{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9, "eval_steps": 500, "global_step": 90000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.6883660554885864, "learning_rate": 1.8e-07, "loss": 1.2807, "step": 10 }, { "grad_norm": 0.7840365171432495, "learning_rate": 3.8e-07, "loss": 1.2965, "step": 20 }, { "grad_norm": 0.5540201663970947, "learning_rate": 5.8e-07, "loss": 1.2855, "step": 30 }, { "grad_norm": 0.523270845413208, "learning_rate": 7.8e-07, "loss": 1.2793, "step": 40 }, { "grad_norm": 0.5578494668006897, "learning_rate": 9.8e-07, "loss": 1.2747, "step": 50 }, { "grad_norm": 0.5746376514434814, "learning_rate": 1.18e-06, "loss": 1.2344, "step": 60 }, { "grad_norm": 0.9343597292900085, "learning_rate": 1.3800000000000001e-06, "loss": 1.2193, "step": 70 }, { "grad_norm": 0.8400384187698364, "learning_rate": 1.5800000000000003e-06, "loss": 1.2066, "step": 80 }, { "grad_norm": 0.6800181865692139, "learning_rate": 1.7800000000000001e-06, "loss": 1.2081, "step": 90 }, { "grad_norm": 2.150812864303589, "learning_rate": 1.98e-06, "loss": 1.1684, "step": 100 }, { "grad_norm": 0.8558348417282104, "learning_rate": 2.1800000000000003e-06, "loss": 1.1608, "step": 110 }, { "grad_norm": 1.0002028942108154, "learning_rate": 2.38e-06, "loss": 1.1539, "step": 120 }, { "grad_norm": 0.8516902923583984, "learning_rate": 2.5800000000000003e-06, "loss": 1.1161, "step": 130 }, { "grad_norm": 0.767479419708252, "learning_rate": 2.78e-06, "loss": 1.1079, "step": 140 }, { "grad_norm": 0.8377130627632141, "learning_rate": 2.9800000000000003e-06, "loss": 1.0936, "step": 150 }, { "grad_norm": 0.7560415267944336, "learning_rate": 3.1800000000000005e-06, "loss": 1.1188, "step": 160 }, { "grad_norm": 0.7722843289375305, "learning_rate": 3.38e-06, "loss": 1.1034, "step": 170 }, { "grad_norm": 0.6019498705863953, "learning_rate": 3.58e-06, "loss": 1.1039, "step": 180 }, { "grad_norm": 0.7488645911216736, "learning_rate": 3.7800000000000002e-06, "loss": 1.1033, "step": 190 }, { "grad_norm": 0.9592899084091187, "learning_rate": 3.98e-06, "loss": 1.1108, "step": 200 }, { "grad_norm": 0.846477746963501, "learning_rate": 4.18e-06, "loss": 1.0966, "step": 210 }, { "grad_norm": 0.7561697959899902, "learning_rate": 4.38e-06, "loss": 1.0819, "step": 220 }, { "grad_norm": 0.719254195690155, "learning_rate": 4.58e-06, "loss": 1.0838, "step": 230 }, { "grad_norm": 0.8098745942115784, "learning_rate": 4.780000000000001e-06, "loss": 1.0886, "step": 240 }, { "grad_norm": 0.8588639497756958, "learning_rate": 4.98e-06, "loss": 1.0779, "step": 250 }, { "grad_norm": 0.7727457880973816, "learning_rate": 5.18e-06, "loss": 1.1002, "step": 260 }, { "grad_norm": 1.1313961744308472, "learning_rate": 5.38e-06, "loss": 1.0856, "step": 270 }, { "grad_norm": 0.7794157862663269, "learning_rate": 5.580000000000001e-06, "loss": 1.0774, "step": 280 }, { "grad_norm": 0.745906412601471, "learning_rate": 5.78e-06, "loss": 1.0868, "step": 290 }, { "grad_norm": 0.7429569959640503, "learning_rate": 5.98e-06, "loss": 1.0864, "step": 300 }, { "grad_norm": 1.3276633024215698, "learning_rate": 6.18e-06, "loss": 1.0832, "step": 310 }, { "grad_norm": 1.0135293006896973, "learning_rate": 6.38e-06, "loss": 1.0578, "step": 320 }, { "grad_norm": 1.4053196907043457, "learning_rate": 6.58e-06, "loss": 1.0287, "step": 330 }, { "grad_norm": 1.2998108863830566, "learning_rate": 6.78e-06, "loss": 0.9905, "step": 340 }, { "grad_norm": 1.6040230989456177, "learning_rate": 6.98e-06, "loss": 0.975, "step": 350 }, { "grad_norm": 1.6671664714813232, "learning_rate": 7.180000000000001e-06, "loss": 0.9516, "step": 360 }, { "grad_norm": 1.8759191036224365, "learning_rate": 7.3800000000000005e-06, "loss": 0.9076, "step": 370 }, { "grad_norm": 1.3168582916259766, "learning_rate": 7.580000000000001e-06, "loss": 0.8728, "step": 380 }, { "grad_norm": 1.7959421873092651, "learning_rate": 7.78e-06, "loss": 0.8454, "step": 390 }, { "grad_norm": 1.3977402448654175, "learning_rate": 7.98e-06, "loss": 0.7903, "step": 400 }, { "grad_norm": 1.6007812023162842, "learning_rate": 8.18e-06, "loss": 0.7685, "step": 410 }, { "grad_norm": 1.5295259952545166, "learning_rate": 8.380000000000001e-06, "loss": 0.7387, "step": 420 }, { "grad_norm": 1.9129302501678467, "learning_rate": 8.580000000000001e-06, "loss": 0.7052, "step": 430 }, { "grad_norm": 2.222552537918091, "learning_rate": 8.78e-06, "loss": 0.6788, "step": 440 }, { "grad_norm": 2.0612330436706543, "learning_rate": 8.98e-06, "loss": 0.6286, "step": 450 }, { "grad_norm": 1.9947823286056519, "learning_rate": 9.180000000000002e-06, "loss": 0.6044, "step": 460 }, { "grad_norm": 2.384538173675537, "learning_rate": 9.38e-06, "loss": 0.5869, "step": 470 }, { "grad_norm": 3.648452043533325, "learning_rate": 9.58e-06, "loss": 0.5525, "step": 480 }, { "grad_norm": 2.17043399810791, "learning_rate": 9.78e-06, "loss": 0.5314, "step": 490 }, { "grad_norm": 2.177079916000366, "learning_rate": 9.980000000000001e-06, "loss": 0.4912, "step": 500 }, { "grad_norm": 2.804192543029785, "learning_rate": 1.018e-05, "loss": 0.4897, "step": 510 }, { "grad_norm": 2.1262621879577637, "learning_rate": 1.038e-05, "loss": 0.4446, "step": 520 }, { "grad_norm": 2.576927661895752, "learning_rate": 1.058e-05, "loss": 0.431, "step": 530 }, { "grad_norm": 1.8892502784729004, "learning_rate": 1.0780000000000002e-05, "loss": 0.4027, "step": 540 }, { "grad_norm": 2.6871118545532227, "learning_rate": 1.098e-05, "loss": 0.3948, "step": 550 }, { "grad_norm": 2.846674680709839, "learning_rate": 1.118e-05, "loss": 0.3606, "step": 560 }, { "grad_norm": 2.3887064456939697, "learning_rate": 1.1380000000000001e-05, "loss": 0.36, "step": 570 }, { "grad_norm": 2.330533027648926, "learning_rate": 1.1580000000000001e-05, "loss": 0.3386, "step": 580 }, { "grad_norm": 1.8280571699142456, "learning_rate": 1.178e-05, "loss": 0.3205, "step": 590 }, { "grad_norm": 2.612867832183838, "learning_rate": 1.198e-05, "loss": 0.3166, "step": 600 }, { "grad_norm": 2.6280078887939453, "learning_rate": 1.2180000000000002e-05, "loss": 0.2888, "step": 610 }, { "grad_norm": 2.428506851196289, "learning_rate": 1.238e-05, "loss": 0.2965, "step": 620 }, { "grad_norm": 2.6343424320220947, "learning_rate": 1.258e-05, "loss": 0.2455, "step": 630 }, { "grad_norm": 2.5483899116516113, "learning_rate": 1.278e-05, "loss": 0.2452, "step": 640 }, { "grad_norm": 2.6522040367126465, "learning_rate": 1.2980000000000001e-05, "loss": 0.2074, "step": 650 }, { "grad_norm": 2.2885582447052, "learning_rate": 1.3180000000000001e-05, "loss": 0.2148, "step": 660 }, { "grad_norm": 2.4678447246551514, "learning_rate": 1.338e-05, "loss": 0.215, "step": 670 }, { "grad_norm": 2.0842604637145996, "learning_rate": 1.358e-05, "loss": 0.1887, "step": 680 }, { "grad_norm": 2.689399480819702, "learning_rate": 1.3780000000000002e-05, "loss": 0.1821, "step": 690 }, { "grad_norm": 2.1430039405822754, "learning_rate": 1.3980000000000002e-05, "loss": 0.1891, "step": 700 }, { "grad_norm": 2.2432680130004883, "learning_rate": 1.4180000000000001e-05, "loss": 0.1841, "step": 710 }, { "grad_norm": 2.246701240539551, "learning_rate": 1.4380000000000001e-05, "loss": 0.1754, "step": 720 }, { "grad_norm": 2.6975557804107666, "learning_rate": 1.4580000000000003e-05, "loss": 0.1808, "step": 730 }, { "grad_norm": 1.9724483489990234, "learning_rate": 1.4779999999999999e-05, "loss": 0.1628, "step": 740 }, { "grad_norm": 2.75388765335083, "learning_rate": 1.4979999999999999e-05, "loss": 0.1705, "step": 750 }, { "grad_norm": 2.0861546993255615, "learning_rate": 1.518e-05, "loss": 0.1539, "step": 760 }, { "grad_norm": 2.0776045322418213, "learning_rate": 1.538e-05, "loss": 0.1616, "step": 770 }, { "grad_norm": 2.283966541290283, "learning_rate": 1.558e-05, "loss": 0.1598, "step": 780 }, { "grad_norm": 2.152937889099121, "learning_rate": 1.578e-05, "loss": 0.1575, "step": 790 }, { "grad_norm": 2.174529552459717, "learning_rate": 1.598e-05, "loss": 0.1593, "step": 800 }, { "grad_norm": 2.43585467338562, "learning_rate": 1.618e-05, "loss": 0.1826, "step": 810 }, { "grad_norm": 2.048943042755127, "learning_rate": 1.6380000000000002e-05, "loss": 0.1642, "step": 820 }, { "grad_norm": 1.8885246515274048, "learning_rate": 1.658e-05, "loss": 0.1516, "step": 830 }, { "grad_norm": 2.0217995643615723, "learning_rate": 1.6780000000000002e-05, "loss": 0.1384, "step": 840 }, { "grad_norm": 1.870848536491394, "learning_rate": 1.698e-05, "loss": 0.1377, "step": 850 }, { "grad_norm": 1.8226431608200073, "learning_rate": 1.718e-05, "loss": 0.1435, "step": 860 }, { "grad_norm": 2.1497371196746826, "learning_rate": 1.7380000000000003e-05, "loss": 0.1441, "step": 870 }, { "grad_norm": 1.8066797256469727, "learning_rate": 1.758e-05, "loss": 0.1378, "step": 880 }, { "grad_norm": 2.5279488563537598, "learning_rate": 1.7780000000000003e-05, "loss": 0.1516, "step": 890 }, { "grad_norm": 2.2115278244018555, "learning_rate": 1.798e-05, "loss": 0.1489, "step": 900 }, { "grad_norm": 2.2102160453796387, "learning_rate": 1.818e-05, "loss": 0.1552, "step": 910 }, { "grad_norm": 2.19504976272583, "learning_rate": 1.838e-05, "loss": 0.1319, "step": 920 }, { "grad_norm": 2.1513240337371826, "learning_rate": 1.858e-05, "loss": 0.1409, "step": 930 }, { "grad_norm": 1.7692351341247559, "learning_rate": 1.878e-05, "loss": 0.1479, "step": 940 }, { "grad_norm": 1.6555390357971191, "learning_rate": 1.898e-05, "loss": 0.1429, "step": 950 }, { "grad_norm": 1.4117289781570435, "learning_rate": 1.918e-05, "loss": 0.1205, "step": 960 }, { "grad_norm": 1.846531629562378, "learning_rate": 1.938e-05, "loss": 0.1369, "step": 970 }, { "grad_norm": 2.47581148147583, "learning_rate": 1.9580000000000002e-05, "loss": 0.1249, "step": 980 }, { "grad_norm": 1.708080530166626, "learning_rate": 1.978e-05, "loss": 0.1248, "step": 990 }, { "grad_norm": 1.5070749521255493, "learning_rate": 1.9980000000000002e-05, "loss": 0.1397, "step": 1000 }, { "grad_norm": 1.7875438928604126, "learning_rate": 2.0180000000000003e-05, "loss": 0.1199, "step": 1010 }, { "grad_norm": 1.9920581579208374, "learning_rate": 2.038e-05, "loss": 0.134, "step": 1020 }, { "grad_norm": 1.80613112449646, "learning_rate": 2.0580000000000003e-05, "loss": 0.126, "step": 1030 }, { "grad_norm": 1.8298254013061523, "learning_rate": 2.078e-05, "loss": 0.1196, "step": 1040 }, { "grad_norm": 1.6983835697174072, "learning_rate": 2.098e-05, "loss": 0.112, "step": 1050 }, { "grad_norm": 1.717942476272583, "learning_rate": 2.118e-05, "loss": 0.1245, "step": 1060 }, { "grad_norm": 1.6316741704940796, "learning_rate": 2.138e-05, "loss": 0.1234, "step": 1070 }, { "grad_norm": 1.9032847881317139, "learning_rate": 2.158e-05, "loss": 0.1299, "step": 1080 }, { "grad_norm": 2.245509624481201, "learning_rate": 2.178e-05, "loss": 0.1246, "step": 1090 }, { "grad_norm": 1.6945971250534058, "learning_rate": 2.198e-05, "loss": 0.1045, "step": 1100 }, { "grad_norm": 2.334597110748291, "learning_rate": 2.218e-05, "loss": 0.1317, "step": 1110 }, { "grad_norm": 1.4601740837097168, "learning_rate": 2.2380000000000003e-05, "loss": 0.1117, "step": 1120 }, { "grad_norm": 1.4437462091445923, "learning_rate": 2.258e-05, "loss": 0.1273, "step": 1130 }, { "grad_norm": 1.777411937713623, "learning_rate": 2.2780000000000002e-05, "loss": 0.1126, "step": 1140 }, { "grad_norm": 1.496631145477295, "learning_rate": 2.298e-05, "loss": 0.1202, "step": 1150 }, { "grad_norm": 1.5807781219482422, "learning_rate": 2.318e-05, "loss": 0.1139, "step": 1160 }, { "grad_norm": 1.3292713165283203, "learning_rate": 2.3380000000000003e-05, "loss": 0.1121, "step": 1170 }, { "grad_norm": 1.8693162202835083, "learning_rate": 2.358e-05, "loss": 0.1131, "step": 1180 }, { "grad_norm": 1.4570600986480713, "learning_rate": 2.3780000000000003e-05, "loss": 0.114, "step": 1190 }, { "grad_norm": 1.4116793870925903, "learning_rate": 2.398e-05, "loss": 0.127, "step": 1200 }, { "grad_norm": 1.578105092048645, "learning_rate": 2.418e-05, "loss": 0.1045, "step": 1210 }, { "grad_norm": 1.3879563808441162, "learning_rate": 2.438e-05, "loss": 0.1105, "step": 1220 }, { "grad_norm": 1.2454206943511963, "learning_rate": 2.4580000000000002e-05, "loss": 0.1098, "step": 1230 }, { "grad_norm": 1.6433202028274536, "learning_rate": 2.478e-05, "loss": 0.1115, "step": 1240 }, { "grad_norm": 1.5981742143630981, "learning_rate": 2.498e-05, "loss": 0.103, "step": 1250 }, { "grad_norm": 1.705410122871399, "learning_rate": 2.5180000000000003e-05, "loss": 0.1112, "step": 1260 }, { "grad_norm": 1.1877483129501343, "learning_rate": 2.5380000000000004e-05, "loss": 0.1061, "step": 1270 }, { "grad_norm": 1.2447994947433472, "learning_rate": 2.5580000000000002e-05, "loss": 0.0941, "step": 1280 }, { "grad_norm": 1.3369661569595337, "learning_rate": 2.5779999999999997e-05, "loss": 0.1013, "step": 1290 }, { "grad_norm": 1.4166675806045532, "learning_rate": 2.598e-05, "loss": 0.1002, "step": 1300 }, { "grad_norm": 1.512630820274353, "learning_rate": 2.618e-05, "loss": 0.1002, "step": 1310 }, { "grad_norm": 1.5484709739685059, "learning_rate": 2.6379999999999998e-05, "loss": 0.1143, "step": 1320 }, { "grad_norm": 1.6830568313598633, "learning_rate": 2.658e-05, "loss": 0.1069, "step": 1330 }, { "grad_norm": 1.267680287361145, "learning_rate": 2.678e-05, "loss": 0.0972, "step": 1340 }, { "grad_norm": 1.6173714399337769, "learning_rate": 2.698e-05, "loss": 0.0967, "step": 1350 }, { "grad_norm": 1.658255696296692, "learning_rate": 2.718e-05, "loss": 0.1082, "step": 1360 }, { "grad_norm": 1.399245262145996, "learning_rate": 2.738e-05, "loss": 0.0961, "step": 1370 }, { "grad_norm": 1.3702698945999146, "learning_rate": 2.758e-05, "loss": 0.1111, "step": 1380 }, { "grad_norm": 1.9542433023452759, "learning_rate": 2.778e-05, "loss": 0.1113, "step": 1390 }, { "grad_norm": 1.1670713424682617, "learning_rate": 2.798e-05, "loss": 0.1002, "step": 1400 }, { "grad_norm": 1.6531063318252563, "learning_rate": 2.818e-05, "loss": 0.1256, "step": 1410 }, { "grad_norm": 1.4461184740066528, "learning_rate": 2.8380000000000003e-05, "loss": 0.1169, "step": 1420 }, { "grad_norm": 1.3577343225479126, "learning_rate": 2.858e-05, "loss": 0.0995, "step": 1430 }, { "grad_norm": 1.3892253637313843, "learning_rate": 2.8780000000000002e-05, "loss": 0.1063, "step": 1440 }, { "grad_norm": 0.9181585311889648, "learning_rate": 2.898e-05, "loss": 0.0868, "step": 1450 }, { "grad_norm": 1.6972522735595703, "learning_rate": 2.9180000000000002e-05, "loss": 0.0947, "step": 1460 }, { "grad_norm": 1.6717156171798706, "learning_rate": 2.9380000000000003e-05, "loss": 0.1023, "step": 1470 }, { "grad_norm": 1.792961835861206, "learning_rate": 2.958e-05, "loss": 0.1116, "step": 1480 }, { "grad_norm": 1.2995253801345825, "learning_rate": 2.9780000000000003e-05, "loss": 0.103, "step": 1490 }, { "grad_norm": 1.6492317914962769, "learning_rate": 2.998e-05, "loss": 0.0999, "step": 1500 }, { "grad_norm": 1.514906883239746, "learning_rate": 3.0180000000000002e-05, "loss": 0.103, "step": 1510 }, { "grad_norm": 1.3722254037857056, "learning_rate": 3.0380000000000004e-05, "loss": 0.0997, "step": 1520 }, { "grad_norm": 1.075011134147644, "learning_rate": 3.058e-05, "loss": 0.0956, "step": 1530 }, { "grad_norm": 1.505475640296936, "learning_rate": 3.078e-05, "loss": 0.0884, "step": 1540 }, { "grad_norm": 1.1576002836227417, "learning_rate": 3.0980000000000005e-05, "loss": 0.1018, "step": 1550 }, { "grad_norm": 1.2101237773895264, "learning_rate": 3.118e-05, "loss": 0.0981, "step": 1560 }, { "grad_norm": 1.1030594110488892, "learning_rate": 3.138e-05, "loss": 0.0952, "step": 1570 }, { "grad_norm": 1.1988615989685059, "learning_rate": 3.1580000000000006e-05, "loss": 0.0963, "step": 1580 }, { "grad_norm": 1.5842784643173218, "learning_rate": 3.1780000000000004e-05, "loss": 0.1012, "step": 1590 }, { "grad_norm": 1.2311127185821533, "learning_rate": 3.198e-05, "loss": 0.1067, "step": 1600 }, { "grad_norm": 1.2537482976913452, "learning_rate": 3.218e-05, "loss": 0.0985, "step": 1610 }, { "grad_norm": 0.8948624730110168, "learning_rate": 3.238e-05, "loss": 0.1001, "step": 1620 }, { "grad_norm": 1.5607901811599731, "learning_rate": 3.2579999999999996e-05, "loss": 0.0948, "step": 1630 }, { "grad_norm": 1.3944425582885742, "learning_rate": 3.278e-05, "loss": 0.0953, "step": 1640 }, { "grad_norm": 1.4110301733016968, "learning_rate": 3.298e-05, "loss": 0.0911, "step": 1650 }, { "grad_norm": 1.6297261714935303, "learning_rate": 3.318e-05, "loss": 0.1019, "step": 1660 }, { "grad_norm": 1.2740572690963745, "learning_rate": 3.338e-05, "loss": 0.101, "step": 1670 }, { "grad_norm": 1.2544468641281128, "learning_rate": 3.358e-05, "loss": 0.0888, "step": 1680 }, { "grad_norm": 1.7073783874511719, "learning_rate": 3.378e-05, "loss": 0.1052, "step": 1690 }, { "grad_norm": 1.088757038116455, "learning_rate": 3.398e-05, "loss": 0.0948, "step": 1700 }, { "grad_norm": 1.2666457891464233, "learning_rate": 3.418e-05, "loss": 0.1071, "step": 1710 }, { "grad_norm": 1.2064040899276733, "learning_rate": 3.438e-05, "loss": 0.1105, "step": 1720 }, { "grad_norm": 1.4516526460647583, "learning_rate": 3.4580000000000004e-05, "loss": 0.0944, "step": 1730 }, { "grad_norm": 1.0957682132720947, "learning_rate": 3.478e-05, "loss": 0.0898, "step": 1740 }, { "grad_norm": 1.3366644382476807, "learning_rate": 3.498e-05, "loss": 0.0963, "step": 1750 }, { "grad_norm": 1.3166744709014893, "learning_rate": 3.518e-05, "loss": 0.0918, "step": 1760 }, { "grad_norm": 1.197719693183899, "learning_rate": 3.5380000000000003e-05, "loss": 0.0863, "step": 1770 }, { "grad_norm": 1.3439559936523438, "learning_rate": 3.558e-05, "loss": 0.0934, "step": 1780 }, { "grad_norm": 1.024295687675476, "learning_rate": 3.578e-05, "loss": 0.0833, "step": 1790 }, { "grad_norm": 0.9455082416534424, "learning_rate": 3.5980000000000004e-05, "loss": 0.08, "step": 1800 }, { "grad_norm": 1.2484533786773682, "learning_rate": 3.618e-05, "loss": 0.0869, "step": 1810 }, { "grad_norm": 0.7660257816314697, "learning_rate": 3.638e-05, "loss": 0.0926, "step": 1820 }, { "grad_norm": 0.9513797760009766, "learning_rate": 3.6580000000000006e-05, "loss": 0.0893, "step": 1830 }, { "grad_norm": 1.0217965841293335, "learning_rate": 3.6780000000000004e-05, "loss": 0.0866, "step": 1840 }, { "grad_norm": 1.061928391456604, "learning_rate": 3.698e-05, "loss": 0.0881, "step": 1850 }, { "grad_norm": 1.4948731660842896, "learning_rate": 3.7180000000000007e-05, "loss": 0.0972, "step": 1860 }, { "grad_norm": 0.9072536826133728, "learning_rate": 3.7380000000000005e-05, "loss": 0.0898, "step": 1870 }, { "grad_norm": 1.2617610692977905, "learning_rate": 3.758e-05, "loss": 0.0843, "step": 1880 }, { "grad_norm": 1.031685471534729, "learning_rate": 3.778000000000001e-05, "loss": 0.0963, "step": 1890 }, { "grad_norm": 0.8029703497886658, "learning_rate": 3.7980000000000006e-05, "loss": 0.0957, "step": 1900 }, { "grad_norm": 0.9904553294181824, "learning_rate": 3.818e-05, "loss": 0.0897, "step": 1910 }, { "grad_norm": 1.0429750680923462, "learning_rate": 3.838e-05, "loss": 0.0863, "step": 1920 }, { "grad_norm": 0.9095914959907532, "learning_rate": 3.858e-05, "loss": 0.0888, "step": 1930 }, { "grad_norm": 1.2869820594787598, "learning_rate": 3.878e-05, "loss": 0.0942, "step": 1940 }, { "grad_norm": 1.0976139307022095, "learning_rate": 3.898e-05, "loss": 0.0907, "step": 1950 }, { "grad_norm": 1.0262178182601929, "learning_rate": 3.918e-05, "loss": 0.0931, "step": 1960 }, { "grad_norm": 1.0651094913482666, "learning_rate": 3.938e-05, "loss": 0.1003, "step": 1970 }, { "grad_norm": 0.9744113683700562, "learning_rate": 3.958e-05, "loss": 0.0834, "step": 1980 }, { "grad_norm": 1.1709868907928467, "learning_rate": 3.978e-05, "loss": 0.0836, "step": 1990 }, { "grad_norm": 1.0895336866378784, "learning_rate": 3.998e-05, "loss": 0.1004, "step": 2000 }, { "grad_norm": 1.3938798904418945, "learning_rate": 4.018e-05, "loss": 0.0983, "step": 2010 }, { "grad_norm": 0.8506650924682617, "learning_rate": 4.038e-05, "loss": 0.0969, "step": 2020 }, { "grad_norm": 0.9700152277946472, "learning_rate": 4.058e-05, "loss": 0.0961, "step": 2030 }, { "grad_norm": 1.0039291381835938, "learning_rate": 4.078e-05, "loss": 0.0881, "step": 2040 }, { "grad_norm": 0.9314598441123962, "learning_rate": 4.0980000000000004e-05, "loss": 0.0922, "step": 2050 }, { "grad_norm": 0.9507551193237305, "learning_rate": 4.118e-05, "loss": 0.0945, "step": 2060 }, { "grad_norm": 0.9628653526306152, "learning_rate": 4.138e-05, "loss": 0.0896, "step": 2070 }, { "grad_norm": 1.103916883468628, "learning_rate": 4.1580000000000005e-05, "loss": 0.0822, "step": 2080 }, { "grad_norm": 0.8272781372070312, "learning_rate": 4.178e-05, "loss": 0.0854, "step": 2090 }, { "grad_norm": 0.7007940411567688, "learning_rate": 4.198e-05, "loss": 0.0837, "step": 2100 }, { "grad_norm": 0.7817788124084473, "learning_rate": 4.2180000000000006e-05, "loss": 0.0805, "step": 2110 }, { "grad_norm": 1.179325819015503, "learning_rate": 4.2380000000000004e-05, "loss": 0.091, "step": 2120 }, { "grad_norm": 0.9275873303413391, "learning_rate": 4.258e-05, "loss": 0.0878, "step": 2130 }, { "grad_norm": 2.008737802505493, "learning_rate": 4.278e-05, "loss": 0.0872, "step": 2140 }, { "grad_norm": 1.0677462816238403, "learning_rate": 4.2980000000000005e-05, "loss": 0.0854, "step": 2150 }, { "grad_norm": 0.9064395427703857, "learning_rate": 4.318e-05, "loss": 0.0847, "step": 2160 }, { "grad_norm": 0.9747989177703857, "learning_rate": 4.338e-05, "loss": 0.086, "step": 2170 }, { "grad_norm": 1.1643424034118652, "learning_rate": 4.3580000000000006e-05, "loss": 0.0897, "step": 2180 }, { "grad_norm": 1.0263400077819824, "learning_rate": 4.3780000000000004e-05, "loss": 0.0951, "step": 2190 }, { "grad_norm": 0.9893798828125, "learning_rate": 4.398e-05, "loss": 0.0955, "step": 2200 }, { "grad_norm": 0.7696988582611084, "learning_rate": 4.418000000000001e-05, "loss": 0.0758, "step": 2210 }, { "grad_norm": 0.7418758273124695, "learning_rate": 4.438e-05, "loss": 0.081, "step": 2220 }, { "grad_norm": 0.8966600894927979, "learning_rate": 4.458e-05, "loss": 0.0865, "step": 2230 }, { "grad_norm": 0.9534932374954224, "learning_rate": 4.478e-05, "loss": 0.0867, "step": 2240 }, { "grad_norm": 0.8894484639167786, "learning_rate": 4.498e-05, "loss": 0.089, "step": 2250 }, { "grad_norm": 0.9902400970458984, "learning_rate": 4.518e-05, "loss": 0.079, "step": 2260 }, { "grad_norm": 0.8538458347320557, "learning_rate": 4.538e-05, "loss": 0.0726, "step": 2270 }, { "grad_norm": 0.7368228435516357, "learning_rate": 4.558e-05, "loss": 0.0785, "step": 2280 }, { "grad_norm": 1.3523796796798706, "learning_rate": 4.578e-05, "loss": 0.0756, "step": 2290 }, { "grad_norm": 1.1000579595565796, "learning_rate": 4.5980000000000004e-05, "loss": 0.0881, "step": 2300 }, { "grad_norm": 0.974621593952179, "learning_rate": 4.618e-05, "loss": 0.0903, "step": 2310 }, { "grad_norm": 0.8171713948249817, "learning_rate": 4.638e-05, "loss": 0.0768, "step": 2320 }, { "grad_norm": 0.7753507494926453, "learning_rate": 4.6580000000000005e-05, "loss": 0.0773, "step": 2330 }, { "grad_norm": 0.8499396443367004, "learning_rate": 4.678e-05, "loss": 0.0809, "step": 2340 }, { "grad_norm": 0.6965720653533936, "learning_rate": 4.698e-05, "loss": 0.0818, "step": 2350 }, { "grad_norm": 0.5629959106445312, "learning_rate": 4.718e-05, "loss": 0.0805, "step": 2360 }, { "grad_norm": 0.8182156085968018, "learning_rate": 4.7380000000000004e-05, "loss": 0.0791, "step": 2370 }, { "grad_norm": 0.9200155138969421, "learning_rate": 4.758e-05, "loss": 0.0785, "step": 2380 }, { "grad_norm": 0.6897949576377869, "learning_rate": 4.778e-05, "loss": 0.0903, "step": 2390 }, { "grad_norm": 0.9367778897285461, "learning_rate": 4.7980000000000005e-05, "loss": 0.0797, "step": 2400 }, { "grad_norm": 1.0222256183624268, "learning_rate": 4.818e-05, "loss": 0.0854, "step": 2410 }, { "grad_norm": 0.7572758793830872, "learning_rate": 4.838e-05, "loss": 0.0849, "step": 2420 }, { "grad_norm": 0.8679062128067017, "learning_rate": 4.8580000000000006e-05, "loss": 0.0876, "step": 2430 }, { "grad_norm": 0.802072286605835, "learning_rate": 4.8780000000000004e-05, "loss": 0.0821, "step": 2440 }, { "grad_norm": 0.8160699009895325, "learning_rate": 4.898e-05, "loss": 0.0799, "step": 2450 }, { "grad_norm": 0.7841086387634277, "learning_rate": 4.918000000000001e-05, "loss": 0.0809, "step": 2460 }, { "grad_norm": 0.8275153636932373, "learning_rate": 4.9380000000000005e-05, "loss": 0.0759, "step": 2470 }, { "grad_norm": 0.8386090993881226, "learning_rate": 4.958e-05, "loss": 0.0847, "step": 2480 }, { "grad_norm": 0.805515706539154, "learning_rate": 4.978e-05, "loss": 0.0754, "step": 2490 }, { "grad_norm": 0.8537918329238892, "learning_rate": 4.9980000000000006e-05, "loss": 0.0744, "step": 2500 }, { "grad_norm": 0.7708742618560791, "learning_rate": 5.0180000000000004e-05, "loss": 0.0684, "step": 2510 }, { "grad_norm": 0.9133518934249878, "learning_rate": 5.038e-05, "loss": 0.0824, "step": 2520 }, { "grad_norm": 0.921848714351654, "learning_rate": 5.058000000000001e-05, "loss": 0.0986, "step": 2530 }, { "grad_norm": 0.6143094301223755, "learning_rate": 5.0780000000000005e-05, "loss": 0.0917, "step": 2540 }, { "grad_norm": 0.7443643808364868, "learning_rate": 5.098e-05, "loss": 0.0819, "step": 2550 }, { "grad_norm": 0.8640915751457214, "learning_rate": 5.118000000000001e-05, "loss": 0.0773, "step": 2560 }, { "grad_norm": 0.938015878200531, "learning_rate": 5.1380000000000006e-05, "loss": 0.0768, "step": 2570 }, { "grad_norm": 0.8717346787452698, "learning_rate": 5.1580000000000004e-05, "loss": 0.0843, "step": 2580 }, { "grad_norm": 0.6445345282554626, "learning_rate": 5.178000000000001e-05, "loss": 0.0739, "step": 2590 }, { "grad_norm": 0.7314874529838562, "learning_rate": 5.198000000000001e-05, "loss": 0.0766, "step": 2600 }, { "grad_norm": 0.953279435634613, "learning_rate": 5.2180000000000005e-05, "loss": 0.076, "step": 2610 }, { "grad_norm": 0.7633270621299744, "learning_rate": 5.238000000000001e-05, "loss": 0.0766, "step": 2620 }, { "grad_norm": 0.5823792219161987, "learning_rate": 5.258000000000001e-05, "loss": 0.0885, "step": 2630 }, { "grad_norm": 0.780953586101532, "learning_rate": 5.2780000000000006e-05, "loss": 0.0957, "step": 2640 }, { "grad_norm": 0.849938154220581, "learning_rate": 5.2980000000000004e-05, "loss": 0.0791, "step": 2650 }, { "grad_norm": 0.7651070952415466, "learning_rate": 5.318000000000001e-05, "loss": 0.0819, "step": 2660 }, { "grad_norm": 0.9543778896331787, "learning_rate": 5.338000000000001e-05, "loss": 0.0706, "step": 2670 }, { "grad_norm": 0.6410221457481384, "learning_rate": 5.3580000000000005e-05, "loss": 0.0712, "step": 2680 }, { "grad_norm": 0.7023550271987915, "learning_rate": 5.378e-05, "loss": 0.0702, "step": 2690 }, { "grad_norm": 0.8601503968238831, "learning_rate": 5.3979999999999995e-05, "loss": 0.0838, "step": 2700 }, { "grad_norm": 1.0795408487319946, "learning_rate": 5.418e-05, "loss": 0.0899, "step": 2710 }, { "grad_norm": 0.636965274810791, "learning_rate": 5.438e-05, "loss": 0.0849, "step": 2720 }, { "grad_norm": 0.9671086668968201, "learning_rate": 5.4579999999999996e-05, "loss": 0.0781, "step": 2730 }, { "grad_norm": 0.9170036911964417, "learning_rate": 5.478e-05, "loss": 0.077, "step": 2740 }, { "grad_norm": 0.7409165501594543, "learning_rate": 5.498e-05, "loss": 0.0773, "step": 2750 }, { "grad_norm": 0.8120000958442688, "learning_rate": 5.518e-05, "loss": 0.078, "step": 2760 }, { "grad_norm": 0.8688880205154419, "learning_rate": 5.538e-05, "loss": 0.0776, "step": 2770 }, { "grad_norm": 0.655924916267395, "learning_rate": 5.558e-05, "loss": 0.0707, "step": 2780 }, { "grad_norm": 0.5772544145584106, "learning_rate": 5.578e-05, "loss": 0.0676, "step": 2790 }, { "grad_norm": 0.8334255814552307, "learning_rate": 5.5979999999999996e-05, "loss": 0.0796, "step": 2800 }, { "grad_norm": 0.6535278558731079, "learning_rate": 5.618e-05, "loss": 0.0734, "step": 2810 }, { "grad_norm": 0.656789243221283, "learning_rate": 5.638e-05, "loss": 0.0762, "step": 2820 }, { "grad_norm": 0.790532648563385, "learning_rate": 5.658e-05, "loss": 0.0746, "step": 2830 }, { "grad_norm": 0.6961337924003601, "learning_rate": 5.678e-05, "loss": 0.0744, "step": 2840 }, { "grad_norm": 0.7104474902153015, "learning_rate": 5.698e-05, "loss": 0.0755, "step": 2850 }, { "grad_norm": 0.6908148527145386, "learning_rate": 5.718e-05, "loss": 0.086, "step": 2860 }, { "grad_norm": 0.7820292115211487, "learning_rate": 5.738e-05, "loss": 0.0658, "step": 2870 }, { "grad_norm": 0.6726694107055664, "learning_rate": 5.758e-05, "loss": 0.07, "step": 2880 }, { "grad_norm": 0.6976621747016907, "learning_rate": 5.778e-05, "loss": 0.0667, "step": 2890 }, { "grad_norm": 0.7265852689743042, "learning_rate": 5.7980000000000004e-05, "loss": 0.0665, "step": 2900 }, { "grad_norm": 0.6408970355987549, "learning_rate": 5.818e-05, "loss": 0.0808, "step": 2910 }, { "grad_norm": 0.9654882550239563, "learning_rate": 5.838e-05, "loss": 0.0766, "step": 2920 }, { "grad_norm": 0.7449786067008972, "learning_rate": 5.858e-05, "loss": 0.0807, "step": 2930 }, { "grad_norm": 0.6091578006744385, "learning_rate": 5.878e-05, "loss": 0.0702, "step": 2940 }, { "grad_norm": 0.768064022064209, "learning_rate": 5.898e-05, "loss": 0.0713, "step": 2950 }, { "grad_norm": 0.6159602999687195, "learning_rate": 5.918e-05, "loss": 0.073, "step": 2960 }, { "grad_norm": 0.6388112902641296, "learning_rate": 5.9380000000000004e-05, "loss": 0.0754, "step": 2970 }, { "grad_norm": 0.6002140045166016, "learning_rate": 5.958e-05, "loss": 0.0766, "step": 2980 }, { "grad_norm": 0.6340391039848328, "learning_rate": 5.978e-05, "loss": 0.0624, "step": 2990 }, { "grad_norm": 0.555962085723877, "learning_rate": 5.9980000000000005e-05, "loss": 0.078, "step": 3000 }, { "grad_norm": 0.8396109342575073, "learning_rate": 6.018e-05, "loss": 0.0744, "step": 3010 }, { "grad_norm": 0.6787938475608826, "learning_rate": 6.038e-05, "loss": 0.0736, "step": 3020 }, { "grad_norm": 0.7448351383209229, "learning_rate": 6.0580000000000006e-05, "loss": 0.0707, "step": 3030 }, { "grad_norm": 0.5366091132164001, "learning_rate": 6.0780000000000004e-05, "loss": 0.0745, "step": 3040 }, { "grad_norm": 0.7980797290802002, "learning_rate": 6.098e-05, "loss": 0.0741, "step": 3050 }, { "grad_norm": 0.7904975414276123, "learning_rate": 6.118000000000001e-05, "loss": 0.0817, "step": 3060 }, { "grad_norm": 0.6393426656723022, "learning_rate": 6.138e-05, "loss": 0.0881, "step": 3070 }, { "grad_norm": 0.6111093759536743, "learning_rate": 6.158e-05, "loss": 0.083, "step": 3080 }, { "grad_norm": 0.7289749383926392, "learning_rate": 6.178000000000001e-05, "loss": 0.0696, "step": 3090 }, { "grad_norm": 0.8339135646820068, "learning_rate": 6.198e-05, "loss": 0.0734, "step": 3100 }, { "grad_norm": 0.7398651838302612, "learning_rate": 6.218e-05, "loss": 0.0815, "step": 3110 }, { "grad_norm": 0.976265013217926, "learning_rate": 6.238000000000001e-05, "loss": 0.0732, "step": 3120 }, { "grad_norm": 0.7855477333068848, "learning_rate": 6.258e-05, "loss": 0.074, "step": 3130 }, { "grad_norm": 0.9038428068161011, "learning_rate": 6.278e-05, "loss": 0.0758, "step": 3140 }, { "grad_norm": 0.5442663431167603, "learning_rate": 6.298000000000001e-05, "loss": 0.0617, "step": 3150 }, { "grad_norm": 0.4420890510082245, "learning_rate": 6.318e-05, "loss": 0.0698, "step": 3160 }, { "grad_norm": 0.5229161381721497, "learning_rate": 6.338e-05, "loss": 0.0594, "step": 3170 }, { "grad_norm": 0.7068722248077393, "learning_rate": 6.358000000000001e-05, "loss": 0.0715, "step": 3180 }, { "grad_norm": 0.869797945022583, "learning_rate": 6.378e-05, "loss": 0.0766, "step": 3190 }, { "grad_norm": 0.555731475353241, "learning_rate": 6.398000000000001e-05, "loss": 0.066, "step": 3200 }, { "grad_norm": 0.6412601470947266, "learning_rate": 6.418000000000001e-05, "loss": 0.0672, "step": 3210 }, { "grad_norm": 0.6252126097679138, "learning_rate": 6.438e-05, "loss": 0.0632, "step": 3220 }, { "grad_norm": 0.6798176169395447, "learning_rate": 6.458000000000001e-05, "loss": 0.0658, "step": 3230 }, { "grad_norm": 0.7401267886161804, "learning_rate": 6.478000000000001e-05, "loss": 0.0631, "step": 3240 }, { "grad_norm": 0.49821287393569946, "learning_rate": 6.498e-05, "loss": 0.0629, "step": 3250 }, { "grad_norm": 0.6256396174430847, "learning_rate": 6.518000000000001e-05, "loss": 0.0639, "step": 3260 }, { "grad_norm": 0.7524405121803284, "learning_rate": 6.538000000000001e-05, "loss": 0.0622, "step": 3270 }, { "grad_norm": 0.6904860138893127, "learning_rate": 6.558e-05, "loss": 0.0783, "step": 3280 }, { "grad_norm": 0.6659765243530273, "learning_rate": 6.578000000000001e-05, "loss": 0.0653, "step": 3290 }, { "grad_norm": 0.7270122766494751, "learning_rate": 6.598e-05, "loss": 0.0606, "step": 3300 }, { "grad_norm": 0.8471141457557678, "learning_rate": 6.618e-05, "loss": 0.0817, "step": 3310 }, { "grad_norm": 0.5619258284568787, "learning_rate": 6.638e-05, "loss": 0.0701, "step": 3320 }, { "grad_norm": 0.6811439394950867, "learning_rate": 6.658e-05, "loss": 0.0744, "step": 3330 }, { "grad_norm": 0.7311329245567322, "learning_rate": 6.678e-05, "loss": 0.074, "step": 3340 }, { "grad_norm": 0.6086929440498352, "learning_rate": 6.698e-05, "loss": 0.0714, "step": 3350 }, { "grad_norm": 0.6338173151016235, "learning_rate": 6.718e-05, "loss": 0.0631, "step": 3360 }, { "grad_norm": 0.5708853602409363, "learning_rate": 6.738e-05, "loss": 0.0657, "step": 3370 }, { "grad_norm": 0.615634560585022, "learning_rate": 6.758e-05, "loss": 0.0705, "step": 3380 }, { "grad_norm": 0.6961477994918823, "learning_rate": 6.778e-05, "loss": 0.0723, "step": 3390 }, { "grad_norm": 0.5648407340049744, "learning_rate": 6.798e-05, "loss": 0.0752, "step": 3400 }, { "grad_norm": 0.6920296549797058, "learning_rate": 6.818e-05, "loss": 0.0652, "step": 3410 }, { "grad_norm": 0.7153940200805664, "learning_rate": 6.838e-05, "loss": 0.0624, "step": 3420 }, { "grad_norm": 0.5403403639793396, "learning_rate": 6.858e-05, "loss": 0.0747, "step": 3430 }, { "grad_norm": 0.5296555757522583, "learning_rate": 6.878e-05, "loss": 0.0675, "step": 3440 }, { "grad_norm": 0.640818178653717, "learning_rate": 6.898e-05, "loss": 0.0784, "step": 3450 }, { "grad_norm": 0.572515070438385, "learning_rate": 6.918e-05, "loss": 0.0636, "step": 3460 }, { "grad_norm": 0.595860481262207, "learning_rate": 6.938e-05, "loss": 0.0638, "step": 3470 }, { "grad_norm": 0.6020174622535706, "learning_rate": 6.958e-05, "loss": 0.0594, "step": 3480 }, { "grad_norm": 0.6687718033790588, "learning_rate": 6.978e-05, "loss": 0.067, "step": 3490 }, { "grad_norm": 0.6983329057693481, "learning_rate": 6.998e-05, "loss": 0.0557, "step": 3500 }, { "grad_norm": 0.5078064203262329, "learning_rate": 7.018e-05, "loss": 0.0757, "step": 3510 }, { "grad_norm": 0.5666691064834595, "learning_rate": 7.038e-05, "loss": 0.0617, "step": 3520 }, { "grad_norm": 0.8062061071395874, "learning_rate": 7.058e-05, "loss": 0.0658, "step": 3530 }, { "grad_norm": 0.45426586270332336, "learning_rate": 7.078e-05, "loss": 0.0851, "step": 3540 }, { "grad_norm": 0.6019713282585144, "learning_rate": 7.098e-05, "loss": 0.0733, "step": 3550 }, { "grad_norm": 0.5904924273490906, "learning_rate": 7.118e-05, "loss": 0.068, "step": 3560 }, { "grad_norm": 0.6614639759063721, "learning_rate": 7.138e-05, "loss": 0.0704, "step": 3570 }, { "grad_norm": 0.48388996720314026, "learning_rate": 7.158e-05, "loss": 0.0613, "step": 3580 }, { "grad_norm": 0.6672912240028381, "learning_rate": 7.178000000000001e-05, "loss": 0.0737, "step": 3590 }, { "grad_norm": 0.5049688816070557, "learning_rate": 7.198e-05, "loss": 0.0641, "step": 3600 }, { "grad_norm": 0.7039470076560974, "learning_rate": 7.218e-05, "loss": 0.0678, "step": 3610 }, { "grad_norm": 0.44673144817352295, "learning_rate": 7.238000000000001e-05, "loss": 0.0692, "step": 3620 }, { "grad_norm": 0.7434121966362, "learning_rate": 7.258e-05, "loss": 0.0704, "step": 3630 }, { "grad_norm": 0.6303529143333435, "learning_rate": 7.278e-05, "loss": 0.0629, "step": 3640 }, { "grad_norm": 0.47923150658607483, "learning_rate": 7.298000000000001e-05, "loss": 0.0703, "step": 3650 }, { "grad_norm": 0.7425010800361633, "learning_rate": 7.318e-05, "loss": 0.0641, "step": 3660 }, { "grad_norm": 0.4620913565158844, "learning_rate": 7.338e-05, "loss": 0.0652, "step": 3670 }, { "grad_norm": 0.5695753693580627, "learning_rate": 7.358000000000001e-05, "loss": 0.0657, "step": 3680 }, { "grad_norm": 0.4551267921924591, "learning_rate": 7.378e-05, "loss": 0.0684, "step": 3690 }, { "grad_norm": 0.9473399519920349, "learning_rate": 7.398e-05, "loss": 0.0772, "step": 3700 }, { "grad_norm": 0.6675644516944885, "learning_rate": 7.418000000000001e-05, "loss": 0.0817, "step": 3710 }, { "grad_norm": 0.432609498500824, "learning_rate": 7.438e-05, "loss": 0.0617, "step": 3720 }, { "grad_norm": 0.5736871957778931, "learning_rate": 7.458000000000001e-05, "loss": 0.0662, "step": 3730 }, { "grad_norm": 0.7466893792152405, "learning_rate": 7.478e-05, "loss": 0.071, "step": 3740 }, { "grad_norm": 0.5600967407226562, "learning_rate": 7.498e-05, "loss": 0.0699, "step": 3750 }, { "grad_norm": 0.5073946118354797, "learning_rate": 7.518000000000001e-05, "loss": 0.0687, "step": 3760 }, { "grad_norm": 0.6482805013656616, "learning_rate": 7.538e-05, "loss": 0.0702, "step": 3770 }, { "grad_norm": 0.6201328635215759, "learning_rate": 7.558e-05, "loss": 0.0695, "step": 3780 }, { "grad_norm": 0.6043463349342346, "learning_rate": 7.578000000000001e-05, "loss": 0.0759, "step": 3790 }, { "grad_norm": 0.6908556818962097, "learning_rate": 7.598e-05, "loss": 0.073, "step": 3800 }, { "grad_norm": 0.7615258693695068, "learning_rate": 7.618e-05, "loss": 0.0756, "step": 3810 }, { "grad_norm": 0.5845540761947632, "learning_rate": 7.638000000000001e-05, "loss": 0.0749, "step": 3820 }, { "grad_norm": 0.5575625896453857, "learning_rate": 7.658e-05, "loss": 0.0799, "step": 3830 }, { "grad_norm": 0.49917203187942505, "learning_rate": 7.678000000000001e-05, "loss": 0.0672, "step": 3840 }, { "grad_norm": 0.5564008355140686, "learning_rate": 7.698000000000001e-05, "loss": 0.0656, "step": 3850 }, { "grad_norm": 0.6430705189704895, "learning_rate": 7.718e-05, "loss": 0.0691, "step": 3860 }, { "grad_norm": 0.6341207027435303, "learning_rate": 7.738000000000001e-05, "loss": 0.0707, "step": 3870 }, { "grad_norm": 0.6521661281585693, "learning_rate": 7.758000000000001e-05, "loss": 0.0808, "step": 3880 }, { "grad_norm": 0.4554823338985443, "learning_rate": 7.778e-05, "loss": 0.0623, "step": 3890 }, { "grad_norm": 0.5013353228569031, "learning_rate": 7.798000000000001e-05, "loss": 0.0686, "step": 3900 }, { "grad_norm": 0.768687903881073, "learning_rate": 7.818000000000001e-05, "loss": 0.0646, "step": 3910 }, { "grad_norm": 0.4638596177101135, "learning_rate": 7.838e-05, "loss": 0.0666, "step": 3920 }, { "grad_norm": 0.6380972862243652, "learning_rate": 7.858000000000001e-05, "loss": 0.0693, "step": 3930 }, { "grad_norm": 0.511837363243103, "learning_rate": 7.878e-05, "loss": 0.0736, "step": 3940 }, { "grad_norm": 0.6213013529777527, "learning_rate": 7.897999999999999e-05, "loss": 0.077, "step": 3950 }, { "grad_norm": 0.560742199420929, "learning_rate": 7.918e-05, "loss": 0.0756, "step": 3960 }, { "grad_norm": 0.6428983211517334, "learning_rate": 7.938e-05, "loss": 0.0692, "step": 3970 }, { "grad_norm": 0.5820050835609436, "learning_rate": 7.958e-05, "loss": 0.0661, "step": 3980 }, { "grad_norm": 0.6279735565185547, "learning_rate": 7.978e-05, "loss": 0.0696, "step": 3990 }, { "grad_norm": 0.5338489413261414, "learning_rate": 7.998e-05, "loss": 0.0691, "step": 4000 }, { "grad_norm": 0.49248600006103516, "learning_rate": 8.018e-05, "loss": 0.062, "step": 4010 }, { "grad_norm": 0.6462330222129822, "learning_rate": 8.038e-05, "loss": 0.0698, "step": 4020 }, { "grad_norm": 0.5014451742172241, "learning_rate": 8.058e-05, "loss": 0.0721, "step": 4030 }, { "grad_norm": 0.5357964038848877, "learning_rate": 8.078e-05, "loss": 0.064, "step": 4040 }, { "grad_norm": 0.6113072037696838, "learning_rate": 8.098e-05, "loss": 0.0634, "step": 4050 }, { "grad_norm": 0.564304769039154, "learning_rate": 8.118e-05, "loss": 0.0631, "step": 4060 }, { "grad_norm": 0.5992838740348816, "learning_rate": 8.138e-05, "loss": 0.0719, "step": 4070 }, { "grad_norm": 0.5132911801338196, "learning_rate": 8.158e-05, "loss": 0.0612, "step": 4080 }, { "grad_norm": 0.47682350873947144, "learning_rate": 8.178e-05, "loss": 0.0719, "step": 4090 }, { "grad_norm": 0.9639641046524048, "learning_rate": 8.198e-05, "loss": 0.071, "step": 4100 }, { "grad_norm": 0.5708599090576172, "learning_rate": 8.218e-05, "loss": 0.0685, "step": 4110 }, { "grad_norm": 0.45716074109077454, "learning_rate": 8.238000000000001e-05, "loss": 0.0743, "step": 4120 }, { "grad_norm": 0.5290718674659729, "learning_rate": 8.258e-05, "loss": 0.0741, "step": 4130 }, { "grad_norm": 0.5355122089385986, "learning_rate": 8.278e-05, "loss": 0.0764, "step": 4140 }, { "grad_norm": 0.5317597389221191, "learning_rate": 8.298000000000001e-05, "loss": 0.0812, "step": 4150 }, { "grad_norm": 0.4867914319038391, "learning_rate": 8.318e-05, "loss": 0.0663, "step": 4160 }, { "grad_norm": 0.61493980884552, "learning_rate": 8.338e-05, "loss": 0.0668, "step": 4170 }, { "grad_norm": 0.5132406949996948, "learning_rate": 8.358e-05, "loss": 0.0677, "step": 4180 }, { "grad_norm": 0.5281847715377808, "learning_rate": 8.378e-05, "loss": 0.0662, "step": 4190 }, { "grad_norm": 0.519325852394104, "learning_rate": 8.398e-05, "loss": 0.0694, "step": 4200 }, { "grad_norm": 0.43491530418395996, "learning_rate": 8.418e-05, "loss": 0.0605, "step": 4210 }, { "grad_norm": 0.643584668636322, "learning_rate": 8.438e-05, "loss": 0.072, "step": 4220 }, { "grad_norm": 0.7019128799438477, "learning_rate": 8.458e-05, "loss": 0.0675, "step": 4230 }, { "grad_norm": 0.475957453250885, "learning_rate": 8.478e-05, "loss": 0.0668, "step": 4240 }, { "grad_norm": 0.5071917176246643, "learning_rate": 8.498e-05, "loss": 0.0622, "step": 4250 }, { "grad_norm": 0.515241801738739, "learning_rate": 8.518000000000001e-05, "loss": 0.0704, "step": 4260 }, { "grad_norm": 0.5725362300872803, "learning_rate": 8.538e-05, "loss": 0.0723, "step": 4270 }, { "grad_norm": 0.5570510625839233, "learning_rate": 8.558e-05, "loss": 0.0696, "step": 4280 }, { "grad_norm": 0.4370979070663452, "learning_rate": 8.578000000000001e-05, "loss": 0.0637, "step": 4290 }, { "grad_norm": 0.5142576694488525, "learning_rate": 8.598e-05, "loss": 0.0627, "step": 4300 }, { "grad_norm": 0.4969320297241211, "learning_rate": 8.618e-05, "loss": 0.0572, "step": 4310 }, { "grad_norm": 0.43779629468917847, "learning_rate": 8.638000000000001e-05, "loss": 0.058, "step": 4320 }, { "grad_norm": 0.40441465377807617, "learning_rate": 8.658e-05, "loss": 0.0564, "step": 4330 }, { "grad_norm": 0.5149533748626709, "learning_rate": 8.678e-05, "loss": 0.0773, "step": 4340 }, { "grad_norm": 0.504152774810791, "learning_rate": 8.698000000000001e-05, "loss": 0.065, "step": 4350 }, { "grad_norm": 0.4215155243873596, "learning_rate": 8.718e-05, "loss": 0.073, "step": 4360 }, { "grad_norm": 0.4108593761920929, "learning_rate": 8.738000000000001e-05, "loss": 0.0618, "step": 4370 }, { "grad_norm": 0.5483785271644592, "learning_rate": 8.758000000000001e-05, "loss": 0.0617, "step": 4380 }, { "grad_norm": 0.5111189484596252, "learning_rate": 8.778e-05, "loss": 0.0608, "step": 4390 }, { "grad_norm": 0.4640977680683136, "learning_rate": 8.798000000000001e-05, "loss": 0.06, "step": 4400 }, { "grad_norm": 0.4681876301765442, "learning_rate": 8.818000000000001e-05, "loss": 0.0529, "step": 4410 }, { "grad_norm": 0.4849129319190979, "learning_rate": 8.838e-05, "loss": 0.0648, "step": 4420 }, { "grad_norm": 0.6449043154716492, "learning_rate": 8.858000000000001e-05, "loss": 0.0573, "step": 4430 }, { "grad_norm": 0.44419485330581665, "learning_rate": 8.878000000000001e-05, "loss": 0.0596, "step": 4440 }, { "grad_norm": 0.7397124767303467, "learning_rate": 8.898e-05, "loss": 0.059, "step": 4450 }, { "grad_norm": 0.5494067072868347, "learning_rate": 8.918000000000001e-05, "loss": 0.0667, "step": 4460 }, { "grad_norm": 0.5438773036003113, "learning_rate": 8.938e-05, "loss": 0.0703, "step": 4470 }, { "grad_norm": 0.6030551791191101, "learning_rate": 8.958e-05, "loss": 0.0669, "step": 4480 }, { "grad_norm": 0.6974079012870789, "learning_rate": 8.978000000000001e-05, "loss": 0.0841, "step": 4490 }, { "grad_norm": 0.4405834376811981, "learning_rate": 8.998e-05, "loss": 0.0711, "step": 4500 }, { "grad_norm": 0.500303328037262, "learning_rate": 9.018000000000001e-05, "loss": 0.078, "step": 4510 }, { "grad_norm": 0.46409645676612854, "learning_rate": 9.038000000000001e-05, "loss": 0.0724, "step": 4520 }, { "grad_norm": 0.5208818912506104, "learning_rate": 9.058e-05, "loss": 0.0822, "step": 4530 }, { "grad_norm": 0.6934541463851929, "learning_rate": 9.078000000000001e-05, "loss": 0.094, "step": 4540 }, { "grad_norm": 0.7676739692687988, "learning_rate": 9.098000000000001e-05, "loss": 0.0888, "step": 4550 }, { "grad_norm": 0.4823222756385803, "learning_rate": 9.118e-05, "loss": 0.0773, "step": 4560 }, { "grad_norm": 0.4664647579193115, "learning_rate": 9.138e-05, "loss": 0.0687, "step": 4570 }, { "grad_norm": 0.5004900097846985, "learning_rate": 9.158e-05, "loss": 0.062, "step": 4580 }, { "grad_norm": 0.5535494089126587, "learning_rate": 9.178e-05, "loss": 0.071, "step": 4590 }, { "grad_norm": 0.5882132649421692, "learning_rate": 9.198e-05, "loss": 0.0703, "step": 4600 }, { "grad_norm": 0.64252108335495, "learning_rate": 9.218e-05, "loss": 0.0772, "step": 4610 }, { "grad_norm": 0.578276515007019, "learning_rate": 9.238e-05, "loss": 0.0691, "step": 4620 }, { "grad_norm": 0.7065967917442322, "learning_rate": 9.258e-05, "loss": 0.0736, "step": 4630 }, { "grad_norm": 0.6085888743400574, "learning_rate": 9.278e-05, "loss": 0.0715, "step": 4640 }, { "grad_norm": 0.4893440306186676, "learning_rate": 9.298e-05, "loss": 0.0708, "step": 4650 }, { "grad_norm": 0.5657037496566772, "learning_rate": 9.318e-05, "loss": 0.078, "step": 4660 }, { "grad_norm": 0.4288959503173828, "learning_rate": 9.338e-05, "loss": 0.066, "step": 4670 }, { "grad_norm": 0.49627625942230225, "learning_rate": 9.358e-05, "loss": 0.0721, "step": 4680 }, { "grad_norm": 0.4905998110771179, "learning_rate": 9.378e-05, "loss": 0.0741, "step": 4690 }, { "grad_norm": 0.5650030970573425, "learning_rate": 9.398e-05, "loss": 0.0694, "step": 4700 }, { "grad_norm": 0.4999682307243347, "learning_rate": 9.418e-05, "loss": 0.0705, "step": 4710 }, { "grad_norm": 0.41093185544013977, "learning_rate": 9.438e-05, "loss": 0.0721, "step": 4720 }, { "grad_norm": 0.44408902525901794, "learning_rate": 9.458e-05, "loss": 0.0735, "step": 4730 }, { "grad_norm": 0.47562336921691895, "learning_rate": 9.478e-05, "loss": 0.0769, "step": 4740 }, { "grad_norm": 0.48277947306632996, "learning_rate": 9.498e-05, "loss": 0.0748, "step": 4750 }, { "grad_norm": 0.43940356373786926, "learning_rate": 9.518000000000001e-05, "loss": 0.0692, "step": 4760 }, { "grad_norm": 0.49489086866378784, "learning_rate": 9.538e-05, "loss": 0.0631, "step": 4770 }, { "grad_norm": 0.42756232619285583, "learning_rate": 9.558e-05, "loss": 0.0627, "step": 4780 }, { "grad_norm": 0.44428911805152893, "learning_rate": 9.578000000000001e-05, "loss": 0.0642, "step": 4790 }, { "grad_norm": 0.6602418422698975, "learning_rate": 9.598e-05, "loss": 0.0665, "step": 4800 }, { "grad_norm": 0.4300300180912018, "learning_rate": 9.618e-05, "loss": 0.0606, "step": 4810 }, { "grad_norm": 0.5329577922821045, "learning_rate": 9.638000000000001e-05, "loss": 0.0591, "step": 4820 }, { "grad_norm": 0.5176931023597717, "learning_rate": 9.658e-05, "loss": 0.0592, "step": 4830 }, { "grad_norm": 0.5057064890861511, "learning_rate": 9.678e-05, "loss": 0.0615, "step": 4840 }, { "grad_norm": 0.43109312653541565, "learning_rate": 9.698000000000001e-05, "loss": 0.0637, "step": 4850 }, { "grad_norm": 0.6387642621994019, "learning_rate": 9.718e-05, "loss": 0.0639, "step": 4860 }, { "grad_norm": 0.3601652681827545, "learning_rate": 9.738e-05, "loss": 0.0652, "step": 4870 }, { "grad_norm": 0.4271051585674286, "learning_rate": 9.758000000000001e-05, "loss": 0.0607, "step": 4880 }, { "grad_norm": 0.5859048366546631, "learning_rate": 9.778e-05, "loss": 0.0713, "step": 4890 }, { "grad_norm": 0.49480822682380676, "learning_rate": 9.798000000000001e-05, "loss": 0.0671, "step": 4900 }, { "grad_norm": 0.5364655256271362, "learning_rate": 9.818000000000001e-05, "loss": 0.0717, "step": 4910 }, { "grad_norm": 0.5142687559127808, "learning_rate": 9.838e-05, "loss": 0.079, "step": 4920 }, { "grad_norm": 0.5951326489448547, "learning_rate": 9.858000000000001e-05, "loss": 0.0706, "step": 4930 }, { "grad_norm": 0.45892420411109924, "learning_rate": 9.878e-05, "loss": 0.0679, "step": 4940 }, { "grad_norm": 0.6323845386505127, "learning_rate": 9.898e-05, "loss": 0.0771, "step": 4950 }, { "grad_norm": 0.5309751629829407, "learning_rate": 9.918000000000001e-05, "loss": 0.0589, "step": 4960 }, { "grad_norm": 0.6187835931777954, "learning_rate": 9.938e-05, "loss": 0.059, "step": 4970 }, { "grad_norm": 0.5344617366790771, "learning_rate": 9.958e-05, "loss": 0.0669, "step": 4980 }, { "grad_norm": 0.363813579082489, "learning_rate": 9.978000000000001e-05, "loss": 0.0689, "step": 4990 }, { "grad_norm": 0.5071659088134766, "learning_rate": 9.998e-05, "loss": 0.0671, "step": 5000 }, { "grad_norm": 0.7433746457099915, "learning_rate": 9.999999778549045e-05, "loss": 0.0674, "step": 5010 }, { "grad_norm": 0.5526266098022461, "learning_rate": 9.999999013039593e-05, "loss": 0.0799, "step": 5020 }, { "grad_norm": 0.4484197795391083, "learning_rate": 9.999997700737766e-05, "loss": 0.073, "step": 5030 }, { "grad_norm": 0.4032785892486572, "learning_rate": 9.999995841643709e-05, "loss": 0.0653, "step": 5040 }, { "grad_norm": 0.542251467704773, "learning_rate": 9.999993435757623e-05, "loss": 0.0684, "step": 5050 }, { "grad_norm": 0.4255059063434601, "learning_rate": 9.999990483079773e-05, "loss": 0.0755, "step": 5060 }, { "grad_norm": 0.4811205565929413, "learning_rate": 9.999986983610481e-05, "loss": 0.0654, "step": 5070 }, { "grad_norm": 0.42558208107948303, "learning_rate": 9.99998293735013e-05, "loss": 0.0647, "step": 5080 }, { "grad_norm": 0.4216177761554718, "learning_rate": 9.999978344299161e-05, "loss": 0.0768, "step": 5090 }, { "grad_norm": 0.48470592498779297, "learning_rate": 9.99997320445808e-05, "loss": 0.0639, "step": 5100 }, { "grad_norm": 0.420318067073822, "learning_rate": 9.999967517827444e-05, "loss": 0.0604, "step": 5110 }, { "grad_norm": 0.38507339358329773, "learning_rate": 9.999961284407879e-05, "loss": 0.0655, "step": 5120 }, { "grad_norm": 0.4606524109840393, "learning_rate": 9.999954504200067e-05, "loss": 0.0621, "step": 5130 }, { "grad_norm": 0.5688247680664062, "learning_rate": 9.999947177204744e-05, "loss": 0.0645, "step": 5140 }, { "grad_norm": 0.41648170351982117, "learning_rate": 9.999939303422718e-05, "loss": 0.0779, "step": 5150 }, { "grad_norm": 0.47624894976615906, "learning_rate": 9.999930882854847e-05, "loss": 0.0591, "step": 5160 }, { "grad_norm": 0.45600005984306335, "learning_rate": 9.999921915502051e-05, "loss": 0.0665, "step": 5170 }, { "grad_norm": 0.37356141209602356, "learning_rate": 9.99991240136531e-05, "loss": 0.0509, "step": 5180 }, { "grad_norm": 0.5923223495483398, "learning_rate": 9.999902340445668e-05, "loss": 0.0614, "step": 5190 }, { "grad_norm": 0.44355371594429016, "learning_rate": 9.999891732744224e-05, "loss": 0.0635, "step": 5200 }, { "grad_norm": 0.44941359758377075, "learning_rate": 9.999880578262135e-05, "loss": 0.0638, "step": 5210 }, { "grad_norm": 0.4975666105747223, "learning_rate": 9.999868877000624e-05, "loss": 0.0669, "step": 5220 }, { "grad_norm": 0.40191012620925903, "learning_rate": 9.99985662896097e-05, "loss": 0.0626, "step": 5230 }, { "grad_norm": 0.38430285453796387, "learning_rate": 9.999843834144513e-05, "loss": 0.0634, "step": 5240 }, { "grad_norm": 0.5221481919288635, "learning_rate": 9.99983049255265e-05, "loss": 0.0751, "step": 5250 }, { "grad_norm": 0.47057437896728516, "learning_rate": 9.999816604186843e-05, "loss": 0.0674, "step": 5260 }, { "grad_norm": 0.7457453012466431, "learning_rate": 9.999802169048609e-05, "loss": 0.0674, "step": 5270 }, { "grad_norm": 0.36367449164390564, "learning_rate": 9.999787187139527e-05, "loss": 0.0597, "step": 5280 }, { "grad_norm": 0.44836220145225525, "learning_rate": 9.999771658461234e-05, "loss": 0.0673, "step": 5290 }, { "grad_norm": 0.5582446455955505, "learning_rate": 9.999755583015431e-05, "loss": 0.0684, "step": 5300 }, { "grad_norm": 0.502018928527832, "learning_rate": 9.999738960803874e-05, "loss": 0.0653, "step": 5310 }, { "grad_norm": 0.5206476449966431, "learning_rate": 9.99972179182838e-05, "loss": 0.0671, "step": 5320 }, { "grad_norm": 0.49442461133003235, "learning_rate": 9.99970407609083e-05, "loss": 0.0643, "step": 5330 }, { "grad_norm": 0.4062844514846802, "learning_rate": 9.999685813593159e-05, "loss": 0.0623, "step": 5340 }, { "grad_norm": 0.6672260761260986, "learning_rate": 9.999667004337362e-05, "loss": 0.0641, "step": 5350 }, { "grad_norm": 0.43625879287719727, "learning_rate": 9.9996476483255e-05, "loss": 0.0605, "step": 5360 }, { "grad_norm": 0.3523743450641632, "learning_rate": 9.999627745559688e-05, "loss": 0.0603, "step": 5370 }, { "grad_norm": 0.5320336222648621, "learning_rate": 9.999607296042101e-05, "loss": 0.0708, "step": 5380 }, { "grad_norm": 0.4860157370567322, "learning_rate": 9.99958629977498e-05, "loss": 0.0652, "step": 5390 }, { "grad_norm": 0.5468100905418396, "learning_rate": 9.999564756760615e-05, "loss": 0.0594, "step": 5400 }, { "grad_norm": 0.404110312461853, "learning_rate": 9.999542667001366e-05, "loss": 0.0693, "step": 5410 }, { "grad_norm": 0.396740585565567, "learning_rate": 9.999520030499647e-05, "loss": 0.065, "step": 5420 }, { "grad_norm": 0.49031761288642883, "learning_rate": 9.999496847257936e-05, "loss": 0.0566, "step": 5430 }, { "grad_norm": 0.5107786655426025, "learning_rate": 9.999473117278764e-05, "loss": 0.0664, "step": 5440 }, { "grad_norm": 0.3887660503387451, "learning_rate": 9.999448840564731e-05, "loss": 0.0631, "step": 5450 }, { "grad_norm": 0.3370972275733948, "learning_rate": 9.999424017118488e-05, "loss": 0.0701, "step": 5460 }, { "grad_norm": 0.43646812438964844, "learning_rate": 9.999398646942751e-05, "loss": 0.0712, "step": 5470 }, { "grad_norm": 0.4867601990699768, "learning_rate": 9.999372730040296e-05, "loss": 0.0648, "step": 5480 }, { "grad_norm": 0.4276811480522156, "learning_rate": 9.999346266413953e-05, "loss": 0.0659, "step": 5490 }, { "grad_norm": 0.4686550498008728, "learning_rate": 9.99931925606662e-05, "loss": 0.0734, "step": 5500 }, { "grad_norm": 0.4515635371208191, "learning_rate": 9.99929169900125e-05, "loss": 0.0654, "step": 5510 }, { "grad_norm": 0.5102119445800781, "learning_rate": 9.999263595220855e-05, "loss": 0.0741, "step": 5520 }, { "grad_norm": 0.5047770142555237, "learning_rate": 9.99923494472851e-05, "loss": 0.0718, "step": 5530 }, { "grad_norm": 0.4990573227405548, "learning_rate": 9.999205747527348e-05, "loss": 0.0679, "step": 5540 }, { "grad_norm": 0.5384724140167236, "learning_rate": 9.999176003620561e-05, "loss": 0.0744, "step": 5550 }, { "grad_norm": 0.5429059863090515, "learning_rate": 9.999145713011405e-05, "loss": 0.0796, "step": 5560 }, { "grad_norm": 0.41509172320365906, "learning_rate": 9.999114875703186e-05, "loss": 0.0696, "step": 5570 }, { "grad_norm": 0.48734670877456665, "learning_rate": 9.999083491699281e-05, "loss": 0.067, "step": 5580 }, { "grad_norm": 0.4351520538330078, "learning_rate": 9.999051561003123e-05, "loss": 0.0672, "step": 5590 }, { "grad_norm": 0.40793558955192566, "learning_rate": 9.999019083618202e-05, "loss": 0.0601, "step": 5600 }, { "grad_norm": 0.3729257583618164, "learning_rate": 9.99898605954807e-05, "loss": 0.0587, "step": 5610 }, { "grad_norm": 0.4855264127254486, "learning_rate": 9.998952488796338e-05, "loss": 0.0651, "step": 5620 }, { "grad_norm": 0.4377804398536682, "learning_rate": 9.998918371366676e-05, "loss": 0.0592, "step": 5630 }, { "grad_norm": 0.48748403787612915, "learning_rate": 9.99888370726282e-05, "loss": 0.0589, "step": 5640 }, { "grad_norm": 0.4366892874240875, "learning_rate": 9.998848496488556e-05, "loss": 0.0621, "step": 5650 }, { "grad_norm": 0.5554282665252686, "learning_rate": 9.998812739047736e-05, "loss": 0.0669, "step": 5660 }, { "grad_norm": 0.48844608664512634, "learning_rate": 9.99877643494427e-05, "loss": 0.0689, "step": 5670 }, { "grad_norm": 0.47503504157066345, "learning_rate": 9.998739584182128e-05, "loss": 0.0689, "step": 5680 }, { "grad_norm": 0.3648061156272888, "learning_rate": 9.998702186765342e-05, "loss": 0.0688, "step": 5690 }, { "grad_norm": 0.4206521213054657, "learning_rate": 9.998664242698e-05, "loss": 0.0644, "step": 5700 }, { "grad_norm": 0.3644725978374481, "learning_rate": 9.998625751984251e-05, "loss": 0.0613, "step": 5710 }, { "grad_norm": 0.5396430492401123, "learning_rate": 9.998586714628307e-05, "loss": 0.0632, "step": 5720 }, { "grad_norm": 0.4130953848361969, "learning_rate": 9.998547130634432e-05, "loss": 0.0616, "step": 5730 }, { "grad_norm": 0.46441417932510376, "learning_rate": 9.99850700000696e-05, "loss": 0.0633, "step": 5740 }, { "grad_norm": 0.48994627594947815, "learning_rate": 9.998466322750278e-05, "loss": 0.0609, "step": 5750 }, { "grad_norm": 0.4709641933441162, "learning_rate": 9.998425098868834e-05, "loss": 0.0619, "step": 5760 }, { "grad_norm": 0.3775669038295746, "learning_rate": 9.998383328367136e-05, "loss": 0.0627, "step": 5770 }, { "grad_norm": 0.46572601795196533, "learning_rate": 9.99834101124975e-05, "loss": 0.0595, "step": 5780 }, { "grad_norm": 0.4232090711593628, "learning_rate": 9.998298147521309e-05, "loss": 0.0769, "step": 5790 }, { "grad_norm": 0.46717292070388794, "learning_rate": 9.998254737186496e-05, "loss": 0.0776, "step": 5800 }, { "grad_norm": 0.4171079993247986, "learning_rate": 9.99821078025006e-05, "loss": 0.0692, "step": 5810 }, { "grad_norm": 0.374250203371048, "learning_rate": 9.998166276716807e-05, "loss": 0.0646, "step": 5820 }, { "grad_norm": 0.3935685455799103, "learning_rate": 9.998121226591606e-05, "loss": 0.0634, "step": 5830 }, { "grad_norm": 0.46879321336746216, "learning_rate": 9.998075629879382e-05, "loss": 0.0574, "step": 5840 }, { "grad_norm": 0.3449655771255493, "learning_rate": 9.99802948658512e-05, "loss": 0.0538, "step": 5850 }, { "grad_norm": 0.4461919963359833, "learning_rate": 9.99798279671387e-05, "loss": 0.0599, "step": 5860 }, { "grad_norm": 0.4163771867752075, "learning_rate": 9.997935560270734e-05, "loss": 0.0651, "step": 5870 }, { "grad_norm": 0.42708706855773926, "learning_rate": 9.997887777260879e-05, "loss": 0.063, "step": 5880 }, { "grad_norm": 0.3964042067527771, "learning_rate": 9.997839447689532e-05, "loss": 0.0652, "step": 5890 }, { "grad_norm": 0.44054168462753296, "learning_rate": 9.997790571561978e-05, "loss": 0.0696, "step": 5900 }, { "grad_norm": 0.3929017186164856, "learning_rate": 9.99774114888356e-05, "loss": 0.06, "step": 5910 }, { "grad_norm": 0.4784611165523529, "learning_rate": 9.997691179659684e-05, "loss": 0.0745, "step": 5920 }, { "grad_norm": 0.3414703607559204, "learning_rate": 9.997640663895815e-05, "loss": 0.0575, "step": 5930 }, { "grad_norm": 0.4983808994293213, "learning_rate": 9.997589601597477e-05, "loss": 0.0621, "step": 5940 }, { "grad_norm": 0.5199042558670044, "learning_rate": 9.997537992770252e-05, "loss": 0.067, "step": 5950 }, { "grad_norm": 0.34579986333847046, "learning_rate": 9.997485837419788e-05, "loss": 0.0691, "step": 5960 }, { "grad_norm": 0.49459174275398254, "learning_rate": 9.997433135551786e-05, "loss": 0.0663, "step": 5970 }, { "grad_norm": 0.42644089460372925, "learning_rate": 9.997379887172009e-05, "loss": 0.0624, "step": 5980 }, { "grad_norm": 0.43488648533821106, "learning_rate": 9.997326092286281e-05, "loss": 0.0623, "step": 5990 }, { "grad_norm": 0.4259776473045349, "learning_rate": 9.997271750900486e-05, "loss": 0.0629, "step": 6000 }, { "grad_norm": 0.3719702661037445, "learning_rate": 9.997216863020565e-05, "loss": 0.0646, "step": 6010 }, { "grad_norm": 0.3895108103752136, "learning_rate": 9.99716142865252e-05, "loss": 0.0695, "step": 6020 }, { "grad_norm": 0.48355576395988464, "learning_rate": 9.997105447802415e-05, "loss": 0.0632, "step": 6030 }, { "grad_norm": 0.48012393712997437, "learning_rate": 9.997048920476373e-05, "loss": 0.061, "step": 6040 }, { "grad_norm": 0.38356640934944153, "learning_rate": 9.996991846680572e-05, "loss": 0.0618, "step": 6050 }, { "grad_norm": 0.44333696365356445, "learning_rate": 9.996934226421257e-05, "loss": 0.0541, "step": 6060 }, { "grad_norm": 0.4967551529407501, "learning_rate": 9.996876059704726e-05, "loss": 0.0579, "step": 6070 }, { "grad_norm": 0.3985315263271332, "learning_rate": 9.996817346537343e-05, "loss": 0.0558, "step": 6080 }, { "grad_norm": 0.3293595016002655, "learning_rate": 9.996758086925526e-05, "loss": 0.0572, "step": 6090 }, { "grad_norm": 0.3672974109649658, "learning_rate": 9.996698280875759e-05, "loss": 0.0544, "step": 6100 }, { "grad_norm": 0.3916005790233612, "learning_rate": 9.99663792839458e-05, "loss": 0.0623, "step": 6110 }, { "grad_norm": 0.35481002926826477, "learning_rate": 9.99657702948859e-05, "loss": 0.0566, "step": 6120 }, { "grad_norm": 0.5451825261116028, "learning_rate": 9.996515584164448e-05, "loss": 0.0602, "step": 6130 }, { "grad_norm": 0.49415236711502075, "learning_rate": 9.996453592428873e-05, "loss": 0.0725, "step": 6140 }, { "grad_norm": 0.42698681354522705, "learning_rate": 9.996391054288646e-05, "loss": 0.0689, "step": 6150 }, { "grad_norm": 0.3919808268547058, "learning_rate": 9.996327969750605e-05, "loss": 0.0633, "step": 6160 }, { "grad_norm": 0.4782186448574066, "learning_rate": 9.996264338821649e-05, "loss": 0.06, "step": 6170 }, { "grad_norm": 0.46997299790382385, "learning_rate": 9.996200161508735e-05, "loss": 0.069, "step": 6180 }, { "grad_norm": 0.4253512918949127, "learning_rate": 9.996135437818885e-05, "loss": 0.0682, "step": 6190 }, { "grad_norm": 0.553621768951416, "learning_rate": 9.996070167759175e-05, "loss": 0.065, "step": 6200 }, { "grad_norm": 0.591406524181366, "learning_rate": 9.996004351336743e-05, "loss": 0.0709, "step": 6210 }, { "grad_norm": 0.4181950092315674, "learning_rate": 9.995937988558785e-05, "loss": 0.0627, "step": 6220 }, { "grad_norm": 0.36624106764793396, "learning_rate": 9.995871079432561e-05, "loss": 0.0571, "step": 6230 }, { "grad_norm": 0.3646184504032135, "learning_rate": 9.995803623965389e-05, "loss": 0.0548, "step": 6240 }, { "grad_norm": 0.31358960270881653, "learning_rate": 9.995735622164641e-05, "loss": 0.0588, "step": 6250 }, { "grad_norm": 0.4624374806880951, "learning_rate": 9.995667074037758e-05, "loss": 0.0634, "step": 6260 }, { "grad_norm": 0.37206241488456726, "learning_rate": 9.995597979592232e-05, "loss": 0.0506, "step": 6270 }, { "grad_norm": 0.4733327329158783, "learning_rate": 9.995528338835625e-05, "loss": 0.0601, "step": 6280 }, { "grad_norm": 0.40108367800712585, "learning_rate": 9.995458151775547e-05, "loss": 0.0566, "step": 6290 }, { "grad_norm": 0.35214388370513916, "learning_rate": 9.995387418419677e-05, "loss": 0.0582, "step": 6300 }, { "grad_norm": 0.3422820270061493, "learning_rate": 9.99531613877575e-05, "loss": 0.0578, "step": 6310 }, { "grad_norm": 0.3432181775569916, "learning_rate": 9.995244312851559e-05, "loss": 0.0554, "step": 6320 }, { "grad_norm": 0.3794559836387634, "learning_rate": 9.995171940654961e-05, "loss": 0.0587, "step": 6330 }, { "grad_norm": 0.4090096354484558, "learning_rate": 9.995099022193871e-05, "loss": 0.0563, "step": 6340 }, { "grad_norm": 0.4796614348888397, "learning_rate": 9.995025557476261e-05, "loss": 0.0583, "step": 6350 }, { "grad_norm": 0.3762759268283844, "learning_rate": 9.994951546510165e-05, "loss": 0.0625, "step": 6360 }, { "grad_norm": 0.4528583586215973, "learning_rate": 9.994876989303679e-05, "loss": 0.0725, "step": 6370 }, { "grad_norm": 0.3811642825603485, "learning_rate": 9.994801885864955e-05, "loss": 0.0609, "step": 6380 }, { "grad_norm": 0.46243947744369507, "learning_rate": 9.994726236202205e-05, "loss": 0.0567, "step": 6390 }, { "grad_norm": 0.3588138520717621, "learning_rate": 9.994650040323704e-05, "loss": 0.0619, "step": 6400 }, { "grad_norm": 0.3962024450302124, "learning_rate": 9.994573298237784e-05, "loss": 0.0581, "step": 6410 }, { "grad_norm": 0.4531124234199524, "learning_rate": 9.994496009952837e-05, "loss": 0.0571, "step": 6420 }, { "grad_norm": 0.3978465795516968, "learning_rate": 9.994418175477316e-05, "loss": 0.0601, "step": 6430 }, { "grad_norm": 0.3600148856639862, "learning_rate": 9.994339794819733e-05, "loss": 0.059, "step": 6440 }, { "grad_norm": 0.8855752348899841, "learning_rate": 9.994260867988658e-05, "loss": 0.0675, "step": 6450 }, { "grad_norm": 0.42935046553611755, "learning_rate": 9.994181394992723e-05, "loss": 0.054, "step": 6460 }, { "grad_norm": 0.579720139503479, "learning_rate": 9.994101375840618e-05, "loss": 0.0582, "step": 6470 }, { "grad_norm": 0.3431868851184845, "learning_rate": 9.994020810541098e-05, "loss": 0.0689, "step": 6480 }, { "grad_norm": 0.34909307956695557, "learning_rate": 9.99393969910297e-05, "loss": 0.0567, "step": 6490 }, { "grad_norm": 0.3394046425819397, "learning_rate": 9.993858041535104e-05, "loss": 0.0571, "step": 6500 }, { "grad_norm": 0.3802911937236786, "learning_rate": 9.99377583784643e-05, "loss": 0.0528, "step": 6510 }, { "grad_norm": 0.4149894714355469, "learning_rate": 9.993693088045939e-05, "loss": 0.0699, "step": 6520 }, { "grad_norm": 0.44902995228767395, "learning_rate": 9.99360979214268e-05, "loss": 0.0655, "step": 6530 }, { "grad_norm": 0.30978062748908997, "learning_rate": 9.99352595014576e-05, "loss": 0.0649, "step": 6540 }, { "grad_norm": 0.3697481155395508, "learning_rate": 9.993441562064354e-05, "loss": 0.058, "step": 6550 }, { "grad_norm": 0.2785720229148865, "learning_rate": 9.993356627907685e-05, "loss": 0.0591, "step": 6560 }, { "grad_norm": 0.3733918070793152, "learning_rate": 9.99327114768504e-05, "loss": 0.0615, "step": 6570 }, { "grad_norm": 0.38373103737831116, "learning_rate": 9.99318512140577e-05, "loss": 0.0549, "step": 6580 }, { "grad_norm": 0.36461296677589417, "learning_rate": 9.993098549079284e-05, "loss": 0.0518, "step": 6590 }, { "grad_norm": 0.34305140376091003, "learning_rate": 9.993011430715047e-05, "loss": 0.0584, "step": 6600 }, { "grad_norm": 0.3500927984714508, "learning_rate": 9.992923766322586e-05, "loss": 0.0512, "step": 6610 }, { "grad_norm": 0.4042293429374695, "learning_rate": 9.99283555591149e-05, "loss": 0.0613, "step": 6620 }, { "grad_norm": 0.4766307473182678, "learning_rate": 9.992746799491404e-05, "loss": 0.0542, "step": 6630 }, { "grad_norm": 0.4741242229938507, "learning_rate": 9.992657497072033e-05, "loss": 0.0651, "step": 6640 }, { "grad_norm": 0.7445639967918396, "learning_rate": 9.992567648663147e-05, "loss": 0.0535, "step": 6650 }, { "grad_norm": 0.43455207347869873, "learning_rate": 9.992477254274568e-05, "loss": 0.0606, "step": 6660 }, { "grad_norm": 0.3437443673610687, "learning_rate": 9.992386313916183e-05, "loss": 0.0603, "step": 6670 }, { "grad_norm": 0.4037517309188843, "learning_rate": 9.992294827597934e-05, "loss": 0.0594, "step": 6680 }, { "grad_norm": 0.462674081325531, "learning_rate": 9.992202795329831e-05, "loss": 0.0557, "step": 6690 }, { "grad_norm": 0.41224080324172974, "learning_rate": 9.992110217121936e-05, "loss": 0.0604, "step": 6700 }, { "grad_norm": 0.40204811096191406, "learning_rate": 9.992017092984372e-05, "loss": 0.0555, "step": 6710 }, { "grad_norm": 0.3218127489089966, "learning_rate": 9.991923422927326e-05, "loss": 0.0561, "step": 6720 }, { "grad_norm": 0.3597697615623474, "learning_rate": 9.991829206961037e-05, "loss": 0.0582, "step": 6730 }, { "grad_norm": 0.43440213799476624, "learning_rate": 9.991734445095813e-05, "loss": 0.0633, "step": 6740 }, { "grad_norm": 0.5111159682273865, "learning_rate": 9.991639137342015e-05, "loss": 0.074, "step": 6750 }, { "grad_norm": 0.49428901076316833, "learning_rate": 9.991543283710064e-05, "loss": 0.0651, "step": 6760 }, { "grad_norm": 0.34428641200065613, "learning_rate": 9.991446884210445e-05, "loss": 0.0662, "step": 6770 }, { "grad_norm": 0.29645782709121704, "learning_rate": 9.9913499388537e-05, "loss": 0.0577, "step": 6780 }, { "grad_norm": 0.39383918046951294, "learning_rate": 9.99125244765043e-05, "loss": 0.0598, "step": 6790 }, { "grad_norm": 0.406437486410141, "learning_rate": 9.991154410611296e-05, "loss": 0.0581, "step": 6800 }, { "grad_norm": 0.42210108041763306, "learning_rate": 9.99105582774702e-05, "loss": 0.0516, "step": 6810 }, { "grad_norm": 0.4119117558002472, "learning_rate": 9.990956699068384e-05, "loss": 0.0683, "step": 6820 }, { "grad_norm": 0.4770869314670563, "learning_rate": 9.990857024586224e-05, "loss": 0.0576, "step": 6830 }, { "grad_norm": 0.3954780101776123, "learning_rate": 9.990756804311446e-05, "loss": 0.0638, "step": 6840 }, { "grad_norm": 0.3692779541015625, "learning_rate": 9.990656038255006e-05, "loss": 0.0524, "step": 6850 }, { "grad_norm": 0.3105061650276184, "learning_rate": 9.990554726427926e-05, "loss": 0.0544, "step": 6860 }, { "grad_norm": 0.4270382225513458, "learning_rate": 9.990452868841284e-05, "loss": 0.0573, "step": 6870 }, { "grad_norm": 0.37302887439727783, "learning_rate": 9.99035046550622e-05, "loss": 0.0534, "step": 6880 }, { "grad_norm": 0.37935730814933777, "learning_rate": 9.99024751643393e-05, "loss": 0.0492, "step": 6890 }, { "grad_norm": 0.378818541765213, "learning_rate": 9.990144021635677e-05, "loss": 0.0586, "step": 6900 }, { "grad_norm": 0.3110908269882202, "learning_rate": 9.990039981122775e-05, "loss": 0.0559, "step": 6910 }, { "grad_norm": 0.4492669999599457, "learning_rate": 9.989935394906602e-05, "loss": 0.0589, "step": 6920 }, { "grad_norm": 0.33923205733299255, "learning_rate": 9.989830262998598e-05, "loss": 0.0559, "step": 6930 }, { "grad_norm": 0.4125036299228668, "learning_rate": 9.989724585410259e-05, "loss": 0.0563, "step": 6940 }, { "grad_norm": 0.32761749625205994, "learning_rate": 9.989618362153139e-05, "loss": 0.0548, "step": 6950 }, { "grad_norm": 0.34355035424232483, "learning_rate": 9.989511593238859e-05, "loss": 0.0524, "step": 6960 }, { "grad_norm": 0.33454227447509766, "learning_rate": 9.98940427867909e-05, "loss": 0.0549, "step": 6970 }, { "grad_norm": 0.3721087574958801, "learning_rate": 9.989296418485573e-05, "loss": 0.0563, "step": 6980 }, { "grad_norm": 0.34017688035964966, "learning_rate": 9.989188012670101e-05, "loss": 0.056, "step": 6990 }, { "grad_norm": 0.39309507608413696, "learning_rate": 9.989079061244528e-05, "loss": 0.0538, "step": 7000 }, { "grad_norm": 0.47921064496040344, "learning_rate": 9.988969564220769e-05, "loss": 0.0596, "step": 7010 }, { "grad_norm": 0.4388445019721985, "learning_rate": 9.988859521610801e-05, "loss": 0.0595, "step": 7020 }, { "grad_norm": 0.39496031403541565, "learning_rate": 9.988748933426656e-05, "loss": 0.0562, "step": 7030 }, { "grad_norm": 0.406279057264328, "learning_rate": 9.988637799680428e-05, "loss": 0.0595, "step": 7040 }, { "grad_norm": 0.27131420373916626, "learning_rate": 9.98852612038427e-05, "loss": 0.0566, "step": 7050 }, { "grad_norm": 0.38236838579177856, "learning_rate": 9.988413895550397e-05, "loss": 0.0597, "step": 7060 }, { "grad_norm": 0.41894471645355225, "learning_rate": 9.98830112519108e-05, "loss": 0.0614, "step": 7070 }, { "grad_norm": 0.45220106840133667, "learning_rate": 9.98818780931865e-05, "loss": 0.0523, "step": 7080 }, { "grad_norm": 0.4364688992500305, "learning_rate": 9.988073947945502e-05, "loss": 0.0574, "step": 7090 }, { "grad_norm": 0.42433053255081177, "learning_rate": 9.987959541084087e-05, "loss": 0.0538, "step": 7100 }, { "grad_norm": 0.4135838747024536, "learning_rate": 9.987844588746915e-05, "loss": 0.0587, "step": 7110 }, { "grad_norm": 0.39177340269088745, "learning_rate": 9.987729090946558e-05, "loss": 0.0531, "step": 7120 }, { "grad_norm": 0.3479836583137512, "learning_rate": 9.987613047695647e-05, "loss": 0.0561, "step": 7130 }, { "grad_norm": 0.33882686495780945, "learning_rate": 9.987496459006871e-05, "loss": 0.0606, "step": 7140 }, { "grad_norm": 0.439231812953949, "learning_rate": 9.987379324892982e-05, "loss": 0.0549, "step": 7150 }, { "grad_norm": 0.322394460439682, "learning_rate": 9.987261645366788e-05, "loss": 0.0522, "step": 7160 }, { "grad_norm": 0.2706003189086914, "learning_rate": 9.987143420441158e-05, "loss": 0.0581, "step": 7170 }, { "grad_norm": 0.48039641976356506, "learning_rate": 9.987024650129022e-05, "loss": 0.059, "step": 7180 }, { "grad_norm": 0.4062601327896118, "learning_rate": 9.986905334443368e-05, "loss": 0.0542, "step": 7190 }, { "grad_norm": 0.4603733718395233, "learning_rate": 9.986785473397245e-05, "loss": 0.0548, "step": 7200 }, { "grad_norm": 0.35424864292144775, "learning_rate": 9.98666506700376e-05, "loss": 0.0549, "step": 7210 }, { "grad_norm": 0.36616429686546326, "learning_rate": 9.986544115276081e-05, "loss": 0.0488, "step": 7220 }, { "grad_norm": 0.4265422224998474, "learning_rate": 9.986422618227433e-05, "loss": 0.0545, "step": 7230 }, { "grad_norm": 0.3361310064792633, "learning_rate": 9.986300575871106e-05, "loss": 0.0553, "step": 7240 }, { "grad_norm": 0.4105847179889679, "learning_rate": 9.986177988220444e-05, "loss": 0.0523, "step": 7250 }, { "grad_norm": 0.3426525294780731, "learning_rate": 9.986054855288856e-05, "loss": 0.0605, "step": 7260 }, { "grad_norm": 0.34440311789512634, "learning_rate": 9.985931177089802e-05, "loss": 0.0512, "step": 7270 }, { "grad_norm": 0.31142204999923706, "learning_rate": 9.985806953636814e-05, "loss": 0.0475, "step": 7280 }, { "grad_norm": 0.4149320125579834, "learning_rate": 9.985682184943471e-05, "loss": 0.0571, "step": 7290 }, { "grad_norm": 1.2899657487869263, "learning_rate": 9.98555687102342e-05, "loss": 0.0482, "step": 7300 }, { "grad_norm": 0.4656522870063782, "learning_rate": 9.985431011890367e-05, "loss": 0.0517, "step": 7310 }, { "grad_norm": 0.4416208863258362, "learning_rate": 9.985304607558075e-05, "loss": 0.0568, "step": 7320 }, { "grad_norm": 0.36493465304374695, "learning_rate": 9.985177658040364e-05, "loss": 0.0539, "step": 7330 }, { "grad_norm": 0.38540276885032654, "learning_rate": 9.985050163351119e-05, "loss": 0.0514, "step": 7340 }, { "grad_norm": 0.30485573410987854, "learning_rate": 9.984922123504286e-05, "loss": 0.0539, "step": 7350 }, { "grad_norm": 0.2798737585544586, "learning_rate": 9.984793538513862e-05, "loss": 0.0484, "step": 7360 }, { "grad_norm": 0.39406710863113403, "learning_rate": 9.984664408393912e-05, "loss": 0.0567, "step": 7370 }, { "grad_norm": 0.36748915910720825, "learning_rate": 9.984534733158556e-05, "loss": 0.056, "step": 7380 }, { "grad_norm": 0.3232600688934326, "learning_rate": 9.984404512821977e-05, "loss": 0.0498, "step": 7390 }, { "grad_norm": 0.28360792994499207, "learning_rate": 9.984273747398411e-05, "loss": 0.0482, "step": 7400 }, { "grad_norm": 0.33140769600868225, "learning_rate": 9.984142436902165e-05, "loss": 0.0461, "step": 7410 }, { "grad_norm": 0.42815154790878296, "learning_rate": 9.984010581347596e-05, "loss": 0.0536, "step": 7420 }, { "grad_norm": 0.4171295166015625, "learning_rate": 9.983878180749121e-05, "loss": 0.0466, "step": 7430 }, { "grad_norm": 0.3419502079486847, "learning_rate": 9.983745235121222e-05, "loss": 0.0475, "step": 7440 }, { "grad_norm": 0.2729214131832123, "learning_rate": 9.983611744478438e-05, "loss": 0.0476, "step": 7450 }, { "grad_norm": 0.35257792472839355, "learning_rate": 9.983477708835365e-05, "loss": 0.0508, "step": 7460 }, { "grad_norm": 0.3224290907382965, "learning_rate": 9.983343128206664e-05, "loss": 0.0567, "step": 7470 }, { "grad_norm": 0.35674312710762024, "learning_rate": 9.983208002607049e-05, "loss": 0.0532, "step": 7480 }, { "grad_norm": 0.3463895618915558, "learning_rate": 9.9830723320513e-05, "loss": 0.0475, "step": 7490 }, { "grad_norm": 0.2987387776374817, "learning_rate": 9.982936116554254e-05, "loss": 0.0539, "step": 7500 }, { "grad_norm": 0.37285444140434265, "learning_rate": 9.982799356130803e-05, "loss": 0.0575, "step": 7510 }, { "grad_norm": 0.5343098044395447, "learning_rate": 9.982662050795908e-05, "loss": 0.0648, "step": 7520 }, { "grad_norm": 0.39690229296684265, "learning_rate": 9.982524200564583e-05, "loss": 0.0568, "step": 7530 }, { "grad_norm": 0.3560376465320587, "learning_rate": 9.982385805451901e-05, "loss": 0.0547, "step": 7540 }, { "grad_norm": 0.4284445643424988, "learning_rate": 9.982246865472998e-05, "loss": 0.0583, "step": 7550 }, { "grad_norm": 0.43841999769210815, "learning_rate": 9.982107380643069e-05, "loss": 0.0551, "step": 7560 }, { "grad_norm": 0.3131433427333832, "learning_rate": 9.981967350977368e-05, "loss": 0.059, "step": 7570 }, { "grad_norm": 0.4279925227165222, "learning_rate": 9.981826776491208e-05, "loss": 0.0615, "step": 7580 }, { "grad_norm": 0.416048526763916, "learning_rate": 9.98168565719996e-05, "loss": 0.0669, "step": 7590 }, { "grad_norm": 0.2891758978366852, "learning_rate": 9.98154399311906e-05, "loss": 0.0526, "step": 7600 }, { "grad_norm": 0.40039491653442383, "learning_rate": 9.981401784263997e-05, "loss": 0.0649, "step": 7610 }, { "grad_norm": 0.36547237634658813, "learning_rate": 9.981259030650326e-05, "loss": 0.0639, "step": 7620 }, { "grad_norm": 0.3870498538017273, "learning_rate": 9.981115732293655e-05, "loss": 0.0583, "step": 7630 }, { "grad_norm": 0.5157169103622437, "learning_rate": 9.980971889209659e-05, "loss": 0.0617, "step": 7640 }, { "grad_norm": 0.4022206962108612, "learning_rate": 9.980827501414064e-05, "loss": 0.0566, "step": 7650 }, { "grad_norm": 0.3635427951812744, "learning_rate": 9.980682568922663e-05, "loss": 0.0581, "step": 7660 }, { "grad_norm": 0.31630444526672363, "learning_rate": 9.980537091751304e-05, "loss": 0.0487, "step": 7670 }, { "grad_norm": 0.2441459596157074, "learning_rate": 9.980391069915897e-05, "loss": 0.0528, "step": 7680 }, { "grad_norm": 0.35991498827934265, "learning_rate": 9.98024450343241e-05, "loss": 0.0515, "step": 7690 }, { "grad_norm": 0.4928025007247925, "learning_rate": 9.980097392316872e-05, "loss": 0.0558, "step": 7700 }, { "grad_norm": 0.33414018154144287, "learning_rate": 9.97994973658537e-05, "loss": 0.0585, "step": 7710 }, { "grad_norm": 0.35812029242515564, "learning_rate": 9.979801536254054e-05, "loss": 0.0588, "step": 7720 }, { "grad_norm": 0.33718398213386536, "learning_rate": 9.979652791339127e-05, "loss": 0.0504, "step": 7730 }, { "grad_norm": 0.32940882444381714, "learning_rate": 9.97950350185686e-05, "loss": 0.0482, "step": 7740 }, { "grad_norm": 0.36531832814216614, "learning_rate": 9.979353667823574e-05, "loss": 0.0548, "step": 7750 }, { "grad_norm": 0.5853579640388489, "learning_rate": 9.979203289255658e-05, "loss": 0.064, "step": 7760 }, { "grad_norm": 0.37784668803215027, "learning_rate": 9.979052366169557e-05, "loss": 0.0748, "step": 7770 }, { "grad_norm": 0.33992260694503784, "learning_rate": 9.978900898581775e-05, "loss": 0.0602, "step": 7780 }, { "grad_norm": 0.3749878704547882, "learning_rate": 9.978748886508875e-05, "loss": 0.0593, "step": 7790 }, { "grad_norm": 0.5218790173530579, "learning_rate": 9.978596329967484e-05, "loss": 0.0665, "step": 7800 }, { "grad_norm": 0.3788638412952423, "learning_rate": 9.978443228974284e-05, "loss": 0.0582, "step": 7810 }, { "grad_norm": 0.5433095693588257, "learning_rate": 9.978289583546015e-05, "loss": 0.0541, "step": 7820 }, { "grad_norm": 0.47182050347328186, "learning_rate": 9.978135393699484e-05, "loss": 0.0625, "step": 7830 }, { "grad_norm": 0.3749289810657501, "learning_rate": 9.977980659451548e-05, "loss": 0.0727, "step": 7840 }, { "grad_norm": 0.40313857793807983, "learning_rate": 9.977825380819135e-05, "loss": 0.0554, "step": 7850 }, { "grad_norm": 0.36464276909828186, "learning_rate": 9.97766955781922e-05, "loss": 0.0504, "step": 7860 }, { "grad_norm": 0.27353474497795105, "learning_rate": 9.977513190468848e-05, "loss": 0.0566, "step": 7870 }, { "grad_norm": 0.2997104227542877, "learning_rate": 9.977356278785116e-05, "loss": 0.0565, "step": 7880 }, { "grad_norm": 0.4946114420890808, "learning_rate": 9.977198822785184e-05, "loss": 0.0646, "step": 7890 }, { "grad_norm": 0.34463584423065186, "learning_rate": 9.977040822486273e-05, "loss": 0.0592, "step": 7900 }, { "grad_norm": 0.3039094805717468, "learning_rate": 9.97688227790566e-05, "loss": 0.0524, "step": 7910 }, { "grad_norm": 0.3805629014968872, "learning_rate": 9.976723189060684e-05, "loss": 0.0553, "step": 7920 }, { "grad_norm": 0.3448832929134369, "learning_rate": 9.976563555968742e-05, "loss": 0.0548, "step": 7930 }, { "grad_norm": 0.3965652883052826, "learning_rate": 9.976403378647292e-05, "loss": 0.0523, "step": 7940 }, { "grad_norm": 0.35010239481925964, "learning_rate": 9.97624265711385e-05, "loss": 0.0529, "step": 7950 }, { "grad_norm": 0.36718225479125977, "learning_rate": 9.976081391385993e-05, "loss": 0.0543, "step": 7960 }, { "grad_norm": 0.3709736168384552, "learning_rate": 9.975919581481356e-05, "loss": 0.0523, "step": 7970 }, { "grad_norm": 0.3098863661289215, "learning_rate": 9.975757227417634e-05, "loss": 0.0492, "step": 7980 }, { "grad_norm": 0.3448545038700104, "learning_rate": 9.975594329212586e-05, "loss": 0.0494, "step": 7990 }, { "grad_norm": 0.4070127308368683, "learning_rate": 9.97543088688402e-05, "loss": 0.0513, "step": 8000 }, { "grad_norm": 0.34667056798934937, "learning_rate": 9.975266900449814e-05, "loss": 0.045, "step": 8010 }, { "grad_norm": 0.4677402377128601, "learning_rate": 9.975102369927898e-05, "loss": 0.0518, "step": 8020 }, { "grad_norm": 0.358427494764328, "learning_rate": 9.974937295336269e-05, "loss": 0.0498, "step": 8030 }, { "grad_norm": 0.3357219696044922, "learning_rate": 9.974771676692975e-05, "loss": 0.0519, "step": 8040 }, { "grad_norm": 0.4464075565338135, "learning_rate": 9.974605514016131e-05, "loss": 0.0678, "step": 8050 }, { "grad_norm": 0.484467089176178, "learning_rate": 9.974438807323907e-05, "loss": 0.0553, "step": 8060 }, { "grad_norm": 0.34758779406547546, "learning_rate": 9.974271556634535e-05, "loss": 0.0556, "step": 8070 }, { "grad_norm": 0.34623315930366516, "learning_rate": 9.974103761966302e-05, "loss": 0.0531, "step": 8080 }, { "grad_norm": 0.33251115679740906, "learning_rate": 9.973935423337563e-05, "loss": 0.0482, "step": 8090 }, { "grad_norm": 0.37532564997673035, "learning_rate": 9.973766540766722e-05, "loss": 0.0541, "step": 8100 }, { "grad_norm": 0.3610958755016327, "learning_rate": 9.97359711427225e-05, "loss": 0.0468, "step": 8110 }, { "grad_norm": 0.27671071887016296, "learning_rate": 9.973427143872677e-05, "loss": 0.0516, "step": 8120 }, { "grad_norm": 0.26188406348228455, "learning_rate": 9.973256629586589e-05, "loss": 0.0419, "step": 8130 }, { "grad_norm": 0.25826677680015564, "learning_rate": 9.973085571432632e-05, "loss": 0.045, "step": 8140 }, { "grad_norm": 0.3591673970222473, "learning_rate": 9.972913969429513e-05, "loss": 0.0456, "step": 8150 }, { "grad_norm": 0.32289543747901917, "learning_rate": 9.972741823596e-05, "loss": 0.0448, "step": 8160 }, { "grad_norm": 0.3078395128250122, "learning_rate": 9.972569133950917e-05, "loss": 0.0477, "step": 8170 }, { "grad_norm": 0.42553237080574036, "learning_rate": 9.972395900513151e-05, "loss": 0.053, "step": 8180 }, { "grad_norm": 0.33506447076797485, "learning_rate": 9.972222123301645e-05, "loss": 0.047, "step": 8190 }, { "grad_norm": 0.23795413970947266, "learning_rate": 9.972047802335403e-05, "loss": 0.0469, "step": 8200 }, { "grad_norm": 0.3333989083766937, "learning_rate": 9.971872937633488e-05, "loss": 0.0551, "step": 8210 }, { "grad_norm": 0.3709401488304138, "learning_rate": 9.971697529215024e-05, "loss": 0.0507, "step": 8220 }, { "grad_norm": 0.38273072242736816, "learning_rate": 9.971521577099192e-05, "loss": 0.0601, "step": 8230 }, { "grad_norm": 0.5433246493339539, "learning_rate": 9.971345081305236e-05, "loss": 0.0591, "step": 8240 }, { "grad_norm": 0.5127394199371338, "learning_rate": 9.971168041852456e-05, "loss": 0.0611, "step": 8250 }, { "grad_norm": 0.4188228249549866, "learning_rate": 9.970990458760215e-05, "loss": 0.0538, "step": 8260 }, { "grad_norm": 0.3494628369808197, "learning_rate": 9.970812332047929e-05, "loss": 0.0527, "step": 8270 }, { "grad_norm": 0.3187815546989441, "learning_rate": 9.97063366173508e-05, "loss": 0.051, "step": 8280 }, { "grad_norm": 0.38027262687683105, "learning_rate": 9.970454447841207e-05, "loss": 0.0462, "step": 8290 }, { "grad_norm": 0.41543835401535034, "learning_rate": 9.970274690385909e-05, "loss": 0.0511, "step": 8300 }, { "grad_norm": 0.3262440860271454, "learning_rate": 9.970094389388844e-05, "loss": 0.0434, "step": 8310 }, { "grad_norm": 0.5659809112548828, "learning_rate": 9.969913544869728e-05, "loss": 0.0535, "step": 8320 }, { "grad_norm": 0.3349798023700714, "learning_rate": 9.96973215684834e-05, "loss": 0.0563, "step": 8330 }, { "grad_norm": 0.438371479511261, "learning_rate": 9.969550225344513e-05, "loss": 0.0564, "step": 8340 }, { "grad_norm": 0.321286678314209, "learning_rate": 9.969367750378147e-05, "loss": 0.0509, "step": 8350 }, { "grad_norm": 0.5098044276237488, "learning_rate": 9.969184731969194e-05, "loss": 0.0607, "step": 8360 }, { "grad_norm": 0.5104397535324097, "learning_rate": 9.96900117013767e-05, "loss": 0.0625, "step": 8370 }, { "grad_norm": 0.3383916914463043, "learning_rate": 9.96881706490365e-05, "loss": 0.0583, "step": 8380 }, { "grad_norm": 0.32280632853507996, "learning_rate": 9.968632416287265e-05, "loss": 0.0506, "step": 8390 }, { "grad_norm": 0.378194123506546, "learning_rate": 9.96844722430871e-05, "loss": 0.055, "step": 8400 }, { "grad_norm": 0.4178676903247833, "learning_rate": 9.968261488988235e-05, "loss": 0.0566, "step": 8410 }, { "grad_norm": 0.32087114453315735, "learning_rate": 9.968075210346155e-05, "loss": 0.0494, "step": 8420 }, { "grad_norm": 0.3402141034603119, "learning_rate": 9.967888388402839e-05, "loss": 0.0591, "step": 8430 }, { "grad_norm": 0.4512805938720703, "learning_rate": 9.967701023178717e-05, "loss": 0.0611, "step": 8440 }, { "grad_norm": 0.36385104060173035, "learning_rate": 9.967513114694282e-05, "loss": 0.0539, "step": 8450 }, { "grad_norm": 0.40217453241348267, "learning_rate": 9.967324662970079e-05, "loss": 0.0543, "step": 8460 }, { "grad_norm": 0.4210166931152344, "learning_rate": 9.96713566802672e-05, "loss": 0.0548, "step": 8470 }, { "grad_norm": 0.3895285725593567, "learning_rate": 9.966946129884873e-05, "loss": 0.0557, "step": 8480 }, { "grad_norm": 0.6160279512405396, "learning_rate": 9.966756048565265e-05, "loss": 0.0652, "step": 8490 }, { "grad_norm": 0.4039120078086853, "learning_rate": 9.966565424088681e-05, "loss": 0.0608, "step": 8500 }, { "grad_norm": 0.34931081533432007, "learning_rate": 9.96637425647597e-05, "loss": 0.0595, "step": 8510 }, { "grad_norm": 0.4621411859989166, "learning_rate": 9.966182545748038e-05, "loss": 0.0582, "step": 8520 }, { "grad_norm": 0.37147819995880127, "learning_rate": 9.96599029192585e-05, "loss": 0.0563, "step": 8530 }, { "grad_norm": 0.31718915700912476, "learning_rate": 9.965797495030428e-05, "loss": 0.0552, "step": 8540 }, { "grad_norm": 0.4140954315662384, "learning_rate": 9.96560415508286e-05, "loss": 0.0488, "step": 8550 }, { "grad_norm": 0.44554436206817627, "learning_rate": 9.965410272104286e-05, "loss": 0.0472, "step": 8560 }, { "grad_norm": 0.30938759446144104, "learning_rate": 9.96521584611591e-05, "loss": 0.0436, "step": 8570 }, { "grad_norm": 0.3950437605381012, "learning_rate": 9.965020877138994e-05, "loss": 0.0493, "step": 8580 }, { "grad_norm": 0.7412221431732178, "learning_rate": 9.964825365194861e-05, "loss": 0.0514, "step": 8590 }, { "grad_norm": 0.3285205066204071, "learning_rate": 9.96462931030489e-05, "loss": 0.0554, "step": 8600 }, { "grad_norm": 0.31424564123153687, "learning_rate": 9.96443271249052e-05, "loss": 0.0467, "step": 8610 }, { "grad_norm": 0.4056779742240906, "learning_rate": 9.964235571773255e-05, "loss": 0.0476, "step": 8620 }, { "grad_norm": 0.3013033866882324, "learning_rate": 9.96403788817465e-05, "loss": 0.0586, "step": 8630 }, { "grad_norm": 0.43724364042282104, "learning_rate": 9.963839661716325e-05, "loss": 0.0531, "step": 8640 }, { "grad_norm": 0.4216685891151428, "learning_rate": 9.963640892419958e-05, "loss": 0.047, "step": 8650 }, { "grad_norm": 0.4104737937450409, "learning_rate": 9.963441580307286e-05, "loss": 0.0549, "step": 8660 }, { "grad_norm": 0.4537009298801422, "learning_rate": 9.963241725400104e-05, "loss": 0.0531, "step": 8670 }, { "grad_norm": 0.40510764718055725, "learning_rate": 9.963041327720271e-05, "loss": 0.0543, "step": 8680 }, { "grad_norm": 0.39879047870635986, "learning_rate": 9.962840387289697e-05, "loss": 0.0574, "step": 8690 }, { "grad_norm": 0.3116472363471985, "learning_rate": 9.962638904130363e-05, "loss": 0.0562, "step": 8700 }, { "grad_norm": 0.34707239270210266, "learning_rate": 9.962436878264298e-05, "loss": 0.0459, "step": 8710 }, { "grad_norm": 0.28739309310913086, "learning_rate": 9.962234309713598e-05, "loss": 0.0463, "step": 8720 }, { "grad_norm": 0.3860626816749573, "learning_rate": 9.962031198500414e-05, "loss": 0.0476, "step": 8730 }, { "grad_norm": 0.290892094373703, "learning_rate": 9.961827544646958e-05, "loss": 0.0478, "step": 8740 }, { "grad_norm": 0.27613189816474915, "learning_rate": 9.961623348175501e-05, "loss": 0.0463, "step": 8750 }, { "grad_norm": 0.3520761728286743, "learning_rate": 9.961418609108377e-05, "loss": 0.0576, "step": 8760 }, { "grad_norm": 0.3346399962902069, "learning_rate": 9.961213327467971e-05, "loss": 0.0557, "step": 8770 }, { "grad_norm": 0.29887470602989197, "learning_rate": 9.961007503276736e-05, "loss": 0.0471, "step": 8780 }, { "grad_norm": 0.30381453037261963, "learning_rate": 9.960801136557179e-05, "loss": 0.0457, "step": 8790 }, { "grad_norm": 0.32883650064468384, "learning_rate": 9.960594227331866e-05, "loss": 0.0441, "step": 8800 }, { "grad_norm": 0.37172043323516846, "learning_rate": 9.960386775623429e-05, "loss": 0.0506, "step": 8810 }, { "grad_norm": 0.43079835176467896, "learning_rate": 9.96017878145455e-05, "loss": 0.0511, "step": 8820 }, { "grad_norm": 0.4526348412036896, "learning_rate": 9.959970244847977e-05, "loss": 0.0485, "step": 8830 }, { "grad_norm": 0.32558780908584595, "learning_rate": 9.959761165826518e-05, "loss": 0.0503, "step": 8840 }, { "grad_norm": 0.35702940821647644, "learning_rate": 9.959551544413033e-05, "loss": 0.0451, "step": 8850 }, { "grad_norm": 0.29934069514274597, "learning_rate": 9.959341380630448e-05, "loss": 0.0421, "step": 8860 }, { "grad_norm": 0.2841739058494568, "learning_rate": 9.959130674501746e-05, "loss": 0.0452, "step": 8870 }, { "grad_norm": 0.3004636764526367, "learning_rate": 9.958919426049968e-05, "loss": 0.0466, "step": 8880 }, { "grad_norm": 0.3598528802394867, "learning_rate": 9.958707635298219e-05, "loss": 0.0438, "step": 8890 }, { "grad_norm": 0.4285837709903717, "learning_rate": 9.958495302269657e-05, "loss": 0.0675, "step": 8900 }, { "grad_norm": 0.39039987325668335, "learning_rate": 9.958282426987503e-05, "loss": 0.0571, "step": 8910 }, { "grad_norm": 0.40131762623786926, "learning_rate": 9.95806900947504e-05, "loss": 0.0589, "step": 8920 }, { "grad_norm": 0.31700900197029114, "learning_rate": 9.957855049755604e-05, "loss": 0.0497, "step": 8930 }, { "grad_norm": 0.28524866700172424, "learning_rate": 9.957640547852593e-05, "loss": 0.0508, "step": 8940 }, { "grad_norm": 0.338105171918869, "learning_rate": 9.957425503789466e-05, "loss": 0.053, "step": 8950 }, { "grad_norm": 0.35920339822769165, "learning_rate": 9.957209917589738e-05, "loss": 0.05, "step": 8960 }, { "grad_norm": 0.24654695391654968, "learning_rate": 9.956993789276987e-05, "loss": 0.0541, "step": 8970 }, { "grad_norm": 0.3354056477546692, "learning_rate": 9.956777118874847e-05, "loss": 0.0592, "step": 8980 }, { "grad_norm": 0.3094399869441986, "learning_rate": 9.956559906407016e-05, "loss": 0.0506, "step": 8990 }, { "grad_norm": 0.21641427278518677, "learning_rate": 9.956342151897245e-05, "loss": 0.0402, "step": 9000 }, { "grad_norm": 0.37807130813598633, "learning_rate": 9.956123855369346e-05, "loss": 0.0417, "step": 9010 }, { "grad_norm": 0.3151463568210602, "learning_rate": 9.955905016847196e-05, "loss": 0.046, "step": 9020 }, { "grad_norm": 0.3066529929637909, "learning_rate": 9.955685636354723e-05, "loss": 0.0472, "step": 9030 }, { "grad_norm": 0.3173881769180298, "learning_rate": 9.95546571391592e-05, "loss": 0.0498, "step": 9040 }, { "grad_norm": 0.3453599810600281, "learning_rate": 9.955245249554837e-05, "loss": 0.0515, "step": 9050 }, { "grad_norm": 0.39574190974235535, "learning_rate": 9.955024243295582e-05, "loss": 0.0552, "step": 9060 }, { "grad_norm": 0.3260830342769623, "learning_rate": 9.954802695162328e-05, "loss": 0.0558, "step": 9070 }, { "grad_norm": 0.26747044920921326, "learning_rate": 9.954580605179302e-05, "loss": 0.0464, "step": 9080 }, { "grad_norm": 0.3389814794063568, "learning_rate": 9.954357973370788e-05, "loss": 0.0543, "step": 9090 }, { "grad_norm": 0.24647721648216248, "learning_rate": 9.954134799761135e-05, "loss": 0.0409, "step": 9100 }, { "grad_norm": 0.2511289417743683, "learning_rate": 9.953911084374748e-05, "loss": 0.0428, "step": 9110 }, { "grad_norm": 0.25334274768829346, "learning_rate": 9.953686827236093e-05, "loss": 0.0492, "step": 9120 }, { "grad_norm": 0.32475417852401733, "learning_rate": 9.953462028369695e-05, "loss": 0.0464, "step": 9130 }, { "grad_norm": 0.3292321264743805, "learning_rate": 9.953236687800136e-05, "loss": 0.0533, "step": 9140 }, { "grad_norm": 0.3445097506046295, "learning_rate": 9.95301080555206e-05, "loss": 0.0541, "step": 9150 }, { "grad_norm": 0.3394151031970978, "learning_rate": 9.952784381650171e-05, "loss": 0.0479, "step": 9160 }, { "grad_norm": 0.32424941658973694, "learning_rate": 9.952557416119226e-05, "loss": 0.0579, "step": 9170 }, { "grad_norm": 0.24789495766162872, "learning_rate": 9.95232990898405e-05, "loss": 0.0493, "step": 9180 }, { "grad_norm": 0.32108813524246216, "learning_rate": 9.95210186026952e-05, "loss": 0.046, "step": 9190 }, { "grad_norm": 0.3659400939941406, "learning_rate": 9.951873270000576e-05, "loss": 0.0466, "step": 9200 }, { "grad_norm": 0.3044338524341583, "learning_rate": 9.951644138202216e-05, "loss": 0.0498, "step": 9210 }, { "grad_norm": 0.34796440601348877, "learning_rate": 9.951414464899498e-05, "loss": 0.0566, "step": 9220 }, { "grad_norm": 0.4812367558479309, "learning_rate": 9.951184250117538e-05, "loss": 0.0494, "step": 9230 }, { "grad_norm": 0.2544437348842621, "learning_rate": 9.950953493881513e-05, "loss": 0.0418, "step": 9240 }, { "grad_norm": 0.2477595955133438, "learning_rate": 9.950722196216658e-05, "loss": 0.0427, "step": 9250 }, { "grad_norm": 0.4015721082687378, "learning_rate": 9.950490357148265e-05, "loss": 0.0476, "step": 9260 }, { "grad_norm": 0.4016554653644562, "learning_rate": 9.950257976701692e-05, "loss": 0.049, "step": 9270 }, { "grad_norm": 0.3135639727115631, "learning_rate": 9.950025054902348e-05, "loss": 0.0447, "step": 9280 }, { "grad_norm": 0.28822916746139526, "learning_rate": 9.949791591775706e-05, "loss": 0.0391, "step": 9290 }, { "grad_norm": 0.3401110768318176, "learning_rate": 9.949557587347298e-05, "loss": 0.0484, "step": 9300 }, { "grad_norm": 0.28594478964805603, "learning_rate": 9.949323041642713e-05, "loss": 0.0451, "step": 9310 }, { "grad_norm": 0.32651007175445557, "learning_rate": 9.949087954687602e-05, "loss": 0.0495, "step": 9320 }, { "grad_norm": 0.3720116913318634, "learning_rate": 9.948852326507672e-05, "loss": 0.0445, "step": 9330 }, { "grad_norm": 0.34307387471199036, "learning_rate": 9.948616157128694e-05, "loss": 0.0526, "step": 9340 }, { "grad_norm": 0.35065245628356934, "learning_rate": 9.948379446576493e-05, "loss": 0.0492, "step": 9350 }, { "grad_norm": 0.25426939129829407, "learning_rate": 9.948142194876952e-05, "loss": 0.0448, "step": 9360 }, { "grad_norm": 0.45780935883522034, "learning_rate": 9.947904402056024e-05, "loss": 0.0483, "step": 9370 }, { "grad_norm": 0.2591663599014282, "learning_rate": 9.947666068139708e-05, "loss": 0.0454, "step": 9380 }, { "grad_norm": 0.36461302638053894, "learning_rate": 9.947427193154071e-05, "loss": 0.0544, "step": 9390 }, { "grad_norm": 0.36081135272979736, "learning_rate": 9.947187777125233e-05, "loss": 0.0586, "step": 9400 }, { "grad_norm": 0.3536134362220764, "learning_rate": 9.946947820079377e-05, "loss": 0.0553, "step": 9410 }, { "grad_norm": 0.30973219871520996, "learning_rate": 9.946707322042747e-05, "loss": 0.0504, "step": 9420 }, { "grad_norm": 0.2670998275279999, "learning_rate": 9.94646628304164e-05, "loss": 0.0495, "step": 9430 }, { "grad_norm": 0.2748047709465027, "learning_rate": 9.946224703102418e-05, "loss": 0.0426, "step": 9440 }, { "grad_norm": 0.27449989318847656, "learning_rate": 9.945982582251498e-05, "loss": 0.0442, "step": 9450 }, { "grad_norm": 0.33771777153015137, "learning_rate": 9.94573992051536e-05, "loss": 0.0434, "step": 9460 }, { "grad_norm": 0.3643036484718323, "learning_rate": 9.94549671792054e-05, "loss": 0.0539, "step": 9470 }, { "grad_norm": 0.3506060838699341, "learning_rate": 9.945252974493635e-05, "loss": 0.0517, "step": 9480 }, { "grad_norm": 0.3782847821712494, "learning_rate": 9.9450086902613e-05, "loss": 0.0458, "step": 9490 }, { "grad_norm": 0.28771066665649414, "learning_rate": 9.944763865250248e-05, "loss": 0.0459, "step": 9500 }, { "grad_norm": 0.3433609902858734, "learning_rate": 9.944518499487254e-05, "loss": 0.0581, "step": 9510 }, { "grad_norm": 0.661567747592926, "learning_rate": 9.944272592999151e-05, "loss": 0.0605, "step": 9520 }, { "grad_norm": 0.38249385356903076, "learning_rate": 9.94402614581283e-05, "loss": 0.0548, "step": 9530 }, { "grad_norm": 0.43852749466896057, "learning_rate": 9.943779157955244e-05, "loss": 0.0563, "step": 9540 }, { "grad_norm": 0.5118108987808228, "learning_rate": 9.943531629453403e-05, "loss": 0.0504, "step": 9550 }, { "grad_norm": 0.3381496071815491, "learning_rate": 9.943283560334375e-05, "loss": 0.0533, "step": 9560 }, { "grad_norm": 0.27538803219795227, "learning_rate": 9.943034950625288e-05, "loss": 0.0479, "step": 9570 }, { "grad_norm": 0.3756561279296875, "learning_rate": 9.942785800353332e-05, "loss": 0.059, "step": 9580 }, { "grad_norm": 0.30475664138793945, "learning_rate": 9.942536109545751e-05, "loss": 0.0584, "step": 9590 }, { "grad_norm": 0.3492286205291748, "learning_rate": 9.942285878229853e-05, "loss": 0.0468, "step": 9600 }, { "grad_norm": 0.27382105588912964, "learning_rate": 9.942035106433001e-05, "loss": 0.0459, "step": 9610 }, { "grad_norm": 0.31384652853012085, "learning_rate": 9.94178379418262e-05, "loss": 0.049, "step": 9620 }, { "grad_norm": 0.2594074010848999, "learning_rate": 9.941531941506194e-05, "loss": 0.0509, "step": 9630 }, { "grad_norm": 0.3835730254650116, "learning_rate": 9.941279548431263e-05, "loss": 0.044, "step": 9640 }, { "grad_norm": 0.266657292842865, "learning_rate": 9.941026614985431e-05, "loss": 0.0489, "step": 9650 }, { "grad_norm": 0.3229462802410126, "learning_rate": 9.940773141196357e-05, "loss": 0.0404, "step": 9660 }, { "grad_norm": 0.361224889755249, "learning_rate": 9.94051912709176e-05, "loss": 0.0428, "step": 9670 }, { "grad_norm": 0.3715215027332306, "learning_rate": 9.940264572699421e-05, "loss": 0.0497, "step": 9680 }, { "grad_norm": 0.3569626808166504, "learning_rate": 9.940009478047174e-05, "loss": 0.0461, "step": 9690 }, { "grad_norm": 0.4101095497608185, "learning_rate": 9.939753843162918e-05, "loss": 0.0488, "step": 9700 }, { "grad_norm": 0.3309900760650635, "learning_rate": 9.939497668074609e-05, "loss": 0.0428, "step": 9710 }, { "grad_norm": 0.33381032943725586, "learning_rate": 9.93924095281026e-05, "loss": 0.0449, "step": 9720 }, { "grad_norm": 0.4034979045391083, "learning_rate": 9.938983697397948e-05, "loss": 0.0581, "step": 9730 }, { "grad_norm": 0.40786072611808777, "learning_rate": 9.938725901865805e-05, "loss": 0.0522, "step": 9740 }, { "grad_norm": 0.3421618640422821, "learning_rate": 9.93846756624202e-05, "loss": 0.0481, "step": 9750 }, { "grad_norm": 0.2962745428085327, "learning_rate": 9.938208690554849e-05, "loss": 0.0443, "step": 9760 }, { "grad_norm": 0.31509339809417725, "learning_rate": 9.9379492748326e-05, "loss": 0.0438, "step": 9770 }, { "grad_norm": 0.2965123951435089, "learning_rate": 9.937689319103641e-05, "loss": 0.0425, "step": 9780 }, { "grad_norm": 0.33145394921302795, "learning_rate": 9.937428823396404e-05, "loss": 0.0456, "step": 9790 }, { "grad_norm": 0.29733631014823914, "learning_rate": 9.937167787739372e-05, "loss": 0.0392, "step": 9800 }, { "grad_norm": 0.2577762007713318, "learning_rate": 9.936906212161095e-05, "loss": 0.0424, "step": 9810 }, { "grad_norm": 0.37732207775115967, "learning_rate": 9.936644096690176e-05, "loss": 0.0462, "step": 9820 }, { "grad_norm": 0.33011695742607117, "learning_rate": 9.936381441355282e-05, "loss": 0.0423, "step": 9830 }, { "grad_norm": 0.3168487846851349, "learning_rate": 9.936118246185136e-05, "loss": 0.0472, "step": 9840 }, { "grad_norm": 0.37966853380203247, "learning_rate": 9.935854511208518e-05, "loss": 0.0487, "step": 9850 }, { "grad_norm": 0.3396546542644501, "learning_rate": 9.935590236454272e-05, "loss": 0.0475, "step": 9860 }, { "grad_norm": 0.2804177701473236, "learning_rate": 9.935325421951298e-05, "loss": 0.0397, "step": 9870 }, { "grad_norm": 0.4371699094772339, "learning_rate": 9.935060067728557e-05, "loss": 0.0534, "step": 9880 }, { "grad_norm": 0.3099108338356018, "learning_rate": 9.934794173815067e-05, "loss": 0.0534, "step": 9890 }, { "grad_norm": 0.35175788402557373, "learning_rate": 9.934527740239906e-05, "loss": 0.0505, "step": 9900 }, { "grad_norm": 0.2777506113052368, "learning_rate": 9.934260767032209e-05, "loss": 0.0538, "step": 9910 }, { "grad_norm": 0.28799837827682495, "learning_rate": 9.933993254221172e-05, "loss": 0.0462, "step": 9920 }, { "grad_norm": 0.4658948481082916, "learning_rate": 9.933725201836053e-05, "loss": 0.0474, "step": 9930 }, { "grad_norm": 0.5022138357162476, "learning_rate": 9.933456609906162e-05, "loss": 0.0425, "step": 9940 }, { "grad_norm": 0.34683817625045776, "learning_rate": 9.933187478460875e-05, "loss": 0.0642, "step": 9950 }, { "grad_norm": 0.3135940432548523, "learning_rate": 9.93291780752962e-05, "loss": 0.0419, "step": 9960 }, { "grad_norm": 0.3394109010696411, "learning_rate": 9.932647597141893e-05, "loss": 0.0466, "step": 9970 }, { "grad_norm": 0.40057817101478577, "learning_rate": 9.932376847327239e-05, "loss": 0.0487, "step": 9980 }, { "grad_norm": 0.33717188239097595, "learning_rate": 9.932105558115268e-05, "loss": 0.0492, "step": 9990 }, { "grad_norm": 0.5357943177223206, "learning_rate": 9.931833729535651e-05, "loss": 0.0476, "step": 10000 }, { "grad_norm": 0.311511754989624, "learning_rate": 9.931561361618111e-05, "loss": 0.0491, "step": 10010 }, { "grad_norm": 0.5051846504211426, "learning_rate": 9.931288454392435e-05, "loss": 0.0648, "step": 10020 }, { "grad_norm": 0.3145377039909363, "learning_rate": 9.931015007888467e-05, "loss": 0.0538, "step": 10030 }, { "grad_norm": 0.4527645409107208, "learning_rate": 9.930741022136112e-05, "loss": 0.0593, "step": 10040 }, { "grad_norm": 0.4046013057231903, "learning_rate": 9.930466497165333e-05, "loss": 0.0558, "step": 10050 }, { "grad_norm": 0.284259170293808, "learning_rate": 9.93019143300615e-05, "loss": 0.049, "step": 10060 }, { "grad_norm": 0.34102416038513184, "learning_rate": 9.929915829688644e-05, "loss": 0.0464, "step": 10070 }, { "grad_norm": 0.2512054443359375, "learning_rate": 9.929639687242955e-05, "loss": 0.0419, "step": 10080 }, { "grad_norm": 0.2920326888561249, "learning_rate": 9.929363005699281e-05, "loss": 0.0398, "step": 10090 }, { "grad_norm": 0.28750643134117126, "learning_rate": 9.92908578508788e-05, "loss": 0.052, "step": 10100 }, { "grad_norm": 0.31845173239707947, "learning_rate": 9.928808025439069e-05, "loss": 0.0518, "step": 10110 }, { "grad_norm": 0.35335490107536316, "learning_rate": 9.928529726783223e-05, "loss": 0.0471, "step": 10120 }, { "grad_norm": 0.3632693588733673, "learning_rate": 9.928250889150774e-05, "loss": 0.0509, "step": 10130 }, { "grad_norm": 0.2561100721359253, "learning_rate": 9.92797151257222e-05, "loss": 0.0478, "step": 10140 }, { "grad_norm": 0.2498445063829422, "learning_rate": 9.927691597078108e-05, "loss": 0.0394, "step": 10150 }, { "grad_norm": 0.29033973813056946, "learning_rate": 9.927411142699053e-05, "loss": 0.0459, "step": 10160 }, { "grad_norm": 0.3020201623439789, "learning_rate": 9.927130149465725e-05, "loss": 0.0434, "step": 10170 }, { "grad_norm": 0.32888248562812805, "learning_rate": 9.92684861740885e-05, "loss": 0.0434, "step": 10180 }, { "grad_norm": 0.3084479868412018, "learning_rate": 9.926566546559217e-05, "loss": 0.0433, "step": 10190 }, { "grad_norm": 0.30985450744628906, "learning_rate": 9.926283936947673e-05, "loss": 0.0441, "step": 10200 }, { "grad_norm": 0.3616635799407959, "learning_rate": 9.926000788605126e-05, "loss": 0.0425, "step": 10210 }, { "grad_norm": 0.27053213119506836, "learning_rate": 9.92571710156254e-05, "loss": 0.0407, "step": 10220 }, { "grad_norm": 0.3056408762931824, "learning_rate": 9.925432875850936e-05, "loss": 0.039, "step": 10230 }, { "grad_norm": 0.31148460507392883, "learning_rate": 9.925148111501396e-05, "loss": 0.043, "step": 10240 }, { "grad_norm": 0.31632500886917114, "learning_rate": 9.924862808545066e-05, "loss": 0.0432, "step": 10250 }, { "grad_norm": 0.32695847749710083, "learning_rate": 9.924576967013141e-05, "loss": 0.0552, "step": 10260 }, { "grad_norm": 0.36162009835243225, "learning_rate": 9.924290586936887e-05, "loss": 0.0433, "step": 10270 }, { "grad_norm": 0.3064323663711548, "learning_rate": 9.924003668347614e-05, "loss": 0.045, "step": 10280 }, { "grad_norm": 0.27675244212150574, "learning_rate": 9.923716211276704e-05, "loss": 0.0396, "step": 10290 }, { "grad_norm": 0.22169049084186554, "learning_rate": 9.923428215755594e-05, "loss": 0.0458, "step": 10300 }, { "grad_norm": 0.4284702241420746, "learning_rate": 9.923139681815775e-05, "loss": 0.0447, "step": 10310 }, { "grad_norm": 0.3766774833202362, "learning_rate": 9.922850609488801e-05, "loss": 0.0462, "step": 10320 }, { "grad_norm": 0.36791786551475525, "learning_rate": 9.922560998806287e-05, "loss": 0.0462, "step": 10330 }, { "grad_norm": 0.3244604468345642, "learning_rate": 9.922270849799905e-05, "loss": 0.039, "step": 10340 }, { "grad_norm": 0.6613219976425171, "learning_rate": 9.92198016250138e-05, "loss": 0.0491, "step": 10350 }, { "grad_norm": 0.32909849286079407, "learning_rate": 9.921688936942506e-05, "loss": 0.0411, "step": 10360 }, { "grad_norm": 0.473920613527298, "learning_rate": 9.921397173155129e-05, "loss": 0.0545, "step": 10370 }, { "grad_norm": 0.2864956855773926, "learning_rate": 9.921104871171157e-05, "loss": 0.0544, "step": 10380 }, { "grad_norm": 0.37068459391593933, "learning_rate": 9.920812031022554e-05, "loss": 0.0512, "step": 10390 }, { "grad_norm": 0.23865318298339844, "learning_rate": 9.920518652741348e-05, "loss": 0.0528, "step": 10400 }, { "grad_norm": 0.39563044905662537, "learning_rate": 9.920224736359618e-05, "loss": 0.0539, "step": 10410 }, { "grad_norm": 0.3405526876449585, "learning_rate": 9.91993028190951e-05, "loss": 0.0628, "step": 10420 }, { "grad_norm": 0.41231292486190796, "learning_rate": 9.919635289423222e-05, "loss": 0.048, "step": 10430 }, { "grad_norm": 0.42819541692733765, "learning_rate": 9.919339758933015e-05, "loss": 0.0604, "step": 10440 }, { "grad_norm": 0.3457185626029968, "learning_rate": 9.919043690471209e-05, "loss": 0.0507, "step": 10450 }, { "grad_norm": 0.3455601930618286, "learning_rate": 9.91874708407018e-05, "loss": 0.0528, "step": 10460 }, { "grad_norm": 0.5402559041976929, "learning_rate": 9.918449939762367e-05, "loss": 0.0494, "step": 10470 }, { "grad_norm": 0.30495402216911316, "learning_rate": 9.91815225758026e-05, "loss": 0.0482, "step": 10480 }, { "grad_norm": 0.4064962565898895, "learning_rate": 9.917854037556419e-05, "loss": 0.0453, "step": 10490 }, { "grad_norm": 0.39685511589050293, "learning_rate": 9.917555279723454e-05, "loss": 0.0601, "step": 10500 }, { "grad_norm": 0.334701269865036, "learning_rate": 9.917255984114036e-05, "loss": 0.0558, "step": 10510 }, { "grad_norm": 0.3300895392894745, "learning_rate": 9.916956150760896e-05, "loss": 0.0486, "step": 10520 }, { "grad_norm": 0.3162599205970764, "learning_rate": 9.916655779696826e-05, "loss": 0.0472, "step": 10530 }, { "grad_norm": 0.4433479905128479, "learning_rate": 9.916354870954671e-05, "loss": 0.0553, "step": 10540 }, { "grad_norm": 0.3162727653980255, "learning_rate": 9.91605342456734e-05, "loss": 0.0546, "step": 10550 }, { "grad_norm": 0.453572541475296, "learning_rate": 9.915751440567795e-05, "loss": 0.0444, "step": 10560 }, { "grad_norm": 0.22061237692832947, "learning_rate": 9.915448918989066e-05, "loss": 0.0441, "step": 10570 }, { "grad_norm": 0.33530187606811523, "learning_rate": 9.915145859864232e-05, "loss": 0.0469, "step": 10580 }, { "grad_norm": 0.23930855095386505, "learning_rate": 9.914842263226437e-05, "loss": 0.0418, "step": 10590 }, { "grad_norm": 0.30023306608200073, "learning_rate": 9.914538129108882e-05, "loss": 0.0509, "step": 10600 }, { "grad_norm": 0.3713613748550415, "learning_rate": 9.914233457544825e-05, "loss": 0.0567, "step": 10610 }, { "grad_norm": 0.3494429290294647, "learning_rate": 9.913928248567586e-05, "loss": 0.0502, "step": 10620 }, { "grad_norm": 0.584362268447876, "learning_rate": 9.913622502210542e-05, "loss": 0.0524, "step": 10630 }, { "grad_norm": 0.31017085909843445, "learning_rate": 9.913316218507128e-05, "loss": 0.051, "step": 10640 }, { "grad_norm": 0.33326902985572815, "learning_rate": 9.91300939749084e-05, "loss": 0.0422, "step": 10650 }, { "grad_norm": 0.31283462047576904, "learning_rate": 9.91270203919523e-05, "loss": 0.045, "step": 10660 }, { "grad_norm": 0.2918432056903839, "learning_rate": 9.912394143653912e-05, "loss": 0.0524, "step": 10670 }, { "grad_norm": 0.42240652441978455, "learning_rate": 9.912085710900555e-05, "loss": 0.0495, "step": 10680 }, { "grad_norm": 0.3048814833164215, "learning_rate": 9.911776740968892e-05, "loss": 0.0477, "step": 10690 }, { "grad_norm": 0.3723277747631073, "learning_rate": 9.911467233892709e-05, "loss": 0.0416, "step": 10700 }, { "grad_norm": 0.42246755957603455, "learning_rate": 9.911157189705853e-05, "loss": 0.042, "step": 10710 }, { "grad_norm": 0.4553736746311188, "learning_rate": 9.910846608442229e-05, "loss": 0.0409, "step": 10720 }, { "grad_norm": 0.29423415660858154, "learning_rate": 9.910535490135805e-05, "loss": 0.0448, "step": 10730 }, { "grad_norm": 0.2975880801677704, "learning_rate": 9.910223834820603e-05, "loss": 0.0385, "step": 10740 }, { "grad_norm": 0.38897088170051575, "learning_rate": 9.909911642530703e-05, "loss": 0.0451, "step": 10750 }, { "grad_norm": 0.31574565172195435, "learning_rate": 9.909598913300249e-05, "loss": 0.045, "step": 10760 }, { "grad_norm": 0.4129357635974884, "learning_rate": 9.909285647163438e-05, "loss": 0.0489, "step": 10770 }, { "grad_norm": 0.4850662350654602, "learning_rate": 9.908971844154531e-05, "loss": 0.0434, "step": 10780 }, { "grad_norm": 0.4261828064918518, "learning_rate": 9.908657504307843e-05, "loss": 0.0432, "step": 10790 }, { "grad_norm": 0.31240910291671753, "learning_rate": 9.908342627657751e-05, "loss": 0.0392, "step": 10800 }, { "grad_norm": 0.2596571445465088, "learning_rate": 9.908027214238689e-05, "loss": 0.0465, "step": 10810 }, { "grad_norm": 0.28224217891693115, "learning_rate": 9.90771126408515e-05, "loss": 0.0421, "step": 10820 }, { "grad_norm": 0.387029230594635, "learning_rate": 9.907394777231685e-05, "loss": 0.0432, "step": 10830 }, { "grad_norm": 0.26686882972717285, "learning_rate": 9.907077753712905e-05, "loss": 0.0383, "step": 10840 }, { "grad_norm": 0.32325759530067444, "learning_rate": 9.906760193563482e-05, "loss": 0.0391, "step": 10850 }, { "grad_norm": 0.3183691203594208, "learning_rate": 9.906442096818139e-05, "loss": 0.0434, "step": 10860 }, { "grad_norm": 0.26637381315231323, "learning_rate": 9.906123463511665e-05, "loss": 0.0453, "step": 10870 }, { "grad_norm": 0.2919099032878876, "learning_rate": 9.905804293678907e-05, "loss": 0.0455, "step": 10880 }, { "grad_norm": 0.24338579177856445, "learning_rate": 9.905484587354766e-05, "loss": 0.0413, "step": 10890 }, { "grad_norm": 0.3484202027320862, "learning_rate": 9.905164344574205e-05, "loss": 0.0427, "step": 10900 }, { "grad_norm": 0.31699541211128235, "learning_rate": 9.904843565372248e-05, "loss": 0.0449, "step": 10910 }, { "grad_norm": 0.3247033357620239, "learning_rate": 9.904522249783972e-05, "loss": 0.0407, "step": 10920 }, { "grad_norm": 0.2536627948284149, "learning_rate": 9.904200397844517e-05, "loss": 0.0407, "step": 10930 }, { "grad_norm": 0.2789875566959381, "learning_rate": 9.903878009589078e-05, "loss": 0.04, "step": 10940 }, { "grad_norm": 0.44149938225746155, "learning_rate": 9.903555085052915e-05, "loss": 0.0409, "step": 10950 }, { "grad_norm": 0.36425596475601196, "learning_rate": 9.903231624271338e-05, "loss": 0.0466, "step": 10960 }, { "grad_norm": 0.3405303955078125, "learning_rate": 9.902907627279724e-05, "loss": 0.0499, "step": 10970 }, { "grad_norm": 1.03971266746521, "learning_rate": 9.902583094113504e-05, "loss": 0.049, "step": 10980 }, { "grad_norm": 0.28018200397491455, "learning_rate": 9.902258024808168e-05, "loss": 0.047, "step": 10990 }, { "grad_norm": 0.2503003180027008, "learning_rate": 9.901932419399264e-05, "loss": 0.0459, "step": 11000 }, { "grad_norm": 0.41409969329833984, "learning_rate": 9.9016062779224e-05, "loss": 0.0464, "step": 11010 }, { "grad_norm": 0.2955301105976105, "learning_rate": 9.901279600413242e-05, "loss": 0.0435, "step": 11020 }, { "grad_norm": 0.3241637647151947, "learning_rate": 9.900952386907518e-05, "loss": 0.048, "step": 11030 }, { "grad_norm": 0.2416229248046875, "learning_rate": 9.90062463744101e-05, "loss": 0.0501, "step": 11040 }, { "grad_norm": 0.29720503091812134, "learning_rate": 9.900296352049558e-05, "loss": 0.0418, "step": 11050 }, { "grad_norm": 0.2785741984844208, "learning_rate": 9.899967530769065e-05, "loss": 0.0377, "step": 11060 }, { "grad_norm": 0.33001822233200073, "learning_rate": 9.899638173635489e-05, "loss": 0.039, "step": 11070 }, { "grad_norm": 0.761306643486023, "learning_rate": 9.899308280684849e-05, "loss": 0.0483, "step": 11080 }, { "grad_norm": 0.2816638648509979, "learning_rate": 9.898977851953222e-05, "loss": 0.0453, "step": 11090 }, { "grad_norm": 0.3542481064796448, "learning_rate": 9.898646887476741e-05, "loss": 0.0456, "step": 11100 }, { "grad_norm": 0.38446682691574097, "learning_rate": 9.898315387291603e-05, "loss": 0.047, "step": 11110 }, { "grad_norm": 0.6102462410926819, "learning_rate": 9.89798335143406e-05, "loss": 0.0418, "step": 11120 }, { "grad_norm": 0.34836822748184204, "learning_rate": 9.897650779940419e-05, "loss": 0.0451, "step": 11130 }, { "grad_norm": 0.30200982093811035, "learning_rate": 9.897317672847054e-05, "loss": 0.0473, "step": 11140 }, { "grad_norm": 0.3533901870250702, "learning_rate": 9.89698403019039e-05, "loss": 0.0448, "step": 11150 }, { "grad_norm": 0.4860115051269531, "learning_rate": 9.896649852006917e-05, "loss": 0.0442, "step": 11160 }, { "grad_norm": 0.30497050285339355, "learning_rate": 9.896315138333177e-05, "loss": 0.0491, "step": 11170 }, { "grad_norm": 0.40287187695503235, "learning_rate": 9.895979889205774e-05, "loss": 0.0492, "step": 11180 }, { "grad_norm": 0.3177461326122284, "learning_rate": 9.895644104661372e-05, "loss": 0.051, "step": 11190 }, { "grad_norm": 0.3429321348667145, "learning_rate": 9.895307784736691e-05, "loss": 0.0386, "step": 11200 }, { "grad_norm": 0.603959321975708, "learning_rate": 9.894970929468512e-05, "loss": 0.0465, "step": 11210 }, { "grad_norm": 0.31250810623168945, "learning_rate": 9.89463353889367e-05, "loss": 0.0496, "step": 11220 }, { "grad_norm": 0.2805803120136261, "learning_rate": 9.894295613049065e-05, "loss": 0.0473, "step": 11230 }, { "grad_norm": 0.34499073028564453, "learning_rate": 9.893957151971649e-05, "loss": 0.0426, "step": 11240 }, { "grad_norm": 0.2509557902812958, "learning_rate": 9.893618155698436e-05, "loss": 0.0378, "step": 11250 }, { "grad_norm": 0.2786099910736084, "learning_rate": 9.8932786242665e-05, "loss": 0.0419, "step": 11260 }, { "grad_norm": 0.3059607148170471, "learning_rate": 9.89293855771297e-05, "loss": 0.0496, "step": 11270 }, { "grad_norm": 0.37167638540267944, "learning_rate": 9.892597956075036e-05, "loss": 0.0493, "step": 11280 }, { "grad_norm": 0.31034785509109497, "learning_rate": 9.892256819389947e-05, "loss": 0.0511, "step": 11290 }, { "grad_norm": 0.3344513475894928, "learning_rate": 9.891915147695006e-05, "loss": 0.049, "step": 11300 }, { "grad_norm": 0.43516308069229126, "learning_rate": 9.891572941027577e-05, "loss": 0.0655, "step": 11310 }, { "grad_norm": 0.32866376638412476, "learning_rate": 9.89123019942509e-05, "loss": 0.0573, "step": 11320 }, { "grad_norm": 0.23229779303073883, "learning_rate": 9.89088692292502e-05, "loss": 0.0399, "step": 11330 }, { "grad_norm": 0.27346837520599365, "learning_rate": 9.89054311156491e-05, "loss": 0.0462, "step": 11340 }, { "grad_norm": 0.3590884208679199, "learning_rate": 9.890198765382357e-05, "loss": 0.0471, "step": 11350 }, { "grad_norm": 0.292706698179245, "learning_rate": 9.889853884415021e-05, "loss": 0.044, "step": 11360 }, { "grad_norm": 0.2841812074184418, "learning_rate": 9.889508468700614e-05, "loss": 0.0451, "step": 11370 }, { "grad_norm": 0.39001333713531494, "learning_rate": 9.889162518276915e-05, "loss": 0.0427, "step": 11380 }, { "grad_norm": 0.39951828122138977, "learning_rate": 9.888816033181752e-05, "loss": 0.0519, "step": 11390 }, { "grad_norm": 0.368928462266922, "learning_rate": 9.888469013453018e-05, "loss": 0.0473, "step": 11400 }, { "grad_norm": 0.25083714723587036, "learning_rate": 9.888121459128663e-05, "loss": 0.044, "step": 11410 }, { "grad_norm": 0.2678552269935608, "learning_rate": 9.887773370246693e-05, "loss": 0.0469, "step": 11420 }, { "grad_norm": 0.5079693794250488, "learning_rate": 9.887424746845177e-05, "loss": 0.044, "step": 11430 }, { "grad_norm": 0.30192941427230835, "learning_rate": 9.887075588962239e-05, "loss": 0.0433, "step": 11440 }, { "grad_norm": 0.24660992622375488, "learning_rate": 9.88672589663606e-05, "loss": 0.0396, "step": 11450 }, { "grad_norm": 0.35197997093200684, "learning_rate": 9.886375669904886e-05, "loss": 0.0421, "step": 11460 }, { "grad_norm": 0.27518364787101746, "learning_rate": 9.886024908807014e-05, "loss": 0.0437, "step": 11470 }, { "grad_norm": 0.4286629557609558, "learning_rate": 9.885673613380806e-05, "loss": 0.0467, "step": 11480 }, { "grad_norm": 0.2837940752506256, "learning_rate": 9.885321783664676e-05, "loss": 0.0431, "step": 11490 }, { "grad_norm": 0.21690455079078674, "learning_rate": 9.884969419697101e-05, "loss": 0.0449, "step": 11500 }, { "grad_norm": 0.5099011063575745, "learning_rate": 9.884616521516614e-05, "loss": 0.0548, "step": 11510 }, { "grad_norm": 0.47115615010261536, "learning_rate": 9.88426308916181e-05, "loss": 0.042, "step": 11520 }, { "grad_norm": 0.27917981147766113, "learning_rate": 9.883909122671335e-05, "loss": 0.0468, "step": 11530 }, { "grad_norm": 0.3570564091205597, "learning_rate": 9.883554622083904e-05, "loss": 0.0472, "step": 11540 }, { "grad_norm": 0.38342490792274475, "learning_rate": 9.88319958743828e-05, "loss": 0.0507, "step": 11550 }, { "grad_norm": 0.4265608489513397, "learning_rate": 9.882844018773291e-05, "loss": 0.0405, "step": 11560 }, { "grad_norm": 0.25653499364852905, "learning_rate": 9.882487916127823e-05, "loss": 0.0443, "step": 11570 }, { "grad_norm": 0.409659743309021, "learning_rate": 9.882131279540815e-05, "loss": 0.0471, "step": 11580 }, { "grad_norm": 0.2292713224887848, "learning_rate": 9.881774109051271e-05, "loss": 0.0409, "step": 11590 }, { "grad_norm": 0.2496616244316101, "learning_rate": 9.881416404698252e-05, "loss": 0.0469, "step": 11600 }, { "grad_norm": 0.25865134596824646, "learning_rate": 9.881058166520873e-05, "loss": 0.0392, "step": 11610 }, { "grad_norm": 0.29765886068344116, "learning_rate": 9.880699394558311e-05, "loss": 0.0397, "step": 11620 }, { "grad_norm": 0.31763705611228943, "learning_rate": 9.880340088849801e-05, "loss": 0.0427, "step": 11630 }, { "grad_norm": 0.26449182629585266, "learning_rate": 9.879980249434637e-05, "loss": 0.0403, "step": 11640 }, { "grad_norm": 0.2736349403858185, "learning_rate": 9.879619876352168e-05, "loss": 0.0411, "step": 11650 }, { "grad_norm": 0.5267704129219055, "learning_rate": 9.879258969641809e-05, "loss": 0.047, "step": 11660 }, { "grad_norm": 0.31450241804122925, "learning_rate": 9.878897529343023e-05, "loss": 0.0456, "step": 11670 }, { "grad_norm": 0.2656871974468231, "learning_rate": 9.878535555495338e-05, "loss": 0.044, "step": 11680 }, { "grad_norm": 0.31532636284828186, "learning_rate": 9.87817304813834e-05, "loss": 0.0416, "step": 11690 }, { "grad_norm": 0.32394999265670776, "learning_rate": 9.877810007311671e-05, "loss": 0.0465, "step": 11700 }, { "grad_norm": 0.5196300745010376, "learning_rate": 9.877446433055035e-05, "loss": 0.0484, "step": 11710 }, { "grad_norm": 0.2772464454174042, "learning_rate": 9.877082325408191e-05, "loss": 0.0498, "step": 11720 }, { "grad_norm": 0.3033939003944397, "learning_rate": 9.876717684410954e-05, "loss": 0.0449, "step": 11730 }, { "grad_norm": 0.40689176321029663, "learning_rate": 9.876352510103204e-05, "loss": 0.0549, "step": 11740 }, { "grad_norm": 0.40160757303237915, "learning_rate": 9.875986802524875e-05, "loss": 0.0433, "step": 11750 }, { "grad_norm": 0.2776072919368744, "learning_rate": 9.87562056171596e-05, "loss": 0.044, "step": 11760 }, { "grad_norm": 0.3141855001449585, "learning_rate": 9.875253787716511e-05, "loss": 0.0431, "step": 11770 }, { "grad_norm": 0.5078995227813721, "learning_rate": 9.874886480566637e-05, "loss": 0.0493, "step": 11780 }, { "grad_norm": 0.39499422907829285, "learning_rate": 9.874518640306507e-05, "loss": 0.0439, "step": 11790 }, { "grad_norm": 0.3634946048259735, "learning_rate": 9.874150266976347e-05, "loss": 0.0476, "step": 11800 }, { "grad_norm": 0.3373687267303467, "learning_rate": 9.873781360616443e-05, "loss": 0.0533, "step": 11810 }, { "grad_norm": 0.3977544605731964, "learning_rate": 9.873411921267137e-05, "loss": 0.0523, "step": 11820 }, { "grad_norm": 0.4957491159439087, "learning_rate": 9.873041948968829e-05, "loss": 0.048, "step": 11830 }, { "grad_norm": 0.31267377734184265, "learning_rate": 9.872671443761981e-05, "loss": 0.0474, "step": 11840 }, { "grad_norm": 0.4250281751155853, "learning_rate": 9.872300405687109e-05, "loss": 0.0503, "step": 11850 }, { "grad_norm": 0.25110793113708496, "learning_rate": 9.871928834784792e-05, "loss": 0.0428, "step": 11860 }, { "grad_norm": 0.3099499046802521, "learning_rate": 9.871556731095661e-05, "loss": 0.0538, "step": 11870 }, { "grad_norm": 0.33874472975730896, "learning_rate": 9.871184094660411e-05, "loss": 0.0421, "step": 11880 }, { "grad_norm": 0.41492319107055664, "learning_rate": 9.870810925519791e-05, "loss": 0.0417, "step": 11890 }, { "grad_norm": 0.3598022758960724, "learning_rate": 9.870437223714612e-05, "loss": 0.0419, "step": 11900 }, { "grad_norm": 1.1085822582244873, "learning_rate": 9.87006298928574e-05, "loss": 0.0395, "step": 11910 }, { "grad_norm": 0.30431678891181946, "learning_rate": 9.869688222274103e-05, "loss": 0.0424, "step": 11920 }, { "grad_norm": 0.27677837014198303, "learning_rate": 9.869312922720681e-05, "loss": 0.0421, "step": 11930 }, { "grad_norm": 0.37601107358932495, "learning_rate": 9.868937090666521e-05, "loss": 0.0449, "step": 11940 }, { "grad_norm": 0.23291006684303284, "learning_rate": 9.86856072615272e-05, "loss": 0.0456, "step": 11950 }, { "grad_norm": 0.4445899724960327, "learning_rate": 9.868183829220438e-05, "loss": 0.0438, "step": 11960 }, { "grad_norm": 0.3156808316707611, "learning_rate": 9.867806399910893e-05, "loss": 0.055, "step": 11970 }, { "grad_norm": 0.38752031326293945, "learning_rate": 9.867428438265356e-05, "loss": 0.0548, "step": 11980 }, { "grad_norm": 0.32201385498046875, "learning_rate": 9.867049944325165e-05, "loss": 0.0441, "step": 11990 }, { "grad_norm": 0.2710117995738983, "learning_rate": 9.86667091813171e-05, "loss": 0.0453, "step": 12000 }, { "grad_norm": 0.2789979577064514, "learning_rate": 9.866291359726438e-05, "loss": 0.0359, "step": 12010 }, { "grad_norm": 0.5155657529830933, "learning_rate": 9.865911269150861e-05, "loss": 0.0407, "step": 12020 }, { "grad_norm": 0.27933356165885925, "learning_rate": 9.865530646446544e-05, "loss": 0.0388, "step": 12030 }, { "grad_norm": 0.26868119835853577, "learning_rate": 9.86514949165511e-05, "loss": 0.0399, "step": 12040 }, { "grad_norm": 0.28064295649528503, "learning_rate": 9.864767804818243e-05, "loss": 0.0409, "step": 12050 }, { "grad_norm": 0.3321366012096405, "learning_rate": 9.86438558597768e-05, "loss": 0.0405, "step": 12060 }, { "grad_norm": 0.2691470980644226, "learning_rate": 9.864002835175225e-05, "loss": 0.048, "step": 12070 }, { "grad_norm": 0.2709594964981079, "learning_rate": 9.863619552452734e-05, "loss": 0.0507, "step": 12080 }, { "grad_norm": 0.3102666437625885, "learning_rate": 9.863235737852119e-05, "loss": 0.0534, "step": 12090 }, { "grad_norm": 0.3138081133365631, "learning_rate": 9.862851391415356e-05, "loss": 0.0456, "step": 12100 }, { "grad_norm": 0.28217223286628723, "learning_rate": 9.862466513184477e-05, "loss": 0.0418, "step": 12110 }, { "grad_norm": 0.30747655034065247, "learning_rate": 9.86208110320157e-05, "loss": 0.0395, "step": 12120 }, { "grad_norm": 0.2678341567516327, "learning_rate": 9.861695161508784e-05, "loss": 0.0432, "step": 12130 }, { "grad_norm": 0.30305325984954834, "learning_rate": 9.861308688148324e-05, "loss": 0.0422, "step": 12140 }, { "grad_norm": 0.26139211654663086, "learning_rate": 9.860921683162455e-05, "loss": 0.0444, "step": 12150 }, { "grad_norm": 0.2946103513240814, "learning_rate": 9.860534146593499e-05, "loss": 0.0412, "step": 12160 }, { "grad_norm": 0.3157806694507599, "learning_rate": 9.860146078483836e-05, "loss": 0.0385, "step": 12170 }, { "grad_norm": 0.29235050082206726, "learning_rate": 9.859757478875905e-05, "loss": 0.0474, "step": 12180 }, { "grad_norm": 0.40888357162475586, "learning_rate": 9.859368347812204e-05, "loss": 0.0398, "step": 12190 }, { "grad_norm": 0.26955097913742065, "learning_rate": 9.858978685335285e-05, "loss": 0.0489, "step": 12200 }, { "grad_norm": 0.3071417510509491, "learning_rate": 9.858588491487763e-05, "loss": 0.0379, "step": 12210 }, { "grad_norm": 0.4398919641971588, "learning_rate": 9.858197766312308e-05, "loss": 0.0398, "step": 12220 }, { "grad_norm": 0.3180479407310486, "learning_rate": 9.857806509851649e-05, "loss": 0.0482, "step": 12230 }, { "grad_norm": 0.3634192943572998, "learning_rate": 9.857414722148574e-05, "loss": 0.0416, "step": 12240 }, { "grad_norm": 0.2626895308494568, "learning_rate": 9.857022403245928e-05, "loss": 0.0382, "step": 12250 }, { "grad_norm": 0.2924368679523468, "learning_rate": 9.856629553186615e-05, "loss": 0.0401, "step": 12260 }, { "grad_norm": 0.3084647059440613, "learning_rate": 9.856236172013595e-05, "loss": 0.0381, "step": 12270 }, { "grad_norm": 0.29309555888175964, "learning_rate": 9.85584225976989e-05, "loss": 0.0402, "step": 12280 }, { "grad_norm": 0.33205288648605347, "learning_rate": 9.855447816498575e-05, "loss": 0.0388, "step": 12290 }, { "grad_norm": 0.3018510937690735, "learning_rate": 9.855052842242787e-05, "loss": 0.043, "step": 12300 }, { "grad_norm": 0.3923811912536621, "learning_rate": 9.85465733704572e-05, "loss": 0.0433, "step": 12310 }, { "grad_norm": 0.24527622759342194, "learning_rate": 9.854261300950624e-05, "loss": 0.0368, "step": 12320 }, { "grad_norm": 0.29544541239738464, "learning_rate": 9.853864734000813e-05, "loss": 0.0382, "step": 12330 }, { "grad_norm": 0.2625126540660858, "learning_rate": 9.85346763623965e-05, "loss": 0.0355, "step": 12340 }, { "grad_norm": 0.24651053547859192, "learning_rate": 9.853070007710564e-05, "loss": 0.037, "step": 12350 }, { "grad_norm": 0.32019132375717163, "learning_rate": 9.85267184845704e-05, "loss": 0.0429, "step": 12360 }, { "grad_norm": 0.2849300801753998, "learning_rate": 9.852273158522616e-05, "loss": 0.0363, "step": 12370 }, { "grad_norm": 0.2891554832458496, "learning_rate": 9.851873937950896e-05, "loss": 0.0443, "step": 12380 }, { "grad_norm": 0.4822487235069275, "learning_rate": 9.851474186785537e-05, "loss": 0.0474, "step": 12390 }, { "grad_norm": 0.39864829182624817, "learning_rate": 9.851073905070254e-05, "loss": 0.047, "step": 12400 }, { "grad_norm": 0.2644566297531128, "learning_rate": 9.850673092848824e-05, "loss": 0.0424, "step": 12410 }, { "grad_norm": 0.2776907980442047, "learning_rate": 9.850271750165077e-05, "loss": 0.0366, "step": 12420 }, { "grad_norm": 0.2307092845439911, "learning_rate": 9.849869877062902e-05, "loss": 0.041, "step": 12430 }, { "grad_norm": 0.36871838569641113, "learning_rate": 9.849467473586252e-05, "loss": 0.0445, "step": 12440 }, { "grad_norm": 0.390664279460907, "learning_rate": 9.849064539779127e-05, "loss": 0.0416, "step": 12450 }, { "grad_norm": 0.23536796867847443, "learning_rate": 9.848661075685594e-05, "loss": 0.0384, "step": 12460 }, { "grad_norm": 0.2804446220397949, "learning_rate": 9.848257081349778e-05, "loss": 0.0405, "step": 12470 }, { "grad_norm": 0.236503005027771, "learning_rate": 9.847852556815856e-05, "loss": 0.0402, "step": 12480 }, { "grad_norm": 0.29394596815109253, "learning_rate": 9.847447502128067e-05, "loss": 0.0488, "step": 12490 }, { "grad_norm": 0.31576645374298096, "learning_rate": 9.847041917330708e-05, "loss": 0.0444, "step": 12500 }, { "grad_norm": 0.43668627738952637, "learning_rate": 9.846635802468132e-05, "loss": 0.0601, "step": 12510 }, { "grad_norm": 0.48177003860473633, "learning_rate": 9.84622915758475e-05, "loss": 0.0526, "step": 12520 }, { "grad_norm": 1.1711492538452148, "learning_rate": 9.845821982725034e-05, "loss": 0.0476, "step": 12530 }, { "grad_norm": 0.38399258255958557, "learning_rate": 9.845414277933514e-05, "loss": 0.0537, "step": 12540 }, { "grad_norm": 0.7303115129470825, "learning_rate": 9.845006043254771e-05, "loss": 0.0546, "step": 12550 }, { "grad_norm": 0.2607549726963043, "learning_rate": 9.844597278733451e-05, "loss": 0.0509, "step": 12560 }, { "grad_norm": 0.3009967505931854, "learning_rate": 9.844187984414259e-05, "loss": 0.0485, "step": 12570 }, { "grad_norm": 0.24072954058647156, "learning_rate": 9.84377816034195e-05, "loss": 0.0467, "step": 12580 }, { "grad_norm": 0.28709664940834045, "learning_rate": 9.843367806561345e-05, "loss": 0.0503, "step": 12590 }, { "grad_norm": 0.29509487748146057, "learning_rate": 9.842956923117317e-05, "loss": 0.0441, "step": 12600 }, { "grad_norm": 0.4191574454307556, "learning_rate": 9.842545510054802e-05, "loss": 0.0464, "step": 12610 }, { "grad_norm": 0.3469180762767792, "learning_rate": 9.842133567418792e-05, "loss": 0.0496, "step": 12620 }, { "grad_norm": 0.43780773878097534, "learning_rate": 9.841721095254333e-05, "loss": 0.0517, "step": 12630 }, { "grad_norm": 0.2449873685836792, "learning_rate": 9.841308093606537e-05, "loss": 0.0499, "step": 12640 }, { "grad_norm": 0.4162309169769287, "learning_rate": 9.840894562520565e-05, "loss": 0.0461, "step": 12650 }, { "grad_norm": 0.33858928084373474, "learning_rate": 9.840480502041642e-05, "loss": 0.0418, "step": 12660 }, { "grad_norm": 0.37841373682022095, "learning_rate": 9.840065912215049e-05, "loss": 0.0529, "step": 12670 }, { "grad_norm": 0.2335795909166336, "learning_rate": 9.839650793086124e-05, "loss": 0.0386, "step": 12680 }, { "grad_norm": 0.2763230800628662, "learning_rate": 9.839235144700265e-05, "loss": 0.0515, "step": 12690 }, { "grad_norm": 0.29129260778427124, "learning_rate": 9.838818967102926e-05, "loss": 0.0441, "step": 12700 }, { "grad_norm": 0.2912296652793884, "learning_rate": 9.83840226033962e-05, "loss": 0.0434, "step": 12710 }, { "grad_norm": 0.35611844062805176, "learning_rate": 9.837985024455918e-05, "loss": 0.0472, "step": 12720 }, { "grad_norm": 0.2780812382698059, "learning_rate": 9.837567259497447e-05, "loss": 0.0475, "step": 12730 }, { "grad_norm": 0.34123513102531433, "learning_rate": 9.837148965509894e-05, "loss": 0.0434, "step": 12740 }, { "grad_norm": 0.2691580355167389, "learning_rate": 9.836730142539001e-05, "loss": 0.0422, "step": 12750 }, { "grad_norm": 0.26031258702278137, "learning_rate": 9.836310790630574e-05, "loss": 0.0406, "step": 12760 }, { "grad_norm": 0.26079046726226807, "learning_rate": 9.83589090983047e-05, "loss": 0.044, "step": 12770 }, { "grad_norm": 0.2854926586151123, "learning_rate": 9.835470500184605e-05, "loss": 0.0408, "step": 12780 }, { "grad_norm": 0.23367847502231598, "learning_rate": 9.835049561738957e-05, "loss": 0.0347, "step": 12790 }, { "grad_norm": 0.3378695845603943, "learning_rate": 9.834628094539558e-05, "loss": 0.0462, "step": 12800 }, { "grad_norm": 0.2900370657444, "learning_rate": 9.834206098632499e-05, "loss": 0.0467, "step": 12810 }, { "grad_norm": 0.3400246798992157, "learning_rate": 9.833783574063931e-05, "loss": 0.0496, "step": 12820 }, { "grad_norm": 0.3291292190551758, "learning_rate": 9.833360520880058e-05, "loss": 0.0435, "step": 12830 }, { "grad_norm": 0.3062918782234192, "learning_rate": 9.832936939127144e-05, "loss": 0.0448, "step": 12840 }, { "grad_norm": 0.3315475285053253, "learning_rate": 9.832512828851515e-05, "loss": 0.0398, "step": 12850 }, { "grad_norm": 0.3131055533885956, "learning_rate": 9.832088190099546e-05, "loss": 0.0402, "step": 12860 }, { "grad_norm": 0.3210456371307373, "learning_rate": 9.831663022917679e-05, "loss": 0.0442, "step": 12870 }, { "grad_norm": 0.2768024206161499, "learning_rate": 9.831237327352407e-05, "loss": 0.0391, "step": 12880 }, { "grad_norm": 0.3392329812049866, "learning_rate": 9.830811103450286e-05, "loss": 0.0348, "step": 12890 }, { "grad_norm": 0.3967742323875427, "learning_rate": 9.830384351257924e-05, "loss": 0.0408, "step": 12900 }, { "grad_norm": 0.28612229228019714, "learning_rate": 9.829957070821993e-05, "loss": 0.0414, "step": 12910 }, { "grad_norm": 0.29774796962738037, "learning_rate": 9.829529262189218e-05, "loss": 0.0419, "step": 12920 }, { "grad_norm": 0.310901015996933, "learning_rate": 9.829100925406385e-05, "loss": 0.0386, "step": 12930 }, { "grad_norm": 0.4330936074256897, "learning_rate": 9.828672060520333e-05, "loss": 0.0478, "step": 12940 }, { "grad_norm": 0.29873397946357727, "learning_rate": 9.828242667577966e-05, "loss": 0.0542, "step": 12950 }, { "grad_norm": 0.3428194224834442, "learning_rate": 9.82781274662624e-05, "loss": 0.0477, "step": 12960 }, { "grad_norm": 0.32352402806282043, "learning_rate": 9.82738229771217e-05, "loss": 0.0407, "step": 12970 }, { "grad_norm": 0.3374667763710022, "learning_rate": 9.826951320882829e-05, "loss": 0.0425, "step": 12980 }, { "grad_norm": 0.2506619393825531, "learning_rate": 9.826519816185351e-05, "loss": 0.0412, "step": 12990 }, { "grad_norm": 0.2849890887737274, "learning_rate": 9.826087783666921e-05, "loss": 0.0466, "step": 13000 }, { "grad_norm": 0.27923858165740967, "learning_rate": 9.825655223374787e-05, "loss": 0.0401, "step": 13010 }, { "grad_norm": 0.28767651319503784, "learning_rate": 9.825222135356253e-05, "loss": 0.0421, "step": 13020 }, { "grad_norm": 0.37721920013427734, "learning_rate": 9.82478851965868e-05, "loss": 0.0396, "step": 13030 }, { "grad_norm": 0.33355772495269775, "learning_rate": 9.82435437632949e-05, "loss": 0.0455, "step": 13040 }, { "grad_norm": 0.35658761858940125, "learning_rate": 9.823919705416158e-05, "loss": 0.0427, "step": 13050 }, { "grad_norm": 0.32112935185432434, "learning_rate": 9.82348450696622e-05, "loss": 0.0373, "step": 13060 }, { "grad_norm": 0.24573490023612976, "learning_rate": 9.823048781027268e-05, "loss": 0.038, "step": 13070 }, { "grad_norm": 0.408862441778183, "learning_rate": 9.822612527646953e-05, "loss": 0.0414, "step": 13080 }, { "grad_norm": 0.3312143087387085, "learning_rate": 9.822175746872984e-05, "loss": 0.0338, "step": 13090 }, { "grad_norm": 0.21569429337978363, "learning_rate": 9.821738438753123e-05, "loss": 0.0438, "step": 13100 }, { "grad_norm": 0.27201470732688904, "learning_rate": 9.821300603335196e-05, "loss": 0.0424, "step": 13110 }, { "grad_norm": 0.37190350890159607, "learning_rate": 9.820862240667085e-05, "loss": 0.0467, "step": 13120 }, { "grad_norm": 0.25431758165359497, "learning_rate": 9.820423350796726e-05, "loss": 0.0489, "step": 13130 }, { "grad_norm": 0.3061331808567047, "learning_rate": 9.819983933772118e-05, "loss": 0.0468, "step": 13140 }, { "grad_norm": 0.29077816009521484, "learning_rate": 9.819543989641314e-05, "loss": 0.0438, "step": 13150 }, { "grad_norm": 0.38374000787734985, "learning_rate": 9.819103518452423e-05, "loss": 0.0394, "step": 13160 }, { "grad_norm": 0.3523827791213989, "learning_rate": 9.818662520253618e-05, "loss": 0.0474, "step": 13170 }, { "grad_norm": 0.2310057282447815, "learning_rate": 9.818220995093126e-05, "loss": 0.0437, "step": 13180 }, { "grad_norm": 0.4769442677497864, "learning_rate": 9.817778943019228e-05, "loss": 0.0429, "step": 13190 }, { "grad_norm": 0.27001550793647766, "learning_rate": 9.81733636408027e-05, "loss": 0.0429, "step": 13200 }, { "grad_norm": 0.24380020797252655, "learning_rate": 9.816893258324649e-05, "loss": 0.0349, "step": 13210 }, { "grad_norm": 0.31082671880722046, "learning_rate": 9.816449625800823e-05, "loss": 0.0495, "step": 13220 }, { "grad_norm": 0.27921390533447266, "learning_rate": 9.816005466557308e-05, "loss": 0.0379, "step": 13230 }, { "grad_norm": 0.25151991844177246, "learning_rate": 9.815560780642674e-05, "loss": 0.0338, "step": 13240 }, { "grad_norm": 0.3260468542575836, "learning_rate": 9.815115568105555e-05, "loss": 0.046, "step": 13250 }, { "grad_norm": 0.3388170003890991, "learning_rate": 9.814669828994638e-05, "loss": 0.0444, "step": 13260 }, { "grad_norm": 0.31696873903274536, "learning_rate": 9.814223563358665e-05, "loss": 0.048, "step": 13270 }, { "grad_norm": 0.37191078066825867, "learning_rate": 9.813776771246443e-05, "loss": 0.0465, "step": 13280 }, { "grad_norm": 0.41589808464050293, "learning_rate": 9.813329452706829e-05, "loss": 0.0467, "step": 13290 }, { "grad_norm": 0.282308965921402, "learning_rate": 9.812881607788744e-05, "loss": 0.0493, "step": 13300 }, { "grad_norm": 0.34423789381980896, "learning_rate": 9.812433236541163e-05, "loss": 0.049, "step": 13310 }, { "grad_norm": 0.295330673456192, "learning_rate": 9.811984339013116e-05, "loss": 0.042, "step": 13320 }, { "grad_norm": 0.29314857721328735, "learning_rate": 9.811534915253698e-05, "loss": 0.0509, "step": 13330 }, { "grad_norm": 0.3103221654891968, "learning_rate": 9.811084965312056e-05, "loss": 0.044, "step": 13340 }, { "grad_norm": 0.3005836606025696, "learning_rate": 9.810634489237396e-05, "loss": 0.046, "step": 13350 }, { "grad_norm": 0.2893420159816742, "learning_rate": 9.81018348707898e-05, "loss": 0.0427, "step": 13360 }, { "grad_norm": 0.23020225763320923, "learning_rate": 9.809731958886131e-05, "loss": 0.0444, "step": 13370 }, { "grad_norm": 0.3316740393638611, "learning_rate": 9.809279904708224e-05, "loss": 0.0478, "step": 13380 }, { "grad_norm": 0.2301543653011322, "learning_rate": 9.808827324594699e-05, "loss": 0.0448, "step": 13390 }, { "grad_norm": 0.30426159501075745, "learning_rate": 9.808374218595046e-05, "loss": 0.0437, "step": 13400 }, { "grad_norm": 0.24157923460006714, "learning_rate": 9.80792058675882e-05, "loss": 0.0416, "step": 13410 }, { "grad_norm": 0.35914933681488037, "learning_rate": 9.807466429135627e-05, "loss": 0.046, "step": 13420 }, { "grad_norm": 0.2907027006149292, "learning_rate": 9.807011745775132e-05, "loss": 0.0424, "step": 13430 }, { "grad_norm": 0.18922829627990723, "learning_rate": 9.806556536727061e-05, "loss": 0.0362, "step": 13440 }, { "grad_norm": 0.2995680272579193, "learning_rate": 9.806100802041193e-05, "loss": 0.04, "step": 13450 }, { "grad_norm": 0.27273309230804443, "learning_rate": 9.805644541767368e-05, "loss": 0.0393, "step": 13460 }, { "grad_norm": 0.30237090587615967, "learning_rate": 9.805187755955478e-05, "loss": 0.0346, "step": 13470 }, { "grad_norm": 0.2536527216434479, "learning_rate": 9.804730444655483e-05, "loss": 0.0414, "step": 13480 }, { "grad_norm": 0.32595089077949524, "learning_rate": 9.804272607917388e-05, "loss": 0.0459, "step": 13490 }, { "grad_norm": 0.26063066720962524, "learning_rate": 9.803814245791265e-05, "loss": 0.0377, "step": 13500 }, { "grad_norm": 0.29079118371009827, "learning_rate": 9.803355358327239e-05, "loss": 0.0407, "step": 13510 }, { "grad_norm": 0.22678792476654053, "learning_rate": 9.802895945575492e-05, "loss": 0.0408, "step": 13520 }, { "grad_norm": 0.2923998236656189, "learning_rate": 9.802436007586266e-05, "loss": 0.0406, "step": 13530 }, { "grad_norm": 0.3130406141281128, "learning_rate": 9.801975544409858e-05, "loss": 0.0382, "step": 13540 }, { "grad_norm": 0.3032777011394501, "learning_rate": 9.801514556096625e-05, "loss": 0.0419, "step": 13550 }, { "grad_norm": 0.2933416962623596, "learning_rate": 9.801053042696977e-05, "loss": 0.0417, "step": 13560 }, { "grad_norm": 0.291189968585968, "learning_rate": 9.800591004261388e-05, "loss": 0.0357, "step": 13570 }, { "grad_norm": 0.6274422407150269, "learning_rate": 9.800128440840385e-05, "loss": 0.0375, "step": 13580 }, { "grad_norm": 0.34813326597213745, "learning_rate": 9.799665352484552e-05, "loss": 0.0395, "step": 13590 }, { "grad_norm": 0.30105918645858765, "learning_rate": 9.799201739244532e-05, "loss": 0.0363, "step": 13600 }, { "grad_norm": 0.21829922497272491, "learning_rate": 9.798737601171025e-05, "loss": 0.0339, "step": 13610 }, { "grad_norm": 0.2657618522644043, "learning_rate": 9.79827293831479e-05, "loss": 0.0347, "step": 13620 }, { "grad_norm": 0.5879572033882141, "learning_rate": 9.797807750726638e-05, "loss": 0.0438, "step": 13630 }, { "grad_norm": 0.3043363690376282, "learning_rate": 9.797342038457446e-05, "loss": 0.0419, "step": 13640 }, { "grad_norm": 0.31797608733177185, "learning_rate": 9.796875801558141e-05, "loss": 0.0365, "step": 13650 }, { "grad_norm": 0.3438408076763153, "learning_rate": 9.79640904007971e-05, "loss": 0.0435, "step": 13660 }, { "grad_norm": 0.20584863424301147, "learning_rate": 9.795941754073199e-05, "loss": 0.0389, "step": 13670 }, { "grad_norm": 0.3303311765193939, "learning_rate": 9.795473943589705e-05, "loss": 0.0423, "step": 13680 }, { "grad_norm": 0.3294457495212555, "learning_rate": 9.795005608680394e-05, "loss": 0.0431, "step": 13690 }, { "grad_norm": 0.3226814568042755, "learning_rate": 9.794536749396477e-05, "loss": 0.0464, "step": 13700 }, { "grad_norm": 0.38860079646110535, "learning_rate": 9.79406736578923e-05, "loss": 0.049, "step": 13710 }, { "grad_norm": 0.279552698135376, "learning_rate": 9.793597457909984e-05, "loss": 0.0508, "step": 13720 }, { "grad_norm": 0.3639621138572693, "learning_rate": 9.793127025810127e-05, "loss": 0.0433, "step": 13730 }, { "grad_norm": 0.2346642017364502, "learning_rate": 9.792656069541104e-05, "loss": 0.0404, "step": 13740 }, { "grad_norm": 0.42764297127723694, "learning_rate": 9.79218458915442e-05, "loss": 0.0583, "step": 13750 }, { "grad_norm": 0.2605259120464325, "learning_rate": 9.791712584701634e-05, "loss": 0.047, "step": 13760 }, { "grad_norm": 0.2562156915664673, "learning_rate": 9.791240056234364e-05, "loss": 0.0407, "step": 13770 }, { "grad_norm": 0.21532519161701202, "learning_rate": 9.790767003804283e-05, "loss": 0.035, "step": 13780 }, { "grad_norm": 0.288505882024765, "learning_rate": 9.790293427463126e-05, "loss": 0.0382, "step": 13790 }, { "grad_norm": 0.30207911133766174, "learning_rate": 9.789819327262684e-05, "loss": 0.0399, "step": 13800 }, { "grad_norm": 0.37445715069770813, "learning_rate": 9.7893447032548e-05, "loss": 0.0477, "step": 13810 }, { "grad_norm": 0.34875479340553284, "learning_rate": 9.78886955549138e-05, "loss": 0.0512, "step": 13820 }, { "grad_norm": 0.3552783727645874, "learning_rate": 9.788393884024387e-05, "loss": 0.0514, "step": 13830 }, { "grad_norm": 0.2838611602783203, "learning_rate": 9.787917688905836e-05, "loss": 0.0427, "step": 13840 }, { "grad_norm": 0.3255988657474518, "learning_rate": 9.787440970187807e-05, "loss": 0.0379, "step": 13850 }, { "grad_norm": 0.2776106894016266, "learning_rate": 9.786963727922429e-05, "loss": 0.0423, "step": 13860 }, { "grad_norm": 0.32953956723213196, "learning_rate": 9.786485962161897e-05, "loss": 0.0403, "step": 13870 }, { "grad_norm": 0.2689048945903778, "learning_rate": 9.786007672958455e-05, "loss": 0.0397, "step": 13880 }, { "grad_norm": 0.2881467044353485, "learning_rate": 9.78552886036441e-05, "loss": 0.0427, "step": 13890 }, { "grad_norm": 0.23218759894371033, "learning_rate": 9.785049524432124e-05, "loss": 0.0392, "step": 13900 }, { "grad_norm": 0.3243424594402313, "learning_rate": 9.784569665214016e-05, "loss": 0.0418, "step": 13910 }, { "grad_norm": 0.30053216218948364, "learning_rate": 9.784089282762563e-05, "loss": 0.04, "step": 13920 }, { "grad_norm": 0.29011839628219604, "learning_rate": 9.7836083771303e-05, "loss": 0.0402, "step": 13930 }, { "grad_norm": 0.2090558409690857, "learning_rate": 9.783126948369817e-05, "loss": 0.0361, "step": 13940 }, { "grad_norm": 0.21426795423030853, "learning_rate": 9.78264499653376e-05, "loss": 0.039, "step": 13950 }, { "grad_norm": 0.2642742991447449, "learning_rate": 9.782162521674838e-05, "loss": 0.0428, "step": 13960 }, { "grad_norm": 0.2571643590927124, "learning_rate": 9.781679523845812e-05, "loss": 0.0423, "step": 13970 }, { "grad_norm": 0.2633998990058899, "learning_rate": 9.781196003099502e-05, "loss": 0.0394, "step": 13980 }, { "grad_norm": 0.30333834886550903, "learning_rate": 9.780711959488786e-05, "loss": 0.05, "step": 13990 }, { "grad_norm": 0.451124906539917, "learning_rate": 9.780227393066599e-05, "loss": 0.0438, "step": 14000 }, { "grad_norm": 0.2502151131629944, "learning_rate": 9.77974230388593e-05, "loss": 0.039, "step": 14010 }, { "grad_norm": 0.2706962525844574, "learning_rate": 9.779256691999829e-05, "loss": 0.0442, "step": 14020 }, { "grad_norm": 0.3546127378940582, "learning_rate": 9.778770557461403e-05, "loss": 0.0355, "step": 14030 }, { "grad_norm": 0.280489057302475, "learning_rate": 9.778283900323812e-05, "loss": 0.0393, "step": 14040 }, { "grad_norm": 0.22038142383098602, "learning_rate": 9.777796720640277e-05, "loss": 0.0379, "step": 14050 }, { "grad_norm": 0.24386246502399445, "learning_rate": 9.777309018464078e-05, "loss": 0.0357, "step": 14060 }, { "grad_norm": 0.22132070362567902, "learning_rate": 9.776820793848547e-05, "loss": 0.0364, "step": 14070 }, { "grad_norm": 0.31869998574256897, "learning_rate": 9.776332046847075e-05, "loss": 0.035, "step": 14080 }, { "grad_norm": 0.3297432065010071, "learning_rate": 9.775842777513111e-05, "loss": 0.0384, "step": 14090 }, { "grad_norm": 0.22806143760681152, "learning_rate": 9.775352985900163e-05, "loss": 0.0412, "step": 14100 }, { "grad_norm": 0.3528369665145874, "learning_rate": 9.774862672061791e-05, "loss": 0.042, "step": 14110 }, { "grad_norm": 0.30983787775039673, "learning_rate": 9.774371836051616e-05, "loss": 0.0474, "step": 14120 }, { "grad_norm": 0.3151307702064514, "learning_rate": 9.773880477923315e-05, "loss": 0.0529, "step": 14130 }, { "grad_norm": 0.3297314941883087, "learning_rate": 9.773388597730623e-05, "loss": 0.0439, "step": 14140 }, { "grad_norm": 0.3018113970756531, "learning_rate": 9.77289619552733e-05, "loss": 0.048, "step": 14150 }, { "grad_norm": 0.3500923812389374, "learning_rate": 9.772403271367285e-05, "loss": 0.0493, "step": 14160 }, { "grad_norm": 0.2851371765136719, "learning_rate": 9.771909825304396e-05, "loss": 0.0469, "step": 14170 }, { "grad_norm": 0.3578742742538452, "learning_rate": 9.771415857392619e-05, "loss": 0.0434, "step": 14180 }, { "grad_norm": 0.263767272233963, "learning_rate": 9.770921367685978e-05, "loss": 0.0327, "step": 14190 }, { "grad_norm": 0.30043232440948486, "learning_rate": 9.770426356238551e-05, "loss": 0.0369, "step": 14200 }, { "grad_norm": 0.2722148597240448, "learning_rate": 9.769930823104469e-05, "loss": 0.0395, "step": 14210 }, { "grad_norm": 0.5369251370429993, "learning_rate": 9.769434768337926e-05, "loss": 0.0396, "step": 14220 }, { "grad_norm": 0.306398868560791, "learning_rate": 9.768938191993164e-05, "loss": 0.039, "step": 14230 }, { "grad_norm": 0.2728380858898163, "learning_rate": 9.768441094124494e-05, "loss": 0.0408, "step": 14240 }, { "grad_norm": 0.26838377118110657, "learning_rate": 9.767943474786275e-05, "loss": 0.041, "step": 14250 }, { "grad_norm": 0.26339641213417053, "learning_rate": 9.767445334032923e-05, "loss": 0.0412, "step": 14260 }, { "grad_norm": 0.2868843972682953, "learning_rate": 9.766946671918919e-05, "loss": 0.038, "step": 14270 }, { "grad_norm": 0.44757142663002014, "learning_rate": 9.766447488498796e-05, "loss": 0.0443, "step": 14280 }, { "grad_norm": 0.34670332074165344, "learning_rate": 9.765947783827139e-05, "loss": 0.0352, "step": 14290 }, { "grad_norm": 0.3103836178779602, "learning_rate": 9.765447557958599e-05, "loss": 0.0466, "step": 14300 }, { "grad_norm": 0.44283434748649597, "learning_rate": 9.764946810947879e-05, "loss": 0.0452, "step": 14310 }, { "grad_norm": 0.3012360632419586, "learning_rate": 9.764445542849738e-05, "loss": 0.0432, "step": 14320 }, { "grad_norm": 0.34938451647758484, "learning_rate": 9.763943753718998e-05, "loss": 0.0425, "step": 14330 }, { "grad_norm": 0.28477251529693604, "learning_rate": 9.76344144361053e-05, "loss": 0.0339, "step": 14340 }, { "grad_norm": 0.26135584712028503, "learning_rate": 9.762938612579269e-05, "loss": 0.0412, "step": 14350 }, { "grad_norm": 0.2714371979236603, "learning_rate": 9.762435260680202e-05, "loss": 0.0398, "step": 14360 }, { "grad_norm": 0.38514620065689087, "learning_rate": 9.761931387968373e-05, "loss": 0.0415, "step": 14370 }, { "grad_norm": 0.30267032980918884, "learning_rate": 9.76142699449889e-05, "loss": 0.0413, "step": 14380 }, { "grad_norm": 0.3067556917667389, "learning_rate": 9.760922080326908e-05, "loss": 0.0395, "step": 14390 }, { "grad_norm": 0.3632432222366333, "learning_rate": 9.760416645507644e-05, "loss": 0.039, "step": 14400 }, { "grad_norm": 0.24422363936901093, "learning_rate": 9.759910690096375e-05, "loss": 0.0401, "step": 14410 }, { "grad_norm": 0.2780703902244568, "learning_rate": 9.759404214148429e-05, "loss": 0.0405, "step": 14420 }, { "grad_norm": 0.3015499711036682, "learning_rate": 9.758897217719191e-05, "loss": 0.04, "step": 14430 }, { "grad_norm": 0.5137867331504822, "learning_rate": 9.758389700864113e-05, "loss": 0.0445, "step": 14440 }, { "grad_norm": 0.34359505772590637, "learning_rate": 9.757881663638688e-05, "loss": 0.0467, "step": 14450 }, { "grad_norm": 0.2510347068309784, "learning_rate": 9.757373106098478e-05, "loss": 0.0442, "step": 14460 }, { "grad_norm": 0.37131306529045105, "learning_rate": 9.756864028299097e-05, "loss": 0.0392, "step": 14470 }, { "grad_norm": 0.25158339738845825, "learning_rate": 9.75635443029622e-05, "loss": 0.0374, "step": 14480 }, { "grad_norm": 0.2397085428237915, "learning_rate": 9.755844312145572e-05, "loss": 0.0352, "step": 14490 }, { "grad_norm": 0.23203833401203156, "learning_rate": 9.755333673902941e-05, "loss": 0.0346, "step": 14500 }, { "grad_norm": 0.2552068531513214, "learning_rate": 9.75482251562417e-05, "loss": 0.0373, "step": 14510 }, { "grad_norm": 0.32669270038604736, "learning_rate": 9.754310837365155e-05, "loss": 0.0391, "step": 14520 }, { "grad_norm": 0.8636684417724609, "learning_rate": 9.753798639181856e-05, "loss": 0.0436, "step": 14530 }, { "grad_norm": 0.30748677253723145, "learning_rate": 9.753285921130286e-05, "loss": 0.0417, "step": 14540 }, { "grad_norm": 0.32175594568252563, "learning_rate": 9.752772683266512e-05, "loss": 0.0402, "step": 14550 }, { "grad_norm": 0.28241434693336487, "learning_rate": 9.752258925646665e-05, "loss": 0.0419, "step": 14560 }, { "grad_norm": 0.2918451428413391, "learning_rate": 9.751744648326926e-05, "loss": 0.0453, "step": 14570 }, { "grad_norm": 0.28916797041893005, "learning_rate": 9.751229851363536e-05, "loss": 0.0421, "step": 14580 }, { "grad_norm": 0.39780646562576294, "learning_rate": 9.750714534812793e-05, "loss": 0.0435, "step": 14590 }, { "grad_norm": 0.2532484829425812, "learning_rate": 9.750198698731053e-05, "loss": 0.0381, "step": 14600 }, { "grad_norm": 0.8770033717155457, "learning_rate": 9.749682343174722e-05, "loss": 0.0419, "step": 14610 }, { "grad_norm": 0.27095749974250793, "learning_rate": 9.749165468200272e-05, "loss": 0.0403, "step": 14620 }, { "grad_norm": 0.3499591052532196, "learning_rate": 9.748648073864229e-05, "loss": 0.0437, "step": 14630 }, { "grad_norm": 0.37834814190864563, "learning_rate": 9.748130160223168e-05, "loss": 0.0449, "step": 14640 }, { "grad_norm": 0.3142000734806061, "learning_rate": 9.747611727333734e-05, "loss": 0.0561, "step": 14650 }, { "grad_norm": 0.29757753014564514, "learning_rate": 9.74709277525262e-05, "loss": 0.0519, "step": 14660 }, { "grad_norm": 0.32495397329330444, "learning_rate": 9.746573304036576e-05, "loss": 0.0489, "step": 14670 }, { "grad_norm": 0.3755247890949249, "learning_rate": 9.746053313742412e-05, "loss": 0.049, "step": 14680 }, { "grad_norm": 0.4149918258190155, "learning_rate": 9.745532804426994e-05, "loss": 0.0457, "step": 14690 }, { "grad_norm": 0.847391664981842, "learning_rate": 9.745011776147242e-05, "loss": 0.0586, "step": 14700 }, { "grad_norm": 0.30256086587905884, "learning_rate": 9.744490228960138e-05, "loss": 0.0526, "step": 14710 }, { "grad_norm": 0.27884095907211304, "learning_rate": 9.743968162922713e-05, "loss": 0.0398, "step": 14720 }, { "grad_norm": 0.3423587381839752, "learning_rate": 9.743445578092064e-05, "loss": 0.0436, "step": 14730 }, { "grad_norm": 0.28879955410957336, "learning_rate": 9.742922474525338e-05, "loss": 0.0407, "step": 14740 }, { "grad_norm": 0.4366281032562256, "learning_rate": 9.742398852279741e-05, "loss": 0.0446, "step": 14750 }, { "grad_norm": 0.31203117966651917, "learning_rate": 9.741874711412535e-05, "loss": 0.0431, "step": 14760 }, { "grad_norm": 0.40611523389816284, "learning_rate": 9.741350051981042e-05, "loss": 0.0477, "step": 14770 }, { "grad_norm": 0.46012023091316223, "learning_rate": 9.740824874042633e-05, "loss": 0.0479, "step": 14780 }, { "grad_norm": 0.4054811894893646, "learning_rate": 9.740299177654746e-05, "loss": 0.0454, "step": 14790 }, { "grad_norm": 0.2982066571712494, "learning_rate": 9.739772962874867e-05, "loss": 0.043, "step": 14800 }, { "grad_norm": 0.39228159189224243, "learning_rate": 9.739246229760541e-05, "loss": 0.0569, "step": 14810 }, { "grad_norm": 0.3382984697818756, "learning_rate": 9.738718978369376e-05, "loss": 0.0434, "step": 14820 }, { "grad_norm": 0.33796176314353943, "learning_rate": 9.738191208759025e-05, "loss": 0.04, "step": 14830 }, { "grad_norm": 0.27693554759025574, "learning_rate": 9.73766292098721e-05, "loss": 0.0381, "step": 14840 }, { "grad_norm": 0.2530422806739807, "learning_rate": 9.737134115111699e-05, "loss": 0.0403, "step": 14850 }, { "grad_norm": 0.5341730713844299, "learning_rate": 9.736604791190323e-05, "loss": 0.0468, "step": 14860 }, { "grad_norm": 0.3224777579307556, "learning_rate": 9.73607494928097e-05, "loss": 0.043, "step": 14870 }, { "grad_norm": 0.29044535756111145, "learning_rate": 9.735544589441581e-05, "loss": 0.0431, "step": 14880 }, { "grad_norm": 0.2471650391817093, "learning_rate": 9.735013711730154e-05, "loss": 0.0417, "step": 14890 }, { "grad_norm": 0.40082409977912903, "learning_rate": 9.734482316204747e-05, "loss": 0.0454, "step": 14900 }, { "grad_norm": 0.28218966722488403, "learning_rate": 9.733950402923473e-05, "loss": 0.0346, "step": 14910 }, { "grad_norm": 0.29190728068351746, "learning_rate": 9.7334179719445e-05, "loss": 0.0422, "step": 14920 }, { "grad_norm": 0.31538674235343933, "learning_rate": 9.732885023326053e-05, "loss": 0.0497, "step": 14930 }, { "grad_norm": 0.384507417678833, "learning_rate": 9.732351557126418e-05, "loss": 0.0522, "step": 14940 }, { "grad_norm": 0.4519524574279785, "learning_rate": 9.731817573403929e-05, "loss": 0.0476, "step": 14950 }, { "grad_norm": 0.2658109962940216, "learning_rate": 9.731283072216985e-05, "loss": 0.0364, "step": 14960 }, { "grad_norm": 0.5292503237724304, "learning_rate": 9.730748053624039e-05, "loss": 0.0462, "step": 14970 }, { "grad_norm": 0.275566041469574, "learning_rate": 9.730212517683598e-05, "loss": 0.0433, "step": 14980 }, { "grad_norm": 0.32696714997291565, "learning_rate": 9.729676464454228e-05, "loss": 0.046, "step": 14990 }, { "grad_norm": 0.2673732340335846, "learning_rate": 9.72913989399455e-05, "loss": 0.0433, "step": 15000 }, { "grad_norm": 0.36184194684028625, "learning_rate": 9.728602806363242e-05, "loss": 0.0384, "step": 15010 }, { "grad_norm": 0.33208417892456055, "learning_rate": 9.728065201619043e-05, "loss": 0.0355, "step": 15020 }, { "grad_norm": 0.24134254455566406, "learning_rate": 9.727527079820742e-05, "loss": 0.0354, "step": 15030 }, { "grad_norm": 0.4159527122974396, "learning_rate": 9.726988441027186e-05, "loss": 0.0406, "step": 15040 }, { "grad_norm": 0.25019675493240356, "learning_rate": 9.726449285297281e-05, "loss": 0.0343, "step": 15050 }, { "grad_norm": 0.3649773895740509, "learning_rate": 9.72590961268999e-05, "loss": 0.044, "step": 15060 }, { "grad_norm": 0.3191484808921814, "learning_rate": 9.725369423264328e-05, "loss": 0.0354, "step": 15070 }, { "grad_norm": 0.3009691834449768, "learning_rate": 9.72482871707937e-05, "loss": 0.0369, "step": 15080 }, { "grad_norm": 0.27893027663230896, "learning_rate": 9.724287494194247e-05, "loss": 0.0324, "step": 15090 }, { "grad_norm": 0.18120907247066498, "learning_rate": 9.723745754668147e-05, "loss": 0.0357, "step": 15100 }, { "grad_norm": 0.2442377209663391, "learning_rate": 9.723203498560313e-05, "loss": 0.0357, "step": 15110 }, { "grad_norm": 0.32900509238243103, "learning_rate": 9.722660725930046e-05, "loss": 0.0449, "step": 15120 }, { "grad_norm": 0.5861304402351379, "learning_rate": 9.722117436836702e-05, "loss": 0.0443, "step": 15130 }, { "grad_norm": 0.31291380524635315, "learning_rate": 9.721573631339696e-05, "loss": 0.0388, "step": 15140 }, { "grad_norm": 0.49981990456581116, "learning_rate": 9.721029309498494e-05, "loss": 0.0391, "step": 15150 }, { "grad_norm": 0.27632802724838257, "learning_rate": 9.720484471372627e-05, "loss": 0.0412, "step": 15160 }, { "grad_norm": 0.2563539147377014, "learning_rate": 9.719939117021673e-05, "loss": 0.0394, "step": 15170 }, { "grad_norm": 0.23698341846466064, "learning_rate": 9.719393246505275e-05, "loss": 0.0357, "step": 15180 }, { "grad_norm": 0.27009713649749756, "learning_rate": 9.718846859883128e-05, "loss": 0.0386, "step": 15190 }, { "grad_norm": 0.2796517014503479, "learning_rate": 9.718299957214982e-05, "loss": 0.0337, "step": 15200 }, { "grad_norm": 0.3736773729324341, "learning_rate": 9.717752538560646e-05, "loss": 0.0381, "step": 15210 }, { "grad_norm": 0.28872156143188477, "learning_rate": 9.717204603979986e-05, "loss": 0.0376, "step": 15220 }, { "grad_norm": 0.42869678139686584, "learning_rate": 9.716656153532922e-05, "loss": 0.0463, "step": 15230 }, { "grad_norm": 0.2753220200538635, "learning_rate": 9.716107187279434e-05, "loss": 0.0441, "step": 15240 }, { "grad_norm": 0.2556781768798828, "learning_rate": 9.715557705279555e-05, "loss": 0.0353, "step": 15250 }, { "grad_norm": 0.3013724386692047, "learning_rate": 9.715007707593372e-05, "loss": 0.0441, "step": 15260 }, { "grad_norm": 0.5189555883407593, "learning_rate": 9.714457194281036e-05, "loss": 0.0545, "step": 15270 }, { "grad_norm": 0.41351038217544556, "learning_rate": 9.713906165402751e-05, "loss": 0.0556, "step": 15280 }, { "grad_norm": 0.45453694462776184, "learning_rate": 9.713354621018774e-05, "loss": 0.0503, "step": 15290 }, { "grad_norm": 0.27930110692977905, "learning_rate": 9.712802561189422e-05, "loss": 0.0485, "step": 15300 }, { "grad_norm": 0.29430875182151794, "learning_rate": 9.712249985975069e-05, "loss": 0.0426, "step": 15310 }, { "grad_norm": 0.26044994592666626, "learning_rate": 9.71169689543614e-05, "loss": 0.0374, "step": 15320 }, { "grad_norm": 0.30679431557655334, "learning_rate": 9.711143289633123e-05, "loss": 0.0398, "step": 15330 }, { "grad_norm": 0.23104727268218994, "learning_rate": 9.710589168626561e-05, "loss": 0.0358, "step": 15340 }, { "grad_norm": 0.29458358883857727, "learning_rate": 9.710034532477048e-05, "loss": 0.0387, "step": 15350 }, { "grad_norm": 0.44156721234321594, "learning_rate": 9.709479381245239e-05, "loss": 0.0412, "step": 15360 }, { "grad_norm": 0.36855030059814453, "learning_rate": 9.708923714991847e-05, "loss": 0.0437, "step": 15370 }, { "grad_norm": 0.32313480973243713, "learning_rate": 9.708367533777638e-05, "loss": 0.0338, "step": 15380 }, { "grad_norm": 0.30193954706192017, "learning_rate": 9.707810837663431e-05, "loss": 0.0357, "step": 15390 }, { "grad_norm": 0.24447765946388245, "learning_rate": 9.707253626710113e-05, "loss": 0.0372, "step": 15400 }, { "grad_norm": 0.31228259205818176, "learning_rate": 9.706695900978613e-05, "loss": 0.0355, "step": 15410 }, { "grad_norm": 0.24583600461483002, "learning_rate": 9.706137660529926e-05, "loss": 0.0417, "step": 15420 }, { "grad_norm": 0.29451823234558105, "learning_rate": 9.705578905425101e-05, "loss": 0.0302, "step": 15430 }, { "grad_norm": 0.22824908792972565, "learning_rate": 9.705019635725241e-05, "loss": 0.0361, "step": 15440 }, { "grad_norm": 0.27998286485671997, "learning_rate": 9.704459851491508e-05, "loss": 0.0411, "step": 15450 }, { "grad_norm": 0.3220147490501404, "learning_rate": 9.703899552785118e-05, "loss": 0.0435, "step": 15460 }, { "grad_norm": 0.2967126667499542, "learning_rate": 9.703338739667346e-05, "loss": 0.0447, "step": 15470 }, { "grad_norm": 0.2628369927406311, "learning_rate": 9.70277741219952e-05, "loss": 0.0395, "step": 15480 }, { "grad_norm": 0.254682332277298, "learning_rate": 9.702215570443027e-05, "loss": 0.0412, "step": 15490 }, { "grad_norm": 0.20976312458515167, "learning_rate": 9.701653214459309e-05, "loss": 0.0433, "step": 15500 }, { "grad_norm": 0.26191452145576477, "learning_rate": 9.701090344309865e-05, "loss": 0.0479, "step": 15510 }, { "grad_norm": 0.3612656891345978, "learning_rate": 9.700526960056247e-05, "loss": 0.0435, "step": 15520 }, { "grad_norm": 0.32214051485061646, "learning_rate": 9.699963061760068e-05, "loss": 0.0393, "step": 15530 }, { "grad_norm": 0.24570608139038086, "learning_rate": 9.699398649482997e-05, "loss": 0.0362, "step": 15540 }, { "grad_norm": 0.2214808464050293, "learning_rate": 9.698833723286753e-05, "loss": 0.0391, "step": 15550 }, { "grad_norm": 0.41092655062675476, "learning_rate": 9.698268283233118e-05, "loss": 0.0375, "step": 15560 }, { "grad_norm": 0.24396882951259613, "learning_rate": 9.697702329383929e-05, "loss": 0.0347, "step": 15570 }, { "grad_norm": 0.24450844526290894, "learning_rate": 9.697135861801074e-05, "loss": 0.0334, "step": 15580 }, { "grad_norm": 0.3383522629737854, "learning_rate": 9.696568880546505e-05, "loss": 0.0399, "step": 15590 }, { "grad_norm": 0.3362794816493988, "learning_rate": 9.696001385682223e-05, "loss": 0.0367, "step": 15600 }, { "grad_norm": 0.28977248072624207, "learning_rate": 9.695433377270291e-05, "loss": 0.0453, "step": 15610 }, { "grad_norm": 0.29530632495880127, "learning_rate": 9.694864855372824e-05, "loss": 0.0344, "step": 15620 }, { "grad_norm": 0.19267421960830688, "learning_rate": 9.694295820051995e-05, "loss": 0.0316, "step": 15630 }, { "grad_norm": 0.2485141158103943, "learning_rate": 9.693726271370032e-05, "loss": 0.0324, "step": 15640 }, { "grad_norm": 0.2635056972503662, "learning_rate": 9.693156209389221e-05, "loss": 0.0337, "step": 15650 }, { "grad_norm": 0.25441643595695496, "learning_rate": 9.692585634171905e-05, "loss": 0.036, "step": 15660 }, { "grad_norm": 0.2757011651992798, "learning_rate": 9.692014545780476e-05, "loss": 0.0363, "step": 15670 }, { "grad_norm": 0.30695775151252747, "learning_rate": 9.691442944277393e-05, "loss": 0.0354, "step": 15680 }, { "grad_norm": 0.22698193788528442, "learning_rate": 9.690870829725162e-05, "loss": 0.0364, "step": 15690 }, { "grad_norm": 0.2864125967025757, "learning_rate": 9.69029820218635e-05, "loss": 0.0364, "step": 15700 }, { "grad_norm": 0.2858966588973999, "learning_rate": 9.689725061723579e-05, "loss": 0.0348, "step": 15710 }, { "grad_norm": 0.36064648628234863, "learning_rate": 9.689151408399527e-05, "loss": 0.0386, "step": 15720 }, { "grad_norm": 0.322225958108902, "learning_rate": 9.688577242276924e-05, "loss": 0.044, "step": 15730 }, { "grad_norm": 0.40042591094970703, "learning_rate": 9.688002563418566e-05, "loss": 0.0452, "step": 15740 }, { "grad_norm": 0.30461955070495605, "learning_rate": 9.687427371887293e-05, "loss": 0.0406, "step": 15750 }, { "grad_norm": 0.3791019320487976, "learning_rate": 9.686851667746012e-05, "loss": 0.0403, "step": 15760 }, { "grad_norm": 0.3452751934528351, "learning_rate": 9.686275451057677e-05, "loss": 0.0394, "step": 15770 }, { "grad_norm": 0.3412947952747345, "learning_rate": 9.685698721885308e-05, "loss": 0.0337, "step": 15780 }, { "grad_norm": 0.28363874554634094, "learning_rate": 9.68512148029197e-05, "loss": 0.0443, "step": 15790 }, { "grad_norm": 0.2597404420375824, "learning_rate": 9.684543726340791e-05, "loss": 0.0347, "step": 15800 }, { "grad_norm": 0.6070683002471924, "learning_rate": 9.683965460094952e-05, "loss": 0.0433, "step": 15810 }, { "grad_norm": 0.31155768036842346, "learning_rate": 9.683386681617694e-05, "loss": 0.0515, "step": 15820 }, { "grad_norm": 0.26583319902420044, "learning_rate": 9.68280739097231e-05, "loss": 0.0539, "step": 15830 }, { "grad_norm": 0.3668738007545471, "learning_rate": 9.682227588222148e-05, "loss": 0.048, "step": 15840 }, { "grad_norm": 0.2869052290916443, "learning_rate": 9.681647273430618e-05, "loss": 0.0517, "step": 15850 }, { "grad_norm": 0.3761570155620575, "learning_rate": 9.681066446661182e-05, "loss": 0.0475, "step": 15860 }, { "grad_norm": 0.20043757557868958, "learning_rate": 9.680485107977357e-05, "loss": 0.0426, "step": 15870 }, { "grad_norm": 0.24509774148464203, "learning_rate": 9.679903257442716e-05, "loss": 0.0388, "step": 15880 }, { "grad_norm": 0.31805989146232605, "learning_rate": 9.679320895120891e-05, "loss": 0.0371, "step": 15890 }, { "grad_norm": 0.34148210287094116, "learning_rate": 9.67873802107557e-05, "loss": 0.0362, "step": 15900 }, { "grad_norm": 0.31961724162101746, "learning_rate": 9.67815463537049e-05, "loss": 0.0361, "step": 15910 }, { "grad_norm": 0.3294147253036499, "learning_rate": 9.677570738069457e-05, "loss": 0.0382, "step": 15920 }, { "grad_norm": 0.4555065333843231, "learning_rate": 9.676986329236318e-05, "loss": 0.0451, "step": 15930 }, { "grad_norm": 0.3084072768688202, "learning_rate": 9.676401408934987e-05, "loss": 0.042, "step": 15940 }, { "grad_norm": 0.885594367980957, "learning_rate": 9.675815977229428e-05, "loss": 0.0446, "step": 15950 }, { "grad_norm": 0.26858338713645935, "learning_rate": 9.675230034183664e-05, "loss": 0.0407, "step": 15960 }, { "grad_norm": 0.2966618835926056, "learning_rate": 9.674643579861773e-05, "loss": 0.0401, "step": 15970 }, { "grad_norm": 0.31391939520835876, "learning_rate": 9.674056614327886e-05, "loss": 0.0398, "step": 15980 }, { "grad_norm": 0.2141139954328537, "learning_rate": 9.673469137646198e-05, "loss": 0.0364, "step": 15990 }, { "grad_norm": 0.2683178782463074, "learning_rate": 9.67288114988095e-05, "loss": 0.0338, "step": 16000 }, { "grad_norm": 0.3328778147697449, "learning_rate": 9.672292651096447e-05, "loss": 0.0407, "step": 16010 }, { "grad_norm": 0.27951905131340027, "learning_rate": 9.671703641357042e-05, "loss": 0.0394, "step": 16020 }, { "grad_norm": 0.28616172075271606, "learning_rate": 9.67111412072715e-05, "loss": 0.0471, "step": 16030 }, { "grad_norm": 0.3493332862854004, "learning_rate": 9.670524089271242e-05, "loss": 0.044, "step": 16040 }, { "grad_norm": 0.36170104146003723, "learning_rate": 9.669933547053842e-05, "loss": 0.0531, "step": 16050 }, { "grad_norm": 0.3833145499229431, "learning_rate": 9.669342494139531e-05, "loss": 0.0487, "step": 16060 }, { "grad_norm": 0.3263576626777649, "learning_rate": 9.668750930592943e-05, "loss": 0.0445, "step": 16070 }, { "grad_norm": 0.5375795960426331, "learning_rate": 9.668158856478775e-05, "loss": 0.0492, "step": 16080 }, { "grad_norm": 0.31985533237457275, "learning_rate": 9.66756627186177e-05, "loss": 0.0426, "step": 16090 }, { "grad_norm": 0.3551134169101715, "learning_rate": 9.666973176806737e-05, "loss": 0.0429, "step": 16100 }, { "grad_norm": 0.31270840764045715, "learning_rate": 9.666379571378534e-05, "loss": 0.0413, "step": 16110 }, { "grad_norm": 0.3748157024383545, "learning_rate": 9.665785455642076e-05, "loss": 0.038, "step": 16120 }, { "grad_norm": 0.2827254831790924, "learning_rate": 9.665190829662337e-05, "loss": 0.0375, "step": 16130 }, { "grad_norm": 0.25181248784065247, "learning_rate": 9.664595693504342e-05, "loss": 0.0375, "step": 16140 }, { "grad_norm": 0.33546939492225647, "learning_rate": 9.664000047233175e-05, "loss": 0.0393, "step": 16150 }, { "grad_norm": 0.32153066992759705, "learning_rate": 9.663403890913976e-05, "loss": 0.0421, "step": 16160 }, { "grad_norm": 0.45045262575149536, "learning_rate": 9.662807224611938e-05, "loss": 0.0451, "step": 16170 }, { "grad_norm": 0.3416809141635895, "learning_rate": 9.662210048392311e-05, "loss": 0.0413, "step": 16180 }, { "grad_norm": 0.32954639196395874, "learning_rate": 9.661612362320405e-05, "loss": 0.0385, "step": 16190 }, { "grad_norm": 0.5092183351516724, "learning_rate": 9.661014166461579e-05, "loss": 0.0415, "step": 16200 }, { "grad_norm": 0.27594485878944397, "learning_rate": 9.66041546088125e-05, "loss": 0.0479, "step": 16210 }, { "grad_norm": 0.24632953107357025, "learning_rate": 9.659816245644895e-05, "loss": 0.0341, "step": 16220 }, { "grad_norm": 0.3044162094593048, "learning_rate": 9.65921652081804e-05, "loss": 0.0362, "step": 16230 }, { "grad_norm": 0.2738407552242279, "learning_rate": 9.658616286466271e-05, "loss": 0.0487, "step": 16240 }, { "grad_norm": 0.24157637357711792, "learning_rate": 9.65801554265523e-05, "loss": 0.0398, "step": 16250 }, { "grad_norm": 0.2710169553756714, "learning_rate": 9.657414289450612e-05, "loss": 0.0345, "step": 16260 }, { "grad_norm": 0.2736155092716217, "learning_rate": 9.656812526918171e-05, "loss": 0.0391, "step": 16270 }, { "grad_norm": 0.32593050599098206, "learning_rate": 9.656210255123712e-05, "loss": 0.0347, "step": 16280 }, { "grad_norm": 0.2943076193332672, "learning_rate": 9.6556074741331e-05, "loss": 0.0365, "step": 16290 }, { "grad_norm": 0.3295634984970093, "learning_rate": 9.655004184012256e-05, "loss": 0.0386, "step": 16300 }, { "grad_norm": 0.23401133716106415, "learning_rate": 9.654400384827152e-05, "loss": 0.037, "step": 16310 }, { "grad_norm": 0.2183016836643219, "learning_rate": 9.653796076643818e-05, "loss": 0.0365, "step": 16320 }, { "grad_norm": 0.3275575041770935, "learning_rate": 9.653191259528344e-05, "loss": 0.0414, "step": 16330 }, { "grad_norm": 0.3449890613555908, "learning_rate": 9.65258593354687e-05, "loss": 0.0367, "step": 16340 }, { "grad_norm": 0.25197798013687134, "learning_rate": 9.651980098765591e-05, "loss": 0.0367, "step": 16350 }, { "grad_norm": 0.35997670888900757, "learning_rate": 9.651373755250765e-05, "loss": 0.038, "step": 16360 }, { "grad_norm": 0.28068217635154724, "learning_rate": 9.650766903068697e-05, "loss": 0.0369, "step": 16370 }, { "grad_norm": 0.36443012952804565, "learning_rate": 9.650159542285753e-05, "loss": 0.0427, "step": 16380 }, { "grad_norm": 0.3562360405921936, "learning_rate": 9.649551672968353e-05, "loss": 0.0378, "step": 16390 }, { "grad_norm": 0.22084178030490875, "learning_rate": 9.648943295182973e-05, "loss": 0.0405, "step": 16400 }, { "grad_norm": 0.31802359223365784, "learning_rate": 9.648334408996144e-05, "loss": 0.0319, "step": 16410 }, { "grad_norm": 0.22803133726119995, "learning_rate": 9.647725014474452e-05, "loss": 0.0309, "step": 16420 }, { "grad_norm": 0.2937418818473816, "learning_rate": 9.64711511168454e-05, "loss": 0.0358, "step": 16430 }, { "grad_norm": 0.3072397708892822, "learning_rate": 9.646504700693108e-05, "loss": 0.0328, "step": 16440 }, { "grad_norm": 0.17744971811771393, "learning_rate": 9.645893781566907e-05, "loss": 0.0336, "step": 16450 }, { "grad_norm": 0.21469014883041382, "learning_rate": 9.645282354372744e-05, "loss": 0.0342, "step": 16460 }, { "grad_norm": 0.3055298924446106, "learning_rate": 9.644670419177491e-05, "loss": 0.035, "step": 16470 }, { "grad_norm": 0.26022273302078247, "learning_rate": 9.644057976048062e-05, "loss": 0.0289, "step": 16480 }, { "grad_norm": 0.3093717396259308, "learning_rate": 9.643445025051435e-05, "loss": 0.0345, "step": 16490 }, { "grad_norm": 0.3367266058921814, "learning_rate": 9.642831566254641e-05, "loss": 0.0351, "step": 16500 }, { "grad_norm": 0.43631693720817566, "learning_rate": 9.642217599724769e-05, "loss": 0.0403, "step": 16510 }, { "grad_norm": 0.2278498113155365, "learning_rate": 9.64160312552896e-05, "loss": 0.0364, "step": 16520 }, { "grad_norm": 0.5348236560821533, "learning_rate": 9.64098814373441e-05, "loss": 0.0403, "step": 16530 }, { "grad_norm": 0.4045640826225281, "learning_rate": 9.640372654408374e-05, "loss": 0.0442, "step": 16540 }, { "grad_norm": 0.25392043590545654, "learning_rate": 9.639756657618162e-05, "loss": 0.034, "step": 16550 }, { "grad_norm": 0.2946036458015442, "learning_rate": 9.639140153431138e-05, "loss": 0.0373, "step": 16560 }, { "grad_norm": 0.265779048204422, "learning_rate": 9.638523141914721e-05, "loss": 0.0336, "step": 16570 }, { "grad_norm": 0.23796555399894714, "learning_rate": 9.637905623136388e-05, "loss": 0.0351, "step": 16580 }, { "grad_norm": 0.2285798192024231, "learning_rate": 9.637287597163669e-05, "loss": 0.03, "step": 16590 }, { "grad_norm": 0.19815309345722198, "learning_rate": 9.63666906406415e-05, "loss": 0.0311, "step": 16600 }, { "grad_norm": 0.3363135755062103, "learning_rate": 9.636050023905473e-05, "loss": 0.0369, "step": 16610 }, { "grad_norm": 0.1903875172138214, "learning_rate": 9.635430476755336e-05, "loss": 0.0316, "step": 16620 }, { "grad_norm": 0.30136749148368835, "learning_rate": 9.63481042268149e-05, "loss": 0.0373, "step": 16630 }, { "grad_norm": 0.257953941822052, "learning_rate": 9.634189861751745e-05, "loss": 0.034, "step": 16640 }, { "grad_norm": 0.22215628623962402, "learning_rate": 9.633568794033967e-05, "loss": 0.036, "step": 16650 }, { "grad_norm": 0.1888757199048996, "learning_rate": 9.63294721959607e-05, "loss": 0.0316, "step": 16660 }, { "grad_norm": 0.28030550479888916, "learning_rate": 9.63232513850603e-05, "loss": 0.0356, "step": 16670 }, { "grad_norm": 0.2963862419128418, "learning_rate": 9.631702550831878e-05, "loss": 0.0336, "step": 16680 }, { "grad_norm": 0.25413578748703003, "learning_rate": 9.631079456641698e-05, "loss": 0.033, "step": 16690 }, { "grad_norm": 0.3033553957939148, "learning_rate": 9.630455856003632e-05, "loss": 0.0348, "step": 16700 }, { "grad_norm": 0.2434293031692505, "learning_rate": 9.629831748985876e-05, "loss": 0.0404, "step": 16710 }, { "grad_norm": 0.2811605930328369, "learning_rate": 9.629207135656679e-05, "loss": 0.0355, "step": 16720 }, { "grad_norm": 0.2140277475118637, "learning_rate": 9.628582016084353e-05, "loss": 0.0375, "step": 16730 }, { "grad_norm": 0.2920760214328766, "learning_rate": 9.627956390337254e-05, "loss": 0.0349, "step": 16740 }, { "grad_norm": 0.2527925670146942, "learning_rate": 9.627330258483802e-05, "loss": 0.0303, "step": 16750 }, { "grad_norm": 0.2648579776287079, "learning_rate": 9.62670362059247e-05, "loss": 0.0372, "step": 16760 }, { "grad_norm": 0.2789354622364044, "learning_rate": 9.626076476731786e-05, "loss": 0.0296, "step": 16770 }, { "grad_norm": 0.3989693224430084, "learning_rate": 9.625448826970336e-05, "loss": 0.0381, "step": 16780 }, { "grad_norm": 0.28207388520240784, "learning_rate": 9.624820671376755e-05, "loss": 0.0365, "step": 16790 }, { "grad_norm": 0.23578354716300964, "learning_rate": 9.62419201001974e-05, "loss": 0.0363, "step": 16800 }, { "grad_norm": 0.28331103920936584, "learning_rate": 9.623562842968037e-05, "loss": 0.0329, "step": 16810 }, { "grad_norm": 0.35145071148872375, "learning_rate": 9.622933170290454e-05, "loss": 0.0371, "step": 16820 }, { "grad_norm": 0.6222256422042847, "learning_rate": 9.622302992055849e-05, "loss": 0.0396, "step": 16830 }, { "grad_norm": 0.36512893438339233, "learning_rate": 9.62167230833314e-05, "loss": 0.0419, "step": 16840 }, { "grad_norm": 0.23719419538974762, "learning_rate": 9.621041119191295e-05, "loss": 0.0344, "step": 16850 }, { "grad_norm": 0.31327494978904724, "learning_rate": 9.620409424699342e-05, "loss": 0.0384, "step": 16860 }, { "grad_norm": 0.32408103346824646, "learning_rate": 9.619777224926359e-05, "loss": 0.0323, "step": 16870 }, { "grad_norm": 0.2708199918270111, "learning_rate": 9.619144519941485e-05, "loss": 0.041, "step": 16880 }, { "grad_norm": 0.20450621843338013, "learning_rate": 9.618511309813912e-05, "loss": 0.0402, "step": 16890 }, { "grad_norm": 0.325431764125824, "learning_rate": 9.617877594612886e-05, "loss": 0.0357, "step": 16900 }, { "grad_norm": 0.2656888961791992, "learning_rate": 9.617243374407707e-05, "loss": 0.0357, "step": 16910 }, { "grad_norm": 0.37342146039009094, "learning_rate": 9.616608649267736e-05, "loss": 0.0421, "step": 16920 }, { "grad_norm": 0.2893260419368744, "learning_rate": 9.615973419262385e-05, "loss": 0.0405, "step": 16930 }, { "grad_norm": 0.4189290702342987, "learning_rate": 9.615337684461119e-05, "loss": 0.0413, "step": 16940 }, { "grad_norm": 0.36699435114860535, "learning_rate": 9.614701444933465e-05, "loss": 0.0379, "step": 16950 }, { "grad_norm": 0.3116486966609955, "learning_rate": 9.614064700748997e-05, "loss": 0.0351, "step": 16960 }, { "grad_norm": 0.23278628289699554, "learning_rate": 9.613427451977352e-05, "loss": 0.0357, "step": 16970 }, { "grad_norm": 0.4036838114261627, "learning_rate": 9.612789698688216e-05, "loss": 0.0431, "step": 16980 }, { "grad_norm": 0.29615020751953125, "learning_rate": 9.612151440951334e-05, "loss": 0.0352, "step": 16990 }, { "grad_norm": 0.345175564289093, "learning_rate": 9.611512678836506e-05, "loss": 0.0392, "step": 17000 }, { "grad_norm": 0.32458987832069397, "learning_rate": 9.610873412413584e-05, "loss": 0.0413, "step": 17010 }, { "grad_norm": 0.25788402557373047, "learning_rate": 9.610233641752476e-05, "loss": 0.0411, "step": 17020 }, { "grad_norm": 0.27099716663360596, "learning_rate": 9.609593366923151e-05, "loss": 0.0399, "step": 17030 }, { "grad_norm": 0.19645260274410248, "learning_rate": 9.608952587995625e-05, "loss": 0.0399, "step": 17040 }, { "grad_norm": 0.5289287567138672, "learning_rate": 9.608311305039972e-05, "loss": 0.0416, "step": 17050 }, { "grad_norm": 0.3305412828922272, "learning_rate": 9.607669518126326e-05, "loss": 0.0405, "step": 17060 }, { "grad_norm": 0.28949689865112305, "learning_rate": 9.607027227324866e-05, "loss": 0.0424, "step": 17070 }, { "grad_norm": 0.3357350528240204, "learning_rate": 9.606384432705837e-05, "loss": 0.0368, "step": 17080 }, { "grad_norm": 0.3265174925327301, "learning_rate": 9.60574113433953e-05, "loss": 0.0408, "step": 17090 }, { "grad_norm": 0.24820464849472046, "learning_rate": 9.6050973322963e-05, "loss": 0.0333, "step": 17100 }, { "grad_norm": 0.26016148924827576, "learning_rate": 9.604453026646547e-05, "loss": 0.0397, "step": 17110 }, { "grad_norm": 0.27775371074676514, "learning_rate": 9.603808217460735e-05, "loss": 0.0359, "step": 17120 }, { "grad_norm": 0.5341002345085144, "learning_rate": 9.603162904809377e-05, "loss": 0.0356, "step": 17130 }, { "grad_norm": 0.275718092918396, "learning_rate": 9.602517088763045e-05, "loss": 0.0326, "step": 17140 }, { "grad_norm": 0.6744633913040161, "learning_rate": 9.601870769392365e-05, "loss": 0.0362, "step": 17150 }, { "grad_norm": 0.2777157127857208, "learning_rate": 9.601223946768017e-05, "loss": 0.0388, "step": 17160 }, { "grad_norm": 0.30920830368995667, "learning_rate": 9.600576620960734e-05, "loss": 0.0398, "step": 17170 }, { "grad_norm": 0.2796286940574646, "learning_rate": 9.599928792041308e-05, "loss": 0.0364, "step": 17180 }, { "grad_norm": 0.2901093065738678, "learning_rate": 9.599280460080587e-05, "loss": 0.0383, "step": 17190 }, { "grad_norm": 0.2078443169593811, "learning_rate": 9.59863162514947e-05, "loss": 0.0333, "step": 17200 }, { "grad_norm": 0.30223143100738525, "learning_rate": 9.597982287318911e-05, "loss": 0.033, "step": 17210 }, { "grad_norm": 0.3546392321586609, "learning_rate": 9.597332446659923e-05, "loss": 0.0372, "step": 17220 }, { "grad_norm": 0.2258136123418808, "learning_rate": 9.59668210324357e-05, "loss": 0.0371, "step": 17230 }, { "grad_norm": 0.2962287366390228, "learning_rate": 9.596031257140974e-05, "loss": 0.0365, "step": 17240 }, { "grad_norm": 0.2678145170211792, "learning_rate": 9.59537990842331e-05, "loss": 0.0354, "step": 17250 }, { "grad_norm": 0.2443270981311798, "learning_rate": 9.594728057161806e-05, "loss": 0.034, "step": 17260 }, { "grad_norm": 0.313748300075531, "learning_rate": 9.594075703427752e-05, "loss": 0.0378, "step": 17270 }, { "grad_norm": 0.25505998730659485, "learning_rate": 9.593422847292486e-05, "loss": 0.034, "step": 17280 }, { "grad_norm": 0.31312936544418335, "learning_rate": 9.592769488827402e-05, "loss": 0.041, "step": 17290 }, { "grad_norm": 0.2647009789943695, "learning_rate": 9.592115628103952e-05, "loss": 0.0337, "step": 17300 }, { "grad_norm": 0.3565068244934082, "learning_rate": 9.591461265193643e-05, "loss": 0.0376, "step": 17310 }, { "grad_norm": 0.26379767060279846, "learning_rate": 9.590806400168032e-05, "loss": 0.037, "step": 17320 }, { "grad_norm": 0.3651156723499298, "learning_rate": 9.590151033098735e-05, "loss": 0.0382, "step": 17330 }, { "grad_norm": 0.3104977309703827, "learning_rate": 9.589495164057423e-05, "loss": 0.0495, "step": 17340 }, { "grad_norm": 0.3502142131328583, "learning_rate": 9.58883879311582e-05, "loss": 0.0439, "step": 17350 }, { "grad_norm": 0.24700196087360382, "learning_rate": 9.588181920345705e-05, "loss": 0.042, "step": 17360 }, { "grad_norm": 0.26235055923461914, "learning_rate": 9.587524545818913e-05, "loss": 0.0376, "step": 17370 }, { "grad_norm": 0.2681730091571808, "learning_rate": 9.586866669607335e-05, "loss": 0.0367, "step": 17380 }, { "grad_norm": 0.5938873291015625, "learning_rate": 9.586208291782915e-05, "loss": 0.0354, "step": 17390 }, { "grad_norm": 0.2931642532348633, "learning_rate": 9.58554941241765e-05, "loss": 0.039, "step": 17400 }, { "grad_norm": 0.2857111990451813, "learning_rate": 9.584890031583596e-05, "loss": 0.0391, "step": 17410 }, { "grad_norm": 0.17620807886123657, "learning_rate": 9.584230149352861e-05, "loss": 0.0327, "step": 17420 }, { "grad_norm": 0.22336536645889282, "learning_rate": 9.58356976579761e-05, "loss": 0.0392, "step": 17430 }, { "grad_norm": 0.3522174656391144, "learning_rate": 9.58290888099006e-05, "loss": 0.0363, "step": 17440 }, { "grad_norm": 0.4090990722179413, "learning_rate": 9.582247495002486e-05, "loss": 0.037, "step": 17450 }, { "grad_norm": 0.28390079736709595, "learning_rate": 9.581585607907214e-05, "loss": 0.0326, "step": 17460 }, { "grad_norm": 0.22202131152153015, "learning_rate": 9.580923219776628e-05, "loss": 0.0361, "step": 17470 }, { "grad_norm": 0.2620343267917633, "learning_rate": 9.580260330683167e-05, "loss": 0.0389, "step": 17480 }, { "grad_norm": 0.26360878348350525, "learning_rate": 9.579596940699322e-05, "loss": 0.0399, "step": 17490 }, { "grad_norm": 0.3026336133480072, "learning_rate": 9.578933049897643e-05, "loss": 0.0418, "step": 17500 }, { "grad_norm": 0.2880202531814575, "learning_rate": 9.578268658350728e-05, "loss": 0.0366, "step": 17510 }, { "grad_norm": 0.29651549458503723, "learning_rate": 9.577603766131235e-05, "loss": 0.036, "step": 17520 }, { "grad_norm": 0.24632523953914642, "learning_rate": 9.576938373311878e-05, "loss": 0.0328, "step": 17530 }, { "grad_norm": 0.2563846707344055, "learning_rate": 9.576272479965421e-05, "loss": 0.0317, "step": 17540 }, { "grad_norm": 0.29533129930496216, "learning_rate": 9.575606086164687e-05, "loss": 0.0309, "step": 17550 }, { "grad_norm": 0.2741614282131195, "learning_rate": 9.57493919198255e-05, "loss": 0.0361, "step": 17560 }, { "grad_norm": 0.2818109691143036, "learning_rate": 9.57427179749194e-05, "loss": 0.0337, "step": 17570 }, { "grad_norm": 0.6774085760116577, "learning_rate": 9.573603902765846e-05, "loss": 0.0325, "step": 17580 }, { "grad_norm": 0.21569544076919556, "learning_rate": 9.572935507877304e-05, "loss": 0.0384, "step": 17590 }, { "grad_norm": 0.5599412322044373, "learning_rate": 9.57226661289941e-05, "loss": 0.0407, "step": 17600 }, { "grad_norm": 0.22125807404518127, "learning_rate": 9.571597217905315e-05, "loss": 0.0385, "step": 17610 }, { "grad_norm": 0.22407743334770203, "learning_rate": 9.57092732296822e-05, "loss": 0.041, "step": 17620 }, { "grad_norm": 0.25463956594467163, "learning_rate": 9.570256928161385e-05, "loss": 0.0372, "step": 17630 }, { "grad_norm": 0.24775663018226624, "learning_rate": 9.569586033558126e-05, "loss": 0.0365, "step": 17640 }, { "grad_norm": 0.23848621547222137, "learning_rate": 9.568914639231807e-05, "loss": 0.0426, "step": 17650 }, { "grad_norm": 0.3257036507129669, "learning_rate": 9.568242745255852e-05, "loss": 0.0384, "step": 17660 }, { "grad_norm": 0.21785897016525269, "learning_rate": 9.567570351703739e-05, "loss": 0.0326, "step": 17670 }, { "grad_norm": 0.25508472323417664, "learning_rate": 9.566897458649001e-05, "loss": 0.042, "step": 17680 }, { "grad_norm": 0.4968360960483551, "learning_rate": 9.566224066165221e-05, "loss": 0.0362, "step": 17690 }, { "grad_norm": 0.25686150789260864, "learning_rate": 9.565550174326043e-05, "loss": 0.0396, "step": 17700 }, { "grad_norm": 0.2690938711166382, "learning_rate": 9.564875783205162e-05, "loss": 0.0316, "step": 17710 }, { "grad_norm": 0.3205534517765045, "learning_rate": 9.564200892876328e-05, "loss": 0.0385, "step": 17720 }, { "grad_norm": 0.3736378848552704, "learning_rate": 9.563525503413348e-05, "loss": 0.0405, "step": 17730 }, { "grad_norm": 0.22816380858421326, "learning_rate": 9.562849614890079e-05, "loss": 0.0334, "step": 17740 }, { "grad_norm": 0.3096333146095276, "learning_rate": 9.562173227380436e-05, "loss": 0.0341, "step": 17750 }, { "grad_norm": 0.5851543545722961, "learning_rate": 9.561496340958389e-05, "loss": 0.0375, "step": 17760 }, { "grad_norm": 0.2879134714603424, "learning_rate": 9.560818955697959e-05, "loss": 0.035, "step": 17770 }, { "grad_norm": 0.3068440854549408, "learning_rate": 9.560141071673228e-05, "loss": 0.0365, "step": 17780 }, { "grad_norm": 0.2431386560201645, "learning_rate": 9.559462688958323e-05, "loss": 0.0329, "step": 17790 }, { "grad_norm": 0.23130716383457184, "learning_rate": 9.558783807627434e-05, "loss": 0.0336, "step": 17800 }, { "grad_norm": 0.25186488032341003, "learning_rate": 9.558104427754801e-05, "loss": 0.0357, "step": 17810 }, { "grad_norm": 0.22116771340370178, "learning_rate": 9.557424549414722e-05, "loss": 0.0356, "step": 17820 }, { "grad_norm": 0.3293743431568146, "learning_rate": 9.556744172681546e-05, "loss": 0.0402, "step": 17830 }, { "grad_norm": 0.19396480917930603, "learning_rate": 9.556063297629677e-05, "loss": 0.0355, "step": 17840 }, { "grad_norm": 0.5272514224052429, "learning_rate": 9.555381924333578e-05, "loss": 0.042, "step": 17850 }, { "grad_norm": 0.2622215747833252, "learning_rate": 9.554700052867758e-05, "loss": 0.0352, "step": 17860 }, { "grad_norm": 0.18956011533737183, "learning_rate": 9.554017683306789e-05, "loss": 0.0341, "step": 17870 }, { "grad_norm": 0.3778064250946045, "learning_rate": 9.553334815725294e-05, "loss": 0.0401, "step": 17880 }, { "grad_norm": 0.5631225109100342, "learning_rate": 9.552651450197949e-05, "loss": 0.0403, "step": 17890 }, { "grad_norm": 0.5421136021614075, "learning_rate": 9.551967586799486e-05, "loss": 0.0429, "step": 17900 }, { "grad_norm": 0.2511286437511444, "learning_rate": 9.551283225604692e-05, "loss": 0.0437, "step": 17910 }, { "grad_norm": 0.4953473210334778, "learning_rate": 9.550598366688406e-05, "loss": 0.045, "step": 17920 }, { "grad_norm": 0.33081135153770447, "learning_rate": 9.549913010125526e-05, "loss": 0.0398, "step": 17930 }, { "grad_norm": 0.26936763525009155, "learning_rate": 9.549227155990999e-05, "loss": 0.0422, "step": 17940 }, { "grad_norm": 0.27118369936943054, "learning_rate": 9.548540804359828e-05, "loss": 0.039, "step": 17950 }, { "grad_norm": 0.2621851861476898, "learning_rate": 9.547853955307077e-05, "loss": 0.0376, "step": 17960 }, { "grad_norm": 0.23396189510822296, "learning_rate": 9.547166608907853e-05, "loss": 0.0318, "step": 17970 }, { "grad_norm": 0.24017536640167236, "learning_rate": 9.546478765237326e-05, "loss": 0.0388, "step": 17980 }, { "grad_norm": 0.24518635869026184, "learning_rate": 9.545790424370715e-05, "loss": 0.0443, "step": 17990 }, { "grad_norm": 0.29183584451675415, "learning_rate": 9.5451015863833e-05, "loss": 0.0459, "step": 18000 }, { "grad_norm": 0.26856425404548645, "learning_rate": 9.544412251350408e-05, "loss": 0.0441, "step": 18010 }, { "grad_norm": 0.3354041278362274, "learning_rate": 9.543722419347422e-05, "loss": 0.0458, "step": 18020 }, { "grad_norm": 0.2808536887168884, "learning_rate": 9.543032090449788e-05, "loss": 0.0382, "step": 18030 }, { "grad_norm": 0.2594447433948517, "learning_rate": 9.542341264732992e-05, "loss": 0.0367, "step": 18040 }, { "grad_norm": 0.3781889081001282, "learning_rate": 9.541649942272585e-05, "loss": 0.0424, "step": 18050 }, { "grad_norm": 0.4078546464443207, "learning_rate": 9.54095812314417e-05, "loss": 0.0488, "step": 18060 }, { "grad_norm": 0.3504484295845032, "learning_rate": 9.540265807423401e-05, "loss": 0.0467, "step": 18070 }, { "grad_norm": 0.36266812682151794, "learning_rate": 9.53957299518599e-05, "loss": 0.0453, "step": 18080 }, { "grad_norm": 0.35023579001426697, "learning_rate": 9.5388796865077e-05, "loss": 0.033, "step": 18090 }, { "grad_norm": 0.3608020842075348, "learning_rate": 9.538185881464353e-05, "loss": 0.041, "step": 18100 }, { "grad_norm": 0.2803036570549011, "learning_rate": 9.537491580131821e-05, "loss": 0.0376, "step": 18110 }, { "grad_norm": 0.31985902786254883, "learning_rate": 9.53679678258603e-05, "loss": 0.0336, "step": 18120 }, { "grad_norm": 0.31992393732070923, "learning_rate": 9.536101488902966e-05, "loss": 0.0353, "step": 18130 }, { "grad_norm": 0.2900584042072296, "learning_rate": 9.535405699158663e-05, "loss": 0.0318, "step": 18140 }, { "grad_norm": 0.2594236135482788, "learning_rate": 9.53470941342921e-05, "loss": 0.0382, "step": 18150 }, { "grad_norm": 0.2821536064147949, "learning_rate": 9.534012631790756e-05, "loss": 0.0409, "step": 18160 }, { "grad_norm": 0.26594942808151245, "learning_rate": 9.533315354319494e-05, "loss": 0.0366, "step": 18170 }, { "grad_norm": 0.2262989580631256, "learning_rate": 9.532617581091682e-05, "loss": 0.0362, "step": 18180 }, { "grad_norm": 0.30615758895874023, "learning_rate": 9.531919312183629e-05, "loss": 0.0364, "step": 18190 }, { "grad_norm": 0.30306628346443176, "learning_rate": 9.531220547671688e-05, "loss": 0.0374, "step": 18200 }, { "grad_norm": 0.3809656500816345, "learning_rate": 9.530521287632285e-05, "loss": 0.0375, "step": 18210 }, { "grad_norm": 0.2327357530593872, "learning_rate": 9.529821532141884e-05, "loss": 0.033, "step": 18220 }, { "grad_norm": 0.33271726965904236, "learning_rate": 9.52912128127701e-05, "loss": 0.0375, "step": 18230 }, { "grad_norm": 0.3484865128993988, "learning_rate": 9.528420535114244e-05, "loss": 0.0376, "step": 18240 }, { "grad_norm": 0.18986013531684875, "learning_rate": 9.527719293730215e-05, "loss": 0.0352, "step": 18250 }, { "grad_norm": 0.23042525351047516, "learning_rate": 9.527017557201611e-05, "loss": 0.0321, "step": 18260 }, { "grad_norm": 0.32960161566734314, "learning_rate": 9.526315325605176e-05, "loss": 0.033, "step": 18270 }, { "grad_norm": 0.21180401742458344, "learning_rate": 9.525612599017699e-05, "loss": 0.0327, "step": 18280 }, { "grad_norm": 0.24472318589687347, "learning_rate": 9.524909377516033e-05, "loss": 0.0315, "step": 18290 }, { "grad_norm": 0.23170526325702667, "learning_rate": 9.524205661177081e-05, "loss": 0.0383, "step": 18300 }, { "grad_norm": 0.23563213646411896, "learning_rate": 9.523501450077801e-05, "loss": 0.0331, "step": 18310 }, { "grad_norm": 0.3525342047214508, "learning_rate": 9.522796744295202e-05, "loss": 0.0364, "step": 18320 }, { "grad_norm": 0.33085814118385315, "learning_rate": 9.522091543906352e-05, "loss": 0.044, "step": 18330 }, { "grad_norm": 0.27269446849823, "learning_rate": 9.521385848988369e-05, "loss": 0.0346, "step": 18340 }, { "grad_norm": 0.2584041953086853, "learning_rate": 9.520679659618428e-05, "loss": 0.0349, "step": 18350 }, { "grad_norm": 0.251450777053833, "learning_rate": 9.519972975873754e-05, "loss": 0.0309, "step": 18360 }, { "grad_norm": 0.27010488510131836, "learning_rate": 9.519265797831633e-05, "loss": 0.0317, "step": 18370 }, { "grad_norm": 0.23226746916770935, "learning_rate": 9.518558125569399e-05, "loss": 0.0307, "step": 18380 }, { "grad_norm": 0.2975824475288391, "learning_rate": 9.517849959164442e-05, "loss": 0.0278, "step": 18390 }, { "grad_norm": 0.29143640398979187, "learning_rate": 9.517141298694205e-05, "loss": 0.038, "step": 18400 }, { "grad_norm": 0.3991782069206238, "learning_rate": 9.516432144236188e-05, "loss": 0.0429, "step": 18410 }, { "grad_norm": 0.3308488428592682, "learning_rate": 9.515722495867941e-05, "loss": 0.0354, "step": 18420 }, { "grad_norm": 0.38312458992004395, "learning_rate": 9.515012353667072e-05, "loss": 0.0371, "step": 18430 }, { "grad_norm": 0.4418298900127411, "learning_rate": 9.51430171771124e-05, "loss": 0.0423, "step": 18440 }, { "grad_norm": 0.28282487392425537, "learning_rate": 9.513590588078159e-05, "loss": 0.0425, "step": 18450 }, { "grad_norm": 0.2816091775894165, "learning_rate": 9.512878964845597e-05, "loss": 0.0366, "step": 18460 }, { "grad_norm": 0.2741175591945648, "learning_rate": 9.512166848091377e-05, "loss": 0.0444, "step": 18470 }, { "grad_norm": 0.28003981709480286, "learning_rate": 9.511454237893376e-05, "loss": 0.0498, "step": 18480 }, { "grad_norm": 0.3492017984390259, "learning_rate": 9.51074113432952e-05, "loss": 0.0464, "step": 18490 }, { "grad_norm": 0.23578009009361267, "learning_rate": 9.510027537477797e-05, "loss": 0.0448, "step": 18500 }, { "grad_norm": 0.2839478552341461, "learning_rate": 9.509313447416242e-05, "loss": 0.0342, "step": 18510 }, { "grad_norm": 0.26176318526268005, "learning_rate": 9.508598864222949e-05, "loss": 0.0368, "step": 18520 }, { "grad_norm": 0.24613095819950104, "learning_rate": 9.507883787976062e-05, "loss": 0.0327, "step": 18530 }, { "grad_norm": 0.3710730969905853, "learning_rate": 9.507168218753781e-05, "loss": 0.0339, "step": 18540 }, { "grad_norm": 0.248247429728508, "learning_rate": 9.506452156634362e-05, "loss": 0.0337, "step": 18550 }, { "grad_norm": 0.33586257696151733, "learning_rate": 9.505735601696109e-05, "loss": 0.0369, "step": 18560 }, { "grad_norm": 0.3196769654750824, "learning_rate": 9.505018554017385e-05, "loss": 0.0355, "step": 18570 }, { "grad_norm": 0.24263830482959747, "learning_rate": 9.504301013676604e-05, "loss": 0.0391, "step": 18580 }, { "grad_norm": 0.2942768931388855, "learning_rate": 9.503582980752238e-05, "loss": 0.0378, "step": 18590 }, { "grad_norm": 0.26942551136016846, "learning_rate": 9.502864455322809e-05, "loss": 0.0355, "step": 18600 }, { "grad_norm": 0.3381030261516571, "learning_rate": 9.502145437466891e-05, "loss": 0.0328, "step": 18610 }, { "grad_norm": 0.3634454607963562, "learning_rate": 9.501425927263116e-05, "loss": 0.0381, "step": 18620 }, { "grad_norm": 0.500244677066803, "learning_rate": 9.500705924790172e-05, "loss": 0.0355, "step": 18630 }, { "grad_norm": 0.8398213386535645, "learning_rate": 9.499985430126794e-05, "loss": 0.039, "step": 18640 }, { "grad_norm": 0.3569626212120056, "learning_rate": 9.499264443351775e-05, "loss": 0.0421, "step": 18650 }, { "grad_norm": 0.20784583687782288, "learning_rate": 9.498542964543961e-05, "loss": 0.0378, "step": 18660 }, { "grad_norm": 0.3152577877044678, "learning_rate": 9.497820993782252e-05, "loss": 0.0351, "step": 18670 }, { "grad_norm": 0.3945213556289673, "learning_rate": 9.497098531145601e-05, "loss": 0.0377, "step": 18680 }, { "grad_norm": 0.4125654697418213, "learning_rate": 9.496375576713017e-05, "loss": 0.0386, "step": 18690 }, { "grad_norm": 0.26200565695762634, "learning_rate": 9.49565213056356e-05, "loss": 0.043, "step": 18700 }, { "grad_norm": 0.29423102736473083, "learning_rate": 9.494928192776342e-05, "loss": 0.0357, "step": 18710 }, { "grad_norm": 0.21092844009399414, "learning_rate": 9.494203763430538e-05, "loss": 0.0329, "step": 18720 }, { "grad_norm": 0.24323497712612152, "learning_rate": 9.493478842605366e-05, "loss": 0.0349, "step": 18730 }, { "grad_norm": 0.27812033891677856, "learning_rate": 9.492753430380105e-05, "loss": 0.0351, "step": 18740 }, { "grad_norm": 0.277724027633667, "learning_rate": 9.492027526834083e-05, "loss": 0.0356, "step": 18750 }, { "grad_norm": 0.4945635497570038, "learning_rate": 9.491301132046684e-05, "loss": 0.0391, "step": 18760 }, { "grad_norm": 0.24113371968269348, "learning_rate": 9.490574246097345e-05, "loss": 0.0321, "step": 18770 }, { "grad_norm": 0.27000895142555237, "learning_rate": 9.48984686906556e-05, "loss": 0.0324, "step": 18780 }, { "grad_norm": 0.28753969073295593, "learning_rate": 9.489119001030871e-05, "loss": 0.0349, "step": 18790 }, { "grad_norm": 0.2894099950790405, "learning_rate": 9.488390642072878e-05, "loss": 0.0346, "step": 18800 }, { "grad_norm": 0.30275648832321167, "learning_rate": 9.48766179227123e-05, "loss": 0.032, "step": 18810 }, { "grad_norm": 0.30426913499832153, "learning_rate": 9.486932451705636e-05, "loss": 0.0337, "step": 18820 }, { "grad_norm": 0.3283526599407196, "learning_rate": 9.486202620455857e-05, "loss": 0.0401, "step": 18830 }, { "grad_norm": 0.26756343245506287, "learning_rate": 9.485472298601704e-05, "loss": 0.0391, "step": 18840 }, { "grad_norm": 0.31432291865348816, "learning_rate": 9.484741486223043e-05, "loss": 0.0445, "step": 18850 }, { "grad_norm": 0.2821570336818695, "learning_rate": 9.484010183399797e-05, "loss": 0.0373, "step": 18860 }, { "grad_norm": 0.2724394202232361, "learning_rate": 9.483278390211938e-05, "loss": 0.0443, "step": 18870 }, { "grad_norm": 0.28581568598747253, "learning_rate": 9.482546106739496e-05, "loss": 0.0367, "step": 18880 }, { "grad_norm": 0.32243776321411133, "learning_rate": 9.48181333306255e-05, "loss": 0.0412, "step": 18890 }, { "grad_norm": 0.22681966423988342, "learning_rate": 9.481080069261237e-05, "loss": 0.0462, "step": 18900 }, { "grad_norm": 0.32301849126815796, "learning_rate": 9.480346315415745e-05, "loss": 0.0365, "step": 18910 }, { "grad_norm": 0.402991384267807, "learning_rate": 9.479612071606314e-05, "loss": 0.0417, "step": 18920 }, { "grad_norm": 0.2416841834783554, "learning_rate": 9.478877337913244e-05, "loss": 0.0478, "step": 18930 }, { "grad_norm": 0.2563101351261139, "learning_rate": 9.478142114416881e-05, "loss": 0.0351, "step": 18940 }, { "grad_norm": 0.32048022747039795, "learning_rate": 9.47740640119763e-05, "loss": 0.0386, "step": 18950 }, { "grad_norm": 0.28299158811569214, "learning_rate": 9.476670198335947e-05, "loss": 0.0394, "step": 18960 }, { "grad_norm": 0.26634716987609863, "learning_rate": 9.47593350591234e-05, "loss": 0.0393, "step": 18970 }, { "grad_norm": 0.21248698234558105, "learning_rate": 9.475196324007376e-05, "loss": 0.0321, "step": 18980 }, { "grad_norm": 0.29211172461509705, "learning_rate": 9.474458652701669e-05, "loss": 0.0377, "step": 18990 }, { "grad_norm": 0.23508568108081818, "learning_rate": 9.473720492075892e-05, "loss": 0.0357, "step": 19000 }, { "grad_norm": 0.2483387440443039, "learning_rate": 9.472981842210768e-05, "loss": 0.0342, "step": 19010 }, { "grad_norm": 0.20832069218158722, "learning_rate": 9.472242703187074e-05, "loss": 0.0352, "step": 19020 }, { "grad_norm": 0.31365108489990234, "learning_rate": 9.471503075085643e-05, "loss": 0.035, "step": 19030 }, { "grad_norm": 0.25761687755584717, "learning_rate": 9.470762957987359e-05, "loss": 0.0325, "step": 19040 }, { "grad_norm": 0.34123748540878296, "learning_rate": 9.470022351973158e-05, "loss": 0.0329, "step": 19050 }, { "grad_norm": 0.4086810052394867, "learning_rate": 9.469281257124034e-05, "loss": 0.0382, "step": 19060 }, { "grad_norm": 0.24719984829425812, "learning_rate": 9.46853967352103e-05, "loss": 0.0349, "step": 19070 }, { "grad_norm": 0.3183881938457489, "learning_rate": 9.467797601245246e-05, "loss": 0.0386, "step": 19080 }, { "grad_norm": 0.3758000135421753, "learning_rate": 9.467055040377834e-05, "loss": 0.042, "step": 19090 }, { "grad_norm": 0.3499791622161865, "learning_rate": 9.466311990999999e-05, "loss": 0.0403, "step": 19100 }, { "grad_norm": 0.3128044903278351, "learning_rate": 9.465568453193e-05, "loss": 0.0421, "step": 19110 }, { "grad_norm": 0.42136335372924805, "learning_rate": 9.464824427038148e-05, "loss": 0.0399, "step": 19120 }, { "grad_norm": 0.7804340124130249, "learning_rate": 9.46407991261681e-05, "loss": 0.0388, "step": 19130 }, { "grad_norm": 0.2601950764656067, "learning_rate": 9.463334910010404e-05, "loss": 0.0401, "step": 19140 }, { "grad_norm": 0.30294761061668396, "learning_rate": 9.462589419300403e-05, "loss": 0.0479, "step": 19150 }, { "grad_norm": 0.24447545409202576, "learning_rate": 9.461843440568333e-05, "loss": 0.036, "step": 19160 }, { "grad_norm": 0.23156693577766418, "learning_rate": 9.461096973895773e-05, "loss": 0.0379, "step": 19170 }, { "grad_norm": 0.28234222531318665, "learning_rate": 9.460350019364355e-05, "loss": 0.0385, "step": 19180 }, { "grad_norm": 0.5186815857887268, "learning_rate": 9.459602577055764e-05, "loss": 0.0302, "step": 19190 }, { "grad_norm": 0.42097413539886475, "learning_rate": 9.45885464705174e-05, "loss": 0.0373, "step": 19200 }, { "grad_norm": 0.2814905345439911, "learning_rate": 9.458106229434076e-05, "loss": 0.0391, "step": 19210 }, { "grad_norm": 0.24432571232318878, "learning_rate": 9.457357324284617e-05, "loss": 0.0346, "step": 19220 }, { "grad_norm": 0.2855350971221924, "learning_rate": 9.456607931685262e-05, "loss": 0.0313, "step": 19230 }, { "grad_norm": 0.22398337721824646, "learning_rate": 9.455858051717965e-05, "loss": 0.0317, "step": 19240 }, { "grad_norm": 0.21849018335342407, "learning_rate": 9.45510768446473e-05, "loss": 0.0402, "step": 19250 }, { "grad_norm": 0.26307401061058044, "learning_rate": 9.454356830007618e-05, "loss": 0.0352, "step": 19260 }, { "grad_norm": 0.30251941084861755, "learning_rate": 9.45360548842874e-05, "loss": 0.0418, "step": 19270 }, { "grad_norm": 0.29165327548980713, "learning_rate": 9.452853659810261e-05, "loss": 0.0356, "step": 19280 }, { "grad_norm": 0.31065768003463745, "learning_rate": 9.452101344234401e-05, "loss": 0.0403, "step": 19290 }, { "grad_norm": 0.22014948725700378, "learning_rate": 9.451348541783431e-05, "loss": 0.0385, "step": 19300 }, { "grad_norm": 0.2705017328262329, "learning_rate": 9.450595252539678e-05, "loss": 0.0472, "step": 19310 }, { "grad_norm": 0.31170737743377686, "learning_rate": 9.449841476585518e-05, "loss": 0.0377, "step": 19320 }, { "grad_norm": 0.41146188974380493, "learning_rate": 9.449087214003384e-05, "loss": 0.0317, "step": 19330 }, { "grad_norm": 0.24328480660915375, "learning_rate": 9.448332464875765e-05, "loss": 0.0383, "step": 19340 }, { "grad_norm": 0.23158079385757446, "learning_rate": 9.447577229285192e-05, "loss": 0.0319, "step": 19350 }, { "grad_norm": 0.6905397772789001, "learning_rate": 9.446821507314261e-05, "loss": 0.0312, "step": 19360 }, { "grad_norm": 0.4732363522052765, "learning_rate": 9.446065299045617e-05, "loss": 0.0354, "step": 19370 }, { "grad_norm": 0.3511650562286377, "learning_rate": 9.445308604561955e-05, "loss": 0.0404, "step": 19380 }, { "grad_norm": 0.40121906995773315, "learning_rate": 9.444551423946028e-05, "loss": 0.0343, "step": 19390 }, { "grad_norm": 0.24541175365447998, "learning_rate": 9.443793757280638e-05, "loss": 0.0409, "step": 19400 }, { "grad_norm": 0.25070199370384216, "learning_rate": 9.443035604648646e-05, "loss": 0.0327, "step": 19410 }, { "grad_norm": 0.2189522683620453, "learning_rate": 9.44227696613296e-05, "loss": 0.0356, "step": 19420 }, { "grad_norm": 0.28641554713249207, "learning_rate": 9.441517841816542e-05, "loss": 0.0382, "step": 19430 }, { "grad_norm": 0.26149553060531616, "learning_rate": 9.440758231782413e-05, "loss": 0.0356, "step": 19440 }, { "grad_norm": 0.407802939414978, "learning_rate": 9.439998136113639e-05, "loss": 0.034, "step": 19450 }, { "grad_norm": 0.32398778200149536, "learning_rate": 9.439237554893344e-05, "loss": 0.0344, "step": 19460 }, { "grad_norm": 0.21948640048503876, "learning_rate": 9.438476488204705e-05, "loss": 0.0298, "step": 19470 }, { "grad_norm": 0.3824406862258911, "learning_rate": 9.43771493613095e-05, "loss": 0.0332, "step": 19480 }, { "grad_norm": 0.19637565314769745, "learning_rate": 9.436952898755362e-05, "loss": 0.0323, "step": 19490 }, { "grad_norm": 0.27110472321510315, "learning_rate": 9.436190376161276e-05, "loss": 0.0418, "step": 19500 }, { "grad_norm": 0.22896769642829895, "learning_rate": 9.43542736843208e-05, "loss": 0.0409, "step": 19510 }, { "grad_norm": 0.412741482257843, "learning_rate": 9.434663875651216e-05, "loss": 0.0354, "step": 19520 }, { "grad_norm": 0.28997841477394104, "learning_rate": 9.433899897902177e-05, "loss": 0.034, "step": 19530 }, { "grad_norm": 0.2877793312072754, "learning_rate": 9.433135435268511e-05, "loss": 0.0303, "step": 19540 }, { "grad_norm": 0.29124391078948975, "learning_rate": 9.432370487833819e-05, "loss": 0.0308, "step": 19550 }, { "grad_norm": 0.22455953061580658, "learning_rate": 9.431605055681756e-05, "loss": 0.0335, "step": 19560 }, { "grad_norm": 0.23222649097442627, "learning_rate": 9.430839138896026e-05, "loss": 0.0336, "step": 19570 }, { "grad_norm": 0.301658034324646, "learning_rate": 9.43007273756039e-05, "loss": 0.0315, "step": 19580 }, { "grad_norm": 0.24620170891284943, "learning_rate": 9.429305851758658e-05, "loss": 0.0327, "step": 19590 }, { "grad_norm": 0.20658065378665924, "learning_rate": 9.428538481574699e-05, "loss": 0.031, "step": 19600 }, { "grad_norm": 0.17547737061977386, "learning_rate": 9.42777062709243e-05, "loss": 0.0436, "step": 19610 }, { "grad_norm": 0.2740391492843628, "learning_rate": 9.427002288395821e-05, "loss": 0.037, "step": 19620 }, { "grad_norm": 0.372526079416275, "learning_rate": 9.426233465568898e-05, "loss": 0.0388, "step": 19630 }, { "grad_norm": 0.27320238947868347, "learning_rate": 9.42546415869574e-05, "loss": 0.0367, "step": 19640 }, { "grad_norm": 0.22494898736476898, "learning_rate": 9.424694367860473e-05, "loss": 0.0381, "step": 19650 }, { "grad_norm": 0.23514610528945923, "learning_rate": 9.423924093147284e-05, "loss": 0.0362, "step": 19660 }, { "grad_norm": 0.27636298537254333, "learning_rate": 9.423153334640407e-05, "loss": 0.0339, "step": 19670 }, { "grad_norm": 0.21057429909706116, "learning_rate": 9.42238209242413e-05, "loss": 0.0345, "step": 19680 }, { "grad_norm": 0.2774020731449127, "learning_rate": 9.421610366582798e-05, "loss": 0.0404, "step": 19690 }, { "grad_norm": 0.26050493121147156, "learning_rate": 9.420838157200803e-05, "loss": 0.0366, "step": 19700 }, { "grad_norm": 0.2509821057319641, "learning_rate": 9.420065464362594e-05, "loss": 0.0363, "step": 19710 }, { "grad_norm": 0.21481692790985107, "learning_rate": 9.419292288152673e-05, "loss": 0.0359, "step": 19720 }, { "grad_norm": 0.5035185813903809, "learning_rate": 9.418518628655588e-05, "loss": 0.0315, "step": 19730 }, { "grad_norm": 0.22664064168930054, "learning_rate": 9.417744485955951e-05, "loss": 0.0317, "step": 19740 }, { "grad_norm": 0.2668111324310303, "learning_rate": 9.41696986013842e-05, "loss": 0.0388, "step": 19750 }, { "grad_norm": 0.2482457309961319, "learning_rate": 9.416194751287705e-05, "loss": 0.0401, "step": 19760 }, { "grad_norm": 0.2667067050933838, "learning_rate": 9.415419159488572e-05, "loss": 0.0367, "step": 19770 }, { "grad_norm": 0.29460954666137695, "learning_rate": 9.414643084825837e-05, "loss": 0.0342, "step": 19780 }, { "grad_norm": 0.24909211695194244, "learning_rate": 9.413866527384372e-05, "loss": 0.0358, "step": 19790 }, { "grad_norm": 0.2896924614906311, "learning_rate": 9.4130894872491e-05, "loss": 0.0378, "step": 19800 }, { "grad_norm": 0.29976728558540344, "learning_rate": 9.412311964504998e-05, "loss": 0.0386, "step": 19810 }, { "grad_norm": 0.2293497771024704, "learning_rate": 9.411533959237091e-05, "loss": 0.0378, "step": 19820 }, { "grad_norm": 0.33182165026664734, "learning_rate": 9.410755471530464e-05, "loss": 0.0432, "step": 19830 }, { "grad_norm": 0.24575351178646088, "learning_rate": 9.40997650147025e-05, "loss": 0.0464, "step": 19840 }, { "grad_norm": 0.33372995257377625, "learning_rate": 9.409197049141637e-05, "loss": 0.0449, "step": 19850 }, { "grad_norm": 0.2945898473262787, "learning_rate": 9.408417114629863e-05, "loss": 0.0354, "step": 19860 }, { "grad_norm": 0.28523436188697815, "learning_rate": 9.40763669802022e-05, "loss": 0.0348, "step": 19870 }, { "grad_norm": 0.4261539578437805, "learning_rate": 9.406855799398056e-05, "loss": 0.0367, "step": 19880 }, { "grad_norm": 0.2481466829776764, "learning_rate": 9.406074418848767e-05, "loss": 0.0395, "step": 19890 }, { "grad_norm": 0.25341296195983887, "learning_rate": 9.405292556457805e-05, "loss": 0.0357, "step": 19900 }, { "grad_norm": 0.2675163447856903, "learning_rate": 9.404510212310671e-05, "loss": 0.0325, "step": 19910 }, { "grad_norm": 0.2940618097782135, "learning_rate": 9.403727386492924e-05, "loss": 0.0362, "step": 19920 }, { "grad_norm": 0.2796855568885803, "learning_rate": 9.40294407909017e-05, "loss": 0.032, "step": 19930 }, { "grad_norm": 0.2433643937110901, "learning_rate": 9.40216029018807e-05, "loss": 0.0307, "step": 19940 }, { "grad_norm": 0.5880536437034607, "learning_rate": 9.401376019872338e-05, "loss": 0.0328, "step": 19950 }, { "grad_norm": 0.1967274397611618, "learning_rate": 9.400591268228746e-05, "loss": 0.0295, "step": 19960 }, { "grad_norm": 0.24252276122570038, "learning_rate": 9.399806035343106e-05, "loss": 0.0322, "step": 19970 }, { "grad_norm": 0.2736755907535553, "learning_rate": 9.399020321301294e-05, "loss": 0.0364, "step": 19980 }, { "grad_norm": 0.2807011306285858, "learning_rate": 9.398234126189234e-05, "loss": 0.0387, "step": 19990 }, { "grad_norm": 0.23783889412879944, "learning_rate": 9.397447450092902e-05, "loss": 0.0354, "step": 20000 }, { "grad_norm": 0.282541960477829, "learning_rate": 9.39666029309833e-05, "loss": 0.0335, "step": 20010 }, { "grad_norm": 0.2750665843486786, "learning_rate": 9.395872655291596e-05, "loss": 0.0344, "step": 20020 }, { "grad_norm": 0.23280787467956543, "learning_rate": 9.395084536758838e-05, "loss": 0.0299, "step": 20030 }, { "grad_norm": 0.2731790244579315, "learning_rate": 9.394295937586243e-05, "loss": 0.0382, "step": 20040 }, { "grad_norm": 0.2644544839859009, "learning_rate": 9.393506857860052e-05, "loss": 0.0375, "step": 20050 }, { "grad_norm": 0.2951790988445282, "learning_rate": 9.392717297666555e-05, "loss": 0.0387, "step": 20060 }, { "grad_norm": 0.26213952898979187, "learning_rate": 9.391927257092101e-05, "loss": 0.033, "step": 20070 }, { "grad_norm": 0.319023460149765, "learning_rate": 9.391136736223085e-05, "loss": 0.0374, "step": 20080 }, { "grad_norm": 0.28476768732070923, "learning_rate": 9.390345735145956e-05, "loss": 0.0329, "step": 20090 }, { "grad_norm": 0.23943157494068146, "learning_rate": 9.389554253947219e-05, "loss": 0.0322, "step": 20100 }, { "grad_norm": 0.29241788387298584, "learning_rate": 9.388762292713428e-05, "loss": 0.0381, "step": 20110 }, { "grad_norm": 0.24895817041397095, "learning_rate": 9.38796985153119e-05, "loss": 0.0343, "step": 20120 }, { "grad_norm": 0.27851903438568115, "learning_rate": 9.387176930487169e-05, "loss": 0.0327, "step": 20130 }, { "grad_norm": 0.30038005113601685, "learning_rate": 9.386383529668072e-05, "loss": 0.033, "step": 20140 }, { "grad_norm": 0.30691757798194885, "learning_rate": 9.385589649160669e-05, "loss": 0.0411, "step": 20150 }, { "grad_norm": 0.3164742588996887, "learning_rate": 9.384795289051775e-05, "loss": 0.0321, "step": 20160 }, { "grad_norm": 0.24213889241218567, "learning_rate": 9.384000449428261e-05, "loss": 0.0352, "step": 20170 }, { "grad_norm": 0.2531956732273102, "learning_rate": 9.383205130377048e-05, "loss": 0.0289, "step": 20180 }, { "grad_norm": 0.32802772521972656, "learning_rate": 9.382409331985114e-05, "loss": 0.0299, "step": 20190 }, { "grad_norm": 0.30092963576316833, "learning_rate": 9.381613054339482e-05, "loss": 0.0299, "step": 20200 }, { "grad_norm": 0.3313518464565277, "learning_rate": 9.380816297527235e-05, "loss": 0.0318, "step": 20210 }, { "grad_norm": 0.21545274555683136, "learning_rate": 9.380019061635506e-05, "loss": 0.034, "step": 20220 }, { "grad_norm": 0.3935839831829071, "learning_rate": 9.379221346751474e-05, "loss": 0.034, "step": 20230 }, { "grad_norm": 0.2704514265060425, "learning_rate": 9.378423152962382e-05, "loss": 0.0326, "step": 20240 }, { "grad_norm": 0.2730118930339813, "learning_rate": 9.377624480355517e-05, "loss": 0.0345, "step": 20250 }, { "grad_norm": 0.2257988303899765, "learning_rate": 9.376825329018219e-05, "loss": 0.0348, "step": 20260 }, { "grad_norm": 0.23634208738803864, "learning_rate": 9.376025699037884e-05, "loss": 0.0348, "step": 20270 }, { "grad_norm": 0.3993222117424011, "learning_rate": 9.37522559050196e-05, "loss": 0.0321, "step": 20280 }, { "grad_norm": 0.3789746165275574, "learning_rate": 9.37442500349794e-05, "loss": 0.0364, "step": 20290 }, { "grad_norm": 0.2546134293079376, "learning_rate": 9.373623938113381e-05, "loss": 0.0323, "step": 20300 }, { "grad_norm": 0.23608627915382385, "learning_rate": 9.372822394435883e-05, "loss": 0.0408, "step": 20310 }, { "grad_norm": 0.26412564516067505, "learning_rate": 9.372020372553102e-05, "loss": 0.0423, "step": 20320 }, { "grad_norm": 0.22007909417152405, "learning_rate": 9.371217872552746e-05, "loss": 0.0417, "step": 20330 }, { "grad_norm": 0.3933965861797333, "learning_rate": 9.370414894522576e-05, "loss": 0.0398, "step": 20340 }, { "grad_norm": 0.36431196331977844, "learning_rate": 9.369611438550406e-05, "loss": 0.0403, "step": 20350 }, { "grad_norm": 0.3124479353427887, "learning_rate": 9.368807504724095e-05, "loss": 0.0381, "step": 20360 }, { "grad_norm": 0.2642557919025421, "learning_rate": 9.368003093131565e-05, "loss": 0.0362, "step": 20370 }, { "grad_norm": 0.3182290494441986, "learning_rate": 9.367198203860785e-05, "loss": 0.0408, "step": 20380 }, { "grad_norm": 0.6148990988731384, "learning_rate": 9.366392836999774e-05, "loss": 0.0382, "step": 20390 }, { "grad_norm": 0.33113184571266174, "learning_rate": 9.365586992636607e-05, "loss": 0.0378, "step": 20400 }, { "grad_norm": 0.4290004372596741, "learning_rate": 9.364780670859412e-05, "loss": 0.0433, "step": 20410 }, { "grad_norm": 0.2600606381893158, "learning_rate": 9.363973871756364e-05, "loss": 0.0406, "step": 20420 }, { "grad_norm": 0.3098616600036621, "learning_rate": 9.363166595415696e-05, "loss": 0.0404, "step": 20430 }, { "grad_norm": 0.24281656742095947, "learning_rate": 9.362358841925686e-05, "loss": 0.0329, "step": 20440 }, { "grad_norm": 0.30852049589157104, "learning_rate": 9.361550611374674e-05, "loss": 0.0359, "step": 20450 }, { "grad_norm": 0.24780547618865967, "learning_rate": 9.360741903851043e-05, "loss": 0.0374, "step": 20460 }, { "grad_norm": 0.29333606362342834, "learning_rate": 9.359932719443236e-05, "loss": 0.038, "step": 20470 }, { "grad_norm": 0.32507213950157166, "learning_rate": 9.35912305823974e-05, "loss": 0.0396, "step": 20480 }, { "grad_norm": 0.2770870625972748, "learning_rate": 9.358312920329101e-05, "loss": 0.0375, "step": 20490 }, { "grad_norm": 0.2456188201904297, "learning_rate": 9.357502305799914e-05, "loss": 0.0364, "step": 20500 }, { "grad_norm": 0.28644734621047974, "learning_rate": 9.356691214740824e-05, "loss": 0.0333, "step": 20510 }, { "grad_norm": 0.2388211339712143, "learning_rate": 9.355879647240535e-05, "loss": 0.0398, "step": 20520 }, { "grad_norm": 0.3268032968044281, "learning_rate": 9.355067603387798e-05, "loss": 0.0356, "step": 20530 }, { "grad_norm": 0.3461467921733856, "learning_rate": 9.354255083271412e-05, "loss": 0.0385, "step": 20540 }, { "grad_norm": 0.26871487498283386, "learning_rate": 9.353442086980239e-05, "loss": 0.0347, "step": 20550 }, { "grad_norm": 0.28997138142585754, "learning_rate": 9.352628614603185e-05, "loss": 0.0351, "step": 20560 }, { "grad_norm": 0.2609593868255615, "learning_rate": 9.351814666229209e-05, "loss": 0.0334, "step": 20570 }, { "grad_norm": 0.32351964712142944, "learning_rate": 9.351000241947324e-05, "loss": 0.0349, "step": 20580 }, { "grad_norm": 0.18937382102012634, "learning_rate": 9.350185341846594e-05, "loss": 0.0357, "step": 20590 }, { "grad_norm": 0.2324487268924713, "learning_rate": 9.349369966016134e-05, "loss": 0.0352, "step": 20600 }, { "grad_norm": 0.24389630556106567, "learning_rate": 9.348554114545117e-05, "loss": 0.0363, "step": 20610 }, { "grad_norm": 0.34050288796424866, "learning_rate": 9.347737787522758e-05, "loss": 0.0434, "step": 20620 }, { "grad_norm": 0.22990930080413818, "learning_rate": 9.346920985038332e-05, "loss": 0.0379, "step": 20630 }, { "grad_norm": 0.35567423701286316, "learning_rate": 9.346103707181162e-05, "loss": 0.0334, "step": 20640 }, { "grad_norm": 0.3207578957080841, "learning_rate": 9.345285954040626e-05, "loss": 0.0388, "step": 20650 }, { "grad_norm": 0.20311346650123596, "learning_rate": 9.34446772570615e-05, "loss": 0.0344, "step": 20660 }, { "grad_norm": 0.3544842600822449, "learning_rate": 9.343649022267214e-05, "loss": 0.0362, "step": 20670 }, { "grad_norm": 0.2411181628704071, "learning_rate": 9.342829843813353e-05, "loss": 0.0303, "step": 20680 }, { "grad_norm": 0.2631221413612366, "learning_rate": 9.342010190434149e-05, "loss": 0.0311, "step": 20690 }, { "grad_norm": 0.2113591432571411, "learning_rate": 9.34119006221924e-05, "loss": 0.0446, "step": 20700 }, { "grad_norm": 0.2799120843410492, "learning_rate": 9.340369459258313e-05, "loss": 0.0334, "step": 20710 }, { "grad_norm": 0.30326560139656067, "learning_rate": 9.339548381641106e-05, "loss": 0.0398, "step": 20720 }, { "grad_norm": 0.3016830086708069, "learning_rate": 9.338726829457413e-05, "loss": 0.03, "step": 20730 }, { "grad_norm": 0.24794691801071167, "learning_rate": 9.337904802797078e-05, "loss": 0.0328, "step": 20740 }, { "grad_norm": 0.268586665391922, "learning_rate": 9.337082301749993e-05, "loss": 0.0337, "step": 20750 }, { "grad_norm": 0.2770121097564697, "learning_rate": 9.336259326406109e-05, "loss": 0.0293, "step": 20760 }, { "grad_norm": 0.23216956853866577, "learning_rate": 9.335435876855427e-05, "loss": 0.037, "step": 20770 }, { "grad_norm": 0.2902499735355377, "learning_rate": 9.334611953187994e-05, "loss": 0.0355, "step": 20780 }, { "grad_norm": 0.3225032389163971, "learning_rate": 9.333787555493914e-05, "loss": 0.0346, "step": 20790 }, { "grad_norm": 0.246049165725708, "learning_rate": 9.332962683863345e-05, "loss": 0.0301, "step": 20800 }, { "grad_norm": 0.25916117429733276, "learning_rate": 9.332137338386489e-05, "loss": 0.0292, "step": 20810 }, { "grad_norm": 0.21656034886837006, "learning_rate": 9.33131151915361e-05, "loss": 0.0293, "step": 20820 }, { "grad_norm": 0.19018620252609253, "learning_rate": 9.330485226255012e-05, "loss": 0.0274, "step": 20830 }, { "grad_norm": 0.20231637358665466, "learning_rate": 9.329658459781061e-05, "loss": 0.0255, "step": 20840 }, { "grad_norm": 0.28615468740463257, "learning_rate": 9.328831219822172e-05, "loss": 0.0283, "step": 20850 }, { "grad_norm": 0.307865709066391, "learning_rate": 9.328003506468808e-05, "loss": 0.0336, "step": 20860 }, { "grad_norm": 0.34467118978500366, "learning_rate": 9.327175319811488e-05, "loss": 0.0285, "step": 20870 }, { "grad_norm": 0.2720252275466919, "learning_rate": 9.326346659940781e-05, "loss": 0.0341, "step": 20880 }, { "grad_norm": 0.336281955242157, "learning_rate": 9.325517526947308e-05, "loss": 0.0316, "step": 20890 }, { "grad_norm": 0.256191223859787, "learning_rate": 9.32468792092174e-05, "loss": 0.0304, "step": 20900 }, { "grad_norm": 0.19917920231819153, "learning_rate": 9.323857841954803e-05, "loss": 0.0316, "step": 20910 }, { "grad_norm": 0.22648721933364868, "learning_rate": 9.323027290137276e-05, "loss": 0.0301, "step": 20920 }, { "grad_norm": 0.27306437492370605, "learning_rate": 9.322196265559981e-05, "loss": 0.0281, "step": 20930 }, { "grad_norm": 0.3531447947025299, "learning_rate": 9.321364768313803e-05, "loss": 0.0421, "step": 20940 }, { "grad_norm": 0.2562566101551056, "learning_rate": 9.32053279848967e-05, "loss": 0.0366, "step": 20950 }, { "grad_norm": 0.2162816822528839, "learning_rate": 9.319700356178567e-05, "loss": 0.0332, "step": 20960 }, { "grad_norm": 0.3643262982368469, "learning_rate": 9.318867441471527e-05, "loss": 0.032, "step": 20970 }, { "grad_norm": 0.3412376046180725, "learning_rate": 9.318034054459637e-05, "loss": 0.0353, "step": 20980 }, { "grad_norm": 0.5664865374565125, "learning_rate": 9.317200195234034e-05, "loss": 0.0354, "step": 20990 }, { "grad_norm": 0.2868340313434601, "learning_rate": 9.316365863885909e-05, "loss": 0.0355, "step": 21000 }, { "grad_norm": 0.27925217151641846, "learning_rate": 9.315531060506502e-05, "loss": 0.0335, "step": 21010 }, { "grad_norm": 0.24672053754329681, "learning_rate": 9.314695785187108e-05, "loss": 0.0349, "step": 21020 }, { "grad_norm": 0.5487945079803467, "learning_rate": 9.313860038019069e-05, "loss": 0.0361, "step": 21030 }, { "grad_norm": 0.29044750332832336, "learning_rate": 9.313023819093782e-05, "loss": 0.0357, "step": 21040 }, { "grad_norm": 0.21962597966194153, "learning_rate": 9.312187128502695e-05, "loss": 0.0357, "step": 21050 }, { "grad_norm": 0.31974563002586365, "learning_rate": 9.311349966337307e-05, "loss": 0.0326, "step": 21060 }, { "grad_norm": 0.2485199123620987, "learning_rate": 9.310512332689169e-05, "loss": 0.0355, "step": 21070 }, { "grad_norm": 0.2702163755893707, "learning_rate": 9.309674227649883e-05, "loss": 0.0361, "step": 21080 }, { "grad_norm": 0.23764780163764954, "learning_rate": 9.308835651311103e-05, "loss": 0.0406, "step": 21090 }, { "grad_norm": 0.3509295582771301, "learning_rate": 9.307996603764533e-05, "loss": 0.0381, "step": 21100 }, { "grad_norm": 0.2840893864631653, "learning_rate": 9.307157085101932e-05, "loss": 0.0373, "step": 21110 }, { "grad_norm": 0.48008570075035095, "learning_rate": 9.306317095415109e-05, "loss": 0.0371, "step": 21120 }, { "grad_norm": 0.2505468726158142, "learning_rate": 9.305476634795922e-05, "loss": 0.042, "step": 21130 }, { "grad_norm": 0.41452792286872864, "learning_rate": 9.304635703336284e-05, "loss": 0.0345, "step": 21140 }, { "grad_norm": 0.22169895470142365, "learning_rate": 9.303794301128157e-05, "loss": 0.0286, "step": 21150 }, { "grad_norm": 0.2634044289588928, "learning_rate": 9.302952428263555e-05, "loss": 0.0396, "step": 21160 }, { "grad_norm": 0.24978362023830414, "learning_rate": 9.302110084834545e-05, "loss": 0.038, "step": 21170 }, { "grad_norm": 0.2931327521800995, "learning_rate": 9.301267270933245e-05, "loss": 0.0385, "step": 21180 }, { "grad_norm": 0.32108891010284424, "learning_rate": 9.300423986651823e-05, "loss": 0.0377, "step": 21190 }, { "grad_norm": 0.2423805296421051, "learning_rate": 9.299580232082501e-05, "loss": 0.0348, "step": 21200 }, { "grad_norm": 0.2205096334218979, "learning_rate": 9.298736007317547e-05, "loss": 0.0363, "step": 21210 }, { "grad_norm": 0.40961942076683044, "learning_rate": 9.297891312449288e-05, "loss": 0.0396, "step": 21220 }, { "grad_norm": 0.2752154469490051, "learning_rate": 9.297046147570094e-05, "loss": 0.0343, "step": 21230 }, { "grad_norm": 0.24538715183734894, "learning_rate": 9.296200512772396e-05, "loss": 0.0413, "step": 21240 }, { "grad_norm": 0.33306458592414856, "learning_rate": 9.295354408148668e-05, "loss": 0.0301, "step": 21250 }, { "grad_norm": 0.27172306180000305, "learning_rate": 9.294507833791441e-05, "loss": 0.0363, "step": 21260 }, { "grad_norm": 0.29594096541404724, "learning_rate": 9.293660789793295e-05, "loss": 0.0312, "step": 21270 }, { "grad_norm": 0.2934913635253906, "learning_rate": 9.292813276246858e-05, "loss": 0.0352, "step": 21280 }, { "grad_norm": 0.24360741674900055, "learning_rate": 9.291965293244816e-05, "loss": 0.036, "step": 21290 }, { "grad_norm": 0.27891242504119873, "learning_rate": 9.291116840879904e-05, "loss": 0.0338, "step": 21300 }, { "grad_norm": 0.27916184067726135, "learning_rate": 9.290267919244904e-05, "loss": 0.0389, "step": 21310 }, { "grad_norm": 0.25501933693885803, "learning_rate": 9.289418528432655e-05, "loss": 0.032, "step": 21320 }, { "grad_norm": 0.41657957434654236, "learning_rate": 9.288568668536045e-05, "loss": 0.0332, "step": 21330 }, { "grad_norm": 0.19228529930114746, "learning_rate": 9.287718339648013e-05, "loss": 0.0303, "step": 21340 }, { "grad_norm": 0.24125203490257263, "learning_rate": 9.28686754186155e-05, "loss": 0.0343, "step": 21350 }, { "grad_norm": 0.213044211268425, "learning_rate": 9.286016275269698e-05, "loss": 0.0299, "step": 21360 }, { "grad_norm": 0.1670873463153839, "learning_rate": 9.285164539965551e-05, "loss": 0.0286, "step": 21370 }, { "grad_norm": 0.20694266259670258, "learning_rate": 9.284312336042251e-05, "loss": 0.0335, "step": 21380 }, { "grad_norm": 0.3052482306957245, "learning_rate": 9.283459663592996e-05, "loss": 0.0349, "step": 21390 }, { "grad_norm": 1.184759259223938, "learning_rate": 9.282606522711033e-05, "loss": 0.0424, "step": 21400 }, { "grad_norm": 0.23368409276008606, "learning_rate": 9.281752913489657e-05, "loss": 0.0375, "step": 21410 }, { "grad_norm": 0.2753153145313263, "learning_rate": 9.280898836022222e-05, "loss": 0.0329, "step": 21420 }, { "grad_norm": 0.34836164116859436, "learning_rate": 9.280044290402126e-05, "loss": 0.0375, "step": 21430 }, { "grad_norm": 0.2974851727485657, "learning_rate": 9.279189276722821e-05, "loss": 0.0364, "step": 21440 }, { "grad_norm": 0.18847639858722687, "learning_rate": 9.278333795077812e-05, "loss": 0.0371, "step": 21450 }, { "grad_norm": 0.2906855046749115, "learning_rate": 9.27747784556065e-05, "loss": 0.036, "step": 21460 }, { "grad_norm": 0.212025448679924, "learning_rate": 9.276621428264942e-05, "loss": 0.028, "step": 21470 }, { "grad_norm": 0.24521702527999878, "learning_rate": 9.275764543284345e-05, "loss": 0.0335, "step": 21480 }, { "grad_norm": 0.22529275715351105, "learning_rate": 9.274907190712566e-05, "loss": 0.03, "step": 21490 }, { "grad_norm": 0.20324617624282837, "learning_rate": 9.274049370643363e-05, "loss": 0.0292, "step": 21500 }, { "grad_norm": 0.20992890000343323, "learning_rate": 9.273191083170547e-05, "loss": 0.0336, "step": 21510 }, { "grad_norm": 0.21054068207740784, "learning_rate": 9.27233232838798e-05, "loss": 0.0318, "step": 21520 }, { "grad_norm": 0.2715648412704468, "learning_rate": 9.27147310638957e-05, "loss": 0.0345, "step": 21530 }, { "grad_norm": 0.24625226855278015, "learning_rate": 9.270613417269286e-05, "loss": 0.0327, "step": 21540 }, { "grad_norm": 0.23223631083965302, "learning_rate": 9.269753261121138e-05, "loss": 0.0318, "step": 21550 }, { "grad_norm": 0.37876829504966736, "learning_rate": 9.268892638039194e-05, "loss": 0.0322, "step": 21560 }, { "grad_norm": 0.19293083250522614, "learning_rate": 9.268031548117569e-05, "loss": 0.0402, "step": 21570 }, { "grad_norm": 0.24927131831645966, "learning_rate": 9.26716999145043e-05, "loss": 0.034, "step": 21580 }, { "grad_norm": 0.22020363807678223, "learning_rate": 9.266307968131998e-05, "loss": 0.0296, "step": 21590 }, { "grad_norm": 0.20669692754745483, "learning_rate": 9.26544547825654e-05, "loss": 0.0368, "step": 21600 }, { "grad_norm": 0.17111247777938843, "learning_rate": 9.264582521918376e-05, "loss": 0.0311, "step": 21610 }, { "grad_norm": 0.19626955687999725, "learning_rate": 9.263719099211881e-05, "loss": 0.0275, "step": 21620 }, { "grad_norm": 0.23984920978546143, "learning_rate": 9.262855210231476e-05, "loss": 0.0294, "step": 21630 }, { "grad_norm": 0.19484852254390717, "learning_rate": 9.261990855071633e-05, "loss": 0.03, "step": 21640 }, { "grad_norm": 0.30557867884635925, "learning_rate": 9.261126033826878e-05, "loss": 0.0307, "step": 21650 }, { "grad_norm": 0.21189765632152557, "learning_rate": 9.260260746591786e-05, "loss": 0.033, "step": 21660 }, { "grad_norm": 0.34568506479263306, "learning_rate": 9.259394993460985e-05, "loss": 0.0293, "step": 21670 }, { "grad_norm": 0.26486456394195557, "learning_rate": 9.258528774529151e-05, "loss": 0.0292, "step": 21680 }, { "grad_norm": 0.24302515387535095, "learning_rate": 9.257662089891013e-05, "loss": 0.0289, "step": 21690 }, { "grad_norm": 0.253318727016449, "learning_rate": 9.25679493964135e-05, "loss": 0.0291, "step": 21700 }, { "grad_norm": 0.2568321228027344, "learning_rate": 9.255927323874994e-05, "loss": 0.0391, "step": 21710 }, { "grad_norm": 0.31046825647354126, "learning_rate": 9.255059242686822e-05, "loss": 0.034, "step": 21720 }, { "grad_norm": 0.22399789094924927, "learning_rate": 9.254190696171769e-05, "loss": 0.0315, "step": 21730 }, { "grad_norm": 0.3161284029483795, "learning_rate": 9.25332168442482e-05, "loss": 0.033, "step": 21740 }, { "grad_norm": 0.24158740043640137, "learning_rate": 9.252452207541004e-05, "loss": 0.032, "step": 21750 }, { "grad_norm": 0.36985480785369873, "learning_rate": 9.251582265615409e-05, "loss": 0.033, "step": 21760 }, { "grad_norm": 0.2768970727920532, "learning_rate": 9.250711858743169e-05, "loss": 0.0335, "step": 21770 }, { "grad_norm": 0.27711114287376404, "learning_rate": 9.24984098701947e-05, "loss": 0.0308, "step": 21780 }, { "grad_norm": 0.19113120436668396, "learning_rate": 9.248969650539552e-05, "loss": 0.0232, "step": 21790 }, { "grad_norm": 0.19666339457035065, "learning_rate": 9.2480978493987e-05, "loss": 0.032, "step": 21800 }, { "grad_norm": 0.26941409707069397, "learning_rate": 9.247225583692256e-05, "loss": 0.0311, "step": 21810 }, { "grad_norm": 0.2941311001777649, "learning_rate": 9.246352853515607e-05, "loss": 0.0287, "step": 21820 }, { "grad_norm": 0.22929200530052185, "learning_rate": 9.245479658964194e-05, "loss": 0.0306, "step": 21830 }, { "grad_norm": 0.29233211278915405, "learning_rate": 9.244606000133507e-05, "loss": 0.0382, "step": 21840 }, { "grad_norm": 0.2891162633895874, "learning_rate": 9.24373187711909e-05, "loss": 0.0293, "step": 21850 }, { "grad_norm": 0.34612417221069336, "learning_rate": 9.242857290016537e-05, "loss": 0.033, "step": 21860 }, { "grad_norm": 0.221953347325325, "learning_rate": 9.241982238921488e-05, "loss": 0.0283, "step": 21870 }, { "grad_norm": 0.2540751099586487, "learning_rate": 9.24110672392964e-05, "loss": 0.0311, "step": 21880 }, { "grad_norm": 0.21682779490947723, "learning_rate": 9.240230745136737e-05, "loss": 0.0283, "step": 21890 }, { "grad_norm": 0.26362407207489014, "learning_rate": 9.239354302638575e-05, "loss": 0.0319, "step": 21900 }, { "grad_norm": 0.3144509792327881, "learning_rate": 9.238477396531e-05, "loss": 0.036, "step": 21910 }, { "grad_norm": 0.30001187324523926, "learning_rate": 9.23760002690991e-05, "loss": 0.0338, "step": 21920 }, { "grad_norm": 0.3475881814956665, "learning_rate": 9.236722193871252e-05, "loss": 0.0329, "step": 21930 }, { "grad_norm": 0.24226142466068268, "learning_rate": 9.235843897511023e-05, "loss": 0.0381, "step": 21940 }, { "grad_norm": 0.3487955331802368, "learning_rate": 9.234965137925276e-05, "loss": 0.039, "step": 21950 }, { "grad_norm": 0.23574021458625793, "learning_rate": 9.234085915210108e-05, "loss": 0.0335, "step": 21960 }, { "grad_norm": 0.3618949055671692, "learning_rate": 9.23320622946167e-05, "loss": 0.036, "step": 21970 }, { "grad_norm": 0.31440070271492004, "learning_rate": 9.232326080776163e-05, "loss": 0.0362, "step": 21980 }, { "grad_norm": 0.1916724443435669, "learning_rate": 9.23144546924984e-05, "loss": 0.0323, "step": 21990 }, { "grad_norm": 0.23577283322811127, "learning_rate": 9.230564394979e-05, "loss": 0.0288, "step": 22000 }, { "grad_norm": 0.21983641386032104, "learning_rate": 9.22968285806e-05, "loss": 0.0275, "step": 22010 }, { "grad_norm": 0.1789676398038864, "learning_rate": 9.228800858589242e-05, "loss": 0.0308, "step": 22020 }, { "grad_norm": 0.20344503223896027, "learning_rate": 9.227918396663179e-05, "loss": 0.0287, "step": 22030 }, { "grad_norm": 0.2368427813053131, "learning_rate": 9.227035472378319e-05, "loss": 0.0286, "step": 22040 }, { "grad_norm": 0.3201330900192261, "learning_rate": 9.226152085831213e-05, "loss": 0.0322, "step": 22050 }, { "grad_norm": 0.2127286195755005, "learning_rate": 9.22526823711847e-05, "loss": 0.0315, "step": 22060 }, { "grad_norm": 0.17431987822055817, "learning_rate": 9.224383926336745e-05, "loss": 0.0254, "step": 22070 }, { "grad_norm": 0.319038987159729, "learning_rate": 9.223499153582744e-05, "loss": 0.0341, "step": 22080 }, { "grad_norm": 0.2155311107635498, "learning_rate": 9.222613918953226e-05, "loss": 0.0323, "step": 22090 }, { "grad_norm": 0.21967588365077972, "learning_rate": 9.221728222544999e-05, "loss": 0.0316, "step": 22100 }, { "grad_norm": 0.25821784138679504, "learning_rate": 9.22084206445492e-05, "loss": 0.0312, "step": 22110 }, { "grad_norm": 0.24136671423912048, "learning_rate": 9.2199554447799e-05, "loss": 0.0265, "step": 22120 }, { "grad_norm": 0.22243903577327728, "learning_rate": 9.219068363616897e-05, "loss": 0.0294, "step": 22130 }, { "grad_norm": 0.210862398147583, "learning_rate": 9.218180821062919e-05, "loss": 0.0278, "step": 22140 }, { "grad_norm": 0.35154134035110474, "learning_rate": 9.21729281721503e-05, "loss": 0.0342, "step": 22150 }, { "grad_norm": 0.25058209896087646, "learning_rate": 9.216404352170339e-05, "loss": 0.0332, "step": 22160 }, { "grad_norm": 0.2809924781322479, "learning_rate": 9.215515426026007e-05, "loss": 0.0341, "step": 22170 }, { "grad_norm": 0.3176141083240509, "learning_rate": 9.214626038879246e-05, "loss": 0.031, "step": 22180 }, { "grad_norm": 0.25593841075897217, "learning_rate": 9.21373619082732e-05, "loss": 0.0347, "step": 22190 }, { "grad_norm": 0.25951749086380005, "learning_rate": 9.212845881967535e-05, "loss": 0.0333, "step": 22200 }, { "grad_norm": 0.23401924967765808, "learning_rate": 9.211955112397262e-05, "loss": 0.0287, "step": 22210 }, { "grad_norm": 0.2549034357070923, "learning_rate": 9.211063882213909e-05, "loss": 0.0287, "step": 22220 }, { "grad_norm": 0.30894598364830017, "learning_rate": 9.210172191514942e-05, "loss": 0.0326, "step": 22230 }, { "grad_norm": 0.3116694390773773, "learning_rate": 9.209280040397874e-05, "loss": 0.0341, "step": 22240 }, { "grad_norm": 0.2266339659690857, "learning_rate": 9.208387428960268e-05, "loss": 0.0329, "step": 22250 }, { "grad_norm": 0.3471604883670807, "learning_rate": 9.20749435729974e-05, "loss": 0.032, "step": 22260 }, { "grad_norm": 0.31277984380722046, "learning_rate": 9.206600825513957e-05, "loss": 0.0391, "step": 22270 }, { "grad_norm": 0.2874821424484253, "learning_rate": 9.20570683370063e-05, "loss": 0.0305, "step": 22280 }, { "grad_norm": 0.358076810836792, "learning_rate": 9.204812381957528e-05, "loss": 0.0331, "step": 22290 }, { "grad_norm": 0.25988054275512695, "learning_rate": 9.203917470382465e-05, "loss": 0.0375, "step": 22300 }, { "grad_norm": 0.2840164601802826, "learning_rate": 9.203022099073309e-05, "loss": 0.0316, "step": 22310 }, { "grad_norm": 0.32209572196006775, "learning_rate": 9.202126268127976e-05, "loss": 0.0355, "step": 22320 }, { "grad_norm": 0.4795507788658142, "learning_rate": 9.20122997764443e-05, "loss": 0.0342, "step": 22330 }, { "grad_norm": 0.471048504114151, "learning_rate": 9.200333227720692e-05, "loss": 0.0375, "step": 22340 }, { "grad_norm": 0.2948104739189148, "learning_rate": 9.199436018454826e-05, "loss": 0.0368, "step": 22350 }, { "grad_norm": 0.272460013628006, "learning_rate": 9.198538349944952e-05, "loss": 0.03, "step": 22360 }, { "grad_norm": 0.4012543857097626, "learning_rate": 9.197640222289234e-05, "loss": 0.0368, "step": 22370 }, { "grad_norm": 0.3287928104400635, "learning_rate": 9.196741635585895e-05, "loss": 0.0367, "step": 22380 }, { "grad_norm": 0.3190231919288635, "learning_rate": 9.195842589933199e-05, "loss": 0.0372, "step": 22390 }, { "grad_norm": 0.2508600056171417, "learning_rate": 9.194943085429466e-05, "loss": 0.0413, "step": 22400 }, { "grad_norm": 0.25800350308418274, "learning_rate": 9.194043122173065e-05, "loss": 0.0407, "step": 22410 }, { "grad_norm": 0.2205089032649994, "learning_rate": 9.193142700262413e-05, "loss": 0.0322, "step": 22420 }, { "grad_norm": 0.2353154867887497, "learning_rate": 9.192241819795979e-05, "loss": 0.0324, "step": 22430 }, { "grad_norm": 0.2688516080379486, "learning_rate": 9.191340480872284e-05, "loss": 0.0315, "step": 22440 }, { "grad_norm": 0.24455901980400085, "learning_rate": 9.190438683589895e-05, "loss": 0.0315, "step": 22450 }, { "grad_norm": 0.2956455647945404, "learning_rate": 9.189536428047432e-05, "loss": 0.0328, "step": 22460 }, { "grad_norm": 0.2186010181903839, "learning_rate": 9.188633714343564e-05, "loss": 0.0325, "step": 22470 }, { "grad_norm": 0.24396714568138123, "learning_rate": 9.18773054257701e-05, "loss": 0.0341, "step": 22480 }, { "grad_norm": 0.40472614765167236, "learning_rate": 9.18682691284654e-05, "loss": 0.0326, "step": 22490 }, { "grad_norm": 0.2861737906932831, "learning_rate": 9.185922825250974e-05, "loss": 0.04, "step": 22500 }, { "grad_norm": 0.5297004580497742, "learning_rate": 9.185018279889181e-05, "loss": 0.0413, "step": 22510 }, { "grad_norm": 0.2645062208175659, "learning_rate": 9.184113276860082e-05, "loss": 0.0364, "step": 22520 }, { "grad_norm": 0.3367396295070648, "learning_rate": 9.183207816262645e-05, "loss": 0.0438, "step": 22530 }, { "grad_norm": 0.35294869542121887, "learning_rate": 9.182301898195891e-05, "loss": 0.0399, "step": 22540 }, { "grad_norm": 0.35833030939102173, "learning_rate": 9.181395522758889e-05, "loss": 0.0369, "step": 22550 }, { "grad_norm": 0.2517184019088745, "learning_rate": 9.180488690050759e-05, "loss": 0.0358, "step": 22560 }, { "grad_norm": 0.34529033303260803, "learning_rate": 9.179581400170671e-05, "loss": 0.0303, "step": 22570 }, { "grad_norm": 0.37647655606269836, "learning_rate": 9.178673653217845e-05, "loss": 0.0354, "step": 22580 }, { "grad_norm": 0.298834890127182, "learning_rate": 9.177765449291551e-05, "loss": 0.037, "step": 22590 }, { "grad_norm": 0.18827305734157562, "learning_rate": 9.176856788491109e-05, "loss": 0.0293, "step": 22600 }, { "grad_norm": 0.25183233618736267, "learning_rate": 9.175947670915887e-05, "loss": 0.0331, "step": 22610 }, { "grad_norm": 0.36598584055900574, "learning_rate": 9.175038096665309e-05, "loss": 0.04, "step": 22620 }, { "grad_norm": 0.2775095999240875, "learning_rate": 9.17412806583884e-05, "loss": 0.0336, "step": 22630 }, { "grad_norm": 0.19347107410430908, "learning_rate": 9.173217578536002e-05, "loss": 0.0326, "step": 22640 }, { "grad_norm": 0.27067238092422485, "learning_rate": 9.172306634856362e-05, "loss": 0.0329, "step": 22650 }, { "grad_norm": 0.3134446442127228, "learning_rate": 9.171395234899545e-05, "loss": 0.0307, "step": 22660 }, { "grad_norm": 0.4275733530521393, "learning_rate": 9.170483378765214e-05, "loss": 0.037, "step": 22670 }, { "grad_norm": 0.1871904730796814, "learning_rate": 9.169571066553091e-05, "loss": 0.0304, "step": 22680 }, { "grad_norm": 0.28322649002075195, "learning_rate": 9.168658298362946e-05, "loss": 0.0347, "step": 22690 }, { "grad_norm": 0.24346423149108887, "learning_rate": 9.167745074294598e-05, "loss": 0.0348, "step": 22700 }, { "grad_norm": 0.19162265956401825, "learning_rate": 9.166831394447913e-05, "loss": 0.0411, "step": 22710 }, { "grad_norm": 1.1769920587539673, "learning_rate": 9.165917258922812e-05, "loss": 0.0412, "step": 22720 }, { "grad_norm": 0.2643350660800934, "learning_rate": 9.165002667819262e-05, "loss": 0.033, "step": 22730 }, { "grad_norm": 0.2919580936431885, "learning_rate": 9.164087621237282e-05, "loss": 0.0362, "step": 22740 }, { "grad_norm": 0.21342577040195465, "learning_rate": 9.163172119276942e-05, "loss": 0.0362, "step": 22750 }, { "grad_norm": 0.2910279631614685, "learning_rate": 9.162256162038358e-05, "loss": 0.037, "step": 22760 }, { "grad_norm": 0.4714600145816803, "learning_rate": 9.161339749621698e-05, "loss": 0.0417, "step": 22770 }, { "grad_norm": 0.3409506678581238, "learning_rate": 9.160422882127177e-05, "loss": 0.0339, "step": 22780 }, { "grad_norm": 0.26413029432296753, "learning_rate": 9.159505559655069e-05, "loss": 0.0326, "step": 22790 }, { "grad_norm": 0.3437291383743286, "learning_rate": 9.158587782305684e-05, "loss": 0.0306, "step": 22800 }, { "grad_norm": 0.27955159544944763, "learning_rate": 9.157669550179391e-05, "loss": 0.0303, "step": 22810 }, { "grad_norm": 0.3038710057735443, "learning_rate": 9.156750863376609e-05, "loss": 0.034, "step": 22820 }, { "grad_norm": 0.30428192019462585, "learning_rate": 9.155831721997801e-05, "loss": 0.0489, "step": 22830 }, { "grad_norm": 0.29279816150665283, "learning_rate": 9.154912126143484e-05, "loss": 0.0413, "step": 22840 }, { "grad_norm": 0.33986619114875793, "learning_rate": 9.153992075914224e-05, "loss": 0.0429, "step": 22850 }, { "grad_norm": 0.38516101241111755, "learning_rate": 9.153071571410635e-05, "loss": 0.0361, "step": 22860 }, { "grad_norm": 0.28435224294662476, "learning_rate": 9.152150612733384e-05, "loss": 0.0351, "step": 22870 }, { "grad_norm": 0.3072422444820404, "learning_rate": 9.151229199983184e-05, "loss": 0.0377, "step": 22880 }, { "grad_norm": 0.5171369314193726, "learning_rate": 9.150307333260802e-05, "loss": 0.0303, "step": 22890 }, { "grad_norm": 0.2852562367916107, "learning_rate": 9.149385012667048e-05, "loss": 0.0313, "step": 22900 }, { "grad_norm": 0.3316056728363037, "learning_rate": 9.148462238302788e-05, "loss": 0.0349, "step": 22910 }, { "grad_norm": 0.2797139585018158, "learning_rate": 9.147539010268936e-05, "loss": 0.035, "step": 22920 }, { "grad_norm": 0.28061118721961975, "learning_rate": 9.14661532866645e-05, "loss": 0.0311, "step": 22930 }, { "grad_norm": 0.2287835031747818, "learning_rate": 9.145691193596348e-05, "loss": 0.0331, "step": 22940 }, { "grad_norm": 0.2216254323720932, "learning_rate": 9.144766605159691e-05, "loss": 0.0369, "step": 22950 }, { "grad_norm": 0.21579250693321228, "learning_rate": 9.14384156345759e-05, "loss": 0.0337, "step": 22960 }, { "grad_norm": 0.4217585325241089, "learning_rate": 9.142916068591204e-05, "loss": 0.0336, "step": 22970 }, { "grad_norm": 0.3374784588813782, "learning_rate": 9.141990120661746e-05, "loss": 0.0328, "step": 22980 }, { "grad_norm": 0.30703186988830566, "learning_rate": 9.141063719770475e-05, "loss": 0.0387, "step": 22990 }, { "grad_norm": 0.20038624107837677, "learning_rate": 9.140136866018704e-05, "loss": 0.0286, "step": 23000 }, { "grad_norm": 0.2363683134317398, "learning_rate": 9.139209559507788e-05, "loss": 0.0323, "step": 23010 }, { "grad_norm": 0.2422236055135727, "learning_rate": 9.13828180033914e-05, "loss": 0.0333, "step": 23020 }, { "grad_norm": 0.2244807630777359, "learning_rate": 9.137353588614212e-05, "loss": 0.0326, "step": 23030 }, { "grad_norm": 0.20712031424045563, "learning_rate": 9.136424924434519e-05, "loss": 0.0349, "step": 23040 }, { "grad_norm": 0.23031477630138397, "learning_rate": 9.135495807901615e-05, "loss": 0.0326, "step": 23050 }, { "grad_norm": 0.28837066888809204, "learning_rate": 9.134566239117108e-05, "loss": 0.0338, "step": 23060 }, { "grad_norm": 0.2620249092578888, "learning_rate": 9.13363621818265e-05, "loss": 0.0327, "step": 23070 }, { "grad_norm": 0.3115917146205902, "learning_rate": 9.132705745199953e-05, "loss": 0.0341, "step": 23080 }, { "grad_norm": 0.28389930725097656, "learning_rate": 9.131774820270768e-05, "loss": 0.0292, "step": 23090 }, { "grad_norm": 0.2488791048526764, "learning_rate": 9.130843443496901e-05, "loss": 0.0332, "step": 23100 }, { "grad_norm": 0.25766998529434204, "learning_rate": 9.129911614980206e-05, "loss": 0.0377, "step": 23110 }, { "grad_norm": 0.21278920769691467, "learning_rate": 9.128979334822584e-05, "loss": 0.0317, "step": 23120 }, { "grad_norm": 0.24379019439220428, "learning_rate": 9.128046603125992e-05, "loss": 0.0329, "step": 23130 }, { "grad_norm": 0.266310453414917, "learning_rate": 9.12711341999243e-05, "loss": 0.0324, "step": 23140 }, { "grad_norm": 0.2986176609992981, "learning_rate": 9.12617978552395e-05, "loss": 0.0297, "step": 23150 }, { "grad_norm": 0.29845479130744934, "learning_rate": 9.12524569982265e-05, "loss": 0.0283, "step": 23160 }, { "grad_norm": 0.3586942255496979, "learning_rate": 9.124311162990684e-05, "loss": 0.0296, "step": 23170 }, { "grad_norm": 0.8173752427101135, "learning_rate": 9.12337617513025e-05, "loss": 0.0289, "step": 23180 }, { "grad_norm": 0.1957702934741974, "learning_rate": 9.122440736343596e-05, "loss": 0.0276, "step": 23190 }, { "grad_norm": 0.43991273641586304, "learning_rate": 9.12150484673302e-05, "loss": 0.0349, "step": 23200 }, { "grad_norm": 0.6849054098129272, "learning_rate": 9.120568506400873e-05, "loss": 0.0329, "step": 23210 }, { "grad_norm": 0.21839381754398346, "learning_rate": 9.119631715449548e-05, "loss": 0.0307, "step": 23220 }, { "grad_norm": 0.282320499420166, "learning_rate": 9.118694473981493e-05, "loss": 0.0335, "step": 23230 }, { "grad_norm": 0.20710068941116333, "learning_rate": 9.117756782099203e-05, "loss": 0.0372, "step": 23240 }, { "grad_norm": 0.28861939907073975, "learning_rate": 9.11681863990522e-05, "loss": 0.0317, "step": 23250 }, { "grad_norm": 0.28658878803253174, "learning_rate": 9.115880047502142e-05, "loss": 0.0288, "step": 23260 }, { "grad_norm": 0.39898353815078735, "learning_rate": 9.114941004992609e-05, "loss": 0.0309, "step": 23270 }, { "grad_norm": 0.24393020570278168, "learning_rate": 9.114001512479317e-05, "loss": 0.0328, "step": 23280 }, { "grad_norm": 0.2858683168888092, "learning_rate": 9.113061570065003e-05, "loss": 0.0311, "step": 23290 }, { "grad_norm": 0.2680872082710266, "learning_rate": 9.112121177852459e-05, "loss": 0.0303, "step": 23300 }, { "grad_norm": 0.32640141248703003, "learning_rate": 9.111180335944527e-05, "loss": 0.032, "step": 23310 }, { "grad_norm": 0.26004457473754883, "learning_rate": 9.110239044444093e-05, "loss": 0.0316, "step": 23320 }, { "grad_norm": 0.25720566511154175, "learning_rate": 9.109297303454099e-05, "loss": 0.0356, "step": 23330 }, { "grad_norm": 0.2828614115715027, "learning_rate": 9.108355113077526e-05, "loss": 0.0335, "step": 23340 }, { "grad_norm": 0.2722148597240448, "learning_rate": 9.107412473417419e-05, "loss": 0.0298, "step": 23350 }, { "grad_norm": 0.29995328187942505, "learning_rate": 9.106469384576858e-05, "loss": 0.0317, "step": 23360 }, { "grad_norm": 0.2848416268825531, "learning_rate": 9.105525846658978e-05, "loss": 0.0331, "step": 23370 }, { "grad_norm": 0.3517891466617584, "learning_rate": 9.104581859766965e-05, "loss": 0.0412, "step": 23380 }, { "grad_norm": 0.24797603487968445, "learning_rate": 9.10363742400405e-05, "loss": 0.0347, "step": 23390 }, { "grad_norm": 0.29945117235183716, "learning_rate": 9.102692539473518e-05, "loss": 0.0324, "step": 23400 }, { "grad_norm": 0.3813876807689667, "learning_rate": 9.101747206278697e-05, "loss": 0.0342, "step": 23410 }, { "grad_norm": 0.40304023027420044, "learning_rate": 9.100801424522968e-05, "loss": 0.0356, "step": 23420 }, { "grad_norm": 0.360796719789505, "learning_rate": 9.099855194309762e-05, "loss": 0.0322, "step": 23430 }, { "grad_norm": 0.2551528811454773, "learning_rate": 9.098908515742554e-05, "loss": 0.0351, "step": 23440 }, { "grad_norm": 0.24853511154651642, "learning_rate": 9.097961388924873e-05, "loss": 0.0303, "step": 23450 }, { "grad_norm": 0.254309743642807, "learning_rate": 9.097013813960298e-05, "loss": 0.0271, "step": 23460 }, { "grad_norm": 0.40606003999710083, "learning_rate": 9.09606579095245e-05, "loss": 0.0311, "step": 23470 }, { "grad_norm": 0.2227974534034729, "learning_rate": 9.095117320005008e-05, "loss": 0.0299, "step": 23480 }, { "grad_norm": 0.21811142563819885, "learning_rate": 9.094168401221691e-05, "loss": 0.025, "step": 23490 }, { "grad_norm": 0.4957847595214844, "learning_rate": 9.093219034706273e-05, "loss": 0.0294, "step": 23500 }, { "grad_norm": 0.21731005609035492, "learning_rate": 9.092269220562577e-05, "loss": 0.0301, "step": 23510 }, { "grad_norm": 0.31905397772789, "learning_rate": 9.09131895889447e-05, "loss": 0.0364, "step": 23520 }, { "grad_norm": 0.2657492160797119, "learning_rate": 9.090368249805873e-05, "loss": 0.0322, "step": 23530 }, { "grad_norm": 0.2847647964954376, "learning_rate": 9.089417093400754e-05, "loss": 0.0319, "step": 23540 }, { "grad_norm": 0.21292082965373993, "learning_rate": 9.088465489783131e-05, "loss": 0.0292, "step": 23550 }, { "grad_norm": 0.24092859029769897, "learning_rate": 9.087513439057068e-05, "loss": 0.0356, "step": 23560 }, { "grad_norm": 0.40184611082077026, "learning_rate": 9.08656094132668e-05, "loss": 0.0303, "step": 23570 }, { "grad_norm": 0.6445716619491577, "learning_rate": 9.085607996696134e-05, "loss": 0.0296, "step": 23580 }, { "grad_norm": 0.29045775532722473, "learning_rate": 9.084654605269639e-05, "loss": 0.0384, "step": 23590 }, { "grad_norm": 0.4137551784515381, "learning_rate": 9.083700767151457e-05, "loss": 0.0367, "step": 23600 }, { "grad_norm": 0.3983955979347229, "learning_rate": 9.082746482445898e-05, "loss": 0.0324, "step": 23610 }, { "grad_norm": 0.2747040092945099, "learning_rate": 9.081791751257325e-05, "loss": 0.0316, "step": 23620 }, { "grad_norm": 0.23360727727413177, "learning_rate": 9.080836573690142e-05, "loss": 0.0349, "step": 23630 }, { "grad_norm": 0.32527193427085876, "learning_rate": 9.079880949848805e-05, "loss": 0.043, "step": 23640 }, { "grad_norm": 0.27729108929634094, "learning_rate": 9.078924879837822e-05, "loss": 0.0354, "step": 23650 }, { "grad_norm": 0.2662149965763092, "learning_rate": 9.077968363761747e-05, "loss": 0.0388, "step": 23660 }, { "grad_norm": 0.2644321918487549, "learning_rate": 9.077011401725182e-05, "loss": 0.0317, "step": 23670 }, { "grad_norm": 0.26162388920783997, "learning_rate": 9.07605399383278e-05, "loss": 0.0354, "step": 23680 }, { "grad_norm": 0.2610011100769043, "learning_rate": 9.075096140189243e-05, "loss": 0.0365, "step": 23690 }, { "grad_norm": 0.28169190883636475, "learning_rate": 9.074137840899318e-05, "loss": 0.0273, "step": 23700 }, { "grad_norm": 0.18956434726715088, "learning_rate": 9.073179096067804e-05, "loss": 0.0308, "step": 23710 }, { "grad_norm": 0.2838855981826782, "learning_rate": 9.072219905799549e-05, "loss": 0.03, "step": 23720 }, { "grad_norm": 0.3021395206451416, "learning_rate": 9.071260270199447e-05, "loss": 0.0327, "step": 23730 }, { "grad_norm": 0.25926798582077026, "learning_rate": 9.070300189372441e-05, "loss": 0.0318, "step": 23740 }, { "grad_norm": 0.30788037180900574, "learning_rate": 9.069339663423528e-05, "loss": 0.0312, "step": 23750 }, { "grad_norm": 0.31756967306137085, "learning_rate": 9.068378692457747e-05, "loss": 0.0343, "step": 23760 }, { "grad_norm": 0.29448723793029785, "learning_rate": 9.067417276580189e-05, "loss": 0.0306, "step": 23770 }, { "grad_norm": 0.1588144749403, "learning_rate": 9.066455415895993e-05, "loss": 0.0342, "step": 23780 }, { "grad_norm": 0.2678372263908386, "learning_rate": 9.065493110510346e-05, "loss": 0.0325, "step": 23790 }, { "grad_norm": 0.38227710127830505, "learning_rate": 9.064530360528484e-05, "loss": 0.0392, "step": 23800 }, { "grad_norm": 0.2251160591840744, "learning_rate": 9.063567166055695e-05, "loss": 0.0316, "step": 23810 }, { "grad_norm": 0.2540663182735443, "learning_rate": 9.062603527197308e-05, "loss": 0.0326, "step": 23820 }, { "grad_norm": 0.26088619232177734, "learning_rate": 9.06163944405871e-05, "loss": 0.0312, "step": 23830 }, { "grad_norm": 0.22945234179496765, "learning_rate": 9.060674916745327e-05, "loss": 0.0311, "step": 23840 }, { "grad_norm": 0.24355974793434143, "learning_rate": 9.05970994536264e-05, "loss": 0.0269, "step": 23850 }, { "grad_norm": 0.27316150069236755, "learning_rate": 9.05874453001618e-05, "loss": 0.0334, "step": 23860 }, { "grad_norm": 0.2226804941892624, "learning_rate": 9.057778670811517e-05, "loss": 0.0284, "step": 23870 }, { "grad_norm": 0.19551366567611694, "learning_rate": 9.056812367854281e-05, "loss": 0.0305, "step": 23880 }, { "grad_norm": 0.3209216892719269, "learning_rate": 9.055845621250143e-05, "loss": 0.0367, "step": 23890 }, { "grad_norm": 0.4380621612071991, "learning_rate": 9.054878431104825e-05, "loss": 0.0322, "step": 23900 }, { "grad_norm": 0.2965327501296997, "learning_rate": 9.0539107975241e-05, "loss": 0.0325, "step": 23910 }, { "grad_norm": 0.4056030809879303, "learning_rate": 9.052942720613784e-05, "loss": 0.0348, "step": 23920 }, { "grad_norm": 0.27235472202301025, "learning_rate": 9.051974200479745e-05, "loss": 0.0309, "step": 23930 }, { "grad_norm": 0.2524045705795288, "learning_rate": 9.051005237227901e-05, "loss": 0.0314, "step": 23940 }, { "grad_norm": 0.20371025800704956, "learning_rate": 9.050035830964215e-05, "loss": 0.0307, "step": 23950 }, { "grad_norm": 0.33859652280807495, "learning_rate": 9.049065981794698e-05, "loss": 0.037, "step": 23960 }, { "grad_norm": 0.30618953704833984, "learning_rate": 9.048095689825414e-05, "loss": 0.029, "step": 23970 }, { "grad_norm": 0.24059723317623138, "learning_rate": 9.047124955162472e-05, "loss": 0.0341, "step": 23980 }, { "grad_norm": 0.24526013433933258, "learning_rate": 9.046153777912028e-05, "loss": 0.0352, "step": 23990 }, { "grad_norm": 0.9088840484619141, "learning_rate": 9.045182158180292e-05, "loss": 0.037, "step": 24000 }, { "grad_norm": 0.5320703983306885, "learning_rate": 9.044210096073516e-05, "loss": 0.042, "step": 24010 }, { "grad_norm": 0.23230469226837158, "learning_rate": 9.043237591698004e-05, "loss": 0.0477, "step": 24020 }, { "grad_norm": 0.37325888872146606, "learning_rate": 9.04226464516011e-05, "loss": 0.0417, "step": 24030 }, { "grad_norm": 0.26924192905426025, "learning_rate": 9.041291256566229e-05, "loss": 0.0404, "step": 24040 }, { "grad_norm": 0.300161749124527, "learning_rate": 9.040317426022814e-05, "loss": 0.0398, "step": 24050 }, { "grad_norm": 0.4587409496307373, "learning_rate": 9.03934315363636e-05, "loss": 0.0357, "step": 24060 }, { "grad_norm": 0.23339450359344482, "learning_rate": 9.038368439513409e-05, "loss": 0.033, "step": 24070 }, { "grad_norm": 0.2632056474685669, "learning_rate": 9.03739328376056e-05, "loss": 0.0328, "step": 24080 }, { "grad_norm": 0.31608492136001587, "learning_rate": 9.036417686484451e-05, "loss": 0.0336, "step": 24090 }, { "grad_norm": 0.25159233808517456, "learning_rate": 9.035441647791773e-05, "loss": 0.0273, "step": 24100 }, { "grad_norm": 0.2615780234336853, "learning_rate": 9.034465167789263e-05, "loss": 0.0314, "step": 24110 }, { "grad_norm": 0.24058881402015686, "learning_rate": 9.033488246583706e-05, "loss": 0.0336, "step": 24120 }, { "grad_norm": 0.211813822388649, "learning_rate": 9.032510884281941e-05, "loss": 0.029, "step": 24130 }, { "grad_norm": 0.27373257279396057, "learning_rate": 9.031533080990848e-05, "loss": 0.0327, "step": 24140 }, { "grad_norm": 0.26738518476486206, "learning_rate": 9.030554836817358e-05, "loss": 0.0279, "step": 24150 }, { "grad_norm": 0.4542975127696991, "learning_rate": 9.029576151868451e-05, "loss": 0.0303, "step": 24160 }, { "grad_norm": 0.2290855497121811, "learning_rate": 9.028597026251155e-05, "loss": 0.0299, "step": 24170 }, { "grad_norm": 0.2461289018392563, "learning_rate": 9.027617460072547e-05, "loss": 0.0303, "step": 24180 }, { "grad_norm": 0.33087459206581116, "learning_rate": 9.026637453439745e-05, "loss": 0.0386, "step": 24190 }, { "grad_norm": 0.22556163370609283, "learning_rate": 9.025657006459927e-05, "loss": 0.0274, "step": 24200 }, { "grad_norm": 0.29171308875083923, "learning_rate": 9.024676119240311e-05, "loss": 0.0376, "step": 24210 }, { "grad_norm": 0.3188537061214447, "learning_rate": 9.023694791888166e-05, "loss": 0.0308, "step": 24220 }, { "grad_norm": 0.3616294264793396, "learning_rate": 9.022713024510808e-05, "loss": 0.0345, "step": 24230 }, { "grad_norm": 0.2545093297958374, "learning_rate": 9.021730817215601e-05, "loss": 0.031, "step": 24240 }, { "grad_norm": 0.2557913064956665, "learning_rate": 9.02074817010996e-05, "loss": 0.0344, "step": 24250 }, { "grad_norm": 0.23256196081638336, "learning_rate": 9.019765083301342e-05, "loss": 0.0328, "step": 24260 }, { "grad_norm": 0.2712070047855377, "learning_rate": 9.01878155689726e-05, "loss": 0.0326, "step": 24270 }, { "grad_norm": 0.38206639885902405, "learning_rate": 9.017797591005268e-05, "loss": 0.0349, "step": 24280 }, { "grad_norm": 0.33119454979896545, "learning_rate": 9.016813185732972e-05, "loss": 0.0339, "step": 24290 }, { "grad_norm": 0.2911086082458496, "learning_rate": 9.015828341188027e-05, "loss": 0.0398, "step": 24300 }, { "grad_norm": 0.3613594174385071, "learning_rate": 9.01484305747813e-05, "loss": 0.0384, "step": 24310 }, { "grad_norm": 0.34015774726867676, "learning_rate": 9.013857334711033e-05, "loss": 0.0427, "step": 24320 }, { "grad_norm": 0.29010263085365295, "learning_rate": 9.012871172994534e-05, "loss": 0.0433, "step": 24330 }, { "grad_norm": 0.255765825510025, "learning_rate": 9.011884572436476e-05, "loss": 0.0356, "step": 24340 }, { "grad_norm": 0.3012145459651947, "learning_rate": 9.010897533144754e-05, "loss": 0.0324, "step": 24350 }, { "grad_norm": 0.2451861947774887, "learning_rate": 9.009910055227306e-05, "loss": 0.0353, "step": 24360 }, { "grad_norm": 0.26250940561294556, "learning_rate": 9.008922138792124e-05, "loss": 0.0312, "step": 24370 }, { "grad_norm": 0.23292024433612823, "learning_rate": 9.007933783947244e-05, "loss": 0.0342, "step": 24380 }, { "grad_norm": 0.21170800924301147, "learning_rate": 9.006944990800752e-05, "loss": 0.035, "step": 24390 }, { "grad_norm": 0.29935821890830994, "learning_rate": 9.005955759460779e-05, "loss": 0.0351, "step": 24400 }, { "grad_norm": 0.26571768522262573, "learning_rate": 9.004966090035508e-05, "loss": 0.0305, "step": 24410 }, { "grad_norm": 0.241848886013031, "learning_rate": 9.003975982633166e-05, "loss": 0.0347, "step": 24420 }, { "grad_norm": 0.2516230046749115, "learning_rate": 9.00298543736203e-05, "loss": 0.0317, "step": 24430 }, { "grad_norm": 0.3223050832748413, "learning_rate": 9.001994454330427e-05, "loss": 0.0336, "step": 24440 }, { "grad_norm": 0.2889840006828308, "learning_rate": 9.001003033646727e-05, "loss": 0.0294, "step": 24450 }, { "grad_norm": 0.33135131001472473, "learning_rate": 9.00001117541935e-05, "loss": 0.0331, "step": 24460 }, { "grad_norm": 0.2161693572998047, "learning_rate": 8.999018879756764e-05, "loss": 0.0261, "step": 24470 }, { "grad_norm": 0.3006588816642761, "learning_rate": 8.998026146767487e-05, "loss": 0.0378, "step": 24480 }, { "grad_norm": 0.22739799320697784, "learning_rate": 8.99703297656008e-05, "loss": 0.0335, "step": 24490 }, { "grad_norm": 0.245121031999588, "learning_rate": 8.996039369243156e-05, "loss": 0.0324, "step": 24500 }, { "grad_norm": 0.28317299485206604, "learning_rate": 8.995045324925378e-05, "loss": 0.0312, "step": 24510 }, { "grad_norm": 0.2325022965669632, "learning_rate": 8.994050843715448e-05, "loss": 0.0312, "step": 24520 }, { "grad_norm": 0.268210232257843, "learning_rate": 8.993055925722121e-05, "loss": 0.0356, "step": 24530 }, { "grad_norm": 0.21585844457149506, "learning_rate": 8.992060571054202e-05, "loss": 0.0316, "step": 24540 }, { "grad_norm": 0.24121367931365967, "learning_rate": 8.991064779820542e-05, "loss": 0.0294, "step": 24550 }, { "grad_norm": 0.2501186728477478, "learning_rate": 8.990068552130036e-05, "loss": 0.0319, "step": 24560 }, { "grad_norm": 0.28821173310279846, "learning_rate": 8.989071888091634e-05, "loss": 0.0382, "step": 24570 }, { "grad_norm": 0.22904659807682037, "learning_rate": 8.988074787814329e-05, "loss": 0.0287, "step": 24580 }, { "grad_norm": 0.27261704206466675, "learning_rate": 8.987077251407158e-05, "loss": 0.0366, "step": 24590 }, { "grad_norm": 0.20802059769630432, "learning_rate": 8.986079278979216e-05, "loss": 0.0325, "step": 24600 }, { "grad_norm": 0.319580078125, "learning_rate": 8.985080870639635e-05, "loss": 0.0356, "step": 24610 }, { "grad_norm": 0.2777060568332672, "learning_rate": 8.984082026497603e-05, "loss": 0.032, "step": 24620 }, { "grad_norm": 0.17927008867263794, "learning_rate": 8.98308274666235e-05, "loss": 0.0255, "step": 24630 }, { "grad_norm": 0.25083285570144653, "learning_rate": 8.982083031243155e-05, "loss": 0.0307, "step": 24640 }, { "grad_norm": 0.19742335379123688, "learning_rate": 8.98108288034935e-05, "loss": 0.0294, "step": 24650 }, { "grad_norm": 0.2864329218864441, "learning_rate": 8.980082294090305e-05, "loss": 0.0278, "step": 24660 }, { "grad_norm": 0.24178071320056915, "learning_rate": 8.979081272575443e-05, "loss": 0.0352, "step": 24670 }, { "grad_norm": 0.2785872220993042, "learning_rate": 8.978079815914236e-05, "loss": 0.0366, "step": 24680 }, { "grad_norm": 0.279186487197876, "learning_rate": 8.977077924216202e-05, "loss": 0.0334, "step": 24690 }, { "grad_norm": 0.224169060587883, "learning_rate": 8.976075597590905e-05, "loss": 0.0305, "step": 24700 }, { "grad_norm": 0.19276246428489685, "learning_rate": 8.975072836147958e-05, "loss": 0.0305, "step": 24710 }, { "grad_norm": 0.693597674369812, "learning_rate": 8.974069639997025e-05, "loss": 0.0382, "step": 24720 }, { "grad_norm": 0.20501282811164856, "learning_rate": 8.973066009247808e-05, "loss": 0.0387, "step": 24730 }, { "grad_norm": 0.24068748950958252, "learning_rate": 8.972061944010066e-05, "loss": 0.0331, "step": 24740 }, { "grad_norm": 0.25833970308303833, "learning_rate": 8.971057444393603e-05, "loss": 0.0326, "step": 24750 }, { "grad_norm": 0.22453945875167847, "learning_rate": 8.970052510508268e-05, "loss": 0.0311, "step": 24760 }, { "grad_norm": 0.24098028242588043, "learning_rate": 8.969047142463959e-05, "loss": 0.029, "step": 24770 }, { "grad_norm": 0.2732883095741272, "learning_rate": 8.968041340370621e-05, "loss": 0.0289, "step": 24780 }, { "grad_norm": 0.24387629330158234, "learning_rate": 8.96703510433825e-05, "loss": 0.029, "step": 24790 }, { "grad_norm": 0.35209739208221436, "learning_rate": 8.966028434476883e-05, "loss": 0.0346, "step": 24800 }, { "grad_norm": 0.2214013785123825, "learning_rate": 8.96502133089661e-05, "loss": 0.0326, "step": 24810 }, { "grad_norm": 0.3673555254936218, "learning_rate": 8.964013793707564e-05, "loss": 0.0322, "step": 24820 }, { "grad_norm": 0.3093537986278534, "learning_rate": 8.963005823019932e-05, "loss": 0.0313, "step": 24830 }, { "grad_norm": 0.36135196685791016, "learning_rate": 8.961997418943939e-05, "loss": 0.0378, "step": 24840 }, { "grad_norm": 0.281444251537323, "learning_rate": 8.960988581589865e-05, "loss": 0.0347, "step": 24850 }, { "grad_norm": 0.18614210188388824, "learning_rate": 8.959979311068037e-05, "loss": 0.032, "step": 24860 }, { "grad_norm": 0.288984090089798, "learning_rate": 8.958969607488823e-05, "loss": 0.0363, "step": 24870 }, { "grad_norm": 0.32773905992507935, "learning_rate": 8.957959470962647e-05, "loss": 0.0388, "step": 24880 }, { "grad_norm": 0.259502649307251, "learning_rate": 8.956948901599971e-05, "loss": 0.0389, "step": 24890 }, { "grad_norm": 0.38800764083862305, "learning_rate": 8.955937899511315e-05, "loss": 0.05, "step": 24900 }, { "grad_norm": 0.28303930163383484, "learning_rate": 8.954926464807238e-05, "loss": 0.0347, "step": 24910 }, { "grad_norm": 0.2372409999370575, "learning_rate": 8.953914597598347e-05, "loss": 0.0325, "step": 24920 }, { "grad_norm": 0.2899225950241089, "learning_rate": 8.952902297995303e-05, "loss": 0.0333, "step": 24930 }, { "grad_norm": 0.24366125464439392, "learning_rate": 8.951889566108804e-05, "loss": 0.033, "step": 24940 }, { "grad_norm": 0.3659985065460205, "learning_rate": 8.950876402049606e-05, "loss": 0.0362, "step": 24950 }, { "grad_norm": 0.31771641969680786, "learning_rate": 8.949862805928504e-05, "loss": 0.0312, "step": 24960 }, { "grad_norm": 0.21951079368591309, "learning_rate": 8.948848777856343e-05, "loss": 0.0279, "step": 24970 }, { "grad_norm": 0.2325342893600464, "learning_rate": 8.947834317944017e-05, "loss": 0.0341, "step": 24980 }, { "grad_norm": 0.4370131492614746, "learning_rate": 8.946819426302466e-05, "loss": 0.0324, "step": 24990 }, { "grad_norm": 0.30843019485473633, "learning_rate": 8.945804103042676e-05, "loss": 0.0357, "step": 25000 }, { "grad_norm": 0.16247452795505524, "learning_rate": 8.944788348275681e-05, "loss": 0.0273, "step": 25010 }, { "grad_norm": 0.23167523741722107, "learning_rate": 8.943772162112565e-05, "loss": 0.0297, "step": 25020 }, { "grad_norm": 0.2098991572856903, "learning_rate": 8.942755544664454e-05, "loss": 0.0311, "step": 25030 }, { "grad_norm": 0.2272920310497284, "learning_rate": 8.941738496042525e-05, "loss": 0.0358, "step": 25040 }, { "grad_norm": 0.21476659178733826, "learning_rate": 8.940721016357999e-05, "loss": 0.0283, "step": 25050 }, { "grad_norm": 0.24612343311309814, "learning_rate": 8.939703105722148e-05, "loss": 0.0398, "step": 25060 }, { "grad_norm": 0.2532403767108917, "learning_rate": 8.93868476424629e-05, "loss": 0.0324, "step": 25070 }, { "grad_norm": 0.1918575018644333, "learning_rate": 8.937665992041786e-05, "loss": 0.032, "step": 25080 }, { "grad_norm": 0.21792742609977722, "learning_rate": 8.93664678922005e-05, "loss": 0.0303, "step": 25090 }, { "grad_norm": 0.3033759593963623, "learning_rate": 8.93562715589254e-05, "loss": 0.0348, "step": 25100 }, { "grad_norm": 0.2555280923843384, "learning_rate": 8.934607092170762e-05, "loss": 0.0326, "step": 25110 }, { "grad_norm": 0.6826279163360596, "learning_rate": 8.933586598166266e-05, "loss": 0.03, "step": 25120 }, { "grad_norm": 0.24072185158729553, "learning_rate": 8.932565673990655e-05, "loss": 0.033, "step": 25130 }, { "grad_norm": 0.19917719066143036, "learning_rate": 8.931544319755574e-05, "loss": 0.0316, "step": 25140 }, { "grad_norm": 0.24364519119262695, "learning_rate": 8.930522535572718e-05, "loss": 0.0318, "step": 25150 }, { "grad_norm": 0.24565736949443817, "learning_rate": 8.929500321553826e-05, "loss": 0.0311, "step": 25160 }, { "grad_norm": 0.26578468084335327, "learning_rate": 8.928477677810686e-05, "loss": 0.0274, "step": 25170 }, { "grad_norm": 0.21420951187610626, "learning_rate": 8.927454604455137e-05, "loss": 0.0337, "step": 25180 }, { "grad_norm": 0.23409266769886017, "learning_rate": 8.926431101599053e-05, "loss": 0.0279, "step": 25190 }, { "grad_norm": 0.20978368818759918, "learning_rate": 8.925407169354369e-05, "loss": 0.0357, "step": 25200 }, { "grad_norm": 0.1929238885641098, "learning_rate": 8.92438280783306e-05, "loss": 0.0254, "step": 25210 }, { "grad_norm": 0.33420413732528687, "learning_rate": 8.923358017147146e-05, "loss": 0.0339, "step": 25220 }, { "grad_norm": 0.20875537395477295, "learning_rate": 8.922332797408697e-05, "loss": 0.0335, "step": 25230 }, { "grad_norm": 0.3054353892803192, "learning_rate": 8.921307148729831e-05, "loss": 0.0373, "step": 25240 }, { "grad_norm": 0.28762826323509216, "learning_rate": 8.920281071222712e-05, "loss": 0.035, "step": 25250 }, { "grad_norm": 0.2251012921333313, "learning_rate": 8.919254564999548e-05, "loss": 0.0322, "step": 25260 }, { "grad_norm": 0.2501591145992279, "learning_rate": 8.918227630172598e-05, "loss": 0.0346, "step": 25270 }, { "grad_norm": 0.3616265654563904, "learning_rate": 8.917200266854165e-05, "loss": 0.0409, "step": 25280 }, { "grad_norm": 0.3137401044368744, "learning_rate": 8.9161724751566e-05, "loss": 0.0337, "step": 25290 }, { "grad_norm": 0.19688448309898376, "learning_rate": 8.915144255192302e-05, "loss": 0.0296, "step": 25300 }, { "grad_norm": 0.2428036481142044, "learning_rate": 8.914115607073714e-05, "loss": 0.0285, "step": 25310 }, { "grad_norm": 0.3219064474105835, "learning_rate": 8.913086530913327e-05, "loss": 0.0265, "step": 25320 }, { "grad_norm": 0.28250256180763245, "learning_rate": 8.912057026823681e-05, "loss": 0.0272, "step": 25330 }, { "grad_norm": 0.2682245373725891, "learning_rate": 8.91102709491736e-05, "loss": 0.0325, "step": 25340 }, { "grad_norm": 0.30779793858528137, "learning_rate": 8.909996735306996e-05, "loss": 0.034, "step": 25350 }, { "grad_norm": 0.315275639295578, "learning_rate": 8.908965948105268e-05, "loss": 0.0335, "step": 25360 }, { "grad_norm": 0.30234047770500183, "learning_rate": 8.907934733424901e-05, "loss": 0.0366, "step": 25370 }, { "grad_norm": 0.29475411772727966, "learning_rate": 8.906903091378666e-05, "loss": 0.0335, "step": 25380 }, { "grad_norm": 0.302876740694046, "learning_rate": 8.905871022079384e-05, "loss": 0.0342, "step": 25390 }, { "grad_norm": 0.47837477922439575, "learning_rate": 8.90483852563992e-05, "loss": 0.0309, "step": 25400 }, { "grad_norm": 0.25497618317604065, "learning_rate": 8.903805602173185e-05, "loss": 0.0294, "step": 25410 }, { "grad_norm": 0.36064913868904114, "learning_rate": 8.902772251792137e-05, "loss": 0.0301, "step": 25420 }, { "grad_norm": 0.2387017160654068, "learning_rate": 8.901738474609786e-05, "loss": 0.0336, "step": 25430 }, { "grad_norm": 0.2798810601234436, "learning_rate": 8.900704270739179e-05, "loss": 0.0288, "step": 25440 }, { "grad_norm": 0.22188973426818848, "learning_rate": 8.89966964029342e-05, "loss": 0.0275, "step": 25450 }, { "grad_norm": 0.7395557761192322, "learning_rate": 8.898634583385652e-05, "loss": 0.0306, "step": 25460 }, { "grad_norm": 0.25370147824287415, "learning_rate": 8.897599100129065e-05, "loss": 0.0299, "step": 25470 }, { "grad_norm": 0.4285461902618408, "learning_rate": 8.896563190636903e-05, "loss": 0.0328, "step": 25480 }, { "grad_norm": 0.26913946866989136, "learning_rate": 8.895526855022448e-05, "loss": 0.0312, "step": 25490 }, { "grad_norm": 0.391262024641037, "learning_rate": 8.894490093399033e-05, "loss": 0.0328, "step": 25500 }, { "grad_norm": 0.474042683839798, "learning_rate": 8.893452905880035e-05, "loss": 0.0323, "step": 25510 }, { "grad_norm": 0.2569560110569, "learning_rate": 8.892415292578883e-05, "loss": 0.03, "step": 25520 }, { "grad_norm": 0.3436797261238098, "learning_rate": 8.891377253609046e-05, "loss": 0.0374, "step": 25530 }, { "grad_norm": 0.26757779717445374, "learning_rate": 8.890338789084043e-05, "loss": 0.0344, "step": 25540 }, { "grad_norm": 0.3264784812927246, "learning_rate": 8.88929989911744e-05, "loss": 0.0296, "step": 25550 }, { "grad_norm": 0.2652585804462433, "learning_rate": 8.888260583822847e-05, "loss": 0.0358, "step": 25560 }, { "grad_norm": 0.22585584223270416, "learning_rate": 8.887220843313921e-05, "loss": 0.0328, "step": 25570 }, { "grad_norm": 0.23679186403751373, "learning_rate": 8.88618067770437e-05, "loss": 0.0345, "step": 25580 }, { "grad_norm": 0.1951114684343338, "learning_rate": 8.885140087107942e-05, "loss": 0.0327, "step": 25590 }, { "grad_norm": 1.0298504829406738, "learning_rate": 8.884099071638436e-05, "loss": 0.0446, "step": 25600 }, { "grad_norm": 0.30616495013237, "learning_rate": 8.883057631409695e-05, "loss": 0.0386, "step": 25610 }, { "grad_norm": 0.3663007915019989, "learning_rate": 8.882015766535608e-05, "loss": 0.0394, "step": 25620 }, { "grad_norm": 0.31297799944877625, "learning_rate": 8.880973477130115e-05, "loss": 0.0393, "step": 25630 }, { "grad_norm": 0.20108665525913239, "learning_rate": 8.879930763307197e-05, "loss": 0.0338, "step": 25640 }, { "grad_norm": 0.2752569913864136, "learning_rate": 8.878887625180884e-05, "loss": 0.0278, "step": 25650 }, { "grad_norm": 0.31561633944511414, "learning_rate": 8.877844062865253e-05, "loss": 0.0417, "step": 25660 }, { "grad_norm": 0.3792337477207184, "learning_rate": 8.876800076474424e-05, "loss": 0.0365, "step": 25670 }, { "grad_norm": 0.32698720693588257, "learning_rate": 8.875755666122568e-05, "loss": 0.0368, "step": 25680 }, { "grad_norm": 0.23906879127025604, "learning_rate": 8.8747108319239e-05, "loss": 0.0312, "step": 25690 }, { "grad_norm": 0.2024315744638443, "learning_rate": 8.87366557399268e-05, "loss": 0.0363, "step": 25700 }, { "grad_norm": 0.21090391278266907, "learning_rate": 8.872619892443217e-05, "loss": 0.0342, "step": 25710 }, { "grad_norm": 0.23791901767253876, "learning_rate": 8.871573787389865e-05, "loss": 0.032, "step": 25720 }, { "grad_norm": 0.2804446220397949, "learning_rate": 8.870527258947024e-05, "loss": 0.0317, "step": 25730 }, { "grad_norm": 0.2968803346157074, "learning_rate": 8.869480307229143e-05, "loss": 0.0323, "step": 25740 }, { "grad_norm": 0.21573799848556519, "learning_rate": 8.868432932350712e-05, "loss": 0.0341, "step": 25750 }, { "grad_norm": 1.3995106220245361, "learning_rate": 8.867385134426272e-05, "loss": 0.0303, "step": 25760 }, { "grad_norm": 0.28417259454727173, "learning_rate": 8.866336913570407e-05, "loss": 0.034, "step": 25770 }, { "grad_norm": 0.22620761394500732, "learning_rate": 8.865288269897751e-05, "loss": 0.0326, "step": 25780 }, { "grad_norm": 0.22193966805934906, "learning_rate": 8.864239203522981e-05, "loss": 0.0329, "step": 25790 }, { "grad_norm": 0.3550494313240051, "learning_rate": 8.863189714560822e-05, "loss": 0.0373, "step": 25800 }, { "grad_norm": 0.3227013051509857, "learning_rate": 8.862139803126043e-05, "loss": 0.0301, "step": 25810 }, { "grad_norm": 0.25397834181785583, "learning_rate": 8.861089469333463e-05, "loss": 0.0289, "step": 25820 }, { "grad_norm": 0.2002033293247223, "learning_rate": 8.860038713297944e-05, "loss": 0.0296, "step": 25830 }, { "grad_norm": 0.23379072546958923, "learning_rate": 8.858987535134394e-05, "loss": 0.0293, "step": 25840 }, { "grad_norm": 0.2513047754764557, "learning_rate": 8.857935934957769e-05, "loss": 0.0287, "step": 25850 }, { "grad_norm": 0.1939777284860611, "learning_rate": 8.856883912883071e-05, "loss": 0.0317, "step": 25860 }, { "grad_norm": 0.22609655559062958, "learning_rate": 8.855831469025346e-05, "loss": 0.0269, "step": 25870 }, { "grad_norm": 0.2491052746772766, "learning_rate": 8.854778603499689e-05, "loss": 0.0325, "step": 25880 }, { "grad_norm": 0.18953697383403778, "learning_rate": 8.85372531642124e-05, "loss": 0.0283, "step": 25890 }, { "grad_norm": 0.22748316824436188, "learning_rate": 8.852671607905185e-05, "loss": 0.0273, "step": 25900 }, { "grad_norm": 0.2510952353477478, "learning_rate": 8.851617478066754e-05, "loss": 0.0285, "step": 25910 }, { "grad_norm": 0.3463064730167389, "learning_rate": 8.850562927021227e-05, "loss": 0.0361, "step": 25920 }, { "grad_norm": 0.2103700041770935, "learning_rate": 8.849507954883928e-05, "loss": 0.0288, "step": 25930 }, { "grad_norm": 0.396975576877594, "learning_rate": 8.848452561770226e-05, "loss": 0.028, "step": 25940 }, { "grad_norm": 0.27164486050605774, "learning_rate": 8.847396747795538e-05, "loss": 0.0305, "step": 25950 }, { "grad_norm": 0.23918740451335907, "learning_rate": 8.846340513075327e-05, "loss": 0.0309, "step": 25960 }, { "grad_norm": 0.2338365912437439, "learning_rate": 8.845283857725099e-05, "loss": 0.0391, "step": 25970 }, { "grad_norm": 0.3416938781738281, "learning_rate": 8.844226781860409e-05, "loss": 0.0344, "step": 25980 }, { "grad_norm": 0.2524322271347046, "learning_rate": 8.84316928559686e-05, "loss": 0.0271, "step": 25990 }, { "grad_norm": 0.3303412199020386, "learning_rate": 8.842111369050094e-05, "loss": 0.031, "step": 26000 }, { "grad_norm": 0.3589782416820526, "learning_rate": 8.841053032335808e-05, "loss": 0.0306, "step": 26010 }, { "grad_norm": 0.24941441416740417, "learning_rate": 8.839994275569735e-05, "loss": 0.0321, "step": 26020 }, { "grad_norm": 0.22470438480377197, "learning_rate": 8.838935098867662e-05, "loss": 0.0278, "step": 26030 }, { "grad_norm": 0.4144664406776428, "learning_rate": 8.837875502345418e-05, "loss": 0.032, "step": 26040 }, { "grad_norm": 0.19069159030914307, "learning_rate": 8.83681548611888e-05, "loss": 0.0311, "step": 26050 }, { "grad_norm": 0.3217277228832245, "learning_rate": 8.835755050303969e-05, "loss": 0.0326, "step": 26060 }, { "grad_norm": 0.28164544701576233, "learning_rate": 8.834694195016653e-05, "loss": 0.0346, "step": 26070 }, { "grad_norm": 0.30182933807373047, "learning_rate": 8.833632920372942e-05, "loss": 0.0378, "step": 26080 }, { "grad_norm": 0.22129704058170319, "learning_rate": 8.832571226488903e-05, "loss": 0.0322, "step": 26090 }, { "grad_norm": 0.2388865053653717, "learning_rate": 8.831509113480634e-05, "loss": 0.0344, "step": 26100 }, { "grad_norm": 0.32306531071662903, "learning_rate": 8.83044658146429e-05, "loss": 0.0347, "step": 26110 }, { "grad_norm": 0.27553293108940125, "learning_rate": 8.829383630556067e-05, "loss": 0.0279, "step": 26120 }, { "grad_norm": 0.4443410038948059, "learning_rate": 8.828320260872207e-05, "loss": 0.0272, "step": 26130 }, { "grad_norm": 0.2130124866962433, "learning_rate": 8.827256472529e-05, "loss": 0.0304, "step": 26140 }, { "grad_norm": 0.4111209511756897, "learning_rate": 8.826192265642778e-05, "loss": 0.0309, "step": 26150 }, { "grad_norm": 0.27593785524368286, "learning_rate": 8.825127640329923e-05, "loss": 0.0315, "step": 26160 }, { "grad_norm": 0.35318273305892944, "learning_rate": 8.824062596706861e-05, "loss": 0.0298, "step": 26170 }, { "grad_norm": 0.36567822098731995, "learning_rate": 8.822997134890062e-05, "loss": 0.0294, "step": 26180 }, { "grad_norm": 0.2287197709083557, "learning_rate": 8.821931254996044e-05, "loss": 0.0266, "step": 26190 }, { "grad_norm": 0.27669280767440796, "learning_rate": 8.82086495714137e-05, "loss": 0.029, "step": 26200 }, { "grad_norm": 0.29687634110450745, "learning_rate": 8.81979824144265e-05, "loss": 0.0346, "step": 26210 }, { "grad_norm": 0.22374294698238373, "learning_rate": 8.818731108016536e-05, "loss": 0.0321, "step": 26220 }, { "grad_norm": 0.24245953559875488, "learning_rate": 8.81766355697973e-05, "loss": 0.0321, "step": 26230 }, { "grad_norm": 0.25977590680122375, "learning_rate": 8.816595588448977e-05, "loss": 0.031, "step": 26240 }, { "grad_norm": 0.2564906179904938, "learning_rate": 8.81552720254107e-05, "loss": 0.0323, "step": 26250 }, { "grad_norm": 0.21373002231121063, "learning_rate": 8.814458399372842e-05, "loss": 0.0317, "step": 26260 }, { "grad_norm": 0.5152993202209473, "learning_rate": 8.813389179061181e-05, "loss": 0.0337, "step": 26270 }, { "grad_norm": 0.2932848632335663, "learning_rate": 8.812319541723012e-05, "loss": 0.0314, "step": 26280 }, { "grad_norm": 0.30883288383483887, "learning_rate": 8.811249487475309e-05, "loss": 0.0324, "step": 26290 }, { "grad_norm": 0.3160147964954376, "learning_rate": 8.810179016435092e-05, "loss": 0.0364, "step": 26300 }, { "grad_norm": 0.3578127324581146, "learning_rate": 8.809108128719428e-05, "loss": 0.0356, "step": 26310 }, { "grad_norm": 0.23893412947654724, "learning_rate": 8.808036824445424e-05, "loss": 0.0321, "step": 26320 }, { "grad_norm": 0.2757492959499359, "learning_rate": 8.806965103730238e-05, "loss": 0.0351, "step": 26330 }, { "grad_norm": 1.173396348953247, "learning_rate": 8.805892966691074e-05, "loss": 0.0371, "step": 26340 }, { "grad_norm": 0.24887384474277496, "learning_rate": 8.804820413445175e-05, "loss": 0.0379, "step": 26350 }, { "grad_norm": 0.2753428816795349, "learning_rate": 8.803747444109837e-05, "loss": 0.0327, "step": 26360 }, { "grad_norm": 0.28234782814979553, "learning_rate": 8.802674058802399e-05, "loss": 0.0407, "step": 26370 }, { "grad_norm": 0.26768484711647034, "learning_rate": 8.801600257640241e-05, "loss": 0.0282, "step": 26380 }, { "grad_norm": 0.7178577780723572, "learning_rate": 8.800526040740795e-05, "loss": 0.0347, "step": 26390 }, { "grad_norm": 0.2825499176979065, "learning_rate": 8.799451408221535e-05, "loss": 0.0319, "step": 26400 }, { "grad_norm": 0.26922866702079773, "learning_rate": 8.798376360199982e-05, "loss": 0.04, "step": 26410 }, { "grad_norm": 0.319354385137558, "learning_rate": 8.797300896793701e-05, "loss": 0.0324, "step": 26420 }, { "grad_norm": 0.27512580156326294, "learning_rate": 8.796225018120302e-05, "loss": 0.0334, "step": 26430 }, { "grad_norm": 0.26400408148765564, "learning_rate": 8.795148724297444e-05, "loss": 0.037, "step": 26440 }, { "grad_norm": 0.28960782289505005, "learning_rate": 8.794072015442825e-05, "loss": 0.0459, "step": 26450 }, { "grad_norm": 0.2641761004924774, "learning_rate": 8.792994891674198e-05, "loss": 0.0394, "step": 26460 }, { "grad_norm": 0.8122524619102478, "learning_rate": 8.79191735310935e-05, "loss": 0.0415, "step": 26470 }, { "grad_norm": 0.2753002643585205, "learning_rate": 8.790839399866122e-05, "loss": 0.0366, "step": 26480 }, { "grad_norm": 0.3567408621311188, "learning_rate": 8.789761032062397e-05, "loss": 0.0349, "step": 26490 }, { "grad_norm": 0.44699177145957947, "learning_rate": 8.788682249816103e-05, "loss": 0.0353, "step": 26500 }, { "grad_norm": 0.28173068165779114, "learning_rate": 8.787603053245215e-05, "loss": 0.0298, "step": 26510 }, { "grad_norm": 0.18520747125148773, "learning_rate": 8.78652344246775e-05, "loss": 0.0305, "step": 26520 }, { "grad_norm": 0.22877447307109833, "learning_rate": 8.785443417601776e-05, "loss": 0.0318, "step": 26530 }, { "grad_norm": 0.29581859707832336, "learning_rate": 8.784362978765401e-05, "loss": 0.0313, "step": 26540 }, { "grad_norm": 0.26703429222106934, "learning_rate": 8.783282126076779e-05, "loss": 0.0329, "step": 26550 }, { "grad_norm": 0.25798097252845764, "learning_rate": 8.782200859654112e-05, "loss": 0.027, "step": 26560 }, { "grad_norm": 0.21269404888153076, "learning_rate": 8.781119179615646e-05, "loss": 0.023, "step": 26570 }, { "grad_norm": 0.21728812158107758, "learning_rate": 8.780037086079674e-05, "loss": 0.0252, "step": 26580 }, { "grad_norm": 0.18755313754081726, "learning_rate": 8.778954579164527e-05, "loss": 0.0281, "step": 26590 }, { "grad_norm": 0.37723231315612793, "learning_rate": 8.777871658988588e-05, "loss": 0.0278, "step": 26600 }, { "grad_norm": 0.23833028972148895, "learning_rate": 8.776788325670285e-05, "loss": 0.0297, "step": 26610 }, { "grad_norm": 1.2990541458129883, "learning_rate": 8.775704579328089e-05, "loss": 0.0269, "step": 26620 }, { "grad_norm": 0.2844148576259613, "learning_rate": 8.774620420080517e-05, "loss": 0.0326, "step": 26630 }, { "grad_norm": 0.3229585587978363, "learning_rate": 8.773535848046131e-05, "loss": 0.0327, "step": 26640 }, { "grad_norm": 0.24408049881458282, "learning_rate": 8.772450863343538e-05, "loss": 0.0316, "step": 26650 }, { "grad_norm": 1.1503833532333374, "learning_rate": 8.77136546609139e-05, "loss": 0.0343, "step": 26660 }, { "grad_norm": 0.4418993294239044, "learning_rate": 8.770279656408385e-05, "loss": 0.0331, "step": 26670 }, { "grad_norm": 0.4226922392845154, "learning_rate": 8.769193434413265e-05, "loss": 0.0335, "step": 26680 }, { "grad_norm": 0.30659106373786926, "learning_rate": 8.76810680022482e-05, "loss": 0.0331, "step": 26690 }, { "grad_norm": 0.9070488810539246, "learning_rate": 8.767019753961878e-05, "loss": 0.0395, "step": 26700 }, { "grad_norm": 0.5963837504386902, "learning_rate": 8.765932295743321e-05, "loss": 0.0399, "step": 26710 }, { "grad_norm": 0.32991698384284973, "learning_rate": 8.764844425688068e-05, "loss": 0.0455, "step": 26720 }, { "grad_norm": 0.37196066975593567, "learning_rate": 8.763756143915092e-05, "loss": 0.0386, "step": 26730 }, { "grad_norm": 0.265656054019928, "learning_rate": 8.7626674505434e-05, "loss": 0.0346, "step": 26740 }, { "grad_norm": 0.27636462450027466, "learning_rate": 8.761578345692053e-05, "loss": 0.0327, "step": 26750 }, { "grad_norm": 0.21693924069404602, "learning_rate": 8.760488829480156e-05, "loss": 0.0306, "step": 26760 }, { "grad_norm": 0.21702195703983307, "learning_rate": 8.759398902026854e-05, "loss": 0.0323, "step": 26770 }, { "grad_norm": 0.23680971562862396, "learning_rate": 8.758308563451339e-05, "loss": 0.0259, "step": 26780 }, { "grad_norm": 0.2416008859872818, "learning_rate": 8.75721781387285e-05, "loss": 0.0334, "step": 26790 }, { "grad_norm": 0.3331574499607086, "learning_rate": 8.75612665341067e-05, "loss": 0.033, "step": 26800 }, { "grad_norm": 0.19011814892292023, "learning_rate": 8.755035082184126e-05, "loss": 0.0337, "step": 26810 }, { "grad_norm": 0.6472266316413879, "learning_rate": 8.753943100312592e-05, "loss": 0.0302, "step": 26820 }, { "grad_norm": 0.260739803314209, "learning_rate": 8.752850707915484e-05, "loss": 0.0291, "step": 26830 }, { "grad_norm": 0.2839896082878113, "learning_rate": 8.751757905112264e-05, "loss": 0.0315, "step": 26840 }, { "grad_norm": 0.18288910388946533, "learning_rate": 8.75066469202244e-05, "loss": 0.0335, "step": 26850 }, { "grad_norm": 0.2715936303138733, "learning_rate": 8.749571068765567e-05, "loss": 0.0324, "step": 26860 }, { "grad_norm": 0.2660522162914276, "learning_rate": 8.748477035461238e-05, "loss": 0.035, "step": 26870 }, { "grad_norm": 0.2842582166194916, "learning_rate": 8.747382592229095e-05, "loss": 0.0377, "step": 26880 }, { "grad_norm": 0.3401779532432556, "learning_rate": 8.746287739188828e-05, "loss": 0.0278, "step": 26890 }, { "grad_norm": 0.37002378702163696, "learning_rate": 8.745192476460165e-05, "loss": 0.0353, "step": 26900 }, { "grad_norm": 0.5137906074523926, "learning_rate": 8.744096804162882e-05, "loss": 0.0344, "step": 26910 }, { "grad_norm": 0.2622338831424713, "learning_rate": 8.743000722416804e-05, "loss": 0.0306, "step": 26920 }, { "grad_norm": 0.3092069625854492, "learning_rate": 8.741904231341793e-05, "loss": 0.0377, "step": 26930 }, { "grad_norm": 0.26295873522758484, "learning_rate": 8.740807331057762e-05, "loss": 0.0336, "step": 26940 }, { "grad_norm": 0.552797257900238, "learning_rate": 8.739710021684667e-05, "loss": 0.0303, "step": 26950 }, { "grad_norm": 0.3002350330352783, "learning_rate": 8.738612303342503e-05, "loss": 0.034, "step": 26960 }, { "grad_norm": 0.22658400237560272, "learning_rate": 8.73751417615132e-05, "loss": 0.0317, "step": 26970 }, { "grad_norm": 0.23565226793289185, "learning_rate": 8.736415640231208e-05, "loss": 0.0308, "step": 26980 }, { "grad_norm": 0.19303563237190247, "learning_rate": 8.735316695702297e-05, "loss": 0.036, "step": 26990 }, { "grad_norm": 0.3028320074081421, "learning_rate": 8.734217342684769e-05, "loss": 0.0287, "step": 27000 }, { "grad_norm": 0.295695960521698, "learning_rate": 8.733117581298847e-05, "loss": 0.0329, "step": 27010 }, { "grad_norm": 0.3063884973526001, "learning_rate": 8.732017411664796e-05, "loss": 0.0269, "step": 27020 }, { "grad_norm": 0.28823617100715637, "learning_rate": 8.730916833902936e-05, "loss": 0.026, "step": 27030 }, { "grad_norm": 0.2460574060678482, "learning_rate": 8.729815848133618e-05, "loss": 0.0272, "step": 27040 }, { "grad_norm": 0.33692073822021484, "learning_rate": 8.728714454477247e-05, "loss": 0.0229, "step": 27050 }, { "grad_norm": 0.2667436897754669, "learning_rate": 8.727612653054269e-05, "loss": 0.0271, "step": 27060 }, { "grad_norm": 0.2755783200263977, "learning_rate": 8.726510443985176e-05, "loss": 0.0278, "step": 27070 }, { "grad_norm": 0.28894513845443726, "learning_rate": 8.725407827390503e-05, "loss": 0.0316, "step": 27080 }, { "grad_norm": 0.27952808141708374, "learning_rate": 8.724304803390833e-05, "loss": 0.0315, "step": 27090 }, { "grad_norm": 0.2412245124578476, "learning_rate": 8.723201372106788e-05, "loss": 0.0311, "step": 27100 }, { "grad_norm": 0.24384598433971405, "learning_rate": 8.722097533659038e-05, "loss": 0.0315, "step": 27110 }, { "grad_norm": 0.38258615136146545, "learning_rate": 8.720993288168299e-05, "loss": 0.0337, "step": 27120 }, { "grad_norm": 0.2634556293487549, "learning_rate": 8.719888635755327e-05, "loss": 0.0259, "step": 27130 }, { "grad_norm": 0.37362605333328247, "learning_rate": 8.718783576540928e-05, "loss": 0.033, "step": 27140 }, { "grad_norm": 0.2527543604373932, "learning_rate": 8.717678110645948e-05, "loss": 0.0286, "step": 27150 }, { "grad_norm": 0.1823730319738388, "learning_rate": 8.716572238191279e-05, "loss": 0.0271, "step": 27160 }, { "grad_norm": 0.2231152206659317, "learning_rate": 8.715465959297857e-05, "loss": 0.0293, "step": 27170 }, { "grad_norm": 0.2777683436870575, "learning_rate": 8.714359274086665e-05, "loss": 0.0353, "step": 27180 }, { "grad_norm": 0.28299739956855774, "learning_rate": 8.713252182678726e-05, "loss": 0.0341, "step": 27190 }, { "grad_norm": 0.2350223809480667, "learning_rate": 8.712144685195112e-05, "loss": 0.0287, "step": 27200 }, { "grad_norm": 0.2989051342010498, "learning_rate": 8.711036781756936e-05, "loss": 0.0304, "step": 27210 }, { "grad_norm": 0.22854968905448914, "learning_rate": 8.709928472485357e-05, "loss": 0.0318, "step": 27220 }, { "grad_norm": 0.2020060420036316, "learning_rate": 8.708819757501579e-05, "loss": 0.0293, "step": 27230 }, { "grad_norm": 0.3125521242618561, "learning_rate": 8.707710636926846e-05, "loss": 0.0311, "step": 27240 }, { "grad_norm": 0.5104705691337585, "learning_rate": 8.706601110882455e-05, "loss": 0.0325, "step": 27250 }, { "grad_norm": 0.3042803108692169, "learning_rate": 8.705491179489738e-05, "loss": 0.0298, "step": 27260 }, { "grad_norm": 0.3975641429424286, "learning_rate": 8.704380842870077e-05, "loss": 0.0351, "step": 27270 }, { "grad_norm": 0.2895070016384125, "learning_rate": 8.703270101144895e-05, "loss": 0.0334, "step": 27280 }, { "grad_norm": 0.29037603735923767, "learning_rate": 8.702158954435664e-05, "loss": 0.035, "step": 27290 }, { "grad_norm": 0.24437327682971954, "learning_rate": 8.701047402863896e-05, "loss": 0.0302, "step": 27300 }, { "grad_norm": 0.3279654085636139, "learning_rate": 8.699935446551148e-05, "loss": 0.0376, "step": 27310 }, { "grad_norm": 0.26681357622146606, "learning_rate": 8.698823085619022e-05, "loss": 0.0283, "step": 27320 }, { "grad_norm": 0.21811378002166748, "learning_rate": 8.697710320189166e-05, "loss": 0.0256, "step": 27330 }, { "grad_norm": 0.4128751754760742, "learning_rate": 8.696597150383268e-05, "loss": 0.0337, "step": 27340 }, { "grad_norm": 0.36059749126434326, "learning_rate": 8.695483576323063e-05, "loss": 0.0269, "step": 27350 }, { "grad_norm": 0.34253618121147156, "learning_rate": 8.69436959813033e-05, "loss": 0.0295, "step": 27360 }, { "grad_norm": 0.2909358739852905, "learning_rate": 8.693255215926892e-05, "loss": 0.0274, "step": 27370 }, { "grad_norm": 0.18981388211250305, "learning_rate": 8.692140429834617e-05, "loss": 0.0354, "step": 27380 }, { "grad_norm": 0.4305386543273926, "learning_rate": 8.691025239975415e-05, "loss": 0.0311, "step": 27390 }, { "grad_norm": 0.5452466607093811, "learning_rate": 8.689909646471243e-05, "loss": 0.0329, "step": 27400 }, { "grad_norm": 0.3425048589706421, "learning_rate": 8.688793649444099e-05, "loss": 0.0359, "step": 27410 }, { "grad_norm": 0.23766908049583435, "learning_rate": 8.687677249016029e-05, "loss": 0.0297, "step": 27420 }, { "grad_norm": 0.2710918188095093, "learning_rate": 8.686560445309118e-05, "loss": 0.033, "step": 27430 }, { "grad_norm": 0.33186396956443787, "learning_rate": 8.685443238445499e-05, "loss": 0.0324, "step": 27440 }, { "grad_norm": 0.34730494022369385, "learning_rate": 8.68432562854735e-05, "loss": 0.034, "step": 27450 }, { "grad_norm": 0.24351908266544342, "learning_rate": 8.683207615736887e-05, "loss": 0.0295, "step": 27460 }, { "grad_norm": 0.32661721110343933, "learning_rate": 8.682089200136379e-05, "loss": 0.0401, "step": 27470 }, { "grad_norm": 0.29456934332847595, "learning_rate": 8.680970381868132e-05, "loss": 0.0358, "step": 27480 }, { "grad_norm": 0.2703889310359955, "learning_rate": 8.679851161054498e-05, "loss": 0.0375, "step": 27490 }, { "grad_norm": 0.28222766518592834, "learning_rate": 8.678731537817873e-05, "loss": 0.0318, "step": 27500 }, { "grad_norm": 0.4135729670524597, "learning_rate": 8.677611512280697e-05, "loss": 0.0335, "step": 27510 }, { "grad_norm": 0.20568449795246124, "learning_rate": 8.676491084565457e-05, "loss": 0.0243, "step": 27520 }, { "grad_norm": 0.26784199476242065, "learning_rate": 8.675370254794678e-05, "loss": 0.0286, "step": 27530 }, { "grad_norm": 0.26401519775390625, "learning_rate": 8.674249023090935e-05, "loss": 0.0329, "step": 27540 }, { "grad_norm": 0.21858565509319305, "learning_rate": 8.673127389576843e-05, "loss": 0.0391, "step": 27550 }, { "grad_norm": 0.28420525789260864, "learning_rate": 8.67200535437506e-05, "loss": 0.0299, "step": 27560 }, { "grad_norm": 0.30193912982940674, "learning_rate": 8.670882917608296e-05, "loss": 0.0293, "step": 27570 }, { "grad_norm": 0.28901106119155884, "learning_rate": 8.669760079399292e-05, "loss": 0.0345, "step": 27580 }, { "grad_norm": 0.28131595253944397, "learning_rate": 8.668636839870845e-05, "loss": 0.0276, "step": 27590 }, { "grad_norm": 0.30362990498542786, "learning_rate": 8.667513199145789e-05, "loss": 0.0322, "step": 27600 }, { "grad_norm": 0.4676235616207123, "learning_rate": 8.666389157347002e-05, "loss": 0.0299, "step": 27610 }, { "grad_norm": 0.2503485381603241, "learning_rate": 8.66526471459741e-05, "loss": 0.0275, "step": 27620 }, { "grad_norm": 0.234984889626503, "learning_rate": 8.66413987101998e-05, "loss": 0.028, "step": 27630 }, { "grad_norm": 0.18450120091438293, "learning_rate": 8.663014626737723e-05, "loss": 0.0302, "step": 27640 }, { "grad_norm": 0.20268501341342926, "learning_rate": 8.661888981873691e-05, "loss": 0.0283, "step": 27650 }, { "grad_norm": 0.3276572525501251, "learning_rate": 8.660762936550988e-05, "loss": 0.0315, "step": 27660 }, { "grad_norm": 0.3321131765842438, "learning_rate": 8.659636490892753e-05, "loss": 0.0297, "step": 27670 }, { "grad_norm": 0.22256524860858917, "learning_rate": 8.658509645022174e-05, "loss": 0.0281, "step": 27680 }, { "grad_norm": 0.3375886380672455, "learning_rate": 8.657382399062481e-05, "loss": 0.0277, "step": 27690 }, { "grad_norm": 0.2472563534975052, "learning_rate": 8.656254753136946e-05, "loss": 0.0317, "step": 27700 }, { "grad_norm": 0.3404884934425354, "learning_rate": 8.655126707368891e-05, "loss": 0.0362, "step": 27710 }, { "grad_norm": 0.20402121543884277, "learning_rate": 8.653998261881672e-05, "loss": 0.0324, "step": 27720 }, { "grad_norm": 0.26784777641296387, "learning_rate": 8.652869416798699e-05, "loss": 0.0369, "step": 27730 }, { "grad_norm": 0.3172225058078766, "learning_rate": 8.651740172243417e-05, "loss": 0.0276, "step": 27740 }, { "grad_norm": 0.3305200934410095, "learning_rate": 8.65061052833932e-05, "loss": 0.0323, "step": 27750 }, { "grad_norm": 0.2977222800254822, "learning_rate": 8.649480485209945e-05, "loss": 0.0286, "step": 27760 }, { "grad_norm": 0.25140634179115295, "learning_rate": 8.64835004297887e-05, "loss": 0.0264, "step": 27770 }, { "grad_norm": 0.2946300208568573, "learning_rate": 8.64721920176972e-05, "loss": 0.0301, "step": 27780 }, { "grad_norm": 0.24048088490962982, "learning_rate": 8.646087961706164e-05, "loss": 0.0357, "step": 27790 }, { "grad_norm": 0.2391626387834549, "learning_rate": 8.644956322911908e-05, "loss": 0.0334, "step": 27800 }, { "grad_norm": 0.1724044531583786, "learning_rate": 8.643824285510709e-05, "loss": 0.0265, "step": 27810 }, { "grad_norm": 0.17231827974319458, "learning_rate": 8.642691849626364e-05, "loss": 0.0332, "step": 27820 }, { "grad_norm": 0.2715262174606323, "learning_rate": 8.641559015382717e-05, "loss": 0.0333, "step": 27830 }, { "grad_norm": 0.42405521869659424, "learning_rate": 8.640425782903649e-05, "loss": 0.0376, "step": 27840 }, { "grad_norm": 0.24021391570568085, "learning_rate": 8.639292152313091e-05, "loss": 0.0302, "step": 27850 }, { "grad_norm": 0.226128488779068, "learning_rate": 8.638158123735015e-05, "loss": 0.0311, "step": 27860 }, { "grad_norm": 0.26922112703323364, "learning_rate": 8.637023697293436e-05, "loss": 0.0286, "step": 27870 }, { "grad_norm": 0.2506416141986847, "learning_rate": 8.635888873112414e-05, "loss": 0.0309, "step": 27880 }, { "grad_norm": 0.26648080348968506, "learning_rate": 8.634753651316052e-05, "loss": 0.0322, "step": 27890 }, { "grad_norm": 0.2624494731426239, "learning_rate": 8.633618032028496e-05, "loss": 0.0339, "step": 27900 }, { "grad_norm": 0.2127520740032196, "learning_rate": 8.632482015373934e-05, "loss": 0.0306, "step": 27910 }, { "grad_norm": 0.23903587460517883, "learning_rate": 8.6313456014766e-05, "loss": 0.03, "step": 27920 }, { "grad_norm": 0.219352588057518, "learning_rate": 8.630208790460771e-05, "loss": 0.027, "step": 27930 }, { "grad_norm": 0.2579032778739929, "learning_rate": 8.629071582450768e-05, "loss": 0.0334, "step": 27940 }, { "grad_norm": 0.38904085755348206, "learning_rate": 8.62793397757095e-05, "loss": 0.0325, "step": 27950 }, { "grad_norm": 0.2682652175426483, "learning_rate": 8.626795975945729e-05, "loss": 0.0338, "step": 27960 }, { "grad_norm": 0.2876797020435333, "learning_rate": 8.625657577699551e-05, "loss": 0.0353, "step": 27970 }, { "grad_norm": 0.2270570546388626, "learning_rate": 8.624518782956914e-05, "loss": 0.0326, "step": 27980 }, { "grad_norm": 0.26014959812164307, "learning_rate": 8.62337959184235e-05, "loss": 0.0318, "step": 27990 }, { "grad_norm": 0.24424798786640167, "learning_rate": 8.622240004480441e-05, "loss": 0.0348, "step": 28000 }, { "grad_norm": 0.2476186603307724, "learning_rate": 8.621100020995814e-05, "loss": 0.0252, "step": 28010 }, { "grad_norm": 0.2891216278076172, "learning_rate": 8.619959641513132e-05, "loss": 0.028, "step": 28020 }, { "grad_norm": 0.3989788293838501, "learning_rate": 8.618818866157105e-05, "loss": 0.0287, "step": 28030 }, { "grad_norm": 0.2667640149593353, "learning_rate": 8.617677695052487e-05, "loss": 0.0277, "step": 28040 }, { "grad_norm": 0.24319517612457275, "learning_rate": 8.616536128324078e-05, "loss": 0.0267, "step": 28050 }, { "grad_norm": 0.25927332043647766, "learning_rate": 8.615394166096712e-05, "loss": 0.0273, "step": 28060 }, { "grad_norm": 0.2912425398826599, "learning_rate": 8.614251808495279e-05, "loss": 0.0306, "step": 28070 }, { "grad_norm": 0.22749726474285126, "learning_rate": 8.6131090556447e-05, "loss": 0.0319, "step": 28080 }, { "grad_norm": 0.2704734802246094, "learning_rate": 8.611965907669947e-05, "loss": 0.0277, "step": 28090 }, { "grad_norm": 0.22235870361328125, "learning_rate": 8.610822364696034e-05, "loss": 0.0357, "step": 28100 }, { "grad_norm": 0.2619883418083191, "learning_rate": 8.609678426848015e-05, "loss": 0.0295, "step": 28110 }, { "grad_norm": 0.22961850464344025, "learning_rate": 8.60853409425099e-05, "loss": 0.0254, "step": 28120 }, { "grad_norm": 0.2421739399433136, "learning_rate": 8.607389367030104e-05, "loss": 0.0271, "step": 28130 }, { "grad_norm": 0.2408486306667328, "learning_rate": 8.606244245310538e-05, "loss": 0.0308, "step": 28140 }, { "grad_norm": 0.2437603771686554, "learning_rate": 8.605098729217525e-05, "loss": 0.0288, "step": 28150 }, { "grad_norm": 0.2349606454372406, "learning_rate": 8.603952818876335e-05, "loss": 0.0289, "step": 28160 }, { "grad_norm": 0.36646243929862976, "learning_rate": 8.602806514412281e-05, "loss": 0.0243, "step": 28170 }, { "grad_norm": 0.25755855441093445, "learning_rate": 8.601659815950726e-05, "loss": 0.0295, "step": 28180 }, { "grad_norm": 0.24886579811573029, "learning_rate": 8.600512723617067e-05, "loss": 0.0259, "step": 28190 }, { "grad_norm": 0.18023419380187988, "learning_rate": 8.59936523753675e-05, "loss": 0.0286, "step": 28200 }, { "grad_norm": 0.24810704588890076, "learning_rate": 8.598217357835264e-05, "loss": 0.0339, "step": 28210 }, { "grad_norm": 0.21766650676727295, "learning_rate": 8.597069084638135e-05, "loss": 0.0268, "step": 28220 }, { "grad_norm": 0.5350215435028076, "learning_rate": 8.595920418070939e-05, "loss": 0.0274, "step": 28230 }, { "grad_norm": 0.18351653218269348, "learning_rate": 8.594771358259295e-05, "loss": 0.0269, "step": 28240 }, { "grad_norm": 0.18570847809314728, "learning_rate": 8.593621905328858e-05, "loss": 0.0268, "step": 28250 }, { "grad_norm": 0.2690558433532715, "learning_rate": 8.592472059405333e-05, "loss": 0.029, "step": 28260 }, { "grad_norm": 0.2494821548461914, "learning_rate": 8.591321820614464e-05, "loss": 0.0262, "step": 28270 }, { "grad_norm": 0.2514760196208954, "learning_rate": 8.590171189082041e-05, "loss": 0.0259, "step": 28280 }, { "grad_norm": 0.2195308804512024, "learning_rate": 8.589020164933894e-05, "loss": 0.0275, "step": 28290 }, { "grad_norm": 0.2339349389076233, "learning_rate": 8.587868748295898e-05, "loss": 0.0285, "step": 28300 }, { "grad_norm": 0.2774259150028229, "learning_rate": 8.586716939293971e-05, "loss": 0.0363, "step": 28310 }, { "grad_norm": 0.4354439377784729, "learning_rate": 8.58556473805407e-05, "loss": 0.0296, "step": 28320 }, { "grad_norm": 0.4557100236415863, "learning_rate": 8.584412144702202e-05, "loss": 0.0263, "step": 28330 }, { "grad_norm": 0.20286141335964203, "learning_rate": 8.58325915936441e-05, "loss": 0.033, "step": 28340 }, { "grad_norm": 0.24434639513492584, "learning_rate": 8.582105782166783e-05, "loss": 0.0283, "step": 28350 }, { "grad_norm": 0.28078335523605347, "learning_rate": 8.580952013235455e-05, "loss": 0.0318, "step": 28360 }, { "grad_norm": 0.23490537703037262, "learning_rate": 8.579797852696596e-05, "loss": 0.0288, "step": 28370 }, { "grad_norm": 0.2603215277194977, "learning_rate": 8.578643300676428e-05, "loss": 0.0257, "step": 28380 }, { "grad_norm": 0.21667003631591797, "learning_rate": 8.577488357301209e-05, "loss": 0.0322, "step": 28390 }, { "grad_norm": 0.265855997800827, "learning_rate": 8.576333022697242e-05, "loss": 0.0346, "step": 28400 }, { "grad_norm": 0.27419984340667725, "learning_rate": 8.575177296990873e-05, "loss": 0.0262, "step": 28410 }, { "grad_norm": 0.45341962575912476, "learning_rate": 8.574021180308489e-05, "loss": 0.0285, "step": 28420 }, { "grad_norm": 0.2771890163421631, "learning_rate": 8.572864672776523e-05, "loss": 0.0297, "step": 28430 }, { "grad_norm": 0.263658732175827, "learning_rate": 8.571707774521447e-05, "loss": 0.0292, "step": 28440 }, { "grad_norm": 0.20231914520263672, "learning_rate": 8.57055048566978e-05, "loss": 0.0321, "step": 28450 }, { "grad_norm": 0.1688213348388672, "learning_rate": 8.569392806348078e-05, "loss": 0.0274, "step": 28460 }, { "grad_norm": 0.2502206861972809, "learning_rate": 8.568234736682947e-05, "loss": 0.0272, "step": 28470 }, { "grad_norm": 0.26770421862602234, "learning_rate": 8.567076276801029e-05, "loss": 0.0308, "step": 28480 }, { "grad_norm": 0.3697242736816406, "learning_rate": 8.565917426829013e-05, "loss": 0.0278, "step": 28490 }, { "grad_norm": 0.24901807308197021, "learning_rate": 8.564758186893628e-05, "loss": 0.0285, "step": 28500 }, { "grad_norm": 0.22666887938976288, "learning_rate": 8.563598557121649e-05, "loss": 0.028, "step": 28510 }, { "grad_norm": 0.2768372893333435, "learning_rate": 8.562438537639888e-05, "loss": 0.0278, "step": 28520 }, { "grad_norm": 0.2888885736465454, "learning_rate": 8.561278128575206e-05, "loss": 0.0306, "step": 28530 }, { "grad_norm": 0.3243831396102905, "learning_rate": 8.5601173300545e-05, "loss": 0.0338, "step": 28540 }, { "grad_norm": 0.31043416261672974, "learning_rate": 8.558956142204717e-05, "loss": 0.0319, "step": 28550 }, { "grad_norm": 1.0798259973526, "learning_rate": 8.55779456515284e-05, "loss": 0.0301, "step": 28560 }, { "grad_norm": 0.26038262248039246, "learning_rate": 8.556632599025898e-05, "loss": 0.0385, "step": 28570 }, { "grad_norm": 0.5393798351287842, "learning_rate": 8.555470243950964e-05, "loss": 0.0387, "step": 28580 }, { "grad_norm": 0.3210098147392273, "learning_rate": 8.554307500055148e-05, "loss": 0.0349, "step": 28590 }, { "grad_norm": 0.3936275541782379, "learning_rate": 8.553144367465609e-05, "loss": 0.0335, "step": 28600 }, { "grad_norm": 0.3798428475856781, "learning_rate": 8.551980846309544e-05, "loss": 0.0315, "step": 28610 }, { "grad_norm": 0.3468596935272217, "learning_rate": 8.550816936714193e-05, "loss": 0.0302, "step": 28620 }, { "grad_norm": 0.2900868356227875, "learning_rate": 8.549652638806841e-05, "loss": 0.031, "step": 28630 }, { "grad_norm": 0.18509185314178467, "learning_rate": 8.548487952714812e-05, "loss": 0.0277, "step": 28640 }, { "grad_norm": 0.2337651252746582, "learning_rate": 8.547322878565478e-05, "loss": 0.03, "step": 28650 }, { "grad_norm": 0.2173721343278885, "learning_rate": 8.546157416486245e-05, "loss": 0.0273, "step": 28660 }, { "grad_norm": 0.23170903325080872, "learning_rate": 8.54499156660457e-05, "loss": 0.0288, "step": 28670 }, { "grad_norm": 0.2642601728439331, "learning_rate": 8.543825329047947e-05, "loss": 0.029, "step": 28680 }, { "grad_norm": 0.2433537244796753, "learning_rate": 8.542658703943913e-05, "loss": 0.0281, "step": 28690 }, { "grad_norm": 0.19736668467521667, "learning_rate": 8.541491691420051e-05, "loss": 0.0249, "step": 28700 }, { "grad_norm": 0.15059097111225128, "learning_rate": 8.54032429160398e-05, "loss": 0.0284, "step": 28710 }, { "grad_norm": 0.2621898055076599, "learning_rate": 8.539156504623369e-05, "loss": 0.0343, "step": 28720 }, { "grad_norm": 0.18889479339122772, "learning_rate": 8.537988330605923e-05, "loss": 0.0369, "step": 28730 }, { "grad_norm": 0.7713217735290527, "learning_rate": 8.536819769679393e-05, "loss": 0.0269, "step": 28740 }, { "grad_norm": 0.6222209930419922, "learning_rate": 8.53565082197157e-05, "loss": 0.0329, "step": 28750 }, { "grad_norm": 0.40443745255470276, "learning_rate": 8.534481487610289e-05, "loss": 0.0377, "step": 28760 }, { "grad_norm": 0.2907406687736511, "learning_rate": 8.533311766723428e-05, "loss": 0.0366, "step": 28770 }, { "grad_norm": 0.28864002227783203, "learning_rate": 8.532141659438901e-05, "loss": 0.033, "step": 28780 }, { "grad_norm": 0.2847539782524109, "learning_rate": 8.530971165884675e-05, "loss": 0.0347, "step": 28790 }, { "grad_norm": 0.4009510278701782, "learning_rate": 8.529800286188752e-05, "loss": 0.0299, "step": 28800 }, { "grad_norm": 0.4220883846282959, "learning_rate": 8.528629020479175e-05, "loss": 0.029, "step": 28810 }, { "grad_norm": 0.26497897505760193, "learning_rate": 8.527457368884033e-05, "loss": 0.0327, "step": 28820 }, { "grad_norm": 0.22616924345493317, "learning_rate": 8.526285331531458e-05, "loss": 0.0279, "step": 28830 }, { "grad_norm": 0.8900409936904907, "learning_rate": 8.525112908549621e-05, "loss": 0.0277, "step": 28840 }, { "grad_norm": 1.3875341415405273, "learning_rate": 8.523940100066735e-05, "loss": 0.0327, "step": 28850 }, { "grad_norm": 0.20025388896465302, "learning_rate": 8.52276690621106e-05, "loss": 0.0279, "step": 28860 }, { "grad_norm": 0.3013012409210205, "learning_rate": 8.521593327110889e-05, "loss": 0.0287, "step": 28870 }, { "grad_norm": 0.7549549341201782, "learning_rate": 8.520419362894569e-05, "loss": 0.0336, "step": 28880 }, { "grad_norm": 0.2913420796394348, "learning_rate": 8.51924501369048e-05, "loss": 0.0314, "step": 28890 }, { "grad_norm": 0.2596592307090759, "learning_rate": 8.518070279627047e-05, "loss": 0.0396, "step": 28900 }, { "grad_norm": 0.3365045487880707, "learning_rate": 8.516895160832737e-05, "loss": 0.0447, "step": 28910 }, { "grad_norm": 0.2505170702934265, "learning_rate": 8.515719657436061e-05, "loss": 0.0317, "step": 28920 }, { "grad_norm": 0.24414736032485962, "learning_rate": 8.514543769565568e-05, "loss": 0.0346, "step": 28930 }, { "grad_norm": 0.2630526125431061, "learning_rate": 8.513367497349853e-05, "loss": 0.0296, "step": 28940 }, { "grad_norm": 0.23890678584575653, "learning_rate": 8.51219084091755e-05, "loss": 0.0263, "step": 28950 }, { "grad_norm": 0.24365076422691345, "learning_rate": 8.511013800397338e-05, "loss": 0.0326, "step": 28960 }, { "grad_norm": 0.3123881220817566, "learning_rate": 8.509836375917937e-05, "loss": 0.0289, "step": 28970 }, { "grad_norm": 0.23860326409339905, "learning_rate": 8.508658567608104e-05, "loss": 0.028, "step": 28980 }, { "grad_norm": 0.22359560430049896, "learning_rate": 8.507480375596647e-05, "loss": 0.0258, "step": 28990 }, { "grad_norm": 0.39879995584487915, "learning_rate": 8.506301800012408e-05, "loss": 0.0375, "step": 29000 }, { "grad_norm": 0.21983397006988525, "learning_rate": 8.505122840984278e-05, "loss": 0.0385, "step": 29010 }, { "grad_norm": 0.31088441610336304, "learning_rate": 8.503943498641182e-05, "loss": 0.0299, "step": 29020 }, { "grad_norm": 0.28061896562576294, "learning_rate": 8.502763773112095e-05, "loss": 0.0264, "step": 29030 }, { "grad_norm": 0.1869557499885559, "learning_rate": 8.501583664526026e-05, "loss": 0.0336, "step": 29040 }, { "grad_norm": 0.2565059959888458, "learning_rate": 8.500403173012032e-05, "loss": 0.0332, "step": 29050 }, { "grad_norm": 0.46432241797447205, "learning_rate": 8.499222298699211e-05, "loss": 0.0279, "step": 29060 }, { "grad_norm": 0.22580713033676147, "learning_rate": 8.498041041716701e-05, "loss": 0.0276, "step": 29070 }, { "grad_norm": 0.30544888973236084, "learning_rate": 8.496859402193681e-05, "loss": 0.0328, "step": 29080 }, { "grad_norm": 0.2284504622220993, "learning_rate": 8.495677380259374e-05, "loss": 0.0282, "step": 29090 }, { "grad_norm": 0.29300785064697266, "learning_rate": 8.494494976043045e-05, "loss": 0.0288, "step": 29100 }, { "grad_norm": 0.2629551589488983, "learning_rate": 8.493312189673998e-05, "loss": 0.0278, "step": 29110 }, { "grad_norm": 0.23976685106754303, "learning_rate": 8.492129021281584e-05, "loss": 0.0329, "step": 29120 }, { "grad_norm": 0.25230872631073, "learning_rate": 8.490945470995188e-05, "loss": 0.0349, "step": 29130 }, { "grad_norm": 0.2826167941093445, "learning_rate": 8.489761538944247e-05, "loss": 0.0304, "step": 29140 }, { "grad_norm": 0.37453579902648926, "learning_rate": 8.48857722525823e-05, "loss": 0.03, "step": 29150 }, { "grad_norm": 0.28486886620521545, "learning_rate": 8.487392530066652e-05, "loss": 0.0298, "step": 29160 }, { "grad_norm": 0.28489401936531067, "learning_rate": 8.486207453499069e-05, "loss": 0.0274, "step": 29170 }, { "grad_norm": 0.2918493151664734, "learning_rate": 8.485021995685082e-05, "loss": 0.028, "step": 29180 }, { "grad_norm": 0.23512202501296997, "learning_rate": 8.483836156754328e-05, "loss": 0.0324, "step": 29190 }, { "grad_norm": 0.46982061862945557, "learning_rate": 8.482649936836491e-05, "loss": 0.0277, "step": 29200 }, { "grad_norm": 0.30595168471336365, "learning_rate": 8.481463336061293e-05, "loss": 0.0331, "step": 29210 }, { "grad_norm": 0.21097488701343536, "learning_rate": 8.480276354558496e-05, "loss": 0.0278, "step": 29220 }, { "grad_norm": 0.33715489506721497, "learning_rate": 8.479088992457913e-05, "loss": 0.0311, "step": 29230 }, { "grad_norm": 0.35770201683044434, "learning_rate": 8.477901249889387e-05, "loss": 0.0312, "step": 29240 }, { "grad_norm": 0.2249845564365387, "learning_rate": 8.47671312698281e-05, "loss": 0.0318, "step": 29250 }, { "grad_norm": 0.20769435167312622, "learning_rate": 8.475524623868112e-05, "loss": 0.0304, "step": 29260 }, { "grad_norm": 0.21104943752288818, "learning_rate": 8.474335740675266e-05, "loss": 0.0286, "step": 29270 }, { "grad_norm": 0.2738797962665558, "learning_rate": 8.473146477534289e-05, "loss": 0.0286, "step": 29280 }, { "grad_norm": 0.18674223124980927, "learning_rate": 8.471956834575232e-05, "loss": 0.0233, "step": 29290 }, { "grad_norm": 0.29267388582229614, "learning_rate": 8.470766811928197e-05, "loss": 0.0298, "step": 29300 }, { "grad_norm": 0.2773039937019348, "learning_rate": 8.469576409723323e-05, "loss": 0.0358, "step": 29310 }, { "grad_norm": 0.27785009145736694, "learning_rate": 8.468385628090788e-05, "loss": 0.0271, "step": 29320 }, { "grad_norm": 0.2314848154783249, "learning_rate": 8.467194467160815e-05, "loss": 0.0305, "step": 29330 }, { "grad_norm": 0.2767505347728729, "learning_rate": 8.466002927063667e-05, "loss": 0.0288, "step": 29340 }, { "grad_norm": 0.36094093322753906, "learning_rate": 8.464811007929651e-05, "loss": 0.0276, "step": 29350 }, { "grad_norm": 0.20757746696472168, "learning_rate": 8.463618709889114e-05, "loss": 0.0283, "step": 29360 }, { "grad_norm": 0.24942319095134735, "learning_rate": 8.462426033072442e-05, "loss": 0.0294, "step": 29370 }, { "grad_norm": 0.26674380898475647, "learning_rate": 8.461232977610061e-05, "loss": 0.0294, "step": 29380 }, { "grad_norm": 0.261844664812088, "learning_rate": 8.46003954363245e-05, "loss": 0.0296, "step": 29390 }, { "grad_norm": 0.2101154625415802, "learning_rate": 8.458845731270115e-05, "loss": 0.0287, "step": 29400 }, { "grad_norm": 0.3518773019313812, "learning_rate": 8.45765154065361e-05, "loss": 0.0315, "step": 29410 }, { "grad_norm": 0.3843096196651459, "learning_rate": 8.456456971913532e-05, "loss": 0.0242, "step": 29420 }, { "grad_norm": 0.22772733867168427, "learning_rate": 8.455262025180517e-05, "loss": 0.0292, "step": 29430 }, { "grad_norm": 0.2613855004310608, "learning_rate": 8.454066700585242e-05, "loss": 0.0273, "step": 29440 }, { "grad_norm": 0.24513167142868042, "learning_rate": 8.452870998258423e-05, "loss": 0.031, "step": 29450 }, { "grad_norm": 0.2585466802120209, "learning_rate": 8.451674918330825e-05, "loss": 0.0287, "step": 29460 }, { "grad_norm": 0.29565152525901794, "learning_rate": 8.450478460933246e-05, "loss": 0.0258, "step": 29470 }, { "grad_norm": 0.2439287155866623, "learning_rate": 8.449281626196532e-05, "loss": 0.0271, "step": 29480 }, { "grad_norm": 0.2646763324737549, "learning_rate": 8.448084414251564e-05, "loss": 0.0259, "step": 29490 }, { "grad_norm": 0.33198314905166626, "learning_rate": 8.446886825229271e-05, "loss": 0.0255, "step": 29500 }, { "grad_norm": 0.2289891242980957, "learning_rate": 8.445688859260615e-05, "loss": 0.0235, "step": 29510 }, { "grad_norm": 0.4691043198108673, "learning_rate": 8.444490516476606e-05, "loss": 0.0252, "step": 29520 }, { "grad_norm": 0.2741444706916809, "learning_rate": 8.443291797008293e-05, "loss": 0.0266, "step": 29530 }, { "grad_norm": 0.15381857752799988, "learning_rate": 8.442092700986765e-05, "loss": 0.0274, "step": 29540 }, { "grad_norm": 0.2006787657737732, "learning_rate": 8.440893228543156e-05, "loss": 0.0275, "step": 29550 }, { "grad_norm": 0.19170501828193665, "learning_rate": 8.439693379808638e-05, "loss": 0.0255, "step": 29560 }, { "grad_norm": 0.3391876220703125, "learning_rate": 8.43849315491442e-05, "loss": 0.0285, "step": 29570 }, { "grad_norm": 0.2244645655155182, "learning_rate": 8.437292553991763e-05, "loss": 0.0265, "step": 29580 }, { "grad_norm": 0.22329850494861603, "learning_rate": 8.436091577171959e-05, "loss": 0.021, "step": 29590 }, { "grad_norm": 0.21307635307312012, "learning_rate": 8.434890224586347e-05, "loss": 0.027, "step": 29600 }, { "grad_norm": 0.187372088432312, "learning_rate": 8.433688496366303e-05, "loss": 0.0248, "step": 29610 }, { "grad_norm": 0.20424127578735352, "learning_rate": 8.432486392643248e-05, "loss": 0.0288, "step": 29620 }, { "grad_norm": 0.23854710161685944, "learning_rate": 8.431283913548643e-05, "loss": 0.0255, "step": 29630 }, { "grad_norm": 0.22941435873508453, "learning_rate": 8.430081059213985e-05, "loss": 0.0279, "step": 29640 }, { "grad_norm": 0.26676708459854126, "learning_rate": 8.428877829770823e-05, "loss": 0.0297, "step": 29650 }, { "grad_norm": 0.2941827178001404, "learning_rate": 8.427674225350735e-05, "loss": 0.0369, "step": 29660 }, { "grad_norm": 0.308377206325531, "learning_rate": 8.426470246085347e-05, "loss": 0.0314, "step": 29670 }, { "grad_norm": 0.5507528781890869, "learning_rate": 8.425265892106324e-05, "loss": 0.0279, "step": 29680 }, { "grad_norm": 0.22892193496227264, "learning_rate": 8.424061163545374e-05, "loss": 0.0271, "step": 29690 }, { "grad_norm": 0.2570532262325287, "learning_rate": 8.422856060534243e-05, "loss": 0.0294, "step": 29700 }, { "grad_norm": 0.22552895545959473, "learning_rate": 8.421650583204718e-05, "loss": 0.0261, "step": 29710 }, { "grad_norm": 0.6274570226669312, "learning_rate": 8.420444731688633e-05, "loss": 0.0309, "step": 29720 }, { "grad_norm": 0.5804569125175476, "learning_rate": 8.419238506117852e-05, "loss": 0.0307, "step": 29730 }, { "grad_norm": 0.27180302143096924, "learning_rate": 8.418031906624289e-05, "loss": 0.0282, "step": 29740 }, { "grad_norm": 0.2128669023513794, "learning_rate": 8.416824933339898e-05, "loss": 0.0336, "step": 29750 }, { "grad_norm": 0.39828407764434814, "learning_rate": 8.415617586396667e-05, "loss": 0.0328, "step": 29760 }, { "grad_norm": 0.2866886854171753, "learning_rate": 8.414409865926632e-05, "loss": 0.0326, "step": 29770 }, { "grad_norm": 0.2016671895980835, "learning_rate": 8.413201772061867e-05, "loss": 0.033, "step": 29780 }, { "grad_norm": 0.3665384352207184, "learning_rate": 8.411993304934488e-05, "loss": 0.029, "step": 29790 }, { "grad_norm": 0.32513561844825745, "learning_rate": 8.410784464676654e-05, "loss": 0.0304, "step": 29800 }, { "grad_norm": 0.43834391236305237, "learning_rate": 8.409575251420556e-05, "loss": 0.0297, "step": 29810 }, { "grad_norm": 0.5241628289222717, "learning_rate": 8.408365665298435e-05, "loss": 0.0288, "step": 29820 }, { "grad_norm": 0.3363141119480133, "learning_rate": 8.40715570644257e-05, "loss": 0.0331, "step": 29830 }, { "grad_norm": 0.23483017086982727, "learning_rate": 8.40594537498528e-05, "loss": 0.0312, "step": 29840 }, { "grad_norm": 0.24172811210155487, "learning_rate": 8.404734671058924e-05, "loss": 0.0287, "step": 29850 }, { "grad_norm": 0.25215572118759155, "learning_rate": 8.403523594795902e-05, "loss": 0.0313, "step": 29860 }, { "grad_norm": 0.19960737228393555, "learning_rate": 8.402312146328659e-05, "loss": 0.0312, "step": 29870 }, { "grad_norm": 0.3058892786502838, "learning_rate": 8.401100325789675e-05, "loss": 0.0325, "step": 29880 }, { "grad_norm": 0.3456609845161438, "learning_rate": 8.399888133311472e-05, "loss": 0.0321, "step": 29890 }, { "grad_norm": 0.2564021348953247, "learning_rate": 8.398675569026613e-05, "loss": 0.0262, "step": 29900 }, { "grad_norm": 0.2412392497062683, "learning_rate": 8.397462633067705e-05, "loss": 0.0279, "step": 29910 }, { "grad_norm": 0.2735379636287689, "learning_rate": 8.396249325567392e-05, "loss": 0.029, "step": 29920 }, { "grad_norm": 0.27841445803642273, "learning_rate": 8.395035646658357e-05, "loss": 0.0367, "step": 29930 }, { "grad_norm": 0.23286984860897064, "learning_rate": 8.39382159647333e-05, "loss": 0.035, "step": 29940 }, { "grad_norm": 0.455066442489624, "learning_rate": 8.392607175145075e-05, "loss": 0.032, "step": 29950 }, { "grad_norm": 0.2941608130931854, "learning_rate": 8.3913923828064e-05, "loss": 0.0284, "step": 29960 }, { "grad_norm": 0.2720646858215332, "learning_rate": 8.390177219590152e-05, "loss": 0.0283, "step": 29970 }, { "grad_norm": 0.21889090538024902, "learning_rate": 8.388961685629222e-05, "loss": 0.0276, "step": 29980 }, { "grad_norm": 0.23796683549880981, "learning_rate": 8.387745781056536e-05, "loss": 0.0366, "step": 29990 }, { "grad_norm": 0.2679022550582886, "learning_rate": 8.386529506005065e-05, "loss": 0.0288, "step": 30000 }, { "grad_norm": 0.24462857842445374, "learning_rate": 8.38531286060782e-05, "loss": 0.0254, "step": 30010 }, { "grad_norm": 0.21085551381111145, "learning_rate": 8.384095844997849e-05, "loss": 0.0286, "step": 30020 }, { "grad_norm": 0.3230235278606415, "learning_rate": 8.382878459308245e-05, "loss": 0.0323, "step": 30030 }, { "grad_norm": 0.22091545164585114, "learning_rate": 8.381660703672138e-05, "loss": 0.0276, "step": 30040 }, { "grad_norm": 0.22620993852615356, "learning_rate": 8.380442578222702e-05, "loss": 0.0232, "step": 30050 }, { "grad_norm": 0.27649950981140137, "learning_rate": 8.379224083093146e-05, "loss": 0.0262, "step": 30060 }, { "grad_norm": 0.29731470346450806, "learning_rate": 8.378005218416727e-05, "loss": 0.027, "step": 30070 }, { "grad_norm": 0.19306378066539764, "learning_rate": 8.376785984326735e-05, "loss": 0.0259, "step": 30080 }, { "grad_norm": 0.18235521018505096, "learning_rate": 8.375566380956506e-05, "loss": 0.0238, "step": 30090 }, { "grad_norm": 0.2304714024066925, "learning_rate": 8.374346408439411e-05, "loss": 0.0277, "step": 30100 }, { "grad_norm": 0.3003185987472534, "learning_rate": 8.373126066908868e-05, "loss": 0.028, "step": 30110 }, { "grad_norm": 0.2288973033428192, "learning_rate": 8.371905356498326e-05, "loss": 0.0273, "step": 30120 }, { "grad_norm": 0.23730571568012238, "learning_rate": 8.370684277341288e-05, "loss": 0.0258, "step": 30130 }, { "grad_norm": 0.24171800911426544, "learning_rate": 8.369462829571282e-05, "loss": 0.0293, "step": 30140 }, { "grad_norm": 0.20931921899318695, "learning_rate": 8.36824101332189e-05, "loss": 0.0282, "step": 30150 }, { "grad_norm": 0.21649226546287537, "learning_rate": 8.367018828726721e-05, "loss": 0.0326, "step": 30160 }, { "grad_norm": 0.20103828608989716, "learning_rate": 8.365796275919438e-05, "loss": 0.0249, "step": 30170 }, { "grad_norm": 0.6226204037666321, "learning_rate": 8.364573355033734e-05, "loss": 0.0397, "step": 30180 }, { "grad_norm": 0.34096837043762207, "learning_rate": 8.363350066203346e-05, "loss": 0.0309, "step": 30190 }, { "grad_norm": 0.29737675189971924, "learning_rate": 8.362126409562053e-05, "loss": 0.0284, "step": 30200 }, { "grad_norm": 0.23535498976707458, "learning_rate": 8.360902385243667e-05, "loss": 0.0309, "step": 30210 }, { "grad_norm": 0.2587186098098755, "learning_rate": 8.359677993382052e-05, "loss": 0.0355, "step": 30220 }, { "grad_norm": 0.2532632350921631, "learning_rate": 8.358453234111103e-05, "loss": 0.025, "step": 30230 }, { "grad_norm": 0.2612595558166504, "learning_rate": 8.357228107564756e-05, "loss": 0.0322, "step": 30240 }, { "grad_norm": 0.22431330382823944, "learning_rate": 8.356002613876993e-05, "loss": 0.0278, "step": 30250 }, { "grad_norm": 0.22470161318778992, "learning_rate": 8.35477675318183e-05, "loss": 0.0264, "step": 30260 }, { "grad_norm": 0.21571294963359833, "learning_rate": 8.353550525613323e-05, "loss": 0.0334, "step": 30270 }, { "grad_norm": 0.30349406599998474, "learning_rate": 8.352323931305572e-05, "loss": 0.0268, "step": 30280 }, { "grad_norm": 0.25535187125205994, "learning_rate": 8.351096970392717e-05, "loss": 0.0285, "step": 30290 }, { "grad_norm": 0.27467653155326843, "learning_rate": 8.349869643008937e-05, "loss": 0.0303, "step": 30300 }, { "grad_norm": 0.29804277420043945, "learning_rate": 8.348641949288449e-05, "loss": 0.0284, "step": 30310 }, { "grad_norm": 0.26030242443084717, "learning_rate": 8.34741388936551e-05, "loss": 0.0334, "step": 30320 }, { "grad_norm": 0.2150169312953949, "learning_rate": 8.346185463374423e-05, "loss": 0.0291, "step": 30330 }, { "grad_norm": 0.24033595621585846, "learning_rate": 8.344956671449524e-05, "loss": 0.0319, "step": 30340 }, { "grad_norm": 0.4210111200809479, "learning_rate": 8.343727513725192e-05, "loss": 0.0323, "step": 30350 }, { "grad_norm": 0.29059576988220215, "learning_rate": 8.342497990335847e-05, "loss": 0.0334, "step": 30360 }, { "grad_norm": 0.35499072074890137, "learning_rate": 8.341268101415946e-05, "loss": 0.0322, "step": 30370 }, { "grad_norm": 0.26296406984329224, "learning_rate": 8.34003784709999e-05, "loss": 0.0353, "step": 30380 }, { "grad_norm": 0.21529628336429596, "learning_rate": 8.338807227522516e-05, "loss": 0.0288, "step": 30390 }, { "grad_norm": 0.22123554348945618, "learning_rate": 8.337576242818103e-05, "loss": 0.0297, "step": 30400 }, { "grad_norm": 0.2601052522659302, "learning_rate": 8.336344893121372e-05, "loss": 0.0297, "step": 30410 }, { "grad_norm": 0.23772042989730835, "learning_rate": 8.335113178566977e-05, "loss": 0.0276, "step": 30420 }, { "grad_norm": 0.7227645516395569, "learning_rate": 8.33388109928962e-05, "loss": 0.0313, "step": 30430 }, { "grad_norm": 0.2550658881664276, "learning_rate": 8.33264865542404e-05, "loss": 0.0328, "step": 30440 }, { "grad_norm": 0.36017483472824097, "learning_rate": 8.331415847105013e-05, "loss": 0.0315, "step": 30450 }, { "grad_norm": 0.2230597883462906, "learning_rate": 8.330182674467357e-05, "loss": 0.0331, "step": 30460 }, { "grad_norm": 0.27259355783462524, "learning_rate": 8.32894913764593e-05, "loss": 0.0342, "step": 30470 }, { "grad_norm": 0.2238520383834839, "learning_rate": 8.327715236775633e-05, "loss": 0.0342, "step": 30480 }, { "grad_norm": 0.35316571593284607, "learning_rate": 8.326480971991398e-05, "loss": 0.032, "step": 30490 }, { "grad_norm": 0.2784489095211029, "learning_rate": 8.325246343428206e-05, "loss": 0.0344, "step": 30500 }, { "grad_norm": 0.32934144139289856, "learning_rate": 8.324011351221072e-05, "loss": 0.0278, "step": 30510 }, { "grad_norm": 0.2687230706214905, "learning_rate": 8.322775995505057e-05, "loss": 0.0316, "step": 30520 }, { "grad_norm": 0.26156923174858093, "learning_rate": 8.321540276415254e-05, "loss": 0.0288, "step": 30530 }, { "grad_norm": 0.3652398884296417, "learning_rate": 8.320304194086798e-05, "loss": 0.0322, "step": 30540 }, { "grad_norm": 0.43239015340805054, "learning_rate": 8.31906774865487e-05, "loss": 0.0358, "step": 30550 }, { "grad_norm": 0.26894819736480713, "learning_rate": 8.317830940254682e-05, "loss": 0.0307, "step": 30560 }, { "grad_norm": 0.3835921287536621, "learning_rate": 8.316593769021491e-05, "loss": 0.0274, "step": 30570 }, { "grad_norm": 0.29059046506881714, "learning_rate": 8.315356235090592e-05, "loss": 0.0245, "step": 30580 }, { "grad_norm": 0.18810583651065826, "learning_rate": 8.314118338597319e-05, "loss": 0.0259, "step": 30590 }, { "grad_norm": 0.2070809006690979, "learning_rate": 8.312880079677048e-05, "loss": 0.0238, "step": 30600 }, { "grad_norm": 0.19727253913879395, "learning_rate": 8.311641458465191e-05, "loss": 0.0258, "step": 30610 }, { "grad_norm": 0.19573570787906647, "learning_rate": 8.310402475097205e-05, "loss": 0.0278, "step": 30620 }, { "grad_norm": 0.2655319273471832, "learning_rate": 8.309163129708581e-05, "loss": 0.0314, "step": 30630 }, { "grad_norm": 0.1981992870569229, "learning_rate": 8.307923422434852e-05, "loss": 0.0295, "step": 30640 }, { "grad_norm": 0.41455739736557007, "learning_rate": 8.30668335341159e-05, "loss": 0.03, "step": 30650 }, { "grad_norm": 0.26810261607170105, "learning_rate": 8.305442922774408e-05, "loss": 0.0248, "step": 30660 }, { "grad_norm": 0.18094244599342346, "learning_rate": 8.304202130658959e-05, "loss": 0.0245, "step": 30670 }, { "grad_norm": 0.219563290476799, "learning_rate": 8.302960977200931e-05, "loss": 0.0305, "step": 30680 }, { "grad_norm": 0.390444815158844, "learning_rate": 8.301719462536058e-05, "loss": 0.027, "step": 30690 }, { "grad_norm": 0.239581897854805, "learning_rate": 8.300477586800108e-05, "loss": 0.0301, "step": 30700 }, { "grad_norm": 0.3399900496006012, "learning_rate": 8.299235350128892e-05, "loss": 0.0262, "step": 30710 }, { "grad_norm": 0.21151325106620789, "learning_rate": 8.297992752658258e-05, "loss": 0.0293, "step": 30720 }, { "grad_norm": 0.33535027503967285, "learning_rate": 8.296749794524094e-05, "loss": 0.0332, "step": 30730 }, { "grad_norm": 0.21826091408729553, "learning_rate": 8.295506475862332e-05, "loss": 0.0248, "step": 30740 }, { "grad_norm": 0.2695063054561615, "learning_rate": 8.294262796808933e-05, "loss": 0.0257, "step": 30750 }, { "grad_norm": 0.31396690011024475, "learning_rate": 8.293018757499909e-05, "loss": 0.0289, "step": 30760 }, { "grad_norm": 0.22788965702056885, "learning_rate": 8.291774358071305e-05, "loss": 0.0275, "step": 30770 }, { "grad_norm": 0.2610624432563782, "learning_rate": 8.290529598659205e-05, "loss": 0.0245, "step": 30780 }, { "grad_norm": 0.2584162652492523, "learning_rate": 8.289284479399738e-05, "loss": 0.028, "step": 30790 }, { "grad_norm": 0.2548536956310272, "learning_rate": 8.288039000429064e-05, "loss": 0.0282, "step": 30800 }, { "grad_norm": 0.5183523297309875, "learning_rate": 8.286793161883388e-05, "loss": 0.0293, "step": 30810 }, { "grad_norm": 0.22560304403305054, "learning_rate": 8.285546963898954e-05, "loss": 0.0275, "step": 30820 }, { "grad_norm": 0.40332460403442383, "learning_rate": 8.284300406612044e-05, "loss": 0.0271, "step": 30830 }, { "grad_norm": 0.25327908992767334, "learning_rate": 8.283053490158978e-05, "loss": 0.0229, "step": 30840 }, { "grad_norm": 0.23686930537223816, "learning_rate": 8.28180621467612e-05, "loss": 0.0273, "step": 30850 }, { "grad_norm": 0.1905195713043213, "learning_rate": 8.280558580299867e-05, "loss": 0.0286, "step": 30860 }, { "grad_norm": 0.33378610014915466, "learning_rate": 8.279310587166661e-05, "loss": 0.033, "step": 30870 }, { "grad_norm": 0.20642340183258057, "learning_rate": 8.278062235412978e-05, "loss": 0.0281, "step": 30880 }, { "grad_norm": 0.339912086725235, "learning_rate": 8.276813525175339e-05, "loss": 0.0287, "step": 30890 }, { "grad_norm": 0.2528540790081024, "learning_rate": 8.2755644565903e-05, "loss": 0.0241, "step": 30900 }, { "grad_norm": 0.2561543583869934, "learning_rate": 8.274315029794454e-05, "loss": 0.0302, "step": 30910 }, { "grad_norm": 0.2721327245235443, "learning_rate": 8.273065244924443e-05, "loss": 0.0217, "step": 30920 }, { "grad_norm": 0.19846367835998535, "learning_rate": 8.271815102116936e-05, "loss": 0.0265, "step": 30930 }, { "grad_norm": 0.30015090107917786, "learning_rate": 8.270564601508648e-05, "loss": 0.0302, "step": 30940 }, { "grad_norm": 0.22480842471122742, "learning_rate": 8.269313743236333e-05, "loss": 0.0274, "step": 30950 }, { "grad_norm": 0.17338413000106812, "learning_rate": 8.268062527436783e-05, "loss": 0.0243, "step": 30960 }, { "grad_norm": 0.2098439335823059, "learning_rate": 8.266810954246828e-05, "loss": 0.0336, "step": 30970 }, { "grad_norm": 0.3042897880077362, "learning_rate": 8.265559023803338e-05, "loss": 0.0306, "step": 30980 }, { "grad_norm": 0.22698119282722473, "learning_rate": 8.264306736243223e-05, "loss": 0.0267, "step": 30990 }, { "grad_norm": 0.25989046692848206, "learning_rate": 8.263054091703432e-05, "loss": 0.0281, "step": 31000 }, { "grad_norm": 0.26023557782173157, "learning_rate": 8.26180109032095e-05, "loss": 0.0265, "step": 31010 }, { "grad_norm": 0.2712937295436859, "learning_rate": 8.260547732232807e-05, "loss": 0.0261, "step": 31020 }, { "grad_norm": 0.2223110944032669, "learning_rate": 8.259294017576063e-05, "loss": 0.0235, "step": 31030 }, { "grad_norm": 0.18663790822029114, "learning_rate": 8.258039946487828e-05, "loss": 0.0259, "step": 31040 }, { "grad_norm": 0.1997617483139038, "learning_rate": 8.256785519105241e-05, "loss": 0.0289, "step": 31050 }, { "grad_norm": 0.27629077434539795, "learning_rate": 8.255530735565488e-05, "loss": 0.0332, "step": 31060 }, { "grad_norm": 0.27308687567710876, "learning_rate": 8.254275596005787e-05, "loss": 0.0294, "step": 31070 }, { "grad_norm": 0.23658064007759094, "learning_rate": 8.2530201005634e-05, "loss": 0.0294, "step": 31080 }, { "grad_norm": 0.3470397889614105, "learning_rate": 8.251764249375626e-05, "loss": 0.0299, "step": 31090 }, { "grad_norm": 0.24547241628170013, "learning_rate": 8.250508042579803e-05, "loss": 0.0318, "step": 31100 }, { "grad_norm": 0.2204509675502777, "learning_rate": 8.249251480313307e-05, "loss": 0.0293, "step": 31110 }, { "grad_norm": 0.2993306815624237, "learning_rate": 8.247994562713555e-05, "loss": 0.0313, "step": 31120 }, { "grad_norm": 0.19431670010089874, "learning_rate": 8.246737289917998e-05, "loss": 0.0287, "step": 31130 }, { "grad_norm": 0.2627181112766266, "learning_rate": 8.245479662064135e-05, "loss": 0.0292, "step": 31140 }, { "grad_norm": 0.26626017689704895, "learning_rate": 8.244221679289496e-05, "loss": 0.0269, "step": 31150 }, { "grad_norm": 0.24241852760314941, "learning_rate": 8.242963341731652e-05, "loss": 0.0264, "step": 31160 }, { "grad_norm": 0.5073975920677185, "learning_rate": 8.24170464952821e-05, "loss": 0.0287, "step": 31170 }, { "grad_norm": 0.24601179361343384, "learning_rate": 8.240445602816824e-05, "loss": 0.0273, "step": 31180 }, { "grad_norm": 0.28981924057006836, "learning_rate": 8.239186201735179e-05, "loss": 0.0334, "step": 31190 }, { "grad_norm": 0.26016345620155334, "learning_rate": 8.237926446420998e-05, "loss": 0.0284, "step": 31200 }, { "grad_norm": 0.2293347269296646, "learning_rate": 8.236666337012052e-05, "loss": 0.027, "step": 31210 }, { "grad_norm": 0.198297917842865, "learning_rate": 8.23540587364614e-05, "loss": 0.0277, "step": 31220 }, { "grad_norm": 0.20400437712669373, "learning_rate": 8.234145056461107e-05, "loss": 0.0299, "step": 31230 }, { "grad_norm": 0.28722742199897766, "learning_rate": 8.232883885594832e-05, "loss": 0.0248, "step": 31240 }, { "grad_norm": 0.17497491836547852, "learning_rate": 8.231622361185236e-05, "loss": 0.0249, "step": 31250 }, { "grad_norm": 0.2796846330165863, "learning_rate": 8.230360483370278e-05, "loss": 0.0301, "step": 31260 }, { "grad_norm": 0.23049958050251007, "learning_rate": 8.229098252287953e-05, "loss": 0.0281, "step": 31270 }, { "grad_norm": 0.6262218356132507, "learning_rate": 8.2278356680763e-05, "loss": 0.0295, "step": 31280 }, { "grad_norm": 0.3872753977775574, "learning_rate": 8.22657273087339e-05, "loss": 0.0278, "step": 31290 }, { "grad_norm": 0.24500428140163422, "learning_rate": 8.225309440817336e-05, "loss": 0.0318, "step": 31300 }, { "grad_norm": 0.23877739906311035, "learning_rate": 8.224045798046293e-05, "loss": 0.0314, "step": 31310 }, { "grad_norm": 0.19522389769554138, "learning_rate": 8.22278180269845e-05, "loss": 0.03, "step": 31320 }, { "grad_norm": 0.21964803338050842, "learning_rate": 8.221517454912031e-05, "loss": 0.0309, "step": 31330 }, { "grad_norm": 0.3399277627468109, "learning_rate": 8.220252754825308e-05, "loss": 0.0295, "step": 31340 }, { "grad_norm": 0.34754449129104614, "learning_rate": 8.218987702576586e-05, "loss": 0.0279, "step": 31350 }, { "grad_norm": 0.27154457569122314, "learning_rate": 8.217722298304207e-05, "loss": 0.0352, "step": 31360 }, { "grad_norm": 0.2061089128255844, "learning_rate": 8.216456542146557e-05, "loss": 0.028, "step": 31370 }, { "grad_norm": 0.20816878974437714, "learning_rate": 8.215190434242055e-05, "loss": 0.0299, "step": 31380 }, { "grad_norm": 0.4089273512363434, "learning_rate": 8.213923974729161e-05, "loss": 0.0297, "step": 31390 }, { "grad_norm": 0.19540813565254211, "learning_rate": 8.212657163746373e-05, "loss": 0.0266, "step": 31400 }, { "grad_norm": 0.2890082597732544, "learning_rate": 8.211390001432227e-05, "loss": 0.0293, "step": 31410 }, { "grad_norm": 0.22142620384693146, "learning_rate": 8.210122487925297e-05, "loss": 0.0265, "step": 31420 }, { "grad_norm": 0.27818363904953003, "learning_rate": 8.208854623364202e-05, "loss": 0.0282, "step": 31430 }, { "grad_norm": 0.2135395109653473, "learning_rate": 8.207586407887589e-05, "loss": 0.0294, "step": 31440 }, { "grad_norm": 0.31916099786758423, "learning_rate": 8.206317841634148e-05, "loss": 0.0268, "step": 31450 }, { "grad_norm": 0.2513880133628845, "learning_rate": 8.205048924742609e-05, "loss": 0.0329, "step": 31460 }, { "grad_norm": 0.21779070794582367, "learning_rate": 8.203779657351738e-05, "loss": 0.03, "step": 31470 }, { "grad_norm": 0.29322895407676697, "learning_rate": 8.20251003960034e-05, "loss": 0.0319, "step": 31480 }, { "grad_norm": 0.25368890166282654, "learning_rate": 8.201240071627258e-05, "loss": 0.0271, "step": 31490 }, { "grad_norm": 0.4304019510746002, "learning_rate": 8.199969753571377e-05, "loss": 0.03, "step": 31500 }, { "grad_norm": 0.6950773000717163, "learning_rate": 8.198699085571615e-05, "loss": 0.0277, "step": 31510 }, { "grad_norm": 0.2822577655315399, "learning_rate": 8.197428067766928e-05, "loss": 0.0329, "step": 31520 }, { "grad_norm": 0.31145867705345154, "learning_rate": 8.196156700296316e-05, "loss": 0.0283, "step": 31530 }, { "grad_norm": 0.2665460407733917, "learning_rate": 8.194884983298814e-05, "loss": 0.0289, "step": 31540 }, { "grad_norm": 0.2988061308860779, "learning_rate": 8.193612916913491e-05, "loss": 0.0311, "step": 31550 }, { "grad_norm": 0.5701351761817932, "learning_rate": 8.192340501279463e-05, "loss": 0.0407, "step": 31560 }, { "grad_norm": 0.24747037887573242, "learning_rate": 8.191067736535876e-05, "loss": 0.0328, "step": 31570 }, { "grad_norm": 0.2387581169605255, "learning_rate": 8.18979462282192e-05, "loss": 0.0314, "step": 31580 }, { "grad_norm": 0.25199317932128906, "learning_rate": 8.188521160276819e-05, "loss": 0.0292, "step": 31590 }, { "grad_norm": 0.2776709496974945, "learning_rate": 8.187247349039837e-05, "loss": 0.0258, "step": 31600 }, { "grad_norm": 1.276877999305725, "learning_rate": 8.185973189250278e-05, "loss": 0.0314, "step": 31610 }, { "grad_norm": 0.3514576852321625, "learning_rate": 8.184698681047482e-05, "loss": 0.0297, "step": 31620 }, { "grad_norm": 0.32253098487854004, "learning_rate": 8.183423824570827e-05, "loss": 0.0364, "step": 31630 }, { "grad_norm": 0.28919753432273865, "learning_rate": 8.182148619959725e-05, "loss": 0.0322, "step": 31640 }, { "grad_norm": 0.29052966833114624, "learning_rate": 8.180873067353636e-05, "loss": 0.0326, "step": 31650 }, { "grad_norm": 0.23045960068702698, "learning_rate": 8.179597166892052e-05, "loss": 0.0291, "step": 31660 }, { "grad_norm": 0.24225488305091858, "learning_rate": 8.178320918714501e-05, "loss": 0.0338, "step": 31670 }, { "grad_norm": 0.2452707141637802, "learning_rate": 8.177044322960554e-05, "loss": 0.0312, "step": 31680 }, { "grad_norm": 0.24380122125148773, "learning_rate": 8.175767379769816e-05, "loss": 0.0318, "step": 31690 }, { "grad_norm": 0.1678653359413147, "learning_rate": 8.174490089281932e-05, "loss": 0.0257, "step": 31700 }, { "grad_norm": 0.1935458481311798, "learning_rate": 8.173212451636584e-05, "loss": 0.0277, "step": 31710 }, { "grad_norm": 0.2446756362915039, "learning_rate": 8.171934466973493e-05, "loss": 0.0314, "step": 31720 }, { "grad_norm": 0.20489291846752167, "learning_rate": 8.170656135432418e-05, "loss": 0.0253, "step": 31730 }, { "grad_norm": 0.23173758387565613, "learning_rate": 8.169377457153155e-05, "loss": 0.0258, "step": 31740 }, { "grad_norm": 0.3479170799255371, "learning_rate": 8.168098432275539e-05, "loss": 0.0295, "step": 31750 }, { "grad_norm": 0.24845977127552032, "learning_rate": 8.166819060939442e-05, "loss": 0.0289, "step": 31760 }, { "grad_norm": 0.6548343300819397, "learning_rate": 8.165539343284772e-05, "loss": 0.027, "step": 31770 }, { "grad_norm": 0.2149624079465866, "learning_rate": 8.16425927945148e-05, "loss": 0.0264, "step": 31780 }, { "grad_norm": 0.19905619323253632, "learning_rate": 8.162978869579551e-05, "loss": 0.0282, "step": 31790 }, { "grad_norm": 0.20111298561096191, "learning_rate": 8.161698113809007e-05, "loss": 0.0267, "step": 31800 }, { "grad_norm": 0.2814646363258362, "learning_rate": 8.160417012279911e-05, "loss": 0.029, "step": 31810 }, { "grad_norm": 0.19036319851875305, "learning_rate": 8.159135565132363e-05, "loss": 0.0281, "step": 31820 }, { "grad_norm": 0.27490681409835815, "learning_rate": 8.157853772506498e-05, "loss": 0.0286, "step": 31830 }, { "grad_norm": 0.25697633624076843, "learning_rate": 8.156571634542494e-05, "loss": 0.0275, "step": 31840 }, { "grad_norm": 0.18058204650878906, "learning_rate": 8.15528915138056e-05, "loss": 0.0261, "step": 31850 }, { "grad_norm": 0.2084827870130539, "learning_rate": 8.154006323160949e-05, "loss": 0.0219, "step": 31860 }, { "grad_norm": 0.29359960556030273, "learning_rate": 8.152723150023949e-05, "loss": 0.0274, "step": 31870 }, { "grad_norm": 0.2213471382856369, "learning_rate": 8.151439632109886e-05, "loss": 0.0237, "step": 31880 }, { "grad_norm": 0.20946937799453735, "learning_rate": 8.150155769559122e-05, "loss": 0.0278, "step": 31890 }, { "grad_norm": 0.261994332075119, "learning_rate": 8.148871562512058e-05, "loss": 0.0285, "step": 31900 }, { "grad_norm": 0.24755670130252838, "learning_rate": 8.147587011109136e-05, "loss": 0.0275, "step": 31910 }, { "grad_norm": 0.19704510271549225, "learning_rate": 8.14630211549083e-05, "loss": 0.0258, "step": 31920 }, { "grad_norm": 0.2380661517381668, "learning_rate": 8.145016875797655e-05, "loss": 0.0243, "step": 31930 }, { "grad_norm": 0.25040915608406067, "learning_rate": 8.143731292170164e-05, "loss": 0.0248, "step": 31940 }, { "grad_norm": 0.2152094691991806, "learning_rate": 8.142445364748944e-05, "loss": 0.0256, "step": 31950 }, { "grad_norm": 0.2087167501449585, "learning_rate": 8.141159093674624e-05, "loss": 0.025, "step": 31960 }, { "grad_norm": 0.22445175051689148, "learning_rate": 8.139872479087869e-05, "loss": 0.0264, "step": 31970 }, { "grad_norm": 0.1663569211959839, "learning_rate": 8.13858552112938e-05, "loss": 0.0251, "step": 31980 }, { "grad_norm": 0.24160318076610565, "learning_rate": 8.137298219939895e-05, "loss": 0.0289, "step": 31990 }, { "grad_norm": 0.19038733839988708, "learning_rate": 8.136010575660196e-05, "loss": 0.0268, "step": 32000 }, { "grad_norm": 0.3190443813800812, "learning_rate": 8.134722588431095e-05, "loss": 0.0268, "step": 32010 }, { "grad_norm": 0.24086816608905792, "learning_rate": 8.133434258393444e-05, "loss": 0.0283, "step": 32020 }, { "grad_norm": 0.2911648154258728, "learning_rate": 8.132145585688134e-05, "loss": 0.0293, "step": 32030 }, { "grad_norm": 0.41467776894569397, "learning_rate": 8.130856570456093e-05, "loss": 0.0296, "step": 32040 }, { "grad_norm": 0.22261501848697662, "learning_rate": 8.129567212838283e-05, "loss": 0.027, "step": 32050 }, { "grad_norm": 0.18209202587604523, "learning_rate": 8.128277512975708e-05, "loss": 0.0268, "step": 32060 }, { "grad_norm": 0.24496625363826752, "learning_rate": 8.126987471009408e-05, "loss": 0.0263, "step": 32070 }, { "grad_norm": 0.28096911311149597, "learning_rate": 8.125697087080459e-05, "loss": 0.0279, "step": 32080 }, { "grad_norm": 0.29730311036109924, "learning_rate": 8.124406361329976e-05, "loss": 0.028, "step": 32090 }, { "grad_norm": 0.21609999239444733, "learning_rate": 8.123115293899111e-05, "loss": 0.0256, "step": 32100 }, { "grad_norm": 0.2843230068683624, "learning_rate": 8.121823884929055e-05, "loss": 0.0283, "step": 32110 }, { "grad_norm": 0.259970486164093, "learning_rate": 8.12053213456103e-05, "loss": 0.0281, "step": 32120 }, { "grad_norm": 0.22190752625465393, "learning_rate": 8.119240042936303e-05, "loss": 0.0245, "step": 32130 }, { "grad_norm": 0.2613413631916046, "learning_rate": 8.117947610196175e-05, "loss": 0.0268, "step": 32140 }, { "grad_norm": 0.537744402885437, "learning_rate": 8.116654836481982e-05, "loss": 0.0242, "step": 32150 }, { "grad_norm": 0.16600079834461212, "learning_rate": 8.115361721935106e-05, "loss": 0.0325, "step": 32160 }, { "grad_norm": 0.2046288400888443, "learning_rate": 8.114068266696953e-05, "loss": 0.0216, "step": 32170 }, { "grad_norm": 0.20560592412948608, "learning_rate": 8.112774470908978e-05, "loss": 0.026, "step": 32180 }, { "grad_norm": 0.3865208029747009, "learning_rate": 8.111480334712665e-05, "loss": 0.0278, "step": 32190 }, { "grad_norm": 0.3496915102005005, "learning_rate": 8.110185858249542e-05, "loss": 0.0258, "step": 32200 }, { "grad_norm": 0.21934466063976288, "learning_rate": 8.108891041661168e-05, "loss": 0.0259, "step": 32210 }, { "grad_norm": 0.2120644748210907, "learning_rate": 8.107595885089146e-05, "loss": 0.0269, "step": 32220 }, { "grad_norm": 0.2139381319284439, "learning_rate": 8.106300388675108e-05, "loss": 0.0291, "step": 32230 }, { "grad_norm": 0.20242182910442352, "learning_rate": 8.105004552560731e-05, "loss": 0.0265, "step": 32240 }, { "grad_norm": 0.2211281955242157, "learning_rate": 8.103708376887724e-05, "loss": 0.0303, "step": 32250 }, { "grad_norm": 0.20054562389850616, "learning_rate": 8.102411861797836e-05, "loss": 0.0266, "step": 32260 }, { "grad_norm": 0.18782463669776917, "learning_rate": 8.101115007432851e-05, "loss": 0.0252, "step": 32270 }, { "grad_norm": 0.33433452248573303, "learning_rate": 8.09981781393459e-05, "loss": 0.0284, "step": 32280 }, { "grad_norm": 0.2757529318332672, "learning_rate": 8.098520281444915e-05, "loss": 0.0279, "step": 32290 }, { "grad_norm": 0.25317248702049255, "learning_rate": 8.09722241010572e-05, "loss": 0.0276, "step": 32300 }, { "grad_norm": 0.1640080362558365, "learning_rate": 8.095924200058939e-05, "loss": 0.0265, "step": 32310 }, { "grad_norm": 0.287469744682312, "learning_rate": 8.094625651446541e-05, "loss": 0.0277, "step": 32320 }, { "grad_norm": 0.22068001329898834, "learning_rate": 8.093326764410536e-05, "loss": 0.0255, "step": 32330 }, { "grad_norm": 0.23884503543376923, "learning_rate": 8.092027539092966e-05, "loss": 0.0259, "step": 32340 }, { "grad_norm": 0.31785282492637634, "learning_rate": 8.090727975635913e-05, "loss": 0.0308, "step": 32350 }, { "grad_norm": 0.25456666946411133, "learning_rate": 8.089428074181497e-05, "loss": 0.0309, "step": 32360 }, { "grad_norm": 0.22235287725925446, "learning_rate": 8.088127834871871e-05, "loss": 0.0316, "step": 32370 }, { "grad_norm": 0.586432695388794, "learning_rate": 8.086827257849226e-05, "loss": 0.0306, "step": 32380 }, { "grad_norm": 0.2865491211414337, "learning_rate": 8.085526343255795e-05, "loss": 0.0258, "step": 32390 }, { "grad_norm": 0.39434418082237244, "learning_rate": 8.084225091233842e-05, "loss": 0.0322, "step": 32400 }, { "grad_norm": 0.33130577206611633, "learning_rate": 8.082923501925668e-05, "loss": 0.0291, "step": 32410 }, { "grad_norm": 0.5018923878669739, "learning_rate": 8.081621575473617e-05, "loss": 0.0294, "step": 32420 }, { "grad_norm": 0.2454419881105423, "learning_rate": 8.080319312020064e-05, "loss": 0.0296, "step": 32430 }, { "grad_norm": 0.2390935719013214, "learning_rate": 8.079016711707421e-05, "loss": 0.0292, "step": 32440 }, { "grad_norm": 0.2655678987503052, "learning_rate": 8.077713774678139e-05, "loss": 0.0281, "step": 32450 }, { "grad_norm": 0.37229979038238525, "learning_rate": 8.076410501074707e-05, "loss": 0.0329, "step": 32460 }, { "grad_norm": 0.21843814849853516, "learning_rate": 8.075106891039647e-05, "loss": 0.0277, "step": 32470 }, { "grad_norm": 0.4111781716346741, "learning_rate": 8.073802944715523e-05, "loss": 0.0274, "step": 32480 }, { "grad_norm": 0.3224298655986786, "learning_rate": 8.072498662244929e-05, "loss": 0.0288, "step": 32490 }, { "grad_norm": 0.19414427876472473, "learning_rate": 8.0711940437705e-05, "loss": 0.0249, "step": 32500 }, { "grad_norm": 0.2426973134279251, "learning_rate": 8.06988908943491e-05, "loss": 0.026, "step": 32510 }, { "grad_norm": 0.24184221029281616, "learning_rate": 8.068583799380863e-05, "loss": 0.034, "step": 32520 }, { "grad_norm": 0.2520565986633301, "learning_rate": 8.067278173751104e-05, "loss": 0.0291, "step": 32530 }, { "grad_norm": 0.28019872307777405, "learning_rate": 8.065972212688417e-05, "loss": 0.025, "step": 32540 }, { "grad_norm": 0.2755696773529053, "learning_rate": 8.064665916335618e-05, "loss": 0.0299, "step": 32550 }, { "grad_norm": 0.2634671628475189, "learning_rate": 8.063359284835564e-05, "loss": 0.0313, "step": 32560 }, { "grad_norm": 0.257937490940094, "learning_rate": 8.062052318331142e-05, "loss": 0.0259, "step": 32570 }, { "grad_norm": 0.19541388750076294, "learning_rate": 8.060745016965283e-05, "loss": 0.0225, "step": 32580 }, { "grad_norm": 0.22814756631851196, "learning_rate": 8.059437380880952e-05, "loss": 0.0284, "step": 32590 }, { "grad_norm": 0.2519519627094269, "learning_rate": 8.058129410221146e-05, "loss": 0.0272, "step": 32600 }, { "grad_norm": 0.34785422682762146, "learning_rate": 8.056821105128908e-05, "loss": 0.0299, "step": 32610 }, { "grad_norm": 0.2801339030265808, "learning_rate": 8.05551246574731e-05, "loss": 0.0349, "step": 32620 }, { "grad_norm": 0.293673038482666, "learning_rate": 8.05420349221946e-05, "loss": 0.0297, "step": 32630 }, { "grad_norm": 0.25277984142303467, "learning_rate": 8.05289418468851e-05, "loss": 0.0308, "step": 32640 }, { "grad_norm": 0.3036763072013855, "learning_rate": 8.051584543297642e-05, "loss": 0.0301, "step": 32650 }, { "grad_norm": 0.23325251042842865, "learning_rate": 8.050274568190074e-05, "loss": 0.0311, "step": 32660 }, { "grad_norm": 0.20325979590415955, "learning_rate": 8.048964259509067e-05, "loss": 0.0298, "step": 32670 }, { "grad_norm": 0.2086714208126068, "learning_rate": 8.047653617397914e-05, "loss": 0.0259, "step": 32680 }, { "grad_norm": 0.503700852394104, "learning_rate": 8.046342641999941e-05, "loss": 0.0285, "step": 32690 }, { "grad_norm": 0.2077518105506897, "learning_rate": 8.045031333458517e-05, "loss": 0.0276, "step": 32700 }, { "grad_norm": 0.37601369619369507, "learning_rate": 8.043719691917047e-05, "loss": 0.0331, "step": 32710 }, { "grad_norm": 0.2770133316516876, "learning_rate": 8.042407717518966e-05, "loss": 0.0333, "step": 32720 }, { "grad_norm": 0.29489612579345703, "learning_rate": 8.041095410407751e-05, "loss": 0.0299, "step": 32730 }, { "grad_norm": 0.2873101234436035, "learning_rate": 8.039782770726913e-05, "loss": 0.0316, "step": 32740 }, { "grad_norm": 0.2727077007293701, "learning_rate": 8.038469798620004e-05, "loss": 0.0285, "step": 32750 }, { "grad_norm": 0.33916059136390686, "learning_rate": 8.037156494230604e-05, "loss": 0.0271, "step": 32760 }, { "grad_norm": 0.23899942636489868, "learning_rate": 8.035842857702338e-05, "loss": 0.0266, "step": 32770 }, { "grad_norm": 0.26284655928611755, "learning_rate": 8.03452888917886e-05, "loss": 0.0279, "step": 32780 }, { "grad_norm": 0.2909035086631775, "learning_rate": 8.033214588803866e-05, "loss": 0.0342, "step": 32790 }, { "grad_norm": 0.21391555666923523, "learning_rate": 8.031899956721083e-05, "loss": 0.0271, "step": 32800 }, { "grad_norm": 0.16750073432922363, "learning_rate": 8.030584993074282e-05, "loss": 0.0274, "step": 32810 }, { "grad_norm": 0.30448633432388306, "learning_rate": 8.02926969800726e-05, "loss": 0.0224, "step": 32820 }, { "grad_norm": 0.23436953127384186, "learning_rate": 8.027954071663859e-05, "loss": 0.027, "step": 32830 }, { "grad_norm": 0.42940717935562134, "learning_rate": 8.026638114187954e-05, "loss": 0.0283, "step": 32840 }, { "grad_norm": 0.1796412467956543, "learning_rate": 8.025321825723456e-05, "loss": 0.0245, "step": 32850 }, { "grad_norm": 0.16986879706382751, "learning_rate": 8.02400520641431e-05, "loss": 0.026, "step": 32860 }, { "grad_norm": 0.18144717812538147, "learning_rate": 8.022688256404501e-05, "loss": 0.026, "step": 32870 }, { "grad_norm": 0.28194427490234375, "learning_rate": 8.02137097583805e-05, "loss": 0.0269, "step": 32880 }, { "grad_norm": 0.19623273611068726, "learning_rate": 8.02005336485901e-05, "loss": 0.0267, "step": 32890 }, { "grad_norm": 0.4471241533756256, "learning_rate": 8.018735423611476e-05, "loss": 0.028, "step": 32900 }, { "grad_norm": 0.26614683866500854, "learning_rate": 8.017417152239574e-05, "loss": 0.0267, "step": 32910 }, { "grad_norm": 0.2287319153547287, "learning_rate": 8.01609855088747e-05, "loss": 0.0304, "step": 32920 }, { "grad_norm": 0.2434036135673523, "learning_rate": 8.014779619699362e-05, "loss": 0.0274, "step": 32930 }, { "grad_norm": 0.25492873787879944, "learning_rate": 8.013460358819489e-05, "loss": 0.0237, "step": 32940 }, { "grad_norm": 0.29296165704727173, "learning_rate": 8.01214076839212e-05, "loss": 0.0315, "step": 32950 }, { "grad_norm": 0.22360312938690186, "learning_rate": 8.010820848561565e-05, "loss": 0.0249, "step": 32960 }, { "grad_norm": 0.30466514825820923, "learning_rate": 8.009500599472171e-05, "loss": 0.033, "step": 32970 }, { "grad_norm": 0.38486793637275696, "learning_rate": 8.008180021268314e-05, "loss": 0.0291, "step": 32980 }, { "grad_norm": 0.30383944511413574, "learning_rate": 8.006859114094414e-05, "loss": 0.0283, "step": 32990 }, { "grad_norm": 0.38698509335517883, "learning_rate": 8.005537878094921e-05, "loss": 0.0302, "step": 33000 }, { "grad_norm": 0.40383705496788025, "learning_rate": 8.004216313414323e-05, "loss": 0.0246, "step": 33010 }, { "grad_norm": 0.22627076506614685, "learning_rate": 8.002894420197149e-05, "loss": 0.0262, "step": 33020 }, { "grad_norm": 0.3417792022228241, "learning_rate": 8.001572198587954e-05, "loss": 0.0285, "step": 33030 }, { "grad_norm": 0.27435556054115295, "learning_rate": 8.000249648731338e-05, "loss": 0.0293, "step": 33040 }, { "grad_norm": 0.2907601594924927, "learning_rate": 7.998926770771928e-05, "loss": 0.0328, "step": 33050 }, { "grad_norm": 0.2869456708431244, "learning_rate": 7.997603564854397e-05, "loss": 0.0299, "step": 33060 }, { "grad_norm": 0.18765895068645477, "learning_rate": 7.996280031123448e-05, "loss": 0.0265, "step": 33070 }, { "grad_norm": 0.2951035499572754, "learning_rate": 7.994956169723818e-05, "loss": 0.0253, "step": 33080 }, { "grad_norm": 0.24824494123458862, "learning_rate": 7.993631980800285e-05, "loss": 0.032, "step": 33090 }, { "grad_norm": 0.25421029329299927, "learning_rate": 7.992307464497659e-05, "loss": 0.0287, "step": 33100 }, { "grad_norm": 0.22540336847305298, "learning_rate": 7.990982620960787e-05, "loss": 0.0285, "step": 33110 }, { "grad_norm": 0.24089710414409637, "learning_rate": 7.989657450334554e-05, "loss": 0.0248, "step": 33120 }, { "grad_norm": 0.23111212253570557, "learning_rate": 7.988331952763877e-05, "loss": 0.0308, "step": 33130 }, { "grad_norm": 0.17492899298667908, "learning_rate": 7.987006128393709e-05, "loss": 0.027, "step": 33140 }, { "grad_norm": 0.17136861383914948, "learning_rate": 7.985679977369043e-05, "loss": 0.0277, "step": 33150 }, { "grad_norm": 0.24609048664569855, "learning_rate": 7.984353499834902e-05, "loss": 0.0224, "step": 33160 }, { "grad_norm": 0.21621862053871155, "learning_rate": 7.983026695936351e-05, "loss": 0.0269, "step": 33170 }, { "grad_norm": 0.1563490331172943, "learning_rate": 7.981699565818486e-05, "loss": 0.0294, "step": 33180 }, { "grad_norm": 0.24286527931690216, "learning_rate": 7.980372109626437e-05, "loss": 0.0285, "step": 33190 }, { "grad_norm": 0.36179807782173157, "learning_rate": 7.979044327505375e-05, "loss": 0.0275, "step": 33200 }, { "grad_norm": 0.2954281270503998, "learning_rate": 7.977716219600506e-05, "loss": 0.0265, "step": 33210 }, { "grad_norm": 0.2427620142698288, "learning_rate": 7.97638778605707e-05, "loss": 0.026, "step": 33220 }, { "grad_norm": 0.22728413343429565, "learning_rate": 7.975059027020338e-05, "loss": 0.0267, "step": 33230 }, { "grad_norm": 0.2466949075460434, "learning_rate": 7.973729942635623e-05, "loss": 0.0302, "step": 33240 }, { "grad_norm": 0.2609292268753052, "learning_rate": 7.972400533048273e-05, "loss": 0.0326, "step": 33250 }, { "grad_norm": 0.1998962163925171, "learning_rate": 7.97107079840367e-05, "loss": 0.0281, "step": 33260 }, { "grad_norm": 0.2462748885154724, "learning_rate": 7.969740738847231e-05, "loss": 0.0328, "step": 33270 }, { "grad_norm": 0.30010849237442017, "learning_rate": 7.968410354524411e-05, "loss": 0.0333, "step": 33280 }, { "grad_norm": 0.33789387345314026, "learning_rate": 7.967079645580697e-05, "loss": 0.0429, "step": 33290 }, { "grad_norm": 0.21596698462963104, "learning_rate": 7.965748612161612e-05, "loss": 0.0356, "step": 33300 }, { "grad_norm": 0.19366754591464996, "learning_rate": 7.96441725441272e-05, "loss": 0.0322, "step": 33310 }, { "grad_norm": 0.2582774758338928, "learning_rate": 7.963085572479614e-05, "loss": 0.0261, "step": 33320 }, { "grad_norm": 0.2390982210636139, "learning_rate": 7.961753566507924e-05, "loss": 0.0288, "step": 33330 }, { "grad_norm": 0.2502402067184448, "learning_rate": 7.960421236643316e-05, "loss": 0.0307, "step": 33340 }, { "grad_norm": 0.16889870166778564, "learning_rate": 7.959088583031496e-05, "loss": 0.0259, "step": 33350 }, { "grad_norm": 0.2236848622560501, "learning_rate": 7.957755605818194e-05, "loss": 0.0307, "step": 33360 }, { "grad_norm": 0.30288368463516235, "learning_rate": 7.956422305149185e-05, "loss": 0.026, "step": 33370 }, { "grad_norm": 0.23729872703552246, "learning_rate": 7.95508868117028e-05, "loss": 0.0286, "step": 33380 }, { "grad_norm": 0.23829606175422668, "learning_rate": 7.953754734027318e-05, "loss": 0.0252, "step": 33390 }, { "grad_norm": 0.25902822613716125, "learning_rate": 7.952420463866182e-05, "loss": 0.0274, "step": 33400 }, { "grad_norm": 0.3082287907600403, "learning_rate": 7.951085870832782e-05, "loss": 0.0273, "step": 33410 }, { "grad_norm": 0.23400568962097168, "learning_rate": 7.949750955073067e-05, "loss": 0.0293, "step": 33420 }, { "grad_norm": 0.3362292945384979, "learning_rate": 7.948415716733022e-05, "loss": 0.0296, "step": 33430 }, { "grad_norm": 0.2432529628276825, "learning_rate": 7.947080155958669e-05, "loss": 0.0297, "step": 33440 }, { "grad_norm": 0.18446384370326996, "learning_rate": 7.94574427289606e-05, "loss": 0.0262, "step": 33450 }, { "grad_norm": 0.454490065574646, "learning_rate": 7.944408067691284e-05, "loss": 0.0274, "step": 33460 }, { "grad_norm": 0.2528035044670105, "learning_rate": 7.943071540490473e-05, "loss": 0.0294, "step": 33470 }, { "grad_norm": 0.41097983717918396, "learning_rate": 7.94173469143978e-05, "loss": 0.0316, "step": 33480 }, { "grad_norm": 0.2168569713830948, "learning_rate": 7.940397520685406e-05, "loss": 0.0301, "step": 33490 }, { "grad_norm": 0.17556385695934296, "learning_rate": 7.939060028373577e-05, "loss": 0.0267, "step": 33500 }, { "grad_norm": 0.15151013433933258, "learning_rate": 7.937722214650565e-05, "loss": 0.0268, "step": 33510 }, { "grad_norm": 0.2602732479572296, "learning_rate": 7.936384079662666e-05, "loss": 0.0253, "step": 33520 }, { "grad_norm": 0.23690727353096008, "learning_rate": 7.93504562355622e-05, "loss": 0.0309, "step": 33530 }, { "grad_norm": 0.18960539996623993, "learning_rate": 7.933706846477599e-05, "loss": 0.0256, "step": 33540 }, { "grad_norm": 0.32141605019569397, "learning_rate": 7.932367748573206e-05, "loss": 0.0294, "step": 33550 }, { "grad_norm": 0.3029852509498596, "learning_rate": 7.931028329989485e-05, "loss": 0.0263, "step": 33560 }, { "grad_norm": 0.16340288519859314, "learning_rate": 7.929688590872913e-05, "loss": 0.0243, "step": 33570 }, { "grad_norm": 0.2952536344528198, "learning_rate": 7.928348531370003e-05, "loss": 0.0334, "step": 33580 }, { "grad_norm": 0.22982379794120789, "learning_rate": 7.927008151627297e-05, "loss": 0.036, "step": 33590 }, { "grad_norm": 0.24164490401744843, "learning_rate": 7.925667451791383e-05, "loss": 0.0297, "step": 33600 }, { "grad_norm": 0.2141484022140503, "learning_rate": 7.924326432008874e-05, "loss": 0.0249, "step": 33610 }, { "grad_norm": 0.22630243003368378, "learning_rate": 7.922985092426422e-05, "loss": 0.0277, "step": 33620 }, { "grad_norm": 0.2261301428079605, "learning_rate": 7.921643433190717e-05, "loss": 0.0283, "step": 33630 }, { "grad_norm": 0.24521678686141968, "learning_rate": 7.920301454448478e-05, "loss": 0.0317, "step": 33640 }, { "grad_norm": 0.33416223526000977, "learning_rate": 7.918959156346461e-05, "loss": 0.027, "step": 33650 }, { "grad_norm": 0.3536335527896881, "learning_rate": 7.91761653903146e-05, "loss": 0.0342, "step": 33660 }, { "grad_norm": 0.2652643918991089, "learning_rate": 7.916273602650302e-05, "loss": 0.032, "step": 33670 }, { "grad_norm": 0.3288285434246063, "learning_rate": 7.914930347349847e-05, "loss": 0.0347, "step": 33680 }, { "grad_norm": 0.2945712208747864, "learning_rate": 7.913586773276992e-05, "loss": 0.0404, "step": 33690 }, { "grad_norm": 0.3900347352027893, "learning_rate": 7.912242880578667e-05, "loss": 0.0341, "step": 33700 }, { "grad_norm": 0.22570204734802246, "learning_rate": 7.910898669401839e-05, "loss": 0.0274, "step": 33710 }, { "grad_norm": 0.20995420217514038, "learning_rate": 7.909554139893511e-05, "loss": 0.0274, "step": 33720 }, { "grad_norm": 0.16317066550254822, "learning_rate": 7.908209292200715e-05, "loss": 0.0263, "step": 33730 }, { "grad_norm": 0.24280256032943726, "learning_rate": 7.906864126470523e-05, "loss": 0.0249, "step": 33740 }, { "grad_norm": 0.46079161763191223, "learning_rate": 7.905518642850041e-05, "loss": 0.0235, "step": 33750 }, { "grad_norm": 0.2309499830007553, "learning_rate": 7.904172841486409e-05, "loss": 0.0271, "step": 33760 }, { "grad_norm": 0.24758832156658173, "learning_rate": 7.902826722526801e-05, "loss": 0.0311, "step": 33770 }, { "grad_norm": 0.16622969508171082, "learning_rate": 7.901480286118427e-05, "loss": 0.0213, "step": 33780 }, { "grad_norm": 0.21348369121551514, "learning_rate": 7.900133532408531e-05, "loss": 0.0262, "step": 33790 }, { "grad_norm": 0.27258580923080444, "learning_rate": 7.898786461544395e-05, "loss": 0.0246, "step": 33800 }, { "grad_norm": 0.3020140528678894, "learning_rate": 7.897439073673325e-05, "loss": 0.0227, "step": 33810 }, { "grad_norm": 0.231680765748024, "learning_rate": 7.896091368942677e-05, "loss": 0.0276, "step": 33820 }, { "grad_norm": 0.24736730754375458, "learning_rate": 7.894743347499832e-05, "loss": 0.0312, "step": 33830 }, { "grad_norm": 0.27520379424095154, "learning_rate": 7.893395009492203e-05, "loss": 0.0265, "step": 33840 }, { "grad_norm": 0.22126851975917816, "learning_rate": 7.892046355067248e-05, "loss": 0.0288, "step": 33850 }, { "grad_norm": 0.2537693381309509, "learning_rate": 7.890697384372451e-05, "loss": 0.026, "step": 33860 }, { "grad_norm": 0.2155495434999466, "learning_rate": 7.889348097555336e-05, "loss": 0.0246, "step": 33870 }, { "grad_norm": 0.19575592875480652, "learning_rate": 7.887998494763455e-05, "loss": 0.0252, "step": 33880 }, { "grad_norm": 0.2508156895637512, "learning_rate": 7.886648576144404e-05, "loss": 0.027, "step": 33890 }, { "grad_norm": 0.29018110036849976, "learning_rate": 7.885298341845802e-05, "loss": 0.0338, "step": 33900 }, { "grad_norm": 0.3651478588581085, "learning_rate": 7.883947792015311e-05, "loss": 0.0281, "step": 33910 }, { "grad_norm": 0.22434179484844208, "learning_rate": 7.882596926800628e-05, "loss": 0.0328, "step": 33920 }, { "grad_norm": 0.21062946319580078, "learning_rate": 7.881245746349477e-05, "loss": 0.0324, "step": 33930 }, { "grad_norm": 0.24329102039337158, "learning_rate": 7.879894250809623e-05, "loss": 0.0287, "step": 33940 }, { "grad_norm": 0.188398540019989, "learning_rate": 7.878542440328865e-05, "loss": 0.0335, "step": 33950 }, { "grad_norm": 0.2079795002937317, "learning_rate": 7.877190315055031e-05, "loss": 0.0259, "step": 33960 }, { "grad_norm": 0.2124105989933014, "learning_rate": 7.875837875135991e-05, "loss": 0.0263, "step": 33970 }, { "grad_norm": 0.2279079109430313, "learning_rate": 7.874485120719646e-05, "loss": 0.0223, "step": 33980 }, { "grad_norm": 0.27733686566352844, "learning_rate": 7.873132051953928e-05, "loss": 0.028, "step": 33990 }, { "grad_norm": 0.35880762338638306, "learning_rate": 7.87177866898681e-05, "loss": 0.0383, "step": 34000 }, { "grad_norm": 0.2770739197731018, "learning_rate": 7.870424971966294e-05, "loss": 0.0364, "step": 34010 }, { "grad_norm": 0.31734541058540344, "learning_rate": 7.869070961040419e-05, "loss": 0.0315, "step": 34020 }, { "grad_norm": 0.2545628845691681, "learning_rate": 7.867716636357257e-05, "loss": 0.0287, "step": 34030 }, { "grad_norm": 0.205474853515625, "learning_rate": 7.866361998064915e-05, "loss": 0.027, "step": 34040 }, { "grad_norm": 0.2770600914955139, "learning_rate": 7.865007046311534e-05, "loss": 0.0263, "step": 34050 }, { "grad_norm": 0.23864229023456573, "learning_rate": 7.86365178124529e-05, "loss": 0.0247, "step": 34060 }, { "grad_norm": 0.2502285838127136, "learning_rate": 7.862296203014394e-05, "loss": 0.0277, "step": 34070 }, { "grad_norm": 0.257387638092041, "learning_rate": 7.860940311767088e-05, "loss": 0.0293, "step": 34080 }, { "grad_norm": 0.3062477707862854, "learning_rate": 7.85958410765165e-05, "loss": 0.0357, "step": 34090 }, { "grad_norm": 0.24063511192798615, "learning_rate": 7.858227590816394e-05, "loss": 0.0326, "step": 34100 }, { "grad_norm": 0.22900550067424774, "learning_rate": 7.856870761409664e-05, "loss": 0.0243, "step": 34110 }, { "grad_norm": 0.247144415974617, "learning_rate": 7.855513619579846e-05, "loss": 0.0275, "step": 34120 }, { "grad_norm": 0.20796655118465424, "learning_rate": 7.85415616547535e-05, "loss": 0.0304, "step": 34130 }, { "grad_norm": 0.21947459876537323, "learning_rate": 7.852798399244627e-05, "loss": 0.0256, "step": 34140 }, { "grad_norm": 0.17214195430278778, "learning_rate": 7.851440321036161e-05, "loss": 0.0239, "step": 34150 }, { "grad_norm": 0.2306337058544159, "learning_rate": 7.850081930998468e-05, "loss": 0.0273, "step": 34160 }, { "grad_norm": 0.19223903119564056, "learning_rate": 7.8487232292801e-05, "loss": 0.0304, "step": 34170 }, { "grad_norm": 0.20475247502326965, "learning_rate": 7.847364216029642e-05, "loss": 0.0259, "step": 34180 }, { "grad_norm": 0.25363999605178833, "learning_rate": 7.846004891395716e-05, "loss": 0.0229, "step": 34190 }, { "grad_norm": 0.18691310286521912, "learning_rate": 7.844645255526972e-05, "loss": 0.027, "step": 34200 }, { "grad_norm": 0.1956210434436798, "learning_rate": 7.843285308572101e-05, "loss": 0.0282, "step": 34210 }, { "grad_norm": 0.2507486045360565, "learning_rate": 7.841925050679823e-05, "loss": 0.0261, "step": 34220 }, { "grad_norm": 0.19809578359127045, "learning_rate": 7.840564481998895e-05, "loss": 0.0282, "step": 34230 }, { "grad_norm": 0.2094634771347046, "learning_rate": 7.839203602678105e-05, "loss": 0.0282, "step": 34240 }, { "grad_norm": 0.23705124855041504, "learning_rate": 7.837842412866279e-05, "loss": 0.0282, "step": 34250 }, { "grad_norm": 0.2193690836429596, "learning_rate": 7.836480912712272e-05, "loss": 0.0263, "step": 34260 }, { "grad_norm": 0.4761947691440582, "learning_rate": 7.835119102364976e-05, "loss": 0.0291, "step": 34270 }, { "grad_norm": 0.18660491704940796, "learning_rate": 7.83375698197332e-05, "loss": 0.0258, "step": 34280 }, { "grad_norm": 0.19504529237747192, "learning_rate": 7.83239455168626e-05, "loss": 0.0295, "step": 34290 }, { "grad_norm": 0.2109588086605072, "learning_rate": 7.83103181165279e-05, "loss": 0.0258, "step": 34300 }, { "grad_norm": 0.22981102764606476, "learning_rate": 7.829668762021937e-05, "loss": 0.0352, "step": 34310 }, { "grad_norm": 0.20910397171974182, "learning_rate": 7.828305402942764e-05, "loss": 0.027, "step": 34320 }, { "grad_norm": 0.3251906931400299, "learning_rate": 7.826941734564363e-05, "loss": 0.0218, "step": 34330 }, { "grad_norm": 0.2192108929157257, "learning_rate": 7.825577757035865e-05, "loss": 0.0354, "step": 34340 }, { "grad_norm": 0.24026568233966827, "learning_rate": 7.824213470506431e-05, "loss": 0.0275, "step": 34350 }, { "grad_norm": 0.27746814489364624, "learning_rate": 7.822848875125257e-05, "loss": 0.0264, "step": 34360 }, { "grad_norm": 0.2968478798866272, "learning_rate": 7.821483971041576e-05, "loss": 0.0251, "step": 34370 }, { "grad_norm": 0.2232314497232437, "learning_rate": 7.820118758404649e-05, "loss": 0.0313, "step": 34380 }, { "grad_norm": 0.21811117231845856, "learning_rate": 7.818753237363776e-05, "loss": 0.0266, "step": 34390 }, { "grad_norm": 0.267100065946579, "learning_rate": 7.817387408068286e-05, "loss": 0.0321, "step": 34400 }, { "grad_norm": 0.31837180256843567, "learning_rate": 7.816021270667544e-05, "loss": 0.0263, "step": 34410 }, { "grad_norm": 0.31011077761650085, "learning_rate": 7.81465482531095e-05, "loss": 0.0253, "step": 34420 }, { "grad_norm": 0.3063839375972748, "learning_rate": 7.813288072147938e-05, "loss": 0.0282, "step": 34430 }, { "grad_norm": 0.3038769066333771, "learning_rate": 7.811921011327972e-05, "loss": 0.0278, "step": 34440 }, { "grad_norm": 0.25009623169898987, "learning_rate": 7.810553643000549e-05, "loss": 0.0356, "step": 34450 }, { "grad_norm": 0.33424851298332214, "learning_rate": 7.809185967315206e-05, "loss": 0.0313, "step": 34460 }, { "grad_norm": 0.35492438077926636, "learning_rate": 7.80781798442151e-05, "loss": 0.035, "step": 34470 }, { "grad_norm": 0.738868236541748, "learning_rate": 7.806449694469061e-05, "loss": 0.0316, "step": 34480 }, { "grad_norm": 3.3715553283691406, "learning_rate": 7.805081097607492e-05, "loss": 0.0349, "step": 34490 }, { "grad_norm": 0.22112742066383362, "learning_rate": 7.803712193986474e-05, "loss": 0.0247, "step": 34500 }, { "grad_norm": 0.2437402606010437, "learning_rate": 7.802342983755702e-05, "loss": 0.0264, "step": 34510 }, { "grad_norm": 0.4878694415092468, "learning_rate": 7.800973467064918e-05, "loss": 0.0314, "step": 34520 }, { "grad_norm": 0.4192260801792145, "learning_rate": 7.799603644063884e-05, "loss": 0.0344, "step": 34530 }, { "grad_norm": 0.19845721125602722, "learning_rate": 7.798233514902405e-05, "loss": 0.0276, "step": 34540 }, { "grad_norm": 0.2952922284603119, "learning_rate": 7.796863079730318e-05, "loss": 0.0319, "step": 34550 }, { "grad_norm": 0.26225966215133667, "learning_rate": 7.795492338697488e-05, "loss": 0.0243, "step": 34560 }, { "grad_norm": 0.26180338859558105, "learning_rate": 7.794121291953819e-05, "loss": 0.0273, "step": 34570 }, { "grad_norm": 0.2225935310125351, "learning_rate": 7.792749939649246e-05, "loss": 0.0217, "step": 34580 }, { "grad_norm": 0.22971083223819733, "learning_rate": 7.79137828193374e-05, "loss": 0.0246, "step": 34590 }, { "grad_norm": 0.21323712170124054, "learning_rate": 7.790006318957301e-05, "loss": 0.0279, "step": 34600 }, { "grad_norm": 0.24948996305465698, "learning_rate": 7.788634050869965e-05, "loss": 0.0238, "step": 34610 }, { "grad_norm": 0.23406435549259186, "learning_rate": 7.787261477821803e-05, "loss": 0.0245, "step": 34620 }, { "grad_norm": 0.25976961851119995, "learning_rate": 7.785888599962916e-05, "loss": 0.0268, "step": 34630 }, { "grad_norm": 0.21204760670661926, "learning_rate": 7.784515417443439e-05, "loss": 0.0253, "step": 34640 }, { "grad_norm": 0.21922054886817932, "learning_rate": 7.783141930413545e-05, "loss": 0.0328, "step": 34650 }, { "grad_norm": 0.3268185555934906, "learning_rate": 7.78176813902343e-05, "loss": 0.0305, "step": 34660 }, { "grad_norm": 0.2906891405582428, "learning_rate": 7.780394043423336e-05, "loss": 0.0256, "step": 34670 }, { "grad_norm": 0.2415902465581894, "learning_rate": 7.77901964376353e-05, "loss": 0.0252, "step": 34680 }, { "grad_norm": 0.33300545811653137, "learning_rate": 7.777644940194316e-05, "loss": 0.0279, "step": 34690 }, { "grad_norm": 0.21400663256645203, "learning_rate": 7.776269932866023e-05, "loss": 0.0291, "step": 34700 }, { "grad_norm": 0.2831251323223114, "learning_rate": 7.774894621929026e-05, "loss": 0.0294, "step": 34710 }, { "grad_norm": 0.2595599293708801, "learning_rate": 7.773519007533725e-05, "loss": 0.024, "step": 34720 }, { "grad_norm": 0.23544451594352722, "learning_rate": 7.772143089830556e-05, "loss": 0.0259, "step": 34730 }, { "grad_norm": 0.2713226079940796, "learning_rate": 7.770766868969985e-05, "loss": 0.0275, "step": 34740 }, { "grad_norm": 0.2701118290424347, "learning_rate": 7.769390345102518e-05, "loss": 0.0263, "step": 34750 }, { "grad_norm": 0.2958100140094757, "learning_rate": 7.768013518378683e-05, "loss": 0.0236, "step": 34760 }, { "grad_norm": 0.21741117537021637, "learning_rate": 7.766636388949053e-05, "loss": 0.0288, "step": 34770 }, { "grad_norm": 0.3269101083278656, "learning_rate": 7.765258956964229e-05, "loss": 0.0261, "step": 34780 }, { "grad_norm": 0.20126663148403168, "learning_rate": 7.76388122257484e-05, "loss": 0.0257, "step": 34790 }, { "grad_norm": 0.4429716467857361, "learning_rate": 7.762503185931558e-05, "loss": 0.0282, "step": 34800 }, { "grad_norm": 0.34071698784828186, "learning_rate": 7.76112484718508e-05, "loss": 0.0283, "step": 34810 }, { "grad_norm": 0.1859736442565918, "learning_rate": 7.75974620648614e-05, "loss": 0.0269, "step": 34820 }, { "grad_norm": 0.2385382056236267, "learning_rate": 7.758367263985503e-05, "loss": 0.0271, "step": 34830 }, { "grad_norm": 0.25077518820762634, "learning_rate": 7.75698801983397e-05, "loss": 0.031, "step": 34840 }, { "grad_norm": 0.2476392686367035, "learning_rate": 7.755608474182372e-05, "loss": 0.0228, "step": 34850 }, { "grad_norm": 0.2115139365196228, "learning_rate": 7.754228627181574e-05, "loss": 0.0257, "step": 34860 }, { "grad_norm": 0.2123793214559555, "learning_rate": 7.752848478982476e-05, "loss": 0.0278, "step": 34870 }, { "grad_norm": 0.24242804944515228, "learning_rate": 7.751468029736006e-05, "loss": 0.0231, "step": 34880 }, { "grad_norm": 0.28528815507888794, "learning_rate": 7.750087279593129e-05, "loss": 0.0253, "step": 34890 }, { "grad_norm": 0.2932968735694885, "learning_rate": 7.748706228704843e-05, "loss": 0.0291, "step": 34900 }, { "grad_norm": 0.23949940502643585, "learning_rate": 7.747324877222176e-05, "loss": 0.0245, "step": 34910 }, { "grad_norm": 0.18869146704673767, "learning_rate": 7.745943225296188e-05, "loss": 0.0249, "step": 34920 }, { "grad_norm": 0.2437322735786438, "learning_rate": 7.744561273077981e-05, "loss": 0.0262, "step": 34930 }, { "grad_norm": 0.6280354857444763, "learning_rate": 7.743179020718678e-05, "loss": 0.0328, "step": 34940 }, { "grad_norm": 0.19773392379283905, "learning_rate": 7.741796468369443e-05, "loss": 0.0286, "step": 34950 }, { "grad_norm": 0.3722853660583496, "learning_rate": 7.740413616181466e-05, "loss": 0.0304, "step": 34960 }, { "grad_norm": 0.3928009569644928, "learning_rate": 7.739030464305978e-05, "loss": 0.0297, "step": 34970 }, { "grad_norm": 0.27973923087120056, "learning_rate": 7.737647012894235e-05, "loss": 0.0249, "step": 34980 }, { "grad_norm": 0.19128209352493286, "learning_rate": 7.736263262097532e-05, "loss": 0.0286, "step": 34990 }, { "grad_norm": 0.46004146337509155, "learning_rate": 7.734879212067192e-05, "loss": 0.0293, "step": 35000 }, { "grad_norm": 0.29495957493782043, "learning_rate": 7.733494862954573e-05, "loss": 0.0278, "step": 35010 }, { "grad_norm": 0.20426428318023682, "learning_rate": 7.732110214911066e-05, "loss": 0.0268, "step": 35020 }, { "grad_norm": 0.25962159037590027, "learning_rate": 7.730725268088094e-05, "loss": 0.023, "step": 35030 }, { "grad_norm": 0.5407551527023315, "learning_rate": 7.729340022637112e-05, "loss": 0.0259, "step": 35040 }, { "grad_norm": 0.2405913919210434, "learning_rate": 7.727954478709607e-05, "loss": 0.0295, "step": 35050 }, { "grad_norm": 0.2652616798877716, "learning_rate": 7.726568636457103e-05, "loss": 0.0267, "step": 35060 }, { "grad_norm": 0.2671838700771332, "learning_rate": 7.725182496031153e-05, "loss": 0.0311, "step": 35070 }, { "grad_norm": 0.1968432068824768, "learning_rate": 7.723796057583341e-05, "loss": 0.0257, "step": 35080 }, { "grad_norm": 0.18304698169231415, "learning_rate": 7.722409321265287e-05, "loss": 0.0261, "step": 35090 }, { "grad_norm": 0.2854052484035492, "learning_rate": 7.721022287228645e-05, "loss": 0.0268, "step": 35100 }, { "grad_norm": 0.5279983282089233, "learning_rate": 7.719634955625094e-05, "loss": 0.0292, "step": 35110 }, { "grad_norm": 0.3999039828777313, "learning_rate": 7.718247326606355e-05, "loss": 0.0337, "step": 35120 }, { "grad_norm": 0.28641051054000854, "learning_rate": 7.716859400324175e-05, "loss": 0.0282, "step": 35130 }, { "grad_norm": 0.2736269533634186, "learning_rate": 7.715471176930335e-05, "loss": 0.0295, "step": 35140 }, { "grad_norm": 0.3388817012310028, "learning_rate": 7.714082656576651e-05, "loss": 0.0254, "step": 35150 }, { "grad_norm": 0.29560115933418274, "learning_rate": 7.712693839414968e-05, "loss": 0.0307, "step": 35160 }, { "grad_norm": 0.2658950090408325, "learning_rate": 7.711304725597164e-05, "loss": 0.0308, "step": 35170 }, { "grad_norm": 0.23477324843406677, "learning_rate": 7.709915315275151e-05, "loss": 0.0278, "step": 35180 }, { "grad_norm": 0.1569388061761856, "learning_rate": 7.708525608600876e-05, "loss": 0.0259, "step": 35190 }, { "grad_norm": 0.25823143124580383, "learning_rate": 7.707135605726311e-05, "loss": 0.0283, "step": 35200 }, { "grad_norm": 0.2484837919473648, "learning_rate": 7.705745306803466e-05, "loss": 0.0298, "step": 35210 }, { "grad_norm": 0.22270992398262024, "learning_rate": 7.704354711984383e-05, "loss": 0.027, "step": 35220 }, { "grad_norm": 0.2222188264131546, "learning_rate": 7.702963821421133e-05, "loss": 0.0247, "step": 35230 }, { "grad_norm": 0.2973632216453552, "learning_rate": 7.701572635265826e-05, "loss": 0.0252, "step": 35240 }, { "grad_norm": 0.22546538710594177, "learning_rate": 7.700181153670596e-05, "loss": 0.0259, "step": 35250 }, { "grad_norm": 0.6216502785682678, "learning_rate": 7.698789376787616e-05, "loss": 0.0328, "step": 35260 }, { "grad_norm": 0.31504276394844055, "learning_rate": 7.697397304769085e-05, "loss": 0.0316, "step": 35270 }, { "grad_norm": 0.27835312485694885, "learning_rate": 7.696004937767241e-05, "loss": 0.0248, "step": 35280 }, { "grad_norm": 0.7726244926452637, "learning_rate": 7.694612275934352e-05, "loss": 0.0264, "step": 35290 }, { "grad_norm": 0.20817767083644867, "learning_rate": 7.693219319422714e-05, "loss": 0.022, "step": 35300 }, { "grad_norm": 0.18596597015857697, "learning_rate": 7.69182606838466e-05, "loss": 0.0256, "step": 35310 }, { "grad_norm": 0.31999167799949646, "learning_rate": 7.690432522972558e-05, "loss": 0.028, "step": 35320 }, { "grad_norm": 0.3237963616847992, "learning_rate": 7.689038683338796e-05, "loss": 0.0301, "step": 35330 }, { "grad_norm": 0.28900474309921265, "learning_rate": 7.687644549635808e-05, "loss": 0.0296, "step": 35340 }, { "grad_norm": 0.26759791374206543, "learning_rate": 7.686250122016053e-05, "loss": 0.0281, "step": 35350 }, { "grad_norm": 0.203399196267128, "learning_rate": 7.684855400632022e-05, "loss": 0.0284, "step": 35360 }, { "grad_norm": 0.21935857832431793, "learning_rate": 7.683460385636243e-05, "loss": 0.0246, "step": 35370 }, { "grad_norm": 0.2918994128704071, "learning_rate": 7.68206507718127e-05, "loss": 0.0276, "step": 35380 }, { "grad_norm": 0.28094482421875, "learning_rate": 7.680669475419693e-05, "loss": 0.0282, "step": 35390 }, { "grad_norm": 0.2171608954668045, "learning_rate": 7.679273580504132e-05, "loss": 0.0268, "step": 35400 }, { "grad_norm": 0.3209032714366913, "learning_rate": 7.67787739258724e-05, "loss": 0.0263, "step": 35410 }, { "grad_norm": 0.3219570219516754, "learning_rate": 7.676480911821705e-05, "loss": 0.0238, "step": 35420 }, { "grad_norm": 0.24227620661258698, "learning_rate": 7.675084138360238e-05, "loss": 0.026, "step": 35430 }, { "grad_norm": 0.33116552233695984, "learning_rate": 7.673687072355592e-05, "loss": 0.0341, "step": 35440 }, { "grad_norm": 0.2012552171945572, "learning_rate": 7.67228971396055e-05, "loss": 0.0278, "step": 35450 }, { "grad_norm": 0.4040912389755249, "learning_rate": 7.670892063327922e-05, "loss": 0.0275, "step": 35460 }, { "grad_norm": 0.22184115648269653, "learning_rate": 7.669494120610552e-05, "loss": 0.03, "step": 35470 }, { "grad_norm": 0.2909298539161682, "learning_rate": 7.66809588596132e-05, "loss": 0.0323, "step": 35480 }, { "grad_norm": 0.22449742257595062, "learning_rate": 7.666697359533132e-05, "loss": 0.0272, "step": 35490 }, { "grad_norm": 0.31914061307907104, "learning_rate": 7.665298541478932e-05, "loss": 0.0259, "step": 35500 }, { "grad_norm": 0.29045790433883667, "learning_rate": 7.66389943195169e-05, "loss": 0.0309, "step": 35510 }, { "grad_norm": 0.32519203424453735, "learning_rate": 7.66250003110441e-05, "loss": 0.0327, "step": 35520 }, { "grad_norm": 0.25146034359931946, "learning_rate": 7.66110033909013e-05, "loss": 0.0274, "step": 35530 }, { "grad_norm": 0.20354393124580383, "learning_rate": 7.659700356061918e-05, "loss": 0.0303, "step": 35540 }, { "grad_norm": 0.3355584144592285, "learning_rate": 7.658300082172875e-05, "loss": 0.026, "step": 35550 }, { "grad_norm": 0.43745970726013184, "learning_rate": 7.65689951757613e-05, "loss": 0.0275, "step": 35560 }, { "grad_norm": 0.24611100554466248, "learning_rate": 7.655498662424849e-05, "loss": 0.0245, "step": 35570 }, { "grad_norm": 0.24571099877357483, "learning_rate": 7.654097516872227e-05, "loss": 0.0263, "step": 35580 }, { "grad_norm": 0.21635353565216064, "learning_rate": 7.65269608107149e-05, "loss": 0.0275, "step": 35590 }, { "grad_norm": 0.20665916800498962, "learning_rate": 7.6512943551759e-05, "loss": 0.0295, "step": 35600 }, { "grad_norm": 0.2469642013311386, "learning_rate": 7.649892339338743e-05, "loss": 0.0291, "step": 35610 }, { "grad_norm": 0.22251927852630615, "learning_rate": 7.648490033713344e-05, "loss": 0.0281, "step": 35620 }, { "grad_norm": 0.2093934267759323, "learning_rate": 7.647087438453058e-05, "loss": 0.0253, "step": 35630 }, { "grad_norm": 0.33758679032325745, "learning_rate": 7.64568455371127e-05, "loss": 0.0284, "step": 35640 }, { "grad_norm": 0.21927876770496368, "learning_rate": 7.644281379641396e-05, "loss": 0.0248, "step": 35650 }, { "grad_norm": 0.26151102781295776, "learning_rate": 7.642877916396887e-05, "loss": 0.0235, "step": 35660 }, { "grad_norm": 0.4507291316986084, "learning_rate": 7.641474164131221e-05, "loss": 0.0297, "step": 35670 }, { "grad_norm": 0.15832407772541046, "learning_rate": 7.640070122997913e-05, "loss": 0.0254, "step": 35680 }, { "grad_norm": 0.2713509202003479, "learning_rate": 7.638665793150505e-05, "loss": 0.0246, "step": 35690 }, { "grad_norm": 0.2908153831958771, "learning_rate": 7.637261174742574e-05, "loss": 0.0246, "step": 35700 }, { "grad_norm": 0.20533528923988342, "learning_rate": 7.635856267927725e-05, "loss": 0.0277, "step": 35710 }, { "grad_norm": 0.23904316127300262, "learning_rate": 7.634451072859597e-05, "loss": 0.0266, "step": 35720 }, { "grad_norm": 0.2320130616426468, "learning_rate": 7.633045589691863e-05, "loss": 0.0279, "step": 35730 }, { "grad_norm": 0.33277803659439087, "learning_rate": 7.63163981857822e-05, "loss": 0.0303, "step": 35740 }, { "grad_norm": 0.20813848078250885, "learning_rate": 7.630233759672403e-05, "loss": 0.0241, "step": 35750 }, { "grad_norm": 0.2265772521495819, "learning_rate": 7.628827413128177e-05, "loss": 0.0269, "step": 35760 }, { "grad_norm": 0.24690832197666168, "learning_rate": 7.627420779099338e-05, "loss": 0.0252, "step": 35770 }, { "grad_norm": 0.24523894488811493, "learning_rate": 7.626013857739711e-05, "loss": 0.028, "step": 35780 }, { "grad_norm": 0.2604420781135559, "learning_rate": 7.624606649203158e-05, "loss": 0.0264, "step": 35790 }, { "grad_norm": 0.2898944616317749, "learning_rate": 7.623199153643569e-05, "loss": 0.0274, "step": 35800 }, { "grad_norm": 0.4666491746902466, "learning_rate": 7.621791371214863e-05, "loss": 0.028, "step": 35810 }, { "grad_norm": 0.23426896333694458, "learning_rate": 7.620383302070995e-05, "loss": 0.0294, "step": 35820 }, { "grad_norm": 0.1937839835882187, "learning_rate": 7.61897494636595e-05, "loss": 0.0247, "step": 35830 }, { "grad_norm": 0.277731329202652, "learning_rate": 7.617566304253739e-05, "loss": 0.0301, "step": 35840 }, { "grad_norm": 0.18192845582962036, "learning_rate": 7.616157375888416e-05, "loss": 0.0273, "step": 35850 }, { "grad_norm": 0.19707421958446503, "learning_rate": 7.614748161424053e-05, "loss": 0.0302, "step": 35860 }, { "grad_norm": 0.27832815051078796, "learning_rate": 7.613338661014763e-05, "loss": 0.0316, "step": 35870 }, { "grad_norm": 0.26370662450790405, "learning_rate": 7.611928874814686e-05, "loss": 0.03, "step": 35880 }, { "grad_norm": 0.26073071360588074, "learning_rate": 7.610518802977996e-05, "loss": 0.0247, "step": 35890 }, { "grad_norm": 0.2371683418750763, "learning_rate": 7.609108445658893e-05, "loss": 0.0229, "step": 35900 }, { "grad_norm": 0.22542120516300201, "learning_rate": 7.607697803011612e-05, "loss": 0.0271, "step": 35910 }, { "grad_norm": 0.23942551016807556, "learning_rate": 7.606286875190421e-05, "loss": 0.0401, "step": 35920 }, { "grad_norm": 0.1598302572965622, "learning_rate": 7.604875662349617e-05, "loss": 0.0269, "step": 35930 }, { "grad_norm": 0.19485077261924744, "learning_rate": 7.603464164643525e-05, "loss": 0.0223, "step": 35940 }, { "grad_norm": 0.22938226163387299, "learning_rate": 7.602052382226507e-05, "loss": 0.0263, "step": 35950 }, { "grad_norm": 0.4567902684211731, "learning_rate": 7.600640315252954e-05, "loss": 0.0262, "step": 35960 }, { "grad_norm": 0.26431554555892944, "learning_rate": 7.599227963877284e-05, "loss": 0.0265, "step": 35970 }, { "grad_norm": 0.21009604632854462, "learning_rate": 7.597815328253954e-05, "loss": 0.029, "step": 35980 }, { "grad_norm": 0.23947016894817352, "learning_rate": 7.596402408537443e-05, "loss": 0.0238, "step": 35990 }, { "grad_norm": 0.2433278113603592, "learning_rate": 7.59498920488227e-05, "loss": 0.0229, "step": 36000 }, { "grad_norm": 0.30145031213760376, "learning_rate": 7.593575717442979e-05, "loss": 0.0258, "step": 36010 }, { "grad_norm": 0.20896285772323608, "learning_rate": 7.592161946374147e-05, "loss": 0.0263, "step": 36020 }, { "grad_norm": 0.21493913233280182, "learning_rate": 7.590747891830381e-05, "loss": 0.0251, "step": 36030 }, { "grad_norm": 0.1946171373128891, "learning_rate": 7.58933355396632e-05, "loss": 0.0281, "step": 36040 }, { "grad_norm": 0.2507425546646118, "learning_rate": 7.587918932936636e-05, "loss": 0.0192, "step": 36050 }, { "grad_norm": 0.22170327603816986, "learning_rate": 7.586504028896028e-05, "loss": 0.0296, "step": 36060 }, { "grad_norm": 0.26003381609916687, "learning_rate": 7.585088841999228e-05, "loss": 0.0238, "step": 36070 }, { "grad_norm": 0.2813134789466858, "learning_rate": 7.583673372400999e-05, "loss": 0.0245, "step": 36080 }, { "grad_norm": 0.2528923451900482, "learning_rate": 7.582257620256134e-05, "loss": 0.0297, "step": 36090 }, { "grad_norm": 0.20451003313064575, "learning_rate": 7.580841585719458e-05, "loss": 0.0258, "step": 36100 }, { "grad_norm": 0.21170908212661743, "learning_rate": 7.579425268945825e-05, "loss": 0.0283, "step": 36110 }, { "grad_norm": 0.1575460135936737, "learning_rate": 7.578008670090127e-05, "loss": 0.0266, "step": 36120 }, { "grad_norm": 0.30262506008148193, "learning_rate": 7.576591789307272e-05, "loss": 0.0345, "step": 36130 }, { "grad_norm": 0.4360332489013672, "learning_rate": 7.575174626752216e-05, "loss": 0.03, "step": 36140 }, { "grad_norm": 0.6903809905052185, "learning_rate": 7.573757182579934e-05, "loss": 0.0301, "step": 36150 }, { "grad_norm": 0.23527826368808746, "learning_rate": 7.572339456945435e-05, "loss": 0.0283, "step": 36160 }, { "grad_norm": 0.2152324765920639, "learning_rate": 7.570921450003762e-05, "loss": 0.0306, "step": 36170 }, { "grad_norm": 0.2549552023410797, "learning_rate": 7.569503161909983e-05, "loss": 0.0316, "step": 36180 }, { "grad_norm": 1.1106256246566772, "learning_rate": 7.568084592819202e-05, "loss": 0.0324, "step": 36190 }, { "grad_norm": 0.27356868982315063, "learning_rate": 7.566665742886551e-05, "loss": 0.0294, "step": 36200 }, { "grad_norm": 0.2532089054584503, "learning_rate": 7.565246612267194e-05, "loss": 0.0281, "step": 36210 }, { "grad_norm": 0.23574627935886383, "learning_rate": 7.563827201116325e-05, "loss": 0.0244, "step": 36220 }, { "grad_norm": 0.30771610140800476, "learning_rate": 7.562407509589164e-05, "loss": 0.0314, "step": 36230 }, { "grad_norm": 0.16329191625118256, "learning_rate": 7.560987537840973e-05, "loss": 0.0254, "step": 36240 }, { "grad_norm": 0.2067585587501526, "learning_rate": 7.559567286027036e-05, "loss": 0.0293, "step": 36250 }, { "grad_norm": 0.1846630573272705, "learning_rate": 7.558146754302668e-05, "loss": 0.0253, "step": 36260 }, { "grad_norm": 0.17228035628795624, "learning_rate": 7.556725942823217e-05, "loss": 0.0236, "step": 36270 }, { "grad_norm": 0.31818094849586487, "learning_rate": 7.555304851744061e-05, "loss": 0.0219, "step": 36280 }, { "grad_norm": 0.23808448016643524, "learning_rate": 7.553883481220608e-05, "loss": 0.0247, "step": 36290 }, { "grad_norm": 0.7203068733215332, "learning_rate": 7.552461831408298e-05, "loss": 0.0251, "step": 36300 }, { "grad_norm": 0.2580442428588867, "learning_rate": 7.551039902462599e-05, "loss": 0.028, "step": 36310 }, { "grad_norm": 0.26868677139282227, "learning_rate": 7.549617694539011e-05, "loss": 0.0299, "step": 36320 }, { "grad_norm": 0.44450101256370544, "learning_rate": 7.548195207793066e-05, "loss": 0.0322, "step": 36330 }, { "grad_norm": 0.23424334824085236, "learning_rate": 7.546772442380323e-05, "loss": 0.031, "step": 36340 }, { "grad_norm": 0.45050084590911865, "learning_rate": 7.545349398456376e-05, "loss": 0.0335, "step": 36350 }, { "grad_norm": 0.2899732291698456, "learning_rate": 7.543926076176845e-05, "loss": 0.0281, "step": 36360 }, { "grad_norm": 0.28319498896598816, "learning_rate": 7.542502475697385e-05, "loss": 0.0261, "step": 36370 }, { "grad_norm": 0.3496146500110626, "learning_rate": 7.541078597173675e-05, "loss": 0.0313, "step": 36380 }, { "grad_norm": 0.2141355723142624, "learning_rate": 7.539654440761431e-05, "loss": 0.0266, "step": 36390 }, { "grad_norm": 0.29195457696914673, "learning_rate": 7.538230006616395e-05, "loss": 0.0304, "step": 36400 }, { "grad_norm": 0.19971241056919098, "learning_rate": 7.536805294894342e-05, "loss": 0.0243, "step": 36410 }, { "grad_norm": 0.4280199110507965, "learning_rate": 7.535380305751076e-05, "loss": 0.0284, "step": 36420 }, { "grad_norm": 0.2501145303249359, "learning_rate": 7.533955039342431e-05, "loss": 0.0249, "step": 36430 }, { "grad_norm": 0.2659464180469513, "learning_rate": 7.532529495824274e-05, "loss": 0.0236, "step": 36440 }, { "grad_norm": 0.23549245297908783, "learning_rate": 7.531103675352497e-05, "loss": 0.0322, "step": 36450 }, { "grad_norm": 0.3412722647190094, "learning_rate": 7.529677578083029e-05, "loss": 0.0279, "step": 36460 }, { "grad_norm": 0.23604175448417664, "learning_rate": 7.528251204171823e-05, "loss": 0.0252, "step": 36470 }, { "grad_norm": 0.2704279124736786, "learning_rate": 7.526824553774867e-05, "loss": 0.0296, "step": 36480 }, { "grad_norm": 0.2564781606197357, "learning_rate": 7.525397627048178e-05, "loss": 0.0307, "step": 36490 }, { "grad_norm": 0.23648224771022797, "learning_rate": 7.523970424147802e-05, "loss": 0.0273, "step": 36500 }, { "grad_norm": 0.32382428646087646, "learning_rate": 7.522542945229813e-05, "loss": 0.0321, "step": 36510 }, { "grad_norm": 0.2591918706893921, "learning_rate": 7.521115190450321e-05, "loss": 0.0336, "step": 36520 }, { "grad_norm": 0.2729097008705139, "learning_rate": 7.519687159965462e-05, "loss": 0.0296, "step": 36530 }, { "grad_norm": 0.2167416214942932, "learning_rate": 7.518258853931403e-05, "loss": 0.0241, "step": 36540 }, { "grad_norm": 0.22459128499031067, "learning_rate": 7.516830272504342e-05, "loss": 0.0241, "step": 36550 }, { "grad_norm": 0.27067580819129944, "learning_rate": 7.515401415840509e-05, "loss": 0.0272, "step": 36560 }, { "grad_norm": 0.2631478011608124, "learning_rate": 7.513972284096155e-05, "loss": 0.0248, "step": 36570 }, { "grad_norm": 0.19097478687763214, "learning_rate": 7.512542877427574e-05, "loss": 0.0284, "step": 36580 }, { "grad_norm": 0.2445133924484253, "learning_rate": 7.51111319599108e-05, "loss": 0.0276, "step": 36590 }, { "grad_norm": 0.3158731758594513, "learning_rate": 7.509683239943024e-05, "loss": 0.0266, "step": 36600 }, { "grad_norm": 0.341460257768631, "learning_rate": 7.50825300943978e-05, "loss": 0.0386, "step": 36610 }, { "grad_norm": 0.2507258653640747, "learning_rate": 7.50682250463776e-05, "loss": 0.0291, "step": 36620 }, { "grad_norm": 0.4218408465385437, "learning_rate": 7.5053917256934e-05, "loss": 0.0268, "step": 36630 }, { "grad_norm": 0.647028923034668, "learning_rate": 7.503960672763166e-05, "loss": 0.0275, "step": 36640 }, { "grad_norm": 0.20512914657592773, "learning_rate": 7.502529346003559e-05, "loss": 0.0289, "step": 36650 }, { "grad_norm": 0.2086680382490158, "learning_rate": 7.501097745571106e-05, "loss": 0.0326, "step": 36660 }, { "grad_norm": 0.2015814185142517, "learning_rate": 7.499665871622361e-05, "loss": 0.0282, "step": 36670 }, { "grad_norm": 0.21767500042915344, "learning_rate": 7.498233724313917e-05, "loss": 0.0258, "step": 36680 }, { "grad_norm": 0.2654217779636383, "learning_rate": 7.496801303802387e-05, "loss": 0.0299, "step": 36690 }, { "grad_norm": 0.49340131878852844, "learning_rate": 7.495368610244423e-05, "loss": 0.0307, "step": 36700 }, { "grad_norm": 0.24850137531757355, "learning_rate": 7.493935643796697e-05, "loss": 0.0259, "step": 36710 }, { "grad_norm": 0.2508627474308014, "learning_rate": 7.49250240461592e-05, "loss": 0.0262, "step": 36720 }, { "grad_norm": 0.297222763299942, "learning_rate": 7.491068892858826e-05, "loss": 0.0255, "step": 36730 }, { "grad_norm": 0.21396921575069427, "learning_rate": 7.489635108682184e-05, "loss": 0.0263, "step": 36740 }, { "grad_norm": 0.30374956130981445, "learning_rate": 7.48820105224279e-05, "loss": 0.0256, "step": 36750 }, { "grad_norm": 0.23782940208911896, "learning_rate": 7.48676672369747e-05, "loss": 0.0254, "step": 36760 }, { "grad_norm": 0.20295937359333038, "learning_rate": 7.485332123203079e-05, "loss": 0.0257, "step": 36770 }, { "grad_norm": 0.22298642992973328, "learning_rate": 7.483897250916504e-05, "loss": 0.0256, "step": 36780 }, { "grad_norm": 0.2674590051174164, "learning_rate": 7.48246210699466e-05, "loss": 0.0254, "step": 36790 }, { "grad_norm": 0.24008212983608246, "learning_rate": 7.481026691594492e-05, "loss": 0.0304, "step": 36800 }, { "grad_norm": 0.24292252957820892, "learning_rate": 7.479591004872975e-05, "loss": 0.023, "step": 36810 }, { "grad_norm": 0.4144851565361023, "learning_rate": 7.478155046987114e-05, "loss": 0.0251, "step": 36820 }, { "grad_norm": 0.5590366125106812, "learning_rate": 7.476718818093944e-05, "loss": 0.0234, "step": 36830 }, { "grad_norm": 0.3152758777141571, "learning_rate": 7.475282318350527e-05, "loss": 0.0287, "step": 36840 }, { "grad_norm": 0.20246228575706482, "learning_rate": 7.473845547913957e-05, "loss": 0.0277, "step": 36850 }, { "grad_norm": 0.22888974845409393, "learning_rate": 7.472408506941357e-05, "loss": 0.0295, "step": 36860 }, { "grad_norm": 0.41434088349342346, "learning_rate": 7.470971195589881e-05, "loss": 0.0309, "step": 36870 }, { "grad_norm": 0.27058443427085876, "learning_rate": 7.46953361401671e-05, "loss": 0.0262, "step": 36880 }, { "grad_norm": 0.4533992409706116, "learning_rate": 7.468095762379055e-05, "loss": 0.0298, "step": 36890 }, { "grad_norm": 0.24167205393314362, "learning_rate": 7.466657640834158e-05, "loss": 0.026, "step": 36900 }, { "grad_norm": 0.29348331689834595, "learning_rate": 7.46521924953929e-05, "loss": 0.0246, "step": 36910 }, { "grad_norm": 0.24811026453971863, "learning_rate": 7.463780588651752e-05, "loss": 0.0259, "step": 36920 }, { "grad_norm": 0.2547079920768738, "learning_rate": 7.462341658328873e-05, "loss": 0.0263, "step": 36930 }, { "grad_norm": 0.20237822830677032, "learning_rate": 7.460902458728012e-05, "loss": 0.0264, "step": 36940 }, { "grad_norm": 0.18465881049633026, "learning_rate": 7.459462990006558e-05, "loss": 0.0226, "step": 36950 }, { "grad_norm": 0.2100450098514557, "learning_rate": 7.458023252321929e-05, "loss": 0.0238, "step": 36960 }, { "grad_norm": 0.2860022783279419, "learning_rate": 7.456583245831573e-05, "loss": 0.0311, "step": 36970 }, { "grad_norm": 0.21615296602249146, "learning_rate": 7.455142970692968e-05, "loss": 0.0246, "step": 36980 }, { "grad_norm": 0.2277778834104538, "learning_rate": 7.453702427063615e-05, "loss": 0.0241, "step": 36990 }, { "grad_norm": 0.25597238540649414, "learning_rate": 7.452261615101057e-05, "loss": 0.0297, "step": 37000 }, { "grad_norm": 0.19803030788898468, "learning_rate": 7.450820534962856e-05, "loss": 0.0257, "step": 37010 }, { "grad_norm": 0.2874189019203186, "learning_rate": 7.449379186806604e-05, "loss": 0.0255, "step": 37020 }, { "grad_norm": 0.31823819875717163, "learning_rate": 7.44793757078993e-05, "loss": 0.0236, "step": 37030 }, { "grad_norm": 0.2189055234193802, "learning_rate": 7.44649568707048e-05, "loss": 0.0251, "step": 37040 }, { "grad_norm": 0.31701117753982544, "learning_rate": 7.445053535805942e-05, "loss": 0.0246, "step": 37050 }, { "grad_norm": 0.24408277869224548, "learning_rate": 7.443611117154024e-05, "loss": 0.0248, "step": 37060 }, { "grad_norm": 0.16586151719093323, "learning_rate": 7.442168431272469e-05, "loss": 0.0211, "step": 37070 }, { "grad_norm": 0.1931115835905075, "learning_rate": 7.440725478319044e-05, "loss": 0.0273, "step": 37080 }, { "grad_norm": 0.2432338148355484, "learning_rate": 7.439282258451553e-05, "loss": 0.025, "step": 37090 }, { "grad_norm": 0.6956571936607361, "learning_rate": 7.43783877182782e-05, "loss": 0.0268, "step": 37100 }, { "grad_norm": 0.18844659626483917, "learning_rate": 7.436395018605707e-05, "loss": 0.0244, "step": 37110 }, { "grad_norm": 0.14124563336372375, "learning_rate": 7.434950998943094e-05, "loss": 0.0247, "step": 37120 }, { "grad_norm": 0.2530193626880646, "learning_rate": 7.433506712997902e-05, "loss": 0.0274, "step": 37130 }, { "grad_norm": 0.2620680630207062, "learning_rate": 7.432062160928077e-05, "loss": 0.0275, "step": 37140 }, { "grad_norm": 0.22669784724712372, "learning_rate": 7.430617342891588e-05, "loss": 0.0239, "step": 37150 }, { "grad_norm": 0.3807457685470581, "learning_rate": 7.429172259046442e-05, "loss": 0.0283, "step": 37160 }, { "grad_norm": 0.5391192436218262, "learning_rate": 7.42772690955067e-05, "loss": 0.0277, "step": 37170 }, { "grad_norm": 0.22825153172016144, "learning_rate": 7.426281294562332e-05, "loss": 0.0298, "step": 37180 }, { "grad_norm": 0.2452598214149475, "learning_rate": 7.424835414239519e-05, "loss": 0.0295, "step": 37190 }, { "grad_norm": 0.22670412063598633, "learning_rate": 7.423389268740352e-05, "loss": 0.0262, "step": 37200 }, { "grad_norm": 0.33265164494514465, "learning_rate": 7.42194285822298e-05, "loss": 0.026, "step": 37210 }, { "grad_norm": 0.22600321471691132, "learning_rate": 7.420496182845576e-05, "loss": 0.0276, "step": 37220 }, { "grad_norm": 0.26473286747932434, "learning_rate": 7.419049242766348e-05, "loss": 0.0289, "step": 37230 }, { "grad_norm": 0.20162011682987213, "learning_rate": 7.417602038143533e-05, "loss": 0.0262, "step": 37240 }, { "grad_norm": 0.3295709490776062, "learning_rate": 7.416154569135393e-05, "loss": 0.0296, "step": 37250 }, { "grad_norm": 0.27173393964767456, "learning_rate": 7.414706835900222e-05, "loss": 0.0291, "step": 37260 }, { "grad_norm": 0.22188343107700348, "learning_rate": 7.413258838596345e-05, "loss": 0.0227, "step": 37270 }, { "grad_norm": 0.17982593178749084, "learning_rate": 7.411810577382106e-05, "loss": 0.0234, "step": 37280 }, { "grad_norm": 0.28044891357421875, "learning_rate": 7.410362052415889e-05, "loss": 0.0252, "step": 37290 }, { "grad_norm": 0.21277916431427002, "learning_rate": 7.408913263856102e-05, "loss": 0.0233, "step": 37300 }, { "grad_norm": 0.2183552384376526, "learning_rate": 7.407464211861183e-05, "loss": 0.024, "step": 37310 }, { "grad_norm": 0.18180276453495026, "learning_rate": 7.406014896589597e-05, "loss": 0.0221, "step": 37320 }, { "grad_norm": 0.19005528092384338, "learning_rate": 7.404565318199838e-05, "loss": 0.0239, "step": 37330 }, { "grad_norm": 0.14589832723140717, "learning_rate": 7.403115476850432e-05, "loss": 0.0218, "step": 37340 }, { "grad_norm": 0.23500725626945496, "learning_rate": 7.401665372699932e-05, "loss": 0.0237, "step": 37350 }, { "grad_norm": 0.33693864941596985, "learning_rate": 7.400215005906916e-05, "loss": 0.0312, "step": 37360 }, { "grad_norm": 0.2172536700963974, "learning_rate": 7.398764376629996e-05, "loss": 0.0259, "step": 37370 }, { "grad_norm": 0.22509202361106873, "learning_rate": 7.39731348502781e-05, "loss": 0.0258, "step": 37380 }, { "grad_norm": 0.2149723619222641, "learning_rate": 7.395862331259028e-05, "loss": 0.0377, "step": 37390 }, { "grad_norm": 0.27281272411346436, "learning_rate": 7.394410915482341e-05, "loss": 0.0365, "step": 37400 }, { "grad_norm": 0.19596543908119202, "learning_rate": 7.392959237856476e-05, "loss": 0.0324, "step": 37410 }, { "grad_norm": 0.27404800057411194, "learning_rate": 7.391507298540187e-05, "loss": 0.0256, "step": 37420 }, { "grad_norm": 0.22779372334480286, "learning_rate": 7.390055097692257e-05, "loss": 0.0302, "step": 37430 }, { "grad_norm": 0.223918616771698, "learning_rate": 7.388602635471494e-05, "loss": 0.0252, "step": 37440 }, { "grad_norm": 0.21451780200004578, "learning_rate": 7.387149912036738e-05, "loss": 0.0276, "step": 37450 }, { "grad_norm": 0.23893530666828156, "learning_rate": 7.385696927546854e-05, "loss": 0.0294, "step": 37460 }, { "grad_norm": 0.3074037432670593, "learning_rate": 7.384243682160743e-05, "loss": 0.026, "step": 37470 }, { "grad_norm": 0.22538597881793976, "learning_rate": 7.382790176037329e-05, "loss": 0.0224, "step": 37480 }, { "grad_norm": 0.22833523154258728, "learning_rate": 7.381336409335562e-05, "loss": 0.0268, "step": 37490 }, { "grad_norm": 0.17424510419368744, "learning_rate": 7.379882382214426e-05, "loss": 0.0251, "step": 37500 }, { "grad_norm": 0.2006746530532837, "learning_rate": 7.378428094832931e-05, "loss": 0.0218, "step": 37510 }, { "grad_norm": 0.28149136900901794, "learning_rate": 7.376973547350114e-05, "loss": 0.022, "step": 37520 }, { "grad_norm": 0.22111600637435913, "learning_rate": 7.375518739925046e-05, "loss": 0.0256, "step": 37530 }, { "grad_norm": 0.3187180757522583, "learning_rate": 7.374063672716818e-05, "loss": 0.0221, "step": 37540 }, { "grad_norm": 0.17499257624149323, "learning_rate": 7.372608345884558e-05, "loss": 0.025, "step": 37550 }, { "grad_norm": 0.19313624501228333, "learning_rate": 7.371152759587415e-05, "loss": 0.0235, "step": 37560 }, { "grad_norm": 0.26100584864616394, "learning_rate": 7.369696913984572e-05, "loss": 0.0219, "step": 37570 }, { "grad_norm": 0.23882299661636353, "learning_rate": 7.368240809235237e-05, "loss": 0.025, "step": 37580 }, { "grad_norm": 0.4982876777648926, "learning_rate": 7.366784445498649e-05, "loss": 0.0237, "step": 37590 }, { "grad_norm": 0.29092636704444885, "learning_rate": 7.36532782293407e-05, "loss": 0.0263, "step": 37600 }, { "grad_norm": 0.28181153535842896, "learning_rate": 7.363870941700797e-05, "loss": 0.0251, "step": 37610 }, { "grad_norm": 0.21312859654426575, "learning_rate": 7.362413801958152e-05, "loss": 0.0247, "step": 37620 }, { "grad_norm": 0.16066399216651917, "learning_rate": 7.360956403865486e-05, "loss": 0.0243, "step": 37630 }, { "grad_norm": 0.2108432799577713, "learning_rate": 7.359498747582177e-05, "loss": 0.0286, "step": 37640 }, { "grad_norm": 0.2699377238750458, "learning_rate": 7.358040833267634e-05, "loss": 0.0251, "step": 37650 }, { "grad_norm": 0.17333519458770752, "learning_rate": 7.356582661081288e-05, "loss": 0.0255, "step": 37660 }, { "grad_norm": 0.2312477082014084, "learning_rate": 7.355124231182607e-05, "loss": 0.0232, "step": 37670 }, { "grad_norm": 0.22431714832782745, "learning_rate": 7.353665543731083e-05, "loss": 0.0255, "step": 37680 }, { "grad_norm": 0.23768667876720428, "learning_rate": 7.352206598886229e-05, "loss": 0.0278, "step": 37690 }, { "grad_norm": 0.21621465682983398, "learning_rate": 7.350747396807601e-05, "loss": 0.0235, "step": 37700 }, { "grad_norm": 0.19538047909736633, "learning_rate": 7.349287937654772e-05, "loss": 0.0273, "step": 37710 }, { "grad_norm": 0.39174020290374756, "learning_rate": 7.347828221587345e-05, "loss": 0.0258, "step": 37720 }, { "grad_norm": 0.290130078792572, "learning_rate": 7.346368248764955e-05, "loss": 0.0231, "step": 37730 }, { "grad_norm": 0.20070980489253998, "learning_rate": 7.34490801934726e-05, "loss": 0.0213, "step": 37740 }, { "grad_norm": 0.3577079772949219, "learning_rate": 7.343447533493947e-05, "loss": 0.0291, "step": 37750 }, { "grad_norm": 0.26959139108657837, "learning_rate": 7.341986791364739e-05, "loss": 0.0261, "step": 37760 }, { "grad_norm": 0.3308905363082886, "learning_rate": 7.340525793119374e-05, "loss": 0.0285, "step": 37770 }, { "grad_norm": 0.3365584909915924, "learning_rate": 7.33906453891763e-05, "loss": 0.0256, "step": 37780 }, { "grad_norm": 0.318972647190094, "learning_rate": 7.337603028919303e-05, "loss": 0.0289, "step": 37790 }, { "grad_norm": 0.2916962504386902, "learning_rate": 7.336141263284225e-05, "loss": 0.0261, "step": 37800 }, { "grad_norm": 0.2072882354259491, "learning_rate": 7.334679242172249e-05, "loss": 0.0218, "step": 37810 }, { "grad_norm": 0.19210167229175568, "learning_rate": 7.33321696574326e-05, "loss": 0.0282, "step": 37820 }, { "grad_norm": 0.21994148194789886, "learning_rate": 7.331754434157174e-05, "loss": 0.0266, "step": 37830 }, { "grad_norm": 0.19454310834407806, "learning_rate": 7.330291647573928e-05, "loss": 0.0229, "step": 37840 }, { "grad_norm": 0.3024064302444458, "learning_rate": 7.32882860615349e-05, "loss": 0.0247, "step": 37850 }, { "grad_norm": 0.2832963466644287, "learning_rate": 7.327365310055858e-05, "loss": 0.0294, "step": 37860 }, { "grad_norm": 0.5191596150398254, "learning_rate": 7.325901759441055e-05, "loss": 0.0277, "step": 37870 }, { "grad_norm": 0.19751790165901184, "learning_rate": 7.324437954469132e-05, "loss": 0.0294, "step": 37880 }, { "grad_norm": 0.25169637799263, "learning_rate": 7.32297389530017e-05, "loss": 0.0279, "step": 37890 }, { "grad_norm": 0.18583719432353973, "learning_rate": 7.321509582094274e-05, "loss": 0.0248, "step": 37900 }, { "grad_norm": 0.2696877121925354, "learning_rate": 7.320045015011582e-05, "loss": 0.0247, "step": 37910 }, { "grad_norm": 0.15933707356452942, "learning_rate": 7.318580194212253e-05, "loss": 0.0255, "step": 37920 }, { "grad_norm": 0.2137979418039322, "learning_rate": 7.317115119856483e-05, "loss": 0.0225, "step": 37930 }, { "grad_norm": 0.2642854154109955, "learning_rate": 7.315649792104485e-05, "loss": 0.0257, "step": 37940 }, { "grad_norm": 0.49429404735565186, "learning_rate": 7.314184211116507e-05, "loss": 0.0311, "step": 37950 }, { "grad_norm": 0.36282917857170105, "learning_rate": 7.312718377052823e-05, "loss": 0.0253, "step": 37960 }, { "grad_norm": 0.2524509131908417, "learning_rate": 7.311252290073737e-05, "loss": 0.0269, "step": 37970 }, { "grad_norm": 0.7283972501754761, "learning_rate": 7.309785950339572e-05, "loss": 0.0279, "step": 37980 }, { "grad_norm": 0.24526865780353546, "learning_rate": 7.308319358010692e-05, "loss": 0.0254, "step": 37990 }, { "grad_norm": 0.2579321265220642, "learning_rate": 7.306852513247474e-05, "loss": 0.0245, "step": 38000 }, { "grad_norm": 0.20466648042201996, "learning_rate": 7.305385416210336e-05, "loss": 0.0252, "step": 38010 }, { "grad_norm": 0.22372837364673615, "learning_rate": 7.303918067059711e-05, "loss": 0.0266, "step": 38020 }, { "grad_norm": 0.21362021565437317, "learning_rate": 7.302450465956075e-05, "loss": 0.0221, "step": 38030 }, { "grad_norm": 0.19775640964508057, "learning_rate": 7.300982613059914e-05, "loss": 0.0252, "step": 38040 }, { "grad_norm": 0.2014751136302948, "learning_rate": 7.299514508531757e-05, "loss": 0.027, "step": 38050 }, { "grad_norm": 0.2110053300857544, "learning_rate": 7.298046152532148e-05, "loss": 0.0246, "step": 38060 }, { "grad_norm": 0.27222153544425964, "learning_rate": 7.296577545221668e-05, "loss": 0.0267, "step": 38070 }, { "grad_norm": 1.1227360963821411, "learning_rate": 7.29510868676092e-05, "loss": 0.026, "step": 38080 }, { "grad_norm": 0.2410486340522766, "learning_rate": 7.293639577310538e-05, "loss": 0.0239, "step": 38090 }, { "grad_norm": 0.21415270864963531, "learning_rate": 7.292170217031179e-05, "loss": 0.0251, "step": 38100 }, { "grad_norm": 0.39608973264694214, "learning_rate": 7.290700606083532e-05, "loss": 0.0297, "step": 38110 }, { "grad_norm": 0.21181201934814453, "learning_rate": 7.28923074462831e-05, "loss": 0.0231, "step": 38120 }, { "grad_norm": 0.33344995975494385, "learning_rate": 7.287760632826255e-05, "loss": 0.0284, "step": 38130 }, { "grad_norm": 0.23103171586990356, "learning_rate": 7.286290270838139e-05, "loss": 0.0255, "step": 38140 }, { "grad_norm": 0.26478856801986694, "learning_rate": 7.284819658824756e-05, "loss": 0.0315, "step": 38150 }, { "grad_norm": 0.6201943159103394, "learning_rate": 7.283348796946929e-05, "loss": 0.0277, "step": 38160 }, { "grad_norm": 0.17822015285491943, "learning_rate": 7.28187768536551e-05, "loss": 0.0291, "step": 38170 }, { "grad_norm": 0.24521617591381073, "learning_rate": 7.280406324241379e-05, "loss": 0.0262, "step": 38180 }, { "grad_norm": 0.278499573469162, "learning_rate": 7.278934713735441e-05, "loss": 0.0246, "step": 38190 }, { "grad_norm": 0.2461688220500946, "learning_rate": 7.277462854008629e-05, "loss": 0.0224, "step": 38200 }, { "grad_norm": 0.27875033020973206, "learning_rate": 7.275990745221904e-05, "loss": 0.0222, "step": 38210 }, { "grad_norm": 0.2962145507335663, "learning_rate": 7.274518387536252e-05, "loss": 0.0296, "step": 38220 }, { "grad_norm": 0.2220366895198822, "learning_rate": 7.273045781112689e-05, "loss": 0.0283, "step": 38230 }, { "grad_norm": 0.21543928980827332, "learning_rate": 7.271572926112257e-05, "loss": 0.0251, "step": 38240 }, { "grad_norm": 0.22658883035182953, "learning_rate": 7.270099822696024e-05, "loss": 0.0281, "step": 38250 }, { "grad_norm": 0.3284333050251007, "learning_rate": 7.26862647102509e-05, "loss": 0.0279, "step": 38260 }, { "grad_norm": 0.17635081708431244, "learning_rate": 7.267152871260573e-05, "loss": 0.0291, "step": 38270 }, { "grad_norm": 0.20924438536167145, "learning_rate": 7.265679023563629e-05, "loss": 0.0257, "step": 38280 }, { "grad_norm": 0.31029149889945984, "learning_rate": 7.264204928095433e-05, "loss": 0.0241, "step": 38290 }, { "grad_norm": 0.26373329758644104, "learning_rate": 7.262730585017188e-05, "loss": 0.0222, "step": 38300 }, { "grad_norm": 0.24601247906684875, "learning_rate": 7.261255994490131e-05, "loss": 0.026, "step": 38310 }, { "grad_norm": 0.21424171328544617, "learning_rate": 7.259781156675517e-05, "loss": 0.0222, "step": 38320 }, { "grad_norm": 0.24939388036727905, "learning_rate": 7.258306071734633e-05, "loss": 0.022, "step": 38330 }, { "grad_norm": 0.19503197073936462, "learning_rate": 7.256830739828792e-05, "loss": 0.0239, "step": 38340 }, { "grad_norm": 0.314269483089447, "learning_rate": 7.255355161119336e-05, "loss": 0.029, "step": 38350 }, { "grad_norm": 0.44527119398117065, "learning_rate": 7.253879335767628e-05, "loss": 0.0246, "step": 38360 }, { "grad_norm": 0.23124906420707703, "learning_rate": 7.252403263935067e-05, "loss": 0.026, "step": 38370 }, { "grad_norm": 0.2813931703567505, "learning_rate": 7.25092694578307e-05, "loss": 0.023, "step": 38380 }, { "grad_norm": 0.23251023888587952, "learning_rate": 7.249450381473087e-05, "loss": 0.0226, "step": 38390 }, { "grad_norm": 0.2522141933441162, "learning_rate": 7.247973571166593e-05, "loss": 0.0229, "step": 38400 }, { "grad_norm": 0.2637780010700226, "learning_rate": 7.246496515025089e-05, "loss": 0.0243, "step": 38410 }, { "grad_norm": 0.1585916131734848, "learning_rate": 7.245019213210101e-05, "loss": 0.0208, "step": 38420 }, { "grad_norm": 0.23489896953105927, "learning_rate": 7.24354166588319e-05, "loss": 0.0216, "step": 38430 }, { "grad_norm": 0.2161092311143875, "learning_rate": 7.242063873205935e-05, "loss": 0.0199, "step": 38440 }, { "grad_norm": 0.23582635819911957, "learning_rate": 7.240585835339946e-05, "loss": 0.0225, "step": 38450 }, { "grad_norm": 0.22569218277931213, "learning_rate": 7.239107552446857e-05, "loss": 0.021, "step": 38460 }, { "grad_norm": 0.17070399224758148, "learning_rate": 7.237629024688334e-05, "loss": 0.023, "step": 38470 }, { "grad_norm": 0.2884660065174103, "learning_rate": 7.236150252226064e-05, "loss": 0.028, "step": 38480 }, { "grad_norm": 0.25992724299430847, "learning_rate": 7.234671235221765e-05, "loss": 0.0262, "step": 38490 }, { "grad_norm": 0.37541642785072327, "learning_rate": 7.233191973837179e-05, "loss": 0.026, "step": 38500 }, { "grad_norm": 0.2685181498527527, "learning_rate": 7.231712468234078e-05, "loss": 0.0227, "step": 38510 }, { "grad_norm": 0.2139342725276947, "learning_rate": 7.230232718574254e-05, "loss": 0.0222, "step": 38520 }, { "grad_norm": 0.3150336742401123, "learning_rate": 7.228752725019535e-05, "loss": 0.0246, "step": 38530 }, { "grad_norm": 0.1912783533334732, "learning_rate": 7.227272487731769e-05, "loss": 0.0227, "step": 38540 }, { "grad_norm": 0.20636452734470367, "learning_rate": 7.225792006872831e-05, "loss": 0.0254, "step": 38550 }, { "grad_norm": 0.21389538049697876, "learning_rate": 7.224311282604628e-05, "loss": 0.0246, "step": 38560 }, { "grad_norm": 0.3126196563243866, "learning_rate": 7.222830315089085e-05, "loss": 0.0294, "step": 38570 }, { "grad_norm": 0.15503805875778198, "learning_rate": 7.22134910448816e-05, "loss": 0.0257, "step": 38580 }, { "grad_norm": 0.19021423161029816, "learning_rate": 7.219867650963839e-05, "loss": 0.0222, "step": 38590 }, { "grad_norm": 0.18940690159797668, "learning_rate": 7.21838595467813e-05, "loss": 0.0222, "step": 38600 }, { "grad_norm": 0.19229182600975037, "learning_rate": 7.216904015793067e-05, "loss": 0.0222, "step": 38610 }, { "grad_norm": 0.19768452644348145, "learning_rate": 7.215421834470713e-05, "loss": 0.0252, "step": 38620 }, { "grad_norm": 0.15394018590450287, "learning_rate": 7.213939410873158e-05, "loss": 0.0216, "step": 38630 }, { "grad_norm": 0.23405690491199493, "learning_rate": 7.212456745162519e-05, "loss": 0.0235, "step": 38640 }, { "grad_norm": 0.35507434606552124, "learning_rate": 7.210973837500937e-05, "loss": 0.0261, "step": 38650 }, { "grad_norm": 0.1968299150466919, "learning_rate": 7.209490688050578e-05, "loss": 0.0237, "step": 38660 }, { "grad_norm": 0.18827936053276062, "learning_rate": 7.208007296973641e-05, "loss": 0.0275, "step": 38670 }, { "grad_norm": 0.22357793152332306, "learning_rate": 7.206523664432345e-05, "loss": 0.023, "step": 38680 }, { "grad_norm": 0.19063004851341248, "learning_rate": 7.205039790588939e-05, "loss": 0.027, "step": 38690 }, { "grad_norm": 0.20438779890537262, "learning_rate": 7.203555675605697e-05, "loss": 0.0259, "step": 38700 }, { "grad_norm": 0.20739637315273285, "learning_rate": 7.202071319644917e-05, "loss": 0.0203, "step": 38710 }, { "grad_norm": 0.22401846945285797, "learning_rate": 7.200586722868932e-05, "loss": 0.0264, "step": 38720 }, { "grad_norm": 0.4045131802558899, "learning_rate": 7.199101885440088e-05, "loss": 0.0259, "step": 38730 }, { "grad_norm": 0.18522804975509644, "learning_rate": 7.197616807520771e-05, "loss": 0.0262, "step": 38740 }, { "grad_norm": 0.20093393325805664, "learning_rate": 7.196131489273381e-05, "loss": 0.0282, "step": 38750 }, { "grad_norm": 0.23258383572101593, "learning_rate": 7.194645930860355e-05, "loss": 0.0269, "step": 38760 }, { "grad_norm": 0.2818751931190491, "learning_rate": 7.19316013244415e-05, "loss": 0.0264, "step": 38770 }, { "grad_norm": 0.24913182854652405, "learning_rate": 7.191674094187248e-05, "loss": 0.0296, "step": 38780 }, { "grad_norm": 0.2452262043952942, "learning_rate": 7.190187816252165e-05, "loss": 0.0242, "step": 38790 }, { "grad_norm": 0.33776015043258667, "learning_rate": 7.188701298801435e-05, "loss": 0.0257, "step": 38800 }, { "grad_norm": 0.27921655774116516, "learning_rate": 7.18721454199762e-05, "loss": 0.0288, "step": 38810 }, { "grad_norm": 0.2097199261188507, "learning_rate": 7.185727546003312e-05, "loss": 0.0289, "step": 38820 }, { "grad_norm": 0.2339717298746109, "learning_rate": 7.184240310981126e-05, "loss": 0.0245, "step": 38830 }, { "grad_norm": 0.1597982794046402, "learning_rate": 7.182752837093704e-05, "loss": 0.0214, "step": 38840 }, { "grad_norm": 0.23447227478027344, "learning_rate": 7.181265124503711e-05, "loss": 0.0287, "step": 38850 }, { "grad_norm": 0.3778410255908966, "learning_rate": 7.179777173373847e-05, "loss": 0.0343, "step": 38860 }, { "grad_norm": 0.2794860005378723, "learning_rate": 7.178288983866826e-05, "loss": 0.0315, "step": 38870 }, { "grad_norm": 0.33874577283859253, "learning_rate": 7.176800556145397e-05, "loss": 0.0289, "step": 38880 }, { "grad_norm": 0.23578611016273499, "learning_rate": 7.175311890372334e-05, "loss": 0.0268, "step": 38890 }, { "grad_norm": 0.2987678050994873, "learning_rate": 7.17382298671043e-05, "loss": 0.0303, "step": 38900 }, { "grad_norm": 0.18329136073589325, "learning_rate": 7.172333845322515e-05, "loss": 0.0239, "step": 38910 }, { "grad_norm": 0.18129359185695648, "learning_rate": 7.170844466371436e-05, "loss": 0.0225, "step": 38920 }, { "grad_norm": 0.2174641191959381, "learning_rate": 7.16935485002007e-05, "loss": 0.0313, "step": 38930 }, { "grad_norm": 0.391633540391922, "learning_rate": 7.167864996431319e-05, "loss": 0.023, "step": 38940 }, { "grad_norm": 0.21980449557304382, "learning_rate": 7.166374905768111e-05, "loss": 0.0258, "step": 38950 }, { "grad_norm": 0.2294236272573471, "learning_rate": 7.164884578193401e-05, "loss": 0.0287, "step": 38960 }, { "grad_norm": 0.30502602458000183, "learning_rate": 7.163394013870168e-05, "loss": 0.0262, "step": 38970 }, { "grad_norm": 0.2319687455892563, "learning_rate": 7.161903212961419e-05, "loss": 0.0258, "step": 38980 }, { "grad_norm": 0.1874619722366333, "learning_rate": 7.160412175630183e-05, "loss": 0.0236, "step": 38990 }, { "grad_norm": 0.23429825901985168, "learning_rate": 7.158920902039521e-05, "loss": 0.0208, "step": 39000 }, { "grad_norm": 0.19329115748405457, "learning_rate": 7.157429392352514e-05, "loss": 0.0276, "step": 39010 }, { "grad_norm": 0.4468952715396881, "learning_rate": 7.155937646732275e-05, "loss": 0.0285, "step": 39020 }, { "grad_norm": 0.18248692154884338, "learning_rate": 7.154445665341933e-05, "loss": 0.0234, "step": 39030 }, { "grad_norm": 0.2564276158809662, "learning_rate": 7.152953448344654e-05, "loss": 0.0238, "step": 39040 }, { "grad_norm": 0.31697386503219604, "learning_rate": 7.151460995903624e-05, "loss": 0.0257, "step": 39050 }, { "grad_norm": 0.2572212219238281, "learning_rate": 7.149968308182052e-05, "loss": 0.0245, "step": 39060 }, { "grad_norm": 0.22052690386772156, "learning_rate": 7.14847538534318e-05, "loss": 0.0237, "step": 39070 }, { "grad_norm": 0.23542873561382294, "learning_rate": 7.14698222755027e-05, "loss": 0.0272, "step": 39080 }, { "grad_norm": 0.27340760827064514, "learning_rate": 7.14548883496661e-05, "loss": 0.0271, "step": 39090 }, { "grad_norm": 0.3471856415271759, "learning_rate": 7.143995207755517e-05, "loss": 0.024, "step": 39100 }, { "grad_norm": 0.21480058133602142, "learning_rate": 7.142501346080333e-05, "loss": 0.0254, "step": 39110 }, { "grad_norm": 0.2913764715194702, "learning_rate": 7.141007250104421e-05, "loss": 0.0276, "step": 39120 }, { "grad_norm": 0.2525986135005951, "learning_rate": 7.139512919991176e-05, "loss": 0.0239, "step": 39130 }, { "grad_norm": 0.17806501686573029, "learning_rate": 7.138018355904014e-05, "loss": 0.0234, "step": 39140 }, { "grad_norm": 0.230929896235466, "learning_rate": 7.13652355800638e-05, "loss": 0.0206, "step": 39150 }, { "grad_norm": 0.1811518520116806, "learning_rate": 7.135028526461743e-05, "loss": 0.0229, "step": 39160 }, { "grad_norm": 0.1811227947473526, "learning_rate": 7.133533261433594e-05, "loss": 0.02, "step": 39170 }, { "grad_norm": 0.221515953540802, "learning_rate": 7.132037763085457e-05, "loss": 0.0248, "step": 39180 }, { "grad_norm": 0.21575860679149628, "learning_rate": 7.130542031580875e-05, "loss": 0.0217, "step": 39190 }, { "grad_norm": 0.27131474018096924, "learning_rate": 7.12904606708342e-05, "loss": 0.0259, "step": 39200 }, { "grad_norm": 0.2877448499202728, "learning_rate": 7.127549869756687e-05, "loss": 0.0263, "step": 39210 }, { "grad_norm": 0.3913337290287018, "learning_rate": 7.126053439764299e-05, "loss": 0.0286, "step": 39220 }, { "grad_norm": 0.26212766766548157, "learning_rate": 7.124556777269904e-05, "loss": 0.0284, "step": 39230 }, { "grad_norm": 0.2926608920097351, "learning_rate": 7.123059882437174e-05, "loss": 0.0365, "step": 39240 }, { "grad_norm": 0.26953309774398804, "learning_rate": 7.121562755429807e-05, "loss": 0.0272, "step": 39250 }, { "grad_norm": 0.4383833706378937, "learning_rate": 7.120065396411528e-05, "loss": 0.0263, "step": 39260 }, { "grad_norm": 0.25053369998931885, "learning_rate": 7.118567805546084e-05, "loss": 0.0271, "step": 39270 }, { "grad_norm": 0.19956038892269135, "learning_rate": 7.117069982997248e-05, "loss": 0.026, "step": 39280 }, { "grad_norm": 0.3312629759311676, "learning_rate": 7.115571928928825e-05, "loss": 0.0259, "step": 39290 }, { "grad_norm": 0.2297217845916748, "learning_rate": 7.114073643504635e-05, "loss": 0.0239, "step": 39300 }, { "grad_norm": 0.23698405921459198, "learning_rate": 7.11257512688853e-05, "loss": 0.0263, "step": 39310 }, { "grad_norm": 0.6125267744064331, "learning_rate": 7.111076379244384e-05, "loss": 0.0245, "step": 39320 }, { "grad_norm": 0.2566932737827301, "learning_rate": 7.109577400736101e-05, "loss": 0.0257, "step": 39330 }, { "grad_norm": 0.21310412883758545, "learning_rate": 7.108078191527605e-05, "loss": 0.0282, "step": 39340 }, { "grad_norm": 0.27850767970085144, "learning_rate": 7.106578751782847e-05, "loss": 0.0282, "step": 39350 }, { "grad_norm": 0.3768277168273926, "learning_rate": 7.105079081665803e-05, "loss": 0.0276, "step": 39360 }, { "grad_norm": 0.18790364265441895, "learning_rate": 7.103579181340476e-05, "loss": 0.0321, "step": 39370 }, { "grad_norm": 0.32262474298477173, "learning_rate": 7.102079050970893e-05, "loss": 0.0337, "step": 39380 }, { "grad_norm": 0.7288718223571777, "learning_rate": 7.100578690721105e-05, "loss": 0.0249, "step": 39390 }, { "grad_norm": 0.5750892758369446, "learning_rate": 7.09907810075519e-05, "loss": 0.0247, "step": 39400 }, { "grad_norm": 0.22581307590007782, "learning_rate": 7.097577281237249e-05, "loss": 0.0266, "step": 39410 }, { "grad_norm": 0.31250789761543274, "learning_rate": 7.09607623233141e-05, "loss": 0.0244, "step": 39420 }, { "grad_norm": 0.17957474291324615, "learning_rate": 7.094574954201827e-05, "loss": 0.0231, "step": 39430 }, { "grad_norm": 0.23125451803207397, "learning_rate": 7.093073447012675e-05, "loss": 0.0242, "step": 39440 }, { "grad_norm": 0.17342790961265564, "learning_rate": 7.09157171092816e-05, "loss": 0.0278, "step": 39450 }, { "grad_norm": 0.357942670583725, "learning_rate": 7.090069746112504e-05, "loss": 0.0286, "step": 39460 }, { "grad_norm": 0.31469881534576416, "learning_rate": 7.088567552729965e-05, "loss": 0.025, "step": 39470 }, { "grad_norm": 0.2623962461948395, "learning_rate": 7.087065130944818e-05, "loss": 0.0229, "step": 39480 }, { "grad_norm": 0.37625011801719666, "learning_rate": 7.085562480921366e-05, "loss": 0.0249, "step": 39490 }, { "grad_norm": 0.19063538312911987, "learning_rate": 7.084059602823937e-05, "loss": 0.0268, "step": 39500 }, { "grad_norm": 0.23482713103294373, "learning_rate": 7.082556496816882e-05, "loss": 0.0226, "step": 39510 }, { "grad_norm": 0.2562401592731476, "learning_rate": 7.081053163064582e-05, "loss": 0.0245, "step": 39520 }, { "grad_norm": 0.28072354197502136, "learning_rate": 7.079549601731434e-05, "loss": 0.028, "step": 39530 }, { "grad_norm": 0.26960036158561707, "learning_rate": 7.07804581298187e-05, "loss": 0.0262, "step": 39540 }, { "grad_norm": 0.21244493126869202, "learning_rate": 7.07654179698034e-05, "loss": 0.0245, "step": 39550 }, { "grad_norm": 0.15951599180698395, "learning_rate": 7.075037553891321e-05, "loss": 0.0215, "step": 39560 }, { "grad_norm": 0.3234632611274719, "learning_rate": 7.073533083879315e-05, "loss": 0.0242, "step": 39570 }, { "grad_norm": 0.35137903690338135, "learning_rate": 7.072028387108849e-05, "loss": 0.0298, "step": 39580 }, { "grad_norm": 0.20183822512626648, "learning_rate": 7.070523463744472e-05, "loss": 0.0262, "step": 39590 }, { "grad_norm": 0.16362804174423218, "learning_rate": 7.069018313950763e-05, "loss": 0.0229, "step": 39600 }, { "grad_norm": 0.25539132952690125, "learning_rate": 7.067512937892322e-05, "loss": 0.0246, "step": 39610 }, { "grad_norm": 0.24654260277748108, "learning_rate": 7.066007335733775e-05, "loss": 0.0264, "step": 39620 }, { "grad_norm": 0.29817071557044983, "learning_rate": 7.064501507639772e-05, "loss": 0.0254, "step": 39630 }, { "grad_norm": 0.2001810371875763, "learning_rate": 7.062995453774987e-05, "loss": 0.0218, "step": 39640 }, { "grad_norm": 0.2450055629014969, "learning_rate": 7.061489174304121e-05, "loss": 0.0252, "step": 39650 }, { "grad_norm": 0.31214651465415955, "learning_rate": 7.0599826693919e-05, "loss": 0.035, "step": 39660 }, { "grad_norm": 0.32455557584762573, "learning_rate": 7.05847593920307e-05, "loss": 0.027, "step": 39670 }, { "grad_norm": 0.2671504616737366, "learning_rate": 7.056968983902406e-05, "loss": 0.0234, "step": 39680 }, { "grad_norm": 0.5561679601669312, "learning_rate": 7.055461803654706e-05, "loss": 0.0254, "step": 39690 }, { "grad_norm": 0.20537708699703217, "learning_rate": 7.053954398624794e-05, "loss": 0.0277, "step": 39700 }, { "grad_norm": 0.22276267409324646, "learning_rate": 7.052446768977518e-05, "loss": 0.0259, "step": 39710 }, { "grad_norm": 0.4990317225456238, "learning_rate": 7.050938914877748e-05, "loss": 0.0256, "step": 39720 }, { "grad_norm": 0.24169029295444489, "learning_rate": 7.049430836490382e-05, "loss": 0.03, "step": 39730 }, { "grad_norm": 0.308695912361145, "learning_rate": 7.047922533980342e-05, "loss": 0.0273, "step": 39740 }, { "grad_norm": 1.803393840789795, "learning_rate": 7.046414007512571e-05, "loss": 0.0274, "step": 39750 }, { "grad_norm": 0.2163211554288864, "learning_rate": 7.044905257252042e-05, "loss": 0.0289, "step": 39760 }, { "grad_norm": 0.16997048258781433, "learning_rate": 7.04339628336375e-05, "loss": 0.0236, "step": 39770 }, { "grad_norm": 0.20541897416114807, "learning_rate": 7.041887086012711e-05, "loss": 0.0257, "step": 39780 }, { "grad_norm": 0.198052778840065, "learning_rate": 7.040377665363969e-05, "loss": 0.0266, "step": 39790 }, { "grad_norm": 0.17785176634788513, "learning_rate": 7.038868021582594e-05, "loss": 0.0275, "step": 39800 }, { "grad_norm": 0.3175993859767914, "learning_rate": 7.037358154833679e-05, "loss": 0.0247, "step": 39810 }, { "grad_norm": 0.19695471227169037, "learning_rate": 7.035848065282339e-05, "loss": 0.0332, "step": 39820 }, { "grad_norm": 0.32650649547576904, "learning_rate": 7.034337753093714e-05, "loss": 0.0326, "step": 39830 }, { "grad_norm": 0.226198211312294, "learning_rate": 7.032827218432972e-05, "loss": 0.0234, "step": 39840 }, { "grad_norm": 0.23231056332588196, "learning_rate": 7.031316461465302e-05, "loss": 0.021, "step": 39850 }, { "grad_norm": 0.1703377217054367, "learning_rate": 7.029805482355915e-05, "loss": 0.022, "step": 39860 }, { "grad_norm": 0.2076253443956375, "learning_rate": 7.028294281270055e-05, "loss": 0.022, "step": 39870 }, { "grad_norm": 0.15008006989955902, "learning_rate": 7.026782858372982e-05, "loss": 0.022, "step": 39880 }, { "grad_norm": 0.20609895884990692, "learning_rate": 7.025271213829982e-05, "loss": 0.0202, "step": 39890 }, { "grad_norm": 0.23691697418689728, "learning_rate": 7.023759347806366e-05, "loss": 0.0231, "step": 39900 }, { "grad_norm": 0.2324151247739792, "learning_rate": 7.022247260467472e-05, "loss": 0.0244, "step": 39910 }, { "grad_norm": 0.30924591422080994, "learning_rate": 7.020734951978658e-05, "loss": 0.0262, "step": 39920 }, { "grad_norm": 0.24291826784610748, "learning_rate": 7.019222422505307e-05, "loss": 0.0254, "step": 39930 }, { "grad_norm": 0.20118382573127747, "learning_rate": 7.017709672212827e-05, "loss": 0.0242, "step": 39940 }, { "grad_norm": 0.2519913613796234, "learning_rate": 7.016196701266652e-05, "loss": 0.0216, "step": 39950 }, { "grad_norm": 0.293841689825058, "learning_rate": 7.014683509832238e-05, "loss": 0.0258, "step": 39960 }, { "grad_norm": 0.223125159740448, "learning_rate": 7.013170098075063e-05, "loss": 0.0241, "step": 39970 }, { "grad_norm": 0.21747393906116486, "learning_rate": 7.011656466160632e-05, "loss": 0.0193, "step": 39980 }, { "grad_norm": 0.23310644924640656, "learning_rate": 7.010142614254475e-05, "loss": 0.024, "step": 39990 }, { "grad_norm": 0.24628129601478577, "learning_rate": 7.008628542522147e-05, "loss": 0.0212, "step": 40000 }, { "grad_norm": 0.21459007263183594, "learning_rate": 7.007114251129218e-05, "loss": 0.0365, "step": 40010 }, { "grad_norm": 0.7778788805007935, "learning_rate": 7.005599740241293e-05, "loss": 0.0254, "step": 40020 }, { "grad_norm": 0.2518751919269562, "learning_rate": 7.004085010023996e-05, "loss": 0.024, "step": 40030 }, { "grad_norm": 0.28313031792640686, "learning_rate": 7.002570060642976e-05, "loss": 0.0247, "step": 40040 }, { "grad_norm": 0.32705625891685486, "learning_rate": 7.001054892263903e-05, "loss": 0.0247, "step": 40050 }, { "grad_norm": 0.18869926035404205, "learning_rate": 6.999539505052477e-05, "loss": 0.024, "step": 40060 }, { "grad_norm": 0.18186427652835846, "learning_rate": 6.998023899174418e-05, "loss": 0.0218, "step": 40070 }, { "grad_norm": 0.2775023877620697, "learning_rate": 6.996508074795467e-05, "loss": 0.0242, "step": 40080 }, { "grad_norm": 0.4985402822494507, "learning_rate": 6.994992032081396e-05, "loss": 0.0313, "step": 40090 }, { "grad_norm": 0.21982897818088531, "learning_rate": 6.993475771197995e-05, "loss": 0.0275, "step": 40100 }, { "grad_norm": 0.4035469889640808, "learning_rate": 6.991959292311082e-05, "loss": 0.022, "step": 40110 }, { "grad_norm": 0.21692916750907898, "learning_rate": 6.990442595586495e-05, "loss": 0.0245, "step": 40120 }, { "grad_norm": 0.5944727659225464, "learning_rate": 6.988925681190098e-05, "loss": 0.0268, "step": 40130 }, { "grad_norm": 0.2593784034252167, "learning_rate": 6.987408549287778e-05, "loss": 0.0301, "step": 40140 }, { "grad_norm": 0.26154717803001404, "learning_rate": 6.985891200045449e-05, "loss": 0.0236, "step": 40150 }, { "grad_norm": 0.19983060657978058, "learning_rate": 6.984373633629045e-05, "loss": 0.0228, "step": 40160 }, { "grad_norm": 0.39192914962768555, "learning_rate": 6.98285585020452e-05, "loss": 0.0276, "step": 40170 }, { "grad_norm": 0.36707285046577454, "learning_rate": 6.981337849937864e-05, "loss": 0.0282, "step": 40180 }, { "grad_norm": 0.3167537450790405, "learning_rate": 6.979819632995078e-05, "loss": 0.0282, "step": 40190 }, { "grad_norm": 0.21551766991615295, "learning_rate": 6.978301199542193e-05, "loss": 0.0328, "step": 40200 }, { "grad_norm": 0.26491743326187134, "learning_rate": 6.976782549745263e-05, "loss": 0.0304, "step": 40210 }, { "grad_norm": 0.18843013048171997, "learning_rate": 6.975263683770367e-05, "loss": 0.0284, "step": 40220 }, { "grad_norm": 0.18571455776691437, "learning_rate": 6.973744601783604e-05, "loss": 0.0246, "step": 40230 }, { "grad_norm": 0.22493845224380493, "learning_rate": 6.972225303951097e-05, "loss": 0.0284, "step": 40240 }, { "grad_norm": 0.1738346815109253, "learning_rate": 6.970705790438998e-05, "loss": 0.0276, "step": 40250 }, { "grad_norm": 0.15711912512779236, "learning_rate": 6.969186061413477e-05, "loss": 0.0235, "step": 40260 }, { "grad_norm": 0.24970605969429016, "learning_rate": 6.967666117040727e-05, "loss": 0.0223, "step": 40270 }, { "grad_norm": 0.20354346930980682, "learning_rate": 6.966145957486972e-05, "loss": 0.0237, "step": 40280 }, { "grad_norm": 0.2169649451971054, "learning_rate": 6.964625582918449e-05, "loss": 0.0213, "step": 40290 }, { "grad_norm": 0.2972636818885803, "learning_rate": 6.963104993501425e-05, "loss": 0.0286, "step": 40300 }, { "grad_norm": 0.23528864979743958, "learning_rate": 6.961584189402192e-05, "loss": 0.0244, "step": 40310 }, { "grad_norm": 0.21848341822624207, "learning_rate": 6.96006317078706e-05, "loss": 0.0255, "step": 40320 }, { "grad_norm": 0.2613653838634491, "learning_rate": 6.958541937822367e-05, "loss": 0.0234, "step": 40330 }, { "grad_norm": 0.29220929741859436, "learning_rate": 6.957020490674473e-05, "loss": 0.026, "step": 40340 }, { "grad_norm": 0.23821604251861572, "learning_rate": 6.95549882950976e-05, "loss": 0.0218, "step": 40350 }, { "grad_norm": 0.1750338077545166, "learning_rate": 6.953976954494635e-05, "loss": 0.025, "step": 40360 }, { "grad_norm": 0.2194233238697052, "learning_rate": 6.952454865795528e-05, "loss": 0.0251, "step": 40370 }, { "grad_norm": 0.16491153836250305, "learning_rate": 6.950932563578892e-05, "loss": 0.0247, "step": 40380 }, { "grad_norm": 0.25335225462913513, "learning_rate": 6.949410048011206e-05, "loss": 0.023, "step": 40390 }, { "grad_norm": 0.17368613183498383, "learning_rate": 6.947887319258966e-05, "loss": 0.0232, "step": 40400 }, { "grad_norm": 0.1996622085571289, "learning_rate": 6.946364377488696e-05, "loss": 0.023, "step": 40410 }, { "grad_norm": 0.24358755350112915, "learning_rate": 6.944841222866947e-05, "loss": 0.0217, "step": 40420 }, { "grad_norm": 0.16446761786937714, "learning_rate": 6.943317855560284e-05, "loss": 0.024, "step": 40430 }, { "grad_norm": 0.2777611315250397, "learning_rate": 6.941794275735302e-05, "loss": 0.0309, "step": 40440 }, { "grad_norm": 0.1749776005744934, "learning_rate": 6.94027048355862e-05, "loss": 0.0229, "step": 40450 }, { "grad_norm": 0.17346788942813873, "learning_rate": 6.938746479196871e-05, "loss": 0.0225, "step": 40460 }, { "grad_norm": 0.17320828139781952, "learning_rate": 6.937222262816724e-05, "loss": 0.0215, "step": 40470 }, { "grad_norm": 0.2823019027709961, "learning_rate": 6.935697834584865e-05, "loss": 0.0272, "step": 40480 }, { "grad_norm": 0.21537449955940247, "learning_rate": 6.934173194667998e-05, "loss": 0.0273, "step": 40490 }, { "grad_norm": 0.15427224338054657, "learning_rate": 6.93264834323286e-05, "loss": 0.0233, "step": 40500 }, { "grad_norm": 0.2653518617153168, "learning_rate": 6.931123280446206e-05, "loss": 0.0261, "step": 40510 }, { "grad_norm": 0.25286349654197693, "learning_rate": 6.929598006474811e-05, "loss": 0.0236, "step": 40520 }, { "grad_norm": 0.2210131734609604, "learning_rate": 6.92807252148548e-05, "loss": 0.0226, "step": 40530 }, { "grad_norm": 0.23874279856681824, "learning_rate": 6.926546825645037e-05, "loss": 0.0224, "step": 40540 }, { "grad_norm": 0.2992170751094818, "learning_rate": 6.92502091912033e-05, "loss": 0.0205, "step": 40550 }, { "grad_norm": 0.22824792563915253, "learning_rate": 6.923494802078229e-05, "loss": 0.0237, "step": 40560 }, { "grad_norm": 0.22187326848506927, "learning_rate": 6.921968474685628e-05, "loss": 0.0244, "step": 40570 }, { "grad_norm": 0.20076662302017212, "learning_rate": 6.920441937109445e-05, "loss": 0.0287, "step": 40580 }, { "grad_norm": 0.42872169613838196, "learning_rate": 6.91891518951662e-05, "loss": 0.0249, "step": 40590 }, { "grad_norm": 0.22794516384601593, "learning_rate": 6.917388232074114e-05, "loss": 0.0262, "step": 40600 }, { "grad_norm": 0.20669159293174744, "learning_rate": 6.915861064948914e-05, "loss": 0.0262, "step": 40610 }, { "grad_norm": 0.2595680058002472, "learning_rate": 6.914333688308025e-05, "loss": 0.0287, "step": 40620 }, { "grad_norm": 0.2520489990711212, "learning_rate": 6.912806102318487e-05, "loss": 0.026, "step": 40630 }, { "grad_norm": 0.2453797310590744, "learning_rate": 6.911278307147347e-05, "loss": 0.0245, "step": 40640 }, { "grad_norm": 0.20866890251636505, "learning_rate": 6.909750302961684e-05, "loss": 0.021, "step": 40650 }, { "grad_norm": 0.19689565896987915, "learning_rate": 6.9082220899286e-05, "loss": 0.0267, "step": 40660 }, { "grad_norm": 0.1909559667110443, "learning_rate": 6.906693668215219e-05, "loss": 0.0236, "step": 40670 }, { "grad_norm": 0.15216214954853058, "learning_rate": 6.905165037988683e-05, "loss": 0.02, "step": 40680 }, { "grad_norm": 0.18199202418327332, "learning_rate": 6.903636199416164e-05, "loss": 0.022, "step": 40690 }, { "grad_norm": 0.2305254191160202, "learning_rate": 6.902107152664851e-05, "loss": 0.0277, "step": 40700 }, { "grad_norm": 0.27909210324287415, "learning_rate": 6.90057789790196e-05, "loss": 0.0271, "step": 40710 }, { "grad_norm": 0.2820298969745636, "learning_rate": 6.899048435294728e-05, "loss": 0.0235, "step": 40720 }, { "grad_norm": 0.24953162670135498, "learning_rate": 6.897518765010415e-05, "loss": 0.0215, "step": 40730 }, { "grad_norm": 0.17470645904541016, "learning_rate": 6.895988887216302e-05, "loss": 0.0212, "step": 40740 }, { "grad_norm": 0.3091484010219574, "learning_rate": 6.894458802079694e-05, "loss": 0.0232, "step": 40750 }, { "grad_norm": 0.27057695388793945, "learning_rate": 6.892928509767922e-05, "loss": 0.0298, "step": 40760 }, { "grad_norm": 0.22912700474262238, "learning_rate": 6.891398010448333e-05, "loss": 0.0294, "step": 40770 }, { "grad_norm": 0.3489782214164734, "learning_rate": 6.8898673042883e-05, "loss": 0.0312, "step": 40780 }, { "grad_norm": 0.21986456215381622, "learning_rate": 6.888336391455222e-05, "loss": 0.0225, "step": 40790 }, { "grad_norm": 0.2514784634113312, "learning_rate": 6.886805272116513e-05, "loss": 0.0321, "step": 40800 }, { "grad_norm": 0.2799781560897827, "learning_rate": 6.885273946439617e-05, "loss": 0.0233, "step": 40810 }, { "grad_norm": 0.18434609472751617, "learning_rate": 6.883742414591998e-05, "loss": 0.0235, "step": 40820 }, { "grad_norm": 0.20363043248653412, "learning_rate": 6.88221067674114e-05, "loss": 0.022, "step": 40830 }, { "grad_norm": 0.23590606451034546, "learning_rate": 6.88067873305455e-05, "loss": 0.0216, "step": 40840 }, { "grad_norm": 0.19869282841682434, "learning_rate": 6.879146583699765e-05, "loss": 0.0201, "step": 40850 }, { "grad_norm": 0.21268217265605927, "learning_rate": 6.877614228844334e-05, "loss": 0.0264, "step": 40860 }, { "grad_norm": 0.20034703612327576, "learning_rate": 6.876081668655832e-05, "loss": 0.0202, "step": 40870 }, { "grad_norm": 0.16221627593040466, "learning_rate": 6.874548903301861e-05, "loss": 0.024, "step": 40880 }, { "grad_norm": 0.20100022852420807, "learning_rate": 6.873015932950039e-05, "loss": 0.0218, "step": 40890 }, { "grad_norm": 0.21808668971061707, "learning_rate": 6.871482757768012e-05, "loss": 0.0223, "step": 40900 }, { "grad_norm": 0.17999622225761414, "learning_rate": 6.869949377923442e-05, "loss": 0.0252, "step": 40910 }, { "grad_norm": 0.2871536910533905, "learning_rate": 6.868415793584022e-05, "loss": 0.0236, "step": 40920 }, { "grad_norm": 0.23893123865127563, "learning_rate": 6.86688200491746e-05, "loss": 0.0227, "step": 40930 }, { "grad_norm": 0.249048113822937, "learning_rate": 6.865348012091488e-05, "loss": 0.0225, "step": 40940 }, { "grad_norm": 0.297698974609375, "learning_rate": 6.86381381527386e-05, "loss": 0.0258, "step": 40950 }, { "grad_norm": 0.23633672297000885, "learning_rate": 6.862279414632359e-05, "loss": 0.0228, "step": 40960 }, { "grad_norm": 0.2562101185321808, "learning_rate": 6.860744810334778e-05, "loss": 0.0286, "step": 40970 }, { "grad_norm": 0.2186896950006485, "learning_rate": 6.859210002548943e-05, "loss": 0.022, "step": 40980 }, { "grad_norm": 0.1860511600971222, "learning_rate": 6.8576749914427e-05, "loss": 0.0234, "step": 40990 }, { "grad_norm": 0.4237487316131592, "learning_rate": 6.85613977718391e-05, "loss": 0.0253, "step": 41000 }, { "grad_norm": 0.20803681015968323, "learning_rate": 6.854604359940464e-05, "loss": 0.0217, "step": 41010 }, { "grad_norm": 0.23751983046531677, "learning_rate": 6.853068739880276e-05, "loss": 0.0296, "step": 41020 }, { "grad_norm": 0.1801528036594391, "learning_rate": 6.851532917171276e-05, "loss": 0.0281, "step": 41030 }, { "grad_norm": 0.19990505278110504, "learning_rate": 6.849996891981418e-05, "loss": 0.0206, "step": 41040 }, { "grad_norm": 0.2434033304452896, "learning_rate": 6.848460664478684e-05, "loss": 0.0299, "step": 41050 }, { "grad_norm": 0.23140640556812286, "learning_rate": 6.846924234831069e-05, "loss": 0.0193, "step": 41060 }, { "grad_norm": 0.18296635150909424, "learning_rate": 6.845387603206597e-05, "loss": 0.0236, "step": 41070 }, { "grad_norm": 0.18189358711242676, "learning_rate": 6.843850769773312e-05, "loss": 0.0178, "step": 41080 }, { "grad_norm": 0.3654903769493103, "learning_rate": 6.842313734699278e-05, "loss": 0.0217, "step": 41090 }, { "grad_norm": 0.19961602985858917, "learning_rate": 6.840776498152584e-05, "loss": 0.0275, "step": 41100 }, { "grad_norm": 0.25558385252952576, "learning_rate": 6.83923906030134e-05, "loss": 0.0245, "step": 41110 }, { "grad_norm": 0.17341755330562592, "learning_rate": 6.837701421313677e-05, "loss": 0.0264, "step": 41120 }, { "grad_norm": 0.20670194923877716, "learning_rate": 6.83616358135775e-05, "loss": 0.0264, "step": 41130 }, { "grad_norm": 0.296172559261322, "learning_rate": 6.834625540601734e-05, "loss": 0.0253, "step": 41140 }, { "grad_norm": 0.26668983697891235, "learning_rate": 6.833087299213829e-05, "loss": 0.0255, "step": 41150 }, { "grad_norm": 0.2543655037879944, "learning_rate": 6.83154885736225e-05, "loss": 0.027, "step": 41160 }, { "grad_norm": 0.5628642439842224, "learning_rate": 6.830010215215244e-05, "loss": 0.0269, "step": 41170 }, { "grad_norm": 0.21864603459835052, "learning_rate": 6.828471372941072e-05, "loss": 0.0222, "step": 41180 }, { "grad_norm": 0.2223510444164276, "learning_rate": 6.826932330708018e-05, "loss": 0.0228, "step": 41190 }, { "grad_norm": 0.5148501992225647, "learning_rate": 6.825393088684393e-05, "loss": 0.0253, "step": 41200 }, { "grad_norm": 0.3555745482444763, "learning_rate": 6.823853647038524e-05, "loss": 0.0308, "step": 41210 }, { "grad_norm": 0.2392127811908722, "learning_rate": 6.822314005938762e-05, "loss": 0.0239, "step": 41220 }, { "grad_norm": 0.2501329481601715, "learning_rate": 6.820774165553479e-05, "loss": 0.0218, "step": 41230 }, { "grad_norm": 0.25134506821632385, "learning_rate": 6.819234126051072e-05, "loss": 0.0223, "step": 41240 }, { "grad_norm": 0.19567129015922546, "learning_rate": 6.817693887599956e-05, "loss": 0.0237, "step": 41250 }, { "grad_norm": 0.9560220837593079, "learning_rate": 6.816153450368568e-05, "loss": 0.0255, "step": 41260 }, { "grad_norm": 0.28273680806159973, "learning_rate": 6.814612814525372e-05, "loss": 0.0277, "step": 41270 }, { "grad_norm": 0.1921355277299881, "learning_rate": 6.813071980238846e-05, "loss": 0.0249, "step": 41280 }, { "grad_norm": 0.26599952578544617, "learning_rate": 6.811530947677492e-05, "loss": 0.0261, "step": 41290 }, { "grad_norm": 0.2602590322494507, "learning_rate": 6.809989717009839e-05, "loss": 0.0268, "step": 41300 }, { "grad_norm": 0.18971988558769226, "learning_rate": 6.808448288404431e-05, "loss": 0.0285, "step": 41310 }, { "grad_norm": 0.2801395356655121, "learning_rate": 6.806906662029838e-05, "loss": 0.0253, "step": 41320 }, { "grad_norm": 0.32163140177726746, "learning_rate": 6.805364838054647e-05, "loss": 0.0268, "step": 41330 }, { "grad_norm": 0.22957353293895721, "learning_rate": 6.803822816647474e-05, "loss": 0.0206, "step": 41340 }, { "grad_norm": 0.24516575038433075, "learning_rate": 6.802280597976949e-05, "loss": 0.0202, "step": 41350 }, { "grad_norm": 0.17125461995601654, "learning_rate": 6.800738182211727e-05, "loss": 0.0253, "step": 41360 }, { "grad_norm": 0.29415732622146606, "learning_rate": 6.799195569520485e-05, "loss": 0.0266, "step": 41370 }, { "grad_norm": 0.28847426176071167, "learning_rate": 6.79765276007192e-05, "loss": 0.0254, "step": 41380 }, { "grad_norm": 0.2147148996591568, "learning_rate": 6.796109754034753e-05, "loss": 0.0284, "step": 41390 }, { "grad_norm": 0.21991169452667236, "learning_rate": 6.794566551577724e-05, "loss": 0.0238, "step": 41400 }, { "grad_norm": 0.2342081069946289, "learning_rate": 6.793023152869594e-05, "loss": 0.0229, "step": 41410 }, { "grad_norm": 0.19770771265029907, "learning_rate": 6.791479558079147e-05, "loss": 0.0271, "step": 41420 }, { "grad_norm": 0.2078704684972763, "learning_rate": 6.789935767375191e-05, "loss": 0.0258, "step": 41430 }, { "grad_norm": 0.2296660989522934, "learning_rate": 6.78839178092655e-05, "loss": 0.0247, "step": 41440 }, { "grad_norm": 0.3125336170196533, "learning_rate": 6.786847598902072e-05, "loss": 0.0278, "step": 41450 }, { "grad_norm": 0.3129180073738098, "learning_rate": 6.785303221470627e-05, "loss": 0.0254, "step": 41460 }, { "grad_norm": 0.29882022738456726, "learning_rate": 6.783758648801108e-05, "loss": 0.0283, "step": 41470 }, { "grad_norm": 0.28827229142189026, "learning_rate": 6.782213881062422e-05, "loss": 0.0254, "step": 41480 }, { "grad_norm": 0.19377543032169342, "learning_rate": 6.780668918423508e-05, "loss": 0.0311, "step": 41490 }, { "grad_norm": 0.30115851759910583, "learning_rate": 6.779123761053317e-05, "loss": 0.0248, "step": 41500 }, { "grad_norm": 0.2326255887746811, "learning_rate": 6.777578409120826e-05, "loss": 0.0263, "step": 41510 }, { "grad_norm": 0.18137796223163605, "learning_rate": 6.776032862795033e-05, "loss": 0.0265, "step": 41520 }, { "grad_norm": 0.28626033663749695, "learning_rate": 6.774487122244956e-05, "loss": 0.0299, "step": 41530 }, { "grad_norm": 0.18536578118801117, "learning_rate": 6.772941187639637e-05, "loss": 0.0251, "step": 41540 }, { "grad_norm": 0.2822231948375702, "learning_rate": 6.771395059148134e-05, "loss": 0.0247, "step": 41550 }, { "grad_norm": 0.3317604959011078, "learning_rate": 6.76984873693953e-05, "loss": 0.0254, "step": 41560 }, { "grad_norm": 0.22872258722782135, "learning_rate": 6.768302221182931e-05, "loss": 0.022, "step": 41570 }, { "grad_norm": 0.24016329646110535, "learning_rate": 6.766755512047457e-05, "loss": 0.0251, "step": 41580 }, { "grad_norm": 0.20918072760105133, "learning_rate": 6.765208609702259e-05, "loss": 0.027, "step": 41590 }, { "grad_norm": 0.25765329599380493, "learning_rate": 6.763661514316499e-05, "loss": 0.0282, "step": 41600 }, { "grad_norm": 0.24107471108436584, "learning_rate": 6.76211422605937e-05, "loss": 0.0255, "step": 41610 }, { "grad_norm": 0.22543273866176605, "learning_rate": 6.760566745100076e-05, "loss": 0.0241, "step": 41620 }, { "grad_norm": 0.3229163587093353, "learning_rate": 6.75901907160785e-05, "loss": 0.0275, "step": 41630 }, { "grad_norm": 0.22968561947345734, "learning_rate": 6.757471205751943e-05, "loss": 0.0231, "step": 41640 }, { "grad_norm": 0.47563597559928894, "learning_rate": 6.75592314770163e-05, "loss": 0.0259, "step": 41650 }, { "grad_norm": 0.22726181149482727, "learning_rate": 6.7543748976262e-05, "loss": 0.0252, "step": 41660 }, { "grad_norm": 0.1976424753665924, "learning_rate": 6.752826455694968e-05, "loss": 0.0253, "step": 41670 }, { "grad_norm": 0.19460119307041168, "learning_rate": 6.751277822077271e-05, "loss": 0.0261, "step": 41680 }, { "grad_norm": 0.19415418803691864, "learning_rate": 6.749728996942465e-05, "loss": 0.0263, "step": 41690 }, { "grad_norm": 0.2408449351787567, "learning_rate": 6.748179980459924e-05, "loss": 0.0257, "step": 41700 }, { "grad_norm": 0.2706194519996643, "learning_rate": 6.746630772799052e-05, "loss": 0.027, "step": 41710 }, { "grad_norm": 0.39695224165916443, "learning_rate": 6.745081374129262e-05, "loss": 0.0278, "step": 41720 }, { "grad_norm": 0.23435647785663605, "learning_rate": 6.743531784619998e-05, "loss": 0.0211, "step": 41730 }, { "grad_norm": 0.26185327768325806, "learning_rate": 6.741982004440719e-05, "loss": 0.0271, "step": 41740 }, { "grad_norm": 0.23665501177310944, "learning_rate": 6.740432033760907e-05, "loss": 0.0222, "step": 41750 }, { "grad_norm": 0.3201254904270172, "learning_rate": 6.738881872750066e-05, "loss": 0.0359, "step": 41760 }, { "grad_norm": 0.7602680325508118, "learning_rate": 6.737331521577715e-05, "loss": 0.042, "step": 41770 }, { "grad_norm": 0.2672409415245056, "learning_rate": 6.735780980413403e-05, "loss": 0.0332, "step": 41780 }, { "grad_norm": 0.2561756372451782, "learning_rate": 6.734230249426692e-05, "loss": 0.0268, "step": 41790 }, { "grad_norm": 0.27316126227378845, "learning_rate": 6.732679328787168e-05, "loss": 0.027, "step": 41800 }, { "grad_norm": 0.42679494619369507, "learning_rate": 6.731128218664438e-05, "loss": 0.0283, "step": 41810 }, { "grad_norm": 0.3396759629249573, "learning_rate": 6.729576919228129e-05, "loss": 0.0246, "step": 41820 }, { "grad_norm": 0.30751174688339233, "learning_rate": 6.728025430647888e-05, "loss": 0.0242, "step": 41830 }, { "grad_norm": 0.263641893863678, "learning_rate": 6.726473753093383e-05, "loss": 0.0274, "step": 41840 }, { "grad_norm": 0.1599244922399521, "learning_rate": 6.724921886734305e-05, "loss": 0.0231, "step": 41850 }, { "grad_norm": 0.3949120342731476, "learning_rate": 6.723369831740363e-05, "loss": 0.0266, "step": 41860 }, { "grad_norm": 0.2829873561859131, "learning_rate": 6.721817588281288e-05, "loss": 0.0276, "step": 41870 }, { "grad_norm": 0.28247302770614624, "learning_rate": 6.720265156526828e-05, "loss": 0.027, "step": 41880 }, { "grad_norm": 0.19690416753292084, "learning_rate": 6.718712536646758e-05, "loss": 0.0245, "step": 41890 }, { "grad_norm": 0.20875462889671326, "learning_rate": 6.71715972881087e-05, "loss": 0.0228, "step": 41900 }, { "grad_norm": 0.19933980703353882, "learning_rate": 6.715606733188974e-05, "loss": 0.022, "step": 41910 }, { "grad_norm": 0.17795592546463013, "learning_rate": 6.714053549950909e-05, "loss": 0.0243, "step": 41920 }, { "grad_norm": 0.35620802640914917, "learning_rate": 6.71250017926652e-05, "loss": 0.0269, "step": 41930 }, { "grad_norm": 0.19078636169433594, "learning_rate": 6.71094662130569e-05, "loss": 0.0228, "step": 41940 }, { "grad_norm": 0.20784500241279602, "learning_rate": 6.709392876238307e-05, "loss": 0.0247, "step": 41950 }, { "grad_norm": 0.2290576547384262, "learning_rate": 6.707838944234289e-05, "loss": 0.0266, "step": 41960 }, { "grad_norm": 0.18441633880138397, "learning_rate": 6.706284825463574e-05, "loss": 0.0204, "step": 41970 }, { "grad_norm": 0.2311619371175766, "learning_rate": 6.704730520096114e-05, "loss": 0.0218, "step": 41980 }, { "grad_norm": 0.27782219648361206, "learning_rate": 6.703176028301888e-05, "loss": 0.0271, "step": 41990 }, { "grad_norm": 0.39164409041404724, "learning_rate": 6.701621350250892e-05, "loss": 0.0258, "step": 42000 }, { "grad_norm": 0.1912223994731903, "learning_rate": 6.700066486113144e-05, "loss": 0.025, "step": 42010 }, { "grad_norm": 0.22402003407478333, "learning_rate": 6.69851143605868e-05, "loss": 0.0203, "step": 42020 }, { "grad_norm": 0.20504610240459442, "learning_rate": 6.696956200257561e-05, "loss": 0.0227, "step": 42030 }, { "grad_norm": 0.2622385323047638, "learning_rate": 6.695400778879863e-05, "loss": 0.0272, "step": 42040 }, { "grad_norm": 0.16285856068134308, "learning_rate": 6.693845172095683e-05, "loss": 0.0278, "step": 42050 }, { "grad_norm": 0.6658533215522766, "learning_rate": 6.692289380075142e-05, "loss": 0.0231, "step": 42060 }, { "grad_norm": 0.372952401638031, "learning_rate": 6.690733402988379e-05, "loss": 0.0241, "step": 42070 }, { "grad_norm": 0.268014520406723, "learning_rate": 6.689177241005553e-05, "loss": 0.023, "step": 42080 }, { "grad_norm": 0.2189565896987915, "learning_rate": 6.687620894296844e-05, "loss": 0.0264, "step": 42090 }, { "grad_norm": 0.24469472467899323, "learning_rate": 6.686064363032451e-05, "loss": 0.0275, "step": 42100 }, { "grad_norm": 0.18845094740390778, "learning_rate": 6.684507647382596e-05, "loss": 0.0193, "step": 42110 }, { "grad_norm": 0.2879035174846649, "learning_rate": 6.682950747517513e-05, "loss": 0.0221, "step": 42120 }, { "grad_norm": 0.22025233507156372, "learning_rate": 6.68139366360747e-05, "loss": 0.024, "step": 42130 }, { "grad_norm": 0.2655129134654999, "learning_rate": 6.679836395822744e-05, "loss": 0.0209, "step": 42140 }, { "grad_norm": 0.21639375388622284, "learning_rate": 6.678278944333633e-05, "loss": 0.0256, "step": 42150 }, { "grad_norm": 0.22478331625461578, "learning_rate": 6.676721309310462e-05, "loss": 0.022, "step": 42160 }, { "grad_norm": 0.25425171852111816, "learning_rate": 6.67516349092357e-05, "loss": 0.0226, "step": 42170 }, { "grad_norm": 0.21740534901618958, "learning_rate": 6.673605489343315e-05, "loss": 0.0229, "step": 42180 }, { "grad_norm": 0.22008687257766724, "learning_rate": 6.672047304740081e-05, "loss": 0.0196, "step": 42190 }, { "grad_norm": 0.22607208788394928, "learning_rate": 6.670488937284268e-05, "loss": 0.0236, "step": 42200 }, { "grad_norm": 0.15119898319244385, "learning_rate": 6.668930387146294e-05, "loss": 0.0202, "step": 42210 }, { "grad_norm": 0.32785382866859436, "learning_rate": 6.667371654496605e-05, "loss": 0.027, "step": 42220 }, { "grad_norm": 0.1345720887184143, "learning_rate": 6.665812739505659e-05, "loss": 0.0248, "step": 42230 }, { "grad_norm": 0.1933020055294037, "learning_rate": 6.664253642343935e-05, "loss": 0.0234, "step": 42240 }, { "grad_norm": 0.18383683264255524, "learning_rate": 6.662694363181935e-05, "loss": 0.0213, "step": 42250 }, { "grad_norm": 0.1626090705394745, "learning_rate": 6.66113490219018e-05, "loss": 0.0215, "step": 42260 }, { "grad_norm": 0.14823776483535767, "learning_rate": 6.659575259539207e-05, "loss": 0.0183, "step": 42270 }, { "grad_norm": 0.2505285143852234, "learning_rate": 6.65801543539958e-05, "loss": 0.0243, "step": 42280 }, { "grad_norm": 0.15526527166366577, "learning_rate": 6.65645542994188e-05, "loss": 0.0245, "step": 42290 }, { "grad_norm": 0.13752232491970062, "learning_rate": 6.654895243336702e-05, "loss": 0.0206, "step": 42300 }, { "grad_norm": 0.24299146234989166, "learning_rate": 6.65333487575467e-05, "loss": 0.0233, "step": 42310 }, { "grad_norm": 0.2010614275932312, "learning_rate": 6.651774327366422e-05, "loss": 0.0228, "step": 42320 }, { "grad_norm": 0.2141333669424057, "learning_rate": 6.650213598342613e-05, "loss": 0.0202, "step": 42330 }, { "grad_norm": 0.2444155216217041, "learning_rate": 6.648652688853931e-05, "loss": 0.0206, "step": 42340 }, { "grad_norm": 0.1913856714963913, "learning_rate": 6.647091599071066e-05, "loss": 0.027, "step": 42350 }, { "grad_norm": 0.17944370210170746, "learning_rate": 6.645530329164743e-05, "loss": 0.0228, "step": 42360 }, { "grad_norm": 0.24800652265548706, "learning_rate": 6.643968879305694e-05, "loss": 0.0288, "step": 42370 }, { "grad_norm": 0.2481638640165329, "learning_rate": 6.642407249664683e-05, "loss": 0.0242, "step": 42380 }, { "grad_norm": 0.1832938939332962, "learning_rate": 6.640845440412483e-05, "loss": 0.0198, "step": 42390 }, { "grad_norm": 0.19546690583229065, "learning_rate": 6.639283451719893e-05, "loss": 0.0265, "step": 42400 }, { "grad_norm": 0.28303006291389465, "learning_rate": 6.637721283757729e-05, "loss": 0.0236, "step": 42410 }, { "grad_norm": 0.4774893820285797, "learning_rate": 6.636158936696828e-05, "loss": 0.022, "step": 42420 }, { "grad_norm": 0.25731945037841797, "learning_rate": 6.634596410708047e-05, "loss": 0.0271, "step": 42430 }, { "grad_norm": 0.3202654719352722, "learning_rate": 6.633033705962257e-05, "loss": 0.0244, "step": 42440 }, { "grad_norm": 0.34908661246299744, "learning_rate": 6.631470822630359e-05, "loss": 0.0242, "step": 42450 }, { "grad_norm": 0.21668170392513275, "learning_rate": 6.629907760883263e-05, "loss": 0.024, "step": 42460 }, { "grad_norm": 0.21060296893119812, "learning_rate": 6.628344520891905e-05, "loss": 0.0215, "step": 42470 }, { "grad_norm": 0.29053619503974915, "learning_rate": 6.62678110282724e-05, "loss": 0.023, "step": 42480 }, { "grad_norm": 0.2094675451517105, "learning_rate": 6.625217506860237e-05, "loss": 0.0247, "step": 42490 }, { "grad_norm": 0.22936998307704926, "learning_rate": 6.623653733161892e-05, "loss": 0.0225, "step": 42500 }, { "grad_norm": 0.1896650195121765, "learning_rate": 6.622089781903217e-05, "loss": 0.0228, "step": 42510 }, { "grad_norm": 0.200737863779068, "learning_rate": 6.620525653255241e-05, "loss": 0.0213, "step": 42520 }, { "grad_norm": 0.34357503056526184, "learning_rate": 6.618961347389015e-05, "loss": 0.0268, "step": 42530 }, { "grad_norm": 0.3355344533920288, "learning_rate": 6.617396864475613e-05, "loss": 0.0238, "step": 42540 }, { "grad_norm": 0.2569078207015991, "learning_rate": 6.61583220468612e-05, "loss": 0.0242, "step": 42550 }, { "grad_norm": 0.2411220818758011, "learning_rate": 6.614267368191645e-05, "loss": 0.0215, "step": 42560 }, { "grad_norm": 0.21064350008964539, "learning_rate": 6.61270235516332e-05, "loss": 0.0196, "step": 42570 }, { "grad_norm": 0.28308242559432983, "learning_rate": 6.611137165772287e-05, "loss": 0.0254, "step": 42580 }, { "grad_norm": 0.24934519827365875, "learning_rate": 6.60957180018972e-05, "loss": 0.0311, "step": 42590 }, { "grad_norm": 0.24497175216674805, "learning_rate": 6.608006258586797e-05, "loss": 0.0222, "step": 42600 }, { "grad_norm": 0.19189949333667755, "learning_rate": 6.606440541134731e-05, "loss": 0.0262, "step": 42610 }, { "grad_norm": 0.2635868489742279, "learning_rate": 6.60487464800474e-05, "loss": 0.0236, "step": 42620 }, { "grad_norm": 0.18359002470970154, "learning_rate": 6.603308579368071e-05, "loss": 0.0216, "step": 42630 }, { "grad_norm": 0.20220158994197845, "learning_rate": 6.601742335395987e-05, "loss": 0.023, "step": 42640 }, { "grad_norm": 0.21704822778701782, "learning_rate": 6.600175916259769e-05, "loss": 0.0212, "step": 42650 }, { "grad_norm": 0.18279637396335602, "learning_rate": 6.598609322130718e-05, "loss": 0.0212, "step": 42660 }, { "grad_norm": 0.1526821255683899, "learning_rate": 6.597042553180155e-05, "loss": 0.0249, "step": 42670 }, { "grad_norm": 0.13801634311676025, "learning_rate": 6.59547560957942e-05, "loss": 0.0207, "step": 42680 }, { "grad_norm": 0.17270413041114807, "learning_rate": 6.59390849149987e-05, "loss": 0.0203, "step": 42690 }, { "grad_norm": 0.21093139052391052, "learning_rate": 6.592341199112886e-05, "loss": 0.0257, "step": 42700 }, { "grad_norm": 0.20415110886096954, "learning_rate": 6.590773732589861e-05, "loss": 0.0229, "step": 42710 }, { "grad_norm": 0.30862465500831604, "learning_rate": 6.589206092102213e-05, "loss": 0.026, "step": 42720 }, { "grad_norm": 0.28044697642326355, "learning_rate": 6.587638277821375e-05, "loss": 0.0248, "step": 42730 }, { "grad_norm": 0.19931092858314514, "learning_rate": 6.586070289918803e-05, "loss": 0.0248, "step": 42740 }, { "grad_norm": 0.3046867549419403, "learning_rate": 6.584502128565968e-05, "loss": 0.0245, "step": 42750 }, { "grad_norm": 0.16780221462249756, "learning_rate": 6.582933793934363e-05, "loss": 0.0212, "step": 42760 }, { "grad_norm": 0.21516063809394836, "learning_rate": 6.581365286195498e-05, "loss": 0.0238, "step": 42770 }, { "grad_norm": 0.24968357384204865, "learning_rate": 6.579796605520903e-05, "loss": 0.0223, "step": 42780 }, { "grad_norm": 0.23597683012485504, "learning_rate": 6.578227752082127e-05, "loss": 0.0219, "step": 42790 }, { "grad_norm": 0.32303914427757263, "learning_rate": 6.576658726050735e-05, "loss": 0.0235, "step": 42800 }, { "grad_norm": 0.2227783352136612, "learning_rate": 6.575089527598316e-05, "loss": 0.023, "step": 42810 }, { "grad_norm": 0.7025042176246643, "learning_rate": 6.573520156896476e-05, "loss": 0.027, "step": 42820 }, { "grad_norm": 0.21042543649673462, "learning_rate": 6.571950614116835e-05, "loss": 0.0272, "step": 42830 }, { "grad_norm": 0.210942804813385, "learning_rate": 6.570380899431039e-05, "loss": 0.0231, "step": 42840 }, { "grad_norm": 0.22441835701465607, "learning_rate": 6.568811013010749e-05, "loss": 0.0227, "step": 42850 }, { "grad_norm": 0.1975577473640442, "learning_rate": 6.567240955027645e-05, "loss": 0.022, "step": 42860 }, { "grad_norm": 0.21182869374752045, "learning_rate": 6.565670725653427e-05, "loss": 0.0246, "step": 42870 }, { "grad_norm": 0.19172623753547668, "learning_rate": 6.564100325059812e-05, "loss": 0.0214, "step": 42880 }, { "grad_norm": 0.2364661693572998, "learning_rate": 6.562529753418538e-05, "loss": 0.0259, "step": 42890 }, { "grad_norm": 0.20083533227443695, "learning_rate": 6.56095901090136e-05, "loss": 0.0234, "step": 42900 }, { "grad_norm": 0.17716455459594727, "learning_rate": 6.55938809768005e-05, "loss": 0.0216, "step": 42910 }, { "grad_norm": 0.25404465198516846, "learning_rate": 6.557817013926402e-05, "loss": 0.0206, "step": 42920 }, { "grad_norm": 0.2008349597454071, "learning_rate": 6.556245759812227e-05, "loss": 0.018, "step": 42930 }, { "grad_norm": 0.24049361050128937, "learning_rate": 6.554674335509357e-05, "loss": 0.0214, "step": 42940 }, { "grad_norm": 0.2645069360733032, "learning_rate": 6.553102741189638e-05, "loss": 0.0216, "step": 42950 }, { "grad_norm": 0.3107368052005768, "learning_rate": 6.55153097702494e-05, "loss": 0.0273, "step": 42960 }, { "grad_norm": 0.20095865428447723, "learning_rate": 6.549959043187146e-05, "loss": 0.025, "step": 42970 }, { "grad_norm": 0.22008675336837769, "learning_rate": 6.54838693984816e-05, "loss": 0.0237, "step": 42980 }, { "grad_norm": 0.19723983108997345, "learning_rate": 6.546814667179909e-05, "loss": 0.0285, "step": 42990 }, { "grad_norm": 0.2533903121948242, "learning_rate": 6.545242225354328e-05, "loss": 0.0252, "step": 43000 }, { "grad_norm": 0.2491704672574997, "learning_rate": 6.543669614543383e-05, "loss": 0.0243, "step": 43010 }, { "grad_norm": 0.15769782662391663, "learning_rate": 6.542096834919049e-05, "loss": 0.0246, "step": 43020 }, { "grad_norm": 0.22514067590236664, "learning_rate": 6.540523886653324e-05, "loss": 0.0245, "step": 43030 }, { "grad_norm": 0.3080812990665436, "learning_rate": 6.538950769918222e-05, "loss": 0.0228, "step": 43040 }, { "grad_norm": 0.17306122183799744, "learning_rate": 6.537377484885779e-05, "loss": 0.0228, "step": 43050 }, { "grad_norm": 0.2047380954027176, "learning_rate": 6.535804031728045e-05, "loss": 0.0197, "step": 43060 }, { "grad_norm": 0.2776724696159363, "learning_rate": 6.534230410617092e-05, "loss": 0.0218, "step": 43070 }, { "grad_norm": 0.17418265342712402, "learning_rate": 6.532656621725007e-05, "loss": 0.0229, "step": 43080 }, { "grad_norm": 0.22037141025066376, "learning_rate": 6.531082665223898e-05, "loss": 0.0214, "step": 43090 }, { "grad_norm": 0.22651661932468414, "learning_rate": 6.529508541285889e-05, "loss": 0.0218, "step": 43100 }, { "grad_norm": 0.21152658760547638, "learning_rate": 6.527934250083128e-05, "loss": 0.0215, "step": 43110 }, { "grad_norm": 0.1988895684480667, "learning_rate": 6.526359791787775e-05, "loss": 0.0254, "step": 43120 }, { "grad_norm": 0.21599674224853516, "learning_rate": 6.524785166572009e-05, "loss": 0.0197, "step": 43130 }, { "grad_norm": 0.43271493911743164, "learning_rate": 6.523210374608029e-05, "loss": 0.0229, "step": 43140 }, { "grad_norm": 0.26525890827178955, "learning_rate": 6.521635416068054e-05, "loss": 0.0251, "step": 43150 }, { "grad_norm": 0.4011283814907074, "learning_rate": 6.520060291124317e-05, "loss": 0.0253, "step": 43160 }, { "grad_norm": 0.2002846747636795, "learning_rate": 6.518484999949072e-05, "loss": 0.0204, "step": 43170 }, { "grad_norm": 0.2684672772884369, "learning_rate": 6.516909542714591e-05, "loss": 0.027, "step": 43180 }, { "grad_norm": 0.23986925184726715, "learning_rate": 6.515333919593161e-05, "loss": 0.0267, "step": 43190 }, { "grad_norm": 0.2039283812046051, "learning_rate": 6.513758130757094e-05, "loss": 0.0258, "step": 43200 }, { "grad_norm": 0.24238499999046326, "learning_rate": 6.512182176378713e-05, "loss": 0.0259, "step": 43210 }, { "grad_norm": 0.2265070229768753, "learning_rate": 6.510606056630362e-05, "loss": 0.025, "step": 43220 }, { "grad_norm": 0.33180779218673706, "learning_rate": 6.509029771684405e-05, "loss": 0.0217, "step": 43230 }, { "grad_norm": 0.32899633049964905, "learning_rate": 6.50745332171322e-05, "loss": 0.0256, "step": 43240 }, { "grad_norm": 0.23434790968894958, "learning_rate": 6.505876706889207e-05, "loss": 0.0213, "step": 43250 }, { "grad_norm": 0.19866855442523956, "learning_rate": 6.504299927384781e-05, "loss": 0.021, "step": 43260 }, { "grad_norm": 0.19232749938964844, "learning_rate": 6.502722983372378e-05, "loss": 0.0223, "step": 43270 }, { "grad_norm": 0.1817840188741684, "learning_rate": 6.501145875024446e-05, "loss": 0.0193, "step": 43280 }, { "grad_norm": 0.15248379111289978, "learning_rate": 6.499568602513459e-05, "loss": 0.0225, "step": 43290 }, { "grad_norm": 0.2588202655315399, "learning_rate": 6.497991166011903e-05, "loss": 0.0207, "step": 43300 }, { "grad_norm": 0.26261013746261597, "learning_rate": 6.49641356569229e-05, "loss": 0.0198, "step": 43310 }, { "grad_norm": 0.25674113631248474, "learning_rate": 6.494835801727135e-05, "loss": 0.028, "step": 43320 }, { "grad_norm": 0.2364465892314911, "learning_rate": 6.493257874288986e-05, "loss": 0.0222, "step": 43330 }, { "grad_norm": 0.23324653506278992, "learning_rate": 6.491679783550399e-05, "loss": 0.0266, "step": 43340 }, { "grad_norm": 0.283186674118042, "learning_rate": 6.490101529683957e-05, "loss": 0.0272, "step": 43350 }, { "grad_norm": 0.19554467499256134, "learning_rate": 6.488523112862248e-05, "loss": 0.0208, "step": 43360 }, { "grad_norm": 0.17981866002082825, "learning_rate": 6.486944533257891e-05, "loss": 0.0232, "step": 43370 }, { "grad_norm": 0.21816295385360718, "learning_rate": 6.485365791043517e-05, "loss": 0.0237, "step": 43380 }, { "grad_norm": 0.19851116836071014, "learning_rate": 6.483786886391772e-05, "loss": 0.0258, "step": 43390 }, { "grad_norm": 0.19321832060813904, "learning_rate": 6.482207819475323e-05, "loss": 0.0239, "step": 43400 }, { "grad_norm": 0.2140846699476242, "learning_rate": 6.480628590466857e-05, "loss": 0.0244, "step": 43410 }, { "grad_norm": 0.25656458735466003, "learning_rate": 6.479049199539074e-05, "loss": 0.0247, "step": 43420 }, { "grad_norm": 0.2602842450141907, "learning_rate": 6.477469646864693e-05, "loss": 0.0267, "step": 43430 }, { "grad_norm": 0.27014341950416565, "learning_rate": 6.475889932616454e-05, "loss": 0.0267, "step": 43440 }, { "grad_norm": 0.19395099580287933, "learning_rate": 6.474310056967111e-05, "loss": 0.0248, "step": 43450 }, { "grad_norm": 0.17918530106544495, "learning_rate": 6.472730020089437e-05, "loss": 0.0244, "step": 43460 }, { "grad_norm": 0.18439047038555145, "learning_rate": 6.471149822156222e-05, "loss": 0.0211, "step": 43470 }, { "grad_norm": 0.2273205816745758, "learning_rate": 6.469569463340275e-05, "loss": 0.0196, "step": 43480 }, { "grad_norm": 0.2180095762014389, "learning_rate": 6.467988943814418e-05, "loss": 0.0235, "step": 43490 }, { "grad_norm": 0.17092491686344147, "learning_rate": 6.4664082637515e-05, "loss": 0.0231, "step": 43500 }, { "grad_norm": 0.244134321808815, "learning_rate": 6.464827423324377e-05, "loss": 0.0201, "step": 43510 }, { "grad_norm": 0.2487846314907074, "learning_rate": 6.463246422705931e-05, "loss": 0.0228, "step": 43520 }, { "grad_norm": 0.2268560230731964, "learning_rate": 6.461665262069055e-05, "loss": 0.0212, "step": 43530 }, { "grad_norm": 0.21145392954349518, "learning_rate": 6.460083941586665e-05, "loss": 0.0262, "step": 43540 }, { "grad_norm": 0.17356577515602112, "learning_rate": 6.45850246143169e-05, "loss": 0.0203, "step": 43550 }, { "grad_norm": 0.33545735478401184, "learning_rate": 6.456920821777077e-05, "loss": 0.0226, "step": 43560 }, { "grad_norm": 0.3069111108779907, "learning_rate": 6.455339022795796e-05, "loss": 0.023, "step": 43570 }, { "grad_norm": 0.37168872356414795, "learning_rate": 6.453757064660825e-05, "loss": 0.0232, "step": 43580 }, { "grad_norm": 0.266202837228775, "learning_rate": 6.452174947545169e-05, "loss": 0.0251, "step": 43590 }, { "grad_norm": 0.21588245034217834, "learning_rate": 6.450592671621842e-05, "loss": 0.0217, "step": 43600 }, { "grad_norm": 0.20213918387889862, "learning_rate": 6.449010237063882e-05, "loss": 0.0235, "step": 43610 }, { "grad_norm": 0.2289634495973587, "learning_rate": 6.447427644044343e-05, "loss": 0.0217, "step": 43620 }, { "grad_norm": 0.20756952464580536, "learning_rate": 6.44584489273629e-05, "loss": 0.0238, "step": 43630 }, { "grad_norm": 0.19651542603969574, "learning_rate": 6.444261983312817e-05, "loss": 0.0198, "step": 43640 }, { "grad_norm": 0.13360020518302917, "learning_rate": 6.442678915947023e-05, "loss": 0.0233, "step": 43650 }, { "grad_norm": 0.160300612449646, "learning_rate": 6.441095690812032e-05, "loss": 0.0224, "step": 43660 }, { "grad_norm": 0.21196326613426208, "learning_rate": 6.439512308080982e-05, "loss": 0.022, "step": 43670 }, { "grad_norm": 0.20707465708255768, "learning_rate": 6.43792876792703e-05, "loss": 0.0188, "step": 43680 }, { "grad_norm": 0.2628493010997772, "learning_rate": 6.43634507052335e-05, "loss": 0.025, "step": 43690 }, { "grad_norm": 0.26791247725486755, "learning_rate": 6.434761216043135e-05, "loss": 0.0256, "step": 43700 }, { "grad_norm": 0.2300543338060379, "learning_rate": 6.433177204659589e-05, "loss": 0.0251, "step": 43710 }, { "grad_norm": 0.28048527240753174, "learning_rate": 6.431593036545938e-05, "loss": 0.0259, "step": 43720 }, { "grad_norm": 0.23190785944461823, "learning_rate": 6.430008711875426e-05, "loss": 0.0251, "step": 43730 }, { "grad_norm": 0.2769015431404114, "learning_rate": 6.42842423082131e-05, "loss": 0.0297, "step": 43740 }, { "grad_norm": 0.15932786464691162, "learning_rate": 6.426839593556868e-05, "loss": 0.0248, "step": 43750 }, { "grad_norm": 0.2253238558769226, "learning_rate": 6.425254800255392e-05, "loss": 0.0246, "step": 43760 }, { "grad_norm": 0.16702647507190704, "learning_rate": 6.423669851090196e-05, "loss": 0.0229, "step": 43770 }, { "grad_norm": 0.20769838988780975, "learning_rate": 6.422084746234604e-05, "loss": 0.0236, "step": 43780 }, { "grad_norm": 0.2020024061203003, "learning_rate": 6.420499485861965e-05, "loss": 0.0239, "step": 43790 }, { "grad_norm": 0.182617649435997, "learning_rate": 6.418914070145636e-05, "loss": 0.0247, "step": 43800 }, { "grad_norm": 0.26122692227363586, "learning_rate": 6.417328499258996e-05, "loss": 0.0247, "step": 43810 }, { "grad_norm": 0.25980454683303833, "learning_rate": 6.415742773375446e-05, "loss": 0.0221, "step": 43820 }, { "grad_norm": 0.28244534134864807, "learning_rate": 6.414156892668393e-05, "loss": 0.0225, "step": 43830 }, { "grad_norm": 0.430751770734787, "learning_rate": 6.412570857311267e-05, "loss": 0.0215, "step": 43840 }, { "grad_norm": 0.232910618185997, "learning_rate": 6.410984667477518e-05, "loss": 0.0227, "step": 43850 }, { "grad_norm": 0.15494883060455322, "learning_rate": 6.409398323340607e-05, "loss": 0.0194, "step": 43860 }, { "grad_norm": 0.18235300481319427, "learning_rate": 6.407811825074012e-05, "loss": 0.0265, "step": 43870 }, { "grad_norm": 0.21159076690673828, "learning_rate": 6.406225172851234e-05, "loss": 0.0235, "step": 43880 }, { "grad_norm": 0.2131901979446411, "learning_rate": 6.404638366845786e-05, "loss": 0.022, "step": 43890 }, { "grad_norm": 0.16487234830856323, "learning_rate": 6.403051407231196e-05, "loss": 0.0233, "step": 43900 }, { "grad_norm": 0.23903989791870117, "learning_rate": 6.401464294181016e-05, "loss": 0.0255, "step": 43910 }, { "grad_norm": 0.20654639601707458, "learning_rate": 6.399877027868808e-05, "loss": 0.0244, "step": 43920 }, { "grad_norm": 0.290409654378891, "learning_rate": 6.39828960846815e-05, "loss": 0.0282, "step": 43930 }, { "grad_norm": 0.1685761958360672, "learning_rate": 6.396702036152644e-05, "loss": 0.0238, "step": 43940 }, { "grad_norm": 0.233814999461174, "learning_rate": 6.395114311095905e-05, "loss": 0.0295, "step": 43950 }, { "grad_norm": 0.21625176072120667, "learning_rate": 6.393526433471559e-05, "loss": 0.0254, "step": 43960 }, { "grad_norm": 0.25147902965545654, "learning_rate": 6.391938403453262e-05, "loss": 0.0238, "step": 43970 }, { "grad_norm": 0.19311214983463287, "learning_rate": 6.390350221214671e-05, "loss": 0.0219, "step": 43980 }, { "grad_norm": 0.4294302463531494, "learning_rate": 6.38876188692947e-05, "loss": 0.0241, "step": 43990 }, { "grad_norm": 0.38484954833984375, "learning_rate": 6.387173400771358e-05, "loss": 0.0264, "step": 44000 }, { "grad_norm": 0.22090353071689606, "learning_rate": 6.385584762914047e-05, "loss": 0.0245, "step": 44010 }, { "grad_norm": 0.30160701274871826, "learning_rate": 6.38399597353127e-05, "loss": 0.0256, "step": 44020 }, { "grad_norm": 0.24661079049110413, "learning_rate": 6.382407032796775e-05, "loss": 0.0273, "step": 44030 }, { "grad_norm": 0.38816317915916443, "learning_rate": 6.380817940884325e-05, "loss": 0.0279, "step": 44040 }, { "grad_norm": 0.2764394283294678, "learning_rate": 6.379228697967702e-05, "loss": 0.028, "step": 44050 }, { "grad_norm": 0.22024592757225037, "learning_rate": 6.3776393042207e-05, "loss": 0.0241, "step": 44060 }, { "grad_norm": 0.7382407784461975, "learning_rate": 6.376049759817137e-05, "loss": 0.0277, "step": 44070 }, { "grad_norm": 0.44563692808151245, "learning_rate": 6.37446006493084e-05, "loss": 0.0247, "step": 44080 }, { "grad_norm": 0.287371963262558, "learning_rate": 6.372870219735656e-05, "loss": 0.0278, "step": 44090 }, { "grad_norm": 0.33127376437187195, "learning_rate": 6.371280224405452e-05, "loss": 0.0261, "step": 44100 }, { "grad_norm": 0.23819339275360107, "learning_rate": 6.369690079114101e-05, "loss": 0.0239, "step": 44110 }, { "grad_norm": 0.7023623585700989, "learning_rate": 6.368099784035504e-05, "loss": 0.0273, "step": 44120 }, { "grad_norm": 0.37681663036346436, "learning_rate": 6.366509339343572e-05, "loss": 0.0232, "step": 44130 }, { "grad_norm": 0.35234397649765015, "learning_rate": 6.364918745212232e-05, "loss": 0.0277, "step": 44140 }, { "grad_norm": 0.20303094387054443, "learning_rate": 6.363328001815431e-05, "loss": 0.0275, "step": 44150 }, { "grad_norm": 0.27011364698410034, "learning_rate": 6.361737109327128e-05, "loss": 0.0233, "step": 44160 }, { "grad_norm": 0.2683143615722656, "learning_rate": 6.360146067921301e-05, "loss": 0.0222, "step": 44170 }, { "grad_norm": 0.3361094295978546, "learning_rate": 6.35855487777195e-05, "loss": 0.0213, "step": 44180 }, { "grad_norm": 0.14189164340496063, "learning_rate": 6.356963539053076e-05, "loss": 0.0218, "step": 44190 }, { "grad_norm": 0.1881239116191864, "learning_rate": 6.355372051938711e-05, "loss": 0.0252, "step": 44200 }, { "grad_norm": 0.4327743649482727, "learning_rate": 6.353780416602894e-05, "loss": 0.0244, "step": 44210 }, { "grad_norm": 0.18715637922286987, "learning_rate": 6.352188633219689e-05, "loss": 0.025, "step": 44220 }, { "grad_norm": 0.25199753046035767, "learning_rate": 6.350596701963166e-05, "loss": 0.0228, "step": 44230 }, { "grad_norm": 0.36719420552253723, "learning_rate": 6.349004623007419e-05, "loss": 0.0208, "step": 44240 }, { "grad_norm": 0.22423577308654785, "learning_rate": 6.347412396526555e-05, "loss": 0.022, "step": 44250 }, { "grad_norm": 0.19106313586235046, "learning_rate": 6.345820022694696e-05, "loss": 0.0313, "step": 44260 }, { "grad_norm": 0.19722560048103333, "learning_rate": 6.344227501685984e-05, "loss": 0.0235, "step": 44270 }, { "grad_norm": 0.187272846698761, "learning_rate": 6.342634833674572e-05, "loss": 0.0224, "step": 44280 }, { "grad_norm": 0.23437651991844177, "learning_rate": 6.341042018834635e-05, "loss": 0.0274, "step": 44290 }, { "grad_norm": 0.31735825538635254, "learning_rate": 6.339449057340359e-05, "loss": 0.026, "step": 44300 }, { "grad_norm": 0.1689332127571106, "learning_rate": 6.337855949365945e-05, "loss": 0.0253, "step": 44310 }, { "grad_norm": 0.22545208036899567, "learning_rate": 6.336262695085619e-05, "loss": 0.0226, "step": 44320 }, { "grad_norm": 0.36260443925857544, "learning_rate": 6.334669294673612e-05, "loss": 0.0239, "step": 44330 }, { "grad_norm": 0.20871011912822723, "learning_rate": 6.333075748304179e-05, "loss": 0.0242, "step": 44340 }, { "grad_norm": 0.20737716555595398, "learning_rate": 6.331482056151585e-05, "loss": 0.0201, "step": 44350 }, { "grad_norm": 0.352289080619812, "learning_rate": 6.329888218390117e-05, "loss": 0.0298, "step": 44360 }, { "grad_norm": 0.22937126457691193, "learning_rate": 6.328294235194072e-05, "loss": 0.0237, "step": 44370 }, { "grad_norm": 0.21380552649497986, "learning_rate": 6.326700106737765e-05, "loss": 0.0215, "step": 44380 }, { "grad_norm": 0.23045791685581207, "learning_rate": 6.325105833195531e-05, "loss": 0.0205, "step": 44390 }, { "grad_norm": 0.20668022334575653, "learning_rate": 6.323511414741715e-05, "loss": 0.0219, "step": 44400 }, { "grad_norm": 0.2074650675058365, "learning_rate": 6.321916851550678e-05, "loss": 0.022, "step": 44410 }, { "grad_norm": 0.1796041578054428, "learning_rate": 6.320322143796806e-05, "loss": 0.0223, "step": 44420 }, { "grad_norm": 0.2647894620895386, "learning_rate": 6.318727291654488e-05, "loss": 0.0232, "step": 44430 }, { "grad_norm": 0.35688328742980957, "learning_rate": 6.317132295298134e-05, "loss": 0.0236, "step": 44440 }, { "grad_norm": 0.20827095210552216, "learning_rate": 6.315537154902173e-05, "loss": 0.0215, "step": 44450 }, { "grad_norm": 0.18144789338111877, "learning_rate": 6.313941870641048e-05, "loss": 0.0208, "step": 44460 }, { "grad_norm": 0.25221025943756104, "learning_rate": 6.312346442689214e-05, "loss": 0.0187, "step": 44470 }, { "grad_norm": 0.2051842212677002, "learning_rate": 6.310750871221147e-05, "loss": 0.0263, "step": 44480 }, { "grad_norm": 0.20276246964931488, "learning_rate": 6.309155156411335e-05, "loss": 0.023, "step": 44490 }, { "grad_norm": 0.17229221761226654, "learning_rate": 6.307559298434284e-05, "loss": 0.0261, "step": 44500 }, { "grad_norm": 0.26464176177978516, "learning_rate": 6.305963297464512e-05, "loss": 0.0245, "step": 44510 }, { "grad_norm": 0.1980924904346466, "learning_rate": 6.304367153676561e-05, "loss": 0.0198, "step": 44520 }, { "grad_norm": 0.1902475655078888, "learning_rate": 6.302770867244978e-05, "loss": 0.0218, "step": 44530 }, { "grad_norm": 0.17259877920150757, "learning_rate": 6.301174438344328e-05, "loss": 0.0225, "step": 44540 }, { "grad_norm": 0.2097300887107849, "learning_rate": 6.299577867149202e-05, "loss": 0.0233, "step": 44550 }, { "grad_norm": 0.1870366930961609, "learning_rate": 6.297981153834192e-05, "loss": 0.0227, "step": 44560 }, { "grad_norm": 0.2924042344093323, "learning_rate": 6.296384298573916e-05, "loss": 0.0273, "step": 44570 }, { "grad_norm": 0.217123344540596, "learning_rate": 6.294787301543001e-05, "loss": 0.0217, "step": 44580 }, { "grad_norm": 0.2490759789943695, "learning_rate": 6.293190162916095e-05, "loss": 0.0233, "step": 44590 }, { "grad_norm": 0.20876026153564453, "learning_rate": 6.291592882867855e-05, "loss": 0.022, "step": 44600 }, { "grad_norm": 0.23578402400016785, "learning_rate": 6.28999546157296e-05, "loss": 0.0234, "step": 44610 }, { "grad_norm": 0.24189390242099762, "learning_rate": 6.288397899206102e-05, "loss": 0.0192, "step": 44620 }, { "grad_norm": 0.20326565206050873, "learning_rate": 6.286800195941984e-05, "loss": 0.0191, "step": 44630 }, { "grad_norm": 0.1916647106409073, "learning_rate": 6.285202351955334e-05, "loss": 0.0195, "step": 44640 }, { "grad_norm": 0.22914615273475647, "learning_rate": 6.283604367420887e-05, "loss": 0.0242, "step": 44650 }, { "grad_norm": 0.2392156422138214, "learning_rate": 6.282006242513394e-05, "loss": 0.0269, "step": 44660 }, { "grad_norm": 0.22084619104862213, "learning_rate": 6.280407977407628e-05, "loss": 0.0229, "step": 44670 }, { "grad_norm": 0.19964449107646942, "learning_rate": 6.27880957227837e-05, "loss": 0.0227, "step": 44680 }, { "grad_norm": 0.2235373854637146, "learning_rate": 6.27721102730042e-05, "loss": 0.0191, "step": 44690 }, { "grad_norm": 0.25477635860443115, "learning_rate": 6.275612342648592e-05, "loss": 0.0227, "step": 44700 }, { "grad_norm": 0.6715540289878845, "learning_rate": 6.274013518497716e-05, "loss": 0.0246, "step": 44710 }, { "grad_norm": 0.2083667665719986, "learning_rate": 6.272414555022636e-05, "loss": 0.0241, "step": 44720 }, { "grad_norm": 0.27136099338531494, "learning_rate": 6.270815452398215e-05, "loss": 0.0286, "step": 44730 }, { "grad_norm": 0.18621453642845154, "learning_rate": 6.269216210799326e-05, "loss": 0.021, "step": 44740 }, { "grad_norm": 0.18709102272987366, "learning_rate": 6.26761683040086e-05, "loss": 0.0233, "step": 44750 }, { "grad_norm": 0.2193632423877716, "learning_rate": 6.266017311377723e-05, "loss": 0.0233, "step": 44760 }, { "grad_norm": 0.25780901312828064, "learning_rate": 6.264417653904839e-05, "loss": 0.0263, "step": 44770 }, { "grad_norm": 0.21387019753456116, "learning_rate": 6.262817858157139e-05, "loss": 0.0205, "step": 44780 }, { "grad_norm": 0.24966369569301605, "learning_rate": 6.261217924309576e-05, "loss": 0.0256, "step": 44790 }, { "grad_norm": 0.2726019620895386, "learning_rate": 6.259617852537118e-05, "loss": 0.0235, "step": 44800 }, { "grad_norm": 0.2395181804895401, "learning_rate": 6.258017643014747e-05, "loss": 0.0233, "step": 44810 }, { "grad_norm": 0.272668719291687, "learning_rate": 6.256417295917456e-05, "loss": 0.0199, "step": 44820 }, { "grad_norm": 0.17068646848201752, "learning_rate": 6.254816811420258e-05, "loss": 0.026, "step": 44830 }, { "grad_norm": 0.18806028366088867, "learning_rate": 6.253216189698183e-05, "loss": 0.0219, "step": 44840 }, { "grad_norm": 0.2238556295633316, "learning_rate": 6.251615430926267e-05, "loss": 0.0222, "step": 44850 }, { "grad_norm": 0.2644675374031067, "learning_rate": 6.25001453527957e-05, "loss": 0.0261, "step": 44860 }, { "grad_norm": 0.23342613875865936, "learning_rate": 6.248413502933164e-05, "loss": 0.0242, "step": 44870 }, { "grad_norm": 0.25783881545066833, "learning_rate": 6.246812334062133e-05, "loss": 0.0233, "step": 44880 }, { "grad_norm": 0.20920312404632568, "learning_rate": 6.245211028841579e-05, "loss": 0.0207, "step": 44890 }, { "grad_norm": 0.19841645658016205, "learning_rate": 6.24360958744662e-05, "loss": 0.0192, "step": 44900 }, { "grad_norm": 0.21664349734783173, "learning_rate": 6.242008010052387e-05, "loss": 0.0193, "step": 44910 }, { "grad_norm": 0.2837234437465668, "learning_rate": 6.240406296834024e-05, "loss": 0.0232, "step": 44920 }, { "grad_norm": 0.1854911893606186, "learning_rate": 6.238804447966694e-05, "loss": 0.0225, "step": 44930 }, { "grad_norm": 0.1989770084619522, "learning_rate": 6.237202463625573e-05, "loss": 0.0206, "step": 44940 }, { "grad_norm": 0.15143975615501404, "learning_rate": 6.235600343985848e-05, "loss": 0.0221, "step": 44950 }, { "grad_norm": 0.21517351269721985, "learning_rate": 6.233998089222729e-05, "loss": 0.0205, "step": 44960 }, { "grad_norm": 0.14567627012729645, "learning_rate": 6.232395699511433e-05, "loss": 0.0228, "step": 44970 }, { "grad_norm": 0.2495778650045395, "learning_rate": 6.230793175027199e-05, "loss": 0.0241, "step": 44980 }, { "grad_norm": 0.19555307924747467, "learning_rate": 6.22919051594527e-05, "loss": 0.0266, "step": 44990 }, { "grad_norm": 0.2628236413002014, "learning_rate": 6.227587722440917e-05, "loss": 0.0207, "step": 45000 }, { "grad_norm": 0.19976215064525604, "learning_rate": 6.225984794689414e-05, "loss": 0.0229, "step": 45010 }, { "grad_norm": 0.25327247381210327, "learning_rate": 6.22438173286606e-05, "loss": 0.0193, "step": 45020 }, { "grad_norm": 0.2884983718395233, "learning_rate": 6.222778537146159e-05, "loss": 0.0233, "step": 45030 }, { "grad_norm": 0.2670062780380249, "learning_rate": 6.221175207705037e-05, "loss": 0.0232, "step": 45040 }, { "grad_norm": 0.20767562091350555, "learning_rate": 6.219571744718027e-05, "loss": 0.02, "step": 45050 }, { "grad_norm": 0.21417023241519928, "learning_rate": 6.217968148360487e-05, "loss": 0.0232, "step": 45060 }, { "grad_norm": 0.26691028475761414, "learning_rate": 6.216364418807782e-05, "loss": 0.0227, "step": 45070 }, { "grad_norm": 0.20419256389141083, "learning_rate": 6.214760556235292e-05, "loss": 0.0208, "step": 45080 }, { "grad_norm": 0.17969366908073425, "learning_rate": 6.213156560818414e-05, "loss": 0.0211, "step": 45090 }, { "grad_norm": 0.2470911741256714, "learning_rate": 6.211552432732559e-05, "loss": 0.0204, "step": 45100 }, { "grad_norm": 0.2306707203388214, "learning_rate": 6.20994817215315e-05, "loss": 0.022, "step": 45110 }, { "grad_norm": 0.1852675825357437, "learning_rate": 6.208343779255628e-05, "loss": 0.0223, "step": 45120 }, { "grad_norm": 0.18288537859916687, "learning_rate": 6.206739254215449e-05, "loss": 0.0222, "step": 45130 }, { "grad_norm": 0.2633810341358185, "learning_rate": 6.205134597208077e-05, "loss": 0.0196, "step": 45140 }, { "grad_norm": 0.25012481212615967, "learning_rate": 6.203529808409e-05, "loss": 0.0203, "step": 45150 }, { "grad_norm": 0.17227542400360107, "learning_rate": 6.201924887993712e-05, "loss": 0.0194, "step": 45160 }, { "grad_norm": 0.4799405634403229, "learning_rate": 6.200319836137725e-05, "loss": 0.0195, "step": 45170 }, { "grad_norm": 0.16971103847026825, "learning_rate": 6.198714653016565e-05, "loss": 0.0226, "step": 45180 }, { "grad_norm": 0.45280444622039795, "learning_rate": 6.197109338805774e-05, "loss": 0.0211, "step": 45190 }, { "grad_norm": 0.1640838235616684, "learning_rate": 6.195503893680903e-05, "loss": 0.0239, "step": 45200 }, { "grad_norm": 0.15851081907749176, "learning_rate": 6.193898317817524e-05, "loss": 0.0177, "step": 45210 }, { "grad_norm": 0.14308571815490723, "learning_rate": 6.192292611391221e-05, "loss": 0.022, "step": 45220 }, { "grad_norm": 0.2570885121822357, "learning_rate": 6.190686774577591e-05, "loss": 0.0209, "step": 45230 }, { "grad_norm": 0.29252079129219055, "learning_rate": 6.189080807552245e-05, "loss": 0.0203, "step": 45240 }, { "grad_norm": 0.282891184091568, "learning_rate": 6.187474710490809e-05, "loss": 0.0219, "step": 45250 }, { "grad_norm": 0.17477741837501526, "learning_rate": 6.185868483568926e-05, "loss": 0.0241, "step": 45260 }, { "grad_norm": 0.32304495573043823, "learning_rate": 6.184262126962245e-05, "loss": 0.0235, "step": 45270 }, { "grad_norm": 0.22383296489715576, "learning_rate": 6.182655640846442e-05, "loss": 0.023, "step": 45280 }, { "grad_norm": 0.16198742389678955, "learning_rate": 6.181049025397196e-05, "loss": 0.0233, "step": 45290 }, { "grad_norm": 0.27256980538368225, "learning_rate": 6.179442280790202e-05, "loss": 0.0237, "step": 45300 }, { "grad_norm": 0.20763526856899261, "learning_rate": 6.177835407201174e-05, "loss": 0.0234, "step": 45310 }, { "grad_norm": 0.241592139005661, "learning_rate": 6.176228404805839e-05, "loss": 0.0257, "step": 45320 }, { "grad_norm": 0.27858129143714905, "learning_rate": 6.174621273779932e-05, "loss": 0.0294, "step": 45330 }, { "grad_norm": 0.3796636462211609, "learning_rate": 6.17301401429921e-05, "loss": 0.0275, "step": 45340 }, { "grad_norm": 0.29101914167404175, "learning_rate": 6.171406626539436e-05, "loss": 0.0209, "step": 45350 }, { "grad_norm": 0.2264062911272049, "learning_rate": 6.169799110676398e-05, "loss": 0.0261, "step": 45360 }, { "grad_norm": 0.27137258648872375, "learning_rate": 6.168191466885885e-05, "loss": 0.0311, "step": 45370 }, { "grad_norm": 0.3371272087097168, "learning_rate": 6.166583695343711e-05, "loss": 0.0226, "step": 45380 }, { "grad_norm": 0.21794986724853516, "learning_rate": 6.164975796225698e-05, "loss": 0.0238, "step": 45390 }, { "grad_norm": 0.17966939508914948, "learning_rate": 6.163367769707683e-05, "loss": 0.0199, "step": 45400 }, { "grad_norm": 0.23182423412799835, "learning_rate": 6.161759615965519e-05, "loss": 0.0213, "step": 45410 }, { "grad_norm": 0.2574281692504883, "learning_rate": 6.16015133517507e-05, "loss": 0.0255, "step": 45420 }, { "grad_norm": 0.2317694127559662, "learning_rate": 6.158542927512214e-05, "loss": 0.0237, "step": 45430 }, { "grad_norm": 0.3990558683872223, "learning_rate": 6.156934393152846e-05, "loss": 0.0217, "step": 45440 }, { "grad_norm": 0.2844204008579254, "learning_rate": 6.15532573227287e-05, "loss": 0.0229, "step": 45450 }, { "grad_norm": 0.188120037317276, "learning_rate": 6.153716945048212e-05, "loss": 0.0218, "step": 45460 }, { "grad_norm": 0.2238900512456894, "learning_rate": 6.152108031654802e-05, "loss": 0.0232, "step": 45470 }, { "grad_norm": 0.1448691040277481, "learning_rate": 6.15049899226859e-05, "loss": 0.025, "step": 45480 }, { "grad_norm": 0.3056561350822449, "learning_rate": 6.148889827065537e-05, "loss": 0.0251, "step": 45490 }, { "grad_norm": 0.2197626680135727, "learning_rate": 6.147280536221622e-05, "loss": 0.0192, "step": 45500 }, { "grad_norm": 0.21484600007534027, "learning_rate": 6.145671119912832e-05, "loss": 0.0203, "step": 45510 }, { "grad_norm": 0.2665480077266693, "learning_rate": 6.144061578315169e-05, "loss": 0.0242, "step": 45520 }, { "grad_norm": 0.1926037222146988, "learning_rate": 6.142451911604654e-05, "loss": 0.0238, "step": 45530 }, { "grad_norm": 0.20501840114593506, "learning_rate": 6.140842119957315e-05, "loss": 0.0215, "step": 45540 }, { "grad_norm": 0.5074490308761597, "learning_rate": 6.139232203549197e-05, "loss": 0.0267, "step": 45550 }, { "grad_norm": 0.1893632411956787, "learning_rate": 6.137622162556357e-05, "loss": 0.0227, "step": 45560 }, { "grad_norm": 0.2029820680618286, "learning_rate": 6.13601199715487e-05, "loss": 0.0262, "step": 45570 }, { "grad_norm": 0.2686406970024109, "learning_rate": 6.134401707520816e-05, "loss": 0.0225, "step": 45580 }, { "grad_norm": 0.4746990501880646, "learning_rate": 6.132791293830298e-05, "loss": 0.0238, "step": 45590 }, { "grad_norm": 0.24573685228824615, "learning_rate": 6.131180756259428e-05, "loss": 0.027, "step": 45600 }, { "grad_norm": 0.30070731043815613, "learning_rate": 6.129570094984331e-05, "loss": 0.0268, "step": 45610 }, { "grad_norm": 0.20428305864334106, "learning_rate": 6.127959310181145e-05, "loss": 0.0301, "step": 45620 }, { "grad_norm": 0.3210342526435852, "learning_rate": 6.126348402026026e-05, "loss": 0.0306, "step": 45630 }, { "grad_norm": 0.31119677424430847, "learning_rate": 6.12473737069514e-05, "loss": 0.0269, "step": 45640 }, { "grad_norm": 0.28226134181022644, "learning_rate": 6.123126216364665e-05, "loss": 0.0241, "step": 45650 }, { "grad_norm": 0.21703822910785675, "learning_rate": 6.121514939210797e-05, "loss": 0.0275, "step": 45660 }, { "grad_norm": 0.17385408282279968, "learning_rate": 6.119903539409741e-05, "loss": 0.0241, "step": 45670 }, { "grad_norm": 0.1871878206729889, "learning_rate": 6.118292017137716e-05, "loss": 0.0212, "step": 45680 }, { "grad_norm": 0.27987098693847656, "learning_rate": 6.116680372570959e-05, "loss": 0.0214, "step": 45690 }, { "grad_norm": 0.32436367869377136, "learning_rate": 6.115068605885713e-05, "loss": 0.029, "step": 45700 }, { "grad_norm": 0.20916172862052917, "learning_rate": 6.113456717258243e-05, "loss": 0.0312, "step": 45710 }, { "grad_norm": 0.2107725739479065, "learning_rate": 6.11184470686482e-05, "loss": 0.0273, "step": 45720 }, { "grad_norm": 0.161626935005188, "learning_rate": 6.110232574881731e-05, "loss": 0.0216, "step": 45730 }, { "grad_norm": 0.25692620873451233, "learning_rate": 6.108620321485277e-05, "loss": 0.0247, "step": 45740 }, { "grad_norm": 0.19868464767932892, "learning_rate": 6.107007946851773e-05, "loss": 0.0225, "step": 45750 }, { "grad_norm": 0.19311872124671936, "learning_rate": 6.105395451157542e-05, "loss": 0.0215, "step": 45760 }, { "grad_norm": 0.20657207071781158, "learning_rate": 6.103782834578928e-05, "loss": 0.0199, "step": 45770 }, { "grad_norm": 0.18255306780338287, "learning_rate": 6.102170097292281e-05, "loss": 0.0232, "step": 45780 }, { "grad_norm": 0.21906226873397827, "learning_rate": 6.1005572394739716e-05, "loss": 0.0228, "step": 45790 }, { "grad_norm": 0.2172698825597763, "learning_rate": 6.0989442613003765e-05, "loss": 0.0248, "step": 45800 }, { "grad_norm": 0.19658714532852173, "learning_rate": 6.097331162947888e-05, "loss": 0.0229, "step": 45810 }, { "grad_norm": 0.1928250640630722, "learning_rate": 6.095717944592914e-05, "loss": 0.0223, "step": 45820 }, { "grad_norm": 0.18749220669269562, "learning_rate": 6.094104606411873e-05, "loss": 0.0211, "step": 45830 }, { "grad_norm": 0.1891990751028061, "learning_rate": 6.0924911485811966e-05, "loss": 0.0235, "step": 45840 }, { "grad_norm": 0.17745698988437653, "learning_rate": 6.090877571277331e-05, "loss": 0.0194, "step": 45850 }, { "grad_norm": 0.22411711513996124, "learning_rate": 6.089263874676736e-05, "loss": 0.021, "step": 45860 }, { "grad_norm": 0.21159246563911438, "learning_rate": 6.0876500589558796e-05, "loss": 0.0247, "step": 45870 }, { "grad_norm": 0.30249685049057007, "learning_rate": 6.086036124291248e-05, "loss": 0.0226, "step": 45880 }, { "grad_norm": 0.18968737125396729, "learning_rate": 6.084422070859339e-05, "loss": 0.0218, "step": 45890 }, { "grad_norm": 0.19616208970546722, "learning_rate": 6.082807898836663e-05, "loss": 0.0212, "step": 45900 }, { "grad_norm": 0.1851501613855362, "learning_rate": 6.081193608399742e-05, "loss": 0.024, "step": 45910 }, { "grad_norm": 0.22602614760398865, "learning_rate": 6.0795791997251164e-05, "loss": 0.0213, "step": 45920 }, { "grad_norm": 0.2197364717721939, "learning_rate": 6.0779646729893294e-05, "loss": 0.0223, "step": 45930 }, { "grad_norm": 0.16434288024902344, "learning_rate": 6.0763500283689476e-05, "loss": 0.0208, "step": 45940 }, { "grad_norm": 0.24832648038864136, "learning_rate": 6.0747352660405455e-05, "loss": 0.0197, "step": 45950 }, { "grad_norm": 0.24992328882217407, "learning_rate": 6.073120386180709e-05, "loss": 0.0187, "step": 45960 }, { "grad_norm": 0.25511983036994934, "learning_rate": 6.0715053889660425e-05, "loss": 0.0189, "step": 45970 }, { "grad_norm": 0.20958828926086426, "learning_rate": 6.069890274573157e-05, "loss": 0.0174, "step": 45980 }, { "grad_norm": 0.23410862684249878, "learning_rate": 6.068275043178679e-05, "loss": 0.027, "step": 45990 }, { "grad_norm": 0.2537403106689453, "learning_rate": 6.066659694959248e-05, "loss": 0.0215, "step": 46000 }, { "grad_norm": 0.2290707677602768, "learning_rate": 6.0650442300915176e-05, "loss": 0.0242, "step": 46010 }, { "grad_norm": 0.2232094407081604, "learning_rate": 6.063428648752152e-05, "loss": 0.0204, "step": 46020 }, { "grad_norm": 0.17388373613357544, "learning_rate": 6.0618129511178266e-05, "loss": 0.021, "step": 46030 }, { "grad_norm": 0.18372377753257751, "learning_rate": 6.060197137365234e-05, "loss": 0.0227, "step": 46040 }, { "grad_norm": 0.17478200793266296, "learning_rate": 6.058581207671077e-05, "loss": 0.0228, "step": 46050 }, { "grad_norm": 0.2911624312400818, "learning_rate": 6.056965162212072e-05, "loss": 0.0232, "step": 46060 }, { "grad_norm": 0.1933620274066925, "learning_rate": 6.055349001164943e-05, "loss": 0.0223, "step": 46070 }, { "grad_norm": 0.24480542540550232, "learning_rate": 6.0537327247064347e-05, "loss": 0.0218, "step": 46080 }, { "grad_norm": 0.2266247272491455, "learning_rate": 6.0521163330133014e-05, "loss": 0.0211, "step": 46090 }, { "grad_norm": 0.24682164192199707, "learning_rate": 6.050499826262306e-05, "loss": 0.0235, "step": 46100 }, { "grad_norm": 0.19148607552051544, "learning_rate": 6.0488832046302294e-05, "loss": 0.0207, "step": 46110 }, { "grad_norm": 0.22611087560653687, "learning_rate": 6.0472664682938626e-05, "loss": 0.0242, "step": 46120 }, { "grad_norm": 0.28363922238349915, "learning_rate": 6.045649617430009e-05, "loss": 0.0262, "step": 46130 }, { "grad_norm": 0.2229962795972824, "learning_rate": 6.0440326522154866e-05, "loss": 0.0262, "step": 46140 }, { "grad_norm": 0.2426486313343048, "learning_rate": 6.0424155728271224e-05, "loss": 0.0261, "step": 46150 }, { "grad_norm": 0.21855147182941437, "learning_rate": 6.040798379441758e-05, "loss": 0.027, "step": 46160 }, { "grad_norm": 0.1376921385526657, "learning_rate": 6.0391810722362485e-05, "loss": 0.0232, "step": 46170 }, { "grad_norm": 0.19261790812015533, "learning_rate": 6.037563651387458e-05, "loss": 0.021, "step": 46180 }, { "grad_norm": 0.18484382331371307, "learning_rate": 6.0359461170722666e-05, "loss": 0.0211, "step": 46190 }, { "grad_norm": 0.22652330994606018, "learning_rate": 6.034328469467566e-05, "loss": 0.0216, "step": 46200 }, { "grad_norm": 0.21866805851459503, "learning_rate": 6.0327107087502596e-05, "loss": 0.0229, "step": 46210 }, { "grad_norm": 0.23709143698215485, "learning_rate": 6.031092835097262e-05, "loss": 0.0238, "step": 46220 }, { "grad_norm": 0.18879978358745575, "learning_rate": 6.0294748486855024e-05, "loss": 0.024, "step": 46230 }, { "grad_norm": 0.28844645619392395, "learning_rate": 6.0278567496919216e-05, "loss": 0.023, "step": 46240 }, { "grad_norm": 0.20926356315612793, "learning_rate": 6.026238538293472e-05, "loss": 0.0218, "step": 46250 }, { "grad_norm": 0.14718465507030487, "learning_rate": 6.024620214667118e-05, "loss": 0.0211, "step": 46260 }, { "grad_norm": 0.20107710361480713, "learning_rate": 6.0230017789898384e-05, "loss": 0.0205, "step": 46270 }, { "grad_norm": 0.2655388116836548, "learning_rate": 6.021383231438622e-05, "loss": 0.0234, "step": 46280 }, { "grad_norm": 0.18962161242961884, "learning_rate": 6.0197645721904704e-05, "loss": 0.0198, "step": 46290 }, { "grad_norm": 0.2602345943450928, "learning_rate": 6.0181458014224e-05, "loss": 0.0268, "step": 46300 }, { "grad_norm": 0.20968830585479736, "learning_rate": 6.016526919311435e-05, "loss": 0.023, "step": 46310 }, { "grad_norm": 0.26273125410079956, "learning_rate": 6.014907926034613e-05, "loss": 0.0222, "step": 46320 }, { "grad_norm": 0.2083510309457779, "learning_rate": 6.0132888217689875e-05, "loss": 0.023, "step": 46330 }, { "grad_norm": 0.1634364277124405, "learning_rate": 6.01166960669162e-05, "loss": 0.0166, "step": 46340 }, { "grad_norm": 0.1833488941192627, "learning_rate": 6.0100502809795845e-05, "loss": 0.0234, "step": 46350 }, { "grad_norm": 0.25044143199920654, "learning_rate": 6.00843084480997e-05, "loss": 0.0206, "step": 46360 }, { "grad_norm": 0.23229119181632996, "learning_rate": 6.0068112983598736e-05, "loss": 0.0254, "step": 46370 }, { "grad_norm": 0.17848743498325348, "learning_rate": 6.005191641806407e-05, "loss": 0.0204, "step": 46380 }, { "grad_norm": 0.28536686301231384, "learning_rate": 6.003571875326694e-05, "loss": 0.0218, "step": 46390 }, { "grad_norm": 0.2711665630340576, "learning_rate": 6.00195199909787e-05, "loss": 0.025, "step": 46400 }, { "grad_norm": 0.16468632221221924, "learning_rate": 6.000332013297082e-05, "loss": 0.0217, "step": 46410 }, { "grad_norm": 0.1778302937746048, "learning_rate": 5.998711918101487e-05, "loss": 0.0199, "step": 46420 }, { "grad_norm": 0.18697048723697662, "learning_rate": 5.997091713688261e-05, "loss": 0.0193, "step": 46430 }, { "grad_norm": 0.20627716183662415, "learning_rate": 5.9954714002345836e-05, "loss": 0.0186, "step": 46440 }, { "grad_norm": 0.1772264540195465, "learning_rate": 5.993850977917649e-05, "loss": 0.0223, "step": 46450 }, { "grad_norm": 0.24975493550300598, "learning_rate": 5.992230446914667e-05, "loss": 0.0189, "step": 46460 }, { "grad_norm": 0.132540762424469, "learning_rate": 5.990609807402855e-05, "loss": 0.0214, "step": 46470 }, { "grad_norm": 0.22107850015163422, "learning_rate": 5.988989059559443e-05, "loss": 0.021, "step": 46480 }, { "grad_norm": 0.1936224102973938, "learning_rate": 5.987368203561675e-05, "loss": 0.0215, "step": 46490 }, { "grad_norm": 0.22365257143974304, "learning_rate": 5.9857472395868055e-05, "loss": 0.019, "step": 46500 }, { "grad_norm": 0.24908329546451569, "learning_rate": 5.9841261678120983e-05, "loss": 0.019, "step": 46510 }, { "grad_norm": 0.47521457076072693, "learning_rate": 5.982504988414834e-05, "loss": 0.0217, "step": 46520 }, { "grad_norm": 0.2717495262622833, "learning_rate": 5.9808837015723015e-05, "loss": 0.0203, "step": 46530 }, { "grad_norm": 0.4116736054420471, "learning_rate": 5.9792623074618016e-05, "loss": 0.0218, "step": 46540 }, { "grad_norm": 0.2045091837644577, "learning_rate": 5.977640806260648e-05, "loss": 0.0169, "step": 46550 }, { "grad_norm": 0.24850738048553467, "learning_rate": 5.976019198146166e-05, "loss": 0.022, "step": 46560 }, { "grad_norm": 0.2210945039987564, "learning_rate": 5.974397483295692e-05, "loss": 0.0223, "step": 46570 }, { "grad_norm": 0.1790705919265747, "learning_rate": 5.972775661886575e-05, "loss": 0.022, "step": 46580 }, { "grad_norm": 0.23740826547145844, "learning_rate": 5.971153734096173e-05, "loss": 0.0218, "step": 46590 }, { "grad_norm": 0.2402390092611313, "learning_rate": 5.969531700101859e-05, "loss": 0.022, "step": 46600 }, { "grad_norm": 0.1749953329563141, "learning_rate": 5.9679095600810155e-05, "loss": 0.018, "step": 46610 }, { "grad_norm": 0.36276790499687195, "learning_rate": 5.9662873142110384e-05, "loss": 0.0235, "step": 46620 }, { "grad_norm": 0.2200687974691391, "learning_rate": 5.964664962669333e-05, "loss": 0.0201, "step": 46630 }, { "grad_norm": 0.4513656795024872, "learning_rate": 5.9630425056333186e-05, "loss": 0.0243, "step": 46640 }, { "grad_norm": 0.3059120774269104, "learning_rate": 5.961419943280422e-05, "loss": 0.0207, "step": 46650 }, { "grad_norm": 0.16110007464885712, "learning_rate": 5.959797275788087e-05, "loss": 0.0202, "step": 46660 }, { "grad_norm": 0.2517288625240326, "learning_rate": 5.958174503333765e-05, "loss": 0.0219, "step": 46670 }, { "grad_norm": 0.23248723149299622, "learning_rate": 5.9565516260949195e-05, "loss": 0.0206, "step": 46680 }, { "grad_norm": 0.21606092154979706, "learning_rate": 5.9549286442490273e-05, "loss": 0.0197, "step": 46690 }, { "grad_norm": 0.23336753249168396, "learning_rate": 5.953305557973572e-05, "loss": 0.0245, "step": 46700 }, { "grad_norm": 0.20490336418151855, "learning_rate": 5.9516823674460564e-05, "loss": 0.0234, "step": 46710 }, { "grad_norm": 0.15774568915367126, "learning_rate": 5.9500590728439875e-05, "loss": 0.0235, "step": 46720 }, { "grad_norm": 0.19981031119823456, "learning_rate": 5.948435674344886e-05, "loss": 0.0229, "step": 46730 }, { "grad_norm": 0.23866087198257446, "learning_rate": 5.946812172126285e-05, "loss": 0.0238, "step": 46740 }, { "grad_norm": 0.2372291386127472, "learning_rate": 5.945188566365729e-05, "loss": 0.0257, "step": 46750 }, { "grad_norm": 0.2753939926624298, "learning_rate": 5.943564857240773e-05, "loss": 0.0237, "step": 46760 }, { "grad_norm": 0.16563758254051208, "learning_rate": 5.941941044928981e-05, "loss": 0.0189, "step": 46770 }, { "grad_norm": 0.18534713983535767, "learning_rate": 5.940317129607935e-05, "loss": 0.0196, "step": 46780 }, { "grad_norm": 0.23166757822036743, "learning_rate": 5.9386931114552204e-05, "loss": 0.0265, "step": 46790 }, { "grad_norm": 0.24199357628822327, "learning_rate": 5.937068990648438e-05, "loss": 0.0224, "step": 46800 }, { "grad_norm": 0.18600670993328094, "learning_rate": 5.935444767365199e-05, "loss": 0.0226, "step": 46810 }, { "grad_norm": 0.1930825561285019, "learning_rate": 5.933820441783129e-05, "loss": 0.0215, "step": 46820 }, { "grad_norm": 0.3988812565803528, "learning_rate": 5.932196014079857e-05, "loss": 0.0279, "step": 46830 }, { "grad_norm": 0.2070576250553131, "learning_rate": 5.930571484433032e-05, "loss": 0.0261, "step": 46840 }, { "grad_norm": 0.2816373109817505, "learning_rate": 5.928946853020309e-05, "loss": 0.0259, "step": 46850 }, { "grad_norm": 0.24900929629802704, "learning_rate": 5.927322120019352e-05, "loss": 0.0226, "step": 46860 }, { "grad_norm": 0.2033550888299942, "learning_rate": 5.925697285607844e-05, "loss": 0.0207, "step": 46870 }, { "grad_norm": 0.17328527569770813, "learning_rate": 5.924072349963472e-05, "loss": 0.0199, "step": 46880 }, { "grad_norm": 0.23261016607284546, "learning_rate": 5.922447313263938e-05, "loss": 0.0286, "step": 46890 }, { "grad_norm": 0.20382918417453766, "learning_rate": 5.920822175686952e-05, "loss": 0.0228, "step": 46900 }, { "grad_norm": 0.24555964767932892, "learning_rate": 5.919196937410239e-05, "loss": 0.0255, "step": 46910 }, { "grad_norm": 0.1751497983932495, "learning_rate": 5.917571598611529e-05, "loss": 0.0268, "step": 46920 }, { "grad_norm": 0.2269761562347412, "learning_rate": 5.9159461594685686e-05, "loss": 0.0218, "step": 46930 }, { "grad_norm": 0.2647487223148346, "learning_rate": 5.914320620159114e-05, "loss": 0.027, "step": 46940 }, { "grad_norm": 0.18416689336299896, "learning_rate": 5.912694980860931e-05, "loss": 0.0202, "step": 46950 }, { "grad_norm": 0.2260766625404358, "learning_rate": 5.9110692417517964e-05, "loss": 0.0232, "step": 46960 }, { "grad_norm": 0.2548942565917969, "learning_rate": 5.9094434030095e-05, "loss": 0.02, "step": 46970 }, { "grad_norm": 0.29571330547332764, "learning_rate": 5.907817464811841e-05, "loss": 0.0191, "step": 46980 }, { "grad_norm": 0.17694896459579468, "learning_rate": 5.906191427336627e-05, "loss": 0.0211, "step": 46990 }, { "grad_norm": 0.2487000972032547, "learning_rate": 5.9045652907616835e-05, "loss": 0.0284, "step": 47000 }, { "grad_norm": 0.3322368860244751, "learning_rate": 5.902939055264838e-05, "loss": 0.0257, "step": 47010 }, { "grad_norm": 0.2557842433452606, "learning_rate": 5.901312721023935e-05, "loss": 0.0225, "step": 47020 }, { "grad_norm": 0.20307962596416473, "learning_rate": 5.8996862882168294e-05, "loss": 0.0193, "step": 47030 }, { "grad_norm": 0.28892645239830017, "learning_rate": 5.8980597570213826e-05, "loss": 0.0239, "step": 47040 }, { "grad_norm": 0.24308504164218903, "learning_rate": 5.896433127615471e-05, "loss": 0.0235, "step": 47050 }, { "grad_norm": 0.23550516366958618, "learning_rate": 5.894806400176981e-05, "loss": 0.0258, "step": 47060 }, { "grad_norm": 0.19249051809310913, "learning_rate": 5.893179574883808e-05, "loss": 0.0232, "step": 47070 }, { "grad_norm": 0.2317677140235901, "learning_rate": 5.8915526519138585e-05, "loss": 0.0227, "step": 47080 }, { "grad_norm": 0.17318996787071228, "learning_rate": 5.889925631445053e-05, "loss": 0.0205, "step": 47090 }, { "grad_norm": 0.16984064877033234, "learning_rate": 5.888298513655318e-05, "loss": 0.0205, "step": 47100 }, { "grad_norm": 0.3144681453704834, "learning_rate": 5.886671298722592e-05, "loss": 0.0191, "step": 47110 }, { "grad_norm": 0.1753852665424347, "learning_rate": 5.885043986824828e-05, "loss": 0.0198, "step": 47120 }, { "grad_norm": 0.25672784447669983, "learning_rate": 5.883416578139982e-05, "loss": 0.0201, "step": 47130 }, { "grad_norm": 0.33781105279922485, "learning_rate": 5.881789072846029e-05, "loss": 0.0218, "step": 47140 }, { "grad_norm": 0.38251712918281555, "learning_rate": 5.8801614711209474e-05, "loss": 0.022, "step": 47150 }, { "grad_norm": 0.2598203718662262, "learning_rate": 5.878533773142732e-05, "loss": 0.0258, "step": 47160 }, { "grad_norm": 0.1757475584745407, "learning_rate": 5.8769059790893834e-05, "loss": 0.0233, "step": 47170 }, { "grad_norm": 0.20623724162578583, "learning_rate": 5.8752780891389145e-05, "loss": 0.0179, "step": 47180 }, { "grad_norm": 0.24343377351760864, "learning_rate": 5.873650103469351e-05, "loss": 0.0256, "step": 47190 }, { "grad_norm": 0.24063624441623688, "learning_rate": 5.872022022258725e-05, "loss": 0.0189, "step": 47200 }, { "grad_norm": 0.2407434731721878, "learning_rate": 5.8703938456850814e-05, "loss": 0.0215, "step": 47210 }, { "grad_norm": 0.33010923862457275, "learning_rate": 5.868765573926477e-05, "loss": 0.0209, "step": 47220 }, { "grad_norm": 0.23803699016571045, "learning_rate": 5.867137207160974e-05, "loss": 0.0204, "step": 47230 }, { "grad_norm": 0.20635630190372467, "learning_rate": 5.86550874556665e-05, "loss": 0.0189, "step": 47240 }, { "grad_norm": 0.21284247934818268, "learning_rate": 5.863880189321592e-05, "loss": 0.0204, "step": 47250 }, { "grad_norm": 0.3267650604248047, "learning_rate": 5.862251538603894e-05, "loss": 0.0235, "step": 47260 }, { "grad_norm": 0.20810800790786743, "learning_rate": 5.8606227935916656e-05, "loss": 0.022, "step": 47270 }, { "grad_norm": 0.7130566835403442, "learning_rate": 5.858993954463021e-05, "loss": 0.0239, "step": 47280 }, { "grad_norm": 0.2894035279750824, "learning_rate": 5.8573650213960896e-05, "loss": 0.0225, "step": 47290 }, { "grad_norm": 0.17817935347557068, "learning_rate": 5.8557359945690084e-05, "loss": 0.0186, "step": 47300 }, { "grad_norm": 0.38818368315696716, "learning_rate": 5.8541068741599246e-05, "loss": 0.0226, "step": 47310 }, { "grad_norm": 0.3024660348892212, "learning_rate": 5.8524776603469985e-05, "loss": 0.0211, "step": 47320 }, { "grad_norm": 0.5650104284286499, "learning_rate": 5.850848353308397e-05, "loss": 0.0235, "step": 47330 }, { "grad_norm": 0.19652001559734344, "learning_rate": 5.849218953222297e-05, "loss": 0.0234, "step": 47340 }, { "grad_norm": 0.40119287371635437, "learning_rate": 5.847589460266891e-05, "loss": 0.0231, "step": 47350 }, { "grad_norm": 0.34932124614715576, "learning_rate": 5.845959874620375e-05, "loss": 0.0206, "step": 47360 }, { "grad_norm": 0.31154415011405945, "learning_rate": 5.84433019646096e-05, "loss": 0.0245, "step": 47370 }, { "grad_norm": 0.24166810512542725, "learning_rate": 5.842700425966863e-05, "loss": 0.0246, "step": 47380 }, { "grad_norm": 0.14967681467533112, "learning_rate": 5.841070563316315e-05, "loss": 0.0224, "step": 47390 }, { "grad_norm": 0.34781327843666077, "learning_rate": 5.839440608687554e-05, "loss": 0.0267, "step": 47400 }, { "grad_norm": 0.723470151424408, "learning_rate": 5.837810562258831e-05, "loss": 0.0224, "step": 47410 }, { "grad_norm": 0.24699260294437408, "learning_rate": 5.836180424208405e-05, "loss": 0.024, "step": 47420 }, { "grad_norm": 0.16988055408000946, "learning_rate": 5.834550194714543e-05, "loss": 0.0233, "step": 47430 }, { "grad_norm": 0.42169302701950073, "learning_rate": 5.832919873955528e-05, "loss": 0.0262, "step": 47440 }, { "grad_norm": 0.2809465527534485, "learning_rate": 5.831289462109648e-05, "loss": 0.0218, "step": 47450 }, { "grad_norm": 0.24625636637210846, "learning_rate": 5.829658959355202e-05, "loss": 0.0228, "step": 47460 }, { "grad_norm": 0.19174587726593018, "learning_rate": 5.8280283658704995e-05, "loss": 0.0223, "step": 47470 }, { "grad_norm": 0.8084394335746765, "learning_rate": 5.826397681833859e-05, "loss": 0.0291, "step": 47480 }, { "grad_norm": 0.18077990412712097, "learning_rate": 5.824766907423612e-05, "loss": 0.022, "step": 47490 }, { "grad_norm": 0.15792140364646912, "learning_rate": 5.8231360428180945e-05, "loss": 0.0245, "step": 47500 }, { "grad_norm": 0.19938254356384277, "learning_rate": 5.821505088195658e-05, "loss": 0.0211, "step": 47510 }, { "grad_norm": 0.23135806620121002, "learning_rate": 5.819874043734661e-05, "loss": 0.0229, "step": 47520 }, { "grad_norm": 0.18234117329120636, "learning_rate": 5.8182429096134695e-05, "loss": 0.0214, "step": 47530 }, { "grad_norm": 0.19231855869293213, "learning_rate": 5.816611686010465e-05, "loss": 0.0214, "step": 47540 }, { "grad_norm": 0.25403037667274475, "learning_rate": 5.814980373104033e-05, "loss": 0.0216, "step": 47550 }, { "grad_norm": 0.21109750866889954, "learning_rate": 5.813348971072572e-05, "loss": 0.0227, "step": 47560 }, { "grad_norm": 0.28619852662086487, "learning_rate": 5.811717480094492e-05, "loss": 0.0235, "step": 47570 }, { "grad_norm": 0.18979229032993317, "learning_rate": 5.810085900348209e-05, "loss": 0.0229, "step": 47580 }, { "grad_norm": 0.19576847553253174, "learning_rate": 5.8084542320121483e-05, "loss": 0.0222, "step": 47590 }, { "grad_norm": 0.30294346809387207, "learning_rate": 5.8068224752647497e-05, "loss": 0.023, "step": 47600 }, { "grad_norm": 0.23733679950237274, "learning_rate": 5.805190630284457e-05, "loss": 0.0211, "step": 47610 }, { "grad_norm": 0.23232479393482208, "learning_rate": 5.803558697249729e-05, "loss": 0.0225, "step": 47620 }, { "grad_norm": 0.25922590494155884, "learning_rate": 5.8019266763390276e-05, "loss": 0.0236, "step": 47630 }, { "grad_norm": 0.22779415547847748, "learning_rate": 5.800294567730833e-05, "loss": 0.0209, "step": 47640 }, { "grad_norm": 0.24596494436264038, "learning_rate": 5.798662371603628e-05, "loss": 0.0202, "step": 47650 }, { "grad_norm": 1.6122747659683228, "learning_rate": 5.797030088135904e-05, "loss": 0.0199, "step": 47660 }, { "grad_norm": 0.2729753255844116, "learning_rate": 5.7953977175061705e-05, "loss": 0.0207, "step": 47670 }, { "grad_norm": 0.26614993810653687, "learning_rate": 5.793765259892938e-05, "loss": 0.0286, "step": 47680 }, { "grad_norm": 0.2799677550792694, "learning_rate": 5.792132715474729e-05, "loss": 0.0243, "step": 47690 }, { "grad_norm": 0.17449010908603668, "learning_rate": 5.790500084430078e-05, "loss": 0.0236, "step": 47700 }, { "grad_norm": 0.22556321322917938, "learning_rate": 5.7888673669375274e-05, "loss": 0.0272, "step": 47710 }, { "grad_norm": 0.22882243990898132, "learning_rate": 5.787234563175625e-05, "loss": 0.0268, "step": 47720 }, { "grad_norm": 0.28355634212493896, "learning_rate": 5.7856016733229355e-05, "loss": 0.0257, "step": 47730 }, { "grad_norm": 0.2220374494791031, "learning_rate": 5.7839686975580297e-05, "loss": 0.0265, "step": 47740 }, { "grad_norm": 0.16704589128494263, "learning_rate": 5.782335636059484e-05, "loss": 0.0212, "step": 47750 }, { "grad_norm": 0.170159712433815, "learning_rate": 5.780702489005889e-05, "loss": 0.0224, "step": 47760 }, { "grad_norm": 0.2229774445295334, "learning_rate": 5.779069256575845e-05, "loss": 0.0221, "step": 47770 }, { "grad_norm": 0.18994836509227753, "learning_rate": 5.7774359389479574e-05, "loss": 0.0202, "step": 47780 }, { "grad_norm": 0.20768855512142181, "learning_rate": 5.775802536300845e-05, "loss": 0.0228, "step": 47790 }, { "grad_norm": 0.19808411598205566, "learning_rate": 5.774169048813134e-05, "loss": 0.0223, "step": 47800 }, { "grad_norm": 0.23848587274551392, "learning_rate": 5.77253547666346e-05, "loss": 0.0233, "step": 47810 }, { "grad_norm": 0.35549354553222656, "learning_rate": 5.770901820030465e-05, "loss": 0.0217, "step": 47820 }, { "grad_norm": 0.17542612552642822, "learning_rate": 5.769268079092809e-05, "loss": 0.018, "step": 47830 }, { "grad_norm": 0.17950297892093658, "learning_rate": 5.767634254029151e-05, "loss": 0.0196, "step": 47840 }, { "grad_norm": 0.1866312175989151, "learning_rate": 5.7660003450181655e-05, "loss": 0.0236, "step": 47850 }, { "grad_norm": 0.23659247159957886, "learning_rate": 5.764366352238534e-05, "loss": 0.0194, "step": 47860 }, { "grad_norm": 0.21674397587776184, "learning_rate": 5.7627322758689474e-05, "loss": 0.0218, "step": 47870 }, { "grad_norm": 0.3030816614627838, "learning_rate": 5.761098116088105e-05, "loss": 0.0219, "step": 47880 }, { "grad_norm": 0.35917404294013977, "learning_rate": 5.759463873074717e-05, "loss": 0.0251, "step": 47890 }, { "grad_norm": 0.20493772625923157, "learning_rate": 5.757829547007504e-05, "loss": 0.0255, "step": 47900 }, { "grad_norm": 0.21038077771663666, "learning_rate": 5.756195138065189e-05, "loss": 0.0261, "step": 47910 }, { "grad_norm": 0.21923142671585083, "learning_rate": 5.754560646426511e-05, "loss": 0.0211, "step": 47920 }, { "grad_norm": 0.2147422730922699, "learning_rate": 5.752926072270216e-05, "loss": 0.0207, "step": 47930 }, { "grad_norm": 0.2949376702308655, "learning_rate": 5.7512914157750563e-05, "loss": 0.0206, "step": 47940 }, { "grad_norm": 0.17570023238658905, "learning_rate": 5.749656677119798e-05, "loss": 0.0208, "step": 47950 }, { "grad_norm": 0.39156585931777954, "learning_rate": 5.7480218564832125e-05, "loss": 0.0213, "step": 47960 }, { "grad_norm": 0.17630013823509216, "learning_rate": 5.746386954044082e-05, "loss": 0.0216, "step": 47970 }, { "grad_norm": 0.2601175606250763, "learning_rate": 5.744751969981195e-05, "loss": 0.0205, "step": 47980 }, { "grad_norm": 0.14942215383052826, "learning_rate": 5.7431169044733526e-05, "loss": 0.0218, "step": 47990 }, { "grad_norm": 0.24637296795845032, "learning_rate": 5.741481757699364e-05, "loss": 0.0226, "step": 48000 }, { "grad_norm": 0.23156915605068207, "learning_rate": 5.7398465298380434e-05, "loss": 0.0231, "step": 48010 }, { "grad_norm": 0.19715069234371185, "learning_rate": 5.7382112210682193e-05, "loss": 0.0181, "step": 48020 }, { "grad_norm": 0.2072894126176834, "learning_rate": 5.7365758315687266e-05, "loss": 0.0246, "step": 48030 }, { "grad_norm": 0.1818406879901886, "learning_rate": 5.734940361518407e-05, "loss": 0.0188, "step": 48040 }, { "grad_norm": 0.5606689453125, "learning_rate": 5.733304811096116e-05, "loss": 0.0202, "step": 48050 }, { "grad_norm": 0.20845350623130798, "learning_rate": 5.731669180480713e-05, "loss": 0.0196, "step": 48060 }, { "grad_norm": 0.3039357364177704, "learning_rate": 5.730033469851067e-05, "loss": 0.0215, "step": 48070 }, { "grad_norm": 0.2705710530281067, "learning_rate": 5.7283976793860607e-05, "loss": 0.0229, "step": 48080 }, { "grad_norm": 0.20351551473140717, "learning_rate": 5.726761809264577e-05, "loss": 0.0214, "step": 48090 }, { "grad_norm": 0.14613837003707886, "learning_rate": 5.7251258596655155e-05, "loss": 0.0227, "step": 48100 }, { "grad_norm": 0.3175255358219147, "learning_rate": 5.72348983076778e-05, "loss": 0.0206, "step": 48110 }, { "grad_norm": 0.16784951090812683, "learning_rate": 5.7218537227502854e-05, "loss": 0.0211, "step": 48120 }, { "grad_norm": 0.2186928242444992, "learning_rate": 5.720217535791951e-05, "loss": 0.0211, "step": 48130 }, { "grad_norm": 0.24772684276103973, "learning_rate": 5.718581270071711e-05, "loss": 0.0208, "step": 48140 }, { "grad_norm": 0.24942997097969055, "learning_rate": 5.716944925768505e-05, "loss": 0.0171, "step": 48150 }, { "grad_norm": 0.1638152152299881, "learning_rate": 5.7153085030612786e-05, "loss": 0.0196, "step": 48160 }, { "grad_norm": 0.21332398056983948, "learning_rate": 5.713672002128989e-05, "loss": 0.0219, "step": 48170 }, { "grad_norm": 0.23432840406894684, "learning_rate": 5.712035423150602e-05, "loss": 0.0209, "step": 48180 }, { "grad_norm": 0.17603878676891327, "learning_rate": 5.710398766305094e-05, "loss": 0.0193, "step": 48190 }, { "grad_norm": 0.22852382063865662, "learning_rate": 5.708762031771442e-05, "loss": 0.0204, "step": 48200 }, { "grad_norm": 0.2239173799753189, "learning_rate": 5.707125219728642e-05, "loss": 0.02, "step": 48210 }, { "grad_norm": 0.2169191539287567, "learning_rate": 5.7054883303556905e-05, "loss": 0.0202, "step": 48220 }, { "grad_norm": 0.15430012345314026, "learning_rate": 5.703851363831595e-05, "loss": 0.02, "step": 48230 }, { "grad_norm": 0.18234455585479736, "learning_rate": 5.702214320335374e-05, "loss": 0.0205, "step": 48240 }, { "grad_norm": 0.34874099493026733, "learning_rate": 5.7005772000460514e-05, "loss": 0.0205, "step": 48250 }, { "grad_norm": 0.14677925407886505, "learning_rate": 5.698940003142656e-05, "loss": 0.0211, "step": 48260 }, { "grad_norm": 0.26005011796951294, "learning_rate": 5.697302729804236e-05, "loss": 0.0168, "step": 48270 }, { "grad_norm": 0.18734130263328552, "learning_rate": 5.695665380209837e-05, "loss": 0.0215, "step": 48280 }, { "grad_norm": 0.38331615924835205, "learning_rate": 5.6940279545385165e-05, "loss": 0.0202, "step": 48290 }, { "grad_norm": 0.15375933051109314, "learning_rate": 5.692390452969344e-05, "loss": 0.0204, "step": 48300 }, { "grad_norm": 0.7752136588096619, "learning_rate": 5.690752875681392e-05, "loss": 0.0212, "step": 48310 }, { "grad_norm": 0.3405587375164032, "learning_rate": 5.6891152228537435e-05, "loss": 0.0252, "step": 48320 }, { "grad_norm": 0.16644473373889923, "learning_rate": 5.687477494665492e-05, "loss": 0.0183, "step": 48330 }, { "grad_norm": 0.19240975379943848, "learning_rate": 5.685839691295733e-05, "loss": 0.0184, "step": 48340 }, { "grad_norm": 0.1754380762577057, "learning_rate": 5.6842018129235786e-05, "loss": 0.023, "step": 48350 }, { "grad_norm": 0.15736083686351776, "learning_rate": 5.6825638597281404e-05, "loss": 0.0254, "step": 48360 }, { "grad_norm": 0.23321811854839325, "learning_rate": 5.680925831888546e-05, "loss": 0.0238, "step": 48370 }, { "grad_norm": 0.15237879753112793, "learning_rate": 5.6792877295839274e-05, "loss": 0.0209, "step": 48380 }, { "grad_norm": 0.26524484157562256, "learning_rate": 5.6776495529934224e-05, "loss": 0.0247, "step": 48390 }, { "grad_norm": 0.178178608417511, "learning_rate": 5.6760113022961824e-05, "loss": 0.0176, "step": 48400 }, { "grad_norm": 0.18575944006443024, "learning_rate": 5.6743729776713617e-05, "loss": 0.0203, "step": 48410 }, { "grad_norm": 0.259090393781662, "learning_rate": 5.672734579298126e-05, "loss": 0.0203, "step": 48420 }, { "grad_norm": 0.20281898975372314, "learning_rate": 5.671096107355649e-05, "loss": 0.0204, "step": 48430 }, { "grad_norm": 0.258068710565567, "learning_rate": 5.66945756202311e-05, "loss": 0.0287, "step": 48440 }, { "grad_norm": 0.21472837030887604, "learning_rate": 5.667818943479699e-05, "loss": 0.0194, "step": 48450 }, { "grad_norm": 0.21071359515190125, "learning_rate": 5.666180251904612e-05, "loss": 0.0208, "step": 48460 }, { "grad_norm": 0.17300091683864594, "learning_rate": 5.6645414874770555e-05, "loss": 0.019, "step": 48470 }, { "grad_norm": 0.2457258254289627, "learning_rate": 5.66290265037624e-05, "loss": 0.0197, "step": 48480 }, { "grad_norm": 0.20683424174785614, "learning_rate": 5.661263740781386e-05, "loss": 0.0185, "step": 48490 }, { "grad_norm": 0.20571470260620117, "learning_rate": 5.6596247588717254e-05, "loss": 0.0198, "step": 48500 }, { "grad_norm": 0.22525329887866974, "learning_rate": 5.6579857048264926e-05, "loss": 0.0198, "step": 48510 }, { "grad_norm": 0.37721797823905945, "learning_rate": 5.6563465788249314e-05, "loss": 0.0178, "step": 48520 }, { "grad_norm": 0.19830791652202606, "learning_rate": 5.6547073810462956e-05, "loss": 0.0164, "step": 48530 }, { "grad_norm": 0.17638325691223145, "learning_rate": 5.653068111669846e-05, "loss": 0.0204, "step": 48540 }, { "grad_norm": 0.4195171892642975, "learning_rate": 5.651428770874848e-05, "loss": 0.0215, "step": 48550 }, { "grad_norm": 0.24332796037197113, "learning_rate": 5.64978935884058e-05, "loss": 0.0186, "step": 48560 }, { "grad_norm": 0.193918377161026, "learning_rate": 5.6481498757463244e-05, "loss": 0.0229, "step": 48570 }, { "grad_norm": 0.17955729365348816, "learning_rate": 5.646510321771373e-05, "loss": 0.0244, "step": 48580 }, { "grad_norm": 0.23565438389778137, "learning_rate": 5.644870697095024e-05, "loss": 0.0203, "step": 48590 }, { "grad_norm": 0.21002615988254547, "learning_rate": 5.643231001896586e-05, "loss": 0.0187, "step": 48600 }, { "grad_norm": 0.3869529068470001, "learning_rate": 5.6415912363553726e-05, "loss": 0.0242, "step": 48610 }, { "grad_norm": 0.2747184932231903, "learning_rate": 5.639951400650706e-05, "loss": 0.027, "step": 48620 }, { "grad_norm": 0.20651844143867493, "learning_rate": 5.6383114949619165e-05, "loss": 0.0203, "step": 48630 }, { "grad_norm": 0.206756129860878, "learning_rate": 5.636671519468342e-05, "loss": 0.0212, "step": 48640 }, { "grad_norm": 0.14892137050628662, "learning_rate": 5.635031474349327e-05, "loss": 0.019, "step": 48650 }, { "grad_norm": 0.15844175219535828, "learning_rate": 5.6333913597842246e-05, "loss": 0.0196, "step": 48660 }, { "grad_norm": 0.1787603348493576, "learning_rate": 5.6317511759523955e-05, "loss": 0.0192, "step": 48670 }, { "grad_norm": 0.20453357696533203, "learning_rate": 5.630110923033207e-05, "loss": 0.0202, "step": 48680 }, { "grad_norm": 0.36268654465675354, "learning_rate": 5.628470601206036e-05, "loss": 0.0206, "step": 48690 }, { "grad_norm": 0.2215774953365326, "learning_rate": 5.626830210650266e-05, "loss": 0.0187, "step": 48700 }, { "grad_norm": 0.2664601504802704, "learning_rate": 5.625189751545285e-05, "loss": 0.0221, "step": 48710 }, { "grad_norm": 0.23268969357013702, "learning_rate": 5.6235492240704936e-05, "loss": 0.0255, "step": 48720 }, { "grad_norm": 0.23264744877815247, "learning_rate": 5.621908628405296e-05, "loss": 0.0217, "step": 48730 }, { "grad_norm": 0.26847925782203674, "learning_rate": 5.620267964729106e-05, "loss": 0.0238, "step": 48740 }, { "grad_norm": 0.3023509085178375, "learning_rate": 5.618627233221344e-05, "loss": 0.0227, "step": 48750 }, { "grad_norm": 0.2042214721441269, "learning_rate": 5.616986434061438e-05, "loss": 0.0202, "step": 48760 }, { "grad_norm": 0.20680752396583557, "learning_rate": 5.615345567428822e-05, "loss": 0.022, "step": 48770 }, { "grad_norm": 0.20712924003601074, "learning_rate": 5.613704633502941e-05, "loss": 0.0228, "step": 48780 }, { "grad_norm": 0.18608812987804413, "learning_rate": 5.612063632463245e-05, "loss": 0.0218, "step": 48790 }, { "grad_norm": 0.18639551103115082, "learning_rate": 5.610422564489188e-05, "loss": 0.022, "step": 48800 }, { "grad_norm": 0.2714344561100006, "learning_rate": 5.608781429760239e-05, "loss": 0.0225, "step": 48810 }, { "grad_norm": 0.2664775550365448, "learning_rate": 5.607140228455866e-05, "loss": 0.0234, "step": 48820 }, { "grad_norm": 0.23826363682746887, "learning_rate": 5.605498960755553e-05, "loss": 0.023, "step": 48830 }, { "grad_norm": 0.1649271845817566, "learning_rate": 5.603857626838782e-05, "loss": 0.0237, "step": 48840 }, { "grad_norm": 0.24659410119056702, "learning_rate": 5.60221622688505e-05, "loss": 0.0221, "step": 48850 }, { "grad_norm": 0.21013423800468445, "learning_rate": 5.6005747610738565e-05, "loss": 0.0209, "step": 48860 }, { "grad_norm": 0.29503482580184937, "learning_rate": 5.59893322958471e-05, "loss": 0.0231, "step": 48870 }, { "grad_norm": 0.23280668258666992, "learning_rate": 5.5972916325971256e-05, "loss": 0.0252, "step": 48880 }, { "grad_norm": 0.22791115939617157, "learning_rate": 5.595649970290628e-05, "loss": 0.0186, "step": 48890 }, { "grad_norm": 0.23515312373638153, "learning_rate": 5.5940082428447426e-05, "loss": 0.0219, "step": 48900 }, { "grad_norm": 0.1773076355457306, "learning_rate": 5.592366450439012e-05, "loss": 0.0173, "step": 48910 }, { "grad_norm": 0.19071000814437866, "learning_rate": 5.590724593252975e-05, "loss": 0.0203, "step": 48920 }, { "grad_norm": 0.2466122955083847, "learning_rate": 5.589082671466184e-05, "loss": 0.0189, "step": 48930 }, { "grad_norm": 0.1797751933336258, "learning_rate": 5.587440685258199e-05, "loss": 0.0224, "step": 48940 }, { "grad_norm": 0.3034866452217102, "learning_rate": 5.585798634808583e-05, "loss": 0.0223, "step": 48950 }, { "grad_norm": 0.18060731887817383, "learning_rate": 5.584156520296909e-05, "loss": 0.0189, "step": 48960 }, { "grad_norm": 0.2503248453140259, "learning_rate": 5.582514341902757e-05, "loss": 0.0225, "step": 48970 }, { "grad_norm": 0.3510168194770813, "learning_rate": 5.580872099805713e-05, "loss": 0.0285, "step": 48980 }, { "grad_norm": 0.4824807047843933, "learning_rate": 5.5792297941853674e-05, "loss": 0.0235, "step": 48990 }, { "grad_norm": 0.2120586335659027, "learning_rate": 5.5775874252213247e-05, "loss": 0.025, "step": 49000 }, { "grad_norm": 0.17702721059322357, "learning_rate": 5.575944993093189e-05, "loss": 0.0234, "step": 49010 }, { "grad_norm": 0.20513229072093964, "learning_rate": 5.574302497980574e-05, "loss": 0.0191, "step": 49020 }, { "grad_norm": 0.19423605501651764, "learning_rate": 5.5726599400631005e-05, "loss": 0.0228, "step": 49030 }, { "grad_norm": 0.31452491879463196, "learning_rate": 5.571017319520399e-05, "loss": 0.0225, "step": 49040 }, { "grad_norm": 0.2250855565071106, "learning_rate": 5.5693746365320984e-05, "loss": 0.0213, "step": 49050 }, { "grad_norm": 0.2595352828502655, "learning_rate": 5.567731891277848e-05, "loss": 0.0222, "step": 49060 }, { "grad_norm": 0.18274502456188202, "learning_rate": 5.566089083937288e-05, "loss": 0.0225, "step": 49070 }, { "grad_norm": 0.22333109378814697, "learning_rate": 5.564446214690079e-05, "loss": 0.0201, "step": 49080 }, { "grad_norm": 0.18888306617736816, "learning_rate": 5.56280328371588e-05, "loss": 0.0217, "step": 49090 }, { "grad_norm": 0.20023031532764435, "learning_rate": 5.5611602911943596e-05, "loss": 0.0268, "step": 49100 }, { "grad_norm": 0.24494057893753052, "learning_rate": 5.559517237305194e-05, "loss": 0.0242, "step": 49110 }, { "grad_norm": 0.19665929675102234, "learning_rate": 5.557874122228064e-05, "loss": 0.0213, "step": 49120 }, { "grad_norm": 0.18165944516658783, "learning_rate": 5.55623094614266e-05, "loss": 0.0229, "step": 49130 }, { "grad_norm": 0.23146794736385345, "learning_rate": 5.5545877092286743e-05, "loss": 0.0225, "step": 49140 }, { "grad_norm": 0.20351682603359222, "learning_rate": 5.552944411665811e-05, "loss": 0.0222, "step": 49150 }, { "grad_norm": 0.18096649646759033, "learning_rate": 5.5513010536337795e-05, "loss": 0.0178, "step": 49160 }, { "grad_norm": 0.2743145823478699, "learning_rate": 5.5496576353122934e-05, "loss": 0.0213, "step": 49170 }, { "grad_norm": 0.26061898469924927, "learning_rate": 5.548014156881074e-05, "loss": 0.0204, "step": 49180 }, { "grad_norm": 0.128851056098938, "learning_rate": 5.5463706185198494e-05, "loss": 0.023, "step": 49190 }, { "grad_norm": 0.2541314363479614, "learning_rate": 5.544727020408358e-05, "loss": 0.0198, "step": 49200 }, { "grad_norm": 0.15604154765605927, "learning_rate": 5.5430833627263366e-05, "loss": 0.0194, "step": 49210 }, { "grad_norm": 0.23373669385910034, "learning_rate": 5.5414396456535354e-05, "loss": 0.0207, "step": 49220 }, { "grad_norm": 0.3052436411380768, "learning_rate": 5.539795869369709e-05, "loss": 0.0233, "step": 49230 }, { "grad_norm": 0.1479836106300354, "learning_rate": 5.538152034054618e-05, "loss": 0.0175, "step": 49240 }, { "grad_norm": 0.15801459550857544, "learning_rate": 5.536508139888028e-05, "loss": 0.0175, "step": 49250 }, { "grad_norm": 0.19466324150562286, "learning_rate": 5.534864187049716e-05, "loss": 0.0229, "step": 49260 }, { "grad_norm": 0.19775179028511047, "learning_rate": 5.5332201757194615e-05, "loss": 0.0184, "step": 49270 }, { "grad_norm": 0.18937894701957703, "learning_rate": 5.531576106077049e-05, "loss": 0.0186, "step": 49280 }, { "grad_norm": 0.27091825008392334, "learning_rate": 5.529931978302272e-05, "loss": 0.0187, "step": 49290 }, { "grad_norm": 0.3239782750606537, "learning_rate": 5.528287792574932e-05, "loss": 0.022, "step": 49300 }, { "grad_norm": 0.3406291902065277, "learning_rate": 5.5266435490748324e-05, "loss": 0.0254, "step": 49310 }, { "grad_norm": 0.5152495503425598, "learning_rate": 5.524999247981787e-05, "loss": 0.0191, "step": 49320 }, { "grad_norm": 0.2116888463497162, "learning_rate": 5.523354889475613e-05, "loss": 0.0222, "step": 49330 }, { "grad_norm": 0.2703687250614166, "learning_rate": 5.521710473736134e-05, "loss": 0.0216, "step": 49340 }, { "grad_norm": 1.3107881546020508, "learning_rate": 5.5200660009431826e-05, "loss": 0.0227, "step": 49350 }, { "grad_norm": 0.31111422181129456, "learning_rate": 5.518421471276596e-05, "loss": 0.0236, "step": 49360 }, { "grad_norm": 0.4307335913181305, "learning_rate": 5.5167768849162174e-05, "loss": 0.0296, "step": 49370 }, { "grad_norm": 0.244975283741951, "learning_rate": 5.515132242041893e-05, "loss": 0.021, "step": 49380 }, { "grad_norm": 0.25802502036094666, "learning_rate": 5.513487542833483e-05, "loss": 0.0217, "step": 49390 }, { "grad_norm": 0.2513744831085205, "learning_rate": 5.5118427874708466e-05, "loss": 0.0218, "step": 49400 }, { "grad_norm": 0.206955224275589, "learning_rate": 5.510197976133853e-05, "loss": 0.0207, "step": 49410 }, { "grad_norm": 0.20225495100021362, "learning_rate": 5.508553109002376e-05, "loss": 0.022, "step": 49420 }, { "grad_norm": 0.22065532207489014, "learning_rate": 5.5069081862562957e-05, "loss": 0.0201, "step": 49430 }, { "grad_norm": 0.23974481225013733, "learning_rate": 5.5052632080754965e-05, "loss": 0.0266, "step": 49440 }, { "grad_norm": 0.32006946206092834, "learning_rate": 5.503618174639874e-05, "loss": 0.0236, "step": 49450 }, { "grad_norm": 0.7927781343460083, "learning_rate": 5.501973086129325e-05, "loss": 0.0228, "step": 49460 }, { "grad_norm": 0.1708076000213623, "learning_rate": 5.500327942723753e-05, "loss": 0.0226, "step": 49470 }, { "grad_norm": 0.18653616309165955, "learning_rate": 5.498682744603071e-05, "loss": 0.0199, "step": 49480 }, { "grad_norm": 0.2130880504846573, "learning_rate": 5.497037491947194e-05, "loss": 0.0212, "step": 49490 }, { "grad_norm": 0.17773860692977905, "learning_rate": 5.4953921849360424e-05, "loss": 0.0196, "step": 49500 }, { "grad_norm": 0.22943589091300964, "learning_rate": 5.493746823749547e-05, "loss": 0.0229, "step": 49510 }, { "grad_norm": 0.17139190435409546, "learning_rate": 5.49210140856764e-05, "loss": 0.0272, "step": 49520 }, { "grad_norm": 0.1992143988609314, "learning_rate": 5.4904559395702647e-05, "loss": 0.02, "step": 49530 }, { "grad_norm": 0.7650502324104309, "learning_rate": 5.488810416937364e-05, "loss": 0.027, "step": 49540 }, { "grad_norm": 0.2932991087436676, "learning_rate": 5.4871648408488926e-05, "loss": 0.0228, "step": 49550 }, { "grad_norm": 0.16946686804294586, "learning_rate": 5.485519211484807e-05, "loss": 0.0252, "step": 49560 }, { "grad_norm": 0.24985425174236298, "learning_rate": 5.4838735290250676e-05, "loss": 0.0263, "step": 49570 }, { "grad_norm": 0.8539372682571411, "learning_rate": 5.482227793649649e-05, "loss": 0.0214, "step": 49580 }, { "grad_norm": 0.2584572434425354, "learning_rate": 5.480582005538524e-05, "loss": 0.0252, "step": 49590 }, { "grad_norm": 0.9505881667137146, "learning_rate": 5.478936164871671e-05, "loss": 0.0247, "step": 49600 }, { "grad_norm": 0.2847954034805298, "learning_rate": 5.477290271829081e-05, "loss": 0.0287, "step": 49610 }, { "grad_norm": 0.30186721682548523, "learning_rate": 5.475644326590744e-05, "loss": 0.0235, "step": 49620 }, { "grad_norm": 0.2377094179391861, "learning_rate": 5.473998329336658e-05, "loss": 0.0252, "step": 49630 }, { "grad_norm": 0.23714089393615723, "learning_rate": 5.4723522802468286e-05, "loss": 0.0256, "step": 49640 }, { "grad_norm": 0.19461590051651, "learning_rate": 5.470706179501264e-05, "loss": 0.0216, "step": 49650 }, { "grad_norm": 0.24092905223369598, "learning_rate": 5.4690600272799776e-05, "loss": 0.0216, "step": 49660 }, { "grad_norm": 0.17638559639453888, "learning_rate": 5.467413823762993e-05, "loss": 0.023, "step": 49670 }, { "grad_norm": 0.27177563309669495, "learning_rate": 5.465767569130335e-05, "loss": 0.023, "step": 49680 }, { "grad_norm": 0.19298014044761658, "learning_rate": 5.464121263562036e-05, "loss": 0.0224, "step": 49690 }, { "grad_norm": 0.2380037009716034, "learning_rate": 5.4624749072381343e-05, "loss": 0.0245, "step": 49700 }, { "grad_norm": 0.3598839044570923, "learning_rate": 5.460828500338672e-05, "loss": 0.0216, "step": 49710 }, { "grad_norm": 0.1510820984840393, "learning_rate": 5.459182043043698e-05, "loss": 0.0194, "step": 49720 }, { "grad_norm": 0.17177608609199524, "learning_rate": 5.457535535533265e-05, "loss": 0.0199, "step": 49730 }, { "grad_norm": 0.25331324338912964, "learning_rate": 5.4558889779874334e-05, "loss": 0.0222, "step": 49740 }, { "grad_norm": 0.2507387101650238, "learning_rate": 5.454242370586269e-05, "loss": 0.0207, "step": 49750 }, { "grad_norm": 0.2264251559972763, "learning_rate": 5.452595713509843e-05, "loss": 0.02, "step": 49760 }, { "grad_norm": 0.20437155663967133, "learning_rate": 5.4509490069382274e-05, "loss": 0.0173, "step": 49770 }, { "grad_norm": 0.1765705943107605, "learning_rate": 5.449302251051509e-05, "loss": 0.0199, "step": 49780 }, { "grad_norm": 0.18686655163764954, "learning_rate": 5.44765544602977e-05, "loss": 0.0202, "step": 49790 }, { "grad_norm": 0.5280807018280029, "learning_rate": 5.446008592053107e-05, "loss": 0.0234, "step": 49800 }, { "grad_norm": 0.2258739471435547, "learning_rate": 5.4443616893016135e-05, "loss": 0.0202, "step": 49810 }, { "grad_norm": 0.1766776740550995, "learning_rate": 5.4427147379553934e-05, "loss": 0.0211, "step": 49820 }, { "grad_norm": 0.20018190145492554, "learning_rate": 5.441067738194556e-05, "loss": 0.022, "step": 49830 }, { "grad_norm": 0.2217615395784378, "learning_rate": 5.439420690199214e-05, "loss": 0.0217, "step": 49840 }, { "grad_norm": 0.3057945966720581, "learning_rate": 5.437773594149484e-05, "loss": 0.0192, "step": 49850 }, { "grad_norm": 0.23598700761795044, "learning_rate": 5.436126450225495e-05, "loss": 0.022, "step": 49860 }, { "grad_norm": 0.24665328860282898, "learning_rate": 5.434479258607371e-05, "loss": 0.0214, "step": 49870 }, { "grad_norm": 0.2503466308116913, "learning_rate": 5.43283201947525e-05, "loss": 0.0203, "step": 49880 }, { "grad_norm": 0.1988745629787445, "learning_rate": 5.431184733009268e-05, "loss": 0.0239, "step": 49890 }, { "grad_norm": 0.20245514810085297, "learning_rate": 5.4295373993895736e-05, "loss": 0.0218, "step": 49900 }, { "grad_norm": 0.24491091072559357, "learning_rate": 5.4278900187963157e-05, "loss": 0.0233, "step": 49910 }, { "grad_norm": 0.16245149075984955, "learning_rate": 5.426242591409646e-05, "loss": 0.0202, "step": 49920 }, { "grad_norm": 0.15843342244625092, "learning_rate": 5.42459511740973e-05, "loss": 0.0207, "step": 49930 }, { "grad_norm": 0.21912552416324615, "learning_rate": 5.422947596976729e-05, "loss": 0.0203, "step": 49940 }, { "grad_norm": 0.23066367208957672, "learning_rate": 5.4213000302908134e-05, "loss": 0.0209, "step": 49950 }, { "grad_norm": 0.23848019540309906, "learning_rate": 5.419652417532162e-05, "loss": 0.0249, "step": 49960 }, { "grad_norm": 0.16534756124019623, "learning_rate": 5.4180047588809534e-05, "loss": 0.0208, "step": 49970 }, { "grad_norm": 0.27443692088127136, "learning_rate": 5.4163570545173704e-05, "loss": 0.0189, "step": 49980 }, { "grad_norm": 0.13816174864768982, "learning_rate": 5.414709304621608e-05, "loss": 0.0254, "step": 49990 }, { "grad_norm": 0.5084081292152405, "learning_rate": 5.413061509373858e-05, "loss": 0.0212, "step": 50000 }, { "grad_norm": 0.18362922966480255, "learning_rate": 5.411413668954324e-05, "loss": 0.019, "step": 50010 }, { "grad_norm": 0.17849482595920563, "learning_rate": 5.4097657835432083e-05, "loss": 0.0187, "step": 50020 }, { "grad_norm": 0.13425329327583313, "learning_rate": 5.408117853320723e-05, "loss": 0.0181, "step": 50030 }, { "grad_norm": 0.26585105061531067, "learning_rate": 5.406469878467084e-05, "loss": 0.0205, "step": 50040 }, { "grad_norm": 0.23796346783638, "learning_rate": 5.404821859162509e-05, "loss": 0.0215, "step": 50050 }, { "grad_norm": 0.2714347243309021, "learning_rate": 5.403173795587225e-05, "loss": 0.0225, "step": 50060 }, { "grad_norm": 0.18756654858589172, "learning_rate": 5.4015256879214606e-05, "loss": 0.02, "step": 50070 }, { "grad_norm": 0.2646290957927704, "learning_rate": 5.39987753634545e-05, "loss": 0.0215, "step": 50080 }, { "grad_norm": 0.20367243885993958, "learning_rate": 5.3982293410394335e-05, "loss": 0.0218, "step": 50090 }, { "grad_norm": 0.18030864000320435, "learning_rate": 5.396581102183655e-05, "loss": 0.0194, "step": 50100 }, { "grad_norm": 0.21288666129112244, "learning_rate": 5.3949328199583634e-05, "loss": 0.0216, "step": 50110 }, { "grad_norm": 0.18792079389095306, "learning_rate": 5.393284494543813e-05, "loss": 0.0237, "step": 50120 }, { "grad_norm": 0.18800948560237885, "learning_rate": 5.391636126120262e-05, "loss": 0.0201, "step": 50130 }, { "grad_norm": 0.5890581011772156, "learning_rate": 5.389987714867971e-05, "loss": 0.0281, "step": 50140 }, { "grad_norm": 0.19326703250408173, "learning_rate": 5.3883392609672114e-05, "loss": 0.0209, "step": 50150 }, { "grad_norm": 0.32914963364601135, "learning_rate": 5.3866907645982545e-05, "loss": 0.0206, "step": 50160 }, { "grad_norm": 0.1972847282886505, "learning_rate": 5.385042225941375e-05, "loss": 0.0188, "step": 50170 }, { "grad_norm": 0.2357843518257141, "learning_rate": 5.3833936451768585e-05, "loss": 0.0186, "step": 50180 }, { "grad_norm": 0.3503594994544983, "learning_rate": 5.3817450224849884e-05, "loss": 0.0191, "step": 50190 }, { "grad_norm": 0.15183450281620026, "learning_rate": 5.380096358046056e-05, "loss": 0.021, "step": 50200 }, { "grad_norm": 0.16407017409801483, "learning_rate": 5.378447652040359e-05, "loss": 0.0193, "step": 50210 }, { "grad_norm": 0.17155368626117706, "learning_rate": 5.376798904648195e-05, "loss": 0.0176, "step": 50220 }, { "grad_norm": 0.2611991763114929, "learning_rate": 5.375150116049869e-05, "loss": 0.0244, "step": 50230 }, { "grad_norm": 0.24502122402191162, "learning_rate": 5.373501286425691e-05, "loss": 0.0235, "step": 50240 }, { "grad_norm": 0.19383612275123596, "learning_rate": 5.3718524159559726e-05, "loss": 0.0199, "step": 50250 }, { "grad_norm": 0.2596430480480194, "learning_rate": 5.370203504821034e-05, "loss": 0.0198, "step": 50260 }, { "grad_norm": 0.2196105718612671, "learning_rate": 5.368554553201196e-05, "loss": 0.022, "step": 50270 }, { "grad_norm": 0.22221669554710388, "learning_rate": 5.3669055612767874e-05, "loss": 0.0193, "step": 50280 }, { "grad_norm": 0.2063293606042862, "learning_rate": 5.3652565292281376e-05, "loss": 0.0208, "step": 50290 }, { "grad_norm": 0.2487500011920929, "learning_rate": 5.3636074572355824e-05, "loss": 0.0196, "step": 50300 }, { "grad_norm": 0.23694458603858948, "learning_rate": 5.361958345479463e-05, "loss": 0.0196, "step": 50310 }, { "grad_norm": 0.29600682854652405, "learning_rate": 5.360309194140123e-05, "loss": 0.0259, "step": 50320 }, { "grad_norm": 0.24049068987369537, "learning_rate": 5.3586600033979106e-05, "loss": 0.0219, "step": 50330 }, { "grad_norm": 0.21827007830142975, "learning_rate": 5.357010773433181e-05, "loss": 0.0207, "step": 50340 }, { "grad_norm": 0.20633019506931305, "learning_rate": 5.3553615044262895e-05, "loss": 0.0216, "step": 50350 }, { "grad_norm": 0.2987828552722931, "learning_rate": 5.3537121965575964e-05, "loss": 0.0205, "step": 50360 }, { "grad_norm": 0.15972605347633362, "learning_rate": 5.352062850007471e-05, "loss": 0.0183, "step": 50370 }, { "grad_norm": 0.2836991548538208, "learning_rate": 5.350413464956282e-05, "loss": 0.0211, "step": 50380 }, { "grad_norm": 0.3399781286716461, "learning_rate": 5.348764041584403e-05, "loss": 0.0205, "step": 50390 }, { "grad_norm": 0.2864149212837219, "learning_rate": 5.347114580072212e-05, "loss": 0.022, "step": 50400 }, { "grad_norm": 0.24122320115566254, "learning_rate": 5.345465080600093e-05, "loss": 0.0199, "step": 50410 }, { "grad_norm": 0.22215503454208374, "learning_rate": 5.3438155433484326e-05, "loss": 0.0187, "step": 50420 }, { "grad_norm": 1.470802903175354, "learning_rate": 5.3421659684976197e-05, "loss": 0.023, "step": 50430 }, { "grad_norm": 0.3002189099788666, "learning_rate": 5.340516356228052e-05, "loss": 0.0211, "step": 50440 }, { "grad_norm": 0.22574719786643982, "learning_rate": 5.338866706720128e-05, "loss": 0.0231, "step": 50450 }, { "grad_norm": 0.2675262689590454, "learning_rate": 5.337217020154249e-05, "loss": 0.0208, "step": 50460 }, { "grad_norm": 0.302083283662796, "learning_rate": 5.335567296710825e-05, "loss": 0.02, "step": 50470 }, { "grad_norm": 0.19742578268051147, "learning_rate": 5.333917536570265e-05, "loss": 0.0225, "step": 50480 }, { "grad_norm": 0.5370926856994629, "learning_rate": 5.332267739912986e-05, "loss": 0.0195, "step": 50490 }, { "grad_norm": 0.16838008165359497, "learning_rate": 5.330617906919405e-05, "loss": 0.0227, "step": 50500 }, { "grad_norm": 0.1896119862794876, "learning_rate": 5.328968037769949e-05, "loss": 0.0188, "step": 50510 }, { "grad_norm": 0.2595866918563843, "learning_rate": 5.327318132645042e-05, "loss": 0.0191, "step": 50520 }, { "grad_norm": 0.3727477192878723, "learning_rate": 5.325668191725116e-05, "loss": 0.0228, "step": 50530 }, { "grad_norm": 0.20801609754562378, "learning_rate": 5.324018215190606e-05, "loss": 0.0216, "step": 50540 }, { "grad_norm": 0.2555764317512512, "learning_rate": 5.3223682032219515e-05, "loss": 0.0192, "step": 50550 }, { "grad_norm": 0.1574224978685379, "learning_rate": 5.320718155999595e-05, "loss": 0.0206, "step": 50560 }, { "grad_norm": 0.2471071183681488, "learning_rate": 5.3190680737039835e-05, "loss": 0.0236, "step": 50570 }, { "grad_norm": 0.18195804953575134, "learning_rate": 5.317417956515567e-05, "loss": 0.0224, "step": 50580 }, { "grad_norm": 0.18759378790855408, "learning_rate": 5.315767804614798e-05, "loss": 0.0216, "step": 50590 }, { "grad_norm": 0.35339343547821045, "learning_rate": 5.3141176181821395e-05, "loss": 0.0207, "step": 50600 }, { "grad_norm": 0.2917289435863495, "learning_rate": 5.31246739739805e-05, "loss": 0.024, "step": 50610 }, { "grad_norm": 0.242282897233963, "learning_rate": 5.310817142442994e-05, "loss": 0.0216, "step": 50620 }, { "grad_norm": 0.39704960584640503, "learning_rate": 5.309166853497445e-05, "loss": 0.0223, "step": 50630 }, { "grad_norm": 0.19591209292411804, "learning_rate": 5.307516530741873e-05, "loss": 0.0198, "step": 50640 }, { "grad_norm": 1.3553509712219238, "learning_rate": 5.305866174356754e-05, "loss": 0.02, "step": 50650 }, { "grad_norm": 0.18951676785945892, "learning_rate": 5.304215784522571e-05, "loss": 0.024, "step": 50660 }, { "grad_norm": 0.15882572531700134, "learning_rate": 5.302565361419808e-05, "loss": 0.0181, "step": 50670 }, { "grad_norm": 0.1990468055009842, "learning_rate": 5.3009149052289507e-05, "loss": 0.0201, "step": 50680 }, { "grad_norm": 0.27316856384277344, "learning_rate": 5.299264416130493e-05, "loss": 0.0193, "step": 50690 }, { "grad_norm": 0.24557168781757355, "learning_rate": 5.297613894304928e-05, "loss": 0.019, "step": 50700 }, { "grad_norm": 0.53412926197052, "learning_rate": 5.2959633399327534e-05, "loss": 0.0215, "step": 50710 }, { "grad_norm": 0.38793322443962097, "learning_rate": 5.294312753194476e-05, "loss": 0.0214, "step": 50720 }, { "grad_norm": 0.17592494189739227, "learning_rate": 5.292662134270596e-05, "loss": 0.0206, "step": 50730 }, { "grad_norm": 0.2418559044599533, "learning_rate": 5.291011483341626e-05, "loss": 0.0273, "step": 50740 }, { "grad_norm": 0.2177807092666626, "learning_rate": 5.2893608005880767e-05, "loss": 0.0242, "step": 50750 }, { "grad_norm": 0.2764587104320526, "learning_rate": 5.287710086190467e-05, "loss": 0.0209, "step": 50760 }, { "grad_norm": 0.21777862310409546, "learning_rate": 5.286059340329314e-05, "loss": 0.0209, "step": 50770 }, { "grad_norm": 0.18467220664024353, "learning_rate": 5.284408563185141e-05, "loss": 0.0199, "step": 50780 }, { "grad_norm": 0.22036685049533844, "learning_rate": 5.282757754938476e-05, "loss": 0.0244, "step": 50790 }, { "grad_norm": 0.1945999264717102, "learning_rate": 5.281106915769849e-05, "loss": 0.0209, "step": 50800 }, { "grad_norm": 0.19399231672286987, "learning_rate": 5.2794560458597897e-05, "loss": 0.026, "step": 50810 }, { "grad_norm": 0.18468821048736572, "learning_rate": 5.27780514538884e-05, "loss": 0.0209, "step": 50820 }, { "grad_norm": 0.26489153504371643, "learning_rate": 5.2761542145375365e-05, "loss": 0.0257, "step": 50830 }, { "grad_norm": 0.15248142182826996, "learning_rate": 5.274503253486421e-05, "loss": 0.0218, "step": 50840 }, { "grad_norm": 0.2789600193500519, "learning_rate": 5.272852262416046e-05, "loss": 0.0181, "step": 50850 }, { "grad_norm": 0.16428346931934357, "learning_rate": 5.2712012415069555e-05, "loss": 0.022, "step": 50860 }, { "grad_norm": 0.23180879652500153, "learning_rate": 5.269550190939705e-05, "loss": 0.0251, "step": 50870 }, { "grad_norm": 0.25701361894607544, "learning_rate": 5.267899110894852e-05, "loss": 0.0217, "step": 50880 }, { "grad_norm": 0.19541649520397186, "learning_rate": 5.266248001552955e-05, "loss": 0.0193, "step": 50890 }, { "grad_norm": 0.20780938863754272, "learning_rate": 5.264596863094575e-05, "loss": 0.0195, "step": 50900 }, { "grad_norm": 0.24163003265857697, "learning_rate": 5.262945695700282e-05, "loss": 0.0255, "step": 50910 }, { "grad_norm": 0.47288277745246887, "learning_rate": 5.261294499550643e-05, "loss": 0.0215, "step": 50920 }, { "grad_norm": 0.2011367380619049, "learning_rate": 5.25964327482623e-05, "loss": 0.0238, "step": 50930 }, { "grad_norm": 0.2240450084209442, "learning_rate": 5.257992021707617e-05, "loss": 0.0245, "step": 50940 }, { "grad_norm": 0.27531448006629944, "learning_rate": 5.256340740375387e-05, "loss": 0.0251, "step": 50950 }, { "grad_norm": 0.20829258859157562, "learning_rate": 5.254689431010117e-05, "loss": 0.0212, "step": 50960 }, { "grad_norm": 0.2026626467704773, "learning_rate": 5.253038093792395e-05, "loss": 0.0221, "step": 50970 }, { "grad_norm": 0.26938503980636597, "learning_rate": 5.251386728902806e-05, "loss": 0.0207, "step": 50980 }, { "grad_norm": 0.28136777877807617, "learning_rate": 5.2497353365219446e-05, "loss": 0.0195, "step": 50990 }, { "grad_norm": 0.227772518992424, "learning_rate": 5.2480839168304e-05, "loss": 0.0171, "step": 51000 }, { "grad_norm": 0.23512578010559082, "learning_rate": 5.246432470008773e-05, "loss": 0.0181, "step": 51010 }, { "grad_norm": 0.24161769449710846, "learning_rate": 5.2447809962376616e-05, "loss": 0.0212, "step": 51020 }, { "grad_norm": 0.2519376277923584, "learning_rate": 5.243129495697666e-05, "loss": 0.0207, "step": 51030 }, { "grad_norm": 0.21513934433460236, "learning_rate": 5.241477968569397e-05, "loss": 0.0237, "step": 51040 }, { "grad_norm": 0.2821178138256073, "learning_rate": 5.23982641503346e-05, "loss": 0.0196, "step": 51050 }, { "grad_norm": 0.22182641923427582, "learning_rate": 5.238174835270465e-05, "loss": 0.0237, "step": 51060 }, { "grad_norm": 0.21832185983657837, "learning_rate": 5.23652322946103e-05, "loss": 0.0191, "step": 51070 }, { "grad_norm": 0.29664021730422974, "learning_rate": 5.234871597785771e-05, "loss": 0.0275, "step": 51080 }, { "grad_norm": 0.20862099528312683, "learning_rate": 5.233219940425306e-05, "loss": 0.0211, "step": 51090 }, { "grad_norm": 0.2176508754491806, "learning_rate": 5.2315682575602576e-05, "loss": 0.0214, "step": 51100 }, { "grad_norm": 0.29363444447517395, "learning_rate": 5.229916549371254e-05, "loss": 0.023, "step": 51110 }, { "grad_norm": 0.16117410361766815, "learning_rate": 5.228264816038924e-05, "loss": 0.0183, "step": 51120 }, { "grad_norm": 0.21131160855293274, "learning_rate": 5.2266130577438935e-05, "loss": 0.0236, "step": 51130 }, { "grad_norm": 0.24134273827075958, "learning_rate": 5.224961274666802e-05, "loss": 0.0192, "step": 51140 }, { "grad_norm": 0.2638857960700989, "learning_rate": 5.223309466988283e-05, "loss": 0.0196, "step": 51150 }, { "grad_norm": 0.22059403359889984, "learning_rate": 5.221657634888976e-05, "loss": 0.0212, "step": 51160 }, { "grad_norm": 0.1787194013595581, "learning_rate": 5.2200057785495235e-05, "loss": 0.0222, "step": 51170 }, { "grad_norm": 0.5012800693511963, "learning_rate": 5.2183538981505706e-05, "loss": 0.0292, "step": 51180 }, { "grad_norm": 0.23485592007637024, "learning_rate": 5.216701993872762e-05, "loss": 0.025, "step": 51190 }, { "grad_norm": 0.1901111900806427, "learning_rate": 5.215050065896752e-05, "loss": 0.0233, "step": 51200 }, { "grad_norm": 0.2053927332162857, "learning_rate": 5.2133981144031875e-05, "loss": 0.0192, "step": 51210 }, { "grad_norm": 0.29864415526390076, "learning_rate": 5.211746139572727e-05, "loss": 0.0199, "step": 51220 }, { "grad_norm": 0.17700007557868958, "learning_rate": 5.210094141586026e-05, "loss": 0.0195, "step": 51230 }, { "grad_norm": 0.19446292519569397, "learning_rate": 5.2084421206237476e-05, "loss": 0.0194, "step": 51240 }, { "grad_norm": 0.2242283970117569, "learning_rate": 5.20679007686655e-05, "loss": 0.0205, "step": 51250 }, { "grad_norm": 0.11011243611574173, "learning_rate": 5.205138010495102e-05, "loss": 0.0197, "step": 51260 }, { "grad_norm": 0.3968088924884796, "learning_rate": 5.20348592169007e-05, "loss": 0.0195, "step": 51270 }, { "grad_norm": 0.2409173548221588, "learning_rate": 5.201833810632123e-05, "loss": 0.02, "step": 51280 }, { "grad_norm": 0.2521456182003021, "learning_rate": 5.200181677501933e-05, "loss": 0.0202, "step": 51290 }, { "grad_norm": 0.25075024366378784, "learning_rate": 5.1985295224801777e-05, "loss": 0.0199, "step": 51300 }, { "grad_norm": 0.18727348744869232, "learning_rate": 5.196877345747531e-05, "loss": 0.0214, "step": 51310 }, { "grad_norm": 0.35182589292526245, "learning_rate": 5.195225147484674e-05, "loss": 0.0208, "step": 51320 }, { "grad_norm": 0.22707244753837585, "learning_rate": 5.193572927872288e-05, "loss": 0.0229, "step": 51330 }, { "grad_norm": 0.12751545011997223, "learning_rate": 5.191920687091059e-05, "loss": 0.0246, "step": 51340 }, { "grad_norm": 0.1745281219482422, "learning_rate": 5.1902684253216695e-05, "loss": 0.0219, "step": 51350 }, { "grad_norm": 1.8300707340240479, "learning_rate": 5.1886161427448134e-05, "loss": 0.0203, "step": 51360 }, { "grad_norm": 0.3602466881275177, "learning_rate": 5.186963839541179e-05, "loss": 0.0193, "step": 51370 }, { "grad_norm": 0.15520140528678894, "learning_rate": 5.185311515891459e-05, "loss": 0.0179, "step": 51380 }, { "grad_norm": 0.17218616604804993, "learning_rate": 5.1836591719763516e-05, "loss": 0.0214, "step": 51390 }, { "grad_norm": 0.1891624927520752, "learning_rate": 5.182006807976554e-05, "loss": 0.0197, "step": 51400 }, { "grad_norm": 0.3852044343948364, "learning_rate": 5.180354424072763e-05, "loss": 0.0271, "step": 51410 }, { "grad_norm": 0.22552131116390228, "learning_rate": 5.178702020445686e-05, "loss": 0.0256, "step": 51420 }, { "grad_norm": 0.263302743434906, "learning_rate": 5.177049597276024e-05, "loss": 0.0187, "step": 51430 }, { "grad_norm": 0.5128048658370972, "learning_rate": 5.175397154744482e-05, "loss": 0.024, "step": 51440 }, { "grad_norm": 0.25112661719322205, "learning_rate": 5.1737446930317736e-05, "loss": 0.0219, "step": 51450 }, { "grad_norm": 0.25176718831062317, "learning_rate": 5.172092212318604e-05, "loss": 0.0208, "step": 51460 }, { "grad_norm": 0.2888098955154419, "learning_rate": 5.170439712785692e-05, "loss": 0.0231, "step": 51470 }, { "grad_norm": 0.19310036301612854, "learning_rate": 5.1687871946137467e-05, "loss": 0.0215, "step": 51480 }, { "grad_norm": 0.24562644958496094, "learning_rate": 5.16713465798349e-05, "loss": 0.0189, "step": 51490 }, { "grad_norm": 0.30732226371765137, "learning_rate": 5.165482103075637e-05, "loss": 0.0207, "step": 51500 }, { "grad_norm": 0.29346799850463867, "learning_rate": 5.163829530070909e-05, "loss": 0.022, "step": 51510 }, { "grad_norm": 0.22166824340820312, "learning_rate": 5.162176939150032e-05, "loss": 0.0209, "step": 51520 }, { "grad_norm": 0.2844517230987549, "learning_rate": 5.160524330493728e-05, "loss": 0.0208, "step": 51530 }, { "grad_norm": 0.32679328322410583, "learning_rate": 5.158871704282724e-05, "loss": 0.0205, "step": 51540 }, { "grad_norm": 0.1776181310415268, "learning_rate": 5.15721906069775e-05, "loss": 0.0201, "step": 51550 }, { "grad_norm": 0.21189187467098236, "learning_rate": 5.155566399919536e-05, "loss": 0.0204, "step": 51560 }, { "grad_norm": 0.24349795281887054, "learning_rate": 5.153913722128813e-05, "loss": 0.0235, "step": 51570 }, { "grad_norm": 0.22552251815795898, "learning_rate": 5.1522610275063176e-05, "loss": 0.0205, "step": 51580 }, { "grad_norm": 0.1996273696422577, "learning_rate": 5.150608316232786e-05, "loss": 0.0214, "step": 51590 }, { "grad_norm": 0.20247040688991547, "learning_rate": 5.148955588488955e-05, "loss": 0.0178, "step": 51600 }, { "grad_norm": 0.24301597476005554, "learning_rate": 5.147302844455564e-05, "loss": 0.0232, "step": 51610 }, { "grad_norm": 0.18861624598503113, "learning_rate": 5.145650084313356e-05, "loss": 0.0191, "step": 51620 }, { "grad_norm": 0.1999921053647995, "learning_rate": 5.1439973082430734e-05, "loss": 0.0178, "step": 51630 }, { "grad_norm": 0.27322787046432495, "learning_rate": 5.142344516425461e-05, "loss": 0.0206, "step": 51640 }, { "grad_norm": 0.15319229662418365, "learning_rate": 5.1406917090412673e-05, "loss": 0.0258, "step": 51650 }, { "grad_norm": 0.22218452394008636, "learning_rate": 5.1390388862712404e-05, "loss": 0.0198, "step": 51660 }, { "grad_norm": 0.15954025089740753, "learning_rate": 5.137386048296129e-05, "loss": 0.0234, "step": 51670 }, { "grad_norm": 0.1903212070465088, "learning_rate": 5.135733195296686e-05, "loss": 0.0175, "step": 51680 }, { "grad_norm": 0.13550131022930145, "learning_rate": 5.134080327453664e-05, "loss": 0.0209, "step": 51690 }, { "grad_norm": 0.16530287265777588, "learning_rate": 5.1324274449478204e-05, "loss": 0.0191, "step": 51700 }, { "grad_norm": 0.23262199759483337, "learning_rate": 5.130774547959909e-05, "loss": 0.0243, "step": 51710 }, { "grad_norm": 0.4477999806404114, "learning_rate": 5.129121636670692e-05, "loss": 0.0223, "step": 51720 }, { "grad_norm": 0.165884330868721, "learning_rate": 5.1274687112609246e-05, "loss": 0.0176, "step": 51730 }, { "grad_norm": 0.18365709483623505, "learning_rate": 5.125815771911372e-05, "loss": 0.0171, "step": 51740 }, { "grad_norm": 0.4589172303676605, "learning_rate": 5.1241628188027956e-05, "loss": 0.0232, "step": 51750 }, { "grad_norm": 0.2658640146255493, "learning_rate": 5.1225098521159585e-05, "loss": 0.0233, "step": 51760 }, { "grad_norm": 0.15745361149311066, "learning_rate": 5.12085687203163e-05, "loss": 0.0183, "step": 51770 }, { "grad_norm": 0.2226429581642151, "learning_rate": 5.119203878730575e-05, "loss": 0.0226, "step": 51780 }, { "grad_norm": 0.2182251513004303, "learning_rate": 5.117550872393564e-05, "loss": 0.022, "step": 51790 }, { "grad_norm": 0.20104391872882843, "learning_rate": 5.1158978532013645e-05, "loss": 0.0237, "step": 51800 }, { "grad_norm": 0.24862775206565857, "learning_rate": 5.114244821334751e-05, "loss": 0.0236, "step": 51810 }, { "grad_norm": 0.18968646228313446, "learning_rate": 5.112591776974497e-05, "loss": 0.0249, "step": 51820 }, { "grad_norm": 0.20793940126895905, "learning_rate": 5.110938720301374e-05, "loss": 0.0247, "step": 51830 }, { "grad_norm": 0.20735399425029755, "learning_rate": 5.1092856514961595e-05, "loss": 0.0215, "step": 51840 }, { "grad_norm": 0.1978490948677063, "learning_rate": 5.107632570739631e-05, "loss": 0.019, "step": 51850 }, { "grad_norm": 0.1390175074338913, "learning_rate": 5.1059794782125645e-05, "loss": 0.0176, "step": 51860 }, { "grad_norm": 0.3112298548221588, "learning_rate": 5.104326374095744e-05, "loss": 0.0196, "step": 51870 }, { "grad_norm": 0.15005318820476532, "learning_rate": 5.102673258569947e-05, "loss": 0.0178, "step": 51880 }, { "grad_norm": 0.22055372595787048, "learning_rate": 5.101020131815956e-05, "loss": 0.0192, "step": 51890 }, { "grad_norm": 0.18986818194389343, "learning_rate": 5.099366994014556e-05, "loss": 0.0187, "step": 51900 }, { "grad_norm": 0.22950422763824463, "learning_rate": 5.097713845346532e-05, "loss": 0.0184, "step": 51910 }, { "grad_norm": 0.19483362138271332, "learning_rate": 5.096060685992668e-05, "loss": 0.0212, "step": 51920 }, { "grad_norm": 0.1729717254638672, "learning_rate": 5.0944075161337514e-05, "loss": 0.0194, "step": 51930 }, { "grad_norm": 0.18992699682712555, "learning_rate": 5.09275433595057e-05, "loss": 0.0182, "step": 51940 }, { "grad_norm": 0.26199328899383545, "learning_rate": 5.0911011456239157e-05, "loss": 0.023, "step": 51950 }, { "grad_norm": 0.25916290283203125, "learning_rate": 5.089447945334574e-05, "loss": 0.0225, "step": 51960 }, { "grad_norm": 0.2047557234764099, "learning_rate": 5.087794735263343e-05, "loss": 0.0173, "step": 51970 }, { "grad_norm": 0.15708379447460175, "learning_rate": 5.086141515591011e-05, "loss": 0.0189, "step": 51980 }, { "grad_norm": 0.2568821907043457, "learning_rate": 5.084488286498371e-05, "loss": 0.0198, "step": 51990 }, { "grad_norm": 0.19841137528419495, "learning_rate": 5.08283504816622e-05, "loss": 0.0181, "step": 52000 }, { "grad_norm": 0.20375698804855347, "learning_rate": 5.081181800775353e-05, "loss": 0.0235, "step": 52010 }, { "grad_norm": 0.20230387151241302, "learning_rate": 5.079528544506565e-05, "loss": 0.0251, "step": 52020 }, { "grad_norm": 0.20972797274589539, "learning_rate": 5.077875279540657e-05, "loss": 0.0179, "step": 52030 }, { "grad_norm": 0.2702392041683197, "learning_rate": 5.076222006058424e-05, "loss": 0.0194, "step": 52040 }, { "grad_norm": 0.17797783017158508, "learning_rate": 5.074568724240667e-05, "loss": 0.021, "step": 52050 }, { "grad_norm": 0.21173378825187683, "learning_rate": 5.072915434268187e-05, "loss": 0.0229, "step": 52060 }, { "grad_norm": 0.1718119978904724, "learning_rate": 5.071262136321785e-05, "loss": 0.02, "step": 52070 }, { "grad_norm": 0.1739574819803238, "learning_rate": 5.069608830582262e-05, "loss": 0.0218, "step": 52080 }, { "grad_norm": 0.11288708448410034, "learning_rate": 5.067955517230423e-05, "loss": 0.019, "step": 52090 }, { "grad_norm": 0.19652871787548065, "learning_rate": 5.06630219644707e-05, "loss": 0.0169, "step": 52100 }, { "grad_norm": 0.243616983294487, "learning_rate": 5.064648868413008e-05, "loss": 0.0176, "step": 52110 }, { "grad_norm": 0.28092634677886963, "learning_rate": 5.062995533309044e-05, "loss": 0.0231, "step": 52120 }, { "grad_norm": 0.28134796023368835, "learning_rate": 5.061342191315983e-05, "loss": 0.0231, "step": 52130 }, { "grad_norm": 0.23750565946102142, "learning_rate": 5.0596888426146325e-05, "loss": 0.019, "step": 52140 }, { "grad_norm": 0.15598462522029877, "learning_rate": 5.058035487385798e-05, "loss": 0.0198, "step": 52150 }, { "grad_norm": 0.17818959057331085, "learning_rate": 5.0563821258102915e-05, "loss": 0.0188, "step": 52160 }, { "grad_norm": 0.11597523093223572, "learning_rate": 5.05472875806892e-05, "loss": 0.0212, "step": 52170 }, { "grad_norm": 0.2051052302122116, "learning_rate": 5.053075384342494e-05, "loss": 0.0235, "step": 52180 }, { "grad_norm": 0.20248574018478394, "learning_rate": 5.0514220048118234e-05, "loss": 0.0196, "step": 52190 }, { "grad_norm": 0.1848280429840088, "learning_rate": 5.049768619657721e-05, "loss": 0.0193, "step": 52200 }, { "grad_norm": 0.20934714376926422, "learning_rate": 5.048115229060996e-05, "loss": 0.0179, "step": 52210 }, { "grad_norm": 0.13713806867599487, "learning_rate": 5.046461833202462e-05, "loss": 0.018, "step": 52220 }, { "grad_norm": 0.1499006599187851, "learning_rate": 5.044808432262932e-05, "loss": 0.0171, "step": 52230 }, { "grad_norm": 0.2151191383600235, "learning_rate": 5.043155026423219e-05, "loss": 0.0198, "step": 52240 }, { "grad_norm": 0.25208041071891785, "learning_rate": 5.041501615864137e-05, "loss": 0.019, "step": 52250 }, { "grad_norm": 0.20061591267585754, "learning_rate": 5.039848200766502e-05, "loss": 0.0179, "step": 52260 }, { "grad_norm": 0.19426824152469635, "learning_rate": 5.038194781311125e-05, "loss": 0.0211, "step": 52270 }, { "grad_norm": 0.17133329808712006, "learning_rate": 5.0365413576788264e-05, "loss": 0.0161, "step": 52280 }, { "grad_norm": 0.369649738073349, "learning_rate": 5.0348879300504194e-05, "loss": 0.0207, "step": 52290 }, { "grad_norm": 0.15345560014247894, "learning_rate": 5.033234498606721e-05, "loss": 0.0179, "step": 52300 }, { "grad_norm": 0.2885396182537079, "learning_rate": 5.0315810635285476e-05, "loss": 0.0166, "step": 52310 }, { "grad_norm": 0.19466952979564667, "learning_rate": 5.0299276249967164e-05, "loss": 0.0154, "step": 52320 }, { "grad_norm": 0.1635052114725113, "learning_rate": 5.0282741831920454e-05, "loss": 0.0209, "step": 52330 }, { "grad_norm": 0.2658454179763794, "learning_rate": 5.026620738295351e-05, "loss": 0.0235, "step": 52340 }, { "grad_norm": 0.31222283840179443, "learning_rate": 5.024967290487455e-05, "loss": 0.0193, "step": 52350 }, { "grad_norm": 0.4080239236354828, "learning_rate": 5.023313839949172e-05, "loss": 0.0205, "step": 52360 }, { "grad_norm": 0.14220570027828217, "learning_rate": 5.0216603868613224e-05, "loss": 0.0176, "step": 52370 }, { "grad_norm": 0.1734410524368286, "learning_rate": 5.020006931404726e-05, "loss": 0.0198, "step": 52380 }, { "grad_norm": 0.3213833272457123, "learning_rate": 5.0183534737602014e-05, "loss": 0.0222, "step": 52390 }, { "grad_norm": 0.1691550761461258, "learning_rate": 5.0167000141085686e-05, "loss": 0.0193, "step": 52400 }, { "grad_norm": 0.1576557606458664, "learning_rate": 5.015046552630647e-05, "loss": 0.0204, "step": 52410 }, { "grad_norm": 0.20765405893325806, "learning_rate": 5.013393089507258e-05, "loss": 0.021, "step": 52420 }, { "grad_norm": 0.35577768087387085, "learning_rate": 5.011739624919219e-05, "loss": 0.02, "step": 52430 }, { "grad_norm": 0.18718355894088745, "learning_rate": 5.0100861590473545e-05, "loss": 0.0171, "step": 52440 }, { "grad_norm": 0.20271378755569458, "learning_rate": 5.008432692072482e-05, "loss": 0.0264, "step": 52450 }, { "grad_norm": 0.2047547847032547, "learning_rate": 5.006779224175423e-05, "loss": 0.0206, "step": 52460 }, { "grad_norm": 0.16140742599964142, "learning_rate": 5.005125755536998e-05, "loss": 0.0193, "step": 52470 }, { "grad_norm": 0.19270654022693634, "learning_rate": 5.003472286338029e-05, "loss": 0.0184, "step": 52480 }, { "grad_norm": 0.14530529081821442, "learning_rate": 5.001818816759335e-05, "loss": 0.016, "step": 52490 }, { "grad_norm": 0.20498965680599213, "learning_rate": 5.000165346981738e-05, "loss": 0.0207, "step": 52500 }, { "grad_norm": 0.3419220745563507, "learning_rate": 4.9985118771860595e-05, "loss": 0.0193, "step": 52510 }, { "grad_norm": 0.2642453908920288, "learning_rate": 4.996858407553119e-05, "loss": 0.0211, "step": 52520 }, { "grad_norm": 0.1955578625202179, "learning_rate": 4.99520493826374e-05, "loss": 0.0175, "step": 52530 }, { "grad_norm": 0.19693318009376526, "learning_rate": 4.9935514694987384e-05, "loss": 0.0228, "step": 52540 }, { "grad_norm": 0.174965038895607, "learning_rate": 4.991898001438941e-05, "loss": 0.0198, "step": 52550 }, { "grad_norm": 0.29144275188446045, "learning_rate": 4.990244534265166e-05, "loss": 0.0222, "step": 52560 }, { "grad_norm": 0.16309650242328644, "learning_rate": 4.988591068158234e-05, "loss": 0.0189, "step": 52570 }, { "grad_norm": 0.2024076133966446, "learning_rate": 4.986937603298965e-05, "loss": 0.0162, "step": 52580 }, { "grad_norm": 0.1908576786518097, "learning_rate": 4.985284139868181e-05, "loss": 0.0198, "step": 52590 }, { "grad_norm": 0.21391190588474274, "learning_rate": 4.9836306780467e-05, "loss": 0.0231, "step": 52600 }, { "grad_norm": 0.2579064965248108, "learning_rate": 4.981977218015343e-05, "loss": 0.018, "step": 52610 }, { "grad_norm": 0.21406061947345734, "learning_rate": 4.980323759954932e-05, "loss": 0.0173, "step": 52620 }, { "grad_norm": 0.21372359991073608, "learning_rate": 4.978670304046284e-05, "loss": 0.0205, "step": 52630 }, { "grad_norm": 0.2222493290901184, "learning_rate": 4.9770168504702195e-05, "loss": 0.02, "step": 52640 }, { "grad_norm": 0.20045435428619385, "learning_rate": 4.975363399407559e-05, "loss": 0.0216, "step": 52650 }, { "grad_norm": 0.2187603861093521, "learning_rate": 4.9737099510391194e-05, "loss": 0.0201, "step": 52660 }, { "grad_norm": 0.2058306336402893, "learning_rate": 4.972056505545719e-05, "loss": 0.0184, "step": 52670 }, { "grad_norm": 0.18374301493167877, "learning_rate": 4.9704030631081775e-05, "loss": 0.0187, "step": 52680 }, { "grad_norm": 0.20717090368270874, "learning_rate": 4.968749623907313e-05, "loss": 0.0237, "step": 52690 }, { "grad_norm": 0.14357315003871918, "learning_rate": 4.967096188123943e-05, "loss": 0.0207, "step": 52700 }, { "grad_norm": 0.19882139563560486, "learning_rate": 4.965442755938884e-05, "loss": 0.0202, "step": 52710 }, { "grad_norm": 0.3442836403846741, "learning_rate": 4.963789327532954e-05, "loss": 0.0241, "step": 52720 }, { "grad_norm": 0.21412226557731628, "learning_rate": 4.962135903086968e-05, "loss": 0.0188, "step": 52730 }, { "grad_norm": 0.18278610706329346, "learning_rate": 4.960482482781741e-05, "loss": 0.0171, "step": 52740 }, { "grad_norm": 0.20082536339759827, "learning_rate": 4.958829066798092e-05, "loss": 0.0191, "step": 52750 }, { "grad_norm": 0.19819515943527222, "learning_rate": 4.957175655316835e-05, "loss": 0.0217, "step": 52760 }, { "grad_norm": 0.18910105526447296, "learning_rate": 4.9555222485187834e-05, "loss": 0.0182, "step": 52770 }, { "grad_norm": 0.22858764231204987, "learning_rate": 4.953868846584752e-05, "loss": 0.0213, "step": 52780 }, { "grad_norm": 0.2148764729499817, "learning_rate": 4.9522154496955544e-05, "loss": 0.0206, "step": 52790 }, { "grad_norm": 0.2019188404083252, "learning_rate": 4.950562058032001e-05, "loss": 0.0202, "step": 52800 }, { "grad_norm": 0.19323910772800446, "learning_rate": 4.948908671774909e-05, "loss": 0.019, "step": 52810 }, { "grad_norm": 0.18071205914020538, "learning_rate": 4.947255291105087e-05, "loss": 0.0196, "step": 52820 }, { "grad_norm": 0.23061734437942505, "learning_rate": 4.945601916203348e-05, "loss": 0.0196, "step": 52830 }, { "grad_norm": 0.24867215752601624, "learning_rate": 4.9439485472505016e-05, "loss": 0.0199, "step": 52840 }, { "grad_norm": 0.24206508696079254, "learning_rate": 4.942295184427356e-05, "loss": 0.0185, "step": 52850 }, { "grad_norm": 0.18040898442268372, "learning_rate": 4.9406418279147234e-05, "loss": 0.0157, "step": 52860 }, { "grad_norm": 0.3000507354736328, "learning_rate": 4.9389884778934116e-05, "loss": 0.0204, "step": 52870 }, { "grad_norm": 0.20308728516101837, "learning_rate": 4.937335134544228e-05, "loss": 0.0205, "step": 52880 }, { "grad_norm": 0.2645111680030823, "learning_rate": 4.935681798047979e-05, "loss": 0.0187, "step": 52890 }, { "grad_norm": 0.17036172747612, "learning_rate": 4.934028468585473e-05, "loss": 0.0247, "step": 52900 }, { "grad_norm": 0.23535622656345367, "learning_rate": 4.932375146337514e-05, "loss": 0.0216, "step": 52910 }, { "grad_norm": 0.23894520103931427, "learning_rate": 4.930721831484908e-05, "loss": 0.0184, "step": 52920 }, { "grad_norm": 0.17478810250759125, "learning_rate": 4.929068524208456e-05, "loss": 0.0222, "step": 52930 }, { "grad_norm": 0.2977915108203888, "learning_rate": 4.927415224688965e-05, "loss": 0.0184, "step": 52940 }, { "grad_norm": 0.18034304678440094, "learning_rate": 4.9257619331072375e-05, "loss": 0.0194, "step": 52950 }, { "grad_norm": 0.2144457995891571, "learning_rate": 4.924108649644072e-05, "loss": 0.0209, "step": 52960 }, { "grad_norm": 0.29564768075942993, "learning_rate": 4.922455374480271e-05, "loss": 0.023, "step": 52970 }, { "grad_norm": 0.4359310269355774, "learning_rate": 4.920802107796634e-05, "loss": 0.0222, "step": 52980 }, { "grad_norm": 0.1755874902009964, "learning_rate": 4.9191488497739564e-05, "loss": 0.0191, "step": 52990 }, { "grad_norm": 0.17002147436141968, "learning_rate": 4.917495600593042e-05, "loss": 0.0187, "step": 53000 }, { "grad_norm": 0.16430255770683289, "learning_rate": 4.915842360434684e-05, "loss": 0.018, "step": 53010 }, { "grad_norm": 0.22759653627872467, "learning_rate": 4.914189129479679e-05, "loss": 0.0194, "step": 53020 }, { "grad_norm": 0.29697710275650024, "learning_rate": 4.912535907908822e-05, "loss": 0.0193, "step": 53030 }, { "grad_norm": 0.21334245800971985, "learning_rate": 4.910882695902907e-05, "loss": 0.0208, "step": 53040 }, { "grad_norm": 0.2215305119752884, "learning_rate": 4.9092294936427234e-05, "loss": 0.0193, "step": 53050 }, { "grad_norm": 0.19689910113811493, "learning_rate": 4.9075763013090685e-05, "loss": 0.0181, "step": 53060 }, { "grad_norm": 0.17538678646087646, "learning_rate": 4.9059231190827305e-05, "loss": 0.0176, "step": 53070 }, { "grad_norm": 0.19501790404319763, "learning_rate": 4.904269947144499e-05, "loss": 0.023, "step": 53080 }, { "grad_norm": 0.31341078877449036, "learning_rate": 4.902616785675161e-05, "loss": 0.0239, "step": 53090 }, { "grad_norm": 0.28017550706863403, "learning_rate": 4.900963634855505e-05, "loss": 0.0238, "step": 53100 }, { "grad_norm": 0.2474769502878189, "learning_rate": 4.8993104948663173e-05, "loss": 0.0202, "step": 53110 }, { "grad_norm": 0.21757060289382935, "learning_rate": 4.8976573658883823e-05, "loss": 0.0215, "step": 53120 }, { "grad_norm": 0.27713295817375183, "learning_rate": 4.896004248102484e-05, "loss": 0.0211, "step": 53130 }, { "grad_norm": 0.2369227409362793, "learning_rate": 4.8943511416894045e-05, "loss": 0.0234, "step": 53140 }, { "grad_norm": 0.25081247091293335, "learning_rate": 4.892698046829926e-05, "loss": 0.0217, "step": 53150 }, { "grad_norm": 0.18665392696857452, "learning_rate": 4.8910449637048275e-05, "loss": 0.0179, "step": 53160 }, { "grad_norm": 0.32828637957572937, "learning_rate": 4.8893918924948884e-05, "loss": 0.0221, "step": 53170 }, { "grad_norm": 0.15221139788627625, "learning_rate": 4.8877388333808835e-05, "loss": 0.0162, "step": 53180 }, { "grad_norm": 0.19201324880123138, "learning_rate": 4.8860857865435925e-05, "loss": 0.0165, "step": 53190 }, { "grad_norm": 0.1762019693851471, "learning_rate": 4.884432752163789e-05, "loss": 0.0222, "step": 53200 }, { "grad_norm": 0.21946817636489868, "learning_rate": 4.8827797304222466e-05, "loss": 0.0232, "step": 53210 }, { "grad_norm": 0.24602442979812622, "learning_rate": 4.8811267214997366e-05, "loss": 0.0201, "step": 53220 }, { "grad_norm": 0.12886689603328705, "learning_rate": 4.879473725577029e-05, "loss": 0.026, "step": 53230 }, { "grad_norm": 0.21702280640602112, "learning_rate": 4.877820742834891e-05, "loss": 0.0169, "step": 53240 }, { "grad_norm": 0.20177723467350006, "learning_rate": 4.8761677734540956e-05, "loss": 0.0201, "step": 53250 }, { "grad_norm": 0.2221832424402237, "learning_rate": 4.874514817615407e-05, "loss": 0.0211, "step": 53260 }, { "grad_norm": 0.29480382800102234, "learning_rate": 4.872861875499588e-05, "loss": 0.0189, "step": 53270 }, { "grad_norm": 0.19862999022006989, "learning_rate": 4.871208947287404e-05, "loss": 0.0196, "step": 53280 }, { "grad_norm": 0.4379562735557556, "learning_rate": 4.869556033159615e-05, "loss": 0.0255, "step": 53290 }, { "grad_norm": 0.1932639479637146, "learning_rate": 4.8679031332969835e-05, "loss": 0.0224, "step": 53300 }, { "grad_norm": 0.2745286822319031, "learning_rate": 4.8662502478802635e-05, "loss": 0.0214, "step": 53310 }, { "grad_norm": 0.19131723046302795, "learning_rate": 4.8645973770902176e-05, "loss": 0.0189, "step": 53320 }, { "grad_norm": 0.2502165734767914, "learning_rate": 4.862944521107599e-05, "loss": 0.0214, "step": 53330 }, { "grad_norm": 0.14258168637752533, "learning_rate": 4.861291680113162e-05, "loss": 0.0153, "step": 53340 }, { "grad_norm": 0.1873372346162796, "learning_rate": 4.8596388542876563e-05, "loss": 0.0183, "step": 53350 }, { "grad_norm": 0.2917800843715668, "learning_rate": 4.857986043811836e-05, "loss": 0.0207, "step": 53360 }, { "grad_norm": 0.4100010097026825, "learning_rate": 4.856333248866448e-05, "loss": 0.0225, "step": 53370 }, { "grad_norm": 0.23769617080688477, "learning_rate": 4.85468046963224e-05, "loss": 0.0226, "step": 53380 }, { "grad_norm": 0.20803095400333405, "learning_rate": 4.853027706289957e-05, "loss": 0.0162, "step": 53390 }, { "grad_norm": 0.23175203800201416, "learning_rate": 4.851374959020344e-05, "loss": 0.0217, "step": 53400 }, { "grad_norm": 0.1701246201992035, "learning_rate": 4.849722228004141e-05, "loss": 0.021, "step": 53410 }, { "grad_norm": 0.24367941915988922, "learning_rate": 4.848069513422091e-05, "loss": 0.0188, "step": 53420 }, { "grad_norm": 0.2171648144721985, "learning_rate": 4.8464168154549294e-05, "loss": 0.0214, "step": 53430 }, { "grad_norm": 0.22215937077999115, "learning_rate": 4.844764134283392e-05, "loss": 0.0157, "step": 53440 }, { "grad_norm": 0.22361603379249573, "learning_rate": 4.843111470088218e-05, "loss": 0.0228, "step": 53450 }, { "grad_norm": 0.18777793645858765, "learning_rate": 4.841458823050137e-05, "loss": 0.0178, "step": 53460 }, { "grad_norm": 0.179940328001976, "learning_rate": 4.8398061933498816e-05, "loss": 0.019, "step": 53470 }, { "grad_norm": 0.1837550699710846, "learning_rate": 4.83815358116818e-05, "loss": 0.0169, "step": 53480 }, { "grad_norm": 0.25791624188423157, "learning_rate": 4.8365009866857584e-05, "loss": 0.0198, "step": 53490 }, { "grad_norm": 0.2770736813545227, "learning_rate": 4.834848410083342e-05, "loss": 0.0184, "step": 53500 }, { "grad_norm": 0.18038125336170197, "learning_rate": 4.833195851541657e-05, "loss": 0.0186, "step": 53510 }, { "grad_norm": 0.18320080637931824, "learning_rate": 4.831543311241423e-05, "loss": 0.0178, "step": 53520 }, { "grad_norm": 0.14043770730495453, "learning_rate": 4.829890789363359e-05, "loss": 0.0199, "step": 53530 }, { "grad_norm": 0.18966975808143616, "learning_rate": 4.828238286088182e-05, "loss": 0.0213, "step": 53540 }, { "grad_norm": 0.24509353935718536, "learning_rate": 4.826585801596609e-05, "loss": 0.0247, "step": 53550 }, { "grad_norm": 0.21213063597679138, "learning_rate": 4.8249333360693486e-05, "loss": 0.02, "step": 53560 }, { "grad_norm": 0.159761443734169, "learning_rate": 4.8232808896871185e-05, "loss": 0.0171, "step": 53570 }, { "grad_norm": 0.19145101308822632, "learning_rate": 4.821628462630624e-05, "loss": 0.0192, "step": 53580 }, { "grad_norm": 0.21861329674720764, "learning_rate": 4.8199760550805735e-05, "loss": 0.0228, "step": 53590 }, { "grad_norm": 0.23504237830638885, "learning_rate": 4.818323667217669e-05, "loss": 0.0205, "step": 53600 }, { "grad_norm": 0.2059590220451355, "learning_rate": 4.8166712992226165e-05, "loss": 0.0189, "step": 53610 }, { "grad_norm": 0.15899336338043213, "learning_rate": 4.8150189512761154e-05, "loss": 0.0179, "step": 53620 }, { "grad_norm": 0.21359162032604218, "learning_rate": 4.813366623558862e-05, "loss": 0.0202, "step": 53630 }, { "grad_norm": 0.22369854152202606, "learning_rate": 4.811714316251554e-05, "loss": 0.0173, "step": 53640 }, { "grad_norm": 0.7204717397689819, "learning_rate": 4.810062029534886e-05, "loss": 0.024, "step": 53650 }, { "grad_norm": 0.2527281939983368, "learning_rate": 4.8084097635895505e-05, "loss": 0.0218, "step": 53660 }, { "grad_norm": 0.2746232748031616, "learning_rate": 4.8067575185962335e-05, "loss": 0.0187, "step": 53670 }, { "grad_norm": 0.15695463120937347, "learning_rate": 4.805105294735625e-05, "loss": 0.0192, "step": 53680 }, { "grad_norm": 0.1827915459871292, "learning_rate": 4.803453092188406e-05, "loss": 0.0219, "step": 53690 }, { "grad_norm": 0.16529437899589539, "learning_rate": 4.801800911135263e-05, "loss": 0.0199, "step": 53700 }, { "grad_norm": 0.19872494041919708, "learning_rate": 4.8001487517568744e-05, "loss": 0.0201, "step": 53710 }, { "grad_norm": 0.24935942888259888, "learning_rate": 4.798496614233918e-05, "loss": 0.017, "step": 53720 }, { "grad_norm": 0.1757669448852539, "learning_rate": 4.796844498747069e-05, "loss": 0.023, "step": 53730 }, { "grad_norm": 0.2720032334327698, "learning_rate": 4.795192405477e-05, "loss": 0.0168, "step": 53740 }, { "grad_norm": 0.3118228614330292, "learning_rate": 4.79354033460438e-05, "loss": 0.0194, "step": 53750 }, { "grad_norm": 0.24822720885276794, "learning_rate": 4.79188828630988e-05, "loss": 0.0176, "step": 53760 }, { "grad_norm": 0.3011195659637451, "learning_rate": 4.790236260774165e-05, "loss": 0.0171, "step": 53770 }, { "grad_norm": 0.15683066844940186, "learning_rate": 4.788584258177896e-05, "loss": 0.0215, "step": 53780 }, { "grad_norm": 0.21964740753173828, "learning_rate": 4.7869322787017356e-05, "loss": 0.0168, "step": 53790 }, { "grad_norm": 0.17938587069511414, "learning_rate": 4.78528032252634e-05, "loss": 0.0198, "step": 53800 }, { "grad_norm": 0.263447105884552, "learning_rate": 4.783628389832366e-05, "loss": 0.0188, "step": 53810 }, { "grad_norm": 0.19718514382839203, "learning_rate": 4.7819764808004644e-05, "loss": 0.0214, "step": 53820 }, { "grad_norm": 0.20828984677791595, "learning_rate": 4.780324595611289e-05, "loss": 0.0215, "step": 53830 }, { "grad_norm": 0.2981850206851959, "learning_rate": 4.7786727344454854e-05, "loss": 0.0204, "step": 53840 }, { "grad_norm": 0.29123327136039734, "learning_rate": 4.777020897483697e-05, "loss": 0.0198, "step": 53850 }, { "grad_norm": 0.22382852435112, "learning_rate": 4.775369084906569e-05, "loss": 0.0183, "step": 53860 }, { "grad_norm": 0.20020796358585358, "learning_rate": 4.773717296894742e-05, "loss": 0.0202, "step": 53870 }, { "grad_norm": 0.16028372943401337, "learning_rate": 4.772065533628848e-05, "loss": 0.0235, "step": 53880 }, { "grad_norm": 0.27721524238586426, "learning_rate": 4.7704137952895265e-05, "loss": 0.0203, "step": 53890 }, { "grad_norm": 0.24251538515090942, "learning_rate": 4.768762082057407e-05, "loss": 0.0218, "step": 53900 }, { "grad_norm": 0.15628138184547424, "learning_rate": 4.7671103941131186e-05, "loss": 0.0193, "step": 53910 }, { "grad_norm": 0.24908770620822906, "learning_rate": 4.765458731637287e-05, "loss": 0.0235, "step": 53920 }, { "grad_norm": 0.24856126308441162, "learning_rate": 4.763807094810537e-05, "loss": 0.0176, "step": 53930 }, { "grad_norm": 0.1656506061553955, "learning_rate": 4.762155483813487e-05, "loss": 0.018, "step": 53940 }, { "grad_norm": 0.196180522441864, "learning_rate": 4.760503898826754e-05, "loss": 0.0192, "step": 53950 }, { "grad_norm": 0.2475215345621109, "learning_rate": 4.758852340030957e-05, "loss": 0.0165, "step": 53960 }, { "grad_norm": 0.2229008823633194, "learning_rate": 4.757200807606705e-05, "loss": 0.0212, "step": 53970 }, { "grad_norm": 0.24608923494815826, "learning_rate": 4.755549301734607e-05, "loss": 0.0207, "step": 53980 }, { "grad_norm": 0.2020183503627777, "learning_rate": 4.753897822595271e-05, "loss": 0.0198, "step": 53990 }, { "grad_norm": 0.21872851252555847, "learning_rate": 4.7522463703692994e-05, "loss": 0.0183, "step": 54000 }, { "grad_norm": 0.21849730610847473, "learning_rate": 4.75059494523729e-05, "loss": 0.0184, "step": 54010 }, { "grad_norm": 0.2410813271999359, "learning_rate": 4.748943547379843e-05, "loss": 0.0218, "step": 54020 }, { "grad_norm": 0.20474955439567566, "learning_rate": 4.7472921769775535e-05, "loss": 0.0193, "step": 54030 }, { "grad_norm": 0.5105623006820679, "learning_rate": 4.745640834211011e-05, "loss": 0.0214, "step": 54040 }, { "grad_norm": 0.17889779806137085, "learning_rate": 4.7439895192608054e-05, "loss": 0.0221, "step": 54050 }, { "grad_norm": 0.19946232438087463, "learning_rate": 4.7423382323075186e-05, "loss": 0.0178, "step": 54060 }, { "grad_norm": 0.2588002383708954, "learning_rate": 4.740686973531737e-05, "loss": 0.0204, "step": 54070 }, { "grad_norm": 0.32150930166244507, "learning_rate": 4.7390357431140375e-05, "loss": 0.0202, "step": 54080 }, { "grad_norm": 0.3828347325325012, "learning_rate": 4.737384541234998e-05, "loss": 0.017, "step": 54090 }, { "grad_norm": 0.21496988832950592, "learning_rate": 4.735733368075189e-05, "loss": 0.0264, "step": 54100 }, { "grad_norm": 0.2609497904777527, "learning_rate": 4.734082223815182e-05, "loss": 0.0198, "step": 54110 }, { "grad_norm": 0.20275944471359253, "learning_rate": 4.7324311086355436e-05, "loss": 0.024, "step": 54120 }, { "grad_norm": 0.22187621891498566, "learning_rate": 4.730780022716836e-05, "loss": 0.0194, "step": 54130 }, { "grad_norm": 0.2108193039894104, "learning_rate": 4.729128966239619e-05, "loss": 0.0163, "step": 54140 }, { "grad_norm": 0.24383342266082764, "learning_rate": 4.7274779393844536e-05, "loss": 0.0212, "step": 54150 }, { "grad_norm": 0.16857025027275085, "learning_rate": 4.725826942331891e-05, "loss": 0.0201, "step": 54160 }, { "grad_norm": 0.2965368330478668, "learning_rate": 4.724175975262481e-05, "loss": 0.0224, "step": 54170 }, { "grad_norm": 0.2375965118408203, "learning_rate": 4.722525038356772e-05, "loss": 0.0195, "step": 54180 }, { "grad_norm": 0.2133924514055252, "learning_rate": 4.7208741317953075e-05, "loss": 0.0235, "step": 54190 }, { "grad_norm": 0.20852898061275482, "learning_rate": 4.7192232557586264e-05, "loss": 0.0215, "step": 54200 }, { "grad_norm": 0.28389841318130493, "learning_rate": 4.717572410427269e-05, "loss": 0.0192, "step": 54210 }, { "grad_norm": 0.25147104263305664, "learning_rate": 4.71592159598177e-05, "loss": 0.0243, "step": 54220 }, { "grad_norm": 0.1756105273962021, "learning_rate": 4.714270812602657e-05, "loss": 0.0217, "step": 54230 }, { "grad_norm": 0.2201753854751587, "learning_rate": 4.712620060470458e-05, "loss": 0.0206, "step": 54240 }, { "grad_norm": 0.26548925042152405, "learning_rate": 4.7109693397656986e-05, "loss": 0.0233, "step": 54250 }, { "grad_norm": 0.23364363610744476, "learning_rate": 4.709318650668894e-05, "loss": 0.0229, "step": 54260 }, { "grad_norm": 0.16417613625526428, "learning_rate": 4.7076679933605676e-05, "loss": 0.0188, "step": 54270 }, { "grad_norm": 0.2041565179824829, "learning_rate": 4.70601736802123e-05, "loss": 0.019, "step": 54280 }, { "grad_norm": 0.16988301277160645, "learning_rate": 4.704366774831391e-05, "loss": 0.0205, "step": 54290 }, { "grad_norm": 0.1395159214735031, "learning_rate": 4.702716213971557e-05, "loss": 0.0153, "step": 54300 }, { "grad_norm": 0.23684561252593994, "learning_rate": 4.70106568562223e-05, "loss": 0.0237, "step": 54310 }, { "grad_norm": 0.24942943453788757, "learning_rate": 4.6994151899639105e-05, "loss": 0.0218, "step": 54320 }, { "grad_norm": 0.19805440306663513, "learning_rate": 4.697764727177093e-05, "loss": 0.0227, "step": 54330 }, { "grad_norm": 0.21562358736991882, "learning_rate": 4.696114297442272e-05, "loss": 0.0238, "step": 54340 }, { "grad_norm": 0.1720276176929474, "learning_rate": 4.694463900939933e-05, "loss": 0.0216, "step": 54350 }, { "grad_norm": 0.25854989886283875, "learning_rate": 4.692813537850564e-05, "loss": 0.02, "step": 54360 }, { "grad_norm": 0.1975284069776535, "learning_rate": 4.6911632083546445e-05, "loss": 0.019, "step": 54370 }, { "grad_norm": 0.16708137094974518, "learning_rate": 4.6895129126326525e-05, "loss": 0.0209, "step": 54380 }, { "grad_norm": 0.3223329484462738, "learning_rate": 4.687862650865059e-05, "loss": 0.0201, "step": 54390 }, { "grad_norm": 0.23544254899024963, "learning_rate": 4.686212423232339e-05, "loss": 0.0176, "step": 54400 }, { "grad_norm": 0.18063412606716156, "learning_rate": 4.6845622299149565e-05, "loss": 0.0201, "step": 54410 }, { "grad_norm": 0.1693425476551056, "learning_rate": 4.682912071093374e-05, "loss": 0.016, "step": 54420 }, { "grad_norm": 0.1765407919883728, "learning_rate": 4.681261946948052e-05, "loss": 0.013, "step": 54430 }, { "grad_norm": 0.286941796541214, "learning_rate": 4.679611857659443e-05, "loss": 0.0188, "step": 54440 }, { "grad_norm": 0.19294473528862, "learning_rate": 4.6779618034079975e-05, "loss": 0.017, "step": 54450 }, { "grad_norm": 0.2234533429145813, "learning_rate": 4.6763117843741675e-05, "loss": 0.0194, "step": 54460 }, { "grad_norm": 0.1798720806837082, "learning_rate": 4.674661800738394e-05, "loss": 0.0181, "step": 54470 }, { "grad_norm": 0.1935950368642807, "learning_rate": 4.673011852681117e-05, "loss": 0.0168, "step": 54480 }, { "grad_norm": 0.2053278386592865, "learning_rate": 4.671361940382772e-05, "loss": 0.019, "step": 54490 }, { "grad_norm": 0.15729928016662598, "learning_rate": 4.669712064023791e-05, "loss": 0.0208, "step": 54500 }, { "grad_norm": 0.20623132586479187, "learning_rate": 4.668062223784602e-05, "loss": 0.0171, "step": 54510 }, { "grad_norm": 0.1849607676267624, "learning_rate": 4.6664124198456274e-05, "loss": 0.0178, "step": 54520 }, { "grad_norm": 0.1603979766368866, "learning_rate": 4.6647626523872905e-05, "loss": 0.0204, "step": 54530 }, { "grad_norm": 0.22810885310173035, "learning_rate": 4.6631129215900066e-05, "loss": 0.0176, "step": 54540 }, { "grad_norm": 0.1717454046010971, "learning_rate": 4.661463227634187e-05, "loss": 0.0213, "step": 54550 }, { "grad_norm": 0.1894560605287552, "learning_rate": 4.659813570700238e-05, "loss": 0.017, "step": 54560 }, { "grad_norm": 0.17375138401985168, "learning_rate": 4.658163950968568e-05, "loss": 0.0179, "step": 54570 }, { "grad_norm": 0.15526561439037323, "learning_rate": 4.6565143686195726e-05, "loss": 0.0229, "step": 54580 }, { "grad_norm": 0.18304869532585144, "learning_rate": 4.654864823833649e-05, "loss": 0.0176, "step": 54590 }, { "grad_norm": 0.15247374773025513, "learning_rate": 4.65321531679119e-05, "loss": 0.0201, "step": 54600 }, { "grad_norm": 0.2116680145263672, "learning_rate": 4.6515658476725834e-05, "loss": 0.0195, "step": 54610 }, { "grad_norm": 0.3479068875312805, "learning_rate": 4.649916416658212e-05, "loss": 0.0176, "step": 54620 }, { "grad_norm": 0.1388036459684372, "learning_rate": 4.648267023928455e-05, "loss": 0.0173, "step": 54630 }, { "grad_norm": 0.20874200761318207, "learning_rate": 4.646617669663687e-05, "loss": 0.0218, "step": 54640 }, { "grad_norm": 0.2129504531621933, "learning_rate": 4.644968354044277e-05, "loss": 0.0191, "step": 54650 }, { "grad_norm": 0.2386414110660553, "learning_rate": 4.643319077250596e-05, "loss": 0.017, "step": 54660 }, { "grad_norm": 0.24216872453689575, "learning_rate": 4.6416698394630055e-05, "loss": 0.0181, "step": 54670 }, { "grad_norm": 0.17372986674308777, "learning_rate": 4.640020640861862e-05, "loss": 0.018, "step": 54680 }, { "grad_norm": 0.1595894694328308, "learning_rate": 4.6383714816275195e-05, "loss": 0.0173, "step": 54690 }, { "grad_norm": 0.17676326632499695, "learning_rate": 4.636722361940329e-05, "loss": 0.0165, "step": 54700 }, { "grad_norm": 0.1978832483291626, "learning_rate": 4.635073281980632e-05, "loss": 0.0195, "step": 54710 }, { "grad_norm": 0.20045112073421478, "learning_rate": 4.633424241928775e-05, "loss": 0.0172, "step": 54720 }, { "grad_norm": 0.19072642922401428, "learning_rate": 4.6317752419650914e-05, "loss": 0.0193, "step": 54730 }, { "grad_norm": 0.18524178862571716, "learning_rate": 4.630126282269914e-05, "loss": 0.0196, "step": 54740 }, { "grad_norm": 0.2521094083786011, "learning_rate": 4.628477363023569e-05, "loss": 0.0189, "step": 54750 }, { "grad_norm": 0.2359103262424469, "learning_rate": 4.626828484406383e-05, "loss": 0.0187, "step": 54760 }, { "grad_norm": 0.25689610838890076, "learning_rate": 4.625179646598669e-05, "loss": 0.0185, "step": 54770 }, { "grad_norm": 0.30565235018730164, "learning_rate": 4.6235308497807487e-05, "loss": 0.0191, "step": 54780 }, { "grad_norm": 0.1828872263431549, "learning_rate": 4.6218820941329276e-05, "loss": 0.0161, "step": 54790 }, { "grad_norm": 0.15559104084968567, "learning_rate": 4.620233379835513e-05, "loss": 0.0155, "step": 54800 }, { "grad_norm": 0.24004918336868286, "learning_rate": 4.618584707068804e-05, "loss": 0.017, "step": 54810 }, { "grad_norm": 0.22522346675395966, "learning_rate": 4.616936076013099e-05, "loss": 0.0203, "step": 54820 }, { "grad_norm": 0.22183959186077118, "learning_rate": 4.615287486848688e-05, "loss": 0.0162, "step": 54830 }, { "grad_norm": 0.1691964715719223, "learning_rate": 4.6136389397558584e-05, "loss": 0.0188, "step": 54840 }, { "grad_norm": 0.24304550886154175, "learning_rate": 4.6119904349148935e-05, "loss": 0.0189, "step": 54850 }, { "grad_norm": 0.2131689041852951, "learning_rate": 4.6103419725060726e-05, "loss": 0.0223, "step": 54860 }, { "grad_norm": 0.22656363248825073, "learning_rate": 4.608693552709667e-05, "loss": 0.0209, "step": 54870 }, { "grad_norm": 0.16573412716388702, "learning_rate": 4.607045175705947e-05, "loss": 0.0187, "step": 54880 }, { "grad_norm": 0.18892893195152283, "learning_rate": 4.605396841675176e-05, "loss": 0.0197, "step": 54890 }, { "grad_norm": 0.21989133954048157, "learning_rate": 4.6037485507976094e-05, "loss": 0.0211, "step": 54900 }, { "grad_norm": 0.7133936285972595, "learning_rate": 4.602100303253509e-05, "loss": 0.0187, "step": 54910 }, { "grad_norm": 0.187444806098938, "learning_rate": 4.600452099223122e-05, "loss": 0.0187, "step": 54920 }, { "grad_norm": 0.28709229826927185, "learning_rate": 4.598803938886691e-05, "loss": 0.0225, "step": 54930 }, { "grad_norm": 0.20698386430740356, "learning_rate": 4.597155822424459e-05, "loss": 0.0183, "step": 54940 }, { "grad_norm": 0.16990351676940918, "learning_rate": 4.59550775001666e-05, "loss": 0.0174, "step": 54950 }, { "grad_norm": 0.2002558708190918, "learning_rate": 4.5938597218435234e-05, "loss": 0.0187, "step": 54960 }, { "grad_norm": 0.2367538958787918, "learning_rate": 4.592211738085279e-05, "loss": 0.0212, "step": 54970 }, { "grad_norm": 0.34883731603622437, "learning_rate": 4.590563798922146e-05, "loss": 0.019, "step": 54980 }, { "grad_norm": 0.2170795202255249, "learning_rate": 4.5889159045343404e-05, "loss": 0.0194, "step": 54990 }, { "grad_norm": 0.2007988691329956, "learning_rate": 4.5872680551020734e-05, "loss": 0.0178, "step": 55000 }, { "grad_norm": 0.20100487768650055, "learning_rate": 4.585620250805552e-05, "loss": 0.0187, "step": 55010 }, { "grad_norm": 0.19549070298671722, "learning_rate": 4.583972491824975e-05, "loss": 0.0175, "step": 55020 }, { "grad_norm": 0.21917946636676788, "learning_rate": 4.582324778340541e-05, "loss": 0.0194, "step": 55030 }, { "grad_norm": 0.5620877742767334, "learning_rate": 4.5806771105324424e-05, "loss": 0.0199, "step": 55040 }, { "grad_norm": 0.22659768164157867, "learning_rate": 4.579029488580865e-05, "loss": 0.0144, "step": 55050 }, { "grad_norm": 0.3219330310821533, "learning_rate": 4.5773819126659876e-05, "loss": 0.017, "step": 55060 }, { "grad_norm": 0.22472961246967316, "learning_rate": 4.575734382967991e-05, "loss": 0.0179, "step": 55070 }, { "grad_norm": 0.2595497667789459, "learning_rate": 4.574086899667044e-05, "loss": 0.0182, "step": 55080 }, { "grad_norm": 0.19913573563098907, "learning_rate": 4.5724394629433106e-05, "loss": 0.0197, "step": 55090 }, { "grad_norm": 0.20850101113319397, "learning_rate": 4.570792072976958e-05, "loss": 0.0234, "step": 55100 }, { "grad_norm": 0.20480483770370483, "learning_rate": 4.569144729948138e-05, "loss": 0.0153, "step": 55110 }, { "grad_norm": 0.20399783551692963, "learning_rate": 4.5674974340370033e-05, "loss": 0.0177, "step": 55120 }, { "grad_norm": 0.22316280007362366, "learning_rate": 4.565850185423699e-05, "loss": 0.021, "step": 55130 }, { "grad_norm": 0.3160359561443329, "learning_rate": 4.564202984288365e-05, "loss": 0.0262, "step": 55140 }, { "grad_norm": 0.18811795115470886, "learning_rate": 4.5625558308111356e-05, "loss": 0.0185, "step": 55150 }, { "grad_norm": 0.1958160251379013, "learning_rate": 4.560908725172145e-05, "loss": 0.021, "step": 55160 }, { "grad_norm": 0.20118825137615204, "learning_rate": 4.559261667551516e-05, "loss": 0.0207, "step": 55170 }, { "grad_norm": 0.16360321640968323, "learning_rate": 4.5576146581293685e-05, "loss": 0.0211, "step": 55180 }, { "grad_norm": 0.22790183126926422, "learning_rate": 4.5559676970858164e-05, "loss": 0.0198, "step": 55190 }, { "grad_norm": 0.14985783398151398, "learning_rate": 4.5543207846009705e-05, "loss": 0.0176, "step": 55200 }, { "grad_norm": 0.17793124914169312, "learning_rate": 4.552673920854933e-05, "loss": 0.021, "step": 55210 }, { "grad_norm": 0.2142205685377121, "learning_rate": 4.5510271060278006e-05, "loss": 0.019, "step": 55220 }, { "grad_norm": 0.17883938550949097, "learning_rate": 4.5493803402996704e-05, "loss": 0.0219, "step": 55230 }, { "grad_norm": 0.18729129433631897, "learning_rate": 4.5477336238506295e-05, "loss": 0.0231, "step": 55240 }, { "grad_norm": 0.18173439800739288, "learning_rate": 4.546086956860759e-05, "loss": 0.0206, "step": 55250 }, { "grad_norm": 0.23880153894424438, "learning_rate": 4.5444403395101374e-05, "loss": 0.0201, "step": 55260 }, { "grad_norm": 0.19059982895851135, "learning_rate": 4.542793771978834e-05, "loss": 0.0178, "step": 55270 }, { "grad_norm": 0.17312975227832794, "learning_rate": 4.541147254446916e-05, "loss": 0.0181, "step": 55280 }, { "grad_norm": 0.19976267218589783, "learning_rate": 4.539500787094445e-05, "loss": 0.0192, "step": 55290 }, { "grad_norm": 0.2142593264579773, "learning_rate": 4.537854370101475e-05, "loss": 0.0188, "step": 55300 }, { "grad_norm": 0.2506832480430603, "learning_rate": 4.5362080036480566e-05, "loss": 0.0169, "step": 55310 }, { "grad_norm": 0.16242676973342896, "learning_rate": 4.534561687914235e-05, "loss": 0.0188, "step": 55320 }, { "grad_norm": 0.20586974918842316, "learning_rate": 4.5329154230800466e-05, "loss": 0.0195, "step": 55330 }, { "grad_norm": 0.25790244340896606, "learning_rate": 4.531269209325525e-05, "loss": 0.0163, "step": 55340 }, { "grad_norm": 0.18919256329536438, "learning_rate": 4.529623046830696e-05, "loss": 0.0188, "step": 55350 }, { "grad_norm": 0.20371688902378082, "learning_rate": 4.5279769357755866e-05, "loss": 0.016, "step": 55360 }, { "grad_norm": 0.23900602757930756, "learning_rate": 4.5263308763402084e-05, "loss": 0.0179, "step": 55370 }, { "grad_norm": 0.21991483867168427, "learning_rate": 4.524684868704574e-05, "loss": 0.0162, "step": 55380 }, { "grad_norm": 0.21018391847610474, "learning_rate": 4.523038913048688e-05, "loss": 0.0171, "step": 55390 }, { "grad_norm": 0.31985509395599365, "learning_rate": 4.521393009552549e-05, "loss": 0.019, "step": 55400 }, { "grad_norm": 0.1499607264995575, "learning_rate": 4.5197471583961484e-05, "loss": 0.0174, "step": 55410 }, { "grad_norm": 0.31051990389823914, "learning_rate": 4.51810135975948e-05, "loss": 0.0186, "step": 55420 }, { "grad_norm": 0.13708311319351196, "learning_rate": 4.516455613822521e-05, "loss": 0.0161, "step": 55430 }, { "grad_norm": 0.20505861937999725, "learning_rate": 4.514809920765249e-05, "loss": 0.0189, "step": 55440 }, { "grad_norm": 0.21018297970294952, "learning_rate": 4.5131642807676356e-05, "loss": 0.0164, "step": 55450 }, { "grad_norm": 0.1804841160774231, "learning_rate": 4.511518694009644e-05, "loss": 0.0188, "step": 55460 }, { "grad_norm": 0.1960485726594925, "learning_rate": 4.509873160671231e-05, "loss": 0.0188, "step": 55470 }, { "grad_norm": 0.40965163707733154, "learning_rate": 4.508227680932356e-05, "loss": 0.0196, "step": 55480 }, { "grad_norm": 0.25903043150901794, "learning_rate": 4.5065822549729605e-05, "loss": 0.0202, "step": 55490 }, { "grad_norm": 0.1875600963830948, "learning_rate": 4.504936882972989e-05, "loss": 0.021, "step": 55500 }, { "grad_norm": 0.2021872103214264, "learning_rate": 4.503291565112374e-05, "loss": 0.0212, "step": 55510 }, { "grad_norm": 0.34789735078811646, "learning_rate": 4.501646301571048e-05, "loss": 0.0172, "step": 55520 }, { "grad_norm": 0.1840897500514984, "learning_rate": 4.500001092528933e-05, "loss": 0.0182, "step": 55530 }, { "grad_norm": 0.23881079256534576, "learning_rate": 4.4983559381659455e-05, "loss": 0.0213, "step": 55540 }, { "grad_norm": 0.22531312704086304, "learning_rate": 4.496710838661999e-05, "loss": 0.0172, "step": 55550 }, { "grad_norm": 0.19552090764045715, "learning_rate": 4.495065794196999e-05, "loss": 0.018, "step": 55560 }, { "grad_norm": 0.21970173716545105, "learning_rate": 4.4934208049508455e-05, "loss": 0.0195, "step": 55570 }, { "grad_norm": 0.18574652075767517, "learning_rate": 4.4917758711034303e-05, "loss": 0.0184, "step": 55580 }, { "grad_norm": 0.13514339923858643, "learning_rate": 4.490130992834642e-05, "loss": 0.0213, "step": 55590 }, { "grad_norm": 0.1541750133037567, "learning_rate": 4.488486170324359e-05, "loss": 0.0169, "step": 55600 }, { "grad_norm": 0.19719688594341278, "learning_rate": 4.486841403752463e-05, "loss": 0.016, "step": 55610 }, { "grad_norm": 0.13268224895000458, "learning_rate": 4.4851966932988174e-05, "loss": 0.0173, "step": 55620 }, { "grad_norm": 0.2018502652645111, "learning_rate": 4.4835520391432894e-05, "loss": 0.0199, "step": 55630 }, { "grad_norm": 0.2194991111755371, "learning_rate": 4.481907441465733e-05, "loss": 0.0188, "step": 55640 }, { "grad_norm": 0.16393809020519257, "learning_rate": 4.480262900446e-05, "loss": 0.0191, "step": 55650 }, { "grad_norm": 0.22310929000377655, "learning_rate": 4.478618416263933e-05, "loss": 0.0157, "step": 55660 }, { "grad_norm": 0.1797362118959427, "learning_rate": 4.476973989099373e-05, "loss": 0.0206, "step": 55670 }, { "grad_norm": 0.17433813214302063, "learning_rate": 4.475329619132153e-05, "loss": 0.0166, "step": 55680 }, { "grad_norm": 0.2073035091161728, "learning_rate": 4.4736853065420956e-05, "loss": 0.0194, "step": 55690 }, { "grad_norm": 0.20424681901931763, "learning_rate": 4.472041051509023e-05, "loss": 0.0205, "step": 55700 }, { "grad_norm": 0.18249860405921936, "learning_rate": 4.470396854212746e-05, "loss": 0.0211, "step": 55710 }, { "grad_norm": 0.20699091255664825, "learning_rate": 4.468752714833072e-05, "loss": 0.0176, "step": 55720 }, { "grad_norm": 0.25672492384910583, "learning_rate": 4.467108633549803e-05, "loss": 0.0193, "step": 55730 }, { "grad_norm": 0.19561539590358734, "learning_rate": 4.4654646105427335e-05, "loss": 0.0216, "step": 55740 }, { "grad_norm": 0.21025818586349487, "learning_rate": 4.463820645991651e-05, "loss": 0.0208, "step": 55750 }, { "grad_norm": 0.168183833360672, "learning_rate": 4.462176740076334e-05, "loss": 0.0227, "step": 55760 }, { "grad_norm": 0.19125762581825256, "learning_rate": 4.4605328929765624e-05, "loss": 0.0251, "step": 55770 }, { "grad_norm": 0.21422244608402252, "learning_rate": 4.458889104872102e-05, "loss": 0.0205, "step": 55780 }, { "grad_norm": 0.2839389145374298, "learning_rate": 4.457245375942715e-05, "loss": 0.0207, "step": 55790 }, { "grad_norm": 0.2554323673248291, "learning_rate": 4.455601706368157e-05, "loss": 0.0178, "step": 55800 }, { "grad_norm": 0.15968947112560272, "learning_rate": 4.453958096328179e-05, "loss": 0.0187, "step": 55810 }, { "grad_norm": 0.18485872447490692, "learning_rate": 4.452314546002523e-05, "loss": 0.022, "step": 55820 }, { "grad_norm": 0.15900041162967682, "learning_rate": 4.4506710555709246e-05, "loss": 0.0179, "step": 55830 }, { "grad_norm": 0.3314917981624603, "learning_rate": 4.449027625213114e-05, "loss": 0.017, "step": 55840 }, { "grad_norm": 0.43338727951049805, "learning_rate": 4.4473842551088135e-05, "loss": 0.0202, "step": 55850 }, { "grad_norm": 0.21573059260845184, "learning_rate": 4.445740945437738e-05, "loss": 0.0193, "step": 55860 }, { "grad_norm": 0.22587163746356964, "learning_rate": 4.444097696379602e-05, "loss": 0.0187, "step": 55870 }, { "grad_norm": 0.2217966914176941, "learning_rate": 4.442454508114106e-05, "loss": 0.0184, "step": 55880 }, { "grad_norm": 0.23730473220348358, "learning_rate": 4.4408113808209465e-05, "loss": 0.0218, "step": 55890 }, { "grad_norm": 0.1815590262413025, "learning_rate": 4.439168314679813e-05, "loss": 0.0224, "step": 55900 }, { "grad_norm": 0.13467702269554138, "learning_rate": 4.43752530987039e-05, "loss": 0.0185, "step": 55910 }, { "grad_norm": 0.17161999642848969, "learning_rate": 4.4358823665723515e-05, "loss": 0.0203, "step": 55920 }, { "grad_norm": 0.22842535376548767, "learning_rate": 4.434239484965371e-05, "loss": 0.0181, "step": 55930 }, { "grad_norm": 0.1497594714164734, "learning_rate": 4.4325966652291103e-05, "loss": 0.0198, "step": 55940 }, { "grad_norm": 0.20385682582855225, "learning_rate": 4.430953907543225e-05, "loss": 0.0197, "step": 55950 }, { "grad_norm": 0.18581129610538483, "learning_rate": 4.4293112120873645e-05, "loss": 0.0179, "step": 55960 }, { "grad_norm": 0.29976263642311096, "learning_rate": 4.427668579041172e-05, "loss": 0.0182, "step": 55970 }, { "grad_norm": 0.22300447523593903, "learning_rate": 4.4260260085842816e-05, "loss": 0.018, "step": 55980 }, { "grad_norm": 0.21297363936901093, "learning_rate": 4.4243835008963256e-05, "loss": 0.0156, "step": 55990 }, { "grad_norm": 0.30444127321243286, "learning_rate": 4.422741056156925e-05, "loss": 0.0183, "step": 56000 }, { "grad_norm": 0.1538270264863968, "learning_rate": 4.421098674545693e-05, "loss": 0.0179, "step": 56010 }, { "grad_norm": 0.24848884344100952, "learning_rate": 4.419456356242241e-05, "loss": 0.0223, "step": 56020 }, { "grad_norm": 0.18314044177532196, "learning_rate": 4.4178141014261685e-05, "loss": 0.0181, "step": 56030 }, { "grad_norm": 0.2477293312549591, "learning_rate": 4.416171910277071e-05, "loss": 0.0196, "step": 56040 }, { "grad_norm": 0.2554730176925659, "learning_rate": 4.414529782974533e-05, "loss": 0.0191, "step": 56050 }, { "grad_norm": 0.19293074309825897, "learning_rate": 4.41288771969814e-05, "loss": 0.0155, "step": 56060 }, { "grad_norm": 0.14681270718574524, "learning_rate": 4.411245720627464e-05, "loss": 0.016, "step": 56070 }, { "grad_norm": 0.21039417386054993, "learning_rate": 4.40960378594207e-05, "loss": 0.0192, "step": 56080 }, { "grad_norm": 0.1739538162946701, "learning_rate": 4.407961915821519e-05, "loss": 0.0228, "step": 56090 }, { "grad_norm": 0.23251979053020477, "learning_rate": 4.406320110445362e-05, "loss": 0.0191, "step": 56100 }, { "grad_norm": 0.2206415981054306, "learning_rate": 4.404678369993144e-05, "loss": 0.0181, "step": 56110 }, { "grad_norm": 0.38844630122184753, "learning_rate": 4.403036694644406e-05, "loss": 0.0215, "step": 56120 }, { "grad_norm": 0.20887386798858643, "learning_rate": 4.4013950845786764e-05, "loss": 0.018, "step": 56130 }, { "grad_norm": 0.18093401193618774, "learning_rate": 4.399753539975482e-05, "loss": 0.019, "step": 56140 }, { "grad_norm": 0.19652625918388367, "learning_rate": 4.398112061014337e-05, "loss": 0.0165, "step": 56150 }, { "grad_norm": 0.1794167160987854, "learning_rate": 4.396470647874753e-05, "loss": 0.0185, "step": 56160 }, { "grad_norm": 0.1899978667497635, "learning_rate": 4.394829300736229e-05, "loss": 0.0165, "step": 56170 }, { "grad_norm": 0.17143514752388, "learning_rate": 4.393188019778265e-05, "loss": 0.0179, "step": 56180 }, { "grad_norm": 0.2218492478132248, "learning_rate": 4.391546805180347e-05, "loss": 0.0194, "step": 56190 }, { "grad_norm": 0.18148273229599, "learning_rate": 4.389905657121955e-05, "loss": 0.0222, "step": 56200 }, { "grad_norm": 0.21370123326778412, "learning_rate": 4.3882645757825635e-05, "loss": 0.0171, "step": 56210 }, { "grad_norm": 0.1407812386751175, "learning_rate": 4.386623561341637e-05, "loss": 0.0164, "step": 56220 }, { "grad_norm": 0.2715083360671997, "learning_rate": 4.384982613978637e-05, "loss": 0.0161, "step": 56230 }, { "grad_norm": 0.26955002546310425, "learning_rate": 4.383341733873012e-05, "loss": 0.0167, "step": 56240 }, { "grad_norm": 0.24198272824287415, "learning_rate": 4.381700921204209e-05, "loss": 0.0173, "step": 56250 }, { "grad_norm": 0.21561288833618164, "learning_rate": 4.380060176151663e-05, "loss": 0.0194, "step": 56260 }, { "grad_norm": 0.15542080998420715, "learning_rate": 4.378419498894805e-05, "loss": 0.0188, "step": 56270 }, { "grad_norm": 0.2163289338350296, "learning_rate": 4.376778889613056e-05, "loss": 0.015, "step": 56280 }, { "grad_norm": 0.21291853487491608, "learning_rate": 4.375138348485831e-05, "loss": 0.0184, "step": 56290 }, { "grad_norm": 0.16247056424617767, "learning_rate": 4.3734978756925346e-05, "loss": 0.0156, "step": 56300 }, { "grad_norm": 0.16769015789031982, "learning_rate": 4.3718574714125706e-05, "loss": 0.0159, "step": 56310 }, { "grad_norm": 0.13461431860923767, "learning_rate": 4.370217135825329e-05, "loss": 0.0201, "step": 56320 }, { "grad_norm": 0.17196869850158691, "learning_rate": 4.368576869110194e-05, "loss": 0.0235, "step": 56330 }, { "grad_norm": 0.23619724810123444, "learning_rate": 4.366936671446544e-05, "loss": 0.0176, "step": 56340 }, { "grad_norm": 0.19539852440357208, "learning_rate": 4.3652965430137474e-05, "loss": 0.0229, "step": 56350 }, { "grad_norm": 0.15080565214157104, "learning_rate": 4.3636564839911646e-05, "loss": 0.0176, "step": 56360 }, { "grad_norm": 0.23530830442905426, "learning_rate": 4.3620164945581545e-05, "loss": 0.0196, "step": 56370 }, { "grad_norm": 0.37971919775009155, "learning_rate": 4.360376574894061e-05, "loss": 0.0212, "step": 56380 }, { "grad_norm": 0.20721545815467834, "learning_rate": 4.358736725178224e-05, "loss": 0.0192, "step": 56390 }, { "grad_norm": 0.13764627277851105, "learning_rate": 4.357096945589974e-05, "loss": 0.0177, "step": 56400 }, { "grad_norm": 0.1583794802427292, "learning_rate": 4.355457236308636e-05, "loss": 0.0174, "step": 56410 }, { "grad_norm": 0.17196357250213623, "learning_rate": 4.353817597513526e-05, "loss": 0.0181, "step": 56420 }, { "grad_norm": 0.17398245632648468, "learning_rate": 4.352178029383948e-05, "loss": 0.0192, "step": 56430 }, { "grad_norm": 0.1542888730764389, "learning_rate": 4.3505385320992105e-05, "loss": 0.0174, "step": 56440 }, { "grad_norm": 0.2478359490633011, "learning_rate": 4.348899105838602e-05, "loss": 0.0196, "step": 56450 }, { "grad_norm": 0.19043873250484467, "learning_rate": 4.3472597507814087e-05, "loss": 0.0172, "step": 56460 }, { "grad_norm": 0.23570020496845245, "learning_rate": 4.3456204671069066e-05, "loss": 0.0164, "step": 56470 }, { "grad_norm": 0.1641254872083664, "learning_rate": 4.343981254994367e-05, "loss": 0.019, "step": 56480 }, { "grad_norm": 0.16940437257289886, "learning_rate": 4.34234211462305e-05, "loss": 0.0165, "step": 56490 }, { "grad_norm": 0.21241600811481476, "learning_rate": 4.340703046172213e-05, "loss": 0.0225, "step": 56500 }, { "grad_norm": 0.22982242703437805, "learning_rate": 4.339064049821097e-05, "loss": 0.0201, "step": 56510 }, { "grad_norm": 0.13724695146083832, "learning_rate": 4.3374251257489446e-05, "loss": 0.0187, "step": 56520 }, { "grad_norm": 0.17880910634994507, "learning_rate": 4.335786274134984e-05, "loss": 0.0151, "step": 56530 }, { "grad_norm": 0.17563396692276, "learning_rate": 4.334147495158438e-05, "loss": 0.0165, "step": 56540 }, { "grad_norm": 0.5560008883476257, "learning_rate": 4.33250878899852e-05, "loss": 0.0181, "step": 56550 }, { "grad_norm": 0.26785799860954285, "learning_rate": 4.330870155834437e-05, "loss": 0.0221, "step": 56560 }, { "grad_norm": 0.18916656076908112, "learning_rate": 4.3292315958453884e-05, "loss": 0.0176, "step": 56570 }, { "grad_norm": 0.22629643976688385, "learning_rate": 4.3275931092105646e-05, "loss": 0.0193, "step": 56580 }, { "grad_norm": 0.22570228576660156, "learning_rate": 4.325954696109147e-05, "loss": 0.0257, "step": 56590 }, { "grad_norm": 0.2062179148197174, "learning_rate": 4.32431635672031e-05, "loss": 0.0173, "step": 56600 }, { "grad_norm": 0.19793324172496796, "learning_rate": 4.3226780912232214e-05, "loss": 0.0207, "step": 56610 }, { "grad_norm": 0.2027864307165146, "learning_rate": 4.321039899797036e-05, "loss": 0.0168, "step": 56620 }, { "grad_norm": 0.3359309434890747, "learning_rate": 4.319401782620908e-05, "loss": 0.0197, "step": 56630 }, { "grad_norm": 0.22077976167201996, "learning_rate": 4.317763739873978e-05, "loss": 0.02, "step": 56640 }, { "grad_norm": 0.18547509610652924, "learning_rate": 4.31612577173538e-05, "loss": 0.0162, "step": 56650 }, { "grad_norm": 0.3582402765750885, "learning_rate": 4.314487878384239e-05, "loss": 0.0191, "step": 56660 }, { "grad_norm": 0.18623431026935577, "learning_rate": 4.3128500599996736e-05, "loss": 0.0153, "step": 56670 }, { "grad_norm": 0.2154698371887207, "learning_rate": 4.31121231676079e-05, "loss": 0.0151, "step": 56680 }, { "grad_norm": 0.17793531715869904, "learning_rate": 4.309574648846694e-05, "loss": 0.0153, "step": 56690 }, { "grad_norm": 0.14438055455684662, "learning_rate": 4.3079370564364755e-05, "loss": 0.0177, "step": 56700 }, { "grad_norm": 0.19932149350643158, "learning_rate": 4.306299539709221e-05, "loss": 0.0183, "step": 56710 }, { "grad_norm": 0.21289217472076416, "learning_rate": 4.3046620988440046e-05, "loss": 0.0175, "step": 56720 }, { "grad_norm": 0.18566715717315674, "learning_rate": 4.303024734019897e-05, "loss": 0.0148, "step": 56730 }, { "grad_norm": 0.22106195986270905, "learning_rate": 4.3013874454159565e-05, "loss": 0.0167, "step": 56740 }, { "grad_norm": 0.1734759360551834, "learning_rate": 4.299750233211233e-05, "loss": 0.023, "step": 56750 }, { "grad_norm": 0.35434892773628235, "learning_rate": 4.2981130975847715e-05, "loss": 0.0181, "step": 56760 }, { "grad_norm": 0.2056816816329956, "learning_rate": 4.2964760387156076e-05, "loss": 0.0195, "step": 56770 }, { "grad_norm": 0.23892024159431458, "learning_rate": 4.294839056782767e-05, "loss": 0.0195, "step": 56780 }, { "grad_norm": 0.15259377658367157, "learning_rate": 4.293202151965266e-05, "loss": 0.0173, "step": 56790 }, { "grad_norm": 0.256651371717453, "learning_rate": 4.291565324442115e-05, "loss": 0.0159, "step": 56800 }, { "grad_norm": 0.1946965754032135, "learning_rate": 4.2899285743923135e-05, "loss": 0.0268, "step": 56810 }, { "grad_norm": 0.2922613322734833, "learning_rate": 4.288291901994857e-05, "loss": 0.018, "step": 56820 }, { "grad_norm": 0.19394133985042572, "learning_rate": 4.286655307428729e-05, "loss": 0.0202, "step": 56830 }, { "grad_norm": 0.15775685012340546, "learning_rate": 4.285018790872903e-05, "loss": 0.0195, "step": 56840 }, { "grad_norm": 0.18436512351036072, "learning_rate": 4.283382352506347e-05, "loss": 0.019, "step": 56850 }, { "grad_norm": 0.17480771243572235, "learning_rate": 4.2817459925080205e-05, "loss": 0.0161, "step": 56860 }, { "grad_norm": 0.20681877434253693, "learning_rate": 4.28010971105687e-05, "loss": 0.0173, "step": 56870 }, { "grad_norm": 0.17499656975269318, "learning_rate": 4.278473508331841e-05, "loss": 0.0168, "step": 56880 }, { "grad_norm": 0.2434036284685135, "learning_rate": 4.276837384511864e-05, "loss": 0.0173, "step": 56890 }, { "grad_norm": 0.3067797124385834, "learning_rate": 4.275201339775864e-05, "loss": 0.0198, "step": 56900 }, { "grad_norm": 0.23677699267864227, "learning_rate": 4.273565374302756e-05, "loss": 0.0179, "step": 56910 }, { "grad_norm": 0.18039855360984802, "learning_rate": 4.2719294882714454e-05, "loss": 0.0169, "step": 56920 }, { "grad_norm": 0.270870178937912, "learning_rate": 4.270293681860831e-05, "loss": 0.0194, "step": 56930 }, { "grad_norm": 0.1833641082048416, "learning_rate": 4.268657955249802e-05, "loss": 0.0189, "step": 56940 }, { "grad_norm": 0.23323585093021393, "learning_rate": 4.267022308617241e-05, "loss": 0.0175, "step": 56950 }, { "grad_norm": 0.20319324731826782, "learning_rate": 4.265386742142018e-05, "loss": 0.0207, "step": 56960 }, { "grad_norm": 0.18146109580993652, "learning_rate": 4.263751256002996e-05, "loss": 0.0154, "step": 56970 }, { "grad_norm": 0.13300368189811707, "learning_rate": 4.26211585037903e-05, "loss": 0.0186, "step": 56980 }, { "grad_norm": 0.15975938737392426, "learning_rate": 4.2604805254489656e-05, "loss": 0.0164, "step": 56990 }, { "grad_norm": 0.15393880009651184, "learning_rate": 4.258845281391638e-05, "loss": 0.0183, "step": 57000 }, { "grad_norm": 0.16847115755081177, "learning_rate": 4.257210118385876e-05, "loss": 0.0144, "step": 57010 }, { "grad_norm": 0.15289857983589172, "learning_rate": 4.2555750366104994e-05, "loss": 0.0163, "step": 57020 }, { "grad_norm": 0.12137490510940552, "learning_rate": 4.2539400362443184e-05, "loss": 0.0179, "step": 57030 }, { "grad_norm": 0.16751477122306824, "learning_rate": 4.252305117466133e-05, "loss": 0.0179, "step": 57040 }, { "grad_norm": 0.1526920199394226, "learning_rate": 4.2506702804547356e-05, "loss": 0.0163, "step": 57050 }, { "grad_norm": 0.16221293807029724, "learning_rate": 4.249035525388907e-05, "loss": 0.0188, "step": 57060 }, { "grad_norm": 0.23846733570098877, "learning_rate": 4.247400852447427e-05, "loss": 0.0187, "step": 57070 }, { "grad_norm": 0.2643286883831024, "learning_rate": 4.245766261809059e-05, "loss": 0.0186, "step": 57080 }, { "grad_norm": 0.20089706778526306, "learning_rate": 4.244131753652558e-05, "loss": 0.0185, "step": 57090 }, { "grad_norm": 0.4024046063423157, "learning_rate": 4.242497328156672e-05, "loss": 0.0162, "step": 57100 }, { "grad_norm": 0.2850140333175659, "learning_rate": 4.240862985500139e-05, "loss": 0.0184, "step": 57110 }, { "grad_norm": 0.21324288845062256, "learning_rate": 4.2392287258616894e-05, "loss": 0.0165, "step": 57120 }, { "grad_norm": 0.322889119386673, "learning_rate": 4.2375945494200396e-05, "loss": 0.0192, "step": 57130 }, { "grad_norm": 0.15192362666130066, "learning_rate": 4.2359604563539054e-05, "loss": 0.0186, "step": 57140 }, { "grad_norm": 0.20941393077373505, "learning_rate": 4.2343264468419885e-05, "loss": 0.0194, "step": 57150 }, { "grad_norm": 0.15494376420974731, "learning_rate": 4.232692521062979e-05, "loss": 0.0172, "step": 57160 }, { "grad_norm": 0.13736727833747864, "learning_rate": 4.231058679195562e-05, "loss": 0.0188, "step": 57170 }, { "grad_norm": 0.35908910632133484, "learning_rate": 4.229424921418411e-05, "loss": 0.0199, "step": 57180 }, { "grad_norm": 0.18780827522277832, "learning_rate": 4.2277912479101916e-05, "loss": 0.0163, "step": 57190 }, { "grad_norm": 0.2007029801607132, "learning_rate": 4.226157658849562e-05, "loss": 0.0187, "step": 57200 }, { "grad_norm": 0.2098010629415512, "learning_rate": 4.2245241544151664e-05, "loss": 0.0173, "step": 57210 }, { "grad_norm": 0.2408423274755478, "learning_rate": 4.2228907347856425e-05, "loss": 0.0191, "step": 57220 }, { "grad_norm": 0.2688041627407074, "learning_rate": 4.2212574001396214e-05, "loss": 0.0204, "step": 57230 }, { "grad_norm": 0.18947133421897888, "learning_rate": 4.21962415065572e-05, "loss": 0.0192, "step": 57240 }, { "grad_norm": 0.21335509419441223, "learning_rate": 4.217990986512548e-05, "loss": 0.024, "step": 57250 }, { "grad_norm": 0.2252698838710785, "learning_rate": 4.216357907888704e-05, "loss": 0.0218, "step": 57260 }, { "grad_norm": 0.24499884247779846, "learning_rate": 4.2147249149627824e-05, "loss": 0.0174, "step": 57270 }, { "grad_norm": 0.20363670587539673, "learning_rate": 4.213092007913364e-05, "loss": 0.0227, "step": 57280 }, { "grad_norm": 0.316353976726532, "learning_rate": 4.2114591869190205e-05, "loss": 0.0236, "step": 57290 }, { "grad_norm": 0.2974039614200592, "learning_rate": 4.209826452158315e-05, "loss": 0.0197, "step": 57300 }, { "grad_norm": 0.22052407264709473, "learning_rate": 4.2081938038098e-05, "loss": 0.0165, "step": 57310 }, { "grad_norm": 0.3252982199192047, "learning_rate": 4.2065612420520185e-05, "loss": 0.0183, "step": 57320 }, { "grad_norm": 0.21789850294589996, "learning_rate": 4.204928767063509e-05, "loss": 0.0177, "step": 57330 }, { "grad_norm": 0.20726484060287476, "learning_rate": 4.2032963790227944e-05, "loss": 0.0162, "step": 57340 }, { "grad_norm": 0.2045115828514099, "learning_rate": 4.2016640781083887e-05, "loss": 0.0227, "step": 57350 }, { "grad_norm": 0.19614654779434204, "learning_rate": 4.2000318644988004e-05, "loss": 0.0196, "step": 57360 }, { "grad_norm": 0.24786044657230377, "learning_rate": 4.198399738372524e-05, "loss": 0.02, "step": 57370 }, { "grad_norm": 0.2619696855545044, "learning_rate": 4.196767699908045e-05, "loss": 0.0204, "step": 57380 }, { "grad_norm": 0.2933409512042999, "learning_rate": 4.195135749283845e-05, "loss": 0.0205, "step": 57390 }, { "grad_norm": 0.2473687082529068, "learning_rate": 4.1935038866783894e-05, "loss": 0.0192, "step": 57400 }, { "grad_norm": 0.17017516493797302, "learning_rate": 4.1918721122701355e-05, "loss": 0.0161, "step": 57410 }, { "grad_norm": 0.18710657954216003, "learning_rate": 4.190240426237533e-05, "loss": 0.018, "step": 57420 }, { "grad_norm": 0.14561069011688232, "learning_rate": 4.188608828759019e-05, "loss": 0.0159, "step": 57430 }, { "grad_norm": 0.48552706837654114, "learning_rate": 4.1869773200130246e-05, "loss": 0.0178, "step": 57440 }, { "grad_norm": 0.14178211987018585, "learning_rate": 4.185345900177967e-05, "loss": 0.0177, "step": 57450 }, { "grad_norm": 0.21628905832767487, "learning_rate": 4.183714569432258e-05, "loss": 0.0182, "step": 57460 }, { "grad_norm": 0.27063480019569397, "learning_rate": 4.182083327954295e-05, "loss": 0.019, "step": 57470 }, { "grad_norm": 0.4527455270290375, "learning_rate": 4.1804521759224714e-05, "loss": 0.02, "step": 57480 }, { "grad_norm": 0.2420421838760376, "learning_rate": 4.178821113515165e-05, "loss": 0.021, "step": 57490 }, { "grad_norm": 0.30052003264427185, "learning_rate": 4.1771901409107476e-05, "loss": 0.0201, "step": 57500 }, { "grad_norm": 0.3126792013645172, "learning_rate": 4.175559258287577e-05, "loss": 0.0159, "step": 57510 }, { "grad_norm": 0.22537682950496674, "learning_rate": 4.173928465824008e-05, "loss": 0.0189, "step": 57520 }, { "grad_norm": 0.16341695189476013, "learning_rate": 4.17229776369838e-05, "loss": 0.0194, "step": 57530 }, { "grad_norm": 0.21690645813941956, "learning_rate": 4.170667152089025e-05, "loss": 0.0194, "step": 57540 }, { "grad_norm": 0.20364896953105927, "learning_rate": 4.1690366311742626e-05, "loss": 0.0174, "step": 57550 }, { "grad_norm": 0.1878155916929245, "learning_rate": 4.1674062011324055e-05, "loss": 0.0222, "step": 57560 }, { "grad_norm": 0.21615158021450043, "learning_rate": 4.165775862141752e-05, "loss": 0.0194, "step": 57570 }, { "grad_norm": 0.17387275397777557, "learning_rate": 4.164145614380598e-05, "loss": 0.0196, "step": 57580 }, { "grad_norm": 0.229639932513237, "learning_rate": 4.162515458027223e-05, "loss": 0.0163, "step": 57590 }, { "grad_norm": 0.2178700715303421, "learning_rate": 4.160885393259899e-05, "loss": 0.0198, "step": 57600 }, { "grad_norm": 0.16617941856384277, "learning_rate": 4.159255420256886e-05, "loss": 0.0153, "step": 57610 }, { "grad_norm": 0.20108315348625183, "learning_rate": 4.157625539196436e-05, "loss": 0.0179, "step": 57620 }, { "grad_norm": 0.2686469554901123, "learning_rate": 4.155995750256791e-05, "loss": 0.0237, "step": 57630 }, { "grad_norm": 0.20330595970153809, "learning_rate": 4.15436605361618e-05, "loss": 0.0183, "step": 57640 }, { "grad_norm": 0.1562120020389557, "learning_rate": 4.152736449452827e-05, "loss": 0.0168, "step": 57650 }, { "grad_norm": 0.17659799754619598, "learning_rate": 4.151106937944943e-05, "loss": 0.0192, "step": 57660 }, { "grad_norm": 0.15038703382015228, "learning_rate": 4.149477519270727e-05, "loss": 0.0218, "step": 57670 }, { "grad_norm": 0.17955386638641357, "learning_rate": 4.147848193608371e-05, "loss": 0.0174, "step": 57680 }, { "grad_norm": 0.2249792218208313, "learning_rate": 4.146218961136056e-05, "loss": 0.0168, "step": 57690 }, { "grad_norm": 0.14557957649230957, "learning_rate": 4.14458982203195e-05, "loss": 0.0175, "step": 57700 }, { "grad_norm": 0.1329835057258606, "learning_rate": 4.1429607764742165e-05, "loss": 0.0167, "step": 57710 }, { "grad_norm": 0.24928374588489532, "learning_rate": 4.141331824641003e-05, "loss": 0.0178, "step": 57720 }, { "grad_norm": 0.20735713839530945, "learning_rate": 4.139702966710452e-05, "loss": 0.0187, "step": 57730 }, { "grad_norm": 0.1932888627052307, "learning_rate": 4.13807420286069e-05, "loss": 0.0179, "step": 57740 }, { "grad_norm": 0.19634848833084106, "learning_rate": 4.136445533269837e-05, "loss": 0.0185, "step": 57750 }, { "grad_norm": 0.24839816987514496, "learning_rate": 4.134816958116003e-05, "loss": 0.0195, "step": 57760 }, { "grad_norm": 0.3053387999534607, "learning_rate": 4.133188477577283e-05, "loss": 0.0193, "step": 57770 }, { "grad_norm": 0.3080415427684784, "learning_rate": 4.131560091831771e-05, "loss": 0.0196, "step": 57780 }, { "grad_norm": 0.22453734278678894, "learning_rate": 4.1299318010575405e-05, "loss": 0.0182, "step": 57790 }, { "grad_norm": 0.22551992535591125, "learning_rate": 4.128303605432661e-05, "loss": 0.0169, "step": 57800 }, { "grad_norm": 0.14756320416927338, "learning_rate": 4.126675505135188e-05, "loss": 0.0174, "step": 57810 }, { "grad_norm": 0.19650128483772278, "learning_rate": 4.125047500343169e-05, "loss": 0.0173, "step": 57820 }, { "grad_norm": 0.1959175020456314, "learning_rate": 4.1234195912346376e-05, "loss": 0.017, "step": 57830 }, { "grad_norm": 0.19751742482185364, "learning_rate": 4.1217917779876235e-05, "loss": 0.0159, "step": 57840 }, { "grad_norm": 0.19740471243858337, "learning_rate": 4.120164060780141e-05, "loss": 0.0201, "step": 57850 }, { "grad_norm": 0.12213601917028427, "learning_rate": 4.1185364397901935e-05, "loss": 0.0158, "step": 57860 }, { "grad_norm": 0.19100388884544373, "learning_rate": 4.116908915195776e-05, "loss": 0.0191, "step": 57870 }, { "grad_norm": 0.17727874219417572, "learning_rate": 4.115281487174872e-05, "loss": 0.0171, "step": 57880 }, { "grad_norm": 0.38224270939826965, "learning_rate": 4.113654155905452e-05, "loss": 0.0213, "step": 57890 }, { "grad_norm": 0.1832171529531479, "learning_rate": 4.112026921565483e-05, "loss": 0.0172, "step": 57900 }, { "grad_norm": 0.2123265117406845, "learning_rate": 4.1103997843329147e-05, "loss": 0.022, "step": 57910 }, { "grad_norm": 0.2552233934402466, "learning_rate": 4.108772744385689e-05, "loss": 0.0175, "step": 57920 }, { "grad_norm": 0.1885574758052826, "learning_rate": 4.1071458019017346e-05, "loss": 0.0177, "step": 57930 }, { "grad_norm": 0.25996458530426025, "learning_rate": 4.1055189570589734e-05, "loss": 0.0183, "step": 57940 }, { "grad_norm": 0.21807131171226501, "learning_rate": 4.103892210035315e-05, "loss": 0.0201, "step": 57950 }, { "grad_norm": 0.25375473499298096, "learning_rate": 4.102265561008656e-05, "loss": 0.0188, "step": 57960 }, { "grad_norm": 0.21620778739452362, "learning_rate": 4.100639010156885e-05, "loss": 0.0181, "step": 57970 }, { "grad_norm": 0.35073575377464294, "learning_rate": 4.099012557657882e-05, "loss": 0.0199, "step": 57980 }, { "grad_norm": 0.20936650037765503, "learning_rate": 4.09738620368951e-05, "loss": 0.0203, "step": 57990 }, { "grad_norm": 0.24882370233535767, "learning_rate": 4.095759948429626e-05, "loss": 0.0182, "step": 58000 }, { "grad_norm": 0.18703985214233398, "learning_rate": 4.094133792056075e-05, "loss": 0.0177, "step": 58010 }, { "grad_norm": 0.2188386470079422, "learning_rate": 4.092507734746688e-05, "loss": 0.0202, "step": 58020 }, { "grad_norm": 0.17435400187969208, "learning_rate": 4.090881776679293e-05, "loss": 0.0197, "step": 58030 }, { "grad_norm": 0.24242186546325684, "learning_rate": 4.0892559180316994e-05, "loss": 0.0184, "step": 58040 }, { "grad_norm": 0.19897884130477905, "learning_rate": 4.08763015898171e-05, "loss": 0.0168, "step": 58050 }, { "grad_norm": 0.3274145722389221, "learning_rate": 4.0860044997071145e-05, "loss": 0.0171, "step": 58060 }, { "grad_norm": 0.2158014178276062, "learning_rate": 4.084378940385693e-05, "loss": 0.0182, "step": 58070 }, { "grad_norm": 0.2506193220615387, "learning_rate": 4.082753481195211e-05, "loss": 0.0204, "step": 58080 }, { "grad_norm": 0.23429684340953827, "learning_rate": 4.0811281223134324e-05, "loss": 0.016, "step": 58090 }, { "grad_norm": 0.22767464816570282, "learning_rate": 4.0795028639181e-05, "loss": 0.0162, "step": 58100 }, { "grad_norm": 0.20993733406066895, "learning_rate": 4.077877706186951e-05, "loss": 0.0203, "step": 58110 }, { "grad_norm": 0.2167709916830063, "learning_rate": 4.076252649297709e-05, "loss": 0.0145, "step": 58120 }, { "grad_norm": 0.2608722150325775, "learning_rate": 4.07462769342809e-05, "loss": 0.0197, "step": 58130 }, { "grad_norm": 0.1769496649503708, "learning_rate": 4.073002838755793e-05, "loss": 0.0153, "step": 58140 }, { "grad_norm": 0.1888217329978943, "learning_rate": 4.0713780854585115e-05, "loss": 0.0167, "step": 58150 }, { "grad_norm": 0.17453806102275848, "learning_rate": 4.069753433713927e-05, "loss": 0.0182, "step": 58160 }, { "grad_norm": 0.21943363547325134, "learning_rate": 4.068128883699709e-05, "loss": 0.0157, "step": 58170 }, { "grad_norm": 0.2517607808113098, "learning_rate": 4.0665044355935155e-05, "loss": 0.0181, "step": 58180 }, { "grad_norm": 0.19112583994865417, "learning_rate": 4.064880089572993e-05, "loss": 0.0177, "step": 58190 }, { "grad_norm": 0.2527977228164673, "learning_rate": 4.063255845815779e-05, "loss": 0.0206, "step": 58200 }, { "grad_norm": 0.2169349044561386, "learning_rate": 4.0616317044994945e-05, "loss": 0.0178, "step": 58210 }, { "grad_norm": 0.19126398861408234, "learning_rate": 4.0600076658017585e-05, "loss": 0.0173, "step": 58220 }, { "grad_norm": 0.1924518197774887, "learning_rate": 4.0583837299001706e-05, "loss": 0.0179, "step": 58230 }, { "grad_norm": 0.2046685814857483, "learning_rate": 4.056759896972322e-05, "loss": 0.0181, "step": 58240 }, { "grad_norm": 0.24404318630695343, "learning_rate": 4.055136167195793e-05, "loss": 0.0198, "step": 58250 }, { "grad_norm": 0.18028052151203156, "learning_rate": 4.053512540748152e-05, "loss": 0.0149, "step": 58260 }, { "grad_norm": 0.19213072955608368, "learning_rate": 4.051889017806955e-05, "loss": 0.0192, "step": 58270 }, { "grad_norm": 0.17843957245349884, "learning_rate": 4.0502655985497516e-05, "loss": 0.0181, "step": 58280 }, { "grad_norm": 0.1773385852575302, "learning_rate": 4.048642283154074e-05, "loss": 0.0179, "step": 58290 }, { "grad_norm": 0.21343502402305603, "learning_rate": 4.0470190717974464e-05, "loss": 0.0195, "step": 58300 }, { "grad_norm": 0.2829630672931671, "learning_rate": 4.045395964657381e-05, "loss": 0.0174, "step": 58310 }, { "grad_norm": 0.2707058787345886, "learning_rate": 4.043772961911376e-05, "loss": 0.0182, "step": 58320 }, { "grad_norm": 0.2344697266817093, "learning_rate": 4.042150063736924e-05, "loss": 0.0194, "step": 58330 }, { "grad_norm": 0.20819301903247833, "learning_rate": 4.0405272703114976e-05, "loss": 0.0184, "step": 58340 }, { "grad_norm": 0.26838430762290955, "learning_rate": 4.038904581812569e-05, "loss": 0.0197, "step": 58350 }, { "grad_norm": 0.12734438478946686, "learning_rate": 4.03728199841759e-05, "loss": 0.0201, "step": 58360 }, { "grad_norm": 0.23396039009094238, "learning_rate": 4.0356595203040044e-05, "loss": 0.017, "step": 58370 }, { "grad_norm": 0.1668984591960907, "learning_rate": 4.0340371476492426e-05, "loss": 0.0212, "step": 58380 }, { "grad_norm": 0.17568302154541016, "learning_rate": 4.032414880630727e-05, "loss": 0.0206, "step": 58390 }, { "grad_norm": 0.27061423659324646, "learning_rate": 4.0307927194258646e-05, "loss": 0.0196, "step": 58400 }, { "grad_norm": 0.21050268411636353, "learning_rate": 4.029170664212054e-05, "loss": 0.016, "step": 58410 }, { "grad_norm": 0.23171532154083252, "learning_rate": 4.0275487151666794e-05, "loss": 0.0148, "step": 58420 }, { "grad_norm": 0.17646056413650513, "learning_rate": 4.025926872467116e-05, "loss": 0.0208, "step": 58430 }, { "grad_norm": 0.22074618935585022, "learning_rate": 4.024305136290725e-05, "loss": 0.0156, "step": 58440 }, { "grad_norm": 0.19062846899032593, "learning_rate": 4.022683506814858e-05, "loss": 0.0157, "step": 58450 }, { "grad_norm": 0.25062310695648193, "learning_rate": 4.021061984216853e-05, "loss": 0.0203, "step": 58460 }, { "grad_norm": 0.18429750204086304, "learning_rate": 4.019440568674035e-05, "loss": 0.0202, "step": 58470 }, { "grad_norm": 0.19442203640937805, "learning_rate": 4.017819260363724e-05, "loss": 0.0202, "step": 58480 }, { "grad_norm": 0.30489352345466614, "learning_rate": 4.016198059463223e-05, "loss": 0.0213, "step": 58490 }, { "grad_norm": 0.19792945683002472, "learning_rate": 4.0145769661498233e-05, "loss": 0.0173, "step": 58500 }, { "grad_norm": 0.21060127019882202, "learning_rate": 4.012955980600804e-05, "loss": 0.016, "step": 58510 }, { "grad_norm": 0.3841991424560547, "learning_rate": 4.0113351029934355e-05, "loss": 0.0186, "step": 58520 }, { "grad_norm": 0.1553727090358734, "learning_rate": 4.0097143335049714e-05, "loss": 0.0161, "step": 58530 }, { "grad_norm": 0.17263972759246826, "learning_rate": 4.0080936723126606e-05, "loss": 0.0188, "step": 58540 }, { "grad_norm": 0.19505168497562408, "learning_rate": 4.006473119593735e-05, "loss": 0.0166, "step": 58550 }, { "grad_norm": 0.1635507494211197, "learning_rate": 4.004852675525415e-05, "loss": 0.0165, "step": 58560 }, { "grad_norm": 0.1838599294424057, "learning_rate": 4.00323234028491e-05, "loss": 0.0164, "step": 58570 }, { "grad_norm": 0.19695869088172913, "learning_rate": 4.0016121140494186e-05, "loss": 0.0185, "step": 58580 }, { "grad_norm": 0.15718185901641846, "learning_rate": 3.999991996996122e-05, "loss": 0.0166, "step": 58590 }, { "grad_norm": 0.2231595665216446, "learning_rate": 3.9983719893022e-05, "loss": 0.02, "step": 58600 }, { "grad_norm": 0.19277530908584595, "learning_rate": 3.996752091144812e-05, "loss": 0.0167, "step": 58610 }, { "grad_norm": 0.20992206037044525, "learning_rate": 3.9951323027011055e-05, "loss": 0.0163, "step": 58620 }, { "grad_norm": 0.1681032031774521, "learning_rate": 3.993512624148219e-05, "loss": 0.0173, "step": 58630 }, { "grad_norm": 0.22421684861183167, "learning_rate": 3.99189305566328e-05, "loss": 0.0174, "step": 58640 }, { "grad_norm": 0.2422105371952057, "learning_rate": 3.990273597423401e-05, "loss": 0.0224, "step": 58650 }, { "grad_norm": 0.23992058634757996, "learning_rate": 3.988654249605681e-05, "loss": 0.0165, "step": 58660 }, { "grad_norm": 0.1556302160024643, "learning_rate": 3.987035012387213e-05, "loss": 0.0173, "step": 58670 }, { "grad_norm": 0.2489759773015976, "learning_rate": 3.9854158859450735e-05, "loss": 0.0161, "step": 58680 }, { "grad_norm": 0.163273885846138, "learning_rate": 3.983796870456328e-05, "loss": 0.0176, "step": 58690 }, { "grad_norm": 0.22600214183330536, "learning_rate": 3.982177966098028e-05, "loss": 0.0171, "step": 58700 }, { "grad_norm": 0.17964443564414978, "learning_rate": 3.980559173047216e-05, "loss": 0.0169, "step": 58710 }, { "grad_norm": 0.2430780678987503, "learning_rate": 3.9789404914809176e-05, "loss": 0.0206, "step": 58720 }, { "grad_norm": 0.385399729013443, "learning_rate": 3.9773219215761544e-05, "loss": 0.0181, "step": 58730 }, { "grad_norm": 0.2160359025001526, "learning_rate": 3.975703463509928e-05, "loss": 0.0184, "step": 58740 }, { "grad_norm": 0.2010682076215744, "learning_rate": 3.974085117459231e-05, "loss": 0.0159, "step": 58750 }, { "grad_norm": 0.20405006408691406, "learning_rate": 3.972466883601043e-05, "loss": 0.0171, "step": 58760 }, { "grad_norm": 0.19927023351192474, "learning_rate": 3.9708487621123316e-05, "loss": 0.0172, "step": 58770 }, { "grad_norm": 0.19208167493343353, "learning_rate": 3.969230753170051e-05, "loss": 0.0192, "step": 58780 }, { "grad_norm": 0.19321957230567932, "learning_rate": 3.967612856951146e-05, "loss": 0.0182, "step": 58790 }, { "grad_norm": 0.25816357135772705, "learning_rate": 3.9659950736325476e-05, "loss": 0.0153, "step": 58800 }, { "grad_norm": 0.20508445799350739, "learning_rate": 3.964377403391174e-05, "loss": 0.0187, "step": 58810 }, { "grad_norm": 0.1573982536792755, "learning_rate": 3.9627598464039303e-05, "loss": 0.0157, "step": 58820 }, { "grad_norm": 0.2272995412349701, "learning_rate": 3.9611424028477096e-05, "loss": 0.0182, "step": 58830 }, { "grad_norm": 0.21737559139728546, "learning_rate": 3.959525072899394e-05, "loss": 0.0179, "step": 58840 }, { "grad_norm": 0.18991869688034058, "learning_rate": 3.9579078567358506e-05, "loss": 0.0161, "step": 58850 }, { "grad_norm": 0.23585176467895508, "learning_rate": 3.956290754533939e-05, "loss": 0.0217, "step": 58860 }, { "grad_norm": 0.33145689964294434, "learning_rate": 3.954673766470501e-05, "loss": 0.0207, "step": 58870 }, { "grad_norm": 0.302279531955719, "learning_rate": 3.9530568927223674e-05, "loss": 0.0191, "step": 58880 }, { "grad_norm": 0.21778637170791626, "learning_rate": 3.95144013346636e-05, "loss": 0.019, "step": 58890 }, { "grad_norm": 0.2098274528980255, "learning_rate": 3.949823488879282e-05, "loss": 0.0162, "step": 58900 }, { "grad_norm": 0.18369074165821075, "learning_rate": 3.9482069591379276e-05, "loss": 0.0178, "step": 58910 }, { "grad_norm": 0.20939227938652039, "learning_rate": 3.9465905444190784e-05, "loss": 0.0171, "step": 58920 }, { "grad_norm": 0.17005684971809387, "learning_rate": 3.944974244899505e-05, "loss": 0.0148, "step": 58930 }, { "grad_norm": 0.1797439604997635, "learning_rate": 3.9433580607559616e-05, "loss": 0.0181, "step": 58940 }, { "grad_norm": 0.24569149315357208, "learning_rate": 3.941741992165193e-05, "loss": 0.0162, "step": 58950 }, { "grad_norm": 0.23325534164905548, "learning_rate": 3.940126039303928e-05, "loss": 0.0195, "step": 58960 }, { "grad_norm": 0.2671443521976471, "learning_rate": 3.938510202348886e-05, "loss": 0.0194, "step": 58970 }, { "grad_norm": 0.18149301409721375, "learning_rate": 3.93689448147677e-05, "loss": 0.0161, "step": 58980 }, { "grad_norm": 0.18939465284347534, "learning_rate": 3.935278876864278e-05, "loss": 0.0195, "step": 58990 }, { "grad_norm": 0.18563054502010345, "learning_rate": 3.933663388688087e-05, "loss": 0.0191, "step": 59000 }, { "grad_norm": 0.21272630989551544, "learning_rate": 3.9320480171248645e-05, "loss": 0.0226, "step": 59010 }, { "grad_norm": 0.18875087797641754, "learning_rate": 3.930432762351265e-05, "loss": 0.0197, "step": 59020 }, { "grad_norm": 0.19099341332912445, "learning_rate": 3.928817624543931e-05, "loss": 0.0179, "step": 59030 }, { "grad_norm": 0.1677238643169403, "learning_rate": 3.9272026038794885e-05, "loss": 0.0192, "step": 59040 }, { "grad_norm": 0.23726129531860352, "learning_rate": 3.9255877005345586e-05, "loss": 0.0214, "step": 59050 }, { "grad_norm": 0.2525598108768463, "learning_rate": 3.923972914685743e-05, "loss": 0.0194, "step": 59060 }, { "grad_norm": 0.2796488106250763, "learning_rate": 3.922358246509631e-05, "loss": 0.0193, "step": 59070 }, { "grad_norm": 0.1681472808122635, "learning_rate": 3.920743696182801e-05, "loss": 0.0193, "step": 59080 }, { "grad_norm": 0.2095167189836502, "learning_rate": 3.9191292638818164e-05, "loss": 0.0164, "step": 59090 }, { "grad_norm": 0.14428792893886566, "learning_rate": 3.9175149497832316e-05, "loss": 0.0192, "step": 59100 }, { "grad_norm": 0.27148741483688354, "learning_rate": 3.915900754063584e-05, "loss": 0.0198, "step": 59110 }, { "grad_norm": 0.15143929421901703, "learning_rate": 3.9142866768994014e-05, "loss": 0.0208, "step": 59120 }, { "grad_norm": 0.22899079322814941, "learning_rate": 3.912672718467193e-05, "loss": 0.016, "step": 59130 }, { "grad_norm": 0.18501409888267517, "learning_rate": 3.911058878943463e-05, "loss": 0.0196, "step": 59140 }, { "grad_norm": 0.22062300145626068, "learning_rate": 3.9094451585046974e-05, "loss": 0.0179, "step": 59150 }, { "grad_norm": 0.20346775650978088, "learning_rate": 3.90783155732737e-05, "loss": 0.0177, "step": 59160 }, { "grad_norm": 0.1657017320394516, "learning_rate": 3.9062180755879373e-05, "loss": 0.0165, "step": 59170 }, { "grad_norm": 0.5689907670021057, "learning_rate": 3.904604713462855e-05, "loss": 0.0152, "step": 59180 }, { "grad_norm": 0.2753983438014984, "learning_rate": 3.902991471128554e-05, "loss": 0.02, "step": 59190 }, { "grad_norm": 0.18185874819755554, "learning_rate": 3.901378348761456e-05, "loss": 0.0203, "step": 59200 }, { "grad_norm": 0.17343297600746155, "learning_rate": 3.899765346537971e-05, "loss": 0.0185, "step": 59210 }, { "grad_norm": 0.2613959014415741, "learning_rate": 3.898152464634493e-05, "loss": 0.0221, "step": 59220 }, { "grad_norm": 0.1597929745912552, "learning_rate": 3.896539703227402e-05, "loss": 0.021, "step": 59230 }, { "grad_norm": 0.15595529973506927, "learning_rate": 3.894927062493073e-05, "loss": 0.0225, "step": 59240 }, { "grad_norm": 0.22586818039417267, "learning_rate": 3.893314542607859e-05, "loss": 0.0141, "step": 59250 }, { "grad_norm": 0.24618104100227356, "learning_rate": 3.891702143748103e-05, "loss": 0.0231, "step": 59260 }, { "grad_norm": 0.1870134174823761, "learning_rate": 3.890089866090133e-05, "loss": 0.0167, "step": 59270 }, { "grad_norm": 0.3289352059364319, "learning_rate": 3.888477709810267e-05, "loss": 0.0187, "step": 59280 }, { "grad_norm": 0.24561533331871033, "learning_rate": 3.886865675084806e-05, "loss": 0.0224, "step": 59290 }, { "grad_norm": 0.24493901431560516, "learning_rate": 3.885253762090043e-05, "loss": 0.0163, "step": 59300 }, { "grad_norm": 0.2095031589269638, "learning_rate": 3.8836419710022534e-05, "loss": 0.0188, "step": 59310 }, { "grad_norm": 0.21406172215938568, "learning_rate": 3.882030301997698e-05, "loss": 0.0179, "step": 59320 }, { "grad_norm": 0.18425223231315613, "learning_rate": 3.880418755252629e-05, "loss": 0.0166, "step": 59330 }, { "grad_norm": 0.18057172000408173, "learning_rate": 3.87880733094328e-05, "loss": 0.0165, "step": 59340 }, { "grad_norm": 0.3916345536708832, "learning_rate": 3.877196029245877e-05, "loss": 0.0154, "step": 59350 }, { "grad_norm": 0.19222979247570038, "learning_rate": 3.875584850336627e-05, "loss": 0.0167, "step": 59360 }, { "grad_norm": 0.2558332085609436, "learning_rate": 3.873973794391728e-05, "loss": 0.0167, "step": 59370 }, { "grad_norm": 0.17922605574131012, "learning_rate": 3.872362861587361e-05, "loss": 0.016, "step": 59380 }, { "grad_norm": 0.15932416915893555, "learning_rate": 3.870752052099698e-05, "loss": 0.0206, "step": 59390 }, { "grad_norm": 0.21944144368171692, "learning_rate": 3.8691413661048916e-05, "loss": 0.0169, "step": 59400 }, { "grad_norm": 0.24382074177265167, "learning_rate": 3.867530803779085e-05, "loss": 0.0194, "step": 59410 }, { "grad_norm": 0.1887091100215912, "learning_rate": 3.8659203652984054e-05, "loss": 0.0164, "step": 59420 }, { "grad_norm": 0.24570561945438385, "learning_rate": 3.864310050838972e-05, "loss": 0.0149, "step": 59430 }, { "grad_norm": 0.19324049353599548, "learning_rate": 3.8626998605768835e-05, "loss": 0.0157, "step": 59440 }, { "grad_norm": 0.18026839196681976, "learning_rate": 3.8610897946882294e-05, "loss": 0.022, "step": 59450 }, { "grad_norm": 0.1631944626569748, "learning_rate": 3.859479853349082e-05, "loss": 0.0142, "step": 59460 }, { "grad_norm": 0.2381250113248825, "learning_rate": 3.857870036735503e-05, "loss": 0.0163, "step": 59470 }, { "grad_norm": 0.16123616695404053, "learning_rate": 3.856260345023538e-05, "loss": 0.0166, "step": 59480 }, { "grad_norm": 0.14016246795654297, "learning_rate": 3.854650778389224e-05, "loss": 0.0194, "step": 59490 }, { "grad_norm": 0.19742673635482788, "learning_rate": 3.8530413370085786e-05, "loss": 0.0175, "step": 59500 }, { "grad_norm": 0.18583780527114868, "learning_rate": 3.851432021057608e-05, "loss": 0.0179, "step": 59510 }, { "grad_norm": 0.2520368993282318, "learning_rate": 3.849822830712306e-05, "loss": 0.024, "step": 59520 }, { "grad_norm": 0.145398810505867, "learning_rate": 3.848213766148649e-05, "loss": 0.0199, "step": 59530 }, { "grad_norm": 0.12582941353321075, "learning_rate": 3.846604827542602e-05, "loss": 0.0164, "step": 59540 }, { "grad_norm": 0.19931870698928833, "learning_rate": 3.844996015070117e-05, "loss": 0.019, "step": 59550 }, { "grad_norm": 0.2125299870967865, "learning_rate": 3.843387328907132e-05, "loss": 0.0206, "step": 59560 }, { "grad_norm": 0.16573083400726318, "learning_rate": 3.8417787692295696e-05, "loss": 0.018, "step": 59570 }, { "grad_norm": 0.24711139500141144, "learning_rate": 3.84017033621334e-05, "loss": 0.0171, "step": 59580 }, { "grad_norm": 0.23267066478729248, "learning_rate": 3.838562030034338e-05, "loss": 0.02, "step": 59590 }, { "grad_norm": 0.22567617893218994, "learning_rate": 3.836953850868447e-05, "loss": 0.0218, "step": 59600 }, { "grad_norm": 0.5320632457733154, "learning_rate": 3.835345798891533e-05, "loss": 0.0184, "step": 59610 }, { "grad_norm": 0.20029261708259583, "learning_rate": 3.833737874279452e-05, "loss": 0.0159, "step": 59620 }, { "grad_norm": 0.1956380009651184, "learning_rate": 3.8321300772080433e-05, "loss": 0.0164, "step": 59630 }, { "grad_norm": 0.24055571854114532, "learning_rate": 3.8305224078531345e-05, "loss": 0.0188, "step": 59640 }, { "grad_norm": 0.24497917294502258, "learning_rate": 3.828914866390537e-05, "loss": 0.0182, "step": 59650 }, { "grad_norm": 0.29021379351615906, "learning_rate": 3.827307452996048e-05, "loss": 0.0208, "step": 59660 }, { "grad_norm": 0.21185871958732605, "learning_rate": 3.825700167845454e-05, "loss": 0.0166, "step": 59670 }, { "grad_norm": 0.22601288557052612, "learning_rate": 3.8240930111145214e-05, "loss": 0.0174, "step": 59680 }, { "grad_norm": 0.3025215268135071, "learning_rate": 3.8224859829790116e-05, "loss": 0.0193, "step": 59690 }, { "grad_norm": 0.2935406565666199, "learning_rate": 3.820879083614664e-05, "loss": 0.0204, "step": 59700 }, { "grad_norm": 0.19624966382980347, "learning_rate": 3.819272313197208e-05, "loss": 0.0162, "step": 59710 }, { "grad_norm": 0.2560350000858307, "learning_rate": 3.817665671902355e-05, "loss": 0.0181, "step": 59720 }, { "grad_norm": 0.19134877622127533, "learning_rate": 3.816059159905808e-05, "loss": 0.0172, "step": 59730 }, { "grad_norm": 0.19372901320457458, "learning_rate": 3.8144527773832476e-05, "loss": 0.0174, "step": 59740 }, { "grad_norm": 0.23920002579689026, "learning_rate": 3.812846524510352e-05, "loss": 0.0192, "step": 59750 }, { "grad_norm": 0.17355087399482727, "learning_rate": 3.811240401462775e-05, "loss": 0.0169, "step": 59760 }, { "grad_norm": 0.1363893747329712, "learning_rate": 3.809634408416162e-05, "loss": 0.0193, "step": 59770 }, { "grad_norm": 0.21291767060756683, "learning_rate": 3.8080285455461394e-05, "loss": 0.0162, "step": 59780 }, { "grad_norm": 0.17066974937915802, "learning_rate": 3.806422813028323e-05, "loss": 0.0172, "step": 59790 }, { "grad_norm": 0.19433428347110748, "learning_rate": 3.8048172110383114e-05, "loss": 0.0171, "step": 59800 }, { "grad_norm": 0.29813724756240845, "learning_rate": 3.8032117397516944e-05, "loss": 0.0179, "step": 59810 }, { "grad_norm": 0.1724732518196106, "learning_rate": 3.801606399344042e-05, "loss": 0.0192, "step": 59820 }, { "grad_norm": 0.18236835300922394, "learning_rate": 3.800001189990913e-05, "loss": 0.0163, "step": 59830 }, { "grad_norm": 0.18134400248527527, "learning_rate": 3.798396111867847e-05, "loss": 0.0168, "step": 59840 }, { "grad_norm": 0.1946028172969818, "learning_rate": 3.796791165150378e-05, "loss": 0.0192, "step": 59850 }, { "grad_norm": 0.6148503422737122, "learning_rate": 3.795186350014018e-05, "loss": 0.0173, "step": 59860 }, { "grad_norm": 0.17285645008087158, "learning_rate": 3.793581666634266e-05, "loss": 0.0152, "step": 59870 }, { "grad_norm": 0.3806810677051544, "learning_rate": 3.791977115186609e-05, "loss": 0.0203, "step": 59880 }, { "grad_norm": 0.1687282770872116, "learning_rate": 3.79037269584652e-05, "loss": 0.0191, "step": 59890 }, { "grad_norm": 0.17679578065872192, "learning_rate": 3.7887684087894545e-05, "loss": 0.0171, "step": 59900 }, { "grad_norm": 0.16695617139339447, "learning_rate": 3.787164254190854e-05, "loss": 0.0166, "step": 59910 }, { "grad_norm": 0.17354083061218262, "learning_rate": 3.7855602322261473e-05, "loss": 0.017, "step": 59920 }, { "grad_norm": 0.18311166763305664, "learning_rate": 3.783956343070746e-05, "loss": 0.0177, "step": 59930 }, { "grad_norm": 0.3528560400009155, "learning_rate": 3.7823525869000526e-05, "loss": 0.0169, "step": 59940 }, { "grad_norm": 0.15392763912677765, "learning_rate": 3.78074896388945e-05, "loss": 0.0167, "step": 59950 }, { "grad_norm": 0.15321482717990875, "learning_rate": 3.7791454742143075e-05, "loss": 0.0148, "step": 59960 }, { "grad_norm": 0.21868059039115906, "learning_rate": 3.7775421180499795e-05, "loss": 0.0164, "step": 59970 }, { "grad_norm": 0.25458207726478577, "learning_rate": 3.7759388955718086e-05, "loss": 0.0183, "step": 59980 }, { "grad_norm": 0.18996410071849823, "learning_rate": 3.774335806955117e-05, "loss": 0.016, "step": 59990 }, { "grad_norm": 0.15759354829788208, "learning_rate": 3.7727328523752215e-05, "loss": 0.0178, "step": 60000 }, { "grad_norm": 0.23357172310352325, "learning_rate": 3.771130032007415e-05, "loss": 0.0163, "step": 60010 }, { "grad_norm": 0.17139898240566254, "learning_rate": 3.769527346026982e-05, "loss": 0.0154, "step": 60020 }, { "grad_norm": 0.22781208157539368, "learning_rate": 3.767924794609188e-05, "loss": 0.0181, "step": 60030 }, { "grad_norm": 0.15529175102710724, "learning_rate": 3.7663223779292854e-05, "loss": 0.0211, "step": 60040 }, { "grad_norm": 0.23603662848472595, "learning_rate": 3.764720096162512e-05, "loss": 0.02, "step": 60050 }, { "grad_norm": 0.2647523283958435, "learning_rate": 3.763117949484092e-05, "loss": 0.0158, "step": 60060 }, { "grad_norm": 0.11277805268764496, "learning_rate": 3.7615159380692344e-05, "loss": 0.018, "step": 60070 }, { "grad_norm": 0.2585466802120209, "learning_rate": 3.759914062093132e-05, "loss": 0.0167, "step": 60080 }, { "grad_norm": 0.1998947262763977, "learning_rate": 3.7583123217309615e-05, "loss": 0.0198, "step": 60090 }, { "grad_norm": 0.3650646209716797, "learning_rate": 3.7567107171578904e-05, "loss": 0.0163, "step": 60100 }, { "grad_norm": 0.21449580788612366, "learning_rate": 3.755109248549066e-05, "loss": 0.019, "step": 60110 }, { "grad_norm": 0.1607198566198349, "learning_rate": 3.75350791607962e-05, "loss": 0.0202, "step": 60120 }, { "grad_norm": 0.44317764043807983, "learning_rate": 3.751906719924676e-05, "loss": 0.0183, "step": 60130 }, { "grad_norm": 0.1665552258491516, "learning_rate": 3.750305660259337e-05, "loss": 0.0148, "step": 60140 }, { "grad_norm": 0.22195172309875488, "learning_rate": 3.748704737258693e-05, "loss": 0.0183, "step": 60150 }, { "grad_norm": 0.16872839629650116, "learning_rate": 3.747103951097816e-05, "loss": 0.0153, "step": 60160 }, { "grad_norm": 0.20095670223236084, "learning_rate": 3.745503301951768e-05, "loss": 0.0181, "step": 60170 }, { "grad_norm": 0.1787799894809723, "learning_rate": 3.743902789995592e-05, "loss": 0.0147, "step": 60180 }, { "grad_norm": 0.15857161581516266, "learning_rate": 3.7423024154043183e-05, "loss": 0.0164, "step": 60190 }, { "grad_norm": 0.1941821128129959, "learning_rate": 3.740702178352963e-05, "loss": 0.0152, "step": 60200 }, { "grad_norm": 0.23825456202030182, "learning_rate": 3.739102079016523e-05, "loss": 0.0202, "step": 60210 }, { "grad_norm": 0.18879453837871552, "learning_rate": 3.7375021175699846e-05, "loss": 0.0172, "step": 60220 }, { "grad_norm": 0.19316519796848297, "learning_rate": 3.735902294188317e-05, "loss": 0.017, "step": 60230 }, { "grad_norm": 0.14965330064296722, "learning_rate": 3.734302609046473e-05, "loss": 0.0158, "step": 60240 }, { "grad_norm": 0.1643630564212799, "learning_rate": 3.732703062319391e-05, "loss": 0.016, "step": 60250 }, { "grad_norm": 0.1868685483932495, "learning_rate": 3.731103654181999e-05, "loss": 0.0165, "step": 60260 }, { "grad_norm": 0.18491412699222565, "learning_rate": 3.729504384809203e-05, "loss": 0.0159, "step": 60270 }, { "grad_norm": 0.16944803297519684, "learning_rate": 3.727905254375898e-05, "loss": 0.0226, "step": 60280 }, { "grad_norm": 0.1810637265443802, "learning_rate": 3.7263062630569614e-05, "loss": 0.015, "step": 60290 }, { "grad_norm": 0.5533208847045898, "learning_rate": 3.724707411027256e-05, "loss": 0.0171, "step": 60300 }, { "grad_norm": 0.14084070920944214, "learning_rate": 3.723108698461631e-05, "loss": 0.0143, "step": 60310 }, { "grad_norm": 0.17297713458538055, "learning_rate": 3.72151012553492e-05, "loss": 0.0186, "step": 60320 }, { "grad_norm": 0.16481637954711914, "learning_rate": 3.719911692421939e-05, "loss": 0.0164, "step": 60330 }, { "grad_norm": 0.20624007284641266, "learning_rate": 3.71831339929749e-05, "loss": 0.0266, "step": 60340 }, { "grad_norm": 0.1315927803516388, "learning_rate": 3.716715246336362e-05, "loss": 0.0167, "step": 60350 }, { "grad_norm": 0.23436357080936432, "learning_rate": 3.715117233713324e-05, "loss": 0.0188, "step": 60360 }, { "grad_norm": 0.18988655507564545, "learning_rate": 3.713519361603135e-05, "loss": 0.0165, "step": 60370 }, { "grad_norm": 0.17067652940750122, "learning_rate": 3.711921630180532e-05, "loss": 0.0187, "step": 60380 }, { "grad_norm": 0.2582125663757324, "learning_rate": 3.710324039620245e-05, "loss": 0.0201, "step": 60390 }, { "grad_norm": 0.20513410866260529, "learning_rate": 3.708726590096982e-05, "loss": 0.0181, "step": 60400 }, { "grad_norm": 0.1612696349620819, "learning_rate": 3.707129281785437e-05, "loss": 0.0184, "step": 60410 }, { "grad_norm": 0.13523949682712555, "learning_rate": 3.705532114860291e-05, "loss": 0.0178, "step": 60420 }, { "grad_norm": 0.18357880413532257, "learning_rate": 3.703935089496207e-05, "loss": 0.016, "step": 60430 }, { "grad_norm": 0.20877142250537872, "learning_rate": 3.702338205867829e-05, "loss": 0.0233, "step": 60440 }, { "grad_norm": 0.24258457124233246, "learning_rate": 3.7007414641497976e-05, "loss": 0.0178, "step": 60450 }, { "grad_norm": 0.18168577551841736, "learning_rate": 3.699144864516726e-05, "loss": 0.0142, "step": 60460 }, { "grad_norm": 0.18418225646018982, "learning_rate": 3.6975484071432153e-05, "loss": 0.0177, "step": 60470 }, { "grad_norm": 0.3554084300994873, "learning_rate": 3.695952092203854e-05, "loss": 0.0175, "step": 60480 }, { "grad_norm": 0.19078467786312103, "learning_rate": 3.69435591987321e-05, "loss": 0.0153, "step": 60490 }, { "grad_norm": 0.2048698514699936, "learning_rate": 3.6927598903258374e-05, "loss": 0.0216, "step": 60500 }, { "grad_norm": 0.1914103627204895, "learning_rate": 3.691164003736281e-05, "loss": 0.0169, "step": 60510 }, { "grad_norm": 0.24937142431735992, "learning_rate": 3.689568260279061e-05, "loss": 0.0168, "step": 60520 }, { "grad_norm": 0.13624949753284454, "learning_rate": 3.687972660128686e-05, "loss": 0.0174, "step": 60530 }, { "grad_norm": 0.21652325987815857, "learning_rate": 3.686377203459648e-05, "loss": 0.0169, "step": 60540 }, { "grad_norm": 0.17308935523033142, "learning_rate": 3.6847818904464226e-05, "loss": 0.0226, "step": 60550 }, { "grad_norm": 0.22724607586860657, "learning_rate": 3.683186721263474e-05, "loss": 0.0164, "step": 60560 }, { "grad_norm": 0.16142527759075165, "learning_rate": 3.6815916960852435e-05, "loss": 0.0173, "step": 60570 }, { "grad_norm": 0.3695622682571411, "learning_rate": 3.679996815086165e-05, "loss": 0.0165, "step": 60580 }, { "grad_norm": 0.20029553771018982, "learning_rate": 3.678402078440649e-05, "loss": 0.0173, "step": 60590 }, { "grad_norm": 0.16414985060691833, "learning_rate": 3.676807486323096e-05, "loss": 0.0173, "step": 60600 }, { "grad_norm": 0.26513057947158813, "learning_rate": 3.6752130389078864e-05, "loss": 0.0188, "step": 60610 }, { "grad_norm": 0.15741859376430511, "learning_rate": 3.673618736369388e-05, "loss": 0.0157, "step": 60620 }, { "grad_norm": 0.16414526104927063, "learning_rate": 3.6720245788819474e-05, "loss": 0.0159, "step": 60630 }, { "grad_norm": 0.19173818826675415, "learning_rate": 3.6704305666199044e-05, "loss": 0.0188, "step": 60640 }, { "grad_norm": 0.2020442932844162, "learning_rate": 3.668836699757576e-05, "loss": 0.0163, "step": 60650 }, { "grad_norm": 0.1741628795862198, "learning_rate": 3.667242978469265e-05, "loss": 0.0195, "step": 60660 }, { "grad_norm": 0.23183217644691467, "learning_rate": 3.6656494029292575e-05, "loss": 0.016, "step": 60670 }, { "grad_norm": 0.29598772525787354, "learning_rate": 3.664055973311825e-05, "loss": 0.0192, "step": 60680 }, { "grad_norm": 0.18496988713741302, "learning_rate": 3.662462689791221e-05, "loss": 0.0166, "step": 60690 }, { "grad_norm": 0.26753249764442444, "learning_rate": 3.660869552541689e-05, "loss": 0.0175, "step": 60700 }, { "grad_norm": 0.1757224202156067, "learning_rate": 3.659276561737448e-05, "loss": 0.0169, "step": 60710 }, { "grad_norm": 0.2765710651874542, "learning_rate": 3.657683717552708e-05, "loss": 0.0174, "step": 60720 }, { "grad_norm": 0.17009295523166656, "learning_rate": 3.6560910201616574e-05, "loss": 0.0202, "step": 60730 }, { "grad_norm": 0.19593314826488495, "learning_rate": 3.654498469738472e-05, "loss": 0.016, "step": 60740 }, { "grad_norm": 0.1546294242143631, "learning_rate": 3.6529060664573126e-05, "loss": 0.0145, "step": 60750 }, { "grad_norm": 0.18965207040309906, "learning_rate": 3.6513138104923176e-05, "loss": 0.0176, "step": 60760 }, { "grad_norm": 0.21798302233219147, "learning_rate": 3.6497217020176176e-05, "loss": 0.0185, "step": 60770 }, { "grad_norm": 0.24315987527370453, "learning_rate": 3.648129741207323e-05, "loss": 0.0182, "step": 60780 }, { "grad_norm": 0.1958000808954239, "learning_rate": 3.646537928235527e-05, "loss": 0.0166, "step": 60790 }, { "grad_norm": 0.18871255218982697, "learning_rate": 3.6449462632763075e-05, "loss": 0.0155, "step": 60800 }, { "grad_norm": 0.17634475231170654, "learning_rate": 3.643354746503729e-05, "loss": 0.0195, "step": 60810 }, { "grad_norm": 0.2045404314994812, "learning_rate": 3.6417633780918345e-05, "loss": 0.0155, "step": 60820 }, { "grad_norm": 0.15964087843894958, "learning_rate": 3.640172158214655e-05, "loss": 0.0177, "step": 60830 }, { "grad_norm": 0.13241222500801086, "learning_rate": 3.6385810870462035e-05, "loss": 0.0167, "step": 60840 }, { "grad_norm": 0.3170444369316101, "learning_rate": 3.636990164760479e-05, "loss": 0.0151, "step": 60850 }, { "grad_norm": 0.14563576877117157, "learning_rate": 3.6353993915314596e-05, "loss": 0.0124, "step": 60860 }, { "grad_norm": 0.19525985419750214, "learning_rate": 3.6338087675331124e-05, "loss": 0.0162, "step": 60870 }, { "grad_norm": 0.33551400899887085, "learning_rate": 3.632218292939383e-05, "loss": 0.0156, "step": 60880 }, { "grad_norm": 0.15198957920074463, "learning_rate": 3.6306279679242033e-05, "loss": 0.0156, "step": 60890 }, { "grad_norm": 0.35835132002830505, "learning_rate": 3.629037792661492e-05, "loss": 0.0184, "step": 60900 }, { "grad_norm": 0.16196966171264648, "learning_rate": 3.627447767325147e-05, "loss": 0.0173, "step": 60910 }, { "grad_norm": 0.3255457580089569, "learning_rate": 3.62585789208905e-05, "loss": 0.0179, "step": 60920 }, { "grad_norm": 0.18746314942836761, "learning_rate": 3.624268167127068e-05, "loss": 0.0164, "step": 60930 }, { "grad_norm": 0.2009943723678589, "learning_rate": 3.622678592613051e-05, "loss": 0.0194, "step": 60940 }, { "grad_norm": 0.42281079292297363, "learning_rate": 3.6210891687208304e-05, "loss": 0.02, "step": 60950 }, { "grad_norm": 0.16800962388515472, "learning_rate": 3.619499895624228e-05, "loss": 0.0173, "step": 60960 }, { "grad_norm": 0.1769821047782898, "learning_rate": 3.61791077349704e-05, "loss": 0.021, "step": 60970 }, { "grad_norm": 0.18066009879112244, "learning_rate": 3.616321802513053e-05, "loss": 0.0171, "step": 60980 }, { "grad_norm": 0.1655922681093216, "learning_rate": 3.6147329828460325e-05, "loss": 0.0185, "step": 60990 }, { "grad_norm": 0.17575031518936157, "learning_rate": 3.6131443146697306e-05, "loss": 0.0196, "step": 61000 }, { "grad_norm": 0.17607828974723816, "learning_rate": 3.611555798157879e-05, "loss": 0.0185, "step": 61010 }, { "grad_norm": 0.3497367799282074, "learning_rate": 3.6099674334842e-05, "loss": 0.0178, "step": 61020 }, { "grad_norm": 0.21443380415439606, "learning_rate": 3.6083792208223934e-05, "loss": 0.0164, "step": 61030 }, { "grad_norm": 0.22330038249492645, "learning_rate": 3.606791160346142e-05, "loss": 0.0175, "step": 61040 }, { "grad_norm": 0.19618061184883118, "learning_rate": 3.6052032522291134e-05, "loss": 0.0206, "step": 61050 }, { "grad_norm": 0.35131368041038513, "learning_rate": 3.603615496644962e-05, "loss": 0.0205, "step": 61060 }, { "grad_norm": 0.2543083727359772, "learning_rate": 3.60202789376732e-05, "loss": 0.0172, "step": 61070 }, { "grad_norm": 0.21175029873847961, "learning_rate": 3.6004404437698036e-05, "loss": 0.0175, "step": 61080 }, { "grad_norm": 0.35584568977355957, "learning_rate": 3.5988531468260184e-05, "loss": 0.0175, "step": 61090 }, { "grad_norm": 0.23235951364040375, "learning_rate": 3.597266003109546e-05, "loss": 0.0171, "step": 61100 }, { "grad_norm": 0.3667970299720764, "learning_rate": 3.5956790127939557e-05, "loss": 0.0194, "step": 61110 }, { "grad_norm": 0.21659836173057556, "learning_rate": 3.594092176052796e-05, "loss": 0.0197, "step": 61120 }, { "grad_norm": 0.19559641182422638, "learning_rate": 3.592505493059603e-05, "loss": 0.0189, "step": 61130 }, { "grad_norm": 0.24597778916358948, "learning_rate": 3.5909189639878905e-05, "loss": 0.0235, "step": 61140 }, { "grad_norm": 0.15530464053153992, "learning_rate": 3.5893325890111626e-05, "loss": 0.0165, "step": 61150 }, { "grad_norm": 0.15905693173408508, "learning_rate": 3.5877463683029034e-05, "loss": 0.0191, "step": 61160 }, { "grad_norm": 0.16725558042526245, "learning_rate": 3.586160302036578e-05, "loss": 0.0177, "step": 61170 }, { "grad_norm": 0.19998131692409515, "learning_rate": 3.584574390385636e-05, "loss": 0.018, "step": 61180 }, { "grad_norm": 0.2742610573768616, "learning_rate": 3.582988633523511e-05, "loss": 0.0176, "step": 61190 }, { "grad_norm": 0.21780149638652802, "learning_rate": 3.581403031623616e-05, "loss": 0.016, "step": 61200 }, { "grad_norm": 0.16664013266563416, "learning_rate": 3.579817584859354e-05, "loss": 0.0174, "step": 61210 }, { "grad_norm": 0.7336285710334778, "learning_rate": 3.578232293404107e-05, "loss": 0.0177, "step": 61220 }, { "grad_norm": 0.15960213541984558, "learning_rate": 3.5766471574312385e-05, "loss": 0.0163, "step": 61230 }, { "grad_norm": 0.20022626221179962, "learning_rate": 3.5750621771140964e-05, "loss": 0.0213, "step": 61240 }, { "grad_norm": 0.18982747197151184, "learning_rate": 3.5734773526260115e-05, "loss": 0.0147, "step": 61250 }, { "grad_norm": 0.49878832697868347, "learning_rate": 3.5718926841402996e-05, "loss": 0.0195, "step": 61260 }, { "grad_norm": 0.14878566563129425, "learning_rate": 3.5703081718302555e-05, "loss": 0.02, "step": 61270 }, { "grad_norm": 0.308169424533844, "learning_rate": 3.568723815869162e-05, "loss": 0.0174, "step": 61280 }, { "grad_norm": 0.23629207909107208, "learning_rate": 3.5671396164302777e-05, "loss": 0.0178, "step": 61290 }, { "grad_norm": 0.17376579344272614, "learning_rate": 3.565555573686853e-05, "loss": 0.0167, "step": 61300 }, { "grad_norm": 0.1907602846622467, "learning_rate": 3.563971687812113e-05, "loss": 0.0213, "step": 61310 }, { "grad_norm": 0.27918320894241333, "learning_rate": 3.5623879589792706e-05, "loss": 0.0145, "step": 61320 }, { "grad_norm": 0.14867722988128662, "learning_rate": 3.5608043873615185e-05, "loss": 0.018, "step": 61330 }, { "grad_norm": 0.2196950614452362, "learning_rate": 3.559220973132035e-05, "loss": 0.0169, "step": 61340 }, { "grad_norm": 0.20361945033073425, "learning_rate": 3.557637716463981e-05, "loss": 0.0173, "step": 61350 }, { "grad_norm": 0.21957699954509735, "learning_rate": 3.556054617530497e-05, "loss": 0.017, "step": 61360 }, { "grad_norm": 0.19037370383739471, "learning_rate": 3.5544716765047106e-05, "loss": 0.0194, "step": 61370 }, { "grad_norm": 0.16258494555950165, "learning_rate": 3.5528888935597274e-05, "loss": 0.0162, "step": 61380 }, { "grad_norm": 0.16651715338230133, "learning_rate": 3.551306268868636e-05, "loss": 0.0149, "step": 61390 }, { "grad_norm": 0.1771157830953598, "learning_rate": 3.5497238026045166e-05, "loss": 0.0163, "step": 61400 }, { "grad_norm": 0.227992981672287, "learning_rate": 3.548141494940421e-05, "loss": 0.0205, "step": 61410 }, { "grad_norm": 0.19555532932281494, "learning_rate": 3.546559346049388e-05, "loss": 0.0165, "step": 61420 }, { "grad_norm": 0.29213228821754456, "learning_rate": 3.544977356104441e-05, "loss": 0.0181, "step": 61430 }, { "grad_norm": 0.16057001054286957, "learning_rate": 3.543395525278582e-05, "loss": 0.015, "step": 61440 }, { "grad_norm": 0.27805501222610474, "learning_rate": 3.5418138537447985e-05, "loss": 0.0188, "step": 61450 }, { "grad_norm": 0.16390404105186462, "learning_rate": 3.5402323416760574e-05, "loss": 0.0155, "step": 61460 }, { "grad_norm": 0.16555216908454895, "learning_rate": 3.538650989245316e-05, "loss": 0.0161, "step": 61470 }, { "grad_norm": 0.12201264500617981, "learning_rate": 3.537069796625504e-05, "loss": 0.0164, "step": 61480 }, { "grad_norm": 0.15169568359851837, "learning_rate": 3.53548876398954e-05, "loss": 0.012, "step": 61490 }, { "grad_norm": 0.16099049150943756, "learning_rate": 3.5339078915103214e-05, "loss": 0.0132, "step": 61500 }, { "grad_norm": 0.21198956668376923, "learning_rate": 3.532327179360733e-05, "loss": 0.017, "step": 61510 }, { "grad_norm": 0.19934259355068207, "learning_rate": 3.530746627713636e-05, "loss": 0.0157, "step": 61520 }, { "grad_norm": 0.2136206179857254, "learning_rate": 3.5291662367418805e-05, "loss": 0.0146, "step": 61530 }, { "grad_norm": 0.1760205626487732, "learning_rate": 3.527586006618293e-05, "loss": 0.0204, "step": 61540 }, { "grad_norm": 0.21025808155536652, "learning_rate": 3.526005937515687e-05, "loss": 0.0185, "step": 61550 }, { "grad_norm": 0.16094718873500824, "learning_rate": 3.524426029606856e-05, "loss": 0.0166, "step": 61560 }, { "grad_norm": 0.1490977555513382, "learning_rate": 3.5228462830645746e-05, "loss": 0.0136, "step": 61570 }, { "grad_norm": 0.19390419125556946, "learning_rate": 3.5212666980616036e-05, "loss": 0.0157, "step": 61580 }, { "grad_norm": 0.13123579323291779, "learning_rate": 3.519687274770681e-05, "loss": 0.0152, "step": 61590 }, { "grad_norm": 0.13688507676124573, "learning_rate": 3.518108013364535e-05, "loss": 0.0173, "step": 61600 }, { "grad_norm": 0.20216251909732819, "learning_rate": 3.516528914015868e-05, "loss": 0.0181, "step": 61610 }, { "grad_norm": 0.16518250107765198, "learning_rate": 3.5149499768973694e-05, "loss": 0.0161, "step": 61620 }, { "grad_norm": 0.2190086394548416, "learning_rate": 3.513371202181708e-05, "loss": 0.015, "step": 61630 }, { "grad_norm": 0.18299362063407898, "learning_rate": 3.511792590041537e-05, "loss": 0.0207, "step": 61640 }, { "grad_norm": 0.2013285756111145, "learning_rate": 3.510214140649488e-05, "loss": 0.0154, "step": 61650 }, { "grad_norm": 0.2822975814342499, "learning_rate": 3.508635854178184e-05, "loss": 0.0182, "step": 61660 }, { "grad_norm": 0.1515417993068695, "learning_rate": 3.50705773080022e-05, "loss": 0.0153, "step": 61670 }, { "grad_norm": 0.3459039330482483, "learning_rate": 3.505479770688178e-05, "loss": 0.0155, "step": 61680 }, { "grad_norm": 0.16100198030471802, "learning_rate": 3.503901974014622e-05, "loss": 0.0175, "step": 61690 }, { "grad_norm": 0.14567239582538605, "learning_rate": 3.502324340952095e-05, "loss": 0.0185, "step": 61700 }, { "grad_norm": 0.3165971040725708, "learning_rate": 3.500746871673126e-05, "loss": 0.018, "step": 61710 }, { "grad_norm": 0.11455876380205154, "learning_rate": 3.499169566350227e-05, "loss": 0.0166, "step": 61720 }, { "grad_norm": 0.1527026891708374, "learning_rate": 3.4975924251558876e-05, "loss": 0.0179, "step": 61730 }, { "grad_norm": 0.1875806599855423, "learning_rate": 3.496015448262582e-05, "loss": 0.0217, "step": 61740 }, { "grad_norm": 0.16966967284679413, "learning_rate": 3.494438635842764e-05, "loss": 0.0163, "step": 61750 }, { "grad_norm": 0.1474917232990265, "learning_rate": 3.492861988068874e-05, "loss": 0.0203, "step": 61760 }, { "grad_norm": 0.25722938776016235, "learning_rate": 3.491285505113332e-05, "loss": 0.0178, "step": 61770 }, { "grad_norm": 0.3672100901603699, "learning_rate": 3.4897091871485376e-05, "loss": 0.0171, "step": 61780 }, { "grad_norm": 0.2626239061355591, "learning_rate": 3.488133034346875e-05, "loss": 0.0164, "step": 61790 }, { "grad_norm": 0.1610945463180542, "learning_rate": 3.486557046880713e-05, "loss": 0.0161, "step": 61800 }, { "grad_norm": 0.16521821916103363, "learning_rate": 3.484981224922395e-05, "loss": 0.0147, "step": 61810 }, { "grad_norm": 0.27572906017303467, "learning_rate": 3.4834055686442544e-05, "loss": 0.0149, "step": 61820 }, { "grad_norm": 0.16997434198856354, "learning_rate": 3.4818300782186e-05, "loss": 0.0173, "step": 61830 }, { "grad_norm": 0.13399638235569, "learning_rate": 3.4802547538177236e-05, "loss": 0.0187, "step": 61840 }, { "grad_norm": 0.15865935385227203, "learning_rate": 3.478679595613905e-05, "loss": 0.0197, "step": 61850 }, { "grad_norm": 0.2436951994895935, "learning_rate": 3.4771046037793975e-05, "loss": 0.0173, "step": 61860 }, { "grad_norm": 0.18551675975322723, "learning_rate": 3.4755297784864416e-05, "loss": 0.0147, "step": 61870 }, { "grad_norm": 0.1872299164533615, "learning_rate": 3.473955119907257e-05, "loss": 0.0157, "step": 61880 }, { "grad_norm": 0.25843679904937744, "learning_rate": 3.472380628214046e-05, "loss": 0.0138, "step": 61890 }, { "grad_norm": 0.17871582508087158, "learning_rate": 3.4708063035789906e-05, "loss": 0.0197, "step": 61900 }, { "grad_norm": 0.15578706562519073, "learning_rate": 3.4692321461742614e-05, "loss": 0.0178, "step": 61910 }, { "grad_norm": 0.1931641846895218, "learning_rate": 3.4676581561720023e-05, "loss": 0.0167, "step": 61920 }, { "grad_norm": 0.19982068240642548, "learning_rate": 3.466084333744344e-05, "loss": 0.0152, "step": 61930 }, { "grad_norm": 0.1842798888683319, "learning_rate": 3.4645106790633964e-05, "loss": 0.017, "step": 61940 }, { "grad_norm": 0.1457556188106537, "learning_rate": 3.4629371923012526e-05, "loss": 0.0146, "step": 61950 }, { "grad_norm": 0.20333102345466614, "learning_rate": 3.4613638736299856e-05, "loss": 0.0167, "step": 61960 }, { "grad_norm": 0.23382329940795898, "learning_rate": 3.459790723221652e-05, "loss": 0.0153, "step": 61970 }, { "grad_norm": 0.23516488075256348, "learning_rate": 3.45821774124829e-05, "loss": 0.0183, "step": 61980 }, { "grad_norm": 0.1703512817621231, "learning_rate": 3.456644927881918e-05, "loss": 0.0155, "step": 61990 }, { "grad_norm": 0.17081409692764282, "learning_rate": 3.4550722832945354e-05, "loss": 0.0205, "step": 62000 }, { "grad_norm": 0.19992195069789886, "learning_rate": 3.453499807658126e-05, "loss": 0.018, "step": 62010 }, { "grad_norm": 0.2221055030822754, "learning_rate": 3.451927501144653e-05, "loss": 0.0159, "step": 62020 }, { "grad_norm": 0.251335471868515, "learning_rate": 3.450355363926059e-05, "loss": 0.0167, "step": 62030 }, { "grad_norm": 0.16472654044628143, "learning_rate": 3.448783396174273e-05, "loss": 0.0183, "step": 62040 }, { "grad_norm": 0.2709249258041382, "learning_rate": 3.4472115980612034e-05, "loss": 0.0146, "step": 62050 }, { "grad_norm": 0.17065133154392242, "learning_rate": 3.4456399697587396e-05, "loss": 0.0142, "step": 62060 }, { "grad_norm": 0.19840000569820404, "learning_rate": 3.4440685114387506e-05, "loss": 0.0174, "step": 62070 }, { "grad_norm": 0.18788915872573853, "learning_rate": 3.442497223273091e-05, "loss": 0.0164, "step": 62080 }, { "grad_norm": 0.1448509246110916, "learning_rate": 3.4409261054335926e-05, "loss": 0.017, "step": 62090 }, { "grad_norm": 0.22715649008750916, "learning_rate": 3.439355158092069e-05, "loss": 0.0163, "step": 62100 }, { "grad_norm": 0.32760509848594666, "learning_rate": 3.437784381420322e-05, "loss": 0.0228, "step": 62110 }, { "grad_norm": 0.20149695873260498, "learning_rate": 3.436213775590126e-05, "loss": 0.0131, "step": 62120 }, { "grad_norm": 0.15756095945835114, "learning_rate": 3.4346433407732405e-05, "loss": 0.0174, "step": 62130 }, { "grad_norm": 0.17939452826976776, "learning_rate": 3.433073077141405e-05, "loss": 0.0153, "step": 62140 }, { "grad_norm": 0.19918009638786316, "learning_rate": 3.431502984866343e-05, "loss": 0.022, "step": 62150 }, { "grad_norm": 0.17207972705364227, "learning_rate": 3.429933064119753e-05, "loss": 0.014, "step": 62160 }, { "grad_norm": 0.22972775995731354, "learning_rate": 3.4283633150733255e-05, "loss": 0.0151, "step": 62170 }, { "grad_norm": 0.1803485006093979, "learning_rate": 3.4267937378987225e-05, "loss": 0.0133, "step": 62180 }, { "grad_norm": 0.1973763406276703, "learning_rate": 3.425224332767592e-05, "loss": 0.0161, "step": 62190 }, { "grad_norm": 0.16974149644374847, "learning_rate": 3.4236550998515595e-05, "loss": 0.016, "step": 62200 }, { "grad_norm": 0.21956531703472137, "learning_rate": 3.422086039322235e-05, "loss": 0.0157, "step": 62210 }, { "grad_norm": 0.17712078988552094, "learning_rate": 3.420517151351209e-05, "loss": 0.0166, "step": 62220 }, { "grad_norm": 0.18455390632152557, "learning_rate": 3.418948436110053e-05, "loss": 0.0165, "step": 62230 }, { "grad_norm": 0.2800106108188629, "learning_rate": 3.417379893770319e-05, "loss": 0.0163, "step": 62240 }, { "grad_norm": 0.19397687911987305, "learning_rate": 3.41581152450354e-05, "loss": 0.0162, "step": 62250 }, { "grad_norm": 0.19725219905376434, "learning_rate": 3.414243328481231e-05, "loss": 0.0155, "step": 62260 }, { "grad_norm": 0.1845681369304657, "learning_rate": 3.412675305874888e-05, "loss": 0.0141, "step": 62270 }, { "grad_norm": 0.2048349529504776, "learning_rate": 3.4111074568559876e-05, "loss": 0.0171, "step": 62280 }, { "grad_norm": 0.22559285163879395, "learning_rate": 3.4095397815959846e-05, "loss": 0.016, "step": 62290 }, { "grad_norm": 0.24510708451271057, "learning_rate": 3.407972280266322e-05, "loss": 0.0169, "step": 62300 }, { "grad_norm": 0.22164276242256165, "learning_rate": 3.4064049530384166e-05, "loss": 0.0149, "step": 62310 }, { "grad_norm": 0.21848423779010773, "learning_rate": 3.40483780008367e-05, "loss": 0.0194, "step": 62320 }, { "grad_norm": 0.16582170128822327, "learning_rate": 3.4032708215734635e-05, "loss": 0.0155, "step": 62330 }, { "grad_norm": 0.21488913893699646, "learning_rate": 3.4017040176791584e-05, "loss": 0.0126, "step": 62340 }, { "grad_norm": 0.19500769674777985, "learning_rate": 3.4001373885720976e-05, "loss": 0.0162, "step": 62350 }, { "grad_norm": 0.18575862050056458, "learning_rate": 3.398570934423608e-05, "loss": 0.0197, "step": 62360 }, { "grad_norm": 0.21074986457824707, "learning_rate": 3.397004655404994e-05, "loss": 0.018, "step": 62370 }, { "grad_norm": 0.29472091794013977, "learning_rate": 3.395438551687541e-05, "loss": 0.0188, "step": 62380 }, { "grad_norm": 0.2600346505641937, "learning_rate": 3.393872623442515e-05, "loss": 0.0162, "step": 62390 }, { "grad_norm": 0.29998844861984253, "learning_rate": 3.392306870841164e-05, "loss": 0.0164, "step": 62400 }, { "grad_norm": 0.179230198264122, "learning_rate": 3.390741294054715e-05, "loss": 0.0165, "step": 62410 }, { "grad_norm": 0.2175447642803192, "learning_rate": 3.38917589325438e-05, "loss": 0.0161, "step": 62420 }, { "grad_norm": 0.17891821265220642, "learning_rate": 3.387610668611349e-05, "loss": 0.0146, "step": 62430 }, { "grad_norm": 0.2517333924770355, "learning_rate": 3.38604562029679e-05, "loss": 0.0191, "step": 62440 }, { "grad_norm": 0.2177516222000122, "learning_rate": 3.384480748481855e-05, "loss": 0.0171, "step": 62450 }, { "grad_norm": 0.15300357341766357, "learning_rate": 3.382916053337676e-05, "loss": 0.0153, "step": 62460 }, { "grad_norm": 0.1644083708524704, "learning_rate": 3.381351535035366e-05, "loss": 0.0197, "step": 62470 }, { "grad_norm": 0.1691470593214035, "learning_rate": 3.379787193746019e-05, "loss": 0.0178, "step": 62480 }, { "grad_norm": 0.1459549516439438, "learning_rate": 3.378223029640709e-05, "loss": 0.016, "step": 62490 }, { "grad_norm": 0.5114812850952148, "learning_rate": 3.376659042890489e-05, "loss": 0.0201, "step": 62500 }, { "grad_norm": 0.18702332675457, "learning_rate": 3.375095233666397e-05, "loss": 0.0165, "step": 62510 }, { "grad_norm": 0.22024452686309814, "learning_rate": 3.373531602139447e-05, "loss": 0.0154, "step": 62520 }, { "grad_norm": 0.1601121425628662, "learning_rate": 3.371968148480635e-05, "loss": 0.0152, "step": 62530 }, { "grad_norm": 0.20123600959777832, "learning_rate": 3.370404872860937e-05, "loss": 0.0196, "step": 62540 }, { "grad_norm": 0.19899730384349823, "learning_rate": 3.3688417754513145e-05, "loss": 0.0163, "step": 62550 }, { "grad_norm": 0.15847502648830414, "learning_rate": 3.367278856422703e-05, "loss": 0.0207, "step": 62560 }, { "grad_norm": 0.341484010219574, "learning_rate": 3.365716115946021e-05, "loss": 0.0204, "step": 62570 }, { "grad_norm": 0.20620906352996826, "learning_rate": 3.3641535541921666e-05, "loss": 0.0184, "step": 62580 }, { "grad_norm": 0.18749722838401794, "learning_rate": 3.36259117133202e-05, "loss": 0.0169, "step": 62590 }, { "grad_norm": 0.2977403998374939, "learning_rate": 3.36102896753644e-05, "loss": 0.0199, "step": 62600 }, { "grad_norm": 0.16901062428951263, "learning_rate": 3.35946694297627e-05, "loss": 0.0184, "step": 62610 }, { "grad_norm": 0.3605242669582367, "learning_rate": 3.357905097822328e-05, "loss": 0.0148, "step": 62620 }, { "grad_norm": 0.18318848311901093, "learning_rate": 3.356343432245416e-05, "loss": 0.0145, "step": 62630 }, { "grad_norm": 0.19511835277080536, "learning_rate": 3.354781946416315e-05, "loss": 0.0171, "step": 62640 }, { "grad_norm": 0.3197980225086212, "learning_rate": 3.3532206405057864e-05, "loss": 0.017, "step": 62650 }, { "grad_norm": 0.23056471347808838, "learning_rate": 3.351659514684573e-05, "loss": 0.0157, "step": 62660 }, { "grad_norm": 0.21047091484069824, "learning_rate": 3.350098569123395e-05, "loss": 0.0167, "step": 62670 }, { "grad_norm": 0.15572209656238556, "learning_rate": 3.348537803992958e-05, "loss": 0.0172, "step": 62680 }, { "grad_norm": 0.15621024370193481, "learning_rate": 3.346977219463945e-05, "loss": 0.0196, "step": 62690 }, { "grad_norm": 0.18160267174243927, "learning_rate": 3.345416815707017e-05, "loss": 0.0167, "step": 62700 }, { "grad_norm": 0.20548012852668762, "learning_rate": 3.3438565928928185e-05, "loss": 0.0187, "step": 62710 }, { "grad_norm": 0.16116496920585632, "learning_rate": 3.3422965511919736e-05, "loss": 0.0167, "step": 62720 }, { "grad_norm": 0.3107767105102539, "learning_rate": 3.3407366907750846e-05, "loss": 0.0174, "step": 62730 }, { "grad_norm": 0.13339823484420776, "learning_rate": 3.339177011812737e-05, "loss": 0.0231, "step": 62740 }, { "grad_norm": 0.2020874321460724, "learning_rate": 3.3376175144754935e-05, "loss": 0.018, "step": 62750 }, { "grad_norm": 0.17767873406410217, "learning_rate": 3.3360581989339004e-05, "loss": 0.017, "step": 62760 }, { "grad_norm": 0.2296893149614334, "learning_rate": 3.3344990653584804e-05, "loss": 0.0182, "step": 62770 }, { "grad_norm": 0.18770970404148102, "learning_rate": 3.332940113919739e-05, "loss": 0.0188, "step": 62780 }, { "grad_norm": 0.21424055099487305, "learning_rate": 3.331381344788161e-05, "loss": 0.0174, "step": 62790 }, { "grad_norm": 0.21065691113471985, "learning_rate": 3.3298227581342076e-05, "loss": 0.0138, "step": 62800 }, { "grad_norm": 0.22248190641403198, "learning_rate": 3.328264354128329e-05, "loss": 0.015, "step": 62810 }, { "grad_norm": 0.22153498232364655, "learning_rate": 3.3267061329409465e-05, "loss": 0.0181, "step": 62820 }, { "grad_norm": 0.3106788098812103, "learning_rate": 3.325148094742466e-05, "loss": 0.0158, "step": 62830 }, { "grad_norm": 0.20891371369361877, "learning_rate": 3.323590239703271e-05, "loss": 0.0185, "step": 62840 }, { "grad_norm": 0.15230432152748108, "learning_rate": 3.322032567993728e-05, "loss": 0.016, "step": 62850 }, { "grad_norm": 0.1715526133775711, "learning_rate": 3.3204750797841774e-05, "loss": 0.0152, "step": 62860 }, { "grad_norm": 0.1794060319662094, "learning_rate": 3.31891777524495e-05, "loss": 0.0175, "step": 62870 }, { "grad_norm": 0.16537347435951233, "learning_rate": 3.317360654546347e-05, "loss": 0.0183, "step": 62880 }, { "grad_norm": 0.12478180974721909, "learning_rate": 3.3158037178586534e-05, "loss": 0.0168, "step": 62890 }, { "grad_norm": 0.17431019246578217, "learning_rate": 3.3142469653521316e-05, "loss": 0.0144, "step": 62900 }, { "grad_norm": 0.23653830587863922, "learning_rate": 3.3126903971970284e-05, "loss": 0.0144, "step": 62910 }, { "grad_norm": 0.24108192324638367, "learning_rate": 3.311134013563564e-05, "loss": 0.0186, "step": 62920 }, { "grad_norm": 0.19255174696445465, "learning_rate": 3.309577814621947e-05, "loss": 0.015, "step": 62930 }, { "grad_norm": 0.20934812724590302, "learning_rate": 3.308021800542359e-05, "loss": 0.0148, "step": 62940 }, { "grad_norm": 0.20236332714557648, "learning_rate": 3.306465971494963e-05, "loss": 0.0171, "step": 62950 }, { "grad_norm": 0.19674202799797058, "learning_rate": 3.304910327649901e-05, "loss": 0.0178, "step": 62960 }, { "grad_norm": 0.11835667490959167, "learning_rate": 3.303354869177297e-05, "loss": 0.0176, "step": 62970 }, { "grad_norm": 0.17790313065052032, "learning_rate": 3.301799596247255e-05, "loss": 0.017, "step": 62980 }, { "grad_norm": 0.17354784905910492, "learning_rate": 3.300244509029854e-05, "loss": 0.018, "step": 62990 }, { "grad_norm": 0.20835739374160767, "learning_rate": 3.2986896076951585e-05, "loss": 0.0182, "step": 63000 }, { "grad_norm": 0.24227328598499298, "learning_rate": 3.297134892413211e-05, "loss": 0.0148, "step": 63010 }, { "grad_norm": 0.37535855174064636, "learning_rate": 3.295580363354031e-05, "loss": 0.0174, "step": 63020 }, { "grad_norm": 0.15177255868911743, "learning_rate": 3.2940260206876194e-05, "loss": 0.0165, "step": 63030 }, { "grad_norm": 0.11679279804229736, "learning_rate": 3.292471864583958e-05, "loss": 0.0151, "step": 63040 }, { "grad_norm": 0.11948675662279129, "learning_rate": 3.290917895213004e-05, "loss": 0.0152, "step": 63050 }, { "grad_norm": 0.17638280987739563, "learning_rate": 3.289364112744701e-05, "loss": 0.0173, "step": 63060 }, { "grad_norm": 0.1971834897994995, "learning_rate": 3.287810517348967e-05, "loss": 0.0188, "step": 63070 }, { "grad_norm": 0.1838873326778412, "learning_rate": 3.2862571091957014e-05, "loss": 0.0169, "step": 63080 }, { "grad_norm": 0.2760923206806183, "learning_rate": 3.284703888454781e-05, "loss": 0.0202, "step": 63090 }, { "grad_norm": 0.2111276537179947, "learning_rate": 3.2831508552960646e-05, "loss": 0.0179, "step": 63100 }, { "grad_norm": 0.1777258813381195, "learning_rate": 3.281598009889387e-05, "loss": 0.0171, "step": 63110 }, { "grad_norm": 0.1684689223766327, "learning_rate": 3.28004535240457e-05, "loss": 0.0152, "step": 63120 }, { "grad_norm": 0.1529151052236557, "learning_rate": 3.278492883011407e-05, "loss": 0.0134, "step": 63130 }, { "grad_norm": 0.1894540637731552, "learning_rate": 3.276940601879675e-05, "loss": 0.0142, "step": 63140 }, { "grad_norm": 0.15209224820137024, "learning_rate": 3.275388509179127e-05, "loss": 0.0155, "step": 63150 }, { "grad_norm": 0.14942698180675507, "learning_rate": 3.2738366050794985e-05, "loss": 0.0137, "step": 63160 }, { "grad_norm": 0.19396653771400452, "learning_rate": 3.2722848897505034e-05, "loss": 0.0166, "step": 63170 }, { "grad_norm": 0.16810956597328186, "learning_rate": 3.2707333633618356e-05, "loss": 0.014, "step": 63180 }, { "grad_norm": 0.145737886428833, "learning_rate": 3.269182026083167e-05, "loss": 0.0157, "step": 63190 }, { "grad_norm": 0.2617458999156952, "learning_rate": 3.2676308780841516e-05, "loss": 0.0161, "step": 63200 }, { "grad_norm": 0.23499101400375366, "learning_rate": 3.266079919534416e-05, "loss": 0.0177, "step": 63210 }, { "grad_norm": 0.23104563355445862, "learning_rate": 3.264529150603575e-05, "loss": 0.0143, "step": 63220 }, { "grad_norm": 0.396010160446167, "learning_rate": 3.262978571461217e-05, "loss": 0.0191, "step": 63230 }, { "grad_norm": 0.40776365995407104, "learning_rate": 3.26142818227691e-05, "loss": 0.0192, "step": 63240 }, { "grad_norm": 0.2261154055595398, "learning_rate": 3.259877983220203e-05, "loss": 0.014, "step": 63250 }, { "grad_norm": 0.15588800609111786, "learning_rate": 3.2583279744606246e-05, "loss": 0.0125, "step": 63260 }, { "grad_norm": 0.2283121794462204, "learning_rate": 3.25677815616768e-05, "loss": 0.017, "step": 63270 }, { "grad_norm": 0.21273541450500488, "learning_rate": 3.255228528510856e-05, "loss": 0.018, "step": 63280 }, { "grad_norm": 0.20199573040008545, "learning_rate": 3.253679091659617e-05, "loss": 0.0161, "step": 63290 }, { "grad_norm": 0.15324261784553528, "learning_rate": 3.252129845783404e-05, "loss": 0.0164, "step": 63300 }, { "grad_norm": 0.17985694110393524, "learning_rate": 3.2505807910516456e-05, "loss": 0.0178, "step": 63310 }, { "grad_norm": 0.2567107379436493, "learning_rate": 3.249031927633742e-05, "loss": 0.0197, "step": 63320 }, { "grad_norm": 0.19362066686153412, "learning_rate": 3.2474832556990737e-05, "loss": 0.0147, "step": 63330 }, { "grad_norm": 0.25314295291900635, "learning_rate": 3.245934775417002e-05, "loss": 0.0177, "step": 63340 }, { "grad_norm": 0.33924275636672974, "learning_rate": 3.244386486956866e-05, "loss": 0.0199, "step": 63350 }, { "grad_norm": 0.29077741503715515, "learning_rate": 3.2428383904879846e-05, "loss": 0.0192, "step": 63360 }, { "grad_norm": 0.16006341576576233, "learning_rate": 3.241290486179652e-05, "loss": 0.0149, "step": 63370 }, { "grad_norm": 0.27504003047943115, "learning_rate": 3.239742774201151e-05, "loss": 0.0152, "step": 63380 }, { "grad_norm": 0.2094273716211319, "learning_rate": 3.238195254721733e-05, "loss": 0.0187, "step": 63390 }, { "grad_norm": 0.1668829321861267, "learning_rate": 3.236647927910633e-05, "loss": 0.0172, "step": 63400 }, { "grad_norm": 0.18151332437992096, "learning_rate": 3.2351007939370646e-05, "loss": 0.0171, "step": 63410 }, { "grad_norm": 0.1881369799375534, "learning_rate": 3.2335538529702184e-05, "loss": 0.0154, "step": 63420 }, { "grad_norm": 0.1568305492401123, "learning_rate": 3.232007105179268e-05, "loss": 0.0165, "step": 63430 }, { "grad_norm": 0.1818186193704605, "learning_rate": 3.230460550733363e-05, "loss": 0.0145, "step": 63440 }, { "grad_norm": 0.20371846854686737, "learning_rate": 3.2289141898016315e-05, "loss": 0.0135, "step": 63450 }, { "grad_norm": 0.21210512518882751, "learning_rate": 3.227368022553181e-05, "loss": 0.0145, "step": 63460 }, { "grad_norm": 0.18494786322116852, "learning_rate": 3.225822049157099e-05, "loss": 0.0151, "step": 63470 }, { "grad_norm": 0.2928085923194885, "learning_rate": 3.2242762697824494e-05, "loss": 0.0169, "step": 63480 }, { "grad_norm": 0.22331885993480682, "learning_rate": 3.222730684598279e-05, "loss": 0.0166, "step": 63490 }, { "grad_norm": 0.14230583608150482, "learning_rate": 3.221185293773605e-05, "loss": 0.017, "step": 63500 }, { "grad_norm": 0.13720542192459106, "learning_rate": 3.2196400974774354e-05, "loss": 0.0182, "step": 63510 }, { "grad_norm": 0.15398164093494415, "learning_rate": 3.2180950958787485e-05, "loss": 0.0156, "step": 63520 }, { "grad_norm": 0.17680372297763824, "learning_rate": 3.2165502891465024e-05, "loss": 0.0157, "step": 63530 }, { "grad_norm": 0.29444554448127747, "learning_rate": 3.215005677449636e-05, "loss": 0.0191, "step": 63540 }, { "grad_norm": 0.16759610176086426, "learning_rate": 3.2134612609570644e-05, "loss": 0.0176, "step": 63550 }, { "grad_norm": 0.17575380206108093, "learning_rate": 3.211917039837682e-05, "loss": 0.0177, "step": 63560 }, { "grad_norm": 0.1978292614221573, "learning_rate": 3.210373014260365e-05, "loss": 0.0154, "step": 63570 }, { "grad_norm": 0.21980410814285278, "learning_rate": 3.208829184393965e-05, "loss": 0.0173, "step": 63580 }, { "grad_norm": 0.20865295827388763, "learning_rate": 3.207285550407312e-05, "loss": 0.0188, "step": 63590 }, { "grad_norm": 0.3666374385356903, "learning_rate": 3.205742112469217e-05, "loss": 0.0164, "step": 63600 }, { "grad_norm": 0.21843424439430237, "learning_rate": 3.204198870748465e-05, "loss": 0.0176, "step": 63610 }, { "grad_norm": 0.19309227168560028, "learning_rate": 3.202655825413824e-05, "loss": 0.0147, "step": 63620 }, { "grad_norm": 0.3531096577644348, "learning_rate": 3.2011129766340416e-05, "loss": 0.0172, "step": 63630 }, { "grad_norm": 0.20731621980667114, "learning_rate": 3.1995703245778384e-05, "loss": 0.0156, "step": 63640 }, { "grad_norm": 0.20111730694770813, "learning_rate": 3.1980278694139186e-05, "loss": 0.02, "step": 63650 }, { "grad_norm": 0.11664839833974838, "learning_rate": 3.196485611310962e-05, "loss": 0.0123, "step": 63660 }, { "grad_norm": 0.5019282698631287, "learning_rate": 3.1949435504376257e-05, "loss": 0.0184, "step": 63670 }, { "grad_norm": 0.20105652511119843, "learning_rate": 3.1934016869625505e-05, "loss": 0.0172, "step": 63680 }, { "grad_norm": 0.7845024466514587, "learning_rate": 3.191860021054349e-05, "loss": 0.0159, "step": 63690 }, { "grad_norm": 0.5165940523147583, "learning_rate": 3.190318552881618e-05, "loss": 0.0174, "step": 63700 }, { "grad_norm": 0.19757455587387085, "learning_rate": 3.1887772826129285e-05, "loss": 0.0167, "step": 63710 }, { "grad_norm": 0.16966867446899414, "learning_rate": 3.1872362104168335e-05, "loss": 0.015, "step": 63720 }, { "grad_norm": 0.21566995978355408, "learning_rate": 3.185695336461861e-05, "loss": 0.0139, "step": 63730 }, { "grad_norm": 0.20078660547733307, "learning_rate": 3.184154660916519e-05, "loss": 0.0207, "step": 63740 }, { "grad_norm": 0.2359900176525116, "learning_rate": 3.1826141839492906e-05, "loss": 0.0151, "step": 63750 }, { "grad_norm": 0.22357098758220673, "learning_rate": 3.181073905728644e-05, "loss": 0.0172, "step": 63760 }, { "grad_norm": 0.2119474560022354, "learning_rate": 3.1795338264230214e-05, "loss": 0.0153, "step": 63770 }, { "grad_norm": 0.19352558255195618, "learning_rate": 3.1779939462008426e-05, "loss": 0.0153, "step": 63780 }, { "grad_norm": 0.17985664308071136, "learning_rate": 3.1764542652305066e-05, "loss": 0.0183, "step": 63790 }, { "grad_norm": 0.31512418389320374, "learning_rate": 3.17491478368039e-05, "loss": 0.0176, "step": 63800 }, { "grad_norm": 0.1537390947341919, "learning_rate": 3.173375501718847e-05, "loss": 0.0155, "step": 63810 }, { "grad_norm": 0.17469194531440735, "learning_rate": 3.171836419514215e-05, "loss": 0.0172, "step": 63820 }, { "grad_norm": 0.167324498295784, "learning_rate": 3.170297537234803e-05, "loss": 0.0146, "step": 63830 }, { "grad_norm": 0.22142542898654938, "learning_rate": 3.168758855048901e-05, "loss": 0.0149, "step": 63840 }, { "grad_norm": 0.2765560448169708, "learning_rate": 3.1672203731247785e-05, "loss": 0.0145, "step": 63850 }, { "grad_norm": 0.16522502899169922, "learning_rate": 3.16568209163068e-05, "loss": 0.0161, "step": 63860 }, { "grad_norm": 0.20300622284412384, "learning_rate": 3.1641440107348295e-05, "loss": 0.0175, "step": 63870 }, { "grad_norm": 0.21006833016872406, "learning_rate": 3.162606130605429e-05, "loss": 0.0191, "step": 63880 }, { "grad_norm": 0.22484157979488373, "learning_rate": 3.16106845141066e-05, "loss": 0.0171, "step": 63890 }, { "grad_norm": 0.25505131483078003, "learning_rate": 3.159530973318682e-05, "loss": 0.0193, "step": 63900 }, { "grad_norm": 0.20809830725193024, "learning_rate": 3.1579936964976286e-05, "loss": 0.0215, "step": 63910 }, { "grad_norm": 0.14619998633861542, "learning_rate": 3.156456621115615e-05, "loss": 0.0158, "step": 63920 }, { "grad_norm": 0.2001958042383194, "learning_rate": 3.154919747340734e-05, "loss": 0.0156, "step": 63930 }, { "grad_norm": 0.17511367797851562, "learning_rate": 3.153383075341054e-05, "loss": 0.0156, "step": 63940 }, { "grad_norm": 0.19135312736034393, "learning_rate": 3.1518466052846264e-05, "loss": 0.0183, "step": 63950 }, { "grad_norm": 0.20903217792510986, "learning_rate": 3.150310337339474e-05, "loss": 0.0171, "step": 63960 }, { "grad_norm": 0.17892573773860931, "learning_rate": 3.148774271673603e-05, "loss": 0.0131, "step": 63970 }, { "grad_norm": 0.25609490275382996, "learning_rate": 3.147238408454995e-05, "loss": 0.0151, "step": 63980 }, { "grad_norm": 0.18531976640224457, "learning_rate": 3.1457027478516086e-05, "loss": 0.015, "step": 63990 }, { "grad_norm": 0.15153276920318604, "learning_rate": 3.144167290031382e-05, "loss": 0.0157, "step": 64000 }, { "grad_norm": 0.2039397954940796, "learning_rate": 3.142632035162227e-05, "loss": 0.0162, "step": 64010 }, { "grad_norm": 0.16404955089092255, "learning_rate": 3.141096983412044e-05, "loss": 0.0168, "step": 64020 }, { "grad_norm": 0.3186003565788269, "learning_rate": 3.139562134948699e-05, "loss": 0.0171, "step": 64030 }, { "grad_norm": 0.18707050383090973, "learning_rate": 3.1380274899400424e-05, "loss": 0.0178, "step": 64040 }, { "grad_norm": 0.5974363684654236, "learning_rate": 3.1364930485539e-05, "loss": 0.0181, "step": 64050 }, { "grad_norm": 0.2608981430530548, "learning_rate": 3.134958810958076e-05, "loss": 0.0154, "step": 64060 }, { "grad_norm": 0.15527428686618805, "learning_rate": 3.13342477732035e-05, "loss": 0.018, "step": 64070 }, { "grad_norm": 0.218560591340065, "learning_rate": 3.131890947808487e-05, "loss": 0.0177, "step": 64080 }, { "grad_norm": 0.19744576513767242, "learning_rate": 3.13035732259022e-05, "loss": 0.018, "step": 64090 }, { "grad_norm": 0.15587006509304047, "learning_rate": 3.1288239018332656e-05, "loss": 0.0135, "step": 64100 }, { "grad_norm": 0.22260966897010803, "learning_rate": 3.1272906857053164e-05, "loss": 0.0186, "step": 64110 }, { "grad_norm": 0.20183490216732025, "learning_rate": 3.125757674374042e-05, "loss": 0.0141, "step": 64120 }, { "grad_norm": 0.17229017615318298, "learning_rate": 3.124224868007088e-05, "loss": 0.013, "step": 64130 }, { "grad_norm": 0.2155783623456955, "learning_rate": 3.1226922667720854e-05, "loss": 0.014, "step": 64140 }, { "grad_norm": 0.18606522679328918, "learning_rate": 3.121159870836633e-05, "loss": 0.0164, "step": 64150 }, { "grad_norm": 0.25111478567123413, "learning_rate": 3.119627680368312e-05, "loss": 0.0199, "step": 64160 }, { "grad_norm": 0.15033189952373505, "learning_rate": 3.1180956955346816e-05, "loss": 0.0158, "step": 64170 }, { "grad_norm": 0.8875977993011475, "learning_rate": 3.116563916503276e-05, "loss": 0.0153, "step": 64180 }, { "grad_norm": 0.21568205952644348, "learning_rate": 3.1150323434416095e-05, "loss": 0.0142, "step": 64190 }, { "grad_norm": 0.1493932604789734, "learning_rate": 3.1135009765171696e-05, "loss": 0.0147, "step": 64200 }, { "grad_norm": 0.15313853323459625, "learning_rate": 3.111969815897429e-05, "loss": 0.0169, "step": 64210 }, { "grad_norm": 0.19872739911079407, "learning_rate": 3.1104388617498295e-05, "loss": 0.0168, "step": 64220 }, { "grad_norm": 0.16926012933254242, "learning_rate": 3.1089081142417966e-05, "loss": 0.0163, "step": 64230 }, { "grad_norm": 0.14978310465812683, "learning_rate": 3.1073775735407285e-05, "loss": 0.0185, "step": 64240 }, { "grad_norm": 0.22507920861244202, "learning_rate": 3.1058472398140035e-05, "loss": 0.0177, "step": 64250 }, { "grad_norm": 0.16884659230709076, "learning_rate": 3.1043171132289736e-05, "loss": 0.0151, "step": 64260 }, { "grad_norm": 0.2181120067834854, "learning_rate": 3.1027871939529774e-05, "loss": 0.0246, "step": 64270 }, { "grad_norm": 0.2105741798877716, "learning_rate": 3.1012574821533214e-05, "loss": 0.0146, "step": 64280 }, { "grad_norm": 0.16471347212791443, "learning_rate": 3.099727977997291e-05, "loss": 0.0195, "step": 64290 }, { "grad_norm": 0.16817834973335266, "learning_rate": 3.098198681652154e-05, "loss": 0.0147, "step": 64300 }, { "grad_norm": 0.21480107307434082, "learning_rate": 3.0966695932851475e-05, "loss": 0.0162, "step": 64310 }, { "grad_norm": 0.12358029931783676, "learning_rate": 3.0951407130634926e-05, "loss": 0.0169, "step": 64320 }, { "grad_norm": 0.1936781257390976, "learning_rate": 3.0936120411543864e-05, "loss": 0.013, "step": 64330 }, { "grad_norm": 0.18534260988235474, "learning_rate": 3.0920835777250015e-05, "loss": 0.0166, "step": 64340 }, { "grad_norm": 0.20205111801624298, "learning_rate": 3.090555322942488e-05, "loss": 0.0132, "step": 64350 }, { "grad_norm": 0.2204243391752243, "learning_rate": 3.089027276973973e-05, "loss": 0.0169, "step": 64360 }, { "grad_norm": 0.17507164180278778, "learning_rate": 3.0874994399865615e-05, "loss": 0.0142, "step": 64370 }, { "grad_norm": 0.19560499489307404, "learning_rate": 3.085971812147337e-05, "loss": 0.0137, "step": 64380 }, { "grad_norm": 0.10965145379304886, "learning_rate": 3.0844443936233557e-05, "loss": 0.0161, "step": 64390 }, { "grad_norm": 0.1850929707288742, "learning_rate": 3.082917184581657e-05, "loss": 0.0164, "step": 64400 }, { "grad_norm": 0.13707394897937775, "learning_rate": 3.0813901851892513e-05, "loss": 0.02, "step": 64410 }, { "grad_norm": 0.15079300105571747, "learning_rate": 3.0798633956131316e-05, "loss": 0.0186, "step": 64420 }, { "grad_norm": 0.20495674014091492, "learning_rate": 3.078336816020264e-05, "loss": 0.0161, "step": 64430 }, { "grad_norm": 0.2258724719285965, "learning_rate": 3.0768104465775935e-05, "loss": 0.0166, "step": 64440 }, { "grad_norm": 0.4000144898891449, "learning_rate": 3.0752842874520396e-05, "loss": 0.0147, "step": 64450 }, { "grad_norm": 0.3070231080055237, "learning_rate": 3.073758338810503e-05, "loss": 0.0174, "step": 64460 }, { "grad_norm": 0.18310192227363586, "learning_rate": 3.072232600819859e-05, "loss": 0.0156, "step": 64470 }, { "grad_norm": 0.20817407965660095, "learning_rate": 3.070707073646959e-05, "loss": 0.0177, "step": 64480 }, { "grad_norm": 0.11981800198554993, "learning_rate": 3.069181757458633e-05, "loss": 0.0127, "step": 64490 }, { "grad_norm": 0.2099969983100891, "learning_rate": 3.067656652421687e-05, "loss": 0.0175, "step": 64500 }, { "grad_norm": 0.24073633551597595, "learning_rate": 3.0661317587029025e-05, "loss": 0.0174, "step": 64510 }, { "grad_norm": 0.16847991943359375, "learning_rate": 3.064607076469043e-05, "loss": 0.0157, "step": 64520 }, { "grad_norm": 0.22236286103725433, "learning_rate": 3.063082605886845e-05, "loss": 0.0228, "step": 64530 }, { "grad_norm": 0.20967519283294678, "learning_rate": 3.061558347123019e-05, "loss": 0.0158, "step": 64540 }, { "grad_norm": 0.17525483667850494, "learning_rate": 3.060034300344259e-05, "loss": 0.0179, "step": 64550 }, { "grad_norm": 0.19489876925945282, "learning_rate": 3.058510465717231e-05, "loss": 0.0179, "step": 64560 }, { "grad_norm": 0.17180243134498596, "learning_rate": 3.0569868434085804e-05, "loss": 0.0158, "step": 64570 }, { "grad_norm": 0.17742282152175903, "learning_rate": 3.055463433584924e-05, "loss": 0.0151, "step": 64580 }, { "grad_norm": 0.22489148378372192, "learning_rate": 3.0539402364128656e-05, "loss": 0.0163, "step": 64590 }, { "grad_norm": 0.2830894887447357, "learning_rate": 3.052417252058977e-05, "loss": 0.018, "step": 64600 }, { "grad_norm": 0.16349707543849945, "learning_rate": 3.0508944806898092e-05, "loss": 0.0138, "step": 64610 }, { "grad_norm": 0.23858508467674255, "learning_rate": 3.04937192247189e-05, "loss": 0.0178, "step": 64620 }, { "grad_norm": 0.13259689509868622, "learning_rate": 3.047849577571725e-05, "loss": 0.0134, "step": 64630 }, { "grad_norm": 0.3590901494026184, "learning_rate": 3.046327446155795e-05, "loss": 0.0174, "step": 64640 }, { "grad_norm": 0.1590566486120224, "learning_rate": 3.0448055283905587e-05, "loss": 0.0156, "step": 64650 }, { "grad_norm": 0.14394830167293549, "learning_rate": 3.0432838244424494e-05, "loss": 0.016, "step": 64660 }, { "grad_norm": 0.3208976089954376, "learning_rate": 3.04176233447788e-05, "loss": 0.0158, "step": 64670 }, { "grad_norm": 0.21567189693450928, "learning_rate": 3.040241058663238e-05, "loss": 0.0166, "step": 64680 }, { "grad_norm": 0.19977885484695435, "learning_rate": 3.038719997164887e-05, "loss": 0.0157, "step": 64690 }, { "grad_norm": 0.20070557296276093, "learning_rate": 3.0371991501491683e-05, "loss": 0.014, "step": 64700 }, { "grad_norm": 0.3812842071056366, "learning_rate": 3.0356785177823975e-05, "loss": 0.0188, "step": 64710 }, { "grad_norm": 0.19442301988601685, "learning_rate": 3.0341581002308726e-05, "loss": 0.0168, "step": 64720 }, { "grad_norm": 0.17927558720111847, "learning_rate": 3.0326378976608627e-05, "loss": 0.0161, "step": 64730 }, { "grad_norm": 0.15611812472343445, "learning_rate": 3.0311179102386146e-05, "loss": 0.0161, "step": 64740 }, { "grad_norm": 0.1485220193862915, "learning_rate": 3.0295981381303513e-05, "loss": 0.0132, "step": 64750 }, { "grad_norm": 0.1848532110452652, "learning_rate": 3.0280785815022727e-05, "loss": 0.0151, "step": 64760 }, { "grad_norm": 0.28983229398727417, "learning_rate": 3.0265592405205544e-05, "loss": 0.019, "step": 64770 }, { "grad_norm": 0.2218097299337387, "learning_rate": 3.0250401153513517e-05, "loss": 0.0167, "step": 64780 }, { "grad_norm": 0.15738293528556824, "learning_rate": 3.0235212061607936e-05, "loss": 0.017, "step": 64790 }, { "grad_norm": 0.17512084543704987, "learning_rate": 3.0220025131149842e-05, "loss": 0.0168, "step": 64800 }, { "grad_norm": 0.1977098286151886, "learning_rate": 3.020484036380007e-05, "loss": 0.0202, "step": 64810 }, { "grad_norm": 0.15304802358150482, "learning_rate": 3.0189657761219188e-05, "loss": 0.0157, "step": 64820 }, { "grad_norm": 0.16942311823368073, "learning_rate": 3.017447732506753e-05, "loss": 0.0153, "step": 64830 }, { "grad_norm": 0.28063076734542847, "learning_rate": 3.015929905700524e-05, "loss": 0.0194, "step": 64840 }, { "grad_norm": 0.16413964331150055, "learning_rate": 3.014412295869218e-05, "loss": 0.0158, "step": 64850 }, { "grad_norm": 0.16655953228473663, "learning_rate": 3.0128949031787977e-05, "loss": 0.0168, "step": 64860 }, { "grad_norm": 0.2113906294107437, "learning_rate": 3.0113777277952022e-05, "loss": 0.0141, "step": 64870 }, { "grad_norm": 0.1752091646194458, "learning_rate": 3.0098607698843496e-05, "loss": 0.0132, "step": 64880 }, { "grad_norm": 0.13192054629325867, "learning_rate": 3.0083440296121312e-05, "loss": 0.0164, "step": 64890 }, { "grad_norm": 0.14697346091270447, "learning_rate": 3.0068275071444142e-05, "loss": 0.014, "step": 64900 }, { "grad_norm": 0.1535046100616455, "learning_rate": 3.0053112026470442e-05, "loss": 0.0173, "step": 64910 }, { "grad_norm": 0.14570632576942444, "learning_rate": 3.0037951162858434e-05, "loss": 0.015, "step": 64920 }, { "grad_norm": 0.17827482521533966, "learning_rate": 3.002279248226607e-05, "loss": 0.0149, "step": 64930 }, { "grad_norm": 0.2154865860939026, "learning_rate": 3.000763598635109e-05, "loss": 0.0171, "step": 64940 }, { "grad_norm": 0.22550511360168457, "learning_rate": 2.999248167677098e-05, "loss": 0.0179, "step": 64950 }, { "grad_norm": 0.1434812992811203, "learning_rate": 2.9977329555182976e-05, "loss": 0.0164, "step": 64960 }, { "grad_norm": 0.15407924354076385, "learning_rate": 2.9962179623244124e-05, "loss": 0.0148, "step": 64970 }, { "grad_norm": 0.22837647795677185, "learning_rate": 2.9947031882611175e-05, "loss": 0.0153, "step": 64980 }, { "grad_norm": 0.1529594361782074, "learning_rate": 2.9931886334940685e-05, "loss": 0.0157, "step": 64990 }, { "grad_norm": 0.21761588752269745, "learning_rate": 2.9916742981888923e-05, "loss": 0.0163, "step": 65000 }, { "grad_norm": 0.16824373602867126, "learning_rate": 2.990160182511196e-05, "loss": 0.0143, "step": 65010 }, { "grad_norm": 0.20121881365776062, "learning_rate": 2.9886462866265574e-05, "loss": 0.0157, "step": 65020 }, { "grad_norm": 0.1814948469400406, "learning_rate": 2.9871326107005392e-05, "loss": 0.0155, "step": 65030 }, { "grad_norm": 0.29414528608322144, "learning_rate": 2.9856191548986735e-05, "loss": 0.0166, "step": 65040 }, { "grad_norm": 0.18458272516727448, "learning_rate": 2.9841059193864672e-05, "loss": 0.0157, "step": 65050 }, { "grad_norm": 0.18007144331932068, "learning_rate": 2.982592904329407e-05, "loss": 0.0141, "step": 65060 }, { "grad_norm": 0.20345579087734222, "learning_rate": 2.981080109892954e-05, "loss": 0.0216, "step": 65070 }, { "grad_norm": 0.226904034614563, "learning_rate": 2.9795675362425434e-05, "loss": 0.0174, "step": 65080 }, { "grad_norm": 0.3197908401489258, "learning_rate": 2.97805518354359e-05, "loss": 0.0176, "step": 65090 }, { "grad_norm": 0.1901017427444458, "learning_rate": 2.9765430519614825e-05, "loss": 0.0169, "step": 65100 }, { "grad_norm": 0.21113638579845428, "learning_rate": 2.975031141661584e-05, "loss": 0.0146, "step": 65110 }, { "grad_norm": 0.12191587686538696, "learning_rate": 2.9735194528092347e-05, "loss": 0.0127, "step": 65120 }, { "grad_norm": 0.20054471492767334, "learning_rate": 2.9720079855697525e-05, "loss": 0.0156, "step": 65130 }, { "grad_norm": 0.20050813257694244, "learning_rate": 2.970496740108427e-05, "loss": 0.0162, "step": 65140 }, { "grad_norm": 0.5809703469276428, "learning_rate": 2.968985716590527e-05, "loss": 0.0162, "step": 65150 }, { "grad_norm": 0.24096354842185974, "learning_rate": 2.9674749151812953e-05, "loss": 0.0179, "step": 65160 }, { "grad_norm": 0.16855469346046448, "learning_rate": 2.965964336045951e-05, "loss": 0.0191, "step": 65170 }, { "grad_norm": 0.3357580304145813, "learning_rate": 2.9644539793496894e-05, "loss": 0.0184, "step": 65180 }, { "grad_norm": 0.43965962529182434, "learning_rate": 2.9629438452576807e-05, "loss": 0.0196, "step": 65190 }, { "grad_norm": 0.22669433057308197, "learning_rate": 2.9614339339350698e-05, "loss": 0.0174, "step": 65200 }, { "grad_norm": 0.14590178430080414, "learning_rate": 2.959924245546979e-05, "loss": 0.0129, "step": 65210 }, { "grad_norm": 0.2141789346933365, "learning_rate": 2.9584147802585038e-05, "loss": 0.0134, "step": 65220 }, { "grad_norm": 0.17567850649356842, "learning_rate": 2.956905538234721e-05, "loss": 0.0163, "step": 65230 }, { "grad_norm": 0.16068941354751587, "learning_rate": 2.9553965196406764e-05, "loss": 0.0135, "step": 65240 }, { "grad_norm": 0.25686442852020264, "learning_rate": 2.9538877246413943e-05, "loss": 0.0154, "step": 65250 }, { "grad_norm": 0.11026354879140854, "learning_rate": 2.952379153401874e-05, "loss": 0.0169, "step": 65260 }, { "grad_norm": 0.1952710747718811, "learning_rate": 2.9508708060870914e-05, "loss": 0.015, "step": 65270 }, { "grad_norm": 0.16448187828063965, "learning_rate": 2.9493626828619948e-05, "loss": 0.0155, "step": 65280 }, { "grad_norm": 0.21798840165138245, "learning_rate": 2.9478547838915137e-05, "loss": 0.0131, "step": 65290 }, { "grad_norm": 0.22778262197971344, "learning_rate": 2.946347109340548e-05, "loss": 0.0157, "step": 65300 }, { "grad_norm": 0.21572333574295044, "learning_rate": 2.944839659373976e-05, "loss": 0.0177, "step": 65310 }, { "grad_norm": 0.2179124355316162, "learning_rate": 2.943332434156647e-05, "loss": 0.0182, "step": 65320 }, { "grad_norm": 0.16765572130680084, "learning_rate": 2.9418254338533912e-05, "loss": 0.0199, "step": 65330 }, { "grad_norm": 0.17862224578857422, "learning_rate": 2.940318658629011e-05, "loss": 0.0154, "step": 65340 }, { "grad_norm": 0.11354763060808182, "learning_rate": 2.938812108648287e-05, "loss": 0.0151, "step": 65350 }, { "grad_norm": 0.2753131091594696, "learning_rate": 2.937305784075972e-05, "loss": 0.0154, "step": 65360 }, { "grad_norm": 0.17395460605621338, "learning_rate": 2.9357996850767937e-05, "loss": 0.0204, "step": 65370 }, { "grad_norm": 0.16580043733119965, "learning_rate": 2.9342938118154595e-05, "loss": 0.0166, "step": 65380 }, { "grad_norm": 0.1686129868030548, "learning_rate": 2.9327881644566485e-05, "loss": 0.0137, "step": 65390 }, { "grad_norm": 0.20537447929382324, "learning_rate": 2.9312827431650157e-05, "loss": 0.0176, "step": 65400 }, { "grad_norm": 0.16621731221675873, "learning_rate": 2.92977754810519e-05, "loss": 0.0143, "step": 65410 }, { "grad_norm": 0.17619779706001282, "learning_rate": 2.928272579441781e-05, "loss": 0.0148, "step": 65420 }, { "grad_norm": 0.18488891422748566, "learning_rate": 2.9267678373393676e-05, "loss": 0.0165, "step": 65430 }, { "grad_norm": 0.3535255491733551, "learning_rate": 2.925263321962507e-05, "loss": 0.0151, "step": 65440 }, { "grad_norm": 0.213094562292099, "learning_rate": 2.9237590334757292e-05, "loss": 0.0154, "step": 65450 }, { "grad_norm": 0.3027937412261963, "learning_rate": 2.922254972043543e-05, "loss": 0.0153, "step": 65460 }, { "grad_norm": 0.17898687720298767, "learning_rate": 2.9207511378304265e-05, "loss": 0.0137, "step": 65470 }, { "grad_norm": 0.2399054765701294, "learning_rate": 2.9192475310008428e-05, "loss": 0.0128, "step": 65480 }, { "grad_norm": 0.18179140985012054, "learning_rate": 2.917744151719219e-05, "loss": 0.0163, "step": 65490 }, { "grad_norm": 0.2265082597732544, "learning_rate": 2.916241000149964e-05, "loss": 0.0128, "step": 65500 }, { "grad_norm": 0.20897352695465088, "learning_rate": 2.9147380764574622e-05, "loss": 0.0158, "step": 65510 }, { "grad_norm": 0.25178882479667664, "learning_rate": 2.9132353808060674e-05, "loss": 0.0159, "step": 65520 }, { "grad_norm": 0.17991086840629578, "learning_rate": 2.9117329133601134e-05, "loss": 0.015, "step": 65530 }, { "grad_norm": 0.16291412711143494, "learning_rate": 2.9102306742839093e-05, "loss": 0.0136, "step": 65540 }, { "grad_norm": 0.21082262694835663, "learning_rate": 2.9087286637417375e-05, "loss": 0.017, "step": 65550 }, { "grad_norm": 0.15866245329380035, "learning_rate": 2.9072268818978537e-05, "loss": 0.0149, "step": 65560 }, { "grad_norm": 0.18342557549476624, "learning_rate": 2.905725328916493e-05, "loss": 0.0147, "step": 65570 }, { "grad_norm": 0.2517344653606415, "learning_rate": 2.9042240049618608e-05, "loss": 0.0144, "step": 65580 }, { "grad_norm": 0.302321195602417, "learning_rate": 2.9027229101981413e-05, "loss": 0.0155, "step": 65590 }, { "grad_norm": 0.17211076617240906, "learning_rate": 2.901222044789488e-05, "loss": 0.0184, "step": 65600 }, { "grad_norm": 0.1905418485403061, "learning_rate": 2.89972140890004e-05, "loss": 0.0143, "step": 65610 }, { "grad_norm": 0.22849875688552856, "learning_rate": 2.8982210026938988e-05, "loss": 0.0151, "step": 65620 }, { "grad_norm": 0.13096770644187927, "learning_rate": 2.896720826335151e-05, "loss": 0.0164, "step": 65630 }, { "grad_norm": 0.13274915516376495, "learning_rate": 2.8952208799878493e-05, "loss": 0.0124, "step": 65640 }, { "grad_norm": 0.1639413982629776, "learning_rate": 2.89372116381603e-05, "loss": 0.0167, "step": 65650 }, { "grad_norm": 0.1546882838010788, "learning_rate": 2.8922216779836932e-05, "loss": 0.0161, "step": 65660 }, { "grad_norm": 0.16233748197555542, "learning_rate": 2.8907224226548285e-05, "loss": 0.0159, "step": 65670 }, { "grad_norm": 0.18587350845336914, "learning_rate": 2.8892233979933868e-05, "loss": 0.0148, "step": 65680 }, { "grad_norm": 0.18030941486358643, "learning_rate": 2.8877246041633026e-05, "loss": 0.0196, "step": 65690 }, { "grad_norm": 0.1661321073770523, "learning_rate": 2.886226041328478e-05, "loss": 0.0169, "step": 65700 }, { "grad_norm": 0.14799565076828003, "learning_rate": 2.8847277096527957e-05, "loss": 0.0146, "step": 65710 }, { "grad_norm": 0.18156343698501587, "learning_rate": 2.8832296093001105e-05, "loss": 0.0159, "step": 65720 }, { "grad_norm": 0.28537583351135254, "learning_rate": 2.8817317404342547e-05, "loss": 0.0194, "step": 65730 }, { "grad_norm": 0.23194696009159088, "learning_rate": 2.880234103219029e-05, "loss": 0.0145, "step": 65740 }, { "grad_norm": 0.38748615980148315, "learning_rate": 2.878736697818215e-05, "loss": 0.0177, "step": 65750 }, { "grad_norm": 0.17001669108867645, "learning_rate": 2.877239524395568e-05, "loss": 0.0149, "step": 65760 }, { "grad_norm": 0.21504150331020355, "learning_rate": 2.8757425831148134e-05, "loss": 0.0148, "step": 65770 }, { "grad_norm": 0.10300613194704056, "learning_rate": 2.874245874139658e-05, "loss": 0.0158, "step": 65780 }, { "grad_norm": 0.5278159379959106, "learning_rate": 2.8727493976337737e-05, "loss": 0.015, "step": 65790 }, { "grad_norm": 0.1509362906217575, "learning_rate": 2.8712531537608218e-05, "loss": 0.0138, "step": 65800 }, { "grad_norm": 0.18403255939483643, "learning_rate": 2.8697571426844217e-05, "loss": 0.0147, "step": 65810 }, { "grad_norm": 0.16838164627552032, "learning_rate": 2.86826136456818e-05, "loss": 0.0163, "step": 65820 }, { "grad_norm": 0.239451065659523, "learning_rate": 2.8667658195756685e-05, "loss": 0.0142, "step": 65830 }, { "grad_norm": 0.16421644389629364, "learning_rate": 2.8652705078704412e-05, "loss": 0.0151, "step": 65840 }, { "grad_norm": 0.1654471904039383, "learning_rate": 2.863775429616018e-05, "loss": 0.0159, "step": 65850 }, { "grad_norm": 0.2525123953819275, "learning_rate": 2.862280584975906e-05, "loss": 0.0143, "step": 65860 }, { "grad_norm": 0.1567484587430954, "learning_rate": 2.8607859741135735e-05, "loss": 0.0174, "step": 65870 }, { "grad_norm": 0.21776653826236725, "learning_rate": 2.8592915971924723e-05, "loss": 0.0215, "step": 65880 }, { "grad_norm": 0.22623085975646973, "learning_rate": 2.8577974543760212e-05, "loss": 0.0168, "step": 65890 }, { "grad_norm": 0.22308874130249023, "learning_rate": 2.8563035458276222e-05, "loss": 0.0198, "step": 65900 }, { "grad_norm": 0.17513824999332428, "learning_rate": 2.8548098717106424e-05, "loss": 0.0155, "step": 65910 }, { "grad_norm": 0.20082546770572662, "learning_rate": 2.85331643218843e-05, "loss": 0.015, "step": 65920 }, { "grad_norm": 0.18487133085727692, "learning_rate": 2.8518232274243046e-05, "loss": 0.0146, "step": 65930 }, { "grad_norm": 0.2541685104370117, "learning_rate": 2.8503302575815638e-05, "loss": 0.0159, "step": 65940 }, { "grad_norm": 0.21490459144115448, "learning_rate": 2.8488375228234716e-05, "loss": 0.0179, "step": 65950 }, { "grad_norm": 0.33287733793258667, "learning_rate": 2.8473450233132738e-05, "loss": 0.0135, "step": 65960 }, { "grad_norm": 0.2337348312139511, "learning_rate": 2.8458527592141898e-05, "loss": 0.0195, "step": 65970 }, { "grad_norm": 0.202619269490242, "learning_rate": 2.8443607306894073e-05, "loss": 0.0174, "step": 65980 }, { "grad_norm": 0.18580226600170135, "learning_rate": 2.8428689379020946e-05, "loss": 0.0174, "step": 65990 }, { "grad_norm": 0.34566980600357056, "learning_rate": 2.841377381015391e-05, "loss": 0.0168, "step": 66000 }, { "grad_norm": 0.20469559729099274, "learning_rate": 2.8398860601924144e-05, "loss": 0.0155, "step": 66010 }, { "grad_norm": 0.39918237924575806, "learning_rate": 2.838394975596248e-05, "loss": 0.0177, "step": 66020 }, { "grad_norm": 0.21410369873046875, "learning_rate": 2.83690412738996e-05, "loss": 0.0168, "step": 66030 }, { "grad_norm": 0.12357798963785172, "learning_rate": 2.835413515736579e-05, "loss": 0.0157, "step": 66040 }, { "grad_norm": 0.21152810752391815, "learning_rate": 2.8339231407991267e-05, "loss": 0.019, "step": 66050 }, { "grad_norm": 0.1876419484615326, "learning_rate": 2.832433002740581e-05, "loss": 0.0133, "step": 66060 }, { "grad_norm": 0.16259093582630157, "learning_rate": 2.8309431017239047e-05, "loss": 0.0185, "step": 66070 }, { "grad_norm": 0.16201817989349365, "learning_rate": 2.829453437912028e-05, "loss": 0.0148, "step": 66080 }, { "grad_norm": 0.18240006268024445, "learning_rate": 2.8279640114678617e-05, "loss": 0.0165, "step": 66090 }, { "grad_norm": 0.19830834865570068, "learning_rate": 2.8264748225542836e-05, "loss": 0.0146, "step": 66100 }, { "grad_norm": 0.1842334121465683, "learning_rate": 2.8249858713341516e-05, "loss": 0.0183, "step": 66110 }, { "grad_norm": 0.16078384220600128, "learning_rate": 2.823497157970294e-05, "loss": 0.0155, "step": 66120 }, { "grad_norm": 0.21620292961597443, "learning_rate": 2.8220086826255164e-05, "loss": 0.0165, "step": 66130 }, { "grad_norm": 0.165835440158844, "learning_rate": 2.8205204454625934e-05, "loss": 0.0145, "step": 66140 }, { "grad_norm": 0.2221686840057373, "learning_rate": 2.8190324466442798e-05, "loss": 0.0153, "step": 66150 }, { "grad_norm": 0.4067649841308594, "learning_rate": 2.8175446863332965e-05, "loss": 0.0223, "step": 66160 }, { "grad_norm": 0.15240591764450073, "learning_rate": 2.816057164692345e-05, "loss": 0.0189, "step": 66170 }, { "grad_norm": 0.23201189935207367, "learning_rate": 2.8145698818840992e-05, "loss": 0.0184, "step": 66180 }, { "grad_norm": 0.19481799006462097, "learning_rate": 2.8130828380712065e-05, "loss": 0.0172, "step": 66190 }, { "grad_norm": 0.16310560703277588, "learning_rate": 2.811596033416285e-05, "loss": 0.0153, "step": 66200 }, { "grad_norm": 0.5344452261924744, "learning_rate": 2.8101094680819307e-05, "loss": 0.015, "step": 66210 }, { "grad_norm": 0.13440941274166107, "learning_rate": 2.808623142230715e-05, "loss": 0.0134, "step": 66220 }, { "grad_norm": 0.2638828456401825, "learning_rate": 2.8071370560251757e-05, "loss": 0.0162, "step": 66230 }, { "grad_norm": 0.15258903801441193, "learning_rate": 2.805651209627831e-05, "loss": 0.015, "step": 66240 }, { "grad_norm": 0.29329124093055725, "learning_rate": 2.804165603201171e-05, "loss": 0.0147, "step": 66250 }, { "grad_norm": 0.15900538861751556, "learning_rate": 2.802680236907661e-05, "loss": 0.0164, "step": 66260 }, { "grad_norm": 0.15932439267635345, "learning_rate": 2.8011951109097344e-05, "loss": 0.0145, "step": 66270 }, { "grad_norm": 0.2267443835735321, "learning_rate": 2.7997102253698072e-05, "loss": 0.0183, "step": 66280 }, { "grad_norm": 0.15096557140350342, "learning_rate": 2.7982255804502588e-05, "loss": 0.0167, "step": 66290 }, { "grad_norm": 0.18564875423908234, "learning_rate": 2.7967411763134516e-05, "loss": 0.0168, "step": 66300 }, { "grad_norm": 0.15690962970256805, "learning_rate": 2.7952570131217166e-05, "loss": 0.0138, "step": 66310 }, { "grad_norm": 0.24928100407123566, "learning_rate": 2.7937730910373618e-05, "loss": 0.0214, "step": 66320 }, { "grad_norm": 0.2004648745059967, "learning_rate": 2.7922894102226628e-05, "loss": 0.0151, "step": 66330 }, { "grad_norm": 0.29224586486816406, "learning_rate": 2.790805970839876e-05, "loss": 0.0156, "step": 66340 }, { "grad_norm": 0.16141894459724426, "learning_rate": 2.789322773051225e-05, "loss": 0.0165, "step": 66350 }, { "grad_norm": 0.13134942948818207, "learning_rate": 2.7878398170189124e-05, "loss": 0.0192, "step": 66360 }, { "grad_norm": 0.1709665209054947, "learning_rate": 2.7863571029051112e-05, "loss": 0.0146, "step": 66370 }, { "grad_norm": 0.18550029397010803, "learning_rate": 2.784874630871971e-05, "loss": 0.0142, "step": 66380 }, { "grad_norm": 0.18133926391601562, "learning_rate": 2.7833924010816086e-05, "loss": 0.0173, "step": 66390 }, { "grad_norm": 0.2022271752357483, "learning_rate": 2.7819104136961226e-05, "loss": 0.0185, "step": 66400 }, { "grad_norm": 0.16569407284259796, "learning_rate": 2.780428668877577e-05, "loss": 0.019, "step": 66410 }, { "grad_norm": 0.18382826447486877, "learning_rate": 2.7789471667880152e-05, "loss": 0.016, "step": 66420 }, { "grad_norm": 0.21951903402805328, "learning_rate": 2.7774659075894523e-05, "loss": 0.0152, "step": 66430 }, { "grad_norm": 0.18521648645401, "learning_rate": 2.7759848914438778e-05, "loss": 0.0141, "step": 66440 }, { "grad_norm": 0.4546058177947998, "learning_rate": 2.7745041185132496e-05, "loss": 0.0173, "step": 66450 }, { "grad_norm": 0.18249541521072388, "learning_rate": 2.773023588959506e-05, "loss": 0.0139, "step": 66460 }, { "grad_norm": 0.17820759117603302, "learning_rate": 2.7715433029445555e-05, "loss": 0.0173, "step": 66470 }, { "grad_norm": 0.16426992416381836, "learning_rate": 2.7700632606302766e-05, "loss": 0.0162, "step": 66480 }, { "grad_norm": 0.21966244280338287, "learning_rate": 2.768583462178528e-05, "loss": 0.0145, "step": 66490 }, { "grad_norm": 0.1904207020998001, "learning_rate": 2.767103907751136e-05, "loss": 0.0162, "step": 66500 }, { "grad_norm": 0.15126706659793854, "learning_rate": 2.7656245975099065e-05, "loss": 0.0194, "step": 66510 }, { "grad_norm": 0.1383824497461319, "learning_rate": 2.7641455316166093e-05, "loss": 0.0144, "step": 66520 }, { "grad_norm": 0.09933829307556152, "learning_rate": 2.7626667102329973e-05, "loss": 0.0131, "step": 66530 }, { "grad_norm": 0.14443457126617432, "learning_rate": 2.761188133520788e-05, "loss": 0.0188, "step": 66540 }, { "grad_norm": 0.20377308130264282, "learning_rate": 2.7597098016416784e-05, "loss": 0.015, "step": 66550 }, { "grad_norm": 0.12377109378576279, "learning_rate": 2.7582317147573368e-05, "loss": 0.0161, "step": 66560 }, { "grad_norm": 0.20065665245056152, "learning_rate": 2.7567538730294064e-05, "loss": 0.0169, "step": 66570 }, { "grad_norm": 0.18074364960193634, "learning_rate": 2.7552762766194973e-05, "loss": 0.0154, "step": 66580 }, { "grad_norm": 0.16844165325164795, "learning_rate": 2.753798925689202e-05, "loss": 0.0131, "step": 66590 }, { "grad_norm": 0.18146495521068573, "learning_rate": 2.7523218204000774e-05, "loss": 0.0188, "step": 66600 }, { "grad_norm": 0.13675040006637573, "learning_rate": 2.7508449609136612e-05, "loss": 0.0163, "step": 66610 }, { "grad_norm": 0.17910708487033844, "learning_rate": 2.749368347391454e-05, "loss": 0.0157, "step": 66620 }, { "grad_norm": 0.17298321425914764, "learning_rate": 2.7478919799949443e-05, "loss": 0.0164, "step": 66630 }, { "grad_norm": 0.2145462930202484, "learning_rate": 2.7464158588855803e-05, "loss": 0.0134, "step": 66640 }, { "grad_norm": 0.19271628558635712, "learning_rate": 2.7449399842247918e-05, "loss": 0.0209, "step": 66650 }, { "grad_norm": 0.18798792362213135, "learning_rate": 2.7434643561739736e-05, "loss": 0.0167, "step": 66660 }, { "grad_norm": 0.2263452112674713, "learning_rate": 2.7419889748945016e-05, "loss": 0.0159, "step": 66670 }, { "grad_norm": 0.18127018213272095, "learning_rate": 2.7405138405477203e-05, "loss": 0.0138, "step": 66680 }, { "grad_norm": 0.14729994535446167, "learning_rate": 2.73903895329495e-05, "loss": 0.0121, "step": 66690 }, { "grad_norm": 0.17890621721744537, "learning_rate": 2.737564313297478e-05, "loss": 0.0131, "step": 66700 }, { "grad_norm": 0.25676780939102173, "learning_rate": 2.736089920716572e-05, "loss": 0.0148, "step": 66710 }, { "grad_norm": 0.2003343552350998, "learning_rate": 2.7346157757134706e-05, "loss": 0.0155, "step": 66720 }, { "grad_norm": 0.18256784975528717, "learning_rate": 2.733141878449379e-05, "loss": 0.0148, "step": 66730 }, { "grad_norm": 0.17486357688903809, "learning_rate": 2.7316682290854844e-05, "loss": 0.0141, "step": 66740 }, { "grad_norm": 0.16744278371334076, "learning_rate": 2.7301948277829414e-05, "loss": 0.0123, "step": 66750 }, { "grad_norm": 0.20627495646476746, "learning_rate": 2.7287216747028814e-05, "loss": 0.0133, "step": 66760 }, { "grad_norm": 0.16196270287036896, "learning_rate": 2.7272487700064025e-05, "loss": 0.0148, "step": 66770 }, { "grad_norm": 0.20917139947414398, "learning_rate": 2.725776113854583e-05, "loss": 0.015, "step": 66780 }, { "grad_norm": 0.1562391221523285, "learning_rate": 2.7243037064084664e-05, "loss": 0.0184, "step": 66790 }, { "grad_norm": 0.24294547736644745, "learning_rate": 2.7228315478290766e-05, "loss": 0.0164, "step": 66800 }, { "grad_norm": 0.14399918913841248, "learning_rate": 2.721359638277402e-05, "loss": 0.0144, "step": 66810 }, { "grad_norm": 0.2581535279750824, "learning_rate": 2.7198879779144148e-05, "loss": 0.0157, "step": 66820 }, { "grad_norm": 0.10817421972751617, "learning_rate": 2.7184165669010485e-05, "loss": 0.013, "step": 66830 }, { "grad_norm": 0.32166028022766113, "learning_rate": 2.716945405398218e-05, "loss": 0.0149, "step": 66840 }, { "grad_norm": 0.157614603638649, "learning_rate": 2.715474493566803e-05, "loss": 0.0162, "step": 66850 }, { "grad_norm": 0.19198350608348846, "learning_rate": 2.7140038315676652e-05, "loss": 0.0153, "step": 66860 }, { "grad_norm": 0.13098251819610596, "learning_rate": 2.7125334195616275e-05, "loss": 0.0172, "step": 66870 }, { "grad_norm": 0.1656109243631363, "learning_rate": 2.7110632577094997e-05, "loss": 0.0167, "step": 66880 }, { "grad_norm": 0.29277122020721436, "learning_rate": 2.709593346172051e-05, "loss": 0.0163, "step": 66890 }, { "grad_norm": 0.20377930998802185, "learning_rate": 2.708123685110032e-05, "loss": 0.0173, "step": 66900 }, { "grad_norm": 0.19090840220451355, "learning_rate": 2.7066542746841588e-05, "loss": 0.0159, "step": 66910 }, { "grad_norm": 0.22076445817947388, "learning_rate": 2.7051851150551267e-05, "loss": 0.0138, "step": 66920 }, { "grad_norm": 0.20972289144992828, "learning_rate": 2.7037162063835996e-05, "loss": 0.0174, "step": 66930 }, { "grad_norm": 0.21415391564369202, "learning_rate": 2.702247548830218e-05, "loss": 0.0169, "step": 66940 }, { "grad_norm": 0.1346004456281662, "learning_rate": 2.7007791425555883e-05, "loss": 0.0178, "step": 66950 }, { "grad_norm": 0.2055961638689041, "learning_rate": 2.6993109877202945e-05, "loss": 0.0164, "step": 66960 }, { "grad_norm": 0.16241641342639923, "learning_rate": 2.6978430844848935e-05, "loss": 0.0125, "step": 66970 }, { "grad_norm": 0.1828388273715973, "learning_rate": 2.69637543300991e-05, "loss": 0.0119, "step": 66980 }, { "grad_norm": 0.22218823432922363, "learning_rate": 2.694908033455848e-05, "loss": 0.013, "step": 66990 }, { "grad_norm": 0.1623559594154358, "learning_rate": 2.6934408859831733e-05, "loss": 0.0159, "step": 67000 }, { "grad_norm": 0.15586182475090027, "learning_rate": 2.69197399075234e-05, "loss": 0.0166, "step": 67010 }, { "grad_norm": 0.17591272294521332, "learning_rate": 2.6905073479237584e-05, "loss": 0.017, "step": 67020 }, { "grad_norm": 0.13624021410942078, "learning_rate": 2.6890409576578235e-05, "loss": 0.0138, "step": 67030 }, { "grad_norm": 0.17532669007778168, "learning_rate": 2.687574820114892e-05, "loss": 0.0152, "step": 67040 }, { "grad_norm": 0.2730664908885956, "learning_rate": 2.686108935455305e-05, "loss": 0.0148, "step": 67050 }, { "grad_norm": 0.19548803567886353, "learning_rate": 2.6846433038393603e-05, "loss": 0.0165, "step": 67060 }, { "grad_norm": 0.33888494968414307, "learning_rate": 2.6831779254273482e-05, "loss": 0.0145, "step": 67070 }, { "grad_norm": 0.17536912858486176, "learning_rate": 2.6817128003795122e-05, "loss": 0.0154, "step": 67080 }, { "grad_norm": 0.138148695230484, "learning_rate": 2.680247928856081e-05, "loss": 0.018, "step": 67090 }, { "grad_norm": 0.24403370916843414, "learning_rate": 2.6787833110172466e-05, "loss": 0.0156, "step": 67100 }, { "grad_norm": 0.2574462294578552, "learning_rate": 2.6773189470231807e-05, "loss": 0.0164, "step": 67110 }, { "grad_norm": 0.1398022472858429, "learning_rate": 2.675854837034022e-05, "loss": 0.0166, "step": 67120 }, { "grad_norm": 0.27325159311294556, "learning_rate": 2.6743909812098833e-05, "loss": 0.0141, "step": 67130 }, { "grad_norm": 0.1642843335866928, "learning_rate": 2.67292737971085e-05, "loss": 0.0173, "step": 67140 }, { "grad_norm": 0.15000587701797485, "learning_rate": 2.671464032696982e-05, "loss": 0.0147, "step": 67150 }, { "grad_norm": 0.12822338938713074, "learning_rate": 2.6700009403283055e-05, "loss": 0.0171, "step": 67160 }, { "grad_norm": 0.20405729115009308, "learning_rate": 2.6685381027648214e-05, "loss": 0.0142, "step": 67170 }, { "grad_norm": 0.15833266079425812, "learning_rate": 2.6670755201665086e-05, "loss": 0.014, "step": 67180 }, { "grad_norm": 0.23631399869918823, "learning_rate": 2.6656131926933066e-05, "loss": 0.0156, "step": 67190 }, { "grad_norm": 0.17175723612308502, "learning_rate": 2.664151120505136e-05, "loss": 0.0136, "step": 67200 }, { "grad_norm": 0.18752248585224152, "learning_rate": 2.6626893037618873e-05, "loss": 0.014, "step": 67210 }, { "grad_norm": 0.1775800734758377, "learning_rate": 2.6612277426234243e-05, "loss": 0.0198, "step": 67220 }, { "grad_norm": 0.1525936722755432, "learning_rate": 2.659766437249577e-05, "loss": 0.0171, "step": 67230 }, { "grad_norm": 0.2105351835489273, "learning_rate": 2.6583053878001557e-05, "loss": 0.0136, "step": 67240 }, { "grad_norm": 0.18311648070812225, "learning_rate": 2.6568445944349323e-05, "loss": 0.0186, "step": 67250 }, { "grad_norm": 0.16260096430778503, "learning_rate": 2.6553840573136657e-05, "loss": 0.0151, "step": 67260 }, { "grad_norm": 0.22768722474575043, "learning_rate": 2.6539237765960706e-05, "loss": 0.0176, "step": 67270 }, { "grad_norm": 0.3866938054561615, "learning_rate": 2.6524637524418465e-05, "loss": 0.0123, "step": 67280 }, { "grad_norm": 0.27690792083740234, "learning_rate": 2.651003985010655e-05, "loss": 0.014, "step": 67290 }, { "grad_norm": 0.18859021365642548, "learning_rate": 2.649544474462138e-05, "loss": 0.0147, "step": 67300 }, { "grad_norm": 0.16720569133758545, "learning_rate": 2.6480852209559015e-05, "loss": 0.0124, "step": 67310 }, { "grad_norm": 0.18727852404117584, "learning_rate": 2.6466262246515296e-05, "loss": 0.0176, "step": 67320 }, { "grad_norm": 0.17017444968223572, "learning_rate": 2.6451674857085746e-05, "loss": 0.0161, "step": 67330 }, { "grad_norm": 0.1595943421125412, "learning_rate": 2.6437090042865655e-05, "loss": 0.0133, "step": 67340 }, { "grad_norm": 0.18120704591274261, "learning_rate": 2.6422507805449947e-05, "loss": 0.015, "step": 67350 }, { "grad_norm": 0.11869905889034271, "learning_rate": 2.6407928146433357e-05, "loss": 0.0132, "step": 67360 }, { "grad_norm": 0.20356015861034393, "learning_rate": 2.639335106741026e-05, "loss": 0.0157, "step": 67370 }, { "grad_norm": 0.2084326148033142, "learning_rate": 2.6378776569974796e-05, "loss": 0.014, "step": 67380 }, { "grad_norm": 0.19283358752727509, "learning_rate": 2.6364204655720814e-05, "loss": 0.0151, "step": 67390 }, { "grad_norm": 0.24740545451641083, "learning_rate": 2.6349635326241894e-05, "loss": 0.0143, "step": 67400 }, { "grad_norm": 0.22640752792358398, "learning_rate": 2.633506858313127e-05, "loss": 0.0131, "step": 67410 }, { "grad_norm": 0.18319909274578094, "learning_rate": 2.6320504427981975e-05, "loss": 0.016, "step": 67420 }, { "grad_norm": 0.18643756210803986, "learning_rate": 2.6305942862386734e-05, "loss": 0.0128, "step": 67430 }, { "grad_norm": 0.14436697959899902, "learning_rate": 2.629138388793793e-05, "loss": 0.0138, "step": 67440 }, { "grad_norm": 0.1477818489074707, "learning_rate": 2.6276827506227743e-05, "loss": 0.0126, "step": 67450 }, { "grad_norm": 0.3095865845680237, "learning_rate": 2.6262273718848037e-05, "loss": 0.0146, "step": 67460 }, { "grad_norm": 0.2428378313779831, "learning_rate": 2.6247722527390407e-05, "loss": 0.016, "step": 67470 }, { "grad_norm": 0.1572265625, "learning_rate": 2.6233173933446104e-05, "loss": 0.0141, "step": 67480 }, { "grad_norm": 0.19335760176181793, "learning_rate": 2.6218627938606188e-05, "loss": 0.0156, "step": 67490 }, { "grad_norm": 0.11798879504203796, "learning_rate": 2.620408454446135e-05, "loss": 0.0156, "step": 67500 }, { "grad_norm": 0.14703397452831268, "learning_rate": 2.6189543752602046e-05, "loss": 0.0149, "step": 67510 }, { "grad_norm": 0.34071725606918335, "learning_rate": 2.6175005564618444e-05, "loss": 0.014, "step": 67520 }, { "grad_norm": 0.1572926640510559, "learning_rate": 2.6160469982100428e-05, "loss": 0.0164, "step": 67530 }, { "grad_norm": 0.2197490781545639, "learning_rate": 2.6145937006637555e-05, "loss": 0.0148, "step": 67540 }, { "grad_norm": 0.19682839512825012, "learning_rate": 2.6131406639819166e-05, "loss": 0.0167, "step": 67550 }, { "grad_norm": 0.1628515124320984, "learning_rate": 2.611687888323424e-05, "loss": 0.0145, "step": 67560 }, { "grad_norm": 0.12919272482395172, "learning_rate": 2.610235373847153e-05, "loss": 0.0129, "step": 67570 }, { "grad_norm": 0.11866725236177444, "learning_rate": 2.6087831207119494e-05, "loss": 0.0171, "step": 67580 }, { "grad_norm": 0.25194457173347473, "learning_rate": 2.60733112907663e-05, "loss": 0.0158, "step": 67590 }, { "grad_norm": 0.20527614653110504, "learning_rate": 2.6058793990999786e-05, "loss": 0.0135, "step": 67600 }, { "grad_norm": 0.12305978685617447, "learning_rate": 2.6044279309407593e-05, "loss": 0.0153, "step": 67610 }, { "grad_norm": 0.2757519483566284, "learning_rate": 2.6029767247576976e-05, "loss": 0.0139, "step": 67620 }, { "grad_norm": 0.1374405026435852, "learning_rate": 2.601525780709497e-05, "loss": 0.0144, "step": 67630 }, { "grad_norm": 0.12236031144857407, "learning_rate": 2.6000750989548318e-05, "loss": 0.0125, "step": 67640 }, { "grad_norm": 0.18125320971012115, "learning_rate": 2.5986246796523462e-05, "loss": 0.0138, "step": 67650 }, { "grad_norm": 0.24535711109638214, "learning_rate": 2.5971745229606546e-05, "loss": 0.018, "step": 67660 }, { "grad_norm": 0.20270785689353943, "learning_rate": 2.5957246290383442e-05, "loss": 0.0139, "step": 67670 }, { "grad_norm": 0.1770065277814865, "learning_rate": 2.594274998043975e-05, "loss": 0.0146, "step": 67680 }, { "grad_norm": 0.16489319503307343, "learning_rate": 2.5928256301360733e-05, "loss": 0.0129, "step": 67690 }, { "grad_norm": 0.2370295375585556, "learning_rate": 2.591376525473142e-05, "loss": 0.017, "step": 67700 }, { "grad_norm": 0.19313204288482666, "learning_rate": 2.5899276842136523e-05, "loss": 0.0149, "step": 67710 }, { "grad_norm": 0.13677936792373657, "learning_rate": 2.5884791065160495e-05, "loss": 0.0124, "step": 67720 }, { "grad_norm": 0.16041043400764465, "learning_rate": 2.5870307925387436e-05, "loss": 0.0165, "step": 67730 }, { "grad_norm": 0.17259542644023895, "learning_rate": 2.585582742440125e-05, "loss": 0.0126, "step": 67740 }, { "grad_norm": 0.131737619638443, "learning_rate": 2.5841349563785455e-05, "loss": 0.0161, "step": 67750 }, { "grad_norm": 0.2183689922094345, "learning_rate": 2.5826874345123355e-05, "loss": 0.0147, "step": 67760 }, { "grad_norm": 0.18689966201782227, "learning_rate": 2.581240176999792e-05, "loss": 0.0147, "step": 67770 }, { "grad_norm": 0.19850283861160278, "learning_rate": 2.579793183999189e-05, "loss": 0.0163, "step": 67780 }, { "grad_norm": 0.17394758760929108, "learning_rate": 2.5783464556687627e-05, "loss": 0.0162, "step": 67790 }, { "grad_norm": 0.17178410291671753, "learning_rate": 2.576899992166729e-05, "loss": 0.018, "step": 67800 }, { "grad_norm": 0.167607381939888, "learning_rate": 2.5754537936512668e-05, "loss": 0.0147, "step": 67810 }, { "grad_norm": 0.14451482892036438, "learning_rate": 2.574007860280535e-05, "loss": 0.0132, "step": 67820 }, { "grad_norm": 0.20934131741523743, "learning_rate": 2.5725621922126518e-05, "loss": 0.0155, "step": 67830 }, { "grad_norm": 0.16213589906692505, "learning_rate": 2.571116789605722e-05, "loss": 0.0148, "step": 67840 }, { "grad_norm": 0.17704594135284424, "learning_rate": 2.5696716526178054e-05, "loss": 0.0151, "step": 67850 }, { "grad_norm": 0.19713838398456573, "learning_rate": 2.568226781406946e-05, "loss": 0.0172, "step": 67860 }, { "grad_norm": 0.16418170928955078, "learning_rate": 2.5667821761311474e-05, "loss": 0.0155, "step": 67870 }, { "grad_norm": 0.20845696330070496, "learning_rate": 2.5653378369483917e-05, "loss": 0.0149, "step": 67880 }, { "grad_norm": 0.17766809463500977, "learning_rate": 2.5638937640166292e-05, "loss": 0.0131, "step": 67890 }, { "grad_norm": 0.1432752162218094, "learning_rate": 2.5624499574937842e-05, "loss": 0.0149, "step": 67900 }, { "grad_norm": 0.19162480533123016, "learning_rate": 2.5610064175377456e-05, "loss": 0.0158, "step": 67910 }, { "grad_norm": 0.3534628450870514, "learning_rate": 2.5595631443063782e-05, "loss": 0.0136, "step": 67920 }, { "grad_norm": 0.23040832579135895, "learning_rate": 2.5581201379575177e-05, "loss": 0.016, "step": 67930 }, { "grad_norm": 0.2504296600818634, "learning_rate": 2.5566773986489655e-05, "loss": 0.0153, "step": 67940 }, { "grad_norm": 0.21267354488372803, "learning_rate": 2.5552349265385e-05, "loss": 0.0144, "step": 67950 }, { "grad_norm": 0.17653311789035797, "learning_rate": 2.5537927217838675e-05, "loss": 0.0145, "step": 67960 }, { "grad_norm": 0.22058862447738647, "learning_rate": 2.5523507845427864e-05, "loss": 0.0156, "step": 67970 }, { "grad_norm": 0.16436812281608582, "learning_rate": 2.5509091149729413e-05, "loss": 0.0168, "step": 67980 }, { "grad_norm": 0.2554989159107208, "learning_rate": 2.5494677132319962e-05, "loss": 0.0151, "step": 67990 }, { "grad_norm": 0.20413751900196075, "learning_rate": 2.548026579477575e-05, "loss": 0.0141, "step": 68000 }, { "grad_norm": 0.18994393944740295, "learning_rate": 2.5465857138672834e-05, "loss": 0.013, "step": 68010 }, { "grad_norm": 0.18893708288669586, "learning_rate": 2.5451451165586848e-05, "loss": 0.0162, "step": 68020 }, { "grad_norm": 0.17114239931106567, "learning_rate": 2.543704787709329e-05, "loss": 0.0154, "step": 68030 }, { "grad_norm": 0.15217572450637817, "learning_rate": 2.542264727476723e-05, "loss": 0.0132, "step": 68040 }, { "grad_norm": 0.1793864518404007, "learning_rate": 2.5408249360183535e-05, "loss": 0.0174, "step": 68050 }, { "grad_norm": 0.2332349717617035, "learning_rate": 2.5393854134916694e-05, "loss": 0.019, "step": 68060 }, { "grad_norm": 0.18741224706172943, "learning_rate": 2.5379461600540993e-05, "loss": 0.015, "step": 68070 }, { "grad_norm": 0.17702823877334595, "learning_rate": 2.5365071758630317e-05, "loss": 0.0172, "step": 68080 }, { "grad_norm": 0.14848576486110687, "learning_rate": 2.535068461075839e-05, "loss": 0.0132, "step": 68090 }, { "grad_norm": 0.21813839673995972, "learning_rate": 2.5336300158498516e-05, "loss": 0.0127, "step": 68100 }, { "grad_norm": 0.25296327471733093, "learning_rate": 2.5321918403423793e-05, "loss": 0.0192, "step": 68110 }, { "grad_norm": 0.21305187046527863, "learning_rate": 2.5307539347106957e-05, "loss": 0.017, "step": 68120 }, { "grad_norm": 0.20426969230175018, "learning_rate": 2.529316299112049e-05, "loss": 0.0138, "step": 68130 }, { "grad_norm": 0.12935179471969604, "learning_rate": 2.5278789337036567e-05, "loss": 0.0146, "step": 68140 }, { "grad_norm": 0.1467795968055725, "learning_rate": 2.52644183864271e-05, "loss": 0.0142, "step": 68150 }, { "grad_norm": 0.15120960772037506, "learning_rate": 2.525005014086363e-05, "loss": 0.0142, "step": 68160 }, { "grad_norm": 0.19871865212917328, "learning_rate": 2.5235684601917465e-05, "loss": 0.0173, "step": 68170 }, { "grad_norm": 0.1359604299068451, "learning_rate": 2.5221321771159618e-05, "loss": 0.0139, "step": 68180 }, { "grad_norm": 0.15543417632579803, "learning_rate": 2.5206961650160743e-05, "loss": 0.0134, "step": 68190 }, { "grad_norm": 0.17150796949863434, "learning_rate": 2.519260424049129e-05, "loss": 0.0144, "step": 68200 }, { "grad_norm": 0.5839594006538391, "learning_rate": 2.5178249543721304e-05, "loss": 0.0191, "step": 68210 }, { "grad_norm": 0.1350836455821991, "learning_rate": 2.5163897561420668e-05, "loss": 0.0168, "step": 68220 }, { "grad_norm": 0.1592787355184555, "learning_rate": 2.514954829515883e-05, "loss": 0.0182, "step": 68230 }, { "grad_norm": 0.17887386679649353, "learning_rate": 2.513520174650505e-05, "loss": 0.0168, "step": 68240 }, { "grad_norm": 0.18596039712429047, "learning_rate": 2.5120857917028207e-05, "loss": 0.0175, "step": 68250 }, { "grad_norm": 0.20683497190475464, "learning_rate": 2.5106516808296953e-05, "loss": 0.0169, "step": 68260 }, { "grad_norm": 0.1734791100025177, "learning_rate": 2.5092178421879552e-05, "loss": 0.0133, "step": 68270 }, { "grad_norm": 0.1735353022813797, "learning_rate": 2.5077842759344116e-05, "loss": 0.015, "step": 68280 }, { "grad_norm": 0.1794137805700302, "learning_rate": 2.5063509822258304e-05, "loss": 0.0138, "step": 68290 }, { "grad_norm": 0.17308984696865082, "learning_rate": 2.5049179612189576e-05, "loss": 0.0161, "step": 68300 }, { "grad_norm": 0.17184136807918549, "learning_rate": 2.5034852130705043e-05, "loss": 0.0136, "step": 68310 }, { "grad_norm": 0.1869049072265625, "learning_rate": 2.5020527379371572e-05, "loss": 0.0126, "step": 68320 }, { "grad_norm": 0.18007835745811462, "learning_rate": 2.500620535975564e-05, "loss": 0.014, "step": 68330 }, { "grad_norm": 0.20769992470741272, "learning_rate": 2.4991886073423515e-05, "loss": 0.0141, "step": 68340 }, { "grad_norm": 0.3225545883178711, "learning_rate": 2.4977569521941134e-05, "loss": 0.0175, "step": 68350 }, { "grad_norm": 0.963822066783905, "learning_rate": 2.4963255706874143e-05, "loss": 0.014, "step": 68360 }, { "grad_norm": 0.11691693216562271, "learning_rate": 2.4948944629787853e-05, "loss": 0.0127, "step": 68370 }, { "grad_norm": 0.11626001447439194, "learning_rate": 2.493463629224731e-05, "loss": 0.0141, "step": 68380 }, { "grad_norm": 0.20144839584827423, "learning_rate": 2.492033069581728e-05, "loss": 0.0161, "step": 68390 }, { "grad_norm": 0.21290411055088043, "learning_rate": 2.4906027842062137e-05, "loss": 0.0166, "step": 68400 }, { "grad_norm": 0.15456360578536987, "learning_rate": 2.48917277325461e-05, "loss": 0.0144, "step": 68410 }, { "grad_norm": 0.17675891518592834, "learning_rate": 2.487743036883295e-05, "loss": 0.013, "step": 68420 }, { "grad_norm": 0.20720933377742767, "learning_rate": 2.486313575248626e-05, "loss": 0.0149, "step": 68430 }, { "grad_norm": 0.13379442691802979, "learning_rate": 2.4848843885069235e-05, "loss": 0.0109, "step": 68440 }, { "grad_norm": 0.2238694578409195, "learning_rate": 2.483455476814484e-05, "loss": 0.0165, "step": 68450 }, { "grad_norm": 0.28847530484199524, "learning_rate": 2.4820268403275672e-05, "loss": 0.0151, "step": 68460 }, { "grad_norm": 0.2669890522956848, "learning_rate": 2.480598479202413e-05, "loss": 0.0141, "step": 68470 }, { "grad_norm": 0.25137045979499817, "learning_rate": 2.4791703935952193e-05, "loss": 0.0158, "step": 68480 }, { "grad_norm": 0.2026055008172989, "learning_rate": 2.477742583662163e-05, "loss": 0.0178, "step": 68490 }, { "grad_norm": 0.18683667480945587, "learning_rate": 2.476315049559384e-05, "loss": 0.0151, "step": 68500 }, { "grad_norm": 0.2759872078895569, "learning_rate": 2.4748877914429987e-05, "loss": 0.0164, "step": 68510 }, { "grad_norm": 0.22396181523799896, "learning_rate": 2.4734608094690857e-05, "loss": 0.0179, "step": 68520 }, { "grad_norm": 0.4524174630641937, "learning_rate": 2.4720341037937006e-05, "loss": 0.0165, "step": 68530 }, { "grad_norm": 0.2232622504234314, "learning_rate": 2.4706076745728645e-05, "loss": 0.015, "step": 68540 }, { "grad_norm": 0.19890113174915314, "learning_rate": 2.469181521962573e-05, "loss": 0.0155, "step": 68550 }, { "grad_norm": 0.16707734763622284, "learning_rate": 2.467755646118783e-05, "loss": 0.0153, "step": 68560 }, { "grad_norm": 0.16863608360290527, "learning_rate": 2.46633004719743e-05, "loss": 0.0144, "step": 68570 }, { "grad_norm": 0.23690447211265564, "learning_rate": 2.4649047253544123e-05, "loss": 0.0169, "step": 68580 }, { "grad_norm": 0.14928430318832397, "learning_rate": 2.4634796807456016e-05, "loss": 0.0148, "step": 68590 }, { "grad_norm": 0.13922612369060516, "learning_rate": 2.46205491352684e-05, "loss": 0.0134, "step": 68600 }, { "grad_norm": 0.22309716045856476, "learning_rate": 2.4606304238539375e-05, "loss": 0.0126, "step": 68610 }, { "grad_norm": 0.19129642844200134, "learning_rate": 2.4592062118826753e-05, "loss": 0.0116, "step": 68620 }, { "grad_norm": 0.2017466425895691, "learning_rate": 2.4577822777688004e-05, "loss": 0.0156, "step": 68630 }, { "grad_norm": 0.2222670465707779, "learning_rate": 2.4563586216680347e-05, "loss": 0.0164, "step": 68640 }, { "grad_norm": 0.17540311813354492, "learning_rate": 2.4549352437360622e-05, "loss": 0.0155, "step": 68650 }, { "grad_norm": 0.22643810510635376, "learning_rate": 2.4535121441285493e-05, "loss": 0.0133, "step": 68660 }, { "grad_norm": 0.40893614292144775, "learning_rate": 2.4520893230011172e-05, "loss": 0.0141, "step": 68670 }, { "grad_norm": 0.20933808386325836, "learning_rate": 2.450666780509368e-05, "loss": 0.0178, "step": 68680 }, { "grad_norm": 0.17109577357769012, "learning_rate": 2.4492445168088656e-05, "loss": 0.0163, "step": 68690 }, { "grad_norm": 0.160757377743721, "learning_rate": 2.447822532055149e-05, "loss": 0.0133, "step": 68700 }, { "grad_norm": 0.19255250692367554, "learning_rate": 2.4464008264037212e-05, "loss": 0.0132, "step": 68710 }, { "grad_norm": 0.2112598568201065, "learning_rate": 2.4449794000100605e-05, "loss": 0.0162, "step": 68720 }, { "grad_norm": 0.20325501263141632, "learning_rate": 2.4435582530296108e-05, "loss": 0.0151, "step": 68730 }, { "grad_norm": 0.16751058399677277, "learning_rate": 2.4421373856177887e-05, "loss": 0.0156, "step": 68740 }, { "grad_norm": 0.13948923349380493, "learning_rate": 2.440716797929975e-05, "loss": 0.0147, "step": 68750 }, { "grad_norm": 0.14011146128177643, "learning_rate": 2.439296490121526e-05, "loss": 0.0143, "step": 68760 }, { "grad_norm": 0.13826681673526764, "learning_rate": 2.4378764623477608e-05, "loss": 0.0189, "step": 68770 }, { "grad_norm": 0.21744172275066376, "learning_rate": 2.436456714763974e-05, "loss": 0.0167, "step": 68780 }, { "grad_norm": 0.28502312302589417, "learning_rate": 2.4350372475254262e-05, "loss": 0.0154, "step": 68790 }, { "grad_norm": 0.16922859847545624, "learning_rate": 2.433618060787351e-05, "loss": 0.0164, "step": 68800 }, { "grad_norm": 0.17438723146915436, "learning_rate": 2.432199154704944e-05, "loss": 0.0145, "step": 68810 }, { "grad_norm": 0.2031174898147583, "learning_rate": 2.430780529433377e-05, "loss": 0.0103, "step": 68820 }, { "grad_norm": 0.13743364810943604, "learning_rate": 2.42936218512779e-05, "loss": 0.0135, "step": 68830 }, { "grad_norm": 0.16365444660186768, "learning_rate": 2.427944121943288e-05, "loss": 0.0147, "step": 68840 }, { "grad_norm": 0.13851726055145264, "learning_rate": 2.4265263400349497e-05, "loss": 0.0139, "step": 68850 }, { "grad_norm": 0.252440482378006, "learning_rate": 2.4251088395578213e-05, "loss": 0.0165, "step": 68860 }, { "grad_norm": 0.11380661278963089, "learning_rate": 2.4236916206669214e-05, "loss": 0.0131, "step": 68870 }, { "grad_norm": 0.22065119445323944, "learning_rate": 2.4222746835172304e-05, "loss": 0.015, "step": 68880 }, { "grad_norm": 0.15568915009498596, "learning_rate": 2.4208580282637065e-05, "loss": 0.0153, "step": 68890 }, { "grad_norm": 0.13790687918663025, "learning_rate": 2.4194416550612687e-05, "loss": 0.0163, "step": 68900 }, { "grad_norm": 0.15114779770374298, "learning_rate": 2.418025564064812e-05, "loss": 0.0147, "step": 68910 }, { "grad_norm": 0.2671175003051758, "learning_rate": 2.4166097554291978e-05, "loss": 0.0136, "step": 68920 }, { "grad_norm": 0.20846973359584808, "learning_rate": 2.4151942293092583e-05, "loss": 0.0145, "step": 68930 }, { "grad_norm": 0.2657669484615326, "learning_rate": 2.41377898585979e-05, "loss": 0.016, "step": 68940 }, { "grad_norm": 0.1878364533185959, "learning_rate": 2.4123640252355652e-05, "loss": 0.0145, "step": 68950 }, { "grad_norm": 0.17361558973789215, "learning_rate": 2.410949347591318e-05, "loss": 0.0138, "step": 68960 }, { "grad_norm": 0.19305627048015594, "learning_rate": 2.4095349530817578e-05, "loss": 0.0137, "step": 68970 }, { "grad_norm": 0.20281828939914703, "learning_rate": 2.4081208418615603e-05, "loss": 0.0156, "step": 68980 }, { "grad_norm": 0.19810278713703156, "learning_rate": 2.4067070140853725e-05, "loss": 0.0171, "step": 68990 }, { "grad_norm": 0.21905238926410675, "learning_rate": 2.405293469907805e-05, "loss": 0.0152, "step": 69000 }, { "grad_norm": 0.18179133534431458, "learning_rate": 2.4038802094834446e-05, "loss": 0.0155, "step": 69010 }, { "grad_norm": 0.184733584523201, "learning_rate": 2.4024672329668395e-05, "loss": 0.015, "step": 69020 }, { "grad_norm": 0.16618266701698303, "learning_rate": 2.4010545405125125e-05, "loss": 0.0164, "step": 69030 }, { "grad_norm": 0.1756504476070404, "learning_rate": 2.3996421322749528e-05, "loss": 0.0134, "step": 69040 }, { "grad_norm": 0.21345973014831543, "learning_rate": 2.3982300084086224e-05, "loss": 0.0156, "step": 69050 }, { "grad_norm": 0.24329951405525208, "learning_rate": 2.396818169067944e-05, "loss": 0.0171, "step": 69060 }, { "grad_norm": 0.16678386926651, "learning_rate": 2.3954066144073178e-05, "loss": 0.0165, "step": 69070 }, { "grad_norm": 0.25520947575569153, "learning_rate": 2.39399534458111e-05, "loss": 0.0183, "step": 69080 }, { "grad_norm": 0.20278657972812653, "learning_rate": 2.392584359743651e-05, "loss": 0.0154, "step": 69090 }, { "grad_norm": 0.5746288299560547, "learning_rate": 2.3911736600492463e-05, "loss": 0.016, "step": 69100 }, { "grad_norm": 0.11496800929307938, "learning_rate": 2.3897632456521672e-05, "loss": 0.0146, "step": 69110 }, { "grad_norm": 0.2791765630245209, "learning_rate": 2.388353116706658e-05, "loss": 0.0152, "step": 69120 }, { "grad_norm": 0.18575909733772278, "learning_rate": 2.3869432733669232e-05, "loss": 0.0162, "step": 69130 }, { "grad_norm": 0.17378513514995575, "learning_rate": 2.385533715787145e-05, "loss": 0.0156, "step": 69140 }, { "grad_norm": 0.1374199241399765, "learning_rate": 2.3841244441214673e-05, "loss": 0.0124, "step": 69150 }, { "grad_norm": 0.20434622466564178, "learning_rate": 2.3827154585240073e-05, "loss": 0.0189, "step": 69160 }, { "grad_norm": 0.15874026715755463, "learning_rate": 2.3813067591488498e-05, "loss": 0.0169, "step": 69170 }, { "grad_norm": 0.26151350140571594, "learning_rate": 2.3798983461500494e-05, "loss": 0.0143, "step": 69180 }, { "grad_norm": 0.15011295676231384, "learning_rate": 2.3784902196816256e-05, "loss": 0.0121, "step": 69190 }, { "grad_norm": 0.1726982444524765, "learning_rate": 2.3770823798975716e-05, "loss": 0.0147, "step": 69200 }, { "grad_norm": 0.17766030132770538, "learning_rate": 2.3756748269518424e-05, "loss": 0.017, "step": 69210 }, { "grad_norm": 0.18375028669834137, "learning_rate": 2.374267560998371e-05, "loss": 0.0124, "step": 69220 }, { "grad_norm": 0.17210999131202698, "learning_rate": 2.3728605821910477e-05, "loss": 0.0155, "step": 69230 }, { "grad_norm": 0.13585923612117767, "learning_rate": 2.3714538906837453e-05, "loss": 0.0151, "step": 69240 }, { "grad_norm": 0.2016706019639969, "learning_rate": 2.3700474866302913e-05, "loss": 0.0141, "step": 69250 }, { "grad_norm": 0.12479736655950546, "learning_rate": 2.3686413701844917e-05, "loss": 0.0165, "step": 69260 }, { "grad_norm": 0.20000194013118744, "learning_rate": 2.367235541500114e-05, "loss": 0.0174, "step": 69270 }, { "grad_norm": 0.15969109535217285, "learning_rate": 2.3658300007308993e-05, "loss": 0.0157, "step": 69280 }, { "grad_norm": 0.5276307463645935, "learning_rate": 2.364424748030555e-05, "loss": 0.0212, "step": 69290 }, { "grad_norm": 0.1849430948495865, "learning_rate": 2.3630197835527595e-05, "loss": 0.0162, "step": 69300 }, { "grad_norm": 0.24028043448925018, "learning_rate": 2.3616151074511546e-05, "loss": 0.0157, "step": 69310 }, { "grad_norm": 0.1632838100194931, "learning_rate": 2.3602107198793545e-05, "loss": 0.0143, "step": 69320 }, { "grad_norm": 0.18587727844715118, "learning_rate": 2.3588066209909427e-05, "loss": 0.0147, "step": 69330 }, { "grad_norm": 0.47458770871162415, "learning_rate": 2.3574028109394662e-05, "loss": 0.014, "step": 69340 }, { "grad_norm": 0.22416560351848602, "learning_rate": 2.3559992898784454e-05, "loss": 0.016, "step": 69350 }, { "grad_norm": 0.21622484922409058, "learning_rate": 2.354596057961367e-05, "loss": 0.0163, "step": 69360 }, { "grad_norm": 0.23886710405349731, "learning_rate": 2.353193115341688e-05, "loss": 0.0146, "step": 69370 }, { "grad_norm": 0.11805958300828934, "learning_rate": 2.3517904621728294e-05, "loss": 0.0141, "step": 69380 }, { "grad_norm": 0.17150194942951202, "learning_rate": 2.3503880986081855e-05, "loss": 0.0163, "step": 69390 }, { "grad_norm": 0.21978650987148285, "learning_rate": 2.3489860248011136e-05, "loss": 0.0161, "step": 69400 }, { "grad_norm": 0.5433674454689026, "learning_rate": 2.3475842409049464e-05, "loss": 0.0155, "step": 69410 }, { "grad_norm": 0.2399958074092865, "learning_rate": 2.346182747072975e-05, "loss": 0.013, "step": 69420 }, { "grad_norm": 0.1418897807598114, "learning_rate": 2.3447815434584718e-05, "loss": 0.0161, "step": 69430 }, { "grad_norm": 0.16349613666534424, "learning_rate": 2.3433806302146656e-05, "loss": 0.0148, "step": 69440 }, { "grad_norm": 0.23240716755390167, "learning_rate": 2.3419800074947612e-05, "loss": 0.0156, "step": 69450 }, { "grad_norm": 0.22581954300403595, "learning_rate": 2.3405796754519248e-05, "loss": 0.016, "step": 69460 }, { "grad_norm": 0.15019801259040833, "learning_rate": 2.3391796342392992e-05, "loss": 0.0153, "step": 69470 }, { "grad_norm": 0.1717495173215866, "learning_rate": 2.337779884009984e-05, "loss": 0.013, "step": 69480 }, { "grad_norm": 0.30095815658569336, "learning_rate": 2.3363804249170624e-05, "loss": 0.0145, "step": 69490 }, { "grad_norm": 0.14388631284236908, "learning_rate": 2.3349812571135703e-05, "loss": 0.015, "step": 69500 }, { "grad_norm": 0.2049727886915207, "learning_rate": 2.333582380752523e-05, "loss": 0.0161, "step": 69510 }, { "grad_norm": 0.147335946559906, "learning_rate": 2.3321837959868963e-05, "loss": 0.0142, "step": 69520 }, { "grad_norm": 0.18392899632453918, "learning_rate": 2.3307855029696383e-05, "loss": 0.0129, "step": 69530 }, { "grad_norm": 0.16917982697486877, "learning_rate": 2.3293875018536648e-05, "loss": 0.014, "step": 69540 }, { "grad_norm": 0.1590634286403656, "learning_rate": 2.327989792791861e-05, "loss": 0.0125, "step": 69550 }, { "grad_norm": 0.1723678708076477, "learning_rate": 2.326592375937074e-05, "loss": 0.0132, "step": 69560 }, { "grad_norm": 0.17582187056541443, "learning_rate": 2.3251952514421248e-05, "loss": 0.0137, "step": 69570 }, { "grad_norm": 0.2657001316547394, "learning_rate": 2.323798419459804e-05, "loss": 0.0141, "step": 69580 }, { "grad_norm": 0.14231927692890167, "learning_rate": 2.3224018801428625e-05, "loss": 0.0141, "step": 69590 }, { "grad_norm": 0.13113057613372803, "learning_rate": 2.3210056336440273e-05, "loss": 0.0118, "step": 69600 }, { "grad_norm": 0.15087315440177917, "learning_rate": 2.3196096801159844e-05, "loss": 0.0148, "step": 69610 }, { "grad_norm": 0.1892079859972, "learning_rate": 2.3182140197114012e-05, "loss": 0.0168, "step": 69620 }, { "grad_norm": 0.22077126801013947, "learning_rate": 2.3168186525828985e-05, "loss": 0.0125, "step": 69630 }, { "grad_norm": 0.15084148943424225, "learning_rate": 2.3154235788830757e-05, "loss": 0.0142, "step": 69640 }, { "grad_norm": 0.14091245830059052, "learning_rate": 2.3140287987644927e-05, "loss": 0.016, "step": 69650 }, { "grad_norm": 0.19040197134017944, "learning_rate": 2.3126343123796833e-05, "loss": 0.015, "step": 69660 }, { "grad_norm": 0.27763527631759644, "learning_rate": 2.3112401198811418e-05, "loss": 0.0138, "step": 69670 }, { "grad_norm": 0.1883411556482315, "learning_rate": 2.3098462214213418e-05, "loss": 0.014, "step": 69680 }, { "grad_norm": 0.2676977515220642, "learning_rate": 2.308452617152712e-05, "loss": 0.0136, "step": 69690 }, { "grad_norm": 0.19911302626132965, "learning_rate": 2.3070593072276603e-05, "loss": 0.0151, "step": 69700 }, { "grad_norm": 0.1351281851530075, "learning_rate": 2.3056662917985518e-05, "loss": 0.0151, "step": 69710 }, { "grad_norm": 0.16582746803760529, "learning_rate": 2.3042735710177283e-05, "loss": 0.0142, "step": 69720 }, { "grad_norm": 0.16834735870361328, "learning_rate": 2.302881145037492e-05, "loss": 0.0164, "step": 69730 }, { "grad_norm": 0.194685161113739, "learning_rate": 2.3014890140101192e-05, "loss": 0.0162, "step": 69740 }, { "grad_norm": 0.19334450364112854, "learning_rate": 2.30009717808785e-05, "loss": 0.0154, "step": 69750 }, { "grad_norm": 0.13450545072555542, "learning_rate": 2.2987056374228967e-05, "loss": 0.0116, "step": 69760 }, { "grad_norm": 0.17209669947624207, "learning_rate": 2.297314392167431e-05, "loss": 0.0164, "step": 69770 }, { "grad_norm": 0.17951035499572754, "learning_rate": 2.295923442473601e-05, "loss": 0.014, "step": 69780 }, { "grad_norm": 0.18374183773994446, "learning_rate": 2.294532788493519e-05, "loss": 0.0138, "step": 69790 }, { "grad_norm": 0.30213943123817444, "learning_rate": 2.293142430379262e-05, "loss": 0.015, "step": 69800 }, { "grad_norm": 0.1404806226491928, "learning_rate": 2.291752368282879e-05, "loss": 0.0131, "step": 69810 }, { "grad_norm": 0.22770142555236816, "learning_rate": 2.290362602356385e-05, "loss": 0.0149, "step": 69820 }, { "grad_norm": 0.1390230804681778, "learning_rate": 2.288973132751765e-05, "loss": 0.0122, "step": 69830 }, { "grad_norm": 0.17653553187847137, "learning_rate": 2.287583959620965e-05, "loss": 0.0175, "step": 69840 }, { "grad_norm": 0.18354469537734985, "learning_rate": 2.2861950831159074e-05, "loss": 0.0116, "step": 69850 }, { "grad_norm": 0.21427525579929352, "learning_rate": 2.284806503388472e-05, "loss": 0.0136, "step": 69860 }, { "grad_norm": 0.18849141895771027, "learning_rate": 2.283418220590518e-05, "loss": 0.0164, "step": 69870 }, { "grad_norm": 0.13039231300354004, "learning_rate": 2.282030234873862e-05, "loss": 0.013, "step": 69880 }, { "grad_norm": 0.22381329536437988, "learning_rate": 2.2806425463902948e-05, "loss": 0.0132, "step": 69890 }, { "grad_norm": 0.18146339058876038, "learning_rate": 2.279255155291568e-05, "loss": 0.0144, "step": 69900 }, { "grad_norm": 0.1848268061876297, "learning_rate": 2.277868061729409e-05, "loss": 0.0139, "step": 69910 }, { "grad_norm": 0.2195015251636505, "learning_rate": 2.276481265855504e-05, "loss": 0.0146, "step": 69920 }, { "grad_norm": 0.17477218806743622, "learning_rate": 2.2750947678215128e-05, "loss": 0.0144, "step": 69930 }, { "grad_norm": 0.13945093750953674, "learning_rate": 2.2737085677790616e-05, "loss": 0.0132, "step": 69940 }, { "grad_norm": 0.12129372358322144, "learning_rate": 2.2723226658797443e-05, "loss": 0.0193, "step": 69950 }, { "grad_norm": 0.274125874042511, "learning_rate": 2.270937062275117e-05, "loss": 0.0167, "step": 69960 }, { "grad_norm": 0.21008843183517456, "learning_rate": 2.2695517571167113e-05, "loss": 0.0144, "step": 69970 }, { "grad_norm": 0.19463030993938446, "learning_rate": 2.268166750556019e-05, "loss": 0.0148, "step": 69980 }, { "grad_norm": 0.4826914668083191, "learning_rate": 2.266782042744503e-05, "loss": 0.0158, "step": 69990 }, { "grad_norm": 0.2406773865222931, "learning_rate": 2.2653976338335935e-05, "loss": 0.0131, "step": 70000 }, { "grad_norm": 0.8992515206336975, "learning_rate": 2.2640135239746895e-05, "loss": 0.0167, "step": 70010 }, { "grad_norm": 0.1700291931629181, "learning_rate": 2.2626297133191515e-05, "loss": 0.0215, "step": 70020 }, { "grad_norm": 0.14715443551540375, "learning_rate": 2.261246202018312e-05, "loss": 0.0145, "step": 70030 }, { "grad_norm": 0.16602152585983276, "learning_rate": 2.2598629902234718e-05, "loss": 0.0133, "step": 70040 }, { "grad_norm": 0.24967056512832642, "learning_rate": 2.258480078085894e-05, "loss": 0.0148, "step": 70050 }, { "grad_norm": 0.1819509118795395, "learning_rate": 2.2570974657568126e-05, "loss": 0.0179, "step": 70060 }, { "grad_norm": 0.20274391770362854, "learning_rate": 2.2557151533874286e-05, "loss": 0.0132, "step": 70070 }, { "grad_norm": 1.2757295370101929, "learning_rate": 2.2543331411289116e-05, "loss": 0.0173, "step": 70080 }, { "grad_norm": 0.16984571516513824, "learning_rate": 2.252951429132392e-05, "loss": 0.0143, "step": 70090 }, { "grad_norm": 0.5422137379646301, "learning_rate": 2.2515700175489762e-05, "loss": 0.0161, "step": 70100 }, { "grad_norm": 0.14238283038139343, "learning_rate": 2.2501889065297288e-05, "loss": 0.0132, "step": 70110 }, { "grad_norm": 0.16367019712924957, "learning_rate": 2.2488080962256887e-05, "loss": 0.0119, "step": 70120 }, { "grad_norm": 0.16043232381343842, "learning_rate": 2.2474275867878585e-05, "loss": 0.0124, "step": 70130 }, { "grad_norm": 0.26287269592285156, "learning_rate": 2.246047378367211e-05, "loss": 0.016, "step": 70140 }, { "grad_norm": 0.16490796208381653, "learning_rate": 2.2446674711146798e-05, "loss": 0.0156, "step": 70150 }, { "grad_norm": 0.21953989565372467, "learning_rate": 2.2432878651811734e-05, "loss": 0.0188, "step": 70160 }, { "grad_norm": 0.3101935386657715, "learning_rate": 2.2419085607175594e-05, "loss": 0.0172, "step": 70170 }, { "grad_norm": 0.23607556521892548, "learning_rate": 2.240529557874678e-05, "loss": 0.0159, "step": 70180 }, { "grad_norm": 0.20890672504901886, "learning_rate": 2.2391508568033358e-05, "loss": 0.0133, "step": 70190 }, { "grad_norm": 0.16353972256183624, "learning_rate": 2.237772457654307e-05, "loss": 0.0165, "step": 70200 }, { "grad_norm": 0.18831147253513336, "learning_rate": 2.236394360578327e-05, "loss": 0.0139, "step": 70210 }, { "grad_norm": 0.20425891876220703, "learning_rate": 2.2350165657261073e-05, "loss": 0.0191, "step": 70220 }, { "grad_norm": 0.19427484273910522, "learning_rate": 2.2336390732483164e-05, "loss": 0.0185, "step": 70230 }, { "grad_norm": 0.18184946477413177, "learning_rate": 2.2322618832955976e-05, "loss": 0.0164, "step": 70240 }, { "grad_norm": 0.13357006013393402, "learning_rate": 2.2308849960185578e-05, "loss": 0.0139, "step": 70250 }, { "grad_norm": 0.17009343206882477, "learning_rate": 2.229508411567774e-05, "loss": 0.0171, "step": 70260 }, { "grad_norm": 0.23388056457042694, "learning_rate": 2.2281321300937823e-05, "loss": 0.0149, "step": 70270 }, { "grad_norm": 0.21953995525836945, "learning_rate": 2.2267561517470937e-05, "loss": 0.014, "step": 70280 }, { "grad_norm": 0.1646827906370163, "learning_rate": 2.2253804766781845e-05, "loss": 0.0123, "step": 70290 }, { "grad_norm": 0.14186300337314606, "learning_rate": 2.2240051050374933e-05, "loss": 0.0141, "step": 70300 }, { "grad_norm": 0.13887065649032593, "learning_rate": 2.22263003697543e-05, "loss": 0.0143, "step": 70310 }, { "grad_norm": 0.16852903366088867, "learning_rate": 2.2212552726423702e-05, "loss": 0.0151, "step": 70320 }, { "grad_norm": 0.22843633592128754, "learning_rate": 2.2198808121886578e-05, "loss": 0.0137, "step": 70330 }, { "grad_norm": 0.1541237086057663, "learning_rate": 2.218506655764598e-05, "loss": 0.0135, "step": 70340 }, { "grad_norm": 0.12619785964488983, "learning_rate": 2.2171328035204704e-05, "loss": 0.0165, "step": 70350 }, { "grad_norm": 0.13112296164035797, "learning_rate": 2.2157592556065133e-05, "loss": 0.0149, "step": 70360 }, { "grad_norm": 0.2870159447193146, "learning_rate": 2.2143860121729382e-05, "loss": 0.0159, "step": 70370 }, { "grad_norm": 0.22676023840904236, "learning_rate": 2.2130130733699206e-05, "loss": 0.0171, "step": 70380 }, { "grad_norm": 0.19804710149765015, "learning_rate": 2.2116404393476052e-05, "loss": 0.0143, "step": 70390 }, { "grad_norm": 0.19662383198738098, "learning_rate": 2.2102681102560967e-05, "loss": 0.0189, "step": 70400 }, { "grad_norm": 0.12444531917572021, "learning_rate": 2.208896086245476e-05, "loss": 0.0147, "step": 70410 }, { "grad_norm": 0.4672756493091583, "learning_rate": 2.2075243674657808e-05, "loss": 0.0127, "step": 70420 }, { "grad_norm": 0.15790992975234985, "learning_rate": 2.206152954067024e-05, "loss": 0.0156, "step": 70430 }, { "grad_norm": 0.2148950695991516, "learning_rate": 2.2047818461991758e-05, "loss": 0.0128, "step": 70440 }, { "grad_norm": 0.21169191598892212, "learning_rate": 2.2034110440121864e-05, "loss": 0.0127, "step": 70450 }, { "grad_norm": 0.16700351238250732, "learning_rate": 2.2020405476559585e-05, "loss": 0.0152, "step": 70460 }, { "grad_norm": 0.1739198863506317, "learning_rate": 2.2006703572803722e-05, "loss": 0.0139, "step": 70470 }, { "grad_norm": 0.1958906650543213, "learning_rate": 2.1993004730352646e-05, "loss": 0.0147, "step": 70480 }, { "grad_norm": 0.17452886700630188, "learning_rate": 2.1979308950704464e-05, "loss": 0.0138, "step": 70490 }, { "grad_norm": 0.23132245242595673, "learning_rate": 2.1965616235356923e-05, "loss": 0.0154, "step": 70500 }, { "grad_norm": 0.1597249060869217, "learning_rate": 2.1951926585807465e-05, "loss": 0.0156, "step": 70510 }, { "grad_norm": 0.25792941451072693, "learning_rate": 2.193824000355313e-05, "loss": 0.014, "step": 70520 }, { "grad_norm": 0.17142394185066223, "learning_rate": 2.192455649009067e-05, "loss": 0.0152, "step": 70530 }, { "grad_norm": 0.16971975564956665, "learning_rate": 2.1910876046916523e-05, "loss": 0.014, "step": 70540 }, { "grad_norm": 0.13380789756774902, "learning_rate": 2.1897198675526725e-05, "loss": 0.0111, "step": 70550 }, { "grad_norm": 0.20106284320354462, "learning_rate": 2.1883524377417024e-05, "loss": 0.0121, "step": 70560 }, { "grad_norm": 0.2140936553478241, "learning_rate": 2.1869853154082825e-05, "loss": 0.0144, "step": 70570 }, { "grad_norm": 0.2643638253211975, "learning_rate": 2.1856185007019215e-05, "loss": 0.0165, "step": 70580 }, { "grad_norm": 0.1697145253419876, "learning_rate": 2.1842519937720874e-05, "loss": 0.0158, "step": 70590 }, { "grad_norm": 0.16410431265830994, "learning_rate": 2.182885794768224e-05, "loss": 0.0133, "step": 70600 }, { "grad_norm": 0.19387036561965942, "learning_rate": 2.1815199038397326e-05, "loss": 0.0164, "step": 70610 }, { "grad_norm": 0.23252946138381958, "learning_rate": 2.1801543211359893e-05, "loss": 0.0135, "step": 70620 }, { "grad_norm": 0.17462146282196045, "learning_rate": 2.178789046806326e-05, "loss": 0.0153, "step": 70630 }, { "grad_norm": 0.16631625592708588, "learning_rate": 2.1774240810000547e-05, "loss": 0.0126, "step": 70640 }, { "grad_norm": 0.15575401484966278, "learning_rate": 2.17605942386644e-05, "loss": 0.0128, "step": 70650 }, { "grad_norm": 0.23765593767166138, "learning_rate": 2.174695075554723e-05, "loss": 0.0145, "step": 70660 }, { "grad_norm": 0.18567971885204315, "learning_rate": 2.173331036214103e-05, "loss": 0.0132, "step": 70670 }, { "grad_norm": 0.14454829692840576, "learning_rate": 2.171967305993753e-05, "loss": 0.015, "step": 70680 }, { "grad_norm": 0.20914453268051147, "learning_rate": 2.1706038850428024e-05, "loss": 0.0156, "step": 70690 }, { "grad_norm": 0.1817326545715332, "learning_rate": 2.169240773510361e-05, "loss": 0.0142, "step": 70700 }, { "grad_norm": 0.1383604258298874, "learning_rate": 2.1678779715454906e-05, "loss": 0.0143, "step": 70710 }, { "grad_norm": 0.25653594732284546, "learning_rate": 2.166515479297229e-05, "loss": 0.0137, "step": 70720 }, { "grad_norm": 0.27098017930984497, "learning_rate": 2.1651532969145732e-05, "loss": 0.0191, "step": 70730 }, { "grad_norm": 0.2973472774028778, "learning_rate": 2.1637914245464903e-05, "loss": 0.0163, "step": 70740 }, { "grad_norm": 0.2749265432357788, "learning_rate": 2.1624298623419133e-05, "loss": 0.0163, "step": 70750 }, { "grad_norm": 0.17390097677707672, "learning_rate": 2.1610686104497413e-05, "loss": 0.0129, "step": 70760 }, { "grad_norm": 0.20304200053215027, "learning_rate": 2.1597076690188366e-05, "loss": 0.0189, "step": 70770 }, { "grad_norm": 0.15835048258304596, "learning_rate": 2.1583470381980315e-05, "loss": 0.0131, "step": 70780 }, { "grad_norm": 0.1609317660331726, "learning_rate": 2.156986718136123e-05, "loss": 0.0148, "step": 70790 }, { "grad_norm": 0.22206878662109375, "learning_rate": 2.155626708981871e-05, "loss": 0.0168, "step": 70800 }, { "grad_norm": 0.23511454463005066, "learning_rate": 2.1542670108840075e-05, "loss": 0.0149, "step": 70810 }, { "grad_norm": 0.1833486407995224, "learning_rate": 2.152907623991221e-05, "loss": 0.0105, "step": 70820 }, { "grad_norm": 0.19527779519557953, "learning_rate": 2.1515485484521802e-05, "loss": 0.0138, "step": 70830 }, { "grad_norm": 0.167558491230011, "learning_rate": 2.150189784415506e-05, "loss": 0.0151, "step": 70840 }, { "grad_norm": 0.1573791354894638, "learning_rate": 2.148831332029794e-05, "loss": 0.0174, "step": 70850 }, { "grad_norm": 0.1479271799325943, "learning_rate": 2.147473191443599e-05, "loss": 0.0126, "step": 70860 }, { "grad_norm": 0.15445809066295624, "learning_rate": 2.1461153628054492e-05, "loss": 0.0185, "step": 70870 }, { "grad_norm": 0.3477948307991028, "learning_rate": 2.1447578462638284e-05, "loss": 0.0141, "step": 70880 }, { "grad_norm": 0.15540239214897156, "learning_rate": 2.143400641967201e-05, "loss": 0.015, "step": 70890 }, { "grad_norm": 0.29920288920402527, "learning_rate": 2.1420437500639823e-05, "loss": 0.0156, "step": 70900 }, { "grad_norm": 0.3027271032333374, "learning_rate": 2.1406871707025646e-05, "loss": 0.0103, "step": 70910 }, { "grad_norm": 0.15472398698329926, "learning_rate": 2.1393309040312963e-05, "loss": 0.0164, "step": 70920 }, { "grad_norm": 0.20351138710975647, "learning_rate": 2.137974950198502e-05, "loss": 0.0159, "step": 70930 }, { "grad_norm": 0.20406556129455566, "learning_rate": 2.1366193093524617e-05, "loss": 0.0167, "step": 70940 }, { "grad_norm": 0.13309350609779358, "learning_rate": 2.1352639816414288e-05, "loss": 0.0143, "step": 70950 }, { "grad_norm": 0.1496395468711853, "learning_rate": 2.1339089672136202e-05, "loss": 0.0134, "step": 70960 }, { "grad_norm": 0.1718449741601944, "learning_rate": 2.1325542662172193e-05, "loss": 0.0109, "step": 70970 }, { "grad_norm": 0.1495838612318039, "learning_rate": 2.1311998788003712e-05, "loss": 0.012, "step": 70980 }, { "grad_norm": 0.1968930959701538, "learning_rate": 2.1298458051111907e-05, "loss": 0.016, "step": 70990 }, { "grad_norm": 0.2309143990278244, "learning_rate": 2.1284920452977602e-05, "loss": 0.0143, "step": 71000 }, { "grad_norm": 0.20689518749713898, "learning_rate": 2.12713859950812e-05, "loss": 0.019, "step": 71010 }, { "grad_norm": 0.17673726379871368, "learning_rate": 2.1257854678902832e-05, "loss": 0.0148, "step": 71020 }, { "grad_norm": 0.1860366016626358, "learning_rate": 2.124432650592227e-05, "loss": 0.0163, "step": 71030 }, { "grad_norm": 0.22340600192546844, "learning_rate": 2.1230801477618937e-05, "loss": 0.0132, "step": 71040 }, { "grad_norm": 0.23155494034290314, "learning_rate": 2.121727959547189e-05, "loss": 0.0141, "step": 71050 }, { "grad_norm": 0.20708145201206207, "learning_rate": 2.1203760860959886e-05, "loss": 0.0152, "step": 71060 }, { "grad_norm": 0.17258739471435547, "learning_rate": 2.119024527556127e-05, "loss": 0.0134, "step": 71070 }, { "grad_norm": 0.12183310091495514, "learning_rate": 2.117673284075415e-05, "loss": 0.0157, "step": 71080 }, { "grad_norm": 0.17556069791316986, "learning_rate": 2.1163223558016165e-05, "loss": 0.0144, "step": 71090 }, { "grad_norm": 0.21175353229045868, "learning_rate": 2.1149717428824722e-05, "loss": 0.0135, "step": 71100 }, { "grad_norm": 0.20423075556755066, "learning_rate": 2.1136214454656784e-05, "loss": 0.0145, "step": 71110 }, { "grad_norm": 0.1887914538383484, "learning_rate": 2.1122714636989056e-05, "loss": 0.0129, "step": 71120 }, { "grad_norm": 0.15262508392333984, "learning_rate": 2.1109217977297823e-05, "loss": 0.014, "step": 71130 }, { "grad_norm": 0.20567303895950317, "learning_rate": 2.1095724477059077e-05, "loss": 0.0136, "step": 71140 }, { "grad_norm": 0.11806818097829819, "learning_rate": 2.1082234137748434e-05, "loss": 0.0131, "step": 71150 }, { "grad_norm": 0.2075841724872589, "learning_rate": 2.106874696084122e-05, "loss": 0.0156, "step": 71160 }, { "grad_norm": 0.23395174741744995, "learning_rate": 2.1055262947812315e-05, "loss": 0.0154, "step": 71170 }, { "grad_norm": 0.16502009332180023, "learning_rate": 2.104178210013636e-05, "loss": 0.0135, "step": 71180 }, { "grad_norm": 0.2013535350561142, "learning_rate": 2.1028304419287554e-05, "loss": 0.0131, "step": 71190 }, { "grad_norm": 0.2127794623374939, "learning_rate": 2.101482990673982e-05, "loss": 0.0137, "step": 71200 }, { "grad_norm": 0.14192542433738708, "learning_rate": 2.1001358563966707e-05, "loss": 0.0128, "step": 71210 }, { "grad_norm": 0.2291981279850006, "learning_rate": 2.098789039244144e-05, "loss": 0.0152, "step": 71220 }, { "grad_norm": 0.14469139277935028, "learning_rate": 2.097442539363685e-05, "loss": 0.018, "step": 71230 }, { "grad_norm": 0.11336059868335724, "learning_rate": 2.0960963569025448e-05, "loss": 0.0136, "step": 71240 }, { "grad_norm": 0.2128087878227234, "learning_rate": 2.0947504920079435e-05, "loss": 0.0162, "step": 71250 }, { "grad_norm": 0.17534011602401733, "learning_rate": 2.093404944827059e-05, "loss": 0.014, "step": 71260 }, { "grad_norm": 0.17696066200733185, "learning_rate": 2.0920597155070394e-05, "loss": 0.014, "step": 71270 }, { "grad_norm": 0.2977273762226105, "learning_rate": 2.090714804194997e-05, "loss": 0.0167, "step": 71280 }, { "grad_norm": 0.223770871758461, "learning_rate": 2.0893702110380114e-05, "loss": 0.0153, "step": 71290 }, { "grad_norm": 0.2875058352947235, "learning_rate": 2.0880259361831216e-05, "loss": 0.0148, "step": 71300 }, { "grad_norm": 0.13507232069969177, "learning_rate": 2.0866819797773397e-05, "loss": 0.0138, "step": 71310 }, { "grad_norm": 0.16731123626232147, "learning_rate": 2.085338341967634e-05, "loss": 0.0128, "step": 71320 }, { "grad_norm": 0.15732066333293915, "learning_rate": 2.083995022900946e-05, "loss": 0.0131, "step": 71330 }, { "grad_norm": 0.34764763712882996, "learning_rate": 2.0826520227241768e-05, "loss": 0.0158, "step": 71340 }, { "grad_norm": 0.14646117389202118, "learning_rate": 2.081309341584199e-05, "loss": 0.0145, "step": 71350 }, { "grad_norm": 0.1172608733177185, "learning_rate": 2.0799669796278413e-05, "loss": 0.0131, "step": 71360 }, { "grad_norm": 0.22727304697036743, "learning_rate": 2.078624937001906e-05, "loss": 0.016, "step": 71370 }, { "grad_norm": 0.15335571765899658, "learning_rate": 2.0772832138531544e-05, "loss": 0.015, "step": 71380 }, { "grad_norm": 0.28981485962867737, "learning_rate": 2.0759418103283158e-05, "loss": 0.0143, "step": 71390 }, { "grad_norm": 0.15197992324829102, "learning_rate": 2.0746007265740846e-05, "loss": 0.0147, "step": 71400 }, { "grad_norm": 0.1515970379114151, "learning_rate": 2.0732599627371215e-05, "loss": 0.0137, "step": 71410 }, { "grad_norm": 0.1627858728170395, "learning_rate": 2.0719195189640468e-05, "loss": 0.0165, "step": 71420 }, { "grad_norm": 0.1826777458190918, "learning_rate": 2.0705793954014528e-05, "loss": 0.0194, "step": 71430 }, { "grad_norm": 0.13980980217456818, "learning_rate": 2.06923959219589e-05, "loss": 0.0135, "step": 71440 }, { "grad_norm": 0.16662092506885529, "learning_rate": 2.0679001094938793e-05, "loss": 0.0167, "step": 71450 }, { "grad_norm": 0.18360775709152222, "learning_rate": 2.066560947441904e-05, "loss": 0.0182, "step": 71460 }, { "grad_norm": 0.1761339157819748, "learning_rate": 2.0652221061864152e-05, "loss": 0.0155, "step": 71470 }, { "grad_norm": 0.19151003658771515, "learning_rate": 2.0638835858738226e-05, "loss": 0.0146, "step": 71480 }, { "grad_norm": 0.19114720821380615, "learning_rate": 2.0625453866505068e-05, "loss": 0.0152, "step": 71490 }, { "grad_norm": 0.12329945713281631, "learning_rate": 2.0612075086628123e-05, "loss": 0.0133, "step": 71500 }, { "grad_norm": 0.22541476786136627, "learning_rate": 2.059869952057045e-05, "loss": 0.0138, "step": 71510 }, { "grad_norm": 0.14084920287132263, "learning_rate": 2.0585327169794793e-05, "loss": 0.0124, "step": 71520 }, { "grad_norm": 0.14978189766407013, "learning_rate": 2.0571958035763534e-05, "loss": 0.018, "step": 71530 }, { "grad_norm": 0.2007283717393875, "learning_rate": 2.0558592119938715e-05, "loss": 0.0149, "step": 71540 }, { "grad_norm": 0.23421898484230042, "learning_rate": 2.054522942378198e-05, "loss": 0.0136, "step": 71550 }, { "grad_norm": 0.13105051219463348, "learning_rate": 2.053186994875469e-05, "loss": 0.0157, "step": 71560 }, { "grad_norm": 0.12909148633480072, "learning_rate": 2.0518513696317775e-05, "loss": 0.0135, "step": 71570 }, { "grad_norm": 0.14407891035079956, "learning_rate": 2.050516066793188e-05, "loss": 0.0141, "step": 71580 }, { "grad_norm": 0.28076159954071045, "learning_rate": 2.0491810865057276e-05, "loss": 0.0173, "step": 71590 }, { "grad_norm": 0.45246952772140503, "learning_rate": 2.0478464289153882e-05, "loss": 0.0168, "step": 71600 }, { "grad_norm": 0.19452673196792603, "learning_rate": 2.0465120941681232e-05, "loss": 0.0153, "step": 71610 }, { "grad_norm": 0.15531066060066223, "learning_rate": 2.0451780824098566e-05, "loss": 0.0155, "step": 71620 }, { "grad_norm": 0.15726032853126526, "learning_rate": 2.043844393786471e-05, "loss": 0.0135, "step": 71630 }, { "grad_norm": 0.22101566195487976, "learning_rate": 2.0425110284438193e-05, "loss": 0.0128, "step": 71640 }, { "grad_norm": 0.21225690841674805, "learning_rate": 2.041177986527712e-05, "loss": 0.0144, "step": 71650 }, { "grad_norm": 0.12724964320659637, "learning_rate": 2.0398452681839347e-05, "loss": 0.0133, "step": 71660 }, { "grad_norm": 0.20521534979343414, "learning_rate": 2.0385128735582272e-05, "loss": 0.0101, "step": 71670 }, { "grad_norm": 0.2081834077835083, "learning_rate": 2.0371808027963003e-05, "loss": 0.0169, "step": 71680 }, { "grad_norm": 0.13879293203353882, "learning_rate": 2.035849056043825e-05, "loss": 0.0144, "step": 71690 }, { "grad_norm": 0.14609020948410034, "learning_rate": 2.03451763344644e-05, "loss": 0.0147, "step": 71700 }, { "grad_norm": 0.16290497779846191, "learning_rate": 2.0331865351497482e-05, "loss": 0.0119, "step": 71710 }, { "grad_norm": 0.28210166096687317, "learning_rate": 2.0318557612993185e-05, "loss": 0.0129, "step": 71720 }, { "grad_norm": 0.1966543197631836, "learning_rate": 2.0305253120406782e-05, "loss": 0.0153, "step": 71730 }, { "grad_norm": 0.17260298132896423, "learning_rate": 2.029195187519325e-05, "loss": 0.0152, "step": 71740 }, { "grad_norm": 0.2004162222146988, "learning_rate": 2.0278653878807217e-05, "loss": 0.0108, "step": 71750 }, { "grad_norm": 0.31386372447013855, "learning_rate": 2.0265359132702892e-05, "loss": 0.0164, "step": 71760 }, { "grad_norm": 0.12466612458229065, "learning_rate": 2.0252067638334182e-05, "loss": 0.0127, "step": 71770 }, { "grad_norm": 0.20216603577136993, "learning_rate": 2.023877939715464e-05, "loss": 0.0163, "step": 71780 }, { "grad_norm": 0.17766572535037994, "learning_rate": 2.0225494410617448e-05, "loss": 0.0119, "step": 71790 }, { "grad_norm": 0.1864210069179535, "learning_rate": 2.02122126801754e-05, "loss": 0.0143, "step": 71800 }, { "grad_norm": 0.14748641848564148, "learning_rate": 2.0198934207281018e-05, "loss": 0.0173, "step": 71810 }, { "grad_norm": 0.18309853971004486, "learning_rate": 2.0185658993386357e-05, "loss": 0.0142, "step": 71820 }, { "grad_norm": 0.46937814354896545, "learning_rate": 2.0172387039943226e-05, "loss": 0.0136, "step": 71830 }, { "grad_norm": 0.15788739919662476, "learning_rate": 2.0159118348402968e-05, "loss": 0.0136, "step": 71840 }, { "grad_norm": 0.11595402657985687, "learning_rate": 2.0145852920216697e-05, "loss": 0.0147, "step": 71850 }, { "grad_norm": 0.1747692972421646, "learning_rate": 2.0132590756835045e-05, "loss": 0.0127, "step": 71860 }, { "grad_norm": 0.10399280488491058, "learning_rate": 2.0119331859708386e-05, "loss": 0.0111, "step": 71870 }, { "grad_norm": 0.16422709822654724, "learning_rate": 2.0106076230286654e-05, "loss": 0.0139, "step": 71880 }, { "grad_norm": 0.12363772094249725, "learning_rate": 2.0092823870019488e-05, "loss": 0.0163, "step": 71890 }, { "grad_norm": 0.13242100179195404, "learning_rate": 2.0079574780356115e-05, "loss": 0.0129, "step": 71900 }, { "grad_norm": 0.2519669234752655, "learning_rate": 2.0066328962745486e-05, "loss": 0.0171, "step": 71910 }, { "grad_norm": 0.17108245193958282, "learning_rate": 2.00530864186361e-05, "loss": 0.0118, "step": 71920 }, { "grad_norm": 0.1835384964942932, "learning_rate": 2.003984714947618e-05, "loss": 0.0168, "step": 71930 }, { "grad_norm": 0.19186139106750488, "learning_rate": 2.0026611156713515e-05, "loss": 0.0146, "step": 71940 }, { "grad_norm": 0.14150094985961914, "learning_rate": 2.0013378441795584e-05, "loss": 0.0161, "step": 71950 }, { "grad_norm": 0.25159168243408203, "learning_rate": 2.0000149006169506e-05, "loss": 0.0147, "step": 71960 }, { "grad_norm": 0.16222795844078064, "learning_rate": 1.9986922851282048e-05, "loss": 0.0144, "step": 71970 }, { "grad_norm": 0.16208955645561218, "learning_rate": 1.997369997857956e-05, "loss": 0.0144, "step": 71980 }, { "grad_norm": 0.2118726521730423, "learning_rate": 1.9960480389508103e-05, "loss": 0.0123, "step": 71990 }, { "grad_norm": 0.1457725465297699, "learning_rate": 1.9947264085513363e-05, "loss": 0.0149, "step": 72000 }, { "grad_norm": 0.5248107314109802, "learning_rate": 1.993405106804062e-05, "loss": 0.0162, "step": 72010 }, { "grad_norm": 0.2693701982498169, "learning_rate": 1.9920841338534867e-05, "loss": 0.0144, "step": 72020 }, { "grad_norm": 0.13911288976669312, "learning_rate": 1.9907634898440646e-05, "loss": 0.0124, "step": 72030 }, { "grad_norm": 0.21838633716106415, "learning_rate": 1.989443174920227e-05, "loss": 0.0149, "step": 72040 }, { "grad_norm": 0.11990159004926682, "learning_rate": 1.988123189226355e-05, "loss": 0.0151, "step": 72050 }, { "grad_norm": 0.19480977952480316, "learning_rate": 1.986803532906805e-05, "loss": 0.0142, "step": 72060 }, { "grad_norm": 0.1581392139196396, "learning_rate": 1.985484206105889e-05, "loss": 0.0129, "step": 72070 }, { "grad_norm": 0.18919220566749573, "learning_rate": 1.9841652089678887e-05, "loss": 0.0193, "step": 72080 }, { "grad_norm": 0.1503480225801468, "learning_rate": 1.9828465416370435e-05, "loss": 0.0159, "step": 72090 }, { "grad_norm": 0.18236887454986572, "learning_rate": 1.9815282042575685e-05, "loss": 0.0127, "step": 72100 }, { "grad_norm": 0.27288660407066345, "learning_rate": 1.9802101969736285e-05, "loss": 0.018, "step": 72110 }, { "grad_norm": 0.29761096835136414, "learning_rate": 1.9788925199293633e-05, "loss": 0.0153, "step": 72120 }, { "grad_norm": 0.15250712633132935, "learning_rate": 1.977575173268868e-05, "loss": 0.014, "step": 72130 }, { "grad_norm": 0.5254316329956055, "learning_rate": 1.9762581571362094e-05, "loss": 0.0148, "step": 72140 }, { "grad_norm": 0.1250302493572235, "learning_rate": 1.9749414716754112e-05, "loss": 0.0163, "step": 72150 }, { "grad_norm": 0.16993507742881775, "learning_rate": 1.973625117030465e-05, "loss": 0.0137, "step": 72160 }, { "grad_norm": 0.17875461280345917, "learning_rate": 1.972309093345326e-05, "loss": 0.0144, "step": 72170 }, { "grad_norm": 0.18893127143383026, "learning_rate": 1.970993400763915e-05, "loss": 0.0159, "step": 72180 }, { "grad_norm": 0.15866655111312866, "learning_rate": 1.969678039430109e-05, "loss": 0.0154, "step": 72190 }, { "grad_norm": 0.22146578133106232, "learning_rate": 1.9683630094877564e-05, "loss": 0.0176, "step": 72200 }, { "grad_norm": 0.13439074158668518, "learning_rate": 1.9670483110806696e-05, "loss": 0.0154, "step": 72210 }, { "grad_norm": 0.17510755360126495, "learning_rate": 1.9657339443526172e-05, "loss": 0.0161, "step": 72220 }, { "grad_norm": 0.09593566507101059, "learning_rate": 1.964419909447339e-05, "loss": 0.0135, "step": 72230 }, { "grad_norm": 0.22107554972171783, "learning_rate": 1.963106206508536e-05, "loss": 0.017, "step": 72240 }, { "grad_norm": 0.160752072930336, "learning_rate": 1.9617928356798738e-05, "loss": 0.0154, "step": 72250 }, { "grad_norm": 0.3251427114009857, "learning_rate": 1.960479797104977e-05, "loss": 0.0128, "step": 72260 }, { "grad_norm": 0.1754661351442337, "learning_rate": 1.9591670909274418e-05, "loss": 0.0126, "step": 72270 }, { "grad_norm": 0.15465857088565826, "learning_rate": 1.9578547172908184e-05, "loss": 0.0142, "step": 72280 }, { "grad_norm": 0.1345251202583313, "learning_rate": 1.9565426763386325e-05, "loss": 0.014, "step": 72290 }, { "grad_norm": 0.22115600109100342, "learning_rate": 1.9552309682143622e-05, "loss": 0.0169, "step": 72300 }, { "grad_norm": 0.1465662717819214, "learning_rate": 1.9539195930614574e-05, "loss": 0.0124, "step": 72310 }, { "grad_norm": 0.17701581120491028, "learning_rate": 1.952608551023324e-05, "loss": 0.0129, "step": 72320 }, { "grad_norm": 0.17773206532001495, "learning_rate": 1.9512978422433397e-05, "loss": 0.0159, "step": 72330 }, { "grad_norm": 0.17254823446273804, "learning_rate": 1.9499874668648376e-05, "loss": 0.0144, "step": 72340 }, { "grad_norm": 0.12841598689556122, "learning_rate": 1.9486774250311213e-05, "loss": 0.0114, "step": 72350 }, { "grad_norm": 0.3157528340816498, "learning_rate": 1.9473677168854537e-05, "loss": 0.0138, "step": 72360 }, { "grad_norm": 0.1630229651927948, "learning_rate": 1.946058342571065e-05, "loss": 0.0132, "step": 72370 }, { "grad_norm": 0.13903681933879852, "learning_rate": 1.9447493022311424e-05, "loss": 0.0133, "step": 72380 }, { "grad_norm": 0.13446855545043945, "learning_rate": 1.9434405960088442e-05, "loss": 0.013, "step": 72390 }, { "grad_norm": 0.12278703600168228, "learning_rate": 1.9421322240472857e-05, "loss": 0.0124, "step": 72400 }, { "grad_norm": 0.17631609737873077, "learning_rate": 1.9408241864895494e-05, "loss": 0.0145, "step": 72410 }, { "grad_norm": 0.1233539953827858, "learning_rate": 1.939516483478681e-05, "loss": 0.0132, "step": 72420 }, { "grad_norm": 0.19227604568004608, "learning_rate": 1.9382091151576902e-05, "loss": 0.011, "step": 72430 }, { "grad_norm": 0.20853689312934875, "learning_rate": 1.9369020816695454e-05, "loss": 0.0143, "step": 72440 }, { "grad_norm": 0.1894008368253708, "learning_rate": 1.935595383157184e-05, "loss": 0.0153, "step": 72450 }, { "grad_norm": 0.17868083715438843, "learning_rate": 1.9342890197635057e-05, "loss": 0.0167, "step": 72460 }, { "grad_norm": 0.19830550253391266, "learning_rate": 1.9329829916313685e-05, "loss": 0.0152, "step": 72470 }, { "grad_norm": 0.2506372928619385, "learning_rate": 1.9316772989036013e-05, "loss": 0.0134, "step": 72480 }, { "grad_norm": 0.17179147899150848, "learning_rate": 1.930371941722991e-05, "loss": 0.0151, "step": 72490 }, { "grad_norm": 0.19848795235157013, "learning_rate": 1.929066920232292e-05, "loss": 0.0153, "step": 72500 }, { "grad_norm": 0.23941664397716522, "learning_rate": 1.9277622345742152e-05, "loss": 0.0149, "step": 72510 }, { "grad_norm": 0.17291133105754852, "learning_rate": 1.926457884891444e-05, "loss": 0.0148, "step": 72520 }, { "grad_norm": 0.20204482972621918, "learning_rate": 1.9251538713266147e-05, "loss": 0.0119, "step": 72530 }, { "grad_norm": 0.19171467423439026, "learning_rate": 1.9238501940223358e-05, "loss": 0.0131, "step": 72540 }, { "grad_norm": 0.13856995105743408, "learning_rate": 1.922546853121175e-05, "loss": 0.0116, "step": 72550 }, { "grad_norm": 0.23501738905906677, "learning_rate": 1.9212438487656647e-05, "loss": 0.0123, "step": 72560 }, { "grad_norm": 0.17836496233940125, "learning_rate": 1.9199411810982965e-05, "loss": 0.0118, "step": 72570 }, { "grad_norm": 0.2049369513988495, "learning_rate": 1.918638850261532e-05, "loss": 0.0141, "step": 72580 }, { "grad_norm": 0.17087435722351074, "learning_rate": 1.9173368563977885e-05, "loss": 0.0142, "step": 72590 }, { "grad_norm": 0.203935444355011, "learning_rate": 1.9160351996494513e-05, "loss": 0.0129, "step": 72600 }, { "grad_norm": 0.27907267212867737, "learning_rate": 1.9147338801588677e-05, "loss": 0.0129, "step": 72610 }, { "grad_norm": 0.1996413618326187, "learning_rate": 1.913432898068351e-05, "loss": 0.0156, "step": 72620 }, { "grad_norm": 0.16959427297115326, "learning_rate": 1.912132253520169e-05, "loss": 0.0132, "step": 72630 }, { "grad_norm": 0.3567604720592499, "learning_rate": 1.910831946656564e-05, "loss": 0.0118, "step": 72640 }, { "grad_norm": 0.22221772372722626, "learning_rate": 1.909531977619731e-05, "loss": 0.0124, "step": 72650 }, { "grad_norm": 0.26689502596855164, "learning_rate": 1.908232346551834e-05, "loss": 0.0158, "step": 72660 }, { "grad_norm": 0.144389808177948, "learning_rate": 1.9069330535949993e-05, "loss": 0.0162, "step": 72670 }, { "grad_norm": 0.19776830077171326, "learning_rate": 1.9056340988913167e-05, "loss": 0.0165, "step": 72680 }, { "grad_norm": 0.2199144959449768, "learning_rate": 1.9043354825828357e-05, "loss": 0.014, "step": 72690 }, { "grad_norm": 0.2617785632610321, "learning_rate": 1.9030372048115724e-05, "loss": 0.0134, "step": 72700 }, { "grad_norm": 0.15552276372909546, "learning_rate": 1.9017392657195056e-05, "loss": 0.0122, "step": 72710 }, { "grad_norm": 0.11894167959690094, "learning_rate": 1.9004416654485724e-05, "loss": 0.0156, "step": 72720 }, { "grad_norm": 0.15959827601909637, "learning_rate": 1.8991444041406785e-05, "loss": 0.0134, "step": 72730 }, { "grad_norm": 0.12076658755540848, "learning_rate": 1.8978474819376912e-05, "loss": 0.0132, "step": 72740 }, { "grad_norm": 0.1349000483751297, "learning_rate": 1.8965508989814406e-05, "loss": 0.0133, "step": 72750 }, { "grad_norm": 0.12977048754692078, "learning_rate": 1.8952546554137158e-05, "loss": 0.0145, "step": 72760 }, { "grad_norm": 0.22622106969356537, "learning_rate": 1.8939587513762757e-05, "loss": 0.014, "step": 72770 }, { "grad_norm": 0.19288745522499084, "learning_rate": 1.892663187010835e-05, "loss": 0.0186, "step": 72780 }, { "grad_norm": 0.17607924342155457, "learning_rate": 1.891367962459077e-05, "loss": 0.013, "step": 72790 }, { "grad_norm": 0.1616939753293991, "learning_rate": 1.890073077862644e-05, "loss": 0.0113, "step": 72800 }, { "grad_norm": 0.2095443457365036, "learning_rate": 1.888778533363146e-05, "loss": 0.0128, "step": 72810 }, { "grad_norm": 0.19239743053913116, "learning_rate": 1.8874843291021478e-05, "loss": 0.0146, "step": 72820 }, { "grad_norm": 0.18446172773838043, "learning_rate": 1.8861904652211853e-05, "loss": 0.0148, "step": 72830 }, { "grad_norm": 0.21746820211410522, "learning_rate": 1.8848969418617507e-05, "loss": 0.0146, "step": 72840 }, { "grad_norm": 0.1300583779811859, "learning_rate": 1.8836037591653044e-05, "loss": 0.0143, "step": 72850 }, { "grad_norm": 0.16960154473781586, "learning_rate": 1.882310917273262e-05, "loss": 0.0161, "step": 72860 }, { "grad_norm": 0.1535622924566269, "learning_rate": 1.881018416327014e-05, "loss": 0.0142, "step": 72870 }, { "grad_norm": 0.23155620694160461, "learning_rate": 1.8797262564679014e-05, "loss": 0.0143, "step": 72880 }, { "grad_norm": 0.1617467701435089, "learning_rate": 1.8784344378372355e-05, "loss": 0.011, "step": 72890 }, { "grad_norm": 0.18569372594356537, "learning_rate": 1.8771429605762843e-05, "loss": 0.0158, "step": 72900 }, { "grad_norm": 0.2100868821144104, "learning_rate": 1.8758518248262835e-05, "loss": 0.0154, "step": 72910 }, { "grad_norm": 0.1351066380739212, "learning_rate": 1.8745610307284305e-05, "loss": 0.0134, "step": 72920 }, { "grad_norm": 0.4645369350910187, "learning_rate": 1.8732705784238862e-05, "loss": 0.0163, "step": 72930 }, { "grad_norm": 0.17283402383327484, "learning_rate": 1.8719804680537682e-05, "loss": 0.011, "step": 72940 }, { "grad_norm": 0.1851184070110321, "learning_rate": 1.870690699759164e-05, "loss": 0.0126, "step": 72950 }, { "grad_norm": 0.17305375635623932, "learning_rate": 1.8694012736811213e-05, "loss": 0.0168, "step": 72960 }, { "grad_norm": 0.16331474483013153, "learning_rate": 1.8681121899606463e-05, "loss": 0.0145, "step": 72970 }, { "grad_norm": 0.15223345160484314, "learning_rate": 1.8668234487387144e-05, "loss": 0.013, "step": 72980 }, { "grad_norm": 0.1879952996969223, "learning_rate": 1.8655350501562592e-05, "loss": 0.0129, "step": 72990 }, { "grad_norm": 0.22963622212409973, "learning_rate": 1.8642469943541796e-05, "loss": 0.0144, "step": 73000 }, { "grad_norm": 0.18692024052143097, "learning_rate": 1.8629592814733328e-05, "loss": 0.0106, "step": 73010 }, { "grad_norm": 0.17262177169322968, "learning_rate": 1.8616719116545438e-05, "loss": 0.0133, "step": 73020 }, { "grad_norm": 0.1973821371793747, "learning_rate": 1.8603848850385952e-05, "loss": 0.0132, "step": 73030 }, { "grad_norm": 0.12264829874038696, "learning_rate": 1.859098201766236e-05, "loss": 0.0149, "step": 73040 }, { "grad_norm": 0.1491314023733139, "learning_rate": 1.8578118619781726e-05, "loss": 0.0107, "step": 73050 }, { "grad_norm": 0.15254022181034088, "learning_rate": 1.856525865815083e-05, "loss": 0.0123, "step": 73060 }, { "grad_norm": 0.12882307171821594, "learning_rate": 1.8552402134175966e-05, "loss": 0.0151, "step": 73070 }, { "grad_norm": 0.2306426763534546, "learning_rate": 1.8539549049263143e-05, "loss": 0.0141, "step": 73080 }, { "grad_norm": 0.31039389967918396, "learning_rate": 1.8526699404817917e-05, "loss": 0.015, "step": 73090 }, { "grad_norm": 0.11151883751153946, "learning_rate": 1.8513853202245546e-05, "loss": 0.0129, "step": 73100 }, { "grad_norm": 0.20804712176322937, "learning_rate": 1.850101044295081e-05, "loss": 0.0123, "step": 73110 }, { "grad_norm": 0.12611007690429688, "learning_rate": 1.848817112833825e-05, "loss": 0.0112, "step": 73120 }, { "grad_norm": 0.16361600160598755, "learning_rate": 1.8475335259811905e-05, "loss": 0.0169, "step": 73130 }, { "grad_norm": 0.33968913555145264, "learning_rate": 1.8462502838775508e-05, "loss": 0.0124, "step": 73140 }, { "grad_norm": 0.16802990436553955, "learning_rate": 1.8449673866632378e-05, "loss": 0.0158, "step": 73150 }, { "grad_norm": 0.22162573039531708, "learning_rate": 1.8436848344785473e-05, "loss": 0.0175, "step": 73160 }, { "grad_norm": 0.16111789643764496, "learning_rate": 1.8424026274637378e-05, "loss": 0.0174, "step": 73170 }, { "grad_norm": 0.1905493438243866, "learning_rate": 1.8411207657590312e-05, "loss": 0.0145, "step": 73180 }, { "grad_norm": 0.12447857856750488, "learning_rate": 1.839839249504607e-05, "loss": 0.0146, "step": 73190 }, { "grad_norm": 0.24845249950885773, "learning_rate": 1.8385580788406108e-05, "loss": 0.017, "step": 73200 }, { "grad_norm": 0.1158752515912056, "learning_rate": 1.837277253907152e-05, "loss": 0.015, "step": 73210 }, { "grad_norm": 0.15900707244873047, "learning_rate": 1.8359967748442957e-05, "loss": 0.0177, "step": 73220 }, { "grad_norm": 0.1721220463514328, "learning_rate": 1.8347166417920774e-05, "loss": 0.0147, "step": 73230 }, { "grad_norm": 0.11065667122602463, "learning_rate": 1.833436854890484e-05, "loss": 0.0136, "step": 73240 }, { "grad_norm": 0.1601477861404419, "learning_rate": 1.8321574142794796e-05, "loss": 0.0134, "step": 73250 }, { "grad_norm": 0.1960432529449463, "learning_rate": 1.8308783200989748e-05, "loss": 0.0142, "step": 73260 }, { "grad_norm": 0.19232818484306335, "learning_rate": 1.829599572488855e-05, "loss": 0.0144, "step": 73270 }, { "grad_norm": 0.17217223346233368, "learning_rate": 1.828321171588957e-05, "loss": 0.0126, "step": 73280 }, { "grad_norm": 0.21253971755504608, "learning_rate": 1.827043117539089e-05, "loss": 0.0161, "step": 73290 }, { "grad_norm": 0.1621832549571991, "learning_rate": 1.8257654104790116e-05, "loss": 0.013, "step": 73300 }, { "grad_norm": 0.17742544412612915, "learning_rate": 1.82448805054846e-05, "loss": 0.0127, "step": 73310 }, { "grad_norm": 0.12229517102241516, "learning_rate": 1.823211037887119e-05, "loss": 0.0127, "step": 73320 }, { "grad_norm": 0.2040814906358719, "learning_rate": 1.821934372634645e-05, "loss": 0.0153, "step": 73330 }, { "grad_norm": 0.1587880551815033, "learning_rate": 1.8206580549306472e-05, "loss": 0.015, "step": 73340 }, { "grad_norm": 0.17099779844284058, "learning_rate": 1.819382084914707e-05, "loss": 0.0135, "step": 73350 }, { "grad_norm": 0.15189798176288605, "learning_rate": 1.8181064627263577e-05, "loss": 0.0128, "step": 73360 }, { "grad_norm": 0.22035056352615356, "learning_rate": 1.8168311885051022e-05, "loss": 0.0142, "step": 73370 }, { "grad_norm": 0.14569909870624542, "learning_rate": 1.815556262390402e-05, "loss": 0.0146, "step": 73380 }, { "grad_norm": 0.20042386651039124, "learning_rate": 1.8142816845216826e-05, "loss": 0.0146, "step": 73390 }, { "grad_norm": 0.10257252305746078, "learning_rate": 1.8130074550383276e-05, "loss": 0.0133, "step": 73400 }, { "grad_norm": 0.17115788161754608, "learning_rate": 1.8117335740796848e-05, "loss": 0.0134, "step": 73410 }, { "grad_norm": 0.13459782302379608, "learning_rate": 1.810460041785067e-05, "loss": 0.0128, "step": 73420 }, { "grad_norm": 0.2442166954278946, "learning_rate": 1.809186858293742e-05, "loss": 0.0138, "step": 73430 }, { "grad_norm": 0.22246335446834564, "learning_rate": 1.8079140237449448e-05, "loss": 0.0147, "step": 73440 }, { "grad_norm": 0.2720659375190735, "learning_rate": 1.8066415382778707e-05, "loss": 0.0166, "step": 73450 }, { "grad_norm": 0.276059627532959, "learning_rate": 1.8053694020316786e-05, "loss": 0.0133, "step": 73460 }, { "grad_norm": 0.13020262122154236, "learning_rate": 1.8040976151454837e-05, "loss": 0.0181, "step": 73470 }, { "grad_norm": 0.13930286467075348, "learning_rate": 1.8028261777583704e-05, "loss": 0.0119, "step": 73480 }, { "grad_norm": 0.1924474686384201, "learning_rate": 1.8015550900093758e-05, "loss": 0.0179, "step": 73490 }, { "grad_norm": 0.15296269953250885, "learning_rate": 1.800284352037512e-05, "loss": 0.0128, "step": 73500 }, { "grad_norm": 0.17683742940425873, "learning_rate": 1.799013963981738e-05, "loss": 0.017, "step": 73510 }, { "grad_norm": 0.16770197451114655, "learning_rate": 1.797743925980987e-05, "loss": 0.0153, "step": 73520 }, { "grad_norm": 0.1951422095298767, "learning_rate": 1.7964742381741435e-05, "loss": 0.0122, "step": 73530 }, { "grad_norm": 0.19654813408851624, "learning_rate": 1.7952049007000627e-05, "loss": 0.0132, "step": 73540 }, { "grad_norm": 0.15890806913375854, "learning_rate": 1.7939359136975535e-05, "loss": 0.0163, "step": 73550 }, { "grad_norm": 0.23648206889629364, "learning_rate": 1.7926672773053932e-05, "loss": 0.0136, "step": 73560 }, { "grad_norm": 0.2646712362766266, "learning_rate": 1.791398991662317e-05, "loss": 0.0116, "step": 73570 }, { "grad_norm": 0.1355496346950531, "learning_rate": 1.7901310569070244e-05, "loss": 0.0119, "step": 73580 }, { "grad_norm": 0.12590140104293823, "learning_rate": 1.7888634731781724e-05, "loss": 0.0124, "step": 73590 }, { "grad_norm": 0.19165584444999695, "learning_rate": 1.7875962406143844e-05, "loss": 0.0144, "step": 73600 }, { "grad_norm": 0.1749468296766281, "learning_rate": 1.7863293593542403e-05, "loss": 0.0156, "step": 73610 }, { "grad_norm": 0.17395280301570892, "learning_rate": 1.7850628295362853e-05, "loss": 0.013, "step": 73620 }, { "grad_norm": 0.15117935836315155, "learning_rate": 1.7837966512990263e-05, "loss": 0.0154, "step": 73630 }, { "grad_norm": 0.12460926175117493, "learning_rate": 1.782530824780931e-05, "loss": 0.0172, "step": 73640 }, { "grad_norm": 0.4411470592021942, "learning_rate": 1.7812653501204263e-05, "loss": 0.0196, "step": 73650 }, { "grad_norm": 0.20383337140083313, "learning_rate": 1.780000227455903e-05, "loss": 0.0168, "step": 73660 }, { "grad_norm": 0.13775376975536346, "learning_rate": 1.778735456925716e-05, "loss": 0.0139, "step": 73670 }, { "grad_norm": 0.1561727225780487, "learning_rate": 1.7774710386681737e-05, "loss": 0.0143, "step": 73680 }, { "grad_norm": 0.22487519681453705, "learning_rate": 1.7762069728215545e-05, "loss": 0.0171, "step": 73690 }, { "grad_norm": 0.175393745303154, "learning_rate": 1.7749432595240932e-05, "loss": 0.0142, "step": 73700 }, { "grad_norm": 0.17497889697551727, "learning_rate": 1.773679898913991e-05, "loss": 0.0134, "step": 73710 }, { "grad_norm": 0.188021719455719, "learning_rate": 1.7724168911294015e-05, "loss": 0.0145, "step": 73720 }, { "grad_norm": 0.18619681894779205, "learning_rate": 1.7711542363084504e-05, "loss": 0.0137, "step": 73730 }, { "grad_norm": 0.23603607714176178, "learning_rate": 1.7698919345892158e-05, "loss": 0.014, "step": 73740 }, { "grad_norm": 0.20591925084590912, "learning_rate": 1.768629986109743e-05, "loss": 0.0136, "step": 73750 }, { "grad_norm": 0.21566736698150635, "learning_rate": 1.7673683910080357e-05, "loss": 0.0153, "step": 73760 }, { "grad_norm": 0.12981626391410828, "learning_rate": 1.7661071494220638e-05, "loss": 0.0131, "step": 73770 }, { "grad_norm": 0.16278530657291412, "learning_rate": 1.7648462614897498e-05, "loss": 0.0142, "step": 73780 }, { "grad_norm": 0.1865449994802475, "learning_rate": 1.7635857273489863e-05, "loss": 0.0141, "step": 73790 }, { "grad_norm": 0.16814523935317993, "learning_rate": 1.7623255471376198e-05, "loss": 0.0128, "step": 73800 }, { "grad_norm": 0.15808261930942535, "learning_rate": 1.7610657209934646e-05, "loss": 0.014, "step": 73810 }, { "grad_norm": 0.23999080061912537, "learning_rate": 1.7598062490542915e-05, "loss": 0.0158, "step": 73820 }, { "grad_norm": 0.15941306948661804, "learning_rate": 1.7585471314578377e-05, "loss": 0.0133, "step": 73830 }, { "grad_norm": 0.11007320135831833, "learning_rate": 1.7572883683417945e-05, "loss": 0.012, "step": 73840 }, { "grad_norm": 0.24786148965358734, "learning_rate": 1.7560299598438212e-05, "loss": 0.015, "step": 73850 }, { "grad_norm": 0.13355769217014313, "learning_rate": 1.7547719061015334e-05, "loss": 0.0143, "step": 73860 }, { "grad_norm": 0.10480976104736328, "learning_rate": 1.75351420725251e-05, "loss": 0.0124, "step": 73870 }, { "grad_norm": 0.21277746558189392, "learning_rate": 1.7522568634342923e-05, "loss": 0.0175, "step": 73880 }, { "grad_norm": 0.24681991338729858, "learning_rate": 1.7509998747843825e-05, "loss": 0.0151, "step": 73890 }, { "grad_norm": 0.1496397703886032, "learning_rate": 1.7497432414402403e-05, "loss": 0.0163, "step": 73900 }, { "grad_norm": 0.15832889080047607, "learning_rate": 1.7484869635392903e-05, "loss": 0.016, "step": 73910 }, { "grad_norm": 0.19385385513305664, "learning_rate": 1.747231041218919e-05, "loss": 0.0172, "step": 73920 }, { "grad_norm": 0.18959768116474152, "learning_rate": 1.7459754746164692e-05, "loss": 0.0142, "step": 73930 }, { "grad_norm": 0.4021744728088379, "learning_rate": 1.744720263869249e-05, "loss": 0.0166, "step": 73940 }, { "grad_norm": 0.22060611844062805, "learning_rate": 1.7434654091145263e-05, "loss": 0.0122, "step": 73950 }, { "grad_norm": 0.11590785533189774, "learning_rate": 1.742210910489532e-05, "loss": 0.0127, "step": 73960 }, { "grad_norm": 0.20698601007461548, "learning_rate": 1.7409567681314532e-05, "loss": 0.0123, "step": 73970 }, { "grad_norm": 0.12469017505645752, "learning_rate": 1.7397029821774434e-05, "loss": 0.0127, "step": 73980 }, { "grad_norm": 0.26521551609039307, "learning_rate": 1.7384495527646126e-05, "loss": 0.016, "step": 73990 }, { "grad_norm": 0.15106365084648132, "learning_rate": 1.737196480030035e-05, "loss": 0.0127, "step": 74000 }, { "grad_norm": 0.12145514786243439, "learning_rate": 1.7359437641107447e-05, "loss": 0.0142, "step": 74010 }, { "grad_norm": 0.14500875771045685, "learning_rate": 1.7346914051437387e-05, "loss": 0.0141, "step": 74020 }, { "grad_norm": 0.17800788581371307, "learning_rate": 1.733439403265969e-05, "loss": 0.0134, "step": 74030 }, { "grad_norm": 0.13245058059692383, "learning_rate": 1.732187758614357e-05, "loss": 0.0109, "step": 74040 }, { "grad_norm": 0.1595151424407959, "learning_rate": 1.730936471325777e-05, "loss": 0.0143, "step": 74050 }, { "grad_norm": 0.19387303292751312, "learning_rate": 1.7296855415370718e-05, "loss": 0.0119, "step": 74060 }, { "grad_norm": 0.21405109763145447, "learning_rate": 1.728434969385035e-05, "loss": 0.0178, "step": 74070 }, { "grad_norm": 0.32538899779319763, "learning_rate": 1.7271847550064346e-05, "loss": 0.0147, "step": 74080 }, { "grad_norm": 0.1514934003353119, "learning_rate": 1.725934898537987e-05, "loss": 0.0156, "step": 74090 }, { "grad_norm": 0.14032122492790222, "learning_rate": 1.7246854001163783e-05, "loss": 0.0112, "step": 74100 }, { "grad_norm": 0.1794218271970749, "learning_rate": 1.7234362598782476e-05, "loss": 0.0113, "step": 74110 }, { "grad_norm": 0.14085924625396729, "learning_rate": 1.722187477960202e-05, "loss": 0.0149, "step": 74120 }, { "grad_norm": 0.15359458327293396, "learning_rate": 1.7209390544988057e-05, "loss": 0.017, "step": 74130 }, { "grad_norm": 0.1230798065662384, "learning_rate": 1.7196909896305856e-05, "loss": 0.013, "step": 74140 }, { "grad_norm": 0.1678561270236969, "learning_rate": 1.718443283492026e-05, "loss": 0.0107, "step": 74150 }, { "grad_norm": 0.19020240008831024, "learning_rate": 1.7171959362195746e-05, "loss": 0.0123, "step": 74160 }, { "grad_norm": 0.13412562012672424, "learning_rate": 1.715948947949642e-05, "loss": 0.0154, "step": 74170 }, { "grad_norm": 0.18208399415016174, "learning_rate": 1.714702318818593e-05, "loss": 0.0148, "step": 74180 }, { "grad_norm": 0.16727448999881744, "learning_rate": 1.7134560489627592e-05, "loss": 0.0146, "step": 74190 }, { "grad_norm": 0.1473810076713562, "learning_rate": 1.712210138518431e-05, "loss": 0.0154, "step": 74200 }, { "grad_norm": 0.2015390843153, "learning_rate": 1.71096458762186e-05, "loss": 0.014, "step": 74210 }, { "grad_norm": 0.2429816871881485, "learning_rate": 1.7097193964092552e-05, "loss": 0.0121, "step": 74220 }, { "grad_norm": 0.22141999006271362, "learning_rate": 1.708474565016792e-05, "loss": 0.0136, "step": 74230 }, { "grad_norm": 0.2345903068780899, "learning_rate": 1.7072300935806e-05, "loss": 0.0118, "step": 74240 }, { "grad_norm": 0.14450332522392273, "learning_rate": 1.705985982236776e-05, "loss": 0.0134, "step": 74250 }, { "grad_norm": 0.15243135392665863, "learning_rate": 1.7047422311213697e-05, "loss": 0.0131, "step": 74260 }, { "grad_norm": 0.1352868229150772, "learning_rate": 1.703498840370401e-05, "loss": 0.0127, "step": 74270 }, { "grad_norm": 0.15879730880260468, "learning_rate": 1.702255810119842e-05, "loss": 0.0092, "step": 74280 }, { "grad_norm": 0.12631049752235413, "learning_rate": 1.7010131405056305e-05, "loss": 0.0139, "step": 74290 }, { "grad_norm": 0.12283727526664734, "learning_rate": 1.6997708316636613e-05, "loss": 0.0145, "step": 74300 }, { "grad_norm": 0.29477185010910034, "learning_rate": 1.6985288837297932e-05, "loss": 0.011, "step": 74310 }, { "grad_norm": 0.1617862582206726, "learning_rate": 1.6972872968398394e-05, "loss": 0.0131, "step": 74320 }, { "grad_norm": 0.14487706124782562, "learning_rate": 1.6960460711295845e-05, "loss": 0.0131, "step": 74330 }, { "grad_norm": 0.10493557900190353, "learning_rate": 1.6948052067347623e-05, "loss": 0.0142, "step": 74340 }, { "grad_norm": 0.2346217930316925, "learning_rate": 1.693564703791075e-05, "loss": 0.015, "step": 74350 }, { "grad_norm": 0.17368946969509125, "learning_rate": 1.6923245624341793e-05, "loss": 0.0127, "step": 74360 }, { "grad_norm": 0.1519494503736496, "learning_rate": 1.6910847827996962e-05, "loss": 0.017, "step": 74370 }, { "grad_norm": 0.22425618767738342, "learning_rate": 1.689845365023206e-05, "loss": 0.0134, "step": 74380 }, { "grad_norm": 0.2184961438179016, "learning_rate": 1.6886063092402505e-05, "loss": 0.013, "step": 74390 }, { "grad_norm": 0.2203354686498642, "learning_rate": 1.6873676155863317e-05, "loss": 0.015, "step": 74400 }, { "grad_norm": 0.12818026542663574, "learning_rate": 1.6861292841969085e-05, "loss": 0.0137, "step": 74410 }, { "grad_norm": 0.1987779140472412, "learning_rate": 1.684891315207406e-05, "loss": 0.0164, "step": 74420 }, { "grad_norm": 0.23993834853172302, "learning_rate": 1.6836537087532034e-05, "loss": 0.0151, "step": 74430 }, { "grad_norm": 0.20482851564884186, "learning_rate": 1.6824164649696465e-05, "loss": 0.0159, "step": 74440 }, { "grad_norm": 0.16784396767616272, "learning_rate": 1.681179583992034e-05, "loss": 0.0185, "step": 74450 }, { "grad_norm": 0.15172110497951508, "learning_rate": 1.679943065955636e-05, "loss": 0.0157, "step": 74460 }, { "grad_norm": 0.13379915058612823, "learning_rate": 1.678706910995671e-05, "loss": 0.013, "step": 74470 }, { "grad_norm": 0.2126183956861496, "learning_rate": 1.6774711192473262e-05, "loss": 0.014, "step": 74480 }, { "grad_norm": 0.19237120449543, "learning_rate": 1.676235690845743e-05, "loss": 0.0144, "step": 74490 }, { "grad_norm": 0.12801887094974518, "learning_rate": 1.6750006259260285e-05, "loss": 0.0134, "step": 74500 }, { "grad_norm": 0.1778131127357483, "learning_rate": 1.673765924623244e-05, "loss": 0.014, "step": 74510 }, { "grad_norm": 0.1669495850801468, "learning_rate": 1.67253158707242e-05, "loss": 0.0125, "step": 74520 }, { "grad_norm": 0.20955997705459595, "learning_rate": 1.671297613408537e-05, "loss": 0.0137, "step": 74530 }, { "grad_norm": 0.16468168795108795, "learning_rate": 1.6700640037665444e-05, "loss": 0.0135, "step": 74540 }, { "grad_norm": 0.1587914079427719, "learning_rate": 1.668830758281344e-05, "loss": 0.0143, "step": 74550 }, { "grad_norm": 0.23336325585842133, "learning_rate": 1.6675978770878052e-05, "loss": 0.0152, "step": 74560 }, { "grad_norm": 0.2242891490459442, "learning_rate": 1.6663653603207506e-05, "loss": 0.0143, "step": 74570 }, { "grad_norm": 0.23155122995376587, "learning_rate": 1.6651332081149685e-05, "loss": 0.0144, "step": 74580 }, { "grad_norm": 0.12114818394184113, "learning_rate": 1.6639014206052046e-05, "loss": 0.0151, "step": 74590 }, { "grad_norm": 0.6179866790771484, "learning_rate": 1.6626699979261657e-05, "loss": 0.017, "step": 74600 }, { "grad_norm": 0.18545137345790863, "learning_rate": 1.6614389402125197e-05, "loss": 0.0126, "step": 74610 }, { "grad_norm": 0.15669560432434082, "learning_rate": 1.66020824759889e-05, "loss": 0.0182, "step": 74620 }, { "grad_norm": 0.1864587962627411, "learning_rate": 1.6589779202198675e-05, "loss": 0.0152, "step": 74630 }, { "grad_norm": 0.3178178668022156, "learning_rate": 1.6577479582099924e-05, "loss": 0.0176, "step": 74640 }, { "grad_norm": 0.24046045541763306, "learning_rate": 1.6565183617037788e-05, "loss": 0.016, "step": 74650 }, { "grad_norm": 0.15854553878307343, "learning_rate": 1.6552891308356894e-05, "loss": 0.0132, "step": 74660 }, { "grad_norm": 0.11208083480596542, "learning_rate": 1.6540602657401526e-05, "loss": 0.0147, "step": 74670 }, { "grad_norm": 0.3372188210487366, "learning_rate": 1.6528317665515537e-05, "loss": 0.0131, "step": 74680 }, { "grad_norm": 0.34790393710136414, "learning_rate": 1.6516036334042416e-05, "loss": 0.0152, "step": 74690 }, { "grad_norm": 0.1242232471704483, "learning_rate": 1.650375866432519e-05, "loss": 0.0122, "step": 74700 }, { "grad_norm": 0.18869610130786896, "learning_rate": 1.649148465770659e-05, "loss": 0.013, "step": 74710 }, { "grad_norm": 0.1872168332338333, "learning_rate": 1.647921431552883e-05, "loss": 0.0153, "step": 74720 }, { "grad_norm": 0.17958861589431763, "learning_rate": 1.646694763913381e-05, "loss": 0.0144, "step": 74730 }, { "grad_norm": 0.14768922328948975, "learning_rate": 1.6454684629862966e-05, "loss": 0.0111, "step": 74740 }, { "grad_norm": 0.15984301269054413, "learning_rate": 1.6442425289057388e-05, "loss": 0.0134, "step": 74750 }, { "grad_norm": 0.2083013653755188, "learning_rate": 1.6430169618057718e-05, "loss": 0.0128, "step": 74760 }, { "grad_norm": 0.1551380455493927, "learning_rate": 1.641791761820422e-05, "loss": 0.0135, "step": 74770 }, { "grad_norm": 0.1638954132795334, "learning_rate": 1.640566929083676e-05, "loss": 0.013, "step": 74780 }, { "grad_norm": 0.17494946718215942, "learning_rate": 1.6393424637294823e-05, "loss": 0.0124, "step": 74790 }, { "grad_norm": 0.13883143663406372, "learning_rate": 1.638118365891742e-05, "loss": 0.0135, "step": 74800 }, { "grad_norm": 0.1630328893661499, "learning_rate": 1.636894635704323e-05, "loss": 0.0135, "step": 74810 }, { "grad_norm": 0.17207390069961548, "learning_rate": 1.6356712733010517e-05, "loss": 0.0146, "step": 74820 }, { "grad_norm": 0.09617836773395538, "learning_rate": 1.6344482788157096e-05, "loss": 0.0111, "step": 74830 }, { "grad_norm": 0.11441405862569809, "learning_rate": 1.6332256523820444e-05, "loss": 0.0133, "step": 74840 }, { "grad_norm": 0.15728192031383514, "learning_rate": 1.6320033941337597e-05, "loss": 0.0145, "step": 74850 }, { "grad_norm": 0.25985702872276306, "learning_rate": 1.6307815042045216e-05, "loss": 0.0131, "step": 74860 }, { "grad_norm": 0.15763850510120392, "learning_rate": 1.6295599827279513e-05, "loss": 0.0148, "step": 74870 }, { "grad_norm": 0.18615195155143738, "learning_rate": 1.6283388298376356e-05, "loss": 0.0129, "step": 74880 }, { "grad_norm": 0.1802818328142166, "learning_rate": 1.6271180456671125e-05, "loss": 0.0139, "step": 74890 }, { "grad_norm": 0.17366740107536316, "learning_rate": 1.625897630349893e-05, "loss": 0.0128, "step": 74900 }, { "grad_norm": 0.18604370951652527, "learning_rate": 1.6246775840194335e-05, "loss": 0.014, "step": 74910 }, { "grad_norm": 0.17387104034423828, "learning_rate": 1.6234579068091604e-05, "loss": 0.0137, "step": 74920 }, { "grad_norm": 0.22506770491600037, "learning_rate": 1.6222385988524524e-05, "loss": 0.0122, "step": 74930 }, { "grad_norm": 0.16237308084964752, "learning_rate": 1.6210196602826544e-05, "loss": 0.0172, "step": 74940 }, { "grad_norm": 0.26156383752822876, "learning_rate": 1.6198010912330646e-05, "loss": 0.0137, "step": 74950 }, { "grad_norm": 0.2290673851966858, "learning_rate": 1.6185828918369456e-05, "loss": 0.0179, "step": 74960 }, { "grad_norm": 0.13487082719802856, "learning_rate": 1.617365062227517e-05, "loss": 0.0134, "step": 74970 }, { "grad_norm": 0.12405338883399963, "learning_rate": 1.616147602537962e-05, "loss": 0.0137, "step": 74980 }, { "grad_norm": 0.1452598124742508, "learning_rate": 1.614930512901415e-05, "loss": 0.0123, "step": 74990 }, { "grad_norm": 0.1426110416650772, "learning_rate": 1.6137137934509804e-05, "loss": 0.0157, "step": 75000 }, { "grad_norm": 0.25866878032684326, "learning_rate": 1.6124974443197123e-05, "loss": 0.0134, "step": 75010 }, { "grad_norm": 0.23839668929576874, "learning_rate": 1.6112814656406306e-05, "loss": 0.0145, "step": 75020 }, { "grad_norm": 0.12560419738292694, "learning_rate": 1.6100658575467125e-05, "loss": 0.0195, "step": 75030 }, { "grad_norm": 0.1441882848739624, "learning_rate": 1.6088506201708974e-05, "loss": 0.0132, "step": 75040 }, { "grad_norm": 0.13947109878063202, "learning_rate": 1.6076357536460785e-05, "loss": 0.0128, "step": 75050 }, { "grad_norm": 0.2934829890727997, "learning_rate": 1.606421258105113e-05, "loss": 0.0161, "step": 75060 }, { "grad_norm": 0.24705679714679718, "learning_rate": 1.6052071336808183e-05, "loss": 0.0142, "step": 75070 }, { "grad_norm": 0.16265319287776947, "learning_rate": 1.6039933805059658e-05, "loss": 0.0116, "step": 75080 }, { "grad_norm": 0.1312064230442047, "learning_rate": 1.6027799987132915e-05, "loss": 0.0142, "step": 75090 }, { "grad_norm": 0.1532869189977646, "learning_rate": 1.6015669884354888e-05, "loss": 0.0105, "step": 75100 }, { "grad_norm": 0.17198337614536285, "learning_rate": 1.6003543498052125e-05, "loss": 0.0123, "step": 75110 }, { "grad_norm": 0.1770513653755188, "learning_rate": 1.5991420829550715e-05, "loss": 0.0148, "step": 75120 }, { "grad_norm": 0.16189050674438477, "learning_rate": 1.5979301880176405e-05, "loss": 0.0167, "step": 75130 }, { "grad_norm": 0.2426275610923767, "learning_rate": 1.5967186651254485e-05, "loss": 0.0177, "step": 75140 }, { "grad_norm": 0.2611396014690399, "learning_rate": 1.5955075144109853e-05, "loss": 0.014, "step": 75150 }, { "grad_norm": 0.4284418821334839, "learning_rate": 1.594296736006703e-05, "loss": 0.0137, "step": 75160 }, { "grad_norm": 0.20697277784347534, "learning_rate": 1.59308633004501e-05, "loss": 0.0129, "step": 75170 }, { "grad_norm": 0.20557118952274323, "learning_rate": 1.5918762966582724e-05, "loss": 0.0126, "step": 75180 }, { "grad_norm": 0.09158487617969513, "learning_rate": 1.5906666359788203e-05, "loss": 0.0119, "step": 75190 }, { "grad_norm": 0.16011455655097961, "learning_rate": 1.589457348138938e-05, "loss": 0.0127, "step": 75200 }, { "grad_norm": 0.14108990132808685, "learning_rate": 1.588248433270872e-05, "loss": 0.0135, "step": 75210 }, { "grad_norm": 0.19015692174434662, "learning_rate": 1.587039891506828e-05, "loss": 0.0158, "step": 75220 }, { "grad_norm": 0.13388994336128235, "learning_rate": 1.5858317229789715e-05, "loss": 0.0139, "step": 75230 }, { "grad_norm": 0.16718894243240356, "learning_rate": 1.584623927819423e-05, "loss": 0.0161, "step": 75240 }, { "grad_norm": 0.24934126436710358, "learning_rate": 1.583416506160269e-05, "loss": 0.0142, "step": 75250 }, { "grad_norm": 0.19251702725887299, "learning_rate": 1.5822094581335472e-05, "loss": 0.013, "step": 75260 }, { "grad_norm": 0.10204874724149704, "learning_rate": 1.5810027838712604e-05, "loss": 0.0127, "step": 75270 }, { "grad_norm": 0.1535707265138626, "learning_rate": 1.5797964835053687e-05, "loss": 0.0153, "step": 75280 }, { "grad_norm": 0.30172091722488403, "learning_rate": 1.5785905571677933e-05, "loss": 0.013, "step": 75290 }, { "grad_norm": 0.29503151774406433, "learning_rate": 1.5773850049904086e-05, "loss": 0.0147, "step": 75300 }, { "grad_norm": 0.21304485201835632, "learning_rate": 1.5761798271050543e-05, "loss": 0.0131, "step": 75310 }, { "grad_norm": 0.19529017806053162, "learning_rate": 1.5749750236435277e-05, "loss": 0.0152, "step": 75320 }, { "grad_norm": 0.21022126078605652, "learning_rate": 1.5737705947375815e-05, "loss": 0.0124, "step": 75330 }, { "grad_norm": 0.12175637483596802, "learning_rate": 1.5725665405189322e-05, "loss": 0.0137, "step": 75340 }, { "grad_norm": 0.7236336469650269, "learning_rate": 1.571362861119253e-05, "loss": 0.0156, "step": 75350 }, { "grad_norm": 0.2063860297203064, "learning_rate": 1.570159556670177e-05, "loss": 0.0164, "step": 75360 }, { "grad_norm": 0.15318183600902557, "learning_rate": 1.5689566273032946e-05, "loss": 0.0163, "step": 75370 }, { "grad_norm": 0.2833911180496216, "learning_rate": 1.5677540731501583e-05, "loss": 0.0152, "step": 75380 }, { "grad_norm": 0.14088355004787445, "learning_rate": 1.5665518943422746e-05, "loss": 0.0155, "step": 75390 }, { "grad_norm": 0.260865718126297, "learning_rate": 1.5653500910111134e-05, "loss": 0.0113, "step": 75400 }, { "grad_norm": 0.18610386550426483, "learning_rate": 1.5641486632881023e-05, "loss": 0.0129, "step": 75410 }, { "grad_norm": 0.1727224886417389, "learning_rate": 1.56294761130463e-05, "loss": 0.0131, "step": 75420 }, { "grad_norm": 0.1293954998254776, "learning_rate": 1.5617469351920373e-05, "loss": 0.0132, "step": 75430 }, { "grad_norm": 0.1325998306274414, "learning_rate": 1.5605466350816315e-05, "loss": 0.0138, "step": 75440 }, { "grad_norm": 0.18959659337997437, "learning_rate": 1.559346711104674e-05, "loss": 0.0147, "step": 75450 }, { "grad_norm": 0.12851864099502563, "learning_rate": 1.5581471633923883e-05, "loss": 0.0109, "step": 75460 }, { "grad_norm": 0.11345122754573822, "learning_rate": 1.5569479920759505e-05, "loss": 0.0123, "step": 75470 }, { "grad_norm": 0.19917050004005432, "learning_rate": 1.5557491972865072e-05, "loss": 0.0135, "step": 75480 }, { "grad_norm": 0.19906014204025269, "learning_rate": 1.554550779155152e-05, "loss": 0.015, "step": 75490 }, { "grad_norm": 0.20665287971496582, "learning_rate": 1.5533527378129453e-05, "loss": 0.0107, "step": 75500 }, { "grad_norm": 0.21456685662269592, "learning_rate": 1.552155073390899e-05, "loss": 0.0162, "step": 75510 }, { "grad_norm": 0.17663374543190002, "learning_rate": 1.5509577860199903e-05, "loss": 0.0109, "step": 75520 }, { "grad_norm": 0.1374804526567459, "learning_rate": 1.5497608758311525e-05, "loss": 0.0139, "step": 75530 }, { "grad_norm": 0.16335709393024445, "learning_rate": 1.5485643429552797e-05, "loss": 0.0113, "step": 75540 }, { "grad_norm": 0.20267429947853088, "learning_rate": 1.5473681875232197e-05, "loss": 0.013, "step": 75550 }, { "grad_norm": 0.15463899075984955, "learning_rate": 1.5461724096657837e-05, "loss": 0.0163, "step": 75560 }, { "grad_norm": 0.1801292896270752, "learning_rate": 1.5449770095137416e-05, "loss": 0.0134, "step": 75570 }, { "grad_norm": 0.11823833733797073, "learning_rate": 1.5437819871978178e-05, "loss": 0.0113, "step": 75580 }, { "grad_norm": 0.14248234033584595, "learning_rate": 1.5425873428486992e-05, "loss": 0.0131, "step": 75590 }, { "grad_norm": 0.16140981018543243, "learning_rate": 1.5413930765970312e-05, "loss": 0.0129, "step": 75600 }, { "grad_norm": 0.14457648992538452, "learning_rate": 1.5401991885734175e-05, "loss": 0.0137, "step": 75610 }, { "grad_norm": 0.1815074235200882, "learning_rate": 1.5390056789084174e-05, "loss": 0.0146, "step": 75620 }, { "grad_norm": 0.2646065056324005, "learning_rate": 1.5378125477325545e-05, "loss": 0.0133, "step": 75630 }, { "grad_norm": 0.14053647220134735, "learning_rate": 1.5366197951763046e-05, "loss": 0.0135, "step": 75640 }, { "grad_norm": 0.14364780485630035, "learning_rate": 1.5354274213701075e-05, "loss": 0.0115, "step": 75650 }, { "grad_norm": 0.3594938814640045, "learning_rate": 1.5342354264443566e-05, "loss": 0.0135, "step": 75660 }, { "grad_norm": 0.19324907660484314, "learning_rate": 1.533043810529411e-05, "loss": 0.0126, "step": 75670 }, { "grad_norm": 0.16221757233142853, "learning_rate": 1.5318525737555807e-05, "loss": 0.0147, "step": 75680 }, { "grad_norm": 0.17183040082454681, "learning_rate": 1.53066171625314e-05, "loss": 0.0159, "step": 75690 }, { "grad_norm": 0.12260116636753082, "learning_rate": 1.529471238152317e-05, "loss": 0.015, "step": 75700 }, { "grad_norm": 0.21623073518276215, "learning_rate": 1.5282811395833023e-05, "loss": 0.0146, "step": 75710 }, { "grad_norm": 0.19724363088607788, "learning_rate": 1.5270914206762397e-05, "loss": 0.0116, "step": 75720 }, { "grad_norm": 0.9716084003448486, "learning_rate": 1.525902081561241e-05, "loss": 0.0128, "step": 75730 }, { "grad_norm": 0.170660138130188, "learning_rate": 1.5247131223683664e-05, "loss": 0.0136, "step": 75740 }, { "grad_norm": 0.17882482707500458, "learning_rate": 1.5235245432276408e-05, "loss": 0.0143, "step": 75750 }, { "grad_norm": 0.1422208845615387, "learning_rate": 1.5223363442690424e-05, "loss": 0.014, "step": 75760 }, { "grad_norm": 0.18156805634498596, "learning_rate": 1.5211485256225133e-05, "loss": 0.0109, "step": 75770 }, { "grad_norm": 0.34382638335227966, "learning_rate": 1.519961087417951e-05, "loss": 0.0129, "step": 75780 }, { "grad_norm": 0.17300160229206085, "learning_rate": 1.5187740297852137e-05, "loss": 0.0115, "step": 75790 }, { "grad_norm": 0.131649449467659, "learning_rate": 1.517587352854113e-05, "loss": 0.0125, "step": 75800 }, { "grad_norm": 0.16598813235759735, "learning_rate": 1.516401056754423e-05, "loss": 0.0119, "step": 75810 }, { "grad_norm": 0.1907566338777542, "learning_rate": 1.5152151416158783e-05, "loss": 0.0127, "step": 75820 }, { "grad_norm": 0.39683613181114197, "learning_rate": 1.5140296075681642e-05, "loss": 0.0135, "step": 75830 }, { "grad_norm": 0.14837510883808136, "learning_rate": 1.5128444547409331e-05, "loss": 0.0125, "step": 75840 }, { "grad_norm": 0.17927229404449463, "learning_rate": 1.5116596832637864e-05, "loss": 0.0122, "step": 75850 }, { "grad_norm": 0.20110110938549042, "learning_rate": 1.5104752932662952e-05, "loss": 0.0125, "step": 75860 }, { "grad_norm": 0.16030697524547577, "learning_rate": 1.5092912848779778e-05, "loss": 0.0111, "step": 75870 }, { "grad_norm": 0.3986435532569885, "learning_rate": 1.5081076582283194e-05, "loss": 0.013, "step": 75880 }, { "grad_norm": 0.16127018630504608, "learning_rate": 1.5069244134467553e-05, "loss": 0.0123, "step": 75890 }, { "grad_norm": 0.12861329317092896, "learning_rate": 1.5057415506626876e-05, "loss": 0.0128, "step": 75900 }, { "grad_norm": 0.1575874239206314, "learning_rate": 1.5045590700054668e-05, "loss": 0.0132, "step": 75910 }, { "grad_norm": 0.3055267632007599, "learning_rate": 1.5033769716044144e-05, "loss": 0.0154, "step": 75920 }, { "grad_norm": 0.19169719517230988, "learning_rate": 1.5021952555887975e-05, "loss": 0.0119, "step": 75930 }, { "grad_norm": 0.33231544494628906, "learning_rate": 1.50101392208785e-05, "loss": 0.0139, "step": 75940 }, { "grad_norm": 0.16349101066589355, "learning_rate": 1.4998329712307579e-05, "loss": 0.0143, "step": 75950 }, { "grad_norm": 0.2031787633895874, "learning_rate": 1.4986524031466708e-05, "loss": 0.0145, "step": 75960 }, { "grad_norm": 0.20437604188919067, "learning_rate": 1.4974722179646916e-05, "loss": 0.0107, "step": 75970 }, { "grad_norm": 0.17118261754512787, "learning_rate": 1.4962924158138847e-05, "loss": 0.0114, "step": 75980 }, { "grad_norm": 0.2429327368736267, "learning_rate": 1.4951129968232714e-05, "loss": 0.0132, "step": 75990 }, { "grad_norm": 0.5867718458175659, "learning_rate": 1.4939339611218329e-05, "loss": 0.0166, "step": 76000 }, { "grad_norm": 0.18900004029273987, "learning_rate": 1.4927553088385032e-05, "loss": 0.0137, "step": 76010 }, { "grad_norm": 0.18248921632766724, "learning_rate": 1.4915770401021801e-05, "loss": 0.0123, "step": 76020 }, { "grad_norm": 0.13952380418777466, "learning_rate": 1.4903991550417185e-05, "loss": 0.0116, "step": 76030 }, { "grad_norm": 0.2730009853839874, "learning_rate": 1.4892216537859277e-05, "loss": 0.0136, "step": 76040 }, { "grad_norm": 0.13425776362419128, "learning_rate": 1.4880445364635775e-05, "loss": 0.0116, "step": 76050 }, { "grad_norm": 0.14853833615779877, "learning_rate": 1.4868678032033973e-05, "loss": 0.0116, "step": 76060 }, { "grad_norm": 0.13707007467746735, "learning_rate": 1.4856914541340739e-05, "loss": 0.0129, "step": 76070 }, { "grad_norm": 0.13716955482959747, "learning_rate": 1.4845154893842472e-05, "loss": 0.0111, "step": 76080 }, { "grad_norm": 0.157281294465065, "learning_rate": 1.4833399090825234e-05, "loss": 0.0102, "step": 76090 }, { "grad_norm": 0.20662258565425873, "learning_rate": 1.4821647133574562e-05, "loss": 0.0141, "step": 76100 }, { "grad_norm": 0.3417339324951172, "learning_rate": 1.4809899023375706e-05, "loss": 0.0134, "step": 76110 }, { "grad_norm": 0.2642117440700531, "learning_rate": 1.4798154761513366e-05, "loss": 0.0117, "step": 76120 }, { "grad_norm": 0.20238488912582397, "learning_rate": 1.4786414349271916e-05, "loss": 0.0103, "step": 76130 }, { "grad_norm": 0.18057110905647278, "learning_rate": 1.4774677787935226e-05, "loss": 0.0132, "step": 76140 }, { "grad_norm": 0.18762615323066711, "learning_rate": 1.4762945078786838e-05, "loss": 0.014, "step": 76150 }, { "grad_norm": 0.1312098652124405, "learning_rate": 1.4751216223109777e-05, "loss": 0.0122, "step": 76160 }, { "grad_norm": 0.17324939370155334, "learning_rate": 1.4739491222186714e-05, "loss": 0.0149, "step": 76170 }, { "grad_norm": 0.20092779397964478, "learning_rate": 1.4727770077299879e-05, "loss": 0.0132, "step": 76180 }, { "grad_norm": 0.34593114256858826, "learning_rate": 1.4716052789731094e-05, "loss": 0.0167, "step": 76190 }, { "grad_norm": 0.19997325539588928, "learning_rate": 1.470433936076171e-05, "loss": 0.0142, "step": 76200 }, { "grad_norm": 0.16320334374904633, "learning_rate": 1.4692629791672719e-05, "loss": 0.0163, "step": 76210 }, { "grad_norm": 0.14876788854599, "learning_rate": 1.468092408374464e-05, "loss": 0.0098, "step": 76220 }, { "grad_norm": 0.14013464748859406, "learning_rate": 1.4669222238257602e-05, "loss": 0.0131, "step": 76230 }, { "grad_norm": 0.18939177691936493, "learning_rate": 1.4657524256491301e-05, "loss": 0.0147, "step": 76240 }, { "grad_norm": 0.16922356188297272, "learning_rate": 1.464583013972502e-05, "loss": 0.0124, "step": 76250 }, { "grad_norm": 0.18187421560287476, "learning_rate": 1.4634139889237591e-05, "loss": 0.0163, "step": 76260 }, { "grad_norm": 0.11143036186695099, "learning_rate": 1.4622453506307448e-05, "loss": 0.011, "step": 76270 }, { "grad_norm": 0.10227710008621216, "learning_rate": 1.461077099221262e-05, "loss": 0.0114, "step": 76280 }, { "grad_norm": 0.12061383575201035, "learning_rate": 1.4599092348230652e-05, "loss": 0.0114, "step": 76290 }, { "grad_norm": 0.21064777672290802, "learning_rate": 1.4587417575638724e-05, "loss": 0.0131, "step": 76300 }, { "grad_norm": 0.17824922502040863, "learning_rate": 1.4575746675713563e-05, "loss": 0.0147, "step": 76310 }, { "grad_norm": 0.09973306208848953, "learning_rate": 1.4564079649731504e-05, "loss": 0.0133, "step": 76320 }, { "grad_norm": 0.25600293278694153, "learning_rate": 1.4552416498968402e-05, "loss": 0.0149, "step": 76330 }, { "grad_norm": 0.12301488220691681, "learning_rate": 1.4540757224699747e-05, "loss": 0.0153, "step": 76340 }, { "grad_norm": 0.12293323874473572, "learning_rate": 1.452910182820056e-05, "loss": 0.012, "step": 76350 }, { "grad_norm": 0.17493395507335663, "learning_rate": 1.4517450310745468e-05, "loss": 0.0148, "step": 76360 }, { "grad_norm": 0.35346007347106934, "learning_rate": 1.450580267360866e-05, "loss": 0.0152, "step": 76370 }, { "grad_norm": 0.160195991396904, "learning_rate": 1.4494158918063922e-05, "loss": 0.0122, "step": 76380 }, { "grad_norm": 0.15846188366413116, "learning_rate": 1.4482519045384568e-05, "loss": 0.0107, "step": 76390 }, { "grad_norm": 0.19519630074501038, "learning_rate": 1.4470883056843548e-05, "loss": 0.0132, "step": 76400 }, { "grad_norm": 0.1855456382036209, "learning_rate": 1.4459250953713322e-05, "loss": 0.0142, "step": 76410 }, { "grad_norm": 0.1816648542881012, "learning_rate": 1.444762273726598e-05, "loss": 0.0123, "step": 76420 }, { "grad_norm": 0.14668947458267212, "learning_rate": 1.443599840877316e-05, "loss": 0.0107, "step": 76430 }, { "grad_norm": 0.08877810835838318, "learning_rate": 1.4424377969506103e-05, "loss": 0.0131, "step": 76440 }, { "grad_norm": 0.24111522734165192, "learning_rate": 1.4412761420735565e-05, "loss": 0.017, "step": 76450 }, { "grad_norm": 0.14640577137470245, "learning_rate": 1.4401148763731953e-05, "loss": 0.0134, "step": 76460 }, { "grad_norm": 0.1195497214794159, "learning_rate": 1.4389539999765173e-05, "loss": 0.0137, "step": 76470 }, { "grad_norm": 0.1497800052165985, "learning_rate": 1.4377935130104758e-05, "loss": 0.0104, "step": 76480 }, { "grad_norm": 0.2294985055923462, "learning_rate": 1.4366334156019801e-05, "loss": 0.0132, "step": 76490 }, { "grad_norm": 0.16132818162441254, "learning_rate": 1.435473707877898e-05, "loss": 0.0121, "step": 76500 }, { "grad_norm": 0.10254614055156708, "learning_rate": 1.4343143899650508e-05, "loss": 0.0122, "step": 76510 }, { "grad_norm": 0.2452293187379837, "learning_rate": 1.4331554619902204e-05, "loss": 0.0131, "step": 76520 }, { "grad_norm": 0.2015308141708374, "learning_rate": 1.4319969240801484e-05, "loss": 0.0139, "step": 76530 }, { "grad_norm": 0.5569625496864319, "learning_rate": 1.430838776361526e-05, "loss": 0.013, "step": 76540 }, { "grad_norm": 0.2419363260269165, "learning_rate": 1.4296810189610093e-05, "loss": 0.0129, "step": 76550 }, { "grad_norm": 0.17006005346775055, "learning_rate": 1.4285236520052087e-05, "loss": 0.0165, "step": 76560 }, { "grad_norm": 0.14821898937225342, "learning_rate": 1.4273666756206933e-05, "loss": 0.0112, "step": 76570 }, { "grad_norm": 0.17638860642910004, "learning_rate": 1.4262100899339848e-05, "loss": 0.0126, "step": 76580 }, { "grad_norm": 0.1968962401151657, "learning_rate": 1.4250538950715696e-05, "loss": 0.0114, "step": 76590 }, { "grad_norm": 0.31182482838630676, "learning_rate": 1.4238980911598842e-05, "loss": 0.0141, "step": 76600 }, { "grad_norm": 0.20660080015659332, "learning_rate": 1.4227426783253267e-05, "loss": 0.0137, "step": 76610 }, { "grad_norm": 0.20174409449100494, "learning_rate": 1.4215876566942521e-05, "loss": 0.0142, "step": 76620 }, { "grad_norm": 0.1705634444952011, "learning_rate": 1.4204330263929721e-05, "loss": 0.0133, "step": 76630 }, { "grad_norm": 0.20695851743221283, "learning_rate": 1.4192787875477537e-05, "loss": 0.0108, "step": 76640 }, { "grad_norm": 0.14966654777526855, "learning_rate": 1.4181249402848246e-05, "loss": 0.0116, "step": 76650 }, { "grad_norm": 0.16378150880336761, "learning_rate": 1.4169714847303656e-05, "loss": 0.0153, "step": 76660 }, { "grad_norm": 0.27447107434272766, "learning_rate": 1.4158184210105197e-05, "loss": 0.0181, "step": 76670 }, { "grad_norm": 0.16702590882778168, "learning_rate": 1.4146657492513787e-05, "loss": 0.0123, "step": 76680 }, { "grad_norm": 0.1330534666776657, "learning_rate": 1.4135134695790047e-05, "loss": 0.0136, "step": 76690 }, { "grad_norm": 0.12377552688121796, "learning_rate": 1.4123615821194037e-05, "loss": 0.0117, "step": 76700 }, { "grad_norm": 0.18314926326274872, "learning_rate": 1.4112100869985473e-05, "loss": 0.0137, "step": 76710 }, { "grad_norm": 0.23506976664066315, "learning_rate": 1.4100589843423589e-05, "loss": 0.0125, "step": 76720 }, { "grad_norm": 0.09765171259641647, "learning_rate": 1.408908274276723e-05, "loss": 0.0147, "step": 76730 }, { "grad_norm": 0.20737196505069733, "learning_rate": 1.4077579569274786e-05, "loss": 0.0168, "step": 76740 }, { "grad_norm": 0.25106534361839294, "learning_rate": 1.4066080324204245e-05, "loss": 0.0144, "step": 76750 }, { "grad_norm": 0.2032269686460495, "learning_rate": 1.405458500881312e-05, "loss": 0.0133, "step": 76760 }, { "grad_norm": 0.19861625134944916, "learning_rate": 1.4043093624358538e-05, "loss": 0.0133, "step": 76770 }, { "grad_norm": 0.15515659749507904, "learning_rate": 1.4031606172097189e-05, "loss": 0.0121, "step": 76780 }, { "grad_norm": 0.2048235386610031, "learning_rate": 1.40201226532853e-05, "loss": 0.0138, "step": 76790 }, { "grad_norm": 0.2268959879875183, "learning_rate": 1.4008643069178701e-05, "loss": 0.0151, "step": 76800 }, { "grad_norm": 0.18298913538455963, "learning_rate": 1.399716742103278e-05, "loss": 0.0119, "step": 76810 }, { "grad_norm": 0.22329296171665192, "learning_rate": 1.398569571010252e-05, "loss": 0.0149, "step": 76820 }, { "grad_norm": 0.16273841261863708, "learning_rate": 1.3974227937642414e-05, "loss": 0.0141, "step": 76830 }, { "grad_norm": 0.1478523313999176, "learning_rate": 1.3962764104906595e-05, "loss": 0.0113, "step": 76840 }, { "grad_norm": 0.16088871657848358, "learning_rate": 1.3951304213148696e-05, "loss": 0.0154, "step": 76850 }, { "grad_norm": 0.1573324203491211, "learning_rate": 1.3939848263621979e-05, "loss": 0.0118, "step": 76860 }, { "grad_norm": 0.09754733741283417, "learning_rate": 1.3928396257579219e-05, "loss": 0.0113, "step": 76870 }, { "grad_norm": 0.16286559402942657, "learning_rate": 1.3916948196272833e-05, "loss": 0.0114, "step": 76880 }, { "grad_norm": 0.15735168755054474, "learning_rate": 1.3905504080954734e-05, "loss": 0.0109, "step": 76890 }, { "grad_norm": 0.21162882447242737, "learning_rate": 1.3894063912876454e-05, "loss": 0.0125, "step": 76900 }, { "grad_norm": 0.27904272079467773, "learning_rate": 1.3882627693289046e-05, "loss": 0.0139, "step": 76910 }, { "grad_norm": 0.15126487612724304, "learning_rate": 1.3871195423443183e-05, "loss": 0.0134, "step": 76920 }, { "grad_norm": 0.1983170211315155, "learning_rate": 1.3859767104589044e-05, "loss": 0.0135, "step": 76930 }, { "grad_norm": 0.1810532957315445, "learning_rate": 1.3848342737976467e-05, "loss": 0.0127, "step": 76940 }, { "grad_norm": 0.2448970228433609, "learning_rate": 1.3836922324854757e-05, "loss": 0.0148, "step": 76950 }, { "grad_norm": 0.17878447473049164, "learning_rate": 1.382550586647287e-05, "loss": 0.012, "step": 76960 }, { "grad_norm": 0.12499593943357468, "learning_rate": 1.3814093364079257e-05, "loss": 0.0155, "step": 76970 }, { "grad_norm": 0.17283140122890472, "learning_rate": 1.3802684818921991e-05, "loss": 0.0145, "step": 76980 }, { "grad_norm": 0.18111403286457062, "learning_rate": 1.3791280232248687e-05, "loss": 0.0159, "step": 76990 }, { "grad_norm": 0.1513400375843048, "learning_rate": 1.3779879605306555e-05, "loss": 0.0123, "step": 77000 }, { "grad_norm": 0.18046008050441742, "learning_rate": 1.376848293934232e-05, "loss": 0.0143, "step": 77010 }, { "grad_norm": 0.1865512728691101, "learning_rate": 1.3757090235602316e-05, "loss": 0.0106, "step": 77020 }, { "grad_norm": 0.19271010160446167, "learning_rate": 1.3745701495332447e-05, "loss": 0.0151, "step": 77030 }, { "grad_norm": 0.2516183853149414, "learning_rate": 1.3734316719778145e-05, "loss": 0.0145, "step": 77040 }, { "grad_norm": 0.24073262512683868, "learning_rate": 1.3722935910184454e-05, "loss": 0.0126, "step": 77050 }, { "grad_norm": 0.20821605622768402, "learning_rate": 1.3711559067795921e-05, "loss": 0.0136, "step": 77060 }, { "grad_norm": 0.3015255928039551, "learning_rate": 1.3700186193856756e-05, "loss": 0.0168, "step": 77070 }, { "grad_norm": 0.1583835333585739, "learning_rate": 1.368881728961064e-05, "loss": 0.0119, "step": 77080 }, { "grad_norm": 0.18550194799900055, "learning_rate": 1.3677452356300885e-05, "loss": 0.0173, "step": 77090 }, { "grad_norm": 0.28238338232040405, "learning_rate": 1.3666091395170321e-05, "loss": 0.0127, "step": 77100 }, { "grad_norm": 0.20445993542671204, "learning_rate": 1.3654734407461384e-05, "loss": 0.0123, "step": 77110 }, { "grad_norm": 0.28535178303718567, "learning_rate": 1.3643381394416022e-05, "loss": 0.0156, "step": 77120 }, { "grad_norm": 0.17857874929904938, "learning_rate": 1.3632032357275838e-05, "loss": 0.0122, "step": 77130 }, { "grad_norm": 0.1474059373140335, "learning_rate": 1.3620687297281898e-05, "loss": 0.0131, "step": 77140 }, { "grad_norm": 0.1412959098815918, "learning_rate": 1.3609346215674917e-05, "loss": 0.016, "step": 77150 }, { "grad_norm": 0.20719121396541595, "learning_rate": 1.3598009113695109e-05, "loss": 0.0116, "step": 77160 }, { "grad_norm": 0.22986406087875366, "learning_rate": 1.358667599258231e-05, "loss": 0.011, "step": 77170 }, { "grad_norm": 0.16771481931209564, "learning_rate": 1.357534685357586e-05, "loss": 0.014, "step": 77180 }, { "grad_norm": 0.16562725603580475, "learning_rate": 1.3564021697914715e-05, "loss": 0.0113, "step": 77190 }, { "grad_norm": 0.360045462846756, "learning_rate": 1.3552700526837381e-05, "loss": 0.011, "step": 77200 }, { "grad_norm": 0.15259894728660583, "learning_rate": 1.3541383341581925e-05, "loss": 0.0149, "step": 77210 }, { "grad_norm": 0.09594298154115677, "learning_rate": 1.3530070143385965e-05, "loss": 0.0117, "step": 77220 }, { "grad_norm": 0.18889668583869934, "learning_rate": 1.3518760933486702e-05, "loss": 0.0128, "step": 77230 }, { "grad_norm": 0.1151437908411026, "learning_rate": 1.3507455713120908e-05, "loss": 0.0144, "step": 77240 }, { "grad_norm": 0.13992398977279663, "learning_rate": 1.3496154483524876e-05, "loss": 0.0113, "step": 77250 }, { "grad_norm": 0.15001381933689117, "learning_rate": 1.3484857245934506e-05, "loss": 0.0147, "step": 77260 }, { "grad_norm": 0.1427600085735321, "learning_rate": 1.3473564001585248e-05, "loss": 0.0139, "step": 77270 }, { "grad_norm": 0.12520238757133484, "learning_rate": 1.3462274751712129e-05, "loss": 0.0132, "step": 77280 }, { "grad_norm": 0.14238882064819336, "learning_rate": 1.3450989497549698e-05, "loss": 0.0117, "step": 77290 }, { "grad_norm": 0.17307060956954956, "learning_rate": 1.3439708240332115e-05, "loss": 0.0136, "step": 77300 }, { "grad_norm": 0.15618717670440674, "learning_rate": 1.3428430981293043e-05, "loss": 0.0144, "step": 77310 }, { "grad_norm": 0.1544395089149475, "learning_rate": 1.3417157721665808e-05, "loss": 0.0139, "step": 77320 }, { "grad_norm": 0.1409813016653061, "learning_rate": 1.3405888462683186e-05, "loss": 0.0125, "step": 77330 }, { "grad_norm": 0.14216651022434235, "learning_rate": 1.3394623205577595e-05, "loss": 0.0133, "step": 77340 }, { "grad_norm": 0.19840945303440094, "learning_rate": 1.3383361951580963e-05, "loss": 0.0166, "step": 77350 }, { "grad_norm": 0.2112005203962326, "learning_rate": 1.337210470192483e-05, "loss": 0.0122, "step": 77360 }, { "grad_norm": 0.18372157216072083, "learning_rate": 1.3360851457840245e-05, "loss": 0.0125, "step": 77370 }, { "grad_norm": 0.16914290189743042, "learning_rate": 1.3349602220557856e-05, "loss": 0.0135, "step": 77380 }, { "grad_norm": 0.17326591908931732, "learning_rate": 1.3338356991307865e-05, "loss": 0.0124, "step": 77390 }, { "grad_norm": 0.1772417575120926, "learning_rate": 1.3327115771320048e-05, "loss": 0.0118, "step": 77400 }, { "grad_norm": 0.18757674098014832, "learning_rate": 1.3315878561823697e-05, "loss": 0.0131, "step": 77410 }, { "grad_norm": 0.21972224116325378, "learning_rate": 1.330464536404773e-05, "loss": 0.0132, "step": 77420 }, { "grad_norm": 0.25685974955558777, "learning_rate": 1.3293416179220563e-05, "loss": 0.0124, "step": 77430 }, { "grad_norm": 0.20625491440296173, "learning_rate": 1.328219100857021e-05, "loss": 0.014, "step": 77440 }, { "grad_norm": 0.2103879451751709, "learning_rate": 1.327096985332425e-05, "loss": 0.0137, "step": 77450 }, { "grad_norm": 0.21006441116333008, "learning_rate": 1.3259752714709816e-05, "loss": 0.0135, "step": 77460 }, { "grad_norm": 0.12808457016944885, "learning_rate": 1.3248539593953569e-05, "loss": 0.0142, "step": 77470 }, { "grad_norm": 0.2852586805820465, "learning_rate": 1.323733049228178e-05, "loss": 0.0153, "step": 77480 }, { "grad_norm": 0.2101057469844818, "learning_rate": 1.3226125410920276e-05, "loss": 0.0125, "step": 77490 }, { "grad_norm": 0.10143065452575684, "learning_rate": 1.3214924351094388e-05, "loss": 0.0163, "step": 77500 }, { "grad_norm": 0.19620195031166077, "learning_rate": 1.3203727314029069e-05, "loss": 0.0121, "step": 77510 }, { "grad_norm": 0.20967763662338257, "learning_rate": 1.3192534300948811e-05, "loss": 0.0111, "step": 77520 }, { "grad_norm": 0.1524069756269455, "learning_rate": 1.3181345313077675e-05, "loss": 0.0112, "step": 77530 }, { "grad_norm": 0.25610998272895813, "learning_rate": 1.3170160351639243e-05, "loss": 0.0158, "step": 77540 }, { "grad_norm": 0.21392594277858734, "learning_rate": 1.3158979417856721e-05, "loss": 0.0114, "step": 77550 }, { "grad_norm": 0.13510434329509735, "learning_rate": 1.31478025129528e-05, "loss": 0.0108, "step": 77560 }, { "grad_norm": 0.17180344462394714, "learning_rate": 1.3136629638149794e-05, "loss": 0.0111, "step": 77570 }, { "grad_norm": 0.10600028932094574, "learning_rate": 1.3125460794669547e-05, "loss": 0.0127, "step": 77580 }, { "grad_norm": 0.18484680354595184, "learning_rate": 1.3114295983733483e-05, "loss": 0.0137, "step": 77590 }, { "grad_norm": 0.2589041590690613, "learning_rate": 1.3103135206562534e-05, "loss": 0.0118, "step": 77600 }, { "grad_norm": 0.7429209351539612, "learning_rate": 1.3091978464377264e-05, "loss": 0.0135, "step": 77610 }, { "grad_norm": 0.18175125122070312, "learning_rate": 1.3080825758397724e-05, "loss": 0.0167, "step": 77620 }, { "grad_norm": 0.10241422057151794, "learning_rate": 1.3069677089843568e-05, "loss": 0.0122, "step": 77630 }, { "grad_norm": 0.07767318934202194, "learning_rate": 1.305853245993401e-05, "loss": 0.0125, "step": 77640 }, { "grad_norm": 0.1698799580335617, "learning_rate": 1.3047391869887809e-05, "loss": 0.0168, "step": 77650 }, { "grad_norm": 0.18198296427726746, "learning_rate": 1.3036255320923263e-05, "loss": 0.014, "step": 77660 }, { "grad_norm": 0.37164852023124695, "learning_rate": 1.3025122814258273e-05, "loss": 0.0141, "step": 77670 }, { "grad_norm": 0.12480700761079788, "learning_rate": 1.3013994351110249e-05, "loss": 0.0113, "step": 77680 }, { "grad_norm": 0.17351622879505157, "learning_rate": 1.3002869932696194e-05, "loss": 0.015, "step": 77690 }, { "grad_norm": 0.22606663405895233, "learning_rate": 1.2991749560232664e-05, "loss": 0.0124, "step": 77700 }, { "grad_norm": 0.1355653703212738, "learning_rate": 1.2980633234935768e-05, "loss": 0.0136, "step": 77710 }, { "grad_norm": 0.18093343079090118, "learning_rate": 1.2969520958021153e-05, "loss": 0.0163, "step": 77720 }, { "grad_norm": 0.24828684329986572, "learning_rate": 1.295841273070405e-05, "loss": 0.0128, "step": 77730 }, { "grad_norm": 0.19856402277946472, "learning_rate": 1.2947308554199255e-05, "loss": 0.0129, "step": 77740 }, { "grad_norm": 0.1991201639175415, "learning_rate": 1.2936208429721075e-05, "loss": 0.0115, "step": 77750 }, { "grad_norm": 0.22498786449432373, "learning_rate": 1.2925112358483416e-05, "loss": 0.0138, "step": 77760 }, { "grad_norm": 0.12023389339447021, "learning_rate": 1.2914020341699728e-05, "loss": 0.013, "step": 77770 }, { "grad_norm": 0.1437968909740448, "learning_rate": 1.2902932380583032e-05, "loss": 0.0117, "step": 77780 }, { "grad_norm": 0.12381412088871002, "learning_rate": 1.2891848476345863e-05, "loss": 0.016, "step": 77790 }, { "grad_norm": 0.16794408857822418, "learning_rate": 1.2880768630200368e-05, "loss": 0.0121, "step": 77800 }, { "grad_norm": 0.20326465368270874, "learning_rate": 1.286969284335819e-05, "loss": 0.015, "step": 77810 }, { "grad_norm": 0.17246469855308533, "learning_rate": 1.2858621117030578e-05, "loss": 0.0116, "step": 77820 }, { "grad_norm": 0.26566770672798157, "learning_rate": 1.2847553452428324e-05, "loss": 0.0126, "step": 77830 }, { "grad_norm": 0.6079802513122559, "learning_rate": 1.283648985076178e-05, "loss": 0.0147, "step": 77840 }, { "grad_norm": 0.2883492410182953, "learning_rate": 1.2825430313240817e-05, "loss": 0.015, "step": 77850 }, { "grad_norm": 0.12627312541007996, "learning_rate": 1.281437484107491e-05, "loss": 0.0125, "step": 77860 }, { "grad_norm": 0.14773769676685333, "learning_rate": 1.2803323435473058e-05, "loss": 0.0111, "step": 77870 }, { "grad_norm": 0.1893405318260193, "learning_rate": 1.2792276097643841e-05, "loss": 0.012, "step": 77880 }, { "grad_norm": 0.14399583637714386, "learning_rate": 1.2781232828795342e-05, "loss": 0.0119, "step": 77890 }, { "grad_norm": 0.20154501497745514, "learning_rate": 1.2770193630135292e-05, "loss": 0.016, "step": 77900 }, { "grad_norm": 0.1506931483745575, "learning_rate": 1.2759158502870877e-05, "loss": 0.0129, "step": 77910 }, { "grad_norm": 0.1606423258781433, "learning_rate": 1.2748127448208913e-05, "loss": 0.0118, "step": 77920 }, { "grad_norm": 0.15068306028842926, "learning_rate": 1.2737100467355706e-05, "loss": 0.0112, "step": 77930 }, { "grad_norm": 0.345824271440506, "learning_rate": 1.2726077561517169e-05, "loss": 0.0153, "step": 77940 }, { "grad_norm": 0.15195079147815704, "learning_rate": 1.2715058731898755e-05, "loss": 0.013, "step": 77950 }, { "grad_norm": 0.15328183770179749, "learning_rate": 1.2704043979705471e-05, "loss": 0.0104, "step": 77960 }, { "grad_norm": 0.1893967092037201, "learning_rate": 1.269303330614185e-05, "loss": 0.0157, "step": 77970 }, { "grad_norm": 0.2472158819437027, "learning_rate": 1.2682026712412015e-05, "loss": 0.0114, "step": 77980 }, { "grad_norm": 0.09159621596336365, "learning_rate": 1.2671024199719644e-05, "loss": 0.0123, "step": 77990 }, { "grad_norm": 0.18036781251430511, "learning_rate": 1.2660025769267936e-05, "loss": 0.011, "step": 78000 }, { "grad_norm": 0.15917202830314636, "learning_rate": 1.2649031422259667e-05, "loss": 0.0144, "step": 78010 }, { "grad_norm": 0.24959640204906464, "learning_rate": 1.2638041159897163e-05, "loss": 0.0141, "step": 78020 }, { "grad_norm": 0.13850782811641693, "learning_rate": 1.262705498338232e-05, "loss": 0.0138, "step": 78030 }, { "grad_norm": 0.14734935760498047, "learning_rate": 1.2616072893916542e-05, "loss": 0.0133, "step": 78040 }, { "grad_norm": 0.14393256604671478, "learning_rate": 1.2605094892700836e-05, "loss": 0.0132, "step": 78050 }, { "grad_norm": 0.17608846724033356, "learning_rate": 1.259412098093572e-05, "loss": 0.0126, "step": 78060 }, { "grad_norm": 0.1821984350681305, "learning_rate": 1.258315115982131e-05, "loss": 0.0129, "step": 78070 }, { "grad_norm": 0.28622201085090637, "learning_rate": 1.2572185430557203e-05, "loss": 0.0146, "step": 78080 }, { "grad_norm": 0.143262580037117, "learning_rate": 1.2561223794342659e-05, "loss": 0.0137, "step": 78090 }, { "grad_norm": 0.15721409022808075, "learning_rate": 1.255026625237637e-05, "loss": 0.0105, "step": 78100 }, { "grad_norm": 0.146615669131279, "learning_rate": 1.2539312805856678e-05, "loss": 0.0162, "step": 78110 }, { "grad_norm": 0.22760321199893951, "learning_rate": 1.25283634559814e-05, "loss": 0.0116, "step": 78120 }, { "grad_norm": 0.27778440713882446, "learning_rate": 1.2517418203947972e-05, "loss": 0.0121, "step": 78130 }, { "grad_norm": 0.21930833160877228, "learning_rate": 1.2506477050953308e-05, "loss": 0.0155, "step": 78140 }, { "grad_norm": 0.1593749076128006, "learning_rate": 1.2495539998193972e-05, "loss": 0.0127, "step": 78150 }, { "grad_norm": 0.2992572784423828, "learning_rate": 1.2484607046865975e-05, "loss": 0.011, "step": 78160 }, { "grad_norm": 0.19872140884399414, "learning_rate": 1.2473678198164968e-05, "loss": 0.0126, "step": 78170 }, { "grad_norm": 0.17688694596290588, "learning_rate": 1.2462753453286075e-05, "loss": 0.0128, "step": 78180 }, { "grad_norm": 0.15239036083221436, "learning_rate": 1.2451832813424031e-05, "loss": 0.0157, "step": 78190 }, { "grad_norm": 0.13001437485218048, "learning_rate": 1.2440916279773095e-05, "loss": 0.0116, "step": 78200 }, { "grad_norm": 0.16824625432491302, "learning_rate": 1.2430003853527101e-05, "loss": 0.0126, "step": 78210 }, { "grad_norm": 0.1831914335489273, "learning_rate": 1.2419095535879383e-05, "loss": 0.0112, "step": 78220 }, { "grad_norm": 0.33050885796546936, "learning_rate": 1.2408191328022878e-05, "loss": 0.0138, "step": 78230 }, { "grad_norm": 0.17126286029815674, "learning_rate": 1.2397291231150066e-05, "loss": 0.0129, "step": 78240 }, { "grad_norm": 0.20034988224506378, "learning_rate": 1.2386395246452937e-05, "loss": 0.0168, "step": 78250 }, { "grad_norm": 0.24721913039684296, "learning_rate": 1.2375503375123083e-05, "loss": 0.0117, "step": 78260 }, { "grad_norm": 0.2386011779308319, "learning_rate": 1.2364615618351583e-05, "loss": 0.0156, "step": 78270 }, { "grad_norm": 0.23079347610473633, "learning_rate": 1.2353731977329169e-05, "loss": 0.0117, "step": 78280 }, { "grad_norm": 0.34856846928596497, "learning_rate": 1.2342852453246007e-05, "loss": 0.014, "step": 78290 }, { "grad_norm": 0.15005260705947876, "learning_rate": 1.2331977047291898e-05, "loss": 0.0109, "step": 78300 }, { "grad_norm": 0.15069368481636047, "learning_rate": 1.2321105760656137e-05, "loss": 0.0145, "step": 78310 }, { "grad_norm": 0.1787184476852417, "learning_rate": 1.2310238594527613e-05, "loss": 0.0136, "step": 78320 }, { "grad_norm": 0.09876874089241028, "learning_rate": 1.2299375550094693e-05, "loss": 0.0108, "step": 78330 }, { "grad_norm": 0.11228509247303009, "learning_rate": 1.2288516628545421e-05, "loss": 0.0125, "step": 78340 }, { "grad_norm": 0.11931423097848892, "learning_rate": 1.2277661831067255e-05, "loss": 0.0113, "step": 78350 }, { "grad_norm": 0.24486815929412842, "learning_rate": 1.2266811158847286e-05, "loss": 0.0104, "step": 78360 }, { "grad_norm": 0.13096679747104645, "learning_rate": 1.2255964613072107e-05, "loss": 0.0162, "step": 78370 }, { "grad_norm": 0.19570143520832062, "learning_rate": 1.2245122194927905e-05, "loss": 0.0114, "step": 78380 }, { "grad_norm": 0.1324922889471054, "learning_rate": 1.2234283905600364e-05, "loss": 0.0121, "step": 78390 }, { "grad_norm": 0.20257867872714996, "learning_rate": 1.222344974627475e-05, "loss": 0.013, "step": 78400 }, { "grad_norm": 0.2061624526977539, "learning_rate": 1.2212619718135875e-05, "loss": 0.0121, "step": 78410 }, { "grad_norm": 0.15875357389450073, "learning_rate": 1.2201793822368118e-05, "loss": 0.0133, "step": 78420 }, { "grad_norm": 0.2964286506175995, "learning_rate": 1.2190972060155337e-05, "loss": 0.0136, "step": 78430 }, { "grad_norm": 0.1688939779996872, "learning_rate": 1.2180154432681013e-05, "loss": 0.0127, "step": 78440 }, { "grad_norm": 0.17071960866451263, "learning_rate": 1.2169340941128148e-05, "loss": 0.013, "step": 78450 }, { "grad_norm": 0.1754390150308609, "learning_rate": 1.2158531586679267e-05, "loss": 0.0116, "step": 78460 }, { "grad_norm": 0.1288251131772995, "learning_rate": 1.2147726370516477e-05, "loss": 0.0157, "step": 78470 }, { "grad_norm": 0.2703435719013214, "learning_rate": 1.2136925293821422e-05, "loss": 0.0182, "step": 78480 }, { "grad_norm": 0.2071598917245865, "learning_rate": 1.2126128357775306e-05, "loss": 0.0132, "step": 78490 }, { "grad_norm": 0.1281784325838089, "learning_rate": 1.2115335563558838e-05, "loss": 0.0126, "step": 78500 }, { "grad_norm": 0.35509705543518066, "learning_rate": 1.2104546912352327e-05, "loss": 0.011, "step": 78510 }, { "grad_norm": 0.14413000643253326, "learning_rate": 1.2093762405335557e-05, "loss": 0.012, "step": 78520 }, { "grad_norm": 0.16277919709682465, "learning_rate": 1.2082982043687974e-05, "loss": 0.0114, "step": 78530 }, { "grad_norm": 0.2635844051837921, "learning_rate": 1.207220582858845e-05, "loss": 0.0111, "step": 78540 }, { "grad_norm": 0.18567602336406708, "learning_rate": 1.206143376121549e-05, "loss": 0.0116, "step": 78550 }, { "grad_norm": 0.2923468351364136, "learning_rate": 1.2050665842747078e-05, "loss": 0.0165, "step": 78560 }, { "grad_norm": 0.139320969581604, "learning_rate": 1.2039902074360804e-05, "loss": 0.012, "step": 78570 }, { "grad_norm": 0.20334792137145996, "learning_rate": 1.2029142457233755e-05, "loss": 0.0167, "step": 78580 }, { "grad_norm": 0.1858392059803009, "learning_rate": 1.2018386992542601e-05, "loss": 0.011, "step": 78590 }, { "grad_norm": 0.1800810545682907, "learning_rate": 1.2007635681463542e-05, "loss": 0.012, "step": 78600 }, { "grad_norm": 0.2043738067150116, "learning_rate": 1.1996888525172334e-05, "loss": 0.0114, "step": 78610 }, { "grad_norm": 0.16912910342216492, "learning_rate": 1.198614552484425e-05, "loss": 0.0119, "step": 78620 }, { "grad_norm": 0.11055796593427658, "learning_rate": 1.197540668165415e-05, "loss": 0.0127, "step": 78630 }, { "grad_norm": 0.1752724051475525, "learning_rate": 1.1964671996776395e-05, "loss": 0.0127, "step": 78640 }, { "grad_norm": 0.1791905015707016, "learning_rate": 1.1953941471384922e-05, "loss": 0.0143, "step": 78650 }, { "grad_norm": 0.11785060167312622, "learning_rate": 1.1943215106653205e-05, "loss": 0.0109, "step": 78660 }, { "grad_norm": 0.1800774484872818, "learning_rate": 1.193249290375429e-05, "loss": 0.013, "step": 78670 }, { "grad_norm": 0.11828453093767166, "learning_rate": 1.1921774863860696e-05, "loss": 0.011, "step": 78680 }, { "grad_norm": 0.33862727880477905, "learning_rate": 1.1911060988144556e-05, "loss": 0.0128, "step": 78690 }, { "grad_norm": 0.20806679129600525, "learning_rate": 1.1900351277777533e-05, "loss": 0.0119, "step": 78700 }, { "grad_norm": 0.22991295158863068, "learning_rate": 1.1889645733930798e-05, "loss": 0.012, "step": 78710 }, { "grad_norm": 0.16794927418231964, "learning_rate": 1.1878944357775107e-05, "loss": 0.0103, "step": 78720 }, { "grad_norm": 0.16378799080848694, "learning_rate": 1.1868247150480749e-05, "loss": 0.0128, "step": 78730 }, { "grad_norm": 0.1692585051059723, "learning_rate": 1.1857554113217567e-05, "loss": 0.0105, "step": 78740 }, { "grad_norm": 0.16520076990127563, "learning_rate": 1.1846865247154903e-05, "loss": 0.0127, "step": 78750 }, { "grad_norm": 0.13215556740760803, "learning_rate": 1.1836180553461711e-05, "loss": 0.0118, "step": 78760 }, { "grad_norm": 0.17016717791557312, "learning_rate": 1.1825500033306419e-05, "loss": 0.0144, "step": 78770 }, { "grad_norm": 0.1562684029340744, "learning_rate": 1.1814823687857052e-05, "loss": 0.0135, "step": 78780 }, { "grad_norm": 0.5202279686927795, "learning_rate": 1.1804151518281153e-05, "loss": 0.0109, "step": 78790 }, { "grad_norm": 0.14435291290283203, "learning_rate": 1.179348352574583e-05, "loss": 0.0131, "step": 78800 }, { "grad_norm": 0.4491676986217499, "learning_rate": 1.1782819711417697e-05, "loss": 0.013, "step": 78810 }, { "grad_norm": 0.23751556873321533, "learning_rate": 1.1772160076462951e-05, "loss": 0.0167, "step": 78820 }, { "grad_norm": 0.13497094810009003, "learning_rate": 1.1761504622047293e-05, "loss": 0.0133, "step": 78830 }, { "grad_norm": 0.6562068462371826, "learning_rate": 1.1750853349335999e-05, "loss": 0.0125, "step": 78840 }, { "grad_norm": 0.156685471534729, "learning_rate": 1.1740206259493874e-05, "loss": 0.0114, "step": 78850 }, { "grad_norm": 0.15149608254432678, "learning_rate": 1.1729563353685286e-05, "loss": 0.0138, "step": 78860 }, { "grad_norm": 0.14489690959453583, "learning_rate": 1.1718924633074091e-05, "loss": 0.0155, "step": 78870 }, { "grad_norm": 0.12231090664863586, "learning_rate": 1.1708290098823765e-05, "loss": 0.0107, "step": 78880 }, { "grad_norm": 0.26533645391464233, "learning_rate": 1.1697659752097245e-05, "loss": 0.0156, "step": 78890 }, { "grad_norm": 0.18284541368484497, "learning_rate": 1.1687033594057068e-05, "loss": 0.0113, "step": 78900 }, { "grad_norm": 0.16401372849941254, "learning_rate": 1.1676411625865297e-05, "loss": 0.0109, "step": 78910 }, { "grad_norm": 0.15973854064941406, "learning_rate": 1.1665793848683542e-05, "loss": 0.0123, "step": 78920 }, { "grad_norm": 0.17707253992557526, "learning_rate": 1.1655180263672927e-05, "loss": 0.0135, "step": 78930 }, { "grad_norm": 0.2221572995185852, "learning_rate": 1.1644570871994142e-05, "loss": 0.0124, "step": 78940 }, { "grad_norm": 0.1747043877840042, "learning_rate": 1.163396567480744e-05, "loss": 0.0118, "step": 78950 }, { "grad_norm": 0.14725928008556366, "learning_rate": 1.1623364673272552e-05, "loss": 0.0128, "step": 78960 }, { "grad_norm": 0.18553867936134338, "learning_rate": 1.1612767868548807e-05, "loss": 0.0117, "step": 78970 }, { "grad_norm": 0.1685522496700287, "learning_rate": 1.1602175261795057e-05, "loss": 0.0149, "step": 78980 }, { "grad_norm": 0.193708136677742, "learning_rate": 1.1591586854169696e-05, "loss": 0.0119, "step": 78990 }, { "grad_norm": 0.1932252049446106, "learning_rate": 1.158100264683064e-05, "loss": 0.0133, "step": 79000 }, { "grad_norm": 0.17204976081848145, "learning_rate": 1.1570422640935386e-05, "loss": 0.0165, "step": 79010 }, { "grad_norm": 0.26586076617240906, "learning_rate": 1.155984683764092e-05, "loss": 0.0145, "step": 79020 }, { "grad_norm": 0.23417076468467712, "learning_rate": 1.1549275238103812e-05, "loss": 0.0138, "step": 79030 }, { "grad_norm": 0.18815597891807556, "learning_rate": 1.1538707843480152e-05, "loss": 0.0129, "step": 79040 }, { "grad_norm": 0.3055475652217865, "learning_rate": 1.1528144654925593e-05, "loss": 0.0111, "step": 79050 }, { "grad_norm": 0.2318369746208191, "learning_rate": 1.1517585673595277e-05, "loss": 0.0137, "step": 79060 }, { "grad_norm": 0.1365744173526764, "learning_rate": 1.150703090064395e-05, "loss": 0.0132, "step": 79070 }, { "grad_norm": 0.18042205274105072, "learning_rate": 1.1496480337225835e-05, "loss": 0.0101, "step": 79080 }, { "grad_norm": 0.10247749090194702, "learning_rate": 1.1485933984494751e-05, "loss": 0.0117, "step": 79090 }, { "grad_norm": 0.26623421907424927, "learning_rate": 1.1475391843604e-05, "loss": 0.0111, "step": 79100 }, { "grad_norm": 0.2296280413866043, "learning_rate": 1.1464853915706503e-05, "loss": 0.0106, "step": 79110 }, { "grad_norm": 0.15252640843391418, "learning_rate": 1.1454320201954626e-05, "loss": 0.0128, "step": 79120 }, { "grad_norm": 0.1835242360830307, "learning_rate": 1.1443790703500356e-05, "loss": 0.012, "step": 79130 }, { "grad_norm": 0.15015441179275513, "learning_rate": 1.1433265421495154e-05, "loss": 0.0105, "step": 79140 }, { "grad_norm": 0.1455603539943695, "learning_rate": 1.1422744357090065e-05, "loss": 0.0115, "step": 79150 }, { "grad_norm": 0.26360490918159485, "learning_rate": 1.1412227511435647e-05, "loss": 0.0139, "step": 79160 }, { "grad_norm": 0.21459342539310455, "learning_rate": 1.1401714885682025e-05, "loss": 0.0117, "step": 79170 }, { "grad_norm": 0.13257646560668945, "learning_rate": 1.1391206480978823e-05, "loss": 0.0137, "step": 79180 }, { "grad_norm": 0.20874163508415222, "learning_rate": 1.1380702298475227e-05, "loss": 0.0138, "step": 79190 }, { "grad_norm": 0.17100568115711212, "learning_rate": 1.1370202339319985e-05, "loss": 0.0088, "step": 79200 }, { "grad_norm": 0.13792265951633453, "learning_rate": 1.1359706604661319e-05, "loss": 0.0129, "step": 79210 }, { "grad_norm": 0.1427517980337143, "learning_rate": 1.1349215095647042e-05, "loss": 0.0105, "step": 79220 }, { "grad_norm": 0.12102405726909637, "learning_rate": 1.1338727813424493e-05, "loss": 0.0109, "step": 79230 }, { "grad_norm": 0.22329320013523102, "learning_rate": 1.132824475914056e-05, "loss": 0.0122, "step": 79240 }, { "grad_norm": 0.18372665345668793, "learning_rate": 1.131776593394162e-05, "loss": 0.0126, "step": 79250 }, { "grad_norm": 0.18615061044692993, "learning_rate": 1.1307291338973657e-05, "loss": 0.0109, "step": 79260 }, { "grad_norm": 0.18838487565517426, "learning_rate": 1.1296820975382121e-05, "loss": 0.0131, "step": 79270 }, { "grad_norm": 0.13098976016044617, "learning_rate": 1.1286354844312074e-05, "loss": 0.0112, "step": 79280 }, { "grad_norm": 0.12163924425840378, "learning_rate": 1.1275892946908023e-05, "loss": 0.0123, "step": 79290 }, { "grad_norm": 0.21697767078876495, "learning_rate": 1.126543528431413e-05, "loss": 0.0119, "step": 79300 }, { "grad_norm": 0.17581042647361755, "learning_rate": 1.125498185767398e-05, "loss": 0.0123, "step": 79310 }, { "grad_norm": 0.13650840520858765, "learning_rate": 1.1244532668130781e-05, "loss": 0.0119, "step": 79320 }, { "grad_norm": 0.18504633009433746, "learning_rate": 1.1234087716827203e-05, "loss": 0.0143, "step": 79330 }, { "grad_norm": 0.16525711119174957, "learning_rate": 1.1223647004905524e-05, "loss": 0.0124, "step": 79340 }, { "grad_norm": 0.23882678151130676, "learning_rate": 1.1213210533507485e-05, "loss": 0.0162, "step": 79350 }, { "grad_norm": 0.1572607308626175, "learning_rate": 1.1202778303774447e-05, "loss": 0.0118, "step": 79360 }, { "grad_norm": 0.30363544821739197, "learning_rate": 1.1192350316847234e-05, "loss": 0.0142, "step": 79370 }, { "grad_norm": 0.14686420559883118, "learning_rate": 1.1181926573866258e-05, "loss": 0.0153, "step": 79380 }, { "grad_norm": 0.16429533064365387, "learning_rate": 1.1171507075971416e-05, "loss": 0.0122, "step": 79390 }, { "grad_norm": 0.14473070204257965, "learning_rate": 1.1161091824302177e-05, "loss": 0.0129, "step": 79400 }, { "grad_norm": 0.17603063583374023, "learning_rate": 1.115068081999755e-05, "loss": 0.0148, "step": 79410 }, { "grad_norm": 0.19041693210601807, "learning_rate": 1.1140274064196071e-05, "loss": 0.0173, "step": 79420 }, { "grad_norm": 0.25045478343963623, "learning_rate": 1.112987155803578e-05, "loss": 0.0127, "step": 79430 }, { "grad_norm": 0.1285318285226822, "learning_rate": 1.1119473302654298e-05, "loss": 0.0108, "step": 79440 }, { "grad_norm": 0.304208904504776, "learning_rate": 1.1109079299188769e-05, "loss": 0.0134, "step": 79450 }, { "grad_norm": 0.17750976979732513, "learning_rate": 1.1098689548775847e-05, "loss": 0.012, "step": 79460 }, { "grad_norm": 0.2767241895198822, "learning_rate": 1.1088304052551762e-05, "loss": 0.0136, "step": 79470 }, { "grad_norm": 0.19929009675979614, "learning_rate": 1.1077922811652215e-05, "loss": 0.0168, "step": 79480 }, { "grad_norm": 0.16829082369804382, "learning_rate": 1.1067545827212534e-05, "loss": 0.0112, "step": 79490 }, { "grad_norm": 0.17927756905555725, "learning_rate": 1.1057173100367496e-05, "loss": 0.0142, "step": 79500 }, { "grad_norm": 0.1834876835346222, "learning_rate": 1.1046804632251474e-05, "loss": 0.0113, "step": 79510 }, { "grad_norm": 0.2524336278438568, "learning_rate": 1.1036440423998318e-05, "loss": 0.0172, "step": 79520 }, { "grad_norm": 0.16589024662971497, "learning_rate": 1.1026080476741469e-05, "loss": 0.0145, "step": 79530 }, { "grad_norm": 0.13482706248760223, "learning_rate": 1.1015724791613836e-05, "loss": 0.0104, "step": 79540 }, { "grad_norm": 0.22101372480392456, "learning_rate": 1.1005373369747951e-05, "loss": 0.0119, "step": 79550 }, { "grad_norm": 0.10512217879295349, "learning_rate": 1.09950262122758e-05, "loss": 0.0135, "step": 79560 }, { "grad_norm": 0.17858687043190002, "learning_rate": 1.0984683320328948e-05, "loss": 0.014, "step": 79570 }, { "grad_norm": 0.15151618421077728, "learning_rate": 1.0974344695038458e-05, "loss": 0.0129, "step": 79580 }, { "grad_norm": 0.16236135363578796, "learning_rate": 1.0964010337534968e-05, "loss": 0.0117, "step": 79590 }, { "grad_norm": 0.1534346640110016, "learning_rate": 1.0953680248948611e-05, "loss": 0.0116, "step": 79600 }, { "grad_norm": 0.2188008725643158, "learning_rate": 1.0943354430409069e-05, "loss": 0.0137, "step": 79610 }, { "grad_norm": 0.22739167511463165, "learning_rate": 1.0933032883045574e-05, "loss": 0.0105, "step": 79620 }, { "grad_norm": 0.22057420015335083, "learning_rate": 1.0922715607986872e-05, "loss": 0.0129, "step": 79630 }, { "grad_norm": 0.13185125589370728, "learning_rate": 1.0912402606361234e-05, "loss": 0.0127, "step": 79640 }, { "grad_norm": 0.16541439294815063, "learning_rate": 1.090209387929647e-05, "loss": 0.0111, "step": 79650 }, { "grad_norm": 0.29182708263397217, "learning_rate": 1.0891789427919957e-05, "loss": 0.0129, "step": 79660 }, { "grad_norm": 0.15213407576084137, "learning_rate": 1.088148925335854e-05, "loss": 0.0128, "step": 79670 }, { "grad_norm": 0.2772608995437622, "learning_rate": 1.087119335673864e-05, "loss": 0.012, "step": 79680 }, { "grad_norm": 0.11971861124038696, "learning_rate": 1.0860901739186208e-05, "loss": 0.0108, "step": 79690 }, { "grad_norm": 0.14243486523628235, "learning_rate": 1.085061440182673e-05, "loss": 0.0163, "step": 79700 }, { "grad_norm": 0.22218719124794006, "learning_rate": 1.0840331345785193e-05, "loss": 0.0141, "step": 79710 }, { "grad_norm": 0.16701960563659668, "learning_rate": 1.083005257218615e-05, "loss": 0.0117, "step": 79720 }, { "grad_norm": 0.26290714740753174, "learning_rate": 1.0819778082153642e-05, "loss": 0.0135, "step": 79730 }, { "grad_norm": 0.18889044225215912, "learning_rate": 1.0809507876811325e-05, "loss": 0.0117, "step": 79740 }, { "grad_norm": 0.14498405158519745, "learning_rate": 1.0799241957282292e-05, "loss": 0.0133, "step": 79750 }, { "grad_norm": 0.1277035027742386, "learning_rate": 1.078898032468923e-05, "loss": 0.0109, "step": 79760 }, { "grad_norm": 0.15802210569381714, "learning_rate": 1.0778722980154315e-05, "loss": 0.0127, "step": 79770 }, { "grad_norm": 0.29106438159942627, "learning_rate": 1.0768469924799302e-05, "loss": 0.0174, "step": 79780 }, { "grad_norm": 0.1456255167722702, "learning_rate": 1.0758221159745419e-05, "loss": 0.0119, "step": 79790 }, { "grad_norm": 0.13751454651355743, "learning_rate": 1.074797668611347e-05, "loss": 0.0134, "step": 79800 }, { "grad_norm": 0.15153850615024567, "learning_rate": 1.0737736505023777e-05, "loss": 0.0127, "step": 79810 }, { "grad_norm": 0.08701730519533157, "learning_rate": 1.0727500617596203e-05, "loss": 0.0122, "step": 79820 }, { "grad_norm": 0.30962812900543213, "learning_rate": 1.0717269024950104e-05, "loss": 0.0129, "step": 79830 }, { "grad_norm": 0.12795040011405945, "learning_rate": 1.070704172820442e-05, "loss": 0.0129, "step": 79840 }, { "grad_norm": 0.09899145364761353, "learning_rate": 1.0696818728477559e-05, "loss": 0.0129, "step": 79850 }, { "grad_norm": 0.2251976579427719, "learning_rate": 1.0686600026887511e-05, "loss": 0.0131, "step": 79860 }, { "grad_norm": 0.11035169661045074, "learning_rate": 1.0676385624551777e-05, "loss": 0.0126, "step": 79870 }, { "grad_norm": 0.16252975165843964, "learning_rate": 1.0666175522587402e-05, "loss": 0.0116, "step": 79880 }, { "grad_norm": 0.17838560044765472, "learning_rate": 1.0655969722110925e-05, "loss": 0.0141, "step": 79890 }, { "grad_norm": 0.23476077616214752, "learning_rate": 1.0645768224238439e-05, "loss": 0.0123, "step": 79900 }, { "grad_norm": 0.3492492735385895, "learning_rate": 1.063557103008559e-05, "loss": 0.0147, "step": 79910 }, { "grad_norm": 0.23260730504989624, "learning_rate": 1.0625378140767494e-05, "loss": 0.0111, "step": 79920 }, { "grad_norm": 0.524021565914154, "learning_rate": 1.0615189557398846e-05, "loss": 0.0131, "step": 79930 }, { "grad_norm": 0.14172083139419556, "learning_rate": 1.060500528109386e-05, "loss": 0.0119, "step": 79940 }, { "grad_norm": 0.12042161077260971, "learning_rate": 1.0594825312966272e-05, "loss": 0.0114, "step": 79950 }, { "grad_norm": 0.13227348029613495, "learning_rate": 1.0584649654129336e-05, "loss": 0.0108, "step": 79960 }, { "grad_norm": 0.13864043354988098, "learning_rate": 1.0574478305695863e-05, "loss": 0.0129, "step": 79970 }, { "grad_norm": 0.2282995879650116, "learning_rate": 1.0564311268778154e-05, "loss": 0.0119, "step": 79980 }, { "grad_norm": 0.17414428293704987, "learning_rate": 1.0554148544488073e-05, "loss": 0.0114, "step": 79990 }, { "grad_norm": 0.19986556470394135, "learning_rate": 1.0543990133937004e-05, "loss": 0.0117, "step": 80000 }, { "grad_norm": 0.21092526614665985, "learning_rate": 1.0533836038235868e-05, "loss": 0.013, "step": 80010 }, { "grad_norm": 0.17438577115535736, "learning_rate": 1.052368625849507e-05, "loss": 0.0129, "step": 80020 }, { "grad_norm": 0.21974240243434906, "learning_rate": 1.05135407958246e-05, "loss": 0.0123, "step": 80030 }, { "grad_norm": 0.1539907604455948, "learning_rate": 1.0503399651333934e-05, "loss": 0.0129, "step": 80040 }, { "grad_norm": 0.2406221628189087, "learning_rate": 1.04932628261321e-05, "loss": 0.014, "step": 80050 }, { "grad_norm": 0.29967200756073, "learning_rate": 1.0483130321327644e-05, "loss": 0.0138, "step": 80060 }, { "grad_norm": 0.19583231210708618, "learning_rate": 1.0473002138028654e-05, "loss": 0.0113, "step": 80070 }, { "grad_norm": 0.15487302839756012, "learning_rate": 1.0462878277342713e-05, "loss": 0.0098, "step": 80080 }, { "grad_norm": 0.22489582002162933, "learning_rate": 1.0452758740376972e-05, "loss": 0.0151, "step": 80090 }, { "grad_norm": 0.12635831534862518, "learning_rate": 1.0442643528238067e-05, "loss": 0.0113, "step": 80100 }, { "grad_norm": 0.1272578239440918, "learning_rate": 1.0432532642032188e-05, "loss": 0.0116, "step": 80110 }, { "grad_norm": 0.13395847380161285, "learning_rate": 1.042242608286505e-05, "loss": 0.0134, "step": 80120 }, { "grad_norm": 0.14443880319595337, "learning_rate": 1.041232385184191e-05, "loss": 0.0109, "step": 80130 }, { "grad_norm": 0.10920335352420807, "learning_rate": 1.0402225950067495e-05, "loss": 0.0128, "step": 80140 }, { "grad_norm": 0.16232144832611084, "learning_rate": 1.0392132378646119e-05, "loss": 0.0137, "step": 80150 }, { "grad_norm": 0.1584547609090805, "learning_rate": 1.0382043138681607e-05, "loss": 0.0119, "step": 80160 }, { "grad_norm": 0.14217019081115723, "learning_rate": 1.0371958231277284e-05, "loss": 0.012, "step": 80170 }, { "grad_norm": 0.1278742402791977, "learning_rate": 1.0361877657536024e-05, "loss": 0.0111, "step": 80180 }, { "grad_norm": 0.18648655712604523, "learning_rate": 1.035180141856023e-05, "loss": 0.0159, "step": 80190 }, { "grad_norm": 0.14956781268119812, "learning_rate": 1.0341729515451836e-05, "loss": 0.0141, "step": 80200 }, { "grad_norm": 0.12426096200942993, "learning_rate": 1.0331661949312265e-05, "loss": 0.013, "step": 80210 }, { "grad_norm": 0.6965517401695251, "learning_rate": 1.0321598721242514e-05, "loss": 0.0107, "step": 80220 }, { "grad_norm": 0.3285578787326813, "learning_rate": 1.0311539832343064e-05, "loss": 0.0111, "step": 80230 }, { "grad_norm": 0.1705550253391266, "learning_rate": 1.0301485283713942e-05, "loss": 0.014, "step": 80240 }, { "grad_norm": 0.16229036450386047, "learning_rate": 1.0291435076454703e-05, "loss": 0.0175, "step": 80250 }, { "grad_norm": 0.12017661333084106, "learning_rate": 1.028138921166444e-05, "loss": 0.0124, "step": 80260 }, { "grad_norm": 0.2012222856283188, "learning_rate": 1.0271347690441719e-05, "loss": 0.016, "step": 80270 }, { "grad_norm": 0.2534787654876709, "learning_rate": 1.0261310513884698e-05, "loss": 0.0127, "step": 80280 }, { "grad_norm": 0.16298985481262207, "learning_rate": 1.0251277683090998e-05, "loss": 0.0154, "step": 80290 }, { "grad_norm": 0.1793881207704544, "learning_rate": 1.0241249199157826e-05, "loss": 0.0116, "step": 80300 }, { "grad_norm": 0.14296111464500427, "learning_rate": 1.0231225063181837e-05, "loss": 0.0101, "step": 80310 }, { "grad_norm": 0.20694150030612946, "learning_rate": 1.0221205276259316e-05, "loss": 0.0102, "step": 80320 }, { "grad_norm": 0.387951523065567, "learning_rate": 1.0211189839485958e-05, "loss": 0.0133, "step": 80330 }, { "grad_norm": 0.24920159578323364, "learning_rate": 1.0201178753957075e-05, "loss": 0.0141, "step": 80340 }, { "grad_norm": 0.15652534365653992, "learning_rate": 1.0191172020767432e-05, "loss": 0.0146, "step": 80350 }, { "grad_norm": 0.410928875207901, "learning_rate": 1.0181169641011362e-05, "loss": 0.0137, "step": 80360 }, { "grad_norm": 0.21056680381298065, "learning_rate": 1.0171171615782721e-05, "loss": 0.0127, "step": 80370 }, { "grad_norm": 0.11018813401460648, "learning_rate": 1.0161177946174866e-05, "loss": 0.0113, "step": 80380 }, { "grad_norm": 0.15035557746887207, "learning_rate": 1.0151188633280706e-05, "loss": 0.011, "step": 80390 }, { "grad_norm": 0.18595540523529053, "learning_rate": 1.0141203678192635e-05, "loss": 0.0162, "step": 80400 }, { "grad_norm": 0.18340669572353363, "learning_rate": 1.0131223082002612e-05, "loss": 0.0136, "step": 80410 }, { "grad_norm": 0.29995009303092957, "learning_rate": 1.012124684580208e-05, "loss": 0.0102, "step": 80420 }, { "grad_norm": 0.1247926652431488, "learning_rate": 1.0111274970682039e-05, "loss": 0.0101, "step": 80430 }, { "grad_norm": 0.10957943648099899, "learning_rate": 1.010130745773299e-05, "loss": 0.0151, "step": 80440 }, { "grad_norm": 0.19852285087108612, "learning_rate": 1.0091344308044986e-05, "loss": 0.0119, "step": 80450 }, { "grad_norm": 0.1509762704372406, "learning_rate": 1.0081385522707554e-05, "loss": 0.0147, "step": 80460 }, { "grad_norm": 0.2047395557165146, "learning_rate": 1.0071431102809803e-05, "loss": 0.0116, "step": 80470 }, { "grad_norm": 0.15446121990680695, "learning_rate": 1.0061481049440297e-05, "loss": 0.0127, "step": 80480 }, { "grad_norm": 0.2396112084388733, "learning_rate": 1.0051535363687187e-05, "loss": 0.0112, "step": 80490 }, { "grad_norm": 0.12466733157634735, "learning_rate": 1.0041594046638087e-05, "loss": 0.009, "step": 80500 }, { "grad_norm": 0.21065545082092285, "learning_rate": 1.0031657099380204e-05, "loss": 0.011, "step": 80510 }, { "grad_norm": 0.23247145116329193, "learning_rate": 1.0021724523000204e-05, "loss": 0.0119, "step": 80520 }, { "grad_norm": 0.18048463761806488, "learning_rate": 1.001179631858431e-05, "loss": 0.01, "step": 80530 }, { "grad_norm": 0.11660120636224747, "learning_rate": 1.0001872487218245e-05, "loss": 0.0121, "step": 80540 }, { "grad_norm": 0.20544622838497162, "learning_rate": 9.991953029987272e-06, "loss": 0.0139, "step": 80550 }, { "grad_norm": 0.23504295945167542, "learning_rate": 9.982037947976147e-06, "loss": 0.0146, "step": 80560 }, { "grad_norm": 0.10827545821666718, "learning_rate": 9.972127242269208e-06, "loss": 0.0101, "step": 80570 }, { "grad_norm": 0.15765300393104553, "learning_rate": 9.962220913950238e-06, "loss": 0.0135, "step": 80580 }, { "grad_norm": 0.19392672181129456, "learning_rate": 9.952318964102591e-06, "loss": 0.0102, "step": 80590 }, { "grad_norm": 0.1750412881374359, "learning_rate": 9.942421393809148e-06, "loss": 0.012, "step": 80600 }, { "grad_norm": 0.14927129447460175, "learning_rate": 9.932528204152264e-06, "loss": 0.0123, "step": 80610 }, { "grad_norm": 0.2010796219110489, "learning_rate": 9.922639396213856e-06, "loss": 0.0125, "step": 80620 }, { "grad_norm": 0.1804199367761612, "learning_rate": 9.912754971075344e-06, "loss": 0.0175, "step": 80630 }, { "grad_norm": 0.1683483123779297, "learning_rate": 9.902874929817696e-06, "loss": 0.0129, "step": 80640 }, { "grad_norm": 0.21029940247535706, "learning_rate": 9.892999273521342e-06, "loss": 0.0133, "step": 80650 }, { "grad_norm": 0.1582086980342865, "learning_rate": 9.883128003266302e-06, "loss": 0.0125, "step": 80660 }, { "grad_norm": 0.14234016835689545, "learning_rate": 9.87326112013206e-06, "loss": 0.011, "step": 80670 }, { "grad_norm": 0.14708299934864044, "learning_rate": 9.863398625197662e-06, "loss": 0.013, "step": 80680 }, { "grad_norm": 0.1353532373905182, "learning_rate": 9.853540519541615e-06, "loss": 0.0129, "step": 80690 }, { "grad_norm": 0.17171551287174225, "learning_rate": 9.843686804242053e-06, "loss": 0.0118, "step": 80700 }, { "grad_norm": 0.4925638735294342, "learning_rate": 9.833837480376506e-06, "loss": 0.0154, "step": 80710 }, { "grad_norm": 0.1818990260362625, "learning_rate": 9.82399254902212e-06, "loss": 0.0159, "step": 80720 }, { "grad_norm": 0.1674192249774933, "learning_rate": 9.814152011255494e-06, "loss": 0.0145, "step": 80730 }, { "grad_norm": 0.2079000174999237, "learning_rate": 9.804315868152797e-06, "loss": 0.0132, "step": 80740 }, { "grad_norm": 0.13472409546375275, "learning_rate": 9.79448412078966e-06, "loss": 0.0124, "step": 80750 }, { "grad_norm": 0.15020862221717834, "learning_rate": 9.784656770241312e-06, "loss": 0.0117, "step": 80760 }, { "grad_norm": 0.15822544693946838, "learning_rate": 9.77483381758243e-06, "loss": 0.0095, "step": 80770 }, { "grad_norm": 0.18302814662456512, "learning_rate": 9.76501526388725e-06, "loss": 0.0116, "step": 80780 }, { "grad_norm": 0.1465338170528412, "learning_rate": 9.7552011102295e-06, "loss": 0.0132, "step": 80790 }, { "grad_norm": 0.25632667541503906, "learning_rate": 9.745391357682449e-06, "loss": 0.0154, "step": 80800 }, { "grad_norm": 0.1387237161397934, "learning_rate": 9.735586007318875e-06, "loss": 0.0121, "step": 80810 }, { "grad_norm": 0.1675298660993576, "learning_rate": 9.725785060211096e-06, "loss": 0.0133, "step": 80820 }, { "grad_norm": 0.22706438601016998, "learning_rate": 9.715988517430896e-06, "loss": 0.0155, "step": 80830 }, { "grad_norm": 0.28156450390815735, "learning_rate": 9.706196380049625e-06, "loss": 0.0119, "step": 80840 }, { "grad_norm": 0.13057665526866913, "learning_rate": 9.69640864913815e-06, "loss": 0.0125, "step": 80850 }, { "grad_norm": 0.19773848354816437, "learning_rate": 9.686625325766818e-06, "loss": 0.0147, "step": 80860 }, { "grad_norm": 0.20010794699192047, "learning_rate": 9.67684641100554e-06, "loss": 0.0109, "step": 80870 }, { "grad_norm": 0.2813275158405304, "learning_rate": 9.667071905923692e-06, "loss": 0.0166, "step": 80880 }, { "grad_norm": 0.17590199410915375, "learning_rate": 9.657301811590242e-06, "loss": 0.0115, "step": 80890 }, { "grad_norm": 0.30286532640457153, "learning_rate": 9.647536129073603e-06, "loss": 0.0116, "step": 80900 }, { "grad_norm": 0.19357329607009888, "learning_rate": 9.637774859441751e-06, "loss": 0.0151, "step": 80910 }, { "grad_norm": 0.15198510885238647, "learning_rate": 9.628018003762152e-06, "loss": 0.0149, "step": 80920 }, { "grad_norm": 0.16447687149047852, "learning_rate": 9.618265563101813e-06, "loss": 0.0114, "step": 80930 }, { "grad_norm": 0.37939560413360596, "learning_rate": 9.608517538527211e-06, "loss": 0.0133, "step": 80940 }, { "grad_norm": 0.15987789630889893, "learning_rate": 9.598773931104431e-06, "loss": 0.0134, "step": 80950 }, { "grad_norm": 0.2234651893377304, "learning_rate": 9.589034741898983e-06, "loss": 0.0144, "step": 80960 }, { "grad_norm": 0.17070788145065308, "learning_rate": 9.579299971975947e-06, "loss": 0.0106, "step": 80970 }, { "grad_norm": 0.12236776947975159, "learning_rate": 9.569569622399882e-06, "loss": 0.0121, "step": 80980 }, { "grad_norm": 0.19891607761383057, "learning_rate": 9.559843694234916e-06, "loss": 0.0129, "step": 80990 }, { "grad_norm": 0.16172103583812714, "learning_rate": 9.550122188544625e-06, "loss": 0.0119, "step": 81000 }, { "grad_norm": 0.15458062291145325, "learning_rate": 9.540405106392152e-06, "loss": 0.0118, "step": 81010 }, { "grad_norm": 0.15799902379512787, "learning_rate": 9.53069244884015e-06, "loss": 0.0124, "step": 81020 }, { "grad_norm": 0.19811904430389404, "learning_rate": 9.52098421695079e-06, "loss": 0.0126, "step": 81030 }, { "grad_norm": 0.10034505277872086, "learning_rate": 9.511280411785728e-06, "loss": 0.0109, "step": 81040 }, { "grad_norm": 0.21448184549808502, "learning_rate": 9.50158103440616e-06, "loss": 0.0129, "step": 81050 }, { "grad_norm": 0.19103847444057465, "learning_rate": 9.491886085872815e-06, "loss": 0.0134, "step": 81060 }, { "grad_norm": 0.20849233865737915, "learning_rate": 9.482195567245895e-06, "loss": 0.018, "step": 81070 }, { "grad_norm": 0.16248561441898346, "learning_rate": 9.472509479585151e-06, "loss": 0.0153, "step": 81080 }, { "grad_norm": 0.17909376323223114, "learning_rate": 9.462827823949833e-06, "loss": 0.0165, "step": 81090 }, { "grad_norm": 0.15513946115970612, "learning_rate": 9.453150601398735e-06, "loss": 0.0118, "step": 81100 }, { "grad_norm": 0.17541705071926117, "learning_rate": 9.44347781299011e-06, "loss": 0.0121, "step": 81110 }, { "grad_norm": 0.15915082395076752, "learning_rate": 9.43380945978179e-06, "loss": 0.0143, "step": 81120 }, { "grad_norm": 0.19711396098136902, "learning_rate": 9.424145542831047e-06, "loss": 0.0124, "step": 81130 }, { "grad_norm": 0.18952327966690063, "learning_rate": 9.414486063194778e-06, "loss": 0.0115, "step": 81140 }, { "grad_norm": 0.17416100203990936, "learning_rate": 9.404831021929273e-06, "loss": 0.0114, "step": 81150 }, { "grad_norm": 0.22015546262264252, "learning_rate": 9.395180420090427e-06, "loss": 0.0107, "step": 81160 }, { "grad_norm": 0.2047221064567566, "learning_rate": 9.385534258733591e-06, "loss": 0.0092, "step": 81170 }, { "grad_norm": 0.1351049393415451, "learning_rate": 9.375892538913677e-06, "loss": 0.0106, "step": 81180 }, { "grad_norm": 0.4693552255630493, "learning_rate": 9.36625526168507e-06, "loss": 0.0132, "step": 81190 }, { "grad_norm": 0.1387934386730194, "learning_rate": 9.356622428101686e-06, "loss": 0.0126, "step": 81200 }, { "grad_norm": 0.26558172702789307, "learning_rate": 9.346994039216972e-06, "loss": 0.0128, "step": 81210 }, { "grad_norm": 0.14645543694496155, "learning_rate": 9.337370096083886e-06, "loss": 0.0125, "step": 81220 }, { "grad_norm": 0.18429264426231384, "learning_rate": 9.327750599754854e-06, "loss": 0.0131, "step": 81230 }, { "grad_norm": 0.1605818271636963, "learning_rate": 9.318135551281875e-06, "loss": 0.0129, "step": 81240 }, { "grad_norm": 0.15057040750980377, "learning_rate": 9.308524951716424e-06, "loss": 0.0106, "step": 81250 }, { "grad_norm": 0.20383915305137634, "learning_rate": 9.298918802109497e-06, "loss": 0.0126, "step": 81260 }, { "grad_norm": 0.18508018553256989, "learning_rate": 9.289317103511625e-06, "loss": 0.0124, "step": 81270 }, { "grad_norm": 0.25971636176109314, "learning_rate": 9.279719856972835e-06, "loss": 0.0119, "step": 81280 }, { "grad_norm": 0.16077974438667297, "learning_rate": 9.270127063542645e-06, "loss": 0.0119, "step": 81290 }, { "grad_norm": 0.10366783291101456, "learning_rate": 9.260538724270124e-06, "loss": 0.0115, "step": 81300 }, { "grad_norm": 0.24489721655845642, "learning_rate": 9.250954840203852e-06, "loss": 0.0135, "step": 81310 }, { "grad_norm": 0.12746137380599976, "learning_rate": 9.241375412391878e-06, "loss": 0.0098, "step": 81320 }, { "grad_norm": 0.30706751346588135, "learning_rate": 9.23180044188181e-06, "loss": 0.0107, "step": 81330 }, { "grad_norm": 0.14672835171222687, "learning_rate": 9.222229929720755e-06, "loss": 0.0102, "step": 81340 }, { "grad_norm": 0.11444977670907974, "learning_rate": 9.212663876955336e-06, "loss": 0.0087, "step": 81350 }, { "grad_norm": 0.1341988444328308, "learning_rate": 9.20310228463166e-06, "loss": 0.012, "step": 81360 }, { "grad_norm": 0.163031667470932, "learning_rate": 9.19354515379539e-06, "loss": 0.0146, "step": 81370 }, { "grad_norm": 0.24131131172180176, "learning_rate": 9.183992485491666e-06, "loss": 0.0119, "step": 81380 }, { "grad_norm": 0.15253208577632904, "learning_rate": 9.174444280765154e-06, "loss": 0.0125, "step": 81390 }, { "grad_norm": 0.1269407421350479, "learning_rate": 9.164900540660032e-06, "loss": 0.0126, "step": 81400 }, { "grad_norm": 0.20693691074848175, "learning_rate": 9.155361266220008e-06, "loss": 0.0117, "step": 81410 }, { "grad_norm": 0.1477879136800766, "learning_rate": 9.145826458488254e-06, "loss": 0.0122, "step": 81420 }, { "grad_norm": 0.11514542996883392, "learning_rate": 9.136296118507509e-06, "loss": 0.0114, "step": 81430 }, { "grad_norm": 0.2855006754398346, "learning_rate": 9.126770247319966e-06, "loss": 0.0135, "step": 81440 }, { "grad_norm": 0.1614503413438797, "learning_rate": 9.11724884596738e-06, "loss": 0.012, "step": 81450 }, { "grad_norm": 0.12827491760253906, "learning_rate": 9.107731915490996e-06, "loss": 0.0105, "step": 81460 }, { "grad_norm": 0.19005481898784637, "learning_rate": 9.098219456931579e-06, "loss": 0.0118, "step": 81470 }, { "grad_norm": 0.21533985435962677, "learning_rate": 9.088711471329376e-06, "loss": 0.011, "step": 81480 }, { "grad_norm": 0.2662101686000824, "learning_rate": 9.079207959724184e-06, "loss": 0.0116, "step": 81490 }, { "grad_norm": 0.15605859458446503, "learning_rate": 9.069708923155279e-06, "loss": 0.0139, "step": 81500 }, { "grad_norm": 0.24035364389419556, "learning_rate": 9.060214362661468e-06, "loss": 0.0146, "step": 81510 }, { "grad_norm": 0.24869269132614136, "learning_rate": 9.050724279281064e-06, "loss": 0.0136, "step": 81520 }, { "grad_norm": 0.625934898853302, "learning_rate": 9.041238674051894e-06, "loss": 0.0119, "step": 81530 }, { "grad_norm": 0.1813727617263794, "learning_rate": 9.031757548011277e-06, "loss": 0.0107, "step": 81540 }, { "grad_norm": 0.21505439281463623, "learning_rate": 9.022280902196051e-06, "loss": 0.0123, "step": 81550 }, { "grad_norm": 0.20844198763370514, "learning_rate": 9.012808737642592e-06, "loss": 0.0106, "step": 81560 }, { "grad_norm": 0.16618962585926056, "learning_rate": 9.003341055386739e-06, "loss": 0.0143, "step": 81570 }, { "grad_norm": 0.13900940120220184, "learning_rate": 8.993877856463862e-06, "loss": 0.0116, "step": 81580 }, { "grad_norm": 0.18450701236724854, "learning_rate": 8.984419141908856e-06, "loss": 0.0114, "step": 81590 }, { "grad_norm": 0.23852746188640594, "learning_rate": 8.974964912756117e-06, "loss": 0.0125, "step": 81600 }, { "grad_norm": 0.2481354922056198, "learning_rate": 8.965515170039523e-06, "loss": 0.0137, "step": 81610 }, { "grad_norm": 0.18745605647563934, "learning_rate": 8.956069914792514e-06, "loss": 0.0127, "step": 81620 }, { "grad_norm": 0.15431277453899384, "learning_rate": 8.946629148047969e-06, "loss": 0.0131, "step": 81630 }, { "grad_norm": 0.12828148901462555, "learning_rate": 8.93719287083834e-06, "loss": 0.0112, "step": 81640 }, { "grad_norm": 0.31586283445358276, "learning_rate": 8.927761084195569e-06, "loss": 0.0116, "step": 81650 }, { "grad_norm": 0.23274792730808258, "learning_rate": 8.918333789151096e-06, "loss": 0.0108, "step": 81660 }, { "grad_norm": 0.2066810578107834, "learning_rate": 8.90891098673587e-06, "loss": 0.0111, "step": 81670 }, { "grad_norm": 0.2924872934818268, "learning_rate": 8.89949267798037e-06, "loss": 0.0113, "step": 81680 }, { "grad_norm": 0.27059826254844666, "learning_rate": 8.890078863914542e-06, "loss": 0.0111, "step": 81690 }, { "grad_norm": 0.2485598921775818, "learning_rate": 8.880669545567894e-06, "loss": 0.0123, "step": 81700 }, { "grad_norm": 0.2525399327278137, "learning_rate": 8.871264723969381e-06, "loss": 0.0111, "step": 81710 }, { "grad_norm": 0.18364182114601135, "learning_rate": 8.86186440014754e-06, "loss": 0.011, "step": 81720 }, { "grad_norm": 0.19363640248775482, "learning_rate": 8.852468575130351e-06, "loss": 0.012, "step": 81730 }, { "grad_norm": 0.1424681693315506, "learning_rate": 8.843077249945347e-06, "loss": 0.0123, "step": 81740 }, { "grad_norm": 0.16816797852516174, "learning_rate": 8.833690425619523e-06, "loss": 0.011, "step": 81750 }, { "grad_norm": 0.15704165399074554, "learning_rate": 8.82430810317942e-06, "loss": 0.0118, "step": 81760 }, { "grad_norm": 0.12726305425167084, "learning_rate": 8.814930283651074e-06, "loss": 0.0136, "step": 81770 }, { "grad_norm": 0.13954009115695953, "learning_rate": 8.805556968060048e-06, "loss": 0.0116, "step": 81780 }, { "grad_norm": 0.11508642882108688, "learning_rate": 8.796188157431362e-06, "loss": 0.0115, "step": 81790 }, { "grad_norm": 0.1906154453754425, "learning_rate": 8.786823852789595e-06, "loss": 0.0128, "step": 81800 }, { "grad_norm": 0.16618052124977112, "learning_rate": 8.777464055158813e-06, "loss": 0.0131, "step": 81810 }, { "grad_norm": 0.20730248093605042, "learning_rate": 8.768108765562578e-06, "loss": 0.0113, "step": 81820 }, { "grad_norm": 0.22857992351055145, "learning_rate": 8.758757985023975e-06, "loss": 0.0157, "step": 81830 }, { "grad_norm": 0.153434157371521, "learning_rate": 8.749411714565598e-06, "loss": 0.0112, "step": 81840 }, { "grad_norm": 0.13676050305366516, "learning_rate": 8.74006995520954e-06, "loss": 0.0107, "step": 81850 }, { "grad_norm": 0.19772332906723022, "learning_rate": 8.730732707977397e-06, "loss": 0.0104, "step": 81860 }, { "grad_norm": 0.19447211921215057, "learning_rate": 8.721399973890281e-06, "loss": 0.0134, "step": 81870 }, { "grad_norm": 0.12007617205381393, "learning_rate": 8.712071753968792e-06, "loss": 0.0118, "step": 81880 }, { "grad_norm": 0.17786529660224915, "learning_rate": 8.702748049233072e-06, "loss": 0.0151, "step": 81890 }, { "grad_norm": 0.13106635212898254, "learning_rate": 8.693428860702718e-06, "loss": 0.0101, "step": 81900 }, { "grad_norm": 0.1632792204618454, "learning_rate": 8.684114189396898e-06, "loss": 0.0127, "step": 81910 }, { "grad_norm": 0.13529229164123535, "learning_rate": 8.674804036334222e-06, "loss": 0.0105, "step": 81920 }, { "grad_norm": 0.21919886767864227, "learning_rate": 8.665498402532857e-06, "loss": 0.0126, "step": 81930 }, { "grad_norm": 0.21710380911827087, "learning_rate": 8.656197289010432e-06, "loss": 0.0149, "step": 81940 }, { "grad_norm": 0.15056544542312622, "learning_rate": 8.646900696784122e-06, "loss": 0.0127, "step": 81950 }, { "grad_norm": 0.3352862000465393, "learning_rate": 8.637608626870553e-06, "loss": 0.0122, "step": 81960 }, { "grad_norm": 0.2097792774438858, "learning_rate": 8.628321080285944e-06, "loss": 0.0126, "step": 81970 }, { "grad_norm": 0.17481178045272827, "learning_rate": 8.619038058045925e-06, "loss": 0.0143, "step": 81980 }, { "grad_norm": 0.5717402100563049, "learning_rate": 8.609759561165704e-06, "loss": 0.0127, "step": 81990 }, { "grad_norm": 0.1865551769733429, "learning_rate": 8.600485590659934e-06, "loss": 0.0143, "step": 82000 }, { "grad_norm": 0.181321918964386, "learning_rate": 8.591216147542813e-06, "loss": 0.0121, "step": 82010 }, { "grad_norm": 0.13046737015247345, "learning_rate": 8.581951232828045e-06, "loss": 0.013, "step": 82020 }, { "grad_norm": 0.4177570044994354, "learning_rate": 8.572690847528825e-06, "loss": 0.012, "step": 82030 }, { "grad_norm": 0.21520724892616272, "learning_rate": 8.563434992657838e-06, "loss": 0.0116, "step": 82040 }, { "grad_norm": 0.15437056124210358, "learning_rate": 8.554183669227307e-06, "loss": 0.0098, "step": 82050 }, { "grad_norm": 0.1948491632938385, "learning_rate": 8.544936878248937e-06, "loss": 0.0125, "step": 82060 }, { "grad_norm": 0.1722467690706253, "learning_rate": 8.535694620733942e-06, "loss": 0.012, "step": 82070 }, { "grad_norm": 0.1771511733531952, "learning_rate": 8.526456897693053e-06, "loss": 0.0142, "step": 82080 }, { "grad_norm": 0.14430858194828033, "learning_rate": 8.517223710136456e-06, "loss": 0.0141, "step": 82090 }, { "grad_norm": 0.12963680922985077, "learning_rate": 8.507995059073931e-06, "loss": 0.0114, "step": 82100 }, { "grad_norm": 0.8321232199668884, "learning_rate": 8.498770945514673e-06, "loss": 0.0123, "step": 82110 }, { "grad_norm": 0.1305752694606781, "learning_rate": 8.489551370467436e-06, "loss": 0.0104, "step": 82120 }, { "grad_norm": 0.16923664510250092, "learning_rate": 8.480336334940442e-06, "loss": 0.0155, "step": 82130 }, { "grad_norm": 0.21351194381713867, "learning_rate": 8.471125839941458e-06, "loss": 0.0129, "step": 82140 }, { "grad_norm": 0.17901764810085297, "learning_rate": 8.461919886477682e-06, "loss": 0.0117, "step": 82150 }, { "grad_norm": 0.13156497478485107, "learning_rate": 8.452718475555927e-06, "loss": 0.0128, "step": 82160 }, { "grad_norm": 0.18147285282611847, "learning_rate": 8.443521608182398e-06, "loss": 0.0113, "step": 82170 }, { "grad_norm": 0.25196921825408936, "learning_rate": 8.43432928536288e-06, "loss": 0.0124, "step": 82180 }, { "grad_norm": 0.16208943724632263, "learning_rate": 8.425141508102607e-06, "loss": 0.0125, "step": 82190 }, { "grad_norm": 0.16592510044574738, "learning_rate": 8.415958277406365e-06, "loss": 0.0124, "step": 82200 }, { "grad_norm": 0.1761370748281479, "learning_rate": 8.406779594278392e-06, "loss": 0.0118, "step": 82210 }, { "grad_norm": 0.23791363835334778, "learning_rate": 8.397605459722468e-06, "loss": 0.0125, "step": 82220 }, { "grad_norm": 0.13727447390556335, "learning_rate": 8.388435874741863e-06, "loss": 0.0103, "step": 82230 }, { "grad_norm": 0.11335580796003342, "learning_rate": 8.37927084033936e-06, "loss": 0.012, "step": 82240 }, { "grad_norm": 0.17001335322856903, "learning_rate": 8.370110357517208e-06, "loss": 0.0103, "step": 82250 }, { "grad_norm": 0.2158564031124115, "learning_rate": 8.360954427277201e-06, "loss": 0.0116, "step": 82260 }, { "grad_norm": 0.18600288033485413, "learning_rate": 8.351803050620627e-06, "loss": 0.0141, "step": 82270 }, { "grad_norm": 0.09940598160028458, "learning_rate": 8.34265622854824e-06, "loss": 0.011, "step": 82280 }, { "grad_norm": 0.19794157147407532, "learning_rate": 8.333513962060335e-06, "loss": 0.014, "step": 82290 }, { "grad_norm": 0.12828902900218964, "learning_rate": 8.324376252156701e-06, "loss": 0.0096, "step": 82300 }, { "grad_norm": 0.11586009711027145, "learning_rate": 8.315243099836633e-06, "loss": 0.0113, "step": 82310 }, { "grad_norm": 0.2037130743265152, "learning_rate": 8.3061145060989e-06, "loss": 0.0094, "step": 82320 }, { "grad_norm": 0.22877298295497894, "learning_rate": 8.296990471941806e-06, "loss": 0.0121, "step": 82330 }, { "grad_norm": 0.16924799978733063, "learning_rate": 8.287870998363112e-06, "loss": 0.0106, "step": 82340 }, { "grad_norm": 0.20791560411453247, "learning_rate": 8.278756086360156e-06, "loss": 0.0101, "step": 82350 }, { "grad_norm": 0.252695232629776, "learning_rate": 8.269645736929693e-06, "loss": 0.0135, "step": 82360 }, { "grad_norm": 0.1933669149875641, "learning_rate": 8.260539951068048e-06, "loss": 0.0155, "step": 82370 }, { "grad_norm": 0.2269088476896286, "learning_rate": 8.251438729770983e-06, "loss": 0.0117, "step": 82380 }, { "grad_norm": 0.13757334649562836, "learning_rate": 8.242342074033826e-06, "loss": 0.0145, "step": 82390 }, { "grad_norm": 0.7299441695213318, "learning_rate": 8.233249984851344e-06, "loss": 0.0155, "step": 82400 }, { "grad_norm": 0.15714837610721588, "learning_rate": 8.224162463217849e-06, "loss": 0.011, "step": 82410 }, { "grad_norm": 0.3163238763809204, "learning_rate": 8.215079510127139e-06, "loss": 0.0131, "step": 82420 }, { "grad_norm": 0.24139408767223358, "learning_rate": 8.206001126572521e-06, "loss": 0.0096, "step": 82430 }, { "grad_norm": 0.15597043931484222, "learning_rate": 8.196927313546776e-06, "loss": 0.0099, "step": 82440 }, { "grad_norm": 0.4324798285961151, "learning_rate": 8.18785807204222e-06, "loss": 0.0122, "step": 82450 }, { "grad_norm": 0.1283116340637207, "learning_rate": 8.178793403050627e-06, "loss": 0.0096, "step": 82460 }, { "grad_norm": 0.330215722322464, "learning_rate": 8.16973330756331e-06, "loss": 0.012, "step": 82470 }, { "grad_norm": 0.24649454653263092, "learning_rate": 8.16067778657107e-06, "loss": 0.0125, "step": 82480 }, { "grad_norm": 0.18837027251720428, "learning_rate": 8.151626841064214e-06, "loss": 0.015, "step": 82490 }, { "grad_norm": 0.152325838804245, "learning_rate": 8.142580472032518e-06, "loss": 0.0109, "step": 82500 }, { "grad_norm": 0.17681926488876343, "learning_rate": 8.133538680465285e-06, "loss": 0.0124, "step": 82510 }, { "grad_norm": 0.14975963532924652, "learning_rate": 8.124501467351336e-06, "loss": 0.0125, "step": 82520 }, { "grad_norm": 0.17413115501403809, "learning_rate": 8.11546883367893e-06, "loss": 0.0113, "step": 82530 }, { "grad_norm": 0.1250261664390564, "learning_rate": 8.106440780435882e-06, "loss": 0.014, "step": 82540 }, { "grad_norm": 0.1628294438123703, "learning_rate": 8.097417308609485e-06, "loss": 0.0146, "step": 82550 }, { "grad_norm": 0.20623143017292023, "learning_rate": 8.088398419186544e-06, "loss": 0.0104, "step": 82560 }, { "grad_norm": 0.15167571604251862, "learning_rate": 8.079384113153321e-06, "loss": 0.0105, "step": 82570 }, { "grad_norm": 0.14544229209423065, "learning_rate": 8.070374391495644e-06, "loss": 0.0102, "step": 82580 }, { "grad_norm": 0.18761491775512695, "learning_rate": 8.061369255198764e-06, "loss": 0.0138, "step": 82590 }, { "grad_norm": 0.1474306434392929, "learning_rate": 8.052368705247493e-06, "loss": 0.0115, "step": 82600 }, { "grad_norm": 0.15378829836845398, "learning_rate": 8.043372742626114e-06, "loss": 0.0107, "step": 82610 }, { "grad_norm": 0.24526523053646088, "learning_rate": 8.034381368318422e-06, "loss": 0.0124, "step": 82620 }, { "grad_norm": 0.11867818981409073, "learning_rate": 8.025394583307677e-06, "loss": 0.0129, "step": 82630 }, { "grad_norm": 0.3503515124320984, "learning_rate": 8.016412388576683e-06, "loss": 0.0147, "step": 82640 }, { "grad_norm": 0.18669570982456207, "learning_rate": 8.007434785107699e-06, "loss": 0.0109, "step": 82650 }, { "grad_norm": 0.11002405732870102, "learning_rate": 7.998461773882509e-06, "loss": 0.0121, "step": 82660 }, { "grad_norm": 0.13414326310157776, "learning_rate": 7.989493355882393e-06, "loss": 0.0127, "step": 82670 }, { "grad_norm": 0.14872848987579346, "learning_rate": 7.980529532088137e-06, "loss": 0.0127, "step": 82680 }, { "grad_norm": 0.2152722328901291, "learning_rate": 7.97157030347998e-06, "loss": 0.0138, "step": 82690 }, { "grad_norm": 0.18729107081890106, "learning_rate": 7.96261567103772e-06, "loss": 0.0133, "step": 82700 }, { "grad_norm": 0.10760790854692459, "learning_rate": 7.953665635740593e-06, "loss": 0.0126, "step": 82710 }, { "grad_norm": 0.1966475397348404, "learning_rate": 7.94472019856738e-06, "loss": 0.0111, "step": 82720 }, { "grad_norm": 0.12132617086172104, "learning_rate": 7.935779360496337e-06, "loss": 0.0122, "step": 82730 }, { "grad_norm": 0.11114512383937836, "learning_rate": 7.92684312250523e-06, "loss": 0.0111, "step": 82740 }, { "grad_norm": 0.1419810801744461, "learning_rate": 7.917911485571289e-06, "loss": 0.0123, "step": 82750 }, { "grad_norm": 0.27345314621925354, "learning_rate": 7.90898445067128e-06, "loss": 0.0106, "step": 82760 }, { "grad_norm": 0.18498308956623077, "learning_rate": 7.900062018781462e-06, "loss": 0.0123, "step": 82770 }, { "grad_norm": 0.26423174142837524, "learning_rate": 7.891144190877553e-06, "loss": 0.0135, "step": 82780 }, { "grad_norm": 0.16744771599769592, "learning_rate": 7.882230967934806e-06, "loss": 0.0121, "step": 82790 }, { "grad_norm": 0.1835915595293045, "learning_rate": 7.873322350927959e-06, "loss": 0.0102, "step": 82800 }, { "grad_norm": 0.24720174074172974, "learning_rate": 7.864418340831253e-06, "loss": 0.0138, "step": 82810 }, { "grad_norm": 0.1688021719455719, "learning_rate": 7.855518938618395e-06, "loss": 0.0103, "step": 82820 }, { "grad_norm": 0.2909137010574341, "learning_rate": 7.846624145262633e-06, "loss": 0.0138, "step": 82830 }, { "grad_norm": 0.1469404101371765, "learning_rate": 7.837733961736672e-06, "loss": 0.0105, "step": 82840 }, { "grad_norm": 0.11897008866071701, "learning_rate": 7.828848389012727e-06, "loss": 0.0108, "step": 82850 }, { "grad_norm": 0.19604092836380005, "learning_rate": 7.819967428062524e-06, "loss": 0.013, "step": 82860 }, { "grad_norm": 0.14065656065940857, "learning_rate": 7.811091079857274e-06, "loss": 0.0154, "step": 82870 }, { "grad_norm": 0.17216050624847412, "learning_rate": 7.802219345367667e-06, "loss": 0.0112, "step": 82880 }, { "grad_norm": 0.159990593791008, "learning_rate": 7.793352225563916e-06, "loss": 0.0121, "step": 82890 }, { "grad_norm": 0.1764681488275528, "learning_rate": 7.7844897214157e-06, "loss": 0.0122, "step": 82900 }, { "grad_norm": 0.1487869918346405, "learning_rate": 7.775631833892228e-06, "loss": 0.0131, "step": 82910 }, { "grad_norm": 0.13829761743545532, "learning_rate": 7.76677856396215e-06, "loss": 0.0125, "step": 82920 }, { "grad_norm": 0.1380368173122406, "learning_rate": 7.757929912593697e-06, "loss": 0.0141, "step": 82930 }, { "grad_norm": 0.1619773507118225, "learning_rate": 7.749085880754509e-06, "loss": 0.0107, "step": 82940 }, { "grad_norm": 0.09218413382768631, "learning_rate": 7.740246469411777e-06, "loss": 0.0117, "step": 82950 }, { "grad_norm": 0.35575175285339355, "learning_rate": 7.731411679532142e-06, "loss": 0.0111, "step": 82960 }, { "grad_norm": 0.1671118289232254, "learning_rate": 7.722581512081779e-06, "loss": 0.0121, "step": 82970 }, { "grad_norm": 0.270058810710907, "learning_rate": 7.713755968026331e-06, "loss": 0.0136, "step": 82980 }, { "grad_norm": 0.17740505933761597, "learning_rate": 7.704935048330969e-06, "loss": 0.0134, "step": 82990 }, { "grad_norm": 0.1797732710838318, "learning_rate": 7.696118753960313e-06, "loss": 0.013, "step": 83000 }, { "grad_norm": 0.2283283770084381, "learning_rate": 7.687307085878503e-06, "loss": 0.0137, "step": 83010 }, { "grad_norm": 0.1520545929670334, "learning_rate": 7.678500045049186e-06, "loss": 0.0116, "step": 83020 }, { "grad_norm": 0.17128384113311768, "learning_rate": 7.669697632435463e-06, "loss": 0.0119, "step": 83030 }, { "grad_norm": 0.15959954261779785, "learning_rate": 7.660899848999963e-06, "loss": 0.0106, "step": 83040 }, { "grad_norm": 0.11625585705041885, "learning_rate": 7.652106695704803e-06, "loss": 0.0122, "step": 83050 }, { "grad_norm": 0.3128902018070221, "learning_rate": 7.643318173511598e-06, "loss": 0.0119, "step": 83060 }, { "grad_norm": 0.12383610755205154, "learning_rate": 7.634534283381423e-06, "loss": 0.0121, "step": 83070 }, { "grad_norm": 0.12492502480745316, "learning_rate": 7.625755026274889e-06, "loss": 0.0115, "step": 83080 }, { "grad_norm": 0.3320217728614807, "learning_rate": 7.616980403152069e-06, "loss": 0.0112, "step": 83090 }, { "grad_norm": 0.15986928343772888, "learning_rate": 7.608210414972561e-06, "loss": 0.0112, "step": 83100 }, { "grad_norm": 0.15084856748580933, "learning_rate": 7.599445062695404e-06, "loss": 0.0115, "step": 83110 }, { "grad_norm": 0.13460318744182587, "learning_rate": 7.590684347279214e-06, "loss": 0.0122, "step": 83120 }, { "grad_norm": 0.18207113444805145, "learning_rate": 7.581928269682004e-06, "loss": 0.0146, "step": 83130 }, { "grad_norm": 0.10258660465478897, "learning_rate": 7.573176830861361e-06, "loss": 0.0123, "step": 83140 }, { "grad_norm": 0.1329035758972168, "learning_rate": 7.5644300317743e-06, "loss": 0.0105, "step": 83150 }, { "grad_norm": 0.20309244096279144, "learning_rate": 7.555687873377376e-06, "loss": 0.0135, "step": 83160 }, { "grad_norm": 0.19734174013137817, "learning_rate": 7.546950356626592e-06, "loss": 0.0151, "step": 83170 }, { "grad_norm": 0.2043158859014511, "learning_rate": 7.538217482477516e-06, "loss": 0.0115, "step": 83180 }, { "grad_norm": 0.18203286826610565, "learning_rate": 7.529489251885119e-06, "loss": 0.0114, "step": 83190 }, { "grad_norm": 0.14777866005897522, "learning_rate": 7.520765665803941e-06, "loss": 0.0109, "step": 83200 }, { "grad_norm": 0.17172646522521973, "learning_rate": 7.512046725187949e-06, "loss": 0.0123, "step": 83210 }, { "grad_norm": 0.19525592029094696, "learning_rate": 7.503332430990645e-06, "loss": 0.0112, "step": 83220 }, { "grad_norm": 0.1284187287092209, "learning_rate": 7.494622784165017e-06, "loss": 0.0111, "step": 83230 }, { "grad_norm": 0.16648077964782715, "learning_rate": 7.485917785663543e-06, "loss": 0.0128, "step": 83240 }, { "grad_norm": 0.10020885616540909, "learning_rate": 7.477217436438172e-06, "loss": 0.0096, "step": 83250 }, { "grad_norm": 0.23233452439308167, "learning_rate": 7.46852173744037e-06, "loss": 0.0131, "step": 83260 }, { "grad_norm": 0.15227244794368744, "learning_rate": 7.4598306896211004e-06, "loss": 0.0119, "step": 83270 }, { "grad_norm": 0.13216061890125275, "learning_rate": 7.451144293930773e-06, "loss": 0.0106, "step": 83280 }, { "grad_norm": 0.11855597794055939, "learning_rate": 7.4424625513193455e-06, "loss": 0.0111, "step": 83290 }, { "grad_norm": 0.20103822648525238, "learning_rate": 7.433785462736209e-06, "loss": 0.0121, "step": 83300 }, { "grad_norm": 0.1064986139535904, "learning_rate": 7.425113029130315e-06, "loss": 0.0124, "step": 83310 }, { "grad_norm": 0.1639123111963272, "learning_rate": 7.416445251450044e-06, "loss": 0.0096, "step": 83320 }, { "grad_norm": 0.13740019500255585, "learning_rate": 7.4077821306433035e-06, "loss": 0.0154, "step": 83330 }, { "grad_norm": 0.23871809244155884, "learning_rate": 7.399123667657465e-06, "loss": 0.0113, "step": 83340 }, { "grad_norm": 0.2082359939813614, "learning_rate": 7.390469863439426e-06, "loss": 0.013, "step": 83350 }, { "grad_norm": 0.17251744866371155, "learning_rate": 7.381820718935512e-06, "loss": 0.0119, "step": 83360 }, { "grad_norm": 0.21455128490924835, "learning_rate": 7.373176235091633e-06, "loss": 0.0114, "step": 83370 }, { "grad_norm": 0.18081659078598022, "learning_rate": 7.364536412853101e-06, "loss": 0.0128, "step": 83380 }, { "grad_norm": 0.17142871022224426, "learning_rate": 7.355901253164776e-06, "loss": 0.0105, "step": 83390 }, { "grad_norm": 0.19242963194847107, "learning_rate": 7.34727075697097e-06, "loss": 0.0115, "step": 83400 }, { "grad_norm": 0.1101682037115097, "learning_rate": 7.3386449252155095e-06, "loss": 0.0106, "step": 83410 }, { "grad_norm": 0.20022937655448914, "learning_rate": 7.330023758841692e-06, "loss": 0.0112, "step": 83420 }, { "grad_norm": 0.17874999344348907, "learning_rate": 7.321407258792323e-06, "loss": 0.0119, "step": 83430 }, { "grad_norm": 0.14412017166614532, "learning_rate": 7.312795426009694e-06, "loss": 0.0109, "step": 83440 }, { "grad_norm": 0.28093478083610535, "learning_rate": 7.304188261435585e-06, "loss": 0.0108, "step": 83450 }, { "grad_norm": 0.1890256553888321, "learning_rate": 7.295585766011242e-06, "loss": 0.011, "step": 83460 }, { "grad_norm": 0.36836758255958557, "learning_rate": 7.286987940677436e-06, "loss": 0.0103, "step": 83470 }, { "grad_norm": 0.21565473079681396, "learning_rate": 7.278394786374426e-06, "loss": 0.0138, "step": 83480 }, { "grad_norm": 0.3429165482521057, "learning_rate": 7.269806304041915e-06, "loss": 0.0103, "step": 83490 }, { "grad_norm": 0.23343828320503235, "learning_rate": 7.261222494619147e-06, "loss": 0.0108, "step": 83500 }, { "grad_norm": 0.11961448192596436, "learning_rate": 7.252643359044825e-06, "loss": 0.0105, "step": 83510 }, { "grad_norm": 0.1875803917646408, "learning_rate": 7.244068898257172e-06, "loss": 0.012, "step": 83520 }, { "grad_norm": 0.2323296070098877, "learning_rate": 7.2354991131938495e-06, "loss": 0.0138, "step": 83530 }, { "grad_norm": 0.3249804973602295, "learning_rate": 7.2269340047920574e-06, "loss": 0.0152, "step": 83540 }, { "grad_norm": 0.1788995862007141, "learning_rate": 7.218373573988435e-06, "loss": 0.0123, "step": 83550 }, { "grad_norm": 0.20319464802742004, "learning_rate": 7.209817821719178e-06, "loss": 0.0116, "step": 83560 }, { "grad_norm": 0.17338936030864716, "learning_rate": 7.201266748919894e-06, "loss": 0.0165, "step": 83570 }, { "grad_norm": 0.11603709310293198, "learning_rate": 7.19272035652575e-06, "loss": 0.0087, "step": 83580 }, { "grad_norm": 0.20385447144508362, "learning_rate": 7.184178645471334e-06, "loss": 0.0147, "step": 83590 }, { "grad_norm": 0.21038590371608734, "learning_rate": 7.175641616690775e-06, "loss": 0.0122, "step": 83600 }, { "grad_norm": 0.12451139092445374, "learning_rate": 7.167109271117661e-06, "loss": 0.0124, "step": 83610 }, { "grad_norm": 0.18473972380161285, "learning_rate": 7.158581609685072e-06, "loss": 0.0107, "step": 83620 }, { "grad_norm": 0.26167944073677063, "learning_rate": 7.150058633325591e-06, "loss": 0.0143, "step": 83630 }, { "grad_norm": 0.10319274663925171, "learning_rate": 7.141540342971287e-06, "loss": 0.0123, "step": 83640 }, { "grad_norm": 0.1574818342924118, "learning_rate": 7.133026739553677e-06, "loss": 0.0116, "step": 83650 }, { "grad_norm": 0.12414370477199554, "learning_rate": 7.124517824003829e-06, "loss": 0.0109, "step": 83660 }, { "grad_norm": 0.2334490865468979, "learning_rate": 7.116013597252236e-06, "loss": 0.012, "step": 83670 }, { "grad_norm": 0.25155267119407654, "learning_rate": 7.107514060228921e-06, "loss": 0.0108, "step": 83680 }, { "grad_norm": 0.16877751052379608, "learning_rate": 7.099019213863378e-06, "loss": 0.0148, "step": 83690 }, { "grad_norm": 0.1851213127374649, "learning_rate": 7.090529059084611e-06, "loss": 0.0127, "step": 83700 }, { "grad_norm": 0.18403276801109314, "learning_rate": 7.082043596821058e-06, "loss": 0.0097, "step": 83710 }, { "grad_norm": 0.3174470067024231, "learning_rate": 7.07356282800069e-06, "loss": 0.0121, "step": 83720 }, { "grad_norm": 0.1934674084186554, "learning_rate": 7.0650867535509645e-06, "loss": 0.0105, "step": 83730 }, { "grad_norm": 1.0287612676620483, "learning_rate": 7.0566153743987895e-06, "loss": 0.0104, "step": 83740 }, { "grad_norm": 0.16184726357460022, "learning_rate": 7.04814869147059e-06, "loss": 0.0126, "step": 83750 }, { "grad_norm": 0.26424622535705566, "learning_rate": 7.0396867056922725e-06, "loss": 0.0137, "step": 83760 }, { "grad_norm": 0.18233248591423035, "learning_rate": 7.031229417989244e-06, "loss": 0.0093, "step": 83770 }, { "grad_norm": 0.17579616606235504, "learning_rate": 7.022776829286348e-06, "loss": 0.0144, "step": 83780 }, { "grad_norm": 0.16235274076461792, "learning_rate": 7.014328940507975e-06, "loss": 0.01, "step": 83790 }, { "grad_norm": 0.1634349226951599, "learning_rate": 7.005885752577945e-06, "loss": 0.0116, "step": 83800 }, { "grad_norm": 0.2535852789878845, "learning_rate": 6.997447266419615e-06, "loss": 0.0156, "step": 83810 }, { "grad_norm": 0.10995502769947052, "learning_rate": 6.989013482955797e-06, "loss": 0.012, "step": 83820 }, { "grad_norm": 0.18594112992286682, "learning_rate": 6.980584403108808e-06, "loss": 0.0111, "step": 83830 }, { "grad_norm": 0.16351059079170227, "learning_rate": 6.972160027800417e-06, "loss": 0.0138, "step": 83840 }, { "grad_norm": 0.21746253967285156, "learning_rate": 6.963740357951931e-06, "loss": 0.0124, "step": 83850 }, { "grad_norm": 0.1252880096435547, "learning_rate": 6.95532539448408e-06, "loss": 0.0122, "step": 83860 }, { "grad_norm": 0.1550704538822174, "learning_rate": 6.946915138317128e-06, "loss": 0.0094, "step": 83870 }, { "grad_norm": 0.18700093030929565, "learning_rate": 6.9385095903708065e-06, "loss": 0.0123, "step": 83880 }, { "grad_norm": 0.1462961733341217, "learning_rate": 6.930108751564346e-06, "loss": 0.0093, "step": 83890 }, { "grad_norm": 0.20974041521549225, "learning_rate": 6.921712622816428e-06, "loss": 0.0101, "step": 83900 }, { "grad_norm": 0.16493551433086395, "learning_rate": 6.913321205045259e-06, "loss": 0.0126, "step": 83910 }, { "grad_norm": 0.14894497394561768, "learning_rate": 6.904934499168491e-06, "loss": 0.0108, "step": 83920 }, { "grad_norm": 0.1546076089143753, "learning_rate": 6.896552506103299e-06, "loss": 0.0126, "step": 83930 }, { "grad_norm": 0.19119712710380554, "learning_rate": 6.888175226766313e-06, "loss": 0.0117, "step": 83940 }, { "grad_norm": 0.42293140292167664, "learning_rate": 6.879802662073686e-06, "loss": 0.0118, "step": 83950 }, { "grad_norm": 0.29641038179397583, "learning_rate": 6.871434812940991e-06, "loss": 0.013, "step": 83960 }, { "grad_norm": 1.2425408363342285, "learning_rate": 6.863071680283345e-06, "loss": 0.0129, "step": 83970 }, { "grad_norm": 0.16299359500408173, "learning_rate": 6.854713265015339e-06, "loss": 0.0129, "step": 83980 }, { "grad_norm": 0.11930516362190247, "learning_rate": 6.8463595680510025e-06, "loss": 0.0112, "step": 83990 }, { "grad_norm": 0.2075105458498001, "learning_rate": 6.838010590303911e-06, "loss": 0.0116, "step": 84000 }, { "grad_norm": 0.14401517808437347, "learning_rate": 6.829666332687085e-06, "loss": 0.0113, "step": 84010 }, { "grad_norm": 0.241710364818573, "learning_rate": 6.821326796113059e-06, "loss": 0.0137, "step": 84020 }, { "grad_norm": 0.11581569164991379, "learning_rate": 6.812991981493799e-06, "loss": 0.0125, "step": 84030 }, { "grad_norm": 0.22721511125564575, "learning_rate": 6.804661889740821e-06, "loss": 0.012, "step": 84040 }, { "grad_norm": 0.17774000763893127, "learning_rate": 6.7963365217650625e-06, "loss": 0.012, "step": 84050 }, { "grad_norm": 0.1329130381345749, "learning_rate": 6.788015878476983e-06, "loss": 0.012, "step": 84060 }, { "grad_norm": 0.23115216195583344, "learning_rate": 6.779699960786523e-06, "loss": 0.0104, "step": 84070 }, { "grad_norm": 0.2827214002609253, "learning_rate": 6.771388769603104e-06, "loss": 0.0108, "step": 84080 }, { "grad_norm": 0.25207510590553284, "learning_rate": 6.763082305835605e-06, "loss": 0.0101, "step": 84090 }, { "grad_norm": 0.2839023768901825, "learning_rate": 6.754780570392433e-06, "loss": 0.0109, "step": 84100 }, { "grad_norm": 0.21629740297794342, "learning_rate": 6.7464835641814215e-06, "loss": 0.0135, "step": 84110 }, { "grad_norm": 0.27747756242752075, "learning_rate": 6.7381912881099515e-06, "loss": 0.0152, "step": 84120 }, { "grad_norm": 0.1536925882101059, "learning_rate": 6.72990374308482e-06, "loss": 0.0111, "step": 84130 }, { "grad_norm": 0.16558662056922913, "learning_rate": 6.721620930012379e-06, "loss": 0.012, "step": 84140 }, { "grad_norm": 0.15909814834594727, "learning_rate": 6.713342849798393e-06, "loss": 0.0106, "step": 84150 }, { "grad_norm": 0.13560135662555695, "learning_rate": 6.705069503348166e-06, "loss": 0.0116, "step": 84160 }, { "grad_norm": 0.16916057467460632, "learning_rate": 6.696800891566435e-06, "loss": 0.0147, "step": 84170 }, { "grad_norm": 0.160344198346138, "learning_rate": 6.688537015357449e-06, "loss": 0.0138, "step": 84180 }, { "grad_norm": 0.10717741400003433, "learning_rate": 6.680277875624941e-06, "loss": 0.0114, "step": 84190 }, { "grad_norm": 0.1819750815629959, "learning_rate": 6.672023473272121e-06, "loss": 0.0135, "step": 84200 }, { "grad_norm": 0.27929726243019104, "learning_rate": 6.663773809201667e-06, "loss": 0.0118, "step": 84210 }, { "grad_norm": 0.19436368346214294, "learning_rate": 6.655528884315748e-06, "loss": 0.0104, "step": 84220 }, { "grad_norm": 0.25332126021385193, "learning_rate": 6.647288699516036e-06, "loss": 0.0132, "step": 84230 }, { "grad_norm": 0.145548015832901, "learning_rate": 6.639053255703642e-06, "loss": 0.0098, "step": 84240 }, { "grad_norm": 0.2768349349498749, "learning_rate": 6.6308225537791925e-06, "loss": 0.0121, "step": 84250 }, { "grad_norm": 0.1705566942691803, "learning_rate": 6.622596594642777e-06, "loss": 0.0135, "step": 84260 }, { "grad_norm": 0.19988499581813812, "learning_rate": 6.614375379193999e-06, "loss": 0.0126, "step": 84270 }, { "grad_norm": 0.28897199034690857, "learning_rate": 6.606158908331889e-06, "loss": 0.0115, "step": 84280 }, { "grad_norm": 0.1518479883670807, "learning_rate": 6.597947182955005e-06, "loss": 0.0138, "step": 84290 }, { "grad_norm": 0.11564639210700989, "learning_rate": 6.589740203961359e-06, "loss": 0.0115, "step": 84300 }, { "grad_norm": 0.19638219475746155, "learning_rate": 6.581537972248464e-06, "loss": 0.0129, "step": 84310 }, { "grad_norm": 0.14682893455028534, "learning_rate": 6.573340488713275e-06, "loss": 0.0126, "step": 84320 }, { "grad_norm": 0.15163759887218475, "learning_rate": 6.565147754252304e-06, "loss": 0.0127, "step": 84330 }, { "grad_norm": 0.2235301285982132, "learning_rate": 6.556959769761462e-06, "loss": 0.0114, "step": 84340 }, { "grad_norm": 0.18901757895946503, "learning_rate": 6.548776536136192e-06, "loss": 0.0128, "step": 84350 }, { "grad_norm": 0.2448417842388153, "learning_rate": 6.54059805427138e-06, "loss": 0.0125, "step": 84360 }, { "grad_norm": 0.18642625212669373, "learning_rate": 6.532424325061437e-06, "loss": 0.0105, "step": 84370 }, { "grad_norm": 0.14271223545074463, "learning_rate": 6.5242553494001935e-06, "loss": 0.0121, "step": 84380 }, { "grad_norm": 0.19295360147953033, "learning_rate": 6.516091128181046e-06, "loss": 0.0114, "step": 84390 }, { "grad_norm": 0.1477447897195816, "learning_rate": 6.507931662296785e-06, "loss": 0.013, "step": 84400 }, { "grad_norm": 0.8841858506202698, "learning_rate": 6.4997769526397354e-06, "loss": 0.014, "step": 84410 }, { "grad_norm": 0.11084119230508804, "learning_rate": 6.491627000101663e-06, "loss": 0.0113, "step": 84420 }, { "grad_norm": 0.11963005363941193, "learning_rate": 6.4834818055738545e-06, "loss": 0.0141, "step": 84430 }, { "grad_norm": 0.1340971440076828, "learning_rate": 6.4753413699470465e-06, "loss": 0.0107, "step": 84440 }, { "grad_norm": 0.20113228261470795, "learning_rate": 6.467205694111478e-06, "loss": 0.0112, "step": 84450 }, { "grad_norm": 0.3786192834377289, "learning_rate": 6.459074778956836e-06, "loss": 0.0104, "step": 84460 }, { "grad_norm": 0.1082519143819809, "learning_rate": 6.450948625372311e-06, "loss": 0.0102, "step": 84470 }, { "grad_norm": 0.17141865193843842, "learning_rate": 6.442827234246585e-06, "loss": 0.0133, "step": 84480 }, { "grad_norm": 0.3395198583602905, "learning_rate": 6.43471060646777e-06, "loss": 0.011, "step": 84490 }, { "grad_norm": 0.15613137185573578, "learning_rate": 6.426598742923512e-06, "loss": 0.0109, "step": 84500 }, { "grad_norm": 0.1658208668231964, "learning_rate": 6.418491644500885e-06, "loss": 0.0127, "step": 84510 }, { "grad_norm": 0.13022209703922272, "learning_rate": 6.410389312086512e-06, "loss": 0.0113, "step": 84520 }, { "grad_norm": 0.24034537374973297, "learning_rate": 6.4022917465664135e-06, "loss": 0.0117, "step": 84530 }, { "grad_norm": 0.20328837633132935, "learning_rate": 6.3941989488261575e-06, "loss": 0.0126, "step": 84540 }, { "grad_norm": 0.20694254338741302, "learning_rate": 6.386110919750726e-06, "loss": 0.01, "step": 84550 }, { "grad_norm": 0.17096473276615143, "learning_rate": 6.378027660224651e-06, "loss": 0.011, "step": 84560 }, { "grad_norm": 0.2004580795764923, "learning_rate": 6.369949171131856e-06, "loss": 0.0091, "step": 84570 }, { "grad_norm": 0.2066282033920288, "learning_rate": 6.361875453355848e-06, "loss": 0.0118, "step": 84580 }, { "grad_norm": 0.126470685005188, "learning_rate": 6.353806507779525e-06, "loss": 0.0108, "step": 84590 }, { "grad_norm": 0.2046724557876587, "learning_rate": 6.3457423352853044e-06, "loss": 0.012, "step": 84600 }, { "grad_norm": 0.19800464808940887, "learning_rate": 6.337682936755062e-06, "loss": 0.0108, "step": 84610 }, { "grad_norm": 0.1541948914527893, "learning_rate": 6.329628313070179e-06, "loss": 0.0117, "step": 84620 }, { "grad_norm": 0.2782445251941681, "learning_rate": 6.321578465111477e-06, "loss": 0.0181, "step": 84630 }, { "grad_norm": 0.17025770246982574, "learning_rate": 6.31353339375928e-06, "loss": 0.0125, "step": 84640 }, { "grad_norm": 0.2703959345817566, "learning_rate": 6.305493099893395e-06, "loss": 0.0107, "step": 84650 }, { "grad_norm": 0.19065064191818237, "learning_rate": 6.297457584393096e-06, "loss": 0.0114, "step": 84660 }, { "grad_norm": 0.43725869059562683, "learning_rate": 6.289426848137126e-06, "loss": 0.012, "step": 84670 }, { "grad_norm": 0.16085118055343628, "learning_rate": 6.281400892003713e-06, "loss": 0.0095, "step": 84680 }, { "grad_norm": 0.2062482386827469, "learning_rate": 6.273379716870581e-06, "loss": 0.0127, "step": 84690 }, { "grad_norm": 0.12204960733652115, "learning_rate": 6.265363323614887e-06, "loss": 0.0131, "step": 84700 }, { "grad_norm": 0.18176795542240143, "learning_rate": 6.257351713113308e-06, "loss": 0.0101, "step": 84710 }, { "grad_norm": 0.16983754932880402, "learning_rate": 6.249344886241981e-06, "loss": 0.0116, "step": 84720 }, { "grad_norm": 0.23615191876888275, "learning_rate": 6.2413428438765295e-06, "loss": 0.014, "step": 84730 }, { "grad_norm": 0.2497939169406891, "learning_rate": 6.233345586892026e-06, "loss": 0.0115, "step": 84740 }, { "grad_norm": 0.2476065754890442, "learning_rate": 6.225353116163052e-06, "loss": 0.0117, "step": 84750 }, { "grad_norm": 0.10678020119667053, "learning_rate": 6.217365432563627e-06, "loss": 0.0123, "step": 84760 }, { "grad_norm": 0.19054222106933594, "learning_rate": 6.209382536967318e-06, "loss": 0.0117, "step": 84770 }, { "grad_norm": 0.2601439654827118, "learning_rate": 6.201404430247082e-06, "loss": 0.0127, "step": 84780 }, { "grad_norm": 0.12925434112548828, "learning_rate": 6.1934311132754165e-06, "loss": 0.0104, "step": 84790 }, { "grad_norm": 0.10752502083778381, "learning_rate": 6.185462586924257e-06, "loss": 0.0107, "step": 84800 }, { "grad_norm": 0.24786849319934845, "learning_rate": 6.177498852065039e-06, "loss": 0.0111, "step": 84810 }, { "grad_norm": 0.1629343181848526, "learning_rate": 6.169539909568655e-06, "loss": 0.0128, "step": 84820 }, { "grad_norm": 0.20870822668075562, "learning_rate": 6.161585760305488e-06, "loss": 0.0111, "step": 84830 }, { "grad_norm": 0.2948463559150696, "learning_rate": 6.153636405145391e-06, "loss": 0.0122, "step": 84840 }, { "grad_norm": 0.17108607292175293, "learning_rate": 6.145691844957702e-06, "loss": 0.0107, "step": 84850 }, { "grad_norm": 0.18554283678531647, "learning_rate": 6.1377520806112145e-06, "loss": 0.0174, "step": 84860 }, { "grad_norm": 0.11398065090179443, "learning_rate": 6.129817112974218e-06, "loss": 0.0097, "step": 84870 }, { "grad_norm": 0.1932564228773117, "learning_rate": 6.121886942914457e-06, "loss": 0.0107, "step": 84880 }, { "grad_norm": 0.28265443444252014, "learning_rate": 6.11396157129917e-06, "loss": 0.0136, "step": 84890 }, { "grad_norm": 0.17333455383777618, "learning_rate": 6.106040998995066e-06, "loss": 0.0112, "step": 84900 }, { "grad_norm": 0.13046947121620178, "learning_rate": 6.098125226868329e-06, "loss": 0.0128, "step": 84910 }, { "grad_norm": 0.24619336426258087, "learning_rate": 6.090214255784604e-06, "loss": 0.0102, "step": 84920 }, { "grad_norm": 0.1448274403810501, "learning_rate": 6.0823080866090346e-06, "loss": 0.0128, "step": 84930 }, { "grad_norm": 0.15332633256912231, "learning_rate": 6.074406720206233e-06, "loss": 0.0119, "step": 84940 }, { "grad_norm": 0.21747849881649017, "learning_rate": 6.066510157440259e-06, "loss": 0.0125, "step": 84950 }, { "grad_norm": 0.1863887906074524, "learning_rate": 6.058618399174676e-06, "loss": 0.0141, "step": 84960 }, { "grad_norm": 0.2350151091814041, "learning_rate": 6.050731446272528e-06, "loss": 0.0087, "step": 84970 }, { "grad_norm": 0.16030241549015045, "learning_rate": 6.042849299596321e-06, "loss": 0.0132, "step": 84980 }, { "grad_norm": 0.14347495138645172, "learning_rate": 6.034971960008013e-06, "loss": 0.0126, "step": 84990 }, { "grad_norm": 0.1243543028831482, "learning_rate": 6.027099428369082e-06, "loss": 0.0127, "step": 85000 }, { "grad_norm": 0.12754414975643158, "learning_rate": 6.019231705540435e-06, "loss": 0.0138, "step": 85010 }, { "grad_norm": 0.1301470249891281, "learning_rate": 6.011368792382488e-06, "loss": 0.0102, "step": 85020 }, { "grad_norm": 0.34729063510894775, "learning_rate": 6.00351068975511e-06, "loss": 0.0128, "step": 85030 }, { "grad_norm": 0.18736036121845245, "learning_rate": 5.9956573985176674e-06, "loss": 0.0122, "step": 85040 }, { "grad_norm": 0.2098030000925064, "learning_rate": 5.987808919528959e-06, "loss": 0.012, "step": 85050 }, { "grad_norm": 0.19752362370491028, "learning_rate": 5.979965253647307e-06, "loss": 0.013, "step": 85060 }, { "grad_norm": 0.142939031124115, "learning_rate": 5.9721264017304655e-06, "loss": 0.0104, "step": 85070 }, { "grad_norm": 0.13789187371730804, "learning_rate": 5.964292364635682e-06, "loss": 0.012, "step": 85080 }, { "grad_norm": 0.14470885694026947, "learning_rate": 5.956463143219682e-06, "loss": 0.0112, "step": 85090 }, { "grad_norm": 0.27355697751045227, "learning_rate": 5.948638738338663e-06, "loss": 0.0118, "step": 85100 }, { "grad_norm": 0.11066148430109024, "learning_rate": 5.940819150848271e-06, "loss": 0.0102, "step": 85110 }, { "grad_norm": 0.150994673371315, "learning_rate": 5.933004381603663e-06, "loss": 0.0119, "step": 85120 }, { "grad_norm": 0.25982651114463806, "learning_rate": 5.92519443145943e-06, "loss": 0.0118, "step": 85130 }, { "grad_norm": 0.1421002596616745, "learning_rate": 5.917389301269671e-06, "loss": 0.0123, "step": 85140 }, { "grad_norm": 0.1965571790933609, "learning_rate": 5.9095889918879435e-06, "loss": 0.0115, "step": 85150 }, { "grad_norm": 0.16950544714927673, "learning_rate": 5.901793504167286e-06, "loss": 0.0099, "step": 85160 }, { "grad_norm": 0.2052336484193802, "learning_rate": 5.894002838960172e-06, "loss": 0.0128, "step": 85170 }, { "grad_norm": 0.17644183337688446, "learning_rate": 5.886216997118604e-06, "loss": 0.0125, "step": 85180 }, { "grad_norm": 0.24983148276805878, "learning_rate": 5.878435979494023e-06, "loss": 0.02, "step": 85190 }, { "grad_norm": 0.14659813046455383, "learning_rate": 5.870659786937344e-06, "loss": 0.0102, "step": 85200 }, { "grad_norm": 0.2530427873134613, "learning_rate": 5.862888420298962e-06, "loss": 0.0135, "step": 85210 }, { "grad_norm": 0.28059762716293335, "learning_rate": 5.855121880428743e-06, "loss": 0.0097, "step": 85220 }, { "grad_norm": 0.2608793079853058, "learning_rate": 5.84736016817603e-06, "loss": 0.0111, "step": 85230 }, { "grad_norm": 0.29306313395500183, "learning_rate": 5.839603284389622e-06, "loss": 0.0119, "step": 85240 }, { "grad_norm": 0.1500820815563202, "learning_rate": 5.831851229917812e-06, "loss": 0.012, "step": 85250 }, { "grad_norm": 0.100191630423069, "learning_rate": 5.824104005608338e-06, "loss": 0.0116, "step": 85260 }, { "grad_norm": 0.11071357876062393, "learning_rate": 5.81636161230843e-06, "loss": 0.0121, "step": 85270 }, { "grad_norm": 0.24734316766262054, "learning_rate": 5.808624050864791e-06, "loss": 0.0129, "step": 85280 }, { "grad_norm": 0.12519970536231995, "learning_rate": 5.800891322123592e-06, "loss": 0.0106, "step": 85290 }, { "grad_norm": 0.12312903255224228, "learning_rate": 5.793163426930454e-06, "loss": 0.0125, "step": 85300 }, { "grad_norm": 0.16576328873634338, "learning_rate": 5.785440366130512e-06, "loss": 0.0111, "step": 85310 }, { "grad_norm": 0.12450428307056427, "learning_rate": 5.777722140568326e-06, "loss": 0.0111, "step": 85320 }, { "grad_norm": 0.08322510868310928, "learning_rate": 5.7700087510879685e-06, "loss": 0.0126, "step": 85330 }, { "grad_norm": 0.14471690356731415, "learning_rate": 5.762300198532933e-06, "loss": 0.0119, "step": 85340 }, { "grad_norm": 0.15394781529903412, "learning_rate": 5.7545964837462606e-06, "loss": 0.0115, "step": 85350 }, { "grad_norm": 0.1946907341480255, "learning_rate": 5.746897607570384e-06, "loss": 0.0102, "step": 85360 }, { "grad_norm": 0.1548563688993454, "learning_rate": 5.7392035708472626e-06, "loss": 0.0107, "step": 85370 }, { "grad_norm": 0.31359001994132996, "learning_rate": 5.731514374418278e-06, "loss": 0.015, "step": 85380 }, { "grad_norm": 0.22398188710212708, "learning_rate": 5.7238300191243325e-06, "loss": 0.0143, "step": 85390 }, { "grad_norm": 0.18725545704364777, "learning_rate": 5.716150505805762e-06, "loss": 0.0118, "step": 85400 }, { "grad_norm": 0.16962014138698578, "learning_rate": 5.708475835302407e-06, "loss": 0.013, "step": 85410 }, { "grad_norm": 0.2019203156232834, "learning_rate": 5.70080600845353e-06, "loss": 0.0119, "step": 85420 }, { "grad_norm": 0.19721584022045135, "learning_rate": 5.693141026097909e-06, "loss": 0.0119, "step": 85430 }, { "grad_norm": 0.16112326085567474, "learning_rate": 5.6854808890737745e-06, "loss": 0.0111, "step": 85440 }, { "grad_norm": 0.18288995325565338, "learning_rate": 5.6778255982188225e-06, "loss": 0.0124, "step": 85450 }, { "grad_norm": 0.1721353530883789, "learning_rate": 5.670175154370217e-06, "loss": 0.0118, "step": 85460 }, { "grad_norm": 0.1662585437297821, "learning_rate": 5.662529558364616e-06, "loss": 0.0122, "step": 85470 }, { "grad_norm": 0.18719345331192017, "learning_rate": 5.654888811038128e-06, "loss": 0.0119, "step": 85480 }, { "grad_norm": 0.16806873679161072, "learning_rate": 5.647252913226314e-06, "loss": 0.0127, "step": 85490 }, { "grad_norm": 0.13515044748783112, "learning_rate": 5.639621865764255e-06, "loss": 0.0113, "step": 85500 }, { "grad_norm": 0.23010480403900146, "learning_rate": 5.63199566948644e-06, "loss": 0.015, "step": 85510 }, { "grad_norm": 0.18099597096443176, "learning_rate": 5.6243743252268835e-06, "loss": 0.0093, "step": 85520 }, { "grad_norm": 0.19242338836193085, "learning_rate": 5.616757833819009e-06, "loss": 0.0109, "step": 85530 }, { "grad_norm": 0.13126298785209656, "learning_rate": 5.609146196095794e-06, "loss": 0.0093, "step": 85540 }, { "grad_norm": 0.14576667547225952, "learning_rate": 5.6015394128896005e-06, "loss": 0.013, "step": 85550 }, { "grad_norm": 0.23618434369564056, "learning_rate": 5.593937485032308e-06, "loss": 0.012, "step": 85560 }, { "grad_norm": 0.15044371783733368, "learning_rate": 5.586340413355246e-06, "loss": 0.01, "step": 85570 }, { "grad_norm": 0.24435491859912872, "learning_rate": 5.578748198689227e-06, "loss": 0.0126, "step": 85580 }, { "grad_norm": 0.15656208992004395, "learning_rate": 5.571160841864504e-06, "loss": 0.012, "step": 85590 }, { "grad_norm": 0.30128714442253113, "learning_rate": 5.563578343710846e-06, "loss": 0.0111, "step": 85600 }, { "grad_norm": 0.16796875, "learning_rate": 5.556000705057446e-06, "loss": 0.0121, "step": 85610 }, { "grad_norm": 0.16327926516532898, "learning_rate": 5.548427926732996e-06, "loss": 0.0097, "step": 85620 }, { "grad_norm": 0.28123676776885986, "learning_rate": 5.540860009565629e-06, "loss": 0.0168, "step": 85630 }, { "grad_norm": 0.20328107476234436, "learning_rate": 5.533296954382966e-06, "loss": 0.0134, "step": 85640 }, { "grad_norm": 0.19192908704280853, "learning_rate": 5.525738762012095e-06, "loss": 0.0123, "step": 85650 }, { "grad_norm": 0.23178896307945251, "learning_rate": 5.518185433279577e-06, "loss": 0.0125, "step": 85660 }, { "grad_norm": 0.13457684218883514, "learning_rate": 5.510636969011407e-06, "loss": 0.0118, "step": 85670 }, { "grad_norm": 0.28217220306396484, "learning_rate": 5.503093370033096e-06, "loss": 0.0119, "step": 85680 }, { "grad_norm": 0.2751167416572571, "learning_rate": 5.495554637169603e-06, "loss": 0.0125, "step": 85690 }, { "grad_norm": 0.17710131406784058, "learning_rate": 5.488020771245328e-06, "loss": 0.0111, "step": 85700 }, { "grad_norm": 0.13031519949436188, "learning_rate": 5.480491773084195e-06, "loss": 0.0113, "step": 85710 }, { "grad_norm": 0.31801995635032654, "learning_rate": 5.472967643509525e-06, "loss": 0.0132, "step": 85720 }, { "grad_norm": 0.1310553401708603, "learning_rate": 5.46544838334419e-06, "loss": 0.0091, "step": 85730 }, { "grad_norm": 0.17358505725860596, "learning_rate": 5.4579339934104525e-06, "loss": 0.0116, "step": 85740 }, { "grad_norm": 0.2228064239025116, "learning_rate": 5.450424474530091e-06, "loss": 0.0105, "step": 85750 }, { "grad_norm": 0.18949368596076965, "learning_rate": 5.442919827524323e-06, "loss": 0.0169, "step": 85760 }, { "grad_norm": 0.17061810195446014, "learning_rate": 5.435420053213863e-06, "loss": 0.0113, "step": 85770 }, { "grad_norm": 0.20963892340660095, "learning_rate": 5.427925152418845e-06, "loss": 0.0118, "step": 85780 }, { "grad_norm": 0.1285812258720398, "learning_rate": 5.420435125958934e-06, "loss": 0.0135, "step": 85790 }, { "grad_norm": 0.13226918876171112, "learning_rate": 5.412949974653209e-06, "loss": 0.012, "step": 85800 }, { "grad_norm": 0.26257774233818054, "learning_rate": 5.405469699320248e-06, "loss": 0.0135, "step": 85810 }, { "grad_norm": 0.21786756813526154, "learning_rate": 5.397994300778064e-06, "loss": 0.0097, "step": 85820 }, { "grad_norm": 0.1869320571422577, "learning_rate": 5.39052377984417e-06, "loss": 0.0113, "step": 85830 }, { "grad_norm": 0.14500635862350464, "learning_rate": 5.383058137335523e-06, "loss": 0.0109, "step": 85840 }, { "grad_norm": 0.4620148539543152, "learning_rate": 5.375597374068553e-06, "loss": 0.0111, "step": 85850 }, { "grad_norm": 0.12624983489513397, "learning_rate": 5.368141490859163e-06, "loss": 0.0099, "step": 85860 }, { "grad_norm": 0.16969050467014313, "learning_rate": 5.360690488522729e-06, "loss": 0.0109, "step": 85870 }, { "grad_norm": 0.46425995230674744, "learning_rate": 5.353244367874055e-06, "loss": 0.0136, "step": 85880 }, { "grad_norm": 0.157388374209404, "learning_rate": 5.345803129727456e-06, "loss": 0.01, "step": 85890 }, { "grad_norm": 0.17125099897384644, "learning_rate": 5.338366774896697e-06, "loss": 0.0113, "step": 85900 }, { "grad_norm": 0.18529047071933746, "learning_rate": 5.330935304194995e-06, "loss": 0.0138, "step": 85910 }, { "grad_norm": 0.18781255185604095, "learning_rate": 5.323508718435044e-06, "loss": 0.0108, "step": 85920 }, { "grad_norm": 0.20639891922473907, "learning_rate": 5.316087018429011e-06, "loss": 0.0123, "step": 85930 }, { "grad_norm": 0.14187712967395782, "learning_rate": 5.30867020498853e-06, "loss": 0.0107, "step": 85940 }, { "grad_norm": 0.7246314883232117, "learning_rate": 5.30125827892467e-06, "loss": 0.0136, "step": 85950 }, { "grad_norm": 0.2029096782207489, "learning_rate": 5.293851241048015e-06, "loss": 0.012, "step": 85960 }, { "grad_norm": 0.7036951184272766, "learning_rate": 5.286449092168555e-06, "loss": 0.0121, "step": 85970 }, { "grad_norm": 0.21597075462341309, "learning_rate": 5.2790518330958105e-06, "loss": 0.0117, "step": 85980 }, { "grad_norm": 0.14373530447483063, "learning_rate": 5.271659464638717e-06, "loss": 0.0104, "step": 85990 }, { "grad_norm": 0.22310185432434082, "learning_rate": 5.264271987605701e-06, "loss": 0.0117, "step": 86000 }, { "grad_norm": 0.13466966152191162, "learning_rate": 5.256889402804632e-06, "loss": 0.0129, "step": 86010 }, { "grad_norm": 0.15675251185894012, "learning_rate": 5.249511711042881e-06, "loss": 0.0102, "step": 86020 }, { "grad_norm": 0.14216825366020203, "learning_rate": 5.242138913127237e-06, "loss": 0.0113, "step": 86030 }, { "grad_norm": 0.1965349316596985, "learning_rate": 5.234771009863987e-06, "loss": 0.0102, "step": 86040 }, { "grad_norm": 0.1635173112154007, "learning_rate": 5.227408002058881e-06, "loss": 0.0096, "step": 86050 }, { "grad_norm": 0.5045143365859985, "learning_rate": 5.220049890517126e-06, "loss": 0.0109, "step": 86060 }, { "grad_norm": 0.11826767027378082, "learning_rate": 5.212696676043377e-06, "loss": 0.0126, "step": 86070 }, { "grad_norm": 0.1857862025499344, "learning_rate": 5.205348359441798e-06, "loss": 0.0116, "step": 86080 }, { "grad_norm": 0.19868797063827515, "learning_rate": 5.198004941515966e-06, "loss": 0.0109, "step": 86090 }, { "grad_norm": 0.12608768045902252, "learning_rate": 5.190666423068952e-06, "loss": 0.0119, "step": 86100 }, { "grad_norm": 0.11100286990404129, "learning_rate": 5.183332804903284e-06, "loss": 0.0101, "step": 86110 }, { "grad_norm": 0.2624679207801819, "learning_rate": 5.176004087820979e-06, "loss": 0.011, "step": 86120 }, { "grad_norm": 0.14650298655033112, "learning_rate": 5.168680272623455e-06, "loss": 0.014, "step": 86130 }, { "grad_norm": 0.16788673400878906, "learning_rate": 5.161361360111661e-06, "loss": 0.0135, "step": 86140 }, { "grad_norm": 0.15805712342262268, "learning_rate": 5.154047351085984e-06, "loss": 0.0119, "step": 86150 }, { "grad_norm": 0.10570982843637466, "learning_rate": 5.1467382463462385e-06, "loss": 0.0128, "step": 86160 }, { "grad_norm": 0.17224901914596558, "learning_rate": 5.139434046691788e-06, "loss": 0.0138, "step": 86170 }, { "grad_norm": 0.14822272956371307, "learning_rate": 5.13213475292137e-06, "loss": 0.0097, "step": 86180 }, { "grad_norm": 0.14489516615867615, "learning_rate": 5.12484036583325e-06, "loss": 0.0114, "step": 86190 }, { "grad_norm": 0.16928717494010925, "learning_rate": 5.117550886225109e-06, "loss": 0.0098, "step": 86200 }, { "grad_norm": 0.14088547229766846, "learning_rate": 5.11026631489413e-06, "loss": 0.0095, "step": 86210 }, { "grad_norm": 0.20322027802467346, "learning_rate": 5.102986652636926e-06, "loss": 0.0115, "step": 86220 }, { "grad_norm": 0.128853902220726, "learning_rate": 5.095711900249606e-06, "loss": 0.012, "step": 86230 }, { "grad_norm": 0.1359286904335022, "learning_rate": 5.088442058527715e-06, "loss": 0.0126, "step": 86240 }, { "grad_norm": 0.13119876384735107, "learning_rate": 5.08117712826629e-06, "loss": 0.012, "step": 86250 }, { "grad_norm": 0.13414442539215088, "learning_rate": 5.073917110259785e-06, "loss": 0.0094, "step": 86260 }, { "grad_norm": 0.14636576175689697, "learning_rate": 5.066662005302175e-06, "loss": 0.0099, "step": 86270 }, { "grad_norm": 0.5022321939468384, "learning_rate": 5.059411814186838e-06, "loss": 0.0107, "step": 86280 }, { "grad_norm": 0.16838055849075317, "learning_rate": 5.05216653770666e-06, "loss": 0.0119, "step": 86290 }, { "grad_norm": 0.09995298832654953, "learning_rate": 5.044926176653969e-06, "loss": 0.01, "step": 86300 }, { "grad_norm": 0.2124633491039276, "learning_rate": 5.037690731820571e-06, "loss": 0.0112, "step": 86310 }, { "grad_norm": 0.282868891954422, "learning_rate": 5.030460203997711e-06, "loss": 0.0112, "step": 86320 }, { "grad_norm": 0.2674475312232971, "learning_rate": 5.0232345939761174e-06, "loss": 0.0126, "step": 86330 }, { "grad_norm": 0.18267004191875458, "learning_rate": 5.016013902545957e-06, "loss": 0.0125, "step": 86340 }, { "grad_norm": 0.16097205877304077, "learning_rate": 5.008798130496889e-06, "loss": 0.0116, "step": 86350 }, { "grad_norm": 0.1277635097503662, "learning_rate": 5.001587278618008e-06, "loss": 0.0122, "step": 86360 }, { "grad_norm": 0.24482481181621552, "learning_rate": 4.994381347697891e-06, "loss": 0.0127, "step": 86370 }, { "grad_norm": 0.2756713330745697, "learning_rate": 4.987180338524572e-06, "loss": 0.0116, "step": 86380 }, { "grad_norm": 0.2276688814163208, "learning_rate": 4.9799842518855275e-06, "loss": 0.0117, "step": 86390 }, { "grad_norm": 0.18499888479709625, "learning_rate": 4.9727930885677285e-06, "loss": 0.0104, "step": 86400 }, { "grad_norm": 0.15788257122039795, "learning_rate": 4.965606849357573e-06, "loss": 0.0101, "step": 86410 }, { "grad_norm": 0.15846435725688934, "learning_rate": 4.9584255350409445e-06, "loss": 0.0108, "step": 86420 }, { "grad_norm": 0.17358878254890442, "learning_rate": 4.951249146403175e-06, "loss": 0.0117, "step": 86430 }, { "grad_norm": 0.6199862360954285, "learning_rate": 4.9440776842290815e-06, "loss": 0.0161, "step": 86440 }, { "grad_norm": 0.1308976709842682, "learning_rate": 4.936911149302903e-06, "loss": 0.0099, "step": 86450 }, { "grad_norm": 0.1769086867570877, "learning_rate": 4.929749542408379e-06, "loss": 0.0142, "step": 86460 }, { "grad_norm": 0.26288366317749023, "learning_rate": 4.922592864328673e-06, "loss": 0.0102, "step": 86470 }, { "grad_norm": 0.20616398751735687, "learning_rate": 4.915441115846436e-06, "loss": 0.0134, "step": 86480 }, { "grad_norm": 0.24784104526042938, "learning_rate": 4.908294297743782e-06, "loss": 0.0135, "step": 86490 }, { "grad_norm": 0.146258145570755, "learning_rate": 4.901152410802273e-06, "loss": 0.0117, "step": 86500 }, { "grad_norm": 0.6349354386329651, "learning_rate": 4.894015455802919e-06, "loss": 0.0125, "step": 86510 }, { "grad_norm": 0.10943485051393509, "learning_rate": 4.8868834335262346e-06, "loss": 0.0112, "step": 86520 }, { "grad_norm": 0.12362827360630035, "learning_rate": 4.87975634475214e-06, "loss": 0.01, "step": 86530 }, { "grad_norm": 0.2397545427083969, "learning_rate": 4.8726341902600614e-06, "loss": 0.0114, "step": 86540 }, { "grad_norm": 0.17248369753360748, "learning_rate": 4.865516970828843e-06, "loss": 0.0101, "step": 86550 }, { "grad_norm": 0.1622486263513565, "learning_rate": 4.85840468723685e-06, "loss": 0.0098, "step": 86560 }, { "grad_norm": 0.21946416795253754, "learning_rate": 4.85129734026184e-06, "loss": 0.0124, "step": 86570 }, { "grad_norm": 0.15426874160766602, "learning_rate": 4.844194930681073e-06, "loss": 0.0114, "step": 86580 }, { "grad_norm": 0.1978517770767212, "learning_rate": 4.837097459271267e-06, "loss": 0.0118, "step": 86590 }, { "grad_norm": 0.16053949296474457, "learning_rate": 4.830004926808573e-06, "loss": 0.0107, "step": 86600 }, { "grad_norm": 0.24329151213169098, "learning_rate": 4.822917334068627e-06, "loss": 0.0109, "step": 86610 }, { "grad_norm": 0.16059622168540955, "learning_rate": 4.81583468182652e-06, "loss": 0.0114, "step": 86620 }, { "grad_norm": 0.39742979407310486, "learning_rate": 4.808756970856809e-06, "loss": 0.0106, "step": 86630 }, { "grad_norm": 0.4565373659133911, "learning_rate": 4.801684201933476e-06, "loss": 0.0118, "step": 86640 }, { "grad_norm": 0.22105972468852997, "learning_rate": 4.794616375830019e-06, "loss": 0.0145, "step": 86650 }, { "grad_norm": 0.18120527267456055, "learning_rate": 4.787553493319341e-06, "loss": 0.0128, "step": 86660 }, { "grad_norm": 0.16880302131175995, "learning_rate": 4.780495555173836e-06, "loss": 0.0111, "step": 86670 }, { "grad_norm": 0.2259494811296463, "learning_rate": 4.7734425621653475e-06, "loss": 0.0137, "step": 86680 }, { "grad_norm": 0.17134739458560944, "learning_rate": 4.766394515065198e-06, "loss": 0.0102, "step": 86690 }, { "grad_norm": 0.2448241263628006, "learning_rate": 4.75935141464412e-06, "loss": 0.0134, "step": 86700 }, { "grad_norm": 0.14209303259849548, "learning_rate": 4.75231326167237e-06, "loss": 0.0099, "step": 86710 }, { "grad_norm": 0.19674138724803925, "learning_rate": 4.745280056919599e-06, "loss": 0.0125, "step": 86720 }, { "grad_norm": 0.14284875988960266, "learning_rate": 4.738251801154975e-06, "loss": 0.0113, "step": 86730 }, { "grad_norm": 0.1253056675195694, "learning_rate": 4.731228495147061e-06, "loss": 0.0108, "step": 86740 }, { "grad_norm": 0.1544121503829956, "learning_rate": 4.7242101396639595e-06, "loss": 0.0127, "step": 86750 }, { "grad_norm": 0.13560499250888824, "learning_rate": 4.717196735473151e-06, "loss": 0.0114, "step": 86760 }, { "grad_norm": 0.23004993796348572, "learning_rate": 4.710188283341643e-06, "loss": 0.0137, "step": 86770 }, { "grad_norm": 0.38126271963119507, "learning_rate": 4.703184784035835e-06, "loss": 0.0094, "step": 86780 }, { "grad_norm": 0.18658040463924408, "learning_rate": 4.696186238321642e-06, "loss": 0.0117, "step": 86790 }, { "grad_norm": 0.08768939226865768, "learning_rate": 4.689192646964408e-06, "loss": 0.0099, "step": 86800 }, { "grad_norm": 0.1251397579908371, "learning_rate": 4.682204010728952e-06, "loss": 0.0119, "step": 86810 }, { "grad_norm": 0.17432567477226257, "learning_rate": 4.67522033037952e-06, "loss": 0.0113, "step": 86820 }, { "grad_norm": 0.09774670004844666, "learning_rate": 4.668241606679846e-06, "loss": 0.0131, "step": 86830 }, { "grad_norm": 0.19941796362400055, "learning_rate": 4.66126784039313e-06, "loss": 0.0151, "step": 86840 }, { "grad_norm": 0.12940345704555511, "learning_rate": 4.6542990322819836e-06, "loss": 0.0134, "step": 86850 }, { "grad_norm": 0.11538306623697281, "learning_rate": 4.64733518310852e-06, "loss": 0.0123, "step": 86860 }, { "grad_norm": 0.12705597281455994, "learning_rate": 4.640376293634291e-06, "loss": 0.0096, "step": 86870 }, { "grad_norm": 0.24010194838047028, "learning_rate": 4.633422364620327e-06, "loss": 0.0121, "step": 86880 }, { "grad_norm": 0.18264392018318176, "learning_rate": 4.6264733968270745e-06, "loss": 0.0124, "step": 86890 }, { "grad_norm": 0.2063046395778656, "learning_rate": 4.619529391014482e-06, "loss": 0.011, "step": 86900 }, { "grad_norm": 0.19146867096424103, "learning_rate": 4.61259034794192e-06, "loss": 0.0111, "step": 86910 }, { "grad_norm": 0.22744089365005493, "learning_rate": 4.605656268368247e-06, "loss": 0.0122, "step": 86920 }, { "grad_norm": 0.21074093878269196, "learning_rate": 4.598727153051735e-06, "loss": 0.0113, "step": 86930 }, { "grad_norm": 0.23063774406909943, "learning_rate": 4.591803002750178e-06, "loss": 0.011, "step": 86940 }, { "grad_norm": 0.1638367623090744, "learning_rate": 4.58488381822077e-06, "loss": 0.0108, "step": 86950 }, { "grad_norm": 0.13683080673217773, "learning_rate": 4.5779696002201946e-06, "loss": 0.0132, "step": 86960 }, { "grad_norm": 0.16019491851329803, "learning_rate": 4.571060349504563e-06, "loss": 0.0125, "step": 86970 }, { "grad_norm": 0.21473339200019836, "learning_rate": 4.564156066829478e-06, "loss": 0.0113, "step": 86980 }, { "grad_norm": 0.14364121854305267, "learning_rate": 4.557256752949957e-06, "loss": 0.0117, "step": 86990 }, { "grad_norm": 0.11079297214746475, "learning_rate": 4.550362408620534e-06, "loss": 0.011, "step": 87000 }, { "grad_norm": 0.2758088707923889, "learning_rate": 4.543473034595136e-06, "loss": 0.0132, "step": 87010 }, { "grad_norm": 0.12474994361400604, "learning_rate": 4.536588631627192e-06, "loss": 0.0101, "step": 87020 }, { "grad_norm": 0.17022395133972168, "learning_rate": 4.5297092004695565e-06, "loss": 0.0116, "step": 87030 }, { "grad_norm": 0.18916364014148712, "learning_rate": 4.522834741874555e-06, "loss": 0.0112, "step": 87040 }, { "grad_norm": 0.16208311915397644, "learning_rate": 4.515965256593974e-06, "loss": 0.0112, "step": 87050 }, { "grad_norm": 0.16492614150047302, "learning_rate": 4.509100745379058e-06, "loss": 0.0141, "step": 87060 }, { "grad_norm": 0.15802450478076935, "learning_rate": 4.502241208980479e-06, "loss": 0.0121, "step": 87070 }, { "grad_norm": 0.1793152540922165, "learning_rate": 4.495386648148398e-06, "loss": 0.0108, "step": 87080 }, { "grad_norm": 0.2073564976453781, "learning_rate": 4.488537063632431e-06, "loss": 0.0113, "step": 87090 }, { "grad_norm": 0.15874265134334564, "learning_rate": 4.481692456181608e-06, "loss": 0.0127, "step": 87100 }, { "grad_norm": 0.24621015787124634, "learning_rate": 4.474852826544479e-06, "loss": 0.0122, "step": 87110 }, { "grad_norm": 0.19027020037174225, "learning_rate": 4.468018175468974e-06, "loss": 0.012, "step": 87120 }, { "grad_norm": 0.17537149786949158, "learning_rate": 4.461188503702568e-06, "loss": 0.0111, "step": 87130 }, { "grad_norm": 0.24859404563903809, "learning_rate": 4.454363811992107e-06, "loss": 0.0101, "step": 87140 }, { "grad_norm": 0.1046869307756424, "learning_rate": 4.447544101083956e-06, "loss": 0.0128, "step": 87150 }, { "grad_norm": 0.23694266378879547, "learning_rate": 4.440729371723884e-06, "loss": 0.0107, "step": 87160 }, { "grad_norm": 0.14550629258155823, "learning_rate": 4.433919624657162e-06, "loss": 0.0116, "step": 87170 }, { "grad_norm": 0.2683970630168915, "learning_rate": 4.427114860628462e-06, "loss": 0.0107, "step": 87180 }, { "grad_norm": 0.1303207278251648, "learning_rate": 4.420315080381982e-06, "loss": 0.0146, "step": 87190 }, { "grad_norm": 0.2223242223262787, "learning_rate": 4.4135202846613045e-06, "loss": 0.0115, "step": 87200 }, { "grad_norm": 0.1484600305557251, "learning_rate": 4.406730474209525e-06, "loss": 0.012, "step": 87210 }, { "grad_norm": 0.13135461509227753, "learning_rate": 4.399945649769138e-06, "loss": 0.0089, "step": 87220 }, { "grad_norm": 0.33124372363090515, "learning_rate": 4.393165812082151e-06, "loss": 0.0116, "step": 87230 }, { "grad_norm": 0.1359155774116516, "learning_rate": 4.386390961889969e-06, "loss": 0.0106, "step": 87240 }, { "grad_norm": 0.13936962187290192, "learning_rate": 4.3796210999334895e-06, "loss": 0.0105, "step": 87250 }, { "grad_norm": 0.17868326604366302, "learning_rate": 4.372856226953065e-06, "loss": 0.0086, "step": 87260 }, { "grad_norm": 0.1367407888174057, "learning_rate": 4.366096343688486e-06, "loss": 0.011, "step": 87270 }, { "grad_norm": 0.15903030335903168, "learning_rate": 4.359341450878995e-06, "loss": 0.011, "step": 87280 }, { "grad_norm": 0.20615316927433014, "learning_rate": 4.352591549263302e-06, "loss": 0.0112, "step": 87290 }, { "grad_norm": 0.12041869014501572, "learning_rate": 4.345846639579582e-06, "loss": 0.0114, "step": 87300 }, { "grad_norm": 0.3265320956707001, "learning_rate": 4.339106722565417e-06, "loss": 0.0114, "step": 87310 }, { "grad_norm": 0.11596549302339554, "learning_rate": 4.332371798957896e-06, "loss": 0.0144, "step": 87320 }, { "grad_norm": 0.15169309079647064, "learning_rate": 4.3256418694935295e-06, "loss": 0.0113, "step": 87330 }, { "grad_norm": 0.12678410112857819, "learning_rate": 4.3189169349083125e-06, "loss": 0.012, "step": 87340 }, { "grad_norm": 0.2029755711555481, "learning_rate": 4.312196995937645e-06, "loss": 0.0127, "step": 87350 }, { "grad_norm": 0.1100812554359436, "learning_rate": 4.305482053316434e-06, "loss": 0.0136, "step": 87360 }, { "grad_norm": 0.11591026186943054, "learning_rate": 4.2987721077789925e-06, "loss": 0.0114, "step": 87370 }, { "grad_norm": 0.15892325341701508, "learning_rate": 4.2920671600591325e-06, "loss": 0.0123, "step": 87380 }, { "grad_norm": 0.18490608036518097, "learning_rate": 4.285367210890084e-06, "loss": 0.0121, "step": 87390 }, { "grad_norm": 0.24226495623588562, "learning_rate": 4.278672261004551e-06, "loss": 0.0125, "step": 87400 }, { "grad_norm": 0.10956081002950668, "learning_rate": 4.2719823111346704e-06, "loss": 0.0131, "step": 87410 }, { "grad_norm": 0.18802610039710999, "learning_rate": 4.265297362012061e-06, "loss": 0.012, "step": 87420 }, { "grad_norm": 0.17127157747745514, "learning_rate": 4.258617414367766e-06, "loss": 0.0114, "step": 87430 }, { "grad_norm": 0.1916985660791397, "learning_rate": 4.251942468932296e-06, "loss": 0.0108, "step": 87440 }, { "grad_norm": 0.15365247428417206, "learning_rate": 4.2452725264356175e-06, "loss": 0.0115, "step": 87450 }, { "grad_norm": 0.141136035323143, "learning_rate": 4.2386075876071575e-06, "loss": 0.0116, "step": 87460 }, { "grad_norm": 0.20781798660755157, "learning_rate": 4.231947653175755e-06, "loss": 0.0123, "step": 87470 }, { "grad_norm": 0.18045605719089508, "learning_rate": 4.225292723869761e-06, "loss": 0.0108, "step": 87480 }, { "grad_norm": 0.11694701015949249, "learning_rate": 4.218642800416916e-06, "loss": 0.0107, "step": 87490 }, { "grad_norm": 0.12576618790626526, "learning_rate": 4.211997883544472e-06, "loss": 0.0104, "step": 87500 }, { "grad_norm": 0.159433513879776, "learning_rate": 4.205357973979096e-06, "loss": 0.0123, "step": 87510 }, { "grad_norm": 0.35796186327934265, "learning_rate": 4.198723072446925e-06, "loss": 0.0118, "step": 87520 }, { "grad_norm": 0.12106771767139435, "learning_rate": 4.192093179673534e-06, "loss": 0.0124, "step": 87530 }, { "grad_norm": 0.15851552784442902, "learning_rate": 4.185468296383959e-06, "loss": 0.0124, "step": 87540 }, { "grad_norm": 0.1488678753376007, "learning_rate": 4.1788484233027005e-06, "loss": 0.0103, "step": 87550 }, { "grad_norm": 0.11828455328941345, "learning_rate": 4.172233561153677e-06, "loss": 0.0111, "step": 87560 }, { "grad_norm": 0.28356462717056274, "learning_rate": 4.165623710660288e-06, "loss": 0.0113, "step": 87570 }, { "grad_norm": 0.14278355240821838, "learning_rate": 4.1590188725453835e-06, "loss": 0.0097, "step": 87580 }, { "grad_norm": 0.23500622808933258, "learning_rate": 4.152419047531259e-06, "loss": 0.0111, "step": 87590 }, { "grad_norm": 0.09100687503814697, "learning_rate": 4.145824236339651e-06, "loss": 0.0113, "step": 87600 }, { "grad_norm": 0.1432458758354187, "learning_rate": 4.1392344396917695e-06, "loss": 0.0108, "step": 87610 }, { "grad_norm": 0.15936963260173798, "learning_rate": 4.132649658308252e-06, "loss": 0.0092, "step": 87620 }, { "grad_norm": 0.17485536634922028, "learning_rate": 4.126069892909213e-06, "loss": 0.0116, "step": 87630 }, { "grad_norm": 0.10109944641590118, "learning_rate": 4.119495144214191e-06, "loss": 0.0109, "step": 87640 }, { "grad_norm": 0.2830912172794342, "learning_rate": 4.112925412942214e-06, "loss": 0.0114, "step": 87650 }, { "grad_norm": 0.17201240360736847, "learning_rate": 4.106360699811712e-06, "loss": 0.0146, "step": 87660 }, { "grad_norm": 0.21628683805465698, "learning_rate": 4.099801005540616e-06, "loss": 0.0135, "step": 87670 }, { "grad_norm": 0.23445503413677216, "learning_rate": 4.093246330846256e-06, "loss": 0.0129, "step": 87680 }, { "grad_norm": 0.1711418479681015, "learning_rate": 4.086696676445467e-06, "loss": 0.0128, "step": 87690 }, { "grad_norm": 0.22814372181892395, "learning_rate": 4.0801520430544914e-06, "loss": 0.0098, "step": 87700 }, { "grad_norm": 0.3192621171474457, "learning_rate": 4.073612431389062e-06, "loss": 0.0123, "step": 87710 }, { "grad_norm": 0.1705179661512375, "learning_rate": 4.067077842164319e-06, "loss": 0.0116, "step": 87720 }, { "grad_norm": 0.1446065455675125, "learning_rate": 4.0605482760948945e-06, "loss": 0.011, "step": 87730 }, { "grad_norm": 0.1494424045085907, "learning_rate": 4.054023733894829e-06, "loss": 0.0107, "step": 87740 }, { "grad_norm": 0.3719222843647003, "learning_rate": 4.047504216277653e-06, "loss": 0.0126, "step": 87750 }, { "grad_norm": 0.15823641419410706, "learning_rate": 4.040989723956323e-06, "loss": 0.0132, "step": 87760 }, { "grad_norm": 0.24101482331752777, "learning_rate": 4.03448025764327e-06, "loss": 0.0118, "step": 87770 }, { "grad_norm": 0.13285480439662933, "learning_rate": 4.027975818050339e-06, "loss": 0.0101, "step": 87780 }, { "grad_norm": 0.24505963921546936, "learning_rate": 4.021476405888858e-06, "loss": 0.0137, "step": 87790 }, { "grad_norm": 0.17482295632362366, "learning_rate": 4.01498202186959e-06, "loss": 0.0105, "step": 87800 }, { "grad_norm": 0.17056135833263397, "learning_rate": 4.008492666702751e-06, "loss": 0.0139, "step": 87810 }, { "grad_norm": 0.16089540719985962, "learning_rate": 4.002008341098001e-06, "loss": 0.011, "step": 87820 }, { "grad_norm": 0.1398255079984665, "learning_rate": 3.995529045764462e-06, "loss": 0.0113, "step": 87830 }, { "grad_norm": 0.17841993272304535, "learning_rate": 3.989054781410712e-06, "loss": 0.0099, "step": 87840 }, { "grad_norm": 0.15692056715488434, "learning_rate": 3.982585548744744e-06, "loss": 0.0111, "step": 87850 }, { "grad_norm": 0.42692336440086365, "learning_rate": 3.9761213484740435e-06, "loss": 0.0097, "step": 87860 }, { "grad_norm": 0.18417726457118988, "learning_rate": 3.9696621813055055e-06, "loss": 0.0126, "step": 87870 }, { "grad_norm": 0.15961919724941254, "learning_rate": 3.9632080479455044e-06, "loss": 0.0105, "step": 87880 }, { "grad_norm": 0.12784194946289062, "learning_rate": 3.956758949099859e-06, "loss": 0.0131, "step": 87890 }, { "grad_norm": 0.15695784986019135, "learning_rate": 3.9503148854738385e-06, "loss": 0.0133, "step": 87900 }, { "grad_norm": 0.21734970808029175, "learning_rate": 3.943875857772139e-06, "loss": 0.0131, "step": 87910 }, { "grad_norm": 0.13484761118888855, "learning_rate": 3.9374418666989375e-06, "loss": 0.0096, "step": 87920 }, { "grad_norm": 0.2734479010105133, "learning_rate": 3.931012912957838e-06, "loss": 0.0098, "step": 87930 }, { "grad_norm": 0.12995384633541107, "learning_rate": 3.9245889972519045e-06, "loss": 0.0116, "step": 87940 }, { "grad_norm": 0.2698175013065338, "learning_rate": 3.9181701202836265e-06, "loss": 0.0133, "step": 87950 }, { "grad_norm": 0.21662785112857819, "learning_rate": 3.911756282755003e-06, "loss": 0.0118, "step": 87960 }, { "grad_norm": 0.18126483261585236, "learning_rate": 3.905347485367411e-06, "loss": 0.0111, "step": 87970 }, { "grad_norm": 0.14485570788383484, "learning_rate": 3.898943728821725e-06, "loss": 0.0118, "step": 87980 }, { "grad_norm": 0.13756559789180756, "learning_rate": 3.892545013818227e-06, "loss": 0.0093, "step": 87990 }, { "grad_norm": 0.2002817690372467, "learning_rate": 3.886151341056687e-06, "loss": 0.0143, "step": 88000 }, { "grad_norm": 0.18675071001052856, "learning_rate": 3.8797627112363066e-06, "loss": 0.0104, "step": 88010 }, { "grad_norm": 0.2019711583852768, "learning_rate": 3.873379125055743e-06, "loss": 0.012, "step": 88020 }, { "grad_norm": 0.27754947543144226, "learning_rate": 3.867000583213082e-06, "loss": 0.0128, "step": 88030 }, { "grad_norm": 0.16364702582359314, "learning_rate": 3.860627086405871e-06, "loss": 0.0114, "step": 88040 }, { "grad_norm": 0.1664266288280487, "learning_rate": 3.854258635331126e-06, "loss": 0.011, "step": 88050 }, { "grad_norm": 0.19547998905181885, "learning_rate": 3.8478952306852665e-06, "loss": 0.0104, "step": 88060 }, { "grad_norm": 0.21304963529109955, "learning_rate": 3.841536873164198e-06, "loss": 0.0099, "step": 88070 }, { "grad_norm": 0.16450846195220947, "learning_rate": 3.8351835634632626e-06, "loss": 0.0119, "step": 88080 }, { "grad_norm": 0.2965136766433716, "learning_rate": 3.828835302277245e-06, "loss": 0.0104, "step": 88090 }, { "grad_norm": 0.10395526140928268, "learning_rate": 3.822492090300378e-06, "loss": 0.0103, "step": 88100 }, { "grad_norm": 0.13434606790542603, "learning_rate": 3.8161539282263566e-06, "loss": 0.012, "step": 88110 }, { "grad_norm": 0.08347323536872864, "learning_rate": 3.8098208167482995e-06, "loss": 0.0115, "step": 88120 }, { "grad_norm": 0.17034228146076202, "learning_rate": 3.803492756558796e-06, "loss": 0.0122, "step": 88130 }, { "grad_norm": 0.11976788192987442, "learning_rate": 3.797169748349855e-06, "loss": 0.012, "step": 88140 }, { "grad_norm": 0.14743828773498535, "learning_rate": 3.790851792812977e-06, "loss": 0.0111, "step": 88150 }, { "grad_norm": 0.1629645675420761, "learning_rate": 3.784538890639072e-06, "loss": 0.0103, "step": 88160 }, { "grad_norm": 0.1710948646068573, "learning_rate": 3.7782310425185153e-06, "loss": 0.0101, "step": 88170 }, { "grad_norm": 0.11447707563638687, "learning_rate": 3.77192824914111e-06, "loss": 0.0088, "step": 88180 }, { "grad_norm": 0.18950854241847992, "learning_rate": 3.7656305111961333e-06, "loss": 0.0103, "step": 88190 }, { "grad_norm": 0.13148732483386993, "learning_rate": 3.7593378293722782e-06, "loss": 0.0104, "step": 88200 }, { "grad_norm": 0.2058882713317871, "learning_rate": 3.7530502043577333e-06, "loss": 0.0133, "step": 88210 }, { "grad_norm": 0.365869402885437, "learning_rate": 3.746767636840076e-06, "loss": 0.0116, "step": 88220 }, { "grad_norm": 0.18115220963954926, "learning_rate": 3.74049012750638e-06, "loss": 0.0114, "step": 88230 }, { "grad_norm": 0.27892524003982544, "learning_rate": 3.734217677043128e-06, "loss": 0.0101, "step": 88240 }, { "grad_norm": 0.15424475073814392, "learning_rate": 3.727950286136267e-06, "loss": 0.0123, "step": 88250 }, { "grad_norm": 0.31493356823921204, "learning_rate": 3.7216879554711925e-06, "loss": 0.0095, "step": 88260 }, { "grad_norm": 0.10745108127593994, "learning_rate": 3.715430685732757e-06, "loss": 0.0105, "step": 88270 }, { "grad_norm": 0.2422521561384201, "learning_rate": 3.7091784776052243e-06, "loss": 0.0126, "step": 88280 }, { "grad_norm": 0.15446500480175018, "learning_rate": 3.7029313317723356e-06, "loss": 0.0109, "step": 88290 }, { "grad_norm": 0.20217972993850708, "learning_rate": 3.6966892489172787e-06, "loss": 0.0125, "step": 88300 }, { "grad_norm": 0.16808119416236877, "learning_rate": 3.6904522297226575e-06, "loss": 0.011, "step": 88310 }, { "grad_norm": 0.1418006718158722, "learning_rate": 3.6842202748705647e-06, "loss": 0.012, "step": 88320 }, { "grad_norm": 0.1329929083585739, "learning_rate": 3.677993385042494e-06, "loss": 0.0105, "step": 88330 }, { "grad_norm": 0.1938605159521103, "learning_rate": 3.67177156091944e-06, "loss": 0.0123, "step": 88340 }, { "grad_norm": 0.18348009884357452, "learning_rate": 3.6655548031817866e-06, "loss": 0.0099, "step": 88350 }, { "grad_norm": 0.1266687959432602, "learning_rate": 3.6593431125094e-06, "loss": 0.0117, "step": 88360 }, { "grad_norm": 0.13942137360572815, "learning_rate": 3.6531364895815766e-06, "loss": 0.0134, "step": 88370 }, { "grad_norm": 0.17674492299556732, "learning_rate": 3.6469349350770677e-06, "loss": 0.0098, "step": 88380 }, { "grad_norm": 0.14968828856945038, "learning_rate": 3.6407384496740525e-06, "loss": 0.0131, "step": 88390 }, { "grad_norm": 0.09695156663656235, "learning_rate": 3.6345470340501954e-06, "loss": 0.0123, "step": 88400 }, { "grad_norm": 0.1112690418958664, "learning_rate": 3.6283606888825537e-06, "loss": 0.0111, "step": 88410 }, { "grad_norm": 0.17693917453289032, "learning_rate": 3.622179414847676e-06, "loss": 0.0107, "step": 88420 }, { "grad_norm": 0.3757992088794708, "learning_rate": 3.616003212621527e-06, "loss": 0.0102, "step": 88430 }, { "grad_norm": 0.18328818678855896, "learning_rate": 3.609832082879533e-06, "loss": 0.0101, "step": 88440 }, { "grad_norm": 0.17086903750896454, "learning_rate": 3.6036660262965485e-06, "loss": 0.013, "step": 88450 }, { "grad_norm": 0.16741015017032623, "learning_rate": 3.597505043546895e-06, "loss": 0.0137, "step": 88460 }, { "grad_norm": 0.1513635665178299, "learning_rate": 3.5913491353043227e-06, "loss": 0.0117, "step": 88470 }, { "grad_norm": 0.16152501106262207, "learning_rate": 3.5851983022420375e-06, "loss": 0.0121, "step": 88480 }, { "grad_norm": 0.2523666322231293, "learning_rate": 3.5790525450326797e-06, "loss": 0.0118, "step": 88490 }, { "grad_norm": 0.22408926486968994, "learning_rate": 3.572911864348344e-06, "loss": 0.0103, "step": 88500 }, { "grad_norm": 0.15421141684055328, "learning_rate": 3.5667762608605717e-06, "loss": 0.0107, "step": 88510 }, { "grad_norm": 0.14001117646694183, "learning_rate": 3.5606457352403255e-06, "loss": 0.0122, "step": 88520 }, { "grad_norm": 0.1684308499097824, "learning_rate": 3.554520288158042e-06, "loss": 0.0116, "step": 88530 }, { "grad_norm": 0.20663394033908844, "learning_rate": 3.5483999202835903e-06, "loss": 0.0126, "step": 88540 }, { "grad_norm": 0.19660915434360504, "learning_rate": 3.5422846322862913e-06, "loss": 0.0123, "step": 88550 }, { "grad_norm": 0.16392312943935394, "learning_rate": 3.5361744248348925e-06, "loss": 0.0111, "step": 88560 }, { "grad_norm": 0.27796927094459534, "learning_rate": 3.5300692985976047e-06, "loss": 0.0116, "step": 88570 }, { "grad_norm": 0.15775738656520844, "learning_rate": 3.52396925424206e-06, "loss": 0.0117, "step": 88580 }, { "grad_norm": 0.14464275538921356, "learning_rate": 3.5178742924353747e-06, "loss": 0.0089, "step": 88590 }, { "grad_norm": 0.18026714026927948, "learning_rate": 3.511784413844066e-06, "loss": 0.0108, "step": 88600 }, { "grad_norm": 0.19849957525730133, "learning_rate": 3.5056996191341287e-06, "loss": 0.0127, "step": 88610 }, { "grad_norm": 1.282426357269287, "learning_rate": 3.4996199089709692e-06, "loss": 0.0125, "step": 88620 }, { "grad_norm": 0.2500821053981781, "learning_rate": 3.4935452840194725e-06, "loss": 0.0113, "step": 88630 }, { "grad_norm": 0.45757582783699036, "learning_rate": 3.48747574494393e-06, "loss": 0.0128, "step": 88640 }, { "grad_norm": 0.149797722697258, "learning_rate": 3.4814112924081154e-06, "loss": 0.012, "step": 88650 }, { "grad_norm": 0.2696639597415924, "learning_rate": 3.475351927075221e-06, "loss": 0.0132, "step": 88660 }, { "grad_norm": 0.26436686515808105, "learning_rate": 3.469297649607894e-06, "loss": 0.0131, "step": 88670 }, { "grad_norm": 0.17147600650787354, "learning_rate": 3.463248460668217e-06, "loss": 0.0101, "step": 88680 }, { "grad_norm": 0.1483064591884613, "learning_rate": 3.457204360917726e-06, "loss": 0.0106, "step": 88690 }, { "grad_norm": 0.16879971325397491, "learning_rate": 3.4511653510173827e-06, "loss": 0.0151, "step": 88700 }, { "grad_norm": 0.11203136295080185, "learning_rate": 3.4451314316276074e-06, "loss": 0.0103, "step": 88710 }, { "grad_norm": 0.14907394349575043, "learning_rate": 3.4391026034082674e-06, "loss": 0.0106, "step": 88720 }, { "grad_norm": 0.22335213422775269, "learning_rate": 3.433078867018674e-06, "loss": 0.0104, "step": 88730 }, { "grad_norm": 0.24597202241420746, "learning_rate": 3.427060223117551e-06, "loss": 0.0121, "step": 88740 }, { "grad_norm": 0.1248566135764122, "learning_rate": 3.4210466723631042e-06, "loss": 0.0103, "step": 88750 }, { "grad_norm": 0.20214489102363586, "learning_rate": 3.415038215412969e-06, "loss": 0.0128, "step": 88760 }, { "grad_norm": 0.14704006910324097, "learning_rate": 3.4090348529242088e-06, "loss": 0.0117, "step": 88770 }, { "grad_norm": 0.229715496301651, "learning_rate": 3.4030365855533487e-06, "loss": 0.0126, "step": 88780 }, { "grad_norm": 0.2337024211883545, "learning_rate": 3.397043413956347e-06, "loss": 0.0115, "step": 88790 }, { "grad_norm": 0.1464434713125229, "learning_rate": 3.3910553387886237e-06, "loss": 0.012, "step": 88800 }, { "grad_norm": 0.22160063683986664, "learning_rate": 3.3850723607049996e-06, "loss": 0.0117, "step": 88810 }, { "grad_norm": 0.26637017726898193, "learning_rate": 3.3790944803597902e-06, "loss": 0.0115, "step": 88820 }, { "grad_norm": 0.1770765781402588, "learning_rate": 3.373121698406706e-06, "loss": 0.0109, "step": 88830 }, { "grad_norm": 0.21659615635871887, "learning_rate": 3.3671540154989302e-06, "loss": 0.0124, "step": 88840 }, { "grad_norm": 0.11998385936021805, "learning_rate": 3.361191432289079e-06, "loss": 0.011, "step": 88850 }, { "grad_norm": 0.31764110922813416, "learning_rate": 3.3552339494292207e-06, "loss": 0.0151, "step": 88860 }, { "grad_norm": 0.1680205762386322, "learning_rate": 3.349281567570839e-06, "loss": 0.0114, "step": 88870 }, { "grad_norm": 0.17789091169834137, "learning_rate": 3.3433342873648965e-06, "loss": 0.0099, "step": 88880 }, { "grad_norm": 0.18481218814849854, "learning_rate": 3.337392109461762e-06, "loss": 0.013, "step": 88890 }, { "grad_norm": 0.14385542273521423, "learning_rate": 3.331455034511266e-06, "loss": 0.0121, "step": 88900 }, { "grad_norm": 0.23056864738464355, "learning_rate": 3.325523063162689e-06, "loss": 0.0123, "step": 88910 }, { "grad_norm": 0.49796247482299805, "learning_rate": 3.3195961960647393e-06, "loss": 0.0117, "step": 88920 }, { "grad_norm": 0.17180119454860687, "learning_rate": 3.31367443386556e-06, "loss": 0.0129, "step": 88930 }, { "grad_norm": 0.2723277509212494, "learning_rate": 3.3077577772127556e-06, "loss": 0.0119, "step": 88940 }, { "grad_norm": 0.15971200168132782, "learning_rate": 3.301846226753358e-06, "loss": 0.0125, "step": 88950 }, { "grad_norm": 0.2647460401058197, "learning_rate": 3.2959397831338444e-06, "loss": 0.0121, "step": 88960 }, { "grad_norm": 0.3125098943710327, "learning_rate": 3.2900384470001432e-06, "loss": 0.0106, "step": 88970 }, { "grad_norm": 0.30730631947517395, "learning_rate": 3.28414221899761e-06, "loss": 0.0112, "step": 88980 }, { "grad_norm": 0.1357000470161438, "learning_rate": 3.278251099771046e-06, "loss": 0.0137, "step": 88990 }, { "grad_norm": 0.22690266370773315, "learning_rate": 3.2723650899646906e-06, "loss": 0.0126, "step": 89000 }, { "grad_norm": 0.1211712583899498, "learning_rate": 3.2664841902222466e-06, "loss": 0.0094, "step": 89010 }, { "grad_norm": 0.16403590142726898, "learning_rate": 3.2606084011868208e-06, "loss": 0.0112, "step": 89020 }, { "grad_norm": 0.12715360522270203, "learning_rate": 3.254737723500989e-06, "loss": 0.0105, "step": 89030 }, { "grad_norm": 0.17027373611927032, "learning_rate": 3.2488721578067595e-06, "loss": 0.0104, "step": 89040 }, { "grad_norm": 0.18023629486560822, "learning_rate": 3.243011704745591e-06, "loss": 0.011, "step": 89050 }, { "grad_norm": 0.15613234043121338, "learning_rate": 3.237156364958355e-06, "loss": 0.0105, "step": 89060 }, { "grad_norm": 0.4025839567184448, "learning_rate": 3.231306139085394e-06, "loss": 0.0103, "step": 89070 }, { "grad_norm": 0.22803287208080292, "learning_rate": 3.2254610277664797e-06, "loss": 0.0123, "step": 89080 }, { "grad_norm": 0.1647782325744629, "learning_rate": 3.2196210316408183e-06, "loss": 0.0107, "step": 89090 }, { "grad_norm": 0.13789215683937073, "learning_rate": 3.2137861513470644e-06, "loss": 0.0119, "step": 89100 }, { "grad_norm": 0.19220374524593353, "learning_rate": 3.2079563875233252e-06, "loss": 0.0119, "step": 89110 }, { "grad_norm": 0.1792694330215454, "learning_rate": 3.2021317408071184e-06, "loss": 0.0108, "step": 89120 }, { "grad_norm": 0.19114691019058228, "learning_rate": 3.196312211835434e-06, "loss": 0.0099, "step": 89130 }, { "grad_norm": 0.2336999624967575, "learning_rate": 3.1904978012446686e-06, "loss": 0.013, "step": 89140 }, { "grad_norm": 0.15734697878360748, "learning_rate": 3.184688509670691e-06, "loss": 0.0119, "step": 89150 }, { "grad_norm": 0.3178386986255646, "learning_rate": 3.178884337748783e-06, "loss": 0.0125, "step": 89160 }, { "grad_norm": 0.2125415802001953, "learning_rate": 3.173085286113697e-06, "loss": 0.0115, "step": 89170 }, { "grad_norm": 0.16068542003631592, "learning_rate": 3.1672913553995986e-06, "loss": 0.0117, "step": 89180 }, { "grad_norm": 0.11267394572496414, "learning_rate": 3.1615025462401138e-06, "loss": 0.0107, "step": 89190 }, { "grad_norm": 0.1635546088218689, "learning_rate": 3.1557188592682808e-06, "loss": 0.0119, "step": 89200 }, { "grad_norm": 0.2016916126012802, "learning_rate": 3.1499402951166045e-06, "loss": 0.0109, "step": 89210 }, { "grad_norm": 0.27126839756965637, "learning_rate": 3.144166854417019e-06, "loss": 0.0106, "step": 89220 }, { "grad_norm": 0.16562163829803467, "learning_rate": 3.1383985378009073e-06, "loss": 0.0112, "step": 89230 }, { "grad_norm": 0.20627343654632568, "learning_rate": 3.1326353458990654e-06, "loss": 0.0144, "step": 89240 }, { "grad_norm": 0.12491533905267715, "learning_rate": 3.126877279341761e-06, "loss": 0.0113, "step": 89250 }, { "grad_norm": 0.5457809567451477, "learning_rate": 3.1211243387586964e-06, "loss": 0.0134, "step": 89260 }, { "grad_norm": 0.22664576768875122, "learning_rate": 3.115376524778979e-06, "loss": 0.0132, "step": 89270 }, { "grad_norm": 0.14014358818531036, "learning_rate": 3.109633838031195e-06, "loss": 0.0097, "step": 89280 }, { "grad_norm": 0.22506114840507507, "learning_rate": 3.103896279143359e-06, "loss": 0.0112, "step": 89290 }, { "grad_norm": 0.16382893919944763, "learning_rate": 3.098163848742924e-06, "loss": 0.0111, "step": 89300 }, { "grad_norm": 0.11314330250024796, "learning_rate": 3.0924365474567675e-06, "loss": 0.0122, "step": 89310 }, { "grad_norm": 0.1644136905670166, "learning_rate": 3.086714375911237e-06, "loss": 0.0119, "step": 89320 }, { "grad_norm": 0.1250627636909485, "learning_rate": 3.0809973347320775e-06, "loss": 0.0094, "step": 89330 }, { "grad_norm": 0.1988528072834015, "learning_rate": 3.075285424544516e-06, "loss": 0.0115, "step": 89340 }, { "grad_norm": 0.1210038959980011, "learning_rate": 3.069578645973176e-06, "loss": 0.0121, "step": 89350 }, { "grad_norm": 0.18039213120937347, "learning_rate": 3.0638769996421735e-06, "loss": 0.0138, "step": 89360 }, { "grad_norm": 0.13271689414978027, "learning_rate": 3.058180486175005e-06, "loss": 0.0126, "step": 89370 }, { "grad_norm": 0.11425379663705826, "learning_rate": 3.052489106194645e-06, "loss": 0.01, "step": 89380 }, { "grad_norm": 0.12381310760974884, "learning_rate": 3.0468028603234946e-06, "loss": 0.0116, "step": 89390 }, { "grad_norm": 0.1444387584924698, "learning_rate": 3.04112174918339e-06, "loss": 0.0121, "step": 89400 }, { "grad_norm": 0.246360644698143, "learning_rate": 3.0354457733955954e-06, "loss": 0.0133, "step": 89410 }, { "grad_norm": 0.2470523864030838, "learning_rate": 3.029774933580859e-06, "loss": 0.0121, "step": 89420 }, { "grad_norm": 0.21254682540893555, "learning_rate": 3.0241092303593065e-06, "loss": 0.0105, "step": 89430 }, { "grad_norm": 0.15468916296958923, "learning_rate": 3.018448664350554e-06, "loss": 0.0115, "step": 89440 }, { "grad_norm": 0.15764032304286957, "learning_rate": 3.0127932361736055e-06, "loss": 0.0095, "step": 89450 }, { "grad_norm": 0.3085237443447113, "learning_rate": 3.00714294644695e-06, "loss": 0.0102, "step": 89460 }, { "grad_norm": 0.1443335860967636, "learning_rate": 3.0014977957884828e-06, "loss": 0.012, "step": 89470 }, { "grad_norm": 0.27974167466163635, "learning_rate": 2.99585778481557e-06, "loss": 0.0123, "step": 89480 }, { "grad_norm": 0.19189387559890747, "learning_rate": 2.990222914144969e-06, "loss": 0.0118, "step": 89490 }, { "grad_norm": 0.22727547585964203, "learning_rate": 2.984593184392914e-06, "loss": 0.0119, "step": 89500 }, { "grad_norm": 0.13499867916107178, "learning_rate": 2.978968596175069e-06, "loss": 0.0108, "step": 89510 }, { "grad_norm": 0.31055113673210144, "learning_rate": 2.9733491501065137e-06, "loss": 0.0102, "step": 89520 }, { "grad_norm": 0.10756167769432068, "learning_rate": 2.9677348468018005e-06, "loss": 0.0123, "step": 89530 }, { "grad_norm": 0.16056190431118011, "learning_rate": 2.9621256868748726e-06, "loss": 0.0101, "step": 89540 }, { "grad_norm": 0.16416269540786743, "learning_rate": 2.956521670939183e-06, "loss": 0.0144, "step": 89550 }, { "grad_norm": 0.2577342987060547, "learning_rate": 2.950922799607542e-06, "loss": 0.0103, "step": 89560 }, { "grad_norm": 0.24093103408813477, "learning_rate": 2.9453290734922535e-06, "loss": 0.01, "step": 89570 }, { "grad_norm": 0.4098810851573944, "learning_rate": 2.939740493205029e-06, "loss": 0.0098, "step": 89580 }, { "grad_norm": 0.1716448813676834, "learning_rate": 2.9341570593570345e-06, "loss": 0.0097, "step": 89590 }, { "grad_norm": 0.14775535464286804, "learning_rate": 2.928578772558843e-06, "loss": 0.0106, "step": 89600 }, { "grad_norm": 0.1839624047279358, "learning_rate": 2.923005633420528e-06, "loss": 0.0101, "step": 89610 }, { "grad_norm": 0.13179226219654083, "learning_rate": 2.9174376425515237e-06, "loss": 0.0078, "step": 89620 }, { "grad_norm": 0.2133285105228424, "learning_rate": 2.91187480056076e-06, "loss": 0.0133, "step": 89630 }, { "grad_norm": 0.14078189432621002, "learning_rate": 2.9063171080565664e-06, "loss": 0.0118, "step": 89640 }, { "grad_norm": 0.1744273453950882, "learning_rate": 2.900764565646741e-06, "loss": 0.0117, "step": 89650 }, { "grad_norm": 0.42300158739089966, "learning_rate": 2.895217173938475e-06, "loss": 0.0132, "step": 89660 }, { "grad_norm": 0.18159745633602142, "learning_rate": 2.8896749335384445e-06, "loss": 0.0113, "step": 89670 }, { "grad_norm": 0.1472124308347702, "learning_rate": 2.884137845052731e-06, "loss": 0.01, "step": 89680 }, { "grad_norm": 0.2032047063112259, "learning_rate": 2.8786059090868734e-06, "loss": 0.0118, "step": 89690 }, { "grad_norm": 0.1421128809452057, "learning_rate": 2.87307912624582e-06, "loss": 0.0113, "step": 89700 }, { "grad_norm": 0.15626361966133118, "learning_rate": 2.867557497133977e-06, "loss": 0.0109, "step": 89710 }, { "grad_norm": 0.190145343542099, "learning_rate": 2.862041022355194e-06, "loss": 0.0123, "step": 89720 }, { "grad_norm": 0.7628502249717712, "learning_rate": 2.8565297025127226e-06, "loss": 0.01, "step": 89730 }, { "grad_norm": 0.30270326137542725, "learning_rate": 2.8510235382092853e-06, "loss": 0.0128, "step": 89740 }, { "grad_norm": 0.09361405670642853, "learning_rate": 2.845522530047029e-06, "loss": 0.0112, "step": 89750 }, { "grad_norm": 0.19502119719982147, "learning_rate": 2.8400266786275387e-06, "loss": 0.0131, "step": 89760 }, { "grad_norm": 0.24784713983535767, "learning_rate": 2.834535984551817e-06, "loss": 0.0091, "step": 89770 }, { "grad_norm": 0.13930219411849976, "learning_rate": 2.8290504484203397e-06, "loss": 0.0084, "step": 89780 }, { "grad_norm": 0.11005205661058426, "learning_rate": 2.8235700708329705e-06, "loss": 0.0106, "step": 89790 }, { "grad_norm": 0.1631484180688858, "learning_rate": 2.818094852389058e-06, "loss": 0.0107, "step": 89800 }, { "grad_norm": 0.2924240529537201, "learning_rate": 2.8126247936873516e-06, "loss": 0.0129, "step": 89810 }, { "grad_norm": 0.2332753837108612, "learning_rate": 2.8071598953260614e-06, "loss": 0.0125, "step": 89820 }, { "grad_norm": 0.2458767294883728, "learning_rate": 2.8017001579028035e-06, "loss": 0.0105, "step": 89830 }, { "grad_norm": 0.1254679411649704, "learning_rate": 2.7962455820146617e-06, "loss": 0.0113, "step": 89840 }, { "grad_norm": 0.12829700112342834, "learning_rate": 2.7907961682581253e-06, "loss": 0.0176, "step": 89850 }, { "grad_norm": 0.31206846237182617, "learning_rate": 2.7853519172291453e-06, "loss": 0.0142, "step": 89860 }, { "grad_norm": 0.31772375106811523, "learning_rate": 2.7799128295230904e-06, "loss": 0.012, "step": 89870 }, { "grad_norm": 0.1949170082807541, "learning_rate": 2.774478905734784e-06, "loss": 0.0093, "step": 89880 }, { "grad_norm": 0.29878872632980347, "learning_rate": 2.7690501464584516e-06, "loss": 0.0108, "step": 89890 }, { "grad_norm": 0.14911289513111115, "learning_rate": 2.763626552287796e-06, "loss": 0.011, "step": 89900 }, { "grad_norm": 0.16724923253059387, "learning_rate": 2.7582081238159083e-06, "loss": 0.0099, "step": 89910 }, { "grad_norm": 0.10518430173397064, "learning_rate": 2.7527948616353548e-06, "loss": 0.012, "step": 89920 }, { "grad_norm": 0.18006527423858643, "learning_rate": 2.7473867663381226e-06, "loss": 0.0098, "step": 89930 }, { "grad_norm": 0.17693842947483063, "learning_rate": 2.7419838385156383e-06, "loss": 0.0109, "step": 89940 }, { "grad_norm": 0.1322336196899414, "learning_rate": 2.7365860787587407e-06, "loss": 0.0115, "step": 89950 }, { "grad_norm": 0.14944905042648315, "learning_rate": 2.7311934876577295e-06, "loss": 0.0111, "step": 89960 }, { "grad_norm": 0.27128544449806213, "learning_rate": 2.725806065802339e-06, "loss": 0.0138, "step": 89970 }, { "grad_norm": 0.18135473132133484, "learning_rate": 2.720423813781714e-06, "loss": 0.0105, "step": 89980 }, { "grad_norm": 0.3603149354457855, "learning_rate": 2.715046732184462e-06, "loss": 0.0125, "step": 89990 }, { "grad_norm": 0.13765022158622742, "learning_rate": 2.7096748215986013e-06, "loss": 0.0114, "step": 90000 } ], "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": 24, "trial_name": null, "trial_params": null }