": 0.0084, "step": 36020 }, { "grad_norm": 0.1263696253299713, "learning_rate": 6.221616136621629e-05, "loss": 0.0114, "step": 36030 }, { "grad_norm": 0.11557792872190475, "learning_rate": 6.219611832918887e-05, "loss": 0.0087, "step": 36040 }, { "grad_norm": 0.14145028591156006, "learning_rate": 6.217607320817838e-05, "loss": 0.0086, "step": 36050 }, { "grad_norm": 0.12399866431951523, "learning_rate": 6.215602600661001e-05, "loss": 0.0095, "step": 36060 }, { "grad_norm": 0.17773868143558502, "learning_rate": 6.213597672790925e-05, "loss": 0.0092, "step": 36070 }, { "grad_norm": 0.1466801017522812, "learning_rate": 6.2115925375502e-05, "loss": 0.0092, "step": 36080 }, { "grad_norm": 0.14560404419898987, "learning_rate": 6.209587195281447e-05, "loss": 0.0098, "step": 36090 }, { "grad_norm": 0.1319192349910736, "learning_rate": 6.207581646327324e-05, "loss": 0.0108, "step": 36100 }, { "grad_norm": 0.13294216990470886, "learning_rate": 6.205575891030526e-05, "loss": 0.0089, "step": 36110 }, { "grad_norm": 0.1382438838481903, "learning_rate": 6.203569929733781e-05, "loss": 0.0095, "step": 36120 }, { "grad_norm": 0.13807913661003113, "learning_rate": 6.201563762779852e-05, "loss": 0.0089, "step": 36130 }, { "grad_norm": 0.14910638332366943, "learning_rate": 6.199557390511538e-05, "loss": 0.008, "step": 36140 }, { "grad_norm": 0.17160305380821228, "learning_rate": 6.197550813271675e-05, "loss": 0.0091, "step": 36150 }, { "grad_norm": 0.24311651289463043, "learning_rate": 6.195544031403131e-05, "loss": 0.0077, "step": 36160 }, { "grad_norm": 0.10487133264541626, "learning_rate": 6.19353704524881e-05, "loss": 0.01, "step": 36170 }, { "grad_norm": 0.1047452837228775, "learning_rate": 6.191529855151652e-05, "loss": 0.008, "step": 36180 }, { "grad_norm": 0.13835157454013824, "learning_rate": 6.189522461454629e-05, "loss": 0.0087, "step": 36190 }, { "grad_norm": 0.118964284658432, "learning_rate": 6.187514864500752e-05, "loss": 0.0076, "step": 36200 }, { "grad_norm": 0.118782639503479, "learning_rate": 6.185507064633062e-05, "loss": 0.0073, "step": 36210 }, { "grad_norm": 0.09283566474914551, "learning_rate": 6.18349906219464e-05, "loss": 0.0072, "step": 36220 }, { "grad_norm": 0.13646258413791656, "learning_rate": 6.181490857528596e-05, "loss": 0.0106, "step": 36230 }, { "grad_norm": 0.14553441107273102, "learning_rate": 6.179482450978077e-05, "loss": 0.0092, "step": 36240 }, { "grad_norm": 0.11927549540996552, "learning_rate": 6.177473842886269e-05, "loss": 0.009, "step": 36250 }, { "grad_norm": 0.14189325273036957, "learning_rate": 6.175465033596382e-05, "loss": 0.0098, "step": 36260 }, { "grad_norm": 0.13475461304187775, "learning_rate": 6.173456023451671e-05, "loss": 0.009, "step": 36270 }, { "grad_norm": 0.14184331893920898, "learning_rate": 6.171446812795422e-05, "loss": 0.0098, "step": 36280 }, { "grad_norm": 0.15708331763744354, "learning_rate": 6.169437401970949e-05, "loss": 0.0079, "step": 36290 }, { "grad_norm": 0.16038750112056732, "learning_rate": 6.16742779132161e-05, "loss": 0.0079, "step": 36300 }, { "grad_norm": 0.13189339637756348, "learning_rate": 6.165417981190789e-05, "loss": 0.0079, "step": 36310 }, { "grad_norm": 0.1413940191268921, "learning_rate": 6.16340797192191e-05, "loss": 0.008, "step": 36320 }, { "grad_norm": 0.15401847660541534, "learning_rate": 6.161397763858427e-05, "loss": 0.0091, "step": 36330 }, { "grad_norm": 0.20080378651618958, "learning_rate": 6.159387357343834e-05, "loss": 0.0108, "step": 36340 }, { "grad_norm": 0.11560485512018204, "learning_rate": 6.157376752721648e-05, "loss": 0.0082, "step": 36350 }, { "grad_norm": 0.10563166439533234, "learning_rate": 6.155365950335428e-05, "loss": 0.0066, "step": 36360 }, { "grad_norm": 0.13204234838485718, "learning_rate": 6.153354950528768e-05, "loss": 0.0075, "step": 36370 }, { "grad_norm": 0.15404249727725983, "learning_rate": 6.151343753645293e-05, "loss": 0.0109, "step": 36380 }, { "grad_norm": 0.13382798433303833, "learning_rate": 6.149332360028657e-05, "loss": 0.0108, "step": 36390 }, { "grad_norm": 0.1667860597372055, "learning_rate": 6.147320770022555e-05, "loss": 0.008, "step": 36400 }, { "grad_norm": 0.15631049871444702, "learning_rate": 6.145308983970715e-05, "loss": 0.0082, "step": 36410 }, { "grad_norm": 0.16441945731639862, "learning_rate": 6.143297002216892e-05, "loss": 0.0085, "step": 36420 }, { "grad_norm": 0.1326589733362198, "learning_rate": 6.141284825104882e-05, "loss": 0.0101, "step": 36430 }, { "grad_norm": 0.14416725933551788, "learning_rate": 6.13927245297851e-05, "loss": 0.0106, "step": 36440 }, { "grad_norm": 0.19273462891578674, "learning_rate": 6.137259886181633e-05, "loss": 0.0101, "step": 36450 }, { "grad_norm": 0.11641886085271835, "learning_rate": 6.135247125058145e-05, "loss": 0.0084, "step": 36460 }, { "grad_norm": 0.1347382813692093, "learning_rate": 6.133234169951974e-05, "loss": 0.0086, "step": 36470 }, { "grad_norm": 0.1042318195104599, "learning_rate": 6.131221021207078e-05, "loss": 0.0081, "step": 36480 }, { "grad_norm": 0.1297278106212616, "learning_rate": 6.129207679167448e-05, "loss": 0.0078, "step": 36490 }, { "grad_norm": 0.12625813484191895, "learning_rate": 6.127194144177109e-05, "loss": 0.0078, "step": 36500 }, { "grad_norm": 0.14293010532855988, "learning_rate": 6.125180416580118e-05, "loss": 0.0073, "step": 36510 }, { "grad_norm": 0.08999461680650711, "learning_rate": 6.123166496720571e-05, "loss": 0.0061, "step": 36520 }, { "grad_norm": 0.0815160796046257, "learning_rate": 6.121152384942588e-05, "loss": 0.007, "step": 36530 }, { "grad_norm": 0.17815925180912018, "learning_rate": 6.119138081590324e-05, "loss": 0.0094, "step": 36540 }, { "grad_norm": 0.12347102165222168, "learning_rate": 6.117123587007971e-05, "loss": 0.0078, "step": 36550 }, { "grad_norm": 0.10589826852083206, "learning_rate": 6.11510890153975e-05, "loss": 0.007, "step": 36560 }, { "grad_norm": 0.08592724055051804, "learning_rate": 6.113094025529916e-05, "loss": 0.0068, "step": 36570 }, { "grad_norm": 0.11877107620239258, "learning_rate": 6.111078959322757e-05, "loss": 0.0116, "step": 36580 }, { "grad_norm": 0.1338849812746048, "learning_rate": 6.109063703262592e-05, "loss": 0.0105, "step": 36590 }, { "grad_norm": 0.15293164551258087, "learning_rate": 6.107048257693772e-05, "loss": 0.0093, "step": 36600 }, { "grad_norm": 0.18705162405967712, "learning_rate": 6.105032622960683e-05, "loss": 0.01, "step": 36610 }, { "grad_norm": 0.1241745725274086, "learning_rate": 6.103016799407743e-05, "loss": 0.0091, "step": 36620 }, { "grad_norm": 0.11430615931749344, "learning_rate": 6.1010007873793984e-05, "loss": 0.007, "step": 36630 }, { "grad_norm": 0.17164456844329834, "learning_rate": 6.098984587220131e-05, "loss": 0.0086, "step": 36640 }, { "grad_norm": 0.12653334438800812, "learning_rate": 6.096968199274456e-05, "loss": 0.0098, "step": 36650 }, { "grad_norm": 0.1482526957988739, "learning_rate": 6.0949516238869166e-05, "loss": 0.0067, "step": 36660 }, { "grad_norm": 0.12870030105113983, "learning_rate": 6.092934861402092e-05, "loss": 0.0077, "step": 36670 }, { "grad_norm": 0.1184038296341896, "learning_rate": 6.0909179121645924e-05, "loss": 0.0088, "step": 36680 }, { "grad_norm": 0.12813368439674377, "learning_rate": 6.0889007765190576e-05, "loss": 0.0101, "step": 36690 }, { "grad_norm": 0.1215892806649208, "learning_rate": 6.086883454810162e-05, "loss": 0.0088, "step": 36700 }, { "grad_norm": 0.13129869103431702, "learning_rate": 6.0848659473826084e-05, "loss": 0.0082, "step": 36710 }, { "grad_norm": 0.13575562834739685, "learning_rate": 6.082848254581138e-05, "loss": 0.0098, "step": 36720 }, { "grad_norm": 0.08187636733055115, "learning_rate": 6.080830376750517e-05, "loss": 0.0078, "step": 36730 }, { "grad_norm": 0.11338767409324646, "learning_rate": 6.0788123142355445e-05, "loss": 0.0081, "step": 36740 }, { "grad_norm": 0.07798078656196594, "learning_rate": 6.076794067381052e-05, "loss": 0.011, "step": 36750 }, { "grad_norm": 0.11147021502256393, "learning_rate": 6.074775636531905e-05, "loss": 0.0104, "step": 36760 }, { "grad_norm": 0.2030383050441742, "learning_rate": 6.072757022032997e-05, "loss": 0.0067, "step": 36770 }, { "grad_norm": 0.35099878907203674, "learning_rate": 6.070738224229253e-05, "loss": 0.0082, "step": 36780 }, { "grad_norm": 0.20001138746738434, "learning_rate": 6.0687192434656314e-05, "loss": 0.0089, "step": 36790 }, { "grad_norm": 0.15731917321681976, "learning_rate": 6.066700080087121e-05, "loss": 0.0104, "step": 36800 }, { "grad_norm": 0.11053655296564102, "learning_rate": 6.0646807344387424e-05, "loss": 0.0079, "step": 36810 }, { "grad_norm": 0.17336848378181458, "learning_rate": 6.062661206865543e-05, "loss": 0.0106, "step": 36820 }, { "grad_norm": 0.1756565123796463, "learning_rate": 6.06064149771261e-05, "loss": 0.0107, "step": 36830 }, { "grad_norm": 0.15864448249340057, "learning_rate": 6.058621607325051e-05, "loss": 0.0084, "step": 36840 }, { "grad_norm": 0.13502535223960876, "learning_rate": 6.056601536048014e-05, "loss": 0.0093, "step": 36850 }, { "grad_norm": 0.13435277342796326, "learning_rate": 6.0545812842266725e-05, "loss": 0.0098, "step": 36860 }, { "grad_norm": 0.15001849830150604, "learning_rate": 6.052560852206232e-05, "loss": 0.0079, "step": 36870 }, { "grad_norm": 0.08428347855806351, "learning_rate": 6.05054024033193e-05, "loss": 0.0093, "step": 36880 }, { "grad_norm": 0.19580420851707458, "learning_rate": 6.048519448949032e-05, "loss": 0.0076, "step": 36890 }, { "grad_norm": 0.1566636711359024, "learning_rate": 6.046498478402839e-05, "loss": 0.009, "step": 36900 }, { "grad_norm": 0.16061735153198242, "learning_rate": 6.044477329038677e-05, "loss": 0.0118, "step": 36910 }, { "grad_norm": 0.11606305837631226, "learning_rate": 6.042456001201906e-05, "loss": 0.0087, "step": 36920 }, { "grad_norm": 0.1324979066848755, "learning_rate": 6.040434495237917e-05, "loss": 0.0073, "step": 36930 }, { "grad_norm": 0.16741158068180084, "learning_rate": 6.0384128114921256e-05, "loss": 0.009, "step": 36940 }, { "grad_norm": 0.15688541531562805, "learning_rate": 6.036390950309987e-05, "loss": 0.0113, "step": 36950 }, { "grad_norm": 0.11922308802604675, "learning_rate": 6.0343689120369805e-05, "loss": 0.0127, "step": 36960 }, { "grad_norm": 0.12019897252321243, "learning_rate": 6.032346697018616e-05, "loss": 0.0088, "step": 36970 }, { "grad_norm": 0.1306859701871872, "learning_rate": 6.0303243056004375e-05, "loss": 0.008, "step": 36980 }, { "grad_norm": 0.13534323871135712, "learning_rate": 6.0283017381280136e-05, "loss": 0.0081, "step": 36990 }, { "grad_norm": 0.13017503917217255, "learning_rate": 6.026278994946948e-05, "loss": 0.0105, "step": 37000 }, { "grad_norm": 0.15190190076828003, "learning_rate": 6.02425607640287e-05, "loss": 0.0101, "step": 37010 }, { "grad_norm": 0.10410691797733307, "learning_rate": 6.022232982841441e-05, "loss": 0.0078, "step": 37020 }, { "grad_norm": 0.14236174523830414, "learning_rate": 6.020209714608355e-05, "loss": 0.0081, "step": 37030 }, { "grad_norm": 0.10985168069601059, "learning_rate": 6.018186272049331e-05, "loss": 0.0111, "step": 37040 }, { "grad_norm": 0.12948596477508545, "learning_rate": 6.01616265551012e-05, "loss": 0.01, "step": 37050 }, { "grad_norm": 0.11598857492208481, "learning_rate": 6.014138865336503e-05, "loss": 0.0114, "step": 37060 }, { "grad_norm": 0.11226200312376022, "learning_rate": 6.0121149018742905e-05, "loss": 0.0081, "step": 37070 }, { "grad_norm": 0.15322993695735931, "learning_rate": 6.010090765469325e-05, "loss": 0.0082, "step": 37080 }, { "grad_norm": 0.13452745974063873, "learning_rate": 6.008066456467473e-05, "loss": 0.0104, "step": 37090 }, { "grad_norm": 0.15817460417747498, "learning_rate": 6.0060419752146335e-05, "loss": 0.0088, "step": 37100 }, { "grad_norm": 0.1191396489739418, "learning_rate": 6.0040173220567353e-05, "loss": 0.0076, "step": 37110 }, { "grad_norm": 0.1486348956823349, "learning_rate": 6.001992497339737e-05, "loss": 0.007, "step": 37120 }, { "grad_norm": 0.10147425532341003, "learning_rate": 5.999967501409626e-05, "loss": 0.0077, "step": 37130 }, { "grad_norm": 0.11298152804374695, "learning_rate": 5.997942334612418e-05, "loss": 0.0064, "step": 37140 }, { "grad_norm": 0.1455271989107132, "learning_rate": 5.995916997294158e-05, "loss": 0.0086, "step": 37150 }, { "grad_norm": 0.12990660965442657, "learning_rate": 5.9938914898009214e-05, "loss": 0.0074, "step": 37160 }, { "grad_norm": 0.21609614789485931, "learning_rate": 5.991865812478813e-05, "loss": 0.0098, "step": 37170 }, { "grad_norm": 0.09510092437267303, "learning_rate": 5.989839965673964e-05, "loss": 0.0094, "step": 37180 }, { "grad_norm": 0.2529117465019226, "learning_rate": 5.987813949732539e-05, "loss": 0.0124, "step": 37190 }, { "grad_norm": 0.1272066980600357, "learning_rate": 5.9857877650007255e-05, "loss": 0.0095, "step": 37200 }, { "grad_norm": 0.15674526989459991, "learning_rate": 5.983761411824744e-05, "loss": 0.0111, "step": 37210 }, { "grad_norm": 0.12619490921497345, "learning_rate": 5.981734890550844e-05, "loss": 0.009, "step": 37220 }, { "grad_norm": 0.11421380192041397, "learning_rate": 5.979708201525301e-05, "loss": 0.0103, "step": 37230 }, { "grad_norm": 0.12182971090078354, "learning_rate": 5.977681345094422e-05, "loss": 0.0081, "step": 37240 }, { "grad_norm": 0.09884937107563019, "learning_rate": 5.97565432160454e-05, "loss": 0.01, "step": 37250 }, { "grad_norm": 0.11972250044345856, "learning_rate": 5.9736271314020186e-05, "loss": 0.0085, "step": 37260 }, { "grad_norm": 0.20047029852867126, "learning_rate": 5.971599774833251e-05, "loss": 0.0084, "step": 37270 }, { "grad_norm": 0.15405529737472534, "learning_rate": 5.9695722522446525e-05, "loss": 0.0101, "step": 37280 }, { "grad_norm": 0.15244567394256592, "learning_rate": 5.9675445639826765e-05, "loss": 0.0073, "step": 37290 }, { "grad_norm": 0.15420591831207275, "learning_rate": 5.965516710393796e-05, "loss": 0.0098, "step": 37300 }, { "grad_norm": 0.11138743162155151, "learning_rate": 5.963488691824516e-05, "loss": 0.0067, "step": 37310 }, { "grad_norm": 0.1329888552427292, "learning_rate": 5.96146050862137e-05, "loss": 0.0115, "step": 37320 }, { "grad_norm": 0.15707866847515106, "learning_rate": 5.959432161130919e-05, "loss": 0.0101, "step": 37330 }, { "grad_norm": 0.17803919315338135, "learning_rate": 5.9574036496997545e-05, "loss": 0.0095, "step": 37340 }, { "grad_norm": 0.11339455097913742, "learning_rate": 5.955374974674488e-05, "loss": 0.0086, "step": 37350 }, { "grad_norm": 0.11669214814901352, "learning_rate": 5.9533461364017696e-05, "loss": 0.0097, "step": 37360 }, { "grad_norm": 0.1185925081372261, "learning_rate": 5.9513171352282716e-05, "loss": 0.0077, "step": 37370 }, { "grad_norm": 0.16285601258277893, "learning_rate": 5.949287971500692e-05, "loss": 0.0086, "step": 37380 }, { "grad_norm": 0.08662686496973038, "learning_rate": 5.947258645565762e-05, "loss": 0.0102, "step": 37390 }, { "grad_norm": 0.14715439081192017, "learning_rate": 5.945229157770237e-05, "loss": 0.0089, "step": 37400 }, { "grad_norm": 0.12303265184164047, "learning_rate": 5.9431995084609006e-05, "loss": 0.0105, "step": 37410 }, { "grad_norm": 0.12778371572494507, "learning_rate": 5.941169697984564e-05, "loss": 0.008, "step": 37420 }, { "grad_norm": 0.16639474034309387, "learning_rate": 5.9391397266880675e-05, "loss": 0.0079, "step": 37430 }, { "grad_norm": 0.11227113753557205, "learning_rate": 5.937109594918279e-05, "loss": 0.0119, "step": 37440 }, { "grad_norm": 0.19009192287921906, "learning_rate": 5.9350793030220884e-05, "loss": 0.0095, "step": 37450 }, { "grad_norm": 0.14439179003238678, "learning_rate": 5.933048851346421e-05, "loss": 0.0109, "step": 37460 }, { "grad_norm": 0.1541055142879486, "learning_rate": 5.931018240238224e-05, "loss": 0.0113, "step": 37470 }, { "grad_norm": 0.18614618480205536, "learning_rate": 5.928987470044471e-05, "loss": 0.0088, "step": 37480 }, { "grad_norm": 0.16060051321983337, "learning_rate": 5.9269565411121695e-05, "loss": 0.0091, "step": 37490 }, { "grad_norm": 0.14087951183319092, "learning_rate": 5.924925453788347e-05, "loss": 0.0097, "step": 37500 }, { "grad_norm": 0.09897221624851227, "learning_rate": 5.92289420842006e-05, "loss": 0.008, "step": 37510 }, { "grad_norm": 0.0998958945274353, "learning_rate": 5.9208628053543945e-05, "loss": 0.0109, "step": 37520 }, { "grad_norm": 0.16085225343704224, "learning_rate": 5.918831244938462e-05, "loss": 0.009, "step": 37530 }, { "grad_norm": 0.17187857627868652, "learning_rate": 5.9167995275194e-05, "loss": 0.0086, "step": 37540 }, { "grad_norm": 0.10537808388471603, "learning_rate": 5.914767653444373e-05, "loss": 0.0086, "step": 37550 }, { "grad_norm": 0.09012789279222488, "learning_rate": 5.912735623060572e-05, "loss": 0.0088, "step": 37560 }, { "grad_norm": 0.1511910855770111, "learning_rate": 5.910703436715217e-05, "loss": 0.0108, "step": 37570 }, { "grad_norm": 0.15326350927352905, "learning_rate": 5.908671094755552e-05, "loss": 0.0085, "step": 37580 }, { "grad_norm": 0.16328464448451996, "learning_rate": 5.906638597528851e-05, "loss": 0.0078, "step": 37590 }, { "grad_norm": 0.1304885894060135, "learning_rate": 5.9046059453824076e-05, "loss": 0.0062, "step": 37600 }, { "grad_norm": 0.14218537509441376, "learning_rate": 5.9025731386635505e-05, "loss": 0.0087, "step": 37610 }, { "grad_norm": 0.12056908756494522, "learning_rate": 5.900540177719629e-05, "loss": 0.0067, "step": 37620 }, { "grad_norm": 0.13604117929935455, "learning_rate": 5.898507062898021e-05, "loss": 0.007, "step": 37630 }, { "grad_norm": 0.13047078251838684, "learning_rate": 5.8964737945461316e-05, "loss": 0.0074, "step": 37640 }, { "grad_norm": 0.12004921585321426, "learning_rate": 5.8944403730113885e-05, "loss": 0.0102, "step": 37650 }, { "grad_norm": 0.09621186554431915, "learning_rate": 5.892406798641248e-05, "loss": 0.0102, "step": 37660 }, { "grad_norm": 0.14175713062286377, "learning_rate": 5.890373071783193e-05, "loss": 0.0101, "step": 37670 }, { "grad_norm": 0.15461236238479614, "learning_rate": 5.888339192784732e-05, "loss": 0.0125, "step": 37680 }, { "grad_norm": 0.10548959672451019, "learning_rate": 5.8863051619934003e-05, "loss": 0.0099, "step": 37690 }, { "grad_norm": 0.08939266949892044, "learning_rate": 5.8842709797567554e-05, "loss": 0.0073, "step": 37700 }, { "grad_norm": 0.11331529915332794, "learning_rate": 5.8822366464223855e-05, "loss": 0.0079, "step": 37710 }, { "grad_norm": 0.1008916050195694, "learning_rate": 5.880202162337901e-05, "loss": 0.0076, "step": 37720 }, { "grad_norm": 0.12318426370620728, "learning_rate": 5.8781675278509405e-05, "loss": 0.0072, "step": 37730 }, { "grad_norm": 0.16664399206638336, "learning_rate": 5.8761327433091696e-05, "loss": 0.0086, "step": 37740 }, { "grad_norm": 0.10037737339735031, "learning_rate": 5.874097809060275e-05, "loss": 0.0087, "step": 37750 }, { "grad_norm": 0.08879107981920242, "learning_rate": 5.87206272545197e-05, "loss": 0.0081, "step": 37760 }, { "grad_norm": 0.14418308436870575, "learning_rate": 5.8700274928319955e-05, "loss": 0.007, "step": 37770 }, { "grad_norm": 0.12458420544862747, "learning_rate": 5.867992111548118e-05, "loss": 0.0102, "step": 37780 }, { "grad_norm": 0.13569511473178864, "learning_rate": 5.865956581948131e-05, "loss": 0.0066, "step": 37790 }, { "grad_norm": 0.11849694699048996, "learning_rate": 5.863920904379845e-05, "loss": 0.0075, "step": 37800 }, { "grad_norm": 0.11545602232217789, "learning_rate": 5.8618850791911064e-05, "loss": 0.0089, "step": 37810 }, { "grad_norm": 0.12054091691970825, "learning_rate": 5.859849106729779e-05, "loss": 0.0083, "step": 37820 }, { "grad_norm": 0.14133670926094055, "learning_rate": 5.857812987343758e-05, "loss": 0.0067, "step": 37830 }, { "grad_norm": 0.15441963076591492, "learning_rate": 5.855776721380957e-05, "loss": 0.0085, "step": 37840 }, { "grad_norm": 0.14087527990341187, "learning_rate": 5.8537403091893217e-05, "loss": 0.01, "step": 37850 }, { "grad_norm": 0.11868144571781158, "learning_rate": 5.851703751116816e-05, "loss": 0.0078, "step": 37860 }, { "grad_norm": 0.13666386902332306, "learning_rate": 5.8496670475114336e-05, "loss": 0.0072, "step": 37870 }, { "grad_norm": 0.09076328575611115, "learning_rate": 5.84763019872119e-05, "loss": 0.0081, "step": 37880 }, { "grad_norm": 0.10146061331033707, "learning_rate": 5.845593205094131e-05, "loss": 0.0071, "step": 37890 }, { "grad_norm": 0.10343500226736069, "learning_rate": 5.843556066978318e-05, "loss": 0.0061, "step": 37900 }, { "grad_norm": 0.12325108051300049, "learning_rate": 5.8415187847218455e-05, "loss": 0.0071, "step": 37910 }, { "grad_norm": 0.1617833524942398, "learning_rate": 5.839481358672827e-05, "loss": 0.0081, "step": 37920 }, { "grad_norm": 0.11655310541391373, "learning_rate": 5.837443789179407e-05, "loss": 0.0081, "step": 37930 }, { "grad_norm": 0.1138712465763092, "learning_rate": 5.8354060765897445e-05, "loss": 0.0071, "step": 37940 }, { "grad_norm": 0.14062096178531647, "learning_rate": 5.8333682212520334e-05, "loss": 0.0073, "step": 37950 }, { "grad_norm": 0.10983752459287643, "learning_rate": 5.831330223514486e-05, "loss": 0.0089, "step": 37960 }, { "grad_norm": 0.08731916546821594, "learning_rate": 5.8292920837253396e-05, "loss": 0.0091, "step": 37970 }, { "grad_norm": 0.16212496161460876, "learning_rate": 5.827253802232857e-05, "loss": 0.0091, "step": 37980 }, { "grad_norm": 0.10819923132658005, "learning_rate": 5.825215379385327e-05, "loss": 0.0066, "step": 37990 }, { "grad_norm": 0.19040760397911072, "learning_rate": 5.823176815531057e-05, "loss": 0.0116, "step": 38000 }, { "grad_norm": 0.15229150652885437, "learning_rate": 5.8211381110183826e-05, "loss": 0.0089, "step": 38010 }, { "grad_norm": 0.16162270307540894, "learning_rate": 5.8190992661956645e-05, "loss": 0.009, "step": 38020 }, { "grad_norm": 0.11785849183797836, "learning_rate": 5.817060281411284e-05, "loss": 0.0057, "step": 38030 }, { "grad_norm": 0.10903224349021912, "learning_rate": 5.815021157013647e-05, "loss": 0.0071, "step": 38040 }, { "grad_norm": 0.15818645060062408, "learning_rate": 5.8129818933511856e-05, "loss": 0.0089, "step": 38050 }, { "grad_norm": 0.10721064358949661, "learning_rate": 5.8109424907723544e-05, "loss": 0.0076, "step": 38060 }, { "grad_norm": 0.08911380171775818, "learning_rate": 5.80890294962563e-05, "loss": 0.0065, "step": 38070 }, { "grad_norm": 0.12134923040866852, "learning_rate": 5.806863270259515e-05, "loss": 0.0079, "step": 38080 }, { "grad_norm": 0.1243322417140007, "learning_rate": 5.804823453022536e-05, "loss": 0.0104, "step": 38090 }, { "grad_norm": 0.1495492309331894, "learning_rate": 5.80278349826324e-05, "loss": 0.0104, "step": 38100 }, { "grad_norm": 0.09700765460729599, "learning_rate": 5.8007434063302014e-05, "loss": 0.0081, "step": 38110 }, { "grad_norm": 0.14584951102733612, "learning_rate": 5.7987031775720136e-05, "loss": 0.0085, "step": 38120 }, { "grad_norm": 0.21376368403434753, "learning_rate": 5.7966628123372976e-05, "loss": 0.0099, "step": 38130 }, { "grad_norm": 0.14489561319351196, "learning_rate": 5.7946223109746956e-05, "loss": 0.0059, "step": 38140 }, { "grad_norm": 0.14646723866462708, "learning_rate": 5.7925816738328754e-05, "loss": 0.0072, "step": 38150 }, { "grad_norm": 0.14656557142734528, "learning_rate": 5.790540901260521e-05, "loss": 0.0091, "step": 38160 }, { "grad_norm": 0.12667915225028992, "learning_rate": 5.788499993606351e-05, "loss": 0.0111, "step": 38170 }, { "grad_norm": 0.11782146245241165, "learning_rate": 5.786458951219096e-05, "loss": 0.0069, "step": 38180 }, { "grad_norm": 0.10613599419593811, "learning_rate": 5.784417774447517e-05, "loss": 0.0086, "step": 38190 }, { "grad_norm": 0.145681232213974, "learning_rate": 5.782376463640393e-05, "loss": 0.0076, "step": 38200 }, { "grad_norm": 0.1189889907836914, "learning_rate": 5.780335019146531e-05, "loss": 0.0081, "step": 38210 }, { "grad_norm": 0.14446274936199188, "learning_rate": 5.778293441314755e-05, "loss": 0.0078, "step": 38220 }, { "grad_norm": 0.12158843129873276, "learning_rate": 5.776251730493917e-05, "loss": 0.0095, "step": 38230 }, { "grad_norm": 0.13937588036060333, "learning_rate": 5.774209887032887e-05, "loss": 0.0077, "step": 38240 }, { "grad_norm": 0.1536688655614853, "learning_rate": 5.772167911280565e-05, "loss": 0.0096, "step": 38250 }, { "grad_norm": 0.13420633971691132, "learning_rate": 5.770125803585864e-05, "loss": 0.0088, "step": 38260 }, { "grad_norm": 0.10775334388017654, "learning_rate": 5.768083564297726e-05, "loss": 0.0079, "step": 38270 }, { "grad_norm": 0.150953009724617, "learning_rate": 5.766041193765114e-05, "loss": 0.0083, "step": 38280 }, { "grad_norm": 0.1292877346277237, "learning_rate": 5.763998692337015e-05, "loss": 0.0081, "step": 38290 }, { "grad_norm": 0.13987629115581512, "learning_rate": 5.761956060362433e-05, "loss": 0.008, "step": 38300 }, { "grad_norm": 0.11851822584867477, "learning_rate": 5.7599132981904e-05, "loss": 0.0083, "step": 38310 }, { "grad_norm": 0.13494737446308136, "learning_rate": 5.75787040616997e-05, "loss": 0.008, "step": 38320 }, { "grad_norm": 0.14906755089759827, "learning_rate": 5.755827384650212e-05, "loss": 0.0099, "step": 38330 }, { "grad_norm": 0.13034789264202118, "learning_rate": 5.753784233980228e-05, "loss": 0.0119, "step": 38340 }, { "grad_norm": 0.1572135090827942, "learning_rate": 5.751740954509135e-05, "loss": 0.0085, "step": 38350 }, { "grad_norm": 0.1252157837152481, "learning_rate": 5.7496975465860715e-05, "loss": 0.0077, "step": 38360 }, { "grad_norm": 0.18310494720935822, "learning_rate": 5.747654010560202e-05, "loss": 0.0102, "step": 38370 }, { "grad_norm": 0.1462484747171402, "learning_rate": 5.7456103467807097e-05, "loss": 0.0088, "step": 38380 }, { "grad_norm": 0.14899590611457825, "learning_rate": 5.7435665555968046e-05, "loss": 0.0072, "step": 38390 }, { "grad_norm": 0.1275605708360672, "learning_rate": 5.74152263735771e-05, "loss": 0.0076, "step": 38400 }, { "grad_norm": 0.14564917981624603, "learning_rate": 5.739478592412677e-05, "loss": 0.0112, "step": 38410 }, { "grad_norm": 0.10581491887569427, "learning_rate": 5.7374344211109766e-05, "loss": 0.0096, "step": 38420 }, { "grad_norm": 0.14637978374958038, "learning_rate": 5.735390123801904e-05, "loss": 0.0088, "step": 38430 }, { "grad_norm": 0.13275489211082458, "learning_rate": 5.7333457008347704e-05, "loss": 0.0077, "step": 38440 }, { "grad_norm": 0.1455148607492447, "learning_rate": 5.7313011525589156e-05, "loss": 0.0081, "step": 38450 }, { "grad_norm": 0.11156168580055237, "learning_rate": 5.729256479323694e-05, "loss": 0.0095, "step": 38460 }, { "grad_norm": 0.11924037337303162, "learning_rate": 5.727211681478485e-05, "loss": 0.0085, "step": 38470 }, { "grad_norm": 0.11956571042537689, "learning_rate": 5.7251667593726886e-05, "loss": 0.0077, "step": 38480 }, { "grad_norm": 0.13100354373455048, "learning_rate": 5.723121713355728e-05, "loss": 0.0088, "step": 38490 }, { "grad_norm": 0.09267068654298782, "learning_rate": 5.721076543777044e-05, "loss": 0.0099, "step": 38500 }, { "grad_norm": 0.13640514016151428, "learning_rate": 5.7190312509860986e-05, "loss": 0.011, "step": 38510 }, { "grad_norm": 0.16515544056892395, "learning_rate": 5.716985835332379e-05, "loss": 0.0072, "step": 38520 }, { "grad_norm": 0.1413503885269165, "learning_rate": 5.714940297165389e-05, "loss": 0.0078, "step": 38530 }, { "grad_norm": 0.1503511220216751, "learning_rate": 5.712894636834656e-05, "loss": 0.0095, "step": 38540 }, { "grad_norm": 0.14185267686843872, "learning_rate": 5.7108488546897276e-05, "loss": 0.0109, "step": 38550 }, { "grad_norm": 0.11283903568983078, "learning_rate": 5.708802951080172e-05, "loss": 0.0079, "step": 38560 }, { "grad_norm": 0.10645119845867157, "learning_rate": 5.706756926355576e-05, "loss": 0.0092, "step": 38570 }, { "grad_norm": 0.12715721130371094, "learning_rate": 5.704710780865554e-05, "loss": 0.0082, "step": 38580 }, { "grad_norm": 0.16293399035930634, "learning_rate": 5.7026645149597325e-05, "loss": 0.0094, "step": 38590 }, { "grad_norm": 0.10009501129388809, "learning_rate": 5.700618128987764e-05, "loss": 0.009, "step": 38600 }, { "grad_norm": 0.15394535660743713, "learning_rate": 5.698571623299317e-05, "loss": 0.009, "step": 38610 }, { "grad_norm": 0.15446050465106964, "learning_rate": 5.696524998244086e-05, "loss": 0.0113, "step": 38620 }, { "grad_norm": 0.14488384127616882, "learning_rate": 5.6944782541717836e-05, "loss": 0.0082, "step": 38630 }, { "grad_norm": 0.15350526571273804, "learning_rate": 5.69243139143214e-05, "loss": 0.009, "step": 38640 }, { "grad_norm": 0.12139123678207397, "learning_rate": 5.6903844103749125e-05, "loss": 0.0086, "step": 38650 }, { "grad_norm": 0.15710584819316864, "learning_rate": 5.688337311349869e-05, "loss": 0.0082, "step": 38660 }, { "grad_norm": 0.1292770355939865, "learning_rate": 5.6862900947068074e-05, "loss": 0.0065, "step": 38670 }, { "grad_norm": 0.2919643223285675, "learning_rate": 5.6842427607955374e-05, "loss": 0.0084, "step": 38680 }, { "grad_norm": 0.10253249853849411, "learning_rate": 5.682195309965893e-05, "loss": 0.0077, "step": 38690 }, { "grad_norm": 0.10271058231592178, "learning_rate": 5.6801477425677294e-05, "loss": 0.007, "step": 38700 }, { "grad_norm": 0.1287459284067154, "learning_rate": 5.678100058950917e-05, "loss": 0.0084, "step": 38710 }, { "grad_norm": 0.09275763481855392, "learning_rate": 5.676052259465352e-05, "loss": 0.0085, "step": 38720 }, { "grad_norm": 0.1535203456878662, "learning_rate": 5.674004344460945e-05, "loss": 0.0092, "step": 38730 }, { "grad_norm": 0.11067065596580505, "learning_rate": 5.6719563142876295e-05, "loss": 0.009, "step": 38740 }, { "grad_norm": 0.17359969019889832, "learning_rate": 5.669908169295359e-05, "loss": 0.0115, "step": 38750 }, { "grad_norm": 0.10886511951684952, "learning_rate": 5.667859909834105e-05, "loss": 0.0073, "step": 38760 }, { "grad_norm": 0.13443799316883087, "learning_rate": 5.6658115362538565e-05, "loss": 0.009, "step": 38770 }, { "grad_norm": 0.14069969952106476, "learning_rate": 5.663763048904628e-05, "loss": 0.0084, "step": 38780 }, { "grad_norm": 0.13734732568264008, "learning_rate": 5.661714448136447e-05, "loss": 0.0076, "step": 38790 }, { "grad_norm": 0.09060356765985489, "learning_rate": 5.659665734299366e-05, "loss": 0.0084, "step": 38800 }, { "grad_norm": 0.12817487120628357, "learning_rate": 5.6576169077434516e-05, "loss": 0.009, "step": 38810 }, { "grad_norm": 0.10469669103622437, "learning_rate": 5.6555679688187944e-05, "loss": 0.009, "step": 38820 }, { "grad_norm": 0.18151813745498657, "learning_rate": 5.6535189178755e-05, "loss": 0.0089, "step": 38830 }, { "grad_norm": 0.11522621661424637, "learning_rate": 5.651469755263698e-05, "loss": 0.0082, "step": 38840 }, { "grad_norm": 0.12134721130132675, "learning_rate": 5.6494204813335316e-05, "loss": 0.0093, "step": 38850 }, { "grad_norm": 0.12419024109840393, "learning_rate": 5.647371096435168e-05, "loss": 0.0111, "step": 38860 }, { "grad_norm": 0.1918681114912033, "learning_rate": 5.645321600918788e-05, "loss": 0.0089, "step": 38870 }, { "grad_norm": 0.17680814862251282, "learning_rate": 5.643271995134597e-05, "loss": 0.0089, "step": 38880 }, { "grad_norm": 0.15578807890415192, "learning_rate": 5.641222279432814e-05, "loss": 0.0087, "step": 38890 }, { "grad_norm": 0.11686801165342331, "learning_rate": 5.6391724541636834e-05, "loss": 0.0107, "step": 38900 }, { "grad_norm": 0.14157019555568695, "learning_rate": 5.6371225196774605e-05, "loss": 0.0088, "step": 38910 }, { "grad_norm": 0.15304018557071686, "learning_rate": 5.635072476324423e-05, "loss": 0.0083, "step": 38920 }, { "grad_norm": 0.0972244068980217, "learning_rate": 5.63302232445487e-05, "loss": 0.0065, "step": 38930 }, { "grad_norm": 0.10337643325328827, "learning_rate": 5.6309720644191144e-05, "loss": 0.0074, "step": 38940 }, { "grad_norm": 0.14854757487773895, "learning_rate": 5.628921696567491e-05, "loss": 0.0101, "step": 38950 }, { "grad_norm": 0.13475851714611053, "learning_rate": 5.62687122125035e-05, "loss": 0.0078, "step": 38960 }, { "grad_norm": 0.1615505963563919, "learning_rate": 5.624820638818062e-05, "loss": 0.0083, "step": 38970 }, { "grad_norm": 0.14421729743480682, "learning_rate": 5.6227699496210164e-05, "loss": 0.0086, "step": 38980 }, { "grad_norm": 0.09995554387569427, "learning_rate": 5.6207191540096195e-05, "loss": 0.0086, "step": 38990 }, { "grad_norm": 0.1661364734172821, "learning_rate": 5.618668252334296e-05, "loss": 0.008, "step": 39000 }, { "grad_norm": 0.12121599912643433, "learning_rate": 5.616617244945488e-05, "loss": 0.0092, "step": 39010 }, { "grad_norm": 0.12706251442432404, "learning_rate": 5.614566132193656e-05, "loss": 0.0074, "step": 39020 }, { "grad_norm": 0.09676417708396912, "learning_rate": 5.612514914429282e-05, "loss": 0.0086, "step": 39030 }, { "grad_norm": 0.15585105121135712, "learning_rate": 5.610463592002863e-05, "loss": 0.0094, "step": 39040 }, { "grad_norm": 0.12207882851362228, "learning_rate": 5.608412165264909e-05, "loss": 0.0073, "step": 39050 }, { "grad_norm": 0.10982402414083481, "learning_rate": 5.606360634565959e-05, "loss": 0.0095, "step": 39060 }, { "grad_norm": 0.16379836201667786, "learning_rate": 5.604309000256559e-05, "loss": 0.0091, "step": 39070 }, { "grad_norm": 0.18224942684173584, "learning_rate": 5.6022572626872785e-05, "loss": 0.0069, "step": 39080 }, { "grad_norm": 0.14626306295394897, "learning_rate": 5.600205422208704e-05, "loss": 0.0087, "step": 39090 }, { "grad_norm": 0.11683216691017151, "learning_rate": 5.59815347917144e-05, "loss": 0.009, "step": 39100 }, { "grad_norm": 0.14130912721157074, "learning_rate": 5.596101433926103e-05, "loss": 0.0067, "step": 39110 }, { "grad_norm": 0.15172576904296875, "learning_rate": 5.5940492868233364e-05, "loss": 0.011, "step": 39120 }, { "grad_norm": 0.1113307848572731, "learning_rate": 5.591997038213793e-05, "loss": 0.0109, "step": 39130 }, { "grad_norm": 0.1196531131863594, "learning_rate": 5.5899446884481475e-05, "loss": 0.0087, "step": 39140 }, { "grad_norm": 0.08722897619009018, "learning_rate": 5.5878922378770906e-05, "loss": 0.0074, "step": 39150 }, { "grad_norm": 0.1512221097946167, "learning_rate": 5.5858396868513285e-05, "loss": 0.0076, "step": 39160 }, { "grad_norm": 0.08866408467292786, "learning_rate": 5.583787035721586e-05, "loss": 0.0074, "step": 39170 }, { "grad_norm": 0.12181077897548676, "learning_rate": 5.581734284838606e-05, "loss": 0.0071, "step": 39180 }, { "grad_norm": 0.13920755684375763, "learning_rate": 5.579681434553147e-05, "loss": 0.0102, "step": 39190 }, { "grad_norm": 0.17291226983070374, "learning_rate": 5.5776284852159854e-05, "loss": 0.0083, "step": 39200 }, { "grad_norm": 0.1738319844007492, "learning_rate": 5.575575437177913e-05, "loss": 0.0093, "step": 39210 }, { "grad_norm": 0.12461821734905243, "learning_rate": 5.573522290789742e-05, "loss": 0.0095, "step": 39220 }, { "grad_norm": 0.1594351828098297, "learning_rate": 5.571469046402298e-05, "loss": 0.0084, "step": 39230 }, { "grad_norm": 0.22578176856040955, "learning_rate": 5.5694157043664205e-05, "loss": 0.0132, "step": 39240 }, { "grad_norm": 0.10471714287996292, "learning_rate": 5.567362265032975e-05, "loss": 0.0065, "step": 39250 }, { "grad_norm": 0.1140238344669342, "learning_rate": 5.565308728752836e-05, "loss": 0.0069, "step": 39260 }, { "grad_norm": 0.15281368792057037, "learning_rate": 5.5632550958768956e-05, "loss": 0.007, "step": 39270 }, { "grad_norm": 0.11702851206064224, "learning_rate": 5.5612013667560636e-05, "loss": 0.0099, "step": 39280 }, { "grad_norm": 0.09266287088394165, "learning_rate": 5.5591475417412676e-05, "loss": 0.0107, "step": 39290 }, { "grad_norm": 0.14222724735736847, "learning_rate": 5.557093621183451e-05, "loss": 0.0076, "step": 39300 }, { "grad_norm": 0.11000329256057739, "learning_rate": 5.55503960543357e-05, "loss": 0.0084, "step": 39310 }, { "grad_norm": 0.18537607789039612, "learning_rate": 5.552985494842601e-05, "loss": 0.009, "step": 39320 }, { "grad_norm": 0.1154063269495964, "learning_rate": 5.550931289761534e-05, "loss": 0.0093, "step": 39330 }, { "grad_norm": 0.10472183674573898, "learning_rate": 5.548876990541378e-05, "loss": 0.0071, "step": 39340 }, { "grad_norm": 0.15944752097129822, "learning_rate": 5.5468225975331564e-05, "loss": 0.0107, "step": 39350 }, { "grad_norm": 0.14817097783088684, "learning_rate": 5.544768111087909e-05, "loss": 0.0093, "step": 39360 }, { "grad_norm": 0.1491386443376541, "learning_rate": 5.54271353155669e-05, "loss": 0.0073, "step": 39370 }, { "grad_norm": 0.1428069919347763, "learning_rate": 5.5406588592905715e-05, "loss": 0.0082, "step": 39380 }, { "grad_norm": 0.09366456419229507, "learning_rate": 5.5386040946406416e-05, "loss": 0.0078, "step": 39390 }, { "grad_norm": 0.11704188585281372, "learning_rate": 5.536549237958004e-05, "loss": 0.0072, "step": 39400 }, { "grad_norm": 0.14049704372882843, "learning_rate": 5.5344942895937744e-05, "loss": 0.0075, "step": 39410 }, { "grad_norm": 0.12625138461589813, "learning_rate": 5.5324392498990904e-05, "loss": 0.0081, "step": 39420 }, { "grad_norm": 0.14307363331317902, "learning_rate": 5.5303841192251e-05, "loss": 0.0094, "step": 39430 }, { "grad_norm": 0.13702213764190674, "learning_rate": 5.52832889792297e-05, "loss": 0.0092, "step": 39440 }, { "grad_norm": 0.06995740532875061, "learning_rate": 5.526273586343881e-05, "loss": 0.008, "step": 39450 }, { "grad_norm": 0.14960017800331116, "learning_rate": 5.5242181848390306e-05, "loss": 0.0081, "step": 39460 }, { "grad_norm": 0.08044662326574326, "learning_rate": 5.5221626937596285e-05, "loss": 0.0081, "step": 39470 }, { "grad_norm": 0.11051976680755615, "learning_rate": 5.520107113456903e-05, "loss": 0.0068, "step": 39480 }, { "grad_norm": 0.12491894513368607, "learning_rate": 5.5180514442820974e-05, "loss": 0.0081, "step": 39490 }, { "grad_norm": 0.12755338847637177, "learning_rate": 5.515995686586469e-05, "loss": 0.007, "step": 39500 }, { "grad_norm": 0.18742917478084564, "learning_rate": 5.5139398407212916e-05, "loss": 0.0121, "step": 39510 }, { "grad_norm": 0.13360752165317535, "learning_rate": 5.511883907037849e-05, "loss": 0.0093, "step": 39520 }, { "grad_norm": 0.11160749197006226, "learning_rate": 5.509827885887449e-05, "loss": 0.007, "step": 39530 }, { "grad_norm": 0.13374558091163635, "learning_rate": 5.507771777621406e-05, "loss": 0.0079, "step": 39540 }, { "grad_norm": 0.0866677314043045, "learning_rate": 5.505715582591052e-05, "loss": 0.0071, "step": 39550 }, { "grad_norm": 0.08887866139411926, "learning_rate": 5.50365930114774e-05, "loss": 0.0062, "step": 39560 }, { "grad_norm": 0.121973417699337, "learning_rate": 5.5016029336428255e-05, "loss": 0.0067, "step": 39570 }, { "grad_norm": 0.16563449800014496, "learning_rate": 5.49954648042769e-05, "loss": 0.0086, "step": 39580 }, { "grad_norm": 0.1367083191871643, "learning_rate": 5.4974899418537226e-05, "loss": 0.0087, "step": 39590 }, { "grad_norm": 0.1579822301864624, "learning_rate": 5.4954333182723316e-05, "loss": 0.0068, "step": 39600 }, { "grad_norm": 0.12907040119171143, "learning_rate": 5.493376610034937e-05, "loss": 0.0083, "step": 39610 }, { "grad_norm": 0.17620742321014404, "learning_rate": 5.4913198174929735e-05, "loss": 0.0063, "step": 39620 }, { "grad_norm": 0.1135924682021141, "learning_rate": 5.489262940997889e-05, "loss": 0.0075, "step": 39630 }, { "grad_norm": 0.13297568261623383, "learning_rate": 5.487205980901151e-05, "loss": 0.0069, "step": 39640 }, { "grad_norm": 0.13931889832019806, "learning_rate": 5.485148937554234e-05, "loss": 0.0063, "step": 39650 }, { "grad_norm": 0.15229706466197968, "learning_rate": 5.483091811308635e-05, "loss": 0.0078, "step": 39660 }, { "grad_norm": 0.13542114198207855, "learning_rate": 5.4810346025158564e-05, "loss": 0.008, "step": 39670 }, { "grad_norm": 0.11902592331171036, "learning_rate": 5.478977311527421e-05, "loss": 0.0085, "step": 39680 }, { "grad_norm": 0.10025172680616379, "learning_rate": 5.476919938694863e-05, "loss": 0.0073, "step": 39690 }, { "grad_norm": 0.11493280529975891, "learning_rate": 5.474862484369733e-05, "loss": 0.0086, "step": 39700 }, { "grad_norm": 0.14116807281970978, "learning_rate": 5.472804948903589e-05, "loss": 0.009, "step": 39710 }, { "grad_norm": 0.15023835003376007, "learning_rate": 5.470747332648013e-05, "loss": 0.0079, "step": 39720 }, { "grad_norm": 0.16408918797969818, "learning_rate": 5.468689635954591e-05, "loss": 0.0089, "step": 39730 }, { "grad_norm": 0.13092447817325592, "learning_rate": 5.46663185917493e-05, "loss": 0.0112, "step": 39740 }, { "grad_norm": 0.16822315752506256, "learning_rate": 5.464574002660645e-05, "loss": 0.0067, "step": 39750 }, { "grad_norm": 0.20103803277015686, "learning_rate": 5.4625160667633724e-05, "loss": 0.0088, "step": 39760 }, { "grad_norm": 0.15128310024738312, "learning_rate": 5.4604580518347505e-05, "loss": 0.008, "step": 39770 }, { "grad_norm": 0.0950423926115036, "learning_rate": 5.458399958226442e-05, "loss": 0.0076, "step": 39780 }, { "grad_norm": 0.1291119009256363, "learning_rate": 5.456341786290119e-05, "loss": 0.01, "step": 39790 }, { "grad_norm": 0.12252869457006454, "learning_rate": 5.454283536377465e-05, "loss": 0.0085, "step": 39800 }, { "grad_norm": 0.08450193703174591, "learning_rate": 5.452225208840179e-05, "loss": 0.0087, "step": 39810 }, { "grad_norm": 0.12146016955375671, "learning_rate": 5.450166804029975e-05, "loss": 0.0087, "step": 39820 }, { "grad_norm": 0.1358206868171692, "learning_rate": 5.448108322298574e-05, "loss": 0.0081, "step": 39830 }, { "grad_norm": 0.19542506337165833, "learning_rate": 5.446049763997717e-05, "loss": 0.0087, "step": 39840 }, { "grad_norm": 0.14479701220989227, "learning_rate": 5.4439911294791546e-05, "loss": 0.0101, "step": 39850 }, { "grad_norm": 0.11676144599914551, "learning_rate": 5.441932419094652e-05, "loss": 0.0088, "step": 39860 }, { "grad_norm": 0.11040101945400238, "learning_rate": 5.439873633195985e-05, "loss": 0.0082, "step": 39870 }, { "grad_norm": 0.13960058987140656, "learning_rate": 5.437814772134947e-05, "loss": 0.0073, "step": 39880 }, { "grad_norm": 0.15142422914505005, "learning_rate": 5.4357558362633366e-05, "loss": 0.0106, "step": 39890 }, { "grad_norm": 0.12508119642734528, "learning_rate": 5.4336968259329726e-05, "loss": 0.0099, "step": 39900 }, { "grad_norm": 0.12421046197414398, "learning_rate": 5.431637741495681e-05, "loss": 0.0079, "step": 39910 }, { "grad_norm": 0.1069563776254654, "learning_rate": 5.429578583303307e-05, "loss": 0.0092, "step": 39920 }, { "grad_norm": 0.15531639754772186, "learning_rate": 5.427519351707701e-05, "loss": 0.0104, "step": 39930 }, { "grad_norm": 0.11766452342271805, "learning_rate": 5.42546004706073e-05, "loss": 0.0075, "step": 39940 }, { "grad_norm": 0.10544676333665848, "learning_rate": 5.4234006697142735e-05, "loss": 0.0065, "step": 39950 }, { "grad_norm": 0.1313561648130417, "learning_rate": 5.421341220020224e-05, "loss": 0.0087, "step": 39960 }, { "grad_norm": 0.09619082510471344, "learning_rate": 5.419281698330482e-05, "loss": 0.0065, "step": 39970 }, { "grad_norm": 0.10002286732196808, "learning_rate": 5.4172221049969665e-05, "loss": 0.0075, "step": 39980 }, { "grad_norm": 0.10516496747732162, "learning_rate": 5.415162440371604e-05, "loss": 0.0063, "step": 39990 }, { "grad_norm": 0.09158113598823547, "learning_rate": 5.413102704806334e-05, "loss": 0.0065, "step": 40000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }