{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 30000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.9198344945907593, "learning_rate": 6.000000000000001e-07, "loss": 1.2618, "step": 10 }, { "grad_norm": 1.5244935750961304, "learning_rate": 1.2666666666666667e-06, "loss": 1.2575, "step": 20 }, { "grad_norm": 1.205666422843933, "learning_rate": 1.9333333333333336e-06, "loss": 1.2703, "step": 30 }, { "grad_norm": 1.2688565254211426, "learning_rate": 2.6e-06, "loss": 1.2461, "step": 40 }, { "grad_norm": 0.6094927787780762, "learning_rate": 3.2666666666666666e-06, "loss": 1.2189, "step": 50 }, { "grad_norm": 0.5283153653144836, "learning_rate": 3.9333333333333335e-06, "loss": 1.1577, "step": 60 }, { "grad_norm": 0.6232669353485107, "learning_rate": 4.6e-06, "loss": 1.1433, "step": 70 }, { "grad_norm": 0.7596688270568848, "learning_rate": 5.266666666666667e-06, "loss": 1.1474, "step": 80 }, { "grad_norm": 1.1742751598358154, "learning_rate": 5.933333333333334e-06, "loss": 1.1305, "step": 90 }, { "grad_norm": 0.4896790087223053, "learning_rate": 6.6e-06, "loss": 1.1268, "step": 100 }, { "grad_norm": 0.5746106505393982, "learning_rate": 7.266666666666668e-06, "loss": 1.1533, "step": 110 }, { "grad_norm": 0.6397786736488342, "learning_rate": 7.933333333333334e-06, "loss": 1.1325, "step": 120 }, { "grad_norm": 0.5786457061767578, "learning_rate": 8.599999999999999e-06, "loss": 1.1155, "step": 130 }, { "grad_norm": 0.5029047727584839, "learning_rate": 9.266666666666667e-06, "loss": 1.0976, "step": 140 }, { "grad_norm": 0.7537752985954285, "learning_rate": 9.933333333333334e-06, "loss": 1.1153, "step": 150 }, { "grad_norm": 0.6609781384468079, "learning_rate": 1.06e-05, "loss": 1.0853, "step": 160 }, { "grad_norm": 0.685785710811615, "learning_rate": 1.1266666666666667e-05, "loss": 1.1122, "step": 170 }, { "grad_norm": 0.5571264624595642, "learning_rate": 1.1933333333333333e-05, "loss": 1.0723, "step": 180 }, { "grad_norm": 0.9104903340339661, "learning_rate": 1.2600000000000001e-05, "loss": 1.0973, "step": 190 }, { "grad_norm": 0.757935106754303, "learning_rate": 1.3266666666666666e-05, "loss": 1.079, "step": 200 }, { "grad_norm": 0.7954821586608887, "learning_rate": 1.3933333333333334e-05, "loss": 1.0966, "step": 210 }, { "grad_norm": 0.6015939712524414, "learning_rate": 1.4599999999999999e-05, "loss": 1.0805, "step": 220 }, { "grad_norm": 1.0411932468414307, "learning_rate": 1.5266666666666667e-05, "loss": 1.0873, "step": 230 }, { "grad_norm": 0.533362627029419, "learning_rate": 1.5933333333333332e-05, "loss": 1.0715, "step": 240 }, { "grad_norm": 0.9060723185539246, "learning_rate": 1.66e-05, "loss": 1.081, "step": 250 }, { "grad_norm": 0.5110937356948853, "learning_rate": 1.726666666666667e-05, "loss": 1.0785, "step": 260 }, { "grad_norm": 0.7501968145370483, "learning_rate": 1.7933333333333337e-05, "loss": 1.0765, "step": 270 }, { "grad_norm": 0.7590754628181458, "learning_rate": 1.86e-05, "loss": 1.0679, "step": 280 }, { "grad_norm": 1.1224004030227661, "learning_rate": 1.926666666666667e-05, "loss": 1.0681, "step": 290 }, { "grad_norm": 1.1518421173095703, "learning_rate": 1.9933333333333334e-05, "loss": 1.0501, "step": 300 }, { "grad_norm": 0.9900417327880859, "learning_rate": 2.06e-05, "loss": 1.0461, "step": 310 }, { "grad_norm": 0.8574306964874268, "learning_rate": 2.1266666666666667e-05, "loss": 1.0458, "step": 320 }, { "grad_norm": 0.7958481907844543, "learning_rate": 2.1933333333333332e-05, "loss": 1.0411, "step": 330 }, { "grad_norm": 0.963300347328186, "learning_rate": 2.26e-05, "loss": 1.0268, "step": 340 }, { "grad_norm": 0.7371602654457092, "learning_rate": 2.326666666666667e-05, "loss": 1.0272, "step": 350 }, { "grad_norm": 0.7280622124671936, "learning_rate": 2.3933333333333337e-05, "loss": 1.0197, "step": 360 }, { "grad_norm": 1.0074427127838135, "learning_rate": 2.46e-05, "loss": 1.013, "step": 370 }, { "grad_norm": 1.0045920610427856, "learning_rate": 2.5266666666666666e-05, "loss": 0.963, "step": 380 }, { "grad_norm": 1.4531843662261963, "learning_rate": 2.5933333333333338e-05, "loss": 0.949, "step": 390 }, { "grad_norm": 1.1588211059570312, "learning_rate": 2.6600000000000003e-05, "loss": 0.9221, "step": 400 }, { "grad_norm": 1.196110725402832, "learning_rate": 2.7266666666666668e-05, "loss": 0.8815, "step": 410 }, { "grad_norm": 1.1931462287902832, "learning_rate": 2.7933333333333332e-05, "loss": 0.8447, "step": 420 }, { "grad_norm": 1.2790521383285522, "learning_rate": 2.86e-05, "loss": 0.834, "step": 430 }, { "grad_norm": 1.1627215147018433, "learning_rate": 2.926666666666667e-05, "loss": 0.7825, "step": 440 }, { "grad_norm": 1.519866704940796, "learning_rate": 2.9933333333333337e-05, "loss": 0.746, "step": 450 }, { "grad_norm": 1.3904961347579956, "learning_rate": 3.06e-05, "loss": 0.7228, "step": 460 }, { "grad_norm": 1.5789117813110352, "learning_rate": 3.126666666666666e-05, "loss": 0.7082, "step": 470 }, { "grad_norm": 1.5974048376083374, "learning_rate": 3.1933333333333335e-05, "loss": 0.6773, "step": 480 }, { "grad_norm": 1.7163608074188232, "learning_rate": 3.26e-05, "loss": 0.641, "step": 490 }, { "grad_norm": 1.6014769077301025, "learning_rate": 3.326666666666667e-05, "loss": 0.5794, "step": 500 }, { "grad_norm": 1.4093105792999268, "learning_rate": 3.3933333333333336e-05, "loss": 0.5711, "step": 510 }, { "grad_norm": 1.2497873306274414, "learning_rate": 3.46e-05, "loss": 0.5459, "step": 520 }, { "grad_norm": 1.6634900569915771, "learning_rate": 3.526666666666667e-05, "loss": 0.5209, "step": 530 }, { "grad_norm": 1.212041974067688, "learning_rate": 3.593333333333334e-05, "loss": 0.4689, "step": 540 }, { "grad_norm": 1.768399715423584, "learning_rate": 3.66e-05, "loss": 0.4704, "step": 550 }, { "grad_norm": 1.5253715515136719, "learning_rate": 3.726666666666667e-05, "loss": 0.4198, "step": 560 }, { "grad_norm": 1.1857234239578247, "learning_rate": 3.793333333333334e-05, "loss": 0.368, "step": 570 }, { "grad_norm": 1.259238600730896, "learning_rate": 3.86e-05, "loss": 0.3468, "step": 580 }, { "grad_norm": 1.6979483366012573, "learning_rate": 3.926666666666667e-05, "loss": 0.3354, "step": 590 }, { "grad_norm": 1.4338759183883667, "learning_rate": 3.993333333333333e-05, "loss": 0.2984, "step": 600 }, { "grad_norm": 1.6504828929901123, "learning_rate": 4.0600000000000004e-05, "loss": 0.2781, "step": 610 }, { "grad_norm": 1.6862783432006836, "learning_rate": 4.126666666666667e-05, "loss": 0.2676, "step": 620 }, { "grad_norm": 1.396815538406372, "learning_rate": 4.1933333333333334e-05, "loss": 0.2685, "step": 630 }, { "grad_norm": 1.3235665559768677, "learning_rate": 4.26e-05, "loss": 0.231, "step": 640 }, { "grad_norm": 1.5948623418807983, "learning_rate": 4.3266666666666664e-05, "loss": 0.2023, "step": 650 }, { "grad_norm": 1.3301814794540405, "learning_rate": 4.3933333333333335e-05, "loss": 0.2315, "step": 660 }, { "grad_norm": 1.280762791633606, "learning_rate": 4.46e-05, "loss": 0.1898, "step": 670 }, { "grad_norm": 1.4321988821029663, "learning_rate": 4.526666666666667e-05, "loss": 0.2063, "step": 680 }, { "grad_norm": 1.4155633449554443, "learning_rate": 4.5933333333333336e-05, "loss": 0.2051, "step": 690 }, { "grad_norm": 1.466268539428711, "learning_rate": 4.660000000000001e-05, "loss": 0.2155, "step": 700 }, { "grad_norm": 1.1416596174240112, "learning_rate": 4.726666666666667e-05, "loss": 0.1925, "step": 710 }, { "grad_norm": 1.2276180982589722, "learning_rate": 4.793333333333334e-05, "loss": 0.2047, "step": 720 }, { "grad_norm": 1.600016713142395, "learning_rate": 4.86e-05, "loss": 0.1996, "step": 730 }, { "grad_norm": 1.258832573890686, "learning_rate": 4.926666666666667e-05, "loss": 0.1699, "step": 740 }, { "grad_norm": 1.147660255432129, "learning_rate": 4.993333333333334e-05, "loss": 0.2025, "step": 750 }, { "grad_norm": 0.9750813245773315, "learning_rate": 5.0600000000000003e-05, "loss": 0.1952, "step": 760 }, { "grad_norm": 0.8782268166542053, "learning_rate": 5.1266666666666675e-05, "loss": 0.1945, "step": 770 }, { "grad_norm": 1.0923429727554321, "learning_rate": 5.193333333333333e-05, "loss": 0.2425, "step": 780 }, { "grad_norm": 0.9937282800674438, "learning_rate": 5.2600000000000005e-05, "loss": 0.1858, "step": 790 }, { "grad_norm": 1.117896556854248, "learning_rate": 5.326666666666666e-05, "loss": 0.177, "step": 800 }, { "grad_norm": 1.2316797971725464, "learning_rate": 5.3933333333333334e-05, "loss": 0.1792, "step": 810 }, { "grad_norm": 0.9455136656761169, "learning_rate": 5.4600000000000006e-05, "loss": 0.1919, "step": 820 }, { "grad_norm": 1.0609848499298096, "learning_rate": 5.5266666666666664e-05, "loss": 0.203, "step": 830 }, { "grad_norm": 0.8871763944625854, "learning_rate": 5.5933333333333335e-05, "loss": 0.1954, "step": 840 }, { "grad_norm": 1.0370177030563354, "learning_rate": 5.66e-05, "loss": 0.2053, "step": 850 }, { "grad_norm": 1.0110114812850952, "learning_rate": 5.726666666666667e-05, "loss": 0.1955, "step": 860 }, { "grad_norm": 1.023281216621399, "learning_rate": 5.7933333333333337e-05, "loss": 0.2003, "step": 870 }, { "grad_norm": 1.0264396667480469, "learning_rate": 5.86e-05, "loss": 0.1684, "step": 880 }, { "grad_norm": 1.0403268337249756, "learning_rate": 5.926666666666667e-05, "loss": 0.1879, "step": 890 }, { "grad_norm": 0.8513246774673462, "learning_rate": 5.9933333333333345e-05, "loss": 0.1844, "step": 900 }, { "grad_norm": 0.9342168569564819, "learning_rate": 6.06e-05, "loss": 0.1789, "step": 910 }, { "grad_norm": 0.9911207556724548, "learning_rate": 6.126666666666667e-05, "loss": 0.1681, "step": 920 }, { "grad_norm": 0.9268270134925842, "learning_rate": 6.193333333333333e-05, "loss": 0.1789, "step": 930 }, { "grad_norm": 0.8090454339981079, "learning_rate": 6.26e-05, "loss": 0.1739, "step": 940 }, { "grad_norm": 0.7379406094551086, "learning_rate": 6.326666666666667e-05, "loss": 0.1636, "step": 950 }, { "grad_norm": 0.8934803009033203, "learning_rate": 6.393333333333333e-05, "loss": 0.1639, "step": 960 }, { "grad_norm": 0.7845222353935242, "learning_rate": 6.460000000000001e-05, "loss": 0.1918, "step": 970 }, { "grad_norm": 0.7984896302223206, "learning_rate": 6.526666666666666e-05, "loss": 0.1587, "step": 980 }, { "grad_norm": 0.9760947227478027, "learning_rate": 6.593333333333334e-05, "loss": 0.1589, "step": 990 }, { "grad_norm": 0.8190470933914185, "learning_rate": 6.66e-05, "loss": 0.167, "step": 1000 }, { "grad_norm": 0.7485285997390747, "learning_rate": 6.726666666666667e-05, "loss": 0.1588, "step": 1010 }, { "grad_norm": 0.8362904191017151, "learning_rate": 6.793333333333334e-05, "loss": 0.1783, "step": 1020 }, { "grad_norm": 0.9342178702354431, "learning_rate": 6.860000000000001e-05, "loss": 0.1484, "step": 1030 }, { "grad_norm": 0.6118074655532837, "learning_rate": 6.926666666666667e-05, "loss": 0.1561, "step": 1040 }, { "grad_norm": 1.0551546812057495, "learning_rate": 6.993333333333334e-05, "loss": 0.1629, "step": 1050 }, { "grad_norm": 0.6967196464538574, "learning_rate": 7.06e-05, "loss": 0.18, "step": 1060 }, { "grad_norm": 0.806589663028717, "learning_rate": 7.126666666666667e-05, "loss": 0.1846, "step": 1070 }, { "grad_norm": 0.980701744556427, "learning_rate": 7.193333333333334e-05, "loss": 0.1723, "step": 1080 }, { "grad_norm": 0.9944673776626587, "learning_rate": 7.26e-05, "loss": 0.1697, "step": 1090 }, { "grad_norm": 0.7328429222106934, "learning_rate": 7.326666666666667e-05, "loss": 0.1371, "step": 1100 }, { "grad_norm": 1.057700753211975, "learning_rate": 7.393333333333333e-05, "loss": 0.1827, "step": 1110 }, { "grad_norm": 0.6237815618515015, "learning_rate": 7.46e-05, "loss": 0.132, "step": 1120 }, { "grad_norm": 0.9634680151939392, "learning_rate": 7.526666666666668e-05, "loss": 0.1494, "step": 1130 }, { "grad_norm": 0.8337220549583435, "learning_rate": 7.593333333333334e-05, "loss": 0.1877, "step": 1140 }, { "grad_norm": 0.8768863677978516, "learning_rate": 7.66e-05, "loss": 0.1764, "step": 1150 }, { "grad_norm": 0.7452356219291687, "learning_rate": 7.726666666666667e-05, "loss": 0.1462, "step": 1160 }, { "grad_norm": 0.755405604839325, "learning_rate": 7.793333333333333e-05, "loss": 0.144, "step": 1170 }, { "grad_norm": 0.6388455629348755, "learning_rate": 7.860000000000001e-05, "loss": 0.1175, "step": 1180 }, { "grad_norm": 0.7085903882980347, "learning_rate": 7.926666666666666e-05, "loss": 0.1457, "step": 1190 }, { "grad_norm": 1.1641480922698975, "learning_rate": 7.993333333333334e-05, "loss": 0.1447, "step": 1200 }, { "grad_norm": 1.1087267398834229, "learning_rate": 8.060000000000001e-05, "loss": 0.1583, "step": 1210 }, { "grad_norm": 0.9150151014328003, "learning_rate": 8.126666666666667e-05, "loss": 0.1473, "step": 1220 }, { "grad_norm": 0.6998509168624878, "learning_rate": 8.193333333333334e-05, "loss": 0.1396, "step": 1230 }, { "grad_norm": 0.6994796991348267, "learning_rate": 8.26e-05, "loss": 0.1476, "step": 1240 }, { "grad_norm": 0.9219644069671631, "learning_rate": 8.326666666666667e-05, "loss": 0.1231, "step": 1250 }, { "grad_norm": 0.7757239937782288, "learning_rate": 8.393333333333335e-05, "loss": 0.134, "step": 1260 }, { "grad_norm": 0.6924928426742554, "learning_rate": 8.46e-05, "loss": 0.1087, "step": 1270 }, { "grad_norm": 0.5475831627845764, "learning_rate": 8.526666666666667e-05, "loss": 0.1293, "step": 1280 }, { "grad_norm": 0.564661979675293, "learning_rate": 8.593333333333333e-05, "loss": 0.1412, "step": 1290 }, { "grad_norm": 0.6156620383262634, "learning_rate": 8.66e-05, "loss": 0.1317, "step": 1300 }, { "grad_norm": 0.6949765086174011, "learning_rate": 8.726666666666667e-05, "loss": 0.1342, "step": 1310 }, { "grad_norm": 0.5939993262290955, "learning_rate": 8.793333333333333e-05, "loss": 0.1243, "step": 1320 }, { "grad_norm": 0.6796340346336365, "learning_rate": 8.86e-05, "loss": 0.1279, "step": 1330 }, { "grad_norm": 0.549228847026825, "learning_rate": 8.926666666666668e-05, "loss": 0.1115, "step": 1340 }, { "grad_norm": 0.6801861524581909, "learning_rate": 8.993333333333334e-05, "loss": 0.1499, "step": 1350 }, { "grad_norm": 0.7394158840179443, "learning_rate": 9.06e-05, "loss": 0.1362, "step": 1360 }, { "grad_norm": 0.7071419358253479, "learning_rate": 9.126666666666667e-05, "loss": 0.1178, "step": 1370 }, { "grad_norm": 0.7034131288528442, "learning_rate": 9.193333333333334e-05, "loss": 0.1115, "step": 1380 }, { "grad_norm": 0.8657644391059875, "learning_rate": 9.260000000000001e-05, "loss": 0.1256, "step": 1390 }, { "grad_norm": 0.6930060386657715, "learning_rate": 9.326666666666667e-05, "loss": 0.1092, "step": 1400 }, { "grad_norm": 0.6042575836181641, "learning_rate": 9.393333333333334e-05, "loss": 0.1259, "step": 1410 }, { "grad_norm": 0.595906674861908, "learning_rate": 9.46e-05, "loss": 0.1291, "step": 1420 }, { "grad_norm": 0.5124180316925049, "learning_rate": 9.526666666666667e-05, "loss": 0.1013, "step": 1430 }, { "grad_norm": 0.6364486813545227, "learning_rate": 9.593333333333334e-05, "loss": 0.1245, "step": 1440 }, { "grad_norm": 0.5574372410774231, "learning_rate": 9.66e-05, "loss": 0.1304, "step": 1450 }, { "grad_norm": 0.7704989314079285, "learning_rate": 9.726666666666667e-05, "loss": 0.1342, "step": 1460 }, { "grad_norm": 0.5697133541107178, "learning_rate": 9.793333333333333e-05, "loss": 0.1372, "step": 1470 }, { "grad_norm": 0.7197838425636292, "learning_rate": 9.86e-05, "loss": 0.1192, "step": 1480 }, { "grad_norm": 0.801982581615448, "learning_rate": 9.926666666666668e-05, "loss": 0.1171, "step": 1490 }, { "grad_norm": 0.6211628913879395, "learning_rate": 9.993333333333334e-05, "loss": 0.1208, "step": 1500 }, { "grad_norm": 0.5781983733177185, "learning_rate": 9.999997539434007e-05, "loss": 0.1055, "step": 1510 }, { "grad_norm": 0.6830707788467407, "learning_rate": 9.999989033776898e-05, "loss": 0.1276, "step": 1520 }, { "grad_norm": 0.5188012719154358, "learning_rate": 9.999974452661641e-05, "loss": 0.1238, "step": 1530 }, { "grad_norm": 0.6128641963005066, "learning_rate": 9.999953796105959e-05, "loss": 0.112, "step": 1540 }, { "grad_norm": 0.4158335328102112, "learning_rate": 9.999927064134949e-05, "loss": 0.1212, "step": 1550 }, { "grad_norm": 0.450379341840744, "learning_rate": 9.999894256781095e-05, "loss": 0.1341, "step": 1560 }, { "grad_norm": 0.604142963886261, "learning_rate": 9.99985537408426e-05, "loss": 0.1161, "step": 1570 }, { "grad_norm": 0.5382503271102905, "learning_rate": 9.999810416091688e-05, "loss": 0.105, "step": 1580 }, { "grad_norm": 0.6099220514297485, "learning_rate": 9.99975938285801e-05, "loss": 0.1242, "step": 1590 }, { "grad_norm": 0.5786386728286743, "learning_rate": 9.999702274445236e-05, "loss": 0.1088, "step": 1600 }, { "grad_norm": 0.40269529819488525, "learning_rate": 9.999639090922756e-05, "loss": 0.1103, "step": 1610 }, { "grad_norm": 0.5732976794242859, "learning_rate": 9.999569832367346e-05, "loss": 0.126, "step": 1620 }, { "grad_norm": 0.5805198550224304, "learning_rate": 9.999494498863162e-05, "loss": 0.1551, "step": 1630 }, { "grad_norm": 0.5416373014450073, "learning_rate": 9.99941309050174e-05, "loss": 0.1409, "step": 1640 }, { "grad_norm": 0.5198296904563904, "learning_rate": 9.999325607381999e-05, "loss": 0.1155, "step": 1650 }, { "grad_norm": 0.5030234456062317, "learning_rate": 9.999232049610238e-05, "loss": 0.1337, "step": 1660 }, { "grad_norm": 0.6114234328269958, "learning_rate": 9.999132417300142e-05, "loss": 0.1093, "step": 1670 }, { "grad_norm": 0.6113016605377197, "learning_rate": 9.99902671057277e-05, "loss": 0.1111, "step": 1680 }, { "grad_norm": 0.5011656284332275, "learning_rate": 9.998914929556569e-05, "loss": 0.1177, "step": 1690 }, { "grad_norm": 0.445525586605072, "learning_rate": 9.998797074387361e-05, "loss": 0.1122, "step": 1700 }, { "grad_norm": 0.5823173522949219, "learning_rate": 9.99867314520835e-05, "loss": 0.1212, "step": 1710 }, { "grad_norm": 0.4142376482486725, "learning_rate": 9.998543142170126e-05, "loss": 0.1052, "step": 1720 }, { "grad_norm": 0.444185346364975, "learning_rate": 9.99840706543065e-05, "loss": 0.1138, "step": 1730 }, { "grad_norm": 0.7114488482475281, "learning_rate": 9.998264915155274e-05, "loss": 0.104, "step": 1740 }, { "grad_norm": 0.456831693649292, "learning_rate": 9.998116691516718e-05, "loss": 0.1245, "step": 1750 }, { "grad_norm": 0.6638155579566956, "learning_rate": 9.997962394695091e-05, "loss": 0.1195, "step": 1760 }, { "grad_norm": 0.4662315547466278, "learning_rate": 9.997802024877875e-05, "loss": 0.1247, "step": 1770 }, { "grad_norm": 0.5707429051399231, "learning_rate": 9.99763558225994e-05, "loss": 0.151, "step": 1780 }, { "grad_norm": 0.746601402759552, "learning_rate": 9.997463067043526e-05, "loss": 0.1361, "step": 1790 }, { "grad_norm": 0.727504551410675, "learning_rate": 9.997284479438253e-05, "loss": 0.1435, "step": 1800 }, { "grad_norm": 0.5973384380340576, "learning_rate": 9.997099819661127e-05, "loss": 0.1312, "step": 1810 }, { "grad_norm": 0.6944571137428284, "learning_rate": 9.996909087936524e-05, "loss": 0.1443, "step": 1820 }, { "grad_norm": 0.5821080803871155, "learning_rate": 9.996712284496201e-05, "loss": 0.1434, "step": 1830 }, { "grad_norm": 0.6013721227645874, "learning_rate": 9.996509409579293e-05, "loss": 0.1513, "step": 1840 }, { "grad_norm": 0.3996436297893524, "learning_rate": 9.996300463432312e-05, "loss": 0.1633, "step": 1850 }, { "grad_norm": 0.6903651356697083, "learning_rate": 9.996085446309148e-05, "loss": 0.1798, "step": 1860 }, { "grad_norm": 0.516541600227356, "learning_rate": 9.995864358471066e-05, "loss": 0.1419, "step": 1870 }, { "grad_norm": 0.6729509830474854, "learning_rate": 9.99563720018671e-05, "loss": 0.1292, "step": 1880 }, { "grad_norm": 0.6056923866271973, "learning_rate": 9.995403971732098e-05, "loss": 0.1321, "step": 1890 }, { "grad_norm": 0.5741896033287048, "learning_rate": 9.995164673390625e-05, "loss": 0.1386, "step": 1900 }, { "grad_norm": 0.663706362247467, "learning_rate": 9.994919305453059e-05, "loss": 0.166, "step": 1910 }, { "grad_norm": 0.6009074449539185, "learning_rate": 9.994667868217548e-05, "loss": 0.1373, "step": 1920 }, { "grad_norm": 0.6618272066116333, "learning_rate": 9.99441036198961e-05, "loss": 0.1273, "step": 1930 }, { "grad_norm": 0.5536155700683594, "learning_rate": 9.99414678708214e-05, "loss": 0.1396, "step": 1940 }, { "grad_norm": 0.7006198763847351, "learning_rate": 9.993877143815407e-05, "loss": 0.1246, "step": 1950 }, { "grad_norm": 0.6220900416374207, "learning_rate": 9.993601432517053e-05, "loss": 0.1556, "step": 1960 }, { "grad_norm": 0.48185935616493225, "learning_rate": 9.993319653522091e-05, "loss": 0.109, "step": 1970 }, { "grad_norm": 0.6907294988632202, "learning_rate": 9.993031807172911e-05, "loss": 0.1474, "step": 1980 }, { "grad_norm": 0.5758239030838013, "learning_rate": 9.992737893819273e-05, "loss": 0.1182, "step": 1990 }, { "grad_norm": 0.5130513906478882, "learning_rate": 9.992437913818312e-05, "loss": 0.1315, "step": 2000 }, { "grad_norm": 0.593991756439209, "learning_rate": 9.992131867534526e-05, "loss": 0.1592, "step": 2010 }, { "grad_norm": 0.5710794925689697, "learning_rate": 9.991819755339796e-05, "loss": 0.1317, "step": 2020 }, { "grad_norm": 0.5791780352592468, "learning_rate": 9.991501577613365e-05, "loss": 0.1148, "step": 2030 }, { "grad_norm": 0.4964178800582886, "learning_rate": 9.99117733474185e-05, "loss": 0.14, "step": 2040 }, { "grad_norm": 0.6173135042190552, "learning_rate": 9.990847027119234e-05, "loss": 0.1419, "step": 2050 }, { "grad_norm": 0.5468834638595581, "learning_rate": 9.990510655146877e-05, "loss": 0.1194, "step": 2060 }, { "grad_norm": 0.5716961622238159, "learning_rate": 9.990168219233496e-05, "loss": 0.1113, "step": 2070 }, { "grad_norm": 0.5471460819244385, "learning_rate": 9.989819719795188e-05, "loss": 0.126, "step": 2080 }, { "grad_norm": 0.4512519836425781, "learning_rate": 9.989465157255412e-05, "loss": 0.1231, "step": 2090 }, { "grad_norm": 0.514064371585846, "learning_rate": 9.989104532044994e-05, "loss": 0.1232, "step": 2100 }, { "grad_norm": 0.5343894362449646, "learning_rate": 9.988737844602128e-05, "loss": 0.1534, "step": 2110 }, { "grad_norm": 0.4953218102455139, "learning_rate": 9.988365095372372e-05, "loss": 0.1205, "step": 2120 }, { "grad_norm": 0.5329728126525879, "learning_rate": 9.987986284808654e-05, "loss": 0.1225, "step": 2130 }, { "grad_norm": 0.42026662826538086, "learning_rate": 9.987601413371264e-05, "loss": 0.1036, "step": 2140 }, { "grad_norm": 0.42160019278526306, "learning_rate": 9.987210481527855e-05, "loss": 0.1212, "step": 2150 }, { "grad_norm": 0.5244167447090149, "learning_rate": 9.98681348975345e-05, "loss": 0.1153, "step": 2160 }, { "grad_norm": 0.5770167112350464, "learning_rate": 9.986410438530427e-05, "loss": 0.1005, "step": 2170 }, { "grad_norm": 0.36518263816833496, "learning_rate": 9.986001328348534e-05, "loss": 0.1179, "step": 2180 }, { "grad_norm": 0.4710124731063843, "learning_rate": 9.985586159704878e-05, "loss": 0.1127, "step": 2190 }, { "grad_norm": 0.3905828595161438, "learning_rate": 9.985164933103929e-05, "loss": 0.1064, "step": 2200 }, { "grad_norm": 0.5291871428489685, "learning_rate": 9.984737649057513e-05, "loss": 0.1027, "step": 2210 }, { "grad_norm": 0.47446969151496887, "learning_rate": 9.984304308084827e-05, "loss": 0.0962, "step": 2220 }, { "grad_norm": 0.4433903992176056, "learning_rate": 9.983864910712416e-05, "loss": 0.1, "step": 2230 }, { "grad_norm": 0.4638219475746155, "learning_rate": 9.98341945747419e-05, "loss": 0.0994, "step": 2240 }, { "grad_norm": 0.4983360767364502, "learning_rate": 9.98296794891142e-05, "loss": 0.1072, "step": 2250 }, { "grad_norm": 0.43558934330940247, "learning_rate": 9.982510385572725e-05, "loss": 0.0818, "step": 2260 }, { "grad_norm": 0.4412870705127716, "learning_rate": 9.982046768014094e-05, "loss": 0.0935, "step": 2270 }, { "grad_norm": 0.4096260964870453, "learning_rate": 9.981577096798863e-05, "loss": 0.0873, "step": 2280 }, { "grad_norm": 0.5714293718338013, "learning_rate": 9.981101372497727e-05, "loss": 0.1321, "step": 2290 }, { "grad_norm": 0.32630112767219543, "learning_rate": 9.980619595688737e-05, "loss": 0.0774, "step": 2300 }, { "grad_norm": 0.4621831774711609, "learning_rate": 9.980131766957295e-05, "loss": 0.0925, "step": 2310 }, { "grad_norm": 0.4845905303955078, "learning_rate": 9.979637886896163e-05, "loss": 0.0837, "step": 2320 }, { "grad_norm": 0.46584513783454895, "learning_rate": 9.979137956105447e-05, "loss": 0.1021, "step": 2330 }, { "grad_norm": 0.44054335355758667, "learning_rate": 9.978631975192613e-05, "loss": 0.0907, "step": 2340 }, { "grad_norm": 0.5335957407951355, "learning_rate": 9.978119944772475e-05, "loss": 0.1118, "step": 2350 }, { "grad_norm": 0.4252392649650574, "learning_rate": 9.977601865467197e-05, "loss": 0.1098, "step": 2360 }, { "grad_norm": 0.5392661690711975, "learning_rate": 9.977077737906297e-05, "loss": 0.1027, "step": 2370 }, { "grad_norm": 0.4112991988658905, "learning_rate": 9.976547562726636e-05, "loss": 0.0948, "step": 2380 }, { "grad_norm": 0.4831346869468689, "learning_rate": 9.976011340572429e-05, "loss": 0.1092, "step": 2390 }, { "grad_norm": 0.41623151302337646, "learning_rate": 9.975469072095237e-05, "loss": 0.0918, "step": 2400 }, { "grad_norm": 0.3423726260662079, "learning_rate": 9.974920757953965e-05, "loss": 0.0882, "step": 2410 }, { "grad_norm": 0.4590810537338257, "learning_rate": 9.97436639881487e-05, "loss": 0.1385, "step": 2420 }, { "grad_norm": 0.4831586182117462, "learning_rate": 9.973805995351545e-05, "loss": 0.097, "step": 2430 }, { "grad_norm": 0.5352029800415039, "learning_rate": 9.973239548244939e-05, "loss": 0.1099, "step": 2440 }, { "grad_norm": 0.5652214884757996, "learning_rate": 9.972667058183333e-05, "loss": 0.1032, "step": 2450 }, { "grad_norm": 0.4202468991279602, "learning_rate": 9.972088525862362e-05, "loss": 0.0848, "step": 2460 }, { "grad_norm": 0.537562906742096, "learning_rate": 9.971503951984995e-05, "loss": 0.0919, "step": 2470 }, { "grad_norm": 0.4361104965209961, "learning_rate": 9.970913337261543e-05, "loss": 0.1091, "step": 2480 }, { "grad_norm": 0.42589330673217773, "learning_rate": 9.97031668240966e-05, "loss": 0.0882, "step": 2490 }, { "grad_norm": 0.3317532539367676, "learning_rate": 9.969713988154339e-05, "loss": 0.081, "step": 2500 }, { "grad_norm": 0.5919957756996155, "learning_rate": 9.969105255227906e-05, "loss": 0.0841, "step": 2510 }, { "grad_norm": 0.5040435194969177, "learning_rate": 9.968490484370035e-05, "loss": 0.0951, "step": 2520 }, { "grad_norm": 0.4618407189846039, "learning_rate": 9.967869676327726e-05, "loss": 0.0927, "step": 2530 }, { "grad_norm": 0.5180523991584778, "learning_rate": 9.96724283185532e-05, "loss": 0.1076, "step": 2540 }, { "grad_norm": 0.5802026987075806, "learning_rate": 9.966609951714494e-05, "loss": 0.1217, "step": 2550 }, { "grad_norm": 0.4124075770378113, "learning_rate": 9.965971036674255e-05, "loss": 0.1121, "step": 2560 }, { "grad_norm": 0.45880916714668274, "learning_rate": 9.965326087510947e-05, "loss": 0.1072, "step": 2570 }, { "grad_norm": 0.45518481731414795, "learning_rate": 9.964675105008244e-05, "loss": 0.1046, "step": 2580 }, { "grad_norm": 0.4403131306171417, "learning_rate": 9.964018089957147e-05, "loss": 0.0937, "step": 2590 }, { "grad_norm": 0.40046045184135437, "learning_rate": 9.963355043155996e-05, "loss": 0.1042, "step": 2600 }, { "grad_norm": 0.4444412887096405, "learning_rate": 9.962685965410455e-05, "loss": 0.0871, "step": 2610 }, { "grad_norm": 0.2661614418029785, "learning_rate": 9.962010857533514e-05, "loss": 0.0758, "step": 2620 }, { "grad_norm": 0.45521771907806396, "learning_rate": 9.961329720345493e-05, "loss": 0.0921, "step": 2630 }, { "grad_norm": 0.38939711451530457, "learning_rate": 9.96064255467404e-05, "loss": 0.1002, "step": 2640 }, { "grad_norm": 0.4601057171821594, "learning_rate": 9.959949361354126e-05, "loss": 0.0984, "step": 2650 }, { "grad_norm": 0.450595498085022, "learning_rate": 9.959250141228045e-05, "loss": 0.0946, "step": 2660 }, { "grad_norm": 0.4462396502494812, "learning_rate": 9.958544895145414e-05, "loss": 0.1067, "step": 2670 }, { "grad_norm": 0.4597097933292389, "learning_rate": 9.957833623963177e-05, "loss": 0.1227, "step": 2680 }, { "grad_norm": 0.5270684957504272, "learning_rate": 9.957116328545593e-05, "loss": 0.0898, "step": 2690 }, { "grad_norm": 0.43505752086639404, "learning_rate": 9.956393009764244e-05, "loss": 0.0883, "step": 2700 }, { "grad_norm": 0.3674320578575134, "learning_rate": 9.955663668498032e-05, "loss": 0.1022, "step": 2710 }, { "grad_norm": 0.4823606014251709, "learning_rate": 9.954928305633173e-05, "loss": 0.1034, "step": 2720 }, { "grad_norm": 0.6190723776817322, "learning_rate": 9.954186922063204e-05, "loss": 0.0949, "step": 2730 }, { "grad_norm": 0.5872604846954346, "learning_rate": 9.953439518688974e-05, "loss": 0.103, "step": 2740 }, { "grad_norm": 0.4424697458744049, "learning_rate": 9.952686096418652e-05, "loss": 0.1004, "step": 2750 }, { "grad_norm": 0.40386030077934265, "learning_rate": 9.951926656167714e-05, "loss": 0.0871, "step": 2760 }, { "grad_norm": 0.5215825438499451, "learning_rate": 9.951161198858953e-05, "loss": 0.087, "step": 2770 }, { "grad_norm": 0.34196022152900696, "learning_rate": 9.950389725422471e-05, "loss": 0.0961, "step": 2780 }, { "grad_norm": 0.5131240487098694, "learning_rate": 9.949612236795682e-05, "loss": 0.1099, "step": 2790 }, { "grad_norm": 0.3456869423389435, "learning_rate": 9.948828733923305e-05, "loss": 0.098, "step": 2800 }, { "grad_norm": 0.4857978820800781, "learning_rate": 9.948039217757374e-05, "loss": 0.1092, "step": 2810 }, { "grad_norm": 0.5231520533561707, "learning_rate": 9.947243689257225e-05, "loss": 0.1083, "step": 2820 }, { "grad_norm": 0.5414676070213318, "learning_rate": 9.946442149389497e-05, "loss": 0.0827, "step": 2830 }, { "grad_norm": 0.442600280046463, "learning_rate": 9.945634599128139e-05, "loss": 0.1031, "step": 2840 }, { "grad_norm": 0.5978485941886902, "learning_rate": 9.944821039454402e-05, "loss": 0.1017, "step": 2850 }, { "grad_norm": 0.3982934057712555, "learning_rate": 9.944001471356835e-05, "loss": 0.1018, "step": 2860 }, { "grad_norm": 0.5722682476043701, "learning_rate": 9.94317589583129e-05, "loss": 0.097, "step": 2870 }, { "grad_norm": 0.40946269035339355, "learning_rate": 9.942344313880922e-05, "loss": 0.0864, "step": 2880 }, { "grad_norm": 0.5635035037994385, "learning_rate": 9.941506726516179e-05, "loss": 0.1084, "step": 2890 }, { "grad_norm": 0.29737797379493713, "learning_rate": 9.94066313475481e-05, "loss": 0.1031, "step": 2900 }, { "grad_norm": 0.43938305974006653, "learning_rate": 9.939813539621857e-05, "loss": 0.0906, "step": 2910 }, { "grad_norm": 0.3778993785381317, "learning_rate": 9.93895794214966e-05, "loss": 0.0918, "step": 2920 }, { "grad_norm": 0.4366512894630432, "learning_rate": 9.938096343377852e-05, "loss": 0.076, "step": 2930 }, { "grad_norm": 0.3742038607597351, "learning_rate": 9.937228744353353e-05, "loss": 0.0849, "step": 2940 }, { "grad_norm": 0.6156976222991943, "learning_rate": 9.936355146130379e-05, "loss": 0.0744, "step": 2950 }, { "grad_norm": 0.4558393061161041, "learning_rate": 9.935475549770435e-05, "loss": 0.0911, "step": 2960 }, { "grad_norm": 0.4664708971977234, "learning_rate": 9.934589956342315e-05, "loss": 0.0839, "step": 2970 }, { "grad_norm": 0.4425683319568634, "learning_rate": 9.933698366922093e-05, "loss": 0.0843, "step": 2980 }, { "grad_norm": 0.6152092814445496, "learning_rate": 9.93280078259314e-05, "loss": 0.1135, "step": 2990 }, { "grad_norm": 0.4604393243789673, "learning_rate": 9.931897204446104e-05, "loss": 0.1016, "step": 3000 }, { "grad_norm": 0.3695294260978699, "learning_rate": 9.930987633578915e-05, "loss": 0.0641, "step": 3010 }, { "grad_norm": 0.3074706196784973, "learning_rate": 9.93007207109679e-05, "loss": 0.0741, "step": 3020 }, { "grad_norm": 0.3816603124141693, "learning_rate": 9.929150518112224e-05, "loss": 0.0712, "step": 3030 }, { "grad_norm": 0.2972683608531952, "learning_rate": 9.928222975744991e-05, "loss": 0.0838, "step": 3040 }, { "grad_norm": 0.34503281116485596, "learning_rate": 9.92728944512214e-05, "loss": 0.0787, "step": 3050 }, { "grad_norm": 0.4534955322742462, "learning_rate": 9.926349927378001e-05, "loss": 0.0785, "step": 3060 }, { "grad_norm": 0.34704679250717163, "learning_rate": 9.925404423654174e-05, "loss": 0.0684, "step": 3070 }, { "grad_norm": 0.33779966831207275, "learning_rate": 9.924452935099537e-05, "loss": 0.0725, "step": 3080 }, { "grad_norm": 0.4560643136501312, "learning_rate": 9.92349546287024e-05, "loss": 0.095, "step": 3090 }, { "grad_norm": 0.4062853157520294, "learning_rate": 9.9225320081297e-05, "loss": 0.0692, "step": 3100 }, { "grad_norm": 0.36584505438804626, "learning_rate": 9.921562572048606e-05, "loss": 0.0646, "step": 3110 }, { "grad_norm": 0.42547282576560974, "learning_rate": 9.920587155804913e-05, "loss": 0.066, "step": 3120 }, { "grad_norm": 0.37031126022338867, "learning_rate": 9.919605760583845e-05, "loss": 0.0766, "step": 3130 }, { "grad_norm": 0.41706132888793945, "learning_rate": 9.91861838757789e-05, "loss": 0.0879, "step": 3140 }, { "grad_norm": 0.3580584228038788, "learning_rate": 9.917625037986798e-05, "loss": 0.071, "step": 3150 }, { "grad_norm": 0.5560547113418579, "learning_rate": 9.916625713017583e-05, "loss": 0.0901, "step": 3160 }, { "grad_norm": 0.4494364559650421, "learning_rate": 9.915620413884519e-05, "loss": 0.0784, "step": 3170 }, { "grad_norm": 0.38397303223609924, "learning_rate": 9.914609141809139e-05, "loss": 0.0908, "step": 3180 }, { "grad_norm": 0.39296671748161316, "learning_rate": 9.913591898020235e-05, "loss": 0.0813, "step": 3190 }, { "grad_norm": 0.3453673720359802, "learning_rate": 9.912568683753853e-05, "loss": 0.0764, "step": 3200 }, { "grad_norm": 0.41762223839759827, "learning_rate": 9.911539500253295e-05, "loss": 0.0709, "step": 3210 }, { "grad_norm": 0.7119028568267822, "learning_rate": 9.910504348769118e-05, "loss": 0.0862, "step": 3220 }, { "grad_norm": 0.4524203836917877, "learning_rate": 9.909463230559127e-05, "loss": 0.0917, "step": 3230 }, { "grad_norm": 0.49764981865882874, "learning_rate": 9.908416146888376e-05, "loss": 0.0721, "step": 3240 }, { "grad_norm": 0.4580518305301666, "learning_rate": 9.907363099029175e-05, "loss": 0.0823, "step": 3250 }, { "grad_norm": 0.3435407280921936, "learning_rate": 9.906304088261073e-05, "loss": 0.0913, "step": 3260 }, { "grad_norm": 0.38877609372138977, "learning_rate": 9.905239115870872e-05, "loss": 0.0802, "step": 3270 }, { "grad_norm": 0.3511050343513489, "learning_rate": 9.90416818315261e-05, "loss": 0.0687, "step": 3280 }, { "grad_norm": 0.37033766508102417, "learning_rate": 9.903091291407573e-05, "loss": 0.0906, "step": 3290 }, { "grad_norm": 0.4122171401977539, "learning_rate": 9.902008441944286e-05, "loss": 0.0781, "step": 3300 }, { "grad_norm": 0.44277137517929077, "learning_rate": 9.900919636078512e-05, "loss": 0.0939, "step": 3310 }, { "grad_norm": 0.48251035809516907, "learning_rate": 9.899824875133255e-05, "loss": 0.1077, "step": 3320 }, { "grad_norm": 0.3936455547809601, "learning_rate": 9.898724160438749e-05, "loss": 0.086, "step": 3330 }, { "grad_norm": 0.3127557039260864, "learning_rate": 9.89761749333247e-05, "loss": 0.0811, "step": 3340 }, { "grad_norm": 0.40884605050086975, "learning_rate": 9.896504875159122e-05, "loss": 0.0914, "step": 3350 }, { "grad_norm": 0.4506255090236664, "learning_rate": 9.89538630727064e-05, "loss": 0.0876, "step": 3360 }, { "grad_norm": 0.31989049911499023, "learning_rate": 9.894261791026189e-05, "loss": 0.0926, "step": 3370 }, { "grad_norm": 0.37960129976272583, "learning_rate": 9.893131327792165e-05, "loss": 0.1215, "step": 3380 }, { "grad_norm": 0.3188486695289612, "learning_rate": 9.891994918942182e-05, "loss": 0.0864, "step": 3390 }, { "grad_norm": 0.3180984854698181, "learning_rate": 9.890852565857092e-05, "loss": 0.0832, "step": 3400 }, { "grad_norm": 0.3113315999507904, "learning_rate": 9.889704269924954e-05, "loss": 0.0888, "step": 3410 }, { "grad_norm": 0.425373911857605, "learning_rate": 9.888550032541059e-05, "loss": 0.0864, "step": 3420 }, { "grad_norm": 0.31875455379486084, "learning_rate": 9.887389855107916e-05, "loss": 0.082, "step": 3430 }, { "grad_norm": 0.5382300615310669, "learning_rate": 9.886223739035248e-05, "loss": 0.1032, "step": 3440 }, { "grad_norm": 0.34398865699768066, "learning_rate": 9.885051685739997e-05, "loss": 0.0917, "step": 3450 }, { "grad_norm": 0.3874058127403259, "learning_rate": 9.883873696646316e-05, "loss": 0.0781, "step": 3460 }, { "grad_norm": 0.4429183006286621, "learning_rate": 9.882689773185575e-05, "loss": 0.1036, "step": 3470 }, { "grad_norm": 0.3680932819843292, "learning_rate": 9.881499916796353e-05, "loss": 0.0924, "step": 3480 }, { "grad_norm": 0.46685484051704407, "learning_rate": 9.880304128924434e-05, "loss": 0.0921, "step": 3490 }, { "grad_norm": 0.4526357054710388, "learning_rate": 9.879102411022817e-05, "loss": 0.0845, "step": 3500 }, { "grad_norm": 0.3050115406513214, "learning_rate": 9.877894764551703e-05, "loss": 0.0851, "step": 3510 }, { "grad_norm": 0.4201575219631195, "learning_rate": 9.876681190978494e-05, "loss": 0.0887, "step": 3520 }, { "grad_norm": 0.3609040677547455, "learning_rate": 9.875461691777797e-05, "loss": 0.0903, "step": 3530 }, { "grad_norm": 0.44264477491378784, "learning_rate": 9.874236268431417e-05, "loss": 0.1031, "step": 3540 }, { "grad_norm": 0.4067551791667938, "learning_rate": 9.873004922428361e-05, "loss": 0.1034, "step": 3550 }, { "grad_norm": 0.4394097328186035, "learning_rate": 9.871767655264829e-05, "loss": 0.0804, "step": 3560 }, { "grad_norm": 0.415182501077652, "learning_rate": 9.87052446844422e-05, "loss": 0.1098, "step": 3570 }, { "grad_norm": 0.35675379633903503, "learning_rate": 9.869275363477122e-05, "loss": 0.1229, "step": 3580 }, { "grad_norm": 0.3474445343017578, "learning_rate": 9.868020341881312e-05, "loss": 0.1104, "step": 3590 }, { "grad_norm": 0.45996662974357605, "learning_rate": 9.866759405181765e-05, "loss": 0.1002, "step": 3600 }, { "grad_norm": 0.4077044725418091, "learning_rate": 9.865492554910633e-05, "loss": 0.1082, "step": 3610 }, { "grad_norm": 0.4369373917579651, "learning_rate": 9.864219792607262e-05, "loss": 0.0938, "step": 3620 }, { "grad_norm": 0.32438090443611145, "learning_rate": 9.862941119818177e-05, "loss": 0.0879, "step": 3630 }, { "grad_norm": 0.46805909276008606, "learning_rate": 9.861656538097086e-05, "loss": 0.0878, "step": 3640 }, { "grad_norm": 0.49294430017471313, "learning_rate": 9.860366049004877e-05, "loss": 0.1027, "step": 3650 }, { "grad_norm": 0.44509759545326233, "learning_rate": 9.859069654109615e-05, "loss": 0.1022, "step": 3660 }, { "grad_norm": 0.4021644592285156, "learning_rate": 9.857767354986545e-05, "loss": 0.0896, "step": 3670 }, { "grad_norm": 0.556555449962616, "learning_rate": 9.856459153218078e-05, "loss": 0.1037, "step": 3680 }, { "grad_norm": 0.39403751492500305, "learning_rate": 9.855145050393808e-05, "loss": 0.0947, "step": 3690 }, { "grad_norm": 0.4582133889198303, "learning_rate": 9.85382504811049e-05, "loss": 0.1028, "step": 3700 }, { "grad_norm": 0.35186225175857544, "learning_rate": 9.852499147972054e-05, "loss": 0.0991, "step": 3710 }, { "grad_norm": 0.45507970452308655, "learning_rate": 9.851167351589592e-05, "loss": 0.1041, "step": 3720 }, { "grad_norm": 0.4589422345161438, "learning_rate": 9.849829660581363e-05, "loss": 0.0841, "step": 3730 }, { "grad_norm": 0.33682236075401306, "learning_rate": 9.848486076572787e-05, "loss": 0.0963, "step": 3740 }, { "grad_norm": 0.4150694012641907, "learning_rate": 9.847136601196446e-05, "loss": 0.1015, "step": 3750 }, { "grad_norm": 0.4066098630428314, "learning_rate": 9.845781236092078e-05, "loss": 0.0858, "step": 3760 }, { "grad_norm": 0.4496650695800781, "learning_rate": 9.844419982906583e-05, "loss": 0.1087, "step": 3770 }, { "grad_norm": 0.4405280351638794, "learning_rate": 9.843052843294008e-05, "loss": 0.1036, "step": 3780 }, { "grad_norm": 0.37568551301956177, "learning_rate": 9.841679818915559e-05, "loss": 0.0833, "step": 3790 }, { "grad_norm": 0.38421598076820374, "learning_rate": 9.840300911439591e-05, "loss": 0.1171, "step": 3800 }, { "grad_norm": 0.3990124762058258, "learning_rate": 9.838916122541603e-05, "loss": 0.0862, "step": 3810 }, { "grad_norm": 0.4300597012042999, "learning_rate": 9.837525453904246e-05, "loss": 0.1049, "step": 3820 }, { "grad_norm": 0.37744995951652527, "learning_rate": 9.836128907217314e-05, "loss": 0.0818, "step": 3830 }, { "grad_norm": 0.37615966796875, "learning_rate": 9.834726484177743e-05, "loss": 0.083, "step": 3840 }, { "grad_norm": 0.3812742531299591, "learning_rate": 9.833318186489609e-05, "loss": 0.1041, "step": 3850 }, { "grad_norm": 0.46358123421669006, "learning_rate": 9.831904015864126e-05, "loss": 0.0858, "step": 3860 }, { "grad_norm": 0.5451292991638184, "learning_rate": 9.830483974019645e-05, "loss": 0.0844, "step": 3870 }, { "grad_norm": 0.5372259616851807, "learning_rate": 9.82905806268165e-05, "loss": 0.0891, "step": 3880 }, { "grad_norm": 0.4205310046672821, "learning_rate": 9.82762628358276e-05, "loss": 0.1028, "step": 3890 }, { "grad_norm": 0.396879106760025, "learning_rate": 9.826188638462718e-05, "loss": 0.1027, "step": 3900 }, { "grad_norm": 0.40621599555015564, "learning_rate": 9.824745129068402e-05, "loss": 0.0888, "step": 3910 }, { "grad_norm": 0.4687616229057312, "learning_rate": 9.82329575715381e-05, "loss": 0.091, "step": 3920 }, { "grad_norm": 0.335933119058609, "learning_rate": 9.821840524480066e-05, "loss": 0.0807, "step": 3930 }, { "grad_norm": 0.43819114565849304, "learning_rate": 9.820379432815414e-05, "loss": 0.0971, "step": 3940 }, { "grad_norm": 0.3565191924571991, "learning_rate": 9.81891248393522e-05, "loss": 0.0824, "step": 3950 }, { "grad_norm": 0.40219879150390625, "learning_rate": 9.817439679621963e-05, "loss": 0.1034, "step": 3960 }, { "grad_norm": 0.4466690123081207, "learning_rate": 9.815961021665243e-05, "loss": 0.0871, "step": 3970 }, { "grad_norm": 0.5360392332077026, "learning_rate": 9.814476511861763e-05, "loss": 0.0936, "step": 3980 }, { "grad_norm": 0.429108202457428, "learning_rate": 9.812986152015348e-05, "loss": 0.0876, "step": 3990 }, { "grad_norm": 0.4050862193107605, "learning_rate": 9.811489943936922e-05, "loss": 0.0952, "step": 4000 }, { "grad_norm": 0.3858043849468231, "learning_rate": 9.809987889444522e-05, "loss": 0.0871, "step": 4010 }, { "grad_norm": 0.3494444489479065, "learning_rate": 9.808479990363282e-05, "loss": 0.0901, "step": 4020 }, { "grad_norm": 0.4903619885444641, "learning_rate": 9.806966248525445e-05, "loss": 0.101, "step": 4030 }, { "grad_norm": 0.41512423753738403, "learning_rate": 9.805446665770348e-05, "loss": 0.0712, "step": 4040 }, { "grad_norm": 0.4817846417427063, "learning_rate": 9.803921243944429e-05, "loss": 0.0793, "step": 4050 }, { "grad_norm": 0.40218764543533325, "learning_rate": 9.802389984901218e-05, "loss": 0.081, "step": 4060 }, { "grad_norm": 0.4342617392539978, "learning_rate": 9.80085289050134e-05, "loss": 0.0742, "step": 4070 }, { "grad_norm": 0.3691234588623047, "learning_rate": 9.799309962612508e-05, "loss": 0.0963, "step": 4080 }, { "grad_norm": 0.4554790258407593, "learning_rate": 9.797761203109527e-05, "loss": 0.0891, "step": 4090 }, { "grad_norm": 0.40157392621040344, "learning_rate": 9.796206613874283e-05, "loss": 0.0802, "step": 4100 }, { "grad_norm": 0.41646337509155273, "learning_rate": 9.794646196795754e-05, "loss": 0.1091, "step": 4110 }, { "grad_norm": 0.47498819231987, "learning_rate": 9.793079953769987e-05, "loss": 0.1018, "step": 4120 }, { "grad_norm": 0.4234955906867981, "learning_rate": 9.79150788670012e-05, "loss": 0.0971, "step": 4130 }, { "grad_norm": 0.47126302123069763, "learning_rate": 9.78992999749636e-05, "loss": 0.0897, "step": 4140 }, { "grad_norm": 0.43909552693367004, "learning_rate": 9.788346288075994e-05, "loss": 0.1014, "step": 4150 }, { "grad_norm": 0.529973030090332, "learning_rate": 9.786756760363373e-05, "loss": 0.0882, "step": 4160 }, { "grad_norm": 0.3986838757991791, "learning_rate": 9.78516141628993e-05, "loss": 0.0885, "step": 4170 }, { "grad_norm": 0.44612765312194824, "learning_rate": 9.783560257794154e-05, "loss": 0.0846, "step": 4180 }, { "grad_norm": 0.35597026348114014, "learning_rate": 9.781953286821603e-05, "loss": 0.0888, "step": 4190 }, { "grad_norm": 0.5065276026725769, "learning_rate": 9.780340505324901e-05, "loss": 0.0866, "step": 4200 }, { "grad_norm": 0.40804746747016907, "learning_rate": 9.778721915263727e-05, "loss": 0.0879, "step": 4210 }, { "grad_norm": 0.5177534222602844, "learning_rate": 9.777097518604824e-05, "loss": 0.1092, "step": 4220 }, { "grad_norm": 0.3254549205303192, "learning_rate": 9.775467317321984e-05, "loss": 0.1069, "step": 4230 }, { "grad_norm": 0.3936311900615692, "learning_rate": 9.773831313396055e-05, "loss": 0.1024, "step": 4240 }, { "grad_norm": 0.4377549886703491, "learning_rate": 9.77218950881494e-05, "loss": 0.0927, "step": 4250 }, { "grad_norm": 0.9518507719039917, "learning_rate": 9.770541905573583e-05, "loss": 0.0996, "step": 4260 }, { "grad_norm": 0.3887861967086792, "learning_rate": 9.768888505673976e-05, "loss": 0.1029, "step": 4270 }, { "grad_norm": 0.3961091935634613, "learning_rate": 9.767229311125162e-05, "loss": 0.0901, "step": 4280 }, { "grad_norm": 0.41934943199157715, "learning_rate": 9.765564323943211e-05, "loss": 0.0859, "step": 4290 }, { "grad_norm": 0.3683587312698364, "learning_rate": 9.763893546151244e-05, "loss": 0.0845, "step": 4300 }, { "grad_norm": 0.4356439411640167, "learning_rate": 9.762216979779412e-05, "loss": 0.0896, "step": 4310 }, { "grad_norm": 0.5015527606010437, "learning_rate": 9.760534626864902e-05, "loss": 0.1138, "step": 4320 }, { "grad_norm": 0.47165796160697937, "learning_rate": 9.758846489451931e-05, "loss": 0.0871, "step": 4330 }, { "grad_norm": 0.43491995334625244, "learning_rate": 9.757152569591748e-05, "loss": 0.0782, "step": 4340 }, { "grad_norm": 0.310588002204895, "learning_rate": 9.75545286934262e-05, "loss": 0.0813, "step": 4350 }, { "grad_norm": 0.4053053557872772, "learning_rate": 9.753747390769847e-05, "loss": 0.0753, "step": 4360 }, { "grad_norm": 0.5220550894737244, "learning_rate": 9.752036135945744e-05, "loss": 0.0952, "step": 4370 }, { "grad_norm": 0.45605993270874023, "learning_rate": 9.750319106949649e-05, "loss": 0.0872, "step": 4380 }, { "grad_norm": 0.4115452468395233, "learning_rate": 9.748596305867913e-05, "loss": 0.0741, "step": 4390 }, { "grad_norm": 0.3586653172969818, "learning_rate": 9.746867734793903e-05, "loss": 0.1018, "step": 4400 }, { "grad_norm": 0.32515329122543335, "learning_rate": 9.745133395827993e-05, "loss": 0.0874, "step": 4410 }, { "grad_norm": 0.4078673720359802, "learning_rate": 9.743393291077572e-05, "loss": 0.1017, "step": 4420 }, { "grad_norm": 0.4147643446922302, "learning_rate": 9.741647422657028e-05, "loss": 0.0911, "step": 4430 }, { "grad_norm": 0.422559529542923, "learning_rate": 9.739895792687758e-05, "loss": 0.0884, "step": 4440 }, { "grad_norm": 0.34459203481674194, "learning_rate": 9.738138403298157e-05, "loss": 0.0949, "step": 4450 }, { "grad_norm": 0.3680853247642517, "learning_rate": 9.736375256623619e-05, "loss": 0.0815, "step": 4460 }, { "grad_norm": 0.40941303968429565, "learning_rate": 9.734606354806533e-05, "loss": 0.08, "step": 4470 }, { "grad_norm": 0.3629305958747864, "learning_rate": 9.73283169999628e-05, "loss": 0.0875, "step": 4480 }, { "grad_norm": 0.41937246918678284, "learning_rate": 9.731051294349238e-05, "loss": 0.093, "step": 4490 }, { "grad_norm": 0.28256770968437195, "learning_rate": 9.729265140028762e-05, "loss": 0.082, "step": 4500 }, { "grad_norm": 0.32177984714508057, "learning_rate": 9.727473239205201e-05, "loss": 0.1063, "step": 4510 }, { "grad_norm": 0.36707842350006104, "learning_rate": 9.725675594055883e-05, "loss": 0.0875, "step": 4520 }, { "grad_norm": 0.39332568645477295, "learning_rate": 9.723872206765116e-05, "loss": 0.0955, "step": 4530 }, { "grad_norm": 0.5022501945495605, "learning_rate": 9.722063079524185e-05, "loss": 0.107, "step": 4540 }, { "grad_norm": 0.32705047726631165, "learning_rate": 9.720248214531351e-05, "loss": 0.0978, "step": 4550 }, { "grad_norm": 0.3915790021419525, "learning_rate": 9.718427613991848e-05, "loss": 0.0821, "step": 4560 }, { "grad_norm": 0.3863784670829773, "learning_rate": 9.716601280117873e-05, "loss": 0.0991, "step": 4570 }, { "grad_norm": 0.4040505290031433, "learning_rate": 9.714769215128596e-05, "loss": 0.0918, "step": 4580 }, { "grad_norm": 0.3403347134590149, "learning_rate": 9.712931421250152e-05, "loss": 0.0788, "step": 4590 }, { "grad_norm": 0.4703873097896576, "learning_rate": 9.711087900715627e-05, "loss": 0.0904, "step": 4600 }, { "grad_norm": 0.3685687780380249, "learning_rate": 9.709238655765078e-05, "loss": 0.0924, "step": 4610 }, { "grad_norm": 0.4195750057697296, "learning_rate": 9.707383688645511e-05, "loss": 0.0889, "step": 4620 }, { "grad_norm": 0.4628749191761017, "learning_rate": 9.705523001610883e-05, "loss": 0.0971, "step": 4630 }, { "grad_norm": 0.3873494267463684, "learning_rate": 9.703656596922107e-05, "loss": 0.0953, "step": 4640 }, { "grad_norm": 0.3203631341457367, "learning_rate": 9.70178447684704e-05, "loss": 0.0805, "step": 4650 }, { "grad_norm": 0.35135212540626526, "learning_rate": 9.699906643660483e-05, "loss": 0.0954, "step": 4660 }, { "grad_norm": 0.5061282515525818, "learning_rate": 9.698023099644185e-05, "loss": 0.096, "step": 4670 }, { "grad_norm": 0.4155007302761078, "learning_rate": 9.696133847086823e-05, "loss": 0.0847, "step": 4680 }, { "grad_norm": 0.3662669360637665, "learning_rate": 9.694238888284022e-05, "loss": 0.1061, "step": 4690 }, { "grad_norm": 0.31039708852767944, "learning_rate": 9.692338225538333e-05, "loss": 0.0915, "step": 4700 }, { "grad_norm": 0.418296217918396, "learning_rate": 9.690431861159241e-05, "loss": 0.0917, "step": 4710 }, { "grad_norm": 0.44139793515205383, "learning_rate": 9.688519797463161e-05, "loss": 0.0885, "step": 4720 }, { "grad_norm": 0.39166197180747986, "learning_rate": 9.686602036773426e-05, "loss": 0.1073, "step": 4730 }, { "grad_norm": 0.41829442977905273, "learning_rate": 9.684678581420302e-05, "loss": 0.0975, "step": 4740 }, { "grad_norm": 0.3788524568080902, "learning_rate": 9.682749433740962e-05, "loss": 0.0841, "step": 4750 }, { "grad_norm": 0.4820230305194855, "learning_rate": 9.680814596079507e-05, "loss": 0.0982, "step": 4760 }, { "grad_norm": 0.3911021053791046, "learning_rate": 9.678874070786945e-05, "loss": 0.0918, "step": 4770 }, { "grad_norm": 0.4559861123561859, "learning_rate": 9.676927860221199e-05, "loss": 0.1136, "step": 4780 }, { "grad_norm": 0.40197858214378357, "learning_rate": 9.674975966747097e-05, "loss": 0.0815, "step": 4790 }, { "grad_norm": 0.3797534704208374, "learning_rate": 9.673018392736374e-05, "loss": 0.0865, "step": 4800 }, { "grad_norm": 0.4899069368839264, "learning_rate": 9.671055140567667e-05, "loss": 0.0829, "step": 4810 }, { "grad_norm": 0.3916506767272949, "learning_rate": 9.669086212626511e-05, "loss": 0.0964, "step": 4820 }, { "grad_norm": 0.36679989099502563, "learning_rate": 9.667111611305341e-05, "loss": 0.0821, "step": 4830 }, { "grad_norm": 0.40763694047927856, "learning_rate": 9.665131339003486e-05, "loss": 0.0932, "step": 4840 }, { "grad_norm": 0.33907774090766907, "learning_rate": 9.663145398127158e-05, "loss": 0.0834, "step": 4850 }, { "grad_norm": 0.3791409432888031, "learning_rate": 9.661153791089467e-05, "loss": 0.0748, "step": 4860 }, { "grad_norm": 0.4505537152290344, "learning_rate": 9.659156520310402e-05, "loss": 0.0842, "step": 4870 }, { "grad_norm": 0.43753907084465027, "learning_rate": 9.657153588216835e-05, "loss": 0.0922, "step": 4880 }, { "grad_norm": 0.46918532252311707, "learning_rate": 9.655144997242516e-05, "loss": 0.1092, "step": 4890 }, { "grad_norm": 0.3426768481731415, "learning_rate": 9.653130749828075e-05, "loss": 0.0791, "step": 4900 }, { "grad_norm": 0.4599316120147705, "learning_rate": 9.65111084842101e-05, "loss": 0.0863, "step": 4910 }, { "grad_norm": 0.44941452145576477, "learning_rate": 9.649085295475695e-05, "loss": 0.1005, "step": 4920 }, { "grad_norm": 0.3883955180644989, "learning_rate": 9.647054093453365e-05, "loss": 0.0782, "step": 4930 }, { "grad_norm": 0.42258569598197937, "learning_rate": 9.645017244822123e-05, "loss": 0.0888, "step": 4940 }, { "grad_norm": 0.3640228807926178, "learning_rate": 9.642974752056931e-05, "loss": 0.0987, "step": 4950 }, { "grad_norm": 0.3292219340801239, "learning_rate": 9.640926617639613e-05, "loss": 0.0927, "step": 4960 }, { "grad_norm": 0.35426995158195496, "learning_rate": 9.638872844058843e-05, "loss": 0.0988, "step": 4970 }, { "grad_norm": 0.37547269463539124, "learning_rate": 9.63681343381015e-05, "loss": 0.0876, "step": 4980 }, { "grad_norm": 0.3024173378944397, "learning_rate": 9.634748389395914e-05, "loss": 0.092, "step": 4990 }, { "grad_norm": 0.35037025809288025, "learning_rate": 9.632677713325353e-05, "loss": 0.1023, "step": 5000 }, { "grad_norm": 0.3905862271785736, "learning_rate": 9.63060140811454e-05, "loss": 0.0926, "step": 5010 }, { "grad_norm": 0.4529273808002472, "learning_rate": 9.628519476286379e-05, "loss": 0.0962, "step": 5020 }, { "grad_norm": 0.3958476483821869, "learning_rate": 9.626431920370612e-05, "loss": 0.0837, "step": 5030 }, { "grad_norm": 0.2599676549434662, "learning_rate": 9.624338742903819e-05, "loss": 0.0757, "step": 5040 }, { "grad_norm": 0.4286079406738281, "learning_rate": 9.622239946429406e-05, "loss": 0.0896, "step": 5050 }, { "grad_norm": 0.3273478150367737, "learning_rate": 9.620135533497609e-05, "loss": 0.0579, "step": 5060 }, { "grad_norm": 0.4061534106731415, "learning_rate": 9.61802550666549e-05, "loss": 0.0992, "step": 5070 }, { "grad_norm": 0.37974825501441956, "learning_rate": 9.615909868496928e-05, "loss": 0.0904, "step": 5080 }, { "grad_norm": 0.36587247252464294, "learning_rate": 9.613788621562622e-05, "loss": 0.0843, "step": 5090 }, { "grad_norm": 0.32553428411483765, "learning_rate": 9.611661768440093e-05, "loss": 0.084, "step": 5100 }, { "grad_norm": 0.33018380403518677, "learning_rate": 9.609529311713661e-05, "loss": 0.0868, "step": 5110 }, { "grad_norm": 0.45552170276641846, "learning_rate": 9.607391253974466e-05, "loss": 0.1101, "step": 5120 }, { "grad_norm": 0.3310212790966034, "learning_rate": 9.605247597820448e-05, "loss": 0.0954, "step": 5130 }, { "grad_norm": 0.4502544105052948, "learning_rate": 9.603098345856354e-05, "loss": 0.0793, "step": 5140 }, { "grad_norm": 0.28509441018104553, "learning_rate": 9.600943500693725e-05, "loss": 0.0733, "step": 5150 }, { "grad_norm": 0.4297519028186798, "learning_rate": 9.598783064950902e-05, "loss": 0.0979, "step": 5160 }, { "grad_norm": 0.3371419906616211, "learning_rate": 9.596617041253018e-05, "loss": 0.0779, "step": 5170 }, { "grad_norm": 0.30251070857048035, "learning_rate": 9.594445432231996e-05, "loss": 0.0758, "step": 5180 }, { "grad_norm": 0.3779813051223755, "learning_rate": 9.592268240526547e-05, "loss": 0.1002, "step": 5190 }, { "grad_norm": 0.3391347825527191, "learning_rate": 9.590085468782162e-05, "loss": 0.0632, "step": 5200 }, { "grad_norm": 0.416698694229126, "learning_rate": 9.587897119651116e-05, "loss": 0.0746, "step": 5210 }, { "grad_norm": 0.34892556071281433, "learning_rate": 9.585703195792459e-05, "loss": 0.0757, "step": 5220 }, { "grad_norm": 0.30562624335289, "learning_rate": 9.583503699872016e-05, "loss": 0.0635, "step": 5230 }, { "grad_norm": 0.39323118329048157, "learning_rate": 9.581298634562381e-05, "loss": 0.0676, "step": 5240 }, { "grad_norm": 0.2651589810848236, "learning_rate": 9.579088002542917e-05, "loss": 0.0814, "step": 5250 }, { "grad_norm": 0.35021722316741943, "learning_rate": 9.57687180649975e-05, "loss": 0.0702, "step": 5260 }, { "grad_norm": 0.3869832754135132, "learning_rate": 9.574650049125768e-05, "loss": 0.0744, "step": 5270 }, { "grad_norm": 0.41678768396377563, "learning_rate": 9.572422733120614e-05, "loss": 0.0868, "step": 5280 }, { "grad_norm": 0.2925581932067871, "learning_rate": 9.570189861190689e-05, "loss": 0.0649, "step": 5290 }, { "grad_norm": 0.3191482722759247, "learning_rate": 9.56795143604914e-05, "loss": 0.0683, "step": 5300 }, { "grad_norm": 0.33472707867622375, "learning_rate": 9.565707460415869e-05, "loss": 0.0692, "step": 5310 }, { "grad_norm": 0.3466562032699585, "learning_rate": 9.563457937017515e-05, "loss": 0.0663, "step": 5320 }, { "grad_norm": 0.42158159613609314, "learning_rate": 9.56120286858746e-05, "loss": 0.0869, "step": 5330 }, { "grad_norm": 0.28790831565856934, "learning_rate": 9.558942257865829e-05, "loss": 0.0718, "step": 5340 }, { "grad_norm": 0.41247522830963135, "learning_rate": 9.556676107599472e-05, "loss": 0.0654, "step": 5350 }, { "grad_norm": 0.26415616273880005, "learning_rate": 9.554404420541978e-05, "loss": 0.0732, "step": 5360 }, { "grad_norm": 0.301887571811676, "learning_rate": 9.55212719945366e-05, "loss": 0.051, "step": 5370 }, { "grad_norm": 0.39712151885032654, "learning_rate": 9.549844447101559e-05, "loss": 0.0568, "step": 5380 }, { "grad_norm": 0.35504209995269775, "learning_rate": 9.547556166259433e-05, "loss": 0.0537, "step": 5390 }, { "grad_norm": 0.31485041975975037, "learning_rate": 9.545262359707756e-05, "loss": 0.0547, "step": 5400 }, { "grad_norm": 0.2986229658126831, "learning_rate": 9.542963030233724e-05, "loss": 0.06, "step": 5410 }, { "grad_norm": 0.24231328070163727, "learning_rate": 9.540658180631237e-05, "loss": 0.0553, "step": 5420 }, { "grad_norm": 0.2972283363342285, "learning_rate": 9.538347813700904e-05, "loss": 0.0629, "step": 5430 }, { "grad_norm": 0.32967084646224976, "learning_rate": 9.536031932250036e-05, "loss": 0.0791, "step": 5440 }, { "grad_norm": 0.368572473526001, "learning_rate": 9.533710539092653e-05, "loss": 0.0508, "step": 5450 }, { "grad_norm": 0.37674200534820557, "learning_rate": 9.531383637049464e-05, "loss": 0.065, "step": 5460 }, { "grad_norm": 0.3544699549674988, "learning_rate": 9.529051228947875e-05, "loss": 0.0507, "step": 5470 }, { "grad_norm": 0.28372296690940857, "learning_rate": 9.52671331762198e-05, "loss": 0.0541, "step": 5480 }, { "grad_norm": 0.27456921339035034, "learning_rate": 9.524369905912565e-05, "loss": 0.0586, "step": 5490 }, { "grad_norm": 0.2676243782043457, "learning_rate": 9.522020996667092e-05, "loss": 0.0629, "step": 5500 }, { "grad_norm": 0.32565543055534363, "learning_rate": 9.519666592739709e-05, "loss": 0.0649, "step": 5510 }, { "grad_norm": 0.2683241665363312, "learning_rate": 9.517306696991241e-05, "loss": 0.0574, "step": 5520 }, { "grad_norm": 0.19389525055885315, "learning_rate": 9.51494131228918e-05, "loss": 0.0546, "step": 5530 }, { "grad_norm": 0.32473140954971313, "learning_rate": 9.512570441507695e-05, "loss": 0.0468, "step": 5540 }, { "grad_norm": 0.2534855306148529, "learning_rate": 9.510194087527615e-05, "loss": 0.0596, "step": 5550 }, { "grad_norm": 0.29308661818504333, "learning_rate": 9.507812253236435e-05, "loss": 0.0646, "step": 5560 }, { "grad_norm": 0.3338563144207001, "learning_rate": 9.505424941528309e-05, "loss": 0.0609, "step": 5570 }, { "grad_norm": 0.4182955026626587, "learning_rate": 9.503032155304046e-05, "loss": 0.0819, "step": 5580 }, { "grad_norm": 0.39998859167099, "learning_rate": 9.500633897471106e-05, "loss": 0.0607, "step": 5590 }, { "grad_norm": 0.34791192412376404, "learning_rate": 9.498230170943596e-05, "loss": 0.0728, "step": 5600 }, { "grad_norm": 0.358845055103302, "learning_rate": 9.495820978642275e-05, "loss": 0.0669, "step": 5610 }, { "grad_norm": 0.3373357057571411, "learning_rate": 9.493406323494535e-05, "loss": 0.0553, "step": 5620 }, { "grad_norm": 0.33360862731933594, "learning_rate": 9.490986208434413e-05, "loss": 0.0532, "step": 5630 }, { "grad_norm": 0.39161887764930725, "learning_rate": 9.488560636402577e-05, "loss": 0.0634, "step": 5640 }, { "grad_norm": 0.2505131661891937, "learning_rate": 9.486129610346321e-05, "loss": 0.0688, "step": 5650 }, { "grad_norm": 0.3050679862499237, "learning_rate": 9.483693133219576e-05, "loss": 0.0613, "step": 5660 }, { "grad_norm": 0.3585796356201172, "learning_rate": 9.481251207982888e-05, "loss": 0.0541, "step": 5670 }, { "grad_norm": 0.3364236652851105, "learning_rate": 9.47880383760343e-05, "loss": 0.0623, "step": 5680 }, { "grad_norm": 0.2877309322357178, "learning_rate": 9.476351025054983e-05, "loss": 0.0517, "step": 5690 }, { "grad_norm": 0.2761077880859375, "learning_rate": 9.473892773317952e-05, "loss": 0.0648, "step": 5700 }, { "grad_norm": 0.22885943949222565, "learning_rate": 9.471429085379338e-05, "loss": 0.0669, "step": 5710 }, { "grad_norm": 0.29400819540023804, "learning_rate": 9.468959964232757e-05, "loss": 0.0603, "step": 5720 }, { "grad_norm": 0.29646435379981995, "learning_rate": 9.466485412878425e-05, "loss": 0.0644, "step": 5730 }, { "grad_norm": 0.42985376715660095, "learning_rate": 9.464005434323154e-05, "loss": 0.0777, "step": 5740 }, { "grad_norm": 0.27659741044044495, "learning_rate": 9.461520031580352e-05, "loss": 0.0595, "step": 5750 }, { "grad_norm": 0.40346115827560425, "learning_rate": 9.459029207670019e-05, "loss": 0.0622, "step": 5760 }, { "grad_norm": 0.3765413165092468, "learning_rate": 9.456532965618737e-05, "loss": 0.0747, "step": 5770 }, { "grad_norm": 0.352042019367218, "learning_rate": 9.454031308459681e-05, "loss": 0.0834, "step": 5780 }, { "grad_norm": 0.2860874831676483, "learning_rate": 9.451524239232595e-05, "loss": 0.0653, "step": 5790 }, { "grad_norm": 0.4556673467159271, "learning_rate": 9.449011760983809e-05, "loss": 0.0886, "step": 5800 }, { "grad_norm": 0.3578868806362152, "learning_rate": 9.446493876766218e-05, "loss": 0.069, "step": 5810 }, { "grad_norm": 0.38888880610466003, "learning_rate": 9.44397058963929e-05, "loss": 0.0647, "step": 5820 }, { "grad_norm": 0.34869369864463806, "learning_rate": 9.441441902669056e-05, "loss": 0.0743, "step": 5830 }, { "grad_norm": 0.33114078640937805, "learning_rate": 9.43890781892811e-05, "loss": 0.0764, "step": 5840 }, { "grad_norm": 0.38846471905708313, "learning_rate": 9.436368341495603e-05, "loss": 0.0545, "step": 5850 }, { "grad_norm": 0.43637844920158386, "learning_rate": 9.43382347345724e-05, "loss": 0.0753, "step": 5860 }, { "grad_norm": 0.29598769545555115, "learning_rate": 9.431273217905273e-05, "loss": 0.0719, "step": 5870 }, { "grad_norm": 0.41488534212112427, "learning_rate": 9.428717577938504e-05, "loss": 0.0661, "step": 5880 }, { "grad_norm": 0.32096976041793823, "learning_rate": 9.426156556662276e-05, "loss": 0.0727, "step": 5890 }, { "grad_norm": 0.3249981999397278, "learning_rate": 9.423590157188474e-05, "loss": 0.0897, "step": 5900 }, { "grad_norm": 0.317863792181015, "learning_rate": 9.421018382635513e-05, "loss": 0.083, "step": 5910 }, { "grad_norm": 0.42008042335510254, "learning_rate": 9.418441236128343e-05, "loss": 0.0746, "step": 5920 }, { "grad_norm": 0.31067824363708496, "learning_rate": 9.41585872079844e-05, "loss": 0.0718, "step": 5930 }, { "grad_norm": 0.405107706785202, "learning_rate": 9.413270839783802e-05, "loss": 0.068, "step": 5940 }, { "grad_norm": 0.3707500398159027, "learning_rate": 9.41067759622895e-05, "loss": 0.074, "step": 5950 }, { "grad_norm": 0.4251691401004791, "learning_rate": 9.408078993284917e-05, "loss": 0.0719, "step": 5960 }, { "grad_norm": 0.3873128592967987, "learning_rate": 9.405475034109254e-05, "loss": 0.0815, "step": 5970 }, { "grad_norm": 0.3870551586151123, "learning_rate": 9.402865721866015e-05, "loss": 0.0795, "step": 5980 }, { "grad_norm": 0.38576966524124146, "learning_rate": 9.400251059725762e-05, "loss": 0.0954, "step": 5990 }, { "grad_norm": 0.3726792335510254, "learning_rate": 9.397631050865554e-05, "loss": 0.1004, "step": 6000 }, { "grad_norm": 0.4123550057411194, "learning_rate": 9.395005698468949e-05, "loss": 0.0769, "step": 6010 }, { "grad_norm": 0.3591300845146179, "learning_rate": 9.392375005726e-05, "loss": 0.0756, "step": 6020 }, { "grad_norm": 0.4204219877719879, "learning_rate": 9.389738975833242e-05, "loss": 0.079, "step": 6030 }, { "grad_norm": 0.4564354419708252, "learning_rate": 9.387097611993707e-05, "loss": 0.0802, "step": 6040 }, { "grad_norm": 0.4113357365131378, "learning_rate": 9.384450917416894e-05, "loss": 0.0719, "step": 6050 }, { "grad_norm": 0.3828178644180298, "learning_rate": 9.381798895318792e-05, "loss": 0.0645, "step": 6060 }, { "grad_norm": 0.4908241629600525, "learning_rate": 9.379141548921853e-05, "loss": 0.0751, "step": 6070 }, { "grad_norm": 0.37096697092056274, "learning_rate": 9.376478881455009e-05, "loss": 0.0836, "step": 6080 }, { "grad_norm": 0.31290072202682495, "learning_rate": 9.373810896153647e-05, "loss": 0.0808, "step": 6090 }, { "grad_norm": 0.28079235553741455, "learning_rate": 9.371137596259623e-05, "loss": 0.088, "step": 6100 }, { "grad_norm": 0.37595871090888977, "learning_rate": 9.368458985021248e-05, "loss": 0.061, "step": 6110 }, { "grad_norm": 0.2815130054950714, "learning_rate": 9.365775065693287e-05, "loss": 0.0692, "step": 6120 }, { "grad_norm": 0.3199518918991089, "learning_rate": 9.363085841536957e-05, "loss": 0.0571, "step": 6130 }, { "grad_norm": 0.24395594000816345, "learning_rate": 9.360391315819917e-05, "loss": 0.0689, "step": 6140 }, { "grad_norm": 0.43718135356903076, "learning_rate": 9.357691491816269e-05, "loss": 0.0871, "step": 6150 }, { "grad_norm": 0.42586368322372437, "learning_rate": 9.354986372806557e-05, "loss": 0.0866, "step": 6160 }, { "grad_norm": 0.38560256361961365, "learning_rate": 9.352275962077752e-05, "loss": 0.078, "step": 6170 }, { "grad_norm": 0.373443603515625, "learning_rate": 9.349560262923262e-05, "loss": 0.0612, "step": 6180 }, { "grad_norm": 0.33997249603271484, "learning_rate": 9.346839278642913e-05, "loss": 0.0691, "step": 6190 }, { "grad_norm": 0.372051477432251, "learning_rate": 9.344113012542963e-05, "loss": 0.0584, "step": 6200 }, { "grad_norm": 0.3784361481666565, "learning_rate": 9.341381467936079e-05, "loss": 0.0688, "step": 6210 }, { "grad_norm": 0.31574758887290955, "learning_rate": 9.338644648141346e-05, "loss": 0.0631, "step": 6220 }, { "grad_norm": 0.42290163040161133, "learning_rate": 9.335902556484257e-05, "loss": 0.0699, "step": 6230 }, { "grad_norm": 0.34759193658828735, "learning_rate": 9.333155196296712e-05, "loss": 0.0714, "step": 6240 }, { "grad_norm": 0.4844174385070801, "learning_rate": 9.330402570917016e-05, "loss": 0.0864, "step": 6250 }, { "grad_norm": 0.2976205050945282, "learning_rate": 9.327644683689865e-05, "loss": 0.0557, "step": 6260 }, { "grad_norm": 0.32384932041168213, "learning_rate": 9.324881537966354e-05, "loss": 0.0594, "step": 6270 }, { "grad_norm": 0.36238422989845276, "learning_rate": 9.322113137103964e-05, "loss": 0.061, "step": 6280 }, { "grad_norm": 0.37662220001220703, "learning_rate": 9.319339484466564e-05, "loss": 0.0658, "step": 6290 }, { "grad_norm": 0.28956839442253113, "learning_rate": 9.316560583424403e-05, "loss": 0.058, "step": 6300 }, { "grad_norm": 0.4654834270477295, "learning_rate": 9.313776437354108e-05, "loss": 0.059, "step": 6310 }, { "grad_norm": 0.3382844030857086, "learning_rate": 9.31098704963868e-05, "loss": 0.0595, "step": 6320 }, { "grad_norm": 0.3910624384880066, "learning_rate": 9.308192423667485e-05, "loss": 0.061, "step": 6330 }, { "grad_norm": 0.3296445906162262, "learning_rate": 9.305392562836262e-05, "loss": 0.0758, "step": 6340 }, { "grad_norm": 0.5302577614784241, "learning_rate": 9.3025874705471e-05, "loss": 0.0701, "step": 6350 }, { "grad_norm": 0.4098237156867981, "learning_rate": 9.299777150208457e-05, "loss": 0.0736, "step": 6360 }, { "grad_norm": 0.4289112687110901, "learning_rate": 9.296961605235133e-05, "loss": 0.0672, "step": 6370 }, { "grad_norm": 0.37219858169555664, "learning_rate": 9.29414083904828e-05, "loss": 0.0658, "step": 6380 }, { "grad_norm": 0.4130930006504059, "learning_rate": 9.291314855075397e-05, "loss": 0.0819, "step": 6390 }, { "grad_norm": 0.330931693315506, "learning_rate": 9.288483656750322e-05, "loss": 0.0736, "step": 6400 }, { "grad_norm": 0.3112224042415619, "learning_rate": 9.285647247513225e-05, "loss": 0.055, "step": 6410 }, { "grad_norm": 0.2981925308704376, "learning_rate": 9.282805630810614e-05, "loss": 0.0738, "step": 6420 }, { "grad_norm": 0.3567010462284088, "learning_rate": 9.279958810095317e-05, "loss": 0.0596, "step": 6430 }, { "grad_norm": 0.29627394676208496, "learning_rate": 9.277106788826494e-05, "loss": 0.0654, "step": 6440 }, { "grad_norm": 0.306034654378891, "learning_rate": 9.274249570469618e-05, "loss": 0.0612, "step": 6450 }, { "grad_norm": 0.2855243384838104, "learning_rate": 9.271387158496476e-05, "loss": 0.0517, "step": 6460 }, { "grad_norm": 0.42064598202705383, "learning_rate": 9.268519556385173e-05, "loss": 0.0653, "step": 6470 }, { "grad_norm": 0.2711760997772217, "learning_rate": 9.265646767620113e-05, "loss": 0.0552, "step": 6480 }, { "grad_norm": 0.4397341310977936, "learning_rate": 9.262768795692006e-05, "loss": 0.0768, "step": 6490 }, { "grad_norm": 0.3796547055244446, "learning_rate": 9.259885644097861e-05, "loss": 0.0623, "step": 6500 }, { "grad_norm": 0.3461489975452423, "learning_rate": 9.256997316340976e-05, "loss": 0.0541, "step": 6510 }, { "grad_norm": 0.30998528003692627, "learning_rate": 9.254103815930943e-05, "loss": 0.0595, "step": 6520 }, { "grad_norm": 0.4400295615196228, "learning_rate": 9.251205146383636e-05, "loss": 0.067, "step": 6530 }, { "grad_norm": 0.21478162705898285, "learning_rate": 9.248301311221216e-05, "loss": 0.0652, "step": 6540 }, { "grad_norm": 0.39470869302749634, "learning_rate": 9.245392313972115e-05, "loss": 0.055, "step": 6550 }, { "grad_norm": 0.35615289211273193, "learning_rate": 9.242478158171037e-05, "loss": 0.0641, "step": 6560 }, { "grad_norm": 0.44015946984291077, "learning_rate": 9.239558847358957e-05, "loss": 0.0672, "step": 6570 }, { "grad_norm": 0.2700004577636719, "learning_rate": 9.236634385083114e-05, "loss": 0.05, "step": 6580 }, { "grad_norm": 0.3225199282169342, "learning_rate": 9.233704774897005e-05, "loss": 0.0879, "step": 6590 }, { "grad_norm": 0.339106947183609, "learning_rate": 9.230770020360382e-05, "loss": 0.0672, "step": 6600 }, { "grad_norm": 0.36039599776268005, "learning_rate": 9.227830125039247e-05, "loss": 0.061, "step": 6610 }, { "grad_norm": 0.26372620463371277, "learning_rate": 9.224885092505853e-05, "loss": 0.0605, "step": 6620 }, { "grad_norm": 0.3742723762989044, "learning_rate": 9.221934926338691e-05, "loss": 0.0658, "step": 6630 }, { "grad_norm": 0.36760544776916504, "learning_rate": 9.21897963012249e-05, "loss": 0.0507, "step": 6640 }, { "grad_norm": 0.2720581591129303, "learning_rate": 9.216019207448217e-05, "loss": 0.0756, "step": 6650 }, { "grad_norm": 0.2975386381149292, "learning_rate": 9.21305366191306e-05, "loss": 0.0438, "step": 6660 }, { "grad_norm": 0.2960805296897888, "learning_rate": 9.21008299712044e-05, "loss": 0.0529, "step": 6670 }, { "grad_norm": 0.3475245535373688, "learning_rate": 9.207107216679993e-05, "loss": 0.055, "step": 6680 }, { "grad_norm": 0.2974264621734619, "learning_rate": 9.204126324207574e-05, "loss": 0.0622, "step": 6690 }, { "grad_norm": 0.31373077630996704, "learning_rate": 9.201140323325248e-05, "loss": 0.0559, "step": 6700 }, { "grad_norm": 0.29098671674728394, "learning_rate": 9.198149217661287e-05, "loss": 0.0652, "step": 6710 }, { "grad_norm": 0.3027670383453369, "learning_rate": 9.195153010850167e-05, "loss": 0.0671, "step": 6720 }, { "grad_norm": 0.31832370162010193, "learning_rate": 9.192151706532562e-05, "loss": 0.0586, "step": 6730 }, { "grad_norm": 0.31084969639778137, "learning_rate": 9.189145308355337e-05, "loss": 0.0887, "step": 6740 }, { "grad_norm": 0.25085651874542236, "learning_rate": 9.186133819971555e-05, "loss": 0.0541, "step": 6750 }, { "grad_norm": 0.2884919345378876, "learning_rate": 9.183117245040455e-05, "loss": 0.0683, "step": 6760 }, { "grad_norm": 0.2646777033805847, "learning_rate": 9.180095587227458e-05, "loss": 0.0451, "step": 6770 }, { "grad_norm": 0.31468841433525085, "learning_rate": 9.177068850204167e-05, "loss": 0.073, "step": 6780 }, { "grad_norm": 0.2562132775783539, "learning_rate": 9.174037037648351e-05, "loss": 0.0541, "step": 6790 }, { "grad_norm": 0.3336126208305359, "learning_rate": 9.171000153243947e-05, "loss": 0.0833, "step": 6800 }, { "grad_norm": 0.2712336480617523, "learning_rate": 9.167958200681057e-05, "loss": 0.0584, "step": 6810 }, { "grad_norm": 0.37498190999031067, "learning_rate": 9.164911183655943e-05, "loss": 0.0761, "step": 6820 }, { "grad_norm": 0.29450586438179016, "learning_rate": 9.161859105871013e-05, "loss": 0.0682, "step": 6830 }, { "grad_norm": 0.32402539253234863, "learning_rate": 9.158801971034832e-05, "loss": 0.0846, "step": 6840 }, { "grad_norm": 0.266650527715683, "learning_rate": 9.155739782862107e-05, "loss": 0.0713, "step": 6850 }, { "grad_norm": 0.3357616662979126, "learning_rate": 9.152672545073687e-05, "loss": 0.062, "step": 6860 }, { "grad_norm": 0.3223040997982025, "learning_rate": 9.149600261396551e-05, "loss": 0.0766, "step": 6870 }, { "grad_norm": 0.4206574261188507, "learning_rate": 9.146522935563817e-05, "loss": 0.075, "step": 6880 }, { "grad_norm": 0.30879107117652893, "learning_rate": 9.143440571314723e-05, "loss": 0.0761, "step": 6890 }, { "grad_norm": 0.40953752398490906, "learning_rate": 9.140353172394636e-05, "loss": 0.086, "step": 6900 }, { "grad_norm": 0.36636731028556824, "learning_rate": 9.137260742555032e-05, "loss": 0.0626, "step": 6910 }, { "grad_norm": 0.26533272862434387, "learning_rate": 9.13416328555351e-05, "loss": 0.075, "step": 6920 }, { "grad_norm": 0.28045469522476196, "learning_rate": 9.131060805153769e-05, "loss": 0.0786, "step": 6930 }, { "grad_norm": 0.27520087361335754, "learning_rate": 9.127953305125618e-05, "loss": 0.0596, "step": 6940 }, { "grad_norm": 0.257928729057312, "learning_rate": 9.124840789244958e-05, "loss": 0.0596, "step": 6950 }, { "grad_norm": 0.25661271810531616, "learning_rate": 9.121723261293794e-05, "loss": 0.0696, "step": 6960 }, { "grad_norm": 0.35159793496131897, "learning_rate": 9.118600725060214e-05, "loss": 0.0727, "step": 6970 }, { "grad_norm": 0.4414897859096527, "learning_rate": 9.115473184338392e-05, "loss": 0.0776, "step": 6980 }, { "grad_norm": 0.2959267497062683, "learning_rate": 9.11234064292859e-05, "loss": 0.0568, "step": 6990 }, { "grad_norm": 0.3596073389053345, "learning_rate": 9.109203104637137e-05, "loss": 0.072, "step": 7000 }, { "grad_norm": 0.32795876264572144, "learning_rate": 9.10606057327644e-05, "loss": 0.0781, "step": 7010 }, { "grad_norm": 0.3796412944793701, "learning_rate": 9.10291305266497e-05, "loss": 0.0738, "step": 7020 }, { "grad_norm": 0.32179585099220276, "learning_rate": 9.099760546627261e-05, "loss": 0.0912, "step": 7030 }, { "grad_norm": 0.3382680416107178, "learning_rate": 9.096603058993907e-05, "loss": 0.0725, "step": 7040 }, { "grad_norm": 0.3623231053352356, "learning_rate": 9.093440593601552e-05, "loss": 0.0585, "step": 7050 }, { "grad_norm": 0.30419784784317017, "learning_rate": 9.090273154292889e-05, "loss": 0.0647, "step": 7060 }, { "grad_norm": 0.27348974347114563, "learning_rate": 9.087100744916656e-05, "loss": 0.0742, "step": 7070 }, { "grad_norm": 0.3035682141780853, "learning_rate": 9.08392336932763e-05, "loss": 0.0912, "step": 7080 }, { "grad_norm": 0.3073776662349701, "learning_rate": 9.080741031386619e-05, "loss": 0.071, "step": 7090 }, { "grad_norm": 0.3244733512401581, "learning_rate": 9.077553734960468e-05, "loss": 0.0693, "step": 7100 }, { "grad_norm": 0.3421766757965088, "learning_rate": 9.074361483922041e-05, "loss": 0.0873, "step": 7110 }, { "grad_norm": 0.46409210562705994, "learning_rate": 9.071164282150224e-05, "loss": 0.0712, "step": 7120 }, { "grad_norm": 0.295052170753479, "learning_rate": 9.067962133529919e-05, "loss": 0.0693, "step": 7130 }, { "grad_norm": 0.2866746783256531, "learning_rate": 9.064755041952036e-05, "loss": 0.0927, "step": 7140 }, { "grad_norm": 0.2507902979850769, "learning_rate": 9.061543011313497e-05, "loss": 0.0886, "step": 7150 }, { "grad_norm": 0.4562154710292816, "learning_rate": 9.05832604551722e-05, "loss": 0.0951, "step": 7160 }, { "grad_norm": 0.29059094190597534, "learning_rate": 9.055104148472124e-05, "loss": 0.0784, "step": 7170 }, { "grad_norm": 0.32866334915161133, "learning_rate": 9.051877324093114e-05, "loss": 0.0752, "step": 7180 }, { "grad_norm": 0.2248542308807373, "learning_rate": 9.04864557630109e-05, "loss": 0.0752, "step": 7190 }, { "grad_norm": 0.3837790787220001, "learning_rate": 9.045408909022928e-05, "loss": 0.0807, "step": 7200 }, { "grad_norm": 0.26797232031822205, "learning_rate": 9.042167326191484e-05, "loss": 0.0714, "step": 7210 }, { "grad_norm": 0.30856582522392273, "learning_rate": 9.038920831745587e-05, "loss": 0.0879, "step": 7220 }, { "grad_norm": 0.342583030462265, "learning_rate": 9.035669429630036e-05, "loss": 0.0742, "step": 7230 }, { "grad_norm": 0.3736778497695923, "learning_rate": 9.032413123795588e-05, "loss": 0.0746, "step": 7240 }, { "grad_norm": 0.3080156743526459, "learning_rate": 9.029151918198962e-05, "loss": 0.0839, "step": 7250 }, { "grad_norm": 0.4206292927265167, "learning_rate": 9.025885816802833e-05, "loss": 0.0733, "step": 7260 }, { "grad_norm": 0.29639485478401184, "learning_rate": 9.02261482357582e-05, "loss": 0.0737, "step": 7270 }, { "grad_norm": 0.3351867198944092, "learning_rate": 9.019338942492485e-05, "loss": 0.0682, "step": 7280 }, { "grad_norm": 0.37301281094551086, "learning_rate": 9.01605817753334e-05, "loss": 0.0794, "step": 7290 }, { "grad_norm": 0.3460105061531067, "learning_rate": 9.012772532684818e-05, "loss": 0.0767, "step": 7300 }, { "grad_norm": 0.35435643792152405, "learning_rate": 9.00948201193929e-05, "loss": 0.0831, "step": 7310 }, { "grad_norm": 0.3447599411010742, "learning_rate": 9.006186619295047e-05, "loss": 0.0622, "step": 7320 }, { "grad_norm": 0.23698163032531738, "learning_rate": 9.002886358756305e-05, "loss": 0.0705, "step": 7330 }, { "grad_norm": 0.41723495721817017, "learning_rate": 8.999581234333188e-05, "loss": 0.0776, "step": 7340 }, { "grad_norm": 0.29103609919548035, "learning_rate": 8.996271250041734e-05, "loss": 0.0644, "step": 7350 }, { "grad_norm": 0.36110755801200867, "learning_rate": 8.99295640990389e-05, "loss": 0.0734, "step": 7360 }, { "grad_norm": 0.25180578231811523, "learning_rate": 8.989636717947494e-05, "loss": 0.0801, "step": 7370 }, { "grad_norm": 0.35290178656578064, "learning_rate": 8.986312178206291e-05, "loss": 0.0708, "step": 7380 }, { "grad_norm": 0.3062017261981964, "learning_rate": 8.982982794719904e-05, "loss": 0.0759, "step": 7390 }, { "grad_norm": 0.3984578549861908, "learning_rate": 8.979648571533852e-05, "loss": 0.0762, "step": 7400 }, { "grad_norm": 0.32276326417922974, "learning_rate": 8.976309512699528e-05, "loss": 0.0823, "step": 7410 }, { "grad_norm": 0.38441768288612366, "learning_rate": 8.972965622274205e-05, "loss": 0.1022, "step": 7420 }, { "grad_norm": 0.4468815326690674, "learning_rate": 8.969616904321026e-05, "loss": 0.0928, "step": 7430 }, { "grad_norm": 0.37163904309272766, "learning_rate": 8.966263362908998e-05, "loss": 0.0809, "step": 7440 }, { "grad_norm": 0.37135565280914307, "learning_rate": 8.962905002112989e-05, "loss": 0.0757, "step": 7450 }, { "grad_norm": 0.37873753905296326, "learning_rate": 8.959541826013724e-05, "loss": 0.0766, "step": 7460 }, { "grad_norm": 0.32542404532432556, "learning_rate": 8.956173838697781e-05, "loss": 0.0853, "step": 7470 }, { "grad_norm": 0.36785489320755005, "learning_rate": 8.952801044257581e-05, "loss": 0.0917, "step": 7480 }, { "grad_norm": 0.325432151556015, "learning_rate": 8.949423446791387e-05, "loss": 0.0901, "step": 7490 }, { "grad_norm": 0.35331645607948303, "learning_rate": 8.9460410504033e-05, "loss": 0.0715, "step": 7500 }, { "grad_norm": 0.33233383297920227, "learning_rate": 8.942653859203248e-05, "loss": 0.0661, "step": 7510 }, { "grad_norm": 0.24059639871120453, "learning_rate": 8.93926187730699e-05, "loss": 0.0658, "step": 7520 }, { "grad_norm": 0.3192104399204254, "learning_rate": 8.935865108836102e-05, "loss": 0.0639, "step": 7530 }, { "grad_norm": 0.29839953780174255, "learning_rate": 8.932463557917981e-05, "loss": 0.0692, "step": 7540 }, { "grad_norm": 0.38585853576660156, "learning_rate": 8.929057228685829e-05, "loss": 0.0732, "step": 7550 }, { "grad_norm": 0.40978941321372986, "learning_rate": 8.925646125278658e-05, "loss": 0.0648, "step": 7560 }, { "grad_norm": 0.3373661935329437, "learning_rate": 8.92223025184128e-05, "loss": 0.0772, "step": 7570 }, { "grad_norm": 0.3103821873664856, "learning_rate": 8.918809612524303e-05, "loss": 0.0745, "step": 7580 }, { "grad_norm": 0.39536982774734497, "learning_rate": 8.91538421148413e-05, "loss": 0.0738, "step": 7590 }, { "grad_norm": 0.24737860262393951, "learning_rate": 8.91195405288294e-05, "loss": 0.0723, "step": 7600 }, { "grad_norm": 0.30484825372695923, "learning_rate": 8.908519140888705e-05, "loss": 0.0792, "step": 7610 }, { "grad_norm": 0.26080337166786194, "learning_rate": 8.905079479675164e-05, "loss": 0.0671, "step": 7620 }, { "grad_norm": 0.26961565017700195, "learning_rate": 8.901635073421831e-05, "loss": 0.0779, "step": 7630 }, { "grad_norm": 0.5087644457817078, "learning_rate": 8.898185926313983e-05, "loss": 0.0628, "step": 7640 }, { "grad_norm": 0.3164035379886627, "learning_rate": 8.894732042542659e-05, "loss": 0.0651, "step": 7650 }, { "grad_norm": 0.3038000464439392, "learning_rate": 8.891273426304654e-05, "loss": 0.0628, "step": 7660 }, { "grad_norm": 0.3564732074737549, "learning_rate": 8.887810081802515e-05, "loss": 0.0733, "step": 7670 }, { "grad_norm": 0.3813934028148651, "learning_rate": 8.884342013244529e-05, "loss": 0.0654, "step": 7680 }, { "grad_norm": 0.3492625951766968, "learning_rate": 8.880869224844725e-05, "loss": 0.0683, "step": 7690 }, { "grad_norm": 0.3102383613586426, "learning_rate": 8.877391720822875e-05, "loss": 0.0499, "step": 7700 }, { "grad_norm": 0.2606445550918579, "learning_rate": 8.873909505404468e-05, "loss": 0.0849, "step": 7710 }, { "grad_norm": 0.34573808312416077, "learning_rate": 8.870422582820726e-05, "loss": 0.0669, "step": 7720 }, { "grad_norm": 0.29785794019699097, "learning_rate": 8.866930957308588e-05, "loss": 0.0805, "step": 7730 }, { "grad_norm": 0.32044291496276855, "learning_rate": 8.86343463311071e-05, "loss": 0.0736, "step": 7740 }, { "grad_norm": 0.3109455704689026, "learning_rate": 8.859933614475452e-05, "loss": 0.0625, "step": 7750 }, { "grad_norm": 0.279070109128952, "learning_rate": 8.85642790565689e-05, "loss": 0.0719, "step": 7760 }, { "grad_norm": 0.3653663992881775, "learning_rate": 8.852917510914782e-05, "loss": 0.0704, "step": 7770 }, { "grad_norm": 0.4109472930431366, "learning_rate": 8.84940243451459e-05, "loss": 0.0718, "step": 7780 }, { "grad_norm": 0.2957683503627777, "learning_rate": 8.845882680727469e-05, "loss": 0.0751, "step": 7790 }, { "grad_norm": 0.29171738028526306, "learning_rate": 8.842358253830246e-05, "loss": 0.0534, "step": 7800 }, { "grad_norm": 0.29225465655326843, "learning_rate": 8.838829158105434e-05, "loss": 0.0646, "step": 7810 }, { "grad_norm": 0.2870543301105499, "learning_rate": 8.835295397841217e-05, "loss": 0.0437, "step": 7820 }, { "grad_norm": 0.3388708531856537, "learning_rate": 8.831756977331447e-05, "loss": 0.067, "step": 7830 }, { "grad_norm": 0.3022441864013672, "learning_rate": 8.828213900875638e-05, "loss": 0.0578, "step": 7840 }, { "grad_norm": 0.31380876898765564, "learning_rate": 8.824666172778964e-05, "loss": 0.0675, "step": 7850 }, { "grad_norm": 0.3840019702911377, "learning_rate": 8.821113797352245e-05, "loss": 0.068, "step": 7860 }, { "grad_norm": 0.3927735686302185, "learning_rate": 8.817556778911956e-05, "loss": 0.0766, "step": 7870 }, { "grad_norm": 0.3057989478111267, "learning_rate": 8.81399512178021e-05, "loss": 0.0619, "step": 7880 }, { "grad_norm": 0.2815624475479126, "learning_rate": 8.810428830284752e-05, "loss": 0.0601, "step": 7890 }, { "grad_norm": 0.3628629744052887, "learning_rate": 8.806857908758967e-05, "loss": 0.0707, "step": 7900 }, { "grad_norm": 0.37212055921554565, "learning_rate": 8.80328236154186e-05, "loss": 0.0816, "step": 7910 }, { "grad_norm": 0.23834164440631866, "learning_rate": 8.799702192978056e-05, "loss": 0.0788, "step": 7920 }, { "grad_norm": 0.3322717845439911, "learning_rate": 8.7961174074178e-05, "loss": 0.0598, "step": 7930 }, { "grad_norm": 0.24324537813663483, "learning_rate": 8.792528009216942e-05, "loss": 0.058, "step": 7940 }, { "grad_norm": 0.40193113684654236, "learning_rate": 8.788934002736943e-05, "loss": 0.0695, "step": 7950 }, { "grad_norm": 0.35611966252326965, "learning_rate": 8.785335392344857e-05, "loss": 0.0546, "step": 7960 }, { "grad_norm": 0.3150451183319092, "learning_rate": 8.781732182413336e-05, "loss": 0.062, "step": 7970 }, { "grad_norm": 0.3062570095062256, "learning_rate": 8.778124377320618e-05, "loss": 0.0655, "step": 7980 }, { "grad_norm": 0.3221885859966278, "learning_rate": 8.774511981450529e-05, "loss": 0.0564, "step": 7990 }, { "grad_norm": 0.31379520893096924, "learning_rate": 8.770894999192467e-05, "loss": 0.052, "step": 8000 }, { "grad_norm": 0.3399493098258972, "learning_rate": 8.767273434941412e-05, "loss": 0.0562, "step": 8010 }, { "grad_norm": 0.22517575323581696, "learning_rate": 8.7636472930979e-05, "loss": 0.059, "step": 8020 }, { "grad_norm": 0.31971481442451477, "learning_rate": 8.76001657806804e-05, "loss": 0.0617, "step": 8030 }, { "grad_norm": 0.33451130986213684, "learning_rate": 8.756381294263491e-05, "loss": 0.0525, "step": 8040 }, { "grad_norm": 0.2606923282146454, "learning_rate": 8.752741446101464e-05, "loss": 0.0685, "step": 8050 }, { "grad_norm": 0.17768029868602753, "learning_rate": 8.749097038004722e-05, "loss": 0.0491, "step": 8060 }, { "grad_norm": 0.3321302831172943, "learning_rate": 8.745448074401561e-05, "loss": 0.0571, "step": 8070 }, { "grad_norm": 0.32249870896339417, "learning_rate": 8.741794559725818e-05, "loss": 0.0745, "step": 8080 }, { "grad_norm": 0.34119081497192383, "learning_rate": 8.738136498416856e-05, "loss": 0.067, "step": 8090 }, { "grad_norm": 0.27629217505455017, "learning_rate": 8.734473894919564e-05, "loss": 0.0667, "step": 8100 }, { "grad_norm": 0.3690558671951294, "learning_rate": 8.730806753684353e-05, "loss": 0.0733, "step": 8110 }, { "grad_norm": 0.27195847034454346, "learning_rate": 8.727135079167144e-05, "loss": 0.0617, "step": 8120 }, { "grad_norm": 0.3272133469581604, "learning_rate": 8.723458875829368e-05, "loss": 0.0756, "step": 8130 }, { "grad_norm": 0.2452179193496704, "learning_rate": 8.719778148137958e-05, "loss": 0.0565, "step": 8140 }, { "grad_norm": 0.30834993720054626, "learning_rate": 8.716092900565346e-05, "loss": 0.0509, "step": 8150 }, { "grad_norm": 0.2717835009098053, "learning_rate": 8.712403137589456e-05, "loss": 0.0537, "step": 8160 }, { "grad_norm": 0.30055272579193115, "learning_rate": 8.708708863693697e-05, "loss": 0.0571, "step": 8170 }, { "grad_norm": 0.18932680785655975, "learning_rate": 8.70501008336696e-05, "loss": 0.046, "step": 8180 }, { "grad_norm": 0.2999027967453003, "learning_rate": 8.701306801103611e-05, "loss": 0.0577, "step": 8190 }, { "grad_norm": 0.28788453340530396, "learning_rate": 8.697599021403489e-05, "loss": 0.0431, "step": 8200 }, { "grad_norm": 0.270668089389801, "learning_rate": 8.693886748771896e-05, "loss": 0.0439, "step": 8210 }, { "grad_norm": 0.2906114459037781, "learning_rate": 8.690169987719593e-05, "loss": 0.0465, "step": 8220 }, { "grad_norm": 0.3025170564651489, "learning_rate": 8.686448742762791e-05, "loss": 0.0566, "step": 8230 }, { "grad_norm": 0.17975570261478424, "learning_rate": 8.682723018423161e-05, "loss": 0.0516, "step": 8240 }, { "grad_norm": 0.24628978967666626, "learning_rate": 8.678992819227804e-05, "loss": 0.0585, "step": 8250 }, { "grad_norm": 0.3163599967956543, "learning_rate": 8.675258149709265e-05, "loss": 0.0468, "step": 8260 }, { "grad_norm": 0.2964513599872589, "learning_rate": 8.67151901440552e-05, "loss": 0.0558, "step": 8270 }, { "grad_norm": 0.4975317716598511, "learning_rate": 8.667775417859971e-05, "loss": 0.0523, "step": 8280 }, { "grad_norm": 0.24812600016593933, "learning_rate": 8.664027364621441e-05, "loss": 0.0421, "step": 8290 }, { "grad_norm": 0.2597090005874634, "learning_rate": 8.660274859244168e-05, "loss": 0.0542, "step": 8300 }, { "grad_norm": 0.3056149184703827, "learning_rate": 8.656517906287799e-05, "loss": 0.0619, "step": 8310 }, { "grad_norm": 0.24847209453582764, "learning_rate": 8.652756510317387e-05, "loss": 0.0407, "step": 8320 }, { "grad_norm": 0.21093441545963287, "learning_rate": 8.648990675903382e-05, "loss": 0.0577, "step": 8330 }, { "grad_norm": 0.278901606798172, "learning_rate": 8.64522040762163e-05, "loss": 0.0433, "step": 8340 }, { "grad_norm": 0.2529560327529907, "learning_rate": 8.64144571005336e-05, "loss": 0.0612, "step": 8350 }, { "grad_norm": 0.23818399012088776, "learning_rate": 8.637666587785184e-05, "loss": 0.0472, "step": 8360 }, { "grad_norm": 0.3120236396789551, "learning_rate": 8.633883045409096e-05, "loss": 0.0459, "step": 8370 }, { "grad_norm": 0.2247028797864914, "learning_rate": 8.630095087522457e-05, "loss": 0.0483, "step": 8380 }, { "grad_norm": 0.39260390400886536, "learning_rate": 8.62630271872799e-05, "loss": 0.0496, "step": 8390 }, { "grad_norm": 0.2796708345413208, "learning_rate": 8.62250594363378e-05, "loss": 0.0495, "step": 8400 }, { "grad_norm": 0.2618634104728699, "learning_rate": 8.61870476685327e-05, "loss": 0.0548, "step": 8410 }, { "grad_norm": 0.250218003988266, "learning_rate": 8.614899193005248e-05, "loss": 0.0465, "step": 8420 }, { "grad_norm": 0.34483805298805237, "learning_rate": 8.611089226713842e-05, "loss": 0.0539, "step": 8430 }, { "grad_norm": 0.3173517882823944, "learning_rate": 8.607274872608521e-05, "loss": 0.0516, "step": 8440 }, { "grad_norm": 0.29788267612457275, "learning_rate": 8.603456135324088e-05, "loss": 0.0517, "step": 8450 }, { "grad_norm": 0.249750018119812, "learning_rate": 8.599633019500665e-05, "loss": 0.0492, "step": 8460 }, { "grad_norm": 0.32733964920043945, "learning_rate": 8.595805529783702e-05, "loss": 0.0528, "step": 8470 }, { "grad_norm": 0.25852009654045105, "learning_rate": 8.591973670823958e-05, "loss": 0.0457, "step": 8480 }, { "grad_norm": 0.3136705160140991, "learning_rate": 8.588137447277503e-05, "loss": 0.0654, "step": 8490 }, { "grad_norm": 0.2908875346183777, "learning_rate": 8.584296863805709e-05, "loss": 0.0726, "step": 8500 }, { "grad_norm": 0.3555467128753662, "learning_rate": 8.580451925075248e-05, "loss": 0.0569, "step": 8510 }, { "grad_norm": 0.2341107875108719, "learning_rate": 8.576602635758085e-05, "loss": 0.0378, "step": 8520 }, { "grad_norm": 0.24696886539459229, "learning_rate": 8.572749000531469e-05, "loss": 0.0507, "step": 8530 }, { "grad_norm": 0.2810973525047302, "learning_rate": 8.568891024077926e-05, "loss": 0.0574, "step": 8540 }, { "grad_norm": 0.24124041199684143, "learning_rate": 8.565028711085265e-05, "loss": 0.0553, "step": 8550 }, { "grad_norm": 0.27259981632232666, "learning_rate": 8.561162066246561e-05, "loss": 0.0514, "step": 8560 }, { "grad_norm": 0.3204583525657654, "learning_rate": 8.557291094260152e-05, "loss": 0.0864, "step": 8570 }, { "grad_norm": 0.3560045063495636, "learning_rate": 8.553415799829629e-05, "loss": 0.0709, "step": 8580 }, { "grad_norm": 0.27009761333465576, "learning_rate": 8.549536187663847e-05, "loss": 0.0604, "step": 8590 }, { "grad_norm": 0.3344092071056366, "learning_rate": 8.545652262476897e-05, "loss": 0.0432, "step": 8600 }, { "grad_norm": 0.3227461278438568, "learning_rate": 8.541764028988115e-05, "loss": 0.0476, "step": 8610 }, { "grad_norm": 0.2818009853363037, "learning_rate": 8.537871491922071e-05, "loss": 0.0541, "step": 8620 }, { "grad_norm": 0.26765769720077515, "learning_rate": 8.533974656008566e-05, "loss": 0.04, "step": 8630 }, { "grad_norm": 0.35619986057281494, "learning_rate": 8.530073525982621e-05, "loss": 0.0593, "step": 8640 }, { "grad_norm": 0.2339169681072235, "learning_rate": 8.526168106584476e-05, "loss": 0.0617, "step": 8650 }, { "grad_norm": 0.2824109196662903, "learning_rate": 8.522258402559586e-05, "loss": 0.0469, "step": 8660 }, { "grad_norm": 0.19321531057357788, "learning_rate": 8.518344418658613e-05, "loss": 0.0391, "step": 8670 }, { "grad_norm": 0.2509182393550873, "learning_rate": 8.51442615963741e-05, "loss": 0.0353, "step": 8680 }, { "grad_norm": 0.25188514590263367, "learning_rate": 8.510503630257034e-05, "loss": 0.0527, "step": 8690 }, { "grad_norm": 0.31468215584754944, "learning_rate": 8.50657683528373e-05, "loss": 0.0382, "step": 8700 }, { "grad_norm": 0.26425349712371826, "learning_rate": 8.502645779488922e-05, "loss": 0.0615, "step": 8710 }, { "grad_norm": 0.34671303629875183, "learning_rate": 8.498710467649213e-05, "loss": 0.0761, "step": 8720 }, { "grad_norm": 0.21573609113693237, "learning_rate": 8.494770904546381e-05, "loss": 0.0497, "step": 8730 }, { "grad_norm": 0.18519026041030884, "learning_rate": 8.490827094967363e-05, "loss": 0.0514, "step": 8740 }, { "grad_norm": 0.2984887659549713, "learning_rate": 8.486879043704263e-05, "loss": 0.0656, "step": 8750 }, { "grad_norm": 0.2905673384666443, "learning_rate": 8.482926755554332e-05, "loss": 0.0516, "step": 8760 }, { "grad_norm": 0.3386414349079132, "learning_rate": 8.478970235319975e-05, "loss": 0.049, "step": 8770 }, { "grad_norm": 0.3081241548061371, "learning_rate": 8.475009487808737e-05, "loss": 0.0603, "step": 8780 }, { "grad_norm": 0.28692418336868286, "learning_rate": 8.471044517833298e-05, "loss": 0.0416, "step": 8790 }, { "grad_norm": 0.24351166188716888, "learning_rate": 8.467075330211474e-05, "loss": 0.043, "step": 8800 }, { "grad_norm": 0.29312369227409363, "learning_rate": 8.463101929766198e-05, "loss": 0.0503, "step": 8810 }, { "grad_norm": 0.3101311922073364, "learning_rate": 8.459124321325528e-05, "loss": 0.0471, "step": 8820 }, { "grad_norm": 0.3281428813934326, "learning_rate": 8.455142509722634e-05, "loss": 0.0617, "step": 8830 }, { "grad_norm": 0.3127662241458893, "learning_rate": 8.451156499795792e-05, "loss": 0.0419, "step": 8840 }, { "grad_norm": 0.2770302891731262, "learning_rate": 8.447166296388381e-05, "loss": 0.0454, "step": 8850 }, { "grad_norm": 0.24282479286193848, "learning_rate": 8.443171904348872e-05, "loss": 0.0423, "step": 8860 }, { "grad_norm": 0.322608083486557, "learning_rate": 8.43917332853083e-05, "loss": 0.051, "step": 8870 }, { "grad_norm": 0.22199705243110657, "learning_rate": 8.435170573792901e-05, "loss": 0.0513, "step": 8880 }, { "grad_norm": 0.20761196315288544, "learning_rate": 8.431163644998808e-05, "loss": 0.055, "step": 8890 }, { "grad_norm": 0.2231435775756836, "learning_rate": 8.42715254701735e-05, "loss": 0.0463, "step": 8900 }, { "grad_norm": 0.2528821527957916, "learning_rate": 8.42313728472239e-05, "loss": 0.0501, "step": 8910 }, { "grad_norm": 0.3470785617828369, "learning_rate": 8.419117862992844e-05, "loss": 0.0561, "step": 8920 }, { "grad_norm": 0.26537367701530457, "learning_rate": 8.415094286712694e-05, "loss": 0.0482, "step": 8930 }, { "grad_norm": 0.17023199796676636, "learning_rate": 8.411066560770965e-05, "loss": 0.0653, "step": 8940 }, { "grad_norm": 0.25650760531425476, "learning_rate": 8.407034690061721e-05, "loss": 0.0521, "step": 8950 }, { "grad_norm": 0.37079691886901855, "learning_rate": 8.402998679484067e-05, "loss": 0.0496, "step": 8960 }, { "grad_norm": 0.24631409347057343, "learning_rate": 8.398958533942135e-05, "loss": 0.0528, "step": 8970 }, { "grad_norm": 0.3600345253944397, "learning_rate": 8.394914258345083e-05, "loss": 0.0595, "step": 8980 }, { "grad_norm": 0.2683735191822052, "learning_rate": 8.390865857607089e-05, "loss": 0.084, "step": 8990 }, { "grad_norm": 0.38127171993255615, "learning_rate": 8.386813336647339e-05, "loss": 0.0681, "step": 9000 }, { "grad_norm": 0.47200074791908264, "learning_rate": 8.38275670039003e-05, "loss": 0.0503, "step": 9010 }, { "grad_norm": 0.3373412787914276, "learning_rate": 8.378695953764356e-05, "loss": 0.0559, "step": 9020 }, { "grad_norm": 0.26795703172683716, "learning_rate": 8.374631101704509e-05, "loss": 0.0708, "step": 9030 }, { "grad_norm": 0.3614940941333771, "learning_rate": 8.370562149149665e-05, "loss": 0.0555, "step": 9040 }, { "grad_norm": 0.3005194664001465, "learning_rate": 8.366489101043989e-05, "loss": 0.0524, "step": 9050 }, { "grad_norm": 0.3441217243671417, "learning_rate": 8.362411962336614e-05, "loss": 0.0657, "step": 9060 }, { "grad_norm": 0.2366398423910141, "learning_rate": 8.358330737981651e-05, "loss": 0.0377, "step": 9070 }, { "grad_norm": 0.35372400283813477, "learning_rate": 8.35424543293817e-05, "loss": 0.0451, "step": 9080 }, { "grad_norm": 0.23972487449645996, "learning_rate": 8.350156052170206e-05, "loss": 0.0549, "step": 9090 }, { "grad_norm": 0.3042377531528473, "learning_rate": 8.346062600646738e-05, "loss": 0.0658, "step": 9100 }, { "grad_norm": 0.2550870478153229, "learning_rate": 8.341965083341697e-05, "loss": 0.0483, "step": 9110 }, { "grad_norm": 0.32105371356010437, "learning_rate": 8.337863505233953e-05, "loss": 0.0584, "step": 9120 }, { "grad_norm": 0.35311374068260193, "learning_rate": 8.33375787130731e-05, "loss": 0.0559, "step": 9130 }, { "grad_norm": 0.31463316082954407, "learning_rate": 8.3296481865505e-05, "loss": 0.0478, "step": 9140 }, { "grad_norm": 0.3628855049610138, "learning_rate": 8.325534455957181e-05, "loss": 0.0647, "step": 9150 }, { "grad_norm": 0.27999362349510193, "learning_rate": 8.321416684525917e-05, "loss": 0.0464, "step": 9160 }, { "grad_norm": 0.21482115983963013, "learning_rate": 8.317294877260193e-05, "loss": 0.0449, "step": 9170 }, { "grad_norm": 0.2623753845691681, "learning_rate": 8.313169039168394e-05, "loss": 0.0491, "step": 9180 }, { "grad_norm": 0.32760393619537354, "learning_rate": 8.3090391752638e-05, "loss": 0.0406, "step": 9190 }, { "grad_norm": 0.24143631756305695, "learning_rate": 8.304905290564586e-05, "loss": 0.0527, "step": 9200 }, { "grad_norm": 0.25226637721061707, "learning_rate": 8.300767390093813e-05, "loss": 0.0662, "step": 9210 }, { "grad_norm": 0.2566879093647003, "learning_rate": 8.296625478879417e-05, "loss": 0.061, "step": 9220 }, { "grad_norm": 0.277544230222702, "learning_rate": 8.292479561954214e-05, "loss": 0.0511, "step": 9230 }, { "grad_norm": 0.2880678176879883, "learning_rate": 8.288329644355883e-05, "loss": 0.0631, "step": 9240 }, { "grad_norm": 0.3591022789478302, "learning_rate": 8.284175731126965e-05, "loss": 0.0479, "step": 9250 }, { "grad_norm": 0.40673813223838806, "learning_rate": 8.280017827314854e-05, "loss": 0.0474, "step": 9260 }, { "grad_norm": 0.2732219696044922, "learning_rate": 8.2758559379718e-05, "loss": 0.058, "step": 9270 }, { "grad_norm": 0.3124210834503174, "learning_rate": 8.271690068154887e-05, "loss": 0.0465, "step": 9280 }, { "grad_norm": 0.29425087571144104, "learning_rate": 8.26752022292604e-05, "loss": 0.0474, "step": 9290 }, { "grad_norm": 0.2728716731071472, "learning_rate": 8.263346407352018e-05, "loss": 0.0591, "step": 9300 }, { "grad_norm": 0.2973133623600006, "learning_rate": 8.259168626504395e-05, "loss": 0.046, "step": 9310 }, { "grad_norm": 0.2720426619052887, "learning_rate": 8.254986885459572e-05, "loss": 0.0567, "step": 9320 }, { "grad_norm": 0.3335602879524231, "learning_rate": 8.250801189298757e-05, "loss": 0.0504, "step": 9330 }, { "grad_norm": 0.29516732692718506, "learning_rate": 8.246611543107967e-05, "loss": 0.0513, "step": 9340 }, { "grad_norm": 0.3028777539730072, "learning_rate": 8.242417951978018e-05, "loss": 0.0453, "step": 9350 }, { "grad_norm": 0.2826080918312073, "learning_rate": 8.238220421004517e-05, "loss": 0.0491, "step": 9360 }, { "grad_norm": 0.27067768573760986, "learning_rate": 8.234018955287859e-05, "loss": 0.0558, "step": 9370 }, { "grad_norm": 0.2205626666545868, "learning_rate": 8.229813559933224e-05, "loss": 0.053, "step": 9380 }, { "grad_norm": 0.2743198275566101, "learning_rate": 8.22560424005056e-05, "loss": 0.0468, "step": 9390 }, { "grad_norm": 0.38924816250801086, "learning_rate": 8.22139100075459e-05, "loss": 0.0596, "step": 9400 }, { "grad_norm": 0.24847157299518585, "learning_rate": 8.217173847164799e-05, "loss": 0.055, "step": 9410 }, { "grad_norm": 0.3238399922847748, "learning_rate": 8.212952784405422e-05, "loss": 0.0491, "step": 9420 }, { "grad_norm": 0.2075970619916916, "learning_rate": 8.208727817605451e-05, "loss": 0.0502, "step": 9430 }, { "grad_norm": 0.3121325373649597, "learning_rate": 8.204498951898618e-05, "loss": 0.0528, "step": 9440 }, { "grad_norm": 0.29418063163757324, "learning_rate": 8.200266192423395e-05, "loss": 0.0434, "step": 9450 }, { "grad_norm": 0.22110435366630554, "learning_rate": 8.196029544322981e-05, "loss": 0.0403, "step": 9460 }, { "grad_norm": 0.38335055112838745, "learning_rate": 8.191789012745305e-05, "loss": 0.0621, "step": 9470 }, { "grad_norm": 0.2737831175327301, "learning_rate": 8.187544602843014e-05, "loss": 0.0492, "step": 9480 }, { "grad_norm": 0.3939169943332672, "learning_rate": 8.183296319773466e-05, "loss": 0.0469, "step": 9490 }, { "grad_norm": 0.2538835108280182, "learning_rate": 8.179044168698721e-05, "loss": 0.04, "step": 9500 }, { "grad_norm": 0.263415664434433, "learning_rate": 8.174788154785548e-05, "loss": 0.0462, "step": 9510 }, { "grad_norm": 0.24150384962558746, "learning_rate": 8.170528283205403e-05, "loss": 0.0583, "step": 9520 }, { "grad_norm": 0.3282778263092041, "learning_rate": 8.166264559134435e-05, "loss": 0.0578, "step": 9530 }, { "grad_norm": 0.3509140610694885, "learning_rate": 8.161996987753467e-05, "loss": 0.0667, "step": 9540 }, { "grad_norm": 0.24726663529872894, "learning_rate": 8.157725574248e-05, "loss": 0.054, "step": 9550 }, { "grad_norm": 0.309364914894104, "learning_rate": 8.153450323808209e-05, "loss": 0.0672, "step": 9560 }, { "grad_norm": 0.37005293369293213, "learning_rate": 8.149171241628923e-05, "loss": 0.0704, "step": 9570 }, { "grad_norm": 0.3555232286453247, "learning_rate": 8.144888332909631e-05, "loss": 0.0645, "step": 9580 }, { "grad_norm": 0.3315240442752838, "learning_rate": 8.14060160285447e-05, "loss": 0.0646, "step": 9590 }, { "grad_norm": 0.317145973443985, "learning_rate": 8.136311056672223e-05, "loss": 0.0542, "step": 9600 }, { "grad_norm": 0.32773029804229736, "learning_rate": 8.132016699576308e-05, "loss": 0.0514, "step": 9610 }, { "grad_norm": 0.270693838596344, "learning_rate": 8.127718536784771e-05, "loss": 0.0573, "step": 9620 }, { "grad_norm": 0.2615382969379425, "learning_rate": 8.12341657352029e-05, "loss": 0.048, "step": 9630 }, { "grad_norm": 0.30986103415489197, "learning_rate": 8.119110815010151e-05, "loss": 0.0487, "step": 9640 }, { "grad_norm": 0.2850775420665741, "learning_rate": 8.114801266486261e-05, "loss": 0.0532, "step": 9650 }, { "grad_norm": 0.25056594610214233, "learning_rate": 8.110487933185122e-05, "loss": 0.0513, "step": 9660 }, { "grad_norm": 0.3580748736858368, "learning_rate": 8.106170820347847e-05, "loss": 0.0575, "step": 9670 }, { "grad_norm": 0.2431454211473465, "learning_rate": 8.101849933220132e-05, "loss": 0.0469, "step": 9680 }, { "grad_norm": 0.259175181388855, "learning_rate": 8.097525277052264e-05, "loss": 0.0532, "step": 9690 }, { "grad_norm": 0.2431984692811966, "learning_rate": 8.093196857099106e-05, "loss": 0.0442, "step": 9700 }, { "grad_norm": 0.3272860050201416, "learning_rate": 8.088864678620096e-05, "loss": 0.0477, "step": 9710 }, { "grad_norm": 0.31705576181411743, "learning_rate": 8.084528746879243e-05, "loss": 0.061, "step": 9720 }, { "grad_norm": 0.3224635720252991, "learning_rate": 8.080189067145108e-05, "loss": 0.0682, "step": 9730 }, { "grad_norm": 0.35277220606803894, "learning_rate": 8.075845644690813e-05, "loss": 0.0544, "step": 9740 }, { "grad_norm": 0.3342515230178833, "learning_rate": 8.07149848479403e-05, "loss": 0.0476, "step": 9750 }, { "grad_norm": 0.28427353501319885, "learning_rate": 8.067147592736962e-05, "loss": 0.0628, "step": 9760 }, { "grad_norm": 0.3272070288658142, "learning_rate": 8.062792973806357e-05, "loss": 0.0604, "step": 9770 }, { "grad_norm": 0.27143317461013794, "learning_rate": 8.058434633293485e-05, "loss": 0.0566, "step": 9780 }, { "grad_norm": 0.31342849135398865, "learning_rate": 8.054072576494143e-05, "loss": 0.054, "step": 9790 }, { "grad_norm": 0.318314790725708, "learning_rate": 8.04970680870864e-05, "loss": 0.0527, "step": 9800 }, { "grad_norm": 0.36012133955955505, "learning_rate": 8.045337335241793e-05, "loss": 0.0595, "step": 9810 }, { "grad_norm": 0.2712692320346832, "learning_rate": 8.040964161402932e-05, "loss": 0.0656, "step": 9820 }, { "grad_norm": 0.26829537749290466, "learning_rate": 8.036587292505869e-05, "loss": 0.0597, "step": 9830 }, { "grad_norm": 0.2871735990047455, "learning_rate": 8.032206733868912e-05, "loss": 0.0661, "step": 9840 }, { "grad_norm": 0.30234649777412415, "learning_rate": 8.027822490814858e-05, "loss": 0.0507, "step": 9850 }, { "grad_norm": 0.2740025818347931, "learning_rate": 8.023434568670972e-05, "loss": 0.0476, "step": 9860 }, { "grad_norm": 0.2501823306083679, "learning_rate": 8.019042972768992e-05, "loss": 0.0468, "step": 9870 }, { "grad_norm": 0.31524837017059326, "learning_rate": 8.014647708445124e-05, "loss": 0.0604, "step": 9880 }, { "grad_norm": 0.25509488582611084, "learning_rate": 8.010248781040027e-05, "loss": 0.0393, "step": 9890 }, { "grad_norm": 0.3586788773536682, "learning_rate": 8.005846195898814e-05, "loss": 0.0558, "step": 9900 }, { "grad_norm": 0.32207879424095154, "learning_rate": 8.00143995837104e-05, "loss": 0.0631, "step": 9910 }, { "grad_norm": 0.27882054448127747, "learning_rate": 7.997030073810698e-05, "loss": 0.0595, "step": 9920 }, { "grad_norm": 0.3689604103565216, "learning_rate": 7.992616547576217e-05, "loss": 0.0446, "step": 9930 }, { "grad_norm": 0.32280346751213074, "learning_rate": 7.988199385030445e-05, "loss": 0.0523, "step": 9940 }, { "grad_norm": 0.28709346055984497, "learning_rate": 7.98377859154065e-05, "loss": 0.0456, "step": 9950 }, { "grad_norm": 0.31783056259155273, "learning_rate": 7.979354172478516e-05, "loss": 0.0457, "step": 9960 }, { "grad_norm": 0.28264951705932617, "learning_rate": 7.974926133220127e-05, "loss": 0.046, "step": 9970 }, { "grad_norm": 0.2878062427043915, "learning_rate": 7.970494479145967e-05, "loss": 0.0514, "step": 9980 }, { "grad_norm": 0.3237844705581665, "learning_rate": 7.966059215640919e-05, "loss": 0.0588, "step": 9990 }, { "grad_norm": 0.3620549440383911, "learning_rate": 7.961620348094241e-05, "loss": 0.0682, "step": 10000 }, { "grad_norm": 0.33132708072662354, "learning_rate": 7.957177881899577e-05, "loss": 0.0563, "step": 10010 }, { "grad_norm": 0.32278794050216675, "learning_rate": 7.952731822454944e-05, "loss": 0.0556, "step": 10020 }, { "grad_norm": 0.2934204339981079, "learning_rate": 7.948282175162722e-05, "loss": 0.0448, "step": 10030 }, { "grad_norm": 0.2591647803783417, "learning_rate": 7.943828945429652e-05, "loss": 0.0723, "step": 10040 }, { "grad_norm": 0.32507428526878357, "learning_rate": 7.939372138666827e-05, "loss": 0.0613, "step": 10050 }, { "grad_norm": 0.32368388772010803, "learning_rate": 7.934911760289692e-05, "loss": 0.0534, "step": 10060 }, { "grad_norm": 0.27588149905204773, "learning_rate": 7.930447815718022e-05, "loss": 0.0429, "step": 10070 }, { "grad_norm": 0.27428266406059265, "learning_rate": 7.925980310375934e-05, "loss": 0.0594, "step": 10080 }, { "grad_norm": 0.40983086824417114, "learning_rate": 7.921509249691865e-05, "loss": 0.0565, "step": 10090 }, { "grad_norm": 0.24723628163337708, "learning_rate": 7.917034639098579e-05, "loss": 0.0618, "step": 10100 }, { "grad_norm": 0.26453569531440735, "learning_rate": 7.912556484033146e-05, "loss": 0.0387, "step": 10110 }, { "grad_norm": 0.3762427568435669, "learning_rate": 7.908074789936952e-05, "loss": 0.0468, "step": 10120 }, { "grad_norm": 0.32893022894859314, "learning_rate": 7.903589562255673e-05, "loss": 0.0546, "step": 10130 }, { "grad_norm": 0.32823359966278076, "learning_rate": 7.899100806439285e-05, "loss": 0.052, "step": 10140 }, { "grad_norm": 0.2984393835067749, "learning_rate": 7.894608527942049e-05, "loss": 0.064, "step": 10150 }, { "grad_norm": 0.3259609639644623, "learning_rate": 7.89011273222251e-05, "loss": 0.057, "step": 10160 }, { "grad_norm": 0.35673999786376953, "learning_rate": 7.885613424743482e-05, "loss": 0.049, "step": 10170 }, { "grad_norm": 0.24055549502372742, "learning_rate": 7.881110610972044e-05, "loss": 0.0688, "step": 10180 }, { "grad_norm": 0.36448580026626587, "learning_rate": 7.876604296379544e-05, "loss": 0.0674, "step": 10190 }, { "grad_norm": 0.2905661463737488, "learning_rate": 7.872094486441579e-05, "loss": 0.0654, "step": 10200 }, { "grad_norm": 0.2830527424812317, "learning_rate": 7.867581186637991e-05, "loss": 0.0523, "step": 10210 }, { "grad_norm": 0.22366714477539062, "learning_rate": 7.863064402452866e-05, "loss": 0.0497, "step": 10220 }, { "grad_norm": 0.31298044323921204, "learning_rate": 7.858544139374524e-05, "loss": 0.0582, "step": 10230 }, { "grad_norm": 0.2092851996421814, "learning_rate": 7.854020402895508e-05, "loss": 0.051, "step": 10240 }, { "grad_norm": 0.2408243864774704, "learning_rate": 7.849493198512587e-05, "loss": 0.0509, "step": 10250 }, { "grad_norm": 0.2615901529788971, "learning_rate": 7.84496253172674e-05, "loss": 0.0362, "step": 10260 }, { "grad_norm": 0.31269392371177673, "learning_rate": 7.840428408043155e-05, "loss": 0.0527, "step": 10270 }, { "grad_norm": 0.34652313590049744, "learning_rate": 7.835890832971218e-05, "loss": 0.0595, "step": 10280 }, { "grad_norm": 0.31965941190719604, "learning_rate": 7.831349812024512e-05, "loss": 0.062, "step": 10290 }, { "grad_norm": 0.29599615931510925, "learning_rate": 7.826805350720807e-05, "loss": 0.0537, "step": 10300 }, { "grad_norm": 0.24278312921524048, "learning_rate": 7.822257454582049e-05, "loss": 0.0463, "step": 10310 }, { "grad_norm": 0.35328978300094604, "learning_rate": 7.817706129134363e-05, "loss": 0.0468, "step": 10320 }, { "grad_norm": 0.24585260450839996, "learning_rate": 7.813151379908036e-05, "loss": 0.0538, "step": 10330 }, { "grad_norm": 0.332724392414093, "learning_rate": 7.808593212437523e-05, "loss": 0.0557, "step": 10340 }, { "grad_norm": 0.3731438219547272, "learning_rate": 7.804031632261421e-05, "loss": 0.0659, "step": 10350 }, { "grad_norm": 0.3650723695755005, "learning_rate": 7.799466644922484e-05, "loss": 0.0422, "step": 10360 }, { "grad_norm": 0.29940640926361084, "learning_rate": 7.794898255967602e-05, "loss": 0.0497, "step": 10370 }, { "grad_norm": 0.3471282124519348, "learning_rate": 7.790326470947795e-05, "loss": 0.0527, "step": 10380 }, { "grad_norm": 0.28322187066078186, "learning_rate": 7.785751295418217e-05, "loss": 0.0556, "step": 10390 }, { "grad_norm": 0.2890983819961548, "learning_rate": 7.781172734938136e-05, "loss": 0.0596, "step": 10400 }, { "grad_norm": 0.27639293670654297, "learning_rate": 7.776590795070933e-05, "loss": 0.0554, "step": 10410 }, { "grad_norm": 0.23798902332782745, "learning_rate": 7.772005481384099e-05, "loss": 0.0539, "step": 10420 }, { "grad_norm": 0.2751089334487915, "learning_rate": 7.767416799449223e-05, "loss": 0.0422, "step": 10430 }, { "grad_norm": 0.3559429347515106, "learning_rate": 7.762824754841984e-05, "loss": 0.0452, "step": 10440 }, { "grad_norm": 0.24810270965099335, "learning_rate": 7.758229353142152e-05, "loss": 0.0661, "step": 10450 }, { "grad_norm": 0.21435704827308655, "learning_rate": 7.753630599933572e-05, "loss": 0.0435, "step": 10460 }, { "grad_norm": 0.21953807771205902, "learning_rate": 7.749028500804161e-05, "loss": 0.0411, "step": 10470 }, { "grad_norm": 0.20052248239517212, "learning_rate": 7.744423061345906e-05, "loss": 0.0344, "step": 10480 }, { "grad_norm": 0.1935633271932602, "learning_rate": 7.73981428715485e-05, "loss": 0.0372, "step": 10490 }, { "grad_norm": 0.2296011596918106, "learning_rate": 7.735202183831085e-05, "loss": 0.0403, "step": 10500 }, { "grad_norm": 0.33665189146995544, "learning_rate": 7.730586756978757e-05, "loss": 0.0553, "step": 10510 }, { "grad_norm": 0.23534074425697327, "learning_rate": 7.72596801220604e-05, "loss": 0.0317, "step": 10520 }, { "grad_norm": 0.27294039726257324, "learning_rate": 7.721345955125147e-05, "loss": 0.044, "step": 10530 }, { "grad_norm": 0.20685341954231262, "learning_rate": 7.71672059135231e-05, "loss": 0.0324, "step": 10540 }, { "grad_norm": 0.2650996744632721, "learning_rate": 7.712091926507787e-05, "loss": 0.0329, "step": 10550 }, { "grad_norm": 0.24565280973911285, "learning_rate": 7.70745996621584e-05, "loss": 0.0351, "step": 10560 }, { "grad_norm": 0.3496357500553131, "learning_rate": 7.702824716104735e-05, "loss": 0.0359, "step": 10570 }, { "grad_norm": 0.20169053971767426, "learning_rate": 7.698186181806743e-05, "loss": 0.0464, "step": 10580 }, { "grad_norm": 0.28644654154777527, "learning_rate": 7.693544368958116e-05, "loss": 0.0523, "step": 10590 }, { "grad_norm": 0.24216139316558838, "learning_rate": 7.688899283199096e-05, "loss": 0.0537, "step": 10600 }, { "grad_norm": 0.30106598138809204, "learning_rate": 7.684250930173901e-05, "loss": 0.0681, "step": 10610 }, { "grad_norm": 0.43666359782218933, "learning_rate": 7.679599315530718e-05, "loss": 0.0639, "step": 10620 }, { "grad_norm": 0.3166155219078064, "learning_rate": 7.674944444921695e-05, "loss": 0.0439, "step": 10630 }, { "grad_norm": 0.2645878791809082, "learning_rate": 7.670286324002944e-05, "loss": 0.0512, "step": 10640 }, { "grad_norm": 0.3639654517173767, "learning_rate": 7.665624958434514e-05, "loss": 0.0602, "step": 10650 }, { "grad_norm": 0.3224162757396698, "learning_rate": 7.66096035388041e-05, "loss": 0.0512, "step": 10660 }, { "grad_norm": 0.31354862451553345, "learning_rate": 7.656292516008563e-05, "loss": 0.0558, "step": 10670 }, { "grad_norm": 0.3189253807067871, "learning_rate": 7.651621450490837e-05, "loss": 0.0332, "step": 10680 }, { "grad_norm": 0.2652711570262909, "learning_rate": 7.646947163003017e-05, "loss": 0.0415, "step": 10690 }, { "grad_norm": 0.3182506263256073, "learning_rate": 7.642269659224803e-05, "loss": 0.0681, "step": 10700 }, { "grad_norm": 0.29999294877052307, "learning_rate": 7.637588944839803e-05, "loss": 0.0546, "step": 10710 }, { "grad_norm": 0.19786204397678375, "learning_rate": 7.632905025535528e-05, "loss": 0.0525, "step": 10720 }, { "grad_norm": 0.25497713685035706, "learning_rate": 7.628217907003378e-05, "loss": 0.0407, "step": 10730 }, { "grad_norm": 0.2241908758878708, "learning_rate": 7.623527594938649e-05, "loss": 0.033, "step": 10740 }, { "grad_norm": 0.32697877287864685, "learning_rate": 7.618834095040509e-05, "loss": 0.0453, "step": 10750 }, { "grad_norm": 0.2551109790802002, "learning_rate": 7.614137413012e-05, "loss": 0.0505, "step": 10760 }, { "grad_norm": 0.3612462878227234, "learning_rate": 7.609437554560042e-05, "loss": 0.0515, "step": 10770 }, { "grad_norm": 0.2806759774684906, "learning_rate": 7.604734525395398e-05, "loss": 0.0491, "step": 10780 }, { "grad_norm": 0.27127817273139954, "learning_rate": 7.600028331232699e-05, "loss": 0.048, "step": 10790 }, { "grad_norm": 0.2838228642940521, "learning_rate": 7.595318977790408e-05, "loss": 0.0457, "step": 10800 }, { "grad_norm": 0.3736947178840637, "learning_rate": 7.590606470790836e-05, "loss": 0.0455, "step": 10810 }, { "grad_norm": 0.19184522330760956, "learning_rate": 7.585890815960125e-05, "loss": 0.0451, "step": 10820 }, { "grad_norm": 0.25505003333091736, "learning_rate": 7.581172019028238e-05, "loss": 0.0379, "step": 10830 }, { "grad_norm": 0.21059545874595642, "learning_rate": 7.576450085728959e-05, "loss": 0.0391, "step": 10840 }, { "grad_norm": 0.3221857249736786, "learning_rate": 7.571725021799885e-05, "loss": 0.0504, "step": 10850 }, { "grad_norm": 0.2043369561433792, "learning_rate": 7.566996832982409e-05, "loss": 0.0345, "step": 10860 }, { "grad_norm": 0.2110092043876648, "learning_rate": 7.56226552502173e-05, "loss": 0.0428, "step": 10870 }, { "grad_norm": 0.23808524012565613, "learning_rate": 7.557531103666833e-05, "loss": 0.0536, "step": 10880 }, { "grad_norm": 0.24966995418071747, "learning_rate": 7.552793574670485e-05, "loss": 0.0448, "step": 10890 }, { "grad_norm": 0.24053335189819336, "learning_rate": 7.54805294378923e-05, "loss": 0.0499, "step": 10900 }, { "grad_norm": 0.31211426854133606, "learning_rate": 7.543309216783384e-05, "loss": 0.0404, "step": 10910 }, { "grad_norm": 0.3212166428565979, "learning_rate": 7.53856239941702e-05, "loss": 0.0492, "step": 10920 }, { "grad_norm": 0.33951273560523987, "learning_rate": 7.533812497457972e-05, "loss": 0.0568, "step": 10930 }, { "grad_norm": 0.316363126039505, "learning_rate": 7.529059516677814e-05, "loss": 0.0446, "step": 10940 }, { "grad_norm": 0.20694230496883392, "learning_rate": 7.524303462851872e-05, "loss": 0.0469, "step": 10950 }, { "grad_norm": 0.2670970857143402, "learning_rate": 7.519544341759192e-05, "loss": 0.0421, "step": 10960 }, { "grad_norm": 0.3025738000869751, "learning_rate": 7.514782159182562e-05, "loss": 0.0481, "step": 10970 }, { "grad_norm": 0.2301190346479416, "learning_rate": 7.510016920908481e-05, "loss": 0.0495, "step": 10980 }, { "grad_norm": 0.234406977891922, "learning_rate": 7.505248632727158e-05, "loss": 0.0393, "step": 10990 }, { "grad_norm": 0.22023805975914001, "learning_rate": 7.50047730043252e-05, "loss": 0.0415, "step": 11000 }, { "grad_norm": 0.3040928840637207, "learning_rate": 7.495702929822183e-05, "loss": 0.0423, "step": 11010 }, { "grad_norm": 0.23407776653766632, "learning_rate": 7.490925526697455e-05, "loss": 0.046, "step": 11020 }, { "grad_norm": 0.30900394916534424, "learning_rate": 7.486145096863334e-05, "loss": 0.0547, "step": 11030 }, { "grad_norm": 0.2530072331428528, "learning_rate": 7.481361646128491e-05, "loss": 0.0471, "step": 11040 }, { "grad_norm": 0.3361715078353882, "learning_rate": 7.476575180305271e-05, "loss": 0.0399, "step": 11050 }, { "grad_norm": 0.24628327786922455, "learning_rate": 7.471785705209682e-05, "loss": 0.0576, "step": 11060 }, { "grad_norm": 0.28060218691825867, "learning_rate": 7.466993226661387e-05, "loss": 0.0424, "step": 11070 }, { "grad_norm": 0.2367982268333435, "learning_rate": 7.462197750483699e-05, "loss": 0.0508, "step": 11080 }, { "grad_norm": 0.3303638696670532, "learning_rate": 7.457399282503574e-05, "loss": 0.0475, "step": 11090 }, { "grad_norm": 0.2684824466705322, "learning_rate": 7.452597828551604e-05, "loss": 0.0545, "step": 11100 }, { "grad_norm": 0.33203116059303284, "learning_rate": 7.447793394462006e-05, "loss": 0.0661, "step": 11110 }, { "grad_norm": 0.27813443541526794, "learning_rate": 7.442985986072624e-05, "loss": 0.0485, "step": 11120 }, { "grad_norm": 0.20397348701953888, "learning_rate": 7.438175609224908e-05, "loss": 0.0651, "step": 11130 }, { "grad_norm": 0.29070982336997986, "learning_rate": 7.433362269763924e-05, "loss": 0.0483, "step": 11140 }, { "grad_norm": 0.30380284786224365, "learning_rate": 7.428545973538329e-05, "loss": 0.0362, "step": 11150 }, { "grad_norm": 0.21173126995563507, "learning_rate": 7.42372672640038e-05, "loss": 0.0441, "step": 11160 }, { "grad_norm": 0.29563266038894653, "learning_rate": 7.418904534205917e-05, "loss": 0.0475, "step": 11170 }, { "grad_norm": 0.30811402201652527, "learning_rate": 7.414079402814356e-05, "loss": 0.0569, "step": 11180 }, { "grad_norm": 0.3020748496055603, "learning_rate": 7.409251338088687e-05, "loss": 0.0615, "step": 11190 }, { "grad_norm": 0.27788203954696655, "learning_rate": 7.404420345895467e-05, "loss": 0.0526, "step": 11200 }, { "grad_norm": 0.3455301821231842, "learning_rate": 7.399586432104804e-05, "loss": 0.0603, "step": 11210 }, { "grad_norm": 0.23239581286907196, "learning_rate": 7.394749602590358e-05, "loss": 0.0494, "step": 11220 }, { "grad_norm": 0.2696135342121124, "learning_rate": 7.389909863229336e-05, "loss": 0.0632, "step": 11230 }, { "grad_norm": 0.2566918134689331, "learning_rate": 7.385067219902477e-05, "loss": 0.0502, "step": 11240 }, { "grad_norm": 0.2538367211818695, "learning_rate": 7.380221678494049e-05, "loss": 0.0372, "step": 11250 }, { "grad_norm": 0.24516938626766205, "learning_rate": 7.37537324489184e-05, "loss": 0.0384, "step": 11260 }, { "grad_norm": 0.22987090051174164, "learning_rate": 7.370521924987155e-05, "loss": 0.05, "step": 11270 }, { "grad_norm": 0.3568144142627716, "learning_rate": 7.365667724674806e-05, "loss": 0.0676, "step": 11280 }, { "grad_norm": 0.3328545093536377, "learning_rate": 7.360810649853105e-05, "loss": 0.0571, "step": 11290 }, { "grad_norm": 0.2935340404510498, "learning_rate": 7.355950706423853e-05, "loss": 0.0539, "step": 11300 }, { "grad_norm": 0.28663113713264465, "learning_rate": 7.351087900292342e-05, "loss": 0.0582, "step": 11310 }, { "grad_norm": 0.24561890959739685, "learning_rate": 7.346222237367339e-05, "loss": 0.0526, "step": 11320 }, { "grad_norm": 0.2559717297554016, "learning_rate": 7.341353723561081e-05, "loss": 0.0526, "step": 11330 }, { "grad_norm": 0.17693690955638885, "learning_rate": 7.336482364789277e-05, "loss": 0.0296, "step": 11340 }, { "grad_norm": 0.22005753219127655, "learning_rate": 7.331608166971082e-05, "loss": 0.0572, "step": 11350 }, { "grad_norm": 0.26099079847335815, "learning_rate": 7.326731136029108e-05, "loss": 0.0399, "step": 11360 }, { "grad_norm": 0.23998157680034637, "learning_rate": 7.321851277889408e-05, "loss": 0.0394, "step": 11370 }, { "grad_norm": 0.236002117395401, "learning_rate": 7.316968598481469e-05, "loss": 0.041, "step": 11380 }, { "grad_norm": 0.29364362359046936, "learning_rate": 7.312083103738206e-05, "loss": 0.0585, "step": 11390 }, { "grad_norm": 0.26120659708976746, "learning_rate": 7.307194799595958e-05, "loss": 0.0564, "step": 11400 }, { "grad_norm": 0.2911699116230011, "learning_rate": 7.302303691994473e-05, "loss": 0.0483, "step": 11410 }, { "grad_norm": 0.21139676868915558, "learning_rate": 7.29740978687691e-05, "loss": 0.0577, "step": 11420 }, { "grad_norm": 0.23853741586208344, "learning_rate": 7.292513090189825e-05, "loss": 0.039, "step": 11430 }, { "grad_norm": 0.25128620862960815, "learning_rate": 7.287613607883163e-05, "loss": 0.0396, "step": 11440 }, { "grad_norm": 0.25657713413238525, "learning_rate": 7.282711345910263e-05, "loss": 0.0418, "step": 11450 }, { "grad_norm": 0.27613696455955505, "learning_rate": 7.27780631022783e-05, "loss": 0.0486, "step": 11460 }, { "grad_norm": 0.3475509583950043, "learning_rate": 7.272898506795948e-05, "loss": 0.043, "step": 11470 }, { "grad_norm": 0.20536865293979645, "learning_rate": 7.267987941578058e-05, "loss": 0.0372, "step": 11480 }, { "grad_norm": 0.2695615589618683, "learning_rate": 7.263074620540962e-05, "loss": 0.0543, "step": 11490 }, { "grad_norm": 0.46835342049598694, "learning_rate": 7.25815854965481e-05, "loss": 0.0447, "step": 11500 }, { "grad_norm": 0.20317630469799042, "learning_rate": 7.253239734893089e-05, "loss": 0.0347, "step": 11510 }, { "grad_norm": 0.21950574219226837, "learning_rate": 7.248318182232623e-05, "loss": 0.0443, "step": 11520 }, { "grad_norm": 0.2229817807674408, "learning_rate": 7.243393897653565e-05, "loss": 0.0389, "step": 11530 }, { "grad_norm": 0.23480597138404846, "learning_rate": 7.23846688713938e-05, "loss": 0.0431, "step": 11540 }, { "grad_norm": 0.2717709243297577, "learning_rate": 7.233537156676854e-05, "loss": 0.0509, "step": 11550 }, { "grad_norm": 0.22844643890857697, "learning_rate": 7.228604712256076e-05, "loss": 0.0433, "step": 11560 }, { "grad_norm": 0.2866854667663574, "learning_rate": 7.223669559870426e-05, "loss": 0.0377, "step": 11570 }, { "grad_norm": 0.2501957416534424, "learning_rate": 7.218731705516585e-05, "loss": 0.044, "step": 11580 }, { "grad_norm": 0.29818347096443176, "learning_rate": 7.21379115519451e-05, "loss": 0.0497, "step": 11590 }, { "grad_norm": 0.2898947298526764, "learning_rate": 7.20884791490743e-05, "loss": 0.0443, "step": 11600 }, { "grad_norm": 0.20588861405849457, "learning_rate": 7.203901990661857e-05, "loss": 0.0505, "step": 11610 }, { "grad_norm": 0.2993820905685425, "learning_rate": 7.198953388467549e-05, "loss": 0.0558, "step": 11620 }, { "grad_norm": 0.2154010832309723, "learning_rate": 7.194002114337526e-05, "loss": 0.0597, "step": 11630 }, { "grad_norm": 0.25284940004348755, "learning_rate": 7.189048174288054e-05, "loss": 0.0369, "step": 11640 }, { "grad_norm": 0.26337292790412903, "learning_rate": 7.184091574338636e-05, "loss": 0.0403, "step": 11650 }, { "grad_norm": 0.30319035053253174, "learning_rate": 7.179132320512009e-05, "loss": 0.035, "step": 11660 }, { "grad_norm": 0.2818997800350189, "learning_rate": 7.174170418834134e-05, "loss": 0.0383, "step": 11670 }, { "grad_norm": 0.24978923797607422, "learning_rate": 7.169205875334189e-05, "loss": 0.0478, "step": 11680 }, { "grad_norm": 0.28280794620513916, "learning_rate": 7.164238696044562e-05, "loss": 0.0449, "step": 11690 }, { "grad_norm": 0.22627778351306915, "learning_rate": 7.159268887000846e-05, "loss": 0.0427, "step": 11700 }, { "grad_norm": 0.24511271715164185, "learning_rate": 7.154296454241827e-05, "loss": 0.0323, "step": 11710 }, { "grad_norm": 0.2262425273656845, "learning_rate": 7.149321403809479e-05, "loss": 0.0398, "step": 11720 }, { "grad_norm": 0.19349408149719238, "learning_rate": 7.144343741748956e-05, "loss": 0.0287, "step": 11730 }, { "grad_norm": 0.20254862308502197, "learning_rate": 7.139363474108589e-05, "loss": 0.0389, "step": 11740 }, { "grad_norm": 0.355533629655838, "learning_rate": 7.134380606939874e-05, "loss": 0.0459, "step": 11750 }, { "grad_norm": 0.21760597825050354, "learning_rate": 7.12939514629746e-05, "loss": 0.0339, "step": 11760 }, { "grad_norm": 0.19992859661579132, "learning_rate": 7.124407098239155e-05, "loss": 0.0534, "step": 11770 }, { "grad_norm": 0.2798827290534973, "learning_rate": 7.119416468825908e-05, "loss": 0.0336, "step": 11780 }, { "grad_norm": 0.25921034812927246, "learning_rate": 7.114423264121804e-05, "loss": 0.0396, "step": 11790 }, { "grad_norm": 0.20574228465557098, "learning_rate": 7.109427490194056e-05, "loss": 0.0449, "step": 11800 }, { "grad_norm": 0.21406352519989014, "learning_rate": 7.104429153113001e-05, "loss": 0.0572, "step": 11810 }, { "grad_norm": 0.24189870059490204, "learning_rate": 7.099428258952092e-05, "loss": 0.0423, "step": 11820 }, { "grad_norm": 0.26482176780700684, "learning_rate": 7.094424813787883e-05, "loss": 0.0447, "step": 11830 }, { "grad_norm": 0.3032524585723877, "learning_rate": 7.089418823700034e-05, "loss": 0.0387, "step": 11840 }, { "grad_norm": 0.17934875190258026, "learning_rate": 7.084410294771298e-05, "loss": 0.042, "step": 11850 }, { "grad_norm": 0.2308870404958725, "learning_rate": 7.079399233087504e-05, "loss": 0.0548, "step": 11860 }, { "grad_norm": 0.2603908181190491, "learning_rate": 7.074385644737568e-05, "loss": 0.0456, "step": 11870 }, { "grad_norm": 0.198285773396492, "learning_rate": 7.069369535813473e-05, "loss": 0.0541, "step": 11880 }, { "grad_norm": 0.19239330291748047, "learning_rate": 7.06435091241026e-05, "loss": 0.0409, "step": 11890 }, { "grad_norm": 0.2985738217830658, "learning_rate": 7.059329780626034e-05, "loss": 0.0548, "step": 11900 }, { "grad_norm": 0.270248681306839, "learning_rate": 7.054306146561944e-05, "loss": 0.0487, "step": 11910 }, { "grad_norm": 0.2968447804450989, "learning_rate": 7.049280016322178e-05, "loss": 0.0448, "step": 11920 }, { "grad_norm": 0.309084415435791, "learning_rate": 7.044251396013957e-05, "loss": 0.0478, "step": 11930 }, { "grad_norm": 0.29188257455825806, "learning_rate": 7.039220291747528e-05, "loss": 0.0404, "step": 11940 }, { "grad_norm": 0.22276915609836578, "learning_rate": 7.034186709636159e-05, "loss": 0.0391, "step": 11950 }, { "grad_norm": 0.24379786849021912, "learning_rate": 7.029150655796129e-05, "loss": 0.0444, "step": 11960 }, { "grad_norm": 0.29198282957077026, "learning_rate": 7.024112136346712e-05, "loss": 0.0495, "step": 11970 }, { "grad_norm": 0.29257962107658386, "learning_rate": 7.01907115741019e-05, "loss": 0.0494, "step": 11980 }, { "grad_norm": 0.19931025803089142, "learning_rate": 7.014027725111825e-05, "loss": 0.0542, "step": 11990 }, { "grad_norm": 0.30612480640411377, "learning_rate": 7.008981845579864e-05, "loss": 0.0532, "step": 12000 }, { "grad_norm": 0.23519417643547058, "learning_rate": 7.003933524945528e-05, "loss": 0.0463, "step": 12010 }, { "grad_norm": 0.32575276494026184, "learning_rate": 6.998882769342998e-05, "loss": 0.0405, "step": 12020 }, { "grad_norm": 0.20692406594753265, "learning_rate": 6.993829584909423e-05, "loss": 0.0465, "step": 12030 }, { "grad_norm": 0.3243485689163208, "learning_rate": 6.988773977784895e-05, "loss": 0.0542, "step": 12040 }, { "grad_norm": 0.35817739367485046, "learning_rate": 6.983715954112454e-05, "loss": 0.0603, "step": 12050 }, { "grad_norm": 0.3146810233592987, "learning_rate": 6.978655520038079e-05, "loss": 0.0587, "step": 12060 }, { "grad_norm": 0.24292369186878204, "learning_rate": 6.97359268171067e-05, "loss": 0.0753, "step": 12070 }, { "grad_norm": 0.3394567370414734, "learning_rate": 6.968527445282056e-05, "loss": 0.052, "step": 12080 }, { "grad_norm": 0.20344555377960205, "learning_rate": 6.963459816906977e-05, "loss": 0.0434, "step": 12090 }, { "grad_norm": 0.26246607303619385, "learning_rate": 6.958389802743077e-05, "loss": 0.0605, "step": 12100 }, { "grad_norm": 0.3500259220600128, "learning_rate": 6.953317408950903e-05, "loss": 0.0553, "step": 12110 }, { "grad_norm": 0.2843182384967804, "learning_rate": 6.94824264169389e-05, "loss": 0.0496, "step": 12120 }, { "grad_norm": 0.2245684415102005, "learning_rate": 6.94316550713836e-05, "loss": 0.0563, "step": 12130 }, { "grad_norm": 0.2734261155128479, "learning_rate": 6.938086011453513e-05, "loss": 0.0493, "step": 12140 }, { "grad_norm": 0.27631089091300964, "learning_rate": 6.93300416081141e-05, "loss": 0.055, "step": 12150 }, { "grad_norm": 0.2564859688282013, "learning_rate": 6.927919961386984e-05, "loss": 0.0466, "step": 12160 }, { "grad_norm": 0.27902746200561523, "learning_rate": 6.922833419358013e-05, "loss": 0.0496, "step": 12170 }, { "grad_norm": 0.281280517578125, "learning_rate": 6.917744540905124e-05, "loss": 0.0658, "step": 12180 }, { "grad_norm": 0.3388251066207886, "learning_rate": 6.912653332211787e-05, "loss": 0.0503, "step": 12190 }, { "grad_norm": 0.25451093912124634, "learning_rate": 6.9075597994643e-05, "loss": 0.0726, "step": 12200 }, { "grad_norm": 0.20979827642440796, "learning_rate": 6.902463948851787e-05, "loss": 0.0761, "step": 12210 }, { "grad_norm": 0.29862359166145325, "learning_rate": 6.897365786566184e-05, "loss": 0.0512, "step": 12220 }, { "grad_norm": 0.22309713065624237, "learning_rate": 6.892265318802242e-05, "loss": 0.0585, "step": 12230 }, { "grad_norm": 0.23608781397342682, "learning_rate": 6.887162551757507e-05, "loss": 0.0467, "step": 12240 }, { "grad_norm": 0.223038449883461, "learning_rate": 6.882057491632326e-05, "loss": 0.0648, "step": 12250 }, { "grad_norm": 0.3235296607017517, "learning_rate": 6.876950144629824e-05, "loss": 0.0674, "step": 12260 }, { "grad_norm": 0.2999056875705719, "learning_rate": 6.87184051695591e-05, "loss": 0.0492, "step": 12270 }, { "grad_norm": 0.27466267347335815, "learning_rate": 6.866728614819268e-05, "loss": 0.0505, "step": 12280 }, { "grad_norm": 0.2942052483558655, "learning_rate": 6.861614444431337e-05, "loss": 0.0544, "step": 12290 }, { "grad_norm": 0.2369472086429596, "learning_rate": 6.856498012006318e-05, "loss": 0.0391, "step": 12300 }, { "grad_norm": 0.2866470217704773, "learning_rate": 6.851379323761157e-05, "loss": 0.0501, "step": 12310 }, { "grad_norm": 0.28425905108451843, "learning_rate": 6.846258385915545e-05, "loss": 0.0558, "step": 12320 }, { "grad_norm": 0.17640024423599243, "learning_rate": 6.841135204691902e-05, "loss": 0.0427, "step": 12330 }, { "grad_norm": 0.26754412055015564, "learning_rate": 6.836009786315377e-05, "loss": 0.0525, "step": 12340 }, { "grad_norm": 0.25522381067276, "learning_rate": 6.830882137013839e-05, "loss": 0.0454, "step": 12350 }, { "grad_norm": 0.2924649119377136, "learning_rate": 6.825752263017863e-05, "loss": 0.0604, "step": 12360 }, { "grad_norm": 0.2609576880931854, "learning_rate": 6.820620170560731e-05, "loss": 0.0429, "step": 12370 }, { "grad_norm": 0.2827361226081848, "learning_rate": 6.815485865878418e-05, "loss": 0.0522, "step": 12380 }, { "grad_norm": 0.31728923320770264, "learning_rate": 6.810349355209588e-05, "loss": 0.0574, "step": 12390 }, { "grad_norm": 0.2583361864089966, "learning_rate": 6.805210644795588e-05, "loss": 0.0594, "step": 12400 }, { "grad_norm": 0.34379732608795166, "learning_rate": 6.800069740880433e-05, "loss": 0.0645, "step": 12410 }, { "grad_norm": 0.2525900900363922, "learning_rate": 6.794926649710807e-05, "loss": 0.0433, "step": 12420 }, { "grad_norm": 0.36740368604660034, "learning_rate": 6.789781377536052e-05, "loss": 0.0633, "step": 12430 }, { "grad_norm": 0.30210140347480774, "learning_rate": 6.784633930608158e-05, "loss": 0.0622, "step": 12440 }, { "grad_norm": 0.2558841109275818, "learning_rate": 6.779484315181759e-05, "loss": 0.044, "step": 12450 }, { "grad_norm": 0.22508147358894348, "learning_rate": 6.774332537514122e-05, "loss": 0.0532, "step": 12460 }, { "grad_norm": 0.3017764389514923, "learning_rate": 6.769178603865143e-05, "loss": 0.0478, "step": 12470 }, { "grad_norm": 0.2352668195962906, "learning_rate": 6.764022520497337e-05, "loss": 0.0396, "step": 12480 }, { "grad_norm": 0.22109867632389069, "learning_rate": 6.758864293675833e-05, "loss": 0.0408, "step": 12490 }, { "grad_norm": 0.24820923805236816, "learning_rate": 6.753703929668362e-05, "loss": 0.048, "step": 12500 }, { "grad_norm": 0.2098865956068039, "learning_rate": 6.748541434745253e-05, "loss": 0.0599, "step": 12510 }, { "grad_norm": 0.31755331158638, "learning_rate": 6.743376815179424e-05, "loss": 0.0452, "step": 12520 }, { "grad_norm": 0.21638964116573334, "learning_rate": 6.738210077246376e-05, "loss": 0.0441, "step": 12530 }, { "grad_norm": 0.3185419738292694, "learning_rate": 6.733041227224181e-05, "loss": 0.0487, "step": 12540 }, { "grad_norm": 0.28742408752441406, "learning_rate": 6.72787027139348e-05, "loss": 0.0435, "step": 12550 }, { "grad_norm": 0.2878827452659607, "learning_rate": 6.72269721603747e-05, "loss": 0.0409, "step": 12560 }, { "grad_norm": 0.21383638679981232, "learning_rate": 6.717522067441904e-05, "loss": 0.0428, "step": 12570 }, { "grad_norm": 0.2186751812696457, "learning_rate": 6.712344831895074e-05, "loss": 0.0423, "step": 12580 }, { "grad_norm": 0.20645152032375336, "learning_rate": 6.707165515687811e-05, "loss": 0.0475, "step": 12590 }, { "grad_norm": 0.2761061489582062, "learning_rate": 6.70198412511347e-05, "loss": 0.0414, "step": 12600 }, { "grad_norm": 0.30351167917251587, "learning_rate": 6.69680066646793e-05, "loss": 0.044, "step": 12610 }, { "grad_norm": 0.22715266048908234, "learning_rate": 6.691615146049583e-05, "loss": 0.0395, "step": 12620 }, { "grad_norm": 0.176068514585495, "learning_rate": 6.686427570159324e-05, "loss": 0.029, "step": 12630 }, { "grad_norm": 0.295550674200058, "learning_rate": 6.681237945100548e-05, "loss": 0.0534, "step": 12640 }, { "grad_norm": 0.2280937284231186, "learning_rate": 6.676046277179139e-05, "loss": 0.0363, "step": 12650 }, { "grad_norm": 0.24965718388557434, "learning_rate": 6.670852572703462e-05, "loss": 0.0508, "step": 12660 }, { "grad_norm": 0.21874406933784485, "learning_rate": 6.665656837984359e-05, "loss": 0.043, "step": 12670 }, { "grad_norm": 0.262835830450058, "learning_rate": 6.660459079335135e-05, "loss": 0.0557, "step": 12680 }, { "grad_norm": 0.26292774081230164, "learning_rate": 6.655259303071558e-05, "loss": 0.0458, "step": 12690 }, { "grad_norm": 0.3098023533821106, "learning_rate": 6.650057515511848e-05, "loss": 0.0428, "step": 12700 }, { "grad_norm": 0.16780833899974823, "learning_rate": 6.644853722976667e-05, "loss": 0.0353, "step": 12710 }, { "grad_norm": 0.2477322518825531, "learning_rate": 6.639647931789113e-05, "loss": 0.0459, "step": 12720 }, { "grad_norm": 0.2533622980117798, "learning_rate": 6.634440148274713e-05, "loss": 0.0369, "step": 12730 }, { "grad_norm": 0.2946830689907074, "learning_rate": 6.629230378761415e-05, "loss": 0.0466, "step": 12740 }, { "grad_norm": 0.2583003044128418, "learning_rate": 6.624018629579581e-05, "loss": 0.0367, "step": 12750 }, { "grad_norm": 0.3376270532608032, "learning_rate": 6.618804907061976e-05, "loss": 0.0665, "step": 12760 }, { "grad_norm": 0.1828286349773407, "learning_rate": 6.613589217543766e-05, "loss": 0.0579, "step": 12770 }, { "grad_norm": 0.3600698411464691, "learning_rate": 6.608371567362504e-05, "loss": 0.0501, "step": 12780 }, { "grad_norm": 0.2311152219772339, "learning_rate": 6.60315196285813e-05, "loss": 0.045, "step": 12790 }, { "grad_norm": 0.2380017638206482, "learning_rate": 6.597930410372954e-05, "loss": 0.0466, "step": 12800 }, { "grad_norm": 0.1733928918838501, "learning_rate": 6.592706916251653e-05, "loss": 0.0452, "step": 12810 }, { "grad_norm": 0.23156650364398956, "learning_rate": 6.587481486841267e-05, "loss": 0.0483, "step": 12820 }, { "grad_norm": 0.2455449253320694, "learning_rate": 6.582254128491184e-05, "loss": 0.0415, "step": 12830 }, { "grad_norm": 0.27493777871131897, "learning_rate": 6.577024847553139e-05, "loss": 0.038, "step": 12840 }, { "grad_norm": 0.17757564783096313, "learning_rate": 6.571793650381201e-05, "loss": 0.0312, "step": 12850 }, { "grad_norm": 0.2842480540275574, "learning_rate": 6.566560543331766e-05, "loss": 0.0525, "step": 12860 }, { "grad_norm": 0.2238079160451889, "learning_rate": 6.561325532763554e-05, "loss": 0.0309, "step": 12870 }, { "grad_norm": 0.26320672035217285, "learning_rate": 6.556088625037597e-05, "loss": 0.033, "step": 12880 }, { "grad_norm": 0.18543098866939545, "learning_rate": 6.55084982651723e-05, "loss": 0.0371, "step": 12890 }, { "grad_norm": 0.26198095083236694, "learning_rate": 6.54560914356809e-05, "loss": 0.0393, "step": 12900 }, { "grad_norm": 0.17788954079151154, "learning_rate": 6.540366582558097e-05, "loss": 0.0296, "step": 12910 }, { "grad_norm": 0.209189772605896, "learning_rate": 6.53512214985746e-05, "loss": 0.0279, "step": 12920 }, { "grad_norm": 0.21414706110954285, "learning_rate": 6.529875851838658e-05, "loss": 0.0421, "step": 12930 }, { "grad_norm": 0.34116730093955994, "learning_rate": 6.52462769487644e-05, "loss": 0.0455, "step": 12940 }, { "grad_norm": 0.15826067328453064, "learning_rate": 6.519377685347807e-05, "loss": 0.0353, "step": 12950 }, { "grad_norm": 0.2656620442867279, "learning_rate": 6.514125829632021e-05, "loss": 0.0328, "step": 12960 }, { "grad_norm": 0.20195116102695465, "learning_rate": 6.508872134110577e-05, "loss": 0.0412, "step": 12970 }, { "grad_norm": 0.16112206876277924, "learning_rate": 6.503616605167213e-05, "loss": 0.0338, "step": 12980 }, { "grad_norm": 0.15051321685314178, "learning_rate": 6.498359249187893e-05, "loss": 0.0354, "step": 12990 }, { "grad_norm": 0.22857053577899933, "learning_rate": 6.493100072560799e-05, "loss": 0.0309, "step": 13000 }, { "grad_norm": 0.27193185687065125, "learning_rate": 6.487839081676327e-05, "loss": 0.06, "step": 13010 }, { "grad_norm": 0.22366248071193695, "learning_rate": 6.482576282927076e-05, "loss": 0.0419, "step": 13020 }, { "grad_norm": 0.3100625276565552, "learning_rate": 6.477311682707845e-05, "loss": 0.0566, "step": 13030 }, { "grad_norm": 0.2141428142786026, "learning_rate": 6.472045287415616e-05, "loss": 0.0348, "step": 13040 }, { "grad_norm": 0.25038474798202515, "learning_rate": 6.466777103449558e-05, "loss": 0.029, "step": 13050 }, { "grad_norm": 0.20111817121505737, "learning_rate": 6.461507137211012e-05, "loss": 0.0343, "step": 13060 }, { "grad_norm": 0.2026173621416092, "learning_rate": 6.456235395103483e-05, "loss": 0.0529, "step": 13070 }, { "grad_norm": 0.2870531678199768, "learning_rate": 6.450961883532634e-05, "loss": 0.0358, "step": 13080 }, { "grad_norm": 0.22877317667007446, "learning_rate": 6.445686608906283e-05, "loss": 0.0438, "step": 13090 }, { "grad_norm": 0.17588594555854797, "learning_rate": 6.44040957763438e-05, "loss": 0.0327, "step": 13100 }, { "grad_norm": 0.17877669632434845, "learning_rate": 6.435130796129018e-05, "loss": 0.0318, "step": 13110 }, { "grad_norm": 0.23388399183750153, "learning_rate": 6.429850270804416e-05, "loss": 0.0315, "step": 13120 }, { "grad_norm": 0.26680952310562134, "learning_rate": 6.424568008076909e-05, "loss": 0.0463, "step": 13130 }, { "grad_norm": 0.2454134076833725, "learning_rate": 6.419284014364943e-05, "loss": 0.0407, "step": 13140 }, { "grad_norm": 0.2529073655605316, "learning_rate": 6.41399829608907e-05, "loss": 0.0359, "step": 13150 }, { "grad_norm": 0.1678391844034195, "learning_rate": 6.408710859671938e-05, "loss": 0.0242, "step": 13160 }, { "grad_norm": 0.2166416347026825, "learning_rate": 6.403421711538278e-05, "loss": 0.0385, "step": 13170 }, { "grad_norm": 0.1579716056585312, "learning_rate": 6.398130858114903e-05, "loss": 0.0391, "step": 13180 }, { "grad_norm": 0.16446398198604584, "learning_rate": 6.392838305830701e-05, "loss": 0.0304, "step": 13190 }, { "grad_norm": 0.24959097802639008, "learning_rate": 6.387544061116621e-05, "loss": 0.0456, "step": 13200 }, { "grad_norm": 0.1959265172481537, "learning_rate": 6.382248130405671e-05, "loss": 0.0251, "step": 13210 }, { "grad_norm": 0.2797122895717621, "learning_rate": 6.376950520132905e-05, "loss": 0.0341, "step": 13220 }, { "grad_norm": 0.23010745644569397, "learning_rate": 6.371651236735418e-05, "loss": 0.0352, "step": 13230 }, { "grad_norm": 0.30470213294029236, "learning_rate": 6.366350286652341e-05, "loss": 0.0436, "step": 13240 }, { "grad_norm": 0.26293066143989563, "learning_rate": 6.361047676324827e-05, "loss": 0.0326, "step": 13250 }, { "grad_norm": 0.2388153076171875, "learning_rate": 6.355743412196047e-05, "loss": 0.0344, "step": 13260 }, { "grad_norm": 0.2736150622367859, "learning_rate": 6.350437500711184e-05, "loss": 0.0354, "step": 13270 }, { "grad_norm": 0.25424832105636597, "learning_rate": 6.345129948317418e-05, "loss": 0.0366, "step": 13280 }, { "grad_norm": 0.23282268643379211, "learning_rate": 6.33982076146393e-05, "loss": 0.036, "step": 13290 }, { "grad_norm": 0.19793224334716797, "learning_rate": 6.334509946601879e-05, "loss": 0.0331, "step": 13300 }, { "grad_norm": 0.22775475680828094, "learning_rate": 6.329197510184406e-05, "loss": 0.0344, "step": 13310 }, { "grad_norm": 0.3382323086261749, "learning_rate": 6.323883458666624e-05, "loss": 0.0475, "step": 13320 }, { "grad_norm": 0.24822965264320374, "learning_rate": 6.318567798505605e-05, "loss": 0.0299, "step": 13330 }, { "grad_norm": 0.30606934428215027, "learning_rate": 6.313250536160378e-05, "loss": 0.0379, "step": 13340 }, { "grad_norm": 0.2354898601770401, "learning_rate": 6.307931678091917e-05, "loss": 0.027, "step": 13350 }, { "grad_norm": 0.14283324778079987, "learning_rate": 6.302611230763138e-05, "loss": 0.0337, "step": 13360 }, { "grad_norm": 0.2780652940273285, "learning_rate": 6.297289200638887e-05, "loss": 0.034, "step": 13370 }, { "grad_norm": 0.3744414448738098, "learning_rate": 6.29196559418593e-05, "loss": 0.0373, "step": 13380 }, { "grad_norm": 0.27270567417144775, "learning_rate": 6.286640417872951e-05, "loss": 0.0363, "step": 13390 }, { "grad_norm": 0.19098936021327972, "learning_rate": 6.281313678170542e-05, "loss": 0.0422, "step": 13400 }, { "grad_norm": 0.21677039563655853, "learning_rate": 6.275985381551194e-05, "loss": 0.0326, "step": 13410 }, { "grad_norm": 0.18409425020217896, "learning_rate": 6.270655534489292e-05, "loss": 0.031, "step": 13420 }, { "grad_norm": 0.2627837359905243, "learning_rate": 6.265324143461098e-05, "loss": 0.0234, "step": 13430 }, { "grad_norm": 0.18557144701480865, "learning_rate": 6.259991214944758e-05, "loss": 0.0307, "step": 13440 }, { "grad_norm": 0.3057144582271576, "learning_rate": 6.254656755420283e-05, "loss": 0.048, "step": 13450 }, { "grad_norm": 0.1474284678697586, "learning_rate": 6.249320771369545e-05, "loss": 0.0496, "step": 13460 }, { "grad_norm": 0.18485069274902344, "learning_rate": 6.243983269276263e-05, "loss": 0.0564, "step": 13470 }, { "grad_norm": 0.22214427590370178, "learning_rate": 6.238644255626012e-05, "loss": 0.0455, "step": 13480 }, { "grad_norm": 0.1736307591199875, "learning_rate": 6.233303736906193e-05, "loss": 0.0355, "step": 13490 }, { "grad_norm": 0.20857229828834534, "learning_rate": 6.227961719606039e-05, "loss": 0.0286, "step": 13500 }, { "grad_norm": 0.20561181008815765, "learning_rate": 6.222618210216608e-05, "loss": 0.041, "step": 13510 }, { "grad_norm": 0.1499837338924408, "learning_rate": 6.217273215230767e-05, "loss": 0.0374, "step": 13520 }, { "grad_norm": 0.18792781233787537, "learning_rate": 6.211926741143188e-05, "loss": 0.0489, "step": 13530 }, { "grad_norm": 0.2103004902601242, "learning_rate": 6.20657879445034e-05, "loss": 0.0426, "step": 13540 }, { "grad_norm": 0.2041206657886505, "learning_rate": 6.201229381650485e-05, "loss": 0.0265, "step": 13550 }, { "grad_norm": 0.2430030256509781, "learning_rate": 6.195878509243661e-05, "loss": 0.0431, "step": 13560 }, { "grad_norm": 0.22608217597007751, "learning_rate": 6.190526183731685e-05, "loss": 0.0364, "step": 13570 }, { "grad_norm": 0.23398171365261078, "learning_rate": 6.185172411618138e-05, "loss": 0.0337, "step": 13580 }, { "grad_norm": 0.21345075964927673, "learning_rate": 6.179817199408355e-05, "loss": 0.039, "step": 13590 }, { "grad_norm": 0.21894922852516174, "learning_rate": 6.174460553609426e-05, "loss": 0.0675, "step": 13600 }, { "grad_norm": 0.20099259912967682, "learning_rate": 6.16910248073018e-05, "loss": 0.0277, "step": 13610 }, { "grad_norm": 0.22411102056503296, "learning_rate": 6.16374298728118e-05, "loss": 0.0502, "step": 13620 }, { "grad_norm": 0.18892306089401245, "learning_rate": 6.158382079774716e-05, "loss": 0.0356, "step": 13630 }, { "grad_norm": 0.23678375780582428, "learning_rate": 6.153019764724799e-05, "loss": 0.0339, "step": 13640 }, { "grad_norm": 0.17481276392936707, "learning_rate": 6.147656048647144e-05, "loss": 0.0258, "step": 13650 }, { "grad_norm": 0.32484325766563416, "learning_rate": 6.142290938059173e-05, "loss": 0.0435, "step": 13660 }, { "grad_norm": 0.20631511509418488, "learning_rate": 6.136924439480001e-05, "loss": 0.0308, "step": 13670 }, { "grad_norm": 0.2821146845817566, "learning_rate": 6.13155655943043e-05, "loss": 0.0435, "step": 13680 }, { "grad_norm": 0.3278275728225708, "learning_rate": 6.126187304432941e-05, "loss": 0.0445, "step": 13690 }, { "grad_norm": 0.25184667110443115, "learning_rate": 6.120816681011684e-05, "loss": 0.0344, "step": 13700 }, { "grad_norm": 0.2556883692741394, "learning_rate": 6.115444695692474e-05, "loss": 0.0392, "step": 13710 }, { "grad_norm": 0.162822425365448, "learning_rate": 6.110071355002779e-05, "loss": 0.0299, "step": 13720 }, { "grad_norm": 0.20809780061244965, "learning_rate": 6.104696665471714e-05, "loss": 0.031, "step": 13730 }, { "grad_norm": 0.19980672001838684, "learning_rate": 6.099320633630036e-05, "loss": 0.036, "step": 13740 }, { "grad_norm": 0.22772574424743652, "learning_rate": 6.093943266010128e-05, "loss": 0.0346, "step": 13750 }, { "grad_norm": 0.26900139451026917, "learning_rate": 6.088564569146e-05, "loss": 0.052, "step": 13760 }, { "grad_norm": 0.23240606486797333, "learning_rate": 6.083184549573274e-05, "loss": 0.0384, "step": 13770 }, { "grad_norm": 0.19179531931877136, "learning_rate": 6.077803213829184e-05, "loss": 0.0355, "step": 13780 }, { "grad_norm": 0.24418708682060242, "learning_rate": 6.0724205684525594e-05, "loss": 0.0396, "step": 13790 }, { "grad_norm": 0.2597847878932953, "learning_rate": 6.0670366199838215e-05, "loss": 0.0332, "step": 13800 }, { "grad_norm": 0.16759580373764038, "learning_rate": 6.061651374964974e-05, "loss": 0.0428, "step": 13810 }, { "grad_norm": 0.16963791847229004, "learning_rate": 6.0562648399396004e-05, "loss": 0.0362, "step": 13820 }, { "grad_norm": 0.23696628212928772, "learning_rate": 6.050877021452845e-05, "loss": 0.0468, "step": 13830 }, { "grad_norm": 0.220236673951149, "learning_rate": 6.04548792605142e-05, "loss": 0.0413, "step": 13840 }, { "grad_norm": 0.2609077990055084, "learning_rate": 6.04009756028358e-05, "loss": 0.0325, "step": 13850 }, { "grad_norm": 0.1692759245634079, "learning_rate": 6.0347059306991294e-05, "loss": 0.0422, "step": 13860 }, { "grad_norm": 0.21413341164588928, "learning_rate": 6.029313043849407e-05, "loss": 0.0441, "step": 13870 }, { "grad_norm": 0.19262272119522095, "learning_rate": 6.023918906287279e-05, "loss": 0.0328, "step": 13880 }, { "grad_norm": 0.33383285999298096, "learning_rate": 6.01852352456713e-05, "loss": 0.0574, "step": 13890 }, { "grad_norm": 0.19612327218055725, "learning_rate": 6.013126905244858e-05, "loss": 0.0324, "step": 13900 }, { "grad_norm": 0.2641826272010803, "learning_rate": 6.007729054877864e-05, "loss": 0.0453, "step": 13910 }, { "grad_norm": 0.2316683828830719, "learning_rate": 6.0023299800250466e-05, "loss": 0.0443, "step": 13920 }, { "grad_norm": 0.29425105452537537, "learning_rate": 5.9969296872467894e-05, "loss": 0.0424, "step": 13930 }, { "grad_norm": 0.23025363683700562, "learning_rate": 5.9915281831049585e-05, "loss": 0.0418, "step": 13940 }, { "grad_norm": 0.24371972680091858, "learning_rate": 5.98612547416289e-05, "loss": 0.0461, "step": 13950 }, { "grad_norm": 0.2405940145254135, "learning_rate": 5.9807215669853855e-05, "loss": 0.0511, "step": 13960 }, { "grad_norm": 0.2209792286157608, "learning_rate": 5.9753164681387e-05, "loss": 0.066, "step": 13970 }, { "grad_norm": 0.25722190737724304, "learning_rate": 5.969910184190539e-05, "loss": 0.051, "step": 13980 }, { "grad_norm": 0.22887277603149414, "learning_rate": 5.9645027217100477e-05, "loss": 0.0515, "step": 13990 }, { "grad_norm": 0.24642027914524078, "learning_rate": 5.959094087267804e-05, "loss": 0.0544, "step": 14000 }, { "grad_norm": 0.18771304190158844, "learning_rate": 5.9536842874358065e-05, "loss": 0.0442, "step": 14010 }, { "grad_norm": 0.2672604024410248, "learning_rate": 5.9482733287874734e-05, "loss": 0.0498, "step": 14020 }, { "grad_norm": 0.24038641154766083, "learning_rate": 5.942861217897631e-05, "loss": 0.0634, "step": 14030 }, { "grad_norm": 0.27619436383247375, "learning_rate": 5.9374479613425e-05, "loss": 0.042, "step": 14040 }, { "grad_norm": 0.23460906744003296, "learning_rate": 5.932033565699704e-05, "loss": 0.0389, "step": 14050 }, { "grad_norm": 0.15453466773033142, "learning_rate": 5.926618037548237e-05, "loss": 0.0505, "step": 14060 }, { "grad_norm": 0.25585466623306274, "learning_rate": 5.9212013834684824e-05, "loss": 0.0533, "step": 14070 }, { "grad_norm": 0.25934070348739624, "learning_rate": 5.9157836100421824e-05, "loss": 0.0479, "step": 14080 }, { "grad_norm": 0.20040905475616455, "learning_rate": 5.910364723852444e-05, "loss": 0.0428, "step": 14090 }, { "grad_norm": 0.1816982477903366, "learning_rate": 5.904944731483723e-05, "loss": 0.0509, "step": 14100 }, { "grad_norm": 0.2069862335920334, "learning_rate": 5.899523639521825e-05, "loss": 0.0439, "step": 14110 }, { "grad_norm": 0.22540567815303802, "learning_rate": 5.894101454553883e-05, "loss": 0.0418, "step": 14120 }, { "grad_norm": 0.33092325925827026, "learning_rate": 5.888678183168368e-05, "loss": 0.0533, "step": 14130 }, { "grad_norm": 0.24165059626102448, "learning_rate": 5.883253831955061e-05, "loss": 0.0537, "step": 14140 }, { "grad_norm": 0.23912487924098969, "learning_rate": 5.8778284075050625e-05, "loss": 0.0391, "step": 14150 }, { "grad_norm": 0.2884124517440796, "learning_rate": 5.872401916410777e-05, "loss": 0.056, "step": 14160 }, { "grad_norm": 0.23026712238788605, "learning_rate": 5.8669743652659014e-05, "loss": 0.048, "step": 14170 }, { "grad_norm": 0.1751270890235901, "learning_rate": 5.86154576066542e-05, "loss": 0.0447, "step": 14180 }, { "grad_norm": 0.2757657766342163, "learning_rate": 5.856116109205602e-05, "loss": 0.0525, "step": 14190 }, { "grad_norm": 0.2318994551897049, "learning_rate": 5.850685417483983e-05, "loss": 0.0443, "step": 14200 }, { "grad_norm": 0.202314093708992, "learning_rate": 5.8452536920993684e-05, "loss": 0.0478, "step": 14210 }, { "grad_norm": 0.2020840346813202, "learning_rate": 5.839820939651817e-05, "loss": 0.0546, "step": 14220 }, { "grad_norm": 0.2128271758556366, "learning_rate": 5.834387166742632e-05, "loss": 0.0474, "step": 14230 }, { "grad_norm": 0.26261764764785767, "learning_rate": 5.828952379974364e-05, "loss": 0.046, "step": 14240 }, { "grad_norm": 0.2828563451766968, "learning_rate": 5.8235165859507864e-05, "loss": 0.0548, "step": 14250 }, { "grad_norm": 0.21041357517242432, "learning_rate": 5.818079791276907e-05, "loss": 0.0435, "step": 14260 }, { "grad_norm": 0.1936400830745697, "learning_rate": 5.8126420025589415e-05, "loss": 0.0536, "step": 14270 }, { "grad_norm": 0.2391810268163681, "learning_rate": 5.807203226404313e-05, "loss": 0.0367, "step": 14280 }, { "grad_norm": 0.18931831419467926, "learning_rate": 5.801763469421652e-05, "loss": 0.0315, "step": 14290 }, { "grad_norm": 0.19822530448436737, "learning_rate": 5.796322738220774e-05, "loss": 0.0463, "step": 14300 }, { "grad_norm": 0.18052971363067627, "learning_rate": 5.79088103941268e-05, "loss": 0.0319, "step": 14310 }, { "grad_norm": 0.18759305775165558, "learning_rate": 5.785438379609549e-05, "loss": 0.0529, "step": 14320 }, { "grad_norm": 0.24209752678871155, "learning_rate": 5.779994765424724e-05, "loss": 0.0379, "step": 14330 }, { "grad_norm": 0.2872958481311798, "learning_rate": 5.7745502034727114e-05, "loss": 0.0387, "step": 14340 }, { "grad_norm": 0.23387618362903595, "learning_rate": 5.769104700369164e-05, "loss": 0.0408, "step": 14350 }, { "grad_norm": 0.2177293300628662, "learning_rate": 5.7636582627308854e-05, "loss": 0.0376, "step": 14360 }, { "grad_norm": 0.16894161701202393, "learning_rate": 5.7582108971758095e-05, "loss": 0.0457, "step": 14370 }, { "grad_norm": 0.28059402108192444, "learning_rate": 5.752762610322999e-05, "loss": 0.043, "step": 14380 }, { "grad_norm": 0.2401747703552246, "learning_rate": 5.747313408792636e-05, "loss": 0.0571, "step": 14390 }, { "grad_norm": 0.22574836015701294, "learning_rate": 5.741863299206014e-05, "loss": 0.0514, "step": 14400 }, { "grad_norm": 0.1901606172323227, "learning_rate": 5.73641228818553e-05, "loss": 0.0415, "step": 14410 }, { "grad_norm": 0.23655609786510468, "learning_rate": 5.730960382354676e-05, "loss": 0.0427, "step": 14420 }, { "grad_norm": 0.20032235980033875, "learning_rate": 5.725507588338035e-05, "loss": 0.0469, "step": 14430 }, { "grad_norm": 0.2565443217754364, "learning_rate": 5.7200539127612604e-05, "loss": 0.0412, "step": 14440 }, { "grad_norm": 0.2287723422050476, "learning_rate": 5.714599362251087e-05, "loss": 0.0448, "step": 14450 }, { "grad_norm": 0.20677787065505981, "learning_rate": 5.709143943435307e-05, "loss": 0.0455, "step": 14460 }, { "grad_norm": 0.2300729900598526, "learning_rate": 5.7036876629427646e-05, "loss": 0.0484, "step": 14470 }, { "grad_norm": 0.2511281371116638, "learning_rate": 5.698230527403361e-05, "loss": 0.0412, "step": 14480 }, { "grad_norm": 0.2553737163543701, "learning_rate": 5.6927725434480274e-05, "loss": 0.0403, "step": 14490 }, { "grad_norm": 0.2141558974981308, "learning_rate": 5.687313717708728e-05, "loss": 0.06, "step": 14500 }, { "grad_norm": 0.19750818610191345, "learning_rate": 5.681854056818453e-05, "loss": 0.0447, "step": 14510 }, { "grad_norm": 0.27140673995018005, "learning_rate": 5.676393567411205e-05, "loss": 0.0391, "step": 14520 }, { "grad_norm": 0.2691078186035156, "learning_rate": 5.670932256121991e-05, "loss": 0.0657, "step": 14530 }, { "grad_norm": 0.3461399972438812, "learning_rate": 5.6654701295868216e-05, "loss": 0.0375, "step": 14540 }, { "grad_norm": 0.20540601015090942, "learning_rate": 5.660007194442697e-05, "loss": 0.0359, "step": 14550 }, { "grad_norm": 0.23647332191467285, "learning_rate": 5.6545434573275946e-05, "loss": 0.0474, "step": 14560 }, { "grad_norm": 0.17516471445560455, "learning_rate": 5.649078924880472e-05, "loss": 0.035, "step": 14570 }, { "grad_norm": 0.281095951795578, "learning_rate": 5.643613603741252e-05, "loss": 0.0524, "step": 14580 }, { "grad_norm": 0.2130465805530548, "learning_rate": 5.6381475005508154e-05, "loss": 0.0599, "step": 14590 }, { "grad_norm": 0.30053281784057617, "learning_rate": 5.632680621950992e-05, "loss": 0.0473, "step": 14600 }, { "grad_norm": 0.21987885236740112, "learning_rate": 5.627212974584555e-05, "loss": 0.0353, "step": 14610 }, { "grad_norm": 0.3048321604728699, "learning_rate": 5.62174456509521e-05, "loss": 0.0377, "step": 14620 }, { "grad_norm": 0.2118852138519287, "learning_rate": 5.616275400127594e-05, "loss": 0.0363, "step": 14630 }, { "grad_norm": 0.22018705308437347, "learning_rate": 5.610805486327254e-05, "loss": 0.0331, "step": 14640 }, { "grad_norm": 0.24166084825992584, "learning_rate": 5.6053348303406536e-05, "loss": 0.0435, "step": 14650 }, { "grad_norm": 0.2740503251552582, "learning_rate": 5.599863438815156e-05, "loss": 0.0393, "step": 14660 }, { "grad_norm": 0.18368415534496307, "learning_rate": 5.594391318399017e-05, "loss": 0.0469, "step": 14670 }, { "grad_norm": 0.2346373349428177, "learning_rate": 5.588918475741377e-05, "loss": 0.0371, "step": 14680 }, { "grad_norm": 0.21182793378829956, "learning_rate": 5.5834449174922586e-05, "loss": 0.0388, "step": 14690 }, { "grad_norm": 0.25930121541023254, "learning_rate": 5.57797065030255e-05, "loss": 0.0447, "step": 14700 }, { "grad_norm": 0.24734878540039062, "learning_rate": 5.5724956808240016e-05, "loss": 0.0519, "step": 14710 }, { "grad_norm": 0.22324860095977783, "learning_rate": 5.567020015709219e-05, "loss": 0.0431, "step": 14720 }, { "grad_norm": 0.17215146124362946, "learning_rate": 5.561543661611649e-05, "loss": 0.0381, "step": 14730 }, { "grad_norm": 0.21854068338871002, "learning_rate": 5.556066625185583e-05, "loss": 0.0447, "step": 14740 }, { "grad_norm": 0.15432070195674896, "learning_rate": 5.550588913086131e-05, "loss": 0.0462, "step": 14750 }, { "grad_norm": 0.25642073154449463, "learning_rate": 5.545110531969234e-05, "loss": 0.0449, "step": 14760 }, { "grad_norm": 0.15306852757930756, "learning_rate": 5.539631488491641e-05, "loss": 0.036, "step": 14770 }, { "grad_norm": 0.22211147844791412, "learning_rate": 5.534151789310904e-05, "loss": 0.0353, "step": 14780 }, { "grad_norm": 0.20247460901737213, "learning_rate": 5.528671441085376e-05, "loss": 0.026, "step": 14790 }, { "grad_norm": 0.19772638380527496, "learning_rate": 5.523190450474197e-05, "loss": 0.0287, "step": 14800 }, { "grad_norm": 0.3126679062843323, "learning_rate": 5.5177088241372865e-05, "loss": 0.0395, "step": 14810 }, { "grad_norm": 0.27556517720222473, "learning_rate": 5.512226568735338e-05, "loss": 0.037, "step": 14820 }, { "grad_norm": 0.20253241062164307, "learning_rate": 5.506743690929809e-05, "loss": 0.0427, "step": 14830 }, { "grad_norm": 0.22430184483528137, "learning_rate": 5.501260197382912e-05, "loss": 0.0442, "step": 14840 }, { "grad_norm": 0.2609712481498718, "learning_rate": 5.49577609475761e-05, "loss": 0.0552, "step": 14850 }, { "grad_norm": 0.18670204281806946, "learning_rate": 5.490291389717603e-05, "loss": 0.0323, "step": 14860 }, { "grad_norm": 0.13871395587921143, "learning_rate": 5.484806088927329e-05, "loss": 0.0301, "step": 14870 }, { "grad_norm": 0.2550557851791382, "learning_rate": 5.479320199051942e-05, "loss": 0.0378, "step": 14880 }, { "grad_norm": 0.2323555052280426, "learning_rate": 5.473833726757314e-05, "loss": 0.0362, "step": 14890 }, { "grad_norm": 0.2695874571800232, "learning_rate": 5.4683466787100304e-05, "loss": 0.0366, "step": 14900 }, { "grad_norm": 0.22652754187583923, "learning_rate": 5.462859061577369e-05, "loss": 0.0285, "step": 14910 }, { "grad_norm": 0.22265128791332245, "learning_rate": 5.4573708820273026e-05, "loss": 0.0296, "step": 14920 }, { "grad_norm": 0.2621091902256012, "learning_rate": 5.451882146728488e-05, "loss": 0.0311, "step": 14930 }, { "grad_norm": 0.1606391966342926, "learning_rate": 5.446392862350255e-05, "loss": 0.0387, "step": 14940 }, { "grad_norm": 0.21763649582862854, "learning_rate": 5.440903035562603e-05, "loss": 0.0515, "step": 14950 }, { "grad_norm": 0.1955883949995041, "learning_rate": 5.435412673036188e-05, "loss": 0.0295, "step": 14960 }, { "grad_norm": 0.30634036660194397, "learning_rate": 5.429921781442318e-05, "loss": 0.0413, "step": 14970 }, { "grad_norm": 0.1893945336341858, "learning_rate": 5.424430367452945e-05, "loss": 0.0384, "step": 14980 }, { "grad_norm": 0.2609647214412689, "learning_rate": 5.418938437740655e-05, "loss": 0.0408, "step": 14990 }, { "grad_norm": 0.22139419615268707, "learning_rate": 5.413445998978658e-05, "loss": 0.0397, "step": 15000 }, { "grad_norm": 0.24316708743572235, "learning_rate": 5.407953057840789e-05, "loss": 0.0361, "step": 15010 }, { "grad_norm": 0.19356046617031097, "learning_rate": 5.4024596210014853e-05, "loss": 0.0326, "step": 15020 }, { "grad_norm": 0.27790606021881104, "learning_rate": 5.3969656951357935e-05, "loss": 0.034, "step": 15030 }, { "grad_norm": 0.25830790400505066, "learning_rate": 5.3914712869193504e-05, "loss": 0.0458, "step": 15040 }, { "grad_norm": 0.18981517851352692, "learning_rate": 5.385976403028381e-05, "loss": 0.0549, "step": 15050 }, { "grad_norm": 0.20615150034427643, "learning_rate": 5.3804810501396875e-05, "loss": 0.0426, "step": 15060 }, { "grad_norm": 0.1733548492193222, "learning_rate": 5.374985234930639e-05, "loss": 0.0356, "step": 15070 }, { "grad_norm": 0.2139529287815094, "learning_rate": 5.3694889640791724e-05, "loss": 0.0423, "step": 15080 }, { "grad_norm": 0.25642672181129456, "learning_rate": 5.3639922442637736e-05, "loss": 0.0474, "step": 15090 }, { "grad_norm": 0.23717907071113586, "learning_rate": 5.358495082163476e-05, "loss": 0.0464, "step": 15100 }, { "grad_norm": 0.24115340411663055, "learning_rate": 5.35299748445785e-05, "loss": 0.037, "step": 15110 }, { "grad_norm": 0.21627314388751984, "learning_rate": 5.3474994578269944e-05, "loss": 0.0277, "step": 15120 }, { "grad_norm": 0.21980252861976624, "learning_rate": 5.34200100895153e-05, "loss": 0.0323, "step": 15130 }, { "grad_norm": 0.18250951170921326, "learning_rate": 5.3365021445125916e-05, "loss": 0.0225, "step": 15140 }, { "grad_norm": 0.23111744225025177, "learning_rate": 5.331002871191817e-05, "loss": 0.0447, "step": 15150 }, { "grad_norm": 0.19363923370838165, "learning_rate": 5.3255031956713443e-05, "loss": 0.0318, "step": 15160 }, { "grad_norm": 0.207148015499115, "learning_rate": 5.320003124633795e-05, "loss": 0.0377, "step": 15170 }, { "grad_norm": 0.20350833237171173, "learning_rate": 5.314502664762274e-05, "loss": 0.0305, "step": 15180 }, { "grad_norm": 0.19085563719272614, "learning_rate": 5.3090018227403604e-05, "loss": 0.0327, "step": 15190 }, { "grad_norm": 0.2257489413022995, "learning_rate": 5.3035006052520955e-05, "loss": 0.0473, "step": 15200 }, { "grad_norm": 0.19037660956382751, "learning_rate": 5.297999018981976e-05, "loss": 0.033, "step": 15210 }, { "grad_norm": 0.22943733632564545, "learning_rate": 5.2924970706149505e-05, "loss": 0.0293, "step": 15220 }, { "grad_norm": 0.17537996172904968, "learning_rate": 5.2869947668364014e-05, "loss": 0.0237, "step": 15230 }, { "grad_norm": 0.1470334827899933, "learning_rate": 5.2814921143321504e-05, "loss": 0.0316, "step": 15240 }, { "grad_norm": 0.31675368547439575, "learning_rate": 5.275989119788436e-05, "loss": 0.0496, "step": 15250 }, { "grad_norm": 0.22736841440200806, "learning_rate": 5.270485789891919e-05, "loss": 0.0424, "step": 15260 }, { "grad_norm": 0.23021981120109558, "learning_rate": 5.264982131329661e-05, "loss": 0.0317, "step": 15270 }, { "grad_norm": 0.17361243069171906, "learning_rate": 5.259478150789128e-05, "loss": 0.0296, "step": 15280 }, { "grad_norm": 0.25106680393218994, "learning_rate": 5.253973854958173e-05, "loss": 0.0489, "step": 15290 }, { "grad_norm": 0.21685688197612762, "learning_rate": 5.2484692505250376e-05, "loss": 0.0373, "step": 15300 }, { "grad_norm": 0.18189367651939392, "learning_rate": 5.242964344178333e-05, "loss": 0.0382, "step": 15310 }, { "grad_norm": 0.19547368586063385, "learning_rate": 5.237459142607041e-05, "loss": 0.0257, "step": 15320 }, { "grad_norm": 0.20228630304336548, "learning_rate": 5.231953652500498e-05, "loss": 0.0315, "step": 15330 }, { "grad_norm": 0.140726238489151, "learning_rate": 5.2264478805483976e-05, "loss": 0.0352, "step": 15340 }, { "grad_norm": 0.13349156081676483, "learning_rate": 5.220941833440768e-05, "loss": 0.0315, "step": 15350 }, { "grad_norm": 0.194534033536911, "learning_rate": 5.215435517867977e-05, "loss": 0.0237, "step": 15360 }, { "grad_norm": 0.14328493177890778, "learning_rate": 5.209928940520719e-05, "loss": 0.0261, "step": 15370 }, { "grad_norm": 0.12424099445343018, "learning_rate": 5.204422108090004e-05, "loss": 0.0325, "step": 15380 }, { "grad_norm": 0.22946031391620636, "learning_rate": 5.19891502726715e-05, "loss": 0.0215, "step": 15390 }, { "grad_norm": 0.1508287787437439, "learning_rate": 5.193407704743782e-05, "loss": 0.0268, "step": 15400 }, { "grad_norm": 0.17863206565380096, "learning_rate": 5.187900147211815e-05, "loss": 0.02, "step": 15410 }, { "grad_norm": 0.18380852043628693, "learning_rate": 5.182392361363453e-05, "loss": 0.0305, "step": 15420 }, { "grad_norm": 0.24871328473091125, "learning_rate": 5.176884353891172e-05, "loss": 0.0528, "step": 15430 }, { "grad_norm": 0.17719528079032898, "learning_rate": 5.171376131487722e-05, "loss": 0.0286, "step": 15440 }, { "grad_norm": 0.1915503293275833, "learning_rate": 5.165867700846113e-05, "loss": 0.0382, "step": 15450 }, { "grad_norm": 0.2137172967195511, "learning_rate": 5.160359068659606e-05, "loss": 0.0302, "step": 15460 }, { "grad_norm": 0.24854391813278198, "learning_rate": 5.154850241621712e-05, "loss": 0.0258, "step": 15470 }, { "grad_norm": 0.18732157349586487, "learning_rate": 5.149341226426172e-05, "loss": 0.0292, "step": 15480 }, { "grad_norm": 0.22397463023662567, "learning_rate": 5.14383202976696e-05, "loss": 0.0313, "step": 15490 }, { "grad_norm": 0.20543788373470306, "learning_rate": 5.1383226583382684e-05, "loss": 0.0229, "step": 15500 }, { "grad_norm": 0.25860118865966797, "learning_rate": 5.132813118834504e-05, "loss": 0.0377, "step": 15510 }, { "grad_norm": 0.2248518019914627, "learning_rate": 5.1273034179502777e-05, "loss": 0.0208, "step": 15520 }, { "grad_norm": 0.26968827843666077, "learning_rate": 5.1217935623803945e-05, "loss": 0.0286, "step": 15530 }, { "grad_norm": 0.22853006422519684, "learning_rate": 5.1162835588198476e-05, "loss": 0.0319, "step": 15540 }, { "grad_norm": 0.1584775149822235, "learning_rate": 5.110773413963813e-05, "loss": 0.0312, "step": 15550 }, { "grad_norm": 0.2278273105621338, "learning_rate": 5.105263134507636e-05, "loss": 0.0311, "step": 15560 }, { "grad_norm": 0.2809124290943146, "learning_rate": 5.099752727146824e-05, "loss": 0.041, "step": 15570 }, { "grad_norm": 0.18524205684661865, "learning_rate": 5.094242198577042e-05, "loss": 0.0344, "step": 15580 }, { "grad_norm": 0.13688844442367554, "learning_rate": 5.088731555494102e-05, "loss": 0.0278, "step": 15590 }, { "grad_norm": 0.19823113083839417, "learning_rate": 5.0832208045939556e-05, "loss": 0.0329, "step": 15600 }, { "grad_norm": 0.23817162215709686, "learning_rate": 5.0777099525726844e-05, "loss": 0.0332, "step": 15610 }, { "grad_norm": 0.1601613610982895, "learning_rate": 5.072199006126494e-05, "loss": 0.0238, "step": 15620 }, { "grad_norm": 0.4294241964817047, "learning_rate": 5.0666879719517026e-05, "loss": 0.0338, "step": 15630 }, { "grad_norm": 0.19021837413311005, "learning_rate": 5.061176856744737e-05, "loss": 0.0297, "step": 15640 }, { "grad_norm": 0.12740124762058258, "learning_rate": 5.0556656672021205e-05, "loss": 0.0205, "step": 15650 }, { "grad_norm": 0.2737583816051483, "learning_rate": 5.050154410020472e-05, "loss": 0.0363, "step": 15660 }, { "grad_norm": 0.17844900488853455, "learning_rate": 5.044643091896485e-05, "loss": 0.0325, "step": 15670 }, { "grad_norm": 0.19980019330978394, "learning_rate": 5.039131719526932e-05, "loss": 0.0225, "step": 15680 }, { "grad_norm": 0.20457105338573456, "learning_rate": 5.03362029960865e-05, "loss": 0.0296, "step": 15690 }, { "grad_norm": 0.22462224960327148, "learning_rate": 5.028108838838533e-05, "loss": 0.0221, "step": 15700 }, { "grad_norm": 0.23789110779762268, "learning_rate": 5.0225973439135276e-05, "loss": 0.0306, "step": 15710 }, { "grad_norm": 0.31748542189598083, "learning_rate": 5.017085821530617e-05, "loss": 0.0262, "step": 15720 }, { "grad_norm": 0.24862779676914215, "learning_rate": 5.011574278386822e-05, "loss": 0.0404, "step": 15730 }, { "grad_norm": 0.22122421860694885, "learning_rate": 5.006062721179189e-05, "loss": 0.0275, "step": 15740 }, { "grad_norm": 0.17756973206996918, "learning_rate": 5.000551156604777e-05, "loss": 0.0322, "step": 15750 }, { "grad_norm": 0.18257424235343933, "learning_rate": 4.9950395913606594e-05, "loss": 0.0292, "step": 15760 }, { "grad_norm": 0.30035004019737244, "learning_rate": 4.989528032143903e-05, "loss": 0.0261, "step": 15770 }, { "grad_norm": 0.2337464839220047, "learning_rate": 4.984016485651578e-05, "loss": 0.0306, "step": 15780 }, { "grad_norm": 0.3577556908130646, "learning_rate": 4.9785049585807274e-05, "loss": 0.0433, "step": 15790 }, { "grad_norm": 0.20973412692546844, "learning_rate": 4.972993457628381e-05, "loss": 0.0371, "step": 15800 }, { "grad_norm": 0.18818731606006622, "learning_rate": 4.9674819894915306e-05, "loss": 0.0291, "step": 15810 }, { "grad_norm": 0.21062414348125458, "learning_rate": 4.9619705608671265e-05, "loss": 0.0335, "step": 15820 }, { "grad_norm": 0.20797160267829895, "learning_rate": 4.956459178452079e-05, "loss": 0.0322, "step": 15830 }, { "grad_norm": 0.18805018067359924, "learning_rate": 4.950947848943235e-05, "loss": 0.0313, "step": 15840 }, { "grad_norm": 0.1985694169998169, "learning_rate": 4.9454365790373805e-05, "loss": 0.0323, "step": 15850 }, { "grad_norm": 0.18034890294075012, "learning_rate": 4.939925375431226e-05, "loss": 0.038, "step": 15860 }, { "grad_norm": 0.18013574182987213, "learning_rate": 4.9344142448214046e-05, "loss": 0.0379, "step": 15870 }, { "grad_norm": 0.18831756711006165, "learning_rate": 4.928903193904461e-05, "loss": 0.031, "step": 15880 }, { "grad_norm": 0.2789424955844879, "learning_rate": 4.923392229376842e-05, "loss": 0.0416, "step": 15890 }, { "grad_norm": 0.2016260176897049, "learning_rate": 4.9178813579348865e-05, "loss": 0.0268, "step": 15900 }, { "grad_norm": 0.2038709968328476, "learning_rate": 4.9123705862748253e-05, "loss": 0.0288, "step": 15910 }, { "grad_norm": 0.1835741251707077, "learning_rate": 4.9068599210927627e-05, "loss": 0.0336, "step": 15920 }, { "grad_norm": 0.14978674054145813, "learning_rate": 4.901349369084681e-05, "loss": 0.0378, "step": 15930 }, { "grad_norm": 0.1662510633468628, "learning_rate": 4.8958389369464156e-05, "loss": 0.0337, "step": 15940 }, { "grad_norm": 0.197488933801651, "learning_rate": 4.890328631373666e-05, "loss": 0.0289, "step": 15950 }, { "grad_norm": 0.1613176167011261, "learning_rate": 4.88481845906197e-05, "loss": 0.0318, "step": 15960 }, { "grad_norm": 0.15729044377803802, "learning_rate": 4.879308426706707e-05, "loss": 0.0418, "step": 15970 }, { "grad_norm": 0.16356970369815826, "learning_rate": 4.873798541003084e-05, "loss": 0.0229, "step": 15980 }, { "grad_norm": 0.22156411409378052, "learning_rate": 4.868288808646136e-05, "loss": 0.0466, "step": 15990 }, { "grad_norm": 0.19999949634075165, "learning_rate": 4.862779236330705e-05, "loss": 0.0354, "step": 16000 }, { "grad_norm": 0.23114009201526642, "learning_rate": 4.8572698307514395e-05, "loss": 0.0473, "step": 16010 }, { "grad_norm": 0.25993239879608154, "learning_rate": 4.85176059860279e-05, "loss": 0.0434, "step": 16020 }, { "grad_norm": 0.20156438648700714, "learning_rate": 4.846251546578989e-05, "loss": 0.0474, "step": 16030 }, { "grad_norm": 0.17206701636314392, "learning_rate": 4.840742681374058e-05, "loss": 0.0413, "step": 16040 }, { "grad_norm": 0.22952349483966827, "learning_rate": 4.835234009681787e-05, "loss": 0.038, "step": 16050 }, { "grad_norm": 0.2570783793926239, "learning_rate": 4.8297255381957296e-05, "loss": 0.0381, "step": 16060 }, { "grad_norm": 0.25052279233932495, "learning_rate": 4.824217273609198e-05, "loss": 0.038, "step": 16070 }, { "grad_norm": 0.1992686539888382, "learning_rate": 4.8187092226152555e-05, "loss": 0.0355, "step": 16080 }, { "grad_norm": 0.25340571999549866, "learning_rate": 4.8132013919067016e-05, "loss": 0.042, "step": 16090 }, { "grad_norm": 0.1953749805688858, "learning_rate": 4.807693788176071e-05, "loss": 0.0428, "step": 16100 }, { "grad_norm": 0.20592743158340454, "learning_rate": 4.8021864181156214e-05, "loss": 0.0423, "step": 16110 }, { "grad_norm": 0.24996092915534973, "learning_rate": 4.796679288417326e-05, "loss": 0.0553, "step": 16120 }, { "grad_norm": 0.2823004722595215, "learning_rate": 4.791172405772866e-05, "loss": 0.0502, "step": 16130 }, { "grad_norm": 0.24368084967136383, "learning_rate": 4.785665776873625e-05, "loss": 0.0339, "step": 16140 }, { "grad_norm": 0.1881355196237564, "learning_rate": 4.7801594084106763e-05, "loss": 0.049, "step": 16150 }, { "grad_norm": 0.18709814548492432, "learning_rate": 4.774653307074775e-05, "loss": 0.041, "step": 16160 }, { "grad_norm": 0.22618357837200165, "learning_rate": 4.769147479556355e-05, "loss": 0.0341, "step": 16170 }, { "grad_norm": 0.21321657299995422, "learning_rate": 4.763641932545515e-05, "loss": 0.0299, "step": 16180 }, { "grad_norm": 0.2529887855052948, "learning_rate": 4.758136672732013e-05, "loss": 0.0443, "step": 16190 }, { "grad_norm": 0.16545626521110535, "learning_rate": 4.752631706805261e-05, "loss": 0.0362, "step": 16200 }, { "grad_norm": 0.18176037073135376, "learning_rate": 4.7471270414543114e-05, "loss": 0.0538, "step": 16210 }, { "grad_norm": 0.19996878504753113, "learning_rate": 4.741622683367849e-05, "loss": 0.0353, "step": 16220 }, { "grad_norm": 0.20248398184776306, "learning_rate": 4.736118639234191e-05, "loss": 0.0556, "step": 16230 }, { "grad_norm": 0.23256130516529083, "learning_rate": 4.730614915741267e-05, "loss": 0.0513, "step": 16240 }, { "grad_norm": 0.17439186573028564, "learning_rate": 4.7251115195766234e-05, "loss": 0.0418, "step": 16250 }, { "grad_norm": 0.17631998658180237, "learning_rate": 4.719608457427404e-05, "loss": 0.0411, "step": 16260 }, { "grad_norm": 0.23560506105422974, "learning_rate": 4.714105735980347e-05, "loss": 0.034, "step": 16270 }, { "grad_norm": 0.16569863259792328, "learning_rate": 4.70860336192178e-05, "loss": 0.0404, "step": 16280 }, { "grad_norm": 0.1920316517353058, "learning_rate": 4.7031013419376034e-05, "loss": 0.0414, "step": 16290 }, { "grad_norm": 0.1938362717628479, "learning_rate": 4.6975996827132926e-05, "loss": 0.029, "step": 16300 }, { "grad_norm": 0.23465505242347717, "learning_rate": 4.692098390933883e-05, "loss": 0.0454, "step": 16310 }, { "grad_norm": 0.22367005050182343, "learning_rate": 4.6865974732839615e-05, "loss": 0.0355, "step": 16320 }, { "grad_norm": 0.24409650266170502, "learning_rate": 4.6810969364476615e-05, "loss": 0.0567, "step": 16330 }, { "grad_norm": 0.196989968419075, "learning_rate": 4.675596787108653e-05, "loss": 0.0377, "step": 16340 }, { "grad_norm": 0.21227724850177765, "learning_rate": 4.670097031950139e-05, "loss": 0.0329, "step": 16350 }, { "grad_norm": 0.1497187614440918, "learning_rate": 4.664597677654838e-05, "loss": 0.0403, "step": 16360 }, { "grad_norm": 0.28369638323783875, "learning_rate": 4.659098730904986e-05, "loss": 0.0472, "step": 16370 }, { "grad_norm": 0.1766453981399536, "learning_rate": 4.6536001983823204e-05, "loss": 0.0376, "step": 16380 }, { "grad_norm": 0.20155072212219238, "learning_rate": 4.648102086768078e-05, "loss": 0.058, "step": 16390 }, { "grad_norm": 0.2339235544204712, "learning_rate": 4.6426044027429786e-05, "loss": 0.0458, "step": 16400 }, { "grad_norm": 0.29515689611434937, "learning_rate": 4.637107152987234e-05, "loss": 0.04, "step": 16410 }, { "grad_norm": 0.17282435297966003, "learning_rate": 4.631610344180515e-05, "loss": 0.0327, "step": 16420 }, { "grad_norm": 0.18782255053520203, "learning_rate": 4.626113983001965e-05, "loss": 0.0451, "step": 16430 }, { "grad_norm": 0.22838972508907318, "learning_rate": 4.620618076130182e-05, "loss": 0.0373, "step": 16440 }, { "grad_norm": 0.13673453032970428, "learning_rate": 4.6151226302432074e-05, "loss": 0.0361, "step": 16450 }, { "grad_norm": 0.1780867576599121, "learning_rate": 4.6096276520185304e-05, "loss": 0.0293, "step": 16460 }, { "grad_norm": 0.22124846279621124, "learning_rate": 4.604133148133066e-05, "loss": 0.0336, "step": 16470 }, { "grad_norm": 0.19676972925662994, "learning_rate": 4.598639125263155e-05, "loss": 0.0454, "step": 16480 }, { "grad_norm": 0.199634850025177, "learning_rate": 4.593145590084553e-05, "loss": 0.0478, "step": 16490 }, { "grad_norm": 0.22204986214637756, "learning_rate": 4.58765254927242e-05, "loss": 0.0288, "step": 16500 }, { "grad_norm": 0.17784124612808228, "learning_rate": 4.582160009501323e-05, "loss": 0.0329, "step": 16510 }, { "grad_norm": 0.22105737030506134, "learning_rate": 4.5766679774452143e-05, "loss": 0.0419, "step": 16520 }, { "grad_norm": 0.20960526168346405, "learning_rate": 4.571176459777431e-05, "loss": 0.0394, "step": 16530 }, { "grad_norm": 0.18201638758182526, "learning_rate": 4.5656854631706846e-05, "loss": 0.037, "step": 16540 }, { "grad_norm": 0.19521449506282806, "learning_rate": 4.560194994297054e-05, "loss": 0.0432, "step": 16550 }, { "grad_norm": 0.18197879195213318, "learning_rate": 4.554705059827974e-05, "loss": 0.0402, "step": 16560 }, { "grad_norm": 0.24373213946819305, "learning_rate": 4.5492156664342365e-05, "loss": 0.0442, "step": 16570 }, { "grad_norm": 0.22170308232307434, "learning_rate": 4.5437268207859695e-05, "loss": 0.0482, "step": 16580 }, { "grad_norm": 0.17027218639850616, "learning_rate": 4.538238529552641e-05, "loss": 0.0253, "step": 16590 }, { "grad_norm": 0.15954935550689697, "learning_rate": 4.5327507994030394e-05, "loss": 0.0291, "step": 16600 }, { "grad_norm": 0.19129158556461334, "learning_rate": 4.527263637005273e-05, "loss": 0.0289, "step": 16610 }, { "grad_norm": 0.14617279171943665, "learning_rate": 4.521777049026766e-05, "loss": 0.0275, "step": 16620 }, { "grad_norm": 0.18538282811641693, "learning_rate": 4.5162910421342386e-05, "loss": 0.0283, "step": 16630 }, { "grad_norm": 0.30161309242248535, "learning_rate": 4.510805622993706e-05, "loss": 0.0341, "step": 16640 }, { "grad_norm": 0.2234259992837906, "learning_rate": 4.505320798270467e-05, "loss": 0.0472, "step": 16650 }, { "grad_norm": 0.15000714361667633, "learning_rate": 4.499836574629105e-05, "loss": 0.035, "step": 16660 }, { "grad_norm": 0.13758152723312378, "learning_rate": 4.494352958733465e-05, "loss": 0.0494, "step": 16670 }, { "grad_norm": 0.2439495176076889, "learning_rate": 4.488869957246663e-05, "loss": 0.0423, "step": 16680 }, { "grad_norm": 0.20449699461460114, "learning_rate": 4.4833875768310575e-05, "loss": 0.0385, "step": 16690 }, { "grad_norm": 0.21864454448223114, "learning_rate": 4.47790582414826e-05, "loss": 0.0337, "step": 16700 }, { "grad_norm": 0.21230706572532654, "learning_rate": 4.472424705859114e-05, "loss": 0.0404, "step": 16710 }, { "grad_norm": 0.20778273046016693, "learning_rate": 4.466944228623701e-05, "loss": 0.0479, "step": 16720 }, { "grad_norm": 0.22977210581302643, "learning_rate": 4.461464399101312e-05, "loss": 0.0434, "step": 16730 }, { "grad_norm": 0.14157815277576447, "learning_rate": 4.45598522395046e-05, "loss": 0.0363, "step": 16740 }, { "grad_norm": 0.23720505833625793, "learning_rate": 4.4505067098288577e-05, "loss": 0.0432, "step": 16750 }, { "grad_norm": 0.1773872822523117, "learning_rate": 4.445028863393417e-05, "loss": 0.0307, "step": 16760 }, { "grad_norm": 0.21853715181350708, "learning_rate": 4.439551691300236e-05, "loss": 0.0329, "step": 16770 }, { "grad_norm": 0.2459811568260193, "learning_rate": 4.434075200204599e-05, "loss": 0.0378, "step": 16780 }, { "grad_norm": 0.16335107386112213, "learning_rate": 4.428599396760956e-05, "loss": 0.0315, "step": 16790 }, { "grad_norm": 0.19703884422779083, "learning_rate": 4.423124287622926e-05, "loss": 0.0322, "step": 16800 }, { "grad_norm": 0.20233698189258575, "learning_rate": 4.417649879443282e-05, "loss": 0.0315, "step": 16810 }, { "grad_norm": 0.2190602719783783, "learning_rate": 4.4121761788739444e-05, "loss": 0.0376, "step": 16820 }, { "grad_norm": 0.18399833142757416, "learning_rate": 4.4067031925659805e-05, "loss": 0.0359, "step": 16830 }, { "grad_norm": 0.1622905433177948, "learning_rate": 4.4012309271695816e-05, "loss": 0.0305, "step": 16840 }, { "grad_norm": 0.194683238863945, "learning_rate": 4.395759389334067e-05, "loss": 0.0372, "step": 16850 }, { "grad_norm": 0.16187703609466553, "learning_rate": 4.3902885857078685e-05, "loss": 0.0448, "step": 16860 }, { "grad_norm": 0.16982007026672363, "learning_rate": 4.384818522938531e-05, "loss": 0.0292, "step": 16870 }, { "grad_norm": 0.21310696005821228, "learning_rate": 4.379349207672695e-05, "loss": 0.0284, "step": 16880 }, { "grad_norm": 0.21731770038604736, "learning_rate": 4.373880646556098e-05, "loss": 0.0352, "step": 16890 }, { "grad_norm": 0.2401011884212494, "learning_rate": 4.3684128462335533e-05, "loss": 0.0269, "step": 16900 }, { "grad_norm": 0.19142329692840576, "learning_rate": 4.362945813348955e-05, "loss": 0.0313, "step": 16910 }, { "grad_norm": 0.17992989718914032, "learning_rate": 4.357479554545262e-05, "loss": 0.0355, "step": 16920 }, { "grad_norm": 0.1972612589597702, "learning_rate": 4.352014076464499e-05, "loss": 0.039, "step": 16930 }, { "grad_norm": 0.175844207406044, "learning_rate": 4.346549385747734e-05, "loss": 0.0279, "step": 16940 }, { "grad_norm": 0.15819713473320007, "learning_rate": 4.34108548903508e-05, "loss": 0.0372, "step": 16950 }, { "grad_norm": 0.2354213446378708, "learning_rate": 4.335622392965689e-05, "loss": 0.0332, "step": 16960 }, { "grad_norm": 0.3464144468307495, "learning_rate": 4.330160104177738e-05, "loss": 0.0342, "step": 16970 }, { "grad_norm": 0.11107303947210312, "learning_rate": 4.32469862930842e-05, "loss": 0.0364, "step": 16980 }, { "grad_norm": 0.17501556873321533, "learning_rate": 4.319237974993946e-05, "loss": 0.0392, "step": 16990 }, { "grad_norm": 0.2296830266714096, "learning_rate": 4.3137781478695236e-05, "loss": 0.0307, "step": 17000 }, { "grad_norm": 0.19337046146392822, "learning_rate": 4.308319154569358e-05, "loss": 0.0449, "step": 17010 }, { "grad_norm": 0.17625300586223602, "learning_rate": 4.302861001726641e-05, "loss": 0.0412, "step": 17020 }, { "grad_norm": 0.15634265542030334, "learning_rate": 4.2974036959735415e-05, "loss": 0.0298, "step": 17030 }, { "grad_norm": 0.2172774374485016, "learning_rate": 4.2919472439412034e-05, "loss": 0.0233, "step": 17040 }, { "grad_norm": 0.21264608204364777, "learning_rate": 4.286491652259729e-05, "loss": 0.0303, "step": 17050 }, { "grad_norm": 0.4149863123893738, "learning_rate": 4.281036927558178e-05, "loss": 0.0361, "step": 17060 }, { "grad_norm": 0.18217696249485016, "learning_rate": 4.275583076464552e-05, "loss": 0.02, "step": 17070 }, { "grad_norm": 0.2094411700963974, "learning_rate": 4.270130105605794e-05, "loss": 0.0254, "step": 17080 }, { "grad_norm": 0.15377216041088104, "learning_rate": 4.264678021607782e-05, "loss": 0.0316, "step": 17090 }, { "grad_norm": 0.1521397829055786, "learning_rate": 4.25922683109531e-05, "loss": 0.0326, "step": 17100 }, { "grad_norm": 0.22790682315826416, "learning_rate": 4.25377654069209e-05, "loss": 0.035, "step": 17110 }, { "grad_norm": 0.19866004586219788, "learning_rate": 4.248327157020737e-05, "loss": 0.0357, "step": 17120 }, { "grad_norm": 0.20211586356163025, "learning_rate": 4.242878686702763e-05, "loss": 0.0316, "step": 17130 }, { "grad_norm": 0.16573238372802734, "learning_rate": 4.23743113635858e-05, "loss": 0.0225, "step": 17140 }, { "grad_norm": 0.20053517818450928, "learning_rate": 4.2319845126074705e-05, "loss": 0.0409, "step": 17150 }, { "grad_norm": 0.5372312068939209, "learning_rate": 4.226538822067598e-05, "loss": 0.0412, "step": 17160 }, { "grad_norm": 0.18660536408424377, "learning_rate": 4.2210940713559896e-05, "loss": 0.045, "step": 17170 }, { "grad_norm": 0.23725625872612, "learning_rate": 4.21565026708853e-05, "loss": 0.0338, "step": 17180 }, { "grad_norm": 0.26244524121284485, "learning_rate": 4.2102074158799535e-05, "loss": 0.0322, "step": 17190 }, { "grad_norm": 0.14086462557315826, "learning_rate": 4.204765524343841e-05, "loss": 0.0347, "step": 17200 }, { "grad_norm": 0.21544791758060455, "learning_rate": 4.1993245990926036e-05, "loss": 0.0315, "step": 17210 }, { "grad_norm": 0.1665424108505249, "learning_rate": 4.193884646737474e-05, "loss": 0.0319, "step": 17220 }, { "grad_norm": 0.2505476772785187, "learning_rate": 4.1884456738885124e-05, "loss": 0.0315, "step": 17230 }, { "grad_norm": 0.22755497694015503, "learning_rate": 4.1830076871545795e-05, "loss": 0.0373, "step": 17240 }, { "grad_norm": 0.21797364950180054, "learning_rate": 4.177570693143346e-05, "loss": 0.0489, "step": 17250 }, { "grad_norm": 0.17442183196544647, "learning_rate": 4.172134698461271e-05, "loss": 0.0248, "step": 17260 }, { "grad_norm": 0.19967007637023926, "learning_rate": 4.166699709713599e-05, "loss": 0.0305, "step": 17270 }, { "grad_norm": 0.1873466670513153, "learning_rate": 4.1612657335043555e-05, "loss": 0.0266, "step": 17280 }, { "grad_norm": 0.22969257831573486, "learning_rate": 4.155832776436331e-05, "loss": 0.0375, "step": 17290 }, { "grad_norm": 0.1707921177148819, "learning_rate": 4.1504008451110845e-05, "loss": 0.0309, "step": 17300 }, { "grad_norm": 0.23129211366176605, "learning_rate": 4.1449699461289225e-05, "loss": 0.0333, "step": 17310 }, { "grad_norm": 0.1480247676372528, "learning_rate": 4.139540086088901e-05, "loss": 0.0253, "step": 17320 }, { "grad_norm": 0.220368430018425, "learning_rate": 4.1341112715888106e-05, "loss": 0.0294, "step": 17330 }, { "grad_norm": 0.14880087971687317, "learning_rate": 4.128683509225172e-05, "loss": 0.0345, "step": 17340 }, { "grad_norm": 0.1543043702840805, "learning_rate": 4.123256805593231e-05, "loss": 0.0288, "step": 17350 }, { "grad_norm": 0.2201840728521347, "learning_rate": 4.117831167286943e-05, "loss": 0.0376, "step": 17360 }, { "grad_norm": 0.2439708560705185, "learning_rate": 4.1124066008989685e-05, "loss": 0.0338, "step": 17370 }, { "grad_norm": 0.09443800151348114, "learning_rate": 4.106983113020669e-05, "loss": 0.0312, "step": 17380 }, { "grad_norm": 0.15740253031253815, "learning_rate": 4.101560710242094e-05, "loss": 0.0295, "step": 17390 }, { "grad_norm": 0.22322557866573334, "learning_rate": 4.096139399151971e-05, "loss": 0.0357, "step": 17400 }, { "grad_norm": 0.20545728504657745, "learning_rate": 4.090719186337709e-05, "loss": 0.0262, "step": 17410 }, { "grad_norm": 0.2304871529340744, "learning_rate": 4.0853000783853746e-05, "loss": 0.0397, "step": 17420 }, { "grad_norm": 0.2542682886123657, "learning_rate": 4.0798820818796956e-05, "loss": 0.0292, "step": 17430 }, { "grad_norm": 0.21827110648155212, "learning_rate": 4.074465203404048e-05, "loss": 0.0361, "step": 17440 }, { "grad_norm": 0.2567199766635895, "learning_rate": 4.0690494495404495e-05, "loss": 0.04, "step": 17450 }, { "grad_norm": 0.1915086954832077, "learning_rate": 4.0636348268695535e-05, "loss": 0.0312, "step": 17460 }, { "grad_norm": 0.1676570624113083, "learning_rate": 4.058221341970637e-05, "loss": 0.0305, "step": 17470 }, { "grad_norm": 0.22271402180194855, "learning_rate": 4.0528090014215945e-05, "loss": 0.0396, "step": 17480 }, { "grad_norm": 0.2074212282896042, "learning_rate": 4.047397811798929e-05, "loss": 0.026, "step": 17490 }, { "grad_norm": 0.19638590514659882, "learning_rate": 4.041987779677745e-05, "loss": 0.03, "step": 17500 }, { "grad_norm": 0.2510003447532654, "learning_rate": 4.036578911631745e-05, "loss": 0.0348, "step": 17510 }, { "grad_norm": 0.17145125567913055, "learning_rate": 4.031171214233211e-05, "loss": 0.0403, "step": 17520 }, { "grad_norm": 0.18289557099342346, "learning_rate": 4.025764694053008e-05, "loss": 0.0358, "step": 17530 }, { "grad_norm": 0.19655637443065643, "learning_rate": 4.020359357660566e-05, "loss": 0.0444, "step": 17540 }, { "grad_norm": 0.20548003911972046, "learning_rate": 4.014955211623875e-05, "loss": 0.0341, "step": 17550 }, { "grad_norm": 0.23746474087238312, "learning_rate": 4.0095522625094895e-05, "loss": 0.0508, "step": 17560 }, { "grad_norm": 0.2323760688304901, "learning_rate": 4.004150516882497e-05, "loss": 0.0471, "step": 17570 }, { "grad_norm": 0.18428923189640045, "learning_rate": 3.99874998130653e-05, "loss": 0.043, "step": 17580 }, { "grad_norm": 0.30359897017478943, "learning_rate": 3.9933506623437455e-05, "loss": 0.0423, "step": 17590 }, { "grad_norm": 0.21229343116283417, "learning_rate": 3.9879525665548276e-05, "loss": 0.0613, "step": 17600 }, { "grad_norm": 0.31361615657806396, "learning_rate": 3.9825557004989704e-05, "loss": 0.0416, "step": 17610 }, { "grad_norm": 0.1819619983434677, "learning_rate": 3.977160070733877e-05, "loss": 0.0415, "step": 17620 }, { "grad_norm": 0.2578895688056946, "learning_rate": 3.971765683815746e-05, "loss": 0.0306, "step": 17630 }, { "grad_norm": 0.21937401592731476, "learning_rate": 3.966372546299265e-05, "loss": 0.0446, "step": 17640 }, { "grad_norm": 0.3193115293979645, "learning_rate": 3.960980664737604e-05, "loss": 0.0364, "step": 17650 }, { "grad_norm": 0.23718389868736267, "learning_rate": 3.955590045682408e-05, "loss": 0.0497, "step": 17660 }, { "grad_norm": 0.2850209176540375, "learning_rate": 3.950200695683788e-05, "loss": 0.0442, "step": 17670 }, { "grad_norm": 0.2615636885166168, "learning_rate": 3.944812621290314e-05, "loss": 0.0447, "step": 17680 }, { "grad_norm": 0.23840856552124023, "learning_rate": 3.939425829049002e-05, "loss": 0.0416, "step": 17690 }, { "grad_norm": 0.18407590687274933, "learning_rate": 3.9340403255053135e-05, "loss": 0.0387, "step": 17700 }, { "grad_norm": 0.21182048320770264, "learning_rate": 3.9286561172031407e-05, "loss": 0.0391, "step": 17710 }, { "grad_norm": 0.17676414549350739, "learning_rate": 3.923273210684809e-05, "loss": 0.0409, "step": 17720 }, { "grad_norm": 0.20453956723213196, "learning_rate": 3.9178916124910555e-05, "loss": 0.0308, "step": 17730 }, { "grad_norm": 0.2249411940574646, "learning_rate": 3.9125113291610274e-05, "loss": 0.0481, "step": 17740 }, { "grad_norm": 0.27679717540740967, "learning_rate": 3.907132367232278e-05, "loss": 0.0564, "step": 17750 }, { "grad_norm": 0.22610853612422943, "learning_rate": 3.901754733240752e-05, "loss": 0.0484, "step": 17760 }, { "grad_norm": 0.24631579220294952, "learning_rate": 3.896378433720785e-05, "loss": 0.0284, "step": 17770 }, { "grad_norm": 0.19653761386871338, "learning_rate": 3.891003475205086e-05, "loss": 0.0512, "step": 17780 }, { "grad_norm": 0.1597522348165512, "learning_rate": 3.885629864224736e-05, "loss": 0.0205, "step": 17790 }, { "grad_norm": 0.19200357794761658, "learning_rate": 3.880257607309178e-05, "loss": 0.0354, "step": 17800 }, { "grad_norm": 0.2562716007232666, "learning_rate": 3.874886710986213e-05, "loss": 0.0495, "step": 17810 }, { "grad_norm": 0.23644733428955078, "learning_rate": 3.8695171817819834e-05, "loss": 0.0369, "step": 17820 }, { "grad_norm": 0.20975366234779358, "learning_rate": 3.8641490262209774e-05, "loss": 0.0369, "step": 17830 }, { "grad_norm": 0.2481648325920105, "learning_rate": 3.8587822508260084e-05, "loss": 0.0342, "step": 17840 }, { "grad_norm": 0.2557702362537384, "learning_rate": 3.853416862118214e-05, "loss": 0.0313, "step": 17850 }, { "grad_norm": 0.18996906280517578, "learning_rate": 3.848052866617049e-05, "loss": 0.0374, "step": 17860 }, { "grad_norm": 0.26094773411750793, "learning_rate": 3.84269027084027e-05, "loss": 0.0383, "step": 17870 }, { "grad_norm": 0.2058882862329483, "learning_rate": 3.83732908130394e-05, "loss": 0.0372, "step": 17880 }, { "grad_norm": 0.2173628807067871, "learning_rate": 3.831969304522409e-05, "loss": 0.0352, "step": 17890 }, { "grad_norm": 0.23674464225769043, "learning_rate": 3.826610947008312e-05, "loss": 0.0383, "step": 17900 }, { "grad_norm": 0.25689318776130676, "learning_rate": 3.821254015272559e-05, "loss": 0.0403, "step": 17910 }, { "grad_norm": 0.17689329385757446, "learning_rate": 3.8158985158243214e-05, "loss": 0.0316, "step": 17920 }, { "grad_norm": 0.16930650174617767, "learning_rate": 3.810544455171044e-05, "loss": 0.0387, "step": 17930 }, { "grad_norm": 0.24286694824695587, "learning_rate": 3.8051918398184114e-05, "loss": 0.0374, "step": 17940 }, { "grad_norm": 0.17894239723682404, "learning_rate": 3.799840676270356e-05, "loss": 0.0441, "step": 17950 }, { "grad_norm": 0.2311716377735138, "learning_rate": 3.794490971029048e-05, "loss": 0.0489, "step": 17960 }, { "grad_norm": 0.23548994958400726, "learning_rate": 3.789142730594881e-05, "loss": 0.0303, "step": 17970 }, { "grad_norm": 0.2196691781282425, "learning_rate": 3.783795961466471e-05, "loss": 0.0503, "step": 17980 }, { "grad_norm": 0.2409685105085373, "learning_rate": 3.778450670140651e-05, "loss": 0.0285, "step": 17990 }, { "grad_norm": 0.16938459873199463, "learning_rate": 3.773106863112451e-05, "loss": 0.0421, "step": 18000 }, { "grad_norm": 0.18907026946544647, "learning_rate": 3.7677645468751e-05, "loss": 0.0506, "step": 18010 }, { "grad_norm": 0.1788053810596466, "learning_rate": 3.762423727920018e-05, "loss": 0.0409, "step": 18020 }, { "grad_norm": 0.18449702858924866, "learning_rate": 3.7570844127367995e-05, "loss": 0.0449, "step": 18030 }, { "grad_norm": 0.18287159502506256, "learning_rate": 3.751746607813221e-05, "loss": 0.044, "step": 18040 }, { "grad_norm": 0.19219540059566498, "learning_rate": 3.746410319635217e-05, "loss": 0.0402, "step": 18050 }, { "grad_norm": 0.18559890985488892, "learning_rate": 3.7410755546868806e-05, "loss": 0.0348, "step": 18060 }, { "grad_norm": 0.20342591404914856, "learning_rate": 3.735742319450454e-05, "loss": 0.0403, "step": 18070 }, { "grad_norm": 0.19558192789554596, "learning_rate": 3.7304106204063184e-05, "loss": 0.0347, "step": 18080 }, { "grad_norm": 0.16370651125907898, "learning_rate": 3.725080464032996e-05, "loss": 0.0383, "step": 18090 }, { "grad_norm": 0.16821178793907166, "learning_rate": 3.719751856807125e-05, "loss": 0.0439, "step": 18100 }, { "grad_norm": 0.2235502302646637, "learning_rate": 3.71442480520347e-05, "loss": 0.0333, "step": 18110 }, { "grad_norm": 0.1981503814458847, "learning_rate": 3.709099315694897e-05, "loss": 0.0375, "step": 18120 }, { "grad_norm": 0.2684493362903595, "learning_rate": 3.703775394752378e-05, "loss": 0.0391, "step": 18130 }, { "grad_norm": 0.1816844344139099, "learning_rate": 3.698453048844983e-05, "loss": 0.03, "step": 18140 }, { "grad_norm": 0.25211620330810547, "learning_rate": 3.693132284439861e-05, "loss": 0.0413, "step": 18150 }, { "grad_norm": 0.16730985045433044, "learning_rate": 3.6878131080022414e-05, "loss": 0.0364, "step": 18160 }, { "grad_norm": 0.20597617328166962, "learning_rate": 3.682495525995429e-05, "loss": 0.0355, "step": 18170 }, { "grad_norm": 0.20978334546089172, "learning_rate": 3.6771795448807846e-05, "loss": 0.0341, "step": 18180 }, { "grad_norm": 0.12720029056072235, "learning_rate": 3.671865171117724e-05, "loss": 0.0385, "step": 18190 }, { "grad_norm": 0.17334280908107758, "learning_rate": 3.666552411163718e-05, "loss": 0.0342, "step": 18200 }, { "grad_norm": 0.19893065094947815, "learning_rate": 3.661241271474269e-05, "loss": 0.0321, "step": 18210 }, { "grad_norm": 0.19480782747268677, "learning_rate": 3.6559317585029116e-05, "loss": 0.0302, "step": 18220 }, { "grad_norm": 0.17633108794689178, "learning_rate": 3.650623878701204e-05, "loss": 0.0329, "step": 18230 }, { "grad_norm": 0.27699360251426697, "learning_rate": 3.645317638518721e-05, "loss": 0.0348, "step": 18240 }, { "grad_norm": 0.2831765413284302, "learning_rate": 3.640013044403046e-05, "loss": 0.0406, "step": 18250 }, { "grad_norm": 0.2677629590034485, "learning_rate": 3.634710102799761e-05, "loss": 0.0337, "step": 18260 }, { "grad_norm": 0.2703036069869995, "learning_rate": 3.6294088201524394e-05, "loss": 0.0393, "step": 18270 }, { "grad_norm": 0.1464790552854538, "learning_rate": 3.62410920290264e-05, "loss": 0.0253, "step": 18280 }, { "grad_norm": 0.25031083822250366, "learning_rate": 3.6188112574898954e-05, "loss": 0.035, "step": 18290 }, { "grad_norm": 0.2124224454164505, "learning_rate": 3.6135149903517115e-05, "loss": 0.0267, "step": 18300 }, { "grad_norm": 0.1874246448278427, "learning_rate": 3.608220407923552e-05, "loss": 0.0371, "step": 18310 }, { "grad_norm": 0.17868053913116455, "learning_rate": 3.602927516638833e-05, "loss": 0.0366, "step": 18320 }, { "grad_norm": 0.15576161444187164, "learning_rate": 3.5976363229289165e-05, "loss": 0.0351, "step": 18330 }, { "grad_norm": 0.2119443565607071, "learning_rate": 3.5923468332231e-05, "loss": 0.0391, "step": 18340 }, { "grad_norm": 0.26227131485939026, "learning_rate": 3.587059053948616e-05, "loss": 0.0351, "step": 18350 }, { "grad_norm": 0.2130778580904007, "learning_rate": 3.5817729915306134e-05, "loss": 0.0346, "step": 18360 }, { "grad_norm": 0.17493723332881927, "learning_rate": 3.5764886523921566e-05, "loss": 0.0291, "step": 18370 }, { "grad_norm": 0.2187938541173935, "learning_rate": 3.571206042954214e-05, "loss": 0.0344, "step": 18380 }, { "grad_norm": 0.13787372410297394, "learning_rate": 3.5659251696356566e-05, "loss": 0.029, "step": 18390 }, { "grad_norm": 0.19648106396198273, "learning_rate": 3.56064603885324e-05, "loss": 0.0433, "step": 18400 }, { "grad_norm": 0.22102369368076324, "learning_rate": 3.5553686570216115e-05, "loss": 0.0442, "step": 18410 }, { "grad_norm": 0.17229019105434418, "learning_rate": 3.550093030553284e-05, "loss": 0.0408, "step": 18420 }, { "grad_norm": 0.26873594522476196, "learning_rate": 3.544819165858642e-05, "loss": 0.0331, "step": 18430 }, { "grad_norm": 0.22624897956848145, "learning_rate": 3.539547069345926e-05, "loss": 0.0286, "step": 18440 }, { "grad_norm": 0.14645111560821533, "learning_rate": 3.5342767474212344e-05, "loss": 0.0429, "step": 18450 }, { "grad_norm": 0.2993749678134918, "learning_rate": 3.529008206488502e-05, "loss": 0.0341, "step": 18460 }, { "grad_norm": 0.3248251974582672, "learning_rate": 3.5237414529495055e-05, "loss": 0.0465, "step": 18470 }, { "grad_norm": 0.11484508216381073, "learning_rate": 3.5184764932038454e-05, "loss": 0.0327, "step": 18480 }, { "grad_norm": 0.1629406213760376, "learning_rate": 3.513213333648945e-05, "loss": 0.0275, "step": 18490 }, { "grad_norm": 0.23732957243919373, "learning_rate": 3.5079519806800374e-05, "loss": 0.0491, "step": 18500 }, { "grad_norm": 0.24712564051151276, "learning_rate": 3.502692440690165e-05, "loss": 0.0385, "step": 18510 }, { "grad_norm": 0.15794216096401215, "learning_rate": 3.497434720070165e-05, "loss": 0.0335, "step": 18520 }, { "grad_norm": 0.21695983409881592, "learning_rate": 3.4921788252086616e-05, "loss": 0.0393, "step": 18530 }, { "grad_norm": 0.10693687200546265, "learning_rate": 3.486924762492065e-05, "loss": 0.0289, "step": 18540 }, { "grad_norm": 0.16470690071582794, "learning_rate": 3.4816725383045534e-05, "loss": 0.0344, "step": 18550 }, { "grad_norm": 0.15006233751773834, "learning_rate": 3.476422159028079e-05, "loss": 0.0547, "step": 18560 }, { "grad_norm": 0.21596089005470276, "learning_rate": 3.471173631042345e-05, "loss": 0.0357, "step": 18570 }, { "grad_norm": 0.17482729256153107, "learning_rate": 3.465926960724808e-05, "loss": 0.0274, "step": 18580 }, { "grad_norm": 0.24598614871501923, "learning_rate": 3.460682154450666e-05, "loss": 0.0356, "step": 18590 }, { "grad_norm": 0.22643035650253296, "learning_rate": 3.4554392185928564e-05, "loss": 0.0523, "step": 18600 }, { "grad_norm": 0.17603425681591034, "learning_rate": 3.450198159522037e-05, "loss": 0.0298, "step": 18610 }, { "grad_norm": 0.15921419858932495, "learning_rate": 3.444958983606592e-05, "loss": 0.053, "step": 18620 }, { "grad_norm": 0.3005395531654358, "learning_rate": 3.439721697212612e-05, "loss": 0.0641, "step": 18630 }, { "grad_norm": 0.16808553040027618, "learning_rate": 3.4344863067038954e-05, "loss": 0.0321, "step": 18640 }, { "grad_norm": 0.24982233345508575, "learning_rate": 3.429252818441935e-05, "loss": 0.0367, "step": 18650 }, { "grad_norm": 0.17254509031772614, "learning_rate": 3.4240212387859094e-05, "loss": 0.0389, "step": 18660 }, { "grad_norm": 0.20754341781139374, "learning_rate": 3.418791574092686e-05, "loss": 0.0391, "step": 18670 }, { "grad_norm": 0.25050297379493713, "learning_rate": 3.413563830716796e-05, "loss": 0.0398, "step": 18680 }, { "grad_norm": 0.1877581626176834, "learning_rate": 3.408338015010445e-05, "loss": 0.0316, "step": 18690 }, { "grad_norm": 0.12823662161827087, "learning_rate": 3.40311413332349e-05, "loss": 0.023, "step": 18700 }, { "grad_norm": 0.1968139111995697, "learning_rate": 3.3978921920034365e-05, "loss": 0.0267, "step": 18710 }, { "grad_norm": 0.12501156330108643, "learning_rate": 3.392672197395441e-05, "loss": 0.0203, "step": 18720 }, { "grad_norm": 0.1986612230539322, "learning_rate": 3.387454155842287e-05, "loss": 0.0687, "step": 18730 }, { "grad_norm": 0.1545063555240631, "learning_rate": 3.382238073684386e-05, "loss": 0.0222, "step": 18740 }, { "grad_norm": 0.18074332177639008, "learning_rate": 3.377023957259771e-05, "loss": 0.0247, "step": 18750 }, { "grad_norm": 0.2664492130279541, "learning_rate": 3.3718118129040835e-05, "loss": 0.0336, "step": 18760 }, { "grad_norm": 0.16315147280693054, "learning_rate": 3.3666016469505724e-05, "loss": 0.0403, "step": 18770 }, { "grad_norm": 0.24761256575584412, "learning_rate": 3.361393465730079e-05, "loss": 0.0451, "step": 18780 }, { "grad_norm": 0.1949421763420105, "learning_rate": 3.3561872755710366e-05, "loss": 0.0304, "step": 18790 }, { "grad_norm": 0.21247145533561707, "learning_rate": 3.350983082799456e-05, "loss": 0.0302, "step": 18800 }, { "grad_norm": 0.22408461570739746, "learning_rate": 3.34578089373892e-05, "loss": 0.033, "step": 18810 }, { "grad_norm": 0.2062862664461136, "learning_rate": 3.340580714710581e-05, "loss": 0.0304, "step": 18820 }, { "grad_norm": 0.20906749367713928, "learning_rate": 3.3353825520331466e-05, "loss": 0.0309, "step": 18830 }, { "grad_norm": 0.14874252676963806, "learning_rate": 3.330186412022876e-05, "loss": 0.0321, "step": 18840 }, { "grad_norm": 0.17219728231430054, "learning_rate": 3.324992300993568e-05, "loss": 0.0263, "step": 18850 }, { "grad_norm": 0.14920870959758759, "learning_rate": 3.319800225256556e-05, "loss": 0.0317, "step": 18860 }, { "grad_norm": 0.20687468349933624, "learning_rate": 3.314610191120702e-05, "loss": 0.0387, "step": 18870 }, { "grad_norm": 0.1286058872938156, "learning_rate": 3.30942220489239e-05, "loss": 0.0277, "step": 18880 }, { "grad_norm": 0.20597349107265472, "learning_rate": 3.3042362728755086e-05, "loss": 0.0352, "step": 18890 }, { "grad_norm": 0.12942515313625336, "learning_rate": 3.299052401371456e-05, "loss": 0.0466, "step": 18900 }, { "grad_norm": 0.1755361109972, "learning_rate": 3.293870596679125e-05, "loss": 0.0243, "step": 18910 }, { "grad_norm": 0.18293610215187073, "learning_rate": 3.288690865094895e-05, "loss": 0.0263, "step": 18920 }, { "grad_norm": 0.19544711709022522, "learning_rate": 3.283513212912632e-05, "loss": 0.0376, "step": 18930 }, { "grad_norm": 0.18266679346561432, "learning_rate": 3.2783376464236684e-05, "loss": 0.024, "step": 18940 }, { "grad_norm": 0.1919373720884323, "learning_rate": 3.273164171916806e-05, "loss": 0.0232, "step": 18950 }, { "grad_norm": 0.16968359053134918, "learning_rate": 3.267992795678306e-05, "loss": 0.0308, "step": 18960 }, { "grad_norm": 0.19283561408519745, "learning_rate": 3.2628235239918744e-05, "loss": 0.0374, "step": 18970 }, { "grad_norm": 0.1951591521501541, "learning_rate": 3.2576563631386695e-05, "loss": 0.0372, "step": 18980 }, { "grad_norm": 0.1839912086725235, "learning_rate": 3.252491319397275e-05, "loss": 0.0256, "step": 18990 }, { "grad_norm": 0.2229239046573639, "learning_rate": 3.247328399043706e-05, "loss": 0.0443, "step": 19000 }, { "grad_norm": 0.20551182329654694, "learning_rate": 3.242167608351399e-05, "loss": 0.041, "step": 19010 }, { "grad_norm": 0.2610750198364258, "learning_rate": 3.2370089535911986e-05, "loss": 0.0314, "step": 19020 }, { "grad_norm": 0.19975534081459045, "learning_rate": 3.23185244103136e-05, "loss": 0.0266, "step": 19030 }, { "grad_norm": 0.17642317712306976, "learning_rate": 3.22669807693753e-05, "loss": 0.0418, "step": 19040 }, { "grad_norm": 0.189057856798172, "learning_rate": 3.2215458675727495e-05, "loss": 0.0348, "step": 19050 }, { "grad_norm": 0.22612616419792175, "learning_rate": 3.216395819197438e-05, "loss": 0.0387, "step": 19060 }, { "grad_norm": 0.22631070017814636, "learning_rate": 3.211247938069387e-05, "loss": 0.0439, "step": 19070 }, { "grad_norm": 0.23374705016613007, "learning_rate": 3.206102230443759e-05, "loss": 0.0278, "step": 19080 }, { "grad_norm": 0.20651166141033173, "learning_rate": 3.2009587025730764e-05, "loss": 0.039, "step": 19090 }, { "grad_norm": 0.17833474278450012, "learning_rate": 3.195817360707207e-05, "loss": 0.0445, "step": 19100 }, { "grad_norm": 0.25085994601249695, "learning_rate": 3.19067821109337e-05, "loss": 0.0311, "step": 19110 }, { "grad_norm": 0.14792053401470184, "learning_rate": 3.185541259976114e-05, "loss": 0.022, "step": 19120 }, { "grad_norm": 0.16901874542236328, "learning_rate": 3.180406513597316e-05, "loss": 0.045, "step": 19130 }, { "grad_norm": 0.18258601427078247, "learning_rate": 3.1752739781961835e-05, "loss": 0.0255, "step": 19140 }, { "grad_norm": 0.19481147825717926, "learning_rate": 3.170143660009228e-05, "loss": 0.0182, "step": 19150 }, { "grad_norm": 0.22523626685142517, "learning_rate": 3.16501556527027e-05, "loss": 0.0347, "step": 19160 }, { "grad_norm": 0.19510923326015472, "learning_rate": 3.1598897002104265e-05, "loss": 0.0266, "step": 19170 }, { "grad_norm": 0.17791448533535004, "learning_rate": 3.154766071058108e-05, "loss": 0.029, "step": 19180 }, { "grad_norm": 0.1949159801006317, "learning_rate": 3.149644684039008e-05, "loss": 0.0473, "step": 19190 }, { "grad_norm": 0.20815108716487885, "learning_rate": 3.144525545376095e-05, "loss": 0.027, "step": 19200 }, { "grad_norm": 0.23945747315883636, "learning_rate": 3.139408661289603e-05, "loss": 0.0297, "step": 19210 }, { "grad_norm": 0.2272346317768097, "learning_rate": 3.134294037997032e-05, "loss": 0.0277, "step": 19220 }, { "grad_norm": 0.24184994399547577, "learning_rate": 3.129181681713127e-05, "loss": 0.0324, "step": 19230 }, { "grad_norm": 0.20474806427955627, "learning_rate": 3.1240715986498855e-05, "loss": 0.0303, "step": 19240 }, { "grad_norm": 0.1872473955154419, "learning_rate": 3.1189637950165394e-05, "loss": 0.0358, "step": 19250 }, { "grad_norm": 0.2087792456150055, "learning_rate": 3.1138582770195544e-05, "loss": 0.0267, "step": 19260 }, { "grad_norm": 0.14861610531806946, "learning_rate": 3.108755050862615e-05, "loss": 0.0231, "step": 19270 }, { "grad_norm": 0.22141896188259125, "learning_rate": 3.1036541227466204e-05, "loss": 0.026, "step": 19280 }, { "grad_norm": 0.2618032693862915, "learning_rate": 3.0985554988696786e-05, "loss": 0.0453, "step": 19290 }, { "grad_norm": 0.17633269727230072, "learning_rate": 3.093459185427102e-05, "loss": 0.0224, "step": 19300 }, { "grad_norm": 0.29034292697906494, "learning_rate": 3.0883651886113905e-05, "loss": 0.0304, "step": 19310 }, { "grad_norm": 0.1869911551475525, "learning_rate": 3.0832735146122294e-05, "loss": 0.0479, "step": 19320 }, { "grad_norm": 0.20567278563976288, "learning_rate": 3.078184169616485e-05, "loss": 0.0477, "step": 19330 }, { "grad_norm": 0.23613309860229492, "learning_rate": 3.073097159808187e-05, "loss": 0.0499, "step": 19340 }, { "grad_norm": 0.17235489189624786, "learning_rate": 3.068012491368537e-05, "loss": 0.041, "step": 19350 }, { "grad_norm": 0.17517909407615662, "learning_rate": 3.062930170475885e-05, "loss": 0.0208, "step": 19360 }, { "grad_norm": 0.2294636368751526, "learning_rate": 3.057850203305729e-05, "loss": 0.0288, "step": 19370 }, { "grad_norm": 0.23882533609867096, "learning_rate": 3.052772596030708e-05, "loss": 0.0366, "step": 19380 }, { "grad_norm": 0.10134632885456085, "learning_rate": 3.0476973548205943e-05, "loss": 0.0296, "step": 19390 }, { "grad_norm": 0.1951560378074646, "learning_rate": 3.0426244858422847e-05, "loss": 0.0301, "step": 19400 }, { "grad_norm": 0.11982168257236481, "learning_rate": 3.0375539952597943e-05, "loss": 0.0184, "step": 19410 }, { "grad_norm": 0.2145329862833023, "learning_rate": 3.0324858892342468e-05, "loss": 0.0215, "step": 19420 }, { "grad_norm": 0.5094833970069885, "learning_rate": 3.0274201739238672e-05, "loss": 0.0401, "step": 19430 }, { "grad_norm": 0.22126208245754242, "learning_rate": 3.0223568554839786e-05, "loss": 0.0242, "step": 19440 }, { "grad_norm": 0.17152267694473267, "learning_rate": 3.0172959400669886e-05, "loss": 0.0223, "step": 19450 }, { "grad_norm": 0.16729439795017242, "learning_rate": 3.0122374338223902e-05, "loss": 0.0206, "step": 19460 }, { "grad_norm": 0.1498456448316574, "learning_rate": 3.0071813428967427e-05, "loss": 0.0329, "step": 19470 }, { "grad_norm": 0.2324570268392563, "learning_rate": 3.0021276734336746e-05, "loss": 0.0259, "step": 19480 }, { "grad_norm": 0.16636350750923157, "learning_rate": 2.997076431573871e-05, "loss": 0.036, "step": 19490 }, { "grad_norm": 0.25444895029067993, "learning_rate": 2.9920276234550636e-05, "loss": 0.0244, "step": 19500 }, { "grad_norm": 0.1194058507680893, "learning_rate": 2.986981255212035e-05, "loss": 0.0407, "step": 19510 }, { "grad_norm": 0.19864346086978912, "learning_rate": 2.9819373329765977e-05, "loss": 0.0267, "step": 19520 }, { "grad_norm": 0.16163180768489838, "learning_rate": 2.9768958628775902e-05, "loss": 0.0178, "step": 19530 }, { "grad_norm": 0.19096767902374268, "learning_rate": 2.9718568510408762e-05, "loss": 0.0218, "step": 19540 }, { "grad_norm": 0.18430452048778534, "learning_rate": 2.9668203035893272e-05, "loss": 0.0268, "step": 19550 }, { "grad_norm": 0.1986236423254013, "learning_rate": 2.9617862266428287e-05, "loss": 0.0361, "step": 19560 }, { "grad_norm": 0.16380731761455536, "learning_rate": 2.9567546263182556e-05, "loss": 0.0258, "step": 19570 }, { "grad_norm": 0.19214165210723877, "learning_rate": 2.951725508729476e-05, "loss": 0.033, "step": 19580 }, { "grad_norm": 0.17830154299736023, "learning_rate": 2.946698879987344e-05, "loss": 0.0318, "step": 19590 }, { "grad_norm": 0.2021605372428894, "learning_rate": 2.9416747461996853e-05, "loss": 0.0325, "step": 19600 }, { "grad_norm": 0.1445900946855545, "learning_rate": 2.9366531134712973e-05, "loss": 0.0246, "step": 19610 }, { "grad_norm": 0.21563471853733063, "learning_rate": 2.9316339879039367e-05, "loss": 0.028, "step": 19620 }, { "grad_norm": 0.12324675917625427, "learning_rate": 2.9266173755963167e-05, "loss": 0.0267, "step": 19630 }, { "grad_norm": 0.18537619709968567, "learning_rate": 2.9216032826440926e-05, "loss": 0.0241, "step": 19640 }, { "grad_norm": 0.1994611769914627, "learning_rate": 2.9165917151398592e-05, "loss": 0.0263, "step": 19650 }, { "grad_norm": 0.1934840977191925, "learning_rate": 2.9115826791731428e-05, "loss": 0.024, "step": 19660 }, { "grad_norm": 0.15007077157497406, "learning_rate": 2.906576180830398e-05, "loss": 0.0205, "step": 19670 }, { "grad_norm": 0.22179685533046722, "learning_rate": 2.9015722261949917e-05, "loss": 0.0252, "step": 19680 }, { "grad_norm": 0.17887121438980103, "learning_rate": 2.8965708213471986e-05, "loss": 0.0396, "step": 19690 }, { "grad_norm": 0.25737902522087097, "learning_rate": 2.8915719723641975e-05, "loss": 0.0304, "step": 19700 }, { "grad_norm": 0.11517304927110672, "learning_rate": 2.8865756853200604e-05, "loss": 0.0321, "step": 19710 }, { "grad_norm": 0.17800123989582062, "learning_rate": 2.8815819662857503e-05, "loss": 0.0332, "step": 19720 }, { "grad_norm": 0.2051839530467987, "learning_rate": 2.876590821329105e-05, "loss": 0.0274, "step": 19730 }, { "grad_norm": 0.173445463180542, "learning_rate": 2.871602256514836e-05, "loss": 0.0267, "step": 19740 }, { "grad_norm": 0.15847386419773102, "learning_rate": 2.8666162779045203e-05, "loss": 0.0187, "step": 19750 }, { "grad_norm": 0.1442551612854004, "learning_rate": 2.8616328915565904e-05, "loss": 0.0436, "step": 19760 }, { "grad_norm": 0.15275998413562775, "learning_rate": 2.856652103526334e-05, "loss": 0.0178, "step": 19770 }, { "grad_norm": 0.25260835886001587, "learning_rate": 2.8516739198658755e-05, "loss": 0.0373, "step": 19780 }, { "grad_norm": 0.19478574395179749, "learning_rate": 2.846698346624177e-05, "loss": 0.0367, "step": 19790 }, { "grad_norm": 0.17324167490005493, "learning_rate": 2.841725389847032e-05, "loss": 0.0289, "step": 19800 }, { "grad_norm": 0.1577959954738617, "learning_rate": 2.8367550555770506e-05, "loss": 0.0237, "step": 19810 }, { "grad_norm": 0.14876766502857208, "learning_rate": 2.831787349853655e-05, "loss": 0.0331, "step": 19820 }, { "grad_norm": 0.15258604288101196, "learning_rate": 2.8268222787130806e-05, "loss": 0.0178, "step": 19830 }, { "grad_norm": 0.17461040616035461, "learning_rate": 2.821859848188355e-05, "loss": 0.0228, "step": 19840 }, { "grad_norm": 0.2127096801996231, "learning_rate": 2.816900064309299e-05, "loss": 0.04, "step": 19850 }, { "grad_norm": 0.2279936522245407, "learning_rate": 2.811942933102517e-05, "loss": 0.0407, "step": 19860 }, { "grad_norm": 0.19612039625644684, "learning_rate": 2.806988460591391e-05, "loss": 0.0237, "step": 19870 }, { "grad_norm": 0.21209228038787842, "learning_rate": 2.802036652796074e-05, "loss": 0.0315, "step": 19880 }, { "grad_norm": 0.12661439180374146, "learning_rate": 2.797087515733478e-05, "loss": 0.0328, "step": 19890 }, { "grad_norm": 0.2136591374874115, "learning_rate": 2.7921410554172723e-05, "loss": 0.0538, "step": 19900 }, { "grad_norm": 0.1741955578327179, "learning_rate": 2.787197277857871e-05, "loss": 0.0237, "step": 19910 }, { "grad_norm": 0.25654467940330505, "learning_rate": 2.782256189062429e-05, "loss": 0.029, "step": 19920 }, { "grad_norm": 0.18921031057834625, "learning_rate": 2.777317795034839e-05, "loss": 0.0386, "step": 19930 }, { "grad_norm": 0.1802884042263031, "learning_rate": 2.7723821017757112e-05, "loss": 0.0292, "step": 19940 }, { "grad_norm": 0.18317203223705292, "learning_rate": 2.7674491152823822e-05, "loss": 0.023, "step": 19950 }, { "grad_norm": 0.20081117749214172, "learning_rate": 2.7625188415488944e-05, "loss": 0.0302, "step": 19960 }, { "grad_norm": 0.17144300043582916, "learning_rate": 2.7575912865659924e-05, "loss": 0.0329, "step": 19970 }, { "grad_norm": 0.25255900621414185, "learning_rate": 2.7526664563211245e-05, "loss": 0.0361, "step": 19980 }, { "grad_norm": 0.20729579031467438, "learning_rate": 2.7477443567984224e-05, "loss": 0.0351, "step": 19990 }, { "grad_norm": 0.23493489623069763, "learning_rate": 2.7428249939787e-05, "loss": 0.0377, "step": 20000 }, { "grad_norm": 0.1688770055770874, "learning_rate": 2.7379083738394484e-05, "loss": 0.0279, "step": 20010 }, { "grad_norm": 0.15197257697582245, "learning_rate": 2.732994502354823e-05, "loss": 0.0319, "step": 20020 }, { "grad_norm": 0.20481829345226288, "learning_rate": 2.72808338549564e-05, "loss": 0.0267, "step": 20030 }, { "grad_norm": 0.21561433374881744, "learning_rate": 2.723175029229374e-05, "loss": 0.0429, "step": 20040 }, { "grad_norm": 0.2024550437927246, "learning_rate": 2.718269439520138e-05, "loss": 0.0364, "step": 20050 }, { "grad_norm": 0.19204632937908173, "learning_rate": 2.7133666223286858e-05, "loss": 0.0278, "step": 20060 }, { "grad_norm": 0.3270433843135834, "learning_rate": 2.7084665836124006e-05, "loss": 0.0486, "step": 20070 }, { "grad_norm": 0.16668082773685455, "learning_rate": 2.703569329325296e-05, "loss": 0.0352, "step": 20080 }, { "grad_norm": 0.1965586245059967, "learning_rate": 2.698674865417994e-05, "loss": 0.0354, "step": 20090 }, { "grad_norm": 0.13878567516803741, "learning_rate": 2.6937831978377325e-05, "loss": 0.0286, "step": 20100 }, { "grad_norm": 0.21392883360385895, "learning_rate": 2.6888943325283485e-05, "loss": 0.0262, "step": 20110 }, { "grad_norm": 0.1801300197839737, "learning_rate": 2.6840082754302733e-05, "loss": 0.0265, "step": 20120 }, { "grad_norm": 0.10979640483856201, "learning_rate": 2.679125032480525e-05, "loss": 0.0334, "step": 20130 }, { "grad_norm": 0.1553611308336258, "learning_rate": 2.674244609612708e-05, "loss": 0.0202, "step": 20140 }, { "grad_norm": 0.14357419312000275, "learning_rate": 2.6693670127569958e-05, "loss": 0.0224, "step": 20150 }, { "grad_norm": 0.13759490847587585, "learning_rate": 2.6644922478401268e-05, "loss": 0.0232, "step": 20160 }, { "grad_norm": 0.2187042087316513, "learning_rate": 2.6596203207854004e-05, "loss": 0.0331, "step": 20170 }, { "grad_norm": 0.1625836342573166, "learning_rate": 2.654751237512666e-05, "loss": 0.0248, "step": 20180 }, { "grad_norm": 0.1496584266424179, "learning_rate": 2.649885003938323e-05, "loss": 0.0242, "step": 20190 }, { "grad_norm": 0.16999220848083496, "learning_rate": 2.6450216259753e-05, "loss": 0.0188, "step": 20200 }, { "grad_norm": 0.15133078396320343, "learning_rate": 2.640161109533063e-05, "loss": 0.027, "step": 20210 }, { "grad_norm": 0.1911870688199997, "learning_rate": 2.6353034605175936e-05, "loss": 0.0249, "step": 20220 }, { "grad_norm": 0.26678910851478577, "learning_rate": 2.630448684831398e-05, "loss": 0.0409, "step": 20230 }, { "grad_norm": 0.1607958972454071, "learning_rate": 2.625596788373482e-05, "loss": 0.0243, "step": 20240 }, { "grad_norm": 0.18567954003810883, "learning_rate": 2.620747777039363e-05, "loss": 0.0216, "step": 20250 }, { "grad_norm": 0.08817169070243835, "learning_rate": 2.6159016567210424e-05, "loss": 0.0201, "step": 20260 }, { "grad_norm": 0.17050345242023468, "learning_rate": 2.6110584333070153e-05, "loss": 0.0226, "step": 20270 }, { "grad_norm": 0.1576932817697525, "learning_rate": 2.6062181126822537e-05, "loss": 0.0186, "step": 20280 }, { "grad_norm": 0.14608705043792725, "learning_rate": 2.601380700728203e-05, "loss": 0.0378, "step": 20290 }, { "grad_norm": 0.18365882337093353, "learning_rate": 2.596546203322777e-05, "loss": 0.0225, "step": 20300 }, { "grad_norm": 0.16317495703697205, "learning_rate": 2.5917146263403457e-05, "loss": 0.0369, "step": 20310 }, { "grad_norm": 0.10982577502727509, "learning_rate": 2.5868859756517296e-05, "loss": 0.0222, "step": 20320 }, { "grad_norm": 0.1903417557477951, "learning_rate": 2.5820602571241947e-05, "loss": 0.0276, "step": 20330 }, { "grad_norm": 0.1603066623210907, "learning_rate": 2.577237476621442e-05, "loss": 0.0268, "step": 20340 }, { "grad_norm": 0.22208483517169952, "learning_rate": 2.572417640003609e-05, "loss": 0.0369, "step": 20350 }, { "grad_norm": 0.1415211707353592, "learning_rate": 2.5676007531272473e-05, "loss": 0.0322, "step": 20360 }, { "grad_norm": 0.18838229775428772, "learning_rate": 2.5627868218453332e-05, "loss": 0.0318, "step": 20370 }, { "grad_norm": 0.13203783333301544, "learning_rate": 2.5579758520072443e-05, "loss": 0.0363, "step": 20380 }, { "grad_norm": 0.22818239033222198, "learning_rate": 2.5531678494587615e-05, "loss": 0.0255, "step": 20390 }, { "grad_norm": 0.11725274473428726, "learning_rate": 2.5483628200420644e-05, "loss": 0.0224, "step": 20400 }, { "grad_norm": 0.2075803130865097, "learning_rate": 2.5435607695957154e-05, "loss": 0.0334, "step": 20410 }, { "grad_norm": 0.14864018559455872, "learning_rate": 2.5387617039546586e-05, "loss": 0.0179, "step": 20420 }, { "grad_norm": 0.12741626799106598, "learning_rate": 2.5339656289502106e-05, "loss": 0.0233, "step": 20430 }, { "grad_norm": 0.2421254813671112, "learning_rate": 2.529172550410056e-05, "loss": 0.0355, "step": 20440 }, { "grad_norm": 0.18888892233371735, "learning_rate": 2.524382474158234e-05, "loss": 0.0278, "step": 20450 }, { "grad_norm": 0.1441790759563446, "learning_rate": 2.5195954060151433e-05, "loss": 0.04, "step": 20460 }, { "grad_norm": 0.1561560183763504, "learning_rate": 2.5148113517975212e-05, "loss": 0.0315, "step": 20470 }, { "grad_norm": 0.18558965623378754, "learning_rate": 2.5100303173184447e-05, "loss": 0.0204, "step": 20480 }, { "grad_norm": 0.20697222650051117, "learning_rate": 2.5052523083873196e-05, "loss": 0.02, "step": 20490 }, { "grad_norm": 0.15609242022037506, "learning_rate": 2.500477330809881e-05, "loss": 0.0253, "step": 20500 }, { "grad_norm": 0.16097944974899292, "learning_rate": 2.4957053903881733e-05, "loss": 0.0323, "step": 20510 }, { "grad_norm": 0.20265859365463257, "learning_rate": 2.4909364929205576e-05, "loss": 0.0244, "step": 20520 }, { "grad_norm": 0.21971498429775238, "learning_rate": 2.4861706442016923e-05, "loss": 0.029, "step": 20530 }, { "grad_norm": 0.20994700491428375, "learning_rate": 2.481407850022533e-05, "loss": 0.018, "step": 20540 }, { "grad_norm": 0.20007985830307007, "learning_rate": 2.4766481161703213e-05, "loss": 0.0301, "step": 20550 }, { "grad_norm": 0.19712957739830017, "learning_rate": 2.4718914484285876e-05, "loss": 0.0247, "step": 20560 }, { "grad_norm": 0.2247798889875412, "learning_rate": 2.467137852577129e-05, "loss": 0.0222, "step": 20570 }, { "grad_norm": 0.16199837625026703, "learning_rate": 2.4623873343920123e-05, "loss": 0.0212, "step": 20580 }, { "grad_norm": 0.19048060476779938, "learning_rate": 2.4576398996455656e-05, "loss": 0.0322, "step": 20590 }, { "grad_norm": 0.16861756145954132, "learning_rate": 2.452895554106368e-05, "loss": 0.0234, "step": 20600 }, { "grad_norm": 0.19529052078723907, "learning_rate": 2.4481543035392506e-05, "loss": 0.0323, "step": 20610 }, { "grad_norm": 0.10229980945587158, "learning_rate": 2.4434161537052774e-05, "loss": 0.027, "step": 20620 }, { "grad_norm": 0.1871950626373291, "learning_rate": 2.4386811103617474e-05, "loss": 0.0318, "step": 20630 }, { "grad_norm": 0.16084523499011993, "learning_rate": 2.4339491792621833e-05, "loss": 0.0255, "step": 20640 }, { "grad_norm": 0.15760231018066406, "learning_rate": 2.4292203661563312e-05, "loss": 0.0208, "step": 20650 }, { "grad_norm": 0.12880602478981018, "learning_rate": 2.424494676790141e-05, "loss": 0.0228, "step": 20660 }, { "grad_norm": 0.23049892485141754, "learning_rate": 2.4197721169057745e-05, "loss": 0.0184, "step": 20670 }, { "grad_norm": 0.12840205430984497, "learning_rate": 2.4150526922415857e-05, "loss": 0.0284, "step": 20680 }, { "grad_norm": 0.20380505919456482, "learning_rate": 2.41033640853212e-05, "loss": 0.0201, "step": 20690 }, { "grad_norm": 0.09281755983829498, "learning_rate": 2.405623271508108e-05, "loss": 0.0316, "step": 20700 }, { "grad_norm": 0.22415830194950104, "learning_rate": 2.4009132868964522e-05, "loss": 0.0208, "step": 20710 }, { "grad_norm": 0.21203817427158356, "learning_rate": 2.3962064604202327e-05, "loss": 0.0215, "step": 20720 }, { "grad_norm": 0.20170597732067108, "learning_rate": 2.3915027977986858e-05, "loss": 0.0345, "step": 20730 }, { "grad_norm": 0.1444002240896225, "learning_rate": 2.386802304747205e-05, "loss": 0.0215, "step": 20740 }, { "grad_norm": 0.2299066036939621, "learning_rate": 2.382104986977332e-05, "loss": 0.0329, "step": 20750 }, { "grad_norm": 0.1904812455177307, "learning_rate": 2.377410850196749e-05, "loss": 0.0287, "step": 20760 }, { "grad_norm": 0.18683667480945587, "learning_rate": 2.3727199001092797e-05, "loss": 0.0378, "step": 20770 }, { "grad_norm": 0.21142247319221497, "learning_rate": 2.3680321424148677e-05, "loss": 0.0426, "step": 20780 }, { "grad_norm": 0.11223860830068588, "learning_rate": 2.3633475828095793e-05, "loss": 0.0216, "step": 20790 }, { "grad_norm": 0.12369458377361298, "learning_rate": 2.358666226985599e-05, "loss": 0.0382, "step": 20800 }, { "grad_norm": 0.19788211584091187, "learning_rate": 2.353988080631213e-05, "loss": 0.0243, "step": 20810 }, { "grad_norm": 0.16268885135650635, "learning_rate": 2.3493131494308142e-05, "loss": 0.038, "step": 20820 }, { "grad_norm": 0.2067040503025055, "learning_rate": 2.344641439064881e-05, "loss": 0.0244, "step": 20830 }, { "grad_norm": 0.1329353302717209, "learning_rate": 2.339972955209984e-05, "loss": 0.0341, "step": 20840 }, { "grad_norm": 0.1376141756772995, "learning_rate": 2.335307703538771e-05, "loss": 0.0316, "step": 20850 }, { "grad_norm": 0.14588366448879242, "learning_rate": 2.330645689719962e-05, "loss": 0.0451, "step": 20860 }, { "grad_norm": 0.2211124747991562, "learning_rate": 2.3259869194183414e-05, "loss": 0.0263, "step": 20870 }, { "grad_norm": 0.12521974742412567, "learning_rate": 2.321331398294759e-05, "loss": 0.0308, "step": 20880 }, { "grad_norm": 0.22582505643367767, "learning_rate": 2.3166791320061094e-05, "loss": 0.0328, "step": 20890 }, { "grad_norm": 0.21417847275733948, "learning_rate": 2.312030126205335e-05, "loss": 0.0301, "step": 20900 }, { "grad_norm": 0.1755513697862625, "learning_rate": 2.3073843865414162e-05, "loss": 0.0332, "step": 20910 }, { "grad_norm": 0.18527251482009888, "learning_rate": 2.302741918659363e-05, "loss": 0.0193, "step": 20920 }, { "grad_norm": 0.17968769371509552, "learning_rate": 2.2981027282002154e-05, "loss": 0.0248, "step": 20930 }, { "grad_norm": 0.0942239910364151, "learning_rate": 2.2934668208010236e-05, "loss": 0.0269, "step": 20940 }, { "grad_norm": 0.21117550134658813, "learning_rate": 2.2888342020948555e-05, "loss": 0.0263, "step": 20950 }, { "grad_norm": 0.16504725813865662, "learning_rate": 2.2842048777107783e-05, "loss": 0.0323, "step": 20960 }, { "grad_norm": 0.15530329942703247, "learning_rate": 2.2795788532738553e-05, "loss": 0.0269, "step": 20970 }, { "grad_norm": 0.15260106325149536, "learning_rate": 2.2749561344051468e-05, "loss": 0.0209, "step": 20980 }, { "grad_norm": 0.17478150129318237, "learning_rate": 2.2703367267216896e-05, "loss": 0.0266, "step": 20990 }, { "grad_norm": 0.14663922786712646, "learning_rate": 2.2657206358364997e-05, "loss": 0.027, "step": 21000 }, { "grad_norm": 0.16895005106925964, "learning_rate": 2.261107867358563e-05, "loss": 0.0412, "step": 21010 }, { "grad_norm": 0.1495346575975418, "learning_rate": 2.2564984268928263e-05, "loss": 0.0225, "step": 21020 }, { "grad_norm": 0.18198394775390625, "learning_rate": 2.251892320040198e-05, "loss": 0.0308, "step": 21030 }, { "grad_norm": 0.16610011458396912, "learning_rate": 2.2472895523975312e-05, "loss": 0.0275, "step": 21040 }, { "grad_norm": 0.22386524081230164, "learning_rate": 2.2426901295576212e-05, "loss": 0.0362, "step": 21050 }, { "grad_norm": 0.18002431094646454, "learning_rate": 2.2380940571092013e-05, "loss": 0.0378, "step": 21060 }, { "grad_norm": 0.12864571809768677, "learning_rate": 2.2335013406369302e-05, "loss": 0.0274, "step": 21070 }, { "grad_norm": 0.17372266948223114, "learning_rate": 2.228911985721397e-05, "loss": 0.0218, "step": 21080 }, { "grad_norm": 0.14629247784614563, "learning_rate": 2.224325997939095e-05, "loss": 0.0285, "step": 21090 }, { "grad_norm": 0.23127131164073944, "learning_rate": 2.219743382862437e-05, "loss": 0.0443, "step": 21100 }, { "grad_norm": 0.24913686513900757, "learning_rate": 2.2151641460597294e-05, "loss": 0.0265, "step": 21110 }, { "grad_norm": 0.1516319364309311, "learning_rate": 2.210588293095177e-05, "loss": 0.0218, "step": 21120 }, { "grad_norm": 0.2286088913679123, "learning_rate": 2.2060158295288714e-05, "loss": 0.0343, "step": 21130 }, { "grad_norm": 0.1560145616531372, "learning_rate": 2.2014467609167905e-05, "loss": 0.0252, "step": 21140 }, { "grad_norm": 0.16035591065883636, "learning_rate": 2.1968810928107806e-05, "loss": 0.027, "step": 21150 }, { "grad_norm": 0.15179751813411713, "learning_rate": 2.1923188307585606e-05, "loss": 0.0238, "step": 21160 }, { "grad_norm": 0.1590225100517273, "learning_rate": 2.187759980303708e-05, "loss": 0.0352, "step": 21170 }, { "grad_norm": 0.2057982087135315, "learning_rate": 2.1832045469856545e-05, "loss": 0.0251, "step": 21180 }, { "grad_norm": 0.25758516788482666, "learning_rate": 2.1786525363396844e-05, "loss": 0.0382, "step": 21190 }, { "grad_norm": 0.196612149477005, "learning_rate": 2.1741039538969183e-05, "loss": 0.0284, "step": 21200 }, { "grad_norm": 0.16738943755626678, "learning_rate": 2.169558805184313e-05, "loss": 0.0249, "step": 21210 }, { "grad_norm": 0.1902991384267807, "learning_rate": 2.165017095724651e-05, "loss": 0.0273, "step": 21220 }, { "grad_norm": 0.17849840223789215, "learning_rate": 2.1604788310365405e-05, "loss": 0.0301, "step": 21230 }, { "grad_norm": 0.1640634685754776, "learning_rate": 2.155944016634398e-05, "loss": 0.0248, "step": 21240 }, { "grad_norm": 0.12986505031585693, "learning_rate": 2.1514126580284538e-05, "loss": 0.0244, "step": 21250 }, { "grad_norm": 0.20847229659557343, "learning_rate": 2.1468847607247344e-05, "loss": 0.0315, "step": 21260 }, { "grad_norm": 0.1402651071548462, "learning_rate": 2.1423603302250623e-05, "loss": 0.0314, "step": 21270 }, { "grad_norm": 0.12852810323238373, "learning_rate": 2.1378393720270468e-05, "loss": 0.03, "step": 21280 }, { "grad_norm": 0.18841251730918884, "learning_rate": 2.133321891624076e-05, "loss": 0.0304, "step": 21290 }, { "grad_norm": 0.15956151485443115, "learning_rate": 2.1288078945053193e-05, "loss": 0.0236, "step": 21300 }, { "grad_norm": 0.20243099331855774, "learning_rate": 2.1242973861557064e-05, "loss": 0.0288, "step": 21310 }, { "grad_norm": 0.0882873609662056, "learning_rate": 2.1197903720559304e-05, "loss": 0.0242, "step": 21320 }, { "grad_norm": 0.17120522260665894, "learning_rate": 2.115286857682438e-05, "loss": 0.0394, "step": 21330 }, { "grad_norm": 0.12689794600009918, "learning_rate": 2.1107868485074228e-05, "loss": 0.0265, "step": 21340 }, { "grad_norm": 0.14227309823036194, "learning_rate": 2.1062903499988234e-05, "loss": 0.0292, "step": 21350 }, { "grad_norm": 0.22940264642238617, "learning_rate": 2.101797367620308e-05, "loss": 0.0293, "step": 21360 }, { "grad_norm": 0.1723417341709137, "learning_rate": 2.0973079068312713e-05, "loss": 0.0359, "step": 21370 }, { "grad_norm": 0.2031976282596588, "learning_rate": 2.0928219730868358e-05, "loss": 0.0228, "step": 21380 }, { "grad_norm": 0.14974458515644073, "learning_rate": 2.0883395718378302e-05, "loss": 0.0274, "step": 21390 }, { "grad_norm": 0.17786921560764313, "learning_rate": 2.083860708530798e-05, "loss": 0.0376, "step": 21400 }, { "grad_norm": 0.13911838829517365, "learning_rate": 2.0793853886079796e-05, "loss": 0.0438, "step": 21410 }, { "grad_norm": 0.16717226803302765, "learning_rate": 2.074913617507309e-05, "loss": 0.0338, "step": 21420 }, { "grad_norm": 0.15858346223831177, "learning_rate": 2.070445400662411e-05, "loss": 0.0318, "step": 21430 }, { "grad_norm": 0.16048508882522583, "learning_rate": 2.0659807435025907e-05, "loss": 0.0275, "step": 21440 }, { "grad_norm": 0.17269977927207947, "learning_rate": 2.061519651452825e-05, "loss": 0.0449, "step": 21450 }, { "grad_norm": 0.09934234619140625, "learning_rate": 2.0570621299337656e-05, "loss": 0.0265, "step": 21460 }, { "grad_norm": 0.27813079953193665, "learning_rate": 2.0526081843617183e-05, "loss": 0.0512, "step": 21470 }, { "grad_norm": 0.1343764364719391, "learning_rate": 2.0481578201486485e-05, "loss": 0.0248, "step": 21480 }, { "grad_norm": 0.20363295078277588, "learning_rate": 2.0437110427021677e-05, "loss": 0.0319, "step": 21490 }, { "grad_norm": 0.145937979221344, "learning_rate": 2.0392678574255282e-05, "loss": 0.034, "step": 21500 }, { "grad_norm": 0.16833585500717163, "learning_rate": 2.0348282697176217e-05, "loss": 0.036, "step": 21510 }, { "grad_norm": 0.16174155473709106, "learning_rate": 2.030392284972964e-05, "loss": 0.0332, "step": 21520 }, { "grad_norm": 0.14360186457633972, "learning_rate": 2.0259599085816973e-05, "loss": 0.0267, "step": 21530 }, { "grad_norm": 0.22250375151634216, "learning_rate": 2.0215311459295755e-05, "loss": 0.03, "step": 21540 }, { "grad_norm": 0.22078818082809448, "learning_rate": 2.0171060023979606e-05, "loss": 0.0337, "step": 21550 }, { "grad_norm": 0.13595589995384216, "learning_rate": 2.012684483363823e-05, "loss": 0.0337, "step": 21560 }, { "grad_norm": 0.1643228679895401, "learning_rate": 2.0082665941997237e-05, "loss": 0.0322, "step": 21570 }, { "grad_norm": 0.12083564698696136, "learning_rate": 2.0038523402738146e-05, "loss": 0.0226, "step": 21580 }, { "grad_norm": 0.2053171843290329, "learning_rate": 1.99944172694983e-05, "loss": 0.0436, "step": 21590 }, { "grad_norm": 0.13193447887897491, "learning_rate": 1.99503475958708e-05, "loss": 0.0259, "step": 21600 }, { "grad_norm": 0.19086401164531708, "learning_rate": 1.9906314435404484e-05, "loss": 0.0388, "step": 21610 }, { "grad_norm": 0.14477041363716125, "learning_rate": 1.986231784160378e-05, "loss": 0.0267, "step": 21620 }, { "grad_norm": 0.15138781070709229, "learning_rate": 1.9818357867928693e-05, "loss": 0.0365, "step": 21630 }, { "grad_norm": 0.14204023778438568, "learning_rate": 1.977443456779474e-05, "loss": 0.0268, "step": 21640 }, { "grad_norm": 0.17319995164871216, "learning_rate": 1.9730547994572858e-05, "loss": 0.0256, "step": 21650 }, { "grad_norm": 0.18385271728038788, "learning_rate": 1.9686698201589393e-05, "loss": 0.0355, "step": 21660 }, { "grad_norm": 0.16126340627670288, "learning_rate": 1.9642885242125964e-05, "loss": 0.0263, "step": 21670 }, { "grad_norm": 0.11147787421941757, "learning_rate": 1.9599109169419466e-05, "loss": 0.0258, "step": 21680 }, { "grad_norm": 0.18592438101768494, "learning_rate": 1.9555370036661948e-05, "loss": 0.0327, "step": 21690 }, { "grad_norm": 0.16050124168395996, "learning_rate": 1.9511667897000574e-05, "loss": 0.0308, "step": 21700 }, { "grad_norm": 0.18323960900306702, "learning_rate": 1.946800280353755e-05, "loss": 0.0267, "step": 21710 }, { "grad_norm": 0.1491936445236206, "learning_rate": 1.9424374809330114e-05, "loss": 0.043, "step": 21720 }, { "grad_norm": 0.18593376874923706, "learning_rate": 1.9380783967390382e-05, "loss": 0.0232, "step": 21730 }, { "grad_norm": 0.17879951000213623, "learning_rate": 1.933723033068533e-05, "loss": 0.0313, "step": 21740 }, { "grad_norm": 0.17527461051940918, "learning_rate": 1.929371395213674e-05, "loss": 0.0251, "step": 21750 }, { "grad_norm": 0.20616242289543152, "learning_rate": 1.9250234884621092e-05, "loss": 0.0362, "step": 21760 }, { "grad_norm": 0.2566409111022949, "learning_rate": 1.920679318096959e-05, "loss": 0.0381, "step": 21770 }, { "grad_norm": 0.14385095238685608, "learning_rate": 1.9163388893967982e-05, "loss": 0.0502, "step": 21780 }, { "grad_norm": 0.19524560868740082, "learning_rate": 1.9120022076356575e-05, "loss": 0.0279, "step": 21790 }, { "grad_norm": 0.2028251737356186, "learning_rate": 1.9076692780830114e-05, "loss": 0.0363, "step": 21800 }, { "grad_norm": 0.17486754059791565, "learning_rate": 1.903340106003782e-05, "loss": 0.0215, "step": 21810 }, { "grad_norm": 0.1146191656589508, "learning_rate": 1.8990146966583183e-05, "loss": 0.0242, "step": 21820 }, { "grad_norm": 0.22738708555698395, "learning_rate": 1.8946930553024035e-05, "loss": 0.0432, "step": 21830 }, { "grad_norm": 0.1487378031015396, "learning_rate": 1.8903751871872378e-05, "loss": 0.0296, "step": 21840 }, { "grad_norm": 0.133095845580101, "learning_rate": 1.8860610975594382e-05, "loss": 0.03, "step": 21850 }, { "grad_norm": 0.17250503599643707, "learning_rate": 1.8817507916610307e-05, "loss": 0.0372, "step": 21860 }, { "grad_norm": 0.13362239301204681, "learning_rate": 1.8774442747294407e-05, "loss": 0.035, "step": 21870 }, { "grad_norm": 0.13076874613761902, "learning_rate": 1.8731415519974965e-05, "loss": 0.0259, "step": 21880 }, { "grad_norm": 0.14679747819900513, "learning_rate": 1.86884262869341e-05, "loss": 0.0268, "step": 21890 }, { "grad_norm": 0.2250286042690277, "learning_rate": 1.8645475100407788e-05, "loss": 0.0265, "step": 21900 }, { "grad_norm": 0.2095121145248413, "learning_rate": 1.8602562012585768e-05, "loss": 0.0348, "step": 21910 }, { "grad_norm": 0.11173725873231888, "learning_rate": 1.8559687075611466e-05, "loss": 0.0228, "step": 21920 }, { "grad_norm": 0.1716179996728897, "learning_rate": 1.8516850341582017e-05, "loss": 0.0244, "step": 21930 }, { "grad_norm": 0.184444397687912, "learning_rate": 1.847405186254807e-05, "loss": 0.0264, "step": 21940 }, { "grad_norm": 0.1327427476644516, "learning_rate": 1.843129169051379e-05, "loss": 0.0188, "step": 21950 }, { "grad_norm": 0.17671984434127808, "learning_rate": 1.8388569877436863e-05, "loss": 0.027, "step": 21960 }, { "grad_norm": 0.1359911561012268, "learning_rate": 1.834588647522828e-05, "loss": 0.0202, "step": 21970 }, { "grad_norm": 0.18165627121925354, "learning_rate": 1.8303241535752435e-05, "loss": 0.0293, "step": 21980 }, { "grad_norm": 0.17452044785022736, "learning_rate": 1.8260635110826934e-05, "loss": 0.025, "step": 21990 }, { "grad_norm": 0.11570931226015091, "learning_rate": 1.8218067252222597e-05, "loss": 0.0222, "step": 22000 }, { "grad_norm": 0.16496850550174713, "learning_rate": 1.817553801166339e-05, "loss": 0.0171, "step": 22010 }, { "grad_norm": 0.10447271913290024, "learning_rate": 1.8133047440826333e-05, "loss": 0.021, "step": 22020 }, { "grad_norm": 0.10792049020528793, "learning_rate": 1.809059559134151e-05, "loss": 0.0299, "step": 22030 }, { "grad_norm": 0.12403341382741928, "learning_rate": 1.80481825147919e-05, "loss": 0.03, "step": 22040 }, { "grad_norm": 0.15826697647571564, "learning_rate": 1.8005808262713396e-05, "loss": 0.0225, "step": 22050 }, { "grad_norm": 0.1319727748632431, "learning_rate": 1.7963472886594714e-05, "loss": 0.0189, "step": 22060 }, { "grad_norm": 0.10074193775653839, "learning_rate": 1.7921176437877302e-05, "loss": 0.0248, "step": 22070 }, { "grad_norm": 0.1626395434141159, "learning_rate": 1.7878918967955366e-05, "loss": 0.0234, "step": 22080 }, { "grad_norm": 0.14710471034049988, "learning_rate": 1.7836700528175693e-05, "loss": 0.0279, "step": 22090 }, { "grad_norm": 0.13400223851203918, "learning_rate": 1.779452116983769e-05, "loss": 0.025, "step": 22100 }, { "grad_norm": 0.17121894657611847, "learning_rate": 1.775238094419325e-05, "loss": 0.0287, "step": 22110 }, { "grad_norm": 0.14161424338817596, "learning_rate": 1.7710279902446708e-05, "loss": 0.0398, "step": 22120 }, { "grad_norm": 0.12955687940120697, "learning_rate": 1.7668218095754795e-05, "loss": 0.021, "step": 22130 }, { "grad_norm": 0.12401886284351349, "learning_rate": 1.7626195575226594e-05, "loss": 0.0246, "step": 22140 }, { "grad_norm": 0.07011739164590836, "learning_rate": 1.758421239192343e-05, "loss": 0.0218, "step": 22150 }, { "grad_norm": 0.14793123304843903, "learning_rate": 1.7542268596858814e-05, "loss": 0.0208, "step": 22160 }, { "grad_norm": 0.09630024433135986, "learning_rate": 1.750036424099841e-05, "loss": 0.0183, "step": 22170 }, { "grad_norm": 0.21261560916900635, "learning_rate": 1.7458499375259956e-05, "loss": 0.0241, "step": 22180 }, { "grad_norm": 0.10704566538333893, "learning_rate": 1.7416674050513245e-05, "loss": 0.0251, "step": 22190 }, { "grad_norm": 0.1621054708957672, "learning_rate": 1.737488831757997e-05, "loss": 0.0271, "step": 22200 }, { "grad_norm": 0.19829992949962616, "learning_rate": 1.733314222723373e-05, "loss": 0.0277, "step": 22210 }, { "grad_norm": 0.15150372684001923, "learning_rate": 1.7291435830199954e-05, "loss": 0.0256, "step": 22220 }, { "grad_norm": 0.13119326531887054, "learning_rate": 1.7249769177155878e-05, "loss": 0.0199, "step": 22230 }, { "grad_norm": 0.2586275339126587, "learning_rate": 1.720814231873038e-05, "loss": 0.0405, "step": 22240 }, { "grad_norm": 0.12112568318843842, "learning_rate": 1.716655530550405e-05, "loss": 0.0182, "step": 22250 }, { "grad_norm": 0.1819823980331421, "learning_rate": 1.7125008188009017e-05, "loss": 0.0322, "step": 22260 }, { "grad_norm": 0.1028791293501854, "learning_rate": 1.708350101672894e-05, "loss": 0.0188, "step": 22270 }, { "grad_norm": 0.17057742178440094, "learning_rate": 1.7042033842098957e-05, "loss": 0.0265, "step": 22280 }, { "grad_norm": 0.12417320907115936, "learning_rate": 1.7000606714505567e-05, "loss": 0.0264, "step": 22290 }, { "grad_norm": 0.21096181869506836, "learning_rate": 1.695921968428668e-05, "loss": 0.0199, "step": 22300 }, { "grad_norm": 0.09061019122600555, "learning_rate": 1.6917872801731416e-05, "loss": 0.0308, "step": 22310 }, { "grad_norm": 0.211623415350914, "learning_rate": 1.687656611708014e-05, "loss": 0.0276, "step": 22320 }, { "grad_norm": 0.11808983236551285, "learning_rate": 1.6835299680524368e-05, "loss": 0.0208, "step": 22330 }, { "grad_norm": 0.12852376699447632, "learning_rate": 1.679407354220669e-05, "loss": 0.0323, "step": 22340 }, { "grad_norm": 0.12886618077754974, "learning_rate": 1.6752887752220793e-05, "loss": 0.0214, "step": 22350 }, { "grad_norm": 0.13725526630878448, "learning_rate": 1.6711742360611277e-05, "loss": 0.02, "step": 22360 }, { "grad_norm": 0.1520010232925415, "learning_rate": 1.6670637417373652e-05, "loss": 0.0306, "step": 22370 }, { "grad_norm": 0.1612754464149475, "learning_rate": 1.6629572972454333e-05, "loss": 0.0268, "step": 22380 }, { "grad_norm": 0.160241037607193, "learning_rate": 1.6588549075750464e-05, "loss": 0.0147, "step": 22390 }, { "grad_norm": 0.21872030198574066, "learning_rate": 1.6547565777109976e-05, "loss": 0.0285, "step": 22400 }, { "grad_norm": 0.12778924405574799, "learning_rate": 1.650662312633143e-05, "loss": 0.0312, "step": 22410 }, { "grad_norm": 0.22537663578987122, "learning_rate": 1.6465721173164002e-05, "loss": 0.0339, "step": 22420 }, { "grad_norm": 0.14660532772541046, "learning_rate": 1.6424859967307428e-05, "loss": 0.0177, "step": 22430 }, { "grad_norm": 0.2062496542930603, "learning_rate": 1.6384039558411903e-05, "loss": 0.0214, "step": 22440 }, { "grad_norm": 0.1851368397474289, "learning_rate": 1.634325999607811e-05, "loss": 0.0278, "step": 22450 }, { "grad_norm": 0.18888473510742188, "learning_rate": 1.6302521329857045e-05, "loss": 0.028, "step": 22460 }, { "grad_norm": 0.10405266284942627, "learning_rate": 1.6261823609250027e-05, "loss": 0.026, "step": 22470 }, { "grad_norm": 0.19816018640995026, "learning_rate": 1.6221166883708627e-05, "loss": 0.0272, "step": 22480 }, { "grad_norm": 0.20710442960262299, "learning_rate": 1.6180551202634603e-05, "loss": 0.0279, "step": 22490 }, { "grad_norm": 0.17816093564033508, "learning_rate": 1.613997661537981e-05, "loss": 0.024, "step": 22500 }, { "grad_norm": 0.14775338768959045, "learning_rate": 1.609944317124624e-05, "loss": 0.0168, "step": 22510 }, { "grad_norm": 0.13392624258995056, "learning_rate": 1.605895091948582e-05, "loss": 0.0284, "step": 22520 }, { "grad_norm": 0.21937009692192078, "learning_rate": 1.6018499909300477e-05, "loss": 0.0211, "step": 22530 }, { "grad_norm": 0.1867634356021881, "learning_rate": 1.597809018984199e-05, "loss": 0.0136, "step": 22540 }, { "grad_norm": 0.09212896227836609, "learning_rate": 1.593772181021196e-05, "loss": 0.0347, "step": 22550 }, { "grad_norm": 0.24576757848262787, "learning_rate": 1.5897394819461813e-05, "loss": 0.0307, "step": 22560 }, { "grad_norm": 0.16472811996936798, "learning_rate": 1.5857109266592625e-05, "loss": 0.0191, "step": 22570 }, { "grad_norm": 0.06850098818540573, "learning_rate": 1.581686520055514e-05, "loss": 0.0117, "step": 22580 }, { "grad_norm": 0.12753242254257202, "learning_rate": 1.5776662670249704e-05, "loss": 0.0106, "step": 22590 }, { "grad_norm": 0.1682201623916626, "learning_rate": 1.573650172452615e-05, "loss": 0.016, "step": 22600 }, { "grad_norm": 0.2092752605676651, "learning_rate": 1.5696382412183852e-05, "loss": 0.0152, "step": 22610 }, { "grad_norm": 0.10181137174367905, "learning_rate": 1.565630478197155e-05, "loss": 0.0204, "step": 22620 }, { "grad_norm": 0.1383846253156662, "learning_rate": 1.5616268882587332e-05, "loss": 0.0193, "step": 22630 }, { "grad_norm": 0.1244281679391861, "learning_rate": 1.5576274762678594e-05, "loss": 0.0328, "step": 22640 }, { "grad_norm": 0.1629749983549118, "learning_rate": 1.5536322470841952e-05, "loss": 0.0196, "step": 22650 }, { "grad_norm": 0.13924728333950043, "learning_rate": 1.549641205562324e-05, "loss": 0.024, "step": 22660 }, { "grad_norm": 0.08170490711927414, "learning_rate": 1.5456543565517334e-05, "loss": 0.0146, "step": 22670 }, { "grad_norm": 0.14456412196159363, "learning_rate": 1.541671704896825e-05, "loss": 0.0245, "step": 22680 }, { "grad_norm": 0.10173783451318741, "learning_rate": 1.5376932554368938e-05, "loss": 0.0163, "step": 22690 }, { "grad_norm": 0.25308695435523987, "learning_rate": 1.5337190130061318e-05, "loss": 0.0208, "step": 22700 }, { "grad_norm": 0.18977485597133636, "learning_rate": 1.5297489824336165e-05, "loss": 0.0281, "step": 22710 }, { "grad_norm": 0.12410629540681839, "learning_rate": 1.5257831685433121e-05, "loss": 0.0111, "step": 22720 }, { "grad_norm": 0.2330123782157898, "learning_rate": 1.521821576154055e-05, "loss": 0.0195, "step": 22730 }, { "grad_norm": 0.1197102963924408, "learning_rate": 1.5178642100795542e-05, "loss": 0.0156, "step": 22740 }, { "grad_norm": 0.14425671100616455, "learning_rate": 1.5139110751283819e-05, "loss": 0.0234, "step": 22750 }, { "grad_norm": 0.16794535517692566, "learning_rate": 1.5099621761039683e-05, "loss": 0.0424, "step": 22760 }, { "grad_norm": 0.12472505867481232, "learning_rate": 1.5060175178046016e-05, "loss": 0.0189, "step": 22770 }, { "grad_norm": 0.08263107389211655, "learning_rate": 1.5020771050234117e-05, "loss": 0.0163, "step": 22780 }, { "grad_norm": 0.12266583740711212, "learning_rate": 1.4981409425483716e-05, "loss": 0.0169, "step": 22790 }, { "grad_norm": 0.12008442729711533, "learning_rate": 1.4942090351622883e-05, "loss": 0.012, "step": 22800 }, { "grad_norm": 0.12352461367845535, "learning_rate": 1.490281387642804e-05, "loss": 0.0191, "step": 22810 }, { "grad_norm": 0.13138462603092194, "learning_rate": 1.4863580047623765e-05, "loss": 0.0222, "step": 22820 }, { "grad_norm": 0.14958550035953522, "learning_rate": 1.4824388912882897e-05, "loss": 0.0137, "step": 22830 }, { "grad_norm": 0.10854235291481018, "learning_rate": 1.4785240519826344e-05, "loss": 0.034, "step": 22840 }, { "grad_norm": 0.20903536677360535, "learning_rate": 1.4746134916023097e-05, "loss": 0.027, "step": 22850 }, { "grad_norm": 0.1701909750699997, "learning_rate": 1.4707072148990141e-05, "loss": 0.0168, "step": 22860 }, { "grad_norm": 0.15308234095573425, "learning_rate": 1.4668052266192423e-05, "loss": 0.0189, "step": 22870 }, { "grad_norm": 0.15250664949417114, "learning_rate": 1.4629075315042795e-05, "loss": 0.0362, "step": 22880 }, { "grad_norm": 0.1451135277748108, "learning_rate": 1.4590141342901925e-05, "loss": 0.0146, "step": 22890 }, { "grad_norm": 0.19500435888767242, "learning_rate": 1.4551250397078253e-05, "loss": 0.0344, "step": 22900 }, { "grad_norm": 0.12066467851400375, "learning_rate": 1.4512402524827945e-05, "loss": 0.0239, "step": 22910 }, { "grad_norm": 0.22603604197502136, "learning_rate": 1.447359777335482e-05, "loss": 0.0271, "step": 22920 }, { "grad_norm": 0.13262906670570374, "learning_rate": 1.4434836189810335e-05, "loss": 0.0152, "step": 22930 }, { "grad_norm": 0.15246163308620453, "learning_rate": 1.4396117821293453e-05, "loss": 0.017, "step": 22940 }, { "grad_norm": 0.1991804540157318, "learning_rate": 1.4357442714850633e-05, "loss": 0.0249, "step": 22950 }, { "grad_norm": 0.13018935918807983, "learning_rate": 1.43188109174758e-05, "loss": 0.0146, "step": 22960 }, { "grad_norm": 0.102576345205307, "learning_rate": 1.4280222476110206e-05, "loss": 0.0173, "step": 22970 }, { "grad_norm": 0.13773828744888306, "learning_rate": 1.4241677437642469e-05, "loss": 0.0313, "step": 22980 }, { "grad_norm": 0.21857167780399323, "learning_rate": 1.420317584890844e-05, "loss": 0.0291, "step": 22990 }, { "grad_norm": 0.15346559882164001, "learning_rate": 1.4164717756691176e-05, "loss": 0.0189, "step": 23000 }, { "grad_norm": 0.1369662880897522, "learning_rate": 1.4126303207720882e-05, "loss": 0.0205, "step": 23010 }, { "grad_norm": 0.12875066697597504, "learning_rate": 1.408793224867484e-05, "loss": 0.0215, "step": 23020 }, { "grad_norm": 0.16170501708984375, "learning_rate": 1.4049604926177423e-05, "loss": 0.0247, "step": 23030 }, { "grad_norm": 0.18139630556106567, "learning_rate": 1.4011321286799916e-05, "loss": 0.0274, "step": 23040 }, { "grad_norm": 0.1735920011997223, "learning_rate": 1.3973081377060565e-05, "loss": 0.0222, "step": 23050 }, { "grad_norm": 0.1900762915611267, "learning_rate": 1.393488524342445e-05, "loss": 0.0181, "step": 23060 }, { "grad_norm": 0.18461140990257263, "learning_rate": 1.3896732932303486e-05, "loss": 0.0202, "step": 23070 }, { "grad_norm": 0.15603899955749512, "learning_rate": 1.3858624490056304e-05, "loss": 0.0202, "step": 23080 }, { "grad_norm": 0.08259700238704681, "learning_rate": 1.3820559962988299e-05, "loss": 0.0226, "step": 23090 }, { "grad_norm": 0.1462545543909073, "learning_rate": 1.3782539397351418e-05, "loss": 0.0288, "step": 23100 }, { "grad_norm": 0.07380040735006332, "learning_rate": 1.3744562839344266e-05, "loss": 0.0302, "step": 23110 }, { "grad_norm": 0.14781923592090607, "learning_rate": 1.370663033511193e-05, "loss": 0.0258, "step": 23120 }, { "grad_norm": 0.07815045863389969, "learning_rate": 1.3668741930745964e-05, "loss": 0.0181, "step": 23130 }, { "grad_norm": 0.14859843254089355, "learning_rate": 1.3630897672284382e-05, "loss": 0.0174, "step": 23140 }, { "grad_norm": 0.1314459592103958, "learning_rate": 1.3593097605711508e-05, "loss": 0.0342, "step": 23150 }, { "grad_norm": 0.12771187722682953, "learning_rate": 1.3555341776957992e-05, "loss": 0.0196, "step": 23160 }, { "grad_norm": 0.16058263182640076, "learning_rate": 1.3517630231900724e-05, "loss": 0.0224, "step": 23170 }, { "grad_norm": 0.09042187035083771, "learning_rate": 1.3479963016362767e-05, "loss": 0.0156, "step": 23180 }, { "grad_norm": 0.1916525512933731, "learning_rate": 1.3442340176113377e-05, "loss": 0.0235, "step": 23190 }, { "grad_norm": 0.0819987952709198, "learning_rate": 1.340476175686784e-05, "loss": 0.0209, "step": 23200 }, { "grad_norm": 0.095334492623806, "learning_rate": 1.3367227804287469e-05, "loss": 0.0202, "step": 23210 }, { "grad_norm": 0.15073801577091217, "learning_rate": 1.332973836397956e-05, "loss": 0.0197, "step": 23220 }, { "grad_norm": 0.18808776140213013, "learning_rate": 1.329229348149731e-05, "loss": 0.0192, "step": 23230 }, { "grad_norm": 0.14016996324062347, "learning_rate": 1.3254893202339796e-05, "loss": 0.0255, "step": 23240 }, { "grad_norm": 0.09992094337940216, "learning_rate": 1.3217537571951872e-05, "loss": 0.0232, "step": 23250 }, { "grad_norm": 0.1241409033536911, "learning_rate": 1.3180226635724169e-05, "loss": 0.0192, "step": 23260 }, { "grad_norm": 0.20864833891391754, "learning_rate": 1.314296043899298e-05, "loss": 0.0286, "step": 23270 }, { "grad_norm": 0.10215700417757034, "learning_rate": 1.3105739027040248e-05, "loss": 0.0273, "step": 23280 }, { "grad_norm": 0.12541939318180084, "learning_rate": 1.3068562445093491e-05, "loss": 0.0295, "step": 23290 }, { "grad_norm": 0.1635899692773819, "learning_rate": 1.3031430738325778e-05, "loss": 0.021, "step": 23300 }, { "grad_norm": 0.18829700350761414, "learning_rate": 1.299434395185563e-05, "loss": 0.0348, "step": 23310 }, { "grad_norm": 0.24512282013893127, "learning_rate": 1.295730213074699e-05, "loss": 0.0228, "step": 23320 }, { "grad_norm": 0.116411492228508, "learning_rate": 1.2920305320009152e-05, "loss": 0.0289, "step": 23330 }, { "grad_norm": 0.1469888538122177, "learning_rate": 1.2883353564596729e-05, "loss": 0.0321, "step": 23340 }, { "grad_norm": 0.17447881400585175, "learning_rate": 1.2846446909409604e-05, "loss": 0.0268, "step": 23350 }, { "grad_norm": 0.21254107356071472, "learning_rate": 1.280958539929284e-05, "loss": 0.0308, "step": 23360 }, { "grad_norm": 0.18992529809474945, "learning_rate": 1.277276907903664e-05, "loss": 0.0255, "step": 23370 }, { "grad_norm": 0.1690300852060318, "learning_rate": 1.2735997993376298e-05, "loss": 0.0423, "step": 23380 }, { "grad_norm": 0.2128286510705948, "learning_rate": 1.2699272186992166e-05, "loss": 0.0308, "step": 23390 }, { "grad_norm": 0.13769066333770752, "learning_rate": 1.2662591704509547e-05, "loss": 0.0257, "step": 23400 }, { "grad_norm": 0.1910003274679184, "learning_rate": 1.2625956590498711e-05, "loss": 0.0423, "step": 23410 }, { "grad_norm": 0.18454618752002716, "learning_rate": 1.2589366889474757e-05, "loss": 0.0202, "step": 23420 }, { "grad_norm": 0.12370724231004715, "learning_rate": 1.2552822645897621e-05, "loss": 0.0227, "step": 23430 }, { "grad_norm": 0.1309249848127365, "learning_rate": 1.2516323904172e-05, "loss": 0.0246, "step": 23440 }, { "grad_norm": 0.1982060670852661, "learning_rate": 1.2479870708647324e-05, "loss": 0.0248, "step": 23450 }, { "grad_norm": 0.0761120468378067, "learning_rate": 1.2443463103617658e-05, "loss": 0.0207, "step": 23460 }, { "grad_norm": 0.18510165810585022, "learning_rate": 1.240710113332167e-05, "loss": 0.0268, "step": 23470 }, { "grad_norm": 0.10180258750915527, "learning_rate": 1.2370784841942584e-05, "loss": 0.0213, "step": 23480 }, { "grad_norm": 0.18450966477394104, "learning_rate": 1.2334514273608117e-05, "loss": 0.0287, "step": 23490 }, { "grad_norm": 0.10581308603286743, "learning_rate": 1.2298289472390417e-05, "loss": 0.0293, "step": 23500 }, { "grad_norm": 0.13073809444904327, "learning_rate": 1.226211048230606e-05, "loss": 0.0282, "step": 23510 }, { "grad_norm": 0.19417846202850342, "learning_rate": 1.2225977347315921e-05, "loss": 0.0501, "step": 23520 }, { "grad_norm": 0.12955628335475922, "learning_rate": 1.2189890111325148e-05, "loss": 0.0211, "step": 23530 }, { "grad_norm": 0.10793203860521317, "learning_rate": 1.2153848818183161e-05, "loss": 0.0276, "step": 23540 }, { "grad_norm": 0.1660565435886383, "learning_rate": 1.2117853511683507e-05, "loss": 0.0191, "step": 23550 }, { "grad_norm": 0.14769227802753448, "learning_rate": 1.2081904235563906e-05, "loss": 0.0214, "step": 23560 }, { "grad_norm": 0.14926324784755707, "learning_rate": 1.20460010335061e-05, "loss": 0.023, "step": 23570 }, { "grad_norm": 0.1650434285402298, "learning_rate": 1.2010143949135865e-05, "loss": 0.0254, "step": 23580 }, { "grad_norm": 0.20322354137897491, "learning_rate": 1.1974333026022938e-05, "loss": 0.0356, "step": 23590 }, { "grad_norm": 0.19024328887462616, "learning_rate": 1.1938568307680964e-05, "loss": 0.0281, "step": 23600 }, { "grad_norm": 0.11561702191829681, "learning_rate": 1.1902849837567464e-05, "loss": 0.0298, "step": 23610 }, { "grad_norm": 0.12755048274993896, "learning_rate": 1.1867177659083739e-05, "loss": 0.0251, "step": 23620 }, { "grad_norm": 0.12279576063156128, "learning_rate": 1.1831551815574848e-05, "loss": 0.0202, "step": 23630 }, { "grad_norm": 0.18917034566402435, "learning_rate": 1.1795972350329554e-05, "loss": 0.0379, "step": 23640 }, { "grad_norm": 0.18388713896274567, "learning_rate": 1.176043930658024e-05, "loss": 0.0271, "step": 23650 }, { "grad_norm": 0.09998369961977005, "learning_rate": 1.1724952727502942e-05, "loss": 0.0236, "step": 23660 }, { "grad_norm": 0.12202000617980957, "learning_rate": 1.1689512656217178e-05, "loss": 0.0261, "step": 23670 }, { "grad_norm": 0.10024774819612503, "learning_rate": 1.1654119135785962e-05, "loss": 0.0315, "step": 23680 }, { "grad_norm": 0.20308011770248413, "learning_rate": 1.1618772209215794e-05, "loss": 0.0188, "step": 23690 }, { "grad_norm": 0.15022382140159607, "learning_rate": 1.1583471919456506e-05, "loss": 0.0275, "step": 23700 }, { "grad_norm": 0.18528737127780914, "learning_rate": 1.1548218309401266e-05, "loss": 0.0206, "step": 23710 }, { "grad_norm": 0.1497962474822998, "learning_rate": 1.1513011421886555e-05, "loss": 0.0239, "step": 23720 }, { "grad_norm": 0.1684035062789917, "learning_rate": 1.1477851299692056e-05, "loss": 0.036, "step": 23730 }, { "grad_norm": 0.10088558495044708, "learning_rate": 1.144273798554063e-05, "loss": 0.0142, "step": 23740 }, { "grad_norm": 0.13090604543685913, "learning_rate": 1.1407671522098262e-05, "loss": 0.0222, "step": 23750 }, { "grad_norm": 0.16485197842121124, "learning_rate": 1.1372651951974e-05, "loss": 0.0275, "step": 23760 }, { "grad_norm": 0.14805015921592712, "learning_rate": 1.1337679317719952e-05, "loss": 0.0198, "step": 23770 }, { "grad_norm": 0.19167007505893707, "learning_rate": 1.1302753661831151e-05, "loss": 0.0235, "step": 23780 }, { "grad_norm": 0.1031511127948761, "learning_rate": 1.1267875026745561e-05, "loss": 0.02, "step": 23790 }, { "grad_norm": 0.10744258761405945, "learning_rate": 1.1233043454844016e-05, "loss": 0.0259, "step": 23800 }, { "grad_norm": 0.12728653848171234, "learning_rate": 1.1198258988450144e-05, "loss": 0.0237, "step": 23810 }, { "grad_norm": 0.12758736312389374, "learning_rate": 1.1163521669830368e-05, "loss": 0.0152, "step": 23820 }, { "grad_norm": 0.11084078252315521, "learning_rate": 1.1128831541193818e-05, "loss": 0.0159, "step": 23830 }, { "grad_norm": 0.21825888752937317, "learning_rate": 1.1094188644692255e-05, "loss": 0.0303, "step": 23840 }, { "grad_norm": 0.12620306015014648, "learning_rate": 1.1059593022420067e-05, "loss": 0.0194, "step": 23850 }, { "grad_norm": 0.17251573503017426, "learning_rate": 1.1025044716414184e-05, "loss": 0.0204, "step": 23860 }, { "grad_norm": 0.18829214572906494, "learning_rate": 1.0990543768654082e-05, "loss": 0.0304, "step": 23870 }, { "grad_norm": 0.14793407917022705, "learning_rate": 1.0956090221061655e-05, "loss": 0.0187, "step": 23880 }, { "grad_norm": 0.10214345157146454, "learning_rate": 1.0921684115501208e-05, "loss": 0.0247, "step": 23890 }, { "grad_norm": 0.08635331690311432, "learning_rate": 1.0887325493779405e-05, "loss": 0.0184, "step": 23900 }, { "grad_norm": 0.1766044944524765, "learning_rate": 1.0853014397645207e-05, "loss": 0.0182, "step": 23910 }, { "grad_norm": 0.15855924785137177, "learning_rate": 1.0818750868789829e-05, "loss": 0.0289, "step": 23920 }, { "grad_norm": 0.11482121795415878, "learning_rate": 1.0784534948846703e-05, "loss": 0.0162, "step": 23930 }, { "grad_norm": 0.14736749231815338, "learning_rate": 1.0750366679391393e-05, "loss": 0.0178, "step": 23940 }, { "grad_norm": 0.11545604467391968, "learning_rate": 1.0716246101941557e-05, "loss": 0.0176, "step": 23950 }, { "grad_norm": 0.22904285788536072, "learning_rate": 1.0682173257956934e-05, "loss": 0.0278, "step": 23960 }, { "grad_norm": 0.14020408689975739, "learning_rate": 1.064814818883922e-05, "loss": 0.0285, "step": 23970 }, { "grad_norm": 0.1511145681142807, "learning_rate": 1.06141709359321e-05, "loss": 0.021, "step": 23980 }, { "grad_norm": 0.12491977214813232, "learning_rate": 1.058024154052114e-05, "loss": 0.0326, "step": 23990 }, { "grad_norm": 0.22263625264167786, "learning_rate": 1.0546360043833747e-05, "loss": 0.0218, "step": 24000 }, { "grad_norm": 0.11706625670194626, "learning_rate": 1.0512526487039138e-05, "loss": 0.0174, "step": 24010 }, { "grad_norm": 0.12343801558017731, "learning_rate": 1.0478740911248258e-05, "loss": 0.0222, "step": 24020 }, { "grad_norm": 0.14112496376037598, "learning_rate": 1.0445003357513793e-05, "loss": 0.0223, "step": 24030 }, { "grad_norm": 0.17291833460330963, "learning_rate": 1.0411313866830042e-05, "loss": 0.0233, "step": 24040 }, { "grad_norm": 0.12825600802898407, "learning_rate": 1.0377672480132915e-05, "loss": 0.0217, "step": 24050 }, { "grad_norm": 0.09165316820144653, "learning_rate": 1.0344079238299864e-05, "loss": 0.026, "step": 24060 }, { "grad_norm": 0.08885050565004349, "learning_rate": 1.0310534182149833e-05, "loss": 0.0191, "step": 24070 }, { "grad_norm": 0.14408963918685913, "learning_rate": 1.0277037352443258e-05, "loss": 0.0321, "step": 24080 }, { "grad_norm": 0.09331661462783813, "learning_rate": 1.024358878988193e-05, "loss": 0.0191, "step": 24090 }, { "grad_norm": 0.09186235815286636, "learning_rate": 1.0210188535108995e-05, "loss": 0.0216, "step": 24100 }, { "grad_norm": 0.07130032032728195, "learning_rate": 1.0176836628708935e-05, "loss": 0.0193, "step": 24110 }, { "grad_norm": 0.25738272070884705, "learning_rate": 1.0143533111207454e-05, "loss": 0.0151, "step": 24120 }, { "grad_norm": 0.09443286061286926, "learning_rate": 1.0110278023071446e-05, "loss": 0.0208, "step": 24130 }, { "grad_norm": 0.13584457337856293, "learning_rate": 1.007707140470901e-05, "loss": 0.0235, "step": 24140 }, { "grad_norm": 0.19587217271327972, "learning_rate": 1.0043913296469298e-05, "loss": 0.0144, "step": 24150 }, { "grad_norm": 0.07818237692117691, "learning_rate": 1.001080373864255e-05, "loss": 0.0165, "step": 24160 }, { "grad_norm": 0.184669628739357, "learning_rate": 9.977742771459991e-06, "loss": 0.0219, "step": 24170 }, { "grad_norm": 0.16990022361278534, "learning_rate": 9.944730435093802e-06, "loss": 0.0186, "step": 24180 }, { "grad_norm": 0.09556028246879578, "learning_rate": 9.911766769657115e-06, "loss": 0.022, "step": 24190 }, { "grad_norm": 0.09026476740837097, "learning_rate": 9.878851815203882e-06, "loss": 0.0176, "step": 24200 }, { "grad_norm": 0.11792144179344177, "learning_rate": 9.845985611728864e-06, "loss": 0.021, "step": 24210 }, { "grad_norm": 0.1016387939453125, "learning_rate": 9.813168199167606e-06, "loss": 0.0293, "step": 24220 }, { "grad_norm": 0.14798597991466522, "learning_rate": 9.780399617396341e-06, "loss": 0.0273, "step": 24230 }, { "grad_norm": 0.103168785572052, "learning_rate": 9.747679906232016e-06, "loss": 0.016, "step": 24240 }, { "grad_norm": 0.07252749800682068, "learning_rate": 9.715009105432138e-06, "loss": 0.0233, "step": 24250 }, { "grad_norm": 0.15792019665241241, "learning_rate": 9.682387254694835e-06, "loss": 0.0184, "step": 24260 }, { "grad_norm": 0.16793875396251678, "learning_rate": 9.649814393658723e-06, "loss": 0.0174, "step": 24270 }, { "grad_norm": 0.23016968369483948, "learning_rate": 9.617290561902881e-06, "loss": 0.0208, "step": 24280 }, { "grad_norm": 0.14824949204921722, "learning_rate": 9.584815798946861e-06, "loss": 0.0334, "step": 24290 }, { "grad_norm": 0.1094171404838562, "learning_rate": 9.55239014425055e-06, "loss": 0.0231, "step": 24300 }, { "grad_norm": 0.25008052587509155, "learning_rate": 9.520013637214176e-06, "loss": 0.0173, "step": 24310 }, { "grad_norm": 0.13960811495780945, "learning_rate": 9.48768631717824e-06, "loss": 0.0201, "step": 24320 }, { "grad_norm": 0.08881165087223053, "learning_rate": 9.455408223423496e-06, "loss": 0.0144, "step": 24330 }, { "grad_norm": 0.16311708092689514, "learning_rate": 9.423179395170845e-06, "loss": 0.018, "step": 24340 }, { "grad_norm": 0.10306548327207565, "learning_rate": 9.390999871581391e-06, "loss": 0.0273, "step": 24350 }, { "grad_norm": 0.08202516287565231, "learning_rate": 9.358869691756273e-06, "loss": 0.0275, "step": 24360 }, { "grad_norm": 0.11817525327205658, "learning_rate": 9.326788894736688e-06, "loss": 0.0168, "step": 24370 }, { "grad_norm": 0.11153603345155716, "learning_rate": 9.294757519503811e-06, "loss": 0.0143, "step": 24380 }, { "grad_norm": 0.08622651547193527, "learning_rate": 9.262775604978819e-06, "loss": 0.0177, "step": 24390 }, { "grad_norm": 0.2208944410085678, "learning_rate": 9.230843190022725e-06, "loss": 0.0274, "step": 24400 }, { "grad_norm": 0.10807009786367416, "learning_rate": 9.198960313436445e-06, "loss": 0.0185, "step": 24410 }, { "grad_norm": 0.14714765548706055, "learning_rate": 9.16712701396067e-06, "loss": 0.0216, "step": 24420 }, { "grad_norm": 0.15060792863368988, "learning_rate": 9.135343330275864e-06, "loss": 0.0153, "step": 24430 }, { "grad_norm": 0.09967207908630371, "learning_rate": 9.10360930100218e-06, "loss": 0.011, "step": 24440 }, { "grad_norm": 0.16901037096977234, "learning_rate": 9.07192496469949e-06, "loss": 0.0283, "step": 24450 }, { "grad_norm": 0.09882207959890366, "learning_rate": 9.040290359867231e-06, "loss": 0.0277, "step": 24460 }, { "grad_norm": 0.06998378038406372, "learning_rate": 9.008705524944438e-06, "loss": 0.0161, "step": 24470 }, { "grad_norm": 0.09123891592025757, "learning_rate": 8.977170498309651e-06, "loss": 0.0237, "step": 24480 }, { "grad_norm": 0.12081175297498703, "learning_rate": 8.945685318280916e-06, "loss": 0.0198, "step": 24490 }, { "grad_norm": 0.13055261969566345, "learning_rate": 8.914250023115672e-06, "loss": 0.0236, "step": 24500 }, { "grad_norm": 0.1326124519109726, "learning_rate": 8.882864651010798e-06, "loss": 0.0133, "step": 24510 }, { "grad_norm": 0.1378897726535797, "learning_rate": 8.851529240102462e-06, "loss": 0.0294, "step": 24520 }, { "grad_norm": 0.15275616943836212, "learning_rate": 8.820243828466134e-06, "loss": 0.0154, "step": 24530 }, { "grad_norm": 0.16484276950359344, "learning_rate": 8.789008454116566e-06, "loss": 0.0205, "step": 24540 }, { "grad_norm": 0.13871724903583527, "learning_rate": 8.757823155007655e-06, "loss": 0.0202, "step": 24550 }, { "grad_norm": 0.08396141976118088, "learning_rate": 8.72668796903251e-06, "loss": 0.0166, "step": 24560 }, { "grad_norm": 0.13495312631130219, "learning_rate": 8.695602934023301e-06, "loss": 0.0199, "step": 24570 }, { "grad_norm": 0.18777784705162048, "learning_rate": 8.664568087751274e-06, "loss": 0.022, "step": 24580 }, { "grad_norm": 0.1697889119386673, "learning_rate": 8.633583467926697e-06, "loss": 0.015, "step": 24590 }, { "grad_norm": 0.10911808907985687, "learning_rate": 8.602649112198796e-06, "loss": 0.0193, "step": 24600 }, { "grad_norm": 0.1463003307580948, "learning_rate": 8.571765058155745e-06, "loss": 0.0245, "step": 24610 }, { "grad_norm": 0.10676056891679764, "learning_rate": 8.540931343324583e-06, "loss": 0.018, "step": 24620 }, { "grad_norm": 0.12294916063547134, "learning_rate": 8.51014800517117e-06, "loss": 0.0251, "step": 24630 }, { "grad_norm": 0.14716637134552002, "learning_rate": 8.479415081100167e-06, "loss": 0.0238, "step": 24640 }, { "grad_norm": 0.12477454543113708, "learning_rate": 8.448732608454967e-06, "loss": 0.0187, "step": 24650 }, { "grad_norm": 0.15064896643161774, "learning_rate": 8.418100624517688e-06, "loss": 0.0238, "step": 24660 }, { "grad_norm": 0.13337920606136322, "learning_rate": 8.387519166509062e-06, "loss": 0.0289, "step": 24670 }, { "grad_norm": 0.13688629865646362, "learning_rate": 8.356988271588445e-06, "loss": 0.0193, "step": 24680 }, { "grad_norm": 0.11503276973962784, "learning_rate": 8.326507976853765e-06, "loss": 0.0264, "step": 24690 }, { "grad_norm": 0.19762413203716278, "learning_rate": 8.296078319341443e-06, "loss": 0.0361, "step": 24700 }, { "grad_norm": 0.1907636970281601, "learning_rate": 8.265699336026383e-06, "loss": 0.0267, "step": 24710 }, { "grad_norm": 0.12171091884374619, "learning_rate": 8.235371063821923e-06, "loss": 0.0276, "step": 24720 }, { "grad_norm": 0.14329156279563904, "learning_rate": 8.205093539579767e-06, "loss": 0.0267, "step": 24730 }, { "grad_norm": 0.14906270802021027, "learning_rate": 8.174866800089963e-06, "loss": 0.0265, "step": 24740 }, { "grad_norm": 0.13819992542266846, "learning_rate": 8.144690882080853e-06, "loss": 0.0219, "step": 24750 }, { "grad_norm": 0.17218300700187683, "learning_rate": 8.114565822219006e-06, "loss": 0.0352, "step": 24760 }, { "grad_norm": 0.12954777479171753, "learning_rate": 8.084491657109233e-06, "loss": 0.029, "step": 24770 }, { "grad_norm": 0.15865136682987213, "learning_rate": 8.054468423294476e-06, "loss": 0.0289, "step": 24780 }, { "grad_norm": 0.11736460030078888, "learning_rate": 8.024496157255785e-06, "loss": 0.0202, "step": 24790 }, { "grad_norm": 0.15524296462535858, "learning_rate": 7.994574895412293e-06, "loss": 0.0176, "step": 24800 }, { "grad_norm": 0.19071373343467712, "learning_rate": 7.964704674121149e-06, "loss": 0.022, "step": 24810 }, { "grad_norm": 0.12949849665164948, "learning_rate": 7.934885529677505e-06, "loss": 0.0147, "step": 24820 }, { "grad_norm": 0.1156974732875824, "learning_rate": 7.905117498314412e-06, "loss": 0.0295, "step": 24830 }, { "grad_norm": 0.1899183839559555, "learning_rate": 7.87540061620286e-06, "loss": 0.0273, "step": 24840 }, { "grad_norm": 0.07511170953512192, "learning_rate": 7.845734919451647e-06, "loss": 0.022, "step": 24850 }, { "grad_norm": 0.09703709185123444, "learning_rate": 7.816120444107383e-06, "loss": 0.0146, "step": 24860 }, { "grad_norm": 0.18824000656604767, "learning_rate": 7.786557226154473e-06, "loss": 0.0255, "step": 24870 }, { "grad_norm": 0.13681839406490326, "learning_rate": 7.757045301514998e-06, "loss": 0.0158, "step": 24880 }, { "grad_norm": 0.10210196673870087, "learning_rate": 7.727584706048735e-06, "loss": 0.0238, "step": 24890 }, { "grad_norm": 0.2133311629295349, "learning_rate": 7.698175475553077e-06, "loss": 0.0296, "step": 24900 }, { "grad_norm": 0.16117626428604126, "learning_rate": 7.668817645763021e-06, "loss": 0.0191, "step": 24910 }, { "grad_norm": 0.11546635627746582, "learning_rate": 7.639511252351088e-06, "loss": 0.0148, "step": 24920 }, { "grad_norm": 0.08173678070306778, "learning_rate": 7.610256330927324e-06, "loss": 0.0181, "step": 24930 }, { "grad_norm": 0.07260394096374512, "learning_rate": 7.581052917039211e-06, "loss": 0.0232, "step": 24940 }, { "grad_norm": 0.15569333732128143, "learning_rate": 7.551901046171645e-06, "loss": 0.0277, "step": 24950 }, { "grad_norm": 0.13387908041477203, "learning_rate": 7.522800753746895e-06, "loss": 0.0177, "step": 24960 }, { "grad_norm": 0.06866181641817093, "learning_rate": 7.493752075124577e-06, "loss": 0.0247, "step": 24970 }, { "grad_norm": 0.13509854674339294, "learning_rate": 7.464755045601557e-06, "loss": 0.0212, "step": 24980 }, { "grad_norm": 0.14304110407829285, "learning_rate": 7.435809700411972e-06, "loss": 0.0345, "step": 24990 }, { "grad_norm": 0.1869986653327942, "learning_rate": 7.40691607472715e-06, "loss": 0.022, "step": 25000 }, { "grad_norm": 0.15831121802330017, "learning_rate": 7.3780742036555606e-06, "loss": 0.0285, "step": 25010 }, { "grad_norm": 0.1576581448316574, "learning_rate": 7.349284122242783e-06, "loss": 0.0234, "step": 25020 }, { "grad_norm": 0.13359220325946808, "learning_rate": 7.320545865471512e-06, "loss": 0.0314, "step": 25030 }, { "grad_norm": 0.12899549305438995, "learning_rate": 7.291859468261425e-06, "loss": 0.0213, "step": 25040 }, { "grad_norm": 0.15319222211837769, "learning_rate": 7.263224965469195e-06, "loss": 0.0168, "step": 25050 }, { "grad_norm": 0.13224028050899506, "learning_rate": 7.234642391888446e-06, "loss": 0.0266, "step": 25060 }, { "grad_norm": 0.15819133818149567, "learning_rate": 7.206111782249697e-06, "loss": 0.0182, "step": 25070 }, { "grad_norm": 0.15242885053157806, "learning_rate": 7.1776331712203385e-06, "loss": 0.0253, "step": 25080 }, { "grad_norm": 0.2167121171951294, "learning_rate": 7.149206593404561e-06, "loss": 0.033, "step": 25090 }, { "grad_norm": 0.14522965252399445, "learning_rate": 7.120832083343337e-06, "loss": 0.0283, "step": 25100 }, { "grad_norm": 0.14494097232818604, "learning_rate": 7.092509675514369e-06, "loss": 0.0264, "step": 25110 }, { "grad_norm": 0.17240257561206818, "learning_rate": 7.064239404332063e-06, "loss": 0.0237, "step": 25120 }, { "grad_norm": 0.15423783659934998, "learning_rate": 7.03602130414745e-06, "loss": 0.0418, "step": 25130 }, { "grad_norm": 0.07620268315076828, "learning_rate": 7.007855409248198e-06, "loss": 0.0299, "step": 25140 }, { "grad_norm": 0.16667534410953522, "learning_rate": 6.979741753858521e-06, "loss": 0.0282, "step": 25150 }, { "grad_norm": 0.0652800127863884, "learning_rate": 6.951680372139158e-06, "loss": 0.0234, "step": 25160 }, { "grad_norm": 0.20300531387329102, "learning_rate": 6.923671298187334e-06, "loss": 0.0155, "step": 25170 }, { "grad_norm": 0.10094204545021057, "learning_rate": 6.895714566036704e-06, "loss": 0.0251, "step": 25180 }, { "grad_norm": 0.12505994737148285, "learning_rate": 6.86781020965736e-06, "loss": 0.0286, "step": 25190 }, { "grad_norm": 0.07362297922372818, "learning_rate": 6.839958262955709e-06, "loss": 0.0165, "step": 25200 }, { "grad_norm": 0.16232089698314667, "learning_rate": 6.8121587597744894e-06, "loss": 0.0269, "step": 25210 }, { "grad_norm": 0.10323423892259598, "learning_rate": 6.784411733892732e-06, "loss": 0.0209, "step": 25220 }, { "grad_norm": 0.17691636085510254, "learning_rate": 6.756717219025666e-06, "loss": 0.0272, "step": 25230 }, { "grad_norm": 0.1320650726556778, "learning_rate": 6.729075248824762e-06, "loss": 0.0214, "step": 25240 }, { "grad_norm": 0.14538827538490295, "learning_rate": 6.701485856877615e-06, "loss": 0.019, "step": 25250 }, { "grad_norm": 0.08397609740495682, "learning_rate": 6.673949076707925e-06, "loss": 0.0158, "step": 25260 }, { "grad_norm": 0.16260674595832825, "learning_rate": 6.646464941775499e-06, "loss": 0.0243, "step": 25270 }, { "grad_norm": 0.11454814672470093, "learning_rate": 6.619033485476128e-06, "loss": 0.0269, "step": 25280 }, { "grad_norm": 0.17012378573417664, "learning_rate": 6.59165474114164e-06, "loss": 0.0261, "step": 25290 }, { "grad_norm": 0.12865573167800903, "learning_rate": 6.564328742039783e-06, "loss": 0.0231, "step": 25300 }, { "grad_norm": 0.20164966583251953, "learning_rate": 6.537055521374219e-06, "loss": 0.0259, "step": 25310 }, { "grad_norm": 0.11626294255256653, "learning_rate": 6.509835112284485e-06, "loss": 0.0175, "step": 25320 }, { "grad_norm": 0.1868552565574646, "learning_rate": 6.482667547845944e-06, "loss": 0.0422, "step": 25330 }, { "grad_norm": 0.16526012122631073, "learning_rate": 6.455552861069736e-06, "loss": 0.0236, "step": 25340 }, { "grad_norm": 0.16755834221839905, "learning_rate": 6.428491084902788e-06, "loss": 0.0277, "step": 25350 }, { "grad_norm": 0.1524128019809723, "learning_rate": 6.401482252227698e-06, "loss": 0.0205, "step": 25360 }, { "grad_norm": 0.21186840534210205, "learning_rate": 6.37452639586274e-06, "loss": 0.0297, "step": 25370 }, { "grad_norm": 0.11582881212234497, "learning_rate": 6.347623548561826e-06, "loss": 0.0187, "step": 25380 }, { "grad_norm": 0.15068554878234863, "learning_rate": 6.320773743014441e-06, "loss": 0.0245, "step": 25390 }, { "grad_norm": 0.10572364926338196, "learning_rate": 6.293977011845648e-06, "loss": 0.0214, "step": 25400 }, { "grad_norm": 0.20672662556171417, "learning_rate": 6.2672333876159835e-06, "loss": 0.0323, "step": 25410 }, { "grad_norm": 0.15808624029159546, "learning_rate": 6.2405429028215e-06, "loss": 0.0213, "step": 25420 }, { "grad_norm": 0.2187930792570114, "learning_rate": 6.213905589893631e-06, "loss": 0.0402, "step": 25430 }, { "grad_norm": 0.1872318834066391, "learning_rate": 6.187321481199221e-06, "loss": 0.0248, "step": 25440 }, { "grad_norm": 0.13730093836784363, "learning_rate": 6.1607906090404895e-06, "loss": 0.0363, "step": 25450 }, { "grad_norm": 0.23654817044734955, "learning_rate": 6.134313005654929e-06, "loss": 0.0253, "step": 25460 }, { "grad_norm": 0.19498980045318604, "learning_rate": 6.107888703215337e-06, "loss": 0.0331, "step": 25470 }, { "grad_norm": 0.1664724200963974, "learning_rate": 6.081517733829722e-06, "loss": 0.0233, "step": 25480 }, { "grad_norm": 0.11545655876398087, "learning_rate": 6.055200129541294e-06, "loss": 0.0247, "step": 25490 }, { "grad_norm": 0.21779179573059082, "learning_rate": 6.0289359223284434e-06, "loss": 0.0305, "step": 25500 }, { "grad_norm": 0.11852090060710907, "learning_rate": 6.002725144104648e-06, "loss": 0.0284, "step": 25510 }, { "grad_norm": 0.17803584039211273, "learning_rate": 5.976567826718477e-06, "loss": 0.0228, "step": 25520 }, { "grad_norm": 0.19529466331005096, "learning_rate": 5.9504640019535315e-06, "loss": 0.0412, "step": 25530 }, { "grad_norm": 0.20274417102336884, "learning_rate": 5.924413701528414e-06, "loss": 0.0279, "step": 25540 }, { "grad_norm": 0.137828066945076, "learning_rate": 5.898416957096703e-06, "loss": 0.0282, "step": 25550 }, { "grad_norm": 0.22871646285057068, "learning_rate": 5.872473800246913e-06, "loss": 0.034, "step": 25560 }, { "grad_norm": 0.19272355735301971, "learning_rate": 5.846584262502402e-06, "loss": 0.033, "step": 25570 }, { "grad_norm": 0.11007756739854813, "learning_rate": 5.820748375321411e-06, "loss": 0.0348, "step": 25580 }, { "grad_norm": 0.13602502644062042, "learning_rate": 5.794966170096977e-06, "loss": 0.0348, "step": 25590 }, { "grad_norm": 0.14325231313705444, "learning_rate": 5.769237678156897e-06, "loss": 0.0253, "step": 25600 }, { "grad_norm": 0.13202248513698578, "learning_rate": 5.743562930763735e-06, "loss": 0.0309, "step": 25610 }, { "grad_norm": 0.18813487887382507, "learning_rate": 5.717941959114726e-06, "loss": 0.0344, "step": 25620 }, { "grad_norm": 0.1618967056274414, "learning_rate": 5.69237479434176e-06, "loss": 0.0353, "step": 25630 }, { "grad_norm": 0.16153164207935333, "learning_rate": 5.666861467511353e-06, "loss": 0.0235, "step": 25640 }, { "grad_norm": 0.1967756301164627, "learning_rate": 5.641402009624591e-06, "loss": 0.0335, "step": 25650 }, { "grad_norm": 0.1538332998752594, "learning_rate": 5.615996451617145e-06, "loss": 0.0189, "step": 25660 }, { "grad_norm": 0.15486447513103485, "learning_rate": 5.590644824359148e-06, "loss": 0.0485, "step": 25670 }, { "grad_norm": 0.15440884232521057, "learning_rate": 5.56534715865521e-06, "loss": 0.0223, "step": 25680 }, { "grad_norm": 0.13140802085399628, "learning_rate": 5.5401034852443965e-06, "loss": 0.0259, "step": 25690 }, { "grad_norm": 0.1872115582227707, "learning_rate": 5.514913834800134e-06, "loss": 0.0264, "step": 25700 }, { "grad_norm": 0.21282802522182465, "learning_rate": 5.489778237930238e-06, "loss": 0.0334, "step": 25710 }, { "grad_norm": 0.23591376841068268, "learning_rate": 5.464696725176827e-06, "loss": 0.0235, "step": 25720 }, { "grad_norm": 0.1985621452331543, "learning_rate": 5.439669327016294e-06, "loss": 0.0394, "step": 25730 }, { "grad_norm": 0.20875617861747742, "learning_rate": 5.4146960738592985e-06, "loss": 0.0269, "step": 25740 }, { "grad_norm": 0.19578462839126587, "learning_rate": 5.389776996050694e-06, "loss": 0.0267, "step": 25750 }, { "grad_norm": 0.16013099253177643, "learning_rate": 5.364912123869492e-06, "loss": 0.0322, "step": 25760 }, { "grad_norm": 0.14962057769298553, "learning_rate": 5.3401014875288864e-06, "loss": 0.0265, "step": 25770 }, { "grad_norm": 0.1571096032857895, "learning_rate": 5.315345117176129e-06, "loss": 0.0209, "step": 25780 }, { "grad_norm": 0.1340540051460266, "learning_rate": 5.2906430428925405e-06, "loss": 0.0285, "step": 25790 }, { "grad_norm": 0.15509329736232758, "learning_rate": 5.265995294693471e-06, "loss": 0.0283, "step": 25800 }, { "grad_norm": 0.18689042329788208, "learning_rate": 5.241401902528254e-06, "loss": 0.0233, "step": 25810 }, { "grad_norm": 0.11003687977790833, "learning_rate": 5.216862896280189e-06, "loss": 0.0219, "step": 25820 }, { "grad_norm": 0.1167622059583664, "learning_rate": 5.192378305766471e-06, "loss": 0.0242, "step": 25830 }, { "grad_norm": 0.11994116753339767, "learning_rate": 5.1679481607382065e-06, "loss": 0.0327, "step": 25840 }, { "grad_norm": 0.1138112023472786, "learning_rate": 5.1435724908803105e-06, "loss": 0.026, "step": 25850 }, { "grad_norm": 0.15986306965351105, "learning_rate": 5.119251325811514e-06, "loss": 0.0289, "step": 25860 }, { "grad_norm": 0.11195272952318192, "learning_rate": 5.094984695084348e-06, "loss": 0.0194, "step": 25870 }, { "grad_norm": 0.19876685738563538, "learning_rate": 5.070772628185039e-06, "loss": 0.0372, "step": 25880 }, { "grad_norm": 0.11463234573602676, "learning_rate": 5.046615154533535e-06, "loss": 0.0214, "step": 25890 }, { "grad_norm": 0.13338130712509155, "learning_rate": 5.022512303483451e-06, "loss": 0.0201, "step": 25900 }, { "grad_norm": 0.15808065235614777, "learning_rate": 4.998464104322015e-06, "loss": 0.0267, "step": 25910 }, { "grad_norm": 0.24102559685707092, "learning_rate": 4.974470586270047e-06, "loss": 0.0218, "step": 25920 }, { "grad_norm": 0.074319027364254, "learning_rate": 4.950531778481965e-06, "loss": 0.0176, "step": 25930 }, { "grad_norm": 0.1671372354030609, "learning_rate": 4.926647710045651e-06, "loss": 0.013, "step": 25940 }, { "grad_norm": 0.14793163537979126, "learning_rate": 4.9028184099825125e-06, "loss": 0.0174, "step": 25950 }, { "grad_norm": 0.0986865982413292, "learning_rate": 4.879043907247382e-06, "loss": 0.0172, "step": 25960 }, { "grad_norm": 0.14476938545703888, "learning_rate": 4.8553242307285414e-06, "loss": 0.0223, "step": 25970 }, { "grad_norm": 0.19566942751407623, "learning_rate": 4.8316594092476185e-06, "loss": 0.0203, "step": 25980 }, { "grad_norm": 0.09693784266710281, "learning_rate": 4.808049471559617e-06, "loss": 0.0224, "step": 25990 }, { "grad_norm": 0.17160534858703613, "learning_rate": 4.784494446352833e-06, "loss": 0.0187, "step": 26000 }, { "grad_norm": 0.15620408952236176, "learning_rate": 4.760994362248833e-06, "loss": 0.0221, "step": 26010 }, { "grad_norm": 0.1922338902950287, "learning_rate": 4.737549247802442e-06, "loss": 0.019, "step": 26020 }, { "grad_norm": 0.20088842511177063, "learning_rate": 4.714159131501689e-06, "loss": 0.0148, "step": 26030 }, { "grad_norm": 0.1840410977602005, "learning_rate": 4.690824041767766e-06, "loss": 0.0272, "step": 26040 }, { "grad_norm": 0.11229259520769119, "learning_rate": 4.6675440069550105e-06, "loss": 0.0166, "step": 26050 }, { "grad_norm": 0.1419122964143753, "learning_rate": 4.64431905535086e-06, "loss": 0.0167, "step": 26060 }, { "grad_norm": 0.08349046856164932, "learning_rate": 4.621149215175808e-06, "loss": 0.0219, "step": 26070 }, { "grad_norm": 0.22587457299232483, "learning_rate": 4.5980345145834154e-06, "loss": 0.0298, "step": 26080 }, { "grad_norm": 0.12168365716934204, "learning_rate": 4.574974981660213e-06, "loss": 0.0293, "step": 26090 }, { "grad_norm": 0.1566368192434311, "learning_rate": 4.551970644425707e-06, "loss": 0.0191, "step": 26100 }, { "grad_norm": 0.09941282868385315, "learning_rate": 4.529021530832328e-06, "loss": 0.0126, "step": 26110 }, { "grad_norm": 0.10314151644706726, "learning_rate": 4.5061276687654306e-06, "loss": 0.0254, "step": 26120 }, { "grad_norm": 0.13186399638652802, "learning_rate": 4.4832890860431965e-06, "loss": 0.0142, "step": 26130 }, { "grad_norm": 0.14384974539279938, "learning_rate": 4.460505810416682e-06, "loss": 0.0188, "step": 26140 }, { "grad_norm": 0.16996712982654572, "learning_rate": 4.4377778695696984e-06, "loss": 0.0228, "step": 26150 }, { "grad_norm": 0.06456530094146729, "learning_rate": 4.415105291118843e-06, "loss": 0.0128, "step": 26160 }, { "grad_norm": 0.1187678650021553, "learning_rate": 4.392488102613435e-06, "loss": 0.0221, "step": 26170 }, { "grad_norm": 0.18632887303829193, "learning_rate": 4.369926331535473e-06, "loss": 0.0169, "step": 26180 }, { "grad_norm": 0.15140432119369507, "learning_rate": 4.347420005299668e-06, "loss": 0.0232, "step": 26190 }, { "grad_norm": 0.11403393745422363, "learning_rate": 4.324969151253317e-06, "loss": 0.017, "step": 26200 }, { "grad_norm": 0.15788282454013824, "learning_rate": 4.302573796676312e-06, "loss": 0.0281, "step": 26210 }, { "grad_norm": 0.13541404902935028, "learning_rate": 4.2802339687811385e-06, "loss": 0.0237, "step": 26220 }, { "grad_norm": 0.1296081691980362, "learning_rate": 4.257949694712771e-06, "loss": 0.0228, "step": 26230 }, { "grad_norm": 0.11451195180416107, "learning_rate": 4.235721001548726e-06, "loss": 0.0102, "step": 26240 }, { "grad_norm": 0.10097282379865646, "learning_rate": 4.21354791629896e-06, "loss": 0.0233, "step": 26250 }, { "grad_norm": 0.14702783524990082, "learning_rate": 4.191430465905843e-06, "loss": 0.0291, "step": 26260 }, { "grad_norm": 0.1210961863398552, "learning_rate": 4.169368677244184e-06, "loss": 0.0148, "step": 26270 }, { "grad_norm": 0.1089329645037651, "learning_rate": 4.147362577121122e-06, "loss": 0.0337, "step": 26280 }, { "grad_norm": 0.07913164049386978, "learning_rate": 4.125412192276157e-06, "loss": 0.0248, "step": 26290 }, { "grad_norm": 0.1079515740275383, "learning_rate": 4.10351754938107e-06, "loss": 0.0283, "step": 26300 }, { "grad_norm": 0.09326153993606567, "learning_rate": 4.081678675039913e-06, "loss": 0.0358, "step": 26310 }, { "grad_norm": 0.12161669135093689, "learning_rate": 4.0598955957889876e-06, "loss": 0.0123, "step": 26320 }, { "grad_norm": 0.10358002036809921, "learning_rate": 4.038168338096776e-06, "loss": 0.0207, "step": 26330 }, { "grad_norm": 0.18642064929008484, "learning_rate": 4.016496928363944e-06, "loss": 0.0204, "step": 26340 }, { "grad_norm": 0.10812433809041977, "learning_rate": 3.994881392923317e-06, "loss": 0.0146, "step": 26350 }, { "grad_norm": 0.10043401271104813, "learning_rate": 3.9733217580397945e-06, "loss": 0.02, "step": 26360 }, { "grad_norm": 0.12395023554563522, "learning_rate": 3.951818049910366e-06, "loss": 0.0136, "step": 26370 }, { "grad_norm": 0.10807546228170395, "learning_rate": 3.930370294664071e-06, "loss": 0.0191, "step": 26380 }, { "grad_norm": 0.09653477370738983, "learning_rate": 3.908978518361938e-06, "loss": 0.0161, "step": 26390 }, { "grad_norm": 0.1807551085948944, "learning_rate": 3.887642746997017e-06, "loss": 0.0142, "step": 26400 }, { "grad_norm": 0.062499865889549255, "learning_rate": 3.866363006494255e-06, "loss": 0.0169, "step": 26410 }, { "grad_norm": 0.15637126564979553, "learning_rate": 3.8451393227105725e-06, "loss": 0.0146, "step": 26420 }, { "grad_norm": 0.09493592381477356, "learning_rate": 3.823971721434727e-06, "loss": 0.0175, "step": 26430 }, { "grad_norm": 0.09523970633745193, "learning_rate": 3.8028602283873505e-06, "loss": 0.0254, "step": 26440 }, { "grad_norm": 0.1386086791753769, "learning_rate": 3.781804869220912e-06, "loss": 0.0181, "step": 26450 }, { "grad_norm": 0.12187820672988892, "learning_rate": 3.760805669519646e-06, "loss": 0.0202, "step": 26460 }, { "grad_norm": 0.10179018974304199, "learning_rate": 3.7398626547995587e-06, "loss": 0.0117, "step": 26470 }, { "grad_norm": 0.1311856508255005, "learning_rate": 3.7189758505083995e-06, "loss": 0.0184, "step": 26480 }, { "grad_norm": 0.14044009149074554, "learning_rate": 3.6981452820255837e-06, "loss": 0.0134, "step": 26490 }, { "grad_norm": 0.12106408178806305, "learning_rate": 3.6773709746622363e-06, "loss": 0.0151, "step": 26500 }, { "grad_norm": 0.09205186367034912, "learning_rate": 3.6566529536610995e-06, "loss": 0.0261, "step": 26510 }, { "grad_norm": 0.12186262011528015, "learning_rate": 3.635991244196513e-06, "loss": 0.0172, "step": 26520 }, { "grad_norm": 0.10872261971235275, "learning_rate": 3.615385871374405e-06, "loss": 0.0253, "step": 26530 }, { "grad_norm": 0.1254366636276245, "learning_rate": 3.5948368602322367e-06, "loss": 0.0215, "step": 26540 }, { "grad_norm": 0.2288755476474762, "learning_rate": 3.5743442357390187e-06, "loss": 0.023, "step": 26550 }, { "grad_norm": 0.1510211080312729, "learning_rate": 3.5539080227951937e-06, "loss": 0.0239, "step": 26560 }, { "grad_norm": 0.13619869947433472, "learning_rate": 3.533528246232709e-06, "loss": 0.0147, "step": 26570 }, { "grad_norm": 0.14232537150382996, "learning_rate": 3.5132049308149117e-06, "loss": 0.0178, "step": 26580 }, { "grad_norm": 0.2194085419178009, "learning_rate": 3.4929381012365425e-06, "loss": 0.0262, "step": 26590 }, { "grad_norm": 0.13797703385353088, "learning_rate": 3.472727782123697e-06, "loss": 0.0223, "step": 26600 }, { "grad_norm": 0.11913114786148071, "learning_rate": 3.4525739980338425e-06, "loss": 0.0175, "step": 26610 }, { "grad_norm": 0.07425662130117416, "learning_rate": 3.432476773455712e-06, "loss": 0.0195, "step": 26620 }, { "grad_norm": 0.16549770534038544, "learning_rate": 3.4124361328093378e-06, "loss": 0.0254, "step": 26630 }, { "grad_norm": 0.1106576919555664, "learning_rate": 3.392452100445975e-06, "loss": 0.0191, "step": 26640 }, { "grad_norm": 0.14090275764465332, "learning_rate": 3.3725247006481163e-06, "loss": 0.0176, "step": 26650 }, { "grad_norm": 0.15371257066726685, "learning_rate": 3.3526539576294313e-06, "loss": 0.0325, "step": 26660 }, { "grad_norm": 0.1316637396812439, "learning_rate": 3.332839895534745e-06, "loss": 0.0202, "step": 26670 }, { "grad_norm": 0.18366529047489166, "learning_rate": 3.3130825384400155e-06, "loss": 0.0253, "step": 26680 }, { "grad_norm": 0.16447412967681885, "learning_rate": 3.293381910352278e-06, "loss": 0.0285, "step": 26690 }, { "grad_norm": 0.21193091571331024, "learning_rate": 3.273738035209678e-06, "loss": 0.0319, "step": 26700 }, { "grad_norm": 0.22279618680477142, "learning_rate": 3.254150936881356e-06, "loss": 0.0324, "step": 26710 }, { "grad_norm": 0.09389699995517731, "learning_rate": 3.234620639167496e-06, "loss": 0.0135, "step": 26720 }, { "grad_norm": 0.1244015097618103, "learning_rate": 3.2151471657992482e-06, "loss": 0.0195, "step": 26730 }, { "grad_norm": 0.18860147893428802, "learning_rate": 3.195730540438718e-06, "loss": 0.019, "step": 26740 }, { "grad_norm": 0.14387300610542297, "learning_rate": 3.1763707866789336e-06, "loss": 0.0247, "step": 26750 }, { "grad_norm": 0.17068225145339966, "learning_rate": 3.1570679280438165e-06, "loss": 0.025, "step": 26760 }, { "grad_norm": 0.1433863341808319, "learning_rate": 3.1378219879881767e-06, "loss": 0.0186, "step": 26770 }, { "grad_norm": 0.09929054230451584, "learning_rate": 3.11863298989763e-06, "loss": 0.0175, "step": 26780 }, { "grad_norm": 0.23583965003490448, "learning_rate": 3.0995009570886303e-06, "loss": 0.0246, "step": 26790 }, { "grad_norm": 0.11341741681098938, "learning_rate": 3.0804259128083936e-06, "loss": 0.0188, "step": 26800 }, { "grad_norm": 0.11855362355709076, "learning_rate": 3.06140788023489e-06, "loss": 0.0211, "step": 26810 }, { "grad_norm": 0.1380843073129654, "learning_rate": 3.042446882476846e-06, "loss": 0.0181, "step": 26820 }, { "grad_norm": 0.13050150871276855, "learning_rate": 3.023542942573643e-06, "loss": 0.023, "step": 26830 }, { "grad_norm": 0.15919415652751923, "learning_rate": 3.004696083495351e-06, "loss": 0.0311, "step": 26840 }, { "grad_norm": 0.21425464749336243, "learning_rate": 2.98590632814269e-06, "loss": 0.0237, "step": 26850 }, { "grad_norm": 0.09748256206512451, "learning_rate": 2.9671736993469702e-06, "loss": 0.0209, "step": 26860 }, { "grad_norm": 0.25342556834220886, "learning_rate": 2.9484982198701216e-06, "loss": 0.0226, "step": 26870 }, { "grad_norm": 0.1645422726869583, "learning_rate": 2.929879912404604e-06, "loss": 0.0269, "step": 26880 }, { "grad_norm": 0.11314207315444946, "learning_rate": 2.9113187995734147e-06, "loss": 0.0167, "step": 26890 }, { "grad_norm": 0.14032837748527527, "learning_rate": 2.8928149039300524e-06, "loss": 0.0207, "step": 26900 }, { "grad_norm": 0.11793795973062515, "learning_rate": 2.8743682479584975e-06, "loss": 0.0193, "step": 26910 }, { "grad_norm": 0.1730591505765915, "learning_rate": 2.8559788540731824e-06, "loss": 0.0202, "step": 26920 }, { "grad_norm": 0.19654609262943268, "learning_rate": 2.8376467446189493e-06, "loss": 0.0215, "step": 26930 }, { "grad_norm": 0.2525007426738739, "learning_rate": 2.8193719418710406e-06, "loss": 0.0295, "step": 26940 }, { "grad_norm": 0.13693009316921234, "learning_rate": 2.8011544680350667e-06, "loss": 0.0275, "step": 26950 }, { "grad_norm": 0.10682759433984756, "learning_rate": 2.7829943452469753e-06, "loss": 0.0134, "step": 26960 }, { "grad_norm": 0.13979457318782806, "learning_rate": 2.7648915955730213e-06, "loss": 0.0205, "step": 26970 }, { "grad_norm": 0.10048786550760269, "learning_rate": 2.746846241009765e-06, "loss": 0.014, "step": 26980 }, { "grad_norm": 0.17828251421451569, "learning_rate": 2.7288583034839945e-06, "loss": 0.0219, "step": 26990 }, { "grad_norm": 0.1249995231628418, "learning_rate": 2.7109278048527752e-06, "loss": 0.0148, "step": 27000 }, { "grad_norm": 0.12605221569538116, "learning_rate": 2.6930547669033413e-06, "loss": 0.0138, "step": 27010 }, { "grad_norm": 0.13817806541919708, "learning_rate": 2.675239211353109e-06, "loss": 0.016, "step": 27020 }, { "grad_norm": 0.1635843813419342, "learning_rate": 2.6574811598496786e-06, "loss": 0.0106, "step": 27030 }, { "grad_norm": 0.07605167478322983, "learning_rate": 2.6397806339707456e-06, "loss": 0.0269, "step": 27040 }, { "grad_norm": 0.16851548850536346, "learning_rate": 2.6221376552241217e-06, "loss": 0.0117, "step": 27050 }, { "grad_norm": 0.08505366742610931, "learning_rate": 2.604552245047681e-06, "loss": 0.0166, "step": 27060 }, { "grad_norm": 0.11488756537437439, "learning_rate": 2.587024424809359e-06, "loss": 0.0173, "step": 27070 }, { "grad_norm": 0.08520681411027908, "learning_rate": 2.5695542158071186e-06, "loss": 0.0202, "step": 27080 }, { "grad_norm": 0.12186137586832047, "learning_rate": 2.5521416392689068e-06, "loss": 0.0156, "step": 27090 }, { "grad_norm": 0.12466062605381012, "learning_rate": 2.5347867163526384e-06, "loss": 0.0128, "step": 27100 }, { "grad_norm": 0.22704464197158813, "learning_rate": 2.517489468146189e-06, "loss": 0.0214, "step": 27110 }, { "grad_norm": 0.15954899787902832, "learning_rate": 2.500249915667341e-06, "loss": 0.0194, "step": 27120 }, { "grad_norm": 0.13348907232284546, "learning_rate": 2.4830680798637817e-06, "loss": 0.0159, "step": 27130 }, { "grad_norm": 0.13258741796016693, "learning_rate": 2.465943981613056e-06, "loss": 0.0308, "step": 27140 }, { "grad_norm": 0.15943306684494019, "learning_rate": 2.448877641722569e-06, "loss": 0.0165, "step": 27150 }, { "grad_norm": 0.09438923001289368, "learning_rate": 2.431869080929522e-06, "loss": 0.0176, "step": 27160 }, { "grad_norm": 0.17481988668441772, "learning_rate": 2.4149183199009216e-06, "loss": 0.0234, "step": 27170 }, { "grad_norm": 0.25590917468070984, "learning_rate": 2.3980253792335427e-06, "loss": 0.0244, "step": 27180 }, { "grad_norm": 0.14255087077617645, "learning_rate": 2.381190279453899e-06, "loss": 0.0253, "step": 27190 }, { "grad_norm": 0.10709479451179504, "learning_rate": 2.364413041018232e-06, "loss": 0.0354, "step": 27200 }, { "grad_norm": 0.1349218338727951, "learning_rate": 2.347693684312463e-06, "loss": 0.0211, "step": 27210 }, { "grad_norm": 0.10535775125026703, "learning_rate": 2.331032229652186e-06, "loss": 0.0146, "step": 27220 }, { "grad_norm": 0.1723625510931015, "learning_rate": 2.314428697282628e-06, "loss": 0.0152, "step": 27230 }, { "grad_norm": 0.10169080644845963, "learning_rate": 2.297883107378673e-06, "loss": 0.0181, "step": 27240 }, { "grad_norm": 0.17933490872383118, "learning_rate": 2.2813954800447513e-06, "loss": 0.0252, "step": 27250 }, { "grad_norm": 0.1021367609500885, "learning_rate": 2.2649658353148974e-06, "loss": 0.0127, "step": 27260 }, { "grad_norm": 0.12624342739582062, "learning_rate": 2.2485941931526645e-06, "loss": 0.0136, "step": 27270 }, { "grad_norm": 0.17649374902248383, "learning_rate": 2.232280573451151e-06, "loss": 0.0219, "step": 27280 }, { "grad_norm": 0.14308635890483856, "learning_rate": 2.216024996032945e-06, "loss": 0.016, "step": 27290 }, { "grad_norm": 0.11447527259588242, "learning_rate": 2.1998274806501074e-06, "loss": 0.0162, "step": 27300 }, { "grad_norm": 0.1462414711713791, "learning_rate": 2.183688046984139e-06, "loss": 0.0175, "step": 27310 }, { "grad_norm": 0.15251697599887848, "learning_rate": 2.167606714645981e-06, "loss": 0.0193, "step": 27320 }, { "grad_norm": 0.12472930550575256, "learning_rate": 2.151583503175958e-06, "loss": 0.0108, "step": 27330 }, { "grad_norm": 0.08705972880125046, "learning_rate": 2.1356184320437957e-06, "loss": 0.0174, "step": 27340 }, { "grad_norm": 0.09962349385023117, "learning_rate": 2.119711520648554e-06, "loss": 0.0159, "step": 27350 }, { "grad_norm": 0.12157399207353592, "learning_rate": 2.103862788318628e-06, "loss": 0.0236, "step": 27360 }, { "grad_norm": 0.05131950601935387, "learning_rate": 2.088072254311729e-06, "loss": 0.0168, "step": 27370 }, { "grad_norm": 0.1433200091123581, "learning_rate": 2.0723399378148435e-06, "loss": 0.0214, "step": 27380 }, { "grad_norm": 0.09750042110681534, "learning_rate": 2.0566658579442065e-06, "loss": 0.0167, "step": 27390 }, { "grad_norm": 0.14732426404953003, "learning_rate": 2.0410500337453176e-06, "loss": 0.0155, "step": 27400 }, { "grad_norm": 0.05296378210186958, "learning_rate": 2.0254924841928645e-06, "loss": 0.0297, "step": 27410 }, { "grad_norm": 0.09149641543626785, "learning_rate": 2.009993228190754e-06, "loss": 0.0205, "step": 27420 }, { "grad_norm": 0.1723482310771942, "learning_rate": 1.9945522845720323e-06, "loss": 0.0237, "step": 27430 }, { "grad_norm": 0.1988799124956131, "learning_rate": 1.9791696720988963e-06, "loss": 0.0317, "step": 27440 }, { "grad_norm": 0.15035851299762726, "learning_rate": 1.9638454094626833e-06, "loss": 0.022, "step": 27450 }, { "grad_norm": 0.06745445728302002, "learning_rate": 1.9485795152838104e-06, "loss": 0.016, "step": 27460 }, { "grad_norm": 0.16435739398002625, "learning_rate": 1.933372008111778e-06, "loss": 0.0344, "step": 27470 }, { "grad_norm": 0.18015924096107483, "learning_rate": 1.918222906425143e-06, "loss": 0.0288, "step": 27480 }, { "grad_norm": 0.1312905251979828, "learning_rate": 1.903132228631488e-06, "loss": 0.026, "step": 27490 }, { "grad_norm": 0.203019380569458, "learning_rate": 1.8880999930674215e-06, "loss": 0.0326, "step": 27500 }, { "grad_norm": 0.2000075727701187, "learning_rate": 1.8731262179985166e-06, "loss": 0.0331, "step": 27510 }, { "grad_norm": 0.15054650604724884, "learning_rate": 1.8582109216193244e-06, "loss": 0.0197, "step": 27520 }, { "grad_norm": 0.0818881019949913, "learning_rate": 1.8433541220533368e-06, "loss": 0.0309, "step": 27530 }, { "grad_norm": 0.17579896748065948, "learning_rate": 1.8285558373529577e-06, "loss": 0.0248, "step": 27540 }, { "grad_norm": 0.10682690143585205, "learning_rate": 1.8138160854995145e-06, "loss": 0.031, "step": 27550 }, { "grad_norm": 0.09102911502122879, "learning_rate": 1.7991348844031864e-06, "loss": 0.0191, "step": 27560 }, { "grad_norm": 0.14145421981811523, "learning_rate": 1.78451225190302e-06, "loss": 0.0179, "step": 27570 }, { "grad_norm": 0.14105238020420074, "learning_rate": 1.7699482057668914e-06, "loss": 0.023, "step": 27580 }, { "grad_norm": 0.0860939472913742, "learning_rate": 1.7554427636914895e-06, "loss": 0.0217, "step": 27590 }, { "grad_norm": 0.16897229850292206, "learning_rate": 1.7409959433022871e-06, "loss": 0.0345, "step": 27600 }, { "grad_norm": 0.126984104514122, "learning_rate": 1.7266077621535481e-06, "loss": 0.0375, "step": 27610 }, { "grad_norm": 0.17528283596038818, "learning_rate": 1.7122782377282597e-06, "loss": 0.0268, "step": 27620 }, { "grad_norm": 0.08236650377511978, "learning_rate": 1.6980073874381497e-06, "loss": 0.0254, "step": 27630 }, { "grad_norm": 0.16962243616580963, "learning_rate": 1.6837952286236413e-06, "loss": 0.0253, "step": 27640 }, { "grad_norm": 0.2031347155570984, "learning_rate": 1.6696417785538488e-06, "loss": 0.0357, "step": 27650 }, { "grad_norm": 0.15325340628623962, "learning_rate": 1.6555470544265538e-06, "loss": 0.037, "step": 27660 }, { "grad_norm": 0.11123790591955185, "learning_rate": 1.6415110733681738e-06, "loss": 0.0371, "step": 27670 }, { "grad_norm": 0.19162175059318542, "learning_rate": 1.6275338524337436e-06, "loss": 0.0292, "step": 27680 }, { "grad_norm": 0.15148788690567017, "learning_rate": 1.6136154086069056e-06, "loss": 0.0215, "step": 27690 }, { "grad_norm": 0.20379361510276794, "learning_rate": 1.5997557587998868e-06, "loss": 0.0337, "step": 27700 }, { "grad_norm": 0.16033822298049927, "learning_rate": 1.5859549198534606e-06, "loss": 0.0202, "step": 27710 }, { "grad_norm": 0.10424648225307465, "learning_rate": 1.572212908536963e-06, "loss": 0.0236, "step": 27720 }, { "grad_norm": 0.13254457712173462, "learning_rate": 1.5585297415482202e-06, "loss": 0.0182, "step": 27730 }, { "grad_norm": 0.09357276558876038, "learning_rate": 1.5449054355135717e-06, "loss": 0.0255, "step": 27740 }, { "grad_norm": 0.1415044516324997, "learning_rate": 1.5313400069878415e-06, "loss": 0.0222, "step": 27750 }, { "grad_norm": 0.1691199690103531, "learning_rate": 1.5178334724542887e-06, "loss": 0.0226, "step": 27760 }, { "grad_norm": 0.15877622365951538, "learning_rate": 1.504385848324641e-06, "loss": 0.0429, "step": 27770 }, { "grad_norm": 0.16009078919887543, "learning_rate": 1.490997150939033e-06, "loss": 0.0249, "step": 27780 }, { "grad_norm": 0.18990758061408997, "learning_rate": 1.4776673965659793e-06, "loss": 0.0341, "step": 27790 }, { "grad_norm": 0.17218820750713348, "learning_rate": 1.4643966014023957e-06, "loss": 0.0264, "step": 27800 }, { "grad_norm": 0.19476017355918884, "learning_rate": 1.4511847815735503e-06, "loss": 0.0369, "step": 27810 }, { "grad_norm": 0.12561151385307312, "learning_rate": 1.438031953133051e-06, "loss": 0.0215, "step": 27820 }, { "grad_norm": 0.10154806822538376, "learning_rate": 1.4249381320628197e-06, "loss": 0.0249, "step": 27830 }, { "grad_norm": 0.18672963976860046, "learning_rate": 1.4119033342730902e-06, "loss": 0.0338, "step": 27840 }, { "grad_norm": 0.04672496020793915, "learning_rate": 1.3989275756023656e-06, "loss": 0.0179, "step": 27850 }, { "grad_norm": 0.19500946998596191, "learning_rate": 1.386010871817417e-06, "loss": 0.0224, "step": 27860 }, { "grad_norm": 0.1297519952058792, "learning_rate": 1.3731532386132616e-06, "loss": 0.0177, "step": 27870 }, { "grad_norm": 0.20687516033649445, "learning_rate": 1.360354691613136e-06, "loss": 0.0389, "step": 27880 }, { "grad_norm": 0.07852189242839813, "learning_rate": 1.3476152463684776e-06, "loss": 0.0246, "step": 27890 }, { "grad_norm": 0.12094691395759583, "learning_rate": 1.3349349183589155e-06, "loss": 0.0225, "step": 27900 }, { "grad_norm": 0.14689454436302185, "learning_rate": 1.3223137229922356e-06, "loss": 0.0275, "step": 27910 }, { "grad_norm": 0.18429847061634064, "learning_rate": 1.3097516756043981e-06, "loss": 0.0238, "step": 27920 }, { "grad_norm": 0.15424267947673798, "learning_rate": 1.29724879145946e-06, "loss": 0.0268, "step": 27930 }, { "grad_norm": 0.1206110343337059, "learning_rate": 1.284805085749613e-06, "loss": 0.0228, "step": 27940 }, { "grad_norm": 0.16988010704517365, "learning_rate": 1.2724205735951288e-06, "loss": 0.0239, "step": 27950 }, { "grad_norm": 0.14286041259765625, "learning_rate": 1.2600952700443591e-06, "loss": 0.0359, "step": 27960 }, { "grad_norm": 0.10331439226865768, "learning_rate": 1.247829190073707e-06, "loss": 0.0213, "step": 27970 }, { "grad_norm": 0.1532374918460846, "learning_rate": 1.2356223485876173e-06, "loss": 0.0169, "step": 27980 }, { "grad_norm": 0.15138868987560272, "learning_rate": 1.2234747604185526e-06, "loss": 0.026, "step": 27990 }, { "grad_norm": 0.28238216042518616, "learning_rate": 1.2113864403269836e-06, "loss": 0.0356, "step": 28000 }, { "grad_norm": 0.1706603318452835, "learning_rate": 1.1993574030013555e-06, "loss": 0.0225, "step": 28010 }, { "grad_norm": 0.21431490778923035, "learning_rate": 1.1873876630580816e-06, "loss": 0.0408, "step": 28020 }, { "grad_norm": 0.20173393189907074, "learning_rate": 1.1754772350415278e-06, "loss": 0.0301, "step": 28030 }, { "grad_norm": 0.12692463397979736, "learning_rate": 1.163626133423984e-06, "loss": 0.0305, "step": 28040 }, { "grad_norm": 0.08535335958003998, "learning_rate": 1.1518343726056591e-06, "loss": 0.026, "step": 28050 }, { "grad_norm": 0.09239307045936584, "learning_rate": 1.1401019669146474e-06, "loss": 0.0233, "step": 28060 }, { "grad_norm": 0.1548362821340561, "learning_rate": 1.128428930606934e-06, "loss": 0.026, "step": 28070 }, { "grad_norm": 0.14434678852558136, "learning_rate": 1.1168152778663621e-06, "loss": 0.0229, "step": 28080 }, { "grad_norm": 0.12999895215034485, "learning_rate": 1.10526102280461e-06, "loss": 0.0297, "step": 28090 }, { "grad_norm": 0.13342730700969696, "learning_rate": 1.0937661794611864e-06, "loss": 0.0262, "step": 28100 }, { "grad_norm": 0.1879059076309204, "learning_rate": 1.082330761803413e-06, "loss": 0.0335, "step": 28110 }, { "grad_norm": 0.1983952820301056, "learning_rate": 1.0709547837263966e-06, "loss": 0.0369, "step": 28120 }, { "grad_norm": 0.13074657320976257, "learning_rate": 1.059638259053025e-06, "loss": 0.0233, "step": 28130 }, { "grad_norm": 0.20608630776405334, "learning_rate": 1.0483812015339434e-06, "loss": 0.0268, "step": 28140 }, { "grad_norm": 0.26983749866485596, "learning_rate": 1.037183624847543e-06, "loss": 0.0507, "step": 28150 }, { "grad_norm": 0.13587841391563416, "learning_rate": 1.0260455425999238e-06, "loss": 0.0352, "step": 28160 }, { "grad_norm": 0.14419224858283997, "learning_rate": 1.0149669683249208e-06, "loss": 0.0266, "step": 28170 }, { "grad_norm": 0.19252878427505493, "learning_rate": 1.003947915484027e-06, "loss": 0.0372, "step": 28180 }, { "grad_norm": 0.1690388172864914, "learning_rate": 9.92988397466449e-07, "loss": 0.0446, "step": 28190 }, { "grad_norm": 0.2574954330921173, "learning_rate": 9.820884275890286e-07, "loss": 0.0391, "step": 28200 }, { "grad_norm": 0.1735139638185501, "learning_rate": 9.712480190962546e-07, "loss": 0.0277, "step": 28210 }, { "grad_norm": 0.15264476835727692, "learning_rate": 9.604671851602464e-07, "loss": 0.0264, "step": 28220 }, { "grad_norm": 0.2571484446525574, "learning_rate": 9.497459388807306e-07, "loss": 0.0321, "step": 28230 }, { "grad_norm": 0.1223103329539299, "learning_rate": 9.390842932850364e-07, "loss": 0.0218, "step": 28240 }, { "grad_norm": 0.20435482263565063, "learning_rate": 9.28482261328073e-07, "loss": 0.0452, "step": 28250 }, { "grad_norm": 0.20132353901863098, "learning_rate": 9.179398558923025e-07, "loss": 0.0342, "step": 28260 }, { "grad_norm": 0.143037348985672, "learning_rate": 9.074570897877388e-07, "loss": 0.021, "step": 28270 }, { "grad_norm": 0.15457434952259064, "learning_rate": 8.970339757519375e-07, "loss": 0.0295, "step": 28280 }, { "grad_norm": 0.10197629779577255, "learning_rate": 8.866705264499619e-07, "loss": 0.0217, "step": 28290 }, { "grad_norm": 0.1186271458864212, "learning_rate": 8.763667544743836e-07, "loss": 0.0235, "step": 28300 }, { "grad_norm": 0.16419963538646698, "learning_rate": 8.661226723452543e-07, "loss": 0.0272, "step": 28310 }, { "grad_norm": 0.13693001866340637, "learning_rate": 8.559382925101001e-07, "loss": 0.0253, "step": 28320 }, { "grad_norm": 0.11380830407142639, "learning_rate": 8.458136273438943e-07, "loss": 0.0324, "step": 28330 }, { "grad_norm": 0.1717309057712555, "learning_rate": 8.357486891490795e-07, "loss": 0.0323, "step": 28340 }, { "grad_norm": 0.16540224850177765, "learning_rate": 8.257434901554895e-07, "loss": 0.018, "step": 28350 }, { "grad_norm": 0.14530305564403534, "learning_rate": 8.157980425203937e-07, "loss": 0.0209, "step": 28360 }, { "grad_norm": 0.15041691064834595, "learning_rate": 8.059123583284367e-07, "loss": 0.0301, "step": 28370 }, { "grad_norm": 0.12378647178411484, "learning_rate": 7.960864495916654e-07, "loss": 0.0261, "step": 28380 }, { "grad_norm": 0.11598815023899078, "learning_rate": 7.863203282494847e-07, "loss": 0.0279, "step": 28390 }, { "grad_norm": 0.18787126243114471, "learning_rate": 7.766140061686522e-07, "loss": 0.02, "step": 28400 }, { "grad_norm": 0.21308490633964539, "learning_rate": 7.669674951432615e-07, "loss": 0.0255, "step": 28410 }, { "grad_norm": 0.15801121294498444, "learning_rate": 7.573808068947363e-07, "loss": 0.0316, "step": 28420 }, { "grad_norm": 0.23403377830982208, "learning_rate": 7.478539530718087e-07, "loss": 0.0259, "step": 28430 }, { "grad_norm": 0.15341304242610931, "learning_rate": 7.383869452504965e-07, "loss": 0.0209, "step": 28440 }, { "grad_norm": 0.11308202147483826, "learning_rate": 7.289797949341204e-07, "loss": 0.0194, "step": 28450 }, { "grad_norm": 0.1578955501317978, "learning_rate": 7.196325135532423e-07, "loss": 0.0243, "step": 28460 }, { "grad_norm": 0.19167807698249817, "learning_rate": 7.10345112465699e-07, "loss": 0.0192, "step": 28470 }, { "grad_norm": 0.21326188743114471, "learning_rate": 7.011176029565525e-07, "loss": 0.0328, "step": 28480 }, { "grad_norm": 0.1551733762025833, "learning_rate": 6.919499962381004e-07, "loss": 0.0138, "step": 28490 }, { "grad_norm": 0.16603098809719086, "learning_rate": 6.828423034498488e-07, "loss": 0.0163, "step": 28500 }, { "grad_norm": 0.16482019424438477, "learning_rate": 6.737945356585007e-07, "loss": 0.0234, "step": 28510 }, { "grad_norm": 0.13511459529399872, "learning_rate": 6.648067038579508e-07, "loss": 0.0343, "step": 28520 }, { "grad_norm": 0.22677332162857056, "learning_rate": 6.558788189692578e-07, "loss": 0.0305, "step": 28530 }, { "grad_norm": 0.17296789586544037, "learning_rate": 6.470108918406492e-07, "loss": 0.0188, "step": 28540 }, { "grad_norm": 0.14698590338230133, "learning_rate": 6.382029332474893e-07, "loss": 0.0174, "step": 28550 }, { "grad_norm": 0.17462453246116638, "learning_rate": 6.294549538922778e-07, "loss": 0.0245, "step": 28560 }, { "grad_norm": 0.1297178566455841, "learning_rate": 6.207669644046344e-07, "loss": 0.0347, "step": 28570 }, { "grad_norm": 0.19662632048130035, "learning_rate": 6.121389753412865e-07, "loss": 0.0281, "step": 28580 }, { "grad_norm": 0.11333277076482773, "learning_rate": 6.035709971860592e-07, "loss": 0.0204, "step": 28590 }, { "grad_norm": 0.14027443528175354, "learning_rate": 5.950630403498469e-07, "loss": 0.0168, "step": 28600 }, { "grad_norm": 0.14043889939785004, "learning_rate": 5.866151151706189e-07, "loss": 0.0289, "step": 28610 }, { "grad_norm": 0.151023730635643, "learning_rate": 5.782272319134086e-07, "loss": 0.03, "step": 28620 }, { "grad_norm": 0.07884180545806885, "learning_rate": 5.698994007702796e-07, "loss": 0.0163, "step": 28630 }, { "grad_norm": 0.0796985775232315, "learning_rate": 5.616316318603321e-07, "loss": 0.0234, "step": 28640 }, { "grad_norm": 0.12429188936948776, "learning_rate": 5.534239352296799e-07, "loss": 0.023, "step": 28650 }, { "grad_norm": 0.20143143832683563, "learning_rate": 5.452763208514621e-07, "loss": 0.0271, "step": 28660 }, { "grad_norm": 0.16091272234916687, "learning_rate": 5.371887986257873e-07, "loss": 0.0156, "step": 28670 }, { "grad_norm": 0.15533478558063507, "learning_rate": 5.291613783797611e-07, "loss": 0.0292, "step": 28680 }, { "grad_norm": 0.16535401344299316, "learning_rate": 5.211940698674534e-07, "loss": 0.0267, "step": 28690 }, { "grad_norm": 0.17011450231075287, "learning_rate": 5.132868827698978e-07, "loss": 0.0224, "step": 28700 }, { "grad_norm": 0.1796565055847168, "learning_rate": 5.054398266950755e-07, "loss": 0.0171, "step": 28710 }, { "grad_norm": 0.12735994160175323, "learning_rate": 4.976529111778872e-07, "loss": 0.0251, "step": 28720 }, { "grad_norm": 0.12516257166862488, "learning_rate": 4.899261456801862e-07, "loss": 0.0421, "step": 28730 }, { "grad_norm": 0.1414061039686203, "learning_rate": 4.822595395907126e-07, "loss": 0.0143, "step": 28740 }, { "grad_norm": 0.08848924189805984, "learning_rate": 4.7465310222510886e-07, "loss": 0.0206, "step": 28750 }, { "grad_norm": 0.1151849702000618, "learning_rate": 4.6710684282593175e-07, "loss": 0.0285, "step": 28760 }, { "grad_norm": 0.12407878786325455, "learning_rate": 4.5962077056257993e-07, "loss": 0.0175, "step": 28770 }, { "grad_norm": 0.1688760221004486, "learning_rate": 4.521948945313492e-07, "loss": 0.0219, "step": 28780 }, { "grad_norm": 0.09942125529050827, "learning_rate": 4.4482922375537195e-07, "loss": 0.0212, "step": 28790 }, { "grad_norm": 0.10096052289009094, "learning_rate": 4.375237671846333e-07, "loss": 0.0153, "step": 28800 }, { "grad_norm": 0.08416605740785599, "learning_rate": 4.302785336959547e-07, "loss": 0.014, "step": 28810 }, { "grad_norm": 0.095969557762146, "learning_rate": 4.2309353209297744e-07, "loss": 0.0205, "step": 28820 }, { "grad_norm": 0.08167442679405212, "learning_rate": 4.159687711061566e-07, "loss": 0.0119, "step": 28830 }, { "grad_norm": 0.19282443821430206, "learning_rate": 4.0890425939275055e-07, "loss": 0.0161, "step": 28840 }, { "grad_norm": 0.2184470295906067, "learning_rate": 4.0190000553679827e-07, "loss": 0.0289, "step": 28850 }, { "grad_norm": 0.15572667121887207, "learning_rate": 3.9495601804913627e-07, "loss": 0.0178, "step": 28860 }, { "grad_norm": 0.22656799852848053, "learning_rate": 3.880723053673652e-07, "loss": 0.0257, "step": 28870 }, { "grad_norm": 0.10000621527433395, "learning_rate": 3.812488758558386e-07, "loss": 0.0274, "step": 28880 }, { "grad_norm": 0.22154702246189117, "learning_rate": 3.744857378056743e-07, "loss": 0.0146, "step": 28890 }, { "grad_norm": 0.15876848995685577, "learning_rate": 3.677828994347154e-07, "loss": 0.0154, "step": 28900 }, { "grad_norm": 0.07978667318820953, "learning_rate": 3.61140368887547e-07, "loss": 0.0223, "step": 28910 }, { "grad_norm": 0.12790051102638245, "learning_rate": 3.545581542354681e-07, "loss": 0.0118, "step": 28920 }, { "grad_norm": 0.09665203839540482, "learning_rate": 3.480362634764922e-07, "loss": 0.014, "step": 28930 }, { "grad_norm": 0.19523866474628448, "learning_rate": 3.4157470453533014e-07, "loss": 0.0191, "step": 28940 }, { "grad_norm": 0.07229378074407578, "learning_rate": 3.3517348526339034e-07, "loss": 0.0256, "step": 28950 }, { "grad_norm": 0.12807172536849976, "learning_rate": 3.288326134387454e-07, "loss": 0.0227, "step": 28960 }, { "grad_norm": 0.11864825338125229, "learning_rate": 3.225520967661655e-07, "loss": 0.0203, "step": 28970 }, { "grad_norm": 0.16337837278842926, "learning_rate": 3.163319428770628e-07, "loss": 0.0173, "step": 28980 }, { "grad_norm": 0.12731149792671204, "learning_rate": 3.1017215932951374e-07, "loss": 0.0132, "step": 28990 }, { "grad_norm": 0.08428743481636047, "learning_rate": 3.040727536082366e-07, "loss": 0.0147, "step": 29000 }, { "grad_norm": 0.05956312641501427, "learning_rate": 2.980337331245864e-07, "loss": 0.0125, "step": 29010 }, { "grad_norm": 0.07293201982975006, "learning_rate": 2.9205510521653213e-07, "loss": 0.0133, "step": 29020 }, { "grad_norm": 0.04868275672197342, "learning_rate": 2.86136877148685e-07, "loss": 0.0177, "step": 29030 }, { "grad_norm": 0.16990557312965393, "learning_rate": 2.8027905611223704e-07, "loss": 0.0114, "step": 29040 }, { "grad_norm": 0.07017213106155396, "learning_rate": 2.7448164922500573e-07, "loss": 0.0129, "step": 29050 }, { "grad_norm": 0.14914347231388092, "learning_rate": 2.687446635313784e-07, "loss": 0.0172, "step": 29060 }, { "grad_norm": 0.12786036729812622, "learning_rate": 2.630681060023343e-07, "loss": 0.0152, "step": 29070 }, { "grad_norm": 0.06468004733324051, "learning_rate": 2.5745198353542833e-07, "loss": 0.0162, "step": 29080 }, { "grad_norm": 0.09800933301448822, "learning_rate": 2.518963029547794e-07, "loss": 0.0128, "step": 29090 }, { "grad_norm": 0.0763760432600975, "learning_rate": 2.464010710110598e-07, "loss": 0.0128, "step": 29100 }, { "grad_norm": 0.08697045594453812, "learning_rate": 2.4096629438150054e-07, "loss": 0.0129, "step": 29110 }, { "grad_norm": 0.10500375181436539, "learning_rate": 2.3559197966985802e-07, "loss": 0.0202, "step": 29120 }, { "grad_norm": 0.09854038804769516, "learning_rate": 2.3027813340644188e-07, "loss": 0.0145, "step": 29130 }, { "grad_norm": 0.12660899758338928, "learning_rate": 2.2502476204807055e-07, "loss": 0.0163, "step": 29140 }, { "grad_norm": 0.10079503804445267, "learning_rate": 2.1983187197808786e-07, "loss": 0.0197, "step": 29150 }, { "grad_norm": 0.12920449674129486, "learning_rate": 2.1469946950634644e-07, "loss": 0.0266, "step": 29160 }, { "grad_norm": 0.14925722777843475, "learning_rate": 2.096275608691911e-07, "loss": 0.035, "step": 29170 }, { "grad_norm": 0.11386192589998245, "learning_rate": 2.046161522294754e-07, "loss": 0.0179, "step": 29180 }, { "grad_norm": 0.14699490368366241, "learning_rate": 1.9966524967653944e-07, "loss": 0.0197, "step": 29190 }, { "grad_norm": 0.1429467797279358, "learning_rate": 1.9477485922618222e-07, "loss": 0.0235, "step": 29200 }, { "grad_norm": 0.12461379170417786, "learning_rate": 1.899449868207004e-07, "loss": 0.0136, "step": 29210 }, { "grad_norm": 0.07267045974731445, "learning_rate": 1.851756383288439e-07, "loss": 0.0347, "step": 29220 }, { "grad_norm": 0.17950975894927979, "learning_rate": 1.8046681954581035e-07, "loss": 0.0167, "step": 29230 }, { "grad_norm": 0.07232677936553955, "learning_rate": 1.7581853619327294e-07, "loss": 0.0241, "step": 29240 }, { "grad_norm": 0.1793254166841507, "learning_rate": 1.7123079391932472e-07, "loss": 0.0154, "step": 29250 }, { "grad_norm": 0.09414488077163696, "learning_rate": 1.6670359829850657e-07, "loss": 0.0238, "step": 29260 }, { "grad_norm": 0.07721440494060516, "learning_rate": 1.6223695483179037e-07, "loss": 0.023, "step": 29270 }, { "grad_norm": 0.11248885095119476, "learning_rate": 1.5783086894656795e-07, "loss": 0.0223, "step": 29280 }, { "grad_norm": 0.16388128697872162, "learning_rate": 1.5348534599665121e-07, "loss": 0.0182, "step": 29290 }, { "grad_norm": 0.1760338395833969, "learning_rate": 1.4920039126225527e-07, "loss": 0.023, "step": 29300 }, { "grad_norm": 0.1520654559135437, "learning_rate": 1.449760099500097e-07, "loss": 0.0256, "step": 29310 }, { "grad_norm": 0.15055181086063385, "learning_rate": 1.4081220719293075e-07, "loss": 0.0192, "step": 29320 }, { "grad_norm": 0.09100976586341858, "learning_rate": 1.3670898805043797e-07, "loss": 0.0178, "step": 29330 }, { "grad_norm": 0.11538500338792801, "learning_rate": 1.326663575083209e-07, "loss": 0.0158, "step": 29340 }, { "grad_norm": 0.1673152595758438, "learning_rate": 1.2868432047876688e-07, "loss": 0.0224, "step": 29350 }, { "grad_norm": 0.12073265016078949, "learning_rate": 1.2476288180032213e-07, "loss": 0.0165, "step": 29360 }, { "grad_norm": 0.13167938590049744, "learning_rate": 1.209020462379029e-07, "loss": 0.0207, "step": 29370 }, { "grad_norm": 0.14568635821342468, "learning_rate": 1.1710181848278435e-07, "loss": 0.0193, "step": 29380 }, { "grad_norm": 0.11018969863653183, "learning_rate": 1.133622031526116e-07, "loss": 0.0107, "step": 29390 }, { "grad_norm": 0.07358010858297348, "learning_rate": 1.0968320479136095e-07, "loss": 0.0252, "step": 29400 }, { "grad_norm": 0.15385667979717255, "learning_rate": 1.0606482786936767e-07, "loss": 0.024, "step": 29410 }, { "grad_norm": 0.13298827409744263, "learning_rate": 1.0250707678329808e-07, "loss": 0.0149, "step": 29420 }, { "grad_norm": 0.14043673872947693, "learning_rate": 9.900995585615525e-08, "loss": 0.0206, "step": 29430 }, { "grad_norm": 0.09016802161931992, "learning_rate": 9.55734693372734e-08, "loss": 0.027, "step": 29440 }, { "grad_norm": 0.10892381519079208, "learning_rate": 9.219762140231236e-08, "loss": 0.02, "step": 29450 }, { "grad_norm": 0.09008079767227173, "learning_rate": 8.888241615322978e-08, "loss": 0.012, "step": 29460 }, { "grad_norm": 0.1377497762441635, "learning_rate": 8.562785761833114e-08, "loss": 0.022, "step": 29470 }, { "grad_norm": 0.08145402371883392, "learning_rate": 8.243394975219753e-08, "loss": 0.0324, "step": 29480 }, { "grad_norm": 0.1940058171749115, "learning_rate": 7.930069643573568e-08, "loss": 0.0256, "step": 29490 }, { "grad_norm": 0.17656837403774261, "learning_rate": 7.622810147614456e-08, "loss": 0.0366, "step": 29500 }, { "grad_norm": 0.1442977786064148, "learning_rate": 7.321616860690995e-08, "loss": 0.0202, "step": 29510 }, { "grad_norm": 0.1665685772895813, "learning_rate": 7.026490148782095e-08, "loss": 0.0241, "step": 29520 }, { "grad_norm": 0.13265162706375122, "learning_rate": 6.737430370494236e-08, "loss": 0.0299, "step": 29530 }, { "grad_norm": 0.0677717924118042, "learning_rate": 6.454437877062569e-08, "loss": 0.0356, "step": 29540 }, { "grad_norm": 0.1174728125333786, "learning_rate": 6.177513012349256e-08, "loss": 0.024, "step": 29550 }, { "grad_norm": 0.07957148551940918, "learning_rate": 5.9066561128445775e-08, "loss": 0.0172, "step": 29560 }, { "grad_norm": 0.13589176535606384, "learning_rate": 5.6418675076641556e-08, "loss": 0.0194, "step": 29570 }, { "grad_norm": 0.060098856687545776, "learning_rate": 5.383147518552845e-08, "loss": 0.0279, "step": 29580 }, { "grad_norm": 0.13348542153835297, "learning_rate": 5.1304964598786196e-08, "loss": 0.0288, "step": 29590 }, { "grad_norm": 0.09491535276174545, "learning_rate": 4.883914638636467e-08, "loss": 0.0189, "step": 29600 }, { "grad_norm": 0.09029200673103333, "learning_rate": 4.643402354446713e-08, "loss": 0.0128, "step": 29610 }, { "grad_norm": 0.12431933730840683, "learning_rate": 4.4089598995544766e-08, "loss": 0.0154, "step": 29620 }, { "grad_norm": 0.16028453409671783, "learning_rate": 4.180587558829663e-08, "loss": 0.0296, "step": 29630 }, { "grad_norm": 0.14220476150512695, "learning_rate": 3.958285609765855e-08, "loss": 0.0165, "step": 29640 }, { "grad_norm": 0.16499176621437073, "learning_rate": 3.74205432248087e-08, "loss": 0.0204, "step": 29650 }, { "grad_norm": 0.0827883630990982, "learning_rate": 3.531893959716204e-08, "loss": 0.0205, "step": 29660 }, { "grad_norm": 0.1946626752614975, "learning_rate": 3.327804776837029e-08, "loss": 0.0266, "step": 29670 }, { "grad_norm": 0.15080881118774414, "learning_rate": 3.129787021829977e-08, "loss": 0.0247, "step": 29680 }, { "grad_norm": 0.13864462077617645, "learning_rate": 2.9378409353059133e-08, "loss": 0.0152, "step": 29690 }, { "grad_norm": 0.13161145150661469, "learning_rate": 2.7519667504971593e-08, "loss": 0.0169, "step": 29700 }, { "grad_norm": 0.20179538428783417, "learning_rate": 2.572164693258605e-08, "loss": 0.026, "step": 29710 }, { "grad_norm": 0.09696594625711441, "learning_rate": 2.3984349820665996e-08, "loss": 0.0303, "step": 29720 }, { "grad_norm": 0.11584252119064331, "learning_rate": 2.2307778280189485e-08, "loss": 0.0364, "step": 29730 }, { "grad_norm": 0.0708511471748352, "learning_rate": 2.0691934348354704e-08, "loss": 0.0217, "step": 29740 }, { "grad_norm": 0.1613028645515442, "learning_rate": 1.9136819988557763e-08, "loss": 0.0292, "step": 29750 }, { "grad_norm": 0.10159213095903397, "learning_rate": 1.7642437090414908e-08, "loss": 0.0212, "step": 29760 }, { "grad_norm": 0.10728535801172256, "learning_rate": 1.6208787469734755e-08, "loss": 0.0338, "step": 29770 }, { "grad_norm": 0.0878261923789978, "learning_rate": 1.483587286854604e-08, "loss": 0.023, "step": 29780 }, { "grad_norm": 0.1006697416305542, "learning_rate": 1.3523694955064336e-08, "loss": 0.0187, "step": 29790 }, { "grad_norm": 0.11944656074047089, "learning_rate": 1.2272255323708681e-08, "loss": 0.0228, "step": 29800 }, { "grad_norm": 0.09436451643705368, "learning_rate": 1.1081555495096042e-08, "loss": 0.0198, "step": 29810 }, { "grad_norm": 0.09309688955545425, "learning_rate": 9.951596916041306e-09, "loss": 0.0183, "step": 29820 }, { "grad_norm": 0.23777581751346588, "learning_rate": 8.882380959551739e-09, "loss": 0.038, "step": 29830 }, { "grad_norm": 0.09651901572942734, "learning_rate": 7.873908924821428e-09, "loss": 0.0259, "step": 29840 }, { "grad_norm": 0.16517260670661926, "learning_rate": 6.926182037242379e-09, "loss": 0.0319, "step": 29850 }, { "grad_norm": 0.08165998756885529, "learning_rate": 6.039201448387876e-09, "loss": 0.019, "step": 29860 }, { "grad_norm": 0.10257837921380997, "learning_rate": 5.212968236029125e-09, "loss": 0.0168, "step": 29870 }, { "grad_norm": 0.14416112005710602, "learning_rate": 4.447483404118602e-09, "loss": 0.0204, "step": 29880 }, { "grad_norm": 0.07579655200242996, "learning_rate": 3.742747882784503e-09, "loss": 0.0158, "step": 29890 }, { "grad_norm": 0.11301553249359131, "learning_rate": 3.0987625283473987e-09, "loss": 0.0355, "step": 29900 }, { "grad_norm": 0.13612067699432373, "learning_rate": 2.515528123320232e-09, "loss": 0.0192, "step": 29910 }, { "grad_norm": 0.0953507199883461, "learning_rate": 1.9930453763750133e-09, "loss": 0.0221, "step": 29920 }, { "grad_norm": 0.15751031041145325, "learning_rate": 1.5313149223872281e-09, "loss": 0.0331, "step": 29930 }, { "grad_norm": 0.13335071504116058, "learning_rate": 1.1303373224025305e-09, "loss": 0.0257, "step": 29940 }, { "grad_norm": 0.10917995125055313, "learning_rate": 7.901130636367438e-10, "loss": 0.0185, "step": 29950 }, { "grad_norm": 0.12676796317100525, "learning_rate": 5.106425595036158e-10, "loss": 0.0195, "step": 29960 }, { "grad_norm": 0.11761768907308578, "learning_rate": 2.9192614958706286e-10, "loss": 0.0204, "step": 29970 }, { "grad_norm": 0.13873013854026794, "learning_rate": 1.3396409964117064e-10, "loss": 0.0258, "step": 29980 }, { "grad_norm": 0.1284443885087967, "learning_rate": 3.6756601606846574e-11, "loss": 0.0283, "step": 29990 }, { "grad_norm": 0.16682197153568268, "learning_rate": 3.0377360626943076e-13, "loss": 0.0168, "step": 30000 } ], "logging_steps": 10, "max_steps": 30000, "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": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }