{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.1, "eval_steps": 500, "global_step": 10000, "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 } ], "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 }