{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.8581995368003845, "learning_rate": 2.25e-07, "loss": 1.4991, "step": 10 }, { "grad_norm": 0.9741699695587158, "learning_rate": 4.75e-07, "loss": 1.545, "step": 20 }, { "grad_norm": 1.0240248441696167, "learning_rate": 7.25e-07, "loss": 1.4918, "step": 30 }, { "grad_norm": 0.7832372188568115, "learning_rate": 9.75e-07, "loss": 1.4685, "step": 40 }, { "grad_norm": 0.8714929819107056, "learning_rate": 1.2250000000000001e-06, "loss": 1.4923, "step": 50 }, { "grad_norm": 1.361850619316101, "learning_rate": 1.475e-06, "loss": 1.4936, "step": 60 }, { "grad_norm": 2.3495960235595703, "learning_rate": 1.7250000000000002e-06, "loss": 1.3747, "step": 70 }, { "grad_norm": 2.2871360778808594, "learning_rate": 1.975e-06, "loss": 1.3452, "step": 80 }, { "grad_norm": 1.6665167808532715, "learning_rate": 2.225e-06, "loss": 1.2959, "step": 90 }, { "grad_norm": 1.6520758867263794, "learning_rate": 2.4750000000000004e-06, "loss": 1.2706, "step": 100 }, { "grad_norm": 1.4286201000213623, "learning_rate": 2.725e-06, "loss": 1.2784, "step": 110 }, { "grad_norm": 1.6470242738723755, "learning_rate": 2.975e-06, "loss": 1.2473, "step": 120 }, { "grad_norm": 1.6853891611099243, "learning_rate": 3.225e-06, "loss": 1.1917, "step": 130 }, { "grad_norm": 1.4710819721221924, "learning_rate": 3.4750000000000006e-06, "loss": 1.2049, "step": 140 }, { "grad_norm": 1.3664950132369995, "learning_rate": 3.725e-06, "loss": 1.195, "step": 150 }, { "grad_norm": 2.6374003887176514, "learning_rate": 3.975e-06, "loss": 1.1601, "step": 160 }, { "grad_norm": 2.030930280685425, "learning_rate": 4.225e-06, "loss": 1.1708, "step": 170 }, { "grad_norm": 2.4633688926696777, "learning_rate": 4.475e-06, "loss": 1.2153, "step": 180 }, { "grad_norm": 2.2116878032684326, "learning_rate": 4.7250000000000005e-06, "loss": 1.2142, "step": 190 }, { "grad_norm": 2.287712812423706, "learning_rate": 4.975000000000001e-06, "loss": 1.1931, "step": 200 }, { "grad_norm": 1.4614124298095703, "learning_rate": 5.225e-06, "loss": 1.1844, "step": 210 }, { "grad_norm": 1.6014972925186157, "learning_rate": 5.475e-06, "loss": 1.1654, "step": 220 }, { "grad_norm": 1.8665953874588013, "learning_rate": 5.725e-06, "loss": 1.1669, "step": 230 }, { "grad_norm": 1.5550140142440796, "learning_rate": 5.975e-06, "loss": 1.1538, "step": 240 }, { "grad_norm": 2.870995044708252, "learning_rate": 6.2250000000000005e-06, "loss": 1.1252, "step": 250 }, { "grad_norm": 1.557847023010254, "learning_rate": 6.475000000000001e-06, "loss": 1.0965, "step": 260 }, { "grad_norm": 1.8374453783035278, "learning_rate": 6.725000000000001e-06, "loss": 1.1086, "step": 270 }, { "grad_norm": 3.5533909797668457, "learning_rate": 6.975000000000001e-06, "loss": 1.0959, "step": 280 }, { "grad_norm": 2.1160459518432617, "learning_rate": 7.2249999999999994e-06, "loss": 1.1066, "step": 290 }, { "grad_norm": 1.598073124885559, "learning_rate": 7.4750000000000004e-06, "loss": 1.1143, "step": 300 }, { "grad_norm": 2.0287368297576904, "learning_rate": 7.725e-06, "loss": 1.1041, "step": 310 }, { "grad_norm": 4.412517547607422, "learning_rate": 7.975e-06, "loss": 1.1093, "step": 320 }, { "grad_norm": 1.7186756134033203, "learning_rate": 8.225e-06, "loss": 1.1047, "step": 330 }, { "grad_norm": 1.9219340085983276, "learning_rate": 8.475000000000001e-06, "loss": 1.1, "step": 340 }, { "grad_norm": 1.4213459491729736, "learning_rate": 8.725e-06, "loss": 1.0898, "step": 350 }, { "grad_norm": 2.2069225311279297, "learning_rate": 8.975e-06, "loss": 1.0805, "step": 360 }, { "grad_norm": 1.8459502458572388, "learning_rate": 9.225e-06, "loss": 1.083, "step": 370 }, { "grad_norm": 1.590882658958435, "learning_rate": 9.475e-06, "loss": 1.079, "step": 380 }, { "grad_norm": 1.472062587738037, "learning_rate": 9.725000000000001e-06, "loss": 1.0426, "step": 390 }, { "grad_norm": 1.450273871421814, "learning_rate": 9.975e-06, "loss": 1.0962, "step": 400 }, { "grad_norm": 1.4977627992630005, "learning_rate": 1.0225e-05, "loss": 1.0917, "step": 410 }, { "grad_norm": 1.1626904010772705, "learning_rate": 1.0475e-05, "loss": 1.072, "step": 420 }, { "grad_norm": 3.4820690155029297, "learning_rate": 1.0725e-05, "loss": 1.0662, "step": 430 }, { "grad_norm": 2.5945799350738525, "learning_rate": 1.0975e-05, "loss": 1.0821, "step": 440 }, { "grad_norm": 1.3138461112976074, "learning_rate": 1.1225e-05, "loss": 1.0556, "step": 450 }, { "grad_norm": 1.446639895439148, "learning_rate": 1.1475000000000001e-05, "loss": 1.0698, "step": 460 }, { "grad_norm": 1.6813281774520874, "learning_rate": 1.1725e-05, "loss": 1.0729, "step": 470 }, { "grad_norm": 3.213225841522217, "learning_rate": 1.1975e-05, "loss": 1.1095, "step": 480 }, { "grad_norm": 1.514164686203003, "learning_rate": 1.2225e-05, "loss": 1.0706, "step": 490 }, { "grad_norm": 1.5521894693374634, "learning_rate": 1.2475e-05, "loss": 1.0797, "step": 500 }, { "grad_norm": 1.566810965538025, "learning_rate": 1.2725000000000001e-05, "loss": 1.0586, "step": 510 }, { "grad_norm": 2.7550506591796875, "learning_rate": 1.2975e-05, "loss": 1.0678, "step": 520 }, { "grad_norm": 1.2056630849838257, "learning_rate": 1.3225000000000001e-05, "loss": 1.0858, "step": 530 }, { "grad_norm": 4.566923141479492, "learning_rate": 1.3475000000000002e-05, "loss": 1.0656, "step": 540 }, { "grad_norm": 1.4410126209259033, "learning_rate": 1.3725000000000002e-05, "loss": 1.0894, "step": 550 }, { "grad_norm": 1.2167634963989258, "learning_rate": 1.3975000000000003e-05, "loss": 1.0712, "step": 560 }, { "grad_norm": 1.2489228248596191, "learning_rate": 1.4225e-05, "loss": 1.0366, "step": 570 }, { "grad_norm": 1.4533782005310059, "learning_rate": 1.4475e-05, "loss": 1.0204, "step": 580 }, { "grad_norm": 1.7476335763931274, "learning_rate": 1.4725e-05, "loss": 1.0109, "step": 590 }, { "grad_norm": 2.476857900619507, "learning_rate": 1.4975e-05, "loss": 0.9847, "step": 600 }, { "grad_norm": 1.9164401292800903, "learning_rate": 1.5225e-05, "loss": 0.9604, "step": 610 }, { "grad_norm": 2.7636940479278564, "learning_rate": 1.5475e-05, "loss": 0.9299, "step": 620 }, { "grad_norm": 1.7850292921066284, "learning_rate": 1.5725e-05, "loss": 0.9068, "step": 630 }, { "grad_norm": 2.172623634338379, "learning_rate": 1.5975000000000002e-05, "loss": 0.8821, "step": 640 }, { "grad_norm": 2.0404021739959717, "learning_rate": 1.6225e-05, "loss": 0.8275, "step": 650 }, { "grad_norm": 2.9238784313201904, "learning_rate": 1.6475e-05, "loss": 0.8009, "step": 660 }, { "grad_norm": 1.7407655715942383, "learning_rate": 1.6725000000000003e-05, "loss": 0.7462, "step": 670 }, { "grad_norm": 2.3808741569519043, "learning_rate": 1.6975000000000003e-05, "loss": 0.7155, "step": 680 }, { "grad_norm": 2.192972421646118, "learning_rate": 1.7225e-05, "loss": 0.7006, "step": 690 }, { "grad_norm": 3.0234670639038086, "learning_rate": 1.7475e-05, "loss": 0.6922, "step": 700 }, { "grad_norm": 2.395650863647461, "learning_rate": 1.7725e-05, "loss": 0.617, "step": 710 }, { "grad_norm": 2.52492618560791, "learning_rate": 1.7975e-05, "loss": 0.6358, "step": 720 }, { "grad_norm": 2.471536636352539, "learning_rate": 1.8225e-05, "loss": 0.5749, "step": 730 }, { "grad_norm": 1.8052228689193726, "learning_rate": 1.8475000000000002e-05, "loss": 0.5299, "step": 740 }, { "grad_norm": 3.4395830631256104, "learning_rate": 1.8725e-05, "loss": 0.5153, "step": 750 }, { "grad_norm": 1.9210350513458252, "learning_rate": 1.8975e-05, "loss": 0.4808, "step": 760 }, { "grad_norm": 2.479142904281616, "learning_rate": 1.9225e-05, "loss": 0.4595, "step": 770 }, { "grad_norm": 2.4353861808776855, "learning_rate": 1.9475000000000002e-05, "loss": 0.4728, "step": 780 }, { "grad_norm": 2.4107065200805664, "learning_rate": 1.9725000000000002e-05, "loss": 0.3942, "step": 790 }, { "grad_norm": 3.2788171768188477, "learning_rate": 1.9975e-05, "loss": 0.3925, "step": 800 }, { "grad_norm": 2.7130208015441895, "learning_rate": 2.0225000000000004e-05, "loss": 0.3946, "step": 810 }, { "grad_norm": 2.5490829944610596, "learning_rate": 2.0475e-05, "loss": 0.3777, "step": 820 }, { "grad_norm": 2.2808547019958496, "learning_rate": 2.0725e-05, "loss": 0.3773, "step": 830 }, { "grad_norm": 3.305499792098999, "learning_rate": 2.0975e-05, "loss": 0.3512, "step": 840 }, { "grad_norm": 2.089221715927124, "learning_rate": 2.1225e-05, "loss": 0.3607, "step": 850 }, { "grad_norm": 2.7198705673217773, "learning_rate": 2.1475e-05, "loss": 0.3201, "step": 860 }, { "grad_norm": 2.3835701942443848, "learning_rate": 2.1725e-05, "loss": 0.3358, "step": 870 }, { "grad_norm": 2.948763132095337, "learning_rate": 2.1975000000000002e-05, "loss": 0.288, "step": 880 }, { "grad_norm": 2.474575996398926, "learning_rate": 2.2225e-05, "loss": 0.2863, "step": 890 }, { "grad_norm": 1.9832484722137451, "learning_rate": 2.2475e-05, "loss": 0.2933, "step": 900 }, { "grad_norm": 2.3598761558532715, "learning_rate": 2.2725000000000003e-05, "loss": 0.2803, "step": 910 }, { "grad_norm": 2.240010976791382, "learning_rate": 2.2975000000000003e-05, "loss": 0.2304, "step": 920 }, { "grad_norm": 2.513833999633789, "learning_rate": 2.3225000000000002e-05, "loss": 0.2581, "step": 930 }, { "grad_norm": 2.5357398986816406, "learning_rate": 2.3475e-05, "loss": 0.2395, "step": 940 }, { "grad_norm": 2.17702579498291, "learning_rate": 2.3725e-05, "loss": 0.2152, "step": 950 }, { "grad_norm": 1.9739844799041748, "learning_rate": 2.3975e-05, "loss": 0.2161, "step": 960 }, { "grad_norm": 2.116798162460327, "learning_rate": 2.4225e-05, "loss": 0.2431, "step": 970 }, { "grad_norm": 2.0366859436035156, "learning_rate": 2.4475000000000002e-05, "loss": 0.2118, "step": 980 }, { "grad_norm": 2.157320022583008, "learning_rate": 2.4725e-05, "loss": 0.1839, "step": 990 }, { "grad_norm": 2.973156452178955, "learning_rate": 2.4975e-05, "loss": 0.2042, "step": 1000 }, { "grad_norm": 1.8952770233154297, "learning_rate": 2.5225e-05, "loss": 0.1918, "step": 1010 }, { "grad_norm": 2.3106913566589355, "learning_rate": 2.5475e-05, "loss": 0.2303, "step": 1020 }, { "grad_norm": 2.367784023284912, "learning_rate": 2.5725e-05, "loss": 0.1775, "step": 1030 }, { "grad_norm": 1.871354579925537, "learning_rate": 2.5974999999999998e-05, "loss": 0.1992, "step": 1040 }, { "grad_norm": 2.38970947265625, "learning_rate": 2.6225e-05, "loss": 0.1948, "step": 1050 }, { "grad_norm": 1.7859092950820923, "learning_rate": 2.6475e-05, "loss": 0.2111, "step": 1060 }, { "grad_norm": 2.166773557662964, "learning_rate": 2.6725e-05, "loss": 0.2302, "step": 1070 }, { "grad_norm": 2.567538261413574, "learning_rate": 2.6975000000000002e-05, "loss": 0.2153, "step": 1080 }, { "grad_norm": 1.8941320180892944, "learning_rate": 2.7225e-05, "loss": 0.2006, "step": 1090 }, { "grad_norm": 2.1873769760131836, "learning_rate": 2.7475e-05, "loss": 0.1583, "step": 1100 }, { "grad_norm": 1.8602272272109985, "learning_rate": 2.7725e-05, "loss": 0.2347, "step": 1110 }, { "grad_norm": 1.764486312866211, "learning_rate": 2.7975000000000002e-05, "loss": 0.1594, "step": 1120 }, { "grad_norm": 1.598453402519226, "learning_rate": 2.8225e-05, "loss": 0.1795, "step": 1130 }, { "grad_norm": 2.597371816635132, "learning_rate": 2.8475e-05, "loss": 0.2446, "step": 1140 }, { "grad_norm": 2.1234471797943115, "learning_rate": 2.8725e-05, "loss": 0.2165, "step": 1150 }, { "grad_norm": 3.2807137966156006, "learning_rate": 2.8975000000000003e-05, "loss": 0.1932, "step": 1160 }, { "grad_norm": 1.7750866413116455, "learning_rate": 2.9225000000000002e-05, "loss": 0.1858, "step": 1170 }, { "grad_norm": 2.291245937347412, "learning_rate": 2.9475e-05, "loss": 0.1684, "step": 1180 }, { "grad_norm": 2.052006959915161, "learning_rate": 2.9725000000000004e-05, "loss": 0.1812, "step": 1190 }, { "grad_norm": 2.604687452316284, "learning_rate": 2.9975000000000004e-05, "loss": 0.1958, "step": 1200 }, { "grad_norm": 2.7578377723693848, "learning_rate": 3.0225000000000003e-05, "loss": 0.2045, "step": 1210 }, { "grad_norm": 2.1554465293884277, "learning_rate": 3.0475000000000002e-05, "loss": 0.19, "step": 1220 }, { "grad_norm": 2.418527126312256, "learning_rate": 3.0725e-05, "loss": 0.1886, "step": 1230 }, { "grad_norm": 2.2285842895507812, "learning_rate": 3.0975e-05, "loss": 0.199, "step": 1240 }, { "grad_norm": 2.2568185329437256, "learning_rate": 3.122500000000001e-05, "loss": 0.166, "step": 1250 }, { "grad_norm": 2.1604762077331543, "learning_rate": 3.1475e-05, "loss": 0.1766, "step": 1260 }, { "grad_norm": 1.6072932481765747, "learning_rate": 3.1725e-05, "loss": 0.146, "step": 1270 }, { "grad_norm": 1.6355475187301636, "learning_rate": 3.1975e-05, "loss": 0.1754, "step": 1280 }, { "grad_norm": 1.5683919191360474, "learning_rate": 3.2225e-05, "loss": 0.1853, "step": 1290 }, { "grad_norm": 1.363614797592163, "learning_rate": 3.2474999999999997e-05, "loss": 0.1535, "step": 1300 }, { "grad_norm": 1.6818902492523193, "learning_rate": 3.2725e-05, "loss": 0.1508, "step": 1310 }, { "grad_norm": 1.7760802507400513, "learning_rate": 3.2975e-05, "loss": 0.1675, "step": 1320 }, { "grad_norm": 1.706284999847412, "learning_rate": 3.3225e-05, "loss": 0.1416, "step": 1330 }, { "grad_norm": 2.172621250152588, "learning_rate": 3.3475e-05, "loss": 0.1471, "step": 1340 }, { "grad_norm": 2.1956584453582764, "learning_rate": 3.3725e-05, "loss": 0.1956, "step": 1350 }, { "grad_norm": 2.2653989791870117, "learning_rate": 3.3975e-05, "loss": 0.1921, "step": 1360 }, { "grad_norm": 1.9143446683883667, "learning_rate": 3.4225e-05, "loss": 0.145, "step": 1370 }, { "grad_norm": 1.892903447151184, "learning_rate": 3.4475000000000005e-05, "loss": 0.1446, "step": 1380 }, { "grad_norm": 1.9731247425079346, "learning_rate": 3.4725000000000004e-05, "loss": 0.1678, "step": 1390 }, { "grad_norm": 1.7971580028533936, "learning_rate": 3.4975e-05, "loss": 0.1421, "step": 1400 }, { "grad_norm": 1.5332306623458862, "learning_rate": 3.5225e-05, "loss": 0.1538, "step": 1410 }, { "grad_norm": 1.8628439903259277, "learning_rate": 3.5475e-05, "loss": 0.159, "step": 1420 }, { "grad_norm": 1.9378061294555664, "learning_rate": 3.5725e-05, "loss": 0.1471, "step": 1430 }, { "grad_norm": 1.7027795314788818, "learning_rate": 3.5975e-05, "loss": 0.165, "step": 1440 }, { "grad_norm": 1.3147341012954712, "learning_rate": 3.6225000000000006e-05, "loss": 0.1703, "step": 1450 }, { "grad_norm": 1.3263579607009888, "learning_rate": 3.6475000000000006e-05, "loss": 0.1606, "step": 1460 }, { "grad_norm": 1.3185330629348755, "learning_rate": 3.6725000000000005e-05, "loss": 0.1664, "step": 1470 }, { "grad_norm": 1.4098179340362549, "learning_rate": 3.6975000000000004e-05, "loss": 0.1514, "step": 1480 }, { "grad_norm": 1.435321569442749, "learning_rate": 3.7225000000000004e-05, "loss": 0.1472, "step": 1490 }, { "grad_norm": 1.424238681793213, "learning_rate": 3.7475e-05, "loss": 0.1616, "step": 1500 }, { "grad_norm": 1.158288836479187, "learning_rate": 3.7725e-05, "loss": 0.1307, "step": 1510 }, { "grad_norm": 1.7149287462234497, "learning_rate": 3.7975e-05, "loss": 0.145, "step": 1520 }, { "grad_norm": 1.5420939922332764, "learning_rate": 3.8225e-05, "loss": 0.1534, "step": 1530 }, { "grad_norm": 1.5442694425582886, "learning_rate": 3.8475e-05, "loss": 0.1595, "step": 1540 }, { "grad_norm": 1.2980259656906128, "learning_rate": 3.8725e-05, "loss": 0.1403, "step": 1550 }, { "grad_norm": 1.8177140951156616, "learning_rate": 3.8975e-05, "loss": 0.1693, "step": 1560 }, { "grad_norm": 1.7758963108062744, "learning_rate": 3.9225e-05, "loss": 0.1296, "step": 1570 }, { "grad_norm": 1.8069531917572021, "learning_rate": 3.9475000000000004e-05, "loss": 0.1497, "step": 1580 }, { "grad_norm": 1.5191307067871094, "learning_rate": 3.9725e-05, "loss": 0.1554, "step": 1590 }, { "grad_norm": 1.233381748199463, "learning_rate": 3.9975e-05, "loss": 0.1342, "step": 1600 }, { "grad_norm": 1.2465975284576416, "learning_rate": 4.0225e-05, "loss": 0.1245, "step": 1610 }, { "grad_norm": 1.2459765672683716, "learning_rate": 4.0475e-05, "loss": 0.1467, "step": 1620 }, { "grad_norm": 1.5190714597702026, "learning_rate": 4.0725e-05, "loss": 0.1879, "step": 1630 }, { "grad_norm": 1.0795170068740845, "learning_rate": 4.0975e-05, "loss": 0.1515, "step": 1640 }, { "grad_norm": 1.471235752105713, "learning_rate": 4.1225e-05, "loss": 0.1589, "step": 1650 }, { "grad_norm": 1.3314272165298462, "learning_rate": 4.1475000000000005e-05, "loss": 0.1573, "step": 1660 }, { "grad_norm": 1.3607292175292969, "learning_rate": 4.1725000000000005e-05, "loss": 0.1305, "step": 1670 }, { "grad_norm": 1.4666807651519775, "learning_rate": 4.1975000000000004e-05, "loss": 0.138, "step": 1680 }, { "grad_norm": 1.231964349746704, "learning_rate": 4.2225e-05, "loss": 0.1393, "step": 1690 }, { "grad_norm": 1.0134303569793701, "learning_rate": 4.2475e-05, "loss": 0.1468, "step": 1700 }, { "grad_norm": 2.1333563327789307, "learning_rate": 4.2725e-05, "loss": 0.1718, "step": 1710 }, { "grad_norm": 1.8224492073059082, "learning_rate": 4.2975e-05, "loss": 0.1514, "step": 1720 }, { "grad_norm": 1.1641024351119995, "learning_rate": 4.322500000000001e-05, "loss": 0.1482, "step": 1730 }, { "grad_norm": 1.241318702697754, "learning_rate": 4.3475000000000006e-05, "loss": 0.1292, "step": 1740 }, { "grad_norm": 1.7894623279571533, "learning_rate": 4.3725000000000006e-05, "loss": 0.1472, "step": 1750 }, { "grad_norm": 1.4755194187164307, "learning_rate": 4.3975e-05, "loss": 0.1485, "step": 1760 }, { "grad_norm": 1.2708015441894531, "learning_rate": 4.4225e-05, "loss": 0.1337, "step": 1770 }, { "grad_norm": 1.3792446851730347, "learning_rate": 4.4475e-05, "loss": 0.1612, "step": 1780 }, { "grad_norm": 1.5497392416000366, "learning_rate": 4.4725e-05, "loss": 0.131, "step": 1790 }, { "grad_norm": 1.1793837547302246, "learning_rate": 4.4975e-05, "loss": 0.136, "step": 1800 }, { "grad_norm": 1.0487171411514282, "learning_rate": 4.5225e-05, "loss": 0.1407, "step": 1810 }, { "grad_norm": 1.573211669921875, "learning_rate": 4.5475e-05, "loss": 0.148, "step": 1820 }, { "grad_norm": 1.8588950634002686, "learning_rate": 4.5725e-05, "loss": 0.1341, "step": 1830 }, { "grad_norm": 1.38381826877594, "learning_rate": 4.5975e-05, "loss": 0.1553, "step": 1840 }, { "grad_norm": 1.4103209972381592, "learning_rate": 4.6225e-05, "loss": 0.1659, "step": 1850 }, { "grad_norm": 1.1400229930877686, "learning_rate": 4.6475000000000005e-05, "loss": 0.1651, "step": 1860 }, { "grad_norm": 0.9964438080787659, "learning_rate": 4.6725000000000004e-05, "loss": 0.1346, "step": 1870 }, { "grad_norm": 1.3773812055587769, "learning_rate": 4.6975000000000003e-05, "loss": 0.1235, "step": 1880 }, { "grad_norm": 1.0925477743148804, "learning_rate": 4.7225e-05, "loss": 0.1573, "step": 1890 }, { "grad_norm": 1.3585104942321777, "learning_rate": 4.7475e-05, "loss": 0.1352, "step": 1900 }, { "grad_norm": 1.3947548866271973, "learning_rate": 4.7725e-05, "loss": 0.1689, "step": 1910 }, { "grad_norm": 1.4609800577163696, "learning_rate": 4.7975e-05, "loss": 0.1532, "step": 1920 }, { "grad_norm": 1.355292797088623, "learning_rate": 4.822500000000001e-05, "loss": 0.1308, "step": 1930 }, { "grad_norm": 1.640354871749878, "learning_rate": 4.8475000000000006e-05, "loss": 0.1469, "step": 1940 }, { "grad_norm": 1.131830096244812, "learning_rate": 4.8725000000000005e-05, "loss": 0.1125, "step": 1950 }, { "grad_norm": 1.4303263425827026, "learning_rate": 4.8975000000000005e-05, "loss": 0.156, "step": 1960 }, { "grad_norm": 1.0991493463516235, "learning_rate": 4.9225000000000004e-05, "loss": 0.1065, "step": 1970 }, { "grad_norm": 1.3911397457122803, "learning_rate": 4.9475e-05, "loss": 0.165, "step": 1980 }, { "grad_norm": 1.1653836965560913, "learning_rate": 4.9725e-05, "loss": 0.1206, "step": 1990 }, { "grad_norm": 1.424317717552185, "learning_rate": 4.9975e-05, "loss": 0.1402, "step": 2000 }, { "grad_norm": 1.1168763637542725, "learning_rate": 5.0225e-05, "loss": 0.1716, "step": 2010 }, { "grad_norm": 1.4345104694366455, "learning_rate": 5.047500000000001e-05, "loss": 0.1463, "step": 2020 }, { "grad_norm": 2.1837961673736572, "learning_rate": 5.0725e-05, "loss": 0.1251, "step": 2030 }, { "grad_norm": 1.1767901182174683, "learning_rate": 5.0975000000000006e-05, "loss": 0.1414, "step": 2040 }, { "grad_norm": 1.5764249563217163, "learning_rate": 5.1225e-05, "loss": 0.1446, "step": 2050 }, { "grad_norm": 1.3659663200378418, "learning_rate": 5.1475000000000004e-05, "loss": 0.1447, "step": 2060 }, { "grad_norm": 1.0032602548599243, "learning_rate": 5.1725000000000004e-05, "loss": 0.1233, "step": 2070 }, { "grad_norm": 1.5089176893234253, "learning_rate": 5.197500000000001e-05, "loss": 0.1436, "step": 2080 }, { "grad_norm": 0.9643181562423706, "learning_rate": 5.2225e-05, "loss": 0.1434, "step": 2090 }, { "grad_norm": 1.0679306983947754, "learning_rate": 5.247500000000001e-05, "loss": 0.137, "step": 2100 }, { "grad_norm": 1.4400620460510254, "learning_rate": 5.2725e-05, "loss": 0.1669, "step": 2110 }, { "grad_norm": 1.127615213394165, "learning_rate": 5.297500000000001e-05, "loss": 0.1341, "step": 2120 }, { "grad_norm": 1.0059804916381836, "learning_rate": 5.3225e-05, "loss": 0.124, "step": 2130 }, { "grad_norm": 0.90712571144104, "learning_rate": 5.3475e-05, "loss": 0.1182, "step": 2140 }, { "grad_norm": 1.2739944458007812, "learning_rate": 5.3725000000000005e-05, "loss": 0.159, "step": 2150 }, { "grad_norm": 1.0408382415771484, "learning_rate": 5.3975e-05, "loss": 0.1384, "step": 2160 }, { "grad_norm": 1.1610506772994995, "learning_rate": 5.4225000000000003e-05, "loss": 0.1344, "step": 2170 }, { "grad_norm": 1.0424929857254028, "learning_rate": 5.4474999999999996e-05, "loss": 0.1466, "step": 2180 }, { "grad_norm": 1.2444555759429932, "learning_rate": 5.4725e-05, "loss": 0.135, "step": 2190 }, { "grad_norm": 1.0922415256500244, "learning_rate": 5.4975e-05, "loss": 0.1227, "step": 2200 }, { "grad_norm": 0.8875325918197632, "learning_rate": 5.522500000000001e-05, "loss": 0.1241, "step": 2210 }, { "grad_norm": 1.325106143951416, "learning_rate": 5.5475e-05, "loss": 0.1225, "step": 2220 }, { "grad_norm": 1.3773280382156372, "learning_rate": 5.5725000000000006e-05, "loss": 0.1232, "step": 2230 }, { "grad_norm": 1.0253798961639404, "learning_rate": 5.5975e-05, "loss": 0.1295, "step": 2240 }, { "grad_norm": 0.8507858514785767, "learning_rate": 5.6225000000000005e-05, "loss": 0.1326, "step": 2250 }, { "grad_norm": 0.8303322196006775, "learning_rate": 5.6475e-05, "loss": 0.1361, "step": 2260 }, { "grad_norm": 1.0879855155944824, "learning_rate": 5.6725e-05, "loss": 0.1248, "step": 2270 }, { "grad_norm": 1.0063351392745972, "learning_rate": 5.6975e-05, "loss": 0.1226, "step": 2280 }, { "grad_norm": 2.0597591400146484, "learning_rate": 5.722500000000001e-05, "loss": 0.162, "step": 2290 }, { "grad_norm": 1.122329831123352, "learning_rate": 5.7475e-05, "loss": 0.1305, "step": 2300 }, { "grad_norm": 1.0106711387634277, "learning_rate": 5.772500000000001e-05, "loss": 0.1272, "step": 2310 }, { "grad_norm": 1.0600337982177734, "learning_rate": 5.7975e-05, "loss": 0.114, "step": 2320 }, { "grad_norm": 1.0931288003921509, "learning_rate": 5.8225000000000006e-05, "loss": 0.1457, "step": 2330 }, { "grad_norm": 0.8238407969474792, "learning_rate": 5.8475000000000005e-05, "loss": 0.1294, "step": 2340 }, { "grad_norm": 1.0070254802703857, "learning_rate": 5.8725000000000004e-05, "loss": 0.1292, "step": 2350 }, { "grad_norm": 0.998976469039917, "learning_rate": 5.8975000000000004e-05, "loss": 0.1294, "step": 2360 }, { "grad_norm": 0.9616894125938416, "learning_rate": 5.922500000000001e-05, "loss": 0.1271, "step": 2370 }, { "grad_norm": 1.0951929092407227, "learning_rate": 5.9475e-05, "loss": 0.14, "step": 2380 }, { "grad_norm": 1.279456377029419, "learning_rate": 5.9724999999999995e-05, "loss": 0.1369, "step": 2390 }, { "grad_norm": 1.0865901708602905, "learning_rate": 5.9975e-05, "loss": 0.139, "step": 2400 }, { "grad_norm": 1.0460479259490967, "learning_rate": 6.0225e-05, "loss": 0.1297, "step": 2410 }, { "grad_norm": 0.9113929271697998, "learning_rate": 6.0475000000000006e-05, "loss": 0.1744, "step": 2420 }, { "grad_norm": 1.1556888818740845, "learning_rate": 6.0725e-05, "loss": 0.1386, "step": 2430 }, { "grad_norm": 1.0478001832962036, "learning_rate": 6.0975000000000005e-05, "loss": 0.1441, "step": 2440 }, { "grad_norm": 1.0841212272644043, "learning_rate": 6.1225e-05, "loss": 0.1171, "step": 2450 }, { "grad_norm": 0.8485478758811951, "learning_rate": 6.1475e-05, "loss": 0.1213, "step": 2460 }, { "grad_norm": 1.1947463750839233, "learning_rate": 6.1725e-05, "loss": 0.1209, "step": 2470 }, { "grad_norm": 1.0034921169281006, "learning_rate": 6.1975e-05, "loss": 0.1462, "step": 2480 }, { "grad_norm": 1.1073071956634521, "learning_rate": 6.2225e-05, "loss": 0.1329, "step": 2490 }, { "grad_norm": 0.7975163459777832, "learning_rate": 6.2475e-05, "loss": 0.1245, "step": 2500 }, { "grad_norm": 0.6894334554672241, "learning_rate": 6.2725e-05, "loss": 0.1159, "step": 2510 }, { "grad_norm": 0.7747728228569031, "learning_rate": 6.297500000000001e-05, "loss": 0.1222, "step": 2520 }, { "grad_norm": 1.0152428150177002, "learning_rate": 6.3225e-05, "loss": 0.113, "step": 2530 }, { "grad_norm": 1.0119216442108154, "learning_rate": 6.347500000000001e-05, "loss": 0.1209, "step": 2540 }, { "grad_norm": 0.8103578090667725, "learning_rate": 6.3725e-05, "loss": 0.1437, "step": 2550 }, { "grad_norm": 1.0274053812026978, "learning_rate": 6.397500000000001e-05, "loss": 0.1282, "step": 2560 }, { "grad_norm": 0.9556519389152527, "learning_rate": 6.4225e-05, "loss": 0.1201, "step": 2570 }, { "grad_norm": 0.8972387313842773, "learning_rate": 6.447500000000001e-05, "loss": 0.1453, "step": 2580 }, { "grad_norm": 0.9351829290390015, "learning_rate": 6.4725e-05, "loss": 0.1266, "step": 2590 }, { "grad_norm": 0.7517243027687073, "learning_rate": 6.497500000000001e-05, "loss": 0.1453, "step": 2600 }, { "grad_norm": 0.8839123845100403, "learning_rate": 6.5225e-05, "loss": 0.1262, "step": 2610 }, { "grad_norm": 0.9453726410865784, "learning_rate": 6.5475e-05, "loss": 0.1179, "step": 2620 }, { "grad_norm": 1.0021945238113403, "learning_rate": 6.5725e-05, "loss": 0.12, "step": 2630 }, { "grad_norm": 1.4119203090667725, "learning_rate": 6.5975e-05, "loss": 0.1437, "step": 2640 }, { "grad_norm": 1.1928406953811646, "learning_rate": 6.6225e-05, "loss": 0.1343, "step": 2650 }, { "grad_norm": 0.9589769244194031, "learning_rate": 6.6475e-05, "loss": 0.1393, "step": 2660 }, { "grad_norm": 0.7767665982246399, "learning_rate": 6.672500000000001e-05, "loss": 0.1505, "step": 2670 }, { "grad_norm": 0.9421648383140564, "learning_rate": 6.6975e-05, "loss": 0.1587, "step": 2680 }, { "grad_norm": 0.8698979020118713, "learning_rate": 6.722500000000001e-05, "loss": 0.1271, "step": 2690 }, { "grad_norm": 1.0022358894348145, "learning_rate": 6.7475e-05, "loss": 0.1119, "step": 2700 }, { "grad_norm": 0.7582162618637085, "learning_rate": 6.7725e-05, "loss": 0.1228, "step": 2710 }, { "grad_norm": 0.8033136129379272, "learning_rate": 6.7975e-05, "loss": 0.1397, "step": 2720 }, { "grad_norm": 1.0386176109313965, "learning_rate": 6.8225e-05, "loss": 0.1152, "step": 2730 }, { "grad_norm": 0.785668134689331, "learning_rate": 6.8475e-05, "loss": 0.1073, "step": 2740 }, { "grad_norm": 1.1740469932556152, "learning_rate": 6.8725e-05, "loss": 0.133, "step": 2750 }, { "grad_norm": 0.9316465258598328, "learning_rate": 6.8975e-05, "loss": 0.1289, "step": 2760 }, { "grad_norm": 0.887520968914032, "learning_rate": 6.9225e-05, "loss": 0.1219, "step": 2770 }, { "grad_norm": 0.8089535236358643, "learning_rate": 6.9475e-05, "loss": 0.1231, "step": 2780 }, { "grad_norm": 0.893642783164978, "learning_rate": 6.9725e-05, "loss": 0.1214, "step": 2790 }, { "grad_norm": 0.7382221221923828, "learning_rate": 6.997500000000001e-05, "loss": 0.1123, "step": 2800 }, { "grad_norm": 1.0548275709152222, "learning_rate": 7.022500000000001e-05, "loss": 0.1574, "step": 2810 }, { "grad_norm": 1.0904161930084229, "learning_rate": 7.0475e-05, "loss": 0.1195, "step": 2820 }, { "grad_norm": 0.7921836376190186, "learning_rate": 7.072500000000001e-05, "loss": 0.1108, "step": 2830 }, { "grad_norm": 0.8059587478637695, "learning_rate": 7.0975e-05, "loss": 0.1472, "step": 2840 }, { "grad_norm": 0.8365418314933777, "learning_rate": 7.122500000000001e-05, "loss": 0.1255, "step": 2850 }, { "grad_norm": 0.9346626400947571, "learning_rate": 7.1475e-05, "loss": 0.1207, "step": 2860 }, { "grad_norm": 0.9122495651245117, "learning_rate": 7.172500000000001e-05, "loss": 0.1106, "step": 2870 }, { "grad_norm": 0.7810640335083008, "learning_rate": 7.1975e-05, "loss": 0.1072, "step": 2880 }, { "grad_norm": 0.9350712299346924, "learning_rate": 7.2225e-05, "loss": 0.1218, "step": 2890 }, { "grad_norm": 0.7595111727714539, "learning_rate": 7.2475e-05, "loss": 0.1242, "step": 2900 }, { "grad_norm": 0.9962909817695618, "learning_rate": 7.272499999999999e-05, "loss": 0.1162, "step": 2910 }, { "grad_norm": 0.7283868193626404, "learning_rate": 7.2975e-05, "loss": 0.1322, "step": 2920 }, { "grad_norm": 0.8760164976119995, "learning_rate": 7.3225e-05, "loss": 0.1013, "step": 2930 }, { "grad_norm": 0.6660860776901245, "learning_rate": 7.347500000000001e-05, "loss": 0.11, "step": 2940 }, { "grad_norm": 0.8151665925979614, "learning_rate": 7.3725e-05, "loss": 0.1095, "step": 2950 }, { "grad_norm": 0.8155245780944824, "learning_rate": 7.397500000000001e-05, "loss": 0.1249, "step": 2960 }, { "grad_norm": 0.8712654113769531, "learning_rate": 7.4225e-05, "loss": 0.1066, "step": 2970 }, { "grad_norm": 0.9757735729217529, "learning_rate": 7.447500000000001e-05, "loss": 0.1261, "step": 2980 }, { "grad_norm": 0.9941118359565735, "learning_rate": 7.4725e-05, "loss": 0.1368, "step": 2990 }, { "grad_norm": 0.7229763865470886, "learning_rate": 7.4975e-05, "loss": 0.1179, "step": 3000 }, { "grad_norm": 0.6808424592018127, "learning_rate": 7.5225e-05, "loss": 0.0996, "step": 3010 }, { "grad_norm": 0.7159308195114136, "learning_rate": 7.5475e-05, "loss": 0.1057, "step": 3020 }, { "grad_norm": 1.1289381980895996, "learning_rate": 7.5725e-05, "loss": 0.1188, "step": 3030 }, { "grad_norm": 1.0929224491119385, "learning_rate": 7.5975e-05, "loss": 0.1156, "step": 3040 }, { "grad_norm": 1.0364917516708374, "learning_rate": 7.6225e-05, "loss": 0.1259, "step": 3050 }, { "grad_norm": 0.9043738842010498, "learning_rate": 7.6475e-05, "loss": 0.1322, "step": 3060 }, { "grad_norm": 0.81471186876297, "learning_rate": 7.672500000000001e-05, "loss": 0.1043, "step": 3070 }, { "grad_norm": 0.6367139220237732, "learning_rate": 7.697500000000001e-05, "loss": 0.1248, "step": 3080 }, { "grad_norm": 1.1086359024047852, "learning_rate": 7.722500000000001e-05, "loss": 0.1308, "step": 3090 }, { "grad_norm": 0.9926832914352417, "learning_rate": 7.747500000000001e-05, "loss": 0.1087, "step": 3100 }, { "grad_norm": 0.686936616897583, "learning_rate": 7.7725e-05, "loss": 0.1073, "step": 3110 }, { "grad_norm": 0.7353771328926086, "learning_rate": 7.797500000000001e-05, "loss": 0.1068, "step": 3120 }, { "grad_norm": 1.034144401550293, "learning_rate": 7.8225e-05, "loss": 0.1196, "step": 3130 }, { "grad_norm": 0.8224286437034607, "learning_rate": 7.8475e-05, "loss": 0.1269, "step": 3140 }, { "grad_norm": 0.8184403777122498, "learning_rate": 7.8725e-05, "loss": 0.1259, "step": 3150 }, { "grad_norm": 0.6854239702224731, "learning_rate": 7.8975e-05, "loss": 0.1261, "step": 3160 }, { "grad_norm": 0.9154037833213806, "learning_rate": 7.9225e-05, "loss": 0.1044, "step": 3170 }, { "grad_norm": 0.7398649454116821, "learning_rate": 7.9475e-05, "loss": 0.1218, "step": 3180 }, { "grad_norm": 0.7885061502456665, "learning_rate": 7.9725e-05, "loss": 0.1192, "step": 3190 }, { "grad_norm": 0.6989965438842773, "learning_rate": 7.9975e-05, "loss": 0.1169, "step": 3200 }, { "grad_norm": 0.8041898608207703, "learning_rate": 8.022500000000001e-05, "loss": 0.0997, "step": 3210 }, { "grad_norm": 0.8450477123260498, "learning_rate": 8.0475e-05, "loss": 0.1377, "step": 3220 }, { "grad_norm": 0.7909006476402283, "learning_rate": 8.072500000000001e-05, "loss": 0.1161, "step": 3230 }, { "grad_norm": 0.6391357183456421, "learning_rate": 8.0975e-05, "loss": 0.1048, "step": 3240 }, { "grad_norm": 0.48351016640663147, "learning_rate": 8.122500000000001e-05, "loss": 0.1118, "step": 3250 }, { "grad_norm": 1.1519356966018677, "learning_rate": 8.1475e-05, "loss": 0.1122, "step": 3260 }, { "grad_norm": 0.8494468331336975, "learning_rate": 8.172500000000001e-05, "loss": 0.1138, "step": 3270 }, { "grad_norm": 0.8854901194572449, "learning_rate": 8.1975e-05, "loss": 0.1037, "step": 3280 }, { "grad_norm": 0.9049690961837769, "learning_rate": 8.2225e-05, "loss": 0.1143, "step": 3290 }, { "grad_norm": 0.734511137008667, "learning_rate": 8.2475e-05, "loss": 0.1186, "step": 3300 }, { "grad_norm": 0.9253223538398743, "learning_rate": 8.2725e-05, "loss": 0.1237, "step": 3310 }, { "grad_norm": 0.7557114362716675, "learning_rate": 8.2975e-05, "loss": 0.1327, "step": 3320 }, { "grad_norm": 0.6484308838844299, "learning_rate": 8.3225e-05, "loss": 0.1224, "step": 3330 }, { "grad_norm": 0.9086417555809021, "learning_rate": 8.347500000000001e-05, "loss": 0.1162, "step": 3340 }, { "grad_norm": 0.6578227877616882, "learning_rate": 8.3725e-05, "loss": 0.1092, "step": 3350 }, { "grad_norm": 0.9530283808708191, "learning_rate": 8.397500000000001e-05, "loss": 0.1089, "step": 3360 }, { "grad_norm": 0.9166907072067261, "learning_rate": 8.422500000000001e-05, "loss": 0.1102, "step": 3370 }, { "grad_norm": 0.8611083030700684, "learning_rate": 8.447500000000001e-05, "loss": 0.1632, "step": 3380 }, { "grad_norm": 1.0131044387817383, "learning_rate": 8.4725e-05, "loss": 0.1072, "step": 3390 }, { "grad_norm": 0.7080921530723572, "learning_rate": 8.4975e-05, "loss": 0.1166, "step": 3400 }, { "grad_norm": 1.1113468408584595, "learning_rate": 8.5225e-05, "loss": 0.1134, "step": 3410 }, { "grad_norm": 0.8787866234779358, "learning_rate": 8.5475e-05, "loss": 0.1087, "step": 3420 }, { "grad_norm": 0.6298444271087646, "learning_rate": 8.5725e-05, "loss": 0.1144, "step": 3430 }, { "grad_norm": 0.7549595832824707, "learning_rate": 8.5975e-05, "loss": 0.1325, "step": 3440 }, { "grad_norm": 0.8853477835655212, "learning_rate": 8.6225e-05, "loss": 0.1144, "step": 3450 }, { "grad_norm": 0.6113834381103516, "learning_rate": 8.6475e-05, "loss": 0.1043, "step": 3460 }, { "grad_norm": 0.6829582452774048, "learning_rate": 8.672500000000001e-05, "loss": 0.1299, "step": 3470 }, { "grad_norm": 0.9447005987167358, "learning_rate": 8.6975e-05, "loss": 0.1277, "step": 3480 }, { "grad_norm": 0.8137087225914001, "learning_rate": 8.7225e-05, "loss": 0.1239, "step": 3490 }, { "grad_norm": 1.4807965755462646, "learning_rate": 8.747500000000001e-05, "loss": 0.1055, "step": 3500 }, { "grad_norm": 1.0811145305633545, "learning_rate": 8.7725e-05, "loss": 0.1186, "step": 3510 }, { "grad_norm": 0.9542176127433777, "learning_rate": 8.797500000000001e-05, "loss": 0.1049, "step": 3520 }, { "grad_norm": 0.984744131565094, "learning_rate": 8.8225e-05, "loss": 0.105, "step": 3530 }, { "grad_norm": 0.7059900164604187, "learning_rate": 8.847500000000001e-05, "loss": 0.1026, "step": 3540 }, { "grad_norm": 0.8489640951156616, "learning_rate": 8.8725e-05, "loss": 0.1157, "step": 3550 }, { "grad_norm": 1.0051528215408325, "learning_rate": 8.897500000000001e-05, "loss": 0.1113, "step": 3560 }, { "grad_norm": 0.6424331665039062, "learning_rate": 8.9225e-05, "loss": 0.123, "step": 3570 }, { "grad_norm": 0.7402273416519165, "learning_rate": 8.9475e-05, "loss": 0.1265, "step": 3580 }, { "grad_norm": 0.807869553565979, "learning_rate": 8.9725e-05, "loss": 0.1153, "step": 3590 }, { "grad_norm": 0.7787632346153259, "learning_rate": 8.9975e-05, "loss": 0.1059, "step": 3600 }, { "grad_norm": 0.9184478521347046, "learning_rate": 9.0225e-05, "loss": 0.1128, "step": 3610 }, { "grad_norm": 0.7678946256637573, "learning_rate": 9.0475e-05, "loss": 0.1036, "step": 3620 }, { "grad_norm": 0.6594859957695007, "learning_rate": 9.072500000000001e-05, "loss": 0.0896, "step": 3630 }, { "grad_norm": 0.6361724734306335, "learning_rate": 9.0975e-05, "loss": 0.1009, "step": 3640 }, { "grad_norm": 0.8869944214820862, "learning_rate": 9.122500000000001e-05, "loss": 0.1109, "step": 3650 }, { "grad_norm": 0.627805233001709, "learning_rate": 9.1475e-05, "loss": 0.1115, "step": 3660 }, { "grad_norm": 0.8018751740455627, "learning_rate": 9.172500000000001e-05, "loss": 0.1354, "step": 3670 }, { "grad_norm": 0.6641882658004761, "learning_rate": 9.1975e-05, "loss": 0.116, "step": 3680 }, { "grad_norm": 0.7983853220939636, "learning_rate": 9.2225e-05, "loss": 0.1092, "step": 3690 }, { "grad_norm": 0.6948412656784058, "learning_rate": 9.2475e-05, "loss": 0.1029, "step": 3700 }, { "grad_norm": 0.6734766364097595, "learning_rate": 9.2725e-05, "loss": 0.1028, "step": 3710 }, { "grad_norm": 0.5633143782615662, "learning_rate": 9.2975e-05, "loss": 0.1103, "step": 3720 }, { "grad_norm": 0.6940491795539856, "learning_rate": 9.3225e-05, "loss": 0.093, "step": 3730 }, { "grad_norm": 0.6606996655464172, "learning_rate": 9.3475e-05, "loss": 0.1043, "step": 3740 }, { "grad_norm": 0.7399952411651611, "learning_rate": 9.3725e-05, "loss": 0.1241, "step": 3750 }, { "grad_norm": 0.5361890196800232, "learning_rate": 9.397500000000001e-05, "loss": 0.0981, "step": 3760 }, { "grad_norm": 0.7456457018852234, "learning_rate": 9.422500000000001e-05, "loss": 0.1092, "step": 3770 }, { "grad_norm": 0.9013688564300537, "learning_rate": 9.4475e-05, "loss": 0.118, "step": 3780 }, { "grad_norm": 0.7201308012008667, "learning_rate": 9.472500000000001e-05, "loss": 0.1088, "step": 3790 }, { "grad_norm": 0.6784569621086121, "learning_rate": 9.4975e-05, "loss": 0.1451, "step": 3800 }, { "grad_norm": 0.5498961806297302, "learning_rate": 9.522500000000001e-05, "loss": 0.0936, "step": 3810 }, { "grad_norm": 0.75736004114151, "learning_rate": 9.5475e-05, "loss": 0.114, "step": 3820 }, { "grad_norm": 0.7642662525177002, "learning_rate": 9.572500000000001e-05, "loss": 0.0888, "step": 3830 }, { "grad_norm": 0.6550065875053406, "learning_rate": 9.5975e-05, "loss": 0.104, "step": 3840 }, { "grad_norm": 0.7708797454833984, "learning_rate": 9.622500000000001e-05, "loss": 0.1218, "step": 3850 }, { "grad_norm": 0.8095201253890991, "learning_rate": 9.6475e-05, "loss": 0.1089, "step": 3860 }, { "grad_norm": 0.6815040707588196, "learning_rate": 9.6725e-05, "loss": 0.1054, "step": 3870 }, { "grad_norm": 0.7240086197853088, "learning_rate": 9.6975e-05, "loss": 0.1029, "step": 3880 }, { "grad_norm": 0.6467761993408203, "learning_rate": 9.7225e-05, "loss": 0.1163, "step": 3890 }, { "grad_norm": 0.7900791764259338, "learning_rate": 9.747500000000001e-05, "loss": 0.1187, "step": 3900 }, { "grad_norm": 0.7642802000045776, "learning_rate": 9.7725e-05, "loss": 0.1039, "step": 3910 }, { "grad_norm": 0.844757616519928, "learning_rate": 9.797500000000001e-05, "loss": 0.1167, "step": 3920 }, { "grad_norm": 0.6663240194320679, "learning_rate": 9.8225e-05, "loss": 0.1012, "step": 3930 }, { "grad_norm": 0.7928993105888367, "learning_rate": 9.847500000000001e-05, "loss": 0.1189, "step": 3940 }, { "grad_norm": 0.5990821719169617, "learning_rate": 9.8725e-05, "loss": 0.1076, "step": 3950 }, { "grad_norm": 0.7413053512573242, "learning_rate": 9.897500000000001e-05, "loss": 0.1178, "step": 3960 }, { "grad_norm": 0.6920686960220337, "learning_rate": 9.9225e-05, "loss": 0.0941, "step": 3970 }, { "grad_norm": 0.5692014098167419, "learning_rate": 9.9475e-05, "loss": 0.1036, "step": 3980 }, { "grad_norm": 0.7956191301345825, "learning_rate": 9.9725e-05, "loss": 0.109, "step": 3990 }, { "grad_norm": 0.7039074897766113, "learning_rate": 9.9975e-05, "loss": 0.1014, "step": 4000 }, { "grad_norm": 0.7981671094894409, "learning_rate": 9.999999653982884e-05, "loss": 0.097, "step": 4010 }, { "grad_norm": 0.6060303449630737, "learning_rate": 9.999998457874392e-05, "loss": 0.1078, "step": 4020 }, { "grad_norm": 0.6691869497299194, "learning_rate": 9.999996407402913e-05, "loss": 0.1123, "step": 4030 }, { "grad_norm": 0.6337980628013611, "learning_rate": 9.999993502568801e-05, "loss": 0.0862, "step": 4040 }, { "grad_norm": 0.8269188404083252, "learning_rate": 9.999989743372548e-05, "loss": 0.1055, "step": 4050 }, { "grad_norm": 0.7657672762870789, "learning_rate": 9.999985129814798e-05, "loss": 0.1016, "step": 4060 }, { "grad_norm": 0.6619076728820801, "learning_rate": 9.99997966189634e-05, "loss": 0.089, "step": 4070 }, { "grad_norm": 0.7781673073768616, "learning_rate": 9.999973339618107e-05, "loss": 0.1073, "step": 4080 }, { "grad_norm": 1.2670925855636597, "learning_rate": 9.999966162981179e-05, "loss": 0.1144, "step": 4090 }, { "grad_norm": 0.9453447461128235, "learning_rate": 9.999958131986784e-05, "loss": 0.0986, "step": 4100 }, { "grad_norm": 0.7231854200363159, "learning_rate": 9.999949246636293e-05, "loss": 0.1155, "step": 4110 }, { "grad_norm": 0.8937472701072693, "learning_rate": 9.999939506931224e-05, "loss": 0.1131, "step": 4120 }, { "grad_norm": 0.7968401312828064, "learning_rate": 9.999928912873243e-05, "loss": 0.1004, "step": 4130 }, { "grad_norm": 0.5259761810302734, "learning_rate": 9.999917464464159e-05, "loss": 0.1218, "step": 4140 }, { "grad_norm": 0.750586748123169, "learning_rate": 9.999905161705929e-05, "loss": 0.1183, "step": 4150 }, { "grad_norm": 0.8341888785362244, "learning_rate": 9.999892004600653e-05, "loss": 0.1007, "step": 4160 }, { "grad_norm": 0.8245435953140259, "learning_rate": 9.999877993150581e-05, "loss": 0.1068, "step": 4170 }, { "grad_norm": 0.5787221789360046, "learning_rate": 9.999863127358108e-05, "loss": 0.0943, "step": 4180 }, { "grad_norm": 0.6376844644546509, "learning_rate": 9.999847407225773e-05, "loss": 0.0933, "step": 4190 }, { "grad_norm": 0.5055151581764221, "learning_rate": 9.999830832756262e-05, "loss": 0.096, "step": 4200 }, { "grad_norm": 0.598585307598114, "learning_rate": 9.999813403952407e-05, "loss": 0.0965, "step": 4210 }, { "grad_norm": 0.8081017136573792, "learning_rate": 9.999795120817187e-05, "loss": 0.1126, "step": 4220 }, { "grad_norm": 0.7285760641098022, "learning_rate": 9.999775983353725e-05, "loss": 0.1201, "step": 4230 }, { "grad_norm": 0.6338053345680237, "learning_rate": 9.999755991565292e-05, "loss": 0.1187, "step": 4240 }, { "grad_norm": 0.6395648717880249, "learning_rate": 9.999735145455303e-05, "loss": 0.1175, "step": 4250 }, { "grad_norm": 0.697553813457489, "learning_rate": 9.99971344502732e-05, "loss": 0.112, "step": 4260 }, { "grad_norm": 0.522239089012146, "learning_rate": 9.999690890285053e-05, "loss": 0.1108, "step": 4270 }, { "grad_norm": 0.5528923869132996, "learning_rate": 9.999667481232356e-05, "loss": 0.0876, "step": 4280 }, { "grad_norm": 0.7470232248306274, "learning_rate": 9.999643217873225e-05, "loss": 0.0997, "step": 4290 }, { "grad_norm": 0.6680892705917358, "learning_rate": 9.999618100211809e-05, "loss": 0.0842, "step": 4300 }, { "grad_norm": 0.7164068222045898, "learning_rate": 9.999592128252402e-05, "loss": 0.096, "step": 4310 }, { "grad_norm": 1.0878950357437134, "learning_rate": 9.999565301999437e-05, "loss": 0.1321, "step": 4320 }, { "grad_norm": 0.5837415456771851, "learning_rate": 9.999537621457502e-05, "loss": 0.0953, "step": 4330 }, { "grad_norm": 0.5510469675064087, "learning_rate": 9.999509086631323e-05, "loss": 0.076, "step": 4340 }, { "grad_norm": 0.688227117061615, "learning_rate": 9.99947969752578e-05, "loss": 0.0851, "step": 4350 }, { "grad_norm": 0.7382494807243347, "learning_rate": 9.999449454145891e-05, "loss": 0.0885, "step": 4360 }, { "grad_norm": 0.7650052309036255, "learning_rate": 9.999418356496827e-05, "loss": 0.0919, "step": 4370 }, { "grad_norm": 0.5685590505599976, "learning_rate": 9.999386404583899e-05, "loss": 0.1065, "step": 4380 }, { "grad_norm": 0.7607053518295288, "learning_rate": 9.999353598412568e-05, "loss": 0.0902, "step": 4390 }, { "grad_norm": 0.565920889377594, "learning_rate": 9.999319937988442e-05, "loss": 0.1069, "step": 4400 }, { "grad_norm": 0.6965380311012268, "learning_rate": 9.999285423317268e-05, "loss": 0.1145, "step": 4410 }, { "grad_norm": 0.5565849542617798, "learning_rate": 9.999250054404947e-05, "loss": 0.1195, "step": 4420 }, { "grad_norm": 0.7990571856498718, "learning_rate": 9.99921383125752e-05, "loss": 0.1051, "step": 4430 }, { "grad_norm": 0.6265918612480164, "learning_rate": 9.99917675388118e-05, "loss": 0.1029, "step": 4440 }, { "grad_norm": 0.6388951539993286, "learning_rate": 9.99913882228226e-05, "loss": 0.1143, "step": 4450 }, { "grad_norm": 0.6089746356010437, "learning_rate": 9.999100036467242e-05, "loss": 0.0953, "step": 4460 }, { "grad_norm": 0.5206828713417053, "learning_rate": 9.999060396442753e-05, "loss": 0.0965, "step": 4470 }, { "grad_norm": 0.7462579011917114, "learning_rate": 9.999019902215566e-05, "loss": 0.1061, "step": 4480 }, { "grad_norm": 0.623490035533905, "learning_rate": 9.998978553792602e-05, "loss": 0.0964, "step": 4490 }, { "grad_norm": 0.6023817658424377, "learning_rate": 9.998936351180926e-05, "loss": 0.1049, "step": 4500 }, { "grad_norm": 0.8331722021102905, "learning_rate": 9.998893294387747e-05, "loss": 0.1196, "step": 4510 }, { "grad_norm": 0.6991809010505676, "learning_rate": 9.998849383420426e-05, "loss": 0.1148, "step": 4520 }, { "grad_norm": 0.675678014755249, "learning_rate": 9.998804618286465e-05, "loss": 0.1, "step": 4530 }, { "grad_norm": 0.7038875222206116, "learning_rate": 9.99875899899351e-05, "loss": 0.1104, "step": 4540 }, { "grad_norm": 0.6484823226928711, "learning_rate": 9.99871252554936e-05, "loss": 0.1052, "step": 4550 }, { "grad_norm": 0.5916885137557983, "learning_rate": 9.998665197961955e-05, "loss": 0.0862, "step": 4560 }, { "grad_norm": 0.5336573123931885, "learning_rate": 9.998617016239379e-05, "loss": 0.1089, "step": 4570 }, { "grad_norm": 0.8801625967025757, "learning_rate": 9.998567980389869e-05, "loss": 0.0985, "step": 4580 }, { "grad_norm": 0.6799409985542297, "learning_rate": 9.998518090421802e-05, "loss": 0.0823, "step": 4590 }, { "grad_norm": 0.7137022614479065, "learning_rate": 9.998467346343703e-05, "loss": 0.0927, "step": 4600 }, { "grad_norm": 0.5891428589820862, "learning_rate": 9.998415748164243e-05, "loss": 0.0839, "step": 4610 }, { "grad_norm": 0.6559416055679321, "learning_rate": 9.998363295892238e-05, "loss": 0.0859, "step": 4620 }, { "grad_norm": 0.7482913732528687, "learning_rate": 9.998309989536652e-05, "loss": 0.0924, "step": 4630 }, { "grad_norm": 0.5690167546272278, "learning_rate": 9.998255829106593e-05, "loss": 0.0986, "step": 4640 }, { "grad_norm": 0.7534058094024658, "learning_rate": 9.998200814611316e-05, "loss": 0.0769, "step": 4650 }, { "grad_norm": 0.5015066862106323, "learning_rate": 9.998144946060219e-05, "loss": 0.0958, "step": 4660 }, { "grad_norm": 0.8920350670814514, "learning_rate": 9.998088223462852e-05, "loss": 0.0918, "step": 4670 }, { "grad_norm": 0.5986049771308899, "learning_rate": 9.998030646828905e-05, "loss": 0.0892, "step": 4680 }, { "grad_norm": 0.5840468406677246, "learning_rate": 9.997972216168217e-05, "loss": 0.0991, "step": 4690 }, { "grad_norm": 0.7178251147270203, "learning_rate": 9.997912931490771e-05, "loss": 0.0901, "step": 4700 }, { "grad_norm": 0.694788932800293, "learning_rate": 9.9978527928067e-05, "loss": 0.0979, "step": 4710 }, { "grad_norm": 0.5742424130439758, "learning_rate": 9.997791800126277e-05, "loss": 0.0829, "step": 4720 }, { "grad_norm": 0.6194410920143127, "learning_rate": 9.997729953459927e-05, "loss": 0.0957, "step": 4730 }, { "grad_norm": 0.601927638053894, "learning_rate": 9.997667252818214e-05, "loss": 0.095, "step": 4740 }, { "grad_norm": 0.7542332410812378, "learning_rate": 9.997603698211855e-05, "loss": 0.0817, "step": 4750 }, { "grad_norm": 0.8742945790290833, "learning_rate": 9.99753928965171e-05, "loss": 0.098, "step": 4760 }, { "grad_norm": 0.9261525869369507, "learning_rate": 9.997474027148781e-05, "loss": 0.1055, "step": 4770 }, { "grad_norm": 0.6791014671325684, "learning_rate": 9.997407910714223e-05, "loss": 0.1075, "step": 4780 }, { "grad_norm": 0.7027056813240051, "learning_rate": 9.997340940359332e-05, "loss": 0.0827, "step": 4790 }, { "grad_norm": 0.6770966649055481, "learning_rate": 9.997273116095552e-05, "loss": 0.0888, "step": 4800 }, { "grad_norm": 0.5753495693206787, "learning_rate": 9.997204437934473e-05, "loss": 0.0915, "step": 4810 }, { "grad_norm": 0.6300247311592102, "learning_rate": 9.997134905887829e-05, "loss": 0.1, "step": 4820 }, { "grad_norm": 0.5409174561500549, "learning_rate": 9.997064519967501e-05, "loss": 0.1032, "step": 4830 }, { "grad_norm": 0.5869407653808594, "learning_rate": 9.996993280185517e-05, "loss": 0.0986, "step": 4840 }, { "grad_norm": 0.7518944144248962, "learning_rate": 9.99692118655405e-05, "loss": 0.0958, "step": 4850 }, { "grad_norm": 0.5565025210380554, "learning_rate": 9.996848239085417e-05, "loss": 0.0784, "step": 4860 }, { "grad_norm": 0.5204944610595703, "learning_rate": 9.996774437792085e-05, "loss": 0.0813, "step": 4870 }, { "grad_norm": 0.8250227570533752, "learning_rate": 9.996699782686664e-05, "loss": 0.0775, "step": 4880 }, { "grad_norm": 0.7036257982254028, "learning_rate": 9.996624273781909e-05, "loss": 0.1066, "step": 4890 }, { "grad_norm": 0.565721869468689, "learning_rate": 9.996547911090725e-05, "loss": 0.0804, "step": 4900 }, { "grad_norm": 0.5985860824584961, "learning_rate": 9.996470694626157e-05, "loss": 0.0876, "step": 4910 }, { "grad_norm": 0.6181182265281677, "learning_rate": 9.996392624401403e-05, "loss": 0.0904, "step": 4920 }, { "grad_norm": 0.7152917981147766, "learning_rate": 9.996313700429801e-05, "loss": 0.0913, "step": 4930 }, { "grad_norm": 0.47154614329338074, "learning_rate": 9.996233922724836e-05, "loss": 0.0739, "step": 4940 }, { "grad_norm": 0.567353367805481, "learning_rate": 9.996153291300141e-05, "loss": 0.0983, "step": 4950 }, { "grad_norm": 0.5772924423217773, "learning_rate": 9.996071806169494e-05, "loss": 0.0823, "step": 4960 }, { "grad_norm": 0.6517527103424072, "learning_rate": 9.995989467346817e-05, "loss": 0.1005, "step": 4970 }, { "grad_norm": 0.6185274720191956, "learning_rate": 9.995906274846183e-05, "loss": 0.086, "step": 4980 }, { "grad_norm": 0.7519074082374573, "learning_rate": 9.995822228681803e-05, "loss": 0.0838, "step": 4990 }, { "grad_norm": 0.5517241358757019, "learning_rate": 9.99573732886804e-05, "loss": 0.1017, "step": 5000 }, { "grad_norm": 0.6155580878257751, "learning_rate": 9.995651575419402e-05, "loss": 0.0911, "step": 5010 }, { "grad_norm": 0.7370482683181763, "learning_rate": 9.995564968350541e-05, "loss": 0.0924, "step": 5020 }, { "grad_norm": 0.5623706579208374, "learning_rate": 9.995477507676256e-05, "loss": 0.0755, "step": 5030 }, { "grad_norm": 0.6741370558738708, "learning_rate": 9.995389193411493e-05, "loss": 0.0883, "step": 5040 }, { "grad_norm": 0.6500552296638489, "learning_rate": 9.995300025571339e-05, "loss": 0.0916, "step": 5050 }, { "grad_norm": 0.48787999153137207, "learning_rate": 9.995210004171034e-05, "loss": 0.0773, "step": 5060 }, { "grad_norm": 0.5634146928787231, "learning_rate": 9.995119129225956e-05, "loss": 0.0962, "step": 5070 }, { "grad_norm": 0.6397677659988403, "learning_rate": 9.995027400751637e-05, "loss": 0.0961, "step": 5080 }, { "grad_norm": 0.5837829113006592, "learning_rate": 9.994934818763751e-05, "loss": 0.085, "step": 5090 }, { "grad_norm": 0.5180878639221191, "learning_rate": 9.994841383278115e-05, "loss": 0.0903, "step": 5100 }, { "grad_norm": 0.919968843460083, "learning_rate": 9.994747094310695e-05, "loss": 0.1023, "step": 5110 }, { "grad_norm": 0.7755216360092163, "learning_rate": 9.994651951877604e-05, "loss": 0.1212, "step": 5120 }, { "grad_norm": 0.7316403388977051, "learning_rate": 9.994555955995099e-05, "loss": 0.1069, "step": 5130 }, { "grad_norm": 0.5562529563903809, "learning_rate": 9.994459106679581e-05, "loss": 0.0957, "step": 5140 }, { "grad_norm": 0.45953264832496643, "learning_rate": 9.994361403947603e-05, "loss": 0.0849, "step": 5150 }, { "grad_norm": 0.6658604741096497, "learning_rate": 9.994262847815854e-05, "loss": 0.1118, "step": 5160 }, { "grad_norm": 0.6145270466804504, "learning_rate": 9.99416343830118e-05, "loss": 0.0902, "step": 5170 }, { "grad_norm": 0.5948000550270081, "learning_rate": 9.994063175420565e-05, "loss": 0.0735, "step": 5180 }, { "grad_norm": 0.6031689643859863, "learning_rate": 9.99396205919114e-05, "loss": 0.1103, "step": 5190 }, { "grad_norm": 0.6850559115409851, "learning_rate": 9.993860089630185e-05, "loss": 0.0794, "step": 5200 }, { "grad_norm": 0.5220316648483276, "learning_rate": 9.993757266755123e-05, "loss": 0.0888, "step": 5210 }, { "grad_norm": 0.7960907816886902, "learning_rate": 9.993653590583522e-05, "loss": 0.0946, "step": 5220 }, { "grad_norm": 0.6132569313049316, "learning_rate": 9.993549061133102e-05, "loss": 0.0864, "step": 5230 }, { "grad_norm": 0.6274350881576538, "learning_rate": 9.993443678421719e-05, "loss": 0.0907, "step": 5240 }, { "grad_norm": 0.7321327924728394, "learning_rate": 9.993337442467384e-05, "loss": 0.0976, "step": 5250 }, { "grad_norm": 0.6260828375816345, "learning_rate": 9.993230353288248e-05, "loss": 0.0881, "step": 5260 }, { "grad_norm": 0.580403208732605, "learning_rate": 9.993122410902608e-05, "loss": 0.0819, "step": 5270 }, { "grad_norm": 0.6990161538124084, "learning_rate": 9.993013615328912e-05, "loss": 0.1005, "step": 5280 }, { "grad_norm": 0.554620623588562, "learning_rate": 9.992903966585747e-05, "loss": 0.0963, "step": 5290 }, { "grad_norm": 0.6714040040969849, "learning_rate": 9.992793464691852e-05, "loss": 0.07, "step": 5300 }, { "grad_norm": 0.5640988349914551, "learning_rate": 9.992682109666105e-05, "loss": 0.0827, "step": 5310 }, { "grad_norm": 0.759824275970459, "learning_rate": 9.992569901527538e-05, "loss": 0.0919, "step": 5320 }, { "grad_norm": 0.6658539175987244, "learning_rate": 9.99245684029532e-05, "loss": 0.095, "step": 5330 }, { "grad_norm": 0.731031060218811, "learning_rate": 9.992342925988774e-05, "loss": 0.0879, "step": 5340 }, { "grad_norm": 0.6614003777503967, "learning_rate": 9.992228158627361e-05, "loss": 0.0868, "step": 5350 }, { "grad_norm": 0.523201584815979, "learning_rate": 9.992112538230693e-05, "loss": 0.095, "step": 5360 }, { "grad_norm": 0.5986416935920715, "learning_rate": 9.991996064818527e-05, "loss": 0.0905, "step": 5370 }, { "grad_norm": 0.5762227177619934, "learning_rate": 9.991878738410768e-05, "loss": 0.0785, "step": 5380 }, { "grad_norm": 0.6274988055229187, "learning_rate": 9.991760559027457e-05, "loss": 0.0816, "step": 5390 }, { "grad_norm": 0.5453078150749207, "learning_rate": 9.991641526688793e-05, "loss": 0.0874, "step": 5400 }, { "grad_norm": 0.5618433356285095, "learning_rate": 9.991521641415113e-05, "loss": 0.0797, "step": 5410 }, { "grad_norm": 0.7659759521484375, "learning_rate": 9.991400903226904e-05, "loss": 0.0762, "step": 5420 }, { "grad_norm": 0.7010557651519775, "learning_rate": 9.991279312144794e-05, "loss": 0.0901, "step": 5430 }, { "grad_norm": 0.7206907868385315, "learning_rate": 9.991156868189564e-05, "loss": 0.0969, "step": 5440 }, { "grad_norm": 0.5613806247711182, "learning_rate": 9.991033571382131e-05, "loss": 0.0686, "step": 5450 }, { "grad_norm": 0.7191039323806763, "learning_rate": 9.990909421743569e-05, "loss": 0.08, "step": 5460 }, { "grad_norm": 0.6218670606613159, "learning_rate": 9.990784419295085e-05, "loss": 0.0843, "step": 5470 }, { "grad_norm": 0.496562659740448, "learning_rate": 9.990658564058044e-05, "loss": 0.0779, "step": 5480 }, { "grad_norm": 0.6657266020774841, "learning_rate": 9.990531856053948e-05, "loss": 0.0868, "step": 5490 }, { "grad_norm": 0.7505048513412476, "learning_rate": 9.99040429530445e-05, "loss": 0.0926, "step": 5500 }, { "grad_norm": 0.6429239511489868, "learning_rate": 9.990275881831346e-05, "loss": 0.0836, "step": 5510 }, { "grad_norm": 0.5243264436721802, "learning_rate": 9.990146615656577e-05, "loss": 0.0715, "step": 5520 }, { "grad_norm": 0.7601678967475891, "learning_rate": 9.990016496802233e-05, "loss": 0.0695, "step": 5530 }, { "grad_norm": 0.5012028217315674, "learning_rate": 9.989885525290548e-05, "loss": 0.0679, "step": 5540 }, { "grad_norm": 0.5842660069465637, "learning_rate": 9.989753701143897e-05, "loss": 0.0791, "step": 5550 }, { "grad_norm": 0.5159702301025391, "learning_rate": 9.989621024384812e-05, "loss": 0.0732, "step": 5560 }, { "grad_norm": 0.6364569067955017, "learning_rate": 9.989487495035959e-05, "loss": 0.0826, "step": 5570 }, { "grad_norm": 0.6001564860343933, "learning_rate": 9.989353113120156e-05, "loss": 0.0993, "step": 5580 }, { "grad_norm": 0.6488908529281616, "learning_rate": 9.989217878660366e-05, "loss": 0.0826, "step": 5590 }, { "grad_norm": 0.6595285534858704, "learning_rate": 9.989081791679695e-05, "loss": 0.0871, "step": 5600 }, { "grad_norm": 0.8973170518875122, "learning_rate": 9.988944852201397e-05, "loss": 0.0784, "step": 5610 }, { "grad_norm": 0.5641223192214966, "learning_rate": 9.988807060248873e-05, "loss": 0.0757, "step": 5620 }, { "grad_norm": 0.7831069231033325, "learning_rate": 9.988668415845665e-05, "loss": 0.0759, "step": 5630 }, { "grad_norm": 0.7538325786590576, "learning_rate": 9.988528919015466e-05, "loss": 0.0852, "step": 5640 }, { "grad_norm": 0.7059245109558105, "learning_rate": 9.988388569782112e-05, "loss": 0.1096, "step": 5650 }, { "grad_norm": 0.6149865984916687, "learning_rate": 9.988247368169583e-05, "loss": 0.091, "step": 5660 }, { "grad_norm": 0.7377924919128418, "learning_rate": 9.988105314202007e-05, "loss": 0.0812, "step": 5670 }, { "grad_norm": 0.515812873840332, "learning_rate": 9.987962407903659e-05, "loss": 0.0941, "step": 5680 }, { "grad_norm": 0.829767107963562, "learning_rate": 9.987818649298957e-05, "loss": 0.0859, "step": 5690 }, { "grad_norm": 0.662843644618988, "learning_rate": 9.987674038412465e-05, "loss": 0.0827, "step": 5700 }, { "grad_norm": 0.5835556983947754, "learning_rate": 9.987528575268891e-05, "loss": 0.0819, "step": 5710 }, { "grad_norm": 0.5776867270469666, "learning_rate": 9.987382259893095e-05, "loss": 0.0784, "step": 5720 }, { "grad_norm": 0.547709047794342, "learning_rate": 9.987235092310074e-05, "loss": 0.0793, "step": 5730 }, { "grad_norm": 0.4801659882068634, "learning_rate": 9.987087072544978e-05, "loss": 0.0836, "step": 5740 }, { "grad_norm": 0.6272757053375244, "learning_rate": 9.9869382006231e-05, "loss": 0.0731, "step": 5750 }, { "grad_norm": 0.499258428812027, "learning_rate": 9.986788476569875e-05, "loss": 0.0797, "step": 5760 }, { "grad_norm": 0.6645480394363403, "learning_rate": 9.986637900410887e-05, "loss": 0.0945, "step": 5770 }, { "grad_norm": 0.6234697699546814, "learning_rate": 9.986486472171869e-05, "loss": 0.093, "step": 5780 }, { "grad_norm": 0.6439487934112549, "learning_rate": 9.986334191878692e-05, "loss": 0.0747, "step": 5790 }, { "grad_norm": 0.5174974203109741, "learning_rate": 9.986181059557378e-05, "loss": 0.0939, "step": 5800 }, { "grad_norm": 0.5600347518920898, "learning_rate": 9.986027075234094e-05, "loss": 0.0767, "step": 5810 }, { "grad_norm": 0.5671840310096741, "learning_rate": 9.985872238935152e-05, "loss": 0.082, "step": 5820 }, { "grad_norm": 0.5315535664558411, "learning_rate": 9.985716550687008e-05, "loss": 0.0862, "step": 5830 }, { "grad_norm": 0.5145463347434998, "learning_rate": 9.985560010516264e-05, "loss": 0.0743, "step": 5840 }, { "grad_norm": 0.5388501882553101, "learning_rate": 9.985402618449668e-05, "loss": 0.0721, "step": 5850 }, { "grad_norm": 0.5255123972892761, "learning_rate": 9.985244374514118e-05, "loss": 0.094, "step": 5860 }, { "grad_norm": 0.6455999612808228, "learning_rate": 9.985085278736651e-05, "loss": 0.0791, "step": 5870 }, { "grad_norm": 0.7137066721916199, "learning_rate": 9.984925331144452e-05, "loss": 0.0922, "step": 5880 }, { "grad_norm": 0.4131578505039215, "learning_rate": 9.984764531764851e-05, "loss": 0.0892, "step": 5890 }, { "grad_norm": 0.6826311349868774, "learning_rate": 9.984602880625326e-05, "loss": 0.1021, "step": 5900 }, { "grad_norm": 0.5684216618537903, "learning_rate": 9.9844403777535e-05, "loss": 0.0879, "step": 5910 }, { "grad_norm": 0.8349085450172424, "learning_rate": 9.984277023177135e-05, "loss": 0.0825, "step": 5920 }, { "grad_norm": 0.5297510623931885, "learning_rate": 9.984112816924148e-05, "loss": 0.0747, "step": 5930 }, { "grad_norm": 0.5735505819320679, "learning_rate": 9.983947759022596e-05, "loss": 0.0797, "step": 5940 }, { "grad_norm": 0.464016854763031, "learning_rate": 9.983781849500682e-05, "loss": 0.0677, "step": 5950 }, { "grad_norm": 0.6817145347595215, "learning_rate": 9.98361508838676e-05, "loss": 0.0672, "step": 5960 }, { "grad_norm": 0.8335007429122925, "learning_rate": 9.98344747570932e-05, "loss": 0.0787, "step": 5970 }, { "grad_norm": 0.6549260020256042, "learning_rate": 9.983279011497004e-05, "loss": 0.0837, "step": 5980 }, { "grad_norm": 0.5789822936058044, "learning_rate": 9.983109695778596e-05, "loss": 0.0986, "step": 5990 }, { "grad_norm": 0.45071253180503845, "learning_rate": 9.982939528583032e-05, "loss": 0.1091, "step": 6000 }, { "grad_norm": 0.5560003519058228, "learning_rate": 9.982768509939385e-05, "loss": 0.0913, "step": 6010 }, { "grad_norm": 0.7087506055831909, "learning_rate": 9.982596639876879e-05, "loss": 0.0866, "step": 6020 }, { "grad_norm": 0.5803112387657166, "learning_rate": 9.982423918424881e-05, "loss": 0.0895, "step": 6030 }, { "grad_norm": 0.5216010212898254, "learning_rate": 9.982250345612908e-05, "loss": 0.0783, "step": 6040 }, { "grad_norm": 0.47466447949409485, "learning_rate": 9.982075921470611e-05, "loss": 0.0898, "step": 6050 }, { "grad_norm": 0.565553605556488, "learning_rate": 9.981900646027802e-05, "loss": 0.0765, "step": 6060 }, { "grad_norm": 0.5241788029670715, "learning_rate": 9.981724519314425e-05, "loss": 0.0738, "step": 6070 }, { "grad_norm": 0.5426858067512512, "learning_rate": 9.981547541360581e-05, "loss": 0.0938, "step": 6080 }, { "grad_norm": 0.7127178907394409, "learning_rate": 9.981369712196508e-05, "loss": 0.0864, "step": 6090 }, { "grad_norm": 0.6573800444602966, "learning_rate": 9.981191031852592e-05, "loss": 0.0937, "step": 6100 }, { "grad_norm": 0.5573520660400391, "learning_rate": 9.981011500359362e-05, "loss": 0.0721, "step": 6110 }, { "grad_norm": 0.5194175839424133, "learning_rate": 9.9808311177475e-05, "loss": 0.0823, "step": 6120 }, { "grad_norm": 0.5058048963546753, "learning_rate": 9.980649884047826e-05, "loss": 0.0607, "step": 6130 }, { "grad_norm": 0.6107575297355652, "learning_rate": 9.980467799291307e-05, "loss": 0.0787, "step": 6140 }, { "grad_norm": 0.5394389033317566, "learning_rate": 9.980284863509058e-05, "loss": 0.109, "step": 6150 }, { "grad_norm": 0.7399730682373047, "learning_rate": 9.980101076732334e-05, "loss": 0.1108, "step": 6160 }, { "grad_norm": 0.6432236433029175, "learning_rate": 9.979916438992544e-05, "loss": 0.0965, "step": 6170 }, { "grad_norm": 0.4074375331401825, "learning_rate": 9.979730950321237e-05, "loss": 0.0823, "step": 6180 }, { "grad_norm": 0.6698418855667114, "learning_rate": 9.979544610750104e-05, "loss": 0.0936, "step": 6190 }, { "grad_norm": 0.5599140524864197, "learning_rate": 9.97935742031099e-05, "loss": 0.0776, "step": 6200 }, { "grad_norm": 0.7415172457695007, "learning_rate": 9.979169379035878e-05, "loss": 0.0846, "step": 6210 }, { "grad_norm": 0.6173596978187561, "learning_rate": 9.978980486956899e-05, "loss": 0.0691, "step": 6220 }, { "grad_norm": 0.5903868079185486, "learning_rate": 9.978790744106332e-05, "loss": 0.0818, "step": 6230 }, { "grad_norm": 0.5365450978279114, "learning_rate": 9.978600150516594e-05, "loss": 0.0898, "step": 6240 }, { "grad_norm": 0.4831831157207489, "learning_rate": 9.978408706220259e-05, "loss": 0.0878, "step": 6250 }, { "grad_norm": 0.7381918430328369, "learning_rate": 9.978216411250032e-05, "loss": 0.0686, "step": 6260 }, { "grad_norm": 0.6360598206520081, "learning_rate": 9.978023265638778e-05, "loss": 0.077, "step": 6270 }, { "grad_norm": 0.5635548830032349, "learning_rate": 9.977829269419495e-05, "loss": 0.0773, "step": 6280 }, { "grad_norm": 0.7385721802711487, "learning_rate": 9.977634422625335e-05, "loss": 0.0833, "step": 6290 }, { "grad_norm": 0.44835302233695984, "learning_rate": 9.97743872528959e-05, "loss": 0.0794, "step": 6300 }, { "grad_norm": 0.8148064017295837, "learning_rate": 9.9772421774457e-05, "loss": 0.0871, "step": 6310 }, { "grad_norm": 0.6062756180763245, "learning_rate": 9.977044779127252e-05, "loss": 0.0784, "step": 6320 }, { "grad_norm": 0.6139128804206848, "learning_rate": 9.976846530367971e-05, "loss": 0.0725, "step": 6330 }, { "grad_norm": 0.603299617767334, "learning_rate": 9.976647431201735e-05, "loss": 0.0863, "step": 6340 }, { "grad_norm": 0.4596889317035675, "learning_rate": 9.976447481662568e-05, "loss": 0.0766, "step": 6350 }, { "grad_norm": 0.5278664827346802, "learning_rate": 9.976246681784629e-05, "loss": 0.0795, "step": 6360 }, { "grad_norm": 0.5469969511032104, "learning_rate": 9.976045031602234e-05, "loss": 0.0695, "step": 6370 }, { "grad_norm": 0.6287055015563965, "learning_rate": 9.975842531149837e-05, "loss": 0.0711, "step": 6380 }, { "grad_norm": 0.653008222579956, "learning_rate": 9.975639180462043e-05, "loss": 0.0831, "step": 6390 }, { "grad_norm": 0.6992461681365967, "learning_rate": 9.975434979573596e-05, "loss": 0.0879, "step": 6400 }, { "grad_norm": 0.5120721459388733, "learning_rate": 9.97522992851939e-05, "loss": 0.0666, "step": 6410 }, { "grad_norm": 0.5056882500648499, "learning_rate": 9.975024027334461e-05, "loss": 0.0723, "step": 6420 }, { "grad_norm": 0.5476727485656738, "learning_rate": 9.974817276053993e-05, "loss": 0.0777, "step": 6430 }, { "grad_norm": 0.6113075017929077, "learning_rate": 9.974609674713315e-05, "loss": 0.0767, "step": 6440 }, { "grad_norm": 0.5964843034744263, "learning_rate": 9.9744012233479e-05, "loss": 0.0721, "step": 6450 }, { "grad_norm": 0.4700183570384979, "learning_rate": 9.974191921993366e-05, "loss": 0.0731, "step": 6460 }, { "grad_norm": 0.4248567521572113, "learning_rate": 9.973981770685474e-05, "loss": 0.0727, "step": 6470 }, { "grad_norm": 0.5567407011985779, "learning_rate": 9.97377076946014e-05, "loss": 0.0682, "step": 6480 }, { "grad_norm": 0.6391996741294861, "learning_rate": 9.973558918353412e-05, "loss": 0.0949, "step": 6490 }, { "grad_norm": 0.6600782871246338, "learning_rate": 9.973346217401494e-05, "loss": 0.0728, "step": 6500 }, { "grad_norm": 0.5791072845458984, "learning_rate": 9.973132666640726e-05, "loss": 0.0662, "step": 6510 }, { "grad_norm": 0.5697438716888428, "learning_rate": 9.972918266107602e-05, "loss": 0.0746, "step": 6520 }, { "grad_norm": 0.5476062297821045, "learning_rate": 9.972703015838756e-05, "loss": 0.072, "step": 6530 }, { "grad_norm": 0.534630537033081, "learning_rate": 9.97248691587097e-05, "loss": 0.0789, "step": 6540 }, { "grad_norm": 0.5791231989860535, "learning_rate": 9.972269966241166e-05, "loss": 0.0749, "step": 6550 }, { "grad_norm": 0.5867448449134827, "learning_rate": 9.972052166986417e-05, "loss": 0.0802, "step": 6560 }, { "grad_norm": 0.6257652044296265, "learning_rate": 9.971833518143938e-05, "loss": 0.0903, "step": 6570 }, { "grad_norm": 0.5633111596107483, "learning_rate": 9.971614019751093e-05, "loss": 0.0692, "step": 6580 }, { "grad_norm": 0.6508511304855347, "learning_rate": 9.971393671845383e-05, "loss": 0.084, "step": 6590 }, { "grad_norm": 0.5435343384742737, "learning_rate": 9.971172474464464e-05, "loss": 0.101, "step": 6600 }, { "grad_norm": 0.5429442524909973, "learning_rate": 9.97095042764613e-05, "loss": 0.0838, "step": 6610 }, { "grad_norm": 0.8443180322647095, "learning_rate": 9.970727531428324e-05, "loss": 0.0745, "step": 6620 }, { "grad_norm": 0.7826460599899292, "learning_rate": 9.970503785849132e-05, "loss": 0.0815, "step": 6630 }, { "grad_norm": 0.5323282480239868, "learning_rate": 9.970279190946788e-05, "loss": 0.0678, "step": 6640 }, { "grad_norm": 0.7596765160560608, "learning_rate": 9.970053746759667e-05, "loss": 0.09, "step": 6650 }, { "grad_norm": 0.577897846698761, "learning_rate": 9.969827453326292e-05, "loss": 0.0749, "step": 6660 }, { "grad_norm": 0.5126219987869263, "learning_rate": 9.969600310685332e-05, "loss": 0.0693, "step": 6670 }, { "grad_norm": 0.654761552810669, "learning_rate": 9.969372318875596e-05, "loss": 0.0834, "step": 6680 }, { "grad_norm": 0.5383556485176086, "learning_rate": 9.969143477936043e-05, "loss": 0.0891, "step": 6690 }, { "grad_norm": 0.8019862174987793, "learning_rate": 9.968913787905775e-05, "loss": 0.0871, "step": 6700 }, { "grad_norm": 0.6262308359146118, "learning_rate": 9.968683248824045e-05, "loss": 0.0804, "step": 6710 }, { "grad_norm": 0.6475951671600342, "learning_rate": 9.968451860730238e-05, "loss": 0.0969, "step": 6720 }, { "grad_norm": 0.6583721041679382, "learning_rate": 9.968219623663896e-05, "loss": 0.0832, "step": 6730 }, { "grad_norm": 0.5326792001724243, "learning_rate": 9.967986537664702e-05, "loss": 0.102, "step": 6740 }, { "grad_norm": 0.5565977692604065, "learning_rate": 9.967752602772483e-05, "loss": 0.0779, "step": 6750 }, { "grad_norm": 0.49016037583351135, "learning_rate": 9.967517819027212e-05, "loss": 0.0873, "step": 6760 }, { "grad_norm": 0.8446739912033081, "learning_rate": 9.967282186469009e-05, "loss": 0.0773, "step": 6770 }, { "grad_norm": 0.6177231073379517, "learning_rate": 9.967045705138135e-05, "loss": 0.0837, "step": 6780 }, { "grad_norm": 0.6173117160797119, "learning_rate": 9.966808375074998e-05, "loss": 0.0685, "step": 6790 }, { "grad_norm": 0.46553051471710205, "learning_rate": 9.966570196320154e-05, "loss": 0.0769, "step": 6800 }, { "grad_norm": 0.5338685512542725, "learning_rate": 9.966331168914299e-05, "loss": 0.0668, "step": 6810 }, { "grad_norm": 0.516286313533783, "learning_rate": 9.966091292898277e-05, "loss": 0.0811, "step": 6820 }, { "grad_norm": 0.5690728425979614, "learning_rate": 9.965850568313076e-05, "loss": 0.0757, "step": 6830 }, { "grad_norm": 0.5627686381340027, "learning_rate": 9.965608995199827e-05, "loss": 0.0906, "step": 6840 }, { "grad_norm": 0.9712328314781189, "learning_rate": 9.965366573599812e-05, "loss": 0.0791, "step": 6850 }, { "grad_norm": 0.887399435043335, "learning_rate": 9.965123303554453e-05, "loss": 0.0742, "step": 6860 }, { "grad_norm": 0.5922119617462158, "learning_rate": 9.964879185105317e-05, "loss": 0.0897, "step": 6870 }, { "grad_norm": 0.6617682576179504, "learning_rate": 9.964634218294119e-05, "loss": 0.0946, "step": 6880 }, { "grad_norm": 0.6934698224067688, "learning_rate": 9.964388403162714e-05, "loss": 0.0788, "step": 6890 }, { "grad_norm": 0.7618352770805359, "learning_rate": 9.96414173975311e-05, "loss": 0.0942, "step": 6900 }, { "grad_norm": 0.5789797306060791, "learning_rate": 9.963894228107451e-05, "loss": 0.0751, "step": 6910 }, { "grad_norm": 0.4329642951488495, "learning_rate": 9.963645868268032e-05, "loss": 0.0835, "step": 6920 }, { "grad_norm": 0.5321944952011108, "learning_rate": 9.963396660277289e-05, "loss": 0.085, "step": 6930 }, { "grad_norm": 0.576295793056488, "learning_rate": 9.963146604177807e-05, "loss": 0.0658, "step": 6940 }, { "grad_norm": 0.6184061765670776, "learning_rate": 9.962895700012311e-05, "loss": 0.0794, "step": 6950 }, { "grad_norm": 0.626836895942688, "learning_rate": 9.962643947823677e-05, "loss": 0.0828, "step": 6960 }, { "grad_norm": 0.6582612991333008, "learning_rate": 9.962391347654921e-05, "loss": 0.099, "step": 6970 }, { "grad_norm": 0.6256773471832275, "learning_rate": 9.962137899549204e-05, "loss": 0.0941, "step": 6980 }, { "grad_norm": 0.7043883800506592, "learning_rate": 9.961883603549835e-05, "loss": 0.0771, "step": 6990 }, { "grad_norm": 0.702126145362854, "learning_rate": 9.961628459700267e-05, "loss": 0.0766, "step": 7000 }, { "grad_norm": 0.5891894698143005, "learning_rate": 9.961372468044095e-05, "loss": 0.0905, "step": 7010 }, { "grad_norm": 0.6090518832206726, "learning_rate": 9.961115628625062e-05, "loss": 0.0741, "step": 7020 }, { "grad_norm": 0.7374948263168335, "learning_rate": 9.960857941487056e-05, "loss": 0.0969, "step": 7030 }, { "grad_norm": 0.5550625324249268, "learning_rate": 9.960599406674106e-05, "loss": 0.0741, "step": 7040 }, { "grad_norm": 0.5954139232635498, "learning_rate": 9.960340024230393e-05, "loss": 0.0637, "step": 7050 }, { "grad_norm": 0.5108658671379089, "learning_rate": 9.960079794200232e-05, "loss": 0.0745, "step": 7060 }, { "grad_norm": 0.4743388891220093, "learning_rate": 9.959818716628096e-05, "loss": 0.0735, "step": 7070 }, { "grad_norm": 0.47525879740715027, "learning_rate": 9.95955679155859e-05, "loss": 0.0845, "step": 7080 }, { "grad_norm": 0.7168765664100647, "learning_rate": 9.959294019036472e-05, "loss": 0.0802, "step": 7090 }, { "grad_norm": 0.7662846446037292, "learning_rate": 9.959030399106646e-05, "loss": 0.0868, "step": 7100 }, { "grad_norm": 0.5188718438148499, "learning_rate": 9.958765931814153e-05, "loss": 0.0888, "step": 7110 }, { "grad_norm": 0.6816150546073914, "learning_rate": 9.958500617204184e-05, "loss": 0.0653, "step": 7120 }, { "grad_norm": 0.49062973260879517, "learning_rate": 9.958234455322075e-05, "loss": 0.0735, "step": 7130 }, { "grad_norm": 0.5240134000778198, "learning_rate": 9.957967446213308e-05, "loss": 0.0912, "step": 7140 }, { "grad_norm": 0.6824215650558472, "learning_rate": 9.957699589923501e-05, "loss": 0.0856, "step": 7150 }, { "grad_norm": 0.725627064704895, "learning_rate": 9.957430886498431e-05, "loss": 0.0798, "step": 7160 }, { "grad_norm": 0.6024031639099121, "learning_rate": 9.957161335984008e-05, "loss": 0.0728, "step": 7170 }, { "grad_norm": 0.5169978737831116, "learning_rate": 9.956890938426291e-05, "loss": 0.0631, "step": 7180 }, { "grad_norm": 0.4276851415634155, "learning_rate": 9.956619693871482e-05, "loss": 0.0836, "step": 7190 }, { "grad_norm": 0.6773342490196228, "learning_rate": 9.956347602365934e-05, "loss": 0.084, "step": 7200 }, { "grad_norm": 0.8517842292785645, "learning_rate": 9.956074663956135e-05, "loss": 0.0937, "step": 7210 }, { "grad_norm": 0.5296301245689392, "learning_rate": 9.955800878688726e-05, "loss": 0.0869, "step": 7220 }, { "grad_norm": 0.5865211486816406, "learning_rate": 9.955526246610489e-05, "loss": 0.0835, "step": 7230 }, { "grad_norm": 0.6299800872802734, "learning_rate": 9.955250767768349e-05, "loss": 0.086, "step": 7240 }, { "grad_norm": 0.5645926594734192, "learning_rate": 9.95497444220938e-05, "loss": 0.0763, "step": 7250 }, { "grad_norm": 0.7391597032546997, "learning_rate": 9.954697269980797e-05, "loss": 0.0796, "step": 7260 }, { "grad_norm": 0.5879502892494202, "learning_rate": 9.954419251129962e-05, "loss": 0.0739, "step": 7270 }, { "grad_norm": 0.5604378581047058, "learning_rate": 9.95414038570438e-05, "loss": 0.069, "step": 7280 }, { "grad_norm": 0.5437883138656616, "learning_rate": 9.953860673751703e-05, "loss": 0.0732, "step": 7290 }, { "grad_norm": 0.6358292102813721, "learning_rate": 9.953580115319725e-05, "loss": 0.0711, "step": 7300 }, { "grad_norm": 0.6106297373771667, "learning_rate": 9.953298710456387e-05, "loss": 0.0848, "step": 7310 }, { "grad_norm": 0.6180039644241333, "learning_rate": 9.953016459209771e-05, "loss": 0.0822, "step": 7320 }, { "grad_norm": 0.6439800262451172, "learning_rate": 9.952733361628108e-05, "loss": 0.0779, "step": 7330 }, { "grad_norm": 0.5294616222381592, "learning_rate": 9.952449417759772e-05, "loss": 0.0836, "step": 7340 }, { "grad_norm": 0.6064625382423401, "learning_rate": 9.952164627653279e-05, "loss": 0.071, "step": 7350 }, { "grad_norm": 0.504880428314209, "learning_rate": 9.951878991357292e-05, "loss": 0.0809, "step": 7360 }, { "grad_norm": 0.6438607573509216, "learning_rate": 9.951592508920622e-05, "loss": 0.0891, "step": 7370 }, { "grad_norm": 0.5234296321868896, "learning_rate": 9.951305180392219e-05, "loss": 0.0712, "step": 7380 }, { "grad_norm": 0.5800169706344604, "learning_rate": 9.951017005821178e-05, "loss": 0.0818, "step": 7390 }, { "grad_norm": 0.5378853678703308, "learning_rate": 9.95072798525674e-05, "loss": 0.0732, "step": 7400 }, { "grad_norm": 0.5393425226211548, "learning_rate": 9.950438118748293e-05, "loss": 0.0747, "step": 7410 }, { "grad_norm": 0.5493202209472656, "learning_rate": 9.950147406345366e-05, "loss": 0.081, "step": 7420 }, { "grad_norm": 0.7139249444007874, "learning_rate": 9.949855848097635e-05, "loss": 0.0829, "step": 7430 }, { "grad_norm": 0.7116929888725281, "learning_rate": 9.949563444054916e-05, "loss": 0.0883, "step": 7440 }, { "grad_norm": 0.553570568561554, "learning_rate": 9.949270194267178e-05, "loss": 0.0719, "step": 7450 }, { "grad_norm": 0.4794083535671234, "learning_rate": 9.948976098784526e-05, "loss": 0.0723, "step": 7460 }, { "grad_norm": 0.5328785181045532, "learning_rate": 9.948681157657213e-05, "loss": 0.0842, "step": 7470 }, { "grad_norm": 0.7829962372779846, "learning_rate": 9.948385370935638e-05, "loss": 0.0958, "step": 7480 }, { "grad_norm": 0.7546810507774353, "learning_rate": 9.94808873867034e-05, "loss": 0.082, "step": 7490 }, { "grad_norm": 0.6584804654121399, "learning_rate": 9.947791260912009e-05, "loss": 0.0752, "step": 7500 }, { "grad_norm": 0.4817871153354645, "learning_rate": 9.947492937711474e-05, "loss": 0.0643, "step": 7510 }, { "grad_norm": 0.47018513083457947, "learning_rate": 9.947193769119707e-05, "loss": 0.0758, "step": 7520 }, { "grad_norm": 0.5537567138671875, "learning_rate": 9.946893755187834e-05, "loss": 0.0714, "step": 7530 }, { "grad_norm": 0.4271013140678406, "learning_rate": 9.946592895967115e-05, "loss": 0.0733, "step": 7540 }, { "grad_norm": 0.4515814185142517, "learning_rate": 9.94629119150896e-05, "loss": 0.0621, "step": 7550 }, { "grad_norm": 0.6678693294525146, "learning_rate": 9.94598864186492e-05, "loss": 0.0705, "step": 7560 }, { "grad_norm": 0.516845703125, "learning_rate": 9.945685247086696e-05, "loss": 0.0838, "step": 7570 }, { "grad_norm": 0.5802726149559021, "learning_rate": 9.945381007226129e-05, "loss": 0.0877, "step": 7580 }, { "grad_norm": 0.4783889055252075, "learning_rate": 9.945075922335203e-05, "loss": 0.0704, "step": 7590 }, { "grad_norm": 0.5927579402923584, "learning_rate": 9.944769992466049e-05, "loss": 0.089, "step": 7600 }, { "grad_norm": 0.5019915699958801, "learning_rate": 9.944463217670945e-05, "loss": 0.0803, "step": 7610 }, { "grad_norm": 0.49065667390823364, "learning_rate": 9.944155598002307e-05, "loss": 0.0725, "step": 7620 }, { "grad_norm": 0.4648829996585846, "learning_rate": 9.943847133512701e-05, "loss": 0.0684, "step": 7630 }, { "grad_norm": 0.5961100459098816, "learning_rate": 9.943537824254834e-05, "loss": 0.0663, "step": 7640 }, { "grad_norm": 0.437222957611084, "learning_rate": 9.943227670281559e-05, "loss": 0.0689, "step": 7650 }, { "grad_norm": 0.4754517078399658, "learning_rate": 9.942916671645873e-05, "loss": 0.0601, "step": 7660 }, { "grad_norm": 0.48666542768478394, "learning_rate": 9.942604828400916e-05, "loss": 0.0778, "step": 7670 }, { "grad_norm": 0.6504423022270203, "learning_rate": 9.942292140599975e-05, "loss": 0.0743, "step": 7680 }, { "grad_norm": 0.4074263870716095, "learning_rate": 9.94197860829648e-05, "loss": 0.072, "step": 7690 }, { "grad_norm": 0.546238899230957, "learning_rate": 9.941664231544004e-05, "loss": 0.0634, "step": 7700 }, { "grad_norm": 0.5568965673446655, "learning_rate": 9.941349010396264e-05, "loss": 0.0879, "step": 7710 }, { "grad_norm": 0.6110197305679321, "learning_rate": 9.941032944907125e-05, "loss": 0.08, "step": 7720 }, { "grad_norm": 0.4494459927082062, "learning_rate": 9.940716035130596e-05, "loss": 0.0741, "step": 7730 }, { "grad_norm": 0.4940345883369446, "learning_rate": 9.940398281120821e-05, "loss": 0.0766, "step": 7740 }, { "grad_norm": 0.4358576834201813, "learning_rate": 9.940079682932102e-05, "loss": 0.0643, "step": 7750 }, { "grad_norm": 0.4967268705368042, "learning_rate": 9.939760240618877e-05, "loss": 0.0753, "step": 7760 }, { "grad_norm": 0.6038383841514587, "learning_rate": 9.939439954235729e-05, "loss": 0.0835, "step": 7770 }, { "grad_norm": 0.5755693912506104, "learning_rate": 9.939118823837387e-05, "loss": 0.0887, "step": 7780 }, { "grad_norm": 0.6131347417831421, "learning_rate": 9.938796849478725e-05, "loss": 0.0797, "step": 7790 }, { "grad_norm": 0.5084652900695801, "learning_rate": 9.938474031214755e-05, "loss": 0.0837, "step": 7800 }, { "grad_norm": 0.7179185152053833, "learning_rate": 9.938150369100643e-05, "loss": 0.0859, "step": 7810 }, { "grad_norm": 0.4404284954071045, "learning_rate": 9.93782586319169e-05, "loss": 0.0661, "step": 7820 }, { "grad_norm": 0.675346851348877, "learning_rate": 9.937500513543348e-05, "loss": 0.0763, "step": 7830 }, { "grad_norm": 0.5501136183738708, "learning_rate": 9.937174320211207e-05, "loss": 0.0779, "step": 7840 }, { "grad_norm": 0.4938749074935913, "learning_rate": 9.936847283251009e-05, "loss": 0.0777, "step": 7850 }, { "grad_norm": 0.6494969129562378, "learning_rate": 9.936519402718632e-05, "loss": 0.0767, "step": 7860 }, { "grad_norm": 0.5523350834846497, "learning_rate": 9.936190678670102e-05, "loss": 0.0846, "step": 7870 }, { "grad_norm": 0.4806918799877167, "learning_rate": 9.935861111161593e-05, "loss": 0.08, "step": 7880 }, { "grad_norm": 0.5385254621505737, "learning_rate": 9.935530700249416e-05, "loss": 0.0668, "step": 7890 }, { "grad_norm": 0.5241627097129822, "learning_rate": 9.935199445990028e-05, "loss": 0.0878, "step": 7900 }, { "grad_norm": 0.5000324249267578, "learning_rate": 9.934867348440033e-05, "loss": 0.0828, "step": 7910 }, { "grad_norm": 0.5535463094711304, "learning_rate": 9.934534407656176e-05, "loss": 0.1049, "step": 7920 }, { "grad_norm": 0.6068828105926514, "learning_rate": 9.93420062369535e-05, "loss": 0.0733, "step": 7930 }, { "grad_norm": 0.5254327654838562, "learning_rate": 9.933865996614589e-05, "loss": 0.0749, "step": 7940 }, { "grad_norm": 0.4975655674934387, "learning_rate": 9.933530526471068e-05, "loss": 0.0753, "step": 7950 }, { "grad_norm": 0.4780024588108063, "learning_rate": 9.933194213322114e-05, "loss": 0.068, "step": 7960 }, { "grad_norm": 0.506361186504364, "learning_rate": 9.932857057225192e-05, "loss": 0.0756, "step": 7970 }, { "grad_norm": 0.48174211382865906, "learning_rate": 9.932519058237912e-05, "loss": 0.0682, "step": 7980 }, { "grad_norm": 0.5110396146774292, "learning_rate": 9.932180216418032e-05, "loss": 0.0766, "step": 7990 }, { "grad_norm": 0.4600239396095276, "learning_rate": 9.931840531823446e-05, "loss": 0.0698, "step": 8000 }, { "grad_norm": 0.5897713303565979, "learning_rate": 9.9315000045122e-05, "loss": 0.0684, "step": 8010 }, { "grad_norm": 0.5294606685638428, "learning_rate": 9.931158634542481e-05, "loss": 0.068, "step": 8020 }, { "grad_norm": 0.3419560194015503, "learning_rate": 9.930816421972617e-05, "loss": 0.0649, "step": 8030 }, { "grad_norm": 0.6536368131637573, "learning_rate": 9.930473366861086e-05, "loss": 0.0627, "step": 8040 }, { "grad_norm": 0.39220818877220154, "learning_rate": 9.930129469266505e-05, "loss": 0.0683, "step": 8050 }, { "grad_norm": 0.5838785171508789, "learning_rate": 9.929784729247638e-05, "loss": 0.0642, "step": 8060 }, { "grad_norm": 0.5132438540458679, "learning_rate": 9.929439146863389e-05, "loss": 0.063, "step": 8070 }, { "grad_norm": 0.7103028893470764, "learning_rate": 9.92909272217281e-05, "loss": 0.0872, "step": 8080 }, { "grad_norm": 0.428720623254776, "learning_rate": 9.928745455235097e-05, "loss": 0.0713, "step": 8090 }, { "grad_norm": 0.5103150010108948, "learning_rate": 9.928397346109588e-05, "loss": 0.0757, "step": 8100 }, { "grad_norm": 0.582655668258667, "learning_rate": 9.928048394855762e-05, "loss": 0.0765, "step": 8110 }, { "grad_norm": 0.49834296107292175, "learning_rate": 9.92769860153325e-05, "loss": 0.0788, "step": 8120 }, { "grad_norm": 0.7876986265182495, "learning_rate": 9.927347966201819e-05, "loss": 0.0893, "step": 8130 }, { "grad_norm": 0.615111231803894, "learning_rate": 9.926996488921383e-05, "loss": 0.0852, "step": 8140 }, { "grad_norm": 0.5312209129333496, "learning_rate": 9.926644169752001e-05, "loss": 0.0694, "step": 8150 }, { "grad_norm": 0.5380465388298035, "learning_rate": 9.926291008753875e-05, "loss": 0.0839, "step": 8160 }, { "grad_norm": 0.4210248589515686, "learning_rate": 9.92593700598735e-05, "loss": 0.0792, "step": 8170 }, { "grad_norm": 0.7208216786384583, "learning_rate": 9.925582161512915e-05, "loss": 0.0701, "step": 8180 }, { "grad_norm": 0.611214816570282, "learning_rate": 9.925226475391205e-05, "loss": 0.0753, "step": 8190 }, { "grad_norm": 0.6348235607147217, "learning_rate": 9.924869947682993e-05, "loss": 0.0662, "step": 8200 }, { "grad_norm": 0.3264731168746948, "learning_rate": 9.924512578449204e-05, "loss": 0.0613, "step": 8210 }, { "grad_norm": 0.5061949491500854, "learning_rate": 9.924154367750901e-05, "loss": 0.0683, "step": 8220 }, { "grad_norm": 0.7734207510948181, "learning_rate": 9.923795315649293e-05, "loss": 0.0756, "step": 8230 }, { "grad_norm": 0.5907043814659119, "learning_rate": 9.92343542220573e-05, "loss": 0.0742, "step": 8240 }, { "grad_norm": 0.5351845026016235, "learning_rate": 9.92307468748171e-05, "loss": 0.073, "step": 8250 }, { "grad_norm": 0.5879158973693848, "learning_rate": 9.922713111538873e-05, "loss": 0.0733, "step": 8260 }, { "grad_norm": 0.5222053527832031, "learning_rate": 9.922350694439003e-05, "loss": 0.0662, "step": 8270 }, { "grad_norm": 0.5050473809242249, "learning_rate": 9.921987436244024e-05, "loss": 0.0744, "step": 8280 }, { "grad_norm": 0.5178161859512329, "learning_rate": 9.921623337016008e-05, "loss": 0.074, "step": 8290 }, { "grad_norm": 0.3676506280899048, "learning_rate": 9.921258396817172e-05, "loss": 0.0789, "step": 8300 }, { "grad_norm": 0.44654151797294617, "learning_rate": 9.920892615709874e-05, "loss": 0.0737, "step": 8310 }, { "grad_norm": 0.7261173725128174, "learning_rate": 9.920525993756612e-05, "loss": 0.0695, "step": 8320 }, { "grad_norm": 0.42489534616470337, "learning_rate": 9.920158531020036e-05, "loss": 0.0713, "step": 8330 }, { "grad_norm": 0.5825843811035156, "learning_rate": 9.919790227562933e-05, "loss": 0.0597, "step": 8340 }, { "grad_norm": 0.5623264908790588, "learning_rate": 9.919421083448237e-05, "loss": 0.0828, "step": 8350 }, { "grad_norm": 0.5851744413375854, "learning_rate": 9.919051098739022e-05, "loss": 0.0582, "step": 8360 }, { "grad_norm": 0.511134147644043, "learning_rate": 9.918680273498514e-05, "loss": 0.0621, "step": 8370 }, { "grad_norm": 0.5255844593048096, "learning_rate": 9.918308607790072e-05, "loss": 0.0646, "step": 8380 }, { "grad_norm": 0.43751171231269836, "learning_rate": 9.917936101677205e-05, "loss": 0.0601, "step": 8390 }, { "grad_norm": 0.4173765480518341, "learning_rate": 9.917562755223564e-05, "loss": 0.0704, "step": 8400 }, { "grad_norm": 0.4687918722629547, "learning_rate": 9.917188568492944e-05, "loss": 0.0712, "step": 8410 }, { "grad_norm": 0.40938591957092285, "learning_rate": 9.916813541549283e-05, "loss": 0.0671, "step": 8420 }, { "grad_norm": 0.6395427584648132, "learning_rate": 9.916437674456663e-05, "loss": 0.0678, "step": 8430 }, { "grad_norm": 0.5449535250663757, "learning_rate": 9.916060967279308e-05, "loss": 0.0685, "step": 8440 }, { "grad_norm": 0.5503254532814026, "learning_rate": 9.91568342008159e-05, "loss": 0.068, "step": 8450 }, { "grad_norm": 0.41232389211654663, "learning_rate": 9.915305032928019e-05, "loss": 0.062, "step": 8460 }, { "grad_norm": 0.40275222063064575, "learning_rate": 9.914925805883253e-05, "loss": 0.059, "step": 8470 }, { "grad_norm": 0.44665074348449707, "learning_rate": 9.914545739012088e-05, "loss": 0.0604, "step": 8480 }, { "grad_norm": 0.6611142158508301, "learning_rate": 9.91416483237947e-05, "loss": 0.0687, "step": 8490 }, { "grad_norm": 0.5017679929733276, "learning_rate": 9.913783086050485e-05, "loss": 0.0694, "step": 8500 }, { "grad_norm": 0.5797958374023438, "learning_rate": 9.913400500090364e-05, "loss": 0.0664, "step": 8510 }, { "grad_norm": 0.5785658359527588, "learning_rate": 9.913017074564479e-05, "loss": 0.063, "step": 8520 }, { "grad_norm": 0.40601539611816406, "learning_rate": 9.912632809538348e-05, "loss": 0.0741, "step": 8530 }, { "grad_norm": 0.48877808451652527, "learning_rate": 9.912247705077629e-05, "loss": 0.0622, "step": 8540 }, { "grad_norm": 0.5853069424629211, "learning_rate": 9.911861761248127e-05, "loss": 0.0755, "step": 8550 }, { "grad_norm": 0.442500501871109, "learning_rate": 9.91147497811579e-05, "loss": 0.0611, "step": 8560 }, { "grad_norm": 0.6065354347229004, "learning_rate": 9.911087355746709e-05, "loss": 0.0952, "step": 8570 }, { "grad_norm": 0.48845237493515015, "learning_rate": 9.910698894207117e-05, "loss": 0.0811, "step": 8580 }, { "grad_norm": 0.5177421569824219, "learning_rate": 9.910309593563392e-05, "loss": 0.0802, "step": 8590 }, { "grad_norm": 0.4258434772491455, "learning_rate": 9.909919453882057e-05, "loss": 0.0644, "step": 8600 }, { "grad_norm": 0.5775660872459412, "learning_rate": 9.90952847522977e-05, "loss": 0.0586, "step": 8610 }, { "grad_norm": 0.5959863066673279, "learning_rate": 9.909136657673346e-05, "loss": 0.0802, "step": 8620 }, { "grad_norm": 0.41902050375938416, "learning_rate": 9.908744001279731e-05, "loss": 0.0647, "step": 8630 }, { "grad_norm": 0.5621774196624756, "learning_rate": 9.90835050611602e-05, "loss": 0.0782, "step": 8640 }, { "grad_norm": 0.44408777356147766, "learning_rate": 9.90795617224945e-05, "loss": 0.0705, "step": 8650 }, { "grad_norm": 0.7510326504707336, "learning_rate": 9.907560999747405e-05, "loss": 0.0664, "step": 8660 }, { "grad_norm": 0.464748352766037, "learning_rate": 9.907164988677408e-05, "loss": 0.0612, "step": 8670 }, { "grad_norm": 0.41710466146469116, "learning_rate": 9.906768139107124e-05, "loss": 0.0605, "step": 8680 }, { "grad_norm": 0.3664914071559906, "learning_rate": 9.906370451104367e-05, "loss": 0.0607, "step": 8690 }, { "grad_norm": 0.4136778712272644, "learning_rate": 9.905971924737088e-05, "loss": 0.0596, "step": 8700 }, { "grad_norm": 0.42648765444755554, "learning_rate": 9.905572560073387e-05, "loss": 0.0653, "step": 8710 }, { "grad_norm": 0.8134230971336365, "learning_rate": 9.905172357181501e-05, "loss": 0.0914, "step": 8720 }, { "grad_norm": 0.42982423305511475, "learning_rate": 9.904771316129817e-05, "loss": 0.0708, "step": 8730 }, { "grad_norm": 0.5396020412445068, "learning_rate": 9.904369436986862e-05, "loss": 0.0724, "step": 8740 }, { "grad_norm": 0.5685049295425415, "learning_rate": 9.903966719821303e-05, "loss": 0.0792, "step": 8750 }, { "grad_norm": 0.6095120310783386, "learning_rate": 9.903563164701956e-05, "loss": 0.0746, "step": 8760 }, { "grad_norm": 0.5686846971511841, "learning_rate": 9.903158771697778e-05, "loss": 0.0723, "step": 8770 }, { "grad_norm": 0.6054487824440002, "learning_rate": 9.902753540877867e-05, "loss": 0.0753, "step": 8780 }, { "grad_norm": 0.4343457818031311, "learning_rate": 9.902347472311466e-05, "loss": 0.0627, "step": 8790 }, { "grad_norm": 0.46223384141921997, "learning_rate": 9.901940566067962e-05, "loss": 0.0555, "step": 8800 }, { "grad_norm": 0.41128039360046387, "learning_rate": 9.901532822216883e-05, "loss": 0.0529, "step": 8810 }, { "grad_norm": 0.4356979727745056, "learning_rate": 9.901124240827904e-05, "loss": 0.0725, "step": 8820 }, { "grad_norm": 0.5240651965141296, "learning_rate": 9.900714821970835e-05, "loss": 0.0638, "step": 8830 }, { "grad_norm": 0.5857155323028564, "learning_rate": 9.900304565715641e-05, "loss": 0.0791, "step": 8840 }, { "grad_norm": 0.5443987846374512, "learning_rate": 9.899893472132419e-05, "loss": 0.074, "step": 8850 }, { "grad_norm": 0.5177403092384338, "learning_rate": 9.899481541291415e-05, "loss": 0.0744, "step": 8860 }, { "grad_norm": 1.0090142488479614, "learning_rate": 9.899068773263016e-05, "loss": 0.0764, "step": 8870 }, { "grad_norm": 0.41456490755081177, "learning_rate": 9.898655168117754e-05, "loss": 0.0722, "step": 8880 }, { "grad_norm": 0.4277929365634918, "learning_rate": 9.898240725926302e-05, "loss": 0.0704, "step": 8890 }, { "grad_norm": 0.38684573769569397, "learning_rate": 9.897825446759478e-05, "loss": 0.0729, "step": 8900 }, { "grad_norm": 0.46542826294898987, "learning_rate": 9.897409330688241e-05, "loss": 0.0665, "step": 8910 }, { "grad_norm": 0.6101300120353699, "learning_rate": 9.896992377783692e-05, "loss": 0.0779, "step": 8920 }, { "grad_norm": 0.4978196918964386, "learning_rate": 9.89657458811708e-05, "loss": 0.0575, "step": 8930 }, { "grad_norm": 0.4265516400337219, "learning_rate": 9.896155961759792e-05, "loss": 0.0694, "step": 8940 }, { "grad_norm": 0.44903284311294556, "learning_rate": 9.895736498783361e-05, "loss": 0.0681, "step": 8950 }, { "grad_norm": 0.5076518654823303, "learning_rate": 9.895316199259462e-05, "loss": 0.0681, "step": 8960 }, { "grad_norm": 0.48139527440071106, "learning_rate": 9.894895063259909e-05, "loss": 0.0553, "step": 8970 }, { "grad_norm": 0.6551302075386047, "learning_rate": 9.894473090856667e-05, "loss": 0.0671, "step": 8980 }, { "grad_norm": 0.6767544150352478, "learning_rate": 9.894050282121839e-05, "loss": 0.0901, "step": 8990 }, { "grad_norm": 0.4141164720058441, "learning_rate": 9.893626637127668e-05, "loss": 0.0762, "step": 9000 }, { "grad_norm": 0.4255976676940918, "learning_rate": 9.893202155946546e-05, "loss": 0.0606, "step": 9010 }, { "grad_norm": 0.5988855361938477, "learning_rate": 9.892776838651006e-05, "loss": 0.0637, "step": 9020 }, { "grad_norm": 0.6005128026008606, "learning_rate": 9.892350685313722e-05, "loss": 0.0805, "step": 9030 }, { "grad_norm": 0.44376248121261597, "learning_rate": 9.891923696007513e-05, "loss": 0.0729, "step": 9040 }, { "grad_norm": 0.35520127415657043, "learning_rate": 9.891495870805336e-05, "loss": 0.0557, "step": 9050 }, { "grad_norm": 0.48309221863746643, "learning_rate": 9.891067209780298e-05, "loss": 0.0747, "step": 9060 }, { "grad_norm": 0.41271379590034485, "learning_rate": 9.890637713005646e-05, "loss": 0.0599, "step": 9070 }, { "grad_norm": 0.5345275402069092, "learning_rate": 9.890207380554767e-05, "loss": 0.0643, "step": 9080 }, { "grad_norm": 0.4146653115749359, "learning_rate": 9.889776212501196e-05, "loss": 0.067, "step": 9090 }, { "grad_norm": 0.48601680994033813, "learning_rate": 9.889344208918605e-05, "loss": 0.0792, "step": 9100 }, { "grad_norm": 0.6131239533424377, "learning_rate": 9.888911369880812e-05, "loss": 0.0694, "step": 9110 }, { "grad_norm": 0.546017587184906, "learning_rate": 9.888477695461777e-05, "loss": 0.0625, "step": 9120 }, { "grad_norm": 0.37981677055358887, "learning_rate": 9.888043185735607e-05, "loss": 0.0691, "step": 9130 }, { "grad_norm": 0.5866047143936157, "learning_rate": 9.887607840776542e-05, "loss": 0.0669, "step": 9140 }, { "grad_norm": 0.48158279061317444, "learning_rate": 9.887171660658975e-05, "loss": 0.0751, "step": 9150 }, { "grad_norm": 0.3893205523490906, "learning_rate": 9.886734645457435e-05, "loss": 0.0527, "step": 9160 }, { "grad_norm": 0.5876193046569824, "learning_rate": 9.886296795246597e-05, "loss": 0.0606, "step": 9170 }, { "grad_norm": 0.4250008761882782, "learning_rate": 9.885858110101276e-05, "loss": 0.0565, "step": 9180 }, { "grad_norm": 0.40307390689849854, "learning_rate": 9.885418590096434e-05, "loss": 0.052, "step": 9190 }, { "grad_norm": 0.6546624302864075, "learning_rate": 9.88497823530717e-05, "loss": 0.0637, "step": 9200 }, { "grad_norm": 0.3923279047012329, "learning_rate": 9.884537045808732e-05, "loss": 0.077, "step": 9210 }, { "grad_norm": 0.9629393219947815, "learning_rate": 9.884095021676502e-05, "loss": 0.0805, "step": 9220 }, { "grad_norm": 0.6723442673683167, "learning_rate": 9.883652162986017e-05, "loss": 0.0702, "step": 9230 }, { "grad_norm": 0.46354588866233826, "learning_rate": 9.883208469812943e-05, "loss": 0.0759, "step": 9240 }, { "grad_norm": 0.6925092339515686, "learning_rate": 9.882763942233098e-05, "loss": 0.076, "step": 9250 }, { "grad_norm": 0.5100873708724976, "learning_rate": 9.882318580322441e-05, "loss": 0.0679, "step": 9260 }, { "grad_norm": 0.538996160030365, "learning_rate": 9.881872384157067e-05, "loss": 0.0735, "step": 9270 }, { "grad_norm": 0.5385605096817017, "learning_rate": 9.881425353813225e-05, "loss": 0.075, "step": 9280 }, { "grad_norm": 0.5665311813354492, "learning_rate": 9.880977489367296e-05, "loss": 0.0781, "step": 9290 }, { "grad_norm": 0.4152820110321045, "learning_rate": 9.88052879089581e-05, "loss": 0.0799, "step": 9300 }, { "grad_norm": 0.431931734085083, "learning_rate": 9.880079258475434e-05, "loss": 0.0571, "step": 9310 }, { "grad_norm": 0.40741243958473206, "learning_rate": 9.879628892182985e-05, "loss": 0.0807, "step": 9320 }, { "grad_norm": 0.4724070727825165, "learning_rate": 9.879177692095416e-05, "loss": 0.0735, "step": 9330 }, { "grad_norm": 0.5877304077148438, "learning_rate": 9.878725658289825e-05, "loss": 0.0715, "step": 9340 }, { "grad_norm": 0.4292556345462799, "learning_rate": 9.878272790843454e-05, "loss": 0.0637, "step": 9350 }, { "grad_norm": 0.4480270445346832, "learning_rate": 9.877819089833682e-05, "loss": 0.0593, "step": 9360 }, { "grad_norm": 0.6365513205528259, "learning_rate": 9.877364555338038e-05, "loss": 0.0785, "step": 9370 }, { "grad_norm": 0.5863914489746094, "learning_rate": 9.876909187434186e-05, "loss": 0.0824, "step": 9380 }, { "grad_norm": 0.525239109992981, "learning_rate": 9.876452986199939e-05, "loss": 0.0681, "step": 9390 }, { "grad_norm": 0.5213145017623901, "learning_rate": 9.875995951713248e-05, "loss": 0.0779, "step": 9400 }, { "grad_norm": 0.4612201452255249, "learning_rate": 9.875538084052207e-05, "loss": 0.0726, "step": 9410 }, { "grad_norm": 0.6503954529762268, "learning_rate": 9.875079383295053e-05, "loss": 0.0717, "step": 9420 }, { "grad_norm": 0.4512424170970917, "learning_rate": 9.874619849520167e-05, "loss": 0.0588, "step": 9430 }, { "grad_norm": 0.5291045308113098, "learning_rate": 9.874159482806069e-05, "loss": 0.0605, "step": 9440 }, { "grad_norm": 0.6714299917221069, "learning_rate": 9.873698283231426e-05, "loss": 0.062, "step": 9450 }, { "grad_norm": 0.4409976005554199, "learning_rate": 9.87323625087504e-05, "loss": 0.0673, "step": 9460 }, { "grad_norm": 0.5546119809150696, "learning_rate": 9.872773385815863e-05, "loss": 0.078, "step": 9470 }, { "grad_norm": 0.6340759992599487, "learning_rate": 9.872309688132986e-05, "loss": 0.0675, "step": 9480 }, { "grad_norm": 0.4595271348953247, "learning_rate": 9.871845157905639e-05, "loss": 0.0657, "step": 9490 }, { "grad_norm": 0.5053815841674805, "learning_rate": 9.871379795213201e-05, "loss": 0.0571, "step": 9500 }, { "grad_norm": 0.5362878441810608, "learning_rate": 9.87091360013519e-05, "loss": 0.0719, "step": 9510 }, { "grad_norm": 0.4862179458141327, "learning_rate": 9.870446572751262e-05, "loss": 0.0739, "step": 9520 }, { "grad_norm": 0.45025500655174255, "learning_rate": 9.869978713141224e-05, "loss": 0.0642, "step": 9530 }, { "grad_norm": 0.6639398336410522, "learning_rate": 9.869510021385016e-05, "loss": 0.0758, "step": 9540 }, { "grad_norm": 0.4317772686481476, "learning_rate": 9.869040497562727e-05, "loss": 0.0578, "step": 9550 }, { "grad_norm": 0.5382212400436401, "learning_rate": 9.868570141754587e-05, "loss": 0.0673, "step": 9560 }, { "grad_norm": 0.5382785201072693, "learning_rate": 9.868098954040965e-05, "loss": 0.0758, "step": 9570 }, { "grad_norm": 0.523288369178772, "learning_rate": 9.867626934502374e-05, "loss": 0.0655, "step": 9580 }, { "grad_norm": 0.5377297401428223, "learning_rate": 9.86715408321947e-05, "loss": 0.0715, "step": 9590 }, { "grad_norm": 0.5680006146430969, "learning_rate": 9.86668040027305e-05, "loss": 0.0621, "step": 9600 }, { "grad_norm": 0.4325830638408661, "learning_rate": 9.866205885744053e-05, "loss": 0.0661, "step": 9610 }, { "grad_norm": 0.580685019493103, "learning_rate": 9.865730539713563e-05, "loss": 0.078, "step": 9620 }, { "grad_norm": 0.4629313051700592, "learning_rate": 9.8652543622628e-05, "loss": 0.0717, "step": 9630 }, { "grad_norm": 0.41758012771606445, "learning_rate": 9.864777353473132e-05, "loss": 0.0603, "step": 9640 }, { "grad_norm": 0.4581223130226135, "learning_rate": 9.864299513426068e-05, "loss": 0.0613, "step": 9650 }, { "grad_norm": 0.5093326568603516, "learning_rate": 9.863820842203254e-05, "loss": 0.0682, "step": 9660 }, { "grad_norm": 0.6064857840538025, "learning_rate": 9.863341339886483e-05, "loss": 0.083, "step": 9670 }, { "grad_norm": 0.5360509157180786, "learning_rate": 9.86286100655769e-05, "loss": 0.0707, "step": 9680 }, { "grad_norm": 0.6415427923202515, "learning_rate": 9.862379842298953e-05, "loss": 0.0846, "step": 9690 }, { "grad_norm": 0.6582838296890259, "learning_rate": 9.861897847192485e-05, "loss": 0.0673, "step": 9700 }, { "grad_norm": 0.5848968029022217, "learning_rate": 9.86141502132065e-05, "loss": 0.0688, "step": 9710 }, { "grad_norm": 0.5979533195495605, "learning_rate": 9.860931364765946e-05, "loss": 0.0657, "step": 9720 }, { "grad_norm": 0.489179402589798, "learning_rate": 9.860446877611021e-05, "loss": 0.081, "step": 9730 }, { "grad_norm": 0.42518261075019836, "learning_rate": 9.859961559938655e-05, "loss": 0.0719, "step": 9740 }, { "grad_norm": 0.6087596416473389, "learning_rate": 9.85947541183178e-05, "loss": 0.0725, "step": 9750 }, { "grad_norm": 0.5844334959983826, "learning_rate": 9.858988433373463e-05, "loss": 0.0681, "step": 9760 }, { "grad_norm": 0.6599464416503906, "learning_rate": 9.858500624646918e-05, "loss": 0.0692, "step": 9770 }, { "grad_norm": 0.4491422474384308, "learning_rate": 9.858011985735497e-05, "loss": 0.0635, "step": 9780 }, { "grad_norm": 0.4760487675666809, "learning_rate": 9.857522516722693e-05, "loss": 0.0633, "step": 9790 }, { "grad_norm": 0.5263728499412537, "learning_rate": 9.857032217692145e-05, "loss": 0.0595, "step": 9800 }, { "grad_norm": 0.546788215637207, "learning_rate": 9.856541088727631e-05, "loss": 0.0654, "step": 9810 }, { "grad_norm": 0.49590855836868286, "learning_rate": 9.856049129913072e-05, "loss": 0.0767, "step": 9820 }, { "grad_norm": 0.4554649293422699, "learning_rate": 9.85555634133253e-05, "loss": 0.0745, "step": 9830 }, { "grad_norm": 0.3591163456439972, "learning_rate": 9.855062723070208e-05, "loss": 0.0737, "step": 9840 }, { "grad_norm": 0.5612030029296875, "learning_rate": 9.854568275210454e-05, "loss": 0.0691, "step": 9850 }, { "grad_norm": 0.48468640446662903, "learning_rate": 9.854072997837754e-05, "loss": 0.0563, "step": 9860 }, { "grad_norm": 0.5175195336341858, "learning_rate": 9.853576891036737e-05, "loss": 0.062, "step": 9870 }, { "grad_norm": 0.6684340834617615, "learning_rate": 9.853079954892177e-05, "loss": 0.0801, "step": 9880 }, { "grad_norm": 0.37183481454849243, "learning_rate": 9.852582189488983e-05, "loss": 0.0643, "step": 9890 }, { "grad_norm": 0.6393813490867615, "learning_rate": 9.852083594912212e-05, "loss": 0.0777, "step": 9900 }, { "grad_norm": 0.39517199993133545, "learning_rate": 9.851584171247058e-05, "loss": 0.0736, "step": 9910 }, { "grad_norm": 0.4990461766719818, "learning_rate": 9.851083918578863e-05, "loss": 0.0788, "step": 9920 }, { "grad_norm": 0.5117393732070923, "learning_rate": 9.850582836993103e-05, "loss": 0.0588, "step": 9930 }, { "grad_norm": 0.50581294298172, "learning_rate": 9.850080926575397e-05, "loss": 0.0677, "step": 9940 }, { "grad_norm": 0.44389814138412476, "learning_rate": 9.849578187411515e-05, "loss": 0.0664, "step": 9950 }, { "grad_norm": 0.40350377559661865, "learning_rate": 9.849074619587354e-05, "loss": 0.0569, "step": 9960 }, { "grad_norm": 0.42525535821914673, "learning_rate": 9.848570223188964e-05, "loss": 0.0643, "step": 9970 }, { "grad_norm": 0.4563453793525696, "learning_rate": 9.848064998302531e-05, "loss": 0.0627, "step": 9980 }, { "grad_norm": 0.8103851079940796, "learning_rate": 9.847558945014386e-05, "loss": 0.0692, "step": 9990 }, { "grad_norm": 0.6957834362983704, "learning_rate": 9.847052063410996e-05, "loss": 0.0799, "step": 10000 }, { "grad_norm": 0.4669095277786255, "learning_rate": 9.846544353578977e-05, "loss": 0.0686, "step": 10010 }, { "grad_norm": 0.4026638865470886, "learning_rate": 9.846035815605081e-05, "loss": 0.0679, "step": 10020 }, { "grad_norm": 1.0156984329223633, "learning_rate": 9.845526449576204e-05, "loss": 0.058, "step": 10030 }, { "grad_norm": 0.5948846936225891, "learning_rate": 9.845016255579383e-05, "loss": 0.0668, "step": 10040 }, { "grad_norm": 0.4538783133029938, "learning_rate": 9.844505233701794e-05, "loss": 0.0656, "step": 10050 }, { "grad_norm": 0.5452332496643066, "learning_rate": 9.843993384030757e-05, "loss": 0.069, "step": 10060 }, { "grad_norm": 0.5180734992027283, "learning_rate": 9.843480706653737e-05, "loss": 0.0535, "step": 10070 }, { "grad_norm": 0.45540088415145874, "learning_rate": 9.84296720165833e-05, "loss": 0.07, "step": 10080 }, { "grad_norm": 0.8065170049667358, "learning_rate": 9.842452869132286e-05, "loss": 0.0674, "step": 10090 }, { "grad_norm": 0.548095703125, "learning_rate": 9.841937709163489e-05, "loss": 0.0717, "step": 10100 }, { "grad_norm": 0.8286959528923035, "learning_rate": 9.841421721839962e-05, "loss": 0.0628, "step": 10110 }, { "grad_norm": 0.4590461254119873, "learning_rate": 9.840904907249879e-05, "loss": 0.0719, "step": 10120 }, { "grad_norm": 0.6425314545631409, "learning_rate": 9.840387265481545e-05, "loss": 0.0881, "step": 10130 }, { "grad_norm": 0.5733399987220764, "learning_rate": 9.839868796623411e-05, "loss": 0.0677, "step": 10140 }, { "grad_norm": 0.6631444692611694, "learning_rate": 9.839349500764072e-05, "loss": 0.0795, "step": 10150 }, { "grad_norm": 0.6871787309646606, "learning_rate": 9.83882937799226e-05, "loss": 0.0788, "step": 10160 }, { "grad_norm": 0.5633203387260437, "learning_rate": 9.838308428396849e-05, "loss": 0.0662, "step": 10170 }, { "grad_norm": 0.5512059330940247, "learning_rate": 9.837786652066854e-05, "loss": 0.0781, "step": 10180 }, { "grad_norm": 0.5731333494186401, "learning_rate": 9.837264049091437e-05, "loss": 0.0829, "step": 10190 }, { "grad_norm": 0.5814641118049622, "learning_rate": 9.836740619559893e-05, "loss": 0.0751, "step": 10200 }, { "grad_norm": 0.5084793567657471, "learning_rate": 9.836216363561659e-05, "loss": 0.0676, "step": 10210 }, { "grad_norm": 0.45760229229927063, "learning_rate": 9.835691281186322e-05, "loss": 0.0685, "step": 10220 }, { "grad_norm": 0.5332320332527161, "learning_rate": 9.8351653725236e-05, "loss": 0.0658, "step": 10230 }, { "grad_norm": 0.5024942755699158, "learning_rate": 9.83463863766336e-05, "loss": 0.0639, "step": 10240 }, { "grad_norm": 0.6340418457984924, "learning_rate": 9.834111076695602e-05, "loss": 0.0692, "step": 10250 }, { "grad_norm": 0.5471330881118774, "learning_rate": 9.833582689710477e-05, "loss": 0.0626, "step": 10260 }, { "grad_norm": 0.7091149687767029, "learning_rate": 9.833053476798268e-05, "loss": 0.07, "step": 10270 }, { "grad_norm": 0.5696122050285339, "learning_rate": 9.832523438049404e-05, "loss": 0.0652, "step": 10280 }, { "grad_norm": 0.5106252431869507, "learning_rate": 9.831992573554454e-05, "loss": 0.0687, "step": 10290 }, { "grad_norm": 0.5924279689788818, "learning_rate": 9.831460883404128e-05, "loss": 0.0617, "step": 10300 }, { "grad_norm": 0.41401541233062744, "learning_rate": 9.830928367689278e-05, "loss": 0.0558, "step": 10310 }, { "grad_norm": 0.45855259895324707, "learning_rate": 9.830395026500896e-05, "loss": 0.0653, "step": 10320 }, { "grad_norm": 0.4788433313369751, "learning_rate": 9.829860859930115e-05, "loss": 0.0675, "step": 10330 }, { "grad_norm": 0.5101725459098816, "learning_rate": 9.829325868068212e-05, "loss": 0.0609, "step": 10340 }, { "grad_norm": 0.4491460621356964, "learning_rate": 9.8287900510066e-05, "loss": 0.0703, "step": 10350 }, { "grad_norm": 0.4552091360092163, "learning_rate": 9.828253408836834e-05, "loss": 0.0642, "step": 10360 }, { "grad_norm": 0.6952123641967773, "learning_rate": 9.827715941650615e-05, "loss": 0.0652, "step": 10370 }, { "grad_norm": 0.6484888792037964, "learning_rate": 9.82717764953978e-05, "loss": 0.0702, "step": 10380 }, { "grad_norm": 0.456266313791275, "learning_rate": 9.826638532596308e-05, "loss": 0.0656, "step": 10390 }, { "grad_norm": 0.43839240074157715, "learning_rate": 9.82609859091232e-05, "loss": 0.0636, "step": 10400 }, { "grad_norm": 0.43464913964271545, "learning_rate": 9.825557824580076e-05, "loss": 0.0663, "step": 10410 }, { "grad_norm": 0.6248151063919067, "learning_rate": 9.82501623369198e-05, "loss": 0.0638, "step": 10420 }, { "grad_norm": 0.5657467246055603, "learning_rate": 9.824473818340574e-05, "loss": 0.0619, "step": 10430 }, { "grad_norm": 0.4778118431568146, "learning_rate": 9.823930578618541e-05, "loss": 0.0565, "step": 10440 }, { "grad_norm": 0.7408611178398132, "learning_rate": 9.823386514618709e-05, "loss": 0.0648, "step": 10450 }, { "grad_norm": 0.5813812613487244, "learning_rate": 9.82284162643404e-05, "loss": 0.0616, "step": 10460 }, { "grad_norm": 0.3946056365966797, "learning_rate": 9.822295914157642e-05, "loss": 0.0617, "step": 10470 }, { "grad_norm": 0.49409157037734985, "learning_rate": 9.821749377882763e-05, "loss": 0.049, "step": 10480 }, { "grad_norm": 0.4353576600551605, "learning_rate": 9.821202017702791e-05, "loss": 0.0632, "step": 10490 }, { "grad_norm": 0.47799748182296753, "learning_rate": 9.820653833711253e-05, "loss": 0.0578, "step": 10500 }, { "grad_norm": 0.4257725477218628, "learning_rate": 9.820104826001822e-05, "loss": 0.069, "step": 10510 }, { "grad_norm": 0.49935081601142883, "learning_rate": 9.819554994668305e-05, "loss": 0.0515, "step": 10520 }, { "grad_norm": 0.44494929909706116, "learning_rate": 9.819004339804654e-05, "loss": 0.0729, "step": 10530 }, { "grad_norm": 0.5533270239830017, "learning_rate": 9.818452861504961e-05, "loss": 0.0634, "step": 10540 }, { "grad_norm": 0.41759026050567627, "learning_rate": 9.81790055986346e-05, "loss": 0.0568, "step": 10550 }, { "grad_norm": 0.42355477809906006, "learning_rate": 9.817347434974523e-05, "loss": 0.0527, "step": 10560 }, { "grad_norm": 0.5861482620239258, "learning_rate": 9.816793486932664e-05, "loss": 0.0537, "step": 10570 }, { "grad_norm": 0.639019787311554, "learning_rate": 9.816238715832538e-05, "loss": 0.0646, "step": 10580 }, { "grad_norm": 0.4599081575870514, "learning_rate": 9.815683121768939e-05, "loss": 0.0672, "step": 10590 }, { "grad_norm": 0.4819299876689911, "learning_rate": 9.815126704836804e-05, "loss": 0.076, "step": 10600 }, { "grad_norm": 0.3247239887714386, "learning_rate": 9.81456946513121e-05, "loss": 0.0795, "step": 10610 }, { "grad_norm": 0.509430468082428, "learning_rate": 9.814011402747373e-05, "loss": 0.0783, "step": 10620 }, { "grad_norm": 0.3927002549171448, "learning_rate": 9.813452517780651e-05, "loss": 0.0582, "step": 10630 }, { "grad_norm": 0.4772464334964752, "learning_rate": 9.81289281032654e-05, "loss": 0.0701, "step": 10640 }, { "grad_norm": 0.611981987953186, "learning_rate": 9.812332280480683e-05, "loss": 0.071, "step": 10650 }, { "grad_norm": 0.5724931955337524, "learning_rate": 9.811770928338854e-05, "loss": 0.0658, "step": 10660 }, { "grad_norm": 0.712836503982544, "learning_rate": 9.811208753996979e-05, "loss": 0.0637, "step": 10670 }, { "grad_norm": 0.6398259997367859, "learning_rate": 9.810645757551113e-05, "loss": 0.0525, "step": 10680 }, { "grad_norm": 0.5359078645706177, "learning_rate": 9.810081939097459e-05, "loss": 0.066, "step": 10690 }, { "grad_norm": 0.37641918659210205, "learning_rate": 9.809517298732356e-05, "loss": 0.0836, "step": 10700 }, { "grad_norm": 0.38651156425476074, "learning_rate": 9.80895183655229e-05, "loss": 0.0734, "step": 10710 }, { "grad_norm": 0.7096900343894958, "learning_rate": 9.808385552653877e-05, "loss": 0.0637, "step": 10720 }, { "grad_norm": 0.47102436423301697, "learning_rate": 9.807818447133886e-05, "loss": 0.0598, "step": 10730 }, { "grad_norm": 0.5547316670417786, "learning_rate": 9.807250520089215e-05, "loss": 0.0582, "step": 10740 }, { "grad_norm": 0.5037075877189636, "learning_rate": 9.806681771616908e-05, "loss": 0.0644, "step": 10750 }, { "grad_norm": 0.45208173990249634, "learning_rate": 9.80611220181415e-05, "loss": 0.0624, "step": 10760 }, { "grad_norm": 0.7549859881401062, "learning_rate": 9.805541810778264e-05, "loss": 0.0697, "step": 10770 }, { "grad_norm": 0.5108388066291809, "learning_rate": 9.804970598606716e-05, "loss": 0.0675, "step": 10780 }, { "grad_norm": 0.752048134803772, "learning_rate": 9.804398565397106e-05, "loss": 0.0798, "step": 10790 }, { "grad_norm": 0.5787680149078369, "learning_rate": 9.803825711247183e-05, "loss": 0.0679, "step": 10800 }, { "grad_norm": 0.6097978949546814, "learning_rate": 9.803252036254831e-05, "loss": 0.0658, "step": 10810 }, { "grad_norm": 0.5212689638137817, "learning_rate": 9.802677540518076e-05, "loss": 0.0679, "step": 10820 }, { "grad_norm": 0.3707226514816284, "learning_rate": 9.802102224135081e-05, "loss": 0.0656, "step": 10830 }, { "grad_norm": 0.5314022302627563, "learning_rate": 9.801526087204155e-05, "loss": 0.0708, "step": 10840 }, { "grad_norm": 0.6238178610801697, "learning_rate": 9.800949129823743e-05, "loss": 0.0687, "step": 10850 }, { "grad_norm": 0.46081098914146423, "learning_rate": 9.80037135209243e-05, "loss": 0.0521, "step": 10860 }, { "grad_norm": 0.7541595101356506, "learning_rate": 9.799792754108946e-05, "loss": 0.0616, "step": 10870 }, { "grad_norm": 0.3772663474082947, "learning_rate": 9.799213335972152e-05, "loss": 0.0667, "step": 10880 }, { "grad_norm": 0.41477084159851074, "learning_rate": 9.798633097781058e-05, "loss": 0.0606, "step": 10890 }, { "grad_norm": 0.48308050632476807, "learning_rate": 9.79805203963481e-05, "loss": 0.0645, "step": 10900 }, { "grad_norm": 0.46908026933670044, "learning_rate": 9.797470161632697e-05, "loss": 0.0539, "step": 10910 }, { "grad_norm": 0.6056347489356995, "learning_rate": 9.796887463874145e-05, "loss": 0.0624, "step": 10920 }, { "grad_norm": 0.612390398979187, "learning_rate": 9.796303946458718e-05, "loss": 0.0785, "step": 10930 }, { "grad_norm": 0.5713821053504944, "learning_rate": 9.795719609486127e-05, "loss": 0.0716, "step": 10940 }, { "grad_norm": 0.5697445273399353, "learning_rate": 9.795134453056219e-05, "loss": 0.0649, "step": 10950 }, { "grad_norm": 0.40633392333984375, "learning_rate": 9.794548477268979e-05, "loss": 0.0529, "step": 10960 }, { "grad_norm": 0.4575723111629486, "learning_rate": 9.793961682224537e-05, "loss": 0.0659, "step": 10970 }, { "grad_norm": 0.42263108491897583, "learning_rate": 9.793374068023156e-05, "loss": 0.0699, "step": 10980 }, { "grad_norm": 0.4736119210720062, "learning_rate": 9.792785634765247e-05, "loss": 0.0641, "step": 10990 }, { "grad_norm": 0.5232572555541992, "learning_rate": 9.792196382551357e-05, "loss": 0.0677, "step": 11000 }, { "grad_norm": 0.6041659712791443, "learning_rate": 9.791606311482171e-05, "loss": 0.0617, "step": 11010 }, { "grad_norm": 0.46415337920188904, "learning_rate": 9.791015421658518e-05, "loss": 0.0588, "step": 11020 }, { "grad_norm": 0.5176757574081421, "learning_rate": 9.790423713181362e-05, "loss": 0.0655, "step": 11030 }, { "grad_norm": 0.6022626161575317, "learning_rate": 9.789831186151814e-05, "loss": 0.0595, "step": 11040 }, { "grad_norm": 0.7424196600914001, "learning_rate": 9.789237840671118e-05, "loss": 0.0636, "step": 11050 }, { "grad_norm": 0.5165503621101379, "learning_rate": 9.78864367684066e-05, "loss": 0.0679, "step": 11060 }, { "grad_norm": 0.5003151893615723, "learning_rate": 9.788048694761968e-05, "loss": 0.0631, "step": 11070 }, { "grad_norm": 0.7319124341011047, "learning_rate": 9.787452894536709e-05, "loss": 0.062, "step": 11080 }, { "grad_norm": 0.6112250089645386, "learning_rate": 9.786856276266685e-05, "loss": 0.0597, "step": 11090 }, { "grad_norm": 0.37059444189071655, "learning_rate": 9.786258840053845e-05, "loss": 0.0706, "step": 11100 }, { "grad_norm": 0.551810085773468, "learning_rate": 9.785660586000273e-05, "loss": 0.0641, "step": 11110 }, { "grad_norm": 0.4697067439556122, "learning_rate": 9.785061514208196e-05, "loss": 0.0585, "step": 11120 }, { "grad_norm": 0.5155469179153442, "learning_rate": 9.784461624779977e-05, "loss": 0.0693, "step": 11130 }, { "grad_norm": 0.4236745536327362, "learning_rate": 9.783860917818123e-05, "loss": 0.067, "step": 11140 }, { "grad_norm": 0.5112512707710266, "learning_rate": 9.783259393425277e-05, "loss": 0.0498, "step": 11150 }, { "grad_norm": 0.6848070025444031, "learning_rate": 9.782657051704221e-05, "loss": 0.0603, "step": 11160 }, { "grad_norm": 0.5311362743377686, "learning_rate": 9.782053892757883e-05, "loss": 0.0621, "step": 11170 }, { "grad_norm": 0.5234346389770508, "learning_rate": 9.781449916689324e-05, "loss": 0.0569, "step": 11180 }, { "grad_norm": 0.4175735414028168, "learning_rate": 9.780845123601746e-05, "loss": 0.0617, "step": 11190 }, { "grad_norm": 0.7847967147827148, "learning_rate": 9.780239513598492e-05, "loss": 0.0677, "step": 11200 }, { "grad_norm": 0.6365655660629272, "learning_rate": 9.779633086783047e-05, "loss": 0.0676, "step": 11210 }, { "grad_norm": 0.5657297372817993, "learning_rate": 9.779025843259031e-05, "loss": 0.0599, "step": 11220 }, { "grad_norm": 0.5235732793807983, "learning_rate": 9.778417783130204e-05, "loss": 0.0741, "step": 11230 }, { "grad_norm": 0.6142932772636414, "learning_rate": 9.777808906500468e-05, "loss": 0.0655, "step": 11240 }, { "grad_norm": 0.5208907127380371, "learning_rate": 9.777199213473862e-05, "loss": 0.051, "step": 11250 }, { "grad_norm": 0.47076234221458435, "learning_rate": 9.77658870415457e-05, "loss": 0.0506, "step": 11260 }, { "grad_norm": 0.43011555075645447, "learning_rate": 9.775977378646906e-05, "loss": 0.0581, "step": 11270 }, { "grad_norm": 0.6224409341812134, "learning_rate": 9.775365237055331e-05, "loss": 0.0809, "step": 11280 }, { "grad_norm": 0.5754622220993042, "learning_rate": 9.774752279484445e-05, "loss": 0.0742, "step": 11290 }, { "grad_norm": 0.5638670921325684, "learning_rate": 9.774138506038984e-05, "loss": 0.0586, "step": 11300 }, { "grad_norm": 0.43084782361984253, "learning_rate": 9.773523916823826e-05, "loss": 0.0619, "step": 11310 }, { "grad_norm": 0.3746253550052643, "learning_rate": 9.772908511943986e-05, "loss": 0.0586, "step": 11320 }, { "grad_norm": 0.3648998439311981, "learning_rate": 9.77229229150462e-05, "loss": 0.0543, "step": 11330 }, { "grad_norm": 0.5629281997680664, "learning_rate": 9.771675255611024e-05, "loss": 0.0577, "step": 11340 }, { "grad_norm": 0.4178617000579834, "learning_rate": 9.771057404368632e-05, "loss": 0.0795, "step": 11350 }, { "grad_norm": 0.3922828137874603, "learning_rate": 9.770438737883018e-05, "loss": 0.0582, "step": 11360 }, { "grad_norm": 0.4178881347179413, "learning_rate": 9.769819256259898e-05, "loss": 0.0478, "step": 11370 }, { "grad_norm": 0.4267699420452118, "learning_rate": 9.769198959605119e-05, "loss": 0.0593, "step": 11380 }, { "grad_norm": 0.3909999430179596, "learning_rate": 9.768577848024678e-05, "loss": 0.0582, "step": 11390 }, { "grad_norm": 0.6157898306846619, "learning_rate": 9.767955921624702e-05, "loss": 0.0687, "step": 11400 }, { "grad_norm": 0.38863953948020935, "learning_rate": 9.767333180511465e-05, "loss": 0.0528, "step": 11410 }, { "grad_norm": 0.38047730922698975, "learning_rate": 9.766709624791373e-05, "loss": 0.067, "step": 11420 }, { "grad_norm": 0.5071837306022644, "learning_rate": 9.766085254570975e-05, "loss": 0.0604, "step": 11430 }, { "grad_norm": 0.45292750000953674, "learning_rate": 9.76546006995696e-05, "loss": 0.0528, "step": 11440 }, { "grad_norm": 0.6294566988945007, "learning_rate": 9.764834071056155e-05, "loss": 0.0563, "step": 11450 }, { "grad_norm": 0.5810850858688354, "learning_rate": 9.764207257975526e-05, "loss": 0.0756, "step": 11460 }, { "grad_norm": 0.4917775094509125, "learning_rate": 9.763579630822179e-05, "loss": 0.0583, "step": 11470 }, { "grad_norm": 0.624000608921051, "learning_rate": 9.762951189703356e-05, "loss": 0.0597, "step": 11480 }, { "grad_norm": 0.5488342642784119, "learning_rate": 9.762321934726442e-05, "loss": 0.0721, "step": 11490 }, { "grad_norm": 0.4656793177127838, "learning_rate": 9.761691865998959e-05, "loss": 0.067, "step": 11500 }, { "grad_norm": 0.5434004068374634, "learning_rate": 9.76106098362857e-05, "loss": 0.0697, "step": 11510 }, { "grad_norm": 0.44162699580192566, "learning_rate": 9.760429287723072e-05, "loss": 0.0657, "step": 11520 }, { "grad_norm": 0.5542402863502502, "learning_rate": 9.759796778390406e-05, "loss": 0.0615, "step": 11530 }, { "grad_norm": 0.40374311804771423, "learning_rate": 9.759163455738653e-05, "loss": 0.0592, "step": 11540 }, { "grad_norm": 0.7553908824920654, "learning_rate": 9.75852931987603e-05, "loss": 0.0679, "step": 11550 }, { "grad_norm": 0.5080470442771912, "learning_rate": 9.757894370910891e-05, "loss": 0.0531, "step": 11560 }, { "grad_norm": 0.526825487613678, "learning_rate": 9.757258608951733e-05, "loss": 0.0566, "step": 11570 }, { "grad_norm": 0.5938364863395691, "learning_rate": 9.75662203410719e-05, "loss": 0.0665, "step": 11580 }, { "grad_norm": 0.4620223343372345, "learning_rate": 9.755984646486034e-05, "loss": 0.0619, "step": 11590 }, { "grad_norm": 0.456038236618042, "learning_rate": 9.75534644619718e-05, "loss": 0.0552, "step": 11600 }, { "grad_norm": 0.37138432264328003, "learning_rate": 9.754707433349676e-05, "loss": 0.0798, "step": 11610 }, { "grad_norm": 0.6290766000747681, "learning_rate": 9.754067608052715e-05, "loss": 0.0797, "step": 11620 }, { "grad_norm": 0.613487184047699, "learning_rate": 9.753426970415622e-05, "loss": 0.0854, "step": 11630 }, { "grad_norm": 0.5281153917312622, "learning_rate": 9.752785520547868e-05, "loss": 0.0561, "step": 11640 }, { "grad_norm": 0.4812183678150177, "learning_rate": 9.752143258559056e-05, "loss": 0.0591, "step": 11650 }, { "grad_norm": 0.37270864844322205, "learning_rate": 9.751500184558933e-05, "loss": 0.0558, "step": 11660 }, { "grad_norm": 0.3536663055419922, "learning_rate": 9.750856298657383e-05, "loss": 0.0469, "step": 11670 }, { "grad_norm": 0.4240564703941345, "learning_rate": 9.750211600964428e-05, "loss": 0.0639, "step": 11680 }, { "grad_norm": 0.48459652066230774, "learning_rate": 9.749566091590226e-05, "loss": 0.0547, "step": 11690 }, { "grad_norm": 0.4205496311187744, "learning_rate": 9.748919770645083e-05, "loss": 0.0596, "step": 11700 }, { "grad_norm": 0.4372355043888092, "learning_rate": 9.748272638239432e-05, "loss": 0.0589, "step": 11710 }, { "grad_norm": 0.6977128982543945, "learning_rate": 9.747624694483855e-05, "loss": 0.056, "step": 11720 }, { "grad_norm": 0.3020981550216675, "learning_rate": 9.746975939489065e-05, "loss": 0.0504, "step": 11730 }, { "grad_norm": 0.5295040011405945, "learning_rate": 9.746326373365918e-05, "loss": 0.0589, "step": 11740 }, { "grad_norm": 0.3744295537471771, "learning_rate": 9.745675996225403e-05, "loss": 0.0619, "step": 11750 }, { "grad_norm": 0.34088510274887085, "learning_rate": 9.745024808178657e-05, "loss": 0.0568, "step": 11760 }, { "grad_norm": 0.6191713213920593, "learning_rate": 9.744372809336947e-05, "loss": 0.0547, "step": 11770 }, { "grad_norm": 0.5692792534828186, "learning_rate": 9.743719999811682e-05, "loss": 0.0551, "step": 11780 }, { "grad_norm": 0.44307947158813477, "learning_rate": 9.743066379714412e-05, "loss": 0.0526, "step": 11790 }, { "grad_norm": 0.43269023299217224, "learning_rate": 9.74241194915682e-05, "loss": 0.0532, "step": 11800 }, { "grad_norm": 0.4063360393047333, "learning_rate": 9.741756708250731e-05, "loss": 0.0631, "step": 11810 }, { "grad_norm": 0.5074233412742615, "learning_rate": 9.741100657108109e-05, "loss": 0.0587, "step": 11820 }, { "grad_norm": 0.4955817461013794, "learning_rate": 9.740443795841054e-05, "loss": 0.0703, "step": 11830 }, { "grad_norm": 0.5872031450271606, "learning_rate": 9.739786124561805e-05, "loss": 0.0499, "step": 11840 }, { "grad_norm": 0.39639800786972046, "learning_rate": 9.73912764338274e-05, "loss": 0.0614, "step": 11850 }, { "grad_norm": 0.5579843521118164, "learning_rate": 9.738468352416377e-05, "loss": 0.0599, "step": 11860 }, { "grad_norm": 0.4942742586135864, "learning_rate": 9.737808251775369e-05, "loss": 0.0548, "step": 11870 }, { "grad_norm": 0.5542868375778198, "learning_rate": 9.737147341572512e-05, "loss": 0.0708, "step": 11880 }, { "grad_norm": 0.4737536907196045, "learning_rate": 9.736485621920735e-05, "loss": 0.0611, "step": 11890 }, { "grad_norm": 0.5329036116600037, "learning_rate": 9.735823092933108e-05, "loss": 0.0622, "step": 11900 }, { "grad_norm": 0.4247475564479828, "learning_rate": 9.735159754722838e-05, "loss": 0.0657, "step": 11910 }, { "grad_norm": 0.5441759824752808, "learning_rate": 9.734495607403275e-05, "loss": 0.056, "step": 11920 }, { "grad_norm": 0.5175656676292419, "learning_rate": 9.733830651087901e-05, "loss": 0.0554, "step": 11930 }, { "grad_norm": 0.460461288690567, "learning_rate": 9.733164885890338e-05, "loss": 0.0503, "step": 11940 }, { "grad_norm": 0.371044784784317, "learning_rate": 9.732498311924349e-05, "loss": 0.0397, "step": 11950 }, { "grad_norm": 0.49817603826522827, "learning_rate": 9.731830929303833e-05, "loss": 0.0514, "step": 11960 }, { "grad_norm": 0.47250136733055115, "learning_rate": 9.731162738142827e-05, "loss": 0.0644, "step": 11970 }, { "grad_norm": 0.6543782353401184, "learning_rate": 9.730493738555506e-05, "loss": 0.0615, "step": 11980 }, { "grad_norm": 0.3876018822193146, "learning_rate": 9.729823930656186e-05, "loss": 0.0631, "step": 11990 }, { "grad_norm": 0.5418081879615784, "learning_rate": 9.729153314559316e-05, "loss": 0.0732, "step": 12000 }, { "grad_norm": 0.43622228503227234, "learning_rate": 9.728481890379486e-05, "loss": 0.0607, "step": 12010 }, { "grad_norm": 0.3299902081489563, "learning_rate": 9.727809658231428e-05, "loss": 0.0604, "step": 12020 }, { "grad_norm": 0.458077073097229, "learning_rate": 9.727136618230003e-05, "loss": 0.0536, "step": 12030 }, { "grad_norm": 0.4471132159233093, "learning_rate": 9.726462770490219e-05, "loss": 0.0648, "step": 12040 }, { "grad_norm": 0.44974154233932495, "learning_rate": 9.725788115127214e-05, "loss": 0.0654, "step": 12050 }, { "grad_norm": 0.6576809287071228, "learning_rate": 9.725112652256274e-05, "loss": 0.0662, "step": 12060 }, { "grad_norm": 0.42775392532348633, "learning_rate": 9.724436381992812e-05, "loss": 0.0823, "step": 12070 }, { "grad_norm": 0.5550362467765808, "learning_rate": 9.723759304452387e-05, "loss": 0.0557, "step": 12080 }, { "grad_norm": 0.7255702018737793, "learning_rate": 9.72308141975069e-05, "loss": 0.0571, "step": 12090 }, { "grad_norm": 0.5541763305664062, "learning_rate": 9.722402728003557e-05, "loss": 0.0674, "step": 12100 }, { "grad_norm": 0.5492673516273499, "learning_rate": 9.721723229326953e-05, "loss": 0.0622, "step": 12110 }, { "grad_norm": 0.5307424068450928, "learning_rate": 9.721042923836992e-05, "loss": 0.0469, "step": 12120 }, { "grad_norm": 0.45154649019241333, "learning_rate": 9.720361811649914e-05, "loss": 0.0601, "step": 12130 }, { "grad_norm": 0.6760947108268738, "learning_rate": 9.719679892882106e-05, "loss": 0.0631, "step": 12140 }, { "grad_norm": 0.49319353699684143, "learning_rate": 9.718997167650085e-05, "loss": 0.0656, "step": 12150 }, { "grad_norm": 0.5592806339263916, "learning_rate": 9.718313636070515e-05, "loss": 0.0512, "step": 12160 }, { "grad_norm": 0.5217500925064087, "learning_rate": 9.717629298260192e-05, "loss": 0.0584, "step": 12170 }, { "grad_norm": 0.6478539109230042, "learning_rate": 9.716944154336047e-05, "loss": 0.0713, "step": 12180 }, { "grad_norm": 0.44025012850761414, "learning_rate": 9.716258204415157e-05, "loss": 0.0554, "step": 12190 }, { "grad_norm": 0.3630196750164032, "learning_rate": 9.715571448614728e-05, "loss": 0.0649, "step": 12200 }, { "grad_norm": 0.3255175054073334, "learning_rate": 9.71488388705211e-05, "loss": 0.0867, "step": 12210 }, { "grad_norm": 0.46375226974487305, "learning_rate": 9.714195519844788e-05, "loss": 0.0558, "step": 12220 }, { "grad_norm": 0.5543634295463562, "learning_rate": 9.713506347110386e-05, "loss": 0.0651, "step": 12230 }, { "grad_norm": 0.31050020456314087, "learning_rate": 9.712816368966662e-05, "loss": 0.0505, "step": 12240 }, { "grad_norm": 0.392966091632843, "learning_rate": 9.712125585531517e-05, "loss": 0.0692, "step": 12250 }, { "grad_norm": 0.48491063714027405, "learning_rate": 9.711433996922988e-05, "loss": 0.0745, "step": 12260 }, { "grad_norm": 0.4704011380672455, "learning_rate": 9.710741603259245e-05, "loss": 0.0602, "step": 12270 }, { "grad_norm": 0.37096163630485535, "learning_rate": 9.710048404658603e-05, "loss": 0.0514, "step": 12280 }, { "grad_norm": 0.46299299597740173, "learning_rate": 9.709354401239508e-05, "loss": 0.0679, "step": 12290 }, { "grad_norm": 0.3902064859867096, "learning_rate": 9.708659593120546e-05, "loss": 0.0498, "step": 12300 }, { "grad_norm": 0.570878267288208, "learning_rate": 9.707963980420443e-05, "loss": 0.0623, "step": 12310 }, { "grad_norm": 0.6994086503982544, "learning_rate": 9.707267563258058e-05, "loss": 0.0659, "step": 12320 }, { "grad_norm": 0.4854696989059448, "learning_rate": 9.70657034175239e-05, "loss": 0.0634, "step": 12330 }, { "grad_norm": 0.3735952377319336, "learning_rate": 9.705872316022577e-05, "loss": 0.0691, "step": 12340 }, { "grad_norm": 0.5831769108772278, "learning_rate": 9.705173486187891e-05, "loss": 0.0648, "step": 12350 }, { "grad_norm": 0.5882343649864197, "learning_rate": 9.704473852367741e-05, "loss": 0.0699, "step": 12360 }, { "grad_norm": 0.5652139782905579, "learning_rate": 9.70377341468168e-05, "loss": 0.0553, "step": 12370 }, { "grad_norm": 0.396382600069046, "learning_rate": 9.703072173249389e-05, "loss": 0.0538, "step": 12380 }, { "grad_norm": 0.5203232765197754, "learning_rate": 9.702370128190693e-05, "loss": 0.0578, "step": 12390 }, { "grad_norm": 0.6023860573768616, "learning_rate": 9.701667279625552e-05, "loss": 0.0723, "step": 12400 }, { "grad_norm": 0.6158231496810913, "learning_rate": 9.700963627674065e-05, "loss": 0.0691, "step": 12410 }, { "grad_norm": 0.42047175765037537, "learning_rate": 9.700259172456466e-05, "loss": 0.0549, "step": 12420 }, { "grad_norm": 0.6913103461265564, "learning_rate": 9.699553914093124e-05, "loss": 0.0731, "step": 12430 }, { "grad_norm": 0.477525532245636, "learning_rate": 9.698847852704553e-05, "loss": 0.0735, "step": 12440 }, { "grad_norm": 0.4234514832496643, "learning_rate": 9.6981409884114e-05, "loss": 0.0524, "step": 12450 }, { "grad_norm": 0.4520117938518524, "learning_rate": 9.697433321334443e-05, "loss": 0.0624, "step": 12460 }, { "grad_norm": 0.6091469526290894, "learning_rate": 9.696724851594607e-05, "loss": 0.0594, "step": 12470 }, { "grad_norm": 0.4559984505176544, "learning_rate": 9.696015579312952e-05, "loss": 0.0544, "step": 12480 }, { "grad_norm": 0.435546338558197, "learning_rate": 9.695305504610668e-05, "loss": 0.0611, "step": 12490 }, { "grad_norm": 0.5215734243392944, "learning_rate": 9.694594627609092e-05, "loss": 0.0605, "step": 12500 }, { "grad_norm": 0.6445625424385071, "learning_rate": 9.693882948429691e-05, "loss": 0.0737, "step": 12510 }, { "grad_norm": 0.40874096751213074, "learning_rate": 9.693170467194071e-05, "loss": 0.0613, "step": 12520 }, { "grad_norm": 0.3979538381099701, "learning_rate": 9.692457184023977e-05, "loss": 0.0609, "step": 12530 }, { "grad_norm": 0.5592889785766602, "learning_rate": 9.691743099041291e-05, "loss": 0.0771, "step": 12540 }, { "grad_norm": 0.6987166404724121, "learning_rate": 9.691028212368027e-05, "loss": 0.069, "step": 12550 }, { "grad_norm": 0.6092842221260071, "learning_rate": 9.690312524126342e-05, "loss": 0.0592, "step": 12560 }, { "grad_norm": 0.38881969451904297, "learning_rate": 9.689596034438527e-05, "loss": 0.0592, "step": 12570 }, { "grad_norm": 0.38199251890182495, "learning_rate": 9.688878743427012e-05, "loss": 0.0583, "step": 12580 }, { "grad_norm": 0.5281520485877991, "learning_rate": 9.688160651214359e-05, "loss": 0.0732, "step": 12590 }, { "grad_norm": 0.6520177125930786, "learning_rate": 9.687441757923273e-05, "loss": 0.0625, "step": 12600 }, { "grad_norm": 0.4363517463207245, "learning_rate": 9.68672206367659e-05, "loss": 0.0588, "step": 12610 }, { "grad_norm": 0.5015581250190735, "learning_rate": 9.686001568597291e-05, "loss": 0.063, "step": 12620 }, { "grad_norm": 0.3825364112854004, "learning_rate": 9.685280272808486e-05, "loss": 0.062, "step": 12630 }, { "grad_norm": 0.4752962589263916, "learning_rate": 9.684558176433424e-05, "loss": 0.0735, "step": 12640 }, { "grad_norm": 0.599179744720459, "learning_rate": 9.683835279595495e-05, "loss": 0.0572, "step": 12650 }, { "grad_norm": 0.5987246036529541, "learning_rate": 9.683111582418216e-05, "loss": 0.0766, "step": 12660 }, { "grad_norm": 0.5026206374168396, "learning_rate": 9.682387085025254e-05, "loss": 0.0572, "step": 12670 }, { "grad_norm": 0.5432601571083069, "learning_rate": 9.681661787540401e-05, "loss": 0.0734, "step": 12680 }, { "grad_norm": 0.5318342447280884, "learning_rate": 9.680935690087593e-05, "loss": 0.0683, "step": 12690 }, { "grad_norm": 0.5372499823570251, "learning_rate": 9.680208792790901e-05, "loss": 0.0627, "step": 12700 }, { "grad_norm": 0.6465361714363098, "learning_rate": 9.679481095774529e-05, "loss": 0.0608, "step": 12710 }, { "grad_norm": 0.5680628418922424, "learning_rate": 9.678752599162822e-05, "loss": 0.0575, "step": 12720 }, { "grad_norm": 0.4520575702190399, "learning_rate": 9.678023303080259e-05, "loss": 0.051, "step": 12730 }, { "grad_norm": 0.4927060604095459, "learning_rate": 9.677293207651459e-05, "loss": 0.0607, "step": 12740 }, { "grad_norm": 0.33354058861732483, "learning_rate": 9.676562313001173e-05, "loss": 0.0447, "step": 12750 }, { "grad_norm": 0.5738013386726379, "learning_rate": 9.675830619254293e-05, "loss": 0.0649, "step": 12760 }, { "grad_norm": 0.5025943517684937, "learning_rate": 9.675098126535843e-05, "loss": 0.0699, "step": 12770 }, { "grad_norm": 0.5431092977523804, "learning_rate": 9.674364834970988e-05, "loss": 0.0717, "step": 12780 }, { "grad_norm": 0.4139796197414398, "learning_rate": 9.673630744685028e-05, "loss": 0.0617, "step": 12790 }, { "grad_norm": 0.39344605803489685, "learning_rate": 9.672895855803397e-05, "loss": 0.0512, "step": 12800 }, { "grad_norm": 0.3709760904312134, "learning_rate": 9.672160168451667e-05, "loss": 0.0519, "step": 12810 }, { "grad_norm": 0.47385406494140625, "learning_rate": 9.671423682755549e-05, "loss": 0.06, "step": 12820 }, { "grad_norm": 0.6502569317817688, "learning_rate": 9.670686398840888e-05, "loss": 0.0651, "step": 12830 }, { "grad_norm": 0.4012832045555115, "learning_rate": 9.669948316833664e-05, "loss": 0.0516, "step": 12840 }, { "grad_norm": 0.3782866597175598, "learning_rate": 9.669209436859997e-05, "loss": 0.0544, "step": 12850 }, { "grad_norm": 0.4118228256702423, "learning_rate": 9.66846975904614e-05, "loss": 0.0602, "step": 12860 }, { "grad_norm": 0.5278518199920654, "learning_rate": 9.667729283518483e-05, "loss": 0.0528, "step": 12870 }, { "grad_norm": 0.49492260813713074, "learning_rate": 9.666988010403557e-05, "loss": 0.0509, "step": 12880 }, { "grad_norm": 0.4736824035644531, "learning_rate": 9.66624593982802e-05, "loss": 0.0546, "step": 12890 }, { "grad_norm": 0.6914287209510803, "learning_rate": 9.665503071918675e-05, "loss": 0.0687, "step": 12900 }, { "grad_norm": 0.4808539152145386, "learning_rate": 9.664759406802456e-05, "loss": 0.0518, "step": 12910 }, { "grad_norm": 0.3406277000904083, "learning_rate": 9.664014944606437e-05, "loss": 0.0427, "step": 12920 }, { "grad_norm": 0.49325811862945557, "learning_rate": 9.663269685457822e-05, "loss": 0.0581, "step": 12930 }, { "grad_norm": 0.59039306640625, "learning_rate": 9.662523629483962e-05, "loss": 0.0595, "step": 12940 }, { "grad_norm": 0.40893319249153137, "learning_rate": 9.661776776812333e-05, "loss": 0.0477, "step": 12950 }, { "grad_norm": 0.5007252097129822, "learning_rate": 9.661029127570553e-05, "loss": 0.0628, "step": 12960 }, { "grad_norm": 0.4636251926422119, "learning_rate": 9.660280681886373e-05, "loss": 0.0522, "step": 12970 }, { "grad_norm": 0.3929549753665924, "learning_rate": 9.659531439887685e-05, "loss": 0.0458, "step": 12980 }, { "grad_norm": 0.6014068126678467, "learning_rate": 9.658781401702511e-05, "loss": 0.0702, "step": 12990 }, { "grad_norm": 0.4872022271156311, "learning_rate": 9.658030567459015e-05, "loss": 0.0618, "step": 13000 }, { "grad_norm": 0.6482728719711304, "learning_rate": 9.65727893728549e-05, "loss": 0.0759, "step": 13010 }, { "grad_norm": 0.5010973811149597, "learning_rate": 9.656526511310375e-05, "loss": 0.072, "step": 13020 }, { "grad_norm": 0.5733979344367981, "learning_rate": 9.655773289662233e-05, "loss": 0.0669, "step": 13030 }, { "grad_norm": 0.5016338229179382, "learning_rate": 9.655019272469772e-05, "loss": 0.0597, "step": 13040 }, { "grad_norm": 0.5386804938316345, "learning_rate": 9.654264459861832e-05, "loss": 0.0475, "step": 13050 }, { "grad_norm": 0.49190303683280945, "learning_rate": 9.653508851967391e-05, "loss": 0.0588, "step": 13060 }, { "grad_norm": 0.536585807800293, "learning_rate": 9.65275244891556e-05, "loss": 0.0666, "step": 13070 }, { "grad_norm": 0.5885763168334961, "learning_rate": 9.651995250835591e-05, "loss": 0.0576, "step": 13080 }, { "grad_norm": 0.45739325881004333, "learning_rate": 9.651237257856862e-05, "loss": 0.0631, "step": 13090 }, { "grad_norm": 1.0664511919021606, "learning_rate": 9.6504784701089e-05, "loss": 0.0689, "step": 13100 }, { "grad_norm": 0.5186640620231628, "learning_rate": 9.649718887721357e-05, "loss": 0.0583, "step": 13110 }, { "grad_norm": 0.5524018406867981, "learning_rate": 9.648958510824028e-05, "loss": 0.0553, "step": 13120 }, { "grad_norm": 0.5533007979393005, "learning_rate": 9.648197339546837e-05, "loss": 0.0685, "step": 13130 }, { "grad_norm": 0.422372430562973, "learning_rate": 9.647435374019851e-05, "loss": 0.0674, "step": 13140 }, { "grad_norm": 0.898463785648346, "learning_rate": 9.646672614373266e-05, "loss": 0.0761, "step": 13150 }, { "grad_norm": 0.7156273722648621, "learning_rate": 9.645909060737418e-05, "loss": 0.0545, "step": 13160 }, { "grad_norm": 0.4020717740058899, "learning_rate": 9.645144713242778e-05, "loss": 0.0627, "step": 13170 }, { "grad_norm": 0.4145359396934509, "learning_rate": 9.64437957201995e-05, "loss": 0.0621, "step": 13180 }, { "grad_norm": 0.46258676052093506, "learning_rate": 9.643613637199678e-05, "loss": 0.0506, "step": 13190 }, { "grad_norm": 0.5313073992729187, "learning_rate": 9.642846908912839e-05, "loss": 0.0713, "step": 13200 }, { "grad_norm": 0.3574063777923584, "learning_rate": 9.642079387290444e-05, "loss": 0.0555, "step": 13210 }, { "grad_norm": 0.5378382802009583, "learning_rate": 9.641311072463644e-05, "loss": 0.0564, "step": 13220 }, { "grad_norm": 0.47971636056900024, "learning_rate": 9.640541964563722e-05, "loss": 0.0604, "step": 13230 }, { "grad_norm": 0.6064529418945312, "learning_rate": 9.639772063722096e-05, "loss": 0.079, "step": 13240 }, { "grad_norm": 0.4725635349750519, "learning_rate": 9.639001370070324e-05, "loss": 0.0771, "step": 13250 }, { "grad_norm": 1.0220065116882324, "learning_rate": 9.638229883740095e-05, "loss": 0.0642, "step": 13260 }, { "grad_norm": 0.4890687167644501, "learning_rate": 9.637457604863233e-05, "loss": 0.0504, "step": 13270 }, { "grad_norm": 0.39473676681518555, "learning_rate": 9.636684533571703e-05, "loss": 0.0612, "step": 13280 }, { "grad_norm": 0.41612759232521057, "learning_rate": 9.635910669997599e-05, "loss": 0.0571, "step": 13290 }, { "grad_norm": 0.39914262294769287, "learning_rate": 9.635136014273154e-05, "loss": 0.0498, "step": 13300 }, { "grad_norm": 0.4691527187824249, "learning_rate": 9.634360566530735e-05, "loss": 0.0617, "step": 13310 }, { "grad_norm": 0.503741979598999, "learning_rate": 9.633584326902845e-05, "loss": 0.0656, "step": 13320 }, { "grad_norm": 0.5121105313301086, "learning_rate": 9.632807295522124e-05, "loss": 0.0594, "step": 13330 }, { "grad_norm": 0.541774570941925, "learning_rate": 9.632029472521342e-05, "loss": 0.0556, "step": 13340 }, { "grad_norm": 0.35952574014663696, "learning_rate": 9.631250858033409e-05, "loss": 0.0464, "step": 13350 }, { "grad_norm": 0.7625659108161926, "learning_rate": 9.630471452191371e-05, "loss": 0.0629, "step": 13360 }, { "grad_norm": 0.3421168327331543, "learning_rate": 9.629691255128405e-05, "loss": 0.0625, "step": 13370 }, { "grad_norm": 0.48923584818840027, "learning_rate": 9.628910266977825e-05, "loss": 0.0568, "step": 13380 }, { "grad_norm": 0.5204852223396301, "learning_rate": 9.628128487873083e-05, "loss": 0.0581, "step": 13390 }, { "grad_norm": 0.4525747001171112, "learning_rate": 9.627345917947761e-05, "loss": 0.0681, "step": 13400 }, { "grad_norm": 0.4484346807003021, "learning_rate": 9.626562557335579e-05, "loss": 0.0567, "step": 13410 }, { "grad_norm": 0.4433538615703583, "learning_rate": 9.625778406170393e-05, "loss": 0.055, "step": 13420 }, { "grad_norm": 0.6699813604354858, "learning_rate": 9.624993464586193e-05, "loss": 0.0615, "step": 13430 }, { "grad_norm": 0.3968525528907776, "learning_rate": 9.624207732717105e-05, "loss": 0.0578, "step": 13440 }, { "grad_norm": 0.5279749035835266, "learning_rate": 9.623421210697386e-05, "loss": 0.0692, "step": 13450 }, { "grad_norm": 0.5403776168823242, "learning_rate": 9.622633898661434e-05, "loss": 0.0737, "step": 13460 }, { "grad_norm": 0.6918983459472656, "learning_rate": 9.621845796743778e-05, "loss": 0.0717, "step": 13470 }, { "grad_norm": 0.47163569927215576, "learning_rate": 9.621056905079082e-05, "loss": 0.0714, "step": 13480 }, { "grad_norm": 0.49229007959365845, "learning_rate": 9.620267223802149e-05, "loss": 0.0598, "step": 13490 }, { "grad_norm": 0.414153128862381, "learning_rate": 9.619476753047911e-05, "loss": 0.0622, "step": 13500 }, { "grad_norm": 0.5330760478973389, "learning_rate": 9.618685492951438e-05, "loss": 0.0659, "step": 13510 }, { "grad_norm": 0.6453769207000732, "learning_rate": 9.617893443647938e-05, "loss": 0.0779, "step": 13520 }, { "grad_norm": 0.5528104901313782, "learning_rate": 9.617100605272746e-05, "loss": 0.0727, "step": 13530 }, { "grad_norm": 0.4125394821166992, "learning_rate": 9.616306977961338e-05, "loss": 0.0572, "step": 13540 }, { "grad_norm": 0.504672646522522, "learning_rate": 9.615512561849326e-05, "loss": 0.0497, "step": 13550 }, { "grad_norm": 0.4895838499069214, "learning_rate": 9.61471735707245e-05, "loss": 0.0637, "step": 13560 }, { "grad_norm": 0.5066535472869873, "learning_rate": 9.613921363766592e-05, "loss": 0.0519, "step": 13570 }, { "grad_norm": 0.5198884606361389, "learning_rate": 9.613124582067763e-05, "loss": 0.0673, "step": 13580 }, { "grad_norm": 0.752364456653595, "learning_rate": 9.612327012112112e-05, "loss": 0.0767, "step": 13590 }, { "grad_norm": 0.7538574934005737, "learning_rate": 9.611528654035921e-05, "loss": 0.086, "step": 13600 }, { "grad_norm": 0.5317166447639465, "learning_rate": 9.610729507975611e-05, "loss": 0.0616, "step": 13610 }, { "grad_norm": 0.4679943323135376, "learning_rate": 9.609929574067731e-05, "loss": 0.0705, "step": 13620 }, { "grad_norm": 0.44071561098098755, "learning_rate": 9.609128852448967e-05, "loss": 0.0489, "step": 13630 }, { "grad_norm": 0.48586443066596985, "learning_rate": 9.608327343256143e-05, "loss": 0.0538, "step": 13640 }, { "grad_norm": 0.5805922746658325, "learning_rate": 9.607525046626216e-05, "loss": 0.0531, "step": 13650 }, { "grad_norm": 0.6554820537567139, "learning_rate": 9.606721962696272e-05, "loss": 0.0651, "step": 13660 }, { "grad_norm": 0.46534785628318787, "learning_rate": 9.60591809160354e-05, "loss": 0.0624, "step": 13670 }, { "grad_norm": 0.43495479226112366, "learning_rate": 9.605113433485378e-05, "loss": 0.0651, "step": 13680 }, { "grad_norm": 0.5713026523590088, "learning_rate": 9.604307988479279e-05, "loss": 0.0561, "step": 13690 }, { "grad_norm": 0.439228355884552, "learning_rate": 9.603501756722876e-05, "loss": 0.0552, "step": 13700 }, { "grad_norm": 0.5127286314964294, "learning_rate": 9.602694738353927e-05, "loss": 0.0611, "step": 13710 }, { "grad_norm": 0.5253855586051941, "learning_rate": 9.601886933510331e-05, "loss": 0.0564, "step": 13720 }, { "grad_norm": 0.617775559425354, "learning_rate": 9.60107834233012e-05, "loss": 0.0567, "step": 13730 }, { "grad_norm": 0.35462382435798645, "learning_rate": 9.60026896495146e-05, "loss": 0.0515, "step": 13740 }, { "grad_norm": 0.48223942518234253, "learning_rate": 9.599458801512652e-05, "loss": 0.0524, "step": 13750 }, { "grad_norm": 0.644154965877533, "learning_rate": 9.598647852152129e-05, "loss": 0.0725, "step": 13760 }, { "grad_norm": 0.47443336248397827, "learning_rate": 9.597836117008462e-05, "loss": 0.0512, "step": 13770 }, { "grad_norm": 0.5640837550163269, "learning_rate": 9.597023596220356e-05, "loss": 0.0501, "step": 13780 }, { "grad_norm": 0.47128409147262573, "learning_rate": 9.596210289926643e-05, "loss": 0.0499, "step": 13790 }, { "grad_norm": 0.41940784454345703, "learning_rate": 9.5953961982663e-05, "loss": 0.0509, "step": 13800 }, { "grad_norm": 0.443432480096817, "learning_rate": 9.594581321378431e-05, "loss": 0.0611, "step": 13810 }, { "grad_norm": 0.4498577117919922, "learning_rate": 9.593765659402276e-05, "loss": 0.0595, "step": 13820 }, { "grad_norm": 0.6021342873573303, "learning_rate": 9.59294921247721e-05, "loss": 0.0584, "step": 13830 }, { "grad_norm": 0.478153258562088, "learning_rate": 9.59213198074274e-05, "loss": 0.0609, "step": 13840 }, { "grad_norm": 0.5000378489494324, "learning_rate": 9.59131396433851e-05, "loss": 0.0478, "step": 13850 }, { "grad_norm": 0.41754505038261414, "learning_rate": 9.590495163404297e-05, "loss": 0.0569, "step": 13860 }, { "grad_norm": 0.551659882068634, "learning_rate": 9.589675578080009e-05, "loss": 0.0584, "step": 13870 }, { "grad_norm": 0.4487088620662689, "learning_rate": 9.588855208505694e-05, "loss": 0.0434, "step": 13880 }, { "grad_norm": 0.5856648087501526, "learning_rate": 9.588034054821529e-05, "loss": 0.0663, "step": 13890 }, { "grad_norm": 0.4702214002609253, "learning_rate": 9.587212117167826e-05, "loss": 0.0511, "step": 13900 }, { "grad_norm": 0.40920186042785645, "learning_rate": 9.586389395685033e-05, "loss": 0.0535, "step": 13910 }, { "grad_norm": 0.5013796091079712, "learning_rate": 9.585565890513733e-05, "loss": 0.0661, "step": 13920 }, { "grad_norm": 0.4895075261592865, "learning_rate": 9.584741601794636e-05, "loss": 0.057, "step": 13930 }, { "grad_norm": 0.549379289150238, "learning_rate": 9.58391652966859e-05, "loss": 0.0621, "step": 13940 }, { "grad_norm": 0.7101083397865295, "learning_rate": 9.583090674276583e-05, "loss": 0.0587, "step": 13950 }, { "grad_norm": 0.4798755943775177, "learning_rate": 9.582264035759726e-05, "loss": 0.0506, "step": 13960 }, { "grad_norm": 0.44113996624946594, "learning_rate": 9.58143661425927e-05, "loss": 0.0841, "step": 13970 }, { "grad_norm": 0.574261486530304, "learning_rate": 9.580608409916601e-05, "loss": 0.0657, "step": 13980 }, { "grad_norm": 0.32795029878616333, "learning_rate": 9.579779422873233e-05, "loss": 0.0669, "step": 13990 }, { "grad_norm": 0.4367734491825104, "learning_rate": 9.578949653270819e-05, "loss": 0.06, "step": 14000 }, { "grad_norm": 0.5028672814369202, "learning_rate": 9.578119101251144e-05, "loss": 0.0517, "step": 14010 }, { "grad_norm": 0.394030898809433, "learning_rate": 9.577287766956127e-05, "loss": 0.0696, "step": 14020 }, { "grad_norm": 0.37235626578330994, "learning_rate": 9.57645565052782e-05, "loss": 0.0632, "step": 14030 }, { "grad_norm": 0.5014033913612366, "learning_rate": 9.575622752108407e-05, "loss": 0.061, "step": 14040 }, { "grad_norm": 0.5652327537536621, "learning_rate": 9.57478907184021e-05, "loss": 0.0554, "step": 14050 }, { "grad_norm": 0.4625941812992096, "learning_rate": 9.573954609865681e-05, "loss": 0.0605, "step": 14060 }, { "grad_norm": 0.6254191398620605, "learning_rate": 9.573119366327408e-05, "loss": 0.0551, "step": 14070 }, { "grad_norm": 0.4674900770187378, "learning_rate": 9.57228334136811e-05, "loss": 0.0528, "step": 14080 }, { "grad_norm": 0.5840640068054199, "learning_rate": 9.571446535130641e-05, "loss": 0.0602, "step": 14090 }, { "grad_norm": 0.3758598566055298, "learning_rate": 9.570608947757988e-05, "loss": 0.061, "step": 14100 }, { "grad_norm": 0.4459385275840759, "learning_rate": 9.569770579393274e-05, "loss": 0.0527, "step": 14110 }, { "grad_norm": 0.4939264953136444, "learning_rate": 9.56893143017975e-05, "loss": 0.0543, "step": 14120 }, { "grad_norm": 0.5227604508399963, "learning_rate": 9.568091500260806e-05, "loss": 0.0623, "step": 14130 }, { "grad_norm": 0.4371969699859619, "learning_rate": 9.567250789779961e-05, "loss": 0.0628, "step": 14140 }, { "grad_norm": 0.5719658136367798, "learning_rate": 9.566409298880872e-05, "loss": 0.0471, "step": 14150 }, { "grad_norm": 0.5261656641960144, "learning_rate": 9.565567027707326e-05, "loss": 0.0734, "step": 14160 }, { "grad_norm": 0.5250770449638367, "learning_rate": 9.56472397640324e-05, "loss": 0.0551, "step": 14170 }, { "grad_norm": 0.6316156387329102, "learning_rate": 9.563880145112675e-05, "loss": 0.0583, "step": 14180 }, { "grad_norm": 0.6214054226875305, "learning_rate": 9.563035533979814e-05, "loss": 0.0625, "step": 14190 }, { "grad_norm": 0.4356718063354492, "learning_rate": 9.562190143148981e-05, "loss": 0.0539, "step": 14200 }, { "grad_norm": 0.45476168394088745, "learning_rate": 9.561343972764627e-05, "loss": 0.0515, "step": 14210 }, { "grad_norm": 0.45684167742729187, "learning_rate": 9.560497022971343e-05, "loss": 0.0619, "step": 14220 }, { "grad_norm": 0.45215028524398804, "learning_rate": 9.559649293913847e-05, "loss": 0.056, "step": 14230 }, { "grad_norm": 0.5025825500488281, "learning_rate": 9.558800785736993e-05, "loss": 0.0504, "step": 14240 }, { "grad_norm": 0.35959112644195557, "learning_rate": 9.557951498585767e-05, "loss": 0.0625, "step": 14250 }, { "grad_norm": 0.3826776444911957, "learning_rate": 9.557101432605293e-05, "loss": 0.0525, "step": 14260 }, { "grad_norm": 0.49708420038223267, "learning_rate": 9.556250587940818e-05, "loss": 0.0591, "step": 14270 }, { "grad_norm": 0.45622673630714417, "learning_rate": 9.555398964737734e-05, "loss": 0.0491, "step": 14280 }, { "grad_norm": 0.631655216217041, "learning_rate": 9.554546563141555e-05, "loss": 0.0566, "step": 14290 }, { "grad_norm": 0.6161102652549744, "learning_rate": 9.553693383297937e-05, "loss": 0.0622, "step": 14300 }, { "grad_norm": 0.43919137120246887, "learning_rate": 9.552839425352663e-05, "loss": 0.0544, "step": 14310 }, { "grad_norm": 0.46581918001174927, "learning_rate": 9.551984689451652e-05, "loss": 0.0679, "step": 14320 }, { "grad_norm": 0.5201224684715271, "learning_rate": 9.551129175740953e-05, "loss": 0.0651, "step": 14330 }, { "grad_norm": 0.5060061812400818, "learning_rate": 9.550272884366754e-05, "loss": 0.0618, "step": 14340 }, { "grad_norm": 0.4970674514770508, "learning_rate": 9.549415815475369e-05, "loss": 0.0541, "step": 14350 }, { "grad_norm": 0.40464210510253906, "learning_rate": 9.548557969213247e-05, "loss": 0.0444, "step": 14360 }, { "grad_norm": 0.44218891859054565, "learning_rate": 9.547699345726972e-05, "loss": 0.0549, "step": 14370 }, { "grad_norm": 0.5873460173606873, "learning_rate": 9.546839945163257e-05, "loss": 0.0526, "step": 14380 }, { "grad_norm": 0.3450511395931244, "learning_rate": 9.545979767668953e-05, "loss": 0.0617, "step": 14390 }, { "grad_norm": 0.4169127345085144, "learning_rate": 9.54511881339104e-05, "loss": 0.0489, "step": 14400 }, { "grad_norm": 0.7831217646598816, "learning_rate": 9.54425708247663e-05, "loss": 0.0513, "step": 14410 }, { "grad_norm": 0.6945239901542664, "learning_rate": 9.543394575072972e-05, "loss": 0.0486, "step": 14420 }, { "grad_norm": 0.3433079123497009, "learning_rate": 9.542531291327441e-05, "loss": 0.0493, "step": 14430 }, { "grad_norm": 0.6531136631965637, "learning_rate": 9.541667231387552e-05, "loss": 0.0428, "step": 14440 }, { "grad_norm": 0.5724310278892517, "learning_rate": 9.540802395400949e-05, "loss": 0.0569, "step": 14450 }, { "grad_norm": 0.5031688809394836, "learning_rate": 9.539936783515406e-05, "loss": 0.0573, "step": 14460 }, { "grad_norm": 0.48783373832702637, "learning_rate": 9.539070395878835e-05, "loss": 0.0616, "step": 14470 }, { "grad_norm": 0.45027631521224976, "learning_rate": 9.538203232639277e-05, "loss": 0.0462, "step": 14480 }, { "grad_norm": 0.40337392687797546, "learning_rate": 9.537335293944907e-05, "loss": 0.0635, "step": 14490 }, { "grad_norm": 0.36269596219062805, "learning_rate": 9.536466579944032e-05, "loss": 0.0712, "step": 14500 }, { "grad_norm": 0.4378304183483124, "learning_rate": 9.535597090785091e-05, "loss": 0.0601, "step": 14510 }, { "grad_norm": 0.5752530097961426, "learning_rate": 9.534726826616656e-05, "loss": 0.0628, "step": 14520 }, { "grad_norm": 0.4455508589744568, "learning_rate": 9.53385578758743e-05, "loss": 0.0796, "step": 14530 }, { "grad_norm": 0.45398589968681335, "learning_rate": 9.532983973846252e-05, "loss": 0.0536, "step": 14540 }, { "grad_norm": 0.34208354353904724, "learning_rate": 9.53211138554209e-05, "loss": 0.0604, "step": 14550 }, { "grad_norm": 0.7783527970314026, "learning_rate": 9.531238022824047e-05, "loss": 0.0598, "step": 14560 }, { "grad_norm": 0.4132677912712097, "learning_rate": 9.530363885841355e-05, "loss": 0.0484, "step": 14570 }, { "grad_norm": 0.46608254313468933, "learning_rate": 9.52948897474338e-05, "loss": 0.066, "step": 14580 }, { "grad_norm": 0.5217907428741455, "learning_rate": 9.528613289679622e-05, "loss": 0.0621, "step": 14590 }, { "grad_norm": 0.41059574484825134, "learning_rate": 9.52773683079971e-05, "loss": 0.0554, "step": 14600 }, { "grad_norm": 0.5236170291900635, "learning_rate": 9.526859598253407e-05, "loss": 0.0499, "step": 14610 }, { "grad_norm": 0.47892269492149353, "learning_rate": 9.525981592190609e-05, "loss": 0.0504, "step": 14620 }, { "grad_norm": 0.596693754196167, "learning_rate": 9.525102812761342e-05, "loss": 0.0508, "step": 14630 }, { "grad_norm": 0.5188814401626587, "learning_rate": 9.524223260115768e-05, "loss": 0.0459, "step": 14640 }, { "grad_norm": 0.5433838963508606, "learning_rate": 9.523342934404175e-05, "loss": 0.0488, "step": 14650 }, { "grad_norm": 0.3700965642929077, "learning_rate": 9.522461835776989e-05, "loss": 0.0472, "step": 14660 }, { "grad_norm": 0.4500696361064911, "learning_rate": 9.521579964384764e-05, "loss": 0.0584, "step": 14670 }, { "grad_norm": 0.6134686470031738, "learning_rate": 9.52069732037819e-05, "loss": 0.0463, "step": 14680 }, { "grad_norm": 0.41970425844192505, "learning_rate": 9.519813903908083e-05, "loss": 0.0489, "step": 14690 }, { "grad_norm": 0.5445101857185364, "learning_rate": 9.5189297151254e-05, "loss": 0.0535, "step": 14700 }, { "grad_norm": 0.5883059501647949, "learning_rate": 9.518044754181218e-05, "loss": 0.0653, "step": 14710 }, { "grad_norm": 0.4356693625450134, "learning_rate": 9.51715902122676e-05, "loss": 0.06, "step": 14720 }, { "grad_norm": 0.3423433303833008, "learning_rate": 9.516272516413368e-05, "loss": 0.054, "step": 14730 }, { "grad_norm": 0.39761367440223694, "learning_rate": 9.515385239892525e-05, "loss": 0.0671, "step": 14740 }, { "grad_norm": 0.5087958574295044, "learning_rate": 9.514497191815839e-05, "loss": 0.0598, "step": 14750 }, { "grad_norm": 0.45307472348213196, "learning_rate": 9.513608372335055e-05, "loss": 0.0574, "step": 14760 }, { "grad_norm": 0.5207860469818115, "learning_rate": 9.512718781602045e-05, "loss": 0.0547, "step": 14770 }, { "grad_norm": 0.44787120819091797, "learning_rate": 9.511828419768823e-05, "loss": 0.0601, "step": 14780 }, { "grad_norm": 0.3703463077545166, "learning_rate": 9.510937286987521e-05, "loss": 0.0393, "step": 14790 }, { "grad_norm": 0.5475189089775085, "learning_rate": 9.510045383410408e-05, "loss": 0.0551, "step": 14800 }, { "grad_norm": 0.3500477075576782, "learning_rate": 9.509152709189892e-05, "loss": 0.0473, "step": 14810 }, { "grad_norm": 0.42863303422927856, "learning_rate": 9.508259264478504e-05, "loss": 0.0558, "step": 14820 }, { "grad_norm": 0.45269834995269775, "learning_rate": 9.507365049428909e-05, "loss": 0.0645, "step": 14830 }, { "grad_norm": 0.8591776490211487, "learning_rate": 9.506470064193902e-05, "loss": 0.0747, "step": 14840 }, { "grad_norm": 0.37637859582901, "learning_rate": 9.505574308926414e-05, "loss": 0.0689, "step": 14850 }, { "grad_norm": 0.455636590719223, "learning_rate": 9.504677783779505e-05, "loss": 0.0516, "step": 14860 }, { "grad_norm": 0.37094956636428833, "learning_rate": 9.503780488906365e-05, "loss": 0.0485, "step": 14870 }, { "grad_norm": 0.49238747358322144, "learning_rate": 9.502882424460319e-05, "loss": 0.054, "step": 14880 }, { "grad_norm": 0.5393867492675781, "learning_rate": 9.501983590594821e-05, "loss": 0.0538, "step": 14890 }, { "grad_norm": 0.6286449432373047, "learning_rate": 9.501083987463455e-05, "loss": 0.0539, "step": 14900 }, { "grad_norm": 0.5057975649833679, "learning_rate": 9.500183615219942e-05, "loss": 0.0577, "step": 14910 }, { "grad_norm": 0.4670957326889038, "learning_rate": 9.49928247401813e-05, "loss": 0.0557, "step": 14920 }, { "grad_norm": 0.5054927468299866, "learning_rate": 9.498380564011997e-05, "loss": 0.0496, "step": 14930 }, { "grad_norm": 0.5878849625587463, "learning_rate": 9.497477885355656e-05, "loss": 0.059, "step": 14940 }, { "grad_norm": 0.43845856189727783, "learning_rate": 9.496574438203353e-05, "loss": 0.063, "step": 14950 }, { "grad_norm": 0.45041292905807495, "learning_rate": 9.495670222709459e-05, "loss": 0.0561, "step": 14960 }, { "grad_norm": 0.4646911323070526, "learning_rate": 9.494765239028483e-05, "loss": 0.0512, "step": 14970 }, { "grad_norm": 0.4453141987323761, "learning_rate": 9.493859487315057e-05, "loss": 0.0482, "step": 14980 }, { "grad_norm": 0.3955843150615692, "learning_rate": 9.492952967723953e-05, "loss": 0.0544, "step": 14990 }, { "grad_norm": 0.4217848479747772, "learning_rate": 9.492045680410068e-05, "loss": 0.0606, "step": 15000 }, { "grad_norm": 0.3853559195995331, "learning_rate": 9.491137625528436e-05, "loss": 0.0522, "step": 15010 }, { "grad_norm": 0.598393440246582, "learning_rate": 9.490228803234215e-05, "loss": 0.0515, "step": 15020 }, { "grad_norm": 0.5804033279418945, "learning_rate": 9.489319213682701e-05, "loss": 0.0538, "step": 15030 }, { "grad_norm": 0.48119306564331055, "learning_rate": 9.488408857029316e-05, "loss": 0.0612, "step": 15040 }, { "grad_norm": 0.4721461534500122, "learning_rate": 9.487497733429616e-05, "loss": 0.0697, "step": 15050 }, { "grad_norm": 0.4699365198612213, "learning_rate": 9.486585843039286e-05, "loss": 0.0615, "step": 15060 }, { "grad_norm": 0.573897123336792, "learning_rate": 9.485673186014143e-05, "loss": 0.0576, "step": 15070 }, { "grad_norm": 0.41530823707580566, "learning_rate": 9.484759762510137e-05, "loss": 0.0659, "step": 15080 }, { "grad_norm": 0.7794461250305176, "learning_rate": 9.483845572683346e-05, "loss": 0.0603, "step": 15090 }, { "grad_norm": 0.3545230031013489, "learning_rate": 9.48293061668998e-05, "loss": 0.0553, "step": 15100 }, { "grad_norm": 0.4892611503601074, "learning_rate": 9.48201489468638e-05, "loss": 0.053, "step": 15110 }, { "grad_norm": 0.37302663922309875, "learning_rate": 9.481098406829016e-05, "loss": 0.047, "step": 15120 }, { "grad_norm": 0.6219609975814819, "learning_rate": 9.480181153274495e-05, "loss": 0.051, "step": 15130 }, { "grad_norm": 0.5452433824539185, "learning_rate": 9.479263134179548e-05, "loss": 0.0489, "step": 15140 }, { "grad_norm": 0.5391332507133484, "learning_rate": 9.478344349701039e-05, "loss": 0.0518, "step": 15150 }, { "grad_norm": 0.4208994507789612, "learning_rate": 9.477424799995964e-05, "loss": 0.0506, "step": 15160 }, { "grad_norm": 0.43583858013153076, "learning_rate": 9.476504485221448e-05, "loss": 0.0511, "step": 15170 }, { "grad_norm": 0.4935840964317322, "learning_rate": 9.475583405534748e-05, "loss": 0.0475, "step": 15180 }, { "grad_norm": 0.5593347549438477, "learning_rate": 9.474661561093251e-05, "loss": 0.0496, "step": 15190 }, { "grad_norm": 0.4071085453033447, "learning_rate": 9.473738952054478e-05, "loss": 0.0563, "step": 15200 }, { "grad_norm": 0.29276221990585327, "learning_rate": 9.472815578576073e-05, "loss": 0.0517, "step": 15210 }, { "grad_norm": 0.4944506287574768, "learning_rate": 9.471891440815817e-05, "loss": 0.0487, "step": 15220 }, { "grad_norm": 0.5545539855957031, "learning_rate": 9.470966538931621e-05, "loss": 0.0402, "step": 15230 }, { "grad_norm": 0.40136751532554626, "learning_rate": 9.470040873081525e-05, "loss": 0.0438, "step": 15240 }, { "grad_norm": 0.4709755480289459, "learning_rate": 9.469114443423698e-05, "loss": 0.0613, "step": 15250 }, { "grad_norm": 0.8785680532455444, "learning_rate": 9.468187250116445e-05, "loss": 0.0521, "step": 15260 }, { "grad_norm": 0.5651416778564453, "learning_rate": 9.467259293318197e-05, "loss": 0.0555, "step": 15270 }, { "grad_norm": 0.4706498682498932, "learning_rate": 9.466330573187514e-05, "loss": 0.0493, "step": 15280 }, { "grad_norm": 0.39575037360191345, "learning_rate": 9.46540108988309e-05, "loss": 0.0686, "step": 15290 }, { "grad_norm": 0.7380779981613159, "learning_rate": 9.46447084356375e-05, "loss": 0.0607, "step": 15300 }, { "grad_norm": 0.3878534734249115, "learning_rate": 9.463539834388447e-05, "loss": 0.0701, "step": 15310 }, { "grad_norm": 0.4243421256542206, "learning_rate": 9.462608062516263e-05, "loss": 0.0419, "step": 15320 }, { "grad_norm": 0.5342383980751038, "learning_rate": 9.461675528106413e-05, "loss": 0.0464, "step": 15330 }, { "grad_norm": 0.35047420859336853, "learning_rate": 9.460742231318244e-05, "loss": 0.0555, "step": 15340 }, { "grad_norm": 0.6395313143730164, "learning_rate": 9.459808172311229e-05, "loss": 0.0616, "step": 15350 }, { "grad_norm": 0.7477926015853882, "learning_rate": 9.458873351244972e-05, "loss": 0.0475, "step": 15360 }, { "grad_norm": 0.8705278038978577, "learning_rate": 9.457937768279211e-05, "loss": 0.0557, "step": 15370 }, { "grad_norm": 0.39385491609573364, "learning_rate": 9.45700142357381e-05, "loss": 0.0492, "step": 15380 }, { "grad_norm": 0.5723256468772888, "learning_rate": 9.456064317288765e-05, "loss": 0.0484, "step": 15390 }, { "grad_norm": 0.3901405334472656, "learning_rate": 9.455126449584201e-05, "loss": 0.0541, "step": 15400 }, { "grad_norm": 0.5488235354423523, "learning_rate": 9.454187820620375e-05, "loss": 0.0456, "step": 15410 }, { "grad_norm": 0.3850806653499603, "learning_rate": 9.453248430557673e-05, "loss": 0.06, "step": 15420 }, { "grad_norm": 0.696812629699707, "learning_rate": 9.452308279556611e-05, "loss": 0.0737, "step": 15430 }, { "grad_norm": 0.4113246500492096, "learning_rate": 9.451367367777835e-05, "loss": 0.0498, "step": 15440 }, { "grad_norm": 0.3668394088745117, "learning_rate": 9.450425695382122e-05, "loss": 0.0575, "step": 15450 }, { "grad_norm": 0.4918607473373413, "learning_rate": 9.449483262530375e-05, "loss": 0.053, "step": 15460 }, { "grad_norm": 0.5017022490501404, "learning_rate": 9.448540069383633e-05, "loss": 0.0539, "step": 15470 }, { "grad_norm": 0.5089495778083801, "learning_rate": 9.447596116103061e-05, "loss": 0.0565, "step": 15480 }, { "grad_norm": 0.41574713587760925, "learning_rate": 9.446651402849955e-05, "loss": 0.0578, "step": 15490 }, { "grad_norm": 0.3968159258365631, "learning_rate": 9.44570592978574e-05, "loss": 0.05, "step": 15500 }, { "grad_norm": 0.296266108751297, "learning_rate": 9.444759697071972e-05, "loss": 0.0555, "step": 15510 }, { "grad_norm": 0.699704647064209, "learning_rate": 9.443812704870336e-05, "loss": 0.0442, "step": 15520 }, { "grad_norm": 0.4064404368400574, "learning_rate": 9.442864953342649e-05, "loss": 0.0469, "step": 15530 }, { "grad_norm": 0.38790979981422424, "learning_rate": 9.441916442650852e-05, "loss": 0.056, "step": 15540 }, { "grad_norm": 0.49546992778778076, "learning_rate": 9.440967172957023e-05, "loss": 0.0477, "step": 15550 }, { "grad_norm": 0.5492532253265381, "learning_rate": 9.440017144423364e-05, "loss": 0.0663, "step": 15560 }, { "grad_norm": 0.5768803358078003, "learning_rate": 9.439066357212209e-05, "loss": 0.0629, "step": 15570 }, { "grad_norm": 0.46429455280303955, "learning_rate": 9.438114811486022e-05, "loss": 0.06, "step": 15580 }, { "grad_norm": 0.35947078466415405, "learning_rate": 9.4371625074074e-05, "loss": 0.0463, "step": 15590 }, { "grad_norm": 0.4345625638961792, "learning_rate": 9.436209445139059e-05, "loss": 0.051, "step": 15600 }, { "grad_norm": 0.3832964301109314, "learning_rate": 9.435255624843855e-05, "loss": 0.0501, "step": 15610 }, { "grad_norm": 0.3888304531574249, "learning_rate": 9.43430104668477e-05, "loss": 0.044, "step": 15620 }, { "grad_norm": 0.376471608877182, "learning_rate": 9.433345710824914e-05, "loss": 0.0543, "step": 15630 }, { "grad_norm": 0.47813355922698975, "learning_rate": 9.432389617427529e-05, "loss": 0.0494, "step": 15640 }, { "grad_norm": 0.37987300753593445, "learning_rate": 9.431432766655984e-05, "loss": 0.0481, "step": 15650 }, { "grad_norm": 0.5421223640441895, "learning_rate": 9.430475158673778e-05, "loss": 0.0557, "step": 15660 }, { "grad_norm": 0.6208243370056152, "learning_rate": 9.429516793644542e-05, "loss": 0.0512, "step": 15670 }, { "grad_norm": 0.31623631715774536, "learning_rate": 9.428557671732034e-05, "loss": 0.0394, "step": 15680 }, { "grad_norm": 0.3435785174369812, "learning_rate": 9.42759779310014e-05, "loss": 0.0492, "step": 15690 }, { "grad_norm": 0.42648348212242126, "learning_rate": 9.426637157912879e-05, "loss": 0.0506, "step": 15700 }, { "grad_norm": 0.371799498796463, "learning_rate": 9.425675766334397e-05, "loss": 0.0503, "step": 15710 }, { "grad_norm": 0.5044535994529724, "learning_rate": 9.424713618528968e-05, "loss": 0.0402, "step": 15720 }, { "grad_norm": 0.4580181837081909, "learning_rate": 9.423750714661e-05, "loss": 0.065, "step": 15730 }, { "grad_norm": 0.5894309878349304, "learning_rate": 9.422787054895022e-05, "loss": 0.0573, "step": 15740 }, { "grad_norm": 0.5065669417381287, "learning_rate": 9.4218226393957e-05, "loss": 0.0656, "step": 15750 }, { "grad_norm": 0.4363287389278412, "learning_rate": 9.420857468327828e-05, "loss": 0.0534, "step": 15760 }, { "grad_norm": 0.5843654870986938, "learning_rate": 9.419891541856323e-05, "loss": 0.0521, "step": 15770 }, { "grad_norm": 0.4300498962402344, "learning_rate": 9.41892486014624e-05, "loss": 0.0442, "step": 15780 }, { "grad_norm": 0.5040595531463623, "learning_rate": 9.417957423362756e-05, "loss": 0.0604, "step": 15790 }, { "grad_norm": 0.4320857524871826, "learning_rate": 9.416989231671178e-05, "loss": 0.0519, "step": 15800 }, { "grad_norm": 0.4675910472869873, "learning_rate": 9.416020285236946e-05, "loss": 0.0517, "step": 15810 }, { "grad_norm": 0.44464945793151855, "learning_rate": 9.415050584225626e-05, "loss": 0.0454, "step": 15820 }, { "grad_norm": 0.4904812276363373, "learning_rate": 9.414080128802914e-05, "loss": 0.0525, "step": 15830 }, { "grad_norm": 0.40677329897880554, "learning_rate": 9.413108919134632e-05, "loss": 0.0475, "step": 15840 }, { "grad_norm": 0.5896919369697571, "learning_rate": 9.412136955386734e-05, "loss": 0.0523, "step": 15850 }, { "grad_norm": 0.48947563767433167, "learning_rate": 9.411164237725303e-05, "loss": 0.0491, "step": 15860 }, { "grad_norm": 0.3660587966442108, "learning_rate": 9.41019076631655e-05, "loss": 0.0527, "step": 15870 }, { "grad_norm": 0.4314135015010834, "learning_rate": 9.409216541326815e-05, "loss": 0.0602, "step": 15880 }, { "grad_norm": 0.5027595162391663, "learning_rate": 9.408241562922564e-05, "loss": 0.0551, "step": 15890 }, { "grad_norm": 0.4491422474384308, "learning_rate": 9.407265831270395e-05, "loss": 0.05, "step": 15900 }, { "grad_norm": 0.4829779863357544, "learning_rate": 9.406289346537035e-05, "loss": 0.0529, "step": 15910 }, { "grad_norm": 0.5555915832519531, "learning_rate": 9.405312108889339e-05, "loss": 0.0539, "step": 15920 }, { "grad_norm": 0.47634339332580566, "learning_rate": 9.404334118494288e-05, "loss": 0.0541, "step": 15930 }, { "grad_norm": 0.3734493851661682, "learning_rate": 9.403355375518995e-05, "loss": 0.0439, "step": 15940 }, { "grad_norm": 0.40348395705223083, "learning_rate": 9.4023758801307e-05, "loss": 0.0507, "step": 15950 }, { "grad_norm": 0.3858911991119385, "learning_rate": 9.401395632496774e-05, "loss": 0.0438, "step": 15960 }, { "grad_norm": 0.39586785435676575, "learning_rate": 9.400414632784711e-05, "loss": 0.0541, "step": 15970 }, { "grad_norm": 0.5144167542457581, "learning_rate": 9.39943288116214e-05, "loss": 0.049, "step": 15980 }, { "grad_norm": 0.537800669670105, "learning_rate": 9.398450377796815e-05, "loss": 0.053, "step": 15990 }, { "grad_norm": 0.43327343463897705, "learning_rate": 9.397467122856616e-05, "loss": 0.0525, "step": 16000 }, { "grad_norm": 0.6503564715385437, "learning_rate": 9.396483116509558e-05, "loss": 0.0578, "step": 16010 }, { "grad_norm": 0.45918136835098267, "learning_rate": 9.39549835892378e-05, "loss": 0.0595, "step": 16020 }, { "grad_norm": 0.45125091075897217, "learning_rate": 9.39451285026755e-05, "loss": 0.0565, "step": 16030 }, { "grad_norm": 0.42736542224884033, "learning_rate": 9.393526590709262e-05, "loss": 0.0561, "step": 16040 }, { "grad_norm": 0.4872633218765259, "learning_rate": 9.392539580417444e-05, "loss": 0.0475, "step": 16050 }, { "grad_norm": 0.5425727367401123, "learning_rate": 9.391551819560747e-05, "loss": 0.0537, "step": 16060 }, { "grad_norm": 0.4475146532058716, "learning_rate": 9.390563308307955e-05, "loss": 0.0415, "step": 16070 }, { "grad_norm": 0.4277302026748657, "learning_rate": 9.389574046827974e-05, "loss": 0.0481, "step": 16080 }, { "grad_norm": 0.4282088279724121, "learning_rate": 9.388584035289845e-05, "loss": 0.0638, "step": 16090 }, { "grad_norm": 0.45165497064590454, "learning_rate": 9.387593273862732e-05, "loss": 0.0606, "step": 16100 }, { "grad_norm": 0.42237478494644165, "learning_rate": 9.386601762715929e-05, "loss": 0.0599, "step": 16110 }, { "grad_norm": 0.4068446159362793, "learning_rate": 9.38560950201886e-05, "loss": 0.062, "step": 16120 }, { "grad_norm": 0.757282018661499, "learning_rate": 9.384616491941071e-05, "loss": 0.0608, "step": 16130 }, { "grad_norm": 0.697412371635437, "learning_rate": 9.383622732652245e-05, "loss": 0.0592, "step": 16140 }, { "grad_norm": 0.41971299052238464, "learning_rate": 9.382628224322187e-05, "loss": 0.0572, "step": 16150 }, { "grad_norm": 0.3989499509334564, "learning_rate": 9.381632967120829e-05, "loss": 0.0491, "step": 16160 }, { "grad_norm": 0.44794875383377075, "learning_rate": 9.380636961218235e-05, "loss": 0.0454, "step": 16170 }, { "grad_norm": 0.5472517013549805, "learning_rate": 9.379640206784597e-05, "loss": 0.0467, "step": 16180 }, { "grad_norm": 0.4878322184085846, "learning_rate": 9.378642703990229e-05, "loss": 0.0491, "step": 16190 }, { "grad_norm": 0.413239061832428, "learning_rate": 9.37764445300558e-05, "loss": 0.0522, "step": 16200 }, { "grad_norm": 0.3681265711784363, "learning_rate": 9.376645454001222e-05, "loss": 0.0569, "step": 16210 }, { "grad_norm": 0.5028650760650635, "learning_rate": 9.375645707147858e-05, "loss": 0.0485, "step": 16220 }, { "grad_norm": 0.5103313326835632, "learning_rate": 9.374645212616316e-05, "loss": 0.0602, "step": 16230 }, { "grad_norm": 0.4375486969947815, "learning_rate": 9.373643970577555e-05, "loss": 0.0499, "step": 16240 }, { "grad_norm": 0.5011734366416931, "learning_rate": 9.372641981202659e-05, "loss": 0.0469, "step": 16250 }, { "grad_norm": 0.48417970538139343, "learning_rate": 9.37163924466284e-05, "loss": 0.0575, "step": 16260 }, { "grad_norm": 0.44506898522377014, "learning_rate": 9.370635761129438e-05, "loss": 0.0521, "step": 16270 }, { "grad_norm": 0.3775123357772827, "learning_rate": 9.36963153077392e-05, "loss": 0.0493, "step": 16280 }, { "grad_norm": 0.3168901205062866, "learning_rate": 9.368626553767888e-05, "loss": 0.044, "step": 16290 }, { "grad_norm": 0.5100103616714478, "learning_rate": 9.367620830283057e-05, "loss": 0.0421, "step": 16300 }, { "grad_norm": 0.5866093039512634, "learning_rate": 9.366614360491281e-05, "loss": 0.0599, "step": 16310 }, { "grad_norm": 0.4777442514896393, "learning_rate": 9.365607144564539e-05, "loss": 0.0481, "step": 16320 }, { "grad_norm": 0.44923558831214905, "learning_rate": 9.364599182674934e-05, "loss": 0.0597, "step": 16330 }, { "grad_norm": 0.42120039463043213, "learning_rate": 9.3635904749947e-05, "loss": 0.0511, "step": 16340 }, { "grad_norm": 0.32138171792030334, "learning_rate": 9.362581021696202e-05, "loss": 0.055, "step": 16350 }, { "grad_norm": 0.5124514102935791, "learning_rate": 9.361570822951921e-05, "loss": 0.0534, "step": 16360 }, { "grad_norm": 0.49569687247276306, "learning_rate": 9.360559878934476e-05, "loss": 0.0571, "step": 16370 }, { "grad_norm": 0.44159579277038574, "learning_rate": 9.359548189816611e-05, "loss": 0.0516, "step": 16380 }, { "grad_norm": 0.38666078448295593, "learning_rate": 9.358535755771193e-05, "loss": 0.0592, "step": 16390 }, { "grad_norm": 0.7684077620506287, "learning_rate": 9.357522576971221e-05, "loss": 0.0611, "step": 16400 }, { "grad_norm": 0.3733054995536804, "learning_rate": 9.356508653589819e-05, "loss": 0.0629, "step": 16410 }, { "grad_norm": 0.41125914454460144, "learning_rate": 9.355493985800237e-05, "loss": 0.0514, "step": 16420 }, { "grad_norm": 0.4141237437725067, "learning_rate": 9.354478573775857e-05, "loss": 0.0541, "step": 16430 }, { "grad_norm": 0.3004266023635864, "learning_rate": 9.353462417690186e-05, "loss": 0.0495, "step": 16440 }, { "grad_norm": 0.5578495860099792, "learning_rate": 9.352445517716853e-05, "loss": 0.05, "step": 16450 }, { "grad_norm": 0.5376029014587402, "learning_rate": 9.351427874029621e-05, "loss": 0.0534, "step": 16460 }, { "grad_norm": 0.618778645992279, "learning_rate": 9.350409486802379e-05, "loss": 0.0542, "step": 16470 }, { "grad_norm": 0.541602373123169, "learning_rate": 9.349390356209138e-05, "loss": 0.0517, "step": 16480 }, { "grad_norm": 0.3557041883468628, "learning_rate": 9.348370482424042e-05, "loss": 0.0522, "step": 16490 }, { "grad_norm": 0.47963613271713257, "learning_rate": 9.347349865621357e-05, "loss": 0.0492, "step": 16500 }, { "grad_norm": 0.40665605664253235, "learning_rate": 9.346328505975481e-05, "loss": 0.0465, "step": 16510 }, { "grad_norm": 0.42446762323379517, "learning_rate": 9.345306403660936e-05, "loss": 0.053, "step": 16520 }, { "grad_norm": 0.5759106278419495, "learning_rate": 9.344283558852371e-05, "loss": 0.054, "step": 16530 }, { "grad_norm": 0.6869299411773682, "learning_rate": 9.343259971724563e-05, "loss": 0.0533, "step": 16540 }, { "grad_norm": 0.7691987752914429, "learning_rate": 9.342235642452413e-05, "loss": 0.0684, "step": 16550 }, { "grad_norm": 0.6961710453033447, "learning_rate": 9.341210571210954e-05, "loss": 0.0576, "step": 16560 }, { "grad_norm": 0.565174400806427, "learning_rate": 9.340184758175338e-05, "loss": 0.0512, "step": 16570 }, { "grad_norm": 0.5464860200881958, "learning_rate": 9.339158203520854e-05, "loss": 0.0627, "step": 16580 }, { "grad_norm": 0.5552876591682434, "learning_rate": 9.338130907422908e-05, "loss": 0.0415, "step": 16590 }, { "grad_norm": 0.5503347516059875, "learning_rate": 9.337102870057037e-05, "loss": 0.0399, "step": 16600 }, { "grad_norm": 0.406472772359848, "learning_rate": 9.336074091598907e-05, "loss": 0.0487, "step": 16610 }, { "grad_norm": 0.49037617444992065, "learning_rate": 9.335044572224306e-05, "loss": 0.0487, "step": 16620 }, { "grad_norm": 0.3868565559387207, "learning_rate": 9.334014312109151e-05, "loss": 0.0393, "step": 16630 }, { "grad_norm": 0.5958299040794373, "learning_rate": 9.332983311429486e-05, "loss": 0.0635, "step": 16640 }, { "grad_norm": 0.38823655247688293, "learning_rate": 9.33195157036148e-05, "loss": 0.0573, "step": 16650 }, { "grad_norm": 0.5335453152656555, "learning_rate": 9.330919089081432e-05, "loss": 0.051, "step": 16660 }, { "grad_norm": 0.4134352505207062, "learning_rate": 9.32988586776576e-05, "loss": 0.0725, "step": 16670 }, { "grad_norm": 0.45132794976234436, "learning_rate": 9.328851906591016e-05, "loss": 0.0531, "step": 16680 }, { "grad_norm": 0.5678310990333557, "learning_rate": 9.327817205733875e-05, "loss": 0.0552, "step": 16690 }, { "grad_norm": 0.4684636890888214, "learning_rate": 9.326781765371142e-05, "loss": 0.0453, "step": 16700 }, { "grad_norm": 0.5771246552467346, "learning_rate": 9.325745585679741e-05, "loss": 0.0525, "step": 16710 }, { "grad_norm": 0.4956587851047516, "learning_rate": 9.32470866683673e-05, "loss": 0.0639, "step": 16720 }, { "grad_norm": 0.5004554986953735, "learning_rate": 9.323671009019288e-05, "loss": 0.0538, "step": 16730 }, { "grad_norm": 0.4519786834716797, "learning_rate": 9.322632612404725e-05, "loss": 0.0512, "step": 16740 }, { "grad_norm": 0.4525945484638214, "learning_rate": 9.321593477170471e-05, "loss": 0.0544, "step": 16750 }, { "grad_norm": 0.5296329259872437, "learning_rate": 9.320553603494088e-05, "loss": 0.0447, "step": 16760 }, { "grad_norm": 0.5996739864349365, "learning_rate": 9.319512991553261e-05, "loss": 0.0449, "step": 16770 }, { "grad_norm": 0.46033531427383423, "learning_rate": 9.318471641525803e-05, "loss": 0.0564, "step": 16780 }, { "grad_norm": 0.525039792060852, "learning_rate": 9.317429553589652e-05, "loss": 0.0474, "step": 16790 }, { "grad_norm": 0.4480270445346832, "learning_rate": 9.316386727922873e-05, "loss": 0.0579, "step": 16800 }, { "grad_norm": 0.38880887627601624, "learning_rate": 9.315343164703656e-05, "loss": 0.0467, "step": 16810 }, { "grad_norm": 0.3870995342731476, "learning_rate": 9.314298864110316e-05, "loss": 0.0417, "step": 16820 }, { "grad_norm": 0.5439058542251587, "learning_rate": 9.313253826321295e-05, "loss": 0.0426, "step": 16830 }, { "grad_norm": 0.3464486598968506, "learning_rate": 9.312208051515165e-05, "loss": 0.0441, "step": 16840 }, { "grad_norm": 0.30791887640953064, "learning_rate": 9.311161539870618e-05, "loss": 0.0461, "step": 16850 }, { "grad_norm": 0.42002958059310913, "learning_rate": 9.310114291566474e-05, "loss": 0.0544, "step": 16860 }, { "grad_norm": 0.4959518611431122, "learning_rate": 9.309066306781679e-05, "loss": 0.0432, "step": 16870 }, { "grad_norm": 0.4774094820022583, "learning_rate": 9.308017585695306e-05, "loss": 0.0465, "step": 16880 }, { "grad_norm": 0.4611276388168335, "learning_rate": 9.306968128486552e-05, "loss": 0.0484, "step": 16890 }, { "grad_norm": 0.5175912976264954, "learning_rate": 9.30591793533474e-05, "loss": 0.0445, "step": 16900 }, { "grad_norm": 0.6268205642700195, "learning_rate": 9.304867006419321e-05, "loss": 0.0495, "step": 16910 }, { "grad_norm": 0.565309464931488, "learning_rate": 9.303815341919868e-05, "loss": 0.0507, "step": 16920 }, { "grad_norm": 0.6268182396888733, "learning_rate": 9.302762942016084e-05, "loss": 0.065, "step": 16930 }, { "grad_norm": 0.4676123559474945, "learning_rate": 9.301709806887792e-05, "loss": 0.0628, "step": 16940 }, { "grad_norm": 0.5744362473487854, "learning_rate": 9.300655936714948e-05, "loss": 0.0779, "step": 16950 }, { "grad_norm": 0.47050711512565613, "learning_rate": 9.299601331677627e-05, "loss": 0.0623, "step": 16960 }, { "grad_norm": 0.4703686535358429, "learning_rate": 9.298545991956033e-05, "loss": 0.0599, "step": 16970 }, { "grad_norm": 0.3992998003959656, "learning_rate": 9.297489917730493e-05, "loss": 0.053, "step": 16980 }, { "grad_norm": 0.5125941634178162, "learning_rate": 9.296433109181464e-05, "loss": 0.0554, "step": 16990 }, { "grad_norm": 0.5622234344482422, "learning_rate": 9.295375566489523e-05, "loss": 0.0504, "step": 17000 }, { "grad_norm": 0.46487340331077576, "learning_rate": 9.294317289835379e-05, "loss": 0.057, "step": 17010 }, { "grad_norm": 0.4427585005760193, "learning_rate": 9.293258279399859e-05, "loss": 0.0638, "step": 17020 }, { "grad_norm": 0.48832058906555176, "learning_rate": 9.292198535363919e-05, "loss": 0.0509, "step": 17030 }, { "grad_norm": 0.5102348923683167, "learning_rate": 9.291138057908641e-05, "loss": 0.0531, "step": 17040 }, { "grad_norm": 0.5096217393875122, "learning_rate": 9.290076847215234e-05, "loss": 0.0487, "step": 17050 }, { "grad_norm": 0.6361953616142273, "learning_rate": 9.289014903465025e-05, "loss": 0.0695, "step": 17060 }, { "grad_norm": 0.46465763449668884, "learning_rate": 9.287952226839475e-05, "loss": 0.0478, "step": 17070 }, { "grad_norm": 0.7516945004463196, "learning_rate": 9.286888817520164e-05, "loss": 0.0503, "step": 17080 }, { "grad_norm": 0.5600035190582275, "learning_rate": 9.285824675688803e-05, "loss": 0.0545, "step": 17090 }, { "grad_norm": 0.4047950208187103, "learning_rate": 9.28475980152722e-05, "loss": 0.059, "step": 17100 }, { "grad_norm": 0.5458607077598572, "learning_rate": 9.283694195217379e-05, "loss": 0.0635, "step": 17110 }, { "grad_norm": 0.5628366470336914, "learning_rate": 9.282627856941356e-05, "loss": 0.0559, "step": 17120 }, { "grad_norm": 0.5002710819244385, "learning_rate": 9.281560786881363e-05, "loss": 0.0511, "step": 17130 }, { "grad_norm": 0.7595816254615784, "learning_rate": 9.280492985219733e-05, "loss": 0.0522, "step": 17140 }, { "grad_norm": 0.533601701259613, "learning_rate": 9.279424452138924e-05, "loss": 0.0549, "step": 17150 }, { "grad_norm": 0.6257062554359436, "learning_rate": 9.278355187821517e-05, "loss": 0.0614, "step": 17160 }, { "grad_norm": 0.4256041646003723, "learning_rate": 9.277285192450224e-05, "loss": 0.0533, "step": 17170 }, { "grad_norm": 0.5279727578163147, "learning_rate": 9.276214466207875e-05, "loss": 0.0544, "step": 17180 }, { "grad_norm": 0.5223628878593445, "learning_rate": 9.275143009277427e-05, "loss": 0.0469, "step": 17190 }, { "grad_norm": 0.5914530158042908, "learning_rate": 9.274070821841964e-05, "loss": 0.0521, "step": 17200 }, { "grad_norm": 0.44542330503463745, "learning_rate": 9.272997904084696e-05, "loss": 0.0487, "step": 17210 }, { "grad_norm": 0.5000082850456238, "learning_rate": 9.271924256188951e-05, "loss": 0.0576, "step": 17220 }, { "grad_norm": 0.5408269762992859, "learning_rate": 9.270849878338189e-05, "loss": 0.0625, "step": 17230 }, { "grad_norm": 0.4904961884021759, "learning_rate": 9.269774770715991e-05, "loss": 0.0475, "step": 17240 }, { "grad_norm": 0.39272966980934143, "learning_rate": 9.268698933506061e-05, "loss": 0.0607, "step": 17250 }, { "grad_norm": 0.3572113513946533, "learning_rate": 9.267622366892235e-05, "loss": 0.0494, "step": 17260 }, { "grad_norm": 0.28196704387664795, "learning_rate": 9.266545071058465e-05, "loss": 0.0445, "step": 17270 }, { "grad_norm": 0.4065991938114166, "learning_rate": 9.265467046188833e-05, "loss": 0.0474, "step": 17280 }, { "grad_norm": 0.39964696764945984, "learning_rate": 9.264388292467543e-05, "loss": 0.0532, "step": 17290 }, { "grad_norm": 0.4774635136127472, "learning_rate": 9.263308810078926e-05, "loss": 0.0613, "step": 17300 }, { "grad_norm": 0.5945859551429749, "learning_rate": 9.262228599207434e-05, "loss": 0.0555, "step": 17310 }, { "grad_norm": 0.45985543727874756, "learning_rate": 9.261147660037647e-05, "loss": 0.0562, "step": 17320 }, { "grad_norm": 0.5763592720031738, "learning_rate": 9.26006599275427e-05, "loss": 0.055, "step": 17330 }, { "grad_norm": 0.5220933556556702, "learning_rate": 9.258983597542124e-05, "loss": 0.0511, "step": 17340 }, { "grad_norm": 0.284938782453537, "learning_rate": 9.257900474586167e-05, "loss": 0.044, "step": 17350 }, { "grad_norm": 0.47578150033950806, "learning_rate": 9.256816624071471e-05, "loss": 0.0577, "step": 17360 }, { "grad_norm": 0.40856578946113586, "learning_rate": 9.25573204618324e-05, "loss": 0.0576, "step": 17370 }, { "grad_norm": 0.4387516975402832, "learning_rate": 9.254646741106796e-05, "loss": 0.0481, "step": 17380 }, { "grad_norm": 0.3015258312225342, "learning_rate": 9.253560709027589e-05, "loss": 0.048, "step": 17390 }, { "grad_norm": 0.5221129059791565, "learning_rate": 9.252473950131192e-05, "loss": 0.0521, "step": 17400 }, { "grad_norm": 0.5460434556007385, "learning_rate": 9.251386464603302e-05, "loss": 0.0455, "step": 17410 }, { "grad_norm": 0.4457576870918274, "learning_rate": 9.250298252629741e-05, "loss": 0.049, "step": 17420 }, { "grad_norm": 0.5870035290718079, "learning_rate": 9.249209314396454e-05, "loss": 0.0525, "step": 17430 }, { "grad_norm": 0.5322816371917725, "learning_rate": 9.248119650089513e-05, "loss": 0.0535, "step": 17440 }, { "grad_norm": 0.4262332320213318, "learning_rate": 9.247029259895108e-05, "loss": 0.0481, "step": 17450 }, { "grad_norm": 0.4736897051334381, "learning_rate": 9.24593814399956e-05, "loss": 0.0533, "step": 17460 }, { "grad_norm": 0.44566425681114197, "learning_rate": 9.244846302589309e-05, "loss": 0.0424, "step": 17470 }, { "grad_norm": 0.4342499077320099, "learning_rate": 9.243753735850923e-05, "loss": 0.0551, "step": 17480 }, { "grad_norm": 0.3538890480995178, "learning_rate": 9.24266044397109e-05, "loss": 0.0409, "step": 17490 }, { "grad_norm": 0.3644617795944214, "learning_rate": 9.241566427136624e-05, "loss": 0.045, "step": 17500 }, { "grad_norm": 0.4828670620918274, "learning_rate": 9.240471685534463e-05, "loss": 0.0551, "step": 17510 }, { "grad_norm": 0.3735526204109192, "learning_rate": 9.239376219351667e-05, "loss": 0.0551, "step": 17520 }, { "grad_norm": 0.43669265508651733, "learning_rate": 9.238280028775425e-05, "loss": 0.0461, "step": 17530 }, { "grad_norm": 0.4874038100242615, "learning_rate": 9.237183113993041e-05, "loss": 0.0485, "step": 17540 }, { "grad_norm": 0.4369938373565674, "learning_rate": 9.236085475191952e-05, "loss": 0.0444, "step": 17550 }, { "grad_norm": 0.43079739809036255, "learning_rate": 9.234987112559709e-05, "loss": 0.0632, "step": 17560 }, { "grad_norm": 0.41289886832237244, "learning_rate": 9.233888026283999e-05, "loss": 0.0501, "step": 17570 }, { "grad_norm": 0.41136911511421204, "learning_rate": 9.232788216552619e-05, "loss": 0.053, "step": 17580 }, { "grad_norm": 0.5089476108551025, "learning_rate": 9.231687683553502e-05, "loss": 0.0516, "step": 17590 }, { "grad_norm": 0.5134636759757996, "learning_rate": 9.230586427474698e-05, "loss": 0.0594, "step": 17600 }, { "grad_norm": 0.42459040880203247, "learning_rate": 9.229484448504379e-05, "loss": 0.0578, "step": 17610 }, { "grad_norm": 0.4205755293369293, "learning_rate": 9.228381746830843e-05, "loss": 0.0491, "step": 17620 }, { "grad_norm": 0.4281677305698395, "learning_rate": 9.227278322642514e-05, "loss": 0.042, "step": 17630 }, { "grad_norm": 0.5511592626571655, "learning_rate": 9.226174176127937e-05, "loss": 0.0491, "step": 17640 }, { "grad_norm": 0.52854323387146, "learning_rate": 9.22506930747578e-05, "loss": 0.0506, "step": 17650 }, { "grad_norm": 0.6166616678237915, "learning_rate": 9.223963716874831e-05, "loss": 0.0646, "step": 17660 }, { "grad_norm": 0.5213914513587952, "learning_rate": 9.222857404514012e-05, "loss": 0.0618, "step": 17670 }, { "grad_norm": 0.5303839445114136, "learning_rate": 9.221750370582355e-05, "loss": 0.0505, "step": 17680 }, { "grad_norm": 0.36492791771888733, "learning_rate": 9.220642615269028e-05, "loss": 0.0575, "step": 17690 }, { "grad_norm": 0.4627252221107483, "learning_rate": 9.219534138763311e-05, "loss": 0.0674, "step": 17700 }, { "grad_norm": 0.4538649916648865, "learning_rate": 9.218424941254613e-05, "loss": 0.0605, "step": 17710 }, { "grad_norm": 0.3922034502029419, "learning_rate": 9.217315022932468e-05, "loss": 0.0607, "step": 17720 }, { "grad_norm": 0.7968274354934692, "learning_rate": 9.216204383986528e-05, "loss": 0.0548, "step": 17730 }, { "grad_norm": 0.4107120931148529, "learning_rate": 9.215093024606574e-05, "loss": 0.0561, "step": 17740 }, { "grad_norm": 0.49138662219047546, "learning_rate": 9.213980944982506e-05, "loss": 0.0674, "step": 17750 }, { "grad_norm": 0.4447906017303467, "learning_rate": 9.212868145304346e-05, "loss": 0.06, "step": 17760 }, { "grad_norm": 0.44064861536026, "learning_rate": 9.211754625762241e-05, "loss": 0.0473, "step": 17770 }, { "grad_norm": 0.4025506377220154, "learning_rate": 9.210640386546463e-05, "loss": 0.0623, "step": 17780 }, { "grad_norm": 0.5083202719688416, "learning_rate": 9.209525427847405e-05, "loss": 0.0405, "step": 17790 }, { "grad_norm": 0.3208185136318207, "learning_rate": 9.208409749855583e-05, "loss": 0.048, "step": 17800 }, { "grad_norm": 0.49161240458488464, "learning_rate": 9.207293352761633e-05, "loss": 0.0537, "step": 17810 }, { "grad_norm": 0.42357125878334045, "learning_rate": 9.206176236756319e-05, "loss": 0.052, "step": 17820 }, { "grad_norm": 0.4170270562171936, "learning_rate": 9.205058402030525e-05, "loss": 0.05, "step": 17830 }, { "grad_norm": 0.3640647232532501, "learning_rate": 9.203939848775259e-05, "loss": 0.0427, "step": 17840 }, { "grad_norm": 0.40281808376312256, "learning_rate": 9.202820577181652e-05, "loss": 0.047, "step": 17850 }, { "grad_norm": 0.48868462443351746, "learning_rate": 9.201700587440953e-05, "loss": 0.0429, "step": 17860 }, { "grad_norm": 0.45118623971939087, "learning_rate": 9.200579879744544e-05, "loss": 0.0497, "step": 17870 }, { "grad_norm": 0.37819474935531616, "learning_rate": 9.199458454283918e-05, "loss": 0.0465, "step": 17880 }, { "grad_norm": 1.3071722984313965, "learning_rate": 9.198336311250697e-05, "loss": 0.0558, "step": 17890 }, { "grad_norm": 0.7020917534828186, "learning_rate": 9.197213450836626e-05, "loss": 0.0498, "step": 17900 }, { "grad_norm": 0.545120358467102, "learning_rate": 9.19608987323357e-05, "loss": 0.0575, "step": 17910 }, { "grad_norm": 0.5155004262924194, "learning_rate": 9.194965578633517e-05, "loss": 0.0486, "step": 17920 }, { "grad_norm": 0.3609928786754608, "learning_rate": 9.193840567228582e-05, "loss": 0.0466, "step": 17930 }, { "grad_norm": 0.3942502737045288, "learning_rate": 9.192714839210994e-05, "loss": 0.0472, "step": 17940 }, { "grad_norm": 0.43178537487983704, "learning_rate": 9.19158839477311e-05, "loss": 0.0589, "step": 17950 }, { "grad_norm": 0.4579649269580841, "learning_rate": 9.190461234107411e-05, "loss": 0.0491, "step": 17960 }, { "grad_norm": 0.5197175741195679, "learning_rate": 9.189333357406496e-05, "loss": 0.0464, "step": 17970 }, { "grad_norm": 0.5425653457641602, "learning_rate": 9.188204764863089e-05, "loss": 0.0623, "step": 17980 }, { "grad_norm": 0.6097222566604614, "learning_rate": 9.187075456670033e-05, "loss": 0.049, "step": 17990 }, { "grad_norm": 0.3613094389438629, "learning_rate": 9.1859454330203e-05, "loss": 0.0486, "step": 18000 }, { "grad_norm": 0.3896371126174927, "learning_rate": 9.18481469410698e-05, "loss": 0.0586, "step": 18010 }, { "grad_norm": 0.4369397461414337, "learning_rate": 9.183683240123281e-05, "loss": 0.0516, "step": 18020 }, { "grad_norm": 0.4605483412742615, "learning_rate": 9.182551071262541e-05, "loss": 0.0486, "step": 18030 }, { "grad_norm": 0.48611921072006226, "learning_rate": 9.181418187718218e-05, "loss": 0.052, "step": 18040 }, { "grad_norm": 0.49920594692230225, "learning_rate": 9.180284589683888e-05, "loss": 0.0566, "step": 18050 }, { "grad_norm": 0.5468576550483704, "learning_rate": 9.17915027735325e-05, "loss": 0.0508, "step": 18060 }, { "grad_norm": 0.4838389754295349, "learning_rate": 9.178015250920133e-05, "loss": 0.0533, "step": 18070 }, { "grad_norm": 0.37922239303588867, "learning_rate": 9.176879510578477e-05, "loss": 0.0518, "step": 18080 }, { "grad_norm": 0.4482899606227875, "learning_rate": 9.17574305652235e-05, "loss": 0.0447, "step": 18090 }, { "grad_norm": 0.5144628882408142, "learning_rate": 9.174605888945942e-05, "loss": 0.0565, "step": 18100 }, { "grad_norm": 0.5234751105308533, "learning_rate": 9.173468008043564e-05, "loss": 0.0484, "step": 18110 }, { "grad_norm": 0.44359201192855835, "learning_rate": 9.172329414009648e-05, "loss": 0.0499, "step": 18120 }, { "grad_norm": 0.48304611444473267, "learning_rate": 9.171190107038747e-05, "loss": 0.0519, "step": 18130 }, { "grad_norm": 0.4560583829879761, "learning_rate": 9.170050087325541e-05, "loss": 0.0473, "step": 18140 }, { "grad_norm": 0.4065675139427185, "learning_rate": 9.168909355064824e-05, "loss": 0.0471, "step": 18150 }, { "grad_norm": 0.3653090298175812, "learning_rate": 9.167767910451519e-05, "loss": 0.0506, "step": 18160 }, { "grad_norm": 0.42117738723754883, "learning_rate": 9.166625753680669e-05, "loss": 0.0489, "step": 18170 }, { "grad_norm": 0.5854431390762329, "learning_rate": 9.165482884947431e-05, "loss": 0.0427, "step": 18180 }, { "grad_norm": 0.36670389771461487, "learning_rate": 9.164339304447098e-05, "loss": 0.0551, "step": 18190 }, { "grad_norm": 0.5163714289665222, "learning_rate": 9.163195012375072e-05, "loss": 0.0427, "step": 18200 }, { "grad_norm": 0.367352694272995, "learning_rate": 9.16205000892688e-05, "loss": 0.0516, "step": 18210 }, { "grad_norm": 0.40028882026672363, "learning_rate": 9.160904294298175e-05, "loss": 0.0454, "step": 18220 }, { "grad_norm": 0.46279072761535645, "learning_rate": 9.159757868684727e-05, "loss": 0.0516, "step": 18230 }, { "grad_norm": 0.5353474020957947, "learning_rate": 9.15861073228243e-05, "loss": 0.0449, "step": 18240 }, { "grad_norm": 0.5797568559646606, "learning_rate": 9.157462885287296e-05, "loss": 0.0554, "step": 18250 }, { "grad_norm": 0.4386034309864044, "learning_rate": 9.156314327895461e-05, "loss": 0.0474, "step": 18260 }, { "grad_norm": 0.38369402289390564, "learning_rate": 9.155165060303185e-05, "loss": 0.0518, "step": 18270 }, { "grad_norm": 0.48126155138015747, "learning_rate": 9.154015082706841e-05, "loss": 0.0433, "step": 18280 }, { "grad_norm": 0.5027033090591431, "learning_rate": 9.152864395302936e-05, "loss": 0.0502, "step": 18290 }, { "grad_norm": 0.5321907997131348, "learning_rate": 9.151712998288085e-05, "loss": 0.0375, "step": 18300 }, { "grad_norm": 0.4437737762928009, "learning_rate": 9.150560891859031e-05, "loss": 0.0453, "step": 18310 }, { "grad_norm": 0.4213466942310333, "learning_rate": 9.14940807621264e-05, "loss": 0.041, "step": 18320 }, { "grad_norm": 0.45462602376937866, "learning_rate": 9.148254551545894e-05, "loss": 0.0441, "step": 18330 }, { "grad_norm": 0.4744494557380676, "learning_rate": 9.147100318055901e-05, "loss": 0.0432, "step": 18340 }, { "grad_norm": 0.5087177157402039, "learning_rate": 9.145945375939888e-05, "loss": 0.046, "step": 18350 }, { "grad_norm": 0.3782236576080322, "learning_rate": 9.144789725395203e-05, "loss": 0.0539, "step": 18360 }, { "grad_norm": 0.42650851607322693, "learning_rate": 9.14363336661931e-05, "loss": 0.0424, "step": 18370 }, { "grad_norm": 0.6664273738861084, "learning_rate": 9.142476299809806e-05, "loss": 0.0547, "step": 18380 }, { "grad_norm": 0.38700589537620544, "learning_rate": 9.1413185251644e-05, "loss": 0.0375, "step": 18390 }, { "grad_norm": 0.5006874799728394, "learning_rate": 9.140160042880923e-05, "loss": 0.049, "step": 18400 }, { "grad_norm": 0.3290030360221863, "learning_rate": 9.139000853157327e-05, "loss": 0.0544, "step": 18410 }, { "grad_norm": 0.6442640423774719, "learning_rate": 9.137840956191688e-05, "loss": 0.0524, "step": 18420 }, { "grad_norm": 0.3481347858905792, "learning_rate": 9.136680352182199e-05, "loss": 0.041, "step": 18430 }, { "grad_norm": 0.5605137944221497, "learning_rate": 9.135519041327177e-05, "loss": 0.0486, "step": 18440 }, { "grad_norm": 0.4528462886810303, "learning_rate": 9.134357023825058e-05, "loss": 0.0514, "step": 18450 }, { "grad_norm": 0.5024856328964233, "learning_rate": 9.133194299874398e-05, "loss": 0.0549, "step": 18460 }, { "grad_norm": 0.5865769982337952, "learning_rate": 9.132030869673876e-05, "loss": 0.0586, "step": 18470 }, { "grad_norm": 0.5129500031471252, "learning_rate": 9.130866733422288e-05, "loss": 0.0592, "step": 18480 }, { "grad_norm": 0.44921958446502686, "learning_rate": 9.129701891318556e-05, "loss": 0.0494, "step": 18490 }, { "grad_norm": 0.406440407037735, "learning_rate": 9.128536343561718e-05, "loss": 0.0611, "step": 18500 }, { "grad_norm": 0.54929119348526, "learning_rate": 9.127370090350934e-05, "loss": 0.0522, "step": 18510 }, { "grad_norm": 0.5049206018447876, "learning_rate": 9.126203131885487e-05, "loss": 0.0519, "step": 18520 }, { "grad_norm": 0.44279390573501587, "learning_rate": 9.125035468364775e-05, "loss": 0.0477, "step": 18530 }, { "grad_norm": 0.4436798095703125, "learning_rate": 9.123867099988322e-05, "loss": 0.0448, "step": 18540 }, { "grad_norm": 0.3686646819114685, "learning_rate": 9.122698026955769e-05, "loss": 0.0443, "step": 18550 }, { "grad_norm": 0.3444031774997711, "learning_rate": 9.12152824946688e-05, "loss": 0.0635, "step": 18560 }, { "grad_norm": 0.5500444769859314, "learning_rate": 9.120357767721538e-05, "loss": 0.051, "step": 18570 }, { "grad_norm": 0.5415595769882202, "learning_rate": 9.119186581919745e-05, "loss": 0.0473, "step": 18580 }, { "grad_norm": 0.5652108788490295, "learning_rate": 9.118014692261624e-05, "loss": 0.0456, "step": 18590 }, { "grad_norm": 0.6486071348190308, "learning_rate": 9.116842098947422e-05, "loss": 0.0667, "step": 18600 }, { "grad_norm": 0.38966795802116394, "learning_rate": 9.115668802177499e-05, "loss": 0.0503, "step": 18610 }, { "grad_norm": 0.3926853537559509, "learning_rate": 9.114494802152342e-05, "loss": 0.0567, "step": 18620 }, { "grad_norm": 0.6575027704238892, "learning_rate": 9.113320099072555e-05, "loss": 0.0787, "step": 18630 }, { "grad_norm": 0.41182297468185425, "learning_rate": 9.112144693138864e-05, "loss": 0.0524, "step": 18640 }, { "grad_norm": 0.5455276370048523, "learning_rate": 9.110968584552111e-05, "loss": 0.0494, "step": 18650 }, { "grad_norm": 0.37656816840171814, "learning_rate": 9.109791773513264e-05, "loss": 0.0505, "step": 18660 }, { "grad_norm": 0.41333433985710144, "learning_rate": 9.108614260223403e-05, "loss": 0.0527, "step": 18670 }, { "grad_norm": 0.4897536337375641, "learning_rate": 9.107436044883738e-05, "loss": 0.051, "step": 18680 }, { "grad_norm": 0.511650025844574, "learning_rate": 9.10625712769559e-05, "loss": 0.0502, "step": 18690 }, { "grad_norm": 0.42007291316986084, "learning_rate": 9.105077508860406e-05, "loss": 0.0459, "step": 18700 }, { "grad_norm": 0.4174588918685913, "learning_rate": 9.103897188579751e-05, "loss": 0.0469, "step": 18710 }, { "grad_norm": 0.3557695150375366, "learning_rate": 9.102716167055308e-05, "loss": 0.0362, "step": 18720 }, { "grad_norm": 0.405350923538208, "learning_rate": 9.10153444448888e-05, "loss": 0.076, "step": 18730 }, { "grad_norm": 0.36841344833374023, "learning_rate": 9.100352021082393e-05, "loss": 0.0401, "step": 18740 }, { "grad_norm": 0.351815789937973, "learning_rate": 9.099168897037891e-05, "loss": 0.0438, "step": 18750 }, { "grad_norm": 0.36858251690864563, "learning_rate": 9.097985072557538e-05, "loss": 0.0428, "step": 18760 }, { "grad_norm": 0.5094027519226074, "learning_rate": 9.096800547843615e-05, "loss": 0.0687, "step": 18770 }, { "grad_norm": 0.5145494341850281, "learning_rate": 9.095615323098526e-05, "loss": 0.0561, "step": 18780 }, { "grad_norm": 0.4538804590702057, "learning_rate": 9.094429398524795e-05, "loss": 0.0431, "step": 18790 }, { "grad_norm": 0.3993886411190033, "learning_rate": 9.093242774325061e-05, "loss": 0.0492, "step": 18800 }, { "grad_norm": 0.3377540409564972, "learning_rate": 9.092055450702088e-05, "loss": 0.0449, "step": 18810 }, { "grad_norm": 0.4279111325740814, "learning_rate": 9.090867427858756e-05, "loss": 0.0511, "step": 18820 }, { "grad_norm": 0.4717673659324646, "learning_rate": 9.089678705998066e-05, "loss": 0.049, "step": 18830 }, { "grad_norm": 0.3518950939178467, "learning_rate": 9.088489285323139e-05, "loss": 0.043, "step": 18840 }, { "grad_norm": 0.37072187662124634, "learning_rate": 9.087299166037212e-05, "loss": 0.0458, "step": 18850 }, { "grad_norm": 0.3438737988471985, "learning_rate": 9.086108348343647e-05, "loss": 0.0494, "step": 18860 }, { "grad_norm": 0.8359073400497437, "learning_rate": 9.08491683244592e-05, "loss": 0.056, "step": 18870 }, { "grad_norm": 0.4355098307132721, "learning_rate": 9.08372461854763e-05, "loss": 0.0408, "step": 18880 }, { "grad_norm": 0.5148125290870667, "learning_rate": 9.082531706852492e-05, "loss": 0.0507, "step": 18890 }, { "grad_norm": 0.3542616665363312, "learning_rate": 9.081338097564342e-05, "loss": 0.057, "step": 18900 }, { "grad_norm": 0.34908294677734375, "learning_rate": 9.080143790887137e-05, "loss": 0.043, "step": 18910 }, { "grad_norm": 0.5832956433296204, "learning_rate": 9.07894878702495e-05, "loss": 0.0441, "step": 18920 }, { "grad_norm": 1.2119907140731812, "learning_rate": 9.077753086181974e-05, "loss": 0.0599, "step": 18930 }, { "grad_norm": 0.5597949624061584, "learning_rate": 9.076556688562524e-05, "loss": 0.0478, "step": 18940 }, { "grad_norm": 0.3556927740573883, "learning_rate": 9.075359594371029e-05, "loss": 0.0419, "step": 18950 }, { "grad_norm": 0.3539232313632965, "learning_rate": 9.07416180381204e-05, "loss": 0.0507, "step": 18960 }, { "grad_norm": 0.36127594113349915, "learning_rate": 9.072963317090228e-05, "loss": 0.0473, "step": 18970 }, { "grad_norm": 0.519733726978302, "learning_rate": 9.071764134410382e-05, "loss": 0.0461, "step": 18980 }, { "grad_norm": 0.4762538969516754, "learning_rate": 9.070564255977407e-05, "loss": 0.0423, "step": 18990 }, { "grad_norm": 0.4297885000705719, "learning_rate": 9.06936368199633e-05, "loss": 0.0469, "step": 19000 }, { "grad_norm": 0.41768378019332886, "learning_rate": 9.0681624126723e-05, "loss": 0.0504, "step": 19010 }, { "grad_norm": 0.4807989299297333, "learning_rate": 9.066960448210576e-05, "loss": 0.0522, "step": 19020 }, { "grad_norm": 0.3274226188659668, "learning_rate": 9.065757788816543e-05, "loss": 0.038, "step": 19030 }, { "grad_norm": 0.4697139263153076, "learning_rate": 9.064554434695705e-05, "loss": 0.0522, "step": 19040 }, { "grad_norm": 0.44769179821014404, "learning_rate": 9.063350386053677e-05, "loss": 0.0487, "step": 19050 }, { "grad_norm": 0.3978944718837738, "learning_rate": 9.062145643096202e-05, "loss": 0.0472, "step": 19060 }, { "grad_norm": 0.4247998595237732, "learning_rate": 9.060940206029136e-05, "loss": 0.0538, "step": 19070 }, { "grad_norm": 0.389443039894104, "learning_rate": 9.059734075058457e-05, "loss": 0.0391, "step": 19080 }, { "grad_norm": 0.5417117476463318, "learning_rate": 9.058527250390257e-05, "loss": 0.0602, "step": 19090 }, { "grad_norm": 0.3712310194969177, "learning_rate": 9.057319732230752e-05, "loss": 0.0546, "step": 19100 }, { "grad_norm": 0.27526646852493286, "learning_rate": 9.056111520786273e-05, "loss": 0.0487, "step": 19110 }, { "grad_norm": 0.3520198464393616, "learning_rate": 9.054902616263268e-05, "loss": 0.0635, "step": 19120 }, { "grad_norm": 0.41551241278648376, "learning_rate": 9.05369301886831e-05, "loss": 0.0627, "step": 19130 }, { "grad_norm": 0.4239307940006256, "learning_rate": 9.052482728808083e-05, "loss": 0.053, "step": 19140 }, { "grad_norm": 0.4171057641506195, "learning_rate": 9.051271746289391e-05, "loss": 0.0458, "step": 19150 }, { "grad_norm": 0.4876154959201813, "learning_rate": 9.050060071519162e-05, "loss": 0.0525, "step": 19160 }, { "grad_norm": 0.5066834688186646, "learning_rate": 9.048847704704437e-05, "loss": 0.0483, "step": 19170 }, { "grad_norm": 0.3518519401550293, "learning_rate": 9.047634646052376e-05, "loss": 0.0538, "step": 19180 }, { "grad_norm": 0.5359349846839905, "learning_rate": 9.046420895770256e-05, "loss": 0.0607, "step": 19190 }, { "grad_norm": 0.46476808190345764, "learning_rate": 9.045206454065473e-05, "loss": 0.047, "step": 19200 }, { "grad_norm": 0.34127089381217957, "learning_rate": 9.043991321145546e-05, "loss": 0.0483, "step": 19210 }, { "grad_norm": 0.414694219827652, "learning_rate": 9.042775497218105e-05, "loss": 0.0394, "step": 19220 }, { "grad_norm": 0.4185863137245178, "learning_rate": 9.041558982490901e-05, "loss": 0.0451, "step": 19230 }, { "grad_norm": 0.6040807366371155, "learning_rate": 9.040341777171805e-05, "loss": 0.041, "step": 19240 }, { "grad_norm": 0.407202810049057, "learning_rate": 9.039123881468802e-05, "loss": 0.0446, "step": 19250 }, { "grad_norm": 0.31718191504478455, "learning_rate": 9.037905295589998e-05, "loss": 0.0402, "step": 19260 }, { "grad_norm": 0.444353848695755, "learning_rate": 9.036686019743617e-05, "loss": 0.0335, "step": 19270 }, { "grad_norm": 0.36236801743507385, "learning_rate": 9.035466054137997e-05, "loss": 0.0433, "step": 19280 }, { "grad_norm": 0.44010230898857117, "learning_rate": 9.0342453989816e-05, "loss": 0.0545, "step": 19290 }, { "grad_norm": 0.3833494782447815, "learning_rate": 9.033024054483e-05, "loss": 0.045, "step": 19300 }, { "grad_norm": 0.44015562534332275, "learning_rate": 9.031802020850894e-05, "loss": 0.0471, "step": 19310 }, { "grad_norm": 0.40508174896240234, "learning_rate": 9.030579298294092e-05, "loss": 0.0573, "step": 19320 }, { "grad_norm": 0.48547810316085815, "learning_rate": 9.029355887021524e-05, "loss": 0.0552, "step": 19330 }, { "grad_norm": 0.5681449174880981, "learning_rate": 9.028131787242238e-05, "loss": 0.0676, "step": 19340 }, { "grad_norm": 0.4897367060184479, "learning_rate": 9.026906999165399e-05, "loss": 0.0559, "step": 19350 }, { "grad_norm": 0.3675989806652069, "learning_rate": 9.025681523000291e-05, "loss": 0.0398, "step": 19360 }, { "grad_norm": 0.4475105404853821, "learning_rate": 9.024455358956315e-05, "loss": 0.0405, "step": 19370 }, { "grad_norm": 0.5167994499206543, "learning_rate": 9.023228507242984e-05, "loss": 0.0539, "step": 19380 }, { "grad_norm": 0.373487263917923, "learning_rate": 9.022000968069937e-05, "loss": 0.0509, "step": 19390 }, { "grad_norm": 0.39619702100753784, "learning_rate": 9.020772741646928e-05, "loss": 0.0444, "step": 19400 }, { "grad_norm": 0.3840530812740326, "learning_rate": 9.019543828183826e-05, "loss": 0.0469, "step": 19410 }, { "grad_norm": 0.45032304525375366, "learning_rate": 9.018314227890616e-05, "loss": 0.039, "step": 19420 }, { "grad_norm": 0.4764549136161804, "learning_rate": 9.017083940977408e-05, "loss": 0.0596, "step": 19430 }, { "grad_norm": 0.5466468930244446, "learning_rate": 9.015852967654422e-05, "loss": 0.0393, "step": 19440 }, { "grad_norm": 0.4737316370010376, "learning_rate": 9.014621308131996e-05, "loss": 0.0484, "step": 19450 }, { "grad_norm": 0.44148001074790955, "learning_rate": 9.01338896262059e-05, "loss": 0.0413, "step": 19460 }, { "grad_norm": 0.3502601981163025, "learning_rate": 9.012155931330777e-05, "loss": 0.0542, "step": 19470 }, { "grad_norm": 0.4198257327079773, "learning_rate": 9.010922214473246e-05, "loss": 0.0428, "step": 19480 }, { "grad_norm": 0.49476537108421326, "learning_rate": 9.009687812258808e-05, "loss": 0.0565, "step": 19490 }, { "grad_norm": 0.5293270349502563, "learning_rate": 9.00845272489839e-05, "loss": 0.0465, "step": 19500 }, { "grad_norm": 0.483898401260376, "learning_rate": 9.007216952603031e-05, "loss": 0.0576, "step": 19510 }, { "grad_norm": 0.43279698491096497, "learning_rate": 9.005980495583894e-05, "loss": 0.044, "step": 19520 }, { "grad_norm": 0.4003146290779114, "learning_rate": 9.004743354052252e-05, "loss": 0.0425, "step": 19530 }, { "grad_norm": 0.5175650715827942, "learning_rate": 9.003505528219503e-05, "loss": 0.047, "step": 19540 }, { "grad_norm": 0.41844749450683594, "learning_rate": 9.002267018297154e-05, "loss": 0.0555, "step": 19550 }, { "grad_norm": 0.3570314645767212, "learning_rate": 9.001027824496834e-05, "loss": 0.059, "step": 19560 }, { "grad_norm": 0.3544917702674866, "learning_rate": 8.999787947030287e-05, "loss": 0.0458, "step": 19570 }, { "grad_norm": 0.3649086654186249, "learning_rate": 8.998547386109376e-05, "loss": 0.0565, "step": 19580 }, { "grad_norm": 0.4696737825870514, "learning_rate": 8.997306141946073e-05, "loss": 0.0444, "step": 19590 }, { "grad_norm": 0.500440239906311, "learning_rate": 8.996064214752481e-05, "loss": 0.0513, "step": 19600 }, { "grad_norm": 0.44714847207069397, "learning_rate": 8.994821604740806e-05, "loss": 0.0482, "step": 19610 }, { "grad_norm": 0.36449092626571655, "learning_rate": 8.993578312123377e-05, "loss": 0.0467, "step": 19620 }, { "grad_norm": 0.4062497913837433, "learning_rate": 8.992334337112639e-05, "loss": 0.0475, "step": 19630 }, { "grad_norm": 0.28016215562820435, "learning_rate": 8.991089679921154e-05, "loss": 0.0384, "step": 19640 }, { "grad_norm": 0.40734273195266724, "learning_rate": 8.989844340761599e-05, "loss": 0.0408, "step": 19650 }, { "grad_norm": 0.41797736287117004, "learning_rate": 8.988598319846768e-05, "loss": 0.044, "step": 19660 }, { "grad_norm": 0.4262579381465912, "learning_rate": 8.987351617389574e-05, "loss": 0.0516, "step": 19670 }, { "grad_norm": 0.5525339245796204, "learning_rate": 8.98610423360304e-05, "loss": 0.0456, "step": 19680 }, { "grad_norm": 0.26115792989730835, "learning_rate": 8.984856168700317e-05, "loss": 0.053, "step": 19690 }, { "grad_norm": 0.34171801805496216, "learning_rate": 8.983607422894658e-05, "loss": 0.0476, "step": 19700 }, { "grad_norm": 0.3503652811050415, "learning_rate": 8.982357996399442e-05, "loss": 0.0538, "step": 19710 }, { "grad_norm": 0.32174813747406006, "learning_rate": 8.981107889428164e-05, "loss": 0.0443, "step": 19720 }, { "grad_norm": 0.48306217789649963, "learning_rate": 8.979857102194428e-05, "loss": 0.0477, "step": 19730 }, { "grad_norm": 0.5729153752326965, "learning_rate": 8.978605634911968e-05, "loss": 0.0479, "step": 19740 }, { "grad_norm": 0.3779255449771881, "learning_rate": 8.977353487794616e-05, "loss": 0.0366, "step": 19750 }, { "grad_norm": 0.3241032361984253, "learning_rate": 8.976100661056334e-05, "loss": 0.0529, "step": 19760 }, { "grad_norm": 0.44802984595298767, "learning_rate": 8.974847154911197e-05, "loss": 0.0404, "step": 19770 }, { "grad_norm": 0.3740008771419525, "learning_rate": 8.973592969573393e-05, "loss": 0.0526, "step": 19780 }, { "grad_norm": 0.49970829486846924, "learning_rate": 8.972338105257228e-05, "loss": 0.0508, "step": 19790 }, { "grad_norm": 0.3409971594810486, "learning_rate": 8.971082562177125e-05, "loss": 0.0429, "step": 19800 }, { "grad_norm": 0.4135831594467163, "learning_rate": 8.96982634054762e-05, "loss": 0.0446, "step": 19810 }, { "grad_norm": 0.654991865158081, "learning_rate": 8.96856944058337e-05, "loss": 0.0563, "step": 19820 }, { "grad_norm": 0.4241202473640442, "learning_rate": 8.967311862499144e-05, "loss": 0.0404, "step": 19830 }, { "grad_norm": 0.8752793073654175, "learning_rate": 8.966053606509825e-05, "loss": 0.0506, "step": 19840 }, { "grad_norm": 0.6595100164413452, "learning_rate": 8.964794672830417e-05, "loss": 0.067, "step": 19850 }, { "grad_norm": 0.5573925971984863, "learning_rate": 8.963535061676038e-05, "loss": 0.0603, "step": 19860 }, { "grad_norm": 0.46343791484832764, "learning_rate": 8.962274773261918e-05, "loss": 0.0509, "step": 19870 }, { "grad_norm": 0.341840922832489, "learning_rate": 8.961013807803409e-05, "loss": 0.0576, "step": 19880 }, { "grad_norm": 0.4414932429790497, "learning_rate": 8.959752165515973e-05, "loss": 0.0526, "step": 19890 }, { "grad_norm": 0.6550657749176025, "learning_rate": 8.958489846615193e-05, "loss": 0.0702, "step": 19900 }, { "grad_norm": 0.5611836910247803, "learning_rate": 8.957226851316762e-05, "loss": 0.0572, "step": 19910 }, { "grad_norm": 0.46033918857574463, "learning_rate": 8.955963179836493e-05, "loss": 0.0505, "step": 19920 }, { "grad_norm": 0.35551145672798157, "learning_rate": 8.954698832390312e-05, "loss": 0.0505, "step": 19930 }, { "grad_norm": 0.47157183289527893, "learning_rate": 8.953433809194263e-05, "loss": 0.0481, "step": 19940 }, { "grad_norm": 0.42084571719169617, "learning_rate": 8.9521681104645e-05, "loss": 0.0467, "step": 19950 }, { "grad_norm": 0.5623788237571716, "learning_rate": 8.9509017364173e-05, "loss": 0.0561, "step": 19960 }, { "grad_norm": 0.39897918701171875, "learning_rate": 8.949634687269052e-05, "loss": 0.0448, "step": 19970 }, { "grad_norm": 0.503510057926178, "learning_rate": 8.948366963236259e-05, "loss": 0.0459, "step": 19980 }, { "grad_norm": 0.6922836899757385, "learning_rate": 8.947098564535538e-05, "loss": 0.0519, "step": 19990 }, { "grad_norm": 0.5520749092102051, "learning_rate": 8.945829491383627e-05, "loss": 0.0457, "step": 20000 }, { "grad_norm": 0.29852357506752014, "learning_rate": 8.944559743997374e-05, "loss": 0.044, "step": 20010 }, { "grad_norm": 0.6737194061279297, "learning_rate": 8.943289322593746e-05, "loss": 0.0467, "step": 20020 }, { "grad_norm": 0.39570537209510803, "learning_rate": 8.942018227389821e-05, "loss": 0.0389, "step": 20030 }, { "grad_norm": 0.5275892019271851, "learning_rate": 8.940746458602795e-05, "loss": 0.0555, "step": 20040 }, { "grad_norm": 0.5714529752731323, "learning_rate": 8.939474016449979e-05, "loss": 0.049, "step": 20050 }, { "grad_norm": 0.5710515379905701, "learning_rate": 8.938200901148799e-05, "loss": 0.0364, "step": 20060 }, { "grad_norm": 0.5907671451568604, "learning_rate": 8.936927112916795e-05, "loss": 0.0484, "step": 20070 }, { "grad_norm": 0.409584105014801, "learning_rate": 8.935652651971622e-05, "loss": 0.0468, "step": 20080 }, { "grad_norm": 0.4306665062904358, "learning_rate": 8.934377518531052e-05, "loss": 0.052, "step": 20090 }, { "grad_norm": 0.579693615436554, "learning_rate": 8.933101712812967e-05, "loss": 0.0596, "step": 20100 }, { "grad_norm": 0.3558349609375, "learning_rate": 8.931825235035374e-05, "loss": 0.0424, "step": 20110 }, { "grad_norm": 0.8314046263694763, "learning_rate": 8.930548085416382e-05, "loss": 0.0474, "step": 20120 }, { "grad_norm": 0.3706062436103821, "learning_rate": 8.92927026417422e-05, "loss": 0.0479, "step": 20130 }, { "grad_norm": 0.5639284253120422, "learning_rate": 8.92799177152724e-05, "loss": 0.0465, "step": 20140 }, { "grad_norm": 0.31074875593185425, "learning_rate": 8.926712607693895e-05, "loss": 0.0446, "step": 20150 }, { "grad_norm": 0.3933444023132324, "learning_rate": 8.925432772892762e-05, "loss": 0.0528, "step": 20160 }, { "grad_norm": 0.4048672318458557, "learning_rate": 8.924152267342529e-05, "loss": 0.056, "step": 20170 }, { "grad_norm": 0.5905811190605164, "learning_rate": 8.922871091261998e-05, "loss": 0.0564, "step": 20180 }, { "grad_norm": 0.388610303401947, "learning_rate": 8.92158924487009e-05, "loss": 0.047, "step": 20190 }, { "grad_norm": 0.558713436126709, "learning_rate": 8.920306728385834e-05, "loss": 0.0423, "step": 20200 }, { "grad_norm": 0.35191136598587036, "learning_rate": 8.919023542028379e-05, "loss": 0.0391, "step": 20210 }, { "grad_norm": 0.4939667880535126, "learning_rate": 8.917739686016988e-05, "loss": 0.0481, "step": 20220 }, { "grad_norm": 0.3642098307609558, "learning_rate": 8.916455160571033e-05, "loss": 0.055, "step": 20230 }, { "grad_norm": 0.3861846625804901, "learning_rate": 8.915169965910008e-05, "loss": 0.037, "step": 20240 }, { "grad_norm": 0.5093153715133667, "learning_rate": 8.913884102253514e-05, "loss": 0.0444, "step": 20250 }, { "grad_norm": 0.624779999256134, "learning_rate": 8.912597569821273e-05, "loss": 0.0381, "step": 20260 }, { "grad_norm": 0.3428029417991638, "learning_rate": 8.911310368833118e-05, "loss": 0.043, "step": 20270 }, { "grad_norm": 0.4260753393173218, "learning_rate": 8.910022499508994e-05, "loss": 0.0389, "step": 20280 }, { "grad_norm": 0.40917497873306274, "learning_rate": 8.908733962068965e-05, "loss": 0.0485, "step": 20290 }, { "grad_norm": 0.5050711631774902, "learning_rate": 8.907444756733207e-05, "loss": 0.0431, "step": 20300 }, { "grad_norm": 0.5744189620018005, "learning_rate": 8.906154883722006e-05, "loss": 0.0487, "step": 20310 }, { "grad_norm": 0.5430971384048462, "learning_rate": 8.904864343255773e-05, "loss": 0.0471, "step": 20320 }, { "grad_norm": 0.39280831813812256, "learning_rate": 8.90357313555502e-05, "loss": 0.0436, "step": 20330 }, { "grad_norm": 0.34743058681488037, "learning_rate": 8.90228126084038e-05, "loss": 0.0512, "step": 20340 }, { "grad_norm": 0.5362100601196289, "learning_rate": 8.900988719332601e-05, "loss": 0.0518, "step": 20350 }, { "grad_norm": 0.5132984519004822, "learning_rate": 8.899695511252542e-05, "loss": 0.0486, "step": 20360 }, { "grad_norm": 0.43222710490226746, "learning_rate": 8.898401636821176e-05, "loss": 0.0554, "step": 20370 }, { "grad_norm": 0.3748404383659363, "learning_rate": 8.897107096259593e-05, "loss": 0.0528, "step": 20380 }, { "grad_norm": 0.4121312201023102, "learning_rate": 8.895811889788994e-05, "loss": 0.0411, "step": 20390 }, { "grad_norm": 0.25994449853897095, "learning_rate": 8.894516017630692e-05, "loss": 0.0361, "step": 20400 }, { "grad_norm": 0.5944382548332214, "learning_rate": 8.893219480006118e-05, "loss": 0.0387, "step": 20410 }, { "grad_norm": 0.4037998616695404, "learning_rate": 8.891922277136817e-05, "loss": 0.0405, "step": 20420 }, { "grad_norm": 0.41794365644454956, "learning_rate": 8.890624409244441e-05, "loss": 0.039, "step": 20430 }, { "grad_norm": 0.4325920343399048, "learning_rate": 8.889325876550763e-05, "loss": 0.0473, "step": 20440 }, { "grad_norm": 0.3741753399372101, "learning_rate": 8.888026679277666e-05, "loss": 0.0467, "step": 20450 }, { "grad_norm": 0.32312700152397156, "learning_rate": 8.886726817647147e-05, "loss": 0.0635, "step": 20460 }, { "grad_norm": 0.5315676331520081, "learning_rate": 8.885426291881319e-05, "loss": 0.061, "step": 20470 }, { "grad_norm": 0.45704400539398193, "learning_rate": 8.884125102202401e-05, "loss": 0.0487, "step": 20480 }, { "grad_norm": 0.42350050806999207, "learning_rate": 8.882823248832736e-05, "loss": 0.0367, "step": 20490 }, { "grad_norm": 0.3679253160953522, "learning_rate": 8.881520731994772e-05, "loss": 0.0434, "step": 20500 }, { "grad_norm": 0.4091085195541382, "learning_rate": 8.880217551911077e-05, "loss": 0.0482, "step": 20510 }, { "grad_norm": 0.4666108191013336, "learning_rate": 8.878913708804323e-05, "loss": 0.0417, "step": 20520 }, { "grad_norm": 0.398070752620697, "learning_rate": 8.877609202897308e-05, "loss": 0.052, "step": 20530 }, { "grad_norm": 0.42559683322906494, "learning_rate": 8.876304034412933e-05, "loss": 0.0388, "step": 20540 }, { "grad_norm": 0.4264746904373169, "learning_rate": 8.874998203574214e-05, "loss": 0.0466, "step": 20550 }, { "grad_norm": 0.4258725345134735, "learning_rate": 8.873691710604284e-05, "loss": 0.039, "step": 20560 }, { "grad_norm": 0.4199366569519043, "learning_rate": 8.872384555726387e-05, "loss": 0.04, "step": 20570 }, { "grad_norm": 0.43731844425201416, "learning_rate": 8.871076739163878e-05, "loss": 0.0427, "step": 20580 }, { "grad_norm": 0.3016221821308136, "learning_rate": 8.86976826114023e-05, "loss": 0.0464, "step": 20590 }, { "grad_norm": 0.42001625895500183, "learning_rate": 8.868459121879023e-05, "loss": 0.0433, "step": 20600 }, { "grad_norm": 0.3688039779663086, "learning_rate": 8.867149321603956e-05, "loss": 0.0557, "step": 20610 }, { "grad_norm": 0.37933969497680664, "learning_rate": 8.865838860538835e-05, "loss": 0.0424, "step": 20620 }, { "grad_norm": 0.42510855197906494, "learning_rate": 8.864527738907585e-05, "loss": 0.0512, "step": 20630 }, { "grad_norm": 0.4819476008415222, "learning_rate": 8.863215956934239e-05, "loss": 0.0485, "step": 20640 }, { "grad_norm": 0.5409268736839294, "learning_rate": 8.861903514842947e-05, "loss": 0.0378, "step": 20650 }, { "grad_norm": 0.35407230257987976, "learning_rate": 8.860590412857966e-05, "loss": 0.041, "step": 20660 }, { "grad_norm": 0.31389927864074707, "learning_rate": 8.85927665120367e-05, "loss": 0.0374, "step": 20670 }, { "grad_norm": 0.4168339669704437, "learning_rate": 8.857962230104546e-05, "loss": 0.0405, "step": 20680 }, { "grad_norm": 0.4264257252216339, "learning_rate": 8.856647149785193e-05, "loss": 0.0414, "step": 20690 }, { "grad_norm": 0.4103916883468628, "learning_rate": 8.855331410470322e-05, "loss": 0.046, "step": 20700 }, { "grad_norm": 0.4178889989852905, "learning_rate": 8.854015012384756e-05, "loss": 0.0428, "step": 20710 }, { "grad_norm": 0.37555378675460815, "learning_rate": 8.852697955753433e-05, "loss": 0.0497, "step": 20720 }, { "grad_norm": 0.4109729528427124, "learning_rate": 8.851380240801399e-05, "loss": 0.0467, "step": 20730 }, { "grad_norm": 0.3595787286758423, "learning_rate": 8.850061867753818e-05, "loss": 0.034, "step": 20740 }, { "grad_norm": 0.5242611765861511, "learning_rate": 8.848742836835963e-05, "loss": 0.0498, "step": 20750 }, { "grad_norm": 0.5097054243087769, "learning_rate": 8.847423148273221e-05, "loss": 0.0492, "step": 20760 }, { "grad_norm": 0.4384428858757019, "learning_rate": 8.846102802291092e-05, "loss": 0.0493, "step": 20770 }, { "grad_norm": 0.4153221845626831, "learning_rate": 8.844781799115183e-05, "loss": 0.047, "step": 20780 }, { "grad_norm": 0.5138063430786133, "learning_rate": 8.84346013897122e-05, "loss": 0.0362, "step": 20790 }, { "grad_norm": 0.38757938146591187, "learning_rate": 8.842137822085038e-05, "loss": 0.0499, "step": 20800 }, { "grad_norm": 0.3771473169326782, "learning_rate": 8.840814848682585e-05, "loss": 0.0405, "step": 20810 }, { "grad_norm": 0.39955657720565796, "learning_rate": 8.83949121898992e-05, "loss": 0.054, "step": 20820 }, { "grad_norm": 0.30099329352378845, "learning_rate": 8.838166933233217e-05, "loss": 0.0391, "step": 20830 }, { "grad_norm": 0.3940200209617615, "learning_rate": 8.83684199163876e-05, "loss": 0.0513, "step": 20840 }, { "grad_norm": 0.34857988357543945, "learning_rate": 8.835516394432943e-05, "loss": 0.0418, "step": 20850 }, { "grad_norm": 0.5164331197738647, "learning_rate": 8.834190141842276e-05, "loss": 0.0536, "step": 20860 }, { "grad_norm": 0.37890616059303284, "learning_rate": 8.83286323409338e-05, "loss": 0.0443, "step": 20870 }, { "grad_norm": 0.3914577066898346, "learning_rate": 8.831535671412986e-05, "loss": 0.0454, "step": 20880 }, { "grad_norm": 0.32176634669303894, "learning_rate": 8.830207454027938e-05, "loss": 0.0437, "step": 20890 }, { "grad_norm": 0.5109374523162842, "learning_rate": 8.828878582165192e-05, "loss": 0.0431, "step": 20900 }, { "grad_norm": 0.4311383366584778, "learning_rate": 8.827549056051818e-05, "loss": 0.0417, "step": 20910 }, { "grad_norm": 0.31568947434425354, "learning_rate": 8.826218875914993e-05, "loss": 0.0404, "step": 20920 }, { "grad_norm": 0.3292481005191803, "learning_rate": 8.82488804198201e-05, "loss": 0.0434, "step": 20930 }, { "grad_norm": 0.3011327087879181, "learning_rate": 8.82355655448027e-05, "loss": 0.0398, "step": 20940 }, { "grad_norm": 0.3994313180446625, "learning_rate": 8.822224413637293e-05, "loss": 0.0431, "step": 20950 }, { "grad_norm": 0.3536672592163086, "learning_rate": 8.820891619680697e-05, "loss": 0.0475, "step": 20960 }, { "grad_norm": 0.934705376625061, "learning_rate": 8.819558172838227e-05, "loss": 0.0537, "step": 20970 }, { "grad_norm": 0.417129248380661, "learning_rate": 8.818224073337731e-05, "loss": 0.0382, "step": 20980 }, { "grad_norm": 0.48901885747909546, "learning_rate": 8.816889321407169e-05, "loss": 0.0424, "step": 20990 }, { "grad_norm": 0.3733789324760437, "learning_rate": 8.815553917274615e-05, "loss": 0.0454, "step": 21000 }, { "grad_norm": 0.3655532896518707, "learning_rate": 8.81421786116825e-05, "loss": 0.0557, "step": 21010 }, { "grad_norm": 0.3808937072753906, "learning_rate": 8.812881153316373e-05, "loss": 0.0358, "step": 21020 }, { "grad_norm": 0.6481075882911682, "learning_rate": 8.81154379394739e-05, "loss": 0.0493, "step": 21030 }, { "grad_norm": 0.40756309032440186, "learning_rate": 8.810205783289818e-05, "loss": 0.0383, "step": 21040 }, { "grad_norm": 0.3835257589817047, "learning_rate": 8.808867121572286e-05, "loss": 0.0459, "step": 21050 }, { "grad_norm": 0.30166518688201904, "learning_rate": 8.807527809023537e-05, "loss": 0.0493, "step": 21060 }, { "grad_norm": 0.2732974588871002, "learning_rate": 8.80618784587242e-05, "loss": 0.0468, "step": 21070 }, { "grad_norm": 0.4826183021068573, "learning_rate": 8.804847232347902e-05, "loss": 0.0431, "step": 21080 }, { "grad_norm": 0.8503368496894836, "learning_rate": 8.803505968679054e-05, "loss": 0.055, "step": 21090 }, { "grad_norm": 0.4431348741054535, "learning_rate": 8.802164055095061e-05, "loss": 0.0528, "step": 21100 }, { "grad_norm": 0.6033118963241577, "learning_rate": 8.80082149182522e-05, "loss": 0.0525, "step": 21110 }, { "grad_norm": 0.4687495827674866, "learning_rate": 8.79947827909894e-05, "loss": 0.0379, "step": 21120 }, { "grad_norm": 0.5920835137367249, "learning_rate": 8.798134417145738e-05, "loss": 0.043, "step": 21130 }, { "grad_norm": 0.3793583810329437, "learning_rate": 8.796789906195243e-05, "loss": 0.0401, "step": 21140 }, { "grad_norm": 0.5052685141563416, "learning_rate": 8.795444746477195e-05, "loss": 0.0371, "step": 21150 }, { "grad_norm": 0.5123188495635986, "learning_rate": 8.794098938221446e-05, "loss": 0.0496, "step": 21160 }, { "grad_norm": 0.34983935952186584, "learning_rate": 8.792752481657957e-05, "loss": 0.0465, "step": 21170 }, { "grad_norm": 0.471817284822464, "learning_rate": 8.791405377016802e-05, "loss": 0.0398, "step": 21180 }, { "grad_norm": 0.4429417848587036, "learning_rate": 8.790057624528163e-05, "loss": 0.0443, "step": 21190 }, { "grad_norm": 0.5061697363853455, "learning_rate": 8.788709224422333e-05, "loss": 0.0451, "step": 21200 }, { "grad_norm": 0.39716437458992004, "learning_rate": 8.787360176929717e-05, "loss": 0.044, "step": 21210 }, { "grad_norm": 0.7377652525901794, "learning_rate": 8.786010482280834e-05, "loss": 0.0472, "step": 21220 }, { "grad_norm": 0.3471709191799164, "learning_rate": 8.784660140706306e-05, "loss": 0.0461, "step": 21230 }, { "grad_norm": 0.3113759160041809, "learning_rate": 8.783309152436872e-05, "loss": 0.038, "step": 21240 }, { "grad_norm": 0.3813178539276123, "learning_rate": 8.781957517703375e-05, "loss": 0.0436, "step": 21250 }, { "grad_norm": 0.45921197533607483, "learning_rate": 8.780605236736776e-05, "loss": 0.0447, "step": 21260 }, { "grad_norm": 0.3268726170063019, "learning_rate": 8.779252309768142e-05, "loss": 0.045, "step": 21270 }, { "grad_norm": 0.3832651376724243, "learning_rate": 8.777898737028652e-05, "loss": 0.0499, "step": 21280 }, { "grad_norm": 0.4508974552154541, "learning_rate": 8.776544518749591e-05, "loss": 0.0386, "step": 21290 }, { "grad_norm": 0.3147774934768677, "learning_rate": 8.775189655162364e-05, "loss": 0.0463, "step": 21300 }, { "grad_norm": 0.40919801592826843, "learning_rate": 8.773834146498474e-05, "loss": 0.044, "step": 21310 }, { "grad_norm": 0.32637542486190796, "learning_rate": 8.772477992989545e-05, "loss": 0.0401, "step": 21320 }, { "grad_norm": 0.33396267890930176, "learning_rate": 8.771121194867304e-05, "loss": 0.057, "step": 21330 }, { "grad_norm": 0.32963165640830994, "learning_rate": 8.769763752363589e-05, "loss": 0.0392, "step": 21340 }, { "grad_norm": 0.37469878792762756, "learning_rate": 8.768405665710352e-05, "loss": 0.0454, "step": 21350 }, { "grad_norm": 0.5446921586990356, "learning_rate": 8.767046935139655e-05, "loss": 0.0527, "step": 21360 }, { "grad_norm": 0.5191640257835388, "learning_rate": 8.765687560883666e-05, "loss": 0.0419, "step": 21370 }, { "grad_norm": 0.37605616450309753, "learning_rate": 8.764327543174664e-05, "loss": 0.0392, "step": 21380 }, { "grad_norm": 0.4214508533477783, "learning_rate": 8.762966882245038e-05, "loss": 0.0495, "step": 21390 }, { "grad_norm": 0.49437984824180603, "learning_rate": 8.761605578327291e-05, "loss": 0.0552, "step": 21400 }, { "grad_norm": 0.3973881006240845, "learning_rate": 8.76024363165403e-05, "loss": 0.0568, "step": 21410 }, { "grad_norm": 0.4171906113624573, "learning_rate": 8.758881042457976e-05, "loss": 0.0489, "step": 21420 }, { "grad_norm": 0.516806423664093, "learning_rate": 8.757517810971957e-05, "loss": 0.0458, "step": 21430 }, { "grad_norm": 0.34378674626350403, "learning_rate": 8.756153937428913e-05, "loss": 0.0436, "step": 21440 }, { "grad_norm": 0.45161381363868713, "learning_rate": 8.754789422061889e-05, "loss": 0.0569, "step": 21450 }, { "grad_norm": 0.34450966119766235, "learning_rate": 8.753424265104052e-05, "loss": 0.0395, "step": 21460 }, { "grad_norm": 0.5367270708084106, "learning_rate": 8.752058466788659e-05, "loss": 0.0579, "step": 21470 }, { "grad_norm": 0.4313832223415375, "learning_rate": 8.750692027349097e-05, "loss": 0.0395, "step": 21480 }, { "grad_norm": 0.46810534596443176, "learning_rate": 8.749324947018847e-05, "loss": 0.041, "step": 21490 }, { "grad_norm": 0.44296929240226746, "learning_rate": 8.747957226031507e-05, "loss": 0.0438, "step": 21500 }, { "grad_norm": 0.30104076862335205, "learning_rate": 8.746588864620787e-05, "loss": 0.0435, "step": 21510 }, { "grad_norm": 0.5795950889587402, "learning_rate": 8.745219863020498e-05, "loss": 0.0343, "step": 21520 }, { "grad_norm": 0.27206045389175415, "learning_rate": 8.743850221464564e-05, "loss": 0.0375, "step": 21530 }, { "grad_norm": 0.36639443039894104, "learning_rate": 8.742479940187026e-05, "loss": 0.0407, "step": 21540 }, { "grad_norm": 0.43487194180488586, "learning_rate": 8.74110901942202e-05, "loss": 0.0418, "step": 21550 }, { "grad_norm": 0.5299597382545471, "learning_rate": 8.739737459403803e-05, "loss": 0.0449, "step": 21560 }, { "grad_norm": 0.45113107562065125, "learning_rate": 8.738365260366737e-05, "loss": 0.0402, "step": 21570 }, { "grad_norm": 0.332548052072525, "learning_rate": 8.736992422545292e-05, "loss": 0.0382, "step": 21580 }, { "grad_norm": 0.42887675762176514, "learning_rate": 8.73561894617405e-05, "loss": 0.0482, "step": 21590 }, { "grad_norm": 0.36182981729507446, "learning_rate": 8.734244831487697e-05, "loss": 0.0378, "step": 21600 }, { "grad_norm": 0.3989037871360779, "learning_rate": 8.732870078721035e-05, "loss": 0.0473, "step": 21610 }, { "grad_norm": 0.3035220503807068, "learning_rate": 8.731494688108972e-05, "loss": 0.034, "step": 21620 }, { "grad_norm": 0.3547845482826233, "learning_rate": 8.730118659886523e-05, "loss": 0.042, "step": 21630 }, { "grad_norm": 0.43288156390190125, "learning_rate": 8.728741994288814e-05, "loss": 0.0394, "step": 21640 }, { "grad_norm": 0.29459065198898315, "learning_rate": 8.727364691551079e-05, "loss": 0.0373, "step": 21650 }, { "grad_norm": 0.3588293492794037, "learning_rate": 8.725986751908661e-05, "loss": 0.0396, "step": 21660 }, { "grad_norm": 0.3254915773868561, "learning_rate": 8.724608175597016e-05, "loss": 0.0355, "step": 21670 }, { "grad_norm": 0.3712998330593109, "learning_rate": 8.723228962851699e-05, "loss": 0.0348, "step": 21680 }, { "grad_norm": 0.38249412178993225, "learning_rate": 8.721849113908385e-05, "loss": 0.0386, "step": 21690 }, { "grad_norm": 0.4789455235004425, "learning_rate": 8.720468629002848e-05, "loss": 0.0406, "step": 21700 }, { "grad_norm": 0.39784204959869385, "learning_rate": 8.719087508370978e-05, "loss": 0.0366, "step": 21710 }, { "grad_norm": 0.5315825939178467, "learning_rate": 8.717705752248772e-05, "loss": 0.0474, "step": 21720 }, { "grad_norm": 0.41814714670181274, "learning_rate": 8.71632336087233e-05, "loss": 0.0306, "step": 21730 }, { "grad_norm": 0.39177337288856506, "learning_rate": 8.71494033447787e-05, "loss": 0.0423, "step": 21740 }, { "grad_norm": 0.7871300578117371, "learning_rate": 8.713556673301708e-05, "loss": 0.0403, "step": 21750 }, { "grad_norm": 0.6278435587882996, "learning_rate": 8.712172377580278e-05, "loss": 0.0485, "step": 21760 }, { "grad_norm": 0.5048931837081909, "learning_rate": 8.710787447550114e-05, "loss": 0.0486, "step": 21770 }, { "grad_norm": 0.3493000864982605, "learning_rate": 8.70940188344787e-05, "loss": 0.0546, "step": 21780 }, { "grad_norm": 0.3646092414855957, "learning_rate": 8.708015685510293e-05, "loss": 0.0437, "step": 21790 }, { "grad_norm": 0.4637819826602936, "learning_rate": 8.706628853974252e-05, "loss": 0.0436, "step": 21800 }, { "grad_norm": 0.3859504759311676, "learning_rate": 8.705241389076715e-05, "loss": 0.0418, "step": 21810 }, { "grad_norm": 0.3622349500656128, "learning_rate": 8.703853291054764e-05, "loss": 0.0447, "step": 21820 }, { "grad_norm": 0.48983144760131836, "learning_rate": 8.702464560145587e-05, "loss": 0.0404, "step": 21830 }, { "grad_norm": 0.6631761789321899, "learning_rate": 8.701075196586476e-05, "loss": 0.0464, "step": 21840 }, { "grad_norm": 0.534059464931488, "learning_rate": 8.699685200614842e-05, "loss": 0.054, "step": 21850 }, { "grad_norm": 0.40180137753486633, "learning_rate": 8.698294572468193e-05, "loss": 0.0488, "step": 21860 }, { "grad_norm": 0.37782153487205505, "learning_rate": 8.696903312384148e-05, "loss": 0.0478, "step": 21870 }, { "grad_norm": 0.36470651626586914, "learning_rate": 8.695511420600439e-05, "loss": 0.0404, "step": 21880 }, { "grad_norm": 0.34426289796829224, "learning_rate": 8.694118897354901e-05, "loss": 0.0399, "step": 21890 }, { "grad_norm": 0.4559626579284668, "learning_rate": 8.692725742885478e-05, "loss": 0.0337, "step": 21900 }, { "grad_norm": 0.4551740288734436, "learning_rate": 8.691331957430221e-05, "loss": 0.0509, "step": 21910 }, { "grad_norm": 0.4661119282245636, "learning_rate": 8.68993754122729e-05, "loss": 0.0521, "step": 21920 }, { "grad_norm": 0.5386781096458435, "learning_rate": 8.688542494514955e-05, "loss": 0.0385, "step": 21930 }, { "grad_norm": 0.40735653042793274, "learning_rate": 8.68714681753159e-05, "loss": 0.0385, "step": 21940 }, { "grad_norm": 0.4812728762626648, "learning_rate": 8.685750510515676e-05, "loss": 0.0411, "step": 21950 }, { "grad_norm": 0.4704895317554474, "learning_rate": 8.684353573705805e-05, "loss": 0.0392, "step": 21960 }, { "grad_norm": 0.398224800825119, "learning_rate": 8.682956007340677e-05, "loss": 0.0322, "step": 21970 }, { "grad_norm": 0.39841508865356445, "learning_rate": 8.681557811659095e-05, "loss": 0.0416, "step": 21980 }, { "grad_norm": 0.43351197242736816, "learning_rate": 8.680158986899974e-05, "loss": 0.0457, "step": 21990 }, { "grad_norm": 0.3305664360523224, "learning_rate": 8.678759533302335e-05, "loss": 0.0365, "step": 22000 }, { "grad_norm": 0.6016598343849182, "learning_rate": 8.677359451105308e-05, "loss": 0.0446, "step": 22010 }, { "grad_norm": 0.37287986278533936, "learning_rate": 8.675958740548123e-05, "loss": 0.0501, "step": 22020 }, { "grad_norm": 0.48030656576156616, "learning_rate": 8.674557401870129e-05, "loss": 0.0499, "step": 22030 }, { "grad_norm": 0.42210036516189575, "learning_rate": 8.673155435310775e-05, "loss": 0.04, "step": 22040 }, { "grad_norm": 0.4408465325832367, "learning_rate": 8.671752841109617e-05, "loss": 0.0384, "step": 22050 }, { "grad_norm": 0.3822059631347656, "learning_rate": 8.670349619506321e-05, "loss": 0.0365, "step": 22060 }, { "grad_norm": 0.5248544216156006, "learning_rate": 8.66894577074066e-05, "loss": 0.0384, "step": 22070 }, { "grad_norm": 0.3025086224079132, "learning_rate": 8.667541295052513e-05, "loss": 0.0375, "step": 22080 }, { "grad_norm": 0.5806949734687805, "learning_rate": 8.666136192681865e-05, "loss": 0.0544, "step": 22090 }, { "grad_norm": 0.4407007098197937, "learning_rate": 8.664730463868811e-05, "loss": 0.0465, "step": 22100 }, { "grad_norm": 0.5279036164283752, "learning_rate": 8.663324108853552e-05, "loss": 0.05, "step": 22110 }, { "grad_norm": 0.3566443622112274, "learning_rate": 8.661917127876395e-05, "loss": 0.0571, "step": 22120 }, { "grad_norm": 0.43629103899002075, "learning_rate": 8.660509521177754e-05, "loss": 0.0441, "step": 22130 }, { "grad_norm": 0.4101337194442749, "learning_rate": 8.65910128899815e-05, "loss": 0.045, "step": 22140 }, { "grad_norm": 0.40569257736206055, "learning_rate": 8.657692431578214e-05, "loss": 0.0522, "step": 22150 }, { "grad_norm": 0.38255923986434937, "learning_rate": 8.656282949158679e-05, "loss": 0.0373, "step": 22160 }, { "grad_norm": 0.2785429060459137, "learning_rate": 8.654872841980388e-05, "loss": 0.0346, "step": 22170 }, { "grad_norm": 0.3282535970211029, "learning_rate": 8.653462110284289e-05, "loss": 0.0365, "step": 22180 }, { "grad_norm": 0.3426849842071533, "learning_rate": 8.652050754311437e-05, "loss": 0.0383, "step": 22190 }, { "grad_norm": 0.40056702494621277, "learning_rate": 8.650638774302995e-05, "loss": 0.0408, "step": 22200 }, { "grad_norm": 2.69040584564209, "learning_rate": 8.649226170500233e-05, "loss": 0.0382, "step": 22210 }, { "grad_norm": 0.5419777035713196, "learning_rate": 8.647812943144524e-05, "loss": 0.0412, "step": 22220 }, { "grad_norm": 0.45407700538635254, "learning_rate": 8.646399092477351e-05, "loss": 0.0362, "step": 22230 }, { "grad_norm": 0.6002230644226074, "learning_rate": 8.644984618740301e-05, "loss": 0.054, "step": 22240 }, { "grad_norm": 0.3236853778362274, "learning_rate": 8.643569522175073e-05, "loss": 0.0348, "step": 22250 }, { "grad_norm": 0.4505665600299835, "learning_rate": 8.642153803023463e-05, "loss": 0.0443, "step": 22260 }, { "grad_norm": 0.2988266944885254, "learning_rate": 8.640737461527383e-05, "loss": 0.0337, "step": 22270 }, { "grad_norm": 0.3516892194747925, "learning_rate": 8.639320497928845e-05, "loss": 0.0351, "step": 22280 }, { "grad_norm": 0.3184296488761902, "learning_rate": 8.637902912469969e-05, "loss": 0.0352, "step": 22290 }, { "grad_norm": 0.6199243664741516, "learning_rate": 8.636484705392982e-05, "loss": 0.0374, "step": 22300 }, { "grad_norm": 0.30302703380584717, "learning_rate": 8.635065876940216e-05, "loss": 0.0378, "step": 22310 }, { "grad_norm": 0.41815677285194397, "learning_rate": 8.633646427354112e-05, "loss": 0.0393, "step": 22320 }, { "grad_norm": 0.3364010453224182, "learning_rate": 8.632226356877213e-05, "loss": 0.0379, "step": 22330 }, { "grad_norm": 0.6135392189025879, "learning_rate": 8.630805665752173e-05, "loss": 0.0468, "step": 22340 }, { "grad_norm": 0.32815980911254883, "learning_rate": 8.629384354221748e-05, "loss": 0.0466, "step": 22350 }, { "grad_norm": 0.42962953448295593, "learning_rate": 8.627962422528797e-05, "loss": 0.0373, "step": 22360 }, { "grad_norm": 0.47471845149993896, "learning_rate": 8.626539870916296e-05, "loss": 0.0579, "step": 22370 }, { "grad_norm": 0.3266737461090088, "learning_rate": 8.625116699627317e-05, "loss": 0.0401, "step": 22380 }, { "grad_norm": 0.3036263883113861, "learning_rate": 8.623692908905041e-05, "loss": 0.0339, "step": 22390 }, { "grad_norm": 0.5571492314338684, "learning_rate": 8.622268498992755e-05, "loss": 0.0536, "step": 22400 }, { "grad_norm": 0.37524130940437317, "learning_rate": 8.620843470133851e-05, "loss": 0.0543, "step": 22410 }, { "grad_norm": 0.49526098370552063, "learning_rate": 8.619417822571829e-05, "loss": 0.0474, "step": 22420 }, { "grad_norm": 0.3905222415924072, "learning_rate": 8.617991556550292e-05, "loss": 0.0373, "step": 22430 }, { "grad_norm": 0.4989441931247711, "learning_rate": 8.616564672312952e-05, "loss": 0.0487, "step": 22440 }, { "grad_norm": 0.3538248538970947, "learning_rate": 8.61513717010362e-05, "loss": 0.0482, "step": 22450 }, { "grad_norm": 0.3726920187473297, "learning_rate": 8.613709050166221e-05, "loss": 0.0532, "step": 22460 }, { "grad_norm": 0.43397364020347595, "learning_rate": 8.61228031274478e-05, "loss": 0.0537, "step": 22470 }, { "grad_norm": 0.48469364643096924, "learning_rate": 8.610850958083431e-05, "loss": 0.0454, "step": 22480 }, { "grad_norm": 0.5369492173194885, "learning_rate": 8.609420986426409e-05, "loss": 0.0517, "step": 22490 }, { "grad_norm": 0.37034866213798523, "learning_rate": 8.60799039801806e-05, "loss": 0.0404, "step": 22500 }, { "grad_norm": 0.3672083914279938, "learning_rate": 8.606559193102828e-05, "loss": 0.0359, "step": 22510 }, { "grad_norm": 0.33486098051071167, "learning_rate": 8.605127371925273e-05, "loss": 0.0467, "step": 22520 }, { "grad_norm": 0.4155297875404358, "learning_rate": 8.603694934730047e-05, "loss": 0.045, "step": 22530 }, { "grad_norm": 0.303344190120697, "learning_rate": 8.602261881761919e-05, "loss": 0.0322, "step": 22540 }, { "grad_norm": 0.24408210813999176, "learning_rate": 8.600828213265759e-05, "loss": 0.0479, "step": 22550 }, { "grad_norm": 0.46526020765304565, "learning_rate": 8.599393929486539e-05, "loss": 0.0573, "step": 22560 }, { "grad_norm": 0.39692407846450806, "learning_rate": 8.59795903066934e-05, "loss": 0.0436, "step": 22570 }, { "grad_norm": 0.430937796831131, "learning_rate": 8.596523517059347e-05, "loss": 0.0441, "step": 22580 }, { "grad_norm": 0.38001805543899536, "learning_rate": 8.59508738890185e-05, "loss": 0.04, "step": 22590 }, { "grad_norm": 0.38507527112960815, "learning_rate": 8.593650646442246e-05, "loss": 0.0421, "step": 22600 }, { "grad_norm": 0.39286500215530396, "learning_rate": 8.59221328992603e-05, "loss": 0.0425, "step": 22610 }, { "grad_norm": 0.417391300201416, "learning_rate": 8.590775319598813e-05, "loss": 0.0407, "step": 22620 }, { "grad_norm": 0.4296852946281433, "learning_rate": 8.589336735706301e-05, "loss": 0.0373, "step": 22630 }, { "grad_norm": 0.4840414822101593, "learning_rate": 8.587897538494307e-05, "loss": 0.0523, "step": 22640 }, { "grad_norm": 0.5608999729156494, "learning_rate": 8.586457728208756e-05, "loss": 0.0479, "step": 22650 }, { "grad_norm": 0.7593539357185364, "learning_rate": 8.585017305095667e-05, "loss": 0.0474, "step": 22660 }, { "grad_norm": 0.41097813844680786, "learning_rate": 8.583576269401173e-05, "loss": 0.0409, "step": 22670 }, { "grad_norm": 0.38418108224868774, "learning_rate": 8.582134621371504e-05, "loss": 0.0435, "step": 22680 }, { "grad_norm": 0.48682281374931335, "learning_rate": 8.580692361253e-05, "loss": 0.0414, "step": 22690 }, { "grad_norm": 0.4232872724533081, "learning_rate": 8.579249489292104e-05, "loss": 0.0392, "step": 22700 }, { "grad_norm": 0.4700431227684021, "learning_rate": 8.577806005735363e-05, "loss": 0.0506, "step": 22710 }, { "grad_norm": 0.36295294761657715, "learning_rate": 8.576361910829429e-05, "loss": 0.0364, "step": 22720 }, { "grad_norm": 0.42601442337036133, "learning_rate": 8.574917204821057e-05, "loss": 0.0418, "step": 22730 }, { "grad_norm": 0.34997811913490295, "learning_rate": 8.57347188795711e-05, "loss": 0.0459, "step": 22740 }, { "grad_norm": 0.44528263807296753, "learning_rate": 8.572025960484551e-05, "loss": 0.0476, "step": 22750 }, { "grad_norm": 0.5088690519332886, "learning_rate": 8.57057942265045e-05, "loss": 0.052, "step": 22760 }, { "grad_norm": 0.3779039978981018, "learning_rate": 8.569132274701984e-05, "loss": 0.0432, "step": 22770 }, { "grad_norm": 0.3384109437465668, "learning_rate": 8.567684516886427e-05, "loss": 0.0383, "step": 22780 }, { "grad_norm": 0.3878732919692993, "learning_rate": 8.56623614945116e-05, "loss": 0.0339, "step": 22790 }, { "grad_norm": 0.38632699847221375, "learning_rate": 8.564787172643675e-05, "loss": 0.0445, "step": 22800 }, { "grad_norm": 0.3784126937389374, "learning_rate": 8.563337586711559e-05, "loss": 0.0399, "step": 22810 }, { "grad_norm": 0.5078145265579224, "learning_rate": 8.561887391902506e-05, "loss": 0.0514, "step": 22820 }, { "grad_norm": 0.5378051400184631, "learning_rate": 8.560436588464316e-05, "loss": 0.0414, "step": 22830 }, { "grad_norm": 0.4054345488548279, "learning_rate": 8.55898517664489e-05, "loss": 0.0454, "step": 22840 }, { "grad_norm": 0.6596713662147522, "learning_rate": 8.557533156692236e-05, "loss": 0.0512, "step": 22850 }, { "grad_norm": 0.3563947379589081, "learning_rate": 8.556080528854467e-05, "loss": 0.0319, "step": 22860 }, { "grad_norm": 0.5582278370857239, "learning_rate": 8.554627293379791e-05, "loss": 0.0344, "step": 22870 }, { "grad_norm": 0.31257686018943787, "learning_rate": 8.553173450516531e-05, "loss": 0.0467, "step": 22880 }, { "grad_norm": 0.4222767651081085, "learning_rate": 8.551719000513108e-05, "loss": 0.0389, "step": 22890 }, { "grad_norm": 0.4174056649208069, "learning_rate": 8.550263943618049e-05, "loss": 0.0584, "step": 22900 }, { "grad_norm": 0.454760879278183, "learning_rate": 8.54880828007998e-05, "loss": 0.0532, "step": 22910 }, { "grad_norm": 0.4487668573856354, "learning_rate": 8.547352010147637e-05, "loss": 0.0616, "step": 22920 }, { "grad_norm": 0.36768054962158203, "learning_rate": 8.545895134069855e-05, "loss": 0.0478, "step": 22930 }, { "grad_norm": 0.7534730434417725, "learning_rate": 8.544437652095576e-05, "loss": 0.0424, "step": 22940 }, { "grad_norm": 0.46705740690231323, "learning_rate": 8.542979564473843e-05, "loss": 0.0584, "step": 22950 }, { "grad_norm": 0.41841843724250793, "learning_rate": 8.541520871453802e-05, "loss": 0.0492, "step": 22960 }, { "grad_norm": 0.5558544993400574, "learning_rate": 8.540061573284705e-05, "loss": 0.041, "step": 22970 }, { "grad_norm": 0.3771643042564392, "learning_rate": 8.538601670215906e-05, "loss": 0.0377, "step": 22980 }, { "grad_norm": 0.28797510266304016, "learning_rate": 8.537141162496864e-05, "loss": 0.0421, "step": 22990 }, { "grad_norm": 0.4019532799720764, "learning_rate": 8.535680050377137e-05, "loss": 0.0374, "step": 23000 }, { "grad_norm": 0.35897761583328247, "learning_rate": 8.534218334106391e-05, "loss": 0.0372, "step": 23010 }, { "grad_norm": 0.342890202999115, "learning_rate": 8.532756013934393e-05, "loss": 0.0374, "step": 23020 }, { "grad_norm": 0.370448499917984, "learning_rate": 8.531293090111012e-05, "loss": 0.0476, "step": 23030 }, { "grad_norm": 0.3708201050758362, "learning_rate": 8.529829562886225e-05, "loss": 0.0342, "step": 23040 }, { "grad_norm": 0.43160977959632874, "learning_rate": 8.528365432510105e-05, "loss": 0.0487, "step": 23050 }, { "grad_norm": 0.3282303512096405, "learning_rate": 8.526900699232833e-05, "loss": 0.0371, "step": 23060 }, { "grad_norm": 0.359165221452713, "learning_rate": 8.525435363304695e-05, "loss": 0.0398, "step": 23070 }, { "grad_norm": 0.4802524447441101, "learning_rate": 8.523969424976072e-05, "loss": 0.04, "step": 23080 }, { "grad_norm": 0.40493232011795044, "learning_rate": 8.522502884497457e-05, "loss": 0.0382, "step": 23090 }, { "grad_norm": 0.41502752900123596, "learning_rate": 8.521035742119437e-05, "loss": 0.0396, "step": 23100 }, { "grad_norm": 0.2997351884841919, "learning_rate": 8.519567998092712e-05, "loss": 0.0528, "step": 23110 }, { "grad_norm": 0.4758051633834839, "learning_rate": 8.518099652668075e-05, "loss": 0.0418, "step": 23120 }, { "grad_norm": 0.3970460593700409, "learning_rate": 8.516630706096429e-05, "loss": 0.0395, "step": 23130 }, { "grad_norm": 0.29240456223487854, "learning_rate": 8.515161158628773e-05, "loss": 0.034, "step": 23140 }, { "grad_norm": 0.41264206171035767, "learning_rate": 8.513691010516216e-05, "loss": 0.0564, "step": 23150 }, { "grad_norm": 0.321532279253006, "learning_rate": 8.512220262009966e-05, "loss": 0.0358, "step": 23160 }, { "grad_norm": 0.48036840558052063, "learning_rate": 8.510748913361332e-05, "loss": 0.0413, "step": 23170 }, { "grad_norm": 0.391921728849411, "learning_rate": 8.509276964821726e-05, "loss": 0.0357, "step": 23180 }, { "grad_norm": 0.38276049494743347, "learning_rate": 8.507804416642669e-05, "loss": 0.0361, "step": 23190 }, { "grad_norm": 0.3325672447681427, "learning_rate": 8.506331269075774e-05, "loss": 0.0372, "step": 23200 }, { "grad_norm": 0.3772946000099182, "learning_rate": 8.504857522372765e-05, "loss": 0.0312, "step": 23210 }, { "grad_norm": 0.2885102331638336, "learning_rate": 8.503383176785461e-05, "loss": 0.0377, "step": 23220 }, { "grad_norm": 0.39247849583625793, "learning_rate": 8.501908232565792e-05, "loss": 0.041, "step": 23230 }, { "grad_norm": 0.36187121272087097, "learning_rate": 8.50043268996578e-05, "loss": 0.0352, "step": 23240 }, { "grad_norm": 0.43386679887771606, "learning_rate": 8.498956549237562e-05, "loss": 0.0529, "step": 23250 }, { "grad_norm": 0.3926471173763275, "learning_rate": 8.497479810633366e-05, "loss": 0.0403, "step": 23260 }, { "grad_norm": 0.5136855840682983, "learning_rate": 8.496002474405525e-05, "loss": 0.0503, "step": 23270 }, { "grad_norm": 0.4527529776096344, "learning_rate": 8.494524540806478e-05, "loss": 0.0513, "step": 23280 }, { "grad_norm": 0.36396780610084534, "learning_rate": 8.493046010088761e-05, "loss": 0.0558, "step": 23290 }, { "grad_norm": 0.36823323369026184, "learning_rate": 8.491566882505018e-05, "loss": 0.0496, "step": 23300 }, { "grad_norm": 0.42871472239494324, "learning_rate": 8.490087158307988e-05, "loss": 0.0442, "step": 23310 }, { "grad_norm": 0.6042630672454834, "learning_rate": 8.488606837750518e-05, "loss": 0.0441, "step": 23320 }, { "grad_norm": 0.3063327670097351, "learning_rate": 8.487125921085552e-05, "loss": 0.0437, "step": 23330 }, { "grad_norm": 0.6023086309432983, "learning_rate": 8.485644408566141e-05, "loss": 0.0473, "step": 23340 }, { "grad_norm": 0.37416860461235046, "learning_rate": 8.484162300445431e-05, "loss": 0.0436, "step": 23350 }, { "grad_norm": 0.401816189289093, "learning_rate": 8.482679596976676e-05, "loss": 0.0382, "step": 23360 }, { "grad_norm": 0.29431721568107605, "learning_rate": 8.48119629841323e-05, "loss": 0.0504, "step": 23370 }, { "grad_norm": 0.4340132772922516, "learning_rate": 8.479712405008547e-05, "loss": 0.0499, "step": 23380 }, { "grad_norm": 0.37912142276763916, "learning_rate": 8.478227917016184e-05, "loss": 0.0497, "step": 23390 }, { "grad_norm": 0.25023263692855835, "learning_rate": 8.476742834689801e-05, "loss": 0.0474, "step": 23400 }, { "grad_norm": 0.5753133296966553, "learning_rate": 8.475257158283157e-05, "loss": 0.0506, "step": 23410 }, { "grad_norm": 0.44364482164382935, "learning_rate": 8.473770888050112e-05, "loss": 0.0353, "step": 23420 }, { "grad_norm": 0.6090152263641357, "learning_rate": 8.47228402424463e-05, "loss": 0.043, "step": 23430 }, { "grad_norm": 0.5679950714111328, "learning_rate": 8.470796567120775e-05, "loss": 0.0372, "step": 23440 }, { "grad_norm": 0.4950231611728668, "learning_rate": 8.469308516932714e-05, "loss": 0.0468, "step": 23450 }, { "grad_norm": 0.44390973448753357, "learning_rate": 8.467819873934714e-05, "loss": 0.0386, "step": 23460 }, { "grad_norm": 0.3728378713130951, "learning_rate": 8.466330638381143e-05, "loss": 0.0505, "step": 23470 }, { "grad_norm": 0.47100764513015747, "learning_rate": 8.464840810526469e-05, "loss": 0.0394, "step": 23480 }, { "grad_norm": 0.41440701484680176, "learning_rate": 8.463350390625264e-05, "loss": 0.0537, "step": 23490 }, { "grad_norm": 0.5011507868766785, "learning_rate": 8.4618593789322e-05, "loss": 0.0601, "step": 23500 }, { "grad_norm": 0.2984727919101715, "learning_rate": 8.46036777570205e-05, "loss": 0.0482, "step": 23510 }, { "grad_norm": 0.3895256221294403, "learning_rate": 8.458875581189688e-05, "loss": 0.0621, "step": 23520 }, { "grad_norm": 0.44451552629470825, "learning_rate": 8.457382795650092e-05, "loss": 0.0397, "step": 23530 }, { "grad_norm": 0.329892098903656, "learning_rate": 8.455889419338335e-05, "loss": 0.0444, "step": 23540 }, { "grad_norm": 0.4613569676876068, "learning_rate": 8.454395452509593e-05, "loss": 0.0418, "step": 23550 }, { "grad_norm": 0.31536760926246643, "learning_rate": 8.452900895419146e-05, "loss": 0.0404, "step": 23560 }, { "grad_norm": 0.2737281918525696, "learning_rate": 8.451405748322376e-05, "loss": 0.0435, "step": 23570 }, { "grad_norm": 0.6157851219177246, "learning_rate": 8.449910011474759e-05, "loss": 0.0466, "step": 23580 }, { "grad_norm": 0.49445274472236633, "learning_rate": 8.448413685131876e-05, "loss": 0.0527, "step": 23590 }, { "grad_norm": 0.5907596945762634, "learning_rate": 8.446916769549407e-05, "loss": 0.0471, "step": 23600 }, { "grad_norm": 0.44507279992103577, "learning_rate": 8.445419264983136e-05, "loss": 0.0477, "step": 23610 }, { "grad_norm": 0.42865753173828125, "learning_rate": 8.443921171688947e-05, "loss": 0.0373, "step": 23620 }, { "grad_norm": 0.3214915692806244, "learning_rate": 8.442422489922819e-05, "loss": 0.0381, "step": 23630 }, { "grad_norm": 0.42132362723350525, "learning_rate": 8.440923219940838e-05, "loss": 0.0462, "step": 23640 }, { "grad_norm": 1.770567774772644, "learning_rate": 8.439423361999189e-05, "loss": 0.0447, "step": 23650 }, { "grad_norm": 0.33255356550216675, "learning_rate": 8.437922916354155e-05, "loss": 0.051, "step": 23660 }, { "grad_norm": 0.3217192590236664, "learning_rate": 8.436421883262123e-05, "loss": 0.0438, "step": 23670 }, { "grad_norm": 0.30361470580101013, "learning_rate": 8.434920262979577e-05, "loss": 0.0521, "step": 23680 }, { "grad_norm": 0.4602224826812744, "learning_rate": 8.433418055763104e-05, "loss": 0.0409, "step": 23690 }, { "grad_norm": 0.42388463020324707, "learning_rate": 8.431915261869389e-05, "loss": 0.0468, "step": 23700 }, { "grad_norm": 0.4611998498439789, "learning_rate": 8.43041188155522e-05, "loss": 0.0451, "step": 23710 }, { "grad_norm": 0.32586854696273804, "learning_rate": 8.428907915077481e-05, "loss": 0.0361, "step": 23720 }, { "grad_norm": 0.4607838988304138, "learning_rate": 8.42740336269316e-05, "loss": 0.0482, "step": 23730 }, { "grad_norm": 0.5241453647613525, "learning_rate": 8.425898224659345e-05, "loss": 0.0366, "step": 23740 }, { "grad_norm": 0.34204018115997314, "learning_rate": 8.42439250123322e-05, "loss": 0.0363, "step": 23750 }, { "grad_norm": 0.37726742029190063, "learning_rate": 8.422886192672076e-05, "loss": 0.0514, "step": 23760 }, { "grad_norm": 0.5203210115432739, "learning_rate": 8.421379299233297e-05, "loss": 0.0394, "step": 23770 }, { "grad_norm": 0.6088079810142517, "learning_rate": 8.419871821174371e-05, "loss": 0.0442, "step": 23780 }, { "grad_norm": 0.40572941303253174, "learning_rate": 8.418363758752884e-05, "loss": 0.0367, "step": 23790 }, { "grad_norm": 0.353787899017334, "learning_rate": 8.416855112226523e-05, "loss": 0.0428, "step": 23800 }, { "grad_norm": 0.3244031071662903, "learning_rate": 8.415345881853075e-05, "loss": 0.0431, "step": 23810 }, { "grad_norm": 0.48374924063682556, "learning_rate": 8.413836067890426e-05, "loss": 0.043, "step": 23820 }, { "grad_norm": 0.32793647050857544, "learning_rate": 8.41232567059656e-05, "loss": 0.0409, "step": 23830 }, { "grad_norm": 0.41914108395576477, "learning_rate": 8.410814690229565e-05, "loss": 0.0517, "step": 23840 }, { "grad_norm": 0.3987395763397217, "learning_rate": 8.409303127047626e-05, "loss": 0.0383, "step": 23850 }, { "grad_norm": 0.4590511918067932, "learning_rate": 8.407790981309028e-05, "loss": 0.0374, "step": 23860 }, { "grad_norm": 0.5330420136451721, "learning_rate": 8.406278253272153e-05, "loss": 0.0477, "step": 23870 }, { "grad_norm": 0.305419921875, "learning_rate": 8.404764943195487e-05, "loss": 0.0372, "step": 23880 }, { "grad_norm": 0.326199471950531, "learning_rate": 8.403251051337613e-05, "loss": 0.0464, "step": 23890 }, { "grad_norm": 0.3373045325279236, "learning_rate": 8.401736577957214e-05, "loss": 0.0358, "step": 23900 }, { "grad_norm": 0.3459758758544922, "learning_rate": 8.40022152331307e-05, "loss": 0.0407, "step": 23910 }, { "grad_norm": 0.37372708320617676, "learning_rate": 8.398705887664064e-05, "loss": 0.0486, "step": 23920 }, { "grad_norm": 0.36430901288986206, "learning_rate": 8.397189671269177e-05, "loss": 0.0354, "step": 23930 }, { "grad_norm": 0.5660068988800049, "learning_rate": 8.395672874387488e-05, "loss": 0.0368, "step": 23940 }, { "grad_norm": 0.36146384477615356, "learning_rate": 8.394155497278177e-05, "loss": 0.0368, "step": 23950 }, { "grad_norm": 0.6243921518325806, "learning_rate": 8.392637540200523e-05, "loss": 0.0542, "step": 23960 }, { "grad_norm": 0.32447633147239685, "learning_rate": 8.391119003413902e-05, "loss": 0.037, "step": 23970 }, { "grad_norm": 0.45473426580429077, "learning_rate": 8.38959988717779e-05, "loss": 0.0382, "step": 23980 }, { "grad_norm": 0.29908502101898193, "learning_rate": 8.388080191751764e-05, "loss": 0.0457, "step": 23990 }, { "grad_norm": 0.34594401717185974, "learning_rate": 8.386559917395496e-05, "loss": 0.0386, "step": 24000 }, { "grad_norm": 0.34009283781051636, "learning_rate": 8.385039064368761e-05, "loss": 0.0292, "step": 24010 }, { "grad_norm": 0.28320959210395813, "learning_rate": 8.383517632931431e-05, "loss": 0.0306, "step": 24020 }, { "grad_norm": 0.43620049953460693, "learning_rate": 8.381995623343477e-05, "loss": 0.0363, "step": 24030 }, { "grad_norm": 0.4266238510608673, "learning_rate": 8.380473035864968e-05, "loss": 0.0444, "step": 24040 }, { "grad_norm": 0.2900135815143585, "learning_rate": 8.378949870756076e-05, "loss": 0.0425, "step": 24050 }, { "grad_norm": 0.3093028962612152, "learning_rate": 8.377426128277063e-05, "loss": 0.038, "step": 24060 }, { "grad_norm": 0.2734031081199646, "learning_rate": 8.375901808688298e-05, "loss": 0.033, "step": 24070 }, { "grad_norm": 0.5459153652191162, "learning_rate": 8.374376912250246e-05, "loss": 0.0432, "step": 24080 }, { "grad_norm": 0.3012552857398987, "learning_rate": 8.372851439223468e-05, "loss": 0.0371, "step": 24090 }, { "grad_norm": 0.4753910303115845, "learning_rate": 8.371325389868627e-05, "loss": 0.0466, "step": 24100 }, { "grad_norm": 0.4813925623893738, "learning_rate": 8.369798764446482e-05, "loss": 0.0462, "step": 24110 }, { "grad_norm": 0.5659438967704773, "learning_rate": 8.368271563217893e-05, "loss": 0.0471, "step": 24120 }, { "grad_norm": 0.30051425099372864, "learning_rate": 8.366743786443817e-05, "loss": 0.0572, "step": 24130 }, { "grad_norm": 0.5529270768165588, "learning_rate": 8.365215434385309e-05, "loss": 0.0515, "step": 24140 }, { "grad_norm": 0.3874393403530121, "learning_rate": 8.36368650730352e-05, "loss": 0.043, "step": 24150 }, { "grad_norm": 0.5898957848548889, "learning_rate": 8.362157005459705e-05, "loss": 0.0477, "step": 24160 }, { "grad_norm": 0.49991676211357117, "learning_rate": 8.360626929115213e-05, "loss": 0.0375, "step": 24170 }, { "grad_norm": 0.6054319143295288, "learning_rate": 8.359096278531492e-05, "loss": 0.0427, "step": 24180 }, { "grad_norm": 0.3386556804180145, "learning_rate": 8.357565053970088e-05, "loss": 0.0385, "step": 24190 }, { "grad_norm": 0.4213273525238037, "learning_rate": 8.356033255692647e-05, "loss": 0.0475, "step": 24200 }, { "grad_norm": 0.663413405418396, "learning_rate": 8.354500883960911e-05, "loss": 0.0417, "step": 24210 }, { "grad_norm": 0.4258385896682739, "learning_rate": 8.352967939036717e-05, "loss": 0.0446, "step": 24220 }, { "grad_norm": 0.4051204025745392, "learning_rate": 8.35143442118201e-05, "loss": 0.0503, "step": 24230 }, { "grad_norm": 0.32984283566474915, "learning_rate": 8.349900330658819e-05, "loss": 0.0442, "step": 24240 }, { "grad_norm": 0.3709541857242584, "learning_rate": 8.348365667729284e-05, "loss": 0.0424, "step": 24250 }, { "grad_norm": 0.35585689544677734, "learning_rate": 8.346830432655633e-05, "loss": 0.0334, "step": 24260 }, { "grad_norm": 0.6238483190536499, "learning_rate": 8.345294625700195e-05, "loss": 0.0391, "step": 24270 }, { "grad_norm": 0.39294755458831787, "learning_rate": 8.343758247125402e-05, "loss": 0.0392, "step": 24280 }, { "grad_norm": 0.454134464263916, "learning_rate": 8.342221297193776e-05, "loss": 0.0512, "step": 24290 }, { "grad_norm": 0.3363328278064728, "learning_rate": 8.34068377616794e-05, "loss": 0.0465, "step": 24300 }, { "grad_norm": 0.5512090921401978, "learning_rate": 8.339145684310615e-05, "loss": 0.0472, "step": 24310 }, { "grad_norm": 0.44952210783958435, "learning_rate": 8.337607021884618e-05, "loss": 0.0444, "step": 24320 }, { "grad_norm": 0.300571084022522, "learning_rate": 8.336067789152867e-05, "loss": 0.0349, "step": 24330 }, { "grad_norm": 0.3845832645893097, "learning_rate": 8.334527986378369e-05, "loss": 0.0494, "step": 24340 }, { "grad_norm": 0.485914409160614, "learning_rate": 8.332987613824239e-05, "loss": 0.0601, "step": 24350 }, { "grad_norm": 0.8619210124015808, "learning_rate": 8.331446671753685e-05, "loss": 0.0476, "step": 24360 }, { "grad_norm": 0.5043625235557556, "learning_rate": 8.329905160430007e-05, "loss": 0.0371, "step": 24370 }, { "grad_norm": 0.37515679001808167, "learning_rate": 8.328363080116611e-05, "loss": 0.0392, "step": 24380 }, { "grad_norm": 0.3518769145011902, "learning_rate": 8.326820431076997e-05, "loss": 0.0389, "step": 24390 }, { "grad_norm": 0.4623965620994568, "learning_rate": 8.325277213574759e-05, "loss": 0.04, "step": 24400 }, { "grad_norm": 0.5741809010505676, "learning_rate": 8.32373342787359e-05, "loss": 0.0394, "step": 24410 }, { "grad_norm": 0.30625757575035095, "learning_rate": 8.322189074237285e-05, "loss": 0.0417, "step": 24420 }, { "grad_norm": 0.3138837218284607, "learning_rate": 8.32064415292973e-05, "loss": 0.0407, "step": 24430 }, { "grad_norm": 0.2364334613084793, "learning_rate": 8.319098664214907e-05, "loss": 0.0378, "step": 24440 }, { "grad_norm": 0.6120840907096863, "learning_rate": 8.3175526083569e-05, "loss": 0.0428, "step": 24450 }, { "grad_norm": 0.5041807293891907, "learning_rate": 8.316005985619889e-05, "loss": 0.0426, "step": 24460 }, { "grad_norm": 0.4394310712814331, "learning_rate": 8.314458796268147e-05, "loss": 0.0434, "step": 24470 }, { "grad_norm": 0.713224470615387, "learning_rate": 8.312911040566047e-05, "loss": 0.0397, "step": 24480 }, { "grad_norm": 0.4452374279499054, "learning_rate": 8.31136271877806e-05, "loss": 0.041, "step": 24490 }, { "grad_norm": 0.33653518557548523, "learning_rate": 8.309813831168748e-05, "loss": 0.0411, "step": 24500 }, { "grad_norm": 0.3617177903652191, "learning_rate": 8.308264378002777e-05, "loss": 0.0332, "step": 24510 }, { "grad_norm": 0.3732021749019623, "learning_rate": 8.306714359544906e-05, "loss": 0.0472, "step": 24520 }, { "grad_norm": 0.3927610218524933, "learning_rate": 8.30516377605999e-05, "loss": 0.0338, "step": 24530 }, { "grad_norm": 0.4759689271450043, "learning_rate": 8.30361262781298e-05, "loss": 0.0323, "step": 24540 }, { "grad_norm": 0.34843629598617554, "learning_rate": 8.302060915068924e-05, "loss": 0.0482, "step": 24550 }, { "grad_norm": 0.49176377058029175, "learning_rate": 8.300508638092972e-05, "loss": 0.0376, "step": 24560 }, { "grad_norm": 0.41627296805381775, "learning_rate": 8.298955797150361e-05, "loss": 0.0457, "step": 24570 }, { "grad_norm": 0.4340144693851471, "learning_rate": 8.297402392506433e-05, "loss": 0.0422, "step": 24580 }, { "grad_norm": 0.5372414588928223, "learning_rate": 8.295848424426617e-05, "loss": 0.0351, "step": 24590 }, { "grad_norm": 0.5618509650230408, "learning_rate": 8.29429389317645e-05, "loss": 0.0395, "step": 24600 }, { "grad_norm": 0.31086814403533936, "learning_rate": 8.292738799021556e-05, "loss": 0.0436, "step": 24610 }, { "grad_norm": 0.5837622284889221, "learning_rate": 8.291183142227656e-05, "loss": 0.0357, "step": 24620 }, { "grad_norm": 0.3934926390647888, "learning_rate": 8.289626923060572e-05, "loss": 0.0519, "step": 24630 }, { "grad_norm": 0.45948198437690735, "learning_rate": 8.288070141786218e-05, "loss": 0.0439, "step": 24640 }, { "grad_norm": 0.40835118293762207, "learning_rate": 8.286512798670605e-05, "loss": 0.0448, "step": 24650 }, { "grad_norm": 0.5159660577774048, "learning_rate": 8.284954893979842e-05, "loss": 0.0416, "step": 24660 }, { "grad_norm": 0.253311425447464, "learning_rate": 8.283396427980131e-05, "loss": 0.0411, "step": 24670 }, { "grad_norm": 0.4078831076622009, "learning_rate": 8.281837400937771e-05, "loss": 0.0413, "step": 24680 }, { "grad_norm": 0.40348246693611145, "learning_rate": 8.28027781311916e-05, "loss": 0.0405, "step": 24690 }, { "grad_norm": 0.3802432119846344, "learning_rate": 8.278717664790785e-05, "loss": 0.0414, "step": 24700 }, { "grad_norm": 0.45685046911239624, "learning_rate": 8.277156956219234e-05, "loss": 0.0425, "step": 24710 }, { "grad_norm": 0.5111185908317566, "learning_rate": 8.275595687671189e-05, "loss": 0.0352, "step": 24720 }, { "grad_norm": 0.6452677249908447, "learning_rate": 8.27403385941343e-05, "loss": 0.0437, "step": 24730 }, { "grad_norm": 0.36411556601524353, "learning_rate": 8.272471471712828e-05, "loss": 0.0394, "step": 24740 }, { "grad_norm": 0.4253517985343933, "learning_rate": 8.270908524836355e-05, "loss": 0.046, "step": 24750 }, { "grad_norm": 0.3707743287086487, "learning_rate": 8.269345019051074e-05, "loss": 0.0532, "step": 24760 }, { "grad_norm": 0.5345566272735596, "learning_rate": 8.267780954624147e-05, "loss": 0.0445, "step": 24770 }, { "grad_norm": 0.3714980185031891, "learning_rate": 8.266216331822827e-05, "loss": 0.0403, "step": 24780 }, { "grad_norm": 0.42374932765960693, "learning_rate": 8.264651150914469e-05, "loss": 0.0379, "step": 24790 }, { "grad_norm": 0.34787195920944214, "learning_rate": 8.263085412166517e-05, "loss": 0.0336, "step": 24800 }, { "grad_norm": 1.0797016620635986, "learning_rate": 8.261519115846514e-05, "loss": 0.0352, "step": 24810 }, { "grad_norm": 0.3686676323413849, "learning_rate": 8.259952262222096e-05, "loss": 0.0369, "step": 24820 }, { "grad_norm": 0.2942864000797272, "learning_rate": 8.258384851560997e-05, "loss": 0.0399, "step": 24830 }, { "grad_norm": 0.4748389422893524, "learning_rate": 8.256816884131044e-05, "loss": 0.0416, "step": 24840 }, { "grad_norm": 0.3025726079940796, "learning_rate": 8.255248360200159e-05, "loss": 0.0319, "step": 24850 }, { "grad_norm": 0.4192485213279724, "learning_rate": 8.253679280036359e-05, "loss": 0.0423, "step": 24860 }, { "grad_norm": 0.4462782144546509, "learning_rate": 8.252109643907762e-05, "loss": 0.0605, "step": 24870 }, { "grad_norm": 0.3381744921207428, "learning_rate": 8.250539452082569e-05, "loss": 0.0413, "step": 24880 }, { "grad_norm": 0.3735997676849365, "learning_rate": 8.248968704829087e-05, "loss": 0.0405, "step": 24890 }, { "grad_norm": 0.3993508815765381, "learning_rate": 8.247397402415714e-05, "loss": 0.0403, "step": 24900 }, { "grad_norm": 0.4055646061897278, "learning_rate": 8.24582554511094e-05, "loss": 0.0372, "step": 24910 }, { "grad_norm": 0.3109389841556549, "learning_rate": 8.244253133183355e-05, "loss": 0.0357, "step": 24920 }, { "grad_norm": 0.31392744183540344, "learning_rate": 8.24268016690164e-05, "loss": 0.0449, "step": 24930 }, { "grad_norm": 0.4562455713748932, "learning_rate": 8.241106646534571e-05, "loss": 0.0372, "step": 24940 }, { "grad_norm": 0.363057941198349, "learning_rate": 8.23953257235102e-05, "loss": 0.037, "step": 24950 }, { "grad_norm": 0.5787076950073242, "learning_rate": 8.237957944619956e-05, "loss": 0.0468, "step": 24960 }, { "grad_norm": 0.5030458569526672, "learning_rate": 8.236382763610437e-05, "loss": 0.0365, "step": 24970 }, { "grad_norm": 0.5137977600097656, "learning_rate": 8.234807029591619e-05, "loss": 0.0395, "step": 24980 }, { "grad_norm": 0.3772628605365753, "learning_rate": 8.233230742832752e-05, "loss": 0.0567, "step": 24990 }, { "grad_norm": 0.4658295214176178, "learning_rate": 8.231653903603178e-05, "loss": 0.0452, "step": 25000 }, { "grad_norm": 0.6219373345375061, "learning_rate": 8.23007651217234e-05, "loss": 0.038, "step": 25010 }, { "grad_norm": 0.3939495384693146, "learning_rate": 8.228498568809769e-05, "loss": 0.0373, "step": 25020 }, { "grad_norm": 0.6112968921661377, "learning_rate": 8.22692007378509e-05, "loss": 0.0496, "step": 25030 }, { "grad_norm": 0.4493216276168823, "learning_rate": 8.225341027368028e-05, "loss": 0.0416, "step": 25040 }, { "grad_norm": 0.294026255607605, "learning_rate": 8.223761429828399e-05, "loss": 0.0402, "step": 25050 }, { "grad_norm": 0.33539602160453796, "learning_rate": 8.22218128143611e-05, "loss": 0.042, "step": 25060 }, { "grad_norm": 0.387373149394989, "learning_rate": 8.220600582461166e-05, "loss": 0.0376, "step": 25070 }, { "grad_norm": 0.2585362195968628, "learning_rate": 8.219019333173668e-05, "loss": 0.0454, "step": 25080 }, { "grad_norm": 0.43708309531211853, "learning_rate": 8.217437533843805e-05, "loss": 0.046, "step": 25090 }, { "grad_norm": 0.38997551798820496, "learning_rate": 8.215855184741867e-05, "loss": 0.0367, "step": 25100 }, { "grad_norm": 0.3966960906982422, "learning_rate": 8.21427228613823e-05, "loss": 0.0441, "step": 25110 }, { "grad_norm": 0.4946901798248291, "learning_rate": 8.21268883830337e-05, "loss": 0.0444, "step": 25120 }, { "grad_norm": 0.40369167923927307, "learning_rate": 8.211104841507855e-05, "loss": 0.052, "step": 25130 }, { "grad_norm": 0.3711593449115753, "learning_rate": 8.209520296022346e-05, "loss": 0.0438, "step": 25140 }, { "grad_norm": 0.3202763497829437, "learning_rate": 8.207935202117599e-05, "loss": 0.0379, "step": 25150 }, { "grad_norm": 0.35567983984947205, "learning_rate": 8.206349560064463e-05, "loss": 0.0426, "step": 25160 }, { "grad_norm": 0.47121143341064453, "learning_rate": 8.204763370133881e-05, "loss": 0.0362, "step": 25170 }, { "grad_norm": 0.38276925683021545, "learning_rate": 8.203176632596892e-05, "loss": 0.0451, "step": 25180 }, { "grad_norm": 0.31392785906791687, "learning_rate": 8.20158934772462e-05, "loss": 0.0395, "step": 25190 }, { "grad_norm": 0.4010296165943146, "learning_rate": 8.200001515788294e-05, "loss": 0.0387, "step": 25200 }, { "grad_norm": 0.4762752056121826, "learning_rate": 8.198413137059228e-05, "loss": 0.0476, "step": 25210 }, { "grad_norm": 0.39659732580184937, "learning_rate": 8.196824211808835e-05, "loss": 0.0322, "step": 25220 }, { "grad_norm": 0.3022807240486145, "learning_rate": 8.195234740308617e-05, "loss": 0.0437, "step": 25230 }, { "grad_norm": 0.35663366317749023, "learning_rate": 8.193644722830171e-05, "loss": 0.0392, "step": 25240 }, { "grad_norm": 0.36201536655426025, "learning_rate": 8.19205415964519e-05, "loss": 0.0448, "step": 25250 }, { "grad_norm": 0.4279096722602844, "learning_rate": 8.190463051025456e-05, "loss": 0.0453, "step": 25260 }, { "grad_norm": 0.8305291533470154, "learning_rate": 8.188871397242843e-05, "loss": 0.0414, "step": 25270 }, { "grad_norm": 0.31688162684440613, "learning_rate": 8.187279198569326e-05, "loss": 0.038, "step": 25280 }, { "grad_norm": 0.4300632178783417, "learning_rate": 8.185686455276966e-05, "loss": 0.0436, "step": 25290 }, { "grad_norm": 0.3479504883289337, "learning_rate": 8.184093167637921e-05, "loss": 0.0396, "step": 25300 }, { "grad_norm": 0.49097293615341187, "learning_rate": 8.182499335924437e-05, "loss": 0.0373, "step": 25310 }, { "grad_norm": 0.6245836615562439, "learning_rate": 8.18090496040886e-05, "loss": 0.0348, "step": 25320 }, { "grad_norm": 0.31295865774154663, "learning_rate": 8.179310041363621e-05, "loss": 0.0446, "step": 25330 }, { "grad_norm": 0.46096429228782654, "learning_rate": 8.17771457906125e-05, "loss": 0.0382, "step": 25340 }, { "grad_norm": 0.552966296672821, "learning_rate": 8.176118573774371e-05, "loss": 0.0322, "step": 25350 }, { "grad_norm": 0.5272948741912842, "learning_rate": 8.174522025775692e-05, "loss": 0.0425, "step": 25360 }, { "grad_norm": 0.47431549429893494, "learning_rate": 8.172924935338022e-05, "loss": 0.0449, "step": 25370 }, { "grad_norm": 0.33069735765457153, "learning_rate": 8.171327302734262e-05, "loss": 0.0409, "step": 25380 }, { "grad_norm": 0.3373515009880066, "learning_rate": 8.169729128237401e-05, "loss": 0.0344, "step": 25390 }, { "grad_norm": 0.38746359944343567, "learning_rate": 8.168130412120525e-05, "loss": 0.035, "step": 25400 }, { "grad_norm": 0.3791331946849823, "learning_rate": 8.16653115465681e-05, "loss": 0.0443, "step": 25410 }, { "grad_norm": 0.4718109965324402, "learning_rate": 8.164931356119526e-05, "loss": 0.0291, "step": 25420 }, { "grad_norm": 0.4531159996986389, "learning_rate": 8.163331016782032e-05, "loss": 0.0466, "step": 25430 }, { "grad_norm": 0.5477817058563232, "learning_rate": 8.161730136917785e-05, "loss": 0.0288, "step": 25440 }, { "grad_norm": 0.5938358902931213, "learning_rate": 8.160128716800333e-05, "loss": 0.0478, "step": 25450 }, { "grad_norm": 0.4364064633846283, "learning_rate": 8.158526756703313e-05, "loss": 0.0386, "step": 25460 }, { "grad_norm": 0.3734825849533081, "learning_rate": 8.156924256900455e-05, "loss": 0.0414, "step": 25470 }, { "grad_norm": 0.40126511454582214, "learning_rate": 8.155321217665584e-05, "loss": 0.0355, "step": 25480 }, { "grad_norm": 0.31305813789367676, "learning_rate": 8.153717639272614e-05, "loss": 0.0374, "step": 25490 }, { "grad_norm": 0.5932291150093079, "learning_rate": 8.152113521995555e-05, "loss": 0.0448, "step": 25500 }, { "grad_norm": 0.33244338631629944, "learning_rate": 8.150508866108505e-05, "loss": 0.0341, "step": 25510 }, { "grad_norm": 0.6022777557373047, "learning_rate": 8.148903671885657e-05, "loss": 0.0392, "step": 25520 }, { "grad_norm": 0.39084547758102417, "learning_rate": 8.147297939601292e-05, "loss": 0.0454, "step": 25530 }, { "grad_norm": 0.43337199091911316, "learning_rate": 8.145691669529792e-05, "loss": 0.0474, "step": 25540 }, { "grad_norm": 0.4025699496269226, "learning_rate": 8.144084861945618e-05, "loss": 0.0426, "step": 25550 }, { "grad_norm": 0.5246708989143372, "learning_rate": 8.142477517123333e-05, "loss": 0.039, "step": 25560 }, { "grad_norm": 0.4639624357223511, "learning_rate": 8.140869635337586e-05, "loss": 0.0385, "step": 25570 }, { "grad_norm": 0.3004366159439087, "learning_rate": 8.139261216863123e-05, "loss": 0.0433, "step": 25580 }, { "grad_norm": 0.30311328172683716, "learning_rate": 8.137652261974776e-05, "loss": 0.0377, "step": 25590 }, { "grad_norm": 0.3632791340351105, "learning_rate": 8.136042770947472e-05, "loss": 0.0392, "step": 25600 }, { "grad_norm": 0.2619684934616089, "learning_rate": 8.134432744056228e-05, "loss": 0.0422, "step": 25610 }, { "grad_norm": 0.33322247862815857, "learning_rate": 8.132822181576158e-05, "loss": 0.0434, "step": 25620 }, { "grad_norm": 0.44487544894218445, "learning_rate": 8.131211083782459e-05, "loss": 0.0462, "step": 25630 }, { "grad_norm": 0.4867202043533325, "learning_rate": 8.129599450950424e-05, "loss": 0.0358, "step": 25640 }, { "grad_norm": 0.537584125995636, "learning_rate": 8.127987283355438e-05, "loss": 0.0425, "step": 25650 }, { "grad_norm": 0.31472933292388916, "learning_rate": 8.126374581272976e-05, "loss": 0.0306, "step": 25660 }, { "grad_norm": 0.527162492275238, "learning_rate": 8.124761344978605e-05, "loss": 0.046, "step": 25670 }, { "grad_norm": 0.3946962356567383, "learning_rate": 8.12314757474798e-05, "loss": 0.0358, "step": 25680 }, { "grad_norm": 0.34521374106407166, "learning_rate": 8.121533270856856e-05, "loss": 0.0432, "step": 25690 }, { "grad_norm": 0.39211505651474, "learning_rate": 8.119918433581069e-05, "loss": 0.0343, "step": 25700 }, { "grad_norm": 0.4261539578437805, "learning_rate": 8.118303063196551e-05, "loss": 0.0372, "step": 25710 }, { "grad_norm": 0.38439828157424927, "learning_rate": 8.116687159979326e-05, "loss": 0.0365, "step": 25720 }, { "grad_norm": 0.40005379915237427, "learning_rate": 8.115070724205508e-05, "loss": 0.0487, "step": 25730 }, { "grad_norm": 0.5223513245582581, "learning_rate": 8.113453756151296e-05, "loss": 0.0443, "step": 25740 }, { "grad_norm": 0.4351767301559448, "learning_rate": 8.111836256092995e-05, "loss": 0.0412, "step": 25750 }, { "grad_norm": 0.5341153740882874, "learning_rate": 8.110218224306985e-05, "loss": 0.0427, "step": 25760 }, { "grad_norm": 0.9625320434570312, "learning_rate": 8.108599661069745e-05, "loss": 0.0457, "step": 25770 }, { "grad_norm": 0.5578693151473999, "learning_rate": 8.106980566657845e-05, "loss": 0.0343, "step": 25780 }, { "grad_norm": 0.4611549377441406, "learning_rate": 8.10536094134794e-05, "loss": 0.0437, "step": 25790 }, { "grad_norm": 0.777712345123291, "learning_rate": 8.103740785416783e-05, "loss": 0.0396, "step": 25800 }, { "grad_norm": 0.4533154368400574, "learning_rate": 8.102120099141212e-05, "loss": 0.0402, "step": 25810 }, { "grad_norm": 1.4653129577636719, "learning_rate": 8.100498882798163e-05, "loss": 0.0353, "step": 25820 }, { "grad_norm": 0.33195289969444275, "learning_rate": 8.09887713666465e-05, "loss": 0.043, "step": 25830 }, { "grad_norm": 0.35258951783180237, "learning_rate": 8.09725486101779e-05, "loss": 0.0404, "step": 25840 }, { "grad_norm": 0.303424596786499, "learning_rate": 8.095632056134784e-05, "loss": 0.0486, "step": 25850 }, { "grad_norm": 0.40286996960639954, "learning_rate": 8.094008722292925e-05, "loss": 0.0389, "step": 25860 }, { "grad_norm": 0.3620084226131439, "learning_rate": 8.092384859769598e-05, "loss": 0.0423, "step": 25870 }, { "grad_norm": 0.4013848602771759, "learning_rate": 8.090760468842275e-05, "loss": 0.0466, "step": 25880 }, { "grad_norm": 0.251365065574646, "learning_rate": 8.089135549788521e-05, "loss": 0.0387, "step": 25890 }, { "grad_norm": 0.29330042004585266, "learning_rate": 8.087510102885987e-05, "loss": 0.0299, "step": 25900 }, { "grad_norm": 0.3226483464241028, "learning_rate": 8.085884128412422e-05, "loss": 0.0406, "step": 25910 }, { "grad_norm": 0.46633434295654297, "learning_rate": 8.084257626645659e-05, "loss": 0.0324, "step": 25920 }, { "grad_norm": 0.39683178067207336, "learning_rate": 8.08263059786362e-05, "loss": 0.0302, "step": 25930 }, { "grad_norm": 0.4056679904460907, "learning_rate": 8.081003042344325e-05, "loss": 0.0331, "step": 25940 }, { "grad_norm": 0.37015560269355774, "learning_rate": 8.079374960365872e-05, "loss": 0.0412, "step": 25950 }, { "grad_norm": 0.5269263982772827, "learning_rate": 8.077746352206463e-05, "loss": 0.037, "step": 25960 }, { "grad_norm": 0.6203103065490723, "learning_rate": 8.076117218144377e-05, "loss": 0.0348, "step": 25970 }, { "grad_norm": 0.4005374610424042, "learning_rate": 8.074487558457991e-05, "loss": 0.0358, "step": 25980 }, { "grad_norm": 0.4095280170440674, "learning_rate": 8.072857373425768e-05, "loss": 0.038, "step": 25990 }, { "grad_norm": 0.4212171137332916, "learning_rate": 8.071226663326264e-05, "loss": 0.0332, "step": 26000 }, { "grad_norm": 0.3945312798023224, "learning_rate": 8.069595428438121e-05, "loss": 0.0432, "step": 26010 }, { "grad_norm": 0.3508782982826233, "learning_rate": 8.067963669040072e-05, "loss": 0.0438, "step": 26020 }, { "grad_norm": 0.4675115942955017, "learning_rate": 8.066331385410942e-05, "loss": 0.0325, "step": 26030 }, { "grad_norm": 0.4156157970428467, "learning_rate": 8.064698577829641e-05, "loss": 0.0387, "step": 26040 }, { "grad_norm": 0.3332407474517822, "learning_rate": 8.063065246575175e-05, "loss": 0.0383, "step": 26050 }, { "grad_norm": 0.2966417074203491, "learning_rate": 8.061431391926631e-05, "loss": 0.0325, "step": 26060 }, { "grad_norm": 0.4364691376686096, "learning_rate": 8.059797014163195e-05, "loss": 0.0427, "step": 26070 }, { "grad_norm": 0.44108930230140686, "learning_rate": 8.058162113564133e-05, "loss": 0.0496, "step": 26080 }, { "grad_norm": 0.5041614174842834, "learning_rate": 8.056526690408806e-05, "loss": 0.0471, "step": 26090 }, { "grad_norm": 0.3599418103694916, "learning_rate": 8.054890744976666e-05, "loss": 0.0407, "step": 26100 }, { "grad_norm": 0.31763955950737, "learning_rate": 8.053254277547248e-05, "loss": 0.0358, "step": 26110 }, { "grad_norm": 0.5715650916099548, "learning_rate": 8.051617288400182e-05, "loss": 0.0416, "step": 26120 }, { "grad_norm": 0.28298091888427734, "learning_rate": 8.049979777815182e-05, "loss": 0.0385, "step": 26130 }, { "grad_norm": 0.4368724822998047, "learning_rate": 8.048341746072054e-05, "loss": 0.0372, "step": 26140 }, { "grad_norm": 0.40701523423194885, "learning_rate": 8.046703193450696e-05, "loss": 0.0409, "step": 26150 }, { "grad_norm": 0.6216837763786316, "learning_rate": 8.04506412023109e-05, "loss": 0.0404, "step": 26160 }, { "grad_norm": 0.2566041350364685, "learning_rate": 8.043424526693306e-05, "loss": 0.0407, "step": 26170 }, { "grad_norm": 0.4718979597091675, "learning_rate": 8.04178441311751e-05, "loss": 0.0317, "step": 26180 }, { "grad_norm": 0.39743611216545105, "learning_rate": 8.04014377978395e-05, "loss": 0.0315, "step": 26190 }, { "grad_norm": 0.7141262292861938, "learning_rate": 8.038502626972967e-05, "loss": 0.0491, "step": 26200 }, { "grad_norm": 0.3735017776489258, "learning_rate": 8.036860954964989e-05, "loss": 0.0432, "step": 26210 }, { "grad_norm": 0.42656445503234863, "learning_rate": 8.035218764040531e-05, "loss": 0.0396, "step": 26220 }, { "grad_norm": 0.5386948585510254, "learning_rate": 8.033576054480199e-05, "loss": 0.0395, "step": 26230 }, { "grad_norm": 0.356005996465683, "learning_rate": 8.031932826564688e-05, "loss": 0.0339, "step": 26240 }, { "grad_norm": 0.37854650616645813, "learning_rate": 8.030289080574782e-05, "loss": 0.0425, "step": 26250 }, { "grad_norm": 0.388591468334198, "learning_rate": 8.028644816791349e-05, "loss": 0.0396, "step": 26260 }, { "grad_norm": 0.34103259444236755, "learning_rate": 8.027000035495351e-05, "loss": 0.0319, "step": 26270 }, { "grad_norm": 0.33470696210861206, "learning_rate": 8.025354736967836e-05, "loss": 0.0481, "step": 26280 }, { "grad_norm": 0.33211979269981384, "learning_rate": 8.023708921489941e-05, "loss": 0.0484, "step": 26290 }, { "grad_norm": 0.46441662311553955, "learning_rate": 8.022062589342887e-05, "loss": 0.0484, "step": 26300 }, { "grad_norm": 0.5146073698997498, "learning_rate": 8.020415740807993e-05, "loss": 0.0459, "step": 26310 }, { "grad_norm": 0.4078150689601898, "learning_rate": 8.018768376166656e-05, "loss": 0.0394, "step": 26320 }, { "grad_norm": 0.5927128195762634, "learning_rate": 8.017120495700368e-05, "loss": 0.0412, "step": 26330 }, { "grad_norm": 0.4733007848262787, "learning_rate": 8.015472099690704e-05, "loss": 0.0358, "step": 26340 }, { "grad_norm": 0.28564873337745667, "learning_rate": 8.013823188419332e-05, "loss": 0.0339, "step": 26350 }, { "grad_norm": 0.4224521815776825, "learning_rate": 8.012173762168006e-05, "loss": 0.039, "step": 26360 }, { "grad_norm": 0.6078730225563049, "learning_rate": 8.010523821218567e-05, "loss": 0.0384, "step": 26370 }, { "grad_norm": 0.41697409749031067, "learning_rate": 8.008873365852945e-05, "loss": 0.0376, "step": 26380 }, { "grad_norm": 0.4343544542789459, "learning_rate": 8.007222396353157e-05, "loss": 0.0415, "step": 26390 }, { "grad_norm": 0.4941118657588959, "learning_rate": 8.00557091300131e-05, "loss": 0.0361, "step": 26400 }, { "grad_norm": 0.3693160116672516, "learning_rate": 8.003918916079597e-05, "loss": 0.0432, "step": 26410 }, { "grad_norm": 0.3596316874027252, "learning_rate": 8.002266405870298e-05, "loss": 0.0318, "step": 26420 }, { "grad_norm": 0.31744280457496643, "learning_rate": 8.000613382655782e-05, "loss": 0.0424, "step": 26430 }, { "grad_norm": 0.45033541321754456, "learning_rate": 7.998959846718505e-05, "loss": 0.0401, "step": 26440 }, { "grad_norm": 0.3565666377544403, "learning_rate": 7.997305798341012e-05, "loss": 0.0426, "step": 26450 }, { "grad_norm": 0.40418365597724915, "learning_rate": 7.995651237805937e-05, "loss": 0.0416, "step": 26460 }, { "grad_norm": 0.44713228940963745, "learning_rate": 7.993996165395996e-05, "loss": 0.0301, "step": 26470 }, { "grad_norm": 0.3272956311702728, "learning_rate": 7.992340581393996e-05, "loss": 0.0371, "step": 26480 }, { "grad_norm": 0.2898213267326355, "learning_rate": 7.990684486082831e-05, "loss": 0.0336, "step": 26490 }, { "grad_norm": 0.48419326543807983, "learning_rate": 7.989027879745482e-05, "loss": 0.0416, "step": 26500 }, { "grad_norm": 0.364582359790802, "learning_rate": 7.98737076266502e-05, "loss": 0.0443, "step": 26510 }, { "grad_norm": 0.38663727045059204, "learning_rate": 7.985713135124598e-05, "loss": 0.0438, "step": 26520 }, { "grad_norm": 0.38244709372520447, "learning_rate": 7.98405499740746e-05, "loss": 0.0393, "step": 26530 }, { "grad_norm": 0.4519927203655243, "learning_rate": 7.98239634979694e-05, "loss": 0.035, "step": 26540 }, { "grad_norm": 0.47237250208854675, "learning_rate": 7.98073719257645e-05, "loss": 0.0372, "step": 26550 }, { "grad_norm": 0.49250760674476624, "learning_rate": 7.979077526029499e-05, "loss": 0.0403, "step": 26560 }, { "grad_norm": 0.26094570755958557, "learning_rate": 7.977417350439675e-05, "loss": 0.0318, "step": 26570 }, { "grad_norm": 0.4182516932487488, "learning_rate": 7.97575666609066e-05, "loss": 0.0377, "step": 26580 }, { "grad_norm": 0.3828883469104767, "learning_rate": 7.974095473266216e-05, "loss": 0.0428, "step": 26590 }, { "grad_norm": 0.3993728756904602, "learning_rate": 7.972433772250198e-05, "loss": 0.0426, "step": 26600 }, { "grad_norm": 0.5001667141914368, "learning_rate": 7.970771563326544e-05, "loss": 0.0422, "step": 26610 }, { "grad_norm": 0.3912253677845001, "learning_rate": 7.96910884677928e-05, "loss": 0.0477, "step": 26620 }, { "grad_norm": 0.4091661870479584, "learning_rate": 7.967445622892523e-05, "loss": 0.0448, "step": 26630 }, { "grad_norm": 0.37208741903305054, "learning_rate": 7.965781891950465e-05, "loss": 0.0361, "step": 26640 }, { "grad_norm": 0.4171026647090912, "learning_rate": 7.964117654237397e-05, "loss": 0.0312, "step": 26650 }, { "grad_norm": 0.40117403864860535, "learning_rate": 7.962452910037692e-05, "loss": 0.0411, "step": 26660 }, { "grad_norm": 0.4204687178134918, "learning_rate": 7.96078765963581e-05, "loss": 0.0344, "step": 26670 }, { "grad_norm": 0.4301338195800781, "learning_rate": 7.95912190331629e-05, "loss": 0.0437, "step": 26680 }, { "grad_norm": 0.3603147566318512, "learning_rate": 7.957455641363772e-05, "loss": 0.0405, "step": 26690 }, { "grad_norm": 0.4767201542854309, "learning_rate": 7.955788874062968e-05, "loss": 0.0559, "step": 26700 }, { "grad_norm": 0.41309934854507446, "learning_rate": 7.95412160169869e-05, "loss": 0.0445, "step": 26710 }, { "grad_norm": 0.47587597370147705, "learning_rate": 7.952453824555824e-05, "loss": 0.0439, "step": 26720 }, { "grad_norm": 0.3753611147403717, "learning_rate": 7.95078554291935e-05, "loss": 0.0336, "step": 26730 }, { "grad_norm": 0.4268398880958557, "learning_rate": 7.94911675707433e-05, "loss": 0.0349, "step": 26740 }, { "grad_norm": 0.36759087443351746, "learning_rate": 7.947447467305915e-05, "loss": 0.0409, "step": 26750 }, { "grad_norm": 0.29062411189079285, "learning_rate": 7.94577767389934e-05, "loss": 0.032, "step": 26760 }, { "grad_norm": 0.43547338247299194, "learning_rate": 7.944107377139928e-05, "loss": 0.0368, "step": 26770 }, { "grad_norm": 0.4598444402217865, "learning_rate": 7.942436577313088e-05, "loss": 0.0426, "step": 26780 }, { "grad_norm": 0.4065611958503723, "learning_rate": 7.940765274704312e-05, "loss": 0.0446, "step": 26790 }, { "grad_norm": 0.3301936984062195, "learning_rate": 7.939093469599181e-05, "loss": 0.0397, "step": 26800 }, { "grad_norm": 0.3900900185108185, "learning_rate": 7.93742116228336e-05, "loss": 0.0366, "step": 26810 }, { "grad_norm": 0.3827505111694336, "learning_rate": 7.935748353042602e-05, "loss": 0.0396, "step": 26820 }, { "grad_norm": 0.444869726896286, "learning_rate": 7.934075042162744e-05, "loss": 0.0351, "step": 26830 }, { "grad_norm": 0.5470142960548401, "learning_rate": 7.932401229929705e-05, "loss": 0.0401, "step": 26840 }, { "grad_norm": 0.6410254836082458, "learning_rate": 7.9307269166295e-05, "loss": 0.0385, "step": 26850 }, { "grad_norm": 0.27170926332473755, "learning_rate": 7.92905210254822e-05, "loss": 0.0347, "step": 26860 }, { "grad_norm": 0.30987977981567383, "learning_rate": 7.927376787972045e-05, "loss": 0.0377, "step": 26870 }, { "grad_norm": 0.44786062836647034, "learning_rate": 7.92570097318724e-05, "loss": 0.0462, "step": 26880 }, { "grad_norm": 0.34665706753730774, "learning_rate": 7.924024658480158e-05, "loss": 0.0357, "step": 26890 }, { "grad_norm": 0.4491877257823944, "learning_rate": 7.922347844137233e-05, "loss": 0.0378, "step": 26900 }, { "grad_norm": 0.2485455423593521, "learning_rate": 7.92067053044499e-05, "loss": 0.0383, "step": 26910 }, { "grad_norm": 0.49866983294487, "learning_rate": 7.918992717690031e-05, "loss": 0.0432, "step": 26920 }, { "grad_norm": 0.49247947335243225, "learning_rate": 7.917314406159053e-05, "loss": 0.0386, "step": 26930 }, { "grad_norm": 0.5605646371841431, "learning_rate": 7.915635596138832e-05, "loss": 0.0481, "step": 26940 }, { "grad_norm": 0.3553234338760376, "learning_rate": 7.913956287916228e-05, "loss": 0.0444, "step": 26950 }, { "grad_norm": 0.3378056585788727, "learning_rate": 7.912276481778193e-05, "loss": 0.0331, "step": 26960 }, { "grad_norm": 0.2941969335079193, "learning_rate": 7.910596178011759e-05, "loss": 0.0421, "step": 26970 }, { "grad_norm": 0.33343425393104553, "learning_rate": 7.908915376904043e-05, "loss": 0.0351, "step": 26980 }, { "grad_norm": 0.33140984177589417, "learning_rate": 7.907234078742247e-05, "loss": 0.0344, "step": 26990 }, { "grad_norm": 0.3600066602230072, "learning_rate": 7.90555228381366e-05, "loss": 0.0333, "step": 27000 }, { "grad_norm": 0.3597949743270874, "learning_rate": 7.903869992405656e-05, "loss": 0.0302, "step": 27010 }, { "grad_norm": 0.4588044583797455, "learning_rate": 7.902187204805691e-05, "loss": 0.0344, "step": 27020 }, { "grad_norm": 0.48208895325660706, "learning_rate": 7.900503921301308e-05, "loss": 0.0319, "step": 27030 }, { "grad_norm": 0.386430948972702, "learning_rate": 7.898820142180133e-05, "loss": 0.0486, "step": 27040 }, { "grad_norm": 0.42877018451690674, "learning_rate": 7.897135867729879e-05, "loss": 0.0332, "step": 27050 }, { "grad_norm": 0.5832762122154236, "learning_rate": 7.89545109823834e-05, "loss": 0.0303, "step": 27060 }, { "grad_norm": 0.3118719458580017, "learning_rate": 7.8937658339934e-05, "loss": 0.0354, "step": 27070 }, { "grad_norm": 0.3495251536369324, "learning_rate": 7.892080075283026e-05, "loss": 0.0392, "step": 27080 }, { "grad_norm": 0.3927771747112274, "learning_rate": 7.890393822395263e-05, "loss": 0.0489, "step": 27090 }, { "grad_norm": 0.38977134227752686, "learning_rate": 7.88870707561825e-05, "loss": 0.0486, "step": 27100 }, { "grad_norm": 0.42120060324668884, "learning_rate": 7.887019835240203e-05, "loss": 0.0513, "step": 27110 }, { "grad_norm": 0.6833174824714661, "learning_rate": 7.885332101549427e-05, "loss": 0.0435, "step": 27120 }, { "grad_norm": 0.3639395534992218, "learning_rate": 7.883643874834308e-05, "loss": 0.0366, "step": 27130 }, { "grad_norm": 0.3598639965057373, "learning_rate": 7.881955155383321e-05, "loss": 0.0443, "step": 27140 }, { "grad_norm": 0.4539203345775604, "learning_rate": 7.880265943485017e-05, "loss": 0.0376, "step": 27150 }, { "grad_norm": 0.32049769163131714, "learning_rate": 7.878576239428038e-05, "loss": 0.036, "step": 27160 }, { "grad_norm": 0.43034568428993225, "learning_rate": 7.87688604350111e-05, "loss": 0.049, "step": 27170 }, { "grad_norm": 0.5858937501907349, "learning_rate": 7.875195355993042e-05, "loss": 0.039, "step": 27180 }, { "grad_norm": 0.30035606026649475, "learning_rate": 7.873504177192724e-05, "loss": 0.0377, "step": 27190 }, { "grad_norm": 0.33951881527900696, "learning_rate": 7.87181250738913e-05, "loss": 0.0428, "step": 27200 }, { "grad_norm": 0.4097718298435211, "learning_rate": 7.870120346871324e-05, "loss": 0.0407, "step": 27210 }, { "grad_norm": 0.45527762174606323, "learning_rate": 7.86842769592845e-05, "loss": 0.0344, "step": 27220 }, { "grad_norm": 0.4651847183704376, "learning_rate": 7.866734554849732e-05, "loss": 0.0317, "step": 27230 }, { "grad_norm": 0.31562164425849915, "learning_rate": 7.865040923924486e-05, "loss": 0.0412, "step": 27240 }, { "grad_norm": 0.4819819927215576, "learning_rate": 7.863346803442104e-05, "loss": 0.0446, "step": 27250 }, { "grad_norm": 0.5429366827011108, "learning_rate": 7.861652193692067e-05, "loss": 0.0363, "step": 27260 }, { "grad_norm": 0.5221893787384033, "learning_rate": 7.859957094963937e-05, "loss": 0.0337, "step": 27270 }, { "grad_norm": 0.3110267221927643, "learning_rate": 7.858261507547357e-05, "loss": 0.0329, "step": 27280 }, { "grad_norm": 0.33439135551452637, "learning_rate": 7.856565431732061e-05, "loss": 0.0323, "step": 27290 }, { "grad_norm": 0.2840114235877991, "learning_rate": 7.854868867807859e-05, "loss": 0.0295, "step": 27300 }, { "grad_norm": 0.4440622925758362, "learning_rate": 7.85317181606465e-05, "loss": 0.0435, "step": 27310 }, { "grad_norm": 0.3823755979537964, "learning_rate": 7.85147427679241e-05, "loss": 0.0301, "step": 27320 }, { "grad_norm": 0.2513831853866577, "learning_rate": 7.849776250281205e-05, "loss": 0.0319, "step": 27330 }, { "grad_norm": 0.4919097125530243, "learning_rate": 7.84807773682118e-05, "loss": 0.0376, "step": 27340 }, { "grad_norm": 0.3664979338645935, "learning_rate": 7.846378736702565e-05, "loss": 0.0428, "step": 27350 }, { "grad_norm": 0.3902571201324463, "learning_rate": 7.844679250215671e-05, "loss": 0.0425, "step": 27360 }, { "grad_norm": 0.27300530672073364, "learning_rate": 7.842979277650898e-05, "loss": 0.033, "step": 27370 }, { "grad_norm": 0.4049846827983856, "learning_rate": 7.84127881929872e-05, "loss": 0.0342, "step": 27380 }, { "grad_norm": 0.36748814582824707, "learning_rate": 7.839577875449704e-05, "loss": 0.0393, "step": 27390 }, { "grad_norm": 0.5243465304374695, "learning_rate": 7.837876446394489e-05, "loss": 0.0357, "step": 27400 }, { "grad_norm": 0.2853851616382599, "learning_rate": 7.836174532423805e-05, "loss": 0.0407, "step": 27410 }, { "grad_norm": 0.49522271752357483, "learning_rate": 7.834472133828466e-05, "loss": 0.0363, "step": 27420 }, { "grad_norm": 0.577283501625061, "learning_rate": 7.832769250899359e-05, "loss": 0.04, "step": 27430 }, { "grad_norm": 0.4521055221557617, "learning_rate": 7.831065883927464e-05, "loss": 0.0431, "step": 27440 }, { "grad_norm": 0.39449992775917053, "learning_rate": 7.829362033203841e-05, "loss": 0.036, "step": 27450 }, { "grad_norm": 0.5884475111961365, "learning_rate": 7.827657699019628e-05, "loss": 0.0402, "step": 27460 }, { "grad_norm": 0.5365828275680542, "learning_rate": 7.825952881666052e-05, "loss": 0.0509, "step": 27470 }, { "grad_norm": 0.5373572111129761, "learning_rate": 7.824247581434418e-05, "loss": 0.0431, "step": 27480 }, { "grad_norm": 0.32571855187416077, "learning_rate": 7.822541798616116e-05, "loss": 0.0416, "step": 27490 }, { "grad_norm": 0.47258636355400085, "learning_rate": 7.820835533502617e-05, "loss": 0.0431, "step": 27500 }, { "grad_norm": 0.5828981399536133, "learning_rate": 7.819128786385475e-05, "loss": 0.0401, "step": 27510 }, { "grad_norm": 0.3546864986419678, "learning_rate": 7.817421557556329e-05, "loss": 0.0425, "step": 27520 }, { "grad_norm": 0.4543490409851074, "learning_rate": 7.815713847306893e-05, "loss": 0.0492, "step": 27530 }, { "grad_norm": 0.33238860964775085, "learning_rate": 7.81400565592897e-05, "loss": 0.0333, "step": 27540 }, { "grad_norm": 0.33048224449157715, "learning_rate": 7.812296983714444e-05, "loss": 0.0414, "step": 27550 }, { "grad_norm": 0.27247297763824463, "learning_rate": 7.810587830955281e-05, "loss": 0.0304, "step": 27560 }, { "grad_norm": 0.4247877895832062, "learning_rate": 7.808878197943528e-05, "loss": 0.0368, "step": 27570 }, { "grad_norm": 0.4266870617866516, "learning_rate": 7.807168084971312e-05, "loss": 0.032, "step": 27580 }, { "grad_norm": 0.4773986339569092, "learning_rate": 7.805457492330849e-05, "loss": 0.0354, "step": 27590 }, { "grad_norm": 0.4190475642681122, "learning_rate": 7.803746420314428e-05, "loss": 0.0419, "step": 27600 }, { "grad_norm": 0.606280505657196, "learning_rate": 7.802034869214428e-05, "loss": 0.047, "step": 27610 }, { "grad_norm": 0.4584100842475891, "learning_rate": 7.800322839323303e-05, "loss": 0.0399, "step": 27620 }, { "grad_norm": 0.4940832257270813, "learning_rate": 7.798610330933593e-05, "loss": 0.0364, "step": 27630 }, { "grad_norm": 0.3631036877632141, "learning_rate": 7.796897344337922e-05, "loss": 0.0426, "step": 27640 }, { "grad_norm": 0.3555448353290558, "learning_rate": 7.795183879828989e-05, "loss": 0.0363, "step": 27650 }, { "grad_norm": 0.43291276693344116, "learning_rate": 7.793469937699579e-05, "loss": 0.05, "step": 27660 }, { "grad_norm": 0.3387986123561859, "learning_rate": 7.791755518242558e-05, "loss": 0.0421, "step": 27670 }, { "grad_norm": 0.47925451397895813, "learning_rate": 7.790040621750876e-05, "loss": 0.043, "step": 27680 }, { "grad_norm": 0.391903281211853, "learning_rate": 7.788325248517558e-05, "loss": 0.0342, "step": 27690 }, { "grad_norm": 0.3981603980064392, "learning_rate": 7.786609398835715e-05, "loss": 0.0375, "step": 27700 }, { "grad_norm": 0.34016019105911255, "learning_rate": 7.784893072998541e-05, "loss": 0.0369, "step": 27710 }, { "grad_norm": 0.3300769329071045, "learning_rate": 7.783176271299306e-05, "loss": 0.0389, "step": 27720 }, { "grad_norm": 0.785724937915802, "learning_rate": 7.781458994031368e-05, "loss": 0.0373, "step": 27730 }, { "grad_norm": 0.4652867615222931, "learning_rate": 7.779741241488161e-05, "loss": 0.0391, "step": 27740 }, { "grad_norm": 0.46242499351501465, "learning_rate": 7.7780230139632e-05, "loss": 0.0426, "step": 27750 }, { "grad_norm": 0.43583932518959045, "learning_rate": 7.776304311750087e-05, "loss": 0.0373, "step": 27760 }, { "grad_norm": 0.40438976883888245, "learning_rate": 7.7745851351425e-05, "loss": 0.0441, "step": 27770 }, { "grad_norm": 0.32588303089141846, "learning_rate": 7.772865484434197e-05, "loss": 0.0327, "step": 27780 }, { "grad_norm": 0.4512597620487213, "learning_rate": 7.77114535991902e-05, "loss": 0.0448, "step": 27790 }, { "grad_norm": 0.46730053424835205, "learning_rate": 7.769424761890893e-05, "loss": 0.0356, "step": 27800 }, { "grad_norm": 0.48889222741127014, "learning_rate": 7.767703690643817e-05, "loss": 0.037, "step": 27810 }, { "grad_norm": 0.2405378669500351, "learning_rate": 7.76598214647188e-05, "loss": 0.0315, "step": 27820 }, { "grad_norm": 0.3401450514793396, "learning_rate": 7.764260129669241e-05, "loss": 0.0325, "step": 27830 }, { "grad_norm": 0.29699158668518066, "learning_rate": 7.76253764053015e-05, "loss": 0.043, "step": 27840 }, { "grad_norm": 0.34951168298721313, "learning_rate": 7.760814679348932e-05, "loss": 0.0311, "step": 27850 }, { "grad_norm": 0.33480098843574524, "learning_rate": 7.759091246419992e-05, "loss": 0.0342, "step": 27860 }, { "grad_norm": 0.7634294629096985, "learning_rate": 7.75736734203782e-05, "loss": 0.0356, "step": 27870 }, { "grad_norm": 0.6557361483573914, "learning_rate": 7.755642966496985e-05, "loss": 0.0488, "step": 27880 }, { "grad_norm": 0.3760750889778137, "learning_rate": 7.753918120092132e-05, "loss": 0.0452, "step": 27890 }, { "grad_norm": 0.4133826494216919, "learning_rate": 7.752192803117993e-05, "loss": 0.0342, "step": 27900 }, { "grad_norm": 0.3725648522377014, "learning_rate": 7.750467015869377e-05, "loss": 0.0347, "step": 27910 }, { "grad_norm": 0.4061618447303772, "learning_rate": 7.748740758641174e-05, "loss": 0.0357, "step": 27920 }, { "grad_norm": 0.3291943669319153, "learning_rate": 7.74701403172835e-05, "loss": 0.0455, "step": 27930 }, { "grad_norm": 0.3040817677974701, "learning_rate": 7.745286835425962e-05, "loss": 0.037, "step": 27940 }, { "grad_norm": 0.4501684010028839, "learning_rate": 7.743559170029138e-05, "loss": 0.0427, "step": 27950 }, { "grad_norm": 0.3093554973602295, "learning_rate": 7.741831035833087e-05, "loss": 0.0416, "step": 27960 }, { "grad_norm": 0.410664826631546, "learning_rate": 7.740102433133102e-05, "loss": 0.0326, "step": 27970 }, { "grad_norm": 0.32457348704338074, "learning_rate": 7.738373362224553e-05, "loss": 0.0346, "step": 27980 }, { "grad_norm": 0.30974262952804565, "learning_rate": 7.73664382340289e-05, "loss": 0.0316, "step": 27990 }, { "grad_norm": 0.27337589859962463, "learning_rate": 7.734913816963647e-05, "loss": 0.0365, "step": 28000 }, { "grad_norm": 0.46555855870246887, "learning_rate": 7.73318334320243e-05, "loss": 0.0335, "step": 28010 }, { "grad_norm": 0.47715768218040466, "learning_rate": 7.731452402414934e-05, "loss": 0.0378, "step": 28020 }, { "grad_norm": 0.5230596661567688, "learning_rate": 7.729720994896928e-05, "loss": 0.0425, "step": 28030 }, { "grad_norm": 0.7002301812171936, "learning_rate": 7.727989120944262e-05, "loss": 0.0425, "step": 28040 }, { "grad_norm": 0.8640162348747253, "learning_rate": 7.726256780852865e-05, "loss": 0.0366, "step": 28050 }, { "grad_norm": 0.347044974565506, "learning_rate": 7.724523974918749e-05, "loss": 0.034, "step": 28060 }, { "grad_norm": 0.3585217297077179, "learning_rate": 7.722790703438002e-05, "loss": 0.0331, "step": 28070 }, { "grad_norm": 0.5204712152481079, "learning_rate": 7.72105696670679e-05, "loss": 0.0361, "step": 28080 }, { "grad_norm": 0.26951414346694946, "learning_rate": 7.719322765021364e-05, "loss": 0.0401, "step": 28090 }, { "grad_norm": 0.49338293075561523, "learning_rate": 7.717588098678051e-05, "loss": 0.0391, "step": 28100 }, { "grad_norm": 0.315275102853775, "learning_rate": 7.715852967973258e-05, "loss": 0.0422, "step": 28110 }, { "grad_norm": 0.4837130606174469, "learning_rate": 7.714117373203474e-05, "loss": 0.0463, "step": 28120 }, { "grad_norm": 0.39666134119033813, "learning_rate": 7.712381314665259e-05, "loss": 0.0382, "step": 28130 }, { "grad_norm": 0.4113397002220154, "learning_rate": 7.710644792655261e-05, "loss": 0.0338, "step": 28140 }, { "grad_norm": 0.5650933980941772, "learning_rate": 7.708907807470207e-05, "loss": 0.0456, "step": 28150 }, { "grad_norm": 0.4150216579437256, "learning_rate": 7.707170359406896e-05, "loss": 0.04, "step": 28160 }, { "grad_norm": 0.44273585081100464, "learning_rate": 7.705432448762213e-05, "loss": 0.0344, "step": 28170 }, { "grad_norm": 0.47570690512657166, "learning_rate": 7.703694075833117e-05, "loss": 0.0488, "step": 28180 }, { "grad_norm": 0.49563056230545044, "learning_rate": 7.70195524091665e-05, "loss": 0.0579, "step": 28190 }, { "grad_norm": 0.618242084980011, "learning_rate": 7.70021594430993e-05, "loss": 0.0474, "step": 28200 }, { "grad_norm": 0.3281741440296173, "learning_rate": 7.698476186310157e-05, "loss": 0.035, "step": 28210 }, { "grad_norm": 0.3958882987499237, "learning_rate": 7.696735967214608e-05, "loss": 0.0342, "step": 28220 }, { "grad_norm": 0.4129568040370941, "learning_rate": 7.694995287320636e-05, "loss": 0.031, "step": 28230 }, { "grad_norm": 0.3748399317264557, "learning_rate": 7.693254146925679e-05, "loss": 0.0343, "step": 28240 }, { "grad_norm": 0.49533259868621826, "learning_rate": 7.691512546327251e-05, "loss": 0.0443, "step": 28250 }, { "grad_norm": 0.32176467776298523, "learning_rate": 7.689770485822939e-05, "loss": 0.04, "step": 28260 }, { "grad_norm": 0.31426048278808594, "learning_rate": 7.688027965710416e-05, "loss": 0.0294, "step": 28270 }, { "grad_norm": 0.3533056974411011, "learning_rate": 7.686284986287433e-05, "loss": 0.0421, "step": 28280 }, { "grad_norm": 0.32459431886672974, "learning_rate": 7.684541547851817e-05, "loss": 0.0357, "step": 28290 }, { "grad_norm": 0.2739104628562927, "learning_rate": 7.68279765070147e-05, "loss": 0.0372, "step": 28300 }, { "grad_norm": 0.4060366451740265, "learning_rate": 7.68105329513438e-05, "loss": 0.0365, "step": 28310 }, { "grad_norm": 0.29416337609291077, "learning_rate": 7.67930848144861e-05, "loss": 0.0324, "step": 28320 }, { "grad_norm": 0.3609600365161896, "learning_rate": 7.6775632099423e-05, "loss": 0.0393, "step": 28330 }, { "grad_norm": 0.38119325041770935, "learning_rate": 7.675817480913667e-05, "loss": 0.035, "step": 28340 }, { "grad_norm": 0.42416709661483765, "learning_rate": 7.674071294661011e-05, "loss": 0.0353, "step": 28350 }, { "grad_norm": 0.2921368181705475, "learning_rate": 7.672324651482707e-05, "loss": 0.0327, "step": 28360 }, { "grad_norm": 0.5051478147506714, "learning_rate": 7.670577551677209e-05, "loss": 0.0416, "step": 28370 }, { "grad_norm": 0.40846648812294006, "learning_rate": 7.668829995543047e-05, "loss": 0.0314, "step": 28380 }, { "grad_norm": 0.3447922170162201, "learning_rate": 7.667081983378832e-05, "loss": 0.0371, "step": 28390 }, { "grad_norm": 0.3805209994316101, "learning_rate": 7.66533351548325e-05, "loss": 0.0318, "step": 28400 }, { "grad_norm": 1.5145738124847412, "learning_rate": 7.663584592155069e-05, "loss": 0.037, "step": 28410 }, { "grad_norm": 0.4750555157661438, "learning_rate": 7.661835213693129e-05, "loss": 0.0397, "step": 28420 }, { "grad_norm": 0.43354716897010803, "learning_rate": 7.660085380396353e-05, "loss": 0.0328, "step": 28430 }, { "grad_norm": 0.33547496795654297, "learning_rate": 7.658335092563738e-05, "loss": 0.0327, "step": 28440 }, { "grad_norm": 0.308672696352005, "learning_rate": 7.656584350494362e-05, "loss": 0.0301, "step": 28450 }, { "grad_norm": 0.4686298966407776, "learning_rate": 7.654833154487378e-05, "loss": 0.0372, "step": 28460 }, { "grad_norm": 0.29736077785491943, "learning_rate": 7.653081504842017e-05, "loss": 0.0321, "step": 28470 }, { "grad_norm": 0.4747299551963806, "learning_rate": 7.65132940185759e-05, "loss": 0.0388, "step": 28480 }, { "grad_norm": 0.3513573408126831, "learning_rate": 7.649576845833481e-05, "loss": 0.0328, "step": 28490 }, { "grad_norm": 0.36267074942588806, "learning_rate": 7.647823837069156e-05, "loss": 0.0462, "step": 28500 }, { "grad_norm": 0.4220057725906372, "learning_rate": 7.646070375864156e-05, "loss": 0.0337, "step": 28510 }, { "grad_norm": 0.3941207528114319, "learning_rate": 7.644316462518099e-05, "loss": 0.0362, "step": 28520 }, { "grad_norm": 0.6231436729431152, "learning_rate": 7.642562097330679e-05, "loss": 0.0407, "step": 28530 }, { "grad_norm": 0.35596171021461487, "learning_rate": 7.640807280601671e-05, "loss": 0.0393, "step": 28540 }, { "grad_norm": 0.3679351806640625, "learning_rate": 7.639052012630927e-05, "loss": 0.0363, "step": 28550 }, { "grad_norm": 0.6070947051048279, "learning_rate": 7.63729629371837e-05, "loss": 0.0445, "step": 28560 }, { "grad_norm": 0.46450719237327576, "learning_rate": 7.635540124164009e-05, "loss": 0.0443, "step": 28570 }, { "grad_norm": 0.29154637455940247, "learning_rate": 7.633783504267922e-05, "loss": 0.0365, "step": 28580 }, { "grad_norm": 0.45513513684272766, "learning_rate": 7.632026434330269e-05, "loss": 0.0374, "step": 28590 }, { "grad_norm": 0.4902134835720062, "learning_rate": 7.630268914651282e-05, "loss": 0.0345, "step": 28600 }, { "grad_norm": 0.4780738949775696, "learning_rate": 7.628510945531278e-05, "loss": 0.0462, "step": 28610 }, { "grad_norm": 0.25704425573349, "learning_rate": 7.626752527270641e-05, "loss": 0.0389, "step": 28620 }, { "grad_norm": 0.29970037937164307, "learning_rate": 7.62499366016984e-05, "loss": 0.0343, "step": 28630 }, { "grad_norm": 0.39846858382225037, "learning_rate": 7.623234344529416e-05, "loss": 0.0379, "step": 28640 }, { "grad_norm": 0.40413519740104675, "learning_rate": 7.62147458064999e-05, "loss": 0.0406, "step": 28650 }, { "grad_norm": 0.5020326375961304, "learning_rate": 7.619714368832254e-05, "loss": 0.0392, "step": 28660 }, { "grad_norm": 0.564156711101532, "learning_rate": 7.61795370937698e-05, "loss": 0.032, "step": 28670 }, { "grad_norm": 0.4247159957885742, "learning_rate": 7.61619260258502e-05, "loss": 0.0392, "step": 28680 }, { "grad_norm": 0.35513055324554443, "learning_rate": 7.614431048757298e-05, "loss": 0.0439, "step": 28690 }, { "grad_norm": 0.43561333417892456, "learning_rate": 7.612669048194814e-05, "loss": 0.0405, "step": 28700 }, { "grad_norm": 0.3424459993839264, "learning_rate": 7.610906601198646e-05, "loss": 0.0341, "step": 28710 }, { "grad_norm": 0.4531015455722809, "learning_rate": 7.60914370806995e-05, "loss": 0.0454, "step": 28720 }, { "grad_norm": 0.563332736492157, "learning_rate": 7.607380369109953e-05, "loss": 0.0596, "step": 28730 }, { "grad_norm": 0.3120252192020416, "learning_rate": 7.605616584619961e-05, "loss": 0.0331, "step": 28740 }, { "grad_norm": 0.42230933904647827, "learning_rate": 7.603852354901362e-05, "loss": 0.044, "step": 28750 }, { "grad_norm": 0.34505823254585266, "learning_rate": 7.602087680255609e-05, "loss": 0.05, "step": 28760 }, { "grad_norm": 0.4922657012939453, "learning_rate": 7.600322560984238e-05, "loss": 0.0344, "step": 28770 }, { "grad_norm": 0.6885706186294556, "learning_rate": 7.598556997388863e-05, "loss": 0.0413, "step": 28780 }, { "grad_norm": 0.34292367100715637, "learning_rate": 7.596790989771166e-05, "loss": 0.0403, "step": 28790 }, { "grad_norm": 0.4591211676597595, "learning_rate": 7.595024538432914e-05, "loss": 0.0356, "step": 28800 }, { "grad_norm": 0.36403313279151917, "learning_rate": 7.59325764367594e-05, "loss": 0.0356, "step": 28810 }, { "grad_norm": 0.40710943937301636, "learning_rate": 7.59149030580216e-05, "loss": 0.0453, "step": 28820 }, { "grad_norm": 0.2747039496898651, "learning_rate": 7.589722525113562e-05, "loss": 0.0359, "step": 28830 }, { "grad_norm": 0.4908159375190735, "learning_rate": 7.587954301912216e-05, "loss": 0.0375, "step": 28840 }, { "grad_norm": 0.4654216468334198, "learning_rate": 7.586185636500263e-05, "loss": 0.0429, "step": 28850 }, { "grad_norm": 0.4941474497318268, "learning_rate": 7.584416529179914e-05, "loss": 0.0443, "step": 28860 }, { "grad_norm": 0.40078940987586975, "learning_rate": 7.582646980253465e-05, "loss": 0.035, "step": 28870 }, { "grad_norm": 0.3124418258666992, "learning_rate": 7.580876990023282e-05, "loss": 0.0421, "step": 28880 }, { "grad_norm": 0.3996962606906891, "learning_rate": 7.579106558791809e-05, "loss": 0.0314, "step": 28890 }, { "grad_norm": 0.3695491850376129, "learning_rate": 7.577335686861565e-05, "loss": 0.0395, "step": 28900 }, { "grad_norm": 0.5721259117126465, "learning_rate": 7.575564374535141e-05, "loss": 0.0419, "step": 28910 }, { "grad_norm": 0.3551567494869232, "learning_rate": 7.573792622115207e-05, "loss": 0.0301, "step": 28920 }, { "grad_norm": 0.34988272190093994, "learning_rate": 7.572020429904507e-05, "loss": 0.0343, "step": 28930 }, { "grad_norm": 0.3992784023284912, "learning_rate": 7.570247798205861e-05, "loss": 0.0418, "step": 28940 }, { "grad_norm": 0.2883925139904022, "learning_rate": 7.568474727322164e-05, "loss": 0.0411, "step": 28950 }, { "grad_norm": 0.2564494013786316, "learning_rate": 7.566701217556384e-05, "loss": 0.0351, "step": 28960 }, { "grad_norm": 0.37056678533554077, "learning_rate": 7.564927269211564e-05, "loss": 0.0373, "step": 28970 }, { "grad_norm": 0.44178497791290283, "learning_rate": 7.563152882590824e-05, "loss": 0.0391, "step": 28980 }, { "grad_norm": 0.3018684983253479, "learning_rate": 7.56137805799736e-05, "loss": 0.0379, "step": 28990 }, { "grad_norm": 0.5553391575813293, "learning_rate": 7.559602795734439e-05, "loss": 0.0379, "step": 29000 }, { "grad_norm": 0.2709180414676666, "learning_rate": 7.557827096105408e-05, "loss": 0.0288, "step": 29010 }, { "grad_norm": 0.3714272081851959, "learning_rate": 7.55605095941368e-05, "loss": 0.0348, "step": 29020 }, { "grad_norm": 0.3675384819507599, "learning_rate": 7.554274385962752e-05, "loss": 0.0397, "step": 29030 }, { "grad_norm": 0.4609202444553375, "learning_rate": 7.552497376056191e-05, "loss": 0.0379, "step": 29040 }, { "grad_norm": 0.3076239824295044, "learning_rate": 7.550719929997639e-05, "loss": 0.0389, "step": 29050 }, { "grad_norm": 0.5705327987670898, "learning_rate": 7.548942048090813e-05, "loss": 0.0344, "step": 29060 }, { "grad_norm": 0.3925672173500061, "learning_rate": 7.547163730639506e-05, "loss": 0.0409, "step": 29070 }, { "grad_norm": 0.4535709619522095, "learning_rate": 7.545384977947583e-05, "loss": 0.0346, "step": 29080 }, { "grad_norm": 0.37799936532974243, "learning_rate": 7.543605790318981e-05, "loss": 0.0361, "step": 29090 }, { "grad_norm": 0.5880883932113647, "learning_rate": 7.54182616805772e-05, "loss": 0.0453, "step": 29100 }, { "grad_norm": 0.32686132192611694, "learning_rate": 7.540046111467885e-05, "loss": 0.0342, "step": 29110 }, { "grad_norm": 0.3074241280555725, "learning_rate": 7.53826562085364e-05, "loss": 0.0359, "step": 29120 }, { "grad_norm": 0.5007890462875366, "learning_rate": 7.536484696519221e-05, "loss": 0.0309, "step": 29130 }, { "grad_norm": 0.33406147360801697, "learning_rate": 7.534703338768942e-05, "loss": 0.0303, "step": 29140 }, { "grad_norm": 0.43587568402290344, "learning_rate": 7.532921547907185e-05, "loss": 0.038, "step": 29150 }, { "grad_norm": 0.3282313942909241, "learning_rate": 7.531139324238412e-05, "loss": 0.0412, "step": 29160 }, { "grad_norm": 0.40551432967185974, "learning_rate": 7.529356668067157e-05, "loss": 0.0489, "step": 29170 }, { "grad_norm": 0.37783947587013245, "learning_rate": 7.527573579698023e-05, "loss": 0.0382, "step": 29180 }, { "grad_norm": 0.3248288929462433, "learning_rate": 7.525790059435693e-05, "loss": 0.0375, "step": 29190 }, { "grad_norm": 0.4557148814201355, "learning_rate": 7.524006107584926e-05, "loss": 0.033, "step": 29200 }, { "grad_norm": 0.4347898066043854, "learning_rate": 7.522221724450544e-05, "loss": 0.0348, "step": 29210 }, { "grad_norm": 0.3682006597518921, "learning_rate": 7.520436910337451e-05, "loss": 0.0459, "step": 29220 }, { "grad_norm": 0.5035701394081116, "learning_rate": 7.518651665550627e-05, "loss": 0.0325, "step": 29230 }, { "grad_norm": 0.32628175616264343, "learning_rate": 7.516865990395117e-05, "loss": 0.0397, "step": 29240 }, { "grad_norm": 0.36347150802612305, "learning_rate": 7.515079885176047e-05, "loss": 0.0348, "step": 29250 }, { "grad_norm": 0.4538796544075012, "learning_rate": 7.513293350198612e-05, "loss": 0.036, "step": 29260 }, { "grad_norm": 0.41752558946609497, "learning_rate": 7.511506385768081e-05, "loss": 0.0346, "step": 29270 }, { "grad_norm": 0.5125117897987366, "learning_rate": 7.509718992189801e-05, "loss": 0.0422, "step": 29280 }, { "grad_norm": 0.4620459973812103, "learning_rate": 7.507931169769182e-05, "loss": 0.0409, "step": 29290 }, { "grad_norm": 0.5224756598472595, "learning_rate": 7.506142918811722e-05, "loss": 0.0412, "step": 29300 }, { "grad_norm": 0.30921801924705505, "learning_rate": 7.504354239622978e-05, "loss": 0.0444, "step": 29310 }, { "grad_norm": 0.34262847900390625, "learning_rate": 7.50256513250859e-05, "loss": 0.0343, "step": 29320 }, { "grad_norm": 0.2779839336872101, "learning_rate": 7.500775597774265e-05, "loss": 0.0336, "step": 29330 }, { "grad_norm": 0.487194687128067, "learning_rate": 7.498985635725788e-05, "loss": 0.0312, "step": 29340 }, { "grad_norm": 0.41995635628700256, "learning_rate": 7.497195246669012e-05, "loss": 0.0354, "step": 29350 }, { "grad_norm": 0.5077736377716064, "learning_rate": 7.495404430909868e-05, "loss": 0.0386, "step": 29360 }, { "grad_norm": 0.5293236374855042, "learning_rate": 7.493613188754356e-05, "loss": 0.044, "step": 29370 }, { "grad_norm": 0.5716099143028259, "learning_rate": 7.49182152050855e-05, "loss": 0.0446, "step": 29380 }, { "grad_norm": 0.6899645924568176, "learning_rate": 7.490029426478598e-05, "loss": 0.0359, "step": 29390 }, { "grad_norm": 0.3307837247848511, "learning_rate": 7.488236906970719e-05, "loss": 0.0407, "step": 29400 }, { "grad_norm": 1.3327211141586304, "learning_rate": 7.486443962291207e-05, "loss": 0.0416, "step": 29410 }, { "grad_norm": 0.3036181628704071, "learning_rate": 7.484650592746424e-05, "loss": 0.0315, "step": 29420 }, { "grad_norm": 0.31023213267326355, "learning_rate": 7.482856798642811e-05, "loss": 0.0406, "step": 29430 }, { "grad_norm": 0.3197069764137268, "learning_rate": 7.481062580286878e-05, "loss": 0.0362, "step": 29440 }, { "grad_norm": 0.29768475890159607, "learning_rate": 7.479267937985208e-05, "loss": 0.0321, "step": 29450 }, { "grad_norm": 0.30940037965774536, "learning_rate": 7.477472872044456e-05, "loss": 0.0282, "step": 29460 }, { "grad_norm": 0.3612639009952545, "learning_rate": 7.475677382771347e-05, "loss": 0.0384, "step": 29470 }, { "grad_norm": 0.40537476539611816, "learning_rate": 7.473881470472683e-05, "loss": 0.0409, "step": 29480 }, { "grad_norm": 0.5707420706748962, "learning_rate": 7.472085135455337e-05, "loss": 0.0353, "step": 29490 }, { "grad_norm": 0.27249783277511597, "learning_rate": 7.470288378026256e-05, "loss": 0.0406, "step": 29500 }, { "grad_norm": 0.3394007384777069, "learning_rate": 7.468491198492451e-05, "loss": 0.0299, "step": 29510 }, { "grad_norm": 0.40895023941993713, "learning_rate": 7.466693597161013e-05, "loss": 0.0288, "step": 29520 }, { "grad_norm": 0.3364728093147278, "learning_rate": 7.464895574339104e-05, "loss": 0.0373, "step": 29530 }, { "grad_norm": 0.34544697403907776, "learning_rate": 7.463097130333958e-05, "loss": 0.0438, "step": 29540 }, { "grad_norm": 0.28057876229286194, "learning_rate": 7.461298265452876e-05, "loss": 0.04, "step": 29550 }, { "grad_norm": 0.4132610559463501, "learning_rate": 7.459498980003239e-05, "loss": 0.0312, "step": 29560 }, { "grad_norm": 0.46794435381889343, "learning_rate": 7.457699274292493e-05, "loss": 0.0445, "step": 29570 }, { "grad_norm": 0.27438709139823914, "learning_rate": 7.455899148628159e-05, "loss": 0.0434, "step": 29580 }, { "grad_norm": 0.30406633019447327, "learning_rate": 7.45409860331783e-05, "loss": 0.0315, "step": 29590 }, { "grad_norm": 0.4006156921386719, "learning_rate": 7.452297638669169e-05, "loss": 0.0324, "step": 29600 }, { "grad_norm": 0.31660202145576477, "learning_rate": 7.450496254989911e-05, "loss": 0.0268, "step": 29610 }, { "grad_norm": 0.40305083990097046, "learning_rate": 7.448694452587866e-05, "loss": 0.0339, "step": 29620 }, { "grad_norm": 0.4720960855484009, "learning_rate": 7.44689223177091e-05, "loss": 0.0403, "step": 29630 }, { "grad_norm": 0.3938717842102051, "learning_rate": 7.445089592846994e-05, "loss": 0.0362, "step": 29640 }, { "grad_norm": 0.4164455235004425, "learning_rate": 7.443286536124141e-05, "loss": 0.0405, "step": 29650 }, { "grad_norm": 0.34746816754341125, "learning_rate": 7.441483061910443e-05, "loss": 0.0432, "step": 29660 }, { "grad_norm": 0.4077763557434082, "learning_rate": 7.439679170514064e-05, "loss": 0.0419, "step": 29670 }, { "grad_norm": 0.4958491921424866, "learning_rate": 7.43787486224324e-05, "loss": 0.0377, "step": 29680 }, { "grad_norm": 0.3651946485042572, "learning_rate": 7.436070137406276e-05, "loss": 0.0345, "step": 29690 }, { "grad_norm": 0.3660089671611786, "learning_rate": 7.434264996311556e-05, "loss": 0.0317, "step": 29700 }, { "grad_norm": 0.6687074303627014, "learning_rate": 7.432459439267525e-05, "loss": 0.0409, "step": 29710 }, { "grad_norm": 0.3627680838108063, "learning_rate": 7.430653466582701e-05, "loss": 0.0335, "step": 29720 }, { "grad_norm": 0.4389142394065857, "learning_rate": 7.42884707856568e-05, "loss": 0.0515, "step": 29730 }, { "grad_norm": 0.5049268007278442, "learning_rate": 7.427040275525122e-05, "loss": 0.0363, "step": 29740 }, { "grad_norm": 0.33122238516807556, "learning_rate": 7.42523305776976e-05, "loss": 0.0362, "step": 29750 }, { "grad_norm": 0.4987010359764099, "learning_rate": 7.4234254256084e-05, "loss": 0.0362, "step": 29760 }, { "grad_norm": 0.44653406739234924, "learning_rate": 7.421617379349915e-05, "loss": 0.0372, "step": 29770 }, { "grad_norm": 0.3917689025402069, "learning_rate": 7.41980891930325e-05, "loss": 0.0312, "step": 29780 }, { "grad_norm": 0.3310330808162689, "learning_rate": 7.418000045777425e-05, "loss": 0.0266, "step": 29790 }, { "grad_norm": 0.291337251663208, "learning_rate": 7.416190759081523e-05, "loss": 0.0289, "step": 29800 }, { "grad_norm": 0.32279884815216064, "learning_rate": 7.414381059524704e-05, "loss": 0.0267, "step": 29810 }, { "grad_norm": 0.3264009356498718, "learning_rate": 7.412570947416195e-05, "loss": 0.029, "step": 29820 }, { "grad_norm": 1.0193150043487549, "learning_rate": 7.410760423065295e-05, "loss": 0.0494, "step": 29830 }, { "grad_norm": 0.49816250801086426, "learning_rate": 7.408949486781372e-05, "loss": 0.045, "step": 29840 }, { "grad_norm": 0.6784267425537109, "learning_rate": 7.407138138873868e-05, "loss": 0.0447, "step": 29850 }, { "grad_norm": 0.47710534930229187, "learning_rate": 7.405326379652292e-05, "loss": 0.0391, "step": 29860 }, { "grad_norm": 0.36474141478538513, "learning_rate": 7.403514209426222e-05, "loss": 0.0343, "step": 29870 }, { "grad_norm": 0.5118702054023743, "learning_rate": 7.40170162850531e-05, "loss": 0.0398, "step": 29880 }, { "grad_norm": 0.39116355776786804, "learning_rate": 7.399888637199278e-05, "loss": 0.0303, "step": 29890 }, { "grad_norm": 0.35389211773872375, "learning_rate": 7.398075235817914e-05, "loss": 0.0513, "step": 29900 }, { "grad_norm": 0.655318021774292, "learning_rate": 7.39626142467108e-05, "loss": 0.0421, "step": 29910 }, { "grad_norm": 0.3186022937297821, "learning_rate": 7.394447204068706e-05, "loss": 0.0305, "step": 29920 }, { "grad_norm": 0.4829877018928528, "learning_rate": 7.392632574320793e-05, "loss": 0.0486, "step": 29930 }, { "grad_norm": 0.32852330803871155, "learning_rate": 7.390817535737411e-05, "loss": 0.0415, "step": 29940 }, { "grad_norm": 0.2675226628780365, "learning_rate": 7.389002088628703e-05, "loss": 0.0362, "step": 29950 }, { "grad_norm": 0.5385065078735352, "learning_rate": 7.387186233304877e-05, "loss": 0.0401, "step": 29960 }, { "grad_norm": 0.5945344567298889, "learning_rate": 7.385369970076212e-05, "loss": 0.0349, "step": 29970 }, { "grad_norm": 0.4838988780975342, "learning_rate": 7.38355329925306e-05, "loss": 0.0381, "step": 29980 }, { "grad_norm": 0.42937299609184265, "learning_rate": 7.381736221145838e-05, "loss": 0.0385, "step": 29990 }, { "grad_norm": 0.394004762172699, "learning_rate": 7.37991873606504e-05, "loss": 0.037, "step": 30000 }, { "grad_norm": 0.419651597738266, "learning_rate": 7.378100844321218e-05, "loss": 0.0398, "step": 30010 }, { "grad_norm": 0.29175639152526855, "learning_rate": 7.376282546225004e-05, "loss": 0.0342, "step": 30020 }, { "grad_norm": 0.31953346729278564, "learning_rate": 7.374463842087094e-05, "loss": 0.0268, "step": 30030 }, { "grad_norm": 0.418753981590271, "learning_rate": 7.372644732218254e-05, "loss": 0.035, "step": 30040 }, { "grad_norm": 0.45080846548080444, "learning_rate": 7.370825216929322e-05, "loss": 0.0508, "step": 30050 }, { "grad_norm": 0.3457571864128113, "learning_rate": 7.369005296531205e-05, "loss": 0.0375, "step": 30060 }, { "grad_norm": 0.3579409718513489, "learning_rate": 7.367184971334873e-05, "loss": 0.038, "step": 30070 }, { "grad_norm": 0.3781866133213043, "learning_rate": 7.365364241651371e-05, "loss": 0.0318, "step": 30080 }, { "grad_norm": 0.41527259349823, "learning_rate": 7.363543107791815e-05, "loss": 0.0327, "step": 30090 }, { "grad_norm": 0.4061008393764496, "learning_rate": 7.361721570067384e-05, "loss": 0.03, "step": 30100 }, { "grad_norm": 0.4045584499835968, "learning_rate": 7.359899628789331e-05, "loss": 0.0384, "step": 30110 }, { "grad_norm": 0.320324569940567, "learning_rate": 7.358077284268974e-05, "loss": 0.0357, "step": 30120 }, { "grad_norm": 0.4156014323234558, "learning_rate": 7.356254536817702e-05, "loss": 0.0351, "step": 30130 }, { "grad_norm": 0.48597049713134766, "learning_rate": 7.354431386746973e-05, "loss": 0.0385, "step": 30140 }, { "grad_norm": 0.45345935225486755, "learning_rate": 7.352607834368316e-05, "loss": 0.0395, "step": 30150 }, { "grad_norm": 0.3337574601173401, "learning_rate": 7.350783879993324e-05, "loss": 0.0315, "step": 30160 }, { "grad_norm": 0.48009008169174194, "learning_rate": 7.348959523933658e-05, "loss": 0.0435, "step": 30170 }, { "grad_norm": 0.35707181692123413, "learning_rate": 7.347134766501057e-05, "loss": 0.0265, "step": 30180 }, { "grad_norm": 0.6865251064300537, "learning_rate": 7.345309608007315e-05, "loss": 0.0343, "step": 30190 }, { "grad_norm": 0.3671325743198395, "learning_rate": 7.343484048764308e-05, "loss": 0.0331, "step": 30200 }, { "grad_norm": 0.4520847797393799, "learning_rate": 7.341658089083972e-05, "loss": 0.0353, "step": 30210 }, { "grad_norm": 0.4257137179374695, "learning_rate": 7.339831729278313e-05, "loss": 0.0295, "step": 30220 }, { "grad_norm": 0.3234575688838959, "learning_rate": 7.338004969659404e-05, "loss": 0.0364, "step": 30230 }, { "grad_norm": 0.40108197927474976, "learning_rate": 7.336177810539391e-05, "loss": 0.0446, "step": 30240 }, { "grad_norm": 0.4964539110660553, "learning_rate": 7.334350252230485e-05, "loss": 0.0325, "step": 30250 }, { "grad_norm": 0.33138683438301086, "learning_rate": 7.332522295044965e-05, "loss": 0.0301, "step": 30260 }, { "grad_norm": 0.3796243965625763, "learning_rate": 7.33069393929518e-05, "loss": 0.0405, "step": 30270 }, { "grad_norm": 0.41732752323150635, "learning_rate": 7.328865185293545e-05, "loss": 0.0489, "step": 30280 }, { "grad_norm": 0.3600330352783203, "learning_rate": 7.327036033352546e-05, "loss": 0.0332, "step": 30290 }, { "grad_norm": 0.3820914924144745, "learning_rate": 7.325206483784733e-05, "loss": 0.0297, "step": 30300 }, { "grad_norm": 0.28308719396591187, "learning_rate": 7.323376536902724e-05, "loss": 0.0435, "step": 30310 }, { "grad_norm": 0.3684673011302948, "learning_rate": 7.321546193019213e-05, "loss": 0.035, "step": 30320 }, { "grad_norm": 0.4501417279243469, "learning_rate": 7.31971545244695e-05, "loss": 0.0374, "step": 30330 }, { "grad_norm": 0.36679092049598694, "learning_rate": 7.31788431549876e-05, "loss": 0.0313, "step": 30340 }, { "grad_norm": 0.43286648392677307, "learning_rate": 7.316052782487534e-05, "loss": 0.0283, "step": 30350 }, { "grad_norm": 0.4734741747379303, "learning_rate": 7.314220853726234e-05, "loss": 0.0454, "step": 30360 }, { "grad_norm": 0.4641866087913513, "learning_rate": 7.312388529527884e-05, "loss": 0.033, "step": 30370 }, { "grad_norm": 0.42445048689842224, "learning_rate": 7.310555810205577e-05, "loss": 0.0357, "step": 30380 }, { "grad_norm": 0.24524350464344025, "learning_rate": 7.308722696072476e-05, "loss": 0.038, "step": 30390 }, { "grad_norm": 0.6128733158111572, "learning_rate": 7.306889187441811e-05, "loss": 0.0379, "step": 30400 }, { "grad_norm": 0.3403359353542328, "learning_rate": 7.305055284626876e-05, "loss": 0.0295, "step": 30410 }, { "grad_norm": 0.41819342970848083, "learning_rate": 7.303220987941037e-05, "loss": 0.0326, "step": 30420 }, { "grad_norm": 0.2877107560634613, "learning_rate": 7.301386297697726e-05, "loss": 0.032, "step": 30430 }, { "grad_norm": 0.4998900592327118, "learning_rate": 7.299551214210438e-05, "loss": 0.0325, "step": 30440 }, { "grad_norm": 0.2920784056186676, "learning_rate": 7.297715737792738e-05, "loss": 0.0324, "step": 30450 }, { "grad_norm": 0.682249128818512, "learning_rate": 7.295879868758265e-05, "loss": 0.0409, "step": 30460 }, { "grad_norm": 0.30541330575942993, "learning_rate": 7.294043607420713e-05, "loss": 0.0285, "step": 30470 }, { "grad_norm": 0.35394832491874695, "learning_rate": 7.292206954093852e-05, "loss": 0.0412, "step": 30480 }, { "grad_norm": 0.4128326177597046, "learning_rate": 7.290369909091515e-05, "loss": 0.033, "step": 30490 }, { "grad_norm": 0.2863662540912628, "learning_rate": 7.2885324727276e-05, "loss": 0.0379, "step": 30500 }, { "grad_norm": 0.4889232814311981, "learning_rate": 7.286694645316076e-05, "loss": 0.0361, "step": 30510 }, { "grad_norm": 0.4663173258304596, "learning_rate": 7.284856427170982e-05, "loss": 0.0302, "step": 30520 }, { "grad_norm": 0.35872331261634827, "learning_rate": 7.283017818606414e-05, "loss": 0.0387, "step": 30530 }, { "grad_norm": 1.332293152809143, "learning_rate": 7.28117881993654e-05, "loss": 0.0395, "step": 30540 }, { "grad_norm": 0.24854226410388947, "learning_rate": 7.279339431475598e-05, "loss": 0.0302, "step": 30550 }, { "grad_norm": 0.3330562710762024, "learning_rate": 7.277499653537887e-05, "loss": 0.0334, "step": 30560 }, { "grad_norm": 0.4191920757293701, "learning_rate": 7.275659486437776e-05, "loss": 0.0394, "step": 30570 }, { "grad_norm": 0.3813830316066742, "learning_rate": 7.273818930489695e-05, "loss": 0.0443, "step": 30580 }, { "grad_norm": 0.3663913607597351, "learning_rate": 7.271977986008151e-05, "loss": 0.0314, "step": 30590 }, { "grad_norm": 0.29492852091789246, "learning_rate": 7.270136653307705e-05, "loss": 0.0354, "step": 30600 }, { "grad_norm": 0.396356463432312, "learning_rate": 7.268294932702994e-05, "loss": 0.0346, "step": 30610 }, { "grad_norm": 0.44570842385292053, "learning_rate": 7.266452824508719e-05, "loss": 0.0391, "step": 30620 }, { "grad_norm": 0.260729044675827, "learning_rate": 7.264610329039643e-05, "loss": 0.0309, "step": 30630 }, { "grad_norm": 0.43839535117149353, "learning_rate": 7.262767446610599e-05, "loss": 0.0336, "step": 30640 }, { "grad_norm": 0.5843912959098816, "learning_rate": 7.260924177536485e-05, "loss": 0.0407, "step": 30650 }, { "grad_norm": 0.488546758890152, "learning_rate": 7.259080522132265e-05, "loss": 0.0331, "step": 30660 }, { "grad_norm": 0.34152480959892273, "learning_rate": 7.257236480712972e-05, "loss": 0.0321, "step": 30670 }, { "grad_norm": 0.4199368953704834, "learning_rate": 7.255392053593697e-05, "loss": 0.035, "step": 30680 }, { "grad_norm": 0.5073906183242798, "learning_rate": 7.253547241089607e-05, "loss": 0.0323, "step": 30690 }, { "grad_norm": 0.38315361738204956, "learning_rate": 7.251702043515927e-05, "loss": 0.0431, "step": 30700 }, { "grad_norm": 0.34607383608818054, "learning_rate": 7.249856461187952e-05, "loss": 0.0305, "step": 30710 }, { "grad_norm": 0.3323642611503601, "learning_rate": 7.248010494421042e-05, "loss": 0.034, "step": 30720 }, { "grad_norm": 0.32081863284111023, "learning_rate": 7.246164143530622e-05, "loss": 0.0464, "step": 30730 }, { "grad_norm": 0.2618185877799988, "learning_rate": 7.244317408832181e-05, "loss": 0.0354, "step": 30740 }, { "grad_norm": 0.2968282401561737, "learning_rate": 7.242470290641279e-05, "loss": 0.0348, "step": 30750 }, { "grad_norm": 0.6119416952133179, "learning_rate": 7.240622789273536e-05, "loss": 0.0347, "step": 30760 }, { "grad_norm": 0.30034735798835754, "learning_rate": 7.238774905044638e-05, "loss": 0.0335, "step": 30770 }, { "grad_norm": 0.37573400139808655, "learning_rate": 7.236926638270341e-05, "loss": 0.0402, "step": 30780 }, { "grad_norm": 0.3779957890510559, "learning_rate": 7.23507798926646e-05, "loss": 0.0465, "step": 30790 }, { "grad_norm": 0.2646234631538391, "learning_rate": 7.23322895834888e-05, "loss": 0.0364, "step": 30800 }, { "grad_norm": 0.271578311920166, "learning_rate": 7.231379545833552e-05, "loss": 0.0335, "step": 30810 }, { "grad_norm": 0.2931413948535919, "learning_rate": 7.229529752036487e-05, "loss": 0.0378, "step": 30820 }, { "grad_norm": 0.36471956968307495, "learning_rate": 7.227679577273765e-05, "loss": 0.0383, "step": 30830 }, { "grad_norm": 0.4084446132183075, "learning_rate": 7.225829021861529e-05, "loss": 0.0366, "step": 30840 }, { "grad_norm": 0.3542843163013458, "learning_rate": 7.223978086115992e-05, "loss": 0.0281, "step": 30850 }, { "grad_norm": 0.29522705078125, "learning_rate": 7.222126770353425e-05, "loss": 0.0388, "step": 30860 }, { "grad_norm": 0.3762660622596741, "learning_rate": 7.22027507489017e-05, "loss": 0.0316, "step": 30870 }, { "grad_norm": 0.36872249841690063, "learning_rate": 7.218423000042627e-05, "loss": 0.0295, "step": 30880 }, { "grad_norm": 0.3605111539363861, "learning_rate": 7.216570546127268e-05, "loss": 0.0353, "step": 30890 }, { "grad_norm": 0.5026079416275024, "learning_rate": 7.214717713460626e-05, "loss": 0.0409, "step": 30900 }, { "grad_norm": 0.37496939301490784, "learning_rate": 7.2128645023593e-05, "loss": 0.0393, "step": 30910 }, { "grad_norm": 0.43468624353408813, "learning_rate": 7.211010913139951e-05, "loss": 0.0403, "step": 30920 }, { "grad_norm": 0.37285736203193665, "learning_rate": 7.209156946119308e-05, "loss": 0.0337, "step": 30930 }, { "grad_norm": 0.4269050657749176, "learning_rate": 7.207302601614166e-05, "loss": 0.0334, "step": 30940 }, { "grad_norm": 0.4367576241493225, "learning_rate": 7.205447879941378e-05, "loss": 0.038, "step": 30950 }, { "grad_norm": 0.36003750562667847, "learning_rate": 7.203592781417866e-05, "loss": 0.0292, "step": 30960 }, { "grad_norm": 0.3972064256668091, "learning_rate": 7.201737306360617e-05, "loss": 0.0349, "step": 30970 }, { "grad_norm": 0.3258184492588043, "learning_rate": 7.19988145508668e-05, "loss": 0.0312, "step": 30980 }, { "grad_norm": 0.38541534543037415, "learning_rate": 7.198025227913168e-05, "loss": 0.0358, "step": 30990 }, { "grad_norm": 0.4499596655368805, "learning_rate": 7.196168625157261e-05, "loss": 0.0366, "step": 31000 }, { "grad_norm": 0.3831482231616974, "learning_rate": 7.194311647136201e-05, "loss": 0.034, "step": 31010 }, { "grad_norm": 0.4915318489074707, "learning_rate": 7.192454294167297e-05, "loss": 0.0301, "step": 31020 }, { "grad_norm": 0.40846946835517883, "learning_rate": 7.190596566567917e-05, "loss": 0.0369, "step": 31030 }, { "grad_norm": 0.45506805181503296, "learning_rate": 7.188738464655496e-05, "loss": 0.0392, "step": 31040 }, { "grad_norm": 0.25226929783821106, "learning_rate": 7.186879988747533e-05, "loss": 0.0316, "step": 31050 }, { "grad_norm": 0.37878888845443726, "learning_rate": 7.185021139161592e-05, "loss": 0.028, "step": 31060 }, { "grad_norm": 0.5236983299255371, "learning_rate": 7.1831619162153e-05, "loss": 0.0291, "step": 31070 }, { "grad_norm": 0.3670037090778351, "learning_rate": 7.181302320226345e-05, "loss": 0.0304, "step": 31080 }, { "grad_norm": 0.3122554123401642, "learning_rate": 7.179442351512482e-05, "loss": 0.0406, "step": 31090 }, { "grad_norm": 0.36126765608787537, "learning_rate": 7.177582010391528e-05, "loss": 0.0336, "step": 31100 }, { "grad_norm": 0.44540145993232727, "learning_rate": 7.175721297181366e-05, "loss": 0.0315, "step": 31110 }, { "grad_norm": 0.3820830285549164, "learning_rate": 7.173860212199942e-05, "loss": 0.0321, "step": 31120 }, { "grad_norm": 0.3152284324169159, "learning_rate": 7.171998755765263e-05, "loss": 0.0326, "step": 31130 }, { "grad_norm": 0.5951884388923645, "learning_rate": 7.170136928195398e-05, "loss": 0.0474, "step": 31140 }, { "grad_norm": 0.4092089831829071, "learning_rate": 7.168274729808489e-05, "loss": 0.0357, "step": 31150 }, { "grad_norm": 0.26811283826828003, "learning_rate": 7.166412160922728e-05, "loss": 0.0389, "step": 31160 }, { "grad_norm": 0.3116966485977173, "learning_rate": 7.164549221856382e-05, "loss": 0.0401, "step": 31170 }, { "grad_norm": 0.3778054714202881, "learning_rate": 7.162685912927775e-05, "loss": 0.0441, "step": 31180 }, { "grad_norm": 0.38452932238578796, "learning_rate": 7.160822234455294e-05, "loss": 0.0334, "step": 31190 }, { "grad_norm": 0.4189922511577606, "learning_rate": 7.158958186757391e-05, "loss": 0.0276, "step": 31200 }, { "grad_norm": 0.4063712954521179, "learning_rate": 7.157093770152582e-05, "loss": 0.0334, "step": 31210 }, { "grad_norm": 0.3139989376068115, "learning_rate": 7.155228984959446e-05, "loss": 0.0301, "step": 31220 }, { "grad_norm": 0.39309757947921753, "learning_rate": 7.153363831496621e-05, "loss": 0.0362, "step": 31230 }, { "grad_norm": 0.33116862177848816, "learning_rate": 7.151498310082811e-05, "loss": 0.0382, "step": 31240 }, { "grad_norm": 0.3043581545352936, "learning_rate": 7.149632421036784e-05, "loss": 0.0306, "step": 31250 }, { "grad_norm": 0.4764583110809326, "learning_rate": 7.147766164677369e-05, "loss": 0.0583, "step": 31260 }, { "grad_norm": 0.24923819303512573, "learning_rate": 7.145899541323459e-05, "loss": 0.0294, "step": 31270 }, { "grad_norm": 0.36385440826416016, "learning_rate": 7.144032551294007e-05, "loss": 0.0454, "step": 31280 }, { "grad_norm": 0.38495969772338867, "learning_rate": 7.14216519490803e-05, "loss": 0.0467, "step": 31290 }, { "grad_norm": 0.4732532501220703, "learning_rate": 7.140297472484609e-05, "loss": 0.0454, "step": 31300 }, { "grad_norm": 0.2618119716644287, "learning_rate": 7.138429384342891e-05, "loss": 0.0535, "step": 31310 }, { "grad_norm": 0.4746036231517792, "learning_rate": 7.136560930802074e-05, "loss": 0.0327, "step": 31320 }, { "grad_norm": 0.42882779240608215, "learning_rate": 7.134692112181431e-05, "loss": 0.0284, "step": 31330 }, { "grad_norm": 0.43366333842277527, "learning_rate": 7.13282292880029e-05, "loss": 0.0356, "step": 31340 }, { "grad_norm": 0.4495382010936737, "learning_rate": 7.130953380978043e-05, "loss": 0.0337, "step": 31350 }, { "grad_norm": 0.31313565373420715, "learning_rate": 7.129083469034144e-05, "loss": 0.0338, "step": 31360 }, { "grad_norm": 0.5295373797416687, "learning_rate": 7.127213193288112e-05, "loss": 0.0399, "step": 31370 }, { "grad_norm": 0.3818603456020355, "learning_rate": 7.125342554059522e-05, "loss": 0.0271, "step": 31380 }, { "grad_norm": 0.3444819748401642, "learning_rate": 7.12347155166802e-05, "loss": 0.0385, "step": 31390 }, { "grad_norm": 0.4388984441757202, "learning_rate": 7.121600186433306e-05, "loss": 0.0322, "step": 31400 }, { "grad_norm": 0.43212077021598816, "learning_rate": 7.119728458675148e-05, "loss": 0.0316, "step": 31410 }, { "grad_norm": 0.5303488373756409, "learning_rate": 7.117856368713369e-05, "loss": 0.0308, "step": 31420 }, { "grad_norm": 0.3980102837085724, "learning_rate": 7.115983916867861e-05, "loss": 0.035, "step": 31430 }, { "grad_norm": 0.28795677423477173, "learning_rate": 7.114111103458574e-05, "loss": 0.0271, "step": 31440 }, { "grad_norm": 0.21450963616371155, "learning_rate": 7.11223792880552e-05, "loss": 0.0297, "step": 31450 }, { "grad_norm": 0.5996628403663635, "learning_rate": 7.110364393228773e-05, "loss": 0.033, "step": 31460 }, { "grad_norm": 0.36934447288513184, "learning_rate": 7.108490497048471e-05, "loss": 0.0322, "step": 31470 }, { "grad_norm": 0.3023853600025177, "learning_rate": 7.10661624058481e-05, "loss": 0.0327, "step": 31480 }, { "grad_norm": 0.341068297624588, "learning_rate": 7.10474162415805e-05, "loss": 0.0348, "step": 31490 }, { "grad_norm": 0.2711678445339203, "learning_rate": 7.102866648088511e-05, "loss": 0.0401, "step": 31500 }, { "grad_norm": 0.3200438320636749, "learning_rate": 7.100991312696576e-05, "loss": 0.0356, "step": 31510 }, { "grad_norm": 0.3826659321784973, "learning_rate": 7.099115618302686e-05, "loss": 0.0438, "step": 31520 }, { "grad_norm": 0.3310851752758026, "learning_rate": 7.097239565227349e-05, "loss": 0.0484, "step": 31530 }, { "grad_norm": 0.43690764904022217, "learning_rate": 7.09536315379113e-05, "loss": 0.0313, "step": 31540 }, { "grad_norm": 0.5221384763717651, "learning_rate": 7.093486384314656e-05, "loss": 0.0343, "step": 31550 }, { "grad_norm": 0.39427030086517334, "learning_rate": 7.091609257118616e-05, "loss": 0.037, "step": 31560 }, { "grad_norm": 0.3849527835845947, "learning_rate": 7.08973177252376e-05, "loss": 0.0333, "step": 31570 }, { "grad_norm": 0.40799570083618164, "learning_rate": 7.087853930850898e-05, "loss": 0.0345, "step": 31580 }, { "grad_norm": 0.340352863073349, "learning_rate": 7.085975732420903e-05, "loss": 0.029, "step": 31590 }, { "grad_norm": 0.6121144890785217, "learning_rate": 7.084097177554706e-05, "loss": 0.0269, "step": 31600 }, { "grad_norm": 1.1031218767166138, "learning_rate": 7.082218266573301e-05, "loss": 0.0398, "step": 31610 }, { "grad_norm": 0.36920326948165894, "learning_rate": 7.080338999797743e-05, "loss": 0.0347, "step": 31620 }, { "grad_norm": 0.38344264030456543, "learning_rate": 7.07845937754915e-05, "loss": 0.0339, "step": 31630 }, { "grad_norm": 0.36143872141838074, "learning_rate": 7.076579400148693e-05, "loss": 0.0396, "step": 31640 }, { "grad_norm": 0.27040839195251465, "learning_rate": 7.074699067917611e-05, "loss": 0.0388, "step": 31650 }, { "grad_norm": 0.5974910259246826, "learning_rate": 7.072818381177201e-05, "loss": 0.0363, "step": 31660 }, { "grad_norm": 0.36990585923194885, "learning_rate": 7.070937340248823e-05, "loss": 0.034, "step": 31670 }, { "grad_norm": 0.31855159997940063, "learning_rate": 7.069055945453893e-05, "loss": 0.0359, "step": 31680 }, { "grad_norm": 0.4019246995449066, "learning_rate": 7.067174197113892e-05, "loss": 0.0354, "step": 31690 }, { "grad_norm": 0.33225005865097046, "learning_rate": 7.065292095550355e-05, "loss": 0.0366, "step": 31700 }, { "grad_norm": 0.3935735821723938, "learning_rate": 7.063409641084887e-05, "loss": 0.0353, "step": 31710 }, { "grad_norm": 0.4064558446407318, "learning_rate": 7.061526834039145e-05, "loss": 0.0364, "step": 31720 }, { "grad_norm": 0.3970738351345062, "learning_rate": 7.05964367473485e-05, "loss": 0.0298, "step": 31730 }, { "grad_norm": 0.2744649350643158, "learning_rate": 7.057760163493783e-05, "loss": 0.0331, "step": 31740 }, { "grad_norm": 0.29575395584106445, "learning_rate": 7.055876300637783e-05, "loss": 0.029, "step": 31750 }, { "grad_norm": 0.3955247402191162, "learning_rate": 7.053992086488753e-05, "loss": 0.0472, "step": 31760 }, { "grad_norm": 0.32191330194473267, "learning_rate": 7.052107521368651e-05, "loss": 0.0316, "step": 31770 }, { "grad_norm": 0.3012225031852722, "learning_rate": 7.0502226055995e-05, "loss": 0.0386, "step": 31780 }, { "grad_norm": 0.27716726064682007, "learning_rate": 7.048337339503379e-05, "loss": 0.0415, "step": 31790 }, { "grad_norm": 0.4232079088687897, "learning_rate": 7.046451723402427e-05, "loss": 0.0374, "step": 31800 }, { "grad_norm": 0.3140341639518738, "learning_rate": 7.044565757618848e-05, "loss": 0.0439, "step": 31810 }, { "grad_norm": 0.37553995847702026, "learning_rate": 7.042679442474899e-05, "loss": 0.0346, "step": 31820 }, { "grad_norm": 0.28409650921821594, "learning_rate": 7.040792778292902e-05, "loss": 0.0472, "step": 31830 }, { "grad_norm": 0.5679172277450562, "learning_rate": 7.038905765395234e-05, "loss": 0.0442, "step": 31840 }, { "grad_norm": 0.4205353260040283, "learning_rate": 7.037018404104334e-05, "loss": 0.0378, "step": 31850 }, { "grad_norm": 0.2988627254962921, "learning_rate": 7.035130694742702e-05, "loss": 0.0322, "step": 31860 }, { "grad_norm": 0.3783476948738098, "learning_rate": 7.033242637632897e-05, "loss": 0.036, "step": 31870 }, { "grad_norm": 0.35858938097953796, "learning_rate": 7.031354233097534e-05, "loss": 0.0338, "step": 31880 }, { "grad_norm": 0.45326852798461914, "learning_rate": 7.029465481459289e-05, "loss": 0.0526, "step": 31890 }, { "grad_norm": 0.48851627111434937, "learning_rate": 7.027576383040898e-05, "loss": 0.0319, "step": 31900 }, { "grad_norm": 0.299921452999115, "learning_rate": 7.025686938165159e-05, "loss": 0.0408, "step": 31910 }, { "grad_norm": 0.4646305441856384, "learning_rate": 7.023797147154924e-05, "loss": 0.0496, "step": 31920 }, { "grad_norm": 0.556547224521637, "learning_rate": 7.021907010333111e-05, "loss": 0.03, "step": 31930 }, { "grad_norm": 0.30064117908477783, "learning_rate": 7.020016528022685e-05, "loss": 0.032, "step": 31940 }, { "grad_norm": 0.3882339596748352, "learning_rate": 7.018125700546683e-05, "loss": 0.046, "step": 31950 }, { "grad_norm": 0.37401995062828064, "learning_rate": 7.016234528228196e-05, "loss": 0.0384, "step": 31960 }, { "grad_norm": 0.44245755672454834, "learning_rate": 7.014343011390372e-05, "loss": 0.0331, "step": 31970 }, { "grad_norm": 0.3565702736377716, "learning_rate": 7.01245115035642e-05, "loss": 0.0276, "step": 31980 }, { "grad_norm": 0.4154418706893921, "learning_rate": 7.010558945449606e-05, "loss": 0.0471, "step": 31990 }, { "grad_norm": 0.33105775713920593, "learning_rate": 7.008666396993258e-05, "loss": 0.0261, "step": 32000 }, { "grad_norm": 0.370629221200943, "learning_rate": 7.006773505310759e-05, "loss": 0.0385, "step": 32010 }, { "grad_norm": 0.5200501680374146, "learning_rate": 7.004880270725553e-05, "loss": 0.0332, "step": 32020 }, { "grad_norm": 0.31594207882881165, "learning_rate": 7.002986693561144e-05, "loss": 0.0367, "step": 32030 }, { "grad_norm": 0.5163987874984741, "learning_rate": 7.001092774141089e-05, "loss": 0.0425, "step": 32040 }, { "grad_norm": 0.3799741268157959, "learning_rate": 6.999198512789009e-05, "loss": 0.0343, "step": 32050 }, { "grad_norm": 0.5031296014785767, "learning_rate": 6.997303909828584e-05, "loss": 0.0401, "step": 32060 }, { "grad_norm": 0.39129501581192017, "learning_rate": 6.995408965583544e-05, "loss": 0.0373, "step": 32070 }, { "grad_norm": 0.3732142746448517, "learning_rate": 6.993513680377688e-05, "loss": 0.0248, "step": 32080 }, { "grad_norm": 0.5201902389526367, "learning_rate": 6.991618054534868e-05, "loss": 0.043, "step": 32090 }, { "grad_norm": 0.4272080361843109, "learning_rate": 6.989722088378991e-05, "loss": 0.0339, "step": 32100 }, { "grad_norm": 0.44010984897613525, "learning_rate": 6.987825782234027e-05, "loss": 0.0405, "step": 32110 }, { "grad_norm": 0.3467625081539154, "learning_rate": 6.985929136424006e-05, "loss": 0.0313, "step": 32120 }, { "grad_norm": 0.5875796675682068, "learning_rate": 6.984032151273012e-05, "loss": 0.0439, "step": 32130 }, { "grad_norm": 0.4986245632171631, "learning_rate": 6.982134827105186e-05, "loss": 0.0371, "step": 32140 }, { "grad_norm": 0.5478833317756653, "learning_rate": 6.980237164244729e-05, "loss": 0.0347, "step": 32150 }, { "grad_norm": 0.4202039837837219, "learning_rate": 6.9783391630159e-05, "loss": 0.0294, "step": 32160 }, { "grad_norm": 0.4130723178386688, "learning_rate": 6.976440823743015e-05, "loss": 0.03, "step": 32170 }, { "grad_norm": 0.3273014724254608, "learning_rate": 6.974542146750451e-05, "loss": 0.0273, "step": 32180 }, { "grad_norm": 0.2950308620929718, "learning_rate": 6.972643132362637e-05, "loss": 0.0266, "step": 32190 }, { "grad_norm": 0.2872341275215149, "learning_rate": 6.970743780904064e-05, "loss": 0.0356, "step": 32200 }, { "grad_norm": 0.2966613173484802, "learning_rate": 6.968844092699277e-05, "loss": 0.0323, "step": 32210 }, { "grad_norm": 0.6963862776756287, "learning_rate": 6.966944068072883e-05, "loss": 0.0324, "step": 32220 }, { "grad_norm": 0.36676278710365295, "learning_rate": 6.965043707349545e-05, "loss": 0.0267, "step": 32230 }, { "grad_norm": 0.4575226902961731, "learning_rate": 6.963143010853982e-05, "loss": 0.0376, "step": 32240 }, { "grad_norm": 0.462315171957016, "learning_rate": 6.961241978910971e-05, "loss": 0.0393, "step": 32250 }, { "grad_norm": 0.39620471000671387, "learning_rate": 6.959340611845344e-05, "loss": 0.0336, "step": 32260 }, { "grad_norm": 0.3739464581012726, "learning_rate": 6.957438909981995e-05, "loss": 0.0386, "step": 32270 }, { "grad_norm": 0.333859384059906, "learning_rate": 6.955536873645872e-05, "loss": 0.0286, "step": 32280 }, { "grad_norm": 0.44981393218040466, "learning_rate": 6.953634503161982e-05, "loss": 0.0339, "step": 32290 }, { "grad_norm": 0.4996778666973114, "learning_rate": 6.951731798855387e-05, "loss": 0.0352, "step": 32300 }, { "grad_norm": 0.2552970349788666, "learning_rate": 6.949828761051208e-05, "loss": 0.039, "step": 32310 }, { "grad_norm": 0.2810278534889221, "learning_rate": 6.947925390074622e-05, "loss": 0.0247, "step": 32320 }, { "grad_norm": 0.32417383790016174, "learning_rate": 6.946021686250863e-05, "loss": 0.0266, "step": 32330 }, { "grad_norm": 0.2726321518421173, "learning_rate": 6.94411764990522e-05, "loss": 0.0308, "step": 32340 }, { "grad_norm": 0.3633575141429901, "learning_rate": 6.942213281363044e-05, "loss": 0.0268, "step": 32350 }, { "grad_norm": 0.5775138139724731, "learning_rate": 6.940308580949737e-05, "loss": 0.0329, "step": 32360 }, { "grad_norm": 0.34421610832214355, "learning_rate": 6.93840354899076e-05, "loss": 0.0308, "step": 32370 }, { "grad_norm": 0.265896737575531, "learning_rate": 6.936498185811633e-05, "loss": 0.0341, "step": 32380 }, { "grad_norm": 0.3363979458808899, "learning_rate": 6.93459249173793e-05, "loss": 0.0282, "step": 32390 }, { "grad_norm": 0.39270898699760437, "learning_rate": 6.932686467095279e-05, "loss": 0.0326, "step": 32400 }, { "grad_norm": 0.47159266471862793, "learning_rate": 6.930780112209373e-05, "loss": 0.0359, "step": 32410 }, { "grad_norm": 0.5416678786277771, "learning_rate": 6.92887342740595e-05, "loss": 0.0318, "step": 32420 }, { "grad_norm": 0.4667346179485321, "learning_rate": 6.926966413010816e-05, "loss": 0.0422, "step": 32430 }, { "grad_norm": 0.3133229613304138, "learning_rate": 6.925059069349824e-05, "loss": 0.0273, "step": 32440 }, { "grad_norm": 0.3318506181240082, "learning_rate": 6.923151396748886e-05, "loss": 0.0328, "step": 32450 }, { "grad_norm": 0.37422975897789, "learning_rate": 6.921243395533974e-05, "loss": 0.0312, "step": 32460 }, { "grad_norm": 0.3894628882408142, "learning_rate": 6.919335066031109e-05, "loss": 0.0407, "step": 32470 }, { "grad_norm": 0.37828120589256287, "learning_rate": 6.917426408566379e-05, "loss": 0.0399, "step": 32480 }, { "grad_norm": 0.4784557521343231, "learning_rate": 6.915517423465916e-05, "loss": 0.0434, "step": 32490 }, { "grad_norm": 0.3351314961910248, "learning_rate": 6.913608111055914e-05, "loss": 0.0328, "step": 32500 }, { "grad_norm": 0.2790713608264923, "learning_rate": 6.911698471662623e-05, "loss": 0.0258, "step": 32510 }, { "grad_norm": 0.3767964243888855, "learning_rate": 6.90978850561235e-05, "loss": 0.0352, "step": 32520 }, { "grad_norm": 0.3591919243335724, "learning_rate": 6.907878213231454e-05, "loss": 0.0303, "step": 32530 }, { "grad_norm": 0.4743526875972748, "learning_rate": 6.90596759484635e-05, "loss": 0.0332, "step": 32540 }, { "grad_norm": 0.38021132349967957, "learning_rate": 6.904056650783514e-05, "loss": 0.0347, "step": 32550 }, { "grad_norm": 0.48176905512809753, "learning_rate": 6.902145381369471e-05, "loss": 0.0368, "step": 32560 }, { "grad_norm": 0.49933746457099915, "learning_rate": 6.900233786930808e-05, "loss": 0.0337, "step": 32570 }, { "grad_norm": 0.455862820148468, "learning_rate": 6.898321867794161e-05, "loss": 0.0362, "step": 32580 }, { "grad_norm": 0.6583614349365234, "learning_rate": 6.896409624286226e-05, "loss": 0.0288, "step": 32590 }, { "grad_norm": 0.36139407753944397, "learning_rate": 6.894497056733754e-05, "loss": 0.0316, "step": 32600 }, { "grad_norm": 0.2632269263267517, "learning_rate": 6.89258416546355e-05, "loss": 0.0355, "step": 32610 }, { "grad_norm": 0.37448009848594666, "learning_rate": 6.890670950802474e-05, "loss": 0.0331, "step": 32620 }, { "grad_norm": 0.3025475740432739, "learning_rate": 6.88875741307744e-05, "loss": 0.0344, "step": 32630 }, { "grad_norm": 0.32383331656455994, "learning_rate": 6.886843552615425e-05, "loss": 0.0311, "step": 32640 }, { "grad_norm": 0.33434563875198364, "learning_rate": 6.884929369743451e-05, "loss": 0.0247, "step": 32650 }, { "grad_norm": 0.333529531955719, "learning_rate": 6.8830148647886e-05, "loss": 0.0309, "step": 32660 }, { "grad_norm": 0.30096325278282166, "learning_rate": 6.88110003807801e-05, "loss": 0.0427, "step": 32670 }, { "grad_norm": 0.36118072271347046, "learning_rate": 6.87918488993887e-05, "loss": 0.0351, "step": 32680 }, { "grad_norm": 0.4170941412448883, "learning_rate": 6.877269420698431e-05, "loss": 0.0379, "step": 32690 }, { "grad_norm": 0.4186846613883972, "learning_rate": 6.875353630683989e-05, "loss": 0.0401, "step": 32700 }, { "grad_norm": 0.48794102668762207, "learning_rate": 6.873437520222905e-05, "loss": 0.0468, "step": 32710 }, { "grad_norm": 0.2970696985721588, "learning_rate": 6.871521089642585e-05, "loss": 0.0318, "step": 32720 }, { "grad_norm": 0.2711103558540344, "learning_rate": 6.869604339270498e-05, "loss": 0.0334, "step": 32730 }, { "grad_norm": 0.3673119843006134, "learning_rate": 6.867687269434164e-05, "loss": 0.0391, "step": 32740 }, { "grad_norm": 0.28704962134361267, "learning_rate": 6.865769880461156e-05, "loss": 0.0313, "step": 32750 }, { "grad_norm": 0.3950086534023285, "learning_rate": 6.863852172679104e-05, "loss": 0.033, "step": 32760 }, { "grad_norm": 0.30308249592781067, "learning_rate": 6.861934146415693e-05, "loss": 0.0335, "step": 32770 }, { "grad_norm": 0.3581182062625885, "learning_rate": 6.86001580199866e-05, "loss": 0.0384, "step": 32780 }, { "grad_norm": 0.23908495903015137, "learning_rate": 6.858097139755798e-05, "loss": 0.0268, "step": 32790 }, { "grad_norm": 0.42139995098114014, "learning_rate": 6.856178160014955e-05, "loss": 0.0335, "step": 32800 }, { "grad_norm": 0.37038877606391907, "learning_rate": 6.85425886310403e-05, "loss": 0.026, "step": 32810 }, { "grad_norm": 0.3953113257884979, "learning_rate": 6.852339249350979e-05, "loss": 0.0332, "step": 32820 }, { "grad_norm": 0.4136682450771332, "learning_rate": 6.850419319083812e-05, "loss": 0.0419, "step": 32830 }, { "grad_norm": 0.35142791271209717, "learning_rate": 6.848499072630592e-05, "loss": 0.029, "step": 32840 }, { "grad_norm": 0.35719361901283264, "learning_rate": 6.846578510319439e-05, "loss": 0.0356, "step": 32850 }, { "grad_norm": 0.31676313281059265, "learning_rate": 6.844657632478519e-05, "loss": 0.0352, "step": 32860 }, { "grad_norm": 0.5051090717315674, "learning_rate": 6.842736439436063e-05, "loss": 0.0264, "step": 32870 }, { "grad_norm": 0.3439368009567261, "learning_rate": 6.84081493152035e-05, "loss": 0.0262, "step": 32880 }, { "grad_norm": 0.30669164657592773, "learning_rate": 6.83889310905971e-05, "loss": 0.0234, "step": 32890 }, { "grad_norm": 0.3939712643623352, "learning_rate": 6.836970972382533e-05, "loss": 0.039, "step": 32900 }, { "grad_norm": 0.29067525267601013, "learning_rate": 6.835048521817257e-05, "loss": 0.034, "step": 32910 }, { "grad_norm": 0.31097787618637085, "learning_rate": 6.833125757692379e-05, "loss": 0.0356, "step": 32920 }, { "grad_norm": 0.3662610948085785, "learning_rate": 6.831202680336441e-05, "loss": 0.0267, "step": 32930 }, { "grad_norm": 0.3025481700897217, "learning_rate": 6.829279290078052e-05, "loss": 0.0293, "step": 32940 }, { "grad_norm": 0.3091421127319336, "learning_rate": 6.827355587245863e-05, "loss": 0.0373, "step": 32950 }, { "grad_norm": 0.38091838359832764, "learning_rate": 6.82543157216858e-05, "loss": 0.0332, "step": 32960 }, { "grad_norm": 0.3159310221672058, "learning_rate": 6.823507245174969e-05, "loss": 0.0352, "step": 32970 }, { "grad_norm": 0.4156155586242676, "learning_rate": 6.821582606593841e-05, "loss": 0.0313, "step": 32980 }, { "grad_norm": 0.3550998866558075, "learning_rate": 6.81965765675407e-05, "loss": 0.0273, "step": 32990 }, { "grad_norm": 0.3556648790836334, "learning_rate": 6.81773239598457e-05, "loss": 0.0347, "step": 33000 }, { "grad_norm": 0.390311062335968, "learning_rate": 6.815806824614319e-05, "loss": 0.0352, "step": 33010 }, { "grad_norm": 0.2593719959259033, "learning_rate": 6.813880942972343e-05, "loss": 0.0244, "step": 33020 }, { "grad_norm": 0.4060942232608795, "learning_rate": 6.811954751387726e-05, "loss": 0.0334, "step": 33030 }, { "grad_norm": 0.349837064743042, "learning_rate": 6.810028250189598e-05, "loss": 0.0287, "step": 33040 }, { "grad_norm": 0.45767316222190857, "learning_rate": 6.808101439707147e-05, "loss": 0.0453, "step": 33050 }, { "grad_norm": 0.37946370244026184, "learning_rate": 6.806174320269609e-05, "loss": 0.0355, "step": 33060 }, { "grad_norm": 0.4056757688522339, "learning_rate": 6.804246892206281e-05, "loss": 0.0394, "step": 33070 }, { "grad_norm": 0.26176270842552185, "learning_rate": 6.802319155846506e-05, "loss": 0.0488, "step": 33080 }, { "grad_norm": 0.33351263403892517, "learning_rate": 6.800391111519679e-05, "loss": 0.0321, "step": 33090 }, { "grad_norm": 0.30584266781806946, "learning_rate": 6.798462759555253e-05, "loss": 0.0375, "step": 33100 }, { "grad_norm": 0.37287142872810364, "learning_rate": 6.79653410028273e-05, "loss": 0.0296, "step": 33110 }, { "grad_norm": 0.3111148178577423, "learning_rate": 6.794605134031663e-05, "loss": 0.0333, "step": 33120 }, { "grad_norm": 0.4856933653354645, "learning_rate": 6.792675861131661e-05, "loss": 0.0319, "step": 33130 }, { "grad_norm": 0.7935153245925903, "learning_rate": 6.790746281912386e-05, "loss": 0.0459, "step": 33140 }, { "grad_norm": 0.39169493317604065, "learning_rate": 6.788816396703546e-05, "loss": 0.0336, "step": 33150 }, { "grad_norm": 0.41080963611602783, "learning_rate": 6.78688620583491e-05, "loss": 0.0367, "step": 33160 }, { "grad_norm": 0.4510292112827301, "learning_rate": 6.784955709636292e-05, "loss": 0.0353, "step": 33170 }, { "grad_norm": 0.4338591992855072, "learning_rate": 6.783024908437564e-05, "loss": 0.0295, "step": 33180 }, { "grad_norm": 0.4420567452907562, "learning_rate": 6.781093802568641e-05, "loss": 0.0263, "step": 33190 }, { "grad_norm": 0.32832545042037964, "learning_rate": 6.779162392359504e-05, "loss": 0.0328, "step": 33200 }, { "grad_norm": 0.4657171368598938, "learning_rate": 6.777230678140172e-05, "loss": 0.0378, "step": 33210 }, { "grad_norm": 0.43794888257980347, "learning_rate": 6.775298660240726e-05, "loss": 0.0381, "step": 33220 }, { "grad_norm": 0.3038850426673889, "learning_rate": 6.773366338991292e-05, "loss": 0.0435, "step": 33230 }, { "grad_norm": 0.3024139106273651, "learning_rate": 6.771433714722052e-05, "loss": 0.0317, "step": 33240 }, { "grad_norm": 0.362456738948822, "learning_rate": 6.769500787763239e-05, "loss": 0.0443, "step": 33250 }, { "grad_norm": 0.3129747807979584, "learning_rate": 6.76756755844514e-05, "loss": 0.0347, "step": 33260 }, { "grad_norm": 0.5799694061279297, "learning_rate": 6.765634027098087e-05, "loss": 0.0391, "step": 33270 }, { "grad_norm": 0.3092932403087616, "learning_rate": 6.763700194052468e-05, "loss": 0.0321, "step": 33280 }, { "grad_norm": 0.4162944257259369, "learning_rate": 6.761766059638723e-05, "loss": 0.0377, "step": 33290 }, { "grad_norm": 0.3792020380496979, "learning_rate": 6.759831624187345e-05, "loss": 0.0307, "step": 33300 }, { "grad_norm": 0.4716881513595581, "learning_rate": 6.757896888028871e-05, "loss": 0.0323, "step": 33310 }, { "grad_norm": 0.3337768316268921, "learning_rate": 6.7559618514939e-05, "loss": 0.0337, "step": 33320 }, { "grad_norm": 0.35728082060813904, "learning_rate": 6.754026514913073e-05, "loss": 0.0339, "step": 33330 }, { "grad_norm": 0.2138880342245102, "learning_rate": 6.752090878617087e-05, "loss": 0.0394, "step": 33340 }, { "grad_norm": 0.2697577476501465, "learning_rate": 6.75015494293669e-05, "loss": 0.0384, "step": 33350 }, { "grad_norm": 0.34790143370628357, "learning_rate": 6.74821870820268e-05, "loss": 0.0258, "step": 33360 }, { "grad_norm": 0.3862864077091217, "learning_rate": 6.746282174745907e-05, "loss": 0.0413, "step": 33370 }, { "grad_norm": 0.39276793599128723, "learning_rate": 6.744345342897271e-05, "loss": 0.037, "step": 33380 }, { "grad_norm": 0.34820252656936646, "learning_rate": 6.742408212987724e-05, "loss": 0.0338, "step": 33390 }, { "grad_norm": 0.9746970534324646, "learning_rate": 6.740470785348269e-05, "loss": 0.0318, "step": 33400 }, { "grad_norm": 0.3358828127384186, "learning_rate": 6.738533060309958e-05, "loss": 0.0353, "step": 33410 }, { "grad_norm": 0.31572479009628296, "learning_rate": 6.736595038203894e-05, "loss": 0.0375, "step": 33420 }, { "grad_norm": 0.3328385055065155, "learning_rate": 6.734656719361236e-05, "loss": 0.0434, "step": 33430 }, { "grad_norm": 0.2986137866973877, "learning_rate": 6.732718104113189e-05, "loss": 0.0401, "step": 33440 }, { "grad_norm": 0.6464989185333252, "learning_rate": 6.730779192791006e-05, "loss": 0.032, "step": 33450 }, { "grad_norm": 0.3572026491165161, "learning_rate": 6.728839985725997e-05, "loss": 0.0349, "step": 33460 }, { "grad_norm": 0.23879499733448029, "learning_rate": 6.726900483249517e-05, "loss": 0.0271, "step": 33470 }, { "grad_norm": 0.4047766625881195, "learning_rate": 6.724960685692976e-05, "loss": 0.0303, "step": 33480 }, { "grad_norm": 0.3055402338504791, "learning_rate": 6.723020593387833e-05, "loss": 0.027, "step": 33490 }, { "grad_norm": 0.37305769324302673, "learning_rate": 6.721080206665593e-05, "loss": 0.0337, "step": 33500 }, { "grad_norm": 0.24436315894126892, "learning_rate": 6.719139525857819e-05, "loss": 0.0376, "step": 33510 }, { "grad_norm": 0.2927365303039551, "learning_rate": 6.717198551296117e-05, "loss": 0.0284, "step": 33520 }, { "grad_norm": 0.2853141129016876, "learning_rate": 6.715257283312148e-05, "loss": 0.0356, "step": 33530 }, { "grad_norm": 0.3659020960330963, "learning_rate": 6.713315722237623e-05, "loss": 0.033, "step": 33540 }, { "grad_norm": 0.4558752477169037, "learning_rate": 6.7113738684043e-05, "loss": 0.0402, "step": 33550 }, { "grad_norm": 0.37797805666923523, "learning_rate": 6.709431722143989e-05, "loss": 0.03, "step": 33560 }, { "grad_norm": 0.2769075930118561, "learning_rate": 6.70748928378855e-05, "loss": 0.0375, "step": 33570 }, { "grad_norm": 0.565754234790802, "learning_rate": 6.705546553669891e-05, "loss": 0.0241, "step": 33580 }, { "grad_norm": 0.275956392288208, "learning_rate": 6.703603532119974e-05, "loss": 0.0274, "step": 33590 }, { "grad_norm": 0.31535640358924866, "learning_rate": 6.701660219470808e-05, "loss": 0.029, "step": 33600 }, { "grad_norm": 0.6076816916465759, "learning_rate": 6.69971661605445e-05, "loss": 0.0352, "step": 33610 }, { "grad_norm": 0.2736791968345642, "learning_rate": 6.697772722203008e-05, "loss": 0.0298, "step": 33620 }, { "grad_norm": 0.3990859091281891, "learning_rate": 6.695828538248643e-05, "loss": 0.0321, "step": 33630 }, { "grad_norm": 0.3477013409137726, "learning_rate": 6.693884064523563e-05, "loss": 0.0385, "step": 33640 }, { "grad_norm": 0.37555375695228577, "learning_rate": 6.691939301360023e-05, "loss": 0.0303, "step": 33650 }, { "grad_norm": 0.4276393949985504, "learning_rate": 6.689994249090333e-05, "loss": 0.0327, "step": 33660 }, { "grad_norm": 0.7237300872802734, "learning_rate": 6.688048908046845e-05, "loss": 0.037, "step": 33670 }, { "grad_norm": 0.29760026931762695, "learning_rate": 6.686103278561969e-05, "loss": 0.0256, "step": 33680 }, { "grad_norm": 0.2960900068283081, "learning_rate": 6.684157360968156e-05, "loss": 0.0353, "step": 33690 }, { "grad_norm": 0.4159163236618042, "learning_rate": 6.682211155597911e-05, "loss": 0.0256, "step": 33700 }, { "grad_norm": 0.46655118465423584, "learning_rate": 6.680264662783789e-05, "loss": 0.0327, "step": 33710 }, { "grad_norm": 0.3027013838291168, "learning_rate": 6.678317882858391e-05, "loss": 0.0325, "step": 33720 }, { "grad_norm": 0.4763803482055664, "learning_rate": 6.67637081615437e-05, "loss": 0.0365, "step": 33730 }, { "grad_norm": 0.3162138760089874, "learning_rate": 6.674423463004427e-05, "loss": 0.0285, "step": 33740 }, { "grad_norm": 0.3938765227794647, "learning_rate": 6.672475823741308e-05, "loss": 0.0338, "step": 33750 }, { "grad_norm": 0.3504049777984619, "learning_rate": 6.670527898697811e-05, "loss": 0.0407, "step": 33760 }, { "grad_norm": 0.44699227809906006, "learning_rate": 6.668579688206788e-05, "loss": 0.0384, "step": 33770 }, { "grad_norm": 0.3510332405567169, "learning_rate": 6.666631192601131e-05, "loss": 0.03, "step": 33780 }, { "grad_norm": 0.4174252152442932, "learning_rate": 6.664682412213785e-05, "loss": 0.0384, "step": 33790 }, { "grad_norm": 0.3458435535430908, "learning_rate": 6.662733347377745e-05, "loss": 0.0305, "step": 33800 }, { "grad_norm": 0.32180196046829224, "learning_rate": 6.660783998426051e-05, "loss": 0.0291, "step": 33810 }, { "grad_norm": 0.41702646017074585, "learning_rate": 6.658834365691794e-05, "loss": 0.0299, "step": 33820 }, { "grad_norm": 0.551552414894104, "learning_rate": 6.656884449508115e-05, "loss": 0.0288, "step": 33830 }, { "grad_norm": 0.37254539132118225, "learning_rate": 6.654934250208198e-05, "loss": 0.0313, "step": 33840 }, { "grad_norm": 0.33772918581962585, "learning_rate": 6.65298376812528e-05, "loss": 0.0301, "step": 33850 }, { "grad_norm": 0.5262866616249084, "learning_rate": 6.651033003592646e-05, "loss": 0.0292, "step": 33860 }, { "grad_norm": 0.4556344151496887, "learning_rate": 6.649081956943626e-05, "loss": 0.0267, "step": 33870 }, { "grad_norm": 0.35534968972206116, "learning_rate": 6.647130628511604e-05, "loss": 0.0275, "step": 33880 }, { "grad_norm": 0.5424228310585022, "learning_rate": 6.645179018630005e-05, "loss": 0.0258, "step": 33890 }, { "grad_norm": 0.38805198669433594, "learning_rate": 6.643227127632309e-05, "loss": 0.0309, "step": 33900 }, { "grad_norm": 0.4894638657569885, "learning_rate": 6.641274955852038e-05, "loss": 0.039, "step": 33910 }, { "grad_norm": 0.8793970942497253, "learning_rate": 6.639322503622768e-05, "loss": 0.059, "step": 33920 }, { "grad_norm": 0.45892587304115295, "learning_rate": 6.637369771278116e-05, "loss": 0.0368, "step": 33930 }, { "grad_norm": 0.3693811893463135, "learning_rate": 6.635416759151751e-05, "loss": 0.0326, "step": 33940 }, { "grad_norm": 0.4806731045246124, "learning_rate": 6.633463467577394e-05, "loss": 0.0485, "step": 33950 }, { "grad_norm": 0.366025447845459, "learning_rate": 6.631509896888803e-05, "loss": 0.0341, "step": 33960 }, { "grad_norm": 0.3680005967617035, "learning_rate": 6.629556047419794e-05, "loss": 0.0264, "step": 33970 }, { "grad_norm": 0.41901084780693054, "learning_rate": 6.627601919504223e-05, "loss": 0.0267, "step": 33980 }, { "grad_norm": 0.32861125469207764, "learning_rate": 6.625647513476001e-05, "loss": 0.0292, "step": 33990 }, { "grad_norm": 0.5864718556404114, "learning_rate": 6.62369282966908e-05, "loss": 0.0429, "step": 34000 }, { "grad_norm": 0.3943950831890106, "learning_rate": 6.621737868417464e-05, "loss": 0.0362, "step": 34010 }, { "grad_norm": 0.3292861878871918, "learning_rate": 6.619782630055198e-05, "loss": 0.0356, "step": 34020 }, { "grad_norm": 0.3029470145702362, "learning_rate": 6.617827114916382e-05, "loss": 0.0395, "step": 34030 }, { "grad_norm": 0.31775328516960144, "learning_rate": 6.615871323335161e-05, "loss": 0.0303, "step": 34040 }, { "grad_norm": 0.25148913264274597, "learning_rate": 6.613915255645725e-05, "loss": 0.0326, "step": 34050 }, { "grad_norm": 0.3675662875175476, "learning_rate": 6.611958912182312e-05, "loss": 0.0292, "step": 34060 }, { "grad_norm": 0.2856113314628601, "learning_rate": 6.610002293279207e-05, "loss": 0.0357, "step": 34070 }, { "grad_norm": 0.3796400725841522, "learning_rate": 6.608045399270746e-05, "loss": 0.0273, "step": 34080 }, { "grad_norm": 0.4076458513736725, "learning_rate": 6.606088230491304e-05, "loss": 0.0323, "step": 34090 }, { "grad_norm": 0.6074095368385315, "learning_rate": 6.604130787275312e-05, "loss": 0.0295, "step": 34100 }, { "grad_norm": 0.31813371181488037, "learning_rate": 6.602173069957242e-05, "loss": 0.0348, "step": 34110 }, { "grad_norm": 0.506299614906311, "learning_rate": 6.600215078871612e-05, "loss": 0.0333, "step": 34120 }, { "grad_norm": 0.3171873688697815, "learning_rate": 6.598256814352992e-05, "loss": 0.0286, "step": 34130 }, { "grad_norm": 0.3181289732456207, "learning_rate": 6.596298276735995e-05, "loss": 0.0294, "step": 34140 }, { "grad_norm": 0.5334560871124268, "learning_rate": 6.594339466355282e-05, "loss": 0.0317, "step": 34150 }, { "grad_norm": 0.34306830167770386, "learning_rate": 6.592380383545558e-05, "loss": 0.0319, "step": 34160 }, { "grad_norm": 0.41631245613098145, "learning_rate": 6.590421028641577e-05, "loss": 0.0366, "step": 34170 }, { "grad_norm": 0.3083113729953766, "learning_rate": 6.588461401978143e-05, "loss": 0.0343, "step": 34180 }, { "grad_norm": 0.41091474890708923, "learning_rate": 6.586501503890099e-05, "loss": 0.0306, "step": 34190 }, { "grad_norm": 0.40012380480766296, "learning_rate": 6.584541334712338e-05, "loss": 0.0267, "step": 34200 }, { "grad_norm": 0.399160772562027, "learning_rate": 6.582580894779802e-05, "loss": 0.0309, "step": 34210 }, { "grad_norm": 0.34523364901542664, "learning_rate": 6.580620184427473e-05, "loss": 0.0274, "step": 34220 }, { "grad_norm": 0.37389039993286133, "learning_rate": 6.578659203990385e-05, "loss": 0.0281, "step": 34230 }, { "grad_norm": 0.26838386058807373, "learning_rate": 6.576697953803615e-05, "loss": 0.032, "step": 34240 }, { "grad_norm": 0.5248852968215942, "learning_rate": 6.57473643420229e-05, "loss": 0.0311, "step": 34250 }, { "grad_norm": 0.3268253207206726, "learning_rate": 6.572774645521574e-05, "loss": 0.0336, "step": 34260 }, { "grad_norm": 0.5103569030761719, "learning_rate": 6.570812588096688e-05, "loss": 0.0442, "step": 34270 }, { "grad_norm": 0.4789632260799408, "learning_rate": 6.568850262262893e-05, "loss": 0.0311, "step": 34280 }, { "grad_norm": 0.3261806070804596, "learning_rate": 6.566887668355497e-05, "loss": 0.0282, "step": 34290 }, { "grad_norm": 0.28362539410591125, "learning_rate": 6.564924806709851e-05, "loss": 0.0258, "step": 34300 }, { "grad_norm": 0.34956255555152893, "learning_rate": 6.562961677661359e-05, "loss": 0.0362, "step": 34310 }, { "grad_norm": 0.5302630066871643, "learning_rate": 6.56099828154546e-05, "loss": 0.0356, "step": 34320 }, { "grad_norm": 0.31316864490509033, "learning_rate": 6.55903461869765e-05, "loss": 0.0277, "step": 34330 }, { "grad_norm": 0.20137861371040344, "learning_rate": 6.557070689453465e-05, "loss": 0.0262, "step": 34340 }, { "grad_norm": 0.2333398014307022, "learning_rate": 6.555106494148482e-05, "loss": 0.0384, "step": 34350 }, { "grad_norm": 0.4862411618232727, "learning_rate": 6.553142033118333e-05, "loss": 0.0366, "step": 34360 }, { "grad_norm": 0.5108123421669006, "learning_rate": 6.551177306698688e-05, "loss": 0.0417, "step": 34370 }, { "grad_norm": 0.5107901692390442, "learning_rate": 6.549212315225267e-05, "loss": 0.0304, "step": 34380 }, { "grad_norm": 0.23619118332862854, "learning_rate": 6.547247059033833e-05, "loss": 0.0322, "step": 34390 }, { "grad_norm": 0.35699325799942017, "learning_rate": 6.545281538460193e-05, "loss": 0.0339, "step": 34400 }, { "grad_norm": 0.5290123820304871, "learning_rate": 6.543315753840202e-05, "loss": 0.0296, "step": 34410 }, { "grad_norm": 0.3134731948375702, "learning_rate": 6.541349705509758e-05, "loss": 0.0285, "step": 34420 }, { "grad_norm": 0.30627453327178955, "learning_rate": 6.539383393804805e-05, "loss": 0.0307, "step": 34430 }, { "grad_norm": 0.25154149532318115, "learning_rate": 6.537416819061333e-05, "loss": 0.0356, "step": 34440 }, { "grad_norm": 0.2981135845184326, "learning_rate": 6.535449981615375e-05, "loss": 0.032, "step": 34450 }, { "grad_norm": 0.3525814116001129, "learning_rate": 6.53348288180301e-05, "loss": 0.0269, "step": 34460 }, { "grad_norm": 0.4495707154273987, "learning_rate": 6.531515519960361e-05, "loss": 0.038, "step": 34470 }, { "grad_norm": 0.3751237988471985, "learning_rate": 6.529547896423597e-05, "loss": 0.0313, "step": 34480 }, { "grad_norm": 0.37937793135643005, "learning_rate": 6.52758001152893e-05, "loss": 0.0416, "step": 34490 }, { "grad_norm": 0.24503357708454132, "learning_rate": 6.525611865612618e-05, "loss": 0.0273, "step": 34500 }, { "grad_norm": 0.41179683804512024, "learning_rate": 6.523643459010966e-05, "loss": 0.0305, "step": 34510 }, { "grad_norm": 0.3749791979789734, "learning_rate": 6.521674792060317e-05, "loss": 0.0293, "step": 34520 }, { "grad_norm": 0.48435521125793457, "learning_rate": 6.519705865097063e-05, "loss": 0.0352, "step": 34530 }, { "grad_norm": 0.4635770916938782, "learning_rate": 6.517736678457641e-05, "loss": 0.0309, "step": 34540 }, { "grad_norm": 0.36322399973869324, "learning_rate": 6.515767232478534e-05, "loss": 0.0311, "step": 34550 }, { "grad_norm": 0.2823254466056824, "learning_rate": 6.51379752749626e-05, "loss": 0.0217, "step": 34560 }, { "grad_norm": 0.34044715762138367, "learning_rate": 6.511827563847393e-05, "loss": 0.032, "step": 34570 }, { "grad_norm": 0.35362744331359863, "learning_rate": 6.509857341868542e-05, "loss": 0.037, "step": 34580 }, { "grad_norm": 0.3429074287414551, "learning_rate": 6.507886861896367e-05, "loss": 0.031, "step": 34590 }, { "grad_norm": 0.42235487699508667, "learning_rate": 6.505916124267567e-05, "loss": 0.0379, "step": 34600 }, { "grad_norm": 0.5626109838485718, "learning_rate": 6.503945129318891e-05, "loss": 0.0359, "step": 34610 }, { "grad_norm": 0.45604342222213745, "learning_rate": 6.501973877387122e-05, "loss": 0.0254, "step": 34620 }, { "grad_norm": 1.3704543113708496, "learning_rate": 6.500002368809098e-05, "loss": 0.0297, "step": 34630 }, { "grad_norm": 0.43851613998413086, "learning_rate": 6.498030603921694e-05, "loss": 0.0384, "step": 34640 }, { "grad_norm": 0.5291946530342102, "learning_rate": 6.496058583061832e-05, "loss": 0.0379, "step": 34650 }, { "grad_norm": 0.3408827483654022, "learning_rate": 6.494086306566475e-05, "loss": 0.0305, "step": 34660 }, { "grad_norm": 0.5653716921806335, "learning_rate": 6.492113774772632e-05, "loss": 0.0332, "step": 34670 }, { "grad_norm": 0.4735427796840668, "learning_rate": 6.490140988017354e-05, "loss": 0.0346, "step": 34680 }, { "grad_norm": 0.44707566499710083, "learning_rate": 6.488167946637736e-05, "loss": 0.0315, "step": 34690 }, { "grad_norm": 0.4532257616519928, "learning_rate": 6.486194650970915e-05, "loss": 0.037, "step": 34700 }, { "grad_norm": 0.40889814496040344, "learning_rate": 6.48422110135408e-05, "loss": 0.0346, "step": 34710 }, { "grad_norm": 0.24466148018836975, "learning_rate": 6.482247298124451e-05, "loss": 0.0338, "step": 34720 }, { "grad_norm": 0.26507773995399475, "learning_rate": 6.480273241619297e-05, "loss": 0.0324, "step": 34730 }, { "grad_norm": 0.3602505922317505, "learning_rate": 6.478298932175933e-05, "loss": 0.0318, "step": 34740 }, { "grad_norm": 0.34007519483566284, "learning_rate": 6.476324370131712e-05, "loss": 0.0341, "step": 34750 }, { "grad_norm": 0.3381635248661041, "learning_rate": 6.474349555824036e-05, "loss": 0.0395, "step": 34760 }, { "grad_norm": 0.3755863606929779, "learning_rate": 6.472374489590342e-05, "loss": 0.0273, "step": 34770 }, { "grad_norm": 0.1879078447818756, "learning_rate": 6.470399171768118e-05, "loss": 0.0468, "step": 34780 }, { "grad_norm": 0.5436182618141174, "learning_rate": 6.468423602694891e-05, "loss": 0.0296, "step": 34790 }, { "grad_norm": 0.37170740962028503, "learning_rate": 6.466447782708232e-05, "loss": 0.0268, "step": 34800 }, { "grad_norm": 0.24001345038414001, "learning_rate": 6.464471712145754e-05, "loss": 0.0279, "step": 34810 }, { "grad_norm": 0.23531654477119446, "learning_rate": 6.462495391345114e-05, "loss": 0.0323, "step": 34820 }, { "grad_norm": 0.3227222263813019, "learning_rate": 6.46051882064401e-05, "loss": 0.0232, "step": 34830 }, { "grad_norm": 0.34535712003707886, "learning_rate": 6.458542000380186e-05, "loss": 0.0339, "step": 34840 }, { "grad_norm": 0.3391430974006653, "learning_rate": 6.456564930891424e-05, "loss": 0.0267, "step": 34850 }, { "grad_norm": 0.45907309651374817, "learning_rate": 6.454587612515555e-05, "loss": 0.0352, "step": 34860 }, { "grad_norm": 0.31403428316116333, "learning_rate": 6.452610045590444e-05, "loss": 0.0284, "step": 34870 }, { "grad_norm": 0.5653931498527527, "learning_rate": 6.450632230454005e-05, "loss": 0.0307, "step": 34880 }, { "grad_norm": 0.3740706145763397, "learning_rate": 6.448654167444195e-05, "loss": 0.0423, "step": 34890 }, { "grad_norm": 0.5237762928009033, "learning_rate": 6.446675856899005e-05, "loss": 0.0319, "step": 34900 }, { "grad_norm": 0.2409091591835022, "learning_rate": 6.444697299156481e-05, "loss": 0.0297, "step": 34910 }, { "grad_norm": 0.4104795455932617, "learning_rate": 6.442718494554701e-05, "loss": 0.0368, "step": 34920 }, { "grad_norm": 0.3199581205844879, "learning_rate": 6.440739443431787e-05, "loss": 0.0248, "step": 34930 }, { "grad_norm": 0.25302475690841675, "learning_rate": 6.438760146125906e-05, "loss": 0.0203, "step": 34940 }, { "grad_norm": 0.3392316699028015, "learning_rate": 6.436780602975267e-05, "loss": 0.0274, "step": 34950 }, { "grad_norm": 0.32736778259277344, "learning_rate": 6.43480081431812e-05, "loss": 0.0289, "step": 34960 }, { "grad_norm": 0.43551701307296753, "learning_rate": 6.432820780492756e-05, "loss": 0.0324, "step": 34970 }, { "grad_norm": 0.3503730297088623, "learning_rate": 6.430840501837506e-05, "loss": 0.0368, "step": 34980 }, { "grad_norm": 0.3475421369075775, "learning_rate": 6.428859978690748e-05, "loss": 0.0332, "step": 34990 }, { "grad_norm": 0.3443954288959503, "learning_rate": 6.426879211390901e-05, "loss": 0.0285, "step": 35000 }, { "grad_norm": 0.44441327452659607, "learning_rate": 6.424898200276422e-05, "loss": 0.0329, "step": 35010 }, { "grad_norm": 0.330588161945343, "learning_rate": 6.42291694568581e-05, "loss": 0.0295, "step": 35020 }, { "grad_norm": 0.4821425676345825, "learning_rate": 6.42093544795761e-05, "loss": 0.0418, "step": 35030 }, { "grad_norm": 0.22746828198432922, "learning_rate": 6.418953707430403e-05, "loss": 0.0292, "step": 35040 }, { "grad_norm": 0.38520872592926025, "learning_rate": 6.416971724442819e-05, "loss": 0.0276, "step": 35050 }, { "grad_norm": 0.24083638191223145, "learning_rate": 6.414989499333519e-05, "loss": 0.0228, "step": 35060 }, { "grad_norm": 0.42938998341560364, "learning_rate": 6.413007032441214e-05, "loss": 0.0302, "step": 35070 }, { "grad_norm": 0.3642318844795227, "learning_rate": 6.411024324104653e-05, "loss": 0.0303, "step": 35080 }, { "grad_norm": 0.2889244258403778, "learning_rate": 6.409041374662628e-05, "loss": 0.0259, "step": 35090 }, { "grad_norm": 0.278328537940979, "learning_rate": 6.407058184453967e-05, "loss": 0.0369, "step": 35100 }, { "grad_norm": 0.4049442410469055, "learning_rate": 6.405074753817548e-05, "loss": 0.0284, "step": 35110 }, { "grad_norm": 0.323723167181015, "learning_rate": 6.40309108309228e-05, "loss": 0.0297, "step": 35120 }, { "grad_norm": 0.41481921076774597, "learning_rate": 6.401107172617122e-05, "loss": 0.0291, "step": 35130 }, { "grad_norm": 0.440049409866333, "learning_rate": 6.399123022731068e-05, "loss": 0.0479, "step": 35140 }, { "grad_norm": 0.37575241923332214, "learning_rate": 6.397138633773157e-05, "loss": 0.0265, "step": 35150 }, { "grad_norm": 0.5694388747215271, "learning_rate": 6.395154006082463e-05, "loss": 0.0304, "step": 35160 }, { "grad_norm": 0.40398383140563965, "learning_rate": 6.393169139998109e-05, "loss": 0.032, "step": 35170 }, { "grad_norm": 0.47221991419792175, "learning_rate": 6.39118403585925e-05, "loss": 0.0298, "step": 35180 }, { "grad_norm": 0.28583526611328125, "learning_rate": 6.38919869400509e-05, "loss": 0.0365, "step": 35190 }, { "grad_norm": 0.367671936750412, "learning_rate": 6.387213114774865e-05, "loss": 0.0311, "step": 35200 }, { "grad_norm": 0.3896581530570984, "learning_rate": 6.385227298507863e-05, "loss": 0.0506, "step": 35210 }, { "grad_norm": 0.33003735542297363, "learning_rate": 6.3832412455434e-05, "loss": 0.0268, "step": 35220 }, { "grad_norm": 0.27215784788131714, "learning_rate": 6.381254956220841e-05, "loss": 0.0268, "step": 35230 }, { "grad_norm": 0.4320337176322937, "learning_rate": 6.379268430879586e-05, "loss": 0.0358, "step": 35240 }, { "grad_norm": 0.38316574692726135, "learning_rate": 6.37728166985908e-05, "loss": 0.0274, "step": 35250 }, { "grad_norm": 0.35016050934791565, "learning_rate": 6.375294673498804e-05, "loss": 0.0287, "step": 35260 }, { "grad_norm": 0.45525652170181274, "learning_rate": 6.373307442138284e-05, "loss": 0.0321, "step": 35270 }, { "grad_norm": 0.3644084632396698, "learning_rate": 6.371319976117081e-05, "loss": 0.0475, "step": 35280 }, { "grad_norm": 0.3256347179412842, "learning_rate": 6.3693322757748e-05, "loss": 0.0345, "step": 35290 }, { "grad_norm": 0.29996559023857117, "learning_rate": 6.367344341451086e-05, "loss": 0.0281, "step": 35300 }, { "grad_norm": 0.4517126977443695, "learning_rate": 6.36535617348562e-05, "loss": 0.033, "step": 35310 }, { "grad_norm": 0.39292052388191223, "learning_rate": 6.363367772218128e-05, "loss": 0.0338, "step": 35320 }, { "grad_norm": 0.3229483664035797, "learning_rate": 6.36137913798837e-05, "loss": 0.0297, "step": 35330 }, { "grad_norm": 0.3429716229438782, "learning_rate": 6.359390271136151e-05, "loss": 0.0279, "step": 35340 }, { "grad_norm": 0.333098828792572, "learning_rate": 6.357401172001314e-05, "loss": 0.0272, "step": 35350 }, { "grad_norm": 0.39443323016166687, "learning_rate": 6.355411840923742e-05, "loss": 0.0311, "step": 35360 }, { "grad_norm": 0.4294198453426361, "learning_rate": 6.353422278243358e-05, "loss": 0.0259, "step": 35370 }, { "grad_norm": 0.31337985396385193, "learning_rate": 6.351432484300121e-05, "loss": 0.0347, "step": 35380 }, { "grad_norm": 0.4471362233161926, "learning_rate": 6.349442459434036e-05, "loss": 0.036, "step": 35390 }, { "grad_norm": 0.6484528183937073, "learning_rate": 6.34745220398514e-05, "loss": 0.0382, "step": 35400 }, { "grad_norm": 0.42421817779541016, "learning_rate": 6.345461718293518e-05, "loss": 0.0307, "step": 35410 }, { "grad_norm": 0.39443203806877136, "learning_rate": 6.343471002699286e-05, "loss": 0.0337, "step": 35420 }, { "grad_norm": 0.3956332206726074, "learning_rate": 6.341480057542602e-05, "loss": 0.0341, "step": 35430 }, { "grad_norm": 0.26539409160614014, "learning_rate": 6.339488883163667e-05, "loss": 0.031, "step": 35440 }, { "grad_norm": 0.38259798288345337, "learning_rate": 6.337497479902716e-05, "loss": 0.0254, "step": 35450 }, { "grad_norm": 0.2983611822128296, "learning_rate": 6.335505848100027e-05, "loss": 0.0295, "step": 35460 }, { "grad_norm": 0.3264286518096924, "learning_rate": 6.333513988095915e-05, "loss": 0.0321, "step": 35470 }, { "grad_norm": 0.29417699575424194, "learning_rate": 6.331521900230735e-05, "loss": 0.0257, "step": 35480 }, { "grad_norm": 0.4941446781158447, "learning_rate": 6.329529584844878e-05, "loss": 0.0288, "step": 35490 }, { "grad_norm": 0.4492447078227997, "learning_rate": 6.327537042278777e-05, "loss": 0.0373, "step": 35500 }, { "grad_norm": 0.3697170615196228, "learning_rate": 6.325544272872905e-05, "loss": 0.0278, "step": 35510 }, { "grad_norm": 0.28603070974349976, "learning_rate": 6.323551276967771e-05, "loss": 0.0347, "step": 35520 }, { "grad_norm": 0.4197061061859131, "learning_rate": 6.321558054903922e-05, "loss": 0.027, "step": 35530 }, { "grad_norm": 0.3689723014831543, "learning_rate": 6.319564607021947e-05, "loss": 0.0274, "step": 35540 }, { "grad_norm": 0.32958781719207764, "learning_rate": 6.31757093366247e-05, "loss": 0.026, "step": 35550 }, { "grad_norm": 0.4100472033023834, "learning_rate": 6.315577035166154e-05, "loss": 0.0285, "step": 35560 }, { "grad_norm": 0.26341164112091064, "learning_rate": 6.313582911873708e-05, "loss": 0.0272, "step": 35570 }, { "grad_norm": 0.3035280108451843, "learning_rate": 6.311588564125865e-05, "loss": 0.0265, "step": 35580 }, { "grad_norm": 0.48089420795440674, "learning_rate": 6.30959399226341e-05, "loss": 0.0306, "step": 35590 }, { "grad_norm": 0.36975008249282837, "learning_rate": 6.30759919662716e-05, "loss": 0.0412, "step": 35600 }, { "grad_norm": 0.304103285074234, "learning_rate": 6.30560417755797e-05, "loss": 0.0306, "step": 35610 }, { "grad_norm": 0.2840994596481323, "learning_rate": 6.303608935396735e-05, "loss": 0.0247, "step": 35620 }, { "grad_norm": 0.21290722489356995, "learning_rate": 6.301613470484386e-05, "loss": 0.022, "step": 35630 }, { "grad_norm": 0.30663585662841797, "learning_rate": 6.299617783161893e-05, "loss": 0.0327, "step": 35640 }, { "grad_norm": 0.30453774333000183, "learning_rate": 6.297621873770266e-05, "loss": 0.0269, "step": 35650 }, { "grad_norm": 0.2959623634815216, "learning_rate": 6.29562574265055e-05, "loss": 0.0266, "step": 35660 }, { "grad_norm": 0.2560519874095917, "learning_rate": 6.293629390143834e-05, "loss": 0.0268, "step": 35670 }, { "grad_norm": 0.2800796926021576, "learning_rate": 6.291632816591232e-05, "loss": 0.0245, "step": 35680 }, { "grad_norm": 0.23591354489326477, "learning_rate": 6.28963602233391e-05, "loss": 0.0321, "step": 35690 }, { "grad_norm": 0.38903549313545227, "learning_rate": 6.287639007713062e-05, "loss": 0.0297, "step": 35700 }, { "grad_norm": 0.2740771472454071, "learning_rate": 6.285641773069926e-05, "loss": 0.0244, "step": 35710 }, { "grad_norm": 0.3026002049446106, "learning_rate": 6.283644318745773e-05, "loss": 0.0233, "step": 35720 }, { "grad_norm": 0.39200669527053833, "learning_rate": 6.281646645081912e-05, "loss": 0.0369, "step": 35730 }, { "grad_norm": 0.4573889970779419, "learning_rate": 6.279648752419693e-05, "loss": 0.0241, "step": 35740 }, { "grad_norm": 0.4690277576446533, "learning_rate": 6.2776506411005e-05, "loss": 0.0312, "step": 35750 }, { "grad_norm": 0.36617645621299744, "learning_rate": 6.275652311465758e-05, "loss": 0.0304, "step": 35760 }, { "grad_norm": 0.482201486825943, "learning_rate": 6.273653763856926e-05, "loss": 0.0342, "step": 35770 }, { "grad_norm": 0.38683241605758667, "learning_rate": 6.271654998615501e-05, "loss": 0.028, "step": 35780 }, { "grad_norm": 0.430409699678421, "learning_rate": 6.269656016083013e-05, "loss": 0.031, "step": 35790 }, { "grad_norm": 0.2289097160100937, "learning_rate": 6.267656816601038e-05, "loss": 0.0332, "step": 35800 }, { "grad_norm": 0.24013598263263702, "learning_rate": 6.265657400511185e-05, "loss": 0.0289, "step": 35810 }, { "grad_norm": 0.36738935112953186, "learning_rate": 6.263657768155098e-05, "loss": 0.029, "step": 35820 }, { "grad_norm": 0.6101770401000977, "learning_rate": 6.261657919874457e-05, "loss": 0.0299, "step": 35830 }, { "grad_norm": 0.2838019132614136, "learning_rate": 6.259657856010986e-05, "loss": 0.042, "step": 35840 }, { "grad_norm": 0.36891281604766846, "learning_rate": 6.257657576906439e-05, "loss": 0.0271, "step": 35850 }, { "grad_norm": 0.26974982023239136, "learning_rate": 6.255657082902609e-05, "loss": 0.0278, "step": 35860 }, { "grad_norm": 0.3160493075847626, "learning_rate": 6.253656374341325e-05, "loss": 0.0266, "step": 35870 }, { "grad_norm": 0.3346896469593048, "learning_rate": 6.251655451564457e-05, "loss": 0.0294, "step": 35880 }, { "grad_norm": 0.35766515135765076, "learning_rate": 6.249654314913902e-05, "loss": 0.0305, "step": 35890 }, { "grad_norm": 0.44915932416915894, "learning_rate": 6.247652964731604e-05, "loss": 0.0286, "step": 35900 }, { "grad_norm": 0.32946255803108215, "learning_rate": 6.245651401359537e-05, "loss": 0.0427, "step": 35910 }, { "grad_norm": 0.2569029927253723, "learning_rate": 6.243649625139715e-05, "loss": 0.035, "step": 35920 }, { "grad_norm": 0.3599663972854614, "learning_rate": 6.241647636414185e-05, "loss": 0.0291, "step": 35930 }, { "grad_norm": 0.3845139741897583, "learning_rate": 6.239645435525034e-05, "loss": 0.0376, "step": 35940 }, { "grad_norm": 0.28759801387786865, "learning_rate": 6.237643022814381e-05, "loss": 0.0337, "step": 35950 }, { "grad_norm": 0.4238048791885376, "learning_rate": 6.235640398624386e-05, "loss": 0.0285, "step": 35960 }, { "grad_norm": 0.2365969568490982, "learning_rate": 6.233637563297243e-05, "loss": 0.0328, "step": 35970 }, { "grad_norm": 0.47208860516548157, "learning_rate": 6.23163451717518e-05, "loss": 0.0377, "step": 35980 }, { "grad_norm": 0.4048742651939392, "learning_rate": 6.229631260600463e-05, "loss": 0.0349, "step": 35990 }, { "grad_norm": 0.3044869899749756, "learning_rate": 6.227627793915392e-05, "loss": 0.0318, "step": 36000 }, { "grad_norm": 0.340962290763855, "learning_rate": 6.225624117462309e-05, "loss": 0.0264, "step": 36010 }, { "grad_norm": 0.3584168255329132, "learning_rate": 6.223620231583586e-05, "loss": 0.0327, "step": 36020 }, { "grad_norm": 0.7758364081382751, "learning_rate": 6.221616136621629e-05, "loss": 0.0284, "step": 36030 }, { "grad_norm": 0.3233591616153717, "learning_rate": 6.219611832918887e-05, "loss": 0.031, "step": 36040 }, { "grad_norm": 0.3997601568698883, "learning_rate": 6.217607320817838e-05, "loss": 0.0253, "step": 36050 }, { "grad_norm": 0.45534613728523254, "learning_rate": 6.215602600661001e-05, "loss": 0.028, "step": 36060 }, { "grad_norm": 0.34744787216186523, "learning_rate": 6.213597672790925e-05, "loss": 0.0326, "step": 36070 }, { "grad_norm": 0.2791145145893097, "learning_rate": 6.2115925375502e-05, "loss": 0.0322, "step": 36080 }, { "grad_norm": 0.33239060640335083, "learning_rate": 6.209587195281447e-05, "loss": 0.0294, "step": 36090 }, { "grad_norm": 0.52198326587677, "learning_rate": 6.207581646327324e-05, "loss": 0.0438, "step": 36100 }, { "grad_norm": 0.28633591532707214, "learning_rate": 6.205575891030526e-05, "loss": 0.0322, "step": 36110 }, { "grad_norm": 0.39109230041503906, "learning_rate": 6.203569929733781e-05, "loss": 0.027, "step": 36120 }, { "grad_norm": 0.45502105355262756, "learning_rate": 6.201563762779852e-05, "loss": 0.0319, "step": 36130 }, { "grad_norm": 0.45598113536834717, "learning_rate": 6.199557390511538e-05, "loss": 0.0279, "step": 36140 }, { "grad_norm": 0.27931272983551025, "learning_rate": 6.197550813271675e-05, "loss": 0.031, "step": 36150 }, { "grad_norm": 1.3029628992080688, "learning_rate": 6.195544031403131e-05, "loss": 0.0294, "step": 36160 }, { "grad_norm": 0.37616246938705444, "learning_rate": 6.19353704524881e-05, "loss": 0.0311, "step": 36170 }, { "grad_norm": 0.343275785446167, "learning_rate": 6.191529855151652e-05, "loss": 0.0261, "step": 36180 }, { "grad_norm": 0.31458616256713867, "learning_rate": 6.189522461454629e-05, "loss": 0.0302, "step": 36190 }, { "grad_norm": 0.3868592083454132, "learning_rate": 6.187514864500752e-05, "loss": 0.0352, "step": 36200 }, { "grad_norm": 0.5351836681365967, "learning_rate": 6.185507064633062e-05, "loss": 0.0328, "step": 36210 }, { "grad_norm": 0.3415111303329468, "learning_rate": 6.18349906219464e-05, "loss": 0.0319, "step": 36220 }, { "grad_norm": 0.44492796063423157, "learning_rate": 6.181490857528596e-05, "loss": 0.0289, "step": 36230 }, { "grad_norm": 0.33382537961006165, "learning_rate": 6.179482450978077e-05, "loss": 0.0239, "step": 36240 }, { "grad_norm": 0.2847612202167511, "learning_rate": 6.177473842886269e-05, "loss": 0.0289, "step": 36250 }, { "grad_norm": 0.3947473168373108, "learning_rate": 6.175465033596382e-05, "loss": 0.0362, "step": 36260 }, { "grad_norm": 0.2128610461950302, "learning_rate": 6.173456023451671e-05, "loss": 0.0312, "step": 36270 }, { "grad_norm": 0.3147664964199066, "learning_rate": 6.171446812795422e-05, "loss": 0.0316, "step": 36280 }, { "grad_norm": 0.38977643847465515, "learning_rate": 6.169437401970949e-05, "loss": 0.0412, "step": 36290 }, { "grad_norm": 0.5738442540168762, "learning_rate": 6.16742779132161e-05, "loss": 0.0286, "step": 36300 }, { "grad_norm": 0.37947338819503784, "learning_rate": 6.165417981190789e-05, "loss": 0.0335, "step": 36310 }, { "grad_norm": 0.39346814155578613, "learning_rate": 6.16340797192191e-05, "loss": 0.0273, "step": 36320 }, { "grad_norm": 0.38065865635871887, "learning_rate": 6.161397763858427e-05, "loss": 0.0344, "step": 36330 }, { "grad_norm": 0.5406926870346069, "learning_rate": 6.159387357343834e-05, "loss": 0.0258, "step": 36340 }, { "grad_norm": 0.34727758169174194, "learning_rate": 6.157376752721648e-05, "loss": 0.0281, "step": 36350 }, { "grad_norm": 0.34113118052482605, "learning_rate": 6.155365950335428e-05, "loss": 0.0375, "step": 36360 }, { "grad_norm": 0.5047626495361328, "learning_rate": 6.153354950528768e-05, "loss": 0.0322, "step": 36370 }, { "grad_norm": 0.30945560336112976, "learning_rate": 6.151343753645293e-05, "loss": 0.0319, "step": 36380 }, { "grad_norm": 0.3746357262134552, "learning_rate": 6.149332360028657e-05, "loss": 0.0358, "step": 36390 }, { "grad_norm": 0.36202266812324524, "learning_rate": 6.147320770022555e-05, "loss": 0.0331, "step": 36400 }, { "grad_norm": 0.24065573513507843, "learning_rate": 6.145308983970715e-05, "loss": 0.0315, "step": 36410 }, { "grad_norm": 0.9120782017707825, "learning_rate": 6.143297002216892e-05, "loss": 0.0343, "step": 36420 }, { "grad_norm": 0.34444937109947205, "learning_rate": 6.141284825104882e-05, "loss": 0.0354, "step": 36430 }, { "grad_norm": 0.29360857605934143, "learning_rate": 6.13927245297851e-05, "loss": 0.0327, "step": 36440 }, { "grad_norm": 0.29379042983055115, "learning_rate": 6.137259886181633e-05, "loss": 0.0369, "step": 36450 }, { "grad_norm": 0.25915563106536865, "learning_rate": 6.135247125058145e-05, "loss": 0.0353, "step": 36460 }, { "grad_norm": 3.541867971420288, "learning_rate": 6.133234169951974e-05, "loss": 0.0222, "step": 36470 }, { "grad_norm": 0.4042527675628662, "learning_rate": 6.131221021207078e-05, "loss": 0.0358, "step": 36480 }, { "grad_norm": 0.4463178515434265, "learning_rate": 6.129207679167448e-05, "loss": 0.0295, "step": 36490 }, { "grad_norm": 0.2903721034526825, "learning_rate": 6.127194144177109e-05, "loss": 0.0272, "step": 36500 }, { "grad_norm": 0.2694246768951416, "learning_rate": 6.125180416580118e-05, "loss": 0.0263, "step": 36510 }, { "grad_norm": 0.29510948061943054, "learning_rate": 6.123166496720571e-05, "loss": 0.0309, "step": 36520 }, { "grad_norm": 0.28525692224502563, "learning_rate": 6.121152384942588e-05, "loss": 0.027, "step": 36530 }, { "grad_norm": 0.29320403933525085, "learning_rate": 6.119138081590324e-05, "loss": 0.0276, "step": 36540 }, { "grad_norm": 0.4361952245235443, "learning_rate": 6.117123587007971e-05, "loss": 0.0278, "step": 36550 }, { "grad_norm": 0.4989640712738037, "learning_rate": 6.11510890153975e-05, "loss": 0.0383, "step": 36560 }, { "grad_norm": 0.33289751410484314, "learning_rate": 6.113094025529916e-05, "loss": 0.0239, "step": 36570 }, { "grad_norm": 0.3277558386325836, "learning_rate": 6.111078959322757e-05, "loss": 0.0409, "step": 36580 }, { "grad_norm": 0.3839488923549652, "learning_rate": 6.109063703262592e-05, "loss": 0.0387, "step": 36590 }, { "grad_norm": 0.6112929582595825, "learning_rate": 6.107048257693772e-05, "loss": 0.0299, "step": 36600 }, { "grad_norm": 0.37606361508369446, "learning_rate": 6.105032622960683e-05, "loss": 0.0293, "step": 36610 }, { "grad_norm": 0.4278032183647156, "learning_rate": 6.103016799407743e-05, "loss": 0.0296, "step": 36620 }, { "grad_norm": 0.33128243684768677, "learning_rate": 6.1010007873793984e-05, "loss": 0.037, "step": 36630 }, { "grad_norm": 0.26379698514938354, "learning_rate": 6.098984587220131e-05, "loss": 0.0305, "step": 36640 }, { "grad_norm": 0.34327957034111023, "learning_rate": 6.096968199274456e-05, "loss": 0.0277, "step": 36650 }, { "grad_norm": 0.3781611919403076, "learning_rate": 6.0949516238869166e-05, "loss": 0.0353, "step": 36660 }, { "grad_norm": 0.4194372594356537, "learning_rate": 6.092934861402092e-05, "loss": 0.0399, "step": 36670 }, { "grad_norm": 0.31545963883399963, "learning_rate": 6.0909179121645924e-05, "loss": 0.0359, "step": 36680 }, { "grad_norm": 0.3504061996936798, "learning_rate": 6.0889007765190576e-05, "loss": 0.038, "step": 36690 }, { "grad_norm": 0.23870298266410828, "learning_rate": 6.086883454810162e-05, "loss": 0.028, "step": 36700 }, { "grad_norm": 0.4750116169452667, "learning_rate": 6.0848659473826084e-05, "loss": 0.0321, "step": 36710 }, { "grad_norm": 0.4328903257846832, "learning_rate": 6.082848254581138e-05, "loss": 0.0311, "step": 36720 }, { "grad_norm": 0.36220112442970276, "learning_rate": 6.080830376750517e-05, "loss": 0.0313, "step": 36730 }, { "grad_norm": 0.29101574420928955, "learning_rate": 6.0788123142355445e-05, "loss": 0.0289, "step": 36740 }, { "grad_norm": 0.4172380566596985, "learning_rate": 6.076794067381052e-05, "loss": 0.0323, "step": 36750 }, { "grad_norm": 0.2954413890838623, "learning_rate": 6.074775636531905e-05, "loss": 0.0296, "step": 36760 }, { "grad_norm": 0.4056990444660187, "learning_rate": 6.072757022032997e-05, "loss": 0.0292, "step": 36770 }, { "grad_norm": 0.2663024365901947, "learning_rate": 6.070738224229253e-05, "loss": 0.0376, "step": 36780 }, { "grad_norm": 0.2804863154888153, "learning_rate": 6.0687192434656314e-05, "loss": 0.0314, "step": 36790 }, { "grad_norm": 0.41566652059555054, "learning_rate": 6.066700080087121e-05, "loss": 0.0314, "step": 36800 }, { "grad_norm": 0.31411096453666687, "learning_rate": 6.0646807344387424e-05, "loss": 0.0307, "step": 36810 }, { "grad_norm": 0.44134801626205444, "learning_rate": 6.062661206865543e-05, "loss": 0.0339, "step": 36820 }, { "grad_norm": 0.28686144948005676, "learning_rate": 6.06064149771261e-05, "loss": 0.0292, "step": 36830 }, { "grad_norm": 0.2835904657840729, "learning_rate": 6.058621607325051e-05, "loss": 0.0346, "step": 36840 }, { "grad_norm": 0.317917138338089, "learning_rate": 6.056601536048014e-05, "loss": 0.0309, "step": 36850 }, { "grad_norm": 0.435265451669693, "learning_rate": 6.0545812842266725e-05, "loss": 0.0392, "step": 36860 }, { "grad_norm": 0.30087152123451233, "learning_rate": 6.052560852206232e-05, "loss": 0.0315, "step": 36870 }, { "grad_norm": 0.34019735455513, "learning_rate": 6.05054024033193e-05, "loss": 0.0417, "step": 36880 }, { "grad_norm": 0.3332081437110901, "learning_rate": 6.048519448949032e-05, "loss": 0.0307, "step": 36890 }, { "grad_norm": 0.5019272565841675, "learning_rate": 6.046498478402839e-05, "loss": 0.0307, "step": 36900 }, { "grad_norm": 0.29471880197525024, "learning_rate": 6.044477329038677e-05, "loss": 0.0271, "step": 36910 }, { "grad_norm": 0.5927501320838928, "learning_rate": 6.042456001201906e-05, "loss": 0.0299, "step": 36920 }, { "grad_norm": 0.6104077100753784, "learning_rate": 6.040434495237917e-05, "loss": 0.0282, "step": 36930 }, { "grad_norm": 0.3731870651245117, "learning_rate": 6.0384128114921256e-05, "loss": 0.0254, "step": 36940 }, { "grad_norm": 0.3166201114654541, "learning_rate": 6.036390950309987e-05, "loss": 0.0249, "step": 36950 }, { "grad_norm": 0.33500197529792786, "learning_rate": 6.0343689120369805e-05, "loss": 0.0326, "step": 36960 }, { "grad_norm": 0.3994349241256714, "learning_rate": 6.032346697018616e-05, "loss": 0.0241, "step": 36970 }, { "grad_norm": 0.24223290383815765, "learning_rate": 6.0303243056004375e-05, "loss": 0.0426, "step": 36980 }, { "grad_norm": 0.4132881164550781, "learning_rate": 6.0283017381280136e-05, "loss": 0.0326, "step": 36990 }, { "grad_norm": 0.4562288224697113, "learning_rate": 6.026278994946948e-05, "loss": 0.0277, "step": 37000 }, { "grad_norm": 0.26865097880363464, "learning_rate": 6.02425607640287e-05, "loss": 0.034, "step": 37010 }, { "grad_norm": 0.3816227614879608, "learning_rate": 6.022232982841441e-05, "loss": 0.0261, "step": 37020 }, { "grad_norm": 0.3296733796596527, "learning_rate": 6.020209714608355e-05, "loss": 0.0283, "step": 37030 }, { "grad_norm": 0.4056899845600128, "learning_rate": 6.018186272049331e-05, "loss": 0.0316, "step": 37040 }, { "grad_norm": 0.2489434778690338, "learning_rate": 6.01616265551012e-05, "loss": 0.0305, "step": 37050 }, { "grad_norm": 0.26176515221595764, "learning_rate": 6.014138865336503e-05, "loss": 0.0278, "step": 37060 }, { "grad_norm": 0.2933286726474762, "learning_rate": 6.0121149018742905e-05, "loss": 0.0325, "step": 37070 }, { "grad_norm": 0.5157597064971924, "learning_rate": 6.010090765469325e-05, "loss": 0.0278, "step": 37080 }, { "grad_norm": 0.28103065490722656, "learning_rate": 6.008066456467473e-05, "loss": 0.03, "step": 37090 }, { "grad_norm": 0.3860423266887665, "learning_rate": 6.0060419752146335e-05, "loss": 0.0269, "step": 37100 }, { "grad_norm": 0.34103497862815857, "learning_rate": 6.0040173220567353e-05, "loss": 0.0264, "step": 37110 }, { "grad_norm": 0.33077946305274963, "learning_rate": 6.001992497339737e-05, "loss": 0.0268, "step": 37120 }, { "grad_norm": 0.34890255331993103, "learning_rate": 5.999967501409626e-05, "loss": 0.0379, "step": 37130 }, { "grad_norm": 0.2469475269317627, "learning_rate": 5.997942334612418e-05, "loss": 0.0253, "step": 37140 }, { "grad_norm": 0.30555734038352966, "learning_rate": 5.995916997294158e-05, "loss": 0.02, "step": 37150 }, { "grad_norm": 0.3244035542011261, "learning_rate": 5.9938914898009214e-05, "loss": 0.0346, "step": 37160 }, { "grad_norm": 0.797245442867279, "learning_rate": 5.991865812478813e-05, "loss": 0.0267, "step": 37170 }, { "grad_norm": 0.39219099283218384, "learning_rate": 5.989839965673964e-05, "loss": 0.0413, "step": 37180 }, { "grad_norm": 0.3081209361553192, "learning_rate": 5.987813949732539e-05, "loss": 0.029, "step": 37190 }, { "grad_norm": 0.2903214991092682, "learning_rate": 5.9857877650007255e-05, "loss": 0.028, "step": 37200 }, { "grad_norm": 0.27829286456108093, "learning_rate": 5.983761411824744e-05, "loss": 0.0246, "step": 37210 }, { "grad_norm": 0.3116350471973419, "learning_rate": 5.981734890550844e-05, "loss": 0.0264, "step": 37220 }, { "grad_norm": 0.35564419627189636, "learning_rate": 5.979708201525301e-05, "loss": 0.0359, "step": 37230 }, { "grad_norm": 0.4371447265148163, "learning_rate": 5.977681345094422e-05, "loss": 0.0329, "step": 37240 }, { "grad_norm": 0.34222427010536194, "learning_rate": 5.97565432160454e-05, "loss": 0.0285, "step": 37250 }, { "grad_norm": 0.263795405626297, "learning_rate": 5.9736271314020186e-05, "loss": 0.0326, "step": 37260 }, { "grad_norm": 0.4301188588142395, "learning_rate": 5.971599774833251e-05, "loss": 0.031, "step": 37270 }, { "grad_norm": 0.2961646020412445, "learning_rate": 5.9695722522446525e-05, "loss": 0.0269, "step": 37280 }, { "grad_norm": 0.36615121364593506, "learning_rate": 5.9675445639826765e-05, "loss": 0.0278, "step": 37290 }, { "grad_norm": 0.24894575774669647, "learning_rate": 5.965516710393796e-05, "loss": 0.0293, "step": 37300 }, { "grad_norm": 0.3121899366378784, "learning_rate": 5.963488691824516e-05, "loss": 0.0342, "step": 37310 }, { "grad_norm": 0.3546648323535919, "learning_rate": 5.96146050862137e-05, "loss": 0.0435, "step": 37320 }, { "grad_norm": 0.32028064131736755, "learning_rate": 5.959432161130919e-05, "loss": 0.0261, "step": 37330 }, { "grad_norm": 0.3400379717350006, "learning_rate": 5.9574036496997545e-05, "loss": 0.0342, "step": 37340 }, { "grad_norm": 0.4180845022201538, "learning_rate": 5.955374974674488e-05, "loss": 0.0268, "step": 37350 }, { "grad_norm": 0.28008678555488586, "learning_rate": 5.9533461364017696e-05, "loss": 0.0274, "step": 37360 }, { "grad_norm": 0.23971226811408997, "learning_rate": 5.9513171352282716e-05, "loss": 0.024, "step": 37370 }, { "grad_norm": 0.31742459535598755, "learning_rate": 5.949287971500692e-05, "loss": 0.0296, "step": 37380 }, { "grad_norm": 0.3308100998401642, "learning_rate": 5.947258645565762e-05, "loss": 0.0247, "step": 37390 }, { "grad_norm": 0.2517683804035187, "learning_rate": 5.945229157770237e-05, "loss": 0.0301, "step": 37400 }, { "grad_norm": 0.32650837302207947, "learning_rate": 5.9431995084609006e-05, "loss": 0.0325, "step": 37410 }, { "grad_norm": 0.36081239581108093, "learning_rate": 5.941169697984564e-05, "loss": 0.0267, "step": 37420 }, { "grad_norm": 0.3771057724952698, "learning_rate": 5.9391397266880675e-05, "loss": 0.0276, "step": 37430 }, { "grad_norm": 0.4915980100631714, "learning_rate": 5.937109594918279e-05, "loss": 0.0466, "step": 37440 }, { "grad_norm": 0.7301997542381287, "learning_rate": 5.9350793030220884e-05, "loss": 0.0482, "step": 37450 }, { "grad_norm": 0.36464717984199524, "learning_rate": 5.933048851346421e-05, "loss": 0.0293, "step": 37460 }, { "grad_norm": 0.3691125810146332, "learning_rate": 5.931018240238224e-05, "loss": 0.0425, "step": 37470 }, { "grad_norm": 0.2750648558139801, "learning_rate": 5.928987470044471e-05, "loss": 0.0341, "step": 37480 }, { "grad_norm": 0.43360233306884766, "learning_rate": 5.9269565411121695e-05, "loss": 0.0295, "step": 37490 }, { "grad_norm": 0.4037584066390991, "learning_rate": 5.924925453788347e-05, "loss": 0.0349, "step": 37500 }, { "grad_norm": 0.5268147587776184, "learning_rate": 5.92289420842006e-05, "loss": 0.0344, "step": 37510 }, { "grad_norm": 0.41911473870277405, "learning_rate": 5.9208628053543945e-05, "loss": 0.0403, "step": 37520 }, { "grad_norm": 0.396742582321167, "learning_rate": 5.918831244938462e-05, "loss": 0.0286, "step": 37530 }, { "grad_norm": 0.48382070660591125, "learning_rate": 5.9167995275194e-05, "loss": 0.0298, "step": 37540 }, { "grad_norm": 0.4300812780857086, "learning_rate": 5.914767653444373e-05, "loss": 0.0304, "step": 37550 }, { "grad_norm": 1.0079452991485596, "learning_rate": 5.912735623060572e-05, "loss": 0.0377, "step": 37560 }, { "grad_norm": 0.47785553336143494, "learning_rate": 5.910703436715217e-05, "loss": 0.0323, "step": 37570 }, { "grad_norm": 0.2887279987335205, "learning_rate": 5.908671094755552e-05, "loss": 0.0245, "step": 37580 }, { "grad_norm": 0.3516062796115875, "learning_rate": 5.906638597528851e-05, "loss": 0.0272, "step": 37590 }, { "grad_norm": 0.25997960567474365, "learning_rate": 5.9046059453824076e-05, "loss": 0.0262, "step": 37600 }, { "grad_norm": 0.302402138710022, "learning_rate": 5.9025731386635505e-05, "loss": 0.033, "step": 37610 }, { "grad_norm": 0.3893722593784332, "learning_rate": 5.900540177719629e-05, "loss": 0.0323, "step": 37620 }, { "grad_norm": 0.33993709087371826, "learning_rate": 5.898507062898021e-05, "loss": 0.0308, "step": 37630 }, { "grad_norm": 0.46362408995628357, "learning_rate": 5.8964737945461316e-05, "loss": 0.0269, "step": 37640 }, { "grad_norm": 0.2239937037229538, "learning_rate": 5.8944403730113885e-05, "loss": 0.0308, "step": 37650 }, { "grad_norm": 0.42838144302368164, "learning_rate": 5.892406798641248e-05, "loss": 0.032, "step": 37660 }, { "grad_norm": 0.4040905237197876, "learning_rate": 5.890373071783193e-05, "loss": 0.0242, "step": 37670 }, { "grad_norm": 0.31504112482070923, "learning_rate": 5.888339192784732e-05, "loss": 0.0314, "step": 37680 }, { "grad_norm": 0.37580418586730957, "learning_rate": 5.8863051619934003e-05, "loss": 0.0253, "step": 37690 }, { "grad_norm": 0.38360440731048584, "learning_rate": 5.8842709797567554e-05, "loss": 0.031, "step": 37700 }, { "grad_norm": 0.34578269720077515, "learning_rate": 5.8822366464223855e-05, "loss": 0.0271, "step": 37710 }, { "grad_norm": 0.22501972317695618, "learning_rate": 5.880202162337901e-05, "loss": 0.036, "step": 37720 }, { "grad_norm": 0.376505970954895, "learning_rate": 5.8781675278509405e-05, "loss": 0.0288, "step": 37730 }, { "grad_norm": 0.3641703724861145, "learning_rate": 5.8761327433091696e-05, "loss": 0.0283, "step": 37740 }, { "grad_norm": 0.28984248638153076, "learning_rate": 5.874097809060275e-05, "loss": 0.0345, "step": 37750 }, { "grad_norm": 0.5384408831596375, "learning_rate": 5.87206272545197e-05, "loss": 0.0475, "step": 37760 }, { "grad_norm": 0.3491349518299103, "learning_rate": 5.8700274928319955e-05, "loss": 0.0413, "step": 37770 }, { "grad_norm": 0.4251168668270111, "learning_rate": 5.867992111548118e-05, "loss": 0.0305, "step": 37780 }, { "grad_norm": 0.345044881105423, "learning_rate": 5.865956581948131e-05, "loss": 0.0294, "step": 37790 }, { "grad_norm": 0.5255987048149109, "learning_rate": 5.863920904379845e-05, "loss": 0.0288, "step": 37800 }, { "grad_norm": 0.5750481486320496, "learning_rate": 5.8618850791911064e-05, "loss": 0.0348, "step": 37810 }, { "grad_norm": 0.32179513573646545, "learning_rate": 5.859849106729779e-05, "loss": 0.03, "step": 37820 }, { "grad_norm": 0.354017049074173, "learning_rate": 5.857812987343758e-05, "loss": 0.0291, "step": 37830 }, { "grad_norm": 0.3954960107803345, "learning_rate": 5.855776721380957e-05, "loss": 0.0297, "step": 37840 }, { "grad_norm": 0.3592546284198761, "learning_rate": 5.8537403091893217e-05, "loss": 0.03, "step": 37850 }, { "grad_norm": 0.3028241693973541, "learning_rate": 5.851703751116816e-05, "loss": 0.0289, "step": 37860 }, { "grad_norm": 0.4646967351436615, "learning_rate": 5.8496670475114336e-05, "loss": 0.0281, "step": 37870 }, { "grad_norm": 0.582629919052124, "learning_rate": 5.84763019872119e-05, "loss": 0.0417, "step": 37880 }, { "grad_norm": 0.38756516575813293, "learning_rate": 5.845593205094131e-05, "loss": 0.0271, "step": 37890 }, { "grad_norm": 0.33045855164527893, "learning_rate": 5.843556066978318e-05, "loss": 0.0276, "step": 37900 }, { "grad_norm": 0.2573586702346802, "learning_rate": 5.8415187847218455e-05, "loss": 0.0341, "step": 37910 }, { "grad_norm": 0.3783807158470154, "learning_rate": 5.839481358672827e-05, "loss": 0.0242, "step": 37920 }, { "grad_norm": 0.3767474591732025, "learning_rate": 5.837443789179407e-05, "loss": 0.0315, "step": 37930 }, { "grad_norm": 0.45145294070243835, "learning_rate": 5.8354060765897445e-05, "loss": 0.0289, "step": 37940 }, { "grad_norm": 1.0102245807647705, "learning_rate": 5.8333682212520334e-05, "loss": 0.0417, "step": 37950 }, { "grad_norm": 0.4155523478984833, "learning_rate": 5.831330223514486e-05, "loss": 0.0318, "step": 37960 }, { "grad_norm": 0.2744986116886139, "learning_rate": 5.8292920837253396e-05, "loss": 0.0255, "step": 37970 }, { "grad_norm": 0.32430100440979004, "learning_rate": 5.827253802232857e-05, "loss": 0.0272, "step": 37980 }, { "grad_norm": 0.5293289422988892, "learning_rate": 5.825215379385327e-05, "loss": 0.0274, "step": 37990 }, { "grad_norm": 0.7055168747901917, "learning_rate": 5.823176815531057e-05, "loss": 0.0466, "step": 38000 }, { "grad_norm": 0.2730239927768707, "learning_rate": 5.8211381110183826e-05, "loss": 0.0376, "step": 38010 }, { "grad_norm": 0.32390159368515015, "learning_rate": 5.8190992661956645e-05, "loss": 0.0277, "step": 38020 }, { "grad_norm": 0.2376878559589386, "learning_rate": 5.817060281411284e-05, "loss": 0.0303, "step": 38030 }, { "grad_norm": 0.30301105976104736, "learning_rate": 5.815021157013647e-05, "loss": 0.0255, "step": 38040 }, { "grad_norm": 0.337772011756897, "learning_rate": 5.8129818933511856e-05, "loss": 0.0361, "step": 38050 }, { "grad_norm": 0.8631503582000732, "learning_rate": 5.8109424907723544e-05, "loss": 0.0317, "step": 38060 }, { "grad_norm": 0.25325754284858704, "learning_rate": 5.80890294962563e-05, "loss": 0.0202, "step": 38070 }, { "grad_norm": 0.43904057145118713, "learning_rate": 5.806863270259515e-05, "loss": 0.0328, "step": 38080 }, { "grad_norm": 0.3182682991027832, "learning_rate": 5.804823453022536e-05, "loss": 0.0297, "step": 38090 }, { "grad_norm": 0.30745550990104675, "learning_rate": 5.80278349826324e-05, "loss": 0.0304, "step": 38100 }, { "grad_norm": 0.3130139112472534, "learning_rate": 5.8007434063302014e-05, "loss": 0.0337, "step": 38110 }, { "grad_norm": 0.3368528485298157, "learning_rate": 5.7987031775720136e-05, "loss": 0.0328, "step": 38120 }, { "grad_norm": 0.2084844559431076, "learning_rate": 5.7966628123372976e-05, "loss": 0.0334, "step": 38130 }, { "grad_norm": 0.49202194809913635, "learning_rate": 5.7946223109746956e-05, "loss": 0.043, "step": 38140 }, { "grad_norm": 0.30729442834854126, "learning_rate": 5.7925816738328754e-05, "loss": 0.0208, "step": 38150 }, { "grad_norm": 0.837124228477478, "learning_rate": 5.790540901260521e-05, "loss": 0.0295, "step": 38160 }, { "grad_norm": 0.6195530891418457, "learning_rate": 5.788499993606351e-05, "loss": 0.0334, "step": 38170 }, { "grad_norm": 0.6430031657218933, "learning_rate": 5.786458951219096e-05, "loss": 0.0447, "step": 38180 }, { "grad_norm": 0.3526427447795868, "learning_rate": 5.784417774447517e-05, "loss": 0.0346, "step": 38190 }, { "grad_norm": 0.4887782633304596, "learning_rate": 5.782376463640393e-05, "loss": 0.0301, "step": 38200 }, { "grad_norm": 0.4550520181655884, "learning_rate": 5.780335019146531e-05, "loss": 0.0289, "step": 38210 }, { "grad_norm": 0.31784787774086, "learning_rate": 5.778293441314755e-05, "loss": 0.0364, "step": 38220 }, { "grad_norm": 0.3925563097000122, "learning_rate": 5.776251730493917e-05, "loss": 0.037, "step": 38230 }, { "grad_norm": 0.4432201385498047, "learning_rate": 5.774209887032887e-05, "loss": 0.0297, "step": 38240 }, { "grad_norm": 0.3750755488872528, "learning_rate": 5.772167911280565e-05, "loss": 0.0351, "step": 38250 }, { "grad_norm": 0.3242511451244354, "learning_rate": 5.770125803585864e-05, "loss": 0.0253, "step": 38260 }, { "grad_norm": 0.30089759826660156, "learning_rate": 5.768083564297726e-05, "loss": 0.0399, "step": 38270 }, { "grad_norm": 0.3343295156955719, "learning_rate": 5.766041193765114e-05, "loss": 0.032, "step": 38280 }, { "grad_norm": 0.3244801163673401, "learning_rate": 5.763998692337015e-05, "loss": 0.0257, "step": 38290 }, { "grad_norm": 0.46534135937690735, "learning_rate": 5.761956060362433e-05, "loss": 0.033, "step": 38300 }, { "grad_norm": 0.41438594460487366, "learning_rate": 5.7599132981904e-05, "loss": 0.0308, "step": 38310 }, { "grad_norm": 0.3411646783351898, "learning_rate": 5.75787040616997e-05, "loss": 0.0339, "step": 38320 }, { "grad_norm": 0.3257233500480652, "learning_rate": 5.755827384650212e-05, "loss": 0.0283, "step": 38330 }, { "grad_norm": 0.31304511427879333, "learning_rate": 5.753784233980228e-05, "loss": 0.0276, "step": 38340 }, { "grad_norm": 0.2837987244129181, "learning_rate": 5.751740954509135e-05, "loss": 0.0233, "step": 38350 }, { "grad_norm": 0.26321500539779663, "learning_rate": 5.7496975465860715e-05, "loss": 0.0225, "step": 38360 }, { "grad_norm": 0.28138381242752075, "learning_rate": 5.747654010560202e-05, "loss": 0.0255, "step": 38370 }, { "grad_norm": 0.4096807837486267, "learning_rate": 5.7456103467807097e-05, "loss": 0.0351, "step": 38380 }, { "grad_norm": 0.36978644132614136, "learning_rate": 5.7435665555968046e-05, "loss": 0.0266, "step": 38390 }, { "grad_norm": 0.31268376111984253, "learning_rate": 5.74152263735771e-05, "loss": 0.0312, "step": 38400 }, { "grad_norm": 0.40667465329170227, "learning_rate": 5.739478592412677e-05, "loss": 0.0282, "step": 38410 }, { "grad_norm": 0.46199601888656616, "learning_rate": 5.7374344211109766e-05, "loss": 0.0321, "step": 38420 }, { "grad_norm": 0.2949558198451996, "learning_rate": 5.735390123801904e-05, "loss": 0.0378, "step": 38430 }, { "grad_norm": 0.4350360333919525, "learning_rate": 5.7333457008347704e-05, "loss": 0.0333, "step": 38440 }, { "grad_norm": 0.34704551100730896, "learning_rate": 5.7313011525589156e-05, "loss": 0.0372, "step": 38450 }, { "grad_norm": 0.2946770489215851, "learning_rate": 5.729256479323694e-05, "loss": 0.0283, "step": 38460 }, { "grad_norm": 0.3311002850532532, "learning_rate": 5.727211681478485e-05, "loss": 0.0268, "step": 38470 }, { "grad_norm": 0.4003371000289917, "learning_rate": 5.7251667593726886e-05, "loss": 0.0325, "step": 38480 }, { "grad_norm": 0.23049381375312805, "learning_rate": 5.723121713355728e-05, "loss": 0.036, "step": 38490 }, { "grad_norm": 0.2694346010684967, "learning_rate": 5.721076543777044e-05, "loss": 0.0289, "step": 38500 }, { "grad_norm": 0.31259360909461975, "learning_rate": 5.7190312509860986e-05, "loss": 0.0257, "step": 38510 }, { "grad_norm": 0.5062063932418823, "learning_rate": 5.716985835332379e-05, "loss": 0.0285, "step": 38520 }, { "grad_norm": 0.49277329444885254, "learning_rate": 5.714940297165389e-05, "loss": 0.0307, "step": 38530 }, { "grad_norm": 0.34848934412002563, "learning_rate": 5.712894636834656e-05, "loss": 0.0287, "step": 38540 }, { "grad_norm": 0.38040441274642944, "learning_rate": 5.7108488546897276e-05, "loss": 0.024, "step": 38550 }, { "grad_norm": 0.187747061252594, "learning_rate": 5.708802951080172e-05, "loss": 0.0252, "step": 38560 }, { "grad_norm": 0.3692777752876282, "learning_rate": 5.706756926355576e-05, "loss": 0.0344, "step": 38570 }, { "grad_norm": 0.42204686999320984, "learning_rate": 5.704710780865554e-05, "loss": 0.0293, "step": 38580 }, { "grad_norm": 0.5214183926582336, "learning_rate": 5.7026645149597325e-05, "loss": 0.0375, "step": 38590 }, { "grad_norm": 0.3371730446815491, "learning_rate": 5.700618128987764e-05, "loss": 0.0334, "step": 38600 }, { "grad_norm": 0.2815958857536316, "learning_rate": 5.698571623299317e-05, "loss": 0.0275, "step": 38610 }, { "grad_norm": 0.44122031331062317, "learning_rate": 5.696524998244086e-05, "loss": 0.0267, "step": 38620 }, { "grad_norm": 0.39596831798553467, "learning_rate": 5.6944782541717836e-05, "loss": 0.0322, "step": 38630 }, { "grad_norm": 0.22158662974834442, "learning_rate": 5.69243139143214e-05, "loss": 0.0314, "step": 38640 }, { "grad_norm": 0.30924105644226074, "learning_rate": 5.6903844103749125e-05, "loss": 0.0294, "step": 38650 }, { "grad_norm": 0.2941252887248993, "learning_rate": 5.688337311349869e-05, "loss": 0.0267, "step": 38660 }, { "grad_norm": 0.26138827204704285, "learning_rate": 5.6862900947068074e-05, "loss": 0.0246, "step": 38670 }, { "grad_norm": 0.38032057881355286, "learning_rate": 5.6842427607955374e-05, "loss": 0.0326, "step": 38680 }, { "grad_norm": 0.3536189794540405, "learning_rate": 5.682195309965893e-05, "loss": 0.0257, "step": 38690 }, { "grad_norm": 0.5137813687324524, "learning_rate": 5.6801477425677294e-05, "loss": 0.0325, "step": 38700 }, { "grad_norm": 0.36035463213920593, "learning_rate": 5.678100058950917e-05, "loss": 0.0376, "step": 38710 }, { "grad_norm": 0.4895799458026886, "learning_rate": 5.676052259465352e-05, "loss": 0.0302, "step": 38720 }, { "grad_norm": 0.41357895731925964, "learning_rate": 5.674004344460945e-05, "loss": 0.0239, "step": 38730 }, { "grad_norm": 0.5754876136779785, "learning_rate": 5.6719563142876295e-05, "loss": 0.026, "step": 38740 }, { "grad_norm": 0.3919346332550049, "learning_rate": 5.669908169295359e-05, "loss": 0.0252, "step": 38750 }, { "grad_norm": 0.2661491334438324, "learning_rate": 5.667859909834105e-05, "loss": 0.0306, "step": 38760 }, { "grad_norm": 0.32267624139785767, "learning_rate": 5.6658115362538565e-05, "loss": 0.0307, "step": 38770 }, { "grad_norm": 0.32478466629981995, "learning_rate": 5.663763048904628e-05, "loss": 0.0289, "step": 38780 }, { "grad_norm": 0.4276479482650757, "learning_rate": 5.661714448136447e-05, "loss": 0.0229, "step": 38790 }, { "grad_norm": 0.2755756974220276, "learning_rate": 5.659665734299366e-05, "loss": 0.0203, "step": 38800 }, { "grad_norm": 0.4520646631717682, "learning_rate": 5.6576169077434516e-05, "loss": 0.0307, "step": 38810 }, { "grad_norm": 0.31668367981910706, "learning_rate": 5.6555679688187944e-05, "loss": 0.0277, "step": 38820 }, { "grad_norm": 0.5325462222099304, "learning_rate": 5.6535189178755e-05, "loss": 0.0296, "step": 38830 }, { "grad_norm": 0.31691786646842957, "learning_rate": 5.651469755263698e-05, "loss": 0.026, "step": 38840 }, { "grad_norm": 0.20497770607471466, "learning_rate": 5.6494204813335316e-05, "loss": 0.0345, "step": 38850 }, { "grad_norm": 0.20394879579544067, "learning_rate": 5.647371096435168e-05, "loss": 0.0227, "step": 38860 }, { "grad_norm": 0.2561553120613098, "learning_rate": 5.645321600918788e-05, "loss": 0.0243, "step": 38870 }, { "grad_norm": 0.47605088353157043, "learning_rate": 5.643271995134597e-05, "loss": 0.0349, "step": 38880 }, { "grad_norm": 0.32952603697776794, "learning_rate": 5.641222279432814e-05, "loss": 0.0276, "step": 38890 }, { "grad_norm": 0.5964363813400269, "learning_rate": 5.6391724541636834e-05, "loss": 0.0264, "step": 38900 }, { "grad_norm": 0.6220915913581848, "learning_rate": 5.6371225196774605e-05, "loss": 0.0319, "step": 38910 }, { "grad_norm": 0.31791016459465027, "learning_rate": 5.635072476324423e-05, "loss": 0.0353, "step": 38920 }, { "grad_norm": 0.3700740337371826, "learning_rate": 5.63302232445487e-05, "loss": 0.0271, "step": 38930 }, { "grad_norm": 0.25947290658950806, "learning_rate": 5.6309720644191144e-05, "loss": 0.0359, "step": 38940 }, { "grad_norm": 0.41005653142929077, "learning_rate": 5.628921696567491e-05, "loss": 0.0247, "step": 38950 }, { "grad_norm": 0.5255761742591858, "learning_rate": 5.62687122125035e-05, "loss": 0.0345, "step": 38960 }, { "grad_norm": 0.33864110708236694, "learning_rate": 5.624820638818062e-05, "loss": 0.0295, "step": 38970 }, { "grad_norm": 0.4948095977306366, "learning_rate": 5.6227699496210164e-05, "loss": 0.0402, "step": 38980 }, { "grad_norm": 0.2696906328201294, "learning_rate": 5.6207191540096195e-05, "loss": 0.029, "step": 38990 }, { "grad_norm": 0.32292330265045166, "learning_rate": 5.618668252334296e-05, "loss": 0.0211, "step": 39000 }, { "grad_norm": 0.3128988742828369, "learning_rate": 5.616617244945488e-05, "loss": 0.0228, "step": 39010 }, { "grad_norm": 0.3558002710342407, "learning_rate": 5.614566132193656e-05, "loss": 0.0266, "step": 39020 }, { "grad_norm": 0.3729517161846161, "learning_rate": 5.612514914429282e-05, "loss": 0.0258, "step": 39030 }, { "grad_norm": 0.3563621938228607, "learning_rate": 5.610463592002863e-05, "loss": 0.0259, "step": 39040 }, { "grad_norm": 0.33741146326065063, "learning_rate": 5.608412165264909e-05, "loss": 0.029, "step": 39050 }, { "grad_norm": 0.4234406054019928, "learning_rate": 5.606360634565959e-05, "loss": 0.0258, "step": 39060 }, { "grad_norm": 0.18913881480693817, "learning_rate": 5.604309000256559e-05, "loss": 0.0293, "step": 39070 }, { "grad_norm": 0.2840020954608917, "learning_rate": 5.6022572626872785e-05, "loss": 0.0292, "step": 39080 }, { "grad_norm": 0.4055667221546173, "learning_rate": 5.600205422208704e-05, "loss": 0.0341, "step": 39090 }, { "grad_norm": 0.2585505545139313, "learning_rate": 5.59815347917144e-05, "loss": 0.0345, "step": 39100 }, { "grad_norm": 0.4129590094089508, "learning_rate": 5.596101433926103e-05, "loss": 0.0328, "step": 39110 }, { "grad_norm": 0.3809181749820709, "learning_rate": 5.5940492868233364e-05, "loss": 0.028, "step": 39120 }, { "grad_norm": 0.41463401913642883, "learning_rate": 5.591997038213793e-05, "loss": 0.0347, "step": 39130 }, { "grad_norm": 0.33214429020881653, "learning_rate": 5.5899446884481475e-05, "loss": 0.0285, "step": 39140 }, { "grad_norm": 0.33700668811798096, "learning_rate": 5.5878922378770906e-05, "loss": 0.0401, "step": 39150 }, { "grad_norm": 0.26555532217025757, "learning_rate": 5.5858396868513285e-05, "loss": 0.0309, "step": 39160 }, { "grad_norm": 0.3599572777748108, "learning_rate": 5.583787035721586e-05, "loss": 0.0223, "step": 39170 }, { "grad_norm": 0.49511057138442993, "learning_rate": 5.581734284838606e-05, "loss": 0.0286, "step": 39180 }, { "grad_norm": 0.34578952193260193, "learning_rate": 5.579681434553147e-05, "loss": 0.0278, "step": 39190 }, { "grad_norm": 1.2374833822250366, "learning_rate": 5.5776284852159854e-05, "loss": 0.025, "step": 39200 }, { "grad_norm": 0.37836888432502747, "learning_rate": 5.575575437177913e-05, "loss": 0.0422, "step": 39210 }, { "grad_norm": 0.37991881370544434, "learning_rate": 5.573522290789742e-05, "loss": 0.0251, "step": 39220 }, { "grad_norm": 0.26898080110549927, "learning_rate": 5.571469046402298e-05, "loss": 0.0276, "step": 39230 }, { "grad_norm": 0.2573591470718384, "learning_rate": 5.5694157043664205e-05, "loss": 0.0346, "step": 39240 }, { "grad_norm": 0.5659496784210205, "learning_rate": 5.567362265032975e-05, "loss": 0.0246, "step": 39250 }, { "grad_norm": 0.33986103534698486, "learning_rate": 5.565308728752836e-05, "loss": 0.029, "step": 39260 }, { "grad_norm": 0.34921789169311523, "learning_rate": 5.5632550958768956e-05, "loss": 0.0283, "step": 39270 }, { "grad_norm": 0.4401642084121704, "learning_rate": 5.5612013667560636e-05, "loss": 0.0301, "step": 39280 }, { "grad_norm": 0.4551072418689728, "learning_rate": 5.5591475417412676e-05, "loss": 0.0431, "step": 39290 }, { "grad_norm": 0.33895277976989746, "learning_rate": 5.557093621183451e-05, "loss": 0.0338, "step": 39300 }, { "grad_norm": 0.2507670521736145, "learning_rate": 5.55503960543357e-05, "loss": 0.0296, "step": 39310 }, { "grad_norm": 0.4414438307285309, "learning_rate": 5.552985494842601e-05, "loss": 0.0289, "step": 39320 }, { "grad_norm": 0.2560226619243622, "learning_rate": 5.550931289761534e-05, "loss": 0.0249, "step": 39330 }, { "grad_norm": 0.4325816333293915, "learning_rate": 5.548876990541378e-05, "loss": 0.0291, "step": 39340 }, { "grad_norm": 0.31073877215385437, "learning_rate": 5.5468225975331564e-05, "loss": 0.0333, "step": 39350 }, { "grad_norm": 0.4945613443851471, "learning_rate": 5.544768111087909e-05, "loss": 0.0274, "step": 39360 }, { "grad_norm": 0.36983299255371094, "learning_rate": 5.54271353155669e-05, "loss": 0.0298, "step": 39370 }, { "grad_norm": 0.20270970463752747, "learning_rate": 5.5406588592905715e-05, "loss": 0.027, "step": 39380 }, { "grad_norm": 0.22939597070217133, "learning_rate": 5.5386040946406416e-05, "loss": 0.0298, "step": 39390 }, { "grad_norm": 0.6300071477890015, "learning_rate": 5.536549237958004e-05, "loss": 0.0388, "step": 39400 }, { "grad_norm": 0.32895854115486145, "learning_rate": 5.5344942895937744e-05, "loss": 0.0342, "step": 39410 }, { "grad_norm": 0.3846641182899475, "learning_rate": 5.5324392498990904e-05, "loss": 0.0275, "step": 39420 }, { "grad_norm": 0.4080331027507782, "learning_rate": 5.5303841192251e-05, "loss": 0.024, "step": 39430 }, { "grad_norm": 0.28327807784080505, "learning_rate": 5.52832889792297e-05, "loss": 0.0296, "step": 39440 }, { "grad_norm": 0.23127210140228271, "learning_rate": 5.526273586343881e-05, "loss": 0.0221, "step": 39450 }, { "grad_norm": 0.23945631086826324, "learning_rate": 5.5242181848390306e-05, "loss": 0.0214, "step": 39460 }, { "grad_norm": 0.359886109828949, "learning_rate": 5.5221626937596285e-05, "loss": 0.0241, "step": 39470 }, { "grad_norm": 0.22337181866168976, "learning_rate": 5.520107113456903e-05, "loss": 0.0335, "step": 39480 }, { "grad_norm": 0.4321826994419098, "learning_rate": 5.5180514442820974e-05, "loss": 0.033, "step": 39490 }, { "grad_norm": 0.29830336570739746, "learning_rate": 5.515995686586469e-05, "loss": 0.0215, "step": 39500 }, { "grad_norm": 0.3358123302459717, "learning_rate": 5.5139398407212916e-05, "loss": 0.0225, "step": 39510 }, { "grad_norm": 0.3648362159729004, "learning_rate": 5.511883907037849e-05, "loss": 0.0277, "step": 39520 }, { "grad_norm": 0.36691829562187195, "learning_rate": 5.509827885887449e-05, "loss": 0.0329, "step": 39530 }, { "grad_norm": 0.33842867612838745, "learning_rate": 5.507771777621406e-05, "loss": 0.0295, "step": 39540 }, { "grad_norm": 0.3960074484348297, "learning_rate": 5.505715582591052e-05, "loss": 0.0473, "step": 39550 }, { "grad_norm": 0.4361366033554077, "learning_rate": 5.50365930114774e-05, "loss": 0.033, "step": 39560 }, { "grad_norm": 0.29366493225097656, "learning_rate": 5.5016029336428255e-05, "loss": 0.0256, "step": 39570 }, { "grad_norm": 0.5122881531715393, "learning_rate": 5.49954648042769e-05, "loss": 0.031, "step": 39580 }, { "grad_norm": 0.23333430290222168, "learning_rate": 5.4974899418537226e-05, "loss": 0.0316, "step": 39590 }, { "grad_norm": 0.39745908975601196, "learning_rate": 5.4954333182723316e-05, "loss": 0.0241, "step": 39600 }, { "grad_norm": 0.33407703042030334, "learning_rate": 5.493376610034937e-05, "loss": 0.0295, "step": 39610 }, { "grad_norm": 0.3638027310371399, "learning_rate": 5.4913198174929735e-05, "loss": 0.0252, "step": 39620 }, { "grad_norm": 0.47239333391189575, "learning_rate": 5.489262940997889e-05, "loss": 0.0321, "step": 39630 }, { "grad_norm": 0.29308652877807617, "learning_rate": 5.487205980901151e-05, "loss": 0.0287, "step": 39640 }, { "grad_norm": 0.29597848653793335, "learning_rate": 5.485148937554234e-05, "loss": 0.0472, "step": 39650 }, { "grad_norm": 0.2772977650165558, "learning_rate": 5.483091811308635e-05, "loss": 0.0249, "step": 39660 }, { "grad_norm": 0.4365379512310028, "learning_rate": 5.4810346025158564e-05, "loss": 0.0317, "step": 39670 }, { "grad_norm": 0.8052005767822266, "learning_rate": 5.478977311527421e-05, "loss": 0.0392, "step": 39680 }, { "grad_norm": 0.27263343334198, "learning_rate": 5.476919938694863e-05, "loss": 0.0265, "step": 39690 }, { "grad_norm": 0.3167801797389984, "learning_rate": 5.474862484369733e-05, "loss": 0.02, "step": 39700 }, { "grad_norm": 0.2834832966327667, "learning_rate": 5.472804948903589e-05, "loss": 0.0295, "step": 39710 }, { "grad_norm": 0.37266775965690613, "learning_rate": 5.470747332648013e-05, "loss": 0.0197, "step": 39720 }, { "grad_norm": 0.4123323857784271, "learning_rate": 5.468689635954591e-05, "loss": 0.0394, "step": 39730 }, { "grad_norm": 0.2768721282482147, "learning_rate": 5.46663185917493e-05, "loss": 0.0266, "step": 39740 }, { "grad_norm": 0.3620891571044922, "learning_rate": 5.464574002660645e-05, "loss": 0.0282, "step": 39750 }, { "grad_norm": 0.2445160150527954, "learning_rate": 5.4625160667633724e-05, "loss": 0.0255, "step": 39760 }, { "grad_norm": 0.2970869243144989, "learning_rate": 5.4604580518347505e-05, "loss": 0.0302, "step": 39770 }, { "grad_norm": 0.2501341700553894, "learning_rate": 5.458399958226442e-05, "loss": 0.0263, "step": 39780 }, { "grad_norm": 0.5890122652053833, "learning_rate": 5.456341786290119e-05, "loss": 0.03, "step": 39790 }, { "grad_norm": 0.5559951066970825, "learning_rate": 5.454283536377465e-05, "loss": 0.0331, "step": 39800 }, { "grad_norm": 0.3619743287563324, "learning_rate": 5.452225208840179e-05, "loss": 0.0303, "step": 39810 }, { "grad_norm": 0.42917174100875854, "learning_rate": 5.450166804029975e-05, "loss": 0.0364, "step": 39820 }, { "grad_norm": 0.4244197905063629, "learning_rate": 5.448108322298574e-05, "loss": 0.0239, "step": 39830 }, { "grad_norm": 0.24637746810913086, "learning_rate": 5.446049763997717e-05, "loss": 0.0257, "step": 39840 }, { "grad_norm": 0.3655984401702881, "learning_rate": 5.4439911294791546e-05, "loss": 0.0236, "step": 39850 }, { "grad_norm": 0.4643425941467285, "learning_rate": 5.441932419094652e-05, "loss": 0.0237, "step": 39860 }, { "grad_norm": 0.3120129406452179, "learning_rate": 5.439873633195985e-05, "loss": 0.0268, "step": 39870 }, { "grad_norm": 0.29872575402259827, "learning_rate": 5.437814772134947e-05, "loss": 0.0329, "step": 39880 }, { "grad_norm": 0.34537971019744873, "learning_rate": 5.4357558362633366e-05, "loss": 0.0306, "step": 39890 }, { "grad_norm": 0.37808936834335327, "learning_rate": 5.4336968259329726e-05, "loss": 0.0219, "step": 39900 }, { "grad_norm": 0.49795034527778625, "learning_rate": 5.431637741495681e-05, "loss": 0.0285, "step": 39910 }, { "grad_norm": 0.32264015078544617, "learning_rate": 5.429578583303307e-05, "loss": 0.0288, "step": 39920 }, { "grad_norm": 0.2808287441730499, "learning_rate": 5.427519351707701e-05, "loss": 0.03, "step": 39930 }, { "grad_norm": 0.2825321853160858, "learning_rate": 5.42546004706073e-05, "loss": 0.029, "step": 39940 }, { "grad_norm": 0.30560603737831116, "learning_rate": 5.4234006697142735e-05, "loss": 0.0288, "step": 39950 }, { "grad_norm": 0.4046770930290222, "learning_rate": 5.421341220020224e-05, "loss": 0.0358, "step": 39960 }, { "grad_norm": 0.2690848410129547, "learning_rate": 5.419281698330482e-05, "loss": 0.0375, "step": 39970 }, { "grad_norm": 0.4142431318759918, "learning_rate": 5.4172221049969665e-05, "loss": 0.029, "step": 39980 }, { "grad_norm": 0.25696203112602234, "learning_rate": 5.415162440371604e-05, "loss": 0.0225, "step": 39990 }, { "grad_norm": 0.2827845513820648, "learning_rate": 5.413102704806334e-05, "loss": 0.0291, "step": 40000 }, { "grad_norm": 1.0127992630004883, "learning_rate": 5.41104289865311e-05, "loss": 0.0322, "step": 40010 }, { "grad_norm": 0.33333224058151245, "learning_rate": 5.408983022263898e-05, "loss": 0.0251, "step": 40020 }, { "grad_norm": 1.9541895389556885, "learning_rate": 5.406923075990673e-05, "loss": 0.0347, "step": 40030 }, { "grad_norm": 0.44437816739082336, "learning_rate": 5.404863060185423e-05, "loss": 0.0332, "step": 40040 }, { "grad_norm": 0.332926481962204, "learning_rate": 5.40280297520015e-05, "loss": 0.0344, "step": 40050 }, { "grad_norm": 0.31015273928642273, "learning_rate": 5.400742821386865e-05, "loss": 0.03, "step": 40060 }, { "grad_norm": 0.3737308084964752, "learning_rate": 5.398682599097592e-05, "loss": 0.03, "step": 40070 }, { "grad_norm": 0.2679122984409332, "learning_rate": 5.396622308684367e-05, "loss": 0.0236, "step": 40080 }, { "grad_norm": 0.4861246943473816, "learning_rate": 5.394561950499236e-05, "loss": 0.0362, "step": 40090 }, { "grad_norm": 0.2378491908311844, "learning_rate": 5.39250152489426e-05, "loss": 0.0291, "step": 40100 }, { "grad_norm": 0.45437419414520264, "learning_rate": 5.390441032221507e-05, "loss": 0.0296, "step": 40110 }, { "grad_norm": 0.315127968788147, "learning_rate": 5.388380472833062e-05, "loss": 0.0279, "step": 40120 }, { "grad_norm": 0.3858359158039093, "learning_rate": 5.386319847081016e-05, "loss": 0.0295, "step": 40130 }, { "grad_norm": 0.4185282588005066, "learning_rate": 5.384259155317473e-05, "loss": 0.029, "step": 40140 }, { "grad_norm": 0.466519296169281, "learning_rate": 5.38219839789455e-05, "loss": 0.0281, "step": 40150 }, { "grad_norm": 0.2372788041830063, "learning_rate": 5.380137575164376e-05, "loss": 0.0349, "step": 40160 }, { "grad_norm": 0.5104637742042542, "learning_rate": 5.378076687479085e-05, "loss": 0.027, "step": 40170 }, { "grad_norm": 0.3137945532798767, "learning_rate": 5.3760157351908305e-05, "loss": 0.0283, "step": 40180 }, { "grad_norm": 0.3316071331501007, "learning_rate": 5.373954718651768e-05, "loss": 0.0276, "step": 40190 }, { "grad_norm": 0.3262283504009247, "learning_rate": 5.371893638214074e-05, "loss": 0.0277, "step": 40200 }, { "grad_norm": 0.3784376084804535, "learning_rate": 5.369832494229927e-05, "loss": 0.0258, "step": 40210 }, { "grad_norm": 0.3973388373851776, "learning_rate": 5.367771287051524e-05, "loss": 0.0283, "step": 40220 }, { "grad_norm": 0.3066853880882263, "learning_rate": 5.365710017031065e-05, "loss": 0.0295, "step": 40230 }, { "grad_norm": 0.3375396132469177, "learning_rate": 5.3636486845207654e-05, "loss": 0.0266, "step": 40240 }, { "grad_norm": 0.3631013333797455, "learning_rate": 5.3615872898728526e-05, "loss": 0.0249, "step": 40250 }, { "grad_norm": 0.6723897457122803, "learning_rate": 5.3595258334395614e-05, "loss": 0.0286, "step": 40260 }, { "grad_norm": 0.38041746616363525, "learning_rate": 5.3574643155731396e-05, "loss": 0.0326, "step": 40270 }, { "grad_norm": 0.3264412581920624, "learning_rate": 5.3554027366258405e-05, "loss": 0.0262, "step": 40280 }, { "grad_norm": 0.35943976044654846, "learning_rate": 5.3533410969499355e-05, "loss": 0.0269, "step": 40290 }, { "grad_norm": 0.3832583427429199, "learning_rate": 5.3512793968977e-05, "loss": 0.0318, "step": 40300 }, { "grad_norm": 0.4392338991165161, "learning_rate": 5.349217636821423e-05, "loss": 0.0279, "step": 40310 }, { "grad_norm": 0.2238094061613083, "learning_rate": 5.347155817073405e-05, "loss": 0.0261, "step": 40320 }, { "grad_norm": 0.4567791521549225, "learning_rate": 5.34509393800595e-05, "loss": 0.0304, "step": 40330 }, { "grad_norm": 0.3106398284435272, "learning_rate": 5.343031999971381e-05, "loss": 0.0242, "step": 40340 }, { "grad_norm": 0.30377352237701416, "learning_rate": 5.340970003322026e-05, "loss": 0.0294, "step": 40350 }, { "grad_norm": 0.8523573875427246, "learning_rate": 5.33890794841022e-05, "loss": 0.0269, "step": 40360 }, { "grad_norm": 0.3154697120189667, "learning_rate": 5.336845835588318e-05, "loss": 0.0268, "step": 40370 }, { "grad_norm": 0.3307287395000458, "learning_rate": 5.334783665208674e-05, "loss": 0.0248, "step": 40380 }, { "grad_norm": 0.3160960078239441, "learning_rate": 5.332721437623657e-05, "loss": 0.0263, "step": 40390 }, { "grad_norm": 1.0455392599105835, "learning_rate": 5.3306591531856464e-05, "loss": 0.0352, "step": 40400 }, { "grad_norm": 0.24933134019374847, "learning_rate": 5.3285968122470295e-05, "loss": 0.0271, "step": 40410 }, { "grad_norm": 0.2539032995700836, "learning_rate": 5.326534415160207e-05, "loss": 0.0341, "step": 40420 }, { "grad_norm": 0.3325704038143158, "learning_rate": 5.324471962277582e-05, "loss": 0.0409, "step": 40430 }, { "grad_norm": 0.39972633123397827, "learning_rate": 5.3224094539515746e-05, "loss": 0.0284, "step": 40440 }, { "grad_norm": 0.3684305250644684, "learning_rate": 5.3203468905346076e-05, "loss": 0.0251, "step": 40450 }, { "grad_norm": 0.3375016748905182, "learning_rate": 5.3182842723791195e-05, "loss": 0.0302, "step": 40460 }, { "grad_norm": 0.36549609899520874, "learning_rate": 5.316221599837554e-05, "loss": 0.024, "step": 40470 }, { "grad_norm": 0.4291869103908539, "learning_rate": 5.314158873262366e-05, "loss": 0.0337, "step": 40480 }, { "grad_norm": 0.3314798176288605, "learning_rate": 5.312096093006018e-05, "loss": 0.027, "step": 40490 }, { "grad_norm": 0.2752379775047302, "learning_rate": 5.3100332594209846e-05, "loss": 0.0272, "step": 40500 }, { "grad_norm": 0.3090914189815521, "learning_rate": 5.3079703728597454e-05, "loss": 0.0276, "step": 40510 }, { "grad_norm": 0.2940375506877899, "learning_rate": 5.305907433674794e-05, "loss": 0.0233, "step": 40520 }, { "grad_norm": 0.33525463938713074, "learning_rate": 5.3038444422186284e-05, "loss": 0.0326, "step": 40530 }, { "grad_norm": 0.3256370723247528, "learning_rate": 5.301781398843757e-05, "loss": 0.024, "step": 40540 }, { "grad_norm": 0.22040869295597076, "learning_rate": 5.299718303902699e-05, "loss": 0.025, "step": 40550 }, { "grad_norm": 0.2636911869049072, "learning_rate": 5.297655157747979e-05, "loss": 0.0243, "step": 40560 }, { "grad_norm": 0.37965187430381775, "learning_rate": 5.295591960732136e-05, "loss": 0.0274, "step": 40570 }, { "grad_norm": 0.3021327257156372, "learning_rate": 5.293528713207708e-05, "loss": 0.0277, "step": 40580 }, { "grad_norm": 0.23065391182899475, "learning_rate": 5.291465415527253e-05, "loss": 0.0266, "step": 40590 }, { "grad_norm": 0.28522688150405884, "learning_rate": 5.2894020680433296e-05, "loss": 0.0239, "step": 40600 }, { "grad_norm": 0.33207815885543823, "learning_rate": 5.287338671108507e-05, "loss": 0.0236, "step": 40610 }, { "grad_norm": 0.4061671793460846, "learning_rate": 5.285275225075367e-05, "loss": 0.0265, "step": 40620 }, { "grad_norm": 0.2787873148918152, "learning_rate": 5.283211730296492e-05, "loss": 0.0245, "step": 40630 }, { "grad_norm": 0.23967798054218292, "learning_rate": 5.281148187124477e-05, "loss": 0.0263, "step": 40640 }, { "grad_norm": 0.3665410876274109, "learning_rate": 5.279084595911927e-05, "loss": 0.0342, "step": 40650 }, { "grad_norm": 0.3865894377231598, "learning_rate": 5.277020957011453e-05, "loss": 0.0367, "step": 40660 }, { "grad_norm": 0.4535720944404602, "learning_rate": 5.274957270775673e-05, "loss": 0.0324, "step": 40670 }, { "grad_norm": 0.30650681257247925, "learning_rate": 5.2728935375572164e-05, "loss": 0.0229, "step": 40680 }, { "grad_norm": 0.20419135689735413, "learning_rate": 5.2708297577087176e-05, "loss": 0.0225, "step": 40690 }, { "grad_norm": 0.3754660487174988, "learning_rate": 5.2687659315828196e-05, "loss": 0.0248, "step": 40700 }, { "grad_norm": 0.34232479333877563, "learning_rate": 5.266702059532175e-05, "loss": 0.0234, "step": 40710 }, { "grad_norm": 0.33576497435569763, "learning_rate": 5.264638141909444e-05, "loss": 0.0291, "step": 40720 }, { "grad_norm": 0.3517645299434662, "learning_rate": 5.2625741790672925e-05, "loss": 0.0252, "step": 40730 }, { "grad_norm": 0.2678612172603607, "learning_rate": 5.2605101713583935e-05, "loss": 0.0252, "step": 40740 }, { "grad_norm": 0.3097687363624573, "learning_rate": 5.2584461191354315e-05, "loss": 0.0335, "step": 40750 }, { "grad_norm": 0.3632507622241974, "learning_rate": 5.2563820227510964e-05, "loss": 0.0256, "step": 40760 }, { "grad_norm": 0.35755154490470886, "learning_rate": 5.2543178825580864e-05, "loss": 0.0236, "step": 40770 }, { "grad_norm": 0.4062873423099518, "learning_rate": 5.2522536989091054e-05, "loss": 0.0309, "step": 40780 }, { "grad_norm": 0.4232248067855835, "learning_rate": 5.2501894721568655e-05, "loss": 0.0301, "step": 40790 }, { "grad_norm": 0.34596481919288635, "learning_rate": 5.248125202654089e-05, "loss": 0.0288, "step": 40800 }, { "grad_norm": 0.36973002552986145, "learning_rate": 5.246060890753501e-05, "loss": 0.0318, "step": 40810 }, { "grad_norm": 0.3701677620410919, "learning_rate": 5.243996536807837e-05, "loss": 0.035, "step": 40820 }, { "grad_norm": 0.3072204887866974, "learning_rate": 5.2419321411698384e-05, "loss": 0.0238, "step": 40830 }, { "grad_norm": 0.45748627185821533, "learning_rate": 5.239867704192253e-05, "loss": 0.0275, "step": 40840 }, { "grad_norm": 0.193656787276268, "learning_rate": 5.237803226227838e-05, "loss": 0.0235, "step": 40850 }, { "grad_norm": 0.2330111712217331, "learning_rate": 5.235738707629354e-05, "loss": 0.0246, "step": 40860 }, { "grad_norm": 0.3812797963619232, "learning_rate": 5.233674148749575e-05, "loss": 0.0289, "step": 40870 }, { "grad_norm": 0.534605085849762, "learning_rate": 5.231609549941272e-05, "loss": 0.0317, "step": 40880 }, { "grad_norm": 0.32443973422050476, "learning_rate": 5.229544911557233e-05, "loss": 0.0316, "step": 40890 }, { "grad_norm": 0.40843451023101807, "learning_rate": 5.227480233950246e-05, "loss": 0.0332, "step": 40900 }, { "grad_norm": 0.3158012628555298, "learning_rate": 5.22541551747311e-05, "loss": 0.0328, "step": 40910 }, { "grad_norm": 0.5076861381530762, "learning_rate": 5.223350762478626e-05, "loss": 0.0273, "step": 40920 }, { "grad_norm": 0.2786528766155243, "learning_rate": 5.221285969319606e-05, "loss": 0.0257, "step": 40930 }, { "grad_norm": 0.27948129177093506, "learning_rate": 5.219221138348865e-05, "loss": 0.0312, "step": 40940 }, { "grad_norm": 0.2903517484664917, "learning_rate": 5.217156269919228e-05, "loss": 0.0268, "step": 40950 }, { "grad_norm": 0.2987803816795349, "learning_rate": 5.215091364383523e-05, "loss": 0.0258, "step": 40960 }, { "grad_norm": 0.40046700835227966, "learning_rate": 5.213026422094588e-05, "loss": 0.033, "step": 40970 }, { "grad_norm": 0.2614728808403015, "learning_rate": 5.210961443405262e-05, "loss": 0.0249, "step": 40980 }, { "grad_norm": 0.2986013889312744, "learning_rate": 5.208896428668396e-05, "loss": 0.0227, "step": 40990 }, { "grad_norm": 0.3713226318359375, "learning_rate": 5.206831378236845e-05, "loss": 0.0306, "step": 41000 }, { "grad_norm": 0.42943552136421204, "learning_rate": 5.2047662924634666e-05, "loss": 0.0244, "step": 41010 }, { "grad_norm": 0.36874356865882874, "learning_rate": 5.20270117170113e-05, "loss": 0.0207, "step": 41020 }, { "grad_norm": 0.4181622266769409, "learning_rate": 5.200636016302707e-05, "loss": 0.0359, "step": 41030 }, { "grad_norm": 0.40180695056915283, "learning_rate": 5.1985708266210754e-05, "loss": 0.0349, "step": 41040 }, { "grad_norm": 0.2757284939289093, "learning_rate": 5.1965056030091206e-05, "loss": 0.0303, "step": 41050 }, { "grad_norm": 0.2444285899400711, "learning_rate": 5.194440345819732e-05, "loss": 0.0233, "step": 41060 }, { "grad_norm": 0.3529561460018158, "learning_rate": 5.1923750554058084e-05, "loss": 0.0297, "step": 41070 }, { "grad_norm": 0.2597268223762512, "learning_rate": 5.1903097321202476e-05, "loss": 0.0231, "step": 41080 }, { "grad_norm": 0.375946968793869, "learning_rate": 5.18824437631596e-05, "loss": 0.0367, "step": 41090 }, { "grad_norm": 0.27136117219924927, "learning_rate": 5.186178988345856e-05, "loss": 0.0266, "step": 41100 }, { "grad_norm": 0.346116840839386, "learning_rate": 5.184113568562855e-05, "loss": 0.0269, "step": 41110 }, { "grad_norm": 0.32991859316825867, "learning_rate": 5.18204811731988e-05, "loss": 0.0338, "step": 41120 }, { "grad_norm": 0.4353448152542114, "learning_rate": 5.179982634969862e-05, "loss": 0.0281, "step": 41130 }, { "grad_norm": 0.43137574195861816, "learning_rate": 5.177917121865734e-05, "loss": 0.0344, "step": 41140 }, { "grad_norm": 0.2722684144973755, "learning_rate": 5.1758515783604346e-05, "loss": 0.026, "step": 41150 }, { "grad_norm": 0.29818645119667053, "learning_rate": 5.173786004806911e-05, "loss": 0.0366, "step": 41160 }, { "grad_norm": 0.49441078305244446, "learning_rate": 5.1717204015581135e-05, "loss": 0.0283, "step": 41170 }, { "grad_norm": 0.294035404920578, "learning_rate": 5.169654768966994e-05, "loss": 0.0261, "step": 41180 }, { "grad_norm": 0.48002225160598755, "learning_rate": 5.1675891073865156e-05, "loss": 0.0337, "step": 41190 }, { "grad_norm": 0.28856638073921204, "learning_rate": 5.1655234171696424e-05, "loss": 0.0217, "step": 41200 }, { "grad_norm": 0.26544418931007385, "learning_rate": 5.163457698669343e-05, "loss": 0.0198, "step": 41210 }, { "grad_norm": 0.359097421169281, "learning_rate": 5.1613919522385936e-05, "loss": 0.0217, "step": 41220 }, { "grad_norm": 0.34465640783309937, "learning_rate": 5.1593261782303746e-05, "loss": 0.0291, "step": 41230 }, { "grad_norm": 0.31933993101119995, "learning_rate": 5.157260376997669e-05, "loss": 0.0252, "step": 41240 }, { "grad_norm": 0.3091995120048523, "learning_rate": 5.155194548893466e-05, "loss": 0.0292, "step": 41250 }, { "grad_norm": 0.48863136768341064, "learning_rate": 5.153128694270759e-05, "loss": 0.0393, "step": 41260 }, { "grad_norm": 0.4901552200317383, "learning_rate": 5.151062813482548e-05, "loss": 0.0375, "step": 41270 }, { "grad_norm": 0.3856066167354584, "learning_rate": 5.148996906881832e-05, "loss": 0.026, "step": 41280 }, { "grad_norm": 0.4740217328071594, "learning_rate": 5.1469309748216196e-05, "loss": 0.0304, "step": 41290 }, { "grad_norm": 0.29847463965415955, "learning_rate": 5.144865017654923e-05, "loss": 0.0216, "step": 41300 }, { "grad_norm": 0.5837292671203613, "learning_rate": 5.1427990357347556e-05, "loss": 0.0316, "step": 41310 }, { "grad_norm": 0.34022945165634155, "learning_rate": 5.140733029414139e-05, "loss": 0.0331, "step": 41320 }, { "grad_norm": 0.379336953163147, "learning_rate": 5.138666999046098e-05, "loss": 0.0313, "step": 41330 }, { "grad_norm": 0.43550071120262146, "learning_rate": 5.136600944983658e-05, "loss": 0.0479, "step": 41340 }, { "grad_norm": 0.3201881945133209, "learning_rate": 5.134534867579853e-05, "loss": 0.0299, "step": 41350 }, { "grad_norm": 0.4284713566303253, "learning_rate": 5.1324687671877184e-05, "loss": 0.0323, "step": 41360 }, { "grad_norm": 0.30911481380462646, "learning_rate": 5.130402644160296e-05, "loss": 0.0298, "step": 41370 }, { "grad_norm": 0.2523999810218811, "learning_rate": 5.128336498850628e-05, "loss": 0.0225, "step": 41380 }, { "grad_norm": 0.3072887063026428, "learning_rate": 5.126270331611761e-05, "loss": 0.0231, "step": 41390 }, { "grad_norm": 0.41369307041168213, "learning_rate": 5.124204142796748e-05, "loss": 0.0229, "step": 41400 }, { "grad_norm": 0.2670312523841858, "learning_rate": 5.122137932758644e-05, "loss": 0.0241, "step": 41410 }, { "grad_norm": 0.2817099094390869, "learning_rate": 5.1200717018505074e-05, "loss": 0.0233, "step": 41420 }, { "grad_norm": 0.31108358502388, "learning_rate": 5.118005450425403e-05, "loss": 0.0275, "step": 41430 }, { "grad_norm": 0.3407426178455353, "learning_rate": 5.115939178836392e-05, "loss": 0.0209, "step": 41440 }, { "grad_norm": 0.4826003313064575, "learning_rate": 5.113872887436547e-05, "loss": 0.0249, "step": 41450 }, { "grad_norm": 0.3513208031654358, "learning_rate": 5.11180657657894e-05, "loss": 0.0313, "step": 41460 }, { "grad_norm": 0.4889087975025177, "learning_rate": 5.109740246616649e-05, "loss": 0.0285, "step": 41470 }, { "grad_norm": 0.4577505886554718, "learning_rate": 5.10767389790275e-05, "loss": 0.0359, "step": 41480 }, { "grad_norm": 0.33628538250923157, "learning_rate": 5.1056075307903265e-05, "loss": 0.0286, "step": 41490 }, { "grad_norm": 0.301107257604599, "learning_rate": 5.1035411456324634e-05, "loss": 0.0249, "step": 41500 }, { "grad_norm": 0.5015422701835632, "learning_rate": 5.1014747427822504e-05, "loss": 0.034, "step": 41510 }, { "grad_norm": 0.40492191910743713, "learning_rate": 5.0994083225927804e-05, "loss": 0.0366, "step": 41520 }, { "grad_norm": 0.27268630266189575, "learning_rate": 5.0973418854171475e-05, "loss": 0.0285, "step": 41530 }, { "grad_norm": 0.29741600155830383, "learning_rate": 5.095275431608447e-05, "loss": 0.0246, "step": 41540 }, { "grad_norm": 0.2562010884284973, "learning_rate": 5.093208961519782e-05, "loss": 0.0247, "step": 41550 }, { "grad_norm": 0.27142134308815, "learning_rate": 5.091142475504255e-05, "loss": 0.0261, "step": 41560 }, { "grad_norm": 0.4787325859069824, "learning_rate": 5.089075973914971e-05, "loss": 0.0293, "step": 41570 }, { "grad_norm": 0.4929867386817932, "learning_rate": 5.08700945710504e-05, "loss": 0.0343, "step": 41580 }, { "grad_norm": 0.29348623752593994, "learning_rate": 5.0849429254275714e-05, "loss": 0.0252, "step": 41590 }, { "grad_norm": 0.25736165046691895, "learning_rate": 5.0828763792356804e-05, "loss": 0.0228, "step": 41600 }, { "grad_norm": 0.4451765716075897, "learning_rate": 5.080809818882483e-05, "loss": 0.0305, "step": 41610 }, { "grad_norm": 0.31214311718940735, "learning_rate": 5.078743244721097e-05, "loss": 0.0251, "step": 41620 }, { "grad_norm": 0.6781065464019775, "learning_rate": 5.0766766571046455e-05, "loss": 0.0274, "step": 41630 }, { "grad_norm": 0.3229592442512512, "learning_rate": 5.07461005638625e-05, "loss": 0.0272, "step": 41640 }, { "grad_norm": 0.3351781964302063, "learning_rate": 5.072543442919037e-05, "loss": 0.0276, "step": 41650 }, { "grad_norm": 0.29431819915771484, "learning_rate": 5.070476817056132e-05, "loss": 0.0291, "step": 41660 }, { "grad_norm": 0.40439966320991516, "learning_rate": 5.068410179150668e-05, "loss": 0.0237, "step": 41670 }, { "grad_norm": 0.32177117466926575, "learning_rate": 5.066343529555775e-05, "loss": 0.0258, "step": 41680 }, { "grad_norm": 0.22977323830127716, "learning_rate": 5.064276868624587e-05, "loss": 0.0243, "step": 41690 }, { "grad_norm": 0.3219558596611023, "learning_rate": 5.0622101967102396e-05, "loss": 0.0214, "step": 41700 }, { "grad_norm": 0.34353649616241455, "learning_rate": 5.0601435141658714e-05, "loss": 0.0265, "step": 41710 }, { "grad_norm": 0.503265917301178, "learning_rate": 5.058076821344622e-05, "loss": 0.0262, "step": 41720 }, { "grad_norm": 0.3624367117881775, "learning_rate": 5.056010118599632e-05, "loss": 0.0348, "step": 41730 }, { "grad_norm": 0.4415563941001892, "learning_rate": 5.053943406284044e-05, "loss": 0.0331, "step": 41740 }, { "grad_norm": 0.3989621102809906, "learning_rate": 5.051876684751006e-05, "loss": 0.03, "step": 41750 }, { "grad_norm": 0.42630818486213684, "learning_rate": 5.0498099543536584e-05, "loss": 0.0269, "step": 41760 }, { "grad_norm": 0.3490816354751587, "learning_rate": 5.047743215445152e-05, "loss": 0.029, "step": 41770 }, { "grad_norm": 0.3060547709465027, "learning_rate": 5.045676468378637e-05, "loss": 0.0281, "step": 41780 }, { "grad_norm": 0.3636080324649811, "learning_rate": 5.0436097135072626e-05, "loss": 0.0323, "step": 41790 }, { "grad_norm": 0.3295642137527466, "learning_rate": 5.041542951184181e-05, "loss": 0.0262, "step": 41800 }, { "grad_norm": 0.2080216258764267, "learning_rate": 5.039476181762545e-05, "loss": 0.0279, "step": 41810 }, { "grad_norm": 0.31913378834724426, "learning_rate": 5.037409405595508e-05, "loss": 0.0212, "step": 41820 }, { "grad_norm": 0.45828977227211, "learning_rate": 5.035342623036229e-05, "loss": 0.0232, "step": 41830 }, { "grad_norm": 0.36096739768981934, "learning_rate": 5.033275834437862e-05, "loss": 0.0271, "step": 41840 }, { "grad_norm": 0.35582026839256287, "learning_rate": 5.031209040153564e-05, "loss": 0.0229, "step": 41850 }, { "grad_norm": 0.29049617052078247, "learning_rate": 5.0291422405364955e-05, "loss": 0.0406, "step": 41860 }, { "grad_norm": 0.3525810241699219, "learning_rate": 5.0270754359398133e-05, "loss": 0.0246, "step": 41870 }, { "grad_norm": 0.24642720818519592, "learning_rate": 5.025008626716682e-05, "loss": 0.0242, "step": 41880 }, { "grad_norm": 0.3433397710323334, "learning_rate": 5.0229418132202585e-05, "loss": 0.0263, "step": 41890 }, { "grad_norm": 0.4130711257457733, "learning_rate": 5.020874995803707e-05, "loss": 0.0328, "step": 41900 }, { "grad_norm": 0.31278589367866516, "learning_rate": 5.01880817482019e-05, "loss": 0.0282, "step": 41910 }, { "grad_norm": 0.3022257089614868, "learning_rate": 5.01674135062287e-05, "loss": 0.0245, "step": 41920 }, { "grad_norm": 0.3782065808773041, "learning_rate": 5.0146745235649115e-05, "loss": 0.0263, "step": 41930 }, { "grad_norm": 0.3190433979034424, "learning_rate": 5.0126076939994795e-05, "loss": 0.0237, "step": 41940 }, { "grad_norm": 0.23771776258945465, "learning_rate": 5.010540862279736e-05, "loss": 0.0259, "step": 41950 }, { "grad_norm": 0.5498961806297302, "learning_rate": 5.008474028758846e-05, "loss": 0.0328, "step": 41960 }, { "grad_norm": 0.28601357340812683, "learning_rate": 5.0064071937899784e-05, "loss": 0.0259, "step": 41970 }, { "grad_norm": 0.5786641836166382, "learning_rate": 5.004340357726296e-05, "loss": 0.0248, "step": 41980 }, { "grad_norm": 0.42241519689559937, "learning_rate": 5.002273520920965e-05, "loss": 0.0254, "step": 41990 }, { "grad_norm": 0.7533131241798401, "learning_rate": 5.000206683727151e-05, "loss": 0.0279, "step": 42000 }, { "grad_norm": 0.4233328402042389, "learning_rate": 4.998139846498021e-05, "loss": 0.0283, "step": 42010 }, { "grad_norm": 0.24081459641456604, "learning_rate": 4.9960730095867405e-05, "loss": 0.0285, "step": 42020 }, { "grad_norm": 0.6471837759017944, "learning_rate": 4.9940061733464755e-05, "loss": 0.0316, "step": 42030 }, { "grad_norm": 0.26490676403045654, "learning_rate": 4.991939338130392e-05, "loss": 0.0312, "step": 42040 }, { "grad_norm": 0.4346080720424652, "learning_rate": 4.989872504291653e-05, "loss": 0.0359, "step": 42050 }, { "grad_norm": 0.35625889897346497, "learning_rate": 4.9878056721834273e-05, "loss": 0.0302, "step": 42060 }, { "grad_norm": 0.4865626096725464, "learning_rate": 4.98573884215888e-05, "loss": 0.0284, "step": 42070 }, { "grad_norm": 0.38270553946495056, "learning_rate": 4.9836720145711715e-05, "loss": 0.0317, "step": 42080 }, { "grad_norm": 0.21241658926010132, "learning_rate": 4.9816051897734725e-05, "loss": 0.033, "step": 42090 }, { "grad_norm": 0.46164578199386597, "learning_rate": 4.979538368118941e-05, "loss": 0.0335, "step": 42100 }, { "grad_norm": 0.3375518321990967, "learning_rate": 4.9774715499607446e-05, "loss": 0.0266, "step": 42110 }, { "grad_norm": 0.36266815662384033, "learning_rate": 4.9754047356520416e-05, "loss": 0.0291, "step": 42120 }, { "grad_norm": 0.38243579864501953, "learning_rate": 4.973337925545999e-05, "loss": 0.0295, "step": 42130 }, { "grad_norm": 0.2977079749107361, "learning_rate": 4.9712711199957746e-05, "loss": 0.0213, "step": 42140 }, { "grad_norm": 0.32238224148750305, "learning_rate": 4.969204319354531e-05, "loss": 0.0324, "step": 42150 }, { "grad_norm": 0.3475271165370941, "learning_rate": 4.9671375239754267e-05, "loss": 0.0266, "step": 42160 }, { "grad_norm": 0.2915974259376526, "learning_rate": 4.9650707342116195e-05, "loss": 0.0198, "step": 42170 }, { "grad_norm": 0.2937251031398773, "learning_rate": 4.96300395041627e-05, "loss": 0.0221, "step": 42180 }, { "grad_norm": 0.43754860758781433, "learning_rate": 4.960937172942532e-05, "loss": 0.0265, "step": 42190 }, { "grad_norm": 0.25191590189933777, "learning_rate": 4.958870402143565e-05, "loss": 0.0229, "step": 42200 }, { "grad_norm": 0.3410957157611847, "learning_rate": 4.9568036383725186e-05, "loss": 0.0251, "step": 42210 }, { "grad_norm": 0.39922162890434265, "learning_rate": 4.9547368819825516e-05, "loss": 0.0217, "step": 42220 }, { "grad_norm": 0.40291890501976013, "learning_rate": 4.952670133326812e-05, "loss": 0.0244, "step": 42230 }, { "grad_norm": 0.37709566950798035, "learning_rate": 4.950603392758453e-05, "loss": 0.0294, "step": 42240 }, { "grad_norm": 0.2601104974746704, "learning_rate": 4.948536660630621e-05, "loss": 0.0198, "step": 42250 }, { "grad_norm": 0.33743342757225037, "learning_rate": 4.9464699372964685e-05, "loss": 0.029, "step": 42260 }, { "grad_norm": 0.2606988549232483, "learning_rate": 4.9444032231091395e-05, "loss": 0.0252, "step": 42270 }, { "grad_norm": 0.29509884119033813, "learning_rate": 4.9423365184217776e-05, "loss": 0.0262, "step": 42280 }, { "grad_norm": 0.29065462946891785, "learning_rate": 4.940269823587529e-05, "loss": 0.0252, "step": 42290 }, { "grad_norm": 0.2434130758047104, "learning_rate": 4.938203138959533e-05, "loss": 0.0224, "step": 42300 }, { "grad_norm": 0.37843406200408936, "learning_rate": 4.9361364648909325e-05, "loss": 0.0265, "step": 42310 }, { "grad_norm": 0.30125153064727783, "learning_rate": 4.9340698017348605e-05, "loss": 0.0213, "step": 42320 }, { "grad_norm": 0.35106950998306274, "learning_rate": 4.932003149844458e-05, "loss": 0.0219, "step": 42330 }, { "grad_norm": 0.28259381651878357, "learning_rate": 4.929936509572857e-05, "loss": 0.0306, "step": 42340 }, { "grad_norm": 0.40443024039268494, "learning_rate": 4.927869881273191e-05, "loss": 0.0253, "step": 42350 }, { "grad_norm": 0.30985552072525024, "learning_rate": 4.9258032652985894e-05, "loss": 0.0291, "step": 42360 }, { "grad_norm": 0.30366799235343933, "learning_rate": 4.9237366620021786e-05, "loss": 0.0311, "step": 42370 }, { "grad_norm": 0.24115678668022156, "learning_rate": 4.921670071737089e-05, "loss": 0.0228, "step": 42380 }, { "grad_norm": 0.27855396270751953, "learning_rate": 4.91960349485644e-05, "loss": 0.0318, "step": 42390 }, { "grad_norm": 0.34523284435272217, "learning_rate": 4.9175369317133566e-05, "loss": 0.0284, "step": 42400 }, { "grad_norm": 0.2582989037036896, "learning_rate": 4.915470382660953e-05, "loss": 0.0231, "step": 42410 }, { "grad_norm": 0.261881560087204, "learning_rate": 4.9134038480523524e-05, "loss": 0.0232, "step": 42420 }, { "grad_norm": 0.33011898398399353, "learning_rate": 4.911337328240664e-05, "loss": 0.0286, "step": 42430 }, { "grad_norm": 0.25830769538879395, "learning_rate": 4.909270823579003e-05, "loss": 0.0277, "step": 42440 }, { "grad_norm": 0.3090258240699768, "learning_rate": 4.907204334420476e-05, "loss": 0.0241, "step": 42450 }, { "grad_norm": 0.24737413227558136, "learning_rate": 4.9051378611181884e-05, "loss": 0.0296, "step": 42460 }, { "grad_norm": 0.24816621840000153, "learning_rate": 4.9030714040252486e-05, "loss": 0.0226, "step": 42470 }, { "grad_norm": 0.19176605343818665, "learning_rate": 4.901004963494752e-05, "loss": 0.0204, "step": 42480 }, { "grad_norm": 0.3543115258216858, "learning_rate": 4.898938539879802e-05, "loss": 0.0289, "step": 42490 }, { "grad_norm": 0.593245267868042, "learning_rate": 4.8968721335334904e-05, "loss": 0.0285, "step": 42500 }, { "grad_norm": 0.39757150411605835, "learning_rate": 4.894805744808912e-05, "loss": 0.025, "step": 42510 }, { "grad_norm": 0.281960666179657, "learning_rate": 4.8927393740591524e-05, "loss": 0.0259, "step": 42520 }, { "grad_norm": 0.32289206981658936, "learning_rate": 4.890673021637302e-05, "loss": 0.0243, "step": 42530 }, { "grad_norm": 0.3158063590526581, "learning_rate": 4.888606687896442e-05, "loss": 0.0236, "step": 42540 }, { "grad_norm": 0.3209039568901062, "learning_rate": 4.886540373189652e-05, "loss": 0.0247, "step": 42550 }, { "grad_norm": 0.2481584995985031, "learning_rate": 4.88447407787001e-05, "loss": 0.0224, "step": 42560 }, { "grad_norm": 0.3109600245952606, "learning_rate": 4.8824078022905864e-05, "loss": 0.0199, "step": 42570 }, { "grad_norm": 0.36035165190696716, "learning_rate": 4.880341546804455e-05, "loss": 0.0223, "step": 42580 }, { "grad_norm": 0.42575207352638245, "learning_rate": 4.878275311764679e-05, "loss": 0.0293, "step": 42590 }, { "grad_norm": 0.32806116342544556, "learning_rate": 4.8762090975243254e-05, "loss": 0.022, "step": 42600 }, { "grad_norm": 0.229302316904068, "learning_rate": 4.874142904436448e-05, "loss": 0.0217, "step": 42610 }, { "grad_norm": 0.4496576488018036, "learning_rate": 4.8720767328541094e-05, "loss": 0.0333, "step": 42620 }, { "grad_norm": 0.5497820973396301, "learning_rate": 4.870010583130358e-05, "loss": 0.0215, "step": 42630 }, { "grad_norm": 0.40782472491264343, "learning_rate": 4.867944455618241e-05, "loss": 0.0269, "step": 42640 }, { "grad_norm": 0.3098592758178711, "learning_rate": 4.865878350670807e-05, "loss": 0.022, "step": 42650 }, { "grad_norm": 0.3350852429866791, "learning_rate": 4.8638122686410914e-05, "loss": 0.0241, "step": 42660 }, { "grad_norm": 0.4033747613430023, "learning_rate": 4.861746209882137e-05, "loss": 0.0219, "step": 42670 }, { "grad_norm": 0.28718122839927673, "learning_rate": 4.859680174746972e-05, "loss": 0.0247, "step": 42680 }, { "grad_norm": 0.2960719168186188, "learning_rate": 4.857614163588629e-05, "loss": 0.0346, "step": 42690 }, { "grad_norm": 0.3239504396915436, "learning_rate": 4.855548176760131e-05, "loss": 0.0227, "step": 42700 }, { "grad_norm": 0.27204015851020813, "learning_rate": 4.8534822146144986e-05, "loss": 0.0305, "step": 42710 }, { "grad_norm": 0.3205743134021759, "learning_rate": 4.851416277504749e-05, "loss": 0.0358, "step": 42720 }, { "grad_norm": 0.34779199957847595, "learning_rate": 4.8493503657838923e-05, "loss": 0.0243, "step": 42730 }, { "grad_norm": 0.41821587085723877, "learning_rate": 4.84728447980494e-05, "loss": 0.0276, "step": 42740 }, { "grad_norm": 0.3901883065700531, "learning_rate": 4.8452186199208914e-05, "loss": 0.0258, "step": 42750 }, { "grad_norm": 0.3369847238063812, "learning_rate": 4.843152786484749e-05, "loss": 0.0264, "step": 42760 }, { "grad_norm": 0.1922902911901474, "learning_rate": 4.8410869798495035e-05, "loss": 0.0234, "step": 42770 }, { "grad_norm": 0.29865381121635437, "learning_rate": 4.8390212003681486e-05, "loss": 0.0242, "step": 42780 }, { "grad_norm": 0.4561345875263214, "learning_rate": 4.836955448393667e-05, "loss": 0.0338, "step": 42790 }, { "grad_norm": 0.2741447389125824, "learning_rate": 4.8348897242790406e-05, "loss": 0.0242, "step": 42800 }, { "grad_norm": 0.22351804375648499, "learning_rate": 4.832824028377243e-05, "loss": 0.0233, "step": 42810 }, { "grad_norm": 0.4828795790672302, "learning_rate": 4.830758361041249e-05, "loss": 0.0312, "step": 42820 }, { "grad_norm": 0.3272923529148102, "learning_rate": 4.828692722624022e-05, "loss": 0.0263, "step": 42830 }, { "grad_norm": 0.29192012548446655, "learning_rate": 4.826627113478522e-05, "loss": 0.0444, "step": 42840 }, { "grad_norm": 0.3694661855697632, "learning_rate": 4.824561533957708e-05, "loss": 0.0229, "step": 42850 }, { "grad_norm": 0.43660879135131836, "learning_rate": 4.822495984414527e-05, "loss": 0.0261, "step": 42860 }, { "grad_norm": 0.3461917042732239, "learning_rate": 4.8204304652019304e-05, "loss": 0.024, "step": 42870 }, { "grad_norm": 1.0449165105819702, "learning_rate": 4.8183649766728537e-05, "loss": 0.0274, "step": 42880 }, { "grad_norm": 0.38088294863700867, "learning_rate": 4.816299519180237e-05, "loss": 0.0286, "step": 42890 }, { "grad_norm": 0.21732959151268005, "learning_rate": 4.8142340930770066e-05, "loss": 0.0258, "step": 42900 }, { "grad_norm": 0.5097055435180664, "learning_rate": 4.81216869871609e-05, "loss": 0.0276, "step": 42910 }, { "grad_norm": 0.3755106031894684, "learning_rate": 4.8101033364504064e-05, "loss": 0.0239, "step": 42920 }, { "grad_norm": 0.3172346353530884, "learning_rate": 4.808038006632866e-05, "loss": 0.0326, "step": 42930 }, { "grad_norm": 0.7169066071510315, "learning_rate": 4.805972709616383e-05, "loss": 0.0207, "step": 42940 }, { "grad_norm": 0.41013017296791077, "learning_rate": 4.8039074457538556e-05, "loss": 0.0218, "step": 42950 }, { "grad_norm": 0.43056604266166687, "learning_rate": 4.801842215398184e-05, "loss": 0.0271, "step": 42960 }, { "grad_norm": 0.32963570952415466, "learning_rate": 4.799777018902256e-05, "loss": 0.0265, "step": 42970 }, { "grad_norm": 0.35077419877052307, "learning_rate": 4.797711856618961e-05, "loss": 0.0184, "step": 42980 }, { "grad_norm": 0.39265120029449463, "learning_rate": 4.7956467289011765e-05, "loss": 0.0227, "step": 42990 }, { "grad_norm": 0.26855409145355225, "learning_rate": 4.793581636101778e-05, "loss": 0.0229, "step": 43000 }, { "grad_norm": 0.35771363973617554, "learning_rate": 4.7915165785736326e-05, "loss": 0.0218, "step": 43010 }, { "grad_norm": 0.3336363136768341, "learning_rate": 4.789451556669599e-05, "loss": 0.0232, "step": 43020 }, { "grad_norm": 0.2652214467525482, "learning_rate": 4.7873865707425384e-05, "loss": 0.0312, "step": 43030 }, { "grad_norm": 0.2736227214336395, "learning_rate": 4.7853216211452975e-05, "loss": 0.0246, "step": 43040 }, { "grad_norm": 0.5383521318435669, "learning_rate": 4.78325670823072e-05, "loss": 0.0297, "step": 43050 }, { "grad_norm": 0.33252936601638794, "learning_rate": 4.781191832351641e-05, "loss": 0.0286, "step": 43060 }, { "grad_norm": 0.42368772625923157, "learning_rate": 4.7791269938608965e-05, "loss": 0.027, "step": 43070 }, { "grad_norm": 0.30822286009788513, "learning_rate": 4.777062193111305e-05, "loss": 0.0244, "step": 43080 }, { "grad_norm": 0.2721115052700043, "learning_rate": 4.77499743045569e-05, "loss": 0.0241, "step": 43090 }, { "grad_norm": 0.4441697597503662, "learning_rate": 4.7729327062468596e-05, "loss": 0.0329, "step": 43100 }, { "grad_norm": 0.3547402620315552, "learning_rate": 4.770868020837619e-05, "loss": 0.0304, "step": 43110 }, { "grad_norm": 0.3083537220954895, "learning_rate": 4.768803374580768e-05, "loss": 0.0192, "step": 43120 }, { "grad_norm": 0.30646443367004395, "learning_rate": 4.766738767829094e-05, "loss": 0.0209, "step": 43130 }, { "grad_norm": 0.2816719114780426, "learning_rate": 4.764674200935388e-05, "loss": 0.0217, "step": 43140 }, { "grad_norm": 0.28070151805877686, "learning_rate": 4.762609674252424e-05, "loss": 0.0214, "step": 43150 }, { "grad_norm": 0.2440311163663864, "learning_rate": 4.760545188132974e-05, "loss": 0.0372, "step": 43160 }, { "grad_norm": 0.2189074456691742, "learning_rate": 4.758480742929801e-05, "loss": 0.0258, "step": 43170 }, { "grad_norm": 0.5992908477783203, "learning_rate": 4.756416338995664e-05, "loss": 0.0297, "step": 43180 }, { "grad_norm": 0.3972327709197998, "learning_rate": 4.7543519766833145e-05, "loss": 0.0207, "step": 43190 }, { "grad_norm": 0.22080057859420776, "learning_rate": 4.752287656345492e-05, "loss": 0.0261, "step": 43200 }, { "grad_norm": 0.4350558817386627, "learning_rate": 4.7502233783349355e-05, "loss": 0.0305, "step": 43210 }, { "grad_norm": 0.5897974371910095, "learning_rate": 4.7481591430043694e-05, "loss": 0.0237, "step": 43220 }, { "grad_norm": 0.2884136438369751, "learning_rate": 4.746094950706521e-05, "loss": 0.0245, "step": 43230 }, { "grad_norm": 0.31580618023872375, "learning_rate": 4.744030801794099e-05, "loss": 0.026, "step": 43240 }, { "grad_norm": 0.3565528392791748, "learning_rate": 4.741966696619813e-05, "loss": 0.0288, "step": 43250 }, { "grad_norm": 0.6340527534484863, "learning_rate": 4.739902635536361e-05, "loss": 0.0312, "step": 43260 }, { "grad_norm": 0.41492608189582825, "learning_rate": 4.737838618896436e-05, "loss": 0.0264, "step": 43270 }, { "grad_norm": 0.3585568070411682, "learning_rate": 4.7357746470527203e-05, "loss": 0.0259, "step": 43280 }, { "grad_norm": 0.7903847694396973, "learning_rate": 4.73371072035789e-05, "loss": 0.0326, "step": 43290 }, { "grad_norm": 0.4805487394332886, "learning_rate": 4.731646839164616e-05, "loss": 0.0285, "step": 43300 }, { "grad_norm": 0.29221832752227783, "learning_rate": 4.729583003825556e-05, "loss": 0.0305, "step": 43310 }, { "grad_norm": 0.3313990533351898, "learning_rate": 4.727519214693367e-05, "loss": 0.0272, "step": 43320 }, { "grad_norm": 0.30797290802001953, "learning_rate": 4.725455472120689e-05, "loss": 0.0242, "step": 43330 }, { "grad_norm": 0.22236332297325134, "learning_rate": 4.723391776460164e-05, "loss": 0.0278, "step": 43340 }, { "grad_norm": 0.6460248827934265, "learning_rate": 4.7213281280644186e-05, "loss": 0.0279, "step": 43350 }, { "grad_norm": 0.3490011394023895, "learning_rate": 4.719264527286075e-05, "loss": 0.0252, "step": 43360 }, { "grad_norm": 0.5523420572280884, "learning_rate": 4.717200974477744e-05, "loss": 0.0283, "step": 43370 }, { "grad_norm": 0.2588449716567993, "learning_rate": 4.715137469992034e-05, "loss": 0.0299, "step": 43380 }, { "grad_norm": 0.4507896602153778, "learning_rate": 4.713074014181539e-05, "loss": 0.0272, "step": 43390 }, { "grad_norm": 0.24004767835140228, "learning_rate": 4.711010607398846e-05, "loss": 0.0214, "step": 43400 }, { "grad_norm": 0.3185681700706482, "learning_rate": 4.708947249996538e-05, "loss": 0.0207, "step": 43410 }, { "grad_norm": 0.23465900123119354, "learning_rate": 4.706883942327183e-05, "loss": 0.0225, "step": 43420 }, { "grad_norm": 0.25222665071487427, "learning_rate": 4.704820684743347e-05, "loss": 0.0252, "step": 43430 }, { "grad_norm": 0.33357056975364685, "learning_rate": 4.702757477597581e-05, "loss": 0.0224, "step": 43440 }, { "grad_norm": 0.3565250039100647, "learning_rate": 4.700694321242435e-05, "loss": 0.0289, "step": 43450 }, { "grad_norm": 0.34783682227134705, "learning_rate": 4.698631216030441e-05, "loss": 0.0222, "step": 43460 }, { "grad_norm": 0.20980502665042877, "learning_rate": 4.6965681623141314e-05, "loss": 0.0214, "step": 43470 }, { "grad_norm": 0.24514666199684143, "learning_rate": 4.694505160446024e-05, "loss": 0.0209, "step": 43480 }, { "grad_norm": 0.39338862895965576, "learning_rate": 4.692442210778628e-05, "loss": 0.025, "step": 43490 }, { "grad_norm": 0.4875231981277466, "learning_rate": 4.690379313664448e-05, "loss": 0.0288, "step": 43500 }, { "grad_norm": 0.3862665295600891, "learning_rate": 4.688316469455973e-05, "loss": 0.023, "step": 43510 }, { "grad_norm": 0.22794555127620697, "learning_rate": 4.6862536785056914e-05, "loss": 0.0249, "step": 43520 }, { "grad_norm": 0.4509010314941406, "learning_rate": 4.684190941166073e-05, "loss": 0.0243, "step": 43530 }, { "grad_norm": 0.35945913195610046, "learning_rate": 4.682128257789587e-05, "loss": 0.0262, "step": 43540 }, { "grad_norm": 0.5007678866386414, "learning_rate": 4.680065628728687e-05, "loss": 0.028, "step": 43550 }, { "grad_norm": 0.34182465076446533, "learning_rate": 4.678003054335822e-05, "loss": 0.0328, "step": 43560 }, { "grad_norm": 0.39618903398513794, "learning_rate": 4.675940534963428e-05, "loss": 0.0319, "step": 43570 }, { "grad_norm": 0.745048999786377, "learning_rate": 4.673878070963931e-05, "loss": 0.0291, "step": 43580 }, { "grad_norm": 0.31921759247779846, "learning_rate": 4.671815662689756e-05, "loss": 0.0198, "step": 43590 }, { "grad_norm": 0.4031966030597687, "learning_rate": 4.669753310493306e-05, "loss": 0.0274, "step": 43600 }, { "grad_norm": 0.27467864751815796, "learning_rate": 4.6676910147269845e-05, "loss": 0.0274, "step": 43610 }, { "grad_norm": 0.3007173240184784, "learning_rate": 4.665628775743177e-05, "loss": 0.028, "step": 43620 }, { "grad_norm": 0.4527752995491028, "learning_rate": 4.66356659389427e-05, "loss": 0.0222, "step": 43630 }, { "grad_norm": 0.35793766379356384, "learning_rate": 4.661504469532629e-05, "loss": 0.0202, "step": 43640 }, { "grad_norm": 0.32311099767684937, "learning_rate": 4.659442403010617e-05, "loss": 0.0236, "step": 43650 }, { "grad_norm": 0.3588756024837494, "learning_rate": 4.6573803946805845e-05, "loss": 0.0231, "step": 43660 }, { "grad_norm": 0.3422597348690033, "learning_rate": 4.6553184448948695e-05, "loss": 0.0263, "step": 43670 }, { "grad_norm": 0.2847678065299988, "learning_rate": 4.653256554005807e-05, "loss": 0.0208, "step": 43680 }, { "grad_norm": 0.21677342057228088, "learning_rate": 4.6511947223657145e-05, "loss": 0.0281, "step": 43690 }, { "grad_norm": 0.4300840497016907, "learning_rate": 4.649132950326906e-05, "loss": 0.0305, "step": 43700 }, { "grad_norm": 0.3780324459075928, "learning_rate": 4.647071238241679e-05, "loss": 0.0271, "step": 43710 }, { "grad_norm": 0.30775341391563416, "learning_rate": 4.6450095864623264e-05, "loss": 0.0231, "step": 43720 }, { "grad_norm": 0.34261804819107056, "learning_rate": 4.642947995341123e-05, "loss": 0.0245, "step": 43730 }, { "grad_norm": 0.18621517717838287, "learning_rate": 4.640886465230345e-05, "loss": 0.0245, "step": 43740 }, { "grad_norm": 0.44345757365226746, "learning_rate": 4.6388249964822485e-05, "loss": 0.027, "step": 43750 }, { "grad_norm": 0.3855520188808441, "learning_rate": 4.63676358944908e-05, "loss": 0.0285, "step": 43760 }, { "grad_norm": 0.4073784351348877, "learning_rate": 4.634702244483081e-05, "loss": 0.0299, "step": 43770 }, { "grad_norm": 0.6339405179023743, "learning_rate": 4.632640961936475e-05, "loss": 0.0266, "step": 43780 }, { "grad_norm": 0.33367323875427246, "learning_rate": 4.6305797421614835e-05, "loss": 0.0244, "step": 43790 }, { "grad_norm": 0.41164907813072205, "learning_rate": 4.6285185855103105e-05, "loss": 0.03, "step": 43800 }, { "grad_norm": 0.5444576740264893, "learning_rate": 4.626457492335151e-05, "loss": 0.0266, "step": 43810 }, { "grad_norm": 0.19352135062217712, "learning_rate": 4.624396462988188e-05, "loss": 0.0198, "step": 43820 }, { "grad_norm": 0.36021149158477783, "learning_rate": 4.6223354978215985e-05, "loss": 0.0255, "step": 43830 }, { "grad_norm": 0.28275027871131897, "learning_rate": 4.620274597187544e-05, "loss": 0.0296, "step": 43840 }, { "grad_norm": 0.23381780087947845, "learning_rate": 4.6182137614381736e-05, "loss": 0.0272, "step": 43850 }, { "grad_norm": 0.2981725335121155, "learning_rate": 4.616152990925631e-05, "loss": 0.0243, "step": 43860 }, { "grad_norm": 0.2270594835281372, "learning_rate": 4.614092286002041e-05, "loss": 0.0288, "step": 43870 }, { "grad_norm": 0.34788405895233154, "learning_rate": 4.612031647019528e-05, "loss": 0.0267, "step": 43880 }, { "grad_norm": 0.46212005615234375, "learning_rate": 4.609971074330193e-05, "loss": 0.0248, "step": 43890 }, { "grad_norm": 0.4220123589038849, "learning_rate": 4.607910568286136e-05, "loss": 0.0278, "step": 43900 }, { "grad_norm": 0.39293304085731506, "learning_rate": 4.605850129239438e-05, "loss": 0.0288, "step": 43910 }, { "grad_norm": 0.3274528384208679, "learning_rate": 4.603789757542174e-05, "loss": 0.0259, "step": 43920 }, { "grad_norm": 0.4452112913131714, "learning_rate": 4.601729453546404e-05, "loss": 0.0242, "step": 43930 }, { "grad_norm": 0.3089178502559662, "learning_rate": 4.599669217604177e-05, "loss": 0.0283, "step": 43940 }, { "grad_norm": 0.3397660553455353, "learning_rate": 4.597609050067532e-05, "loss": 0.031, "step": 43950 }, { "grad_norm": 0.40521085262298584, "learning_rate": 4.595548951288495e-05, "loss": 0.0305, "step": 43960 }, { "grad_norm": 0.26163604855537415, "learning_rate": 4.593488921619081e-05, "loss": 0.0278, "step": 43970 }, { "grad_norm": 0.36578822135925293, "learning_rate": 4.591428961411289e-05, "loss": 0.0274, "step": 43980 }, { "grad_norm": 0.44285377860069275, "learning_rate": 4.589369071017117e-05, "loss": 0.032, "step": 43990 }, { "grad_norm": 0.32977980375289917, "learning_rate": 4.587309250788538e-05, "loss": 0.0307, "step": 44000 }, { "grad_norm": 0.2113930881023407, "learning_rate": 4.585249501077522e-05, "loss": 0.0238, "step": 44010 }, { "grad_norm": 0.2192375212907791, "learning_rate": 4.583189822236021e-05, "loss": 0.0298, "step": 44020 }, { "grad_norm": 0.31479156017303467, "learning_rate": 4.5811302146159816e-05, "loss": 0.0247, "step": 44030 }, { "grad_norm": 0.40219077467918396, "learning_rate": 4.579070678569332e-05, "loss": 0.0245, "step": 44040 }, { "grad_norm": 0.37870141863822937, "learning_rate": 4.5770112144479904e-05, "loss": 0.0256, "step": 44050 }, { "grad_norm": 0.3469471037387848, "learning_rate": 4.5749518226038645e-05, "loss": 0.0214, "step": 44060 }, { "grad_norm": 0.36300331354141235, "learning_rate": 4.572892503388845e-05, "loss": 0.0196, "step": 44070 }, { "grad_norm": 0.4573170840740204, "learning_rate": 4.570833257154817e-05, "loss": 0.0274, "step": 44080 }, { "grad_norm": 0.3568297028541565, "learning_rate": 4.568774084253646e-05, "loss": 0.0283, "step": 44090 }, { "grad_norm": 0.3438381254673004, "learning_rate": 4.566714985037191e-05, "loss": 0.031, "step": 44100 }, { "grad_norm": 0.4346735179424286, "learning_rate": 4.564655959857295e-05, "loss": 0.0243, "step": 44110 }, { "grad_norm": 0.45567983388900757, "learning_rate": 4.5625970090657893e-05, "loss": 0.0268, "step": 44120 }, { "grad_norm": 0.3400024175643921, "learning_rate": 4.560538133014491e-05, "loss": 0.0246, "step": 44130 }, { "grad_norm": 0.510796308517456, "learning_rate": 4.5584793320552055e-05, "loss": 0.0246, "step": 44140 }, { "grad_norm": 0.3488231599330902, "learning_rate": 4.556420606539728e-05, "loss": 0.0277, "step": 44150 }, { "grad_norm": 0.2726303040981293, "learning_rate": 4.554361956819836e-05, "loss": 0.0272, "step": 44160 }, { "grad_norm": 0.31148749589920044, "learning_rate": 4.552303383247299e-05, "loss": 0.0239, "step": 44170 }, { "grad_norm": 0.23254112899303436, "learning_rate": 4.5502448861738664e-05, "loss": 0.0289, "step": 44180 }, { "grad_norm": 0.25390857458114624, "learning_rate": 4.548186465951284e-05, "loss": 0.0342, "step": 44190 }, { "grad_norm": 0.22752459347248077, "learning_rate": 4.546128122931277e-05, "loss": 0.0307, "step": 44200 }, { "grad_norm": 0.3186878561973572, "learning_rate": 4.544069857465561e-05, "loss": 0.028, "step": 44210 }, { "grad_norm": 0.3167416751384735, "learning_rate": 4.542011669905837e-05, "loss": 0.0224, "step": 44220 }, { "grad_norm": 0.37822696566581726, "learning_rate": 4.53995356060379e-05, "loss": 0.0266, "step": 44230 }, { "grad_norm": 0.2949184477329254, "learning_rate": 4.5378955299110994e-05, "loss": 0.0234, "step": 44240 }, { "grad_norm": 0.40822094678878784, "learning_rate": 4.5358375781794233e-05, "loss": 0.031, "step": 44250 }, { "grad_norm": 1.0742026567459106, "learning_rate": 4.533779705760411e-05, "loss": 0.0318, "step": 44260 }, { "grad_norm": 0.28853777050971985, "learning_rate": 4.5317219130056934e-05, "loss": 0.0246, "step": 44270 }, { "grad_norm": 0.6442931890487671, "learning_rate": 4.5296642002668964e-05, "loss": 0.027, "step": 44280 }, { "grad_norm": 0.5174352526664734, "learning_rate": 4.5276065678956216e-05, "loss": 0.0297, "step": 44290 }, { "grad_norm": 0.3031100034713745, "learning_rate": 4.525549016243466e-05, "loss": 0.0249, "step": 44300 }, { "grad_norm": 0.3041139245033264, "learning_rate": 4.523491545662008e-05, "loss": 0.0238, "step": 44310 }, { "grad_norm": 0.2889525890350342, "learning_rate": 4.5214341565028104e-05, "loss": 0.0322, "step": 44320 }, { "grad_norm": 0.4863657057285309, "learning_rate": 4.519376849117428e-05, "loss": 0.0271, "step": 44330 }, { "grad_norm": 0.1873551607131958, "learning_rate": 4.517319623857395e-05, "loss": 0.0385, "step": 44340 }, { "grad_norm": 0.35430294275283813, "learning_rate": 4.5152624810742394e-05, "loss": 0.0285, "step": 44350 }, { "grad_norm": 0.33462217450141907, "learning_rate": 4.5132054211194664e-05, "loss": 0.029, "step": 44360 }, { "grad_norm": 0.24123398959636688, "learning_rate": 4.511148444344574e-05, "loss": 0.0246, "step": 44370 }, { "grad_norm": 0.3755353093147278, "learning_rate": 4.509091551101041e-05, "loss": 0.0297, "step": 44380 }, { "grad_norm": 0.30573132634162903, "learning_rate": 4.507034741740338e-05, "loss": 0.0242, "step": 44390 }, { "grad_norm": 0.3017348647117615, "learning_rate": 4.5049780166139145e-05, "loss": 0.0269, "step": 44400 }, { "grad_norm": 0.4356144070625305, "learning_rate": 4.5029213760732075e-05, "loss": 0.0356, "step": 44410 }, { "grad_norm": 0.2880639135837555, "learning_rate": 4.5008648204696434e-05, "loss": 0.0301, "step": 44420 }, { "grad_norm": 0.19064360857009888, "learning_rate": 4.4988083501546284e-05, "loss": 0.0225, "step": 44430 }, { "grad_norm": 0.3209873139858246, "learning_rate": 4.4967519654795606e-05, "loss": 0.0245, "step": 44440 }, { "grad_norm": 0.32135695219039917, "learning_rate": 4.494695666795816e-05, "loss": 0.0202, "step": 44450 }, { "grad_norm": 0.43832001090049744, "learning_rate": 4.4926394544547644e-05, "loss": 0.0268, "step": 44460 }, { "grad_norm": 0.9098397493362427, "learning_rate": 4.490583328807752e-05, "loss": 0.0286, "step": 44470 }, { "grad_norm": 0.30299919843673706, "learning_rate": 4.488527290206117e-05, "loss": 0.0336, "step": 44480 }, { "grad_norm": 0.3025049865245819, "learning_rate": 4.48647133900118e-05, "loss": 0.0231, "step": 44490 }, { "grad_norm": 0.2504251003265381, "learning_rate": 4.484415475544244e-05, "loss": 0.0245, "step": 44500 }, { "grad_norm": 0.2381928712129593, "learning_rate": 4.4823597001866035e-05, "loss": 0.0243, "step": 44510 }, { "grad_norm": 0.22899644076824188, "learning_rate": 4.480304013279532e-05, "loss": 0.022, "step": 44520 }, { "grad_norm": 0.3299967646598816, "learning_rate": 4.478248415174292e-05, "loss": 0.0197, "step": 44530 }, { "grad_norm": 0.2508828341960907, "learning_rate": 4.476192906222126e-05, "loss": 0.0294, "step": 44540 }, { "grad_norm": 0.4764086902141571, "learning_rate": 4.474137486774268e-05, "loss": 0.0393, "step": 44550 }, { "grad_norm": 0.2854417860507965, "learning_rate": 4.4720821571819296e-05, "loss": 0.0292, "step": 44560 }, { "grad_norm": 0.22320102155208588, "learning_rate": 4.470026917796314e-05, "loss": 0.0314, "step": 44570 }, { "grad_norm": 0.3525116741657257, "learning_rate": 4.4679717689686005e-05, "loss": 0.0419, "step": 44580 }, { "grad_norm": 0.2903261184692383, "learning_rate": 4.465916711049963e-05, "loss": 0.0222, "step": 44590 }, { "grad_norm": 0.3343314230442047, "learning_rate": 4.4638617443915524e-05, "loss": 0.0242, "step": 44600 }, { "grad_norm": 0.4556993842124939, "learning_rate": 4.4618068693445055e-05, "loss": 0.0279, "step": 44610 }, { "grad_norm": 0.24023909866809845, "learning_rate": 4.459752086259946e-05, "loss": 0.0182, "step": 44620 }, { "grad_norm": 0.5254234075546265, "learning_rate": 4.457697395488977e-05, "loss": 0.0364, "step": 44630 }, { "grad_norm": 0.33781006932258606, "learning_rate": 4.455642797382693e-05, "loss": 0.0302, "step": 44640 }, { "grad_norm": 0.288385272026062, "learning_rate": 4.4535882922921643e-05, "loss": 0.0303, "step": 44650 }, { "grad_norm": 0.24448883533477783, "learning_rate": 4.451533880568455e-05, "loss": 0.0235, "step": 44660 }, { "grad_norm": 0.3193679451942444, "learning_rate": 4.449479562562603e-05, "loss": 0.0322, "step": 44670 }, { "grad_norm": 0.3230924606323242, "learning_rate": 4.447425338625639e-05, "loss": 0.0269, "step": 44680 }, { "grad_norm": 0.36048728227615356, "learning_rate": 4.4453712091085705e-05, "loss": 0.0252, "step": 44690 }, { "grad_norm": 0.30821096897125244, "learning_rate": 4.443317174362392e-05, "loss": 0.0273, "step": 44700 }, { "grad_norm": 0.5082061290740967, "learning_rate": 4.441263234738085e-05, "loss": 0.0266, "step": 44710 }, { "grad_norm": 0.2902437746524811, "learning_rate": 4.4392093905866086e-05, "loss": 0.0243, "step": 44720 }, { "grad_norm": 0.4212825298309326, "learning_rate": 4.437155642258911e-05, "loss": 0.0244, "step": 44730 }, { "grad_norm": 0.21940961480140686, "learning_rate": 4.4351019901059177e-05, "loss": 0.0277, "step": 44740 }, { "grad_norm": 0.5203031301498413, "learning_rate": 4.433048434478547e-05, "loss": 0.0336, "step": 44750 }, { "grad_norm": 0.32277557253837585, "learning_rate": 4.4309949757276906e-05, "loss": 0.0341, "step": 44760 }, { "grad_norm": 0.304217129945755, "learning_rate": 4.428941614204233e-05, "loss": 0.0313, "step": 44770 }, { "grad_norm": 0.253303200006485, "learning_rate": 4.426888350259034e-05, "loss": 0.0293, "step": 44780 }, { "grad_norm": 0.4893215596675873, "learning_rate": 4.424835184242939e-05, "loss": 0.038, "step": 44790 }, { "grad_norm": 0.33819353580474854, "learning_rate": 4.422782116506784e-05, "loss": 0.02, "step": 44800 }, { "grad_norm": 0.41093114018440247, "learning_rate": 4.4207291474013754e-05, "loss": 0.0265, "step": 44810 }, { "grad_norm": 0.25717154145240784, "learning_rate": 4.418676277277514e-05, "loss": 0.0358, "step": 44820 }, { "grad_norm": 0.35729730129241943, "learning_rate": 4.416623506485976e-05, "loss": 0.0291, "step": 44830 }, { "grad_norm": 0.3411773443222046, "learning_rate": 4.414570835377527e-05, "loss": 0.0269, "step": 44840 }, { "grad_norm": 0.37050384283065796, "learning_rate": 4.412518264302909e-05, "loss": 0.0257, "step": 44850 }, { "grad_norm": 0.4671110510826111, "learning_rate": 4.4104657936128535e-05, "loss": 0.0227, "step": 44860 }, { "grad_norm": 0.29101672768592834, "learning_rate": 4.408413423658071e-05, "loss": 0.0278, "step": 44870 }, { "grad_norm": 0.30142608284950256, "learning_rate": 4.406361154789253e-05, "loss": 0.0257, "step": 44880 }, { "grad_norm": 0.44113394618034363, "learning_rate": 4.4043089873570776e-05, "loss": 0.0323, "step": 44890 }, { "grad_norm": 0.4190467298030853, "learning_rate": 4.4022569217122025e-05, "loss": 0.0291, "step": 44900 }, { "grad_norm": 0.2520582675933838, "learning_rate": 4.400204958205274e-05, "loss": 0.0231, "step": 44910 }, { "grad_norm": 0.32399222254753113, "learning_rate": 4.3981530971869125e-05, "loss": 0.0307, "step": 44920 }, { "grad_norm": 0.2589002549648285, "learning_rate": 4.396101339007727e-05, "loss": 0.0227, "step": 44930 }, { "grad_norm": 0.5422623157501221, "learning_rate": 4.394049684018304e-05, "loss": 0.0296, "step": 44940 }, { "grad_norm": 0.37238043546676636, "learning_rate": 4.3919981325692186e-05, "loss": 0.0293, "step": 44950 }, { "grad_norm": 0.18343821167945862, "learning_rate": 4.389946685011024e-05, "loss": 0.0206, "step": 44960 }, { "grad_norm": 0.25633707642555237, "learning_rate": 4.387895341694255e-05, "loss": 0.0278, "step": 44970 }, { "grad_norm": 0.2196756899356842, "learning_rate": 4.3858441029694324e-05, "loss": 0.0323, "step": 44980 }, { "grad_norm": 0.46013668179512024, "learning_rate": 4.3837929691870527e-05, "loss": 0.026, "step": 44990 }, { "grad_norm": 0.4775107204914093, "learning_rate": 4.381741940697604e-05, "loss": 0.0327, "step": 45000 }, { "grad_norm": 0.300311803817749, "learning_rate": 4.379691017851547e-05, "loss": 0.0228, "step": 45010 }, { "grad_norm": 0.567525327205658, "learning_rate": 4.3776402009993304e-05, "loss": 0.0266, "step": 45020 }, { "grad_norm": 0.334858775138855, "learning_rate": 4.3755894904913794e-05, "loss": 0.0268, "step": 45030 }, { "grad_norm": 0.29364529252052307, "learning_rate": 4.373538886678109e-05, "loss": 0.0217, "step": 45040 }, { "grad_norm": 0.26820117235183716, "learning_rate": 4.371488389909909e-05, "loss": 0.0202, "step": 45050 }, { "grad_norm": 0.25886955857276917, "learning_rate": 4.3694380005371515e-05, "loss": 0.0297, "step": 45060 }, { "grad_norm": 0.3762136697769165, "learning_rate": 4.367387718910196e-05, "loss": 0.0215, "step": 45070 }, { "grad_norm": 0.2956061363220215, "learning_rate": 4.3653375453793764e-05, "loss": 0.0243, "step": 45080 }, { "grad_norm": 0.5125382542610168, "learning_rate": 4.3632874802950136e-05, "loss": 0.0273, "step": 45090 }, { "grad_norm": 0.3732176125049591, "learning_rate": 4.3612375240074034e-05, "loss": 0.0301, "step": 45100 }, { "grad_norm": 0.8329445719718933, "learning_rate": 4.3591876768668325e-05, "loss": 0.0267, "step": 45110 }, { "grad_norm": 0.2642480134963989, "learning_rate": 4.3571379392235605e-05, "loss": 0.0218, "step": 45120 }, { "grad_norm": 0.37139126658439636, "learning_rate": 4.3550883114278335e-05, "loss": 0.0235, "step": 45130 }, { "grad_norm": 0.3435308337211609, "learning_rate": 4.353038793829876e-05, "loss": 0.0216, "step": 45140 }, { "grad_norm": 0.35966119170188904, "learning_rate": 4.350989386779891e-05, "loss": 0.0203, "step": 45150 }, { "grad_norm": 0.2786075472831726, "learning_rate": 4.3489400906280724e-05, "loss": 0.0328, "step": 45160 }, { "grad_norm": 0.5217087268829346, "learning_rate": 4.3468909057245845e-05, "loss": 0.0355, "step": 45170 }, { "grad_norm": 0.2371935397386551, "learning_rate": 4.3448418324195794e-05, "loss": 0.0249, "step": 45180 }, { "grad_norm": 0.2446201592683792, "learning_rate": 4.342792871063184e-05, "loss": 0.0227, "step": 45190 }, { "grad_norm": 0.282664954662323, "learning_rate": 4.3407440220055145e-05, "loss": 0.0264, "step": 45200 }, { "grad_norm": 0.2878715991973877, "learning_rate": 4.33869528559666e-05, "loss": 0.022, "step": 45210 }, { "grad_norm": 0.33196383714675903, "learning_rate": 4.336646662186696e-05, "loss": 0.0195, "step": 45220 }, { "grad_norm": 0.259542316198349, "learning_rate": 4.334598152125672e-05, "loss": 0.0319, "step": 45230 }, { "grad_norm": 0.29311397671699524, "learning_rate": 4.3325497557636276e-05, "loss": 0.0289, "step": 45240 }, { "grad_norm": 0.3129536211490631, "learning_rate": 4.330501473450574e-05, "loss": 0.0265, "step": 45250 }, { "grad_norm": 0.22861820459365845, "learning_rate": 4.328453305536507e-05, "loss": 0.0227, "step": 45260 }, { "grad_norm": 0.2619404196739197, "learning_rate": 4.326405252371404e-05, "loss": 0.0237, "step": 45270 }, { "grad_norm": 0.9402861595153809, "learning_rate": 4.324357314305221e-05, "loss": 0.0247, "step": 45280 }, { "grad_norm": 0.3019530773162842, "learning_rate": 4.3223094916878945e-05, "loss": 0.0223, "step": 45290 }, { "grad_norm": 0.25123193860054016, "learning_rate": 4.320261784869338e-05, "loss": 0.0345, "step": 45300 }, { "grad_norm": 0.4204620122909546, "learning_rate": 4.318214194199455e-05, "loss": 0.0289, "step": 45310 }, { "grad_norm": 0.19113734364509583, "learning_rate": 4.316166720028118e-05, "loss": 0.0188, "step": 45320 }, { "grad_norm": 0.37020909786224365, "learning_rate": 4.3141193627051864e-05, "loss": 0.0223, "step": 45330 }, { "grad_norm": 0.39170464873313904, "learning_rate": 4.312072122580496e-05, "loss": 0.0245, "step": 45340 }, { "grad_norm": 0.2560989558696747, "learning_rate": 4.3100250000038646e-05, "loss": 0.0228, "step": 45350 }, { "grad_norm": 0.4170224964618683, "learning_rate": 4.307977995325091e-05, "loss": 0.0391, "step": 45360 }, { "grad_norm": 0.36432698369026184, "learning_rate": 4.30593110889395e-05, "loss": 0.026, "step": 45370 }, { "grad_norm": 0.20597098767757416, "learning_rate": 4.3038843410602016e-05, "loss": 0.021, "step": 45380 }, { "grad_norm": 0.27625739574432373, "learning_rate": 4.3018376921735774e-05, "loss": 0.0242, "step": 45390 }, { "grad_norm": 0.22721248865127563, "learning_rate": 4.299791162583799e-05, "loss": 0.0235, "step": 45400 }, { "grad_norm": 0.2460644245147705, "learning_rate": 4.29774475264056e-05, "loss": 0.03, "step": 45410 }, { "grad_norm": 0.2835613489151001, "learning_rate": 4.2956984626935365e-05, "loss": 0.0217, "step": 45420 }, { "grad_norm": 0.2371666580438614, "learning_rate": 4.293652293092383e-05, "loss": 0.022, "step": 45430 }, { "grad_norm": 0.27821239829063416, "learning_rate": 4.2916062441867324e-05, "loss": 0.024, "step": 45440 }, { "grad_norm": 0.23712806403636932, "learning_rate": 4.289560316326201e-05, "loss": 0.0279, "step": 45450 }, { "grad_norm": 0.40062829852104187, "learning_rate": 4.28751450986038e-05, "loss": 0.0267, "step": 45460 }, { "grad_norm": 1.0451390743255615, "learning_rate": 4.2854688251388444e-05, "loss": 0.0293, "step": 45470 }, { "grad_norm": 0.25949421525001526, "learning_rate": 4.2834232625111425e-05, "loss": 0.0319, "step": 45480 }, { "grad_norm": 0.2880914509296417, "learning_rate": 4.2813778223268086e-05, "loss": 0.0246, "step": 45490 }, { "grad_norm": 0.7577462792396545, "learning_rate": 4.2793325049353477e-05, "loss": 0.0219, "step": 45500 }, { "grad_norm": 0.41850653290748596, "learning_rate": 4.2772873106862535e-05, "loss": 0.0231, "step": 45510 }, { "grad_norm": 0.2121780514717102, "learning_rate": 4.275242239928993e-05, "loss": 0.0205, "step": 45520 }, { "grad_norm": 0.21805354952812195, "learning_rate": 4.273197293013009e-05, "loss": 0.02, "step": 45530 }, { "grad_norm": 0.38403210043907166, "learning_rate": 4.271152470287731e-05, "loss": 0.0206, "step": 45540 }, { "grad_norm": 0.3632057309150696, "learning_rate": 4.2691077721025594e-05, "loss": 0.0226, "step": 45550 }, { "grad_norm": 0.23130202293395996, "learning_rate": 4.267063198806883e-05, "loss": 0.021, "step": 45560 }, { "grad_norm": 0.28351712226867676, "learning_rate": 4.2650187507500574e-05, "loss": 0.0231, "step": 45570 }, { "grad_norm": 0.35881784558296204, "learning_rate": 4.2629744282814275e-05, "loss": 0.0194, "step": 45580 }, { "grad_norm": 0.19710388779640198, "learning_rate": 4.2609302317503074e-05, "loss": 0.0179, "step": 45590 }, { "grad_norm": 0.1826043576002121, "learning_rate": 4.258886161505999e-05, "loss": 0.0209, "step": 45600 }, { "grad_norm": 0.2345539927482605, "learning_rate": 4.2568422178977775e-05, "loss": 0.0226, "step": 45610 }, { "grad_norm": 0.3172560930252075, "learning_rate": 4.254798401274894e-05, "loss": 0.0292, "step": 45620 }, { "grad_norm": 0.4295717179775238, "learning_rate": 4.252754711986583e-05, "loss": 0.0261, "step": 45630 }, { "grad_norm": 0.294609934091568, "learning_rate": 4.250711150382052e-05, "loss": 0.0203, "step": 45640 }, { "grad_norm": 0.6314966678619385, "learning_rate": 4.248667716810495e-05, "loss": 0.0246, "step": 45650 }, { "grad_norm": 0.2799989581108093, "learning_rate": 4.246624411621074e-05, "loss": 0.0221, "step": 45660 }, { "grad_norm": 0.37374308705329895, "learning_rate": 4.244581235162938e-05, "loss": 0.0212, "step": 45670 }, { "grad_norm": 0.23617291450500488, "learning_rate": 4.2425381877852075e-05, "loss": 0.0277, "step": 45680 }, { "grad_norm": 0.27005431056022644, "learning_rate": 4.2404952698369856e-05, "loss": 0.0227, "step": 45690 }, { "grad_norm": 0.2666109800338745, "learning_rate": 4.238452481667349e-05, "loss": 0.0232, "step": 45700 }, { "grad_norm": 0.4037668704986572, "learning_rate": 4.2364098236253526e-05, "loss": 0.031, "step": 45710 }, { "grad_norm": 0.27731865644454956, "learning_rate": 4.2343672960600356e-05, "loss": 0.0232, "step": 45720 }, { "grad_norm": 0.32114872336387634, "learning_rate": 4.232324899320406e-05, "loss": 0.018, "step": 45730 }, { "grad_norm": 0.3031271994113922, "learning_rate": 4.230282633755457e-05, "loss": 0.0235, "step": 45740 }, { "grad_norm": 0.3300268352031708, "learning_rate": 4.2282404997141515e-05, "loss": 0.0227, "step": 45750 }, { "grad_norm": 0.34205809235572815, "learning_rate": 4.2261984975454397e-05, "loss": 0.0257, "step": 45760 }, { "grad_norm": 0.26743194460868835, "learning_rate": 4.224156627598239e-05, "loss": 0.0208, "step": 45770 }, { "grad_norm": 0.3067971467971802, "learning_rate": 4.222114890221453e-05, "loss": 0.0261, "step": 45780 }, { "grad_norm": 0.6052331328392029, "learning_rate": 4.2200732857639546e-05, "loss": 0.0283, "step": 45790 }, { "grad_norm": 0.25444936752319336, "learning_rate": 4.2180318145746035e-05, "loss": 0.0305, "step": 45800 }, { "grad_norm": 0.33335113525390625, "learning_rate": 4.215990477002227e-05, "loss": 0.0312, "step": 45810 }, { "grad_norm": 0.3287651538848877, "learning_rate": 4.2139492733956356e-05, "loss": 0.0211, "step": 45820 }, { "grad_norm": 0.4129413068294525, "learning_rate": 4.211908204103615e-05, "loss": 0.0337, "step": 45830 }, { "grad_norm": 0.3206149935722351, "learning_rate": 4.2098672694749265e-05, "loss": 0.0377, "step": 45840 }, { "grad_norm": 0.32648584246635437, "learning_rate": 4.2078264698583133e-05, "loss": 0.0288, "step": 45850 }, { "grad_norm": 0.29902032017707825, "learning_rate": 4.205785805602488e-05, "loss": 0.0255, "step": 45860 }, { "grad_norm": 0.45251092314720154, "learning_rate": 4.203745277056149e-05, "loss": 0.0297, "step": 45870 }, { "grad_norm": 0.2564891576766968, "learning_rate": 4.201704884567964e-05, "loss": 0.0243, "step": 45880 }, { "grad_norm": 0.3857046663761139, "learning_rate": 4.1996646284865816e-05, "loss": 0.0398, "step": 45890 }, { "grad_norm": 0.4172080457210541, "learning_rate": 4.197624509160625e-05, "loss": 0.0232, "step": 45900 }, { "grad_norm": 0.2986234128475189, "learning_rate": 4.195584526938692e-05, "loss": 0.0251, "step": 45910 }, { "grad_norm": 0.4167120158672333, "learning_rate": 4.193544682169365e-05, "loss": 0.0283, "step": 45920 }, { "grad_norm": 0.32598766684532166, "learning_rate": 4.1915049752011946e-05, "loss": 0.0205, "step": 45930 }, { "grad_norm": 0.32996034622192383, "learning_rate": 4.189465406382712e-05, "loss": 0.0231, "step": 45940 }, { "grad_norm": 0.2740716338157654, "learning_rate": 4.187425976062422e-05, "loss": 0.0322, "step": 45950 }, { "grad_norm": 0.417696475982666, "learning_rate": 4.18538668458881e-05, "loss": 0.0239, "step": 45960 }, { "grad_norm": 0.30587664246559143, "learning_rate": 4.183347532310333e-05, "loss": 0.0226, "step": 45970 }, { "grad_norm": 0.26021385192871094, "learning_rate": 4.181308519575429e-05, "loss": 0.0229, "step": 45980 }, { "grad_norm": 0.30485159158706665, "learning_rate": 4.179269646732507e-05, "loss": 0.0322, "step": 45990 }, { "grad_norm": 0.33209559321403503, "learning_rate": 4.177230914129954e-05, "loss": 0.0215, "step": 46000 }, { "grad_norm": 0.36650383472442627, "learning_rate": 4.175192322116136e-05, "loss": 0.0236, "step": 46010 }, { "grad_norm": 0.3831484913825989, "learning_rate": 4.173153871039391e-05, "loss": 0.0207, "step": 46020 }, { "grad_norm": 0.47404858469963074, "learning_rate": 4.171115561248036e-05, "loss": 0.0255, "step": 46030 }, { "grad_norm": 0.2569223642349243, "learning_rate": 4.16907739309036e-05, "loss": 0.0234, "step": 46040 }, { "grad_norm": 0.438471257686615, "learning_rate": 4.167039366914633e-05, "loss": 0.0236, "step": 46050 }, { "grad_norm": 0.3732922375202179, "learning_rate": 4.165001483069096e-05, "loss": 0.0254, "step": 46060 }, { "grad_norm": 0.2971816658973694, "learning_rate": 4.16296374190197e-05, "loss": 0.0291, "step": 46070 }, { "grad_norm": 0.28446316719055176, "learning_rate": 4.1609261437614464e-05, "loss": 0.0235, "step": 46080 }, { "grad_norm": 0.4730260372161865, "learning_rate": 4.158888688995696e-05, "loss": 0.0322, "step": 46090 }, { "grad_norm": 0.357930064201355, "learning_rate": 4.1568513779528645e-05, "loss": 0.0211, "step": 46100 }, { "grad_norm": 0.3520495891571045, "learning_rate": 4.1548142109810704e-05, "loss": 0.0315, "step": 46110 }, { "grad_norm": 0.4182930290699005, "learning_rate": 4.152777188428414e-05, "loss": 0.0212, "step": 46120 }, { "grad_norm": 0.2434624433517456, "learning_rate": 4.1507403106429646e-05, "loss": 0.0252, "step": 46130 }, { "grad_norm": 0.34298545122146606, "learning_rate": 4.148703577972768e-05, "loss": 0.0263, "step": 46140 }, { "grad_norm": 0.46850770711898804, "learning_rate": 4.146666990765846e-05, "loss": 0.0229, "step": 46150 }, { "grad_norm": 0.26305514574050903, "learning_rate": 4.1446305493701986e-05, "loss": 0.023, "step": 46160 }, { "grad_norm": 0.375844806432724, "learning_rate": 4.142594254133796e-05, "loss": 0.0245, "step": 46170 }, { "grad_norm": 0.3991389274597168, "learning_rate": 4.1405581054045835e-05, "loss": 0.032, "step": 46180 }, { "grad_norm": 0.3780858814716339, "learning_rate": 4.1385221035304864e-05, "loss": 0.0235, "step": 46190 }, { "grad_norm": 0.2639467716217041, "learning_rate": 4.136486248859398e-05, "loss": 0.0287, "step": 46200 }, { "grad_norm": 0.8839953541755676, "learning_rate": 4.1344505417391955e-05, "loss": 0.0259, "step": 46210 }, { "grad_norm": 0.2374337613582611, "learning_rate": 4.132414982517721e-05, "loss": 0.0228, "step": 46220 }, { "grad_norm": 0.26224789023399353, "learning_rate": 4.130379571542798e-05, "loss": 0.0247, "step": 46230 }, { "grad_norm": 0.24289128184318542, "learning_rate": 4.12834430916222e-05, "loss": 0.0243, "step": 46240 }, { "grad_norm": 0.24779947102069855, "learning_rate": 4.126309195723763e-05, "loss": 0.0195, "step": 46250 }, { "grad_norm": 0.35292506217956543, "learning_rate": 4.124274231575168e-05, "loss": 0.0287, "step": 46260 }, { "grad_norm": 0.2436390072107315, "learning_rate": 4.122239417064154e-05, "loss": 0.0198, "step": 46270 }, { "grad_norm": 0.2946391999721527, "learning_rate": 4.1202047525384184e-05, "loss": 0.019, "step": 46280 }, { "grad_norm": 0.2691001296043396, "learning_rate": 4.118170238345627e-05, "loss": 0.0208, "step": 46290 }, { "grad_norm": 0.28391605615615845, "learning_rate": 4.1161358748334256e-05, "loss": 0.0253, "step": 46300 }, { "grad_norm": 0.24490487575531006, "learning_rate": 4.1141016623494266e-05, "loss": 0.0237, "step": 46310 }, { "grad_norm": 0.3549215495586395, "learning_rate": 4.112067601241227e-05, "loss": 0.0235, "step": 46320 }, { "grad_norm": 0.3144702911376953, "learning_rate": 4.110033691856387e-05, "loss": 0.0278, "step": 46330 }, { "grad_norm": 0.590678334236145, "learning_rate": 4.107999934542451e-05, "loss": 0.0352, "step": 46340 }, { "grad_norm": 0.29720309376716614, "learning_rate": 4.105966329646928e-05, "loss": 0.0256, "step": 46350 }, { "grad_norm": 0.6726054549217224, "learning_rate": 4.103932877517308e-05, "loss": 0.0293, "step": 46360 }, { "grad_norm": 0.5098319053649902, "learning_rate": 4.101899578501052e-05, "loss": 0.0216, "step": 46370 }, { "grad_norm": 0.24195723235607147, "learning_rate": 4.099866432945595e-05, "loss": 0.025, "step": 46380 }, { "grad_norm": 0.39126092195510864, "learning_rate": 4.097833441198346e-05, "loss": 0.0221, "step": 46390 }, { "grad_norm": 0.1815720647573471, "learning_rate": 4.095800603606685e-05, "loss": 0.0192, "step": 46400 }, { "grad_norm": 0.2960331439971924, "learning_rate": 4.093767920517975e-05, "loss": 0.0246, "step": 46410 }, { "grad_norm": 0.2913340628147125, "learning_rate": 4.091735392279539e-05, "loss": 0.0253, "step": 46420 }, { "grad_norm": 0.20327366888523102, "learning_rate": 4.089703019238685e-05, "loss": 0.0227, "step": 46430 }, { "grad_norm": 0.32288986444473267, "learning_rate": 4.0876708017426866e-05, "loss": 0.0195, "step": 46440 }, { "grad_norm": 0.9407678246498108, "learning_rate": 4.085638740138798e-05, "loss": 0.027, "step": 46450 }, { "grad_norm": 0.324373334646225, "learning_rate": 4.08360683477424e-05, "loss": 0.0265, "step": 46460 }, { "grad_norm": 0.3169960677623749, "learning_rate": 4.0815750859962085e-05, "loss": 0.0256, "step": 46470 }, { "grad_norm": 0.362544447183609, "learning_rate": 4.079543494151879e-05, "loss": 0.0304, "step": 46480 }, { "grad_norm": 0.34243518114089966, "learning_rate": 4.07751205958839e-05, "loss": 0.0219, "step": 46490 }, { "grad_norm": 0.35876530408859253, "learning_rate": 4.0754807826528615e-05, "loss": 0.0215, "step": 46500 }, { "grad_norm": 0.2808051109313965, "learning_rate": 4.073449663692379e-05, "loss": 0.0287, "step": 46510 }, { "grad_norm": 0.22954650223255157, "learning_rate": 4.07141870305401e-05, "loss": 0.0216, "step": 46520 }, { "grad_norm": 0.40430423617362976, "learning_rate": 4.0693879010847866e-05, "loss": 0.0347, "step": 46530 }, { "grad_norm": 0.38008421659469604, "learning_rate": 4.06735725813172e-05, "loss": 0.0252, "step": 46540 }, { "grad_norm": 0.36999744176864624, "learning_rate": 4.0653267745417903e-05, "loss": 0.02, "step": 46550 }, { "grad_norm": 0.2459939867258072, "learning_rate": 4.063296450661949e-05, "loss": 0.018, "step": 46560 }, { "grad_norm": 0.22838759422302246, "learning_rate": 4.061266286839128e-05, "loss": 0.0213, "step": 46570 }, { "grad_norm": 1.0822336673736572, "learning_rate": 4.0592362834202225e-05, "loss": 0.0306, "step": 46580 }, { "grad_norm": 0.43785059452056885, "learning_rate": 4.057206440752107e-05, "loss": 0.0232, "step": 46590 }, { "grad_norm": 0.27147459983825684, "learning_rate": 4.0551767591816245e-05, "loss": 0.0197, "step": 46600 }, { "grad_norm": 0.24262608587741852, "learning_rate": 4.0531472390555935e-05, "loss": 0.0292, "step": 46610 }, { "grad_norm": 0.4117071032524109, "learning_rate": 4.051117880720802e-05, "loss": 0.0231, "step": 46620 }, { "grad_norm": 0.23064987361431122, "learning_rate": 4.049088684524015e-05, "loss": 0.0214, "step": 46630 }, { "grad_norm": 0.38007622957229614, "learning_rate": 4.0470596508119636e-05, "loss": 0.0214, "step": 46640 }, { "grad_norm": 0.26797688007354736, "learning_rate": 4.0450307799313524e-05, "loss": 0.0225, "step": 46650 }, { "grad_norm": 0.3011663556098938, "learning_rate": 4.0430020722288656e-05, "loss": 0.0208, "step": 46660 }, { "grad_norm": 0.34940269589424133, "learning_rate": 4.040973528051148e-05, "loss": 0.0273, "step": 46670 }, { "grad_norm": 0.2371700257062912, "learning_rate": 4.038945147744827e-05, "loss": 0.0263, "step": 46680 }, { "grad_norm": 0.2393054962158203, "learning_rate": 4.0369169316564945e-05, "loss": 0.02, "step": 46690 }, { "grad_norm": 1.470335841178894, "learning_rate": 4.034888880132718e-05, "loss": 0.03, "step": 46700 }, { "grad_norm": 0.8028993606567383, "learning_rate": 4.032860993520035e-05, "loss": 0.0231, "step": 46710 }, { "grad_norm": 0.3422888219356537, "learning_rate": 4.030833272164959e-05, "loss": 0.0232, "step": 46720 }, { "grad_norm": 0.24311359226703644, "learning_rate": 4.0288057164139705e-05, "loss": 0.0221, "step": 46730 }, { "grad_norm": 0.3606208562850952, "learning_rate": 4.0267783266135205e-05, "loss": 0.0199, "step": 46740 }, { "grad_norm": 0.6600911617279053, "learning_rate": 4.024751103110039e-05, "loss": 0.0261, "step": 46750 }, { "grad_norm": 0.18246503174304962, "learning_rate": 4.0227240462499176e-05, "loss": 0.0264, "step": 46760 }, { "grad_norm": 0.25659289956092834, "learning_rate": 4.020697156379531e-05, "loss": 0.0236, "step": 46770 }, { "grad_norm": 0.3213455379009247, "learning_rate": 4.018670433845215e-05, "loss": 0.0278, "step": 46780 }, { "grad_norm": 0.22075650095939636, "learning_rate": 4.016643878993284e-05, "loss": 0.0194, "step": 46790 }, { "grad_norm": 0.282693088054657, "learning_rate": 4.014617492170017e-05, "loss": 0.0229, "step": 46800 }, { "grad_norm": 0.2779712378978729, "learning_rate": 4.0125912737216726e-05, "loss": 0.0298, "step": 46810 }, { "grad_norm": 0.29723161458969116, "learning_rate": 4.0105652239944735e-05, "loss": 0.0283, "step": 46820 }, { "grad_norm": 0.24865910410881042, "learning_rate": 4.0085393433346144e-05, "loss": 0.0243, "step": 46830 }, { "grad_norm": 0.2046583890914917, "learning_rate": 4.006513632088268e-05, "loss": 0.0187, "step": 46840 }, { "grad_norm": 0.28349974751472473, "learning_rate": 4.004488090601567e-05, "loss": 0.0177, "step": 46850 }, { "grad_norm": 0.269297331571579, "learning_rate": 4.002462719220626e-05, "loss": 0.0211, "step": 46860 }, { "grad_norm": 0.18381556868553162, "learning_rate": 4.000437518291522e-05, "loss": 0.0273, "step": 46870 }, { "grad_norm": 0.3383823335170746, "learning_rate": 3.9984124881603094e-05, "loss": 0.0257, "step": 46880 }, { "grad_norm": 0.30722442269325256, "learning_rate": 3.9963876291730086e-05, "loss": 0.0284, "step": 46890 }, { "grad_norm": 0.22891266644001007, "learning_rate": 3.994362941675614e-05, "loss": 0.0225, "step": 46900 }, { "grad_norm": 0.31250470876693726, "learning_rate": 3.992338426014088e-05, "loss": 0.0267, "step": 46910 }, { "grad_norm": 0.5509258508682251, "learning_rate": 3.9903140825343636e-05, "loss": 0.0276, "step": 46920 }, { "grad_norm": 1.0218775272369385, "learning_rate": 3.98828991158235e-05, "loss": 0.0231, "step": 46930 }, { "grad_norm": 0.665651798248291, "learning_rate": 3.9862659135039185e-05, "loss": 0.0334, "step": 46940 }, { "grad_norm": 0.3911222517490387, "learning_rate": 3.984242088644918e-05, "loss": 0.0239, "step": 46950 }, { "grad_norm": 0.2982257306575775, "learning_rate": 3.9822184373511615e-05, "loss": 0.0254, "step": 46960 }, { "grad_norm": 0.1868986338376999, "learning_rate": 3.98019495996844e-05, "loss": 0.0233, "step": 46970 }, { "grad_norm": 0.2819393277168274, "learning_rate": 3.978171656842507e-05, "loss": 0.0209, "step": 46980 }, { "grad_norm": 0.35584723949432373, "learning_rate": 3.976148528319091e-05, "loss": 0.0251, "step": 46990 }, { "grad_norm": 0.2667015790939331, "learning_rate": 3.974125574743888e-05, "loss": 0.0294, "step": 47000 }, { "grad_norm": 0.29894721508026123, "learning_rate": 3.9721027964625686e-05, "loss": 0.0241, "step": 47010 }, { "grad_norm": 0.22648662328720093, "learning_rate": 3.9700801938207676e-05, "loss": 0.0254, "step": 47020 }, { "grad_norm": 0.31689178943634033, "learning_rate": 3.9680577671640916e-05, "loss": 0.0258, "step": 47030 }, { "grad_norm": 0.3635892868041992, "learning_rate": 3.966035516838121e-05, "loss": 0.0347, "step": 47040 }, { "grad_norm": 0.3567042350769043, "learning_rate": 3.9640134431884014e-05, "loss": 0.0256, "step": 47050 }, { "grad_norm": 0.2771196663379669, "learning_rate": 3.961991546560451e-05, "loss": 0.0273, "step": 47060 }, { "grad_norm": 0.2413787692785263, "learning_rate": 3.959969827299753e-05, "loss": 0.0266, "step": 47070 }, { "grad_norm": 0.47932836413383484, "learning_rate": 3.9579482857517684e-05, "loss": 0.0229, "step": 47080 }, { "grad_norm": 0.3384127914905548, "learning_rate": 3.955926922261921e-05, "loss": 0.0239, "step": 47090 }, { "grad_norm": 0.720106303691864, "learning_rate": 3.9539057371756084e-05, "loss": 0.0289, "step": 47100 }, { "grad_norm": 0.3347909450531006, "learning_rate": 3.951884730838195e-05, "loss": 0.0182, "step": 47110 }, { "grad_norm": 0.35886672139167786, "learning_rate": 3.949863903595012e-05, "loss": 0.018, "step": 47120 }, { "grad_norm": 0.27777376770973206, "learning_rate": 3.947843255791369e-05, "loss": 0.0191, "step": 47130 }, { "grad_norm": 0.46576088666915894, "learning_rate": 3.9458227877725364e-05, "loss": 0.0303, "step": 47140 }, { "grad_norm": 0.28811877965927124, "learning_rate": 3.943802499883758e-05, "loss": 0.0167, "step": 47150 }, { "grad_norm": 0.2554353177547455, "learning_rate": 3.9417823924702437e-05, "loss": 0.0302, "step": 47160 }, { "grad_norm": 0.44172418117523193, "learning_rate": 3.939762465877178e-05, "loss": 0.0269, "step": 47170 }, { "grad_norm": 0.3854711949825287, "learning_rate": 3.937742720449708e-05, "loss": 0.0249, "step": 47180 }, { "grad_norm": 0.2710590064525604, "learning_rate": 3.9357231565329563e-05, "loss": 0.0233, "step": 47190 }, { "grad_norm": 0.6206216216087341, "learning_rate": 3.933703774472008e-05, "loss": 0.0272, "step": 47200 }, { "grad_norm": 0.2427666038274765, "learning_rate": 3.93168457461192e-05, "loss": 0.0179, "step": 47210 }, { "grad_norm": 0.2627488970756531, "learning_rate": 3.9296655572977216e-05, "loss": 0.0183, "step": 47220 }, { "grad_norm": 0.35919302701950073, "learning_rate": 3.927646722874404e-05, "loss": 0.0241, "step": 47230 }, { "grad_norm": 0.502452552318573, "learning_rate": 3.925628071686934e-05, "loss": 0.0247, "step": 47240 }, { "grad_norm": 0.2587682902812958, "learning_rate": 3.9236096040802415e-05, "loss": 0.0221, "step": 47250 }, { "grad_norm": 0.47167059779167175, "learning_rate": 3.9215913203992294e-05, "loss": 0.0291, "step": 47260 }, { "grad_norm": 0.30951792001724243, "learning_rate": 3.9195732209887645e-05, "loss": 0.028, "step": 47270 }, { "grad_norm": 0.1892513632774353, "learning_rate": 3.9175553061936875e-05, "loss": 0.0187, "step": 47280 }, { "grad_norm": 0.30319011211395264, "learning_rate": 3.9155375763588045e-05, "loss": 0.0304, "step": 47290 }, { "grad_norm": 0.2248193472623825, "learning_rate": 3.913520031828889e-05, "loss": 0.0182, "step": 47300 }, { "grad_norm": 0.36621493101119995, "learning_rate": 3.911502672948685e-05, "loss": 0.0246, "step": 47310 }, { "grad_norm": 0.3195318281650543, "learning_rate": 3.9094855000629014e-05, "loss": 0.0196, "step": 47320 }, { "grad_norm": 0.25337743759155273, "learning_rate": 3.907468513516223e-05, "loss": 0.017, "step": 47330 }, { "grad_norm": 0.3627724051475525, "learning_rate": 3.905451713653294e-05, "loss": 0.023, "step": 47340 }, { "grad_norm": 0.2607434093952179, "learning_rate": 3.903435100818731e-05, "loss": 0.0174, "step": 47350 }, { "grad_norm": 0.3127043843269348, "learning_rate": 3.901418675357117e-05, "loss": 0.0259, "step": 47360 }, { "grad_norm": 0.2917044162750244, "learning_rate": 3.8994024376130075e-05, "loss": 0.0266, "step": 47370 }, { "grad_norm": 0.35167980194091797, "learning_rate": 3.8973863879309194e-05, "loss": 0.0242, "step": 47380 }, { "grad_norm": 0.4336336553096771, "learning_rate": 3.8953705266553394e-05, "loss": 0.0268, "step": 47390 }, { "grad_norm": 0.32118159532546997, "learning_rate": 3.893354854130727e-05, "loss": 0.0272, "step": 47400 }, { "grad_norm": 0.2761676609516144, "learning_rate": 3.8913393707015006e-05, "loss": 0.024, "step": 47410 }, { "grad_norm": 0.38966095447540283, "learning_rate": 3.889324076712056e-05, "loss": 0.0248, "step": 47420 }, { "grad_norm": 0.29429730772972107, "learning_rate": 3.8873089725067476e-05, "loss": 0.0206, "step": 47430 }, { "grad_norm": 0.4101874530315399, "learning_rate": 3.885294058429905e-05, "loss": 0.0313, "step": 47440 }, { "grad_norm": 0.22036126255989075, "learning_rate": 3.8832793348258206e-05, "loss": 0.0207, "step": 47450 }, { "grad_norm": 0.2386276125907898, "learning_rate": 3.881264802038756e-05, "loss": 0.0331, "step": 47460 }, { "grad_norm": 0.343379408121109, "learning_rate": 3.879250460412939e-05, "loss": 0.023, "step": 47470 }, { "grad_norm": 0.23626936972141266, "learning_rate": 3.8772363102925644e-05, "loss": 0.0206, "step": 47480 }, { "grad_norm": 0.2334052324295044, "learning_rate": 3.875222352021798e-05, "loss": 0.0195, "step": 47490 }, { "grad_norm": 0.23465049266815186, "learning_rate": 3.8732085859447686e-05, "loss": 0.0225, "step": 47500 }, { "grad_norm": 0.37113675475120544, "learning_rate": 3.871195012405575e-05, "loss": 0.0206, "step": 47510 }, { "grad_norm": 0.28150689601898193, "learning_rate": 3.869181631748278e-05, "loss": 0.0234, "step": 47520 }, { "grad_norm": 0.5903366208076477, "learning_rate": 3.867168444316915e-05, "loss": 0.022, "step": 47530 }, { "grad_norm": 0.17146694660186768, "learning_rate": 3.865155450455481e-05, "loss": 0.0152, "step": 47540 }, { "grad_norm": 0.5704384446144104, "learning_rate": 3.8631426505079426e-05, "loss": 0.0303, "step": 47550 }, { "grad_norm": 0.46249204874038696, "learning_rate": 3.8611300448182304e-05, "loss": 0.0228, "step": 47560 }, { "grad_norm": 0.34667545557022095, "learning_rate": 3.859117633730248e-05, "loss": 0.0256, "step": 47570 }, { "grad_norm": 0.3921087384223938, "learning_rate": 3.857105417587858e-05, "loss": 0.026, "step": 47580 }, { "grad_norm": 0.43830186128616333, "learning_rate": 3.855093396734894e-05, "loss": 0.0281, "step": 47590 }, { "grad_norm": 0.2828051447868347, "learning_rate": 3.8530815715151545e-05, "loss": 0.033, "step": 47600 }, { "grad_norm": 0.33708828687667847, "learning_rate": 3.851069942272405e-05, "loss": 0.0227, "step": 47610 }, { "grad_norm": 0.37838104367256165, "learning_rate": 3.849058509350382e-05, "loss": 0.0231, "step": 47620 }, { "grad_norm": 0.3438190519809723, "learning_rate": 3.8470472730927783e-05, "loss": 0.0283, "step": 47630 }, { "grad_norm": 0.37696734070777893, "learning_rate": 3.845036233843264e-05, "loss": 0.0198, "step": 47640 }, { "grad_norm": 0.2532922923564911, "learning_rate": 3.843025391945469e-05, "loss": 0.0232, "step": 47650 }, { "grad_norm": 0.541826069355011, "learning_rate": 3.841014747742992e-05, "loss": 0.0213, "step": 47660 }, { "grad_norm": 0.2688407897949219, "learning_rate": 3.839004301579397e-05, "loss": 0.0184, "step": 47670 }, { "grad_norm": 0.44521236419677734, "learning_rate": 3.8369940537982097e-05, "loss": 0.0254, "step": 47680 }, { "grad_norm": 0.3483290672302246, "learning_rate": 3.834984004742933e-05, "loss": 0.0185, "step": 47690 }, { "grad_norm": 0.4504656493663788, "learning_rate": 3.832974154757026e-05, "loss": 0.0204, "step": 47700 }, { "grad_norm": 0.2940061688423157, "learning_rate": 3.830964504183919e-05, "loss": 0.0241, "step": 47710 }, { "grad_norm": 0.18993239104747772, "learning_rate": 3.828955053367003e-05, "loss": 0.0204, "step": 47720 }, { "grad_norm": 0.3792122006416321, "learning_rate": 3.826945802649642e-05, "loss": 0.0237, "step": 47730 }, { "grad_norm": 0.31774553656578064, "learning_rate": 3.824936752375159e-05, "loss": 0.0304, "step": 47740 }, { "grad_norm": 0.2878068685531616, "learning_rate": 3.822927902886848e-05, "loss": 0.021, "step": 47750 }, { "grad_norm": 0.3172931671142578, "learning_rate": 3.8209192545279653e-05, "loss": 0.0269, "step": 47760 }, { "grad_norm": 0.20194976031780243, "learning_rate": 3.8189108076417326e-05, "loss": 0.0207, "step": 47770 }, { "grad_norm": 0.2689431607723236, "learning_rate": 3.816902562571342e-05, "loss": 0.0356, "step": 47780 }, { "grad_norm": 0.335183709859848, "learning_rate": 3.814894519659944e-05, "loss": 0.0237, "step": 47790 }, { "grad_norm": 0.1862332820892334, "learning_rate": 3.812886679250661e-05, "loss": 0.0258, "step": 47800 }, { "grad_norm": 0.16246899962425232, "learning_rate": 3.810879041686575e-05, "loss": 0.0206, "step": 47810 }, { "grad_norm": 0.39164280891418457, "learning_rate": 3.808871607310741e-05, "loss": 0.0276, "step": 47820 }, { "grad_norm": 0.30069562792778015, "learning_rate": 3.806864376466169e-05, "loss": 0.0228, "step": 47830 }, { "grad_norm": 0.19546297192573547, "learning_rate": 3.804857349495845e-05, "loss": 0.0205, "step": 47840 }, { "grad_norm": 0.35244590044021606, "learning_rate": 3.802850526742713e-05, "loss": 0.0401, "step": 47850 }, { "grad_norm": 0.24646824598312378, "learning_rate": 3.800843908549683e-05, "loss": 0.0156, "step": 47860 }, { "grad_norm": 0.2991666793823242, "learning_rate": 3.7988374952596325e-05, "loss": 0.0193, "step": 47870 }, { "grad_norm": 0.3149835169315338, "learning_rate": 3.7968312872154e-05, "loss": 0.025, "step": 47880 }, { "grad_norm": 0.22927725315093994, "learning_rate": 3.7948252847597965e-05, "loss": 0.0214, "step": 47890 }, { "grad_norm": 0.3001899719238281, "learning_rate": 3.7928194882355885e-05, "loss": 0.0258, "step": 47900 }, { "grad_norm": 0.3253174126148224, "learning_rate": 3.790813897985515e-05, "loss": 0.0185, "step": 47910 }, { "grad_norm": 0.42964231967926025, "learning_rate": 3.7888085143522726e-05, "loss": 0.0278, "step": 47920 }, { "grad_norm": 0.27871087193489075, "learning_rate": 3.7868033376785314e-05, "loss": 0.0208, "step": 47930 }, { "grad_norm": 0.4172138571739197, "learning_rate": 3.784798368306919e-05, "loss": 0.0212, "step": 47940 }, { "grad_norm": 0.6460532546043396, "learning_rate": 3.782793606580029e-05, "loss": 0.0226, "step": 47950 }, { "grad_norm": 0.4397587478160858, "learning_rate": 3.7807890528404205e-05, "loss": 0.0278, "step": 47960 }, { "grad_norm": 0.1784612536430359, "learning_rate": 3.778784707430616e-05, "loss": 0.026, "step": 47970 }, { "grad_norm": 0.2722201347351074, "learning_rate": 3.776780570693107e-05, "loss": 0.0197, "step": 47980 }, { "grad_norm": 0.21755319833755493, "learning_rate": 3.774776642970342e-05, "loss": 0.0279, "step": 47990 }, { "grad_norm": 0.2455749809741974, "learning_rate": 3.77277292460474e-05, "loss": 0.0235, "step": 48000 }, { "grad_norm": 0.27859410643577576, "learning_rate": 3.770769415938678e-05, "loss": 0.0203, "step": 48010 }, { "grad_norm": 0.2615619897842407, "learning_rate": 3.768766117314506e-05, "loss": 0.0285, "step": 48020 }, { "grad_norm": 0.31238189339637756, "learning_rate": 3.76676302907453e-05, "loss": 0.0211, "step": 48030 }, { "grad_norm": 0.25894227623939514, "learning_rate": 3.764760151561021e-05, "loss": 0.0193, "step": 48040 }, { "grad_norm": 0.38014358282089233, "learning_rate": 3.76275748511622e-05, "loss": 0.021, "step": 48050 }, { "grad_norm": 0.3120250403881073, "learning_rate": 3.7607550300823255e-05, "loss": 0.0187, "step": 48060 }, { "grad_norm": 0.3153393268585205, "learning_rate": 3.7587527868015044e-05, "loss": 0.0329, "step": 48070 }, { "grad_norm": 0.314225971698761, "learning_rate": 3.756750755615881e-05, "loss": 0.0242, "step": 48080 }, { "grad_norm": 0.2732482850551605, "learning_rate": 3.754748936867553e-05, "loss": 0.0219, "step": 48090 }, { "grad_norm": 0.26865655183792114, "learning_rate": 3.752747330898573e-05, "loss": 0.021, "step": 48100 }, { "grad_norm": 0.3935214877128601, "learning_rate": 3.750745938050962e-05, "loss": 0.0224, "step": 48110 }, { "grad_norm": 0.5144045948982239, "learning_rate": 3.7487447586667025e-05, "loss": 0.0297, "step": 48120 }, { "grad_norm": 0.1675814390182495, "learning_rate": 3.7467437930877426e-05, "loss": 0.0165, "step": 48130 }, { "grad_norm": 0.2892982065677643, "learning_rate": 3.744743041655992e-05, "loss": 0.026, "step": 48140 }, { "grad_norm": 0.25553762912750244, "learning_rate": 3.742742504713324e-05, "loss": 0.0243, "step": 48150 }, { "grad_norm": 0.34816229343414307, "learning_rate": 3.740742182601576e-05, "loss": 0.0233, "step": 48160 }, { "grad_norm": 0.22328783571720123, "learning_rate": 3.7387420756625464e-05, "loss": 0.0266, "step": 48170 }, { "grad_norm": 0.2855086624622345, "learning_rate": 3.736742184238002e-05, "loss": 0.0228, "step": 48180 }, { "grad_norm": 0.23044754564762115, "learning_rate": 3.7347425086696684e-05, "loss": 0.0233, "step": 48190 }, { "grad_norm": 0.44720345735549927, "learning_rate": 3.732743049299235e-05, "loss": 0.0286, "step": 48200 }, { "grad_norm": 0.32097944617271423, "learning_rate": 3.730743806468354e-05, "loss": 0.0293, "step": 48210 }, { "grad_norm": 0.29519909620285034, "learning_rate": 3.7287447805186436e-05, "loss": 0.0209, "step": 48220 }, { "grad_norm": 0.2973046600818634, "learning_rate": 3.726745971791682e-05, "loss": 0.0245, "step": 48230 }, { "grad_norm": 0.3981448709964752, "learning_rate": 3.724747380629008e-05, "loss": 0.0372, "step": 48240 }, { "grad_norm": 0.26177453994750977, "learning_rate": 3.72274900737213e-05, "loss": 0.0266, "step": 48250 }, { "grad_norm": 0.23071405291557312, "learning_rate": 3.7207508523625123e-05, "loss": 0.019, "step": 48260 }, { "grad_norm": 0.2697070240974426, "learning_rate": 3.718752915941588e-05, "loss": 0.0202, "step": 48270 }, { "grad_norm": 0.34653350710868835, "learning_rate": 3.7167551984507464e-05, "loss": 0.0211, "step": 48280 }, { "grad_norm": 0.2828068733215332, "learning_rate": 3.714757700231346e-05, "loss": 0.0211, "step": 48290 }, { "grad_norm": 0.21979711949825287, "learning_rate": 3.712760421624703e-05, "loss": 0.019, "step": 48300 }, { "grad_norm": 0.28837552666664124, "learning_rate": 3.710763362972099e-05, "loss": 0.0239, "step": 48310 }, { "grad_norm": 0.3682553470134735, "learning_rate": 3.708766524614774e-05, "loss": 0.0247, "step": 48320 }, { "grad_norm": 0.33427250385284424, "learning_rate": 3.7067699068939335e-05, "loss": 0.0213, "step": 48330 }, { "grad_norm": 0.21575316786766052, "learning_rate": 3.704773510150748e-05, "loss": 0.0233, "step": 48340 }, { "grad_norm": 0.38682350516319275, "learning_rate": 3.702777334726344e-05, "loss": 0.0253, "step": 48350 }, { "grad_norm": 0.2143716663122177, "learning_rate": 3.7007813809618164e-05, "loss": 0.0168, "step": 48360 }, { "grad_norm": 0.3942444622516632, "learning_rate": 3.6987856491982145e-05, "loss": 0.0268, "step": 48370 }, { "grad_norm": 0.21689894795417786, "learning_rate": 3.696790139776558e-05, "loss": 0.0253, "step": 48380 }, { "grad_norm": 0.4107270836830139, "learning_rate": 3.6947948530378235e-05, "loss": 0.0263, "step": 48390 }, { "grad_norm": 0.2597987651824951, "learning_rate": 3.6927997893229516e-05, "loss": 0.0192, "step": 48400 }, { "grad_norm": 0.21038828790187836, "learning_rate": 3.690804948972845e-05, "loss": 0.0157, "step": 48410 }, { "grad_norm": 0.23222452402114868, "learning_rate": 3.688810332328363e-05, "loss": 0.0236, "step": 48420 }, { "grad_norm": 0.3931509256362915, "learning_rate": 3.686815939730336e-05, "loss": 0.023, "step": 48430 }, { "grad_norm": 0.27755385637283325, "learning_rate": 3.684821771519548e-05, "loss": 0.0211, "step": 48440 }, { "grad_norm": 0.298615038394928, "learning_rate": 3.68282782803675e-05, "loss": 0.0251, "step": 48450 }, { "grad_norm": 0.2004171460866928, "learning_rate": 3.6808341096226504e-05, "loss": 0.03, "step": 48460 }, { "grad_norm": 0.4295656085014343, "learning_rate": 3.678840616617924e-05, "loss": 0.0266, "step": 48470 }, { "grad_norm": 0.3401945233345032, "learning_rate": 3.6768473493632e-05, "loss": 0.0238, "step": 48480 }, { "grad_norm": 0.3452712297439575, "learning_rate": 3.6748543081990783e-05, "loss": 0.0247, "step": 48490 }, { "grad_norm": 0.2758553624153137, "learning_rate": 3.672861493466112e-05, "loss": 0.0188, "step": 48500 }, { "grad_norm": 0.1919887661933899, "learning_rate": 3.670868905504818e-05, "loss": 0.0185, "step": 48510 }, { "grad_norm": 0.7245864272117615, "learning_rate": 3.6688765446556784e-05, "loss": 0.0203, "step": 48520 }, { "grad_norm": 0.42061230540275574, "learning_rate": 3.6668844112591276e-05, "loss": 0.0219, "step": 48530 }, { "grad_norm": 0.4113631546497345, "learning_rate": 3.664892505655573e-05, "loss": 0.0268, "step": 48540 }, { "grad_norm": 0.4729655981063843, "learning_rate": 3.662900828185373e-05, "loss": 0.0286, "step": 48550 }, { "grad_norm": 0.2559208869934082, "learning_rate": 3.6609093791888516e-05, "loss": 0.0247, "step": 48560 }, { "grad_norm": 0.3451659381389618, "learning_rate": 3.658918159006292e-05, "loss": 0.0308, "step": 48570 }, { "grad_norm": 0.5468297004699707, "learning_rate": 3.656927167977942e-05, "loss": 0.0286, "step": 48580 }, { "grad_norm": 0.2751002013683319, "learning_rate": 3.654936406444006e-05, "loss": 0.0231, "step": 48590 }, { "grad_norm": 0.28332415223121643, "learning_rate": 3.65294587474465e-05, "loss": 0.019, "step": 48600 }, { "grad_norm": 0.461311012506485, "learning_rate": 3.650955573220002e-05, "loss": 0.0529, "step": 48610 }, { "grad_norm": 0.3727886974811554, "learning_rate": 3.648965502210149e-05, "loss": 0.019, "step": 48620 }, { "grad_norm": 0.31957584619522095, "learning_rate": 3.646975662055142e-05, "loss": 0.0243, "step": 48630 }, { "grad_norm": 0.2588523328304291, "learning_rate": 3.644986053094987e-05, "loss": 0.0232, "step": 48640 }, { "grad_norm": 0.3733220398426056, "learning_rate": 3.642996675669659e-05, "loss": 0.0224, "step": 48650 }, { "grad_norm": 0.23100575804710388, "learning_rate": 3.641007530119083e-05, "loss": 0.0165, "step": 48660 }, { "grad_norm": 0.2533051669597626, "learning_rate": 3.639018616783153e-05, "loss": 0.0231, "step": 48670 }, { "grad_norm": 0.33196863532066345, "learning_rate": 3.637029936001719e-05, "loss": 0.0189, "step": 48680 }, { "grad_norm": 0.24872000515460968, "learning_rate": 3.6350414881145886e-05, "loss": 0.0266, "step": 48690 }, { "grad_norm": 0.18505848944187164, "learning_rate": 3.6330532734615386e-05, "loss": 0.0253, "step": 48700 }, { "grad_norm": 0.26115426421165466, "learning_rate": 3.6310652923822975e-05, "loss": 0.0249, "step": 48710 }, { "grad_norm": 0.29265934228897095, "learning_rate": 3.629077545216558e-05, "loss": 0.0247, "step": 48720 }, { "grad_norm": 0.41174623370170593, "learning_rate": 3.627090032303969e-05, "loss": 0.0274, "step": 48730 }, { "grad_norm": 0.28996044397354126, "learning_rate": 3.625102753984146e-05, "loss": 0.0228, "step": 48740 }, { "grad_norm": 0.285264253616333, "learning_rate": 3.623115710596659e-05, "loss": 0.0182, "step": 48750 }, { "grad_norm": 0.503265917301178, "learning_rate": 3.6211289024810395e-05, "loss": 0.0303, "step": 48760 }, { "grad_norm": 0.21242383122444153, "learning_rate": 3.619142329976777e-05, "loss": 0.0216, "step": 48770 }, { "grad_norm": 0.3276682198047638, "learning_rate": 3.6171559934233247e-05, "loss": 0.0233, "step": 48780 }, { "grad_norm": 0.2828875780105591, "learning_rate": 3.615169893160093e-05, "loss": 0.0166, "step": 48790 }, { "grad_norm": 0.2693047523498535, "learning_rate": 3.61318402952645e-05, "loss": 0.0199, "step": 48800 }, { "grad_norm": 0.4370678961277008, "learning_rate": 3.6111984028617285e-05, "loss": 0.0233, "step": 48810 }, { "grad_norm": 0.8308297395706177, "learning_rate": 3.6092130135052134e-05, "loss": 0.0234, "step": 48820 }, { "grad_norm": 0.2792102098464966, "learning_rate": 3.6072278617961584e-05, "loss": 0.0305, "step": 48830 }, { "grad_norm": 0.28404998779296875, "learning_rate": 3.605242948073767e-05, "loss": 0.0291, "step": 48840 }, { "grad_norm": 0.29468226432800293, "learning_rate": 3.603258272677212e-05, "loss": 0.0224, "step": 48850 }, { "grad_norm": 0.2171405404806137, "learning_rate": 3.601273835945616e-05, "loss": 0.0224, "step": 48860 }, { "grad_norm": 0.2801445722579956, "learning_rate": 3.5992896382180664e-05, "loss": 0.0233, "step": 48870 }, { "grad_norm": 0.3271803557872772, "learning_rate": 3.597305679833609e-05, "loss": 0.0277, "step": 48880 }, { "grad_norm": 0.26697880029678345, "learning_rate": 3.595321961131245e-05, "loss": 0.0221, "step": 48890 }, { "grad_norm": 0.3023359477519989, "learning_rate": 3.593338482449942e-05, "loss": 0.0248, "step": 48900 }, { "grad_norm": 0.3127893805503845, "learning_rate": 3.591355244128618e-05, "loss": 0.0326, "step": 48910 }, { "grad_norm": 0.26596951484680176, "learning_rate": 3.589372246506158e-05, "loss": 0.0276, "step": 48920 }, { "grad_norm": 0.33089733123779297, "learning_rate": 3.5873894899213984e-05, "loss": 0.0181, "step": 48930 }, { "grad_norm": 0.3023843765258789, "learning_rate": 3.5854069747131416e-05, "loss": 0.0175, "step": 48940 }, { "grad_norm": 0.3200451731681824, "learning_rate": 3.583424701220143e-05, "loss": 0.027, "step": 48950 }, { "grad_norm": 0.317472368478775, "learning_rate": 3.581442669781121e-05, "loss": 0.0248, "step": 48960 }, { "grad_norm": 0.2425830215215683, "learning_rate": 3.579460880734749e-05, "loss": 0.0236, "step": 48970 }, { "grad_norm": 0.3672862648963928, "learning_rate": 3.577479334419657e-05, "loss": 0.0247, "step": 48980 }, { "grad_norm": 0.44924411177635193, "learning_rate": 3.575498031174444e-05, "loss": 0.0236, "step": 48990 }, { "grad_norm": 0.33463534712791443, "learning_rate": 3.573516971337657e-05, "loss": 0.0256, "step": 49000 }, { "grad_norm": 0.4587945342063904, "learning_rate": 3.5715361552478046e-05, "loss": 0.0288, "step": 49010 }, { "grad_norm": 0.23790185153484344, "learning_rate": 3.5695555832433536e-05, "loss": 0.0213, "step": 49020 }, { "grad_norm": 0.35377800464630127, "learning_rate": 3.5675752556627325e-05, "loss": 0.0235, "step": 49030 }, { "grad_norm": 0.3007887303829193, "learning_rate": 3.565595172844322e-05, "loss": 0.0223, "step": 49040 }, { "grad_norm": 0.39484265446662903, "learning_rate": 3.5636153351264666e-05, "loss": 0.0281, "step": 49050 }, { "grad_norm": 0.2890808582305908, "learning_rate": 3.5616357428474655e-05, "loss": 0.0256, "step": 49060 }, { "grad_norm": 0.3281765878200531, "learning_rate": 3.559656396345575e-05, "loss": 0.0261, "step": 49070 }, { "grad_norm": 0.3379916548728943, "learning_rate": 3.5576772959590146e-05, "loss": 0.0267, "step": 49080 }, { "grad_norm": 0.1798686534166336, "learning_rate": 3.5556984420259545e-05, "loss": 0.0225, "step": 49090 }, { "grad_norm": 0.2357570379972458, "learning_rate": 3.5537198348845305e-05, "loss": 0.026, "step": 49100 }, { "grad_norm": 0.3100479543209076, "learning_rate": 3.551741474872831e-05, "loss": 0.0261, "step": 49110 }, { "grad_norm": 0.5338011980056763, "learning_rate": 3.549763362328903e-05, "loss": 0.0326, "step": 49120 }, { "grad_norm": 0.28045350313186646, "learning_rate": 3.5477854975907515e-05, "loss": 0.021, "step": 49130 }, { "grad_norm": 0.2633286118507385, "learning_rate": 3.5458078809963416e-05, "loss": 0.0194, "step": 49140 }, { "grad_norm": 0.23652350902557373, "learning_rate": 3.543830512883594e-05, "loss": 0.0234, "step": 49150 }, { "grad_norm": 0.2969265282154083, "learning_rate": 3.5418533935903824e-05, "loss": 0.0267, "step": 49160 }, { "grad_norm": 0.35061919689178467, "learning_rate": 3.539876523454547e-05, "loss": 0.0243, "step": 49170 }, { "grad_norm": 0.2916000187397003, "learning_rate": 3.537899902813878e-05, "loss": 0.0234, "step": 49180 }, { "grad_norm": 0.2581939399242401, "learning_rate": 3.5359235320061293e-05, "loss": 0.0201, "step": 49190 }, { "grad_norm": 0.2919550836086273, "learning_rate": 3.533947411369003e-05, "loss": 0.0212, "step": 49200 }, { "grad_norm": 0.3632827401161194, "learning_rate": 3.53197154124017e-05, "loss": 0.021, "step": 49210 }, { "grad_norm": 0.4243754744529724, "learning_rate": 3.52999592195725e-05, "loss": 0.0238, "step": 49220 }, { "grad_norm": 0.30379652976989746, "learning_rate": 3.5280205538578224e-05, "loss": 0.0201, "step": 49230 }, { "grad_norm": 0.2976398169994354, "learning_rate": 3.5260454372794236e-05, "loss": 0.021, "step": 49240 }, { "grad_norm": 0.301293283700943, "learning_rate": 3.524070572559545e-05, "loss": 0.0244, "step": 49250 }, { "grad_norm": 0.28511080145835876, "learning_rate": 3.5220959600356395e-05, "loss": 0.0207, "step": 49260 }, { "grad_norm": 0.1880822628736496, "learning_rate": 3.5201216000451145e-05, "loss": 0.0234, "step": 49270 }, { "grad_norm": 0.7075817584991455, "learning_rate": 3.5181474929253335e-05, "loss": 0.0176, "step": 49280 }, { "grad_norm": 0.24155518412590027, "learning_rate": 3.516173639013615e-05, "loss": 0.0235, "step": 49290 }, { "grad_norm": 0.20124810934066772, "learning_rate": 3.5142000386472406e-05, "loss": 0.0209, "step": 49300 }, { "grad_norm": 0.227478489279747, "learning_rate": 3.5122266921634427e-05, "loss": 0.0192, "step": 49310 }, { "grad_norm": 0.3685767650604248, "learning_rate": 3.510253599899413e-05, "loss": 0.0244, "step": 49320 }, { "grad_norm": 0.2558806836605072, "learning_rate": 3.5082807621922965e-05, "loss": 0.0204, "step": 49330 }, { "grad_norm": 0.2822859287261963, "learning_rate": 3.506308179379201e-05, "loss": 0.0244, "step": 49340 }, { "grad_norm": 0.3876174986362457, "learning_rate": 3.5043358517971844e-05, "loss": 0.0232, "step": 49350 }, { "grad_norm": 0.21640072762966156, "learning_rate": 3.502363779783264e-05, "loss": 0.023, "step": 49360 }, { "grad_norm": 0.2952844798564911, "learning_rate": 3.500391963674415e-05, "loss": 0.0242, "step": 49370 }, { "grad_norm": 0.26105326414108276, "learning_rate": 3.4984204038075615e-05, "loss": 0.0305, "step": 49380 }, { "grad_norm": 0.2644352912902832, "learning_rate": 3.496449100519595e-05, "loss": 0.0245, "step": 49390 }, { "grad_norm": 0.29980629682540894, "learning_rate": 3.494478054147354e-05, "loss": 0.02, "step": 49400 }, { "grad_norm": 0.455386757850647, "learning_rate": 3.4925072650276395e-05, "loss": 0.0207, "step": 49410 }, { "grad_norm": 0.3584372103214264, "learning_rate": 3.4905367334972016e-05, "loss": 0.0237, "step": 49420 }, { "grad_norm": 0.2699681520462036, "learning_rate": 3.488566459892752e-05, "loss": 0.0241, "step": 49430 }, { "grad_norm": 0.7048608064651489, "learning_rate": 3.4865964445509585e-05, "loss": 0.0246, "step": 49440 }, { "grad_norm": 0.25741252303123474, "learning_rate": 3.484626687808438e-05, "loss": 0.02, "step": 49450 }, { "grad_norm": 0.292211651802063, "learning_rate": 3.4826571900017735e-05, "loss": 0.016, "step": 49460 }, { "grad_norm": 0.43495434522628784, "learning_rate": 3.480687951467495e-05, "loss": 0.0224, "step": 49470 }, { "grad_norm": 0.2592117488384247, "learning_rate": 3.4787189725420925e-05, "loss": 0.0193, "step": 49480 }, { "grad_norm": 0.2244434356689453, "learning_rate": 3.4767502535620086e-05, "loss": 0.019, "step": 49490 }, { "grad_norm": 0.28624773025512695, "learning_rate": 3.474781794863648e-05, "loss": 0.0282, "step": 49500 }, { "grad_norm": 0.3259296417236328, "learning_rate": 3.472813596783363e-05, "loss": 0.0305, "step": 49510 }, { "grad_norm": 1.8010181188583374, "learning_rate": 3.470845659657466e-05, "loss": 0.0257, "step": 49520 }, { "grad_norm": 0.24403667449951172, "learning_rate": 3.468877983822223e-05, "loss": 0.0217, "step": 49530 }, { "grad_norm": 0.2601079046726227, "learning_rate": 3.466910569613855e-05, "loss": 0.0217, "step": 49540 }, { "grad_norm": 0.3190058469772339, "learning_rate": 3.464943417368542e-05, "loss": 0.0247, "step": 49550 }, { "grad_norm": 0.254667729139328, "learning_rate": 3.462976527422415e-05, "loss": 0.02, "step": 49560 }, { "grad_norm": 0.32247158885002136, "learning_rate": 3.461009900111562e-05, "loss": 0.0243, "step": 49570 }, { "grad_norm": 0.38060757517814636, "learning_rate": 3.459043535772023e-05, "loss": 0.0226, "step": 49580 }, { "grad_norm": 0.2947445511817932, "learning_rate": 3.4570774347398014e-05, "loss": 0.0246, "step": 49590 }, { "grad_norm": 0.39812058210372925, "learning_rate": 3.4551115973508454e-05, "loss": 0.0282, "step": 49600 }, { "grad_norm": 0.5366798043251038, "learning_rate": 3.453146023941066e-05, "loss": 0.031, "step": 49610 }, { "grad_norm": 0.27987995743751526, "learning_rate": 3.451180714846325e-05, "loss": 0.0206, "step": 49620 }, { "grad_norm": 0.7046130299568176, "learning_rate": 3.449215670402438e-05, "loss": 0.0229, "step": 49630 }, { "grad_norm": 0.4225248396396637, "learning_rate": 3.447250890945181e-05, "loss": 0.0218, "step": 49640 }, { "grad_norm": 0.3611215651035309, "learning_rate": 3.4452863768102754e-05, "loss": 0.0229, "step": 49650 }, { "grad_norm": 0.25762447714805603, "learning_rate": 3.443322128333409e-05, "loss": 0.02, "step": 49660 }, { "grad_norm": 0.44944486021995544, "learning_rate": 3.441358145850215e-05, "loss": 0.0248, "step": 49670 }, { "grad_norm": 0.4086132347583771, "learning_rate": 3.439394429696286e-05, "loss": 0.0255, "step": 49680 }, { "grad_norm": 0.2653915584087372, "learning_rate": 3.4374309802071644e-05, "loss": 0.0221, "step": 49690 }, { "grad_norm": 0.577492892742157, "learning_rate": 3.435467797718353e-05, "loss": 0.0209, "step": 49700 }, { "grad_norm": 0.2913721799850464, "learning_rate": 3.433504882565306e-05, "loss": 0.0173, "step": 49710 }, { "grad_norm": 0.26853570342063904, "learning_rate": 3.43154223508343e-05, "loss": 0.0253, "step": 49720 }, { "grad_norm": 0.23148202896118164, "learning_rate": 3.429579855608089e-05, "loss": 0.0227, "step": 49730 }, { "grad_norm": 0.2844250798225403, "learning_rate": 3.427617744474597e-05, "loss": 0.019, "step": 49740 }, { "grad_norm": 0.2745659053325653, "learning_rate": 3.425655902018231e-05, "loss": 0.0239, "step": 49750 }, { "grad_norm": 0.25585392117500305, "learning_rate": 3.423694328574211e-05, "loss": 0.0219, "step": 49760 }, { "grad_norm": 0.27833127975463867, "learning_rate": 3.42173302447772e-05, "loss": 0.0224, "step": 49770 }, { "grad_norm": 0.2972194254398346, "learning_rate": 3.419771990063886e-05, "loss": 0.017, "step": 49780 }, { "grad_norm": 0.40866878628730774, "learning_rate": 3.417811225667803e-05, "loss": 0.0155, "step": 49790 }, { "grad_norm": 0.3448387682437897, "learning_rate": 3.415850731624508e-05, "loss": 0.0225, "step": 49800 }, { "grad_norm": 0.21690665185451508, "learning_rate": 3.4138905082689945e-05, "loss": 0.025, "step": 49810 }, { "grad_norm": 0.46114569902420044, "learning_rate": 3.4119305559362145e-05, "loss": 0.0234, "step": 49820 }, { "grad_norm": 0.22747360169887543, "learning_rate": 3.4099708749610684e-05, "loss": 0.0173, "step": 49830 }, { "grad_norm": 0.35244688391685486, "learning_rate": 3.408011465678413e-05, "loss": 0.0339, "step": 49840 }, { "grad_norm": 0.26139360666275024, "learning_rate": 3.406052328423055e-05, "loss": 0.0215, "step": 49850 }, { "grad_norm": 0.2590314447879791, "learning_rate": 3.4040934635297615e-05, "loss": 0.0197, "step": 49860 }, { "grad_norm": 0.33779454231262207, "learning_rate": 3.4021348713332466e-05, "loss": 0.0217, "step": 49870 }, { "grad_norm": 0.3173362910747528, "learning_rate": 3.4001765521681807e-05, "loss": 0.022, "step": 49880 }, { "grad_norm": 0.23263321816921234, "learning_rate": 3.398218506369188e-05, "loss": 0.0295, "step": 49890 }, { "grad_norm": 0.2118072658777237, "learning_rate": 3.3962607342708404e-05, "loss": 0.0193, "step": 49900 }, { "grad_norm": 0.3468252420425415, "learning_rate": 3.394303236207673e-05, "loss": 0.0213, "step": 49910 }, { "grad_norm": 0.27509739995002747, "learning_rate": 3.392346012514166e-05, "loss": 0.0206, "step": 49920 }, { "grad_norm": 0.28037917613983154, "learning_rate": 3.390389063524757e-05, "loss": 0.02, "step": 49930 }, { "grad_norm": 0.318632572889328, "learning_rate": 3.3884323895738324e-05, "loss": 0.0179, "step": 49940 }, { "grad_norm": 0.4387553334236145, "learning_rate": 3.386475990995738e-05, "loss": 0.024, "step": 49950 }, { "grad_norm": 0.32378652691841125, "learning_rate": 3.384519868124765e-05, "loss": 0.0179, "step": 49960 }, { "grad_norm": 0.21704666316509247, "learning_rate": 3.3825640212951645e-05, "loss": 0.0219, "step": 49970 }, { "grad_norm": 0.6900607943534851, "learning_rate": 3.380608450841134e-05, "loss": 0.025, "step": 49980 }, { "grad_norm": 0.3408602178096771, "learning_rate": 3.3786531570968305e-05, "loss": 0.0202, "step": 49990 }, { "grad_norm": 0.3393857181072235, "learning_rate": 3.3766981403963584e-05, "loss": 0.0264, "step": 50000 }, { "grad_norm": 0.41574084758758545, "learning_rate": 3.374743401073775e-05, "loss": 0.0251, "step": 50010 }, { "grad_norm": 0.31421807408332825, "learning_rate": 3.372788939463095e-05, "loss": 0.0191, "step": 50020 }, { "grad_norm": 0.2674720585346222, "learning_rate": 3.370834755898281e-05, "loss": 0.0164, "step": 50030 }, { "grad_norm": 0.3026706278324127, "learning_rate": 3.3688808507132493e-05, "loss": 0.0269, "step": 50040 }, { "grad_norm": 0.3198016583919525, "learning_rate": 3.3669272242418685e-05, "loss": 0.0245, "step": 50050 }, { "grad_norm": 0.2516564130783081, "learning_rate": 3.364973876817961e-05, "loss": 0.0232, "step": 50060 }, { "grad_norm": 0.3159831166267395, "learning_rate": 3.363020808775299e-05, "loss": 0.0206, "step": 50070 }, { "grad_norm": 0.2765820026397705, "learning_rate": 3.361068020447611e-05, "loss": 0.0222, "step": 50080 }, { "grad_norm": 0.24243779480457306, "learning_rate": 3.3591155121685724e-05, "loss": 0.0182, "step": 50090 }, { "grad_norm": 0.5163854360580444, "learning_rate": 3.357163284271814e-05, "loss": 0.0253, "step": 50100 }, { "grad_norm": 0.3859004080295563, "learning_rate": 3.355211337090919e-05, "loss": 0.0175, "step": 50110 }, { "grad_norm": 0.2841583490371704, "learning_rate": 3.353259670959421e-05, "loss": 0.0281, "step": 50120 }, { "grad_norm": 0.24211002886295319, "learning_rate": 3.351308286210808e-05, "loss": 0.0244, "step": 50130 }, { "grad_norm": 0.23450319468975067, "learning_rate": 3.3493571831785156e-05, "loss": 0.0232, "step": 50140 }, { "grad_norm": 0.2717736065387726, "learning_rate": 3.347406362195936e-05, "loss": 0.018, "step": 50150 }, { "grad_norm": 0.48325997591018677, "learning_rate": 3.345455823596411e-05, "loss": 0.0276, "step": 50160 }, { "grad_norm": 0.3747532069683075, "learning_rate": 3.3435055677132346e-05, "loss": 0.0252, "step": 50170 }, { "grad_norm": 0.3584616184234619, "learning_rate": 3.3415555948796505e-05, "loss": 0.0246, "step": 50180 }, { "grad_norm": 0.30887991189956665, "learning_rate": 3.3396059054288556e-05, "loss": 0.0221, "step": 50190 }, { "grad_norm": 0.31344619393348694, "learning_rate": 3.3376564996940015e-05, "loss": 0.0347, "step": 50200 }, { "grad_norm": 0.3344951272010803, "learning_rate": 3.3357073780081836e-05, "loss": 0.0247, "step": 50210 }, { "grad_norm": 0.24827562272548676, "learning_rate": 3.333758540704459e-05, "loss": 0.0204, "step": 50220 }, { "grad_norm": 0.35012489557266235, "learning_rate": 3.331809988115827e-05, "loss": 0.0194, "step": 50230 }, { "grad_norm": 0.3580482006072998, "learning_rate": 3.329861720575244e-05, "loss": 0.0201, "step": 50240 }, { "grad_norm": 0.2091917246580124, "learning_rate": 3.3279137384156126e-05, "loss": 0.0198, "step": 50250 }, { "grad_norm": 0.284660279750824, "learning_rate": 3.3259660419697934e-05, "loss": 0.0288, "step": 50260 }, { "grad_norm": 0.15769243240356445, "learning_rate": 3.3240186315705926e-05, "loss": 0.0261, "step": 50270 }, { "grad_norm": 0.3125869035720825, "learning_rate": 3.322071507550769e-05, "loss": 0.0276, "step": 50280 }, { "grad_norm": 0.340805321931839, "learning_rate": 3.320124670243033e-05, "loss": 0.0221, "step": 50290 }, { "grad_norm": 0.5278860926628113, "learning_rate": 3.318178119980045e-05, "loss": 0.0229, "step": 50300 }, { "grad_norm": 0.38234028220176697, "learning_rate": 3.316231857094421e-05, "loss": 0.0248, "step": 50310 }, { "grad_norm": 0.33424821496009827, "learning_rate": 3.31428588191872e-05, "loss": 0.0289, "step": 50320 }, { "grad_norm": 0.2954716682434082, "learning_rate": 3.312340194785458e-05, "loss": 0.0221, "step": 50330 }, { "grad_norm": 0.28242459893226624, "learning_rate": 3.3103947960270975e-05, "loss": 0.016, "step": 50340 }, { "grad_norm": 0.3041694462299347, "learning_rate": 3.308449685976058e-05, "loss": 0.0195, "step": 50350 }, { "grad_norm": 0.2891097962856293, "learning_rate": 3.3065048649647024e-05, "loss": 0.0238, "step": 50360 }, { "grad_norm": 0.3477853536605835, "learning_rate": 3.304560333325348e-05, "loss": 0.0231, "step": 50370 }, { "grad_norm": 0.3714648485183716, "learning_rate": 3.3026160913902624e-05, "loss": 0.0212, "step": 50380 }, { "grad_norm": 0.42098256945610046, "learning_rate": 3.300672139491662e-05, "loss": 0.0232, "step": 50390 }, { "grad_norm": 0.27910178899765015, "learning_rate": 3.298728477961717e-05, "loss": 0.027, "step": 50400 }, { "grad_norm": 0.37677517533302307, "learning_rate": 3.296785107132544e-05, "loss": 0.0223, "step": 50410 }, { "grad_norm": 0.35482555627822876, "learning_rate": 3.2948420273362156e-05, "loss": 0.0173, "step": 50420 }, { "grad_norm": 0.19759057462215424, "learning_rate": 3.292899238904747e-05, "loss": 0.0185, "step": 50430 }, { "grad_norm": 0.3710730969905853, "learning_rate": 3.29095674217011e-05, "loss": 0.0218, "step": 50440 }, { "grad_norm": 0.2461991012096405, "learning_rate": 3.289014537464224e-05, "loss": 0.0253, "step": 50450 }, { "grad_norm": 0.36194878816604614, "learning_rate": 3.287072625118955e-05, "loss": 0.0198, "step": 50460 }, { "grad_norm": 0.28133946657180786, "learning_rate": 3.285131005466129e-05, "loss": 0.0205, "step": 50470 }, { "grad_norm": 0.279837429523468, "learning_rate": 3.2831896788375105e-05, "loss": 0.0187, "step": 50480 }, { "grad_norm": 0.432104229927063, "learning_rate": 3.281248645564822e-05, "loss": 0.0164, "step": 50490 }, { "grad_norm": 0.17612743377685547, "learning_rate": 3.2793079059797306e-05, "loss": 0.0277, "step": 50500 }, { "grad_norm": 0.2361811250448227, "learning_rate": 3.277367460413859e-05, "loss": 0.0201, "step": 50510 }, { "grad_norm": 0.2721240520477295, "learning_rate": 3.275427309198773e-05, "loss": 0.0217, "step": 50520 }, { "grad_norm": 0.2470233142375946, "learning_rate": 3.273487452665993e-05, "loss": 0.0177, "step": 50530 }, { "grad_norm": 0.229965478181839, "learning_rate": 3.271547891146986e-05, "loss": 0.0191, "step": 50540 }, { "grad_norm": 0.3636346757411957, "learning_rate": 3.269608624973173e-05, "loss": 0.0178, "step": 50550 }, { "grad_norm": 0.34993457794189453, "learning_rate": 3.26766965447592e-05, "loss": 0.0217, "step": 50560 }, { "grad_norm": 0.3566839396953583, "learning_rate": 3.2657309799865424e-05, "loss": 0.0229, "step": 50570 }, { "grad_norm": 0.22412246465682983, "learning_rate": 3.2637926018363084e-05, "loss": 0.0178, "step": 50580 }, { "grad_norm": 0.3881116211414337, "learning_rate": 3.261854520356432e-05, "loss": 0.0184, "step": 50590 }, { "grad_norm": 0.5964478254318237, "learning_rate": 3.2599167358780816e-05, "loss": 0.0253, "step": 50600 }, { "grad_norm": 0.33314234018325806, "learning_rate": 3.2579792487323676e-05, "loss": 0.0161, "step": 50610 }, { "grad_norm": 0.2531227171421051, "learning_rate": 3.256042059250358e-05, "loss": 0.0274, "step": 50620 }, { "grad_norm": 0.424332857131958, "learning_rate": 3.254105167763062e-05, "loss": 0.0238, "step": 50630 }, { "grad_norm": 0.346358984708786, "learning_rate": 3.252168574601443e-05, "loss": 0.0234, "step": 50640 }, { "grad_norm": 0.37419164180755615, "learning_rate": 3.2502322800964115e-05, "loss": 0.0223, "step": 50650 }, { "grad_norm": 0.2659653425216675, "learning_rate": 3.248296284578825e-05, "loss": 0.0206, "step": 50660 }, { "grad_norm": 0.20441973209381104, "learning_rate": 3.246360588379497e-05, "loss": 0.0278, "step": 50670 }, { "grad_norm": 0.21358738839626312, "learning_rate": 3.24442519182918e-05, "loss": 0.03, "step": 50680 }, { "grad_norm": 0.21937501430511475, "learning_rate": 3.2424900952585845e-05, "loss": 0.0183, "step": 50690 }, { "grad_norm": 0.2452497035264969, "learning_rate": 3.240555298998362e-05, "loss": 0.0224, "step": 50700 }, { "grad_norm": 0.2732005715370178, "learning_rate": 3.2386208033791194e-05, "loss": 0.021, "step": 50710 }, { "grad_norm": 0.3167650103569031, "learning_rate": 3.2366866087314064e-05, "loss": 0.023, "step": 50720 }, { "grad_norm": 0.25638145208358765, "learning_rate": 3.234752715385727e-05, "loss": 0.0255, "step": 50730 }, { "grad_norm": 0.3179123103618622, "learning_rate": 3.232819123672529e-05, "loss": 0.0292, "step": 50740 }, { "grad_norm": 0.7144988179206848, "learning_rate": 3.2308858339222076e-05, "loss": 0.0304, "step": 50750 }, { "grad_norm": 0.4205479323863983, "learning_rate": 3.2289528464651144e-05, "loss": 0.0264, "step": 50760 }, { "grad_norm": 0.3071274757385254, "learning_rate": 3.22702016163154e-05, "loss": 0.019, "step": 50770 }, { "grad_norm": 0.40659719705581665, "learning_rate": 3.2250877797517295e-05, "loss": 0.0265, "step": 50780 }, { "grad_norm": 0.8812669515609741, "learning_rate": 3.223155701155872e-05, "loss": 0.0227, "step": 50790 }, { "grad_norm": 0.28131812810897827, "learning_rate": 3.2212239261741105e-05, "loss": 0.0239, "step": 50800 }, { "grad_norm": 0.21234345436096191, "learning_rate": 3.219292455136528e-05, "loss": 0.0191, "step": 50810 }, { "grad_norm": 0.2362825870513916, "learning_rate": 3.217361288373165e-05, "loss": 0.0274, "step": 50820 }, { "grad_norm": 0.2718563973903656, "learning_rate": 3.215430426214002e-05, "loss": 0.0188, "step": 50830 }, { "grad_norm": 0.3023160696029663, "learning_rate": 3.2134998689889705e-05, "loss": 0.0217, "step": 50840 }, { "grad_norm": 0.39810606837272644, "learning_rate": 3.211569617027952e-05, "loss": 0.0199, "step": 50850 }, { "grad_norm": 0.37610527873039246, "learning_rate": 3.209639670660769e-05, "loss": 0.0259, "step": 50860 }, { "grad_norm": 0.5506680011749268, "learning_rate": 3.207710030217202e-05, "loss": 0.0176, "step": 50870 }, { "grad_norm": 0.43705087900161743, "learning_rate": 3.205780696026972e-05, "loss": 0.019, "step": 50880 }, { "grad_norm": 0.2236938625574112, "learning_rate": 3.203851668419749e-05, "loss": 0.0196, "step": 50890 }, { "grad_norm": 0.3372909426689148, "learning_rate": 3.201922947725149e-05, "loss": 0.0209, "step": 50900 }, { "grad_norm": 0.31367427110671997, "learning_rate": 3.199994534272742e-05, "loss": 0.0207, "step": 50910 }, { "grad_norm": 0.26642903685569763, "learning_rate": 3.1980664283920394e-05, "loss": 0.0237, "step": 50920 }, { "grad_norm": 0.2788082957267761, "learning_rate": 3.196138630412499e-05, "loss": 0.0193, "step": 50930 }, { "grad_norm": 0.3180633783340454, "learning_rate": 3.194211140663532e-05, "loss": 0.0241, "step": 50940 }, { "grad_norm": 0.28462356328964233, "learning_rate": 3.1922839594744914e-05, "loss": 0.0264, "step": 50950 }, { "grad_norm": 0.31794461607933044, "learning_rate": 3.190357087174683e-05, "loss": 0.0256, "step": 50960 }, { "grad_norm": 0.15922993421554565, "learning_rate": 3.188430524093353e-05, "loss": 0.0191, "step": 50970 }, { "grad_norm": 0.2922731935977936, "learning_rate": 3.1865042705597014e-05, "loss": 0.0198, "step": 50980 }, { "grad_norm": 0.30332860350608826, "learning_rate": 3.184578326902868e-05, "loss": 0.0215, "step": 50990 }, { "grad_norm": 1.3175636529922485, "learning_rate": 3.182652693451949e-05, "loss": 0.0211, "step": 51000 }, { "grad_norm": 0.2765343487262726, "learning_rate": 3.180727370535981e-05, "loss": 0.0227, "step": 51010 }, { "grad_norm": 0.1963014006614685, "learning_rate": 3.178802358483945e-05, "loss": 0.0187, "step": 51020 }, { "grad_norm": 0.3510403037071228, "learning_rate": 3.17687765762478e-05, "loss": 0.022, "step": 51030 }, { "grad_norm": 0.2090887874364853, "learning_rate": 3.174953268287357e-05, "loss": 0.0237, "step": 51040 }, { "grad_norm": 0.4169033467769623, "learning_rate": 3.173029190800508e-05, "loss": 0.0195, "step": 51050 }, { "grad_norm": 0.3792291581630707, "learning_rate": 3.171105425493e-05, "loss": 0.0245, "step": 51060 }, { "grad_norm": 0.25420570373535156, "learning_rate": 3.169181972693557e-05, "loss": 0.0139, "step": 51070 }, { "grad_norm": 0.2902293801307678, "learning_rate": 3.1672588327308404e-05, "loss": 0.0165, "step": 51080 }, { "grad_norm": 0.34720396995544434, "learning_rate": 3.1653360059334646e-05, "loss": 0.0225, "step": 51090 }, { "grad_norm": 0.46508610248565674, "learning_rate": 3.163413492629985e-05, "loss": 0.0374, "step": 51100 }, { "grad_norm": 0.379316508769989, "learning_rate": 3.161491293148912e-05, "loss": 0.0183, "step": 51110 }, { "grad_norm": 0.26753607392311096, "learning_rate": 3.1595694078186925e-05, "loss": 0.0206, "step": 51120 }, { "grad_norm": 0.3312463164329529, "learning_rate": 3.157647836967724e-05, "loss": 0.0246, "step": 51130 }, { "grad_norm": 0.35985007882118225, "learning_rate": 3.155726580924354e-05, "loss": 0.0227, "step": 51140 }, { "grad_norm": 0.22262315452098846, "learning_rate": 3.1538056400168676e-05, "loss": 0.0267, "step": 51150 }, { "grad_norm": 0.2641167938709259, "learning_rate": 3.151885014573506e-05, "loss": 0.0249, "step": 51160 }, { "grad_norm": 0.3064456582069397, "learning_rate": 3.149964704922448e-05, "loss": 0.0257, "step": 51170 }, { "grad_norm": 0.2346615195274353, "learning_rate": 3.148044711391825e-05, "loss": 0.0184, "step": 51180 }, { "grad_norm": 0.21992333233356476, "learning_rate": 3.146125034309707e-05, "loss": 0.0207, "step": 51190 }, { "grad_norm": 0.22679853439331055, "learning_rate": 3.1442056740041195e-05, "loss": 0.0285, "step": 51200 }, { "grad_norm": 0.40488579869270325, "learning_rate": 3.142286630803026e-05, "loss": 0.0358, "step": 51210 }, { "grad_norm": 0.30023038387298584, "learning_rate": 3.140367905034337e-05, "loss": 0.0195, "step": 51220 }, { "grad_norm": 0.20069724321365356, "learning_rate": 3.138449497025914e-05, "loss": 0.0246, "step": 51230 }, { "grad_norm": 0.4426819682121277, "learning_rate": 3.136531407105557e-05, "loss": 0.0248, "step": 51240 }, { "grad_norm": 0.26626065373420715, "learning_rate": 3.1346136356010184e-05, "loss": 0.0286, "step": 51250 }, { "grad_norm": 0.22936572134494781, "learning_rate": 3.132696182839988e-05, "loss": 0.0217, "step": 51260 }, { "grad_norm": 0.34923315048217773, "learning_rate": 3.130779049150111e-05, "loss": 0.0275, "step": 51270 }, { "grad_norm": 0.38020333647727966, "learning_rate": 3.128862234858971e-05, "loss": 0.0212, "step": 51280 }, { "grad_norm": 0.307142972946167, "learning_rate": 3.1269457402941005e-05, "loss": 0.0216, "step": 51290 }, { "grad_norm": 0.2590712904930115, "learning_rate": 3.125029565782974e-05, "loss": 0.0176, "step": 51300 }, { "grad_norm": 0.4807068407535553, "learning_rate": 3.123113711653013e-05, "loss": 0.0251, "step": 51310 }, { "grad_norm": 0.3696208894252777, "learning_rate": 3.121198178231587e-05, "loss": 0.0245, "step": 51320 }, { "grad_norm": 0.374551385641098, "learning_rate": 3.119282965846006e-05, "loss": 0.0205, "step": 51330 }, { "grad_norm": 0.3857969641685486, "learning_rate": 3.1173680748235303e-05, "loss": 0.0225, "step": 51340 }, { "grad_norm": 0.34222543239593506, "learning_rate": 3.115453505491358e-05, "loss": 0.0262, "step": 51350 }, { "grad_norm": 0.3169116973876953, "learning_rate": 3.1135392581766404e-05, "loss": 0.0203, "step": 51360 }, { "grad_norm": 0.2519746720790863, "learning_rate": 3.111625333206467e-05, "loss": 0.0223, "step": 51370 }, { "grad_norm": 0.27307578921318054, "learning_rate": 3.109711730907881e-05, "loss": 0.0185, "step": 51380 }, { "grad_norm": 0.3556232154369354, "learning_rate": 3.107798451607859e-05, "loss": 0.0203, "step": 51390 }, { "grad_norm": 0.2660761773586273, "learning_rate": 3.105885495633328e-05, "loss": 0.0199, "step": 51400 }, { "grad_norm": 0.19702357053756714, "learning_rate": 3.103972863311163e-05, "loss": 0.0167, "step": 51410 }, { "grad_norm": 0.33644986152648926, "learning_rate": 3.102060554968178e-05, "loss": 0.0266, "step": 51420 }, { "grad_norm": 0.26645857095718384, "learning_rate": 3.100148570931137e-05, "loss": 0.0172, "step": 51430 }, { "grad_norm": 0.23061440885066986, "learning_rate": 3.098236911526744e-05, "loss": 0.0235, "step": 51440 }, { "grad_norm": 0.3349817395210266, "learning_rate": 3.0963255770816495e-05, "loss": 0.0148, "step": 51450 }, { "grad_norm": 0.21815283596515656, "learning_rate": 3.094414567922447e-05, "loss": 0.0231, "step": 51460 }, { "grad_norm": 0.2238413542509079, "learning_rate": 3.092503884375678e-05, "loss": 0.0214, "step": 51470 }, { "grad_norm": 0.6423467993736267, "learning_rate": 3.0905935267678255e-05, "loss": 0.0284, "step": 51480 }, { "grad_norm": 0.30558323860168457, "learning_rate": 3.0886834954253166e-05, "loss": 0.0183, "step": 51490 }, { "grad_norm": 0.33870893716812134, "learning_rate": 3.086773790674524e-05, "loss": 0.0222, "step": 51500 }, { "grad_norm": 0.44836801290512085, "learning_rate": 3.0848644128417617e-05, "loss": 0.0241, "step": 51510 }, { "grad_norm": 0.33803778886795044, "learning_rate": 3.082955362253294e-05, "loss": 0.0244, "step": 51520 }, { "grad_norm": 0.3591398298740387, "learning_rate": 3.081046639235323e-05, "loss": 0.0268, "step": 51530 }, { "grad_norm": 0.27394503355026245, "learning_rate": 3.0791382441139986e-05, "loss": 0.0158, "step": 51540 }, { "grad_norm": 0.35322001576423645, "learning_rate": 3.077230177215411e-05, "loss": 0.0218, "step": 51550 }, { "grad_norm": 0.25188684463500977, "learning_rate": 3.0753224388656e-05, "loss": 0.0175, "step": 51560 }, { "grad_norm": 0.2660154700279236, "learning_rate": 3.073415029390544e-05, "loss": 0.0312, "step": 51570 }, { "grad_norm": 0.20711781084537506, "learning_rate": 3.071507949116166e-05, "loss": 0.0198, "step": 51580 }, { "grad_norm": 0.370594322681427, "learning_rate": 3.069601198368337e-05, "loss": 0.0217, "step": 51590 }, { "grad_norm": 0.3329165279865265, "learning_rate": 3.067694777472864e-05, "loss": 0.0185, "step": 51600 }, { "grad_norm": 0.2061898410320282, "learning_rate": 3.065788686755508e-05, "loss": 0.0186, "step": 51610 }, { "grad_norm": 0.2431284785270691, "learning_rate": 3.063882926541961e-05, "loss": 0.0232, "step": 51620 }, { "grad_norm": 0.3559480905532837, "learning_rate": 3.061977497157872e-05, "loss": 0.0189, "step": 51630 }, { "grad_norm": 0.42887213826179504, "learning_rate": 3.0600723989288224e-05, "loss": 0.0227, "step": 51640 }, { "grad_norm": 0.3433935344219208, "learning_rate": 3.058167632180343e-05, "loss": 0.0304, "step": 51650 }, { "grad_norm": 0.8588144779205322, "learning_rate": 3.0562631972379074e-05, "loss": 0.0262, "step": 51660 }, { "grad_norm": 0.2896077632904053, "learning_rate": 3.0543590944269276e-05, "loss": 0.0221, "step": 51670 }, { "grad_norm": 0.3763158619403839, "learning_rate": 3.052455324072766e-05, "loss": 0.0287, "step": 51680 }, { "grad_norm": 0.3386867046356201, "learning_rate": 3.0505518865007244e-05, "loss": 0.0211, "step": 51690 }, { "grad_norm": 0.2657149136066437, "learning_rate": 3.048648782036048e-05, "loss": 0.0183, "step": 51700 }, { "grad_norm": 0.25328704714775085, "learning_rate": 3.0467460110039236e-05, "loss": 0.0199, "step": 51710 }, { "grad_norm": 0.35279276967048645, "learning_rate": 3.0448435737294866e-05, "loss": 0.0245, "step": 51720 }, { "grad_norm": 0.3247304856777191, "learning_rate": 3.042941470537808e-05, "loss": 0.0184, "step": 51730 }, { "grad_norm": 0.37053385376930237, "learning_rate": 3.0410397017539072e-05, "loss": 0.0228, "step": 51740 }, { "grad_norm": 0.3826713562011719, "learning_rate": 3.039138267702742e-05, "loss": 0.02, "step": 51750 }, { "grad_norm": 0.257843941450119, "learning_rate": 3.0372371687092193e-05, "loss": 0.0162, "step": 51760 }, { "grad_norm": 0.34304672479629517, "learning_rate": 3.0353364050981823e-05, "loss": 0.0185, "step": 51770 }, { "grad_norm": 0.4211587607860565, "learning_rate": 3.033435977194418e-05, "loss": 0.0213, "step": 51780 }, { "grad_norm": 0.35023507475852966, "learning_rate": 3.0315358853226618e-05, "loss": 0.0239, "step": 51790 }, { "grad_norm": 0.6556239724159241, "learning_rate": 3.0296361298075844e-05, "loss": 0.0229, "step": 51800 }, { "grad_norm": 0.3732854425907135, "learning_rate": 3.027736710973803e-05, "loss": 0.025, "step": 51810 }, { "grad_norm": 0.1796884834766388, "learning_rate": 3.025837629145875e-05, "loss": 0.0167, "step": 51820 }, { "grad_norm": 0.29912689328193665, "learning_rate": 3.0239388846483048e-05, "loss": 0.0223, "step": 51830 }, { "grad_norm": 0.22075064480304718, "learning_rate": 3.022040477805532e-05, "loss": 0.0254, "step": 51840 }, { "grad_norm": 0.3778507709503174, "learning_rate": 3.020142408941946e-05, "loss": 0.0221, "step": 51850 }, { "grad_norm": 0.4462396204471588, "learning_rate": 3.018244678381873e-05, "loss": 0.0195, "step": 51860 }, { "grad_norm": 0.47878164052963257, "learning_rate": 3.0163472864495812e-05, "loss": 0.0254, "step": 51870 }, { "grad_norm": 0.27276864647865295, "learning_rate": 3.014450233469287e-05, "loss": 0.0233, "step": 51880 }, { "grad_norm": 0.3448357582092285, "learning_rate": 3.0125535197651422e-05, "loss": 0.0269, "step": 51890 }, { "grad_norm": 0.2665356993675232, "learning_rate": 3.0106571456612448e-05, "loss": 0.0191, "step": 51900 }, { "grad_norm": 0.3818536698818207, "learning_rate": 3.0087611114816305e-05, "loss": 0.0291, "step": 51910 }, { "grad_norm": 0.31711289286613464, "learning_rate": 3.006865417550284e-05, "loss": 0.0202, "step": 51920 }, { "grad_norm": 0.3174024820327759, "learning_rate": 3.0049700641911242e-05, "loss": 0.0224, "step": 51930 }, { "grad_norm": 0.2689805328845978, "learning_rate": 3.0030750517280175e-05, "loss": 0.0239, "step": 51940 }, { "grad_norm": 0.26834043860435486, "learning_rate": 3.0011803804847682e-05, "loss": 0.0257, "step": 51950 }, { "grad_norm": 0.37480413913726807, "learning_rate": 2.9992860507851224e-05, "loss": 0.0217, "step": 51960 }, { "grad_norm": 0.26678362488746643, "learning_rate": 2.9973920629527723e-05, "loss": 0.0148, "step": 51970 }, { "grad_norm": 0.19037486612796783, "learning_rate": 2.9954984173113453e-05, "loss": 0.0193, "step": 51980 }, { "grad_norm": 0.34430280327796936, "learning_rate": 2.993605114184418e-05, "loss": 0.0267, "step": 51990 }, { "grad_norm": 0.44645488262176514, "learning_rate": 2.9917121538955005e-05, "loss": 0.0269, "step": 52000 }, { "grad_norm": 0.3428857922554016, "learning_rate": 2.98981953676805e-05, "loss": 0.0296, "step": 52010 }, { "grad_norm": 0.24703921377658844, "learning_rate": 2.9879272631254594e-05, "loss": 0.02, "step": 52020 }, { "grad_norm": 0.21372202038764954, "learning_rate": 2.986035333291073e-05, "loss": 0.02, "step": 52030 }, { "grad_norm": 0.23518787324428558, "learning_rate": 2.9841437475881646e-05, "loss": 0.0237, "step": 52040 }, { "grad_norm": 0.2948774993419647, "learning_rate": 2.9822525063399555e-05, "loss": 0.0199, "step": 52050 }, { "grad_norm": 0.2129734605550766, "learning_rate": 2.9803616098696087e-05, "loss": 0.0245, "step": 52060 }, { "grad_norm": 0.31137368083000183, "learning_rate": 2.9784710585002228e-05, "loss": 0.0262, "step": 52070 }, { "grad_norm": 0.3095332086086273, "learning_rate": 2.9765808525548467e-05, "loss": 0.0224, "step": 52080 }, { "grad_norm": 0.31393447518348694, "learning_rate": 2.974690992356461e-05, "loss": 0.0215, "step": 52090 }, { "grad_norm": 0.17326979339122772, "learning_rate": 2.9728014782279934e-05, "loss": 0.0238, "step": 52100 }, { "grad_norm": 0.2788352072238922, "learning_rate": 2.970912310492307e-05, "loss": 0.0213, "step": 52110 }, { "grad_norm": 0.2454557716846466, "learning_rate": 2.9690234894722134e-05, "loss": 0.0209, "step": 52120 }, { "grad_norm": 0.17827725410461426, "learning_rate": 2.9671350154904577e-05, "loss": 0.0266, "step": 52130 }, { "grad_norm": 0.27429187297821045, "learning_rate": 2.9652468888697282e-05, "loss": 0.0207, "step": 52140 }, { "grad_norm": 0.44009286165237427, "learning_rate": 2.9633591099326562e-05, "loss": 0.0183, "step": 52150 }, { "grad_norm": 0.2762565314769745, "learning_rate": 2.9614716790018072e-05, "loss": 0.0165, "step": 52160 }, { "grad_norm": 0.28561532497406006, "learning_rate": 2.959584596399697e-05, "loss": 0.0144, "step": 52170 }, { "grad_norm": 0.37664708495140076, "learning_rate": 2.9576978624487717e-05, "loss": 0.0233, "step": 52180 }, { "grad_norm": 0.31211185455322266, "learning_rate": 2.9558114774714263e-05, "loss": 0.02, "step": 52190 }, { "grad_norm": 0.36318081617355347, "learning_rate": 2.9539254417899897e-05, "loss": 0.0241, "step": 52200 }, { "grad_norm": 0.3560534417629242, "learning_rate": 2.9520397557267365e-05, "loss": 0.0206, "step": 52210 }, { "grad_norm": 0.2453194558620453, "learning_rate": 2.9501544196038765e-05, "loss": 0.024, "step": 52220 }, { "grad_norm": 0.26365381479263306, "learning_rate": 2.9482694337435614e-05, "loss": 0.0305, "step": 52230 }, { "grad_norm": 0.23052704334259033, "learning_rate": 2.946384798467887e-05, "loss": 0.0179, "step": 52240 }, { "grad_norm": 0.2461378574371338, "learning_rate": 2.9445005140988824e-05, "loss": 0.0163, "step": 52250 }, { "grad_norm": 0.29220858216285706, "learning_rate": 2.942616580958524e-05, "loss": 0.019, "step": 52260 }, { "grad_norm": 0.2726554274559021, "learning_rate": 2.9407329993687193e-05, "loss": 0.0171, "step": 52270 }, { "grad_norm": 0.4578266441822052, "learning_rate": 2.938849769651326e-05, "loss": 0.0154, "step": 52280 }, { "grad_norm": 0.3617180287837982, "learning_rate": 2.9369668921281345e-05, "loss": 0.0199, "step": 52290 }, { "grad_norm": 0.24691690504550934, "learning_rate": 2.9350843671208773e-05, "loss": 0.017, "step": 52300 }, { "grad_norm": 0.3347949683666229, "learning_rate": 2.9332021949512245e-05, "loss": 0.019, "step": 52310 }, { "grad_norm": 0.4140705168247223, "learning_rate": 2.9313203759407908e-05, "loss": 0.0209, "step": 52320 }, { "grad_norm": 0.23069480061531067, "learning_rate": 2.929438910411127e-05, "loss": 0.0334, "step": 52330 }, { "grad_norm": 0.45314154028892517, "learning_rate": 2.927557798683722e-05, "loss": 0.0256, "step": 52340 }, { "grad_norm": 0.22497820854187012, "learning_rate": 2.9256770410800095e-05, "loss": 0.0197, "step": 52350 }, { "grad_norm": 0.2697463631629944, "learning_rate": 2.9237966379213554e-05, "loss": 0.0185, "step": 52360 }, { "grad_norm": 0.3623933792114258, "learning_rate": 2.9219165895290736e-05, "loss": 0.0194, "step": 52370 }, { "grad_norm": 0.19473415613174438, "learning_rate": 2.9200368962244083e-05, "loss": 0.0236, "step": 52380 }, { "grad_norm": 0.2218369096517563, "learning_rate": 2.9181575583285536e-05, "loss": 0.0191, "step": 52390 }, { "grad_norm": 0.2643360197544098, "learning_rate": 2.916278576162632e-05, "loss": 0.0233, "step": 52400 }, { "grad_norm": 0.24232570827007294, "learning_rate": 2.9143999500477116e-05, "loss": 0.0191, "step": 52410 }, { "grad_norm": 0.2782343029975891, "learning_rate": 2.9125216803048004e-05, "loss": 0.0285, "step": 52420 }, { "grad_norm": 0.22106638550758362, "learning_rate": 2.9106437672548403e-05, "loss": 0.0185, "step": 52430 }, { "grad_norm": 0.21129940450191498, "learning_rate": 2.9087662112187154e-05, "loss": 0.0172, "step": 52440 }, { "grad_norm": 0.5191906690597534, "learning_rate": 2.9068890125172498e-05, "loss": 0.0173, "step": 52450 }, { "grad_norm": 0.2031264454126358, "learning_rate": 2.9050121714712058e-05, "loss": 0.0231, "step": 52460 }, { "grad_norm": 0.2764514982700348, "learning_rate": 2.9031356884012835e-05, "loss": 0.0241, "step": 52470 }, { "grad_norm": 0.17647285759449005, "learning_rate": 2.9012595636281247e-05, "loss": 0.0231, "step": 52480 }, { "grad_norm": 0.2345162034034729, "learning_rate": 2.8993837974723016e-05, "loss": 0.0273, "step": 52490 }, { "grad_norm": 0.27297845482826233, "learning_rate": 2.8975083902543394e-05, "loss": 0.0178, "step": 52500 }, { "grad_norm": 0.2863997220993042, "learning_rate": 2.895633342294688e-05, "loss": 0.0235, "step": 52510 }, { "grad_norm": 0.2803654670715332, "learning_rate": 2.893758653913744e-05, "loss": 0.0167, "step": 52520 }, { "grad_norm": 0.19907119870185852, "learning_rate": 2.8918843254318423e-05, "loss": 0.0179, "step": 52530 }, { "grad_norm": 0.3534562289714813, "learning_rate": 2.8900103571692483e-05, "loss": 0.0231, "step": 52540 }, { "grad_norm": 0.3349807560443878, "learning_rate": 2.8881367494461796e-05, "loss": 0.0183, "step": 52550 }, { "grad_norm": 0.31097185611724854, "learning_rate": 2.886263502582779e-05, "loss": 0.0225, "step": 52560 }, { "grad_norm": 0.3889528214931488, "learning_rate": 2.8843906168991353e-05, "loss": 0.0213, "step": 52570 }, { "grad_norm": 0.2292400598526001, "learning_rate": 2.8825180927152728e-05, "loss": 0.0216, "step": 52580 }, { "grad_norm": 0.35590821504592896, "learning_rate": 2.8806459303511547e-05, "loss": 0.0151, "step": 52590 }, { "grad_norm": 0.29287928342819214, "learning_rate": 2.8787741301266852e-05, "loss": 0.0202, "step": 52600 }, { "grad_norm": 0.2805315852165222, "learning_rate": 2.876902692361697e-05, "loss": 0.0183, "step": 52610 }, { "grad_norm": 0.2435590624809265, "learning_rate": 2.875031617375976e-05, "loss": 0.0188, "step": 52620 }, { "grad_norm": 0.21906599402427673, "learning_rate": 2.8731609054892318e-05, "loss": 0.015, "step": 52630 }, { "grad_norm": 0.3438796401023865, "learning_rate": 2.871290557021119e-05, "loss": 0.02, "step": 52640 }, { "grad_norm": 0.3139813542366028, "learning_rate": 2.8694205722912298e-05, "loss": 0.0174, "step": 52650 }, { "grad_norm": 0.4041745960712433, "learning_rate": 2.8675509516190936e-05, "loss": 0.023, "step": 52660 }, { "grad_norm": 0.21831712126731873, "learning_rate": 2.8656816953241772e-05, "loss": 0.0266, "step": 52670 }, { "grad_norm": 0.5042135715484619, "learning_rate": 2.863812803725887e-05, "loss": 0.0185, "step": 52680 }, { "grad_norm": 0.2521030306816101, "learning_rate": 2.8619442771435623e-05, "loss": 0.0192, "step": 52690 }, { "grad_norm": 0.23893946409225464, "learning_rate": 2.860076115896484e-05, "loss": 0.025, "step": 52700 }, { "grad_norm": 0.28548914194107056, "learning_rate": 2.8582083203038712e-05, "loss": 0.019, "step": 52710 }, { "grad_norm": 0.2987220585346222, "learning_rate": 2.8563408906848778e-05, "loss": 0.019, "step": 52720 }, { "grad_norm": 0.3011718690395355, "learning_rate": 2.8544738273585993e-05, "loss": 0.0211, "step": 52730 }, { "grad_norm": 0.2748377025127411, "learning_rate": 2.8526071306440595e-05, "loss": 0.0161, "step": 52740 }, { "grad_norm": 0.2862199544906616, "learning_rate": 2.850740800860233e-05, "loss": 0.0207, "step": 52750 }, { "grad_norm": 0.23430679738521576, "learning_rate": 2.8488748383260178e-05, "loss": 0.0188, "step": 52760 }, { "grad_norm": 0.6926658749580383, "learning_rate": 2.8470092433602635e-05, "loss": 0.0239, "step": 52770 }, { "grad_norm": 0.3304423689842224, "learning_rate": 2.845144016281745e-05, "loss": 0.0199, "step": 52780 }, { "grad_norm": 0.33999723196029663, "learning_rate": 2.8432791574091744e-05, "loss": 0.0302, "step": 52790 }, { "grad_norm": 0.31489163637161255, "learning_rate": 2.8414146670612134e-05, "loss": 0.0185, "step": 52800 }, { "grad_norm": 0.1902848780155182, "learning_rate": 2.8395505455564446e-05, "loss": 0.0274, "step": 52810 }, { "grad_norm": 0.29726916551589966, "learning_rate": 2.837686793213403e-05, "loss": 0.0217, "step": 52820 }, { "grad_norm": 0.2866336703300476, "learning_rate": 2.8358234103505477e-05, "loss": 0.0248, "step": 52830 }, { "grad_norm": 0.30434030294418335, "learning_rate": 2.8339603972862806e-05, "loss": 0.0269, "step": 52840 }, { "grad_norm": 0.5860031843185425, "learning_rate": 2.83209775433894e-05, "loss": 0.0246, "step": 52850 }, { "grad_norm": 0.21703417599201202, "learning_rate": 2.8302354818268013e-05, "loss": 0.0233, "step": 52860 }, { "grad_norm": 0.6316068768501282, "learning_rate": 2.8283735800680754e-05, "loss": 0.0259, "step": 52870 }, { "grad_norm": 0.19473858177661896, "learning_rate": 2.826512049380913e-05, "loss": 0.0188, "step": 52880 }, { "grad_norm": 0.29354941844940186, "learning_rate": 2.8246508900833935e-05, "loss": 0.0217, "step": 52890 }, { "grad_norm": 0.37713149189949036, "learning_rate": 2.822790102493541e-05, "loss": 0.0243, "step": 52900 }, { "grad_norm": 0.2617332339286804, "learning_rate": 2.820929686929314e-05, "loss": 0.0228, "step": 52910 }, { "grad_norm": 0.2739585340023041, "learning_rate": 2.819069643708605e-05, "loss": 0.0204, "step": 52920 }, { "grad_norm": 0.25615060329437256, "learning_rate": 2.8172099731492484e-05, "loss": 0.0205, "step": 52930 }, { "grad_norm": 0.3278346359729767, "learning_rate": 2.815350675569003e-05, "loss": 0.0305, "step": 52940 }, { "grad_norm": 0.2912069261074066, "learning_rate": 2.8134917512855807e-05, "loss": 0.0292, "step": 52950 }, { "grad_norm": 0.3308424651622772, "learning_rate": 2.8116332006166134e-05, "loss": 0.0133, "step": 52960 }, { "grad_norm": 0.30009597539901733, "learning_rate": 2.809775023879685e-05, "loss": 0.0307, "step": 52970 }, { "grad_norm": 0.496578574180603, "learning_rate": 2.807917221392299e-05, "loss": 0.0221, "step": 52980 }, { "grad_norm": 0.2651825249195099, "learning_rate": 2.8060597934719067e-05, "loss": 0.0181, "step": 52990 }, { "grad_norm": 0.23909297585487366, "learning_rate": 2.8042027404358927e-05, "loss": 0.019, "step": 53000 }, { "grad_norm": 0.3071237802505493, "learning_rate": 2.802346062601572e-05, "loss": 0.0232, "step": 53010 }, { "grad_norm": 0.4470633268356323, "learning_rate": 2.800489760286207e-05, "loss": 0.0367, "step": 53020 }, { "grad_norm": 0.3401888608932495, "learning_rate": 2.7986338338069834e-05, "loss": 0.0207, "step": 53030 }, { "grad_norm": 0.28348541259765625, "learning_rate": 2.7967782834810297e-05, "loss": 0.0235, "step": 53040 }, { "grad_norm": 0.25106456875801086, "learning_rate": 2.794923109625409e-05, "loss": 0.0211, "step": 53050 }, { "grad_norm": 0.32477664947509766, "learning_rate": 2.7930683125571205e-05, "loss": 0.0177, "step": 53060 }, { "grad_norm": 0.26618707180023193, "learning_rate": 2.7912138925930997e-05, "loss": 0.0313, "step": 53070 }, { "grad_norm": 0.30532217025756836, "learning_rate": 2.7893598500502117e-05, "loss": 0.0435, "step": 53080 }, { "grad_norm": 0.2980576157569885, "learning_rate": 2.7875061852452644e-05, "loss": 0.0191, "step": 53090 }, { "grad_norm": 0.3939913213253021, "learning_rate": 2.7856528984949982e-05, "loss": 0.027, "step": 53100 }, { "grad_norm": 0.264618456363678, "learning_rate": 2.7837999901160888e-05, "loss": 0.0212, "step": 53110 }, { "grad_norm": 0.3919631242752075, "learning_rate": 2.7819474604251484e-05, "loss": 0.019, "step": 53120 }, { "grad_norm": 0.25825929641723633, "learning_rate": 2.780095309738725e-05, "loss": 0.0225, "step": 53130 }, { "grad_norm": 0.3318645656108856, "learning_rate": 2.778243538373294e-05, "loss": 0.0196, "step": 53140 }, { "grad_norm": 0.2111615240573883, "learning_rate": 2.7763921466452826e-05, "loss": 0.0211, "step": 53150 }, { "grad_norm": 0.35491839051246643, "learning_rate": 2.7745411348710336e-05, "loss": 0.0275, "step": 53160 }, { "grad_norm": 0.3554823398590088, "learning_rate": 2.7726905033668395e-05, "loss": 0.0252, "step": 53170 }, { "grad_norm": 0.25006312131881714, "learning_rate": 2.7708402524489214e-05, "loss": 0.0228, "step": 53180 }, { "grad_norm": 0.3253636062145233, "learning_rate": 2.7689903824334364e-05, "loss": 0.0236, "step": 53190 }, { "grad_norm": 0.29609203338623047, "learning_rate": 2.7671408936364785e-05, "loss": 0.0203, "step": 53200 }, { "grad_norm": 0.2780236005783081, "learning_rate": 2.765291786374069e-05, "loss": 0.0187, "step": 53210 }, { "grad_norm": 0.24999484419822693, "learning_rate": 2.7634430609621775e-05, "loss": 0.0163, "step": 53220 }, { "grad_norm": 0.6713378429412842, "learning_rate": 2.7615947177166956e-05, "loss": 0.0208, "step": 53230 }, { "grad_norm": 0.2926099896430969, "learning_rate": 2.7597467569534553e-05, "loss": 0.016, "step": 53240 }, { "grad_norm": 0.22773519158363342, "learning_rate": 2.757899178988226e-05, "loss": 0.0191, "step": 53250 }, { "grad_norm": 0.2042379230260849, "learning_rate": 2.7560519841367005e-05, "loss": 0.0218, "step": 53260 }, { "grad_norm": 0.17655636370182037, "learning_rate": 2.7542051727145237e-05, "loss": 0.0204, "step": 53270 }, { "grad_norm": 0.3106299638748169, "learning_rate": 2.7523587450372578e-05, "loss": 0.0281, "step": 53280 }, { "grad_norm": 0.3571084439754486, "learning_rate": 2.750512701420409e-05, "loss": 0.0162, "step": 53290 }, { "grad_norm": 0.14081481099128723, "learning_rate": 2.7486670421794158e-05, "loss": 0.0182, "step": 53300 }, { "grad_norm": 0.24362407624721527, "learning_rate": 2.7468217676296515e-05, "loss": 0.0249, "step": 53310 }, { "grad_norm": 0.3254942297935486, "learning_rate": 2.7449768780864226e-05, "loss": 0.0245, "step": 53320 }, { "grad_norm": 0.25623050332069397, "learning_rate": 2.7431323738649724e-05, "loss": 0.0204, "step": 53330 }, { "grad_norm": 0.4665789306163788, "learning_rate": 2.7412882552804713e-05, "loss": 0.022, "step": 53340 }, { "grad_norm": 0.23213893175125122, "learning_rate": 2.739444522648032e-05, "loss": 0.0155, "step": 53350 }, { "grad_norm": 0.22154764831066132, "learning_rate": 2.737601176282697e-05, "loss": 0.0204, "step": 53360 }, { "grad_norm": 0.39481836557388306, "learning_rate": 2.735758216499445e-05, "loss": 0.0202, "step": 53370 }, { "grad_norm": 0.32213670015335083, "learning_rate": 2.7339156436131864e-05, "loss": 0.0174, "step": 53380 }, { "grad_norm": 0.3144828975200653, "learning_rate": 2.7320734579387663e-05, "loss": 0.0164, "step": 53390 }, { "grad_norm": 0.2590526044368744, "learning_rate": 2.730231659790966e-05, "loss": 0.0159, "step": 53400 }, { "grad_norm": 0.26552069187164307, "learning_rate": 2.728390249484494e-05, "loss": 0.0225, "step": 53410 }, { "grad_norm": 0.22673362493515015, "learning_rate": 2.7265492273340032e-05, "loss": 0.0177, "step": 53420 }, { "grad_norm": 0.25482699275016785, "learning_rate": 2.7247085936540678e-05, "loss": 0.0177, "step": 53430 }, { "grad_norm": 0.42329731583595276, "learning_rate": 2.722868348759205e-05, "loss": 0.0252, "step": 53440 }, { "grad_norm": 0.270103394985199, "learning_rate": 2.7210284929638635e-05, "loss": 0.0279, "step": 53450 }, { "grad_norm": 0.2499268501996994, "learning_rate": 2.7191890265824183e-05, "loss": 0.0232, "step": 53460 }, { "grad_norm": 0.39740315079689026, "learning_rate": 2.7173499499291926e-05, "loss": 0.0226, "step": 53470 }, { "grad_norm": 0.3758704662322998, "learning_rate": 2.7155112633184277e-05, "loss": 0.024, "step": 53480 }, { "grad_norm": 0.2749707102775574, "learning_rate": 2.713672967064307e-05, "loss": 0.0231, "step": 53490 }, { "grad_norm": 0.19513648748397827, "learning_rate": 2.711835061480945e-05, "loss": 0.0174, "step": 53500 }, { "grad_norm": 0.2672918140888214, "learning_rate": 2.7099975468823896e-05, "loss": 0.027, "step": 53510 }, { "grad_norm": 0.32067084312438965, "learning_rate": 2.708160423582622e-05, "loss": 0.0247, "step": 53520 }, { "grad_norm": 0.6226522922515869, "learning_rate": 2.706323691895557e-05, "loss": 0.0289, "step": 53530 }, { "grad_norm": 0.3571699261665344, "learning_rate": 2.70448735213504e-05, "loss": 0.0233, "step": 53540 }, { "grad_norm": 0.27899235486984253, "learning_rate": 2.702651404614852e-05, "loss": 0.0205, "step": 53550 }, { "grad_norm": 0.3340490758419037, "learning_rate": 2.7008158496487056e-05, "loss": 0.0201, "step": 53560 }, { "grad_norm": 1.5228098630905151, "learning_rate": 2.6989806875502487e-05, "loss": 0.0186, "step": 53570 }, { "grad_norm": 0.27271655201911926, "learning_rate": 2.6971459186330584e-05, "loss": 0.0156, "step": 53580 }, { "grad_norm": 0.26953116059303284, "learning_rate": 2.695311543210648e-05, "loss": 0.0195, "step": 53590 }, { "grad_norm": 0.3073422610759735, "learning_rate": 2.693477561596463e-05, "loss": 0.0175, "step": 53600 }, { "grad_norm": 0.2994403839111328, "learning_rate": 2.6916439741038756e-05, "loss": 0.0185, "step": 53610 }, { "grad_norm": 0.47686442732810974, "learning_rate": 2.689810781046203e-05, "loss": 0.0224, "step": 53620 }, { "grad_norm": 0.22362923622131348, "learning_rate": 2.6879779827366823e-05, "loss": 0.0176, "step": 53630 }, { "grad_norm": 0.2960676848888397, "learning_rate": 2.6861455794884904e-05, "loss": 0.0183, "step": 53640 }, { "grad_norm": 0.3138507902622223, "learning_rate": 2.6843135716147373e-05, "loss": 0.0226, "step": 53650 }, { "grad_norm": 0.254511296749115, "learning_rate": 2.6824819594284556e-05, "loss": 0.0241, "step": 53660 }, { "grad_norm": 0.26849299669265747, "learning_rate": 2.6806507432426275e-05, "loss": 0.0231, "step": 53670 }, { "grad_norm": 0.25649216771125793, "learning_rate": 2.6788199233701512e-05, "loss": 0.0213, "step": 53680 }, { "grad_norm": 0.34818512201309204, "learning_rate": 2.6769895001238652e-05, "loss": 0.0232, "step": 53690 }, { "grad_norm": 0.3783120810985565, "learning_rate": 2.67515947381654e-05, "loss": 0.0207, "step": 53700 }, { "grad_norm": 0.37686505913734436, "learning_rate": 2.6733298447608745e-05, "loss": 0.0185, "step": 53710 }, { "grad_norm": 0.6710466742515564, "learning_rate": 2.6715006132695074e-05, "loss": 0.029, "step": 53720 }, { "grad_norm": 0.18523992598056793, "learning_rate": 2.6696717796549985e-05, "loss": 0.0253, "step": 53730 }, { "grad_norm": 0.4445994794368744, "learning_rate": 2.667843344229848e-05, "loss": 0.0217, "step": 53740 }, { "grad_norm": 0.31929972767829895, "learning_rate": 2.6660153073064846e-05, "loss": 0.0242, "step": 53750 }, { "grad_norm": 0.3238218426704407, "learning_rate": 2.664187669197271e-05, "loss": 0.0277, "step": 53760 }, { "grad_norm": 0.4057847857475281, "learning_rate": 2.6623604302145005e-05, "loss": 0.0205, "step": 53770 }, { "grad_norm": 0.31290555000305176, "learning_rate": 2.6605335906703975e-05, "loss": 0.0198, "step": 53780 }, { "grad_norm": 0.3182675540447235, "learning_rate": 2.65870715087712e-05, "loss": 0.0183, "step": 53790 }, { "grad_norm": 0.28331995010375977, "learning_rate": 2.6568811111467573e-05, "loss": 0.0194, "step": 53800 }, { "grad_norm": 0.2996138036251068, "learning_rate": 2.6550554717913258e-05, "loss": 0.019, "step": 53810 }, { "grad_norm": 0.2986421585083008, "learning_rate": 2.6532302331227805e-05, "loss": 0.0206, "step": 53820 }, { "grad_norm": 0.3589681386947632, "learning_rate": 2.651405395453004e-05, "loss": 0.0222, "step": 53830 }, { "grad_norm": 0.29406043887138367, "learning_rate": 2.6495809590938115e-05, "loss": 0.0259, "step": 53840 }, { "grad_norm": 0.1964416205883026, "learning_rate": 2.64775692435695e-05, "loss": 0.0213, "step": 53850 }, { "grad_norm": 0.25795289874076843, "learning_rate": 2.6459332915540928e-05, "loss": 0.018, "step": 53860 }, { "grad_norm": 0.24752835929393768, "learning_rate": 2.644110060996856e-05, "loss": 0.0257, "step": 53870 }, { "grad_norm": 0.17813464999198914, "learning_rate": 2.642287232996774e-05, "loss": 0.0196, "step": 53880 }, { "grad_norm": 0.33042216300964355, "learning_rate": 2.6404648078653205e-05, "loss": 0.0167, "step": 53890 }, { "grad_norm": 0.35655099153518677, "learning_rate": 2.6386427859139002e-05, "loss": 0.0245, "step": 53900 }, { "grad_norm": 0.4315664768218994, "learning_rate": 2.63682116745384e-05, "loss": 0.0204, "step": 53910 }, { "grad_norm": 0.2794186770915985, "learning_rate": 2.6349999527964138e-05, "loss": 0.0197, "step": 53920 }, { "grad_norm": 0.26779693365097046, "learning_rate": 2.63317914225281e-05, "loss": 0.0249, "step": 53930 }, { "grad_norm": 0.18410611152648926, "learning_rate": 2.6313587361341585e-05, "loss": 0.0248, "step": 53940 }, { "grad_norm": 0.5997313261032104, "learning_rate": 2.6295387347515165e-05, "loss": 0.0271, "step": 53950 }, { "grad_norm": 0.35571831464767456, "learning_rate": 2.6277191384158727e-05, "loss": 0.0277, "step": 53960 }, { "grad_norm": 0.24761909246444702, "learning_rate": 2.6258999474381457e-05, "loss": 0.0239, "step": 53970 }, { "grad_norm": 0.33022457361221313, "learning_rate": 2.624081162129186e-05, "loss": 0.0239, "step": 53980 }, { "grad_norm": 0.2060290426015854, "learning_rate": 2.6222627827997765e-05, "loss": 0.0321, "step": 53990 }, { "grad_norm": 0.24283456802368164, "learning_rate": 2.6204448097606236e-05, "loss": 0.0203, "step": 54000 }, { "grad_norm": 0.30772286653518677, "learning_rate": 2.6186272433223726e-05, "loss": 0.0236, "step": 54010 }, { "grad_norm": 0.3142155110836029, "learning_rate": 2.6168100837955943e-05, "loss": 0.0217, "step": 54020 }, { "grad_norm": 0.23053641617298126, "learning_rate": 2.6149933314907926e-05, "loss": 0.0228, "step": 54030 }, { "grad_norm": 0.3462408185005188, "learning_rate": 2.6131769867184e-05, "loss": 0.0264, "step": 54040 }, { "grad_norm": 0.27507680654525757, "learning_rate": 2.611361049788783e-05, "loss": 0.0223, "step": 54050 }, { "grad_norm": 0.3742872476577759, "learning_rate": 2.6095455210122292e-05, "loss": 0.0222, "step": 54060 }, { "grad_norm": 0.26681211590766907, "learning_rate": 2.6077304006989712e-05, "loss": 0.0153, "step": 54070 }, { "grad_norm": 0.29506853222846985, "learning_rate": 2.6059156891591562e-05, "loss": 0.0225, "step": 54080 }, { "grad_norm": 0.3070394992828369, "learning_rate": 2.6041013867028718e-05, "loss": 0.0235, "step": 54090 }, { "grad_norm": 0.35768142342567444, "learning_rate": 2.6022874936401347e-05, "loss": 0.0143, "step": 54100 }, { "grad_norm": 0.23974935710430145, "learning_rate": 2.6004740102808832e-05, "loss": 0.02, "step": 54110 }, { "grad_norm": 0.28600648045539856, "learning_rate": 2.598660936935e-05, "loss": 0.0149, "step": 54120 }, { "grad_norm": 0.4544752836227417, "learning_rate": 2.5968482739122845e-05, "loss": 0.0193, "step": 54130 }, { "grad_norm": 0.35440221428871155, "learning_rate": 2.595036021522472e-05, "loss": 0.0171, "step": 54140 }, { "grad_norm": 0.5061582922935486, "learning_rate": 2.5932241800752278e-05, "loss": 0.0248, "step": 54150 }, { "grad_norm": 0.43632370233535767, "learning_rate": 2.5914127498801453e-05, "loss": 0.0324, "step": 54160 }, { "grad_norm": 0.1971825659275055, "learning_rate": 2.5896017312467497e-05, "loss": 0.0184, "step": 54170 }, { "grad_norm": 0.3665805757045746, "learning_rate": 2.587791124484493e-05, "loss": 0.0171, "step": 54180 }, { "grad_norm": 0.29506057500839233, "learning_rate": 2.5859809299027615e-05, "loss": 0.0179, "step": 54190 }, { "grad_norm": 0.25672680139541626, "learning_rate": 2.5841711478108632e-05, "loss": 0.029, "step": 54200 }, { "grad_norm": 0.49539992213249207, "learning_rate": 2.582361778518043e-05, "loss": 0.0216, "step": 54210 }, { "grad_norm": 0.27136364579200745, "learning_rate": 2.580552822333472e-05, "loss": 0.0223, "step": 54220 }, { "grad_norm": 0.3140889108181, "learning_rate": 2.578744279566252e-05, "loss": 0.023, "step": 54230 }, { "grad_norm": 0.2066611796617508, "learning_rate": 2.576936150525413e-05, "loss": 0.0247, "step": 54240 }, { "grad_norm": 0.20245394110679626, "learning_rate": 2.5751284355199168e-05, "loss": 0.0165, "step": 54250 }, { "grad_norm": 0.1875944584608078, "learning_rate": 2.573321134858646e-05, "loss": 0.0135, "step": 54260 }, { "grad_norm": 0.2130141705274582, "learning_rate": 2.5715142488504286e-05, "loss": 0.018, "step": 54270 }, { "grad_norm": 0.24264907836914062, "learning_rate": 2.5697077778040042e-05, "loss": 0.0175, "step": 54280 }, { "grad_norm": 0.24600531160831451, "learning_rate": 2.5679017220280522e-05, "loss": 0.0376, "step": 54290 }, { "grad_norm": 0.33199286460876465, "learning_rate": 2.5660960818311796e-05, "loss": 0.0222, "step": 54300 }, { "grad_norm": 0.43560776114463806, "learning_rate": 2.564290857521915e-05, "loss": 0.0202, "step": 54310 }, { "grad_norm": 0.21395756304264069, "learning_rate": 2.5624860494087298e-05, "loss": 0.0204, "step": 54320 }, { "grad_norm": 0.17979387938976288, "learning_rate": 2.5606816578000115e-05, "loss": 0.0228, "step": 54330 }, { "grad_norm": 0.35502612590789795, "learning_rate": 2.558877683004082e-05, "loss": 0.0184, "step": 54340 }, { "grad_norm": 0.1737358123064041, "learning_rate": 2.557074125329192e-05, "loss": 0.019, "step": 54350 }, { "grad_norm": 0.2772735357284546, "learning_rate": 2.5552709850835195e-05, "loss": 0.0182, "step": 54360 }, { "grad_norm": 0.2513229548931122, "learning_rate": 2.5534682625751738e-05, "loss": 0.0138, "step": 54370 }, { "grad_norm": 0.2985042631626129, "learning_rate": 2.551665958112186e-05, "loss": 0.024, "step": 54380 }, { "grad_norm": 0.35844528675079346, "learning_rate": 2.549864072002527e-05, "loss": 0.019, "step": 54390 }, { "grad_norm": 0.32547393441200256, "learning_rate": 2.5480626045540858e-05, "loss": 0.0216, "step": 54400 }, { "grad_norm": 0.4221634566783905, "learning_rate": 2.546261556074684e-05, "loss": 0.015, "step": 54410 }, { "grad_norm": 0.26382866501808167, "learning_rate": 2.5444609268720726e-05, "loss": 0.0201, "step": 54420 }, { "grad_norm": 0.2586909532546997, "learning_rate": 2.5426607172539297e-05, "loss": 0.0171, "step": 54430 }, { "grad_norm": 0.20833536982536316, "learning_rate": 2.5408609275278617e-05, "loss": 0.0232, "step": 54440 }, { "grad_norm": 0.3538796603679657, "learning_rate": 2.5390615580014055e-05, "loss": 0.021, "step": 54450 }, { "grad_norm": 0.2222040891647339, "learning_rate": 2.5372626089820207e-05, "loss": 0.0223, "step": 54460 }, { "grad_norm": 0.33059313893318176, "learning_rate": 2.5354640807770997e-05, "loss": 0.0184, "step": 54470 }, { "grad_norm": 0.35082879662513733, "learning_rate": 2.5336659736939622e-05, "loss": 0.0216, "step": 54480 }, { "grad_norm": 0.27348607778549194, "learning_rate": 2.5318682880398554e-05, "loss": 0.0239, "step": 54490 }, { "grad_norm": 0.2346682995557785, "learning_rate": 2.530071024121956e-05, "loss": 0.0256, "step": 54500 }, { "grad_norm": 0.2849014103412628, "learning_rate": 2.5282741822473627e-05, "loss": 0.0172, "step": 54510 }, { "grad_norm": 0.3423350751399994, "learning_rate": 2.526477762723114e-05, "loss": 0.0188, "step": 54520 }, { "grad_norm": 0.35498830676078796, "learning_rate": 2.5246817658561618e-05, "loss": 0.0202, "step": 54530 }, { "grad_norm": 0.2669502794742584, "learning_rate": 2.5228861919533965e-05, "loss": 0.0222, "step": 54540 }, { "grad_norm": 0.30772092938423157, "learning_rate": 2.5210910413216326e-05, "loss": 0.0345, "step": 54550 }, { "grad_norm": 0.21890012919902802, "learning_rate": 2.5192963142676086e-05, "loss": 0.0157, "step": 54560 }, { "grad_norm": 0.19963741302490234, "learning_rate": 2.517502011098001e-05, "loss": 0.0166, "step": 54570 }, { "grad_norm": 0.1567308008670807, "learning_rate": 2.5157081321193987e-05, "loss": 0.0177, "step": 54580 }, { "grad_norm": 0.34325307607650757, "learning_rate": 2.5139146776383356e-05, "loss": 0.0191, "step": 54590 }, { "grad_norm": 0.20812854170799255, "learning_rate": 2.5121216479612575e-05, "loss": 0.0169, "step": 54600 }, { "grad_norm": 0.2312985062599182, "learning_rate": 2.510329043394546e-05, "loss": 0.0352, "step": 54610 }, { "grad_norm": 0.2941783368587494, "learning_rate": 2.508536864244508e-05, "loss": 0.0218, "step": 54620 }, { "grad_norm": 0.32522010803222656, "learning_rate": 2.5067451108173778e-05, "loss": 0.0176, "step": 54630 }, { "grad_norm": 0.40626952052116394, "learning_rate": 2.5049537834193204e-05, "loss": 0.0195, "step": 54640 }, { "grad_norm": 0.29076045751571655, "learning_rate": 2.5031628823564194e-05, "loss": 0.0186, "step": 54650 }, { "grad_norm": 0.47841575741767883, "learning_rate": 2.5013724079346933e-05, "loss": 0.023, "step": 54660 }, { "grad_norm": 0.23902221024036407, "learning_rate": 2.4995823604600854e-05, "loss": 0.0145, "step": 54670 }, { "grad_norm": 0.40047532320022583, "learning_rate": 2.497792740238465e-05, "loss": 0.0189, "step": 54680 }, { "grad_norm": 0.2764413058757782, "learning_rate": 2.49600354757563e-05, "loss": 0.0201, "step": 54690 }, { "grad_norm": 0.38315194845199585, "learning_rate": 2.494214782777306e-05, "loss": 0.0192, "step": 54700 }, { "grad_norm": 0.5352681875228882, "learning_rate": 2.4924264461491386e-05, "loss": 0.0224, "step": 54710 }, { "grad_norm": 0.35549163818359375, "learning_rate": 2.4906385379967133e-05, "loss": 0.0253, "step": 54720 }, { "grad_norm": 0.32092180848121643, "learning_rate": 2.4888510586255285e-05, "loss": 0.018, "step": 54730 }, { "grad_norm": 0.23897776007652283, "learning_rate": 2.487064008341018e-05, "loss": 0.0239, "step": 54740 }, { "grad_norm": 0.32946255803108215, "learning_rate": 2.4852773874485407e-05, "loss": 0.0203, "step": 54750 }, { "grad_norm": 0.2769763469696045, "learning_rate": 2.483491196253377e-05, "loss": 0.0218, "step": 54760 }, { "grad_norm": 0.2545347511768341, "learning_rate": 2.4817054350607443e-05, "loss": 0.0202, "step": 54770 }, { "grad_norm": 0.3096960484981537, "learning_rate": 2.4799201041757742e-05, "loss": 0.0226, "step": 54780 }, { "grad_norm": 0.2989618480205536, "learning_rate": 2.4781352039035373e-05, "loss": 0.0223, "step": 54790 }, { "grad_norm": 0.258954256772995, "learning_rate": 2.4763507345490194e-05, "loss": 0.0182, "step": 54800 }, { "grad_norm": 0.3598124384880066, "learning_rate": 2.4745666964171386e-05, "loss": 0.0195, "step": 54810 }, { "grad_norm": 0.4759398102760315, "learning_rate": 2.4727830898127387e-05, "loss": 0.0211, "step": 54820 }, { "grad_norm": 0.31979140639305115, "learning_rate": 2.4709999150405895e-05, "loss": 0.0238, "step": 54830 }, { "grad_norm": 0.3552182614803314, "learning_rate": 2.469217172405388e-05, "loss": 0.0195, "step": 54840 }, { "grad_norm": 0.2327723205089569, "learning_rate": 2.4674348622117527e-05, "loss": 0.0165, "step": 54850 }, { "grad_norm": 0.7669501304626465, "learning_rate": 2.465652984764234e-05, "loss": 0.0238, "step": 54860 }, { "grad_norm": 0.3553162217140198, "learning_rate": 2.4638715403673056e-05, "loss": 0.0157, "step": 54870 }, { "grad_norm": 0.3982006013393402, "learning_rate": 2.4620905293253676e-05, "loss": 0.0195, "step": 54880 }, { "grad_norm": 0.3211422264575958, "learning_rate": 2.4603099519427463e-05, "loss": 0.0208, "step": 54890 }, { "grad_norm": 0.3139670789241791, "learning_rate": 2.458529808523695e-05, "loss": 0.0183, "step": 54900 }, { "grad_norm": 0.2976415753364563, "learning_rate": 2.456750099372387e-05, "loss": 0.0214, "step": 54910 }, { "grad_norm": 0.33531272411346436, "learning_rate": 2.4549708247929327e-05, "loss": 0.0244, "step": 54920 }, { "grad_norm": 0.24154090881347656, "learning_rate": 2.4531919850893554e-05, "loss": 0.0189, "step": 54930 }, { "grad_norm": 0.21455250680446625, "learning_rate": 2.4514135805656125e-05, "loss": 0.0165, "step": 54940 }, { "grad_norm": 0.3376013934612274, "learning_rate": 2.449635611525587e-05, "loss": 0.0207, "step": 54950 }, { "grad_norm": 0.27896761894226074, "learning_rate": 2.447858078273079e-05, "loss": 0.0229, "step": 54960 }, { "grad_norm": 0.34891772270202637, "learning_rate": 2.4460809811118275e-05, "loss": 0.0271, "step": 54970 }, { "grad_norm": 0.3543878197669983, "learning_rate": 2.444304320345483e-05, "loss": 0.019, "step": 54980 }, { "grad_norm": 0.22244609892368317, "learning_rate": 2.4425280962776354e-05, "loss": 0.0221, "step": 54990 }, { "grad_norm": 0.30655795335769653, "learning_rate": 2.4407523092117875e-05, "loss": 0.0207, "step": 55000 }, { "grad_norm": 0.5800658464431763, "learning_rate": 2.438976959451374e-05, "loss": 0.0192, "step": 55010 }, { "grad_norm": 0.30553582310676575, "learning_rate": 2.4372020472997565e-05, "loss": 0.0175, "step": 55020 }, { "grad_norm": 0.25180017948150635, "learning_rate": 2.4354275730602122e-05, "loss": 0.0201, "step": 55030 }, { "grad_norm": 0.23060651123523712, "learning_rate": 2.433653537035958e-05, "loss": 0.0195, "step": 55040 }, { "grad_norm": 0.2427147626876831, "learning_rate": 2.431879939530123e-05, "loss": 0.0185, "step": 55050 }, { "grad_norm": 0.35008013248443604, "learning_rate": 2.4301067808457684e-05, "loss": 0.0179, "step": 55060 }, { "grad_norm": 0.24733677506446838, "learning_rate": 2.428334061285878e-05, "loss": 0.0186, "step": 55070 }, { "grad_norm": 0.2579108476638794, "learning_rate": 2.426561781153361e-05, "loss": 0.019, "step": 55080 }, { "grad_norm": 0.22511065006256104, "learning_rate": 2.424789940751052e-05, "loss": 0.0167, "step": 55090 }, { "grad_norm": 0.2285076379776001, "learning_rate": 2.423018540381712e-05, "loss": 0.0233, "step": 55100 }, { "grad_norm": 0.26494261622428894, "learning_rate": 2.421247580348021e-05, "loss": 0.0199, "step": 55110 }, { "grad_norm": 0.24524249136447906, "learning_rate": 2.4194770609525897e-05, "loss": 0.0157, "step": 55120 }, { "grad_norm": 0.2535676956176758, "learning_rate": 2.417706982497951e-05, "loss": 0.0204, "step": 55130 }, { "grad_norm": 0.4944702088832855, "learning_rate": 2.415937345286563e-05, "loss": 0.0277, "step": 55140 }, { "grad_norm": 0.3719961643218994, "learning_rate": 2.4141681496208087e-05, "loss": 0.0222, "step": 55150 }, { "grad_norm": 0.2392708957195282, "learning_rate": 2.4123993958029946e-05, "loss": 0.0155, "step": 55160 }, { "grad_norm": 0.6669548153877258, "learning_rate": 2.4106310841353548e-05, "loss": 0.0184, "step": 55170 }, { "grad_norm": 0.2224811613559723, "learning_rate": 2.4088632149200398e-05, "loss": 0.0261, "step": 55180 }, { "grad_norm": 0.21530357003211975, "learning_rate": 2.4070957884591367e-05, "loss": 0.0163, "step": 55190 }, { "grad_norm": 0.4741274416446686, "learning_rate": 2.4053288050546464e-05, "loss": 0.0237, "step": 55200 }, { "grad_norm": 0.21380458772182465, "learning_rate": 2.403562265008498e-05, "loss": 0.0233, "step": 55210 }, { "grad_norm": 0.3639364242553711, "learning_rate": 2.4017961686225483e-05, "loss": 0.0216, "step": 55220 }, { "grad_norm": 0.25945359468460083, "learning_rate": 2.400030516198568e-05, "loss": 0.025, "step": 55230 }, { "grad_norm": 0.2261158674955368, "learning_rate": 2.3982653080382673e-05, "loss": 0.0208, "step": 55240 }, { "grad_norm": 0.30598756670951843, "learning_rate": 2.396500544443266e-05, "loss": 0.0204, "step": 55250 }, { "grad_norm": 0.6207901239395142, "learning_rate": 2.3947362257151156e-05, "loss": 0.018, "step": 55260 }, { "grad_norm": 0.34505510330200195, "learning_rate": 2.39297235215529e-05, "loss": 0.0243, "step": 55270 }, { "grad_norm": 0.44295063614845276, "learning_rate": 2.3912089240651873e-05, "loss": 0.0283, "step": 55280 }, { "grad_norm": 0.23090402781963348, "learning_rate": 2.389445941746129e-05, "loss": 0.0247, "step": 55290 }, { "grad_norm": 0.2968457341194153, "learning_rate": 2.3876834054993625e-05, "loss": 0.0205, "step": 55300 }, { "grad_norm": 0.209225133061409, "learning_rate": 2.3859213156260522e-05, "loss": 0.018, "step": 55310 }, { "grad_norm": 0.2404959350824356, "learning_rate": 2.3841596724272948e-05, "loss": 0.0189, "step": 55320 }, { "grad_norm": 0.2631383538246155, "learning_rate": 2.382398476204106e-05, "loss": 0.0211, "step": 55330 }, { "grad_norm": 0.2843589782714844, "learning_rate": 2.3806377272574254e-05, "loss": 0.0157, "step": 55340 }, { "grad_norm": 0.3558405339717865, "learning_rate": 2.3788774258881174e-05, "loss": 0.0211, "step": 55350 }, { "grad_norm": 0.6377124190330505, "learning_rate": 2.37711757239697e-05, "loss": 0.0248, "step": 55360 }, { "grad_norm": 0.3165874183177948, "learning_rate": 2.3753581670846954e-05, "loss": 0.0129, "step": 55370 }, { "grad_norm": 0.15375366806983948, "learning_rate": 2.3735992102519216e-05, "loss": 0.0166, "step": 55380 }, { "grad_norm": 0.3953658938407898, "learning_rate": 2.371840702199215e-05, "loss": 0.0193, "step": 55390 }, { "grad_norm": 0.3225082755088806, "learning_rate": 2.3700826432270494e-05, "loss": 0.0157, "step": 55400 }, { "grad_norm": 0.3334067463874817, "learning_rate": 2.3683250336358326e-05, "loss": 0.0255, "step": 55410 }, { "grad_norm": 0.1962735503911972, "learning_rate": 2.3665678737258923e-05, "loss": 0.0255, "step": 55420 }, { "grad_norm": 0.2659006416797638, "learning_rate": 2.3648111637974745e-05, "loss": 0.0205, "step": 55430 }, { "grad_norm": 0.23650991916656494, "learning_rate": 2.36305490415076e-05, "loss": 0.0192, "step": 55440 }, { "grad_norm": 0.44771480560302734, "learning_rate": 2.3612990950858392e-05, "loss": 0.0242, "step": 55450 }, { "grad_norm": 0.23912270367145538, "learning_rate": 2.359543736902735e-05, "loss": 0.0208, "step": 55460 }, { "grad_norm": 0.3025442957878113, "learning_rate": 2.3577888299013896e-05, "loss": 0.0177, "step": 55470 }, { "grad_norm": 0.29559049010276794, "learning_rate": 2.3560343743816683e-05, "loss": 0.0196, "step": 55480 }, { "grad_norm": 0.33558422327041626, "learning_rate": 2.354280370643362e-05, "loss": 0.0271, "step": 55490 }, { "grad_norm": 0.384222149848938, "learning_rate": 2.3525268189861777e-05, "loss": 0.0182, "step": 55500 }, { "grad_norm": 0.38863879442214966, "learning_rate": 2.3507737197097513e-05, "loss": 0.0169, "step": 55510 }, { "grad_norm": 0.4753800630569458, "learning_rate": 2.34902107311364e-05, "loss": 0.0233, "step": 55520 }, { "grad_norm": 0.23648644983768463, "learning_rate": 2.3472688794973225e-05, "loss": 0.0209, "step": 55530 }, { "grad_norm": 0.27366286516189575, "learning_rate": 2.3455171391602016e-05, "loss": 0.0207, "step": 55540 }, { "grad_norm": 0.40983137488365173, "learning_rate": 2.343765852401601e-05, "loss": 0.0143, "step": 55550 }, { "grad_norm": 0.26699191331863403, "learning_rate": 2.342015019520768e-05, "loss": 0.0212, "step": 55560 }, { "grad_norm": 0.1888400763273239, "learning_rate": 2.3402646408168742e-05, "loss": 0.017, "step": 55570 }, { "grad_norm": 0.3715038299560547, "learning_rate": 2.3385147165890074e-05, "loss": 0.0241, "step": 55580 }, { "grad_norm": 0.27623769640922546, "learning_rate": 2.336765247136184e-05, "loss": 0.0184, "step": 55590 }, { "grad_norm": 0.39898815751075745, "learning_rate": 2.33501623275734e-05, "loss": 0.0309, "step": 55600 }, { "grad_norm": 0.2750771939754486, "learning_rate": 2.333267673751334e-05, "loss": 0.0206, "step": 55610 }, { "grad_norm": 0.30137380957603455, "learning_rate": 2.33151957041695e-05, "loss": 0.0203, "step": 55620 }, { "grad_norm": 0.25982218980789185, "learning_rate": 2.329771923052884e-05, "loss": 0.0199, "step": 55630 }, { "grad_norm": 0.6069042682647705, "learning_rate": 2.3280247319577697e-05, "loss": 0.0201, "step": 55640 }, { "grad_norm": 0.4346301853656769, "learning_rate": 2.3262779974301473e-05, "loss": 0.0275, "step": 55650 }, { "grad_norm": 0.34255772829055786, "learning_rate": 2.3245317197684895e-05, "loss": 0.0203, "step": 55660 }, { "grad_norm": 0.3190346658229828, "learning_rate": 2.322785899271188e-05, "loss": 0.0183, "step": 55670 }, { "grad_norm": 0.2145766317844391, "learning_rate": 2.3210405362365507e-05, "loss": 0.0237, "step": 55680 }, { "grad_norm": 0.2831368148326874, "learning_rate": 2.31929563096282e-05, "loss": 0.0256, "step": 55690 }, { "grad_norm": 0.25846901535987854, "learning_rate": 2.317551183748146e-05, "loss": 0.0186, "step": 55700 }, { "grad_norm": 0.19690780341625214, "learning_rate": 2.3158071948906103e-05, "loss": 0.0186, "step": 55710 }, { "grad_norm": 0.300209105014801, "learning_rate": 2.314063664688212e-05, "loss": 0.0163, "step": 55720 }, { "grad_norm": 0.2957000136375427, "learning_rate": 2.3123205934388725e-05, "loss": 0.018, "step": 55730 }, { "grad_norm": 0.2109033763408661, "learning_rate": 2.310577981440436e-05, "loss": 0.0236, "step": 55740 }, { "grad_norm": 0.38496339321136475, "learning_rate": 2.3088358289906665e-05, "loss": 0.0171, "step": 55750 }, { "grad_norm": 0.28212496638298035, "learning_rate": 2.307094136387252e-05, "loss": 0.0168, "step": 55760 }, { "grad_norm": 0.28389742970466614, "learning_rate": 2.305352903927796e-05, "loss": 0.0257, "step": 55770 }, { "grad_norm": 0.30146166682243347, "learning_rate": 2.303612131909831e-05, "loss": 0.0174, "step": 55780 }, { "grad_norm": 0.27036893367767334, "learning_rate": 2.3018718206308054e-05, "loss": 0.0259, "step": 55790 }, { "grad_norm": 0.34864503145217896, "learning_rate": 2.3001319703880925e-05, "loss": 0.0222, "step": 55800 }, { "grad_norm": 0.37466856837272644, "learning_rate": 2.2983925814789835e-05, "loss": 0.0174, "step": 55810 }, { "grad_norm": 0.28129807114601135, "learning_rate": 2.2966536542006957e-05, "loss": 0.0189, "step": 55820 }, { "grad_norm": 0.2887532114982605, "learning_rate": 2.294915188850358e-05, "loss": 0.023, "step": 55830 }, { "grad_norm": 0.22083039581775665, "learning_rate": 2.2931771857250333e-05, "loss": 0.0226, "step": 55840 }, { "grad_norm": 0.15431855618953705, "learning_rate": 2.2914396451216946e-05, "loss": 0.0196, "step": 55850 }, { "grad_norm": 0.34401676058769226, "learning_rate": 2.2897025673372412e-05, "loss": 0.019, "step": 55860 }, { "grad_norm": 0.25317829847335815, "learning_rate": 2.287965952668494e-05, "loss": 0.0209, "step": 55870 }, { "grad_norm": 0.24765925109386444, "learning_rate": 2.2862298014121873e-05, "loss": 0.02, "step": 55880 }, { "grad_norm": 0.3978746235370636, "learning_rate": 2.28449411386499e-05, "loss": 0.0164, "step": 55890 }, { "grad_norm": 0.17498590052127838, "learning_rate": 2.282758890323477e-05, "loss": 0.0189, "step": 55900 }, { "grad_norm": 0.16055946052074432, "learning_rate": 2.2810241310841528e-05, "loss": 0.0211, "step": 55910 }, { "grad_norm": 0.20481665432453156, "learning_rate": 2.2792898364434412e-05, "loss": 0.0144, "step": 55920 }, { "grad_norm": 0.30108630657196045, "learning_rate": 2.2775560066976843e-05, "loss": 0.0248, "step": 55930 }, { "grad_norm": 0.268616259098053, "learning_rate": 2.275822642143147e-05, "loss": 0.019, "step": 55940 }, { "grad_norm": 1.453619360923767, "learning_rate": 2.274089743076014e-05, "loss": 0.0209, "step": 55950 }, { "grad_norm": 0.3419024646282196, "learning_rate": 2.2723573097923917e-05, "loss": 0.0196, "step": 55960 }, { "grad_norm": 0.23970194160938263, "learning_rate": 2.2706253425883024e-05, "loss": 0.024, "step": 55970 }, { "grad_norm": 0.4795912504196167, "learning_rate": 2.2688938417596933e-05, "loss": 0.0188, "step": 55980 }, { "grad_norm": 0.15405823290348053, "learning_rate": 2.2671628076024305e-05, "loss": 0.0177, "step": 55990 }, { "grad_norm": 0.3179848790168762, "learning_rate": 2.2654322404123008e-05, "loss": 0.0169, "step": 56000 }, { "grad_norm": 0.21699434518814087, "learning_rate": 2.2637021404850105e-05, "loss": 0.0218, "step": 56010 }, { "grad_norm": 0.3141169548034668, "learning_rate": 2.2619725081161885e-05, "loss": 0.0194, "step": 56020 }, { "grad_norm": 0.18176312744617462, "learning_rate": 2.260243343601376e-05, "loss": 0.0162, "step": 56030 }, { "grad_norm": 0.32368507981300354, "learning_rate": 2.2585146472360473e-05, "loss": 0.0199, "step": 56040 }, { "grad_norm": 0.2526595890522003, "learning_rate": 2.2567864193155834e-05, "loss": 0.0155, "step": 56050 }, { "grad_norm": 0.25350940227508545, "learning_rate": 2.255058660135294e-05, "loss": 0.023, "step": 56060 }, { "grad_norm": 0.5235154628753662, "learning_rate": 2.253331369990407e-05, "loss": 0.0268, "step": 56070 }, { "grad_norm": 0.1554935872554779, "learning_rate": 2.251604549176063e-05, "loss": 0.0239, "step": 56080 }, { "grad_norm": 0.2518109679222107, "learning_rate": 2.249878197987337e-05, "loss": 0.0194, "step": 56090 }, { "grad_norm": 0.27096205949783325, "learning_rate": 2.2481523167192087e-05, "loss": 0.0175, "step": 56100 }, { "grad_norm": 0.2444128394126892, "learning_rate": 2.2464269056665867e-05, "loss": 0.0198, "step": 56110 }, { "grad_norm": 0.19742408394813538, "learning_rate": 2.2447019651242958e-05, "loss": 0.0166, "step": 56120 }, { "grad_norm": 0.39579030871391296, "learning_rate": 2.2429774953870818e-05, "loss": 0.019, "step": 56130 }, { "grad_norm": 0.3547098934650421, "learning_rate": 2.241253496749611e-05, "loss": 0.0195, "step": 56140 }, { "grad_norm": 0.27533984184265137, "learning_rate": 2.2395299695064614e-05, "loss": 0.016, "step": 56150 }, { "grad_norm": 0.3788303732872009, "learning_rate": 2.237806913952145e-05, "loss": 0.0207, "step": 56160 }, { "grad_norm": 0.2913672626018524, "learning_rate": 2.2360843303810798e-05, "loss": 0.0141, "step": 56170 }, { "grad_norm": 0.2809270918369293, "learning_rate": 2.2343622190876084e-05, "loss": 0.0186, "step": 56180 }, { "grad_norm": 0.4021417200565338, "learning_rate": 2.2326405803659935e-05, "loss": 0.0191, "step": 56190 }, { "grad_norm": 0.23225438594818115, "learning_rate": 2.230919414510416e-05, "loss": 0.0161, "step": 56200 }, { "grad_norm": 0.3179782032966614, "learning_rate": 2.229198721814976e-05, "loss": 0.0211, "step": 56210 }, { "grad_norm": 0.22960342466831207, "learning_rate": 2.2274785025736948e-05, "loss": 0.018, "step": 56220 }, { "grad_norm": 0.29084521532058716, "learning_rate": 2.225758757080507e-05, "loss": 0.02, "step": 56230 }, { "grad_norm": 0.3832225799560547, "learning_rate": 2.2240394856292723e-05, "loss": 0.0198, "step": 56240 }, { "grad_norm": 0.21322819590568542, "learning_rate": 2.2223206885137664e-05, "loss": 0.0179, "step": 56250 }, { "grad_norm": 0.40625566244125366, "learning_rate": 2.2206023660276853e-05, "loss": 0.0165, "step": 56260 }, { "grad_norm": 0.35196274518966675, "learning_rate": 2.218884518464645e-05, "loss": 0.0201, "step": 56270 }, { "grad_norm": 0.3871203660964966, "learning_rate": 2.2171671461181732e-05, "loss": 0.0181, "step": 56280 }, { "grad_norm": 0.270929753780365, "learning_rate": 2.2154502492817292e-05, "loss": 0.0202, "step": 56290 }, { "grad_norm": 0.2690573036670685, "learning_rate": 2.2137338282486782e-05, "loss": 0.0145, "step": 56300 }, { "grad_norm": 0.13534410297870636, "learning_rate": 2.2120178833123113e-05, "loss": 0.0172, "step": 56310 }, { "grad_norm": 0.3952072858810425, "learning_rate": 2.210302414765838e-05, "loss": 0.0216, "step": 56320 }, { "grad_norm": 0.22431953251361847, "learning_rate": 2.2085874229023808e-05, "loss": 0.0137, "step": 56330 }, { "grad_norm": 0.23665687441825867, "learning_rate": 2.2068729080149907e-05, "loss": 0.0168, "step": 56340 }, { "grad_norm": 0.4256700873374939, "learning_rate": 2.205158870396625e-05, "loss": 0.024, "step": 56350 }, { "grad_norm": 0.2090609222650528, "learning_rate": 2.2034453103401732e-05, "loss": 0.018, "step": 56360 }, { "grad_norm": 0.15831075608730316, "learning_rate": 2.20173222813843e-05, "loss": 0.012, "step": 56370 }, { "grad_norm": 0.23555666208267212, "learning_rate": 2.200019624084116e-05, "loss": 0.0193, "step": 56380 }, { "grad_norm": 0.1976366639137268, "learning_rate": 2.1983074984698687e-05, "loss": 0.016, "step": 56390 }, { "grad_norm": 0.39273157715797424, "learning_rate": 2.1965958515882433e-05, "loss": 0.0232, "step": 56400 }, { "grad_norm": 0.34590771794319153, "learning_rate": 2.1948846837317162e-05, "loss": 0.0183, "step": 56410 }, { "grad_norm": 0.26531684398651123, "learning_rate": 2.1931739951926738e-05, "loss": 0.0144, "step": 56420 }, { "grad_norm": 0.22449618577957153, "learning_rate": 2.191463786263429e-05, "loss": 0.0147, "step": 56430 }, { "grad_norm": 0.2619568705558777, "learning_rate": 2.1897540572362095e-05, "loss": 0.0173, "step": 56440 }, { "grad_norm": 0.2166738510131836, "learning_rate": 2.1880448084031614e-05, "loss": 0.0161, "step": 56450 }, { "grad_norm": 0.381616473197937, "learning_rate": 2.1863360400563482e-05, "loss": 0.0183, "step": 56460 }, { "grad_norm": 0.3570810854434967, "learning_rate": 2.184627752487754e-05, "loss": 0.0225, "step": 56470 }, { "grad_norm": 0.20380379259586334, "learning_rate": 2.1829199459892725e-05, "loss": 0.018, "step": 56480 }, { "grad_norm": 0.20739246904850006, "learning_rate": 2.1812126208527282e-05, "loss": 0.0206, "step": 56490 }, { "grad_norm": 0.17624378204345703, "learning_rate": 2.179505777369852e-05, "loss": 0.0247, "step": 56500 }, { "grad_norm": 0.5964787602424622, "learning_rate": 2.1777994158322974e-05, "loss": 0.0216, "step": 56510 }, { "grad_norm": 0.26133206486701965, "learning_rate": 2.176093536531637e-05, "loss": 0.0218, "step": 56520 }, { "grad_norm": 0.19566532969474792, "learning_rate": 2.1743881397593537e-05, "loss": 0.0181, "step": 56530 }, { "grad_norm": 0.349357008934021, "learning_rate": 2.1726832258068595e-05, "loss": 0.0263, "step": 56540 }, { "grad_norm": 0.4240031838417053, "learning_rate": 2.170978794965472e-05, "loss": 0.019, "step": 56550 }, { "grad_norm": 0.662280261516571, "learning_rate": 2.169274847526438e-05, "loss": 0.0164, "step": 56560 }, { "grad_norm": 0.2790379524230957, "learning_rate": 2.1675713837809103e-05, "loss": 0.0183, "step": 56570 }, { "grad_norm": 0.25822147727012634, "learning_rate": 2.1658684040199655e-05, "loss": 0.014, "step": 56580 }, { "grad_norm": 0.3416319489479065, "learning_rate": 2.1641659085345974e-05, "loss": 0.0153, "step": 56590 }, { "grad_norm": 0.24391977488994598, "learning_rate": 2.1624638976157154e-05, "loss": 0.0216, "step": 56600 }, { "grad_norm": 0.34910619258880615, "learning_rate": 2.1607623715541476e-05, "loss": 0.0265, "step": 56610 }, { "grad_norm": 0.25624606013298035, "learning_rate": 2.159061330640636e-05, "loss": 0.0174, "step": 56620 }, { "grad_norm": 0.3315650522708893, "learning_rate": 2.1573607751658425e-05, "loss": 0.015, "step": 56630 }, { "grad_norm": 0.2618037462234497, "learning_rate": 2.155660705420347e-05, "loss": 0.0202, "step": 56640 }, { "grad_norm": 0.293500691652298, "learning_rate": 2.153961121694644e-05, "loss": 0.0214, "step": 56650 }, { "grad_norm": 0.2596195340156555, "learning_rate": 2.152262024279145e-05, "loss": 0.0268, "step": 56660 }, { "grad_norm": 0.5724645853042603, "learning_rate": 2.150563413464183e-05, "loss": 0.019, "step": 56670 }, { "grad_norm": 0.18135951459407806, "learning_rate": 2.1488652895399974e-05, "loss": 0.0154, "step": 56680 }, { "grad_norm": 0.48932215571403503, "learning_rate": 2.1471676527967587e-05, "loss": 0.0206, "step": 56690 }, { "grad_norm": 0.2888866066932678, "learning_rate": 2.1454705035245414e-05, "loss": 0.0237, "step": 56700 }, { "grad_norm": 0.22892123460769653, "learning_rate": 2.143773842013343e-05, "loss": 0.0173, "step": 56710 }, { "grad_norm": 0.2240576297044754, "learning_rate": 2.1420776685530796e-05, "loss": 0.0176, "step": 56720 }, { "grad_norm": 0.3525671064853668, "learning_rate": 2.1403819834335742e-05, "loss": 0.0155, "step": 56730 }, { "grad_norm": 0.21313735842704773, "learning_rate": 2.1386867869445808e-05, "loss": 0.0217, "step": 56740 }, { "grad_norm": 0.273802787065506, "learning_rate": 2.1369920793757548e-05, "loss": 0.0224, "step": 56750 }, { "grad_norm": 0.4302075207233429, "learning_rate": 2.1352978610166828e-05, "loss": 0.0173, "step": 56760 }, { "grad_norm": 0.3400571644306183, "learning_rate": 2.1336041321568546e-05, "loss": 0.0181, "step": 56770 }, { "grad_norm": 0.21857407689094543, "learning_rate": 2.131910893085684e-05, "loss": 0.0173, "step": 56780 }, { "grad_norm": 0.23704268038272858, "learning_rate": 2.1302181440925012e-05, "loss": 0.0188, "step": 56790 }, { "grad_norm": 0.27580365538597107, "learning_rate": 2.128525885466546e-05, "loss": 0.0151, "step": 56800 }, { "grad_norm": 0.23445501923561096, "learning_rate": 2.1268341174969847e-05, "loss": 0.0183, "step": 56810 }, { "grad_norm": 0.2754036486148834, "learning_rate": 2.1251428404728894e-05, "loss": 0.0155, "step": 56820 }, { "grad_norm": 0.41763803362846375, "learning_rate": 2.123452054683256e-05, "loss": 0.0212, "step": 56830 }, { "grad_norm": 0.4577544033527374, "learning_rate": 2.1217617604169927e-05, "loss": 0.0219, "step": 56840 }, { "grad_norm": 0.3787376284599304, "learning_rate": 2.120071957962924e-05, "loss": 0.0183, "step": 56850 }, { "grad_norm": 0.30765455961227417, "learning_rate": 2.1183826476097917e-05, "loss": 0.0231, "step": 56860 }, { "grad_norm": 0.2904318869113922, "learning_rate": 2.116693829646254e-05, "loss": 0.0253, "step": 56870 }, { "grad_norm": 0.298929363489151, "learning_rate": 2.1150055043608806e-05, "loss": 0.018, "step": 56880 }, { "grad_norm": 0.4459921419620514, "learning_rate": 2.113317672042162e-05, "loss": 0.0255, "step": 56890 }, { "grad_norm": 0.2789374887943268, "learning_rate": 2.1116303329785025e-05, "loss": 0.0158, "step": 56900 }, { "grad_norm": 0.26063603162765503, "learning_rate": 2.109943487458222e-05, "loss": 0.022, "step": 56910 }, { "grad_norm": 0.2514863908290863, "learning_rate": 2.1082571357695574e-05, "loss": 0.0149, "step": 56920 }, { "grad_norm": 0.26285693049430847, "learning_rate": 2.1065712782006557e-05, "loss": 0.0203, "step": 56930 }, { "grad_norm": 0.1314210295677185, "learning_rate": 2.104885915039591e-05, "loss": 0.0139, "step": 56940 }, { "grad_norm": 0.26138415932655334, "learning_rate": 2.103201046574338e-05, "loss": 0.0154, "step": 56950 }, { "grad_norm": 0.319856733083725, "learning_rate": 2.1015166730928022e-05, "loss": 0.0151, "step": 56960 }, { "grad_norm": 0.48283255100250244, "learning_rate": 2.0998327948827912e-05, "loss": 0.0222, "step": 56970 }, { "grad_norm": 0.24697022140026093, "learning_rate": 2.0981494122320363e-05, "loss": 0.0183, "step": 56980 }, { "grad_norm": 0.3874140977859497, "learning_rate": 2.0964665254281822e-05, "loss": 0.0208, "step": 56990 }, { "grad_norm": 0.24065841734409332, "learning_rate": 2.094784134758784e-05, "loss": 0.0161, "step": 57000 }, { "grad_norm": 0.23216721415519714, "learning_rate": 2.0931022405113226e-05, "loss": 0.0274, "step": 57010 }, { "grad_norm": 0.30544474720954895, "learning_rate": 2.091420842973183e-05, "loss": 0.018, "step": 57020 }, { "grad_norm": 0.1715836077928543, "learning_rate": 2.0897399424316715e-05, "loss": 0.0122, "step": 57030 }, { "grad_norm": 0.40309956669807434, "learning_rate": 2.0880595391740078e-05, "loss": 0.0178, "step": 57040 }, { "grad_norm": 0.20937681198120117, "learning_rate": 2.0863796334873277e-05, "loss": 0.0166, "step": 57050 }, { "grad_norm": 0.24073854088783264, "learning_rate": 2.08470022565868e-05, "loss": 0.0185, "step": 57060 }, { "grad_norm": 0.24705985188484192, "learning_rate": 2.0830213159750317e-05, "loss": 0.0141, "step": 57070 }, { "grad_norm": 0.626124918460846, "learning_rate": 2.0813429047232596e-05, "loss": 0.0173, "step": 57080 }, { "grad_norm": 0.17303821444511414, "learning_rate": 2.0796649921901594e-05, "loss": 0.0166, "step": 57090 }, { "grad_norm": 0.2294790893793106, "learning_rate": 2.077987578662441e-05, "loss": 0.0153, "step": 57100 }, { "grad_norm": 0.27669757604599, "learning_rate": 2.0763106644267277e-05, "loss": 0.0212, "step": 57110 }, { "grad_norm": 0.2562471926212311, "learning_rate": 2.0746342497695607e-05, "loss": 0.0124, "step": 57120 }, { "grad_norm": 1.4786940813064575, "learning_rate": 2.0729583349773886e-05, "loss": 0.0164, "step": 57130 }, { "grad_norm": 0.37698620557785034, "learning_rate": 2.0712829203365853e-05, "loss": 0.0215, "step": 57140 }, { "grad_norm": 0.31940317153930664, "learning_rate": 2.0696080061334267e-05, "loss": 0.0195, "step": 57150 }, { "grad_norm": 0.24025532603263855, "learning_rate": 2.067933592654117e-05, "loss": 0.0176, "step": 57160 }, { "grad_norm": 0.3034813702106476, "learning_rate": 2.066259680184763e-05, "loss": 0.0182, "step": 57170 }, { "grad_norm": 0.22449292242527008, "learning_rate": 2.0645862690113908e-05, "loss": 0.0148, "step": 57180 }, { "grad_norm": 0.3494115173816681, "learning_rate": 2.0629133594199436e-05, "loss": 0.0276, "step": 57190 }, { "grad_norm": 0.2155926674604416, "learning_rate": 2.0612409516962704e-05, "loss": 0.0153, "step": 57200 }, { "grad_norm": 0.31690850853919983, "learning_rate": 2.0595690461261467e-05, "loss": 0.0199, "step": 57210 }, { "grad_norm": 0.23454450070858002, "learning_rate": 2.0578976429952503e-05, "loss": 0.0175, "step": 57220 }, { "grad_norm": 0.2957124710083008, "learning_rate": 2.0562267425891802e-05, "loss": 0.0192, "step": 57230 }, { "grad_norm": 0.2862665057182312, "learning_rate": 2.0545563451934467e-05, "loss": 0.0177, "step": 57240 }, { "grad_norm": 0.21835780143737793, "learning_rate": 2.0528864510934764e-05, "loss": 0.0194, "step": 57250 }, { "grad_norm": 0.1869666427373886, "learning_rate": 2.0512170605746096e-05, "loss": 0.0274, "step": 57260 }, { "grad_norm": 0.2619611620903015, "learning_rate": 2.0495481739220963e-05, "loss": 0.018, "step": 57270 }, { "grad_norm": 0.48202717304229736, "learning_rate": 2.0478797914211045e-05, "loss": 0.0217, "step": 57280 }, { "grad_norm": 0.20023798942565918, "learning_rate": 2.046211913356716e-05, "loss": 0.0229, "step": 57290 }, { "grad_norm": 0.4816255271434784, "learning_rate": 2.0445445400139247e-05, "loss": 0.0245, "step": 57300 }, { "grad_norm": 0.39606377482414246, "learning_rate": 2.0428776716776405e-05, "loss": 0.017, "step": 57310 }, { "grad_norm": 0.3680092394351959, "learning_rate": 2.0412113086326856e-05, "loss": 0.0197, "step": 57320 }, { "grad_norm": 0.34527045488357544, "learning_rate": 2.0395454511637918e-05, "loss": 0.0142, "step": 57330 }, { "grad_norm": 0.3955511450767517, "learning_rate": 2.037880099555616e-05, "loss": 0.0176, "step": 57340 }, { "grad_norm": 0.28757455945014954, "learning_rate": 2.0362152540927144e-05, "loss": 0.0184, "step": 57350 }, { "grad_norm": 0.27160125970840454, "learning_rate": 2.0345509150595666e-05, "loss": 0.0265, "step": 57360 }, { "grad_norm": 0.29266664385795593, "learning_rate": 2.0328870827405617e-05, "loss": 0.0185, "step": 57370 }, { "grad_norm": 0.21552181243896484, "learning_rate": 2.0312237574200043e-05, "loss": 0.0176, "step": 57380 }, { "grad_norm": 0.25913742184638977, "learning_rate": 2.029560939382112e-05, "loss": 0.0166, "step": 57390 }, { "grad_norm": 0.3250068128108978, "learning_rate": 2.0278986289110097e-05, "loss": 0.0195, "step": 57400 }, { "grad_norm": 0.2172064632177353, "learning_rate": 2.0262368262907484e-05, "loss": 0.0191, "step": 57410 }, { "grad_norm": 0.3111938536167145, "learning_rate": 2.024575531805279e-05, "loss": 0.0217, "step": 57420 }, { "grad_norm": 0.21961857378482819, "learning_rate": 2.0229147457384735e-05, "loss": 0.0192, "step": 57430 }, { "grad_norm": 0.33651646971702576, "learning_rate": 2.0212544683741157e-05, "loss": 0.028, "step": 57440 }, { "grad_norm": 0.3349016010761261, "learning_rate": 2.0195946999958976e-05, "loss": 0.0278, "step": 57450 }, { "grad_norm": 0.7105945944786072, "learning_rate": 2.017935440887434e-05, "loss": 0.0222, "step": 57460 }, { "grad_norm": 0.34678635001182556, "learning_rate": 2.0162766913322423e-05, "loss": 0.0182, "step": 57470 }, { "grad_norm": 0.19521012902259827, "learning_rate": 2.0146184516137588e-05, "loss": 0.0213, "step": 57480 }, { "grad_norm": 0.17917785048484802, "learning_rate": 2.012960722015332e-05, "loss": 0.02, "step": 57490 }, { "grad_norm": 0.21070173382759094, "learning_rate": 2.0113035028202214e-05, "loss": 0.0263, "step": 57500 }, { "grad_norm": 0.3327701687812805, "learning_rate": 2.009646794311602e-05, "loss": 0.0198, "step": 57510 }, { "grad_norm": 0.27380138635635376, "learning_rate": 2.007990596772559e-05, "loss": 0.0209, "step": 57520 }, { "grad_norm": 0.16567663848400116, "learning_rate": 2.0063349104860923e-05, "loss": 0.0176, "step": 57530 }, { "grad_norm": 0.27417266368865967, "learning_rate": 2.0046797357351116e-05, "loss": 0.0231, "step": 57540 }, { "grad_norm": 0.16334408521652222, "learning_rate": 2.0030250728024412e-05, "loss": 0.0138, "step": 57550 }, { "grad_norm": 0.23694077134132385, "learning_rate": 2.001370921970819e-05, "loss": 0.0136, "step": 57560 }, { "grad_norm": 0.17177928984165192, "learning_rate": 1.9997172835228932e-05, "loss": 0.0152, "step": 57570 }, { "grad_norm": 0.27575862407684326, "learning_rate": 1.9980641577412262e-05, "loss": 0.015, "step": 57580 }, { "grad_norm": 0.394377201795578, "learning_rate": 1.9964115449082925e-05, "loss": 0.0187, "step": 57590 }, { "grad_norm": 0.2511231005191803, "learning_rate": 1.9947594453064742e-05, "loss": 0.0201, "step": 57600 }, { "grad_norm": 0.23485590517520905, "learning_rate": 1.9931078592180774e-05, "loss": 0.0169, "step": 57610 }, { "grad_norm": 0.26320669054985046, "learning_rate": 1.9914567869253065e-05, "loss": 0.0184, "step": 57620 }, { "grad_norm": 0.15364228188991547, "learning_rate": 1.989806228710287e-05, "loss": 0.0247, "step": 57630 }, { "grad_norm": 0.19796344637870789, "learning_rate": 1.9881561848550555e-05, "loss": 0.0148, "step": 57640 }, { "grad_norm": 0.21151845157146454, "learning_rate": 1.9865066556415544e-05, "loss": 0.0224, "step": 57650 }, { "grad_norm": 0.2699463367462158, "learning_rate": 1.98485764135165e-05, "loss": 0.0174, "step": 57660 }, { "grad_norm": 0.1757328063249588, "learning_rate": 1.983209142267109e-05, "loss": 0.0142, "step": 57670 }, { "grad_norm": 0.2638110816478729, "learning_rate": 1.9815611586696165e-05, "loss": 0.0157, "step": 57680 }, { "grad_norm": 0.26389074325561523, "learning_rate": 1.9799136908407667e-05, "loss": 0.0204, "step": 57690 }, { "grad_norm": 0.3567333519458771, "learning_rate": 1.9782667390620678e-05, "loss": 0.0189, "step": 57700 }, { "grad_norm": 0.19078461825847626, "learning_rate": 1.976620303614939e-05, "loss": 0.0184, "step": 57710 }, { "grad_norm": 0.25256097316741943, "learning_rate": 1.9749743847807108e-05, "loss": 0.0254, "step": 57720 }, { "grad_norm": 0.32394495606422424, "learning_rate": 1.9733289828406272e-05, "loss": 0.0144, "step": 57730 }, { "grad_norm": 0.2738199830055237, "learning_rate": 1.9716840980758382e-05, "loss": 0.0211, "step": 57740 }, { "grad_norm": 0.24743054807186127, "learning_rate": 1.9700397307674134e-05, "loss": 0.0213, "step": 57750 }, { "grad_norm": 0.2043345868587494, "learning_rate": 1.968395881196328e-05, "loss": 0.0131, "step": 57760 }, { "grad_norm": 0.4945933222770691, "learning_rate": 1.966752549643473e-05, "loss": 0.0202, "step": 57770 }, { "grad_norm": 0.18987639248371124, "learning_rate": 1.965109736389647e-05, "loss": 0.0175, "step": 57780 }, { "grad_norm": 0.290530800819397, "learning_rate": 1.9634674417155645e-05, "loss": 0.0159, "step": 57790 }, { "grad_norm": 0.25754377245903015, "learning_rate": 1.9618256659018434e-05, "loss": 0.0351, "step": 57800 }, { "grad_norm": 0.33791980147361755, "learning_rate": 1.9601844092290257e-05, "loss": 0.0209, "step": 57810 }, { "grad_norm": 0.38199612498283386, "learning_rate": 1.9585436719775512e-05, "loss": 0.0196, "step": 57820 }, { "grad_norm": 0.517098605632782, "learning_rate": 1.9569034544277793e-05, "loss": 0.022, "step": 57830 }, { "grad_norm": 0.4262183904647827, "learning_rate": 1.9552637568599798e-05, "loss": 0.0182, "step": 57840 }, { "grad_norm": 0.2266940474510193, "learning_rate": 1.953624579554327e-05, "loss": 0.0224, "step": 57850 }, { "grad_norm": 0.23061053454875946, "learning_rate": 1.951985922790918e-05, "loss": 0.024, "step": 57860 }, { "grad_norm": 0.2100294530391693, "learning_rate": 1.9503477868497505e-05, "loss": 0.0122, "step": 57870 }, { "grad_norm": 0.4839475154876709, "learning_rate": 1.9487101720107375e-05, "loss": 0.0233, "step": 57880 }, { "grad_norm": 0.29257115721702576, "learning_rate": 1.9470730785537032e-05, "loss": 0.0175, "step": 57890 }, { "grad_norm": 0.16500116884708405, "learning_rate": 1.9454365067583823e-05, "loss": 0.0197, "step": 57900 }, { "grad_norm": 0.23219381272792816, "learning_rate": 1.9438004569044215e-05, "loss": 0.0146, "step": 57910 }, { "grad_norm": 0.6744224429130554, "learning_rate": 1.9421649292713724e-05, "loss": 0.0327, "step": 57920 }, { "grad_norm": 0.2716515362262726, "learning_rate": 1.9405299241387076e-05, "loss": 0.0189, "step": 57930 }, { "grad_norm": 0.25572678446769714, "learning_rate": 1.9388954417858007e-05, "loss": 0.0143, "step": 57940 }, { "grad_norm": 0.2966194152832031, "learning_rate": 1.9372614824919417e-05, "loss": 0.0193, "step": 57950 }, { "grad_norm": 0.3421739935874939, "learning_rate": 1.9356280465363284e-05, "loss": 0.0217, "step": 57960 }, { "grad_norm": 0.260947585105896, "learning_rate": 1.9339951341980723e-05, "loss": 0.0218, "step": 57970 }, { "grad_norm": 0.3085862398147583, "learning_rate": 1.9323627457561916e-05, "loss": 0.0177, "step": 57980 }, { "grad_norm": 0.21901483833789825, "learning_rate": 1.9307308814896198e-05, "loss": 0.0128, "step": 57990 }, { "grad_norm": 0.3618226945400238, "learning_rate": 1.9290995416771935e-05, "loss": 0.0178, "step": 58000 }, { "grad_norm": 0.23417501151561737, "learning_rate": 1.9274687265976665e-05, "loss": 0.0165, "step": 58010 }, { "grad_norm": 0.37511056661605835, "learning_rate": 1.9258384365297e-05, "loss": 0.0205, "step": 58020 }, { "grad_norm": 0.31768515706062317, "learning_rate": 1.924208671751866e-05, "loss": 0.0151, "step": 58030 }, { "grad_norm": 0.2546702027320862, "learning_rate": 1.9225794325426492e-05, "loss": 0.0166, "step": 58040 }, { "grad_norm": 0.1709662824869156, "learning_rate": 1.920950719180436e-05, "loss": 0.0143, "step": 58050 }, { "grad_norm": 0.18583321571350098, "learning_rate": 1.919322531943536e-05, "loss": 0.0174, "step": 58060 }, { "grad_norm": 0.3054315745830536, "learning_rate": 1.917694871110157e-05, "loss": 0.0162, "step": 58070 }, { "grad_norm": 0.37407711148262024, "learning_rate": 1.9160677369584234e-05, "loss": 0.016, "step": 58080 }, { "grad_norm": 0.2474830001592636, "learning_rate": 1.9144411297663694e-05, "loss": 0.014, "step": 58090 }, { "grad_norm": 0.42740491032600403, "learning_rate": 1.9128150498119328e-05, "loss": 0.0192, "step": 58100 }, { "grad_norm": 0.24220047891139984, "learning_rate": 1.9111894973729726e-05, "loss": 0.0137, "step": 58110 }, { "grad_norm": 0.4398978054523468, "learning_rate": 1.9095644727272454e-05, "loss": 0.0215, "step": 58120 }, { "grad_norm": 0.33605054020881653, "learning_rate": 1.907939976152429e-05, "loss": 0.0187, "step": 58130 }, { "grad_norm": 0.36537089943885803, "learning_rate": 1.906316007926101e-05, "loss": 0.0219, "step": 58140 }, { "grad_norm": 0.24723371863365173, "learning_rate": 1.904692568325755e-05, "loss": 0.0258, "step": 58150 }, { "grad_norm": 0.28856271505355835, "learning_rate": 1.9030696576287925e-05, "loss": 0.0177, "step": 58160 }, { "grad_norm": 0.33068832755088806, "learning_rate": 1.9014472761125242e-05, "loss": 0.0251, "step": 58170 }, { "grad_norm": 0.24302075803279877, "learning_rate": 1.899825424054172e-05, "loss": 0.0192, "step": 58180 }, { "grad_norm": 0.2542843222618103, "learning_rate": 1.898204101730863e-05, "loss": 0.0168, "step": 58190 }, { "grad_norm": 0.2834097146987915, "learning_rate": 1.8965833094196394e-05, "loss": 0.0313, "step": 58200 }, { "grad_norm": 0.254432737827301, "learning_rate": 1.8949630473974495e-05, "loss": 0.0177, "step": 58210 }, { "grad_norm": 0.18185381591320038, "learning_rate": 1.8933433159411517e-05, "loss": 0.019, "step": 58220 }, { "grad_norm": 0.14475442469120026, "learning_rate": 1.891724115327514e-05, "loss": 0.0133, "step": 58230 }, { "grad_norm": 0.220383882522583, "learning_rate": 1.8901054458332152e-05, "loss": 0.0175, "step": 58240 }, { "grad_norm": 0.20293521881103516, "learning_rate": 1.8884873077348364e-05, "loss": 0.0184, "step": 58250 }, { "grad_norm": 0.5476036667823792, "learning_rate": 1.88686970130888e-05, "loss": 0.0231, "step": 58260 }, { "grad_norm": 0.3407403528690338, "learning_rate": 1.8852526268317456e-05, "loss": 0.0195, "step": 58270 }, { "grad_norm": 0.3289220333099365, "learning_rate": 1.883636084579749e-05, "loss": 0.0174, "step": 58280 }, { "grad_norm": 0.3464517295360565, "learning_rate": 1.8820200748291145e-05, "loss": 0.0214, "step": 58290 }, { "grad_norm": 0.2599075436592102, "learning_rate": 1.8804045978559686e-05, "loss": 0.0196, "step": 58300 }, { "grad_norm": 0.2526929974555969, "learning_rate": 1.8787896539363594e-05, "loss": 0.0189, "step": 58310 }, { "grad_norm": 0.2966950535774231, "learning_rate": 1.8771752433462298e-05, "loss": 0.0186, "step": 58320 }, { "grad_norm": 0.3300835192203522, "learning_rate": 1.8755613663614447e-05, "loss": 0.0212, "step": 58330 }, { "grad_norm": 0.4047725796699524, "learning_rate": 1.873948023257767e-05, "loss": 0.0186, "step": 58340 }, { "grad_norm": 0.2151694893836975, "learning_rate": 1.8723352143108747e-05, "loss": 0.0136, "step": 58350 }, { "grad_norm": 0.2544015645980835, "learning_rate": 1.870722939796352e-05, "loss": 0.0162, "step": 58360 }, { "grad_norm": 0.3418322503566742, "learning_rate": 1.869111199989693e-05, "loss": 0.0207, "step": 58370 }, { "grad_norm": 0.29358333349227905, "learning_rate": 1.8674999951663018e-05, "loss": 0.017, "step": 58380 }, { "grad_norm": 0.3434312641620636, "learning_rate": 1.8658893256014857e-05, "loss": 0.0204, "step": 58390 }, { "grad_norm": 0.18842703104019165, "learning_rate": 1.8642791915704655e-05, "loss": 0.0135, "step": 58400 }, { "grad_norm": 0.4146954417228699, "learning_rate": 1.8626695933483697e-05, "loss": 0.015, "step": 58410 }, { "grad_norm": 0.309000164270401, "learning_rate": 1.8610605312102335e-05, "loss": 0.0173, "step": 58420 }, { "grad_norm": 0.2785181403160095, "learning_rate": 1.8594520054310032e-05, "loss": 0.0209, "step": 58430 }, { "grad_norm": 0.21074244379997253, "learning_rate": 1.8578440162855326e-05, "loss": 0.0169, "step": 58440 }, { "grad_norm": 0.27628180384635925, "learning_rate": 1.8562365640485774e-05, "loss": 0.0194, "step": 58450 }, { "grad_norm": 0.19428429007530212, "learning_rate": 1.854629648994815e-05, "loss": 0.0155, "step": 58460 }, { "grad_norm": 0.3414735198020935, "learning_rate": 1.8530232713988183e-05, "loss": 0.0172, "step": 58470 }, { "grad_norm": 0.3027762174606323, "learning_rate": 1.851417431535074e-05, "loss": 0.0184, "step": 58480 }, { "grad_norm": 0.28029632568359375, "learning_rate": 1.8498121296779785e-05, "loss": 0.0298, "step": 58490 }, { "grad_norm": 0.1980539858341217, "learning_rate": 1.848207366101829e-05, "loss": 0.0195, "step": 58500 }, { "grad_norm": 0.38608166575431824, "learning_rate": 1.8466031410808422e-05, "loss": 0.0204, "step": 58510 }, { "grad_norm": 0.31243735551834106, "learning_rate": 1.844999454889129e-05, "loss": 0.0189, "step": 58520 }, { "grad_norm": 0.29402387142181396, "learning_rate": 1.8433963078007237e-05, "loss": 0.0192, "step": 58530 }, { "grad_norm": 0.21806137263774872, "learning_rate": 1.841793700089554e-05, "loss": 0.0155, "step": 58540 }, { "grad_norm": 0.2557189464569092, "learning_rate": 1.8401916320294644e-05, "loss": 0.015, "step": 58550 }, { "grad_norm": 0.3211047351360321, "learning_rate": 1.838590103894205e-05, "loss": 0.0148, "step": 58560 }, { "grad_norm": 0.18292513489723206, "learning_rate": 1.8369891159574292e-05, "loss": 0.0183, "step": 58570 }, { "grad_norm": 0.3708590567111969, "learning_rate": 1.835388668492708e-05, "loss": 0.0229, "step": 58580 }, { "grad_norm": 0.3397960066795349, "learning_rate": 1.8337887617735095e-05, "loss": 0.017, "step": 58590 }, { "grad_norm": 0.26577919721603394, "learning_rate": 1.8321893960732157e-05, "loss": 0.0151, "step": 58600 }, { "grad_norm": 0.3800511658191681, "learning_rate": 1.8305905716651138e-05, "loss": 0.0162, "step": 58610 }, { "grad_norm": 0.17515650391578674, "learning_rate": 1.8289922888224e-05, "loss": 0.0188, "step": 58620 }, { "grad_norm": 0.22938252985477448, "learning_rate": 1.8273945478181765e-05, "loss": 0.0213, "step": 58630 }, { "grad_norm": 0.2808637320995331, "learning_rate": 1.8257973489254558e-05, "loss": 0.019, "step": 58640 }, { "grad_norm": 0.3106546700000763, "learning_rate": 1.824200692417152e-05, "loss": 0.0186, "step": 58650 }, { "grad_norm": 0.2117556631565094, "learning_rate": 1.8226045785660912e-05, "loss": 0.0157, "step": 58660 }, { "grad_norm": 0.253357470035553, "learning_rate": 1.821009007645006e-05, "loss": 0.0165, "step": 58670 }, { "grad_norm": 0.13973522186279297, "learning_rate": 1.8194139799265357e-05, "loss": 0.0143, "step": 58680 }, { "grad_norm": 0.30740922689437866, "learning_rate": 1.8178194956832295e-05, "loss": 0.021, "step": 58690 }, { "grad_norm": 0.32933667302131653, "learning_rate": 1.8162255551875346e-05, "loss": 0.0165, "step": 58700 }, { "grad_norm": 0.23195898532867432, "learning_rate": 1.8146321587118194e-05, "loss": 0.0243, "step": 58710 }, { "grad_norm": 0.32587945461273193, "learning_rate": 1.8130393065283448e-05, "loss": 0.0178, "step": 58720 }, { "grad_norm": 0.13281762599945068, "learning_rate": 1.8114469989092925e-05, "loss": 0.0143, "step": 58730 }, { "grad_norm": 0.38662874698638916, "learning_rate": 1.8098552361267397e-05, "loss": 0.0194, "step": 58740 }, { "grad_norm": 0.24607504904270172, "learning_rate": 1.8082640184526763e-05, "loss": 0.0209, "step": 58750 }, { "grad_norm": 0.42685577273368835, "learning_rate": 1.806673346158999e-05, "loss": 0.0189, "step": 58760 }, { "grad_norm": 0.2592494487762451, "learning_rate": 1.8050832195175067e-05, "loss": 0.0194, "step": 58770 }, { "grad_norm": 0.4850131869316101, "learning_rate": 1.8034936387999136e-05, "loss": 0.0231, "step": 58780 }, { "grad_norm": 0.14478087425231934, "learning_rate": 1.8019046042778315e-05, "loss": 0.0166, "step": 58790 }, { "grad_norm": 0.39682626724243164, "learning_rate": 1.800316116222785e-05, "loss": 0.0173, "step": 58800 }, { "grad_norm": 0.16121168434619904, "learning_rate": 1.7987281749062018e-05, "loss": 0.0345, "step": 58810 }, { "grad_norm": 0.2761099338531494, "learning_rate": 1.7971407805994195e-05, "loss": 0.0368, "step": 58820 }, { "grad_norm": 0.2701604962348938, "learning_rate": 1.7955539335736787e-05, "loss": 0.0182, "step": 58830 }, { "grad_norm": 0.2725178897380829, "learning_rate": 1.7939676341001304e-05, "loss": 0.0262, "step": 58840 }, { "grad_norm": 0.2893586754798889, "learning_rate": 1.7923818824498275e-05, "loss": 0.0164, "step": 58850 }, { "grad_norm": 0.27293524146080017, "learning_rate": 1.7907966788937315e-05, "loss": 0.0195, "step": 58860 }, { "grad_norm": 0.28973084688186646, "learning_rate": 1.7892120237027116e-05, "loss": 0.0247, "step": 58870 }, { "grad_norm": 0.25491389632225037, "learning_rate": 1.7876279171475413e-05, "loss": 0.0184, "step": 58880 }, { "grad_norm": 0.1945183128118515, "learning_rate": 1.7860443594989028e-05, "loss": 0.0171, "step": 58890 }, { "grad_norm": 0.1638597548007965, "learning_rate": 1.784461351027379e-05, "loss": 0.0121, "step": 58900 }, { "grad_norm": 0.35725173354148865, "learning_rate": 1.7828788920034677e-05, "loss": 0.0201, "step": 58910 }, { "grad_norm": 0.3243994414806366, "learning_rate": 1.7812969826975623e-05, "loss": 0.0167, "step": 58920 }, { "grad_norm": 0.25723135471343994, "learning_rate": 1.7797156233799738e-05, "loss": 0.0168, "step": 58930 }, { "grad_norm": 0.23525370657444, "learning_rate": 1.7781348143209094e-05, "loss": 0.0187, "step": 58940 }, { "grad_norm": 0.26670974493026733, "learning_rate": 1.7765545557904862e-05, "loss": 0.0276, "step": 58950 }, { "grad_norm": 0.28913530707359314, "learning_rate": 1.7749748480587302e-05, "loss": 0.0185, "step": 58960 }, { "grad_norm": 0.16923454403877258, "learning_rate": 1.773395691395564e-05, "loss": 0.0166, "step": 58970 }, { "grad_norm": 0.22061270475387573, "learning_rate": 1.7718170860708305e-05, "loss": 0.0159, "step": 58980 }, { "grad_norm": 0.2424461394548416, "learning_rate": 1.770239032354264e-05, "loss": 0.0138, "step": 58990 }, { "grad_norm": 0.26783064007759094, "learning_rate": 1.7686615305155124e-05, "loss": 0.034, "step": 59000 }, { "grad_norm": 0.31900864839553833, "learning_rate": 1.767084580824128e-05, "loss": 0.0154, "step": 59010 }, { "grad_norm": 0.5577208995819092, "learning_rate": 1.765508183549569e-05, "loss": 0.0218, "step": 59020 }, { "grad_norm": 0.31379586458206177, "learning_rate": 1.763932338961199e-05, "loss": 0.0189, "step": 59030 }, { "grad_norm": 0.2807890474796295, "learning_rate": 1.762357047328284e-05, "loss": 0.0139, "step": 59040 }, { "grad_norm": 0.24599413573741913, "learning_rate": 1.7607823089199997e-05, "loss": 0.0216, "step": 59050 }, { "grad_norm": 0.2544068694114685, "learning_rate": 1.7592081240054265e-05, "loss": 0.0221, "step": 59060 }, { "grad_norm": 0.346453458070755, "learning_rate": 1.7576344928535488e-05, "loss": 0.0173, "step": 59070 }, { "grad_norm": 0.341790109872818, "learning_rate": 1.7560614157332573e-05, "loss": 0.0209, "step": 59080 }, { "grad_norm": 0.40092733502388, "learning_rate": 1.7544888929133495e-05, "loss": 0.0197, "step": 59090 }, { "grad_norm": 0.21509559452533722, "learning_rate": 1.752916924662522e-05, "loss": 0.0173, "step": 59100 }, { "grad_norm": 0.1735782027244568, "learning_rate": 1.751345511249387e-05, "loss": 0.0176, "step": 59110 }, { "grad_norm": 0.2298157960176468, "learning_rate": 1.749774652942452e-05, "loss": 0.0207, "step": 59120 }, { "grad_norm": 0.20068883895874023, "learning_rate": 1.748204350010135e-05, "loss": 0.0209, "step": 59130 }, { "grad_norm": 0.17631804943084717, "learning_rate": 1.7466346027207574e-05, "loss": 0.0146, "step": 59140 }, { "grad_norm": 0.22543863952159882, "learning_rate": 1.745065411342547e-05, "loss": 0.0202, "step": 59150 }, { "grad_norm": 0.1600237786769867, "learning_rate": 1.7434967761436366e-05, "loss": 0.0213, "step": 59160 }, { "grad_norm": 0.35343584418296814, "learning_rate": 1.741928697392058e-05, "loss": 0.0202, "step": 59170 }, { "grad_norm": 0.271083801984787, "learning_rate": 1.7403611753557597e-05, "loss": 0.0151, "step": 59180 }, { "grad_norm": 0.24494676291942596, "learning_rate": 1.738794210302584e-05, "loss": 0.0165, "step": 59190 }, { "grad_norm": 0.30381497740745544, "learning_rate": 1.7372278025002837e-05, "loss": 0.0161, "step": 59200 }, { "grad_norm": 0.30102238059043884, "learning_rate": 1.7356619522165164e-05, "loss": 0.0193, "step": 59210 }, { "grad_norm": 0.2514151930809021, "learning_rate": 1.7340966597188378e-05, "loss": 0.0114, "step": 59220 }, { "grad_norm": 0.20096801221370697, "learning_rate": 1.732531925274722e-05, "loss": 0.0173, "step": 59230 }, { "grad_norm": 0.2678171694278717, "learning_rate": 1.7309677491515318e-05, "loss": 0.0178, "step": 59240 }, { "grad_norm": 0.19878922402858734, "learning_rate": 1.7294041316165455e-05, "loss": 0.0134, "step": 59250 }, { "grad_norm": 0.21002574265003204, "learning_rate": 1.727841072936942e-05, "loss": 0.0177, "step": 59260 }, { "grad_norm": 0.27125275135040283, "learning_rate": 1.7262785733798058e-05, "loss": 0.0146, "step": 59270 }, { "grad_norm": 0.3194488286972046, "learning_rate": 1.7247166332121246e-05, "loss": 0.0189, "step": 59280 }, { "grad_norm": 0.21761691570281982, "learning_rate": 1.7231552527007933e-05, "loss": 0.0179, "step": 59290 }, { "grad_norm": 0.24758696556091309, "learning_rate": 1.721594432112606e-05, "loss": 0.0174, "step": 59300 }, { "grad_norm": 0.23845823109149933, "learning_rate": 1.7200341717142655e-05, "loss": 0.0209, "step": 59310 }, { "grad_norm": 0.3425320088863373, "learning_rate": 1.718474471772377e-05, "loss": 0.0195, "step": 59320 }, { "grad_norm": 0.30825528502464294, "learning_rate": 1.7169153325534528e-05, "loss": 0.0176, "step": 59330 }, { "grad_norm": 0.1809777319431305, "learning_rate": 1.7153567543239045e-05, "loss": 0.0195, "step": 59340 }, { "grad_norm": 0.21636204421520233, "learning_rate": 1.7137987373500525e-05, "loss": 0.0174, "step": 59350 }, { "grad_norm": 0.21649904549121857, "learning_rate": 1.7122412818981198e-05, "loss": 0.0211, "step": 59360 }, { "grad_norm": 0.32142961025238037, "learning_rate": 1.7106843882342283e-05, "loss": 0.0178, "step": 59370 }, { "grad_norm": 0.2472977191209793, "learning_rate": 1.709128056624415e-05, "loss": 0.0179, "step": 59380 }, { "grad_norm": 0.3031267523765564, "learning_rate": 1.70757228733461e-05, "loss": 0.0177, "step": 59390 }, { "grad_norm": 0.17076383531093597, "learning_rate": 1.706017080630653e-05, "loss": 0.0157, "step": 59400 }, { "grad_norm": 0.20778940618038177, "learning_rate": 1.7044624367782873e-05, "loss": 0.016, "step": 59410 }, { "grad_norm": 0.2689976096153259, "learning_rate": 1.7029083560431553e-05, "loss": 0.0212, "step": 59420 }, { "grad_norm": 0.2684507369995117, "learning_rate": 1.701354838690813e-05, "loss": 0.0206, "step": 59430 }, { "grad_norm": 0.26042303442955017, "learning_rate": 1.6998018849867087e-05, "loss": 0.0126, "step": 59440 }, { "grad_norm": 0.307182639837265, "learning_rate": 1.698249495196202e-05, "loss": 0.021, "step": 59450 }, { "grad_norm": 0.27523645758628845, "learning_rate": 1.6966976695845528e-05, "loss": 0.0191, "step": 59460 }, { "grad_norm": 0.16972243785858154, "learning_rate": 1.6951464084169268e-05, "loss": 0.0154, "step": 59470 }, { "grad_norm": 0.33612561225891113, "learning_rate": 1.6935957119583916e-05, "loss": 0.024, "step": 59480 }, { "grad_norm": 0.19259363412857056, "learning_rate": 1.6920455804739205e-05, "loss": 0.0155, "step": 59490 }, { "grad_norm": 0.2350740134716034, "learning_rate": 1.6904960142283855e-05, "loss": 0.0219, "step": 59500 }, { "grad_norm": 0.20368850231170654, "learning_rate": 1.6889470134865666e-05, "loss": 0.0147, "step": 59510 }, { "grad_norm": 0.17259477078914642, "learning_rate": 1.687398578513145e-05, "loss": 0.0225, "step": 59520 }, { "grad_norm": 0.38379189372062683, "learning_rate": 1.6858507095727066e-05, "loss": 0.0187, "step": 59530 }, { "grad_norm": 0.26296260952949524, "learning_rate": 1.6843034069297403e-05, "loss": 0.0156, "step": 59540 }, { "grad_norm": 0.29641199111938477, "learning_rate": 1.682756670848637e-05, "loss": 0.0229, "step": 59550 }, { "grad_norm": 0.2537968158721924, "learning_rate": 1.6812105015936937e-05, "loss": 0.0128, "step": 59560 }, { "grad_norm": 0.376257985830307, "learning_rate": 1.6796648994291027e-05, "loss": 0.0194, "step": 59570 }, { "grad_norm": 0.24899452924728394, "learning_rate": 1.678119864618973e-05, "loss": 0.0148, "step": 59580 }, { "grad_norm": 0.25549304485321045, "learning_rate": 1.6765753974273023e-05, "loss": 0.0181, "step": 59590 }, { "grad_norm": 0.5415204167366028, "learning_rate": 1.675031498118001e-05, "loss": 0.0187, "step": 59600 }, { "grad_norm": 0.44205403327941895, "learning_rate": 1.6734881669548804e-05, "loss": 0.0138, "step": 59610 }, { "grad_norm": 0.1606810837984085, "learning_rate": 1.671945404201648e-05, "loss": 0.0253, "step": 59620 }, { "grad_norm": 0.31587913632392883, "learning_rate": 1.670403210121927e-05, "loss": 0.0172, "step": 59630 }, { "grad_norm": 0.2882581651210785, "learning_rate": 1.6688615849792312e-05, "loss": 0.0144, "step": 59640 }, { "grad_norm": 0.4062470495700836, "learning_rate": 1.6673205290369832e-05, "loss": 0.0252, "step": 59650 }, { "grad_norm": 0.47193777561187744, "learning_rate": 1.6657800425585086e-05, "loss": 0.0243, "step": 59660 }, { "grad_norm": 0.22227855026721954, "learning_rate": 1.6642401258070327e-05, "loss": 0.024, "step": 59670 }, { "grad_norm": 0.3131685256958008, "learning_rate": 1.6627007790456884e-05, "loss": 0.0202, "step": 59680 }, { "grad_norm": 0.1385500729084015, "learning_rate": 1.6611620025375037e-05, "loss": 0.0149, "step": 59690 }, { "grad_norm": 0.18795104324817657, "learning_rate": 1.6596237965454154e-05, "loss": 0.0193, "step": 59700 }, { "grad_norm": 0.40828099846839905, "learning_rate": 1.6580861613322607e-05, "loss": 0.0218, "step": 59710 }, { "grad_norm": 0.21704670786857605, "learning_rate": 1.6565490971607796e-05, "loss": 0.0106, "step": 59720 }, { "grad_norm": 0.29762622714042664, "learning_rate": 1.655012604293614e-05, "loss": 0.0191, "step": 59730 }, { "grad_norm": 0.1952352672815323, "learning_rate": 1.653476682993309e-05, "loss": 0.0209, "step": 59740 }, { "grad_norm": 0.2662971615791321, "learning_rate": 1.651941333522311e-05, "loss": 0.0193, "step": 59750 }, { "grad_norm": 0.28175026178359985, "learning_rate": 1.6504065561429715e-05, "loss": 0.0175, "step": 59760 }, { "grad_norm": 0.25366711616516113, "learning_rate": 1.6488723511175385e-05, "loss": 0.0242, "step": 59770 }, { "grad_norm": 0.3074069917201996, "learning_rate": 1.6473387187081668e-05, "loss": 0.0194, "step": 59780 }, { "grad_norm": 0.19725237786769867, "learning_rate": 1.6458056591769123e-05, "loss": 0.0158, "step": 59790 }, { "grad_norm": 0.29938599467277527, "learning_rate": 1.6442731727857335e-05, "loss": 0.0171, "step": 59800 }, { "grad_norm": 0.27174532413482666, "learning_rate": 1.642741259796492e-05, "loss": 0.0169, "step": 59810 }, { "grad_norm": 0.2828049957752228, "learning_rate": 1.641209920470944e-05, "loss": 0.016, "step": 59820 }, { "grad_norm": 0.2832939922809601, "learning_rate": 1.639679155070762e-05, "loss": 0.015, "step": 59830 }, { "grad_norm": 0.2640364468097687, "learning_rate": 1.6381489638575048e-05, "loss": 0.0166, "step": 59840 }, { "grad_norm": 0.2886745035648346, "learning_rate": 1.636619347092643e-05, "loss": 0.0175, "step": 59850 }, { "grad_norm": 0.3375096023082733, "learning_rate": 1.6350903050375476e-05, "loss": 0.0165, "step": 59860 }, { "grad_norm": 0.3589394986629486, "learning_rate": 1.6335618379534856e-05, "loss": 0.0174, "step": 59870 }, { "grad_norm": 0.29199302196502686, "learning_rate": 1.6320339461016364e-05, "loss": 0.0143, "step": 59880 }, { "grad_norm": 0.457783043384552, "learning_rate": 1.6305066297430687e-05, "loss": 0.0191, "step": 59890 }, { "grad_norm": 0.37436509132385254, "learning_rate": 1.6289798891387654e-05, "loss": 0.0247, "step": 59900 }, { "grad_norm": 0.21633394062519073, "learning_rate": 1.6274537245495995e-05, "loss": 0.0167, "step": 59910 }, { "grad_norm": 0.27157390117645264, "learning_rate": 1.6259281362363527e-05, "loss": 0.0144, "step": 59920 }, { "grad_norm": 0.2446240931749344, "learning_rate": 1.6244031244597068e-05, "loss": 0.0234, "step": 59930 }, { "grad_norm": 0.36042308807373047, "learning_rate": 1.622878689480244e-05, "loss": 0.0169, "step": 59940 }, { "grad_norm": 0.33536943793296814, "learning_rate": 1.6213548315584498e-05, "loss": 0.0212, "step": 59950 }, { "grad_norm": 0.32261916995048523, "learning_rate": 1.6198315509547074e-05, "loss": 0.0153, "step": 59960 }, { "grad_norm": 0.19975169003009796, "learning_rate": 1.6183088479293056e-05, "loss": 0.0143, "step": 59970 }, { "grad_norm": 0.24516183137893677, "learning_rate": 1.6167867227424316e-05, "loss": 0.0151, "step": 59980 }, { "grad_norm": 0.1411381959915161, "learning_rate": 1.615265175654176e-05, "loss": 0.017, "step": 59990 }, { "grad_norm": 0.29535672068595886, "learning_rate": 1.613744206924529e-05, "loss": 0.0184, "step": 60000 }, { "grad_norm": 0.2760394811630249, "learning_rate": 1.6122238168133845e-05, "loss": 0.0235, "step": 60010 }, { "grad_norm": 0.24764259159564972, "learning_rate": 1.6107040055805305e-05, "loss": 0.0164, "step": 60020 }, { "grad_norm": 0.30726343393325806, "learning_rate": 1.6091847734856673e-05, "loss": 0.0181, "step": 60030 }, { "grad_norm": 0.29370224475860596, "learning_rate": 1.6076661207883865e-05, "loss": 0.0165, "step": 60040 }, { "grad_norm": 0.17591461539268494, "learning_rate": 1.6061480477481848e-05, "loss": 0.014, "step": 60050 }, { "grad_norm": 0.3302326798439026, "learning_rate": 1.604630554624461e-05, "loss": 0.0145, "step": 60060 }, { "grad_norm": 0.2882370948791504, "learning_rate": 1.603113641676509e-05, "loss": 0.0144, "step": 60070 }, { "grad_norm": 0.18246735632419586, "learning_rate": 1.6015973091635338e-05, "loss": 0.0156, "step": 60080 }, { "grad_norm": 0.19736991822719574, "learning_rate": 1.6000815573446288e-05, "loss": 0.0151, "step": 60090 }, { "grad_norm": 0.2767229378223419, "learning_rate": 1.5985663864788002e-05, "loss": 0.0211, "step": 60100 }, { "grad_norm": 0.3998051583766937, "learning_rate": 1.597051796824946e-05, "loss": 0.0213, "step": 60110 }, { "grad_norm": 0.2679370641708374, "learning_rate": 1.5955377886418682e-05, "loss": 0.0142, "step": 60120 }, { "grad_norm": 0.23248951137065887, "learning_rate": 1.5940243621882704e-05, "loss": 0.0134, "step": 60130 }, { "grad_norm": 0.27120736241340637, "learning_rate": 1.5925115177227555e-05, "loss": 0.0227, "step": 60140 }, { "grad_norm": 0.27250561118125916, "learning_rate": 1.5909992555038288e-05, "loss": 0.0178, "step": 60150 }, { "grad_norm": 0.27100592851638794, "learning_rate": 1.5894875757898912e-05, "loss": 0.0227, "step": 60160 }, { "grad_norm": 0.2584637403488159, "learning_rate": 1.58797647883925e-05, "loss": 0.0188, "step": 60170 }, { "grad_norm": 0.4047964811325073, "learning_rate": 1.586465964910109e-05, "loss": 0.0176, "step": 60180 }, { "grad_norm": 0.3359230160713196, "learning_rate": 1.5849560342605734e-05, "loss": 0.0189, "step": 60190 }, { "grad_norm": 0.25717824697494507, "learning_rate": 1.583446687148651e-05, "loss": 0.0183, "step": 60200 }, { "grad_norm": 0.2190369963645935, "learning_rate": 1.581937923832248e-05, "loss": 0.0178, "step": 60210 }, { "grad_norm": 0.2821587920188904, "learning_rate": 1.5804297445691663e-05, "loss": 0.015, "step": 60220 }, { "grad_norm": 0.34128502011299133, "learning_rate": 1.578922149617119e-05, "loss": 0.0202, "step": 60230 }, { "grad_norm": 0.19308297336101532, "learning_rate": 1.5774151392337084e-05, "loss": 0.0136, "step": 60240 }, { "grad_norm": 0.25178730487823486, "learning_rate": 1.575908713676442e-05, "loss": 0.0148, "step": 60250 }, { "grad_norm": 0.4411889314651489, "learning_rate": 1.5744028732027293e-05, "loss": 0.019, "step": 60260 }, { "grad_norm": 0.25929924845695496, "learning_rate": 1.5728976180698723e-05, "loss": 0.0178, "step": 60270 }, { "grad_norm": 0.2505747675895691, "learning_rate": 1.5713929485350836e-05, "loss": 0.0208, "step": 60280 }, { "grad_norm": 0.22160066664218903, "learning_rate": 1.569888864855465e-05, "loss": 0.0146, "step": 60290 }, { "grad_norm": 0.2076391577720642, "learning_rate": 1.568385367288028e-05, "loss": 0.0134, "step": 60300 }, { "grad_norm": 0.23038537800312042, "learning_rate": 1.5668824560896755e-05, "loss": 0.0192, "step": 60310 }, { "grad_norm": 0.4016944468021393, "learning_rate": 1.5653801315172156e-05, "loss": 0.0164, "step": 60320 }, { "grad_norm": 0.2513532340526581, "learning_rate": 1.5638783938273554e-05, "loss": 0.0149, "step": 60330 }, { "grad_norm": 0.3334217071533203, "learning_rate": 1.5623772432766966e-05, "loss": 0.0225, "step": 60340 }, { "grad_norm": 0.2331857979297638, "learning_rate": 1.5608766801217507e-05, "loss": 0.019, "step": 60350 }, { "grad_norm": 0.18719199299812317, "learning_rate": 1.5593767046189184e-05, "loss": 0.0137, "step": 60360 }, { "grad_norm": 0.20533300936222076, "learning_rate": 1.5578773170245064e-05, "loss": 0.0145, "step": 60370 }, { "grad_norm": 0.32898423075675964, "learning_rate": 1.5563785175947182e-05, "loss": 0.0191, "step": 60380 }, { "grad_norm": 0.2422778308391571, "learning_rate": 1.5548803065856582e-05, "loss": 0.0125, "step": 60390 }, { "grad_norm": 0.4240127205848694, "learning_rate": 1.55338268425333e-05, "loss": 0.0197, "step": 60400 }, { "grad_norm": 0.21339094638824463, "learning_rate": 1.5518856508536373e-05, "loss": 0.0217, "step": 60410 }, { "grad_norm": 0.36667001247406006, "learning_rate": 1.5503892066423786e-05, "loss": 0.0184, "step": 60420 }, { "grad_norm": 0.526227593421936, "learning_rate": 1.548893351875258e-05, "loss": 0.0267, "step": 60430 }, { "grad_norm": 0.2961614727973938, "learning_rate": 1.547398086807875e-05, "loss": 0.0164, "step": 60440 }, { "grad_norm": 0.3416142761707306, "learning_rate": 1.5459034116957305e-05, "loss": 0.0279, "step": 60450 }, { "grad_norm": 0.2551792562007904, "learning_rate": 1.544409326794225e-05, "loss": 0.033, "step": 60460 }, { "grad_norm": 0.20779898762702942, "learning_rate": 1.542915832358651e-05, "loss": 0.0189, "step": 60470 }, { "grad_norm": 0.20263341069221497, "learning_rate": 1.541422928644214e-05, "loss": 0.0198, "step": 60480 }, { "grad_norm": 0.3150947093963623, "learning_rate": 1.5399306159060024e-05, "loss": 0.0236, "step": 60490 }, { "grad_norm": 0.14416566491127014, "learning_rate": 1.5384388943990185e-05, "loss": 0.0143, "step": 60500 }, { "grad_norm": 0.2657409906387329, "learning_rate": 1.5369477643781526e-05, "loss": 0.0184, "step": 60510 }, { "grad_norm": 0.19368427991867065, "learning_rate": 1.5354572260981985e-05, "loss": 0.0207, "step": 60520 }, { "grad_norm": 0.2427494376897812, "learning_rate": 1.533967279813851e-05, "loss": 0.0144, "step": 60530 }, { "grad_norm": 0.44258007407188416, "learning_rate": 1.5324779257796956e-05, "loss": 0.0172, "step": 60540 }, { "grad_norm": 0.2147538959980011, "learning_rate": 1.5309891642502293e-05, "loss": 0.0163, "step": 60550 }, { "grad_norm": 0.27716711163520813, "learning_rate": 1.5295009954798357e-05, "loss": 0.0148, "step": 60560 }, { "grad_norm": 0.2255203127861023, "learning_rate": 1.5280134197228036e-05, "loss": 0.0197, "step": 60570 }, { "grad_norm": 0.37567436695098877, "learning_rate": 1.526526437233319e-05, "loss": 0.0206, "step": 60580 }, { "grad_norm": 0.21380391716957092, "learning_rate": 1.525040048265466e-05, "loss": 0.0139, "step": 60590 }, { "grad_norm": 0.13784097135066986, "learning_rate": 1.5235542530732288e-05, "loss": 0.0162, "step": 60600 }, { "grad_norm": 0.24655000865459442, "learning_rate": 1.5220690519104901e-05, "loss": 0.0187, "step": 60610 }, { "grad_norm": 0.22926585376262665, "learning_rate": 1.5205844450310275e-05, "loss": 0.0169, "step": 60620 }, { "grad_norm": 0.3502432107925415, "learning_rate": 1.519100432688521e-05, "loss": 0.0197, "step": 60630 }, { "grad_norm": 0.2156250923871994, "learning_rate": 1.5176170151365476e-05, "loss": 0.0179, "step": 60640 }, { "grad_norm": 0.13281869888305664, "learning_rate": 1.5161341926285833e-05, "loss": 0.0144, "step": 60650 }, { "grad_norm": 0.4311956465244293, "learning_rate": 1.5146519654180025e-05, "loss": 0.0184, "step": 60660 }, { "grad_norm": 0.38392338156700134, "learning_rate": 1.5131703337580739e-05, "loss": 0.0228, "step": 60670 }, { "grad_norm": 0.1983446329832077, "learning_rate": 1.5116892979019731e-05, "loss": 0.0179, "step": 60680 }, { "grad_norm": 0.1838809698820114, "learning_rate": 1.5102088581027623e-05, "loss": 0.0143, "step": 60690 }, { "grad_norm": 0.3069838285446167, "learning_rate": 1.5087290146134154e-05, "loss": 0.0147, "step": 60700 }, { "grad_norm": 0.2315262407064438, "learning_rate": 1.5072497676867915e-05, "loss": 0.0164, "step": 60710 }, { "grad_norm": 0.8201242685317993, "learning_rate": 1.505771117575655e-05, "loss": 0.0233, "step": 60720 }, { "grad_norm": 0.3384436070919037, "learning_rate": 1.5042930645326691e-05, "loss": 0.0145, "step": 60730 }, { "grad_norm": 0.26493245363235474, "learning_rate": 1.5028156088103879e-05, "loss": 0.02, "step": 60740 }, { "grad_norm": 0.3783290684223175, "learning_rate": 1.5013387506612735e-05, "loss": 0.0207, "step": 60750 }, { "grad_norm": 0.2749718725681305, "learning_rate": 1.4998624903376767e-05, "loss": 0.0195, "step": 60760 }, { "grad_norm": 0.3510986864566803, "learning_rate": 1.4983868280918517e-05, "loss": 0.018, "step": 60770 }, { "grad_norm": 0.27377691864967346, "learning_rate": 1.4969117641759478e-05, "loss": 0.0215, "step": 60780 }, { "grad_norm": 0.20228150486946106, "learning_rate": 1.495437298842014e-05, "loss": 0.017, "step": 60790 }, { "grad_norm": 0.29187074303627014, "learning_rate": 1.4939634323419976e-05, "loss": 0.0176, "step": 60800 }, { "grad_norm": 0.3034822642803192, "learning_rate": 1.4924901649277384e-05, "loss": 0.0207, "step": 60810 }, { "grad_norm": 0.24454757571220398, "learning_rate": 1.4910174968509793e-05, "loss": 0.0128, "step": 60820 }, { "grad_norm": 0.2779881954193115, "learning_rate": 1.4895454283633598e-05, "loss": 0.0135, "step": 60830 }, { "grad_norm": 0.3014433979988098, "learning_rate": 1.488073959716415e-05, "loss": 0.0166, "step": 60840 }, { "grad_norm": 0.2589792013168335, "learning_rate": 1.4866030911615791e-05, "loss": 0.0212, "step": 60850 }, { "grad_norm": 0.16170106828212738, "learning_rate": 1.4851328229501849e-05, "loss": 0.0145, "step": 60860 }, { "grad_norm": 0.17404478788375854, "learning_rate": 1.4836631553334562e-05, "loss": 0.0162, "step": 60870 }, { "grad_norm": 0.19648444652557373, "learning_rate": 1.4821940885625251e-05, "loss": 0.015, "step": 60880 }, { "grad_norm": 0.3195236623287201, "learning_rate": 1.4807256228884109e-05, "loss": 0.018, "step": 60890 }, { "grad_norm": 0.21383921802043915, "learning_rate": 1.4792577585620353e-05, "loss": 0.0129, "step": 60900 }, { "grad_norm": 0.29446887969970703, "learning_rate": 1.4777904958342164e-05, "loss": 0.0166, "step": 60910 }, { "grad_norm": 0.34349149465560913, "learning_rate": 1.4763238349556691e-05, "loss": 0.0137, "step": 60920 }, { "grad_norm": 0.2176813781261444, "learning_rate": 1.4748577761770072e-05, "loss": 0.0195, "step": 60930 }, { "grad_norm": 0.2263011783361435, "learning_rate": 1.4733923197487354e-05, "loss": 0.0176, "step": 60940 }, { "grad_norm": 0.20140549540519714, "learning_rate": 1.4719274659212662e-05, "loss": 0.0192, "step": 60950 }, { "grad_norm": 0.15176637470722198, "learning_rate": 1.4704632149448983e-05, "loss": 0.017, "step": 60960 }, { "grad_norm": 0.25229695439338684, "learning_rate": 1.4689995670698343e-05, "loss": 0.0194, "step": 60970 }, { "grad_norm": 0.28130903840065, "learning_rate": 1.4675365225461728e-05, "loss": 0.018, "step": 60980 }, { "grad_norm": 0.2623971104621887, "learning_rate": 1.4660740816239032e-05, "loss": 0.0152, "step": 60990 }, { "grad_norm": 0.3340604901313782, "learning_rate": 1.464612244552923e-05, "loss": 0.0159, "step": 61000 }, { "grad_norm": 0.1641353815793991, "learning_rate": 1.4631510115830161e-05, "loss": 0.0142, "step": 61010 }, { "grad_norm": 0.259081095457077, "learning_rate": 1.4616903829638679e-05, "loss": 0.0156, "step": 61020 }, { "grad_norm": 0.15371611714363098, "learning_rate": 1.46023035894506e-05, "loss": 0.0172, "step": 61030 }, { "grad_norm": 0.36661234498023987, "learning_rate": 1.4587709397760713e-05, "loss": 0.02, "step": 61040 }, { "grad_norm": 0.2245754599571228, "learning_rate": 1.4573121257062755e-05, "loss": 0.0163, "step": 61050 }, { "grad_norm": 0.5562722086906433, "learning_rate": 1.4558539169849472e-05, "loss": 0.0254, "step": 61060 }, { "grad_norm": 0.3860163688659668, "learning_rate": 1.4543963138612499e-05, "loss": 0.0186, "step": 61070 }, { "grad_norm": 0.24029351770877838, "learning_rate": 1.4529393165842498e-05, "loss": 0.0155, "step": 61080 }, { "grad_norm": 0.2921229302883148, "learning_rate": 1.4514829254029084e-05, "loss": 0.0145, "step": 61090 }, { "grad_norm": 0.2704012393951416, "learning_rate": 1.450027140566083e-05, "loss": 0.0184, "step": 61100 }, { "grad_norm": 0.262210488319397, "learning_rate": 1.4485719623225268e-05, "loss": 0.0165, "step": 61110 }, { "grad_norm": 0.2069641798734665, "learning_rate": 1.4471173909208912e-05, "loss": 0.0179, "step": 61120 }, { "grad_norm": 0.19496318697929382, "learning_rate": 1.4456634266097236e-05, "loss": 0.0184, "step": 61130 }, { "grad_norm": 0.3779647648334503, "learning_rate": 1.444210069637461e-05, "loss": 0.0207, "step": 61140 }, { "grad_norm": 0.33493226766586304, "learning_rate": 1.4427573202524502e-05, "loss": 0.0176, "step": 61150 }, { "grad_norm": 0.24294641613960266, "learning_rate": 1.4413051787029208e-05, "loss": 0.0234, "step": 61160 }, { "grad_norm": 0.23903359472751617, "learning_rate": 1.439853645237006e-05, "loss": 0.0202, "step": 61170 }, { "grad_norm": 0.17709507048130035, "learning_rate": 1.4384027201027344e-05, "loss": 0.0151, "step": 61180 }, { "grad_norm": 0.14252352714538574, "learning_rate": 1.4369524035480253e-05, "loss": 0.0149, "step": 61190 }, { "grad_norm": 0.6118283867835999, "learning_rate": 1.4355026958207035e-05, "loss": 0.0206, "step": 61200 }, { "grad_norm": 0.2110971361398697, "learning_rate": 1.434053597168481e-05, "loss": 0.0176, "step": 61210 }, { "grad_norm": 0.4262000620365143, "learning_rate": 1.432605107838969e-05, "loss": 0.0208, "step": 61220 }, { "grad_norm": 0.6569964289665222, "learning_rate": 1.4311572280796759e-05, "loss": 0.0206, "step": 61230 }, { "grad_norm": 0.29573842883110046, "learning_rate": 1.4297099581380047e-05, "loss": 0.0231, "step": 61240 }, { "grad_norm": 0.25185465812683105, "learning_rate": 1.4282632982612538e-05, "loss": 0.0205, "step": 61250 }, { "grad_norm": 0.23340532183647156, "learning_rate": 1.4268172486966197e-05, "loss": 0.0236, "step": 61260 }, { "grad_norm": 0.21801786124706268, "learning_rate": 1.4253718096911894e-05, "loss": 0.0219, "step": 61270 }, { "grad_norm": 0.3115274906158447, "learning_rate": 1.4239269814919504e-05, "loss": 0.0248, "step": 61280 }, { "grad_norm": 0.21358603239059448, "learning_rate": 1.4224827643457843e-05, "loss": 0.0203, "step": 61290 }, { "grad_norm": 0.2376098781824112, "learning_rate": 1.4210391584994686e-05, "loss": 0.0178, "step": 61300 }, { "grad_norm": 0.3345540463924408, "learning_rate": 1.4195961641996763e-05, "loss": 0.0163, "step": 61310 }, { "grad_norm": 0.23144815862178802, "learning_rate": 1.4181537816929751e-05, "loss": 0.0182, "step": 61320 }, { "grad_norm": 0.2865256667137146, "learning_rate": 1.4167120112258303e-05, "loss": 0.023, "step": 61330 }, { "grad_norm": 0.18033424019813538, "learning_rate": 1.4152708530445963e-05, "loss": 0.0145, "step": 61340 }, { "grad_norm": 0.180026113986969, "learning_rate": 1.4138303073955344e-05, "loss": 0.014, "step": 61350 }, { "grad_norm": 0.2668270170688629, "learning_rate": 1.4123903745247897e-05, "loss": 0.0174, "step": 61360 }, { "grad_norm": 0.13380075991153717, "learning_rate": 1.4109510546784082e-05, "loss": 0.0114, "step": 61370 }, { "grad_norm": 0.36691758036613464, "learning_rate": 1.4095123481023325e-05, "loss": 0.0227, "step": 61380 }, { "grad_norm": 0.4402371644973755, "learning_rate": 1.4080742550423936e-05, "loss": 0.0171, "step": 61390 }, { "grad_norm": 0.17406314611434937, "learning_rate": 1.4066367757443278e-05, "loss": 0.0175, "step": 61400 }, { "grad_norm": 0.23294919729232788, "learning_rate": 1.4051999104537572e-05, "loss": 0.0151, "step": 61410 }, { "grad_norm": 0.29345938563346863, "learning_rate": 1.4037636594162034e-05, "loss": 0.0164, "step": 61420 }, { "grad_norm": 0.2609432637691498, "learning_rate": 1.402328022877083e-05, "loss": 0.0148, "step": 61430 }, { "grad_norm": 0.17642538249492645, "learning_rate": 1.4008930010817073e-05, "loss": 0.0194, "step": 61440 }, { "grad_norm": 0.32691487669944763, "learning_rate": 1.3994585942752831e-05, "loss": 0.0124, "step": 61450 }, { "grad_norm": 0.3482919931411743, "learning_rate": 1.398024802702909e-05, "loss": 0.0158, "step": 61460 }, { "grad_norm": 0.3173793852329254, "learning_rate": 1.3965916266095819e-05, "loss": 0.0165, "step": 61470 }, { "grad_norm": 0.24933448433876038, "learning_rate": 1.395159066240192e-05, "loss": 0.0146, "step": 61480 }, { "grad_norm": 0.22866122424602509, "learning_rate": 1.3937271218395248e-05, "loss": 0.0127, "step": 61490 }, { "grad_norm": 0.30060017108917236, "learning_rate": 1.3922957936522613e-05, "loss": 0.0199, "step": 61500 }, { "grad_norm": 0.2810000479221344, "learning_rate": 1.3908650819229758e-05, "loss": 0.0156, "step": 61510 }, { "grad_norm": 0.2567741870880127, "learning_rate": 1.389434986896137e-05, "loss": 0.0156, "step": 61520 }, { "grad_norm": 0.2057054191827774, "learning_rate": 1.388005508816112e-05, "loss": 0.0123, "step": 61530 }, { "grad_norm": 0.2654685974121094, "learning_rate": 1.3865766479271557e-05, "loss": 0.0174, "step": 61540 }, { "grad_norm": 0.27061939239501953, "learning_rate": 1.3851484044734225e-05, "loss": 0.0253, "step": 61550 }, { "grad_norm": 0.22319887578487396, "learning_rate": 1.3837207786989608e-05, "loss": 0.0261, "step": 61560 }, { "grad_norm": 0.23345033824443817, "learning_rate": 1.382293770847713e-05, "loss": 0.0161, "step": 61570 }, { "grad_norm": 0.26154637336730957, "learning_rate": 1.3808673811635159e-05, "loss": 0.0165, "step": 61580 }, { "grad_norm": 0.18182121217250824, "learning_rate": 1.3794416098900975e-05, "loss": 0.0183, "step": 61590 }, { "grad_norm": 0.22989219427108765, "learning_rate": 1.3780164572710886e-05, "loss": 0.0247, "step": 61600 }, { "grad_norm": 0.2661822736263275, "learning_rate": 1.3765919235500035e-05, "loss": 0.0242, "step": 61610 }, { "grad_norm": 0.3048069477081299, "learning_rate": 1.375168008970259e-05, "loss": 0.018, "step": 61620 }, { "grad_norm": 0.3685881197452545, "learning_rate": 1.3737447137751635e-05, "loss": 0.0142, "step": 61630 }, { "grad_norm": 0.3028658330440521, "learning_rate": 1.372322038207915e-05, "loss": 0.0154, "step": 61640 }, { "grad_norm": 0.2916730046272278, "learning_rate": 1.3708999825116159e-05, "loss": 0.0281, "step": 61650 }, { "grad_norm": 0.3106381893157959, "learning_rate": 1.3694785469292526e-05, "loss": 0.022, "step": 61660 }, { "grad_norm": 0.2825448513031006, "learning_rate": 1.3680577317037101e-05, "loss": 0.0206, "step": 61670 }, { "grad_norm": 0.25984352827072144, "learning_rate": 1.3666375370777678e-05, "loss": 0.0139, "step": 61680 }, { "grad_norm": 0.3142218589782715, "learning_rate": 1.3652179632940981e-05, "loss": 0.0201, "step": 61690 }, { "grad_norm": 0.20512577891349792, "learning_rate": 1.3637990105952664e-05, "loss": 0.018, "step": 61700 }, { "grad_norm": 0.3056464195251465, "learning_rate": 1.3623806792237337e-05, "loss": 0.0189, "step": 61710 }, { "grad_norm": 0.1810392290353775, "learning_rate": 1.3609629694218551e-05, "loss": 0.0211, "step": 61720 }, { "grad_norm": 0.268768310546875, "learning_rate": 1.3595458814318762e-05, "loss": 0.0155, "step": 61730 }, { "grad_norm": 0.21900922060012817, "learning_rate": 1.3581294154959389e-05, "loss": 0.0149, "step": 61740 }, { "grad_norm": 0.583968997001648, "learning_rate": 1.3567135718560792e-05, "loss": 0.0201, "step": 61750 }, { "grad_norm": 0.26497459411621094, "learning_rate": 1.3552983507542261e-05, "loss": 0.0173, "step": 61760 }, { "grad_norm": 0.513507604598999, "learning_rate": 1.3538837524322023e-05, "loss": 0.0226, "step": 61770 }, { "grad_norm": 0.1651206910610199, "learning_rate": 1.3524697771317251e-05, "loss": 0.0205, "step": 61780 }, { "grad_norm": 0.14796790480613708, "learning_rate": 1.3510564250943997e-05, "loss": 0.0164, "step": 61790 }, { "grad_norm": 0.169406920671463, "learning_rate": 1.3496436965617353e-05, "loss": 0.0153, "step": 61800 }, { "grad_norm": 0.18929865956306458, "learning_rate": 1.3482315917751243e-05, "loss": 0.0223, "step": 61810 }, { "grad_norm": 0.33623382449150085, "learning_rate": 1.3468201109758582e-05, "loss": 0.0174, "step": 61820 }, { "grad_norm": 0.4779675006866455, "learning_rate": 1.3454092544051222e-05, "loss": 0.0192, "step": 61830 }, { "grad_norm": 0.25761914253234863, "learning_rate": 1.3439990223039878e-05, "loss": 0.0164, "step": 61840 }, { "grad_norm": 0.1829572319984436, "learning_rate": 1.3425894149134321e-05, "loss": 0.015, "step": 61850 }, { "grad_norm": 0.28861159086227417, "learning_rate": 1.3411804324743132e-05, "loss": 0.02, "step": 61860 }, { "grad_norm": 0.18716126680374146, "learning_rate": 1.3397720752273896e-05, "loss": 0.019, "step": 61870 }, { "grad_norm": 0.24069230258464813, "learning_rate": 1.3383643434133108e-05, "loss": 0.0146, "step": 61880 }, { "grad_norm": 1.0954079627990723, "learning_rate": 1.3369572372726196e-05, "loss": 0.0186, "step": 61890 }, { "grad_norm": 0.21062150597572327, "learning_rate": 1.3355507570457525e-05, "loss": 0.0164, "step": 61900 }, { "grad_norm": 0.29313236474990845, "learning_rate": 1.334144902973038e-05, "loss": 0.0159, "step": 61910 }, { "grad_norm": 0.2407456785440445, "learning_rate": 1.3327396752946997e-05, "loss": 0.0169, "step": 61920 }, { "grad_norm": 0.23712192475795746, "learning_rate": 1.3313350742508496e-05, "loss": 0.0213, "step": 61930 }, { "grad_norm": 0.17822878062725067, "learning_rate": 1.3299311000814973e-05, "loss": 0.0207, "step": 61940 }, { "grad_norm": 0.10896643996238708, "learning_rate": 1.3285277530265432e-05, "loss": 0.0183, "step": 61950 }, { "grad_norm": 0.32196512818336487, "learning_rate": 1.3271250333257813e-05, "loss": 0.0172, "step": 61960 }, { "grad_norm": 0.3103334307670593, "learning_rate": 1.3257229412188977e-05, "loss": 0.0112, "step": 61970 }, { "grad_norm": 0.3161102533340454, "learning_rate": 1.3243214769454726e-05, "loss": 0.0174, "step": 61980 }, { "grad_norm": 0.1793273389339447, "learning_rate": 1.3229206407449751e-05, "loss": 0.0125, "step": 61990 }, { "grad_norm": 0.21252396702766418, "learning_rate": 1.3215204328567742e-05, "loss": 0.0158, "step": 62000 }, { "grad_norm": 0.2855333387851715, "learning_rate": 1.3201208535201232e-05, "loss": 0.0185, "step": 62010 }, { "grad_norm": 0.24034033715724945, "learning_rate": 1.3187219029741737e-05, "loss": 0.0174, "step": 62020 }, { "grad_norm": 0.3225245773792267, "learning_rate": 1.3173235814579693e-05, "loss": 0.0157, "step": 62030 }, { "grad_norm": 0.25978541374206543, "learning_rate": 1.3159258892104398e-05, "loss": 0.0182, "step": 62040 }, { "grad_norm": 0.2973618805408478, "learning_rate": 1.3145288264704198e-05, "loss": 0.0167, "step": 62050 }, { "grad_norm": 0.1479216068983078, "learning_rate": 1.3131323934766237e-05, "loss": 0.0161, "step": 62060 }, { "grad_norm": 0.23334459960460663, "learning_rate": 1.311736590467666e-05, "loss": 0.0131, "step": 62070 }, { "grad_norm": 0.19652032852172852, "learning_rate": 1.3103414176820505e-05, "loss": 0.0168, "step": 62080 }, { "grad_norm": 0.3359741270542145, "learning_rate": 1.308946875358174e-05, "loss": 0.0283, "step": 62090 }, { "grad_norm": 0.37073618173599243, "learning_rate": 1.3075529637343276e-05, "loss": 0.0259, "step": 62100 }, { "grad_norm": 0.33044397830963135, "learning_rate": 1.3061596830486883e-05, "loss": 0.0181, "step": 62110 }, { "grad_norm": 0.16811631619930267, "learning_rate": 1.3047670335393353e-05, "loss": 0.0148, "step": 62120 }, { "grad_norm": 0.4556082785129547, "learning_rate": 1.3033750154442298e-05, "loss": 0.0235, "step": 62130 }, { "grad_norm": 0.23953665792942047, "learning_rate": 1.3019836290012316e-05, "loss": 0.0147, "step": 62140 }, { "grad_norm": 0.19922161102294922, "learning_rate": 1.30059287444809e-05, "loss": 0.017, "step": 62150 }, { "grad_norm": 0.465288370847702, "learning_rate": 1.299202752022447e-05, "loss": 0.0287, "step": 62160 }, { "grad_norm": 0.2115333378314972, "learning_rate": 1.2978132619618371e-05, "loss": 0.0149, "step": 62170 }, { "grad_norm": 0.23409555852413177, "learning_rate": 1.2964244045036866e-05, "loss": 0.0206, "step": 62180 }, { "grad_norm": 0.2836620807647705, "learning_rate": 1.295036179885311e-05, "loss": 0.0149, "step": 62190 }, { "grad_norm": 0.21689681708812714, "learning_rate": 1.293648588343922e-05, "loss": 0.0179, "step": 62200 }, { "grad_norm": 0.4077065587043762, "learning_rate": 1.2922616301166196e-05, "loss": 0.0161, "step": 62210 }, { "grad_norm": 0.27859723567962646, "learning_rate": 1.2908753054403976e-05, "loss": 0.0152, "step": 62220 }, { "grad_norm": 0.31615397334098816, "learning_rate": 1.2894896145521429e-05, "loss": 0.0218, "step": 62230 }, { "grad_norm": 0.2296808660030365, "learning_rate": 1.2881045576886275e-05, "loss": 0.0171, "step": 62240 }, { "grad_norm": 0.25201308727264404, "learning_rate": 1.2867201350865254e-05, "loss": 0.0182, "step": 62250 }, { "grad_norm": 0.2713209390640259, "learning_rate": 1.2853363469823914e-05, "loss": 0.0185, "step": 62260 }, { "grad_norm": 0.25746437907218933, "learning_rate": 1.283953193612682e-05, "loss": 0.0124, "step": 62270 }, { "grad_norm": 0.2976439297199249, "learning_rate": 1.2825706752137372e-05, "loss": 0.0158, "step": 62280 }, { "grad_norm": 0.39690664410591125, "learning_rate": 1.2811887920217896e-05, "loss": 0.0221, "step": 62290 }, { "grad_norm": 0.3761309087276459, "learning_rate": 1.2798075442729707e-05, "loss": 0.0171, "step": 62300 }, { "grad_norm": 0.24404583871364594, "learning_rate": 1.2784269322032922e-05, "loss": 0.02, "step": 62310 }, { "grad_norm": 0.16070042550563812, "learning_rate": 1.2770469560486686e-05, "loss": 0.0191, "step": 62320 }, { "grad_norm": 0.3594089150428772, "learning_rate": 1.2756676160448956e-05, "loss": 0.0194, "step": 62330 }, { "grad_norm": 0.3618527948856354, "learning_rate": 1.2742889124276663e-05, "loss": 0.0191, "step": 62340 }, { "grad_norm": 0.3324425518512726, "learning_rate": 1.2729108454325639e-05, "loss": 0.0169, "step": 62350 }, { "grad_norm": 0.3815461993217468, "learning_rate": 1.2715334152950614e-05, "loss": 0.0188, "step": 62360 }, { "grad_norm": 0.37803927063941956, "learning_rate": 1.2701566222505246e-05, "loss": 0.0163, "step": 62370 }, { "grad_norm": 0.3062388598918915, "learning_rate": 1.2687804665342107e-05, "loss": 0.0202, "step": 62380 }, { "grad_norm": 0.25367432832717896, "learning_rate": 1.267404948381265e-05, "loss": 0.0161, "step": 62390 }, { "grad_norm": 0.36688151955604553, "learning_rate": 1.2660300680267267e-05, "loss": 0.0128, "step": 62400 }, { "grad_norm": 0.4293707609176636, "learning_rate": 1.2646558257055257e-05, "loss": 0.0183, "step": 62410 }, { "grad_norm": 0.20091630518436432, "learning_rate": 1.263282221652482e-05, "loss": 0.0134, "step": 62420 }, { "grad_norm": 0.23649179935455322, "learning_rate": 1.2619092561023088e-05, "loss": 0.0141, "step": 62430 }, { "grad_norm": 0.21725435554981232, "learning_rate": 1.2605369292896036e-05, "loss": 0.0141, "step": 62440 }, { "grad_norm": 0.18525581061840057, "learning_rate": 1.2591652414488658e-05, "loss": 0.015, "step": 62450 }, { "grad_norm": 0.1633904129266739, "learning_rate": 1.2577941928144732e-05, "loss": 0.0126, "step": 62460 }, { "grad_norm": 0.41132134199142456, "learning_rate": 1.256423783620706e-05, "loss": 0.0168, "step": 62470 }, { "grad_norm": 0.21646706759929657, "learning_rate": 1.2550540141017264e-05, "loss": 0.0135, "step": 62480 }, { "grad_norm": 0.19565151631832123, "learning_rate": 1.253684884491591e-05, "loss": 0.0113, "step": 62490 }, { "grad_norm": 0.32140451669692993, "learning_rate": 1.2523163950242483e-05, "loss": 0.0136, "step": 62500 }, { "grad_norm": 0.28502917289733887, "learning_rate": 1.2509485459335313e-05, "loss": 0.0147, "step": 62510 }, { "grad_norm": 0.21303164958953857, "learning_rate": 1.2495813374531739e-05, "loss": 0.02, "step": 62520 }, { "grad_norm": 0.4194597899913788, "learning_rate": 1.2482147698167907e-05, "loss": 0.0158, "step": 62530 }, { "grad_norm": 0.1973615437746048, "learning_rate": 1.2468488432578911e-05, "loss": 0.0135, "step": 62540 }, { "grad_norm": 0.20066477358341217, "learning_rate": 1.2454835580098761e-05, "loss": 0.0169, "step": 62550 }, { "grad_norm": 0.20157364010810852, "learning_rate": 1.2441189143060338e-05, "loss": 0.0163, "step": 62560 }, { "grad_norm": 0.3341899514198303, "learning_rate": 1.242754912379548e-05, "loss": 0.0173, "step": 62570 }, { "grad_norm": 0.23108917474746704, "learning_rate": 1.2413915524634844e-05, "loss": 0.0249, "step": 62580 }, { "grad_norm": 0.2580057978630066, "learning_rate": 1.2400288347908073e-05, "loss": 0.0176, "step": 62590 }, { "grad_norm": 0.38369154930114746, "learning_rate": 1.2386667595943663e-05, "loss": 0.0174, "step": 62600 }, { "grad_norm": 0.19238600134849548, "learning_rate": 1.2373053271069035e-05, "loss": 0.0157, "step": 62610 }, { "grad_norm": 0.21484917402267456, "learning_rate": 1.2359445375610501e-05, "loss": 0.016, "step": 62620 }, { "grad_norm": 1.1631269454956055, "learning_rate": 1.2345843911893301e-05, "loss": 0.0173, "step": 62630 }, { "grad_norm": 0.24644547700881958, "learning_rate": 1.2332248882241498e-05, "loss": 0.0276, "step": 62640 }, { "grad_norm": 0.22542650997638702, "learning_rate": 1.2318660288978178e-05, "loss": 0.0164, "step": 62650 }, { "grad_norm": 0.3275107145309448, "learning_rate": 1.2305078134425213e-05, "loss": 0.0164, "step": 62660 }, { "grad_norm": 0.21001185476779938, "learning_rate": 1.2291502420903434e-05, "loss": 0.016, "step": 62670 }, { "grad_norm": 0.22074812650680542, "learning_rate": 1.2277933150732567e-05, "loss": 0.0154, "step": 62680 }, { "grad_norm": 0.11984089016914368, "learning_rate": 1.2264370326231216e-05, "loss": 0.0212, "step": 62690 }, { "grad_norm": 0.2264547199010849, "learning_rate": 1.2250813949716927e-05, "loss": 0.0152, "step": 62700 }, { "grad_norm": 0.2641107738018036, "learning_rate": 1.2237264023506063e-05, "loss": 0.0181, "step": 62710 }, { "grad_norm": 0.14446768164634705, "learning_rate": 1.2223720549913987e-05, "loss": 0.0143, "step": 62720 }, { "grad_norm": 0.5374935865402222, "learning_rate": 1.221018353125487e-05, "loss": 0.016, "step": 62730 }, { "grad_norm": 0.16902951896190643, "learning_rate": 1.2196652969841837e-05, "loss": 0.0175, "step": 62740 }, { "grad_norm": 0.37410223484039307, "learning_rate": 1.2183128867986904e-05, "loss": 0.0193, "step": 62750 }, { "grad_norm": 0.1647883951663971, "learning_rate": 1.2169611228000927e-05, "loss": 0.0141, "step": 62760 }, { "grad_norm": 0.2997024357318878, "learning_rate": 1.2156100052193752e-05, "loss": 0.0167, "step": 62770 }, { "grad_norm": 0.2586030960083008, "learning_rate": 1.214259534287403e-05, "loss": 0.0177, "step": 62780 }, { "grad_norm": 0.14864341914653778, "learning_rate": 1.2129097102349363e-05, "loss": 0.0158, "step": 62790 }, { "grad_norm": 0.24601732194423676, "learning_rate": 1.2115605332926227e-05, "loss": 0.0162, "step": 62800 }, { "grad_norm": 0.21790355443954468, "learning_rate": 1.2102120036910003e-05, "loss": 0.0284, "step": 62810 }, { "grad_norm": 0.24515855312347412, "learning_rate": 1.2088641216604956e-05, "loss": 0.0177, "step": 62820 }, { "grad_norm": 0.2131652683019638, "learning_rate": 1.2075168874314264e-05, "loss": 0.0185, "step": 62830 }, { "grad_norm": 0.2021835893392563, "learning_rate": 1.2061703012339942e-05, "loss": 0.0174, "step": 62840 }, { "grad_norm": 0.3221898078918457, "learning_rate": 1.204824363298297e-05, "loss": 0.0179, "step": 62850 }, { "grad_norm": 0.37786445021629333, "learning_rate": 1.2034790738543173e-05, "loss": 0.0182, "step": 62860 }, { "grad_norm": 0.2261970192193985, "learning_rate": 1.2021344331319284e-05, "loss": 0.0149, "step": 62870 }, { "grad_norm": 0.25624293088912964, "learning_rate": 1.200790441360894e-05, "loss": 0.0175, "step": 62880 }, { "grad_norm": 0.23698678612709045, "learning_rate": 1.199447098770864e-05, "loss": 0.0177, "step": 62890 }, { "grad_norm": 0.2691859006881714, "learning_rate": 1.198104405591381e-05, "loss": 0.0203, "step": 62900 }, { "grad_norm": 0.2166471779346466, "learning_rate": 1.1967623620518697e-05, "loss": 0.013, "step": 62910 }, { "grad_norm": 0.39798983931541443, "learning_rate": 1.1954209683816554e-05, "loss": 0.0174, "step": 62920 }, { "grad_norm": 0.2595937252044678, "learning_rate": 1.1940802248099402e-05, "loss": 0.0163, "step": 62930 }, { "grad_norm": 0.30607637763023376, "learning_rate": 1.1927401315658232e-05, "loss": 0.0145, "step": 62940 }, { "grad_norm": 0.3012508153915405, "learning_rate": 1.1914006888782898e-05, "loss": 0.0215, "step": 62950 }, { "grad_norm": 0.2731192111968994, "learning_rate": 1.1900618969762106e-05, "loss": 0.0171, "step": 62960 }, { "grad_norm": 0.27806344628334045, "learning_rate": 1.188723756088354e-05, "loss": 0.0128, "step": 62970 }, { "grad_norm": 0.24909336864948273, "learning_rate": 1.1873862664433672e-05, "loss": 0.0146, "step": 62980 }, { "grad_norm": 0.23457176983356476, "learning_rate": 1.1860494282697927e-05, "loss": 0.0148, "step": 62990 }, { "grad_norm": 0.2589491307735443, "learning_rate": 1.1847132417960588e-05, "loss": 0.0135, "step": 63000 }, { "grad_norm": 0.42278721928596497, "learning_rate": 1.1833777072504832e-05, "loss": 0.0182, "step": 63010 }, { "grad_norm": 0.24000492691993713, "learning_rate": 1.1820428248612731e-05, "loss": 0.0133, "step": 63020 }, { "grad_norm": 0.33266180753707886, "learning_rate": 1.1807085948565245e-05, "loss": 0.0201, "step": 63030 }, { "grad_norm": 0.18715977668762207, "learning_rate": 1.1793750174642172e-05, "loss": 0.0152, "step": 63040 }, { "grad_norm": 0.18860164284706116, "learning_rate": 1.1780420929122254e-05, "loss": 0.0177, "step": 63050 }, { "grad_norm": 0.2677530348300934, "learning_rate": 1.176709821428309e-05, "loss": 0.0178, "step": 63060 }, { "grad_norm": 0.2836453914642334, "learning_rate": 1.1753782032401173e-05, "loss": 0.0202, "step": 63070 }, { "grad_norm": 0.321682333946228, "learning_rate": 1.1740472385751866e-05, "loss": 0.0175, "step": 63080 }, { "grad_norm": 0.20452283322811127, "learning_rate": 1.172716927660943e-05, "loss": 0.0158, "step": 63090 }, { "grad_norm": 0.35245537757873535, "learning_rate": 1.1713872707247015e-05, "loss": 0.0201, "step": 63100 }, { "grad_norm": 0.2624383568763733, "learning_rate": 1.1700582679936595e-05, "loss": 0.0162, "step": 63110 }, { "grad_norm": 0.1932969093322754, "learning_rate": 1.1687299196949136e-05, "loss": 0.0238, "step": 63120 }, { "grad_norm": 0.3898608982563019, "learning_rate": 1.1674022260554374e-05, "loss": 0.0185, "step": 63130 }, { "grad_norm": 0.2631186544895172, "learning_rate": 1.1660751873020987e-05, "loss": 0.0167, "step": 63140 }, { "grad_norm": 0.2365065962076187, "learning_rate": 1.1647488036616538e-05, "loss": 0.0203, "step": 63150 }, { "grad_norm": 0.272798091173172, "learning_rate": 1.1634230753607417e-05, "loss": 0.0228, "step": 63160 }, { "grad_norm": 0.2828902006149292, "learning_rate": 1.1620980026258982e-05, "loss": 0.0149, "step": 63170 }, { "grad_norm": 0.22114403545856476, "learning_rate": 1.1607735856835373e-05, "loss": 0.0117, "step": 63180 }, { "grad_norm": 0.15938015282154083, "learning_rate": 1.1594498247599677e-05, "loss": 0.0154, "step": 63190 }, { "grad_norm": 0.41295671463012695, "learning_rate": 1.158126720081384e-05, "loss": 0.0224, "step": 63200 }, { "grad_norm": 0.41712868213653564, "learning_rate": 1.156804271873868e-05, "loss": 0.0182, "step": 63210 }, { "grad_norm": 0.31827643513679504, "learning_rate": 1.1554824803633924e-05, "loss": 0.0165, "step": 63220 }, { "grad_norm": 0.3214035928249359, "learning_rate": 1.154161345775811e-05, "loss": 0.0162, "step": 63230 }, { "grad_norm": 0.35979893803596497, "learning_rate": 1.1528408683368724e-05, "loss": 0.0145, "step": 63240 }, { "grad_norm": 0.2995612323284149, "learning_rate": 1.1515210482722088e-05, "loss": 0.0194, "step": 63250 }, { "grad_norm": 0.18395650386810303, "learning_rate": 1.150201885807342e-05, "loss": 0.0238, "step": 63260 }, { "grad_norm": 0.23157437145709991, "learning_rate": 1.1488833811676807e-05, "loss": 0.0181, "step": 63270 }, { "grad_norm": 0.29230570793151855, "learning_rate": 1.1475655345785213e-05, "loss": 0.0188, "step": 63280 }, { "grad_norm": 0.223676398396492, "learning_rate": 1.1462483462650481e-05, "loss": 0.0174, "step": 63290 }, { "grad_norm": 0.3052264451980591, "learning_rate": 1.1449318164523331e-05, "loss": 0.0202, "step": 63300 }, { "grad_norm": 0.3008245825767517, "learning_rate": 1.1436159453653334e-05, "loss": 0.0184, "step": 63310 }, { "grad_norm": 0.3811858594417572, "learning_rate": 1.1423007332288955e-05, "loss": 0.017, "step": 63320 }, { "grad_norm": 0.4099268913269043, "learning_rate": 1.1409861802677546e-05, "loss": 0.0167, "step": 63330 }, { "grad_norm": 0.2871021330356598, "learning_rate": 1.1396722867065313e-05, "loss": 0.0145, "step": 63340 }, { "grad_norm": 0.2572419345378876, "learning_rate": 1.1383590527697352e-05, "loss": 0.0192, "step": 63350 }, { "grad_norm": 0.18757393956184387, "learning_rate": 1.137046478681758e-05, "loss": 0.015, "step": 63360 }, { "grad_norm": 0.26909324526786804, "learning_rate": 1.1357345646668888e-05, "loss": 0.0304, "step": 63370 }, { "grad_norm": 0.272197961807251, "learning_rate": 1.1344233109492924e-05, "loss": 0.0169, "step": 63380 }, { "grad_norm": 0.22506862878799438, "learning_rate": 1.1331127177530292e-05, "loss": 0.0151, "step": 63390 }, { "grad_norm": 0.2808849811553955, "learning_rate": 1.1318027853020441e-05, "loss": 0.0144, "step": 63400 }, { "grad_norm": 0.2607607841491699, "learning_rate": 1.1304935138201645e-05, "loss": 0.0203, "step": 63410 }, { "grad_norm": 0.18497198820114136, "learning_rate": 1.1291849035311153e-05, "loss": 0.0148, "step": 63420 }, { "grad_norm": 0.23495736718177795, "learning_rate": 1.1278769546584972e-05, "loss": 0.0144, "step": 63430 }, { "grad_norm": 0.34834542870521545, "learning_rate": 1.1265696674258052e-05, "loss": 0.0131, "step": 63440 }, { "grad_norm": 0.19876641035079956, "learning_rate": 1.1252630420564186e-05, "loss": 0.0184, "step": 63450 }, { "grad_norm": 0.2690967619419098, "learning_rate": 1.1239570787736036e-05, "loss": 0.0176, "step": 63460 }, { "grad_norm": 0.4461914896965027, "learning_rate": 1.1226517778005135e-05, "loss": 0.0226, "step": 63470 }, { "grad_norm": 0.37316662073135376, "learning_rate": 1.1213471393601893e-05, "loss": 0.0129, "step": 63480 }, { "grad_norm": 0.5487045049667358, "learning_rate": 1.1200431636755587e-05, "loss": 0.027, "step": 63490 }, { "grad_norm": 0.33204108476638794, "learning_rate": 1.1187398509694336e-05, "loss": 0.0184, "step": 63500 }, { "grad_norm": 0.20742616057395935, "learning_rate": 1.1174372014645146e-05, "loss": 0.0151, "step": 63510 }, { "grad_norm": 0.21470117568969727, "learning_rate": 1.1161352153833899e-05, "loss": 0.0257, "step": 63520 }, { "grad_norm": 0.1410530060529709, "learning_rate": 1.1148338929485325e-05, "loss": 0.0172, "step": 63530 }, { "grad_norm": 0.2853304147720337, "learning_rate": 1.113533234382304e-05, "loss": 0.0186, "step": 63540 }, { "grad_norm": 0.2439045011997223, "learning_rate": 1.1122332399069513e-05, "loss": 0.0159, "step": 63550 }, { "grad_norm": 0.2433633953332901, "learning_rate": 1.1109339097446047e-05, "loss": 0.0172, "step": 63560 }, { "grad_norm": 0.35282668471336365, "learning_rate": 1.1096352441172897e-05, "loss": 0.0216, "step": 63570 }, { "grad_norm": 0.2772885262966156, "learning_rate": 1.1083372432469086e-05, "loss": 0.0164, "step": 63580 }, { "grad_norm": 0.22239699959754944, "learning_rate": 1.107039907355255e-05, "loss": 0.0242, "step": 63590 }, { "grad_norm": 0.31725576519966125, "learning_rate": 1.1057432366640103e-05, "loss": 0.0133, "step": 63600 }, { "grad_norm": 0.4244128167629242, "learning_rate": 1.1044472313947352e-05, "loss": 0.0234, "step": 63610 }, { "grad_norm": 0.25164350867271423, "learning_rate": 1.1031518917688877e-05, "loss": 0.018, "step": 63620 }, { "grad_norm": 0.4828771650791168, "learning_rate": 1.1018572180078007e-05, "loss": 0.0251, "step": 63630 }, { "grad_norm": 0.36271563172340393, "learning_rate": 1.1005632103327018e-05, "loss": 0.0148, "step": 63640 }, { "grad_norm": 0.19409918785095215, "learning_rate": 1.0992698689646996e-05, "loss": 0.0173, "step": 63650 }, { "grad_norm": 0.2929823100566864, "learning_rate": 1.0979771941247919e-05, "loss": 0.0222, "step": 63660 }, { "grad_norm": 0.25059202313423157, "learning_rate": 1.0966851860338611e-05, "loss": 0.0168, "step": 63670 }, { "grad_norm": 0.2732638418674469, "learning_rate": 1.0953938449126766e-05, "loss": 0.0214, "step": 63680 }, { "grad_norm": 0.5177809000015259, "learning_rate": 1.0941031709818933e-05, "loss": 0.0268, "step": 63690 }, { "grad_norm": 0.2749631404876709, "learning_rate": 1.0928131644620509e-05, "loss": 0.0251, "step": 63700 }, { "grad_norm": 0.21829509735107422, "learning_rate": 1.0915238255735766e-05, "loss": 0.0158, "step": 63710 }, { "grad_norm": 0.22570139169692993, "learning_rate": 1.0902351545367833e-05, "loss": 0.0155, "step": 63720 }, { "grad_norm": 0.24269357323646545, "learning_rate": 1.0889471515718702e-05, "loss": 0.0135, "step": 63730 }, { "grad_norm": 0.15012432634830475, "learning_rate": 1.087659816898921e-05, "loss": 0.0113, "step": 63740 }, { "grad_norm": 0.2206564098596573, "learning_rate": 1.0863731507379082e-05, "loss": 0.0145, "step": 63750 }, { "grad_norm": 0.29129719734191895, "learning_rate": 1.0850871533086827e-05, "loss": 0.0163, "step": 63760 }, { "grad_norm": 0.3051804006099701, "learning_rate": 1.0838018248309927e-05, "loss": 0.0132, "step": 63770 }, { "grad_norm": 0.39053449034690857, "learning_rate": 1.0825171655244615e-05, "loss": 0.0234, "step": 63780 }, { "grad_norm": 0.2706219255924225, "learning_rate": 1.0812331756086025e-05, "loss": 0.0168, "step": 63790 }, { "grad_norm": 0.25749966502189636, "learning_rate": 1.079949855302817e-05, "loss": 0.0157, "step": 63800 }, { "grad_norm": 0.3452991545200348, "learning_rate": 1.0786672048263852e-05, "loss": 0.0197, "step": 63810 }, { "grad_norm": 0.24587464332580566, "learning_rate": 1.0773852243984817e-05, "loss": 0.0215, "step": 63820 }, { "grad_norm": 0.25078222155570984, "learning_rate": 1.0761039142381586e-05, "loss": 0.0171, "step": 63830 }, { "grad_norm": 0.4914715886116028, "learning_rate": 1.0748232745643577e-05, "loss": 0.0208, "step": 63840 }, { "grad_norm": 0.2311461865901947, "learning_rate": 1.0735433055959055e-05, "loss": 0.0176, "step": 63850 }, { "grad_norm": 0.23868562281131744, "learning_rate": 1.0722640075515133e-05, "loss": 0.0144, "step": 63860 }, { "grad_norm": 0.24228964745998383, "learning_rate": 1.0709853806497795e-05, "loss": 0.0156, "step": 63870 }, { "grad_norm": 0.3089058995246887, "learning_rate": 1.0697074251091831e-05, "loss": 0.0216, "step": 63880 }, { "grad_norm": 0.20887166261672974, "learning_rate": 1.0684301411480962e-05, "loss": 0.0141, "step": 63890 }, { "grad_norm": 0.21054856479167938, "learning_rate": 1.067153528984769e-05, "loss": 0.0138, "step": 63900 }, { "grad_norm": 0.3178527057170868, "learning_rate": 1.0658775888373395e-05, "loss": 0.0158, "step": 63910 }, { "grad_norm": 0.2386292964220047, "learning_rate": 1.0646023209238314e-05, "loss": 0.0162, "step": 63920 }, { "grad_norm": 0.1798522025346756, "learning_rate": 1.0633277254621537e-05, "loss": 0.0121, "step": 63930 }, { "grad_norm": 0.29468902945518494, "learning_rate": 1.0620538026700994e-05, "loss": 0.0188, "step": 63940 }, { "grad_norm": 0.23073504865169525, "learning_rate": 1.0607805527653486e-05, "loss": 0.0161, "step": 63950 }, { "grad_norm": 0.49995970726013184, "learning_rate": 1.059507975965462e-05, "loss": 0.0188, "step": 63960 }, { "grad_norm": 0.16478055715560913, "learning_rate": 1.0582360724878898e-05, "loss": 0.0108, "step": 63970 }, { "grad_norm": 0.3902246952056885, "learning_rate": 1.0569648425499651e-05, "loss": 0.0215, "step": 63980 }, { "grad_norm": 0.2267196774482727, "learning_rate": 1.0556942863689063e-05, "loss": 0.0171, "step": 63990 }, { "grad_norm": 0.39210501313209534, "learning_rate": 1.054424404161819e-05, "loss": 0.0148, "step": 64000 }, { "grad_norm": 0.3247936964035034, "learning_rate": 1.053155196145686e-05, "loss": 0.0176, "step": 64010 }, { "grad_norm": 0.324434369802475, "learning_rate": 1.0518866625373863e-05, "loss": 0.0149, "step": 64020 }, { "grad_norm": 0.16690020263195038, "learning_rate": 1.0506188035536735e-05, "loss": 0.0186, "step": 64030 }, { "grad_norm": 0.2103331983089447, "learning_rate": 1.0493516194111919e-05, "loss": 0.0206, "step": 64040 }, { "grad_norm": 0.4214366376399994, "learning_rate": 1.0480851103264688e-05, "loss": 0.014, "step": 64050 }, { "grad_norm": 0.6148564219474792, "learning_rate": 1.046819276515913e-05, "loss": 0.0157, "step": 64060 }, { "grad_norm": 0.33010023832321167, "learning_rate": 1.0455541181958256e-05, "loss": 0.0146, "step": 64070 }, { "grad_norm": 0.18465740978717804, "learning_rate": 1.0442896355823822e-05, "loss": 0.0144, "step": 64080 }, { "grad_norm": 0.24868375062942505, "learning_rate": 1.0430258288916539e-05, "loss": 0.0132, "step": 64090 }, { "grad_norm": 0.1558198481798172, "learning_rate": 1.0417626983395868e-05, "loss": 0.0164, "step": 64100 }, { "grad_norm": 0.2053273320198059, "learning_rate": 1.0405002441420165e-05, "loss": 0.0171, "step": 64110 }, { "grad_norm": 0.2677065134048462, "learning_rate": 1.039238466514662e-05, "loss": 0.0187, "step": 64120 }, { "grad_norm": 0.1477670967578888, "learning_rate": 1.0379773656731262e-05, "loss": 0.0173, "step": 64130 }, { "grad_norm": 0.2577088475227356, "learning_rate": 1.0367169418328986e-05, "loss": 0.0153, "step": 64140 }, { "grad_norm": 0.37136080861091614, "learning_rate": 1.0354571952093484e-05, "loss": 0.0194, "step": 64150 }, { "grad_norm": 0.1740599423646927, "learning_rate": 1.034198126017733e-05, "loss": 0.0136, "step": 64160 }, { "grad_norm": 0.23537679016590118, "learning_rate": 1.0329397344731928e-05, "loss": 0.0173, "step": 64170 }, { "grad_norm": 0.24197600781917572, "learning_rate": 1.0316820207907524e-05, "loss": 0.0182, "step": 64180 }, { "grad_norm": 0.3667101263999939, "learning_rate": 1.0304249851853215e-05, "loss": 0.0194, "step": 64190 }, { "grad_norm": 0.3409103453159332, "learning_rate": 1.0291686278716933e-05, "loss": 0.0162, "step": 64200 }, { "grad_norm": 0.3152071237564087, "learning_rate": 1.0279129490645418e-05, "loss": 0.0189, "step": 64210 }, { "grad_norm": 0.2863869369029999, "learning_rate": 1.0266579489784328e-05, "loss": 0.0175, "step": 64220 }, { "grad_norm": 0.2599349915981293, "learning_rate": 1.0254036278278084e-05, "loss": 0.0204, "step": 64230 }, { "grad_norm": 0.20073071122169495, "learning_rate": 1.0241499858269982e-05, "loss": 0.0133, "step": 64240 }, { "grad_norm": 0.25717976689338684, "learning_rate": 1.0228970231902169e-05, "loss": 0.0174, "step": 64250 }, { "grad_norm": 0.25794604420661926, "learning_rate": 1.0216447401315582e-05, "loss": 0.0156, "step": 64260 }, { "grad_norm": 0.21831446886062622, "learning_rate": 1.020393136865007e-05, "loss": 0.0172, "step": 64270 }, { "grad_norm": 0.41839930415153503, "learning_rate": 1.0191422136044242e-05, "loss": 0.0282, "step": 64280 }, { "grad_norm": 0.36781689524650574, "learning_rate": 1.017891970563563e-05, "loss": 0.0172, "step": 64290 }, { "grad_norm": 0.18959398567676544, "learning_rate": 1.0166424079560516e-05, "loss": 0.0162, "step": 64300 }, { "grad_norm": 0.28082457184791565, "learning_rate": 1.0153935259954078e-05, "loss": 0.0121, "step": 64310 }, { "grad_norm": 0.2284751981496811, "learning_rate": 1.0141453248950311e-05, "loss": 0.0182, "step": 64320 }, { "grad_norm": 0.29028916358947754, "learning_rate": 1.0128978048682054e-05, "loss": 0.0173, "step": 64330 }, { "grad_norm": 0.2817510664463043, "learning_rate": 1.0116509661280982e-05, "loss": 0.0205, "step": 64340 }, { "grad_norm": 0.23748625814914703, "learning_rate": 1.0104048088877576e-05, "loss": 0.0137, "step": 64350 }, { "grad_norm": 0.2780531942844391, "learning_rate": 1.0091593333601201e-05, "loss": 0.0166, "step": 64360 }, { "grad_norm": 0.463874876499176, "learning_rate": 1.0079145397580031e-05, "loss": 0.0187, "step": 64370 }, { "grad_norm": 0.5545872449874878, "learning_rate": 1.006670428294107e-05, "loss": 0.0243, "step": 64380 }, { "grad_norm": 0.26173481345176697, "learning_rate": 1.0054269991810166e-05, "loss": 0.0178, "step": 64390 }, { "grad_norm": 0.16030749678611755, "learning_rate": 1.0041842526312024e-05, "loss": 0.0135, "step": 64400 }, { "grad_norm": 0.3329751789569855, "learning_rate": 1.0029421888570101e-05, "loss": 0.0119, "step": 64410 }, { "grad_norm": 0.25799667835235596, "learning_rate": 1.0017008080706813e-05, "loss": 0.0229, "step": 64420 }, { "grad_norm": 0.4152376353740692, "learning_rate": 1.0004601104843287e-05, "loss": 0.0146, "step": 64430 }, { "grad_norm": 0.33702224493026733, "learning_rate": 9.992200963099562e-06, "loss": 0.0179, "step": 64440 }, { "grad_norm": 0.17836113274097443, "learning_rate": 9.979807657594486e-06, "loss": 0.0156, "step": 64450 }, { "grad_norm": 0.3713698089122772, "learning_rate": 9.967421190445703e-06, "loss": 0.0137, "step": 64460 }, { "grad_norm": 0.1967012584209442, "learning_rate": 9.955041563769769e-06, "loss": 0.0175, "step": 64470 }, { "grad_norm": 0.5767173767089844, "learning_rate": 9.942668779681974e-06, "loss": 0.0178, "step": 64480 }, { "grad_norm": 0.2502671480178833, "learning_rate": 9.930302840296541e-06, "loss": 0.0127, "step": 64490 }, { "grad_norm": 0.17910391092300415, "learning_rate": 9.917943747726426e-06, "loss": 0.015, "step": 64500 }, { "grad_norm": 0.6994702816009521, "learning_rate": 9.905591504083484e-06, "loss": 0.0187, "step": 64510 }, { "grad_norm": 0.3103151321411133, "learning_rate": 9.893246111478382e-06, "loss": 0.0195, "step": 64520 }, { "grad_norm": 0.37502771615982056, "learning_rate": 9.88090757202057e-06, "loss": 0.0155, "step": 64530 }, { "grad_norm": 0.20350761711597443, "learning_rate": 9.868575887818421e-06, "loss": 0.0231, "step": 64540 }, { "grad_norm": 0.31563737988471985, "learning_rate": 9.856251060979044e-06, "loss": 0.0157, "step": 64550 }, { "grad_norm": 0.24019040167331696, "learning_rate": 9.843933093608426e-06, "loss": 0.0222, "step": 64560 }, { "grad_norm": 0.1942615956068039, "learning_rate": 9.831621987811368e-06, "loss": 0.0184, "step": 64570 }, { "grad_norm": 0.3808468282222748, "learning_rate": 9.819317745691509e-06, "loss": 0.0162, "step": 64580 }, { "grad_norm": 0.2293674200773239, "learning_rate": 9.8070203693513e-06, "loss": 0.0138, "step": 64590 }, { "grad_norm": 0.25332775712013245, "learning_rate": 9.794729860892048e-06, "loss": 0.0128, "step": 64600 }, { "grad_norm": 0.3475145101547241, "learning_rate": 9.782446222413827e-06, "loss": 0.0223, "step": 64610 }, { "grad_norm": 0.2846350371837616, "learning_rate": 9.770169456015598e-06, "loss": 0.0222, "step": 64620 }, { "grad_norm": 0.14950476586818695, "learning_rate": 9.75789956379512e-06, "loss": 0.0154, "step": 64630 }, { "grad_norm": 0.29145896434783936, "learning_rate": 9.74563654784898e-06, "loss": 0.0182, "step": 64640 }, { "grad_norm": 0.28052449226379395, "learning_rate": 9.733380410272596e-06, "loss": 0.0166, "step": 64650 }, { "grad_norm": 0.2214379906654358, "learning_rate": 9.721131153160207e-06, "loss": 0.0223, "step": 64660 }, { "grad_norm": 0.16283836960792542, "learning_rate": 9.708888778604886e-06, "loss": 0.0113, "step": 64670 }, { "grad_norm": 0.1821616142988205, "learning_rate": 9.696653288698488e-06, "loss": 0.0147, "step": 64680 }, { "grad_norm": 0.20318801701068878, "learning_rate": 9.684424685531763e-06, "loss": 0.0178, "step": 64690 }, { "grad_norm": 0.2530680298805237, "learning_rate": 9.672202971194216e-06, "loss": 0.021, "step": 64700 }, { "grad_norm": 0.12538401782512665, "learning_rate": 9.659988147774213e-06, "loss": 0.0148, "step": 64710 }, { "grad_norm": 0.24823656678199768, "learning_rate": 9.647780217358942e-06, "loss": 0.015, "step": 64720 }, { "grad_norm": 0.300834596157074, "learning_rate": 9.635579182034376e-06, "loss": 0.0128, "step": 64730 }, { "grad_norm": 0.40328526496887207, "learning_rate": 9.623385043885386e-06, "loss": 0.0211, "step": 64740 }, { "grad_norm": 0.20750556886196136, "learning_rate": 9.61119780499557e-06, "loss": 0.0117, "step": 64750 }, { "grad_norm": 0.23900561034679413, "learning_rate": 9.599017467447418e-06, "loss": 0.0161, "step": 64760 }, { "grad_norm": 0.21291495859622955, "learning_rate": 9.586844033322206e-06, "loss": 0.0135, "step": 64770 }, { "grad_norm": 0.3428187668323517, "learning_rate": 9.574677504700052e-06, "loss": 0.0187, "step": 64780 }, { "grad_norm": 0.2399308830499649, "learning_rate": 9.562517883659877e-06, "loss": 0.0243, "step": 64790 }, { "grad_norm": 0.28280436992645264, "learning_rate": 9.55036517227944e-06, "loss": 0.0153, "step": 64800 }, { "grad_norm": 0.24080556631088257, "learning_rate": 9.538219372635282e-06, "loss": 0.0161, "step": 64810 }, { "grad_norm": 0.4399721324443817, "learning_rate": 9.526080486802802e-06, "loss": 0.0193, "step": 64820 }, { "grad_norm": 0.20381410419940948, "learning_rate": 9.513948516856203e-06, "loss": 0.0151, "step": 64830 }, { "grad_norm": 0.21513986587524414, "learning_rate": 9.501823464868504e-06, "loss": 0.0149, "step": 64840 }, { "grad_norm": 0.17987874150276184, "learning_rate": 9.489705332911547e-06, "loss": 0.0139, "step": 64850 }, { "grad_norm": 0.25727134943008423, "learning_rate": 9.477594123055994e-06, "loss": 0.0146, "step": 64860 }, { "grad_norm": 0.2141689956188202, "learning_rate": 9.465489837371321e-06, "loss": 0.0226, "step": 64870 }, { "grad_norm": 0.24877838790416718, "learning_rate": 9.453392477925794e-06, "loss": 0.0125, "step": 64880 }, { "grad_norm": 0.20710189640522003, "learning_rate": 9.441302046786566e-06, "loss": 0.0133, "step": 64890 }, { "grad_norm": 0.5346659421920776, "learning_rate": 9.429218546019519e-06, "loss": 0.0155, "step": 64900 }, { "grad_norm": 0.6107189655303955, "learning_rate": 9.41714197768941e-06, "loss": 0.0265, "step": 64910 }, { "grad_norm": 0.2525344789028168, "learning_rate": 9.405072343859805e-06, "loss": 0.0212, "step": 64920 }, { "grad_norm": 0.22782151401042938, "learning_rate": 9.393009646593043e-06, "loss": 0.0135, "step": 64930 }, { "grad_norm": 0.25980016589164734, "learning_rate": 9.38095388795035e-06, "loss": 0.0154, "step": 64940 }, { "grad_norm": 0.4028307795524597, "learning_rate": 9.36890506999169e-06, "loss": 0.0143, "step": 64950 }, { "grad_norm": 0.20137915015220642, "learning_rate": 9.356863194775894e-06, "loss": 0.0147, "step": 64960 }, { "grad_norm": 0.3550245761871338, "learning_rate": 9.344828264360583e-06, "loss": 0.0156, "step": 64970 }, { "grad_norm": 0.2367393672466278, "learning_rate": 9.332800280802201e-06, "loss": 0.0165, "step": 64980 }, { "grad_norm": 0.2274729460477829, "learning_rate": 9.32077924615602e-06, "loss": 0.013, "step": 64990 }, { "grad_norm": 0.42954927682876587, "learning_rate": 9.308765162476063e-06, "loss": 0.0175, "step": 65000 }, { "grad_norm": 0.4200420379638672, "learning_rate": 9.296758031815239e-06, "loss": 0.0221, "step": 65010 }, { "grad_norm": 0.357928067445755, "learning_rate": 9.284757856225229e-06, "loss": 0.0133, "step": 65020 }, { "grad_norm": 0.23698152601718903, "learning_rate": 9.272764637756538e-06, "loss": 0.0175, "step": 65030 }, { "grad_norm": 0.3038569986820221, "learning_rate": 9.260778378458479e-06, "loss": 0.0172, "step": 65040 }, { "grad_norm": 0.21921323239803314, "learning_rate": 9.248799080379172e-06, "loss": 0.0186, "step": 65050 }, { "grad_norm": 0.2344379723072052, "learning_rate": 9.236826745565558e-06, "loss": 0.0214, "step": 65060 }, { "grad_norm": 0.24972373247146606, "learning_rate": 9.224861376063388e-06, "loss": 0.0163, "step": 65070 }, { "grad_norm": 0.1791289746761322, "learning_rate": 9.212902973917192e-06, "loss": 0.0136, "step": 65080 }, { "grad_norm": 0.19108812510967255, "learning_rate": 9.20095154117035e-06, "loss": 0.0141, "step": 65090 }, { "grad_norm": 0.2515420913696289, "learning_rate": 9.189007079865036e-06, "loss": 0.0138, "step": 65100 }, { "grad_norm": 0.2319505214691162, "learning_rate": 9.177069592042226e-06, "loss": 0.019, "step": 65110 }, { "grad_norm": 0.4502476751804352, "learning_rate": 9.165139079741724e-06, "loss": 0.0176, "step": 65120 }, { "grad_norm": 0.3333813548088074, "learning_rate": 9.153215545002098e-06, "loss": 0.0223, "step": 65130 }, { "grad_norm": 0.21351100504398346, "learning_rate": 9.141298989860798e-06, "loss": 0.0187, "step": 65140 }, { "grad_norm": 0.09384582191705704, "learning_rate": 9.129389416353994e-06, "loss": 0.011, "step": 65150 }, { "grad_norm": 0.24663440883159637, "learning_rate": 9.11748682651673e-06, "loss": 0.0128, "step": 65160 }, { "grad_norm": 0.2870166301727295, "learning_rate": 9.105591222382837e-06, "loss": 0.0144, "step": 65170 }, { "grad_norm": 0.16412167251110077, "learning_rate": 9.093702605984915e-06, "loss": 0.0195, "step": 65180 }, { "grad_norm": 0.2096939980983734, "learning_rate": 9.081820979354455e-06, "loss": 0.0158, "step": 65190 }, { "grad_norm": 0.2593206763267517, "learning_rate": 9.069946344521663e-06, "loss": 0.0167, "step": 65200 }, { "grad_norm": 0.20498795807361603, "learning_rate": 9.058078703515598e-06, "loss": 0.0146, "step": 65210 }, { "grad_norm": 0.21369363367557526, "learning_rate": 9.046218058364125e-06, "loss": 0.0222, "step": 65220 }, { "grad_norm": 0.7846679091453552, "learning_rate": 9.034364411093893e-06, "loss": 0.0156, "step": 65230 }, { "grad_norm": 0.22698485851287842, "learning_rate": 9.022517763730371e-06, "loss": 0.0181, "step": 65240 }, { "grad_norm": 0.326740562915802, "learning_rate": 9.010678118297827e-06, "loss": 0.0158, "step": 65250 }, { "grad_norm": 0.29565924406051636, "learning_rate": 8.998845476819345e-06, "loss": 0.017, "step": 65260 }, { "grad_norm": 0.32988885045051575, "learning_rate": 8.987019841316773e-06, "loss": 0.0158, "step": 65270 }, { "grad_norm": 0.19843557476997375, "learning_rate": 8.975201213810802e-06, "loss": 0.0127, "step": 65280 }, { "grad_norm": 0.23209188878536224, "learning_rate": 8.963389596320915e-06, "loss": 0.0188, "step": 65290 }, { "grad_norm": 0.19222410023212433, "learning_rate": 8.951584990865391e-06, "loss": 0.014, "step": 65300 }, { "grad_norm": 0.35103437304496765, "learning_rate": 8.939787399461319e-06, "loss": 0.0164, "step": 65310 }, { "grad_norm": 0.4060095250606537, "learning_rate": 8.927996824124591e-06, "loss": 0.0184, "step": 65320 }, { "grad_norm": 0.4188278317451477, "learning_rate": 8.916213266869854e-06, "loss": 0.0199, "step": 65330 }, { "grad_norm": 0.31621038913726807, "learning_rate": 8.904436729710658e-06, "loss": 0.0169, "step": 65340 }, { "grad_norm": 0.24448366463184357, "learning_rate": 8.892667214659245e-06, "loss": 0.0134, "step": 65350 }, { "grad_norm": 0.2329087257385254, "learning_rate": 8.880904723726713e-06, "loss": 0.0149, "step": 65360 }, { "grad_norm": 0.23170454800128937, "learning_rate": 8.869149258922971e-06, "loss": 0.0128, "step": 65370 }, { "grad_norm": 0.3207460343837738, "learning_rate": 8.857400822256662e-06, "loss": 0.0135, "step": 65380 }, { "grad_norm": 0.19865283370018005, "learning_rate": 8.845659415735324e-06, "loss": 0.0209, "step": 65390 }, { "grad_norm": 0.28764501214027405, "learning_rate": 8.83392504136521e-06, "loss": 0.0166, "step": 65400 }, { "grad_norm": 0.11383118480443954, "learning_rate": 8.822197701151407e-06, "loss": 0.0163, "step": 65410 }, { "grad_norm": 0.17864874005317688, "learning_rate": 8.810477397097804e-06, "loss": 0.0108, "step": 65420 }, { "grad_norm": 0.21390096843242645, "learning_rate": 8.798764131207077e-06, "loss": 0.0163, "step": 65430 }, { "grad_norm": 0.17731726169586182, "learning_rate": 8.787057905480706e-06, "loss": 0.0186, "step": 65440 }, { "grad_norm": 0.22265209257602692, "learning_rate": 8.775358721918958e-06, "loss": 0.0161, "step": 65450 }, { "grad_norm": 0.2529190480709076, "learning_rate": 8.763666582520923e-06, "loss": 0.0149, "step": 65460 }, { "grad_norm": 0.2403067797422409, "learning_rate": 8.751981489284445e-06, "loss": 0.0134, "step": 65470 }, { "grad_norm": 0.2693171501159668, "learning_rate": 8.740303444206188e-06, "loss": 0.0152, "step": 65480 }, { "grad_norm": 0.14843802154064178, "learning_rate": 8.72863244928162e-06, "loss": 0.011, "step": 65490 }, { "grad_norm": 0.2260286509990692, "learning_rate": 8.716968506504991e-06, "loss": 0.0215, "step": 65500 }, { "grad_norm": 0.255069762468338, "learning_rate": 8.70531161786935e-06, "loss": 0.0119, "step": 65510 }, { "grad_norm": 0.34236761927604675, "learning_rate": 8.693661785366558e-06, "loss": 0.0255, "step": 65520 }, { "grad_norm": 0.19976863265037537, "learning_rate": 8.682019010987208e-06, "loss": 0.0215, "step": 65530 }, { "grad_norm": 0.1422872096300125, "learning_rate": 8.670383296720786e-06, "loss": 0.0139, "step": 65540 }, { "grad_norm": 0.24653099477291107, "learning_rate": 8.658754644555478e-06, "loss": 0.0124, "step": 65550 }, { "grad_norm": 0.2517816424369812, "learning_rate": 8.647133056478313e-06, "loss": 0.0255, "step": 65560 }, { "grad_norm": 0.27635714411735535, "learning_rate": 8.635518534475123e-06, "loss": 0.019, "step": 65570 }, { "grad_norm": 0.21896430850028992, "learning_rate": 8.623911080530467e-06, "loss": 0.0184, "step": 65580 }, { "grad_norm": 0.3016926050186157, "learning_rate": 8.61231069662779e-06, "loss": 0.0191, "step": 65590 }, { "grad_norm": 0.2026699334383011, "learning_rate": 8.600717384749252e-06, "loss": 0.0154, "step": 65600 }, { "grad_norm": 0.2500942647457123, "learning_rate": 8.58913114687584e-06, "loss": 0.0139, "step": 65610 }, { "grad_norm": 0.347674697637558, "learning_rate": 8.57755198498732e-06, "loss": 0.0168, "step": 65620 }, { "grad_norm": 0.4275586009025574, "learning_rate": 8.565979901062265e-06, "loss": 0.03, "step": 65630 }, { "grad_norm": 0.2317688763141632, "learning_rate": 8.554414897078033e-06, "loss": 0.0156, "step": 65640 }, { "grad_norm": 0.20715008676052094, "learning_rate": 8.542856975010727e-06, "loss": 0.0128, "step": 65650 }, { "grad_norm": 0.22645539045333862, "learning_rate": 8.531306136835337e-06, "loss": 0.0152, "step": 65660 }, { "grad_norm": 0.20204026997089386, "learning_rate": 8.519762384525548e-06, "loss": 0.0167, "step": 65670 }, { "grad_norm": 0.43678414821624756, "learning_rate": 8.508225720053875e-06, "loss": 0.0175, "step": 65680 }, { "grad_norm": 0.3277323246002197, "learning_rate": 8.496696145391625e-06, "loss": 0.0135, "step": 65690 }, { "grad_norm": 0.28023865818977356, "learning_rate": 8.485173662508889e-06, "loss": 0.0122, "step": 65700 }, { "grad_norm": 0.26813623309135437, "learning_rate": 8.473658273374536e-06, "loss": 0.0151, "step": 65710 }, { "grad_norm": 0.2721710205078125, "learning_rate": 8.462149979956253e-06, "loss": 0.0137, "step": 65720 }, { "grad_norm": 0.384883314371109, "learning_rate": 8.450648784220461e-06, "loss": 0.0227, "step": 65730 }, { "grad_norm": 0.4442656338214874, "learning_rate": 8.439154688132417e-06, "loss": 0.0162, "step": 65740 }, { "grad_norm": 0.31714579463005066, "learning_rate": 8.427667693656143e-06, "loss": 0.014, "step": 65750 }, { "grad_norm": 0.23629747331142426, "learning_rate": 8.416187802754454e-06, "loss": 0.0176, "step": 65760 }, { "grad_norm": 0.21111898124217987, "learning_rate": 8.404715017388965e-06, "loss": 0.0167, "step": 65770 }, { "grad_norm": 0.18890875577926636, "learning_rate": 8.393249339520015e-06, "loss": 0.0169, "step": 65780 }, { "grad_norm": 0.1604517102241516, "learning_rate": 8.381790771106834e-06, "loss": 0.0162, "step": 65790 }, { "grad_norm": 0.2639438509941101, "learning_rate": 8.370339314107339e-06, "loss": 0.0111, "step": 65800 }, { "grad_norm": 0.19683173298835754, "learning_rate": 8.358894970478281e-06, "loss": 0.0125, "step": 65810 }, { "grad_norm": 0.1920863837003708, "learning_rate": 8.347457742175196e-06, "loss": 0.0149, "step": 65820 }, { "grad_norm": 0.33173319697380066, "learning_rate": 8.33602763115236e-06, "loss": 0.0163, "step": 65830 }, { "grad_norm": 0.2589060366153717, "learning_rate": 8.324604639362916e-06, "loss": 0.0179, "step": 65840 }, { "grad_norm": 0.2630505859851837, "learning_rate": 8.31318876875869e-06, "loss": 0.0139, "step": 65850 }, { "grad_norm": 0.2123727649450302, "learning_rate": 8.30178002129039e-06, "loss": 0.0118, "step": 65860 }, { "grad_norm": 0.24403466284275055, "learning_rate": 8.290378398907423e-06, "loss": 0.0134, "step": 65870 }, { "grad_norm": 0.26760947704315186, "learning_rate": 8.278983903558029e-06, "loss": 0.0231, "step": 65880 }, { "grad_norm": 0.29143378138542175, "learning_rate": 8.26759653718921e-06, "loss": 0.0154, "step": 65890 }, { "grad_norm": 0.2079913467168808, "learning_rate": 8.25621630174676e-06, "loss": 0.0124, "step": 65900 }, { "grad_norm": 0.20247921347618103, "learning_rate": 8.244843199175261e-06, "loss": 0.0189, "step": 65910 }, { "grad_norm": 0.1315285861492157, "learning_rate": 8.233477231418046e-06, "loss": 0.0122, "step": 65920 }, { "grad_norm": 0.3931730389595032, "learning_rate": 8.22211840041725e-06, "loss": 0.02, "step": 65930 }, { "grad_norm": 0.4111490845680237, "learning_rate": 8.210766708113792e-06, "loss": 0.0205, "step": 65940 }, { "grad_norm": 0.3345244228839874, "learning_rate": 8.199422156447367e-06, "loss": 0.0185, "step": 65950 }, { "grad_norm": 0.2800315320491791, "learning_rate": 8.188084747356451e-06, "loss": 0.0193, "step": 65960 }, { "grad_norm": 0.23825225234031677, "learning_rate": 8.176754482778299e-06, "loss": 0.0122, "step": 65970 }, { "grad_norm": 0.29068824648857117, "learning_rate": 8.165431364648918e-06, "loss": 0.0217, "step": 65980 }, { "grad_norm": 0.24965932965278625, "learning_rate": 8.154115394903162e-06, "loss": 0.0188, "step": 65990 }, { "grad_norm": 0.2426062673330307, "learning_rate": 8.142806575474582e-06, "loss": 0.0159, "step": 66000 }, { "grad_norm": 0.32095468044281006, "learning_rate": 8.131504908295562e-06, "loss": 0.0154, "step": 66010 }, { "grad_norm": 0.26555266976356506, "learning_rate": 8.120210395297262e-06, "loss": 0.011, "step": 66020 }, { "grad_norm": 0.20020903646945953, "learning_rate": 8.108923038409565e-06, "loss": 0.0168, "step": 66030 }, { "grad_norm": 0.4193006753921509, "learning_rate": 8.097642839561226e-06, "loss": 0.0176, "step": 66040 }, { "grad_norm": 0.34610041975975037, "learning_rate": 8.08636980067966e-06, "loss": 0.0157, "step": 66050 }, { "grad_norm": 0.1842603087425232, "learning_rate": 8.075103923691186e-06, "loss": 0.0125, "step": 66060 }, { "grad_norm": 0.21037693321704865, "learning_rate": 8.063845210520793e-06, "loss": 0.0167, "step": 66070 }, { "grad_norm": 0.25786301493644714, "learning_rate": 8.052593663092295e-06, "loss": 0.0196, "step": 66080 }, { "grad_norm": 0.25906896591186523, "learning_rate": 8.04134928332827e-06, "loss": 0.0145, "step": 66090 }, { "grad_norm": 0.2613047957420349, "learning_rate": 8.030112073150086e-06, "loss": 0.0159, "step": 66100 }, { "grad_norm": 0.25141921639442444, "learning_rate": 8.018882034477881e-06, "loss": 0.0153, "step": 66110 }, { "grad_norm": 0.2886759638786316, "learning_rate": 8.007659169230541e-06, "loss": 0.0151, "step": 66120 }, { "grad_norm": 0.4863952398300171, "learning_rate": 7.996443479325755e-06, "loss": 0.0201, "step": 66130 }, { "grad_norm": 0.12500576674938202, "learning_rate": 7.985234966679977e-06, "loss": 0.0134, "step": 66140 }, { "grad_norm": 0.2311258763074875, "learning_rate": 7.974033633208438e-06, "loss": 0.0125, "step": 66150 }, { "grad_norm": 0.19332677125930786, "learning_rate": 7.962839480825135e-06, "loss": 0.0167, "step": 66160 }, { "grad_norm": 0.5061447024345398, "learning_rate": 7.951652511442858e-06, "loss": 0.018, "step": 66170 }, { "grad_norm": 0.25065740942955017, "learning_rate": 7.940472726973125e-06, "loss": 0.0178, "step": 66180 }, { "grad_norm": 0.32697805762290955, "learning_rate": 7.929300129326289e-06, "loss": 0.0182, "step": 66190 }, { "grad_norm": 0.2717570662498474, "learning_rate": 7.91813472041142e-06, "loss": 0.014, "step": 66200 }, { "grad_norm": 0.2210778295993805, "learning_rate": 7.906976502136376e-06, "loss": 0.0162, "step": 66210 }, { "grad_norm": 0.250993013381958, "learning_rate": 7.89582547640782e-06, "loss": 0.0194, "step": 66220 }, { "grad_norm": 0.32821911573410034, "learning_rate": 7.884681645131115e-06, "loss": 0.0167, "step": 66230 }, { "grad_norm": 0.5304642915725708, "learning_rate": 7.87354501021048e-06, "loss": 0.0164, "step": 66240 }, { "grad_norm": 0.28927889466285706, "learning_rate": 7.86241557354882e-06, "loss": 0.0149, "step": 66250 }, { "grad_norm": 0.21912765502929688, "learning_rate": 7.8512933370479e-06, "loss": 0.0203, "step": 66260 }, { "grad_norm": 0.2443932443857193, "learning_rate": 7.840178302608158e-06, "loss": 0.0151, "step": 66270 }, { "grad_norm": 0.4864855706691742, "learning_rate": 7.82907047212888e-06, "loss": 0.018, "step": 66280 }, { "grad_norm": 0.26337969303131104, "learning_rate": 7.81796984750809e-06, "loss": 0.014, "step": 66290 }, { "grad_norm": 0.27896448969841003, "learning_rate": 7.806876430642546e-06, "loss": 0.016, "step": 66300 }, { "grad_norm": 0.4028293788433075, "learning_rate": 7.795790223427862e-06, "loss": 0.0164, "step": 66310 }, { "grad_norm": 0.37326231598854065, "learning_rate": 7.784711227758324e-06, "loss": 0.0188, "step": 66320 }, { "grad_norm": 0.31598562002182007, "learning_rate": 7.773639445527053e-06, "loss": 0.0166, "step": 66330 }, { "grad_norm": 0.22890529036521912, "learning_rate": 7.762574878625905e-06, "loss": 0.0213, "step": 66340 }, { "grad_norm": 0.6822071671485901, "learning_rate": 7.751517528945513e-06, "loss": 0.0197, "step": 66350 }, { "grad_norm": 0.4404529333114624, "learning_rate": 7.740467398375278e-06, "loss": 0.0201, "step": 66360 }, { "grad_norm": 0.1622535139322281, "learning_rate": 7.729424488803378e-06, "loss": 0.0169, "step": 66370 }, { "grad_norm": 0.27451491355895996, "learning_rate": 7.71838880211671e-06, "loss": 0.0209, "step": 66380 }, { "grad_norm": 0.3657079339027405, "learning_rate": 7.707360340200997e-06, "loss": 0.0101, "step": 66390 }, { "grad_norm": 0.20862631499767303, "learning_rate": 7.696339104940697e-06, "loss": 0.0168, "step": 66400 }, { "grad_norm": 0.1657666563987732, "learning_rate": 7.685325098219038e-06, "loss": 0.0151, "step": 66410 }, { "grad_norm": 0.18202102184295654, "learning_rate": 7.674318321918017e-06, "loss": 0.0166, "step": 66420 }, { "grad_norm": 0.18023666739463806, "learning_rate": 7.663318777918366e-06, "loss": 0.0182, "step": 66430 }, { "grad_norm": 0.23553167283535004, "learning_rate": 7.652326468099647e-06, "loss": 0.016, "step": 66440 }, { "grad_norm": 0.1895611584186554, "learning_rate": 7.641341394340096e-06, "loss": 0.0112, "step": 66450 }, { "grad_norm": 0.33800557255744934, "learning_rate": 7.630363558516818e-06, "loss": 0.0196, "step": 66460 }, { "grad_norm": 0.18402963876724243, "learning_rate": 7.619392962505578e-06, "loss": 0.018, "step": 66470 }, { "grad_norm": 0.22236397862434387, "learning_rate": 7.6084296081809725e-06, "loss": 0.0133, "step": 66480 }, { "grad_norm": 0.4570190906524658, "learning_rate": 7.597473497416347e-06, "loss": 0.0179, "step": 66490 }, { "grad_norm": 0.20722800493240356, "learning_rate": 7.586524632083764e-06, "loss": 0.0124, "step": 66500 }, { "grad_norm": 0.31315159797668457, "learning_rate": 7.5755830140541326e-06, "loss": 0.0144, "step": 66510 }, { "grad_norm": 0.17986637353897095, "learning_rate": 7.5646486451970475e-06, "loss": 0.0144, "step": 66520 }, { "grad_norm": 0.1674865037202835, "learning_rate": 7.553721527380897e-06, "loss": 0.0176, "step": 66530 }, { "grad_norm": 0.23164737224578857, "learning_rate": 7.542801662472826e-06, "loss": 0.0151, "step": 66540 }, { "grad_norm": 0.21141473948955536, "learning_rate": 7.5318890523387475e-06, "loss": 0.0141, "step": 66550 }, { "grad_norm": 0.18181172013282776, "learning_rate": 7.52098369884332e-06, "loss": 0.0124, "step": 66560 }, { "grad_norm": 0.244593545794487, "learning_rate": 7.510085603849992e-06, "loss": 0.0209, "step": 66570 }, { "grad_norm": 0.1912321001291275, "learning_rate": 7.499194769220918e-06, "loss": 0.0136, "step": 66580 }, { "grad_norm": 0.398888498544693, "learning_rate": 7.48831119681706e-06, "loss": 0.0225, "step": 66590 }, { "grad_norm": 0.408434122800827, "learning_rate": 7.477434888498119e-06, "loss": 0.0165, "step": 66600 }, { "grad_norm": 0.26353687047958374, "learning_rate": 7.466565846122564e-06, "loss": 0.019, "step": 66610 }, { "grad_norm": 0.25310271978378296, "learning_rate": 7.455704071547626e-06, "loss": 0.015, "step": 66620 }, { "grad_norm": 0.19831544160842896, "learning_rate": 7.444849566629247e-06, "loss": 0.0137, "step": 66630 }, { "grad_norm": 0.20596906542778015, "learning_rate": 7.434002333222212e-06, "loss": 0.0137, "step": 66640 }, { "grad_norm": 0.3344584107398987, "learning_rate": 7.423162373179976e-06, "loss": 0.0173, "step": 66650 }, { "grad_norm": 0.2751316726207733, "learning_rate": 7.412329688354835e-06, "loss": 0.0163, "step": 66660 }, { "grad_norm": 0.21083615720272064, "learning_rate": 7.40150428059776e-06, "loss": 0.0135, "step": 66670 }, { "grad_norm": 0.35942110419273376, "learning_rate": 7.3906861517585354e-06, "loss": 0.0161, "step": 66680 }, { "grad_norm": 0.21353112161159515, "learning_rate": 7.37987530368569e-06, "loss": 0.0162, "step": 66690 }, { "grad_norm": 0.21284504234790802, "learning_rate": 7.369071738226474e-06, "loss": 0.017, "step": 66700 }, { "grad_norm": 0.24642512202262878, "learning_rate": 7.358275457226954e-06, "loss": 0.0137, "step": 66710 }, { "grad_norm": 0.3257758617401123, "learning_rate": 7.347486462531899e-06, "loss": 0.0155, "step": 66720 }, { "grad_norm": 0.25837963819503784, "learning_rate": 7.336704755984858e-06, "loss": 0.0127, "step": 66730 }, { "grad_norm": 0.1719171702861786, "learning_rate": 7.325930339428133e-06, "loss": 0.0151, "step": 66740 }, { "grad_norm": 0.6625431776046753, "learning_rate": 7.315163214702769e-06, "loss": 0.018, "step": 66750 }, { "grad_norm": 0.31902173161506653, "learning_rate": 7.304403383648595e-06, "loss": 0.0143, "step": 66760 }, { "grad_norm": 0.16955989599227905, "learning_rate": 7.293650848104139e-06, "loss": 0.0133, "step": 66770 }, { "grad_norm": 0.2934669852256775, "learning_rate": 7.2829056099067374e-06, "loss": 0.0153, "step": 66780 }, { "grad_norm": 0.3715202808380127, "learning_rate": 7.2721676708924494e-06, "loss": 0.0141, "step": 66790 }, { "grad_norm": 0.47028017044067383, "learning_rate": 7.261437032896096e-06, "loss": 0.026, "step": 66800 }, { "grad_norm": 0.3096470534801483, "learning_rate": 7.250713697751255e-06, "loss": 0.018, "step": 66810 }, { "grad_norm": 0.19932010769844055, "learning_rate": 7.239997667290255e-06, "loss": 0.0144, "step": 66820 }, { "grad_norm": 0.11978494375944138, "learning_rate": 7.2292889433441425e-06, "loss": 0.013, "step": 66830 }, { "grad_norm": 0.2859113812446594, "learning_rate": 7.218587527742793e-06, "loss": 0.0132, "step": 66840 }, { "grad_norm": 0.233157679438591, "learning_rate": 7.207893422314749e-06, "loss": 0.0167, "step": 66850 }, { "grad_norm": 0.1578710675239563, "learning_rate": 7.1972066288873545e-06, "loss": 0.0153, "step": 66860 }, { "grad_norm": 0.7288541197776794, "learning_rate": 7.186527149286687e-06, "loss": 0.03, "step": 66870 }, { "grad_norm": 0.22513046860694885, "learning_rate": 7.175854985337576e-06, "loss": 0.0152, "step": 66880 }, { "grad_norm": 0.30604231357574463, "learning_rate": 7.16519013886362e-06, "loss": 0.018, "step": 66890 }, { "grad_norm": 0.2875951826572418, "learning_rate": 7.154532611687109e-06, "loss": 0.0168, "step": 66900 }, { "grad_norm": 0.37635737657546997, "learning_rate": 7.143882405629177e-06, "loss": 0.019, "step": 66910 }, { "grad_norm": 0.14578424394130707, "learning_rate": 7.133239522509605e-06, "loss": 0.0119, "step": 66920 }, { "grad_norm": 0.1738225370645523, "learning_rate": 7.1226039641469956e-06, "loss": 0.0118, "step": 66930 }, { "grad_norm": 0.17403574287891388, "learning_rate": 7.111975732358678e-06, "loss": 0.0096, "step": 66940 }, { "grad_norm": 0.10213946551084518, "learning_rate": 7.101354828960693e-06, "loss": 0.0139, "step": 66950 }, { "grad_norm": 0.3777185380458832, "learning_rate": 7.090741255767918e-06, "loss": 0.0211, "step": 66960 }, { "grad_norm": 0.24926894903182983, "learning_rate": 7.080135014593875e-06, "loss": 0.0145, "step": 66970 }, { "grad_norm": 0.29881346225738525, "learning_rate": 7.069536107250896e-06, "loss": 0.0124, "step": 66980 }, { "grad_norm": 0.2573882043361664, "learning_rate": 7.058944535550049e-06, "loss": 0.0136, "step": 66990 }, { "grad_norm": 0.233493834733963, "learning_rate": 7.048360301301138e-06, "loss": 0.0146, "step": 67000 }, { "grad_norm": 0.20161579549312592, "learning_rate": 7.03778340631272e-06, "loss": 0.017, "step": 67010 }, { "grad_norm": 0.2811221480369568, "learning_rate": 7.0272138523921e-06, "loss": 0.0161, "step": 67020 }, { "grad_norm": 0.24478156864643097, "learning_rate": 7.016651641345334e-06, "loss": 0.0217, "step": 67030 }, { "grad_norm": 0.3060799837112427, "learning_rate": 7.0060967749772e-06, "loss": 0.015, "step": 67040 }, { "grad_norm": 0.435418039560318, "learning_rate": 6.995549255091238e-06, "loss": 0.0265, "step": 67050 }, { "grad_norm": 0.2172946333885193, "learning_rate": 6.98500908348973e-06, "loss": 0.015, "step": 67060 }, { "grad_norm": 0.21759532392024994, "learning_rate": 6.974476261973711e-06, "loss": 0.0136, "step": 67070 }, { "grad_norm": 0.44533151388168335, "learning_rate": 6.963950792342949e-06, "loss": 0.0137, "step": 67080 }, { "grad_norm": 0.200670525431633, "learning_rate": 6.9534326763959715e-06, "loss": 0.0095, "step": 67090 }, { "grad_norm": 0.1776338815689087, "learning_rate": 6.942921915929995e-06, "loss": 0.0114, "step": 67100 }, { "grad_norm": 0.12617610394954681, "learning_rate": 6.9324185127410735e-06, "loss": 0.0211, "step": 67110 }, { "grad_norm": 0.28555819392204285, "learning_rate": 6.921922468623921e-06, "loss": 0.0147, "step": 67120 }, { "grad_norm": 0.21049878001213074, "learning_rate": 6.911433785372023e-06, "loss": 0.0169, "step": 67130 }, { "grad_norm": 0.2343442738056183, "learning_rate": 6.900952464777632e-06, "loss": 0.0203, "step": 67140 }, { "grad_norm": 0.2618185579776764, "learning_rate": 6.8904785086316815e-06, "loss": 0.0114, "step": 67150 }, { "grad_norm": 0.5072823166847229, "learning_rate": 6.880011918723927e-06, "loss": 0.0182, "step": 67160 }, { "grad_norm": 0.21868474781513214, "learning_rate": 6.86955269684279e-06, "loss": 0.0139, "step": 67170 }, { "grad_norm": 0.3535196781158447, "learning_rate": 6.859100844775473e-06, "loss": 0.0279, "step": 67180 }, { "grad_norm": 0.20273703336715698, "learning_rate": 6.84865636430792e-06, "loss": 0.0156, "step": 67190 }, { "grad_norm": 0.26328858733177185, "learning_rate": 6.838219257224804e-06, "loss": 0.0129, "step": 67200 }, { "grad_norm": 0.23006677627563477, "learning_rate": 6.827789525309536e-06, "loss": 0.0177, "step": 67210 }, { "grad_norm": 0.33647289872169495, "learning_rate": 6.81736717034428e-06, "loss": 0.0188, "step": 67220 }, { "grad_norm": 0.24790571630001068, "learning_rate": 6.806952194109933e-06, "loss": 0.0248, "step": 67230 }, { "grad_norm": 0.14802303910255432, "learning_rate": 6.796544598386112e-06, "loss": 0.012, "step": 67240 }, { "grad_norm": 0.5669457316398621, "learning_rate": 6.786144384951204e-06, "loss": 0.0211, "step": 67250 }, { "grad_norm": 0.19344420731067657, "learning_rate": 6.775751555582322e-06, "loss": 0.0182, "step": 67260 }, { "grad_norm": 0.6076120138168335, "learning_rate": 6.76536611205531e-06, "loss": 0.0182, "step": 67270 }, { "grad_norm": 0.27057385444641113, "learning_rate": 6.754988056144762e-06, "loss": 0.0178, "step": 67280 }, { "grad_norm": 0.19372116029262543, "learning_rate": 6.744617389624014e-06, "loss": 0.0126, "step": 67290 }, { "grad_norm": 0.3374011516571045, "learning_rate": 6.734254114265087e-06, "loss": 0.015, "step": 67300 }, { "grad_norm": 0.26220932602882385, "learning_rate": 6.723898231838843e-06, "loss": 0.013, "step": 67310 }, { "grad_norm": 0.2305428385734558, "learning_rate": 6.713549744114766e-06, "loss": 0.0166, "step": 67320 }, { "grad_norm": 0.244368314743042, "learning_rate": 6.703208652861159e-06, "loss": 0.0151, "step": 67330 }, { "grad_norm": 0.1733650118112564, "learning_rate": 6.69287495984503e-06, "loss": 0.0184, "step": 67340 }, { "grad_norm": 0.2685113847255707, "learning_rate": 6.6825486668321e-06, "loss": 0.0134, "step": 67350 }, { "grad_norm": 0.2644938826560974, "learning_rate": 6.672229775586886e-06, "loss": 0.0181, "step": 67360 }, { "grad_norm": 0.23372916877269745, "learning_rate": 6.661918287872576e-06, "loss": 0.0149, "step": 67370 }, { "grad_norm": 0.21636155247688293, "learning_rate": 6.6516142054511346e-06, "loss": 0.0171, "step": 67380 }, { "grad_norm": 0.2544962167739868, "learning_rate": 6.641317530083241e-06, "loss": 0.0126, "step": 67390 }, { "grad_norm": 0.4109073877334595, "learning_rate": 6.631028263528322e-06, "loss": 0.0177, "step": 67400 }, { "grad_norm": 0.1807810366153717, "learning_rate": 6.620746407544537e-06, "loss": 0.016, "step": 67410 }, { "grad_norm": 0.2641541063785553, "learning_rate": 6.610471963888742e-06, "loss": 0.0218, "step": 67420 }, { "grad_norm": 0.19155529141426086, "learning_rate": 6.600204934316606e-06, "loss": 0.0103, "step": 67430 }, { "grad_norm": 0.2402494102716446, "learning_rate": 6.589945320582452e-06, "loss": 0.0131, "step": 67440 }, { "grad_norm": 0.3478727340698242, "learning_rate": 6.579693124439374e-06, "loss": 0.0163, "step": 67450 }, { "grad_norm": 0.23726949095726013, "learning_rate": 6.56944834763919e-06, "loss": 0.0113, "step": 67460 }, { "grad_norm": 0.26490306854248047, "learning_rate": 6.5592109919324575e-06, "loss": 0.0138, "step": 67470 }, { "grad_norm": 0.19151298701763153, "learning_rate": 6.54898105906846e-06, "loss": 0.0176, "step": 67480 }, { "grad_norm": 0.22535355389118195, "learning_rate": 6.5387585507952155e-06, "loss": 0.0234, "step": 67490 }, { "grad_norm": 0.2837262451648712, "learning_rate": 6.528543468859461e-06, "loss": 0.0146, "step": 67500 }, { "grad_norm": 0.18501272797584534, "learning_rate": 6.518335815006682e-06, "loss": 0.0187, "step": 67510 }, { "grad_norm": 0.3169989585876465, "learning_rate": 6.5081355909810845e-06, "loss": 0.0149, "step": 67520 }, { "grad_norm": 0.2283203899860382, "learning_rate": 6.497942798525608e-06, "loss": 0.0187, "step": 67530 }, { "grad_norm": 0.43214040994644165, "learning_rate": 6.487757439381936e-06, "loss": 0.0198, "step": 67540 }, { "grad_norm": 0.5806844830513, "learning_rate": 6.477579515290433e-06, "loss": 0.018, "step": 67550 }, { "grad_norm": 0.21633116900920868, "learning_rate": 6.467409027990273e-06, "loss": 0.019, "step": 67560 }, { "grad_norm": 0.21101707220077515, "learning_rate": 6.457245979219279e-06, "loss": 0.0213, "step": 67570 }, { "grad_norm": 0.3549993932247162, "learning_rate": 6.447090370714054e-06, "loss": 0.0252, "step": 67580 }, { "grad_norm": 0.2644118070602417, "learning_rate": 6.436942204209917e-06, "loss": 0.0218, "step": 67590 }, { "grad_norm": 0.25551941990852356, "learning_rate": 6.4268014814408804e-06, "loss": 0.0245, "step": 67600 }, { "grad_norm": 0.3303214907646179, "learning_rate": 6.41666820413977e-06, "loss": 0.0174, "step": 67610 }, { "grad_norm": 0.3008260428905487, "learning_rate": 6.406542374038033e-06, "loss": 0.0199, "step": 67620 }, { "grad_norm": 0.25374317169189453, "learning_rate": 6.396423992865935e-06, "loss": 0.0181, "step": 67630 }, { "grad_norm": 0.2868001461029053, "learning_rate": 6.386313062352412e-06, "loss": 0.0148, "step": 67640 }, { "grad_norm": 0.233977809548378, "learning_rate": 6.376209584225152e-06, "loss": 0.0149, "step": 67650 }, { "grad_norm": 0.2505989372730255, "learning_rate": 6.366113560210557e-06, "loss": 0.0107, "step": 67660 }, { "grad_norm": 0.4115070402622223, "learning_rate": 6.356024992033766e-06, "loss": 0.0235, "step": 67670 }, { "grad_norm": 0.2551666498184204, "learning_rate": 6.345943881418648e-06, "loss": 0.0142, "step": 67680 }, { "grad_norm": 0.17425423860549927, "learning_rate": 6.33587023008777e-06, "loss": 0.0149, "step": 67690 }, { "grad_norm": 0.28828781843185425, "learning_rate": 6.32580403976245e-06, "loss": 0.018, "step": 67700 }, { "grad_norm": 0.2572808861732483, "learning_rate": 6.31574531216273e-06, "loss": 0.0162, "step": 67710 }, { "grad_norm": 0.24098734557628632, "learning_rate": 6.305694049007371e-06, "loss": 0.0193, "step": 67720 }, { "grad_norm": 0.26256585121154785, "learning_rate": 6.2956502520138575e-06, "loss": 0.0161, "step": 67730 }, { "grad_norm": 0.22472895681858063, "learning_rate": 6.285613922898409e-06, "loss": 0.0117, "step": 67740 }, { "grad_norm": 0.37095797061920166, "learning_rate": 6.275585063375927e-06, "loss": 0.0202, "step": 67750 }, { "grad_norm": 0.1662856489419937, "learning_rate": 6.265563675160113e-06, "loss": 0.0139, "step": 67760 }, { "grad_norm": 0.2805190980434418, "learning_rate": 6.25554975996332e-06, "loss": 0.0123, "step": 67770 }, { "grad_norm": 0.2637392580509186, "learning_rate": 6.245543319496661e-06, "loss": 0.0146, "step": 67780 }, { "grad_norm": 0.23049135506153107, "learning_rate": 6.2355443554699685e-06, "loss": 0.0166, "step": 67790 }, { "grad_norm": 0.25831329822540283, "learning_rate": 6.225552869591766e-06, "loss": 0.0124, "step": 67800 }, { "grad_norm": 0.16874738037586212, "learning_rate": 6.215568863569365e-06, "loss": 0.0197, "step": 67810 }, { "grad_norm": 0.35364529490470886, "learning_rate": 6.205592339108712e-06, "loss": 0.0177, "step": 67820 }, { "grad_norm": 0.19865815341472626, "learning_rate": 6.195623297914577e-06, "loss": 0.0168, "step": 67830 }, { "grad_norm": 0.3460718095302582, "learning_rate": 6.185661741690357e-06, "loss": 0.0136, "step": 67840 }, { "grad_norm": 0.32159268856048584, "learning_rate": 6.1757076721382145e-06, "loss": 0.0202, "step": 67850 }, { "grad_norm": 0.15726003050804138, "learning_rate": 6.165761090959038e-06, "loss": 0.0209, "step": 67860 }, { "grad_norm": 0.27726709842681885, "learning_rate": 6.155821999852424e-06, "loss": 0.0158, "step": 67870 }, { "grad_norm": 0.22908522188663483, "learning_rate": 6.145890400516696e-06, "loss": 0.0168, "step": 67880 }, { "grad_norm": 0.30983737111091614, "learning_rate": 6.1359662946488816e-06, "loss": 0.0155, "step": 67890 }, { "grad_norm": 0.422805517911911, "learning_rate": 6.12604968394474e-06, "loss": 0.0155, "step": 67900 }, { "grad_norm": 0.3041369616985321, "learning_rate": 6.11614057009875e-06, "loss": 0.0199, "step": 67910 }, { "grad_norm": 0.2266404628753662, "learning_rate": 6.106238954804111e-06, "loss": 0.0146, "step": 67920 }, { "grad_norm": 0.2610297203063965, "learning_rate": 6.096344839752738e-06, "loss": 0.0171, "step": 67930 }, { "grad_norm": 0.11611741036176682, "learning_rate": 6.086458226635278e-06, "loss": 0.0149, "step": 67940 }, { "grad_norm": 0.1970275491476059, "learning_rate": 6.076579117141046e-06, "loss": 0.0259, "step": 67950 }, { "grad_norm": 0.3253474831581116, "learning_rate": 6.066707512958153e-06, "loss": 0.0163, "step": 67960 }, { "grad_norm": 0.2908165156841278, "learning_rate": 6.056843415773361e-06, "loss": 0.0169, "step": 67970 }, { "grad_norm": 0.202759250998497, "learning_rate": 6.0469868272721776e-06, "loss": 0.0156, "step": 67980 }, { "grad_norm": 0.25655481219291687, "learning_rate": 6.037137749138844e-06, "loss": 0.0189, "step": 67990 }, { "grad_norm": 0.3570430874824524, "learning_rate": 6.027296183056252e-06, "loss": 0.0234, "step": 68000 }, { "grad_norm": 0.3230566680431366, "learning_rate": 6.017462130706114e-06, "loss": 0.0167, "step": 68010 }, { "grad_norm": 0.1903209537267685, "learning_rate": 6.0076355937687516e-06, "loss": 0.02, "step": 68020 }, { "grad_norm": 0.1497894823551178, "learning_rate": 5.9978165739232925e-06, "loss": 0.0138, "step": 68030 }, { "grad_norm": 0.26170599460601807, "learning_rate": 5.988005072847508e-06, "loss": 0.0124, "step": 68040 }, { "grad_norm": 0.27391114830970764, "learning_rate": 5.978201092217928e-06, "loss": 0.0228, "step": 68050 }, { "grad_norm": 0.26475438475608826, "learning_rate": 5.9684046337097895e-06, "loss": 0.0141, "step": 68060 }, { "grad_norm": 0.16553595662117004, "learning_rate": 5.958615698997016e-06, "loss": 0.014, "step": 68070 }, { "grad_norm": 0.3310909867286682, "learning_rate": 5.948834289752303e-06, "loss": 0.0136, "step": 68080 }, { "grad_norm": 0.30935004353523254, "learning_rate": 5.939060407646996e-06, "loss": 0.0209, "step": 68090 }, { "grad_norm": 0.24390505254268646, "learning_rate": 5.929294054351198e-06, "loss": 0.0189, "step": 68100 }, { "grad_norm": 0.22091448307037354, "learning_rate": 5.919535231533707e-06, "loss": 0.0153, "step": 68110 }, { "grad_norm": 0.2443723827600479, "learning_rate": 5.909783940862046e-06, "loss": 0.0156, "step": 68120 }, { "grad_norm": 0.43821966648101807, "learning_rate": 5.900040184002436e-06, "loss": 0.0185, "step": 68130 }, { "grad_norm": 0.3067741394042969, "learning_rate": 5.890303962619831e-06, "loss": 0.0124, "step": 68140 }, { "grad_norm": 0.3072137236595154, "learning_rate": 5.88057527837787e-06, "loss": 0.0127, "step": 68150 }, { "grad_norm": 0.20542754232883453, "learning_rate": 5.870854132938924e-06, "loss": 0.0148, "step": 68160 }, { "grad_norm": 0.2839219868183136, "learning_rate": 5.861140527964071e-06, "loss": 0.0164, "step": 68170 }, { "grad_norm": 0.15634086728096008, "learning_rate": 5.851434465113098e-06, "loss": 0.0114, "step": 68180 }, { "grad_norm": 0.2594242990016937, "learning_rate": 5.841735946044524e-06, "loss": 0.0145, "step": 68190 }, { "grad_norm": 0.17485521733760834, "learning_rate": 5.832044972415523e-06, "loss": 0.018, "step": 68200 }, { "grad_norm": 0.2890956401824951, "learning_rate": 5.8223615458820655e-06, "loss": 0.0149, "step": 68210 }, { "grad_norm": 0.14784614741802216, "learning_rate": 5.812685668098733e-06, "loss": 0.0164, "step": 68220 }, { "grad_norm": 0.3182300627231598, "learning_rate": 5.803017340718913e-06, "loss": 0.0212, "step": 68230 }, { "grad_norm": 0.2468278557062149, "learning_rate": 5.793356565394636e-06, "loss": 0.016, "step": 68240 }, { "grad_norm": 0.2636201083660126, "learning_rate": 5.783703343776658e-06, "loss": 0.015, "step": 68250 }, { "grad_norm": 0.2354794293642044, "learning_rate": 5.774057677514477e-06, "loss": 0.0122, "step": 68260 }, { "grad_norm": 0.24448665976524353, "learning_rate": 5.764419568256235e-06, "loss": 0.0144, "step": 68270 }, { "grad_norm": 0.35377037525177, "learning_rate": 5.754789017648865e-06, "loss": 0.0245, "step": 68280 }, { "grad_norm": 0.11524893343448639, "learning_rate": 5.745166027337934e-06, "loss": 0.0182, "step": 68290 }, { "grad_norm": 0.18620312213897705, "learning_rate": 5.735550598967754e-06, "loss": 0.0164, "step": 68300 }, { "grad_norm": 0.490882009267807, "learning_rate": 5.725942734181339e-06, "loss": 0.0211, "step": 68310 }, { "grad_norm": 0.3855549395084381, "learning_rate": 5.7163424346204146e-06, "loss": 0.02, "step": 68320 }, { "grad_norm": 0.27529841661453247, "learning_rate": 5.706749701925407e-06, "loss": 0.0149, "step": 68330 }, { "grad_norm": 0.19573086500167847, "learning_rate": 5.697164537735461e-06, "loss": 0.0122, "step": 68340 }, { "grad_norm": 0.18793147802352905, "learning_rate": 5.687586943688406e-06, "loss": 0.0137, "step": 68350 }, { "grad_norm": 0.20805208384990692, "learning_rate": 5.678016921420792e-06, "loss": 0.0156, "step": 68360 }, { "grad_norm": 0.2852635979652405, "learning_rate": 5.668454472567875e-06, "loss": 0.0213, "step": 68370 }, { "grad_norm": 0.3124929666519165, "learning_rate": 5.658899598763617e-06, "loss": 0.0164, "step": 68380 }, { "grad_norm": 0.2770560681819916, "learning_rate": 5.6493523016407035e-06, "loss": 0.0135, "step": 68390 }, { "grad_norm": 0.15043015778064728, "learning_rate": 5.6398125828304615e-06, "loss": 0.0161, "step": 68400 }, { "grad_norm": 0.1224273219704628, "learning_rate": 5.630280443963015e-06, "loss": 0.0185, "step": 68410 }, { "grad_norm": 0.21652215719223022, "learning_rate": 5.620755886667112e-06, "loss": 0.0168, "step": 68420 }, { "grad_norm": 0.2783757746219635, "learning_rate": 5.611238912570271e-06, "loss": 0.0148, "step": 68430 }, { "grad_norm": 0.30153167247772217, "learning_rate": 5.601729523298649e-06, "loss": 0.016, "step": 68440 }, { "grad_norm": 0.1966380476951599, "learning_rate": 5.592227720477161e-06, "loss": 0.0145, "step": 68450 }, { "grad_norm": 0.25412511825561523, "learning_rate": 5.582733505729415e-06, "loss": 0.0167, "step": 68460 }, { "grad_norm": 0.342060387134552, "learning_rate": 5.573246880677668e-06, "loss": 0.0212, "step": 68470 }, { "grad_norm": 0.30012205243110657, "learning_rate": 5.56376784694298e-06, "loss": 0.0177, "step": 68480 }, { "grad_norm": 0.28744223713874817, "learning_rate": 5.554296406145027e-06, "loss": 0.0213, "step": 68490 }, { "grad_norm": 0.3142659366130829, "learning_rate": 5.544832559902219e-06, "loss": 0.0155, "step": 68500 }, { "grad_norm": 0.2983895242214203, "learning_rate": 5.53537630983168e-06, "loss": 0.0144, "step": 68510 }, { "grad_norm": 0.24124367535114288, "learning_rate": 5.5259276575492125e-06, "loss": 0.0193, "step": 68520 }, { "grad_norm": 0.3240358829498291, "learning_rate": 5.516486604669357e-06, "loss": 0.023, "step": 68530 }, { "grad_norm": 0.38533759117126465, "learning_rate": 5.507053152805303e-06, "loss": 0.0222, "step": 68540 }, { "grad_norm": 0.2976173758506775, "learning_rate": 5.497627303568975e-06, "loss": 0.0179, "step": 68550 }, { "grad_norm": 0.099936842918396, "learning_rate": 5.488209058571003e-06, "loss": 0.0108, "step": 68560 }, { "grad_norm": 0.21180562674999237, "learning_rate": 5.4787984194207054e-06, "loss": 0.0174, "step": 68570 }, { "grad_norm": 0.38963454961776733, "learning_rate": 5.469395387726095e-06, "loss": 0.0131, "step": 68580 }, { "grad_norm": 0.2905188798904419, "learning_rate": 5.45999996509391e-06, "loss": 0.0142, "step": 68590 }, { "grad_norm": 0.26866012811660767, "learning_rate": 5.450612153129536e-06, "loss": 0.0188, "step": 68600 }, { "grad_norm": 0.3141629695892334, "learning_rate": 5.4412319534371426e-06, "loss": 0.0159, "step": 68610 }, { "grad_norm": 0.3119954466819763, "learning_rate": 5.431859367619513e-06, "loss": 0.0133, "step": 68620 }, { "grad_norm": 0.2584798336029053, "learning_rate": 5.422494397278172e-06, "loss": 0.0164, "step": 68630 }, { "grad_norm": 0.5727072954177856, "learning_rate": 5.413137044013344e-06, "loss": 0.0166, "step": 68640 }, { "grad_norm": 0.16306106746196747, "learning_rate": 5.403787309423941e-06, "loss": 0.0168, "step": 68650 }, { "grad_norm": 0.19875971972942352, "learning_rate": 5.3944451951075835e-06, "loss": 0.0139, "step": 68660 }, { "grad_norm": 0.31471288204193115, "learning_rate": 5.385110702660562e-06, "loss": 0.0143, "step": 68670 }, { "grad_norm": 0.20689134299755096, "learning_rate": 5.375783833677922e-06, "loss": 0.0193, "step": 68680 }, { "grad_norm": 0.22059357166290283, "learning_rate": 5.36646458975334e-06, "loss": 0.0146, "step": 68690 }, { "grad_norm": 0.22708719968795776, "learning_rate": 5.3571529724792294e-06, "loss": 0.0094, "step": 68700 }, { "grad_norm": 0.41756895184516907, "learning_rate": 5.347848983446702e-06, "loss": 0.0166, "step": 68710 }, { "grad_norm": 0.41342592239379883, "learning_rate": 5.3385526242455185e-06, "loss": 0.0181, "step": 68720 }, { "grad_norm": 0.8198671936988831, "learning_rate": 5.329263896464226e-06, "loss": 0.014, "step": 68730 }, { "grad_norm": 0.17335620522499084, "learning_rate": 5.3199828016899715e-06, "loss": 0.0155, "step": 68740 }, { "grad_norm": 0.27114495635032654, "learning_rate": 5.310709341508657e-06, "loss": 0.0154, "step": 68750 }, { "grad_norm": 0.21571767330169678, "learning_rate": 5.301443517504861e-06, "loss": 0.0118, "step": 68760 }, { "grad_norm": 0.148734912276268, "learning_rate": 5.292185331261862e-06, "loss": 0.012, "step": 68770 }, { "grad_norm": 0.288320928812027, "learning_rate": 5.28293478436162e-06, "loss": 0.0159, "step": 68780 }, { "grad_norm": 0.699846625328064, "learning_rate": 5.273691878384829e-06, "loss": 0.0164, "step": 68790 }, { "grad_norm": 0.2486359030008316, "learning_rate": 5.264456614910812e-06, "loss": 0.0155, "step": 68800 }, { "grad_norm": 0.234064981341362, "learning_rate": 5.255228995517647e-06, "loss": 0.0228, "step": 68810 }, { "grad_norm": 0.19003620743751526, "learning_rate": 5.246009021782067e-06, "loss": 0.0137, "step": 68820 }, { "grad_norm": 0.16519637405872345, "learning_rate": 5.2367966952795225e-06, "loss": 0.0234, "step": 68830 }, { "grad_norm": 0.1882932037115097, "learning_rate": 5.2275920175841485e-06, "loss": 0.0143, "step": 68840 }, { "grad_norm": 0.31380918622016907, "learning_rate": 5.218394990268771e-06, "loss": 0.0187, "step": 68850 }, { "grad_norm": 0.4411848187446594, "learning_rate": 5.209205614904916e-06, "loss": 0.0173, "step": 68860 }, { "grad_norm": 0.3517546057701111, "learning_rate": 5.200023893062777e-06, "loss": 0.0147, "step": 68870 }, { "grad_norm": 0.2029130607843399, "learning_rate": 5.190849826311289e-06, "loss": 0.0119, "step": 68880 }, { "grad_norm": 0.3400532603263855, "learning_rate": 5.181683416218025e-06, "loss": 0.019, "step": 68890 }, { "grad_norm": 0.18555086851119995, "learning_rate": 5.172524664349276e-06, "loss": 0.0132, "step": 68900 }, { "grad_norm": 0.6282731890678406, "learning_rate": 5.1633735722700416e-06, "loss": 0.0124, "step": 68910 }, { "grad_norm": 0.1658242791891098, "learning_rate": 5.154230141543958e-06, "loss": 0.0162, "step": 68920 }, { "grad_norm": 0.36317020654678345, "learning_rate": 5.145094373733433e-06, "loss": 0.0173, "step": 68930 }, { "grad_norm": 0.33079177141189575, "learning_rate": 5.135966270399478e-06, "loss": 0.0117, "step": 68940 }, { "grad_norm": 0.4720246195793152, "learning_rate": 5.126845833101857e-06, "loss": 0.0182, "step": 68950 }, { "grad_norm": 0.47042644023895264, "learning_rate": 5.1177330633989994e-06, "loss": 0.0178, "step": 68960 }, { "grad_norm": 0.41379907727241516, "learning_rate": 5.108627962848033e-06, "loss": 0.0316, "step": 68970 }, { "grad_norm": 0.19264128804206848, "learning_rate": 5.099530533004759e-06, "loss": 0.0168, "step": 68980 }, { "grad_norm": 0.12861740589141846, "learning_rate": 5.090440775423699e-06, "loss": 0.0147, "step": 68990 }, { "grad_norm": 0.2332793027162552, "learning_rate": 5.081358691658022e-06, "loss": 0.0131, "step": 69000 }, { "grad_norm": 0.25233447551727295, "learning_rate": 5.072284283259621e-06, "loss": 0.0173, "step": 69010 }, { "grad_norm": 0.17368189990520477, "learning_rate": 5.063217551779053e-06, "loss": 0.0129, "step": 69020 }, { "grad_norm": 0.23296959698200226, "learning_rate": 5.054158498765588e-06, "loss": 0.0127, "step": 69030 }, { "grad_norm": 0.22535179555416107, "learning_rate": 5.0451071257671625e-06, "loss": 0.014, "step": 69040 }, { "grad_norm": 0.35317927598953247, "learning_rate": 5.036063434330407e-06, "loss": 0.0212, "step": 69050 }, { "grad_norm": 0.16281485557556152, "learning_rate": 5.027027426000652e-06, "loss": 0.0152, "step": 69060 }, { "grad_norm": 0.17111483216285706, "learning_rate": 5.017999102321886e-06, "loss": 0.0143, "step": 69070 }, { "grad_norm": 0.3815719485282898, "learning_rate": 5.0089784648368224e-06, "loss": 0.0171, "step": 69080 }, { "grad_norm": 0.2537890672683716, "learning_rate": 4.999965515086829e-06, "loss": 0.0147, "step": 69090 }, { "grad_norm": 0.12530182301998138, "learning_rate": 4.990960254611976e-06, "loss": 0.0105, "step": 69100 }, { "grad_norm": 0.38944095373153687, "learning_rate": 4.9819626849510194e-06, "loss": 0.0157, "step": 69110 }, { "grad_norm": 0.21173886954784393, "learning_rate": 4.97297280764138e-06, "loss": 0.0205, "step": 69120 }, { "grad_norm": 0.18675340712070465, "learning_rate": 4.963990624219211e-06, "loss": 0.0195, "step": 69130 }, { "grad_norm": 0.302240788936615, "learning_rate": 4.955016136219298e-06, "loss": 0.0126, "step": 69140 }, { "grad_norm": 0.22266356647014618, "learning_rate": 4.946049345175146e-06, "loss": 0.0122, "step": 69150 }, { "grad_norm": 0.3755554258823395, "learning_rate": 4.937090252618937e-06, "loss": 0.0132, "step": 69160 }, { "grad_norm": 0.3686714172363281, "learning_rate": 4.928138860081521e-06, "loss": 0.0224, "step": 69170 }, { "grad_norm": 0.2838473320007324, "learning_rate": 4.919195169092472e-06, "loss": 0.017, "step": 69180 }, { "grad_norm": 0.12120558321475983, "learning_rate": 4.910259181179994e-06, "loss": 0.0116, "step": 69190 }, { "grad_norm": 0.1682298481464386, "learning_rate": 4.901330897871015e-06, "loss": 0.0165, "step": 69200 }, { "grad_norm": 0.1645282357931137, "learning_rate": 4.8924103206911375e-06, "loss": 0.0125, "step": 69210 }, { "grad_norm": 0.1970892995595932, "learning_rate": 4.883497451164637e-06, "loss": 0.0165, "step": 69220 }, { "grad_norm": 0.2995406687259674, "learning_rate": 4.874592290814484e-06, "loss": 0.0196, "step": 69230 }, { "grad_norm": 0.4244095981121063, "learning_rate": 4.865694841162327e-06, "loss": 0.0247, "step": 69240 }, { "grad_norm": 0.1661580353975296, "learning_rate": 4.856805103728496e-06, "loss": 0.011, "step": 69250 }, { "grad_norm": 0.27922263741493225, "learning_rate": 4.847923080032007e-06, "loss": 0.0106, "step": 69260 }, { "grad_norm": 0.29639774560928345, "learning_rate": 4.8390487715905486e-06, "loss": 0.0171, "step": 69270 }, { "grad_norm": 0.2814950942993164, "learning_rate": 4.8301821799205e-06, "loss": 0.0146, "step": 69280 }, { "grad_norm": 0.14127425849437714, "learning_rate": 4.821323306536918e-06, "loss": 0.0113, "step": 69290 }, { "grad_norm": 0.18485385179519653, "learning_rate": 4.8124721529535455e-06, "loss": 0.0132, "step": 69300 }, { "grad_norm": 0.23902836441993713, "learning_rate": 4.803628720682807e-06, "loss": 0.0187, "step": 69310 }, { "grad_norm": 0.3144749402999878, "learning_rate": 4.794793011235776e-06, "loss": 0.0148, "step": 69320 }, { "grad_norm": 0.1602446436882019, "learning_rate": 4.78596502612228e-06, "loss": 0.0177, "step": 69330 }, { "grad_norm": 0.27195271849632263, "learning_rate": 4.777144766850738e-06, "loss": 0.0194, "step": 69340 }, { "grad_norm": 0.23563054203987122, "learning_rate": 4.768332234928313e-06, "loss": 0.0193, "step": 69350 }, { "grad_norm": 0.18459013104438782, "learning_rate": 4.759527431860828e-06, "loss": 0.0121, "step": 69360 }, { "grad_norm": 0.2341909408569336, "learning_rate": 4.750730359152755e-06, "loss": 0.0227, "step": 69370 }, { "grad_norm": 0.26839104294776917, "learning_rate": 4.741941018307311e-06, "loss": 0.0184, "step": 69380 }, { "grad_norm": 0.19541656970977783, "learning_rate": 4.733159410826321e-06, "loss": 0.0205, "step": 69390 }, { "grad_norm": 0.26774683594703674, "learning_rate": 4.724385538210357e-06, "loss": 0.0127, "step": 69400 }, { "grad_norm": 0.2871279716491699, "learning_rate": 4.715619401958599e-06, "loss": 0.0172, "step": 69410 }, { "grad_norm": 0.2336159199476242, "learning_rate": 4.706861003568958e-06, "loss": 0.0182, "step": 69420 }, { "grad_norm": 0.18385489284992218, "learning_rate": 4.698110344538003e-06, "loss": 0.0142, "step": 69430 }, { "grad_norm": 0.43366971611976624, "learning_rate": 4.689367426360975e-06, "loss": 0.0147, "step": 69440 }, { "grad_norm": 0.3145371675491333, "learning_rate": 4.680632250531819e-06, "loss": 0.0117, "step": 69450 }, { "grad_norm": 0.14983446896076202, "learning_rate": 4.671904818543115e-06, "loss": 0.0179, "step": 69460 }, { "grad_norm": 0.24840041995048523, "learning_rate": 4.66318513188615e-06, "loss": 0.0118, "step": 69470 }, { "grad_norm": 0.3672240078449249, "learning_rate": 4.654473192050884e-06, "loss": 0.0171, "step": 69480 }, { "grad_norm": 0.1682683825492859, "learning_rate": 4.64576900052594e-06, "loss": 0.0162, "step": 69490 }, { "grad_norm": 0.2642570436000824, "learning_rate": 4.637072558798638e-06, "loss": 0.0113, "step": 69500 }, { "grad_norm": 0.2713377773761749, "learning_rate": 4.628383868354969e-06, "loss": 0.0161, "step": 69510 }, { "grad_norm": 0.18960799276828766, "learning_rate": 4.6197029306795595e-06, "loss": 0.0131, "step": 69520 }, { "grad_norm": 0.1764051467180252, "learning_rate": 4.611029747255779e-06, "loss": 0.0129, "step": 69530 }, { "grad_norm": 0.18934442102909088, "learning_rate": 4.6023643195656164e-06, "loss": 0.0118, "step": 69540 }, { "grad_norm": 0.17478899657726288, "learning_rate": 4.593706649089768e-06, "loss": 0.0142, "step": 69550 }, { "grad_norm": 0.3358723521232605, "learning_rate": 4.585056737307597e-06, "loss": 0.0185, "step": 69560 }, { "grad_norm": 0.24123825132846832, "learning_rate": 4.576414585697103e-06, "loss": 0.0154, "step": 69570 }, { "grad_norm": 0.3442382216453552, "learning_rate": 4.567780195735044e-06, "loss": 0.0168, "step": 69580 }, { "grad_norm": 0.18260441720485687, "learning_rate": 4.559153568896757e-06, "loss": 0.0122, "step": 69590 }, { "grad_norm": 0.7963977456092834, "learning_rate": 4.550534706656329e-06, "loss": 0.0217, "step": 69600 }, { "grad_norm": 0.1555248498916626, "learning_rate": 4.541923610486465e-06, "loss": 0.0151, "step": 69610 }, { "grad_norm": 0.2577589154243469, "learning_rate": 4.533320281858578e-06, "loss": 0.0128, "step": 69620 }, { "grad_norm": 0.4251551926136017, "learning_rate": 4.52472472224274e-06, "loss": 0.0207, "step": 69630 }, { "grad_norm": 0.3658243417739868, "learning_rate": 4.51613693310769e-06, "loss": 0.0173, "step": 69640 }, { "grad_norm": 1.1682599782943726, "learning_rate": 4.507556915920868e-06, "loss": 0.0171, "step": 69650 }, { "grad_norm": 0.1803291290998459, "learning_rate": 4.498984672148332e-06, "loss": 0.0136, "step": 69660 }, { "grad_norm": 0.3700800836086273, "learning_rate": 4.490420203254864e-06, "loss": 0.0158, "step": 69670 }, { "grad_norm": 0.38490039110183716, "learning_rate": 4.481863510703893e-06, "loss": 0.0185, "step": 69680 }, { "grad_norm": 0.1652081310749054, "learning_rate": 4.473314595957523e-06, "loss": 0.0117, "step": 69690 }, { "grad_norm": 0.259714275598526, "learning_rate": 4.464773460476535e-06, "loss": 0.0154, "step": 69700 }, { "grad_norm": 0.13926538825035095, "learning_rate": 4.456240105720372e-06, "loss": 0.0141, "step": 69710 }, { "grad_norm": 0.2008162885904312, "learning_rate": 4.4477145331471405e-06, "loss": 0.0256, "step": 69720 }, { "grad_norm": 0.29486456513404846, "learning_rate": 4.439196744213653e-06, "loss": 0.014, "step": 69730 }, { "grad_norm": 0.5807153582572937, "learning_rate": 4.430686740375339e-06, "loss": 0.0153, "step": 69740 }, { "grad_norm": 0.19718192517757416, "learning_rate": 4.422184523086342e-06, "loss": 0.0144, "step": 69750 }, { "grad_norm": 0.27575382590293884, "learning_rate": 4.41369009379946e-06, "loss": 0.0148, "step": 69760 }, { "grad_norm": 0.44800955057144165, "learning_rate": 4.405203453966139e-06, "loss": 0.0179, "step": 69770 }, { "grad_norm": 0.3064155876636505, "learning_rate": 4.396724605036539e-06, "loss": 0.0132, "step": 69780 }, { "grad_norm": 0.21019329130649567, "learning_rate": 4.388253548459437e-06, "loss": 0.0153, "step": 69790 }, { "grad_norm": 0.29201045632362366, "learning_rate": 4.3797902856823395e-06, "loss": 0.0182, "step": 69800 }, { "grad_norm": 0.24324578046798706, "learning_rate": 4.371334818151357e-06, "loss": 0.0123, "step": 69810 }, { "grad_norm": 0.2527848184108734, "learning_rate": 4.362887147311306e-06, "loss": 0.0136, "step": 69820 }, { "grad_norm": 0.18307878077030182, "learning_rate": 4.354447274605672e-06, "loss": 0.0121, "step": 69830 }, { "grad_norm": 0.23917169868946075, "learning_rate": 4.346015201476572e-06, "loss": 0.021, "step": 69840 }, { "grad_norm": 0.15144318342208862, "learning_rate": 4.337590929364855e-06, "loss": 0.0164, "step": 69850 }, { "grad_norm": 0.2534600496292114, "learning_rate": 4.329174459709973e-06, "loss": 0.017, "step": 69860 }, { "grad_norm": 0.29711058735847473, "learning_rate": 4.320765793950071e-06, "loss": 0.0204, "step": 69870 }, { "grad_norm": 0.4043082594871521, "learning_rate": 4.312364933521962e-06, "loss": 0.0165, "step": 69880 }, { "grad_norm": 0.2220499962568283, "learning_rate": 4.303971879861129e-06, "loss": 0.0126, "step": 69890 }, { "grad_norm": 0.35313981771469116, "learning_rate": 4.295586634401716e-06, "loss": 0.0194, "step": 69900 }, { "grad_norm": 0.31686529517173767, "learning_rate": 4.287209198576536e-06, "loss": 0.013, "step": 69910 }, { "grad_norm": 0.2519712746143341, "learning_rate": 4.278839573817045e-06, "loss": 0.0166, "step": 69920 }, { "grad_norm": 0.6077377796173096, "learning_rate": 4.270477761553399e-06, "loss": 0.021, "step": 69930 }, { "grad_norm": 0.19662772119045258, "learning_rate": 4.262123763214393e-06, "loss": 0.0147, "step": 69940 }, { "grad_norm": 0.21348465979099274, "learning_rate": 4.253777580227508e-06, "loss": 0.0125, "step": 69950 }, { "grad_norm": 0.29234838485717773, "learning_rate": 4.2454392140188755e-06, "loss": 0.0231, "step": 69960 }, { "grad_norm": 0.5695111751556396, "learning_rate": 4.2371086660132785e-06, "loss": 0.0215, "step": 69970 }, { "grad_norm": 0.344123899936676, "learning_rate": 4.228785937634205e-06, "loss": 0.011, "step": 69980 }, { "grad_norm": 0.2774571478366852, "learning_rate": 4.220471030303758e-06, "loss": 0.0124, "step": 69990 }, { "grad_norm": 0.1565295159816742, "learning_rate": 4.212163945442754e-06, "loss": 0.0111, "step": 70000 }, { "grad_norm": 0.5071172118186951, "learning_rate": 4.203864684470621e-06, "loss": 0.0154, "step": 70010 }, { "grad_norm": 0.16479745507240295, "learning_rate": 4.19557324880549e-06, "loss": 0.0162, "step": 70020 }, { "grad_norm": 0.26127898693084717, "learning_rate": 4.187289639864145e-06, "loss": 0.0157, "step": 70030 }, { "grad_norm": 0.18977084755897522, "learning_rate": 4.1790138590619974e-06, "loss": 0.0204, "step": 70040 }, { "grad_norm": 0.2183503657579422, "learning_rate": 4.170745907813195e-06, "loss": 0.0136, "step": 70050 }, { "grad_norm": 0.18377158045768738, "learning_rate": 4.162485787530479e-06, "loss": 0.014, "step": 70060 }, { "grad_norm": 0.4644523561000824, "learning_rate": 4.154233499625282e-06, "loss": 0.0146, "step": 70070 }, { "grad_norm": 0.2276068925857544, "learning_rate": 4.145989045507692e-06, "loss": 0.0127, "step": 70080 }, { "grad_norm": 0.20784790813922882, "learning_rate": 4.1377524265864666e-06, "loss": 0.0156, "step": 70090 }, { "grad_norm": 0.21591700613498688, "learning_rate": 4.1295236442690175e-06, "loss": 0.0151, "step": 70100 }, { "grad_norm": 0.15799100697040558, "learning_rate": 4.121302699961421e-06, "loss": 0.0135, "step": 70110 }, { "grad_norm": 0.2947741150856018, "learning_rate": 4.113089595068403e-06, "loss": 0.0167, "step": 70120 }, { "grad_norm": 0.31473487615585327, "learning_rate": 4.104884330993364e-06, "loss": 0.0189, "step": 70130 }, { "grad_norm": 0.43922126293182373, "learning_rate": 4.0966869091383585e-06, "loss": 0.0219, "step": 70140 }, { "grad_norm": 0.21868419647216797, "learning_rate": 4.088497330904101e-06, "loss": 0.0173, "step": 70150 }, { "grad_norm": 0.243094339966774, "learning_rate": 4.080315597689976e-06, "loss": 0.0143, "step": 70160 }, { "grad_norm": 0.16373783349990845, "learning_rate": 4.072141710893995e-06, "loss": 0.0248, "step": 70170 }, { "grad_norm": 0.12703834474086761, "learning_rate": 4.063975671912879e-06, "loss": 0.0229, "step": 70180 }, { "grad_norm": 0.29875653982162476, "learning_rate": 4.055817482141949e-06, "loss": 0.0142, "step": 70190 }, { "grad_norm": 0.23157982528209686, "learning_rate": 4.047667142975259e-06, "loss": 0.012, "step": 70200 }, { "grad_norm": 0.3751567304134369, "learning_rate": 4.039524655805443e-06, "loss": 0.0188, "step": 70210 }, { "grad_norm": 0.19280663132667542, "learning_rate": 4.03139002202384e-06, "loss": 0.0106, "step": 70220 }, { "grad_norm": 0.22215458750724792, "learning_rate": 4.023263243020447e-06, "loss": 0.0161, "step": 70230 }, { "grad_norm": 0.15658098459243774, "learning_rate": 4.015144320183884e-06, "loss": 0.0165, "step": 70240 }, { "grad_norm": 0.27039191126823425, "learning_rate": 4.007033254901482e-06, "loss": 0.0122, "step": 70250 }, { "grad_norm": 0.31786438822746277, "learning_rate": 3.9989300485591795e-06, "loss": 0.0192, "step": 70260 }, { "grad_norm": 0.2340158373117447, "learning_rate": 3.990834702541601e-06, "loss": 0.0171, "step": 70270 }, { "grad_norm": 0.20111499726772308, "learning_rate": 3.982747218232019e-06, "loss": 0.012, "step": 70280 }, { "grad_norm": 0.3850453794002533, "learning_rate": 3.97466759701236e-06, "loss": 0.0191, "step": 70290 }, { "grad_norm": 0.36008015275001526, "learning_rate": 3.966595840263226e-06, "loss": 0.0136, "step": 70300 }, { "grad_norm": 0.3487856090068817, "learning_rate": 3.958531949363836e-06, "loss": 0.016, "step": 70310 }, { "grad_norm": 0.19800545275211334, "learning_rate": 3.950475925692098e-06, "loss": 0.0178, "step": 70320 }, { "grad_norm": 0.2517266571521759, "learning_rate": 3.9424277706245685e-06, "loss": 0.0129, "step": 70330 }, { "grad_norm": 0.11803917586803436, "learning_rate": 3.934387485536451e-06, "loss": 0.0134, "step": 70340 }, { "grad_norm": 0.20696741342544556, "learning_rate": 3.926355071801619e-06, "loss": 0.0167, "step": 70350 }, { "grad_norm": 0.19362559914588928, "learning_rate": 3.9183305307925965e-06, "loss": 0.0173, "step": 70360 }, { "grad_norm": 0.27859318256378174, "learning_rate": 3.910313863880533e-06, "loss": 0.0115, "step": 70370 }, { "grad_norm": 0.18076373636722565, "learning_rate": 3.902305072435292e-06, "loss": 0.0129, "step": 70380 }, { "grad_norm": 0.28843051195144653, "learning_rate": 3.894304157825329e-06, "loss": 0.0152, "step": 70390 }, { "grad_norm": 0.28568020462989807, "learning_rate": 3.886311121417791e-06, "loss": 0.0146, "step": 70400 }, { "grad_norm": 0.1616780161857605, "learning_rate": 3.878325964578472e-06, "loss": 0.0135, "step": 70410 }, { "grad_norm": 0.159089133143425, "learning_rate": 3.870348688671815e-06, "loss": 0.0097, "step": 70420 }, { "grad_norm": 0.3096481263637543, "learning_rate": 3.862379295060931e-06, "loss": 0.0164, "step": 70430 }, { "grad_norm": 0.14507907629013062, "learning_rate": 3.854417785107534e-06, "loss": 0.0186, "step": 70440 }, { "grad_norm": 0.28860485553741455, "learning_rate": 3.8464641601720755e-06, "loss": 0.0149, "step": 70450 }, { "grad_norm": 0.3202371299266815, "learning_rate": 3.838518421613579e-06, "loss": 0.0222, "step": 70460 }, { "grad_norm": 0.2734992802143097, "learning_rate": 3.830580570789766e-06, "loss": 0.0172, "step": 70470 }, { "grad_norm": 0.1892903745174408, "learning_rate": 3.822650609057005e-06, "loss": 0.0146, "step": 70480 }, { "grad_norm": 0.16975563764572144, "learning_rate": 3.814728537770285e-06, "loss": 0.0146, "step": 70490 }, { "grad_norm": 0.21498802304267883, "learning_rate": 3.806814358283306e-06, "loss": 0.0148, "step": 70500 }, { "grad_norm": 0.21203118562698364, "learning_rate": 3.7989080719483596e-06, "loss": 0.0132, "step": 70510 }, { "grad_norm": 0.3039420247077942, "learning_rate": 3.7910096801164143e-06, "loss": 0.0152, "step": 70520 }, { "grad_norm": 0.20796120166778564, "learning_rate": 3.7831191841371016e-06, "loss": 0.0148, "step": 70530 }, { "grad_norm": 0.2701936364173889, "learning_rate": 3.775236585358688e-06, "loss": 0.0136, "step": 70540 }, { "grad_norm": 0.252679705619812, "learning_rate": 3.7673618851280843e-06, "loss": 0.0166, "step": 70550 }, { "grad_norm": 0.3489258587360382, "learning_rate": 3.759495084790887e-06, "loss": 0.0174, "step": 70560 }, { "grad_norm": 0.302142858505249, "learning_rate": 3.751636185691282e-06, "loss": 0.0139, "step": 70570 }, { "grad_norm": 0.14862693846225739, "learning_rate": 3.7437851891721607e-06, "loss": 0.0122, "step": 70580 }, { "grad_norm": 0.2019873410463333, "learning_rate": 3.7359420965750404e-06, "loss": 0.0164, "step": 70590 }, { "grad_norm": 0.20243826508522034, "learning_rate": 3.7281069092400922e-06, "loss": 0.0151, "step": 70600 }, { "grad_norm": 0.29359695315361023, "learning_rate": 3.7202796285061348e-06, "loss": 0.0134, "step": 70610 }, { "grad_norm": 0.17095451056957245, "learning_rate": 3.712460255710637e-06, "loss": 0.0181, "step": 70620 }, { "grad_norm": 0.2627260684967041, "learning_rate": 3.704648792189719e-06, "loss": 0.0125, "step": 70630 }, { "grad_norm": 0.3349855840206146, "learning_rate": 3.696845239278124e-06, "loss": 0.0183, "step": 70640 }, { "grad_norm": 0.2876140773296356, "learning_rate": 3.689049598309302e-06, "loss": 0.0155, "step": 70650 }, { "grad_norm": 0.23442596197128296, "learning_rate": 3.681261870615288e-06, "loss": 0.0232, "step": 70660 }, { "grad_norm": 0.3208776116371155, "learning_rate": 3.6734820575268004e-06, "loss": 0.0253, "step": 70670 }, { "grad_norm": 0.18352888524532318, "learning_rate": 3.665710160373209e-06, "loss": 0.0155, "step": 70680 }, { "grad_norm": 0.228524848818779, "learning_rate": 3.65794618048248e-06, "loss": 0.0108, "step": 70690 }, { "grad_norm": 0.34632131457328796, "learning_rate": 3.6501901191813125e-06, "loss": 0.0103, "step": 70700 }, { "grad_norm": 0.3865699768066406, "learning_rate": 3.642441977794975e-06, "loss": 0.0123, "step": 70710 }, { "grad_norm": 0.19082343578338623, "learning_rate": 3.634701757647424e-06, "loss": 0.0136, "step": 70720 }, { "grad_norm": 0.1291298270225525, "learning_rate": 3.6269694600612468e-06, "loss": 0.0092, "step": 70730 }, { "grad_norm": 0.32305315136909485, "learning_rate": 3.619245086357681e-06, "loss": 0.024, "step": 70740 }, { "grad_norm": 0.4192352890968323, "learning_rate": 3.611528637856615e-06, "loss": 0.014, "step": 70750 }, { "grad_norm": 0.3200034201145172, "learning_rate": 3.6038201158765884e-06, "loss": 0.0156, "step": 70760 }, { "grad_norm": 0.4680930972099304, "learning_rate": 3.5961195217347534e-06, "loss": 0.0212, "step": 70770 }, { "grad_norm": 0.18538060784339905, "learning_rate": 3.588426856746946e-06, "loss": 0.0167, "step": 70780 }, { "grad_norm": 0.1776563674211502, "learning_rate": 3.5807421222276316e-06, "loss": 0.0216, "step": 70790 }, { "grad_norm": 0.3198292851448059, "learning_rate": 3.5730653194899156e-06, "loss": 0.0182, "step": 70800 }, { "grad_norm": 0.27856042981147766, "learning_rate": 3.565396449845554e-06, "loss": 0.015, "step": 70810 }, { "grad_norm": 0.421676367521286, "learning_rate": 3.557735514604954e-06, "loss": 0.0145, "step": 70820 }, { "grad_norm": 0.2962959408760071, "learning_rate": 3.5500825150771633e-06, "loss": 0.0137, "step": 70830 }, { "grad_norm": 0.29526692628860474, "learning_rate": 3.5424374525698466e-06, "loss": 0.0149, "step": 70840 }, { "grad_norm": 0.21922802925109863, "learning_rate": 3.5348003283893704e-06, "loss": 0.0129, "step": 70850 }, { "grad_norm": 0.12883758544921875, "learning_rate": 3.527171143840685e-06, "loss": 0.0169, "step": 70860 }, { "grad_norm": 0.2577982544898987, "learning_rate": 3.519549900227409e-06, "loss": 0.0174, "step": 70870 }, { "grad_norm": 0.34683260321617126, "learning_rate": 3.511936598851828e-06, "loss": 0.0169, "step": 70880 }, { "grad_norm": 0.1633867472410202, "learning_rate": 3.504331241014813e-06, "loss": 0.013, "step": 70890 }, { "grad_norm": 0.2777668237686157, "learning_rate": 3.4967338280159414e-06, "loss": 0.0112, "step": 70900 }, { "grad_norm": 0.1871706247329712, "learning_rate": 3.4891443611533846e-06, "loss": 0.018, "step": 70910 }, { "grad_norm": 0.23908820748329163, "learning_rate": 3.481562841723984e-06, "loss": 0.0181, "step": 70920 }, { "grad_norm": 0.14546538889408112, "learning_rate": 3.473989271023215e-06, "loss": 0.0134, "step": 70930 }, { "grad_norm": 0.2465282380580902, "learning_rate": 3.4664236503451854e-06, "loss": 0.0152, "step": 70940 }, { "grad_norm": 0.3608216643333435, "learning_rate": 3.4588659809826674e-06, "loss": 0.0127, "step": 70950 }, { "grad_norm": 0.382667601108551, "learning_rate": 3.451316264227039e-06, "loss": 0.018, "step": 70960 }, { "grad_norm": 0.28885334730148315, "learning_rate": 3.4437745013683574e-06, "loss": 0.0177, "step": 70970 }, { "grad_norm": 0.1729244738817215, "learning_rate": 3.4362406936952916e-06, "loss": 0.0149, "step": 70980 }, { "grad_norm": 0.20028764009475708, "learning_rate": 3.4287148424951786e-06, "loss": 0.015, "step": 70990 }, { "grad_norm": 0.23688466846942902, "learning_rate": 3.4211969490539617e-06, "loss": 0.0157, "step": 71000 }, { "grad_norm": 0.22497186064720154, "learning_rate": 3.4136870146562584e-06, "loss": 0.0183, "step": 71010 }, { "grad_norm": 0.20849131047725677, "learning_rate": 3.406185040585308e-06, "loss": 0.0119, "step": 71020 }, { "grad_norm": 0.32601654529571533, "learning_rate": 3.3986910281229966e-06, "loss": 0.0182, "step": 71030 }, { "grad_norm": 0.28131943941116333, "learning_rate": 3.391204978549828e-06, "loss": 0.0171, "step": 71040 }, { "grad_norm": 0.25015953183174133, "learning_rate": 3.3837268931449785e-06, "loss": 0.0221, "step": 71050 }, { "grad_norm": 0.246967613697052, "learning_rate": 3.376256773186248e-06, "loss": 0.02, "step": 71060 }, { "grad_norm": 0.355022132396698, "learning_rate": 3.3687946199500664e-06, "loss": 0.0212, "step": 71070 }, { "grad_norm": 0.3350532650947571, "learning_rate": 3.3613404347115295e-06, "loss": 0.0099, "step": 71080 }, { "grad_norm": 0.17212313413619995, "learning_rate": 3.35389421874433e-06, "loss": 0.0151, "step": 71090 }, { "grad_norm": 0.21527020633220673, "learning_rate": 3.3464559733208446e-06, "loss": 0.0156, "step": 71100 }, { "grad_norm": 0.4799298644065857, "learning_rate": 3.3390256997120505e-06, "loss": 0.0201, "step": 71110 }, { "grad_norm": 0.1584549844264984, "learning_rate": 3.331603399187583e-06, "loss": 0.0163, "step": 71120 }, { "grad_norm": 0.17341682314872742, "learning_rate": 3.324189073015721e-06, "loss": 0.0119, "step": 71130 }, { "grad_norm": 0.2118910253047943, "learning_rate": 3.3167827224633465e-06, "loss": 0.0164, "step": 71140 }, { "grad_norm": 0.36924228072166443, "learning_rate": 3.309384348796024e-06, "loss": 0.0189, "step": 71150 }, { "grad_norm": 0.21530498564243317, "learning_rate": 3.301993953277921e-06, "loss": 0.0169, "step": 71160 }, { "grad_norm": 0.1688607931137085, "learning_rate": 3.29461153717186e-06, "loss": 0.0142, "step": 71170 }, { "grad_norm": 0.26710325479507446, "learning_rate": 3.287237101739293e-06, "loss": 0.0191, "step": 71180 }, { "grad_norm": 0.3772251009941101, "learning_rate": 3.2798706482403075e-06, "loss": 0.0218, "step": 71190 }, { "grad_norm": 0.43741270899772644, "learning_rate": 3.2725121779336285e-06, "loss": 0.0183, "step": 71200 }, { "grad_norm": 0.2917245626449585, "learning_rate": 3.265161692076618e-06, "loss": 0.0186, "step": 71210 }, { "grad_norm": 0.2649518549442291, "learning_rate": 3.257819191925282e-06, "loss": 0.0124, "step": 71220 }, { "grad_norm": 0.16110485792160034, "learning_rate": 3.2504846787342392e-06, "loss": 0.0147, "step": 71230 }, { "grad_norm": 0.271019846200943, "learning_rate": 3.243158153756759e-06, "loss": 0.014, "step": 71240 }, { "grad_norm": 0.11786150187253952, "learning_rate": 3.2358396182447514e-06, "loss": 0.0124, "step": 71250 }, { "grad_norm": 0.15012866258621216, "learning_rate": 3.2285290734487496e-06, "loss": 0.0136, "step": 71260 }, { "grad_norm": 0.147919163107872, "learning_rate": 3.221226520617926e-06, "loss": 0.0141, "step": 71270 }, { "grad_norm": 0.29556915163993835, "learning_rate": 3.213931961000094e-06, "loss": 0.0202, "step": 71280 }, { "grad_norm": 0.3146267831325531, "learning_rate": 3.2066453958416733e-06, "loss": 0.0114, "step": 71290 }, { "grad_norm": 0.9378736615180969, "learning_rate": 3.199366826387773e-06, "loss": 0.0199, "step": 71300 }, { "grad_norm": 0.3661051392555237, "learning_rate": 3.192096253882071e-06, "loss": 0.017, "step": 71310 }, { "grad_norm": 0.2705579996109009, "learning_rate": 3.1848336795669177e-06, "loss": 0.0139, "step": 71320 }, { "grad_norm": 0.17014802992343903, "learning_rate": 3.1775791046833035e-06, "loss": 0.0147, "step": 71330 }, { "grad_norm": 0.2075785994529724, "learning_rate": 3.170332530470804e-06, "loss": 0.0137, "step": 71340 }, { "grad_norm": 0.28118348121643066, "learning_rate": 3.1630939581677012e-06, "loss": 0.0188, "step": 71350 }, { "grad_norm": 0.2067214846611023, "learning_rate": 3.155863389010838e-06, "loss": 0.0202, "step": 71360 }, { "grad_norm": 0.21609804034233093, "learning_rate": 3.1486408242357323e-06, "loss": 0.0178, "step": 71370 }, { "grad_norm": 0.3638851046562195, "learning_rate": 3.1414262650765248e-06, "loss": 0.0309, "step": 71380 }, { "grad_norm": 0.31755393743515015, "learning_rate": 3.134219712765979e-06, "loss": 0.0223, "step": 71390 }, { "grad_norm": 0.2056376338005066, "learning_rate": 3.1270211685355044e-06, "loss": 0.0193, "step": 71400 }, { "grad_norm": 0.31937485933303833, "learning_rate": 3.1198306336151338e-06, "loss": 0.0165, "step": 71410 }, { "grad_norm": 0.2974391281604767, "learning_rate": 3.11264810923354e-06, "loss": 0.014, "step": 71420 }, { "grad_norm": 0.08513962477445602, "learning_rate": 3.105473596618008e-06, "loss": 0.0124, "step": 71430 }, { "grad_norm": 0.22971104085445404, "learning_rate": 3.0983070969944683e-06, "loss": 0.0136, "step": 71440 }, { "grad_norm": 0.6680285334587097, "learning_rate": 3.09114861158748e-06, "loss": 0.0163, "step": 71450 }, { "grad_norm": 0.24947160482406616, "learning_rate": 3.0839981416202314e-06, "loss": 0.0151, "step": 71460 }, { "grad_norm": 1.11565363407135, "learning_rate": 3.0768556883145507e-06, "loss": 0.0154, "step": 71470 }, { "grad_norm": 0.17862848937511444, "learning_rate": 3.0697212528908834e-06, "loss": 0.0145, "step": 71480 }, { "grad_norm": 0.3838747441768646, "learning_rate": 3.0625948365683e-06, "loss": 0.0174, "step": 71490 }, { "grad_norm": 0.19885455071926117, "learning_rate": 3.055476440564531e-06, "loss": 0.0183, "step": 71500 }, { "grad_norm": 0.12599822878837585, "learning_rate": 3.0483660660958924e-06, "loss": 0.0134, "step": 71510 }, { "grad_norm": 0.3276005983352661, "learning_rate": 3.0412637143773624e-06, "loss": 0.02, "step": 71520 }, { "grad_norm": 0.251803457736969, "learning_rate": 3.034169386622554e-06, "loss": 0.0099, "step": 71530 }, { "grad_norm": 0.20454935729503632, "learning_rate": 3.027083084043658e-06, "loss": 0.0131, "step": 71540 }, { "grad_norm": 0.22756719589233398, "learning_rate": 3.0200048078515676e-06, "loss": 0.0121, "step": 71550 }, { "grad_norm": 0.24557848274707794, "learning_rate": 3.012934559255742e-06, "loss": 0.016, "step": 71560 }, { "grad_norm": 0.2554757595062256, "learning_rate": 3.0058723394642994e-06, "loss": 0.0154, "step": 71570 }, { "grad_norm": 0.1988402158021927, "learning_rate": 2.998818149683985e-06, "loss": 0.0215, "step": 71580 }, { "grad_norm": 0.17177091538906097, "learning_rate": 2.9917719911201627e-06, "loss": 0.0136, "step": 71590 }, { "grad_norm": 0.19816522300243378, "learning_rate": 2.9847338649768352e-06, "loss": 0.0139, "step": 71600 }, { "grad_norm": 0.2331814467906952, "learning_rate": 2.977703772456608e-06, "loss": 0.0163, "step": 71610 }, { "grad_norm": 0.3165324926376343, "learning_rate": 2.9706817147607535e-06, "loss": 0.0153, "step": 71620 }, { "grad_norm": 0.2598128914833069, "learning_rate": 2.9636676930891393e-06, "loss": 0.0122, "step": 71630 }, { "grad_norm": 0.6375889182090759, "learning_rate": 2.9566617086402625e-06, "loss": 0.0168, "step": 71640 }, { "grad_norm": 0.3333715796470642, "learning_rate": 2.9496637626112655e-06, "loss": 0.0121, "step": 71650 }, { "grad_norm": 0.23561447858810425, "learning_rate": 2.9426738561979027e-06, "loss": 0.0151, "step": 71660 }, { "grad_norm": 0.3167489767074585, "learning_rate": 2.9356919905945524e-06, "loss": 0.0262, "step": 71670 }, { "grad_norm": 0.21891482174396515, "learning_rate": 2.928718166994243e-06, "loss": 0.0117, "step": 71680 }, { "grad_norm": 0.8919470310211182, "learning_rate": 2.9217523865885833e-06, "loss": 0.0189, "step": 71690 }, { "grad_norm": 0.35286903381347656, "learning_rate": 2.9147946505678537e-06, "loss": 0.0173, "step": 71700 }, { "grad_norm": 0.18526162207126617, "learning_rate": 2.9078449601209313e-06, "loss": 0.0139, "step": 71710 }, { "grad_norm": 0.3677004873752594, "learning_rate": 2.900903316435333e-06, "loss": 0.0151, "step": 71720 }, { "grad_norm": 0.2557407021522522, "learning_rate": 2.8939697206971983e-06, "loss": 0.0158, "step": 71730 }, { "grad_norm": 0.2166275978088379, "learning_rate": 2.8870441740912746e-06, "loss": 0.0151, "step": 71740 }, { "grad_norm": 0.2222713679075241, "learning_rate": 2.8801266778009762e-06, "loss": 0.0145, "step": 71750 }, { "grad_norm": 0.22839082777500153, "learning_rate": 2.8732172330082798e-06, "loss": 0.0154, "step": 71760 }, { "grad_norm": 0.14176228642463684, "learning_rate": 2.8663158408938517e-06, "loss": 0.0158, "step": 71770 }, { "grad_norm": 0.1730652153491974, "learning_rate": 2.859422502636938e-06, "loss": 0.0151, "step": 71780 }, { "grad_norm": 0.21500909328460693, "learning_rate": 2.8525372194154076e-06, "loss": 0.0134, "step": 71790 }, { "grad_norm": 0.1716233789920807, "learning_rate": 2.8456599924057913e-06, "loss": 0.0126, "step": 71800 }, { "grad_norm": 0.1941504329442978, "learning_rate": 2.8387908227831995e-06, "loss": 0.0142, "step": 71810 }, { "grad_norm": 0.2248494178056717, "learning_rate": 2.831929711721404e-06, "loss": 0.0156, "step": 71820 }, { "grad_norm": 0.20483292639255524, "learning_rate": 2.825076660392767e-06, "loss": 0.0129, "step": 71830 }, { "grad_norm": 0.14890962839126587, "learning_rate": 2.8182316699682908e-06, "loss": 0.0197, "step": 71840 }, { "grad_norm": 0.12503816187381744, "learning_rate": 2.811394741617601e-06, "loss": 0.0141, "step": 71850 }, { "grad_norm": 0.5270735621452332, "learning_rate": 2.8045658765089356e-06, "loss": 0.0194, "step": 71860 }, { "grad_norm": 0.2860957086086273, "learning_rate": 2.7977450758091605e-06, "loss": 0.0202, "step": 71870 }, { "grad_norm": 0.5891755223274231, "learning_rate": 2.790932340683783e-06, "loss": 0.019, "step": 71880 }, { "grad_norm": 0.41301262378692627, "learning_rate": 2.7841276722968823e-06, "loss": 0.0154, "step": 71890 }, { "grad_norm": 0.22910326719284058, "learning_rate": 2.7773310718112067e-06, "loss": 0.0137, "step": 71900 }, { "grad_norm": 0.23545686900615692, "learning_rate": 2.7705425403881102e-06, "loss": 0.0164, "step": 71910 }, { "grad_norm": 0.3502074182033539, "learning_rate": 2.763762079187565e-06, "loss": 0.0154, "step": 71920 }, { "grad_norm": 0.20447450876235962, "learning_rate": 2.756989689368178e-06, "loss": 0.016, "step": 71930 }, { "grad_norm": 0.4939071834087372, "learning_rate": 2.750225372087134e-06, "loss": 0.0152, "step": 71940 }, { "grad_norm": 0.2404349446296692, "learning_rate": 2.7434691285003033e-06, "loss": 0.0089, "step": 71950 }, { "grad_norm": 0.20347608625888824, "learning_rate": 2.7367209597621178e-06, "loss": 0.0147, "step": 71960 }, { "grad_norm": 0.32799863815307617, "learning_rate": 2.729980867025683e-06, "loss": 0.0157, "step": 71970 }, { "grad_norm": 0.27597373723983765, "learning_rate": 2.7232488514426724e-06, "loss": 0.0176, "step": 71980 }, { "grad_norm": 0.31909939646720886, "learning_rate": 2.7165249141634097e-06, "loss": 0.0138, "step": 71990 }, { "grad_norm": 0.27712687849998474, "learning_rate": 2.709809056336837e-06, "loss": 0.0164, "step": 72000 }, { "grad_norm": 0.3720592260360718, "learning_rate": 2.703101279110498e-06, "loss": 0.0144, "step": 72010 }, { "grad_norm": 0.21687647700309753, "learning_rate": 2.696401583630587e-06, "loss": 0.0139, "step": 72020 }, { "grad_norm": 0.22840255498886108, "learning_rate": 2.689709971041887e-06, "loss": 0.0123, "step": 72030 }, { "grad_norm": 0.2767457664012909, "learning_rate": 2.6830264424878126e-06, "loss": 0.0119, "step": 72040 }, { "grad_norm": 0.23787462711334229, "learning_rate": 2.676350999110394e-06, "loss": 0.0237, "step": 72050 }, { "grad_norm": 0.29992231726646423, "learning_rate": 2.6696836420502856e-06, "loss": 0.0233, "step": 72060 }, { "grad_norm": 0.18649423122406006, "learning_rate": 2.663024372446765e-06, "loss": 0.0124, "step": 72070 }, { "grad_norm": 0.31069397926330566, "learning_rate": 2.6563731914377055e-06, "loss": 0.0107, "step": 72080 }, { "grad_norm": 0.1562502086162567, "learning_rate": 2.6497301001596087e-06, "loss": 0.0181, "step": 72090 }, { "grad_norm": 0.2918016314506531, "learning_rate": 2.643095099747611e-06, "loss": 0.015, "step": 72100 }, { "grad_norm": 0.19509285688400269, "learning_rate": 2.636468191335445e-06, "loss": 0.0151, "step": 72110 }, { "grad_norm": 0.25476235151290894, "learning_rate": 2.629849376055471e-06, "loss": 0.0201, "step": 72120 }, { "grad_norm": 0.23321574926376343, "learning_rate": 2.6232386550386678e-06, "loss": 0.0159, "step": 72130 }, { "grad_norm": 0.2178511917591095, "learning_rate": 2.6166360294146097e-06, "loss": 0.0127, "step": 72140 }, { "grad_norm": 0.30669042468070984, "learning_rate": 2.6100415003115275e-06, "loss": 0.0205, "step": 72150 }, { "grad_norm": 0.2248264104127884, "learning_rate": 2.603455068856225e-06, "loss": 0.0156, "step": 72160 }, { "grad_norm": 0.15748664736747742, "learning_rate": 2.5968767361741584e-06, "loss": 0.0157, "step": 72170 }, { "grad_norm": 0.24428915977478027, "learning_rate": 2.590306503389378e-06, "loss": 0.0227, "step": 72180 }, { "grad_norm": 0.20275838673114777, "learning_rate": 2.583744371624558e-06, "loss": 0.0235, "step": 72190 }, { "grad_norm": 0.2894664406776428, "learning_rate": 2.57719034200099e-06, "loss": 0.0158, "step": 72200 }, { "grad_norm": 0.21400870382785797, "learning_rate": 2.570644415638568e-06, "loss": 0.0169, "step": 72210 }, { "grad_norm": 0.3761967718601227, "learning_rate": 2.5641065936558295e-06, "loss": 0.0189, "step": 72220 }, { "grad_norm": 0.22524011135101318, "learning_rate": 2.557576877169898e-06, "loss": 0.0167, "step": 72230 }, { "grad_norm": 0.3935399651527405, "learning_rate": 2.5510552672965205e-06, "loss": 0.0121, "step": 72240 }, { "grad_norm": 0.2270367592573166, "learning_rate": 2.544541765150077e-06, "loss": 0.012, "step": 72250 }, { "grad_norm": 0.6813557744026184, "learning_rate": 2.5380363718435165e-06, "loss": 0.0185, "step": 72260 }, { "grad_norm": 0.11975834518671036, "learning_rate": 2.5315390884884714e-06, "loss": 0.0118, "step": 72270 }, { "grad_norm": 0.20836453139781952, "learning_rate": 2.5250499161951214e-06, "loss": 0.0139, "step": 72280 }, { "grad_norm": 0.19022604823112488, "learning_rate": 2.518568856072301e-06, "loss": 0.01, "step": 72290 }, { "grad_norm": 0.19262297451496124, "learning_rate": 2.512095909227441e-06, "loss": 0.0133, "step": 72300 }, { "grad_norm": 0.15730734169483185, "learning_rate": 2.50563107676659e-06, "loss": 0.0148, "step": 72310 }, { "grad_norm": 0.22148168087005615, "learning_rate": 2.4991743597944152e-06, "loss": 0.013, "step": 72320 }, { "grad_norm": 0.33621668815612793, "learning_rate": 2.492725759414205e-06, "loss": 0.0165, "step": 72330 }, { "grad_norm": 0.26594477891921997, "learning_rate": 2.4862852767278234e-06, "loss": 0.0114, "step": 72340 }, { "grad_norm": 0.21199049055576324, "learning_rate": 2.479852912835784e-06, "loss": 0.0129, "step": 72350 }, { "grad_norm": 0.33186739683151245, "learning_rate": 2.4734286688372075e-06, "loss": 0.0154, "step": 72360 }, { "grad_norm": 0.9866282939910889, "learning_rate": 2.467012545829811e-06, "loss": 0.0193, "step": 72370 }, { "grad_norm": 0.22115235030651093, "learning_rate": 2.460604544909945e-06, "loss": 0.0143, "step": 72380 }, { "grad_norm": 0.32893508672714233, "learning_rate": 2.454204667172555e-06, "loss": 0.0158, "step": 72390 }, { "grad_norm": 0.24819032847881317, "learning_rate": 2.447812913711217e-06, "loss": 0.0152, "step": 72400 }, { "grad_norm": 0.22546358406543732, "learning_rate": 2.44142928561808e-06, "loss": 0.0172, "step": 72410 }, { "grad_norm": 0.3617135286331177, "learning_rate": 2.435053783983965e-06, "loss": 0.0153, "step": 72420 }, { "grad_norm": 0.23019473254680634, "learning_rate": 2.4286864098982453e-06, "loss": 0.0158, "step": 72430 }, { "grad_norm": 0.12317579239606857, "learning_rate": 2.4223271644489397e-06, "loss": 0.0183, "step": 72440 }, { "grad_norm": 0.21809643507003784, "learning_rate": 2.415976048722679e-06, "loss": 0.0112, "step": 72450 }, { "grad_norm": 0.1579858660697937, "learning_rate": 2.4096330638046673e-06, "loss": 0.0105, "step": 72460 }, { "grad_norm": 0.3197469115257263, "learning_rate": 2.4032982107787816e-06, "loss": 0.0144, "step": 72470 }, { "grad_norm": 0.24666468799114227, "learning_rate": 2.396971490727451e-06, "loss": 0.018, "step": 72480 }, { "grad_norm": 0.17027141153812408, "learning_rate": 2.3906529047317493e-06, "loss": 0.0103, "step": 72490 }, { "grad_norm": 0.19731153547763824, "learning_rate": 2.3843424538713465e-06, "loss": 0.0123, "step": 72500 }, { "grad_norm": 0.21644167602062225, "learning_rate": 2.3780401392245298e-06, "loss": 0.0159, "step": 72510 }, { "grad_norm": 0.20611123740673065, "learning_rate": 2.3717459618681883e-06, "loss": 0.0184, "step": 72520 }, { "grad_norm": 0.20279230177402496, "learning_rate": 2.3654599228778274e-06, "loss": 0.0134, "step": 72530 }, { "grad_norm": 0.37078043818473816, "learning_rate": 2.3591820233275607e-06, "loss": 0.019, "step": 72540 }, { "grad_norm": 0.3032154440879822, "learning_rate": 2.3529122642901024e-06, "loss": 0.0232, "step": 72550 }, { "grad_norm": 0.21971388161182404, "learning_rate": 2.3466506468367845e-06, "loss": 0.0213, "step": 72560 }, { "grad_norm": 0.7291905283927917, "learning_rate": 2.3403971720375506e-06, "loss": 0.0176, "step": 72570 }, { "grad_norm": 0.28479379415512085, "learning_rate": 2.334151840960952e-06, "loss": 0.013, "step": 72580 }, { "grad_norm": 0.5941724181175232, "learning_rate": 2.3279146546741347e-06, "loss": 0.0143, "step": 72590 }, { "grad_norm": 0.21301598846912384, "learning_rate": 2.32168561424288e-06, "loss": 0.0128, "step": 72600 }, { "grad_norm": 0.17943038046360016, "learning_rate": 2.3154647207315307e-06, "loss": 0.0155, "step": 72610 }, { "grad_norm": 0.1435638964176178, "learning_rate": 2.309251975203103e-06, "loss": 0.0166, "step": 72620 }, { "grad_norm": 0.534783661365509, "learning_rate": 2.303047378719159e-06, "loss": 0.0152, "step": 72630 }, { "grad_norm": 0.2369757890701294, "learning_rate": 2.2968509323399e-06, "loss": 0.0208, "step": 72640 }, { "grad_norm": 0.2548963725566864, "learning_rate": 2.290662637124147e-06, "loss": 0.0137, "step": 72650 }, { "grad_norm": 0.12763521075248718, "learning_rate": 2.2844824941292807e-06, "loss": 0.0089, "step": 72660 }, { "grad_norm": 0.316490113735199, "learning_rate": 2.2783105044113406e-06, "loss": 0.015, "step": 72670 }, { "grad_norm": 0.17660348117351532, "learning_rate": 2.272146669024944e-06, "loss": 0.012, "step": 72680 }, { "grad_norm": 0.19457441568374634, "learning_rate": 2.26599098902332e-06, "loss": 0.0161, "step": 72690 }, { "grad_norm": 0.3212920129299164, "learning_rate": 2.2598434654583113e-06, "loss": 0.0159, "step": 72700 }, { "grad_norm": 0.31730714440345764, "learning_rate": 2.253704099380355e-06, "loss": 0.0227, "step": 72710 }, { "grad_norm": 0.35895511507987976, "learning_rate": 2.247572891838512e-06, "loss": 0.0147, "step": 72720 }, { "grad_norm": 0.3358481824398041, "learning_rate": 2.241449843880422e-06, "loss": 0.0178, "step": 72730 }, { "grad_norm": 0.10220763832330704, "learning_rate": 2.235334956552354e-06, "loss": 0.0164, "step": 72740 }, { "grad_norm": 0.26328060030937195, "learning_rate": 2.2292282308991775e-06, "loss": 0.019, "step": 72750 }, { "grad_norm": 0.2907232940196991, "learning_rate": 2.223129667964363e-06, "loss": 0.0141, "step": 72760 }, { "grad_norm": 0.24843667447566986, "learning_rate": 2.2170392687899834e-06, "loss": 0.0116, "step": 72770 }, { "grad_norm": 0.304185152053833, "learning_rate": 2.210957034416733e-06, "loss": 0.0138, "step": 72780 }, { "grad_norm": 0.27330300211906433, "learning_rate": 2.2048829658838867e-06, "loss": 0.0156, "step": 72790 }, { "grad_norm": 0.585191011428833, "learning_rate": 2.1988170642293525e-06, "loss": 0.0112, "step": 72800 }, { "grad_norm": 0.24795441329479218, "learning_rate": 2.1927593304896075e-06, "loss": 0.013, "step": 72810 }, { "grad_norm": 0.09399066865444183, "learning_rate": 2.1867097656997626e-06, "loss": 0.0148, "step": 72820 }, { "grad_norm": 0.25492361187934875, "learning_rate": 2.180668370893524e-06, "loss": 0.0149, "step": 72830 }, { "grad_norm": 0.35681062936782837, "learning_rate": 2.174635147103199e-06, "loss": 0.0121, "step": 72840 }, { "grad_norm": 0.2434069961309433, "learning_rate": 2.1686100953597075e-06, "loss": 0.0142, "step": 72850 }, { "grad_norm": 0.21227681636810303, "learning_rate": 2.1625932166925433e-06, "loss": 0.0133, "step": 72860 }, { "grad_norm": 0.3471039831638336, "learning_rate": 2.1565845121298556e-06, "loss": 0.0165, "step": 72870 }, { "grad_norm": 0.21840158104896545, "learning_rate": 2.150583982698351e-06, "loss": 0.0197, "step": 72880 }, { "grad_norm": 0.16471895575523376, "learning_rate": 2.144591629423359e-06, "loss": 0.0162, "step": 72890 }, { "grad_norm": 0.20510059595108032, "learning_rate": 2.138607453328817e-06, "loss": 0.0139, "step": 72900 }, { "grad_norm": 0.2385919839143753, "learning_rate": 2.132631455437234e-06, "loss": 0.0157, "step": 72910 }, { "grad_norm": 0.24360550940036774, "learning_rate": 2.1266636367697714e-06, "loss": 0.021, "step": 72920 }, { "grad_norm": 0.2776602506637573, "learning_rate": 2.120703998346152e-06, "loss": 0.0147, "step": 72930 }, { "grad_norm": 0.2516339421272278, "learning_rate": 2.1147525411847114e-06, "loss": 0.0152, "step": 72940 }, { "grad_norm": 0.1390431970357895, "learning_rate": 2.108809266302403e-06, "loss": 0.0114, "step": 72950 }, { "grad_norm": 0.28731080889701843, "learning_rate": 2.102874174714764e-06, "loss": 0.0164, "step": 72960 }, { "grad_norm": 0.26799070835113525, "learning_rate": 2.0969472674359393e-06, "loss": 0.0141, "step": 72970 }, { "grad_norm": 0.09352976828813553, "learning_rate": 2.091028545478668e-06, "loss": 0.0148, "step": 72980 }, { "grad_norm": 0.1952284425497055, "learning_rate": 2.08511800985432e-06, "loss": 0.0117, "step": 72990 }, { "grad_norm": 0.16886445879936218, "learning_rate": 2.07921566157282e-06, "loss": 0.0247, "step": 73000 }, { "grad_norm": 0.2919987142086029, "learning_rate": 2.073321501642728e-06, "loss": 0.013, "step": 73010 }, { "grad_norm": 0.13968709111213684, "learning_rate": 2.0674355310711937e-06, "loss": 0.0153, "step": 73020 }, { "grad_norm": 0.19032616913318634, "learning_rate": 2.0615577508639682e-06, "loss": 0.0104, "step": 73030 }, { "grad_norm": 0.18309657275676727, "learning_rate": 2.055688162025399e-06, "loss": 0.0131, "step": 73040 }, { "grad_norm": 0.11618810892105103, "learning_rate": 2.0498267655584546e-06, "loss": 0.0177, "step": 73050 }, { "grad_norm": 0.20857974886894226, "learning_rate": 2.0439735624646626e-06, "loss": 0.0194, "step": 73060 }, { "grad_norm": 0.1716356873512268, "learning_rate": 2.0381285537442008e-06, "loss": 0.0124, "step": 73070 }, { "grad_norm": 0.18802717328071594, "learning_rate": 2.032291740395803e-06, "loss": 0.0152, "step": 73080 }, { "grad_norm": 0.278831422328949, "learning_rate": 2.0264631234168276e-06, "loss": 0.0148, "step": 73090 }, { "grad_norm": 0.3631025552749634, "learning_rate": 2.0206427038032326e-06, "loss": 0.0204, "step": 73100 }, { "grad_norm": 0.3078126013278961, "learning_rate": 2.0148304825495457e-06, "loss": 0.0095, "step": 73110 }, { "grad_norm": 0.3558884263038635, "learning_rate": 2.0090264606489496e-06, "loss": 0.0137, "step": 73120 }, { "grad_norm": 0.5505145192146301, "learning_rate": 2.003230639093162e-06, "loss": 0.0153, "step": 73130 }, { "grad_norm": 0.31050699949264526, "learning_rate": 1.9974430188725525e-06, "loss": 0.0156, "step": 73140 }, { "grad_norm": 0.2803749442100525, "learning_rate": 1.9916636009760513e-06, "loss": 0.0198, "step": 73150 }, { "grad_norm": 0.13834460079669952, "learning_rate": 1.985892386391208e-06, "loss": 0.0146, "step": 73160 }, { "grad_norm": 0.2358432561159134, "learning_rate": 1.980129376104173e-06, "loss": 0.0117, "step": 73170 }, { "grad_norm": 0.18065562844276428, "learning_rate": 1.9743745710996796e-06, "loss": 0.0136, "step": 73180 }, { "grad_norm": 0.2242133468389511, "learning_rate": 1.9686279723610757e-06, "loss": 0.0147, "step": 73190 }, { "grad_norm": 0.19986677169799805, "learning_rate": 1.962889580870281e-06, "loss": 0.011, "step": 73200 }, { "grad_norm": 0.1824013590812683, "learning_rate": 1.9571593976078385e-06, "loss": 0.0183, "step": 73210 }, { "grad_norm": 0.2531324028968811, "learning_rate": 1.9514374235528876e-06, "loss": 0.0148, "step": 73220 }, { "grad_norm": 0.2323574274778366, "learning_rate": 1.9457236596831408e-06, "loss": 0.0103, "step": 73230 }, { "grad_norm": 0.1584121733903885, "learning_rate": 1.940018106974939e-06, "loss": 0.0148, "step": 73240 }, { "grad_norm": 0.27036285400390625, "learning_rate": 1.9343207664032025e-06, "loss": 0.013, "step": 73250 }, { "grad_norm": 0.1666731834411621, "learning_rate": 1.928631638941436e-06, "loss": 0.0135, "step": 73260 }, { "grad_norm": 0.23819945752620697, "learning_rate": 1.922950725561784e-06, "loss": 0.0131, "step": 73270 }, { "grad_norm": 0.3258039951324463, "learning_rate": 1.9172780272349312e-06, "loss": 0.0144, "step": 73280 }, { "grad_norm": 0.20202499628067017, "learning_rate": 1.911613544930202e-06, "loss": 0.0133, "step": 73290 }, { "grad_norm": 0.24107857048511505, "learning_rate": 1.9059572796155e-06, "loss": 0.015, "step": 73300 }, { "grad_norm": 0.2166711986064911, "learning_rate": 1.9003092322573192e-06, "loss": 0.016, "step": 73310 }, { "grad_norm": 0.16206085681915283, "learning_rate": 1.8946694038207647e-06, "loss": 0.0151, "step": 73320 }, { "grad_norm": 0.284033864736557, "learning_rate": 1.8890377952695215e-06, "loss": 0.0129, "step": 73330 }, { "grad_norm": 0.25093626976013184, "learning_rate": 1.8834144075658865e-06, "loss": 0.012, "step": 73340 }, { "grad_norm": 0.38129186630249023, "learning_rate": 1.8777992416707357e-06, "loss": 0.0123, "step": 73350 }, { "grad_norm": 0.18404032289981842, "learning_rate": 1.8721922985435458e-06, "loss": 0.0109, "step": 73360 }, { "grad_norm": 0.43144503235816956, "learning_rate": 1.8665935791424006e-06, "loss": 0.018, "step": 73370 }, { "grad_norm": 0.2791576385498047, "learning_rate": 1.8610030844239512e-06, "loss": 0.0123, "step": 73380 }, { "grad_norm": 0.21309314668178558, "learning_rate": 1.8554208153434838e-06, "loss": 0.0143, "step": 73390 }, { "grad_norm": 0.15043939650058746, "learning_rate": 1.8498467728548296e-06, "loss": 0.0119, "step": 73400 }, { "grad_norm": 0.3219263255596161, "learning_rate": 1.8442809579104547e-06, "loss": 0.0128, "step": 73410 }, { "grad_norm": 0.2665400505065918, "learning_rate": 1.8387233714614039e-06, "loss": 0.0169, "step": 73420 }, { "grad_norm": 0.15042108297348022, "learning_rate": 1.8331740144573174e-06, "loss": 0.0233, "step": 73430 }, { "grad_norm": 0.25629350543022156, "learning_rate": 1.8276328878464255e-06, "loss": 0.0178, "step": 73440 }, { "grad_norm": 0.41348275542259216, "learning_rate": 1.82209999257556e-06, "loss": 0.0186, "step": 73450 }, { "grad_norm": 0.1890062838792801, "learning_rate": 1.8165753295901311e-06, "loss": 0.0144, "step": 73460 }, { "grad_norm": 0.24430479109287262, "learning_rate": 1.8110588998341616e-06, "loss": 0.0154, "step": 73470 }, { "grad_norm": 0.18815842270851135, "learning_rate": 1.805550704250253e-06, "loss": 0.0206, "step": 73480 }, { "grad_norm": 0.26936623454093933, "learning_rate": 1.8000507437796077e-06, "loss": 0.0176, "step": 73490 }, { "grad_norm": 0.23777177929878235, "learning_rate": 1.7945590193620242e-06, "loss": 0.0163, "step": 73500 }, { "grad_norm": 0.21529676020145416, "learning_rate": 1.7890755319358742e-06, "loss": 0.0129, "step": 73510 }, { "grad_norm": 0.3308732807636261, "learning_rate": 1.7836002824381525e-06, "loss": 0.0165, "step": 73520 }, { "grad_norm": 0.1608096808195114, "learning_rate": 1.7781332718044165e-06, "loss": 0.0136, "step": 73530 }, { "grad_norm": 0.31931033730506897, "learning_rate": 1.772674500968835e-06, "loss": 0.0135, "step": 73540 }, { "grad_norm": 0.3014717698097229, "learning_rate": 1.7672239708641624e-06, "loss": 0.0191, "step": 73550 }, { "grad_norm": 0.31550776958465576, "learning_rate": 1.7617816824217371e-06, "loss": 0.0105, "step": 73560 }, { "grad_norm": 0.21669870615005493, "learning_rate": 1.7563476365715148e-06, "loss": 0.0217, "step": 73570 }, { "grad_norm": 0.11024957150220871, "learning_rate": 1.750921834242003e-06, "loss": 0.0177, "step": 73580 }, { "grad_norm": 0.27022990584373474, "learning_rate": 1.7455042763603436e-06, "loss": 0.0169, "step": 73590 }, { "grad_norm": 0.26356545090675354, "learning_rate": 1.7400949638522345e-06, "loss": 0.0111, "step": 73600 }, { "grad_norm": 0.25836455821990967, "learning_rate": 1.734693897641987e-06, "loss": 0.0285, "step": 73610 }, { "grad_norm": 0.2549266219139099, "learning_rate": 1.7293010786524955e-06, "loss": 0.0134, "step": 73620 }, { "grad_norm": 0.282126247882843, "learning_rate": 1.72391650780524e-06, "loss": 0.0173, "step": 73630 }, { "grad_norm": 0.4954054355621338, "learning_rate": 1.718540186020301e-06, "loss": 0.0184, "step": 73640 }, { "grad_norm": 0.17101337015628815, "learning_rate": 1.7131721142163437e-06, "loss": 0.0198, "step": 73650 }, { "grad_norm": 0.22954098880290985, "learning_rate": 1.7078122933106233e-06, "loss": 0.0157, "step": 73660 }, { "grad_norm": 0.15806607902050018, "learning_rate": 1.7024607242189905e-06, "loss": 0.0187, "step": 73670 }, { "grad_norm": 0.24803152680397034, "learning_rate": 1.6971174078558749e-06, "loss": 0.0126, "step": 73680 }, { "grad_norm": 0.3133683204650879, "learning_rate": 1.6917823451343073e-06, "loss": 0.0182, "step": 73690 }, { "grad_norm": 0.18597464263439178, "learning_rate": 1.6864555369659142e-06, "loss": 0.0118, "step": 73700 }, { "grad_norm": 0.19604021310806274, "learning_rate": 1.681136984260878e-06, "loss": 0.0153, "step": 73710 }, { "grad_norm": 0.3563467860221863, "learning_rate": 1.6758266879280172e-06, "loss": 0.0142, "step": 73720 }, { "grad_norm": 0.5173802971839905, "learning_rate": 1.6705246488746996e-06, "loss": 0.0135, "step": 73730 }, { "grad_norm": 0.2049895077943802, "learning_rate": 1.6652308680069062e-06, "loss": 0.0135, "step": 73740 }, { "grad_norm": 0.20033760368824005, "learning_rate": 1.6599453462292081e-06, "loss": 0.0141, "step": 73750 }, { "grad_norm": 0.5365527868270874, "learning_rate": 1.6546680844447326e-06, "loss": 0.0166, "step": 73760 }, { "grad_norm": 0.17671555280685425, "learning_rate": 1.6493990835552475e-06, "loss": 0.0161, "step": 73770 }, { "grad_norm": 0.16448253393173218, "learning_rate": 1.6441383444610547e-06, "loss": 0.0124, "step": 73780 }, { "grad_norm": 0.22846326231956482, "learning_rate": 1.6388858680610908e-06, "loss": 0.0177, "step": 73790 }, { "grad_norm": 0.32441607117652893, "learning_rate": 1.6336416552528543e-06, "loss": 0.0133, "step": 73800 }, { "grad_norm": 0.29975801706314087, "learning_rate": 1.6284057069324288e-06, "loss": 0.0174, "step": 73810 }, { "grad_norm": 0.24160514771938324, "learning_rate": 1.623178023994504e-06, "loss": 0.0155, "step": 73820 }, { "grad_norm": 0.4666728377342224, "learning_rate": 1.6179586073323483e-06, "loss": 0.0172, "step": 73830 }, { "grad_norm": 0.34554192423820496, "learning_rate": 1.612747457837821e-06, "loss": 0.0181, "step": 73840 }, { "grad_norm": 0.4251670837402344, "learning_rate": 1.607544576401354e-06, "loss": 0.0191, "step": 73850 }, { "grad_norm": 0.25144726037979126, "learning_rate": 1.6023499639119754e-06, "loss": 0.0147, "step": 73860 }, { "grad_norm": 0.1514735072851181, "learning_rate": 1.5971636212573138e-06, "loss": 0.0143, "step": 73870 }, { "grad_norm": 0.2944766879081726, "learning_rate": 1.5919855493235714e-06, "loss": 0.0176, "step": 73880 }, { "grad_norm": 0.22444364428520203, "learning_rate": 1.586815748995535e-06, "loss": 0.0119, "step": 73890 }, { "grad_norm": 0.2104773074388504, "learning_rate": 1.5816542211565866e-06, "loss": 0.0158, "step": 73900 }, { "grad_norm": 0.5184223055839539, "learning_rate": 1.5765009666886765e-06, "loss": 0.0193, "step": 73910 }, { "grad_norm": 0.3228665292263031, "learning_rate": 1.5713559864723781e-06, "loss": 0.0139, "step": 73920 }, { "grad_norm": 0.3092217445373535, "learning_rate": 1.5662192813868104e-06, "loss": 0.0135, "step": 73930 }, { "grad_norm": 0.26167693734169006, "learning_rate": 1.561090852309699e-06, "loss": 0.0218, "step": 73940 }, { "grad_norm": 0.25162792205810547, "learning_rate": 1.5559707001173651e-06, "loss": 0.0175, "step": 73950 }, { "grad_norm": 0.3106610178947449, "learning_rate": 1.5508588256846757e-06, "loss": 0.0147, "step": 73960 }, { "grad_norm": 0.14318402111530304, "learning_rate": 1.5457552298851319e-06, "loss": 0.0129, "step": 73970 }, { "grad_norm": 0.3024682402610779, "learning_rate": 1.5406599135907918e-06, "loss": 0.0153, "step": 73980 }, { "grad_norm": 0.20037852227687836, "learning_rate": 1.5355728776723088e-06, "loss": 0.0132, "step": 73990 }, { "grad_norm": 0.14455224573612213, "learning_rate": 1.5304941229989155e-06, "loss": 0.0184, "step": 74000 }, { "grad_norm": 0.2887812554836273, "learning_rate": 1.5254236504384345e-06, "loss": 0.0153, "step": 74010 }, { "grad_norm": 0.16781438887119293, "learning_rate": 1.5203614608572726e-06, "loss": 0.0142, "step": 74020 }, { "grad_norm": 0.3737504780292511, "learning_rate": 1.5153075551204044e-06, "loss": 0.0115, "step": 74030 }, { "grad_norm": 0.2887272536754608, "learning_rate": 1.5102619340914225e-06, "loss": 0.0242, "step": 74040 }, { "grad_norm": 0.4114674925804138, "learning_rate": 1.505224598632482e-06, "loss": 0.0157, "step": 74050 }, { "grad_norm": 0.17451776564121246, "learning_rate": 1.5001955496043164e-06, "loss": 0.0173, "step": 74060 }, { "grad_norm": 0.38892799615859985, "learning_rate": 1.4951747878662604e-06, "loss": 0.0119, "step": 74070 }, { "grad_norm": 0.14494571089744568, "learning_rate": 1.4901623142762221e-06, "loss": 0.0109, "step": 74080 }, { "grad_norm": 0.1947280764579773, "learning_rate": 1.4851581296907001e-06, "loss": 0.0176, "step": 74090 }, { "grad_norm": 0.6666743755340576, "learning_rate": 1.4801622349647714e-06, "loss": 0.0153, "step": 74100 }, { "grad_norm": 0.2710045874118805, "learning_rate": 1.4751746309520976e-06, "loss": 0.0143, "step": 74110 }, { "grad_norm": 0.32296517491340637, "learning_rate": 1.4701953185049132e-06, "loss": 0.0184, "step": 74120 }, { "grad_norm": 0.18633021414279938, "learning_rate": 1.4652242984740661e-06, "loss": 0.0179, "step": 74130 }, { "grad_norm": 0.2500033676624298, "learning_rate": 1.4602615717089484e-06, "loss": 0.0216, "step": 74140 }, { "grad_norm": 0.26547959446907043, "learning_rate": 1.4553071390575657e-06, "loss": 0.0158, "step": 74150 }, { "grad_norm": 0.22385814785957336, "learning_rate": 1.450361001366496e-06, "loss": 0.017, "step": 74160 }, { "grad_norm": 0.28847724199295044, "learning_rate": 1.4454231594809021e-06, "loss": 0.0148, "step": 74170 }, { "grad_norm": 0.4704039990901947, "learning_rate": 1.4404936142445036e-06, "loss": 0.0198, "step": 74180 }, { "grad_norm": 0.4298340380191803, "learning_rate": 1.4355723664996546e-06, "loss": 0.015, "step": 74190 }, { "grad_norm": 0.253109872341156, "learning_rate": 1.4306594170872433e-06, "loss": 0.0161, "step": 74200 }, { "grad_norm": 0.24318274855613708, "learning_rate": 1.4257547668467598e-06, "loss": 0.019, "step": 74210 }, { "grad_norm": 0.18575814366340637, "learning_rate": 1.4208584166162886e-06, "loss": 0.0124, "step": 74220 }, { "grad_norm": 0.6245232224464417, "learning_rate": 1.4159703672324553e-06, "loss": 0.0148, "step": 74230 }, { "grad_norm": 0.22606156766414642, "learning_rate": 1.4110906195305252e-06, "loss": 0.0158, "step": 74240 }, { "grad_norm": 0.2847045958042145, "learning_rate": 1.406219174344292e-06, "loss": 0.0142, "step": 74250 }, { "grad_norm": 0.2162850946187973, "learning_rate": 1.401356032506157e-06, "loss": 0.015, "step": 74260 }, { "grad_norm": 0.2767638564109802, "learning_rate": 1.3965011948471051e-06, "loss": 0.0191, "step": 74270 }, { "grad_norm": 0.26989060640335083, "learning_rate": 1.3916546621966842e-06, "loss": 0.0124, "step": 74280 }, { "grad_norm": 0.16944608092308044, "learning_rate": 1.3868164353830482e-06, "loss": 0.0097, "step": 74290 }, { "grad_norm": 0.3019021153450012, "learning_rate": 1.3819865152329082e-06, "loss": 0.0136, "step": 74300 }, { "grad_norm": 0.19749954342842102, "learning_rate": 1.377164902571565e-06, "loss": 0.0145, "step": 74310 }, { "grad_norm": 0.20022229850292206, "learning_rate": 1.372351598222904e-06, "loss": 0.0136, "step": 74320 }, { "grad_norm": 0.23813705146312714, "learning_rate": 1.3675466030093843e-06, "loss": 0.0147, "step": 74330 }, { "grad_norm": 0.3674720227718353, "learning_rate": 1.3627499177520486e-06, "loss": 0.0151, "step": 74340 }, { "grad_norm": 0.329642117023468, "learning_rate": 1.3579615432705196e-06, "loss": 0.0189, "step": 74350 }, { "grad_norm": 0.2098066359758377, "learning_rate": 1.3531814803829978e-06, "loss": 0.0136, "step": 74360 }, { "grad_norm": 0.25390443205833435, "learning_rate": 1.3484097299062803e-06, "loss": 0.0138, "step": 74370 }, { "grad_norm": 0.24526409804821014, "learning_rate": 1.343646292655698e-06, "loss": 0.0179, "step": 74380 }, { "grad_norm": 0.2589835822582245, "learning_rate": 1.3388911694452277e-06, "loss": 0.013, "step": 74390 }, { "grad_norm": 0.39731860160827637, "learning_rate": 1.3341443610873638e-06, "loss": 0.016, "step": 74400 }, { "grad_norm": 0.21788328886032104, "learning_rate": 1.329405868393213e-06, "loss": 0.0123, "step": 74410 }, { "grad_norm": 0.2731983959674835, "learning_rate": 1.3246756921724613e-06, "loss": 0.0207, "step": 74420 }, { "grad_norm": 0.11549431830644608, "learning_rate": 1.3199538332333506e-06, "loss": 0.0119, "step": 74430 }, { "grad_norm": 0.2694450318813324, "learning_rate": 1.3152402923827411e-06, "loss": 0.0236, "step": 74440 }, { "grad_norm": 0.1960960030555725, "learning_rate": 1.3105350704260277e-06, "loss": 0.0128, "step": 74450 }, { "grad_norm": 0.11062859743833542, "learning_rate": 1.305838168167206e-06, "loss": 0.0128, "step": 74460 }, { "grad_norm": 0.14791415631771088, "learning_rate": 1.301149586408862e-06, "loss": 0.0106, "step": 74470 }, { "grad_norm": 0.18954303860664368, "learning_rate": 1.2964693259521322e-06, "loss": 0.0176, "step": 74480 }, { "grad_norm": 0.27105453610420227, "learning_rate": 1.2917973875967548e-06, "loss": 0.014, "step": 74490 }, { "grad_norm": 0.5326432585716248, "learning_rate": 1.2871337721410249e-06, "loss": 0.0198, "step": 74500 }, { "grad_norm": 0.21600216627120972, "learning_rate": 1.2824784803818379e-06, "loss": 0.017, "step": 74510 }, { "grad_norm": 0.34716546535491943, "learning_rate": 1.2778315131146467e-06, "loss": 0.0129, "step": 74520 }, { "grad_norm": 0.2795942723751068, "learning_rate": 1.2731928711334994e-06, "loss": 0.0127, "step": 74530 }, { "grad_norm": 0.25039637088775635, "learning_rate": 1.2685625552310122e-06, "loss": 0.0143, "step": 74540 }, { "grad_norm": 0.2556111812591553, "learning_rate": 1.263940566198374e-06, "loss": 0.0152, "step": 74550 }, { "grad_norm": 0.2520044445991516, "learning_rate": 1.259326904825353e-06, "loss": 0.0167, "step": 74560 }, { "grad_norm": 0.3334246277809143, "learning_rate": 1.2547215719003137e-06, "loss": 0.0186, "step": 74570 }, { "grad_norm": 0.2594083845615387, "learning_rate": 1.2501245682101703e-06, "loss": 0.0134, "step": 74580 }, { "grad_norm": 1.9649664163589478, "learning_rate": 1.2455358945404171e-06, "loss": 0.0169, "step": 74590 }, { "grad_norm": 0.17763040959835052, "learning_rate": 1.2409555516751493e-06, "loss": 0.0165, "step": 74600 }, { "grad_norm": 0.4595348834991455, "learning_rate": 1.2363835403970125e-06, "loss": 0.0147, "step": 74610 }, { "grad_norm": 0.30997374653816223, "learning_rate": 1.2318198614872489e-06, "loss": 0.0157, "step": 74620 }, { "grad_norm": 0.3057887852191925, "learning_rate": 1.2272645157256457e-06, "loss": 0.014, "step": 74630 }, { "grad_norm": 0.22570489346981049, "learning_rate": 1.222717503890608e-06, "loss": 0.019, "step": 74640 }, { "grad_norm": 0.40746283531188965, "learning_rate": 1.218178826759081e-06, "loss": 0.0302, "step": 74650 }, { "grad_norm": 0.3217718303203583, "learning_rate": 1.213648485106611e-06, "loss": 0.0212, "step": 74660 }, { "grad_norm": 0.18831251561641693, "learning_rate": 1.2091264797073066e-06, "loss": 0.0223, "step": 74670 }, { "grad_norm": 0.2908083200454712, "learning_rate": 1.2046128113338494e-06, "loss": 0.0242, "step": 74680 }, { "grad_norm": 0.13256852328777313, "learning_rate": 1.200107480757512e-06, "loss": 0.0175, "step": 74690 }, { "grad_norm": 0.1553049236536026, "learning_rate": 1.1956104887481168e-06, "loss": 0.0148, "step": 74700 }, { "grad_norm": 0.17882724106311798, "learning_rate": 1.191121836074094e-06, "loss": 0.0113, "step": 74710 }, { "grad_norm": 0.3027585446834564, "learning_rate": 1.1866415235024186e-06, "loss": 0.0125, "step": 74720 }, { "grad_norm": 0.2484479546546936, "learning_rate": 1.182169551798662e-06, "loss": 0.0184, "step": 74730 }, { "grad_norm": 0.17849649488925934, "learning_rate": 1.177705921726957e-06, "loss": 0.0115, "step": 74740 }, { "grad_norm": 0.18038013577461243, "learning_rate": 1.173250634050016e-06, "loss": 0.013, "step": 74750 }, { "grad_norm": 0.33187323808670044, "learning_rate": 1.1688036895291354e-06, "loss": 0.0136, "step": 74760 }, { "grad_norm": 0.27713361382484436, "learning_rate": 1.1643650889241574e-06, "loss": 0.0124, "step": 74770 }, { "grad_norm": 0.22317475080490112, "learning_rate": 1.1599348329935311e-06, "loss": 0.0153, "step": 74780 }, { "grad_norm": 0.16662700474262238, "learning_rate": 1.1555129224942673e-06, "loss": 0.0207, "step": 74790 }, { "grad_norm": 0.20613454282283783, "learning_rate": 1.1510993581819396e-06, "loss": 0.0164, "step": 74800 }, { "grad_norm": 0.24818532168865204, "learning_rate": 1.1466941408107112e-06, "loss": 0.0174, "step": 74810 }, { "grad_norm": 0.23511026799678802, "learning_rate": 1.1422972711333247e-06, "loss": 0.0119, "step": 74820 }, { "grad_norm": 0.2825203537940979, "learning_rate": 1.1379087499010565e-06, "loss": 0.0185, "step": 74830 }, { "grad_norm": 0.31481269001960754, "learning_rate": 1.1335285778638127e-06, "loss": 0.012, "step": 74840 }, { "grad_norm": 0.3081441819667816, "learning_rate": 1.1291567557700333e-06, "loss": 0.016, "step": 74850 }, { "grad_norm": 0.3186565339565277, "learning_rate": 1.124793284366743e-06, "loss": 0.0153, "step": 74860 }, { "grad_norm": 0.37295466661453247, "learning_rate": 1.1204381643995455e-06, "loss": 0.0135, "step": 74870 }, { "grad_norm": 0.29995670914649963, "learning_rate": 1.116091396612595e-06, "loss": 0.0164, "step": 74880 }, { "grad_norm": 0.2580283582210541, "learning_rate": 1.1117529817486538e-06, "loss": 0.0103, "step": 74890 }, { "grad_norm": 0.2577809989452362, "learning_rate": 1.107422920549034e-06, "loss": 0.0179, "step": 74900 }, { "grad_norm": 0.27102935314178467, "learning_rate": 1.1031012137536157e-06, "loss": 0.0169, "step": 74910 }, { "grad_norm": 0.28478795289993286, "learning_rate": 1.0987878621008695e-06, "loss": 0.0138, "step": 74920 }, { "grad_norm": 0.3347522020339966, "learning_rate": 1.0944828663278284e-06, "loss": 0.0193, "step": 74930 }, { "grad_norm": 0.1573619246482849, "learning_rate": 1.0901862271700925e-06, "loss": 0.0173, "step": 74940 }, { "grad_norm": 0.2437313050031662, "learning_rate": 1.0858979453618467e-06, "loss": 0.0181, "step": 74950 }, { "grad_norm": 0.2706845998764038, "learning_rate": 1.0816180216358441e-06, "loss": 0.0206, "step": 74960 }, { "grad_norm": 0.7155145406723022, "learning_rate": 1.0773464567233937e-06, "loss": 0.0144, "step": 74970 }, { "grad_norm": 0.40006470680236816, "learning_rate": 1.0730832513543953e-06, "loss": 0.014, "step": 74980 }, { "grad_norm": 0.1713392734527588, "learning_rate": 1.0688284062573218e-06, "loss": 0.0134, "step": 74990 }, { "grad_norm": 0.34191879630088806, "learning_rate": 1.0645819221591969e-06, "loss": 0.0166, "step": 75000 }, { "grad_norm": 0.1821497082710266, "learning_rate": 1.0603437997856347e-06, "loss": 0.0138, "step": 75010 }, { "grad_norm": 0.11460693180561066, "learning_rate": 1.0561140398608228e-06, "loss": 0.0148, "step": 75020 }, { "grad_norm": 0.0908254086971283, "learning_rate": 1.0518926431074937e-06, "loss": 0.0126, "step": 75030 }, { "grad_norm": 0.3110290467739105, "learning_rate": 1.0476796102469877e-06, "loss": 0.0167, "step": 75040 }, { "grad_norm": 0.1951356679201126, "learning_rate": 1.0434749419991786e-06, "loss": 0.0142, "step": 75050 }, { "grad_norm": 0.17171494662761688, "learning_rate": 1.0392786390825415e-06, "loss": 0.0141, "step": 75060 }, { "grad_norm": 0.339395135641098, "learning_rate": 1.0350907022141087e-06, "loss": 0.0153, "step": 75070 }, { "grad_norm": 0.3317647874355316, "learning_rate": 1.0309111321094744e-06, "loss": 0.0192, "step": 75080 }, { "grad_norm": 0.24339769780635834, "learning_rate": 1.026739929482834e-06, "loss": 0.0144, "step": 75090 }, { "grad_norm": 0.3493650555610657, "learning_rate": 1.022577095046906e-06, "loss": 0.019, "step": 75100 }, { "grad_norm": 0.4090566039085388, "learning_rate": 1.0184226295130217e-06, "loss": 0.0158, "step": 75110 }, { "grad_norm": 0.29164016246795654, "learning_rate": 1.0142765335910575e-06, "loss": 0.0174, "step": 75120 }, { "grad_norm": 0.3664390742778778, "learning_rate": 1.0101388079894746e-06, "loss": 0.0161, "step": 75130 }, { "grad_norm": 0.11007663607597351, "learning_rate": 1.0060094534152908e-06, "loss": 0.0131, "step": 75140 }, { "grad_norm": 0.1598302572965622, "learning_rate": 1.0018884705741028e-06, "loss": 0.0118, "step": 75150 }, { "grad_norm": 0.6675712466239929, "learning_rate": 9.977758601700805e-07, "loss": 0.0194, "step": 75160 }, { "grad_norm": 0.3748452067375183, "learning_rate": 9.93671622905945e-07, "loss": 0.0222, "step": 75170 }, { "grad_norm": 0.3590743839740753, "learning_rate": 9.895757594830024e-07, "loss": 0.0123, "step": 75180 }, { "grad_norm": 0.16914036870002747, "learning_rate": 9.854882706011204e-07, "loss": 0.011, "step": 75190 }, { "grad_norm": 0.20678333938121796, "learning_rate": 9.814091569587513e-07, "loss": 0.0142, "step": 75200 }, { "grad_norm": 0.34671831130981445, "learning_rate": 9.77338419252888e-07, "loss": 0.0159, "step": 75210 }, { "grad_norm": 0.2925364375114441, "learning_rate": 9.732760581791289e-07, "loss": 0.0134, "step": 75220 }, { "grad_norm": 0.24193155765533447, "learning_rate": 9.692220744315972e-07, "loss": 0.0184, "step": 75230 }, { "grad_norm": 0.12992113828659058, "learning_rate": 9.651764687030162e-07, "loss": 0.0148, "step": 75240 }, { "grad_norm": 0.2464318573474884, "learning_rate": 9.611392416846776e-07, "loss": 0.015, "step": 75250 }, { "grad_norm": 0.23113936185836792, "learning_rate": 9.571103940664238e-07, "loss": 0.0166, "step": 75260 }, { "grad_norm": 0.1581936925649643, "learning_rate": 9.530899265366822e-07, "loss": 0.0178, "step": 75270 }, { "grad_norm": 0.28986454010009766, "learning_rate": 9.490778397824307e-07, "loss": 0.0133, "step": 75280 }, { "grad_norm": 0.3574711084365845, "learning_rate": 9.450741344892322e-07, "loss": 0.0152, "step": 75290 }, { "grad_norm": 0.3832803964614868, "learning_rate": 9.410788113412116e-07, "loss": 0.0149, "step": 75300 }, { "grad_norm": 0.25390395522117615, "learning_rate": 9.370918710210563e-07, "loss": 0.0136, "step": 75310 }, { "grad_norm": 0.2344391644001007, "learning_rate": 9.331133142100268e-07, "loss": 0.0156, "step": 75320 }, { "grad_norm": 0.2855849862098694, "learning_rate": 9.291431415879459e-07, "loss": 0.0135, "step": 75330 }, { "grad_norm": 0.22875358164310455, "learning_rate": 9.251813538332155e-07, "loss": 0.0194, "step": 75340 }, { "grad_norm": 0.18525570631027222, "learning_rate": 9.212279516227884e-07, "loss": 0.0164, "step": 75350 }, { "grad_norm": 0.2778867781162262, "learning_rate": 9.172829356322021e-07, "loss": 0.0166, "step": 75360 }, { "grad_norm": 0.14507976174354553, "learning_rate": 9.133463065355452e-07, "loss": 0.0088, "step": 75370 }, { "grad_norm": 0.26153823733329773, "learning_rate": 9.094180650054796e-07, "loss": 0.0145, "step": 75380 }, { "grad_norm": 0.2987646162509918, "learning_rate": 9.054982117132404e-07, "loss": 0.0157, "step": 75390 }, { "grad_norm": 0.270569384098053, "learning_rate": 9.015867473286143e-07, "loss": 0.015, "step": 75400 }, { "grad_norm": 0.42297476530075073, "learning_rate": 8.976836725199778e-07, "loss": 0.0125, "step": 75410 }, { "grad_norm": 0.33329933881759644, "learning_rate": 8.937889879542416e-07, "loss": 0.0145, "step": 75420 }, { "grad_norm": 0.1978358030319214, "learning_rate": 8.899026942969068e-07, "loss": 0.0112, "step": 75430 }, { "grad_norm": 0.22531722486019135, "learning_rate": 8.860247922120424e-07, "loss": 0.018, "step": 75440 }, { "grad_norm": 0.3356478214263916, "learning_rate": 8.821552823622737e-07, "loss": 0.0145, "step": 75450 }, { "grad_norm": 0.31696614623069763, "learning_rate": 8.782941654087884e-07, "loss": 0.0196, "step": 75460 }, { "grad_norm": 0.30746451020240784, "learning_rate": 8.744414420113478e-07, "loss": 0.0158, "step": 75470 }, { "grad_norm": 0.22354888916015625, "learning_rate": 8.705971128282753e-07, "loss": 0.012, "step": 75480 }, { "grad_norm": 0.2256380319595337, "learning_rate": 8.66761178516473e-07, "loss": 0.0146, "step": 75490 }, { "grad_norm": 0.18585631251335144, "learning_rate": 8.629336397313781e-07, "loss": 0.012, "step": 75500 }, { "grad_norm": 0.15533465147018433, "learning_rate": 8.591144971270282e-07, "loss": 0.0169, "step": 75510 }, { "grad_norm": 0.2204105406999588, "learning_rate": 8.553037513560069e-07, "loss": 0.0166, "step": 75520 }, { "grad_norm": 0.2152707874774933, "learning_rate": 8.515014030694546e-07, "loss": 0.0108, "step": 75530 }, { "grad_norm": 0.25963330268859863, "learning_rate": 8.47707452917107e-07, "loss": 0.0163, "step": 75540 }, { "grad_norm": 0.17789226770401, "learning_rate": 8.439219015472343e-07, "loss": 0.0174, "step": 75550 }, { "grad_norm": 1.248380184173584, "learning_rate": 8.401447496066861e-07, "loss": 0.0095, "step": 75560 }, { "grad_norm": 0.463314950466156, "learning_rate": 8.363759977408792e-07, "loss": 0.0118, "step": 75570 }, { "grad_norm": 0.26938334107398987, "learning_rate": 8.326156465937817e-07, "loss": 0.0183, "step": 75580 }, { "grad_norm": 0.4789147973060608, "learning_rate": 8.28863696807941e-07, "loss": 0.0211, "step": 75590 }, { "grad_norm": 0.2110677808523178, "learning_rate": 8.251201490244609e-07, "loss": 0.0135, "step": 75600 }, { "grad_norm": 0.18654808402061462, "learning_rate": 8.213850038830129e-07, "loss": 0.0162, "step": 75610 }, { "grad_norm": 0.23060478270053864, "learning_rate": 8.176582620218309e-07, "loss": 0.0141, "step": 75620 }, { "grad_norm": 0.19509707391262054, "learning_rate": 8.139399240777057e-07, "loss": 0.0128, "step": 75630 }, { "grad_norm": 0.24400198459625244, "learning_rate": 8.102299906860122e-07, "loss": 0.0143, "step": 75640 }, { "grad_norm": 0.33992716670036316, "learning_rate": 8.065284624806657e-07, "loss": 0.0179, "step": 75650 }, { "grad_norm": 0.19022680819034576, "learning_rate": 8.028353400941601e-07, "loss": 0.0144, "step": 75660 }, { "grad_norm": 0.10684002935886383, "learning_rate": 7.991506241575575e-07, "loss": 0.0166, "step": 75670 }, { "grad_norm": 0.3770938217639923, "learning_rate": 7.954743153004541e-07, "loss": 0.0223, "step": 75680 }, { "grad_norm": 0.26641348004341125, "learning_rate": 7.918064141510528e-07, "loss": 0.0186, "step": 75690 }, { "grad_norm": 0.18020176887512207, "learning_rate": 7.881469213360859e-07, "loss": 0.021, "step": 75700 }, { "grad_norm": 0.1431182473897934, "learning_rate": 7.844958374808642e-07, "loss": 0.0096, "step": 75710 }, { "grad_norm": 0.35668906569480896, "learning_rate": 7.808531632092608e-07, "loss": 0.0131, "step": 75720 }, { "grad_norm": 0.22171305119991302, "learning_rate": 7.772188991436946e-07, "loss": 0.0144, "step": 75730 }, { "grad_norm": 0.21585987508296967, "learning_rate": 7.735930459051799e-07, "loss": 0.0122, "step": 75740 }, { "grad_norm": 0.20028220117092133, "learning_rate": 7.699756041132655e-07, "loss": 0.0152, "step": 75750 }, { "grad_norm": 0.22489818930625916, "learning_rate": 7.663665743860793e-07, "loss": 0.0115, "step": 75760 }, { "grad_norm": 0.3155202567577362, "learning_rate": 7.627659573403001e-07, "loss": 0.0155, "step": 75770 }, { "grad_norm": 0.2333240807056427, "learning_rate": 7.591737535911802e-07, "loss": 0.0143, "step": 75780 }, { "grad_norm": 0.43817999958992004, "learning_rate": 7.555899637525288e-07, "loss": 0.0145, "step": 75790 }, { "grad_norm": 0.27784019708633423, "learning_rate": 7.520145884367058e-07, "loss": 0.0104, "step": 75800 }, { "grad_norm": 0.27853426337242126, "learning_rate": 7.484476282546615e-07, "loss": 0.0167, "step": 75810 }, { "grad_norm": 0.3277638256549835, "learning_rate": 7.448890838158806e-07, "loss": 0.0196, "step": 75820 }, { "grad_norm": 0.26044613122940063, "learning_rate": 7.413389557284267e-07, "loss": 0.0158, "step": 75830 }, { "grad_norm": 0.13679417967796326, "learning_rate": 7.377972445989201e-07, "loss": 0.0168, "step": 75840 }, { "grad_norm": 0.3101333677768707, "learning_rate": 7.342639510325377e-07, "loss": 0.0153, "step": 75850 }, { "grad_norm": 0.2799355089664459, "learning_rate": 7.307390756330246e-07, "loss": 0.011, "step": 75860 }, { "grad_norm": 0.29439064860343933, "learning_rate": 7.272226190026876e-07, "loss": 0.0125, "step": 75870 }, { "grad_norm": 0.3828200697898865, "learning_rate": 7.237145817423907e-07, "loss": 0.0125, "step": 75880 }, { "grad_norm": 0.11996679753065109, "learning_rate": 7.202149644515654e-07, "loss": 0.0139, "step": 75890 }, { "grad_norm": 0.19991964101791382, "learning_rate": 7.167237677281946e-07, "loss": 0.0161, "step": 75900 }, { "grad_norm": 0.3032054603099823, "learning_rate": 7.132409921688288e-07, "loss": 0.0132, "step": 75910 }, { "grad_norm": 0.14106933772563934, "learning_rate": 7.09766638368592e-07, "loss": 0.0102, "step": 75920 }, { "grad_norm": 0.33446624875068665, "learning_rate": 7.063007069211313e-07, "loss": 0.0148, "step": 75930 }, { "grad_norm": 0.2566397786140442, "learning_rate": 7.028431984187067e-07, "loss": 0.0115, "step": 75940 }, { "grad_norm": 0.3054681718349457, "learning_rate": 6.993941134520898e-07, "loss": 0.0135, "step": 75950 }, { "grad_norm": 0.3077864646911621, "learning_rate": 6.959534526106537e-07, "loss": 0.0127, "step": 75960 }, { "grad_norm": 0.43154484033584595, "learning_rate": 6.925212164822948e-07, "loss": 0.0114, "step": 75970 }, { "grad_norm": 0.3508009910583496, "learning_rate": 6.890974056535049e-07, "loss": 0.0207, "step": 75980 }, { "grad_norm": 0.9656876921653748, "learning_rate": 6.856820207093106e-07, "loss": 0.0105, "step": 75990 }, { "grad_norm": 0.1967490315437317, "learning_rate": 6.822750622333063e-07, "loss": 0.0122, "step": 76000 }, { "grad_norm": 0.6327170729637146, "learning_rate": 6.788765308076539e-07, "loss": 0.0156, "step": 76010 }, { "grad_norm": 0.3140597641468048, "learning_rate": 6.754864270130668e-07, "loss": 0.0113, "step": 76020 }, { "grad_norm": 0.23029929399490356, "learning_rate": 6.721047514288203e-07, "loss": 0.0105, "step": 76030 }, { "grad_norm": 0.40777143836021423, "learning_rate": 6.687315046327469e-07, "loss": 0.0161, "step": 76040 }, { "grad_norm": 0.27751126885414124, "learning_rate": 6.65366687201252e-07, "loss": 0.0162, "step": 76050 }, { "grad_norm": 0.36794111132621765, "learning_rate": 6.620102997092814e-07, "loss": 0.0186, "step": 76060 }, { "grad_norm": 0.21818730235099792, "learning_rate": 6.586623427303596e-07, "loss": 0.018, "step": 76070 }, { "grad_norm": 0.21306981146335602, "learning_rate": 6.553228168365455e-07, "loss": 0.0204, "step": 76080 }, { "grad_norm": 0.49026262760162354, "learning_rate": 6.519917225984884e-07, "loss": 0.0163, "step": 76090 }, { "grad_norm": 0.2918927073478699, "learning_rate": 6.486690605853718e-07, "loss": 0.0146, "step": 76100 }, { "grad_norm": 0.28787660598754883, "learning_rate": 6.453548313649527e-07, "loss": 0.015, "step": 76110 }, { "grad_norm": 0.35281258821487427, "learning_rate": 6.420490355035446e-07, "loss": 0.0165, "step": 76120 }, { "grad_norm": 0.3179427981376648, "learning_rate": 6.387516735660071e-07, "loss": 0.0169, "step": 76130 }, { "grad_norm": 0.2352532297372818, "learning_rate": 6.35462746115778e-07, "loss": 0.0177, "step": 76140 }, { "grad_norm": 0.1999804824590683, "learning_rate": 6.321822537148358e-07, "loss": 0.0138, "step": 76150 }, { "grad_norm": 0.1662026345729828, "learning_rate": 6.289101969237432e-07, "loss": 0.0112, "step": 76160 }, { "grad_norm": 0.2535083293914795, "learning_rate": 6.256465763015918e-07, "loss": 0.0124, "step": 76170 }, { "grad_norm": 0.24163781106472015, "learning_rate": 6.22391392406052e-07, "loss": 0.0168, "step": 76180 }, { "grad_norm": 0.21071751415729523, "learning_rate": 6.191446457933403e-07, "loss": 0.0118, "step": 76190 }, { "grad_norm": 0.1275974065065384, "learning_rate": 6.159063370182406e-07, "loss": 0.0108, "step": 76200 }, { "grad_norm": 0.14945484697818756, "learning_rate": 6.126764666340879e-07, "loss": 0.0106, "step": 76210 }, { "grad_norm": 0.19247815012931824, "learning_rate": 6.094550351927852e-07, "loss": 0.0145, "step": 76220 }, { "grad_norm": 0.3185141682624817, "learning_rate": 6.062420432447757e-07, "loss": 0.0144, "step": 76230 }, { "grad_norm": 0.28344374895095825, "learning_rate": 6.030374913390813e-07, "loss": 0.0133, "step": 76240 }, { "grad_norm": 0.32081177830696106, "learning_rate": 5.998413800232694e-07, "loss": 0.0143, "step": 76250 }, { "grad_norm": 0.3985006511211395, "learning_rate": 5.966537098434755e-07, "loss": 0.0145, "step": 76260 }, { "grad_norm": 0.1934584677219391, "learning_rate": 5.934744813443694e-07, "loss": 0.0114, "step": 76270 }, { "grad_norm": 0.3281009793281555, "learning_rate": 5.903036950692054e-07, "loss": 0.0142, "step": 76280 }, { "grad_norm": 0.27497485280036926, "learning_rate": 5.871413515597835e-07, "loss": 0.0156, "step": 76290 }, { "grad_norm": 0.2705085277557373, "learning_rate": 5.839874513564547e-07, "loss": 0.0112, "step": 76300 }, { "grad_norm": 0.32641875743865967, "learning_rate": 5.808419949981436e-07, "loss": 0.0201, "step": 76310 }, { "grad_norm": 0.6881817579269409, "learning_rate": 5.777049830223257e-07, "loss": 0.0202, "step": 76320 }, { "grad_norm": 0.40484780073165894, "learning_rate": 5.745764159650114e-07, "loss": 0.0126, "step": 76330 }, { "grad_norm": 0.213429257273674, "learning_rate": 5.71456294360806e-07, "loss": 0.0125, "step": 76340 }, { "grad_norm": 0.3665253221988678, "learning_rate": 5.683446187428443e-07, "loss": 0.0148, "step": 76350 }, { "grad_norm": 0.27581217885017395, "learning_rate": 5.652413896428288e-07, "loss": 0.0163, "step": 76360 }, { "grad_norm": 0.2223767340183258, "learning_rate": 5.621466075910131e-07, "loss": 0.0135, "step": 76370 }, { "grad_norm": 0.11575795710086823, "learning_rate": 5.590602731162187e-07, "loss": 0.012, "step": 76380 }, { "grad_norm": 0.22731833159923553, "learning_rate": 5.559823867458125e-07, "loss": 0.016, "step": 76390 }, { "grad_norm": 0.31929919123649597, "learning_rate": 5.529129490057128e-07, "loss": 0.0146, "step": 76400 }, { "grad_norm": 0.32152754068374634, "learning_rate": 5.498519604204167e-07, "loss": 0.0182, "step": 76410 }, { "grad_norm": 0.18798643350601196, "learning_rate": 5.467994215129501e-07, "loss": 0.0151, "step": 76420 }, { "grad_norm": 0.3031480610370636, "learning_rate": 5.437553328049183e-07, "loss": 0.015, "step": 76430 }, { "grad_norm": 0.25644534826278687, "learning_rate": 5.407196948164661e-07, "loss": 0.0147, "step": 76440 }, { "grad_norm": 0.20033317804336548, "learning_rate": 5.376925080663009e-07, "loss": 0.0142, "step": 76450 }, { "grad_norm": 0.28981292247772217, "learning_rate": 5.346737730716977e-07, "loss": 0.0133, "step": 76460 }, { "grad_norm": 0.14094871282577515, "learning_rate": 5.316634903484607e-07, "loss": 0.0105, "step": 76470 }, { "grad_norm": 0.16150884330272675, "learning_rate": 5.286616604109728e-07, "loss": 0.0162, "step": 76480 }, { "grad_norm": 0.2572460174560547, "learning_rate": 5.256682837721627e-07, "loss": 0.023, "step": 76490 }, { "grad_norm": 0.39808234572410583, "learning_rate": 5.226833609435156e-07, "loss": 0.0145, "step": 76500 }, { "grad_norm": 0.2241683304309845, "learning_rate": 5.19706892435079e-07, "loss": 0.0119, "step": 76510 }, { "grad_norm": 0.28494516015052795, "learning_rate": 5.167388787554406e-07, "loss": 0.013, "step": 76520 }, { "grad_norm": 0.25258302688598633, "learning_rate": 5.137793204117614e-07, "loss": 0.0171, "step": 76530 }, { "grad_norm": 0.32477840781211853, "learning_rate": 5.108282179097479e-07, "loss": 0.0142, "step": 76540 }, { "grad_norm": 0.31809377670288086, "learning_rate": 5.078855717536523e-07, "loss": 0.0147, "step": 76550 }, { "grad_norm": 0.22208625078201294, "learning_rate": 5.049513824463059e-07, "loss": 0.0115, "step": 76560 }, { "grad_norm": 0.14870457351207733, "learning_rate": 5.020256504890742e-07, "loss": 0.0144, "step": 76570 }, { "grad_norm": 0.1982165277004242, "learning_rate": 4.991083763818849e-07, "loss": 0.0163, "step": 76580 }, { "grad_norm": 0.23899810016155243, "learning_rate": 4.961995606232228e-07, "loss": 0.0181, "step": 76590 }, { "grad_norm": 0.2758757174015045, "learning_rate": 4.932992037101236e-07, "loss": 0.0125, "step": 76600 }, { "grad_norm": 0.2821520268917084, "learning_rate": 4.904073061381798e-07, "loss": 0.0151, "step": 76610 }, { "grad_norm": 0.4070322811603546, "learning_rate": 4.875238684015349e-07, "loss": 0.0196, "step": 76620 }, { "grad_norm": 0.22934944927692413, "learning_rate": 4.846488909928948e-07, "loss": 0.0115, "step": 76630 }, { "grad_norm": 0.16186770796775818, "learning_rate": 4.81782374403511e-07, "loss": 0.0147, "step": 76640 }, { "grad_norm": 0.2336055338382721, "learning_rate": 4.789243191231918e-07, "loss": 0.0124, "step": 76650 }, { "grad_norm": 0.20011070370674133, "learning_rate": 4.76074725640302e-07, "loss": 0.013, "step": 76660 }, { "grad_norm": 0.19748128950595856, "learning_rate": 4.732335944417632e-07, "loss": 0.0221, "step": 76670 }, { "grad_norm": 0.19995160400867462, "learning_rate": 4.704009260130371e-07, "loss": 0.0175, "step": 76680 }, { "grad_norm": 0.1707441359758377, "learning_rate": 4.675767208381587e-07, "loss": 0.0139, "step": 76690 }, { "grad_norm": 0.29661205410957336, "learning_rate": 4.647609793997032e-07, "loss": 0.0152, "step": 76700 }, { "grad_norm": 0.286032110452652, "learning_rate": 4.619537021788023e-07, "loss": 0.0118, "step": 76710 }, { "grad_norm": 0.3379957675933838, "learning_rate": 4.5915488965515005e-07, "loss": 0.0171, "step": 76720 }, { "grad_norm": 0.369708389043808, "learning_rate": 4.563645423069807e-07, "loss": 0.0189, "step": 76730 }, { "grad_norm": 0.332925409078598, "learning_rate": 4.535826606110849e-07, "loss": 0.0146, "step": 76740 }, { "grad_norm": 0.3350381851196289, "learning_rate": 4.508092450428214e-07, "loss": 0.0201, "step": 76750 }, { "grad_norm": 0.2783515453338623, "learning_rate": 4.480442960760778e-07, "loss": 0.0138, "step": 76760 }, { "grad_norm": 0.1755526065826416, "learning_rate": 4.4528781418332053e-07, "loss": 0.0139, "step": 76770 }, { "grad_norm": 0.34096136689186096, "learning_rate": 4.4253979983555073e-07, "loss": 0.02, "step": 76780 }, { "grad_norm": 0.19527988135814667, "learning_rate": 4.3980025350233154e-07, "loss": 0.0144, "step": 76790 }, { "grad_norm": 0.20235992968082428, "learning_rate": 4.370691756517664e-07, "loss": 0.0125, "step": 76800 }, { "grad_norm": 0.20346197485923767, "learning_rate": 4.3434656675053753e-07, "loss": 0.0097, "step": 76810 }, { "grad_norm": 0.2676005959510803, "learning_rate": 4.3163242726385613e-07, "loss": 0.0125, "step": 76820 }, { "grad_norm": 0.3507152199745178, "learning_rate": 4.289267576554956e-07, "loss": 0.0141, "step": 76830 }, { "grad_norm": 0.20074047148227692, "learning_rate": 4.262295583877807e-07, "loss": 0.0141, "step": 76840 }, { "grad_norm": 0.2655695378780365, "learning_rate": 4.235408299215815e-07, "loss": 0.0171, "step": 76850 }, { "grad_norm": 0.19189883768558502, "learning_rate": 4.2086057271634173e-07, "loss": 0.013, "step": 76860 }, { "grad_norm": 0.4152809679508209, "learning_rate": 4.181887872300394e-07, "loss": 0.013, "step": 76870 }, { "grad_norm": 0.35934239625930786, "learning_rate": 4.155254739192038e-07, "loss": 0.0166, "step": 76880 }, { "grad_norm": 0.21303389966487885, "learning_rate": 4.128706332389265e-07, "loss": 0.0195, "step": 76890 }, { "grad_norm": 0.2860944867134094, "learning_rate": 4.1022426564284453e-07, "loss": 0.014, "step": 76900 }, { "grad_norm": 0.1453598439693451, "learning_rate": 4.075863715831574e-07, "loss": 0.0131, "step": 76910 }, { "grad_norm": 0.2888581156730652, "learning_rate": 4.0495695151059887e-07, "loss": 0.0164, "step": 76920 }, { "grad_norm": 0.4119842052459717, "learning_rate": 4.023360058744763e-07, "loss": 0.0108, "step": 76930 }, { "grad_norm": 0.09073875844478607, "learning_rate": 3.997235351226314e-07, "loss": 0.011, "step": 76940 }, { "grad_norm": 0.506533145904541, "learning_rate": 3.971195397014571e-07, "loss": 0.0179, "step": 76950 }, { "grad_norm": 0.2953524589538574, "learning_rate": 3.9452402005591417e-07, "loss": 0.0115, "step": 76960 }, { "grad_norm": 0.16191531717777252, "learning_rate": 3.9193697662950334e-07, "loss": 0.0112, "step": 76970 }, { "grad_norm": 0.3582862317562103, "learning_rate": 3.893584098642822e-07, "loss": 0.0146, "step": 76980 }, { "grad_norm": 0.28156834840774536, "learning_rate": 3.867883202008538e-07, "loss": 0.0163, "step": 76990 }, { "grad_norm": 0.3252975344657898, "learning_rate": 3.842267080783779e-07, "loss": 0.0132, "step": 77000 }, { "grad_norm": 0.22672417759895325, "learning_rate": 3.816735739345656e-07, "loss": 0.0144, "step": 77010 }, { "grad_norm": 0.22196172177791595, "learning_rate": 3.791289182056679e-07, "loss": 0.0182, "step": 77020 }, { "grad_norm": 0.181911900639534, "learning_rate": 3.7659274132650913e-07, "loss": 0.0176, "step": 77030 }, { "grad_norm": 0.23536299169063568, "learning_rate": 3.7406504373044826e-07, "loss": 0.0131, "step": 77040 }, { "grad_norm": 0.3045678734779358, "learning_rate": 3.7154582584939533e-07, "loss": 0.0147, "step": 77050 }, { "grad_norm": 0.13934987783432007, "learning_rate": 3.690350881138227e-07, "loss": 0.0149, "step": 77060 }, { "grad_norm": 0.19049367308616638, "learning_rate": 3.665328309527427e-07, "loss": 0.014, "step": 77070 }, { "grad_norm": 0.2574869990348816, "learning_rate": 3.640390547937245e-07, "loss": 0.0136, "step": 77080 }, { "grad_norm": 0.14388896524906158, "learning_rate": 3.615537600628882e-07, "loss": 0.0158, "step": 77090 }, { "grad_norm": 0.19768160581588745, "learning_rate": 3.5907694718489427e-07, "loss": 0.0161, "step": 77100 }, { "grad_norm": 0.26185745000839233, "learning_rate": 3.566086165829707e-07, "loss": 0.0188, "step": 77110 }, { "grad_norm": 0.40806668996810913, "learning_rate": 3.5414876867888024e-07, "loss": 0.0181, "step": 77120 }, { "grad_norm": 0.17116335034370422, "learning_rate": 3.5169740389295326e-07, "loss": 0.0181, "step": 77130 }, { "grad_norm": 0.25140005350112915, "learning_rate": 3.4925452264405466e-07, "loss": 0.0149, "step": 77140 }, { "grad_norm": 0.16248339414596558, "learning_rate": 3.468201253496062e-07, "loss": 0.0117, "step": 77150 }, { "grad_norm": 0.2700302004814148, "learning_rate": 3.443942124255861e-07, "loss": 0.0132, "step": 77160 }, { "grad_norm": 0.2660820782184601, "learning_rate": 3.4197678428650183e-07, "loss": 0.0169, "step": 77170 }, { "grad_norm": 0.16181246936321259, "learning_rate": 3.3956784134544504e-07, "loss": 0.014, "step": 77180 }, { "grad_norm": 0.3075138032436371, "learning_rate": 3.3716738401402547e-07, "loss": 0.0164, "step": 77190 }, { "grad_norm": 0.18311834335327148, "learning_rate": 3.3477541270241495e-07, "loss": 0.012, "step": 77200 }, { "grad_norm": 0.23125328123569489, "learning_rate": 3.3239192781934215e-07, "loss": 0.0128, "step": 77210 }, { "grad_norm": 0.2075543999671936, "learning_rate": 3.3001692977207563e-07, "loss": 0.013, "step": 77220 }, { "grad_norm": 0.1464175432920456, "learning_rate": 3.276504189664409e-07, "loss": 0.0128, "step": 77230 }, { "grad_norm": 0.31207460165023804, "learning_rate": 3.252923958068088e-07, "loss": 0.0175, "step": 77240 }, { "grad_norm": 0.08990037441253662, "learning_rate": 3.2294286069609046e-07, "loss": 0.015, "step": 77250 }, { "grad_norm": 0.14673897624015808, "learning_rate": 3.2060181403577583e-07, "loss": 0.0123, "step": 77260 }, { "grad_norm": 0.2744594216346741, "learning_rate": 3.1826925622587823e-07, "loss": 0.0125, "step": 77270 }, { "grad_norm": 0.44422647356987, "learning_rate": 3.159451876649622e-07, "loss": 0.0152, "step": 77280 }, { "grad_norm": 0.2396896481513977, "learning_rate": 3.1362960875015444e-07, "loss": 0.0145, "step": 77290 }, { "grad_norm": 0.4054839313030243, "learning_rate": 3.1132251987711637e-07, "loss": 0.0186, "step": 77300 }, { "grad_norm": 0.261940062046051, "learning_rate": 3.0902392144007695e-07, "loss": 0.017, "step": 77310 }, { "grad_norm": 0.21353910863399506, "learning_rate": 3.0673381383179436e-07, "loss": 0.0152, "step": 77320 }, { "grad_norm": 0.2309817373752594, "learning_rate": 3.0445219744358875e-07, "loss": 0.0117, "step": 77330 }, { "grad_norm": 0.20096227526664734, "learning_rate": 3.021790726653262e-07, "loss": 0.0137, "step": 77340 }, { "grad_norm": 0.45328059792518616, "learning_rate": 2.9991443988542366e-07, "loss": 0.0133, "step": 77350 }, { "grad_norm": 0.17470601201057434, "learning_rate": 2.976582994908439e-07, "loss": 0.0128, "step": 77360 }, { "grad_norm": 0.2190714031457901, "learning_rate": 2.954106518671007e-07, "loss": 0.0131, "step": 77370 }, { "grad_norm": 0.2239532321691513, "learning_rate": 2.9317149739825356e-07, "loss": 0.0122, "step": 77380 }, { "grad_norm": 0.37549272179603577, "learning_rate": 2.909408364669075e-07, "loss": 0.0149, "step": 77390 }, { "grad_norm": 0.32106366753578186, "learning_rate": 2.887186694542299e-07, "loss": 0.0202, "step": 77400 }, { "grad_norm": 0.2711494565010071, "learning_rate": 2.865049967399225e-07, "loss": 0.0147, "step": 77410 }, { "grad_norm": 0.3108512759208679, "learning_rate": 2.842998187022439e-07, "loss": 0.0117, "step": 77420 }, { "grad_norm": 0.2945519685745239, "learning_rate": 2.8210313571800374e-07, "loss": 0.0158, "step": 77430 }, { "grad_norm": 0.35927659273147583, "learning_rate": 2.7991494816255184e-07, "loss": 0.0117, "step": 77440 }, { "grad_norm": 0.23974932730197906, "learning_rate": 2.777352564097779e-07, "loss": 0.0181, "step": 77450 }, { "grad_norm": 0.3062439560890198, "learning_rate": 2.755640608321508e-07, "loss": 0.0133, "step": 77460 }, { "grad_norm": 0.3932819068431854, "learning_rate": 2.73401361800657e-07, "loss": 0.0149, "step": 77470 }, { "grad_norm": 0.1829487383365631, "learning_rate": 2.7124715968484537e-07, "loss": 0.0141, "step": 77480 }, { "grad_norm": 0.23206296563148499, "learning_rate": 2.691014548528104e-07, "loss": 0.0164, "step": 77490 }, { "grad_norm": 0.20366598665714264, "learning_rate": 2.669642476711864e-07, "loss": 0.0127, "step": 77500 }, { "grad_norm": 0.159492626786232, "learning_rate": 2.648355385051815e-07, "loss": 0.0119, "step": 77510 }, { "grad_norm": 0.19848918914794922, "learning_rate": 2.627153277185157e-07, "loss": 0.0117, "step": 77520 }, { "grad_norm": 0.23442095518112183, "learning_rate": 2.606036156734881e-07, "loss": 0.0138, "step": 77530 }, { "grad_norm": 0.1540299952030182, "learning_rate": 2.5850040273092127e-07, "loss": 0.015, "step": 77540 }, { "grad_norm": 0.21528658270835876, "learning_rate": 2.5640568925020536e-07, "loss": 0.0211, "step": 77550 }, { "grad_norm": 0.18051175773143768, "learning_rate": 2.5431947558927083e-07, "loss": 0.0149, "step": 77560 }, { "grad_norm": 0.2646220326423645, "learning_rate": 2.522417621045825e-07, "loss": 0.0125, "step": 77570 }, { "grad_norm": 0.2314745932817459, "learning_rate": 2.501725491511786e-07, "loss": 0.0176, "step": 77580 }, { "grad_norm": 0.20790208876132965, "learning_rate": 2.481118370826263e-07, "loss": 0.0178, "step": 77590 }, { "grad_norm": 0.2793470025062561, "learning_rate": 2.4605962625104393e-07, "loss": 0.0128, "step": 77600 }, { "grad_norm": 0.2784343659877777, "learning_rate": 2.440159170070955e-07, "loss": 0.0201, "step": 77610 }, { "grad_norm": 0.22315239906311035, "learning_rate": 2.41980709699996e-07, "loss": 0.0147, "step": 77620 }, { "grad_norm": 0.4989861249923706, "learning_rate": 2.3995400467751174e-07, "loss": 0.0123, "step": 77630 }, { "grad_norm": 0.2546156346797943, "learning_rate": 2.3793580228594902e-07, "loss": 0.0229, "step": 77640 }, { "grad_norm": 0.24151477217674255, "learning_rate": 2.3592610287015982e-07, "loss": 0.0104, "step": 77650 }, { "grad_norm": 0.1948913186788559, "learning_rate": 2.3392490677354718e-07, "loss": 0.0167, "step": 77660 }, { "grad_norm": 0.5846874117851257, "learning_rate": 2.3193221433806533e-07, "loss": 0.0101, "step": 77670 }, { "grad_norm": 0.47944194078445435, "learning_rate": 2.299480259042086e-07, "loss": 0.016, "step": 77680 }, { "grad_norm": 0.1311970353126526, "learning_rate": 2.2797234181102244e-07, "loss": 0.0138, "step": 77690 }, { "grad_norm": 0.300400972366333, "learning_rate": 2.2600516239609238e-07, "loss": 0.0155, "step": 77700 }, { "grad_norm": 0.5483819842338562, "learning_rate": 2.240464879955606e-07, "loss": 0.0174, "step": 77710 }, { "grad_norm": 0.24476128816604614, "learning_rate": 2.2209631894410387e-07, "loss": 0.0152, "step": 77720 }, { "grad_norm": 0.2639368176460266, "learning_rate": 2.2015465557496117e-07, "loss": 0.0114, "step": 77730 }, { "grad_norm": 0.2792101204395294, "learning_rate": 2.1822149821990602e-07, "loss": 0.0131, "step": 77740 }, { "grad_norm": 0.4730912148952484, "learning_rate": 2.1629684720926303e-07, "loss": 0.0157, "step": 77750 }, { "grad_norm": 0.23075537383556366, "learning_rate": 2.1438070287189693e-07, "loss": 0.0196, "step": 77760 }, { "grad_norm": 0.20264509320259094, "learning_rate": 2.1247306553523472e-07, "loss": 0.0153, "step": 77770 }, { "grad_norm": 0.36927783489227295, "learning_rate": 2.105739355252323e-07, "loss": 0.0131, "step": 77780 }, { "grad_norm": 0.1842103749513626, "learning_rate": 2.0868331316639678e-07, "loss": 0.0151, "step": 77790 }, { "grad_norm": 0.19742093980312347, "learning_rate": 2.0680119878179193e-07, "loss": 0.0132, "step": 77800 }, { "grad_norm": 0.19481199979782104, "learning_rate": 2.0492759269301054e-07, "loss": 0.0162, "step": 77810 }, { "grad_norm": 0.4303101897239685, "learning_rate": 2.0306249522021315e-07, "loss": 0.017, "step": 77820 }, { "grad_norm": 0.3984906077384949, "learning_rate": 2.0120590668207816e-07, "loss": 0.0138, "step": 77830 }, { "grad_norm": 0.2441922426223755, "learning_rate": 1.993578273958574e-07, "loss": 0.0098, "step": 77840 }, { "grad_norm": 0.19442124664783478, "learning_rate": 1.975182576773371e-07, "loss": 0.0125, "step": 77850 }, { "grad_norm": 0.4675902724266052, "learning_rate": 1.9568719784083812e-07, "loss": 0.0186, "step": 77860 }, { "grad_norm": 0.1663823276758194, "learning_rate": 1.9386464819924898e-07, "loss": 0.0174, "step": 77870 }, { "grad_norm": 0.18750852346420288, "learning_rate": 1.9205060906399285e-07, "loss": 0.0144, "step": 77880 }, { "grad_norm": 0.3401205241680145, "learning_rate": 1.9024508074503845e-07, "loss": 0.0126, "step": 77890 }, { "grad_norm": 0.29752352833747864, "learning_rate": 1.8844806355089452e-07, "loss": 0.0218, "step": 77900 }, { "grad_norm": 0.20316152274608612, "learning_rate": 1.8665955778863208e-07, "loss": 0.0247, "step": 77910 }, { "grad_norm": 1.2882012128829956, "learning_rate": 1.848795637638512e-07, "loss": 0.0178, "step": 77920 }, { "grad_norm": 0.3865475058555603, "learning_rate": 1.831080817807085e-07, "loss": 0.0175, "step": 77930 }, { "grad_norm": 0.3853815197944641, "learning_rate": 1.813451121418952e-07, "loss": 0.0248, "step": 77940 }, { "grad_norm": 0.18530839681625366, "learning_rate": 1.795906551486648e-07, "loss": 0.01, "step": 77950 }, { "grad_norm": 0.2538071870803833, "learning_rate": 1.7784471110079414e-07, "loss": 0.0116, "step": 77960 }, { "grad_norm": 0.30571651458740234, "learning_rate": 1.7610728029662793e-07, "loss": 0.0143, "step": 77970 }, { "grad_norm": 0.1813371479511261, "learning_rate": 1.7437836303303979e-07, "loss": 0.0149, "step": 77980 }, { "grad_norm": 0.28224411606788635, "learning_rate": 1.726579596054545e-07, "loss": 0.0136, "step": 77990 }, { "grad_norm": 0.27590957283973694, "learning_rate": 1.7094607030784803e-07, "loss": 0.0161, "step": 78000 }, { "grad_norm": 0.26526257395744324, "learning_rate": 1.6924269543272532e-07, "loss": 0.0143, "step": 78010 }, { "grad_norm": 0.216426819562912, "learning_rate": 1.6754783527115348e-07, "loss": 0.0126, "step": 78020 }, { "grad_norm": 0.35232973098754883, "learning_rate": 1.6586149011273422e-07, "loss": 0.0129, "step": 78030 }, { "grad_norm": 0.27259305119514465, "learning_rate": 1.6418366024562038e-07, "loss": 0.0218, "step": 78040 }, { "grad_norm": 0.34674668312072754, "learning_rate": 1.6251434595651037e-07, "loss": 0.0151, "step": 78050 }, { "grad_norm": 0.2374202460050583, "learning_rate": 1.608535475306372e-07, "loss": 0.0132, "step": 78060 }, { "grad_norm": 0.3238479793071747, "learning_rate": 1.5920126525179048e-07, "loss": 0.0129, "step": 78070 }, { "grad_norm": 0.19310234487056732, "learning_rate": 1.5755749940229435e-07, "loss": 0.0108, "step": 78080 }, { "grad_norm": 0.23609985411167145, "learning_rate": 1.559222502630353e-07, "loss": 0.0134, "step": 78090 }, { "grad_norm": 0.3138655424118042, "learning_rate": 1.5429551811341757e-07, "loss": 0.0158, "step": 78100 }, { "grad_norm": 0.21629536151885986, "learning_rate": 1.5267730323141882e-07, "loss": 0.0165, "step": 78110 }, { "grad_norm": 0.1441717892885208, "learning_rate": 1.5106760589353454e-07, "loss": 0.0116, "step": 78120 }, { "grad_norm": 0.21196207404136658, "learning_rate": 1.4946642637483355e-07, "loss": 0.0112, "step": 78130 }, { "grad_norm": 0.726184070110321, "learning_rate": 1.4787376494889703e-07, "loss": 0.0145, "step": 78140 }, { "grad_norm": 0.31421542167663574, "learning_rate": 1.4628962188787955e-07, "loss": 0.0132, "step": 78150 }, { "grad_norm": 0.2632908225059509, "learning_rate": 1.4471399746247006e-07, "loss": 0.0162, "step": 78160 }, { "grad_norm": 0.15854986011981964, "learning_rate": 1.4314689194188103e-07, "loss": 0.0156, "step": 78170 }, { "grad_norm": 0.45505404472351074, "learning_rate": 1.4158830559390933e-07, "loss": 0.0177, "step": 78180 }, { "grad_norm": 0.2432737797498703, "learning_rate": 1.4003823868486422e-07, "loss": 0.0295, "step": 78190 }, { "grad_norm": 0.2723196744918823, "learning_rate": 1.3849669147960598e-07, "loss": 0.0168, "step": 78200 }, { "grad_norm": 0.14861243963241577, "learning_rate": 1.3696366424155726e-07, "loss": 0.013, "step": 78210 }, { "grad_norm": 0.6030508875846863, "learning_rate": 1.35439157232653e-07, "loss": 0.0224, "step": 78220 }, { "grad_norm": 0.5109241008758545, "learning_rate": 1.3392317071340144e-07, "loss": 0.0147, "step": 78230 }, { "grad_norm": 0.4637286961078644, "learning_rate": 1.3241570494283984e-07, "loss": 0.0135, "step": 78240 }, { "grad_norm": 0.4758991003036499, "learning_rate": 1.3091676017855103e-07, "loss": 0.0188, "step": 78250 }, { "grad_norm": 0.22832833230495453, "learning_rate": 1.2942633667666904e-07, "loss": 0.0135, "step": 78260 }, { "grad_norm": 0.27507710456848145, "learning_rate": 1.2794443469185679e-07, "loss": 0.0175, "step": 78270 }, { "grad_norm": 0.2543404996395111, "learning_rate": 1.2647105447734508e-07, "loss": 0.0122, "step": 78280 }, { "grad_norm": 0.4546054005622864, "learning_rate": 1.2500619628488254e-07, "loss": 0.0131, "step": 78290 }, { "grad_norm": 1.166865348815918, "learning_rate": 1.2354986036477446e-07, "loss": 0.0194, "step": 78300 }, { "grad_norm": 0.2735418677330017, "learning_rate": 1.221020469658718e-07, "loss": 0.0129, "step": 78310 }, { "grad_norm": 0.2757149338722229, "learning_rate": 1.2066275633556556e-07, "loss": 0.0159, "step": 78320 }, { "grad_norm": 0.2304448038339615, "learning_rate": 1.192319887197979e-07, "loss": 0.0138, "step": 78330 }, { "grad_norm": 0.22024759650230408, "learning_rate": 1.1780974436303438e-07, "loss": 0.0122, "step": 78340 }, { "grad_norm": 0.1930255889892578, "learning_rate": 1.1639602350830836e-07, "loss": 0.0163, "step": 78350 }, { "grad_norm": 0.22049757838249207, "learning_rate": 1.1499082639718217e-07, "loss": 0.0133, "step": 78360 }, { "grad_norm": 0.33840999007225037, "learning_rate": 1.1359415326976374e-07, "loss": 0.0136, "step": 78370 }, { "grad_norm": 0.13610468804836273, "learning_rate": 1.1220600436470663e-07, "loss": 0.0172, "step": 78380 }, { "grad_norm": 0.2175500988960266, "learning_rate": 1.1082637991920996e-07, "loss": 0.0142, "step": 78390 }, { "grad_norm": 0.22534309327602386, "learning_rate": 1.0945528016901851e-07, "loss": 0.0148, "step": 78400 }, { "grad_norm": 0.2679053843021393, "learning_rate": 1.0809270534840599e-07, "loss": 0.0135, "step": 78410 }, { "grad_norm": 0.2052120715379715, "learning_rate": 1.067386556902028e-07, "loss": 0.0118, "step": 78420 }, { "grad_norm": 0.21356995403766632, "learning_rate": 1.05393131425785e-07, "loss": 0.0119, "step": 78430 }, { "grad_norm": 0.296938955783844, "learning_rate": 1.0405613278505199e-07, "loss": 0.0155, "step": 78440 }, { "grad_norm": 0.3726133704185486, "learning_rate": 1.027276599964766e-07, "loss": 0.0144, "step": 78450 }, { "grad_norm": 0.20011021196842194, "learning_rate": 1.0140771328704391e-07, "loss": 0.0121, "step": 78460 }, { "grad_norm": 0.15050385892391205, "learning_rate": 1.0009629288231237e-07, "loss": 0.0129, "step": 78470 }, { "grad_norm": 0.18247640132904053, "learning_rate": 9.879339900635276e-08, "loss": 0.0156, "step": 78480 }, { "grad_norm": 0.5817012190818787, "learning_rate": 9.74990318817981e-08, "loss": 0.0148, "step": 78490 }, { "grad_norm": 0.3784177005290985, "learning_rate": 9.621319172982701e-08, "loss": 0.0116, "step": 78500 }, { "grad_norm": 0.13344237208366394, "learning_rate": 9.493587877015264e-08, "loss": 0.0132, "step": 78510 }, { "grad_norm": 0.19839772582054138, "learning_rate": 9.366709322102263e-08, "loss": 0.0097, "step": 78520 }, { "grad_norm": 0.3384036123752594, "learning_rate": 9.240683529924688e-08, "loss": 0.0123, "step": 78530 }, { "grad_norm": 0.23208290338516235, "learning_rate": 9.115510522016979e-08, "loss": 0.0107, "step": 78540 }, { "grad_norm": 0.22985512018203735, "learning_rate": 8.991190319767583e-08, "loss": 0.0129, "step": 78550 }, { "grad_norm": 0.1540190875530243, "learning_rate": 8.867722944419509e-08, "loss": 0.0133, "step": 78560 }, { "grad_norm": 0.2757311165332794, "learning_rate": 8.745108417069214e-08, "loss": 0.0094, "step": 78570 }, { "grad_norm": 0.1254572868347168, "learning_rate": 8.623346758669381e-08, "loss": 0.012, "step": 78580 }, { "grad_norm": 0.14928682148456573, "learning_rate": 8.502437990025037e-08, "loss": 0.0125, "step": 78590 }, { "grad_norm": 0.16927215456962585, "learning_rate": 8.382382131795764e-08, "loss": 0.013, "step": 78600 }, { "grad_norm": 0.12212350219488144, "learning_rate": 8.263179204496818e-08, "loss": 0.0205, "step": 78610 }, { "grad_norm": 0.3270401060581207, "learning_rate": 8.144829228496354e-08, "loss": 0.0164, "step": 78620 }, { "grad_norm": 0.28699547052383423, "learning_rate": 8.027332224016526e-08, "loss": 0.0154, "step": 78630 }, { "grad_norm": 0.25864657759666443, "learning_rate": 7.910688211135164e-08, "loss": 0.0155, "step": 78640 }, { "grad_norm": 0.18404939770698547, "learning_rate": 7.794897209783546e-08, "loss": 0.0195, "step": 78650 }, { "grad_norm": 0.19161495566368103, "learning_rate": 7.679959239746403e-08, "loss": 0.0161, "step": 78660 }, { "grad_norm": 0.1314968466758728, "learning_rate": 7.565874320664135e-08, "loss": 0.015, "step": 78670 }, { "grad_norm": 0.2490159422159195, "learning_rate": 7.45264247203059e-08, "loss": 0.013, "step": 78680 }, { "grad_norm": 0.46533504128456116, "learning_rate": 7.340263713194184e-08, "loss": 0.0116, "step": 78690 }, { "grad_norm": 0.4503345191478729, "learning_rate": 7.228738063356777e-08, "loss": 0.0166, "step": 78700 }, { "grad_norm": 0.4962347149848938, "learning_rate": 7.118065541575903e-08, "loss": 0.0189, "step": 78710 }, { "grad_norm": 0.201047882437706, "learning_rate": 7.008246166761435e-08, "loss": 0.0165, "step": 78720 }, { "grad_norm": 0.381550669670105, "learning_rate": 6.899279957679472e-08, "loss": 0.0123, "step": 78730 }, { "grad_norm": 0.29096436500549316, "learning_rate": 6.791166932949011e-08, "loss": 0.0152, "step": 78740 }, { "grad_norm": 0.28991201519966125, "learning_rate": 6.683907111043608e-08, "loss": 0.0121, "step": 78750 }, { "grad_norm": 0.14020423591136932, "learning_rate": 6.577500510290824e-08, "loss": 0.017, "step": 78760 }, { "grad_norm": 0.2819254994392395, "learning_rate": 6.47194714887278e-08, "loss": 0.0247, "step": 78770 }, { "grad_norm": 0.18512016534805298, "learning_rate": 6.367247044825608e-08, "loss": 0.0126, "step": 78780 }, { "grad_norm": 0.2574141025543213, "learning_rate": 6.263400216039994e-08, "loss": 0.0145, "step": 78790 }, { "grad_norm": 0.3679850697517395, "learning_rate": 6.160406680260078e-08, "loss": 0.0255, "step": 78800 }, { "grad_norm": 0.06944174319505692, "learning_rate": 6.058266455084561e-08, "loss": 0.0164, "step": 78810 }, { "grad_norm": 0.1913866251707077, "learning_rate": 5.956979557967257e-08, "loss": 0.0126, "step": 78820 }, { "grad_norm": 0.19845439493656158, "learning_rate": 5.856546006214325e-08, "loss": 0.0144, "step": 78830 }, { "grad_norm": 0.2273171991109848, "learning_rate": 5.7569658169881466e-08, "loss": 0.0097, "step": 78840 }, { "grad_norm": 0.19015997648239136, "learning_rate": 5.658239007303445e-08, "loss": 0.0162, "step": 78850 }, { "grad_norm": 0.282539039850235, "learning_rate": 5.560365594030059e-08, "loss": 0.0127, "step": 78860 }, { "grad_norm": 0.17978766560554504, "learning_rate": 5.463345593891833e-08, "loss": 0.0148, "step": 78870 }, { "grad_norm": 0.15787045657634735, "learning_rate": 5.367179023467173e-08, "loss": 0.0188, "step": 78880 }, { "grad_norm": 0.2447507381439209, "learning_rate": 5.271865899187378e-08, "loss": 0.014, "step": 78890 }, { "grad_norm": 0.34302207827568054, "learning_rate": 5.177406237340532e-08, "loss": 0.0314, "step": 78900 }, { "grad_norm": 0.20033317804336548, "learning_rate": 5.083800054065946e-08, "loss": 0.0093, "step": 78910 }, { "grad_norm": 0.2195563018321991, "learning_rate": 4.991047365358603e-08, "loss": 0.0129, "step": 78920 }, { "grad_norm": 0.6174765229225159, "learning_rate": 4.899148187067493e-08, "loss": 0.0226, "step": 78930 }, { "grad_norm": 0.16654208302497864, "learning_rate": 4.808102534895609e-08, "loss": 0.0124, "step": 78940 }, { "grad_norm": 0.170892596244812, "learning_rate": 4.717910424400507e-08, "loss": 0.0154, "step": 78950 }, { "grad_norm": 0.34133246541023254, "learning_rate": 4.628571870993193e-08, "loss": 0.0159, "step": 78960 }, { "grad_norm": 0.1920427680015564, "learning_rate": 4.5400868899392324e-08, "loss": 0.0172, "step": 78970 }, { "grad_norm": 0.3257465958595276, "learning_rate": 4.4524554963576436e-08, "loss": 0.0186, "step": 78980 }, { "grad_norm": 0.3103378713130951, "learning_rate": 4.36567770522367e-08, "loss": 0.0145, "step": 78990 }, { "grad_norm": 0.2302328646183014, "learning_rate": 4.279753531364339e-08, "loss": 0.0133, "step": 79000 }, { "grad_norm": 0.3491663932800293, "learning_rate": 4.1946829894617955e-08, "loss": 0.0205, "step": 79010 }, { "grad_norm": 0.18437460064888, "learning_rate": 4.11046609405219e-08, "loss": 0.0219, "step": 79020 }, { "grad_norm": 0.37003299593925476, "learning_rate": 4.027102859526233e-08, "loss": 0.0131, "step": 79030 }, { "grad_norm": 0.5207569003105164, "learning_rate": 3.9445933001280856e-08, "loss": 0.0169, "step": 79040 }, { "grad_norm": 0.21439863741397858, "learning_rate": 3.8629374299564704e-08, "loss": 0.0179, "step": 79050 }, { "grad_norm": 0.21313373744487762, "learning_rate": 3.78213526296467e-08, "loss": 0.0179, "step": 79060 }, { "grad_norm": 0.13133755326271057, "learning_rate": 3.702186812958308e-08, "loss": 0.0147, "step": 79070 }, { "grad_norm": 0.22926586866378784, "learning_rate": 3.623092093599789e-08, "loss": 0.0186, "step": 79080 }, { "grad_norm": 0.37078043818473816, "learning_rate": 3.544851118403303e-08, "loss": 0.0154, "step": 79090 }, { "grad_norm": 0.25800982117652893, "learning_rate": 3.46746390073871e-08, "loss": 0.018, "step": 79100 }, { "grad_norm": 0.3297860026359558, "learning_rate": 3.390930453829322e-08, "loss": 0.0186, "step": 79110 }, { "grad_norm": 0.32206082344055176, "learning_rate": 3.3152507907519005e-08, "loss": 0.0131, "step": 79120 }, { "grad_norm": 0.15925948321819305, "learning_rate": 3.240424924438323e-08, "loss": 0.0153, "step": 79130 }, { "grad_norm": 0.2007281333208084, "learning_rate": 3.1664528676750295e-08, "loss": 0.0156, "step": 79140 }, { "grad_norm": 0.21717338263988495, "learning_rate": 3.093334633100797e-08, "loss": 0.0241, "step": 79150 }, { "grad_norm": 0.2716045677661896, "learning_rate": 3.021070233210077e-08, "loss": 0.0159, "step": 79160 }, { "grad_norm": 0.31255093216896057, "learning_rate": 2.9496596803507692e-08, "loss": 0.0146, "step": 79170 }, { "grad_norm": 0.2309613674879074, "learning_rate": 2.879102986725335e-08, "loss": 0.0116, "step": 79180 }, { "grad_norm": 0.21463607251644135, "learning_rate": 2.809400164389131e-08, "loss": 0.0146, "step": 79190 }, { "grad_norm": 0.29783356189727783, "learning_rate": 2.740551225253185e-08, "loss": 0.018, "step": 79200 }, { "grad_norm": 0.21663692593574524, "learning_rate": 2.6725561810819754e-08, "loss": 0.016, "step": 79210 }, { "grad_norm": 0.2564619183540344, "learning_rate": 2.605415043493431e-08, "loss": 0.0143, "step": 79220 }, { "grad_norm": 0.29587522149086, "learning_rate": 2.5391278239605965e-08, "loss": 0.014, "step": 79230 }, { "grad_norm": 0.17025981843471527, "learning_rate": 2.4736945338094118e-08, "loss": 0.0132, "step": 79240 }, { "grad_norm": 0.28611600399017334, "learning_rate": 2.4091151842214887e-08, "loss": 0.0137, "step": 79250 }, { "grad_norm": 0.28681480884552, "learning_rate": 2.3453897862318884e-08, "loss": 0.0146, "step": 79260 }, { "grad_norm": 0.20111192762851715, "learning_rate": 2.2825183507285686e-08, "loss": 0.0128, "step": 79270 }, { "grad_norm": 0.3334943652153015, "learning_rate": 2.220500888455157e-08, "loss": 0.0198, "step": 79280 }, { "grad_norm": 0.2015855610370636, "learning_rate": 2.1593374100081777e-08, "loss": 0.0196, "step": 79290 }, { "grad_norm": 0.3340941071510315, "learning_rate": 2.09902792583927e-08, "loss": 0.0219, "step": 79300 }, { "grad_norm": 0.18196716904640198, "learning_rate": 2.0395724462540788e-08, "loss": 0.0136, "step": 79310 }, { "grad_norm": 0.25093257427215576, "learning_rate": 1.9809709814111453e-08, "loss": 0.0137, "step": 79320 }, { "grad_norm": 0.27493125200271606, "learning_rate": 1.923223541324126e-08, "loss": 0.0169, "step": 79330 }, { "grad_norm": 0.23570026457309723, "learning_rate": 1.8663301358606833e-08, "loss": 0.0171, "step": 79340 }, { "grad_norm": 0.7333954572677612, "learning_rate": 1.8102907747419295e-08, "loss": 0.0179, "step": 79350 }, { "grad_norm": 0.5487416386604309, "learning_rate": 1.7551054675435385e-08, "loss": 0.0141, "step": 79360 }, { "grad_norm": 0.18844902515411377, "learning_rate": 1.7007742236957447e-08, "loss": 0.0152, "step": 79370 }, { "grad_norm": 0.19833669066429138, "learning_rate": 1.647297052481678e-08, "loss": 0.0145, "step": 79380 }, { "grad_norm": 0.1530066281557083, "learning_rate": 1.5946739630390285e-08, "loss": 0.0204, "step": 79390 }, { "grad_norm": 0.17445455491542816, "learning_rate": 1.5429049643606032e-08, "loss": 0.0203, "step": 79400 }, { "grad_norm": 0.3014028072357178, "learning_rate": 1.4919900652909935e-08, "loss": 0.0154, "step": 79410 }, { "grad_norm": 0.29394254088401794, "learning_rate": 1.4419292745310175e-08, "loss": 0.0178, "step": 79420 }, { "grad_norm": 0.2508317530155182, "learning_rate": 1.3927226006343885e-08, "loss": 0.0142, "step": 79430 }, { "grad_norm": 0.3747885525226593, "learning_rate": 1.3443700520093805e-08, "loss": 0.017, "step": 79440 }, { "grad_norm": 0.19489115476608276, "learning_rate": 1.296871636917718e-08, "loss": 0.0145, "step": 79450 }, { "grad_norm": 0.2016768455505371, "learning_rate": 1.2502273634762419e-08, "loss": 0.0143, "step": 79460 }, { "grad_norm": 0.35121574997901917, "learning_rate": 1.2044372396546876e-08, "loss": 0.0158, "step": 79470 }, { "grad_norm": 0.4677515923976898, "learning_rate": 1.1595012732773524e-08, "loss": 0.0154, "step": 79480 }, { "grad_norm": 0.13361871242523193, "learning_rate": 1.1154194720225386e-08, "loss": 0.0178, "step": 79490 }, { "grad_norm": 0.2359245866537094, "learning_rate": 1.0721918434231093e-08, "loss": 0.014, "step": 79500 }, { "grad_norm": 0.196902334690094, "learning_rate": 1.0298183948648232e-08, "loss": 0.0128, "step": 79510 }, { "grad_norm": 0.11775412410497665, "learning_rate": 9.88299133588555e-09, "loss": 0.0126, "step": 79520 }, { "grad_norm": 0.2312798649072647, "learning_rate": 9.476340666891847e-09, "loss": 0.0225, "step": 79530 }, { "grad_norm": 0.19068492949008942, "learning_rate": 9.078232011139332e-09, "loss": 0.0141, "step": 79540 }, { "grad_norm": 0.19501647353172302, "learning_rate": 8.68866543666802e-09, "loss": 0.0133, "step": 79550 }, { "grad_norm": 0.22054314613342285, "learning_rate": 8.307641010035782e-09, "loss": 0.0145, "step": 79560 }, { "grad_norm": 0.6899367570877075, "learning_rate": 7.93515879635165e-09, "loss": 0.0177, "step": 79570 }, { "grad_norm": 0.3013553321361542, "learning_rate": 7.571218859264706e-09, "loss": 0.0127, "step": 79580 }, { "grad_norm": 0.5194281935691833, "learning_rate": 7.215821260958544e-09, "loss": 0.0166, "step": 79590 }, { "grad_norm": 0.285050630569458, "learning_rate": 6.868966062162363e-09, "loss": 0.0157, "step": 79600 }, { "grad_norm": 0.16565701365470886, "learning_rate": 6.530653322145419e-09, "loss": 0.0119, "step": 79610 }, { "grad_norm": 0.18107955157756805, "learning_rate": 6.200883098717025e-09, "loss": 0.0121, "step": 79620 }, { "grad_norm": 0.39241528511047363, "learning_rate": 5.879655448226551e-09, "loss": 0.0169, "step": 79630 }, { "grad_norm": 0.22372779250144958, "learning_rate": 5.566970425557872e-09, "loss": 0.0175, "step": 79640 }, { "grad_norm": 0.405134916305542, "learning_rate": 5.26282808414047e-09, "loss": 0.0127, "step": 79650 }, { "grad_norm": 0.2126895785331726, "learning_rate": 4.967228475949437e-09, "loss": 0.0157, "step": 79660 }, { "grad_norm": 0.37857380509376526, "learning_rate": 4.680171651494369e-09, "loss": 0.0128, "step": 79670 }, { "grad_norm": 0.21075549721717834, "learning_rate": 4.40165765981937e-09, "loss": 0.0104, "step": 79680 }, { "grad_norm": 0.30616629123687744, "learning_rate": 4.131686548519698e-09, "loss": 0.0147, "step": 79690 }, { "grad_norm": 0.19079291820526123, "learning_rate": 3.8702583637251214e-09, "loss": 0.0134, "step": 79700 }, { "grad_norm": 0.1364702582359314, "learning_rate": 3.617373150105463e-09, "loss": 0.0133, "step": 79710 }, { "grad_norm": 0.27482691407203674, "learning_rate": 3.3730309508706036e-09, "loss": 0.0118, "step": 79720 }, { "grad_norm": 0.29341384768486023, "learning_rate": 3.137231807781582e-09, "loss": 0.0128, "step": 79730 }, { "grad_norm": 0.21177509427070618, "learning_rate": 2.9099757611172894e-09, "loss": 0.0174, "step": 79740 }, { "grad_norm": 0.20375467836856842, "learning_rate": 2.6912628497133275e-09, "loss": 0.0131, "step": 79750 }, { "grad_norm": 0.3018665313720703, "learning_rate": 2.481093110945354e-09, "loss": 0.0165, "step": 79760 }, { "grad_norm": 0.2195078730583191, "learning_rate": 2.2794665807235327e-09, "loss": 0.0144, "step": 79770 }, { "grad_norm": 0.15542343258857727, "learning_rate": 2.0863832934980843e-09, "loss": 0.0116, "step": 79780 }, { "grad_norm": 0.2209658920764923, "learning_rate": 1.901843282264837e-09, "loss": 0.0151, "step": 79790 }, { "grad_norm": 0.18847385048866272, "learning_rate": 1.7258465785541245e-09, "loss": 0.0178, "step": 79800 }, { "grad_norm": 0.12748630344867706, "learning_rate": 1.5583932124418887e-09, "loss": 0.0167, "step": 79810 }, { "grad_norm": 0.19130533933639526, "learning_rate": 1.3994832125441282e-09, "loss": 0.0154, "step": 79820 }, { "grad_norm": 0.2485083043575287, "learning_rate": 1.2491166060057958e-09, "loss": 0.0153, "step": 79830 }, { "grad_norm": 0.15285533666610718, "learning_rate": 1.1072934185230032e-09, "loss": 0.0143, "step": 79840 }, { "grad_norm": 0.18955183029174805, "learning_rate": 9.740136743319195e-10, "loss": 0.0145, "step": 79850 }, { "grad_norm": 0.2326972782611847, "learning_rate": 8.492773962087696e-10, "loss": 0.0123, "step": 79860 }, { "grad_norm": 0.393393874168396, "learning_rate": 7.330846054587337e-10, "loss": 0.0164, "step": 79870 }, { "grad_norm": 0.27857813239097595, "learning_rate": 6.254353219492526e-10, "loss": 0.0161, "step": 79880 }, { "grad_norm": 0.3712920546531677, "learning_rate": 5.263295640600685e-10, "loss": 0.0169, "step": 79890 }, { "grad_norm": 0.16923780739307404, "learning_rate": 4.357673487387359e-10, "loss": 0.0172, "step": 79900 }, { "grad_norm": 0.33952048420906067, "learning_rate": 3.537486914506616e-10, "loss": 0.0145, "step": 79910 }, { "grad_norm": 0.33845311403274536, "learning_rate": 2.8027360621241117e-10, "loss": 0.0163, "step": 79920 }, { "grad_norm": 0.2373988926410675, "learning_rate": 2.153421055806071e-10, "loss": 0.0135, "step": 79930 }, { "grad_norm": 0.916003942489624, "learning_rate": 1.5895420064637733e-10, "loss": 0.0183, "step": 79940 }, { "grad_norm": 0.12277350574731827, "learning_rate": 1.1110990105200891e-10, "loss": 0.0126, "step": 79950 }, { "grad_norm": 0.3783034682273865, "learning_rate": 7.180921496874326e-11, "loss": 0.0179, "step": 79960 }, { "grad_norm": 0.32122886180877686, "learning_rate": 4.10521491134297e-11, "loss": 0.0135, "step": 79970 }, { "grad_norm": 0.3007763922214508, "learning_rate": 1.883870874297422e-11, "loss": 0.0135, "step": 79980 }, { "grad_norm": 0.6216977834701538, "learning_rate": 5.1688976432373584e-12, "loss": 0.0203, "step": 79990 }, { "grad_norm": 0.22566671669483185, "learning_rate": 4.271816234080461e-14, "loss": 0.0134, "step": 80000 } ], "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": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }