diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,6034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.3333333333333333, + "eval_steps": 500, + "global_step": 10000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "grad_norm": 1.9198344945907593, + "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 + } + ], + "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": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}