{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.023407220840454, "learning_rate": 1.8e-07, "loss": 1.5965, "step": 10 }, { "grad_norm": 1.0144346952438354, "learning_rate": 3.8e-07, "loss": 1.5916, "step": 20 }, { "grad_norm": 1.0747151374816895, "learning_rate": 5.8e-07, "loss": 1.5944, "step": 30 }, { "grad_norm": 1.044240117073059, "learning_rate": 7.8e-07, "loss": 1.575, "step": 40 }, { "grad_norm": 1.022773265838623, "learning_rate": 9.8e-07, "loss": 1.5741, "step": 50 }, { "grad_norm": 1.2083618640899658, "learning_rate": 1.18e-06, "loss": 1.4859, "step": 60 }, { "grad_norm": 0.9550134539604187, "learning_rate": 1.3800000000000001e-06, "loss": 1.4091, "step": 70 }, { "grad_norm": 0.9244861006736755, "learning_rate": 1.5800000000000003e-06, "loss": 1.3669, "step": 80 }, { "grad_norm": 0.98438960313797, "learning_rate": 1.7800000000000001e-06, "loss": 1.3455, "step": 90 }, { "grad_norm": 1.1071566343307495, "learning_rate": 1.98e-06, "loss": 1.3037, "step": 100 }, { "grad_norm": 0.8282345533370972, "learning_rate": 2.1800000000000003e-06, "loss": 1.2386, "step": 110 }, { "grad_norm": 0.727868914604187, "learning_rate": 2.38e-06, "loss": 1.2012, "step": 120 }, { "grad_norm": 0.5978755354881287, "learning_rate": 2.5800000000000003e-06, "loss": 1.1686, "step": 130 }, { "grad_norm": 0.7166430354118347, "learning_rate": 2.78e-06, "loss": 1.1564, "step": 140 }, { "grad_norm": 0.6924200654029846, "learning_rate": 2.9800000000000003e-06, "loss": 1.1471, "step": 150 }, { "grad_norm": 0.8938196301460266, "learning_rate": 3.1800000000000005e-06, "loss": 1.1273, "step": 160 }, { "grad_norm": 1.0831204652786255, "learning_rate": 3.38e-06, "loss": 1.1158, "step": 170 }, { "grad_norm": 0.713912308216095, "learning_rate": 3.58e-06, "loss": 1.0982, "step": 180 }, { "grad_norm": 0.9929366111755371, "learning_rate": 3.7800000000000002e-06, "loss": 1.0949, "step": 190 }, { "grad_norm": 0.8392705917358398, "learning_rate": 3.98e-06, "loss": 1.0954, "step": 200 }, { "grad_norm": 0.8849682807922363, "learning_rate": 4.18e-06, "loss": 1.0883, "step": 210 }, { "grad_norm": 0.8804838061332703, "learning_rate": 4.38e-06, "loss": 1.0905, "step": 220 }, { "grad_norm": 1.0088492631912231, "learning_rate": 4.58e-06, "loss": 1.0854, "step": 230 }, { "grad_norm": 0.9600774645805359, "learning_rate": 4.780000000000001e-06, "loss": 1.076, "step": 240 }, { "grad_norm": 0.8432613015174866, "learning_rate": 4.98e-06, "loss": 1.0739, "step": 250 }, { "grad_norm": 1.2461122274398804, "learning_rate": 5.18e-06, "loss": 1.0747, "step": 260 }, { "grad_norm": 0.6867871284484863, "learning_rate": 5.38e-06, "loss": 1.0716, "step": 270 }, { "grad_norm": 1.0343058109283447, "learning_rate": 5.580000000000001e-06, "loss": 1.0662, "step": 280 }, { "grad_norm": 0.9265359044075012, "learning_rate": 5.78e-06, "loss": 1.0622, "step": 290 }, { "grad_norm": 1.0751471519470215, "learning_rate": 5.98e-06, "loss": 1.0473, "step": 300 }, { "grad_norm": 1.1314613819122314, "learning_rate": 6.18e-06, "loss": 1.0467, "step": 310 }, { "grad_norm": 1.5254580974578857, "learning_rate": 6.38e-06, "loss": 1.032, "step": 320 }, { "grad_norm": 1.2172986268997192, "learning_rate": 6.58e-06, "loss": 1.0052, "step": 330 }, { "grad_norm": 1.1255155801773071, "learning_rate": 6.78e-06, "loss": 0.9781, "step": 340 }, { "grad_norm": 1.3154104948043823, "learning_rate": 6.98e-06, "loss": 0.9246, "step": 350 }, { "grad_norm": 1.204949975013733, "learning_rate": 7.180000000000001e-06, "loss": 0.8836, "step": 360 }, { "grad_norm": 1.4141113758087158, "learning_rate": 7.3800000000000005e-06, "loss": 0.8256, "step": 370 }, { "grad_norm": 1.3229793310165405, "learning_rate": 7.580000000000001e-06, "loss": 0.7748, "step": 380 }, { "grad_norm": 1.6299710273742676, "learning_rate": 7.78e-06, "loss": 0.7244, "step": 390 }, { "grad_norm": 1.5823315382003784, "learning_rate": 7.98e-06, "loss": 0.6677, "step": 400 }, { "grad_norm": 1.6841987371444702, "learning_rate": 8.18e-06, "loss": 0.6278, "step": 410 }, { "grad_norm": 1.647455096244812, "learning_rate": 8.380000000000001e-06, "loss": 0.5761, "step": 420 }, { "grad_norm": 2.1119654178619385, "learning_rate": 8.580000000000001e-06, "loss": 0.55, "step": 430 }, { "grad_norm": 1.6627613306045532, "learning_rate": 8.78e-06, "loss": 0.4985, "step": 440 }, { "grad_norm": 1.794103980064392, "learning_rate": 8.98e-06, "loss": 0.4613, "step": 450 }, { "grad_norm": 1.912825345993042, "learning_rate": 9.180000000000002e-06, "loss": 0.4324, "step": 460 }, { "grad_norm": 1.816710114479065, "learning_rate": 9.38e-06, "loss": 0.3914, "step": 470 }, { "grad_norm": 2.246534585952759, "learning_rate": 9.58e-06, "loss": 0.3614, "step": 480 }, { "grad_norm": 1.900060772895813, "learning_rate": 9.78e-06, "loss": 0.329, "step": 490 }, { "grad_norm": 2.258197069168091, "learning_rate": 9.980000000000001e-06, "loss": 0.2975, "step": 500 }, { "grad_norm": 2.057415008544922, "learning_rate": 1.018e-05, "loss": 0.2603, "step": 510 }, { "grad_norm": 1.9273967742919922, "learning_rate": 1.038e-05, "loss": 0.2509, "step": 520 }, { "grad_norm": 2.1379237174987793, "learning_rate": 1.058e-05, "loss": 0.2369, "step": 530 }, { "grad_norm": 1.9979346990585327, "learning_rate": 1.0780000000000002e-05, "loss": 0.2294, "step": 540 }, { "grad_norm": 2.0828936100006104, "learning_rate": 1.098e-05, "loss": 0.2056, "step": 550 }, { "grad_norm": 2.0891759395599365, "learning_rate": 1.118e-05, "loss": 0.2165, "step": 560 }, { "grad_norm": 1.9882935285568237, "learning_rate": 1.1380000000000001e-05, "loss": 0.2129, "step": 570 }, { "grad_norm": 1.874115228652954, "learning_rate": 1.1580000000000001e-05, "loss": 0.2084, "step": 580 }, { "grad_norm": 1.9283794164657593, "learning_rate": 1.178e-05, "loss": 0.1974, "step": 590 }, { "grad_norm": 2.071951150894165, "learning_rate": 1.198e-05, "loss": 0.1956, "step": 600 }, { "grad_norm": 2.2661900520324707, "learning_rate": 1.2180000000000002e-05, "loss": 0.1871, "step": 610 }, { "grad_norm": 2.5992724895477295, "learning_rate": 1.238e-05, "loss": 0.1865, "step": 620 }, { "grad_norm": 2.2891383171081543, "learning_rate": 1.258e-05, "loss": 0.1801, "step": 630 }, { "grad_norm": 1.8812342882156372, "learning_rate": 1.278e-05, "loss": 0.178, "step": 640 }, { "grad_norm": 1.887235403060913, "learning_rate": 1.2980000000000001e-05, "loss": 0.1725, "step": 650 }, { "grad_norm": 2.0349955558776855, "learning_rate": 1.3180000000000001e-05, "loss": 0.1722, "step": 660 }, { "grad_norm": 1.94405198097229, "learning_rate": 1.338e-05, "loss": 0.1607, "step": 670 }, { "grad_norm": 2.1140260696411133, "learning_rate": 1.358e-05, "loss": 0.1545, "step": 680 }, { "grad_norm": 1.9190728664398193, "learning_rate": 1.3780000000000002e-05, "loss": 0.1548, "step": 690 }, { "grad_norm": 1.8172975778579712, "learning_rate": 1.3980000000000002e-05, "loss": 0.1555, "step": 700 }, { "grad_norm": 1.8845738172531128, "learning_rate": 1.4180000000000001e-05, "loss": 0.1542, "step": 710 }, { "grad_norm": 1.44257652759552, "learning_rate": 1.4380000000000001e-05, "loss": 0.1484, "step": 720 }, { "grad_norm": 1.897770881652832, "learning_rate": 1.4580000000000003e-05, "loss": 0.1531, "step": 730 }, { "grad_norm": 1.9622619152069092, "learning_rate": 1.4779999999999999e-05, "loss": 0.1491, "step": 740 }, { "grad_norm": 1.5351799726486206, "learning_rate": 1.4979999999999999e-05, "loss": 0.1446, "step": 750 }, { "grad_norm": 2.384256601333618, "learning_rate": 1.518e-05, "loss": 0.1436, "step": 760 }, { "grad_norm": 1.6154429912567139, "learning_rate": 1.538e-05, "loss": 0.1424, "step": 770 }, { "grad_norm": 2.0050137042999268, "learning_rate": 1.558e-05, "loss": 0.1402, "step": 780 }, { "grad_norm": 1.837302565574646, "learning_rate": 1.578e-05, "loss": 0.1272, "step": 790 }, { "grad_norm": 2.037637948989868, "learning_rate": 1.598e-05, "loss": 0.1367, "step": 800 }, { "grad_norm": 1.6250948905944824, "learning_rate": 1.618e-05, "loss": 0.1211, "step": 810 }, { "grad_norm": 2.130748987197876, "learning_rate": 1.6380000000000002e-05, "loss": 0.1257, "step": 820 }, { "grad_norm": 1.5630332231521606, "learning_rate": 1.658e-05, "loss": 0.1248, "step": 830 }, { "grad_norm": 1.6828413009643555, "learning_rate": 1.6780000000000002e-05, "loss": 0.1278, "step": 840 }, { "grad_norm": 1.7852076292037964, "learning_rate": 1.698e-05, "loss": 0.1218, "step": 850 }, { "grad_norm": 1.4798585176467896, "learning_rate": 1.718e-05, "loss": 0.1241, "step": 860 }, { "grad_norm": 2.11409068107605, "learning_rate": 1.7380000000000003e-05, "loss": 0.1206, "step": 870 }, { "grad_norm": 1.6466084718704224, "learning_rate": 1.758e-05, "loss": 0.1289, "step": 880 }, { "grad_norm": 1.529663324356079, "learning_rate": 1.7780000000000003e-05, "loss": 0.1191, "step": 890 }, { "grad_norm": 1.532305359840393, "learning_rate": 1.798e-05, "loss": 0.1226, "step": 900 }, { "grad_norm": 1.686171054840088, "learning_rate": 1.818e-05, "loss": 0.1128, "step": 910 }, { "grad_norm": 1.683931589126587, "learning_rate": 1.838e-05, "loss": 0.1164, "step": 920 }, { "grad_norm": 1.6764130592346191, "learning_rate": 1.858e-05, "loss": 0.1206, "step": 930 }, { "grad_norm": 1.5179399251937866, "learning_rate": 1.878e-05, "loss": 0.1182, "step": 940 }, { "grad_norm": 1.6653531789779663, "learning_rate": 1.898e-05, "loss": 0.1135, "step": 950 }, { "grad_norm": 1.6094447374343872, "learning_rate": 1.918e-05, "loss": 0.1117, "step": 960 }, { "grad_norm": 1.467617392539978, "learning_rate": 1.938e-05, "loss": 0.1111, "step": 970 }, { "grad_norm": 1.6445685625076294, "learning_rate": 1.9580000000000002e-05, "loss": 0.1092, "step": 980 }, { "grad_norm": 1.6433254480361938, "learning_rate": 1.978e-05, "loss": 0.1243, "step": 990 }, { "grad_norm": 1.6924450397491455, "learning_rate": 1.9980000000000002e-05, "loss": 0.1079, "step": 1000 }, { "grad_norm": 1.297592043876648, "learning_rate": 2.0180000000000003e-05, "loss": 0.1014, "step": 1010 }, { "grad_norm": 1.5432696342468262, "learning_rate": 2.038e-05, "loss": 0.1166, "step": 1020 }, { "grad_norm": 1.8124585151672363, "learning_rate": 2.0580000000000003e-05, "loss": 0.1023, "step": 1030 }, { "grad_norm": 1.8783527612686157, "learning_rate": 2.078e-05, "loss": 0.096, "step": 1040 }, { "grad_norm": 1.7120258808135986, "learning_rate": 2.098e-05, "loss": 0.1026, "step": 1050 }, { "grad_norm": 1.3108700513839722, "learning_rate": 2.118e-05, "loss": 0.0975, "step": 1060 }, { "grad_norm": 1.533889651298523, "learning_rate": 2.138e-05, "loss": 0.0998, "step": 1070 }, { "grad_norm": 1.9446218013763428, "learning_rate": 2.158e-05, "loss": 0.1086, "step": 1080 }, { "grad_norm": 1.6157476902008057, "learning_rate": 2.178e-05, "loss": 0.0996, "step": 1090 }, { "grad_norm": 1.4963892698287964, "learning_rate": 2.198e-05, "loss": 0.0953, "step": 1100 }, { "grad_norm": 1.583634853363037, "learning_rate": 2.218e-05, "loss": 0.0926, "step": 1110 }, { "grad_norm": 1.6390748023986816, "learning_rate": 2.2380000000000003e-05, "loss": 0.096, "step": 1120 }, { "grad_norm": 1.5786094665527344, "learning_rate": 2.258e-05, "loss": 0.0991, "step": 1130 }, { "grad_norm": 1.6103034019470215, "learning_rate": 2.2780000000000002e-05, "loss": 0.0971, "step": 1140 }, { "grad_norm": 1.6258008480072021, "learning_rate": 2.298e-05, "loss": 0.0915, "step": 1150 }, { "grad_norm": 1.5124375820159912, "learning_rate": 2.318e-05, "loss": 0.0959, "step": 1160 }, { "grad_norm": 1.1083059310913086, "learning_rate": 2.3380000000000003e-05, "loss": 0.097, "step": 1170 }, { "grad_norm": 1.5992770195007324, "learning_rate": 2.358e-05, "loss": 0.0989, "step": 1180 }, { "grad_norm": 1.512790322303772, "learning_rate": 2.3780000000000003e-05, "loss": 0.0902, "step": 1190 }, { "grad_norm": 1.7025574445724487, "learning_rate": 2.398e-05, "loss": 0.0976, "step": 1200 }, { "grad_norm": 1.438381552696228, "learning_rate": 2.418e-05, "loss": 0.0905, "step": 1210 }, { "grad_norm": 1.6402215957641602, "learning_rate": 2.438e-05, "loss": 0.0902, "step": 1220 }, { "grad_norm": 1.7492594718933105, "learning_rate": 2.4580000000000002e-05, "loss": 0.095, "step": 1230 }, { "grad_norm": 1.1890685558319092, "learning_rate": 2.478e-05, "loss": 0.0889, "step": 1240 }, { "grad_norm": 1.4870145320892334, "learning_rate": 2.498e-05, "loss": 0.086, "step": 1250 }, { "grad_norm": 1.2789709568023682, "learning_rate": 2.5180000000000003e-05, "loss": 0.0875, "step": 1260 }, { "grad_norm": 1.2717643976211548, "learning_rate": 2.5380000000000004e-05, "loss": 0.0894, "step": 1270 }, { "grad_norm": 1.2912788391113281, "learning_rate": 2.5580000000000002e-05, "loss": 0.0951, "step": 1280 }, { "grad_norm": 1.7258468866348267, "learning_rate": 2.5779999999999997e-05, "loss": 0.0932, "step": 1290 }, { "grad_norm": 1.3919613361358643, "learning_rate": 2.598e-05, "loss": 0.0911, "step": 1300 }, { "grad_norm": 1.3964340686798096, "learning_rate": 2.618e-05, "loss": 0.0866, "step": 1310 }, { "grad_norm": 1.5288422107696533, "learning_rate": 2.6379999999999998e-05, "loss": 0.0839, "step": 1320 }, { "grad_norm": 1.55057954788208, "learning_rate": 2.658e-05, "loss": 0.0968, "step": 1330 }, { "grad_norm": 1.156997799873352, "learning_rate": 2.678e-05, "loss": 0.0909, "step": 1340 }, { "grad_norm": 1.1974120140075684, "learning_rate": 2.698e-05, "loss": 0.0875, "step": 1350 }, { "grad_norm": 1.3488972187042236, "learning_rate": 2.718e-05, "loss": 0.0882, "step": 1360 }, { "grad_norm": 1.027553915977478, "learning_rate": 2.738e-05, "loss": 0.0899, "step": 1370 }, { "grad_norm": 1.47719407081604, "learning_rate": 2.758e-05, "loss": 0.0813, "step": 1380 }, { "grad_norm": 2.190035343170166, "learning_rate": 2.778e-05, "loss": 0.0823, "step": 1390 }, { "grad_norm": 1.5670576095581055, "learning_rate": 2.798e-05, "loss": 0.0841, "step": 1400 }, { "grad_norm": 1.2989089488983154, "learning_rate": 2.818e-05, "loss": 0.0869, "step": 1410 }, { "grad_norm": 1.1555933952331543, "learning_rate": 2.8380000000000003e-05, "loss": 0.0862, "step": 1420 }, { "grad_norm": 1.3815077543258667, "learning_rate": 2.858e-05, "loss": 0.0847, "step": 1430 }, { "grad_norm": 1.2403993606567383, "learning_rate": 2.8780000000000002e-05, "loss": 0.08, "step": 1440 }, { "grad_norm": 1.0357697010040283, "learning_rate": 2.898e-05, "loss": 0.0837, "step": 1450 }, { "grad_norm": 1.2772727012634277, "learning_rate": 2.9180000000000002e-05, "loss": 0.097, "step": 1460 }, { "grad_norm": 1.2637625932693481, "learning_rate": 2.9380000000000003e-05, "loss": 0.0812, "step": 1470 }, { "grad_norm": 1.5035655498504639, "learning_rate": 2.958e-05, "loss": 0.0858, "step": 1480 }, { "grad_norm": 1.8738455772399902, "learning_rate": 2.9780000000000003e-05, "loss": 0.0831, "step": 1490 }, { "grad_norm": 1.579250693321228, "learning_rate": 2.998e-05, "loss": 0.0823, "step": 1500 }, { "grad_norm": 1.357506275177002, "learning_rate": 3.0180000000000002e-05, "loss": 0.0892, "step": 1510 }, { "grad_norm": 1.1160790920257568, "learning_rate": 3.0380000000000004e-05, "loss": 0.0786, "step": 1520 }, { "grad_norm": 1.7116159200668335, "learning_rate": 3.058e-05, "loss": 0.0831, "step": 1530 }, { "grad_norm": 1.301310658454895, "learning_rate": 3.078e-05, "loss": 0.0885, "step": 1540 }, { "grad_norm": 1.3242789506912231, "learning_rate": 3.0980000000000005e-05, "loss": 0.0839, "step": 1550 }, { "grad_norm": 1.338824987411499, "learning_rate": 3.118e-05, "loss": 0.0841, "step": 1560 }, { "grad_norm": 1.2521884441375732, "learning_rate": 3.138e-05, "loss": 0.0887, "step": 1570 }, { "grad_norm": 1.3713350296020508, "learning_rate": 3.1580000000000006e-05, "loss": 0.0818, "step": 1580 }, { "grad_norm": 1.1202232837677002, "learning_rate": 3.1780000000000004e-05, "loss": 0.0803, "step": 1590 }, { "grad_norm": 1.2341430187225342, "learning_rate": 3.198e-05, "loss": 0.0835, "step": 1600 }, { "grad_norm": 1.3516989946365356, "learning_rate": 3.218e-05, "loss": 0.0785, "step": 1610 }, { "grad_norm": 1.1741987466812134, "learning_rate": 3.238e-05, "loss": 0.0749, "step": 1620 }, { "grad_norm": 1.2558022737503052, "learning_rate": 3.2579999999999996e-05, "loss": 0.0802, "step": 1630 }, { "grad_norm": 1.0546863079071045, "learning_rate": 3.278e-05, "loss": 0.0801, "step": 1640 }, { "grad_norm": 1.4028059244155884, "learning_rate": 3.298e-05, "loss": 0.0813, "step": 1650 }, { "grad_norm": 1.3283647298812866, "learning_rate": 3.318e-05, "loss": 0.0791, "step": 1660 }, { "grad_norm": 1.0877794027328491, "learning_rate": 3.338e-05, "loss": 0.0781, "step": 1670 }, { "grad_norm": 0.9140385389328003, "learning_rate": 3.358e-05, "loss": 0.0845, "step": 1680 }, { "grad_norm": 1.686955451965332, "learning_rate": 3.378e-05, "loss": 0.0886, "step": 1690 }, { "grad_norm": 1.2632222175598145, "learning_rate": 3.398e-05, "loss": 0.0877, "step": 1700 }, { "grad_norm": 1.3196531534194946, "learning_rate": 3.418e-05, "loss": 0.0767, "step": 1710 }, { "grad_norm": 1.2255463600158691, "learning_rate": 3.438e-05, "loss": 0.0768, "step": 1720 }, { "grad_norm": 1.4275097846984863, "learning_rate": 3.4580000000000004e-05, "loss": 0.0792, "step": 1730 }, { "grad_norm": 1.1105725765228271, "learning_rate": 3.478e-05, "loss": 0.0743, "step": 1740 }, { "grad_norm": 1.0376957654953003, "learning_rate": 3.498e-05, "loss": 0.0815, "step": 1750 }, { "grad_norm": 1.1636037826538086, "learning_rate": 3.518e-05, "loss": 0.0806, "step": 1760 }, { "grad_norm": 1.2055813074111938, "learning_rate": 3.5380000000000003e-05, "loss": 0.0785, "step": 1770 }, { "grad_norm": 1.2160841226577759, "learning_rate": 3.558e-05, "loss": 0.078, "step": 1780 }, { "grad_norm": 1.4266114234924316, "learning_rate": 3.578e-05, "loss": 0.0844, "step": 1790 }, { "grad_norm": 1.2084457874298096, "learning_rate": 3.5980000000000004e-05, "loss": 0.0795, "step": 1800 }, { "grad_norm": 1.107145071029663, "learning_rate": 3.618e-05, "loss": 0.0723, "step": 1810 }, { "grad_norm": 1.1719261407852173, "learning_rate": 3.638e-05, "loss": 0.0805, "step": 1820 }, { "grad_norm": 1.4169918298721313, "learning_rate": 3.6580000000000006e-05, "loss": 0.0759, "step": 1830 }, { "grad_norm": 1.2762503623962402, "learning_rate": 3.6780000000000004e-05, "loss": 0.0759, "step": 1840 }, { "grad_norm": 1.4640109539031982, "learning_rate": 3.698e-05, "loss": 0.0773, "step": 1850 }, { "grad_norm": 1.4663450717926025, "learning_rate": 3.7180000000000007e-05, "loss": 0.0782, "step": 1860 }, { "grad_norm": 1.264577031135559, "learning_rate": 3.7380000000000005e-05, "loss": 0.074, "step": 1870 }, { "grad_norm": 1.291122317314148, "learning_rate": 3.758e-05, "loss": 0.0752, "step": 1880 }, { "grad_norm": 0.9727973341941833, "learning_rate": 3.778000000000001e-05, "loss": 0.0771, "step": 1890 }, { "grad_norm": 0.8625799417495728, "learning_rate": 3.7980000000000006e-05, "loss": 0.0717, "step": 1900 }, { "grad_norm": 1.0067858695983887, "learning_rate": 3.818e-05, "loss": 0.0737, "step": 1910 }, { "grad_norm": 1.0855491161346436, "learning_rate": 3.838e-05, "loss": 0.0758, "step": 1920 }, { "grad_norm": 1.0056196451187134, "learning_rate": 3.858e-05, "loss": 0.0757, "step": 1930 }, { "grad_norm": 1.107189655303955, "learning_rate": 3.878e-05, "loss": 0.0766, "step": 1940 }, { "grad_norm": 0.6971107125282288, "learning_rate": 3.898e-05, "loss": 0.0756, "step": 1950 }, { "grad_norm": 1.145263910293579, "learning_rate": 3.918e-05, "loss": 0.0742, "step": 1960 }, { "grad_norm": 1.1210030317306519, "learning_rate": 3.938e-05, "loss": 0.0769, "step": 1970 }, { "grad_norm": 1.2341188192367554, "learning_rate": 3.958e-05, "loss": 0.073, "step": 1980 }, { "grad_norm": 1.2768328189849854, "learning_rate": 3.978e-05, "loss": 0.0755, "step": 1990 }, { "grad_norm": 1.1977512836456299, "learning_rate": 3.998e-05, "loss": 0.075, "step": 2000 }, { "grad_norm": 0.9498600959777832, "learning_rate": 4.018e-05, "loss": 0.0693, "step": 2010 }, { "grad_norm": 1.345524787902832, "learning_rate": 4.038e-05, "loss": 0.079, "step": 2020 }, { "grad_norm": 1.136863112449646, "learning_rate": 4.058e-05, "loss": 0.0737, "step": 2030 }, { "grad_norm": 1.365377426147461, "learning_rate": 4.078e-05, "loss": 0.0688, "step": 2040 }, { "grad_norm": 1.0300182104110718, "learning_rate": 4.0980000000000004e-05, "loss": 0.075, "step": 2050 }, { "grad_norm": 0.8991629481315613, "learning_rate": 4.118e-05, "loss": 0.0716, "step": 2060 }, { "grad_norm": 1.1588538885116577, "learning_rate": 4.138e-05, "loss": 0.0743, "step": 2070 }, { "grad_norm": 0.7864369750022888, "learning_rate": 4.1580000000000005e-05, "loss": 0.0674, "step": 2080 }, { "grad_norm": 1.0741896629333496, "learning_rate": 4.178e-05, "loss": 0.0723, "step": 2090 }, { "grad_norm": 0.9592103362083435, "learning_rate": 4.198e-05, "loss": 0.068, "step": 2100 }, { "grad_norm": 0.8487007021903992, "learning_rate": 4.2180000000000006e-05, "loss": 0.0637, "step": 2110 }, { "grad_norm": 0.6948816180229187, "learning_rate": 4.2380000000000004e-05, "loss": 0.0715, "step": 2120 }, { "grad_norm": 0.8854918479919434, "learning_rate": 4.258e-05, "loss": 0.08, "step": 2130 }, { "grad_norm": 1.1349329948425293, "learning_rate": 4.278e-05, "loss": 0.0696, "step": 2140 }, { "grad_norm": 0.9232969880104065, "learning_rate": 4.2980000000000005e-05, "loss": 0.0713, "step": 2150 }, { "grad_norm": 0.9127100110054016, "learning_rate": 4.318e-05, "loss": 0.0663, "step": 2160 }, { "grad_norm": 0.9009758234024048, "learning_rate": 4.338e-05, "loss": 0.0673, "step": 2170 }, { "grad_norm": 0.8359229564666748, "learning_rate": 4.3580000000000006e-05, "loss": 0.0683, "step": 2180 }, { "grad_norm": 0.8087654709815979, "learning_rate": 4.3780000000000004e-05, "loss": 0.0659, "step": 2190 }, { "grad_norm": 1.043697714805603, "learning_rate": 4.398e-05, "loss": 0.0777, "step": 2200 }, { "grad_norm": 1.0830026865005493, "learning_rate": 4.418000000000001e-05, "loss": 0.0748, "step": 2210 }, { "grad_norm": 1.0016895532608032, "learning_rate": 4.438e-05, "loss": 0.0744, "step": 2220 }, { "grad_norm": 0.872761070728302, "learning_rate": 4.458e-05, "loss": 0.0684, "step": 2230 }, { "grad_norm": 1.0914145708084106, "learning_rate": 4.478e-05, "loss": 0.0694, "step": 2240 }, { "grad_norm": 1.0603686571121216, "learning_rate": 4.498e-05, "loss": 0.0664, "step": 2250 }, { "grad_norm": 0.8586265444755554, "learning_rate": 4.518e-05, "loss": 0.0727, "step": 2260 }, { "grad_norm": 0.8644089102745056, "learning_rate": 4.538e-05, "loss": 0.0676, "step": 2270 }, { "grad_norm": 0.7633647918701172, "learning_rate": 4.558e-05, "loss": 0.0616, "step": 2280 }, { "grad_norm": 1.1022799015045166, "learning_rate": 4.578e-05, "loss": 0.0677, "step": 2290 }, { "grad_norm": 0.8648777604103088, "learning_rate": 4.5980000000000004e-05, "loss": 0.0733, "step": 2300 }, { "grad_norm": 1.0946836471557617, "learning_rate": 4.618e-05, "loss": 0.0742, "step": 2310 }, { "grad_norm": 0.8195168375968933, "learning_rate": 4.638e-05, "loss": 0.0759, "step": 2320 }, { "grad_norm": 1.2489172220230103, "learning_rate": 4.6580000000000005e-05, "loss": 0.0758, "step": 2330 }, { "grad_norm": 1.1511355638504028, "learning_rate": 4.678e-05, "loss": 0.0668, "step": 2340 }, { "grad_norm": 0.9578553438186646, "learning_rate": 4.698e-05, "loss": 0.0779, "step": 2350 }, { "grad_norm": 0.6806579828262329, "learning_rate": 4.718e-05, "loss": 0.0696, "step": 2360 }, { "grad_norm": 0.7447581887245178, "learning_rate": 4.7380000000000004e-05, "loss": 0.0687, "step": 2370 }, { "grad_norm": 1.2171930074691772, "learning_rate": 4.758e-05, "loss": 0.0711, "step": 2380 }, { "grad_norm": 0.8890669345855713, "learning_rate": 4.778e-05, "loss": 0.066, "step": 2390 }, { "grad_norm": 0.7680468559265137, "learning_rate": 4.7980000000000005e-05, "loss": 0.069, "step": 2400 }, { "grad_norm": 0.8701220750808716, "learning_rate": 4.818e-05, "loss": 0.0634, "step": 2410 }, { "grad_norm": 0.7631871700286865, "learning_rate": 4.838e-05, "loss": 0.0754, "step": 2420 }, { "grad_norm": 0.791573703289032, "learning_rate": 4.8580000000000006e-05, "loss": 0.0676, "step": 2430 }, { "grad_norm": 1.0542250871658325, "learning_rate": 4.8780000000000004e-05, "loss": 0.0711, "step": 2440 }, { "grad_norm": 0.8592124581336975, "learning_rate": 4.898e-05, "loss": 0.0685, "step": 2450 }, { "grad_norm": 0.8662562966346741, "learning_rate": 4.918000000000001e-05, "loss": 0.0642, "step": 2460 }, { "grad_norm": 0.8102145791053772, "learning_rate": 4.9380000000000005e-05, "loss": 0.0678, "step": 2470 }, { "grad_norm": 0.9714106917381287, "learning_rate": 4.958e-05, "loss": 0.068, "step": 2480 }, { "grad_norm": 0.9829491972923279, "learning_rate": 4.978e-05, "loss": 0.0735, "step": 2490 }, { "grad_norm": 0.8910744786262512, "learning_rate": 4.9980000000000006e-05, "loss": 0.0651, "step": 2500 }, { "grad_norm": 0.8933273553848267, "learning_rate": 5.0180000000000004e-05, "loss": 0.0642, "step": 2510 }, { "grad_norm": 0.8313777446746826, "learning_rate": 5.038e-05, "loss": 0.0635, "step": 2520 }, { "grad_norm": 0.737359344959259, "learning_rate": 5.058000000000001e-05, "loss": 0.0642, "step": 2530 }, { "grad_norm": 0.9815673828125, "learning_rate": 5.0780000000000005e-05, "loss": 0.0651, "step": 2540 }, { "grad_norm": 0.9378631711006165, "learning_rate": 5.098e-05, "loss": 0.0734, "step": 2550 }, { "grad_norm": 0.7085853219032288, "learning_rate": 5.118000000000001e-05, "loss": 0.0738, "step": 2560 }, { "grad_norm": 0.852828860282898, "learning_rate": 5.1380000000000006e-05, "loss": 0.0641, "step": 2570 }, { "grad_norm": 0.9445362687110901, "learning_rate": 5.1580000000000004e-05, "loss": 0.0659, "step": 2580 }, { "grad_norm": 0.8272019028663635, "learning_rate": 5.178000000000001e-05, "loss": 0.0606, "step": 2590 }, { "grad_norm": 0.8419458866119385, "learning_rate": 5.198000000000001e-05, "loss": 0.0698, "step": 2600 }, { "grad_norm": 0.7736785411834717, "learning_rate": 5.2180000000000005e-05, "loss": 0.0653, "step": 2610 }, { "grad_norm": 0.6895586252212524, "learning_rate": 5.238000000000001e-05, "loss": 0.0593, "step": 2620 }, { "grad_norm": 0.8367090821266174, "learning_rate": 5.258000000000001e-05, "loss": 0.0605, "step": 2630 }, { "grad_norm": 0.658772885799408, "learning_rate": 5.2780000000000006e-05, "loss": 0.0661, "step": 2640 }, { "grad_norm": 0.6160585284233093, "learning_rate": 5.2980000000000004e-05, "loss": 0.0627, "step": 2650 }, { "grad_norm": 0.8624112010002136, "learning_rate": 5.318000000000001e-05, "loss": 0.0615, "step": 2660 }, { "grad_norm": 0.6816967129707336, "learning_rate": 5.338000000000001e-05, "loss": 0.0618, "step": 2670 }, { "grad_norm": 0.8099098801612854, "learning_rate": 5.3580000000000005e-05, "loss": 0.0646, "step": 2680 }, { "grad_norm": 1.0644522905349731, "learning_rate": 5.378e-05, "loss": 0.0672, "step": 2690 }, { "grad_norm": 0.6442034244537354, "learning_rate": 5.3979999999999995e-05, "loss": 0.0631, "step": 2700 }, { "grad_norm": 0.6626574993133545, "learning_rate": 5.418e-05, "loss": 0.0619, "step": 2710 }, { "grad_norm": 0.835218071937561, "learning_rate": 5.438e-05, "loss": 0.065, "step": 2720 }, { "grad_norm": 0.9066065549850464, "learning_rate": 5.4579999999999996e-05, "loss": 0.0637, "step": 2730 }, { "grad_norm": 0.8744216561317444, "learning_rate": 5.478e-05, "loss": 0.0614, "step": 2740 }, { "grad_norm": 0.8036755323410034, "learning_rate": 5.498e-05, "loss": 0.0576, "step": 2750 }, { "grad_norm": 0.6660366058349609, "learning_rate": 5.518e-05, "loss": 0.0572, "step": 2760 }, { "grad_norm": 0.6567858457565308, "learning_rate": 5.538e-05, "loss": 0.0625, "step": 2770 }, { "grad_norm": 0.680991530418396, "learning_rate": 5.558e-05, "loss": 0.0579, "step": 2780 }, { "grad_norm": 0.9679102897644043, "learning_rate": 5.578e-05, "loss": 0.0584, "step": 2790 }, { "grad_norm": 1.031913161277771, "learning_rate": 5.5979999999999996e-05, "loss": 0.0598, "step": 2800 }, { "grad_norm": 0.952414333820343, "learning_rate": 5.618e-05, "loss": 0.0653, "step": 2810 }, { "grad_norm": 0.8968058824539185, "learning_rate": 5.638e-05, "loss": 0.0643, "step": 2820 }, { "grad_norm": 0.7751467823982239, "learning_rate": 5.658e-05, "loss": 0.0608, "step": 2830 }, { "grad_norm": 0.6936320066452026, "learning_rate": 5.678e-05, "loss": 0.0595, "step": 2840 }, { "grad_norm": 0.7646088600158691, "learning_rate": 5.698e-05, "loss": 0.0583, "step": 2850 }, { "grad_norm": 0.8810792565345764, "learning_rate": 5.718e-05, "loss": 0.0639, "step": 2860 }, { "grad_norm": 0.8340914845466614, "learning_rate": 5.738e-05, "loss": 0.0625, "step": 2870 }, { "grad_norm": 0.6831308007240295, "learning_rate": 5.758e-05, "loss": 0.0603, "step": 2880 }, { "grad_norm": 0.8145996332168579, "learning_rate": 5.778e-05, "loss": 0.0588, "step": 2890 }, { "grad_norm": 0.9561071395874023, "learning_rate": 5.7980000000000004e-05, "loss": 0.0566, "step": 2900 }, { "grad_norm": 0.9915790557861328, "learning_rate": 5.818e-05, "loss": 0.0571, "step": 2910 }, { "grad_norm": 0.7041525840759277, "learning_rate": 5.838e-05, "loss": 0.0585, "step": 2920 }, { "grad_norm": 0.8726305961608887, "learning_rate": 5.858e-05, "loss": 0.0634, "step": 2930 }, { "grad_norm": 0.8526113033294678, "learning_rate": 5.878e-05, "loss": 0.0659, "step": 2940 }, { "grad_norm": 0.9721672534942627, "learning_rate": 5.898e-05, "loss": 0.0585, "step": 2950 }, { "grad_norm": 0.6632972359657288, "learning_rate": 5.918e-05, "loss": 0.0595, "step": 2960 }, { "grad_norm": 0.7097289562225342, "learning_rate": 5.9380000000000004e-05, "loss": 0.055, "step": 2970 }, { "grad_norm": 0.47635555267333984, "learning_rate": 5.958e-05, "loss": 0.0531, "step": 2980 }, { "grad_norm": 0.7958394885063171, "learning_rate": 5.978e-05, "loss": 0.0551, "step": 2990 }, { "grad_norm": 0.683406412601471, "learning_rate": 5.9980000000000005e-05, "loss": 0.0599, "step": 3000 }, { "grad_norm": 0.6262069940567017, "learning_rate": 6.018e-05, "loss": 0.0583, "step": 3010 }, { "grad_norm": 0.7562952041625977, "learning_rate": 6.038e-05, "loss": 0.0579, "step": 3020 }, { "grad_norm": 0.6233190298080444, "learning_rate": 6.0580000000000006e-05, "loss": 0.0547, "step": 3030 }, { "grad_norm": 0.6364975571632385, "learning_rate": 6.0780000000000004e-05, "loss": 0.0581, "step": 3040 }, { "grad_norm": 0.7921048998832703, "learning_rate": 6.098e-05, "loss": 0.0542, "step": 3050 }, { "grad_norm": 0.8053972125053406, "learning_rate": 6.118000000000001e-05, "loss": 0.0551, "step": 3060 }, { "grad_norm": 0.9008715748786926, "learning_rate": 6.138e-05, "loss": 0.0561, "step": 3070 }, { "grad_norm": 0.8635597229003906, "learning_rate": 6.158e-05, "loss": 0.0612, "step": 3080 }, { "grad_norm": 0.619961678981781, "learning_rate": 6.178000000000001e-05, "loss": 0.0629, "step": 3090 }, { "grad_norm": 0.591262698173523, "learning_rate": 6.198e-05, "loss": 0.0643, "step": 3100 }, { "grad_norm": 0.566543459892273, "learning_rate": 6.218e-05, "loss": 0.0573, "step": 3110 }, { "grad_norm": 0.7321069240570068, "learning_rate": 6.238000000000001e-05, "loss": 0.0616, "step": 3120 }, { "grad_norm": 0.6816449165344238, "learning_rate": 6.258e-05, "loss": 0.0623, "step": 3130 }, { "grad_norm": 0.6710074543952942, "learning_rate": 6.278e-05, "loss": 0.0568, "step": 3140 }, { "grad_norm": 0.6734244227409363, "learning_rate": 6.298000000000001e-05, "loss": 0.0558, "step": 3150 }, { "grad_norm": 0.8777141571044922, "learning_rate": 6.318e-05, "loss": 0.0593, "step": 3160 }, { "grad_norm": 0.7409667372703552, "learning_rate": 6.338e-05, "loss": 0.0586, "step": 3170 }, { "grad_norm": 0.6026716828346252, "learning_rate": 6.358000000000001e-05, "loss": 0.0555, "step": 3180 }, { "grad_norm": 0.6889919638633728, "learning_rate": 6.378e-05, "loss": 0.059, "step": 3190 }, { "grad_norm": 0.6342737078666687, "learning_rate": 6.398000000000001e-05, "loss": 0.0593, "step": 3200 }, { "grad_norm": 0.6677775979042053, "learning_rate": 6.418000000000001e-05, "loss": 0.0603, "step": 3210 }, { "grad_norm": 0.5305030941963196, "learning_rate": 6.438e-05, "loss": 0.0511, "step": 3220 }, { "grad_norm": 0.7689367532730103, "learning_rate": 6.458000000000001e-05, "loss": 0.0517, "step": 3230 }, { "grad_norm": 0.6426669359207153, "learning_rate": 6.478000000000001e-05, "loss": 0.0571, "step": 3240 }, { "grad_norm": 0.6198694109916687, "learning_rate": 6.498e-05, "loss": 0.055, "step": 3250 }, { "grad_norm": 0.6614668965339661, "learning_rate": 6.518000000000001e-05, "loss": 0.0575, "step": 3260 }, { "grad_norm": 0.816674530506134, "learning_rate": 6.538000000000001e-05, "loss": 0.0608, "step": 3270 }, { "grad_norm": 0.6837171912193298, "learning_rate": 6.558e-05, "loss": 0.0751, "step": 3280 }, { "grad_norm": 0.6544035077095032, "learning_rate": 6.578000000000001e-05, "loss": 0.071, "step": 3290 }, { "grad_norm": 0.5683621168136597, "learning_rate": 6.598e-05, "loss": 0.0595, "step": 3300 }, { "grad_norm": 0.5263592600822449, "learning_rate": 6.618e-05, "loss": 0.0594, "step": 3310 }, { "grad_norm": 0.5660032629966736, "learning_rate": 6.638e-05, "loss": 0.0593, "step": 3320 }, { "grad_norm": 0.8180565237998962, "learning_rate": 6.658e-05, "loss": 0.058, "step": 3330 }, { "grad_norm": 0.5847870707511902, "learning_rate": 6.678e-05, "loss": 0.0599, "step": 3340 }, { "grad_norm": 0.6807529330253601, "learning_rate": 6.698e-05, "loss": 0.0563, "step": 3350 }, { "grad_norm": 0.6235266923904419, "learning_rate": 6.718e-05, "loss": 0.0568, "step": 3360 }, { "grad_norm": 0.623821496963501, "learning_rate": 6.738e-05, "loss": 0.0626, "step": 3370 }, { "grad_norm": 0.7032490968704224, "learning_rate": 6.758e-05, "loss": 0.0575, "step": 3380 }, { "grad_norm": 0.8420250415802002, "learning_rate": 6.778e-05, "loss": 0.0579, "step": 3390 }, { "grad_norm": 0.8289609551429749, "learning_rate": 6.798e-05, "loss": 0.0683, "step": 3400 }, { "grad_norm": 0.6110767126083374, "learning_rate": 6.818e-05, "loss": 0.0596, "step": 3410 }, { "grad_norm": 0.49623391032218933, "learning_rate": 6.838e-05, "loss": 0.0576, "step": 3420 }, { "grad_norm": 0.6309054493904114, "learning_rate": 6.858e-05, "loss": 0.0582, "step": 3430 }, { "grad_norm": 0.7614262700080872, "learning_rate": 6.878e-05, "loss": 0.0643, "step": 3440 }, { "grad_norm": 0.7703408002853394, "learning_rate": 6.898e-05, "loss": 0.0601, "step": 3450 }, { "grad_norm": 0.761204719543457, "learning_rate": 6.918e-05, "loss": 0.0628, "step": 3460 }, { "grad_norm": 0.7938456535339355, "learning_rate": 6.938e-05, "loss": 0.053, "step": 3470 }, { "grad_norm": 0.7667320370674133, "learning_rate": 6.958e-05, "loss": 0.0591, "step": 3480 }, { "grad_norm": 0.7216870784759521, "learning_rate": 6.978e-05, "loss": 0.0645, "step": 3490 }, { "grad_norm": 0.6868987083435059, "learning_rate": 6.998e-05, "loss": 0.0615, "step": 3500 }, { "grad_norm": 0.8749945759773254, "learning_rate": 7.018e-05, "loss": 0.0559, "step": 3510 }, { "grad_norm": 0.5575623512268066, "learning_rate": 7.038e-05, "loss": 0.0603, "step": 3520 }, { "grad_norm": 0.5109047889709473, "learning_rate": 7.058e-05, "loss": 0.0587, "step": 3530 }, { "grad_norm": 0.6596612930297852, "learning_rate": 7.078e-05, "loss": 0.0551, "step": 3540 }, { "grad_norm": 0.7300117611885071, "learning_rate": 7.098e-05, "loss": 0.053, "step": 3550 }, { "grad_norm": 0.5924200415611267, "learning_rate": 7.118e-05, "loss": 0.053, "step": 3560 }, { "grad_norm": 0.600793719291687, "learning_rate": 7.138e-05, "loss": 0.0529, "step": 3570 }, { "grad_norm": 0.6228575110435486, "learning_rate": 7.158e-05, "loss": 0.0573, "step": 3580 }, { "grad_norm": 0.7635850310325623, "learning_rate": 7.178000000000001e-05, "loss": 0.055, "step": 3590 }, { "grad_norm": 0.6586592197418213, "learning_rate": 7.198e-05, "loss": 0.0627, "step": 3600 }, { "grad_norm": 0.6907250881195068, "learning_rate": 7.218e-05, "loss": 0.0587, "step": 3610 }, { "grad_norm": 0.6046475172042847, "learning_rate": 7.238000000000001e-05, "loss": 0.0511, "step": 3620 }, { "grad_norm": 0.4910253882408142, "learning_rate": 7.258e-05, "loss": 0.0529, "step": 3630 }, { "grad_norm": 0.58733069896698, "learning_rate": 7.278e-05, "loss": 0.053, "step": 3640 }, { "grad_norm": 0.6628770232200623, "learning_rate": 7.298000000000001e-05, "loss": 0.055, "step": 3650 }, { "grad_norm": 0.5377957224845886, "learning_rate": 7.318e-05, "loss": 0.0559, "step": 3660 }, { "grad_norm": 0.6379234194755554, "learning_rate": 7.338e-05, "loss": 0.0522, "step": 3670 }, { "grad_norm": 0.6253263354301453, "learning_rate": 7.358000000000001e-05, "loss": 0.0494, "step": 3680 }, { "grad_norm": 0.6193934082984924, "learning_rate": 7.378e-05, "loss": 0.0528, "step": 3690 }, { "grad_norm": 0.600163459777832, "learning_rate": 7.398e-05, "loss": 0.0528, "step": 3700 }, { "grad_norm": 0.6069223284721375, "learning_rate": 7.418000000000001e-05, "loss": 0.0532, "step": 3710 }, { "grad_norm": 0.5121361017227173, "learning_rate": 7.438e-05, "loss": 0.0506, "step": 3720 }, { "grad_norm": 0.6166149377822876, "learning_rate": 7.458000000000001e-05, "loss": 0.0493, "step": 3730 }, { "grad_norm": 0.5227261185646057, "learning_rate": 7.478e-05, "loss": 0.0479, "step": 3740 }, { "grad_norm": 0.565218985080719, "learning_rate": 7.498e-05, "loss": 0.0524, "step": 3750 }, { "grad_norm": 0.44434353709220886, "learning_rate": 7.518000000000001e-05, "loss": 0.05, "step": 3760 }, { "grad_norm": 0.5096724629402161, "learning_rate": 7.538e-05, "loss": 0.0516, "step": 3770 }, { "grad_norm": 0.7262697219848633, "learning_rate": 7.558e-05, "loss": 0.052, "step": 3780 }, { "grad_norm": 0.8016819357872009, "learning_rate": 7.578000000000001e-05, "loss": 0.0562, "step": 3790 }, { "grad_norm": 0.6487345099449158, "learning_rate": 7.598e-05, "loss": 0.0583, "step": 3800 }, { "grad_norm": 0.7248115539550781, "learning_rate": 7.618e-05, "loss": 0.0522, "step": 3810 }, { "grad_norm": 0.7238807082176208, "learning_rate": 7.638000000000001e-05, "loss": 0.0547, "step": 3820 }, { "grad_norm": 0.47141319513320923, "learning_rate": 7.658e-05, "loss": 0.0554, "step": 3830 }, { "grad_norm": 0.49793267250061035, "learning_rate": 7.678000000000001e-05, "loss": 0.0553, "step": 3840 }, { "grad_norm": 0.6460472941398621, "learning_rate": 7.698000000000001e-05, "loss": 0.0533, "step": 3850 }, { "grad_norm": 0.5305652618408203, "learning_rate": 7.718e-05, "loss": 0.0529, "step": 3860 }, { "grad_norm": 0.6242535710334778, "learning_rate": 7.738000000000001e-05, "loss": 0.0532, "step": 3870 }, { "grad_norm": 0.4848077595233917, "learning_rate": 7.758000000000001e-05, "loss": 0.0508, "step": 3880 }, { "grad_norm": 0.5770440697669983, "learning_rate": 7.778e-05, "loss": 0.0496, "step": 3890 }, { "grad_norm": 0.6591318249702454, "learning_rate": 7.798000000000001e-05, "loss": 0.0518, "step": 3900 }, { "grad_norm": 0.746138334274292, "learning_rate": 7.818000000000001e-05, "loss": 0.0573, "step": 3910 }, { "grad_norm": 0.621242344379425, "learning_rate": 7.838e-05, "loss": 0.0578, "step": 3920 }, { "grad_norm": 0.843133270740509, "learning_rate": 7.858000000000001e-05, "loss": 0.0529, "step": 3930 }, { "grad_norm": 0.6237853765487671, "learning_rate": 7.878e-05, "loss": 0.0594, "step": 3940 }, { "grad_norm": 0.5796158909797668, "learning_rate": 7.897999999999999e-05, "loss": 0.0584, "step": 3950 }, { "grad_norm": 0.5388773679733276, "learning_rate": 7.918e-05, "loss": 0.0568, "step": 3960 }, { "grad_norm": 0.6417845487594604, "learning_rate": 7.938e-05, "loss": 0.0556, "step": 3970 }, { "grad_norm": 0.7016259431838989, "learning_rate": 7.958e-05, "loss": 0.0628, "step": 3980 }, { "grad_norm": 0.7382046580314636, "learning_rate": 7.978e-05, "loss": 0.0627, "step": 3990 }, { "grad_norm": 0.5609980821609497, "learning_rate": 7.998e-05, "loss": 0.0551, "step": 4000 }, { "grad_norm": 0.41588395833969116, "learning_rate": 8.018e-05, "loss": 0.0535, "step": 4010 }, { "grad_norm": 0.6821397542953491, "learning_rate": 8.038e-05, "loss": 0.0548, "step": 4020 }, { "grad_norm": 0.5176898837089539, "learning_rate": 8.058e-05, "loss": 0.0523, "step": 4030 }, { "grad_norm": 0.4781378209590912, "learning_rate": 8.078e-05, "loss": 0.0528, "step": 4040 }, { "grad_norm": 0.5734107494354248, "learning_rate": 8.098e-05, "loss": 0.0486, "step": 4050 }, { "grad_norm": 0.5004308819770813, "learning_rate": 8.118e-05, "loss": 0.0498, "step": 4060 }, { "grad_norm": 0.4947890341281891, "learning_rate": 8.138e-05, "loss": 0.0556, "step": 4070 }, { "grad_norm": 0.5038651823997498, "learning_rate": 8.158e-05, "loss": 0.0535, "step": 4080 }, { "grad_norm": 0.49191778898239136, "learning_rate": 8.178e-05, "loss": 0.0525, "step": 4090 }, { "grad_norm": 0.5439951419830322, "learning_rate": 8.198e-05, "loss": 0.0555, "step": 4100 }, { "grad_norm": 0.6942946910858154, "learning_rate": 8.218e-05, "loss": 0.0586, "step": 4110 }, { "grad_norm": 0.6302281618118286, "learning_rate": 8.238000000000001e-05, "loss": 0.0605, "step": 4120 }, { "grad_norm": 0.5296270847320557, "learning_rate": 8.258e-05, "loss": 0.0588, "step": 4130 }, { "grad_norm": 0.5611451864242554, "learning_rate": 8.278e-05, "loss": 0.0635, "step": 4140 }, { "grad_norm": 0.5132719874382019, "learning_rate": 8.298000000000001e-05, "loss": 0.0552, "step": 4150 }, { "grad_norm": 0.5582407116889954, "learning_rate": 8.318e-05, "loss": 0.0584, "step": 4160 }, { "grad_norm": 0.5598846673965454, "learning_rate": 8.338e-05, "loss": 0.0577, "step": 4170 }, { "grad_norm": 0.5535320043563843, "learning_rate": 8.358e-05, "loss": 0.0564, "step": 4180 }, { "grad_norm": 0.5314646363258362, "learning_rate": 8.378e-05, "loss": 0.0494, "step": 4190 }, { "grad_norm": 0.5318379998207092, "learning_rate": 8.398e-05, "loss": 0.0495, "step": 4200 }, { "grad_norm": 0.5729165077209473, "learning_rate": 8.418e-05, "loss": 0.0508, "step": 4210 }, { "grad_norm": 0.5059546828269958, "learning_rate": 8.438e-05, "loss": 0.051, "step": 4220 }, { "grad_norm": 0.44463542103767395, "learning_rate": 8.458e-05, "loss": 0.05, "step": 4230 }, { "grad_norm": 0.6675571799278259, "learning_rate": 8.478e-05, "loss": 0.055, "step": 4240 }, { "grad_norm": 0.39588463306427, "learning_rate": 8.498e-05, "loss": 0.0492, "step": 4250 }, { "grad_norm": 0.5677847862243652, "learning_rate": 8.518000000000001e-05, "loss": 0.0495, "step": 4260 }, { "grad_norm": 0.5766999125480652, "learning_rate": 8.538e-05, "loss": 0.0453, "step": 4270 }, { "grad_norm": 0.5542147159576416, "learning_rate": 8.558e-05, "loss": 0.0497, "step": 4280 }, { "grad_norm": 0.6002422571182251, "learning_rate": 8.578000000000001e-05, "loss": 0.0536, "step": 4290 }, { "grad_norm": 0.4922868013381958, "learning_rate": 8.598e-05, "loss": 0.0524, "step": 4300 }, { "grad_norm": 0.6979566216468811, "learning_rate": 8.618e-05, "loss": 0.0558, "step": 4310 }, { "grad_norm": 0.48377490043640137, "learning_rate": 8.638000000000001e-05, "loss": 0.0534, "step": 4320 }, { "grad_norm": 0.4448906183242798, "learning_rate": 8.658e-05, "loss": 0.0545, "step": 4330 }, { "grad_norm": 0.618208110332489, "learning_rate": 8.678e-05, "loss": 0.0553, "step": 4340 }, { "grad_norm": 0.4908168911933899, "learning_rate": 8.698000000000001e-05, "loss": 0.0508, "step": 4350 }, { "grad_norm": 0.4588809013366699, "learning_rate": 8.718e-05, "loss": 0.0465, "step": 4360 }, { "grad_norm": 0.49169349670410156, "learning_rate": 8.738000000000001e-05, "loss": 0.0516, "step": 4370 }, { "grad_norm": 0.555799126625061, "learning_rate": 8.758000000000001e-05, "loss": 0.0534, "step": 4380 }, { "grad_norm": 0.5574902296066284, "learning_rate": 8.778e-05, "loss": 0.0526, "step": 4390 }, { "grad_norm": 0.3789088726043701, "learning_rate": 8.798000000000001e-05, "loss": 0.0547, "step": 4400 }, { "grad_norm": 0.4919670522212982, "learning_rate": 8.818000000000001e-05, "loss": 0.0515, "step": 4410 }, { "grad_norm": 0.5687004923820496, "learning_rate": 8.838e-05, "loss": 0.0471, "step": 4420 }, { "grad_norm": 0.6057011485099792, "learning_rate": 8.858000000000001e-05, "loss": 0.0488, "step": 4430 }, { "grad_norm": 0.5082142949104309, "learning_rate": 8.878000000000001e-05, "loss": 0.0463, "step": 4440 }, { "grad_norm": 0.4625323414802551, "learning_rate": 8.898e-05, "loss": 0.0451, "step": 4450 }, { "grad_norm": 0.5827093720436096, "learning_rate": 8.918000000000001e-05, "loss": 0.0449, "step": 4460 }, { "grad_norm": 0.5986301898956299, "learning_rate": 8.938e-05, "loss": 0.0494, "step": 4470 }, { "grad_norm": 0.5436224937438965, "learning_rate": 8.958e-05, "loss": 0.0536, "step": 4480 }, { "grad_norm": 0.4994352161884308, "learning_rate": 8.978000000000001e-05, "loss": 0.0492, "step": 4490 }, { "grad_norm": 0.6261689066886902, "learning_rate": 8.998e-05, "loss": 0.0519, "step": 4500 }, { "grad_norm": 0.5060194134712219, "learning_rate": 9.018000000000001e-05, "loss": 0.0481, "step": 4510 }, { "grad_norm": 0.6227735877037048, "learning_rate": 9.038000000000001e-05, "loss": 0.0476, "step": 4520 }, { "grad_norm": 0.4782765805721283, "learning_rate": 9.058e-05, "loss": 0.0478, "step": 4530 }, { "grad_norm": 0.570697009563446, "learning_rate": 9.078000000000001e-05, "loss": 0.05, "step": 4540 }, { "grad_norm": 0.5338484048843384, "learning_rate": 9.098000000000001e-05, "loss": 0.0507, "step": 4550 }, { "grad_norm": 0.41564828157424927, "learning_rate": 9.118e-05, "loss": 0.0476, "step": 4560 }, { "grad_norm": 0.3966507315635681, "learning_rate": 9.138e-05, "loss": 0.0437, "step": 4570 }, { "grad_norm": 0.5009426474571228, "learning_rate": 9.158e-05, "loss": 0.052, "step": 4580 }, { "grad_norm": 0.5358611345291138, "learning_rate": 9.178e-05, "loss": 0.0538, "step": 4590 }, { "grad_norm": 0.4125902056694031, "learning_rate": 9.198e-05, "loss": 0.0484, "step": 4600 }, { "grad_norm": 0.47420263290405273, "learning_rate": 9.218e-05, "loss": 0.0513, "step": 4610 }, { "grad_norm": 0.4198904037475586, "learning_rate": 9.238e-05, "loss": 0.049, "step": 4620 }, { "grad_norm": 0.5217586755752563, "learning_rate": 9.258e-05, "loss": 0.0491, "step": 4630 }, { "grad_norm": 0.4766596853733063, "learning_rate": 9.278e-05, "loss": 0.0566, "step": 4640 }, { "grad_norm": 0.4432125985622406, "learning_rate": 9.298e-05, "loss": 0.0553, "step": 4650 }, { "grad_norm": 0.6569498777389526, "learning_rate": 9.318e-05, "loss": 0.0557, "step": 4660 }, { "grad_norm": 0.691231369972229, "learning_rate": 9.338e-05, "loss": 0.0554, "step": 4670 }, { "grad_norm": 0.69048011302948, "learning_rate": 9.358e-05, "loss": 0.0515, "step": 4680 }, { "grad_norm": 0.4077949821949005, "learning_rate": 9.378e-05, "loss": 0.0553, "step": 4690 }, { "grad_norm": 0.47461405396461487, "learning_rate": 9.398e-05, "loss": 0.0506, "step": 4700 }, { "grad_norm": 0.4784652590751648, "learning_rate": 9.418e-05, "loss": 0.0525, "step": 4710 }, { "grad_norm": 0.48425722122192383, "learning_rate": 9.438e-05, "loss": 0.0539, "step": 4720 }, { "grad_norm": 0.45854929089546204, "learning_rate": 9.458e-05, "loss": 0.0488, "step": 4730 }, { "grad_norm": 0.48525556921958923, "learning_rate": 9.478e-05, "loss": 0.0497, "step": 4740 }, { "grad_norm": 0.5272268056869507, "learning_rate": 9.498e-05, "loss": 0.0522, "step": 4750 }, { "grad_norm": 0.6498347520828247, "learning_rate": 9.518000000000001e-05, "loss": 0.0594, "step": 4760 }, { "grad_norm": 0.5023589134216309, "learning_rate": 9.538e-05, "loss": 0.0541, "step": 4770 }, { "grad_norm": 0.4810951352119446, "learning_rate": 9.558e-05, "loss": 0.0552, "step": 4780 }, { "grad_norm": 0.42483121156692505, "learning_rate": 9.578000000000001e-05, "loss": 0.052, "step": 4790 }, { "grad_norm": 0.47179487347602844, "learning_rate": 9.598e-05, "loss": 0.0493, "step": 4800 }, { "grad_norm": 0.4293002784252167, "learning_rate": 9.618e-05, "loss": 0.0535, "step": 4810 }, { "grad_norm": 0.38633114099502563, "learning_rate": 9.638000000000001e-05, "loss": 0.0516, "step": 4820 }, { "grad_norm": 0.48634254932403564, "learning_rate": 9.658e-05, "loss": 0.051, "step": 4830 }, { "grad_norm": 0.5972772240638733, "learning_rate": 9.678e-05, "loss": 0.0482, "step": 4840 }, { "grad_norm": 0.46994972229003906, "learning_rate": 9.698000000000001e-05, "loss": 0.0451, "step": 4850 }, { "grad_norm": 0.48280075192451477, "learning_rate": 9.718e-05, "loss": 0.0522, "step": 4860 }, { "grad_norm": 0.4391314685344696, "learning_rate": 9.738e-05, "loss": 0.0498, "step": 4870 }, { "grad_norm": 0.49570968747138977, "learning_rate": 9.758000000000001e-05, "loss": 0.0451, "step": 4880 }, { "grad_norm": 0.39652219414711, "learning_rate": 9.778e-05, "loss": 0.049, "step": 4890 }, { "grad_norm": 0.4702596664428711, "learning_rate": 9.798000000000001e-05, "loss": 0.0512, "step": 4900 }, { "grad_norm": 0.4860886037349701, "learning_rate": 9.818000000000001e-05, "loss": 0.0462, "step": 4910 }, { "grad_norm": 0.5121074318885803, "learning_rate": 9.838e-05, "loss": 0.0523, "step": 4920 }, { "grad_norm": 0.5406942963600159, "learning_rate": 9.858000000000001e-05, "loss": 0.0548, "step": 4930 }, { "grad_norm": 0.4115707278251648, "learning_rate": 9.878e-05, "loss": 0.0568, "step": 4940 }, { "grad_norm": 0.4699900448322296, "learning_rate": 9.898e-05, "loss": 0.0567, "step": 4950 }, { "grad_norm": 0.575563371181488, "learning_rate": 9.918000000000001e-05, "loss": 0.0538, "step": 4960 }, { "grad_norm": 0.5196059942245483, "learning_rate": 9.938e-05, "loss": 0.0494, "step": 4970 }, { "grad_norm": 0.5048707127571106, "learning_rate": 9.958e-05, "loss": 0.0533, "step": 4980 }, { "grad_norm": 0.40335613489151, "learning_rate": 9.978000000000001e-05, "loss": 0.0521, "step": 4990 }, { "grad_norm": 0.525229811668396, "learning_rate": 9.998e-05, "loss": 0.0544, "step": 5000 }, { "grad_norm": 0.3658559024333954, "learning_rate": 9.999999778549045e-05, "loss": 0.052, "step": 5010 }, { "grad_norm": 0.5470863580703735, "learning_rate": 9.999999013039593e-05, "loss": 0.0533, "step": 5020 }, { "grad_norm": 0.4454970955848694, "learning_rate": 9.999997700737766e-05, "loss": 0.0523, "step": 5030 }, { "grad_norm": 0.4391884207725525, "learning_rate": 9.999995841643709e-05, "loss": 0.0542, "step": 5040 }, { "grad_norm": 0.43698152899742126, "learning_rate": 9.999993435757623e-05, "loss": 0.0483, "step": 5050 }, { "grad_norm": 0.7144546508789062, "learning_rate": 9.999990483079773e-05, "loss": 0.0497, "step": 5060 }, { "grad_norm": 0.4188135266304016, "learning_rate": 9.999986983610481e-05, "loss": 0.0452, "step": 5070 }, { "grad_norm": 0.5300975441932678, "learning_rate": 9.99998293735013e-05, "loss": 0.0458, "step": 5080 }, { "grad_norm": 0.5083855986595154, "learning_rate": 9.999978344299161e-05, "loss": 0.0512, "step": 5090 }, { "grad_norm": 0.459313303232193, "learning_rate": 9.99997320445808e-05, "loss": 0.0489, "step": 5100 }, { "grad_norm": 0.5342221856117249, "learning_rate": 9.999967517827444e-05, "loss": 0.0493, "step": 5110 }, { "grad_norm": 0.44480857253074646, "learning_rate": 9.999961284407879e-05, "loss": 0.0496, "step": 5120 }, { "grad_norm": 0.4076370298862457, "learning_rate": 9.999954504200067e-05, "loss": 0.0541, "step": 5130 }, { "grad_norm": 0.44458991289138794, "learning_rate": 9.999947177204744e-05, "loss": 0.0596, "step": 5140 }, { "grad_norm": 0.34042423963546753, "learning_rate": 9.999939303422718e-05, "loss": 0.0499, "step": 5150 }, { "grad_norm": 0.5053701996803284, "learning_rate": 9.999930882854847e-05, "loss": 0.0465, "step": 5160 }, { "grad_norm": 0.35064205527305603, "learning_rate": 9.999921915502051e-05, "loss": 0.0493, "step": 5170 }, { "grad_norm": 0.4442894458770752, "learning_rate": 9.99991240136531e-05, "loss": 0.0477, "step": 5180 }, { "grad_norm": 0.4766870141029358, "learning_rate": 9.999902340445668e-05, "loss": 0.0483, "step": 5190 }, { "grad_norm": 0.4450841546058655, "learning_rate": 9.999891732744224e-05, "loss": 0.0475, "step": 5200 }, { "grad_norm": 0.44025328755378723, "learning_rate": 9.999880578262135e-05, "loss": 0.048, "step": 5210 }, { "grad_norm": 0.46304789185523987, "learning_rate": 9.999868877000624e-05, "loss": 0.0506, "step": 5220 }, { "grad_norm": 0.5258839726448059, "learning_rate": 9.99985662896097e-05, "loss": 0.0522, "step": 5230 }, { "grad_norm": 0.4481131136417389, "learning_rate": 9.999843834144513e-05, "loss": 0.0487, "step": 5240 }, { "grad_norm": 0.4749087691307068, "learning_rate": 9.99983049255265e-05, "loss": 0.0526, "step": 5250 }, { "grad_norm": 0.3961416482925415, "learning_rate": 9.999816604186843e-05, "loss": 0.0536, "step": 5260 }, { "grad_norm": 0.3592783808708191, "learning_rate": 9.999802169048609e-05, "loss": 0.0487, "step": 5270 }, { "grad_norm": 0.4256177842617035, "learning_rate": 9.999787187139527e-05, "loss": 0.0465, "step": 5280 }, { "grad_norm": 0.6069108843803406, "learning_rate": 9.999771658461234e-05, "loss": 0.0468, "step": 5290 }, { "grad_norm": 0.43303659558296204, "learning_rate": 9.999755583015431e-05, "loss": 0.0496, "step": 5300 }, { "grad_norm": 0.4158823490142822, "learning_rate": 9.999738960803874e-05, "loss": 0.0459, "step": 5310 }, { "grad_norm": 0.4422527849674225, "learning_rate": 9.99972179182838e-05, "loss": 0.0484, "step": 5320 }, { "grad_norm": 0.4332650899887085, "learning_rate": 9.99970407609083e-05, "loss": 0.0446, "step": 5330 }, { "grad_norm": 0.49419739842414856, "learning_rate": 9.999685813593159e-05, "loss": 0.0451, "step": 5340 }, { "grad_norm": 0.41046595573425293, "learning_rate": 9.999667004337362e-05, "loss": 0.043, "step": 5350 }, { "grad_norm": 0.42968448996543884, "learning_rate": 9.9996476483255e-05, "loss": 0.0471, "step": 5360 }, { "grad_norm": 0.47571057081222534, "learning_rate": 9.999627745559688e-05, "loss": 0.0426, "step": 5370 }, { "grad_norm": 0.4825800657272339, "learning_rate": 9.999607296042101e-05, "loss": 0.0448, "step": 5380 }, { "grad_norm": 0.3722878098487854, "learning_rate": 9.99958629977498e-05, "loss": 0.0432, "step": 5390 }, { "grad_norm": 0.4174627959728241, "learning_rate": 9.999564756760615e-05, "loss": 0.046, "step": 5400 }, { "grad_norm": 0.36422812938690186, "learning_rate": 9.999542667001366e-05, "loss": 0.0453, "step": 5410 }, { "grad_norm": 0.4035061001777649, "learning_rate": 9.999520030499647e-05, "loss": 0.0465, "step": 5420 }, { "grad_norm": 0.42024675011634827, "learning_rate": 9.999496847257936e-05, "loss": 0.0442, "step": 5430 }, { "grad_norm": 0.4647481143474579, "learning_rate": 9.999473117278764e-05, "loss": 0.0414, "step": 5440 }, { "grad_norm": 0.42623838782310486, "learning_rate": 9.999448840564731e-05, "loss": 0.0441, "step": 5450 }, { "grad_norm": 0.5000053644180298, "learning_rate": 9.999424017118488e-05, "loss": 0.0434, "step": 5460 }, { "grad_norm": 0.4975408911705017, "learning_rate": 9.999398646942751e-05, "loss": 0.0447, "step": 5470 }, { "grad_norm": 0.42931172251701355, "learning_rate": 9.999372730040296e-05, "loss": 0.0427, "step": 5480 }, { "grad_norm": 0.35775893926620483, "learning_rate": 9.999346266413953e-05, "loss": 0.0428, "step": 5490 }, { "grad_norm": 0.6007570624351501, "learning_rate": 9.99931925606662e-05, "loss": 0.0473, "step": 5500 }, { "grad_norm": 0.4346855878829956, "learning_rate": 9.99929169900125e-05, "loss": 0.0433, "step": 5510 }, { "grad_norm": 0.3324742317199707, "learning_rate": 9.999263595220855e-05, "loss": 0.0463, "step": 5520 }, { "grad_norm": 0.39381101727485657, "learning_rate": 9.99923494472851e-05, "loss": 0.0405, "step": 5530 }, { "grad_norm": 0.5623794198036194, "learning_rate": 9.999205747527348e-05, "loss": 0.044, "step": 5540 }, { "grad_norm": 0.48194000124931335, "learning_rate": 9.999176003620561e-05, "loss": 0.0431, "step": 5550 }, { "grad_norm": 0.5383957028388977, "learning_rate": 9.999145713011405e-05, "loss": 0.0452, "step": 5560 }, { "grad_norm": 0.40431448817253113, "learning_rate": 9.999114875703186e-05, "loss": 0.0429, "step": 5570 }, { "grad_norm": 0.4551243484020233, "learning_rate": 9.999083491699281e-05, "loss": 0.0519, "step": 5580 }, { "grad_norm": 0.3347439765930176, "learning_rate": 9.999051561003123e-05, "loss": 0.044, "step": 5590 }, { "grad_norm": 0.3645385503768921, "learning_rate": 9.999019083618202e-05, "loss": 0.0473, "step": 5600 }, { "grad_norm": 0.530754566192627, "learning_rate": 9.99898605954807e-05, "loss": 0.0475, "step": 5610 }, { "grad_norm": 0.4295206069946289, "learning_rate": 9.998952488796338e-05, "loss": 0.0444, "step": 5620 }, { "grad_norm": 0.3763508200645447, "learning_rate": 9.998918371366676e-05, "loss": 0.0483, "step": 5630 }, { "grad_norm": 0.3945596218109131, "learning_rate": 9.99888370726282e-05, "loss": 0.0397, "step": 5640 }, { "grad_norm": 0.57344651222229, "learning_rate": 9.998848496488556e-05, "loss": 0.0466, "step": 5650 }, { "grad_norm": 0.3529115915298462, "learning_rate": 9.998812739047736e-05, "loss": 0.0483, "step": 5660 }, { "grad_norm": 0.4287039637565613, "learning_rate": 9.99877643494427e-05, "loss": 0.0495, "step": 5670 }, { "grad_norm": 0.4954642951488495, "learning_rate": 9.998739584182128e-05, "loss": 0.0469, "step": 5680 }, { "grad_norm": 0.409316748380661, "learning_rate": 9.998702186765342e-05, "loss": 0.0476, "step": 5690 }, { "grad_norm": 0.39564546942710876, "learning_rate": 9.998664242698e-05, "loss": 0.0428, "step": 5700 }, { "grad_norm": 0.32685163617134094, "learning_rate": 9.998625751984251e-05, "loss": 0.0443, "step": 5710 }, { "grad_norm": 0.4228897988796234, "learning_rate": 9.998586714628307e-05, "loss": 0.0416, "step": 5720 }, { "grad_norm": 0.3521948456764221, "learning_rate": 9.998547130634432e-05, "loss": 0.049, "step": 5730 }, { "grad_norm": 0.39936092495918274, "learning_rate": 9.99850700000696e-05, "loss": 0.0474, "step": 5740 }, { "grad_norm": 0.493640661239624, "learning_rate": 9.998466322750278e-05, "loss": 0.0448, "step": 5750 }, { "grad_norm": 0.48169785737991333, "learning_rate": 9.998425098868834e-05, "loss": 0.0427, "step": 5760 }, { "grad_norm": 0.5155724883079529, "learning_rate": 9.998383328367136e-05, "loss": 0.0456, "step": 5770 }, { "grad_norm": 0.3724830746650696, "learning_rate": 9.99834101124975e-05, "loss": 0.0478, "step": 5780 }, { "grad_norm": 0.2759036421775818, "learning_rate": 9.998298147521309e-05, "loss": 0.0402, "step": 5790 }, { "grad_norm": 0.43618062138557434, "learning_rate": 9.998254737186496e-05, "loss": 0.0485, "step": 5800 }, { "grad_norm": 0.49104613065719604, "learning_rate": 9.99821078025006e-05, "loss": 0.0473, "step": 5810 }, { "grad_norm": 0.49191972613334656, "learning_rate": 9.998166276716807e-05, "loss": 0.0501, "step": 5820 }, { "grad_norm": 0.33800873160362244, "learning_rate": 9.998121226591606e-05, "loss": 0.0422, "step": 5830 }, { "grad_norm": 0.4540763199329376, "learning_rate": 9.998075629879382e-05, "loss": 0.0454, "step": 5840 }, { "grad_norm": 0.3630666136741638, "learning_rate": 9.99802948658512e-05, "loss": 0.0475, "step": 5850 }, { "grad_norm": 0.36544278264045715, "learning_rate": 9.99798279671387e-05, "loss": 0.0516, "step": 5860 }, { "grad_norm": 0.49979346990585327, "learning_rate": 9.997935560270734e-05, "loss": 0.0511, "step": 5870 }, { "grad_norm": 0.5003499388694763, "learning_rate": 9.997887777260879e-05, "loss": 0.0511, "step": 5880 }, { "grad_norm": 0.44371894001960754, "learning_rate": 9.997839447689532e-05, "loss": 0.0424, "step": 5890 }, { "grad_norm": 0.4969932436943054, "learning_rate": 9.997790571561978e-05, "loss": 0.0468, "step": 5900 }, { "grad_norm": 0.4473041296005249, "learning_rate": 9.99774114888356e-05, "loss": 0.046, "step": 5910 }, { "grad_norm": 0.37948039174079895, "learning_rate": 9.997691179659684e-05, "loss": 0.0411, "step": 5920 }, { "grad_norm": 0.42956411838531494, "learning_rate": 9.997640663895815e-05, "loss": 0.0471, "step": 5930 }, { "grad_norm": 0.36295339465141296, "learning_rate": 9.997589601597477e-05, "loss": 0.0435, "step": 5940 }, { "grad_norm": 0.3883329927921295, "learning_rate": 9.997537992770252e-05, "loss": 0.0462, "step": 5950 }, { "grad_norm": 0.5018167495727539, "learning_rate": 9.997485837419788e-05, "loss": 0.0452, "step": 5960 }, { "grad_norm": 0.48685222864151, "learning_rate": 9.997433135551786e-05, "loss": 0.0424, "step": 5970 }, { "grad_norm": 0.5604907274246216, "learning_rate": 9.997379887172009e-05, "loss": 0.0432, "step": 5980 }, { "grad_norm": 0.4297420084476471, "learning_rate": 9.997326092286281e-05, "loss": 0.0476, "step": 5990 }, { "grad_norm": 0.4050474762916565, "learning_rate": 9.997271750900486e-05, "loss": 0.041, "step": 6000 }, { "grad_norm": 0.4848781228065491, "learning_rate": 9.997216863020565e-05, "loss": 0.0468, "step": 6010 }, { "grad_norm": 0.3925725519657135, "learning_rate": 9.99716142865252e-05, "loss": 0.0424, "step": 6020 }, { "grad_norm": 0.46111273765563965, "learning_rate": 9.997105447802415e-05, "loss": 0.0463, "step": 6030 }, { "grad_norm": 0.40027838945388794, "learning_rate": 9.997048920476373e-05, "loss": 0.0445, "step": 6040 }, { "grad_norm": 0.5421060919761658, "learning_rate": 9.996991846680572e-05, "loss": 0.0451, "step": 6050 }, { "grad_norm": 0.470480740070343, "learning_rate": 9.996934226421257e-05, "loss": 0.0464, "step": 6060 }, { "grad_norm": 0.41886311769485474, "learning_rate": 9.996876059704726e-05, "loss": 0.0449, "step": 6070 }, { "grad_norm": 0.3874169886112213, "learning_rate": 9.996817346537343e-05, "loss": 0.0494, "step": 6080 }, { "grad_norm": 0.35877591371536255, "learning_rate": 9.996758086925526e-05, "loss": 0.0414, "step": 6090 }, { "grad_norm": 0.34747883677482605, "learning_rate": 9.996698280875759e-05, "loss": 0.0422, "step": 6100 }, { "grad_norm": 0.4304755926132202, "learning_rate": 9.99663792839458e-05, "loss": 0.0463, "step": 6110 }, { "grad_norm": 0.33798518776893616, "learning_rate": 9.99657702948859e-05, "loss": 0.0449, "step": 6120 }, { "grad_norm": 0.3961312770843506, "learning_rate": 9.996515584164448e-05, "loss": 0.0426, "step": 6130 }, { "grad_norm": 0.44295117259025574, "learning_rate": 9.996453592428873e-05, "loss": 0.0434, "step": 6140 }, { "grad_norm": 0.4295002222061157, "learning_rate": 9.996391054288646e-05, "loss": 0.0396, "step": 6150 }, { "grad_norm": 0.3667892515659332, "learning_rate": 9.996327969750605e-05, "loss": 0.0421, "step": 6160 }, { "grad_norm": 0.39658573269844055, "learning_rate": 9.996264338821649e-05, "loss": 0.04, "step": 6170 }, { "grad_norm": 0.4937622845172882, "learning_rate": 9.996200161508735e-05, "loss": 0.0419, "step": 6180 }, { "grad_norm": 0.3774109184741974, "learning_rate": 9.996135437818885e-05, "loss": 0.042, "step": 6190 }, { "grad_norm": 0.4610752463340759, "learning_rate": 9.996070167759175e-05, "loss": 0.0428, "step": 6200 }, { "grad_norm": 0.40503203868865967, "learning_rate": 9.996004351336743e-05, "loss": 0.0416, "step": 6210 }, { "grad_norm": 0.4107005298137665, "learning_rate": 9.995937988558785e-05, "loss": 0.042, "step": 6220 }, { "grad_norm": 0.5781493782997131, "learning_rate": 9.995871079432561e-05, "loss": 0.0412, "step": 6230 }, { "grad_norm": 0.5061692595481873, "learning_rate": 9.995803623965389e-05, "loss": 0.0466, "step": 6240 }, { "grad_norm": 0.47709959745407104, "learning_rate": 9.995735622164641e-05, "loss": 0.0425, "step": 6250 }, { "grad_norm": 0.37679824233055115, "learning_rate": 9.995667074037758e-05, "loss": 0.0427, "step": 6260 }, { "grad_norm": 0.38907456398010254, "learning_rate": 9.995597979592232e-05, "loss": 0.0454, "step": 6270 }, { "grad_norm": 0.3636194169521332, "learning_rate": 9.995528338835625e-05, "loss": 0.0479, "step": 6280 }, { "grad_norm": 0.43156540393829346, "learning_rate": 9.995458151775547e-05, "loss": 0.0521, "step": 6290 }, { "grad_norm": 0.42730769515037537, "learning_rate": 9.995387418419677e-05, "loss": 0.0437, "step": 6300 }, { "grad_norm": 0.3445398807525635, "learning_rate": 9.99531613877575e-05, "loss": 0.0442, "step": 6310 }, { "grad_norm": 0.3624733090400696, "learning_rate": 9.995244312851559e-05, "loss": 0.0435, "step": 6320 }, { "grad_norm": 0.44414016604423523, "learning_rate": 9.995171940654961e-05, "loss": 0.0409, "step": 6330 }, { "grad_norm": 0.3389238715171814, "learning_rate": 9.995099022193871e-05, "loss": 0.0438, "step": 6340 }, { "grad_norm": 0.3924068808555603, "learning_rate": 9.995025557476261e-05, "loss": 0.0475, "step": 6350 }, { "grad_norm": 0.3816244304180145, "learning_rate": 9.994951546510165e-05, "loss": 0.0386, "step": 6360 }, { "grad_norm": 0.42926403880119324, "learning_rate": 9.994876989303679e-05, "loss": 0.0424, "step": 6370 }, { "grad_norm": 0.41241276264190674, "learning_rate": 9.994801885864955e-05, "loss": 0.041, "step": 6380 }, { "grad_norm": 0.31168287992477417, "learning_rate": 9.994726236202205e-05, "loss": 0.0416, "step": 6390 }, { "grad_norm": 0.397734135389328, "learning_rate": 9.994650040323704e-05, "loss": 0.0455, "step": 6400 }, { "grad_norm": 0.38081714510917664, "learning_rate": 9.994573298237784e-05, "loss": 0.0453, "step": 6410 }, { "grad_norm": 0.4594847559928894, "learning_rate": 9.994496009952837e-05, "loss": 0.0441, "step": 6420 }, { "grad_norm": 0.398040235042572, "learning_rate": 9.994418175477316e-05, "loss": 0.0441, "step": 6430 }, { "grad_norm": 0.5997098684310913, "learning_rate": 9.994339794819733e-05, "loss": 0.0427, "step": 6440 }, { "grad_norm": 0.4005016088485718, "learning_rate": 9.994260867988658e-05, "loss": 0.0417, "step": 6450 }, { "grad_norm": 0.4010942280292511, "learning_rate": 9.994181394992723e-05, "loss": 0.0458, "step": 6460 }, { "grad_norm": 0.3506339192390442, "learning_rate": 9.994101375840618e-05, "loss": 0.0405, "step": 6470 }, { "grad_norm": 0.45583704113960266, "learning_rate": 9.994020810541098e-05, "loss": 0.0465, "step": 6480 }, { "grad_norm": 0.36557522416114807, "learning_rate": 9.99393969910297e-05, "loss": 0.0473, "step": 6490 }, { "grad_norm": 0.3728505074977875, "learning_rate": 9.993858041535104e-05, "loss": 0.0415, "step": 6500 }, { "grad_norm": 0.3070676028728485, "learning_rate": 9.99377583784643e-05, "loss": 0.0474, "step": 6510 }, { "grad_norm": 0.38519003987312317, "learning_rate": 9.993693088045939e-05, "loss": 0.0426, "step": 6520 }, { "grad_norm": 0.47714534401893616, "learning_rate": 9.99360979214268e-05, "loss": 0.0414, "step": 6530 }, { "grad_norm": 0.3421889841556549, "learning_rate": 9.99352595014576e-05, "loss": 0.035, "step": 6540 }, { "grad_norm": 0.3566729724407196, "learning_rate": 9.993441562064354e-05, "loss": 0.042, "step": 6550 }, { "grad_norm": 0.3861135244369507, "learning_rate": 9.993356627907685e-05, "loss": 0.0423, "step": 6560 }, { "grad_norm": 0.36380061507225037, "learning_rate": 9.99327114768504e-05, "loss": 0.0423, "step": 6570 }, { "grad_norm": 0.29547184705734253, "learning_rate": 9.99318512140577e-05, "loss": 0.0383, "step": 6580 }, { "grad_norm": 0.3583581745624542, "learning_rate": 9.993098549079284e-05, "loss": 0.0404, "step": 6590 }, { "grad_norm": 0.3332158029079437, "learning_rate": 9.993011430715047e-05, "loss": 0.0398, "step": 6600 }, { "grad_norm": 0.403486967086792, "learning_rate": 9.992923766322586e-05, "loss": 0.0395, "step": 6610 }, { "grad_norm": 0.37257328629493713, "learning_rate": 9.99283555591149e-05, "loss": 0.0439, "step": 6620 }, { "grad_norm": 0.4917003810405731, "learning_rate": 9.992746799491404e-05, "loss": 0.0434, "step": 6630 }, { "grad_norm": 0.39685913920402527, "learning_rate": 9.992657497072033e-05, "loss": 0.0436, "step": 6640 }, { "grad_norm": 0.3636392652988434, "learning_rate": 9.992567648663147e-05, "loss": 0.0467, "step": 6650 }, { "grad_norm": 0.4119523763656616, "learning_rate": 9.992477254274568e-05, "loss": 0.0407, "step": 6660 }, { "grad_norm": 0.37879374623298645, "learning_rate": 9.992386313916183e-05, "loss": 0.0411, "step": 6670 }, { "grad_norm": 0.5874691009521484, "learning_rate": 9.992294827597934e-05, "loss": 0.04, "step": 6680 }, { "grad_norm": 0.44693827629089355, "learning_rate": 9.992202795329831e-05, "loss": 0.0415, "step": 6690 }, { "grad_norm": 0.3760095536708832, "learning_rate": 9.992110217121936e-05, "loss": 0.0402, "step": 6700 }, { "grad_norm": 0.30414894223213196, "learning_rate": 9.992017092984372e-05, "loss": 0.0419, "step": 6710 }, { "grad_norm": 0.40662941336631775, "learning_rate": 9.991923422927326e-05, "loss": 0.0451, "step": 6720 }, { "grad_norm": 0.34419509768486023, "learning_rate": 9.991829206961037e-05, "loss": 0.0398, "step": 6730 }, { "grad_norm": 0.37239280343055725, "learning_rate": 9.991734445095813e-05, "loss": 0.0379, "step": 6740 }, { "grad_norm": 0.37669482827186584, "learning_rate": 9.991639137342015e-05, "loss": 0.046, "step": 6750 }, { "grad_norm": 0.3112808167934418, "learning_rate": 9.991543283710064e-05, "loss": 0.0374, "step": 6760 }, { "grad_norm": 0.3077293038368225, "learning_rate": 9.991446884210445e-05, "loss": 0.0419, "step": 6770 }, { "grad_norm": 0.31474313139915466, "learning_rate": 9.9913499388537e-05, "loss": 0.0401, "step": 6780 }, { "grad_norm": 0.37316757440567017, "learning_rate": 9.99125244765043e-05, "loss": 0.0361, "step": 6790 }, { "grad_norm": 0.4476058781147003, "learning_rate": 9.991154410611296e-05, "loss": 0.0387, "step": 6800 }, { "grad_norm": 0.38009148836135864, "learning_rate": 9.99105582774702e-05, "loss": 0.0409, "step": 6810 }, { "grad_norm": 0.4715578556060791, "learning_rate": 9.990956699068384e-05, "loss": 0.0432, "step": 6820 }, { "grad_norm": 0.4372938275337219, "learning_rate": 9.990857024586224e-05, "loss": 0.0408, "step": 6830 }, { "grad_norm": 0.45401811599731445, "learning_rate": 9.990756804311446e-05, "loss": 0.0405, "step": 6840 }, { "grad_norm": 0.3592391610145569, "learning_rate": 9.990656038255006e-05, "loss": 0.0379, "step": 6850 }, { "grad_norm": 0.33883222937583923, "learning_rate": 9.990554726427926e-05, "loss": 0.0363, "step": 6860 }, { "grad_norm": 0.3013467788696289, "learning_rate": 9.990452868841284e-05, "loss": 0.0376, "step": 6870 }, { "grad_norm": 0.30562013387680054, "learning_rate": 9.99035046550622e-05, "loss": 0.0404, "step": 6880 }, { "grad_norm": 0.2772228419780731, "learning_rate": 9.99024751643393e-05, "loss": 0.0373, "step": 6890 }, { "grad_norm": 0.4050418734550476, "learning_rate": 9.990144021635677e-05, "loss": 0.0409, "step": 6900 }, { "grad_norm": 0.3891512453556061, "learning_rate": 9.990039981122775e-05, "loss": 0.0448, "step": 6910 }, { "grad_norm": 0.37614214420318604, "learning_rate": 9.989935394906602e-05, "loss": 0.0441, "step": 6920 }, { "grad_norm": 0.45809170603752136, "learning_rate": 9.989830262998598e-05, "loss": 0.0475, "step": 6930 }, { "grad_norm": 0.3788416087627411, "learning_rate": 9.989724585410259e-05, "loss": 0.0399, "step": 6940 }, { "grad_norm": 0.3665698766708374, "learning_rate": 9.989618362153139e-05, "loss": 0.0438, "step": 6950 }, { "grad_norm": 0.4882635176181793, "learning_rate": 9.989511593238859e-05, "loss": 0.042, "step": 6960 }, { "grad_norm": 0.35360443592071533, "learning_rate": 9.98940427867909e-05, "loss": 0.0404, "step": 6970 }, { "grad_norm": 0.3567802309989929, "learning_rate": 9.989296418485573e-05, "loss": 0.0411, "step": 6980 }, { "grad_norm": 0.27692416310310364, "learning_rate": 9.989188012670101e-05, "loss": 0.0381, "step": 6990 }, { "grad_norm": 0.323670357465744, "learning_rate": 9.989079061244528e-05, "loss": 0.038, "step": 7000 }, { "grad_norm": 0.4186215102672577, "learning_rate": 9.988969564220769e-05, "loss": 0.0385, "step": 7010 }, { "grad_norm": 0.3390868604183197, "learning_rate": 9.988859521610801e-05, "loss": 0.0381, "step": 7020 }, { "grad_norm": 0.4397937059402466, "learning_rate": 9.988748933426656e-05, "loss": 0.0421, "step": 7030 }, { "grad_norm": 0.5886811017990112, "learning_rate": 9.988637799680428e-05, "loss": 0.0451, "step": 7040 }, { "grad_norm": 0.36719822883605957, "learning_rate": 9.98852612038427e-05, "loss": 0.0428, "step": 7050 }, { "grad_norm": 0.42500531673431396, "learning_rate": 9.988413895550397e-05, "loss": 0.0411, "step": 7060 }, { "grad_norm": 0.4673559069633484, "learning_rate": 9.98830112519108e-05, "loss": 0.0481, "step": 7070 }, { "grad_norm": 0.48875316977500916, "learning_rate": 9.98818780931865e-05, "loss": 0.044, "step": 7080 }, { "grad_norm": 0.47799134254455566, "learning_rate": 9.988073947945502e-05, "loss": 0.0429, "step": 7090 }, { "grad_norm": 0.4701554775238037, "learning_rate": 9.987959541084087e-05, "loss": 0.0432, "step": 7100 }, { "grad_norm": 0.37925657629966736, "learning_rate": 9.987844588746915e-05, "loss": 0.0429, "step": 7110 }, { "grad_norm": 0.42091089487075806, "learning_rate": 9.987729090946558e-05, "loss": 0.042, "step": 7120 }, { "grad_norm": 0.3974548280239105, "learning_rate": 9.987613047695647e-05, "loss": 0.0445, "step": 7130 }, { "grad_norm": 0.3176124095916748, "learning_rate": 9.987496459006871e-05, "loss": 0.0378, "step": 7140 }, { "grad_norm": 0.3066345453262329, "learning_rate": 9.987379324892982e-05, "loss": 0.037, "step": 7150 }, { "grad_norm": 0.3308650255203247, "learning_rate": 9.987261645366788e-05, "loss": 0.0409, "step": 7160 }, { "grad_norm": 0.3854585289955139, "learning_rate": 9.987143420441158e-05, "loss": 0.0415, "step": 7170 }, { "grad_norm": 0.3625917136669159, "learning_rate": 9.987024650129022e-05, "loss": 0.0408, "step": 7180 }, { "grad_norm": 0.36707189679145813, "learning_rate": 9.986905334443368e-05, "loss": 0.0417, "step": 7190 }, { "grad_norm": 0.27074235677719116, "learning_rate": 9.986785473397245e-05, "loss": 0.0366, "step": 7200 }, { "grad_norm": 0.30905768275260925, "learning_rate": 9.98666506700376e-05, "loss": 0.0391, "step": 7210 }, { "grad_norm": 0.29661455750465393, "learning_rate": 9.986544115276081e-05, "loss": 0.0356, "step": 7220 }, { "grad_norm": 0.3555152416229248, "learning_rate": 9.986422618227433e-05, "loss": 0.0426, "step": 7230 }, { "grad_norm": 0.4472492039203644, "learning_rate": 9.986300575871106e-05, "loss": 0.043, "step": 7240 }, { "grad_norm": 0.3495745062828064, "learning_rate": 9.986177988220444e-05, "loss": 0.0357, "step": 7250 }, { "grad_norm": 0.35166609287261963, "learning_rate": 9.986054855288856e-05, "loss": 0.037, "step": 7260 }, { "grad_norm": 0.3287995159626007, "learning_rate": 9.985931177089802e-05, "loss": 0.0387, "step": 7270 }, { "grad_norm": 0.3657487630844116, "learning_rate": 9.985806953636814e-05, "loss": 0.0363, "step": 7280 }, { "grad_norm": 0.3720487952232361, "learning_rate": 9.985682184943471e-05, "loss": 0.0391, "step": 7290 }, { "grad_norm": 0.3987261652946472, "learning_rate": 9.98555687102342e-05, "loss": 0.0367, "step": 7300 }, { "grad_norm": 0.3779740333557129, "learning_rate": 9.985431011890367e-05, "loss": 0.0369, "step": 7310 }, { "grad_norm": 0.37092921137809753, "learning_rate": 9.985304607558075e-05, "loss": 0.0403, "step": 7320 }, { "grad_norm": 0.3640802800655365, "learning_rate": 9.985177658040364e-05, "loss": 0.0362, "step": 7330 }, { "grad_norm": 0.2470647692680359, "learning_rate": 9.985050163351119e-05, "loss": 0.034, "step": 7340 }, { "grad_norm": 0.4090140163898468, "learning_rate": 9.984922123504286e-05, "loss": 0.0357, "step": 7350 }, { "grad_norm": 0.40210703015327454, "learning_rate": 9.984793538513862e-05, "loss": 0.0462, "step": 7360 }, { "grad_norm": 0.401187002658844, "learning_rate": 9.984664408393912e-05, "loss": 0.0416, "step": 7370 }, { "grad_norm": 0.2860203683376312, "learning_rate": 9.984534733158556e-05, "loss": 0.0376, "step": 7380 }, { "grad_norm": 0.4811350107192993, "learning_rate": 9.984404512821977e-05, "loss": 0.0398, "step": 7390 }, { "grad_norm": 0.3572821021080017, "learning_rate": 9.984273747398411e-05, "loss": 0.0379, "step": 7400 }, { "grad_norm": 0.37029004096984863, "learning_rate": 9.984142436902165e-05, "loss": 0.0363, "step": 7410 }, { "grad_norm": 0.42745885252952576, "learning_rate": 9.984010581347596e-05, "loss": 0.0376, "step": 7420 }, { "grad_norm": 0.34809520840644836, "learning_rate": 9.983878180749121e-05, "loss": 0.0373, "step": 7430 }, { "grad_norm": 0.4131905734539032, "learning_rate": 9.983745235121222e-05, "loss": 0.0386, "step": 7440 }, { "grad_norm": 0.36994072794914246, "learning_rate": 9.983611744478438e-05, "loss": 0.0353, "step": 7450 }, { "grad_norm": 0.4677395522594452, "learning_rate": 9.983477708835365e-05, "loss": 0.0398, "step": 7460 }, { "grad_norm": 0.4013541340827942, "learning_rate": 9.983343128206664e-05, "loss": 0.037, "step": 7470 }, { "grad_norm": 0.3050335943698883, "learning_rate": 9.983208002607049e-05, "loss": 0.0339, "step": 7480 }, { "grad_norm": 0.30793461203575134, "learning_rate": 9.9830723320513e-05, "loss": 0.0322, "step": 7490 }, { "grad_norm": 0.2839100956916809, "learning_rate": 9.982936116554254e-05, "loss": 0.0393, "step": 7500 }, { "grad_norm": 0.3169781565666199, "learning_rate": 9.982799356130803e-05, "loss": 0.0345, "step": 7510 }, { "grad_norm": 0.3204275369644165, "learning_rate": 9.982662050795908e-05, "loss": 0.0372, "step": 7520 }, { "grad_norm": 0.3425615429878235, "learning_rate": 9.982524200564583e-05, "loss": 0.0358, "step": 7530 }, { "grad_norm": 0.36306801438331604, "learning_rate": 9.982385805451901e-05, "loss": 0.036, "step": 7540 }, { "grad_norm": 0.40491846203804016, "learning_rate": 9.982246865472998e-05, "loss": 0.0363, "step": 7550 }, { "grad_norm": 0.26186713576316833, "learning_rate": 9.982107380643069e-05, "loss": 0.0409, "step": 7560 }, { "grad_norm": 0.38968366384506226, "learning_rate": 9.981967350977368e-05, "loss": 0.0348, "step": 7570 }, { "grad_norm": 0.38742193579673767, "learning_rate": 9.981826776491208e-05, "loss": 0.0346, "step": 7580 }, { "grad_norm": 0.35307925939559937, "learning_rate": 9.98168565719996e-05, "loss": 0.0357, "step": 7590 }, { "grad_norm": 0.3132733106613159, "learning_rate": 9.98154399311906e-05, "loss": 0.0368, "step": 7600 }, { "grad_norm": 0.3508443832397461, "learning_rate": 9.981401784263997e-05, "loss": 0.0359, "step": 7610 }, { "grad_norm": 0.4050815999507904, "learning_rate": 9.981259030650326e-05, "loss": 0.0398, "step": 7620 }, { "grad_norm": 0.368773877620697, "learning_rate": 9.981115732293655e-05, "loss": 0.0418, "step": 7630 }, { "grad_norm": 0.46134522557258606, "learning_rate": 9.980971889209659e-05, "loss": 0.0425, "step": 7640 }, { "grad_norm": 0.37935248017311096, "learning_rate": 9.980827501414064e-05, "loss": 0.0419, "step": 7650 }, { "grad_norm": 0.345134437084198, "learning_rate": 9.980682568922663e-05, "loss": 0.0463, "step": 7660 }, { "grad_norm": 0.45526233315467834, "learning_rate": 9.980537091751304e-05, "loss": 0.0416, "step": 7670 }, { "grad_norm": 0.44727104902267456, "learning_rate": 9.980391069915897e-05, "loss": 0.0393, "step": 7680 }, { "grad_norm": 0.3317968547344208, "learning_rate": 9.98024450343241e-05, "loss": 0.0391, "step": 7690 }, { "grad_norm": 0.33676543831825256, "learning_rate": 9.980097392316872e-05, "loss": 0.0443, "step": 7700 }, { "grad_norm": 0.35649561882019043, "learning_rate": 9.97994973658537e-05, "loss": 0.0433, "step": 7710 }, { "grad_norm": 0.4213303029537201, "learning_rate": 9.979801536254054e-05, "loss": 0.0362, "step": 7720 }, { "grad_norm": 0.39798110723495483, "learning_rate": 9.979652791339127e-05, "loss": 0.0393, "step": 7730 }, { "grad_norm": 0.3370610773563385, "learning_rate": 9.97950350185686e-05, "loss": 0.0353, "step": 7740 }, { "grad_norm": 0.4083384573459625, "learning_rate": 9.979353667823574e-05, "loss": 0.0366, "step": 7750 }, { "grad_norm": 0.399888813495636, "learning_rate": 9.979203289255658e-05, "loss": 0.0384, "step": 7760 }, { "grad_norm": 0.34757864475250244, "learning_rate": 9.979052366169557e-05, "loss": 0.0457, "step": 7770 }, { "grad_norm": 0.38708895444869995, "learning_rate": 9.978900898581775e-05, "loss": 0.0383, "step": 7780 }, { "grad_norm": 0.2952060401439667, "learning_rate": 9.978748886508875e-05, "loss": 0.0352, "step": 7790 }, { "grad_norm": 0.37674736976623535, "learning_rate": 9.978596329967484e-05, "loss": 0.0385, "step": 7800 }, { "grad_norm": 0.28590720891952515, "learning_rate": 9.978443228974284e-05, "loss": 0.0355, "step": 7810 }, { "grad_norm": 0.33124130964279175, "learning_rate": 9.978289583546015e-05, "loss": 0.0345, "step": 7820 }, { "grad_norm": 0.3541032671928406, "learning_rate": 9.978135393699484e-05, "loss": 0.0381, "step": 7830 }, { "grad_norm": 0.3403785526752472, "learning_rate": 9.977980659451548e-05, "loss": 0.0354, "step": 7840 }, { "grad_norm": 0.38421159982681274, "learning_rate": 9.977825380819135e-05, "loss": 0.0358, "step": 7850 }, { "grad_norm": 0.3698849678039551, "learning_rate": 9.97766955781922e-05, "loss": 0.0369, "step": 7860 }, { "grad_norm": 0.33067119121551514, "learning_rate": 9.977513190468848e-05, "loss": 0.0365, "step": 7870 }, { "grad_norm": 0.35419970750808716, "learning_rate": 9.977356278785116e-05, "loss": 0.0364, "step": 7880 }, { "grad_norm": 0.3208509683609009, "learning_rate": 9.977198822785184e-05, "loss": 0.0379, "step": 7890 }, { "grad_norm": 0.3492433428764343, "learning_rate": 9.977040822486273e-05, "loss": 0.0358, "step": 7900 }, { "grad_norm": 0.310691237449646, "learning_rate": 9.97688227790566e-05, "loss": 0.0355, "step": 7910 }, { "grad_norm": 0.28883108496665955, "learning_rate": 9.976723189060684e-05, "loss": 0.034, "step": 7920 }, { "grad_norm": 0.3978407084941864, "learning_rate": 9.976563555968742e-05, "loss": 0.0351, "step": 7930 }, { "grad_norm": 0.3962571322917938, "learning_rate": 9.976403378647292e-05, "loss": 0.0376, "step": 7940 }, { "grad_norm": 0.36237773299217224, "learning_rate": 9.97624265711385e-05, "loss": 0.0355, "step": 7950 }, { "grad_norm": 0.4317905306816101, "learning_rate": 9.976081391385993e-05, "loss": 0.0416, "step": 7960 }, { "grad_norm": 0.3746381103992462, "learning_rate": 9.975919581481356e-05, "loss": 0.0446, "step": 7970 }, { "grad_norm": 0.340983510017395, "learning_rate": 9.975757227417634e-05, "loss": 0.0437, "step": 7980 }, { "grad_norm": 0.4021090865135193, "learning_rate": 9.975594329212586e-05, "loss": 0.045, "step": 7990 }, { "grad_norm": 0.32568812370300293, "learning_rate": 9.97543088688402e-05, "loss": 0.0418, "step": 8000 }, { "grad_norm": 0.3566283881664276, "learning_rate": 9.975266900449814e-05, "loss": 0.039, "step": 8010 }, { "grad_norm": 0.32030782103538513, "learning_rate": 9.975102369927898e-05, "loss": 0.0397, "step": 8020 }, { "grad_norm": 0.2850058972835541, "learning_rate": 9.974937295336269e-05, "loss": 0.0357, "step": 8030 }, { "grad_norm": 0.3288343846797943, "learning_rate": 9.974771676692975e-05, "loss": 0.039, "step": 8040 }, { "grad_norm": 0.426302433013916, "learning_rate": 9.974605514016131e-05, "loss": 0.0498, "step": 8050 }, { "grad_norm": 0.434253066778183, "learning_rate": 9.974438807323907e-05, "loss": 0.0502, "step": 8060 }, { "grad_norm": 0.38354426622390747, "learning_rate": 9.974271556634535e-05, "loss": 0.044, "step": 8070 }, { "grad_norm": 0.34531429409980774, "learning_rate": 9.974103761966302e-05, "loss": 0.0445, "step": 8080 }, { "grad_norm": 0.42348185181617737, "learning_rate": 9.973935423337563e-05, "loss": 0.0438, "step": 8090 }, { "grad_norm": 0.3048744797706604, "learning_rate": 9.973766540766722e-05, "loss": 0.0417, "step": 8100 }, { "grad_norm": 0.40481072664260864, "learning_rate": 9.97359711427225e-05, "loss": 0.0432, "step": 8110 }, { "grad_norm": 0.37501201033592224, "learning_rate": 9.973427143872677e-05, "loss": 0.042, "step": 8120 }, { "grad_norm": 0.3102879822254181, "learning_rate": 9.973256629586589e-05, "loss": 0.0413, "step": 8130 }, { "grad_norm": 0.36110785603523254, "learning_rate": 9.973085571432632e-05, "loss": 0.0372, "step": 8140 }, { "grad_norm": 0.4225434958934784, "learning_rate": 9.972913969429513e-05, "loss": 0.0414, "step": 8150 }, { "grad_norm": 0.3022976517677307, "learning_rate": 9.972741823596e-05, "loss": 0.0392, "step": 8160 }, { "grad_norm": 0.3167498707771301, "learning_rate": 9.972569133950917e-05, "loss": 0.038, "step": 8170 }, { "grad_norm": 0.40276527404785156, "learning_rate": 9.972395900513151e-05, "loss": 0.0376, "step": 8180 }, { "grad_norm": 0.32671651244163513, "learning_rate": 9.972222123301645e-05, "loss": 0.0373, "step": 8190 }, { "grad_norm": 0.3896413743495941, "learning_rate": 9.972047802335403e-05, "loss": 0.0391, "step": 8200 }, { "grad_norm": 0.3608238697052002, "learning_rate": 9.971872937633488e-05, "loss": 0.0369, "step": 8210 }, { "grad_norm": 0.4188782870769501, "learning_rate": 9.971697529215024e-05, "loss": 0.039, "step": 8220 }, { "grad_norm": 0.35578981041908264, "learning_rate": 9.971521577099192e-05, "loss": 0.0367, "step": 8230 }, { "grad_norm": 0.3127240836620331, "learning_rate": 9.971345081305236e-05, "loss": 0.0386, "step": 8240 }, { "grad_norm": 0.3173069655895233, "learning_rate": 9.971168041852456e-05, "loss": 0.0369, "step": 8250 }, { "grad_norm": 0.6071885824203491, "learning_rate": 9.970990458760215e-05, "loss": 0.0373, "step": 8260 }, { "grad_norm": 0.3957213759422302, "learning_rate": 9.970812332047929e-05, "loss": 0.042, "step": 8270 }, { "grad_norm": 0.3203412592411041, "learning_rate": 9.97063366173508e-05, "loss": 0.0417, "step": 8280 }, { "grad_norm": 0.36269858479499817, "learning_rate": 9.970454447841207e-05, "loss": 0.0417, "step": 8290 }, { "grad_norm": 0.4221937656402588, "learning_rate": 9.970274690385909e-05, "loss": 0.0385, "step": 8300 }, { "grad_norm": 0.4388277232646942, "learning_rate": 9.970094389388844e-05, "loss": 0.0454, "step": 8310 }, { "grad_norm": 0.2926861643791199, "learning_rate": 9.969913544869728e-05, "loss": 0.041, "step": 8320 }, { "grad_norm": 0.4013611674308777, "learning_rate": 9.96973215684834e-05, "loss": 0.0429, "step": 8330 }, { "grad_norm": 0.3635760545730591, "learning_rate": 9.969550225344513e-05, "loss": 0.0364, "step": 8340 }, { "grad_norm": 0.369133323431015, "learning_rate": 9.969367750378147e-05, "loss": 0.0398, "step": 8350 }, { "grad_norm": 0.3521933853626251, "learning_rate": 9.969184731969194e-05, "loss": 0.0405, "step": 8360 }, { "grad_norm": 0.34550580382347107, "learning_rate": 9.96900117013767e-05, "loss": 0.0359, "step": 8370 }, { "grad_norm": 0.3099754750728607, "learning_rate": 9.96881706490365e-05, "loss": 0.036, "step": 8380 }, { "grad_norm": 0.29338470101356506, "learning_rate": 9.968632416287265e-05, "loss": 0.0383, "step": 8390 }, { "grad_norm": 0.3571929931640625, "learning_rate": 9.96844722430871e-05, "loss": 0.0391, "step": 8400 }, { "grad_norm": 0.3742605447769165, "learning_rate": 9.968261488988235e-05, "loss": 0.0391, "step": 8410 }, { "grad_norm": 0.4118841588497162, "learning_rate": 9.968075210346155e-05, "loss": 0.0366, "step": 8420 }, { "grad_norm": 0.3805902898311615, "learning_rate": 9.967888388402839e-05, "loss": 0.0357, "step": 8430 }, { "grad_norm": 0.37633606791496277, "learning_rate": 9.967701023178717e-05, "loss": 0.0393, "step": 8440 }, { "grad_norm": 0.4364272952079773, "learning_rate": 9.967513114694282e-05, "loss": 0.0445, "step": 8450 }, { "grad_norm": 0.33562126755714417, "learning_rate": 9.967324662970079e-05, "loss": 0.0354, "step": 8460 }, { "grad_norm": 0.2914096713066101, "learning_rate": 9.96713566802672e-05, "loss": 0.0339, "step": 8470 }, { "grad_norm": 0.3154333233833313, "learning_rate": 9.966946129884873e-05, "loss": 0.0371, "step": 8480 }, { "grad_norm": 0.31936290860176086, "learning_rate": 9.966756048565265e-05, "loss": 0.0361, "step": 8490 }, { "grad_norm": 0.2649095356464386, "learning_rate": 9.966565424088681e-05, "loss": 0.0406, "step": 8500 }, { "grad_norm": 0.30959999561309814, "learning_rate": 9.96637425647597e-05, "loss": 0.0362, "step": 8510 }, { "grad_norm": 0.30532306432724, "learning_rate": 9.966182545748038e-05, "loss": 0.0348, "step": 8520 }, { "grad_norm": 0.30468493700027466, "learning_rate": 9.96599029192585e-05, "loss": 0.0331, "step": 8530 }, { "grad_norm": 0.44197046756744385, "learning_rate": 9.965797495030428e-05, "loss": 0.0395, "step": 8540 }, { "grad_norm": 0.31918153166770935, "learning_rate": 9.96560415508286e-05, "loss": 0.0364, "step": 8550 }, { "grad_norm": 0.3919902443885803, "learning_rate": 9.965410272104286e-05, "loss": 0.0376, "step": 8560 }, { "grad_norm": 0.325273722410202, "learning_rate": 9.96521584611591e-05, "loss": 0.0435, "step": 8570 }, { "grad_norm": 0.3456231951713562, "learning_rate": 9.965020877138994e-05, "loss": 0.0371, "step": 8580 }, { "grad_norm": 0.33465850353240967, "learning_rate": 9.964825365194861e-05, "loss": 0.0352, "step": 8590 }, { "grad_norm": 0.30921387672424316, "learning_rate": 9.96462931030489e-05, "loss": 0.0332, "step": 8600 }, { "grad_norm": 0.2752140164375305, "learning_rate": 9.96443271249052e-05, "loss": 0.0357, "step": 8610 }, { "grad_norm": 0.29347705841064453, "learning_rate": 9.964235571773255e-05, "loss": 0.0325, "step": 8620 }, { "grad_norm": 0.2942414879798889, "learning_rate": 9.96403788817465e-05, "loss": 0.0366, "step": 8630 }, { "grad_norm": 0.3742513060569763, "learning_rate": 9.963839661716325e-05, "loss": 0.0351, "step": 8640 }, { "grad_norm": 0.35027018189430237, "learning_rate": 9.963640892419958e-05, "loss": 0.0348, "step": 8650 }, { "grad_norm": 0.38498130440711975, "learning_rate": 9.963441580307286e-05, "loss": 0.0356, "step": 8660 }, { "grad_norm": 0.47406965494155884, "learning_rate": 9.963241725400104e-05, "loss": 0.0495, "step": 8670 }, { "grad_norm": 0.4779032766819, "learning_rate": 9.963041327720271e-05, "loss": 0.0437, "step": 8680 }, { "grad_norm": 0.9222510457038879, "learning_rate": 9.962840387289697e-05, "loss": 0.0384, "step": 8690 }, { "grad_norm": 0.48106569051742554, "learning_rate": 9.962638904130363e-05, "loss": 0.0375, "step": 8700 }, { "grad_norm": 0.3867816627025604, "learning_rate": 9.962436878264298e-05, "loss": 0.0435, "step": 8710 }, { "grad_norm": 0.38184475898742676, "learning_rate": 9.962234309713598e-05, "loss": 0.0424, "step": 8720 }, { "grad_norm": 0.443427175283432, "learning_rate": 9.962031198500414e-05, "loss": 0.0386, "step": 8730 }, { "grad_norm": 0.40488898754119873, "learning_rate": 9.961827544646958e-05, "loss": 0.0388, "step": 8740 }, { "grad_norm": 0.3284037411212921, "learning_rate": 9.961623348175501e-05, "loss": 0.0373, "step": 8750 }, { "grad_norm": 0.31114766001701355, "learning_rate": 9.961418609108377e-05, "loss": 0.0348, "step": 8760 }, { "grad_norm": 0.3138147294521332, "learning_rate": 9.961213327467971e-05, "loss": 0.038, "step": 8770 }, { "grad_norm": 0.2432270050048828, "learning_rate": 9.961007503276736e-05, "loss": 0.0365, "step": 8780 }, { "grad_norm": 0.2557448446750641, "learning_rate": 9.960801136557179e-05, "loss": 0.0354, "step": 8790 }, { "grad_norm": 0.3617704510688782, "learning_rate": 9.960594227331866e-05, "loss": 0.0328, "step": 8800 }, { "grad_norm": 0.3344348669052124, "learning_rate": 9.960386775623429e-05, "loss": 0.0319, "step": 8810 }, { "grad_norm": 0.38413378596305847, "learning_rate": 9.96017878145455e-05, "loss": 0.0404, "step": 8820 }, { "grad_norm": 0.3654346466064453, "learning_rate": 9.959970244847977e-05, "loss": 0.0359, "step": 8830 }, { "grad_norm": 0.3626062870025635, "learning_rate": 9.959761165826518e-05, "loss": 0.0455, "step": 8840 }, { "grad_norm": 0.3835219144821167, "learning_rate": 9.959551544413033e-05, "loss": 0.0462, "step": 8850 }, { "grad_norm": 0.2980332374572754, "learning_rate": 9.959341380630448e-05, "loss": 0.039, "step": 8860 }, { "grad_norm": 0.3447471559047699, "learning_rate": 9.959130674501746e-05, "loss": 0.0379, "step": 8870 }, { "grad_norm": 0.280758798122406, "learning_rate": 9.958919426049968e-05, "loss": 0.0367, "step": 8880 }, { "grad_norm": 0.42634084820747375, "learning_rate": 9.958707635298219e-05, "loss": 0.0408, "step": 8890 }, { "grad_norm": 0.2963109314441681, "learning_rate": 9.958495302269657e-05, "loss": 0.0377, "step": 8900 }, { "grad_norm": 0.3270301818847656, "learning_rate": 9.958282426987503e-05, "loss": 0.0384, "step": 8910 }, { "grad_norm": 0.4291597008705139, "learning_rate": 9.95806900947504e-05, "loss": 0.0381, "step": 8920 }, { "grad_norm": 0.3275458812713623, "learning_rate": 9.957855049755604e-05, "loss": 0.0415, "step": 8930 }, { "grad_norm": 0.3470776379108429, "learning_rate": 9.957640547852593e-05, "loss": 0.036, "step": 8940 }, { "grad_norm": 0.3215741217136383, "learning_rate": 9.957425503789466e-05, "loss": 0.0353, "step": 8950 }, { "grad_norm": 0.34001925587654114, "learning_rate": 9.957209917589738e-05, "loss": 0.0343, "step": 8960 }, { "grad_norm": 0.4093154966831207, "learning_rate": 9.956993789276987e-05, "loss": 0.0383, "step": 8970 }, { "grad_norm": 0.36916446685791016, "learning_rate": 9.956777118874847e-05, "loss": 0.0363, "step": 8980 }, { "grad_norm": 0.2599657475948334, "learning_rate": 9.956559906407016e-05, "loss": 0.0382, "step": 8990 }, { "grad_norm": 0.34207358956336975, "learning_rate": 9.956342151897245e-05, "loss": 0.0364, "step": 9000 }, { "grad_norm": 0.29294824600219727, "learning_rate": 9.956123855369346e-05, "loss": 0.0371, "step": 9010 }, { "grad_norm": 0.28708624839782715, "learning_rate": 9.955905016847196e-05, "loss": 0.0349, "step": 9020 }, { "grad_norm": 0.45953887701034546, "learning_rate": 9.955685636354723e-05, "loss": 0.037, "step": 9030 }, { "grad_norm": 0.3637121319770813, "learning_rate": 9.95546571391592e-05, "loss": 0.0337, "step": 9040 }, { "grad_norm": 0.33417582511901855, "learning_rate": 9.955245249554837e-05, "loss": 0.0328, "step": 9050 }, { "grad_norm": 0.3595345616340637, "learning_rate": 9.955024243295582e-05, "loss": 0.0366, "step": 9060 }, { "grad_norm": 0.38165169954299927, "learning_rate": 9.954802695162328e-05, "loss": 0.0369, "step": 9070 }, { "grad_norm": 0.3333317041397095, "learning_rate": 9.954580605179302e-05, "loss": 0.0365, "step": 9080 }, { "grad_norm": 0.34941139817237854, "learning_rate": 9.954357973370788e-05, "loss": 0.034, "step": 9090 }, { "grad_norm": 0.3329305648803711, "learning_rate": 9.954134799761135e-05, "loss": 0.0356, "step": 9100 }, { "grad_norm": 0.3539453446865082, "learning_rate": 9.953911084374748e-05, "loss": 0.0322, "step": 9110 }, { "grad_norm": 0.2939789891242981, "learning_rate": 9.953686827236093e-05, "loss": 0.0342, "step": 9120 }, { "grad_norm": 0.35512256622314453, "learning_rate": 9.953462028369695e-05, "loss": 0.0352, "step": 9130 }, { "grad_norm": 0.3332633972167969, "learning_rate": 9.953236687800136e-05, "loss": 0.0319, "step": 9140 }, { "grad_norm": 0.37493473291397095, "learning_rate": 9.95301080555206e-05, "loss": 0.0378, "step": 9150 }, { "grad_norm": 0.28818580508232117, "learning_rate": 9.952784381650171e-05, "loss": 0.0321, "step": 9160 }, { "grad_norm": 0.32731640338897705, "learning_rate": 9.952557416119226e-05, "loss": 0.0385, "step": 9170 }, { "grad_norm": 0.32316359877586365, "learning_rate": 9.95232990898405e-05, "loss": 0.0366, "step": 9180 }, { "grad_norm": 0.2593715190887451, "learning_rate": 9.95210186026952e-05, "loss": 0.036, "step": 9190 }, { "grad_norm": 0.32954204082489014, "learning_rate": 9.951873270000576e-05, "loss": 0.0352, "step": 9200 }, { "grad_norm": 0.3812485933303833, "learning_rate": 9.951644138202216e-05, "loss": 0.037, "step": 9210 }, { "grad_norm": 0.2747339904308319, "learning_rate": 9.951414464899498e-05, "loss": 0.0371, "step": 9220 }, { "grad_norm": 0.3944198787212372, "learning_rate": 9.951184250117538e-05, "loss": 0.0368, "step": 9230 }, { "grad_norm": 0.36068013310432434, "learning_rate": 9.950953493881513e-05, "loss": 0.0345, "step": 9240 }, { "grad_norm": 0.3201057016849518, "learning_rate": 9.950722196216658e-05, "loss": 0.0369, "step": 9250 }, { "grad_norm": 0.38178539276123047, "learning_rate": 9.950490357148265e-05, "loss": 0.0394, "step": 9260 }, { "grad_norm": 0.2571437656879425, "learning_rate": 9.950257976701692e-05, "loss": 0.0388, "step": 9270 }, { "grad_norm": 0.283446729183197, "learning_rate": 9.950025054902348e-05, "loss": 0.0368, "step": 9280 }, { "grad_norm": 0.3338736891746521, "learning_rate": 9.949791591775706e-05, "loss": 0.0342, "step": 9290 }, { "grad_norm": 0.4056965708732605, "learning_rate": 9.949557587347298e-05, "loss": 0.0369, "step": 9300 }, { "grad_norm": 0.4688659608364105, "learning_rate": 9.949323041642713e-05, "loss": 0.0333, "step": 9310 }, { "grad_norm": 0.3645583391189575, "learning_rate": 9.949087954687602e-05, "loss": 0.0328, "step": 9320 }, { "grad_norm": 0.32847312092781067, "learning_rate": 9.948852326507672e-05, "loss": 0.0347, "step": 9330 }, { "grad_norm": 0.26625528931617737, "learning_rate": 9.948616157128694e-05, "loss": 0.0376, "step": 9340 }, { "grad_norm": 0.34900861978530884, "learning_rate": 9.948379446576493e-05, "loss": 0.0352, "step": 9350 }, { "grad_norm": 0.2740466892719269, "learning_rate": 9.948142194876952e-05, "loss": 0.0355, "step": 9360 }, { "grad_norm": 0.2750706374645233, "learning_rate": 9.947904402056024e-05, "loss": 0.0306, "step": 9370 }, { "grad_norm": 0.2410459667444229, "learning_rate": 9.947666068139708e-05, "loss": 0.0311, "step": 9380 }, { "grad_norm": 0.4686475396156311, "learning_rate": 9.947427193154071e-05, "loss": 0.0318, "step": 9390 }, { "grad_norm": 0.3232269287109375, "learning_rate": 9.947187777125233e-05, "loss": 0.0321, "step": 9400 }, { "grad_norm": 0.25719785690307617, "learning_rate": 9.946947820079377e-05, "loss": 0.0333, "step": 9410 }, { "grad_norm": 0.32380932569503784, "learning_rate": 9.946707322042747e-05, "loss": 0.0286, "step": 9420 }, { "grad_norm": 0.37362805008888245, "learning_rate": 9.94646628304164e-05, "loss": 0.0305, "step": 9430 }, { "grad_norm": 0.2921845316886902, "learning_rate": 9.946224703102418e-05, "loss": 0.0322, "step": 9440 }, { "grad_norm": 0.3241710960865021, "learning_rate": 9.945982582251498e-05, "loss": 0.0311, "step": 9450 }, { "grad_norm": 0.32611921429634094, "learning_rate": 9.94573992051536e-05, "loss": 0.0333, "step": 9460 }, { "grad_norm": 0.24411600828170776, "learning_rate": 9.94549671792054e-05, "loss": 0.0323, "step": 9470 }, { "grad_norm": 0.22718602418899536, "learning_rate": 9.945252974493635e-05, "loss": 0.0326, "step": 9480 }, { "grad_norm": 0.40117084980010986, "learning_rate": 9.9450086902613e-05, "loss": 0.0354, "step": 9490 }, { "grad_norm": 0.35078710317611694, "learning_rate": 9.944763865250248e-05, "loss": 0.0354, "step": 9500 }, { "grad_norm": 0.3720484673976898, "learning_rate": 9.944518499487254e-05, "loss": 0.0354, "step": 9510 }, { "grad_norm": 0.388527512550354, "learning_rate": 9.944272592999151e-05, "loss": 0.0369, "step": 9520 }, { "grad_norm": 0.30234429240226746, "learning_rate": 9.94402614581283e-05, "loss": 0.0316, "step": 9530 }, { "grad_norm": 0.3057595491409302, "learning_rate": 9.943779157955244e-05, "loss": 0.0343, "step": 9540 }, { "grad_norm": 0.30402788519859314, "learning_rate": 9.943531629453403e-05, "loss": 0.032, "step": 9550 }, { "grad_norm": 0.266257643699646, "learning_rate": 9.943283560334375e-05, "loss": 0.0303, "step": 9560 }, { "grad_norm": 0.29738351702690125, "learning_rate": 9.943034950625288e-05, "loss": 0.0335, "step": 9570 }, { "grad_norm": 0.3179686367511749, "learning_rate": 9.942785800353332e-05, "loss": 0.0366, "step": 9580 }, { "grad_norm": 0.28078991174697876, "learning_rate": 9.942536109545751e-05, "loss": 0.0331, "step": 9590 }, { "grad_norm": 0.28731569647789, "learning_rate": 9.942285878229853e-05, "loss": 0.0326, "step": 9600 }, { "grad_norm": 0.5066277980804443, "learning_rate": 9.942035106433001e-05, "loss": 0.0338, "step": 9610 }, { "grad_norm": 0.3594561219215393, "learning_rate": 9.94178379418262e-05, "loss": 0.0343, "step": 9620 }, { "grad_norm": 0.3230012357234955, "learning_rate": 9.941531941506194e-05, "loss": 0.0352, "step": 9630 }, { "grad_norm": 0.32053256034851074, "learning_rate": 9.941279548431263e-05, "loss": 0.0339, "step": 9640 }, { "grad_norm": 0.3391490578651428, "learning_rate": 9.941026614985431e-05, "loss": 0.0324, "step": 9650 }, { "grad_norm": 0.36488357186317444, "learning_rate": 9.940773141196357e-05, "loss": 0.0298, "step": 9660 }, { "grad_norm": 0.3177427053451538, "learning_rate": 9.94051912709176e-05, "loss": 0.0325, "step": 9670 }, { "grad_norm": 0.3427729606628418, "learning_rate": 9.940264572699421e-05, "loss": 0.0299, "step": 9680 }, { "grad_norm": 0.34901517629623413, "learning_rate": 9.940009478047174e-05, "loss": 0.0313, "step": 9690 }, { "grad_norm": 0.28164124488830566, "learning_rate": 9.939753843162918e-05, "loss": 0.032, "step": 9700 }, { "grad_norm": 0.26857128739356995, "learning_rate": 9.939497668074609e-05, "loss": 0.0319, "step": 9710 }, { "grad_norm": 0.33862999081611633, "learning_rate": 9.93924095281026e-05, "loss": 0.0328, "step": 9720 }, { "grad_norm": 0.2928948402404785, "learning_rate": 9.938983697397948e-05, "loss": 0.0295, "step": 9730 }, { "grad_norm": 0.39114782214164734, "learning_rate": 9.938725901865805e-05, "loss": 0.0373, "step": 9740 }, { "grad_norm": 0.3667900562286377, "learning_rate": 9.93846756624202e-05, "loss": 0.0352, "step": 9750 }, { "grad_norm": 0.3345792591571808, "learning_rate": 9.938208690554849e-05, "loss": 0.0368, "step": 9760 }, { "grad_norm": 0.3453544080257416, "learning_rate": 9.9379492748326e-05, "loss": 0.038, "step": 9770 }, { "grad_norm": 0.29146498441696167, "learning_rate": 9.937689319103641e-05, "loss": 0.0375, "step": 9780 }, { "grad_norm": 0.34296727180480957, "learning_rate": 9.937428823396404e-05, "loss": 0.0378, "step": 9790 }, { "grad_norm": 0.246039018034935, "learning_rate": 9.937167787739372e-05, "loss": 0.0312, "step": 9800 }, { "grad_norm": 0.3207804560661316, "learning_rate": 9.936906212161095e-05, "loss": 0.0302, "step": 9810 }, { "grad_norm": 0.36071428656578064, "learning_rate": 9.936644096690176e-05, "loss": 0.0342, "step": 9820 }, { "grad_norm": 0.30326923727989197, "learning_rate": 9.936381441355282e-05, "loss": 0.0339, "step": 9830 }, { "grad_norm": 0.29173949360847473, "learning_rate": 9.936118246185136e-05, "loss": 0.0326, "step": 9840 }, { "grad_norm": 0.40712273120880127, "learning_rate": 9.935854511208518e-05, "loss": 0.0364, "step": 9850 }, { "grad_norm": 0.2747460603713989, "learning_rate": 9.935590236454272e-05, "loss": 0.0323, "step": 9860 }, { "grad_norm": 0.2912367880344391, "learning_rate": 9.935325421951298e-05, "loss": 0.0357, "step": 9870 }, { "grad_norm": 0.2978017330169678, "learning_rate": 9.935060067728557e-05, "loss": 0.0334, "step": 9880 }, { "grad_norm": 0.33327674865722656, "learning_rate": 9.934794173815067e-05, "loss": 0.0361, "step": 9890 }, { "grad_norm": 0.4101509749889374, "learning_rate": 9.934527740239906e-05, "loss": 0.034, "step": 9900 }, { "grad_norm": 0.33305782079696655, "learning_rate": 9.934260767032209e-05, "loss": 0.0316, "step": 9910 }, { "grad_norm": 0.39365455508232117, "learning_rate": 9.933993254221172e-05, "loss": 0.0336, "step": 9920 }, { "grad_norm": 0.3521905541419983, "learning_rate": 9.933725201836053e-05, "loss": 0.0347, "step": 9930 }, { "grad_norm": 0.3184158205986023, "learning_rate": 9.933456609906162e-05, "loss": 0.0318, "step": 9940 }, { "grad_norm": 0.3135952651500702, "learning_rate": 9.933187478460875e-05, "loss": 0.0311, "step": 9950 }, { "grad_norm": 0.26484426856040955, "learning_rate": 9.93291780752962e-05, "loss": 0.0302, "step": 9960 }, { "grad_norm": 0.2928025722503662, "learning_rate": 9.932647597141893e-05, "loss": 0.029, "step": 9970 }, { "grad_norm": 0.35675105452537537, "learning_rate": 9.932376847327239e-05, "loss": 0.0299, "step": 9980 }, { "grad_norm": 0.2400476485490799, "learning_rate": 9.932105558115268e-05, "loss": 0.0295, "step": 9990 }, { "grad_norm": 0.39233291149139404, "learning_rate": 9.931833729535651e-05, "loss": 0.033, "step": 10000 }, { "grad_norm": 0.24066513776779175, "learning_rate": 9.931561361618111e-05, "loss": 0.0307, "step": 10010 }, { "grad_norm": 0.3552831709384918, "learning_rate": 9.931288454392435e-05, "loss": 0.032, "step": 10020 }, { "grad_norm": 0.30845776200294495, "learning_rate": 9.931015007888467e-05, "loss": 0.0317, "step": 10030 }, { "grad_norm": 0.25123292207717896, "learning_rate": 9.930741022136112e-05, "loss": 0.0291, "step": 10040 }, { "grad_norm": 0.32528916001319885, "learning_rate": 9.930466497165333e-05, "loss": 0.0294, "step": 10050 }, { "grad_norm": 0.29371872544288635, "learning_rate": 9.93019143300615e-05, "loss": 0.0298, "step": 10060 }, { "grad_norm": 0.26397377252578735, "learning_rate": 9.929915829688644e-05, "loss": 0.0313, "step": 10070 }, { "grad_norm": 0.33419284224510193, "learning_rate": 9.929639687242955e-05, "loss": 0.033, "step": 10080 }, { "grad_norm": 0.45851945877075195, "learning_rate": 9.929363005699281e-05, "loss": 0.0321, "step": 10090 }, { "grad_norm": 0.32043737173080444, "learning_rate": 9.92908578508788e-05, "loss": 0.0331, "step": 10100 }, { "grad_norm": 0.2783166170120239, "learning_rate": 9.928808025439069e-05, "loss": 0.0334, "step": 10110 }, { "grad_norm": 0.3390862047672272, "learning_rate": 9.928529726783223e-05, "loss": 0.031, "step": 10120 }, { "grad_norm": 0.26752230525016785, "learning_rate": 9.928250889150774e-05, "loss": 0.0325, "step": 10130 }, { "grad_norm": 0.3265356421470642, "learning_rate": 9.92797151257222e-05, "loss": 0.0352, "step": 10140 }, { "grad_norm": 0.3824738562107086, "learning_rate": 9.927691597078108e-05, "loss": 0.0349, "step": 10150 }, { "grad_norm": 0.3332644999027252, "learning_rate": 9.927411142699053e-05, "loss": 0.0336, "step": 10160 }, { "grad_norm": 0.3521289825439453, "learning_rate": 9.927130149465725e-05, "loss": 0.0386, "step": 10170 }, { "grad_norm": 0.32756859064102173, "learning_rate": 9.92684861740885e-05, "loss": 0.0411, "step": 10180 }, { "grad_norm": 0.35266920924186707, "learning_rate": 9.926566546559217e-05, "loss": 0.0375, "step": 10190 }, { "grad_norm": 0.34406888484954834, "learning_rate": 9.926283936947673e-05, "loss": 0.037, "step": 10200 }, { "grad_norm": 0.2548351585865021, "learning_rate": 9.926000788605126e-05, "loss": 0.0352, "step": 10210 }, { "grad_norm": 0.268652081489563, "learning_rate": 9.92571710156254e-05, "loss": 0.0333, "step": 10220 }, { "grad_norm": 0.452004611492157, "learning_rate": 9.925432875850936e-05, "loss": 0.0319, "step": 10230 }, { "grad_norm": 0.369170218706131, "learning_rate": 9.925148111501396e-05, "loss": 0.0333, "step": 10240 }, { "grad_norm": 0.30802246928215027, "learning_rate": 9.924862808545066e-05, "loss": 0.0344, "step": 10250 }, { "grad_norm": 0.3677777051925659, "learning_rate": 9.924576967013141e-05, "loss": 0.0387, "step": 10260 }, { "grad_norm": 0.35358870029449463, "learning_rate": 9.924290586936887e-05, "loss": 0.0355, "step": 10270 }, { "grad_norm": 0.31397247314453125, "learning_rate": 9.924003668347614e-05, "loss": 0.0347, "step": 10280 }, { "grad_norm": 0.2376248687505722, "learning_rate": 9.923716211276704e-05, "loss": 0.0359, "step": 10290 }, { "grad_norm": 0.29578685760498047, "learning_rate": 9.923428215755594e-05, "loss": 0.0334, "step": 10300 }, { "grad_norm": 0.29201072454452515, "learning_rate": 9.923139681815775e-05, "loss": 0.0331, "step": 10310 }, { "grad_norm": 0.2227524369955063, "learning_rate": 9.922850609488801e-05, "loss": 0.0295, "step": 10320 }, { "grad_norm": 0.327105313539505, "learning_rate": 9.922560998806287e-05, "loss": 0.0324, "step": 10330 }, { "grad_norm": 0.3411119878292084, "learning_rate": 9.922270849799905e-05, "loss": 0.0312, "step": 10340 }, { "grad_norm": 0.3219653367996216, "learning_rate": 9.92198016250138e-05, "loss": 0.0334, "step": 10350 }, { "grad_norm": 0.34273070096969604, "learning_rate": 9.921688936942506e-05, "loss": 0.0338, "step": 10360 }, { "grad_norm": 0.23557701706886292, "learning_rate": 9.921397173155129e-05, "loss": 0.0339, "step": 10370 }, { "grad_norm": 0.2720183730125427, "learning_rate": 9.921104871171157e-05, "loss": 0.0295, "step": 10380 }, { "grad_norm": 0.2698567509651184, "learning_rate": 9.920812031022554e-05, "loss": 0.0327, "step": 10390 }, { "grad_norm": 0.2946777939796448, "learning_rate": 9.920518652741348e-05, "loss": 0.0332, "step": 10400 }, { "grad_norm": 0.2911076247692108, "learning_rate": 9.920224736359618e-05, "loss": 0.0313, "step": 10410 }, { "grad_norm": 0.2436981499195099, "learning_rate": 9.91993028190951e-05, "loss": 0.0318, "step": 10420 }, { "grad_norm": 0.23558758199214935, "learning_rate": 9.919635289423222e-05, "loss": 0.0287, "step": 10430 }, { "grad_norm": 0.26465725898742676, "learning_rate": 9.919339758933015e-05, "loss": 0.0315, "step": 10440 }, { "grad_norm": 0.27271705865859985, "learning_rate": 9.919043690471209e-05, "loss": 0.0303, "step": 10450 }, { "grad_norm": 0.2631959021091461, "learning_rate": 9.91874708407018e-05, "loss": 0.0321, "step": 10460 }, { "grad_norm": 0.35559791326522827, "learning_rate": 9.918449939762367e-05, "loss": 0.0305, "step": 10470 }, { "grad_norm": 0.5247363448143005, "learning_rate": 9.91815225758026e-05, "loss": 0.0302, "step": 10480 }, { "grad_norm": 0.34403926134109497, "learning_rate": 9.917854037556419e-05, "loss": 0.0322, "step": 10490 }, { "grad_norm": 0.33544495701789856, "learning_rate": 9.917555279723454e-05, "loss": 0.0355, "step": 10500 }, { "grad_norm": 0.3411572575569153, "learning_rate": 9.917255984114036e-05, "loss": 0.028, "step": 10510 }, { "grad_norm": 0.3140174150466919, "learning_rate": 9.916956150760896e-05, "loss": 0.031, "step": 10520 }, { "grad_norm": 0.6343084573745728, "learning_rate": 9.916655779696826e-05, "loss": 0.0356, "step": 10530 }, { "grad_norm": 0.3670981526374817, "learning_rate": 9.916354870954671e-05, "loss": 0.04, "step": 10540 }, { "grad_norm": 0.3828842341899872, "learning_rate": 9.91605342456734e-05, "loss": 0.0405, "step": 10550 }, { "grad_norm": 0.31018945574760437, "learning_rate": 9.915751440567795e-05, "loss": 0.0385, "step": 10560 }, { "grad_norm": 0.42551836371421814, "learning_rate": 9.915448918989066e-05, "loss": 0.0385, "step": 10570 }, { "grad_norm": 0.2611962556838989, "learning_rate": 9.915145859864232e-05, "loss": 0.0334, "step": 10580 }, { "grad_norm": 0.4193207919597626, "learning_rate": 9.914842263226437e-05, "loss": 0.0381, "step": 10590 }, { "grad_norm": 0.41713747382164, "learning_rate": 9.914538129108882e-05, "loss": 0.0407, "step": 10600 }, { "grad_norm": 0.2642456889152527, "learning_rate": 9.914233457544825e-05, "loss": 0.0357, "step": 10610 }, { "grad_norm": 0.39146319031715393, "learning_rate": 9.913928248567586e-05, "loss": 0.0437, "step": 10620 }, { "grad_norm": 0.31910642981529236, "learning_rate": 9.913622502210542e-05, "loss": 0.0382, "step": 10630 }, { "grad_norm": 0.33256101608276367, "learning_rate": 9.913316218507128e-05, "loss": 0.0393, "step": 10640 }, { "grad_norm": 0.2755177915096283, "learning_rate": 9.91300939749084e-05, "loss": 0.0382, "step": 10650 }, { "grad_norm": 0.34197843074798584, "learning_rate": 9.91270203919523e-05, "loss": 0.0334, "step": 10660 }, { "grad_norm": 0.34795114398002625, "learning_rate": 9.912394143653912e-05, "loss": 0.0314, "step": 10670 }, { "grad_norm": 0.35572922229766846, "learning_rate": 9.912085710900555e-05, "loss": 0.0352, "step": 10680 }, { "grad_norm": 0.38576728105545044, "learning_rate": 9.911776740968892e-05, "loss": 0.0364, "step": 10690 }, { "grad_norm": 0.3273185193538666, "learning_rate": 9.911467233892709e-05, "loss": 0.0345, "step": 10700 }, { "grad_norm": 0.25870177149772644, "learning_rate": 9.911157189705853e-05, "loss": 0.0319, "step": 10710 }, { "grad_norm": 0.23544853925704956, "learning_rate": 9.910846608442229e-05, "loss": 0.0335, "step": 10720 }, { "grad_norm": 0.22508946061134338, "learning_rate": 9.910535490135805e-05, "loss": 0.0314, "step": 10730 }, { "grad_norm": 0.27264872193336487, "learning_rate": 9.910223834820603e-05, "loss": 0.0316, "step": 10740 }, { "grad_norm": 0.32031744718551636, "learning_rate": 9.909911642530703e-05, "loss": 0.0325, "step": 10750 }, { "grad_norm": 0.42184069752693176, "learning_rate": 9.909598913300249e-05, "loss": 0.0341, "step": 10760 }, { "grad_norm": 0.316181480884552, "learning_rate": 9.909285647163438e-05, "loss": 0.0333, "step": 10770 }, { "grad_norm": 0.3553553521633148, "learning_rate": 9.908971844154531e-05, "loss": 0.0287, "step": 10780 }, { "grad_norm": 0.27938687801361084, "learning_rate": 9.908657504307843e-05, "loss": 0.0334, "step": 10790 }, { "grad_norm": 0.36440569162368774, "learning_rate": 9.908342627657751e-05, "loss": 0.0348, "step": 10800 }, { "grad_norm": 0.2794523239135742, "learning_rate": 9.908027214238689e-05, "loss": 0.0318, "step": 10810 }, { "grad_norm": 0.26414674520492554, "learning_rate": 9.90771126408515e-05, "loss": 0.034, "step": 10820 }, { "grad_norm": 0.27125081419944763, "learning_rate": 9.907394777231685e-05, "loss": 0.0321, "step": 10830 }, { "grad_norm": 0.2811761796474457, "learning_rate": 9.907077753712905e-05, "loss": 0.0341, "step": 10840 }, { "grad_norm": 0.42747771739959717, "learning_rate": 9.906760193563482e-05, "loss": 0.0347, "step": 10850 }, { "grad_norm": 0.3871193826198578, "learning_rate": 9.906442096818139e-05, "loss": 0.0296, "step": 10860 }, { "grad_norm": 0.4959660768508911, "learning_rate": 9.906123463511665e-05, "loss": 0.0337, "step": 10870 }, { "grad_norm": 0.315382182598114, "learning_rate": 9.905804293678907e-05, "loss": 0.0334, "step": 10880 }, { "grad_norm": 0.33155518770217896, "learning_rate": 9.905484587354766e-05, "loss": 0.0301, "step": 10890 }, { "grad_norm": 0.2999240458011627, "learning_rate": 9.905164344574205e-05, "loss": 0.0312, "step": 10900 }, { "grad_norm": 0.2919544577598572, "learning_rate": 9.904843565372248e-05, "loss": 0.0311, "step": 10910 }, { "grad_norm": 0.24710136651992798, "learning_rate": 9.904522249783972e-05, "loss": 0.0305, "step": 10920 }, { "grad_norm": 0.2945749759674072, "learning_rate": 9.904200397844517e-05, "loss": 0.0364, "step": 10930 }, { "grad_norm": 0.2783537209033966, "learning_rate": 9.903878009589078e-05, "loss": 0.0344, "step": 10940 }, { "grad_norm": 0.277155339717865, "learning_rate": 9.903555085052915e-05, "loss": 0.0299, "step": 10950 }, { "grad_norm": 0.3966974914073944, "learning_rate": 9.903231624271338e-05, "loss": 0.0296, "step": 10960 }, { "grad_norm": 0.3335888981819153, "learning_rate": 9.902907627279724e-05, "loss": 0.0352, "step": 10970 }, { "grad_norm": 0.24168270826339722, "learning_rate": 9.902583094113504e-05, "loss": 0.0326, "step": 10980 }, { "grad_norm": 0.3407548666000366, "learning_rate": 9.902258024808168e-05, "loss": 0.036, "step": 10990 }, { "grad_norm": 0.3652121126651764, "learning_rate": 9.901932419399264e-05, "loss": 0.0347, "step": 11000 }, { "grad_norm": 0.3411984145641327, "learning_rate": 9.9016062779224e-05, "loss": 0.0366, "step": 11010 }, { "grad_norm": 0.3080407381057739, "learning_rate": 9.901279600413242e-05, "loss": 0.0353, "step": 11020 }, { "grad_norm": 0.2795518636703491, "learning_rate": 9.900952386907518e-05, "loss": 0.0316, "step": 11030 }, { "grad_norm": 0.3327888250350952, "learning_rate": 9.90062463744101e-05, "loss": 0.0366, "step": 11040 }, { "grad_norm": 0.33777448534965515, "learning_rate": 9.900296352049558e-05, "loss": 0.0322, "step": 11050 }, { "grad_norm": 0.37660154700279236, "learning_rate": 9.899967530769065e-05, "loss": 0.0343, "step": 11060 }, { "grad_norm": 0.2720431685447693, "learning_rate": 9.899638173635489e-05, "loss": 0.0299, "step": 11070 }, { "grad_norm": 0.29247578978538513, "learning_rate": 9.899308280684849e-05, "loss": 0.0291, "step": 11080 }, { "grad_norm": 0.35772547125816345, "learning_rate": 9.898977851953222e-05, "loss": 0.0304, "step": 11090 }, { "grad_norm": 0.2825777530670166, "learning_rate": 9.898646887476741e-05, "loss": 0.0315, "step": 11100 }, { "grad_norm": 0.2946970760822296, "learning_rate": 9.898315387291603e-05, "loss": 0.0326, "step": 11110 }, { "grad_norm": 0.3964313268661499, "learning_rate": 9.89798335143406e-05, "loss": 0.0322, "step": 11120 }, { "grad_norm": 0.30745354294776917, "learning_rate": 9.897650779940419e-05, "loss": 0.0327, "step": 11130 }, { "grad_norm": 0.30488577485084534, "learning_rate": 9.897317672847054e-05, "loss": 0.0309, "step": 11140 }, { "grad_norm": 0.7053833603858948, "learning_rate": 9.89698403019039e-05, "loss": 0.0322, "step": 11150 }, { "grad_norm": 0.26122501492500305, "learning_rate": 9.896649852006917e-05, "loss": 0.036, "step": 11160 }, { "grad_norm": 0.20986796915531158, "learning_rate": 9.896315138333177e-05, "loss": 0.0336, "step": 11170 }, { "grad_norm": 0.30337849259376526, "learning_rate": 9.895979889205774e-05, "loss": 0.0313, "step": 11180 }, { "grad_norm": 0.38645046949386597, "learning_rate": 9.895644104661372e-05, "loss": 0.0378, "step": 11190 }, { "grad_norm": 0.36395785212516785, "learning_rate": 9.895307784736691e-05, "loss": 0.0345, "step": 11200 }, { "grad_norm": 0.29018348455429077, "learning_rate": 9.894970929468512e-05, "loss": 0.0343, "step": 11210 }, { "grad_norm": 0.24500784277915955, "learning_rate": 9.89463353889367e-05, "loss": 0.0324, "step": 11220 }, { "grad_norm": 0.42400339245796204, "learning_rate": 9.894295613049065e-05, "loss": 0.0397, "step": 11230 }, { "grad_norm": 0.2667916715145111, "learning_rate": 9.893957151971649e-05, "loss": 0.0313, "step": 11240 }, { "grad_norm": 0.31172871589660645, "learning_rate": 9.893618155698436e-05, "loss": 0.0317, "step": 11250 }, { "grad_norm": 0.28509441018104553, "learning_rate": 9.8932786242665e-05, "loss": 0.029, "step": 11260 }, { "grad_norm": 0.23811271786689758, "learning_rate": 9.89293855771297e-05, "loss": 0.0291, "step": 11270 }, { "grad_norm": 0.36342573165893555, "learning_rate": 9.892597956075036e-05, "loss": 0.0264, "step": 11280 }, { "grad_norm": 0.2438478320837021, "learning_rate": 9.892256819389947e-05, "loss": 0.0274, "step": 11290 }, { "grad_norm": 0.30911633372306824, "learning_rate": 9.891915147695006e-05, "loss": 0.0264, "step": 11300 }, { "grad_norm": 0.2426052987575531, "learning_rate": 9.891572941027577e-05, "loss": 0.0276, "step": 11310 }, { "grad_norm": 0.25202879309654236, "learning_rate": 9.89123019942509e-05, "loss": 0.0263, "step": 11320 }, { "grad_norm": 0.32665055990219116, "learning_rate": 9.89088692292502e-05, "loss": 0.0315, "step": 11330 }, { "grad_norm": 0.4137662649154663, "learning_rate": 9.89054311156491e-05, "loss": 0.0402, "step": 11340 }, { "grad_norm": 0.35892918705940247, "learning_rate": 9.890198765382357e-05, "loss": 0.0368, "step": 11350 }, { "grad_norm": 0.24309398233890533, "learning_rate": 9.889853884415021e-05, "loss": 0.0343, "step": 11360 }, { "grad_norm": 0.24872298538684845, "learning_rate": 9.889508468700614e-05, "loss": 0.036, "step": 11370 }, { "grad_norm": 0.28711068630218506, "learning_rate": 9.889162518276915e-05, "loss": 0.036, "step": 11380 }, { "grad_norm": 0.331210732460022, "learning_rate": 9.888816033181752e-05, "loss": 0.0326, "step": 11390 }, { "grad_norm": 0.2928112745285034, "learning_rate": 9.888469013453018e-05, "loss": 0.0354, "step": 11400 }, { "grad_norm": 0.28699102997779846, "learning_rate": 9.888121459128663e-05, "loss": 0.0317, "step": 11410 }, { "grad_norm": 0.28452304005622864, "learning_rate": 9.887773370246693e-05, "loss": 0.029, "step": 11420 }, { "grad_norm": 0.2603856325149536, "learning_rate": 9.887424746845177e-05, "loss": 0.0303, "step": 11430 }, { "grad_norm": 0.2144809067249298, "learning_rate": 9.887075588962239e-05, "loss": 0.0263, "step": 11440 }, { "grad_norm": 0.2866925597190857, "learning_rate": 9.88672589663606e-05, "loss": 0.0311, "step": 11450 }, { "grad_norm": 0.23436139523983002, "learning_rate": 9.886375669904886e-05, "loss": 0.0311, "step": 11460 }, { "grad_norm": 0.2831539511680603, "learning_rate": 9.886024908807014e-05, "loss": 0.0309, "step": 11470 }, { "grad_norm": 0.2446422129869461, "learning_rate": 9.885673613380806e-05, "loss": 0.0277, "step": 11480 }, { "grad_norm": 0.31884878873825073, "learning_rate": 9.885321783664676e-05, "loss": 0.0302, "step": 11490 }, { "grad_norm": 0.30045533180236816, "learning_rate": 9.884969419697101e-05, "loss": 0.0306, "step": 11500 }, { "grad_norm": 0.30592960119247437, "learning_rate": 9.884616521516614e-05, "loss": 0.0354, "step": 11510 }, { "grad_norm": 0.2850479185581207, "learning_rate": 9.88426308916181e-05, "loss": 0.0316, "step": 11520 }, { "grad_norm": 0.406877726316452, "learning_rate": 9.883909122671335e-05, "loss": 0.035, "step": 11530 }, { "grad_norm": 0.24110694229602814, "learning_rate": 9.883554622083904e-05, "loss": 0.0309, "step": 11540 }, { "grad_norm": 0.28337886929512024, "learning_rate": 9.88319958743828e-05, "loss": 0.0289, "step": 11550 }, { "grad_norm": 0.2681351900100708, "learning_rate": 9.882844018773291e-05, "loss": 0.03, "step": 11560 }, { "grad_norm": 0.28094810247421265, "learning_rate": 9.882487916127823e-05, "loss": 0.0281, "step": 11570 }, { "grad_norm": 0.33127880096435547, "learning_rate": 9.882131279540815e-05, "loss": 0.0318, "step": 11580 }, { "grad_norm": 0.30856451392173767, "learning_rate": 9.881774109051271e-05, "loss": 0.0312, "step": 11590 }, { "grad_norm": 0.24583080410957336, "learning_rate": 9.881416404698252e-05, "loss": 0.028, "step": 11600 }, { "grad_norm": 0.22277143597602844, "learning_rate": 9.881058166520873e-05, "loss": 0.0311, "step": 11610 }, { "grad_norm": 0.27891454100608826, "learning_rate": 9.880699394558311e-05, "loss": 0.0326, "step": 11620 }, { "grad_norm": 0.2919318377971649, "learning_rate": 9.880340088849801e-05, "loss": 0.0368, "step": 11630 }, { "grad_norm": 0.26877421140670776, "learning_rate": 9.879980249434637e-05, "loss": 0.0304, "step": 11640 }, { "grad_norm": 0.2925218939781189, "learning_rate": 9.879619876352168e-05, "loss": 0.0312, "step": 11650 }, { "grad_norm": 0.2311372458934784, "learning_rate": 9.879258969641809e-05, "loss": 0.0293, "step": 11660 }, { "grad_norm": 0.3274837136268616, "learning_rate": 9.878897529343023e-05, "loss": 0.0303, "step": 11670 }, { "grad_norm": 0.3064956068992615, "learning_rate": 9.878535555495338e-05, "loss": 0.0282, "step": 11680 }, { "grad_norm": 0.3240980803966522, "learning_rate": 9.87817304813834e-05, "loss": 0.0294, "step": 11690 }, { "grad_norm": 0.2995535135269165, "learning_rate": 9.877810007311671e-05, "loss": 0.0295, "step": 11700 }, { "grad_norm": 0.26659801602363586, "learning_rate": 9.877446433055035e-05, "loss": 0.0274, "step": 11710 }, { "grad_norm": 0.23300999402999878, "learning_rate": 9.877082325408191e-05, "loss": 0.0274, "step": 11720 }, { "grad_norm": 0.3468973934650421, "learning_rate": 9.876717684410954e-05, "loss": 0.0289, "step": 11730 }, { "grad_norm": 0.29549887776374817, "learning_rate": 9.876352510103204e-05, "loss": 0.0304, "step": 11740 }, { "grad_norm": 0.2805496156215668, "learning_rate": 9.875986802524875e-05, "loss": 0.0301, "step": 11750 }, { "grad_norm": 0.382571816444397, "learning_rate": 9.87562056171596e-05, "loss": 0.0314, "step": 11760 }, { "grad_norm": 0.41318532824516296, "learning_rate": 9.875253787716511e-05, "loss": 0.0301, "step": 11770 }, { "grad_norm": 0.24480850994586945, "learning_rate": 9.874886480566637e-05, "loss": 0.0266, "step": 11780 }, { "grad_norm": 0.31495559215545654, "learning_rate": 9.874518640306507e-05, "loss": 0.03, "step": 11790 }, { "grad_norm": 0.26663702726364136, "learning_rate": 9.874150266976347e-05, "loss": 0.0313, "step": 11800 }, { "grad_norm": 0.24700815975666046, "learning_rate": 9.873781360616443e-05, "loss": 0.0295, "step": 11810 }, { "grad_norm": 0.29899346828460693, "learning_rate": 9.873411921267137e-05, "loss": 0.0328, "step": 11820 }, { "grad_norm": 0.24820800125598907, "learning_rate": 9.873041948968829e-05, "loss": 0.0304, "step": 11830 }, { "grad_norm": 0.32520291209220886, "learning_rate": 9.872671443761981e-05, "loss": 0.0278, "step": 11840 }, { "grad_norm": 0.2321546971797943, "learning_rate": 9.872300405687109e-05, "loss": 0.0291, "step": 11850 }, { "grad_norm": 0.2785046100616455, "learning_rate": 9.871928834784792e-05, "loss": 0.0294, "step": 11860 }, { "grad_norm": 0.2704564929008484, "learning_rate": 9.871556731095661e-05, "loss": 0.0317, "step": 11870 }, { "grad_norm": 0.2936965823173523, "learning_rate": 9.871184094660411e-05, "loss": 0.0313, "step": 11880 }, { "grad_norm": 0.31777456402778625, "learning_rate": 9.870810925519791e-05, "loss": 0.028, "step": 11890 }, { "grad_norm": 0.3072994351387024, "learning_rate": 9.870437223714612e-05, "loss": 0.0309, "step": 11900 }, { "grad_norm": 0.26418203115463257, "learning_rate": 9.87006298928574e-05, "loss": 0.0288, "step": 11910 }, { "grad_norm": 0.2785259783267975, "learning_rate": 9.869688222274103e-05, "loss": 0.0318, "step": 11920 }, { "grad_norm": 0.3408079445362091, "learning_rate": 9.869312922720681e-05, "loss": 0.0329, "step": 11930 }, { "grad_norm": 0.28353238105773926, "learning_rate": 9.868937090666521e-05, "loss": 0.0341, "step": 11940 }, { "grad_norm": 0.2503509819507599, "learning_rate": 9.86856072615272e-05, "loss": 0.028, "step": 11950 }, { "grad_norm": 0.2510024309158325, "learning_rate": 9.868183829220438e-05, "loss": 0.0329, "step": 11960 }, { "grad_norm": 0.357908695936203, "learning_rate": 9.867806399910893e-05, "loss": 0.0298, "step": 11970 }, { "grad_norm": 0.7028481364250183, "learning_rate": 9.867428438265356e-05, "loss": 0.0291, "step": 11980 }, { "grad_norm": 0.23903116583824158, "learning_rate": 9.867049944325165e-05, "loss": 0.0315, "step": 11990 }, { "grad_norm": 0.33873364329338074, "learning_rate": 9.86667091813171e-05, "loss": 0.034, "step": 12000 }, { "grad_norm": 0.31432899832725525, "learning_rate": 9.866291359726438e-05, "loss": 0.034, "step": 12010 }, { "grad_norm": 0.3113586902618408, "learning_rate": 9.865911269150861e-05, "loss": 0.0298, "step": 12020 }, { "grad_norm": 0.3035186231136322, "learning_rate": 9.865530646446544e-05, "loss": 0.0313, "step": 12030 }, { "grad_norm": 0.3457525074481964, "learning_rate": 9.86514949165511e-05, "loss": 0.0319, "step": 12040 }, { "grad_norm": 0.37277066707611084, "learning_rate": 9.864767804818243e-05, "loss": 0.0332, "step": 12050 }, { "grad_norm": 0.2696834206581116, "learning_rate": 9.86438558597768e-05, "loss": 0.0317, "step": 12060 }, { "grad_norm": 0.33301451802253723, "learning_rate": 9.864002835175225e-05, "loss": 0.0296, "step": 12070 }, { "grad_norm": 0.26416921615600586, "learning_rate": 9.863619552452734e-05, "loss": 0.0301, "step": 12080 }, { "grad_norm": 0.23444032669067383, "learning_rate": 9.863235737852119e-05, "loss": 0.0255, "step": 12090 }, { "grad_norm": 0.23341479897499084, "learning_rate": 9.862851391415356e-05, "loss": 0.0299, "step": 12100 }, { "grad_norm": 0.2764512598514557, "learning_rate": 9.862466513184477e-05, "loss": 0.0274, "step": 12110 }, { "grad_norm": 0.2435770183801651, "learning_rate": 9.86208110320157e-05, "loss": 0.0297, "step": 12120 }, { "grad_norm": 0.3064531683921814, "learning_rate": 9.861695161508784e-05, "loss": 0.0281, "step": 12130 }, { "grad_norm": 0.25749170780181885, "learning_rate": 9.861308688148324e-05, "loss": 0.0274, "step": 12140 }, { "grad_norm": 0.23647122085094452, "learning_rate": 9.860921683162455e-05, "loss": 0.027, "step": 12150 }, { "grad_norm": 0.25491082668304443, "learning_rate": 9.860534146593499e-05, "loss": 0.0274, "step": 12160 }, { "grad_norm": 0.29446500539779663, "learning_rate": 9.860146078483836e-05, "loss": 0.0266, "step": 12170 }, { "grad_norm": 0.2897636294364929, "learning_rate": 9.859757478875905e-05, "loss": 0.0285, "step": 12180 }, { "grad_norm": 0.24061544239521027, "learning_rate": 9.859368347812204e-05, "loss": 0.0261, "step": 12190 }, { "grad_norm": 0.21952679753303528, "learning_rate": 9.858978685335285e-05, "loss": 0.0314, "step": 12200 }, { "grad_norm": 0.28628602623939514, "learning_rate": 9.858588491487763e-05, "loss": 0.0321, "step": 12210 }, { "grad_norm": 0.34040144085884094, "learning_rate": 9.858197766312308e-05, "loss": 0.0249, "step": 12220 }, { "grad_norm": 0.29905787110328674, "learning_rate": 9.857806509851649e-05, "loss": 0.0304, "step": 12230 }, { "grad_norm": 0.2661277651786804, "learning_rate": 9.857414722148574e-05, "loss": 0.0303, "step": 12240 }, { "grad_norm": 0.3244817852973938, "learning_rate": 9.857022403245928e-05, "loss": 0.0258, "step": 12250 }, { "grad_norm": 0.23159995675086975, "learning_rate": 9.856629553186615e-05, "loss": 0.029, "step": 12260 }, { "grad_norm": 0.3328641653060913, "learning_rate": 9.856236172013595e-05, "loss": 0.0316, "step": 12270 }, { "grad_norm": 0.2700420022010803, "learning_rate": 9.85584225976989e-05, "loss": 0.0258, "step": 12280 }, { "grad_norm": 0.3571769893169403, "learning_rate": 9.855447816498575e-05, "loss": 0.0237, "step": 12290 }, { "grad_norm": 0.2360246479511261, "learning_rate": 9.855052842242787e-05, "loss": 0.0292, "step": 12300 }, { "grad_norm": 0.3403230309486389, "learning_rate": 9.85465733704572e-05, "loss": 0.0301, "step": 12310 }, { "grad_norm": 0.22212450206279755, "learning_rate": 9.854261300950624e-05, "loss": 0.0265, "step": 12320 }, { "grad_norm": 0.2780228853225708, "learning_rate": 9.853864734000813e-05, "loss": 0.0265, "step": 12330 }, { "grad_norm": 0.2582284212112427, "learning_rate": 9.85346763623965e-05, "loss": 0.03, "step": 12340 }, { "grad_norm": 0.3481437861919403, "learning_rate": 9.853070007710564e-05, "loss": 0.0318, "step": 12350 }, { "grad_norm": 0.2725103795528412, "learning_rate": 9.85267184845704e-05, "loss": 0.0278, "step": 12360 }, { "grad_norm": 0.2570069432258606, "learning_rate": 9.852273158522616e-05, "loss": 0.0259, "step": 12370 }, { "grad_norm": 0.2245282083749771, "learning_rate": 9.851873937950896e-05, "loss": 0.0261, "step": 12380 }, { "grad_norm": 0.290382444858551, "learning_rate": 9.851474186785537e-05, "loss": 0.0275, "step": 12390 }, { "grad_norm": 0.3187590539455414, "learning_rate": 9.851073905070254e-05, "loss": 0.0291, "step": 12400 }, { "grad_norm": 0.2873249351978302, "learning_rate": 9.850673092848824e-05, "loss": 0.0263, "step": 12410 }, { "grad_norm": 0.34618911147117615, "learning_rate": 9.850271750165077e-05, "loss": 0.028, "step": 12420 }, { "grad_norm": 0.2600424587726593, "learning_rate": 9.849869877062902e-05, "loss": 0.0308, "step": 12430 }, { "grad_norm": 0.289774090051651, "learning_rate": 9.849467473586252e-05, "loss": 0.0274, "step": 12440 }, { "grad_norm": 0.2742133140563965, "learning_rate": 9.849064539779127e-05, "loss": 0.0286, "step": 12450 }, { "grad_norm": 0.3591507375240326, "learning_rate": 9.848661075685594e-05, "loss": 0.0265, "step": 12460 }, { "grad_norm": 0.3671013414859772, "learning_rate": 9.848257081349778e-05, "loss": 0.0328, "step": 12470 }, { "grad_norm": 0.44096988439559937, "learning_rate": 9.847852556815856e-05, "loss": 0.0324, "step": 12480 }, { "grad_norm": 0.2689562737941742, "learning_rate": 9.847447502128067e-05, "loss": 0.0328, "step": 12490 }, { "grad_norm": 0.370286762714386, "learning_rate": 9.847041917330708e-05, "loss": 0.0273, "step": 12500 }, { "grad_norm": 0.37345996499061584, "learning_rate": 9.846635802468132e-05, "loss": 0.0326, "step": 12510 }, { "grad_norm": 0.40669766068458557, "learning_rate": 9.84622915758475e-05, "loss": 0.0276, "step": 12520 }, { "grad_norm": 0.2931649386882782, "learning_rate": 9.845821982725034e-05, "loss": 0.0283, "step": 12530 }, { "grad_norm": 0.2989388406276703, "learning_rate": 9.845414277933514e-05, "loss": 0.027, "step": 12540 }, { "grad_norm": 0.2481948286294937, "learning_rate": 9.845006043254771e-05, "loss": 0.0282, "step": 12550 }, { "grad_norm": 0.2054479569196701, "learning_rate": 9.844597278733451e-05, "loss": 0.0322, "step": 12560 }, { "grad_norm": 0.319269061088562, "learning_rate": 9.844187984414259e-05, "loss": 0.0386, "step": 12570 }, { "grad_norm": 0.2610693573951721, "learning_rate": 9.84377816034195e-05, "loss": 0.0383, "step": 12580 }, { "grad_norm": 0.29544004797935486, "learning_rate": 9.843367806561345e-05, "loss": 0.0333, "step": 12590 }, { "grad_norm": 0.7491469979286194, "learning_rate": 9.842956923117317e-05, "loss": 0.032, "step": 12600 }, { "grad_norm": 0.26052916049957275, "learning_rate": 9.842545510054802e-05, "loss": 0.0347, "step": 12610 }, { "grad_norm": 0.23960623145103455, "learning_rate": 9.842133567418792e-05, "loss": 0.0321, "step": 12620 }, { "grad_norm": 0.20660869777202606, "learning_rate": 9.841721095254333e-05, "loss": 0.0307, "step": 12630 }, { "grad_norm": 0.3334543704986572, "learning_rate": 9.841308093606537e-05, "loss": 0.0323, "step": 12640 }, { "grad_norm": 0.3099224865436554, "learning_rate": 9.840894562520565e-05, "loss": 0.0304, "step": 12650 }, { "grad_norm": 0.3841031491756439, "learning_rate": 9.840480502041642e-05, "loss": 0.0374, "step": 12660 }, { "grad_norm": 0.30218327045440674, "learning_rate": 9.840065912215049e-05, "loss": 0.042, "step": 12670 }, { "grad_norm": 0.39909225702285767, "learning_rate": 9.839650793086124e-05, "loss": 0.0342, "step": 12680 }, { "grad_norm": 0.3083195388317108, "learning_rate": 9.839235144700265e-05, "loss": 0.0365, "step": 12690 }, { "grad_norm": 0.3122695982456207, "learning_rate": 9.838818967102926e-05, "loss": 0.031, "step": 12700 }, { "grad_norm": 0.26696404814720154, "learning_rate": 9.83840226033962e-05, "loss": 0.0322, "step": 12710 }, { "grad_norm": 0.28039416670799255, "learning_rate": 9.837985024455918e-05, "loss": 0.0325, "step": 12720 }, { "grad_norm": 0.3580036461353302, "learning_rate": 9.837567259497447e-05, "loss": 0.0318, "step": 12730 }, { "grad_norm": 0.280336856842041, "learning_rate": 9.837148965509894e-05, "loss": 0.0288, "step": 12740 }, { "grad_norm": 0.22634652256965637, "learning_rate": 9.836730142539001e-05, "loss": 0.0291, "step": 12750 }, { "grad_norm": 0.27783605456352234, "learning_rate": 9.836310790630574e-05, "loss": 0.032, "step": 12760 }, { "grad_norm": 0.2561364471912384, "learning_rate": 9.83589090983047e-05, "loss": 0.0316, "step": 12770 }, { "grad_norm": 0.2458530217409134, "learning_rate": 9.835470500184605e-05, "loss": 0.0284, "step": 12780 }, { "grad_norm": 0.3657557964324951, "learning_rate": 9.835049561738957e-05, "loss": 0.0297, "step": 12790 }, { "grad_norm": 0.3425447940826416, "learning_rate": 9.834628094539558e-05, "loss": 0.0306, "step": 12800 }, { "grad_norm": 0.29536768794059753, "learning_rate": 9.834206098632499e-05, "loss": 0.0289, "step": 12810 }, { "grad_norm": 0.20483297109603882, "learning_rate": 9.833783574063931e-05, "loss": 0.0299, "step": 12820 }, { "grad_norm": 0.23044821619987488, "learning_rate": 9.833360520880058e-05, "loss": 0.0295, "step": 12830 }, { "grad_norm": 0.3162164092063904, "learning_rate": 9.832936939127144e-05, "loss": 0.0277, "step": 12840 }, { "grad_norm": 0.27418994903564453, "learning_rate": 9.832512828851515e-05, "loss": 0.0324, "step": 12850 }, { "grad_norm": 0.2635737657546997, "learning_rate": 9.832088190099546e-05, "loss": 0.0291, "step": 12860 }, { "grad_norm": 0.26308611035346985, "learning_rate": 9.831663022917679e-05, "loss": 0.0291, "step": 12870 }, { "grad_norm": 0.49325093626976013, "learning_rate": 9.831237327352407e-05, "loss": 0.0277, "step": 12880 }, { "grad_norm": 0.28930798172950745, "learning_rate": 9.830811103450286e-05, "loss": 0.0332, "step": 12890 }, { "grad_norm": 0.2797831594944, "learning_rate": 9.830384351257924e-05, "loss": 0.0288, "step": 12900 }, { "grad_norm": 0.22978238761425018, "learning_rate": 9.829957070821993e-05, "loss": 0.0289, "step": 12910 }, { "grad_norm": 0.2119012176990509, "learning_rate": 9.829529262189218e-05, "loss": 0.0299, "step": 12920 }, { "grad_norm": 0.26578977704048157, "learning_rate": 9.829100925406385e-05, "loss": 0.0314, "step": 12930 }, { "grad_norm": 0.30753880739212036, "learning_rate": 9.828672060520333e-05, "loss": 0.0293, "step": 12940 }, { "grad_norm": 0.2932325005531311, "learning_rate": 9.828242667577966e-05, "loss": 0.0315, "step": 12950 }, { "grad_norm": 0.2503829598426819, "learning_rate": 9.82781274662624e-05, "loss": 0.0285, "step": 12960 }, { "grad_norm": 0.41675084829330444, "learning_rate": 9.82738229771217e-05, "loss": 0.0265, "step": 12970 }, { "grad_norm": 0.3060649633407593, "learning_rate": 9.826951320882829e-05, "loss": 0.0296, "step": 12980 }, { "grad_norm": 0.34081849455833435, "learning_rate": 9.826519816185351e-05, "loss": 0.0304, "step": 12990 }, { "grad_norm": 0.25545740127563477, "learning_rate": 9.826087783666921e-05, "loss": 0.0288, "step": 13000 }, { "grad_norm": 0.23892076313495636, "learning_rate": 9.825655223374787e-05, "loss": 0.0282, "step": 13010 }, { "grad_norm": 0.2632860243320465, "learning_rate": 9.825222135356253e-05, "loss": 0.0294, "step": 13020 }, { "grad_norm": 0.2793518900871277, "learning_rate": 9.82478851965868e-05, "loss": 0.0278, "step": 13030 }, { "grad_norm": 0.36557912826538086, "learning_rate": 9.82435437632949e-05, "loss": 0.0252, "step": 13040 }, { "grad_norm": 0.28008192777633667, "learning_rate": 9.823919705416158e-05, "loss": 0.0286, "step": 13050 }, { "grad_norm": 0.29844537377357483, "learning_rate": 9.82348450696622e-05, "loss": 0.0287, "step": 13060 }, { "grad_norm": 0.3132396936416626, "learning_rate": 9.823048781027268e-05, "loss": 0.0247, "step": 13070 }, { "grad_norm": 0.29415959119796753, "learning_rate": 9.822612527646953e-05, "loss": 0.0287, "step": 13080 }, { "grad_norm": 0.31978222727775574, "learning_rate": 9.822175746872984e-05, "loss": 0.0295, "step": 13090 }, { "grad_norm": 0.2840472161769867, "learning_rate": 9.821738438753123e-05, "loss": 0.0279, "step": 13100 }, { "grad_norm": 0.3484739661216736, "learning_rate": 9.821300603335196e-05, "loss": 0.028, "step": 13110 }, { "grad_norm": 0.20763424038887024, "learning_rate": 9.820862240667085e-05, "loss": 0.029, "step": 13120 }, { "grad_norm": 0.26649394631385803, "learning_rate": 9.820423350796726e-05, "loss": 0.0286, "step": 13130 }, { "grad_norm": 0.30691468715667725, "learning_rate": 9.819983933772118e-05, "loss": 0.0266, "step": 13140 }, { "grad_norm": 0.259863942861557, "learning_rate": 9.819543989641314e-05, "loss": 0.0273, "step": 13150 }, { "grad_norm": 0.2591104805469513, "learning_rate": 9.819103518452423e-05, "loss": 0.0265, "step": 13160 }, { "grad_norm": 0.3903253674507141, "learning_rate": 9.818662520253618e-05, "loss": 0.0282, "step": 13170 }, { "grad_norm": 0.21888665854930878, "learning_rate": 9.818220995093126e-05, "loss": 0.0288, "step": 13180 }, { "grad_norm": 0.2954021096229553, "learning_rate": 9.817778943019228e-05, "loss": 0.0298, "step": 13190 }, { "grad_norm": 0.2789255976676941, "learning_rate": 9.81733636408027e-05, "loss": 0.0233, "step": 13200 }, { "grad_norm": 0.2330876588821411, "learning_rate": 9.816893258324649e-05, "loss": 0.0257, "step": 13210 }, { "grad_norm": 0.36955535411834717, "learning_rate": 9.816449625800823e-05, "loss": 0.0268, "step": 13220 }, { "grad_norm": 0.21922053396701813, "learning_rate": 9.816005466557308e-05, "loss": 0.0275, "step": 13230 }, { "grad_norm": 0.250992089509964, "learning_rate": 9.815560780642674e-05, "loss": 0.0243, "step": 13240 }, { "grad_norm": 0.25758421421051025, "learning_rate": 9.815115568105555e-05, "loss": 0.0274, "step": 13250 }, { "grad_norm": 0.2431662678718567, "learning_rate": 9.814669828994638e-05, "loss": 0.0283, "step": 13260 }, { "grad_norm": 0.28598082065582275, "learning_rate": 9.814223563358665e-05, "loss": 0.0259, "step": 13270 }, { "grad_norm": 0.33157023787498474, "learning_rate": 9.813776771246443e-05, "loss": 0.0289, "step": 13280 }, { "grad_norm": 0.2712765038013458, "learning_rate": 9.813329452706829e-05, "loss": 0.0284, "step": 13290 }, { "grad_norm": 0.2299700528383255, "learning_rate": 9.812881607788744e-05, "loss": 0.0299, "step": 13300 }, { "grad_norm": 0.2990049719810486, "learning_rate": 9.812433236541163e-05, "loss": 0.0281, "step": 13310 }, { "grad_norm": 0.3478328585624695, "learning_rate": 9.811984339013116e-05, "loss": 0.0326, "step": 13320 }, { "grad_norm": 0.24790965020656586, "learning_rate": 9.811534915253698e-05, "loss": 0.0323, "step": 13330 }, { "grad_norm": 0.2626650631427765, "learning_rate": 9.811084965312056e-05, "loss": 0.0278, "step": 13340 }, { "grad_norm": 0.30897101759910583, "learning_rate": 9.810634489237396e-05, "loss": 0.0292, "step": 13350 }, { "grad_norm": 0.31656885147094727, "learning_rate": 9.81018348707898e-05, "loss": 0.0354, "step": 13360 }, { "grad_norm": 0.28134584426879883, "learning_rate": 9.809731958886131e-05, "loss": 0.0272, "step": 13370 }, { "grad_norm": 0.22056497633457184, "learning_rate": 9.809279904708224e-05, "loss": 0.0292, "step": 13380 }, { "grad_norm": 0.3755132555961609, "learning_rate": 9.808827324594699e-05, "loss": 0.0284, "step": 13390 }, { "grad_norm": 0.2492716908454895, "learning_rate": 9.808374218595046e-05, "loss": 0.0282, "step": 13400 }, { "grad_norm": 0.28557100892066956, "learning_rate": 9.80792058675882e-05, "loss": 0.0283, "step": 13410 }, { "grad_norm": 0.3788284659385681, "learning_rate": 9.807466429135627e-05, "loss": 0.0305, "step": 13420 }, { "grad_norm": 0.3045898377895355, "learning_rate": 9.807011745775132e-05, "loss": 0.0329, "step": 13430 }, { "grad_norm": 0.24230900406837463, "learning_rate": 9.806556536727061e-05, "loss": 0.0307, "step": 13440 }, { "grad_norm": 0.239655002951622, "learning_rate": 9.806100802041193e-05, "loss": 0.0271, "step": 13450 }, { "grad_norm": 0.31737855076789856, "learning_rate": 9.805644541767368e-05, "loss": 0.029, "step": 13460 }, { "grad_norm": 0.30794671177864075, "learning_rate": 9.805187755955478e-05, "loss": 0.0307, "step": 13470 }, { "grad_norm": 0.24954520165920258, "learning_rate": 9.804730444655483e-05, "loss": 0.0283, "step": 13480 }, { "grad_norm": 0.3675327003002167, "learning_rate": 9.804272607917388e-05, "loss": 0.0273, "step": 13490 }, { "grad_norm": 0.3177080452442169, "learning_rate": 9.803814245791265e-05, "loss": 0.0313, "step": 13500 }, { "grad_norm": 0.2457638531923294, "learning_rate": 9.803355358327239e-05, "loss": 0.0316, "step": 13510 }, { "grad_norm": 0.2858870029449463, "learning_rate": 9.802895945575492e-05, "loss": 0.0272, "step": 13520 }, { "grad_norm": 0.29847264289855957, "learning_rate": 9.802436007586266e-05, "loss": 0.0293, "step": 13530 }, { "grad_norm": 0.2852008044719696, "learning_rate": 9.801975544409858e-05, "loss": 0.0285, "step": 13540 }, { "grad_norm": 0.22683930397033691, "learning_rate": 9.801514556096625e-05, "loss": 0.0316, "step": 13550 }, { "grad_norm": 0.2886556386947632, "learning_rate": 9.801053042696977e-05, "loss": 0.0283, "step": 13560 }, { "grad_norm": 0.31018295884132385, "learning_rate": 9.800591004261388e-05, "loss": 0.0273, "step": 13570 }, { "grad_norm": 0.29426896572113037, "learning_rate": 9.800128440840385e-05, "loss": 0.0276, "step": 13580 }, { "grad_norm": 0.23416514694690704, "learning_rate": 9.799665352484552e-05, "loss": 0.0304, "step": 13590 }, { "grad_norm": 0.2586263418197632, "learning_rate": 9.799201739244532e-05, "loss": 0.0284, "step": 13600 }, { "grad_norm": 0.3369411528110504, "learning_rate": 9.798737601171025e-05, "loss": 0.0255, "step": 13610 }, { "grad_norm": 0.28444117307662964, "learning_rate": 9.79827293831479e-05, "loss": 0.026, "step": 13620 }, { "grad_norm": 0.2636748254299164, "learning_rate": 9.797807750726638e-05, "loss": 0.0291, "step": 13630 }, { "grad_norm": 0.27649861574172974, "learning_rate": 9.797342038457446e-05, "loss": 0.028, "step": 13640 }, { "grad_norm": 0.29361677169799805, "learning_rate": 9.796875801558141e-05, "loss": 0.0306, "step": 13650 }, { "grad_norm": 0.24288217723369598, "learning_rate": 9.79640904007971e-05, "loss": 0.0291, "step": 13660 }, { "grad_norm": 0.2612079381942749, "learning_rate": 9.795941754073199e-05, "loss": 0.0286, "step": 13670 }, { "grad_norm": 0.3055292069911957, "learning_rate": 9.795473943589705e-05, "loss": 0.0333, "step": 13680 }, { "grad_norm": 0.2975929379463196, "learning_rate": 9.795005608680394e-05, "loss": 0.0292, "step": 13690 }, { "grad_norm": 0.22235698997974396, "learning_rate": 9.794536749396477e-05, "loss": 0.0291, "step": 13700 }, { "grad_norm": 0.32491835951805115, "learning_rate": 9.79406736578923e-05, "loss": 0.0289, "step": 13710 }, { "grad_norm": 0.2358362376689911, "learning_rate": 9.793597457909984e-05, "loss": 0.0268, "step": 13720 }, { "grad_norm": 0.29517844319343567, "learning_rate": 9.793127025810127e-05, "loss": 0.0297, "step": 13730 }, { "grad_norm": 0.31216681003570557, "learning_rate": 9.792656069541104e-05, "loss": 0.0262, "step": 13740 }, { "grad_norm": 0.32026612758636475, "learning_rate": 9.79218458915442e-05, "loss": 0.0331, "step": 13750 }, { "grad_norm": 0.2948898375034332, "learning_rate": 9.791712584701634e-05, "loss": 0.0298, "step": 13760 }, { "grad_norm": 0.2506902515888214, "learning_rate": 9.791240056234364e-05, "loss": 0.0276, "step": 13770 }, { "grad_norm": 0.2658625841140747, "learning_rate": 9.790767003804283e-05, "loss": 0.0258, "step": 13780 }, { "grad_norm": 0.30689138174057007, "learning_rate": 9.790293427463126e-05, "loss": 0.0265, "step": 13790 }, { "grad_norm": 0.24957992136478424, "learning_rate": 9.789819327262684e-05, "loss": 0.0262, "step": 13800 }, { "grad_norm": 0.2922487258911133, "learning_rate": 9.7893447032548e-05, "loss": 0.0267, "step": 13810 }, { "grad_norm": 0.34596654772758484, "learning_rate": 9.78886955549138e-05, "loss": 0.0276, "step": 13820 }, { "grad_norm": 0.23931993544101715, "learning_rate": 9.788393884024387e-05, "loss": 0.0296, "step": 13830 }, { "grad_norm": 0.29264068603515625, "learning_rate": 9.787917688905836e-05, "loss": 0.0276, "step": 13840 }, { "grad_norm": 0.3396528959274292, "learning_rate": 9.787440970187807e-05, "loss": 0.0255, "step": 13850 }, { "grad_norm": 0.3646264672279358, "learning_rate": 9.786963727922429e-05, "loss": 0.0233, "step": 13860 }, { "grad_norm": 0.2844822406768799, "learning_rate": 9.786485962161897e-05, "loss": 0.028, "step": 13870 }, { "grad_norm": 0.29303380846977234, "learning_rate": 9.786007672958455e-05, "loss": 0.0256, "step": 13880 }, { "grad_norm": 0.31622326374053955, "learning_rate": 9.78552886036441e-05, "loss": 0.0259, "step": 13890 }, { "grad_norm": 0.2680067718029022, "learning_rate": 9.785049524432124e-05, "loss": 0.0267, "step": 13900 }, { "grad_norm": 0.2422163188457489, "learning_rate": 9.784569665214016e-05, "loss": 0.0287, "step": 13910 }, { "grad_norm": 0.28418999910354614, "learning_rate": 9.784089282762563e-05, "loss": 0.027, "step": 13920 }, { "grad_norm": 0.35586822032928467, "learning_rate": 9.7836083771303e-05, "loss": 0.0247, "step": 13930 }, { "grad_norm": 0.23201899230480194, "learning_rate": 9.783126948369817e-05, "loss": 0.0267, "step": 13940 }, { "grad_norm": 0.2459719330072403, "learning_rate": 9.78264499653376e-05, "loss": 0.0266, "step": 13950 }, { "grad_norm": 0.21741433441638947, "learning_rate": 9.782162521674838e-05, "loss": 0.0281, "step": 13960 }, { "grad_norm": 0.3036254644393921, "learning_rate": 9.781679523845812e-05, "loss": 0.0292, "step": 13970 }, { "grad_norm": 0.2582082748413086, "learning_rate": 9.781196003099502e-05, "loss": 0.0268, "step": 13980 }, { "grad_norm": 0.21821224689483643, "learning_rate": 9.780711959488786e-05, "loss": 0.0263, "step": 13990 }, { "grad_norm": 0.30690041184425354, "learning_rate": 9.780227393066599e-05, "loss": 0.0248, "step": 14000 }, { "grad_norm": 0.26556453108787537, "learning_rate": 9.77974230388593e-05, "loss": 0.0289, "step": 14010 }, { "grad_norm": 0.3673281967639923, "learning_rate": 9.779256691999829e-05, "loss": 0.0302, "step": 14020 }, { "grad_norm": 0.2483862191438675, "learning_rate": 9.778770557461403e-05, "loss": 0.0282, "step": 14030 }, { "grad_norm": 0.23091097176074982, "learning_rate": 9.778283900323812e-05, "loss": 0.0284, "step": 14040 }, { "grad_norm": 0.2870122492313385, "learning_rate": 9.777796720640277e-05, "loss": 0.0327, "step": 14050 }, { "grad_norm": 0.2165539264678955, "learning_rate": 9.777309018464078e-05, "loss": 0.0281, "step": 14060 }, { "grad_norm": 0.2946593165397644, "learning_rate": 9.776820793848547e-05, "loss": 0.0275, "step": 14070 }, { "grad_norm": 0.3705272078514099, "learning_rate": 9.776332046847075e-05, "loss": 0.0272, "step": 14080 }, { "grad_norm": 0.303608775138855, "learning_rate": 9.775842777513111e-05, "loss": 0.0251, "step": 14090 }, { "grad_norm": 0.2711888551712036, "learning_rate": 9.775352985900163e-05, "loss": 0.0271, "step": 14100 }, { "grad_norm": 0.2570894658565521, "learning_rate": 9.774862672061791e-05, "loss": 0.027, "step": 14110 }, { "grad_norm": 0.3103223145008087, "learning_rate": 9.774371836051616e-05, "loss": 0.0296, "step": 14120 }, { "grad_norm": 0.29827746748924255, "learning_rate": 9.773880477923315e-05, "loss": 0.0274, "step": 14130 }, { "grad_norm": 0.3159186542034149, "learning_rate": 9.773388597730623e-05, "loss": 0.0257, "step": 14140 }, { "grad_norm": 0.2636828124523163, "learning_rate": 9.77289619552733e-05, "loss": 0.0227, "step": 14150 }, { "grad_norm": 0.2392222136259079, "learning_rate": 9.772403271367285e-05, "loss": 0.0263, "step": 14160 }, { "grad_norm": 0.3110879361629486, "learning_rate": 9.771909825304396e-05, "loss": 0.0269, "step": 14170 }, { "grad_norm": 0.2784460783004761, "learning_rate": 9.771415857392619e-05, "loss": 0.0256, "step": 14180 }, { "grad_norm": 0.2535569369792938, "learning_rate": 9.770921367685978e-05, "loss": 0.0279, "step": 14190 }, { "grad_norm": 0.22352798283100128, "learning_rate": 9.770426356238551e-05, "loss": 0.0243, "step": 14200 }, { "grad_norm": 0.3302006125450134, "learning_rate": 9.769930823104469e-05, "loss": 0.0268, "step": 14210 }, { "grad_norm": 0.24799920618534088, "learning_rate": 9.769434768337926e-05, "loss": 0.03, "step": 14220 }, { "grad_norm": 0.2763763666152954, "learning_rate": 9.768938191993164e-05, "loss": 0.0283, "step": 14230 }, { "grad_norm": 0.28458482027053833, "learning_rate": 9.768441094124494e-05, "loss": 0.0274, "step": 14240 }, { "grad_norm": 0.307743102312088, "learning_rate": 9.767943474786275e-05, "loss": 0.0301, "step": 14250 }, { "grad_norm": 0.2991558909416199, "learning_rate": 9.767445334032923e-05, "loss": 0.0278, "step": 14260 }, { "grad_norm": 0.3228144645690918, "learning_rate": 9.766946671918919e-05, "loss": 0.0283, "step": 14270 }, { "grad_norm": 0.30026939511299133, "learning_rate": 9.766447488498796e-05, "loss": 0.0243, "step": 14280 }, { "grad_norm": 0.18967847526073456, "learning_rate": 9.765947783827139e-05, "loss": 0.0245, "step": 14290 }, { "grad_norm": 0.2389570027589798, "learning_rate": 9.765447557958599e-05, "loss": 0.0256, "step": 14300 }, { "grad_norm": 0.25559550523757935, "learning_rate": 9.764946810947879e-05, "loss": 0.0286, "step": 14310 }, { "grad_norm": 0.3009314239025116, "learning_rate": 9.764445542849738e-05, "loss": 0.0276, "step": 14320 }, { "grad_norm": 0.23873288929462433, "learning_rate": 9.763943753718998e-05, "loss": 0.026, "step": 14330 }, { "grad_norm": 0.19441446661949158, "learning_rate": 9.76344144361053e-05, "loss": 0.0288, "step": 14340 }, { "grad_norm": 0.21243105828762054, "learning_rate": 9.762938612579269e-05, "loss": 0.0257, "step": 14350 }, { "grad_norm": 0.22886276245117188, "learning_rate": 9.762435260680202e-05, "loss": 0.0266, "step": 14360 }, { "grad_norm": 0.23507462441921234, "learning_rate": 9.761931387968373e-05, "loss": 0.0267, "step": 14370 }, { "grad_norm": 0.2951992452144623, "learning_rate": 9.76142699449889e-05, "loss": 0.0287, "step": 14380 }, { "grad_norm": 0.2539185583591461, "learning_rate": 9.760922080326908e-05, "loss": 0.0299, "step": 14390 }, { "grad_norm": 0.215994730591774, "learning_rate": 9.760416645507644e-05, "loss": 0.0263, "step": 14400 }, { "grad_norm": 0.346110463142395, "learning_rate": 9.759910690096375e-05, "loss": 0.0285, "step": 14410 }, { "grad_norm": 0.23869460821151733, "learning_rate": 9.759404214148429e-05, "loss": 0.0246, "step": 14420 }, { "grad_norm": 0.22138464450836182, "learning_rate": 9.758897217719191e-05, "loss": 0.0278, "step": 14430 }, { "grad_norm": 0.28041037917137146, "learning_rate": 9.758389700864113e-05, "loss": 0.0268, "step": 14440 }, { "grad_norm": 0.22608302533626556, "learning_rate": 9.757881663638688e-05, "loss": 0.0262, "step": 14450 }, { "grad_norm": 0.2418254166841507, "learning_rate": 9.757373106098478e-05, "loss": 0.026, "step": 14460 }, { "grad_norm": 0.21685732901096344, "learning_rate": 9.756864028299097e-05, "loss": 0.0245, "step": 14470 }, { "grad_norm": 0.23711875081062317, "learning_rate": 9.75635443029622e-05, "loss": 0.0236, "step": 14480 }, { "grad_norm": 0.28745558857917786, "learning_rate": 9.755844312145572e-05, "loss": 0.0253, "step": 14490 }, { "grad_norm": 0.2547065317630768, "learning_rate": 9.755333673902941e-05, "loss": 0.0282, "step": 14500 }, { "grad_norm": 0.371177613735199, "learning_rate": 9.75482251562417e-05, "loss": 0.0267, "step": 14510 }, { "grad_norm": 0.293840229511261, "learning_rate": 9.754310837365155e-05, "loss": 0.0302, "step": 14520 }, { "grad_norm": 0.2704843580722809, "learning_rate": 9.753798639181856e-05, "loss": 0.0294, "step": 14530 }, { "grad_norm": 0.2714877426624298, "learning_rate": 9.753285921130286e-05, "loss": 0.0277, "step": 14540 }, { "grad_norm": 0.25895676016807556, "learning_rate": 9.752772683266512e-05, "loss": 0.0289, "step": 14550 }, { "grad_norm": 0.3167363703250885, "learning_rate": 9.752258925646665e-05, "loss": 0.0287, "step": 14560 }, { "grad_norm": 0.2814650237560272, "learning_rate": 9.751744648326926e-05, "loss": 0.0242, "step": 14570 }, { "grad_norm": 0.3629542589187622, "learning_rate": 9.751229851363536e-05, "loss": 0.0259, "step": 14580 }, { "grad_norm": 0.30948013067245483, "learning_rate": 9.750714534812793e-05, "loss": 0.0271, "step": 14590 }, { "grad_norm": 0.2360362410545349, "learning_rate": 9.750198698731053e-05, "loss": 0.025, "step": 14600 }, { "grad_norm": 0.19960370659828186, "learning_rate": 9.749682343174722e-05, "loss": 0.0267, "step": 14610 }, { "grad_norm": 0.2320825457572937, "learning_rate": 9.749165468200272e-05, "loss": 0.0249, "step": 14620 }, { "grad_norm": 0.2052048295736313, "learning_rate": 9.748648073864229e-05, "loss": 0.0232, "step": 14630 }, { "grad_norm": 0.24731223285198212, "learning_rate": 9.748130160223168e-05, "loss": 0.026, "step": 14640 }, { "grad_norm": 0.2553905248641968, "learning_rate": 9.747611727333734e-05, "loss": 0.0266, "step": 14650 }, { "grad_norm": 0.2402094602584839, "learning_rate": 9.74709277525262e-05, "loss": 0.026, "step": 14660 }, { "grad_norm": 0.273367702960968, "learning_rate": 9.746573304036576e-05, "loss": 0.0249, "step": 14670 }, { "grad_norm": 0.24164967238903046, "learning_rate": 9.746053313742412e-05, "loss": 0.0246, "step": 14680 }, { "grad_norm": 0.2228330671787262, "learning_rate": 9.745532804426994e-05, "loss": 0.0229, "step": 14690 }, { "grad_norm": 0.25918543338775635, "learning_rate": 9.745011776147242e-05, "loss": 0.0265, "step": 14700 }, { "grad_norm": 0.29919809103012085, "learning_rate": 9.744490228960138e-05, "loss": 0.0245, "step": 14710 }, { "grad_norm": 0.280224084854126, "learning_rate": 9.743968162922713e-05, "loss": 0.0268, "step": 14720 }, { "grad_norm": 0.3180544972419739, "learning_rate": 9.743445578092064e-05, "loss": 0.0275, "step": 14730 }, { "grad_norm": 0.27898263931274414, "learning_rate": 9.742922474525338e-05, "loss": 0.0239, "step": 14740 }, { "grad_norm": 0.21459542214870453, "learning_rate": 9.742398852279741e-05, "loss": 0.0227, "step": 14750 }, { "grad_norm": 0.20613984763622284, "learning_rate": 9.741874711412535e-05, "loss": 0.0276, "step": 14760 }, { "grad_norm": 0.21722948551177979, "learning_rate": 9.741350051981042e-05, "loss": 0.0238, "step": 14770 }, { "grad_norm": 0.2510583698749542, "learning_rate": 9.740824874042633e-05, "loss": 0.0238, "step": 14780 }, { "grad_norm": 0.24857421219348907, "learning_rate": 9.740299177654746e-05, "loss": 0.0228, "step": 14790 }, { "grad_norm": 0.2684057950973511, "learning_rate": 9.739772962874867e-05, "loss": 0.0265, "step": 14800 }, { "grad_norm": 0.1794997602701187, "learning_rate": 9.739246229760541e-05, "loss": 0.0239, "step": 14810 }, { "grad_norm": 0.21764685213565826, "learning_rate": 9.738718978369376e-05, "loss": 0.0234, "step": 14820 }, { "grad_norm": 0.24092507362365723, "learning_rate": 9.738191208759025e-05, "loss": 0.0242, "step": 14830 }, { "grad_norm": 0.20457983016967773, "learning_rate": 9.73766292098721e-05, "loss": 0.0258, "step": 14840 }, { "grad_norm": 0.19467096030712128, "learning_rate": 9.737134115111699e-05, "loss": 0.023, "step": 14850 }, { "grad_norm": 0.20268258452415466, "learning_rate": 9.736604791190323e-05, "loss": 0.0255, "step": 14860 }, { "grad_norm": 0.2857716977596283, "learning_rate": 9.73607494928097e-05, "loss": 0.0278, "step": 14870 }, { "grad_norm": 0.2278226912021637, "learning_rate": 9.735544589441581e-05, "loss": 0.0262, "step": 14880 }, { "grad_norm": 0.18258064985275269, "learning_rate": 9.735013711730154e-05, "loss": 0.024, "step": 14890 }, { "grad_norm": 0.3791099786758423, "learning_rate": 9.734482316204747e-05, "loss": 0.0282, "step": 14900 }, { "grad_norm": 0.3237806260585785, "learning_rate": 9.733950402923473e-05, "loss": 0.0262, "step": 14910 }, { "grad_norm": 0.31332725286483765, "learning_rate": 9.7334179719445e-05, "loss": 0.0259, "step": 14920 }, { "grad_norm": 0.8750212788581848, "learning_rate": 9.732885023326053e-05, "loss": 0.0288, "step": 14930 }, { "grad_norm": 0.23649397492408752, "learning_rate": 9.732351557126418e-05, "loss": 0.0271, "step": 14940 }, { "grad_norm": 0.2617371678352356, "learning_rate": 9.731817573403929e-05, "loss": 0.0274, "step": 14950 }, { "grad_norm": 0.15922439098358154, "learning_rate": 9.731283072216985e-05, "loss": 0.0242, "step": 14960 }, { "grad_norm": 0.18040475249290466, "learning_rate": 9.730748053624039e-05, "loss": 0.0266, "step": 14970 }, { "grad_norm": 0.25573572516441345, "learning_rate": 9.730212517683598e-05, "loss": 0.0253, "step": 14980 }, { "grad_norm": 0.218913733959198, "learning_rate": 9.729676464454228e-05, "loss": 0.0251, "step": 14990 }, { "grad_norm": 0.3169909715652466, "learning_rate": 9.72913989399455e-05, "loss": 0.0282, "step": 15000 }, { "grad_norm": 0.23736126720905304, "learning_rate": 9.728602806363242e-05, "loss": 0.0309, "step": 15010 }, { "grad_norm": 0.30006545782089233, "learning_rate": 9.728065201619043e-05, "loss": 0.0308, "step": 15020 }, { "grad_norm": 0.21722596883773804, "learning_rate": 9.727527079820742e-05, "loss": 0.0237, "step": 15030 }, { "grad_norm": 0.2828999161720276, "learning_rate": 9.726988441027186e-05, "loss": 0.0253, "step": 15040 }, { "grad_norm": 0.2314092218875885, "learning_rate": 9.726449285297281e-05, "loss": 0.0262, "step": 15050 }, { "grad_norm": 0.45644837617874146, "learning_rate": 9.72590961268999e-05, "loss": 0.0257, "step": 15060 }, { "grad_norm": 0.38552969694137573, "learning_rate": 9.725369423264328e-05, "loss": 0.0296, "step": 15070 }, { "grad_norm": 0.2039709836244583, "learning_rate": 9.72482871707937e-05, "loss": 0.0293, "step": 15080 }, { "grad_norm": 0.6773009896278381, "learning_rate": 9.724287494194247e-05, "loss": 0.0292, "step": 15090 }, { "grad_norm": 0.4016192555427551, "learning_rate": 9.723745754668147e-05, "loss": 0.0358, "step": 15100 }, { "grad_norm": 0.40302902460098267, "learning_rate": 9.723203498560313e-05, "loss": 0.0374, "step": 15110 }, { "grad_norm": 0.3084258437156677, "learning_rate": 9.722660725930046e-05, "loss": 0.0301, "step": 15120 }, { "grad_norm": 0.3071240782737732, "learning_rate": 9.722117436836702e-05, "loss": 0.03, "step": 15130 }, { "grad_norm": 0.30342426896095276, "learning_rate": 9.721573631339696e-05, "loss": 0.032, "step": 15140 }, { "grad_norm": 0.3329104483127594, "learning_rate": 9.721029309498494e-05, "loss": 0.0343, "step": 15150 }, { "grad_norm": 0.28422683477401733, "learning_rate": 9.720484471372627e-05, "loss": 0.0333, "step": 15160 }, { "grad_norm": 0.27779167890548706, "learning_rate": 9.719939117021673e-05, "loss": 0.0313, "step": 15170 }, { "grad_norm": 0.29406648874282837, "learning_rate": 9.719393246505275e-05, "loss": 0.0285, "step": 15180 }, { "grad_norm": 0.2636095881462097, "learning_rate": 9.718846859883128e-05, "loss": 0.0312, "step": 15190 }, { "grad_norm": 0.20598973333835602, "learning_rate": 9.718299957214982e-05, "loss": 0.0287, "step": 15200 }, { "grad_norm": 0.5159875750541687, "learning_rate": 9.717752538560646e-05, "loss": 0.0286, "step": 15210 }, { "grad_norm": 0.3080006241798401, "learning_rate": 9.717204603979986e-05, "loss": 0.0283, "step": 15220 }, { "grad_norm": 0.23664507269859314, "learning_rate": 9.716656153532922e-05, "loss": 0.0292, "step": 15230 }, { "grad_norm": 0.27141323685646057, "learning_rate": 9.716107187279434e-05, "loss": 0.0279, "step": 15240 }, { "grad_norm": 0.21128486096858978, "learning_rate": 9.715557705279555e-05, "loss": 0.0258, "step": 15250 }, { "grad_norm": 0.2491716891527176, "learning_rate": 9.715007707593372e-05, "loss": 0.0247, "step": 15260 }, { "grad_norm": 0.3089141547679901, "learning_rate": 9.714457194281036e-05, "loss": 0.0276, "step": 15270 }, { "grad_norm": 0.2869704067707062, "learning_rate": 9.713906165402751e-05, "loss": 0.0253, "step": 15280 }, { "grad_norm": 0.23893170058727264, "learning_rate": 9.713354621018774e-05, "loss": 0.0267, "step": 15290 }, { "grad_norm": 0.22661148011684418, "learning_rate": 9.712802561189422e-05, "loss": 0.0267, "step": 15300 }, { "grad_norm": 0.23626399040222168, "learning_rate": 9.712249985975069e-05, "loss": 0.0249, "step": 15310 }, { "grad_norm": 0.24257296323776245, "learning_rate": 9.71169689543614e-05, "loss": 0.0231, "step": 15320 }, { "grad_norm": 0.2603594660758972, "learning_rate": 9.711143289633123e-05, "loss": 0.0247, "step": 15330 }, { "grad_norm": 0.27820709347724915, "learning_rate": 9.710589168626561e-05, "loss": 0.0242, "step": 15340 }, { "grad_norm": 0.3044043779373169, "learning_rate": 9.710034532477048e-05, "loss": 0.0255, "step": 15350 }, { "grad_norm": 0.17697829008102417, "learning_rate": 9.709479381245239e-05, "loss": 0.0247, "step": 15360 }, { "grad_norm": 0.2389088124036789, "learning_rate": 9.708923714991847e-05, "loss": 0.0235, "step": 15370 }, { "grad_norm": 0.2567511796951294, "learning_rate": 9.708367533777638e-05, "loss": 0.0243, "step": 15380 }, { "grad_norm": 0.3237888514995575, "learning_rate": 9.707810837663431e-05, "loss": 0.0287, "step": 15390 }, { "grad_norm": 0.25600695610046387, "learning_rate": 9.707253626710113e-05, "loss": 0.0268, "step": 15400 }, { "grad_norm": 0.22382090985774994, "learning_rate": 9.706695900978613e-05, "loss": 0.0226, "step": 15410 }, { "grad_norm": 0.2455943077802658, "learning_rate": 9.706137660529926e-05, "loss": 0.0279, "step": 15420 }, { "grad_norm": 0.25378745794296265, "learning_rate": 9.705578905425101e-05, "loss": 0.0258, "step": 15430 }, { "grad_norm": 0.3113659918308258, "learning_rate": 9.705019635725241e-05, "loss": 0.0275, "step": 15440 }, { "grad_norm": 0.317214697599411, "learning_rate": 9.704459851491508e-05, "loss": 0.0243, "step": 15450 }, { "grad_norm": 0.24126562476158142, "learning_rate": 9.703899552785118e-05, "loss": 0.0239, "step": 15460 }, { "grad_norm": 0.2407691776752472, "learning_rate": 9.703338739667346e-05, "loss": 0.0216, "step": 15470 }, { "grad_norm": 0.23882310092449188, "learning_rate": 9.70277741219952e-05, "loss": 0.0261, "step": 15480 }, { "grad_norm": 0.21090766787528992, "learning_rate": 9.702215570443027e-05, "loss": 0.0255, "step": 15490 }, { "grad_norm": 0.2377903014421463, "learning_rate": 9.701653214459309e-05, "loss": 0.0274, "step": 15500 }, { "grad_norm": 0.3233996331691742, "learning_rate": 9.701090344309865e-05, "loss": 0.0269, "step": 15510 }, { "grad_norm": 0.3248761296272278, "learning_rate": 9.700526960056247e-05, "loss": 0.0307, "step": 15520 }, { "grad_norm": 0.26304394006729126, "learning_rate": 9.699963061760068e-05, "loss": 0.0262, "step": 15530 }, { "grad_norm": 0.2672145366668701, "learning_rate": 9.699398649482997e-05, "loss": 0.0282, "step": 15540 }, { "grad_norm": 0.27459922432899475, "learning_rate": 9.698833723286753e-05, "loss": 0.0271, "step": 15550 }, { "grad_norm": 0.2720268666744232, "learning_rate": 9.698268283233118e-05, "loss": 0.027, "step": 15560 }, { "grad_norm": 0.2100878655910492, "learning_rate": 9.697702329383929e-05, "loss": 0.0263, "step": 15570 }, { "grad_norm": 0.3461340069770813, "learning_rate": 9.697135861801074e-05, "loss": 0.0283, "step": 15580 }, { "grad_norm": 0.21751131117343903, "learning_rate": 9.696568880546505e-05, "loss": 0.023, "step": 15590 }, { "grad_norm": 0.2761978507041931, "learning_rate": 9.696001385682223e-05, "loss": 0.0248, "step": 15600 }, { "grad_norm": 0.23768578469753265, "learning_rate": 9.695433377270291e-05, "loss": 0.0252, "step": 15610 }, { "grad_norm": 0.25364890694618225, "learning_rate": 9.694864855372824e-05, "loss": 0.0248, "step": 15620 }, { "grad_norm": 0.26440510153770447, "learning_rate": 9.694295820051995e-05, "loss": 0.0232, "step": 15630 }, { "grad_norm": 0.3544156551361084, "learning_rate": 9.693726271370032e-05, "loss": 0.0245, "step": 15640 }, { "grad_norm": 0.21234455704689026, "learning_rate": 9.693156209389221e-05, "loss": 0.025, "step": 15650 }, { "grad_norm": 0.25144025683403015, "learning_rate": 9.692585634171905e-05, "loss": 0.0235, "step": 15660 }, { "grad_norm": 0.21116316318511963, "learning_rate": 9.692014545780476e-05, "loss": 0.0236, "step": 15670 }, { "grad_norm": 0.42943719029426575, "learning_rate": 9.691442944277393e-05, "loss": 0.0259, "step": 15680 }, { "grad_norm": 0.2703126072883606, "learning_rate": 9.690870829725162e-05, "loss": 0.0235, "step": 15690 }, { "grad_norm": 0.2559621036052704, "learning_rate": 9.69029820218635e-05, "loss": 0.0239, "step": 15700 }, { "grad_norm": 0.2158404141664505, "learning_rate": 9.689725061723579e-05, "loss": 0.0253, "step": 15710 }, { "grad_norm": 0.24819320440292358, "learning_rate": 9.689151408399527e-05, "loss": 0.0243, "step": 15720 }, { "grad_norm": 0.25043392181396484, "learning_rate": 9.688577242276924e-05, "loss": 0.0257, "step": 15730 }, { "grad_norm": 0.29676613211631775, "learning_rate": 9.688002563418566e-05, "loss": 0.0249, "step": 15740 }, { "grad_norm": 0.34713512659072876, "learning_rate": 9.687427371887293e-05, "loss": 0.0259, "step": 15750 }, { "grad_norm": 0.38211914896965027, "learning_rate": 9.686851667746012e-05, "loss": 0.0249, "step": 15760 }, { "grad_norm": 0.3440866768360138, "learning_rate": 9.686275451057677e-05, "loss": 0.0267, "step": 15770 }, { "grad_norm": 0.23102571070194244, "learning_rate": 9.685698721885308e-05, "loss": 0.0252, "step": 15780 }, { "grad_norm": 0.26415324211120605, "learning_rate": 9.68512148029197e-05, "loss": 0.0259, "step": 15790 }, { "grad_norm": 0.24765951931476593, "learning_rate": 9.684543726340791e-05, "loss": 0.0269, "step": 15800 }, { "grad_norm": 0.35656413435935974, "learning_rate": 9.683965460094952e-05, "loss": 0.0233, "step": 15810 }, { "grad_norm": 0.22515933215618134, "learning_rate": 9.683386681617694e-05, "loss": 0.0246, "step": 15820 }, { "grad_norm": 0.25189223885536194, "learning_rate": 9.68280739097231e-05, "loss": 0.0263, "step": 15830 }, { "grad_norm": 0.2874066233634949, "learning_rate": 9.682227588222148e-05, "loss": 0.0331, "step": 15840 }, { "grad_norm": 0.26723015308380127, "learning_rate": 9.681647273430618e-05, "loss": 0.0296, "step": 15850 }, { "grad_norm": 0.3699802756309509, "learning_rate": 9.681066446661182e-05, "loss": 0.0321, "step": 15860 }, { "grad_norm": 0.3043660521507263, "learning_rate": 9.680485107977357e-05, "loss": 0.0288, "step": 15870 }, { "grad_norm": 0.25948309898376465, "learning_rate": 9.679903257442716e-05, "loss": 0.0277, "step": 15880 }, { "grad_norm": 0.2729259133338928, "learning_rate": 9.679320895120891e-05, "loss": 0.0297, "step": 15890 }, { "grad_norm": 0.3144776225090027, "learning_rate": 9.67873802107557e-05, "loss": 0.027, "step": 15900 }, { "grad_norm": 0.237562894821167, "learning_rate": 9.67815463537049e-05, "loss": 0.0257, "step": 15910 }, { "grad_norm": 0.30666384100914, "learning_rate": 9.677570738069457e-05, "loss": 0.026, "step": 15920 }, { "grad_norm": 0.24640695750713348, "learning_rate": 9.676986329236318e-05, "loss": 0.0268, "step": 15930 }, { "grad_norm": 0.2833688259124756, "learning_rate": 9.676401408934987e-05, "loss": 0.0244, "step": 15940 }, { "grad_norm": 0.3118480145931244, "learning_rate": 9.675815977229428e-05, "loss": 0.0283, "step": 15950 }, { "grad_norm": 0.25482019782066345, "learning_rate": 9.675230034183664e-05, "loss": 0.0277, "step": 15960 }, { "grad_norm": 0.2685750424861908, "learning_rate": 9.674643579861773e-05, "loss": 0.0309, "step": 15970 }, { "grad_norm": 0.3360207676887512, "learning_rate": 9.674056614327886e-05, "loss": 0.0305, "step": 15980 }, { "grad_norm": 0.24730023741722107, "learning_rate": 9.673469137646198e-05, "loss": 0.0324, "step": 15990 }, { "grad_norm": 0.3209632337093353, "learning_rate": 9.67288114988095e-05, "loss": 0.028, "step": 16000 }, { "grad_norm": 0.2786257863044739, "learning_rate": 9.672292651096447e-05, "loss": 0.0265, "step": 16010 }, { "grad_norm": 0.34927091002464294, "learning_rate": 9.671703641357042e-05, "loss": 0.0288, "step": 16020 }, { "grad_norm": 0.23046036064624786, "learning_rate": 9.67111412072715e-05, "loss": 0.0256, "step": 16030 }, { "grad_norm": 0.320446252822876, "learning_rate": 9.670524089271242e-05, "loss": 0.0254, "step": 16040 }, { "grad_norm": 0.24565909802913666, "learning_rate": 9.669933547053842e-05, "loss": 0.026, "step": 16050 }, { "grad_norm": 0.2877464294433594, "learning_rate": 9.669342494139531e-05, "loss": 0.0268, "step": 16060 }, { "grad_norm": 0.310435026884079, "learning_rate": 9.668750930592943e-05, "loss": 0.0226, "step": 16070 }, { "grad_norm": 0.35193443298339844, "learning_rate": 9.668158856478775e-05, "loss": 0.0248, "step": 16080 }, { "grad_norm": 0.2796499729156494, "learning_rate": 9.66756627186177e-05, "loss": 0.027, "step": 16090 }, { "grad_norm": 0.2640926241874695, "learning_rate": 9.666973176806737e-05, "loss": 0.0223, "step": 16100 }, { "grad_norm": 0.21608005464076996, "learning_rate": 9.666379571378534e-05, "loss": 0.0227, "step": 16110 }, { "grad_norm": 0.2604767978191376, "learning_rate": 9.665785455642076e-05, "loss": 0.0245, "step": 16120 }, { "grad_norm": 0.3694501221179962, "learning_rate": 9.665190829662337e-05, "loss": 0.0246, "step": 16130 }, { "grad_norm": 0.2327137589454651, "learning_rate": 9.664595693504342e-05, "loss": 0.024, "step": 16140 }, { "grad_norm": 0.19723820686340332, "learning_rate": 9.664000047233175e-05, "loss": 0.0231, "step": 16150 }, { "grad_norm": 0.3166446387767792, "learning_rate": 9.663403890913976e-05, "loss": 0.0246, "step": 16160 }, { "grad_norm": 0.24948540329933167, "learning_rate": 9.662807224611938e-05, "loss": 0.0247, "step": 16170 }, { "grad_norm": 0.2522730231285095, "learning_rate": 9.662210048392311e-05, "loss": 0.0249, "step": 16180 }, { "grad_norm": 0.3129309415817261, "learning_rate": 9.661612362320405e-05, "loss": 0.0237, "step": 16190 }, { "grad_norm": 0.24813595414161682, "learning_rate": 9.661014166461579e-05, "loss": 0.026, "step": 16200 }, { "grad_norm": 0.397930771112442, "learning_rate": 9.66041546088125e-05, "loss": 0.0267, "step": 16210 }, { "grad_norm": 0.2584666609764099, "learning_rate": 9.659816245644895e-05, "loss": 0.028, "step": 16220 }, { "grad_norm": 0.2035432904958725, "learning_rate": 9.65921652081804e-05, "loss": 0.0248, "step": 16230 }, { "grad_norm": 0.22770380973815918, "learning_rate": 9.658616286466271e-05, "loss": 0.0228, "step": 16240 }, { "grad_norm": 0.25026941299438477, "learning_rate": 9.65801554265523e-05, "loss": 0.0261, "step": 16250 }, { "grad_norm": 0.21433694660663605, "learning_rate": 9.657414289450612e-05, "loss": 0.026, "step": 16260 }, { "grad_norm": 0.2963101267814636, "learning_rate": 9.656812526918171e-05, "loss": 0.027, "step": 16270 }, { "grad_norm": 0.26199445128440857, "learning_rate": 9.656210255123712e-05, "loss": 0.025, "step": 16280 }, { "grad_norm": 0.2822970747947693, "learning_rate": 9.6556074741331e-05, "loss": 0.0289, "step": 16290 }, { "grad_norm": 0.2604855000972748, "learning_rate": 9.655004184012256e-05, "loss": 0.0261, "step": 16300 }, { "grad_norm": 0.27265664935112, "learning_rate": 9.654400384827152e-05, "loss": 0.0273, "step": 16310 }, { "grad_norm": 0.30491742491722107, "learning_rate": 9.653796076643818e-05, "loss": 0.0337, "step": 16320 }, { "grad_norm": 0.2581965923309326, "learning_rate": 9.653191259528344e-05, "loss": 0.0307, "step": 16330 }, { "grad_norm": 0.22685234248638153, "learning_rate": 9.65258593354687e-05, "loss": 0.0285, "step": 16340 }, { "grad_norm": 0.2650866210460663, "learning_rate": 9.651980098765591e-05, "loss": 0.0265, "step": 16350 }, { "grad_norm": 0.2608899474143982, "learning_rate": 9.651373755250765e-05, "loss": 0.0267, "step": 16360 }, { "grad_norm": 0.2218223512172699, "learning_rate": 9.650766903068697e-05, "loss": 0.0252, "step": 16370 }, { "grad_norm": 0.2446441352367401, "learning_rate": 9.650159542285753e-05, "loss": 0.0242, "step": 16380 }, { "grad_norm": 0.29566824436187744, "learning_rate": 9.649551672968353e-05, "loss": 0.0275, "step": 16390 }, { "grad_norm": 0.26149746775627136, "learning_rate": 9.648943295182973e-05, "loss": 0.0255, "step": 16400 }, { "grad_norm": 0.26039376854896545, "learning_rate": 9.648334408996144e-05, "loss": 0.0246, "step": 16410 }, { "grad_norm": 0.3348585367202759, "learning_rate": 9.647725014474452e-05, "loss": 0.0269, "step": 16420 }, { "grad_norm": 0.3292948305606842, "learning_rate": 9.64711511168454e-05, "loss": 0.0285, "step": 16430 }, { "grad_norm": 0.24014924466609955, "learning_rate": 9.646504700693108e-05, "loss": 0.0266, "step": 16440 }, { "grad_norm": 0.224380224943161, "learning_rate": 9.645893781566907e-05, "loss": 0.027, "step": 16450 }, { "grad_norm": 0.29588115215301514, "learning_rate": 9.645282354372744e-05, "loss": 0.0252, "step": 16460 }, { "grad_norm": 0.24583475291728973, "learning_rate": 9.644670419177491e-05, "loss": 0.025, "step": 16470 }, { "grad_norm": 0.2722179889678955, "learning_rate": 9.644057976048062e-05, "loss": 0.0274, "step": 16480 }, { "grad_norm": 0.25196564197540283, "learning_rate": 9.643445025051435e-05, "loss": 0.0265, "step": 16490 }, { "grad_norm": 0.3103501498699188, "learning_rate": 9.642831566254641e-05, "loss": 0.0228, "step": 16500 }, { "grad_norm": 0.265829473733902, "learning_rate": 9.642217599724769e-05, "loss": 0.0246, "step": 16510 }, { "grad_norm": 0.22827103734016418, "learning_rate": 9.64160312552896e-05, "loss": 0.0222, "step": 16520 }, { "grad_norm": 0.19888588786125183, "learning_rate": 9.64098814373441e-05, "loss": 0.0233, "step": 16530 }, { "grad_norm": 0.22777217626571655, "learning_rate": 9.640372654408374e-05, "loss": 0.0228, "step": 16540 }, { "grad_norm": 0.281894713640213, "learning_rate": 9.639756657618162e-05, "loss": 0.0228, "step": 16550 }, { "grad_norm": 0.2682880163192749, "learning_rate": 9.639140153431138e-05, "loss": 0.0237, "step": 16560 }, { "grad_norm": 0.26230159401893616, "learning_rate": 9.638523141914721e-05, "loss": 0.0243, "step": 16570 }, { "grad_norm": 0.2752503752708435, "learning_rate": 9.637905623136388e-05, "loss": 0.0253, "step": 16580 }, { "grad_norm": 0.23685240745544434, "learning_rate": 9.637287597163669e-05, "loss": 0.0235, "step": 16590 }, { "grad_norm": 0.19962050020694733, "learning_rate": 9.63666906406415e-05, "loss": 0.0269, "step": 16600 }, { "grad_norm": 0.23826679587364197, "learning_rate": 9.636050023905473e-05, "loss": 0.0224, "step": 16610 }, { "grad_norm": 0.22987520694732666, "learning_rate": 9.635430476755336e-05, "loss": 0.0224, "step": 16620 }, { "grad_norm": 0.22022569179534912, "learning_rate": 9.63481042268149e-05, "loss": 0.0233, "step": 16630 }, { "grad_norm": 0.20053786039352417, "learning_rate": 9.634189861751745e-05, "loss": 0.0236, "step": 16640 }, { "grad_norm": 0.4208022356033325, "learning_rate": 9.633568794033967e-05, "loss": 0.0237, "step": 16650 }, { "grad_norm": 0.2143082171678543, "learning_rate": 9.63294721959607e-05, "loss": 0.0239, "step": 16660 }, { "grad_norm": 0.286199152469635, "learning_rate": 9.63232513850603e-05, "loss": 0.026, "step": 16670 }, { "grad_norm": 0.2065124362707138, "learning_rate": 9.631702550831878e-05, "loss": 0.0249, "step": 16680 }, { "grad_norm": 0.2712841331958771, "learning_rate": 9.631079456641698e-05, "loss": 0.0303, "step": 16690 }, { "grad_norm": 0.252485066652298, "learning_rate": 9.630455856003632e-05, "loss": 0.0247, "step": 16700 }, { "grad_norm": 0.22359341382980347, "learning_rate": 9.629831748985876e-05, "loss": 0.0246, "step": 16710 }, { "grad_norm": 0.23502680659294128, "learning_rate": 9.629207135656679e-05, "loss": 0.024, "step": 16720 }, { "grad_norm": 0.28517085313796997, "learning_rate": 9.628582016084353e-05, "loss": 0.0238, "step": 16730 }, { "grad_norm": 0.24674977362155914, "learning_rate": 9.627956390337254e-05, "loss": 0.0241, "step": 16740 }, { "grad_norm": 0.22086431086063385, "learning_rate": 9.627330258483802e-05, "loss": 0.0265, "step": 16750 }, { "grad_norm": 0.22116635739803314, "learning_rate": 9.62670362059247e-05, "loss": 0.0233, "step": 16760 }, { "grad_norm": 0.2560024559497833, "learning_rate": 9.626076476731786e-05, "loss": 0.0244, "step": 16770 }, { "grad_norm": 0.26768654584884644, "learning_rate": 9.625448826970336e-05, "loss": 0.0266, "step": 16780 }, { "grad_norm": 0.24242940545082092, "learning_rate": 9.624820671376755e-05, "loss": 0.0255, "step": 16790 }, { "grad_norm": 0.2415277063846588, "learning_rate": 9.62419201001974e-05, "loss": 0.0288, "step": 16800 }, { "grad_norm": 0.28798121213912964, "learning_rate": 9.623562842968037e-05, "loss": 0.0271, "step": 16810 }, { "grad_norm": 0.21918408572673798, "learning_rate": 9.622933170290454e-05, "loss": 0.027, "step": 16820 }, { "grad_norm": 0.2818245589733124, "learning_rate": 9.622302992055849e-05, "loss": 0.0259, "step": 16830 }, { "grad_norm": 0.1963837742805481, "learning_rate": 9.62167230833314e-05, "loss": 0.0243, "step": 16840 }, { "grad_norm": 0.2762114405632019, "learning_rate": 9.621041119191295e-05, "loss": 0.0228, "step": 16850 }, { "grad_norm": 0.2676524221897125, "learning_rate": 9.620409424699342e-05, "loss": 0.0239, "step": 16860 }, { "grad_norm": 0.24672770500183105, "learning_rate": 9.619777224926359e-05, "loss": 0.0233, "step": 16870 }, { "grad_norm": 0.19182518124580383, "learning_rate": 9.619144519941485e-05, "loss": 0.0226, "step": 16880 }, { "grad_norm": 0.23151932656764984, "learning_rate": 9.618511309813912e-05, "loss": 0.0242, "step": 16890 }, { "grad_norm": 0.23240935802459717, "learning_rate": 9.617877594612886e-05, "loss": 0.0251, "step": 16900 }, { "grad_norm": 0.21326564252376556, "learning_rate": 9.617243374407707e-05, "loss": 0.0231, "step": 16910 }, { "grad_norm": 0.23542742431163788, "learning_rate": 9.616608649267736e-05, "loss": 0.026, "step": 16920 }, { "grad_norm": 0.2393019050359726, "learning_rate": 9.615973419262385e-05, "loss": 0.023, "step": 16930 }, { "grad_norm": 0.2116704136133194, "learning_rate": 9.615337684461119e-05, "loss": 0.0212, "step": 16940 }, { "grad_norm": 0.2640947103500366, "learning_rate": 9.614701444933465e-05, "loss": 0.0216, "step": 16950 }, { "grad_norm": 0.2307904213666916, "learning_rate": 9.614064700748997e-05, "loss": 0.0213, "step": 16960 }, { "grad_norm": 0.21338951587677002, "learning_rate": 9.613427451977352e-05, "loss": 0.0197, "step": 16970 }, { "grad_norm": 0.2637673020362854, "learning_rate": 9.612789698688216e-05, "loss": 0.0227, "step": 16980 }, { "grad_norm": 0.388332724571228, "learning_rate": 9.612151440951334e-05, "loss": 0.0258, "step": 16990 }, { "grad_norm": 0.23311002552509308, "learning_rate": 9.611512678836506e-05, "loss": 0.0242, "step": 17000 }, { "grad_norm": 0.22417277097702026, "learning_rate": 9.610873412413584e-05, "loss": 0.0233, "step": 17010 }, { "grad_norm": 0.31707462668418884, "learning_rate": 9.610233641752476e-05, "loss": 0.0259, "step": 17020 }, { "grad_norm": 0.21677745878696442, "learning_rate": 9.609593366923151e-05, "loss": 0.0262, "step": 17030 }, { "grad_norm": 0.26738184690475464, "learning_rate": 9.608952587995625e-05, "loss": 0.0257, "step": 17040 }, { "grad_norm": 0.24097108840942383, "learning_rate": 9.608311305039972e-05, "loss": 0.0225, "step": 17050 }, { "grad_norm": 0.26996615529060364, "learning_rate": 9.607669518126326e-05, "loss": 0.0285, "step": 17060 }, { "grad_norm": 0.24188795685768127, "learning_rate": 9.607027227324866e-05, "loss": 0.0229, "step": 17070 }, { "grad_norm": 0.35718002915382385, "learning_rate": 9.606384432705837e-05, "loss": 0.0261, "step": 17080 }, { "grad_norm": 0.30223774909973145, "learning_rate": 9.60574113433953e-05, "loss": 0.0254, "step": 17090 }, { "grad_norm": 0.26171353459358215, "learning_rate": 9.6050973322963e-05, "loss": 0.0264, "step": 17100 }, { "grad_norm": 0.19304120540618896, "learning_rate": 9.604453026646547e-05, "loss": 0.0254, "step": 17110 }, { "grad_norm": 0.19676756858825684, "learning_rate": 9.603808217460735e-05, "loss": 0.0213, "step": 17120 }, { "grad_norm": 0.1761104315519333, "learning_rate": 9.603162904809377e-05, "loss": 0.0234, "step": 17130 }, { "grad_norm": 0.27313050627708435, "learning_rate": 9.602517088763045e-05, "loss": 0.025, "step": 17140 }, { "grad_norm": 0.3402903974056244, "learning_rate": 9.601870769392365e-05, "loss": 0.0243, "step": 17150 }, { "grad_norm": 0.46713787317276, "learning_rate": 9.601223946768017e-05, "loss": 0.0246, "step": 17160 }, { "grad_norm": 0.27218377590179443, "learning_rate": 9.600576620960734e-05, "loss": 0.0246, "step": 17170 }, { "grad_norm": 0.21105411648750305, "learning_rate": 9.599928792041308e-05, "loss": 0.0271, "step": 17180 }, { "grad_norm": 0.1683046519756317, "learning_rate": 9.599280460080587e-05, "loss": 0.0245, "step": 17190 }, { "grad_norm": 0.18594104051589966, "learning_rate": 9.59863162514947e-05, "loss": 0.022, "step": 17200 }, { "grad_norm": 0.2276935577392578, "learning_rate": 9.597982287318911e-05, "loss": 0.0247, "step": 17210 }, { "grad_norm": 0.327534943819046, "learning_rate": 9.597332446659923e-05, "loss": 0.0247, "step": 17220 }, { "grad_norm": 0.250415176153183, "learning_rate": 9.59668210324357e-05, "loss": 0.0249, "step": 17230 }, { "grad_norm": 0.2805004119873047, "learning_rate": 9.596031257140974e-05, "loss": 0.0218, "step": 17240 }, { "grad_norm": 0.23513087630271912, "learning_rate": 9.59537990842331e-05, "loss": 0.0231, "step": 17250 }, { "grad_norm": 0.27084770798683167, "learning_rate": 9.594728057161806e-05, "loss": 0.0238, "step": 17260 }, { "grad_norm": 0.25677821040153503, "learning_rate": 9.594075703427752e-05, "loss": 0.0213, "step": 17270 }, { "grad_norm": 0.22199273109436035, "learning_rate": 9.593422847292486e-05, "loss": 0.0239, "step": 17280 }, { "grad_norm": 0.24294215440750122, "learning_rate": 9.592769488827402e-05, "loss": 0.0235, "step": 17290 }, { "grad_norm": 0.2270650863647461, "learning_rate": 9.592115628103952e-05, "loss": 0.0245, "step": 17300 }, { "grad_norm": 0.22610215842723846, "learning_rate": 9.591461265193643e-05, "loss": 0.0252, "step": 17310 }, { "grad_norm": 0.2031104862689972, "learning_rate": 9.590806400168032e-05, "loss": 0.0246, "step": 17320 }, { "grad_norm": 0.22986462712287903, "learning_rate": 9.590151033098735e-05, "loss": 0.0238, "step": 17330 }, { "grad_norm": 0.2180652767419815, "learning_rate": 9.589495164057423e-05, "loss": 0.0209, "step": 17340 }, { "grad_norm": 0.2166883796453476, "learning_rate": 9.58883879311582e-05, "loss": 0.0231, "step": 17350 }, { "grad_norm": 0.27155372500419617, "learning_rate": 9.588181920345705e-05, "loss": 0.022, "step": 17360 }, { "grad_norm": 0.33735010027885437, "learning_rate": 9.587524545818913e-05, "loss": 0.024, "step": 17370 }, { "grad_norm": 0.22960583865642548, "learning_rate": 9.586866669607335e-05, "loss": 0.0253, "step": 17380 }, { "grad_norm": 0.23630207777023315, "learning_rate": 9.586208291782915e-05, "loss": 0.0234, "step": 17390 }, { "grad_norm": 0.20598384737968445, "learning_rate": 9.58554941241765e-05, "loss": 0.0249, "step": 17400 }, { "grad_norm": 0.19727346301078796, "learning_rate": 9.584890031583596e-05, "loss": 0.0221, "step": 17410 }, { "grad_norm": 0.5024396181106567, "learning_rate": 9.584230149352861e-05, "loss": 0.0213, "step": 17420 }, { "grad_norm": 0.321027010679245, "learning_rate": 9.58356976579761e-05, "loss": 0.0208, "step": 17430 }, { "grad_norm": 0.2179991900920868, "learning_rate": 9.58290888099006e-05, "loss": 0.0199, "step": 17440 }, { "grad_norm": 0.19251413643360138, "learning_rate": 9.582247495002486e-05, "loss": 0.0194, "step": 17450 }, { "grad_norm": 0.24372045695781708, "learning_rate": 9.581585607907214e-05, "loss": 0.0237, "step": 17460 }, { "grad_norm": 0.21552768349647522, "learning_rate": 9.580923219776628e-05, "loss": 0.0238, "step": 17470 }, { "grad_norm": 0.235016331076622, "learning_rate": 9.580260330683167e-05, "loss": 0.0209, "step": 17480 }, { "grad_norm": 0.3210858106613159, "learning_rate": 9.579596940699322e-05, "loss": 0.0239, "step": 17490 }, { "grad_norm": 0.1990034282207489, "learning_rate": 9.578933049897643e-05, "loss": 0.0219, "step": 17500 }, { "grad_norm": 0.2865016460418701, "learning_rate": 9.578268658350728e-05, "loss": 0.023, "step": 17510 }, { "grad_norm": 0.2781977653503418, "learning_rate": 9.577603766131235e-05, "loss": 0.0242, "step": 17520 }, { "grad_norm": 0.24119605123996735, "learning_rate": 9.576938373311878e-05, "loss": 0.0264, "step": 17530 }, { "grad_norm": 0.3290185034275055, "learning_rate": 9.576272479965421e-05, "loss": 0.0248, "step": 17540 }, { "grad_norm": 0.386508584022522, "learning_rate": 9.575606086164687e-05, "loss": 0.0238, "step": 17550 }, { "grad_norm": 0.2657894194126129, "learning_rate": 9.57493919198255e-05, "loss": 0.0243, "step": 17560 }, { "grad_norm": 0.24038036167621613, "learning_rate": 9.57427179749194e-05, "loss": 0.0262, "step": 17570 }, { "grad_norm": 0.23333968222141266, "learning_rate": 9.573603902765846e-05, "loss": 0.0232, "step": 17580 }, { "grad_norm": 0.34850049018859863, "learning_rate": 9.572935507877304e-05, "loss": 0.0244, "step": 17590 }, { "grad_norm": 0.19649425148963928, "learning_rate": 9.57226661289941e-05, "loss": 0.0242, "step": 17600 }, { "grad_norm": 0.19206057488918304, "learning_rate": 9.571597217905315e-05, "loss": 0.0226, "step": 17610 }, { "grad_norm": 0.22125647962093353, "learning_rate": 9.57092732296822e-05, "loss": 0.0196, "step": 17620 }, { "grad_norm": 0.2613804042339325, "learning_rate": 9.570256928161385e-05, "loss": 0.024, "step": 17630 }, { "grad_norm": 0.25522682070732117, "learning_rate": 9.569586033558126e-05, "loss": 0.0238, "step": 17640 }, { "grad_norm": 0.17198459804058075, "learning_rate": 9.568914639231807e-05, "loss": 0.0219, "step": 17650 }, { "grad_norm": 0.22572024166584015, "learning_rate": 9.568242745255852e-05, "loss": 0.023, "step": 17660 }, { "grad_norm": 0.20444509387016296, "learning_rate": 9.567570351703739e-05, "loss": 0.021, "step": 17670 }, { "grad_norm": 0.22120845317840576, "learning_rate": 9.566897458649001e-05, "loss": 0.0212, "step": 17680 }, { "grad_norm": 0.2405664324760437, "learning_rate": 9.566224066165221e-05, "loss": 0.0245, "step": 17690 }, { "grad_norm": 0.32750168442726135, "learning_rate": 9.565550174326043e-05, "loss": 0.0195, "step": 17700 }, { "grad_norm": 0.25123676657676697, "learning_rate": 9.564875783205162e-05, "loss": 0.0208, "step": 17710 }, { "grad_norm": 0.17565476894378662, "learning_rate": 9.564200892876328e-05, "loss": 0.0202, "step": 17720 }, { "grad_norm": 0.22269076108932495, "learning_rate": 9.563525503413348e-05, "loss": 0.0212, "step": 17730 }, { "grad_norm": 0.24798209965229034, "learning_rate": 9.562849614890079e-05, "loss": 0.0212, "step": 17740 }, { "grad_norm": 0.24992252886295319, "learning_rate": 9.562173227380436e-05, "loss": 0.0228, "step": 17750 }, { "grad_norm": 0.25668802857398987, "learning_rate": 9.561496340958389e-05, "loss": 0.0206, "step": 17760 }, { "grad_norm": 0.3293153643608093, "learning_rate": 9.560818955697959e-05, "loss": 0.0208, "step": 17770 }, { "grad_norm": 0.26507166028022766, "learning_rate": 9.560141071673228e-05, "loss": 0.0219, "step": 17780 }, { "grad_norm": 0.24451033771038055, "learning_rate": 9.559462688958323e-05, "loss": 0.0228, "step": 17790 }, { "grad_norm": 0.26554611325263977, "learning_rate": 9.558783807627434e-05, "loss": 0.0228, "step": 17800 }, { "grad_norm": 0.19520488381385803, "learning_rate": 9.558104427754801e-05, "loss": 0.0255, "step": 17810 }, { "grad_norm": 0.20334450900554657, "learning_rate": 9.557424549414722e-05, "loss": 0.0247, "step": 17820 }, { "grad_norm": 0.2341405749320984, "learning_rate": 9.556744172681546e-05, "loss": 0.0212, "step": 17830 }, { "grad_norm": 0.20894761383533478, "learning_rate": 9.556063297629677e-05, "loss": 0.022, "step": 17840 }, { "grad_norm": 0.2793353796005249, "learning_rate": 9.555381924333578e-05, "loss": 0.0231, "step": 17850 }, { "grad_norm": 0.3258225619792938, "learning_rate": 9.554700052867758e-05, "loss": 0.0216, "step": 17860 }, { "grad_norm": 0.20317138731479645, "learning_rate": 9.554017683306789e-05, "loss": 0.0212, "step": 17870 }, { "grad_norm": 0.2248917669057846, "learning_rate": 9.553334815725294e-05, "loss": 0.0212, "step": 17880 }, { "grad_norm": 0.18075726926326752, "learning_rate": 9.552651450197949e-05, "loss": 0.0213, "step": 17890 }, { "grad_norm": 0.22469072043895721, "learning_rate": 9.551967586799486e-05, "loss": 0.0223, "step": 17900 }, { "grad_norm": 0.22589485347270966, "learning_rate": 9.551283225604692e-05, "loss": 0.0221, "step": 17910 }, { "grad_norm": 0.3141343295574188, "learning_rate": 9.550598366688406e-05, "loss": 0.0242, "step": 17920 }, { "grad_norm": 0.205326646566391, "learning_rate": 9.549913010125526e-05, "loss": 0.023, "step": 17930 }, { "grad_norm": 0.3506966829299927, "learning_rate": 9.549227155990999e-05, "loss": 0.0219, "step": 17940 }, { "grad_norm": 0.2802010476589203, "learning_rate": 9.548540804359828e-05, "loss": 0.0208, "step": 17950 }, { "grad_norm": 0.22112701833248138, "learning_rate": 9.547853955307077e-05, "loss": 0.0203, "step": 17960 }, { "grad_norm": 0.34386736154556274, "learning_rate": 9.547166608907853e-05, "loss": 0.0277, "step": 17970 }, { "grad_norm": 0.24378110468387604, "learning_rate": 9.546478765237326e-05, "loss": 0.0216, "step": 17980 }, { "grad_norm": 0.28674012422561646, "learning_rate": 9.545790424370715e-05, "loss": 0.0226, "step": 17990 }, { "grad_norm": 0.25143179297447205, "learning_rate": 9.5451015863833e-05, "loss": 0.0246, "step": 18000 }, { "grad_norm": 0.1920253187417984, "learning_rate": 9.544412251350408e-05, "loss": 0.0237, "step": 18010 }, { "grad_norm": 0.2604922950267792, "learning_rate": 9.543722419347422e-05, "loss": 0.0236, "step": 18020 }, { "grad_norm": 0.2445686161518097, "learning_rate": 9.543032090449788e-05, "loss": 0.0223, "step": 18030 }, { "grad_norm": 0.2534137964248657, "learning_rate": 9.542341264732992e-05, "loss": 0.0234, "step": 18040 }, { "grad_norm": 0.3583275079727173, "learning_rate": 9.541649942272585e-05, "loss": 0.0252, "step": 18050 }, { "grad_norm": 0.20784522593021393, "learning_rate": 9.54095812314417e-05, "loss": 0.0252, "step": 18060 }, { "grad_norm": 0.44119077920913696, "learning_rate": 9.540265807423401e-05, "loss": 0.0277, "step": 18070 }, { "grad_norm": 1.066701889038086, "learning_rate": 9.53957299518599e-05, "loss": 0.0304, "step": 18080 }, { "grad_norm": 0.3774380683898926, "learning_rate": 9.5388796865077e-05, "loss": 0.033, "step": 18090 }, { "grad_norm": 0.2707652449607849, "learning_rate": 9.538185881464353e-05, "loss": 0.0298, "step": 18100 }, { "grad_norm": 0.3327087163925171, "learning_rate": 9.537491580131821e-05, "loss": 0.0328, "step": 18110 }, { "grad_norm": 0.28254660964012146, "learning_rate": 9.53679678258603e-05, "loss": 0.0294, "step": 18120 }, { "grad_norm": 0.21239914000034332, "learning_rate": 9.536101488902966e-05, "loss": 0.0273, "step": 18130 }, { "grad_norm": 0.3198986351490021, "learning_rate": 9.535405699158663e-05, "loss": 0.0324, "step": 18140 }, { "grad_norm": 0.3249439299106598, "learning_rate": 9.53470941342921e-05, "loss": 0.0248, "step": 18150 }, { "grad_norm": 0.20441840589046478, "learning_rate": 9.534012631790756e-05, "loss": 0.0252, "step": 18160 }, { "grad_norm": 0.3796848952770233, "learning_rate": 9.533315354319494e-05, "loss": 0.0301, "step": 18170 }, { "grad_norm": 0.2418632060289383, "learning_rate": 9.532617581091682e-05, "loss": 0.0242, "step": 18180 }, { "grad_norm": 0.34850332140922546, "learning_rate": 9.531919312183629e-05, "loss": 0.0247, "step": 18190 }, { "grad_norm": 0.29874131083488464, "learning_rate": 9.531220547671688e-05, "loss": 0.0229, "step": 18200 }, { "grad_norm": 0.2269114851951599, "learning_rate": 9.530521287632285e-05, "loss": 0.0231, "step": 18210 }, { "grad_norm": 0.3533179461956024, "learning_rate": 9.529821532141884e-05, "loss": 0.025, "step": 18220 }, { "grad_norm": 0.2764745056629181, "learning_rate": 9.52912128127701e-05, "loss": 0.0235, "step": 18230 }, { "grad_norm": 0.22081492841243744, "learning_rate": 9.528420535114244e-05, "loss": 0.0238, "step": 18240 }, { "grad_norm": 0.2331462949514389, "learning_rate": 9.527719293730215e-05, "loss": 0.0221, "step": 18250 }, { "grad_norm": 0.28506821393966675, "learning_rate": 9.527017557201611e-05, "loss": 0.0238, "step": 18260 }, { "grad_norm": 0.32479384541511536, "learning_rate": 9.526315325605176e-05, "loss": 0.0233, "step": 18270 }, { "grad_norm": 0.27140703797340393, "learning_rate": 9.525612599017699e-05, "loss": 0.0242, "step": 18280 }, { "grad_norm": 0.20539113879203796, "learning_rate": 9.524909377516033e-05, "loss": 0.0217, "step": 18290 }, { "grad_norm": 0.1882798671722412, "learning_rate": 9.524205661177081e-05, "loss": 0.0241, "step": 18300 }, { "grad_norm": 0.20942731201648712, "learning_rate": 9.523501450077801e-05, "loss": 0.0241, "step": 18310 }, { "grad_norm": 0.21696892380714417, "learning_rate": 9.522796744295202e-05, "loss": 0.0213, "step": 18320 }, { "grad_norm": 0.22154690325260162, "learning_rate": 9.522091543906352e-05, "loss": 0.0251, "step": 18330 }, { "grad_norm": 0.25328341126441956, "learning_rate": 9.521385848988369e-05, "loss": 0.0272, "step": 18340 }, { "grad_norm": 0.31507205963134766, "learning_rate": 9.520679659618428e-05, "loss": 0.0228, "step": 18350 }, { "grad_norm": 0.2591251730918884, "learning_rate": 9.519972975873754e-05, "loss": 0.0241, "step": 18360 }, { "grad_norm": 0.22377559542655945, "learning_rate": 9.519265797831633e-05, "loss": 0.0248, "step": 18370 }, { "grad_norm": 0.26566004753112793, "learning_rate": 9.518558125569399e-05, "loss": 0.0259, "step": 18380 }, { "grad_norm": 0.24506549537181854, "learning_rate": 9.517849959164442e-05, "loss": 0.0223, "step": 18390 }, { "grad_norm": 0.21176107227802277, "learning_rate": 9.517141298694205e-05, "loss": 0.0218, "step": 18400 }, { "grad_norm": 0.5165038704872131, "learning_rate": 9.516432144236188e-05, "loss": 0.0251, "step": 18410 }, { "grad_norm": 0.2544362545013428, "learning_rate": 9.515722495867941e-05, "loss": 0.0228, "step": 18420 }, { "grad_norm": 0.2518576681613922, "learning_rate": 9.515012353667072e-05, "loss": 0.0249, "step": 18430 }, { "grad_norm": 0.21046465635299683, "learning_rate": 9.51430171771124e-05, "loss": 0.0202, "step": 18440 }, { "grad_norm": 0.2410619705915451, "learning_rate": 9.513590588078159e-05, "loss": 0.023, "step": 18450 }, { "grad_norm": 0.2680246829986572, "learning_rate": 9.512878964845597e-05, "loss": 0.0227, "step": 18460 }, { "grad_norm": 0.27075400948524475, "learning_rate": 9.512166848091377e-05, "loss": 0.0204, "step": 18470 }, { "grad_norm": 0.265807181596756, "learning_rate": 9.511454237893376e-05, "loss": 0.0232, "step": 18480 }, { "grad_norm": 0.21283045411109924, "learning_rate": 9.51074113432952e-05, "loss": 0.0224, "step": 18490 }, { "grad_norm": 0.15854966640472412, "learning_rate": 9.510027537477797e-05, "loss": 0.0254, "step": 18500 }, { "grad_norm": 0.22560591995716095, "learning_rate": 9.509313447416242e-05, "loss": 0.0266, "step": 18510 }, { "grad_norm": 0.29266127943992615, "learning_rate": 9.508598864222949e-05, "loss": 0.0264, "step": 18520 }, { "grad_norm": 0.24330446124076843, "learning_rate": 9.507883787976062e-05, "loss": 0.0216, "step": 18530 }, { "grad_norm": 0.2583327293395996, "learning_rate": 9.507168218753781e-05, "loss": 0.0258, "step": 18540 }, { "grad_norm": 0.2532365024089813, "learning_rate": 9.506452156634362e-05, "loss": 0.0218, "step": 18550 }, { "grad_norm": 0.2085486650466919, "learning_rate": 9.505735601696109e-05, "loss": 0.0212, "step": 18560 }, { "grad_norm": 0.2421131283044815, "learning_rate": 9.505018554017385e-05, "loss": 0.0242, "step": 18570 }, { "grad_norm": 0.21535076200962067, "learning_rate": 9.504301013676604e-05, "loss": 0.0211, "step": 18580 }, { "grad_norm": 0.2746066153049469, "learning_rate": 9.503582980752238e-05, "loss": 0.0236, "step": 18590 }, { "grad_norm": 0.19975489377975464, "learning_rate": 9.502864455322809e-05, "loss": 0.0225, "step": 18600 }, { "grad_norm": 0.2297179400920868, "learning_rate": 9.502145437466891e-05, "loss": 0.026, "step": 18610 }, { "grad_norm": 0.23884403705596924, "learning_rate": 9.501425927263116e-05, "loss": 0.0241, "step": 18620 }, { "grad_norm": 0.20262007415294647, "learning_rate": 9.500705924790172e-05, "loss": 0.0213, "step": 18630 }, { "grad_norm": 0.29565906524658203, "learning_rate": 9.499985430126794e-05, "loss": 0.0218, "step": 18640 }, { "grad_norm": 0.22833754122257233, "learning_rate": 9.499264443351775e-05, "loss": 0.0208, "step": 18650 }, { "grad_norm": 0.29763713479042053, "learning_rate": 9.498542964543961e-05, "loss": 0.022, "step": 18660 }, { "grad_norm": 0.22396151721477509, "learning_rate": 9.497820993782252e-05, "loss": 0.0193, "step": 18670 }, { "grad_norm": 0.264909952878952, "learning_rate": 9.497098531145601e-05, "loss": 0.0223, "step": 18680 }, { "grad_norm": 0.27597665786743164, "learning_rate": 9.496375576713017e-05, "loss": 0.0235, "step": 18690 }, { "grad_norm": 0.4456389248371124, "learning_rate": 9.49565213056356e-05, "loss": 0.0219, "step": 18700 }, { "grad_norm": 0.23507870733737946, "learning_rate": 9.494928192776342e-05, "loss": 0.0205, "step": 18710 }, { "grad_norm": 0.3462526798248291, "learning_rate": 9.494203763430538e-05, "loss": 0.0231, "step": 18720 }, { "grad_norm": 0.22252900898456573, "learning_rate": 9.493478842605366e-05, "loss": 0.021, "step": 18730 }, { "grad_norm": 0.23266178369522095, "learning_rate": 9.492753430380105e-05, "loss": 0.0231, "step": 18740 }, { "grad_norm": 0.22860103845596313, "learning_rate": 9.492027526834083e-05, "loss": 0.0219, "step": 18750 }, { "grad_norm": 0.21622909605503082, "learning_rate": 9.491301132046684e-05, "loss": 0.0215, "step": 18760 }, { "grad_norm": 0.18044453859329224, "learning_rate": 9.490574246097345e-05, "loss": 0.0211, "step": 18770 }, { "grad_norm": 0.19482356309890747, "learning_rate": 9.48984686906556e-05, "loss": 0.021, "step": 18780 }, { "grad_norm": 0.32475507259368896, "learning_rate": 9.489119001030871e-05, "loss": 0.022, "step": 18790 }, { "grad_norm": 0.24364714324474335, "learning_rate": 9.488390642072878e-05, "loss": 0.025, "step": 18800 }, { "grad_norm": 0.3588086664676666, "learning_rate": 9.48766179227123e-05, "loss": 0.0234, "step": 18810 }, { "grad_norm": 0.28600814938545227, "learning_rate": 9.486932451705636e-05, "loss": 0.0231, "step": 18820 }, { "grad_norm": 0.19523906707763672, "learning_rate": 9.486202620455857e-05, "loss": 0.0227, "step": 18830 }, { "grad_norm": 0.21120499074459076, "learning_rate": 9.485472298601704e-05, "loss": 0.0228, "step": 18840 }, { "grad_norm": 0.3603631556034088, "learning_rate": 9.484741486223043e-05, "loss": 0.0247, "step": 18850 }, { "grad_norm": 0.26373884081840515, "learning_rate": 9.484010183399797e-05, "loss": 0.0236, "step": 18860 }, { "grad_norm": 0.21348145604133606, "learning_rate": 9.483278390211938e-05, "loss": 0.0221, "step": 18870 }, { "grad_norm": 0.2998644709587097, "learning_rate": 9.482546106739496e-05, "loss": 0.0231, "step": 18880 }, { "grad_norm": 0.2275707870721817, "learning_rate": 9.48181333306255e-05, "loss": 0.0235, "step": 18890 }, { "grad_norm": 0.30889463424682617, "learning_rate": 9.481080069261237e-05, "loss": 0.0217, "step": 18900 }, { "grad_norm": 0.2468756139278412, "learning_rate": 9.480346315415745e-05, "loss": 0.0231, "step": 18910 }, { "grad_norm": 0.36622849106788635, "learning_rate": 9.479612071606314e-05, "loss": 0.0238, "step": 18920 }, { "grad_norm": 0.19599808752536774, "learning_rate": 9.478877337913244e-05, "loss": 0.0212, "step": 18930 }, { "grad_norm": 0.2137073278427124, "learning_rate": 9.478142114416881e-05, "loss": 0.0201, "step": 18940 }, { "grad_norm": 0.2176896631717682, "learning_rate": 9.47740640119763e-05, "loss": 0.0232, "step": 18950 }, { "grad_norm": 0.24040786921977997, "learning_rate": 9.476670198335947e-05, "loss": 0.0216, "step": 18960 }, { "grad_norm": 0.24577543139457703, "learning_rate": 9.47593350591234e-05, "loss": 0.0234, "step": 18970 }, { "grad_norm": 0.2439728081226349, "learning_rate": 9.475196324007376e-05, "loss": 0.0222, "step": 18980 }, { "grad_norm": 0.22058546543121338, "learning_rate": 9.474458652701669e-05, "loss": 0.0223, "step": 18990 }, { "grad_norm": 0.2198343276977539, "learning_rate": 9.473720492075892e-05, "loss": 0.0208, "step": 19000 }, { "grad_norm": 0.21346378326416016, "learning_rate": 9.472981842210768e-05, "loss": 0.022, "step": 19010 }, { "grad_norm": 0.2672499716281891, "learning_rate": 9.472242703187074e-05, "loss": 0.0198, "step": 19020 }, { "grad_norm": 0.24208249151706696, "learning_rate": 9.471503075085643e-05, "loss": 0.0217, "step": 19030 }, { "grad_norm": 0.23026493191719055, "learning_rate": 9.470762957987359e-05, "loss": 0.0248, "step": 19040 }, { "grad_norm": 0.26312145590782166, "learning_rate": 9.470022351973158e-05, "loss": 0.0214, "step": 19050 }, { "grad_norm": 0.2002219259738922, "learning_rate": 9.469281257124034e-05, "loss": 0.0251, "step": 19060 }, { "grad_norm": 0.22645960748195648, "learning_rate": 9.46853967352103e-05, "loss": 0.0222, "step": 19070 }, { "grad_norm": 0.25809523463249207, "learning_rate": 9.467797601245246e-05, "loss": 0.0219, "step": 19080 }, { "grad_norm": 0.2455539107322693, "learning_rate": 9.467055040377834e-05, "loss": 0.0208, "step": 19090 }, { "grad_norm": 0.23688216507434845, "learning_rate": 9.466311990999999e-05, "loss": 0.0248, "step": 19100 }, { "grad_norm": 0.30204886198043823, "learning_rate": 9.465568453193e-05, "loss": 0.0273, "step": 19110 }, { "grad_norm": 0.23431698977947235, "learning_rate": 9.464824427038148e-05, "loss": 0.022, "step": 19120 }, { "grad_norm": 0.27962440252304077, "learning_rate": 9.46407991261681e-05, "loss": 0.0245, "step": 19130 }, { "grad_norm": 0.2596489489078522, "learning_rate": 9.463334910010404e-05, "loss": 0.0233, "step": 19140 }, { "grad_norm": 0.22074833512306213, "learning_rate": 9.462589419300403e-05, "loss": 0.0221, "step": 19150 }, { "grad_norm": 0.16833321750164032, "learning_rate": 9.461843440568333e-05, "loss": 0.0203, "step": 19160 }, { "grad_norm": 0.19562499225139618, "learning_rate": 9.461096973895773e-05, "loss": 0.0191, "step": 19170 }, { "grad_norm": 0.20448479056358337, "learning_rate": 9.460350019364355e-05, "loss": 0.0212, "step": 19180 }, { "grad_norm": 0.2233932465314865, "learning_rate": 9.459602577055764e-05, "loss": 0.0229, "step": 19190 }, { "grad_norm": 0.2613198459148407, "learning_rate": 9.45885464705174e-05, "loss": 0.0248, "step": 19200 }, { "grad_norm": 0.22998052835464478, "learning_rate": 9.458106229434076e-05, "loss": 0.0192, "step": 19210 }, { "grad_norm": 0.26320889592170715, "learning_rate": 9.457357324284617e-05, "loss": 0.0211, "step": 19220 }, { "grad_norm": 0.2336733192205429, "learning_rate": 9.456607931685262e-05, "loss": 0.0217, "step": 19230 }, { "grad_norm": 0.23727920651435852, "learning_rate": 9.455858051717965e-05, "loss": 0.0213, "step": 19240 }, { "grad_norm": 0.24371838569641113, "learning_rate": 9.45510768446473e-05, "loss": 0.0217, "step": 19250 }, { "grad_norm": 0.19433003664016724, "learning_rate": 9.454356830007618e-05, "loss": 0.0224, "step": 19260 }, { "grad_norm": 0.24489638209342957, "learning_rate": 9.45360548842874e-05, "loss": 0.0185, "step": 19270 }, { "grad_norm": 0.24936501681804657, "learning_rate": 9.452853659810261e-05, "loss": 0.0236, "step": 19280 }, { "grad_norm": 0.2094735950231552, "learning_rate": 9.452101344234401e-05, "loss": 0.0194, "step": 19290 }, { "grad_norm": 0.15973708033561707, "learning_rate": 9.451348541783431e-05, "loss": 0.0214, "step": 19300 }, { "grad_norm": 0.2495991587638855, "learning_rate": 9.450595252539678e-05, "loss": 0.0205, "step": 19310 }, { "grad_norm": 0.2431386560201645, "learning_rate": 9.449841476585518e-05, "loss": 0.0246, "step": 19320 }, { "grad_norm": 0.2125258594751358, "learning_rate": 9.449087214003384e-05, "loss": 0.0263, "step": 19330 }, { "grad_norm": 0.25479796528816223, "learning_rate": 9.448332464875765e-05, "loss": 0.0281, "step": 19340 }, { "grad_norm": 0.242653489112854, "learning_rate": 9.447577229285192e-05, "loss": 0.0271, "step": 19350 }, { "grad_norm": 0.27732086181640625, "learning_rate": 9.446821507314261e-05, "loss": 0.0232, "step": 19360 }, { "grad_norm": 0.21046428382396698, "learning_rate": 9.446065299045617e-05, "loss": 0.0255, "step": 19370 }, { "grad_norm": 0.6523874402046204, "learning_rate": 9.445308604561955e-05, "loss": 0.0281, "step": 19380 }, { "grad_norm": 0.25254812836647034, "learning_rate": 9.444551423946028e-05, "loss": 0.0288, "step": 19390 }, { "grad_norm": 0.26231202483177185, "learning_rate": 9.443793757280638e-05, "loss": 0.0275, "step": 19400 }, { "grad_norm": 0.3784092962741852, "learning_rate": 9.443035604648646e-05, "loss": 0.0278, "step": 19410 }, { "grad_norm": 0.26971328258514404, "learning_rate": 9.44227696613296e-05, "loss": 0.0241, "step": 19420 }, { "grad_norm": 0.2730194628238678, "learning_rate": 9.441517841816542e-05, "loss": 0.0223, "step": 19430 }, { "grad_norm": 0.22039024531841278, "learning_rate": 9.440758231782413e-05, "loss": 0.0251, "step": 19440 }, { "grad_norm": 0.25240853428840637, "learning_rate": 9.439998136113639e-05, "loss": 0.0209, "step": 19450 }, { "grad_norm": 0.27188414335250854, "learning_rate": 9.439237554893344e-05, "loss": 0.0223, "step": 19460 }, { "grad_norm": 0.20255202054977417, "learning_rate": 9.438476488204705e-05, "loss": 0.0219, "step": 19470 }, { "grad_norm": 0.18518120050430298, "learning_rate": 9.43771493613095e-05, "loss": 0.0219, "step": 19480 }, { "grad_norm": 0.22802291810512543, "learning_rate": 9.436952898755362e-05, "loss": 0.0208, "step": 19490 }, { "grad_norm": 0.2590443193912506, "learning_rate": 9.436190376161276e-05, "loss": 0.0211, "step": 19500 }, { "grad_norm": 0.2245737612247467, "learning_rate": 9.43542736843208e-05, "loss": 0.0265, "step": 19510 }, { "grad_norm": 0.20055396854877472, "learning_rate": 9.434663875651216e-05, "loss": 0.0213, "step": 19520 }, { "grad_norm": 0.20340150594711304, "learning_rate": 9.433899897902177e-05, "loss": 0.0223, "step": 19530 }, { "grad_norm": 0.16878344118595123, "learning_rate": 9.433135435268511e-05, "loss": 0.0191, "step": 19540 }, { "grad_norm": 0.2652433216571808, "learning_rate": 9.432370487833819e-05, "loss": 0.0252, "step": 19550 }, { "grad_norm": 0.22957982122898102, "learning_rate": 9.431605055681756e-05, "loss": 0.0237, "step": 19560 }, { "grad_norm": 0.2495337873697281, "learning_rate": 9.430839138896026e-05, "loss": 0.0247, "step": 19570 }, { "grad_norm": 0.26398545503616333, "learning_rate": 9.43007273756039e-05, "loss": 0.0236, "step": 19580 }, { "grad_norm": 0.21786381304264069, "learning_rate": 9.429305851758658e-05, "loss": 0.0197, "step": 19590 }, { "grad_norm": 0.19482825696468353, "learning_rate": 9.428538481574699e-05, "loss": 0.0207, "step": 19600 }, { "grad_norm": 0.2402459979057312, "learning_rate": 9.42777062709243e-05, "loss": 0.0213, "step": 19610 }, { "grad_norm": 0.1990462690591812, "learning_rate": 9.427002288395821e-05, "loss": 0.0216, "step": 19620 }, { "grad_norm": 0.23267970979213715, "learning_rate": 9.426233465568898e-05, "loss": 0.023, "step": 19630 }, { "grad_norm": 0.25233346223831177, "learning_rate": 9.42546415869574e-05, "loss": 0.0217, "step": 19640 }, { "grad_norm": 0.2220161408185959, "learning_rate": 9.424694367860473e-05, "loss": 0.0242, "step": 19650 }, { "grad_norm": 0.25422805547714233, "learning_rate": 9.423924093147284e-05, "loss": 0.0244, "step": 19660 }, { "grad_norm": 0.2207653969526291, "learning_rate": 9.423153334640407e-05, "loss": 0.0246, "step": 19670 }, { "grad_norm": 0.2965943217277527, "learning_rate": 9.42238209242413e-05, "loss": 0.0237, "step": 19680 }, { "grad_norm": 0.2421577125787735, "learning_rate": 9.421610366582798e-05, "loss": 0.0285, "step": 19690 }, { "grad_norm": 0.24192653596401215, "learning_rate": 9.420838157200803e-05, "loss": 0.0246, "step": 19700 }, { "grad_norm": 0.24668298661708832, "learning_rate": 9.420065464362594e-05, "loss": 0.0226, "step": 19710 }, { "grad_norm": 0.19565334916114807, "learning_rate": 9.419292288152673e-05, "loss": 0.0247, "step": 19720 }, { "grad_norm": 0.23667645454406738, "learning_rate": 9.418518628655588e-05, "loss": 0.0272, "step": 19730 }, { "grad_norm": 0.22387023270130157, "learning_rate": 9.417744485955951e-05, "loss": 0.0261, "step": 19740 }, { "grad_norm": 0.23020219802856445, "learning_rate": 9.41696986013842e-05, "loss": 0.0249, "step": 19750 }, { "grad_norm": 0.2644073963165283, "learning_rate": 9.416194751287705e-05, "loss": 0.0292, "step": 19760 }, { "grad_norm": 0.21615242958068848, "learning_rate": 9.415419159488572e-05, "loss": 0.0241, "step": 19770 }, { "grad_norm": 0.199700266122818, "learning_rate": 9.414643084825837e-05, "loss": 0.0215, "step": 19780 }, { "grad_norm": 0.22798673808574677, "learning_rate": 9.413866527384372e-05, "loss": 0.0243, "step": 19790 }, { "grad_norm": 0.21134185791015625, "learning_rate": 9.4130894872491e-05, "loss": 0.0226, "step": 19800 }, { "grad_norm": 0.21067167818546295, "learning_rate": 9.412311964504998e-05, "loss": 0.0219, "step": 19810 }, { "grad_norm": 0.21502991020679474, "learning_rate": 9.411533959237091e-05, "loss": 0.0214, "step": 19820 }, { "grad_norm": 0.1936536580324173, "learning_rate": 9.410755471530464e-05, "loss": 0.021, "step": 19830 }, { "grad_norm": 0.2522929906845093, "learning_rate": 9.40997650147025e-05, "loss": 0.0211, "step": 19840 }, { "grad_norm": 0.18569731712341309, "learning_rate": 9.409197049141637e-05, "loss": 0.021, "step": 19850 }, { "grad_norm": 0.20073658227920532, "learning_rate": 9.408417114629863e-05, "loss": 0.0229, "step": 19860 }, { "grad_norm": 0.19089166820049286, "learning_rate": 9.40763669802022e-05, "loss": 0.0237, "step": 19870 }, { "grad_norm": 0.2693052589893341, "learning_rate": 9.406855799398056e-05, "loss": 0.0223, "step": 19880 }, { "grad_norm": 0.2101280391216278, "learning_rate": 9.406074418848767e-05, "loss": 0.0229, "step": 19890 }, { "grad_norm": 0.2340124398469925, "learning_rate": 9.405292556457805e-05, "loss": 0.0216, "step": 19900 }, { "grad_norm": 0.21311773359775543, "learning_rate": 9.404510212310671e-05, "loss": 0.0249, "step": 19910 }, { "grad_norm": 0.3301328718662262, "learning_rate": 9.403727386492924e-05, "loss": 0.023, "step": 19920 }, { "grad_norm": 0.22269703447818756, "learning_rate": 9.40294407909017e-05, "loss": 0.0283, "step": 19930 }, { "grad_norm": 0.2001333087682724, "learning_rate": 9.40216029018807e-05, "loss": 0.0228, "step": 19940 }, { "grad_norm": 0.1929970681667328, "learning_rate": 9.401376019872338e-05, "loss": 0.0232, "step": 19950 }, { "grad_norm": 0.26091983914375305, "learning_rate": 9.400591268228746e-05, "loss": 0.0252, "step": 19960 }, { "grad_norm": 0.2314077764749527, "learning_rate": 9.399806035343106e-05, "loss": 0.0203, "step": 19970 }, { "grad_norm": 0.23584550619125366, "learning_rate": 9.399020321301294e-05, "loss": 0.0213, "step": 19980 }, { "grad_norm": 0.21853981912136078, "learning_rate": 9.398234126189234e-05, "loss": 0.0192, "step": 19990 }, { "grad_norm": 0.25677773356437683, "learning_rate": 9.397447450092902e-05, "loss": 0.0221, "step": 20000 }, { "grad_norm": 0.15416528284549713, "learning_rate": 9.39666029309833e-05, "loss": 0.018, "step": 20010 }, { "grad_norm": 0.18838632106781006, "learning_rate": 9.395872655291596e-05, "loss": 0.0212, "step": 20020 }, { "grad_norm": 0.2157837301492691, "learning_rate": 9.395084536758838e-05, "loss": 0.021, "step": 20030 }, { "grad_norm": 0.24035927653312683, "learning_rate": 9.394295937586243e-05, "loss": 0.021, "step": 20040 }, { "grad_norm": 0.17793293297290802, "learning_rate": 9.393506857860052e-05, "loss": 0.0188, "step": 20050 }, { "grad_norm": 0.19675174355506897, "learning_rate": 9.392717297666555e-05, "loss": 0.0184, "step": 20060 }, { "grad_norm": 0.23177149891853333, "learning_rate": 9.391927257092101e-05, "loss": 0.0202, "step": 20070 }, { "grad_norm": 0.26282230019569397, "learning_rate": 9.391136736223085e-05, "loss": 0.0209, "step": 20080 }, { "grad_norm": 0.1927979439496994, "learning_rate": 9.390345735145956e-05, "loss": 0.018, "step": 20090 }, { "grad_norm": 0.1619037389755249, "learning_rate": 9.389554253947219e-05, "loss": 0.0182, "step": 20100 }, { "grad_norm": 0.25512030720710754, "learning_rate": 9.388762292713428e-05, "loss": 0.0206, "step": 20110 }, { "grad_norm": 0.29142341017723083, "learning_rate": 9.38796985153119e-05, "loss": 0.0223, "step": 20120 }, { "grad_norm": 0.20412079989910126, "learning_rate": 9.387176930487169e-05, "loss": 0.0215, "step": 20130 }, { "grad_norm": 0.22849659621715546, "learning_rate": 9.386383529668072e-05, "loss": 0.0197, "step": 20140 }, { "grad_norm": 0.18416406214237213, "learning_rate": 9.385589649160669e-05, "loss": 0.0205, "step": 20150 }, { "grad_norm": 0.18926498293876648, "learning_rate": 9.384795289051775e-05, "loss": 0.0213, "step": 20160 }, { "grad_norm": 0.21505942940711975, "learning_rate": 9.384000449428261e-05, "loss": 0.0199, "step": 20170 }, { "grad_norm": 0.2091224491596222, "learning_rate": 9.383205130377048e-05, "loss": 0.0192, "step": 20180 }, { "grad_norm": 0.21459093689918518, "learning_rate": 9.382409331985114e-05, "loss": 0.0182, "step": 20190 }, { "grad_norm": 0.20315444469451904, "learning_rate": 9.381613054339482e-05, "loss": 0.02, "step": 20200 }, { "grad_norm": 0.2441018521785736, "learning_rate": 9.380816297527235e-05, "loss": 0.0253, "step": 20210 }, { "grad_norm": 0.23270854353904724, "learning_rate": 9.380019061635506e-05, "loss": 0.0228, "step": 20220 }, { "grad_norm": 0.2500436305999756, "learning_rate": 9.379221346751474e-05, "loss": 0.0213, "step": 20230 }, { "grad_norm": 0.2550198435783386, "learning_rate": 9.378423152962382e-05, "loss": 0.02, "step": 20240 }, { "grad_norm": 0.22051836550235748, "learning_rate": 9.377624480355517e-05, "loss": 0.0202, "step": 20250 }, { "grad_norm": 0.25415635108947754, "learning_rate": 9.376825329018219e-05, "loss": 0.0238, "step": 20260 }, { "grad_norm": 0.22581031918525696, "learning_rate": 9.376025699037884e-05, "loss": 0.0218, "step": 20270 }, { "grad_norm": 0.22216485440731049, "learning_rate": 9.37522559050196e-05, "loss": 0.0209, "step": 20280 }, { "grad_norm": 0.2612174451351166, "learning_rate": 9.37442500349794e-05, "loss": 0.0209, "step": 20290 }, { "grad_norm": 0.21305961906909943, "learning_rate": 9.373623938113381e-05, "loss": 0.0218, "step": 20300 }, { "grad_norm": 0.31270837783813477, "learning_rate": 9.372822394435883e-05, "loss": 0.0199, "step": 20310 }, { "grad_norm": 0.2060169279575348, "learning_rate": 9.372020372553102e-05, "loss": 0.018, "step": 20320 }, { "grad_norm": 0.1715165078639984, "learning_rate": 9.371217872552746e-05, "loss": 0.0181, "step": 20330 }, { "grad_norm": 0.2078849822282791, "learning_rate": 9.370414894522576e-05, "loss": 0.0206, "step": 20340 }, { "grad_norm": 0.23553648591041565, "learning_rate": 9.369611438550406e-05, "loss": 0.0193, "step": 20350 }, { "grad_norm": 0.21271111071109772, "learning_rate": 9.368807504724095e-05, "loss": 0.0225, "step": 20360 }, { "grad_norm": 0.1967170536518097, "learning_rate": 9.368003093131565e-05, "loss": 0.0228, "step": 20370 }, { "grad_norm": 0.19272568821907043, "learning_rate": 9.367198203860785e-05, "loss": 0.0236, "step": 20380 }, { "grad_norm": 0.225735142827034, "learning_rate": 9.366392836999774e-05, "loss": 0.0189, "step": 20390 }, { "grad_norm": 0.3650871217250824, "learning_rate": 9.365586992636607e-05, "loss": 0.0222, "step": 20400 }, { "grad_norm": 0.24170827865600586, "learning_rate": 9.364780670859412e-05, "loss": 0.0205, "step": 20410 }, { "grad_norm": 0.2548307180404663, "learning_rate": 9.363973871756364e-05, "loss": 0.0223, "step": 20420 }, { "grad_norm": 0.2619922161102295, "learning_rate": 9.363166595415696e-05, "loss": 0.019, "step": 20430 }, { "grad_norm": 0.22596265375614166, "learning_rate": 9.362358841925686e-05, "loss": 0.0214, "step": 20440 }, { "grad_norm": 0.19198077917099, "learning_rate": 9.361550611374674e-05, "loss": 0.0204, "step": 20450 }, { "grad_norm": 0.21987278759479523, "learning_rate": 9.360741903851043e-05, "loss": 0.0197, "step": 20460 }, { "grad_norm": 0.2573544681072235, "learning_rate": 9.359932719443236e-05, "loss": 0.021, "step": 20470 }, { "grad_norm": 0.2287181317806244, "learning_rate": 9.35912305823974e-05, "loss": 0.0227, "step": 20480 }, { "grad_norm": 0.27857401967048645, "learning_rate": 9.358312920329101e-05, "loss": 0.0213, "step": 20490 }, { "grad_norm": 0.1990620493888855, "learning_rate": 9.357502305799914e-05, "loss": 0.0169, "step": 20500 }, { "grad_norm": 0.20919981598854065, "learning_rate": 9.356691214740824e-05, "loss": 0.0184, "step": 20510 }, { "grad_norm": 0.14442004263401031, "learning_rate": 9.355879647240535e-05, "loss": 0.017, "step": 20520 }, { "grad_norm": 0.18376193940639496, "learning_rate": 9.355067603387798e-05, "loss": 0.0217, "step": 20530 }, { "grad_norm": 0.22586561739444733, "learning_rate": 9.354255083271412e-05, "loss": 0.0214, "step": 20540 }, { "grad_norm": 0.1819133758544922, "learning_rate": 9.353442086980239e-05, "loss": 0.0198, "step": 20550 }, { "grad_norm": 0.2208467274904251, "learning_rate": 9.352628614603185e-05, "loss": 0.0198, "step": 20560 }, { "grad_norm": 0.24433372914791107, "learning_rate": 9.351814666229209e-05, "loss": 0.02, "step": 20570 }, { "grad_norm": 0.2266346663236618, "learning_rate": 9.351000241947324e-05, "loss": 0.0207, "step": 20580 }, { "grad_norm": 0.22464966773986816, "learning_rate": 9.350185341846594e-05, "loss": 0.0191, "step": 20590 }, { "grad_norm": 0.18825601041316986, "learning_rate": 9.349369966016134e-05, "loss": 0.0196, "step": 20600 }, { "grad_norm": 0.17956936359405518, "learning_rate": 9.348554114545117e-05, "loss": 0.0207, "step": 20610 }, { "grad_norm": 0.21905086934566498, "learning_rate": 9.347737787522758e-05, "loss": 0.0187, "step": 20620 }, { "grad_norm": 0.1930258572101593, "learning_rate": 9.346920985038332e-05, "loss": 0.0212, "step": 20630 }, { "grad_norm": 0.273381769657135, "learning_rate": 9.346103707181162e-05, "loss": 0.0178, "step": 20640 }, { "grad_norm": 0.25589925050735474, "learning_rate": 9.345285954040626e-05, "loss": 0.018, "step": 20650 }, { "grad_norm": 0.24975769221782684, "learning_rate": 9.34446772570615e-05, "loss": 0.0181, "step": 20660 }, { "grad_norm": 0.2649098336696625, "learning_rate": 9.343649022267214e-05, "loss": 0.0185, "step": 20670 }, { "grad_norm": 0.1823466569185257, "learning_rate": 9.342829843813353e-05, "loss": 0.0189, "step": 20680 }, { "grad_norm": 0.19772593677043915, "learning_rate": 9.342010190434149e-05, "loss": 0.0203, "step": 20690 }, { "grad_norm": 0.19189874827861786, "learning_rate": 9.34119006221924e-05, "loss": 0.0202, "step": 20700 }, { "grad_norm": 0.18837551772594452, "learning_rate": 9.340369459258313e-05, "loss": 0.0185, "step": 20710 }, { "grad_norm": 0.2363632470369339, "learning_rate": 9.339548381641106e-05, "loss": 0.0233, "step": 20720 }, { "grad_norm": 0.23287343978881836, "learning_rate": 9.338726829457413e-05, "loss": 0.0191, "step": 20730 }, { "grad_norm": 0.22903621196746826, "learning_rate": 9.337904802797078e-05, "loss": 0.022, "step": 20740 }, { "grad_norm": 0.45187440514564514, "learning_rate": 9.337082301749993e-05, "loss": 0.0234, "step": 20750 }, { "grad_norm": 0.177354633808136, "learning_rate": 9.336259326406109e-05, "loss": 0.0191, "step": 20760 }, { "grad_norm": 0.2390439212322235, "learning_rate": 9.335435876855427e-05, "loss": 0.0219, "step": 20770 }, { "grad_norm": 0.19856074452400208, "learning_rate": 9.334611953187994e-05, "loss": 0.021, "step": 20780 }, { "grad_norm": 0.25166046619415283, "learning_rate": 9.333787555493914e-05, "loss": 0.0192, "step": 20790 }, { "grad_norm": 0.1849314421415329, "learning_rate": 9.332962683863345e-05, "loss": 0.0199, "step": 20800 }, { "grad_norm": 0.2120225429534912, "learning_rate": 9.332137338386489e-05, "loss": 0.0208, "step": 20810 }, { "grad_norm": 0.22593297064304352, "learning_rate": 9.33131151915361e-05, "loss": 0.0207, "step": 20820 }, { "grad_norm": 0.21701577305793762, "learning_rate": 9.330485226255012e-05, "loss": 0.0232, "step": 20830 }, { "grad_norm": 0.28302139043807983, "learning_rate": 9.329658459781061e-05, "loss": 0.0196, "step": 20840 }, { "grad_norm": 0.23735590279102325, "learning_rate": 9.328831219822172e-05, "loss": 0.0219, "step": 20850 }, { "grad_norm": 0.23402267694473267, "learning_rate": 9.328003506468808e-05, "loss": 0.0211, "step": 20860 }, { "grad_norm": 0.2009536474943161, "learning_rate": 9.327175319811488e-05, "loss": 0.0218, "step": 20870 }, { "grad_norm": 0.18931087851524353, "learning_rate": 9.326346659940781e-05, "loss": 0.0198, "step": 20880 }, { "grad_norm": 0.18510538339614868, "learning_rate": 9.325517526947308e-05, "loss": 0.0205, "step": 20890 }, { "grad_norm": 0.16925963759422302, "learning_rate": 9.32468792092174e-05, "loss": 0.0206, "step": 20900 }, { "grad_norm": 0.18541207909584045, "learning_rate": 9.323857841954803e-05, "loss": 0.02, "step": 20910 }, { "grad_norm": 0.16612504422664642, "learning_rate": 9.323027290137276e-05, "loss": 0.0212, "step": 20920 }, { "grad_norm": 0.1782609224319458, "learning_rate": 9.322196265559981e-05, "loss": 0.0194, "step": 20930 }, { "grad_norm": 0.21207395195960999, "learning_rate": 9.321364768313803e-05, "loss": 0.0202, "step": 20940 }, { "grad_norm": 0.1630401313304901, "learning_rate": 9.32053279848967e-05, "loss": 0.0198, "step": 20950 }, { "grad_norm": 0.3112931251525879, "learning_rate": 9.319700356178567e-05, "loss": 0.0196, "step": 20960 }, { "grad_norm": 0.2166532576084137, "learning_rate": 9.318867441471527e-05, "loss": 0.021, "step": 20970 }, { "grad_norm": 0.17810800671577454, "learning_rate": 9.318034054459637e-05, "loss": 0.0202, "step": 20980 }, { "grad_norm": 0.25721365213394165, "learning_rate": 9.317200195234034e-05, "loss": 0.021, "step": 20990 }, { "grad_norm": 0.19647277891635895, "learning_rate": 9.316365863885909e-05, "loss": 0.019, "step": 21000 }, { "grad_norm": 0.23620396852493286, "learning_rate": 9.315531060506502e-05, "loss": 0.0202, "step": 21010 }, { "grad_norm": 0.2323158234357834, "learning_rate": 9.314695785187108e-05, "loss": 0.0206, "step": 21020 }, { "grad_norm": 0.18410591781139374, "learning_rate": 9.313860038019069e-05, "loss": 0.0189, "step": 21030 }, { "grad_norm": 0.2661606967449188, "learning_rate": 9.313023819093782e-05, "loss": 0.0216, "step": 21040 }, { "grad_norm": 0.23726262152194977, "learning_rate": 9.312187128502695e-05, "loss": 0.0185, "step": 21050 }, { "grad_norm": 0.23109526932239532, "learning_rate": 9.311349966337307e-05, "loss": 0.0183, "step": 21060 }, { "grad_norm": 0.21964679658412933, "learning_rate": 9.310512332689169e-05, "loss": 0.0184, "step": 21070 }, { "grad_norm": 0.2484031617641449, "learning_rate": 9.309674227649883e-05, "loss": 0.0207, "step": 21080 }, { "grad_norm": 0.3017354905605316, "learning_rate": 9.308835651311103e-05, "loss": 0.0187, "step": 21090 }, { "grad_norm": 0.17381635308265686, "learning_rate": 9.307996603764533e-05, "loss": 0.0185, "step": 21100 }, { "grad_norm": 0.31229856610298157, "learning_rate": 9.307157085101932e-05, "loss": 0.0182, "step": 21110 }, { "grad_norm": 0.2649416923522949, "learning_rate": 9.306317095415109e-05, "loss": 0.0194, "step": 21120 }, { "grad_norm": 0.25890353322029114, "learning_rate": 9.305476634795922e-05, "loss": 0.0207, "step": 21130 }, { "grad_norm": 0.22925637662410736, "learning_rate": 9.304635703336284e-05, "loss": 0.0192, "step": 21140 }, { "grad_norm": 0.22044314444065094, "learning_rate": 9.303794301128157e-05, "loss": 0.0204, "step": 21150 }, { "grad_norm": 0.21514050662517548, "learning_rate": 9.302952428263555e-05, "loss": 0.0185, "step": 21160 }, { "grad_norm": 0.23470480740070343, "learning_rate": 9.302110084834545e-05, "loss": 0.0198, "step": 21170 }, { "grad_norm": 0.18910859525203705, "learning_rate": 9.301267270933245e-05, "loss": 0.0193, "step": 21180 }, { "grad_norm": 0.278852254152298, "learning_rate": 9.300423986651823e-05, "loss": 0.0204, "step": 21190 }, { "grad_norm": 0.2923794686794281, "learning_rate": 9.299580232082501e-05, "loss": 0.0183, "step": 21200 }, { "grad_norm": 0.19968609511852264, "learning_rate": 9.298736007317547e-05, "loss": 0.0198, "step": 21210 }, { "grad_norm": 0.18685926496982574, "learning_rate": 9.297891312449288e-05, "loss": 0.0177, "step": 21220 }, { "grad_norm": 0.2348412126302719, "learning_rate": 9.297046147570094e-05, "loss": 0.0196, "step": 21230 }, { "grad_norm": 0.2972121238708496, "learning_rate": 9.296200512772396e-05, "loss": 0.0191, "step": 21240 }, { "grad_norm": 0.24952183663845062, "learning_rate": 9.295354408148668e-05, "loss": 0.0196, "step": 21250 }, { "grad_norm": 0.18268950283527374, "learning_rate": 9.294507833791441e-05, "loss": 0.0176, "step": 21260 }, { "grad_norm": 0.20999327301979065, "learning_rate": 9.293660789793295e-05, "loss": 0.0183, "step": 21270 }, { "grad_norm": 0.18971119821071625, "learning_rate": 9.292813276246858e-05, "loss": 0.018, "step": 21280 }, { "grad_norm": 0.24593515694141388, "learning_rate": 9.291965293244816e-05, "loss": 0.0195, "step": 21290 }, { "grad_norm": 0.1814117133617401, "learning_rate": 9.291116840879904e-05, "loss": 0.019, "step": 21300 }, { "grad_norm": 0.24541176855564117, "learning_rate": 9.290267919244904e-05, "loss": 0.0218, "step": 21310 }, { "grad_norm": 0.19267527759075165, "learning_rate": 9.289418528432655e-05, "loss": 0.0187, "step": 21320 }, { "grad_norm": 0.2519819140434265, "learning_rate": 9.288568668536045e-05, "loss": 0.0197, "step": 21330 }, { "grad_norm": 0.2040303349494934, "learning_rate": 9.287718339648013e-05, "loss": 0.0196, "step": 21340 }, { "grad_norm": 0.20593132078647614, "learning_rate": 9.28686754186155e-05, "loss": 0.0199, "step": 21350 }, { "grad_norm": 0.18193665146827698, "learning_rate": 9.286016275269698e-05, "loss": 0.0193, "step": 21360 }, { "grad_norm": 0.2946946322917938, "learning_rate": 9.285164539965551e-05, "loss": 0.0186, "step": 21370 }, { "grad_norm": 0.18554513156414032, "learning_rate": 9.284312336042251e-05, "loss": 0.0184, "step": 21380 }, { "grad_norm": 0.31187835335731506, "learning_rate": 9.283459663592996e-05, "loss": 0.0184, "step": 21390 }, { "grad_norm": 0.1783784180879593, "learning_rate": 9.282606522711033e-05, "loss": 0.0222, "step": 21400 }, { "grad_norm": 0.1809946894645691, "learning_rate": 9.281752913489657e-05, "loss": 0.0198, "step": 21410 }, { "grad_norm": 0.2050960659980774, "learning_rate": 9.280898836022222e-05, "loss": 0.0208, "step": 21420 }, { "grad_norm": 0.24679382145404816, "learning_rate": 9.280044290402126e-05, "loss": 0.018, "step": 21430 }, { "grad_norm": 0.13724152743816376, "learning_rate": 9.279189276722821e-05, "loss": 0.0183, "step": 21440 }, { "grad_norm": 0.20221000909805298, "learning_rate": 9.278333795077812e-05, "loss": 0.0174, "step": 21450 }, { "grad_norm": 0.24894553422927856, "learning_rate": 9.27747784556065e-05, "loss": 0.0183, "step": 21460 }, { "grad_norm": 0.2109840214252472, "learning_rate": 9.276621428264942e-05, "loss": 0.0208, "step": 21470 }, { "grad_norm": 0.30075255036354065, "learning_rate": 9.275764543284345e-05, "loss": 0.0193, "step": 21480 }, { "grad_norm": 0.2035180926322937, "learning_rate": 9.274907190712566e-05, "loss": 0.0199, "step": 21490 }, { "grad_norm": 0.19927673041820526, "learning_rate": 9.274049370643363e-05, "loss": 0.0191, "step": 21500 }, { "grad_norm": 0.2413325309753418, "learning_rate": 9.273191083170547e-05, "loss": 0.0195, "step": 21510 }, { "grad_norm": 0.23581498861312866, "learning_rate": 9.27233232838798e-05, "loss": 0.0215, "step": 21520 }, { "grad_norm": 0.1811482161283493, "learning_rate": 9.27147310638957e-05, "loss": 0.018, "step": 21530 }, { "grad_norm": 0.22487525641918182, "learning_rate": 9.270613417269286e-05, "loss": 0.0185, "step": 21540 }, { "grad_norm": 0.2829262912273407, "learning_rate": 9.269753261121138e-05, "loss": 0.0235, "step": 21550 }, { "grad_norm": 0.20744818449020386, "learning_rate": 9.268892638039194e-05, "loss": 0.0221, "step": 21560 }, { "grad_norm": 0.24132061004638672, "learning_rate": 9.268031548117569e-05, "loss": 0.0215, "step": 21570 }, { "grad_norm": 0.2940122187137604, "learning_rate": 9.26716999145043e-05, "loss": 0.0225, "step": 21580 }, { "grad_norm": 0.19680510461330414, "learning_rate": 9.266307968131998e-05, "loss": 0.0214, "step": 21590 }, { "grad_norm": 0.21686838567256927, "learning_rate": 9.26544547825654e-05, "loss": 0.0209, "step": 21600 }, { "grad_norm": 0.26226532459259033, "learning_rate": 9.264582521918376e-05, "loss": 0.0225, "step": 21610 }, { "grad_norm": 0.22969871759414673, "learning_rate": 9.263719099211881e-05, "loss": 0.0207, "step": 21620 }, { "grad_norm": 0.23706041276454926, "learning_rate": 9.262855210231476e-05, "loss": 0.0226, "step": 21630 }, { "grad_norm": 0.2256821244955063, "learning_rate": 9.261990855071633e-05, "loss": 0.0211, "step": 21640 }, { "grad_norm": 0.2825314700603485, "learning_rate": 9.261126033826878e-05, "loss": 0.0231, "step": 21650 }, { "grad_norm": 0.19865559041500092, "learning_rate": 9.260260746591786e-05, "loss": 0.0205, "step": 21660 }, { "grad_norm": 0.30570945143699646, "learning_rate": 9.259394993460985e-05, "loss": 0.021, "step": 21670 }, { "grad_norm": 0.315049409866333, "learning_rate": 9.258528774529151e-05, "loss": 0.0206, "step": 21680 }, { "grad_norm": 0.2392207682132721, "learning_rate": 9.257662089891013e-05, "loss": 0.0195, "step": 21690 }, { "grad_norm": 0.17972856760025024, "learning_rate": 9.25679493964135e-05, "loss": 0.0237, "step": 21700 }, { "grad_norm": 0.2852858304977417, "learning_rate": 9.255927323874994e-05, "loss": 0.0218, "step": 21710 }, { "grad_norm": 0.23528459668159485, "learning_rate": 9.255059242686822e-05, "loss": 0.0236, "step": 21720 }, { "grad_norm": 0.21151335537433624, "learning_rate": 9.254190696171769e-05, "loss": 0.0228, "step": 21730 }, { "grad_norm": 0.2999119758605957, "learning_rate": 9.25332168442482e-05, "loss": 0.0213, "step": 21740 }, { "grad_norm": 0.254161536693573, "learning_rate": 9.252452207541004e-05, "loss": 0.0226, "step": 21750 }, { "grad_norm": 0.1985405832529068, "learning_rate": 9.251582265615409e-05, "loss": 0.0224, "step": 21760 }, { "grad_norm": 0.25158798694610596, "learning_rate": 9.250711858743169e-05, "loss": 0.0196, "step": 21770 }, { "grad_norm": 0.1779521405696869, "learning_rate": 9.24984098701947e-05, "loss": 0.021, "step": 21780 }, { "grad_norm": 0.1730724573135376, "learning_rate": 9.248969650539552e-05, "loss": 0.0172, "step": 21790 }, { "grad_norm": 0.17682074010372162, "learning_rate": 9.2480978493987e-05, "loss": 0.0196, "step": 21800 }, { "grad_norm": 0.21101966500282288, "learning_rate": 9.247225583692256e-05, "loss": 0.0202, "step": 21810 }, { "grad_norm": 0.24424663186073303, "learning_rate": 9.246352853515607e-05, "loss": 0.02, "step": 21820 }, { "grad_norm": 0.2748229503631592, "learning_rate": 9.245479658964194e-05, "loss": 0.0205, "step": 21830 }, { "grad_norm": 0.2887852191925049, "learning_rate": 9.244606000133507e-05, "loss": 0.0197, "step": 21840 }, { "grad_norm": 0.20692239701747894, "learning_rate": 9.24373187711909e-05, "loss": 0.0214, "step": 21850 }, { "grad_norm": 0.22678712010383606, "learning_rate": 9.242857290016537e-05, "loss": 0.02, "step": 21860 }, { "grad_norm": 0.23649711906909943, "learning_rate": 9.241982238921488e-05, "loss": 0.0214, "step": 21870 }, { "grad_norm": 0.2562544345855713, "learning_rate": 9.24110672392964e-05, "loss": 0.0195, "step": 21880 }, { "grad_norm": 0.1942525953054428, "learning_rate": 9.240230745136737e-05, "loss": 0.0189, "step": 21890 }, { "grad_norm": 0.2588883638381958, "learning_rate": 9.239354302638575e-05, "loss": 0.0196, "step": 21900 }, { "grad_norm": 0.4536897540092468, "learning_rate": 9.238477396531e-05, "loss": 0.0196, "step": 21910 }, { "grad_norm": 0.22959479689598083, "learning_rate": 9.23760002690991e-05, "loss": 0.0189, "step": 21920 }, { "grad_norm": 0.2614370286464691, "learning_rate": 9.236722193871252e-05, "loss": 0.0275, "step": 21930 }, { "grad_norm": 0.23670409619808197, "learning_rate": 9.235843897511023e-05, "loss": 0.0268, "step": 21940 }, { "grad_norm": 0.24925965070724487, "learning_rate": 9.234965137925276e-05, "loss": 0.0232, "step": 21950 }, { "grad_norm": 0.2510637938976288, "learning_rate": 9.234085915210108e-05, "loss": 0.0234, "step": 21960 }, { "grad_norm": 0.31439757347106934, "learning_rate": 9.23320622946167e-05, "loss": 0.0253, "step": 21970 }, { "grad_norm": 0.2700434625148773, "learning_rate": 9.232326080776163e-05, "loss": 0.0287, "step": 21980 }, { "grad_norm": 0.21274307370185852, "learning_rate": 9.23144546924984e-05, "loss": 0.0271, "step": 21990 }, { "grad_norm": 0.2683340311050415, "learning_rate": 9.230564394979e-05, "loss": 0.024, "step": 22000 }, { "grad_norm": 0.2736203074455261, "learning_rate": 9.22968285806e-05, "loss": 0.0253, "step": 22010 }, { "grad_norm": 0.2161010503768921, "learning_rate": 9.228800858589242e-05, "loss": 0.0263, "step": 22020 }, { "grad_norm": 0.2527533173561096, "learning_rate": 9.227918396663179e-05, "loss": 0.0241, "step": 22030 }, { "grad_norm": 0.1978859156370163, "learning_rate": 9.227035472378319e-05, "loss": 0.0211, "step": 22040 }, { "grad_norm": 0.2048119753599167, "learning_rate": 9.226152085831213e-05, "loss": 0.0205, "step": 22050 }, { "grad_norm": 0.3331061899662018, "learning_rate": 9.22526823711847e-05, "loss": 0.0179, "step": 22060 }, { "grad_norm": 0.20466451346874237, "learning_rate": 9.224383926336745e-05, "loss": 0.0192, "step": 22070 }, { "grad_norm": 0.1846906691789627, "learning_rate": 9.223499153582744e-05, "loss": 0.0197, "step": 22080 }, { "grad_norm": 0.20317932963371277, "learning_rate": 9.222613918953226e-05, "loss": 0.0201, "step": 22090 }, { "grad_norm": 0.21641986072063446, "learning_rate": 9.221728222544999e-05, "loss": 0.0187, "step": 22100 }, { "grad_norm": 0.19780072569847107, "learning_rate": 9.22084206445492e-05, "loss": 0.02, "step": 22110 }, { "grad_norm": 0.21150338649749756, "learning_rate": 9.2199554447799e-05, "loss": 0.0177, "step": 22120 }, { "grad_norm": 0.20848733186721802, "learning_rate": 9.219068363616897e-05, "loss": 0.0221, "step": 22130 }, { "grad_norm": 0.25272491574287415, "learning_rate": 9.218180821062919e-05, "loss": 0.0203, "step": 22140 }, { "grad_norm": 0.26488903164863586, "learning_rate": 9.21729281721503e-05, "loss": 0.0205, "step": 22150 }, { "grad_norm": 0.19474224746227264, "learning_rate": 9.216404352170339e-05, "loss": 0.0218, "step": 22160 }, { "grad_norm": 0.18256081640720367, "learning_rate": 9.215515426026007e-05, "loss": 0.0201, "step": 22170 }, { "grad_norm": 0.2902123034000397, "learning_rate": 9.214626038879246e-05, "loss": 0.0191, "step": 22180 }, { "grad_norm": 0.3646388053894043, "learning_rate": 9.21373619082732e-05, "loss": 0.0208, "step": 22190 }, { "grad_norm": 0.20545418560504913, "learning_rate": 9.212845881967535e-05, "loss": 0.0196, "step": 22200 }, { "grad_norm": 0.15933144092559814, "learning_rate": 9.211955112397262e-05, "loss": 0.0208, "step": 22210 }, { "grad_norm": 0.22843801975250244, "learning_rate": 9.211063882213909e-05, "loss": 0.0178, "step": 22220 }, { "grad_norm": 0.21285215020179749, "learning_rate": 9.210172191514942e-05, "loss": 0.0197, "step": 22230 }, { "grad_norm": 0.16934067010879517, "learning_rate": 9.209280040397874e-05, "loss": 0.0185, "step": 22240 }, { "grad_norm": 0.3007216453552246, "learning_rate": 9.208387428960268e-05, "loss": 0.0194, "step": 22250 }, { "grad_norm": 0.23681402206420898, "learning_rate": 9.20749435729974e-05, "loss": 0.0177, "step": 22260 }, { "grad_norm": 0.23048090934753418, "learning_rate": 9.206600825513957e-05, "loss": 0.0191, "step": 22270 }, { "grad_norm": 0.21228088438510895, "learning_rate": 9.20570683370063e-05, "loss": 0.0218, "step": 22280 }, { "grad_norm": 0.2686452567577362, "learning_rate": 9.204812381957528e-05, "loss": 0.0193, "step": 22290 }, { "grad_norm": 0.28303202986717224, "learning_rate": 9.203917470382465e-05, "loss": 0.0218, "step": 22300 }, { "grad_norm": 0.23210057616233826, "learning_rate": 9.203022099073309e-05, "loss": 0.0182, "step": 22310 }, { "grad_norm": 0.20889413356781006, "learning_rate": 9.202126268127976e-05, "loss": 0.0183, "step": 22320 }, { "grad_norm": 0.261064350605011, "learning_rate": 9.20122997764443e-05, "loss": 0.0202, "step": 22330 }, { "grad_norm": 0.23054741322994232, "learning_rate": 9.200333227720692e-05, "loss": 0.0197, "step": 22340 }, { "grad_norm": 0.213895782828331, "learning_rate": 9.199436018454826e-05, "loss": 0.0196, "step": 22350 }, { "grad_norm": 0.15304048359394073, "learning_rate": 9.198538349944952e-05, "loss": 0.0173, "step": 22360 }, { "grad_norm": 0.2762056589126587, "learning_rate": 9.197640222289234e-05, "loss": 0.0207, "step": 22370 }, { "grad_norm": 0.27545085549354553, "learning_rate": 9.196741635585895e-05, "loss": 0.0193, "step": 22380 }, { "grad_norm": 0.21156053245067596, "learning_rate": 9.195842589933199e-05, "loss": 0.0209, "step": 22390 }, { "grad_norm": 0.21953754127025604, "learning_rate": 9.194943085429466e-05, "loss": 0.0206, "step": 22400 }, { "grad_norm": 0.24033111333847046, "learning_rate": 9.194043122173065e-05, "loss": 0.0232, "step": 22410 }, { "grad_norm": 0.287862628698349, "learning_rate": 9.193142700262413e-05, "loss": 0.0233, "step": 22420 }, { "grad_norm": 0.18244381248950958, "learning_rate": 9.192241819795979e-05, "loss": 0.0197, "step": 22430 }, { "grad_norm": 0.18805480003356934, "learning_rate": 9.191340480872284e-05, "loss": 0.0213, "step": 22440 }, { "grad_norm": 0.2822752892971039, "learning_rate": 9.190438683589895e-05, "loss": 0.0222, "step": 22450 }, { "grad_norm": 0.22316688299179077, "learning_rate": 9.189536428047432e-05, "loss": 0.0228, "step": 22460 }, { "grad_norm": 0.16943594813346863, "learning_rate": 9.188633714343564e-05, "loss": 0.0191, "step": 22470 }, { "grad_norm": 0.19448387622833252, "learning_rate": 9.18773054257701e-05, "loss": 0.0185, "step": 22480 }, { "grad_norm": 0.21856354176998138, "learning_rate": 9.18682691284654e-05, "loss": 0.0161, "step": 22490 }, { "grad_norm": 0.21052294969558716, "learning_rate": 9.185922825250974e-05, "loss": 0.0198, "step": 22500 }, { "grad_norm": 0.2779821455478668, "learning_rate": 9.185018279889181e-05, "loss": 0.0194, "step": 22510 }, { "grad_norm": 0.20186050236225128, "learning_rate": 9.184113276860082e-05, "loss": 0.0219, "step": 22520 }, { "grad_norm": 0.2034783959388733, "learning_rate": 9.183207816262645e-05, "loss": 0.0171, "step": 22530 }, { "grad_norm": 0.2050912082195282, "learning_rate": 9.182301898195891e-05, "loss": 0.0185, "step": 22540 }, { "grad_norm": 0.20238909125328064, "learning_rate": 9.181395522758889e-05, "loss": 0.0199, "step": 22550 }, { "grad_norm": 0.2814842462539673, "learning_rate": 9.180488690050759e-05, "loss": 0.0196, "step": 22560 }, { "grad_norm": 0.21841780841350555, "learning_rate": 9.179581400170671e-05, "loss": 0.021, "step": 22570 }, { "grad_norm": 0.19045710563659668, "learning_rate": 9.178673653217845e-05, "loss": 0.0192, "step": 22580 }, { "grad_norm": 0.175189808011055, "learning_rate": 9.177765449291551e-05, "loss": 0.0179, "step": 22590 }, { "grad_norm": 0.23661896586418152, "learning_rate": 9.176856788491109e-05, "loss": 0.0182, "step": 22600 }, { "grad_norm": 0.31312960386276245, "learning_rate": 9.175947670915887e-05, "loss": 0.021, "step": 22610 }, { "grad_norm": 0.18218488991260529, "learning_rate": 9.175038096665309e-05, "loss": 0.0189, "step": 22620 }, { "grad_norm": 0.21691769361495972, "learning_rate": 9.17412806583884e-05, "loss": 0.021, "step": 22630 }, { "grad_norm": 0.24953779578208923, "learning_rate": 9.173217578536002e-05, "loss": 0.0187, "step": 22640 }, { "grad_norm": 0.226119726896286, "learning_rate": 9.172306634856362e-05, "loss": 0.0205, "step": 22650 }, { "grad_norm": 0.2171160876750946, "learning_rate": 9.171395234899545e-05, "loss": 0.0209, "step": 22660 }, { "grad_norm": 0.2551782727241516, "learning_rate": 9.170483378765214e-05, "loss": 0.0219, "step": 22670 }, { "grad_norm": 0.3354468047618866, "learning_rate": 9.169571066553091e-05, "loss": 0.0203, "step": 22680 }, { "grad_norm": 0.24855011701583862, "learning_rate": 9.168658298362946e-05, "loss": 0.021, "step": 22690 }, { "grad_norm": 0.20644117891788483, "learning_rate": 9.167745074294598e-05, "loss": 0.0194, "step": 22700 }, { "grad_norm": 0.22370672225952148, "learning_rate": 9.166831394447913e-05, "loss": 0.0206, "step": 22710 }, { "grad_norm": 0.2119012326002121, "learning_rate": 9.165917258922812e-05, "loss": 0.0204, "step": 22720 }, { "grad_norm": 0.2112649828195572, "learning_rate": 9.165002667819262e-05, "loss": 0.0178, "step": 22730 }, { "grad_norm": 0.19605867564678192, "learning_rate": 9.164087621237282e-05, "loss": 0.0191, "step": 22740 }, { "grad_norm": 0.2395203858613968, "learning_rate": 9.163172119276942e-05, "loss": 0.0192, "step": 22750 }, { "grad_norm": 0.1722906231880188, "learning_rate": 9.162256162038358e-05, "loss": 0.0183, "step": 22760 }, { "grad_norm": 0.1691320836544037, "learning_rate": 9.161339749621698e-05, "loss": 0.0191, "step": 22770 }, { "grad_norm": 0.24630679190158844, "learning_rate": 9.160422882127177e-05, "loss": 0.0188, "step": 22780 }, { "grad_norm": 0.30066969990730286, "learning_rate": 9.159505559655069e-05, "loss": 0.0186, "step": 22790 }, { "grad_norm": 0.26856598258018494, "learning_rate": 9.158587782305684e-05, "loss": 0.018, "step": 22800 }, { "grad_norm": 0.23575624823570251, "learning_rate": 9.157669550179391e-05, "loss": 0.0203, "step": 22810 }, { "grad_norm": 0.23467205464839935, "learning_rate": 9.156750863376609e-05, "loss": 0.0164, "step": 22820 }, { "grad_norm": 0.3746332824230194, "learning_rate": 9.155831721997801e-05, "loss": 0.0178, "step": 22830 }, { "grad_norm": 0.23993150889873505, "learning_rate": 9.154912126143484e-05, "loss": 0.0185, "step": 22840 }, { "grad_norm": 0.16908077895641327, "learning_rate": 9.153992075914224e-05, "loss": 0.0183, "step": 22850 }, { "grad_norm": 0.17745167016983032, "learning_rate": 9.153071571410635e-05, "loss": 0.0181, "step": 22860 }, { "grad_norm": 0.17436498403549194, "learning_rate": 9.152150612733384e-05, "loss": 0.0185, "step": 22870 }, { "grad_norm": 0.18214991688728333, "learning_rate": 9.151229199983184e-05, "loss": 0.0187, "step": 22880 }, { "grad_norm": 0.2580423057079315, "learning_rate": 9.150307333260802e-05, "loss": 0.0184, "step": 22890 }, { "grad_norm": 0.19146394729614258, "learning_rate": 9.149385012667048e-05, "loss": 0.0175, "step": 22900 }, { "grad_norm": 0.21045534312725067, "learning_rate": 9.148462238302788e-05, "loss": 0.0182, "step": 22910 }, { "grad_norm": 0.1819983273744583, "learning_rate": 9.147539010268936e-05, "loss": 0.0189, "step": 22920 }, { "grad_norm": 0.28856563568115234, "learning_rate": 9.14661532866645e-05, "loss": 0.0207, "step": 22930 }, { "grad_norm": 0.2065095752477646, "learning_rate": 9.145691193596348e-05, "loss": 0.0198, "step": 22940 }, { "grad_norm": 0.36090797185897827, "learning_rate": 9.144766605159691e-05, "loss": 0.0175, "step": 22950 }, { "grad_norm": 0.19218483567237854, "learning_rate": 9.14384156345759e-05, "loss": 0.0179, "step": 22960 }, { "grad_norm": 0.2927502989768982, "learning_rate": 9.142916068591204e-05, "loss": 0.0176, "step": 22970 }, { "grad_norm": 0.2495180070400238, "learning_rate": 9.141990120661746e-05, "loss": 0.021, "step": 22980 }, { "grad_norm": 0.20300106704235077, "learning_rate": 9.141063719770475e-05, "loss": 0.0202, "step": 22990 }, { "grad_norm": 0.21560733020305634, "learning_rate": 9.140136866018704e-05, "loss": 0.0164, "step": 23000 }, { "grad_norm": 0.2047557234764099, "learning_rate": 9.139209559507788e-05, "loss": 0.0187, "step": 23010 }, { "grad_norm": 0.2688652276992798, "learning_rate": 9.13828180033914e-05, "loss": 0.0187, "step": 23020 }, { "grad_norm": 0.32849326729774475, "learning_rate": 9.137353588614212e-05, "loss": 0.0192, "step": 23030 }, { "grad_norm": 0.472410649061203, "learning_rate": 9.136424924434519e-05, "loss": 0.0204, "step": 23040 }, { "grad_norm": 0.2372087687253952, "learning_rate": 9.135495807901615e-05, "loss": 0.0214, "step": 23050 }, { "grad_norm": 0.26460933685302734, "learning_rate": 9.134566239117108e-05, "loss": 0.0202, "step": 23060 }, { "grad_norm": 0.26572316884994507, "learning_rate": 9.13363621818265e-05, "loss": 0.0201, "step": 23070 }, { "grad_norm": 0.24463839828968048, "learning_rate": 9.132705745199953e-05, "loss": 0.0205, "step": 23080 }, { "grad_norm": 0.23979438841342926, "learning_rate": 9.131774820270768e-05, "loss": 0.0206, "step": 23090 }, { "grad_norm": 0.23389533162117004, "learning_rate": 9.130843443496901e-05, "loss": 0.0202, "step": 23100 }, { "grad_norm": 0.19435901939868927, "learning_rate": 9.129911614980206e-05, "loss": 0.019, "step": 23110 }, { "grad_norm": 0.1890551894903183, "learning_rate": 9.128979334822584e-05, "loss": 0.0191, "step": 23120 }, { "grad_norm": 0.1958026885986328, "learning_rate": 9.128046603125992e-05, "loss": 0.0193, "step": 23130 }, { "grad_norm": 0.2293623983860016, "learning_rate": 9.12711341999243e-05, "loss": 0.0191, "step": 23140 }, { "grad_norm": 0.29990217089653015, "learning_rate": 9.12617978552395e-05, "loss": 0.0215, "step": 23150 }, { "grad_norm": 0.2101905196905136, "learning_rate": 9.12524569982265e-05, "loss": 0.0202, "step": 23160 }, { "grad_norm": 0.2836284339427948, "learning_rate": 9.124311162990684e-05, "loss": 0.019, "step": 23170 }, { "grad_norm": 0.2160191535949707, "learning_rate": 9.12337617513025e-05, "loss": 0.0204, "step": 23180 }, { "grad_norm": 0.19839808344841003, "learning_rate": 9.122440736343596e-05, "loss": 0.0172, "step": 23190 }, { "grad_norm": 0.2139412760734558, "learning_rate": 9.12150484673302e-05, "loss": 0.017, "step": 23200 }, { "grad_norm": 0.17483177781105042, "learning_rate": 9.120568506400873e-05, "loss": 0.0163, "step": 23210 }, { "grad_norm": 0.21111391484737396, "learning_rate": 9.119631715449548e-05, "loss": 0.0178, "step": 23220 }, { "grad_norm": 0.2115418165922165, "learning_rate": 9.118694473981493e-05, "loss": 0.0169, "step": 23230 }, { "grad_norm": 0.23387153446674347, "learning_rate": 9.117756782099203e-05, "loss": 0.0191, "step": 23240 }, { "grad_norm": 0.2049015313386917, "learning_rate": 9.11681863990522e-05, "loss": 0.0178, "step": 23250 }, { "grad_norm": 0.17177025973796844, "learning_rate": 9.115880047502142e-05, "loss": 0.0183, "step": 23260 }, { "grad_norm": 0.2255864441394806, "learning_rate": 9.114941004992609e-05, "loss": 0.0175, "step": 23270 }, { "grad_norm": 0.28769204020500183, "learning_rate": 9.114001512479317e-05, "loss": 0.0195, "step": 23280 }, { "grad_norm": 0.14655551314353943, "learning_rate": 9.113061570065003e-05, "loss": 0.0175, "step": 23290 }, { "grad_norm": 0.20252922177314758, "learning_rate": 9.112121177852459e-05, "loss": 0.0174, "step": 23300 }, { "grad_norm": 0.21081748604774475, "learning_rate": 9.111180335944527e-05, "loss": 0.0177, "step": 23310 }, { "grad_norm": 0.19955652952194214, "learning_rate": 9.110239044444093e-05, "loss": 0.0165, "step": 23320 }, { "grad_norm": 0.3315296173095703, "learning_rate": 9.109297303454099e-05, "loss": 0.0168, "step": 23330 }, { "grad_norm": 0.22602318227291107, "learning_rate": 9.108355113077526e-05, "loss": 0.0204, "step": 23340 }, { "grad_norm": 0.23176738619804382, "learning_rate": 9.107412473417419e-05, "loss": 0.0172, "step": 23350 }, { "grad_norm": 0.2577732503414154, "learning_rate": 9.106469384576858e-05, "loss": 0.0187, "step": 23360 }, { "grad_norm": 0.24860094487667084, "learning_rate": 9.105525846658978e-05, "loss": 0.0194, "step": 23370 }, { "grad_norm": 0.25743624567985535, "learning_rate": 9.104581859766965e-05, "loss": 0.0185, "step": 23380 }, { "grad_norm": 0.2834676206111908, "learning_rate": 9.10363742400405e-05, "loss": 0.0179, "step": 23390 }, { "grad_norm": 0.26762500405311584, "learning_rate": 9.102692539473518e-05, "loss": 0.0185, "step": 23400 }, { "grad_norm": 0.254963219165802, "learning_rate": 9.101747206278697e-05, "loss": 0.0181, "step": 23410 }, { "grad_norm": 0.20761364698410034, "learning_rate": 9.100801424522968e-05, "loss": 0.0218, "step": 23420 }, { "grad_norm": 0.2040819525718689, "learning_rate": 9.099855194309762e-05, "loss": 0.0214, "step": 23430 }, { "grad_norm": 0.30207720398902893, "learning_rate": 9.098908515742554e-05, "loss": 0.0197, "step": 23440 }, { "grad_norm": 0.24352405965328217, "learning_rate": 9.097961388924873e-05, "loss": 0.0182, "step": 23450 }, { "grad_norm": 0.2627069652080536, "learning_rate": 9.097013813960298e-05, "loss": 0.0218, "step": 23460 }, { "grad_norm": 0.19492800533771515, "learning_rate": 9.09606579095245e-05, "loss": 0.0214, "step": 23470 }, { "grad_norm": 0.5310347080230713, "learning_rate": 9.095117320005008e-05, "loss": 0.0218, "step": 23480 }, { "grad_norm": 0.3329215943813324, "learning_rate": 9.094168401221691e-05, "loss": 0.02, "step": 23490 }, { "grad_norm": 0.23966597020626068, "learning_rate": 9.093219034706273e-05, "loss": 0.0213, "step": 23500 }, { "grad_norm": 0.21168732643127441, "learning_rate": 9.092269220562577e-05, "loss": 0.0189, "step": 23510 }, { "grad_norm": 0.21082942187786102, "learning_rate": 9.09131895889447e-05, "loss": 0.0205, "step": 23520 }, { "grad_norm": 0.2047518640756607, "learning_rate": 9.090368249805873e-05, "loss": 0.0195, "step": 23530 }, { "grad_norm": 0.27887895703315735, "learning_rate": 9.089417093400754e-05, "loss": 0.0213, "step": 23540 }, { "grad_norm": 0.193916454911232, "learning_rate": 9.088465489783131e-05, "loss": 0.0209, "step": 23550 }, { "grad_norm": 0.19401387870311737, "learning_rate": 9.087513439057068e-05, "loss": 0.0218, "step": 23560 }, { "grad_norm": 0.2993350625038147, "learning_rate": 9.08656094132668e-05, "loss": 0.0218, "step": 23570 }, { "grad_norm": 0.2851417362689972, "learning_rate": 9.085607996696134e-05, "loss": 0.0181, "step": 23580 }, { "grad_norm": 0.2938733696937561, "learning_rate": 9.084654605269639e-05, "loss": 0.0193, "step": 23590 }, { "grad_norm": 0.21597394347190857, "learning_rate": 9.083700767151457e-05, "loss": 0.0198, "step": 23600 }, { "grad_norm": 0.2787494361400604, "learning_rate": 9.082746482445898e-05, "loss": 0.0192, "step": 23610 }, { "grad_norm": 0.19830094277858734, "learning_rate": 9.081791751257325e-05, "loss": 0.0209, "step": 23620 }, { "grad_norm": 0.29773592948913574, "learning_rate": 9.080836573690142e-05, "loss": 0.0208, "step": 23630 }, { "grad_norm": 0.23745635151863098, "learning_rate": 9.079880949848805e-05, "loss": 0.0178, "step": 23640 }, { "grad_norm": 0.22531019151210785, "learning_rate": 9.078924879837822e-05, "loss": 0.0197, "step": 23650 }, { "grad_norm": 0.4900243282318115, "learning_rate": 9.077968363761747e-05, "loss": 0.0238, "step": 23660 }, { "grad_norm": 0.21661925315856934, "learning_rate": 9.077011401725182e-05, "loss": 0.0219, "step": 23670 }, { "grad_norm": 0.26309648156166077, "learning_rate": 9.07605399383278e-05, "loss": 0.0186, "step": 23680 }, { "grad_norm": 0.1789505034685135, "learning_rate": 9.075096140189243e-05, "loss": 0.0182, "step": 23690 }, { "grad_norm": 0.19843465089797974, "learning_rate": 9.074137840899318e-05, "loss": 0.0181, "step": 23700 }, { "grad_norm": 0.25787267088890076, "learning_rate": 9.073179096067804e-05, "loss": 0.0191, "step": 23710 }, { "grad_norm": 0.20753391087055206, "learning_rate": 9.072219905799549e-05, "loss": 0.0212, "step": 23720 }, { "grad_norm": 0.16827692091464996, "learning_rate": 9.071260270199447e-05, "loss": 0.0207, "step": 23730 }, { "grad_norm": 0.19511806964874268, "learning_rate": 9.070300189372441e-05, "loss": 0.0194, "step": 23740 }, { "grad_norm": 0.2910584807395935, "learning_rate": 9.069339663423528e-05, "loss": 0.0208, "step": 23750 }, { "grad_norm": 0.18363627791404724, "learning_rate": 9.068378692457747e-05, "loss": 0.023, "step": 23760 }, { "grad_norm": 0.1920173615217209, "learning_rate": 9.067417276580189e-05, "loss": 0.0211, "step": 23770 }, { "grad_norm": 0.22475263476371765, "learning_rate": 9.066455415895993e-05, "loss": 0.0218, "step": 23780 }, { "grad_norm": 0.286459356546402, "learning_rate": 9.065493110510346e-05, "loss": 0.0204, "step": 23790 }, { "grad_norm": 0.2084839940071106, "learning_rate": 9.064530360528484e-05, "loss": 0.0189, "step": 23800 }, { "grad_norm": 0.26682502031326294, "learning_rate": 9.063567166055695e-05, "loss": 0.0207, "step": 23810 }, { "grad_norm": 0.2786731421947479, "learning_rate": 9.062603527197308e-05, "loss": 0.0189, "step": 23820 }, { "grad_norm": 0.2583346366882324, "learning_rate": 9.06163944405871e-05, "loss": 0.0209, "step": 23830 }, { "grad_norm": 0.16253244876861572, "learning_rate": 9.060674916745327e-05, "loss": 0.0192, "step": 23840 }, { "grad_norm": 0.2752414643764496, "learning_rate": 9.05970994536264e-05, "loss": 0.0209, "step": 23850 }, { "grad_norm": 0.22276267409324646, "learning_rate": 9.05874453001618e-05, "loss": 0.0199, "step": 23860 }, { "grad_norm": 0.2034042924642563, "learning_rate": 9.057778670811517e-05, "loss": 0.0183, "step": 23870 }, { "grad_norm": 0.25128674507141113, "learning_rate": 9.056812367854281e-05, "loss": 0.0164, "step": 23880 }, { "grad_norm": 0.1813226044178009, "learning_rate": 9.055845621250143e-05, "loss": 0.0173, "step": 23890 }, { "grad_norm": 0.1932104378938675, "learning_rate": 9.054878431104825e-05, "loss": 0.0204, "step": 23900 }, { "grad_norm": 0.2272658795118332, "learning_rate": 9.0539107975241e-05, "loss": 0.0183, "step": 23910 }, { "grad_norm": 0.2552947700023651, "learning_rate": 9.052942720613784e-05, "loss": 0.0178, "step": 23920 }, { "grad_norm": 0.3506920039653778, "learning_rate": 9.051974200479745e-05, "loss": 0.0187, "step": 23930 }, { "grad_norm": 0.24877630174160004, "learning_rate": 9.051005237227901e-05, "loss": 0.0175, "step": 23940 }, { "grad_norm": 0.2112339437007904, "learning_rate": 9.050035830964215e-05, "loss": 0.0189, "step": 23950 }, { "grad_norm": 0.2066977322101593, "learning_rate": 9.049065981794698e-05, "loss": 0.0198, "step": 23960 }, { "grad_norm": 0.24541038274765015, "learning_rate": 9.048095689825414e-05, "loss": 0.0218, "step": 23970 }, { "grad_norm": 0.23728010058403015, "learning_rate": 9.047124955162472e-05, "loss": 0.0194, "step": 23980 }, { "grad_norm": 0.17998908460140228, "learning_rate": 9.046153777912028e-05, "loss": 0.0161, "step": 23990 }, { "grad_norm": 0.1979314386844635, "learning_rate": 9.045182158180292e-05, "loss": 0.0168, "step": 24000 }, { "grad_norm": 0.20176732540130615, "learning_rate": 9.044210096073516e-05, "loss": 0.0205, "step": 24010 }, { "grad_norm": 0.21898102760314941, "learning_rate": 9.043237591698004e-05, "loss": 0.0178, "step": 24020 }, { "grad_norm": 0.22609035670757294, "learning_rate": 9.04226464516011e-05, "loss": 0.0177, "step": 24030 }, { "grad_norm": 0.2411140501499176, "learning_rate": 9.041291256566229e-05, "loss": 0.017, "step": 24040 }, { "grad_norm": 0.21940977871418, "learning_rate": 9.040317426022814e-05, "loss": 0.0172, "step": 24050 }, { "grad_norm": 0.2091631144285202, "learning_rate": 9.03934315363636e-05, "loss": 0.0174, "step": 24060 }, { "grad_norm": 0.2149558812379837, "learning_rate": 9.038368439513409e-05, "loss": 0.017, "step": 24070 }, { "grad_norm": 0.23023469746112823, "learning_rate": 9.03739328376056e-05, "loss": 0.0192, "step": 24080 }, { "grad_norm": 0.23555050790309906, "learning_rate": 9.036417686484451e-05, "loss": 0.0174, "step": 24090 }, { "grad_norm": 0.20598511397838593, "learning_rate": 9.035441647791773e-05, "loss": 0.0192, "step": 24100 }, { "grad_norm": 0.21860207617282867, "learning_rate": 9.034465167789263e-05, "loss": 0.0174, "step": 24110 }, { "grad_norm": 0.1956498920917511, "learning_rate": 9.033488246583706e-05, "loss": 0.0172, "step": 24120 }, { "grad_norm": 0.20429834723472595, "learning_rate": 9.032510884281941e-05, "loss": 0.0191, "step": 24130 }, { "grad_norm": 0.2592304050922394, "learning_rate": 9.031533080990848e-05, "loss": 0.0164, "step": 24140 }, { "grad_norm": 0.1548309326171875, "learning_rate": 9.030554836817358e-05, "loss": 0.0156, "step": 24150 }, { "grad_norm": 0.14394110441207886, "learning_rate": 9.029576151868451e-05, "loss": 0.0201, "step": 24160 }, { "grad_norm": 0.22224494814872742, "learning_rate": 9.028597026251155e-05, "loss": 0.0174, "step": 24170 }, { "grad_norm": 0.22434544563293457, "learning_rate": 9.027617460072547e-05, "loss": 0.0176, "step": 24180 }, { "grad_norm": 0.2420431226491928, "learning_rate": 9.026637453439745e-05, "loss": 0.015, "step": 24190 }, { "grad_norm": 0.2713932991027832, "learning_rate": 9.025657006459927e-05, "loss": 0.0194, "step": 24200 }, { "grad_norm": 0.1980292797088623, "learning_rate": 9.024676119240311e-05, "loss": 0.0179, "step": 24210 }, { "grad_norm": 0.16814936697483063, "learning_rate": 9.023694791888166e-05, "loss": 0.0181, "step": 24220 }, { "grad_norm": 0.17311538755893707, "learning_rate": 9.022713024510808e-05, "loss": 0.016, "step": 24230 }, { "grad_norm": 0.20137499272823334, "learning_rate": 9.021730817215601e-05, "loss": 0.0161, "step": 24240 }, { "grad_norm": 0.196821391582489, "learning_rate": 9.02074817010996e-05, "loss": 0.0158, "step": 24250 }, { "grad_norm": 0.23541173338890076, "learning_rate": 9.019765083301342e-05, "loss": 0.0164, "step": 24260 }, { "grad_norm": 0.2389344722032547, "learning_rate": 9.01878155689726e-05, "loss": 0.0175, "step": 24270 }, { "grad_norm": 0.2529628872871399, "learning_rate": 9.017797591005268e-05, "loss": 0.0185, "step": 24280 }, { "grad_norm": 0.17782871425151825, "learning_rate": 9.016813185732972e-05, "loss": 0.0189, "step": 24290 }, { "grad_norm": 0.25916770100593567, "learning_rate": 9.015828341188027e-05, "loss": 0.0213, "step": 24300 }, { "grad_norm": 0.27403199672698975, "learning_rate": 9.01484305747813e-05, "loss": 0.0234, "step": 24310 }, { "grad_norm": 0.24688169360160828, "learning_rate": 9.013857334711033e-05, "loss": 0.0212, "step": 24320 }, { "grad_norm": 0.22575555741786957, "learning_rate": 9.012871172994534e-05, "loss": 0.0205, "step": 24330 }, { "grad_norm": 0.1942281424999237, "learning_rate": 9.011884572436476e-05, "loss": 0.0203, "step": 24340 }, { "grad_norm": 0.20253553986549377, "learning_rate": 9.010897533144754e-05, "loss": 0.0206, "step": 24350 }, { "grad_norm": 0.25751784443855286, "learning_rate": 9.009910055227306e-05, "loss": 0.0195, "step": 24360 }, { "grad_norm": 0.2258516252040863, "learning_rate": 9.008922138792124e-05, "loss": 0.0198, "step": 24370 }, { "grad_norm": 0.21450236439704895, "learning_rate": 9.007933783947244e-05, "loss": 0.0197, "step": 24380 }, { "grad_norm": 0.2969930171966553, "learning_rate": 9.006944990800752e-05, "loss": 0.0206, "step": 24390 }, { "grad_norm": 0.17319528758525848, "learning_rate": 9.005955759460779e-05, "loss": 0.0205, "step": 24400 }, { "grad_norm": 0.23955367505550385, "learning_rate": 9.004966090035508e-05, "loss": 0.0228, "step": 24410 }, { "grad_norm": 0.22459325194358826, "learning_rate": 9.003975982633166e-05, "loss": 0.0198, "step": 24420 }, { "grad_norm": 0.20125970244407654, "learning_rate": 9.00298543736203e-05, "loss": 0.0191, "step": 24430 }, { "grad_norm": 0.22103732824325562, "learning_rate": 9.001994454330427e-05, "loss": 0.0214, "step": 24440 }, { "grad_norm": 0.21064947545528412, "learning_rate": 9.001003033646727e-05, "loss": 0.0222, "step": 24450 }, { "grad_norm": 0.23335854709148407, "learning_rate": 9.00001117541935e-05, "loss": 0.0161, "step": 24460 }, { "grad_norm": 0.16314640641212463, "learning_rate": 8.999018879756764e-05, "loss": 0.0162, "step": 24470 }, { "grad_norm": 0.19262070953845978, "learning_rate": 8.998026146767487e-05, "loss": 0.0187, "step": 24480 }, { "grad_norm": 0.27746835350990295, "learning_rate": 8.99703297656008e-05, "loss": 0.0164, "step": 24490 }, { "grad_norm": 0.2693499028682709, "learning_rate": 8.996039369243156e-05, "loss": 0.0196, "step": 24500 }, { "grad_norm": 0.3466659188270569, "learning_rate": 8.995045324925378e-05, "loss": 0.0202, "step": 24510 }, { "grad_norm": 0.15904086828231812, "learning_rate": 8.994050843715448e-05, "loss": 0.0201, "step": 24520 }, { "grad_norm": 0.29160773754119873, "learning_rate": 8.993055925722121e-05, "loss": 0.021, "step": 24530 }, { "grad_norm": 0.19170278310775757, "learning_rate": 8.992060571054202e-05, "loss": 0.0189, "step": 24540 }, { "grad_norm": 0.2939615249633789, "learning_rate": 8.991064779820542e-05, "loss": 0.0216, "step": 24550 }, { "grad_norm": 0.29023808240890503, "learning_rate": 8.990068552130036e-05, "loss": 0.0221, "step": 24560 }, { "grad_norm": 0.25106188654899597, "learning_rate": 8.989071888091634e-05, "loss": 0.0198, "step": 24570 }, { "grad_norm": 0.17705632746219635, "learning_rate": 8.988074787814329e-05, "loss": 0.0224, "step": 24580 }, { "grad_norm": 0.29151779413223267, "learning_rate": 8.987077251407158e-05, "loss": 0.0206, "step": 24590 }, { "grad_norm": 0.2137976884841919, "learning_rate": 8.986079278979216e-05, "loss": 0.0212, "step": 24600 }, { "grad_norm": 0.1679777055978775, "learning_rate": 8.985080870639635e-05, "loss": 0.0186, "step": 24610 }, { "grad_norm": 0.26577457785606384, "learning_rate": 8.984082026497603e-05, "loss": 0.022, "step": 24620 }, { "grad_norm": 0.3260345160961151, "learning_rate": 8.98308274666235e-05, "loss": 0.0186, "step": 24630 }, { "grad_norm": 0.1953631490468979, "learning_rate": 8.982083031243155e-05, "loss": 0.018, "step": 24640 }, { "grad_norm": 0.1962016522884369, "learning_rate": 8.98108288034935e-05, "loss": 0.0168, "step": 24650 }, { "grad_norm": 0.20729686319828033, "learning_rate": 8.980082294090305e-05, "loss": 0.0204, "step": 24660 }, { "grad_norm": 0.23199765384197235, "learning_rate": 8.979081272575443e-05, "loss": 0.0176, "step": 24670 }, { "grad_norm": 0.19400595128536224, "learning_rate": 8.978079815914236e-05, "loss": 0.0171, "step": 24680 }, { "grad_norm": 0.27937495708465576, "learning_rate": 8.977077924216202e-05, "loss": 0.0202, "step": 24690 }, { "grad_norm": 0.2940768599510193, "learning_rate": 8.976075597590905e-05, "loss": 0.0185, "step": 24700 }, { "grad_norm": 0.2650950849056244, "learning_rate": 8.975072836147958e-05, "loss": 0.0227, "step": 24710 }, { "grad_norm": 0.18581554293632507, "learning_rate": 8.974069639997025e-05, "loss": 0.0195, "step": 24720 }, { "grad_norm": 0.20594966411590576, "learning_rate": 8.973066009247808e-05, "loss": 0.0198, "step": 24730 }, { "grad_norm": 0.22203120589256287, "learning_rate": 8.972061944010066e-05, "loss": 0.0201, "step": 24740 }, { "grad_norm": 0.22308504581451416, "learning_rate": 8.971057444393603e-05, "loss": 0.0204, "step": 24750 }, { "grad_norm": 0.22077609598636627, "learning_rate": 8.970052510508268e-05, "loss": 0.0194, "step": 24760 }, { "grad_norm": 0.2103995829820633, "learning_rate": 8.969047142463959e-05, "loss": 0.0202, "step": 24770 }, { "grad_norm": 0.2744598984718323, "learning_rate": 8.968041340370621e-05, "loss": 0.0192, "step": 24780 }, { "grad_norm": 0.2170085310935974, "learning_rate": 8.96703510433825e-05, "loss": 0.0202, "step": 24790 }, { "grad_norm": 0.2429370880126953, "learning_rate": 8.966028434476883e-05, "loss": 0.0192, "step": 24800 }, { "grad_norm": 0.16122227907180786, "learning_rate": 8.96502133089661e-05, "loss": 0.0174, "step": 24810 }, { "grad_norm": 0.19414478540420532, "learning_rate": 8.964013793707564e-05, "loss": 0.0195, "step": 24820 }, { "grad_norm": 0.17131242156028748, "learning_rate": 8.963005823019932e-05, "loss": 0.0177, "step": 24830 }, { "grad_norm": 0.26894789934158325, "learning_rate": 8.961997418943939e-05, "loss": 0.0162, "step": 24840 }, { "grad_norm": 0.27557554841041565, "learning_rate": 8.960988581589865e-05, "loss": 0.0202, "step": 24850 }, { "grad_norm": 0.27644094824790955, "learning_rate": 8.959979311068037e-05, "loss": 0.0314, "step": 24860 }, { "grad_norm": 0.2365650087594986, "learning_rate": 8.958969607488823e-05, "loss": 0.0249, "step": 24870 }, { "grad_norm": 0.2839988172054291, "learning_rate": 8.957959470962647e-05, "loss": 0.0347, "step": 24880 }, { "grad_norm": 0.2665570080280304, "learning_rate": 8.956948901599971e-05, "loss": 0.0299, "step": 24890 }, { "grad_norm": 0.20026157796382904, "learning_rate": 8.955937899511315e-05, "loss": 0.0225, "step": 24900 }, { "grad_norm": 0.2621588408946991, "learning_rate": 8.954926464807238e-05, "loss": 0.0246, "step": 24910 }, { "grad_norm": 0.20782876014709473, "learning_rate": 8.953914597598347e-05, "loss": 0.0206, "step": 24920 }, { "grad_norm": 0.2707827389240265, "learning_rate": 8.952902297995303e-05, "loss": 0.0211, "step": 24930 }, { "grad_norm": 0.26839110255241394, "learning_rate": 8.951889566108804e-05, "loss": 0.026, "step": 24940 }, { "grad_norm": 0.25718212127685547, "learning_rate": 8.950876402049606e-05, "loss": 0.0182, "step": 24950 }, { "grad_norm": 0.20133984088897705, "learning_rate": 8.949862805928504e-05, "loss": 0.0222, "step": 24960 }, { "grad_norm": 0.24067118763923645, "learning_rate": 8.948848777856343e-05, "loss": 0.0202, "step": 24970 }, { "grad_norm": 0.29515984654426575, "learning_rate": 8.947834317944017e-05, "loss": 0.0217, "step": 24980 }, { "grad_norm": 0.3292388617992401, "learning_rate": 8.946819426302466e-05, "loss": 0.0233, "step": 24990 }, { "grad_norm": 0.2648528814315796, "learning_rate": 8.945804103042676e-05, "loss": 0.0273, "step": 25000 }, { "grad_norm": 0.21771226823329926, "learning_rate": 8.944788348275681e-05, "loss": 0.0208, "step": 25010 }, { "grad_norm": 0.1658046543598175, "learning_rate": 8.943772162112565e-05, "loss": 0.0193, "step": 25020 }, { "grad_norm": 0.2326301634311676, "learning_rate": 8.942755544664454e-05, "loss": 0.0206, "step": 25030 }, { "grad_norm": 0.1870817393064499, "learning_rate": 8.941738496042525e-05, "loss": 0.0216, "step": 25040 }, { "grad_norm": 0.22656410932540894, "learning_rate": 8.940721016357999e-05, "loss": 0.0183, "step": 25050 }, { "grad_norm": 0.25858020782470703, "learning_rate": 8.939703105722148e-05, "loss": 0.0188, "step": 25060 }, { "grad_norm": 0.20796625316143036, "learning_rate": 8.93868476424629e-05, "loss": 0.0167, "step": 25070 }, { "grad_norm": 0.20156912505626678, "learning_rate": 8.937665992041786e-05, "loss": 0.0223, "step": 25080 }, { "grad_norm": 0.1907363384962082, "learning_rate": 8.93664678922005e-05, "loss": 0.0173, "step": 25090 }, { "grad_norm": 0.18910861015319824, "learning_rate": 8.93562715589254e-05, "loss": 0.0176, "step": 25100 }, { "grad_norm": 0.3886392116546631, "learning_rate": 8.934607092170762e-05, "loss": 0.0206, "step": 25110 }, { "grad_norm": 0.1414688378572464, "learning_rate": 8.933586598166266e-05, "loss": 0.0166, "step": 25120 }, { "grad_norm": 0.21863359212875366, "learning_rate": 8.932565673990655e-05, "loss": 0.0196, "step": 25130 }, { "grad_norm": 0.2235213667154312, "learning_rate": 8.931544319755574e-05, "loss": 0.0171, "step": 25140 }, { "grad_norm": 0.19372420012950897, "learning_rate": 8.930522535572718e-05, "loss": 0.0172, "step": 25150 }, { "grad_norm": 0.19875337183475494, "learning_rate": 8.929500321553826e-05, "loss": 0.0191, "step": 25160 }, { "grad_norm": 0.2728528380393982, "learning_rate": 8.928477677810686e-05, "loss": 0.0182, "step": 25170 }, { "grad_norm": 0.2093803584575653, "learning_rate": 8.927454604455137e-05, "loss": 0.0194, "step": 25180 }, { "grad_norm": 0.2509765326976776, "learning_rate": 8.926431101599053e-05, "loss": 0.0179, "step": 25190 }, { "grad_norm": 0.15305382013320923, "learning_rate": 8.925407169354369e-05, "loss": 0.0177, "step": 25200 }, { "grad_norm": 0.2141573429107666, "learning_rate": 8.92438280783306e-05, "loss": 0.0205, "step": 25210 }, { "grad_norm": 0.230525940656662, "learning_rate": 8.923358017147146e-05, "loss": 0.0216, "step": 25220 }, { "grad_norm": 0.18549276888370514, "learning_rate": 8.922332797408697e-05, "loss": 0.0189, "step": 25230 }, { "grad_norm": 0.1939157247543335, "learning_rate": 8.921307148729831e-05, "loss": 0.0176, "step": 25240 }, { "grad_norm": 0.18578481674194336, "learning_rate": 8.920281071222712e-05, "loss": 0.0187, "step": 25250 }, { "grad_norm": 0.2437255084514618, "learning_rate": 8.919254564999548e-05, "loss": 0.0182, "step": 25260 }, { "grad_norm": 0.19134601950645447, "learning_rate": 8.918227630172598e-05, "loss": 0.0191, "step": 25270 }, { "grad_norm": 0.20529493689537048, "learning_rate": 8.917200266854165e-05, "loss": 0.0175, "step": 25280 }, { "grad_norm": 0.35322684049606323, "learning_rate": 8.9161724751566e-05, "loss": 0.0193, "step": 25290 }, { "grad_norm": 0.2516476511955261, "learning_rate": 8.915144255192302e-05, "loss": 0.0167, "step": 25300 }, { "grad_norm": 0.23310241103172302, "learning_rate": 8.914115607073714e-05, "loss": 0.0176, "step": 25310 }, { "grad_norm": 0.24903613328933716, "learning_rate": 8.913086530913327e-05, "loss": 0.0171, "step": 25320 }, { "grad_norm": 0.22144685685634613, "learning_rate": 8.912057026823681e-05, "loss": 0.0176, "step": 25330 }, { "grad_norm": 0.19926409423351288, "learning_rate": 8.91102709491736e-05, "loss": 0.0185, "step": 25340 }, { "grad_norm": 0.29720309376716614, "learning_rate": 8.909996735306996e-05, "loss": 0.0183, "step": 25350 }, { "grad_norm": 0.1794201135635376, "learning_rate": 8.908965948105268e-05, "loss": 0.0178, "step": 25360 }, { "grad_norm": 0.20975840091705322, "learning_rate": 8.907934733424901e-05, "loss": 0.0177, "step": 25370 }, { "grad_norm": 0.18894782662391663, "learning_rate": 8.906903091378666e-05, "loss": 0.0168, "step": 25380 }, { "grad_norm": 0.21298091113567352, "learning_rate": 8.905871022079384e-05, "loss": 0.0177, "step": 25390 }, { "grad_norm": 0.17764656245708466, "learning_rate": 8.90483852563992e-05, "loss": 0.0166, "step": 25400 }, { "grad_norm": 0.32033947110176086, "learning_rate": 8.903805602173185e-05, "loss": 0.017, "step": 25410 }, { "grad_norm": 0.1827925443649292, "learning_rate": 8.902772251792137e-05, "loss": 0.0188, "step": 25420 }, { "grad_norm": 0.1870199590921402, "learning_rate": 8.901738474609786e-05, "loss": 0.0193, "step": 25430 }, { "grad_norm": 0.215688094496727, "learning_rate": 8.900704270739179e-05, "loss": 0.0165, "step": 25440 }, { "grad_norm": 0.2085178792476654, "learning_rate": 8.89966964029342e-05, "loss": 0.0207, "step": 25450 }, { "grad_norm": 0.19147172570228577, "learning_rate": 8.898634583385652e-05, "loss": 0.017, "step": 25460 }, { "grad_norm": 0.1701255738735199, "learning_rate": 8.897599100129065e-05, "loss": 0.0191, "step": 25470 }, { "grad_norm": 0.22569678723812103, "learning_rate": 8.896563190636903e-05, "loss": 0.0224, "step": 25480 }, { "grad_norm": 0.18371888995170593, "learning_rate": 8.895526855022448e-05, "loss": 0.0193, "step": 25490 }, { "grad_norm": 0.2588895857334137, "learning_rate": 8.894490093399033e-05, "loss": 0.0208, "step": 25500 }, { "grad_norm": 0.24677877128124237, "learning_rate": 8.893452905880035e-05, "loss": 0.0175, "step": 25510 }, { "grad_norm": 0.20772334933280945, "learning_rate": 8.892415292578883e-05, "loss": 0.0189, "step": 25520 }, { "grad_norm": 0.17819510400295258, "learning_rate": 8.891377253609046e-05, "loss": 0.0189, "step": 25530 }, { "grad_norm": 0.22641977667808533, "learning_rate": 8.890338789084043e-05, "loss": 0.0185, "step": 25540 }, { "grad_norm": 0.504706084728241, "learning_rate": 8.88929989911744e-05, "loss": 0.0211, "step": 25550 }, { "grad_norm": 0.17011114954948425, "learning_rate": 8.888260583822847e-05, "loss": 0.0203, "step": 25560 }, { "grad_norm": 0.2784031927585602, "learning_rate": 8.887220843313921e-05, "loss": 0.0198, "step": 25570 }, { "grad_norm": 0.20726557075977325, "learning_rate": 8.88618067770437e-05, "loss": 0.0214, "step": 25580 }, { "grad_norm": 0.16358257830142975, "learning_rate": 8.885140087107942e-05, "loss": 0.0211, "step": 25590 }, { "grad_norm": 0.3335874676704407, "learning_rate": 8.884099071638436e-05, "loss": 0.0188, "step": 25600 }, { "grad_norm": 0.2271752655506134, "learning_rate": 8.883057631409695e-05, "loss": 0.0182, "step": 25610 }, { "grad_norm": 0.19745980203151703, "learning_rate": 8.882015766535608e-05, "loss": 0.0186, "step": 25620 }, { "grad_norm": 0.20270800590515137, "learning_rate": 8.880973477130115e-05, "loss": 0.0172, "step": 25630 }, { "grad_norm": 0.23595012724399567, "learning_rate": 8.879930763307197e-05, "loss": 0.0159, "step": 25640 }, { "grad_norm": 0.22204601764678955, "learning_rate": 8.878887625180884e-05, "loss": 0.0162, "step": 25650 }, { "grad_norm": 0.22899092733860016, "learning_rate": 8.877844062865253e-05, "loss": 0.0163, "step": 25660 }, { "grad_norm": 0.17441195249557495, "learning_rate": 8.876800076474424e-05, "loss": 0.0176, "step": 25670 }, { "grad_norm": 0.18282228708267212, "learning_rate": 8.875755666122568e-05, "loss": 0.0175, "step": 25680 }, { "grad_norm": 0.1844286471605301, "learning_rate": 8.8747108319239e-05, "loss": 0.0173, "step": 25690 }, { "grad_norm": 0.19414448738098145, "learning_rate": 8.87366557399268e-05, "loss": 0.0177, "step": 25700 }, { "grad_norm": 0.1716853827238083, "learning_rate": 8.872619892443217e-05, "loss": 0.0188, "step": 25710 }, { "grad_norm": 0.20715369284152985, "learning_rate": 8.871573787389865e-05, "loss": 0.016, "step": 25720 }, { "grad_norm": 0.16496944427490234, "learning_rate": 8.870527258947024e-05, "loss": 0.0173, "step": 25730 }, { "grad_norm": 0.20398032665252686, "learning_rate": 8.869480307229143e-05, "loss": 0.0202, "step": 25740 }, { "grad_norm": 0.20146672427654266, "learning_rate": 8.868432932350712e-05, "loss": 0.0179, "step": 25750 }, { "grad_norm": 0.19679518043994904, "learning_rate": 8.867385134426272e-05, "loss": 0.0179, "step": 25760 }, { "grad_norm": 0.191228449344635, "learning_rate": 8.866336913570407e-05, "loss": 0.0195, "step": 25770 }, { "grad_norm": 0.23341263830661774, "learning_rate": 8.865288269897751e-05, "loss": 0.0179, "step": 25780 }, { "grad_norm": 0.19826431572437286, "learning_rate": 8.864239203522981e-05, "loss": 0.0205, "step": 25790 }, { "grad_norm": 0.19821473956108093, "learning_rate": 8.863189714560822e-05, "loss": 0.0166, "step": 25800 }, { "grad_norm": 0.1793479472398758, "learning_rate": 8.862139803126043e-05, "loss": 0.0167, "step": 25810 }, { "grad_norm": 0.18333327770233154, "learning_rate": 8.861089469333463e-05, "loss": 0.0168, "step": 25820 }, { "grad_norm": 0.1792839616537094, "learning_rate": 8.860038713297944e-05, "loss": 0.0152, "step": 25830 }, { "grad_norm": 0.3224908113479614, "learning_rate": 8.858987535134394e-05, "loss": 0.0176, "step": 25840 }, { "grad_norm": 0.2550164759159088, "learning_rate": 8.857935934957769e-05, "loss": 0.0165, "step": 25850 }, { "grad_norm": 0.2925122082233429, "learning_rate": 8.856883912883071e-05, "loss": 0.0162, "step": 25860 }, { "grad_norm": 0.18689432740211487, "learning_rate": 8.855831469025346e-05, "loss": 0.0178, "step": 25870 }, { "grad_norm": 0.20873266458511353, "learning_rate": 8.854778603499689e-05, "loss": 0.0169, "step": 25880 }, { "grad_norm": 0.18966591358184814, "learning_rate": 8.85372531642124e-05, "loss": 0.0176, "step": 25890 }, { "grad_norm": 0.20600706338882446, "learning_rate": 8.852671607905185e-05, "loss": 0.0171, "step": 25900 }, { "grad_norm": 0.2094002217054367, "learning_rate": 8.851617478066754e-05, "loss": 0.0159, "step": 25910 }, { "grad_norm": 0.2146744430065155, "learning_rate": 8.850562927021227e-05, "loss": 0.0151, "step": 25920 }, { "grad_norm": 0.22807098925113678, "learning_rate": 8.849507954883928e-05, "loss": 0.0202, "step": 25930 }, { "grad_norm": 0.2896987497806549, "learning_rate": 8.848452561770226e-05, "loss": 0.02, "step": 25940 }, { "grad_norm": 0.19469353556632996, "learning_rate": 8.847396747795538e-05, "loss": 0.0187, "step": 25950 }, { "grad_norm": 0.21910469233989716, "learning_rate": 8.846340513075327e-05, "loss": 0.0181, "step": 25960 }, { "grad_norm": 0.2025061994791031, "learning_rate": 8.845283857725099e-05, "loss": 0.0177, "step": 25970 }, { "grad_norm": 0.184409037232399, "learning_rate": 8.844226781860409e-05, "loss": 0.0195, "step": 25980 }, { "grad_norm": 0.20737065374851227, "learning_rate": 8.84316928559686e-05, "loss": 0.0196, "step": 25990 }, { "grad_norm": 0.2065090388059616, "learning_rate": 8.842111369050094e-05, "loss": 0.0164, "step": 26000 }, { "grad_norm": 0.21476688981056213, "learning_rate": 8.841053032335808e-05, "loss": 0.0186, "step": 26010 }, { "grad_norm": 0.2767253518104553, "learning_rate": 8.839994275569735e-05, "loss": 0.0171, "step": 26020 }, { "grad_norm": 0.22054217755794525, "learning_rate": 8.838935098867662e-05, "loss": 0.0164, "step": 26030 }, { "grad_norm": 0.2465364634990692, "learning_rate": 8.837875502345418e-05, "loss": 0.0208, "step": 26040 }, { "grad_norm": 0.21527646481990814, "learning_rate": 8.83681548611888e-05, "loss": 0.0154, "step": 26050 }, { "grad_norm": 0.20852670073509216, "learning_rate": 8.835755050303969e-05, "loss": 0.0208, "step": 26060 }, { "grad_norm": 0.2162906974554062, "learning_rate": 8.834694195016653e-05, "loss": 0.0174, "step": 26070 }, { "grad_norm": 0.18421775102615356, "learning_rate": 8.833632920372942e-05, "loss": 0.0177, "step": 26080 }, { "grad_norm": 0.18778517842292786, "learning_rate": 8.832571226488903e-05, "loss": 0.0158, "step": 26090 }, { "grad_norm": 0.27141058444976807, "learning_rate": 8.831509113480634e-05, "loss": 0.016, "step": 26100 }, { "grad_norm": 0.2707579731941223, "learning_rate": 8.83044658146429e-05, "loss": 0.0182, "step": 26110 }, { "grad_norm": 0.2018037736415863, "learning_rate": 8.829383630556067e-05, "loss": 0.018, "step": 26120 }, { "grad_norm": 0.23272058367729187, "learning_rate": 8.828320260872207e-05, "loss": 0.0157, "step": 26130 }, { "grad_norm": 0.23791466653347015, "learning_rate": 8.827256472529e-05, "loss": 0.0176, "step": 26140 }, { "grad_norm": 0.17612114548683167, "learning_rate": 8.826192265642778e-05, "loss": 0.0175, "step": 26150 }, { "grad_norm": 0.1480785310268402, "learning_rate": 8.825127640329923e-05, "loss": 0.0163, "step": 26160 }, { "grad_norm": 0.18027931451797485, "learning_rate": 8.824062596706861e-05, "loss": 0.0171, "step": 26170 }, { "grad_norm": 0.17052271962165833, "learning_rate": 8.822997134890062e-05, "loss": 0.0158, "step": 26180 }, { "grad_norm": 0.1812104731798172, "learning_rate": 8.821931254996044e-05, "loss": 0.0157, "step": 26190 }, { "grad_norm": 0.1775035858154297, "learning_rate": 8.82086495714137e-05, "loss": 0.0184, "step": 26200 }, { "grad_norm": 0.21803507208824158, "learning_rate": 8.81979824144265e-05, "loss": 0.0165, "step": 26210 }, { "grad_norm": 0.2045460045337677, "learning_rate": 8.818731108016536e-05, "loss": 0.0176, "step": 26220 }, { "grad_norm": 0.5187933444976807, "learning_rate": 8.81766355697973e-05, "loss": 0.0182, "step": 26230 }, { "grad_norm": 0.23194386065006256, "learning_rate": 8.816595588448977e-05, "loss": 0.017, "step": 26240 }, { "grad_norm": 0.18783196806907654, "learning_rate": 8.81552720254107e-05, "loss": 0.0149, "step": 26250 }, { "grad_norm": 0.17136061191558838, "learning_rate": 8.814458399372842e-05, "loss": 0.0207, "step": 26260 }, { "grad_norm": 0.5479950308799744, "learning_rate": 8.813389179061181e-05, "loss": 0.0173, "step": 26270 }, { "grad_norm": 0.20570117235183716, "learning_rate": 8.812319541723012e-05, "loss": 0.018, "step": 26280 }, { "grad_norm": 0.18767857551574707, "learning_rate": 8.811249487475309e-05, "loss": 0.018, "step": 26290 }, { "grad_norm": 0.2173362523317337, "learning_rate": 8.810179016435092e-05, "loss": 0.0166, "step": 26300 }, { "grad_norm": 0.189387708902359, "learning_rate": 8.809108128719428e-05, "loss": 0.0201, "step": 26310 }, { "grad_norm": 0.3224426805973053, "learning_rate": 8.808036824445424e-05, "loss": 0.0167, "step": 26320 }, { "grad_norm": 0.15286387503147125, "learning_rate": 8.806965103730238e-05, "loss": 0.0185, "step": 26330 }, { "grad_norm": 0.18874453008174896, "learning_rate": 8.805892966691074e-05, "loss": 0.021, "step": 26340 }, { "grad_norm": 0.20088891685009003, "learning_rate": 8.804820413445175e-05, "loss": 0.0187, "step": 26350 }, { "grad_norm": 0.18177199363708496, "learning_rate": 8.803747444109837e-05, "loss": 0.015, "step": 26360 }, { "grad_norm": 0.1742883324623108, "learning_rate": 8.802674058802399e-05, "loss": 0.0182, "step": 26370 }, { "grad_norm": 0.20218612253665924, "learning_rate": 8.801600257640241e-05, "loss": 0.0183, "step": 26380 }, { "grad_norm": 0.21146206557750702, "learning_rate": 8.800526040740795e-05, "loss": 0.0185, "step": 26390 }, { "grad_norm": 0.20694391429424286, "learning_rate": 8.799451408221535e-05, "loss": 0.0192, "step": 26400 }, { "grad_norm": 0.17555727064609528, "learning_rate": 8.798376360199982e-05, "loss": 0.0172, "step": 26410 }, { "grad_norm": 0.3518270254135132, "learning_rate": 8.797300896793701e-05, "loss": 0.0169, "step": 26420 }, { "grad_norm": 0.24620138108730316, "learning_rate": 8.796225018120302e-05, "loss": 0.0174, "step": 26430 }, { "grad_norm": 0.42319023609161377, "learning_rate": 8.795148724297444e-05, "loss": 0.0213, "step": 26440 }, { "grad_norm": 0.353120356798172, "learning_rate": 8.794072015442825e-05, "loss": 0.0181, "step": 26450 }, { "grad_norm": 0.3883807063102722, "learning_rate": 8.792994891674198e-05, "loss": 0.0174, "step": 26460 }, { "grad_norm": 0.21286509931087494, "learning_rate": 8.79191735310935e-05, "loss": 0.0195, "step": 26470 }, { "grad_norm": 0.2983117699623108, "learning_rate": 8.790839399866122e-05, "loss": 0.0183, "step": 26480 }, { "grad_norm": 0.20188021659851074, "learning_rate": 8.789761032062397e-05, "loss": 0.0172, "step": 26490 }, { "grad_norm": 0.24523133039474487, "learning_rate": 8.788682249816103e-05, "loss": 0.0187, "step": 26500 }, { "grad_norm": 0.14022843539714813, "learning_rate": 8.787603053245215e-05, "loss": 0.0186, "step": 26510 }, { "grad_norm": 0.2701816260814667, "learning_rate": 8.78652344246775e-05, "loss": 0.0152, "step": 26520 }, { "grad_norm": 0.20707707107067108, "learning_rate": 8.785443417601776e-05, "loss": 0.0193, "step": 26530 }, { "grad_norm": 0.13264930248260498, "learning_rate": 8.784362978765401e-05, "loss": 0.0159, "step": 26540 }, { "grad_norm": 0.34316205978393555, "learning_rate": 8.783282126076779e-05, "loss": 0.0181, "step": 26550 }, { "grad_norm": 0.21674616634845734, "learning_rate": 8.782200859654112e-05, "loss": 0.0162, "step": 26560 }, { "grad_norm": 0.17729122936725616, "learning_rate": 8.781119179615646e-05, "loss": 0.0192, "step": 26570 }, { "grad_norm": 0.2258809208869934, "learning_rate": 8.780037086079674e-05, "loss": 0.019, "step": 26580 }, { "grad_norm": 0.5347390174865723, "learning_rate": 8.778954579164527e-05, "loss": 0.0165, "step": 26590 }, { "grad_norm": 0.20401595532894135, "learning_rate": 8.777871658988588e-05, "loss": 0.0188, "step": 26600 }, { "grad_norm": 0.2452555149793625, "learning_rate": 8.776788325670285e-05, "loss": 0.0186, "step": 26610 }, { "grad_norm": 0.21545006334781647, "learning_rate": 8.775704579328089e-05, "loss": 0.0161, "step": 26620 }, { "grad_norm": 0.22286607325077057, "learning_rate": 8.774620420080517e-05, "loss": 0.0182, "step": 26630 }, { "grad_norm": 0.19988439977169037, "learning_rate": 8.773535848046131e-05, "loss": 0.0183, "step": 26640 }, { "grad_norm": 0.28874602913856506, "learning_rate": 8.772450863343538e-05, "loss": 0.0185, "step": 26650 }, { "grad_norm": 0.22713996469974518, "learning_rate": 8.77136546609139e-05, "loss": 0.0167, "step": 26660 }, { "grad_norm": 0.15403448045253754, "learning_rate": 8.770279656408385e-05, "loss": 0.0176, "step": 26670 }, { "grad_norm": 0.2562839984893799, "learning_rate": 8.769193434413265e-05, "loss": 0.0173, "step": 26680 }, { "grad_norm": 0.20721808075904846, "learning_rate": 8.76810680022482e-05, "loss": 0.0161, "step": 26690 }, { "grad_norm": 0.1788371354341507, "learning_rate": 8.767019753961878e-05, "loss": 0.0167, "step": 26700 }, { "grad_norm": 0.1719205230474472, "learning_rate": 8.765932295743321e-05, "loss": 0.0167, "step": 26710 }, { "grad_norm": 0.3953075408935547, "learning_rate": 8.764844425688068e-05, "loss": 0.0167, "step": 26720 }, { "grad_norm": 0.3256988525390625, "learning_rate": 8.763756143915092e-05, "loss": 0.0205, "step": 26730 }, { "grad_norm": 0.14542755484580994, "learning_rate": 8.7626674505434e-05, "loss": 0.0156, "step": 26740 }, { "grad_norm": 0.173537939786911, "learning_rate": 8.761578345692053e-05, "loss": 0.0156, "step": 26750 }, { "grad_norm": 0.20995813608169556, "learning_rate": 8.760488829480156e-05, "loss": 0.0167, "step": 26760 }, { "grad_norm": 0.18681612610816956, "learning_rate": 8.759398902026854e-05, "loss": 0.0184, "step": 26770 }, { "grad_norm": 0.14137858152389526, "learning_rate": 8.758308563451339e-05, "loss": 0.0148, "step": 26780 }, { "grad_norm": 0.2253120392560959, "learning_rate": 8.75721781387285e-05, "loss": 0.0189, "step": 26790 }, { "grad_norm": 0.16712279617786407, "learning_rate": 8.75612665341067e-05, "loss": 0.0177, "step": 26800 }, { "grad_norm": 0.20886540412902832, "learning_rate": 8.755035082184126e-05, "loss": 0.0163, "step": 26810 }, { "grad_norm": 0.20373818278312683, "learning_rate": 8.753943100312592e-05, "loss": 0.0155, "step": 26820 }, { "grad_norm": 0.18039584159851074, "learning_rate": 8.752850707915484e-05, "loss": 0.0186, "step": 26830 }, { "grad_norm": 0.17356042563915253, "learning_rate": 8.751757905112264e-05, "loss": 0.0152, "step": 26840 }, { "grad_norm": 0.22939333319664001, "learning_rate": 8.75066469202244e-05, "loss": 0.0176, "step": 26850 }, { "grad_norm": 0.15411920845508575, "learning_rate": 8.749571068765567e-05, "loss": 0.0184, "step": 26860 }, { "grad_norm": 0.22624240815639496, "learning_rate": 8.748477035461238e-05, "loss": 0.0173, "step": 26870 }, { "grad_norm": 0.17247864603996277, "learning_rate": 8.747382592229095e-05, "loss": 0.0156, "step": 26880 }, { "grad_norm": 0.1767461597919464, "learning_rate": 8.746287739188828e-05, "loss": 0.0159, "step": 26890 }, { "grad_norm": 0.19323907792568207, "learning_rate": 8.745192476460165e-05, "loss": 0.0156, "step": 26900 }, { "grad_norm": 0.20555132627487183, "learning_rate": 8.744096804162882e-05, "loss": 0.0165, "step": 26910 }, { "grad_norm": 0.23756933212280273, "learning_rate": 8.743000722416804e-05, "loss": 0.0151, "step": 26920 }, { "grad_norm": 0.24802853167057037, "learning_rate": 8.741904231341793e-05, "loss": 0.0173, "step": 26930 }, { "grad_norm": 0.28591492772102356, "learning_rate": 8.740807331057762e-05, "loss": 0.0157, "step": 26940 }, { "grad_norm": 0.258820503950119, "learning_rate": 8.739710021684667e-05, "loss": 0.0158, "step": 26950 }, { "grad_norm": 0.1988866627216339, "learning_rate": 8.738612303342503e-05, "loss": 0.0164, "step": 26960 }, { "grad_norm": 0.2129291445016861, "learning_rate": 8.73751417615132e-05, "loss": 0.0155, "step": 26970 }, { "grad_norm": 0.1842135637998581, "learning_rate": 8.736415640231208e-05, "loss": 0.018, "step": 26980 }, { "grad_norm": 0.26267069578170776, "learning_rate": 8.735316695702297e-05, "loss": 0.017, "step": 26990 }, { "grad_norm": 0.1909966915845871, "learning_rate": 8.734217342684769e-05, "loss": 0.0177, "step": 27000 }, { "grad_norm": 0.1957731992006302, "learning_rate": 8.733117581298847e-05, "loss": 0.0154, "step": 27010 }, { "grad_norm": 0.19738224148750305, "learning_rate": 8.732017411664796e-05, "loss": 0.0172, "step": 27020 }, { "grad_norm": 0.18437376618385315, "learning_rate": 8.730916833902936e-05, "loss": 0.0185, "step": 27030 }, { "grad_norm": 0.22922475636005402, "learning_rate": 8.729815848133618e-05, "loss": 0.0182, "step": 27040 }, { "grad_norm": 0.20981404185295105, "learning_rate": 8.728714454477247e-05, "loss": 0.0152, "step": 27050 }, { "grad_norm": 0.22498886287212372, "learning_rate": 8.727612653054269e-05, "loss": 0.0156, "step": 27060 }, { "grad_norm": 0.16370633244514465, "learning_rate": 8.726510443985176e-05, "loss": 0.0182, "step": 27070 }, { "grad_norm": 0.23753812909126282, "learning_rate": 8.725407827390503e-05, "loss": 0.0185, "step": 27080 }, { "grad_norm": 0.19863517582416534, "learning_rate": 8.724304803390833e-05, "loss": 0.0187, "step": 27090 }, { "grad_norm": 0.18510892987251282, "learning_rate": 8.723201372106788e-05, "loss": 0.018, "step": 27100 }, { "grad_norm": 0.2107222080230713, "learning_rate": 8.722097533659038e-05, "loss": 0.0168, "step": 27110 }, { "grad_norm": 0.13347725570201874, "learning_rate": 8.720993288168299e-05, "loss": 0.016, "step": 27120 }, { "grad_norm": 0.19508540630340576, "learning_rate": 8.719888635755327e-05, "loss": 0.0192, "step": 27130 }, { "grad_norm": 0.21517819166183472, "learning_rate": 8.718783576540928e-05, "loss": 0.0157, "step": 27140 }, { "grad_norm": 0.22111055254936218, "learning_rate": 8.717678110645948e-05, "loss": 0.0169, "step": 27150 }, { "grad_norm": 0.22402255237102509, "learning_rate": 8.716572238191279e-05, "loss": 0.0152, "step": 27160 }, { "grad_norm": 0.2967386543750763, "learning_rate": 8.715465959297857e-05, "loss": 0.017, "step": 27170 }, { "grad_norm": 0.23773421347141266, "learning_rate": 8.714359274086665e-05, "loss": 0.0197, "step": 27180 }, { "grad_norm": 0.22466208040714264, "learning_rate": 8.713252182678726e-05, "loss": 0.0171, "step": 27190 }, { "grad_norm": 0.3848622441291809, "learning_rate": 8.712144685195112e-05, "loss": 0.0178, "step": 27200 }, { "grad_norm": 0.2314654141664505, "learning_rate": 8.711036781756936e-05, "loss": 0.0192, "step": 27210 }, { "grad_norm": 0.1813499629497528, "learning_rate": 8.709928472485357e-05, "loss": 0.0171, "step": 27220 }, { "grad_norm": 0.26362013816833496, "learning_rate": 8.708819757501579e-05, "loss": 0.0168, "step": 27230 }, { "grad_norm": 0.15221384167671204, "learning_rate": 8.707710636926846e-05, "loss": 0.0191, "step": 27240 }, { "grad_norm": 0.206645667552948, "learning_rate": 8.706601110882455e-05, "loss": 0.0201, "step": 27250 }, { "grad_norm": 0.19994573295116425, "learning_rate": 8.705491179489738e-05, "loss": 0.0172, "step": 27260 }, { "grad_norm": 0.1991277039051056, "learning_rate": 8.704380842870077e-05, "loss": 0.0162, "step": 27270 }, { "grad_norm": 0.16969700157642365, "learning_rate": 8.703270101144895e-05, "loss": 0.0187, "step": 27280 }, { "grad_norm": 0.17218177020549774, "learning_rate": 8.702158954435664e-05, "loss": 0.0173, "step": 27290 }, { "grad_norm": 0.15894053876399994, "learning_rate": 8.701047402863896e-05, "loss": 0.0184, "step": 27300 }, { "grad_norm": 0.2828819155693054, "learning_rate": 8.699935446551148e-05, "loss": 0.0171, "step": 27310 }, { "grad_norm": 0.2123371958732605, "learning_rate": 8.698823085619022e-05, "loss": 0.0171, "step": 27320 }, { "grad_norm": 0.30488333106040955, "learning_rate": 8.697710320189166e-05, "loss": 0.018, "step": 27330 }, { "grad_norm": 0.355317622423172, "learning_rate": 8.696597150383268e-05, "loss": 0.0168, "step": 27340 }, { "grad_norm": 0.22528043389320374, "learning_rate": 8.695483576323063e-05, "loss": 0.0176, "step": 27350 }, { "grad_norm": 0.15558096766471863, "learning_rate": 8.69436959813033e-05, "loss": 0.0161, "step": 27360 }, { "grad_norm": 0.2588522732257843, "learning_rate": 8.693255215926892e-05, "loss": 0.0156, "step": 27370 }, { "grad_norm": 0.22847217321395874, "learning_rate": 8.692140429834617e-05, "loss": 0.0166, "step": 27380 }, { "grad_norm": 0.14392104744911194, "learning_rate": 8.691025239975415e-05, "loss": 0.0154, "step": 27390 }, { "grad_norm": 0.19466789066791534, "learning_rate": 8.689909646471243e-05, "loss": 0.0167, "step": 27400 }, { "grad_norm": 0.3656151294708252, "learning_rate": 8.688793649444099e-05, "loss": 0.0171, "step": 27410 }, { "grad_norm": 0.15420453250408173, "learning_rate": 8.687677249016029e-05, "loss": 0.0183, "step": 27420 }, { "grad_norm": 0.22900019586086273, "learning_rate": 8.686560445309118e-05, "loss": 0.0178, "step": 27430 }, { "grad_norm": 0.17388245463371277, "learning_rate": 8.685443238445499e-05, "loss": 0.018, "step": 27440 }, { "grad_norm": 0.2321457713842392, "learning_rate": 8.68432562854735e-05, "loss": 0.0196, "step": 27450 }, { "grad_norm": 0.2382582575082779, "learning_rate": 8.683207615736887e-05, "loss": 0.0189, "step": 27460 }, { "grad_norm": 0.18440242111682892, "learning_rate": 8.682089200136379e-05, "loss": 0.0167, "step": 27470 }, { "grad_norm": 0.23330137133598328, "learning_rate": 8.680970381868132e-05, "loss": 0.0176, "step": 27480 }, { "grad_norm": 0.17272786796092987, "learning_rate": 8.679851161054498e-05, "loss": 0.0172, "step": 27490 }, { "grad_norm": 0.19381029903888702, "learning_rate": 8.678731537817873e-05, "loss": 0.0162, "step": 27500 }, { "grad_norm": 0.1666240096092224, "learning_rate": 8.677611512280697e-05, "loss": 0.017, "step": 27510 }, { "grad_norm": 0.18958590924739838, "learning_rate": 8.676491084565457e-05, "loss": 0.0203, "step": 27520 }, { "grad_norm": 0.2657327353954315, "learning_rate": 8.675370254794678e-05, "loss": 0.0169, "step": 27530 }, { "grad_norm": 0.2059464305639267, "learning_rate": 8.674249023090935e-05, "loss": 0.0223, "step": 27540 }, { "grad_norm": 0.18777060508728027, "learning_rate": 8.673127389576843e-05, "loss": 0.0164, "step": 27550 }, { "grad_norm": 0.256387859582901, "learning_rate": 8.67200535437506e-05, "loss": 0.0189, "step": 27560 }, { "grad_norm": 0.2416752576828003, "learning_rate": 8.670882917608296e-05, "loss": 0.019, "step": 27570 }, { "grad_norm": 0.16827568411827087, "learning_rate": 8.669760079399292e-05, "loss": 0.016, "step": 27580 }, { "grad_norm": 0.1836307942867279, "learning_rate": 8.668636839870845e-05, "loss": 0.0165, "step": 27590 }, { "grad_norm": 0.24412785470485687, "learning_rate": 8.667513199145789e-05, "loss": 0.018, "step": 27600 }, { "grad_norm": 0.15604639053344727, "learning_rate": 8.666389157347002e-05, "loss": 0.0157, "step": 27610 }, { "grad_norm": 0.21197177469730377, "learning_rate": 8.66526471459741e-05, "loss": 0.0194, "step": 27620 }, { "grad_norm": 0.16269411146640778, "learning_rate": 8.66413987101998e-05, "loss": 0.0172, "step": 27630 }, { "grad_norm": 0.18772535026073456, "learning_rate": 8.663014626737723e-05, "loss": 0.0187, "step": 27640 }, { "grad_norm": 0.1889438033103943, "learning_rate": 8.661888981873691e-05, "loss": 0.0172, "step": 27650 }, { "grad_norm": 0.1696411669254303, "learning_rate": 8.660762936550988e-05, "loss": 0.0183, "step": 27660 }, { "grad_norm": 0.18297140300273895, "learning_rate": 8.659636490892753e-05, "loss": 0.0163, "step": 27670 }, { "grad_norm": 0.22823452949523926, "learning_rate": 8.658509645022174e-05, "loss": 0.0171, "step": 27680 }, { "grad_norm": 0.2949545383453369, "learning_rate": 8.657382399062481e-05, "loss": 0.0183, "step": 27690 }, { "grad_norm": 0.3037419021129608, "learning_rate": 8.656254753136946e-05, "loss": 0.015, "step": 27700 }, { "grad_norm": 0.19053290784358978, "learning_rate": 8.655126707368891e-05, "loss": 0.0186, "step": 27710 }, { "grad_norm": 0.22088022530078888, "learning_rate": 8.653998261881672e-05, "loss": 0.0218, "step": 27720 }, { "grad_norm": 0.24111391603946686, "learning_rate": 8.652869416798699e-05, "loss": 0.0186, "step": 27730 }, { "grad_norm": 0.18667475879192352, "learning_rate": 8.651740172243417e-05, "loss": 0.0175, "step": 27740 }, { "grad_norm": 0.20266394317150116, "learning_rate": 8.65061052833932e-05, "loss": 0.0194, "step": 27750 }, { "grad_norm": 0.2011716663837433, "learning_rate": 8.649480485209945e-05, "loss": 0.0182, "step": 27760 }, { "grad_norm": 0.16770105063915253, "learning_rate": 8.64835004297887e-05, "loss": 0.0154, "step": 27770 }, { "grad_norm": 0.19219553470611572, "learning_rate": 8.64721920176972e-05, "loss": 0.0186, "step": 27780 }, { "grad_norm": 0.26552730798721313, "learning_rate": 8.646087961706164e-05, "loss": 0.0179, "step": 27790 }, { "grad_norm": 0.21404053270816803, "learning_rate": 8.644956322911908e-05, "loss": 0.0192, "step": 27800 }, { "grad_norm": 0.21910087764263153, "learning_rate": 8.643824285510709e-05, "loss": 0.0163, "step": 27810 }, { "grad_norm": 0.1941283494234085, "learning_rate": 8.642691849626364e-05, "loss": 0.0184, "step": 27820 }, { "grad_norm": 0.23236405849456787, "learning_rate": 8.641559015382717e-05, "loss": 0.0178, "step": 27830 }, { "grad_norm": 0.19437898695468903, "learning_rate": 8.640425782903649e-05, "loss": 0.0133, "step": 27840 }, { "grad_norm": 0.13379091024398804, "learning_rate": 8.639292152313091e-05, "loss": 0.0151, "step": 27850 }, { "grad_norm": 0.2585764527320862, "learning_rate": 8.638158123735015e-05, "loss": 0.0138, "step": 27860 }, { "grad_norm": 0.1865893304347992, "learning_rate": 8.637023697293436e-05, "loss": 0.016, "step": 27870 }, { "grad_norm": 0.16152633726596832, "learning_rate": 8.635888873112414e-05, "loss": 0.0182, "step": 27880 }, { "grad_norm": 0.1897214949131012, "learning_rate": 8.634753651316052e-05, "loss": 0.0167, "step": 27890 }, { "grad_norm": 0.22263593971729279, "learning_rate": 8.633618032028496e-05, "loss": 0.0172, "step": 27900 }, { "grad_norm": 0.19260290265083313, "learning_rate": 8.632482015373934e-05, "loss": 0.019, "step": 27910 }, { "grad_norm": 0.17490823566913605, "learning_rate": 8.6313456014766e-05, "loss": 0.0186, "step": 27920 }, { "grad_norm": 0.19046053290367126, "learning_rate": 8.630208790460771e-05, "loss": 0.0143, "step": 27930 }, { "grad_norm": 0.23149938881397247, "learning_rate": 8.629071582450768e-05, "loss": 0.0163, "step": 27940 }, { "grad_norm": 0.179966002702713, "learning_rate": 8.62793397757095e-05, "loss": 0.0146, "step": 27950 }, { "grad_norm": 0.24225777387619019, "learning_rate": 8.626795975945729e-05, "loss": 0.0181, "step": 27960 }, { "grad_norm": 0.1765284240245819, "learning_rate": 8.625657577699551e-05, "loss": 0.0145, "step": 27970 }, { "grad_norm": 0.16061538457870483, "learning_rate": 8.624518782956914e-05, "loss": 0.0152, "step": 27980 }, { "grad_norm": 0.2178659588098526, "learning_rate": 8.62337959184235e-05, "loss": 0.0187, "step": 27990 }, { "grad_norm": 0.18577995896339417, "learning_rate": 8.622240004480441e-05, "loss": 0.0165, "step": 28000 }, { "grad_norm": 0.19300240278244019, "learning_rate": 8.621100020995814e-05, "loss": 0.0165, "step": 28010 }, { "grad_norm": 0.2133711725473404, "learning_rate": 8.619959641513132e-05, "loss": 0.0174, "step": 28020 }, { "grad_norm": 0.15558280050754547, "learning_rate": 8.618818866157105e-05, "loss": 0.0157, "step": 28030 }, { "grad_norm": 0.20971634984016418, "learning_rate": 8.617677695052487e-05, "loss": 0.0183, "step": 28040 }, { "grad_norm": 0.1740049421787262, "learning_rate": 8.616536128324078e-05, "loss": 0.0168, "step": 28050 }, { "grad_norm": 0.19802536070346832, "learning_rate": 8.615394166096712e-05, "loss": 0.0176, "step": 28060 }, { "grad_norm": 0.20941126346588135, "learning_rate": 8.614251808495279e-05, "loss": 0.0146, "step": 28070 }, { "grad_norm": 0.18545092642307281, "learning_rate": 8.6131090556447e-05, "loss": 0.0182, "step": 28080 }, { "grad_norm": 0.23976747691631317, "learning_rate": 8.611965907669947e-05, "loss": 0.018, "step": 28090 }, { "grad_norm": 0.1689208298921585, "learning_rate": 8.610822364696034e-05, "loss": 0.016, "step": 28100 }, { "grad_norm": 0.12217213958501816, "learning_rate": 8.609678426848015e-05, "loss": 0.0181, "step": 28110 }, { "grad_norm": 0.1869659423828125, "learning_rate": 8.60853409425099e-05, "loss": 0.016, "step": 28120 }, { "grad_norm": 0.2697962522506714, "learning_rate": 8.607389367030104e-05, "loss": 0.0169, "step": 28130 }, { "grad_norm": 0.17372383177280426, "learning_rate": 8.606244245310538e-05, "loss": 0.0188, "step": 28140 }, { "grad_norm": 0.19742979109287262, "learning_rate": 8.605098729217525e-05, "loss": 0.0182, "step": 28150 }, { "grad_norm": 0.18096791207790375, "learning_rate": 8.603952818876335e-05, "loss": 0.0161, "step": 28160 }, { "grad_norm": 0.19391094148159027, "learning_rate": 8.602806514412281e-05, "loss": 0.0157, "step": 28170 }, { "grad_norm": 0.19888603687286377, "learning_rate": 8.601659815950726e-05, "loss": 0.0169, "step": 28180 }, { "grad_norm": 0.2508901357650757, "learning_rate": 8.600512723617067e-05, "loss": 0.0155, "step": 28190 }, { "grad_norm": 0.20291279256343842, "learning_rate": 8.59936523753675e-05, "loss": 0.0147, "step": 28200 }, { "grad_norm": 0.18315620720386505, "learning_rate": 8.598217357835264e-05, "loss": 0.0158, "step": 28210 }, { "grad_norm": 0.14931941032409668, "learning_rate": 8.597069084638135e-05, "loss": 0.0148, "step": 28220 }, { "grad_norm": 0.20080512762069702, "learning_rate": 8.595920418070939e-05, "loss": 0.0147, "step": 28230 }, { "grad_norm": 0.21598879992961884, "learning_rate": 8.594771358259295e-05, "loss": 0.0183, "step": 28240 }, { "grad_norm": 0.47480422258377075, "learning_rate": 8.593621905328858e-05, "loss": 0.0218, "step": 28250 }, { "grad_norm": 0.3241274952888489, "learning_rate": 8.592472059405333e-05, "loss": 0.0233, "step": 28260 }, { "grad_norm": 0.16610610485076904, "learning_rate": 8.591321820614464e-05, "loss": 0.0209, "step": 28270 }, { "grad_norm": 0.18360863626003265, "learning_rate": 8.590171189082041e-05, "loss": 0.0168, "step": 28280 }, { "grad_norm": 0.23034389317035675, "learning_rate": 8.589020164933894e-05, "loss": 0.0159, "step": 28290 }, { "grad_norm": 0.12810666859149933, "learning_rate": 8.587868748295898e-05, "loss": 0.0159, "step": 28300 }, { "grad_norm": 0.18662354350090027, "learning_rate": 8.586716939293971e-05, "loss": 0.0173, "step": 28310 }, { "grad_norm": 0.19100861251354218, "learning_rate": 8.58556473805407e-05, "loss": 0.0155, "step": 28320 }, { "grad_norm": 0.17832130193710327, "learning_rate": 8.584412144702202e-05, "loss": 0.016, "step": 28330 }, { "grad_norm": 0.1806049942970276, "learning_rate": 8.58325915936441e-05, "loss": 0.0166, "step": 28340 }, { "grad_norm": 0.2665362060070038, "learning_rate": 8.582105782166783e-05, "loss": 0.0153, "step": 28350 }, { "grad_norm": 0.17280034720897675, "learning_rate": 8.580952013235455e-05, "loss": 0.0174, "step": 28360 }, { "grad_norm": 0.2700539231300354, "learning_rate": 8.579797852696596e-05, "loss": 0.0167, "step": 28370 }, { "grad_norm": 0.18370424211025238, "learning_rate": 8.578643300676428e-05, "loss": 0.0163, "step": 28380 }, { "grad_norm": 0.2516357898712158, "learning_rate": 8.577488357301209e-05, "loss": 0.0181, "step": 28390 }, { "grad_norm": 0.1652558147907257, "learning_rate": 8.576333022697242e-05, "loss": 0.0139, "step": 28400 }, { "grad_norm": 0.20761322975158691, "learning_rate": 8.575177296990873e-05, "loss": 0.0156, "step": 28410 }, { "grad_norm": 0.29786065220832825, "learning_rate": 8.574021180308489e-05, "loss": 0.0145, "step": 28420 }, { "grad_norm": 0.170712411403656, "learning_rate": 8.572864672776523e-05, "loss": 0.016, "step": 28430 }, { "grad_norm": 0.23306132853031158, "learning_rate": 8.571707774521447e-05, "loss": 0.0161, "step": 28440 }, { "grad_norm": 0.16349278390407562, "learning_rate": 8.57055048566978e-05, "loss": 0.0164, "step": 28450 }, { "grad_norm": 0.2112007737159729, "learning_rate": 8.569392806348078e-05, "loss": 0.0153, "step": 28460 }, { "grad_norm": 0.21944792568683624, "learning_rate": 8.568234736682947e-05, "loss": 0.0158, "step": 28470 }, { "grad_norm": 0.1698455512523651, "learning_rate": 8.567076276801029e-05, "loss": 0.0157, "step": 28480 }, { "grad_norm": 0.24528363347053528, "learning_rate": 8.565917426829013e-05, "loss": 0.0146, "step": 28490 }, { "grad_norm": 0.23803697526454926, "learning_rate": 8.564758186893628e-05, "loss": 0.0179, "step": 28500 }, { "grad_norm": 0.1943441480398178, "learning_rate": 8.563598557121649e-05, "loss": 0.0146, "step": 28510 }, { "grad_norm": 0.14670605957508087, "learning_rate": 8.562438537639888e-05, "loss": 0.0136, "step": 28520 }, { "grad_norm": 0.12962770462036133, "learning_rate": 8.561278128575206e-05, "loss": 0.0158, "step": 28530 }, { "grad_norm": 0.18960995972156525, "learning_rate": 8.5601173300545e-05, "loss": 0.0168, "step": 28540 }, { "grad_norm": 0.19605840742588043, "learning_rate": 8.558956142204717e-05, "loss": 0.0153, "step": 28550 }, { "grad_norm": 0.2135748267173767, "learning_rate": 8.55779456515284e-05, "loss": 0.015, "step": 28560 }, { "grad_norm": 0.20777557790279388, "learning_rate": 8.556632599025898e-05, "loss": 0.0172, "step": 28570 }, { "grad_norm": 0.194222092628479, "learning_rate": 8.555470243950964e-05, "loss": 0.0167, "step": 28580 }, { "grad_norm": 0.2037520408630371, "learning_rate": 8.554307500055148e-05, "loss": 0.0138, "step": 28590 }, { "grad_norm": 0.1611642986536026, "learning_rate": 8.553144367465609e-05, "loss": 0.0162, "step": 28600 }, { "grad_norm": 0.2397400438785553, "learning_rate": 8.551980846309544e-05, "loss": 0.014, "step": 28610 }, { "grad_norm": 0.16486673057079315, "learning_rate": 8.550816936714193e-05, "loss": 0.0158, "step": 28620 }, { "grad_norm": 0.2047390639781952, "learning_rate": 8.549652638806841e-05, "loss": 0.0167, "step": 28630 }, { "grad_norm": 0.21652787923812866, "learning_rate": 8.548487952714812e-05, "loss": 0.0151, "step": 28640 }, { "grad_norm": 0.13029615581035614, "learning_rate": 8.547322878565478e-05, "loss": 0.0154, "step": 28650 }, { "grad_norm": 0.22648277878761292, "learning_rate": 8.546157416486245e-05, "loss": 0.0142, "step": 28660 }, { "grad_norm": 0.18048866093158722, "learning_rate": 8.54499156660457e-05, "loss": 0.0167, "step": 28670 }, { "grad_norm": 0.21208728849887848, "learning_rate": 8.543825329047947e-05, "loss": 0.0144, "step": 28680 }, { "grad_norm": 0.18778294324874878, "learning_rate": 8.542658703943913e-05, "loss": 0.0177, "step": 28690 }, { "grad_norm": 0.17855222523212433, "learning_rate": 8.541491691420051e-05, "loss": 0.0161, "step": 28700 }, { "grad_norm": 0.17550842463970184, "learning_rate": 8.54032429160398e-05, "loss": 0.016, "step": 28710 }, { "grad_norm": 0.15490666031837463, "learning_rate": 8.539156504623369e-05, "loss": 0.0159, "step": 28720 }, { "grad_norm": 0.27327197790145874, "learning_rate": 8.537988330605923e-05, "loss": 0.0147, "step": 28730 }, { "grad_norm": 0.20147554576396942, "learning_rate": 8.536819769679393e-05, "loss": 0.0153, "step": 28740 }, { "grad_norm": 0.14771069586277008, "learning_rate": 8.53565082197157e-05, "loss": 0.0195, "step": 28750 }, { "grad_norm": 0.3115522563457489, "learning_rate": 8.534481487610289e-05, "loss": 0.0181, "step": 28760 }, { "grad_norm": 0.19639559090137482, "learning_rate": 8.533311766723428e-05, "loss": 0.017, "step": 28770 }, { "grad_norm": 0.14861233532428741, "learning_rate": 8.532141659438901e-05, "loss": 0.0162, "step": 28780 }, { "grad_norm": 0.1813897341489792, "learning_rate": 8.530971165884675e-05, "loss": 0.0151, "step": 28790 }, { "grad_norm": 0.18709081411361694, "learning_rate": 8.529800286188752e-05, "loss": 0.0153, "step": 28800 }, { "grad_norm": 0.16279856860637665, "learning_rate": 8.528629020479175e-05, "loss": 0.0178, "step": 28810 }, { "grad_norm": 0.23019632697105408, "learning_rate": 8.527457368884033e-05, "loss": 0.0183, "step": 28820 }, { "grad_norm": 0.22128920257091522, "learning_rate": 8.526285331531458e-05, "loss": 0.0173, "step": 28830 }, { "grad_norm": 0.18840385973453522, "learning_rate": 8.525112908549621e-05, "loss": 0.017, "step": 28840 }, { "grad_norm": 0.2150164693593979, "learning_rate": 8.523940100066735e-05, "loss": 0.0173, "step": 28850 }, { "grad_norm": 0.23210270702838898, "learning_rate": 8.52276690621106e-05, "loss": 0.0175, "step": 28860 }, { "grad_norm": 0.2923337519168854, "learning_rate": 8.521593327110889e-05, "loss": 0.0211, "step": 28870 }, { "grad_norm": 0.19980767369270325, "learning_rate": 8.520419362894569e-05, "loss": 0.0208, "step": 28880 }, { "grad_norm": 0.281789630651474, "learning_rate": 8.51924501369048e-05, "loss": 0.017, "step": 28890 }, { "grad_norm": 0.17081686854362488, "learning_rate": 8.518070279627047e-05, "loss": 0.0194, "step": 28900 }, { "grad_norm": 0.21862748265266418, "learning_rate": 8.516895160832737e-05, "loss": 0.0216, "step": 28910 }, { "grad_norm": 0.18205741047859192, "learning_rate": 8.515719657436061e-05, "loss": 0.0176, "step": 28920 }, { "grad_norm": 0.17509987950325012, "learning_rate": 8.514543769565568e-05, "loss": 0.0165, "step": 28930 }, { "grad_norm": 0.1931624859571457, "learning_rate": 8.513367497349853e-05, "loss": 0.0168, "step": 28940 }, { "grad_norm": 0.2106117457151413, "learning_rate": 8.51219084091755e-05, "loss": 0.0184, "step": 28950 }, { "grad_norm": 0.19298452138900757, "learning_rate": 8.511013800397338e-05, "loss": 0.0191, "step": 28960 }, { "grad_norm": 0.18704751133918762, "learning_rate": 8.509836375917937e-05, "loss": 0.0175, "step": 28970 }, { "grad_norm": 0.2027941793203354, "learning_rate": 8.508658567608104e-05, "loss": 0.0164, "step": 28980 }, { "grad_norm": 0.2116352617740631, "learning_rate": 8.507480375596647e-05, "loss": 0.016, "step": 28990 }, { "grad_norm": 0.18839190900325775, "learning_rate": 8.506301800012408e-05, "loss": 0.0172, "step": 29000 }, { "grad_norm": 0.18313510715961456, "learning_rate": 8.505122840984278e-05, "loss": 0.0157, "step": 29010 }, { "grad_norm": 0.2273177206516266, "learning_rate": 8.503943498641182e-05, "loss": 0.0163, "step": 29020 }, { "grad_norm": 0.20208775997161865, "learning_rate": 8.502763773112095e-05, "loss": 0.0159, "step": 29030 }, { "grad_norm": 0.22056028246879578, "learning_rate": 8.501583664526026e-05, "loss": 0.0163, "step": 29040 }, { "grad_norm": 0.23071467876434326, "learning_rate": 8.500403173012032e-05, "loss": 0.0179, "step": 29050 }, { "grad_norm": 0.3072208762168884, "learning_rate": 8.499222298699211e-05, "loss": 0.0188, "step": 29060 }, { "grad_norm": 0.17611803114414215, "learning_rate": 8.498041041716701e-05, "loss": 0.0165, "step": 29070 }, { "grad_norm": 0.1524905264377594, "learning_rate": 8.496859402193681e-05, "loss": 0.0161, "step": 29080 }, { "grad_norm": 0.19298158586025238, "learning_rate": 8.495677380259374e-05, "loss": 0.0147, "step": 29090 }, { "grad_norm": 0.29846519231796265, "learning_rate": 8.494494976043045e-05, "loss": 0.0162, "step": 29100 }, { "grad_norm": 0.2442857325077057, "learning_rate": 8.493312189673998e-05, "loss": 0.0165, "step": 29110 }, { "grad_norm": 0.2049129456281662, "learning_rate": 8.492129021281584e-05, "loss": 0.0173, "step": 29120 }, { "grad_norm": 0.20971287786960602, "learning_rate": 8.490945470995188e-05, "loss": 0.0163, "step": 29130 }, { "grad_norm": 0.32431066036224365, "learning_rate": 8.489761538944247e-05, "loss": 0.0171, "step": 29140 }, { "grad_norm": 0.17046748101711273, "learning_rate": 8.48857722525823e-05, "loss": 0.0161, "step": 29150 }, { "grad_norm": 0.19394712150096893, "learning_rate": 8.487392530066652e-05, "loss": 0.017, "step": 29160 }, { "grad_norm": 0.17718535661697388, "learning_rate": 8.486207453499069e-05, "loss": 0.0167, "step": 29170 }, { "grad_norm": 0.16919279098510742, "learning_rate": 8.485021995685082e-05, "loss": 0.0184, "step": 29180 }, { "grad_norm": 0.31104588508605957, "learning_rate": 8.483836156754328e-05, "loss": 0.0175, "step": 29190 }, { "grad_norm": 0.187296524643898, "learning_rate": 8.482649936836491e-05, "loss": 0.019, "step": 29200 }, { "grad_norm": 0.4323046803474426, "learning_rate": 8.481463336061293e-05, "loss": 0.0203, "step": 29210 }, { "grad_norm": 0.2045111209154129, "learning_rate": 8.480276354558496e-05, "loss": 0.0175, "step": 29220 }, { "grad_norm": 0.18166925013065338, "learning_rate": 8.479088992457913e-05, "loss": 0.019, "step": 29230 }, { "grad_norm": 0.2783075273036957, "learning_rate": 8.477901249889387e-05, "loss": 0.0194, "step": 29240 }, { "grad_norm": 0.1629195213317871, "learning_rate": 8.47671312698281e-05, "loss": 0.0153, "step": 29250 }, { "grad_norm": 0.2108171284198761, "learning_rate": 8.475524623868112e-05, "loss": 0.0203, "step": 29260 }, { "grad_norm": 0.20692197978496552, "learning_rate": 8.474335740675266e-05, "loss": 0.0198, "step": 29270 }, { "grad_norm": 0.19881561398506165, "learning_rate": 8.473146477534289e-05, "loss": 0.0181, "step": 29280 }, { "grad_norm": 0.22076310217380524, "learning_rate": 8.471956834575232e-05, "loss": 0.0174, "step": 29290 }, { "grad_norm": 0.20926344394683838, "learning_rate": 8.470766811928197e-05, "loss": 0.0162, "step": 29300 }, { "grad_norm": 0.24882300198078156, "learning_rate": 8.469576409723323e-05, "loss": 0.0154, "step": 29310 }, { "grad_norm": 0.28848952054977417, "learning_rate": 8.468385628090788e-05, "loss": 0.0168, "step": 29320 }, { "grad_norm": 0.17874957621097565, "learning_rate": 8.467194467160815e-05, "loss": 0.0155, "step": 29330 }, { "grad_norm": 0.26133790612220764, "learning_rate": 8.466002927063667e-05, "loss": 0.018, "step": 29340 }, { "grad_norm": 0.21091993153095245, "learning_rate": 8.464811007929651e-05, "loss": 0.0184, "step": 29350 }, { "grad_norm": 0.21280647814273834, "learning_rate": 8.463618709889114e-05, "loss": 0.0154, "step": 29360 }, { "grad_norm": 0.1899724155664444, "learning_rate": 8.462426033072442e-05, "loss": 0.0158, "step": 29370 }, { "grad_norm": 0.27694299817085266, "learning_rate": 8.461232977610061e-05, "loss": 0.014, "step": 29380 }, { "grad_norm": 0.2195480763912201, "learning_rate": 8.46003954363245e-05, "loss": 0.0163, "step": 29390 }, { "grad_norm": 0.19594670832157135, "learning_rate": 8.458845731270115e-05, "loss": 0.0206, "step": 29400 }, { "grad_norm": 0.22800222039222717, "learning_rate": 8.45765154065361e-05, "loss": 0.0227, "step": 29410 }, { "grad_norm": 0.2500700056552887, "learning_rate": 8.456456971913532e-05, "loss": 0.022, "step": 29420 }, { "grad_norm": 0.2289842665195465, "learning_rate": 8.455262025180517e-05, "loss": 0.0244, "step": 29430 }, { "grad_norm": 0.15248490869998932, "learning_rate": 8.454066700585242e-05, "loss": 0.0178, "step": 29440 }, { "grad_norm": 0.26687902212142944, "learning_rate": 8.452870998258423e-05, "loss": 0.0174, "step": 29450 }, { "grad_norm": 0.19411364197731018, "learning_rate": 8.451674918330825e-05, "loss": 0.0215, "step": 29460 }, { "grad_norm": 0.183277890086174, "learning_rate": 8.450478460933246e-05, "loss": 0.0187, "step": 29470 }, { "grad_norm": 0.17035944759845734, "learning_rate": 8.449281626196532e-05, "loss": 0.0179, "step": 29480 }, { "grad_norm": 0.3581453859806061, "learning_rate": 8.448084414251564e-05, "loss": 0.0191, "step": 29490 }, { "grad_norm": 0.18219509720802307, "learning_rate": 8.446886825229271e-05, "loss": 0.0176, "step": 29500 }, { "grad_norm": 0.20339597761631012, "learning_rate": 8.445688859260615e-05, "loss": 0.0177, "step": 29510 }, { "grad_norm": 0.2231944054365158, "learning_rate": 8.444490516476606e-05, "loss": 0.0197, "step": 29520 }, { "grad_norm": 0.30724915862083435, "learning_rate": 8.443291797008293e-05, "loss": 0.0218, "step": 29530 }, { "grad_norm": 0.1968832165002823, "learning_rate": 8.442092700986765e-05, "loss": 0.02, "step": 29540 }, { "grad_norm": 0.19913232326507568, "learning_rate": 8.440893228543156e-05, "loss": 0.0171, "step": 29550 }, { "grad_norm": 0.2305249720811844, "learning_rate": 8.439693379808638e-05, "loss": 0.0173, "step": 29560 }, { "grad_norm": 0.17794553935527802, "learning_rate": 8.43849315491442e-05, "loss": 0.0172, "step": 29570 }, { "grad_norm": 0.1615249216556549, "learning_rate": 8.437292553991763e-05, "loss": 0.0213, "step": 29580 }, { "grad_norm": 0.28894245624542236, "learning_rate": 8.436091577171959e-05, "loss": 0.0175, "step": 29590 }, { "grad_norm": 0.2320665419101715, "learning_rate": 8.434890224586347e-05, "loss": 0.0202, "step": 29600 }, { "grad_norm": 0.22378382086753845, "learning_rate": 8.433688496366303e-05, "loss": 0.0183, "step": 29610 }, { "grad_norm": 0.19484205543994904, "learning_rate": 8.432486392643248e-05, "loss": 0.0175, "step": 29620 }, { "grad_norm": 0.27237504720687866, "learning_rate": 8.431283913548643e-05, "loss": 0.0206, "step": 29630 }, { "grad_norm": 0.2038906216621399, "learning_rate": 8.430081059213985e-05, "loss": 0.0185, "step": 29640 }, { "grad_norm": 0.2686065435409546, "learning_rate": 8.428877829770823e-05, "loss": 0.0162, "step": 29650 }, { "grad_norm": 0.24294577538967133, "learning_rate": 8.427674225350735e-05, "loss": 0.0165, "step": 29660 }, { "grad_norm": 0.17139488458633423, "learning_rate": 8.426470246085347e-05, "loss": 0.0143, "step": 29670 }, { "grad_norm": 0.151569664478302, "learning_rate": 8.425265892106324e-05, "loss": 0.0165, "step": 29680 }, { "grad_norm": 0.19058075547218323, "learning_rate": 8.424061163545374e-05, "loss": 0.0168, "step": 29690 }, { "grad_norm": 0.1564943641424179, "learning_rate": 8.422856060534243e-05, "loss": 0.0163, "step": 29700 }, { "grad_norm": 0.18367940187454224, "learning_rate": 8.421650583204718e-05, "loss": 0.0159, "step": 29710 }, { "grad_norm": 0.19064974784851074, "learning_rate": 8.420444731688633e-05, "loss": 0.0147, "step": 29720 }, { "grad_norm": 0.17906031012535095, "learning_rate": 8.419238506117852e-05, "loss": 0.0146, "step": 29730 }, { "grad_norm": 0.15799397230148315, "learning_rate": 8.418031906624289e-05, "loss": 0.0159, "step": 29740 }, { "grad_norm": 0.22171637415885925, "learning_rate": 8.416824933339898e-05, "loss": 0.0156, "step": 29750 }, { "grad_norm": 0.1649218499660492, "learning_rate": 8.415617586396667e-05, "loss": 0.0152, "step": 29760 }, { "grad_norm": 0.19930462539196014, "learning_rate": 8.414409865926632e-05, "loss": 0.015, "step": 29770 }, { "grad_norm": 0.21855305135250092, "learning_rate": 8.413201772061867e-05, "loss": 0.0163, "step": 29780 }, { "grad_norm": 0.16088613867759705, "learning_rate": 8.411993304934488e-05, "loss": 0.0153, "step": 29790 }, { "grad_norm": 0.2374972403049469, "learning_rate": 8.410784464676654e-05, "loss": 0.016, "step": 29800 }, { "grad_norm": 0.305396169424057, "learning_rate": 8.409575251420556e-05, "loss": 0.0176, "step": 29810 }, { "grad_norm": 0.1832430511713028, "learning_rate": 8.408365665298435e-05, "loss": 0.0156, "step": 29820 }, { "grad_norm": 0.36221328377723694, "learning_rate": 8.40715570644257e-05, "loss": 0.0147, "step": 29830 }, { "grad_norm": 0.20998847484588623, "learning_rate": 8.40594537498528e-05, "loss": 0.0162, "step": 29840 }, { "grad_norm": 0.22337688505649567, "learning_rate": 8.404734671058924e-05, "loss": 0.0161, "step": 29850 }, { "grad_norm": 0.17822018265724182, "learning_rate": 8.403523594795902e-05, "loss": 0.017, "step": 29860 }, { "grad_norm": 0.1451435387134552, "learning_rate": 8.402312146328659e-05, "loss": 0.0152, "step": 29870 }, { "grad_norm": 0.23524482548236847, "learning_rate": 8.401100325789675e-05, "loss": 0.0168, "step": 29880 }, { "grad_norm": 0.19674408435821533, "learning_rate": 8.399888133311472e-05, "loss": 0.0177, "step": 29890 }, { "grad_norm": 0.16990669071674347, "learning_rate": 8.398675569026613e-05, "loss": 0.016, "step": 29900 }, { "grad_norm": 0.19891469180583954, "learning_rate": 8.397462633067705e-05, "loss": 0.02, "step": 29910 }, { "grad_norm": 0.2926709055900574, "learning_rate": 8.396249325567392e-05, "loss": 0.022, "step": 29920 }, { "grad_norm": 0.24104996025562286, "learning_rate": 8.395035646658357e-05, "loss": 0.02, "step": 29930 }, { "grad_norm": 0.2006831169128418, "learning_rate": 8.39382159647333e-05, "loss": 0.0177, "step": 29940 }, { "grad_norm": 0.19167117774486542, "learning_rate": 8.392607175145075e-05, "loss": 0.0179, "step": 29950 }, { "grad_norm": 0.21217229962348938, "learning_rate": 8.3913923828064e-05, "loss": 0.015, "step": 29960 }, { "grad_norm": 0.22693006694316864, "learning_rate": 8.390177219590152e-05, "loss": 0.0169, "step": 29970 }, { "grad_norm": 0.2504196763038635, "learning_rate": 8.388961685629222e-05, "loss": 0.0179, "step": 29980 }, { "grad_norm": 0.20017872750759125, "learning_rate": 8.387745781056536e-05, "loss": 0.0162, "step": 29990 }, { "grad_norm": 0.17403241991996765, "learning_rate": 8.386529506005065e-05, "loss": 0.0156, "step": 30000 }, { "grad_norm": 0.13879135251045227, "learning_rate": 8.38531286060782e-05, "loss": 0.0171, "step": 30010 }, { "grad_norm": 0.1786445528268814, "learning_rate": 8.384095844997849e-05, "loss": 0.0163, "step": 30020 }, { "grad_norm": 0.33265063166618347, "learning_rate": 8.382878459308245e-05, "loss": 0.0154, "step": 30030 }, { "grad_norm": 0.21190357208251953, "learning_rate": 8.381660703672138e-05, "loss": 0.0171, "step": 30040 }, { "grad_norm": 0.18219758570194244, "learning_rate": 8.380442578222702e-05, "loss": 0.0166, "step": 30050 }, { "grad_norm": 0.26443102955818176, "learning_rate": 8.379224083093146e-05, "loss": 0.0157, "step": 30060 }, { "grad_norm": 0.2241586297750473, "learning_rate": 8.378005218416727e-05, "loss": 0.021, "step": 30070 }, { "grad_norm": 0.179153710603714, "learning_rate": 8.376785984326735e-05, "loss": 0.0156, "step": 30080 }, { "grad_norm": 0.3715363144874573, "learning_rate": 8.375566380956506e-05, "loss": 0.0176, "step": 30090 }, { "grad_norm": 0.1932556927204132, "learning_rate": 8.374346408439411e-05, "loss": 0.022, "step": 30100 }, { "grad_norm": 0.31936901807785034, "learning_rate": 8.373126066908868e-05, "loss": 0.0176, "step": 30110 }, { "grad_norm": 0.1857139766216278, "learning_rate": 8.371905356498326e-05, "loss": 0.0174, "step": 30120 }, { "grad_norm": 0.1637657880783081, "learning_rate": 8.370684277341288e-05, "loss": 0.0197, "step": 30130 }, { "grad_norm": 0.25786375999450684, "learning_rate": 8.369462829571282e-05, "loss": 0.0161, "step": 30140 }, { "grad_norm": 0.26090389490127563, "learning_rate": 8.36824101332189e-05, "loss": 0.017, "step": 30150 }, { "grad_norm": 0.144041508436203, "learning_rate": 8.367018828726721e-05, "loss": 0.0139, "step": 30160 }, { "grad_norm": 0.2182389795780182, "learning_rate": 8.365796275919438e-05, "loss": 0.0152, "step": 30170 }, { "grad_norm": 0.21100442111492157, "learning_rate": 8.364573355033734e-05, "loss": 0.0153, "step": 30180 }, { "grad_norm": 0.3434284031391144, "learning_rate": 8.363350066203346e-05, "loss": 0.0122, "step": 30190 }, { "grad_norm": 0.171081081032753, "learning_rate": 8.362126409562053e-05, "loss": 0.0165, "step": 30200 }, { "grad_norm": 0.19179506599903107, "learning_rate": 8.360902385243667e-05, "loss": 0.0177, "step": 30210 }, { "grad_norm": 0.19560255110263824, "learning_rate": 8.359677993382052e-05, "loss": 0.0188, "step": 30220 }, { "grad_norm": 0.2090933918952942, "learning_rate": 8.358453234111103e-05, "loss": 0.0187, "step": 30230 }, { "grad_norm": 0.2822667062282562, "learning_rate": 8.357228107564756e-05, "loss": 0.0187, "step": 30240 }, { "grad_norm": 0.19122877717018127, "learning_rate": 8.356002613876993e-05, "loss": 0.0179, "step": 30250 }, { "grad_norm": 0.20970427989959717, "learning_rate": 8.35477675318183e-05, "loss": 0.0163, "step": 30260 }, { "grad_norm": 0.2549876868724823, "learning_rate": 8.353550525613323e-05, "loss": 0.0176, "step": 30270 }, { "grad_norm": 0.17667600512504578, "learning_rate": 8.352323931305572e-05, "loss": 0.0157, "step": 30280 }, { "grad_norm": 0.18238693475723267, "learning_rate": 8.351096970392717e-05, "loss": 0.0144, "step": 30290 }, { "grad_norm": 0.19193638861179352, "learning_rate": 8.349869643008937e-05, "loss": 0.0166, "step": 30300 }, { "grad_norm": 0.23242110013961792, "learning_rate": 8.348641949288449e-05, "loss": 0.0154, "step": 30310 }, { "grad_norm": 0.19046439230442047, "learning_rate": 8.34741388936551e-05, "loss": 0.0153, "step": 30320 }, { "grad_norm": 0.19403420388698578, "learning_rate": 8.346185463374423e-05, "loss": 0.0178, "step": 30330 }, { "grad_norm": 0.29232168197631836, "learning_rate": 8.344956671449524e-05, "loss": 0.0156, "step": 30340 }, { "grad_norm": 0.18312670290470123, "learning_rate": 8.343727513725192e-05, "loss": 0.016, "step": 30350 }, { "grad_norm": 0.14935302734375, "learning_rate": 8.342497990335847e-05, "loss": 0.0155, "step": 30360 }, { "grad_norm": 0.15473370254039764, "learning_rate": 8.341268101415946e-05, "loss": 0.0153, "step": 30370 }, { "grad_norm": 0.17044758796691895, "learning_rate": 8.34003784709999e-05, "loss": 0.0145, "step": 30380 }, { "grad_norm": 0.4573553204536438, "learning_rate": 8.338807227522516e-05, "loss": 0.0151, "step": 30390 }, { "grad_norm": 0.15342964231967926, "learning_rate": 8.337576242818103e-05, "loss": 0.0201, "step": 30400 }, { "grad_norm": 0.16665400564670563, "learning_rate": 8.336344893121372e-05, "loss": 0.019, "step": 30410 }, { "grad_norm": 0.21818353235721588, "learning_rate": 8.335113178566977e-05, "loss": 0.016, "step": 30420 }, { "grad_norm": 0.2819567918777466, "learning_rate": 8.33388109928962e-05, "loss": 0.0147, "step": 30430 }, { "grad_norm": 0.14713330566883087, "learning_rate": 8.33264865542404e-05, "loss": 0.0153, "step": 30440 }, { "grad_norm": 0.2696647644042969, "learning_rate": 8.331415847105013e-05, "loss": 0.0159, "step": 30450 }, { "grad_norm": 0.19368930160999298, "learning_rate": 8.330182674467357e-05, "loss": 0.0151, "step": 30460 }, { "grad_norm": 0.19426292181015015, "learning_rate": 8.32894913764593e-05, "loss": 0.014, "step": 30470 }, { "grad_norm": 0.1831308752298355, "learning_rate": 8.327715236775633e-05, "loss": 0.0147, "step": 30480 }, { "grad_norm": 0.12989914417266846, "learning_rate": 8.326480971991398e-05, "loss": 0.0139, "step": 30490 }, { "grad_norm": 0.17040564119815826, "learning_rate": 8.325246343428206e-05, "loss": 0.013, "step": 30500 }, { "grad_norm": 0.20456033945083618, "learning_rate": 8.324011351221072e-05, "loss": 0.0129, "step": 30510 }, { "grad_norm": 0.4503275156021118, "learning_rate": 8.322775995505057e-05, "loss": 0.0162, "step": 30520 }, { "grad_norm": 0.21625708043575287, "learning_rate": 8.321540276415254e-05, "loss": 0.0156, "step": 30530 }, { "grad_norm": 0.1957470327615738, "learning_rate": 8.320304194086798e-05, "loss": 0.0164, "step": 30540 }, { "grad_norm": 0.2899341583251953, "learning_rate": 8.31906774865487e-05, "loss": 0.0157, "step": 30550 }, { "grad_norm": 0.22872135043144226, "learning_rate": 8.317830940254682e-05, "loss": 0.0166, "step": 30560 }, { "grad_norm": 0.21358565986156464, "learning_rate": 8.316593769021491e-05, "loss": 0.0162, "step": 30570 }, { "grad_norm": 0.20166802406311035, "learning_rate": 8.315356235090592e-05, "loss": 0.014, "step": 30580 }, { "grad_norm": 0.18711329996585846, "learning_rate": 8.314118338597319e-05, "loss": 0.016, "step": 30590 }, { "grad_norm": 0.18394623696804047, "learning_rate": 8.312880079677048e-05, "loss": 0.0157, "step": 30600 }, { "grad_norm": 0.1912607103586197, "learning_rate": 8.311641458465191e-05, "loss": 0.0159, "step": 30610 }, { "grad_norm": 0.18674235045909882, "learning_rate": 8.310402475097205e-05, "loss": 0.0148, "step": 30620 }, { "grad_norm": 0.24858827888965607, "learning_rate": 8.309163129708581e-05, "loss": 0.0151, "step": 30630 }, { "grad_norm": 0.2006826549768448, "learning_rate": 8.307923422434852e-05, "loss": 0.014, "step": 30640 }, { "grad_norm": 0.1644306629896164, "learning_rate": 8.30668335341159e-05, "loss": 0.014, "step": 30650 }, { "grad_norm": 0.23960460722446442, "learning_rate": 8.305442922774408e-05, "loss": 0.0149, "step": 30660 }, { "grad_norm": 0.2853119969367981, "learning_rate": 8.304202130658959e-05, "loss": 0.0159, "step": 30670 }, { "grad_norm": 0.17924416065216064, "learning_rate": 8.302960977200931e-05, "loss": 0.0186, "step": 30680 }, { "grad_norm": 0.2370741218328476, "learning_rate": 8.301719462536058e-05, "loss": 0.0156, "step": 30690 }, { "grad_norm": 0.18199126422405243, "learning_rate": 8.300477586800108e-05, "loss": 0.0163, "step": 30700 }, { "grad_norm": 0.17212572693824768, "learning_rate": 8.299235350128892e-05, "loss": 0.0165, "step": 30710 }, { "grad_norm": 0.24427874386310577, "learning_rate": 8.297992752658258e-05, "loss": 0.0149, "step": 30720 }, { "grad_norm": 0.17878401279449463, "learning_rate": 8.296749794524094e-05, "loss": 0.0172, "step": 30730 }, { "grad_norm": 0.21492832899093628, "learning_rate": 8.295506475862332e-05, "loss": 0.0147, "step": 30740 }, { "grad_norm": 0.1946839839220047, "learning_rate": 8.294262796808933e-05, "loss": 0.016, "step": 30750 }, { "grad_norm": 0.17737945914268494, "learning_rate": 8.293018757499909e-05, "loss": 0.0151, "step": 30760 }, { "grad_norm": 0.2108818143606186, "learning_rate": 8.291774358071305e-05, "loss": 0.014, "step": 30770 }, { "grad_norm": 0.16614514589309692, "learning_rate": 8.290529598659205e-05, "loss": 0.0141, "step": 30780 }, { "grad_norm": 0.18944382667541504, "learning_rate": 8.289284479399738e-05, "loss": 0.0142, "step": 30790 }, { "grad_norm": 0.4380946755409241, "learning_rate": 8.288039000429064e-05, "loss": 0.014, "step": 30800 }, { "grad_norm": 0.15062794089317322, "learning_rate": 8.286793161883388e-05, "loss": 0.0148, "step": 30810 }, { "grad_norm": 0.17547203600406647, "learning_rate": 8.285546963898954e-05, "loss": 0.0151, "step": 30820 }, { "grad_norm": 0.22963695228099823, "learning_rate": 8.284300406612044e-05, "loss": 0.015, "step": 30830 }, { "grad_norm": 0.19747111201286316, "learning_rate": 8.283053490158978e-05, "loss": 0.014, "step": 30840 }, { "grad_norm": 0.2513428032398224, "learning_rate": 8.28180621467612e-05, "loss": 0.0144, "step": 30850 }, { "grad_norm": 0.2064928263425827, "learning_rate": 8.280558580299867e-05, "loss": 0.0148, "step": 30860 }, { "grad_norm": 0.1705877184867859, "learning_rate": 8.279310587166661e-05, "loss": 0.0125, "step": 30870 }, { "grad_norm": 0.198860302567482, "learning_rate": 8.278062235412978e-05, "loss": 0.0165, "step": 30880 }, { "grad_norm": 0.2095491886138916, "learning_rate": 8.276813525175339e-05, "loss": 0.0165, "step": 30890 }, { "grad_norm": 0.24300922453403473, "learning_rate": 8.2755644565903e-05, "loss": 0.0188, "step": 30900 }, { "grad_norm": 0.23150691390037537, "learning_rate": 8.274315029794454e-05, "loss": 0.0135, "step": 30910 }, { "grad_norm": 0.20929741859436035, "learning_rate": 8.273065244924443e-05, "loss": 0.014, "step": 30920 }, { "grad_norm": 0.22238720953464508, "learning_rate": 8.271815102116936e-05, "loss": 0.0145, "step": 30930 }, { "grad_norm": 0.16854345798492432, "learning_rate": 8.270564601508648e-05, "loss": 0.0171, "step": 30940 }, { "grad_norm": 0.1652965396642685, "learning_rate": 8.269313743236333e-05, "loss": 0.016, "step": 30950 }, { "grad_norm": 0.15949316322803497, "learning_rate": 8.268062527436783e-05, "loss": 0.0163, "step": 30960 }, { "grad_norm": 0.27414292097091675, "learning_rate": 8.266810954246828e-05, "loss": 0.0126, "step": 30970 }, { "grad_norm": 0.20123140513896942, "learning_rate": 8.265559023803338e-05, "loss": 0.014, "step": 30980 }, { "grad_norm": 0.1595940738916397, "learning_rate": 8.264306736243223e-05, "loss": 0.0153, "step": 30990 }, { "grad_norm": 0.12555673718452454, "learning_rate": 8.263054091703432e-05, "loss": 0.014, "step": 31000 }, { "grad_norm": 0.19921283423900604, "learning_rate": 8.26180109032095e-05, "loss": 0.0154, "step": 31010 }, { "grad_norm": 0.20205244421958923, "learning_rate": 8.260547732232807e-05, "loss": 0.0142, "step": 31020 }, { "grad_norm": 0.16471143066883087, "learning_rate": 8.259294017576063e-05, "loss": 0.0152, "step": 31030 }, { "grad_norm": 0.18327133357524872, "learning_rate": 8.258039946487828e-05, "loss": 0.0148, "step": 31040 }, { "grad_norm": 0.13789792358875275, "learning_rate": 8.256785519105241e-05, "loss": 0.0146, "step": 31050 }, { "grad_norm": 0.23307105898857117, "learning_rate": 8.255530735565488e-05, "loss": 0.0172, "step": 31060 }, { "grad_norm": 0.19204117357730865, "learning_rate": 8.254275596005787e-05, "loss": 0.0135, "step": 31070 }, { "grad_norm": 0.18557728826999664, "learning_rate": 8.2530201005634e-05, "loss": 0.0165, "step": 31080 }, { "grad_norm": 0.22984085977077484, "learning_rate": 8.251764249375626e-05, "loss": 0.0148, "step": 31090 }, { "grad_norm": 0.19271793961524963, "learning_rate": 8.250508042579803e-05, "loss": 0.0149, "step": 31100 }, { "grad_norm": 0.2794313430786133, "learning_rate": 8.249251480313307e-05, "loss": 0.015, "step": 31110 }, { "grad_norm": 0.20499694347381592, "learning_rate": 8.247994562713555e-05, "loss": 0.0155, "step": 31120 }, { "grad_norm": 0.17671813070774078, "learning_rate": 8.246737289917998e-05, "loss": 0.0147, "step": 31130 }, { "grad_norm": 0.1842368096113205, "learning_rate": 8.245479662064135e-05, "loss": 0.0136, "step": 31140 }, { "grad_norm": 0.17221760749816895, "learning_rate": 8.244221679289496e-05, "loss": 0.014, "step": 31150 }, { "grad_norm": 0.2054382562637329, "learning_rate": 8.242963341731652e-05, "loss": 0.0139, "step": 31160 }, { "grad_norm": 0.16275468468666077, "learning_rate": 8.24170464952821e-05, "loss": 0.0145, "step": 31170 }, { "grad_norm": 0.1515771597623825, "learning_rate": 8.240445602816824e-05, "loss": 0.0158, "step": 31180 }, { "grad_norm": 0.30884161591529846, "learning_rate": 8.239186201735179e-05, "loss": 0.0179, "step": 31190 }, { "grad_norm": 0.18953925371170044, "learning_rate": 8.237926446420998e-05, "loss": 0.0143, "step": 31200 }, { "grad_norm": 0.14093653857707977, "learning_rate": 8.236666337012052e-05, "loss": 0.0124, "step": 31210 }, { "grad_norm": 0.15027040243148804, "learning_rate": 8.23540587364614e-05, "loss": 0.0154, "step": 31220 }, { "grad_norm": 0.1677892953157425, "learning_rate": 8.234145056461107e-05, "loss": 0.0127, "step": 31230 }, { "grad_norm": 0.23288752138614655, "learning_rate": 8.232883885594832e-05, "loss": 0.0131, "step": 31240 }, { "grad_norm": 0.24128517508506775, "learning_rate": 8.231622361185236e-05, "loss": 0.0159, "step": 31250 }, { "grad_norm": 0.20237715542316437, "learning_rate": 8.230360483370278e-05, "loss": 0.0136, "step": 31260 }, { "grad_norm": 0.16789525747299194, "learning_rate": 8.229098252287953e-05, "loss": 0.0154, "step": 31270 }, { "grad_norm": 0.15873755514621735, "learning_rate": 8.2278356680763e-05, "loss": 0.0137, "step": 31280 }, { "grad_norm": 0.18880341947078705, "learning_rate": 8.22657273087339e-05, "loss": 0.0146, "step": 31290 }, { "grad_norm": 0.17241108417510986, "learning_rate": 8.225309440817336e-05, "loss": 0.0139, "step": 31300 }, { "grad_norm": 0.2578720152378082, "learning_rate": 8.224045798046293e-05, "loss": 0.0151, "step": 31310 }, { "grad_norm": 0.1987481415271759, "learning_rate": 8.22278180269845e-05, "loss": 0.0147, "step": 31320 }, { "grad_norm": 0.17915533483028412, "learning_rate": 8.221517454912031e-05, "loss": 0.016, "step": 31330 }, { "grad_norm": 0.1488218903541565, "learning_rate": 8.220252754825308e-05, "loss": 0.0156, "step": 31340 }, { "grad_norm": 0.2625662386417389, "learning_rate": 8.218987702576586e-05, "loss": 0.0147, "step": 31350 }, { "grad_norm": 0.19235318899154663, "learning_rate": 8.217722298304207e-05, "loss": 0.0129, "step": 31360 }, { "grad_norm": 0.19127386808395386, "learning_rate": 8.216456542146557e-05, "loss": 0.0168, "step": 31370 }, { "grad_norm": 0.19504208862781525, "learning_rate": 8.215190434242055e-05, "loss": 0.0124, "step": 31380 }, { "grad_norm": 0.27707627415657043, "learning_rate": 8.213923974729161e-05, "loss": 0.017, "step": 31390 }, { "grad_norm": 0.14170123636722565, "learning_rate": 8.212657163746373e-05, "loss": 0.0146, "step": 31400 }, { "grad_norm": 0.21388785541057587, "learning_rate": 8.211390001432227e-05, "loss": 0.0171, "step": 31410 }, { "grad_norm": 0.17587055265903473, "learning_rate": 8.210122487925297e-05, "loss": 0.014, "step": 31420 }, { "grad_norm": 0.19202503561973572, "learning_rate": 8.208854623364202e-05, "loss": 0.0147, "step": 31430 }, { "grad_norm": 0.22657103836536407, "learning_rate": 8.207586407887589e-05, "loss": 0.0142, "step": 31440 }, { "grad_norm": 0.19195227324962616, "learning_rate": 8.206317841634148e-05, "loss": 0.0143, "step": 31450 }, { "grad_norm": 0.19487513601779938, "learning_rate": 8.205048924742609e-05, "loss": 0.0136, "step": 31460 }, { "grad_norm": 0.36557117104530334, "learning_rate": 8.203779657351738e-05, "loss": 0.0146, "step": 31470 }, { "grad_norm": 0.186284601688385, "learning_rate": 8.20251003960034e-05, "loss": 0.015, "step": 31480 }, { "grad_norm": 0.14875918626785278, "learning_rate": 8.201240071627258e-05, "loss": 0.014, "step": 31490 }, { "grad_norm": 0.1436607837677002, "learning_rate": 8.199969753571377e-05, "loss": 0.0131, "step": 31500 }, { "grad_norm": 0.30373066663742065, "learning_rate": 8.198699085571615e-05, "loss": 0.0154, "step": 31510 }, { "grad_norm": 0.18658815324306488, "learning_rate": 8.197428067766928e-05, "loss": 0.0149, "step": 31520 }, { "grad_norm": 0.16879895329475403, "learning_rate": 8.196156700296316e-05, "loss": 0.0149, "step": 31530 }, { "grad_norm": 0.15596631169319153, "learning_rate": 8.194884983298814e-05, "loss": 0.0145, "step": 31540 }, { "grad_norm": 0.2808322310447693, "learning_rate": 8.193612916913491e-05, "loss": 0.0159, "step": 31550 }, { "grad_norm": 0.1394181251525879, "learning_rate": 8.192340501279463e-05, "loss": 0.0148, "step": 31560 }, { "grad_norm": 0.19231191277503967, "learning_rate": 8.191067736535876e-05, "loss": 0.0148, "step": 31570 }, { "grad_norm": 0.21075478196144104, "learning_rate": 8.18979462282192e-05, "loss": 0.0137, "step": 31580 }, { "grad_norm": 0.20330046117305756, "learning_rate": 8.188521160276819e-05, "loss": 0.0141, "step": 31590 }, { "grad_norm": 0.20497842133045197, "learning_rate": 8.187247349039837e-05, "loss": 0.0134, "step": 31600 }, { "grad_norm": 0.18575721979141235, "learning_rate": 8.185973189250278e-05, "loss": 0.0149, "step": 31610 }, { "grad_norm": 0.17709340155124664, "learning_rate": 8.184698681047482e-05, "loss": 0.0158, "step": 31620 }, { "grad_norm": 0.1889292448759079, "learning_rate": 8.183423824570827e-05, "loss": 0.0157, "step": 31630 }, { "grad_norm": 0.1663990318775177, "learning_rate": 8.182148619959725e-05, "loss": 0.0141, "step": 31640 }, { "grad_norm": 0.1681874543428421, "learning_rate": 8.180873067353636e-05, "loss": 0.0155, "step": 31650 }, { "grad_norm": 0.1575501263141632, "learning_rate": 8.179597166892052e-05, "loss": 0.0141, "step": 31660 }, { "grad_norm": 0.25214245915412903, "learning_rate": 8.178320918714501e-05, "loss": 0.015, "step": 31670 }, { "grad_norm": 0.35732781887054443, "learning_rate": 8.177044322960554e-05, "loss": 0.0144, "step": 31680 }, { "grad_norm": 0.22109632194042206, "learning_rate": 8.175767379769816e-05, "loss": 0.0141, "step": 31690 }, { "grad_norm": 0.19789929687976837, "learning_rate": 8.174490089281932e-05, "loss": 0.0154, "step": 31700 }, { "grad_norm": 0.21391257643699646, "learning_rate": 8.173212451636584e-05, "loss": 0.0154, "step": 31710 }, { "grad_norm": 0.1423940360546112, "learning_rate": 8.171934466973493e-05, "loss": 0.0133, "step": 31720 }, { "grad_norm": 0.17232581973075867, "learning_rate": 8.170656135432418e-05, "loss": 0.0124, "step": 31730 }, { "grad_norm": 0.20470330119132996, "learning_rate": 8.169377457153155e-05, "loss": 0.0154, "step": 31740 }, { "grad_norm": 0.16715079545974731, "learning_rate": 8.168098432275539e-05, "loss": 0.0165, "step": 31750 }, { "grad_norm": 0.18925043940544128, "learning_rate": 8.166819060939442e-05, "loss": 0.0133, "step": 31760 }, { "grad_norm": 0.23602844774723053, "learning_rate": 8.165539343284772e-05, "loss": 0.0141, "step": 31770 }, { "grad_norm": 0.26529887318611145, "learning_rate": 8.16425927945148e-05, "loss": 0.0171, "step": 31780 }, { "grad_norm": 0.2880735695362091, "learning_rate": 8.162978869579551e-05, "loss": 0.0186, "step": 31790 }, { "grad_norm": 0.16661979258060455, "learning_rate": 8.161698113809007e-05, "loss": 0.0158, "step": 31800 }, { "grad_norm": 0.16155306994915009, "learning_rate": 8.160417012279911e-05, "loss": 0.0144, "step": 31810 }, { "grad_norm": 0.21387769281864166, "learning_rate": 8.159135565132363e-05, "loss": 0.0155, "step": 31820 }, { "grad_norm": 0.1807890087366104, "learning_rate": 8.157853772506498e-05, "loss": 0.0176, "step": 31830 }, { "grad_norm": 0.2058364897966385, "learning_rate": 8.156571634542494e-05, "loss": 0.0165, "step": 31840 }, { "grad_norm": 0.47873654961586, "learning_rate": 8.15528915138056e-05, "loss": 0.0171, "step": 31850 }, { "grad_norm": 0.17549777030944824, "learning_rate": 8.154006323160949e-05, "loss": 0.0159, "step": 31860 }, { "grad_norm": 0.17928844690322876, "learning_rate": 8.152723150023949e-05, "loss": 0.0145, "step": 31870 }, { "grad_norm": 0.2697289288043976, "learning_rate": 8.151439632109886e-05, "loss": 0.0157, "step": 31880 }, { "grad_norm": 0.19121412932872772, "learning_rate": 8.150155769559122e-05, "loss": 0.0175, "step": 31890 }, { "grad_norm": 0.2948274314403534, "learning_rate": 8.148871562512058e-05, "loss": 0.0157, "step": 31900 }, { "grad_norm": 0.1829981505870819, "learning_rate": 8.147587011109136e-05, "loss": 0.0145, "step": 31910 }, { "grad_norm": 0.20122554898262024, "learning_rate": 8.14630211549083e-05, "loss": 0.012, "step": 31920 }, { "grad_norm": 0.15800006687641144, "learning_rate": 8.145016875797655e-05, "loss": 0.0137, "step": 31930 }, { "grad_norm": 0.29930034279823303, "learning_rate": 8.143731292170164e-05, "loss": 0.0158, "step": 31940 }, { "grad_norm": 0.1961706131696701, "learning_rate": 8.142445364748944e-05, "loss": 0.0191, "step": 31950 }, { "grad_norm": 0.2766253352165222, "learning_rate": 8.141159093674624e-05, "loss": 0.0138, "step": 31960 }, { "grad_norm": 0.1821073442697525, "learning_rate": 8.139872479087869e-05, "loss": 0.0165, "step": 31970 }, { "grad_norm": 0.1925250142812729, "learning_rate": 8.13858552112938e-05, "loss": 0.0152, "step": 31980 }, { "grad_norm": 0.16974662244319916, "learning_rate": 8.137298219939895e-05, "loss": 0.0153, "step": 31990 }, { "grad_norm": 0.19100281596183777, "learning_rate": 8.136010575660196e-05, "loss": 0.0134, "step": 32000 }, { "grad_norm": 0.22067341208457947, "learning_rate": 8.134722588431095e-05, "loss": 0.0148, "step": 32010 }, { "grad_norm": 0.15999667346477509, "learning_rate": 8.133434258393444e-05, "loss": 0.0157, "step": 32020 }, { "grad_norm": 0.1908731758594513, "learning_rate": 8.132145585688134e-05, "loss": 0.0165, "step": 32030 }, { "grad_norm": 0.17660440504550934, "learning_rate": 8.130856570456093e-05, "loss": 0.0161, "step": 32040 }, { "grad_norm": 0.22436575591564178, "learning_rate": 8.129567212838283e-05, "loss": 0.0157, "step": 32050 }, { "grad_norm": 0.2061539888381958, "learning_rate": 8.128277512975708e-05, "loss": 0.0136, "step": 32060 }, { "grad_norm": 0.15352144837379456, "learning_rate": 8.126987471009408e-05, "loss": 0.0142, "step": 32070 }, { "grad_norm": 0.2322005331516266, "learning_rate": 8.125697087080459e-05, "loss": 0.0145, "step": 32080 }, { "grad_norm": 0.20305092632770538, "learning_rate": 8.124406361329976e-05, "loss": 0.0144, "step": 32090 }, { "grad_norm": 0.17709480226039886, "learning_rate": 8.123115293899111e-05, "loss": 0.0154, "step": 32100 }, { "grad_norm": 0.18527568876743317, "learning_rate": 8.121823884929055e-05, "loss": 0.0149, "step": 32110 }, { "grad_norm": 0.19049356877803802, "learning_rate": 8.12053213456103e-05, "loss": 0.0159, "step": 32120 }, { "grad_norm": 0.17354336380958557, "learning_rate": 8.119240042936303e-05, "loss": 0.0154, "step": 32130 }, { "grad_norm": 0.18075034022331238, "learning_rate": 8.117947610196175e-05, "loss": 0.0147, "step": 32140 }, { "grad_norm": 0.22349391877651215, "learning_rate": 8.116654836481982e-05, "loss": 0.0186, "step": 32150 }, { "grad_norm": 0.19230878353118896, "learning_rate": 8.115361721935106e-05, "loss": 0.0204, "step": 32160 }, { "grad_norm": 0.2281779646873474, "learning_rate": 8.114068266696953e-05, "loss": 0.0181, "step": 32170 }, { "grad_norm": 0.2537374794483185, "learning_rate": 8.112774470908978e-05, "loss": 0.0168, "step": 32180 }, { "grad_norm": 0.2125997543334961, "learning_rate": 8.111480334712665e-05, "loss": 0.0167, "step": 32190 }, { "grad_norm": 0.16096018254756927, "learning_rate": 8.110185858249542e-05, "loss": 0.0157, "step": 32200 }, { "grad_norm": 0.18575741350650787, "learning_rate": 8.108891041661168e-05, "loss": 0.0147, "step": 32210 }, { "grad_norm": 0.14566177129745483, "learning_rate": 8.107595885089146e-05, "loss": 0.0157, "step": 32220 }, { "grad_norm": 0.1839437633752823, "learning_rate": 8.106300388675108e-05, "loss": 0.0165, "step": 32230 }, { "grad_norm": 0.21250095963478088, "learning_rate": 8.105004552560731e-05, "loss": 0.015, "step": 32240 }, { "grad_norm": 0.20721715688705444, "learning_rate": 8.103708376887724e-05, "loss": 0.0142, "step": 32250 }, { "grad_norm": 0.2351553738117218, "learning_rate": 8.102411861797836e-05, "loss": 0.0153, "step": 32260 }, { "grad_norm": 0.1706009805202484, "learning_rate": 8.101115007432851e-05, "loss": 0.0146, "step": 32270 }, { "grad_norm": 0.23246216773986816, "learning_rate": 8.09981781393459e-05, "loss": 0.015, "step": 32280 }, { "grad_norm": 0.19942182302474976, "learning_rate": 8.098520281444915e-05, "loss": 0.0159, "step": 32290 }, { "grad_norm": 0.17314311861991882, "learning_rate": 8.09722241010572e-05, "loss": 0.016, "step": 32300 }, { "grad_norm": 0.19237282872200012, "learning_rate": 8.095924200058939e-05, "loss": 0.0152, "step": 32310 }, { "grad_norm": 0.2549073100090027, "learning_rate": 8.094625651446541e-05, "loss": 0.0166, "step": 32320 }, { "grad_norm": 0.17140960693359375, "learning_rate": 8.093326764410536e-05, "loss": 0.0142, "step": 32330 }, { "grad_norm": 0.20056742429733276, "learning_rate": 8.092027539092966e-05, "loss": 0.0168, "step": 32340 }, { "grad_norm": 0.27569979429244995, "learning_rate": 8.090727975635913e-05, "loss": 0.0158, "step": 32350 }, { "grad_norm": 0.18329432606697083, "learning_rate": 8.089428074181497e-05, "loss": 0.0148, "step": 32360 }, { "grad_norm": 0.14645802974700928, "learning_rate": 8.088127834871871e-05, "loss": 0.0133, "step": 32370 }, { "grad_norm": 0.1524561047554016, "learning_rate": 8.086827257849226e-05, "loss": 0.0148, "step": 32380 }, { "grad_norm": 0.2531762421131134, "learning_rate": 8.085526343255795e-05, "loss": 0.0141, "step": 32390 }, { "grad_norm": 0.1930985450744629, "learning_rate": 8.084225091233842e-05, "loss": 0.0154, "step": 32400 }, { "grad_norm": 0.18700100481510162, "learning_rate": 8.082923501925668e-05, "loss": 0.0144, "step": 32410 }, { "grad_norm": 0.1807379573583603, "learning_rate": 8.081621575473617e-05, "loss": 0.0196, "step": 32420 }, { "grad_norm": 0.19270822405815125, "learning_rate": 8.080319312020064e-05, "loss": 0.0172, "step": 32430 }, { "grad_norm": 0.3145652115345001, "learning_rate": 8.079016711707421e-05, "loss": 0.0146, "step": 32440 }, { "grad_norm": 0.2026902139186859, "learning_rate": 8.077713774678139e-05, "loss": 0.0144, "step": 32450 }, { "grad_norm": 0.17437566816806793, "learning_rate": 8.076410501074707e-05, "loss": 0.0146, "step": 32460 }, { "grad_norm": 0.18491198122501373, "learning_rate": 8.075106891039647e-05, "loss": 0.0156, "step": 32470 }, { "grad_norm": 0.16671493649482727, "learning_rate": 8.073802944715523e-05, "loss": 0.014, "step": 32480 }, { "grad_norm": 0.2245194911956787, "learning_rate": 8.072498662244929e-05, "loss": 0.0138, "step": 32490 }, { "grad_norm": 0.16628359258174896, "learning_rate": 8.0711940437705e-05, "loss": 0.017, "step": 32500 }, { "grad_norm": 0.16243356466293335, "learning_rate": 8.06988908943491e-05, "loss": 0.0148, "step": 32510 }, { "grad_norm": 0.1594965010881424, "learning_rate": 8.068583799380863e-05, "loss": 0.0164, "step": 32520 }, { "grad_norm": 0.2248980849981308, "learning_rate": 8.067278173751104e-05, "loss": 0.0147, "step": 32530 }, { "grad_norm": 0.2439408302307129, "learning_rate": 8.065972212688417e-05, "loss": 0.0161, "step": 32540 }, { "grad_norm": 0.21827660501003265, "learning_rate": 8.064665916335618e-05, "loss": 0.0159, "step": 32550 }, { "grad_norm": 0.19188211858272552, "learning_rate": 8.063359284835564e-05, "loss": 0.0198, "step": 32560 }, { "grad_norm": 0.3851165473461151, "learning_rate": 8.062052318331142e-05, "loss": 0.0202, "step": 32570 }, { "grad_norm": 0.2407860904932022, "learning_rate": 8.060745016965283e-05, "loss": 0.017, "step": 32580 }, { "grad_norm": 0.2120017260313034, "learning_rate": 8.059437380880952e-05, "loss": 0.0145, "step": 32590 }, { "grad_norm": 0.14085812866687775, "learning_rate": 8.058129410221146e-05, "loss": 0.014, "step": 32600 }, { "grad_norm": 0.13787929713726044, "learning_rate": 8.056821105128908e-05, "loss": 0.0168, "step": 32610 }, { "grad_norm": 0.19749538600444794, "learning_rate": 8.05551246574731e-05, "loss": 0.0165, "step": 32620 }, { "grad_norm": 0.2095981240272522, "learning_rate": 8.05420349221946e-05, "loss": 0.0145, "step": 32630 }, { "grad_norm": 0.1613462120294571, "learning_rate": 8.05289418468851e-05, "loss": 0.0141, "step": 32640 }, { "grad_norm": 0.21715974807739258, "learning_rate": 8.051584543297642e-05, "loss": 0.0152, "step": 32650 }, { "grad_norm": 0.1860501915216446, "learning_rate": 8.050274568190074e-05, "loss": 0.0127, "step": 32660 }, { "grad_norm": 0.2447880655527115, "learning_rate": 8.048964259509067e-05, "loss": 0.0147, "step": 32670 }, { "grad_norm": 0.25716307759284973, "learning_rate": 8.047653617397914e-05, "loss": 0.0147, "step": 32680 }, { "grad_norm": 0.1574353724718094, "learning_rate": 8.046342641999941e-05, "loss": 0.0149, "step": 32690 }, { "grad_norm": 0.1744568943977356, "learning_rate": 8.045031333458517e-05, "loss": 0.0153, "step": 32700 }, { "grad_norm": 0.2611774802207947, "learning_rate": 8.043719691917047e-05, "loss": 0.0142, "step": 32710 }, { "grad_norm": 0.14838804304599762, "learning_rate": 8.042407717518966e-05, "loss": 0.0129, "step": 32720 }, { "grad_norm": 0.30857613682746887, "learning_rate": 8.041095410407751e-05, "loss": 0.015, "step": 32730 }, { "grad_norm": 0.23515020310878754, "learning_rate": 8.039782770726913e-05, "loss": 0.0143, "step": 32740 }, { "grad_norm": 0.17952324450016022, "learning_rate": 8.038469798620004e-05, "loss": 0.0142, "step": 32750 }, { "grad_norm": 0.19322234392166138, "learning_rate": 8.037156494230604e-05, "loss": 0.0146, "step": 32760 }, { "grad_norm": 0.25562089681625366, "learning_rate": 8.035842857702338e-05, "loss": 0.0169, "step": 32770 }, { "grad_norm": 0.1902308464050293, "learning_rate": 8.03452888917886e-05, "loss": 0.0157, "step": 32780 }, { "grad_norm": 0.2419404238462448, "learning_rate": 8.033214588803866e-05, "loss": 0.0171, "step": 32790 }, { "grad_norm": 0.18080490827560425, "learning_rate": 8.031899956721083e-05, "loss": 0.0167, "step": 32800 }, { "grad_norm": 0.17142394185066223, "learning_rate": 8.030584993074282e-05, "loss": 0.016, "step": 32810 }, { "grad_norm": 0.27425652742385864, "learning_rate": 8.02926969800726e-05, "loss": 0.0143, "step": 32820 }, { "grad_norm": 0.18007510900497437, "learning_rate": 8.027954071663859e-05, "loss": 0.0125, "step": 32830 }, { "grad_norm": 0.19138646125793457, "learning_rate": 8.026638114187954e-05, "loss": 0.0164, "step": 32840 }, { "grad_norm": 0.23646299540996552, "learning_rate": 8.025321825723456e-05, "loss": 0.0162, "step": 32850 }, { "grad_norm": 0.18773236870765686, "learning_rate": 8.02400520641431e-05, "loss": 0.0143, "step": 32860 }, { "grad_norm": 0.21527086198329926, "learning_rate": 8.022688256404501e-05, "loss": 0.0166, "step": 32870 }, { "grad_norm": 0.19517311453819275, "learning_rate": 8.02137097583805e-05, "loss": 0.0149, "step": 32880 }, { "grad_norm": 0.1570393592119217, "learning_rate": 8.02005336485901e-05, "loss": 0.0157, "step": 32890 }, { "grad_norm": 0.18015208840370178, "learning_rate": 8.018735423611476e-05, "loss": 0.0154, "step": 32900 }, { "grad_norm": 0.16937389969825745, "learning_rate": 8.017417152239574e-05, "loss": 0.0161, "step": 32910 }, { "grad_norm": 0.1582004427909851, "learning_rate": 8.01609855088747e-05, "loss": 0.0117, "step": 32920 }, { "grad_norm": 0.18299753963947296, "learning_rate": 8.014779619699362e-05, "loss": 0.0147, "step": 32930 }, { "grad_norm": 0.22957256436347961, "learning_rate": 8.013460358819489e-05, "loss": 0.012, "step": 32940 }, { "grad_norm": 0.2129698544740677, "learning_rate": 8.01214076839212e-05, "loss": 0.0134, "step": 32950 }, { "grad_norm": 0.2578720450401306, "learning_rate": 8.010820848561565e-05, "loss": 0.0158, "step": 32960 }, { "grad_norm": 0.4217696189880371, "learning_rate": 8.009500599472171e-05, "loss": 0.0157, "step": 32970 }, { "grad_norm": 0.15206405520439148, "learning_rate": 8.008180021268314e-05, "loss": 0.0122, "step": 32980 }, { "grad_norm": 0.16525375843048096, "learning_rate": 8.006859114094414e-05, "loss": 0.0157, "step": 32990 }, { "grad_norm": 0.18171975016593933, "learning_rate": 8.005537878094921e-05, "loss": 0.0147, "step": 33000 }, { "grad_norm": 0.1342182457447052, "learning_rate": 8.004216313414323e-05, "loss": 0.0139, "step": 33010 }, { "grad_norm": 0.17318937182426453, "learning_rate": 8.002894420197149e-05, "loss": 0.0127, "step": 33020 }, { "grad_norm": 0.15160582959651947, "learning_rate": 8.001572198587954e-05, "loss": 0.0136, "step": 33030 }, { "grad_norm": 0.15623739361763, "learning_rate": 8.000249648731338e-05, "loss": 0.0123, "step": 33040 }, { "grad_norm": 0.21765993535518646, "learning_rate": 7.998926770771928e-05, "loss": 0.013, "step": 33050 }, { "grad_norm": 0.213839590549469, "learning_rate": 7.997603564854397e-05, "loss": 0.0126, "step": 33060 }, { "grad_norm": 0.21422769129276276, "learning_rate": 7.996280031123448e-05, "loss": 0.0129, "step": 33070 }, { "grad_norm": 0.22956092655658722, "learning_rate": 7.994956169723818e-05, "loss": 0.0155, "step": 33080 }, { "grad_norm": 0.14383624494075775, "learning_rate": 7.993631980800285e-05, "loss": 0.0153, "step": 33090 }, { "grad_norm": 0.22335553169250488, "learning_rate": 7.992307464497659e-05, "loss": 0.0143, "step": 33100 }, { "grad_norm": 0.21257537603378296, "learning_rate": 7.990982620960787e-05, "loss": 0.0123, "step": 33110 }, { "grad_norm": 0.1425086408853531, "learning_rate": 7.989657450334554e-05, "loss": 0.0118, "step": 33120 }, { "grad_norm": 0.15067432820796967, "learning_rate": 7.988331952763877e-05, "loss": 0.0144, "step": 33130 }, { "grad_norm": 0.24128805100917816, "learning_rate": 7.987006128393709e-05, "loss": 0.0128, "step": 33140 }, { "grad_norm": 0.2006748616695404, "learning_rate": 7.985679977369043e-05, "loss": 0.016, "step": 33150 }, { "grad_norm": 0.18504676222801208, "learning_rate": 7.984353499834902e-05, "loss": 0.0136, "step": 33160 }, { "grad_norm": 0.1933891326189041, "learning_rate": 7.983026695936351e-05, "loss": 0.0153, "step": 33170 }, { "grad_norm": 0.2528460621833801, "learning_rate": 7.981699565818486e-05, "loss": 0.0181, "step": 33180 }, { "grad_norm": 0.24825254082679749, "learning_rate": 7.980372109626437e-05, "loss": 0.0168, "step": 33190 }, { "grad_norm": 0.29120808839797974, "learning_rate": 7.979044327505375e-05, "loss": 0.0167, "step": 33200 }, { "grad_norm": 0.33964285254478455, "learning_rate": 7.977716219600506e-05, "loss": 0.0201, "step": 33210 }, { "grad_norm": 0.2705593407154083, "learning_rate": 7.97638778605707e-05, "loss": 0.0183, "step": 33220 }, { "grad_norm": 0.23823478817939758, "learning_rate": 7.975059027020338e-05, "loss": 0.0246, "step": 33230 }, { "grad_norm": 0.20452572405338287, "learning_rate": 7.973729942635623e-05, "loss": 0.0221, "step": 33240 }, { "grad_norm": 0.28355324268341064, "learning_rate": 7.972400533048273e-05, "loss": 0.0181, "step": 33250 }, { "grad_norm": 0.18344375491142273, "learning_rate": 7.97107079840367e-05, "loss": 0.02, "step": 33260 }, { "grad_norm": 0.18905992805957794, "learning_rate": 7.969740738847231e-05, "loss": 0.0166, "step": 33270 }, { "grad_norm": 0.19741664826869965, "learning_rate": 7.968410354524411e-05, "loss": 0.0146, "step": 33280 }, { "grad_norm": 0.1236821636557579, "learning_rate": 7.967079645580697e-05, "loss": 0.0165, "step": 33290 }, { "grad_norm": 0.1869746893644333, "learning_rate": 7.965748612161612e-05, "loss": 0.0143, "step": 33300 }, { "grad_norm": 0.19462406635284424, "learning_rate": 7.96441725441272e-05, "loss": 0.0153, "step": 33310 }, { "grad_norm": 0.2153601497411728, "learning_rate": 7.963085572479614e-05, "loss": 0.0153, "step": 33320 }, { "grad_norm": 0.23543189465999603, "learning_rate": 7.961753566507924e-05, "loss": 0.0157, "step": 33330 }, { "grad_norm": 0.25910091400146484, "learning_rate": 7.960421236643316e-05, "loss": 0.0154, "step": 33340 }, { "grad_norm": 0.19754132628440857, "learning_rate": 7.959088583031496e-05, "loss": 0.0182, "step": 33350 }, { "grad_norm": 0.23399482667446136, "learning_rate": 7.957755605818194e-05, "loss": 0.0148, "step": 33360 }, { "grad_norm": 0.2067166119813919, "learning_rate": 7.956422305149185e-05, "loss": 0.0128, "step": 33370 }, { "grad_norm": 0.14428424835205078, "learning_rate": 7.95508868117028e-05, "loss": 0.0127, "step": 33380 }, { "grad_norm": 0.2516128718852997, "learning_rate": 7.953754734027318e-05, "loss": 0.013, "step": 33390 }, { "grad_norm": 0.26221272349357605, "learning_rate": 7.952420463866182e-05, "loss": 0.0133, "step": 33400 }, { "grad_norm": 0.16925865411758423, "learning_rate": 7.951085870832782e-05, "loss": 0.0157, "step": 33410 }, { "grad_norm": 0.1761651635169983, "learning_rate": 7.949750955073067e-05, "loss": 0.0148, "step": 33420 }, { "grad_norm": 0.19259454309940338, "learning_rate": 7.948415716733022e-05, "loss": 0.0143, "step": 33430 }, { "grad_norm": 0.23488840460777283, "learning_rate": 7.947080155958669e-05, "loss": 0.0178, "step": 33440 }, { "grad_norm": 0.3298773467540741, "learning_rate": 7.94574427289606e-05, "loss": 0.0165, "step": 33450 }, { "grad_norm": 0.18988226354122162, "learning_rate": 7.944408067691284e-05, "loss": 0.0161, "step": 33460 }, { "grad_norm": 0.15651270747184753, "learning_rate": 7.943071540490473e-05, "loss": 0.0147, "step": 33470 }, { "grad_norm": 0.2586215138435364, "learning_rate": 7.94173469143978e-05, "loss": 0.0153, "step": 33480 }, { "grad_norm": 0.17230476438999176, "learning_rate": 7.940397520685406e-05, "loss": 0.016, "step": 33490 }, { "grad_norm": 0.19516991078853607, "learning_rate": 7.939060028373577e-05, "loss": 0.016, "step": 33500 }, { "grad_norm": 0.1868295669555664, "learning_rate": 7.937722214650565e-05, "loss": 0.0153, "step": 33510 }, { "grad_norm": 0.15680250525474548, "learning_rate": 7.936384079662666e-05, "loss": 0.0157, "step": 33520 }, { "grad_norm": 0.1375376135110855, "learning_rate": 7.93504562355622e-05, "loss": 0.0153, "step": 33530 }, { "grad_norm": 0.2590911388397217, "learning_rate": 7.933706846477599e-05, "loss": 0.0147, "step": 33540 }, { "grad_norm": 0.18257370591163635, "learning_rate": 7.932367748573206e-05, "loss": 0.0148, "step": 33550 }, { "grad_norm": 0.22045841813087463, "learning_rate": 7.931028329989485e-05, "loss": 0.0142, "step": 33560 }, { "grad_norm": 0.14214679598808289, "learning_rate": 7.929688590872913e-05, "loss": 0.0127, "step": 33570 }, { "grad_norm": 0.1772661805152893, "learning_rate": 7.928348531370003e-05, "loss": 0.0141, "step": 33580 }, { "grad_norm": 0.24867331981658936, "learning_rate": 7.927008151627297e-05, "loss": 0.0135, "step": 33590 }, { "grad_norm": 0.18208634853363037, "learning_rate": 7.925667451791383e-05, "loss": 0.0138, "step": 33600 }, { "grad_norm": 0.23483234643936157, "learning_rate": 7.924326432008874e-05, "loss": 0.0142, "step": 33610 }, { "grad_norm": 0.20520539581775665, "learning_rate": 7.922985092426422e-05, "loss": 0.0173, "step": 33620 }, { "grad_norm": 0.17959336936473846, "learning_rate": 7.921643433190717e-05, "loss": 0.0185, "step": 33630 }, { "grad_norm": 0.23940575122833252, "learning_rate": 7.920301454448478e-05, "loss": 0.0154, "step": 33640 }, { "grad_norm": 0.16268165409564972, "learning_rate": 7.918959156346461e-05, "loss": 0.0169, "step": 33650 }, { "grad_norm": 0.25019028782844543, "learning_rate": 7.91761653903146e-05, "loss": 0.014, "step": 33660 }, { "grad_norm": 0.24057146906852722, "learning_rate": 7.916273602650302e-05, "loss": 0.0155, "step": 33670 }, { "grad_norm": 0.36964184045791626, "learning_rate": 7.914930347349847e-05, "loss": 0.0183, "step": 33680 }, { "grad_norm": 0.1976652890443802, "learning_rate": 7.913586773276992e-05, "loss": 0.0166, "step": 33690 }, { "grad_norm": 0.5858215689659119, "learning_rate": 7.912242880578667e-05, "loss": 0.0235, "step": 33700 }, { "grad_norm": 0.6809754967689514, "learning_rate": 7.910898669401839e-05, "loss": 0.018, "step": 33710 }, { "grad_norm": 0.21206794679164886, "learning_rate": 7.909554139893511e-05, "loss": 0.0175, "step": 33720 }, { "grad_norm": 0.21229664981365204, "learning_rate": 7.908209292200715e-05, "loss": 0.0162, "step": 33730 }, { "grad_norm": 0.27434253692626953, "learning_rate": 7.906864126470523e-05, "loss": 0.0165, "step": 33740 }, { "grad_norm": 0.16323935985565186, "learning_rate": 7.905518642850041e-05, "loss": 0.0153, "step": 33750 }, { "grad_norm": 0.3180159628391266, "learning_rate": 7.904172841486409e-05, "loss": 0.0157, "step": 33760 }, { "grad_norm": 0.21458293497562408, "learning_rate": 7.902826722526801e-05, "loss": 0.0138, "step": 33770 }, { "grad_norm": 0.15874379873275757, "learning_rate": 7.901480286118427e-05, "loss": 0.0184, "step": 33780 }, { "grad_norm": 0.25391069054603577, "learning_rate": 7.900133532408531e-05, "loss": 0.0162, "step": 33790 }, { "grad_norm": 0.16717834770679474, "learning_rate": 7.898786461544395e-05, "loss": 0.0171, "step": 33800 }, { "grad_norm": 0.2304871529340744, "learning_rate": 7.897439073673325e-05, "loss": 0.0159, "step": 33810 }, { "grad_norm": 0.18988870084285736, "learning_rate": 7.896091368942677e-05, "loss": 0.0143, "step": 33820 }, { "grad_norm": 0.18577750027179718, "learning_rate": 7.894743347499832e-05, "loss": 0.0142, "step": 33830 }, { "grad_norm": 0.18397323787212372, "learning_rate": 7.893395009492203e-05, "loss": 0.0157, "step": 33840 }, { "grad_norm": 0.1585904210805893, "learning_rate": 7.892046355067248e-05, "loss": 0.0158, "step": 33850 }, { "grad_norm": 0.1556302011013031, "learning_rate": 7.890697384372451e-05, "loss": 0.0153, "step": 33860 }, { "grad_norm": 0.15755967795848846, "learning_rate": 7.889348097555336e-05, "loss": 0.0164, "step": 33870 }, { "grad_norm": 0.21415413916110992, "learning_rate": 7.887998494763455e-05, "loss": 0.0139, "step": 33880 }, { "grad_norm": 0.18773198127746582, "learning_rate": 7.886648576144404e-05, "loss": 0.0157, "step": 33890 }, { "grad_norm": 0.22406244277954102, "learning_rate": 7.885298341845802e-05, "loss": 0.0161, "step": 33900 }, { "grad_norm": 0.14831411838531494, "learning_rate": 7.883947792015311e-05, "loss": 0.0134, "step": 33910 }, { "grad_norm": 0.1731807142496109, "learning_rate": 7.882596926800628e-05, "loss": 0.0127, "step": 33920 }, { "grad_norm": 0.32257989048957825, "learning_rate": 7.881245746349477e-05, "loss": 0.0129, "step": 33930 }, { "grad_norm": 0.17282743752002716, "learning_rate": 7.879894250809623e-05, "loss": 0.0157, "step": 33940 }, { "grad_norm": 0.2691267430782318, "learning_rate": 7.878542440328865e-05, "loss": 0.0149, "step": 33950 }, { "grad_norm": 0.24329333007335663, "learning_rate": 7.877190315055031e-05, "loss": 0.0153, "step": 33960 }, { "grad_norm": 0.18319915235042572, "learning_rate": 7.875837875135991e-05, "loss": 0.0136, "step": 33970 }, { "grad_norm": 0.12814858555793762, "learning_rate": 7.874485120719646e-05, "loss": 0.0158, "step": 33980 }, { "grad_norm": 0.12879590690135956, "learning_rate": 7.873132051953928e-05, "loss": 0.0127, "step": 33990 }, { "grad_norm": 0.21200688183307648, "learning_rate": 7.87177866898681e-05, "loss": 0.0134, "step": 34000 }, { "grad_norm": 0.23141524195671082, "learning_rate": 7.870424971966294e-05, "loss": 0.0152, "step": 34010 }, { "grad_norm": 0.19673915207386017, "learning_rate": 7.869070961040419e-05, "loss": 0.0134, "step": 34020 }, { "grad_norm": 0.15452761948108673, "learning_rate": 7.867716636357257e-05, "loss": 0.0158, "step": 34030 }, { "grad_norm": 0.16540107131004333, "learning_rate": 7.866361998064915e-05, "loss": 0.0153, "step": 34040 }, { "grad_norm": 0.22345033288002014, "learning_rate": 7.865007046311534e-05, "loss": 0.0169, "step": 34050 }, { "grad_norm": 0.20118285715579987, "learning_rate": 7.86365178124529e-05, "loss": 0.0148, "step": 34060 }, { "grad_norm": 0.26105716824531555, "learning_rate": 7.862296203014394e-05, "loss": 0.012, "step": 34070 }, { "grad_norm": 0.17803475260734558, "learning_rate": 7.860940311767088e-05, "loss": 0.0139, "step": 34080 }, { "grad_norm": 0.1803297996520996, "learning_rate": 7.85958410765165e-05, "loss": 0.0155, "step": 34090 }, { "grad_norm": 0.18221443891525269, "learning_rate": 7.858227590816394e-05, "loss": 0.0137, "step": 34100 }, { "grad_norm": 0.19159075617790222, "learning_rate": 7.856870761409664e-05, "loss": 0.0139, "step": 34110 }, { "grad_norm": 0.25666919350624084, "learning_rate": 7.855513619579846e-05, "loss": 0.0127, "step": 34120 }, { "grad_norm": 0.19441914558410645, "learning_rate": 7.85415616547535e-05, "loss": 0.0142, "step": 34130 }, { "grad_norm": 0.29338446259498596, "learning_rate": 7.852798399244627e-05, "loss": 0.0156, "step": 34140 }, { "grad_norm": 0.2039041817188263, "learning_rate": 7.851440321036161e-05, "loss": 0.0127, "step": 34150 }, { "grad_norm": 0.12942245602607727, "learning_rate": 7.850081930998468e-05, "loss": 0.0133, "step": 34160 }, { "grad_norm": 0.3120049834251404, "learning_rate": 7.8487232292801e-05, "loss": 0.0128, "step": 34170 }, { "grad_norm": 0.1762479692697525, "learning_rate": 7.847364216029642e-05, "loss": 0.0147, "step": 34180 }, { "grad_norm": 0.19836385548114777, "learning_rate": 7.846004891395716e-05, "loss": 0.0138, "step": 34190 }, { "grad_norm": 0.19334137439727783, "learning_rate": 7.844645255526972e-05, "loss": 0.0169, "step": 34200 }, { "grad_norm": 0.1739654690027237, "learning_rate": 7.843285308572101e-05, "loss": 0.0159, "step": 34210 }, { "grad_norm": 0.15824176371097565, "learning_rate": 7.841925050679823e-05, "loss": 0.014, "step": 34220 }, { "grad_norm": 0.1493171900510788, "learning_rate": 7.840564481998895e-05, "loss": 0.0141, "step": 34230 }, { "grad_norm": 0.16635778546333313, "learning_rate": 7.839203602678105e-05, "loss": 0.0118, "step": 34240 }, { "grad_norm": 0.1976456493139267, "learning_rate": 7.837842412866279e-05, "loss": 0.013, "step": 34250 }, { "grad_norm": 0.1518588662147522, "learning_rate": 7.836480912712272e-05, "loss": 0.0133, "step": 34260 }, { "grad_norm": 0.16842970252037048, "learning_rate": 7.835119102364976e-05, "loss": 0.0132, "step": 34270 }, { "grad_norm": 0.17714303731918335, "learning_rate": 7.83375698197332e-05, "loss": 0.0146, "step": 34280 }, { "grad_norm": 0.18184487521648407, "learning_rate": 7.83239455168626e-05, "loss": 0.0145, "step": 34290 }, { "grad_norm": 0.1788043975830078, "learning_rate": 7.83103181165279e-05, "loss": 0.0155, "step": 34300 }, { "grad_norm": 0.13596096634864807, "learning_rate": 7.829668762021937e-05, "loss": 0.0161, "step": 34310 }, { "grad_norm": 0.20826348662376404, "learning_rate": 7.828305402942764e-05, "loss": 0.0151, "step": 34320 }, { "grad_norm": 0.1793394386768341, "learning_rate": 7.826941734564363e-05, "loss": 0.0159, "step": 34330 }, { "grad_norm": 0.24528057873249054, "learning_rate": 7.825577757035865e-05, "loss": 0.0154, "step": 34340 }, { "grad_norm": 0.24463903903961182, "learning_rate": 7.824213470506431e-05, "loss": 0.0164, "step": 34350 }, { "grad_norm": 0.18951715528964996, "learning_rate": 7.822848875125257e-05, "loss": 0.0137, "step": 34360 }, { "grad_norm": 0.19295568764209747, "learning_rate": 7.821483971041576e-05, "loss": 0.0135, "step": 34370 }, { "grad_norm": 0.1736035943031311, "learning_rate": 7.820118758404649e-05, "loss": 0.0131, "step": 34380 }, { "grad_norm": 0.2107958048582077, "learning_rate": 7.818753237363776e-05, "loss": 0.014, "step": 34390 }, { "grad_norm": 0.25267404317855835, "learning_rate": 7.817387408068286e-05, "loss": 0.0157, "step": 34400 }, { "grad_norm": 0.2177947461605072, "learning_rate": 7.816021270667544e-05, "loss": 0.0175, "step": 34410 }, { "grad_norm": 0.23910246789455414, "learning_rate": 7.81465482531095e-05, "loss": 0.0157, "step": 34420 }, { "grad_norm": 0.22380590438842773, "learning_rate": 7.813288072147938e-05, "loss": 0.0139, "step": 34430 }, { "grad_norm": 0.20902584493160248, "learning_rate": 7.811921011327972e-05, "loss": 0.0169, "step": 34440 }, { "grad_norm": 0.21146079897880554, "learning_rate": 7.810553643000549e-05, "loss": 0.0149, "step": 34450 }, { "grad_norm": 0.20444819331169128, "learning_rate": 7.809185967315206e-05, "loss": 0.0123, "step": 34460 }, { "grad_norm": 0.19201615452766418, "learning_rate": 7.80781798442151e-05, "loss": 0.0142, "step": 34470 }, { "grad_norm": 0.23271435499191284, "learning_rate": 7.806449694469061e-05, "loss": 0.0139, "step": 34480 }, { "grad_norm": 0.1349760890007019, "learning_rate": 7.805081097607492e-05, "loss": 0.0122, "step": 34490 }, { "grad_norm": 0.18802577257156372, "learning_rate": 7.803712193986474e-05, "loss": 0.0161, "step": 34500 }, { "grad_norm": 0.22999636828899384, "learning_rate": 7.802342983755702e-05, "loss": 0.0149, "step": 34510 }, { "grad_norm": 0.2187061905860901, "learning_rate": 7.800973467064918e-05, "loss": 0.0134, "step": 34520 }, { "grad_norm": 0.20488715171813965, "learning_rate": 7.799603644063884e-05, "loss": 0.0156, "step": 34530 }, { "grad_norm": 0.17300809919834137, "learning_rate": 7.798233514902405e-05, "loss": 0.0171, "step": 34540 }, { "grad_norm": 0.16967755556106567, "learning_rate": 7.796863079730318e-05, "loss": 0.0165, "step": 34550 }, { "grad_norm": 0.20722155272960663, "learning_rate": 7.795492338697488e-05, "loss": 0.0144, "step": 34560 }, { "grad_norm": 0.1471472531557083, "learning_rate": 7.794121291953819e-05, "loss": 0.0124, "step": 34570 }, { "grad_norm": 0.16564251482486725, "learning_rate": 7.792749939649246e-05, "loss": 0.0137, "step": 34580 }, { "grad_norm": 0.1801975965499878, "learning_rate": 7.79137828193374e-05, "loss": 0.0171, "step": 34590 }, { "grad_norm": 0.18947437405586243, "learning_rate": 7.790006318957301e-05, "loss": 0.0146, "step": 34600 }, { "grad_norm": 0.1978093981742859, "learning_rate": 7.788634050869965e-05, "loss": 0.0143, "step": 34610 }, { "grad_norm": 0.16695286333560944, "learning_rate": 7.787261477821803e-05, "loss": 0.0127, "step": 34620 }, { "grad_norm": 0.17449654638767242, "learning_rate": 7.785888599962916e-05, "loss": 0.0127, "step": 34630 }, { "grad_norm": 0.16676296293735504, "learning_rate": 7.784515417443439e-05, "loss": 0.0133, "step": 34640 }, { "grad_norm": 0.25771084427833557, "learning_rate": 7.783141930413545e-05, "loss": 0.0138, "step": 34650 }, { "grad_norm": 0.31519657373428345, "learning_rate": 7.78176813902343e-05, "loss": 0.0125, "step": 34660 }, { "grad_norm": 0.1796058863401413, "learning_rate": 7.780394043423336e-05, "loss": 0.0131, "step": 34670 }, { "grad_norm": 0.16969798505306244, "learning_rate": 7.77901964376353e-05, "loss": 0.0129, "step": 34680 }, { "grad_norm": 0.13761882483959198, "learning_rate": 7.777644940194316e-05, "loss": 0.0117, "step": 34690 }, { "grad_norm": 0.23583875596523285, "learning_rate": 7.776269932866023e-05, "loss": 0.0135, "step": 34700 }, { "grad_norm": 0.19776272773742676, "learning_rate": 7.774894621929026e-05, "loss": 0.013, "step": 34710 }, { "grad_norm": 0.23025108873844147, "learning_rate": 7.773519007533725e-05, "loss": 0.0118, "step": 34720 }, { "grad_norm": 0.2432890683412552, "learning_rate": 7.772143089830556e-05, "loss": 0.0122, "step": 34730 }, { "grad_norm": 0.20800507068634033, "learning_rate": 7.770766868969985e-05, "loss": 0.0142, "step": 34740 }, { "grad_norm": 0.2522314488887787, "learning_rate": 7.769390345102518e-05, "loss": 0.0145, "step": 34750 }, { "grad_norm": 0.24823753535747528, "learning_rate": 7.768013518378683e-05, "loss": 0.0163, "step": 34760 }, { "grad_norm": 0.22926147282123566, "learning_rate": 7.766636388949053e-05, "loss": 0.0146, "step": 34770 }, { "grad_norm": 0.2013798952102661, "learning_rate": 7.765258956964229e-05, "loss": 0.0149, "step": 34780 }, { "grad_norm": 0.17547816038131714, "learning_rate": 7.76388122257484e-05, "loss": 0.0145, "step": 34790 }, { "grad_norm": 0.2115090936422348, "learning_rate": 7.762503185931558e-05, "loss": 0.013, "step": 34800 }, { "grad_norm": 0.1842299997806549, "learning_rate": 7.76112484718508e-05, "loss": 0.0151, "step": 34810 }, { "grad_norm": 0.1449650377035141, "learning_rate": 7.75974620648614e-05, "loss": 0.0134, "step": 34820 }, { "grad_norm": 0.27147865295410156, "learning_rate": 7.758367263985503e-05, "loss": 0.0137, "step": 34830 }, { "grad_norm": 0.16380546987056732, "learning_rate": 7.75698801983397e-05, "loss": 0.0129, "step": 34840 }, { "grad_norm": 0.15936067700386047, "learning_rate": 7.755608474182372e-05, "loss": 0.0149, "step": 34850 }, { "grad_norm": 0.20782676339149475, "learning_rate": 7.754228627181574e-05, "loss": 0.0144, "step": 34860 }, { "grad_norm": 0.15562213957309723, "learning_rate": 7.752848478982476e-05, "loss": 0.0128, "step": 34870 }, { "grad_norm": 0.20332540571689606, "learning_rate": 7.751468029736006e-05, "loss": 0.0143, "step": 34880 }, { "grad_norm": 0.20690536499023438, "learning_rate": 7.750087279593129e-05, "loss": 0.0131, "step": 34890 }, { "grad_norm": 0.20968757569789886, "learning_rate": 7.748706228704843e-05, "loss": 0.014, "step": 34900 }, { "grad_norm": 0.23159454762935638, "learning_rate": 7.747324877222176e-05, "loss": 0.0134, "step": 34910 }, { "grad_norm": 0.17180851101875305, "learning_rate": 7.745943225296188e-05, "loss": 0.0132, "step": 34920 }, { "grad_norm": 0.1459161341190338, "learning_rate": 7.744561273077981e-05, "loss": 0.0124, "step": 34930 }, { "grad_norm": 0.1771521121263504, "learning_rate": 7.743179020718678e-05, "loss": 0.0135, "step": 34940 }, { "grad_norm": 0.27416831254959106, "learning_rate": 7.741796468369443e-05, "loss": 0.014, "step": 34950 }, { "grad_norm": 0.21191823482513428, "learning_rate": 7.740413616181466e-05, "loss": 0.0123, "step": 34960 }, { "grad_norm": 0.20002329349517822, "learning_rate": 7.739030464305978e-05, "loss": 0.0143, "step": 34970 }, { "grad_norm": 0.22960147261619568, "learning_rate": 7.737647012894235e-05, "loss": 0.0133, "step": 34980 }, { "grad_norm": 0.24806782603263855, "learning_rate": 7.736263262097532e-05, "loss": 0.0146, "step": 34990 }, { "grad_norm": 0.16941063106060028, "learning_rate": 7.734879212067192e-05, "loss": 0.0128, "step": 35000 }, { "grad_norm": 0.18622495234012604, "learning_rate": 7.733494862954573e-05, "loss": 0.0123, "step": 35010 }, { "grad_norm": 0.15668419003486633, "learning_rate": 7.732110214911066e-05, "loss": 0.0145, "step": 35020 }, { "grad_norm": 0.1540497988462448, "learning_rate": 7.730725268088094e-05, "loss": 0.0155, "step": 35030 }, { "grad_norm": 0.14644695818424225, "learning_rate": 7.729340022637112e-05, "loss": 0.0123, "step": 35040 }, { "grad_norm": 0.1571272313594818, "learning_rate": 7.727954478709607e-05, "loss": 0.0151, "step": 35050 }, { "grad_norm": 0.18562033772468567, "learning_rate": 7.726568636457103e-05, "loss": 0.013, "step": 35060 }, { "grad_norm": 0.17446020245552063, "learning_rate": 7.725182496031153e-05, "loss": 0.014, "step": 35070 }, { "grad_norm": 0.18405038118362427, "learning_rate": 7.723796057583341e-05, "loss": 0.0126, "step": 35080 }, { "grad_norm": 0.2189495712518692, "learning_rate": 7.722409321265287e-05, "loss": 0.0136, "step": 35090 }, { "grad_norm": 0.236107736825943, "learning_rate": 7.721022287228645e-05, "loss": 0.0131, "step": 35100 }, { "grad_norm": 0.16556601226329803, "learning_rate": 7.719634955625094e-05, "loss": 0.0123, "step": 35110 }, { "grad_norm": 0.18848241865634918, "learning_rate": 7.718247326606355e-05, "loss": 0.0137, "step": 35120 }, { "grad_norm": 0.19402426481246948, "learning_rate": 7.716859400324175e-05, "loss": 0.0121, "step": 35130 }, { "grad_norm": 0.14631222188472748, "learning_rate": 7.715471176930335e-05, "loss": 0.0122, "step": 35140 }, { "grad_norm": 0.14765478670597076, "learning_rate": 7.714082656576651e-05, "loss": 0.012, "step": 35150 }, { "grad_norm": 0.19518688321113586, "learning_rate": 7.712693839414968e-05, "loss": 0.0152, "step": 35160 }, { "grad_norm": 0.1539297252893448, "learning_rate": 7.711304725597164e-05, "loss": 0.0128, "step": 35170 }, { "grad_norm": 0.20842717587947845, "learning_rate": 7.709915315275151e-05, "loss": 0.0129, "step": 35180 }, { "grad_norm": 0.3102170526981354, "learning_rate": 7.708525608600876e-05, "loss": 0.0121, "step": 35190 }, { "grad_norm": 0.20488302409648895, "learning_rate": 7.707135605726311e-05, "loss": 0.0122, "step": 35200 }, { "grad_norm": 0.1770157814025879, "learning_rate": 7.705745306803466e-05, "loss": 0.0148, "step": 35210 }, { "grad_norm": 0.2148047387599945, "learning_rate": 7.704354711984383e-05, "loss": 0.0131, "step": 35220 }, { "grad_norm": 0.1518579125404358, "learning_rate": 7.702963821421133e-05, "loss": 0.0143, "step": 35230 }, { "grad_norm": 0.17083431780338287, "learning_rate": 7.701572635265826e-05, "loss": 0.0137, "step": 35240 }, { "grad_norm": 0.18047122657299042, "learning_rate": 7.700181153670596e-05, "loss": 0.0126, "step": 35250 }, { "grad_norm": 0.23003476858139038, "learning_rate": 7.698789376787616e-05, "loss": 0.0171, "step": 35260 }, { "grad_norm": 0.3273559808731079, "learning_rate": 7.697397304769085e-05, "loss": 0.0162, "step": 35270 }, { "grad_norm": 0.19566710293293, "learning_rate": 7.696004937767241e-05, "loss": 0.0188, "step": 35280 }, { "grad_norm": 0.2320820689201355, "learning_rate": 7.694612275934352e-05, "loss": 0.016, "step": 35290 }, { "grad_norm": 0.263996422290802, "learning_rate": 7.693219319422714e-05, "loss": 0.0163, "step": 35300 }, { "grad_norm": 0.1696343570947647, "learning_rate": 7.69182606838466e-05, "loss": 0.014, "step": 35310 }, { "grad_norm": 0.22061039507389069, "learning_rate": 7.690432522972558e-05, "loss": 0.015, "step": 35320 }, { "grad_norm": 0.19745665788650513, "learning_rate": 7.689038683338796e-05, "loss": 0.0152, "step": 35330 }, { "grad_norm": 0.2337128072977066, "learning_rate": 7.687644549635808e-05, "loss": 0.0135, "step": 35340 }, { "grad_norm": 0.17943723499774933, "learning_rate": 7.686250122016053e-05, "loss": 0.013, "step": 35350 }, { "grad_norm": 0.307399719953537, "learning_rate": 7.684855400632022e-05, "loss": 0.014, "step": 35360 }, { "grad_norm": 0.1696184277534485, "learning_rate": 7.683460385636243e-05, "loss": 0.0162, "step": 35370 }, { "grad_norm": 0.1605580449104309, "learning_rate": 7.68206507718127e-05, "loss": 0.0138, "step": 35380 }, { "grad_norm": 0.20956121385097504, "learning_rate": 7.680669475419693e-05, "loss": 0.0132, "step": 35390 }, { "grad_norm": 0.19914864003658295, "learning_rate": 7.679273580504132e-05, "loss": 0.0154, "step": 35400 }, { "grad_norm": 0.15766380727291107, "learning_rate": 7.67787739258724e-05, "loss": 0.0146, "step": 35410 }, { "grad_norm": 0.15518896281719208, "learning_rate": 7.676480911821705e-05, "loss": 0.0145, "step": 35420 }, { "grad_norm": 0.14571307599544525, "learning_rate": 7.675084138360238e-05, "loss": 0.0151, "step": 35430 }, { "grad_norm": 0.12657438218593597, "learning_rate": 7.673687072355592e-05, "loss": 0.0138, "step": 35440 }, { "grad_norm": 0.18883252143859863, "learning_rate": 7.67228971396055e-05, "loss": 0.0141, "step": 35450 }, { "grad_norm": 0.12579850852489471, "learning_rate": 7.670892063327922e-05, "loss": 0.0133, "step": 35460 }, { "grad_norm": 0.21976718306541443, "learning_rate": 7.669494120610552e-05, "loss": 0.0133, "step": 35470 }, { "grad_norm": 0.2827867865562439, "learning_rate": 7.66809588596132e-05, "loss": 0.0128, "step": 35480 }, { "grad_norm": 0.20585548877716064, "learning_rate": 7.666697359533132e-05, "loss": 0.0131, "step": 35490 }, { "grad_norm": 0.20248430967330933, "learning_rate": 7.665298541478932e-05, "loss": 0.0137, "step": 35500 }, { "grad_norm": 0.17564058303833008, "learning_rate": 7.66389943195169e-05, "loss": 0.0133, "step": 35510 }, { "grad_norm": 0.1782148778438568, "learning_rate": 7.66250003110441e-05, "loss": 0.0149, "step": 35520 }, { "grad_norm": 0.20466995239257812, "learning_rate": 7.66110033909013e-05, "loss": 0.0132, "step": 35530 }, { "grad_norm": 0.21285226941108704, "learning_rate": 7.659700356061918e-05, "loss": 0.0141, "step": 35540 }, { "grad_norm": 0.22474315762519836, "learning_rate": 7.658300082172875e-05, "loss": 0.0126, "step": 35550 }, { "grad_norm": 0.21748706698417664, "learning_rate": 7.65689951757613e-05, "loss": 0.0144, "step": 35560 }, { "grad_norm": 0.15691937506198883, "learning_rate": 7.655498662424849e-05, "loss": 0.0118, "step": 35570 }, { "grad_norm": 0.15498945116996765, "learning_rate": 7.654097516872227e-05, "loss": 0.0141, "step": 35580 }, { "grad_norm": 0.24915456771850586, "learning_rate": 7.65269608107149e-05, "loss": 0.0132, "step": 35590 }, { "grad_norm": 0.19062887132167816, "learning_rate": 7.6512943551759e-05, "loss": 0.0146, "step": 35600 }, { "grad_norm": 0.17206108570098877, "learning_rate": 7.649892339338743e-05, "loss": 0.0134, "step": 35610 }, { "grad_norm": 0.22871649265289307, "learning_rate": 7.648490033713344e-05, "loss": 0.0125, "step": 35620 }, { "grad_norm": 0.23073045909404755, "learning_rate": 7.647087438453058e-05, "loss": 0.0126, "step": 35630 }, { "grad_norm": 0.14398863911628723, "learning_rate": 7.64568455371127e-05, "loss": 0.0146, "step": 35640 }, { "grad_norm": 0.14060743153095245, "learning_rate": 7.644281379641396e-05, "loss": 0.0132, "step": 35650 }, { "grad_norm": 0.19743074476718903, "learning_rate": 7.642877916396887e-05, "loss": 0.0133, "step": 35660 }, { "grad_norm": 0.15640316903591156, "learning_rate": 7.641474164131221e-05, "loss": 0.0144, "step": 35670 }, { "grad_norm": 0.19033971428871155, "learning_rate": 7.640070122997913e-05, "loss": 0.0131, "step": 35680 }, { "grad_norm": 0.1989922821521759, "learning_rate": 7.638665793150505e-05, "loss": 0.0133, "step": 35690 }, { "grad_norm": 0.19741320610046387, "learning_rate": 7.637261174742574e-05, "loss": 0.0123, "step": 35700 }, { "grad_norm": 0.16923409700393677, "learning_rate": 7.635856267927725e-05, "loss": 0.0133, "step": 35710 }, { "grad_norm": 0.1619126945734024, "learning_rate": 7.634451072859597e-05, "loss": 0.013, "step": 35720 }, { "grad_norm": 0.20175644755363464, "learning_rate": 7.633045589691863e-05, "loss": 0.0111, "step": 35730 }, { "grad_norm": 0.24163438379764557, "learning_rate": 7.63163981857822e-05, "loss": 0.0142, "step": 35740 }, { "grad_norm": 0.19356048107147217, "learning_rate": 7.630233759672403e-05, "loss": 0.0129, "step": 35750 }, { "grad_norm": 0.2329729199409485, "learning_rate": 7.628827413128177e-05, "loss": 0.0123, "step": 35760 }, { "grad_norm": 0.1591929793357849, "learning_rate": 7.627420779099338e-05, "loss": 0.0133, "step": 35770 }, { "grad_norm": 0.14876911044120789, "learning_rate": 7.626013857739711e-05, "loss": 0.0121, "step": 35780 }, { "grad_norm": 0.1577405482530594, "learning_rate": 7.624606649203158e-05, "loss": 0.0117, "step": 35790 }, { "grad_norm": 0.24563764035701752, "learning_rate": 7.623199153643569e-05, "loss": 0.0122, "step": 35800 }, { "grad_norm": 0.19358424842357635, "learning_rate": 7.621791371214863e-05, "loss": 0.0133, "step": 35810 }, { "grad_norm": 0.2235284149646759, "learning_rate": 7.620383302070995e-05, "loss": 0.0135, "step": 35820 }, { "grad_norm": 0.19260407984256744, "learning_rate": 7.61897494636595e-05, "loss": 0.0118, "step": 35830 }, { "grad_norm": 0.21678569912910461, "learning_rate": 7.617566304253739e-05, "loss": 0.0135, "step": 35840 }, { "grad_norm": 0.15535801649093628, "learning_rate": 7.616157375888416e-05, "loss": 0.0141, "step": 35850 }, { "grad_norm": 0.15573091804981232, "learning_rate": 7.614748161424053e-05, "loss": 0.0117, "step": 35860 }, { "grad_norm": 0.2781222462654114, "learning_rate": 7.613338661014763e-05, "loss": 0.0148, "step": 35870 }, { "grad_norm": 0.13393045961856842, "learning_rate": 7.611928874814686e-05, "loss": 0.011, "step": 35880 }, { "grad_norm": 0.1509118378162384, "learning_rate": 7.610518802977996e-05, "loss": 0.0119, "step": 35890 }, { "grad_norm": 0.17695929110050201, "learning_rate": 7.609108445658893e-05, "loss": 0.0138, "step": 35900 }, { "grad_norm": 0.1834191083908081, "learning_rate": 7.607697803011612e-05, "loss": 0.013, "step": 35910 }, { "grad_norm": 0.18712902069091797, "learning_rate": 7.606286875190421e-05, "loss": 0.0145, "step": 35920 }, { "grad_norm": 0.1574079692363739, "learning_rate": 7.604875662349617e-05, "loss": 0.0131, "step": 35930 }, { "grad_norm": 0.1949128359556198, "learning_rate": 7.603464164643525e-05, "loss": 0.0116, "step": 35940 }, { "grad_norm": 0.2158299833536148, "learning_rate": 7.602052382226507e-05, "loss": 0.0137, "step": 35950 }, { "grad_norm": 0.2136075347661972, "learning_rate": 7.600640315252954e-05, "loss": 0.0111, "step": 35960 }, { "grad_norm": 0.3346234858036041, "learning_rate": 7.599227963877284e-05, "loss": 0.0142, "step": 35970 }, { "grad_norm": 0.23756848275661469, "learning_rate": 7.597815328253954e-05, "loss": 0.0149, "step": 35980 }, { "grad_norm": 0.18811781704425812, "learning_rate": 7.596402408537443e-05, "loss": 0.0124, "step": 35990 }, { "grad_norm": 0.17611804604530334, "learning_rate": 7.59498920488227e-05, "loss": 0.0148, "step": 36000 }, { "grad_norm": 0.18040861189365387, "learning_rate": 7.593575717442979e-05, "loss": 0.0116, "step": 36010 }, { "grad_norm": 0.1546013206243515, "learning_rate": 7.592161946374147e-05, "loss": 0.0128, "step": 36020 }, { "grad_norm": 0.26923856139183044, "learning_rate": 7.590747891830381e-05, "loss": 0.0127, "step": 36030 }, { "grad_norm": 0.1620236337184906, "learning_rate": 7.58933355396632e-05, "loss": 0.0131, "step": 36040 }, { "grad_norm": 0.1645830124616623, "learning_rate": 7.587918932936636e-05, "loss": 0.0131, "step": 36050 }, { "grad_norm": 0.170966237783432, "learning_rate": 7.586504028896028e-05, "loss": 0.0137, "step": 36060 }, { "grad_norm": 0.20277638733386993, "learning_rate": 7.585088841999228e-05, "loss": 0.0143, "step": 36070 }, { "grad_norm": 0.19828148186206818, "learning_rate": 7.583673372400999e-05, "loss": 0.0138, "step": 36080 }, { "grad_norm": 0.38757917284965515, "learning_rate": 7.582257620256134e-05, "loss": 0.0131, "step": 36090 }, { "grad_norm": 0.20047858357429504, "learning_rate": 7.580841585719458e-05, "loss": 0.013, "step": 36100 }, { "grad_norm": 0.3489008843898773, "learning_rate": 7.579425268945825e-05, "loss": 0.0132, "step": 36110 }, { "grad_norm": 0.18059225380420685, "learning_rate": 7.578008670090127e-05, "loss": 0.0134, "step": 36120 }, { "grad_norm": 0.16144877672195435, "learning_rate": 7.576591789307272e-05, "loss": 0.0125, "step": 36130 }, { "grad_norm": 0.15588955581188202, "learning_rate": 7.575174626752216e-05, "loss": 0.0125, "step": 36140 }, { "grad_norm": 0.30751681327819824, "learning_rate": 7.573757182579934e-05, "loss": 0.0123, "step": 36150 }, { "grad_norm": 0.16683432459831238, "learning_rate": 7.572339456945435e-05, "loss": 0.0126, "step": 36160 }, { "grad_norm": 0.201649472117424, "learning_rate": 7.570921450003762e-05, "loss": 0.0142, "step": 36170 }, { "grad_norm": 0.15452010929584503, "learning_rate": 7.569503161909983e-05, "loss": 0.0139, "step": 36180 }, { "grad_norm": 0.14769063889980316, "learning_rate": 7.568084592819202e-05, "loss": 0.0127, "step": 36190 }, { "grad_norm": 0.3185429871082306, "learning_rate": 7.566665742886551e-05, "loss": 0.0144, "step": 36200 }, { "grad_norm": 0.1414683610200882, "learning_rate": 7.565246612267194e-05, "loss": 0.014, "step": 36210 }, { "grad_norm": 0.19957002997398376, "learning_rate": 7.563827201116325e-05, "loss": 0.0131, "step": 36220 }, { "grad_norm": 0.16913489997386932, "learning_rate": 7.562407509589164e-05, "loss": 0.0128, "step": 36230 }, { "grad_norm": 0.23176276683807373, "learning_rate": 7.560987537840973e-05, "loss": 0.015, "step": 36240 }, { "grad_norm": 0.22687262296676636, "learning_rate": 7.559567286027036e-05, "loss": 0.0142, "step": 36250 }, { "grad_norm": 0.15549613535404205, "learning_rate": 7.558146754302668e-05, "loss": 0.0125, "step": 36260 }, { "grad_norm": 0.23016607761383057, "learning_rate": 7.556725942823217e-05, "loss": 0.0151, "step": 36270 }, { "grad_norm": 0.24894455075263977, "learning_rate": 7.555304851744061e-05, "loss": 0.0141, "step": 36280 }, { "grad_norm": 0.15683510899543762, "learning_rate": 7.553883481220608e-05, "loss": 0.0135, "step": 36290 }, { "grad_norm": 0.21749356389045715, "learning_rate": 7.552461831408298e-05, "loss": 0.0126, "step": 36300 }, { "grad_norm": 0.21102780103683472, "learning_rate": 7.551039902462599e-05, "loss": 0.0147, "step": 36310 }, { "grad_norm": 0.20130513608455658, "learning_rate": 7.549617694539011e-05, "loss": 0.0125, "step": 36320 }, { "grad_norm": 0.16470742225646973, "learning_rate": 7.548195207793066e-05, "loss": 0.0142, "step": 36330 }, { "grad_norm": 0.1342773288488388, "learning_rate": 7.546772442380323e-05, "loss": 0.011, "step": 36340 }, { "grad_norm": 0.14572851359844208, "learning_rate": 7.545349398456376e-05, "loss": 0.0134, "step": 36350 }, { "grad_norm": 0.15562695264816284, "learning_rate": 7.543926076176845e-05, "loss": 0.0113, "step": 36360 }, { "grad_norm": 0.2792660593986511, "learning_rate": 7.542502475697385e-05, "loss": 0.0147, "step": 36370 }, { "grad_norm": 0.20800018310546875, "learning_rate": 7.541078597173675e-05, "loss": 0.0142, "step": 36380 }, { "grad_norm": 0.19783884286880493, "learning_rate": 7.539654440761431e-05, "loss": 0.0172, "step": 36390 }, { "grad_norm": 0.2183620035648346, "learning_rate": 7.538230006616395e-05, "loss": 0.0157, "step": 36400 }, { "grad_norm": 0.20623691380023956, "learning_rate": 7.536805294894342e-05, "loss": 0.0136, "step": 36410 }, { "grad_norm": 0.17340625822544098, "learning_rate": 7.535380305751076e-05, "loss": 0.0139, "step": 36420 }, { "grad_norm": 0.18789611756801605, "learning_rate": 7.533955039342431e-05, "loss": 0.0123, "step": 36430 }, { "grad_norm": 0.26388347148895264, "learning_rate": 7.532529495824274e-05, "loss": 0.0141, "step": 36440 }, { "grad_norm": 0.19791005551815033, "learning_rate": 7.531103675352497e-05, "loss": 0.014, "step": 36450 }, { "grad_norm": 0.17200691998004913, "learning_rate": 7.529677578083029e-05, "loss": 0.0142, "step": 36460 }, { "grad_norm": 0.16321244835853577, "learning_rate": 7.528251204171823e-05, "loss": 0.0109, "step": 36470 }, { "grad_norm": 0.20662808418273926, "learning_rate": 7.526824553774867e-05, "loss": 0.0139, "step": 36480 }, { "grad_norm": 0.22274143993854523, "learning_rate": 7.525397627048178e-05, "loss": 0.017, "step": 36490 }, { "grad_norm": 0.19898808002471924, "learning_rate": 7.523970424147802e-05, "loss": 0.0129, "step": 36500 }, { "grad_norm": 0.15078234672546387, "learning_rate": 7.522542945229813e-05, "loss": 0.0132, "step": 36510 }, { "grad_norm": 0.14521251618862152, "learning_rate": 7.521115190450321e-05, "loss": 0.013, "step": 36520 }, { "grad_norm": 0.17453806102275848, "learning_rate": 7.519687159965462e-05, "loss": 0.0151, "step": 36530 }, { "grad_norm": 0.17406755685806274, "learning_rate": 7.518258853931403e-05, "loss": 0.0128, "step": 36540 }, { "grad_norm": 0.17514032125473022, "learning_rate": 7.516830272504342e-05, "loss": 0.0142, "step": 36550 }, { "grad_norm": 0.17873132228851318, "learning_rate": 7.515401415840509e-05, "loss": 0.0109, "step": 36560 }, { "grad_norm": 0.1338733285665512, "learning_rate": 7.513972284096155e-05, "loss": 0.0114, "step": 36570 }, { "grad_norm": 0.21599850058555603, "learning_rate": 7.512542877427574e-05, "loss": 0.0122, "step": 36580 }, { "grad_norm": 0.1938782036304474, "learning_rate": 7.51111319599108e-05, "loss": 0.0133, "step": 36590 }, { "grad_norm": 0.14786726236343384, "learning_rate": 7.509683239943024e-05, "loss": 0.0119, "step": 36600 }, { "grad_norm": 0.17792291939258575, "learning_rate": 7.50825300943978e-05, "loss": 0.0109, "step": 36610 }, { "grad_norm": 0.20565709471702576, "learning_rate": 7.50682250463776e-05, "loss": 0.0135, "step": 36620 }, { "grad_norm": 0.1651357114315033, "learning_rate": 7.5053917256934e-05, "loss": 0.0143, "step": 36630 }, { "grad_norm": 0.20430755615234375, "learning_rate": 7.503960672763166e-05, "loss": 0.015, "step": 36640 }, { "grad_norm": 0.18384608626365662, "learning_rate": 7.502529346003559e-05, "loss": 0.0128, "step": 36650 }, { "grad_norm": 0.15805526077747345, "learning_rate": 7.501097745571106e-05, "loss": 0.0152, "step": 36660 }, { "grad_norm": 0.1510954052209854, "learning_rate": 7.499665871622361e-05, "loss": 0.0123, "step": 36670 }, { "grad_norm": 0.19269850850105286, "learning_rate": 7.498233724313917e-05, "loss": 0.0119, "step": 36680 }, { "grad_norm": 0.17574606835842133, "learning_rate": 7.496801303802387e-05, "loss": 0.012, "step": 36690 }, { "grad_norm": 0.27214327454566956, "learning_rate": 7.495368610244423e-05, "loss": 0.0131, "step": 36700 }, { "grad_norm": 0.22889116406440735, "learning_rate": 7.493935643796697e-05, "loss": 0.0164, "step": 36710 }, { "grad_norm": 0.24745048582553864, "learning_rate": 7.49250240461592e-05, "loss": 0.012, "step": 36720 }, { "grad_norm": 0.2323371171951294, "learning_rate": 7.491068892858826e-05, "loss": 0.0129, "step": 36730 }, { "grad_norm": 0.16877616941928864, "learning_rate": 7.489635108682184e-05, "loss": 0.0128, "step": 36740 }, { "grad_norm": 0.15298834443092346, "learning_rate": 7.48820105224279e-05, "loss": 0.0142, "step": 36750 }, { "grad_norm": 0.1895694136619568, "learning_rate": 7.48676672369747e-05, "loss": 0.0128, "step": 36760 }, { "grad_norm": 0.2281782180070877, "learning_rate": 7.485332123203079e-05, "loss": 0.0163, "step": 36770 }, { "grad_norm": 0.21791523694992065, "learning_rate": 7.483897250916504e-05, "loss": 0.0137, "step": 36780 }, { "grad_norm": 0.14829106628894806, "learning_rate": 7.48246210699466e-05, "loss": 0.0137, "step": 36790 }, { "grad_norm": 0.24493417143821716, "learning_rate": 7.481026691594492e-05, "loss": 0.0118, "step": 36800 }, { "grad_norm": 0.19690640270709991, "learning_rate": 7.479591004872975e-05, "loss": 0.015, "step": 36810 }, { "grad_norm": 0.1834094375371933, "learning_rate": 7.478155046987114e-05, "loss": 0.0119, "step": 36820 }, { "grad_norm": 0.15414412319660187, "learning_rate": 7.476718818093944e-05, "loss": 0.0115, "step": 36830 }, { "grad_norm": 0.168484166264534, "learning_rate": 7.475282318350527e-05, "loss": 0.0121, "step": 36840 }, { "grad_norm": 0.2025376856327057, "learning_rate": 7.473845547913957e-05, "loss": 0.0119, "step": 36850 }, { "grad_norm": 0.17588023841381073, "learning_rate": 7.472408506941357e-05, "loss": 0.014, "step": 36860 }, { "grad_norm": 0.2074722796678543, "learning_rate": 7.470971195589881e-05, "loss": 0.0144, "step": 36870 }, { "grad_norm": 0.18760521709918976, "learning_rate": 7.46953361401671e-05, "loss": 0.0127, "step": 36880 }, { "grad_norm": 0.21523766219615936, "learning_rate": 7.468095762379055e-05, "loss": 0.014, "step": 36890 }, { "grad_norm": 0.13802561163902283, "learning_rate": 7.466657640834158e-05, "loss": 0.0135, "step": 36900 }, { "grad_norm": 0.14010831713676453, "learning_rate": 7.46521924953929e-05, "loss": 0.0133, "step": 36910 }, { "grad_norm": 0.18096812069416046, "learning_rate": 7.463780588651752e-05, "loss": 0.0133, "step": 36920 }, { "grad_norm": 0.20943105220794678, "learning_rate": 7.462341658328873e-05, "loss": 0.0127, "step": 36930 }, { "grad_norm": 0.1739991009235382, "learning_rate": 7.460902458728012e-05, "loss": 0.0129, "step": 36940 }, { "grad_norm": 0.16946257650852203, "learning_rate": 7.459462990006558e-05, "loss": 0.0124, "step": 36950 }, { "grad_norm": 0.43448132276535034, "learning_rate": 7.458023252321929e-05, "loss": 0.0126, "step": 36960 }, { "grad_norm": 0.2771506905555725, "learning_rate": 7.456583245831573e-05, "loss": 0.0117, "step": 36970 }, { "grad_norm": 0.16319887340068817, "learning_rate": 7.455142970692968e-05, "loss": 0.0131, "step": 36980 }, { "grad_norm": 0.20949485898017883, "learning_rate": 7.453702427063615e-05, "loss": 0.0123, "step": 36990 }, { "grad_norm": 0.22637154161930084, "learning_rate": 7.452261615101057e-05, "loss": 0.0147, "step": 37000 }, { "grad_norm": 0.17444317042827606, "learning_rate": 7.450820534962856e-05, "loss": 0.0141, "step": 37010 }, { "grad_norm": 0.18734070658683777, "learning_rate": 7.449379186806604e-05, "loss": 0.0117, "step": 37020 }, { "grad_norm": 0.1768452376127243, "learning_rate": 7.44793757078993e-05, "loss": 0.0128, "step": 37030 }, { "grad_norm": 0.16105665266513824, "learning_rate": 7.44649568707048e-05, "loss": 0.016, "step": 37040 }, { "grad_norm": 0.17581753432750702, "learning_rate": 7.445053535805942e-05, "loss": 0.0136, "step": 37050 }, { "grad_norm": 0.18502940237522125, "learning_rate": 7.443611117154024e-05, "loss": 0.0148, "step": 37060 }, { "grad_norm": 0.17997917532920837, "learning_rate": 7.442168431272469e-05, "loss": 0.0139, "step": 37070 }, { "grad_norm": 0.22058342397212982, "learning_rate": 7.440725478319044e-05, "loss": 0.0131, "step": 37080 }, { "grad_norm": 0.1760770082473755, "learning_rate": 7.439282258451553e-05, "loss": 0.0137, "step": 37090 }, { "grad_norm": 0.14342260360717773, "learning_rate": 7.43783877182782e-05, "loss": 0.0132, "step": 37100 }, { "grad_norm": 0.16658611595630646, "learning_rate": 7.436395018605707e-05, "loss": 0.0127, "step": 37110 }, { "grad_norm": 0.27705103158950806, "learning_rate": 7.434950998943094e-05, "loss": 0.0136, "step": 37120 }, { "grad_norm": 0.18565025925636292, "learning_rate": 7.433506712997902e-05, "loss": 0.0142, "step": 37130 }, { "grad_norm": 0.15502573549747467, "learning_rate": 7.432062160928077e-05, "loss": 0.0124, "step": 37140 }, { "grad_norm": 0.29245561361312866, "learning_rate": 7.430617342891588e-05, "loss": 0.0122, "step": 37150 }, { "grad_norm": 0.1427404284477234, "learning_rate": 7.429172259046442e-05, "loss": 0.0137, "step": 37160 }, { "grad_norm": 0.23707039654254913, "learning_rate": 7.42772690955067e-05, "loss": 0.013, "step": 37170 }, { "grad_norm": 0.22493015229701996, "learning_rate": 7.426281294562332e-05, "loss": 0.0141, "step": 37180 }, { "grad_norm": 0.2678237855434418, "learning_rate": 7.424835414239519e-05, "loss": 0.0162, "step": 37190 }, { "grad_norm": 0.3856179416179657, "learning_rate": 7.423389268740352e-05, "loss": 0.014, "step": 37200 }, { "grad_norm": 0.23522697389125824, "learning_rate": 7.42194285822298e-05, "loss": 0.0127, "step": 37210 }, { "grad_norm": 0.19374898076057434, "learning_rate": 7.420496182845576e-05, "loss": 0.0127, "step": 37220 }, { "grad_norm": 0.24091798067092896, "learning_rate": 7.419049242766348e-05, "loss": 0.0135, "step": 37230 }, { "grad_norm": 0.1720581203699112, "learning_rate": 7.417602038143533e-05, "loss": 0.0138, "step": 37240 }, { "grad_norm": 0.16366179287433624, "learning_rate": 7.416154569135393e-05, "loss": 0.0125, "step": 37250 }, { "grad_norm": 0.15652860701084137, "learning_rate": 7.414706835900222e-05, "loss": 0.0125, "step": 37260 }, { "grad_norm": 0.1692887544631958, "learning_rate": 7.413258838596345e-05, "loss": 0.0129, "step": 37270 }, { "grad_norm": 0.14336811006069183, "learning_rate": 7.411810577382106e-05, "loss": 0.0131, "step": 37280 }, { "grad_norm": 0.13293389976024628, "learning_rate": 7.410362052415889e-05, "loss": 0.0147, "step": 37290 }, { "grad_norm": 0.14648686349391937, "learning_rate": 7.408913263856102e-05, "loss": 0.0147, "step": 37300 }, { "grad_norm": 0.23808905482292175, "learning_rate": 7.407464211861183e-05, "loss": 0.0137, "step": 37310 }, { "grad_norm": 0.14988309144973755, "learning_rate": 7.406014896589597e-05, "loss": 0.0125, "step": 37320 }, { "grad_norm": 0.20822454988956451, "learning_rate": 7.404565318199838e-05, "loss": 0.0138, "step": 37330 }, { "grad_norm": 0.19516563415527344, "learning_rate": 7.403115476850432e-05, "loss": 0.0137, "step": 37340 }, { "grad_norm": 0.29432424902915955, "learning_rate": 7.401665372699932e-05, "loss": 0.0136, "step": 37350 }, { "grad_norm": 0.1784030646085739, "learning_rate": 7.400215005906916e-05, "loss": 0.012, "step": 37360 }, { "grad_norm": 0.17039883136749268, "learning_rate": 7.398764376629996e-05, "loss": 0.0121, "step": 37370 }, { "grad_norm": 0.1626720130443573, "learning_rate": 7.39731348502781e-05, "loss": 0.0113, "step": 37380 }, { "grad_norm": 0.19112205505371094, "learning_rate": 7.395862331259028e-05, "loss": 0.0124, "step": 37390 }, { "grad_norm": 0.16021336615085602, "learning_rate": 7.394410915482341e-05, "loss": 0.0125, "step": 37400 }, { "grad_norm": 0.16830569505691528, "learning_rate": 7.392959237856476e-05, "loss": 0.0141, "step": 37410 }, { "grad_norm": 0.18819506466388702, "learning_rate": 7.391507298540187e-05, "loss": 0.0144, "step": 37420 }, { "grad_norm": 0.18839740753173828, "learning_rate": 7.390055097692257e-05, "loss": 0.014, "step": 37430 }, { "grad_norm": 0.16389340162277222, "learning_rate": 7.388602635471494e-05, "loss": 0.0144, "step": 37440 }, { "grad_norm": 0.19162216782569885, "learning_rate": 7.387149912036738e-05, "loss": 0.0178, "step": 37450 }, { "grad_norm": 0.14232584834098816, "learning_rate": 7.385696927546854e-05, "loss": 0.0155, "step": 37460 }, { "grad_norm": 0.13947196304798126, "learning_rate": 7.384243682160743e-05, "loss": 0.0123, "step": 37470 }, { "grad_norm": 0.1335354894399643, "learning_rate": 7.382790176037329e-05, "loss": 0.012, "step": 37480 }, { "grad_norm": 0.16330158710479736, "learning_rate": 7.381336409335562e-05, "loss": 0.0128, "step": 37490 }, { "grad_norm": 0.15542668104171753, "learning_rate": 7.379882382214426e-05, "loss": 0.0151, "step": 37500 }, { "grad_norm": 0.15400800108909607, "learning_rate": 7.378428094832931e-05, "loss": 0.0164, "step": 37510 }, { "grad_norm": 0.18545480072498322, "learning_rate": 7.376973547350114e-05, "loss": 0.014, "step": 37520 }, { "grad_norm": 0.23644661903381348, "learning_rate": 7.375518739925046e-05, "loss": 0.0138, "step": 37530 }, { "grad_norm": 0.181821808218956, "learning_rate": 7.374063672716818e-05, "loss": 0.0142, "step": 37540 }, { "grad_norm": 0.20438063144683838, "learning_rate": 7.372608345884558e-05, "loss": 0.0116, "step": 37550 }, { "grad_norm": 0.1825239658355713, "learning_rate": 7.371152759587415e-05, "loss": 0.0135, "step": 37560 }, { "grad_norm": 0.2021765410900116, "learning_rate": 7.369696913984572e-05, "loss": 0.0142, "step": 37570 }, { "grad_norm": 0.1580180525779724, "learning_rate": 7.368240809235237e-05, "loss": 0.0123, "step": 37580 }, { "grad_norm": 0.2787756025791168, "learning_rate": 7.366784445498649e-05, "loss": 0.0147, "step": 37590 }, { "grad_norm": 0.20095494389533997, "learning_rate": 7.36532782293407e-05, "loss": 0.0163, "step": 37600 }, { "grad_norm": 0.1824163943529129, "learning_rate": 7.363870941700797e-05, "loss": 0.0129, "step": 37610 }, { "grad_norm": 0.18653319776058197, "learning_rate": 7.362413801958152e-05, "loss": 0.0132, "step": 37620 }, { "grad_norm": 0.22481468319892883, "learning_rate": 7.360956403865486e-05, "loss": 0.0151, "step": 37630 }, { "grad_norm": 0.20175907015800476, "learning_rate": 7.359498747582177e-05, "loss": 0.012, "step": 37640 }, { "grad_norm": 0.27841174602508545, "learning_rate": 7.358040833267634e-05, "loss": 0.0123, "step": 37650 }, { "grad_norm": 0.21487461030483246, "learning_rate": 7.356582661081288e-05, "loss": 0.0138, "step": 37660 }, { "grad_norm": 0.18452604115009308, "learning_rate": 7.355124231182607e-05, "loss": 0.0122, "step": 37670 }, { "grad_norm": 0.14171215891838074, "learning_rate": 7.353665543731083e-05, "loss": 0.0127, "step": 37680 }, { "grad_norm": 0.18974103033542633, "learning_rate": 7.352206598886229e-05, "loss": 0.0137, "step": 37690 }, { "grad_norm": 0.28413453698158264, "learning_rate": 7.350747396807601e-05, "loss": 0.013, "step": 37700 }, { "grad_norm": 0.17982454597949982, "learning_rate": 7.349287937654772e-05, "loss": 0.0163, "step": 37710 }, { "grad_norm": 0.22603729367256165, "learning_rate": 7.347828221587345e-05, "loss": 0.0131, "step": 37720 }, { "grad_norm": 0.15604065358638763, "learning_rate": 7.346368248764955e-05, "loss": 0.0141, "step": 37730 }, { "grad_norm": 0.17310883104801178, "learning_rate": 7.34490801934726e-05, "loss": 0.0154, "step": 37740 }, { "grad_norm": 0.15669101476669312, "learning_rate": 7.343447533493947e-05, "loss": 0.0127, "step": 37750 }, { "grad_norm": 0.16924543678760529, "learning_rate": 7.341986791364739e-05, "loss": 0.0123, "step": 37760 }, { "grad_norm": 0.22946760058403015, "learning_rate": 7.340525793119374e-05, "loss": 0.0147, "step": 37770 }, { "grad_norm": 0.1807970106601715, "learning_rate": 7.33906453891763e-05, "loss": 0.0132, "step": 37780 }, { "grad_norm": 0.17110596597194672, "learning_rate": 7.337603028919303e-05, "loss": 0.0127, "step": 37790 }, { "grad_norm": 0.1542062759399414, "learning_rate": 7.336141263284225e-05, "loss": 0.0126, "step": 37800 }, { "grad_norm": 0.20737527310848236, "learning_rate": 7.334679242172249e-05, "loss": 0.0133, "step": 37810 }, { "grad_norm": 0.1729678064584732, "learning_rate": 7.33321696574326e-05, "loss": 0.0122, "step": 37820 }, { "grad_norm": 0.17805275321006775, "learning_rate": 7.331754434157174e-05, "loss": 0.014, "step": 37830 }, { "grad_norm": 0.25056663155555725, "learning_rate": 7.330291647573928e-05, "loss": 0.0127, "step": 37840 }, { "grad_norm": 0.17322340607643127, "learning_rate": 7.32882860615349e-05, "loss": 0.0135, "step": 37850 }, { "grad_norm": 0.16465041041374207, "learning_rate": 7.327365310055858e-05, "loss": 0.0111, "step": 37860 }, { "grad_norm": 0.18590088188648224, "learning_rate": 7.325901759441055e-05, "loss": 0.0146, "step": 37870 }, { "grad_norm": 0.23969201743602753, "learning_rate": 7.324437954469132e-05, "loss": 0.0144, "step": 37880 }, { "grad_norm": 0.14526215195655823, "learning_rate": 7.32297389530017e-05, "loss": 0.0121, "step": 37890 }, { "grad_norm": 0.19730761647224426, "learning_rate": 7.321509582094274e-05, "loss": 0.0125, "step": 37900 }, { "grad_norm": 0.20291346311569214, "learning_rate": 7.320045015011582e-05, "loss": 0.0114, "step": 37910 }, { "grad_norm": 0.17615273594856262, "learning_rate": 7.318580194212253e-05, "loss": 0.014, "step": 37920 }, { "grad_norm": 0.1899668425321579, "learning_rate": 7.317115119856483e-05, "loss": 0.0138, "step": 37930 }, { "grad_norm": 0.13393111526966095, "learning_rate": 7.315649792104485e-05, "loss": 0.0117, "step": 37940 }, { "grad_norm": 0.20444056391716003, "learning_rate": 7.314184211116507e-05, "loss": 0.0133, "step": 37950 }, { "grad_norm": 0.17362834513187408, "learning_rate": 7.312718377052823e-05, "loss": 0.0142, "step": 37960 }, { "grad_norm": 0.1496487706899643, "learning_rate": 7.311252290073737e-05, "loss": 0.0138, "step": 37970 }, { "grad_norm": 0.18587027490139008, "learning_rate": 7.309785950339572e-05, "loss": 0.0126, "step": 37980 }, { "grad_norm": 0.2600376307964325, "learning_rate": 7.308319358010692e-05, "loss": 0.0123, "step": 37990 }, { "grad_norm": 0.17783813178539276, "learning_rate": 7.306852513247474e-05, "loss": 0.0126, "step": 38000 }, { "grad_norm": 0.19239144027233124, "learning_rate": 7.305385416210336e-05, "loss": 0.0115, "step": 38010 }, { "grad_norm": 0.1449950635433197, "learning_rate": 7.303918067059711e-05, "loss": 0.0123, "step": 38020 }, { "grad_norm": 0.14169883728027344, "learning_rate": 7.302450465956075e-05, "loss": 0.0109, "step": 38030 }, { "grad_norm": 0.19662585854530334, "learning_rate": 7.300982613059914e-05, "loss": 0.0114, "step": 38040 }, { "grad_norm": 0.1555565744638443, "learning_rate": 7.299514508531757e-05, "loss": 0.012, "step": 38050 }, { "grad_norm": 0.19569776952266693, "learning_rate": 7.298046152532148e-05, "loss": 0.0125, "step": 38060 }, { "grad_norm": 0.2538878321647644, "learning_rate": 7.296577545221668e-05, "loss": 0.0124, "step": 38070 }, { "grad_norm": 0.21309392154216766, "learning_rate": 7.29510868676092e-05, "loss": 0.0139, "step": 38080 }, { "grad_norm": 0.12184181064367294, "learning_rate": 7.293639577310538e-05, "loss": 0.0117, "step": 38090 }, { "grad_norm": 0.1751752346754074, "learning_rate": 7.292170217031179e-05, "loss": 0.012, "step": 38100 }, { "grad_norm": 0.20185859501361847, "learning_rate": 7.290700606083532e-05, "loss": 0.0122, "step": 38110 }, { "grad_norm": 0.1996064931154251, "learning_rate": 7.28923074462831e-05, "loss": 0.0113, "step": 38120 }, { "grad_norm": 0.45731037855148315, "learning_rate": 7.287760632826255e-05, "loss": 0.0141, "step": 38130 }, { "grad_norm": 0.2109917849302292, "learning_rate": 7.286290270838139e-05, "loss": 0.013, "step": 38140 }, { "grad_norm": 0.14079786837100983, "learning_rate": 7.284819658824756e-05, "loss": 0.0133, "step": 38150 }, { "grad_norm": 0.16864438354969025, "learning_rate": 7.283348796946929e-05, "loss": 0.0134, "step": 38160 }, { "grad_norm": 0.2029370665550232, "learning_rate": 7.28187768536551e-05, "loss": 0.0125, "step": 38170 }, { "grad_norm": 0.26527976989746094, "learning_rate": 7.280406324241379e-05, "loss": 0.0122, "step": 38180 }, { "grad_norm": 0.18350493907928467, "learning_rate": 7.278934713735441e-05, "loss": 0.0138, "step": 38190 }, { "grad_norm": 0.3382343053817749, "learning_rate": 7.277462854008629e-05, "loss": 0.013, "step": 38200 }, { "grad_norm": 0.2258516103029251, "learning_rate": 7.275990745221904e-05, "loss": 0.0132, "step": 38210 }, { "grad_norm": 0.18593943119049072, "learning_rate": 7.274518387536252e-05, "loss": 0.0122, "step": 38220 }, { "grad_norm": 0.1643323302268982, "learning_rate": 7.273045781112689e-05, "loss": 0.0121, "step": 38230 }, { "grad_norm": 0.19041208922863007, "learning_rate": 7.271572926112257e-05, "loss": 0.0115, "step": 38240 }, { "grad_norm": 0.17395533621311188, "learning_rate": 7.270099822696024e-05, "loss": 0.0132, "step": 38250 }, { "grad_norm": 0.1804037243127823, "learning_rate": 7.26862647102509e-05, "loss": 0.0138, "step": 38260 }, { "grad_norm": 0.20186379551887512, "learning_rate": 7.267152871260573e-05, "loss": 0.0133, "step": 38270 }, { "grad_norm": 0.22373363375663757, "learning_rate": 7.265679023563629e-05, "loss": 0.014, "step": 38280 }, { "grad_norm": 0.1850835084915161, "learning_rate": 7.264204928095433e-05, "loss": 0.0137, "step": 38290 }, { "grad_norm": 0.19695501029491425, "learning_rate": 7.262730585017188e-05, "loss": 0.0166, "step": 38300 }, { "grad_norm": 0.18203821778297424, "learning_rate": 7.261255994490131e-05, "loss": 0.0145, "step": 38310 }, { "grad_norm": 0.21827229857444763, "learning_rate": 7.259781156675517e-05, "loss": 0.0136, "step": 38320 }, { "grad_norm": 0.15261584520339966, "learning_rate": 7.258306071734633e-05, "loss": 0.0128, "step": 38330 }, { "grad_norm": 0.18834426999092102, "learning_rate": 7.256830739828792e-05, "loss": 0.0142, "step": 38340 }, { "grad_norm": 0.18592701852321625, "learning_rate": 7.255355161119336e-05, "loss": 0.0127, "step": 38350 }, { "grad_norm": 0.22856993973255157, "learning_rate": 7.253879335767628e-05, "loss": 0.0121, "step": 38360 }, { "grad_norm": 0.1726876199245453, "learning_rate": 7.252403263935067e-05, "loss": 0.0125, "step": 38370 }, { "grad_norm": 0.16781102120876312, "learning_rate": 7.25092694578307e-05, "loss": 0.0107, "step": 38380 }, { "grad_norm": 0.19220097362995148, "learning_rate": 7.249450381473087e-05, "loss": 0.0138, "step": 38390 }, { "grad_norm": 0.21029530465602875, "learning_rate": 7.247973571166593e-05, "loss": 0.0116, "step": 38400 }, { "grad_norm": 0.15648606419563293, "learning_rate": 7.246496515025089e-05, "loss": 0.0111, "step": 38410 }, { "grad_norm": 0.19204553961753845, "learning_rate": 7.245019213210101e-05, "loss": 0.0125, "step": 38420 }, { "grad_norm": 0.18451005220413208, "learning_rate": 7.24354166588319e-05, "loss": 0.0146, "step": 38430 }, { "grad_norm": 0.1549244374036789, "learning_rate": 7.242063873205935e-05, "loss": 0.0121, "step": 38440 }, { "grad_norm": 0.36733996868133545, "learning_rate": 7.240585835339946e-05, "loss": 0.0157, "step": 38450 }, { "grad_norm": 0.18779948353767395, "learning_rate": 7.239107552446857e-05, "loss": 0.0133, "step": 38460 }, { "grad_norm": 0.24015609920024872, "learning_rate": 7.237629024688334e-05, "loss": 0.0139, "step": 38470 }, { "grad_norm": 0.1705866903066635, "learning_rate": 7.236150252226064e-05, "loss": 0.0155, "step": 38480 }, { "grad_norm": 0.43798181414604187, "learning_rate": 7.234671235221765e-05, "loss": 0.0145, "step": 38490 }, { "grad_norm": 0.187627375125885, "learning_rate": 7.233191973837179e-05, "loss": 0.0148, "step": 38500 }, { "grad_norm": 0.1715877652168274, "learning_rate": 7.231712468234078e-05, "loss": 0.0146, "step": 38510 }, { "grad_norm": 0.17175479233264923, "learning_rate": 7.230232718574254e-05, "loss": 0.0132, "step": 38520 }, { "grad_norm": 0.16247409582138062, "learning_rate": 7.228752725019535e-05, "loss": 0.0152, "step": 38530 }, { "grad_norm": 0.17449620366096497, "learning_rate": 7.227272487731769e-05, "loss": 0.0129, "step": 38540 }, { "grad_norm": 0.19525188207626343, "learning_rate": 7.225792006872831e-05, "loss": 0.0137, "step": 38550 }, { "grad_norm": 0.1884385645389557, "learning_rate": 7.224311282604628e-05, "loss": 0.0145, "step": 38560 }, { "grad_norm": 0.2825694680213928, "learning_rate": 7.222830315089085e-05, "loss": 0.0144, "step": 38570 }, { "grad_norm": 0.22691865265369415, "learning_rate": 7.22134910448816e-05, "loss": 0.0132, "step": 38580 }, { "grad_norm": 0.24001343548297882, "learning_rate": 7.219867650963839e-05, "loss": 0.0144, "step": 38590 }, { "grad_norm": 0.27735358476638794, "learning_rate": 7.21838595467813e-05, "loss": 0.0146, "step": 38600 }, { "grad_norm": 0.2485819309949875, "learning_rate": 7.216904015793067e-05, "loss": 0.0154, "step": 38610 }, { "grad_norm": 0.1836766004562378, "learning_rate": 7.215421834470713e-05, "loss": 0.0131, "step": 38620 }, { "grad_norm": 0.10473271459341049, "learning_rate": 7.213939410873158e-05, "loss": 0.0114, "step": 38630 }, { "grad_norm": 0.20191366970539093, "learning_rate": 7.212456745162519e-05, "loss": 0.0121, "step": 38640 }, { "grad_norm": 0.26541534066200256, "learning_rate": 7.210973837500937e-05, "loss": 0.0131, "step": 38650 }, { "grad_norm": 0.15096715092658997, "learning_rate": 7.209490688050578e-05, "loss": 0.0122, "step": 38660 }, { "grad_norm": 0.17005521059036255, "learning_rate": 7.208007296973641e-05, "loss": 0.0123, "step": 38670 }, { "grad_norm": 0.19774529337882996, "learning_rate": 7.206523664432345e-05, "loss": 0.0152, "step": 38680 }, { "grad_norm": 0.1950654238462448, "learning_rate": 7.205039790588939e-05, "loss": 0.0118, "step": 38690 }, { "grad_norm": 0.20366990566253662, "learning_rate": 7.203555675605697e-05, "loss": 0.0146, "step": 38700 }, { "grad_norm": 0.19004665315151215, "learning_rate": 7.202071319644917e-05, "loss": 0.0116, "step": 38710 }, { "grad_norm": 0.183390274643898, "learning_rate": 7.200586722868932e-05, "loss": 0.0131, "step": 38720 }, { "grad_norm": 0.1632307469844818, "learning_rate": 7.199101885440088e-05, "loss": 0.0135, "step": 38730 }, { "grad_norm": 0.16482767462730408, "learning_rate": 7.197616807520771e-05, "loss": 0.013, "step": 38740 }, { "grad_norm": 0.21084553003311157, "learning_rate": 7.196131489273381e-05, "loss": 0.014, "step": 38750 }, { "grad_norm": 0.16267208755016327, "learning_rate": 7.194645930860355e-05, "loss": 0.0164, "step": 38760 }, { "grad_norm": 0.19023659825325012, "learning_rate": 7.19316013244415e-05, "loss": 0.0138, "step": 38770 }, { "grad_norm": 0.23810835182666779, "learning_rate": 7.191674094187248e-05, "loss": 0.0129, "step": 38780 }, { "grad_norm": 0.22879508137702942, "learning_rate": 7.190187816252165e-05, "loss": 0.0128, "step": 38790 }, { "grad_norm": 0.1911182552576065, "learning_rate": 7.188701298801435e-05, "loss": 0.0116, "step": 38800 }, { "grad_norm": 0.16617438197135925, "learning_rate": 7.18721454199762e-05, "loss": 0.0122, "step": 38810 }, { "grad_norm": 0.23531529307365417, "learning_rate": 7.185727546003312e-05, "loss": 0.014, "step": 38820 }, { "grad_norm": 0.13801167905330658, "learning_rate": 7.184240310981126e-05, "loss": 0.0153, "step": 38830 }, { "grad_norm": 0.1895289421081543, "learning_rate": 7.182752837093704e-05, "loss": 0.0122, "step": 38840 }, { "grad_norm": 0.176600843667984, "learning_rate": 7.181265124503711e-05, "loss": 0.0143, "step": 38850 }, { "grad_norm": 0.18366549909114838, "learning_rate": 7.179777173373847e-05, "loss": 0.0111, "step": 38860 }, { "grad_norm": 0.17686066031455994, "learning_rate": 7.178288983866826e-05, "loss": 0.0147, "step": 38870 }, { "grad_norm": 0.14082671701908112, "learning_rate": 7.176800556145397e-05, "loss": 0.0136, "step": 38880 }, { "grad_norm": 0.1347304731607437, "learning_rate": 7.175311890372334e-05, "loss": 0.0123, "step": 38890 }, { "grad_norm": 0.16192255914211273, "learning_rate": 7.17382298671043e-05, "loss": 0.0141, "step": 38900 }, { "grad_norm": 0.22313660383224487, "learning_rate": 7.172333845322515e-05, "loss": 0.0136, "step": 38910 }, { "grad_norm": 0.16200578212738037, "learning_rate": 7.170844466371436e-05, "loss": 0.012, "step": 38920 }, { "grad_norm": 0.22498241066932678, "learning_rate": 7.16935485002007e-05, "loss": 0.0131, "step": 38930 }, { "grad_norm": 0.1510016918182373, "learning_rate": 7.167864996431319e-05, "loss": 0.0117, "step": 38940 }, { "grad_norm": 0.36190855503082275, "learning_rate": 7.166374905768111e-05, "loss": 0.0115, "step": 38950 }, { "grad_norm": 0.27390068769454956, "learning_rate": 7.164884578193401e-05, "loss": 0.0125, "step": 38960 }, { "grad_norm": 0.20231780409812927, "learning_rate": 7.163394013870168e-05, "loss": 0.0118, "step": 38970 }, { "grad_norm": 0.201507568359375, "learning_rate": 7.161903212961419e-05, "loss": 0.0118, "step": 38980 }, { "grad_norm": 0.18181033432483673, "learning_rate": 7.160412175630183e-05, "loss": 0.0144, "step": 38990 }, { "grad_norm": 0.16262969374656677, "learning_rate": 7.158920902039521e-05, "loss": 0.013, "step": 39000 }, { "grad_norm": 0.19913099706172943, "learning_rate": 7.157429392352514e-05, "loss": 0.0124, "step": 39010 }, { "grad_norm": 0.1629124879837036, "learning_rate": 7.155937646732275e-05, "loss": 0.0113, "step": 39020 }, { "grad_norm": 0.18440435826778412, "learning_rate": 7.154445665341933e-05, "loss": 0.0137, "step": 39030 }, { "grad_norm": 0.17609015107154846, "learning_rate": 7.152953448344654e-05, "loss": 0.012, "step": 39040 }, { "grad_norm": 0.1880524754524231, "learning_rate": 7.151460995903624e-05, "loss": 0.012, "step": 39050 }, { "grad_norm": 0.236612930893898, "learning_rate": 7.149968308182052e-05, "loss": 0.0123, "step": 39060 }, { "grad_norm": 0.1683746725320816, "learning_rate": 7.14847538534318e-05, "loss": 0.0127, "step": 39070 }, { "grad_norm": 0.2691209316253662, "learning_rate": 7.14698222755027e-05, "loss": 0.0117, "step": 39080 }, { "grad_norm": 0.21302202343940735, "learning_rate": 7.14548883496661e-05, "loss": 0.0112, "step": 39090 }, { "grad_norm": 0.1610792875289917, "learning_rate": 7.143995207755517e-05, "loss": 0.0123, "step": 39100 }, { "grad_norm": 0.19417351484298706, "learning_rate": 7.142501346080333e-05, "loss": 0.013, "step": 39110 }, { "grad_norm": 0.15374961495399475, "learning_rate": 7.141007250104421e-05, "loss": 0.0125, "step": 39120 }, { "grad_norm": 0.21538180112838745, "learning_rate": 7.139512919991176e-05, "loss": 0.0122, "step": 39130 }, { "grad_norm": 0.15265806019306183, "learning_rate": 7.138018355904014e-05, "loss": 0.0115, "step": 39140 }, { "grad_norm": 0.20237983763217926, "learning_rate": 7.13652355800638e-05, "loss": 0.0126, "step": 39150 }, { "grad_norm": 0.27029454708099365, "learning_rate": 7.135028526461743e-05, "loss": 0.0122, "step": 39160 }, { "grad_norm": 0.17369426786899567, "learning_rate": 7.133533261433594e-05, "loss": 0.0127, "step": 39170 }, { "grad_norm": 0.15374289453029633, "learning_rate": 7.132037763085457e-05, "loss": 0.0166, "step": 39180 }, { "grad_norm": 0.1783214807510376, "learning_rate": 7.130542031580875e-05, "loss": 0.012, "step": 39190 }, { "grad_norm": 0.21155999600887299, "learning_rate": 7.12904606708342e-05, "loss": 0.0136, "step": 39200 }, { "grad_norm": 0.16865363717079163, "learning_rate": 7.127549869756687e-05, "loss": 0.0135, "step": 39210 }, { "grad_norm": 0.24105779826641083, "learning_rate": 7.126053439764299e-05, "loss": 0.0148, "step": 39220 }, { "grad_norm": 0.13412494957447052, "learning_rate": 7.124556777269904e-05, "loss": 0.0133, "step": 39230 }, { "grad_norm": 0.16611827909946442, "learning_rate": 7.123059882437174e-05, "loss": 0.012, "step": 39240 }, { "grad_norm": 0.17212340235710144, "learning_rate": 7.121562755429807e-05, "loss": 0.0123, "step": 39250 }, { "grad_norm": 0.1375150978565216, "learning_rate": 7.120065396411528e-05, "loss": 0.0138, "step": 39260 }, { "grad_norm": 0.15543711185455322, "learning_rate": 7.118567805546084e-05, "loss": 0.0124, "step": 39270 }, { "grad_norm": 0.15245011448860168, "learning_rate": 7.117069982997248e-05, "loss": 0.012, "step": 39280 }, { "grad_norm": 0.15878607332706451, "learning_rate": 7.115571928928825e-05, "loss": 0.0123, "step": 39290 }, { "grad_norm": 0.2550060451030731, "learning_rate": 7.114073643504635e-05, "loss": 0.0158, "step": 39300 }, { "grad_norm": 0.20105446875095367, "learning_rate": 7.11257512688853e-05, "loss": 0.0195, "step": 39310 }, { "grad_norm": 0.1930614858865738, "learning_rate": 7.111076379244384e-05, "loss": 0.0145, "step": 39320 }, { "grad_norm": 0.3283994495868683, "learning_rate": 7.109577400736101e-05, "loss": 0.0137, "step": 39330 }, { "grad_norm": 0.1595119833946228, "learning_rate": 7.108078191527605e-05, "loss": 0.0154, "step": 39340 }, { "grad_norm": 0.2418384701013565, "learning_rate": 7.106578751782847e-05, "loss": 0.0145, "step": 39350 }, { "grad_norm": 0.16943466663360596, "learning_rate": 7.105079081665803e-05, "loss": 0.0152, "step": 39360 }, { "grad_norm": 0.1753351241350174, "learning_rate": 7.103579181340476e-05, "loss": 0.0147, "step": 39370 }, { "grad_norm": 0.20031684637069702, "learning_rate": 7.102079050970893e-05, "loss": 0.0126, "step": 39380 }, { "grad_norm": 0.1866557002067566, "learning_rate": 7.100578690721105e-05, "loss": 0.0124, "step": 39390 }, { "grad_norm": 0.1935325711965561, "learning_rate": 7.09907810075519e-05, "loss": 0.013, "step": 39400 }, { "grad_norm": 0.20110389590263367, "learning_rate": 7.097577281237249e-05, "loss": 0.0136, "step": 39410 }, { "grad_norm": 0.17882610857486725, "learning_rate": 7.09607623233141e-05, "loss": 0.0141, "step": 39420 }, { "grad_norm": 0.2294963002204895, "learning_rate": 7.094574954201827e-05, "loss": 0.0129, "step": 39430 }, { "grad_norm": 0.24013668298721313, "learning_rate": 7.093073447012675e-05, "loss": 0.012, "step": 39440 }, { "grad_norm": 0.2945171594619751, "learning_rate": 7.09157171092816e-05, "loss": 0.0131, "step": 39450 }, { "grad_norm": 0.15352162718772888, "learning_rate": 7.090069746112504e-05, "loss": 0.013, "step": 39460 }, { "grad_norm": 0.16901631653308868, "learning_rate": 7.088567552729965e-05, "loss": 0.0117, "step": 39470 }, { "grad_norm": 0.16378644108772278, "learning_rate": 7.087065130944818e-05, "loss": 0.0121, "step": 39480 }, { "grad_norm": 0.11460917443037033, "learning_rate": 7.085562480921366e-05, "loss": 0.0108, "step": 39490 }, { "grad_norm": 0.15942110121250153, "learning_rate": 7.084059602823937e-05, "loss": 0.0108, "step": 39500 }, { "grad_norm": 0.173787921667099, "learning_rate": 7.082556496816882e-05, "loss": 0.0123, "step": 39510 }, { "grad_norm": 0.16078616678714752, "learning_rate": 7.081053163064582e-05, "loss": 0.0108, "step": 39520 }, { "grad_norm": 0.16533023118972778, "learning_rate": 7.079549601731434e-05, "loss": 0.0122, "step": 39530 }, { "grad_norm": 0.205962136387825, "learning_rate": 7.07804581298187e-05, "loss": 0.0137, "step": 39540 }, { "grad_norm": 0.20077885687351227, "learning_rate": 7.07654179698034e-05, "loss": 0.0108, "step": 39550 }, { "grad_norm": 0.19671352207660675, "learning_rate": 7.075037553891321e-05, "loss": 0.014, "step": 39560 }, { "grad_norm": 0.1729849874973297, "learning_rate": 7.073533083879315e-05, "loss": 0.0155, "step": 39570 }, { "grad_norm": 0.20287476480007172, "learning_rate": 7.072028387108849e-05, "loss": 0.0128, "step": 39580 }, { "grad_norm": 0.17227330803871155, "learning_rate": 7.070523463744472e-05, "loss": 0.0123, "step": 39590 }, { "grad_norm": 0.18596622347831726, "learning_rate": 7.069018313950763e-05, "loss": 0.0126, "step": 39600 }, { "grad_norm": 0.19682227075099945, "learning_rate": 7.067512937892322e-05, "loss": 0.0144, "step": 39610 }, { "grad_norm": 0.16512250900268555, "learning_rate": 7.066007335733775e-05, "loss": 0.0144, "step": 39620 }, { "grad_norm": 0.17536737024784088, "learning_rate": 7.064501507639772e-05, "loss": 0.011, "step": 39630 }, { "grad_norm": 0.2194562703371048, "learning_rate": 7.062995453774987e-05, "loss": 0.0144, "step": 39640 }, { "grad_norm": 0.17529992759227753, "learning_rate": 7.061489174304121e-05, "loss": 0.0162, "step": 39650 }, { "grad_norm": 0.20069214701652527, "learning_rate": 7.0599826693919e-05, "loss": 0.0159, "step": 39660 }, { "grad_norm": 0.3358708918094635, "learning_rate": 7.05847593920307e-05, "loss": 0.0163, "step": 39670 }, { "grad_norm": 0.18985936045646667, "learning_rate": 7.056968983902406e-05, "loss": 0.0155, "step": 39680 }, { "grad_norm": 0.15590836107730865, "learning_rate": 7.055461803654706e-05, "loss": 0.0164, "step": 39690 }, { "grad_norm": 0.22691398859024048, "learning_rate": 7.053954398624794e-05, "loss": 0.0149, "step": 39700 }, { "grad_norm": 0.18364425003528595, "learning_rate": 7.052446768977518e-05, "loss": 0.0128, "step": 39710 }, { "grad_norm": 0.17416156828403473, "learning_rate": 7.050938914877748e-05, "loss": 0.0123, "step": 39720 }, { "grad_norm": 0.12676776945590973, "learning_rate": 7.049430836490382e-05, "loss": 0.012, "step": 39730 }, { "grad_norm": 0.14537708461284637, "learning_rate": 7.047922533980342e-05, "loss": 0.0128, "step": 39740 }, { "grad_norm": 0.15596209466457367, "learning_rate": 7.046414007512571e-05, "loss": 0.0122, "step": 39750 }, { "grad_norm": 0.21643134951591492, "learning_rate": 7.044905257252042e-05, "loss": 0.0138, "step": 39760 }, { "grad_norm": 0.1999949961900711, "learning_rate": 7.04339628336375e-05, "loss": 0.0127, "step": 39770 }, { "grad_norm": 0.32105523347854614, "learning_rate": 7.041887086012711e-05, "loss": 0.0134, "step": 39780 }, { "grad_norm": 0.21667532622814178, "learning_rate": 7.040377665363969e-05, "loss": 0.0135, "step": 39790 }, { "grad_norm": 0.2149299830198288, "learning_rate": 7.038868021582594e-05, "loss": 0.0152, "step": 39800 }, { "grad_norm": 0.19507020711898804, "learning_rate": 7.037358154833679e-05, "loss": 0.0118, "step": 39810 }, { "grad_norm": 0.15854482352733612, "learning_rate": 7.035848065282339e-05, "loss": 0.0175, "step": 39820 }, { "grad_norm": 0.1884242296218872, "learning_rate": 7.034337753093714e-05, "loss": 0.0131, "step": 39830 }, { "grad_norm": 0.1590212881565094, "learning_rate": 7.032827218432972e-05, "loss": 0.0112, "step": 39840 }, { "grad_norm": 0.23563526570796967, "learning_rate": 7.031316461465302e-05, "loss": 0.0123, "step": 39850 }, { "grad_norm": 0.11438380926847458, "learning_rate": 7.029805482355915e-05, "loss": 0.0104, "step": 39860 }, { "grad_norm": 0.13912846148014069, "learning_rate": 7.028294281270055e-05, "loss": 0.0106, "step": 39870 }, { "grad_norm": 0.2178177684545517, "learning_rate": 7.026782858372982e-05, "loss": 0.014, "step": 39880 }, { "grad_norm": 0.22089554369449615, "learning_rate": 7.025271213829982e-05, "loss": 0.0125, "step": 39890 }, { "grad_norm": 0.18020306527614594, "learning_rate": 7.023759347806366e-05, "loss": 0.012, "step": 39900 }, { "grad_norm": 0.19303704798221588, "learning_rate": 7.022247260467472e-05, "loss": 0.0115, "step": 39910 }, { "grad_norm": 0.19601112604141235, "learning_rate": 7.020734951978658e-05, "loss": 0.0123, "step": 39920 }, { "grad_norm": 0.21463178098201752, "learning_rate": 7.019222422505307e-05, "loss": 0.0132, "step": 39930 }, { "grad_norm": 0.16115069389343262, "learning_rate": 7.017709672212827e-05, "loss": 0.0112, "step": 39940 }, { "grad_norm": 0.12688294053077698, "learning_rate": 7.016196701266652e-05, "loss": 0.011, "step": 39950 }, { "grad_norm": 0.15351024270057678, "learning_rate": 7.014683509832238e-05, "loss": 0.0119, "step": 39960 }, { "grad_norm": 0.16277755796909332, "learning_rate": 7.013170098075063e-05, "loss": 0.0115, "step": 39970 }, { "grad_norm": 0.2627684473991394, "learning_rate": 7.011656466160632e-05, "loss": 0.0137, "step": 39980 }, { "grad_norm": 0.1896129548549652, "learning_rate": 7.010142614254475e-05, "loss": 0.0131, "step": 39990 }, { "grad_norm": 0.17298708856105804, "learning_rate": 7.008628542522147e-05, "loss": 0.015, "step": 40000 }, { "grad_norm": 0.21072907745838165, "learning_rate": 7.007114251129218e-05, "loss": 0.013, "step": 40010 }, { "grad_norm": 0.162358820438385, "learning_rate": 7.005599740241293e-05, "loss": 0.0109, "step": 40020 }, { "grad_norm": 0.18949665129184723, "learning_rate": 7.004085010023996e-05, "loss": 0.013, "step": 40030 }, { "grad_norm": 0.15722428262233734, "learning_rate": 7.002570060642976e-05, "loss": 0.0114, "step": 40040 }, { "grad_norm": 0.21256376802921295, "learning_rate": 7.001054892263903e-05, "loss": 0.0126, "step": 40050 }, { "grad_norm": 0.14575187861919403, "learning_rate": 6.999539505052477e-05, "loss": 0.0114, "step": 40060 }, { "grad_norm": 0.17855191230773926, "learning_rate": 6.998023899174418e-05, "loss": 0.0134, "step": 40070 }, { "grad_norm": 0.17217028141021729, "learning_rate": 6.996508074795467e-05, "loss": 0.0159, "step": 40080 }, { "grad_norm": 0.3106304705142975, "learning_rate": 6.994992032081396e-05, "loss": 0.0127, "step": 40090 }, { "grad_norm": 0.18180224299430847, "learning_rate": 6.993475771197995e-05, "loss": 0.0124, "step": 40100 }, { "grad_norm": 0.12861886620521545, "learning_rate": 6.991959292311082e-05, "loss": 0.0119, "step": 40110 }, { "grad_norm": 0.12611907720565796, "learning_rate": 6.990442595586495e-05, "loss": 0.0106, "step": 40120 }, { "grad_norm": 0.2853662669658661, "learning_rate": 6.988925681190098e-05, "loss": 0.0128, "step": 40130 }, { "grad_norm": 0.21132443845272064, "learning_rate": 6.987408549287778e-05, "loss": 0.0125, "step": 40140 }, { "grad_norm": 0.1850934624671936, "learning_rate": 6.985891200045449e-05, "loss": 0.0133, "step": 40150 }, { "grad_norm": 0.17936940491199493, "learning_rate": 6.984373633629045e-05, "loss": 0.0138, "step": 40160 }, { "grad_norm": 0.1799830198287964, "learning_rate": 6.98285585020452e-05, "loss": 0.0132, "step": 40170 }, { "grad_norm": 0.2315894067287445, "learning_rate": 6.981337849937864e-05, "loss": 0.0098, "step": 40180 }, { "grad_norm": 0.21029916405677795, "learning_rate": 6.979819632995078e-05, "loss": 0.0136, "step": 40190 }, { "grad_norm": 0.27869924902915955, "learning_rate": 6.978301199542193e-05, "loss": 0.0109, "step": 40200 }, { "grad_norm": 0.1667386144399643, "learning_rate": 6.976782549745263e-05, "loss": 0.0125, "step": 40210 }, { "grad_norm": 0.25117167830467224, "learning_rate": 6.975263683770367e-05, "loss": 0.011, "step": 40220 }, { "grad_norm": 0.15984518826007843, "learning_rate": 6.973744601783604e-05, "loss": 0.0108, "step": 40230 }, { "grad_norm": 0.19810797274112701, "learning_rate": 6.972225303951097e-05, "loss": 0.0127, "step": 40240 }, { "grad_norm": 0.10306602716445923, "learning_rate": 6.970705790438998e-05, "loss": 0.0146, "step": 40250 }, { "grad_norm": 0.24486233294010162, "learning_rate": 6.969186061413477e-05, "loss": 0.0122, "step": 40260 }, { "grad_norm": 0.1719643622636795, "learning_rate": 6.967666117040727e-05, "loss": 0.0109, "step": 40270 }, { "grad_norm": 0.34879112243652344, "learning_rate": 6.966145957486972e-05, "loss": 0.0114, "step": 40280 }, { "grad_norm": 0.1773223876953125, "learning_rate": 6.964625582918449e-05, "loss": 0.0128, "step": 40290 }, { "grad_norm": 0.3199135363101959, "learning_rate": 6.963104993501425e-05, "loss": 0.0141, "step": 40300 }, { "grad_norm": 0.21038714051246643, "learning_rate": 6.961584189402192e-05, "loss": 0.0139, "step": 40310 }, { "grad_norm": 0.14372454583644867, "learning_rate": 6.96006317078706e-05, "loss": 0.0109, "step": 40320 }, { "grad_norm": 0.18505272269248962, "learning_rate": 6.958541937822367e-05, "loss": 0.0123, "step": 40330 }, { "grad_norm": 0.19414184987545013, "learning_rate": 6.957020490674473e-05, "loss": 0.0117, "step": 40340 }, { "grad_norm": 0.17315800487995148, "learning_rate": 6.95549882950976e-05, "loss": 0.0131, "step": 40350 }, { "grad_norm": 0.201814204454422, "learning_rate": 6.953976954494635e-05, "loss": 0.0124, "step": 40360 }, { "grad_norm": 0.20114196836948395, "learning_rate": 6.952454865795528e-05, "loss": 0.0132, "step": 40370 }, { "grad_norm": 0.1779671460390091, "learning_rate": 6.950932563578892e-05, "loss": 0.0113, "step": 40380 }, { "grad_norm": 0.2901234030723572, "learning_rate": 6.949410048011206e-05, "loss": 0.0117, "step": 40390 }, { "grad_norm": 0.25198426842689514, "learning_rate": 6.947887319258966e-05, "loss": 0.0106, "step": 40400 }, { "grad_norm": 0.20898327231407166, "learning_rate": 6.946364377488696e-05, "loss": 0.0133, "step": 40410 }, { "grad_norm": 0.2550463378429413, "learning_rate": 6.944841222866947e-05, "loss": 0.0124, "step": 40420 }, { "grad_norm": 0.137750044465065, "learning_rate": 6.943317855560284e-05, "loss": 0.0139, "step": 40430 }, { "grad_norm": 0.18879470229148865, "learning_rate": 6.941794275735302e-05, "loss": 0.0132, "step": 40440 }, { "grad_norm": 0.167707622051239, "learning_rate": 6.94027048355862e-05, "loss": 0.0104, "step": 40450 }, { "grad_norm": 0.18292991816997528, "learning_rate": 6.938746479196871e-05, "loss": 0.0116, "step": 40460 }, { "grad_norm": 0.16400887072086334, "learning_rate": 6.937222262816724e-05, "loss": 0.0137, "step": 40470 }, { "grad_norm": 0.12606088817119598, "learning_rate": 6.935697834584865e-05, "loss": 0.0115, "step": 40480 }, { "grad_norm": 0.1034439206123352, "learning_rate": 6.934173194667998e-05, "loss": 0.0113, "step": 40490 }, { "grad_norm": 0.14427942037582397, "learning_rate": 6.93264834323286e-05, "loss": 0.0109, "step": 40500 }, { "grad_norm": 0.22193238139152527, "learning_rate": 6.931123280446206e-05, "loss": 0.0106, "step": 40510 }, { "grad_norm": 0.14822039008140564, "learning_rate": 6.929598006474811e-05, "loss": 0.0113, "step": 40520 }, { "grad_norm": 0.17743965983390808, "learning_rate": 6.92807252148548e-05, "loss": 0.0114, "step": 40530 }, { "grad_norm": 0.19017577171325684, "learning_rate": 6.926546825645037e-05, "loss": 0.014, "step": 40540 }, { "grad_norm": 0.1983756124973297, "learning_rate": 6.92502091912033e-05, "loss": 0.0107, "step": 40550 }, { "grad_norm": 0.19260233640670776, "learning_rate": 6.923494802078229e-05, "loss": 0.0129, "step": 40560 }, { "grad_norm": 0.15817944705486298, "learning_rate": 6.921968474685628e-05, "loss": 0.0122, "step": 40570 }, { "grad_norm": 0.16506320238113403, "learning_rate": 6.920441937109445e-05, "loss": 0.0113, "step": 40580 }, { "grad_norm": 0.15279296040534973, "learning_rate": 6.91891518951662e-05, "loss": 0.01, "step": 40590 }, { "grad_norm": 0.20878179371356964, "learning_rate": 6.917388232074114e-05, "loss": 0.0119, "step": 40600 }, { "grad_norm": 0.21011753380298615, "learning_rate": 6.915861064948914e-05, "loss": 0.0132, "step": 40610 }, { "grad_norm": 0.1562407910823822, "learning_rate": 6.914333688308025e-05, "loss": 0.0124, "step": 40620 }, { "grad_norm": 0.15152181684970856, "learning_rate": 6.912806102318487e-05, "loss": 0.011, "step": 40630 }, { "grad_norm": 0.14842849969863892, "learning_rate": 6.911278307147347e-05, "loss": 0.0109, "step": 40640 }, { "grad_norm": 0.17441780865192413, "learning_rate": 6.909750302961684e-05, "loss": 0.0098, "step": 40650 }, { "grad_norm": 0.21254456043243408, "learning_rate": 6.9082220899286e-05, "loss": 0.0117, "step": 40660 }, { "grad_norm": 0.18938380479812622, "learning_rate": 6.906693668215219e-05, "loss": 0.0111, "step": 40670 }, { "grad_norm": 0.17631696164608002, "learning_rate": 6.905165037988683e-05, "loss": 0.012, "step": 40680 }, { "grad_norm": 0.24452625215053558, "learning_rate": 6.903636199416164e-05, "loss": 0.012, "step": 40690 }, { "grad_norm": 0.1339995265007019, "learning_rate": 6.902107152664851e-05, "loss": 0.0112, "step": 40700 }, { "grad_norm": 0.19261498749256134, "learning_rate": 6.90057789790196e-05, "loss": 0.0127, "step": 40710 }, { "grad_norm": 0.1456945240497589, "learning_rate": 6.899048435294728e-05, "loss": 0.0135, "step": 40720 }, { "grad_norm": 0.1520567089319229, "learning_rate": 6.897518765010415e-05, "loss": 0.0139, "step": 40730 }, { "grad_norm": 0.18918660283088684, "learning_rate": 6.895988887216302e-05, "loss": 0.0122, "step": 40740 }, { "grad_norm": 0.2672680914402008, "learning_rate": 6.894458802079694e-05, "loss": 0.0108, "step": 40750 }, { "grad_norm": 0.20987479388713837, "learning_rate": 6.892928509767922e-05, "loss": 0.0148, "step": 40760 }, { "grad_norm": 0.1722005009651184, "learning_rate": 6.891398010448333e-05, "loss": 0.0132, "step": 40770 }, { "grad_norm": 0.16708076000213623, "learning_rate": 6.8898673042883e-05, "loss": 0.0126, "step": 40780 }, { "grad_norm": 0.14569935202598572, "learning_rate": 6.888336391455222e-05, "loss": 0.0123, "step": 40790 }, { "grad_norm": 0.25348910689353943, "learning_rate": 6.886805272116513e-05, "loss": 0.0157, "step": 40800 }, { "grad_norm": 0.13259075582027435, "learning_rate": 6.885273946439617e-05, "loss": 0.013, "step": 40810 }, { "grad_norm": 0.19157201051712036, "learning_rate": 6.883742414591998e-05, "loss": 0.0137, "step": 40820 }, { "grad_norm": 0.17722438275814056, "learning_rate": 6.88221067674114e-05, "loss": 0.0148, "step": 40830 }, { "grad_norm": 0.2603098750114441, "learning_rate": 6.88067873305455e-05, "loss": 0.0127, "step": 40840 }, { "grad_norm": 0.11850878596305847, "learning_rate": 6.879146583699765e-05, "loss": 0.0108, "step": 40850 }, { "grad_norm": 0.149306982755661, "learning_rate": 6.877614228844334e-05, "loss": 0.011, "step": 40860 }, { "grad_norm": 0.2396678626537323, "learning_rate": 6.876081668655832e-05, "loss": 0.0122, "step": 40870 }, { "grad_norm": 0.16201967000961304, "learning_rate": 6.874548903301861e-05, "loss": 0.0142, "step": 40880 }, { "grad_norm": 0.1819058656692505, "learning_rate": 6.873015932950039e-05, "loss": 0.0126, "step": 40890 }, { "grad_norm": 0.1938324123620987, "learning_rate": 6.871482757768012e-05, "loss": 0.0123, "step": 40900 }, { "grad_norm": 0.19978374242782593, "learning_rate": 6.869949377923442e-05, "loss": 0.0116, "step": 40910 }, { "grad_norm": 0.19412609934806824, "learning_rate": 6.868415793584022e-05, "loss": 0.0116, "step": 40920 }, { "grad_norm": 0.2057330310344696, "learning_rate": 6.86688200491746e-05, "loss": 0.0125, "step": 40930 }, { "grad_norm": 0.15991835296154022, "learning_rate": 6.865348012091488e-05, "loss": 0.0115, "step": 40940 }, { "grad_norm": 0.21992501616477966, "learning_rate": 6.86381381527386e-05, "loss": 0.0139, "step": 40950 }, { "grad_norm": 0.17071880400180817, "learning_rate": 6.862279414632359e-05, "loss": 0.0122, "step": 40960 }, { "grad_norm": 0.3226985037326813, "learning_rate": 6.860744810334778e-05, "loss": 0.0137, "step": 40970 }, { "grad_norm": 0.17184001207351685, "learning_rate": 6.859210002548943e-05, "loss": 0.0131, "step": 40980 }, { "grad_norm": 0.20300208032131195, "learning_rate": 6.8576749914427e-05, "loss": 0.0126, "step": 40990 }, { "grad_norm": 0.2425946742296219, "learning_rate": 6.85613977718391e-05, "loss": 0.0139, "step": 41000 }, { "grad_norm": 0.1758807748556137, "learning_rate": 6.854604359940464e-05, "loss": 0.0123, "step": 41010 }, { "grad_norm": 0.33620861172676086, "learning_rate": 6.853068739880276e-05, "loss": 0.0118, "step": 41020 }, { "grad_norm": 0.1436256468296051, "learning_rate": 6.851532917171276e-05, "loss": 0.011, "step": 41030 }, { "grad_norm": 0.1487385630607605, "learning_rate": 6.849996891981418e-05, "loss": 0.0137, "step": 41040 }, { "grad_norm": 0.1230132132768631, "learning_rate": 6.848460664478684e-05, "loss": 0.0127, "step": 41050 }, { "grad_norm": 0.2971569895744324, "learning_rate": 6.846924234831069e-05, "loss": 0.0107, "step": 41060 }, { "grad_norm": 0.1873117983341217, "learning_rate": 6.845387603206597e-05, "loss": 0.0111, "step": 41070 }, { "grad_norm": 0.3812579810619354, "learning_rate": 6.843850769773312e-05, "loss": 0.0114, "step": 41080 }, { "grad_norm": 0.18198086321353912, "learning_rate": 6.842313734699278e-05, "loss": 0.0127, "step": 41090 }, { "grad_norm": 0.15503405034542084, "learning_rate": 6.840776498152584e-05, "loss": 0.0116, "step": 41100 }, { "grad_norm": 0.17865994572639465, "learning_rate": 6.83923906030134e-05, "loss": 0.0144, "step": 41110 }, { "grad_norm": 0.1474749594926834, "learning_rate": 6.837701421313677e-05, "loss": 0.0143, "step": 41120 }, { "grad_norm": 0.13080047070980072, "learning_rate": 6.83616358135775e-05, "loss": 0.0109, "step": 41130 }, { "grad_norm": 0.27154967188835144, "learning_rate": 6.834625540601734e-05, "loss": 0.0121, "step": 41140 }, { "grad_norm": 0.13567225635051727, "learning_rate": 6.833087299213829e-05, "loss": 0.0112, "step": 41150 }, { "grad_norm": 0.1911352574825287, "learning_rate": 6.83154885736225e-05, "loss": 0.0119, "step": 41160 }, { "grad_norm": 0.18992088735103607, "learning_rate": 6.830010215215244e-05, "loss": 0.011, "step": 41170 }, { "grad_norm": 0.18799559772014618, "learning_rate": 6.828471372941072e-05, "loss": 0.0153, "step": 41180 }, { "grad_norm": 0.16401100158691406, "learning_rate": 6.826932330708018e-05, "loss": 0.012, "step": 41190 }, { "grad_norm": 0.15502259135246277, "learning_rate": 6.825393088684393e-05, "loss": 0.013, "step": 41200 }, { "grad_norm": 0.14570558071136475, "learning_rate": 6.823853647038524e-05, "loss": 0.0112, "step": 41210 }, { "grad_norm": 0.2003837376832962, "learning_rate": 6.822314005938762e-05, "loss": 0.0114, "step": 41220 }, { "grad_norm": 0.2095889449119568, "learning_rate": 6.820774165553479e-05, "loss": 0.0118, "step": 41230 }, { "grad_norm": 0.18454422056674957, "learning_rate": 6.819234126051072e-05, "loss": 0.0109, "step": 41240 }, { "grad_norm": 0.1518162339925766, "learning_rate": 6.817693887599956e-05, "loss": 0.0106, "step": 41250 }, { "grad_norm": 0.25546520948410034, "learning_rate": 6.816153450368568e-05, "loss": 0.0135, "step": 41260 }, { "grad_norm": 0.20061227679252625, "learning_rate": 6.814612814525372e-05, "loss": 0.0125, "step": 41270 }, { "grad_norm": 0.24426805973052979, "learning_rate": 6.813071980238846e-05, "loss": 0.0129, "step": 41280 }, { "grad_norm": 0.20066866278648376, "learning_rate": 6.811530947677492e-05, "loss": 0.0131, "step": 41290 }, { "grad_norm": 0.16949190199375153, "learning_rate": 6.809989717009839e-05, "loss": 0.0142, "step": 41300 }, { "grad_norm": 0.18954996764659882, "learning_rate": 6.808448288404431e-05, "loss": 0.0115, "step": 41310 }, { "grad_norm": 0.17836597561836243, "learning_rate": 6.806906662029838e-05, "loss": 0.0114, "step": 41320 }, { "grad_norm": 0.19508720934391022, "learning_rate": 6.805364838054647e-05, "loss": 0.0138, "step": 41330 }, { "grad_norm": 0.23274771869182587, "learning_rate": 6.803822816647474e-05, "loss": 0.0115, "step": 41340 }, { "grad_norm": 0.16289466619491577, "learning_rate": 6.802280597976949e-05, "loss": 0.0124, "step": 41350 }, { "grad_norm": 0.4219755232334137, "learning_rate": 6.800738182211727e-05, "loss": 0.0111, "step": 41360 }, { "grad_norm": 0.20008613169193268, "learning_rate": 6.799195569520485e-05, "loss": 0.0117, "step": 41370 }, { "grad_norm": 0.1439870297908783, "learning_rate": 6.79765276007192e-05, "loss": 0.0127, "step": 41380 }, { "grad_norm": 0.21108579635620117, "learning_rate": 6.796109754034753e-05, "loss": 0.013, "step": 41390 }, { "grad_norm": 0.132668599486351, "learning_rate": 6.794566551577724e-05, "loss": 0.013, "step": 41400 }, { "grad_norm": 0.1793561726808548, "learning_rate": 6.793023152869594e-05, "loss": 0.0102, "step": 41410 }, { "grad_norm": 0.1403461992740631, "learning_rate": 6.791479558079147e-05, "loss": 0.0099, "step": 41420 }, { "grad_norm": 0.22848433256149292, "learning_rate": 6.789935767375191e-05, "loss": 0.0104, "step": 41430 }, { "grad_norm": 0.2134680151939392, "learning_rate": 6.78839178092655e-05, "loss": 0.0125, "step": 41440 }, { "grad_norm": 0.18974652886390686, "learning_rate": 6.786847598902072e-05, "loss": 0.0124, "step": 41450 }, { "grad_norm": 0.18001508712768555, "learning_rate": 6.785303221470627e-05, "loss": 0.0107, "step": 41460 }, { "grad_norm": 0.19935810565948486, "learning_rate": 6.783758648801108e-05, "loss": 0.0133, "step": 41470 }, { "grad_norm": 0.17617255449295044, "learning_rate": 6.782213881062422e-05, "loss": 0.0117, "step": 41480 }, { "grad_norm": 0.16083450615406036, "learning_rate": 6.780668918423508e-05, "loss": 0.0126, "step": 41490 }, { "grad_norm": 0.14318646490573883, "learning_rate": 6.779123761053317e-05, "loss": 0.0098, "step": 41500 }, { "grad_norm": 0.18531401455402374, "learning_rate": 6.777578409120826e-05, "loss": 0.0114, "step": 41510 }, { "grad_norm": 0.18141497671604156, "learning_rate": 6.776032862795033e-05, "loss": 0.0111, "step": 41520 }, { "grad_norm": 0.20587414503097534, "learning_rate": 6.774487122244956e-05, "loss": 0.0113, "step": 41530 }, { "grad_norm": 0.17649151384830475, "learning_rate": 6.772941187639637e-05, "loss": 0.0093, "step": 41540 }, { "grad_norm": 0.17705844342708588, "learning_rate": 6.771395059148134e-05, "loss": 0.0122, "step": 41550 }, { "grad_norm": 0.21234829723834991, "learning_rate": 6.76984873693953e-05, "loss": 0.012, "step": 41560 }, { "grad_norm": 0.151505708694458, "learning_rate": 6.768302221182931e-05, "loss": 0.0113, "step": 41570 }, { "grad_norm": 0.2358543872833252, "learning_rate": 6.766755512047457e-05, "loss": 0.0115, "step": 41580 }, { "grad_norm": 0.18678231537342072, "learning_rate": 6.765208609702259e-05, "loss": 0.0123, "step": 41590 }, { "grad_norm": 0.17370262742042542, "learning_rate": 6.763661514316499e-05, "loss": 0.0093, "step": 41600 }, { "grad_norm": 0.31848639249801636, "learning_rate": 6.76211422605937e-05, "loss": 0.0096, "step": 41610 }, { "grad_norm": 0.1609271913766861, "learning_rate": 6.760566745100076e-05, "loss": 0.0118, "step": 41620 }, { "grad_norm": 0.15692929923534393, "learning_rate": 6.75901907160785e-05, "loss": 0.0122, "step": 41630 }, { "grad_norm": 0.19279785454273224, "learning_rate": 6.757471205751943e-05, "loss": 0.0113, "step": 41640 }, { "grad_norm": 0.19921410083770752, "learning_rate": 6.75592314770163e-05, "loss": 0.0094, "step": 41650 }, { "grad_norm": 0.22534185647964478, "learning_rate": 6.7543748976262e-05, "loss": 0.0112, "step": 41660 }, { "grad_norm": 0.1889408826828003, "learning_rate": 6.752826455694968e-05, "loss": 0.0123, "step": 41670 }, { "grad_norm": 0.28598347306251526, "learning_rate": 6.751277822077271e-05, "loss": 0.0107, "step": 41680 }, { "grad_norm": 0.3408801555633545, "learning_rate": 6.749728996942465e-05, "loss": 0.0123, "step": 41690 }, { "grad_norm": 0.19104185700416565, "learning_rate": 6.748179980459924e-05, "loss": 0.0143, "step": 41700 }, { "grad_norm": 0.21308110654354095, "learning_rate": 6.746630772799052e-05, "loss": 0.0123, "step": 41710 }, { "grad_norm": 0.18007203936576843, "learning_rate": 6.745081374129262e-05, "loss": 0.0114, "step": 41720 }, { "grad_norm": 0.2337840497493744, "learning_rate": 6.743531784619998e-05, "loss": 0.0118, "step": 41730 }, { "grad_norm": 0.18536292016506195, "learning_rate": 6.741982004440719e-05, "loss": 0.0123, "step": 41740 }, { "grad_norm": 0.1865270882844925, "learning_rate": 6.740432033760907e-05, "loss": 0.0121, "step": 41750 }, { "grad_norm": 0.2419637143611908, "learning_rate": 6.738881872750066e-05, "loss": 0.0119, "step": 41760 }, { "grad_norm": 0.16338570415973663, "learning_rate": 6.737331521577715e-05, "loss": 0.0111, "step": 41770 }, { "grad_norm": 0.1779428869485855, "learning_rate": 6.735780980413403e-05, "loss": 0.0146, "step": 41780 }, { "grad_norm": 0.15805178880691528, "learning_rate": 6.734230249426692e-05, "loss": 0.0118, "step": 41790 }, { "grad_norm": 0.23832963407039642, "learning_rate": 6.732679328787168e-05, "loss": 0.0113, "step": 41800 }, { "grad_norm": 0.15428325533866882, "learning_rate": 6.731128218664438e-05, "loss": 0.0105, "step": 41810 }, { "grad_norm": 0.17917251586914062, "learning_rate": 6.729576919228129e-05, "loss": 0.0115, "step": 41820 }, { "grad_norm": 0.15574094653129578, "learning_rate": 6.728025430647888e-05, "loss": 0.0132, "step": 41830 }, { "grad_norm": 0.13869531452655792, "learning_rate": 6.726473753093383e-05, "loss": 0.0116, "step": 41840 }, { "grad_norm": 0.15068823099136353, "learning_rate": 6.724921886734305e-05, "loss": 0.013, "step": 41850 }, { "grad_norm": 0.14420390129089355, "learning_rate": 6.723369831740363e-05, "loss": 0.011, "step": 41860 }, { "grad_norm": 0.17296746373176575, "learning_rate": 6.721817588281288e-05, "loss": 0.0105, "step": 41870 }, { "grad_norm": 0.2193671017885208, "learning_rate": 6.720265156526828e-05, "loss": 0.0129, "step": 41880 }, { "grad_norm": 0.22941017150878906, "learning_rate": 6.718712536646758e-05, "loss": 0.0117, "step": 41890 }, { "grad_norm": 0.20184838771820068, "learning_rate": 6.71715972881087e-05, "loss": 0.0134, "step": 41900 }, { "grad_norm": 0.17320884764194489, "learning_rate": 6.715606733188974e-05, "loss": 0.011, "step": 41910 }, { "grad_norm": 0.18011854588985443, "learning_rate": 6.714053549950909e-05, "loss": 0.0101, "step": 41920 }, { "grad_norm": 0.13750216364860535, "learning_rate": 6.71250017926652e-05, "loss": 0.0129, "step": 41930 }, { "grad_norm": 0.1566983461380005, "learning_rate": 6.71094662130569e-05, "loss": 0.011, "step": 41940 }, { "grad_norm": 0.1746659129858017, "learning_rate": 6.709392876238307e-05, "loss": 0.0128, "step": 41950 }, { "grad_norm": 0.2704715430736542, "learning_rate": 6.707838944234289e-05, "loss": 0.0124, "step": 41960 }, { "grad_norm": 0.199392631649971, "learning_rate": 6.706284825463574e-05, "loss": 0.0121, "step": 41970 }, { "grad_norm": 0.14113749563694, "learning_rate": 6.704730520096114e-05, "loss": 0.0107, "step": 41980 }, { "grad_norm": 0.2249227613210678, "learning_rate": 6.703176028301888e-05, "loss": 0.0099, "step": 41990 }, { "grad_norm": 0.2633192241191864, "learning_rate": 6.701621350250892e-05, "loss": 0.0104, "step": 42000 }, { "grad_norm": 0.3087642192840576, "learning_rate": 6.700066486113144e-05, "loss": 0.0124, "step": 42010 }, { "grad_norm": 0.13161154091358185, "learning_rate": 6.69851143605868e-05, "loss": 0.0108, "step": 42020 }, { "grad_norm": 0.21481427550315857, "learning_rate": 6.696956200257561e-05, "loss": 0.0101, "step": 42030 }, { "grad_norm": 0.1625482439994812, "learning_rate": 6.695400778879863e-05, "loss": 0.0102, "step": 42040 }, { "grad_norm": 0.14090022444725037, "learning_rate": 6.693845172095683e-05, "loss": 0.0116, "step": 42050 }, { "grad_norm": 0.2343263328075409, "learning_rate": 6.692289380075142e-05, "loss": 0.0146, "step": 42060 }, { "grad_norm": 0.2820674777030945, "learning_rate": 6.690733402988379e-05, "loss": 0.0171, "step": 42070 }, { "grad_norm": 0.1485965996980667, "learning_rate": 6.689177241005553e-05, "loss": 0.0131, "step": 42080 }, { "grad_norm": 0.2650907039642334, "learning_rate": 6.687620894296844e-05, "loss": 0.011, "step": 42090 }, { "grad_norm": 0.16287405788898468, "learning_rate": 6.686064363032451e-05, "loss": 0.011, "step": 42100 }, { "grad_norm": 0.18936260044574738, "learning_rate": 6.684507647382596e-05, "loss": 0.0129, "step": 42110 }, { "grad_norm": 0.13603685796260834, "learning_rate": 6.682950747517513e-05, "loss": 0.012, "step": 42120 }, { "grad_norm": 0.15316691994667053, "learning_rate": 6.68139366360747e-05, "loss": 0.0107, "step": 42130 }, { "grad_norm": 0.148874431848526, "learning_rate": 6.679836395822744e-05, "loss": 0.0117, "step": 42140 }, { "grad_norm": 0.26840540766716003, "learning_rate": 6.678278944333633e-05, "loss": 0.012, "step": 42150 }, { "grad_norm": 0.18288809061050415, "learning_rate": 6.676721309310462e-05, "loss": 0.0127, "step": 42160 }, { "grad_norm": 0.43005117774009705, "learning_rate": 6.67516349092357e-05, "loss": 0.0135, "step": 42170 }, { "grad_norm": 0.18939924240112305, "learning_rate": 6.673605489343315e-05, "loss": 0.0111, "step": 42180 }, { "grad_norm": 0.18387964367866516, "learning_rate": 6.672047304740081e-05, "loss": 0.012, "step": 42190 }, { "grad_norm": 0.14422547817230225, "learning_rate": 6.670488937284268e-05, "loss": 0.0128, "step": 42200 }, { "grad_norm": 0.19432474672794342, "learning_rate": 6.668930387146294e-05, "loss": 0.0109, "step": 42210 }, { "grad_norm": 0.16280606389045715, "learning_rate": 6.667371654496605e-05, "loss": 0.0121, "step": 42220 }, { "grad_norm": 0.1562511920928955, "learning_rate": 6.665812739505659e-05, "loss": 0.0105, "step": 42230 }, { "grad_norm": 0.19902846217155457, "learning_rate": 6.664253642343935e-05, "loss": 0.0108, "step": 42240 }, { "grad_norm": 0.12362900376319885, "learning_rate": 6.662694363181935e-05, "loss": 0.0122, "step": 42250 }, { "grad_norm": 0.15149962902069092, "learning_rate": 6.66113490219018e-05, "loss": 0.0145, "step": 42260 }, { "grad_norm": 0.1866254061460495, "learning_rate": 6.659575259539207e-05, "loss": 0.0116, "step": 42270 }, { "grad_norm": 0.22548188269138336, "learning_rate": 6.65801543539958e-05, "loss": 0.0132, "step": 42280 }, { "grad_norm": 0.23535187542438507, "learning_rate": 6.65645542994188e-05, "loss": 0.0127, "step": 42290 }, { "grad_norm": 0.19983318448066711, "learning_rate": 6.654895243336702e-05, "loss": 0.0107, "step": 42300 }, { "grad_norm": 0.20228426158428192, "learning_rate": 6.65333487575467e-05, "loss": 0.012, "step": 42310 }, { "grad_norm": 0.15896707773208618, "learning_rate": 6.651774327366422e-05, "loss": 0.0113, "step": 42320 }, { "grad_norm": 0.19902348518371582, "learning_rate": 6.650213598342613e-05, "loss": 0.0135, "step": 42330 }, { "grad_norm": 0.1759963035583496, "learning_rate": 6.648652688853931e-05, "loss": 0.0115, "step": 42340 }, { "grad_norm": 0.17475803196430206, "learning_rate": 6.647091599071066e-05, "loss": 0.0121, "step": 42350 }, { "grad_norm": 0.16111791133880615, "learning_rate": 6.645530329164743e-05, "loss": 0.0111, "step": 42360 }, { "grad_norm": 0.22906115651130676, "learning_rate": 6.643968879305694e-05, "loss": 0.0145, "step": 42370 }, { "grad_norm": 0.19087699055671692, "learning_rate": 6.642407249664683e-05, "loss": 0.0126, "step": 42380 }, { "grad_norm": 0.17367859184741974, "learning_rate": 6.640845440412483e-05, "loss": 0.0123, "step": 42390 }, { "grad_norm": 0.1911129206418991, "learning_rate": 6.639283451719893e-05, "loss": 0.0109, "step": 42400 }, { "grad_norm": 0.22358903288841248, "learning_rate": 6.637721283757729e-05, "loss": 0.0115, "step": 42410 }, { "grad_norm": 0.38469359278678894, "learning_rate": 6.636158936696828e-05, "loss": 0.0132, "step": 42420 }, { "grad_norm": 0.15672148764133453, "learning_rate": 6.634596410708047e-05, "loss": 0.0135, "step": 42430 }, { "grad_norm": 0.35990282893180847, "learning_rate": 6.633033705962257e-05, "loss": 0.0141, "step": 42440 }, { "grad_norm": 0.13310080766677856, "learning_rate": 6.631470822630359e-05, "loss": 0.0124, "step": 42450 }, { "grad_norm": 0.1997712403535843, "learning_rate": 6.629907760883263e-05, "loss": 0.0103, "step": 42460 }, { "grad_norm": 0.19031397998332977, "learning_rate": 6.628344520891905e-05, "loss": 0.0131, "step": 42470 }, { "grad_norm": 0.1509297490119934, "learning_rate": 6.62678110282724e-05, "loss": 0.0116, "step": 42480 }, { "grad_norm": 0.2308516502380371, "learning_rate": 6.625217506860237e-05, "loss": 0.0116, "step": 42490 }, { "grad_norm": 0.18636322021484375, "learning_rate": 6.623653733161892e-05, "loss": 0.0119, "step": 42500 }, { "grad_norm": 0.13616783916950226, "learning_rate": 6.622089781903217e-05, "loss": 0.0104, "step": 42510 }, { "grad_norm": 0.3543224036693573, "learning_rate": 6.620525653255241e-05, "loss": 0.0128, "step": 42520 }, { "grad_norm": 0.1592467874288559, "learning_rate": 6.618961347389015e-05, "loss": 0.0117, "step": 42530 }, { "grad_norm": 0.14615772664546967, "learning_rate": 6.617396864475613e-05, "loss": 0.0122, "step": 42540 }, { "grad_norm": 0.14589037001132965, "learning_rate": 6.61583220468612e-05, "loss": 0.0116, "step": 42550 }, { "grad_norm": 0.19004791975021362, "learning_rate": 6.614267368191645e-05, "loss": 0.0121, "step": 42560 }, { "grad_norm": 0.1612478494644165, "learning_rate": 6.61270235516332e-05, "loss": 0.011, "step": 42570 }, { "grad_norm": 0.2255704551935196, "learning_rate": 6.611137165772287e-05, "loss": 0.0106, "step": 42580 }, { "grad_norm": 0.16219565272331238, "learning_rate": 6.60957180018972e-05, "loss": 0.0128, "step": 42590 }, { "grad_norm": 0.22176115214824677, "learning_rate": 6.608006258586797e-05, "loss": 0.0109, "step": 42600 }, { "grad_norm": 0.15057726204395294, "learning_rate": 6.606440541134731e-05, "loss": 0.0113, "step": 42610 }, { "grad_norm": 0.1766381710767746, "learning_rate": 6.60487464800474e-05, "loss": 0.0105, "step": 42620 }, { "grad_norm": 0.21913111209869385, "learning_rate": 6.603308579368071e-05, "loss": 0.0122, "step": 42630 }, { "grad_norm": 0.21500378847122192, "learning_rate": 6.601742335395987e-05, "loss": 0.0118, "step": 42640 }, { "grad_norm": 0.1975707858800888, "learning_rate": 6.600175916259769e-05, "loss": 0.0105, "step": 42650 }, { "grad_norm": 0.17485415935516357, "learning_rate": 6.598609322130718e-05, "loss": 0.0144, "step": 42660 }, { "grad_norm": 0.20885445177555084, "learning_rate": 6.597042553180155e-05, "loss": 0.0131, "step": 42670 }, { "grad_norm": 0.18102693557739258, "learning_rate": 6.59547560957942e-05, "loss": 0.0119, "step": 42680 }, { "grad_norm": 0.19351628422737122, "learning_rate": 6.59390849149987e-05, "loss": 0.0133, "step": 42690 }, { "grad_norm": 0.19640955328941345, "learning_rate": 6.592341199112886e-05, "loss": 0.0128, "step": 42700 }, { "grad_norm": 0.132390096783638, "learning_rate": 6.590773732589861e-05, "loss": 0.0115, "step": 42710 }, { "grad_norm": 0.20372308790683746, "learning_rate": 6.589206092102213e-05, "loss": 0.0118, "step": 42720 }, { "grad_norm": 0.1854451596736908, "learning_rate": 6.587638277821375e-05, "loss": 0.0125, "step": 42730 }, { "grad_norm": 0.12028245627880096, "learning_rate": 6.586070289918803e-05, "loss": 0.0109, "step": 42740 }, { "grad_norm": 0.14324447512626648, "learning_rate": 6.584502128565968e-05, "loss": 0.0123, "step": 42750 }, { "grad_norm": 0.10278015583753586, "learning_rate": 6.582933793934363e-05, "loss": 0.0088, "step": 42760 }, { "grad_norm": 0.13288789987564087, "learning_rate": 6.581365286195498e-05, "loss": 0.0107, "step": 42770 }, { "grad_norm": 0.1503102034330368, "learning_rate": 6.579796605520903e-05, "loss": 0.0124, "step": 42780 }, { "grad_norm": 0.24452869594097137, "learning_rate": 6.578227752082127e-05, "loss": 0.0125, "step": 42790 }, { "grad_norm": 0.14017201960086823, "learning_rate": 6.576658726050735e-05, "loss": 0.0103, "step": 42800 }, { "grad_norm": 0.1704624891281128, "learning_rate": 6.575089527598316e-05, "loss": 0.0113, "step": 42810 }, { "grad_norm": 0.19548512995243073, "learning_rate": 6.573520156896476e-05, "loss": 0.0122, "step": 42820 }, { "grad_norm": 0.25846338272094727, "learning_rate": 6.571950614116835e-05, "loss": 0.0105, "step": 42830 }, { "grad_norm": 0.16117890179157257, "learning_rate": 6.570380899431039e-05, "loss": 0.0113, "step": 42840 }, { "grad_norm": 0.16639815270900726, "learning_rate": 6.568811013010749e-05, "loss": 0.013, "step": 42850 }, { "grad_norm": 0.2381771206855774, "learning_rate": 6.567240955027645e-05, "loss": 0.0104, "step": 42860 }, { "grad_norm": 0.18900513648986816, "learning_rate": 6.565670725653427e-05, "loss": 0.0105, "step": 42870 }, { "grad_norm": 0.17884357273578644, "learning_rate": 6.564100325059812e-05, "loss": 0.0108, "step": 42880 }, { "grad_norm": 0.20801497995853424, "learning_rate": 6.562529753418538e-05, "loss": 0.0121, "step": 42890 }, { "grad_norm": 0.21456189453601837, "learning_rate": 6.56095901090136e-05, "loss": 0.0113, "step": 42900 }, { "grad_norm": 0.1452694684267044, "learning_rate": 6.55938809768005e-05, "loss": 0.0108, "step": 42910 }, { "grad_norm": 0.17567507922649384, "learning_rate": 6.557817013926402e-05, "loss": 0.0122, "step": 42920 }, { "grad_norm": 0.12656474113464355, "learning_rate": 6.556245759812227e-05, "loss": 0.0106, "step": 42930 }, { "grad_norm": 0.16777031123638153, "learning_rate": 6.554674335509357e-05, "loss": 0.0106, "step": 42940 }, { "grad_norm": 0.26294273138046265, "learning_rate": 6.553102741189638e-05, "loss": 0.0111, "step": 42950 }, { "grad_norm": 0.14769446849822998, "learning_rate": 6.55153097702494e-05, "loss": 0.01, "step": 42960 }, { "grad_norm": 0.13094022870063782, "learning_rate": 6.549959043187146e-05, "loss": 0.0115, "step": 42970 }, { "grad_norm": 0.34080004692077637, "learning_rate": 6.54838693984816e-05, "loss": 0.0097, "step": 42980 }, { "grad_norm": 0.13497045636177063, "learning_rate": 6.546814667179909e-05, "loss": 0.0118, "step": 42990 }, { "grad_norm": 0.15734243392944336, "learning_rate": 6.545242225354328e-05, "loss": 0.0126, "step": 43000 }, { "grad_norm": 0.18154621124267578, "learning_rate": 6.543669614543383e-05, "loss": 0.0112, "step": 43010 }, { "grad_norm": 0.15098612010478973, "learning_rate": 6.542096834919049e-05, "loss": 0.0115, "step": 43020 }, { "grad_norm": 0.30053648352622986, "learning_rate": 6.540523886653324e-05, "loss": 0.0105, "step": 43030 }, { "grad_norm": 0.15517540276050568, "learning_rate": 6.538950769918222e-05, "loss": 0.0132, "step": 43040 }, { "grad_norm": 0.1667247712612152, "learning_rate": 6.537377484885779e-05, "loss": 0.0115, "step": 43050 }, { "grad_norm": 0.1975499987602234, "learning_rate": 6.535804031728045e-05, "loss": 0.011, "step": 43060 }, { "grad_norm": 0.1597767174243927, "learning_rate": 6.534230410617092e-05, "loss": 0.0126, "step": 43070 }, { "grad_norm": 0.381073921918869, "learning_rate": 6.532656621725007e-05, "loss": 0.0115, "step": 43080 }, { "grad_norm": 0.25237956643104553, "learning_rate": 6.531082665223898e-05, "loss": 0.0134, "step": 43090 }, { "grad_norm": 0.21719686686992645, "learning_rate": 6.529508541285889e-05, "loss": 0.0123, "step": 43100 }, { "grad_norm": 0.16153466701507568, "learning_rate": 6.527934250083128e-05, "loss": 0.0094, "step": 43110 }, { "grad_norm": 0.20271988213062286, "learning_rate": 6.526359791787775e-05, "loss": 0.0123, "step": 43120 }, { "grad_norm": 0.202531099319458, "learning_rate": 6.524785166572009e-05, "loss": 0.0116, "step": 43130 }, { "grad_norm": 0.22815416753292084, "learning_rate": 6.523210374608029e-05, "loss": 0.0111, "step": 43140 }, { "grad_norm": 0.18554899096488953, "learning_rate": 6.521635416068054e-05, "loss": 0.0123, "step": 43150 }, { "grad_norm": 0.1545945256948471, "learning_rate": 6.520060291124317e-05, "loss": 0.0122, "step": 43160 }, { "grad_norm": 0.14440010488033295, "learning_rate": 6.518484999949072e-05, "loss": 0.0149, "step": 43170 }, { "grad_norm": 0.1786210983991623, "learning_rate": 6.516909542714591e-05, "loss": 0.0115, "step": 43180 }, { "grad_norm": 0.2552957832813263, "learning_rate": 6.515333919593161e-05, "loss": 0.01, "step": 43190 }, { "grad_norm": 0.1334383189678192, "learning_rate": 6.513758130757094e-05, "loss": 0.0125, "step": 43200 }, { "grad_norm": 0.1638573855161667, "learning_rate": 6.512182176378713e-05, "loss": 0.0101, "step": 43210 }, { "grad_norm": 0.165035218000412, "learning_rate": 6.510606056630362e-05, "loss": 0.0119, "step": 43220 }, { "grad_norm": 0.17006497085094452, "learning_rate": 6.509029771684405e-05, "loss": 0.0128, "step": 43230 }, { "grad_norm": 0.15245886147022247, "learning_rate": 6.50745332171322e-05, "loss": 0.0113, "step": 43240 }, { "grad_norm": 0.14222122728824615, "learning_rate": 6.505876706889207e-05, "loss": 0.01, "step": 43250 }, { "grad_norm": 0.27051177620887756, "learning_rate": 6.504299927384781e-05, "loss": 0.0109, "step": 43260 }, { "grad_norm": 0.15373627841472626, "learning_rate": 6.502722983372378e-05, "loss": 0.0123, "step": 43270 }, { "grad_norm": 0.18218721449375153, "learning_rate": 6.501145875024446e-05, "loss": 0.0134, "step": 43280 }, { "grad_norm": 0.1475587636232376, "learning_rate": 6.499568602513459e-05, "loss": 0.0092, "step": 43290 }, { "grad_norm": 0.21315541863441467, "learning_rate": 6.497991166011903e-05, "loss": 0.0109, "step": 43300 }, { "grad_norm": 0.13165892660617828, "learning_rate": 6.49641356569229e-05, "loss": 0.0105, "step": 43310 }, { "grad_norm": 0.19549629092216492, "learning_rate": 6.494835801727135e-05, "loss": 0.012, "step": 43320 }, { "grad_norm": 0.3746725022792816, "learning_rate": 6.493257874288986e-05, "loss": 0.0113, "step": 43330 }, { "grad_norm": 0.22871650755405426, "learning_rate": 6.491679783550399e-05, "loss": 0.0112, "step": 43340 }, { "grad_norm": 0.138735830783844, "learning_rate": 6.490101529683957e-05, "loss": 0.0111, "step": 43350 }, { "grad_norm": 0.175825297832489, "learning_rate": 6.488523112862248e-05, "loss": 0.0104, "step": 43360 }, { "grad_norm": 0.14675097167491913, "learning_rate": 6.486944533257891e-05, "loss": 0.0121, "step": 43370 }, { "grad_norm": 0.21160365641117096, "learning_rate": 6.485365791043517e-05, "loss": 0.0129, "step": 43380 }, { "grad_norm": 0.18378165364265442, "learning_rate": 6.483786886391772e-05, "loss": 0.0112, "step": 43390 }, { "grad_norm": 0.15977142751216888, "learning_rate": 6.482207819475323e-05, "loss": 0.0098, "step": 43400 }, { "grad_norm": 0.23476071655750275, "learning_rate": 6.480628590466857e-05, "loss": 0.0116, "step": 43410 }, { "grad_norm": 0.14118781685829163, "learning_rate": 6.479049199539074e-05, "loss": 0.0103, "step": 43420 }, { "grad_norm": 0.17422568798065186, "learning_rate": 6.477469646864693e-05, "loss": 0.0106, "step": 43430 }, { "grad_norm": 0.14313775300979614, "learning_rate": 6.475889932616454e-05, "loss": 0.0107, "step": 43440 }, { "grad_norm": 0.14662612974643707, "learning_rate": 6.474310056967111e-05, "loss": 0.0097, "step": 43450 }, { "grad_norm": 0.11806558072566986, "learning_rate": 6.472730020089437e-05, "loss": 0.0105, "step": 43460 }, { "grad_norm": 0.19246791303157806, "learning_rate": 6.471149822156222e-05, "loss": 0.0106, "step": 43470 }, { "grad_norm": 0.21989816427230835, "learning_rate": 6.469569463340275e-05, "loss": 0.0122, "step": 43480 }, { "grad_norm": 0.36873912811279297, "learning_rate": 6.467988943814418e-05, "loss": 0.0123, "step": 43490 }, { "grad_norm": 0.16295574605464935, "learning_rate": 6.4664082637515e-05, "loss": 0.012, "step": 43500 }, { "grad_norm": 0.2577858567237854, "learning_rate": 6.464827423324377e-05, "loss": 0.0121, "step": 43510 }, { "grad_norm": 0.35291263461112976, "learning_rate": 6.463246422705931e-05, "loss": 0.0106, "step": 43520 }, { "grad_norm": 0.15382921695709229, "learning_rate": 6.461665262069055e-05, "loss": 0.0118, "step": 43530 }, { "grad_norm": 0.12346448749303818, "learning_rate": 6.460083941586665e-05, "loss": 0.0112, "step": 43540 }, { "grad_norm": 0.2751938998699188, "learning_rate": 6.45850246143169e-05, "loss": 0.0124, "step": 43550 }, { "grad_norm": 0.17218977212905884, "learning_rate": 6.456920821777077e-05, "loss": 0.013, "step": 43560 }, { "grad_norm": 0.16855385899543762, "learning_rate": 6.455339022795796e-05, "loss": 0.0114, "step": 43570 }, { "grad_norm": 0.28605109453201294, "learning_rate": 6.453757064660825e-05, "loss": 0.0101, "step": 43580 }, { "grad_norm": 0.26200029253959656, "learning_rate": 6.452174947545169e-05, "loss": 0.0125, "step": 43590 }, { "grad_norm": 0.28831809759140015, "learning_rate": 6.450592671621842e-05, "loss": 0.0134, "step": 43600 }, { "grad_norm": 0.2189503014087677, "learning_rate": 6.449010237063882e-05, "loss": 0.0118, "step": 43610 }, { "grad_norm": 0.22185583412647247, "learning_rate": 6.447427644044343e-05, "loss": 0.0119, "step": 43620 }, { "grad_norm": 0.19820024073123932, "learning_rate": 6.44584489273629e-05, "loss": 0.0119, "step": 43630 }, { "grad_norm": 0.1729305386543274, "learning_rate": 6.444261983312817e-05, "loss": 0.0101, "step": 43640 }, { "grad_norm": 0.1903572976589203, "learning_rate": 6.442678915947023e-05, "loss": 0.0121, "step": 43650 }, { "grad_norm": 0.1690843403339386, "learning_rate": 6.441095690812032e-05, "loss": 0.0108, "step": 43660 }, { "grad_norm": 0.14633680880069733, "learning_rate": 6.439512308080982e-05, "loss": 0.0096, "step": 43670 }, { "grad_norm": 0.1593473106622696, "learning_rate": 6.43792876792703e-05, "loss": 0.0105, "step": 43680 }, { "grad_norm": 0.2125176191329956, "learning_rate": 6.43634507052335e-05, "loss": 0.0117, "step": 43690 }, { "grad_norm": 0.2794451415538788, "learning_rate": 6.434761216043135e-05, "loss": 0.0112, "step": 43700 }, { "grad_norm": 0.12019964307546616, "learning_rate": 6.433177204659589e-05, "loss": 0.0114, "step": 43710 }, { "grad_norm": 0.16411474347114563, "learning_rate": 6.431593036545938e-05, "loss": 0.0114, "step": 43720 }, { "grad_norm": 0.45930373668670654, "learning_rate": 6.430008711875426e-05, "loss": 0.0108, "step": 43730 }, { "grad_norm": 0.16009125113487244, "learning_rate": 6.42842423082131e-05, "loss": 0.0097, "step": 43740 }, { "grad_norm": 0.1599859595298767, "learning_rate": 6.426839593556868e-05, "loss": 0.0118, "step": 43750 }, { "grad_norm": 0.2446577548980713, "learning_rate": 6.425254800255392e-05, "loss": 0.0124, "step": 43760 }, { "grad_norm": 0.18504095077514648, "learning_rate": 6.423669851090196e-05, "loss": 0.012, "step": 43770 }, { "grad_norm": 0.16024859249591827, "learning_rate": 6.422084746234604e-05, "loss": 0.0113, "step": 43780 }, { "grad_norm": 0.1747094690799713, "learning_rate": 6.420499485861965e-05, "loss": 0.0112, "step": 43790 }, { "grad_norm": 0.23062002658843994, "learning_rate": 6.418914070145636e-05, "loss": 0.0141, "step": 43800 }, { "grad_norm": 0.1335931420326233, "learning_rate": 6.417328499258996e-05, "loss": 0.0111, "step": 43810 }, { "grad_norm": 0.19176815450191498, "learning_rate": 6.415742773375446e-05, "loss": 0.0108, "step": 43820 }, { "grad_norm": 0.14541812241077423, "learning_rate": 6.414156892668393e-05, "loss": 0.0096, "step": 43830 }, { "grad_norm": 0.14294104278087616, "learning_rate": 6.412570857311267e-05, "loss": 0.0118, "step": 43840 }, { "grad_norm": 0.1726667582988739, "learning_rate": 6.410984667477518e-05, "loss": 0.0119, "step": 43850 }, { "grad_norm": 0.18340949714183807, "learning_rate": 6.409398323340607e-05, "loss": 0.0115, "step": 43860 }, { "grad_norm": 0.17562372982501984, "learning_rate": 6.407811825074012e-05, "loss": 0.0116, "step": 43870 }, { "grad_norm": 0.22320061922073364, "learning_rate": 6.406225172851234e-05, "loss": 0.0116, "step": 43880 }, { "grad_norm": 0.1981271505355835, "learning_rate": 6.404638366845786e-05, "loss": 0.0133, "step": 43890 }, { "grad_norm": 0.2044650763273239, "learning_rate": 6.403051407231196e-05, "loss": 0.0112, "step": 43900 }, { "grad_norm": 0.17460395395755768, "learning_rate": 6.401464294181016e-05, "loss": 0.0124, "step": 43910 }, { "grad_norm": 0.1477164924144745, "learning_rate": 6.399877027868808e-05, "loss": 0.0115, "step": 43920 }, { "grad_norm": 0.1658921241760254, "learning_rate": 6.39828960846815e-05, "loss": 0.0119, "step": 43930 }, { "grad_norm": 0.16896666586399078, "learning_rate": 6.396702036152644e-05, "loss": 0.0108, "step": 43940 }, { "grad_norm": 0.20953449606895447, "learning_rate": 6.395114311095905e-05, "loss": 0.0115, "step": 43950 }, { "grad_norm": 0.17657563090324402, "learning_rate": 6.393526433471559e-05, "loss": 0.0113, "step": 43960 }, { "grad_norm": 0.2586477994918823, "learning_rate": 6.391938403453262e-05, "loss": 0.0118, "step": 43970 }, { "grad_norm": 0.15295422077178955, "learning_rate": 6.390350221214671e-05, "loss": 0.0125, "step": 43980 }, { "grad_norm": 0.1982533484697342, "learning_rate": 6.38876188692947e-05, "loss": 0.0117, "step": 43990 }, { "grad_norm": 0.22515997290611267, "learning_rate": 6.387173400771358e-05, "loss": 0.0101, "step": 44000 }, { "grad_norm": 0.13119995594024658, "learning_rate": 6.385584762914047e-05, "loss": 0.0091, "step": 44010 }, { "grad_norm": 0.13589943945407867, "learning_rate": 6.38399597353127e-05, "loss": 0.0098, "step": 44020 }, { "grad_norm": 0.19570788741111755, "learning_rate": 6.382407032796775e-05, "loss": 0.0108, "step": 44030 }, { "grad_norm": 0.21715056896209717, "learning_rate": 6.380817940884325e-05, "loss": 0.0109, "step": 44040 }, { "grad_norm": 0.13524240255355835, "learning_rate": 6.379228697967702e-05, "loss": 0.0105, "step": 44050 }, { "grad_norm": 0.1826813817024231, "learning_rate": 6.3776393042207e-05, "loss": 0.0114, "step": 44060 }, { "grad_norm": 0.1466422826051712, "learning_rate": 6.376049759817137e-05, "loss": 0.0106, "step": 44070 }, { "grad_norm": 0.20997360348701477, "learning_rate": 6.37446006493084e-05, "loss": 0.0101, "step": 44080 }, { "grad_norm": 0.26489609479904175, "learning_rate": 6.372870219735656e-05, "loss": 0.0122, "step": 44090 }, { "grad_norm": 0.15895141661167145, "learning_rate": 6.371280224405452e-05, "loss": 0.0121, "step": 44100 }, { "grad_norm": 0.13433034718036652, "learning_rate": 6.369690079114101e-05, "loss": 0.0123, "step": 44110 }, { "grad_norm": 0.17554712295532227, "learning_rate": 6.368099784035504e-05, "loss": 0.0109, "step": 44120 }, { "grad_norm": 0.16494512557983398, "learning_rate": 6.366509339343572e-05, "loss": 0.012, "step": 44130 }, { "grad_norm": 0.1523682326078415, "learning_rate": 6.364918745212232e-05, "loss": 0.0112, "step": 44140 }, { "grad_norm": 0.14903020858764648, "learning_rate": 6.363328001815431e-05, "loss": 0.0134, "step": 44150 }, { "grad_norm": 0.18917223811149597, "learning_rate": 6.361737109327128e-05, "loss": 0.0142, "step": 44160 }, { "grad_norm": 0.18752378225326538, "learning_rate": 6.360146067921301e-05, "loss": 0.013, "step": 44170 }, { "grad_norm": 0.19359755516052246, "learning_rate": 6.35855487777195e-05, "loss": 0.0111, "step": 44180 }, { "grad_norm": 0.16606584191322327, "learning_rate": 6.356963539053076e-05, "loss": 0.0106, "step": 44190 }, { "grad_norm": 0.13423000276088715, "learning_rate": 6.355372051938711e-05, "loss": 0.0112, "step": 44200 }, { "grad_norm": 0.18702001869678497, "learning_rate": 6.353780416602894e-05, "loss": 0.0127, "step": 44210 }, { "grad_norm": 0.17669092118740082, "learning_rate": 6.352188633219689e-05, "loss": 0.0119, "step": 44220 }, { "grad_norm": 0.1797039955854416, "learning_rate": 6.350596701963166e-05, "loss": 0.0105, "step": 44230 }, { "grad_norm": 0.16168060898780823, "learning_rate": 6.349004623007419e-05, "loss": 0.0139, "step": 44240 }, { "grad_norm": 0.1917695701122284, "learning_rate": 6.347412396526555e-05, "loss": 0.011, "step": 44250 }, { "grad_norm": 0.1537591516971588, "learning_rate": 6.345820022694696e-05, "loss": 0.0124, "step": 44260 }, { "grad_norm": 0.21921873092651367, "learning_rate": 6.344227501685984e-05, "loss": 0.0126, "step": 44270 }, { "grad_norm": 0.17059148848056793, "learning_rate": 6.342634833674572e-05, "loss": 0.0121, "step": 44280 }, { "grad_norm": 0.1894262582063675, "learning_rate": 6.341042018834635e-05, "loss": 0.0113, "step": 44290 }, { "grad_norm": 0.19702182710170746, "learning_rate": 6.339449057340359e-05, "loss": 0.0099, "step": 44300 }, { "grad_norm": 0.2137576788663864, "learning_rate": 6.337855949365945e-05, "loss": 0.0094, "step": 44310 }, { "grad_norm": 0.15336710214614868, "learning_rate": 6.336262695085619e-05, "loss": 0.0121, "step": 44320 }, { "grad_norm": 0.15363775193691254, "learning_rate": 6.334669294673612e-05, "loss": 0.0112, "step": 44330 }, { "grad_norm": 0.16884084045886993, "learning_rate": 6.333075748304179e-05, "loss": 0.0117, "step": 44340 }, { "grad_norm": 0.20173625648021698, "learning_rate": 6.331482056151585e-05, "loss": 0.0102, "step": 44350 }, { "grad_norm": 0.20023277401924133, "learning_rate": 6.329888218390117e-05, "loss": 0.0109, "step": 44360 }, { "grad_norm": 0.2044862061738968, "learning_rate": 6.328294235194072e-05, "loss": 0.0102, "step": 44370 }, { "grad_norm": 0.12860482931137085, "learning_rate": 6.326700106737765e-05, "loss": 0.011, "step": 44380 }, { "grad_norm": 0.21759316325187683, "learning_rate": 6.325105833195531e-05, "loss": 0.0101, "step": 44390 }, { "grad_norm": 0.13351692259311676, "learning_rate": 6.323511414741715e-05, "loss": 0.0109, "step": 44400 }, { "grad_norm": 0.16238687932491302, "learning_rate": 6.321916851550678e-05, "loss": 0.0101, "step": 44410 }, { "grad_norm": 0.13477176427841187, "learning_rate": 6.320322143796806e-05, "loss": 0.0127, "step": 44420 }, { "grad_norm": 0.26057368516921997, "learning_rate": 6.318727291654488e-05, "loss": 0.0105, "step": 44430 }, { "grad_norm": 0.25396478176116943, "learning_rate": 6.317132295298134e-05, "loss": 0.0109, "step": 44440 }, { "grad_norm": 0.15094895660877228, "learning_rate": 6.315537154902173e-05, "loss": 0.0119, "step": 44450 }, { "grad_norm": 0.19612260162830353, "learning_rate": 6.313941870641048e-05, "loss": 0.0104, "step": 44460 }, { "grad_norm": 0.19584408402442932, "learning_rate": 6.312346442689214e-05, "loss": 0.0098, "step": 44470 }, { "grad_norm": 0.29976966977119446, "learning_rate": 6.310750871221147e-05, "loss": 0.0102, "step": 44480 }, { "grad_norm": 0.15101061761379242, "learning_rate": 6.309155156411335e-05, "loss": 0.0091, "step": 44490 }, { "grad_norm": 0.21475879848003387, "learning_rate": 6.307559298434284e-05, "loss": 0.0128, "step": 44500 }, { "grad_norm": 0.14256693422794342, "learning_rate": 6.305963297464512e-05, "loss": 0.0093, "step": 44510 }, { "grad_norm": 0.11758948862552643, "learning_rate": 6.304367153676561e-05, "loss": 0.0103, "step": 44520 }, { "grad_norm": 0.49185389280319214, "learning_rate": 6.302770867244978e-05, "loss": 0.0121, "step": 44530 }, { "grad_norm": 0.16338972747325897, "learning_rate": 6.301174438344328e-05, "loss": 0.0106, "step": 44540 }, { "grad_norm": 0.18738430738449097, "learning_rate": 6.299577867149202e-05, "loss": 0.0117, "step": 44550 }, { "grad_norm": 0.13613750040531158, "learning_rate": 6.297981153834192e-05, "loss": 0.0107, "step": 44560 }, { "grad_norm": 0.13420189917087555, "learning_rate": 6.296384298573916e-05, "loss": 0.0138, "step": 44570 }, { "grad_norm": 0.1929696947336197, "learning_rate": 6.294787301543001e-05, "loss": 0.0112, "step": 44580 }, { "grad_norm": 0.18490128219127655, "learning_rate": 6.293190162916095e-05, "loss": 0.01, "step": 44590 }, { "grad_norm": 0.12023797631263733, "learning_rate": 6.291592882867855e-05, "loss": 0.0087, "step": 44600 }, { "grad_norm": 0.13292306661605835, "learning_rate": 6.28999546157296e-05, "loss": 0.0118, "step": 44610 }, { "grad_norm": 0.21578535437583923, "learning_rate": 6.288397899206102e-05, "loss": 0.0119, "step": 44620 }, { "grad_norm": 0.2368614673614502, "learning_rate": 6.286800195941984e-05, "loss": 0.0113, "step": 44630 }, { "grad_norm": 0.24560223519802094, "learning_rate": 6.285202351955334e-05, "loss": 0.0099, "step": 44640 }, { "grad_norm": 0.18452969193458557, "learning_rate": 6.283604367420887e-05, "loss": 0.0117, "step": 44650 }, { "grad_norm": 0.14610181748867035, "learning_rate": 6.282006242513394e-05, "loss": 0.0107, "step": 44660 }, { "grad_norm": 0.17375756800174713, "learning_rate": 6.280407977407628e-05, "loss": 0.0103, "step": 44670 }, { "grad_norm": 0.1746370792388916, "learning_rate": 6.27880957227837e-05, "loss": 0.0139, "step": 44680 }, { "grad_norm": 0.1590835452079773, "learning_rate": 6.27721102730042e-05, "loss": 0.0126, "step": 44690 }, { "grad_norm": 0.21425358951091766, "learning_rate": 6.275612342648592e-05, "loss": 0.0123, "step": 44700 }, { "grad_norm": 0.15058296918869019, "learning_rate": 6.274013518497716e-05, "loss": 0.0099, "step": 44710 }, { "grad_norm": 0.17212173342704773, "learning_rate": 6.272414555022636e-05, "loss": 0.0126, "step": 44720 }, { "grad_norm": 0.20297221839427948, "learning_rate": 6.270815452398215e-05, "loss": 0.012, "step": 44730 }, { "grad_norm": 0.3953636884689331, "learning_rate": 6.269216210799326e-05, "loss": 0.0105, "step": 44740 }, { "grad_norm": 0.17469638586044312, "learning_rate": 6.26761683040086e-05, "loss": 0.0108, "step": 44750 }, { "grad_norm": 0.2683488130569458, "learning_rate": 6.266017311377723e-05, "loss": 0.01, "step": 44760 }, { "grad_norm": 0.30613672733306885, "learning_rate": 6.264417653904839e-05, "loss": 0.0103, "step": 44770 }, { "grad_norm": 0.1391700804233551, "learning_rate": 6.262817858157139e-05, "loss": 0.0105, "step": 44780 }, { "grad_norm": 0.2671572268009186, "learning_rate": 6.261217924309576e-05, "loss": 0.0099, "step": 44790 }, { "grad_norm": 0.22809334099292755, "learning_rate": 6.259617852537118e-05, "loss": 0.01, "step": 44800 }, { "grad_norm": 0.18971221148967743, "learning_rate": 6.258017643014747e-05, "loss": 0.0133, "step": 44810 }, { "grad_norm": 0.12651537358760834, "learning_rate": 6.256417295917456e-05, "loss": 0.0102, "step": 44820 }, { "grad_norm": 0.22053900361061096, "learning_rate": 6.254816811420258e-05, "loss": 0.011, "step": 44830 }, { "grad_norm": 0.32897403836250305, "learning_rate": 6.253216189698183e-05, "loss": 0.0109, "step": 44840 }, { "grad_norm": 0.18229517340660095, "learning_rate": 6.251615430926267e-05, "loss": 0.0111, "step": 44850 }, { "grad_norm": 0.16120584309101105, "learning_rate": 6.25001453527957e-05, "loss": 0.0102, "step": 44860 }, { "grad_norm": 0.17557750642299652, "learning_rate": 6.248413502933164e-05, "loss": 0.0095, "step": 44870 }, { "grad_norm": 0.2125440239906311, "learning_rate": 6.246812334062133e-05, "loss": 0.0095, "step": 44880 }, { "grad_norm": 0.18992160260677338, "learning_rate": 6.245211028841579e-05, "loss": 0.0115, "step": 44890 }, { "grad_norm": 0.38854584097862244, "learning_rate": 6.24360958744662e-05, "loss": 0.0133, "step": 44900 }, { "grad_norm": 0.18159402906894684, "learning_rate": 6.242008010052387e-05, "loss": 0.0109, "step": 44910 }, { "grad_norm": 0.13049645721912384, "learning_rate": 6.240406296834024e-05, "loss": 0.0119, "step": 44920 }, { "grad_norm": 0.19059307873249054, "learning_rate": 6.238804447966694e-05, "loss": 0.0108, "step": 44930 }, { "grad_norm": 0.23003734648227692, "learning_rate": 6.237202463625573e-05, "loss": 0.0111, "step": 44940 }, { "grad_norm": 0.25928494334220886, "learning_rate": 6.235600343985848e-05, "loss": 0.0112, "step": 44950 }, { "grad_norm": 0.15783050656318665, "learning_rate": 6.233998089222729e-05, "loss": 0.0093, "step": 44960 }, { "grad_norm": 0.189896821975708, "learning_rate": 6.232395699511433e-05, "loss": 0.0101, "step": 44970 }, { "grad_norm": 0.12937766313552856, "learning_rate": 6.230793175027199e-05, "loss": 0.0088, "step": 44980 }, { "grad_norm": 0.26014238595962524, "learning_rate": 6.22919051594527e-05, "loss": 0.0122, "step": 44990 }, { "grad_norm": 0.30064529180526733, "learning_rate": 6.227587722440917e-05, "loss": 0.0122, "step": 45000 }, { "grad_norm": 0.15775035321712494, "learning_rate": 6.225984794689414e-05, "loss": 0.0123, "step": 45010 }, { "grad_norm": 0.1844812035560608, "learning_rate": 6.22438173286606e-05, "loss": 0.0132, "step": 45020 }, { "grad_norm": 0.15564380586147308, "learning_rate": 6.222778537146159e-05, "loss": 0.0099, "step": 45030 }, { "grad_norm": 0.2063804268836975, "learning_rate": 6.221175207705037e-05, "loss": 0.0138, "step": 45040 }, { "grad_norm": 0.1705736219882965, "learning_rate": 6.219571744718027e-05, "loss": 0.0109, "step": 45050 }, { "grad_norm": 0.13323895633220673, "learning_rate": 6.217968148360487e-05, "loss": 0.0104, "step": 45060 }, { "grad_norm": 0.21472644805908203, "learning_rate": 6.216364418807782e-05, "loss": 0.0123, "step": 45070 }, { "grad_norm": 0.14701849222183228, "learning_rate": 6.214760556235292e-05, "loss": 0.01, "step": 45080 }, { "grad_norm": 0.2203749120235443, "learning_rate": 6.213156560818414e-05, "loss": 0.0109, "step": 45090 }, { "grad_norm": 0.1440374255180359, "learning_rate": 6.211552432732559e-05, "loss": 0.0102, "step": 45100 }, { "grad_norm": 0.11917582899332047, "learning_rate": 6.20994817215315e-05, "loss": 0.011, "step": 45110 }, { "grad_norm": 0.18674726784229279, "learning_rate": 6.208343779255628e-05, "loss": 0.0111, "step": 45120 }, { "grad_norm": 0.15679624676704407, "learning_rate": 6.206739254215449e-05, "loss": 0.0115, "step": 45130 }, { "grad_norm": 0.1877715140581131, "learning_rate": 6.205134597208077e-05, "loss": 0.0115, "step": 45140 }, { "grad_norm": 0.13761231303215027, "learning_rate": 6.203529808409e-05, "loss": 0.0103, "step": 45150 }, { "grad_norm": 0.16327938437461853, "learning_rate": 6.201924887993712e-05, "loss": 0.0103, "step": 45160 }, { "grad_norm": 0.1498405784368515, "learning_rate": 6.200319836137725e-05, "loss": 0.0096, "step": 45170 }, { "grad_norm": 0.14285175502300262, "learning_rate": 6.198714653016565e-05, "loss": 0.0087, "step": 45180 }, { "grad_norm": 0.16745625436306, "learning_rate": 6.197109338805774e-05, "loss": 0.0101, "step": 45190 }, { "grad_norm": 0.14956697821617126, "learning_rate": 6.195503893680903e-05, "loss": 0.0111, "step": 45200 }, { "grad_norm": 0.19755126535892487, "learning_rate": 6.193898317817524e-05, "loss": 0.012, "step": 45210 }, { "grad_norm": 0.15516145527362823, "learning_rate": 6.192292611391221e-05, "loss": 0.0104, "step": 45220 }, { "grad_norm": 0.13637115061283112, "learning_rate": 6.190686774577591e-05, "loss": 0.0093, "step": 45230 }, { "grad_norm": 0.160176083445549, "learning_rate": 6.189080807552245e-05, "loss": 0.0101, "step": 45240 }, { "grad_norm": 0.23452599346637726, "learning_rate": 6.187474710490809e-05, "loss": 0.011, "step": 45250 }, { "grad_norm": 0.12283074110746384, "learning_rate": 6.185868483568926e-05, "loss": 0.0119, "step": 45260 }, { "grad_norm": 0.1512291431427002, "learning_rate": 6.184262126962245e-05, "loss": 0.0118, "step": 45270 }, { "grad_norm": 0.20275968313217163, "learning_rate": 6.182655640846442e-05, "loss": 0.0108, "step": 45280 }, { "grad_norm": 0.1744193136692047, "learning_rate": 6.181049025397196e-05, "loss": 0.0125, "step": 45290 }, { "grad_norm": 0.19396309554576874, "learning_rate": 6.179442280790202e-05, "loss": 0.012, "step": 45300 }, { "grad_norm": 0.17057140171527863, "learning_rate": 6.177835407201174e-05, "loss": 0.014, "step": 45310 }, { "grad_norm": 0.1433049440383911, "learning_rate": 6.176228404805839e-05, "loss": 0.0107, "step": 45320 }, { "grad_norm": 0.1833249181509018, "learning_rate": 6.174621273779932e-05, "loss": 0.0106, "step": 45330 }, { "grad_norm": 0.19042594730854034, "learning_rate": 6.17301401429921e-05, "loss": 0.0104, "step": 45340 }, { "grad_norm": 0.2204723209142685, "learning_rate": 6.171406626539436e-05, "loss": 0.0121, "step": 45350 }, { "grad_norm": 0.19650954008102417, "learning_rate": 6.169799110676398e-05, "loss": 0.0097, "step": 45360 }, { "grad_norm": 0.28705111145973206, "learning_rate": 6.168191466885885e-05, "loss": 0.0115, "step": 45370 }, { "grad_norm": 0.14049853384494781, "learning_rate": 6.166583695343711e-05, "loss": 0.011, "step": 45380 }, { "grad_norm": 0.17902886867523193, "learning_rate": 6.164975796225698e-05, "loss": 0.0103, "step": 45390 }, { "grad_norm": 0.19477733969688416, "learning_rate": 6.163367769707683e-05, "loss": 0.0087, "step": 45400 }, { "grad_norm": 0.15327729284763336, "learning_rate": 6.161759615965519e-05, "loss": 0.0098, "step": 45410 }, { "grad_norm": 0.147743359208107, "learning_rate": 6.16015133517507e-05, "loss": 0.011, "step": 45420 }, { "grad_norm": 0.16636750102043152, "learning_rate": 6.158542927512214e-05, "loss": 0.0092, "step": 45430 }, { "grad_norm": 0.17021185159683228, "learning_rate": 6.156934393152846e-05, "loss": 0.0103, "step": 45440 }, { "grad_norm": 0.18367698788642883, "learning_rate": 6.15532573227287e-05, "loss": 0.0105, "step": 45450 }, { "grad_norm": 0.18678022921085358, "learning_rate": 6.153716945048212e-05, "loss": 0.0125, "step": 45460 }, { "grad_norm": 0.21829593181610107, "learning_rate": 6.152108031654802e-05, "loss": 0.0107, "step": 45470 }, { "grad_norm": 0.12809664011001587, "learning_rate": 6.15049899226859e-05, "loss": 0.0113, "step": 45480 }, { "grad_norm": 0.287412166595459, "learning_rate": 6.148889827065537e-05, "loss": 0.0108, "step": 45490 }, { "grad_norm": 0.21269731223583221, "learning_rate": 6.147280536221622e-05, "loss": 0.0103, "step": 45500 }, { "grad_norm": 0.12166441231966019, "learning_rate": 6.145671119912832e-05, "loss": 0.0106, "step": 45510 }, { "grad_norm": 0.14951547980308533, "learning_rate": 6.144061578315169e-05, "loss": 0.0125, "step": 45520 }, { "grad_norm": 0.14584119617938995, "learning_rate": 6.142451911604654e-05, "loss": 0.0098, "step": 45530 }, { "grad_norm": 0.15854044258594513, "learning_rate": 6.140842119957315e-05, "loss": 0.0101, "step": 45540 }, { "grad_norm": 0.1589893400669098, "learning_rate": 6.139232203549197e-05, "loss": 0.0126, "step": 45550 }, { "grad_norm": 0.19791968166828156, "learning_rate": 6.137622162556357e-05, "loss": 0.0105, "step": 45560 }, { "grad_norm": 0.17906206846237183, "learning_rate": 6.13601199715487e-05, "loss": 0.0114, "step": 45570 }, { "grad_norm": 0.1976114809513092, "learning_rate": 6.134401707520816e-05, "loss": 0.0098, "step": 45580 }, { "grad_norm": 0.22279298305511475, "learning_rate": 6.132791293830298e-05, "loss": 0.0105, "step": 45590 }, { "grad_norm": 0.15970243513584137, "learning_rate": 6.131180756259428e-05, "loss": 0.0126, "step": 45600 }, { "grad_norm": 0.1420339047908783, "learning_rate": 6.129570094984331e-05, "loss": 0.0098, "step": 45610 }, { "grad_norm": 0.2426605075597763, "learning_rate": 6.127959310181145e-05, "loss": 0.0098, "step": 45620 }, { "grad_norm": 0.18957459926605225, "learning_rate": 6.126348402026026e-05, "loss": 0.0115, "step": 45630 }, { "grad_norm": 0.16152848303318024, "learning_rate": 6.12473737069514e-05, "loss": 0.0122, "step": 45640 }, { "grad_norm": 0.1416550874710083, "learning_rate": 6.123126216364665e-05, "loss": 0.0095, "step": 45650 }, { "grad_norm": 0.14660419523715973, "learning_rate": 6.121514939210797e-05, "loss": 0.0111, "step": 45660 }, { "grad_norm": 0.3018069863319397, "learning_rate": 6.119903539409741e-05, "loss": 0.0094, "step": 45670 }, { "grad_norm": 0.1634959876537323, "learning_rate": 6.118292017137716e-05, "loss": 0.0108, "step": 45680 }, { "grad_norm": 0.11932899057865143, "learning_rate": 6.116680372570959e-05, "loss": 0.0089, "step": 45690 }, { "grad_norm": 0.1276397854089737, "learning_rate": 6.115068605885713e-05, "loss": 0.0091, "step": 45700 }, { "grad_norm": 0.17815859615802765, "learning_rate": 6.113456717258243e-05, "loss": 0.0115, "step": 45710 }, { "grad_norm": 0.16558681428432465, "learning_rate": 6.11184470686482e-05, "loss": 0.01, "step": 45720 }, { "grad_norm": 0.14533300697803497, "learning_rate": 6.110232574881731e-05, "loss": 0.0096, "step": 45730 }, { "grad_norm": 0.2496521919965744, "learning_rate": 6.108620321485277e-05, "loss": 0.0104, "step": 45740 }, { "grad_norm": 0.1690020114183426, "learning_rate": 6.107007946851773e-05, "loss": 0.0099, "step": 45750 }, { "grad_norm": 0.3139280378818512, "learning_rate": 6.105395451157542e-05, "loss": 0.0128, "step": 45760 }, { "grad_norm": 0.17726774513721466, "learning_rate": 6.103782834578928e-05, "loss": 0.01, "step": 45770 }, { "grad_norm": 0.16394464671611786, "learning_rate": 6.102170097292281e-05, "loss": 0.0103, "step": 45780 }, { "grad_norm": 0.15045981109142303, "learning_rate": 6.1005572394739716e-05, "loss": 0.0092, "step": 45790 }, { "grad_norm": 0.16829697787761688, "learning_rate": 6.0989442613003765e-05, "loss": 0.0103, "step": 45800 }, { "grad_norm": 0.12974585592746735, "learning_rate": 6.097331162947888e-05, "loss": 0.0108, "step": 45810 }, { "grad_norm": 0.24268391728401184, "learning_rate": 6.095717944592914e-05, "loss": 0.0104, "step": 45820 }, { "grad_norm": 0.29154878854751587, "learning_rate": 6.094104606411873e-05, "loss": 0.0091, "step": 45830 }, { "grad_norm": 0.22792048752307892, "learning_rate": 6.0924911485811966e-05, "loss": 0.0101, "step": 45840 }, { "grad_norm": 0.21400441229343414, "learning_rate": 6.090877571277331e-05, "loss": 0.0103, "step": 45850 }, { "grad_norm": 0.24942907691001892, "learning_rate": 6.089263874676736e-05, "loss": 0.0094, "step": 45860 }, { "grad_norm": 0.17052128911018372, "learning_rate": 6.0876500589558796e-05, "loss": 0.0099, "step": 45870 }, { "grad_norm": 0.170857235789299, "learning_rate": 6.086036124291248e-05, "loss": 0.0093, "step": 45880 }, { "grad_norm": 0.17441724240779877, "learning_rate": 6.084422070859339e-05, "loss": 0.0095, "step": 45890 }, { "grad_norm": 0.1677822768688202, "learning_rate": 6.082807898836663e-05, "loss": 0.0117, "step": 45900 }, { "grad_norm": 0.2122288942337036, "learning_rate": 6.081193608399742e-05, "loss": 0.0114, "step": 45910 }, { "grad_norm": 0.22271546721458435, "learning_rate": 6.0795791997251164e-05, "loss": 0.0117, "step": 45920 }, { "grad_norm": 0.19239211082458496, "learning_rate": 6.0779646729893294e-05, "loss": 0.0106, "step": 45930 }, { "grad_norm": 0.17276135087013245, "learning_rate": 6.0763500283689476e-05, "loss": 0.0111, "step": 45940 }, { "grad_norm": 0.17255254089832306, "learning_rate": 6.0747352660405455e-05, "loss": 0.01, "step": 45950 }, { "grad_norm": 0.17557361721992493, "learning_rate": 6.073120386180709e-05, "loss": 0.0099, "step": 45960 }, { "grad_norm": 0.14974547922611237, "learning_rate": 6.0715053889660425e-05, "loss": 0.0115, "step": 45970 }, { "grad_norm": 0.22427059710025787, "learning_rate": 6.069890274573157e-05, "loss": 0.0138, "step": 45980 }, { "grad_norm": 0.25681018829345703, "learning_rate": 6.068275043178679e-05, "loss": 0.0104, "step": 45990 }, { "grad_norm": 0.23092445731163025, "learning_rate": 6.066659694959248e-05, "loss": 0.0122, "step": 46000 }, { "grad_norm": 0.26660943031311035, "learning_rate": 6.0650442300915176e-05, "loss": 0.0106, "step": 46010 }, { "grad_norm": 0.16794227063655853, "learning_rate": 6.063428648752152e-05, "loss": 0.0121, "step": 46020 }, { "grad_norm": 0.2567921280860901, "learning_rate": 6.0618129511178266e-05, "loss": 0.0099, "step": 46030 }, { "grad_norm": 0.15218539535999298, "learning_rate": 6.060197137365234e-05, "loss": 0.013, "step": 46040 }, { "grad_norm": 0.15042740106582642, "learning_rate": 6.058581207671077e-05, "loss": 0.0121, "step": 46050 }, { "grad_norm": 0.17811378836631775, "learning_rate": 6.056965162212072e-05, "loss": 0.0094, "step": 46060 }, { "grad_norm": 0.1281435638666153, "learning_rate": 6.055349001164943e-05, "loss": 0.0091, "step": 46070 }, { "grad_norm": 0.12432152777910233, "learning_rate": 6.0537327247064347e-05, "loss": 0.0092, "step": 46080 }, { "grad_norm": 0.1469188779592514, "learning_rate": 6.0521163330133014e-05, "loss": 0.0084, "step": 46090 }, { "grad_norm": 0.2422737181186676, "learning_rate": 6.050499826262306e-05, "loss": 0.0087, "step": 46100 }, { "grad_norm": 0.15533135831356049, "learning_rate": 6.0488832046302294e-05, "loss": 0.0115, "step": 46110 }, { "grad_norm": 0.12277168780565262, "learning_rate": 6.0472664682938626e-05, "loss": 0.0092, "step": 46120 }, { "grad_norm": 0.23167462646961212, "learning_rate": 6.045649617430009e-05, "loss": 0.0085, "step": 46130 }, { "grad_norm": 0.2842685878276825, "learning_rate": 6.0440326522154866e-05, "loss": 0.0099, "step": 46140 }, { "grad_norm": 0.17459505796432495, "learning_rate": 6.0424155728271224e-05, "loss": 0.0107, "step": 46150 }, { "grad_norm": 0.169732004404068, "learning_rate": 6.040798379441758e-05, "loss": 0.0103, "step": 46160 }, { "grad_norm": 0.16573618352413177, "learning_rate": 6.0391810722362485e-05, "loss": 0.0122, "step": 46170 }, { "grad_norm": 0.1994306892156601, "learning_rate": 6.037563651387458e-05, "loss": 0.0109, "step": 46180 }, { "grad_norm": 0.11995881050825119, "learning_rate": 6.0359461170722666e-05, "loss": 0.0078, "step": 46190 }, { "grad_norm": 0.18102972209453583, "learning_rate": 6.034328469467566e-05, "loss": 0.01, "step": 46200 }, { "grad_norm": 0.13600099086761475, "learning_rate": 6.0327107087502596e-05, "loss": 0.0101, "step": 46210 }, { "grad_norm": 0.228434219956398, "learning_rate": 6.031092835097262e-05, "loss": 0.0106, "step": 46220 }, { "grad_norm": 0.18012574315071106, "learning_rate": 6.0294748486855024e-05, "loss": 0.0116, "step": 46230 }, { "grad_norm": 0.160194993019104, "learning_rate": 6.0278567496919216e-05, "loss": 0.0122, "step": 46240 }, { "grad_norm": 0.1802082657814026, "learning_rate": 6.026238538293472e-05, "loss": 0.009, "step": 46250 }, { "grad_norm": 0.2955716848373413, "learning_rate": 6.024620214667118e-05, "loss": 0.0092, "step": 46260 }, { "grad_norm": 0.23438967764377594, "learning_rate": 6.0230017789898384e-05, "loss": 0.0101, "step": 46270 }, { "grad_norm": 0.18373991549015045, "learning_rate": 6.021383231438622e-05, "loss": 0.0094, "step": 46280 }, { "grad_norm": 0.2683351933956146, "learning_rate": 6.0197645721904704e-05, "loss": 0.0096, "step": 46290 }, { "grad_norm": 0.24969178438186646, "learning_rate": 6.0181458014224e-05, "loss": 0.0129, "step": 46300 }, { "grad_norm": 0.17123763263225555, "learning_rate": 6.016526919311435e-05, "loss": 0.0094, "step": 46310 }, { "grad_norm": 0.408145934343338, "learning_rate": 6.014907926034613e-05, "loss": 0.0111, "step": 46320 }, { "grad_norm": 0.1621820330619812, "learning_rate": 6.0132888217689875e-05, "loss": 0.0097, "step": 46330 }, { "grad_norm": 0.2242395281791687, "learning_rate": 6.01166960669162e-05, "loss": 0.0109, "step": 46340 }, { "grad_norm": 0.2792883515357971, "learning_rate": 6.0100502809795845e-05, "loss": 0.0095, "step": 46350 }, { "grad_norm": 0.11683236807584763, "learning_rate": 6.00843084480997e-05, "loss": 0.0105, "step": 46360 }, { "grad_norm": 0.19917719066143036, "learning_rate": 6.0068112983598736e-05, "loss": 0.0105, "step": 46370 }, { "grad_norm": 0.24803757667541504, "learning_rate": 6.005191641806407e-05, "loss": 0.0107, "step": 46380 }, { "grad_norm": 0.15344037115573883, "learning_rate": 6.003571875326694e-05, "loss": 0.0111, "step": 46390 }, { "grad_norm": 0.17951855063438416, "learning_rate": 6.00195199909787e-05, "loss": 0.0113, "step": 46400 }, { "grad_norm": 0.17997217178344727, "learning_rate": 6.000332013297082e-05, "loss": 0.0107, "step": 46410 }, { "grad_norm": 0.171053946018219, "learning_rate": 5.998711918101487e-05, "loss": 0.0115, "step": 46420 }, { "grad_norm": 0.2502965033054352, "learning_rate": 5.997091713688261e-05, "loss": 0.011, "step": 46430 }, { "grad_norm": 0.14286328852176666, "learning_rate": 5.9954714002345836e-05, "loss": 0.0099, "step": 46440 }, { "grad_norm": 0.18230581283569336, "learning_rate": 5.993850977917649e-05, "loss": 0.0109, "step": 46450 }, { "grad_norm": 0.24085365235805511, "learning_rate": 5.992230446914667e-05, "loss": 0.0096, "step": 46460 }, { "grad_norm": 0.261655330657959, "learning_rate": 5.990609807402855e-05, "loss": 0.0139, "step": 46470 }, { "grad_norm": 0.18111374974250793, "learning_rate": 5.988989059559443e-05, "loss": 0.0136, "step": 46480 }, { "grad_norm": 0.15353401005268097, "learning_rate": 5.987368203561675e-05, "loss": 0.0096, "step": 46490 }, { "grad_norm": 0.14432477951049805, "learning_rate": 5.9857472395868055e-05, "loss": 0.0103, "step": 46500 }, { "grad_norm": 0.15711772441864014, "learning_rate": 5.9841261678120983e-05, "loss": 0.0102, "step": 46510 }, { "grad_norm": 0.12263917177915573, "learning_rate": 5.982504988414834e-05, "loss": 0.0092, "step": 46520 }, { "grad_norm": 0.2657581865787506, "learning_rate": 5.9808837015723015e-05, "loss": 0.0104, "step": 46530 }, { "grad_norm": 0.174923837184906, "learning_rate": 5.9792623074618016e-05, "loss": 0.0129, "step": 46540 }, { "grad_norm": 0.18242326378822327, "learning_rate": 5.977640806260648e-05, "loss": 0.011, "step": 46550 }, { "grad_norm": 0.14479006826877594, "learning_rate": 5.976019198146166e-05, "loss": 0.0111, "step": 46560 }, { "grad_norm": 0.12477831542491913, "learning_rate": 5.974397483295692e-05, "loss": 0.0093, "step": 46570 }, { "grad_norm": 0.10658815503120422, "learning_rate": 5.972775661886575e-05, "loss": 0.0101, "step": 46580 }, { "grad_norm": 0.1277141571044922, "learning_rate": 5.971153734096173e-05, "loss": 0.0093, "step": 46590 }, { "grad_norm": 0.17282569408416748, "learning_rate": 5.969531700101859e-05, "loss": 0.0115, "step": 46600 }, { "grad_norm": 0.14885807037353516, "learning_rate": 5.9679095600810155e-05, "loss": 0.0094, "step": 46610 }, { "grad_norm": 0.145564004778862, "learning_rate": 5.9662873142110384e-05, "loss": 0.0119, "step": 46620 }, { "grad_norm": 0.23009787499904633, "learning_rate": 5.964664962669333e-05, "loss": 0.0103, "step": 46630 }, { "grad_norm": 0.24165424704551697, "learning_rate": 5.9630425056333186e-05, "loss": 0.011, "step": 46640 }, { "grad_norm": 0.17118415236473083, "learning_rate": 5.961419943280422e-05, "loss": 0.0116, "step": 46650 }, { "grad_norm": 0.13949312269687653, "learning_rate": 5.959797275788087e-05, "loss": 0.012, "step": 46660 }, { "grad_norm": 0.386322021484375, "learning_rate": 5.958174503333765e-05, "loss": 0.0101, "step": 46670 }, { "grad_norm": 0.29139479994773865, "learning_rate": 5.9565516260949195e-05, "loss": 0.0128, "step": 46680 }, { "grad_norm": 0.123387411236763, "learning_rate": 5.9549286442490273e-05, "loss": 0.009, "step": 46690 }, { "grad_norm": 0.17210988700389862, "learning_rate": 5.953305557973572e-05, "loss": 0.0096, "step": 46700 }, { "grad_norm": 0.19882263243198395, "learning_rate": 5.9516823674460564e-05, "loss": 0.0108, "step": 46710 }, { "grad_norm": 0.20851540565490723, "learning_rate": 5.9500590728439875e-05, "loss": 0.0113, "step": 46720 }, { "grad_norm": 0.23149719834327698, "learning_rate": 5.948435674344886e-05, "loss": 0.0128, "step": 46730 }, { "grad_norm": 0.1963629275560379, "learning_rate": 5.946812172126285e-05, "loss": 0.0093, "step": 46740 }, { "grad_norm": 0.16084344685077667, "learning_rate": 5.945188566365729e-05, "loss": 0.0145, "step": 46750 }, { "grad_norm": 0.2751719653606415, "learning_rate": 5.943564857240773e-05, "loss": 0.0121, "step": 46760 }, { "grad_norm": 0.1796863079071045, "learning_rate": 5.941941044928981e-05, "loss": 0.0102, "step": 46770 }, { "grad_norm": 0.18852585554122925, "learning_rate": 5.940317129607935e-05, "loss": 0.0141, "step": 46780 }, { "grad_norm": 0.17168328166007996, "learning_rate": 5.9386931114552204e-05, "loss": 0.0094, "step": 46790 }, { "grad_norm": 0.2048845738172531, "learning_rate": 5.937068990648438e-05, "loss": 0.0101, "step": 46800 }, { "grad_norm": 0.16462209820747375, "learning_rate": 5.935444767365199e-05, "loss": 0.0118, "step": 46810 }, { "grad_norm": 0.13793480396270752, "learning_rate": 5.933820441783129e-05, "loss": 0.0118, "step": 46820 }, { "grad_norm": 0.24796569347381592, "learning_rate": 5.932196014079857e-05, "loss": 0.0092, "step": 46830 }, { "grad_norm": 0.12102602422237396, "learning_rate": 5.930571484433032e-05, "loss": 0.0092, "step": 46840 }, { "grad_norm": 0.15338505804538727, "learning_rate": 5.928946853020309e-05, "loss": 0.0105, "step": 46850 }, { "grad_norm": 0.1953265517950058, "learning_rate": 5.927322120019352e-05, "loss": 0.0101, "step": 46860 }, { "grad_norm": 0.14635321497917175, "learning_rate": 5.925697285607844e-05, "loss": 0.0099, "step": 46870 }, { "grad_norm": 0.2636595070362091, "learning_rate": 5.924072349963472e-05, "loss": 0.0114, "step": 46880 }, { "grad_norm": 0.14527122676372528, "learning_rate": 5.922447313263938e-05, "loss": 0.0086, "step": 46890 }, { "grad_norm": 0.17745362222194672, "learning_rate": 5.920822175686952e-05, "loss": 0.0125, "step": 46900 }, { "grad_norm": 0.6902895569801331, "learning_rate": 5.919196937410239e-05, "loss": 0.0115, "step": 46910 }, { "grad_norm": 0.15976232290267944, "learning_rate": 5.917571598611529e-05, "loss": 0.0126, "step": 46920 }, { "grad_norm": 0.1637459099292755, "learning_rate": 5.9159461594685686e-05, "loss": 0.0108, "step": 46930 }, { "grad_norm": 0.11792755126953125, "learning_rate": 5.914320620159114e-05, "loss": 0.0106, "step": 46940 }, { "grad_norm": 0.15104937553405762, "learning_rate": 5.912694980860931e-05, "loss": 0.0109, "step": 46950 }, { "grad_norm": 0.17348496615886688, "learning_rate": 5.9110692417517964e-05, "loss": 0.0095, "step": 46960 }, { "grad_norm": 0.154026597738266, "learning_rate": 5.9094434030095e-05, "loss": 0.0106, "step": 46970 }, { "grad_norm": 0.1165858805179596, "learning_rate": 5.907817464811841e-05, "loss": 0.009, "step": 46980 }, { "grad_norm": 0.16049803793430328, "learning_rate": 5.906191427336627e-05, "loss": 0.0107, "step": 46990 }, { "grad_norm": 0.12494637072086334, "learning_rate": 5.9045652907616835e-05, "loss": 0.0108, "step": 47000 }, { "grad_norm": 0.17321306467056274, "learning_rate": 5.902939055264838e-05, "loss": 0.0102, "step": 47010 }, { "grad_norm": 0.16879868507385254, "learning_rate": 5.901312721023935e-05, "loss": 0.0101, "step": 47020 }, { "grad_norm": 0.15949858725070953, "learning_rate": 5.8996862882168294e-05, "loss": 0.0103, "step": 47030 }, { "grad_norm": 0.21364904940128326, "learning_rate": 5.8980597570213826e-05, "loss": 0.0096, "step": 47040 }, { "grad_norm": 0.1640610545873642, "learning_rate": 5.896433127615471e-05, "loss": 0.0091, "step": 47050 }, { "grad_norm": 0.2052421271800995, "learning_rate": 5.894806400176981e-05, "loss": 0.0112, "step": 47060 }, { "grad_norm": 0.16236640512943268, "learning_rate": 5.893179574883808e-05, "loss": 0.0098, "step": 47070 }, { "grad_norm": 0.16712281107902527, "learning_rate": 5.8915526519138585e-05, "loss": 0.0109, "step": 47080 }, { "grad_norm": 0.2788027226924896, "learning_rate": 5.889925631445053e-05, "loss": 0.0097, "step": 47090 }, { "grad_norm": 0.14331525564193726, "learning_rate": 5.888298513655318e-05, "loss": 0.0101, "step": 47100 }, { "grad_norm": 0.23338215053081512, "learning_rate": 5.886671298722592e-05, "loss": 0.01, "step": 47110 }, { "grad_norm": 0.1838691532611847, "learning_rate": 5.885043986824828e-05, "loss": 0.0121, "step": 47120 }, { "grad_norm": 0.13351841270923615, "learning_rate": 5.883416578139982e-05, "loss": 0.0104, "step": 47130 }, { "grad_norm": 0.32138708233833313, "learning_rate": 5.881789072846029e-05, "loss": 0.012, "step": 47140 }, { "grad_norm": 0.16289180517196655, "learning_rate": 5.8801614711209474e-05, "loss": 0.0111, "step": 47150 }, { "grad_norm": 0.1332765817642212, "learning_rate": 5.878533773142732e-05, "loss": 0.0097, "step": 47160 }, { "grad_norm": 0.1918114721775055, "learning_rate": 5.8769059790893834e-05, "loss": 0.0128, "step": 47170 }, { "grad_norm": 0.22158633172512054, "learning_rate": 5.8752780891389145e-05, "loss": 0.0122, "step": 47180 }, { "grad_norm": 0.12884780764579773, "learning_rate": 5.873650103469351e-05, "loss": 0.01, "step": 47190 }, { "grad_norm": 0.19262447953224182, "learning_rate": 5.872022022258725e-05, "loss": 0.0104, "step": 47200 }, { "grad_norm": 0.156339630484581, "learning_rate": 5.8703938456850814e-05, "loss": 0.0111, "step": 47210 }, { "grad_norm": 0.2171017974615097, "learning_rate": 5.868765573926477e-05, "loss": 0.0093, "step": 47220 }, { "grad_norm": 0.2093435376882553, "learning_rate": 5.867137207160974e-05, "loss": 0.011, "step": 47230 }, { "grad_norm": 0.1660645753145218, "learning_rate": 5.86550874556665e-05, "loss": 0.0095, "step": 47240 }, { "grad_norm": 0.2578086256980896, "learning_rate": 5.863880189321592e-05, "loss": 0.0094, "step": 47250 }, { "grad_norm": 0.17456687986850739, "learning_rate": 5.862251538603894e-05, "loss": 0.0095, "step": 47260 }, { "grad_norm": 0.21514087915420532, "learning_rate": 5.8606227935916656e-05, "loss": 0.0124, "step": 47270 }, { "grad_norm": 0.24747683107852936, "learning_rate": 5.858993954463021e-05, "loss": 0.0104, "step": 47280 }, { "grad_norm": 0.40184611082077026, "learning_rate": 5.8573650213960896e-05, "loss": 0.0112, "step": 47290 }, { "grad_norm": 0.14563868939876556, "learning_rate": 5.8557359945690084e-05, "loss": 0.0111, "step": 47300 }, { "grad_norm": 0.3288412392139435, "learning_rate": 5.8541068741599246e-05, "loss": 0.0103, "step": 47310 }, { "grad_norm": 0.18256616592407227, "learning_rate": 5.8524776603469985e-05, "loss": 0.0127, "step": 47320 }, { "grad_norm": 0.21108730137348175, "learning_rate": 5.850848353308397e-05, "loss": 0.0124, "step": 47330 }, { "grad_norm": 0.24683772027492523, "learning_rate": 5.849218953222297e-05, "loss": 0.0109, "step": 47340 }, { "grad_norm": 0.2120806723833084, "learning_rate": 5.847589460266891e-05, "loss": 0.0112, "step": 47350 }, { "grad_norm": 0.18737033009529114, "learning_rate": 5.845959874620375e-05, "loss": 0.0095, "step": 47360 }, { "grad_norm": 0.17580915987491608, "learning_rate": 5.84433019646096e-05, "loss": 0.0129, "step": 47370 }, { "grad_norm": 0.27688702940940857, "learning_rate": 5.842700425966863e-05, "loss": 0.013, "step": 47380 }, { "grad_norm": 0.20370827615261078, "learning_rate": 5.841070563316315e-05, "loss": 0.0135, "step": 47390 }, { "grad_norm": 0.2604011595249176, "learning_rate": 5.839440608687554e-05, "loss": 0.0123, "step": 47400 }, { "grad_norm": 0.22974681854248047, "learning_rate": 5.837810562258831e-05, "loss": 0.0111, "step": 47410 }, { "grad_norm": 0.1255769282579422, "learning_rate": 5.836180424208405e-05, "loss": 0.0094, "step": 47420 }, { "grad_norm": 0.15858212113380432, "learning_rate": 5.834550194714543e-05, "loss": 0.0125, "step": 47430 }, { "grad_norm": 0.11676037311553955, "learning_rate": 5.832919873955528e-05, "loss": 0.0123, "step": 47440 }, { "grad_norm": 0.19980257749557495, "learning_rate": 5.831289462109648e-05, "loss": 0.0104, "step": 47450 }, { "grad_norm": 0.15747228264808655, "learning_rate": 5.829658959355202e-05, "loss": 0.0096, "step": 47460 }, { "grad_norm": 0.15476983785629272, "learning_rate": 5.8280283658704995e-05, "loss": 0.0123, "step": 47470 }, { "grad_norm": 0.23896285891532898, "learning_rate": 5.826397681833859e-05, "loss": 0.0106, "step": 47480 }, { "grad_norm": 0.12713110446929932, "learning_rate": 5.824766907423612e-05, "loss": 0.0126, "step": 47490 }, { "grad_norm": 0.1250912994146347, "learning_rate": 5.8231360428180945e-05, "loss": 0.0101, "step": 47500 }, { "grad_norm": 0.17019791901111603, "learning_rate": 5.821505088195658e-05, "loss": 0.0114, "step": 47510 }, { "grad_norm": 0.1566593497991562, "learning_rate": 5.819874043734661e-05, "loss": 0.0093, "step": 47520 }, { "grad_norm": 0.14284038543701172, "learning_rate": 5.8182429096134695e-05, "loss": 0.0106, "step": 47530 }, { "grad_norm": 0.13679082691669464, "learning_rate": 5.816611686010465e-05, "loss": 0.0101, "step": 47540 }, { "grad_norm": 0.2211105078458786, "learning_rate": 5.814980373104033e-05, "loss": 0.0139, "step": 47550 }, { "grad_norm": 0.18539324402809143, "learning_rate": 5.813348971072572e-05, "loss": 0.011, "step": 47560 }, { "grad_norm": 0.18086181581020355, "learning_rate": 5.811717480094492e-05, "loss": 0.01, "step": 47570 }, { "grad_norm": 0.18004749715328217, "learning_rate": 5.810085900348209e-05, "loss": 0.0135, "step": 47580 }, { "grad_norm": 0.14355091750621796, "learning_rate": 5.8084542320121483e-05, "loss": 0.0126, "step": 47590 }, { "grad_norm": 0.34758564829826355, "learning_rate": 5.8068224752647497e-05, "loss": 0.0113, "step": 47600 }, { "grad_norm": 0.1730327159166336, "learning_rate": 5.805190630284457e-05, "loss": 0.0111, "step": 47610 }, { "grad_norm": 0.20968671143054962, "learning_rate": 5.803558697249729e-05, "loss": 0.0108, "step": 47620 }, { "grad_norm": 0.16832028329372406, "learning_rate": 5.8019266763390276e-05, "loss": 0.0119, "step": 47630 }, { "grad_norm": 0.18825994431972504, "learning_rate": 5.800294567730833e-05, "loss": 0.0107, "step": 47640 }, { "grad_norm": 0.12713059782981873, "learning_rate": 5.798662371603628e-05, "loss": 0.0102, "step": 47650 }, { "grad_norm": 0.1641111969947815, "learning_rate": 5.797030088135904e-05, "loss": 0.0094, "step": 47660 }, { "grad_norm": 0.17621386051177979, "learning_rate": 5.7953977175061705e-05, "loss": 0.0098, "step": 47670 }, { "grad_norm": 0.18579021096229553, "learning_rate": 5.793765259892938e-05, "loss": 0.0105, "step": 47680 }, { "grad_norm": 0.13192272186279297, "learning_rate": 5.792132715474729e-05, "loss": 0.0111, "step": 47690 }, { "grad_norm": 0.18150298297405243, "learning_rate": 5.790500084430078e-05, "loss": 0.0103, "step": 47700 }, { "grad_norm": 0.20407922565937042, "learning_rate": 5.7888673669375274e-05, "loss": 0.0092, "step": 47710 }, { "grad_norm": 0.16803313791751862, "learning_rate": 5.787234563175625e-05, "loss": 0.0109, "step": 47720 }, { "grad_norm": 0.15720254182815552, "learning_rate": 5.7856016733229355e-05, "loss": 0.011, "step": 47730 }, { "grad_norm": 0.20277512073516846, "learning_rate": 5.7839686975580297e-05, "loss": 0.011, "step": 47740 }, { "grad_norm": 0.16633141040802002, "learning_rate": 5.782335636059484e-05, "loss": 0.0095, "step": 47750 }, { "grad_norm": 0.125382661819458, "learning_rate": 5.780702489005889e-05, "loss": 0.0095, "step": 47760 }, { "grad_norm": 0.262433260679245, "learning_rate": 5.779069256575845e-05, "loss": 0.0112, "step": 47770 }, { "grad_norm": 0.14686283469200134, "learning_rate": 5.7774359389479574e-05, "loss": 0.0108, "step": 47780 }, { "grad_norm": 0.4192540645599365, "learning_rate": 5.775802536300845e-05, "loss": 0.01, "step": 47790 }, { "grad_norm": 0.21508929133415222, "learning_rate": 5.774169048813134e-05, "loss": 0.01, "step": 47800 }, { "grad_norm": 0.18124395608901978, "learning_rate": 5.77253547666346e-05, "loss": 0.0105, "step": 47810 }, { "grad_norm": 0.16110259294509888, "learning_rate": 5.770901820030465e-05, "loss": 0.0109, "step": 47820 }, { "grad_norm": 0.18405044078826904, "learning_rate": 5.769268079092809e-05, "loss": 0.0101, "step": 47830 }, { "grad_norm": 0.11974041908979416, "learning_rate": 5.767634254029151e-05, "loss": 0.0094, "step": 47840 }, { "grad_norm": 0.2849363088607788, "learning_rate": 5.7660003450181655e-05, "loss": 0.0088, "step": 47850 }, { "grad_norm": 0.2180161029100418, "learning_rate": 5.764366352238534e-05, "loss": 0.0097, "step": 47860 }, { "grad_norm": 0.1296442300081253, "learning_rate": 5.7627322758689474e-05, "loss": 0.0105, "step": 47870 }, { "grad_norm": 0.1853271722793579, "learning_rate": 5.761098116088105e-05, "loss": 0.0089, "step": 47880 }, { "grad_norm": 0.15548589825630188, "learning_rate": 5.759463873074717e-05, "loss": 0.0115, "step": 47890 }, { "grad_norm": 0.17119678854942322, "learning_rate": 5.757829547007504e-05, "loss": 0.0122, "step": 47900 }, { "grad_norm": 0.20656102895736694, "learning_rate": 5.756195138065189e-05, "loss": 0.0104, "step": 47910 }, { "grad_norm": 0.13693520426750183, "learning_rate": 5.754560646426511e-05, "loss": 0.0107, "step": 47920 }, { "grad_norm": 0.11827566474676132, "learning_rate": 5.752926072270216e-05, "loss": 0.0106, "step": 47930 }, { "grad_norm": 0.15694354474544525, "learning_rate": 5.7512914157750563e-05, "loss": 0.0119, "step": 47940 }, { "grad_norm": 0.165932759642601, "learning_rate": 5.749656677119798e-05, "loss": 0.0102, "step": 47950 }, { "grad_norm": 0.17530113458633423, "learning_rate": 5.7480218564832125e-05, "loss": 0.0103, "step": 47960 }, { "grad_norm": 0.1649465709924698, "learning_rate": 5.746386954044082e-05, "loss": 0.0105, "step": 47970 }, { "grad_norm": 0.17298021912574768, "learning_rate": 5.744751969981195e-05, "loss": 0.0095, "step": 47980 }, { "grad_norm": 0.18623539805412292, "learning_rate": 5.7431169044733526e-05, "loss": 0.0091, "step": 47990 }, { "grad_norm": 0.19026534259319305, "learning_rate": 5.741481757699364e-05, "loss": 0.0099, "step": 48000 }, { "grad_norm": 0.22511252760887146, "learning_rate": 5.7398465298380434e-05, "loss": 0.0091, "step": 48010 }, { "grad_norm": 0.24065792560577393, "learning_rate": 5.7382112210682193e-05, "loss": 0.0107, "step": 48020 }, { "grad_norm": 0.12108837068080902, "learning_rate": 5.7365758315687266e-05, "loss": 0.0086, "step": 48030 }, { "grad_norm": 0.1152382642030716, "learning_rate": 5.734940361518407e-05, "loss": 0.0101, "step": 48040 }, { "grad_norm": 0.1847250908613205, "learning_rate": 5.733304811096116e-05, "loss": 0.0086, "step": 48050 }, { "grad_norm": 0.14635668694972992, "learning_rate": 5.731669180480713e-05, "loss": 0.0116, "step": 48060 }, { "grad_norm": 0.161012202501297, "learning_rate": 5.730033469851067e-05, "loss": 0.0097, "step": 48070 }, { "grad_norm": 0.1898537129163742, "learning_rate": 5.7283976793860607e-05, "loss": 0.009, "step": 48080 }, { "grad_norm": 0.1375192105770111, "learning_rate": 5.726761809264577e-05, "loss": 0.0081, "step": 48090 }, { "grad_norm": 0.14261247217655182, "learning_rate": 5.7251258596655155e-05, "loss": 0.0109, "step": 48100 }, { "grad_norm": 0.28043094277381897, "learning_rate": 5.72348983076778e-05, "loss": 0.0115, "step": 48110 }, { "grad_norm": 0.15793277323246002, "learning_rate": 5.7218537227502854e-05, "loss": 0.0099, "step": 48120 }, { "grad_norm": 0.13659614324569702, "learning_rate": 5.720217535791951e-05, "loss": 0.0083, "step": 48130 }, { "grad_norm": 0.14706619083881378, "learning_rate": 5.718581270071711e-05, "loss": 0.0114, "step": 48140 }, { "grad_norm": 0.23224182426929474, "learning_rate": 5.716944925768505e-05, "loss": 0.0116, "step": 48150 }, { "grad_norm": 0.1455198973417282, "learning_rate": 5.7153085030612786e-05, "loss": 0.0094, "step": 48160 }, { "grad_norm": 0.21218451857566833, "learning_rate": 5.713672002128989e-05, "loss": 0.0101, "step": 48170 }, { "grad_norm": 0.1364808827638626, "learning_rate": 5.712035423150602e-05, "loss": 0.012, "step": 48180 }, { "grad_norm": 0.11078570038080215, "learning_rate": 5.710398766305094e-05, "loss": 0.0088, "step": 48190 }, { "grad_norm": 0.13914205133914948, "learning_rate": 5.708762031771442e-05, "loss": 0.0078, "step": 48200 }, { "grad_norm": 0.15836431086063385, "learning_rate": 5.707125219728642e-05, "loss": 0.0101, "step": 48210 }, { "grad_norm": 0.1276102215051651, "learning_rate": 5.7054883303556905e-05, "loss": 0.0097, "step": 48220 }, { "grad_norm": 0.1956780105829239, "learning_rate": 5.703851363831595e-05, "loss": 0.0101, "step": 48230 }, { "grad_norm": 0.17267490923404694, "learning_rate": 5.702214320335374e-05, "loss": 0.0101, "step": 48240 }, { "grad_norm": 0.29912856221199036, "learning_rate": 5.7005772000460514e-05, "loss": 0.0105, "step": 48250 }, { "grad_norm": 0.14743906259536743, "learning_rate": 5.698940003142656e-05, "loss": 0.0092, "step": 48260 }, { "grad_norm": 0.1532669961452484, "learning_rate": 5.697302729804236e-05, "loss": 0.0101, "step": 48270 }, { "grad_norm": 0.36389607191085815, "learning_rate": 5.695665380209837e-05, "loss": 0.0092, "step": 48280 }, { "grad_norm": 0.1486683040857315, "learning_rate": 5.6940279545385165e-05, "loss": 0.0118, "step": 48290 }, { "grad_norm": 0.16775576770305634, "learning_rate": 5.692390452969344e-05, "loss": 0.0097, "step": 48300 }, { "grad_norm": 0.24536879360675812, "learning_rate": 5.690752875681392e-05, "loss": 0.0109, "step": 48310 }, { "grad_norm": 0.15127335488796234, "learning_rate": 5.6891152228537435e-05, "loss": 0.0109, "step": 48320 }, { "grad_norm": 0.2397259920835495, "learning_rate": 5.687477494665492e-05, "loss": 0.0089, "step": 48330 }, { "grad_norm": 0.1385311484336853, "learning_rate": 5.685839691295733e-05, "loss": 0.009, "step": 48340 }, { "grad_norm": 0.14814041554927826, "learning_rate": 5.6842018129235786e-05, "loss": 0.0092, "step": 48350 }, { "grad_norm": 0.14149151742458344, "learning_rate": 5.6825638597281404e-05, "loss": 0.0087, "step": 48360 }, { "grad_norm": 0.18287628889083862, "learning_rate": 5.680925831888546e-05, "loss": 0.01, "step": 48370 }, { "grad_norm": 0.19508138298988342, "learning_rate": 5.6792877295839274e-05, "loss": 0.0102, "step": 48380 }, { "grad_norm": 0.13991376757621765, "learning_rate": 5.6776495529934224e-05, "loss": 0.0089, "step": 48390 }, { "grad_norm": 0.12087976187467575, "learning_rate": 5.6760113022961824e-05, "loss": 0.0093, "step": 48400 }, { "grad_norm": 0.16805054247379303, "learning_rate": 5.6743729776713617e-05, "loss": 0.0103, "step": 48410 }, { "grad_norm": 0.1770739108324051, "learning_rate": 5.672734579298126e-05, "loss": 0.0086, "step": 48420 }, { "grad_norm": 0.32330313324928284, "learning_rate": 5.671096107355649e-05, "loss": 0.0118, "step": 48430 }, { "grad_norm": 0.19957682490348816, "learning_rate": 5.66945756202311e-05, "loss": 0.0114, "step": 48440 }, { "grad_norm": 0.19505704939365387, "learning_rate": 5.667818943479699e-05, "loss": 0.0129, "step": 48450 }, { "grad_norm": 0.12992845475673676, "learning_rate": 5.666180251904612e-05, "loss": 0.0102, "step": 48460 }, { "grad_norm": 0.16000260412693024, "learning_rate": 5.6645414874770555e-05, "loss": 0.0101, "step": 48470 }, { "grad_norm": 0.26266518235206604, "learning_rate": 5.66290265037624e-05, "loss": 0.0106, "step": 48480 }, { "grad_norm": 0.2101874202489853, "learning_rate": 5.661263740781386e-05, "loss": 0.0101, "step": 48490 }, { "grad_norm": 0.3296186923980713, "learning_rate": 5.6596247588717254e-05, "loss": 0.0122, "step": 48500 }, { "grad_norm": 0.20256635546684265, "learning_rate": 5.6579857048264926e-05, "loss": 0.0111, "step": 48510 }, { "grad_norm": 0.14348676800727844, "learning_rate": 5.6563465788249314e-05, "loss": 0.0095, "step": 48520 }, { "grad_norm": 0.24438093602657318, "learning_rate": 5.6547073810462956e-05, "loss": 0.0101, "step": 48530 }, { "grad_norm": 0.20998947322368622, "learning_rate": 5.653068111669846e-05, "loss": 0.0091, "step": 48540 }, { "grad_norm": 0.1759289652109146, "learning_rate": 5.651428770874848e-05, "loss": 0.0116, "step": 48550 }, { "grad_norm": 0.1887320578098297, "learning_rate": 5.64978935884058e-05, "loss": 0.0104, "step": 48560 }, { "grad_norm": 0.16739125549793243, "learning_rate": 5.6481498757463244e-05, "loss": 0.0109, "step": 48570 }, { "grad_norm": 0.15245893597602844, "learning_rate": 5.646510321771373e-05, "loss": 0.0105, "step": 48580 }, { "grad_norm": 0.21085792779922485, "learning_rate": 5.644870697095024e-05, "loss": 0.0092, "step": 48590 }, { "grad_norm": 0.25160518288612366, "learning_rate": 5.643231001896586e-05, "loss": 0.011, "step": 48600 }, { "grad_norm": 0.16625095903873444, "learning_rate": 5.6415912363553726e-05, "loss": 0.0104, "step": 48610 }, { "grad_norm": 0.2135075479745865, "learning_rate": 5.639951400650706e-05, "loss": 0.0114, "step": 48620 }, { "grad_norm": 0.20758557319641113, "learning_rate": 5.6383114949619165e-05, "loss": 0.0102, "step": 48630 }, { "grad_norm": 0.1311665028333664, "learning_rate": 5.636671519468342e-05, "loss": 0.0103, "step": 48640 }, { "grad_norm": 0.2899967432022095, "learning_rate": 5.635031474349327e-05, "loss": 0.0124, "step": 48650 }, { "grad_norm": 0.16173051297664642, "learning_rate": 5.6333913597842246e-05, "loss": 0.0104, "step": 48660 }, { "grad_norm": 0.11512326449155807, "learning_rate": 5.6317511759523955e-05, "loss": 0.0086, "step": 48670 }, { "grad_norm": 0.15674223005771637, "learning_rate": 5.630110923033207e-05, "loss": 0.0104, "step": 48680 }, { "grad_norm": 0.2459816038608551, "learning_rate": 5.628470601206036e-05, "loss": 0.0099, "step": 48690 }, { "grad_norm": 0.21108070015907288, "learning_rate": 5.626830210650266e-05, "loss": 0.0104, "step": 48700 }, { "grad_norm": 0.15609288215637207, "learning_rate": 5.625189751545285e-05, "loss": 0.0096, "step": 48710 }, { "grad_norm": 0.1365005075931549, "learning_rate": 5.6235492240704936e-05, "loss": 0.0091, "step": 48720 }, { "grad_norm": 0.2543545365333557, "learning_rate": 5.621908628405296e-05, "loss": 0.0118, "step": 48730 }, { "grad_norm": 0.11684141308069229, "learning_rate": 5.620267964729106e-05, "loss": 0.0089, "step": 48740 }, { "grad_norm": 0.17967750132083893, "learning_rate": 5.618627233221344e-05, "loss": 0.0109, "step": 48750 }, { "grad_norm": 0.16902290284633636, "learning_rate": 5.616986434061438e-05, "loss": 0.0108, "step": 48760 }, { "grad_norm": 0.14507238566875458, "learning_rate": 5.615345567428822e-05, "loss": 0.01, "step": 48770 }, { "grad_norm": 0.17247480154037476, "learning_rate": 5.613704633502941e-05, "loss": 0.0122, "step": 48780 }, { "grad_norm": 0.2079649418592453, "learning_rate": 5.612063632463245e-05, "loss": 0.0126, "step": 48790 }, { "grad_norm": 0.153224378824234, "learning_rate": 5.610422564489188e-05, "loss": 0.0121, "step": 48800 }, { "grad_norm": 0.1906474232673645, "learning_rate": 5.608781429760239e-05, "loss": 0.0111, "step": 48810 }, { "grad_norm": 0.15235461294651031, "learning_rate": 5.607140228455866e-05, "loss": 0.0094, "step": 48820 }, { "grad_norm": 0.17481622099876404, "learning_rate": 5.605498960755553e-05, "loss": 0.0106, "step": 48830 }, { "grad_norm": 0.2282295823097229, "learning_rate": 5.603857626838782e-05, "loss": 0.0106, "step": 48840 }, { "grad_norm": 0.223977729678154, "learning_rate": 5.60221622688505e-05, "loss": 0.0096, "step": 48850 }, { "grad_norm": 0.1865808665752411, "learning_rate": 5.6005747610738565e-05, "loss": 0.0103, "step": 48860 }, { "grad_norm": 0.20645320415496826, "learning_rate": 5.59893322958471e-05, "loss": 0.0094, "step": 48870 }, { "grad_norm": 0.125182643532753, "learning_rate": 5.5972916325971256e-05, "loss": 0.0109, "step": 48880 }, { "grad_norm": 0.18764953315258026, "learning_rate": 5.595649970290628e-05, "loss": 0.0094, "step": 48890 }, { "grad_norm": 0.1533372402191162, "learning_rate": 5.5940082428447426e-05, "loss": 0.0094, "step": 48900 }, { "grad_norm": 0.1576266884803772, "learning_rate": 5.592366450439012e-05, "loss": 0.0082, "step": 48910 }, { "grad_norm": 0.2483738213777542, "learning_rate": 5.590724593252975e-05, "loss": 0.0085, "step": 48920 }, { "grad_norm": 0.19291892647743225, "learning_rate": 5.589082671466184e-05, "loss": 0.0109, "step": 48930 }, { "grad_norm": 0.23523865640163422, "learning_rate": 5.587440685258199e-05, "loss": 0.0095, "step": 48940 }, { "grad_norm": 0.15335071086883545, "learning_rate": 5.585798634808583e-05, "loss": 0.0096, "step": 48950 }, { "grad_norm": 0.16500186920166016, "learning_rate": 5.584156520296909e-05, "loss": 0.0102, "step": 48960 }, { "grad_norm": 0.16067014634609222, "learning_rate": 5.582514341902757e-05, "loss": 0.0088, "step": 48970 }, { "grad_norm": 0.17422451078891754, "learning_rate": 5.580872099805713e-05, "loss": 0.0118, "step": 48980 }, { "grad_norm": 0.1616881787776947, "learning_rate": 5.5792297941853674e-05, "loss": 0.0103, "step": 48990 }, { "grad_norm": 0.11493687331676483, "learning_rate": 5.5775874252213247e-05, "loss": 0.0093, "step": 49000 }, { "grad_norm": 0.33275106549263, "learning_rate": 5.575944993093189e-05, "loss": 0.0115, "step": 49010 }, { "grad_norm": 0.11521100252866745, "learning_rate": 5.574302497980574e-05, "loss": 0.0136, "step": 49020 }, { "grad_norm": 0.2877221703529358, "learning_rate": 5.5726599400631005e-05, "loss": 0.0115, "step": 49030 }, { "grad_norm": 0.2006857544183731, "learning_rate": 5.571017319520399e-05, "loss": 0.009, "step": 49040 }, { "grad_norm": 0.18975454568862915, "learning_rate": 5.5693746365320984e-05, "loss": 0.0137, "step": 49050 }, { "grad_norm": 0.18289990723133087, "learning_rate": 5.567731891277848e-05, "loss": 0.017, "step": 49060 }, { "grad_norm": 0.19797144830226898, "learning_rate": 5.566089083937288e-05, "loss": 0.0143, "step": 49070 }, { "grad_norm": 0.255066454410553, "learning_rate": 5.564446214690079e-05, "loss": 0.0121, "step": 49080 }, { "grad_norm": 0.14325380325317383, "learning_rate": 5.56280328371588e-05, "loss": 0.0105, "step": 49090 }, { "grad_norm": 0.17860575020313263, "learning_rate": 5.5611602911943596e-05, "loss": 0.0117, "step": 49100 }, { "grad_norm": 0.4959457516670227, "learning_rate": 5.559517237305194e-05, "loss": 0.0116, "step": 49110 }, { "grad_norm": 0.2962120473384857, "learning_rate": 5.557874122228064e-05, "loss": 0.0097, "step": 49120 }, { "grad_norm": 0.28467443585395813, "learning_rate": 5.55623094614266e-05, "loss": 0.0099, "step": 49130 }, { "grad_norm": 0.17235228419303894, "learning_rate": 5.5545877092286743e-05, "loss": 0.0093, "step": 49140 }, { "grad_norm": 0.13501675426959991, "learning_rate": 5.552944411665811e-05, "loss": 0.0102, "step": 49150 }, { "grad_norm": 0.2062538117170334, "learning_rate": 5.5513010536337795e-05, "loss": 0.0107, "step": 49160 }, { "grad_norm": 0.28432244062423706, "learning_rate": 5.5496576353122934e-05, "loss": 0.0097, "step": 49170 }, { "grad_norm": 0.18551099300384521, "learning_rate": 5.548014156881074e-05, "loss": 0.0096, "step": 49180 }, { "grad_norm": 0.31906601786613464, "learning_rate": 5.5463706185198494e-05, "loss": 0.0106, "step": 49190 }, { "grad_norm": 0.1275237649679184, "learning_rate": 5.544727020408358e-05, "loss": 0.0114, "step": 49200 }, { "grad_norm": 0.17922942340373993, "learning_rate": 5.5430833627263366e-05, "loss": 0.0119, "step": 49210 }, { "grad_norm": 0.19845929741859436, "learning_rate": 5.5414396456535354e-05, "loss": 0.0082, "step": 49220 }, { "grad_norm": 0.1887039691209793, "learning_rate": 5.539795869369709e-05, "loss": 0.0097, "step": 49230 }, { "grad_norm": 0.15536479651927948, "learning_rate": 5.538152034054618e-05, "loss": 0.0113, "step": 49240 }, { "grad_norm": 0.18463629484176636, "learning_rate": 5.536508139888028e-05, "loss": 0.0101, "step": 49250 }, { "grad_norm": 0.17111898958683014, "learning_rate": 5.534864187049716e-05, "loss": 0.0093, "step": 49260 }, { "grad_norm": 0.26256734132766724, "learning_rate": 5.5332201757194615e-05, "loss": 0.0127, "step": 49270 }, { "grad_norm": 0.1315091997385025, "learning_rate": 5.531576106077049e-05, "loss": 0.0086, "step": 49280 }, { "grad_norm": 0.1289834827184677, "learning_rate": 5.529931978302272e-05, "loss": 0.0103, "step": 49290 }, { "grad_norm": 0.24179521203041077, "learning_rate": 5.528287792574932e-05, "loss": 0.0105, "step": 49300 }, { "grad_norm": 0.16333508491516113, "learning_rate": 5.5266435490748324e-05, "loss": 0.0085, "step": 49310 }, { "grad_norm": 0.08774235099554062, "learning_rate": 5.524999247981787e-05, "loss": 0.0091, "step": 49320 }, { "grad_norm": 0.23622386157512665, "learning_rate": 5.523354889475613e-05, "loss": 0.0104, "step": 49330 }, { "grad_norm": 0.11254753917455673, "learning_rate": 5.521710473736134e-05, "loss": 0.0094, "step": 49340 }, { "grad_norm": 0.19471082091331482, "learning_rate": 5.5200660009431826e-05, "loss": 0.0095, "step": 49350 }, { "grad_norm": 0.23898757994174957, "learning_rate": 5.518421471276596e-05, "loss": 0.0107, "step": 49360 }, { "grad_norm": 0.1523103564977646, "learning_rate": 5.5167768849162174e-05, "loss": 0.0095, "step": 49370 }, { "grad_norm": 0.13199377059936523, "learning_rate": 5.515132242041893e-05, "loss": 0.0097, "step": 49380 }, { "grad_norm": 0.1470002830028534, "learning_rate": 5.513487542833483e-05, "loss": 0.0092, "step": 49390 }, { "grad_norm": 0.15630067884922028, "learning_rate": 5.5118427874708466e-05, "loss": 0.0084, "step": 49400 }, { "grad_norm": 0.13840332627296448, "learning_rate": 5.510197976133853e-05, "loss": 0.0109, "step": 49410 }, { "grad_norm": 0.23009471595287323, "learning_rate": 5.508553109002376e-05, "loss": 0.0092, "step": 49420 }, { "grad_norm": 0.16043634712696075, "learning_rate": 5.5069081862562957e-05, "loss": 0.01, "step": 49430 }, { "grad_norm": 0.14084526896476746, "learning_rate": 5.5052632080754965e-05, "loss": 0.0104, "step": 49440 }, { "grad_norm": 0.1527092158794403, "learning_rate": 5.503618174639874e-05, "loss": 0.0087, "step": 49450 }, { "grad_norm": 0.2808614671230316, "learning_rate": 5.501973086129325e-05, "loss": 0.0091, "step": 49460 }, { "grad_norm": 0.22003857791423798, "learning_rate": 5.500327942723753e-05, "loss": 0.0086, "step": 49470 }, { "grad_norm": 0.21938666701316833, "learning_rate": 5.498682744603071e-05, "loss": 0.0103, "step": 49480 }, { "grad_norm": 0.15060637891292572, "learning_rate": 5.497037491947194e-05, "loss": 0.0102, "step": 49490 }, { "grad_norm": 0.2128259539604187, "learning_rate": 5.4953921849360424e-05, "loss": 0.0088, "step": 49500 }, { "grad_norm": 0.1784733086824417, "learning_rate": 5.493746823749547e-05, "loss": 0.0092, "step": 49510 }, { "grad_norm": 0.27124765515327454, "learning_rate": 5.49210140856764e-05, "loss": 0.0095, "step": 49520 }, { "grad_norm": 0.2034704089164734, "learning_rate": 5.4904559395702647e-05, "loss": 0.0102, "step": 49530 }, { "grad_norm": 0.3249327540397644, "learning_rate": 5.488810416937364e-05, "loss": 0.0103, "step": 49540 }, { "grad_norm": 0.2537223696708679, "learning_rate": 5.4871648408488926e-05, "loss": 0.0118, "step": 49550 }, { "grad_norm": 0.14341036975383759, "learning_rate": 5.485519211484807e-05, "loss": 0.0093, "step": 49560 }, { "grad_norm": 0.1751876026391983, "learning_rate": 5.4838735290250676e-05, "loss": 0.0098, "step": 49570 }, { "grad_norm": 0.17694583535194397, "learning_rate": 5.482227793649649e-05, "loss": 0.01, "step": 49580 }, { "grad_norm": 0.20933954417705536, "learning_rate": 5.480582005538524e-05, "loss": 0.0112, "step": 49590 }, { "grad_norm": 0.23190723359584808, "learning_rate": 5.478936164871671e-05, "loss": 0.0094, "step": 49600 }, { "grad_norm": 0.13724540174007416, "learning_rate": 5.477290271829081e-05, "loss": 0.01, "step": 49610 }, { "grad_norm": 0.20793557167053223, "learning_rate": 5.475644326590744e-05, "loss": 0.0091, "step": 49620 }, { "grad_norm": 0.14432422816753387, "learning_rate": 5.473998329336658e-05, "loss": 0.0096, "step": 49630 }, { "grad_norm": 0.17734099924564362, "learning_rate": 5.4723522802468286e-05, "loss": 0.0122, "step": 49640 }, { "grad_norm": 0.1579577475786209, "learning_rate": 5.470706179501264e-05, "loss": 0.0095, "step": 49650 }, { "grad_norm": 0.1296856850385666, "learning_rate": 5.4690600272799776e-05, "loss": 0.0088, "step": 49660 }, { "grad_norm": 0.15072540938854218, "learning_rate": 5.467413823762993e-05, "loss": 0.009, "step": 49670 }, { "grad_norm": 0.17484590411186218, "learning_rate": 5.465767569130335e-05, "loss": 0.0083, "step": 49680 }, { "grad_norm": 0.1212497428059578, "learning_rate": 5.464121263562036e-05, "loss": 0.0101, "step": 49690 }, { "grad_norm": 0.18129514157772064, "learning_rate": 5.4624749072381343e-05, "loss": 0.0104, "step": 49700 }, { "grad_norm": 0.14615242183208466, "learning_rate": 5.460828500338672e-05, "loss": 0.0092, "step": 49710 }, { "grad_norm": 0.19891230762004852, "learning_rate": 5.459182043043698e-05, "loss": 0.0095, "step": 49720 }, { "grad_norm": 0.13500772416591644, "learning_rate": 5.457535535533265e-05, "loss": 0.0096, "step": 49730 }, { "grad_norm": 0.19354091584682465, "learning_rate": 5.4558889779874334e-05, "loss": 0.0084, "step": 49740 }, { "grad_norm": 0.12780708074569702, "learning_rate": 5.454242370586269e-05, "loss": 0.0093, "step": 49750 }, { "grad_norm": 0.1206326112151146, "learning_rate": 5.452595713509843e-05, "loss": 0.0099, "step": 49760 }, { "grad_norm": 0.23471838235855103, "learning_rate": 5.4509490069382274e-05, "loss": 0.0079, "step": 49770 }, { "grad_norm": 0.1629716455936432, "learning_rate": 5.449302251051509e-05, "loss": 0.0096, "step": 49780 }, { "grad_norm": 0.24686148762702942, "learning_rate": 5.44765544602977e-05, "loss": 0.0087, "step": 49790 }, { "grad_norm": 0.2525096535682678, "learning_rate": 5.446008592053107e-05, "loss": 0.0101, "step": 49800 }, { "grad_norm": 0.25819480419158936, "learning_rate": 5.4443616893016135e-05, "loss": 0.009, "step": 49810 }, { "grad_norm": 0.2294527292251587, "learning_rate": 5.4427147379553934e-05, "loss": 0.011, "step": 49820 }, { "grad_norm": 0.16722144186496735, "learning_rate": 5.441067738194556e-05, "loss": 0.0112, "step": 49830 }, { "grad_norm": 0.14256823062896729, "learning_rate": 5.439420690199214e-05, "loss": 0.0098, "step": 49840 }, { "grad_norm": 0.1211332455277443, "learning_rate": 5.437773594149484e-05, "loss": 0.0103, "step": 49850 }, { "grad_norm": 0.19927777349948883, "learning_rate": 5.436126450225495e-05, "loss": 0.0113, "step": 49860 }, { "grad_norm": 0.2638828754425049, "learning_rate": 5.434479258607371e-05, "loss": 0.0107, "step": 49870 }, { "grad_norm": 0.21242041885852814, "learning_rate": 5.43283201947525e-05, "loss": 0.0099, "step": 49880 }, { "grad_norm": 0.16035936772823334, "learning_rate": 5.431184733009268e-05, "loss": 0.0085, "step": 49890 }, { "grad_norm": 0.15246866643428802, "learning_rate": 5.4295373993895736e-05, "loss": 0.0092, "step": 49900 }, { "grad_norm": 0.18964919447898865, "learning_rate": 5.4278900187963157e-05, "loss": 0.0089, "step": 49910 }, { "grad_norm": 0.18547284603118896, "learning_rate": 5.426242591409646e-05, "loss": 0.0105, "step": 49920 }, { "grad_norm": 0.17110265791416168, "learning_rate": 5.42459511740973e-05, "loss": 0.0092, "step": 49930 }, { "grad_norm": 0.1770339459180832, "learning_rate": 5.422947596976729e-05, "loss": 0.0118, "step": 49940 }, { "grad_norm": 0.1602790653705597, "learning_rate": 5.4213000302908134e-05, "loss": 0.0115, "step": 49950 }, { "grad_norm": 0.1671534925699234, "learning_rate": 5.419652417532162e-05, "loss": 0.0075, "step": 49960 }, { "grad_norm": 0.1518758237361908, "learning_rate": 5.4180047588809534e-05, "loss": 0.0087, "step": 49970 }, { "grad_norm": 0.1527891606092453, "learning_rate": 5.4163570545173704e-05, "loss": 0.0088, "step": 49980 }, { "grad_norm": 0.17880786955356598, "learning_rate": 5.414709304621608e-05, "loss": 0.0093, "step": 49990 }, { "grad_norm": 0.3460046947002411, "learning_rate": 5.413061509373858e-05, "loss": 0.0093, "step": 50000 }, { "grad_norm": 0.14409217238426208, "learning_rate": 5.411413668954324e-05, "loss": 0.0091, "step": 50010 }, { "grad_norm": 0.18430009484291077, "learning_rate": 5.4097657835432083e-05, "loss": 0.0098, "step": 50020 }, { "grad_norm": 0.16519367694854736, "learning_rate": 5.408117853320723e-05, "loss": 0.008, "step": 50030 }, { "grad_norm": 0.14772537350654602, "learning_rate": 5.406469878467084e-05, "loss": 0.01, "step": 50040 }, { "grad_norm": 0.15694692730903625, "learning_rate": 5.404821859162509e-05, "loss": 0.0093, "step": 50050 }, { "grad_norm": 0.14357957243919373, "learning_rate": 5.403173795587225e-05, "loss": 0.011, "step": 50060 }, { "grad_norm": 0.134568989276886, "learning_rate": 5.4015256879214606e-05, "loss": 0.0116, "step": 50070 }, { "grad_norm": 0.18768475949764252, "learning_rate": 5.39987753634545e-05, "loss": 0.0098, "step": 50080 }, { "grad_norm": 0.1319604367017746, "learning_rate": 5.3982293410394335e-05, "loss": 0.0097, "step": 50090 }, { "grad_norm": 0.1442621797323227, "learning_rate": 5.396581102183655e-05, "loss": 0.0091, "step": 50100 }, { "grad_norm": 0.1396232545375824, "learning_rate": 5.3949328199583634e-05, "loss": 0.0098, "step": 50110 }, { "grad_norm": 0.17236685752868652, "learning_rate": 5.393284494543813e-05, "loss": 0.0096, "step": 50120 }, { "grad_norm": 0.12278735637664795, "learning_rate": 5.391636126120262e-05, "loss": 0.0098, "step": 50130 }, { "grad_norm": 0.1477465182542801, "learning_rate": 5.389987714867971e-05, "loss": 0.0092, "step": 50140 }, { "grad_norm": 0.15739701688289642, "learning_rate": 5.3883392609672114e-05, "loss": 0.0105, "step": 50150 }, { "grad_norm": 0.14583677053451538, "learning_rate": 5.3866907645982545e-05, "loss": 0.0084, "step": 50160 }, { "grad_norm": 0.30898013710975647, "learning_rate": 5.385042225941375e-05, "loss": 0.0121, "step": 50170 }, { "grad_norm": 0.15822359919548035, "learning_rate": 5.3833936451768585e-05, "loss": 0.0107, "step": 50180 }, { "grad_norm": 0.11612649261951447, "learning_rate": 5.3817450224849884e-05, "loss": 0.0108, "step": 50190 }, { "grad_norm": 0.21819405257701874, "learning_rate": 5.380096358046056e-05, "loss": 0.0128, "step": 50200 }, { "grad_norm": 0.17855000495910645, "learning_rate": 5.378447652040359e-05, "loss": 0.0093, "step": 50210 }, { "grad_norm": 0.15031445026397705, "learning_rate": 5.376798904648195e-05, "loss": 0.0107, "step": 50220 }, { "grad_norm": 0.18227490782737732, "learning_rate": 5.375150116049869e-05, "loss": 0.009, "step": 50230 }, { "grad_norm": 0.21196730434894562, "learning_rate": 5.373501286425691e-05, "loss": 0.009, "step": 50240 }, { "grad_norm": 0.14049772918224335, "learning_rate": 5.3718524159559726e-05, "loss": 0.0077, "step": 50250 }, { "grad_norm": 0.17152342200279236, "learning_rate": 5.370203504821034e-05, "loss": 0.0093, "step": 50260 }, { "grad_norm": 0.27798956632614136, "learning_rate": 5.368554553201196e-05, "loss": 0.0105, "step": 50270 }, { "grad_norm": 0.17591355741024017, "learning_rate": 5.3669055612767874e-05, "loss": 0.0101, "step": 50280 }, { "grad_norm": 0.1363784670829773, "learning_rate": 5.3652565292281376e-05, "loss": 0.0094, "step": 50290 }, { "grad_norm": 0.1339588314294815, "learning_rate": 5.3636074572355824e-05, "loss": 0.0094, "step": 50300 }, { "grad_norm": 0.17324349284172058, "learning_rate": 5.361958345479463e-05, "loss": 0.0086, "step": 50310 }, { "grad_norm": 0.16480936110019684, "learning_rate": 5.360309194140123e-05, "loss": 0.0089, "step": 50320 }, { "grad_norm": 0.16656620800495148, "learning_rate": 5.3586600033979106e-05, "loss": 0.0099, "step": 50330 }, { "grad_norm": 0.22566993534564972, "learning_rate": 5.357010773433181e-05, "loss": 0.0094, "step": 50340 }, { "grad_norm": 0.13061313331127167, "learning_rate": 5.3553615044262895e-05, "loss": 0.0083, "step": 50350 }, { "grad_norm": 0.1891506314277649, "learning_rate": 5.3537121965575964e-05, "loss": 0.0084, "step": 50360 }, { "grad_norm": 0.16591797769069672, "learning_rate": 5.352062850007471e-05, "loss": 0.0101, "step": 50370 }, { "grad_norm": 0.1301375776529312, "learning_rate": 5.350413464956282e-05, "loss": 0.0093, "step": 50380 }, { "grad_norm": 0.20164068043231964, "learning_rate": 5.348764041584403e-05, "loss": 0.0089, "step": 50390 }, { "grad_norm": 0.18161600828170776, "learning_rate": 5.347114580072212e-05, "loss": 0.0098, "step": 50400 }, { "grad_norm": 0.1357002854347229, "learning_rate": 5.345465080600093e-05, "loss": 0.0116, "step": 50410 }, { "grad_norm": 0.16485589742660522, "learning_rate": 5.3438155433484326e-05, "loss": 0.0103, "step": 50420 }, { "grad_norm": 0.2364126443862915, "learning_rate": 5.3421659684976197e-05, "loss": 0.0092, "step": 50430 }, { "grad_norm": 0.19214558601379395, "learning_rate": 5.340516356228052e-05, "loss": 0.008, "step": 50440 }, { "grad_norm": 0.24972032010555267, "learning_rate": 5.338866706720128e-05, "loss": 0.0104, "step": 50450 }, { "grad_norm": 0.2342538982629776, "learning_rate": 5.337217020154249e-05, "loss": 0.0106, "step": 50460 }, { "grad_norm": 0.34448665380477905, "learning_rate": 5.335567296710825e-05, "loss": 0.0114, "step": 50470 }, { "grad_norm": 0.1685565561056137, "learning_rate": 5.333917536570265e-05, "loss": 0.0091, "step": 50480 }, { "grad_norm": 0.19756121933460236, "learning_rate": 5.332267739912986e-05, "loss": 0.0095, "step": 50490 }, { "grad_norm": 0.15805959701538086, "learning_rate": 5.330617906919405e-05, "loss": 0.0084, "step": 50500 }, { "grad_norm": 0.1502276510000229, "learning_rate": 5.328968037769949e-05, "loss": 0.011, "step": 50510 }, { "grad_norm": 0.18652021884918213, "learning_rate": 5.327318132645042e-05, "loss": 0.009, "step": 50520 }, { "grad_norm": 0.16499777138233185, "learning_rate": 5.325668191725116e-05, "loss": 0.0088, "step": 50530 }, { "grad_norm": 0.17984707653522491, "learning_rate": 5.324018215190606e-05, "loss": 0.0091, "step": 50540 }, { "grad_norm": 0.14197231829166412, "learning_rate": 5.3223682032219515e-05, "loss": 0.0098, "step": 50550 }, { "grad_norm": 0.32791078090667725, "learning_rate": 5.320718155999595e-05, "loss": 0.0101, "step": 50560 }, { "grad_norm": 0.18109092116355896, "learning_rate": 5.3190680737039835e-05, "loss": 0.0115, "step": 50570 }, { "grad_norm": 0.17931252717971802, "learning_rate": 5.317417956515567e-05, "loss": 0.0091, "step": 50580 }, { "grad_norm": 0.12720361351966858, "learning_rate": 5.315767804614798e-05, "loss": 0.0095, "step": 50590 }, { "grad_norm": 0.15880152583122253, "learning_rate": 5.3141176181821395e-05, "loss": 0.01, "step": 50600 }, { "grad_norm": 0.11388847976922989, "learning_rate": 5.31246739739805e-05, "loss": 0.0107, "step": 50610 }, { "grad_norm": 0.23510198295116425, "learning_rate": 5.310817142442994e-05, "loss": 0.0104, "step": 50620 }, { "grad_norm": 0.17208291590213776, "learning_rate": 5.309166853497445e-05, "loss": 0.0092, "step": 50630 }, { "grad_norm": 0.19078657031059265, "learning_rate": 5.307516530741873e-05, "loss": 0.0107, "step": 50640 }, { "grad_norm": 0.11700933426618576, "learning_rate": 5.305866174356754e-05, "loss": 0.01, "step": 50650 }, { "grad_norm": 0.1568746715784073, "learning_rate": 5.304215784522571e-05, "loss": 0.0087, "step": 50660 }, { "grad_norm": 0.19977030158042908, "learning_rate": 5.302565361419808e-05, "loss": 0.0087, "step": 50670 }, { "grad_norm": 0.14386771619319916, "learning_rate": 5.3009149052289507e-05, "loss": 0.0089, "step": 50680 }, { "grad_norm": 0.18500393629074097, "learning_rate": 5.299264416130493e-05, "loss": 0.0085, "step": 50690 }, { "grad_norm": 0.17917723953723907, "learning_rate": 5.297613894304928e-05, "loss": 0.0098, "step": 50700 }, { "grad_norm": 0.32785212993621826, "learning_rate": 5.2959633399327534e-05, "loss": 0.0107, "step": 50710 }, { "grad_norm": 0.16070692241191864, "learning_rate": 5.294312753194476e-05, "loss": 0.0123, "step": 50720 }, { "grad_norm": 0.17883041501045227, "learning_rate": 5.292662134270596e-05, "loss": 0.0082, "step": 50730 }, { "grad_norm": 0.18532903492450714, "learning_rate": 5.291011483341626e-05, "loss": 0.0092, "step": 50740 }, { "grad_norm": 0.16035261750221252, "learning_rate": 5.2893608005880767e-05, "loss": 0.0113, "step": 50750 }, { "grad_norm": 0.15199849009513855, "learning_rate": 5.287710086190467e-05, "loss": 0.011, "step": 50760 }, { "grad_norm": 0.14063483476638794, "learning_rate": 5.286059340329314e-05, "loss": 0.0098, "step": 50770 }, { "grad_norm": 0.17500406503677368, "learning_rate": 5.284408563185141e-05, "loss": 0.0118, "step": 50780 }, { "grad_norm": 0.22060255706310272, "learning_rate": 5.282757754938476e-05, "loss": 0.0107, "step": 50790 }, { "grad_norm": 0.22756953537464142, "learning_rate": 5.281106915769849e-05, "loss": 0.0109, "step": 50800 }, { "grad_norm": 0.19213804602622986, "learning_rate": 5.2794560458597897e-05, "loss": 0.0087, "step": 50810 }, { "grad_norm": 0.15577614307403564, "learning_rate": 5.27780514538884e-05, "loss": 0.0105, "step": 50820 }, { "grad_norm": 0.1253642737865448, "learning_rate": 5.2761542145375365e-05, "loss": 0.0092, "step": 50830 }, { "grad_norm": 0.14584238827228546, "learning_rate": 5.274503253486421e-05, "loss": 0.0088, "step": 50840 }, { "grad_norm": 0.12658663094043732, "learning_rate": 5.272852262416046e-05, "loss": 0.008, "step": 50850 }, { "grad_norm": 0.13190124928951263, "learning_rate": 5.2712012415069555e-05, "loss": 0.0108, "step": 50860 }, { "grad_norm": 0.15021578967571259, "learning_rate": 5.269550190939705e-05, "loss": 0.0089, "step": 50870 }, { "grad_norm": 0.18545642495155334, "learning_rate": 5.267899110894852e-05, "loss": 0.009, "step": 50880 }, { "grad_norm": 0.21037696301937103, "learning_rate": 5.266248001552955e-05, "loss": 0.0093, "step": 50890 }, { "grad_norm": 0.2669650912284851, "learning_rate": 5.264596863094575e-05, "loss": 0.0101, "step": 50900 }, { "grad_norm": 0.2299928367137909, "learning_rate": 5.262945695700282e-05, "loss": 0.0103, "step": 50910 }, { "grad_norm": 0.16686749458312988, "learning_rate": 5.261294499550643e-05, "loss": 0.013, "step": 50920 }, { "grad_norm": 0.20045721530914307, "learning_rate": 5.25964327482623e-05, "loss": 0.0118, "step": 50930 }, { "grad_norm": 0.20620253682136536, "learning_rate": 5.257992021707617e-05, "loss": 0.0097, "step": 50940 }, { "grad_norm": 0.20445042848587036, "learning_rate": 5.256340740375387e-05, "loss": 0.0103, "step": 50950 }, { "grad_norm": 0.19352494180202484, "learning_rate": 5.254689431010117e-05, "loss": 0.0094, "step": 50960 }, { "grad_norm": 0.1971249133348465, "learning_rate": 5.253038093792395e-05, "loss": 0.0088, "step": 50970 }, { "grad_norm": 0.1974291056394577, "learning_rate": 5.251386728902806e-05, "loss": 0.0109, "step": 50980 }, { "grad_norm": 0.2586503028869629, "learning_rate": 5.2497353365219446e-05, "loss": 0.0109, "step": 50990 }, { "grad_norm": 0.33544859290122986, "learning_rate": 5.2480839168304e-05, "loss": 0.0113, "step": 51000 }, { "grad_norm": 0.1367875635623932, "learning_rate": 5.246432470008773e-05, "loss": 0.0088, "step": 51010 }, { "grad_norm": 0.21750110387802124, "learning_rate": 5.2447809962376616e-05, "loss": 0.0084, "step": 51020 }, { "grad_norm": 0.3985450267791748, "learning_rate": 5.243129495697666e-05, "loss": 0.0095, "step": 51030 }, { "grad_norm": 0.21807274222373962, "learning_rate": 5.241477968569397e-05, "loss": 0.0097, "step": 51040 }, { "grad_norm": 0.28157174587249756, "learning_rate": 5.23982641503346e-05, "loss": 0.0098, "step": 51050 }, { "grad_norm": 0.3782063126564026, "learning_rate": 5.238174835270465e-05, "loss": 0.0108, "step": 51060 }, { "grad_norm": 0.21495342254638672, "learning_rate": 5.23652322946103e-05, "loss": 0.0092, "step": 51070 }, { "grad_norm": 0.13268698751926422, "learning_rate": 5.234871597785771e-05, "loss": 0.01, "step": 51080 }, { "grad_norm": 0.1444823145866394, "learning_rate": 5.233219940425306e-05, "loss": 0.008, "step": 51090 }, { "grad_norm": 0.13181231915950775, "learning_rate": 5.2315682575602576e-05, "loss": 0.0104, "step": 51100 }, { "grad_norm": 0.18035440146923065, "learning_rate": 5.229916549371254e-05, "loss": 0.01, "step": 51110 }, { "grad_norm": 0.11848296970129013, "learning_rate": 5.228264816038924e-05, "loss": 0.0093, "step": 51120 }, { "grad_norm": 0.23542794585227966, "learning_rate": 5.2266130577438935e-05, "loss": 0.0097, "step": 51130 }, { "grad_norm": 0.16684666275978088, "learning_rate": 5.224961274666802e-05, "loss": 0.0085, "step": 51140 }, { "grad_norm": 0.17607922852039337, "learning_rate": 5.223309466988283e-05, "loss": 0.0099, "step": 51150 }, { "grad_norm": 0.16469788551330566, "learning_rate": 5.221657634888976e-05, "loss": 0.0094, "step": 51160 }, { "grad_norm": 0.1700674593448639, "learning_rate": 5.2200057785495235e-05, "loss": 0.0117, "step": 51170 }, { "grad_norm": 0.16417469084262848, "learning_rate": 5.2183538981505706e-05, "loss": 0.0106, "step": 51180 }, { "grad_norm": 0.15616858005523682, "learning_rate": 5.216701993872762e-05, "loss": 0.0095, "step": 51190 }, { "grad_norm": 0.159697487950325, "learning_rate": 5.215050065896752e-05, "loss": 0.0078, "step": 51200 }, { "grad_norm": 0.20332244038581848, "learning_rate": 5.2133981144031875e-05, "loss": 0.0076, "step": 51210 }, { "grad_norm": 0.23451977968215942, "learning_rate": 5.211746139572727e-05, "loss": 0.01, "step": 51220 }, { "grad_norm": 0.1581573784351349, "learning_rate": 5.210094141586026e-05, "loss": 0.0091, "step": 51230 }, { "grad_norm": 0.2060880959033966, "learning_rate": 5.2084421206237476e-05, "loss": 0.0086, "step": 51240 }, { "grad_norm": 0.11891312897205353, "learning_rate": 5.20679007686655e-05, "loss": 0.0081, "step": 51250 }, { "grad_norm": 0.13010798394680023, "learning_rate": 5.205138010495102e-05, "loss": 0.0093, "step": 51260 }, { "grad_norm": 0.1690797060728073, "learning_rate": 5.20348592169007e-05, "loss": 0.0074, "step": 51270 }, { "grad_norm": 0.16077561676502228, "learning_rate": 5.201833810632123e-05, "loss": 0.0088, "step": 51280 }, { "grad_norm": 0.19164034724235535, "learning_rate": 5.200181677501933e-05, "loss": 0.0085, "step": 51290 }, { "grad_norm": 0.37639540433883667, "learning_rate": 5.1985295224801777e-05, "loss": 0.0091, "step": 51300 }, { "grad_norm": 0.17741000652313232, "learning_rate": 5.196877345747531e-05, "loss": 0.0098, "step": 51310 }, { "grad_norm": 0.1431676298379898, "learning_rate": 5.195225147484674e-05, "loss": 0.0081, "step": 51320 }, { "grad_norm": 0.16591793298721313, "learning_rate": 5.193572927872288e-05, "loss": 0.0094, "step": 51330 }, { "grad_norm": 0.15258292853832245, "learning_rate": 5.191920687091059e-05, "loss": 0.0093, "step": 51340 }, { "grad_norm": 0.162583127617836, "learning_rate": 5.1902684253216695e-05, "loss": 0.0086, "step": 51350 }, { "grad_norm": 0.20253436267375946, "learning_rate": 5.1886161427448134e-05, "loss": 0.0095, "step": 51360 }, { "grad_norm": 0.2708319425582886, "learning_rate": 5.186963839541179e-05, "loss": 0.0094, "step": 51370 }, { "grad_norm": 0.13930058479309082, "learning_rate": 5.185311515891459e-05, "loss": 0.012, "step": 51380 }, { "grad_norm": 0.30371636152267456, "learning_rate": 5.1836591719763516e-05, "loss": 0.0086, "step": 51390 }, { "grad_norm": 0.19266772270202637, "learning_rate": 5.182006807976554e-05, "loss": 0.0081, "step": 51400 }, { "grad_norm": 0.2179098129272461, "learning_rate": 5.180354424072763e-05, "loss": 0.0087, "step": 51410 }, { "grad_norm": 0.13511958718299866, "learning_rate": 5.178702020445686e-05, "loss": 0.009, "step": 51420 }, { "grad_norm": 0.3006044328212738, "learning_rate": 5.177049597276024e-05, "loss": 0.0105, "step": 51430 }, { "grad_norm": 0.18473446369171143, "learning_rate": 5.175397154744482e-05, "loss": 0.0119, "step": 51440 }, { "grad_norm": 0.13122697174549103, "learning_rate": 5.1737446930317736e-05, "loss": 0.0117, "step": 51450 }, { "grad_norm": 0.46058183908462524, "learning_rate": 5.172092212318604e-05, "loss": 0.0105, "step": 51460 }, { "grad_norm": 0.15450617671012878, "learning_rate": 5.170439712785692e-05, "loss": 0.0102, "step": 51470 }, { "grad_norm": 0.15867310762405396, "learning_rate": 5.1687871946137467e-05, "loss": 0.01, "step": 51480 }, { "grad_norm": 0.2316945195198059, "learning_rate": 5.16713465798349e-05, "loss": 0.0112, "step": 51490 }, { "grad_norm": 0.1370791643857956, "learning_rate": 5.165482103075637e-05, "loss": 0.009, "step": 51500 }, { "grad_norm": 0.17790478467941284, "learning_rate": 5.163829530070909e-05, "loss": 0.0103, "step": 51510 }, { "grad_norm": 0.145208939909935, "learning_rate": 5.162176939150032e-05, "loss": 0.0092, "step": 51520 }, { "grad_norm": 0.175679549574852, "learning_rate": 5.160524330493728e-05, "loss": 0.0083, "step": 51530 }, { "grad_norm": 0.10568757355213165, "learning_rate": 5.158871704282724e-05, "loss": 0.0129, "step": 51540 }, { "grad_norm": 0.44641247391700745, "learning_rate": 5.15721906069775e-05, "loss": 0.0101, "step": 51550 }, { "grad_norm": 0.1808238923549652, "learning_rate": 5.155566399919536e-05, "loss": 0.0108, "step": 51560 }, { "grad_norm": 0.22924251854419708, "learning_rate": 5.153913722128813e-05, "loss": 0.0088, "step": 51570 }, { "grad_norm": 0.17091476917266846, "learning_rate": 5.1522610275063176e-05, "loss": 0.0098, "step": 51580 }, { "grad_norm": 0.19003276526927948, "learning_rate": 5.150608316232786e-05, "loss": 0.0093, "step": 51590 }, { "grad_norm": 0.09866583347320557, "learning_rate": 5.148955588488955e-05, "loss": 0.0087, "step": 51600 }, { "grad_norm": 0.23069192469120026, "learning_rate": 5.147302844455564e-05, "loss": 0.0111, "step": 51610 }, { "grad_norm": 0.24319209158420563, "learning_rate": 5.145650084313356e-05, "loss": 0.011, "step": 51620 }, { "grad_norm": 0.12213748693466187, "learning_rate": 5.1439973082430734e-05, "loss": 0.0119, "step": 51630 }, { "grad_norm": 0.16560645401477814, "learning_rate": 5.142344516425461e-05, "loss": 0.0104, "step": 51640 }, { "grad_norm": 0.18531808257102966, "learning_rate": 5.1406917090412673e-05, "loss": 0.0101, "step": 51650 }, { "grad_norm": 0.1730813980102539, "learning_rate": 5.1390388862712404e-05, "loss": 0.0093, "step": 51660 }, { "grad_norm": 0.15252114832401276, "learning_rate": 5.137386048296129e-05, "loss": 0.0107, "step": 51670 }, { "grad_norm": 0.21394912898540497, "learning_rate": 5.135733195296686e-05, "loss": 0.0102, "step": 51680 }, { "grad_norm": 0.23353569209575653, "learning_rate": 5.134080327453664e-05, "loss": 0.0103, "step": 51690 }, { "grad_norm": 0.17222163081169128, "learning_rate": 5.1324274449478204e-05, "loss": 0.0086, "step": 51700 }, { "grad_norm": 0.11934859305620193, "learning_rate": 5.130774547959909e-05, "loss": 0.0092, "step": 51710 }, { "grad_norm": 0.16020797193050385, "learning_rate": 5.129121636670692e-05, "loss": 0.0097, "step": 51720 }, { "grad_norm": 0.18172188103199005, "learning_rate": 5.1274687112609246e-05, "loss": 0.0093, "step": 51730 }, { "grad_norm": 0.1472824662923813, "learning_rate": 5.125815771911372e-05, "loss": 0.0093, "step": 51740 }, { "grad_norm": 0.22437629103660583, "learning_rate": 5.1241628188027956e-05, "loss": 0.009, "step": 51750 }, { "grad_norm": 0.28644460439682007, "learning_rate": 5.1225098521159585e-05, "loss": 0.0091, "step": 51760 }, { "grad_norm": 0.14755836129188538, "learning_rate": 5.12085687203163e-05, "loss": 0.0089, "step": 51770 }, { "grad_norm": 0.26728588342666626, "learning_rate": 5.119203878730575e-05, "loss": 0.0072, "step": 51780 }, { "grad_norm": 0.17952430248260498, "learning_rate": 5.117550872393564e-05, "loss": 0.011, "step": 51790 }, { "grad_norm": 0.11612691730260849, "learning_rate": 5.1158978532013645e-05, "loss": 0.0075, "step": 51800 }, { "grad_norm": 0.14830532670021057, "learning_rate": 5.114244821334751e-05, "loss": 0.0084, "step": 51810 }, { "grad_norm": 0.2406485676765442, "learning_rate": 5.112591776974497e-05, "loss": 0.0084, "step": 51820 }, { "grad_norm": 0.2135319858789444, "learning_rate": 5.110938720301374e-05, "loss": 0.0096, "step": 51830 }, { "grad_norm": 0.1296909898519516, "learning_rate": 5.1092856514961595e-05, "loss": 0.0089, "step": 51840 }, { "grad_norm": 0.14329344034194946, "learning_rate": 5.107632570739631e-05, "loss": 0.0091, "step": 51850 }, { "grad_norm": 0.21818552911281586, "learning_rate": 5.1059794782125645e-05, "loss": 0.0102, "step": 51860 }, { "grad_norm": 0.27818775177001953, "learning_rate": 5.104326374095744e-05, "loss": 0.0107, "step": 51870 }, { "grad_norm": 0.23083850741386414, "learning_rate": 5.102673258569947e-05, "loss": 0.0088, "step": 51880 }, { "grad_norm": 0.13344824314117432, "learning_rate": 5.101020131815956e-05, "loss": 0.0072, "step": 51890 }, { "grad_norm": 0.14781583845615387, "learning_rate": 5.099366994014556e-05, "loss": 0.0085, "step": 51900 }, { "grad_norm": 0.15806721150875092, "learning_rate": 5.097713845346532e-05, "loss": 0.008, "step": 51910 }, { "grad_norm": 0.1436925083398819, "learning_rate": 5.096060685992668e-05, "loss": 0.0096, "step": 51920 }, { "grad_norm": 0.16603249311447144, "learning_rate": 5.0944075161337514e-05, "loss": 0.0096, "step": 51930 }, { "grad_norm": 0.23275810480117798, "learning_rate": 5.09275433595057e-05, "loss": 0.0096, "step": 51940 }, { "grad_norm": 0.17597579956054688, "learning_rate": 5.0911011456239157e-05, "loss": 0.0088, "step": 51950 }, { "grad_norm": 0.15292273461818695, "learning_rate": 5.089447945334574e-05, "loss": 0.0108, "step": 51960 }, { "grad_norm": 0.11878424137830734, "learning_rate": 5.087794735263343e-05, "loss": 0.0095, "step": 51970 }, { "grad_norm": 0.18594487011432648, "learning_rate": 5.086141515591011e-05, "loss": 0.0105, "step": 51980 }, { "grad_norm": 0.1752600520849228, "learning_rate": 5.084488286498371e-05, "loss": 0.0092, "step": 51990 }, { "grad_norm": 0.14803959429264069, "learning_rate": 5.08283504816622e-05, "loss": 0.0112, "step": 52000 }, { "grad_norm": 0.11283333599567413, "learning_rate": 5.081181800775353e-05, "loss": 0.0083, "step": 52010 }, { "grad_norm": 0.21806451678276062, "learning_rate": 5.079528544506565e-05, "loss": 0.0077, "step": 52020 }, { "grad_norm": 0.14452597498893738, "learning_rate": 5.077875279540657e-05, "loss": 0.0085, "step": 52030 }, { "grad_norm": 0.20002521574497223, "learning_rate": 5.076222006058424e-05, "loss": 0.0107, "step": 52040 }, { "grad_norm": 0.18235516548156738, "learning_rate": 5.074568724240667e-05, "loss": 0.008, "step": 52050 }, { "grad_norm": 0.19046881794929504, "learning_rate": 5.072915434268187e-05, "loss": 0.0099, "step": 52060 }, { "grad_norm": 0.13894695043563843, "learning_rate": 5.071262136321785e-05, "loss": 0.0081, "step": 52070 }, { "grad_norm": 0.13900096714496613, "learning_rate": 5.069608830582262e-05, "loss": 0.0067, "step": 52080 }, { "grad_norm": 0.1573532223701477, "learning_rate": 5.067955517230423e-05, "loss": 0.0101, "step": 52090 }, { "grad_norm": 0.1724412590265274, "learning_rate": 5.06630219644707e-05, "loss": 0.0077, "step": 52100 }, { "grad_norm": 0.1389234960079193, "learning_rate": 5.064648868413008e-05, "loss": 0.0083, "step": 52110 }, { "grad_norm": 0.15547886490821838, "learning_rate": 5.062995533309044e-05, "loss": 0.0087, "step": 52120 }, { "grad_norm": 0.15465764701366425, "learning_rate": 5.061342191315983e-05, "loss": 0.0085, "step": 52130 }, { "grad_norm": 0.09381694346666336, "learning_rate": 5.0596888426146325e-05, "loss": 0.0072, "step": 52140 }, { "grad_norm": 0.18659253418445587, "learning_rate": 5.058035487385798e-05, "loss": 0.0093, "step": 52150 }, { "grad_norm": 0.14150919020175934, "learning_rate": 5.0563821258102915e-05, "loss": 0.008, "step": 52160 }, { "grad_norm": 0.11624336987733841, "learning_rate": 5.05472875806892e-05, "loss": 0.0081, "step": 52170 }, { "grad_norm": 0.25709229707717896, "learning_rate": 5.053075384342494e-05, "loss": 0.008, "step": 52180 }, { "grad_norm": 0.16362889111042023, "learning_rate": 5.0514220048118234e-05, "loss": 0.0083, "step": 52190 }, { "grad_norm": 0.16252213716506958, "learning_rate": 5.049768619657721e-05, "loss": 0.0094, "step": 52200 }, { "grad_norm": 0.16832564771175385, "learning_rate": 5.048115229060996e-05, "loss": 0.0089, "step": 52210 }, { "grad_norm": 0.14234331250190735, "learning_rate": 5.046461833202462e-05, "loss": 0.0092, "step": 52220 }, { "grad_norm": 0.2310757040977478, "learning_rate": 5.044808432262932e-05, "loss": 0.0108, "step": 52230 }, { "grad_norm": 0.1045774444937706, "learning_rate": 5.043155026423219e-05, "loss": 0.0078, "step": 52240 }, { "grad_norm": 0.15250948071479797, "learning_rate": 5.041501615864137e-05, "loss": 0.0075, "step": 52250 }, { "grad_norm": 0.1356799453496933, "learning_rate": 5.039848200766502e-05, "loss": 0.0084, "step": 52260 }, { "grad_norm": 0.17409475147724152, "learning_rate": 5.038194781311125e-05, "loss": 0.0082, "step": 52270 }, { "grad_norm": 0.11965559422969818, "learning_rate": 5.0365413576788264e-05, "loss": 0.0078, "step": 52280 }, { "grad_norm": 0.24430164694786072, "learning_rate": 5.0348879300504194e-05, "loss": 0.0086, "step": 52290 }, { "grad_norm": 0.1368962675333023, "learning_rate": 5.033234498606721e-05, "loss": 0.0092, "step": 52300 }, { "grad_norm": 0.41010549664497375, "learning_rate": 5.0315810635285476e-05, "loss": 0.0085, "step": 52310 }, { "grad_norm": 0.19153136014938354, "learning_rate": 5.0299276249967164e-05, "loss": 0.0102, "step": 52320 }, { "grad_norm": 0.13893823325634003, "learning_rate": 5.0282741831920454e-05, "loss": 0.0113, "step": 52330 }, { "grad_norm": 0.1678752601146698, "learning_rate": 5.026620738295351e-05, "loss": 0.0082, "step": 52340 }, { "grad_norm": 0.1708896905183792, "learning_rate": 5.024967290487455e-05, "loss": 0.0087, "step": 52350 }, { "grad_norm": 0.1347200870513916, "learning_rate": 5.023313839949172e-05, "loss": 0.0086, "step": 52360 }, { "grad_norm": 0.19767335057258606, "learning_rate": 5.0216603868613224e-05, "loss": 0.0088, "step": 52370 }, { "grad_norm": 0.19052450358867645, "learning_rate": 5.020006931404726e-05, "loss": 0.0088, "step": 52380 }, { "grad_norm": 0.1384531706571579, "learning_rate": 5.0183534737602014e-05, "loss": 0.0088, "step": 52390 }, { "grad_norm": 0.17865099012851715, "learning_rate": 5.0167000141085686e-05, "loss": 0.0106, "step": 52400 }, { "grad_norm": 0.16056212782859802, "learning_rate": 5.015046552630647e-05, "loss": 0.0091, "step": 52410 }, { "grad_norm": 0.15726710855960846, "learning_rate": 5.013393089507258e-05, "loss": 0.011, "step": 52420 }, { "grad_norm": 0.29911208152770996, "learning_rate": 5.011739624919219e-05, "loss": 0.0098, "step": 52430 }, { "grad_norm": 0.16442734003067017, "learning_rate": 5.0100861590473545e-05, "loss": 0.0094, "step": 52440 }, { "grad_norm": 0.18164688348770142, "learning_rate": 5.008432692072482e-05, "loss": 0.0117, "step": 52450 }, { "grad_norm": 0.13638857007026672, "learning_rate": 5.006779224175423e-05, "loss": 0.009, "step": 52460 }, { "grad_norm": 0.15979282557964325, "learning_rate": 5.005125755536998e-05, "loss": 0.0087, "step": 52470 }, { "grad_norm": 0.1170167401432991, "learning_rate": 5.003472286338029e-05, "loss": 0.0074, "step": 52480 }, { "grad_norm": 0.15148644149303436, "learning_rate": 5.001818816759335e-05, "loss": 0.0093, "step": 52490 }, { "grad_norm": 0.1973772943019867, "learning_rate": 5.000165346981738e-05, "loss": 0.0093, "step": 52500 }, { "grad_norm": 0.15143896639347076, "learning_rate": 4.9985118771860595e-05, "loss": 0.0075, "step": 52510 }, { "grad_norm": 0.13253092765808105, "learning_rate": 4.996858407553119e-05, "loss": 0.0095, "step": 52520 }, { "grad_norm": 0.1372934728860855, "learning_rate": 4.99520493826374e-05, "loss": 0.0089, "step": 52530 }, { "grad_norm": 0.19879873096942902, "learning_rate": 4.9935514694987384e-05, "loss": 0.0093, "step": 52540 }, { "grad_norm": 0.20385469496250153, "learning_rate": 4.991898001438941e-05, "loss": 0.0078, "step": 52550 }, { "grad_norm": 0.1614236682653427, "learning_rate": 4.990244534265166e-05, "loss": 0.0087, "step": 52560 }, { "grad_norm": 0.20913970470428467, "learning_rate": 4.988591068158234e-05, "loss": 0.0098, "step": 52570 }, { "grad_norm": 0.14495909214019775, "learning_rate": 4.986937603298965e-05, "loss": 0.0094, "step": 52580 }, { "grad_norm": 0.21755138039588928, "learning_rate": 4.985284139868181e-05, "loss": 0.0102, "step": 52590 }, { "grad_norm": 0.16169272363185883, "learning_rate": 4.9836306780467e-05, "loss": 0.0102, "step": 52600 }, { "grad_norm": 0.20668233931064606, "learning_rate": 4.981977218015343e-05, "loss": 0.0089, "step": 52610 }, { "grad_norm": 0.20729957520961761, "learning_rate": 4.980323759954932e-05, "loss": 0.0107, "step": 52620 }, { "grad_norm": 0.32696908712387085, "learning_rate": 4.978670304046284e-05, "loss": 0.0089, "step": 52630 }, { "grad_norm": 0.2567012906074524, "learning_rate": 4.9770168504702195e-05, "loss": 0.0105, "step": 52640 }, { "grad_norm": 0.15226559340953827, "learning_rate": 4.975363399407559e-05, "loss": 0.009, "step": 52650 }, { "grad_norm": 0.19065266847610474, "learning_rate": 4.9737099510391194e-05, "loss": 0.0099, "step": 52660 }, { "grad_norm": 0.4026826322078705, "learning_rate": 4.972056505545719e-05, "loss": 0.0094, "step": 52670 }, { "grad_norm": 0.2514607906341553, "learning_rate": 4.9704030631081775e-05, "loss": 0.0115, "step": 52680 }, { "grad_norm": 0.12183798849582672, "learning_rate": 4.968749623907313e-05, "loss": 0.009, "step": 52690 }, { "grad_norm": 0.1383153349161148, "learning_rate": 4.967096188123943e-05, "loss": 0.0091, "step": 52700 }, { "grad_norm": 0.33266857266426086, "learning_rate": 4.965442755938884e-05, "loss": 0.0087, "step": 52710 }, { "grad_norm": 0.2244693636894226, "learning_rate": 4.963789327532954e-05, "loss": 0.0105, "step": 52720 }, { "grad_norm": 0.1694115698337555, "learning_rate": 4.962135903086968e-05, "loss": 0.0086, "step": 52730 }, { "grad_norm": 0.1642761081457138, "learning_rate": 4.960482482781741e-05, "loss": 0.009, "step": 52740 }, { "grad_norm": 0.15450170636177063, "learning_rate": 4.958829066798092e-05, "loss": 0.0092, "step": 52750 }, { "grad_norm": 0.15499383211135864, "learning_rate": 4.957175655316835e-05, "loss": 0.009, "step": 52760 }, { "grad_norm": 0.2320733666419983, "learning_rate": 4.9555222485187834e-05, "loss": 0.0083, "step": 52770 }, { "grad_norm": 0.15148429572582245, "learning_rate": 4.953868846584752e-05, "loss": 0.0087, "step": 52780 }, { "grad_norm": 0.19953450560569763, "learning_rate": 4.9522154496955544e-05, "loss": 0.009, "step": 52790 }, { "grad_norm": 0.19031411409378052, "learning_rate": 4.950562058032001e-05, "loss": 0.0091, "step": 52800 }, { "grad_norm": 0.22904781997203827, "learning_rate": 4.948908671774909e-05, "loss": 0.0083, "step": 52810 }, { "grad_norm": 0.15580618381500244, "learning_rate": 4.947255291105087e-05, "loss": 0.0105, "step": 52820 }, { "grad_norm": 0.15573358535766602, "learning_rate": 4.945601916203348e-05, "loss": 0.0084, "step": 52830 }, { "grad_norm": 0.17283210158348083, "learning_rate": 4.9439485472505016e-05, "loss": 0.0101, "step": 52840 }, { "grad_norm": 0.1592063456773758, "learning_rate": 4.942295184427356e-05, "loss": 0.01, "step": 52850 }, { "grad_norm": 0.1710454225540161, "learning_rate": 4.9406418279147234e-05, "loss": 0.0087, "step": 52860 }, { "grad_norm": 0.12778286635875702, "learning_rate": 4.9389884778934116e-05, "loss": 0.0081, "step": 52870 }, { "grad_norm": 0.17298327386379242, "learning_rate": 4.937335134544228e-05, "loss": 0.008, "step": 52880 }, { "grad_norm": 0.1861627697944641, "learning_rate": 4.935681798047979e-05, "loss": 0.0085, "step": 52890 }, { "grad_norm": 0.2244822084903717, "learning_rate": 4.934028468585473e-05, "loss": 0.012, "step": 52900 }, { "grad_norm": 0.15072111785411835, "learning_rate": 4.932375146337514e-05, "loss": 0.0084, "step": 52910 }, { "grad_norm": 0.1635695844888687, "learning_rate": 4.930721831484908e-05, "loss": 0.0079, "step": 52920 }, { "grad_norm": 0.1658220887184143, "learning_rate": 4.929068524208456e-05, "loss": 0.0084, "step": 52930 }, { "grad_norm": 0.13789087533950806, "learning_rate": 4.927415224688965e-05, "loss": 0.0113, "step": 52940 }, { "grad_norm": 0.21833351254463196, "learning_rate": 4.9257619331072375e-05, "loss": 0.0087, "step": 52950 }, { "grad_norm": 0.13272550702095032, "learning_rate": 4.924108649644072e-05, "loss": 0.0089, "step": 52960 }, { "grad_norm": 0.26244568824768066, "learning_rate": 4.922455374480271e-05, "loss": 0.0088, "step": 52970 }, { "grad_norm": 0.1488068848848343, "learning_rate": 4.920802107796634e-05, "loss": 0.0079, "step": 52980 }, { "grad_norm": 0.20326367020606995, "learning_rate": 4.9191488497739564e-05, "loss": 0.0086, "step": 52990 }, { "grad_norm": 0.1810016930103302, "learning_rate": 4.917495600593042e-05, "loss": 0.0088, "step": 53000 }, { "grad_norm": 0.13764354586601257, "learning_rate": 4.915842360434684e-05, "loss": 0.0081, "step": 53010 }, { "grad_norm": 0.20331008732318878, "learning_rate": 4.914189129479679e-05, "loss": 0.0085, "step": 53020 }, { "grad_norm": 0.12323523312807083, "learning_rate": 4.912535907908822e-05, "loss": 0.0094, "step": 53030 }, { "grad_norm": 0.232511505484581, "learning_rate": 4.910882695902907e-05, "loss": 0.012, "step": 53040 }, { "grad_norm": 0.16654250025749207, "learning_rate": 4.9092294936427234e-05, "loss": 0.0091, "step": 53050 }, { "grad_norm": 0.1662164181470871, "learning_rate": 4.9075763013090685e-05, "loss": 0.0089, "step": 53060 }, { "grad_norm": 0.2693789005279541, "learning_rate": 4.9059231190827305e-05, "loss": 0.0088, "step": 53070 }, { "grad_norm": 0.12441147118806839, "learning_rate": 4.904269947144499e-05, "loss": 0.0098, "step": 53080 }, { "grad_norm": 0.16037073731422424, "learning_rate": 4.902616785675161e-05, "loss": 0.0078, "step": 53090 }, { "grad_norm": 0.17890866100788116, "learning_rate": 4.900963634855505e-05, "loss": 0.0094, "step": 53100 }, { "grad_norm": 0.16724100708961487, "learning_rate": 4.8993104948663173e-05, "loss": 0.0083, "step": 53110 }, { "grad_norm": 0.29376906156539917, "learning_rate": 4.8976573658883823e-05, "loss": 0.011, "step": 53120 }, { "grad_norm": 0.17769576609134674, "learning_rate": 4.896004248102484e-05, "loss": 0.0095, "step": 53130 }, { "grad_norm": 0.17817498743534088, "learning_rate": 4.8943511416894045e-05, "loss": 0.0093, "step": 53140 }, { "grad_norm": 0.1847953498363495, "learning_rate": 4.892698046829926e-05, "loss": 0.0074, "step": 53150 }, { "grad_norm": 0.18018311262130737, "learning_rate": 4.8910449637048275e-05, "loss": 0.0078, "step": 53160 }, { "grad_norm": 0.12237035483121872, "learning_rate": 4.8893918924948884e-05, "loss": 0.0074, "step": 53170 }, { "grad_norm": 0.1779976338148117, "learning_rate": 4.8877388333808835e-05, "loss": 0.0097, "step": 53180 }, { "grad_norm": 0.18020300567150116, "learning_rate": 4.8860857865435925e-05, "loss": 0.0081, "step": 53190 }, { "grad_norm": 0.1585456281900406, "learning_rate": 4.884432752163789e-05, "loss": 0.0102, "step": 53200 }, { "grad_norm": 0.22590723633766174, "learning_rate": 4.8827797304222466e-05, "loss": 0.0089, "step": 53210 }, { "grad_norm": 0.18789348006248474, "learning_rate": 4.8811267214997366e-05, "loss": 0.0082, "step": 53220 }, { "grad_norm": 0.1786959022283554, "learning_rate": 4.879473725577029e-05, "loss": 0.0087, "step": 53230 }, { "grad_norm": 0.1983676701784134, "learning_rate": 4.877820742834891e-05, "loss": 0.0112, "step": 53240 }, { "grad_norm": 0.14790673553943634, "learning_rate": 4.8761677734540956e-05, "loss": 0.0093, "step": 53250 }, { "grad_norm": 0.11986754089593887, "learning_rate": 4.874514817615407e-05, "loss": 0.0085, "step": 53260 }, { "grad_norm": 0.15527823567390442, "learning_rate": 4.872861875499588e-05, "loss": 0.0104, "step": 53270 }, { "grad_norm": 0.3012607991695404, "learning_rate": 4.871208947287404e-05, "loss": 0.0079, "step": 53280 }, { "grad_norm": 0.22692175209522247, "learning_rate": 4.869556033159615e-05, "loss": 0.0093, "step": 53290 }, { "grad_norm": 0.20287804305553436, "learning_rate": 4.8679031332969835e-05, "loss": 0.0086, "step": 53300 }, { "grad_norm": 0.11869819462299347, "learning_rate": 4.8662502478802635e-05, "loss": 0.0099, "step": 53310 }, { "grad_norm": 0.23776546120643616, "learning_rate": 4.8645973770902176e-05, "loss": 0.0126, "step": 53320 }, { "grad_norm": 0.13575708866119385, "learning_rate": 4.862944521107599e-05, "loss": 0.0088, "step": 53330 }, { "grad_norm": 0.2851475179195404, "learning_rate": 4.861291680113162e-05, "loss": 0.0088, "step": 53340 }, { "grad_norm": 0.17869630455970764, "learning_rate": 4.8596388542876563e-05, "loss": 0.0081, "step": 53350 }, { "grad_norm": 0.13135212659835815, "learning_rate": 4.857986043811836e-05, "loss": 0.0099, "step": 53360 }, { "grad_norm": 0.13530391454696655, "learning_rate": 4.856333248866448e-05, "loss": 0.0095, "step": 53370 }, { "grad_norm": 0.19660434126853943, "learning_rate": 4.85468046963224e-05, "loss": 0.0087, "step": 53380 }, { "grad_norm": 0.1338316947221756, "learning_rate": 4.853027706289957e-05, "loss": 0.0084, "step": 53390 }, { "grad_norm": 0.11812184005975723, "learning_rate": 4.851374959020344e-05, "loss": 0.0092, "step": 53400 }, { "grad_norm": 0.18571124970912933, "learning_rate": 4.849722228004141e-05, "loss": 0.0086, "step": 53410 }, { "grad_norm": 0.1696515679359436, "learning_rate": 4.848069513422091e-05, "loss": 0.0091, "step": 53420 }, { "grad_norm": 0.13650213181972504, "learning_rate": 4.8464168154549294e-05, "loss": 0.0074, "step": 53430 }, { "grad_norm": 0.25238505005836487, "learning_rate": 4.844764134283392e-05, "loss": 0.0097, "step": 53440 }, { "grad_norm": 0.13456900417804718, "learning_rate": 4.843111470088218e-05, "loss": 0.0089, "step": 53450 }, { "grad_norm": 0.1402331292629242, "learning_rate": 4.841458823050137e-05, "loss": 0.0077, "step": 53460 }, { "grad_norm": 0.2962239682674408, "learning_rate": 4.8398061933498816e-05, "loss": 0.0096, "step": 53470 }, { "grad_norm": 0.24709591269493103, "learning_rate": 4.83815358116818e-05, "loss": 0.0091, "step": 53480 }, { "grad_norm": 0.12256748974323273, "learning_rate": 4.8365009866857584e-05, "loss": 0.0076, "step": 53490 }, { "grad_norm": 0.16024094820022583, "learning_rate": 4.834848410083342e-05, "loss": 0.0085, "step": 53500 }, { "grad_norm": 0.16842831671237946, "learning_rate": 4.833195851541657e-05, "loss": 0.0088, "step": 53510 }, { "grad_norm": 0.2515808641910553, "learning_rate": 4.831543311241423e-05, "loss": 0.0098, "step": 53520 }, { "grad_norm": 0.16811171174049377, "learning_rate": 4.829890789363359e-05, "loss": 0.0078, "step": 53530 }, { "grad_norm": 0.13540184497833252, "learning_rate": 4.828238286088182e-05, "loss": 0.0091, "step": 53540 }, { "grad_norm": 0.16871394217014313, "learning_rate": 4.826585801596609e-05, "loss": 0.0088, "step": 53550 }, { "grad_norm": 0.13632647693157196, "learning_rate": 4.8249333360693486e-05, "loss": 0.0094, "step": 53560 }, { "grad_norm": 0.31477200984954834, "learning_rate": 4.8232808896871185e-05, "loss": 0.0088, "step": 53570 }, { "grad_norm": 0.11860362440347672, "learning_rate": 4.821628462630624e-05, "loss": 0.0096, "step": 53580 }, { "grad_norm": 0.12810611724853516, "learning_rate": 4.8199760550805735e-05, "loss": 0.0088, "step": 53590 }, { "grad_norm": 0.16974395513534546, "learning_rate": 4.818323667217669e-05, "loss": 0.0087, "step": 53600 }, { "grad_norm": 0.12419985979795456, "learning_rate": 4.8166712992226165e-05, "loss": 0.0099, "step": 53610 }, { "grad_norm": 0.20124951004981995, "learning_rate": 4.8150189512761154e-05, "loss": 0.0088, "step": 53620 }, { "grad_norm": 0.17647665739059448, "learning_rate": 4.813366623558862e-05, "loss": 0.0101, "step": 53630 }, { "grad_norm": 0.16070665419101715, "learning_rate": 4.811714316251554e-05, "loss": 0.0086, "step": 53640 }, { "grad_norm": 0.2038649171590805, "learning_rate": 4.810062029534886e-05, "loss": 0.0088, "step": 53650 }, { "grad_norm": 0.12932433187961578, "learning_rate": 4.8084097635895505e-05, "loss": 0.0096, "step": 53660 }, { "grad_norm": 0.14816199243068695, "learning_rate": 4.8067575185962335e-05, "loss": 0.0084, "step": 53670 }, { "grad_norm": 0.15524521470069885, "learning_rate": 4.805105294735625e-05, "loss": 0.0117, "step": 53680 }, { "grad_norm": 0.18085989356040955, "learning_rate": 4.803453092188406e-05, "loss": 0.0096, "step": 53690 }, { "grad_norm": 0.21801862120628357, "learning_rate": 4.801800911135263e-05, "loss": 0.009, "step": 53700 }, { "grad_norm": 0.12703199684619904, "learning_rate": 4.8001487517568744e-05, "loss": 0.0096, "step": 53710 }, { "grad_norm": 0.1702260673046112, "learning_rate": 4.798496614233918e-05, "loss": 0.0085, "step": 53720 }, { "grad_norm": 0.23268768191337585, "learning_rate": 4.796844498747069e-05, "loss": 0.0098, "step": 53730 }, { "grad_norm": 0.2247762680053711, "learning_rate": 4.795192405477e-05, "loss": 0.0092, "step": 53740 }, { "grad_norm": 0.11633018404245377, "learning_rate": 4.79354033460438e-05, "loss": 0.0086, "step": 53750 }, { "grad_norm": 0.16974525153636932, "learning_rate": 4.79188828630988e-05, "loss": 0.0097, "step": 53760 }, { "grad_norm": 0.16868728399276733, "learning_rate": 4.790236260774165e-05, "loss": 0.0113, "step": 53770 }, { "grad_norm": 0.15652304887771606, "learning_rate": 4.788584258177896e-05, "loss": 0.0075, "step": 53780 }, { "grad_norm": 0.1866999715566635, "learning_rate": 4.7869322787017356e-05, "loss": 0.0079, "step": 53790 }, { "grad_norm": 0.13142754137516022, "learning_rate": 4.78528032252634e-05, "loss": 0.0091, "step": 53800 }, { "grad_norm": 0.17326462268829346, "learning_rate": 4.783628389832366e-05, "loss": 0.0098, "step": 53810 }, { "grad_norm": 0.20098961889743805, "learning_rate": 4.7819764808004644e-05, "loss": 0.0083, "step": 53820 }, { "grad_norm": 0.1713862419128418, "learning_rate": 4.780324595611289e-05, "loss": 0.0087, "step": 53830 }, { "grad_norm": 0.15205228328704834, "learning_rate": 4.7786727344454854e-05, "loss": 0.0086, "step": 53840 }, { "grad_norm": 0.20275281369686127, "learning_rate": 4.777020897483697e-05, "loss": 0.009, "step": 53850 }, { "grad_norm": 0.1137469932436943, "learning_rate": 4.775369084906569e-05, "loss": 0.0079, "step": 53860 }, { "grad_norm": 0.35028424859046936, "learning_rate": 4.773717296894742e-05, "loss": 0.0107, "step": 53870 }, { "grad_norm": 0.17574739456176758, "learning_rate": 4.772065533628848e-05, "loss": 0.0087, "step": 53880 }, { "grad_norm": 0.1486491709947586, "learning_rate": 4.7704137952895265e-05, "loss": 0.0084, "step": 53890 }, { "grad_norm": 0.1963261216878891, "learning_rate": 4.768762082057407e-05, "loss": 0.0095, "step": 53900 }, { "grad_norm": 0.17943400144577026, "learning_rate": 4.7671103941131186e-05, "loss": 0.0089, "step": 53910 }, { "grad_norm": 0.10538921505212784, "learning_rate": 4.765458731637287e-05, "loss": 0.0102, "step": 53920 }, { "grad_norm": 0.20966894924640656, "learning_rate": 4.763807094810537e-05, "loss": 0.0094, "step": 53930 }, { "grad_norm": 0.16842733323574066, "learning_rate": 4.762155483813487e-05, "loss": 0.0089, "step": 53940 }, { "grad_norm": 0.24014413356781006, "learning_rate": 4.760503898826754e-05, "loss": 0.0087, "step": 53950 }, { "grad_norm": 0.1507401317358017, "learning_rate": 4.758852340030957e-05, "loss": 0.0091, "step": 53960 }, { "grad_norm": 0.1787853240966797, "learning_rate": 4.757200807606705e-05, "loss": 0.0119, "step": 53970 }, { "grad_norm": 0.18213418126106262, "learning_rate": 4.755549301734607e-05, "loss": 0.0099, "step": 53980 }, { "grad_norm": 0.22296041250228882, "learning_rate": 4.753897822595271e-05, "loss": 0.0098, "step": 53990 }, { "grad_norm": 0.21111217141151428, "learning_rate": 4.7522463703692994e-05, "loss": 0.0096, "step": 54000 }, { "grad_norm": 0.17935967445373535, "learning_rate": 4.75059494523729e-05, "loss": 0.0092, "step": 54010 }, { "grad_norm": 0.10449506342411041, "learning_rate": 4.748943547379843e-05, "loss": 0.0082, "step": 54020 }, { "grad_norm": 0.16004833579063416, "learning_rate": 4.7472921769775535e-05, "loss": 0.0099, "step": 54030 }, { "grad_norm": 0.26319706439971924, "learning_rate": 4.745640834211011e-05, "loss": 0.009, "step": 54040 }, { "grad_norm": 0.14265525341033936, "learning_rate": 4.7439895192608054e-05, "loss": 0.0083, "step": 54050 }, { "grad_norm": 0.31049710512161255, "learning_rate": 4.7423382323075186e-05, "loss": 0.0083, "step": 54060 }, { "grad_norm": 0.2302018105983734, "learning_rate": 4.740686973531737e-05, "loss": 0.0089, "step": 54070 }, { "grad_norm": 0.22100859880447388, "learning_rate": 4.7390357431140375e-05, "loss": 0.0081, "step": 54080 }, { "grad_norm": 0.1817288100719452, "learning_rate": 4.737384541234998e-05, "loss": 0.0106, "step": 54090 }, { "grad_norm": 0.21064671874046326, "learning_rate": 4.735733368075189e-05, "loss": 0.0086, "step": 54100 }, { "grad_norm": 0.24106426537036896, "learning_rate": 4.734082223815182e-05, "loss": 0.0101, "step": 54110 }, { "grad_norm": 0.13431225717067719, "learning_rate": 4.7324311086355436e-05, "loss": 0.01, "step": 54120 }, { "grad_norm": 0.10219278186559677, "learning_rate": 4.730780022716836e-05, "loss": 0.0099, "step": 54130 }, { "grad_norm": 0.18509499728679657, "learning_rate": 4.729128966239619e-05, "loss": 0.0092, "step": 54140 }, { "grad_norm": 0.15031175315380096, "learning_rate": 4.7274779393844536e-05, "loss": 0.0095, "step": 54150 }, { "grad_norm": 0.13997869193553925, "learning_rate": 4.725826942331891e-05, "loss": 0.0088, "step": 54160 }, { "grad_norm": 0.19064971804618835, "learning_rate": 4.724175975262481e-05, "loss": 0.0092, "step": 54170 }, { "grad_norm": 0.15610262751579285, "learning_rate": 4.722525038356772e-05, "loss": 0.0075, "step": 54180 }, { "grad_norm": 0.21468190848827362, "learning_rate": 4.7208741317953075e-05, "loss": 0.0072, "step": 54190 }, { "grad_norm": 0.20116817951202393, "learning_rate": 4.7192232557586264e-05, "loss": 0.0096, "step": 54200 }, { "grad_norm": 0.12143872678279877, "learning_rate": 4.717572410427269e-05, "loss": 0.0081, "step": 54210 }, { "grad_norm": 0.17294810712337494, "learning_rate": 4.71592159598177e-05, "loss": 0.0086, "step": 54220 }, { "grad_norm": 0.1483832597732544, "learning_rate": 4.714270812602657e-05, "loss": 0.0079, "step": 54230 }, { "grad_norm": 0.24716489017009735, "learning_rate": 4.712620060470458e-05, "loss": 0.01, "step": 54240 }, { "grad_norm": 0.11340416967868805, "learning_rate": 4.7109693397656986e-05, "loss": 0.0094, "step": 54250 }, { "grad_norm": 0.18733160197734833, "learning_rate": 4.709318650668894e-05, "loss": 0.0083, "step": 54260 }, { "grad_norm": 0.16610415279865265, "learning_rate": 4.7076679933605676e-05, "loss": 0.0086, "step": 54270 }, { "grad_norm": 0.1778581291437149, "learning_rate": 4.70601736802123e-05, "loss": 0.0084, "step": 54280 }, { "grad_norm": 0.1705697476863861, "learning_rate": 4.704366774831391e-05, "loss": 0.0085, "step": 54290 }, { "grad_norm": 0.12739664316177368, "learning_rate": 4.702716213971557e-05, "loss": 0.0096, "step": 54300 }, { "grad_norm": 0.1356143057346344, "learning_rate": 4.70106568562223e-05, "loss": 0.0074, "step": 54310 }, { "grad_norm": 0.17759950459003448, "learning_rate": 4.6994151899639105e-05, "loss": 0.0096, "step": 54320 }, { "grad_norm": 0.13086551427841187, "learning_rate": 4.697764727177093e-05, "loss": 0.0082, "step": 54330 }, { "grad_norm": 0.17798614501953125, "learning_rate": 4.696114297442272e-05, "loss": 0.0086, "step": 54340 }, { "grad_norm": 0.13853241503238678, "learning_rate": 4.694463900939933e-05, "loss": 0.0107, "step": 54350 }, { "grad_norm": 0.1464114487171173, "learning_rate": 4.692813537850564e-05, "loss": 0.0085, "step": 54360 }, { "grad_norm": 0.1994583010673523, "learning_rate": 4.6911632083546445e-05, "loss": 0.0086, "step": 54370 }, { "grad_norm": 0.13620293140411377, "learning_rate": 4.6895129126326525e-05, "loss": 0.0078, "step": 54380 }, { "grad_norm": 0.1352306604385376, "learning_rate": 4.687862650865059e-05, "loss": 0.0088, "step": 54390 }, { "grad_norm": 0.12983672320842743, "learning_rate": 4.686212423232339e-05, "loss": 0.0099, "step": 54400 }, { "grad_norm": 0.15040917694568634, "learning_rate": 4.6845622299149565e-05, "loss": 0.0086, "step": 54410 }, { "grad_norm": 0.154151052236557, "learning_rate": 4.682912071093374e-05, "loss": 0.0097, "step": 54420 }, { "grad_norm": 0.1688084900379181, "learning_rate": 4.681261946948052e-05, "loss": 0.0077, "step": 54430 }, { "grad_norm": 0.2670186460018158, "learning_rate": 4.679611857659443e-05, "loss": 0.0093, "step": 54440 }, { "grad_norm": 0.15217429399490356, "learning_rate": 4.6779618034079975e-05, "loss": 0.0087, "step": 54450 }, { "grad_norm": 0.18476629257202148, "learning_rate": 4.6763117843741675e-05, "loss": 0.0089, "step": 54460 }, { "grad_norm": 0.11590804904699326, "learning_rate": 4.674661800738394e-05, "loss": 0.0093, "step": 54470 }, { "grad_norm": 0.1584976762533188, "learning_rate": 4.673011852681117e-05, "loss": 0.0098, "step": 54480 }, { "grad_norm": 0.27240586280822754, "learning_rate": 4.671361940382772e-05, "loss": 0.0082, "step": 54490 }, { "grad_norm": 0.27885913848876953, "learning_rate": 4.669712064023791e-05, "loss": 0.0083, "step": 54500 }, { "grad_norm": 0.14133408665657043, "learning_rate": 4.668062223784602e-05, "loss": 0.0092, "step": 54510 }, { "grad_norm": 0.11828310042619705, "learning_rate": 4.6664124198456274e-05, "loss": 0.0074, "step": 54520 }, { "grad_norm": 0.16005831956863403, "learning_rate": 4.6647626523872905e-05, "loss": 0.0068, "step": 54530 }, { "grad_norm": 0.18403260409832, "learning_rate": 4.6631129215900066e-05, "loss": 0.0074, "step": 54540 }, { "grad_norm": 0.15135140717029572, "learning_rate": 4.661463227634187e-05, "loss": 0.0093, "step": 54550 }, { "grad_norm": 0.1843244582414627, "learning_rate": 4.659813570700238e-05, "loss": 0.0088, "step": 54560 }, { "grad_norm": 0.17614467442035675, "learning_rate": 4.658163950968568e-05, "loss": 0.0085, "step": 54570 }, { "grad_norm": 0.12030186504125595, "learning_rate": 4.6565143686195726e-05, "loss": 0.0085, "step": 54580 }, { "grad_norm": 0.11672601103782654, "learning_rate": 4.654864823833649e-05, "loss": 0.008, "step": 54590 }, { "grad_norm": 0.13651473820209503, "learning_rate": 4.65321531679119e-05, "loss": 0.0101, "step": 54600 }, { "grad_norm": 0.24731594324111938, "learning_rate": 4.6515658476725834e-05, "loss": 0.0083, "step": 54610 }, { "grad_norm": 0.16841964423656464, "learning_rate": 4.649916416658212e-05, "loss": 0.0109, "step": 54620 }, { "grad_norm": 0.1912013590335846, "learning_rate": 4.648267023928455e-05, "loss": 0.0077, "step": 54630 }, { "grad_norm": 0.12748302519321442, "learning_rate": 4.646617669663687e-05, "loss": 0.0101, "step": 54640 }, { "grad_norm": 0.5616087317466736, "learning_rate": 4.644968354044277e-05, "loss": 0.0086, "step": 54650 }, { "grad_norm": 0.1251097023487091, "learning_rate": 4.643319077250596e-05, "loss": 0.0086, "step": 54660 }, { "grad_norm": 0.1601683348417282, "learning_rate": 4.6416698394630055e-05, "loss": 0.008, "step": 54670 }, { "grad_norm": 0.2495725154876709, "learning_rate": 4.640020640861862e-05, "loss": 0.009, "step": 54680 }, { "grad_norm": 0.1706072837114334, "learning_rate": 4.6383714816275195e-05, "loss": 0.0076, "step": 54690 }, { "grad_norm": 0.1440340131521225, "learning_rate": 4.636722361940329e-05, "loss": 0.0085, "step": 54700 }, { "grad_norm": 0.1278286874294281, "learning_rate": 4.635073281980632e-05, "loss": 0.0086, "step": 54710 }, { "grad_norm": 0.12891201674938202, "learning_rate": 4.633424241928775e-05, "loss": 0.0071, "step": 54720 }, { "grad_norm": 0.1909947693347931, "learning_rate": 4.6317752419650914e-05, "loss": 0.0108, "step": 54730 }, { "grad_norm": 0.2298499494791031, "learning_rate": 4.630126282269914e-05, "loss": 0.0083, "step": 54740 }, { "grad_norm": 0.1398344337940216, "learning_rate": 4.628477363023569e-05, "loss": 0.0098, "step": 54750 }, { "grad_norm": 0.20970278978347778, "learning_rate": 4.626828484406383e-05, "loss": 0.0094, "step": 54760 }, { "grad_norm": 0.14203596115112305, "learning_rate": 4.625179646598669e-05, "loss": 0.0093, "step": 54770 }, { "grad_norm": 0.173088937997818, "learning_rate": 4.6235308497807487e-05, "loss": 0.0092, "step": 54780 }, { "grad_norm": 0.14675229787826538, "learning_rate": 4.6218820941329276e-05, "loss": 0.0075, "step": 54790 }, { "grad_norm": 0.12326502799987793, "learning_rate": 4.620233379835513e-05, "loss": 0.0086, "step": 54800 }, { "grad_norm": 0.15275003015995026, "learning_rate": 4.618584707068804e-05, "loss": 0.0087, "step": 54810 }, { "grad_norm": 0.15713079273700714, "learning_rate": 4.616936076013099e-05, "loss": 0.008, "step": 54820 }, { "grad_norm": 0.20141881704330444, "learning_rate": 4.615287486848688e-05, "loss": 0.0088, "step": 54830 }, { "grad_norm": 0.2768256664276123, "learning_rate": 4.6136389397558584e-05, "loss": 0.0094, "step": 54840 }, { "grad_norm": 0.17084597051143646, "learning_rate": 4.6119904349148935e-05, "loss": 0.0075, "step": 54850 }, { "grad_norm": 0.31372907757759094, "learning_rate": 4.6103419725060726e-05, "loss": 0.0082, "step": 54860 }, { "grad_norm": 0.22753684222698212, "learning_rate": 4.608693552709667e-05, "loss": 0.0095, "step": 54870 }, { "grad_norm": 0.12868352234363556, "learning_rate": 4.607045175705947e-05, "loss": 0.009, "step": 54880 }, { "grad_norm": 0.15421196818351746, "learning_rate": 4.605396841675176e-05, "loss": 0.009, "step": 54890 }, { "grad_norm": 0.17778819799423218, "learning_rate": 4.6037485507976094e-05, "loss": 0.0097, "step": 54900 }, { "grad_norm": 0.15538373589515686, "learning_rate": 4.602100303253509e-05, "loss": 0.0082, "step": 54910 }, { "grad_norm": 0.13563016057014465, "learning_rate": 4.600452099223122e-05, "loss": 0.0096, "step": 54920 }, { "grad_norm": 0.13881371915340424, "learning_rate": 4.598803938886691e-05, "loss": 0.0086, "step": 54930 }, { "grad_norm": 0.16850972175598145, "learning_rate": 4.597155822424459e-05, "loss": 0.0079, "step": 54940 }, { "grad_norm": 0.12973526120185852, "learning_rate": 4.59550775001666e-05, "loss": 0.01, "step": 54950 }, { "grad_norm": 0.1776888519525528, "learning_rate": 4.5938597218435234e-05, "loss": 0.0085, "step": 54960 }, { "grad_norm": 0.14212769269943237, "learning_rate": 4.592211738085279e-05, "loss": 0.008, "step": 54970 }, { "grad_norm": 0.25513532757759094, "learning_rate": 4.590563798922146e-05, "loss": 0.0081, "step": 54980 }, { "grad_norm": 0.15110564231872559, "learning_rate": 4.5889159045343404e-05, "loss": 0.0074, "step": 54990 }, { "grad_norm": 0.2417529970407486, "learning_rate": 4.5872680551020734e-05, "loss": 0.01, "step": 55000 }, { "grad_norm": 0.16515786945819855, "learning_rate": 4.585620250805552e-05, "loss": 0.0084, "step": 55010 }, { "grad_norm": 0.17274045944213867, "learning_rate": 4.583972491824975e-05, "loss": 0.0097, "step": 55020 }, { "grad_norm": 0.13862760365009308, "learning_rate": 4.582324778340541e-05, "loss": 0.008, "step": 55030 }, { "grad_norm": 0.1174725815653801, "learning_rate": 4.5806771105324424e-05, "loss": 0.0093, "step": 55040 }, { "grad_norm": 0.1338774412870407, "learning_rate": 4.579029488580865e-05, "loss": 0.007, "step": 55050 }, { "grad_norm": 0.12256554514169693, "learning_rate": 4.5773819126659876e-05, "loss": 0.0078, "step": 55060 }, { "grad_norm": 0.14277756214141846, "learning_rate": 4.575734382967991e-05, "loss": 0.0083, "step": 55070 }, { "grad_norm": 0.1369277387857437, "learning_rate": 4.574086899667044e-05, "loss": 0.0088, "step": 55080 }, { "grad_norm": 0.16503387689590454, "learning_rate": 4.5724394629433106e-05, "loss": 0.0099, "step": 55090 }, { "grad_norm": 0.25321757793426514, "learning_rate": 4.570792072976958e-05, "loss": 0.0081, "step": 55100 }, { "grad_norm": 0.13736340403556824, "learning_rate": 4.569144729948138e-05, "loss": 0.008, "step": 55110 }, { "grad_norm": 0.14998988807201385, "learning_rate": 4.5674974340370033e-05, "loss": 0.0097, "step": 55120 }, { "grad_norm": 0.11272420734167099, "learning_rate": 4.565850185423699e-05, "loss": 0.0073, "step": 55130 }, { "grad_norm": 0.15456894040107727, "learning_rate": 4.564202984288365e-05, "loss": 0.0083, "step": 55140 }, { "grad_norm": 0.1453762948513031, "learning_rate": 4.5625558308111356e-05, "loss": 0.0091, "step": 55150 }, { "grad_norm": 0.14382100105285645, "learning_rate": 4.560908725172145e-05, "loss": 0.0079, "step": 55160 }, { "grad_norm": 0.12553070485591888, "learning_rate": 4.559261667551516e-05, "loss": 0.0088, "step": 55170 }, { "grad_norm": 0.13643813133239746, "learning_rate": 4.5576146581293685e-05, "loss": 0.0082, "step": 55180 }, { "grad_norm": 0.17181140184402466, "learning_rate": 4.5559676970858164e-05, "loss": 0.0105, "step": 55190 }, { "grad_norm": 0.1719026118516922, "learning_rate": 4.5543207846009705e-05, "loss": 0.0091, "step": 55200 }, { "grad_norm": 0.17582833766937256, "learning_rate": 4.552673920854933e-05, "loss": 0.0072, "step": 55210 }, { "grad_norm": 0.21098162233829498, "learning_rate": 4.5510271060278006e-05, "loss": 0.0086, "step": 55220 }, { "grad_norm": 0.15352359414100647, "learning_rate": 4.5493803402996704e-05, "loss": 0.0081, "step": 55230 }, { "grad_norm": 0.17276673018932343, "learning_rate": 4.5477336238506295e-05, "loss": 0.0076, "step": 55240 }, { "grad_norm": 0.1603473722934723, "learning_rate": 4.546086956860759e-05, "loss": 0.0088, "step": 55250 }, { "grad_norm": 0.13799802958965302, "learning_rate": 4.5444403395101374e-05, "loss": 0.0077, "step": 55260 }, { "grad_norm": 0.16495047509670258, "learning_rate": 4.542793771978834e-05, "loss": 0.0076, "step": 55270 }, { "grad_norm": 0.17258118093013763, "learning_rate": 4.541147254446916e-05, "loss": 0.0098, "step": 55280 }, { "grad_norm": 0.18347281217575073, "learning_rate": 4.539500787094445e-05, "loss": 0.0101, "step": 55290 }, { "grad_norm": 0.3514164388179779, "learning_rate": 4.537854370101475e-05, "loss": 0.0098, "step": 55300 }, { "grad_norm": 0.19276821613311768, "learning_rate": 4.5362080036480566e-05, "loss": 0.0115, "step": 55310 }, { "grad_norm": 0.14719712734222412, "learning_rate": 4.534561687914235e-05, "loss": 0.0098, "step": 55320 }, { "grad_norm": 0.2715766429901123, "learning_rate": 4.5329154230800466e-05, "loss": 0.0092, "step": 55330 }, { "grad_norm": 0.2411908358335495, "learning_rate": 4.531269209325525e-05, "loss": 0.0082, "step": 55340 }, { "grad_norm": 0.1561289131641388, "learning_rate": 4.529623046830696e-05, "loss": 0.0101, "step": 55350 }, { "grad_norm": 0.24576616287231445, "learning_rate": 4.5279769357755866e-05, "loss": 0.0083, "step": 55360 }, { "grad_norm": 0.28019800782203674, "learning_rate": 4.5263308763402084e-05, "loss": 0.0093, "step": 55370 }, { "grad_norm": 0.2181839793920517, "learning_rate": 4.524684868704574e-05, "loss": 0.0089, "step": 55380 }, { "grad_norm": 0.34666621685028076, "learning_rate": 4.523038913048688e-05, "loss": 0.0103, "step": 55390 }, { "grad_norm": 0.1711053103208542, "learning_rate": 4.521393009552549e-05, "loss": 0.0104, "step": 55400 }, { "grad_norm": 0.1525254249572754, "learning_rate": 4.5197471583961484e-05, "loss": 0.01, "step": 55410 }, { "grad_norm": 0.16726355254650116, "learning_rate": 4.51810135975948e-05, "loss": 0.009, "step": 55420 }, { "grad_norm": 0.17095282673835754, "learning_rate": 4.516455613822521e-05, "loss": 0.0077, "step": 55430 }, { "grad_norm": 0.17463701963424683, "learning_rate": 4.514809920765249e-05, "loss": 0.007, "step": 55440 }, { "grad_norm": 0.14510318636894226, "learning_rate": 4.5131642807676356e-05, "loss": 0.0082, "step": 55450 }, { "grad_norm": 0.1476687788963318, "learning_rate": 4.511518694009644e-05, "loss": 0.0087, "step": 55460 }, { "grad_norm": 0.13694027066230774, "learning_rate": 4.509873160671231e-05, "loss": 0.0093, "step": 55470 }, { "grad_norm": 0.19568583369255066, "learning_rate": 4.508227680932356e-05, "loss": 0.0084, "step": 55480 }, { "grad_norm": 0.1452018767595291, "learning_rate": 4.5065822549729605e-05, "loss": 0.0095, "step": 55490 }, { "grad_norm": 0.14904001355171204, "learning_rate": 4.504936882972989e-05, "loss": 0.0097, "step": 55500 }, { "grad_norm": 0.1671815812587738, "learning_rate": 4.503291565112374e-05, "loss": 0.0077, "step": 55510 }, { "grad_norm": 0.19546599686145782, "learning_rate": 4.501646301571048e-05, "loss": 0.0082, "step": 55520 }, { "grad_norm": 0.12580254673957825, "learning_rate": 4.500001092528933e-05, "loss": 0.0077, "step": 55530 }, { "grad_norm": 0.22680509090423584, "learning_rate": 4.4983559381659455e-05, "loss": 0.0086, "step": 55540 }, { "grad_norm": 0.14260873198509216, "learning_rate": 4.496710838661999e-05, "loss": 0.0079, "step": 55550 }, { "grad_norm": 0.1749783456325531, "learning_rate": 4.495065794196999e-05, "loss": 0.0077, "step": 55560 }, { "grad_norm": 0.18556316196918488, "learning_rate": 4.4934208049508455e-05, "loss": 0.0105, "step": 55570 }, { "grad_norm": 0.18993733823299408, "learning_rate": 4.4917758711034303e-05, "loss": 0.0094, "step": 55580 }, { "grad_norm": 0.1284426748752594, "learning_rate": 4.490130992834642e-05, "loss": 0.0091, "step": 55590 }, { "grad_norm": 0.12868952751159668, "learning_rate": 4.488486170324359e-05, "loss": 0.0084, "step": 55600 }, { "grad_norm": 0.1855451762676239, "learning_rate": 4.486841403752463e-05, "loss": 0.0127, "step": 55610 }, { "grad_norm": 0.17477844655513763, "learning_rate": 4.4851966932988174e-05, "loss": 0.0102, "step": 55620 }, { "grad_norm": 0.14436013996601105, "learning_rate": 4.4835520391432894e-05, "loss": 0.0095, "step": 55630 }, { "grad_norm": 0.14217421412467957, "learning_rate": 4.481907441465733e-05, "loss": 0.0078, "step": 55640 }, { "grad_norm": 0.1545625627040863, "learning_rate": 4.480262900446e-05, "loss": 0.0084, "step": 55650 }, { "grad_norm": 0.17048582434654236, "learning_rate": 4.478618416263933e-05, "loss": 0.0088, "step": 55660 }, { "grad_norm": 0.10758195072412491, "learning_rate": 4.476973989099373e-05, "loss": 0.0091, "step": 55670 }, { "grad_norm": 0.4441896677017212, "learning_rate": 4.475329619132153e-05, "loss": 0.0093, "step": 55680 }, { "grad_norm": 0.12756791710853577, "learning_rate": 4.4736853065420956e-05, "loss": 0.0081, "step": 55690 }, { "grad_norm": 0.1551753431558609, "learning_rate": 4.472041051509023e-05, "loss": 0.0082, "step": 55700 }, { "grad_norm": 0.13489529490470886, "learning_rate": 4.470396854212746e-05, "loss": 0.0077, "step": 55710 }, { "grad_norm": 0.20059828460216522, "learning_rate": 4.468752714833072e-05, "loss": 0.0077, "step": 55720 }, { "grad_norm": 0.11506374925374985, "learning_rate": 4.467108633549803e-05, "loss": 0.0105, "step": 55730 }, { "grad_norm": 0.15897904336452484, "learning_rate": 4.4654646105427335e-05, "loss": 0.0091, "step": 55740 }, { "grad_norm": 0.188357412815094, "learning_rate": 4.463820645991651e-05, "loss": 0.0082, "step": 55750 }, { "grad_norm": 0.14439913630485535, "learning_rate": 4.462176740076334e-05, "loss": 0.0085, "step": 55760 }, { "grad_norm": 0.2943333387374878, "learning_rate": 4.4605328929765624e-05, "loss": 0.0087, "step": 55770 }, { "grad_norm": 0.25833436846733093, "learning_rate": 4.458889104872102e-05, "loss": 0.0082, "step": 55780 }, { "grad_norm": 0.18352673947811127, "learning_rate": 4.457245375942715e-05, "loss": 0.0087, "step": 55790 }, { "grad_norm": 0.11158076673746109, "learning_rate": 4.455601706368157e-05, "loss": 0.0083, "step": 55800 }, { "grad_norm": 0.12752506136894226, "learning_rate": 4.453958096328179e-05, "loss": 0.0071, "step": 55810 }, { "grad_norm": 0.16469886898994446, "learning_rate": 4.452314546002523e-05, "loss": 0.0086, "step": 55820 }, { "grad_norm": 0.20411798357963562, "learning_rate": 4.4506710555709246e-05, "loss": 0.0088, "step": 55830 }, { "grad_norm": 0.2693549692630768, "learning_rate": 4.449027625213114e-05, "loss": 0.0085, "step": 55840 }, { "grad_norm": 0.13304533064365387, "learning_rate": 4.4473842551088135e-05, "loss": 0.0071, "step": 55850 }, { "grad_norm": 0.13971081376075745, "learning_rate": 4.445740945437738e-05, "loss": 0.0085, "step": 55860 }, { "grad_norm": 0.14358966052532196, "learning_rate": 4.444097696379602e-05, "loss": 0.0096, "step": 55870 }, { "grad_norm": 0.24742239713668823, "learning_rate": 4.442454508114106e-05, "loss": 0.0077, "step": 55880 }, { "grad_norm": 0.11622568219900131, "learning_rate": 4.4408113808209465e-05, "loss": 0.0077, "step": 55890 }, { "grad_norm": 0.16754305362701416, "learning_rate": 4.439168314679813e-05, "loss": 0.0075, "step": 55900 }, { "grad_norm": 0.21266144514083862, "learning_rate": 4.43752530987039e-05, "loss": 0.0081, "step": 55910 }, { "grad_norm": 0.141554594039917, "learning_rate": 4.4358823665723515e-05, "loss": 0.0089, "step": 55920 }, { "grad_norm": 0.15342170000076294, "learning_rate": 4.434239484965371e-05, "loss": 0.0087, "step": 55930 }, { "grad_norm": 0.13038939237594604, "learning_rate": 4.4325966652291103e-05, "loss": 0.0085, "step": 55940 }, { "grad_norm": 0.14778760075569153, "learning_rate": 4.430953907543225e-05, "loss": 0.0092, "step": 55950 }, { "grad_norm": 0.18137843906879425, "learning_rate": 4.4293112120873645e-05, "loss": 0.0089, "step": 55960 }, { "grad_norm": 0.144325852394104, "learning_rate": 4.427668579041172e-05, "loss": 0.0077, "step": 55970 }, { "grad_norm": 0.1533881276845932, "learning_rate": 4.4260260085842816e-05, "loss": 0.0096, "step": 55980 }, { "grad_norm": 0.18807558715343475, "learning_rate": 4.4243835008963256e-05, "loss": 0.0084, "step": 55990 }, { "grad_norm": 0.14488187432289124, "learning_rate": 4.422741056156925e-05, "loss": 0.0085, "step": 56000 }, { "grad_norm": 0.18125219643115997, "learning_rate": 4.421098674545693e-05, "loss": 0.0094, "step": 56010 }, { "grad_norm": 0.15674974024295807, "learning_rate": 4.419456356242241e-05, "loss": 0.0084, "step": 56020 }, { "grad_norm": 0.18217305839061737, "learning_rate": 4.4178141014261685e-05, "loss": 0.0082, "step": 56030 }, { "grad_norm": 0.17110088467597961, "learning_rate": 4.416171910277071e-05, "loss": 0.0105, "step": 56040 }, { "grad_norm": 0.19117699563503265, "learning_rate": 4.414529782974533e-05, "loss": 0.0091, "step": 56050 }, { "grad_norm": 0.17785494029521942, "learning_rate": 4.41288771969814e-05, "loss": 0.0092, "step": 56060 }, { "grad_norm": 0.2006179839372635, "learning_rate": 4.411245720627464e-05, "loss": 0.0094, "step": 56070 }, { "grad_norm": 0.2192118912935257, "learning_rate": 4.40960378594207e-05, "loss": 0.0072, "step": 56080 }, { "grad_norm": 0.11451538652181625, "learning_rate": 4.407961915821519e-05, "loss": 0.0107, "step": 56090 }, { "grad_norm": 0.17866083979606628, "learning_rate": 4.406320110445362e-05, "loss": 0.0094, "step": 56100 }, { "grad_norm": 0.1941496878862381, "learning_rate": 4.404678369993144e-05, "loss": 0.009, "step": 56110 }, { "grad_norm": 0.16189399361610413, "learning_rate": 4.403036694644406e-05, "loss": 0.0093, "step": 56120 }, { "grad_norm": 0.15780620276927948, "learning_rate": 4.4013950845786764e-05, "loss": 0.0097, "step": 56130 }, { "grad_norm": 0.24339158833026886, "learning_rate": 4.399753539975482e-05, "loss": 0.009, "step": 56140 }, { "grad_norm": 0.14276066422462463, "learning_rate": 4.398112061014337e-05, "loss": 0.0093, "step": 56150 }, { "grad_norm": 0.10006009787321091, "learning_rate": 4.396470647874753e-05, "loss": 0.0072, "step": 56160 }, { "grad_norm": 0.14580610394477844, "learning_rate": 4.394829300736229e-05, "loss": 0.0075, "step": 56170 }, { "grad_norm": 0.1228165253996849, "learning_rate": 4.393188019778265e-05, "loss": 0.0091, "step": 56180 }, { "grad_norm": 0.14851047098636627, "learning_rate": 4.391546805180347e-05, "loss": 0.0079, "step": 56190 }, { "grad_norm": 0.157237708568573, "learning_rate": 4.389905657121955e-05, "loss": 0.0078, "step": 56200 }, { "grad_norm": 0.13801456987857819, "learning_rate": 4.3882645757825635e-05, "loss": 0.0086, "step": 56210 }, { "grad_norm": 0.17615610361099243, "learning_rate": 4.386623561341637e-05, "loss": 0.0104, "step": 56220 }, { "grad_norm": 0.1837981790304184, "learning_rate": 4.384982613978637e-05, "loss": 0.0091, "step": 56230 }, { "grad_norm": 0.20354241132736206, "learning_rate": 4.383341733873012e-05, "loss": 0.0077, "step": 56240 }, { "grad_norm": 0.12479415535926819, "learning_rate": 4.381700921204209e-05, "loss": 0.0077, "step": 56250 }, { "grad_norm": 0.24614600837230682, "learning_rate": 4.380060176151663e-05, "loss": 0.0086, "step": 56260 }, { "grad_norm": 0.19625361263751984, "learning_rate": 4.378419498894805e-05, "loss": 0.0099, "step": 56270 }, { "grad_norm": 0.17932866513729095, "learning_rate": 4.376778889613056e-05, "loss": 0.0068, "step": 56280 }, { "grad_norm": 0.14490938186645508, "learning_rate": 4.375138348485831e-05, "loss": 0.0077, "step": 56290 }, { "grad_norm": 0.2270965725183487, "learning_rate": 4.3734978756925346e-05, "loss": 0.0094, "step": 56300 }, { "grad_norm": 0.15802137553691864, "learning_rate": 4.3718574714125706e-05, "loss": 0.0109, "step": 56310 }, { "grad_norm": 0.32819345593452454, "learning_rate": 4.370217135825329e-05, "loss": 0.0106, "step": 56320 }, { "grad_norm": 0.20571577548980713, "learning_rate": 4.368576869110194e-05, "loss": 0.0095, "step": 56330 }, { "grad_norm": 0.18674516677856445, "learning_rate": 4.366936671446544e-05, "loss": 0.008, "step": 56340 }, { "grad_norm": 0.179727703332901, "learning_rate": 4.3652965430137474e-05, "loss": 0.0094, "step": 56350 }, { "grad_norm": 0.18793267011642456, "learning_rate": 4.3636564839911646e-05, "loss": 0.0089, "step": 56360 }, { "grad_norm": 0.14208665490150452, "learning_rate": 4.3620164945581545e-05, "loss": 0.0092, "step": 56370 }, { "grad_norm": 0.16964666545391083, "learning_rate": 4.360376574894061e-05, "loss": 0.0081, "step": 56380 }, { "grad_norm": 0.21146464347839355, "learning_rate": 4.358736725178224e-05, "loss": 0.0091, "step": 56390 }, { "grad_norm": 0.15055428445339203, "learning_rate": 4.357096945589974e-05, "loss": 0.0106, "step": 56400 }, { "grad_norm": 0.2067677527666092, "learning_rate": 4.355457236308636e-05, "loss": 0.009, "step": 56410 }, { "grad_norm": 0.1729612797498703, "learning_rate": 4.353817597513526e-05, "loss": 0.0101, "step": 56420 }, { "grad_norm": 0.18121376633644104, "learning_rate": 4.352178029383948e-05, "loss": 0.0091, "step": 56430 }, { "grad_norm": 0.20473067462444305, "learning_rate": 4.3505385320992105e-05, "loss": 0.0096, "step": 56440 }, { "grad_norm": 0.13617263734340668, "learning_rate": 4.348899105838602e-05, "loss": 0.0087, "step": 56450 }, { "grad_norm": 0.21209341287612915, "learning_rate": 4.3472597507814087e-05, "loss": 0.0079, "step": 56460 }, { "grad_norm": 0.17402049899101257, "learning_rate": 4.3456204671069066e-05, "loss": 0.008, "step": 56470 }, { "grad_norm": 0.19485719501972198, "learning_rate": 4.343981254994367e-05, "loss": 0.0079, "step": 56480 }, { "grad_norm": 0.1606254279613495, "learning_rate": 4.34234211462305e-05, "loss": 0.0083, "step": 56490 }, { "grad_norm": 0.23376628756523132, "learning_rate": 4.340703046172213e-05, "loss": 0.0079, "step": 56500 }, { "grad_norm": 0.12725695967674255, "learning_rate": 4.339064049821097e-05, "loss": 0.0077, "step": 56510 }, { "grad_norm": 0.2974849045276642, "learning_rate": 4.3374251257489446e-05, "loss": 0.0084, "step": 56520 }, { "grad_norm": 0.15535013377666473, "learning_rate": 4.335786274134984e-05, "loss": 0.0078, "step": 56530 }, { "grad_norm": 0.17233620584011078, "learning_rate": 4.334147495158438e-05, "loss": 0.0091, "step": 56540 }, { "grad_norm": 0.23029375076293945, "learning_rate": 4.33250878899852e-05, "loss": 0.0084, "step": 56550 }, { "grad_norm": 0.15790171921253204, "learning_rate": 4.330870155834437e-05, "loss": 0.0068, "step": 56560 }, { "grad_norm": 0.2536129653453827, "learning_rate": 4.3292315958453884e-05, "loss": 0.008, "step": 56570 }, { "grad_norm": 0.1983443945646286, "learning_rate": 4.3275931092105646e-05, "loss": 0.0069, "step": 56580 }, { "grad_norm": 0.13167496025562286, "learning_rate": 4.325954696109147e-05, "loss": 0.0077, "step": 56590 }, { "grad_norm": 0.17945332825183868, "learning_rate": 4.32431635672031e-05, "loss": 0.0069, "step": 56600 }, { "grad_norm": 0.1311894953250885, "learning_rate": 4.3226780912232214e-05, "loss": 0.0073, "step": 56610 }, { "grad_norm": 0.11625690013170242, "learning_rate": 4.321039899797036e-05, "loss": 0.0092, "step": 56620 }, { "grad_norm": 0.1391272395849228, "learning_rate": 4.319401782620908e-05, "loss": 0.0073, "step": 56630 }, { "grad_norm": 0.12263011932373047, "learning_rate": 4.317763739873978e-05, "loss": 0.0088, "step": 56640 }, { "grad_norm": 0.13148778676986694, "learning_rate": 4.31612577173538e-05, "loss": 0.0083, "step": 56650 }, { "grad_norm": 0.2183859795331955, "learning_rate": 4.314487878384239e-05, "loss": 0.0087, "step": 56660 }, { "grad_norm": 0.12740612030029297, "learning_rate": 4.3128500599996736e-05, "loss": 0.0096, "step": 56670 }, { "grad_norm": 0.132699117064476, "learning_rate": 4.31121231676079e-05, "loss": 0.0074, "step": 56680 }, { "grad_norm": 0.19328324496746063, "learning_rate": 4.309574648846694e-05, "loss": 0.0075, "step": 56690 }, { "grad_norm": 0.12368563562631607, "learning_rate": 4.3079370564364755e-05, "loss": 0.0076, "step": 56700 }, { "grad_norm": 0.23979945480823517, "learning_rate": 4.306299539709221e-05, "loss": 0.0088, "step": 56710 }, { "grad_norm": 0.18966038525104523, "learning_rate": 4.3046620988440046e-05, "loss": 0.0097, "step": 56720 }, { "grad_norm": 0.31389695405960083, "learning_rate": 4.303024734019897e-05, "loss": 0.0106, "step": 56730 }, { "grad_norm": 0.12446216493844986, "learning_rate": 4.3013874454159565e-05, "loss": 0.0088, "step": 56740 }, { "grad_norm": 0.17576579749584198, "learning_rate": 4.299750233211233e-05, "loss": 0.0082, "step": 56750 }, { "grad_norm": 0.12167984992265701, "learning_rate": 4.2981130975847715e-05, "loss": 0.0085, "step": 56760 }, { "grad_norm": 0.24088126420974731, "learning_rate": 4.2964760387156076e-05, "loss": 0.0092, "step": 56770 }, { "grad_norm": 0.20496250689029694, "learning_rate": 4.294839056782767e-05, "loss": 0.0073, "step": 56780 }, { "grad_norm": 0.220744326710701, "learning_rate": 4.293202151965266e-05, "loss": 0.0086, "step": 56790 }, { "grad_norm": 0.17380906641483307, "learning_rate": 4.291565324442115e-05, "loss": 0.0085, "step": 56800 }, { "grad_norm": 0.18589496612548828, "learning_rate": 4.2899285743923135e-05, "loss": 0.008, "step": 56810 }, { "grad_norm": 0.14252102375030518, "learning_rate": 4.288291901994857e-05, "loss": 0.0076, "step": 56820 }, { "grad_norm": 0.23947668075561523, "learning_rate": 4.286655307428729e-05, "loss": 0.0086, "step": 56830 }, { "grad_norm": 0.18216672539710999, "learning_rate": 4.285018790872903e-05, "loss": 0.0096, "step": 56840 }, { "grad_norm": 0.2056744247674942, "learning_rate": 4.283382352506347e-05, "loss": 0.008, "step": 56850 }, { "grad_norm": 0.18465034663677216, "learning_rate": 4.2817459925080205e-05, "loss": 0.0079, "step": 56860 }, { "grad_norm": 0.1909976452589035, "learning_rate": 4.28010971105687e-05, "loss": 0.0078, "step": 56870 }, { "grad_norm": 0.14188911020755768, "learning_rate": 4.278473508331841e-05, "loss": 0.0093, "step": 56880 }, { "grad_norm": 0.1448967307806015, "learning_rate": 4.276837384511864e-05, "loss": 0.0088, "step": 56890 }, { "grad_norm": 0.1499028354883194, "learning_rate": 4.275201339775864e-05, "loss": 0.0077, "step": 56900 }, { "grad_norm": 0.16409645974636078, "learning_rate": 4.273565374302756e-05, "loss": 0.0077, "step": 56910 }, { "grad_norm": 0.18211659789085388, "learning_rate": 4.2719294882714454e-05, "loss": 0.0083, "step": 56920 }, { "grad_norm": 0.1401965618133545, "learning_rate": 4.270293681860831e-05, "loss": 0.008, "step": 56930 }, { "grad_norm": 0.2396916151046753, "learning_rate": 4.268657955249802e-05, "loss": 0.0087, "step": 56940 }, { "grad_norm": 0.44265738129615784, "learning_rate": 4.267022308617241e-05, "loss": 0.0083, "step": 56950 }, { "grad_norm": 0.13697291910648346, "learning_rate": 4.265386742142018e-05, "loss": 0.0078, "step": 56960 }, { "grad_norm": 0.2227954864501953, "learning_rate": 4.263751256002996e-05, "loss": 0.0087, "step": 56970 }, { "grad_norm": 0.13938738405704498, "learning_rate": 4.26211585037903e-05, "loss": 0.0085, "step": 56980 }, { "grad_norm": 0.11951158195734024, "learning_rate": 4.2604805254489656e-05, "loss": 0.0071, "step": 56990 }, { "grad_norm": 0.12882912158966064, "learning_rate": 4.258845281391638e-05, "loss": 0.0099, "step": 57000 }, { "grad_norm": 0.20497654378414154, "learning_rate": 4.257210118385876e-05, "loss": 0.0078, "step": 57010 }, { "grad_norm": 0.1506042629480362, "learning_rate": 4.2555750366104994e-05, "loss": 0.0081, "step": 57020 }, { "grad_norm": 0.3874027729034424, "learning_rate": 4.2539400362443184e-05, "loss": 0.0089, "step": 57030 }, { "grad_norm": 0.16135822236537933, "learning_rate": 4.252305117466133e-05, "loss": 0.0098, "step": 57040 }, { "grad_norm": 0.17773140966892242, "learning_rate": 4.2506702804547356e-05, "loss": 0.008, "step": 57050 }, { "grad_norm": 0.1908087581396103, "learning_rate": 4.249035525388907e-05, "loss": 0.0078, "step": 57060 }, { "grad_norm": 0.1749008595943451, "learning_rate": 4.247400852447427e-05, "loss": 0.0088, "step": 57070 }, { "grad_norm": 0.32303133606910706, "learning_rate": 4.245766261809059e-05, "loss": 0.0082, "step": 57080 }, { "grad_norm": 0.25207817554473877, "learning_rate": 4.244131753652558e-05, "loss": 0.0086, "step": 57090 }, { "grad_norm": 0.2670714557170868, "learning_rate": 4.242497328156672e-05, "loss": 0.0091, "step": 57100 }, { "grad_norm": 0.16057024896144867, "learning_rate": 4.240862985500139e-05, "loss": 0.0088, "step": 57110 }, { "grad_norm": 0.3090117871761322, "learning_rate": 4.2392287258616894e-05, "loss": 0.0078, "step": 57120 }, { "grad_norm": 0.1306847780942917, "learning_rate": 4.2375945494200396e-05, "loss": 0.0081, "step": 57130 }, { "grad_norm": 0.15386530756950378, "learning_rate": 4.2359604563539054e-05, "loss": 0.0068, "step": 57140 }, { "grad_norm": 0.13740451633930206, "learning_rate": 4.2343264468419885e-05, "loss": 0.0076, "step": 57150 }, { "grad_norm": 0.1423378735780716, "learning_rate": 4.232692521062979e-05, "loss": 0.0097, "step": 57160 }, { "grad_norm": 0.24235855042934418, "learning_rate": 4.231058679195562e-05, "loss": 0.0091, "step": 57170 }, { "grad_norm": 0.16804806888103485, "learning_rate": 4.229424921418411e-05, "loss": 0.0074, "step": 57180 }, { "grad_norm": 0.19719849526882172, "learning_rate": 4.2277912479101916e-05, "loss": 0.0076, "step": 57190 }, { "grad_norm": 0.16503126919269562, "learning_rate": 4.226157658849562e-05, "loss": 0.0087, "step": 57200 }, { "grad_norm": 0.1933974027633667, "learning_rate": 4.2245241544151664e-05, "loss": 0.0078, "step": 57210 }, { "grad_norm": 0.1702367663383484, "learning_rate": 4.2228907347856425e-05, "loss": 0.0079, "step": 57220 }, { "grad_norm": 0.1307903677225113, "learning_rate": 4.2212574001396214e-05, "loss": 0.0108, "step": 57230 }, { "grad_norm": 0.2414083182811737, "learning_rate": 4.21962415065572e-05, "loss": 0.0074, "step": 57240 }, { "grad_norm": 0.15764518082141876, "learning_rate": 4.217990986512548e-05, "loss": 0.009, "step": 57250 }, { "grad_norm": 0.1962674856185913, "learning_rate": 4.216357907888704e-05, "loss": 0.0089, "step": 57260 }, { "grad_norm": 0.13447709381580353, "learning_rate": 4.2147249149627824e-05, "loss": 0.0073, "step": 57270 }, { "grad_norm": 0.14767542481422424, "learning_rate": 4.213092007913364e-05, "loss": 0.0093, "step": 57280 }, { "grad_norm": 0.12382348626852036, "learning_rate": 4.2114591869190205e-05, "loss": 0.0067, "step": 57290 }, { "grad_norm": 0.16236399114131927, "learning_rate": 4.209826452158315e-05, "loss": 0.0073, "step": 57300 }, { "grad_norm": 0.11255193501710892, "learning_rate": 4.2081938038098e-05, "loss": 0.0091, "step": 57310 }, { "grad_norm": 0.1433887779712677, "learning_rate": 4.2065612420520185e-05, "loss": 0.0078, "step": 57320 }, { "grad_norm": 0.138625830411911, "learning_rate": 4.204928767063509e-05, "loss": 0.0077, "step": 57330 }, { "grad_norm": 0.14474357664585114, "learning_rate": 4.2032963790227944e-05, "loss": 0.0087, "step": 57340 }, { "grad_norm": 0.13072308897972107, "learning_rate": 4.2016640781083887e-05, "loss": 0.0073, "step": 57350 }, { "grad_norm": 0.20044881105422974, "learning_rate": 4.2000318644988004e-05, "loss": 0.0099, "step": 57360 }, { "grad_norm": 0.31878405809402466, "learning_rate": 4.198399738372524e-05, "loss": 0.0097, "step": 57370 }, { "grad_norm": 0.1754048764705658, "learning_rate": 4.196767699908045e-05, "loss": 0.0093, "step": 57380 }, { "grad_norm": 0.14450930058956146, "learning_rate": 4.195135749283845e-05, "loss": 0.0087, "step": 57390 }, { "grad_norm": 0.1618386209011078, "learning_rate": 4.1935038866783894e-05, "loss": 0.0086, "step": 57400 }, { "grad_norm": 0.15773186087608337, "learning_rate": 4.1918721122701355e-05, "loss": 0.0084, "step": 57410 }, { "grad_norm": 0.2021162062883377, "learning_rate": 4.190240426237533e-05, "loss": 0.0083, "step": 57420 }, { "grad_norm": 0.11265247315168381, "learning_rate": 4.188608828759019e-05, "loss": 0.0077, "step": 57430 }, { "grad_norm": 0.13892987370491028, "learning_rate": 4.1869773200130246e-05, "loss": 0.0081, "step": 57440 }, { "grad_norm": 0.2236001342535019, "learning_rate": 4.185345900177967e-05, "loss": 0.0097, "step": 57450 }, { "grad_norm": 0.11434030532836914, "learning_rate": 4.183714569432258e-05, "loss": 0.007, "step": 57460 }, { "grad_norm": 0.16563917696475983, "learning_rate": 4.182083327954295e-05, "loss": 0.0078, "step": 57470 }, { "grad_norm": 0.1679747998714447, "learning_rate": 4.1804521759224714e-05, "loss": 0.0082, "step": 57480 }, { "grad_norm": 0.17881952226161957, "learning_rate": 4.178821113515165e-05, "loss": 0.007, "step": 57490 }, { "grad_norm": 0.35393157601356506, "learning_rate": 4.1771901409107476e-05, "loss": 0.0074, "step": 57500 }, { "grad_norm": 0.22390857338905334, "learning_rate": 4.175559258287577e-05, "loss": 0.0066, "step": 57510 }, { "grad_norm": 0.19212239980697632, "learning_rate": 4.173928465824008e-05, "loss": 0.0079, "step": 57520 }, { "grad_norm": 0.12089286744594574, "learning_rate": 4.17229776369838e-05, "loss": 0.0076, "step": 57530 }, { "grad_norm": 0.13965149223804474, "learning_rate": 4.170667152089025e-05, "loss": 0.0077, "step": 57540 }, { "grad_norm": 0.17223085463047028, "learning_rate": 4.1690366311742626e-05, "loss": 0.0076, "step": 57550 }, { "grad_norm": 0.151687353849411, "learning_rate": 4.1674062011324055e-05, "loss": 0.009, "step": 57560 }, { "grad_norm": 0.15026235580444336, "learning_rate": 4.165775862141752e-05, "loss": 0.0079, "step": 57570 }, { "grad_norm": 0.15733017027378082, "learning_rate": 4.164145614380598e-05, "loss": 0.0077, "step": 57580 }, { "grad_norm": 0.20394660532474518, "learning_rate": 4.162515458027223e-05, "loss": 0.0082, "step": 57590 }, { "grad_norm": 0.2816041111946106, "learning_rate": 4.160885393259899e-05, "loss": 0.0071, "step": 57600 }, { "grad_norm": 0.13656547665596008, "learning_rate": 4.159255420256886e-05, "loss": 0.0078, "step": 57610 }, { "grad_norm": 0.1522543877363205, "learning_rate": 4.157625539196436e-05, "loss": 0.007, "step": 57620 }, { "grad_norm": 0.2828614413738251, "learning_rate": 4.155995750256791e-05, "loss": 0.0067, "step": 57630 }, { "grad_norm": 0.11963091790676117, "learning_rate": 4.15436605361618e-05, "loss": 0.0082, "step": 57640 }, { "grad_norm": 0.143487349152565, "learning_rate": 4.152736449452827e-05, "loss": 0.0063, "step": 57650 }, { "grad_norm": 0.1954926699399948, "learning_rate": 4.151106937944943e-05, "loss": 0.0071, "step": 57660 }, { "grad_norm": 0.1289529800415039, "learning_rate": 4.149477519270727e-05, "loss": 0.0076, "step": 57670 }, { "grad_norm": 0.22247432172298431, "learning_rate": 4.147848193608371e-05, "loss": 0.008, "step": 57680 }, { "grad_norm": 0.3117714524269104, "learning_rate": 4.146218961136056e-05, "loss": 0.0066, "step": 57690 }, { "grad_norm": 0.179881751537323, "learning_rate": 4.14458982203195e-05, "loss": 0.0091, "step": 57700 }, { "grad_norm": 0.15246137976646423, "learning_rate": 4.1429607764742165e-05, "loss": 0.0069, "step": 57710 }, { "grad_norm": 0.15727496147155762, "learning_rate": 4.141331824641003e-05, "loss": 0.0072, "step": 57720 }, { "grad_norm": 0.17453625798225403, "learning_rate": 4.139702966710452e-05, "loss": 0.0082, "step": 57730 }, { "grad_norm": 0.167343407869339, "learning_rate": 4.13807420286069e-05, "loss": 0.0075, "step": 57740 }, { "grad_norm": 0.13826872408390045, "learning_rate": 4.136445533269837e-05, "loss": 0.0084, "step": 57750 }, { "grad_norm": 0.18990010023117065, "learning_rate": 4.134816958116003e-05, "loss": 0.0083, "step": 57760 }, { "grad_norm": 0.12995854020118713, "learning_rate": 4.133188477577283e-05, "loss": 0.0075, "step": 57770 }, { "grad_norm": 0.22822999954223633, "learning_rate": 4.131560091831771e-05, "loss": 0.0071, "step": 57780 }, { "grad_norm": 0.2500489354133606, "learning_rate": 4.1299318010575405e-05, "loss": 0.0081, "step": 57790 }, { "grad_norm": 0.14377175271511078, "learning_rate": 4.128303605432661e-05, "loss": 0.0085, "step": 57800 }, { "grad_norm": 0.246780127286911, "learning_rate": 4.126675505135188e-05, "loss": 0.0076, "step": 57810 }, { "grad_norm": 0.14498770236968994, "learning_rate": 4.125047500343169e-05, "loss": 0.0069, "step": 57820 }, { "grad_norm": 0.2096666842699051, "learning_rate": 4.1234195912346376e-05, "loss": 0.0071, "step": 57830 }, { "grad_norm": 0.11936012655496597, "learning_rate": 4.1217917779876235e-05, "loss": 0.0076, "step": 57840 }, { "grad_norm": 0.15456442534923553, "learning_rate": 4.120164060780141e-05, "loss": 0.0087, "step": 57850 }, { "grad_norm": 0.1289590746164322, "learning_rate": 4.1185364397901935e-05, "loss": 0.0079, "step": 57860 }, { "grad_norm": 0.18346351385116577, "learning_rate": 4.116908915195776e-05, "loss": 0.0067, "step": 57870 }, { "grad_norm": 0.15068921446800232, "learning_rate": 4.115281487174872e-05, "loss": 0.0063, "step": 57880 }, { "grad_norm": 0.21011893451213837, "learning_rate": 4.113654155905452e-05, "loss": 0.0094, "step": 57890 }, { "grad_norm": 0.2634514272212982, "learning_rate": 4.112026921565483e-05, "loss": 0.0084, "step": 57900 }, { "grad_norm": 0.19278393685817719, "learning_rate": 4.1103997843329147e-05, "loss": 0.007, "step": 57910 }, { "grad_norm": 0.1357666403055191, "learning_rate": 4.108772744385689e-05, "loss": 0.0086, "step": 57920 }, { "grad_norm": 0.19540773332118988, "learning_rate": 4.1071458019017346e-05, "loss": 0.0077, "step": 57930 }, { "grad_norm": 0.19184811413288116, "learning_rate": 4.1055189570589734e-05, "loss": 0.0073, "step": 57940 }, { "grad_norm": 0.1561758667230606, "learning_rate": 4.103892210035315e-05, "loss": 0.0071, "step": 57950 }, { "grad_norm": 0.22840149700641632, "learning_rate": 4.102265561008656e-05, "loss": 0.0085, "step": 57960 }, { "grad_norm": 0.2533034086227417, "learning_rate": 4.100639010156885e-05, "loss": 0.008, "step": 57970 }, { "grad_norm": 0.2329883873462677, "learning_rate": 4.099012557657882e-05, "loss": 0.0073, "step": 57980 }, { "grad_norm": 0.12638044357299805, "learning_rate": 4.09738620368951e-05, "loss": 0.0077, "step": 57990 }, { "grad_norm": 0.12872575223445892, "learning_rate": 4.095759948429626e-05, "loss": 0.0075, "step": 58000 }, { "grad_norm": 0.12559781968593597, "learning_rate": 4.094133792056075e-05, "loss": 0.0077, "step": 58010 }, { "grad_norm": 0.1429169923067093, "learning_rate": 4.092507734746688e-05, "loss": 0.008, "step": 58020 }, { "grad_norm": 0.12631447613239288, "learning_rate": 4.090881776679293e-05, "loss": 0.0068, "step": 58030 }, { "grad_norm": 0.15468333661556244, "learning_rate": 4.0892559180316994e-05, "loss": 0.0076, "step": 58040 }, { "grad_norm": 0.13599511981010437, "learning_rate": 4.08763015898171e-05, "loss": 0.0083, "step": 58050 }, { "grad_norm": 0.12927918136119843, "learning_rate": 4.0860044997071145e-05, "loss": 0.0079, "step": 58060 }, { "grad_norm": 0.2986169755458832, "learning_rate": 4.084378940385693e-05, "loss": 0.0085, "step": 58070 }, { "grad_norm": 0.16049306094646454, "learning_rate": 4.082753481195211e-05, "loss": 0.0073, "step": 58080 }, { "grad_norm": 0.21090754866600037, "learning_rate": 4.0811281223134324e-05, "loss": 0.0071, "step": 58090 }, { "grad_norm": 0.14682736992835999, "learning_rate": 4.0795028639181e-05, "loss": 0.0112, "step": 58100 }, { "grad_norm": 0.13139575719833374, "learning_rate": 4.077877706186951e-05, "loss": 0.0067, "step": 58110 }, { "grad_norm": 0.15297259390354156, "learning_rate": 4.076252649297709e-05, "loss": 0.0078, "step": 58120 }, { "grad_norm": 0.14569880068302155, "learning_rate": 4.07462769342809e-05, "loss": 0.0076, "step": 58130 }, { "grad_norm": 0.2201182097196579, "learning_rate": 4.073002838755793e-05, "loss": 0.0072, "step": 58140 }, { "grad_norm": 0.24265213310718536, "learning_rate": 4.0713780854585115e-05, "loss": 0.0074, "step": 58150 }, { "grad_norm": 0.20337523519992828, "learning_rate": 4.069753433713927e-05, "loss": 0.0098, "step": 58160 }, { "grad_norm": 0.1396406888961792, "learning_rate": 4.068128883699709e-05, "loss": 0.0071, "step": 58170 }, { "grad_norm": 0.10374059528112411, "learning_rate": 4.0665044355935155e-05, "loss": 0.0083, "step": 58180 }, { "grad_norm": 0.2874423861503601, "learning_rate": 4.064880089572993e-05, "loss": 0.0083, "step": 58190 }, { "grad_norm": 0.15123385190963745, "learning_rate": 4.063255845815779e-05, "loss": 0.0098, "step": 58200 }, { "grad_norm": 0.2815217971801758, "learning_rate": 4.0616317044994945e-05, "loss": 0.0105, "step": 58210 }, { "grad_norm": 0.15079763531684875, "learning_rate": 4.0600076658017585e-05, "loss": 0.0086, "step": 58220 }, { "grad_norm": 0.1657152622938156, "learning_rate": 4.0583837299001706e-05, "loss": 0.0076, "step": 58230 }, { "grad_norm": 0.17564421892166138, "learning_rate": 4.056759896972322e-05, "loss": 0.0072, "step": 58240 }, { "grad_norm": 0.14154906570911407, "learning_rate": 4.055136167195793e-05, "loss": 0.0087, "step": 58250 }, { "grad_norm": 0.19575363397598267, "learning_rate": 4.053512540748152e-05, "loss": 0.0067, "step": 58260 }, { "grad_norm": 0.09996209293603897, "learning_rate": 4.051889017806955e-05, "loss": 0.0064, "step": 58270 }, { "grad_norm": 0.1398632675409317, "learning_rate": 4.0502655985497516e-05, "loss": 0.0055, "step": 58280 }, { "grad_norm": 0.19778139889240265, "learning_rate": 4.048642283154074e-05, "loss": 0.0077, "step": 58290 }, { "grad_norm": 0.18504442274570465, "learning_rate": 4.0470190717974464e-05, "loss": 0.0095, "step": 58300 }, { "grad_norm": 0.3660354018211365, "learning_rate": 4.045395964657381e-05, "loss": 0.0083, "step": 58310 }, { "grad_norm": 0.17017720639705658, "learning_rate": 4.043772961911376e-05, "loss": 0.0077, "step": 58320 }, { "grad_norm": 0.17165310680866241, "learning_rate": 4.042150063736924e-05, "loss": 0.0076, "step": 58330 }, { "grad_norm": 0.1937498301267624, "learning_rate": 4.0405272703114976e-05, "loss": 0.0076, "step": 58340 }, { "grad_norm": 0.1227049008011818, "learning_rate": 4.038904581812569e-05, "loss": 0.0072, "step": 58350 }, { "grad_norm": 0.19709722697734833, "learning_rate": 4.03728199841759e-05, "loss": 0.0082, "step": 58360 }, { "grad_norm": 0.17821104824543, "learning_rate": 4.0356595203040044e-05, "loss": 0.0078, "step": 58370 }, { "grad_norm": 0.11471787840127945, "learning_rate": 4.0340371476492426e-05, "loss": 0.0069, "step": 58380 }, { "grad_norm": 0.15329404175281525, "learning_rate": 4.032414880630727e-05, "loss": 0.0074, "step": 58390 }, { "grad_norm": 0.29427963495254517, "learning_rate": 4.0307927194258646e-05, "loss": 0.0088, "step": 58400 }, { "grad_norm": 0.10935578495264053, "learning_rate": 4.029170664212054e-05, "loss": 0.0085, "step": 58410 }, { "grad_norm": 0.08040900528430939, "learning_rate": 4.0275487151666794e-05, "loss": 0.0071, "step": 58420 }, { "grad_norm": 0.15509963035583496, "learning_rate": 4.025926872467116e-05, "loss": 0.0082, "step": 58430 }, { "grad_norm": 0.12426441162824631, "learning_rate": 4.024305136290725e-05, "loss": 0.0085, "step": 58440 }, { "grad_norm": 0.23051638901233673, "learning_rate": 4.022683506814858e-05, "loss": 0.0076, "step": 58450 }, { "grad_norm": 0.13634991645812988, "learning_rate": 4.021061984216853e-05, "loss": 0.0082, "step": 58460 }, { "grad_norm": 0.15056833624839783, "learning_rate": 4.019440568674035e-05, "loss": 0.0075, "step": 58470 }, { "grad_norm": 0.22050084173679352, "learning_rate": 4.017819260363724e-05, "loss": 0.0086, "step": 58480 }, { "grad_norm": 0.1271389126777649, "learning_rate": 4.016198059463223e-05, "loss": 0.0071, "step": 58490 }, { "grad_norm": 0.22495295107364655, "learning_rate": 4.0145769661498233e-05, "loss": 0.0073, "step": 58500 }, { "grad_norm": 0.2574532926082611, "learning_rate": 4.012955980600804e-05, "loss": 0.0078, "step": 58510 }, { "grad_norm": 0.16589821875095367, "learning_rate": 4.0113351029934355e-05, "loss": 0.0077, "step": 58520 }, { "grad_norm": 0.204730823636055, "learning_rate": 4.0097143335049714e-05, "loss": 0.0081, "step": 58530 }, { "grad_norm": 0.15233172476291656, "learning_rate": 4.0080936723126606e-05, "loss": 0.0096, "step": 58540 }, { "grad_norm": 0.2671334147453308, "learning_rate": 4.006473119593735e-05, "loss": 0.0085, "step": 58550 }, { "grad_norm": 0.12139949202537537, "learning_rate": 4.004852675525415e-05, "loss": 0.0077, "step": 58560 }, { "grad_norm": 0.28895387053489685, "learning_rate": 4.00323234028491e-05, "loss": 0.0091, "step": 58570 }, { "grad_norm": 0.15513694286346436, "learning_rate": 4.0016121140494186e-05, "loss": 0.0078, "step": 58580 }, { "grad_norm": 0.2565425932407379, "learning_rate": 3.999991996996122e-05, "loss": 0.0072, "step": 58590 }, { "grad_norm": 0.11474662274122238, "learning_rate": 3.9983719893022e-05, "loss": 0.0069, "step": 58600 }, { "grad_norm": 0.34768378734588623, "learning_rate": 3.996752091144812e-05, "loss": 0.0078, "step": 58610 }, { "grad_norm": 0.21486350893974304, "learning_rate": 3.9951323027011055e-05, "loss": 0.0076, "step": 58620 }, { "grad_norm": 0.24232646822929382, "learning_rate": 3.993512624148219e-05, "loss": 0.0086, "step": 58630 }, { "grad_norm": 0.13518111407756805, "learning_rate": 3.99189305566328e-05, "loss": 0.0067, "step": 58640 }, { "grad_norm": 0.13915960490703583, "learning_rate": 3.990273597423401e-05, "loss": 0.0061, "step": 58650 }, { "grad_norm": 0.27588894963264465, "learning_rate": 3.988654249605681e-05, "loss": 0.008, "step": 58660 }, { "grad_norm": 0.26223111152648926, "learning_rate": 3.987035012387213e-05, "loss": 0.0088, "step": 58670 }, { "grad_norm": 0.15614263713359833, "learning_rate": 3.9854158859450735e-05, "loss": 0.0091, "step": 58680 }, { "grad_norm": 0.3173205554485321, "learning_rate": 3.983796870456328e-05, "loss": 0.0101, "step": 58690 }, { "grad_norm": 0.26237815618515015, "learning_rate": 3.982177966098028e-05, "loss": 0.0069, "step": 58700 }, { "grad_norm": 0.18850350379943848, "learning_rate": 3.980559173047216e-05, "loss": 0.009, "step": 58710 }, { "grad_norm": 0.16400454938411713, "learning_rate": 3.9789404914809176e-05, "loss": 0.0081, "step": 58720 }, { "grad_norm": 0.0859561637043953, "learning_rate": 3.9773219215761544e-05, "loss": 0.0065, "step": 58730 }, { "grad_norm": 0.1560143232345581, "learning_rate": 3.975703463509928e-05, "loss": 0.0088, "step": 58740 }, { "grad_norm": 0.2494698315858841, "learning_rate": 3.974085117459231e-05, "loss": 0.0078, "step": 58750 }, { "grad_norm": 0.14365890622138977, "learning_rate": 3.972466883601043e-05, "loss": 0.0075, "step": 58760 }, { "grad_norm": 0.17988884449005127, "learning_rate": 3.9708487621123316e-05, "loss": 0.0086, "step": 58770 }, { "grad_norm": 0.2707323431968689, "learning_rate": 3.969230753170051e-05, "loss": 0.0079, "step": 58780 }, { "grad_norm": 0.17137488722801208, "learning_rate": 3.967612856951146e-05, "loss": 0.0083, "step": 58790 }, { "grad_norm": 0.12247268110513687, "learning_rate": 3.9659950736325476e-05, "loss": 0.0101, "step": 58800 }, { "grad_norm": 0.20234636962413788, "learning_rate": 3.964377403391174e-05, "loss": 0.0084, "step": 58810 }, { "grad_norm": 0.11658269912004471, "learning_rate": 3.9627598464039303e-05, "loss": 0.0079, "step": 58820 }, { "grad_norm": 0.248533234000206, "learning_rate": 3.9611424028477096e-05, "loss": 0.0066, "step": 58830 }, { "grad_norm": 0.12527336180210114, "learning_rate": 3.959525072899394e-05, "loss": 0.0084, "step": 58840 }, { "grad_norm": 0.12127886712551117, "learning_rate": 3.9579078567358506e-05, "loss": 0.0081, "step": 58850 }, { "grad_norm": 0.12820212543010712, "learning_rate": 3.956290754533939e-05, "loss": 0.0062, "step": 58860 }, { "grad_norm": 0.19144418835639954, "learning_rate": 3.954673766470501e-05, "loss": 0.0065, "step": 58870 }, { "grad_norm": 0.13370481133460999, "learning_rate": 3.9530568927223674e-05, "loss": 0.0082, "step": 58880 }, { "grad_norm": 0.126048743724823, "learning_rate": 3.95144013346636e-05, "loss": 0.0088, "step": 58890 }, { "grad_norm": 0.20016886293888092, "learning_rate": 3.949823488879282e-05, "loss": 0.0074, "step": 58900 }, { "grad_norm": 0.15479522943496704, "learning_rate": 3.9482069591379276e-05, "loss": 0.008, "step": 58910 }, { "grad_norm": 0.15129126608371735, "learning_rate": 3.9465905444190784e-05, "loss": 0.0075, "step": 58920 }, { "grad_norm": 0.17200037837028503, "learning_rate": 3.944974244899505e-05, "loss": 0.0098, "step": 58930 }, { "grad_norm": 0.15180152654647827, "learning_rate": 3.9433580607559616e-05, "loss": 0.0091, "step": 58940 }, { "grad_norm": 0.1550995409488678, "learning_rate": 3.941741992165193e-05, "loss": 0.0087, "step": 58950 }, { "grad_norm": 0.13814899325370789, "learning_rate": 3.940126039303928e-05, "loss": 0.007, "step": 58960 }, { "grad_norm": 0.14846068620681763, "learning_rate": 3.938510202348886e-05, "loss": 0.0066, "step": 58970 }, { "grad_norm": 0.2172233909368515, "learning_rate": 3.93689448147677e-05, "loss": 0.007, "step": 58980 }, { "grad_norm": 0.13000746071338654, "learning_rate": 3.935278876864278e-05, "loss": 0.007, "step": 58990 }, { "grad_norm": 0.15039250254631042, "learning_rate": 3.933663388688087e-05, "loss": 0.0076, "step": 59000 }, { "grad_norm": 0.12471645325422287, "learning_rate": 3.9320480171248645e-05, "loss": 0.0089, "step": 59010 }, { "grad_norm": 0.19372540712356567, "learning_rate": 3.930432762351265e-05, "loss": 0.0075, "step": 59020 }, { "grad_norm": 0.14746646583080292, "learning_rate": 3.928817624543931e-05, "loss": 0.0076, "step": 59030 }, { "grad_norm": 0.21596620976924896, "learning_rate": 3.9272026038794885e-05, "loss": 0.0091, "step": 59040 }, { "grad_norm": 0.10750047862529755, "learning_rate": 3.9255877005345586e-05, "loss": 0.0078, "step": 59050 }, { "grad_norm": 0.14923368394374847, "learning_rate": 3.923972914685743e-05, "loss": 0.0084, "step": 59060 }, { "grad_norm": 0.20585376024246216, "learning_rate": 3.922358246509631e-05, "loss": 0.0075, "step": 59070 }, { "grad_norm": 0.18428575992584229, "learning_rate": 3.920743696182801e-05, "loss": 0.0095, "step": 59080 }, { "grad_norm": 0.13718703389167786, "learning_rate": 3.9191292638818164e-05, "loss": 0.0081, "step": 59090 }, { "grad_norm": 0.14869838953018188, "learning_rate": 3.9175149497832316e-05, "loss": 0.0084, "step": 59100 }, { "grad_norm": 0.19643373787403107, "learning_rate": 3.915900754063584e-05, "loss": 0.0083, "step": 59110 }, { "grad_norm": 0.1633133590221405, "learning_rate": 3.9142866768994014e-05, "loss": 0.0066, "step": 59120 }, { "grad_norm": 0.1724398136138916, "learning_rate": 3.912672718467193e-05, "loss": 0.0076, "step": 59130 }, { "grad_norm": 0.11183283478021622, "learning_rate": 3.911058878943463e-05, "loss": 0.0086, "step": 59140 }, { "grad_norm": 0.14339181780815125, "learning_rate": 3.9094451585046974e-05, "loss": 0.0067, "step": 59150 }, { "grad_norm": 0.1418873369693756, "learning_rate": 3.90783155732737e-05, "loss": 0.008, "step": 59160 }, { "grad_norm": 0.19730859994888306, "learning_rate": 3.9062180755879373e-05, "loss": 0.0084, "step": 59170 }, { "grad_norm": 0.15010704100131989, "learning_rate": 3.904604713462855e-05, "loss": 0.0076, "step": 59180 }, { "grad_norm": 0.13475395739078522, "learning_rate": 3.902991471128554e-05, "loss": 0.0075, "step": 59190 }, { "grad_norm": 0.16265985369682312, "learning_rate": 3.901378348761456e-05, "loss": 0.0122, "step": 59200 }, { "grad_norm": 0.15912185609340668, "learning_rate": 3.899765346537971e-05, "loss": 0.0092, "step": 59210 }, { "grad_norm": 0.1321820318698883, "learning_rate": 3.898152464634493e-05, "loss": 0.0071, "step": 59220 }, { "grad_norm": 0.12590359151363373, "learning_rate": 3.896539703227402e-05, "loss": 0.007, "step": 59230 }, { "grad_norm": 0.12092378735542297, "learning_rate": 3.894927062493073e-05, "loss": 0.0066, "step": 59240 }, { "grad_norm": 0.1183379665017128, "learning_rate": 3.893314542607859e-05, "loss": 0.0079, "step": 59250 }, { "grad_norm": 0.15583376586437225, "learning_rate": 3.891702143748103e-05, "loss": 0.0067, "step": 59260 }, { "grad_norm": 0.13543415069580078, "learning_rate": 3.890089866090133e-05, "loss": 0.0085, "step": 59270 }, { "grad_norm": 0.25997498631477356, "learning_rate": 3.888477709810267e-05, "loss": 0.0079, "step": 59280 }, { "grad_norm": 0.1991759091615677, "learning_rate": 3.886865675084806e-05, "loss": 0.0094, "step": 59290 }, { "grad_norm": 0.11168617010116577, "learning_rate": 3.885253762090043e-05, "loss": 0.0063, "step": 59300 }, { "grad_norm": 0.20722456276416779, "learning_rate": 3.8836419710022534e-05, "loss": 0.0075, "step": 59310 }, { "grad_norm": 0.1500064730644226, "learning_rate": 3.882030301997698e-05, "loss": 0.0076, "step": 59320 }, { "grad_norm": 0.17358408868312836, "learning_rate": 3.880418755252629e-05, "loss": 0.0076, "step": 59330 }, { "grad_norm": 0.15291541814804077, "learning_rate": 3.87880733094328e-05, "loss": 0.0087, "step": 59340 }, { "grad_norm": 0.16214551031589508, "learning_rate": 3.877196029245877e-05, "loss": 0.0076, "step": 59350 }, { "grad_norm": 0.12506617605686188, "learning_rate": 3.875584850336627e-05, "loss": 0.0091, "step": 59360 }, { "grad_norm": 0.14057384431362152, "learning_rate": 3.873973794391728e-05, "loss": 0.0078, "step": 59370 }, { "grad_norm": 0.16080346703529358, "learning_rate": 3.872362861587361e-05, "loss": 0.007, "step": 59380 }, { "grad_norm": 0.17816922068595886, "learning_rate": 3.870752052099698e-05, "loss": 0.0068, "step": 59390 }, { "grad_norm": 0.1850382387638092, "learning_rate": 3.8691413661048916e-05, "loss": 0.0088, "step": 59400 }, { "grad_norm": 0.15213972330093384, "learning_rate": 3.867530803779085e-05, "loss": 0.0085, "step": 59410 }, { "grad_norm": 0.11824560910463333, "learning_rate": 3.8659203652984054e-05, "loss": 0.0086, "step": 59420 }, { "grad_norm": 0.19104547798633575, "learning_rate": 3.864310050838972e-05, "loss": 0.0062, "step": 59430 }, { "grad_norm": 0.16807006299495697, "learning_rate": 3.8626998605768835e-05, "loss": 0.008, "step": 59440 }, { "grad_norm": 0.18975715339183807, "learning_rate": 3.8610897946882294e-05, "loss": 0.0071, "step": 59450 }, { "grad_norm": 0.11230595409870148, "learning_rate": 3.859479853349082e-05, "loss": 0.007, "step": 59460 }, { "grad_norm": 0.1841219961643219, "learning_rate": 3.857870036735503e-05, "loss": 0.0082, "step": 59470 }, { "grad_norm": 0.16473855078220367, "learning_rate": 3.856260345023538e-05, "loss": 0.0093, "step": 59480 }, { "grad_norm": 0.13314686715602875, "learning_rate": 3.854650778389224e-05, "loss": 0.0078, "step": 59490 }, { "grad_norm": 0.22599180042743683, "learning_rate": 3.8530413370085786e-05, "loss": 0.0079, "step": 59500 }, { "grad_norm": 0.11989755928516388, "learning_rate": 3.851432021057608e-05, "loss": 0.0073, "step": 59510 }, { "grad_norm": 0.2500419020652771, "learning_rate": 3.849822830712306e-05, "loss": 0.0085, "step": 59520 }, { "grad_norm": 0.2644507884979248, "learning_rate": 3.848213766148649e-05, "loss": 0.0091, "step": 59530 }, { "grad_norm": 0.1593526005744934, "learning_rate": 3.846604827542602e-05, "loss": 0.0074, "step": 59540 }, { "grad_norm": 0.15399909019470215, "learning_rate": 3.844996015070117e-05, "loss": 0.008, "step": 59550 }, { "grad_norm": 0.11073912680149078, "learning_rate": 3.843387328907132e-05, "loss": 0.008, "step": 59560 }, { "grad_norm": 0.21241506934165955, "learning_rate": 3.8417787692295696e-05, "loss": 0.0082, "step": 59570 }, { "grad_norm": 0.18432991206645966, "learning_rate": 3.84017033621334e-05, "loss": 0.008, "step": 59580 }, { "grad_norm": 0.21904762089252472, "learning_rate": 3.838562030034338e-05, "loss": 0.0081, "step": 59590 }, { "grad_norm": 0.1693149209022522, "learning_rate": 3.836953850868447e-05, "loss": 0.008, "step": 59600 }, { "grad_norm": 0.13366137444972992, "learning_rate": 3.835345798891533e-05, "loss": 0.008, "step": 59610 }, { "grad_norm": 0.14724422991275787, "learning_rate": 3.833737874279452e-05, "loss": 0.0085, "step": 59620 }, { "grad_norm": 0.2007935792207718, "learning_rate": 3.8321300772080433e-05, "loss": 0.0081, "step": 59630 }, { "grad_norm": 0.13635727763175964, "learning_rate": 3.8305224078531345e-05, "loss": 0.0068, "step": 59640 }, { "grad_norm": 0.23712904751300812, "learning_rate": 3.828914866390537e-05, "loss": 0.008, "step": 59650 }, { "grad_norm": 0.11257744580507278, "learning_rate": 3.827307452996048e-05, "loss": 0.0071, "step": 59660 }, { "grad_norm": 0.121807761490345, "learning_rate": 3.825700167845454e-05, "loss": 0.0078, "step": 59670 }, { "grad_norm": 0.14248496294021606, "learning_rate": 3.8240930111145214e-05, "loss": 0.0069, "step": 59680 }, { "grad_norm": 0.11797546595335007, "learning_rate": 3.8224859829790116e-05, "loss": 0.0067, "step": 59690 }, { "grad_norm": 0.1015324592590332, "learning_rate": 3.820879083614664e-05, "loss": 0.007, "step": 59700 }, { "grad_norm": 0.16398455202579498, "learning_rate": 3.819272313197208e-05, "loss": 0.0072, "step": 59710 }, { "grad_norm": 0.17479172348976135, "learning_rate": 3.817665671902355e-05, "loss": 0.0076, "step": 59720 }, { "grad_norm": 0.22879068553447723, "learning_rate": 3.816059159905808e-05, "loss": 0.0065, "step": 59730 }, { "grad_norm": 0.15876780450344086, "learning_rate": 3.8144527773832476e-05, "loss": 0.0077, "step": 59740 }, { "grad_norm": 0.12469470500946045, "learning_rate": 3.812846524510352e-05, "loss": 0.0069, "step": 59750 }, { "grad_norm": 0.20060470700263977, "learning_rate": 3.811240401462775e-05, "loss": 0.0081, "step": 59760 }, { "grad_norm": 0.13020265102386475, "learning_rate": 3.809634408416162e-05, "loss": 0.0088, "step": 59770 }, { "grad_norm": 0.15196992456912994, "learning_rate": 3.8080285455461394e-05, "loss": 0.0075, "step": 59780 }, { "grad_norm": 0.6169661283493042, "learning_rate": 3.806422813028323e-05, "loss": 0.008, "step": 59790 }, { "grad_norm": 0.18272987008094788, "learning_rate": 3.8048172110383114e-05, "loss": 0.0067, "step": 59800 }, { "grad_norm": 0.204068124294281, "learning_rate": 3.8032117397516944e-05, "loss": 0.0088, "step": 59810 }, { "grad_norm": 0.1949353814125061, "learning_rate": 3.801606399344042e-05, "loss": 0.0074, "step": 59820 }, { "grad_norm": 0.2214309275150299, "learning_rate": 3.800001189990913e-05, "loss": 0.0087, "step": 59830 }, { "grad_norm": 0.19814114272594452, "learning_rate": 3.798396111867847e-05, "loss": 0.0077, "step": 59840 }, { "grad_norm": 0.1734561324119568, "learning_rate": 3.796791165150378e-05, "loss": 0.0068, "step": 59850 }, { "grad_norm": 0.14415594935417175, "learning_rate": 3.795186350014018e-05, "loss": 0.0083, "step": 59860 }, { "grad_norm": 0.25100719928741455, "learning_rate": 3.793581666634266e-05, "loss": 0.0094, "step": 59870 }, { "grad_norm": 0.11723122745752335, "learning_rate": 3.791977115186609e-05, "loss": 0.0077, "step": 59880 }, { "grad_norm": 0.19686852395534515, "learning_rate": 3.79037269584652e-05, "loss": 0.0078, "step": 59890 }, { "grad_norm": 0.15391117334365845, "learning_rate": 3.7887684087894545e-05, "loss": 0.0087, "step": 59900 }, { "grad_norm": 0.14096565544605255, "learning_rate": 3.787164254190854e-05, "loss": 0.008, "step": 59910 }, { "grad_norm": 0.13589929044246674, "learning_rate": 3.7855602322261473e-05, "loss": 0.0078, "step": 59920 }, { "grad_norm": 0.29468855261802673, "learning_rate": 3.783956343070746e-05, "loss": 0.007, "step": 59930 }, { "grad_norm": 0.14141221344470978, "learning_rate": 3.7823525869000526e-05, "loss": 0.0071, "step": 59940 }, { "grad_norm": 0.15071754157543182, "learning_rate": 3.78074896388945e-05, "loss": 0.0077, "step": 59950 }, { "grad_norm": 0.1250554919242859, "learning_rate": 3.7791454742143075e-05, "loss": 0.0069, "step": 59960 }, { "grad_norm": 0.1322864145040512, "learning_rate": 3.7775421180499795e-05, "loss": 0.0092, "step": 59970 }, { "grad_norm": 0.1678038090467453, "learning_rate": 3.7759388955718086e-05, "loss": 0.0069, "step": 59980 }, { "grad_norm": 0.16703703999519348, "learning_rate": 3.774335806955117e-05, "loss": 0.007, "step": 59990 }, { "grad_norm": 0.11594940721988678, "learning_rate": 3.7727328523752215e-05, "loss": 0.0081, "step": 60000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }