{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.8765634894371033, "learning_rate": 1.8e-07, "loss": 1.4958, "step": 10 }, { "grad_norm": 1.0620571374893188, "learning_rate": 3.8e-07, "loss": 1.5115, "step": 20 }, { "grad_norm": 0.9524933695793152, "learning_rate": 5.8e-07, "loss": 1.439, "step": 30 }, { "grad_norm": 1.171401023864746, "learning_rate": 7.8e-07, "loss": 1.4575, "step": 40 }, { "grad_norm": 0.9207823276519775, "learning_rate": 9.8e-07, "loss": 1.4575, "step": 50 }, { "grad_norm": 0.914197564125061, "learning_rate": 1.18e-06, "loss": 1.4399, "step": 60 }, { "grad_norm": 1.621327519416809, "learning_rate": 1.3800000000000001e-06, "loss": 1.3879, "step": 70 }, { "grad_norm": 1.662928819656372, "learning_rate": 1.5800000000000003e-06, "loss": 1.3377, "step": 80 }, { "grad_norm": 1.2807358503341675, "learning_rate": 1.7800000000000001e-06, "loss": 1.2488, "step": 90 }, { "grad_norm": 1.532633900642395, "learning_rate": 1.98e-06, "loss": 1.2275, "step": 100 }, { "grad_norm": 1.2939066886901855, "learning_rate": 2.1800000000000003e-06, "loss": 1.2604, "step": 110 }, { "grad_norm": 1.9265021085739136, "learning_rate": 2.38e-06, "loss": 1.2307, "step": 120 }, { "grad_norm": 1.6161824464797974, "learning_rate": 2.5800000000000003e-06, "loss": 1.1629, "step": 130 }, { "grad_norm": 1.5722763538360596, "learning_rate": 2.78e-06, "loss": 1.1738, "step": 140 }, { "grad_norm": 1.4991711378097534, "learning_rate": 2.9800000000000003e-06, "loss": 1.1192, "step": 150 }, { "grad_norm": 1.1773643493652344, "learning_rate": 3.1800000000000005e-06, "loss": 1.1243, "step": 160 }, { "grad_norm": 2.319329023361206, "learning_rate": 3.38e-06, "loss": 1.1062, "step": 170 }, { "grad_norm": 1.269884705543518, "learning_rate": 3.58e-06, "loss": 1.1124, "step": 180 }, { "grad_norm": 1.265794038772583, "learning_rate": 3.7800000000000002e-06, "loss": 1.0996, "step": 190 }, { "grad_norm": 1.8343150615692139, "learning_rate": 3.98e-06, "loss": 1.1058, "step": 200 }, { "grad_norm": 1.4203811883926392, "learning_rate": 4.18e-06, "loss": 1.0919, "step": 210 }, { "grad_norm": 1.3591406345367432, "learning_rate": 4.38e-06, "loss": 1.0995, "step": 220 }, { "grad_norm": 1.337270736694336, "learning_rate": 4.58e-06, "loss": 1.0859, "step": 230 }, { "grad_norm": 1.552187442779541, "learning_rate": 4.780000000000001e-06, "loss": 1.0999, "step": 240 }, { "grad_norm": 1.2191849946975708, "learning_rate": 4.98e-06, "loss": 1.0886, "step": 250 }, { "grad_norm": 1.9007881879806519, "learning_rate": 5.18e-06, "loss": 1.0696, "step": 260 }, { "grad_norm": 1.5075709819793701, "learning_rate": 5.38e-06, "loss": 1.0744, "step": 270 }, { "grad_norm": 1.3360384702682495, "learning_rate": 5.580000000000001e-06, "loss": 1.0526, "step": 280 }, { "grad_norm": 1.6621371507644653, "learning_rate": 5.78e-06, "loss": 1.089, "step": 290 }, { "grad_norm": 1.1322021484375, "learning_rate": 5.98e-06, "loss": 1.0773, "step": 300 }, { "grad_norm": 1.3267987966537476, "learning_rate": 6.18e-06, "loss": 1.0862, "step": 310 }, { "grad_norm": 1.3511558771133423, "learning_rate": 6.38e-06, "loss": 1.0801, "step": 320 }, { "grad_norm": 1.2178771495819092, "learning_rate": 6.58e-06, "loss": 1.0702, "step": 330 }, { "grad_norm": 1.5012621879577637, "learning_rate": 6.78e-06, "loss": 1.0686, "step": 340 }, { "grad_norm": 1.4785507917404175, "learning_rate": 6.98e-06, "loss": 1.0619, "step": 350 }, { "grad_norm": 1.2449885606765747, "learning_rate": 7.180000000000001e-06, "loss": 1.0758, "step": 360 }, { "grad_norm": 1.6656064987182617, "learning_rate": 7.3800000000000005e-06, "loss": 1.0616, "step": 370 }, { "grad_norm": 1.4091850519180298, "learning_rate": 7.580000000000001e-06, "loss": 1.0654, "step": 380 }, { "grad_norm": 1.368632435798645, "learning_rate": 7.78e-06, "loss": 1.0374, "step": 390 }, { "grad_norm": 1.6620794534683228, "learning_rate": 7.98e-06, "loss": 1.0633, "step": 400 }, { "grad_norm": 1.6231037378311157, "learning_rate": 8.18e-06, "loss": 1.0512, "step": 410 }, { "grad_norm": 1.475819706916809, "learning_rate": 8.380000000000001e-06, "loss": 1.0144, "step": 420 }, { "grad_norm": 2.2538859844207764, "learning_rate": 8.580000000000001e-06, "loss": 0.9873, "step": 430 }, { "grad_norm": 1.891244888305664, "learning_rate": 8.78e-06, "loss": 0.9916, "step": 440 }, { "grad_norm": 2.2251787185668945, "learning_rate": 8.98e-06, "loss": 0.9401, "step": 450 }, { "grad_norm": 2.1357831954956055, "learning_rate": 9.180000000000002e-06, "loss": 0.9312, "step": 460 }, { "grad_norm": 2.406728982925415, "learning_rate": 9.38e-06, "loss": 0.918, "step": 470 }, { "grad_norm": 2.3847551345825195, "learning_rate": 9.58e-06, "loss": 0.9273, "step": 480 }, { "grad_norm": 2.4059605598449707, "learning_rate": 9.78e-06, "loss": 0.88, "step": 490 }, { "grad_norm": 2.132392644882202, "learning_rate": 9.980000000000001e-06, "loss": 0.8781, "step": 500 }, { "grad_norm": 2.876146078109741, "learning_rate": 1.018e-05, "loss": 0.8388, "step": 510 }, { "grad_norm": 2.1167542934417725, "learning_rate": 1.038e-05, "loss": 0.8157, "step": 520 }, { "grad_norm": 2.298130989074707, "learning_rate": 1.058e-05, "loss": 0.7938, "step": 530 }, { "grad_norm": 2.3994998931884766, "learning_rate": 1.0780000000000002e-05, "loss": 0.7666, "step": 540 }, { "grad_norm": 2.429746627807617, "learning_rate": 1.098e-05, "loss": 0.7721, "step": 550 }, { "grad_norm": 2.3058602809906006, "learning_rate": 1.118e-05, "loss": 0.7494, "step": 560 }, { "grad_norm": 2.140563726425171, "learning_rate": 1.1380000000000001e-05, "loss": 0.7013, "step": 570 }, { "grad_norm": 2.0576517581939697, "learning_rate": 1.1580000000000001e-05, "loss": 0.6715, "step": 580 }, { "grad_norm": 2.6971747875213623, "learning_rate": 1.178e-05, "loss": 0.6761, "step": 590 }, { "grad_norm": 2.3286659717559814, "learning_rate": 1.198e-05, "loss": 0.6482, "step": 600 }, { "grad_norm": 2.5937130451202393, "learning_rate": 1.2180000000000002e-05, "loss": 0.6314, "step": 610 }, { "grad_norm": 2.564598560333252, "learning_rate": 1.238e-05, "loss": 0.6174, "step": 620 }, { "grad_norm": 2.4552674293518066, "learning_rate": 1.258e-05, "loss": 0.609, "step": 630 }, { "grad_norm": 3.0940825939178467, "learning_rate": 1.278e-05, "loss": 0.5835, "step": 640 }, { "grad_norm": 2.679537057876587, "learning_rate": 1.2980000000000001e-05, "loss": 0.5585, "step": 650 }, { "grad_norm": 3.146684169769287, "learning_rate": 1.3180000000000001e-05, "loss": 0.5316, "step": 660 }, { "grad_norm": 2.2390682697296143, "learning_rate": 1.338e-05, "loss": 0.5068, "step": 670 }, { "grad_norm": 3.0834591388702393, "learning_rate": 1.358e-05, "loss": 0.4899, "step": 680 }, { "grad_norm": 2.7289109230041504, "learning_rate": 1.3780000000000002e-05, "loss": 0.5001, "step": 690 }, { "grad_norm": 3.7889773845672607, "learning_rate": 1.3980000000000002e-05, "loss": 0.5046, "step": 700 }, { "grad_norm": 3.451676368713379, "learning_rate": 1.4180000000000001e-05, "loss": 0.439, "step": 710 }, { "grad_norm": 3.732891798019409, "learning_rate": 1.4380000000000001e-05, "loss": 0.442, "step": 720 }, { "grad_norm": 3.9701459407806396, "learning_rate": 1.4580000000000003e-05, "loss": 0.4158, "step": 730 }, { "grad_norm": 2.712930202484131, "learning_rate": 1.4779999999999999e-05, "loss": 0.3816, "step": 740 }, { "grad_norm": 3.6617367267608643, "learning_rate": 1.4979999999999999e-05, "loss": 0.398, "step": 750 }, { "grad_norm": 2.8146374225616455, "learning_rate": 1.518e-05, "loss": 0.386, "step": 760 }, { "grad_norm": 2.637340784072876, "learning_rate": 1.538e-05, "loss": 0.3705, "step": 770 }, { "grad_norm": 2.1509950160980225, "learning_rate": 1.558e-05, "loss": 0.3835, "step": 780 }, { "grad_norm": 3.2695505619049072, "learning_rate": 1.578e-05, "loss": 0.3194, "step": 790 }, { "grad_norm": 3.4243319034576416, "learning_rate": 1.598e-05, "loss": 0.3112, "step": 800 }, { "grad_norm": 2.88128399848938, "learning_rate": 1.618e-05, "loss": 0.3124, "step": 810 }, { "grad_norm": 2.5241973400115967, "learning_rate": 1.6380000000000002e-05, "loss": 0.2808, "step": 820 }, { "grad_norm": 2.8489606380462646, "learning_rate": 1.658e-05, "loss": 0.3057, "step": 830 }, { "grad_norm": 2.675304889678955, "learning_rate": 1.6780000000000002e-05, "loss": 0.2932, "step": 840 }, { "grad_norm": 2.44004487991333, "learning_rate": 1.698e-05, "loss": 0.3081, "step": 850 }, { "grad_norm": 3.0543479919433594, "learning_rate": 1.718e-05, "loss": 0.2624, "step": 860 }, { "grad_norm": 2.7427821159362793, "learning_rate": 1.7380000000000003e-05, "loss": 0.2915, "step": 870 }, { "grad_norm": 3.456601858139038, "learning_rate": 1.758e-05, "loss": 0.263, "step": 880 }, { "grad_norm": 2.626004219055176, "learning_rate": 1.7780000000000003e-05, "loss": 0.2533, "step": 890 }, { "grad_norm": 2.9026503562927246, "learning_rate": 1.798e-05, "loss": 0.2492, "step": 900 }, { "grad_norm": 2.2049880027770996, "learning_rate": 1.818e-05, "loss": 0.2398, "step": 910 }, { "grad_norm": 2.2458372116088867, "learning_rate": 1.838e-05, "loss": 0.2023, "step": 920 }, { "grad_norm": 2.4705677032470703, "learning_rate": 1.858e-05, "loss": 0.216, "step": 930 }, { "grad_norm": 3.1009089946746826, "learning_rate": 1.878e-05, "loss": 0.2129, "step": 940 }, { "grad_norm": 2.562770128250122, "learning_rate": 1.898e-05, "loss": 0.2005, "step": 950 }, { "grad_norm": 2.713258743286133, "learning_rate": 1.918e-05, "loss": 0.2184, "step": 960 }, { "grad_norm": 3.1045894622802734, "learning_rate": 1.938e-05, "loss": 0.2338, "step": 970 }, { "grad_norm": 2.9466073513031006, "learning_rate": 1.9580000000000002e-05, "loss": 0.1948, "step": 980 }, { "grad_norm": 2.7939882278442383, "learning_rate": 1.978e-05, "loss": 0.1806, "step": 990 }, { "grad_norm": 3.3212344646453857, "learning_rate": 1.9980000000000002e-05, "loss": 0.1792, "step": 1000 }, { "grad_norm": 2.2782654762268066, "learning_rate": 2.0180000000000003e-05, "loss": 0.2033, "step": 1010 }, { "grad_norm": 2.4946494102478027, "learning_rate": 2.038e-05, "loss": 0.2226, "step": 1020 }, { "grad_norm": 2.3669347763061523, "learning_rate": 2.0580000000000003e-05, "loss": 0.1793, "step": 1030 }, { "grad_norm": 2.4426004886627197, "learning_rate": 2.078e-05, "loss": 0.2045, "step": 1040 }, { "grad_norm": 2.412241220474243, "learning_rate": 2.098e-05, "loss": 0.1927, "step": 1050 }, { "grad_norm": 2.84100341796875, "learning_rate": 2.118e-05, "loss": 0.2043, "step": 1060 }, { "grad_norm": 2.2923145294189453, "learning_rate": 2.138e-05, "loss": 0.207, "step": 1070 }, { "grad_norm": 3.419609546661377, "learning_rate": 2.158e-05, "loss": 0.2139, "step": 1080 }, { "grad_norm": 1.8782737255096436, "learning_rate": 2.178e-05, "loss": 0.1876, "step": 1090 }, { "grad_norm": 2.664591073989868, "learning_rate": 2.198e-05, "loss": 0.1558, "step": 1100 }, { "grad_norm": 3.162869930267334, "learning_rate": 2.218e-05, "loss": 0.2375, "step": 1110 }, { "grad_norm": 1.9275411367416382, "learning_rate": 2.2380000000000003e-05, "loss": 0.1599, "step": 1120 }, { "grad_norm": 2.0433754920959473, "learning_rate": 2.258e-05, "loss": 0.1809, "step": 1130 }, { "grad_norm": 2.7762274742126465, "learning_rate": 2.2780000000000002e-05, "loss": 0.226, "step": 1140 }, { "grad_norm": 2.3303182125091553, "learning_rate": 2.298e-05, "loss": 0.2175, "step": 1150 }, { "grad_norm": 2.023251533508301, "learning_rate": 2.318e-05, "loss": 0.1883, "step": 1160 }, { "grad_norm": 2.1792819499969482, "learning_rate": 2.3380000000000003e-05, "loss": 0.1763, "step": 1170 }, { "grad_norm": 2.269564151763916, "learning_rate": 2.358e-05, "loss": 0.1591, "step": 1180 }, { "grad_norm": 2.9537906646728516, "learning_rate": 2.3780000000000003e-05, "loss": 0.1755, "step": 1190 }, { "grad_norm": 3.7030911445617676, "learning_rate": 2.398e-05, "loss": 0.1841, "step": 1200 }, { "grad_norm": 3.3902029991149902, "learning_rate": 2.418e-05, "loss": 0.1981, "step": 1210 }, { "grad_norm": 2.4147660732269287, "learning_rate": 2.438e-05, "loss": 0.1589, "step": 1220 }, { "grad_norm": 2.4440174102783203, "learning_rate": 2.4580000000000002e-05, "loss": 0.1686, "step": 1230 }, { "grad_norm": 1.7562744617462158, "learning_rate": 2.478e-05, "loss": 0.176, "step": 1240 }, { "grad_norm": 2.881347894668579, "learning_rate": 2.498e-05, "loss": 0.1648, "step": 1250 }, { "grad_norm": 3.338353395462036, "learning_rate": 2.5180000000000003e-05, "loss": 0.1721, "step": 1260 }, { "grad_norm": 2.2503535747528076, "learning_rate": 2.5380000000000004e-05, "loss": 0.1482, "step": 1270 }, { "grad_norm": 2.068032741546631, "learning_rate": 2.5580000000000002e-05, "loss": 0.176, "step": 1280 }, { "grad_norm": 1.7723352909088135, "learning_rate": 2.5779999999999997e-05, "loss": 0.1696, "step": 1290 }, { "grad_norm": 2.324462890625, "learning_rate": 2.598e-05, "loss": 0.1574, "step": 1300 }, { "grad_norm": 2.895449161529541, "learning_rate": 2.618e-05, "loss": 0.1717, "step": 1310 }, { "grad_norm": 2.3051817417144775, "learning_rate": 2.6379999999999998e-05, "loss": 0.1634, "step": 1320 }, { "grad_norm": 1.9681099653244019, "learning_rate": 2.658e-05, "loss": 0.1415, "step": 1330 }, { "grad_norm": 2.2310211658477783, "learning_rate": 2.678e-05, "loss": 0.1368, "step": 1340 }, { "grad_norm": 5.104952812194824, "learning_rate": 2.698e-05, "loss": 0.1795, "step": 1350 }, { "grad_norm": 2.9009690284729004, "learning_rate": 2.718e-05, "loss": 0.1634, "step": 1360 }, { "grad_norm": 2.0108234882354736, "learning_rate": 2.738e-05, "loss": 0.1459, "step": 1370 }, { "grad_norm": 2.0635056495666504, "learning_rate": 2.758e-05, "loss": 0.1452, "step": 1380 }, { "grad_norm": 2.1579325199127197, "learning_rate": 2.778e-05, "loss": 0.1506, "step": 1390 }, { "grad_norm": 2.032189130783081, "learning_rate": 2.798e-05, "loss": 0.1471, "step": 1400 }, { "grad_norm": 2.1149802207946777, "learning_rate": 2.818e-05, "loss": 0.1648, "step": 1410 }, { "grad_norm": 2.3838729858398438, "learning_rate": 2.8380000000000003e-05, "loss": 0.1786, "step": 1420 }, { "grad_norm": 1.474086046218872, "learning_rate": 2.858e-05, "loss": 0.1259, "step": 1430 }, { "grad_norm": 2.1022074222564697, "learning_rate": 2.8780000000000002e-05, "loss": 0.1566, "step": 1440 }, { "grad_norm": 1.8406864404678345, "learning_rate": 2.898e-05, "loss": 0.1693, "step": 1450 }, { "grad_norm": 1.9055906534194946, "learning_rate": 2.9180000000000002e-05, "loss": 0.1715, "step": 1460 }, { "grad_norm": 1.886681318283081, "learning_rate": 2.9380000000000003e-05, "loss": 0.1607, "step": 1470 }, { "grad_norm": 1.5404877662658691, "learning_rate": 2.958e-05, "loss": 0.1349, "step": 1480 }, { "grad_norm": 1.6678810119628906, "learning_rate": 2.9780000000000003e-05, "loss": 0.1445, "step": 1490 }, { "grad_norm": 1.6152527332305908, "learning_rate": 2.998e-05, "loss": 0.1649, "step": 1500 }, { "grad_norm": 1.292999505996704, "learning_rate": 3.0180000000000002e-05, "loss": 0.145, "step": 1510 }, { "grad_norm": 2.4206838607788086, "learning_rate": 3.0380000000000004e-05, "loss": 0.1679, "step": 1520 }, { "grad_norm": 1.5930098295211792, "learning_rate": 3.058e-05, "loss": 0.1495, "step": 1530 }, { "grad_norm": 1.631879448890686, "learning_rate": 3.078e-05, "loss": 0.1547, "step": 1540 }, { "grad_norm": 1.5346232652664185, "learning_rate": 3.0980000000000005e-05, "loss": 0.1338, "step": 1550 }, { "grad_norm": 1.5971662998199463, "learning_rate": 3.118e-05, "loss": 0.1692, "step": 1560 }, { "grad_norm": 1.7635058164596558, "learning_rate": 3.138e-05, "loss": 0.1356, "step": 1570 }, { "grad_norm": 1.8980156183242798, "learning_rate": 3.1580000000000006e-05, "loss": 0.1503, "step": 1580 }, { "grad_norm": 2.027613401412964, "learning_rate": 3.1780000000000004e-05, "loss": 0.1659, "step": 1590 }, { "grad_norm": 1.8548860549926758, "learning_rate": 3.198e-05, "loss": 0.1447, "step": 1600 }, { "grad_norm": 1.1391115188598633, "learning_rate": 3.218e-05, "loss": 0.1238, "step": 1610 }, { "grad_norm": 2.0970590114593506, "learning_rate": 3.238e-05, "loss": 0.1527, "step": 1620 }, { "grad_norm": 1.8569316864013672, "learning_rate": 3.2579999999999996e-05, "loss": 0.1867, "step": 1630 }, { "grad_norm": 1.6760543584823608, "learning_rate": 3.278e-05, "loss": 0.1462, "step": 1640 }, { "grad_norm": 1.9324018955230713, "learning_rate": 3.298e-05, "loss": 0.159, "step": 1650 }, { "grad_norm": 1.618060827255249, "learning_rate": 3.318e-05, "loss": 0.1518, "step": 1660 }, { "grad_norm": 1.5037598609924316, "learning_rate": 3.338e-05, "loss": 0.1264, "step": 1670 }, { "grad_norm": 2.053913116455078, "learning_rate": 3.358e-05, "loss": 0.1272, "step": 1680 }, { "grad_norm": 1.9638161659240723, "learning_rate": 3.378e-05, "loss": 0.144, "step": 1690 }, { "grad_norm": 1.4484996795654297, "learning_rate": 3.398e-05, "loss": 0.1397, "step": 1700 }, { "grad_norm": 2.0244300365448, "learning_rate": 3.418e-05, "loss": 0.1696, "step": 1710 }, { "grad_norm": 1.5042684078216553, "learning_rate": 3.438e-05, "loss": 0.1452, "step": 1720 }, { "grad_norm": 1.6254792213439941, "learning_rate": 3.4580000000000004e-05, "loss": 0.1416, "step": 1730 }, { "grad_norm": 1.8259013891220093, "learning_rate": 3.478e-05, "loss": 0.135, "step": 1740 }, { "grad_norm": 2.110121488571167, "learning_rate": 3.498e-05, "loss": 0.1517, "step": 1750 }, { "grad_norm": 2.0193302631378174, "learning_rate": 3.518e-05, "loss": 0.14, "step": 1760 }, { "grad_norm": 1.808740496635437, "learning_rate": 3.5380000000000003e-05, "loss": 0.1408, "step": 1770 }, { "grad_norm": 1.6526018381118774, "learning_rate": 3.558e-05, "loss": 0.1496, "step": 1780 }, { "grad_norm": 1.6004718542099, "learning_rate": 3.578e-05, "loss": 0.1255, "step": 1790 }, { "grad_norm": 1.6326842308044434, "learning_rate": 3.5980000000000004e-05, "loss": 0.1427, "step": 1800 }, { "grad_norm": 1.645699143409729, "learning_rate": 3.618e-05, "loss": 0.1348, "step": 1810 }, { "grad_norm": 2.237473487854004, "learning_rate": 3.638e-05, "loss": 0.1567, "step": 1820 }, { "grad_norm": 2.2555627822875977, "learning_rate": 3.6580000000000006e-05, "loss": 0.1545, "step": 1830 }, { "grad_norm": 2.025695562362671, "learning_rate": 3.6780000000000004e-05, "loss": 0.1531, "step": 1840 }, { "grad_norm": 1.4468196630477905, "learning_rate": 3.698e-05, "loss": 0.1745, "step": 1850 }, { "grad_norm": 1.617296814918518, "learning_rate": 3.7180000000000007e-05, "loss": 0.1621, "step": 1860 }, { "grad_norm": 1.3272672891616821, "learning_rate": 3.7380000000000005e-05, "loss": 0.1268, "step": 1870 }, { "grad_norm": 1.5455641746520996, "learning_rate": 3.758e-05, "loss": 0.1163, "step": 1880 }, { "grad_norm": 1.517259120941162, "learning_rate": 3.778000000000001e-05, "loss": 0.1517, "step": 1890 }, { "grad_norm": 1.3489035367965698, "learning_rate": 3.7980000000000006e-05, "loss": 0.1343, "step": 1900 }, { "grad_norm": 2.095867156982422, "learning_rate": 3.818e-05, "loss": 0.1567, "step": 1910 }, { "grad_norm": 2.0543196201324463, "learning_rate": 3.838e-05, "loss": 0.1429, "step": 1920 }, { "grad_norm": 1.316035509109497, "learning_rate": 3.858e-05, "loss": 0.135, "step": 1930 }, { "grad_norm": 2.084826707839966, "learning_rate": 3.878e-05, "loss": 0.1309, "step": 1940 }, { "grad_norm": 1.4024131298065186, "learning_rate": 3.898e-05, "loss": 0.1248, "step": 1950 }, { "grad_norm": 1.7635453939437866, "learning_rate": 3.918e-05, "loss": 0.1738, "step": 1960 }, { "grad_norm": 1.3917206525802612, "learning_rate": 3.938e-05, "loss": 0.1296, "step": 1970 }, { "grad_norm": 1.4463366270065308, "learning_rate": 3.958e-05, "loss": 0.1531, "step": 1980 }, { "grad_norm": 1.4400206804275513, "learning_rate": 3.978e-05, "loss": 0.127, "step": 1990 }, { "grad_norm": 1.0920056104660034, "learning_rate": 3.998e-05, "loss": 0.1357, "step": 2000 }, { "grad_norm": 1.6441763639450073, "learning_rate": 4.018e-05, "loss": 0.1651, "step": 2010 }, { "grad_norm": 1.8122658729553223, "learning_rate": 4.038e-05, "loss": 0.1432, "step": 2020 }, { "grad_norm": 1.8747669458389282, "learning_rate": 4.058e-05, "loss": 0.1335, "step": 2030 }, { "grad_norm": 1.6175390481948853, "learning_rate": 4.078e-05, "loss": 0.156, "step": 2040 }, { "grad_norm": 1.7609783411026, "learning_rate": 4.0980000000000004e-05, "loss": 0.1502, "step": 2050 }, { "grad_norm": 1.5865976810455322, "learning_rate": 4.118e-05, "loss": 0.1289, "step": 2060 }, { "grad_norm": 1.4003666639328003, "learning_rate": 4.138e-05, "loss": 0.1204, "step": 2070 }, { "grad_norm": 1.4310914278030396, "learning_rate": 4.1580000000000005e-05, "loss": 0.1278, "step": 2080 }, { "grad_norm": 0.9741531610488892, "learning_rate": 4.178e-05, "loss": 0.1427, "step": 2090 }, { "grad_norm": 1.4184911251068115, "learning_rate": 4.198e-05, "loss": 0.1286, "step": 2100 }, { "grad_norm": 1.3844940662384033, "learning_rate": 4.2180000000000006e-05, "loss": 0.1564, "step": 2110 }, { "grad_norm": 1.1641044616699219, "learning_rate": 4.2380000000000004e-05, "loss": 0.1273, "step": 2120 }, { "grad_norm": 1.1882753372192383, "learning_rate": 4.258e-05, "loss": 0.1371, "step": 2130 }, { "grad_norm": 1.1859841346740723, "learning_rate": 4.278e-05, "loss": 0.1042, "step": 2140 }, { "grad_norm": 1.1855552196502686, "learning_rate": 4.2980000000000005e-05, "loss": 0.1429, "step": 2150 }, { "grad_norm": 1.1952366828918457, "learning_rate": 4.318e-05, "loss": 0.1258, "step": 2160 }, { "grad_norm": 1.125609278678894, "learning_rate": 4.338e-05, "loss": 0.1347, "step": 2170 }, { "grad_norm": 1.1549887657165527, "learning_rate": 4.3580000000000006e-05, "loss": 0.1313, "step": 2180 }, { "grad_norm": 1.5188688039779663, "learning_rate": 4.3780000000000004e-05, "loss": 0.1255, "step": 2190 }, { "grad_norm": 1.484584093093872, "learning_rate": 4.398e-05, "loss": 0.1413, "step": 2200 }, { "grad_norm": 1.109092354774475, "learning_rate": 4.418000000000001e-05, "loss": 0.1364, "step": 2210 }, { "grad_norm": 1.7425835132598877, "learning_rate": 4.438e-05, "loss": 0.1373, "step": 2220 }, { "grad_norm": 1.4852831363677979, "learning_rate": 4.458e-05, "loss": 0.1311, "step": 2230 }, { "grad_norm": 1.280571699142456, "learning_rate": 4.478e-05, "loss": 0.1376, "step": 2240 }, { "grad_norm": 1.0935192108154297, "learning_rate": 4.498e-05, "loss": 0.1442, "step": 2250 }, { "grad_norm": 1.2401803731918335, "learning_rate": 4.518e-05, "loss": 0.1156, "step": 2260 }, { "grad_norm": 1.247541069984436, "learning_rate": 4.538e-05, "loss": 0.1216, "step": 2270 }, { "grad_norm": 1.394266128540039, "learning_rate": 4.558e-05, "loss": 0.1275, "step": 2280 }, { "grad_norm": 1.371987223625183, "learning_rate": 4.578e-05, "loss": 0.1555, "step": 2290 }, { "grad_norm": 1.193678617477417, "learning_rate": 4.5980000000000004e-05, "loss": 0.1167, "step": 2300 }, { "grad_norm": 0.9604241847991943, "learning_rate": 4.618e-05, "loss": 0.1254, "step": 2310 }, { "grad_norm": 1.0916401147842407, "learning_rate": 4.638e-05, "loss": 0.1132, "step": 2320 }, { "grad_norm": 1.2437673807144165, "learning_rate": 4.6580000000000005e-05, "loss": 0.1467, "step": 2330 }, { "grad_norm": 1.2683857679367065, "learning_rate": 4.678e-05, "loss": 0.1173, "step": 2340 }, { "grad_norm": 1.289565920829773, "learning_rate": 4.698e-05, "loss": 0.128, "step": 2350 }, { "grad_norm": 1.3278840780258179, "learning_rate": 4.718e-05, "loss": 0.1323, "step": 2360 }, { "grad_norm": 0.8305469155311584, "learning_rate": 4.7380000000000004e-05, "loss": 0.1243, "step": 2370 }, { "grad_norm": 1.3830838203430176, "learning_rate": 4.758e-05, "loss": 0.1375, "step": 2380 }, { "grad_norm": 1.1192916631698608, "learning_rate": 4.778e-05, "loss": 0.1469, "step": 2390 }, { "grad_norm": 1.2486354112625122, "learning_rate": 4.7980000000000005e-05, "loss": 0.1331, "step": 2400 }, { "grad_norm": 1.0245213508605957, "learning_rate": 4.818e-05, "loss": 0.1314, "step": 2410 }, { "grad_norm": 1.6839102506637573, "learning_rate": 4.838e-05, "loss": 0.1747, "step": 2420 }, { "grad_norm": 1.2584189176559448, "learning_rate": 4.8580000000000006e-05, "loss": 0.1421, "step": 2430 }, { "grad_norm": 1.3957077264785767, "learning_rate": 4.8780000000000004e-05, "loss": 0.1478, "step": 2440 }, { "grad_norm": 1.1468400955200195, "learning_rate": 4.898e-05, "loss": 0.1121, "step": 2450 }, { "grad_norm": 0.954909086227417, "learning_rate": 4.918000000000001e-05, "loss": 0.1409, "step": 2460 }, { "grad_norm": 1.1152756214141846, "learning_rate": 4.9380000000000005e-05, "loss": 0.1063, "step": 2470 }, { "grad_norm": 1.368930697441101, "learning_rate": 4.958e-05, "loss": 0.1447, "step": 2480 }, { "grad_norm": 0.9763432741165161, "learning_rate": 4.978e-05, "loss": 0.1142, "step": 2490 }, { "grad_norm": 1.0341376066207886, "learning_rate": 4.9980000000000006e-05, "loss": 0.119, "step": 2500 }, { "grad_norm": 1.2655800580978394, "learning_rate": 5.0180000000000004e-05, "loss": 0.1182, "step": 2510 }, { "grad_norm": 1.3022370338439941, "learning_rate": 5.038e-05, "loss": 0.1295, "step": 2520 }, { "grad_norm": 0.9598504900932312, "learning_rate": 5.058000000000001e-05, "loss": 0.1083, "step": 2530 }, { "grad_norm": 1.0005040168762207, "learning_rate": 5.0780000000000005e-05, "loss": 0.1324, "step": 2540 }, { "grad_norm": 1.265838384628296, "learning_rate": 5.098e-05, "loss": 0.1361, "step": 2550 }, { "grad_norm": 1.1728522777557373, "learning_rate": 5.118000000000001e-05, "loss": 0.13, "step": 2560 }, { "grad_norm": 1.0428338050842285, "learning_rate": 5.1380000000000006e-05, "loss": 0.1421, "step": 2570 }, { "grad_norm": 1.221566081047058, "learning_rate": 5.1580000000000004e-05, "loss": 0.1441, "step": 2580 }, { "grad_norm": 0.9935548305511475, "learning_rate": 5.178000000000001e-05, "loss": 0.1254, "step": 2590 }, { "grad_norm": 1.16371488571167, "learning_rate": 5.198000000000001e-05, "loss": 0.1321, "step": 2600 }, { "grad_norm": 1.0349863767623901, "learning_rate": 5.2180000000000005e-05, "loss": 0.1383, "step": 2610 }, { "grad_norm": 1.4387259483337402, "learning_rate": 5.238000000000001e-05, "loss": 0.1169, "step": 2620 }, { "grad_norm": 1.1318260431289673, "learning_rate": 5.258000000000001e-05, "loss": 0.1195, "step": 2630 }, { "grad_norm": 0.9872164130210876, "learning_rate": 5.2780000000000006e-05, "loss": 0.135, "step": 2640 }, { "grad_norm": 1.0824757814407349, "learning_rate": 5.2980000000000004e-05, "loss": 0.1367, "step": 2650 }, { "grad_norm": 1.522147297859192, "learning_rate": 5.318000000000001e-05, "loss": 0.1418, "step": 2660 }, { "grad_norm": 1.240170955657959, "learning_rate": 5.338000000000001e-05, "loss": 0.1408, "step": 2670 }, { "grad_norm": 1.224696397781372, "learning_rate": 5.3580000000000005e-05, "loss": 0.1464, "step": 2680 }, { "grad_norm": 1.114503264427185, "learning_rate": 5.378e-05, "loss": 0.1212, "step": 2690 }, { "grad_norm": 1.0838764905929565, "learning_rate": 5.3979999999999995e-05, "loss": 0.1103, "step": 2700 }, { "grad_norm": 1.1536508798599243, "learning_rate": 5.418e-05, "loss": 0.1309, "step": 2710 }, { "grad_norm": 1.006576418876648, "learning_rate": 5.438e-05, "loss": 0.1488, "step": 2720 }, { "grad_norm": 1.208191156387329, "learning_rate": 5.4579999999999996e-05, "loss": 0.1092, "step": 2730 }, { "grad_norm": 0.8975704312324524, "learning_rate": 5.478e-05, "loss": 0.1094, "step": 2740 }, { "grad_norm": 1.1106466054916382, "learning_rate": 5.498e-05, "loss": 0.1345, "step": 2750 }, { "grad_norm": 1.267430305480957, "learning_rate": 5.518e-05, "loss": 0.121, "step": 2760 }, { "grad_norm": 1.2714762687683105, "learning_rate": 5.538e-05, "loss": 0.1323, "step": 2770 }, { "grad_norm": 0.7652685046195984, "learning_rate": 5.558e-05, "loss": 0.1381, "step": 2780 }, { "grad_norm": 1.269576907157898, "learning_rate": 5.578e-05, "loss": 0.1329, "step": 2790 }, { "grad_norm": 1.0322692394256592, "learning_rate": 5.5979999999999996e-05, "loss": 0.1163, "step": 2800 }, { "grad_norm": 1.0551064014434814, "learning_rate": 5.618e-05, "loss": 0.1531, "step": 2810 }, { "grad_norm": 1.107733130455017, "learning_rate": 5.638e-05, "loss": 0.1308, "step": 2820 }, { "grad_norm": 0.8512546420097351, "learning_rate": 5.658e-05, "loss": 0.1118, "step": 2830 }, { "grad_norm": 1.2256578207015991, "learning_rate": 5.678e-05, "loss": 0.1326, "step": 2840 }, { "grad_norm": 1.0716038942337036, "learning_rate": 5.698e-05, "loss": 0.134, "step": 2850 }, { "grad_norm": 0.847619354724884, "learning_rate": 5.718e-05, "loss": 0.1147, "step": 2860 }, { "grad_norm": 0.8640828728675842, "learning_rate": 5.738e-05, "loss": 0.1219, "step": 2870 }, { "grad_norm": 0.9737288355827332, "learning_rate": 5.758e-05, "loss": 0.1098, "step": 2880 }, { "grad_norm": 0.8534566164016724, "learning_rate": 5.778e-05, "loss": 0.1156, "step": 2890 }, { "grad_norm": 0.9717745184898376, "learning_rate": 5.7980000000000004e-05, "loss": 0.1276, "step": 2900 }, { "grad_norm": 1.3601703643798828, "learning_rate": 5.818e-05, "loss": 0.1101, "step": 2910 }, { "grad_norm": 1.0254127979278564, "learning_rate": 5.838e-05, "loss": 0.1233, "step": 2920 }, { "grad_norm": 1.1259608268737793, "learning_rate": 5.858e-05, "loss": 0.0999, "step": 2930 }, { "grad_norm": 0.7405943870544434, "learning_rate": 5.878e-05, "loss": 0.1039, "step": 2940 }, { "grad_norm": 1.2386385202407837, "learning_rate": 5.898e-05, "loss": 0.1122, "step": 2950 }, { "grad_norm": 0.934509813785553, "learning_rate": 5.918e-05, "loss": 0.1469, "step": 2960 }, { "grad_norm": 1.3294734954833984, "learning_rate": 5.9380000000000004e-05, "loss": 0.1214, "step": 2970 }, { "grad_norm": 1.1562579870224, "learning_rate": 5.958e-05, "loss": 0.132, "step": 2980 }, { "grad_norm": 1.0790714025497437, "learning_rate": 5.978e-05, "loss": 0.1472, "step": 2990 }, { "grad_norm": 0.7977883815765381, "learning_rate": 5.9980000000000005e-05, "loss": 0.1226, "step": 3000 }, { "grad_norm": 0.9226921200752258, "learning_rate": 6.018e-05, "loss": 0.1053, "step": 3010 }, { "grad_norm": 0.8704375624656677, "learning_rate": 6.038e-05, "loss": 0.1147, "step": 3020 }, { "grad_norm": 1.1094822883605957, "learning_rate": 6.0580000000000006e-05, "loss": 0.107, "step": 3030 }, { "grad_norm": 0.6657067537307739, "learning_rate": 6.0780000000000004e-05, "loss": 0.1068, "step": 3040 }, { "grad_norm": 0.7606215476989746, "learning_rate": 6.098e-05, "loss": 0.1215, "step": 3050 }, { "grad_norm": 0.9039419889450073, "learning_rate": 6.118000000000001e-05, "loss": 0.1266, "step": 3060 }, { "grad_norm": 1.0196330547332764, "learning_rate": 6.138e-05, "loss": 0.102, "step": 3070 }, { "grad_norm": 0.8795500993728638, "learning_rate": 6.158e-05, "loss": 0.1241, "step": 3080 }, { "grad_norm": 1.1170536279678345, "learning_rate": 6.178000000000001e-05, "loss": 0.1338, "step": 3090 }, { "grad_norm": 1.0536253452301025, "learning_rate": 6.198e-05, "loss": 0.1037, "step": 3100 }, { "grad_norm": 1.0507588386535645, "learning_rate": 6.218e-05, "loss": 0.1055, "step": 3110 }, { "grad_norm": 0.9194474816322327, "learning_rate": 6.238000000000001e-05, "loss": 0.1069, "step": 3120 }, { "grad_norm": 1.113611102104187, "learning_rate": 6.258e-05, "loss": 0.1209, "step": 3130 }, { "grad_norm": 1.1744225025177002, "learning_rate": 6.278e-05, "loss": 0.1201, "step": 3140 }, { "grad_norm": 1.157328486442566, "learning_rate": 6.298000000000001e-05, "loss": 0.1161, "step": 3150 }, { "grad_norm": 0.9268437623977661, "learning_rate": 6.318e-05, "loss": 0.1256, "step": 3160 }, { "grad_norm": 0.9406293034553528, "learning_rate": 6.338e-05, "loss": 0.1074, "step": 3170 }, { "grad_norm": 0.8025528192520142, "learning_rate": 6.358000000000001e-05, "loss": 0.134, "step": 3180 }, { "grad_norm": 1.1105653047561646, "learning_rate": 6.378e-05, "loss": 0.1255, "step": 3190 }, { "grad_norm": 0.8199275732040405, "learning_rate": 6.398000000000001e-05, "loss": 0.1129, "step": 3200 }, { "grad_norm": 0.8296595215797424, "learning_rate": 6.418000000000001e-05, "loss": 0.1064, "step": 3210 }, { "grad_norm": 1.246260166168213, "learning_rate": 6.438e-05, "loss": 0.1568, "step": 3220 }, { "grad_norm": 1.0188353061676025, "learning_rate": 6.458000000000001e-05, "loss": 0.1337, "step": 3230 }, { "grad_norm": 1.0506720542907715, "learning_rate": 6.478000000000001e-05, "loss": 0.1086, "step": 3240 }, { "grad_norm": 0.968858540058136, "learning_rate": 6.498e-05, "loss": 0.1258, "step": 3250 }, { "grad_norm": 1.152908444404602, "learning_rate": 6.518000000000001e-05, "loss": 0.1255, "step": 3260 }, { "grad_norm": 1.50462007522583, "learning_rate": 6.538000000000001e-05, "loss": 0.1196, "step": 3270 }, { "grad_norm": 0.9636582732200623, "learning_rate": 6.558e-05, "loss": 0.1115, "step": 3280 }, { "grad_norm": 0.6987828612327576, "learning_rate": 6.578000000000001e-05, "loss": 0.1142, "step": 3290 }, { "grad_norm": 0.7157345414161682, "learning_rate": 6.598e-05, "loss": 0.1022, "step": 3300 }, { "grad_norm": 0.7561734318733215, "learning_rate": 6.618e-05, "loss": 0.1131, "step": 3310 }, { "grad_norm": 0.7367609739303589, "learning_rate": 6.638e-05, "loss": 0.1284, "step": 3320 }, { "grad_norm": 0.9949061274528503, "learning_rate": 6.658e-05, "loss": 0.107, "step": 3330 }, { "grad_norm": 0.9798692464828491, "learning_rate": 6.678e-05, "loss": 0.1091, "step": 3340 }, { "grad_norm": 0.7761991620063782, "learning_rate": 6.698e-05, "loss": 0.1113, "step": 3350 }, { "grad_norm": 1.0490413904190063, "learning_rate": 6.718e-05, "loss": 0.108, "step": 3360 }, { "grad_norm": 0.8816936612129211, "learning_rate": 6.738e-05, "loss": 0.1188, "step": 3370 }, { "grad_norm": 1.0105409622192383, "learning_rate": 6.758e-05, "loss": 0.1513, "step": 3380 }, { "grad_norm": 0.7741963267326355, "learning_rate": 6.778e-05, "loss": 0.112, "step": 3390 }, { "grad_norm": 0.9761982560157776, "learning_rate": 6.798e-05, "loss": 0.1209, "step": 3400 }, { "grad_norm": 0.9732022285461426, "learning_rate": 6.818e-05, "loss": 0.1066, "step": 3410 }, { "grad_norm": 0.9676787853240967, "learning_rate": 6.838e-05, "loss": 0.1207, "step": 3420 }, { "grad_norm": 0.779762864112854, "learning_rate": 6.858e-05, "loss": 0.105, "step": 3430 }, { "grad_norm": 0.6680462956428528, "learning_rate": 6.878e-05, "loss": 0.1263, "step": 3440 }, { "grad_norm": 1.0314974784851074, "learning_rate": 6.898e-05, "loss": 0.1124, "step": 3450 }, { "grad_norm": 0.6555045247077942, "learning_rate": 6.918e-05, "loss": 0.096, "step": 3460 }, { "grad_norm": 0.822640061378479, "learning_rate": 6.938e-05, "loss": 0.1306, "step": 3470 }, { "grad_norm": 0.8375845551490784, "learning_rate": 6.958e-05, "loss": 0.1139, "step": 3480 }, { "grad_norm": 0.8774402737617493, "learning_rate": 6.978e-05, "loss": 0.1091, "step": 3490 }, { "grad_norm": 0.8289646506309509, "learning_rate": 6.998e-05, "loss": 0.1008, "step": 3500 }, { "grad_norm": 0.7811492681503296, "learning_rate": 7.018e-05, "loss": 0.1123, "step": 3510 }, { "grad_norm": 1.0667214393615723, "learning_rate": 7.038e-05, "loss": 0.1288, "step": 3520 }, { "grad_norm": 1.027891993522644, "learning_rate": 7.058e-05, "loss": 0.1106, "step": 3530 }, { "grad_norm": 0.9904118180274963, "learning_rate": 7.078e-05, "loss": 0.1032, "step": 3540 }, { "grad_norm": 0.7684602737426758, "learning_rate": 7.098e-05, "loss": 0.1141, "step": 3550 }, { "grad_norm": 0.7688536643981934, "learning_rate": 7.118e-05, "loss": 0.0973, "step": 3560 }, { "grad_norm": 0.6606555581092834, "learning_rate": 7.138e-05, "loss": 0.1235, "step": 3570 }, { "grad_norm": 0.7975412011146545, "learning_rate": 7.158e-05, "loss": 0.1242, "step": 3580 }, { "grad_norm": 0.7788171172142029, "learning_rate": 7.178000000000001e-05, "loss": 0.1217, "step": 3590 }, { "grad_norm": 0.829694390296936, "learning_rate": 7.198e-05, "loss": 0.1058, "step": 3600 }, { "grad_norm": 0.9177959561347961, "learning_rate": 7.218e-05, "loss": 0.1277, "step": 3610 }, { "grad_norm": 0.7017259001731873, "learning_rate": 7.238000000000001e-05, "loss": 0.1037, "step": 3620 }, { "grad_norm": 0.9037215113639832, "learning_rate": 7.258e-05, "loss": 0.0927, "step": 3630 }, { "grad_norm": 0.5603802800178528, "learning_rate": 7.278e-05, "loss": 0.0946, "step": 3640 }, { "grad_norm": 0.8582494258880615, "learning_rate": 7.298000000000001e-05, "loss": 0.1255, "step": 3650 }, { "grad_norm": 1.6430299282073975, "learning_rate": 7.318e-05, "loss": 0.1141, "step": 3660 }, { "grad_norm": 1.4471076726913452, "learning_rate": 7.338e-05, "loss": 0.1425, "step": 3670 }, { "grad_norm": 0.9581396579742432, "learning_rate": 7.358000000000001e-05, "loss": 0.1279, "step": 3680 }, { "grad_norm": 0.8626473546028137, "learning_rate": 7.378e-05, "loss": 0.1103, "step": 3690 }, { "grad_norm": 0.9219233393669128, "learning_rate": 7.398e-05, "loss": 0.1068, "step": 3700 }, { "grad_norm": 0.7603467106819153, "learning_rate": 7.418000000000001e-05, "loss": 0.1159, "step": 3710 }, { "grad_norm": 0.695432186126709, "learning_rate": 7.438e-05, "loss": 0.1021, "step": 3720 }, { "grad_norm": 0.8505820631980896, "learning_rate": 7.458000000000001e-05, "loss": 0.0942, "step": 3730 }, { "grad_norm": 1.0521332025527954, "learning_rate": 7.478e-05, "loss": 0.0992, "step": 3740 }, { "grad_norm": 0.8170650601387024, "learning_rate": 7.498e-05, "loss": 0.1117, "step": 3750 }, { "grad_norm": 0.7179012298583984, "learning_rate": 7.518000000000001e-05, "loss": 0.0956, "step": 3760 }, { "grad_norm": 0.8247060775756836, "learning_rate": 7.538e-05, "loss": 0.1102, "step": 3770 }, { "grad_norm": 0.9358729124069214, "learning_rate": 7.558e-05, "loss": 0.1077, "step": 3780 }, { "grad_norm": 0.8919908404350281, "learning_rate": 7.578000000000001e-05, "loss": 0.1141, "step": 3790 }, { "grad_norm": 0.7915765643119812, "learning_rate": 7.598e-05, "loss": 0.1484, "step": 3800 }, { "grad_norm": 0.9018880128860474, "learning_rate": 7.618e-05, "loss": 0.1158, "step": 3810 }, { "grad_norm": 1.014735221862793, "learning_rate": 7.638000000000001e-05, "loss": 0.1184, "step": 3820 }, { "grad_norm": 1.1713080406188965, "learning_rate": 7.658e-05, "loss": 0.1362, "step": 3830 }, { "grad_norm": 0.8115696310997009, "learning_rate": 7.678000000000001e-05, "loss": 0.1182, "step": 3840 }, { "grad_norm": 0.9017682671546936, "learning_rate": 7.698000000000001e-05, "loss": 0.1237, "step": 3850 }, { "grad_norm": 1.152012825012207, "learning_rate": 7.718e-05, "loss": 0.0982, "step": 3860 }, { "grad_norm": 0.908958375453949, "learning_rate": 7.738000000000001e-05, "loss": 0.1164, "step": 3870 }, { "grad_norm": 0.7509891390800476, "learning_rate": 7.758000000000001e-05, "loss": 0.1091, "step": 3880 }, { "grad_norm": 0.9932035207748413, "learning_rate": 7.778e-05, "loss": 0.117, "step": 3890 }, { "grad_norm": 0.9337891340255737, "learning_rate": 7.798000000000001e-05, "loss": 0.1072, "step": 3900 }, { "grad_norm": 0.8214189410209656, "learning_rate": 7.818000000000001e-05, "loss": 0.1085, "step": 3910 }, { "grad_norm": 1.1743890047073364, "learning_rate": 7.838e-05, "loss": 0.1198, "step": 3920 }, { "grad_norm": 0.7139511108398438, "learning_rate": 7.858000000000001e-05, "loss": 0.104, "step": 3930 }, { "grad_norm": 0.6941994428634644, "learning_rate": 7.878e-05, "loss": 0.1305, "step": 3940 }, { "grad_norm": 0.8458877205848694, "learning_rate": 7.897999999999999e-05, "loss": 0.1162, "step": 3950 }, { "grad_norm": 0.8671753406524658, "learning_rate": 7.918e-05, "loss": 0.1276, "step": 3960 }, { "grad_norm": 0.8299127221107483, "learning_rate": 7.938e-05, "loss": 0.1066, "step": 3970 }, { "grad_norm": 1.007913589477539, "learning_rate": 7.958e-05, "loss": 0.1317, "step": 3980 }, { "grad_norm": 1.11934494972229, "learning_rate": 7.978e-05, "loss": 0.132, "step": 3990 }, { "grad_norm": 1.0316047668457031, "learning_rate": 7.998e-05, "loss": 0.1031, "step": 4000 }, { "grad_norm": 1.0056008100509644, "learning_rate": 8.018e-05, "loss": 0.1127, "step": 4010 }, { "grad_norm": 0.6722298264503479, "learning_rate": 8.038e-05, "loss": 0.1024, "step": 4020 }, { "grad_norm": 0.7355875372886658, "learning_rate": 8.058e-05, "loss": 0.1221, "step": 4030 }, { "grad_norm": 1.009796380996704, "learning_rate": 8.078e-05, "loss": 0.1037, "step": 4040 }, { "grad_norm": 0.5862364172935486, "learning_rate": 8.098e-05, "loss": 0.1064, "step": 4050 }, { "grad_norm": 0.7323852181434631, "learning_rate": 8.118e-05, "loss": 0.116, "step": 4060 }, { "grad_norm": 0.8675004839897156, "learning_rate": 8.138e-05, "loss": 0.102, "step": 4070 }, { "grad_norm": 0.7603096961975098, "learning_rate": 8.158e-05, "loss": 0.1133, "step": 4080 }, { "grad_norm": 0.983374297618866, "learning_rate": 8.178e-05, "loss": 0.1117, "step": 4090 }, { "grad_norm": 0.8899638652801514, "learning_rate": 8.198e-05, "loss": 0.1139, "step": 4100 }, { "grad_norm": 0.6955955624580383, "learning_rate": 8.218e-05, "loss": 0.1229, "step": 4110 }, { "grad_norm": 0.8813236355781555, "learning_rate": 8.238000000000001e-05, "loss": 0.1195, "step": 4120 }, { "grad_norm": 0.8934229612350464, "learning_rate": 8.258e-05, "loss": 0.1095, "step": 4130 }, { "grad_norm": 0.9318566918373108, "learning_rate": 8.278e-05, "loss": 0.1149, "step": 4140 }, { "grad_norm": 0.7634225487709045, "learning_rate": 8.298000000000001e-05, "loss": 0.1305, "step": 4150 }, { "grad_norm": 0.6699257493019104, "learning_rate": 8.318e-05, "loss": 0.0921, "step": 4160 }, { "grad_norm": 0.719491720199585, "learning_rate": 8.338e-05, "loss": 0.096, "step": 4170 }, { "grad_norm": 0.5478181838989258, "learning_rate": 8.358e-05, "loss": 0.0896, "step": 4180 }, { "grad_norm": 1.0340334177017212, "learning_rate": 8.378e-05, "loss": 0.1002, "step": 4190 }, { "grad_norm": 0.6812309622764587, "learning_rate": 8.398e-05, "loss": 0.1059, "step": 4200 }, { "grad_norm": 0.8953475952148438, "learning_rate": 8.418e-05, "loss": 0.1099, "step": 4210 }, { "grad_norm": 1.2144333124160767, "learning_rate": 8.438e-05, "loss": 0.1242, "step": 4220 }, { "grad_norm": 1.0883917808532715, "learning_rate": 8.458e-05, "loss": 0.1515, "step": 4230 }, { "grad_norm": 0.7937644124031067, "learning_rate": 8.478e-05, "loss": 0.1143, "step": 4240 }, { "grad_norm": 0.7586570382118225, "learning_rate": 8.498e-05, "loss": 0.1199, "step": 4250 }, { "grad_norm": 0.7288032174110413, "learning_rate": 8.518000000000001e-05, "loss": 0.1272, "step": 4260 }, { "grad_norm": 0.6654855012893677, "learning_rate": 8.538e-05, "loss": 0.115, "step": 4270 }, { "grad_norm": 0.8437451720237732, "learning_rate": 8.558e-05, "loss": 0.0959, "step": 4280 }, { "grad_norm": 1.1201205253601074, "learning_rate": 8.578000000000001e-05, "loss": 0.1088, "step": 4290 }, { "grad_norm": 0.779449462890625, "learning_rate": 8.598e-05, "loss": 0.0965, "step": 4300 }, { "grad_norm": 0.8934101462364197, "learning_rate": 8.618e-05, "loss": 0.1176, "step": 4310 }, { "grad_norm": 1.0906873941421509, "learning_rate": 8.638000000000001e-05, "loss": 0.1213, "step": 4320 }, { "grad_norm": 0.7026787996292114, "learning_rate": 8.658e-05, "loss": 0.102, "step": 4330 }, { "grad_norm": 0.7005046606063843, "learning_rate": 8.678e-05, "loss": 0.0971, "step": 4340 }, { "grad_norm": 0.8272811770439148, "learning_rate": 8.698000000000001e-05, "loss": 0.1077, "step": 4350 }, { "grad_norm": 0.7187658548355103, "learning_rate": 8.718e-05, "loss": 0.0949, "step": 4360 }, { "grad_norm": 1.0183064937591553, "learning_rate": 8.738000000000001e-05, "loss": 0.1288, "step": 4370 }, { "grad_norm": 0.684727668762207, "learning_rate": 8.758000000000001e-05, "loss": 0.0944, "step": 4380 }, { "grad_norm": 0.5910325050354004, "learning_rate": 8.778e-05, "loss": 0.1003, "step": 4390 }, { "grad_norm": 0.7680476903915405, "learning_rate": 8.798000000000001e-05, "loss": 0.1336, "step": 4400 }, { "grad_norm": 0.7688000798225403, "learning_rate": 8.818000000000001e-05, "loss": 0.1067, "step": 4410 }, { "grad_norm": 0.6766910552978516, "learning_rate": 8.838e-05, "loss": 0.131, "step": 4420 }, { "grad_norm": 0.7121308445930481, "learning_rate": 8.858000000000001e-05, "loss": 0.1029, "step": 4430 }, { "grad_norm": 0.7657518982887268, "learning_rate": 8.878000000000001e-05, "loss": 0.1026, "step": 4440 }, { "grad_norm": 0.6849581599235535, "learning_rate": 8.898e-05, "loss": 0.12, "step": 4450 }, { "grad_norm": 0.7673609852790833, "learning_rate": 8.918000000000001e-05, "loss": 0.1084, "step": 4460 }, { "grad_norm": 0.9891780018806458, "learning_rate": 8.938e-05, "loss": 0.0999, "step": 4470 }, { "grad_norm": 0.6689172387123108, "learning_rate": 8.958e-05, "loss": 0.1262, "step": 4480 }, { "grad_norm": 0.6024765372276306, "learning_rate": 8.978000000000001e-05, "loss": 0.1055, "step": 4490 }, { "grad_norm": 0.7818013429641724, "learning_rate": 8.998e-05, "loss": 0.1106, "step": 4500 }, { "grad_norm": 0.8750324249267578, "learning_rate": 9.018000000000001e-05, "loss": 0.1317, "step": 4510 }, { "grad_norm": 0.9395912885665894, "learning_rate": 9.038000000000001e-05, "loss": 0.1138, "step": 4520 }, { "grad_norm": 0.6798073053359985, "learning_rate": 9.058e-05, "loss": 0.1022, "step": 4530 }, { "grad_norm": 0.9360458254814148, "learning_rate": 9.078000000000001e-05, "loss": 0.1244, "step": 4540 }, { "grad_norm": 0.735204815864563, "learning_rate": 9.098000000000001e-05, "loss": 0.1368, "step": 4550 }, { "grad_norm": 0.630633533000946, "learning_rate": 9.118e-05, "loss": 0.1041, "step": 4560 }, { "grad_norm": 0.7147504687309265, "learning_rate": 9.138e-05, "loss": 0.1204, "step": 4570 }, { "grad_norm": 1.1372507810592651, "learning_rate": 9.158e-05, "loss": 0.1235, "step": 4580 }, { "grad_norm": 0.8042636513710022, "learning_rate": 9.178e-05, "loss": 0.1081, "step": 4590 }, { "grad_norm": 0.9635946154594421, "learning_rate": 9.198e-05, "loss": 0.1118, "step": 4600 }, { "grad_norm": 0.9054326415061951, "learning_rate": 9.218e-05, "loss": 0.1144, "step": 4610 }, { "grad_norm": 1.0424317121505737, "learning_rate": 9.238e-05, "loss": 0.0964, "step": 4620 }, { "grad_norm": 0.9291507005691528, "learning_rate": 9.258e-05, "loss": 0.1108, "step": 4630 }, { "grad_norm": 0.6265617609024048, "learning_rate": 9.278e-05, "loss": 0.1186, "step": 4640 }, { "grad_norm": 0.6758886575698853, "learning_rate": 9.298e-05, "loss": 0.0941, "step": 4650 }, { "grad_norm": 0.6008509397506714, "learning_rate": 9.318e-05, "loss": 0.0953, "step": 4660 }, { "grad_norm": 0.938416600227356, "learning_rate": 9.338e-05, "loss": 0.1219, "step": 4670 }, { "grad_norm": 0.8265025019645691, "learning_rate": 9.358e-05, "loss": 0.1116, "step": 4680 }, { "grad_norm": 0.6957197785377502, "learning_rate": 9.378e-05, "loss": 0.1298, "step": 4690 }, { "grad_norm": 0.6880128383636475, "learning_rate": 9.398e-05, "loss": 0.1081, "step": 4700 }, { "grad_norm": 0.6141018271446228, "learning_rate": 9.418e-05, "loss": 0.1134, "step": 4710 }, { "grad_norm": 0.821724534034729, "learning_rate": 9.438e-05, "loss": 0.0982, "step": 4720 }, { "grad_norm": 0.7338222861289978, "learning_rate": 9.458e-05, "loss": 0.1246, "step": 4730 }, { "grad_norm": 0.8533934354782104, "learning_rate": 9.478e-05, "loss": 0.1145, "step": 4740 }, { "grad_norm": 0.5947041511535645, "learning_rate": 9.498e-05, "loss": 0.0993, "step": 4750 }, { "grad_norm": 0.786138653755188, "learning_rate": 9.518000000000001e-05, "loss": 0.11, "step": 4760 }, { "grad_norm": 0.8529260158538818, "learning_rate": 9.538e-05, "loss": 0.1143, "step": 4770 }, { "grad_norm": 0.9582460522651672, "learning_rate": 9.558e-05, "loss": 0.1323, "step": 4780 }, { "grad_norm": 0.8673290610313416, "learning_rate": 9.578000000000001e-05, "loss": 0.1003, "step": 4790 }, { "grad_norm": 0.7391735315322876, "learning_rate": 9.598e-05, "loss": 0.1065, "step": 4800 }, { "grad_norm": 0.7067121267318726, "learning_rate": 9.618e-05, "loss": 0.1009, "step": 4810 }, { "grad_norm": 0.7269254326820374, "learning_rate": 9.638000000000001e-05, "loss": 0.1064, "step": 4820 }, { "grad_norm": 0.7221915125846863, "learning_rate": 9.658e-05, "loss": 0.1118, "step": 4830 }, { "grad_norm": 0.7598391175270081, "learning_rate": 9.678e-05, "loss": 0.1206, "step": 4840 }, { "grad_norm": 0.8106645941734314, "learning_rate": 9.698000000000001e-05, "loss": 0.1122, "step": 4850 }, { "grad_norm": 0.9392073750495911, "learning_rate": 9.718e-05, "loss": 0.1031, "step": 4860 }, { "grad_norm": 0.6523623466491699, "learning_rate": 9.738e-05, "loss": 0.1061, "step": 4870 }, { "grad_norm": 1.0489426851272583, "learning_rate": 9.758000000000001e-05, "loss": 0.1045, "step": 4880 }, { "grad_norm": 1.010888695716858, "learning_rate": 9.778e-05, "loss": 0.1313, "step": 4890 }, { "grad_norm": 0.7509266138076782, "learning_rate": 9.798000000000001e-05, "loss": 0.1057, "step": 4900 }, { "grad_norm": 0.8605509400367737, "learning_rate": 9.818000000000001e-05, "loss": 0.1102, "step": 4910 }, { "grad_norm": 0.7096688747406006, "learning_rate": 9.838e-05, "loss": 0.1146, "step": 4920 }, { "grad_norm": 1.1225011348724365, "learning_rate": 9.858000000000001e-05, "loss": 0.1233, "step": 4930 }, { "grad_norm": 0.7138217091560364, "learning_rate": 9.878e-05, "loss": 0.116, "step": 4940 }, { "grad_norm": 0.7680518627166748, "learning_rate": 9.898e-05, "loss": 0.1235, "step": 4950 }, { "grad_norm": 0.6603628993034363, "learning_rate": 9.918000000000001e-05, "loss": 0.1166, "step": 4960 }, { "grad_norm": 0.7387930750846863, "learning_rate": 9.938e-05, "loss": 0.1219, "step": 4970 }, { "grad_norm": 0.6730515360832214, "learning_rate": 9.958e-05, "loss": 0.1214, "step": 4980 }, { "grad_norm": 0.6425665616989136, "learning_rate": 9.978000000000001e-05, "loss": 0.1133, "step": 4990 }, { "grad_norm": 0.6319780945777893, "learning_rate": 9.998e-05, "loss": 0.1294, "step": 5000 }, { "grad_norm": 0.8604909181594849, "learning_rate": 9.999999778549045e-05, "loss": 0.1228, "step": 5010 }, { "grad_norm": 0.8282168507575989, "learning_rate": 9.999999013039593e-05, "loss": 0.1199, "step": 5020 }, { "grad_norm": 0.9022985696792603, "learning_rate": 9.999997700737766e-05, "loss": 0.1218, "step": 5030 }, { "grad_norm": 0.5979546308517456, "learning_rate": 9.999995841643709e-05, "loss": 0.105, "step": 5040 }, { "grad_norm": 0.9257036447525024, "learning_rate": 9.999993435757623e-05, "loss": 0.1195, "step": 5050 }, { "grad_norm": 0.797887921333313, "learning_rate": 9.999990483079773e-05, "loss": 0.1095, "step": 5060 }, { "grad_norm": 0.7497470378875732, "learning_rate": 9.999986983610481e-05, "loss": 0.1364, "step": 5070 }, { "grad_norm": 1.2889230251312256, "learning_rate": 9.99998293735013e-05, "loss": 0.1265, "step": 5080 }, { "grad_norm": 1.020838975906372, "learning_rate": 9.999978344299161e-05, "loss": 0.1187, "step": 5090 }, { "grad_norm": 1.3007416725158691, "learning_rate": 9.99997320445808e-05, "loss": 0.1285, "step": 5100 }, { "grad_norm": 0.9498431086540222, "learning_rate": 9.999967517827444e-05, "loss": 0.1244, "step": 5110 }, { "grad_norm": 0.8019992113113403, "learning_rate": 9.999961284407879e-05, "loss": 0.1388, "step": 5120 }, { "grad_norm": 0.6894762516021729, "learning_rate": 9.999954504200067e-05, "loss": 0.1224, "step": 5130 }, { "grad_norm": 0.7588593363761902, "learning_rate": 9.999947177204744e-05, "loss": 0.108, "step": 5140 }, { "grad_norm": 0.5550956726074219, "learning_rate": 9.999939303422718e-05, "loss": 0.0868, "step": 5150 }, { "grad_norm": 0.9398595094680786, "learning_rate": 9.999930882854847e-05, "loss": 0.1187, "step": 5160 }, { "grad_norm": 0.7749357223510742, "learning_rate": 9.999921915502051e-05, "loss": 0.1167, "step": 5170 }, { "grad_norm": 0.5549615025520325, "learning_rate": 9.99991240136531e-05, "loss": 0.1193, "step": 5180 }, { "grad_norm": 0.7127315998077393, "learning_rate": 9.999902340445668e-05, "loss": 0.1408, "step": 5190 }, { "grad_norm": 0.6758066415786743, "learning_rate": 9.999891732744224e-05, "loss": 0.1014, "step": 5200 }, { "grad_norm": 0.5796706676483154, "learning_rate": 9.999880578262135e-05, "loss": 0.1118, "step": 5210 }, { "grad_norm": 0.7226642966270447, "learning_rate": 9.999868877000624e-05, "loss": 0.1103, "step": 5220 }, { "grad_norm": 0.46086716651916504, "learning_rate": 9.99985662896097e-05, "loss": 0.1005, "step": 5230 }, { "grad_norm": 0.7970030903816223, "learning_rate": 9.999843834144513e-05, "loss": 0.1045, "step": 5240 }, { "grad_norm": 0.7538625001907349, "learning_rate": 9.99983049255265e-05, "loss": 0.1219, "step": 5250 }, { "grad_norm": 0.9820960164070129, "learning_rate": 9.999816604186843e-05, "loss": 0.1151, "step": 5260 }, { "grad_norm": 0.8765256404876709, "learning_rate": 9.999802169048609e-05, "loss": 0.1069, "step": 5270 }, { "grad_norm": 0.7082048058509827, "learning_rate": 9.999787187139527e-05, "loss": 0.1008, "step": 5280 }, { "grad_norm": 0.6925408244132996, "learning_rate": 9.999771658461234e-05, "loss": 0.104, "step": 5290 }, { "grad_norm": 0.6882164478302002, "learning_rate": 9.999755583015431e-05, "loss": 0.1035, "step": 5300 }, { "grad_norm": 0.9490835070610046, "learning_rate": 9.999738960803874e-05, "loss": 0.0968, "step": 5310 }, { "grad_norm": 0.6714410781860352, "learning_rate": 9.99972179182838e-05, "loss": 0.1097, "step": 5320 }, { "grad_norm": 0.8686289191246033, "learning_rate": 9.99970407609083e-05, "loss": 0.1263, "step": 5330 }, { "grad_norm": 0.8806028366088867, "learning_rate": 9.999685813593159e-05, "loss": 0.1029, "step": 5340 }, { "grad_norm": 0.47920846939086914, "learning_rate": 9.999667004337362e-05, "loss": 0.0932, "step": 5350 }, { "grad_norm": 1.119980812072754, "learning_rate": 9.9996476483255e-05, "loss": 0.1184, "step": 5360 }, { "grad_norm": 0.6605285406112671, "learning_rate": 9.999627745559688e-05, "loss": 0.0927, "step": 5370 }, { "grad_norm": 0.7120414972305298, "learning_rate": 9.999607296042101e-05, "loss": 0.0957, "step": 5380 }, { "grad_norm": 0.524343729019165, "learning_rate": 9.99958629977498e-05, "loss": 0.0927, "step": 5390 }, { "grad_norm": 0.7755628228187561, "learning_rate": 9.999564756760615e-05, "loss": 0.0972, "step": 5400 }, { "grad_norm": 0.7669888138771057, "learning_rate": 9.999542667001366e-05, "loss": 0.0894, "step": 5410 }, { "grad_norm": 0.708515465259552, "learning_rate": 9.999520030499647e-05, "loss": 0.1013, "step": 5420 }, { "grad_norm": 0.8446861505508423, "learning_rate": 9.999496847257936e-05, "loss": 0.0861, "step": 5430 }, { "grad_norm": 0.7299346327781677, "learning_rate": 9.999473117278764e-05, "loss": 0.1155, "step": 5440 }, { "grad_norm": 0.7009485960006714, "learning_rate": 9.999448840564731e-05, "loss": 0.0891, "step": 5450 }, { "grad_norm": 0.6754295825958252, "learning_rate": 9.999424017118488e-05, "loss": 0.1152, "step": 5460 }, { "grad_norm": 0.7163801789283752, "learning_rate": 9.999398646942751e-05, "loss": 0.0921, "step": 5470 }, { "grad_norm": 0.8402428030967712, "learning_rate": 9.999372730040296e-05, "loss": 0.096, "step": 5480 }, { "grad_norm": 0.6390902996063232, "learning_rate": 9.999346266413953e-05, "loss": 0.1125, "step": 5490 }, { "grad_norm": 0.7182981967926025, "learning_rate": 9.99931925606662e-05, "loss": 0.113, "step": 5500 }, { "grad_norm": 0.5521588325500488, "learning_rate": 9.99929169900125e-05, "loss": 0.1064, "step": 5510 }, { "grad_norm": 0.7146857380867004, "learning_rate": 9.999263595220855e-05, "loss": 0.086, "step": 5520 }, { "grad_norm": 0.8081373572349548, "learning_rate": 9.99923494472851e-05, "loss": 0.0975, "step": 5530 }, { "grad_norm": 0.7025677561759949, "learning_rate": 9.999205747527348e-05, "loss": 0.0922, "step": 5540 }, { "grad_norm": 0.6586804986000061, "learning_rate": 9.999176003620561e-05, "loss": 0.0996, "step": 5550 }, { "grad_norm": 1.389208436012268, "learning_rate": 9.999145713011405e-05, "loss": 0.1021, "step": 5560 }, { "grad_norm": 0.8332874774932861, "learning_rate": 9.999114875703186e-05, "loss": 0.1142, "step": 5570 }, { "grad_norm": 0.8978160619735718, "learning_rate": 9.999083491699281e-05, "loss": 0.1231, "step": 5580 }, { "grad_norm": 0.7711489200592041, "learning_rate": 9.999051561003123e-05, "loss": 0.1133, "step": 5590 }, { "grad_norm": 0.7761319279670715, "learning_rate": 9.999019083618202e-05, "loss": 0.1128, "step": 5600 }, { "grad_norm": 0.8201305866241455, "learning_rate": 9.99898605954807e-05, "loss": 0.1103, "step": 5610 }, { "grad_norm": 0.6711265444755554, "learning_rate": 9.998952488796338e-05, "loss": 0.099, "step": 5620 }, { "grad_norm": 1.3161933422088623, "learning_rate": 9.998918371366676e-05, "loss": 0.0951, "step": 5630 }, { "grad_norm": 0.6727063655853271, "learning_rate": 9.99888370726282e-05, "loss": 0.0977, "step": 5640 }, { "grad_norm": 0.8321560621261597, "learning_rate": 9.998848496488556e-05, "loss": 0.1351, "step": 5650 }, { "grad_norm": 0.8548291921615601, "learning_rate": 9.998812739047736e-05, "loss": 0.1011, "step": 5660 }, { "grad_norm": 0.6506791710853577, "learning_rate": 9.99877643494427e-05, "loss": 0.1019, "step": 5670 }, { "grad_norm": 0.7242605686187744, "learning_rate": 9.998739584182128e-05, "loss": 0.1165, "step": 5680 }, { "grad_norm": 0.8968840837478638, "learning_rate": 9.998702186765342e-05, "loss": 0.096, "step": 5690 }, { "grad_norm": 0.5357623100280762, "learning_rate": 9.998664242698e-05, "loss": 0.0919, "step": 5700 }, { "grad_norm": 0.6170088052749634, "learning_rate": 9.998625751984251e-05, "loss": 0.1042, "step": 5710 }, { "grad_norm": 0.643439531326294, "learning_rate": 9.998586714628307e-05, "loss": 0.0933, "step": 5720 }, { "grad_norm": 0.5583825707435608, "learning_rate": 9.998547130634432e-05, "loss": 0.111, "step": 5730 }, { "grad_norm": 0.5879292488098145, "learning_rate": 9.99850700000696e-05, "loss": 0.1259, "step": 5740 }, { "grad_norm": 0.4762560725212097, "learning_rate": 9.998466322750278e-05, "loss": 0.1004, "step": 5750 }, { "grad_norm": 0.6041658520698547, "learning_rate": 9.998425098868834e-05, "loss": 0.0936, "step": 5760 }, { "grad_norm": 0.814990758895874, "learning_rate": 9.998383328367136e-05, "loss": 0.1146, "step": 5770 }, { "grad_norm": 0.6164788603782654, "learning_rate": 9.99834101124975e-05, "loss": 0.1146, "step": 5780 }, { "grad_norm": 0.7220621705055237, "learning_rate": 9.998298147521309e-05, "loss": 0.0923, "step": 5790 }, { "grad_norm": 0.9235259890556335, "learning_rate": 9.998254737186496e-05, "loss": 0.1231, "step": 5800 }, { "grad_norm": 1.2647294998168945, "learning_rate": 9.99821078025006e-05, "loss": 0.0981, "step": 5810 }, { "grad_norm": 0.6153743267059326, "learning_rate": 9.998166276716807e-05, "loss": 0.0978, "step": 5820 }, { "grad_norm": 0.7219897508621216, "learning_rate": 9.998121226591606e-05, "loss": 0.096, "step": 5830 }, { "grad_norm": 0.7985761761665344, "learning_rate": 9.998075629879382e-05, "loss": 0.1, "step": 5840 }, { "grad_norm": 0.7574349641799927, "learning_rate": 9.99802948658512e-05, "loss": 0.0872, "step": 5850 }, { "grad_norm": 0.6678763031959534, "learning_rate": 9.99798279671387e-05, "loss": 0.1178, "step": 5860 }, { "grad_norm": 0.7679030299186707, "learning_rate": 9.997935560270734e-05, "loss": 0.095, "step": 5870 }, { "grad_norm": 0.6229738593101501, "learning_rate": 9.997887777260879e-05, "loss": 0.1123, "step": 5880 }, { "grad_norm": 0.5428259968757629, "learning_rate": 9.997839447689532e-05, "loss": 0.1157, "step": 5890 }, { "grad_norm": 0.5553159713745117, "learning_rate": 9.997790571561978e-05, "loss": 0.119, "step": 5900 }, { "grad_norm": 0.5576863288879395, "learning_rate": 9.99774114888356e-05, "loss": 0.1238, "step": 5910 }, { "grad_norm": 0.7290027141571045, "learning_rate": 9.997691179659684e-05, "loss": 0.1109, "step": 5920 }, { "grad_norm": 0.7472510933876038, "learning_rate": 9.997640663895815e-05, "loss": 0.0985, "step": 5930 }, { "grad_norm": 0.670094907283783, "learning_rate": 9.997589601597477e-05, "loss": 0.117, "step": 5940 }, { "grad_norm": 0.8533096313476562, "learning_rate": 9.997537992770252e-05, "loss": 0.0957, "step": 5950 }, { "grad_norm": 0.8324368000030518, "learning_rate": 9.997485837419788e-05, "loss": 0.1088, "step": 5960 }, { "grad_norm": 0.8326006531715393, "learning_rate": 9.997433135551786e-05, "loss": 0.1163, "step": 5970 }, { "grad_norm": 0.7505078911781311, "learning_rate": 9.997379887172009e-05, "loss": 0.1082, "step": 5980 }, { "grad_norm": 0.7591656446456909, "learning_rate": 9.997326092286281e-05, "loss": 0.1141, "step": 5990 }, { "grad_norm": 0.6566003561019897, "learning_rate": 9.997271750900486e-05, "loss": 0.1266, "step": 6000 }, { "grad_norm": 0.770088255405426, "learning_rate": 9.997216863020565e-05, "loss": 0.1285, "step": 6010 }, { "grad_norm": 0.5619138479232788, "learning_rate": 9.99716142865252e-05, "loss": 0.1229, "step": 6020 }, { "grad_norm": 0.7300375699996948, "learning_rate": 9.997105447802415e-05, "loss": 0.1063, "step": 6030 }, { "grad_norm": 0.6042665839195251, "learning_rate": 9.997048920476373e-05, "loss": 0.0909, "step": 6040 }, { "grad_norm": 0.9744282960891724, "learning_rate": 9.996991846680572e-05, "loss": 0.1227, "step": 6050 }, { "grad_norm": 0.5679469704627991, "learning_rate": 9.996934226421257e-05, "loss": 0.1009, "step": 6060 }, { "grad_norm": 0.6971574425697327, "learning_rate": 9.996876059704726e-05, "loss": 0.1128, "step": 6070 }, { "grad_norm": 0.6403180360794067, "learning_rate": 9.996817346537343e-05, "loss": 0.121, "step": 6080 }, { "grad_norm": 0.6733441352844238, "learning_rate": 9.996758086925526e-05, "loss": 0.1079, "step": 6090 }, { "grad_norm": 0.6164957284927368, "learning_rate": 9.996698280875759e-05, "loss": 0.1234, "step": 6100 }, { "grad_norm": 0.5615900754928589, "learning_rate": 9.99663792839458e-05, "loss": 0.0889, "step": 6110 }, { "grad_norm": 0.6344520449638367, "learning_rate": 9.99657702948859e-05, "loss": 0.1061, "step": 6120 }, { "grad_norm": 0.5878110527992249, "learning_rate": 9.996515584164448e-05, "loss": 0.0996, "step": 6130 }, { "grad_norm": 0.5932937264442444, "learning_rate": 9.996453592428873e-05, "loss": 0.1028, "step": 6140 }, { "grad_norm": 0.7645179629325867, "learning_rate": 9.996391054288646e-05, "loss": 0.1327, "step": 6150 }, { "grad_norm": 0.6094264984130859, "learning_rate": 9.996327969750605e-05, "loss": 0.1261, "step": 6160 }, { "grad_norm": 0.605821430683136, "learning_rate": 9.996264338821649e-05, "loss": 0.1105, "step": 6170 }, { "grad_norm": 0.6260783076286316, "learning_rate": 9.996200161508735e-05, "loss": 0.0991, "step": 6180 }, { "grad_norm": 0.8201457858085632, "learning_rate": 9.996135437818885e-05, "loss": 0.1153, "step": 6190 }, { "grad_norm": 0.5286712050437927, "learning_rate": 9.996070167759175e-05, "loss": 0.0939, "step": 6200 }, { "grad_norm": 0.7666762471199036, "learning_rate": 9.996004351336743e-05, "loss": 0.1115, "step": 6210 }, { "grad_norm": 0.9563229084014893, "learning_rate": 9.995937988558785e-05, "loss": 0.0858, "step": 6220 }, { "grad_norm": 0.6808724403381348, "learning_rate": 9.995871079432561e-05, "loss": 0.1067, "step": 6230 }, { "grad_norm": 0.7143781781196594, "learning_rate": 9.995803623965389e-05, "loss": 0.1157, "step": 6240 }, { "grad_norm": 0.7543137669563293, "learning_rate": 9.995735622164641e-05, "loss": 0.104, "step": 6250 }, { "grad_norm": 0.6310076713562012, "learning_rate": 9.995667074037758e-05, "loss": 0.085, "step": 6260 }, { "grad_norm": 0.5839231610298157, "learning_rate": 9.995597979592232e-05, "loss": 0.0956, "step": 6270 }, { "grad_norm": 0.7693612575531006, "learning_rate": 9.995528338835625e-05, "loss": 0.1062, "step": 6280 }, { "grad_norm": 0.6617515683174133, "learning_rate": 9.995458151775547e-05, "loss": 0.0957, "step": 6290 }, { "grad_norm": 0.5011305212974548, "learning_rate": 9.995387418419677e-05, "loss": 0.0907, "step": 6300 }, { "grad_norm": 0.5772294998168945, "learning_rate": 9.99531613877575e-05, "loss": 0.0861, "step": 6310 }, { "grad_norm": 0.6712604761123657, "learning_rate": 9.995244312851559e-05, "loss": 0.0892, "step": 6320 }, { "grad_norm": 0.6594446897506714, "learning_rate": 9.995171940654961e-05, "loss": 0.1083, "step": 6330 }, { "grad_norm": 0.7391983270645142, "learning_rate": 9.995099022193871e-05, "loss": 0.1059, "step": 6340 }, { "grad_norm": 0.6550357341766357, "learning_rate": 9.995025557476261e-05, "loss": 0.1004, "step": 6350 }, { "grad_norm": 0.9565126895904541, "learning_rate": 9.994951546510165e-05, "loss": 0.1039, "step": 6360 }, { "grad_norm": 0.5794399976730347, "learning_rate": 9.994876989303679e-05, "loss": 0.1036, "step": 6370 }, { "grad_norm": 0.8127283453941345, "learning_rate": 9.994801885864955e-05, "loss": 0.0952, "step": 6380 }, { "grad_norm": 1.166524887084961, "learning_rate": 9.994726236202205e-05, "loss": 0.1186, "step": 6390 }, { "grad_norm": 1.0056968927383423, "learning_rate": 9.994650040323704e-05, "loss": 0.1218, "step": 6400 }, { "grad_norm": 0.6561670899391174, "learning_rate": 9.994573298237784e-05, "loss": 0.088, "step": 6410 }, { "grad_norm": 0.7483327984809875, "learning_rate": 9.994496009952837e-05, "loss": 0.1042, "step": 6420 }, { "grad_norm": 0.6052358150482178, "learning_rate": 9.994418175477316e-05, "loss": 0.1037, "step": 6430 }, { "grad_norm": 0.5880987048149109, "learning_rate": 9.994339794819733e-05, "loss": 0.1052, "step": 6440 }, { "grad_norm": 0.7224973440170288, "learning_rate": 9.994260867988658e-05, "loss": 0.0987, "step": 6450 }, { "grad_norm": 0.5686720609664917, "learning_rate": 9.994181394992723e-05, "loss": 0.0988, "step": 6460 }, { "grad_norm": 0.6323820352554321, "learning_rate": 9.994101375840618e-05, "loss": 0.1082, "step": 6470 }, { "grad_norm": 0.8163219690322876, "learning_rate": 9.994020810541098e-05, "loss": 0.0956, "step": 6480 }, { "grad_norm": 1.1480573415756226, "learning_rate": 9.99393969910297e-05, "loss": 0.1234, "step": 6490 }, { "grad_norm": 0.6476849913597107, "learning_rate": 9.993858041535104e-05, "loss": 0.0875, "step": 6500 }, { "grad_norm": 0.6189082860946655, "learning_rate": 9.99377583784643e-05, "loss": 0.0852, "step": 6510 }, { "grad_norm": 1.039001226425171, "learning_rate": 9.993693088045939e-05, "loss": 0.102, "step": 6520 }, { "grad_norm": 0.7602106332778931, "learning_rate": 9.99360979214268e-05, "loss": 0.0969, "step": 6530 }, { "grad_norm": 1.4141795635223389, "learning_rate": 9.99352595014576e-05, "loss": 0.106, "step": 6540 }, { "grad_norm": 1.0775688886642456, "learning_rate": 9.993441562064354e-05, "loss": 0.0987, "step": 6550 }, { "grad_norm": 0.770470917224884, "learning_rate": 9.993356627907685e-05, "loss": 0.1021, "step": 6560 }, { "grad_norm": 0.6459960341453552, "learning_rate": 9.99327114768504e-05, "loss": 0.1096, "step": 6570 }, { "grad_norm": 0.6573818325996399, "learning_rate": 9.99318512140577e-05, "loss": 0.085, "step": 6580 }, { "grad_norm": 0.7510082125663757, "learning_rate": 9.993098549079284e-05, "loss": 0.115, "step": 6590 }, { "grad_norm": 0.7353570461273193, "learning_rate": 9.993011430715047e-05, "loss": 0.1042, "step": 6600 }, { "grad_norm": 0.5842042565345764, "learning_rate": 9.992923766322586e-05, "loss": 0.0893, "step": 6610 }, { "grad_norm": 0.5906859040260315, "learning_rate": 9.99283555591149e-05, "loss": 0.0786, "step": 6620 }, { "grad_norm": 0.5615403652191162, "learning_rate": 9.992746799491404e-05, "loss": 0.0942, "step": 6630 }, { "grad_norm": 0.7024739384651184, "learning_rate": 9.992657497072033e-05, "loss": 0.0955, "step": 6640 }, { "grad_norm": 0.5610791444778442, "learning_rate": 9.992567648663147e-05, "loss": 0.1131, "step": 6650 }, { "grad_norm": 0.6308814287185669, "learning_rate": 9.992477254274568e-05, "loss": 0.0896, "step": 6660 }, { "grad_norm": 0.5123977661132812, "learning_rate": 9.992386313916183e-05, "loss": 0.0841, "step": 6670 }, { "grad_norm": 0.8263581991195679, "learning_rate": 9.992294827597934e-05, "loss": 0.1192, "step": 6680 }, { "grad_norm": 0.7703801393508911, "learning_rate": 9.992202795329831e-05, "loss": 0.1016, "step": 6690 }, { "grad_norm": 0.6111246347427368, "learning_rate": 9.992110217121936e-05, "loss": 0.0878, "step": 6700 }, { "grad_norm": 0.7748539447784424, "learning_rate": 9.992017092984372e-05, "loss": 0.1117, "step": 6710 }, { "grad_norm": 0.5877307653427124, "learning_rate": 9.991923422927326e-05, "loss": 0.1054, "step": 6720 }, { "grad_norm": 0.8457234501838684, "learning_rate": 9.991829206961037e-05, "loss": 0.1106, "step": 6730 }, { "grad_norm": 0.6558064818382263, "learning_rate": 9.991734445095813e-05, "loss": 0.1289, "step": 6740 }, { "grad_norm": 0.659062922000885, "learning_rate": 9.991639137342015e-05, "loss": 0.0874, "step": 6750 }, { "grad_norm": 0.4353247880935669, "learning_rate": 9.991543283710064e-05, "loss": 0.0939, "step": 6760 }, { "grad_norm": 0.6302896738052368, "learning_rate": 9.991446884210445e-05, "loss": 0.082, "step": 6770 }, { "grad_norm": 0.6957215666770935, "learning_rate": 9.9913499388537e-05, "loss": 0.113, "step": 6780 }, { "grad_norm": 0.9839098453521729, "learning_rate": 9.99125244765043e-05, "loss": 0.0892, "step": 6790 }, { "grad_norm": 1.0163569450378418, "learning_rate": 9.991154410611296e-05, "loss": 0.1153, "step": 6800 }, { "grad_norm": 0.7063432335853577, "learning_rate": 9.99105582774702e-05, "loss": 0.0869, "step": 6810 }, { "grad_norm": 0.6865578889846802, "learning_rate": 9.990956699068384e-05, "loss": 0.107, "step": 6820 }, { "grad_norm": 0.6000558733940125, "learning_rate": 9.990857024586224e-05, "loss": 0.0928, "step": 6830 }, { "grad_norm": 0.6600949168205261, "learning_rate": 9.990756804311446e-05, "loss": 0.1086, "step": 6840 }, { "grad_norm": 0.5418925285339355, "learning_rate": 9.990656038255006e-05, "loss": 0.0901, "step": 6850 }, { "grad_norm": 0.8290911316871643, "learning_rate": 9.990554726427926e-05, "loss": 0.0972, "step": 6860 }, { "grad_norm": 0.952492356300354, "learning_rate": 9.990452868841284e-05, "loss": 0.125, "step": 6870 }, { "grad_norm": 0.6254262924194336, "learning_rate": 9.99035046550622e-05, "loss": 0.1174, "step": 6880 }, { "grad_norm": 0.5422330498695374, "learning_rate": 9.99024751643393e-05, "loss": 0.098, "step": 6890 }, { "grad_norm": 1.1304442882537842, "learning_rate": 9.990144021635677e-05, "loss": 0.1069, "step": 6900 }, { "grad_norm": 0.8532564640045166, "learning_rate": 9.990039981122775e-05, "loss": 0.0927, "step": 6910 }, { "grad_norm": 0.5907381176948547, "learning_rate": 9.989935394906602e-05, "loss": 0.1124, "step": 6920 }, { "grad_norm": 0.8322775959968567, "learning_rate": 9.989830262998598e-05, "loss": 0.1192, "step": 6930 }, { "grad_norm": 0.7001019716262817, "learning_rate": 9.989724585410259e-05, "loss": 0.0934, "step": 6940 }, { "grad_norm": 0.5838519334793091, "learning_rate": 9.989618362153139e-05, "loss": 0.0945, "step": 6950 }, { "grad_norm": 0.6038575768470764, "learning_rate": 9.989511593238859e-05, "loss": 0.0901, "step": 6960 }, { "grad_norm": 0.6375845074653625, "learning_rate": 9.98940427867909e-05, "loss": 0.1093, "step": 6970 }, { "grad_norm": 0.6844246983528137, "learning_rate": 9.989296418485573e-05, "loss": 0.0984, "step": 6980 }, { "grad_norm": 0.4949099123477936, "learning_rate": 9.989188012670101e-05, "loss": 0.0831, "step": 6990 }, { "grad_norm": 0.6451395153999329, "learning_rate": 9.989079061244528e-05, "loss": 0.0898, "step": 7000 }, { "grad_norm": 0.5991080403327942, "learning_rate": 9.988969564220769e-05, "loss": 0.1078, "step": 7010 }, { "grad_norm": 0.7537737488746643, "learning_rate": 9.988859521610801e-05, "loss": 0.0981, "step": 7020 }, { "grad_norm": 0.6206802129745483, "learning_rate": 9.988748933426656e-05, "loss": 0.1244, "step": 7030 }, { "grad_norm": 0.4943234324455261, "learning_rate": 9.988637799680428e-05, "loss": 0.094, "step": 7040 }, { "grad_norm": 0.7357549667358398, "learning_rate": 9.98852612038427e-05, "loss": 0.0807, "step": 7050 }, { "grad_norm": 0.7080848217010498, "learning_rate": 9.988413895550397e-05, "loss": 0.086, "step": 7060 }, { "grad_norm": 0.6427303552627563, "learning_rate": 9.98830112519108e-05, "loss": 0.0954, "step": 7070 }, { "grad_norm": 0.5763518810272217, "learning_rate": 9.98818780931865e-05, "loss": 0.1233, "step": 7080 }, { "grad_norm": 0.8465977311134338, "learning_rate": 9.988073947945502e-05, "loss": 0.0889, "step": 7090 }, { "grad_norm": 0.6811879873275757, "learning_rate": 9.987959541084087e-05, "loss": 0.0967, "step": 7100 }, { "grad_norm": 0.5223190188407898, "learning_rate": 9.987844588746915e-05, "loss": 0.0985, "step": 7110 }, { "grad_norm": 0.5652349591255188, "learning_rate": 9.987729090946558e-05, "loss": 0.0873, "step": 7120 }, { "grad_norm": 0.5345380902290344, "learning_rate": 9.987613047695647e-05, "loss": 0.0853, "step": 7130 }, { "grad_norm": 0.6143341660499573, "learning_rate": 9.987496459006871e-05, "loss": 0.1019, "step": 7140 }, { "grad_norm": 0.6565670967102051, "learning_rate": 9.987379324892982e-05, "loss": 0.1165, "step": 7150 }, { "grad_norm": 0.9053892493247986, "learning_rate": 9.987261645366788e-05, "loss": 0.0977, "step": 7160 }, { "grad_norm": 0.5278805494308472, "learning_rate": 9.987143420441158e-05, "loss": 0.1018, "step": 7170 }, { "grad_norm": 0.8068172335624695, "learning_rate": 9.987024650129022e-05, "loss": 0.1046, "step": 7180 }, { "grad_norm": 0.5487474203109741, "learning_rate": 9.986905334443368e-05, "loss": 0.1045, "step": 7190 }, { "grad_norm": 0.7631041407585144, "learning_rate": 9.986785473397245e-05, "loss": 0.1055, "step": 7200 }, { "grad_norm": 0.6041258573532104, "learning_rate": 9.98666506700376e-05, "loss": 0.0929, "step": 7210 }, { "grad_norm": 0.8155827522277832, "learning_rate": 9.986544115276081e-05, "loss": 0.1017, "step": 7220 }, { "grad_norm": 0.6797366738319397, "learning_rate": 9.986422618227433e-05, "loss": 0.1019, "step": 7230 }, { "grad_norm": 0.6313496232032776, "learning_rate": 9.986300575871106e-05, "loss": 0.1083, "step": 7240 }, { "grad_norm": 0.6571186780929565, "learning_rate": 9.986177988220444e-05, "loss": 0.0987, "step": 7250 }, { "grad_norm": 0.7712662220001221, "learning_rate": 9.986054855288856e-05, "loss": 0.0922, "step": 7260 }, { "grad_norm": 0.5436675548553467, "learning_rate": 9.985931177089802e-05, "loss": 0.1068, "step": 7270 }, { "grad_norm": 0.6379490494728088, "learning_rate": 9.985806953636814e-05, "loss": 0.1012, "step": 7280 }, { "grad_norm": 0.673747181892395, "learning_rate": 9.985682184943471e-05, "loss": 0.1057, "step": 7290 }, { "grad_norm": 0.4692671597003937, "learning_rate": 9.98555687102342e-05, "loss": 0.092, "step": 7300 }, { "grad_norm": 0.7382160425186157, "learning_rate": 9.985431011890367e-05, "loss": 0.099, "step": 7310 }, { "grad_norm": 0.6772993803024292, "learning_rate": 9.985304607558075e-05, "loss": 0.0994, "step": 7320 }, { "grad_norm": 0.6051153540611267, "learning_rate": 9.985177658040364e-05, "loss": 0.104, "step": 7330 }, { "grad_norm": 0.6612464189529419, "learning_rate": 9.985050163351119e-05, "loss": 0.097, "step": 7340 }, { "grad_norm": 0.8772600293159485, "learning_rate": 9.984922123504286e-05, "loss": 0.0808, "step": 7350 }, { "grad_norm": 0.6873916387557983, "learning_rate": 9.984793538513862e-05, "loss": 0.0924, "step": 7360 }, { "grad_norm": 0.6329360008239746, "learning_rate": 9.984664408393912e-05, "loss": 0.1027, "step": 7370 }, { "grad_norm": 0.5684174299240112, "learning_rate": 9.984534733158556e-05, "loss": 0.093, "step": 7380 }, { "grad_norm": 0.7733368277549744, "learning_rate": 9.984404512821977e-05, "loss": 0.1056, "step": 7390 }, { "grad_norm": 0.8307917714118958, "learning_rate": 9.984273747398411e-05, "loss": 0.105, "step": 7400 }, { "grad_norm": 0.6236806511878967, "learning_rate": 9.984142436902165e-05, "loss": 0.0936, "step": 7410 }, { "grad_norm": 0.5347803831100464, "learning_rate": 9.984010581347596e-05, "loss": 0.1043, "step": 7420 }, { "grad_norm": 0.4756137430667877, "learning_rate": 9.983878180749121e-05, "loss": 0.1011, "step": 7430 }, { "grad_norm": 0.6466806530952454, "learning_rate": 9.983745235121222e-05, "loss": 0.0924, "step": 7440 }, { "grad_norm": 0.8909441828727722, "learning_rate": 9.983611744478438e-05, "loss": 0.0944, "step": 7450 }, { "grad_norm": 0.7893890738487244, "learning_rate": 9.983477708835365e-05, "loss": 0.0886, "step": 7460 }, { "grad_norm": 0.6608121395111084, "learning_rate": 9.983343128206664e-05, "loss": 0.1058, "step": 7470 }, { "grad_norm": 0.7588350176811218, "learning_rate": 9.983208002607049e-05, "loss": 0.1033, "step": 7480 }, { "grad_norm": 0.7149489521980286, "learning_rate": 9.9830723320513e-05, "loss": 0.0957, "step": 7490 }, { "grad_norm": 0.8894827961921692, "learning_rate": 9.982936116554254e-05, "loss": 0.1042, "step": 7500 }, { "grad_norm": 0.5682381987571716, "learning_rate": 9.982799356130803e-05, "loss": 0.0843, "step": 7510 }, { "grad_norm": 0.5854442715644836, "learning_rate": 9.982662050795908e-05, "loss": 0.0794, "step": 7520 }, { "grad_norm": 0.5659598708152771, "learning_rate": 9.982524200564583e-05, "loss": 0.0864, "step": 7530 }, { "grad_norm": 0.47159212827682495, "learning_rate": 9.982385805451901e-05, "loss": 0.0894, "step": 7540 }, { "grad_norm": 0.6125784516334534, "learning_rate": 9.982246865472998e-05, "loss": 0.0829, "step": 7550 }, { "grad_norm": 0.6560521721839905, "learning_rate": 9.982107380643069e-05, "loss": 0.093, "step": 7560 }, { "grad_norm": 0.711185097694397, "learning_rate": 9.981967350977368e-05, "loss": 0.1038, "step": 7570 }, { "grad_norm": 0.5370371341705322, "learning_rate": 9.981826776491208e-05, "loss": 0.0902, "step": 7580 }, { "grad_norm": 0.6764140725135803, "learning_rate": 9.98168565719996e-05, "loss": 0.1093, "step": 7590 }, { "grad_norm": 0.5166385173797607, "learning_rate": 9.98154399311906e-05, "loss": 0.097, "step": 7600 }, { "grad_norm": 0.6591076850891113, "learning_rate": 9.981401784263997e-05, "loss": 0.1062, "step": 7610 }, { "grad_norm": 0.5749446749687195, "learning_rate": 9.981259030650326e-05, "loss": 0.0939, "step": 7620 }, { "grad_norm": 0.6148253679275513, "learning_rate": 9.981115732293655e-05, "loss": 0.0904, "step": 7630 }, { "grad_norm": 0.6917336583137512, "learning_rate": 9.980971889209659e-05, "loss": 0.0936, "step": 7640 }, { "grad_norm": 0.7410808801651001, "learning_rate": 9.980827501414064e-05, "loss": 0.0832, "step": 7650 }, { "grad_norm": 0.5946872234344482, "learning_rate": 9.980682568922663e-05, "loss": 0.0893, "step": 7660 }, { "grad_norm": 0.522490918636322, "learning_rate": 9.980537091751304e-05, "loss": 0.0934, "step": 7670 }, { "grad_norm": 0.7246273159980774, "learning_rate": 9.980391069915897e-05, "loss": 0.0866, "step": 7680 }, { "grad_norm": 0.5346115231513977, "learning_rate": 9.98024450343241e-05, "loss": 0.0907, "step": 7690 }, { "grad_norm": 0.6516171097755432, "learning_rate": 9.980097392316872e-05, "loss": 0.0834, "step": 7700 }, { "grad_norm": 0.6127486824989319, "learning_rate": 9.97994973658537e-05, "loss": 0.1087, "step": 7710 }, { "grad_norm": 0.6003605723381042, "learning_rate": 9.979801536254054e-05, "loss": 0.0989, "step": 7720 }, { "grad_norm": 0.6274243593215942, "learning_rate": 9.979652791339127e-05, "loss": 0.1022, "step": 7730 }, { "grad_norm": 0.5296905636787415, "learning_rate": 9.97950350185686e-05, "loss": 0.0889, "step": 7740 }, { "grad_norm": 0.5329187512397766, "learning_rate": 9.979353667823574e-05, "loss": 0.0945, "step": 7750 }, { "grad_norm": 0.8069365620613098, "learning_rate": 9.979203289255658e-05, "loss": 0.0987, "step": 7760 }, { "grad_norm": 0.652104914188385, "learning_rate": 9.979052366169557e-05, "loss": 0.0976, "step": 7770 }, { "grad_norm": 0.6561832427978516, "learning_rate": 9.978900898581775e-05, "loss": 0.0967, "step": 7780 }, { "grad_norm": 0.47995439171791077, "learning_rate": 9.978748886508875e-05, "loss": 0.0834, "step": 7790 }, { "grad_norm": 0.6460239291191101, "learning_rate": 9.978596329967484e-05, "loss": 0.0955, "step": 7800 }, { "grad_norm": 0.7025154829025269, "learning_rate": 9.978443228974284e-05, "loss": 0.0903, "step": 7810 }, { "grad_norm": 0.524189829826355, "learning_rate": 9.978289583546015e-05, "loss": 0.0827, "step": 7820 }, { "grad_norm": 0.8371972441673279, "learning_rate": 9.978135393699484e-05, "loss": 0.0963, "step": 7830 }, { "grad_norm": 0.7038826942443848, "learning_rate": 9.977980659451548e-05, "loss": 0.0945, "step": 7840 }, { "grad_norm": 0.6923335790634155, "learning_rate": 9.977825380819135e-05, "loss": 0.0802, "step": 7850 }, { "grad_norm": 0.5949808955192566, "learning_rate": 9.97766955781922e-05, "loss": 0.0932, "step": 7860 }, { "grad_norm": 0.8989306092262268, "learning_rate": 9.977513190468848e-05, "loss": 0.1034, "step": 7870 }, { "grad_norm": 0.5960517525672913, "learning_rate": 9.977356278785116e-05, "loss": 0.1018, "step": 7880 }, { "grad_norm": 0.7032139301300049, "learning_rate": 9.977198822785184e-05, "loss": 0.0905, "step": 7890 }, { "grad_norm": 0.6542237997055054, "learning_rate": 9.977040822486273e-05, "loss": 0.1009, "step": 7900 }, { "grad_norm": 0.6000580787658691, "learning_rate": 9.97688227790566e-05, "loss": 0.1138, "step": 7910 }, { "grad_norm": 0.6113623976707458, "learning_rate": 9.976723189060684e-05, "loss": 0.1235, "step": 7920 }, { "grad_norm": 0.6417939066886902, "learning_rate": 9.976563555968742e-05, "loss": 0.0903, "step": 7930 }, { "grad_norm": 0.7084187865257263, "learning_rate": 9.976403378647292e-05, "loss": 0.0955, "step": 7940 }, { "grad_norm": 0.683756947517395, "learning_rate": 9.97624265711385e-05, "loss": 0.0956, "step": 7950 }, { "grad_norm": 1.19693922996521, "learning_rate": 9.976081391385993e-05, "loss": 0.1017, "step": 7960 }, { "grad_norm": 0.7504327893257141, "learning_rate": 9.975919581481356e-05, "loss": 0.0968, "step": 7970 }, { "grad_norm": 0.7753810286521912, "learning_rate": 9.975757227417634e-05, "loss": 0.0903, "step": 7980 }, { "grad_norm": 0.6166303753852844, "learning_rate": 9.975594329212586e-05, "loss": 0.0873, "step": 7990 }, { "grad_norm": 0.7241845726966858, "learning_rate": 9.97543088688402e-05, "loss": 0.0884, "step": 8000 }, { "grad_norm": 0.7647027373313904, "learning_rate": 9.975266900449814e-05, "loss": 0.0766, "step": 8010 }, { "grad_norm": 0.8660218119621277, "learning_rate": 9.975102369927898e-05, "loss": 0.0874, "step": 8020 }, { "grad_norm": 0.556987464427948, "learning_rate": 9.974937295336269e-05, "loss": 0.0858, "step": 8030 }, { "grad_norm": 0.6474213004112244, "learning_rate": 9.974771676692975e-05, "loss": 0.0836, "step": 8040 }, { "grad_norm": 0.6072742938995361, "learning_rate": 9.974605514016131e-05, "loss": 0.0946, "step": 8050 }, { "grad_norm": 0.5150613784790039, "learning_rate": 9.974438807323907e-05, "loss": 0.0872, "step": 8060 }, { "grad_norm": 0.7115942239761353, "learning_rate": 9.974271556634535e-05, "loss": 0.0874, "step": 8070 }, { "grad_norm": 0.7376955151557922, "learning_rate": 9.974103761966302e-05, "loss": 0.1115, "step": 8080 }, { "grad_norm": 1.3083995580673218, "learning_rate": 9.973935423337563e-05, "loss": 0.0894, "step": 8090 }, { "grad_norm": 0.5249735713005066, "learning_rate": 9.973766540766722e-05, "loss": 0.1075, "step": 8100 }, { "grad_norm": 0.7112849950790405, "learning_rate": 9.97359711427225e-05, "loss": 0.1084, "step": 8110 }, { "grad_norm": 0.6294620037078857, "learning_rate": 9.973427143872677e-05, "loss": 0.1, "step": 8120 }, { "grad_norm": 0.6726016998291016, "learning_rate": 9.973256629586589e-05, "loss": 0.1006, "step": 8130 }, { "grad_norm": 0.5573478937149048, "learning_rate": 9.973085571432632e-05, "loss": 0.1068, "step": 8140 }, { "grad_norm": 0.5994419455528259, "learning_rate": 9.972913969429513e-05, "loss": 0.0917, "step": 8150 }, { "grad_norm": 0.461438924074173, "learning_rate": 9.972741823596e-05, "loss": 0.0845, "step": 8160 }, { "grad_norm": 0.5559910535812378, "learning_rate": 9.972569133950917e-05, "loss": 0.0974, "step": 8170 }, { "grad_norm": 0.5132887363433838, "learning_rate": 9.972395900513151e-05, "loss": 0.075, "step": 8180 }, { "grad_norm": 0.6110216975212097, "learning_rate": 9.972222123301645e-05, "loss": 0.0823, "step": 8190 }, { "grad_norm": 0.6914598345756531, "learning_rate": 9.972047802335403e-05, "loss": 0.082, "step": 8200 }, { "grad_norm": 0.32948923110961914, "learning_rate": 9.971872937633488e-05, "loss": 0.0729, "step": 8210 }, { "grad_norm": 0.5425401329994202, "learning_rate": 9.971697529215024e-05, "loss": 0.0982, "step": 8220 }, { "grad_norm": 0.5672141313552856, "learning_rate": 9.971521577099192e-05, "loss": 0.0827, "step": 8230 }, { "grad_norm": 0.6183205246925354, "learning_rate": 9.971345081305236e-05, "loss": 0.0962, "step": 8240 }, { "grad_norm": 0.6904867887496948, "learning_rate": 9.971168041852456e-05, "loss": 0.0902, "step": 8250 }, { "grad_norm": 0.6018035411834717, "learning_rate": 9.970990458760215e-05, "loss": 0.0677, "step": 8260 }, { "grad_norm": 0.4531612694263458, "learning_rate": 9.970812332047929e-05, "loss": 0.0879, "step": 8270 }, { "grad_norm": 0.558463990688324, "learning_rate": 9.97063366173508e-05, "loss": 0.099, "step": 8280 }, { "grad_norm": 0.6389732956886292, "learning_rate": 9.970454447841207e-05, "loss": 0.0856, "step": 8290 }, { "grad_norm": 0.6240937113761902, "learning_rate": 9.970274690385909e-05, "loss": 0.0945, "step": 8300 }, { "grad_norm": 0.522860050201416, "learning_rate": 9.970094389388844e-05, "loss": 0.0822, "step": 8310 }, { "grad_norm": 0.5938932299613953, "learning_rate": 9.969913544869728e-05, "loss": 0.0821, "step": 8320 }, { "grad_norm": 0.788172721862793, "learning_rate": 9.96973215684834e-05, "loss": 0.087, "step": 8330 }, { "grad_norm": 0.5587490797042847, "learning_rate": 9.969550225344513e-05, "loss": 0.0708, "step": 8340 }, { "grad_norm": 0.524427592754364, "learning_rate": 9.969367750378147e-05, "loss": 0.0927, "step": 8350 }, { "grad_norm": 0.5726475715637207, "learning_rate": 9.969184731969194e-05, "loss": 0.0794, "step": 8360 }, { "grad_norm": 0.6126094460487366, "learning_rate": 9.96900117013767e-05, "loss": 0.1027, "step": 8370 }, { "grad_norm": 0.668870210647583, "learning_rate": 9.96881706490365e-05, "loss": 0.0931, "step": 8380 }, { "grad_norm": 1.030137538909912, "learning_rate": 9.968632416287265e-05, "loss": 0.0854, "step": 8390 }, { "grad_norm": 0.4805718660354614, "learning_rate": 9.96844722430871e-05, "loss": 0.0803, "step": 8400 }, { "grad_norm": 0.5945214033126831, "learning_rate": 9.968261488988235e-05, "loss": 0.094, "step": 8410 }, { "grad_norm": 0.8402397632598877, "learning_rate": 9.968075210346155e-05, "loss": 0.0837, "step": 8420 }, { "grad_norm": 0.7409793734550476, "learning_rate": 9.967888388402839e-05, "loss": 0.0898, "step": 8430 }, { "grad_norm": 0.5769925713539124, "learning_rate": 9.967701023178717e-05, "loss": 0.0851, "step": 8440 }, { "grad_norm": 0.6869195699691772, "learning_rate": 9.967513114694282e-05, "loss": 0.0933, "step": 8450 }, { "grad_norm": 0.5837861895561218, "learning_rate": 9.967324662970079e-05, "loss": 0.0768, "step": 8460 }, { "grad_norm": 0.8106063008308411, "learning_rate": 9.96713566802672e-05, "loss": 0.0868, "step": 8470 }, { "grad_norm": 0.6190804243087769, "learning_rate": 9.966946129884873e-05, "loss": 0.0978, "step": 8480 }, { "grad_norm": 0.8968300819396973, "learning_rate": 9.966756048565265e-05, "loss": 0.0965, "step": 8490 }, { "grad_norm": 0.813123881816864, "learning_rate": 9.966565424088681e-05, "loss": 0.1125, "step": 8500 }, { "grad_norm": 0.9142129421234131, "learning_rate": 9.96637425647597e-05, "loss": 0.0941, "step": 8510 }, { "grad_norm": 0.5496432781219482, "learning_rate": 9.966182545748038e-05, "loss": 0.0859, "step": 8520 }, { "grad_norm": 0.8053575158119202, "learning_rate": 9.96599029192585e-05, "loss": 0.0911, "step": 8530 }, { "grad_norm": 0.6220606565475464, "learning_rate": 9.965797495030428e-05, "loss": 0.1036, "step": 8540 }, { "grad_norm": 0.5458433628082275, "learning_rate": 9.96560415508286e-05, "loss": 0.0928, "step": 8550 }, { "grad_norm": 0.6482126712799072, "learning_rate": 9.965410272104286e-05, "loss": 0.0858, "step": 8560 }, { "grad_norm": 0.6146712303161621, "learning_rate": 9.96521584611591e-05, "loss": 0.1158, "step": 8570 }, { "grad_norm": 0.472806841135025, "learning_rate": 9.965020877138994e-05, "loss": 0.0984, "step": 8580 }, { "grad_norm": 0.557946503162384, "learning_rate": 9.964825365194861e-05, "loss": 0.0912, "step": 8590 }, { "grad_norm": 0.5416070222854614, "learning_rate": 9.96462931030489e-05, "loss": 0.0856, "step": 8600 }, { "grad_norm": 0.7060956358909607, "learning_rate": 9.96443271249052e-05, "loss": 0.086, "step": 8610 }, { "grad_norm": 0.568074643611908, "learning_rate": 9.964235571773255e-05, "loss": 0.1015, "step": 8620 }, { "grad_norm": 0.6039186120033264, "learning_rate": 9.96403788817465e-05, "loss": 0.0893, "step": 8630 }, { "grad_norm": 0.5524296760559082, "learning_rate": 9.963839661716325e-05, "loss": 0.0945, "step": 8640 }, { "grad_norm": 0.5347482562065125, "learning_rate": 9.963640892419958e-05, "loss": 0.0969, "step": 8650 }, { "grad_norm": 0.8608977198600769, "learning_rate": 9.963441580307286e-05, "loss": 0.0782, "step": 8660 }, { "grad_norm": 0.541918158531189, "learning_rate": 9.963241725400104e-05, "loss": 0.0796, "step": 8670 }, { "grad_norm": 0.6551148891448975, "learning_rate": 9.963041327720271e-05, "loss": 0.0841, "step": 8680 }, { "grad_norm": 0.5885763168334961, "learning_rate": 9.962840387289697e-05, "loss": 0.0937, "step": 8690 }, { "grad_norm": 0.690066933631897, "learning_rate": 9.962638904130363e-05, "loss": 0.1049, "step": 8700 }, { "grad_norm": 0.666325032711029, "learning_rate": 9.962436878264298e-05, "loss": 0.1096, "step": 8710 }, { "grad_norm": 0.6494601964950562, "learning_rate": 9.962234309713598e-05, "loss": 0.1094, "step": 8720 }, { "grad_norm": 0.6895642876625061, "learning_rate": 9.962031198500414e-05, "loss": 0.0913, "step": 8730 }, { "grad_norm": 0.4759531617164612, "learning_rate": 9.961827544646958e-05, "loss": 0.0829, "step": 8740 }, { "grad_norm": 0.5238782167434692, "learning_rate": 9.961623348175501e-05, "loss": 0.0999, "step": 8750 }, { "grad_norm": 0.552483856678009, "learning_rate": 9.961418609108377e-05, "loss": 0.0813, "step": 8760 }, { "grad_norm": 0.7166662216186523, "learning_rate": 9.961213327467971e-05, "loss": 0.0991, "step": 8770 }, { "grad_norm": 0.6160249710083008, "learning_rate": 9.961007503276736e-05, "loss": 0.092, "step": 8780 }, { "grad_norm": 0.7764288187026978, "learning_rate": 9.960801136557179e-05, "loss": 0.0827, "step": 8790 }, { "grad_norm": 0.5544486045837402, "learning_rate": 9.960594227331866e-05, "loss": 0.0754, "step": 8800 }, { "grad_norm": 0.48469477891921997, "learning_rate": 9.960386775623429e-05, "loss": 0.0843, "step": 8810 }, { "grad_norm": 0.6173731088638306, "learning_rate": 9.96017878145455e-05, "loss": 0.0903, "step": 8820 }, { "grad_norm": 0.9675564169883728, "learning_rate": 9.959970244847977e-05, "loss": 0.0927, "step": 8830 }, { "grad_norm": 0.565208375453949, "learning_rate": 9.959761165826518e-05, "loss": 0.0851, "step": 8840 }, { "grad_norm": 0.8072904944419861, "learning_rate": 9.959551544413033e-05, "loss": 0.0879, "step": 8850 }, { "grad_norm": 0.6944670677185059, "learning_rate": 9.959341380630448e-05, "loss": 0.0895, "step": 8860 }, { "grad_norm": 0.6589226722717285, "learning_rate": 9.959130674501746e-05, "loss": 0.1004, "step": 8870 }, { "grad_norm": 0.4864596128463745, "learning_rate": 9.958919426049968e-05, "loss": 0.1029, "step": 8880 }, { "grad_norm": 0.7217707633972168, "learning_rate": 9.958707635298219e-05, "loss": 0.0815, "step": 8890 }, { "grad_norm": 0.5275627374649048, "learning_rate": 9.958495302269657e-05, "loss": 0.1006, "step": 8900 }, { "grad_norm": 0.5520925521850586, "learning_rate": 9.958282426987503e-05, "loss": 0.094, "step": 8910 }, { "grad_norm": 0.7662851214408875, "learning_rate": 9.95806900947504e-05, "loss": 0.1113, "step": 8920 }, { "grad_norm": 0.7199344635009766, "learning_rate": 9.957855049755604e-05, "loss": 0.0918, "step": 8930 }, { "grad_norm": 0.5782783627510071, "learning_rate": 9.957640547852593e-05, "loss": 0.1003, "step": 8940 }, { "grad_norm": 0.611681342124939, "learning_rate": 9.957425503789466e-05, "loss": 0.0911, "step": 8950 }, { "grad_norm": 0.5930809378623962, "learning_rate": 9.957209917589738e-05, "loss": 0.0759, "step": 8960 }, { "grad_norm": 0.4951609969139099, "learning_rate": 9.956993789276987e-05, "loss": 0.0813, "step": 8970 }, { "grad_norm": 0.4332138001918793, "learning_rate": 9.956777118874847e-05, "loss": 0.0842, "step": 8980 }, { "grad_norm": 0.6668884754180908, "learning_rate": 9.956559906407016e-05, "loss": 0.1108, "step": 8990 }, { "grad_norm": 0.5502238869667053, "learning_rate": 9.956342151897245e-05, "loss": 0.0905, "step": 9000 }, { "grad_norm": 1.0058236122131348, "learning_rate": 9.956123855369346e-05, "loss": 0.0772, "step": 9010 }, { "grad_norm": 0.8214946389198303, "learning_rate": 9.955905016847196e-05, "loss": 0.0898, "step": 9020 }, { "grad_norm": 0.6388535499572754, "learning_rate": 9.955685636354723e-05, "loss": 0.1198, "step": 9030 }, { "grad_norm": 0.7376430034637451, "learning_rate": 9.95546571391592e-05, "loss": 0.1049, "step": 9040 }, { "grad_norm": 0.4444321393966675, "learning_rate": 9.955245249554837e-05, "loss": 0.0782, "step": 9050 }, { "grad_norm": 0.48383596539497375, "learning_rate": 9.955024243295582e-05, "loss": 0.1094, "step": 9060 }, { "grad_norm": 0.6617894172668457, "learning_rate": 9.954802695162328e-05, "loss": 0.086, "step": 9070 }, { "grad_norm": 0.7176649570465088, "learning_rate": 9.954580605179302e-05, "loss": 0.0841, "step": 9080 }, { "grad_norm": 0.6815929412841797, "learning_rate": 9.954357973370788e-05, "loss": 0.0935, "step": 9090 }, { "grad_norm": 0.5308022499084473, "learning_rate": 9.954134799761135e-05, "loss": 0.1045, "step": 9100 }, { "grad_norm": 0.6897048354148865, "learning_rate": 9.953911084374748e-05, "loss": 0.0974, "step": 9110 }, { "grad_norm": 0.7320863604545593, "learning_rate": 9.953686827236093e-05, "loss": 0.0942, "step": 9120 }, { "grad_norm": 0.8499615788459778, "learning_rate": 9.953462028369695e-05, "loss": 0.109, "step": 9130 }, { "grad_norm": 0.605301022529602, "learning_rate": 9.953236687800136e-05, "loss": 0.0851, "step": 9140 }, { "grad_norm": 0.3583480715751648, "learning_rate": 9.95301080555206e-05, "loss": 0.1051, "step": 9150 }, { "grad_norm": 0.5078926682472229, "learning_rate": 9.952784381650171e-05, "loss": 0.078, "step": 9160 }, { "grad_norm": 0.663666844367981, "learning_rate": 9.952557416119226e-05, "loss": 0.0774, "step": 9170 }, { "grad_norm": 0.7463130950927734, "learning_rate": 9.95232990898405e-05, "loss": 0.0795, "step": 9180 }, { "grad_norm": 0.5454906225204468, "learning_rate": 9.95210186026952e-05, "loss": 0.0707, "step": 9190 }, { "grad_norm": 0.494364470243454, "learning_rate": 9.951873270000576e-05, "loss": 0.0977, "step": 9200 }, { "grad_norm": 0.5406431555747986, "learning_rate": 9.951644138202216e-05, "loss": 0.0973, "step": 9210 }, { "grad_norm": 0.48436808586120605, "learning_rate": 9.951414464899498e-05, "loss": 0.0917, "step": 9220 }, { "grad_norm": 0.7828488349914551, "learning_rate": 9.951184250117538e-05, "loss": 0.0776, "step": 9230 }, { "grad_norm": 0.6244980692863464, "learning_rate": 9.950953493881513e-05, "loss": 0.0825, "step": 9240 }, { "grad_norm": 0.6579099297523499, "learning_rate": 9.950722196216658e-05, "loss": 0.0792, "step": 9250 }, { "grad_norm": 0.8876604437828064, "learning_rate": 9.950490357148265e-05, "loss": 0.0749, "step": 9260 }, { "grad_norm": 0.4535823464393616, "learning_rate": 9.950257976701692e-05, "loss": 0.0885, "step": 9270 }, { "grad_norm": 0.6193718314170837, "learning_rate": 9.950025054902348e-05, "loss": 0.0845, "step": 9280 }, { "grad_norm": 0.8390649557113647, "learning_rate": 9.949791591775706e-05, "loss": 0.0781, "step": 9290 }, { "grad_norm": 0.5389679670333862, "learning_rate": 9.949557587347298e-05, "loss": 0.1013, "step": 9300 }, { "grad_norm": 0.8840243220329285, "learning_rate": 9.949323041642713e-05, "loss": 0.097, "step": 9310 }, { "grad_norm": 0.8211491107940674, "learning_rate": 9.949087954687602e-05, "loss": 0.1152, "step": 9320 }, { "grad_norm": 0.6408978700637817, "learning_rate": 9.948852326507672e-05, "loss": 0.0941, "step": 9330 }, { "grad_norm": 0.8277348279953003, "learning_rate": 9.948616157128694e-05, "loss": 0.0865, "step": 9340 }, { "grad_norm": 0.4767294228076935, "learning_rate": 9.948379446576493e-05, "loss": 0.0812, "step": 9350 }, { "grad_norm": 0.5744130611419678, "learning_rate": 9.948142194876952e-05, "loss": 0.0796, "step": 9360 }, { "grad_norm": 0.4656582772731781, "learning_rate": 9.947904402056024e-05, "loss": 0.0964, "step": 9370 }, { "grad_norm": 0.7713728547096252, "learning_rate": 9.947666068139708e-05, "loss": 0.105, "step": 9380 }, { "grad_norm": 0.6988658308982849, "learning_rate": 9.947427193154071e-05, "loss": 0.0856, "step": 9390 }, { "grad_norm": 0.5721918344497681, "learning_rate": 9.947187777125233e-05, "loss": 0.1077, "step": 9400 }, { "grad_norm": 0.5009539127349854, "learning_rate": 9.946947820079377e-05, "loss": 0.0844, "step": 9410 }, { "grad_norm": 0.7176943421363831, "learning_rate": 9.946707322042747e-05, "loss": 0.0881, "step": 9420 }, { "grad_norm": 0.5430594682693481, "learning_rate": 9.94646628304164e-05, "loss": 0.0943, "step": 9430 }, { "grad_norm": 0.6054408550262451, "learning_rate": 9.946224703102418e-05, "loss": 0.0945, "step": 9440 }, { "grad_norm": 0.5622678399085999, "learning_rate": 9.945982582251498e-05, "loss": 0.0808, "step": 9450 }, { "grad_norm": 0.7575258016586304, "learning_rate": 9.94573992051536e-05, "loss": 0.0797, "step": 9460 }, { "grad_norm": 0.575436532497406, "learning_rate": 9.94549671792054e-05, "loss": 0.0903, "step": 9470 }, { "grad_norm": 0.8145809173583984, "learning_rate": 9.945252974493635e-05, "loss": 0.0828, "step": 9480 }, { "grad_norm": 0.5354515910148621, "learning_rate": 9.9450086902613e-05, "loss": 0.0862, "step": 9490 }, { "grad_norm": 0.4464588761329651, "learning_rate": 9.944763865250248e-05, "loss": 0.0761, "step": 9500 }, { "grad_norm": 0.5739827156066895, "learning_rate": 9.944518499487254e-05, "loss": 0.077, "step": 9510 }, { "grad_norm": 0.5676437616348267, "learning_rate": 9.944272592999151e-05, "loss": 0.0985, "step": 9520 }, { "grad_norm": 0.6252583265304565, "learning_rate": 9.94402614581283e-05, "loss": 0.0874, "step": 9530 }, { "grad_norm": 0.6852480173110962, "learning_rate": 9.943779157955244e-05, "loss": 0.0914, "step": 9540 }, { "grad_norm": 0.5487833023071289, "learning_rate": 9.943531629453403e-05, "loss": 0.0841, "step": 9550 }, { "grad_norm": 0.7214652895927429, "learning_rate": 9.943283560334375e-05, "loss": 0.095, "step": 9560 }, { "grad_norm": 0.6364503502845764, "learning_rate": 9.943034950625288e-05, "loss": 0.0947, "step": 9570 }, { "grad_norm": 0.646955668926239, "learning_rate": 9.942785800353332e-05, "loss": 0.0908, "step": 9580 }, { "grad_norm": 0.6002282500267029, "learning_rate": 9.942536109545751e-05, "loss": 0.0744, "step": 9590 }, { "grad_norm": 0.6362773180007935, "learning_rate": 9.942285878229853e-05, "loss": 0.0781, "step": 9600 }, { "grad_norm": 0.8537871241569519, "learning_rate": 9.942035106433001e-05, "loss": 0.0788, "step": 9610 }, { "grad_norm": 0.6363987922668457, "learning_rate": 9.94178379418262e-05, "loss": 0.0906, "step": 9620 }, { "grad_norm": 0.8278512358665466, "learning_rate": 9.941531941506194e-05, "loss": 0.0757, "step": 9630 }, { "grad_norm": 0.5313975811004639, "learning_rate": 9.941279548431263e-05, "loss": 0.076, "step": 9640 }, { "grad_norm": 0.4507749378681183, "learning_rate": 9.941026614985431e-05, "loss": 0.0853, "step": 9650 }, { "grad_norm": 0.5015511512756348, "learning_rate": 9.940773141196357e-05, "loss": 0.0787, "step": 9660 }, { "grad_norm": 0.6032697558403015, "learning_rate": 9.94051912709176e-05, "loss": 0.0801, "step": 9670 }, { "grad_norm": 1.1132056713104248, "learning_rate": 9.940264572699421e-05, "loss": 0.0712, "step": 9680 }, { "grad_norm": 0.5808382630348206, "learning_rate": 9.940009478047174e-05, "loss": 0.0775, "step": 9690 }, { "grad_norm": 0.8695408701896667, "learning_rate": 9.939753843162918e-05, "loss": 0.0763, "step": 9700 }, { "grad_norm": 0.5962451696395874, "learning_rate": 9.939497668074609e-05, "loss": 0.0744, "step": 9710 }, { "grad_norm": 0.6342706084251404, "learning_rate": 9.93924095281026e-05, "loss": 0.089, "step": 9720 }, { "grad_norm": 0.5514721870422363, "learning_rate": 9.938983697397948e-05, "loss": 0.0917, "step": 9730 }, { "grad_norm": 0.824354350566864, "learning_rate": 9.938725901865805e-05, "loss": 0.0852, "step": 9740 }, { "grad_norm": 0.5597810745239258, "learning_rate": 9.93846756624202e-05, "loss": 0.0689, "step": 9750 }, { "grad_norm": 0.4741126298904419, "learning_rate": 9.938208690554849e-05, "loss": 0.0838, "step": 9760 }, { "grad_norm": 0.6660110950469971, "learning_rate": 9.9379492748326e-05, "loss": 0.0888, "step": 9770 }, { "grad_norm": 0.5458052158355713, "learning_rate": 9.937689319103641e-05, "loss": 0.0813, "step": 9780 }, { "grad_norm": 0.6220511198043823, "learning_rate": 9.937428823396404e-05, "loss": 0.0778, "step": 9790 }, { "grad_norm": 0.5495402216911316, "learning_rate": 9.937167787739372e-05, "loss": 0.0618, "step": 9800 }, { "grad_norm": 0.5899592638015747, "learning_rate": 9.936906212161095e-05, "loss": 0.0692, "step": 9810 }, { "grad_norm": 0.5436626672744751, "learning_rate": 9.936644096690176e-05, "loss": 0.0864, "step": 9820 }, { "grad_norm": 0.5026155114173889, "learning_rate": 9.936381441355282e-05, "loss": 0.0831, "step": 9830 }, { "grad_norm": 0.5765945315361023, "learning_rate": 9.936118246185136e-05, "loss": 0.0919, "step": 9840 }, { "grad_norm": 0.8476254343986511, "learning_rate": 9.935854511208518e-05, "loss": 0.0876, "step": 9850 }, { "grad_norm": 0.5517253279685974, "learning_rate": 9.935590236454272e-05, "loss": 0.0802, "step": 9860 }, { "grad_norm": 0.5408807992935181, "learning_rate": 9.935325421951298e-05, "loss": 0.0716, "step": 9870 }, { "grad_norm": 0.539560854434967, "learning_rate": 9.935060067728557e-05, "loss": 0.0757, "step": 9880 }, { "grad_norm": 0.494273841381073, "learning_rate": 9.934794173815067e-05, "loss": 0.0629, "step": 9890 }, { "grad_norm": 0.48403140902519226, "learning_rate": 9.934527740239906e-05, "loss": 0.0835, "step": 9900 }, { "grad_norm": 0.49516528844833374, "learning_rate": 9.934260767032209e-05, "loss": 0.0917, "step": 9910 }, { "grad_norm": 0.6511855721473694, "learning_rate": 9.933993254221172e-05, "loss": 0.0832, "step": 9920 }, { "grad_norm": 0.8444227576255798, "learning_rate": 9.933725201836053e-05, "loss": 0.0776, "step": 9930 }, { "grad_norm": 0.7767167687416077, "learning_rate": 9.933456609906162e-05, "loss": 0.0798, "step": 9940 }, { "grad_norm": 0.5875813961029053, "learning_rate": 9.933187478460875e-05, "loss": 0.0781, "step": 9950 }, { "grad_norm": 0.48895856738090515, "learning_rate": 9.93291780752962e-05, "loss": 0.0763, "step": 9960 }, { "grad_norm": 0.45695731043815613, "learning_rate": 9.932647597141893e-05, "loss": 0.0742, "step": 9970 }, { "grad_norm": 0.5026891827583313, "learning_rate": 9.932376847327239e-05, "loss": 0.0815, "step": 9980 }, { "grad_norm": 0.48540690541267395, "learning_rate": 9.932105558115268e-05, "loss": 0.0882, "step": 9990 }, { "grad_norm": 0.6621312499046326, "learning_rate": 9.931833729535651e-05, "loss": 0.0831, "step": 10000 }, { "grad_norm": 0.516537070274353, "learning_rate": 9.931561361618111e-05, "loss": 0.0862, "step": 10010 }, { "grad_norm": 0.4320722222328186, "learning_rate": 9.931288454392435e-05, "loss": 0.0854, "step": 10020 }, { "grad_norm": 0.7458211183547974, "learning_rate": 9.931015007888467e-05, "loss": 0.0851, "step": 10030 }, { "grad_norm": 0.6426944136619568, "learning_rate": 9.930741022136112e-05, "loss": 0.0902, "step": 10040 }, { "grad_norm": 0.45064058899879456, "learning_rate": 9.930466497165333e-05, "loss": 0.0787, "step": 10050 }, { "grad_norm": 0.688895046710968, "learning_rate": 9.93019143300615e-05, "loss": 0.0862, "step": 10060 }, { "grad_norm": 0.429140567779541, "learning_rate": 9.929915829688644e-05, "loss": 0.0751, "step": 10070 }, { "grad_norm": 0.4878551661968231, "learning_rate": 9.929639687242955e-05, "loss": 0.0829, "step": 10080 }, { "grad_norm": 0.5047134160995483, "learning_rate": 9.929363005699281e-05, "loss": 0.0922, "step": 10090 }, { "grad_norm": 0.4964630603790283, "learning_rate": 9.92908578508788e-05, "loss": 0.0808, "step": 10100 }, { "grad_norm": 0.5457819104194641, "learning_rate": 9.928808025439069e-05, "loss": 0.0832, "step": 10110 }, { "grad_norm": 0.5602174401283264, "learning_rate": 9.928529726783223e-05, "loss": 0.0733, "step": 10120 }, { "grad_norm": 0.819243848323822, "learning_rate": 9.928250889150774e-05, "loss": 0.0927, "step": 10130 }, { "grad_norm": 0.833644688129425, "learning_rate": 9.92797151257222e-05, "loss": 0.0862, "step": 10140 }, { "grad_norm": 0.5266880989074707, "learning_rate": 9.927691597078108e-05, "loss": 0.0838, "step": 10150 }, { "grad_norm": 0.6420581936836243, "learning_rate": 9.927411142699053e-05, "loss": 0.0818, "step": 10160 }, { "grad_norm": 0.49410003423690796, "learning_rate": 9.927130149465725e-05, "loss": 0.084, "step": 10170 }, { "grad_norm": 0.6869926452636719, "learning_rate": 9.92684861740885e-05, "loss": 0.0998, "step": 10180 }, { "grad_norm": 0.7708936333656311, "learning_rate": 9.926566546559217e-05, "loss": 0.0999, "step": 10190 }, { "grad_norm": 0.5446496605873108, "learning_rate": 9.926283936947673e-05, "loss": 0.08, "step": 10200 }, { "grad_norm": 0.5995550751686096, "learning_rate": 9.926000788605126e-05, "loss": 0.0865, "step": 10210 }, { "grad_norm": 0.42985326051712036, "learning_rate": 9.92571710156254e-05, "loss": 0.0831, "step": 10220 }, { "grad_norm": 0.6566508412361145, "learning_rate": 9.925432875850936e-05, "loss": 0.0926, "step": 10230 }, { "grad_norm": 0.5227659344673157, "learning_rate": 9.925148111501396e-05, "loss": 0.088, "step": 10240 }, { "grad_norm": 0.47822073101997375, "learning_rate": 9.924862808545066e-05, "loss": 0.0918, "step": 10250 }, { "grad_norm": 0.5251911878585815, "learning_rate": 9.924576967013141e-05, "loss": 0.0692, "step": 10260 }, { "grad_norm": 0.43617871403694153, "learning_rate": 9.924290586936887e-05, "loss": 0.0837, "step": 10270 }, { "grad_norm": 0.6751570105552673, "learning_rate": 9.924003668347614e-05, "loss": 0.0819, "step": 10280 }, { "grad_norm": 0.6115188598632812, "learning_rate": 9.923716211276704e-05, "loss": 0.0882, "step": 10290 }, { "grad_norm": 0.6012777090072632, "learning_rate": 9.923428215755594e-05, "loss": 0.0759, "step": 10300 }, { "grad_norm": 0.6580430269241333, "learning_rate": 9.923139681815775e-05, "loss": 0.0811, "step": 10310 }, { "grad_norm": 0.7115763425827026, "learning_rate": 9.922850609488801e-05, "loss": 0.0753, "step": 10320 }, { "grad_norm": 0.5205172896385193, "learning_rate": 9.922560998806287e-05, "loss": 0.0818, "step": 10330 }, { "grad_norm": 0.6962857842445374, "learning_rate": 9.922270849799905e-05, "loss": 0.0874, "step": 10340 }, { "grad_norm": 0.5513988137245178, "learning_rate": 9.92198016250138e-05, "loss": 0.0862, "step": 10350 }, { "grad_norm": 0.6360200047492981, "learning_rate": 9.921688936942506e-05, "loss": 0.0688, "step": 10360 }, { "grad_norm": 0.4752225875854492, "learning_rate": 9.921397173155129e-05, "loss": 0.0792, "step": 10370 }, { "grad_norm": 0.44294652342796326, "learning_rate": 9.921104871171157e-05, "loss": 0.073, "step": 10380 }, { "grad_norm": 0.5096495151519775, "learning_rate": 9.920812031022554e-05, "loss": 0.0918, "step": 10390 }, { "grad_norm": 0.5660115480422974, "learning_rate": 9.920518652741348e-05, "loss": 0.0844, "step": 10400 }, { "grad_norm": 0.8108341097831726, "learning_rate": 9.920224736359618e-05, "loss": 0.0938, "step": 10410 }, { "grad_norm": 0.4943835437297821, "learning_rate": 9.91993028190951e-05, "loss": 0.0993, "step": 10420 }, { "grad_norm": 0.6355475783348083, "learning_rate": 9.919635289423222e-05, "loss": 0.0862, "step": 10430 }, { "grad_norm": 0.6573145985603333, "learning_rate": 9.919339758933015e-05, "loss": 0.0797, "step": 10440 }, { "grad_norm": 0.4943205416202545, "learning_rate": 9.919043690471209e-05, "loss": 0.0789, "step": 10450 }, { "grad_norm": 0.5130327939987183, "learning_rate": 9.91874708407018e-05, "loss": 0.0834, "step": 10460 }, { "grad_norm": 0.6741731762886047, "learning_rate": 9.918449939762367e-05, "loss": 0.0898, "step": 10470 }, { "grad_norm": 0.5514954924583435, "learning_rate": 9.91815225758026e-05, "loss": 0.0822, "step": 10480 }, { "grad_norm": 0.6400750279426575, "learning_rate": 9.917854037556419e-05, "loss": 0.0856, "step": 10490 }, { "grad_norm": 0.5460151433944702, "learning_rate": 9.917555279723454e-05, "loss": 0.0854, "step": 10500 }, { "grad_norm": 0.6682739853858948, "learning_rate": 9.917255984114036e-05, "loss": 0.1023, "step": 10510 }, { "grad_norm": 0.4153427183628082, "learning_rate": 9.916956150760896e-05, "loss": 0.0709, "step": 10520 }, { "grad_norm": 0.37631964683532715, "learning_rate": 9.916655779696826e-05, "loss": 0.0812, "step": 10530 }, { "grad_norm": 0.40171900391578674, "learning_rate": 9.916354870954671e-05, "loss": 0.073, "step": 10540 }, { "grad_norm": 0.4713350832462311, "learning_rate": 9.91605342456734e-05, "loss": 0.0753, "step": 10550 }, { "grad_norm": 0.7660355567932129, "learning_rate": 9.915751440567795e-05, "loss": 0.0789, "step": 10560 }, { "grad_norm": 0.5332461595535278, "learning_rate": 9.915448918989066e-05, "loss": 0.0735, "step": 10570 }, { "grad_norm": 0.46113747358322144, "learning_rate": 9.915145859864232e-05, "loss": 0.0729, "step": 10580 }, { "grad_norm": 0.6291323304176331, "learning_rate": 9.914842263226437e-05, "loss": 0.0887, "step": 10590 }, { "grad_norm": 0.42865434288978577, "learning_rate": 9.914538129108882e-05, "loss": 0.0853, "step": 10600 }, { "grad_norm": 0.5176200270652771, "learning_rate": 9.914233457544825e-05, "loss": 0.1155, "step": 10610 }, { "grad_norm": 0.8097776770591736, "learning_rate": 9.913928248567586e-05, "loss": 0.1142, "step": 10620 }, { "grad_norm": 0.5262043476104736, "learning_rate": 9.913622502210542e-05, "loss": 0.0752, "step": 10630 }, { "grad_norm": 0.5095224380493164, "learning_rate": 9.913316218507128e-05, "loss": 0.0949, "step": 10640 }, { "grad_norm": 0.5562161207199097, "learning_rate": 9.91300939749084e-05, "loss": 0.0871, "step": 10650 }, { "grad_norm": 0.6482716798782349, "learning_rate": 9.91270203919523e-05, "loss": 0.0934, "step": 10660 }, { "grad_norm": 0.6963366270065308, "learning_rate": 9.912394143653912e-05, "loss": 0.0855, "step": 10670 }, { "grad_norm": 0.5846670269966125, "learning_rate": 9.912085710900555e-05, "loss": 0.0789, "step": 10680 }, { "grad_norm": 0.4007623493671417, "learning_rate": 9.911776740968892e-05, "loss": 0.0763, "step": 10690 }, { "grad_norm": 0.5952415466308594, "learning_rate": 9.911467233892709e-05, "loss": 0.104, "step": 10700 }, { "grad_norm": 0.5176829099655151, "learning_rate": 9.911157189705853e-05, "loss": 0.0914, "step": 10710 }, { "grad_norm": 0.49960756301879883, "learning_rate": 9.910846608442229e-05, "loss": 0.0808, "step": 10720 }, { "grad_norm": 0.6218332648277283, "learning_rate": 9.910535490135805e-05, "loss": 0.0733, "step": 10730 }, { "grad_norm": 0.7229193449020386, "learning_rate": 9.910223834820603e-05, "loss": 0.0837, "step": 10740 }, { "grad_norm": 0.7969045639038086, "learning_rate": 9.909911642530703e-05, "loss": 0.0923, "step": 10750 }, { "grad_norm": 0.44849318265914917, "learning_rate": 9.909598913300249e-05, "loss": 0.096, "step": 10760 }, { "grad_norm": 0.9165327548980713, "learning_rate": 9.909285647163438e-05, "loss": 0.098, "step": 10770 }, { "grad_norm": 0.6488931179046631, "learning_rate": 9.908971844154531e-05, "loss": 0.081, "step": 10780 }, { "grad_norm": 0.5177534818649292, "learning_rate": 9.908657504307843e-05, "loss": 0.0786, "step": 10790 }, { "grad_norm": 0.7712331414222717, "learning_rate": 9.908342627657751e-05, "loss": 0.095, "step": 10800 }, { "grad_norm": 0.5943015217781067, "learning_rate": 9.908027214238689e-05, "loss": 0.0907, "step": 10810 }, { "grad_norm": 0.5738390684127808, "learning_rate": 9.90771126408515e-05, "loss": 0.0997, "step": 10820 }, { "grad_norm": 0.5010881423950195, "learning_rate": 9.907394777231685e-05, "loss": 0.0947, "step": 10830 }, { "grad_norm": 0.5991066694259644, "learning_rate": 9.907077753712905e-05, "loss": 0.0964, "step": 10840 }, { "grad_norm": 0.7017480134963989, "learning_rate": 9.906760193563482e-05, "loss": 0.0953, "step": 10850 }, { "grad_norm": 0.473619669675827, "learning_rate": 9.906442096818139e-05, "loss": 0.0798, "step": 10860 }, { "grad_norm": 0.8293148875236511, "learning_rate": 9.906123463511665e-05, "loss": 0.1086, "step": 10870 }, { "grad_norm": 0.5454365611076355, "learning_rate": 9.905804293678907e-05, "loss": 0.0856, "step": 10880 }, { "grad_norm": 0.627204418182373, "learning_rate": 9.905484587354766e-05, "loss": 0.0859, "step": 10890 }, { "grad_norm": 0.4462340772151947, "learning_rate": 9.905164344574205e-05, "loss": 0.0809, "step": 10900 }, { "grad_norm": 0.44011393189430237, "learning_rate": 9.904843565372248e-05, "loss": 0.0703, "step": 10910 }, { "grad_norm": 0.5496638417243958, "learning_rate": 9.904522249783972e-05, "loss": 0.0912, "step": 10920 }, { "grad_norm": 0.6039366722106934, "learning_rate": 9.904200397844517e-05, "loss": 0.1126, "step": 10930 }, { "grad_norm": 0.6299394965171814, "learning_rate": 9.903878009589078e-05, "loss": 0.0942, "step": 10940 }, { "grad_norm": 0.6623157262802124, "learning_rate": 9.903555085052915e-05, "loss": 0.0783, "step": 10950 }, { "grad_norm": 0.6253601908683777, "learning_rate": 9.903231624271338e-05, "loss": 0.0807, "step": 10960 }, { "grad_norm": 0.4816386103630066, "learning_rate": 9.902907627279724e-05, "loss": 0.08, "step": 10970 }, { "grad_norm": 0.5095792412757874, "learning_rate": 9.902583094113504e-05, "loss": 0.078, "step": 10980 }, { "grad_norm": 0.6091337203979492, "learning_rate": 9.902258024808168e-05, "loss": 0.0855, "step": 10990 }, { "grad_norm": 0.5693967342376709, "learning_rate": 9.901932419399264e-05, "loss": 0.0825, "step": 11000 }, { "grad_norm": 0.604579508304596, "learning_rate": 9.9016062779224e-05, "loss": 0.0944, "step": 11010 }, { "grad_norm": 0.5930746793746948, "learning_rate": 9.901279600413242e-05, "loss": 0.0873, "step": 11020 }, { "grad_norm": 0.5976130366325378, "learning_rate": 9.900952386907518e-05, "loss": 0.0769, "step": 11030 }, { "grad_norm": 0.62961745262146, "learning_rate": 9.90062463744101e-05, "loss": 0.071, "step": 11040 }, { "grad_norm": 0.5194308161735535, "learning_rate": 9.900296352049558e-05, "loss": 0.063, "step": 11050 }, { "grad_norm": 0.43834224343299866, "learning_rate": 9.899967530769065e-05, "loss": 0.0858, "step": 11060 }, { "grad_norm": 0.5944440364837646, "learning_rate": 9.899638173635489e-05, "loss": 0.0832, "step": 11070 }, { "grad_norm": 0.4522581994533539, "learning_rate": 9.899308280684849e-05, "loss": 0.079, "step": 11080 }, { "grad_norm": 0.5712122917175293, "learning_rate": 9.898977851953222e-05, "loss": 0.0831, "step": 11090 }, { "grad_norm": 0.601091742515564, "learning_rate": 9.898646887476741e-05, "loss": 0.0901, "step": 11100 }, { "grad_norm": 0.5317007303237915, "learning_rate": 9.898315387291603e-05, "loss": 0.0983, "step": 11110 }, { "grad_norm": 0.6720441579818726, "learning_rate": 9.89798335143406e-05, "loss": 0.0842, "step": 11120 }, { "grad_norm": 0.5066652894020081, "learning_rate": 9.897650779940419e-05, "loss": 0.0972, "step": 11130 }, { "grad_norm": 0.41278499364852905, "learning_rate": 9.897317672847054e-05, "loss": 0.0825, "step": 11140 }, { "grad_norm": 0.5975252985954285, "learning_rate": 9.89698403019039e-05, "loss": 0.084, "step": 11150 }, { "grad_norm": 0.5463293194770813, "learning_rate": 9.896649852006917e-05, "loss": 0.0811, "step": 11160 }, { "grad_norm": 0.7491188645362854, "learning_rate": 9.896315138333177e-05, "loss": 0.0825, "step": 11170 }, { "grad_norm": 0.7451046109199524, "learning_rate": 9.895979889205774e-05, "loss": 0.077, "step": 11180 }, { "grad_norm": 0.6440878510475159, "learning_rate": 9.895644104661372e-05, "loss": 0.0778, "step": 11190 }, { "grad_norm": 0.7182832360267639, "learning_rate": 9.895307784736691e-05, "loss": 0.0903, "step": 11200 }, { "grad_norm": 0.5519384741783142, "learning_rate": 9.894970929468512e-05, "loss": 0.081, "step": 11210 }, { "grad_norm": 0.5525704026222229, "learning_rate": 9.89463353889367e-05, "loss": 0.0918, "step": 11220 }, { "grad_norm": 0.5765106678009033, "learning_rate": 9.894295613049065e-05, "loss": 0.097, "step": 11230 }, { "grad_norm": 0.6272479891777039, "learning_rate": 9.893957151971649e-05, "loss": 0.0846, "step": 11240 }, { "grad_norm": 0.5863951444625854, "learning_rate": 9.893618155698436e-05, "loss": 0.0834, "step": 11250 }, { "grad_norm": 0.5036810636520386, "learning_rate": 9.8932786242665e-05, "loss": 0.0744, "step": 11260 }, { "grad_norm": 0.47056084871292114, "learning_rate": 9.89293855771297e-05, "loss": 0.0882, "step": 11270 }, { "grad_norm": 0.6462076306343079, "learning_rate": 9.892597956075036e-05, "loss": 0.0877, "step": 11280 }, { "grad_norm": 0.4810498058795929, "learning_rate": 9.892256819389947e-05, "loss": 0.0735, "step": 11290 }, { "grad_norm": 0.5351367592811584, "learning_rate": 9.891915147695006e-05, "loss": 0.0729, "step": 11300 }, { "grad_norm": 0.684328556060791, "learning_rate": 9.891572941027577e-05, "loss": 0.0837, "step": 11310 }, { "grad_norm": 0.5485689043998718, "learning_rate": 9.89123019942509e-05, "loss": 0.0762, "step": 11320 }, { "grad_norm": 0.5012220740318298, "learning_rate": 9.89088692292502e-05, "loss": 0.0638, "step": 11330 }, { "grad_norm": 0.5707051753997803, "learning_rate": 9.89054311156491e-05, "loss": 0.0663, "step": 11340 }, { "grad_norm": 0.529293417930603, "learning_rate": 9.890198765382357e-05, "loss": 0.1025, "step": 11350 }, { "grad_norm": 0.7136170268058777, "learning_rate": 9.889853884415021e-05, "loss": 0.084, "step": 11360 }, { "grad_norm": 0.4247002899646759, "learning_rate": 9.889508468700614e-05, "loss": 0.0739, "step": 11370 }, { "grad_norm": 0.4208688735961914, "learning_rate": 9.889162518276915e-05, "loss": 0.0738, "step": 11380 }, { "grad_norm": 0.4527597725391388, "learning_rate": 9.888816033181752e-05, "loss": 0.0824, "step": 11390 }, { "grad_norm": 0.5641005039215088, "learning_rate": 9.888469013453018e-05, "loss": 0.0988, "step": 11400 }, { "grad_norm": 0.813460648059845, "learning_rate": 9.888121459128663e-05, "loss": 0.0837, "step": 11410 }, { "grad_norm": 0.5169568061828613, "learning_rate": 9.887773370246693e-05, "loss": 0.1002, "step": 11420 }, { "grad_norm": 0.5852852463722229, "learning_rate": 9.887424746845177e-05, "loss": 0.0883, "step": 11430 }, { "grad_norm": 0.4748859703540802, "learning_rate": 9.887075588962239e-05, "loss": 0.0809, "step": 11440 }, { "grad_norm": 0.475252628326416, "learning_rate": 9.88672589663606e-05, "loss": 0.0837, "step": 11450 }, { "grad_norm": 0.4868931472301483, "learning_rate": 9.886375669904886e-05, "loss": 0.0908, "step": 11460 }, { "grad_norm": 0.5544337034225464, "learning_rate": 9.886024908807014e-05, "loss": 0.081, "step": 11470 }, { "grad_norm": 0.8217172026634216, "learning_rate": 9.885673613380806e-05, "loss": 0.0766, "step": 11480 }, { "grad_norm": 0.6146049499511719, "learning_rate": 9.885321783664676e-05, "loss": 0.0805, "step": 11490 }, { "grad_norm": 0.5955825448036194, "learning_rate": 9.884969419697101e-05, "loss": 0.0806, "step": 11500 }, { "grad_norm": 0.6638538837432861, "learning_rate": 9.884616521516614e-05, "loss": 0.0776, "step": 11510 }, { "grad_norm": 0.48598289489746094, "learning_rate": 9.88426308916181e-05, "loss": 0.0822, "step": 11520 }, { "grad_norm": 0.5924763679504395, "learning_rate": 9.883909122671335e-05, "loss": 0.0723, "step": 11530 }, { "grad_norm": 0.5727740526199341, "learning_rate": 9.883554622083904e-05, "loss": 0.0873, "step": 11540 }, { "grad_norm": 0.5983588695526123, "learning_rate": 9.88319958743828e-05, "loss": 0.0924, "step": 11550 }, { "grad_norm": 0.6990658640861511, "learning_rate": 9.882844018773291e-05, "loss": 0.0895, "step": 11560 }, { "grad_norm": 0.6099040508270264, "learning_rate": 9.882487916127823e-05, "loss": 0.0885, "step": 11570 }, { "grad_norm": 0.7639977335929871, "learning_rate": 9.882131279540815e-05, "loss": 0.1097, "step": 11580 }, { "grad_norm": 0.6457202434539795, "learning_rate": 9.881774109051271e-05, "loss": 0.0981, "step": 11590 }, { "grad_norm": 0.521210789680481, "learning_rate": 9.881416404698252e-05, "loss": 0.0931, "step": 11600 }, { "grad_norm": 0.44795116782188416, "learning_rate": 9.881058166520873e-05, "loss": 0.0918, "step": 11610 }, { "grad_norm": 0.6452341079711914, "learning_rate": 9.880699394558311e-05, "loss": 0.1049, "step": 11620 }, { "grad_norm": 0.57034832239151, "learning_rate": 9.880340088849801e-05, "loss": 0.0874, "step": 11630 }, { "grad_norm": 0.5036583542823792, "learning_rate": 9.879980249434637e-05, "loss": 0.0811, "step": 11640 }, { "grad_norm": 1.0690698623657227, "learning_rate": 9.879619876352168e-05, "loss": 0.0903, "step": 11650 }, { "grad_norm": 0.483304888010025, "learning_rate": 9.879258969641809e-05, "loss": 0.0837, "step": 11660 }, { "grad_norm": 0.6132918000221252, "learning_rate": 9.878897529343023e-05, "loss": 0.0743, "step": 11670 }, { "grad_norm": 0.5660738348960876, "learning_rate": 9.878535555495338e-05, "loss": 0.0818, "step": 11680 }, { "grad_norm": 0.6829743385314941, "learning_rate": 9.87817304813834e-05, "loss": 0.0934, "step": 11690 }, { "grad_norm": 0.4344624876976013, "learning_rate": 9.877810007311671e-05, "loss": 0.0798, "step": 11700 }, { "grad_norm": 0.8933051824569702, "learning_rate": 9.877446433055035e-05, "loss": 0.0863, "step": 11710 }, { "grad_norm": 0.6534217000007629, "learning_rate": 9.877082325408191e-05, "loss": 0.0836, "step": 11720 }, { "grad_norm": 0.4363388419151306, "learning_rate": 9.876717684410954e-05, "loss": 0.0661, "step": 11730 }, { "grad_norm": 0.6308813691139221, "learning_rate": 9.876352510103204e-05, "loss": 0.078, "step": 11740 }, { "grad_norm": 0.5235469341278076, "learning_rate": 9.875986802524875e-05, "loss": 0.0773, "step": 11750 }, { "grad_norm": 0.5535505414009094, "learning_rate": 9.87562056171596e-05, "loss": 0.0707, "step": 11760 }, { "grad_norm": 0.5149846076965332, "learning_rate": 9.875253787716511e-05, "loss": 0.0826, "step": 11770 }, { "grad_norm": 0.6763333082199097, "learning_rate": 9.874886480566637e-05, "loss": 0.0884, "step": 11780 }, { "grad_norm": 0.5187081098556519, "learning_rate": 9.874518640306507e-05, "loss": 0.0767, "step": 11790 }, { "grad_norm": 0.5387938022613525, "learning_rate": 9.874150266976347e-05, "loss": 0.0867, "step": 11800 }, { "grad_norm": 0.4834096133708954, "learning_rate": 9.873781360616443e-05, "loss": 0.0933, "step": 11810 }, { "grad_norm": 0.7360605001449585, "learning_rate": 9.873411921267137e-05, "loss": 0.0785, "step": 11820 }, { "grad_norm": 0.5921955704689026, "learning_rate": 9.873041948968829e-05, "loss": 0.0864, "step": 11830 }, { "grad_norm": 0.5398076772689819, "learning_rate": 9.872671443761981e-05, "loss": 0.0805, "step": 11840 }, { "grad_norm": 0.49753114581108093, "learning_rate": 9.872300405687109e-05, "loss": 0.0776, "step": 11850 }, { "grad_norm": 0.6207697987556458, "learning_rate": 9.871928834784792e-05, "loss": 0.0932, "step": 11860 }, { "grad_norm": 0.5310640931129456, "learning_rate": 9.871556731095661e-05, "loss": 0.094, "step": 11870 }, { "grad_norm": 0.7285999059677124, "learning_rate": 9.871184094660411e-05, "loss": 0.0967, "step": 11880 }, { "grad_norm": 0.4897264242172241, "learning_rate": 9.870810925519791e-05, "loss": 0.0781, "step": 11890 }, { "grad_norm": 0.4513561725616455, "learning_rate": 9.870437223714612e-05, "loss": 0.0796, "step": 11900 }, { "grad_norm": 0.4380989372730255, "learning_rate": 9.87006298928574e-05, "loss": 0.085, "step": 11910 }, { "grad_norm": 0.4259394407272339, "learning_rate": 9.869688222274103e-05, "loss": 0.0689, "step": 11920 }, { "grad_norm": 0.5459949970245361, "learning_rate": 9.869312922720681e-05, "loss": 0.0826, "step": 11930 }, { "grad_norm": 0.3801622986793518, "learning_rate": 9.868937090666521e-05, "loss": 0.0799, "step": 11940 }, { "grad_norm": 0.4986346662044525, "learning_rate": 9.86856072615272e-05, "loss": 0.0698, "step": 11950 }, { "grad_norm": 0.6687256693840027, "learning_rate": 9.868183829220438e-05, "loss": 0.0731, "step": 11960 }, { "grad_norm": 0.7880929708480835, "learning_rate": 9.867806399910893e-05, "loss": 0.0722, "step": 11970 }, { "grad_norm": 0.602558970451355, "learning_rate": 9.867428438265356e-05, "loss": 0.0775, "step": 11980 }, { "grad_norm": 1.3546957969665527, "learning_rate": 9.867049944325165e-05, "loss": 0.0907, "step": 11990 }, { "grad_norm": 0.6055338382720947, "learning_rate": 9.86667091813171e-05, "loss": 0.101, "step": 12000 }, { "grad_norm": 0.6841878890991211, "learning_rate": 9.866291359726438e-05, "loss": 0.0919, "step": 12010 }, { "grad_norm": 0.6077026724815369, "learning_rate": 9.865911269150861e-05, "loss": 0.0835, "step": 12020 }, { "grad_norm": 0.4594138264656067, "learning_rate": 9.865530646446544e-05, "loss": 0.0758, "step": 12030 }, { "grad_norm": 0.5658211708068848, "learning_rate": 9.86514949165511e-05, "loss": 0.0861, "step": 12040 }, { "grad_norm": 0.4562493860721588, "learning_rate": 9.864767804818243e-05, "loss": 0.0806, "step": 12050 }, { "grad_norm": 0.7539212703704834, "learning_rate": 9.86438558597768e-05, "loss": 0.093, "step": 12060 }, { "grad_norm": 0.4500163793563843, "learning_rate": 9.864002835175225e-05, "loss": 0.0937, "step": 12070 }, { "grad_norm": 0.60288006067276, "learning_rate": 9.863619552452734e-05, "loss": 0.0725, "step": 12080 }, { "grad_norm": 0.556868851184845, "learning_rate": 9.863235737852119e-05, "loss": 0.0625, "step": 12090 }, { "grad_norm": 0.8608462810516357, "learning_rate": 9.862851391415356e-05, "loss": 0.0757, "step": 12100 }, { "grad_norm": 0.4816674590110779, "learning_rate": 9.862466513184477e-05, "loss": 0.0662, "step": 12110 }, { "grad_norm": 0.5961044430732727, "learning_rate": 9.86208110320157e-05, "loss": 0.0718, "step": 12120 }, { "grad_norm": 0.5063208341598511, "learning_rate": 9.861695161508784e-05, "loss": 0.075, "step": 12130 }, { "grad_norm": 0.6242472529411316, "learning_rate": 9.861308688148324e-05, "loss": 0.0744, "step": 12140 }, { "grad_norm": 0.49087828397750854, "learning_rate": 9.860921683162455e-05, "loss": 0.0743, "step": 12150 }, { "grad_norm": 0.5700323581695557, "learning_rate": 9.860534146593499e-05, "loss": 0.0712, "step": 12160 }, { "grad_norm": 0.5374398827552795, "learning_rate": 9.860146078483836e-05, "loss": 0.0697, "step": 12170 }, { "grad_norm": 0.6397815346717834, "learning_rate": 9.859757478875905e-05, "loss": 0.0788, "step": 12180 }, { "grad_norm": 0.5545065999031067, "learning_rate": 9.859368347812204e-05, "loss": 0.0762, "step": 12190 }, { "grad_norm": 0.5914297699928284, "learning_rate": 9.858978685335285e-05, "loss": 0.072, "step": 12200 }, { "grad_norm": 0.5797119140625, "learning_rate": 9.858588491487763e-05, "loss": 0.0946, "step": 12210 }, { "grad_norm": 0.6797155141830444, "learning_rate": 9.858197766312308e-05, "loss": 0.0692, "step": 12220 }, { "grad_norm": 0.6685819625854492, "learning_rate": 9.857806509851649e-05, "loss": 0.0766, "step": 12230 }, { "grad_norm": 0.49248912930488586, "learning_rate": 9.857414722148574e-05, "loss": 0.0835, "step": 12240 }, { "grad_norm": 0.4545009136199951, "learning_rate": 9.857022403245928e-05, "loss": 0.0914, "step": 12250 }, { "grad_norm": 0.48573940992355347, "learning_rate": 9.856629553186615e-05, "loss": 0.0873, "step": 12260 }, { "grad_norm": 0.5570688843727112, "learning_rate": 9.856236172013595e-05, "loss": 0.079, "step": 12270 }, { "grad_norm": 0.3622426688671112, "learning_rate": 9.85584225976989e-05, "loss": 0.07, "step": 12280 }, { "grad_norm": 0.7347283363342285, "learning_rate": 9.855447816498575e-05, "loss": 0.0958, "step": 12290 }, { "grad_norm": 0.6739410758018494, "learning_rate": 9.855052842242787e-05, "loss": 0.0677, "step": 12300 }, { "grad_norm": 0.6592539548873901, "learning_rate": 9.85465733704572e-05, "loss": 0.0778, "step": 12310 }, { "grad_norm": 0.6191876530647278, "learning_rate": 9.854261300950624e-05, "loss": 0.0868, "step": 12320 }, { "grad_norm": 0.47310853004455566, "learning_rate": 9.853864734000813e-05, "loss": 0.0777, "step": 12330 }, { "grad_norm": 0.4334065914154053, "learning_rate": 9.85346763623965e-05, "loss": 0.0983, "step": 12340 }, { "grad_norm": 0.5016307234764099, "learning_rate": 9.853070007710564e-05, "loss": 0.082, "step": 12350 }, { "grad_norm": 0.6650091409683228, "learning_rate": 9.85267184845704e-05, "loss": 0.0642, "step": 12360 }, { "grad_norm": 0.5416267514228821, "learning_rate": 9.852273158522616e-05, "loss": 0.0912, "step": 12370 }, { "grad_norm": 0.44643503427505493, "learning_rate": 9.851873937950896e-05, "loss": 0.0698, "step": 12380 }, { "grad_norm": 0.6392109394073486, "learning_rate": 9.851474186785537e-05, "loss": 0.086, "step": 12390 }, { "grad_norm": 0.5668228268623352, "learning_rate": 9.851073905070254e-05, "loss": 0.0845, "step": 12400 }, { "grad_norm": 0.8909609913825989, "learning_rate": 9.850673092848824e-05, "loss": 0.0921, "step": 12410 }, { "grad_norm": 0.5182873606681824, "learning_rate": 9.850271750165077e-05, "loss": 0.0686, "step": 12420 }, { "grad_norm": 0.7841343283653259, "learning_rate": 9.849869877062902e-05, "loss": 0.0889, "step": 12430 }, { "grad_norm": 0.5364990830421448, "learning_rate": 9.849467473586252e-05, "loss": 0.0921, "step": 12440 }, { "grad_norm": 0.49353450536727905, "learning_rate": 9.849064539779127e-05, "loss": 0.0785, "step": 12450 }, { "grad_norm": 0.5108940005302429, "learning_rate": 9.848661075685594e-05, "loss": 0.084, "step": 12460 }, { "grad_norm": 0.6341320276260376, "learning_rate": 9.848257081349778e-05, "loss": 0.0844, "step": 12470 }, { "grad_norm": 0.5468721985816956, "learning_rate": 9.847852556815856e-05, "loss": 0.0801, "step": 12480 }, { "grad_norm": 0.7140069603919983, "learning_rate": 9.847447502128067e-05, "loss": 0.082, "step": 12490 }, { "grad_norm": 0.5091925263404846, "learning_rate": 9.847041917330708e-05, "loss": 0.0803, "step": 12500 }, { "grad_norm": 0.5240424871444702, "learning_rate": 9.846635802468132e-05, "loss": 0.093, "step": 12510 }, { "grad_norm": 0.5008531212806702, "learning_rate": 9.84622915758475e-05, "loss": 0.0731, "step": 12520 }, { "grad_norm": 0.4017907381057739, "learning_rate": 9.845821982725034e-05, "loss": 0.0741, "step": 12530 }, { "grad_norm": 0.5694137215614319, "learning_rate": 9.845414277933514e-05, "loss": 0.0844, "step": 12540 }, { "grad_norm": 0.6281148791313171, "learning_rate": 9.845006043254771e-05, "loss": 0.0928, "step": 12550 }, { "grad_norm": 0.7290291786193848, "learning_rate": 9.844597278733451e-05, "loss": 0.0916, "step": 12560 }, { "grad_norm": 0.4454376697540283, "learning_rate": 9.844187984414259e-05, "loss": 0.0786, "step": 12570 }, { "grad_norm": 0.5022451281547546, "learning_rate": 9.84377816034195e-05, "loss": 0.0759, "step": 12580 }, { "grad_norm": 0.5686302781105042, "learning_rate": 9.843367806561345e-05, "loss": 0.0833, "step": 12590 }, { "grad_norm": 0.6162812113761902, "learning_rate": 9.842956923117317e-05, "loss": 0.0826, "step": 12600 }, { "grad_norm": 0.670318067073822, "learning_rate": 9.842545510054802e-05, "loss": 0.0781, "step": 12610 }, { "grad_norm": 0.5844078660011292, "learning_rate": 9.842133567418792e-05, "loss": 0.0713, "step": 12620 }, { "grad_norm": 0.388114869594574, "learning_rate": 9.841721095254333e-05, "loss": 0.0767, "step": 12630 }, { "grad_norm": 0.4268347918987274, "learning_rate": 9.841308093606537e-05, "loss": 0.0778, "step": 12640 }, { "grad_norm": 0.4855157136917114, "learning_rate": 9.840894562520565e-05, "loss": 0.0777, "step": 12650 }, { "grad_norm": 0.722163736820221, "learning_rate": 9.840480502041642e-05, "loss": 0.0904, "step": 12660 }, { "grad_norm": 0.6759570240974426, "learning_rate": 9.840065912215049e-05, "loss": 0.0783, "step": 12670 }, { "grad_norm": 0.6503040194511414, "learning_rate": 9.839650793086124e-05, "loss": 0.0874, "step": 12680 }, { "grad_norm": 0.3876195251941681, "learning_rate": 9.839235144700265e-05, "loss": 0.0728, "step": 12690 }, { "grad_norm": 0.47139772772789, "learning_rate": 9.838818967102926e-05, "loss": 0.0704, "step": 12700 }, { "grad_norm": 0.48805293440818787, "learning_rate": 9.83840226033962e-05, "loss": 0.076, "step": 12710 }, { "grad_norm": 0.4641490578651428, "learning_rate": 9.837985024455918e-05, "loss": 0.0768, "step": 12720 }, { "grad_norm": 0.5043613910675049, "learning_rate": 9.837567259497447e-05, "loss": 0.0755, "step": 12730 }, { "grad_norm": 0.4850758910179138, "learning_rate": 9.837148965509894e-05, "loss": 0.0732, "step": 12740 }, { "grad_norm": 0.584766685962677, "learning_rate": 9.836730142539001e-05, "loss": 0.0711, "step": 12750 }, { "grad_norm": 0.7461034059524536, "learning_rate": 9.836310790630574e-05, "loss": 0.0957, "step": 12760 }, { "grad_norm": 0.4829444885253906, "learning_rate": 9.83589090983047e-05, "loss": 0.0818, "step": 12770 }, { "grad_norm": 0.578623354434967, "learning_rate": 9.835470500184605e-05, "loss": 0.081, "step": 12780 }, { "grad_norm": 0.4859966039657593, "learning_rate": 9.835049561738957e-05, "loss": 0.0796, "step": 12790 }, { "grad_norm": 0.5657700300216675, "learning_rate": 9.834628094539558e-05, "loss": 0.0702, "step": 12800 }, { "grad_norm": 0.601646363735199, "learning_rate": 9.834206098632499e-05, "loss": 0.0768, "step": 12810 }, { "grad_norm": 0.5524895191192627, "learning_rate": 9.833783574063931e-05, "loss": 0.0869, "step": 12820 }, { "grad_norm": 0.5630719661712646, "learning_rate": 9.833360520880058e-05, "loss": 0.0827, "step": 12830 }, { "grad_norm": 0.4295395612716675, "learning_rate": 9.832936939127144e-05, "loss": 0.0753, "step": 12840 }, { "grad_norm": 0.5541462302207947, "learning_rate": 9.832512828851515e-05, "loss": 0.0781, "step": 12850 }, { "grad_norm": 0.6019662022590637, "learning_rate": 9.832088190099546e-05, "loss": 0.0743, "step": 12860 }, { "grad_norm": 0.5625163316726685, "learning_rate": 9.831663022917679e-05, "loss": 0.0803, "step": 12870 }, { "grad_norm": 0.5413423776626587, "learning_rate": 9.831237327352407e-05, "loss": 0.0667, "step": 12880 }, { "grad_norm": 0.505974292755127, "learning_rate": 9.830811103450286e-05, "loss": 0.0657, "step": 12890 }, { "grad_norm": 0.5561848878860474, "learning_rate": 9.830384351257924e-05, "loss": 0.0766, "step": 12900 }, { "grad_norm": 0.7537384033203125, "learning_rate": 9.829957070821993e-05, "loss": 0.0793, "step": 12910 }, { "grad_norm": 0.4585701823234558, "learning_rate": 9.829529262189218e-05, "loss": 0.0697, "step": 12920 }, { "grad_norm": 0.4626939594745636, "learning_rate": 9.829100925406385e-05, "loss": 0.0854, "step": 12930 }, { "grad_norm": 0.47735995054244995, "learning_rate": 9.828672060520333e-05, "loss": 0.0797, "step": 12940 }, { "grad_norm": 0.3454096019268036, "learning_rate": 9.828242667577966e-05, "loss": 0.0605, "step": 12950 }, { "grad_norm": 0.5190862417221069, "learning_rate": 9.82781274662624e-05, "loss": 0.0686, "step": 12960 }, { "grad_norm": 0.64633709192276, "learning_rate": 9.82738229771217e-05, "loss": 0.074, "step": 12970 }, { "grad_norm": 0.4933917224407196, "learning_rate": 9.826951320882829e-05, "loss": 0.0621, "step": 12980 }, { "grad_norm": 0.43880677223205566, "learning_rate": 9.826519816185351e-05, "loss": 0.0657, "step": 12990 }, { "grad_norm": 0.8313888311386108, "learning_rate": 9.826087783666921e-05, "loss": 0.0757, "step": 13000 }, { "grad_norm": 0.9943388104438782, "learning_rate": 9.825655223374787e-05, "loss": 0.0972, "step": 13010 }, { "grad_norm": 0.7819786667823792, "learning_rate": 9.825222135356253e-05, "loss": 0.1018, "step": 13020 }, { "grad_norm": 0.6547978520393372, "learning_rate": 9.82478851965868e-05, "loss": 0.0933, "step": 13030 }, { "grad_norm": 0.4551723897457123, "learning_rate": 9.82435437632949e-05, "loss": 0.0676, "step": 13040 }, { "grad_norm": 0.45861607789993286, "learning_rate": 9.823919705416158e-05, "loss": 0.0654, "step": 13050 }, { "grad_norm": 0.9488147497177124, "learning_rate": 9.82348450696622e-05, "loss": 0.103, "step": 13060 }, { "grad_norm": 0.541519820690155, "learning_rate": 9.823048781027268e-05, "loss": 0.097, "step": 13070 }, { "grad_norm": 0.6400446891784668, "learning_rate": 9.822612527646953e-05, "loss": 0.0774, "step": 13080 }, { "grad_norm": 0.6430671811103821, "learning_rate": 9.822175746872984e-05, "loss": 0.0798, "step": 13090 }, { "grad_norm": 0.7728685736656189, "learning_rate": 9.821738438753123e-05, "loss": 0.0841, "step": 13100 }, { "grad_norm": 0.5840721130371094, "learning_rate": 9.821300603335196e-05, "loss": 0.0722, "step": 13110 }, { "grad_norm": 0.5489844083786011, "learning_rate": 9.820862240667085e-05, "loss": 0.0731, "step": 13120 }, { "grad_norm": 0.4177665710449219, "learning_rate": 9.820423350796726e-05, "loss": 0.0823, "step": 13130 }, { "grad_norm": 0.527338981628418, "learning_rate": 9.819983933772118e-05, "loss": 0.0891, "step": 13140 }, { "grad_norm": 0.8413153886795044, "learning_rate": 9.819543989641314e-05, "loss": 0.0874, "step": 13150 }, { "grad_norm": 0.41759881377220154, "learning_rate": 9.819103518452423e-05, "loss": 0.0756, "step": 13160 }, { "grad_norm": 0.6006781458854675, "learning_rate": 9.818662520253618e-05, "loss": 0.0762, "step": 13170 }, { "grad_norm": 0.415576696395874, "learning_rate": 9.818220995093126e-05, "loss": 0.0695, "step": 13180 }, { "grad_norm": 1.0904539823532104, "learning_rate": 9.817778943019228e-05, "loss": 0.0886, "step": 13190 }, { "grad_norm": 0.6746097803115845, "learning_rate": 9.81733636408027e-05, "loss": 0.0927, "step": 13200 }, { "grad_norm": 0.5642114877700806, "learning_rate": 9.816893258324649e-05, "loss": 0.0717, "step": 13210 }, { "grad_norm": 0.49027636647224426, "learning_rate": 9.816449625800823e-05, "loss": 0.0728, "step": 13220 }, { "grad_norm": 0.5996671915054321, "learning_rate": 9.816005466557308e-05, "loss": 0.0913, "step": 13230 }, { "grad_norm": 0.5535121560096741, "learning_rate": 9.815560780642674e-05, "loss": 0.0902, "step": 13240 }, { "grad_norm": 0.4093320965766907, "learning_rate": 9.815115568105555e-05, "loss": 0.0673, "step": 13250 }, { "grad_norm": 0.800905704498291, "learning_rate": 9.814669828994638e-05, "loss": 0.0833, "step": 13260 }, { "grad_norm": 0.544589102268219, "learning_rate": 9.814223563358665e-05, "loss": 0.0721, "step": 13270 }, { "grad_norm": 0.47249093651771545, "learning_rate": 9.813776771246443e-05, "loss": 0.0765, "step": 13280 }, { "grad_norm": 0.6613267064094543, "learning_rate": 9.813329452706829e-05, "loss": 0.0767, "step": 13290 }, { "grad_norm": 0.6387119889259338, "learning_rate": 9.812881607788744e-05, "loss": 0.0744, "step": 13300 }, { "grad_norm": 0.5069112181663513, "learning_rate": 9.812433236541163e-05, "loss": 0.0802, "step": 13310 }, { "grad_norm": 0.6942349076271057, "learning_rate": 9.811984339013116e-05, "loss": 0.0847, "step": 13320 }, { "grad_norm": 0.403864324092865, "learning_rate": 9.811534915253698e-05, "loss": 0.0736, "step": 13330 }, { "grad_norm": 0.862331748008728, "learning_rate": 9.811084965312056e-05, "loss": 0.0822, "step": 13340 }, { "grad_norm": 0.5834261178970337, "learning_rate": 9.810634489237396e-05, "loss": 0.0691, "step": 13350 }, { "grad_norm": 0.5535306930541992, "learning_rate": 9.81018348707898e-05, "loss": 0.0924, "step": 13360 }, { "grad_norm": 0.8998295068740845, "learning_rate": 9.809731958886131e-05, "loss": 0.0826, "step": 13370 }, { "grad_norm": 0.6111286878585815, "learning_rate": 9.809279904708224e-05, "loss": 0.096, "step": 13380 }, { "grad_norm": 1.476174235343933, "learning_rate": 9.808827324594699e-05, "loss": 0.0881, "step": 13390 }, { "grad_norm": 0.4637415409088135, "learning_rate": 9.808374218595046e-05, "loss": 0.0932, "step": 13400 }, { "grad_norm": 0.5619790554046631, "learning_rate": 9.80792058675882e-05, "loss": 0.0795, "step": 13410 }, { "grad_norm": 0.6132788062095642, "learning_rate": 9.807466429135627e-05, "loss": 0.0788, "step": 13420 }, { "grad_norm": 0.51682448387146, "learning_rate": 9.807011745775132e-05, "loss": 0.0802, "step": 13430 }, { "grad_norm": 0.457828551530838, "learning_rate": 9.806556536727061e-05, "loss": 0.0798, "step": 13440 }, { "grad_norm": 0.6041321754455566, "learning_rate": 9.806100802041193e-05, "loss": 0.0993, "step": 13450 }, { "grad_norm": 0.6036583185195923, "learning_rate": 9.805644541767368e-05, "loss": 0.0887, "step": 13460 }, { "grad_norm": 0.5058721303939819, "learning_rate": 9.805187755955478e-05, "loss": 0.0862, "step": 13470 }, { "grad_norm": 0.45575982332229614, "learning_rate": 9.804730444655483e-05, "loss": 0.0849, "step": 13480 }, { "grad_norm": 0.6083956956863403, "learning_rate": 9.804272607917388e-05, "loss": 0.0832, "step": 13490 }, { "grad_norm": 0.4926019012928009, "learning_rate": 9.803814245791265e-05, "loss": 0.0759, "step": 13500 }, { "grad_norm": 0.5168884992599487, "learning_rate": 9.803355358327239e-05, "loss": 0.0847, "step": 13510 }, { "grad_norm": 0.5945696830749512, "learning_rate": 9.802895945575492e-05, "loss": 0.0857, "step": 13520 }, { "grad_norm": 0.6143903732299805, "learning_rate": 9.802436007586266e-05, "loss": 0.093, "step": 13530 }, { "grad_norm": 0.6144698262214661, "learning_rate": 9.801975544409858e-05, "loss": 0.0722, "step": 13540 }, { "grad_norm": 0.6364634037017822, "learning_rate": 9.801514556096625e-05, "loss": 0.0839, "step": 13550 }, { "grad_norm": 0.5193337798118591, "learning_rate": 9.801053042696977e-05, "loss": 0.0856, "step": 13560 }, { "grad_norm": 0.6630189418792725, "learning_rate": 9.800591004261388e-05, "loss": 0.0721, "step": 13570 }, { "grad_norm": 0.5353966355323792, "learning_rate": 9.800128440840385e-05, "loss": 0.0801, "step": 13580 }, { "grad_norm": 0.595807671546936, "learning_rate": 9.799665352484552e-05, "loss": 0.0826, "step": 13590 }, { "grad_norm": 0.6436368823051453, "learning_rate": 9.799201739244532e-05, "loss": 0.0971, "step": 13600 }, { "grad_norm": 0.3708060085773468, "learning_rate": 9.798737601171025e-05, "loss": 0.0706, "step": 13610 }, { "grad_norm": 0.7293341755867004, "learning_rate": 9.79827293831479e-05, "loss": 0.0885, "step": 13620 }, { "grad_norm": 0.6000480651855469, "learning_rate": 9.797807750726638e-05, "loss": 0.0734, "step": 13630 }, { "grad_norm": 0.6671106815338135, "learning_rate": 9.797342038457446e-05, "loss": 0.0758, "step": 13640 }, { "grad_norm": 0.5529289841651917, "learning_rate": 9.796875801558141e-05, "loss": 0.0929, "step": 13650 }, { "grad_norm": 0.5508347153663635, "learning_rate": 9.79640904007971e-05, "loss": 0.0876, "step": 13660 }, { "grad_norm": 0.5700951814651489, "learning_rate": 9.795941754073199e-05, "loss": 0.0712, "step": 13670 }, { "grad_norm": 0.4744967520236969, "learning_rate": 9.795473943589705e-05, "loss": 0.0688, "step": 13680 }, { "grad_norm": 0.7103068232536316, "learning_rate": 9.795005608680394e-05, "loss": 0.0785, "step": 13690 }, { "grad_norm": 0.9004734754562378, "learning_rate": 9.794536749396477e-05, "loss": 0.0814, "step": 13700 }, { "grad_norm": 0.5983040928840637, "learning_rate": 9.79406736578923e-05, "loss": 0.0792, "step": 13710 }, { "grad_norm": 0.6115719079971313, "learning_rate": 9.793597457909984e-05, "loss": 0.0753, "step": 13720 }, { "grad_norm": 0.45037564635276794, "learning_rate": 9.793127025810127e-05, "loss": 0.0709, "step": 13730 }, { "grad_norm": 0.5863540172576904, "learning_rate": 9.792656069541104e-05, "loss": 0.0652, "step": 13740 }, { "grad_norm": 0.6274346709251404, "learning_rate": 9.79218458915442e-05, "loss": 0.0769, "step": 13750 }, { "grad_norm": 0.576034426689148, "learning_rate": 9.791712584701634e-05, "loss": 0.0816, "step": 13760 }, { "grad_norm": 0.5853741765022278, "learning_rate": 9.791240056234364e-05, "loss": 0.0763, "step": 13770 }, { "grad_norm": 0.5357245206832886, "learning_rate": 9.790767003804283e-05, "loss": 0.0769, "step": 13780 }, { "grad_norm": 0.4838200509548187, "learning_rate": 9.790293427463126e-05, "loss": 0.0707, "step": 13790 }, { "grad_norm": 0.5394423007965088, "learning_rate": 9.789819327262684e-05, "loss": 0.0638, "step": 13800 }, { "grad_norm": 0.4126184582710266, "learning_rate": 9.7893447032548e-05, "loss": 0.0762, "step": 13810 }, { "grad_norm": 0.7029460668563843, "learning_rate": 9.78886955549138e-05, "loss": 0.0777, "step": 13820 }, { "grad_norm": 0.4163617193698883, "learning_rate": 9.788393884024387e-05, "loss": 0.0751, "step": 13830 }, { "grad_norm": 0.5864786505699158, "learning_rate": 9.787917688905836e-05, "loss": 0.0799, "step": 13840 }, { "grad_norm": 0.5789608955383301, "learning_rate": 9.787440970187807e-05, "loss": 0.0667, "step": 13850 }, { "grad_norm": 0.740035891532898, "learning_rate": 9.786963727922429e-05, "loss": 0.0866, "step": 13860 }, { "grad_norm": 0.4376808702945709, "learning_rate": 9.786485962161897e-05, "loss": 0.0782, "step": 13870 }, { "grad_norm": 0.6392340064048767, "learning_rate": 9.786007672958455e-05, "loss": 0.0755, "step": 13880 }, { "grad_norm": 0.4942294657230377, "learning_rate": 9.78552886036441e-05, "loss": 0.0868, "step": 13890 }, { "grad_norm": 0.611576497554779, "learning_rate": 9.785049524432124e-05, "loss": 0.0634, "step": 13900 }, { "grad_norm": 0.5902743935585022, "learning_rate": 9.784569665214016e-05, "loss": 0.0808, "step": 13910 }, { "grad_norm": 0.515721321105957, "learning_rate": 9.784089282762563e-05, "loss": 0.088, "step": 13920 }, { "grad_norm": 0.4732820689678192, "learning_rate": 9.7836083771303e-05, "loss": 0.0774, "step": 13930 }, { "grad_norm": 0.5180055499076843, "learning_rate": 9.783126948369817e-05, "loss": 0.0892, "step": 13940 }, { "grad_norm": 0.45635685324668884, "learning_rate": 9.78264499653376e-05, "loss": 0.0677, "step": 13950 }, { "grad_norm": 0.5417400002479553, "learning_rate": 9.782162521674838e-05, "loss": 0.0762, "step": 13960 }, { "grad_norm": 0.5261611342430115, "learning_rate": 9.781679523845812e-05, "loss": 0.0915, "step": 13970 }, { "grad_norm": 0.527816116809845, "learning_rate": 9.781196003099502e-05, "loss": 0.0846, "step": 13980 }, { "grad_norm": 0.34829413890838623, "learning_rate": 9.780711959488786e-05, "loss": 0.0678, "step": 13990 }, { "grad_norm": 0.3633996844291687, "learning_rate": 9.780227393066599e-05, "loss": 0.0775, "step": 14000 }, { "grad_norm": 0.5901584029197693, "learning_rate": 9.77974230388593e-05, "loss": 0.0962, "step": 14010 }, { "grad_norm": 0.5166484713554382, "learning_rate": 9.779256691999829e-05, "loss": 0.095, "step": 14020 }, { "grad_norm": 0.6712213158607483, "learning_rate": 9.778770557461403e-05, "loss": 0.0916, "step": 14030 }, { "grad_norm": 0.5248945355415344, "learning_rate": 9.778283900323812e-05, "loss": 0.0765, "step": 14040 }, { "grad_norm": 0.6024680137634277, "learning_rate": 9.777796720640277e-05, "loss": 0.0893, "step": 14050 }, { "grad_norm": 0.7272139191627502, "learning_rate": 9.777309018464078e-05, "loss": 0.0751, "step": 14060 }, { "grad_norm": 0.5392565131187439, "learning_rate": 9.776820793848547e-05, "loss": 0.0718, "step": 14070 }, { "grad_norm": 0.975104808807373, "learning_rate": 9.776332046847075e-05, "loss": 0.0743, "step": 14080 }, { "grad_norm": 0.4540884494781494, "learning_rate": 9.775842777513111e-05, "loss": 0.0725, "step": 14090 }, { "grad_norm": 0.6065924763679504, "learning_rate": 9.775352985900163e-05, "loss": 0.0983, "step": 14100 }, { "grad_norm": 1.095752477645874, "learning_rate": 9.774862672061791e-05, "loss": 0.0931, "step": 14110 }, { "grad_norm": 0.5061771273612976, "learning_rate": 9.774371836051616e-05, "loss": 0.0766, "step": 14120 }, { "grad_norm": 0.7868569493293762, "learning_rate": 9.773880477923315e-05, "loss": 0.0758, "step": 14130 }, { "grad_norm": 0.49442118406295776, "learning_rate": 9.773388597730623e-05, "loss": 0.0754, "step": 14140 }, { "grad_norm": 0.45948731899261475, "learning_rate": 9.77289619552733e-05, "loss": 0.0657, "step": 14150 }, { "grad_norm": 0.7507746815681458, "learning_rate": 9.772403271367285e-05, "loss": 0.0869, "step": 14160 }, { "grad_norm": 0.8488110303878784, "learning_rate": 9.771909825304396e-05, "loss": 0.0749, "step": 14170 }, { "grad_norm": 0.5273248553276062, "learning_rate": 9.771415857392619e-05, "loss": 0.0847, "step": 14180 }, { "grad_norm": 0.6748102903366089, "learning_rate": 9.770921367685978e-05, "loss": 0.0911, "step": 14190 }, { "grad_norm": 0.610677182674408, "learning_rate": 9.770426356238551e-05, "loss": 0.0735, "step": 14200 }, { "grad_norm": 0.4530596137046814, "learning_rate": 9.769930823104469e-05, "loss": 0.0721, "step": 14210 }, { "grad_norm": 0.5094717144966125, "learning_rate": 9.769434768337926e-05, "loss": 0.0776, "step": 14220 }, { "grad_norm": 0.4738421142101288, "learning_rate": 9.768938191993164e-05, "loss": 0.0665, "step": 14230 }, { "grad_norm": 0.5670470595359802, "learning_rate": 9.768441094124494e-05, "loss": 0.0646, "step": 14240 }, { "grad_norm": 0.5111590027809143, "learning_rate": 9.767943474786275e-05, "loss": 0.0781, "step": 14250 }, { "grad_norm": 0.36630091071128845, "learning_rate": 9.767445334032923e-05, "loss": 0.0566, "step": 14260 }, { "grad_norm": 0.8730248808860779, "learning_rate": 9.766946671918919e-05, "loss": 0.0828, "step": 14270 }, { "grad_norm": 0.5784919857978821, "learning_rate": 9.766447488498796e-05, "loss": 0.0667, "step": 14280 }, { "grad_norm": 0.3750152885913849, "learning_rate": 9.765947783827139e-05, "loss": 0.0685, "step": 14290 }, { "grad_norm": 0.49299782514572144, "learning_rate": 9.765447557958599e-05, "loss": 0.0754, "step": 14300 }, { "grad_norm": 0.5287960767745972, "learning_rate": 9.764946810947879e-05, "loss": 0.0779, "step": 14310 }, { "grad_norm": 0.45752137899398804, "learning_rate": 9.764445542849738e-05, "loss": 0.0816, "step": 14320 }, { "grad_norm": 0.4387013018131256, "learning_rate": 9.763943753718998e-05, "loss": 0.064, "step": 14330 }, { "grad_norm": 0.40840867161750793, "learning_rate": 9.76344144361053e-05, "loss": 0.0671, "step": 14340 }, { "grad_norm": 0.6225373148918152, "learning_rate": 9.762938612579269e-05, "loss": 0.0778, "step": 14350 }, { "grad_norm": 0.466119647026062, "learning_rate": 9.762435260680202e-05, "loss": 0.0721, "step": 14360 }, { "grad_norm": 0.4971848130226135, "learning_rate": 9.761931387968373e-05, "loss": 0.0767, "step": 14370 }, { "grad_norm": 0.5276655554771423, "learning_rate": 9.76142699449889e-05, "loss": 0.0729, "step": 14380 }, { "grad_norm": 0.5783877968788147, "learning_rate": 9.760922080326908e-05, "loss": 0.086, "step": 14390 }, { "grad_norm": 0.6006457805633545, "learning_rate": 9.760416645507644e-05, "loss": 0.0776, "step": 14400 }, { "grad_norm": 0.6225519776344299, "learning_rate": 9.759910690096375e-05, "loss": 0.0708, "step": 14410 }, { "grad_norm": 0.4319007396697998, "learning_rate": 9.759404214148429e-05, "loss": 0.0823, "step": 14420 }, { "grad_norm": 0.7686729431152344, "learning_rate": 9.758897217719191e-05, "loss": 0.0782, "step": 14430 }, { "grad_norm": 0.6474239826202393, "learning_rate": 9.758389700864113e-05, "loss": 0.069, "step": 14440 }, { "grad_norm": 0.7214471697807312, "learning_rate": 9.757881663638688e-05, "loss": 0.0842, "step": 14450 }, { "grad_norm": 0.4995248317718506, "learning_rate": 9.757373106098478e-05, "loss": 0.0746, "step": 14460 }, { "grad_norm": 0.5762956142425537, "learning_rate": 9.756864028299097e-05, "loss": 0.0802, "step": 14470 }, { "grad_norm": 0.5768554210662842, "learning_rate": 9.75635443029622e-05, "loss": 0.0707, "step": 14480 }, { "grad_norm": 0.5730699300765991, "learning_rate": 9.755844312145572e-05, "loss": 0.063, "step": 14490 }, { "grad_norm": 0.7095434069633484, "learning_rate": 9.755333673902941e-05, "loss": 0.0915, "step": 14500 }, { "grad_norm": 0.5004151463508606, "learning_rate": 9.75482251562417e-05, "loss": 0.0694, "step": 14510 }, { "grad_norm": 0.5283318758010864, "learning_rate": 9.754310837365155e-05, "loss": 0.0682, "step": 14520 }, { "grad_norm": 0.5496636033058167, "learning_rate": 9.753798639181856e-05, "loss": 0.0927, "step": 14530 }, { "grad_norm": 0.6106097102165222, "learning_rate": 9.753285921130286e-05, "loss": 0.069, "step": 14540 }, { "grad_norm": 0.4967171549797058, "learning_rate": 9.752772683266512e-05, "loss": 0.0652, "step": 14550 }, { "grad_norm": 0.48979753255844116, "learning_rate": 9.752258925646665e-05, "loss": 0.0725, "step": 14560 }, { "grad_norm": 0.4804650843143463, "learning_rate": 9.751744648326926e-05, "loss": 0.0771, "step": 14570 }, { "grad_norm": 0.8227003812789917, "learning_rate": 9.751229851363536e-05, "loss": 0.0944, "step": 14580 }, { "grad_norm": 0.6295769810676575, "learning_rate": 9.750714534812793e-05, "loss": 0.1015, "step": 14590 }, { "grad_norm": 0.4578128457069397, "learning_rate": 9.750198698731053e-05, "loss": 0.0828, "step": 14600 }, { "grad_norm": 0.4628613591194153, "learning_rate": 9.749682343174722e-05, "loss": 0.0729, "step": 14610 }, { "grad_norm": 0.3837808072566986, "learning_rate": 9.749165468200272e-05, "loss": 0.0623, "step": 14620 }, { "grad_norm": 0.5277756452560425, "learning_rate": 9.748648073864229e-05, "loss": 0.0781, "step": 14630 }, { "grad_norm": 0.7530875205993652, "learning_rate": 9.748130160223168e-05, "loss": 0.0735, "step": 14640 }, { "grad_norm": 0.6868932247161865, "learning_rate": 9.747611727333734e-05, "loss": 0.077, "step": 14650 }, { "grad_norm": 0.6172138452529907, "learning_rate": 9.74709277525262e-05, "loss": 0.0755, "step": 14660 }, { "grad_norm": 0.6006320118904114, "learning_rate": 9.746573304036576e-05, "loss": 0.0777, "step": 14670 }, { "grad_norm": 0.6999114751815796, "learning_rate": 9.746053313742412e-05, "loss": 0.0695, "step": 14680 }, { "grad_norm": 0.7388207912445068, "learning_rate": 9.745532804426994e-05, "loss": 0.0747, "step": 14690 }, { "grad_norm": 0.4450586438179016, "learning_rate": 9.745011776147242e-05, "loss": 0.0739, "step": 14700 }, { "grad_norm": 0.7616071105003357, "learning_rate": 9.744490228960138e-05, "loss": 0.0867, "step": 14710 }, { "grad_norm": 0.5290884971618652, "learning_rate": 9.743968162922713e-05, "loss": 0.0817, "step": 14720 }, { "grad_norm": 0.5306396484375, "learning_rate": 9.743445578092064e-05, "loss": 0.0893, "step": 14730 }, { "grad_norm": 0.5203820466995239, "learning_rate": 9.742922474525338e-05, "loss": 0.0929, "step": 14740 }, { "grad_norm": 1.0422475337982178, "learning_rate": 9.742398852279741e-05, "loss": 0.0749, "step": 14750 }, { "grad_norm": 0.6143375039100647, "learning_rate": 9.741874711412535e-05, "loss": 0.092, "step": 14760 }, { "grad_norm": 0.7049707174301147, "learning_rate": 9.741350051981042e-05, "loss": 0.0764, "step": 14770 }, { "grad_norm": 0.5300312638282776, "learning_rate": 9.740824874042633e-05, "loss": 0.0768, "step": 14780 }, { "grad_norm": 0.5451170802116394, "learning_rate": 9.740299177654746e-05, "loss": 0.0661, "step": 14790 }, { "grad_norm": 0.5698673129081726, "learning_rate": 9.739772962874867e-05, "loss": 0.0808, "step": 14800 }, { "grad_norm": 0.509153425693512, "learning_rate": 9.739246229760541e-05, "loss": 0.0699, "step": 14810 }, { "grad_norm": 0.5774678587913513, "learning_rate": 9.738718978369376e-05, "loss": 0.0822, "step": 14820 }, { "grad_norm": 0.4140837490558624, "learning_rate": 9.738191208759025e-05, "loss": 0.0808, "step": 14830 }, { "grad_norm": 0.721657395362854, "learning_rate": 9.73766292098721e-05, "loss": 0.0975, "step": 14840 }, { "grad_norm": 0.5350748300552368, "learning_rate": 9.737134115111699e-05, "loss": 0.0925, "step": 14850 }, { "grad_norm": 0.6943306922912598, "learning_rate": 9.736604791190323e-05, "loss": 0.0735, "step": 14860 }, { "grad_norm": 0.3201597332954407, "learning_rate": 9.73607494928097e-05, "loss": 0.0592, "step": 14870 }, { "grad_norm": 0.42889681458473206, "learning_rate": 9.735544589441581e-05, "loss": 0.0738, "step": 14880 }, { "grad_norm": 0.741159200668335, "learning_rate": 9.735013711730154e-05, "loss": 0.0747, "step": 14890 }, { "grad_norm": 0.7066861391067505, "learning_rate": 9.734482316204747e-05, "loss": 0.0712, "step": 14900 }, { "grad_norm": 0.8576223254203796, "learning_rate": 9.733950402923473e-05, "loss": 0.0721, "step": 14910 }, { "grad_norm": 0.44212606549263, "learning_rate": 9.7334179719445e-05, "loss": 0.0648, "step": 14920 }, { "grad_norm": 0.7736199498176575, "learning_rate": 9.732885023326053e-05, "loss": 0.0818, "step": 14930 }, { "grad_norm": 0.40850332379341125, "learning_rate": 9.732351557126418e-05, "loss": 0.0894, "step": 14940 }, { "grad_norm": 0.5902693271636963, "learning_rate": 9.731817573403929e-05, "loss": 0.0726, "step": 14950 }, { "grad_norm": 0.35765811800956726, "learning_rate": 9.731283072216985e-05, "loss": 0.0733, "step": 14960 }, { "grad_norm": 0.5164172649383545, "learning_rate": 9.730748053624039e-05, "loss": 0.0641, "step": 14970 }, { "grad_norm": 0.4903305470943451, "learning_rate": 9.730212517683598e-05, "loss": 0.0806, "step": 14980 }, { "grad_norm": 0.7028083801269531, "learning_rate": 9.729676464454228e-05, "loss": 0.0705, "step": 14990 }, { "grad_norm": 0.40136224031448364, "learning_rate": 9.72913989399455e-05, "loss": 0.0799, "step": 15000 }, { "grad_norm": 0.5158106684684753, "learning_rate": 9.728602806363242e-05, "loss": 0.0645, "step": 15010 }, { "grad_norm": 0.5988296270370483, "learning_rate": 9.728065201619043e-05, "loss": 0.079, "step": 15020 }, { "grad_norm": 0.6708066463470459, "learning_rate": 9.727527079820742e-05, "loss": 0.0735, "step": 15030 }, { "grad_norm": 0.4001227915287018, "learning_rate": 9.726988441027186e-05, "loss": 0.0847, "step": 15040 }, { "grad_norm": 0.5876034498214722, "learning_rate": 9.726449285297281e-05, "loss": 0.085, "step": 15050 }, { "grad_norm": 0.6935211420059204, "learning_rate": 9.72590961268999e-05, "loss": 0.0897, "step": 15060 }, { "grad_norm": 0.3970889747142792, "learning_rate": 9.725369423264328e-05, "loss": 0.0701, "step": 15070 }, { "grad_norm": 0.5683480501174927, "learning_rate": 9.72482871707937e-05, "loss": 0.0799, "step": 15080 }, { "grad_norm": 0.6293565630912781, "learning_rate": 9.724287494194247e-05, "loss": 0.0825, "step": 15090 }, { "grad_norm": 0.352095365524292, "learning_rate": 9.723745754668147e-05, "loss": 0.081, "step": 15100 }, { "grad_norm": 0.5649920105934143, "learning_rate": 9.723203498560313e-05, "loss": 0.0828, "step": 15110 }, { "grad_norm": 0.40021729469299316, "learning_rate": 9.722660725930046e-05, "loss": 0.0616, "step": 15120 }, { "grad_norm": 0.5392590761184692, "learning_rate": 9.722117436836702e-05, "loss": 0.0709, "step": 15130 }, { "grad_norm": 0.39395371079444885, "learning_rate": 9.721573631339696e-05, "loss": 0.071, "step": 15140 }, { "grad_norm": 0.9223935604095459, "learning_rate": 9.721029309498494e-05, "loss": 0.0929, "step": 15150 }, { "grad_norm": 0.5908088684082031, "learning_rate": 9.720484471372627e-05, "loss": 0.0862, "step": 15160 }, { "grad_norm": 0.5712197422981262, "learning_rate": 9.719939117021673e-05, "loss": 0.0801, "step": 15170 }, { "grad_norm": 0.6159675717353821, "learning_rate": 9.719393246505275e-05, "loss": 0.0907, "step": 15180 }, { "grad_norm": 0.5225736498832703, "learning_rate": 9.718846859883128e-05, "loss": 0.0762, "step": 15190 }, { "grad_norm": 0.5338141322135925, "learning_rate": 9.718299957214982e-05, "loss": 0.0809, "step": 15200 }, { "grad_norm": 0.5345048904418945, "learning_rate": 9.717752538560646e-05, "loss": 0.0734, "step": 15210 }, { "grad_norm": 0.6538162231445312, "learning_rate": 9.717204603979986e-05, "loss": 0.0698, "step": 15220 }, { "grad_norm": 0.4592253267765045, "learning_rate": 9.716656153532922e-05, "loss": 0.0696, "step": 15230 }, { "grad_norm": 0.6688017249107361, "learning_rate": 9.716107187279434e-05, "loss": 0.0726, "step": 15240 }, { "grad_norm": 0.684309184551239, "learning_rate": 9.715557705279555e-05, "loss": 0.1, "step": 15250 }, { "grad_norm": 0.6037682294845581, "learning_rate": 9.715007707593372e-05, "loss": 0.0972, "step": 15260 }, { "grad_norm": 0.4790114164352417, "learning_rate": 9.714457194281036e-05, "loss": 0.0763, "step": 15270 }, { "grad_norm": 0.6553101539611816, "learning_rate": 9.713906165402751e-05, "loss": 0.0923, "step": 15280 }, { "grad_norm": 0.6375662088394165, "learning_rate": 9.713354621018774e-05, "loss": 0.0983, "step": 15290 }, { "grad_norm": 0.5524880886077881, "learning_rate": 9.712802561189422e-05, "loss": 0.0804, "step": 15300 }, { "grad_norm": 0.4337279200553894, "learning_rate": 9.712249985975069e-05, "loss": 0.0857, "step": 15310 }, { "grad_norm": 0.48609426617622375, "learning_rate": 9.71169689543614e-05, "loss": 0.0667, "step": 15320 }, { "grad_norm": 0.5758464932441711, "learning_rate": 9.711143289633123e-05, "loss": 0.0651, "step": 15330 }, { "grad_norm": 0.5404796600341797, "learning_rate": 9.710589168626561e-05, "loss": 0.075, "step": 15340 }, { "grad_norm": 0.45598694682121277, "learning_rate": 9.710034532477048e-05, "loss": 0.0774, "step": 15350 }, { "grad_norm": 0.5294922590255737, "learning_rate": 9.709479381245239e-05, "loss": 0.0673, "step": 15360 }, { "grad_norm": 0.43271514773368835, "learning_rate": 9.708923714991847e-05, "loss": 0.0779, "step": 15370 }, { "grad_norm": 0.555586040019989, "learning_rate": 9.708367533777638e-05, "loss": 0.079, "step": 15380 }, { "grad_norm": 0.551166832447052, "learning_rate": 9.707810837663431e-05, "loss": 0.0663, "step": 15390 }, { "grad_norm": 0.7704518437385559, "learning_rate": 9.707253626710113e-05, "loss": 0.0811, "step": 15400 }, { "grad_norm": 0.5643988847732544, "learning_rate": 9.706695900978613e-05, "loss": 0.0607, "step": 15410 }, { "grad_norm": 0.4876280128955841, "learning_rate": 9.706137660529926e-05, "loss": 0.0716, "step": 15420 }, { "grad_norm": 0.5924584865570068, "learning_rate": 9.705578905425101e-05, "loss": 0.0995, "step": 15430 }, { "grad_norm": 0.5952944159507751, "learning_rate": 9.705019635725241e-05, "loss": 0.0709, "step": 15440 }, { "grad_norm": 0.546444833278656, "learning_rate": 9.704459851491508e-05, "loss": 0.0832, "step": 15450 }, { "grad_norm": 0.4716174602508545, "learning_rate": 9.703899552785118e-05, "loss": 0.0672, "step": 15460 }, { "grad_norm": 0.6018290519714355, "learning_rate": 9.703338739667346e-05, "loss": 0.0691, "step": 15470 }, { "grad_norm": 0.5226109623908997, "learning_rate": 9.70277741219952e-05, "loss": 0.0704, "step": 15480 }, { "grad_norm": 0.7526702880859375, "learning_rate": 9.702215570443027e-05, "loss": 0.0866, "step": 15490 }, { "grad_norm": 0.5232195258140564, "learning_rate": 9.701653214459309e-05, "loss": 0.0843, "step": 15500 }, { "grad_norm": 0.3794201910495758, "learning_rate": 9.701090344309865e-05, "loss": 0.0746, "step": 15510 }, { "grad_norm": 0.8728850483894348, "learning_rate": 9.700526960056247e-05, "loss": 0.0739, "step": 15520 }, { "grad_norm": 0.5314946174621582, "learning_rate": 9.699963061760068e-05, "loss": 0.0628, "step": 15530 }, { "grad_norm": 0.6233184337615967, "learning_rate": 9.699398649482997e-05, "loss": 0.0748, "step": 15540 }, { "grad_norm": 0.4945515990257263, "learning_rate": 9.698833723286753e-05, "loss": 0.0634, "step": 15550 }, { "grad_norm": 0.4784224033355713, "learning_rate": 9.698268283233118e-05, "loss": 0.0782, "step": 15560 }, { "grad_norm": 0.7493632435798645, "learning_rate": 9.697702329383929e-05, "loss": 0.0847, "step": 15570 }, { "grad_norm": 0.538897693157196, "learning_rate": 9.697135861801074e-05, "loss": 0.0804, "step": 15580 }, { "grad_norm": 0.4827257990837097, "learning_rate": 9.696568880546505e-05, "loss": 0.0708, "step": 15590 }, { "grad_norm": 0.4874379336833954, "learning_rate": 9.696001385682223e-05, "loss": 0.0678, "step": 15600 }, { "grad_norm": 0.5247143507003784, "learning_rate": 9.695433377270291e-05, "loss": 0.0759, "step": 15610 }, { "grad_norm": 0.4423062205314636, "learning_rate": 9.694864855372824e-05, "loss": 0.0624, "step": 15620 }, { "grad_norm": 0.6352530121803284, "learning_rate": 9.694295820051995e-05, "loss": 0.0705, "step": 15630 }, { "grad_norm": 0.725475549697876, "learning_rate": 9.693726271370032e-05, "loss": 0.0704, "step": 15640 }, { "grad_norm": 0.6886624693870544, "learning_rate": 9.693156209389221e-05, "loss": 0.0728, "step": 15650 }, { "grad_norm": 0.5565219521522522, "learning_rate": 9.692585634171905e-05, "loss": 0.0691, "step": 15660 }, { "grad_norm": 0.5635090470314026, "learning_rate": 9.692014545780476e-05, "loss": 0.0781, "step": 15670 }, { "grad_norm": 0.4193912744522095, "learning_rate": 9.691442944277393e-05, "loss": 0.067, "step": 15680 }, { "grad_norm": 0.4625684320926666, "learning_rate": 9.690870829725162e-05, "loss": 0.068, "step": 15690 }, { "grad_norm": 0.5548231601715088, "learning_rate": 9.69029820218635e-05, "loss": 0.0712, "step": 15700 }, { "grad_norm": 0.4183714985847473, "learning_rate": 9.689725061723579e-05, "loss": 0.0847, "step": 15710 }, { "grad_norm": 0.5129842758178711, "learning_rate": 9.689151408399527e-05, "loss": 0.0718, "step": 15720 }, { "grad_norm": 0.49788999557495117, "learning_rate": 9.688577242276924e-05, "loss": 0.0771, "step": 15730 }, { "grad_norm": 0.6057044267654419, "learning_rate": 9.688002563418566e-05, "loss": 0.0817, "step": 15740 }, { "grad_norm": 0.5321861505508423, "learning_rate": 9.687427371887293e-05, "loss": 0.067, "step": 15750 }, { "grad_norm": 0.38816091418266296, "learning_rate": 9.686851667746012e-05, "loss": 0.0699, "step": 15760 }, { "grad_norm": 0.4683145582675934, "learning_rate": 9.686275451057677e-05, "loss": 0.0646, "step": 15770 }, { "grad_norm": 0.5140568017959595, "learning_rate": 9.685698721885308e-05, "loss": 0.068, "step": 15780 }, { "grad_norm": 0.5388494729995728, "learning_rate": 9.68512148029197e-05, "loss": 0.0796, "step": 15790 }, { "grad_norm": 0.5649490356445312, "learning_rate": 9.684543726340791e-05, "loss": 0.0726, "step": 15800 }, { "grad_norm": 0.555555522441864, "learning_rate": 9.683965460094952e-05, "loss": 0.0627, "step": 15810 }, { "grad_norm": 0.578871488571167, "learning_rate": 9.683386681617694e-05, "loss": 0.07, "step": 15820 }, { "grad_norm": 0.5831742882728577, "learning_rate": 9.68280739097231e-05, "loss": 0.0745, "step": 15830 }, { "grad_norm": 0.4547956585884094, "learning_rate": 9.682227588222148e-05, "loss": 0.0627, "step": 15840 }, { "grad_norm": 0.9226658940315247, "learning_rate": 9.681647273430618e-05, "loss": 0.0728, "step": 15850 }, { "grad_norm": 0.9282383918762207, "learning_rate": 9.681066446661182e-05, "loss": 0.0831, "step": 15860 }, { "grad_norm": 0.6451922655105591, "learning_rate": 9.680485107977357e-05, "loss": 0.0835, "step": 15870 }, { "grad_norm": 0.8235684037208557, "learning_rate": 9.679903257442716e-05, "loss": 0.0904, "step": 15880 }, { "grad_norm": 0.7119265794754028, "learning_rate": 9.679320895120891e-05, "loss": 0.0741, "step": 15890 }, { "grad_norm": 0.5036584138870239, "learning_rate": 9.67873802107557e-05, "loss": 0.0679, "step": 15900 }, { "grad_norm": 0.4797326326370239, "learning_rate": 9.67815463537049e-05, "loss": 0.0667, "step": 15910 }, { "grad_norm": 0.7738030552864075, "learning_rate": 9.677570738069457e-05, "loss": 0.0681, "step": 15920 }, { "grad_norm": 0.4315359592437744, "learning_rate": 9.676986329236318e-05, "loss": 0.0803, "step": 15930 }, { "grad_norm": 0.5098230838775635, "learning_rate": 9.676401408934987e-05, "loss": 0.0696, "step": 15940 }, { "grad_norm": 0.5633373856544495, "learning_rate": 9.675815977229428e-05, "loss": 0.0773, "step": 15950 }, { "grad_norm": 0.6089991331100464, "learning_rate": 9.675230034183664e-05, "loss": 0.0725, "step": 15960 }, { "grad_norm": 0.4958235025405884, "learning_rate": 9.674643579861773e-05, "loss": 0.0888, "step": 15970 }, { "grad_norm": 0.5476804971694946, "learning_rate": 9.674056614327886e-05, "loss": 0.0623, "step": 15980 }, { "grad_norm": 0.5192184448242188, "learning_rate": 9.673469137646198e-05, "loss": 0.0814, "step": 15990 }, { "grad_norm": 0.4465128779411316, "learning_rate": 9.67288114988095e-05, "loss": 0.0598, "step": 16000 }, { "grad_norm": 0.5075980424880981, "learning_rate": 9.672292651096447e-05, "loss": 0.0651, "step": 16010 }, { "grad_norm": 0.4822046756744385, "learning_rate": 9.671703641357042e-05, "loss": 0.0628, "step": 16020 }, { "grad_norm": 0.5315526127815247, "learning_rate": 9.67111412072715e-05, "loss": 0.0851, "step": 16030 }, { "grad_norm": 0.5168359279632568, "learning_rate": 9.670524089271242e-05, "loss": 0.0702, "step": 16040 }, { "grad_norm": 0.6630574464797974, "learning_rate": 9.669933547053842e-05, "loss": 0.0791, "step": 16050 }, { "grad_norm": 0.4953194260597229, "learning_rate": 9.669342494139531e-05, "loss": 0.0757, "step": 16060 }, { "grad_norm": 0.6424949169158936, "learning_rate": 9.668750930592943e-05, "loss": 0.0719, "step": 16070 }, { "grad_norm": 0.4823874533176422, "learning_rate": 9.668158856478775e-05, "loss": 0.0674, "step": 16080 }, { "grad_norm": 0.6857396960258484, "learning_rate": 9.66756627186177e-05, "loss": 0.0801, "step": 16090 }, { "grad_norm": 0.4514603316783905, "learning_rate": 9.666973176806737e-05, "loss": 0.0825, "step": 16100 }, { "grad_norm": 0.549562931060791, "learning_rate": 9.666379571378534e-05, "loss": 0.0722, "step": 16110 }, { "grad_norm": 0.6954529285430908, "learning_rate": 9.665785455642076e-05, "loss": 0.0925, "step": 16120 }, { "grad_norm": 0.5872132778167725, "learning_rate": 9.665190829662337e-05, "loss": 0.0734, "step": 16130 }, { "grad_norm": 0.4399334788322449, "learning_rate": 9.664595693504342e-05, "loss": 0.0695, "step": 16140 }, { "grad_norm": 0.42363953590393066, "learning_rate": 9.664000047233175e-05, "loss": 0.0931, "step": 16150 }, { "grad_norm": 0.40098559856414795, "learning_rate": 9.663403890913976e-05, "loss": 0.0795, "step": 16160 }, { "grad_norm": 0.5142579674720764, "learning_rate": 9.662807224611938e-05, "loss": 0.0746, "step": 16170 }, { "grad_norm": 0.38361093401908875, "learning_rate": 9.662210048392311e-05, "loss": 0.0675, "step": 16180 }, { "grad_norm": 0.5409215092658997, "learning_rate": 9.661612362320405e-05, "loss": 0.0759, "step": 16190 }, { "grad_norm": 0.5773565173149109, "learning_rate": 9.661014166461579e-05, "loss": 0.077, "step": 16200 }, { "grad_norm": 0.48001131415367126, "learning_rate": 9.66041546088125e-05, "loss": 0.0775, "step": 16210 }, { "grad_norm": 0.46057364344596863, "learning_rate": 9.659816245644895e-05, "loss": 0.0668, "step": 16220 }, { "grad_norm": 0.48876628279685974, "learning_rate": 9.65921652081804e-05, "loss": 0.0722, "step": 16230 }, { "grad_norm": 0.5931878685951233, "learning_rate": 9.658616286466271e-05, "loss": 0.0779, "step": 16240 }, { "grad_norm": 0.4165976047515869, "learning_rate": 9.65801554265523e-05, "loss": 0.0823, "step": 16250 }, { "grad_norm": 0.6620944142341614, "learning_rate": 9.657414289450612e-05, "loss": 0.0716, "step": 16260 }, { "grad_norm": 0.6712444424629211, "learning_rate": 9.656812526918171e-05, "loss": 0.0835, "step": 16270 }, { "grad_norm": 0.48818492889404297, "learning_rate": 9.656210255123712e-05, "loss": 0.0708, "step": 16280 }, { "grad_norm": 0.5405653715133667, "learning_rate": 9.6556074741331e-05, "loss": 0.0641, "step": 16290 }, { "grad_norm": 0.3543417155742645, "learning_rate": 9.655004184012256e-05, "loss": 0.062, "step": 16300 }, { "grad_norm": 0.6300846934318542, "learning_rate": 9.654400384827152e-05, "loss": 0.0816, "step": 16310 }, { "grad_norm": 0.5774674415588379, "learning_rate": 9.653796076643818e-05, "loss": 0.0612, "step": 16320 }, { "grad_norm": 0.7600505352020264, "learning_rate": 9.653191259528344e-05, "loss": 0.0749, "step": 16330 }, { "grad_norm": 0.5556977391242981, "learning_rate": 9.65258593354687e-05, "loss": 0.0692, "step": 16340 }, { "grad_norm": 0.3689909279346466, "learning_rate": 9.651980098765591e-05, "loss": 0.0606, "step": 16350 }, { "grad_norm": 0.6298580765724182, "learning_rate": 9.651373755250765e-05, "loss": 0.0764, "step": 16360 }, { "grad_norm": 0.540372908115387, "learning_rate": 9.650766903068697e-05, "loss": 0.073, "step": 16370 }, { "grad_norm": 0.6374687552452087, "learning_rate": 9.650159542285753e-05, "loss": 0.0694, "step": 16380 }, { "grad_norm": 0.6393612027168274, "learning_rate": 9.649551672968353e-05, "loss": 0.0834, "step": 16390 }, { "grad_norm": 0.41214925050735474, "learning_rate": 9.648943295182973e-05, "loss": 0.0675, "step": 16400 }, { "grad_norm": 0.4872303307056427, "learning_rate": 9.648334408996144e-05, "loss": 0.0688, "step": 16410 }, { "grad_norm": 0.46042099595069885, "learning_rate": 9.647725014474452e-05, "loss": 0.0694, "step": 16420 }, { "grad_norm": 0.4603288173675537, "learning_rate": 9.64711511168454e-05, "loss": 0.078, "step": 16430 }, { "grad_norm": 0.30205219984054565, "learning_rate": 9.646504700693108e-05, "loss": 0.0755, "step": 16440 }, { "grad_norm": 0.5361512303352356, "learning_rate": 9.645893781566907e-05, "loss": 0.0735, "step": 16450 }, { "grad_norm": 0.6253567337989807, "learning_rate": 9.645282354372744e-05, "loss": 0.0893, "step": 16460 }, { "grad_norm": 1.114519715309143, "learning_rate": 9.644670419177491e-05, "loss": 0.0848, "step": 16470 }, { "grad_norm": 0.5571073293685913, "learning_rate": 9.644057976048062e-05, "loss": 0.0774, "step": 16480 }, { "grad_norm": 1.0204784870147705, "learning_rate": 9.643445025051435e-05, "loss": 0.0845, "step": 16490 }, { "grad_norm": 0.6198946833610535, "learning_rate": 9.642831566254641e-05, "loss": 0.0697, "step": 16500 }, { "grad_norm": 0.5566926002502441, "learning_rate": 9.642217599724769e-05, "loss": 0.0658, "step": 16510 }, { "grad_norm": 0.5118930339813232, "learning_rate": 9.64160312552896e-05, "loss": 0.0712, "step": 16520 }, { "grad_norm": 0.4759662449359894, "learning_rate": 9.64098814373441e-05, "loss": 0.0676, "step": 16530 }, { "grad_norm": 0.5163881778717041, "learning_rate": 9.640372654408374e-05, "loss": 0.0653, "step": 16540 }, { "grad_norm": 0.43153634667396545, "learning_rate": 9.639756657618162e-05, "loss": 0.0746, "step": 16550 }, { "grad_norm": 0.5223915576934814, "learning_rate": 9.639140153431138e-05, "loss": 0.0746, "step": 16560 }, { "grad_norm": 0.5404914617538452, "learning_rate": 9.638523141914721e-05, "loss": 0.0789, "step": 16570 }, { "grad_norm": 0.5026710033416748, "learning_rate": 9.637905623136388e-05, "loss": 0.075, "step": 16580 }, { "grad_norm": 0.6130282878875732, "learning_rate": 9.637287597163669e-05, "loss": 0.0609, "step": 16590 }, { "grad_norm": 0.4989463984966278, "learning_rate": 9.63666906406415e-05, "loss": 0.0705, "step": 16600 }, { "grad_norm": 0.39746055006980896, "learning_rate": 9.636050023905473e-05, "loss": 0.0569, "step": 16610 }, { "grad_norm": 0.4404814541339874, "learning_rate": 9.635430476755336e-05, "loss": 0.0602, "step": 16620 }, { "grad_norm": 0.6815595030784607, "learning_rate": 9.63481042268149e-05, "loss": 0.0746, "step": 16630 }, { "grad_norm": 0.7576341032981873, "learning_rate": 9.634189861751745e-05, "loss": 0.0795, "step": 16640 }, { "grad_norm": 0.42923420667648315, "learning_rate": 9.633568794033967e-05, "loss": 0.0784, "step": 16650 }, { "grad_norm": 0.39967745542526245, "learning_rate": 9.63294721959607e-05, "loss": 0.0625, "step": 16660 }, { "grad_norm": 0.5029999613761902, "learning_rate": 9.63232513850603e-05, "loss": 0.085, "step": 16670 }, { "grad_norm": 0.5373580455780029, "learning_rate": 9.631702550831878e-05, "loss": 0.08, "step": 16680 }, { "grad_norm": 0.5666658282279968, "learning_rate": 9.631079456641698e-05, "loss": 0.0698, "step": 16690 }, { "grad_norm": 0.6103858351707458, "learning_rate": 9.630455856003632e-05, "loss": 0.0787, "step": 16700 }, { "grad_norm": 0.451408326625824, "learning_rate": 9.629831748985876e-05, "loss": 0.0873, "step": 16710 }, { "grad_norm": 0.6143738627433777, "learning_rate": 9.629207135656679e-05, "loss": 0.0842, "step": 16720 }, { "grad_norm": 0.6015369892120361, "learning_rate": 9.628582016084353e-05, "loss": 0.0696, "step": 16730 }, { "grad_norm": 0.6302627921104431, "learning_rate": 9.627956390337254e-05, "loss": 0.0605, "step": 16740 }, { "grad_norm": 0.505425751209259, "learning_rate": 9.627330258483802e-05, "loss": 0.0723, "step": 16750 }, { "grad_norm": 0.48181048035621643, "learning_rate": 9.62670362059247e-05, "loss": 0.0599, "step": 16760 }, { "grad_norm": 0.5087931752204895, "learning_rate": 9.626076476731786e-05, "loss": 0.0777, "step": 16770 }, { "grad_norm": 0.5029780268669128, "learning_rate": 9.625448826970336e-05, "loss": 0.0776, "step": 16780 }, { "grad_norm": 0.9884427785873413, "learning_rate": 9.624820671376755e-05, "loss": 0.0753, "step": 16790 }, { "grad_norm": 0.5664339661598206, "learning_rate": 9.62419201001974e-05, "loss": 0.0752, "step": 16800 }, { "grad_norm": 0.5554674863815308, "learning_rate": 9.623562842968037e-05, "loss": 0.0619, "step": 16810 }, { "grad_norm": 0.5577620267868042, "learning_rate": 9.622933170290454e-05, "loss": 0.0661, "step": 16820 }, { "grad_norm": 0.5595262050628662, "learning_rate": 9.622302992055849e-05, "loss": 0.0647, "step": 16830 }, { "grad_norm": 0.38565775752067566, "learning_rate": 9.62167230833314e-05, "loss": 0.0637, "step": 16840 }, { "grad_norm": 0.4914967715740204, "learning_rate": 9.621041119191295e-05, "loss": 0.0678, "step": 16850 }, { "grad_norm": 0.6655109524726868, "learning_rate": 9.620409424699342e-05, "loss": 0.081, "step": 16860 }, { "grad_norm": 0.4503035247325897, "learning_rate": 9.619777224926359e-05, "loss": 0.0583, "step": 16870 }, { "grad_norm": 0.41106683015823364, "learning_rate": 9.619144519941485e-05, "loss": 0.0697, "step": 16880 }, { "grad_norm": 0.5527094006538391, "learning_rate": 9.618511309813912e-05, "loss": 0.0767, "step": 16890 }, { "grad_norm": 0.8314101099967957, "learning_rate": 9.617877594612886e-05, "loss": 0.0658, "step": 16900 }, { "grad_norm": 0.5722299218177795, "learning_rate": 9.617243374407707e-05, "loss": 0.0715, "step": 16910 }, { "grad_norm": 0.4569547474384308, "learning_rate": 9.616608649267736e-05, "loss": 0.0795, "step": 16920 }, { "grad_norm": 0.6006328463554382, "learning_rate": 9.615973419262385e-05, "loss": 0.0673, "step": 16930 }, { "grad_norm": 0.9165986776351929, "learning_rate": 9.615337684461119e-05, "loss": 0.06, "step": 16940 }, { "grad_norm": 0.5622009038925171, "learning_rate": 9.614701444933465e-05, "loss": 0.0724, "step": 16950 }, { "grad_norm": 0.6022998094558716, "learning_rate": 9.614064700748997e-05, "loss": 0.0747, "step": 16960 }, { "grad_norm": 0.7125872373580933, "learning_rate": 9.613427451977352e-05, "loss": 0.0682, "step": 16970 }, { "grad_norm": 0.5220851898193359, "learning_rate": 9.612789698688216e-05, "loss": 0.082, "step": 16980 }, { "grad_norm": 0.4220495820045471, "learning_rate": 9.612151440951334e-05, "loss": 0.075, "step": 16990 }, { "grad_norm": 0.8837787508964539, "learning_rate": 9.611512678836506e-05, "loss": 0.0751, "step": 17000 }, { "grad_norm": 0.48239368200302124, "learning_rate": 9.610873412413584e-05, "loss": 0.074, "step": 17010 }, { "grad_norm": 1.166532278060913, "learning_rate": 9.610233641752476e-05, "loss": 0.0867, "step": 17020 }, { "grad_norm": 0.8927943110466003, "learning_rate": 9.609593366923151e-05, "loss": 0.076, "step": 17030 }, { "grad_norm": 0.5689838528633118, "learning_rate": 9.608952587995625e-05, "loss": 0.0641, "step": 17040 }, { "grad_norm": 0.4745612144470215, "learning_rate": 9.608311305039972e-05, "loss": 0.0751, "step": 17050 }, { "grad_norm": 0.7827146053314209, "learning_rate": 9.607669518126326e-05, "loss": 0.0856, "step": 17060 }, { "grad_norm": 0.5778455138206482, "learning_rate": 9.607027227324866e-05, "loss": 0.0677, "step": 17070 }, { "grad_norm": 0.48608025908470154, "learning_rate": 9.606384432705837e-05, "loss": 0.0744, "step": 17080 }, { "grad_norm": 0.8319346904754639, "learning_rate": 9.60574113433953e-05, "loss": 0.0829, "step": 17090 }, { "grad_norm": 0.4135114252567291, "learning_rate": 9.6050973322963e-05, "loss": 0.0785, "step": 17100 }, { "grad_norm": 0.6829720735549927, "learning_rate": 9.604453026646547e-05, "loss": 0.0723, "step": 17110 }, { "grad_norm": 0.471762090921402, "learning_rate": 9.603808217460735e-05, "loss": 0.069, "step": 17120 }, { "grad_norm": 0.4918140470981598, "learning_rate": 9.603162904809377e-05, "loss": 0.0622, "step": 17130 }, { "grad_norm": 0.5629061460494995, "learning_rate": 9.602517088763045e-05, "loss": 0.0582, "step": 17140 }, { "grad_norm": 0.5439116954803467, "learning_rate": 9.601870769392365e-05, "loss": 0.086, "step": 17150 }, { "grad_norm": 0.4746783375740051, "learning_rate": 9.601223946768017e-05, "loss": 0.073, "step": 17160 }, { "grad_norm": 0.4330379366874695, "learning_rate": 9.600576620960734e-05, "loss": 0.0647, "step": 17170 }, { "grad_norm": 0.47664856910705566, "learning_rate": 9.599928792041308e-05, "loss": 0.064, "step": 17180 }, { "grad_norm": 0.7555559277534485, "learning_rate": 9.599280460080587e-05, "loss": 0.0717, "step": 17190 }, { "grad_norm": 0.5846897959709167, "learning_rate": 9.59863162514947e-05, "loss": 0.0776, "step": 17200 }, { "grad_norm": 0.7446384429931641, "learning_rate": 9.597982287318911e-05, "loss": 0.0743, "step": 17210 }, { "grad_norm": 0.5213159322738647, "learning_rate": 9.597332446659923e-05, "loss": 0.0699, "step": 17220 }, { "grad_norm": 0.46430495381355286, "learning_rate": 9.59668210324357e-05, "loss": 0.0688, "step": 17230 }, { "grad_norm": 0.48328346014022827, "learning_rate": 9.596031257140974e-05, "loss": 0.0629, "step": 17240 }, { "grad_norm": 1.0300382375717163, "learning_rate": 9.59537990842331e-05, "loss": 0.085, "step": 17250 }, { "grad_norm": 0.4155627191066742, "learning_rate": 9.594728057161806e-05, "loss": 0.061, "step": 17260 }, { "grad_norm": 0.6038023829460144, "learning_rate": 9.594075703427752e-05, "loss": 0.0693, "step": 17270 }, { "grad_norm": 0.38685134053230286, "learning_rate": 9.593422847292486e-05, "loss": 0.068, "step": 17280 }, { "grad_norm": 0.44087284803390503, "learning_rate": 9.592769488827402e-05, "loss": 0.0758, "step": 17290 }, { "grad_norm": 0.5521239638328552, "learning_rate": 9.592115628103952e-05, "loss": 0.071, "step": 17300 }, { "grad_norm": 0.5565788149833679, "learning_rate": 9.591461265193643e-05, "loss": 0.0646, "step": 17310 }, { "grad_norm": 1.065803050994873, "learning_rate": 9.590806400168032e-05, "loss": 0.053, "step": 17320 }, { "grad_norm": 0.40402716398239136, "learning_rate": 9.590151033098735e-05, "loss": 0.0626, "step": 17330 }, { "grad_norm": 0.41262122988700867, "learning_rate": 9.589495164057423e-05, "loss": 0.0723, "step": 17340 }, { "grad_norm": 0.5142912268638611, "learning_rate": 9.58883879311582e-05, "loss": 0.0673, "step": 17350 }, { "grad_norm": 0.42111465334892273, "learning_rate": 9.588181920345705e-05, "loss": 0.0684, "step": 17360 }, { "grad_norm": 0.5460363626480103, "learning_rate": 9.587524545818913e-05, "loss": 0.0687, "step": 17370 }, { "grad_norm": 0.4247434139251709, "learning_rate": 9.586866669607335e-05, "loss": 0.0616, "step": 17380 }, { "grad_norm": 0.6183528304100037, "learning_rate": 9.586208291782915e-05, "loss": 0.0699, "step": 17390 }, { "grad_norm": 0.506868302822113, "learning_rate": 9.58554941241765e-05, "loss": 0.0694, "step": 17400 }, { "grad_norm": 0.5239683389663696, "learning_rate": 9.584890031583596e-05, "loss": 0.0585, "step": 17410 }, { "grad_norm": 0.40216064453125, "learning_rate": 9.584230149352861e-05, "loss": 0.0666, "step": 17420 }, { "grad_norm": 0.6166157722473145, "learning_rate": 9.58356976579761e-05, "loss": 0.0623, "step": 17430 }, { "grad_norm": 0.6274353861808777, "learning_rate": 9.58290888099006e-05, "loss": 0.0745, "step": 17440 }, { "grad_norm": 0.5492579340934753, "learning_rate": 9.582247495002486e-05, "loss": 0.0711, "step": 17450 }, { "grad_norm": 0.4914036989212036, "learning_rate": 9.581585607907214e-05, "loss": 0.0686, "step": 17460 }, { "grad_norm": 0.5030112266540527, "learning_rate": 9.580923219776628e-05, "loss": 0.0655, "step": 17470 }, { "grad_norm": 0.4328047037124634, "learning_rate": 9.580260330683167e-05, "loss": 0.0591, "step": 17480 }, { "grad_norm": 0.5364432334899902, "learning_rate": 9.579596940699322e-05, "loss": 0.0593, "step": 17490 }, { "grad_norm": 0.7501578330993652, "learning_rate": 9.578933049897643e-05, "loss": 0.0592, "step": 17500 }, { "grad_norm": 0.4699084758758545, "learning_rate": 9.578268658350728e-05, "loss": 0.0676, "step": 17510 }, { "grad_norm": 0.44906866550445557, "learning_rate": 9.577603766131235e-05, "loss": 0.0739, "step": 17520 }, { "grad_norm": 0.8278393149375916, "learning_rate": 9.576938373311878e-05, "loss": 0.0715, "step": 17530 }, { "grad_norm": 0.5784448981285095, "learning_rate": 9.576272479965421e-05, "loss": 0.0736, "step": 17540 }, { "grad_norm": 0.4761884808540344, "learning_rate": 9.575606086164687e-05, "loss": 0.0703, "step": 17550 }, { "grad_norm": 0.48090454936027527, "learning_rate": 9.57493919198255e-05, "loss": 0.0821, "step": 17560 }, { "grad_norm": 0.5989962816238403, "learning_rate": 9.57427179749194e-05, "loss": 0.0825, "step": 17570 }, { "grad_norm": 0.5036561489105225, "learning_rate": 9.573603902765846e-05, "loss": 0.0706, "step": 17580 }, { "grad_norm": 0.4124634563922882, "learning_rate": 9.572935507877304e-05, "loss": 0.072, "step": 17590 }, { "grad_norm": 0.6452357769012451, "learning_rate": 9.57226661289941e-05, "loss": 0.0876, "step": 17600 }, { "grad_norm": 1.049068570137024, "learning_rate": 9.571597217905315e-05, "loss": 0.0825, "step": 17610 }, { "grad_norm": 0.8339008092880249, "learning_rate": 9.57092732296822e-05, "loss": 0.0711, "step": 17620 }, { "grad_norm": 1.043476939201355, "learning_rate": 9.570256928161385e-05, "loss": 0.0605, "step": 17630 }, { "grad_norm": 0.41376981139183044, "learning_rate": 9.569586033558126e-05, "loss": 0.0699, "step": 17640 }, { "grad_norm": 0.4042176902294159, "learning_rate": 9.568914639231807e-05, "loss": 0.0594, "step": 17650 }, { "grad_norm": 0.629799485206604, "learning_rate": 9.568242745255852e-05, "loss": 0.0714, "step": 17660 }, { "grad_norm": 0.4645596742630005, "learning_rate": 9.567570351703739e-05, "loss": 0.073, "step": 17670 }, { "grad_norm": 0.6222571730613708, "learning_rate": 9.566897458649001e-05, "loss": 0.0775, "step": 17680 }, { "grad_norm": 0.6272324919700623, "learning_rate": 9.566224066165221e-05, "loss": 0.0856, "step": 17690 }, { "grad_norm": 0.5091826319694519, "learning_rate": 9.565550174326043e-05, "loss": 0.0816, "step": 17700 }, { "grad_norm": 0.5472010374069214, "learning_rate": 9.564875783205162e-05, "loss": 0.0694, "step": 17710 }, { "grad_norm": 0.6081889867782593, "learning_rate": 9.564200892876328e-05, "loss": 0.0758, "step": 17720 }, { "grad_norm": 0.8190525770187378, "learning_rate": 9.563525503413348e-05, "loss": 0.0665, "step": 17730 }, { "grad_norm": 0.47386249899864197, "learning_rate": 9.562849614890079e-05, "loss": 0.0799, "step": 17740 }, { "grad_norm": 0.46540987491607666, "learning_rate": 9.562173227380436e-05, "loss": 0.0818, "step": 17750 }, { "grad_norm": 0.5269386768341064, "learning_rate": 9.561496340958389e-05, "loss": 0.0689, "step": 17760 }, { "grad_norm": 0.5351634621620178, "learning_rate": 9.560818955697959e-05, "loss": 0.073, "step": 17770 }, { "grad_norm": 0.4957621097564697, "learning_rate": 9.560141071673228e-05, "loss": 0.068, "step": 17780 }, { "grad_norm": 0.43186110258102417, "learning_rate": 9.559462688958323e-05, "loss": 0.046, "step": 17790 }, { "grad_norm": 0.3506406247615814, "learning_rate": 9.558783807627434e-05, "loss": 0.0524, "step": 17800 }, { "grad_norm": 0.48380544781684875, "learning_rate": 9.558104427754801e-05, "loss": 0.0832, "step": 17810 }, { "grad_norm": 0.4996330440044403, "learning_rate": 9.557424549414722e-05, "loss": 0.08, "step": 17820 }, { "grad_norm": 0.422016441822052, "learning_rate": 9.556744172681546e-05, "loss": 0.067, "step": 17830 }, { "grad_norm": 0.9329498410224915, "learning_rate": 9.556063297629677e-05, "loss": 0.0804, "step": 17840 }, { "grad_norm": 0.5428825616836548, "learning_rate": 9.555381924333578e-05, "loss": 0.0575, "step": 17850 }, { "grad_norm": 0.5015822649002075, "learning_rate": 9.554700052867758e-05, "loss": 0.0607, "step": 17860 }, { "grad_norm": 0.646460235118866, "learning_rate": 9.554017683306789e-05, "loss": 0.0702, "step": 17870 }, { "grad_norm": 0.6348096132278442, "learning_rate": 9.553334815725294e-05, "loss": 0.0664, "step": 17880 }, { "grad_norm": 0.4140279293060303, "learning_rate": 9.552651450197949e-05, "loss": 0.0691, "step": 17890 }, { "grad_norm": 0.547721266746521, "learning_rate": 9.551967586799486e-05, "loss": 0.0679, "step": 17900 }, { "grad_norm": 0.5002989172935486, "learning_rate": 9.551283225604692e-05, "loss": 0.0667, "step": 17910 }, { "grad_norm": 0.46830758452415466, "learning_rate": 9.550598366688406e-05, "loss": 0.0725, "step": 17920 }, { "grad_norm": 0.536196768283844, "learning_rate": 9.549913010125526e-05, "loss": 0.0687, "step": 17930 }, { "grad_norm": 0.886650800704956, "learning_rate": 9.549227155990999e-05, "loss": 0.0623, "step": 17940 }, { "grad_norm": 0.4250960648059845, "learning_rate": 9.548540804359828e-05, "loss": 0.0689, "step": 17950 }, { "grad_norm": 0.6739856004714966, "learning_rate": 9.547853955307077e-05, "loss": 0.0749, "step": 17960 }, { "grad_norm": 0.4349808990955353, "learning_rate": 9.547166608907853e-05, "loss": 0.0612, "step": 17970 }, { "grad_norm": 0.5495092272758484, "learning_rate": 9.546478765237326e-05, "loss": 0.0746, "step": 17980 }, { "grad_norm": 0.7227523326873779, "learning_rate": 9.545790424370715e-05, "loss": 0.0738, "step": 17990 }, { "grad_norm": 0.36208271980285645, "learning_rate": 9.5451015863833e-05, "loss": 0.0806, "step": 18000 }, { "grad_norm": 0.46430504322052, "learning_rate": 9.544412251350408e-05, "loss": 0.0757, "step": 18010 }, { "grad_norm": 0.43252643942832947, "learning_rate": 9.543722419347422e-05, "loss": 0.0653, "step": 18020 }, { "grad_norm": 0.4392254650592804, "learning_rate": 9.543032090449788e-05, "loss": 0.0724, "step": 18030 }, { "grad_norm": 0.5682642459869385, "learning_rate": 9.542341264732992e-05, "loss": 0.0798, "step": 18040 }, { "grad_norm": 0.766707181930542, "learning_rate": 9.541649942272585e-05, "loss": 0.0788, "step": 18050 }, { "grad_norm": 0.5059413313865662, "learning_rate": 9.54095812314417e-05, "loss": 0.0606, "step": 18060 }, { "grad_norm": 0.3945586681365967, "learning_rate": 9.540265807423401e-05, "loss": 0.0627, "step": 18070 }, { "grad_norm": 0.46316802501678467, "learning_rate": 9.53957299518599e-05, "loss": 0.068, "step": 18080 }, { "grad_norm": 0.29692232608795166, "learning_rate": 9.5388796865077e-05, "loss": 0.0587, "step": 18090 }, { "grad_norm": 0.35992005467414856, "learning_rate": 9.538185881464353e-05, "loss": 0.0703, "step": 18100 }, { "grad_norm": 0.6246164441108704, "learning_rate": 9.537491580131821e-05, "loss": 0.0636, "step": 18110 }, { "grad_norm": 0.5883521437644958, "learning_rate": 9.53679678258603e-05, "loss": 0.0679, "step": 18120 }, { "grad_norm": 0.580936849117279, "learning_rate": 9.536101488902966e-05, "loss": 0.0696, "step": 18130 }, { "grad_norm": 0.415667861700058, "learning_rate": 9.535405699158663e-05, "loss": 0.0596, "step": 18140 }, { "grad_norm": 0.5602635741233826, "learning_rate": 9.53470941342921e-05, "loss": 0.0539, "step": 18150 }, { "grad_norm": 0.4210849106311798, "learning_rate": 9.534012631790756e-05, "loss": 0.0614, "step": 18160 }, { "grad_norm": 0.6027644276618958, "learning_rate": 9.533315354319494e-05, "loss": 0.0666, "step": 18170 }, { "grad_norm": 0.3531906306743622, "learning_rate": 9.532617581091682e-05, "loss": 0.0617, "step": 18180 }, { "grad_norm": 0.40534135699272156, "learning_rate": 9.531919312183629e-05, "loss": 0.0592, "step": 18190 }, { "grad_norm": 0.5826106071472168, "learning_rate": 9.531220547671688e-05, "loss": 0.0621, "step": 18200 }, { "grad_norm": 0.5806258320808411, "learning_rate": 9.530521287632285e-05, "loss": 0.0674, "step": 18210 }, { "grad_norm": 0.49333471059799194, "learning_rate": 9.529821532141884e-05, "loss": 0.0697, "step": 18220 }, { "grad_norm": 0.5756195783615112, "learning_rate": 9.52912128127701e-05, "loss": 0.0618, "step": 18230 }, { "grad_norm": 0.554771363735199, "learning_rate": 9.528420535114244e-05, "loss": 0.0696, "step": 18240 }, { "grad_norm": 0.533973217010498, "learning_rate": 9.527719293730215e-05, "loss": 0.0783, "step": 18250 }, { "grad_norm": 0.4975667893886566, "learning_rate": 9.527017557201611e-05, "loss": 0.0736, "step": 18260 }, { "grad_norm": 0.4797881841659546, "learning_rate": 9.526315325605176e-05, "loss": 0.0827, "step": 18270 }, { "grad_norm": 0.42011237144470215, "learning_rate": 9.525612599017699e-05, "loss": 0.0564, "step": 18280 }, { "grad_norm": 0.4829649329185486, "learning_rate": 9.524909377516033e-05, "loss": 0.0666, "step": 18290 }, { "grad_norm": 0.5006723999977112, "learning_rate": 9.524205661177081e-05, "loss": 0.0611, "step": 18300 }, { "grad_norm": 0.5433349013328552, "learning_rate": 9.523501450077801e-05, "loss": 0.0749, "step": 18310 }, { "grad_norm": 0.4477561414241791, "learning_rate": 9.522796744295202e-05, "loss": 0.0673, "step": 18320 }, { "grad_norm": 0.5881814956665039, "learning_rate": 9.522091543906352e-05, "loss": 0.0683, "step": 18330 }, { "grad_norm": 0.6248624324798584, "learning_rate": 9.521385848988369e-05, "loss": 0.0613, "step": 18340 }, { "grad_norm": 0.3908185064792633, "learning_rate": 9.520679659618428e-05, "loss": 0.0582, "step": 18350 }, { "grad_norm": 0.4726005792617798, "learning_rate": 9.519972975873754e-05, "loss": 0.0674, "step": 18360 }, { "grad_norm": 0.5145224928855896, "learning_rate": 9.519265797831633e-05, "loss": 0.0611, "step": 18370 }, { "grad_norm": 0.4324630796909332, "learning_rate": 9.518558125569399e-05, "loss": 0.0664, "step": 18380 }, { "grad_norm": 0.5190486907958984, "learning_rate": 9.517849959164442e-05, "loss": 0.0703, "step": 18390 }, { "grad_norm": 0.668719470500946, "learning_rate": 9.517141298694205e-05, "loss": 0.0816, "step": 18400 }, { "grad_norm": 0.4311954975128174, "learning_rate": 9.516432144236188e-05, "loss": 0.0682, "step": 18410 }, { "grad_norm": 0.39830437302589417, "learning_rate": 9.515722495867941e-05, "loss": 0.0648, "step": 18420 }, { "grad_norm": 0.6171777844429016, "learning_rate": 9.515012353667072e-05, "loss": 0.0604, "step": 18430 }, { "grad_norm": 0.7820175886154175, "learning_rate": 9.51430171771124e-05, "loss": 0.065, "step": 18440 }, { "grad_norm": 0.4510914087295532, "learning_rate": 9.513590588078159e-05, "loss": 0.0719, "step": 18450 }, { "grad_norm": 0.6479971408843994, "learning_rate": 9.512878964845597e-05, "loss": 0.0729, "step": 18460 }, { "grad_norm": 0.5577250719070435, "learning_rate": 9.512166848091377e-05, "loss": 0.0844, "step": 18470 }, { "grad_norm": 0.430751234292984, "learning_rate": 9.511454237893376e-05, "loss": 0.0781, "step": 18480 }, { "grad_norm": 0.4996657967567444, "learning_rate": 9.51074113432952e-05, "loss": 0.0569, "step": 18490 }, { "grad_norm": 0.5505456328392029, "learning_rate": 9.510027537477797e-05, "loss": 0.0767, "step": 18500 }, { "grad_norm": 0.467774897813797, "learning_rate": 9.509313447416242e-05, "loss": 0.0759, "step": 18510 }, { "grad_norm": 0.44099298119544983, "learning_rate": 9.508598864222949e-05, "loss": 0.0661, "step": 18520 }, { "grad_norm": 0.5759658217430115, "learning_rate": 9.507883787976062e-05, "loss": 0.0648, "step": 18530 }, { "grad_norm": 0.4964067041873932, "learning_rate": 9.507168218753781e-05, "loss": 0.0642, "step": 18540 }, { "grad_norm": 0.5514315962791443, "learning_rate": 9.506452156634362e-05, "loss": 0.0664, "step": 18550 }, { "grad_norm": 0.45402875542640686, "learning_rate": 9.505735601696109e-05, "loss": 0.0817, "step": 18560 }, { "grad_norm": 0.601093053817749, "learning_rate": 9.505018554017385e-05, "loss": 0.0636, "step": 18570 }, { "grad_norm": 0.7129327654838562, "learning_rate": 9.504301013676604e-05, "loss": 0.0631, "step": 18580 }, { "grad_norm": 0.5367977023124695, "learning_rate": 9.503582980752238e-05, "loss": 0.069, "step": 18590 }, { "grad_norm": 0.6556358933448792, "learning_rate": 9.502864455322809e-05, "loss": 0.0798, "step": 18600 }, { "grad_norm": 0.4361817538738251, "learning_rate": 9.502145437466891e-05, "loss": 0.0728, "step": 18610 }, { "grad_norm": 0.5230607390403748, "learning_rate": 9.501425927263116e-05, "loss": 0.0812, "step": 18620 }, { "grad_norm": 0.7104546427726746, "learning_rate": 9.500705924790172e-05, "loss": 0.0941, "step": 18630 }, { "grad_norm": 0.6961524486541748, "learning_rate": 9.499985430126794e-05, "loss": 0.0729, "step": 18640 }, { "grad_norm": 0.40665486454963684, "learning_rate": 9.499264443351775e-05, "loss": 0.0679, "step": 18650 }, { "grad_norm": 0.47826850414276123, "learning_rate": 9.498542964543961e-05, "loss": 0.0682, "step": 18660 }, { "grad_norm": 0.5255335569381714, "learning_rate": 9.497820993782252e-05, "loss": 0.0686, "step": 18670 }, { "grad_norm": 0.6071751117706299, "learning_rate": 9.497098531145601e-05, "loss": 0.0721, "step": 18680 }, { "grad_norm": 0.637004554271698, "learning_rate": 9.496375576713017e-05, "loss": 0.07, "step": 18690 }, { "grad_norm": 0.38453108072280884, "learning_rate": 9.49565213056356e-05, "loss": 0.0595, "step": 18700 }, { "grad_norm": 0.3436588943004608, "learning_rate": 9.494928192776342e-05, "loss": 0.0617, "step": 18710 }, { "grad_norm": 0.4897957742214203, "learning_rate": 9.494203763430538e-05, "loss": 0.0678, "step": 18720 }, { "grad_norm": 0.5844289660453796, "learning_rate": 9.493478842605366e-05, "loss": 0.098, "step": 18730 }, { "grad_norm": 0.6706228256225586, "learning_rate": 9.492753430380105e-05, "loss": 0.0695, "step": 18740 }, { "grad_norm": 0.49066686630249023, "learning_rate": 9.492027526834083e-05, "loss": 0.0701, "step": 18750 }, { "grad_norm": 0.5518116354942322, "learning_rate": 9.491301132046684e-05, "loss": 0.0755, "step": 18760 }, { "grad_norm": 0.6706849932670593, "learning_rate": 9.490574246097345e-05, "loss": 0.085, "step": 18770 }, { "grad_norm": 0.5056029558181763, "learning_rate": 9.48984686906556e-05, "loss": 0.0783, "step": 18780 }, { "grad_norm": 0.44892942905426025, "learning_rate": 9.489119001030871e-05, "loss": 0.0638, "step": 18790 }, { "grad_norm": 0.4384152293205261, "learning_rate": 9.488390642072878e-05, "loss": 0.0793, "step": 18800 }, { "grad_norm": 0.44314098358154297, "learning_rate": 9.48766179227123e-05, "loss": 0.0668, "step": 18810 }, { "grad_norm": 0.6360728144645691, "learning_rate": 9.486932451705636e-05, "loss": 0.0803, "step": 18820 }, { "grad_norm": 0.6055306196212769, "learning_rate": 9.486202620455857e-05, "loss": 0.0707, "step": 18830 }, { "grad_norm": 0.4198613464832306, "learning_rate": 9.485472298601704e-05, "loss": 0.0678, "step": 18840 }, { "grad_norm": 0.6581206321716309, "learning_rate": 9.484741486223043e-05, "loss": 0.0835, "step": 18850 }, { "grad_norm": 0.6032405495643616, "learning_rate": 9.484010183399797e-05, "loss": 0.0688, "step": 18860 }, { "grad_norm": 0.6927609443664551, "learning_rate": 9.483278390211938e-05, "loss": 0.0747, "step": 18870 }, { "grad_norm": 0.7095508575439453, "learning_rate": 9.482546106739496e-05, "loss": 0.0623, "step": 18880 }, { "grad_norm": 0.5308000445365906, "learning_rate": 9.48181333306255e-05, "loss": 0.0669, "step": 18890 }, { "grad_norm": 0.4934544265270233, "learning_rate": 9.481080069261237e-05, "loss": 0.0821, "step": 18900 }, { "grad_norm": 0.6458343863487244, "learning_rate": 9.480346315415745e-05, "loss": 0.0668, "step": 18910 }, { "grad_norm": 0.5427819490432739, "learning_rate": 9.479612071606314e-05, "loss": 0.0524, "step": 18920 }, { "grad_norm": 0.3463391959667206, "learning_rate": 9.478877337913244e-05, "loss": 0.0628, "step": 18930 }, { "grad_norm": 0.49976781010627747, "learning_rate": 9.478142114416881e-05, "loss": 0.0548, "step": 18940 }, { "grad_norm": 0.4187658727169037, "learning_rate": 9.47740640119763e-05, "loss": 0.0583, "step": 18950 }, { "grad_norm": 0.5274521708488464, "learning_rate": 9.476670198335947e-05, "loss": 0.0763, "step": 18960 }, { "grad_norm": 0.521754801273346, "learning_rate": 9.47593350591234e-05, "loss": 0.0661, "step": 18970 }, { "grad_norm": 0.519373893737793, "learning_rate": 9.475196324007376e-05, "loss": 0.0773, "step": 18980 }, { "grad_norm": 0.6086052060127258, "learning_rate": 9.474458652701669e-05, "loss": 0.0674, "step": 18990 }, { "grad_norm": 0.49530982971191406, "learning_rate": 9.473720492075892e-05, "loss": 0.0674, "step": 19000 }, { "grad_norm": 0.502227783203125, "learning_rate": 9.472981842210768e-05, "loss": 0.0757, "step": 19010 }, { "grad_norm": 0.9225828647613525, "learning_rate": 9.472242703187074e-05, "loss": 0.0739, "step": 19020 }, { "grad_norm": 0.5886405110359192, "learning_rate": 9.471503075085643e-05, "loss": 0.065, "step": 19030 }, { "grad_norm": 0.4411524832248688, "learning_rate": 9.470762957987359e-05, "loss": 0.066, "step": 19040 }, { "grad_norm": 0.39598652720451355, "learning_rate": 9.470022351973158e-05, "loss": 0.0643, "step": 19050 }, { "grad_norm": 0.5870083570480347, "learning_rate": 9.469281257124034e-05, "loss": 0.0643, "step": 19060 }, { "grad_norm": 0.3418735861778259, "learning_rate": 9.46853967352103e-05, "loss": 0.0681, "step": 19070 }, { "grad_norm": 1.0445024967193604, "learning_rate": 9.467797601245246e-05, "loss": 0.0764, "step": 19080 }, { "grad_norm": 0.6294799447059631, "learning_rate": 9.467055040377834e-05, "loss": 0.0721, "step": 19090 }, { "grad_norm": 0.4053105413913727, "learning_rate": 9.466311990999999e-05, "loss": 0.077, "step": 19100 }, { "grad_norm": 0.3645547926425934, "learning_rate": 9.465568453193e-05, "loss": 0.0635, "step": 19110 }, { "grad_norm": 0.34432047605514526, "learning_rate": 9.464824427038148e-05, "loss": 0.0572, "step": 19120 }, { "grad_norm": 0.4694799482822418, "learning_rate": 9.46407991261681e-05, "loss": 0.0793, "step": 19130 }, { "grad_norm": 0.4999270737171173, "learning_rate": 9.463334910010404e-05, "loss": 0.0633, "step": 19140 }, { "grad_norm": 0.3968120813369751, "learning_rate": 9.462589419300403e-05, "loss": 0.0553, "step": 19150 }, { "grad_norm": 0.41254502534866333, "learning_rate": 9.461843440568333e-05, "loss": 0.0713, "step": 19160 }, { "grad_norm": 0.4547724723815918, "learning_rate": 9.461096973895773e-05, "loss": 0.0673, "step": 19170 }, { "grad_norm": 0.4102112352848053, "learning_rate": 9.460350019364355e-05, "loss": 0.0703, "step": 19180 }, { "grad_norm": 0.6573747992515564, "learning_rate": 9.459602577055764e-05, "loss": 0.0831, "step": 19190 }, { "grad_norm": 0.43096184730529785, "learning_rate": 9.45885464705174e-05, "loss": 0.0596, "step": 19200 }, { "grad_norm": 0.5673261284828186, "learning_rate": 9.458106229434076e-05, "loss": 0.0649, "step": 19210 }, { "grad_norm": 0.6647611856460571, "learning_rate": 9.457357324284617e-05, "loss": 0.0608, "step": 19220 }, { "grad_norm": 0.5640179514884949, "learning_rate": 9.456607931685262e-05, "loss": 0.073, "step": 19230 }, { "grad_norm": 0.44001519680023193, "learning_rate": 9.455858051717965e-05, "loss": 0.0636, "step": 19240 }, { "grad_norm": 0.47476354241371155, "learning_rate": 9.45510768446473e-05, "loss": 0.0569, "step": 19250 }, { "grad_norm": 0.6338544487953186, "learning_rate": 9.454356830007618e-05, "loss": 0.0707, "step": 19260 }, { "grad_norm": 0.5022072196006775, "learning_rate": 9.45360548842874e-05, "loss": 0.0523, "step": 19270 }, { "grad_norm": 0.4211325943470001, "learning_rate": 9.452853659810261e-05, "loss": 0.0625, "step": 19280 }, { "grad_norm": 0.48725494742393494, "learning_rate": 9.452101344234401e-05, "loss": 0.0812, "step": 19290 }, { "grad_norm": 0.4794239103794098, "learning_rate": 9.451348541783431e-05, "loss": 0.0695, "step": 19300 }, { "grad_norm": 0.70965975522995, "learning_rate": 9.450595252539678e-05, "loss": 0.0693, "step": 19310 }, { "grad_norm": 0.7072372436523438, "learning_rate": 9.449841476585518e-05, "loss": 0.0754, "step": 19320 }, { "grad_norm": 0.5664085149765015, "learning_rate": 9.449087214003384e-05, "loss": 0.0837, "step": 19330 }, { "grad_norm": 0.6574370861053467, "learning_rate": 9.448332464875765e-05, "loss": 0.0917, "step": 19340 }, { "grad_norm": 0.7239291667938232, "learning_rate": 9.447577229285192e-05, "loss": 0.0878, "step": 19350 }, { "grad_norm": 0.4272611141204834, "learning_rate": 9.446821507314261e-05, "loss": 0.0661, "step": 19360 }, { "grad_norm": 0.4888044595718384, "learning_rate": 9.446065299045617e-05, "loss": 0.0728, "step": 19370 }, { "grad_norm": 0.6561870574951172, "learning_rate": 9.445308604561955e-05, "loss": 0.0777, "step": 19380 }, { "grad_norm": 0.35956934094429016, "learning_rate": 9.444551423946028e-05, "loss": 0.0647, "step": 19390 }, { "grad_norm": 0.4292927086353302, "learning_rate": 9.443793757280638e-05, "loss": 0.0681, "step": 19400 }, { "grad_norm": 0.5523106455802917, "learning_rate": 9.443035604648646e-05, "loss": 0.0657, "step": 19410 }, { "grad_norm": 0.5519589185714722, "learning_rate": 9.44227696613296e-05, "loss": 0.0591, "step": 19420 }, { "grad_norm": 0.7963359951972961, "learning_rate": 9.441517841816542e-05, "loss": 0.0995, "step": 19430 }, { "grad_norm": 0.5314453840255737, "learning_rate": 9.440758231782413e-05, "loss": 0.067, "step": 19440 }, { "grad_norm": 0.4523450434207916, "learning_rate": 9.439998136113639e-05, "loss": 0.0669, "step": 19450 }, { "grad_norm": 0.49138763546943665, "learning_rate": 9.439237554893344e-05, "loss": 0.0654, "step": 19460 }, { "grad_norm": 0.6260332465171814, "learning_rate": 9.438476488204705e-05, "loss": 0.0777, "step": 19470 }, { "grad_norm": 0.4176487326622009, "learning_rate": 9.43771493613095e-05, "loss": 0.0589, "step": 19480 }, { "grad_norm": 0.5080544948577881, "learning_rate": 9.436952898755362e-05, "loss": 0.0794, "step": 19490 }, { "grad_norm": 0.5187861323356628, "learning_rate": 9.436190376161276e-05, "loss": 0.0628, "step": 19500 }, { "grad_norm": 0.4499661326408386, "learning_rate": 9.43542736843208e-05, "loss": 0.0877, "step": 19510 }, { "grad_norm": 0.6525847911834717, "learning_rate": 9.434663875651216e-05, "loss": 0.0606, "step": 19520 }, { "grad_norm": 0.4210343062877655, "learning_rate": 9.433899897902177e-05, "loss": 0.0674, "step": 19530 }, { "grad_norm": 0.3795991837978363, "learning_rate": 9.433135435268511e-05, "loss": 0.0501, "step": 19540 }, { "grad_norm": 0.4327637553215027, "learning_rate": 9.432370487833819e-05, "loss": 0.0652, "step": 19550 }, { "grad_norm": 0.5227122902870178, "learning_rate": 9.431605055681756e-05, "loss": 0.0614, "step": 19560 }, { "grad_norm": 0.4083177447319031, "learning_rate": 9.430839138896026e-05, "loss": 0.0649, "step": 19570 }, { "grad_norm": 0.5659145712852478, "learning_rate": 9.43007273756039e-05, "loss": 0.0682, "step": 19580 }, { "grad_norm": 0.5540937781333923, "learning_rate": 9.429305851758658e-05, "loss": 0.0814, "step": 19590 }, { "grad_norm": 0.555274248123169, "learning_rate": 9.428538481574699e-05, "loss": 0.0771, "step": 19600 }, { "grad_norm": 0.5606322884559631, "learning_rate": 9.42777062709243e-05, "loss": 0.0721, "step": 19610 }, { "grad_norm": 0.4125257432460785, "learning_rate": 9.427002288395821e-05, "loss": 0.06, "step": 19620 }, { "grad_norm": 0.46977993845939636, "learning_rate": 9.426233465568898e-05, "loss": 0.0652, "step": 19630 }, { "grad_norm": 0.3496641516685486, "learning_rate": 9.42546415869574e-05, "loss": 0.058, "step": 19640 }, { "grad_norm": 0.37597784399986267, "learning_rate": 9.424694367860473e-05, "loss": 0.0493, "step": 19650 }, { "grad_norm": 0.399285227060318, "learning_rate": 9.423924093147284e-05, "loss": 0.057, "step": 19660 }, { "grad_norm": 0.4054982364177704, "learning_rate": 9.423153334640407e-05, "loss": 0.0574, "step": 19670 }, { "grad_norm": 0.8054131269454956, "learning_rate": 9.42238209242413e-05, "loss": 0.0514, "step": 19680 }, { "grad_norm": 0.5680443048477173, "learning_rate": 9.421610366582798e-05, "loss": 0.0765, "step": 19690 }, { "grad_norm": 0.6099408864974976, "learning_rate": 9.420838157200803e-05, "loss": 0.0702, "step": 19700 }, { "grad_norm": 0.39724940061569214, "learning_rate": 9.420065464362594e-05, "loss": 0.0768, "step": 19710 }, { "grad_norm": 0.4952755272388458, "learning_rate": 9.419292288152673e-05, "loss": 0.0624, "step": 19720 }, { "grad_norm": 0.45053133368492126, "learning_rate": 9.418518628655588e-05, "loss": 0.0641, "step": 19730 }, { "grad_norm": 0.4753767251968384, "learning_rate": 9.417744485955951e-05, "loss": 0.0579, "step": 19740 }, { "grad_norm": 0.42855849862098694, "learning_rate": 9.41696986013842e-05, "loss": 0.0561, "step": 19750 }, { "grad_norm": 0.48491737246513367, "learning_rate": 9.416194751287705e-05, "loss": 0.0637, "step": 19760 }, { "grad_norm": 0.5457795858383179, "learning_rate": 9.415419159488572e-05, "loss": 0.0608, "step": 19770 }, { "grad_norm": 0.547687292098999, "learning_rate": 9.414643084825837e-05, "loss": 0.071, "step": 19780 }, { "grad_norm": 0.5245007872581482, "learning_rate": 9.413866527384372e-05, "loss": 0.077, "step": 19790 }, { "grad_norm": 0.486406534910202, "learning_rate": 9.4130894872491e-05, "loss": 0.0675, "step": 19800 }, { "grad_norm": 0.5824450850486755, "learning_rate": 9.412311964504998e-05, "loss": 0.0752, "step": 19810 }, { "grad_norm": 0.5437961220741272, "learning_rate": 9.411533959237091e-05, "loss": 0.0773, "step": 19820 }, { "grad_norm": 0.523821234703064, "learning_rate": 9.410755471530464e-05, "loss": 0.0509, "step": 19830 }, { "grad_norm": 0.5382046699523926, "learning_rate": 9.40997650147025e-05, "loss": 0.0667, "step": 19840 }, { "grad_norm": 0.6217731833457947, "learning_rate": 9.409197049141637e-05, "loss": 0.0786, "step": 19850 }, { "grad_norm": 0.43506622314453125, "learning_rate": 9.408417114629863e-05, "loss": 0.0685, "step": 19860 }, { "grad_norm": 0.5220632553100586, "learning_rate": 9.40763669802022e-05, "loss": 0.0575, "step": 19870 }, { "grad_norm": 0.5165217518806458, "learning_rate": 9.406855799398056e-05, "loss": 0.0681, "step": 19880 }, { "grad_norm": 0.6993988156318665, "learning_rate": 9.406074418848767e-05, "loss": 0.0746, "step": 19890 }, { "grad_norm": 0.6302319765090942, "learning_rate": 9.405292556457805e-05, "loss": 0.0878, "step": 19900 }, { "grad_norm": 0.37397128343582153, "learning_rate": 9.404510212310671e-05, "loss": 0.0672, "step": 19910 }, { "grad_norm": 0.4230000376701355, "learning_rate": 9.403727386492924e-05, "loss": 0.062, "step": 19920 }, { "grad_norm": 0.3228847086429596, "learning_rate": 9.40294407909017e-05, "loss": 0.0598, "step": 19930 }, { "grad_norm": 0.5518735647201538, "learning_rate": 9.40216029018807e-05, "loss": 0.0606, "step": 19940 }, { "grad_norm": 0.44024571776390076, "learning_rate": 9.401376019872338e-05, "loss": 0.0568, "step": 19950 }, { "grad_norm": 0.4612334966659546, "learning_rate": 9.400591268228746e-05, "loss": 0.0579, "step": 19960 }, { "grad_norm": 0.555946409702301, "learning_rate": 9.399806035343106e-05, "loss": 0.0667, "step": 19970 }, { "grad_norm": 0.5079373717308044, "learning_rate": 9.399020321301294e-05, "loss": 0.0656, "step": 19980 }, { "grad_norm": 0.5450343489646912, "learning_rate": 9.398234126189234e-05, "loss": 0.0675, "step": 19990 }, { "grad_norm": 0.5182767510414124, "learning_rate": 9.397447450092902e-05, "loss": 0.0718, "step": 20000 }, { "grad_norm": 0.33238485455513, "learning_rate": 9.39666029309833e-05, "loss": 0.0557, "step": 20010 }, { "grad_norm": 0.6462755799293518, "learning_rate": 9.395872655291596e-05, "loss": 0.0588, "step": 20020 }, { "grad_norm": 0.5736261010169983, "learning_rate": 9.395084536758838e-05, "loss": 0.0606, "step": 20030 }, { "grad_norm": 0.528832733631134, "learning_rate": 9.394295937586243e-05, "loss": 0.0787, "step": 20040 }, { "grad_norm": 0.42831650376319885, "learning_rate": 9.393506857860052e-05, "loss": 0.0661, "step": 20050 }, { "grad_norm": 0.428859680891037, "learning_rate": 9.392717297666555e-05, "loss": 0.0527, "step": 20060 }, { "grad_norm": 0.5841279625892639, "learning_rate": 9.391927257092101e-05, "loss": 0.0672, "step": 20070 }, { "grad_norm": 0.3728869557380676, "learning_rate": 9.391136736223085e-05, "loss": 0.0589, "step": 20080 }, { "grad_norm": 0.4842534363269806, "learning_rate": 9.390345735145956e-05, "loss": 0.0687, "step": 20090 }, { "grad_norm": 0.8596099019050598, "learning_rate": 9.389554253947219e-05, "loss": 0.0616, "step": 20100 }, { "grad_norm": 0.7268933653831482, "learning_rate": 9.388762292713428e-05, "loss": 0.0692, "step": 20110 }, { "grad_norm": 0.6018178462982178, "learning_rate": 9.38796985153119e-05, "loss": 0.0748, "step": 20120 }, { "grad_norm": 0.3741055727005005, "learning_rate": 9.387176930487169e-05, "loss": 0.0753, "step": 20130 }, { "grad_norm": 0.9559333920478821, "learning_rate": 9.386383529668072e-05, "loss": 0.0638, "step": 20140 }, { "grad_norm": 0.45440638065338135, "learning_rate": 9.385589649160669e-05, "loss": 0.0676, "step": 20150 }, { "grad_norm": 0.5323059558868408, "learning_rate": 9.384795289051775e-05, "loss": 0.0788, "step": 20160 }, { "grad_norm": 0.6279969215393066, "learning_rate": 9.384000449428261e-05, "loss": 0.0755, "step": 20170 }, { "grad_norm": 0.507439911365509, "learning_rate": 9.383205130377048e-05, "loss": 0.0745, "step": 20180 }, { "grad_norm": 0.4340563416481018, "learning_rate": 9.382409331985114e-05, "loss": 0.0654, "step": 20190 }, { "grad_norm": 0.5271851420402527, "learning_rate": 9.381613054339482e-05, "loss": 0.0579, "step": 20200 }, { "grad_norm": 0.45369717478752136, "learning_rate": 9.380816297527235e-05, "loss": 0.0555, "step": 20210 }, { "grad_norm": 0.5358178615570068, "learning_rate": 9.380019061635506e-05, "loss": 0.0714, "step": 20220 }, { "grad_norm": 0.38014328479766846, "learning_rate": 9.379221346751474e-05, "loss": 0.0677, "step": 20230 }, { "grad_norm": 0.5451798439025879, "learning_rate": 9.378423152962382e-05, "loss": 0.0581, "step": 20240 }, { "grad_norm": 0.5724231004714966, "learning_rate": 9.377624480355517e-05, "loss": 0.0593, "step": 20250 }, { "grad_norm": 0.6366037726402283, "learning_rate": 9.376825329018219e-05, "loss": 0.0693, "step": 20260 }, { "grad_norm": 0.3161980211734772, "learning_rate": 9.376025699037884e-05, "loss": 0.059, "step": 20270 }, { "grad_norm": 0.3925594985485077, "learning_rate": 9.37522559050196e-05, "loss": 0.0644, "step": 20280 }, { "grad_norm": 0.37452834844589233, "learning_rate": 9.37442500349794e-05, "loss": 0.0697, "step": 20290 }, { "grad_norm": 0.5962302088737488, "learning_rate": 9.373623938113381e-05, "loss": 0.0574, "step": 20300 }, { "grad_norm": 0.5403497219085693, "learning_rate": 9.372822394435883e-05, "loss": 0.0718, "step": 20310 }, { "grad_norm": 0.4526095688343048, "learning_rate": 9.372020372553102e-05, "loss": 0.0708, "step": 20320 }, { "grad_norm": 0.5241491794586182, "learning_rate": 9.371217872552746e-05, "loss": 0.0709, "step": 20330 }, { "grad_norm": 0.6870008111000061, "learning_rate": 9.370414894522576e-05, "loss": 0.0669, "step": 20340 }, { "grad_norm": 0.4598480463027954, "learning_rate": 9.369611438550406e-05, "loss": 0.0644, "step": 20350 }, { "grad_norm": 0.546442449092865, "learning_rate": 9.368807504724095e-05, "loss": 0.06, "step": 20360 }, { "grad_norm": 0.4714142680168152, "learning_rate": 9.368003093131565e-05, "loss": 0.0657, "step": 20370 }, { "grad_norm": 0.5158296823501587, "learning_rate": 9.367198203860785e-05, "loss": 0.0829, "step": 20380 }, { "grad_norm": 0.49582791328430176, "learning_rate": 9.366392836999774e-05, "loss": 0.0625, "step": 20390 }, { "grad_norm": 0.5709788203239441, "learning_rate": 9.365586992636607e-05, "loss": 0.0633, "step": 20400 }, { "grad_norm": 0.47246700525283813, "learning_rate": 9.364780670859412e-05, "loss": 0.0594, "step": 20410 }, { "grad_norm": 0.3965008854866028, "learning_rate": 9.363973871756364e-05, "loss": 0.0633, "step": 20420 }, { "grad_norm": 0.8275465965270996, "learning_rate": 9.363166595415696e-05, "loss": 0.0618, "step": 20430 }, { "grad_norm": 0.5784851908683777, "learning_rate": 9.362358841925686e-05, "loss": 0.0704, "step": 20440 }, { "grad_norm": 0.42129072546958923, "learning_rate": 9.361550611374674e-05, "loss": 0.0717, "step": 20450 }, { "grad_norm": 0.6546772122383118, "learning_rate": 9.360741903851043e-05, "loss": 0.0894, "step": 20460 }, { "grad_norm": 0.4859507083892822, "learning_rate": 9.359932719443236e-05, "loss": 0.0874, "step": 20470 }, { "grad_norm": 0.3878924548625946, "learning_rate": 9.35912305823974e-05, "loss": 0.0618, "step": 20480 }, { "grad_norm": 0.4334200322628021, "learning_rate": 9.358312920329101e-05, "loss": 0.0662, "step": 20490 }, { "grad_norm": 0.5441988706588745, "learning_rate": 9.357502305799914e-05, "loss": 0.0642, "step": 20500 }, { "grad_norm": 0.48914217948913574, "learning_rate": 9.356691214740824e-05, "loss": 0.0562, "step": 20510 }, { "grad_norm": 0.4064009487628937, "learning_rate": 9.355879647240535e-05, "loss": 0.0645, "step": 20520 }, { "grad_norm": 0.5163916945457458, "learning_rate": 9.355067603387798e-05, "loss": 0.0702, "step": 20530 }, { "grad_norm": 0.6084264516830444, "learning_rate": 9.354255083271412e-05, "loss": 0.075, "step": 20540 }, { "grad_norm": 0.7935769557952881, "learning_rate": 9.353442086980239e-05, "loss": 0.0677, "step": 20550 }, { "grad_norm": 0.5193598866462708, "learning_rate": 9.352628614603185e-05, "loss": 0.0701, "step": 20560 }, { "grad_norm": 1.5429050922393799, "learning_rate": 9.351814666229209e-05, "loss": 0.0675, "step": 20570 }, { "grad_norm": 0.6123541593551636, "learning_rate": 9.351000241947324e-05, "loss": 0.0692, "step": 20580 }, { "grad_norm": 0.519178569316864, "learning_rate": 9.350185341846594e-05, "loss": 0.0705, "step": 20590 }, { "grad_norm": 0.701431930065155, "learning_rate": 9.349369966016134e-05, "loss": 0.0728, "step": 20600 }, { "grad_norm": 0.678678572177887, "learning_rate": 9.348554114545117e-05, "loss": 0.0755, "step": 20610 }, { "grad_norm": 0.4145781993865967, "learning_rate": 9.347737787522758e-05, "loss": 0.0659, "step": 20620 }, { "grad_norm": 0.5237511396408081, "learning_rate": 9.346920985038332e-05, "loss": 0.0784, "step": 20630 }, { "grad_norm": 0.6725168824195862, "learning_rate": 9.346103707181162e-05, "loss": 0.0725, "step": 20640 }, { "grad_norm": 0.4008435308933258, "learning_rate": 9.345285954040626e-05, "loss": 0.0577, "step": 20650 }, { "grad_norm": 0.5669885873794556, "learning_rate": 9.34446772570615e-05, "loss": 0.0645, "step": 20660 }, { "grad_norm": 0.5810914039611816, "learning_rate": 9.343649022267214e-05, "loss": 0.0623, "step": 20670 }, { "grad_norm": 0.5154998302459717, "learning_rate": 9.342829843813353e-05, "loss": 0.0629, "step": 20680 }, { "grad_norm": 0.5423049330711365, "learning_rate": 9.342010190434149e-05, "loss": 0.0591, "step": 20690 }, { "grad_norm": 0.411165714263916, "learning_rate": 9.34119006221924e-05, "loss": 0.0592, "step": 20700 }, { "grad_norm": 0.4185875952243805, "learning_rate": 9.340369459258313e-05, "loss": 0.0571, "step": 20710 }, { "grad_norm": 0.5203137397766113, "learning_rate": 9.339548381641106e-05, "loss": 0.073, "step": 20720 }, { "grad_norm": 0.31981489062309265, "learning_rate": 9.338726829457413e-05, "loss": 0.0719, "step": 20730 }, { "grad_norm": 0.5808950066566467, "learning_rate": 9.337904802797078e-05, "loss": 0.0586, "step": 20740 }, { "grad_norm": 0.5431593060493469, "learning_rate": 9.337082301749993e-05, "loss": 0.0656, "step": 20750 }, { "grad_norm": 0.4847167432308197, "learning_rate": 9.336259326406109e-05, "loss": 0.0673, "step": 20760 }, { "grad_norm": 0.5118856430053711, "learning_rate": 9.335435876855427e-05, "loss": 0.0836, "step": 20770 }, { "grad_norm": 0.44758981466293335, "learning_rate": 9.334611953187994e-05, "loss": 0.0786, "step": 20780 }, { "grad_norm": 0.45647576451301575, "learning_rate": 9.333787555493914e-05, "loss": 0.0664, "step": 20790 }, { "grad_norm": 1.2298418283462524, "learning_rate": 9.332962683863345e-05, "loss": 0.0782, "step": 20800 }, { "grad_norm": 0.4104004204273224, "learning_rate": 9.332137338386489e-05, "loss": 0.0593, "step": 20810 }, { "grad_norm": 0.4852927327156067, "learning_rate": 9.33131151915361e-05, "loss": 0.0778, "step": 20820 }, { "grad_norm": 0.5593551993370056, "learning_rate": 9.330485226255012e-05, "loss": 0.0665, "step": 20830 }, { "grad_norm": 0.4759643077850342, "learning_rate": 9.329658459781061e-05, "loss": 0.0689, "step": 20840 }, { "grad_norm": 0.46428483724594116, "learning_rate": 9.328831219822172e-05, "loss": 0.0682, "step": 20850 }, { "grad_norm": 0.5296090841293335, "learning_rate": 9.328003506468808e-05, "loss": 0.0746, "step": 20860 }, { "grad_norm": 0.4991951882839203, "learning_rate": 9.327175319811488e-05, "loss": 0.0635, "step": 20870 }, { "grad_norm": 0.4973968267440796, "learning_rate": 9.326346659940781e-05, "loss": 0.0665, "step": 20880 }, { "grad_norm": 0.5054158568382263, "learning_rate": 9.325517526947308e-05, "loss": 0.0665, "step": 20890 }, { "grad_norm": 0.5331205725669861, "learning_rate": 9.32468792092174e-05, "loss": 0.0709, "step": 20900 }, { "grad_norm": 0.5367253422737122, "learning_rate": 9.323857841954803e-05, "loss": 0.0759, "step": 20910 }, { "grad_norm": 0.5159862041473389, "learning_rate": 9.323027290137276e-05, "loss": 0.0601, "step": 20920 }, { "grad_norm": 0.40441641211509705, "learning_rate": 9.322196265559981e-05, "loss": 0.0578, "step": 20930 }, { "grad_norm": 0.5525608062744141, "learning_rate": 9.321364768313803e-05, "loss": 0.0624, "step": 20940 }, { "grad_norm": 0.49481523036956787, "learning_rate": 9.32053279848967e-05, "loss": 0.0687, "step": 20950 }, { "grad_norm": 0.44514065980911255, "learning_rate": 9.319700356178567e-05, "loss": 0.0701, "step": 20960 }, { "grad_norm": 0.44578754901885986, "learning_rate": 9.318867441471527e-05, "loss": 0.064, "step": 20970 }, { "grad_norm": 0.5005937814712524, "learning_rate": 9.318034054459637e-05, "loss": 0.0554, "step": 20980 }, { "grad_norm": 0.46398329734802246, "learning_rate": 9.317200195234034e-05, "loss": 0.0685, "step": 20990 }, { "grad_norm": 0.5304820537567139, "learning_rate": 9.316365863885909e-05, "loss": 0.0788, "step": 21000 }, { "grad_norm": 0.4378073811531067, "learning_rate": 9.315531060506502e-05, "loss": 0.0761, "step": 21010 }, { "grad_norm": 0.534635603427887, "learning_rate": 9.314695785187108e-05, "loss": 0.0621, "step": 21020 }, { "grad_norm": 0.49147531390190125, "learning_rate": 9.313860038019069e-05, "loss": 0.0647, "step": 21030 }, { "grad_norm": 0.7745248079299927, "learning_rate": 9.313023819093782e-05, "loss": 0.0865, "step": 21040 }, { "grad_norm": 0.4827457368373871, "learning_rate": 9.312187128502695e-05, "loss": 0.0712, "step": 21050 }, { "grad_norm": 0.5440381169319153, "learning_rate": 9.311349966337307e-05, "loss": 0.0743, "step": 21060 }, { "grad_norm": 0.6058638691902161, "learning_rate": 9.310512332689169e-05, "loss": 0.0704, "step": 21070 }, { "grad_norm": 0.5407007932662964, "learning_rate": 9.309674227649883e-05, "loss": 0.0578, "step": 21080 }, { "grad_norm": 0.5636566281318665, "learning_rate": 9.308835651311103e-05, "loss": 0.0767, "step": 21090 }, { "grad_norm": 0.5687350034713745, "learning_rate": 9.307996603764533e-05, "loss": 0.0688, "step": 21100 }, { "grad_norm": 0.49709552526474, "learning_rate": 9.307157085101932e-05, "loss": 0.0656, "step": 21110 }, { "grad_norm": 0.5063421130180359, "learning_rate": 9.306317095415109e-05, "loss": 0.0553, "step": 21120 }, { "grad_norm": 0.43830132484436035, "learning_rate": 9.305476634795922e-05, "loss": 0.0666, "step": 21130 }, { "grad_norm": 0.4807336628437042, "learning_rate": 9.304635703336284e-05, "loss": 0.0642, "step": 21140 }, { "grad_norm": 0.4034610390663147, "learning_rate": 9.303794301128157e-05, "loss": 0.0599, "step": 21150 }, { "grad_norm": 0.8195182085037231, "learning_rate": 9.302952428263555e-05, "loss": 0.0609, "step": 21160 }, { "grad_norm": 0.4720141589641571, "learning_rate": 9.302110084834545e-05, "loss": 0.0627, "step": 21170 }, { "grad_norm": 0.46184295415878296, "learning_rate": 9.301267270933245e-05, "loss": 0.0561, "step": 21180 }, { "grad_norm": 0.4151599407196045, "learning_rate": 9.300423986651823e-05, "loss": 0.0602, "step": 21190 }, { "grad_norm": 0.5357250571250916, "learning_rate": 9.299580232082501e-05, "loss": 0.0641, "step": 21200 }, { "grad_norm": 0.44359469413757324, "learning_rate": 9.298736007317547e-05, "loss": 0.0559, "step": 21210 }, { "grad_norm": 0.5036507248878479, "learning_rate": 9.297891312449288e-05, "loss": 0.0623, "step": 21220 }, { "grad_norm": 0.5683504343032837, "learning_rate": 9.297046147570094e-05, "loss": 0.0745, "step": 21230 }, { "grad_norm": 0.4012134373188019, "learning_rate": 9.296200512772396e-05, "loss": 0.0578, "step": 21240 }, { "grad_norm": 0.2769872844219208, "learning_rate": 9.295354408148668e-05, "loss": 0.0536, "step": 21250 }, { "grad_norm": 0.4550328850746155, "learning_rate": 9.294507833791441e-05, "loss": 0.0598, "step": 21260 }, { "grad_norm": 0.7846542000770569, "learning_rate": 9.293660789793295e-05, "loss": 0.0712, "step": 21270 }, { "grad_norm": 0.40943142771720886, "learning_rate": 9.292813276246858e-05, "loss": 0.0624, "step": 21280 }, { "grad_norm": 0.5126705169677734, "learning_rate": 9.291965293244816e-05, "loss": 0.0689, "step": 21290 }, { "grad_norm": 0.55120450258255, "learning_rate": 9.291116840879904e-05, "loss": 0.0539, "step": 21300 }, { "grad_norm": 0.5127356052398682, "learning_rate": 9.290267919244904e-05, "loss": 0.0663, "step": 21310 }, { "grad_norm": 0.6967123746871948, "learning_rate": 9.289418528432655e-05, "loss": 0.0629, "step": 21320 }, { "grad_norm": 0.6446226239204407, "learning_rate": 9.288568668536045e-05, "loss": 0.0843, "step": 21330 }, { "grad_norm": 0.3915710747241974, "learning_rate": 9.287718339648013e-05, "loss": 0.062, "step": 21340 }, { "grad_norm": 0.5041828155517578, "learning_rate": 9.28686754186155e-05, "loss": 0.0602, "step": 21350 }, { "grad_norm": 0.5175696015357971, "learning_rate": 9.286016275269698e-05, "loss": 0.0784, "step": 21360 }, { "grad_norm": 0.44520145654678345, "learning_rate": 9.285164539965551e-05, "loss": 0.0602, "step": 21370 }, { "grad_norm": 0.4205974042415619, "learning_rate": 9.284312336042251e-05, "loss": 0.059, "step": 21380 }, { "grad_norm": 0.53826904296875, "learning_rate": 9.283459663592996e-05, "loss": 0.0654, "step": 21390 }, { "grad_norm": 0.48663949966430664, "learning_rate": 9.282606522711033e-05, "loss": 0.0668, "step": 21400 }, { "grad_norm": 0.5370773673057556, "learning_rate": 9.281752913489657e-05, "loss": 0.0763, "step": 21410 }, { "grad_norm": 0.48987969756126404, "learning_rate": 9.280898836022222e-05, "loss": 0.0681, "step": 21420 }, { "grad_norm": 0.35743939876556396, "learning_rate": 9.280044290402126e-05, "loss": 0.0573, "step": 21430 }, { "grad_norm": 0.4883826971054077, "learning_rate": 9.279189276722821e-05, "loss": 0.0692, "step": 21440 }, { "grad_norm": 0.4195312559604645, "learning_rate": 9.278333795077812e-05, "loss": 0.0698, "step": 21450 }, { "grad_norm": 0.5044364333152771, "learning_rate": 9.27747784556065e-05, "loss": 0.0559, "step": 21460 }, { "grad_norm": 0.41507887840270996, "learning_rate": 9.276621428264942e-05, "loss": 0.074, "step": 21470 }, { "grad_norm": 0.7040372490882874, "learning_rate": 9.275764543284345e-05, "loss": 0.0786, "step": 21480 }, { "grad_norm": 0.4950430691242218, "learning_rate": 9.274907190712566e-05, "loss": 0.0728, "step": 21490 }, { "grad_norm": 0.7275922894477844, "learning_rate": 9.274049370643363e-05, "loss": 0.0669, "step": 21500 }, { "grad_norm": 0.45210209488868713, "learning_rate": 9.273191083170547e-05, "loss": 0.0717, "step": 21510 }, { "grad_norm": 0.5742630362510681, "learning_rate": 9.27233232838798e-05, "loss": 0.0601, "step": 21520 }, { "grad_norm": 0.4416639804840088, "learning_rate": 9.27147310638957e-05, "loss": 0.0613, "step": 21530 }, { "grad_norm": 0.41307348012924194, "learning_rate": 9.270613417269286e-05, "loss": 0.0685, "step": 21540 }, { "grad_norm": 0.42268261313438416, "learning_rate": 9.269753261121138e-05, "loss": 0.0667, "step": 21550 }, { "grad_norm": 0.30660706758499146, "learning_rate": 9.268892638039194e-05, "loss": 0.066, "step": 21560 }, { "grad_norm": 0.3949110805988312, "learning_rate": 9.268031548117569e-05, "loss": 0.0691, "step": 21570 }, { "grad_norm": 0.42437922954559326, "learning_rate": 9.26716999145043e-05, "loss": 0.0548, "step": 21580 }, { "grad_norm": 0.4452384114265442, "learning_rate": 9.266307968131998e-05, "loss": 0.0572, "step": 21590 }, { "grad_norm": 0.4734337329864502, "learning_rate": 9.26544547825654e-05, "loss": 0.0544, "step": 21600 }, { "grad_norm": 0.4926172196865082, "learning_rate": 9.264582521918376e-05, "loss": 0.0576, "step": 21610 }, { "grad_norm": 0.5594095587730408, "learning_rate": 9.263719099211881e-05, "loss": 0.0566, "step": 21620 }, { "grad_norm": 0.977135419845581, "learning_rate": 9.262855210231476e-05, "loss": 0.0722, "step": 21630 }, { "grad_norm": 0.39924025535583496, "learning_rate": 9.261990855071633e-05, "loss": 0.0565, "step": 21640 }, { "grad_norm": 0.44638288021087646, "learning_rate": 9.261126033826878e-05, "loss": 0.0591, "step": 21650 }, { "grad_norm": 0.4501495957374573, "learning_rate": 9.260260746591786e-05, "loss": 0.0577, "step": 21660 }, { "grad_norm": 0.5706306099891663, "learning_rate": 9.259394993460985e-05, "loss": 0.0531, "step": 21670 }, { "grad_norm": 0.39773231744766235, "learning_rate": 9.258528774529151e-05, "loss": 0.0608, "step": 21680 }, { "grad_norm": 0.6713458895683289, "learning_rate": 9.257662089891013e-05, "loss": 0.0617, "step": 21690 }, { "grad_norm": 0.6341748237609863, "learning_rate": 9.25679493964135e-05, "loss": 0.0638, "step": 21700 }, { "grad_norm": 0.3548760712146759, "learning_rate": 9.255927323874994e-05, "loss": 0.053, "step": 21710 }, { "grad_norm": 0.5773940086364746, "learning_rate": 9.255059242686822e-05, "loss": 0.0722, "step": 21720 }, { "grad_norm": 0.4975755512714386, "learning_rate": 9.254190696171769e-05, "loss": 0.0631, "step": 21730 }, { "grad_norm": 0.5899263024330139, "learning_rate": 9.25332168442482e-05, "loss": 0.0769, "step": 21740 }, { "grad_norm": 0.5085222125053406, "learning_rate": 9.252452207541004e-05, "loss": 0.0508, "step": 21750 }, { "grad_norm": 0.5854556560516357, "learning_rate": 9.251582265615409e-05, "loss": 0.0798, "step": 21760 }, { "grad_norm": 0.49656087160110474, "learning_rate": 9.250711858743169e-05, "loss": 0.0664, "step": 21770 }, { "grad_norm": 0.5834074020385742, "learning_rate": 9.24984098701947e-05, "loss": 0.0765, "step": 21780 }, { "grad_norm": 0.5190762281417847, "learning_rate": 9.248969650539552e-05, "loss": 0.066, "step": 21790 }, { "grad_norm": 0.9187321662902832, "learning_rate": 9.2480978493987e-05, "loss": 0.0678, "step": 21800 }, { "grad_norm": 0.7532273530960083, "learning_rate": 9.247225583692256e-05, "loss": 0.0618, "step": 21810 }, { "grad_norm": 0.5244541764259338, "learning_rate": 9.246352853515607e-05, "loss": 0.0606, "step": 21820 }, { "grad_norm": 0.5541168451309204, "learning_rate": 9.245479658964194e-05, "loss": 0.0583, "step": 21830 }, { "grad_norm": 0.5586326718330383, "learning_rate": 9.244606000133507e-05, "loss": 0.0631, "step": 21840 }, { "grad_norm": 0.4540179669857025, "learning_rate": 9.24373187711909e-05, "loss": 0.0623, "step": 21850 }, { "grad_norm": 0.4609430730342865, "learning_rate": 9.242857290016537e-05, "loss": 0.0659, "step": 21860 }, { "grad_norm": 0.43486618995666504, "learning_rate": 9.241982238921488e-05, "loss": 0.0691, "step": 21870 }, { "grad_norm": 0.4080575108528137, "learning_rate": 9.24110672392964e-05, "loss": 0.0469, "step": 21880 }, { "grad_norm": 0.44749921560287476, "learning_rate": 9.240230745136737e-05, "loss": 0.0579, "step": 21890 }, { "grad_norm": 0.4399515986442566, "learning_rate": 9.239354302638575e-05, "loss": 0.0542, "step": 21900 }, { "grad_norm": 0.6806101202964783, "learning_rate": 9.238477396531e-05, "loss": 0.0636, "step": 21910 }, { "grad_norm": 0.6983631253242493, "learning_rate": 9.23760002690991e-05, "loss": 0.0611, "step": 21920 }, { "grad_norm": 0.4270850718021393, "learning_rate": 9.236722193871252e-05, "loss": 0.0565, "step": 21930 }, { "grad_norm": 0.4097515642642975, "learning_rate": 9.235843897511023e-05, "loss": 0.0674, "step": 21940 }, { "grad_norm": 0.5550841689109802, "learning_rate": 9.234965137925276e-05, "loss": 0.0574, "step": 21950 }, { "grad_norm": 0.5012274980545044, "learning_rate": 9.234085915210108e-05, "loss": 0.0614, "step": 21960 }, { "grad_norm": 0.5214057564735413, "learning_rate": 9.23320622946167e-05, "loss": 0.0572, "step": 21970 }, { "grad_norm": 0.4146561622619629, "learning_rate": 9.232326080776163e-05, "loss": 0.0453, "step": 21980 }, { "grad_norm": 0.5621159076690674, "learning_rate": 9.23144546924984e-05, "loss": 0.0495, "step": 21990 }, { "grad_norm": 0.5131868124008179, "learning_rate": 9.230564394979e-05, "loss": 0.0534, "step": 22000 }, { "grad_norm": 0.4331507086753845, "learning_rate": 9.22968285806e-05, "loss": 0.0537, "step": 22010 }, { "grad_norm": 0.38420864939689636, "learning_rate": 9.228800858589242e-05, "loss": 0.0637, "step": 22020 }, { "grad_norm": 0.5491302609443665, "learning_rate": 9.227918396663179e-05, "loss": 0.0603, "step": 22030 }, { "grad_norm": 0.6150916814804077, "learning_rate": 9.227035472378319e-05, "loss": 0.0567, "step": 22040 }, { "grad_norm": 0.40700873732566833, "learning_rate": 9.226152085831213e-05, "loss": 0.0606, "step": 22050 }, { "grad_norm": 0.6345173120498657, "learning_rate": 9.22526823711847e-05, "loss": 0.0659, "step": 22060 }, { "grad_norm": 0.4260636568069458, "learning_rate": 9.224383926336745e-05, "loss": 0.061, "step": 22070 }, { "grad_norm": 0.5113006830215454, "learning_rate": 9.223499153582744e-05, "loss": 0.0537, "step": 22080 }, { "grad_norm": 0.3773685097694397, "learning_rate": 9.222613918953226e-05, "loss": 0.0684, "step": 22090 }, { "grad_norm": 0.6821680665016174, "learning_rate": 9.221728222544999e-05, "loss": 0.0695, "step": 22100 }, { "grad_norm": 0.49978190660476685, "learning_rate": 9.22084206445492e-05, "loss": 0.0648, "step": 22110 }, { "grad_norm": 0.4753914475440979, "learning_rate": 9.2199554447799e-05, "loss": 0.0717, "step": 22120 }, { "grad_norm": 0.5829538106918335, "learning_rate": 9.219068363616897e-05, "loss": 0.0685, "step": 22130 }, { "grad_norm": 0.39439696073532104, "learning_rate": 9.218180821062919e-05, "loss": 0.0662, "step": 22140 }, { "grad_norm": 0.3890421688556671, "learning_rate": 9.21729281721503e-05, "loss": 0.0742, "step": 22150 }, { "grad_norm": 0.4060479998588562, "learning_rate": 9.216404352170339e-05, "loss": 0.068, "step": 22160 }, { "grad_norm": 0.38178351521492004, "learning_rate": 9.215515426026007e-05, "loss": 0.0651, "step": 22170 }, { "grad_norm": 0.546821117401123, "learning_rate": 9.214626038879246e-05, "loss": 0.0643, "step": 22180 }, { "grad_norm": 0.6745181083679199, "learning_rate": 9.21373619082732e-05, "loss": 0.0652, "step": 22190 }, { "grad_norm": 0.6177012324333191, "learning_rate": 9.212845881967535e-05, "loss": 0.0705, "step": 22200 }, { "grad_norm": 0.4824872314929962, "learning_rate": 9.211955112397262e-05, "loss": 0.0657, "step": 22210 }, { "grad_norm": 0.40461239218711853, "learning_rate": 9.211063882213909e-05, "loss": 0.0601, "step": 22220 }, { "grad_norm": 0.43974411487579346, "learning_rate": 9.210172191514942e-05, "loss": 0.0607, "step": 22230 }, { "grad_norm": 0.5052618980407715, "learning_rate": 9.209280040397874e-05, "loss": 0.065, "step": 22240 }, { "grad_norm": 0.46483686566352844, "learning_rate": 9.208387428960268e-05, "loss": 0.0479, "step": 22250 }, { "grad_norm": 0.4165145754814148, "learning_rate": 9.20749435729974e-05, "loss": 0.0616, "step": 22260 }, { "grad_norm": 0.37855100631713867, "learning_rate": 9.206600825513957e-05, "loss": 0.0559, "step": 22270 }, { "grad_norm": 0.5701457262039185, "learning_rate": 9.20570683370063e-05, "loss": 0.0702, "step": 22280 }, { "grad_norm": 0.4444626569747925, "learning_rate": 9.204812381957528e-05, "loss": 0.0601, "step": 22290 }, { "grad_norm": 0.5496653914451599, "learning_rate": 9.203917470382465e-05, "loss": 0.0641, "step": 22300 }, { "grad_norm": 0.45631083846092224, "learning_rate": 9.203022099073309e-05, "loss": 0.0588, "step": 22310 }, { "grad_norm": 0.3973020613193512, "learning_rate": 9.202126268127976e-05, "loss": 0.061, "step": 22320 }, { "grad_norm": 0.5128887295722961, "learning_rate": 9.20122997764443e-05, "loss": 0.0671, "step": 22330 }, { "grad_norm": 0.60735023021698, "learning_rate": 9.200333227720692e-05, "loss": 0.0674, "step": 22340 }, { "grad_norm": 0.6042155623435974, "learning_rate": 9.199436018454826e-05, "loss": 0.0686, "step": 22350 }, { "grad_norm": 0.5201776027679443, "learning_rate": 9.198538349944952e-05, "loss": 0.0583, "step": 22360 }, { "grad_norm": 0.39716511964797974, "learning_rate": 9.197640222289234e-05, "loss": 0.0765, "step": 22370 }, { "grad_norm": 0.41750290989875793, "learning_rate": 9.196741635585895e-05, "loss": 0.0619, "step": 22380 }, { "grad_norm": 0.6679397225379944, "learning_rate": 9.195842589933199e-05, "loss": 0.0519, "step": 22390 }, { "grad_norm": 0.4621851444244385, "learning_rate": 9.194943085429466e-05, "loss": 0.0649, "step": 22400 }, { "grad_norm": 0.5833415389060974, "learning_rate": 9.194043122173065e-05, "loss": 0.0821, "step": 22410 }, { "grad_norm": 0.4555233418941498, "learning_rate": 9.193142700262413e-05, "loss": 0.0634, "step": 22420 }, { "grad_norm": 0.5201921463012695, "learning_rate": 9.192241819795979e-05, "loss": 0.0538, "step": 22430 }, { "grad_norm": 0.47694405913352966, "learning_rate": 9.191340480872284e-05, "loss": 0.0643, "step": 22440 }, { "grad_norm": 0.4297240376472473, "learning_rate": 9.190438683589895e-05, "loss": 0.0704, "step": 22450 }, { "grad_norm": 0.4983634352684021, "learning_rate": 9.189536428047432e-05, "loss": 0.0682, "step": 22460 }, { "grad_norm": 0.40941545367240906, "learning_rate": 9.188633714343564e-05, "loss": 0.0588, "step": 22470 }, { "grad_norm": 1.2445271015167236, "learning_rate": 9.18773054257701e-05, "loss": 0.0794, "step": 22480 }, { "grad_norm": 0.7082418203353882, "learning_rate": 9.18682691284654e-05, "loss": 0.0592, "step": 22490 }, { "grad_norm": 0.5538649559020996, "learning_rate": 9.185922825250974e-05, "loss": 0.0688, "step": 22500 }, { "grad_norm": 0.4409986138343811, "learning_rate": 9.185018279889181e-05, "loss": 0.0557, "step": 22510 }, { "grad_norm": 0.4371313154697418, "learning_rate": 9.184113276860082e-05, "loss": 0.0626, "step": 22520 }, { "grad_norm": 0.48195627331733704, "learning_rate": 9.183207816262645e-05, "loss": 0.0691, "step": 22530 }, { "grad_norm": 0.7125113606452942, "learning_rate": 9.182301898195891e-05, "loss": 0.0599, "step": 22540 }, { "grad_norm": 0.5717142820358276, "learning_rate": 9.181395522758889e-05, "loss": 0.0857, "step": 22550 }, { "grad_norm": 0.6887423992156982, "learning_rate": 9.180488690050759e-05, "loss": 0.0844, "step": 22560 }, { "grad_norm": 0.5563712120056152, "learning_rate": 9.179581400170671e-05, "loss": 0.0697, "step": 22570 }, { "grad_norm": 0.3743048310279846, "learning_rate": 9.178673653217845e-05, "loss": 0.0621, "step": 22580 }, { "grad_norm": 0.5241791605949402, "learning_rate": 9.177765449291551e-05, "loss": 0.0537, "step": 22590 }, { "grad_norm": 0.895290195941925, "learning_rate": 9.176856788491109e-05, "loss": 0.0668, "step": 22600 }, { "grad_norm": 0.5318543910980225, "learning_rate": 9.175947670915887e-05, "loss": 0.0673, "step": 22610 }, { "grad_norm": 0.5802831053733826, "learning_rate": 9.175038096665309e-05, "loss": 0.0632, "step": 22620 }, { "grad_norm": 0.404263973236084, "learning_rate": 9.17412806583884e-05, "loss": 0.0537, "step": 22630 }, { "grad_norm": 0.5910850167274475, "learning_rate": 9.173217578536002e-05, "loss": 0.0797, "step": 22640 }, { "grad_norm": 0.5235990881919861, "learning_rate": 9.172306634856362e-05, "loss": 0.0573, "step": 22650 }, { "grad_norm": 0.5086278915405273, "learning_rate": 9.171395234899545e-05, "loss": 0.0699, "step": 22660 }, { "grad_norm": 0.3468773066997528, "learning_rate": 9.170483378765214e-05, "loss": 0.0591, "step": 22670 }, { "grad_norm": 0.4708344340324402, "learning_rate": 9.169571066553091e-05, "loss": 0.059, "step": 22680 }, { "grad_norm": 0.34104421734809875, "learning_rate": 9.168658298362946e-05, "loss": 0.0568, "step": 22690 }, { "grad_norm": 0.4122150242328644, "learning_rate": 9.167745074294598e-05, "loss": 0.0632, "step": 22700 }, { "grad_norm": 0.48152974247932434, "learning_rate": 9.166831394447913e-05, "loss": 0.0595, "step": 22710 }, { "grad_norm": 0.3727230727672577, "learning_rate": 9.165917258922812e-05, "loss": 0.0444, "step": 22720 }, { "grad_norm": 0.5166029334068298, "learning_rate": 9.165002667819262e-05, "loss": 0.0626, "step": 22730 }, { "grad_norm": 0.46988314390182495, "learning_rate": 9.164087621237282e-05, "loss": 0.0486, "step": 22740 }, { "grad_norm": 0.5309601426124573, "learning_rate": 9.163172119276942e-05, "loss": 0.0674, "step": 22750 }, { "grad_norm": 0.43688103556632996, "learning_rate": 9.162256162038358e-05, "loss": 0.072, "step": 22760 }, { "grad_norm": 0.3427642285823822, "learning_rate": 9.161339749621698e-05, "loss": 0.0636, "step": 22770 }, { "grad_norm": 0.39693424105644226, "learning_rate": 9.160422882127177e-05, "loss": 0.0569, "step": 22780 }, { "grad_norm": 0.4395610988140106, "learning_rate": 9.159505559655069e-05, "loss": 0.0552, "step": 22790 }, { "grad_norm": 1.7221814393997192, "learning_rate": 9.158587782305684e-05, "loss": 0.0573, "step": 22800 }, { "grad_norm": 0.6546832323074341, "learning_rate": 9.157669550179391e-05, "loss": 0.061, "step": 22810 }, { "grad_norm": 0.4333503842353821, "learning_rate": 9.156750863376609e-05, "loss": 0.0678, "step": 22820 }, { "grad_norm": 0.43848761916160583, "learning_rate": 9.155831721997801e-05, "loss": 0.062, "step": 22830 }, { "grad_norm": 0.5441872477531433, "learning_rate": 9.154912126143484e-05, "loss": 0.061, "step": 22840 }, { "grad_norm": 0.5952993631362915, "learning_rate": 9.153992075914224e-05, "loss": 0.0746, "step": 22850 }, { "grad_norm": 0.6056516170501709, "learning_rate": 9.153071571410635e-05, "loss": 0.0598, "step": 22860 }, { "grad_norm": 0.44644346833229065, "learning_rate": 9.152150612733384e-05, "loss": 0.0512, "step": 22870 }, { "grad_norm": 0.39946237206459045, "learning_rate": 9.151229199983184e-05, "loss": 0.0581, "step": 22880 }, { "grad_norm": 0.5658367872238159, "learning_rate": 9.150307333260802e-05, "loss": 0.0615, "step": 22890 }, { "grad_norm": 0.4612862467765808, "learning_rate": 9.149385012667048e-05, "loss": 0.0719, "step": 22900 }, { "grad_norm": 1.0332438945770264, "learning_rate": 9.148462238302788e-05, "loss": 0.0764, "step": 22910 }, { "grad_norm": 0.5278430581092834, "learning_rate": 9.147539010268936e-05, "loss": 0.0663, "step": 22920 }, { "grad_norm": 0.8405671715736389, "learning_rate": 9.14661532866645e-05, "loss": 0.0676, "step": 22930 }, { "grad_norm": 0.4384159743785858, "learning_rate": 9.145691193596348e-05, "loss": 0.0565, "step": 22940 }, { "grad_norm": 0.6737565398216248, "learning_rate": 9.144766605159691e-05, "loss": 0.0859, "step": 22950 }, { "grad_norm": 0.5090729594230652, "learning_rate": 9.14384156345759e-05, "loss": 0.0597, "step": 22960 }, { "grad_norm": 0.5972506999969482, "learning_rate": 9.142916068591204e-05, "loss": 0.0662, "step": 22970 }, { "grad_norm": 0.4203348755836487, "learning_rate": 9.141990120661746e-05, "loss": 0.0628, "step": 22980 }, { "grad_norm": 0.4435538351535797, "learning_rate": 9.141063719770475e-05, "loss": 0.0625, "step": 22990 }, { "grad_norm": 0.5618798732757568, "learning_rate": 9.140136866018704e-05, "loss": 0.0584, "step": 23000 }, { "grad_norm": 0.3989633619785309, "learning_rate": 9.139209559507788e-05, "loss": 0.0507, "step": 23010 }, { "grad_norm": 0.45404529571533203, "learning_rate": 9.13828180033914e-05, "loss": 0.0582, "step": 23020 }, { "grad_norm": 0.5660241842269897, "learning_rate": 9.137353588614212e-05, "loss": 0.0631, "step": 23030 }, { "grad_norm": 0.3712165355682373, "learning_rate": 9.136424924434519e-05, "loss": 0.0573, "step": 23040 }, { "grad_norm": 0.47505396604537964, "learning_rate": 9.135495807901615e-05, "loss": 0.0694, "step": 23050 }, { "grad_norm": 0.3455374240875244, "learning_rate": 9.134566239117108e-05, "loss": 0.0644, "step": 23060 }, { "grad_norm": 0.5885894894599915, "learning_rate": 9.13363621818265e-05, "loss": 0.0471, "step": 23070 }, { "grad_norm": 0.4139184057712555, "learning_rate": 9.132705745199953e-05, "loss": 0.0607, "step": 23080 }, { "grad_norm": 0.4314488470554352, "learning_rate": 9.131774820270768e-05, "loss": 0.0521, "step": 23090 }, { "grad_norm": 0.5321400165557861, "learning_rate": 9.130843443496901e-05, "loss": 0.0648, "step": 23100 }, { "grad_norm": 0.4520702362060547, "learning_rate": 9.129911614980206e-05, "loss": 0.0782, "step": 23110 }, { "grad_norm": 0.48439303040504456, "learning_rate": 9.128979334822584e-05, "loss": 0.064, "step": 23120 }, { "grad_norm": 0.38302257657051086, "learning_rate": 9.128046603125992e-05, "loss": 0.0574, "step": 23130 }, { "grad_norm": 0.48404452204704285, "learning_rate": 9.12711341999243e-05, "loss": 0.0528, "step": 23140 }, { "grad_norm": 0.46273893117904663, "learning_rate": 9.12617978552395e-05, "loss": 0.0826, "step": 23150 }, { "grad_norm": 0.4647936224937439, "learning_rate": 9.12524569982265e-05, "loss": 0.0619, "step": 23160 }, { "grad_norm": 0.6793041825294495, "learning_rate": 9.124311162990684e-05, "loss": 0.0708, "step": 23170 }, { "grad_norm": 0.4280964434146881, "learning_rate": 9.12337617513025e-05, "loss": 0.0621, "step": 23180 }, { "grad_norm": 0.4658789336681366, "learning_rate": 9.122440736343596e-05, "loss": 0.0536, "step": 23190 }, { "grad_norm": 0.382561594247818, "learning_rate": 9.12150484673302e-05, "loss": 0.0554, "step": 23200 }, { "grad_norm": 0.3524417281150818, "learning_rate": 9.120568506400873e-05, "loss": 0.0553, "step": 23210 }, { "grad_norm": 0.36175233125686646, "learning_rate": 9.119631715449548e-05, "loss": 0.0489, "step": 23220 }, { "grad_norm": 0.5297385454177856, "learning_rate": 9.118694473981493e-05, "loss": 0.052, "step": 23230 }, { "grad_norm": 0.7056158185005188, "learning_rate": 9.117756782099203e-05, "loss": 0.0668, "step": 23240 }, { "grad_norm": 0.4544129967689514, "learning_rate": 9.11681863990522e-05, "loss": 0.0653, "step": 23250 }, { "grad_norm": 0.7680746912956238, "learning_rate": 9.115880047502142e-05, "loss": 0.0594, "step": 23260 }, { "grad_norm": 0.7421503067016602, "learning_rate": 9.114941004992609e-05, "loss": 0.0706, "step": 23270 }, { "grad_norm": 0.4863531291484833, "learning_rate": 9.114001512479317e-05, "loss": 0.0616, "step": 23280 }, { "grad_norm": 0.6408441066741943, "learning_rate": 9.113061570065003e-05, "loss": 0.0663, "step": 23290 }, { "grad_norm": 0.4377519488334656, "learning_rate": 9.112121177852459e-05, "loss": 0.0654, "step": 23300 }, { "grad_norm": 0.3976939916610718, "learning_rate": 9.111180335944527e-05, "loss": 0.0667, "step": 23310 }, { "grad_norm": 0.5908895134925842, "learning_rate": 9.110239044444093e-05, "loss": 0.0645, "step": 23320 }, { "grad_norm": 0.5299714207649231, "learning_rate": 9.109297303454099e-05, "loss": 0.0603, "step": 23330 }, { "grad_norm": 0.44910478591918945, "learning_rate": 9.108355113077526e-05, "loss": 0.0592, "step": 23340 }, { "grad_norm": 0.401115357875824, "learning_rate": 9.107412473417419e-05, "loss": 0.0677, "step": 23350 }, { "grad_norm": 0.5067315101623535, "learning_rate": 9.106469384576858e-05, "loss": 0.0655, "step": 23360 }, { "grad_norm": 0.7796066403388977, "learning_rate": 9.105525846658978e-05, "loss": 0.0781, "step": 23370 }, { "grad_norm": 0.4703899919986725, "learning_rate": 9.104581859766965e-05, "loss": 0.0707, "step": 23380 }, { "grad_norm": 0.5516767501831055, "learning_rate": 9.10363742400405e-05, "loss": 0.0693, "step": 23390 }, { "grad_norm": 0.5041712522506714, "learning_rate": 9.102692539473518e-05, "loss": 0.0627, "step": 23400 }, { "grad_norm": 0.555923342704773, "learning_rate": 9.101747206278697e-05, "loss": 0.0721, "step": 23410 }, { "grad_norm": 0.5144757032394409, "learning_rate": 9.100801424522968e-05, "loss": 0.0593, "step": 23420 }, { "grad_norm": 0.38996604084968567, "learning_rate": 9.099855194309762e-05, "loss": 0.0585, "step": 23430 }, { "grad_norm": 0.40190431475639343, "learning_rate": 9.098908515742554e-05, "loss": 0.0551, "step": 23440 }, { "grad_norm": 0.49032536149024963, "learning_rate": 9.097961388924873e-05, "loss": 0.0631, "step": 23450 }, { "grad_norm": 0.5103908777236938, "learning_rate": 9.097013813960298e-05, "loss": 0.0519, "step": 23460 }, { "grad_norm": 0.6675397753715515, "learning_rate": 9.09606579095245e-05, "loss": 0.0608, "step": 23470 }, { "grad_norm": 0.47645360231399536, "learning_rate": 9.095117320005008e-05, "loss": 0.0595, "step": 23480 }, { "grad_norm": 0.48872044682502747, "learning_rate": 9.094168401221691e-05, "loss": 0.061, "step": 23490 }, { "grad_norm": 0.43116840720176697, "learning_rate": 9.093219034706273e-05, "loss": 0.0687, "step": 23500 }, { "grad_norm": 0.4258229732513428, "learning_rate": 9.092269220562577e-05, "loss": 0.0692, "step": 23510 }, { "grad_norm": 0.4610588252544403, "learning_rate": 9.09131895889447e-05, "loss": 0.0844, "step": 23520 }, { "grad_norm": 0.4911282956600189, "learning_rate": 9.090368249805873e-05, "loss": 0.0545, "step": 23530 }, { "grad_norm": 0.4030728340148926, "learning_rate": 9.089417093400754e-05, "loss": 0.052, "step": 23540 }, { "grad_norm": 0.49672406911849976, "learning_rate": 9.088465489783131e-05, "loss": 0.0481, "step": 23550 }, { "grad_norm": 0.4740546941757202, "learning_rate": 9.087513439057068e-05, "loss": 0.0676, "step": 23560 }, { "grad_norm": 0.39316436648368835, "learning_rate": 9.08656094132668e-05, "loss": 0.0647, "step": 23570 }, { "grad_norm": 0.33772608637809753, "learning_rate": 9.085607996696134e-05, "loss": 0.0648, "step": 23580 }, { "grad_norm": 0.5363231301307678, "learning_rate": 9.084654605269639e-05, "loss": 0.0697, "step": 23590 }, { "grad_norm": 0.49399206042289734, "learning_rate": 9.083700767151457e-05, "loss": 0.0631, "step": 23600 }, { "grad_norm": 0.3127589225769043, "learning_rate": 9.082746482445898e-05, "loss": 0.0522, "step": 23610 }, { "grad_norm": 0.3781695067882538, "learning_rate": 9.081791751257325e-05, "loss": 0.0576, "step": 23620 }, { "grad_norm": 0.41436946392059326, "learning_rate": 9.080836573690142e-05, "loss": 0.0579, "step": 23630 }, { "grad_norm": 0.5415264368057251, "learning_rate": 9.079880949848805e-05, "loss": 0.0719, "step": 23640 }, { "grad_norm": 0.7092150449752808, "learning_rate": 9.078924879837822e-05, "loss": 0.0546, "step": 23650 }, { "grad_norm": 1.0000646114349365, "learning_rate": 9.077968363761747e-05, "loss": 0.0567, "step": 23660 }, { "grad_norm": 0.5468481779098511, "learning_rate": 9.077011401725182e-05, "loss": 0.0571, "step": 23670 }, { "grad_norm": 0.5335949659347534, "learning_rate": 9.07605399383278e-05, "loss": 0.0659, "step": 23680 }, { "grad_norm": 0.7227996587753296, "learning_rate": 9.075096140189243e-05, "loss": 0.0546, "step": 23690 }, { "grad_norm": 0.5376784205436707, "learning_rate": 9.074137840899318e-05, "loss": 0.0613, "step": 23700 }, { "grad_norm": 0.43379467725753784, "learning_rate": 9.073179096067804e-05, "loss": 0.0542, "step": 23710 }, { "grad_norm": 0.37219467759132385, "learning_rate": 9.072219905799549e-05, "loss": 0.0562, "step": 23720 }, { "grad_norm": 0.5027692317962646, "learning_rate": 9.071260270199447e-05, "loss": 0.0717, "step": 23730 }, { "grad_norm": 0.4913193881511688, "learning_rate": 9.070300189372441e-05, "loss": 0.0566, "step": 23740 }, { "grad_norm": 0.660525918006897, "learning_rate": 9.069339663423528e-05, "loss": 0.0521, "step": 23750 }, { "grad_norm": 0.44499674439430237, "learning_rate": 9.068378692457747e-05, "loss": 0.0606, "step": 23760 }, { "grad_norm": 0.3395563066005707, "learning_rate": 9.067417276580189e-05, "loss": 0.052, "step": 23770 }, { "grad_norm": 0.4688517153263092, "learning_rate": 9.066455415895993e-05, "loss": 0.0489, "step": 23780 }, { "grad_norm": 0.533699631690979, "learning_rate": 9.065493110510346e-05, "loss": 0.054, "step": 23790 }, { "grad_norm": 0.44988152384757996, "learning_rate": 9.064530360528484e-05, "loss": 0.0548, "step": 23800 }, { "grad_norm": 0.4133418798446655, "learning_rate": 9.063567166055695e-05, "loss": 0.0472, "step": 23810 }, { "grad_norm": 0.5997833609580994, "learning_rate": 9.062603527197308e-05, "loss": 0.0554, "step": 23820 }, { "grad_norm": 0.3905545473098755, "learning_rate": 9.06163944405871e-05, "loss": 0.0549, "step": 23830 }, { "grad_norm": 0.42916929721832275, "learning_rate": 9.060674916745327e-05, "loss": 0.0669, "step": 23840 }, { "grad_norm": 0.3447053134441376, "learning_rate": 9.05970994536264e-05, "loss": 0.0563, "step": 23850 }, { "grad_norm": 0.6158588528633118, "learning_rate": 9.05874453001618e-05, "loss": 0.0525, "step": 23860 }, { "grad_norm": 0.5748797059059143, "learning_rate": 9.057778670811517e-05, "loss": 0.0607, "step": 23870 }, { "grad_norm": 0.6084691286087036, "learning_rate": 9.056812367854281e-05, "loss": 0.0536, "step": 23880 }, { "grad_norm": 0.406383216381073, "learning_rate": 9.055845621250143e-05, "loss": 0.0535, "step": 23890 }, { "grad_norm": 0.44111648201942444, "learning_rate": 9.054878431104825e-05, "loss": 0.0687, "step": 23900 }, { "grad_norm": 0.6221328973770142, "learning_rate": 9.0539107975241e-05, "loss": 0.0702, "step": 23910 }, { "grad_norm": 0.5552685260772705, "learning_rate": 9.052942720613784e-05, "loss": 0.0543, "step": 23920 }, { "grad_norm": 0.40090978145599365, "learning_rate": 9.051974200479745e-05, "loss": 0.0556, "step": 23930 }, { "grad_norm": 0.6521667242050171, "learning_rate": 9.051005237227901e-05, "loss": 0.0499, "step": 23940 }, { "grad_norm": 0.4503183960914612, "learning_rate": 9.050035830964215e-05, "loss": 0.0592, "step": 23950 }, { "grad_norm": 0.5305981040000916, "learning_rate": 9.049065981794698e-05, "loss": 0.0726, "step": 23960 }, { "grad_norm": 0.8119130730628967, "learning_rate": 9.048095689825414e-05, "loss": 0.0582, "step": 23970 }, { "grad_norm": 0.4652225375175476, "learning_rate": 9.047124955162472e-05, "loss": 0.0531, "step": 23980 }, { "grad_norm": 0.4877890944480896, "learning_rate": 9.046153777912028e-05, "loss": 0.0605, "step": 23990 }, { "grad_norm": 0.5310857892036438, "learning_rate": 9.045182158180292e-05, "loss": 0.0556, "step": 24000 }, { "grad_norm": 0.4384762942790985, "learning_rate": 9.044210096073516e-05, "loss": 0.0454, "step": 24010 }, { "grad_norm": 0.3576744496822357, "learning_rate": 9.043237591698004e-05, "loss": 0.0435, "step": 24020 }, { "grad_norm": 0.39000219106674194, "learning_rate": 9.04226464516011e-05, "loss": 0.0543, "step": 24030 }, { "grad_norm": 0.4869169592857361, "learning_rate": 9.041291256566229e-05, "loss": 0.0716, "step": 24040 }, { "grad_norm": 0.5319550037384033, "learning_rate": 9.040317426022814e-05, "loss": 0.0541, "step": 24050 }, { "grad_norm": 0.5133241415023804, "learning_rate": 9.03934315363636e-05, "loss": 0.0694, "step": 24060 }, { "grad_norm": 0.5980368852615356, "learning_rate": 9.038368439513409e-05, "loss": 0.0659, "step": 24070 }, { "grad_norm": 0.5230408906936646, "learning_rate": 9.03739328376056e-05, "loss": 0.069, "step": 24080 }, { "grad_norm": 0.4861423373222351, "learning_rate": 9.036417686484451e-05, "loss": 0.0596, "step": 24090 }, { "grad_norm": 0.5290957093238831, "learning_rate": 9.035441647791773e-05, "loss": 0.0565, "step": 24100 }, { "grad_norm": 0.6373981833457947, "learning_rate": 9.034465167789263e-05, "loss": 0.0551, "step": 24110 }, { "grad_norm": 0.5036478042602539, "learning_rate": 9.033488246583706e-05, "loss": 0.0608, "step": 24120 }, { "grad_norm": 0.42754825949668884, "learning_rate": 9.032510884281941e-05, "loss": 0.0609, "step": 24130 }, { "grad_norm": 0.3308945596218109, "learning_rate": 9.031533080990848e-05, "loss": 0.0554, "step": 24140 }, { "grad_norm": 0.5330603122711182, "learning_rate": 9.030554836817358e-05, "loss": 0.0655, "step": 24150 }, { "grad_norm": 0.5890163779258728, "learning_rate": 9.029576151868451e-05, "loss": 0.0589, "step": 24160 }, { "grad_norm": 0.5377951860427856, "learning_rate": 9.028597026251155e-05, "loss": 0.0528, "step": 24170 }, { "grad_norm": 0.43254226446151733, "learning_rate": 9.027617460072547e-05, "loss": 0.0571, "step": 24180 }, { "grad_norm": 0.4240052402019501, "learning_rate": 9.026637453439745e-05, "loss": 0.0599, "step": 24190 }, { "grad_norm": 0.4022323489189148, "learning_rate": 9.025657006459927e-05, "loss": 0.0576, "step": 24200 }, { "grad_norm": 0.8431634902954102, "learning_rate": 9.024676119240311e-05, "loss": 0.0502, "step": 24210 }, { "grad_norm": 0.4695717692375183, "learning_rate": 9.023694791888166e-05, "loss": 0.0569, "step": 24220 }, { "grad_norm": 0.4731932282447815, "learning_rate": 9.022713024510808e-05, "loss": 0.0645, "step": 24230 }, { "grad_norm": 0.45742157101631165, "learning_rate": 9.021730817215601e-05, "loss": 0.0537, "step": 24240 }, { "grad_norm": 0.505864143371582, "learning_rate": 9.02074817010996e-05, "loss": 0.0638, "step": 24250 }, { "grad_norm": 0.6183704137802124, "learning_rate": 9.019765083301342e-05, "loss": 0.0496, "step": 24260 }, { "grad_norm": 0.8737030625343323, "learning_rate": 9.01878155689726e-05, "loss": 0.0687, "step": 24270 }, { "grad_norm": 0.45545628666877747, "learning_rate": 9.017797591005268e-05, "loss": 0.0513, "step": 24280 }, { "grad_norm": 0.40169432759284973, "learning_rate": 9.016813185732972e-05, "loss": 0.0674, "step": 24290 }, { "grad_norm": 0.5842092037200928, "learning_rate": 9.015828341188027e-05, "loss": 0.0682, "step": 24300 }, { "grad_norm": 0.7293090224266052, "learning_rate": 9.01484305747813e-05, "loss": 0.06, "step": 24310 }, { "grad_norm": 0.5104977488517761, "learning_rate": 9.013857334711033e-05, "loss": 0.0671, "step": 24320 }, { "grad_norm": 1.5046944618225098, "learning_rate": 9.012871172994534e-05, "loss": 0.0562, "step": 24330 }, { "grad_norm": 0.841035783290863, "learning_rate": 9.011884572436476e-05, "loss": 0.0629, "step": 24340 }, { "grad_norm": 0.5904817581176758, "learning_rate": 9.010897533144754e-05, "loss": 0.0627, "step": 24350 }, { "grad_norm": 0.5664510130882263, "learning_rate": 9.009910055227306e-05, "loss": 0.0622, "step": 24360 }, { "grad_norm": 0.43207499384880066, "learning_rate": 9.008922138792124e-05, "loss": 0.0534, "step": 24370 }, { "grad_norm": 0.8731858730316162, "learning_rate": 9.007933783947244e-05, "loss": 0.0553, "step": 24380 }, { "grad_norm": 0.4352192282676697, "learning_rate": 9.006944990800752e-05, "loss": 0.0484, "step": 24390 }, { "grad_norm": 0.4817826449871063, "learning_rate": 9.005955759460779e-05, "loss": 0.0641, "step": 24400 }, { "grad_norm": 0.48990461230278015, "learning_rate": 9.004966090035508e-05, "loss": 0.0508, "step": 24410 }, { "grad_norm": 0.37799179553985596, "learning_rate": 9.003975982633166e-05, "loss": 0.0534, "step": 24420 }, { "grad_norm": 0.5328590869903564, "learning_rate": 9.00298543736203e-05, "loss": 0.0623, "step": 24430 }, { "grad_norm": 0.4073645770549774, "learning_rate": 9.001994454330427e-05, "loss": 0.0477, "step": 24440 }, { "grad_norm": 0.5131226181983948, "learning_rate": 9.001003033646727e-05, "loss": 0.0602, "step": 24450 }, { "grad_norm": 0.4332994818687439, "learning_rate": 9.00001117541935e-05, "loss": 0.0518, "step": 24460 }, { "grad_norm": 0.39696645736694336, "learning_rate": 8.999018879756764e-05, "loss": 0.0601, "step": 24470 }, { "grad_norm": 0.5126440525054932, "learning_rate": 8.998026146767487e-05, "loss": 0.047, "step": 24480 }, { "grad_norm": 0.8396978974342346, "learning_rate": 8.99703297656008e-05, "loss": 0.0513, "step": 24490 }, { "grad_norm": 0.4336380064487457, "learning_rate": 8.996039369243156e-05, "loss": 0.0592, "step": 24500 }, { "grad_norm": 0.5401436686515808, "learning_rate": 8.995045324925378e-05, "loss": 0.0528, "step": 24510 }, { "grad_norm": 0.3622351586818695, "learning_rate": 8.994050843715448e-05, "loss": 0.0655, "step": 24520 }, { "grad_norm": 0.4301915168762207, "learning_rate": 8.993055925722121e-05, "loss": 0.0616, "step": 24530 }, { "grad_norm": 0.5483637452125549, "learning_rate": 8.992060571054202e-05, "loss": 0.0583, "step": 24540 }, { "grad_norm": 0.3956848680973053, "learning_rate": 8.991064779820542e-05, "loss": 0.0519, "step": 24550 }, { "grad_norm": 0.4342867434024811, "learning_rate": 8.990068552130036e-05, "loss": 0.0493, "step": 24560 }, { "grad_norm": 0.3412664234638214, "learning_rate": 8.989071888091634e-05, "loss": 0.0526, "step": 24570 }, { "grad_norm": 0.4655981659889221, "learning_rate": 8.988074787814329e-05, "loss": 0.0584, "step": 24580 }, { "grad_norm": 0.4316984713077545, "learning_rate": 8.987077251407158e-05, "loss": 0.0479, "step": 24590 }, { "grad_norm": 0.41967591643333435, "learning_rate": 8.986079278979216e-05, "loss": 0.0522, "step": 24600 }, { "grad_norm": 0.4418926239013672, "learning_rate": 8.985080870639635e-05, "loss": 0.067, "step": 24610 }, { "grad_norm": 0.4447384476661682, "learning_rate": 8.984082026497603e-05, "loss": 0.0497, "step": 24620 }, { "grad_norm": 0.9443802237510681, "learning_rate": 8.98308274666235e-05, "loss": 0.072, "step": 24630 }, { "grad_norm": 0.47959595918655396, "learning_rate": 8.982083031243155e-05, "loss": 0.058, "step": 24640 }, { "grad_norm": 0.5608507990837097, "learning_rate": 8.98108288034935e-05, "loss": 0.0551, "step": 24650 }, { "grad_norm": 0.44838669896125793, "learning_rate": 8.980082294090305e-05, "loss": 0.0619, "step": 24660 }, { "grad_norm": 0.4787115156650543, "learning_rate": 8.979081272575443e-05, "loss": 0.0609, "step": 24670 }, { "grad_norm": 0.728425920009613, "learning_rate": 8.978079815914236e-05, "loss": 0.0625, "step": 24680 }, { "grad_norm": 0.5140852928161621, "learning_rate": 8.977077924216202e-05, "loss": 0.0629, "step": 24690 }, { "grad_norm": 0.6180113554000854, "learning_rate": 8.976075597590905e-05, "loss": 0.0711, "step": 24700 }, { "grad_norm": 0.4143078029155731, "learning_rate": 8.975072836147958e-05, "loss": 0.0489, "step": 24710 }, { "grad_norm": 0.647502601146698, "learning_rate": 8.974069639997025e-05, "loss": 0.0584, "step": 24720 }, { "grad_norm": 0.42404600977897644, "learning_rate": 8.973066009247808e-05, "loss": 0.066, "step": 24730 }, { "grad_norm": 0.3541921377182007, "learning_rate": 8.972061944010066e-05, "loss": 0.0577, "step": 24740 }, { "grad_norm": 0.4819613993167877, "learning_rate": 8.971057444393603e-05, "loss": 0.0546, "step": 24750 }, { "grad_norm": 0.3860936760902405, "learning_rate": 8.970052510508268e-05, "loss": 0.058, "step": 24760 }, { "grad_norm": 0.48202070593833923, "learning_rate": 8.969047142463959e-05, "loss": 0.0595, "step": 24770 }, { "grad_norm": 0.41037559509277344, "learning_rate": 8.968041340370621e-05, "loss": 0.0528, "step": 24780 }, { "grad_norm": 0.275651216506958, "learning_rate": 8.96703510433825e-05, "loss": 0.0555, "step": 24790 }, { "grad_norm": 0.5989981889724731, "learning_rate": 8.966028434476883e-05, "loss": 0.047, "step": 24800 }, { "grad_norm": 0.47526219487190247, "learning_rate": 8.96502133089661e-05, "loss": 0.0517, "step": 24810 }, { "grad_norm": 0.47834300994873047, "learning_rate": 8.964013793707564e-05, "loss": 0.0496, "step": 24820 }, { "grad_norm": 0.5607186555862427, "learning_rate": 8.963005823019932e-05, "loss": 0.0662, "step": 24830 }, { "grad_norm": 0.5608795881271362, "learning_rate": 8.961997418943939e-05, "loss": 0.0526, "step": 24840 }, { "grad_norm": 0.5382494926452637, "learning_rate": 8.960988581589865e-05, "loss": 0.0588, "step": 24850 }, { "grad_norm": 0.4906909465789795, "learning_rate": 8.959979311068037e-05, "loss": 0.0468, "step": 24860 }, { "grad_norm": 0.5159239768981934, "learning_rate": 8.958969607488823e-05, "loss": 0.0714, "step": 24870 }, { "grad_norm": 0.6068128347396851, "learning_rate": 8.957959470962647e-05, "loss": 0.0482, "step": 24880 }, { "grad_norm": 0.3880009055137634, "learning_rate": 8.956948901599971e-05, "loss": 0.052, "step": 24890 }, { "grad_norm": 0.3743504285812378, "learning_rate": 8.955937899511315e-05, "loss": 0.0616, "step": 24900 }, { "grad_norm": 0.556198000907898, "learning_rate": 8.954926464807238e-05, "loss": 0.0485, "step": 24910 }, { "grad_norm": 0.6225593090057373, "learning_rate": 8.953914597598347e-05, "loss": 0.0595, "step": 24920 }, { "grad_norm": 0.3561125695705414, "learning_rate": 8.952902297995303e-05, "loss": 0.0536, "step": 24930 }, { "grad_norm": 0.30819737911224365, "learning_rate": 8.951889566108804e-05, "loss": 0.0599, "step": 24940 }, { "grad_norm": 0.5718841552734375, "learning_rate": 8.950876402049606e-05, "loss": 0.0637, "step": 24950 }, { "grad_norm": 0.5133184790611267, "learning_rate": 8.949862805928504e-05, "loss": 0.0505, "step": 24960 }, { "grad_norm": 0.4168543815612793, "learning_rate": 8.948848777856343e-05, "loss": 0.0613, "step": 24970 }, { "grad_norm": 0.4721180200576782, "learning_rate": 8.947834317944017e-05, "loss": 0.052, "step": 24980 }, { "grad_norm": 0.4455648362636566, "learning_rate": 8.946819426302466e-05, "loss": 0.068, "step": 24990 }, { "grad_norm": 0.5525137186050415, "learning_rate": 8.945804103042676e-05, "loss": 0.0531, "step": 25000 }, { "grad_norm": 1.2066569328308105, "learning_rate": 8.944788348275681e-05, "loss": 0.0522, "step": 25010 }, { "grad_norm": 0.5065156817436218, "learning_rate": 8.943772162112565e-05, "loss": 0.0544, "step": 25020 }, { "grad_norm": 0.4674818515777588, "learning_rate": 8.942755544664454e-05, "loss": 0.0553, "step": 25030 }, { "grad_norm": 0.6462596654891968, "learning_rate": 8.941738496042525e-05, "loss": 0.0494, "step": 25040 }, { "grad_norm": 0.34755417704582214, "learning_rate": 8.940721016357999e-05, "loss": 0.0555, "step": 25050 }, { "grad_norm": 0.5011102557182312, "learning_rate": 8.939703105722148e-05, "loss": 0.0634, "step": 25060 }, { "grad_norm": 0.44811153411865234, "learning_rate": 8.93868476424629e-05, "loss": 0.0485, "step": 25070 }, { "grad_norm": 0.5739628672599792, "learning_rate": 8.937665992041786e-05, "loss": 0.06, "step": 25080 }, { "grad_norm": 0.5826124548912048, "learning_rate": 8.93664678922005e-05, "loss": 0.0572, "step": 25090 }, { "grad_norm": 0.41592156887054443, "learning_rate": 8.93562715589254e-05, "loss": 0.0473, "step": 25100 }, { "grad_norm": 0.4191981852054596, "learning_rate": 8.934607092170762e-05, "loss": 0.054, "step": 25110 }, { "grad_norm": 0.5371633172035217, "learning_rate": 8.933586598166266e-05, "loss": 0.0595, "step": 25120 }, { "grad_norm": 0.5346769690513611, "learning_rate": 8.932565673990655e-05, "loss": 0.0694, "step": 25130 }, { "grad_norm": 0.3733953535556793, "learning_rate": 8.931544319755574e-05, "loss": 0.051, "step": 25140 }, { "grad_norm": 0.47367063164711, "learning_rate": 8.930522535572718e-05, "loss": 0.0586, "step": 25150 }, { "grad_norm": 0.49690571427345276, "learning_rate": 8.929500321553826e-05, "loss": 0.046, "step": 25160 }, { "grad_norm": 0.4814334511756897, "learning_rate": 8.928477677810686e-05, "loss": 0.0605, "step": 25170 }, { "grad_norm": 0.42536798119544983, "learning_rate": 8.927454604455137e-05, "loss": 0.0531, "step": 25180 }, { "grad_norm": 0.2930739223957062, "learning_rate": 8.926431101599053e-05, "loss": 0.0572, "step": 25190 }, { "grad_norm": 0.7167423367500305, "learning_rate": 8.925407169354369e-05, "loss": 0.0522, "step": 25200 }, { "grad_norm": 0.45256876945495605, "learning_rate": 8.92438280783306e-05, "loss": 0.0676, "step": 25210 }, { "grad_norm": 0.4381050765514374, "learning_rate": 8.923358017147146e-05, "loss": 0.0667, "step": 25220 }, { "grad_norm": 0.49449455738067627, "learning_rate": 8.922332797408697e-05, "loss": 0.062, "step": 25230 }, { "grad_norm": 0.35275089740753174, "learning_rate": 8.921307148729831e-05, "loss": 0.0498, "step": 25240 }, { "grad_norm": 0.45072996616363525, "learning_rate": 8.920281071222712e-05, "loss": 0.044, "step": 25250 }, { "grad_norm": 0.41664132475852966, "learning_rate": 8.919254564999548e-05, "loss": 0.0598, "step": 25260 }, { "grad_norm": 0.4015515446662903, "learning_rate": 8.918227630172598e-05, "loss": 0.0586, "step": 25270 }, { "grad_norm": 0.6558960676193237, "learning_rate": 8.917200266854165e-05, "loss": 0.0557, "step": 25280 }, { "grad_norm": 0.5310849547386169, "learning_rate": 8.9161724751566e-05, "loss": 0.0542, "step": 25290 }, { "grad_norm": 0.43658414483070374, "learning_rate": 8.915144255192302e-05, "loss": 0.0672, "step": 25300 }, { "grad_norm": 0.7468444108963013, "learning_rate": 8.914115607073714e-05, "loss": 0.0617, "step": 25310 }, { "grad_norm": 0.4509240984916687, "learning_rate": 8.913086530913327e-05, "loss": 0.0511, "step": 25320 }, { "grad_norm": 0.43368807435035706, "learning_rate": 8.912057026823681e-05, "loss": 0.0674, "step": 25330 }, { "grad_norm": 0.39633530378341675, "learning_rate": 8.91102709491736e-05, "loss": 0.0608, "step": 25340 }, { "grad_norm": 0.5489696860313416, "learning_rate": 8.909996735306996e-05, "loss": 0.0566, "step": 25350 }, { "grad_norm": 0.5831400156021118, "learning_rate": 8.908965948105268e-05, "loss": 0.0589, "step": 25360 }, { "grad_norm": 0.5981163382530212, "learning_rate": 8.907934733424901e-05, "loss": 0.061, "step": 25370 }, { "grad_norm": 0.2566716969013214, "learning_rate": 8.906903091378666e-05, "loss": 0.0564, "step": 25380 }, { "grad_norm": 0.3231639564037323, "learning_rate": 8.905871022079384e-05, "loss": 0.0623, "step": 25390 }, { "grad_norm": 0.3721787929534912, "learning_rate": 8.90483852563992e-05, "loss": 0.059, "step": 25400 }, { "grad_norm": 0.4686378538608551, "learning_rate": 8.903805602173185e-05, "loss": 0.0504, "step": 25410 }, { "grad_norm": 0.36906537413597107, "learning_rate": 8.902772251792137e-05, "loss": 0.0484, "step": 25420 }, { "grad_norm": 0.5989737510681152, "learning_rate": 8.901738474609786e-05, "loss": 0.06, "step": 25430 }, { "grad_norm": 0.3866930902004242, "learning_rate": 8.900704270739179e-05, "loss": 0.0675, "step": 25440 }, { "grad_norm": 0.4283473491668701, "learning_rate": 8.89966964029342e-05, "loss": 0.0787, "step": 25450 }, { "grad_norm": 0.4578416645526886, "learning_rate": 8.898634583385652e-05, "loss": 0.0514, "step": 25460 }, { "grad_norm": 0.4591713845729828, "learning_rate": 8.897599100129065e-05, "loss": 0.062, "step": 25470 }, { "grad_norm": 0.42147040367126465, "learning_rate": 8.896563190636903e-05, "loss": 0.05, "step": 25480 }, { "grad_norm": 0.7130882143974304, "learning_rate": 8.895526855022448e-05, "loss": 0.0526, "step": 25490 }, { "grad_norm": 0.5974505543708801, "learning_rate": 8.894490093399033e-05, "loss": 0.0613, "step": 25500 }, { "grad_norm": 0.32617059350013733, "learning_rate": 8.893452905880035e-05, "loss": 0.0596, "step": 25510 }, { "grad_norm": 0.5335826277732849, "learning_rate": 8.892415292578883e-05, "loss": 0.0587, "step": 25520 }, { "grad_norm": 0.35001340508461, "learning_rate": 8.891377253609046e-05, "loss": 0.0493, "step": 25530 }, { "grad_norm": 0.4911864101886749, "learning_rate": 8.890338789084043e-05, "loss": 0.0534, "step": 25540 }, { "grad_norm": 0.6031152606010437, "learning_rate": 8.88929989911744e-05, "loss": 0.0494, "step": 25550 }, { "grad_norm": 0.6019364595413208, "learning_rate": 8.888260583822847e-05, "loss": 0.0582, "step": 25560 }, { "grad_norm": 0.43687617778778076, "learning_rate": 8.887220843313921e-05, "loss": 0.0608, "step": 25570 }, { "grad_norm": 0.5713849663734436, "learning_rate": 8.88618067770437e-05, "loss": 0.0512, "step": 25580 }, { "grad_norm": 0.3123026490211487, "learning_rate": 8.885140087107942e-05, "loss": 0.0501, "step": 25590 }, { "grad_norm": 0.7404258847236633, "learning_rate": 8.884099071638436e-05, "loss": 0.0596, "step": 25600 }, { "grad_norm": 0.39191049337387085, "learning_rate": 8.883057631409695e-05, "loss": 0.0627, "step": 25610 }, { "grad_norm": 0.4782814085483551, "learning_rate": 8.882015766535608e-05, "loss": 0.0549, "step": 25620 }, { "grad_norm": 0.542247474193573, "learning_rate": 8.880973477130115e-05, "loss": 0.0541, "step": 25630 }, { "grad_norm": 0.5553238987922668, "learning_rate": 8.879930763307197e-05, "loss": 0.0438, "step": 25640 }, { "grad_norm": 0.3943326771259308, "learning_rate": 8.878887625180884e-05, "loss": 0.0601, "step": 25650 }, { "grad_norm": 0.37497493624687195, "learning_rate": 8.877844062865253e-05, "loss": 0.0406, "step": 25660 }, { "grad_norm": 0.6379011273384094, "learning_rate": 8.876800076474424e-05, "loss": 0.0539, "step": 25670 }, { "grad_norm": 0.378713995218277, "learning_rate": 8.875755666122568e-05, "loss": 0.0459, "step": 25680 }, { "grad_norm": 0.3960517644882202, "learning_rate": 8.8747108319239e-05, "loss": 0.0595, "step": 25690 }, { "grad_norm": 0.4542575776576996, "learning_rate": 8.87366557399268e-05, "loss": 0.053, "step": 25700 }, { "grad_norm": 0.6318703889846802, "learning_rate": 8.872619892443217e-05, "loss": 0.0602, "step": 25710 }, { "grad_norm": 0.4477789103984833, "learning_rate": 8.871573787389865e-05, "loss": 0.0513, "step": 25720 }, { "grad_norm": 0.5971600413322449, "learning_rate": 8.870527258947024e-05, "loss": 0.0642, "step": 25730 }, { "grad_norm": 0.48592209815979004, "learning_rate": 8.869480307229143e-05, "loss": 0.0474, "step": 25740 }, { "grad_norm": 0.5482323169708252, "learning_rate": 8.868432932350712e-05, "loss": 0.0508, "step": 25750 }, { "grad_norm": 0.4011293053627014, "learning_rate": 8.867385134426272e-05, "loss": 0.0578, "step": 25760 }, { "grad_norm": 0.3328964412212372, "learning_rate": 8.866336913570407e-05, "loss": 0.0455, "step": 25770 }, { "grad_norm": 0.32184967398643494, "learning_rate": 8.865288269897751e-05, "loss": 0.0413, "step": 25780 }, { "grad_norm": 0.48382481932640076, "learning_rate": 8.864239203522981e-05, "loss": 0.0511, "step": 25790 }, { "grad_norm": 0.3647262454032898, "learning_rate": 8.863189714560822e-05, "loss": 0.0573, "step": 25800 }, { "grad_norm": 0.39900776743888855, "learning_rate": 8.862139803126043e-05, "loss": 0.0481, "step": 25810 }, { "grad_norm": 0.48286330699920654, "learning_rate": 8.861089469333463e-05, "loss": 0.0563, "step": 25820 }, { "grad_norm": 0.43705201148986816, "learning_rate": 8.860038713297944e-05, "loss": 0.0395, "step": 25830 }, { "grad_norm": 0.5135835409164429, "learning_rate": 8.858987535134394e-05, "loss": 0.0605, "step": 25840 }, { "grad_norm": 0.45088833570480347, "learning_rate": 8.857935934957769e-05, "loss": 0.0571, "step": 25850 }, { "grad_norm": 0.4321674108505249, "learning_rate": 8.856883912883071e-05, "loss": 0.0452, "step": 25860 }, { "grad_norm": 0.379427045583725, "learning_rate": 8.855831469025346e-05, "loss": 0.0472, "step": 25870 }, { "grad_norm": 0.7411864399909973, "learning_rate": 8.854778603499689e-05, "loss": 0.0756, "step": 25880 }, { "grad_norm": 0.6773577332496643, "learning_rate": 8.85372531642124e-05, "loss": 0.0543, "step": 25890 }, { "grad_norm": 0.4224303662776947, "learning_rate": 8.852671607905185e-05, "loss": 0.045, "step": 25900 }, { "grad_norm": 0.34241995215415955, "learning_rate": 8.851617478066754e-05, "loss": 0.0635, "step": 25910 }, { "grad_norm": 0.4987640380859375, "learning_rate": 8.850562927021227e-05, "loss": 0.0605, "step": 25920 }, { "grad_norm": 0.360404908657074, "learning_rate": 8.849507954883928e-05, "loss": 0.05, "step": 25930 }, { "grad_norm": 0.37272605299949646, "learning_rate": 8.848452561770226e-05, "loss": 0.0491, "step": 25940 }, { "grad_norm": 0.4384654760360718, "learning_rate": 8.847396747795538e-05, "loss": 0.0544, "step": 25950 }, { "grad_norm": 0.44403499364852905, "learning_rate": 8.846340513075327e-05, "loss": 0.0485, "step": 25960 }, { "grad_norm": 0.649899423122406, "learning_rate": 8.845283857725099e-05, "loss": 0.0575, "step": 25970 }, { "grad_norm": 0.5132439732551575, "learning_rate": 8.844226781860409e-05, "loss": 0.0403, "step": 25980 }, { "grad_norm": 0.4090859889984131, "learning_rate": 8.84316928559686e-05, "loss": 0.0513, "step": 25990 }, { "grad_norm": 0.8041571378707886, "learning_rate": 8.842111369050094e-05, "loss": 0.0734, "step": 26000 }, { "grad_norm": 0.4606921076774597, "learning_rate": 8.841053032335808e-05, "loss": 0.0561, "step": 26010 }, { "grad_norm": 0.4297226071357727, "learning_rate": 8.839994275569735e-05, "loss": 0.0591, "step": 26020 }, { "grad_norm": 0.7251154780387878, "learning_rate": 8.838935098867662e-05, "loss": 0.0599, "step": 26030 }, { "grad_norm": 0.6076520085334778, "learning_rate": 8.837875502345418e-05, "loss": 0.0639, "step": 26040 }, { "grad_norm": 0.538133442401886, "learning_rate": 8.83681548611888e-05, "loss": 0.0568, "step": 26050 }, { "grad_norm": 0.37339603900909424, "learning_rate": 8.835755050303969e-05, "loss": 0.055, "step": 26060 }, { "grad_norm": 0.5165026187896729, "learning_rate": 8.834694195016653e-05, "loss": 0.0581, "step": 26070 }, { "grad_norm": 0.6184883117675781, "learning_rate": 8.833632920372942e-05, "loss": 0.0707, "step": 26080 }, { "grad_norm": 0.48280492424964905, "learning_rate": 8.832571226488903e-05, "loss": 0.0586, "step": 26090 }, { "grad_norm": 0.5342346429824829, "learning_rate": 8.831509113480634e-05, "loss": 0.0524, "step": 26100 }, { "grad_norm": 0.695767343044281, "learning_rate": 8.83044658146429e-05, "loss": 0.0622, "step": 26110 }, { "grad_norm": 0.34808096289634705, "learning_rate": 8.829383630556067e-05, "loss": 0.0574, "step": 26120 }, { "grad_norm": 1.2916747331619263, "learning_rate": 8.828320260872207e-05, "loss": 0.0645, "step": 26130 }, { "grad_norm": 0.4443622827529907, "learning_rate": 8.827256472529e-05, "loss": 0.0536, "step": 26140 }, { "grad_norm": 0.38067036867141724, "learning_rate": 8.826192265642778e-05, "loss": 0.0541, "step": 26150 }, { "grad_norm": 0.33627960085868835, "learning_rate": 8.825127640329923e-05, "loss": 0.0489, "step": 26160 }, { "grad_norm": 0.5353906154632568, "learning_rate": 8.824062596706861e-05, "loss": 0.0523, "step": 26170 }, { "grad_norm": 0.7591243386268616, "learning_rate": 8.822997134890062e-05, "loss": 0.0583, "step": 26180 }, { "grad_norm": 0.6362511515617371, "learning_rate": 8.821931254996044e-05, "loss": 0.0524, "step": 26190 }, { "grad_norm": 0.409416526556015, "learning_rate": 8.82086495714137e-05, "loss": 0.0666, "step": 26200 }, { "grad_norm": 0.5636258125305176, "learning_rate": 8.81979824144265e-05, "loss": 0.065, "step": 26210 }, { "grad_norm": 0.47669801115989685, "learning_rate": 8.818731108016536e-05, "loss": 0.0583, "step": 26220 }, { "grad_norm": 0.4028814733028412, "learning_rate": 8.81766355697973e-05, "loss": 0.051, "step": 26230 }, { "grad_norm": 0.4524916410446167, "learning_rate": 8.816595588448977e-05, "loss": 0.0538, "step": 26240 }, { "grad_norm": 0.8110694289207458, "learning_rate": 8.81552720254107e-05, "loss": 0.0648, "step": 26250 }, { "grad_norm": 0.5179081559181213, "learning_rate": 8.814458399372842e-05, "loss": 0.0685, "step": 26260 }, { "grad_norm": 0.5975717902183533, "learning_rate": 8.813389179061181e-05, "loss": 0.0453, "step": 26270 }, { "grad_norm": 0.5344196557998657, "learning_rate": 8.812319541723012e-05, "loss": 0.0689, "step": 26280 }, { "grad_norm": 0.4016927182674408, "learning_rate": 8.811249487475309e-05, "loss": 0.0648, "step": 26290 }, { "grad_norm": 0.46044740080833435, "learning_rate": 8.810179016435092e-05, "loss": 0.0558, "step": 26300 }, { "grad_norm": 0.7554464936256409, "learning_rate": 8.809108128719428e-05, "loss": 0.0705, "step": 26310 }, { "grad_norm": 0.6320742964744568, "learning_rate": 8.808036824445424e-05, "loss": 0.0679, "step": 26320 }, { "grad_norm": 0.5795919895172119, "learning_rate": 8.806965103730238e-05, "loss": 0.0626, "step": 26330 }, { "grad_norm": 0.4631526470184326, "learning_rate": 8.805892966691074e-05, "loss": 0.0492, "step": 26340 }, { "grad_norm": 0.42220571637153625, "learning_rate": 8.804820413445175e-05, "loss": 0.0562, "step": 26350 }, { "grad_norm": 0.3720625042915344, "learning_rate": 8.803747444109837e-05, "loss": 0.0602, "step": 26360 }, { "grad_norm": 0.5364465117454529, "learning_rate": 8.802674058802399e-05, "loss": 0.0438, "step": 26370 }, { "grad_norm": 0.4161129295825958, "learning_rate": 8.801600257640241e-05, "loss": 0.0586, "step": 26380 }, { "grad_norm": 0.4842606782913208, "learning_rate": 8.800526040740795e-05, "loss": 0.0603, "step": 26390 }, { "grad_norm": 0.6128057241439819, "learning_rate": 8.799451408221535e-05, "loss": 0.0547, "step": 26400 }, { "grad_norm": 0.5349730849266052, "learning_rate": 8.798376360199982e-05, "loss": 0.0488, "step": 26410 }, { "grad_norm": 0.48097142577171326, "learning_rate": 8.797300896793701e-05, "loss": 0.0478, "step": 26420 }, { "grad_norm": 0.6132275462150574, "learning_rate": 8.796225018120302e-05, "loss": 0.0606, "step": 26430 }, { "grad_norm": 0.4259835183620453, "learning_rate": 8.795148724297444e-05, "loss": 0.0516, "step": 26440 }, { "grad_norm": 0.48423364758491516, "learning_rate": 8.794072015442825e-05, "loss": 0.0566, "step": 26450 }, { "grad_norm": 0.8644121289253235, "learning_rate": 8.792994891674198e-05, "loss": 0.0624, "step": 26460 }, { "grad_norm": 0.34144192934036255, "learning_rate": 8.79191735310935e-05, "loss": 0.0495, "step": 26470 }, { "grad_norm": 0.4898563325405121, "learning_rate": 8.790839399866122e-05, "loss": 0.0658, "step": 26480 }, { "grad_norm": 0.5280744433403015, "learning_rate": 8.789761032062397e-05, "loss": 0.0523, "step": 26490 }, { "grad_norm": 0.4956854283809662, "learning_rate": 8.788682249816103e-05, "loss": 0.0553, "step": 26500 }, { "grad_norm": 0.3032841682434082, "learning_rate": 8.787603053245215e-05, "loss": 0.071, "step": 26510 }, { "grad_norm": 0.6476224064826965, "learning_rate": 8.78652344246775e-05, "loss": 0.0581, "step": 26520 }, { "grad_norm": 0.38137733936309814, "learning_rate": 8.785443417601776e-05, "loss": 0.0462, "step": 26530 }, { "grad_norm": 0.5127217173576355, "learning_rate": 8.784362978765401e-05, "loss": 0.0468, "step": 26540 }, { "grad_norm": 0.613917887210846, "learning_rate": 8.783282126076779e-05, "loss": 0.0542, "step": 26550 }, { "grad_norm": 0.6014720797538757, "learning_rate": 8.782200859654112e-05, "loss": 0.0529, "step": 26560 }, { "grad_norm": 0.5438354015350342, "learning_rate": 8.781119179615646e-05, "loss": 0.0497, "step": 26570 }, { "grad_norm": 0.3823682963848114, "learning_rate": 8.780037086079674e-05, "loss": 0.0525, "step": 26580 }, { "grad_norm": 0.5063014626502991, "learning_rate": 8.778954579164527e-05, "loss": 0.0621, "step": 26590 }, { "grad_norm": 0.7034493684768677, "learning_rate": 8.777871658988588e-05, "loss": 0.0611, "step": 26600 }, { "grad_norm": 0.4062739610671997, "learning_rate": 8.776788325670285e-05, "loss": 0.0616, "step": 26610 }, { "grad_norm": 0.659925103187561, "learning_rate": 8.775704579328089e-05, "loss": 0.0589, "step": 26620 }, { "grad_norm": 0.5940634608268738, "learning_rate": 8.774620420080517e-05, "loss": 0.0576, "step": 26630 }, { "grad_norm": 0.3875425457954407, "learning_rate": 8.773535848046131e-05, "loss": 0.0486, "step": 26640 }, { "grad_norm": 0.5592520236968994, "learning_rate": 8.772450863343538e-05, "loss": 0.0516, "step": 26650 }, { "grad_norm": 0.48080670833587646, "learning_rate": 8.77136546609139e-05, "loss": 0.0611, "step": 26660 }, { "grad_norm": 0.6693307161331177, "learning_rate": 8.770279656408385e-05, "loss": 0.0458, "step": 26670 }, { "grad_norm": 0.4108430743217468, "learning_rate": 8.769193434413265e-05, "loss": 0.0493, "step": 26680 }, { "grad_norm": 0.5468411445617676, "learning_rate": 8.76810680022482e-05, "loss": 0.0551, "step": 26690 }, { "grad_norm": 0.5543123483657837, "learning_rate": 8.767019753961878e-05, "loss": 0.0673, "step": 26700 }, { "grad_norm": 0.6261656880378723, "learning_rate": 8.765932295743321e-05, "loss": 0.0533, "step": 26710 }, { "grad_norm": 0.43343785405158997, "learning_rate": 8.764844425688068e-05, "loss": 0.0562, "step": 26720 }, { "grad_norm": 0.36756396293640137, "learning_rate": 8.763756143915092e-05, "loss": 0.048, "step": 26730 }, { "grad_norm": 0.35756805539131165, "learning_rate": 8.7626674505434e-05, "loss": 0.0495, "step": 26740 }, { "grad_norm": 0.540833592414856, "learning_rate": 8.761578345692053e-05, "loss": 0.0555, "step": 26750 }, { "grad_norm": 0.4595375657081604, "learning_rate": 8.760488829480156e-05, "loss": 0.0494, "step": 26760 }, { "grad_norm": 0.38681378960609436, "learning_rate": 8.759398902026854e-05, "loss": 0.0408, "step": 26770 }, { "grad_norm": 0.5751475095748901, "learning_rate": 8.758308563451339e-05, "loss": 0.0606, "step": 26780 }, { "grad_norm": 0.4966032803058624, "learning_rate": 8.75721781387285e-05, "loss": 0.0603, "step": 26790 }, { "grad_norm": 0.47094640135765076, "learning_rate": 8.75612665341067e-05, "loss": 0.0494, "step": 26800 }, { "grad_norm": 0.4430443346500397, "learning_rate": 8.755035082184126e-05, "loss": 0.0541, "step": 26810 }, { "grad_norm": 0.5206425189971924, "learning_rate": 8.753943100312592e-05, "loss": 0.0604, "step": 26820 }, { "grad_norm": 0.3824484348297119, "learning_rate": 8.752850707915484e-05, "loss": 0.0506, "step": 26830 }, { "grad_norm": 0.4714037775993347, "learning_rate": 8.751757905112264e-05, "loss": 0.0636, "step": 26840 }, { "grad_norm": 0.4845499098300934, "learning_rate": 8.75066469202244e-05, "loss": 0.0524, "step": 26850 }, { "grad_norm": 0.4316638708114624, "learning_rate": 8.749571068765567e-05, "loss": 0.0465, "step": 26860 }, { "grad_norm": 0.4303172528743744, "learning_rate": 8.748477035461238e-05, "loss": 0.0457, "step": 26870 }, { "grad_norm": 0.4312204122543335, "learning_rate": 8.747382592229095e-05, "loss": 0.0539, "step": 26880 }, { "grad_norm": 0.3896985948085785, "learning_rate": 8.746287739188828e-05, "loss": 0.0611, "step": 26890 }, { "grad_norm": 0.4376102685928345, "learning_rate": 8.745192476460165e-05, "loss": 0.0478, "step": 26900 }, { "grad_norm": 0.5115715265274048, "learning_rate": 8.744096804162882e-05, "loss": 0.0642, "step": 26910 }, { "grad_norm": 0.5584380626678467, "learning_rate": 8.743000722416804e-05, "loss": 0.0565, "step": 26920 }, { "grad_norm": 0.5167731046676636, "learning_rate": 8.741904231341793e-05, "loss": 0.0539, "step": 26930 }, { "grad_norm": 0.6842197775840759, "learning_rate": 8.740807331057762e-05, "loss": 0.0619, "step": 26940 }, { "grad_norm": 0.5223310589790344, "learning_rate": 8.739710021684667e-05, "loss": 0.0533, "step": 26950 }, { "grad_norm": 0.33584070205688477, "learning_rate": 8.738612303342503e-05, "loss": 0.0496, "step": 26960 }, { "grad_norm": 0.4729961156845093, "learning_rate": 8.73751417615132e-05, "loss": 0.0472, "step": 26970 }, { "grad_norm": 0.3380998969078064, "learning_rate": 8.736415640231208e-05, "loss": 0.0406, "step": 26980 }, { "grad_norm": 0.3083130121231079, "learning_rate": 8.735316695702297e-05, "loss": 0.0464, "step": 26990 }, { "grad_norm": 0.3296808898448944, "learning_rate": 8.734217342684769e-05, "loss": 0.0378, "step": 27000 }, { "grad_norm": 0.4545842409133911, "learning_rate": 8.733117581298847e-05, "loss": 0.0349, "step": 27010 }, { "grad_norm": 0.40129807591438293, "learning_rate": 8.732017411664796e-05, "loss": 0.0504, "step": 27020 }, { "grad_norm": 0.36272767186164856, "learning_rate": 8.730916833902936e-05, "loss": 0.0468, "step": 27030 }, { "grad_norm": 0.45421186089515686, "learning_rate": 8.729815848133618e-05, "loss": 0.0666, "step": 27040 }, { "grad_norm": 0.402863085269928, "learning_rate": 8.728714454477247e-05, "loss": 0.0425, "step": 27050 }, { "grad_norm": 0.4636208117008209, "learning_rate": 8.727612653054269e-05, "loss": 0.0567, "step": 27060 }, { "grad_norm": 0.43674150109291077, "learning_rate": 8.726510443985176e-05, "loss": 0.0468, "step": 27070 }, { "grad_norm": 0.5032486319541931, "learning_rate": 8.725407827390503e-05, "loss": 0.0564, "step": 27080 }, { "grad_norm": 0.3556166887283325, "learning_rate": 8.724304803390833e-05, "loss": 0.0498, "step": 27090 }, { "grad_norm": 0.38231196999549866, "learning_rate": 8.723201372106788e-05, "loss": 0.05, "step": 27100 }, { "grad_norm": 0.5485209822654724, "learning_rate": 8.722097533659038e-05, "loss": 0.0487, "step": 27110 }, { "grad_norm": 0.5317662954330444, "learning_rate": 8.720993288168299e-05, "loss": 0.0516, "step": 27120 }, { "grad_norm": 0.7194883227348328, "learning_rate": 8.719888635755327e-05, "loss": 0.0441, "step": 27130 }, { "grad_norm": 0.5240634679794312, "learning_rate": 8.718783576540928e-05, "loss": 0.0593, "step": 27140 }, { "grad_norm": 0.5149894952774048, "learning_rate": 8.717678110645948e-05, "loss": 0.0626, "step": 27150 }, { "grad_norm": 0.36310410499572754, "learning_rate": 8.716572238191279e-05, "loss": 0.0503, "step": 27160 }, { "grad_norm": 0.6957537531852722, "learning_rate": 8.715465959297857e-05, "loss": 0.056, "step": 27170 }, { "grad_norm": 0.6715856790542603, "learning_rate": 8.714359274086665e-05, "loss": 0.0733, "step": 27180 }, { "grad_norm": 0.3932790756225586, "learning_rate": 8.713252182678726e-05, "loss": 0.0528, "step": 27190 }, { "grad_norm": 0.4347960352897644, "learning_rate": 8.712144685195112e-05, "loss": 0.0726, "step": 27200 }, { "grad_norm": 0.40218472480773926, "learning_rate": 8.711036781756936e-05, "loss": 0.0514, "step": 27210 }, { "grad_norm": 0.5098747611045837, "learning_rate": 8.709928472485357e-05, "loss": 0.0502, "step": 27220 }, { "grad_norm": 0.47997772693634033, "learning_rate": 8.708819757501579e-05, "loss": 0.047, "step": 27230 }, { "grad_norm": 0.34840333461761475, "learning_rate": 8.707710636926846e-05, "loss": 0.0468, "step": 27240 }, { "grad_norm": 0.4461507201194763, "learning_rate": 8.706601110882455e-05, "loss": 0.0583, "step": 27250 }, { "grad_norm": 0.5934039950370789, "learning_rate": 8.705491179489738e-05, "loss": 0.049, "step": 27260 }, { "grad_norm": 0.392635703086853, "learning_rate": 8.704380842870077e-05, "loss": 0.0476, "step": 27270 }, { "grad_norm": 0.3904500901699066, "learning_rate": 8.703270101144895e-05, "loss": 0.0473, "step": 27280 }, { "grad_norm": 0.3586684763431549, "learning_rate": 8.702158954435664e-05, "loss": 0.0464, "step": 27290 }, { "grad_norm": 0.401117742061615, "learning_rate": 8.701047402863896e-05, "loss": 0.0484, "step": 27300 }, { "grad_norm": 0.7000163793563843, "learning_rate": 8.699935446551148e-05, "loss": 0.0511, "step": 27310 }, { "grad_norm": 0.4151580035686493, "learning_rate": 8.698823085619022e-05, "loss": 0.0548, "step": 27320 }, { "grad_norm": 0.515868604183197, "learning_rate": 8.697710320189166e-05, "loss": 0.0468, "step": 27330 }, { "grad_norm": 0.5515232086181641, "learning_rate": 8.696597150383268e-05, "loss": 0.0549, "step": 27340 }, { "grad_norm": 0.5369244813919067, "learning_rate": 8.695483576323063e-05, "loss": 0.0529, "step": 27350 }, { "grad_norm": 0.46810564398765564, "learning_rate": 8.69436959813033e-05, "loss": 0.0525, "step": 27360 }, { "grad_norm": 0.49745815992355347, "learning_rate": 8.693255215926892e-05, "loss": 0.0507, "step": 27370 }, { "grad_norm": 0.6561667323112488, "learning_rate": 8.692140429834617e-05, "loss": 0.0511, "step": 27380 }, { "grad_norm": 0.48287028074264526, "learning_rate": 8.691025239975415e-05, "loss": 0.0587, "step": 27390 }, { "grad_norm": 0.5171939730644226, "learning_rate": 8.689909646471243e-05, "loss": 0.0464, "step": 27400 }, { "grad_norm": 0.7416695356369019, "learning_rate": 8.688793649444099e-05, "loss": 0.0602, "step": 27410 }, { "grad_norm": 0.4653405547142029, "learning_rate": 8.687677249016029e-05, "loss": 0.0549, "step": 27420 }, { "grad_norm": 0.623237133026123, "learning_rate": 8.686560445309118e-05, "loss": 0.0513, "step": 27430 }, { "grad_norm": 0.6572905778884888, "learning_rate": 8.685443238445499e-05, "loss": 0.0648, "step": 27440 }, { "grad_norm": 1.2824223041534424, "learning_rate": 8.68432562854735e-05, "loss": 0.0604, "step": 27450 }, { "grad_norm": 0.31819429993629456, "learning_rate": 8.683207615736887e-05, "loss": 0.0557, "step": 27460 }, { "grad_norm": 0.8570733666419983, "learning_rate": 8.682089200136379e-05, "loss": 0.0704, "step": 27470 }, { "grad_norm": 0.6193891763687134, "learning_rate": 8.680970381868132e-05, "loss": 0.0661, "step": 27480 }, { "grad_norm": 0.39721420407295227, "learning_rate": 8.679851161054498e-05, "loss": 0.0425, "step": 27490 }, { "grad_norm": 0.37816762924194336, "learning_rate": 8.678731537817873e-05, "loss": 0.061, "step": 27500 }, { "grad_norm": 0.6023706793785095, "learning_rate": 8.677611512280697e-05, "loss": 0.0588, "step": 27510 }, { "grad_norm": 0.7266370058059692, "learning_rate": 8.676491084565457e-05, "loss": 0.0653, "step": 27520 }, { "grad_norm": 0.44815537333488464, "learning_rate": 8.675370254794678e-05, "loss": 0.0641, "step": 27530 }, { "grad_norm": 0.3717030882835388, "learning_rate": 8.674249023090935e-05, "loss": 0.052, "step": 27540 }, { "grad_norm": 0.42931661009788513, "learning_rate": 8.673127389576843e-05, "loss": 0.0578, "step": 27550 }, { "grad_norm": 0.4181097149848938, "learning_rate": 8.67200535437506e-05, "loss": 0.0466, "step": 27560 }, { "grad_norm": 0.5907521843910217, "learning_rate": 8.670882917608296e-05, "loss": 0.0639, "step": 27570 }, { "grad_norm": 0.4198938012123108, "learning_rate": 8.669760079399292e-05, "loss": 0.0524, "step": 27580 }, { "grad_norm": 0.4232648015022278, "learning_rate": 8.668636839870845e-05, "loss": 0.0498, "step": 27590 }, { "grad_norm": 0.5734748244285583, "learning_rate": 8.667513199145789e-05, "loss": 0.063, "step": 27600 }, { "grad_norm": 0.5128574371337891, "learning_rate": 8.666389157347002e-05, "loss": 0.0601, "step": 27610 }, { "grad_norm": 0.6653026342391968, "learning_rate": 8.66526471459741e-05, "loss": 0.0531, "step": 27620 }, { "grad_norm": 0.35851895809173584, "learning_rate": 8.66413987101998e-05, "loss": 0.0562, "step": 27630 }, { "grad_norm": 0.4357742667198181, "learning_rate": 8.663014626737723e-05, "loss": 0.054, "step": 27640 }, { "grad_norm": 0.40320372581481934, "learning_rate": 8.661888981873691e-05, "loss": 0.0626, "step": 27650 }, { "grad_norm": 0.3839113712310791, "learning_rate": 8.660762936550988e-05, "loss": 0.0598, "step": 27660 }, { "grad_norm": 1.4453051090240479, "learning_rate": 8.659636490892753e-05, "loss": 0.0641, "step": 27670 }, { "grad_norm": 0.42929190397262573, "learning_rate": 8.658509645022174e-05, "loss": 0.0539, "step": 27680 }, { "grad_norm": 0.39659106731414795, "learning_rate": 8.657382399062481e-05, "loss": 0.0436, "step": 27690 }, { "grad_norm": 0.7689045071601868, "learning_rate": 8.656254753136946e-05, "loss": 0.0556, "step": 27700 }, { "grad_norm": 0.4234004318714142, "learning_rate": 8.655126707368891e-05, "loss": 0.0389, "step": 27710 }, { "grad_norm": 0.5774377584457397, "learning_rate": 8.653998261881672e-05, "loss": 0.0502, "step": 27720 }, { "grad_norm": 0.45680463314056396, "learning_rate": 8.652869416798699e-05, "loss": 0.0475, "step": 27730 }, { "grad_norm": 0.5074104070663452, "learning_rate": 8.651740172243417e-05, "loss": 0.0518, "step": 27740 }, { "grad_norm": 0.3990098834037781, "learning_rate": 8.65061052833932e-05, "loss": 0.055, "step": 27750 }, { "grad_norm": 0.37539294362068176, "learning_rate": 8.649480485209945e-05, "loss": 0.0551, "step": 27760 }, { "grad_norm": 0.5559013485908508, "learning_rate": 8.64835004297887e-05, "loss": 0.0672, "step": 27770 }, { "grad_norm": 0.5015509128570557, "learning_rate": 8.64721920176972e-05, "loss": 0.0552, "step": 27780 }, { "grad_norm": 0.36001718044281006, "learning_rate": 8.646087961706164e-05, "loss": 0.0619, "step": 27790 }, { "grad_norm": 0.3881569802761078, "learning_rate": 8.644956322911908e-05, "loss": 0.0609, "step": 27800 }, { "grad_norm": 0.5147930979728699, "learning_rate": 8.643824285510709e-05, "loss": 0.0602, "step": 27810 }, { "grad_norm": 0.4038097858428955, "learning_rate": 8.642691849626364e-05, "loss": 0.042, "step": 27820 }, { "grad_norm": 0.3720580041408539, "learning_rate": 8.641559015382717e-05, "loss": 0.0465, "step": 27830 }, { "grad_norm": 0.6232574582099915, "learning_rate": 8.640425782903649e-05, "loss": 0.0557, "step": 27840 }, { "grad_norm": 0.4048125445842743, "learning_rate": 8.639292152313091e-05, "loss": 0.0467, "step": 27850 }, { "grad_norm": 0.43353548645973206, "learning_rate": 8.638158123735015e-05, "loss": 0.0493, "step": 27860 }, { "grad_norm": 0.36848679184913635, "learning_rate": 8.637023697293436e-05, "loss": 0.0476, "step": 27870 }, { "grad_norm": 0.5283417105674744, "learning_rate": 8.635888873112414e-05, "loss": 0.0627, "step": 27880 }, { "grad_norm": 0.3661089241504669, "learning_rate": 8.634753651316052e-05, "loss": 0.0545, "step": 27890 }, { "grad_norm": 0.4187943935394287, "learning_rate": 8.633618032028496e-05, "loss": 0.0518, "step": 27900 }, { "grad_norm": 0.45978742837905884, "learning_rate": 8.632482015373934e-05, "loss": 0.0471, "step": 27910 }, { "grad_norm": 0.47754085063934326, "learning_rate": 8.6313456014766e-05, "loss": 0.0486, "step": 27920 }, { "grad_norm": 0.5675164461135864, "learning_rate": 8.630208790460771e-05, "loss": 0.0509, "step": 27930 }, { "grad_norm": 0.3589920699596405, "learning_rate": 8.629071582450768e-05, "loss": 0.053, "step": 27940 }, { "grad_norm": 0.5904591083526611, "learning_rate": 8.62793397757095e-05, "loss": 0.0556, "step": 27950 }, { "grad_norm": 0.6070259213447571, "learning_rate": 8.626795975945729e-05, "loss": 0.0505, "step": 27960 }, { "grad_norm": 0.5982393622398376, "learning_rate": 8.625657577699551e-05, "loss": 0.0472, "step": 27970 }, { "grad_norm": 0.3433419466018677, "learning_rate": 8.624518782956914e-05, "loss": 0.0476, "step": 27980 }, { "grad_norm": 0.4784255623817444, "learning_rate": 8.62337959184235e-05, "loss": 0.0526, "step": 27990 }, { "grad_norm": 0.4840092658996582, "learning_rate": 8.622240004480441e-05, "loss": 0.0598, "step": 28000 }, { "grad_norm": 0.6733340620994568, "learning_rate": 8.621100020995814e-05, "loss": 0.0571, "step": 28010 }, { "grad_norm": 0.4914548397064209, "learning_rate": 8.619959641513132e-05, "loss": 0.0597, "step": 28020 }, { "grad_norm": 0.5851929783821106, "learning_rate": 8.618818866157105e-05, "loss": 0.055, "step": 28030 }, { "grad_norm": 0.7809267640113831, "learning_rate": 8.617677695052487e-05, "loss": 0.0504, "step": 28040 }, { "grad_norm": 0.4031941592693329, "learning_rate": 8.616536128324078e-05, "loss": 0.0521, "step": 28050 }, { "grad_norm": 0.4736310541629791, "learning_rate": 8.615394166096712e-05, "loss": 0.048, "step": 28060 }, { "grad_norm": 0.33170169591903687, "learning_rate": 8.614251808495279e-05, "loss": 0.0465, "step": 28070 }, { "grad_norm": 0.3491293489933014, "learning_rate": 8.6131090556447e-05, "loss": 0.0469, "step": 28080 }, { "grad_norm": 0.5227440595626831, "learning_rate": 8.611965907669947e-05, "loss": 0.0595, "step": 28090 }, { "grad_norm": 0.4154420793056488, "learning_rate": 8.610822364696034e-05, "loss": 0.0492, "step": 28100 }, { "grad_norm": 0.31570589542388916, "learning_rate": 8.609678426848015e-05, "loss": 0.0513, "step": 28110 }, { "grad_norm": 0.4787535071372986, "learning_rate": 8.60853409425099e-05, "loss": 0.0468, "step": 28120 }, { "grad_norm": 0.4462425708770752, "learning_rate": 8.607389367030104e-05, "loss": 0.0437, "step": 28130 }, { "grad_norm": 0.3632185757160187, "learning_rate": 8.606244245310538e-05, "loss": 0.052, "step": 28140 }, { "grad_norm": 0.671917736530304, "learning_rate": 8.605098729217525e-05, "loss": 0.0534, "step": 28150 }, { "grad_norm": 0.3889899253845215, "learning_rate": 8.603952818876335e-05, "loss": 0.062, "step": 28160 }, { "grad_norm": 0.36419668793678284, "learning_rate": 8.602806514412281e-05, "loss": 0.0508, "step": 28170 }, { "grad_norm": 0.4554409682750702, "learning_rate": 8.601659815950726e-05, "loss": 0.0591, "step": 28180 }, { "grad_norm": 0.4638432562351227, "learning_rate": 8.600512723617067e-05, "loss": 0.0695, "step": 28190 }, { "grad_norm": 0.49122655391693115, "learning_rate": 8.59936523753675e-05, "loss": 0.0533, "step": 28200 }, { "grad_norm": 0.27906063199043274, "learning_rate": 8.598217357835264e-05, "loss": 0.0484, "step": 28210 }, { "grad_norm": 0.5831577777862549, "learning_rate": 8.597069084638135e-05, "loss": 0.0532, "step": 28220 }, { "grad_norm": 0.442907452583313, "learning_rate": 8.595920418070939e-05, "loss": 0.0439, "step": 28230 }, { "grad_norm": 0.3309646546840668, "learning_rate": 8.594771358259295e-05, "loss": 0.0483, "step": 28240 }, { "grad_norm": 0.680077314376831, "learning_rate": 8.593621905328858e-05, "loss": 0.0537, "step": 28250 }, { "grad_norm": 0.3747406005859375, "learning_rate": 8.592472059405333e-05, "loss": 0.0598, "step": 28260 }, { "grad_norm": 0.5778444409370422, "learning_rate": 8.591321820614464e-05, "loss": 0.0571, "step": 28270 }, { "grad_norm": 0.5623263716697693, "learning_rate": 8.590171189082041e-05, "loss": 0.0684, "step": 28280 }, { "grad_norm": 0.3788997232913971, "learning_rate": 8.589020164933894e-05, "loss": 0.0532, "step": 28290 }, { "grad_norm": 0.43718570470809937, "learning_rate": 8.587868748295898e-05, "loss": 0.0447, "step": 28300 }, { "grad_norm": 0.41630905866622925, "learning_rate": 8.586716939293971e-05, "loss": 0.0441, "step": 28310 }, { "grad_norm": 0.38078394532203674, "learning_rate": 8.58556473805407e-05, "loss": 0.0467, "step": 28320 }, { "grad_norm": 0.3858703374862671, "learning_rate": 8.584412144702202e-05, "loss": 0.0506, "step": 28330 }, { "grad_norm": 0.3971234858036041, "learning_rate": 8.58325915936441e-05, "loss": 0.0531, "step": 28340 }, { "grad_norm": 0.36153993010520935, "learning_rate": 8.582105782166783e-05, "loss": 0.0504, "step": 28350 }, { "grad_norm": 0.39918941259384155, "learning_rate": 8.580952013235455e-05, "loss": 0.0494, "step": 28360 }, { "grad_norm": 0.3931688070297241, "learning_rate": 8.579797852696596e-05, "loss": 0.0425, "step": 28370 }, { "grad_norm": 0.40007805824279785, "learning_rate": 8.578643300676428e-05, "loss": 0.0611, "step": 28380 }, { "grad_norm": 0.3632367253303528, "learning_rate": 8.577488357301209e-05, "loss": 0.0561, "step": 28390 }, { "grad_norm": 0.6490442752838135, "learning_rate": 8.576333022697242e-05, "loss": 0.0439, "step": 28400 }, { "grad_norm": 0.728557825088501, "learning_rate": 8.575177296990873e-05, "loss": 0.0487, "step": 28410 }, { "grad_norm": 0.5839288234710693, "learning_rate": 8.574021180308489e-05, "loss": 0.0608, "step": 28420 }, { "grad_norm": 0.5730709433555603, "learning_rate": 8.572864672776523e-05, "loss": 0.0569, "step": 28430 }, { "grad_norm": 0.3494274914264679, "learning_rate": 8.571707774521447e-05, "loss": 0.0472, "step": 28440 }, { "grad_norm": 0.3509444296360016, "learning_rate": 8.57055048566978e-05, "loss": 0.0493, "step": 28450 }, { "grad_norm": 0.6082726120948792, "learning_rate": 8.569392806348078e-05, "loss": 0.0544, "step": 28460 }, { "grad_norm": 0.44634199142456055, "learning_rate": 8.568234736682947e-05, "loss": 0.0461, "step": 28470 }, { "grad_norm": 0.47823596000671387, "learning_rate": 8.567076276801029e-05, "loss": 0.062, "step": 28480 }, { "grad_norm": 0.33799123764038086, "learning_rate": 8.565917426829013e-05, "loss": 0.0352, "step": 28490 }, { "grad_norm": 0.5574173331260681, "learning_rate": 8.564758186893628e-05, "loss": 0.0469, "step": 28500 }, { "grad_norm": 0.5743476748466492, "learning_rate": 8.563598557121649e-05, "loss": 0.0609, "step": 28510 }, { "grad_norm": 0.37686648964881897, "learning_rate": 8.562438537639888e-05, "loss": 0.0533, "step": 28520 }, { "grad_norm": 0.5298989415168762, "learning_rate": 8.561278128575206e-05, "loss": 0.0612, "step": 28530 }, { "grad_norm": 0.804802656173706, "learning_rate": 8.5601173300545e-05, "loss": 0.0639, "step": 28540 }, { "grad_norm": 0.7546474933624268, "learning_rate": 8.558956142204717e-05, "loss": 0.0662, "step": 28550 }, { "grad_norm": 0.4718514382839203, "learning_rate": 8.55779456515284e-05, "loss": 0.0663, "step": 28560 }, { "grad_norm": 0.3692357838153839, "learning_rate": 8.556632599025898e-05, "loss": 0.0622, "step": 28570 }, { "grad_norm": 0.39728614687919617, "learning_rate": 8.555470243950964e-05, "loss": 0.0517, "step": 28580 }, { "grad_norm": 0.44635650515556335, "learning_rate": 8.554307500055148e-05, "loss": 0.0553, "step": 28590 }, { "grad_norm": 0.5781970620155334, "learning_rate": 8.553144367465609e-05, "loss": 0.0465, "step": 28600 }, { "grad_norm": 0.4024858772754669, "learning_rate": 8.551980846309544e-05, "loss": 0.0631, "step": 28610 }, { "grad_norm": 0.5221481323242188, "learning_rate": 8.550816936714193e-05, "loss": 0.0607, "step": 28620 }, { "grad_norm": 0.4479677677154541, "learning_rate": 8.549652638806841e-05, "loss": 0.0529, "step": 28630 }, { "grad_norm": 0.7738714218139648, "learning_rate": 8.548487952714812e-05, "loss": 0.0599, "step": 28640 }, { "grad_norm": 0.43302106857299805, "learning_rate": 8.547322878565478e-05, "loss": 0.0491, "step": 28650 }, { "grad_norm": 0.48681947588920593, "learning_rate": 8.546157416486245e-05, "loss": 0.0571, "step": 28660 }, { "grad_norm": 0.4301939606666565, "learning_rate": 8.54499156660457e-05, "loss": 0.0488, "step": 28670 }, { "grad_norm": 0.42647379636764526, "learning_rate": 8.543825329047947e-05, "loss": 0.0391, "step": 28680 }, { "grad_norm": 0.4121210277080536, "learning_rate": 8.542658703943913e-05, "loss": 0.0567, "step": 28690 }, { "grad_norm": 0.4128122627735138, "learning_rate": 8.541491691420051e-05, "loss": 0.0545, "step": 28700 }, { "grad_norm": 0.5193259119987488, "learning_rate": 8.54032429160398e-05, "loss": 0.0483, "step": 28710 }, { "grad_norm": 0.4340468943119049, "learning_rate": 8.539156504623369e-05, "loss": 0.0596, "step": 28720 }, { "grad_norm": 0.4154853820800781, "learning_rate": 8.537988330605923e-05, "loss": 0.0713, "step": 28730 }, { "grad_norm": 0.4195815324783325, "learning_rate": 8.536819769679393e-05, "loss": 0.0458, "step": 28740 }, { "grad_norm": 0.4386506676673889, "learning_rate": 8.53565082197157e-05, "loss": 0.0464, "step": 28750 }, { "grad_norm": 0.473531037569046, "learning_rate": 8.534481487610289e-05, "loss": 0.0698, "step": 28760 }, { "grad_norm": 0.3046623170375824, "learning_rate": 8.533311766723428e-05, "loss": 0.0397, "step": 28770 }, { "grad_norm": 0.4102590084075928, "learning_rate": 8.532141659438901e-05, "loss": 0.0501, "step": 28780 }, { "grad_norm": 0.4138195514678955, "learning_rate": 8.530971165884675e-05, "loss": 0.0527, "step": 28790 }, { "grad_norm": 0.33881616592407227, "learning_rate": 8.529800286188752e-05, "loss": 0.047, "step": 28800 }, { "grad_norm": 0.3336627185344696, "learning_rate": 8.528629020479175e-05, "loss": 0.0509, "step": 28810 }, { "grad_norm": 0.7218307852745056, "learning_rate": 8.527457368884033e-05, "loss": 0.0628, "step": 28820 }, { "grad_norm": 0.44469940662384033, "learning_rate": 8.526285331531458e-05, "loss": 0.0467, "step": 28830 }, { "grad_norm": 0.43909206986427307, "learning_rate": 8.525112908549621e-05, "loss": 0.0452, "step": 28840 }, { "grad_norm": 0.4587121605873108, "learning_rate": 8.523940100066735e-05, "loss": 0.0611, "step": 28850 }, { "grad_norm": 0.3578479588031769, "learning_rate": 8.52276690621106e-05, "loss": 0.045, "step": 28860 }, { "grad_norm": 0.5441311597824097, "learning_rate": 8.521593327110889e-05, "loss": 0.0503, "step": 28870 }, { "grad_norm": 0.49261149764060974, "learning_rate": 8.520419362894569e-05, "loss": 0.0684, "step": 28880 }, { "grad_norm": 0.4677455425262451, "learning_rate": 8.51924501369048e-05, "loss": 0.0414, "step": 28890 }, { "grad_norm": 0.4748176634311676, "learning_rate": 8.518070279627047e-05, "loss": 0.0578, "step": 28900 }, { "grad_norm": 0.5681449770927429, "learning_rate": 8.516895160832737e-05, "loss": 0.0521, "step": 28910 }, { "grad_norm": 0.5554496645927429, "learning_rate": 8.515719657436061e-05, "loss": 0.0466, "step": 28920 }, { "grad_norm": 0.43244117498397827, "learning_rate": 8.514543769565568e-05, "loss": 0.051, "step": 28930 }, { "grad_norm": 0.6741898655891418, "learning_rate": 8.513367497349853e-05, "loss": 0.0515, "step": 28940 }, { "grad_norm": 0.2898496985435486, "learning_rate": 8.51219084091755e-05, "loss": 0.0563, "step": 28950 }, { "grad_norm": 0.6604868769645691, "learning_rate": 8.511013800397338e-05, "loss": 0.0524, "step": 28960 }, { "grad_norm": 0.5429691076278687, "learning_rate": 8.509836375917937e-05, "loss": 0.0613, "step": 28970 }, { "grad_norm": 0.490251362323761, "learning_rate": 8.508658567608104e-05, "loss": 0.0515, "step": 28980 }, { "grad_norm": 0.4559248685836792, "learning_rate": 8.507480375596647e-05, "loss": 0.0432, "step": 28990 }, { "grad_norm": 0.5303736925125122, "learning_rate": 8.506301800012408e-05, "loss": 0.0454, "step": 29000 }, { "grad_norm": 0.4037920832633972, "learning_rate": 8.505122840984278e-05, "loss": 0.051, "step": 29010 }, { "grad_norm": 0.4594244956970215, "learning_rate": 8.503943498641182e-05, "loss": 0.0624, "step": 29020 }, { "grad_norm": 1.0376161336898804, "learning_rate": 8.502763773112095e-05, "loss": 0.0554, "step": 29030 }, { "grad_norm": 0.6223750710487366, "learning_rate": 8.501583664526026e-05, "loss": 0.0566, "step": 29040 }, { "grad_norm": 0.44091787934303284, "learning_rate": 8.500403173012032e-05, "loss": 0.0615, "step": 29050 }, { "grad_norm": 0.4064379930496216, "learning_rate": 8.499222298699211e-05, "loss": 0.0654, "step": 29060 }, { "grad_norm": 0.5372986197471619, "learning_rate": 8.498041041716701e-05, "loss": 0.0508, "step": 29070 }, { "grad_norm": 0.47415804862976074, "learning_rate": 8.496859402193681e-05, "loss": 0.0517, "step": 29080 }, { "grad_norm": 0.44911572337150574, "learning_rate": 8.495677380259374e-05, "loss": 0.0486, "step": 29090 }, { "grad_norm": 0.3765147030353546, "learning_rate": 8.494494976043045e-05, "loss": 0.0466, "step": 29100 }, { "grad_norm": 0.302749902009964, "learning_rate": 8.493312189673998e-05, "loss": 0.0433, "step": 29110 }, { "grad_norm": 0.3661864697933197, "learning_rate": 8.492129021281584e-05, "loss": 0.0457, "step": 29120 }, { "grad_norm": 0.5323320031166077, "learning_rate": 8.490945470995188e-05, "loss": 0.0471, "step": 29130 }, { "grad_norm": 0.39177778363227844, "learning_rate": 8.489761538944247e-05, "loss": 0.048, "step": 29140 }, { "grad_norm": 0.7157452702522278, "learning_rate": 8.48857722525823e-05, "loss": 0.0486, "step": 29150 }, { "grad_norm": 0.3142099678516388, "learning_rate": 8.487392530066652e-05, "loss": 0.0501, "step": 29160 }, { "grad_norm": 0.45918065309524536, "learning_rate": 8.486207453499069e-05, "loss": 0.0535, "step": 29170 }, { "grad_norm": 0.44302836060523987, "learning_rate": 8.485021995685082e-05, "loss": 0.0496, "step": 29180 }, { "grad_norm": 0.3927454352378845, "learning_rate": 8.483836156754328e-05, "loss": 0.0442, "step": 29190 }, { "grad_norm": 0.47266867756843567, "learning_rate": 8.482649936836491e-05, "loss": 0.0469, "step": 29200 }, { "grad_norm": 0.40847310423851013, "learning_rate": 8.481463336061293e-05, "loss": 0.0439, "step": 29210 }, { "grad_norm": 0.34317007660865784, "learning_rate": 8.480276354558496e-05, "loss": 0.0684, "step": 29220 }, { "grad_norm": 1.2443597316741943, "learning_rate": 8.479088992457913e-05, "loss": 0.0632, "step": 29230 }, { "grad_norm": 0.3872075378894806, "learning_rate": 8.477901249889387e-05, "loss": 0.0585, "step": 29240 }, { "grad_norm": 0.4452645182609558, "learning_rate": 8.47671312698281e-05, "loss": 0.048, "step": 29250 }, { "grad_norm": 0.7418755292892456, "learning_rate": 8.475524623868112e-05, "loss": 0.0574, "step": 29260 }, { "grad_norm": 0.43510326743125916, "learning_rate": 8.474335740675266e-05, "loss": 0.0491, "step": 29270 }, { "grad_norm": 0.45386001467704773, "learning_rate": 8.473146477534289e-05, "loss": 0.0575, "step": 29280 }, { "grad_norm": 0.5297321677207947, "learning_rate": 8.471956834575232e-05, "loss": 0.054, "step": 29290 }, { "grad_norm": 0.704128086566925, "learning_rate": 8.470766811928197e-05, "loss": 0.0487, "step": 29300 }, { "grad_norm": 0.5362609624862671, "learning_rate": 8.469576409723323e-05, "loss": 0.0477, "step": 29310 }, { "grad_norm": 0.4561554193496704, "learning_rate": 8.468385628090788e-05, "loss": 0.0485, "step": 29320 }, { "grad_norm": 0.48697629570961, "learning_rate": 8.467194467160815e-05, "loss": 0.0489, "step": 29330 }, { "grad_norm": 0.8888503909111023, "learning_rate": 8.466002927063667e-05, "loss": 0.0505, "step": 29340 }, { "grad_norm": 0.4173516631126404, "learning_rate": 8.464811007929651e-05, "loss": 0.0526, "step": 29350 }, { "grad_norm": 0.531941294670105, "learning_rate": 8.463618709889114e-05, "loss": 0.0473, "step": 29360 }, { "grad_norm": 0.4140050709247589, "learning_rate": 8.462426033072442e-05, "loss": 0.0598, "step": 29370 }, { "grad_norm": 0.8411747813224792, "learning_rate": 8.461232977610061e-05, "loss": 0.0542, "step": 29380 }, { "grad_norm": 0.5101422667503357, "learning_rate": 8.46003954363245e-05, "loss": 0.0518, "step": 29390 }, { "grad_norm": 0.37427785992622375, "learning_rate": 8.458845731270115e-05, "loss": 0.0524, "step": 29400 }, { "grad_norm": 0.4126964807510376, "learning_rate": 8.45765154065361e-05, "loss": 0.0555, "step": 29410 }, { "grad_norm": 0.46308422088623047, "learning_rate": 8.456456971913532e-05, "loss": 0.0503, "step": 29420 }, { "grad_norm": 0.7742332816123962, "learning_rate": 8.455262025180517e-05, "loss": 0.0548, "step": 29430 }, { "grad_norm": 0.3297988772392273, "learning_rate": 8.454066700585242e-05, "loss": 0.0475, "step": 29440 }, { "grad_norm": 0.4129036068916321, "learning_rate": 8.452870998258423e-05, "loss": 0.0597, "step": 29450 }, { "grad_norm": 0.35242992639541626, "learning_rate": 8.451674918330825e-05, "loss": 0.0476, "step": 29460 }, { "grad_norm": 0.48722296953201294, "learning_rate": 8.450478460933246e-05, "loss": 0.0552, "step": 29470 }, { "grad_norm": 0.4220893681049347, "learning_rate": 8.449281626196532e-05, "loss": 0.0588, "step": 29480 }, { "grad_norm": 0.4362422227859497, "learning_rate": 8.448084414251564e-05, "loss": 0.0584, "step": 29490 }, { "grad_norm": 0.3650757670402527, "learning_rate": 8.446886825229271e-05, "loss": 0.0468, "step": 29500 }, { "grad_norm": 0.4731654226779938, "learning_rate": 8.445688859260615e-05, "loss": 0.0502, "step": 29510 }, { "grad_norm": 0.35793352127075195, "learning_rate": 8.444490516476606e-05, "loss": 0.0444, "step": 29520 }, { "grad_norm": 0.35657384991645813, "learning_rate": 8.443291797008293e-05, "loss": 0.0549, "step": 29530 }, { "grad_norm": 0.36916983127593994, "learning_rate": 8.442092700986765e-05, "loss": 0.0593, "step": 29540 }, { "grad_norm": 0.32606810331344604, "learning_rate": 8.440893228543156e-05, "loss": 0.048, "step": 29550 }, { "grad_norm": 0.4626712501049042, "learning_rate": 8.439693379808638e-05, "loss": 0.0467, "step": 29560 }, { "grad_norm": 0.3609310984611511, "learning_rate": 8.43849315491442e-05, "loss": 0.0526, "step": 29570 }, { "grad_norm": 0.4806734621524811, "learning_rate": 8.437292553991763e-05, "loss": 0.0488, "step": 29580 }, { "grad_norm": 0.7168096899986267, "learning_rate": 8.436091577171959e-05, "loss": 0.0611, "step": 29590 }, { "grad_norm": 0.46289023756980896, "learning_rate": 8.434890224586347e-05, "loss": 0.0537, "step": 29600 }, { "grad_norm": 0.6197232604026794, "learning_rate": 8.433688496366303e-05, "loss": 0.0544, "step": 29610 }, { "grad_norm": 0.431845486164093, "learning_rate": 8.432486392643248e-05, "loss": 0.0502, "step": 29620 }, { "grad_norm": 0.6005504727363586, "learning_rate": 8.431283913548643e-05, "loss": 0.0574, "step": 29630 }, { "grad_norm": 0.5090936422348022, "learning_rate": 8.430081059213985e-05, "loss": 0.0581, "step": 29640 }, { "grad_norm": 0.35803985595703125, "learning_rate": 8.428877829770823e-05, "loss": 0.0582, "step": 29650 }, { "grad_norm": 0.43865251541137695, "learning_rate": 8.427674225350735e-05, "loss": 0.0571, "step": 29660 }, { "grad_norm": 0.448697030544281, "learning_rate": 8.426470246085347e-05, "loss": 0.0508, "step": 29670 }, { "grad_norm": 0.38278934359550476, "learning_rate": 8.425265892106324e-05, "loss": 0.054, "step": 29680 }, { "grad_norm": 0.2838868796825409, "learning_rate": 8.424061163545374e-05, "loss": 0.0541, "step": 29690 }, { "grad_norm": 0.4101909101009369, "learning_rate": 8.422856060534243e-05, "loss": 0.0463, "step": 29700 }, { "grad_norm": 0.4814646244049072, "learning_rate": 8.421650583204718e-05, "loss": 0.0439, "step": 29710 }, { "grad_norm": 0.4603212773799896, "learning_rate": 8.420444731688633e-05, "loss": 0.0511, "step": 29720 }, { "grad_norm": 0.4160260856151581, "learning_rate": 8.419238506117852e-05, "loss": 0.0671, "step": 29730 }, { "grad_norm": 0.7664984464645386, "learning_rate": 8.418031906624289e-05, "loss": 0.0463, "step": 29740 }, { "grad_norm": 0.421132355928421, "learning_rate": 8.416824933339898e-05, "loss": 0.0559, "step": 29750 }, { "grad_norm": 0.46961209177970886, "learning_rate": 8.415617586396667e-05, "loss": 0.0426, "step": 29760 }, { "grad_norm": 0.533136248588562, "learning_rate": 8.414409865926632e-05, "loss": 0.0531, "step": 29770 }, { "grad_norm": 0.4927309453487396, "learning_rate": 8.413201772061867e-05, "loss": 0.0549, "step": 29780 }, { "grad_norm": 0.4104955792427063, "learning_rate": 8.411993304934488e-05, "loss": 0.0402, "step": 29790 }, { "grad_norm": 0.5180081129074097, "learning_rate": 8.410784464676654e-05, "loss": 0.0525, "step": 29800 }, { "grad_norm": 0.31322428584098816, "learning_rate": 8.409575251420556e-05, "loss": 0.0489, "step": 29810 }, { "grad_norm": 0.44039610028266907, "learning_rate": 8.408365665298435e-05, "loss": 0.0536, "step": 29820 }, { "grad_norm": 0.7680988311767578, "learning_rate": 8.40715570644257e-05, "loss": 0.06, "step": 29830 }, { "grad_norm": 0.442139595746994, "learning_rate": 8.40594537498528e-05, "loss": 0.0673, "step": 29840 }, { "grad_norm": 0.3953278064727783, "learning_rate": 8.404734671058924e-05, "loss": 0.0668, "step": 29850 }, { "grad_norm": 0.4031200408935547, "learning_rate": 8.403523594795902e-05, "loss": 0.0592, "step": 29860 }, { "grad_norm": 0.5008127689361572, "learning_rate": 8.402312146328659e-05, "loss": 0.0501, "step": 29870 }, { "grad_norm": 0.4197515845298767, "learning_rate": 8.401100325789675e-05, "loss": 0.0539, "step": 29880 }, { "grad_norm": 0.4987172484397888, "learning_rate": 8.399888133311472e-05, "loss": 0.0526, "step": 29890 }, { "grad_norm": 0.4629811644554138, "learning_rate": 8.398675569026613e-05, "loss": 0.0588, "step": 29900 }, { "grad_norm": 0.44301488995552063, "learning_rate": 8.397462633067705e-05, "loss": 0.0478, "step": 29910 }, { "grad_norm": 0.41269397735595703, "learning_rate": 8.396249325567392e-05, "loss": 0.0476, "step": 29920 }, { "grad_norm": 0.6140600442886353, "learning_rate": 8.395035646658357e-05, "loss": 0.0718, "step": 29930 }, { "grad_norm": 0.5575695633888245, "learning_rate": 8.39382159647333e-05, "loss": 0.057, "step": 29940 }, { "grad_norm": 0.44968143105506897, "learning_rate": 8.392607175145075e-05, "loss": 0.0481, "step": 29950 }, { "grad_norm": 0.4676527976989746, "learning_rate": 8.3913923828064e-05, "loss": 0.061, "step": 29960 }, { "grad_norm": 0.43395188450813293, "learning_rate": 8.390177219590152e-05, "loss": 0.047, "step": 29970 }, { "grad_norm": 0.7097350358963013, "learning_rate": 8.388961685629222e-05, "loss": 0.0632, "step": 29980 }, { "grad_norm": 0.5616375803947449, "learning_rate": 8.387745781056536e-05, "loss": 0.0526, "step": 29990 }, { "grad_norm": 0.4514564871788025, "learning_rate": 8.386529506005065e-05, "loss": 0.0434, "step": 30000 }, { "grad_norm": 0.3903607130050659, "learning_rate": 8.38531286060782e-05, "loss": 0.0492, "step": 30010 }, { "grad_norm": 0.6668137907981873, "learning_rate": 8.384095844997849e-05, "loss": 0.042, "step": 30020 }, { "grad_norm": 0.4142272472381592, "learning_rate": 8.382878459308245e-05, "loss": 0.0433, "step": 30030 }, { "grad_norm": 0.41090986132621765, "learning_rate": 8.381660703672138e-05, "loss": 0.0474, "step": 30040 }, { "grad_norm": 0.520655632019043, "learning_rate": 8.380442578222702e-05, "loss": 0.0605, "step": 30050 }, { "grad_norm": 0.6087146997451782, "learning_rate": 8.379224083093146e-05, "loss": 0.0484, "step": 30060 }, { "grad_norm": 0.3311358094215393, "learning_rate": 8.378005218416727e-05, "loss": 0.0552, "step": 30070 }, { "grad_norm": 0.6028467416763306, "learning_rate": 8.376785984326735e-05, "loss": 0.0593, "step": 30080 }, { "grad_norm": 0.39345961809158325, "learning_rate": 8.375566380956506e-05, "loss": 0.0501, "step": 30090 }, { "grad_norm": 0.6351670622825623, "learning_rate": 8.374346408439411e-05, "loss": 0.0578, "step": 30100 }, { "grad_norm": 0.40846237540245056, "learning_rate": 8.373126066908868e-05, "loss": 0.0499, "step": 30110 }, { "grad_norm": 0.5278502702713013, "learning_rate": 8.371905356498326e-05, "loss": 0.0505, "step": 30120 }, { "grad_norm": 0.40284767746925354, "learning_rate": 8.370684277341288e-05, "loss": 0.0482, "step": 30130 }, { "grad_norm": 0.501997709274292, "learning_rate": 8.369462829571282e-05, "loss": 0.0502, "step": 30140 }, { "grad_norm": 0.5416826009750366, "learning_rate": 8.36824101332189e-05, "loss": 0.0473, "step": 30150 }, { "grad_norm": 0.38333743810653687, "learning_rate": 8.367018828726721e-05, "loss": 0.0437, "step": 30160 }, { "grad_norm": 0.49730992317199707, "learning_rate": 8.365796275919438e-05, "loss": 0.0559, "step": 30170 }, { "grad_norm": 0.5616708993911743, "learning_rate": 8.364573355033734e-05, "loss": 0.043, "step": 30180 }, { "grad_norm": 0.49373331665992737, "learning_rate": 8.363350066203346e-05, "loss": 0.04, "step": 30190 }, { "grad_norm": 1.2000761032104492, "learning_rate": 8.362126409562053e-05, "loss": 0.0534, "step": 30200 }, { "grad_norm": 0.5377982258796692, "learning_rate": 8.360902385243667e-05, "loss": 0.0462, "step": 30210 }, { "grad_norm": 0.5093345046043396, "learning_rate": 8.359677993382052e-05, "loss": 0.0467, "step": 30220 }, { "grad_norm": 0.3650617003440857, "learning_rate": 8.358453234111103e-05, "loss": 0.0422, "step": 30230 }, { "grad_norm": 0.3778090178966522, "learning_rate": 8.357228107564756e-05, "loss": 0.0521, "step": 30240 }, { "grad_norm": 0.5234905481338501, "learning_rate": 8.356002613876993e-05, "loss": 0.0448, "step": 30250 }, { "grad_norm": 0.3684072494506836, "learning_rate": 8.35477675318183e-05, "loss": 0.035, "step": 30260 }, { "grad_norm": 0.5012183785438538, "learning_rate": 8.353550525613323e-05, "loss": 0.0482, "step": 30270 }, { "grad_norm": 0.35334035754203796, "learning_rate": 8.352323931305572e-05, "loss": 0.0562, "step": 30280 }, { "grad_norm": 0.43851643800735474, "learning_rate": 8.351096970392717e-05, "loss": 0.0577, "step": 30290 }, { "grad_norm": 0.3007447421550751, "learning_rate": 8.349869643008937e-05, "loss": 0.0412, "step": 30300 }, { "grad_norm": 0.3303740620613098, "learning_rate": 8.348641949288449e-05, "loss": 0.0543, "step": 30310 }, { "grad_norm": 0.46392711997032166, "learning_rate": 8.34741388936551e-05, "loss": 0.0546, "step": 30320 }, { "grad_norm": 0.5264238119125366, "learning_rate": 8.346185463374423e-05, "loss": 0.0597, "step": 30330 }, { "grad_norm": 0.3571186661720276, "learning_rate": 8.344956671449524e-05, "loss": 0.0512, "step": 30340 }, { "grad_norm": 0.4873559772968292, "learning_rate": 8.343727513725192e-05, "loss": 0.0491, "step": 30350 }, { "grad_norm": 0.5440251231193542, "learning_rate": 8.342497990335847e-05, "loss": 0.0526, "step": 30360 }, { "grad_norm": 0.7361340522766113, "learning_rate": 8.341268101415946e-05, "loss": 0.058, "step": 30370 }, { "grad_norm": 0.3806553781032562, "learning_rate": 8.34003784709999e-05, "loss": 0.0485, "step": 30380 }, { "grad_norm": 0.3016939163208008, "learning_rate": 8.338807227522516e-05, "loss": 0.0491, "step": 30390 }, { "grad_norm": 0.3435434103012085, "learning_rate": 8.337576242818103e-05, "loss": 0.055, "step": 30400 }, { "grad_norm": 0.28907182812690735, "learning_rate": 8.336344893121372e-05, "loss": 0.0497, "step": 30410 }, { "grad_norm": 0.45642560720443726, "learning_rate": 8.335113178566977e-05, "loss": 0.0426, "step": 30420 }, { "grad_norm": 0.6169692277908325, "learning_rate": 8.33388109928962e-05, "loss": 0.0508, "step": 30430 }, { "grad_norm": 0.5566345453262329, "learning_rate": 8.33264865542404e-05, "loss": 0.0548, "step": 30440 }, { "grad_norm": 0.2755270004272461, "learning_rate": 8.331415847105013e-05, "loss": 0.0567, "step": 30450 }, { "grad_norm": 0.41834214329719543, "learning_rate": 8.330182674467357e-05, "loss": 0.0442, "step": 30460 }, { "grad_norm": 0.5175498127937317, "learning_rate": 8.32894913764593e-05, "loss": 0.051, "step": 30470 }, { "grad_norm": 0.4837392568588257, "learning_rate": 8.327715236775633e-05, "loss": 0.0495, "step": 30480 }, { "grad_norm": 0.3863295018672943, "learning_rate": 8.326480971991398e-05, "loss": 0.0454, "step": 30490 }, { "grad_norm": 0.4190269112586975, "learning_rate": 8.325246343428206e-05, "loss": 0.054, "step": 30500 }, { "grad_norm": 0.39434099197387695, "learning_rate": 8.324011351221072e-05, "loss": 0.0421, "step": 30510 }, { "grad_norm": 0.30424168705940247, "learning_rate": 8.322775995505057e-05, "loss": 0.0472, "step": 30520 }, { "grad_norm": 0.41859668493270874, "learning_rate": 8.321540276415254e-05, "loss": 0.0651, "step": 30530 }, { "grad_norm": 0.41262638568878174, "learning_rate": 8.320304194086798e-05, "loss": 0.0538, "step": 30540 }, { "grad_norm": 0.32466188073158264, "learning_rate": 8.31906774865487e-05, "loss": 0.0394, "step": 30550 }, { "grad_norm": 0.4101989269256592, "learning_rate": 8.317830940254682e-05, "loss": 0.0474, "step": 30560 }, { "grad_norm": 0.58473801612854, "learning_rate": 8.316593769021491e-05, "loss": 0.0575, "step": 30570 }, { "grad_norm": 0.554638147354126, "learning_rate": 8.315356235090592e-05, "loss": 0.0465, "step": 30580 }, { "grad_norm": 0.6300260424613953, "learning_rate": 8.314118338597319e-05, "loss": 0.0496, "step": 30590 }, { "grad_norm": 0.69258713722229, "learning_rate": 8.312880079677048e-05, "loss": 0.0507, "step": 30600 }, { "grad_norm": 0.39293888211250305, "learning_rate": 8.311641458465191e-05, "loss": 0.0546, "step": 30610 }, { "grad_norm": 0.421220988035202, "learning_rate": 8.310402475097205e-05, "loss": 0.0512, "step": 30620 }, { "grad_norm": 0.3751610517501831, "learning_rate": 8.309163129708581e-05, "loss": 0.0458, "step": 30630 }, { "grad_norm": 0.40850338339805603, "learning_rate": 8.307923422434852e-05, "loss": 0.0445, "step": 30640 }, { "grad_norm": 0.4301648437976837, "learning_rate": 8.30668335341159e-05, "loss": 0.0576, "step": 30650 }, { "grad_norm": 0.4495495557785034, "learning_rate": 8.305442922774408e-05, "loss": 0.0543, "step": 30660 }, { "grad_norm": 0.774531900882721, "learning_rate": 8.304202130658959e-05, "loss": 0.0506, "step": 30670 }, { "grad_norm": 0.49243786931037903, "learning_rate": 8.302960977200931e-05, "loss": 0.0422, "step": 30680 }, { "grad_norm": 0.5068665146827698, "learning_rate": 8.301719462536058e-05, "loss": 0.0515, "step": 30690 }, { "grad_norm": 0.5218016505241394, "learning_rate": 8.300477586800108e-05, "loss": 0.0471, "step": 30700 }, { "grad_norm": 0.29302334785461426, "learning_rate": 8.299235350128892e-05, "loss": 0.0498, "step": 30710 }, { "grad_norm": 0.5312260985374451, "learning_rate": 8.297992752658258e-05, "loss": 0.0359, "step": 30720 }, { "grad_norm": 0.3934924602508545, "learning_rate": 8.296749794524094e-05, "loss": 0.0521, "step": 30730 }, { "grad_norm": 0.39109906554222107, "learning_rate": 8.295506475862332e-05, "loss": 0.0369, "step": 30740 }, { "grad_norm": 0.42063748836517334, "learning_rate": 8.294262796808933e-05, "loss": 0.0391, "step": 30750 }, { "grad_norm": 0.541274905204773, "learning_rate": 8.293018757499909e-05, "loss": 0.0478, "step": 30760 }, { "grad_norm": 0.4038959741592407, "learning_rate": 8.291774358071305e-05, "loss": 0.0453, "step": 30770 }, { "grad_norm": 0.9503039121627808, "learning_rate": 8.290529598659205e-05, "loss": 0.0564, "step": 30780 }, { "grad_norm": 0.40034040808677673, "learning_rate": 8.289284479399738e-05, "loss": 0.0564, "step": 30790 }, { "grad_norm": 0.7142107486724854, "learning_rate": 8.288039000429064e-05, "loss": 0.0471, "step": 30800 }, { "grad_norm": 0.5499605536460876, "learning_rate": 8.286793161883388e-05, "loss": 0.0521, "step": 30810 }, { "grad_norm": 0.3663977086544037, "learning_rate": 8.285546963898954e-05, "loss": 0.0501, "step": 30820 }, { "grad_norm": 0.48174917697906494, "learning_rate": 8.284300406612044e-05, "loss": 0.0509, "step": 30830 }, { "grad_norm": 0.5385723114013672, "learning_rate": 8.283053490158978e-05, "loss": 0.0442, "step": 30840 }, { "grad_norm": 0.4437209963798523, "learning_rate": 8.28180621467612e-05, "loss": 0.0437, "step": 30850 }, { "grad_norm": 0.499977707862854, "learning_rate": 8.280558580299867e-05, "loss": 0.0525, "step": 30860 }, { "grad_norm": 0.6016925573348999, "learning_rate": 8.279310587166661e-05, "loss": 0.0452, "step": 30870 }, { "grad_norm": 0.7153274416923523, "learning_rate": 8.278062235412978e-05, "loss": 0.0517, "step": 30880 }, { "grad_norm": 0.6496900916099548, "learning_rate": 8.276813525175339e-05, "loss": 0.0507, "step": 30890 }, { "grad_norm": 0.42501699924468994, "learning_rate": 8.2755644565903e-05, "loss": 0.049, "step": 30900 }, { "grad_norm": 0.3954797685146332, "learning_rate": 8.274315029794454e-05, "loss": 0.0451, "step": 30910 }, { "grad_norm": 0.9810815453529358, "learning_rate": 8.273065244924443e-05, "loss": 0.0539, "step": 30920 }, { "grad_norm": 1.5201597213745117, "learning_rate": 8.271815102116936e-05, "loss": 0.0405, "step": 30930 }, { "grad_norm": 0.5427749156951904, "learning_rate": 8.270564601508648e-05, "loss": 0.0443, "step": 30940 }, { "grad_norm": 0.4608166813850403, "learning_rate": 8.269313743236333e-05, "loss": 0.053, "step": 30950 }, { "grad_norm": 0.5282706618309021, "learning_rate": 8.268062527436783e-05, "loss": 0.0547, "step": 30960 }, { "grad_norm": 0.4853528141975403, "learning_rate": 8.266810954246828e-05, "loss": 0.064, "step": 30970 }, { "grad_norm": 0.6378244161605835, "learning_rate": 8.265559023803338e-05, "loss": 0.045, "step": 30980 }, { "grad_norm": 0.4113699197769165, "learning_rate": 8.264306736243223e-05, "loss": 0.0484, "step": 30990 }, { "grad_norm": 0.6932719945907593, "learning_rate": 8.263054091703432e-05, "loss": 0.0562, "step": 31000 }, { "grad_norm": 0.4830980896949768, "learning_rate": 8.26180109032095e-05, "loss": 0.0437, "step": 31010 }, { "grad_norm": 0.8732920289039612, "learning_rate": 8.260547732232807e-05, "loss": 0.0423, "step": 31020 }, { "grad_norm": 0.43510669469833374, "learning_rate": 8.259294017576063e-05, "loss": 0.0485, "step": 31030 }, { "grad_norm": 0.45190539956092834, "learning_rate": 8.258039946487828e-05, "loss": 0.0505, "step": 31040 }, { "grad_norm": 0.35740289092063904, "learning_rate": 8.256785519105241e-05, "loss": 0.0457, "step": 31050 }, { "grad_norm": 0.7174563407897949, "learning_rate": 8.255530735565488e-05, "loss": 0.0504, "step": 31060 }, { "grad_norm": 0.3322531580924988, "learning_rate": 8.254275596005787e-05, "loss": 0.0408, "step": 31070 }, { "grad_norm": 0.4620373547077179, "learning_rate": 8.2530201005634e-05, "loss": 0.0543, "step": 31080 }, { "grad_norm": 0.36191821098327637, "learning_rate": 8.251764249375626e-05, "loss": 0.0555, "step": 31090 }, { "grad_norm": 0.5291821360588074, "learning_rate": 8.250508042579803e-05, "loss": 0.0495, "step": 31100 }, { "grad_norm": 0.8920841217041016, "learning_rate": 8.249251480313307e-05, "loss": 0.0657, "step": 31110 }, { "grad_norm": 0.39982256293296814, "learning_rate": 8.247994562713555e-05, "loss": 0.0488, "step": 31120 }, { "grad_norm": 0.4118933081626892, "learning_rate": 8.246737289917998e-05, "loss": 0.0544, "step": 31130 }, { "grad_norm": 0.37749919295310974, "learning_rate": 8.245479662064135e-05, "loss": 0.0503, "step": 31140 }, { "grad_norm": 0.5813236832618713, "learning_rate": 8.244221679289496e-05, "loss": 0.0461, "step": 31150 }, { "grad_norm": 0.4426780641078949, "learning_rate": 8.242963341731652e-05, "loss": 0.0528, "step": 31160 }, { "grad_norm": 0.6484577655792236, "learning_rate": 8.24170464952821e-05, "loss": 0.0452, "step": 31170 }, { "grad_norm": 0.377900630235672, "learning_rate": 8.240445602816824e-05, "loss": 0.049, "step": 31180 }, { "grad_norm": 1.4678020477294922, "learning_rate": 8.239186201735179e-05, "loss": 0.0495, "step": 31190 }, { "grad_norm": 0.40354108810424805, "learning_rate": 8.237926446420998e-05, "loss": 0.0429, "step": 31200 }, { "grad_norm": 0.42909571528434753, "learning_rate": 8.236666337012052e-05, "loss": 0.0463, "step": 31210 }, { "grad_norm": 0.6048870086669922, "learning_rate": 8.23540587364614e-05, "loss": 0.0577, "step": 31220 }, { "grad_norm": 0.28732138872146606, "learning_rate": 8.234145056461107e-05, "loss": 0.0468, "step": 31230 }, { "grad_norm": 0.43484947085380554, "learning_rate": 8.232883885594832e-05, "loss": 0.0572, "step": 31240 }, { "grad_norm": 0.4839116632938385, "learning_rate": 8.231622361185236e-05, "loss": 0.0661, "step": 31250 }, { "grad_norm": 0.5668026208877563, "learning_rate": 8.230360483370278e-05, "loss": 0.0693, "step": 31260 }, { "grad_norm": 0.6231167316436768, "learning_rate": 8.229098252287953e-05, "loss": 0.0504, "step": 31270 }, { "grad_norm": 0.5482605695724487, "learning_rate": 8.2278356680763e-05, "loss": 0.0578, "step": 31280 }, { "grad_norm": 0.5354491472244263, "learning_rate": 8.22657273087339e-05, "loss": 0.0592, "step": 31290 }, { "grad_norm": 0.6537299156188965, "learning_rate": 8.225309440817336e-05, "loss": 0.0596, "step": 31300 }, { "grad_norm": 0.5999959111213684, "learning_rate": 8.224045798046293e-05, "loss": 0.0721, "step": 31310 }, { "grad_norm": 0.42878323793411255, "learning_rate": 8.22278180269845e-05, "loss": 0.0465, "step": 31320 }, { "grad_norm": 0.5876472592353821, "learning_rate": 8.221517454912031e-05, "loss": 0.0487, "step": 31330 }, { "grad_norm": 0.6776873469352722, "learning_rate": 8.220252754825308e-05, "loss": 0.0416, "step": 31340 }, { "grad_norm": 0.26873791217803955, "learning_rate": 8.218987702576586e-05, "loss": 0.0484, "step": 31350 }, { "grad_norm": 0.28065741062164307, "learning_rate": 8.217722298304207e-05, "loss": 0.0501, "step": 31360 }, { "grad_norm": 0.3579304814338684, "learning_rate": 8.216456542146557e-05, "loss": 0.0461, "step": 31370 }, { "grad_norm": 0.5100035667419434, "learning_rate": 8.215190434242055e-05, "loss": 0.0397, "step": 31380 }, { "grad_norm": 0.5744490027427673, "learning_rate": 8.213923974729161e-05, "loss": 0.05, "step": 31390 }, { "grad_norm": 0.4083622097969055, "learning_rate": 8.212657163746373e-05, "loss": 0.0414, "step": 31400 }, { "grad_norm": 0.39620113372802734, "learning_rate": 8.211390001432227e-05, "loss": 0.0453, "step": 31410 }, { "grad_norm": 0.3512226343154907, "learning_rate": 8.210122487925297e-05, "loss": 0.0466, "step": 31420 }, { "grad_norm": 0.46151983737945557, "learning_rate": 8.208854623364202e-05, "loss": 0.0512, "step": 31430 }, { "grad_norm": 0.323737233877182, "learning_rate": 8.207586407887589e-05, "loss": 0.0411, "step": 31440 }, { "grad_norm": 0.4499281942844391, "learning_rate": 8.206317841634148e-05, "loss": 0.0414, "step": 31450 }, { "grad_norm": 0.2958822250366211, "learning_rate": 8.205048924742609e-05, "loss": 0.0484, "step": 31460 }, { "grad_norm": 0.5052536129951477, "learning_rate": 8.203779657351738e-05, "loss": 0.0411, "step": 31470 }, { "grad_norm": 0.4588641822338104, "learning_rate": 8.20251003960034e-05, "loss": 0.046, "step": 31480 }, { "grad_norm": 0.418128103017807, "learning_rate": 8.201240071627258e-05, "loss": 0.0464, "step": 31490 }, { "grad_norm": 0.34696128964424133, "learning_rate": 8.199969753571377e-05, "loss": 0.0505, "step": 31500 }, { "grad_norm": 0.5519976615905762, "learning_rate": 8.198699085571615e-05, "loss": 0.0482, "step": 31510 }, { "grad_norm": 0.3882805407047272, "learning_rate": 8.197428067766928e-05, "loss": 0.0518, "step": 31520 }, { "grad_norm": 0.41540464758872986, "learning_rate": 8.196156700296316e-05, "loss": 0.0592, "step": 31530 }, { "grad_norm": 0.662615180015564, "learning_rate": 8.194884983298814e-05, "loss": 0.0442, "step": 31540 }, { "grad_norm": 0.36365240812301636, "learning_rate": 8.193612916913491e-05, "loss": 0.0444, "step": 31550 }, { "grad_norm": 0.48298829793930054, "learning_rate": 8.192340501279463e-05, "loss": 0.0489, "step": 31560 }, { "grad_norm": 0.3696316182613373, "learning_rate": 8.191067736535876e-05, "loss": 0.0553, "step": 31570 }, { "grad_norm": 0.559843122959137, "learning_rate": 8.18979462282192e-05, "loss": 0.0541, "step": 31580 }, { "grad_norm": 0.42596593499183655, "learning_rate": 8.188521160276819e-05, "loss": 0.0573, "step": 31590 }, { "grad_norm": 0.41223233938217163, "learning_rate": 8.187247349039837e-05, "loss": 0.0475, "step": 31600 }, { "grad_norm": 0.3727808892726898, "learning_rate": 8.185973189250278e-05, "loss": 0.0652, "step": 31610 }, { "grad_norm": 0.4736608564853668, "learning_rate": 8.184698681047482e-05, "loss": 0.0608, "step": 31620 }, { "grad_norm": 0.4940617084503174, "learning_rate": 8.183423824570827e-05, "loss": 0.0499, "step": 31630 }, { "grad_norm": 1.6737161874771118, "learning_rate": 8.182148619959725e-05, "loss": 0.0591, "step": 31640 }, { "grad_norm": 0.35922446846961975, "learning_rate": 8.180873067353636e-05, "loss": 0.0537, "step": 31650 }, { "grad_norm": 0.3272759020328522, "learning_rate": 8.179597166892052e-05, "loss": 0.0456, "step": 31660 }, { "grad_norm": 0.35177159309387207, "learning_rate": 8.178320918714501e-05, "loss": 0.0467, "step": 31670 }, { "grad_norm": 0.5837201476097107, "learning_rate": 8.177044322960554e-05, "loss": 0.0521, "step": 31680 }, { "grad_norm": 0.49444305896759033, "learning_rate": 8.175767379769816e-05, "loss": 0.0429, "step": 31690 }, { "grad_norm": 0.4291529655456543, "learning_rate": 8.174490089281932e-05, "loss": 0.0545, "step": 31700 }, { "grad_norm": 0.4442196786403656, "learning_rate": 8.173212451636584e-05, "loss": 0.0573, "step": 31710 }, { "grad_norm": 0.9602887034416199, "learning_rate": 8.171934466973493e-05, "loss": 0.063, "step": 31720 }, { "grad_norm": 0.3748042583465576, "learning_rate": 8.170656135432418e-05, "loss": 0.0403, "step": 31730 }, { "grad_norm": 0.4045080542564392, "learning_rate": 8.169377457153155e-05, "loss": 0.0549, "step": 31740 }, { "grad_norm": 0.513693630695343, "learning_rate": 8.168098432275539e-05, "loss": 0.057, "step": 31750 }, { "grad_norm": 0.5664114952087402, "learning_rate": 8.166819060939442e-05, "loss": 0.0539, "step": 31760 }, { "grad_norm": 0.6377497315406799, "learning_rate": 8.165539343284772e-05, "loss": 0.063, "step": 31770 }, { "grad_norm": 0.35574883222579956, "learning_rate": 8.16425927945148e-05, "loss": 0.0585, "step": 31780 }, { "grad_norm": 0.741290271282196, "learning_rate": 8.162978869579551e-05, "loss": 0.056, "step": 31790 }, { "grad_norm": 0.36671051383018494, "learning_rate": 8.161698113809007e-05, "loss": 0.0443, "step": 31800 }, { "grad_norm": 0.4179953336715698, "learning_rate": 8.160417012279911e-05, "loss": 0.063, "step": 31810 }, { "grad_norm": 0.46140986680984497, "learning_rate": 8.159135565132363e-05, "loss": 0.0539, "step": 31820 }, { "grad_norm": 0.5078698396682739, "learning_rate": 8.157853772506498e-05, "loss": 0.0649, "step": 31830 }, { "grad_norm": 0.42560163140296936, "learning_rate": 8.156571634542494e-05, "loss": 0.054, "step": 31840 }, { "grad_norm": 1.4684314727783203, "learning_rate": 8.15528915138056e-05, "loss": 0.0523, "step": 31850 }, { "grad_norm": 0.389362096786499, "learning_rate": 8.154006323160949e-05, "loss": 0.0528, "step": 31860 }, { "grad_norm": 0.5711032748222351, "learning_rate": 8.152723150023949e-05, "loss": 0.0552, "step": 31870 }, { "grad_norm": 0.45738986134529114, "learning_rate": 8.151439632109886e-05, "loss": 0.0571, "step": 31880 }, { "grad_norm": 0.4552816152572632, "learning_rate": 8.150155769559122e-05, "loss": 0.0623, "step": 31890 }, { "grad_norm": 0.9273929595947266, "learning_rate": 8.148871562512058e-05, "loss": 0.0607, "step": 31900 }, { "grad_norm": 0.4162987768650055, "learning_rate": 8.147587011109136e-05, "loss": 0.0626, "step": 31910 }, { "grad_norm": 1.2395552396774292, "learning_rate": 8.14630211549083e-05, "loss": 0.0703, "step": 31920 }, { "grad_norm": 0.4861903190612793, "learning_rate": 8.145016875797655e-05, "loss": 0.0496, "step": 31930 }, { "grad_norm": 0.41847923398017883, "learning_rate": 8.143731292170164e-05, "loss": 0.0579, "step": 31940 }, { "grad_norm": 0.37280401587486267, "learning_rate": 8.142445364748944e-05, "loss": 0.0514, "step": 31950 }, { "grad_norm": 0.38612908124923706, "learning_rate": 8.141159093674624e-05, "loss": 0.0621, "step": 31960 }, { "grad_norm": 0.49754074215888977, "learning_rate": 8.139872479087869e-05, "loss": 0.0478, "step": 31970 }, { "grad_norm": 0.4345981180667877, "learning_rate": 8.13858552112938e-05, "loss": 0.0414, "step": 31980 }, { "grad_norm": 0.4832928776741028, "learning_rate": 8.137298219939895e-05, "loss": 0.0583, "step": 31990 }, { "grad_norm": 0.5845583081245422, "learning_rate": 8.136010575660196e-05, "loss": 0.0529, "step": 32000 }, { "grad_norm": 0.5469772815704346, "learning_rate": 8.134722588431095e-05, "loss": 0.0579, "step": 32010 }, { "grad_norm": 0.8022509813308716, "learning_rate": 8.133434258393444e-05, "loss": 0.0648, "step": 32020 }, { "grad_norm": 0.4173130393028259, "learning_rate": 8.132145585688134e-05, "loss": 0.0533, "step": 32030 }, { "grad_norm": 0.43685632944107056, "learning_rate": 8.130856570456093e-05, "loss": 0.0584, "step": 32040 }, { "grad_norm": 0.45214414596557617, "learning_rate": 8.129567212838283e-05, "loss": 0.0559, "step": 32050 }, { "grad_norm": 0.4653778076171875, "learning_rate": 8.128277512975708e-05, "loss": 0.0576, "step": 32060 }, { "grad_norm": 0.589156985282898, "learning_rate": 8.126987471009408e-05, "loss": 0.0534, "step": 32070 }, { "grad_norm": 0.3467020094394684, "learning_rate": 8.125697087080459e-05, "loss": 0.0443, "step": 32080 }, { "grad_norm": 0.5728257894515991, "learning_rate": 8.124406361329976e-05, "loss": 0.0527, "step": 32090 }, { "grad_norm": 0.47385841608047485, "learning_rate": 8.123115293899111e-05, "loss": 0.048, "step": 32100 }, { "grad_norm": 0.5151092410087585, "learning_rate": 8.121823884929055e-05, "loss": 0.0472, "step": 32110 }, { "grad_norm": 0.48576492071151733, "learning_rate": 8.12053213456103e-05, "loss": 0.036, "step": 32120 }, { "grad_norm": 0.407953679561615, "learning_rate": 8.119240042936303e-05, "loss": 0.0378, "step": 32130 }, { "grad_norm": 0.2964521646499634, "learning_rate": 8.117947610196175e-05, "loss": 0.0416, "step": 32140 }, { "grad_norm": 0.5431731939315796, "learning_rate": 8.116654836481982e-05, "loss": 0.0417, "step": 32150 }, { "grad_norm": 0.3906078636646271, "learning_rate": 8.115361721935106e-05, "loss": 0.0512, "step": 32160 }, { "grad_norm": 0.5028926134109497, "learning_rate": 8.114068266696953e-05, "loss": 0.0544, "step": 32170 }, { "grad_norm": 0.2988712787628174, "learning_rate": 8.112774470908978e-05, "loss": 0.0357, "step": 32180 }, { "grad_norm": 0.28514689207077026, "learning_rate": 8.111480334712665e-05, "loss": 0.0404, "step": 32190 }, { "grad_norm": 0.8543626070022583, "learning_rate": 8.110185858249542e-05, "loss": 0.0454, "step": 32200 }, { "grad_norm": 0.44733911752700806, "learning_rate": 8.108891041661168e-05, "loss": 0.0563, "step": 32210 }, { "grad_norm": 0.4455825984477997, "learning_rate": 8.107595885089146e-05, "loss": 0.0517, "step": 32220 }, { "grad_norm": 0.5019581913948059, "learning_rate": 8.106300388675108e-05, "loss": 0.0525, "step": 32230 }, { "grad_norm": 0.48763805627822876, "learning_rate": 8.105004552560731e-05, "loss": 0.0536, "step": 32240 }, { "grad_norm": 0.5998095273971558, "learning_rate": 8.103708376887724e-05, "loss": 0.0524, "step": 32250 }, { "grad_norm": 0.4446137845516205, "learning_rate": 8.102411861797836e-05, "loss": 0.0505, "step": 32260 }, { "grad_norm": 0.47345060110092163, "learning_rate": 8.101115007432851e-05, "loss": 0.0433, "step": 32270 }, { "grad_norm": 0.6513727903366089, "learning_rate": 8.09981781393459e-05, "loss": 0.0401, "step": 32280 }, { "grad_norm": 0.7657007575035095, "learning_rate": 8.098520281444915e-05, "loss": 0.0447, "step": 32290 }, { "grad_norm": 0.5711519718170166, "learning_rate": 8.09722241010572e-05, "loss": 0.0447, "step": 32300 }, { "grad_norm": 0.293989360332489, "learning_rate": 8.095924200058939e-05, "loss": 0.056, "step": 32310 }, { "grad_norm": 0.4971294701099396, "learning_rate": 8.094625651446541e-05, "loss": 0.0384, "step": 32320 }, { "grad_norm": 0.29264217615127563, "learning_rate": 8.093326764410536e-05, "loss": 0.0498, "step": 32330 }, { "grad_norm": 0.4417336583137512, "learning_rate": 8.092027539092966e-05, "loss": 0.0584, "step": 32340 }, { "grad_norm": 0.39934206008911133, "learning_rate": 8.090727975635913e-05, "loss": 0.0509, "step": 32350 }, { "grad_norm": 0.49319905042648315, "learning_rate": 8.089428074181497e-05, "loss": 0.058, "step": 32360 }, { "grad_norm": 0.3252504765987396, "learning_rate": 8.088127834871871e-05, "loss": 0.0406, "step": 32370 }, { "grad_norm": 0.8331494331359863, "learning_rate": 8.086827257849226e-05, "loss": 0.0552, "step": 32380 }, { "grad_norm": 0.3705076575279236, "learning_rate": 8.085526343255795e-05, "loss": 0.0397, "step": 32390 }, { "grad_norm": 0.5337597727775574, "learning_rate": 8.084225091233842e-05, "loss": 0.059, "step": 32400 }, { "grad_norm": 0.5505791902542114, "learning_rate": 8.082923501925668e-05, "loss": 0.0398, "step": 32410 }, { "grad_norm": 0.49567845463752747, "learning_rate": 8.081621575473617e-05, "loss": 0.054, "step": 32420 }, { "grad_norm": 0.31178608536720276, "learning_rate": 8.080319312020064e-05, "loss": 0.0516, "step": 32430 }, { "grad_norm": 0.4632526934146881, "learning_rate": 8.079016711707421e-05, "loss": 0.0516, "step": 32440 }, { "grad_norm": 0.41924530267715454, "learning_rate": 8.077713774678139e-05, "loss": 0.045, "step": 32450 }, { "grad_norm": 0.5515884160995483, "learning_rate": 8.076410501074707e-05, "loss": 0.0434, "step": 32460 }, { "grad_norm": 0.4408961236476898, "learning_rate": 8.075106891039647e-05, "loss": 0.0529, "step": 32470 }, { "grad_norm": 0.4462811052799225, "learning_rate": 8.073802944715523e-05, "loss": 0.0587, "step": 32480 }, { "grad_norm": 0.5688735842704773, "learning_rate": 8.072498662244929e-05, "loss": 0.0451, "step": 32490 }, { "grad_norm": 0.40515872836112976, "learning_rate": 8.0711940437705e-05, "loss": 0.0439, "step": 32500 }, { "grad_norm": 0.7564606070518494, "learning_rate": 8.06988908943491e-05, "loss": 0.0442, "step": 32510 }, { "grad_norm": 0.3637714087963104, "learning_rate": 8.068583799380863e-05, "loss": 0.042, "step": 32520 }, { "grad_norm": 0.4259677231311798, "learning_rate": 8.067278173751104e-05, "loss": 0.043, "step": 32530 }, { "grad_norm": 0.5058186650276184, "learning_rate": 8.065972212688417e-05, "loss": 0.0422, "step": 32540 }, { "grad_norm": 0.44742289185523987, "learning_rate": 8.064665916335618e-05, "loss": 0.053, "step": 32550 }, { "grad_norm": 0.3063575327396393, "learning_rate": 8.063359284835564e-05, "loss": 0.0491, "step": 32560 }, { "grad_norm": 0.3273659646511078, "learning_rate": 8.062052318331142e-05, "loss": 0.0495, "step": 32570 }, { "grad_norm": 0.4018343389034271, "learning_rate": 8.060745016965283e-05, "loss": 0.0466, "step": 32580 }, { "grad_norm": 0.7672650814056396, "learning_rate": 8.059437380880952e-05, "loss": 0.0428, "step": 32590 }, { "grad_norm": 0.3898192048072815, "learning_rate": 8.058129410221146e-05, "loss": 0.0434, "step": 32600 }, { "grad_norm": 0.4333326816558838, "learning_rate": 8.056821105128908e-05, "loss": 0.0463, "step": 32610 }, { "grad_norm": 0.3822775185108185, "learning_rate": 8.05551246574731e-05, "loss": 0.0439, "step": 32620 }, { "grad_norm": 0.42185595631599426, "learning_rate": 8.05420349221946e-05, "loss": 0.0469, "step": 32630 }, { "grad_norm": 0.7970381379127502, "learning_rate": 8.05289418468851e-05, "loss": 0.039, "step": 32640 }, { "grad_norm": 0.38245248794555664, "learning_rate": 8.051584543297642e-05, "loss": 0.0426, "step": 32650 }, { "grad_norm": 0.526069164276123, "learning_rate": 8.050274568190074e-05, "loss": 0.0446, "step": 32660 }, { "grad_norm": 0.3668428659439087, "learning_rate": 8.048964259509067e-05, "loss": 0.0467, "step": 32670 }, { "grad_norm": 0.49160492420196533, "learning_rate": 8.047653617397914e-05, "loss": 0.052, "step": 32680 }, { "grad_norm": 0.44194164872169495, "learning_rate": 8.046342641999941e-05, "loss": 0.0518, "step": 32690 }, { "grad_norm": 0.38043031096458435, "learning_rate": 8.045031333458517e-05, "loss": 0.0444, "step": 32700 }, { "grad_norm": 0.6059784293174744, "learning_rate": 8.043719691917047e-05, "loss": 0.0542, "step": 32710 }, { "grad_norm": 0.4071055054664612, "learning_rate": 8.042407717518966e-05, "loss": 0.0386, "step": 32720 }, { "grad_norm": 0.4742932617664337, "learning_rate": 8.041095410407751e-05, "loss": 0.0475, "step": 32730 }, { "grad_norm": 0.3248631954193115, "learning_rate": 8.039782770726913e-05, "loss": 0.0429, "step": 32740 }, { "grad_norm": 0.22334085404872894, "learning_rate": 8.038469798620004e-05, "loss": 0.0418, "step": 32750 }, { "grad_norm": 0.4051937460899353, "learning_rate": 8.037156494230604e-05, "loss": 0.0489, "step": 32760 }, { "grad_norm": 0.5607492327690125, "learning_rate": 8.035842857702338e-05, "loss": 0.0414, "step": 32770 }, { "grad_norm": 0.3706533908843994, "learning_rate": 8.03452888917886e-05, "loss": 0.05, "step": 32780 }, { "grad_norm": 0.3095357418060303, "learning_rate": 8.033214588803866e-05, "loss": 0.0404, "step": 32790 }, { "grad_norm": 0.39251387119293213, "learning_rate": 8.031899956721083e-05, "loss": 0.0393, "step": 32800 }, { "grad_norm": 0.4264254570007324, "learning_rate": 8.030584993074282e-05, "loss": 0.0426, "step": 32810 }, { "grad_norm": 0.32053685188293457, "learning_rate": 8.02926969800726e-05, "loss": 0.0499, "step": 32820 }, { "grad_norm": 0.38481977581977844, "learning_rate": 8.027954071663859e-05, "loss": 0.0543, "step": 32830 }, { "grad_norm": 0.43363532423973083, "learning_rate": 8.026638114187954e-05, "loss": 0.042, "step": 32840 }, { "grad_norm": 0.445101797580719, "learning_rate": 8.025321825723456e-05, "loss": 0.0506, "step": 32850 }, { "grad_norm": 0.4225292503833771, "learning_rate": 8.02400520641431e-05, "loss": 0.0589, "step": 32860 }, { "grad_norm": 0.3182205259799957, "learning_rate": 8.022688256404501e-05, "loss": 0.0424, "step": 32870 }, { "grad_norm": 0.5796409249305725, "learning_rate": 8.02137097583805e-05, "loss": 0.0466, "step": 32880 }, { "grad_norm": 0.5222898721694946, "learning_rate": 8.02005336485901e-05, "loss": 0.0479, "step": 32890 }, { "grad_norm": 0.33419308066368103, "learning_rate": 8.018735423611476e-05, "loss": 0.0572, "step": 32900 }, { "grad_norm": 0.9366806745529175, "learning_rate": 8.017417152239574e-05, "loss": 0.0483, "step": 32910 }, { "grad_norm": 0.49374234676361084, "learning_rate": 8.01609855088747e-05, "loss": 0.051, "step": 32920 }, { "grad_norm": 0.48216015100479126, "learning_rate": 8.014779619699362e-05, "loss": 0.0449, "step": 32930 }, { "grad_norm": 0.472956120967865, "learning_rate": 8.013460358819489e-05, "loss": 0.0518, "step": 32940 }, { "grad_norm": 0.4699660837650299, "learning_rate": 8.01214076839212e-05, "loss": 0.0538, "step": 32950 }, { "grad_norm": 0.5160893797874451, "learning_rate": 8.010820848561565e-05, "loss": 0.0423, "step": 32960 }, { "grad_norm": 0.3800444006919861, "learning_rate": 8.009500599472171e-05, "loss": 0.039, "step": 32970 }, { "grad_norm": 0.4379352629184723, "learning_rate": 8.008180021268314e-05, "loss": 0.0421, "step": 32980 }, { "grad_norm": 0.3449932336807251, "learning_rate": 8.006859114094414e-05, "loss": 0.0404, "step": 32990 }, { "grad_norm": 0.3705138862133026, "learning_rate": 8.005537878094921e-05, "loss": 0.0546, "step": 33000 }, { "grad_norm": 0.34100624918937683, "learning_rate": 8.004216313414323e-05, "loss": 0.0544, "step": 33010 }, { "grad_norm": 0.3394896984100342, "learning_rate": 8.002894420197149e-05, "loss": 0.0428, "step": 33020 }, { "grad_norm": 0.3387562930583954, "learning_rate": 8.001572198587954e-05, "loss": 0.0432, "step": 33030 }, { "grad_norm": 0.8581933379173279, "learning_rate": 8.000249648731338e-05, "loss": 0.0452, "step": 33040 }, { "grad_norm": 0.49672365188598633, "learning_rate": 7.998926770771928e-05, "loss": 0.0527, "step": 33050 }, { "grad_norm": 0.5754536390304565, "learning_rate": 7.997603564854397e-05, "loss": 0.066, "step": 33060 }, { "grad_norm": 0.4733847677707672, "learning_rate": 7.996280031123448e-05, "loss": 0.0503, "step": 33070 }, { "grad_norm": 0.3886656165122986, "learning_rate": 7.994956169723818e-05, "loss": 0.061, "step": 33080 }, { "grad_norm": 0.3878486454486847, "learning_rate": 7.993631980800285e-05, "loss": 0.0427, "step": 33090 }, { "grad_norm": 0.6109824180603027, "learning_rate": 7.992307464497659e-05, "loss": 0.0536, "step": 33100 }, { "grad_norm": 0.3212401270866394, "learning_rate": 7.990982620960787e-05, "loss": 0.0357, "step": 33110 }, { "grad_norm": 0.3736859858036041, "learning_rate": 7.989657450334554e-05, "loss": 0.0493, "step": 33120 }, { "grad_norm": 0.3870473802089691, "learning_rate": 7.988331952763877e-05, "loss": 0.0385, "step": 33130 }, { "grad_norm": 0.542404294013977, "learning_rate": 7.987006128393709e-05, "loss": 0.0473, "step": 33140 }, { "grad_norm": 0.5698131322860718, "learning_rate": 7.985679977369043e-05, "loss": 0.045, "step": 33150 }, { "grad_norm": 0.42835569381713867, "learning_rate": 7.984353499834902e-05, "loss": 0.041, "step": 33160 }, { "grad_norm": 0.5233089327812195, "learning_rate": 7.983026695936351e-05, "loss": 0.0525, "step": 33170 }, { "grad_norm": 0.5531136989593506, "learning_rate": 7.981699565818486e-05, "loss": 0.0454, "step": 33180 }, { "grad_norm": 0.5756231546401978, "learning_rate": 7.980372109626437e-05, "loss": 0.0517, "step": 33190 }, { "grad_norm": 0.3792772591114044, "learning_rate": 7.979044327505375e-05, "loss": 0.0418, "step": 33200 }, { "grad_norm": 0.47414129972457886, "learning_rate": 7.977716219600506e-05, "loss": 0.0633, "step": 33210 }, { "grad_norm": 0.5658290982246399, "learning_rate": 7.97638778605707e-05, "loss": 0.0495, "step": 33220 }, { "grad_norm": 0.5965588092803955, "learning_rate": 7.975059027020338e-05, "loss": 0.0615, "step": 33230 }, { "grad_norm": 0.5852965116500854, "learning_rate": 7.973729942635623e-05, "loss": 0.0488, "step": 33240 }, { "grad_norm": 0.4815567433834076, "learning_rate": 7.972400533048273e-05, "loss": 0.0504, "step": 33250 }, { "grad_norm": 0.5932809710502625, "learning_rate": 7.97107079840367e-05, "loss": 0.0515, "step": 33260 }, { "grad_norm": 0.5599609017372131, "learning_rate": 7.969740738847231e-05, "loss": 0.0424, "step": 33270 }, { "grad_norm": 0.590862512588501, "learning_rate": 7.968410354524411e-05, "loss": 0.0469, "step": 33280 }, { "grad_norm": 0.2664354145526886, "learning_rate": 7.967079645580697e-05, "loss": 0.0376, "step": 33290 }, { "grad_norm": 0.45332422852516174, "learning_rate": 7.965748612161612e-05, "loss": 0.0462, "step": 33300 }, { "grad_norm": 0.5084046125411987, "learning_rate": 7.96441725441272e-05, "loss": 0.0596, "step": 33310 }, { "grad_norm": 0.3912438452243805, "learning_rate": 7.963085572479614e-05, "loss": 0.0568, "step": 33320 }, { "grad_norm": 0.36709100008010864, "learning_rate": 7.961753566507924e-05, "loss": 0.0448, "step": 33330 }, { "grad_norm": 0.6974602937698364, "learning_rate": 7.960421236643316e-05, "loss": 0.0533, "step": 33340 }, { "grad_norm": 0.42597755789756775, "learning_rate": 7.959088583031496e-05, "loss": 0.0435, "step": 33350 }, { "grad_norm": 0.3985748589038849, "learning_rate": 7.957755605818194e-05, "loss": 0.0514, "step": 33360 }, { "grad_norm": 0.6292387247085571, "learning_rate": 7.956422305149185e-05, "loss": 0.0579, "step": 33370 }, { "grad_norm": 0.5527913570404053, "learning_rate": 7.95508868117028e-05, "loss": 0.0517, "step": 33380 }, { "grad_norm": 0.40894269943237305, "learning_rate": 7.953754734027318e-05, "loss": 0.0531, "step": 33390 }, { "grad_norm": 0.29157310724258423, "learning_rate": 7.952420463866182e-05, "loss": 0.0453, "step": 33400 }, { "grad_norm": 0.38592812418937683, "learning_rate": 7.951085870832782e-05, "loss": 0.0586, "step": 33410 }, { "grad_norm": 0.43490326404571533, "learning_rate": 7.949750955073067e-05, "loss": 0.0439, "step": 33420 }, { "grad_norm": 0.4629709720611572, "learning_rate": 7.948415716733022e-05, "loss": 0.0551, "step": 33430 }, { "grad_norm": 0.5001355409622192, "learning_rate": 7.947080155958669e-05, "loss": 0.051, "step": 33440 }, { "grad_norm": 0.414530485868454, "learning_rate": 7.94574427289606e-05, "loss": 0.0476, "step": 33450 }, { "grad_norm": 0.4437897503376007, "learning_rate": 7.944408067691284e-05, "loss": 0.0446, "step": 33460 }, { "grad_norm": 0.4637609124183655, "learning_rate": 7.943071540490473e-05, "loss": 0.0484, "step": 33470 }, { "grad_norm": 0.511934220790863, "learning_rate": 7.94173469143978e-05, "loss": 0.0559, "step": 33480 }, { "grad_norm": 0.35276737809181213, "learning_rate": 7.940397520685406e-05, "loss": 0.0422, "step": 33490 }, { "grad_norm": 0.734566330909729, "learning_rate": 7.939060028373577e-05, "loss": 0.0587, "step": 33500 }, { "grad_norm": 0.6158895492553711, "learning_rate": 7.937722214650565e-05, "loss": 0.0661, "step": 33510 }, { "grad_norm": 0.3772863447666168, "learning_rate": 7.936384079662666e-05, "loss": 0.0541, "step": 33520 }, { "grad_norm": 0.4106425940990448, "learning_rate": 7.93504562355622e-05, "loss": 0.0498, "step": 33530 }, { "grad_norm": 0.5123608708381653, "learning_rate": 7.933706846477599e-05, "loss": 0.0585, "step": 33540 }, { "grad_norm": 0.5634910464286804, "learning_rate": 7.932367748573206e-05, "loss": 0.0528, "step": 33550 }, { "grad_norm": 0.3612403869628906, "learning_rate": 7.931028329989485e-05, "loss": 0.0511, "step": 33560 }, { "grad_norm": 0.4803374409675598, "learning_rate": 7.929688590872913e-05, "loss": 0.0541, "step": 33570 }, { "grad_norm": 0.2931687831878662, "learning_rate": 7.928348531370003e-05, "loss": 0.0329, "step": 33580 }, { "grad_norm": 0.5845024585723877, "learning_rate": 7.927008151627297e-05, "loss": 0.0522, "step": 33590 }, { "grad_norm": 0.4003939926624298, "learning_rate": 7.925667451791383e-05, "loss": 0.0457, "step": 33600 }, { "grad_norm": 0.48111122846603394, "learning_rate": 7.924326432008874e-05, "loss": 0.0459, "step": 33610 }, { "grad_norm": 0.34924450516700745, "learning_rate": 7.922985092426422e-05, "loss": 0.0525, "step": 33620 }, { "grad_norm": 0.48443686962127686, "learning_rate": 7.921643433190717e-05, "loss": 0.0539, "step": 33630 }, { "grad_norm": 0.3821417987346649, "learning_rate": 7.920301454448478e-05, "loss": 0.0565, "step": 33640 }, { "grad_norm": 0.28138455748558044, "learning_rate": 7.918959156346461e-05, "loss": 0.041, "step": 33650 }, { "grad_norm": 0.4525277018547058, "learning_rate": 7.91761653903146e-05, "loss": 0.0409, "step": 33660 }, { "grad_norm": 0.4026389420032501, "learning_rate": 7.916273602650302e-05, "loss": 0.0443, "step": 33670 }, { "grad_norm": 0.39438876509666443, "learning_rate": 7.914930347349847e-05, "loss": 0.0445, "step": 33680 }, { "grad_norm": 0.49942532181739807, "learning_rate": 7.913586773276992e-05, "loss": 0.0489, "step": 33690 }, { "grad_norm": 0.5294843912124634, "learning_rate": 7.912242880578667e-05, "loss": 0.0415, "step": 33700 }, { "grad_norm": 0.5645593404769897, "learning_rate": 7.910898669401839e-05, "loss": 0.0465, "step": 33710 }, { "grad_norm": 0.39778923988342285, "learning_rate": 7.909554139893511e-05, "loss": 0.0479, "step": 33720 }, { "grad_norm": 0.3474898338317871, "learning_rate": 7.908209292200715e-05, "loss": 0.0355, "step": 33730 }, { "grad_norm": 0.6360020041465759, "learning_rate": 7.906864126470523e-05, "loss": 0.0469, "step": 33740 }, { "grad_norm": 0.7933972477912903, "learning_rate": 7.905518642850041e-05, "loss": 0.0459, "step": 33750 }, { "grad_norm": 0.47076642513275146, "learning_rate": 7.904172841486409e-05, "loss": 0.0528, "step": 33760 }, { "grad_norm": 0.5259966850280762, "learning_rate": 7.902826722526801e-05, "loss": 0.0506, "step": 33770 }, { "grad_norm": 0.33310315012931824, "learning_rate": 7.901480286118427e-05, "loss": 0.0493, "step": 33780 }, { "grad_norm": 0.3944852352142334, "learning_rate": 7.900133532408531e-05, "loss": 0.0425, "step": 33790 }, { "grad_norm": 0.49256470799446106, "learning_rate": 7.898786461544395e-05, "loss": 0.0422, "step": 33800 }, { "grad_norm": 0.4656261205673218, "learning_rate": 7.897439073673325e-05, "loss": 0.0557, "step": 33810 }, { "grad_norm": 0.40218985080718994, "learning_rate": 7.896091368942677e-05, "loss": 0.057, "step": 33820 }, { "grad_norm": 0.7266269326210022, "learning_rate": 7.894743347499832e-05, "loss": 0.0446, "step": 33830 }, { "grad_norm": 0.38690534234046936, "learning_rate": 7.893395009492203e-05, "loss": 0.0531, "step": 33840 }, { "grad_norm": 0.5575576424598694, "learning_rate": 7.892046355067248e-05, "loss": 0.0476, "step": 33850 }, { "grad_norm": 0.7899539470672607, "learning_rate": 7.890697384372451e-05, "loss": 0.0508, "step": 33860 }, { "grad_norm": 0.4147166609764099, "learning_rate": 7.889348097555336e-05, "loss": 0.0396, "step": 33870 }, { "grad_norm": 0.4370059072971344, "learning_rate": 7.887998494763455e-05, "loss": 0.0504, "step": 33880 }, { "grad_norm": 0.45541784167289734, "learning_rate": 7.886648576144404e-05, "loss": 0.0513, "step": 33890 }, { "grad_norm": 0.35173147916793823, "learning_rate": 7.885298341845802e-05, "loss": 0.0438, "step": 33900 }, { "grad_norm": 0.5171680450439453, "learning_rate": 7.883947792015311e-05, "loss": 0.0475, "step": 33910 }, { "grad_norm": 0.5957611799240112, "learning_rate": 7.882596926800628e-05, "loss": 0.0776, "step": 33920 }, { "grad_norm": 0.4137927293777466, "learning_rate": 7.881245746349477e-05, "loss": 0.0567, "step": 33930 }, { "grad_norm": 0.29667577147483826, "learning_rate": 7.879894250809623e-05, "loss": 0.0482, "step": 33940 }, { "grad_norm": 0.41057083010673523, "learning_rate": 7.878542440328865e-05, "loss": 0.0608, "step": 33950 }, { "grad_norm": 0.39349842071533203, "learning_rate": 7.877190315055031e-05, "loss": 0.0409, "step": 33960 }, { "grad_norm": 0.40907934308052063, "learning_rate": 7.875837875135991e-05, "loss": 0.0521, "step": 33970 }, { "grad_norm": 0.4907262623310089, "learning_rate": 7.874485120719646e-05, "loss": 0.0527, "step": 33980 }, { "grad_norm": 0.33513182401657104, "learning_rate": 7.873132051953928e-05, "loss": 0.0462, "step": 33990 }, { "grad_norm": 0.36741510033607483, "learning_rate": 7.87177866898681e-05, "loss": 0.0508, "step": 34000 }, { "grad_norm": 0.29943764209747314, "learning_rate": 7.870424971966294e-05, "loss": 0.0458, "step": 34010 }, { "grad_norm": 0.5381449460983276, "learning_rate": 7.869070961040419e-05, "loss": 0.0403, "step": 34020 }, { "grad_norm": 0.47942447662353516, "learning_rate": 7.867716636357257e-05, "loss": 0.0442, "step": 34030 }, { "grad_norm": 0.3303493857383728, "learning_rate": 7.866361998064915e-05, "loss": 0.0496, "step": 34040 }, { "grad_norm": 0.3492787480354309, "learning_rate": 7.865007046311534e-05, "loss": 0.0404, "step": 34050 }, { "grad_norm": 0.4494578540325165, "learning_rate": 7.86365178124529e-05, "loss": 0.0466, "step": 34060 }, { "grad_norm": 0.4431302547454834, "learning_rate": 7.862296203014394e-05, "loss": 0.0418, "step": 34070 }, { "grad_norm": 0.49026861786842346, "learning_rate": 7.860940311767088e-05, "loss": 0.0431, "step": 34080 }, { "grad_norm": 0.40405240654945374, "learning_rate": 7.85958410765165e-05, "loss": 0.0432, "step": 34090 }, { "grad_norm": 0.5029217600822449, "learning_rate": 7.858227590816394e-05, "loss": 0.0465, "step": 34100 }, { "grad_norm": 0.3709109425544739, "learning_rate": 7.856870761409664e-05, "loss": 0.0477, "step": 34110 }, { "grad_norm": 0.5431434512138367, "learning_rate": 7.855513619579846e-05, "loss": 0.0475, "step": 34120 }, { "grad_norm": 0.3817933201789856, "learning_rate": 7.85415616547535e-05, "loss": 0.0432, "step": 34130 }, { "grad_norm": 0.2972397208213806, "learning_rate": 7.852798399244627e-05, "loss": 0.0332, "step": 34140 }, { "grad_norm": 0.36700689792633057, "learning_rate": 7.851440321036161e-05, "loss": 0.042, "step": 34150 }, { "grad_norm": 0.36994653940200806, "learning_rate": 7.850081930998468e-05, "loss": 0.0518, "step": 34160 }, { "grad_norm": 0.5314522385597229, "learning_rate": 7.8487232292801e-05, "loss": 0.0492, "step": 34170 }, { "grad_norm": 0.5008684396743774, "learning_rate": 7.847364216029642e-05, "loss": 0.0554, "step": 34180 }, { "grad_norm": 0.47039884328842163, "learning_rate": 7.846004891395716e-05, "loss": 0.0478, "step": 34190 }, { "grad_norm": 0.28902965784072876, "learning_rate": 7.844645255526972e-05, "loss": 0.0432, "step": 34200 }, { "grad_norm": 0.2683808505535126, "learning_rate": 7.843285308572101e-05, "loss": 0.0466, "step": 34210 }, { "grad_norm": 0.6047657132148743, "learning_rate": 7.841925050679823e-05, "loss": 0.0373, "step": 34220 }, { "grad_norm": 0.5072739720344543, "learning_rate": 7.840564481998895e-05, "loss": 0.0536, "step": 34230 }, { "grad_norm": 0.5581194758415222, "learning_rate": 7.839203602678105e-05, "loss": 0.0532, "step": 34240 }, { "grad_norm": 0.37248140573501587, "learning_rate": 7.837842412866279e-05, "loss": 0.0491, "step": 34250 }, { "grad_norm": 0.2989107072353363, "learning_rate": 7.836480912712272e-05, "loss": 0.0501, "step": 34260 }, { "grad_norm": 0.5885781049728394, "learning_rate": 7.835119102364976e-05, "loss": 0.0537, "step": 34270 }, { "grad_norm": 0.3954021632671356, "learning_rate": 7.83375698197332e-05, "loss": 0.0475, "step": 34280 }, { "grad_norm": 0.4400220215320587, "learning_rate": 7.83239455168626e-05, "loss": 0.0472, "step": 34290 }, { "grad_norm": 0.48280903697013855, "learning_rate": 7.83103181165279e-05, "loss": 0.0474, "step": 34300 }, { "grad_norm": 0.4468707740306854, "learning_rate": 7.829668762021937e-05, "loss": 0.0535, "step": 34310 }, { "grad_norm": 0.5975221395492554, "learning_rate": 7.828305402942764e-05, "loss": 0.0489, "step": 34320 }, { "grad_norm": 0.7367064952850342, "learning_rate": 7.826941734564363e-05, "loss": 0.0389, "step": 34330 }, { "grad_norm": 0.5074065327644348, "learning_rate": 7.825577757035865e-05, "loss": 0.0411, "step": 34340 }, { "grad_norm": 0.3242857754230499, "learning_rate": 7.824213470506431e-05, "loss": 0.0412, "step": 34350 }, { "grad_norm": 0.4085344672203064, "learning_rate": 7.822848875125257e-05, "loss": 0.0476, "step": 34360 }, { "grad_norm": 0.34252694249153137, "learning_rate": 7.821483971041576e-05, "loss": 0.0554, "step": 34370 }, { "grad_norm": 0.4534715712070465, "learning_rate": 7.820118758404649e-05, "loss": 0.0451, "step": 34380 }, { "grad_norm": 0.49073195457458496, "learning_rate": 7.818753237363776e-05, "loss": 0.0487, "step": 34390 }, { "grad_norm": 0.5091300010681152, "learning_rate": 7.817387408068286e-05, "loss": 0.051, "step": 34400 }, { "grad_norm": 0.41975751519203186, "learning_rate": 7.816021270667544e-05, "loss": 0.0467, "step": 34410 }, { "grad_norm": 0.38788214325904846, "learning_rate": 7.81465482531095e-05, "loss": 0.0462, "step": 34420 }, { "grad_norm": 0.7282098531723022, "learning_rate": 7.813288072147938e-05, "loss": 0.0548, "step": 34430 }, { "grad_norm": 0.5506125092506409, "learning_rate": 7.811921011327972e-05, "loss": 0.0576, "step": 34440 }, { "grad_norm": 0.3342238664627075, "learning_rate": 7.810553643000549e-05, "loss": 0.0494, "step": 34450 }, { "grad_norm": 0.45616987347602844, "learning_rate": 7.809185967315206e-05, "loss": 0.0477, "step": 34460 }, { "grad_norm": 0.4841248095035553, "learning_rate": 7.80781798442151e-05, "loss": 0.0539, "step": 34470 }, { "grad_norm": 0.6028610467910767, "learning_rate": 7.806449694469061e-05, "loss": 0.0404, "step": 34480 }, { "grad_norm": 0.42462030053138733, "learning_rate": 7.805081097607492e-05, "loss": 0.0559, "step": 34490 }, { "grad_norm": 0.40196821093559265, "learning_rate": 7.803712193986474e-05, "loss": 0.0517, "step": 34500 }, { "grad_norm": 0.47365397214889526, "learning_rate": 7.802342983755702e-05, "loss": 0.0485, "step": 34510 }, { "grad_norm": 0.6081027984619141, "learning_rate": 7.800973467064918e-05, "loss": 0.0557, "step": 34520 }, { "grad_norm": 0.477933406829834, "learning_rate": 7.799603644063884e-05, "loss": 0.0454, "step": 34530 }, { "grad_norm": 0.3979645371437073, "learning_rate": 7.798233514902405e-05, "loss": 0.0432, "step": 34540 }, { "grad_norm": 0.280062198638916, "learning_rate": 7.796863079730318e-05, "loss": 0.0419, "step": 34550 }, { "grad_norm": 0.32316362857818604, "learning_rate": 7.795492338697488e-05, "loss": 0.0445, "step": 34560 }, { "grad_norm": 0.3598358929157257, "learning_rate": 7.794121291953819e-05, "loss": 0.0563, "step": 34570 }, { "grad_norm": 0.46619728207588196, "learning_rate": 7.792749939649246e-05, "loss": 0.057, "step": 34580 }, { "grad_norm": 0.33734557032585144, "learning_rate": 7.79137828193374e-05, "loss": 0.0325, "step": 34590 }, { "grad_norm": 0.5630620718002319, "learning_rate": 7.790006318957301e-05, "loss": 0.0548, "step": 34600 }, { "grad_norm": 0.355230450630188, "learning_rate": 7.788634050869965e-05, "loss": 0.0405, "step": 34610 }, { "grad_norm": 0.4793010652065277, "learning_rate": 7.787261477821803e-05, "loss": 0.0478, "step": 34620 }, { "grad_norm": 0.5398831367492676, "learning_rate": 7.785888599962916e-05, "loss": 0.0414, "step": 34630 }, { "grad_norm": 0.529009222984314, "learning_rate": 7.784515417443439e-05, "loss": 0.0454, "step": 34640 }, { "grad_norm": 0.8506942987442017, "learning_rate": 7.783141930413545e-05, "loss": 0.0439, "step": 34650 }, { "grad_norm": 0.374304860830307, "learning_rate": 7.78176813902343e-05, "loss": 0.0404, "step": 34660 }, { "grad_norm": 0.5379013419151306, "learning_rate": 7.780394043423336e-05, "loss": 0.0498, "step": 34670 }, { "grad_norm": 0.6178775429725647, "learning_rate": 7.77901964376353e-05, "loss": 0.0588, "step": 34680 }, { "grad_norm": 0.5868763327598572, "learning_rate": 7.777644940194316e-05, "loss": 0.051, "step": 34690 }, { "grad_norm": 0.5164398550987244, "learning_rate": 7.776269932866023e-05, "loss": 0.0476, "step": 34700 }, { "grad_norm": 0.4359200894832611, "learning_rate": 7.774894621929026e-05, "loss": 0.0523, "step": 34710 }, { "grad_norm": 0.3336486518383026, "learning_rate": 7.773519007533725e-05, "loss": 0.0554, "step": 34720 }, { "grad_norm": 0.4609352648258209, "learning_rate": 7.772143089830556e-05, "loss": 0.0466, "step": 34730 }, { "grad_norm": 0.3789648711681366, "learning_rate": 7.770766868969985e-05, "loss": 0.0499, "step": 34740 }, { "grad_norm": 0.3583551347255707, "learning_rate": 7.769390345102518e-05, "loss": 0.0477, "step": 34750 }, { "grad_norm": 0.44509705901145935, "learning_rate": 7.768013518378683e-05, "loss": 0.0445, "step": 34760 }, { "grad_norm": 0.33597204089164734, "learning_rate": 7.766636388949053e-05, "loss": 0.0398, "step": 34770 }, { "grad_norm": 0.2572876513004303, "learning_rate": 7.765258956964229e-05, "loss": 0.0617, "step": 34780 }, { "grad_norm": 0.44620609283447266, "learning_rate": 7.76388122257484e-05, "loss": 0.0549, "step": 34790 }, { "grad_norm": 0.3319796919822693, "learning_rate": 7.762503185931558e-05, "loss": 0.0394, "step": 34800 }, { "grad_norm": 0.7509305477142334, "learning_rate": 7.76112484718508e-05, "loss": 0.0492, "step": 34810 }, { "grad_norm": 0.3601037859916687, "learning_rate": 7.75974620648614e-05, "loss": 0.0538, "step": 34820 }, { "grad_norm": 0.44688716530799866, "learning_rate": 7.758367263985503e-05, "loss": 0.0381, "step": 34830 }, { "grad_norm": 0.4857219159603119, "learning_rate": 7.75698801983397e-05, "loss": 0.0417, "step": 34840 }, { "grad_norm": 0.544068455696106, "learning_rate": 7.755608474182372e-05, "loss": 0.0447, "step": 34850 }, { "grad_norm": 0.514374852180481, "learning_rate": 7.754228627181574e-05, "loss": 0.0551, "step": 34860 }, { "grad_norm": 0.29188376665115356, "learning_rate": 7.752848478982476e-05, "loss": 0.0478, "step": 34870 }, { "grad_norm": 0.4324628710746765, "learning_rate": 7.751468029736006e-05, "loss": 0.0451, "step": 34880 }, { "grad_norm": 0.6060751080513, "learning_rate": 7.750087279593129e-05, "loss": 0.0541, "step": 34890 }, { "grad_norm": 0.43202003836631775, "learning_rate": 7.748706228704843e-05, "loss": 0.0455, "step": 34900 }, { "grad_norm": 0.5313200354576111, "learning_rate": 7.747324877222176e-05, "loss": 0.0483, "step": 34910 }, { "grad_norm": 0.4131883382797241, "learning_rate": 7.745943225296188e-05, "loss": 0.0624, "step": 34920 }, { "grad_norm": 0.40019339323043823, "learning_rate": 7.744561273077981e-05, "loss": 0.0384, "step": 34930 }, { "grad_norm": 0.27813011407852173, "learning_rate": 7.743179020718678e-05, "loss": 0.0403, "step": 34940 }, { "grad_norm": 0.45210084319114685, "learning_rate": 7.741796468369443e-05, "loss": 0.0471, "step": 34950 }, { "grad_norm": 0.5150464773178101, "learning_rate": 7.740413616181466e-05, "loss": 0.0541, "step": 34960 }, { "grad_norm": 0.7580070495605469, "learning_rate": 7.739030464305978e-05, "loss": 0.0514, "step": 34970 }, { "grad_norm": 0.8241508603096008, "learning_rate": 7.737647012894235e-05, "loss": 0.065, "step": 34980 }, { "grad_norm": 0.4996550977230072, "learning_rate": 7.736263262097532e-05, "loss": 0.0537, "step": 34990 }, { "grad_norm": 0.4321005642414093, "learning_rate": 7.734879212067192e-05, "loss": 0.0525, "step": 35000 }, { "grad_norm": 0.40131112933158875, "learning_rate": 7.733494862954573e-05, "loss": 0.0452, "step": 35010 }, { "grad_norm": 0.36941322684288025, "learning_rate": 7.732110214911066e-05, "loss": 0.036, "step": 35020 }, { "grad_norm": 0.6387546062469482, "learning_rate": 7.730725268088094e-05, "loss": 0.0527, "step": 35030 }, { "grad_norm": 0.43486782908439636, "learning_rate": 7.729340022637112e-05, "loss": 0.044, "step": 35040 }, { "grad_norm": 0.5050111413002014, "learning_rate": 7.727954478709607e-05, "loss": 0.0433, "step": 35050 }, { "grad_norm": 0.9686017632484436, "learning_rate": 7.726568636457103e-05, "loss": 0.0501, "step": 35060 }, { "grad_norm": 0.46671825647354126, "learning_rate": 7.725182496031153e-05, "loss": 0.0489, "step": 35070 }, { "grad_norm": 0.5217548608779907, "learning_rate": 7.723796057583341e-05, "loss": 0.0588, "step": 35080 }, { "grad_norm": 0.4742969274520874, "learning_rate": 7.722409321265287e-05, "loss": 0.0486, "step": 35090 }, { "grad_norm": 0.4117817282676697, "learning_rate": 7.721022287228645e-05, "loss": 0.0563, "step": 35100 }, { "grad_norm": 0.48779040575027466, "learning_rate": 7.719634955625094e-05, "loss": 0.0436, "step": 35110 }, { "grad_norm": 0.3648832440376282, "learning_rate": 7.718247326606355e-05, "loss": 0.0368, "step": 35120 }, { "grad_norm": 0.44080811738967896, "learning_rate": 7.716859400324175e-05, "loss": 0.0441, "step": 35130 }, { "grad_norm": 0.4627735912799835, "learning_rate": 7.715471176930335e-05, "loss": 0.0698, "step": 35140 }, { "grad_norm": 0.395024836063385, "learning_rate": 7.714082656576651e-05, "loss": 0.0369, "step": 35150 }, { "grad_norm": 0.5617326498031616, "learning_rate": 7.712693839414968e-05, "loss": 0.0413, "step": 35160 }, { "grad_norm": 0.5042747855186462, "learning_rate": 7.711304725597164e-05, "loss": 0.0378, "step": 35170 }, { "grad_norm": 0.513579785823822, "learning_rate": 7.709915315275151e-05, "loss": 0.0484, "step": 35180 }, { "grad_norm": 0.4077295958995819, "learning_rate": 7.708525608600876e-05, "loss": 0.0412, "step": 35190 }, { "grad_norm": 0.4668717682361603, "learning_rate": 7.707135605726311e-05, "loss": 0.0455, "step": 35200 }, { "grad_norm": 0.33722084760665894, "learning_rate": 7.705745306803466e-05, "loss": 0.0697, "step": 35210 }, { "grad_norm": 0.5626246929168701, "learning_rate": 7.704354711984383e-05, "loss": 0.0405, "step": 35220 }, { "grad_norm": 0.3142634332180023, "learning_rate": 7.702963821421133e-05, "loss": 0.0386, "step": 35230 }, { "grad_norm": 0.3048192858695984, "learning_rate": 7.701572635265826e-05, "loss": 0.043, "step": 35240 }, { "grad_norm": 0.6261733174324036, "learning_rate": 7.700181153670596e-05, "loss": 0.0421, "step": 35250 }, { "grad_norm": 0.380493700504303, "learning_rate": 7.698789376787616e-05, "loss": 0.0544, "step": 35260 }, { "grad_norm": 1.023847222328186, "learning_rate": 7.697397304769085e-05, "loss": 0.0486, "step": 35270 }, { "grad_norm": 0.5489987134933472, "learning_rate": 7.696004937767241e-05, "loss": 0.0574, "step": 35280 }, { "grad_norm": 0.4879579246044159, "learning_rate": 7.694612275934352e-05, "loss": 0.0548, "step": 35290 }, { "grad_norm": 0.3479152023792267, "learning_rate": 7.693219319422714e-05, "loss": 0.0548, "step": 35300 }, { "grad_norm": 0.3365383744239807, "learning_rate": 7.69182606838466e-05, "loss": 0.0541, "step": 35310 }, { "grad_norm": 0.558617353439331, "learning_rate": 7.690432522972558e-05, "loss": 0.0523, "step": 35320 }, { "grad_norm": 0.2725426256656647, "learning_rate": 7.689038683338796e-05, "loss": 0.0454, "step": 35330 }, { "grad_norm": 0.5245867967605591, "learning_rate": 7.687644549635808e-05, "loss": 0.0512, "step": 35340 }, { "grad_norm": 0.4330216646194458, "learning_rate": 7.686250122016053e-05, "loss": 0.0416, "step": 35350 }, { "grad_norm": 0.5088164806365967, "learning_rate": 7.684855400632022e-05, "loss": 0.053, "step": 35360 }, { "grad_norm": 0.3747466206550598, "learning_rate": 7.683460385636243e-05, "loss": 0.0392, "step": 35370 }, { "grad_norm": 0.3983733057975769, "learning_rate": 7.68206507718127e-05, "loss": 0.0416, "step": 35380 }, { "grad_norm": 0.46837756037712097, "learning_rate": 7.680669475419693e-05, "loss": 0.0542, "step": 35390 }, { "grad_norm": 0.4982193410396576, "learning_rate": 7.679273580504132e-05, "loss": 0.0384, "step": 35400 }, { "grad_norm": 0.32480067014694214, "learning_rate": 7.67787739258724e-05, "loss": 0.0416, "step": 35410 }, { "grad_norm": 0.3701542019844055, "learning_rate": 7.676480911821705e-05, "loss": 0.0517, "step": 35420 }, { "grad_norm": 0.9435655474662781, "learning_rate": 7.675084138360238e-05, "loss": 0.0478, "step": 35430 }, { "grad_norm": 0.5199537873268127, "learning_rate": 7.673687072355592e-05, "loss": 0.0518, "step": 35440 }, { "grad_norm": 0.46694010496139526, "learning_rate": 7.67228971396055e-05, "loss": 0.0427, "step": 35450 }, { "grad_norm": 0.4407758116722107, "learning_rate": 7.670892063327922e-05, "loss": 0.0424, "step": 35460 }, { "grad_norm": 0.3262564539909363, "learning_rate": 7.669494120610552e-05, "loss": 0.0352, "step": 35470 }, { "grad_norm": 0.22321805357933044, "learning_rate": 7.66809588596132e-05, "loss": 0.0369, "step": 35480 }, { "grad_norm": 0.32900404930114746, "learning_rate": 7.666697359533132e-05, "loss": 0.0428, "step": 35490 }, { "grad_norm": 0.654274046421051, "learning_rate": 7.665298541478932e-05, "loss": 0.0458, "step": 35500 }, { "grad_norm": 0.35149675607681274, "learning_rate": 7.66389943195169e-05, "loss": 0.0419, "step": 35510 }, { "grad_norm": 0.3660120368003845, "learning_rate": 7.66250003110441e-05, "loss": 0.0442, "step": 35520 }, { "grad_norm": 0.38772591948509216, "learning_rate": 7.66110033909013e-05, "loss": 0.044, "step": 35530 }, { "grad_norm": 0.5738263726234436, "learning_rate": 7.659700356061918e-05, "loss": 0.0506, "step": 35540 }, { "grad_norm": 0.8388594388961792, "learning_rate": 7.658300082172875e-05, "loss": 0.0482, "step": 35550 }, { "grad_norm": 0.3714146614074707, "learning_rate": 7.65689951757613e-05, "loss": 0.0476, "step": 35560 }, { "grad_norm": 0.4191496670246124, "learning_rate": 7.655498662424849e-05, "loss": 0.045, "step": 35570 }, { "grad_norm": 0.6694394946098328, "learning_rate": 7.654097516872227e-05, "loss": 0.0433, "step": 35580 }, { "grad_norm": 0.4659855365753174, "learning_rate": 7.65269608107149e-05, "loss": 0.0539, "step": 35590 }, { "grad_norm": 0.6699754595756531, "learning_rate": 7.6512943551759e-05, "loss": 0.069, "step": 35600 }, { "grad_norm": 0.48414283990859985, "learning_rate": 7.649892339338743e-05, "loss": 0.0429, "step": 35610 }, { "grad_norm": 0.8192163109779358, "learning_rate": 7.648490033713344e-05, "loss": 0.0372, "step": 35620 }, { "grad_norm": 0.49945178627967834, "learning_rate": 7.647087438453058e-05, "loss": 0.0488, "step": 35630 }, { "grad_norm": 0.5203830003738403, "learning_rate": 7.64568455371127e-05, "loss": 0.0461, "step": 35640 }, { "grad_norm": 0.4040709435939789, "learning_rate": 7.644281379641396e-05, "loss": 0.0367, "step": 35650 }, { "grad_norm": 0.5318917036056519, "learning_rate": 7.642877916396887e-05, "loss": 0.041, "step": 35660 }, { "grad_norm": 0.41394028067588806, "learning_rate": 7.641474164131221e-05, "loss": 0.0549, "step": 35670 }, { "grad_norm": 0.3593783676624298, "learning_rate": 7.640070122997913e-05, "loss": 0.0367, "step": 35680 }, { "grad_norm": 0.4482988715171814, "learning_rate": 7.638665793150505e-05, "loss": 0.0323, "step": 35690 }, { "grad_norm": 0.4352227449417114, "learning_rate": 7.637261174742574e-05, "loss": 0.0434, "step": 35700 }, { "grad_norm": 0.3144766688346863, "learning_rate": 7.635856267927725e-05, "loss": 0.0328, "step": 35710 }, { "grad_norm": 0.3245919942855835, "learning_rate": 7.634451072859597e-05, "loss": 0.0376, "step": 35720 }, { "grad_norm": 0.40647298097610474, "learning_rate": 7.633045589691863e-05, "loss": 0.0293, "step": 35730 }, { "grad_norm": 0.2455597221851349, "learning_rate": 7.63163981857822e-05, "loss": 0.038, "step": 35740 }, { "grad_norm": 0.4842749834060669, "learning_rate": 7.630233759672403e-05, "loss": 0.037, "step": 35750 }, { "grad_norm": 0.5311565399169922, "learning_rate": 7.628827413128177e-05, "loss": 0.0522, "step": 35760 }, { "grad_norm": 0.6882854104042053, "learning_rate": 7.627420779099338e-05, "loss": 0.0458, "step": 35770 }, { "grad_norm": 0.8013772368431091, "learning_rate": 7.626013857739711e-05, "loss": 0.0436, "step": 35780 }, { "grad_norm": 0.3889625668525696, "learning_rate": 7.624606649203158e-05, "loss": 0.0354, "step": 35790 }, { "grad_norm": 0.5005040168762207, "learning_rate": 7.623199153643569e-05, "loss": 0.0406, "step": 35800 }, { "grad_norm": 0.563899576663971, "learning_rate": 7.621791371214863e-05, "loss": 0.0497, "step": 35810 }, { "grad_norm": 0.4142126441001892, "learning_rate": 7.620383302070995e-05, "loss": 0.0415, "step": 35820 }, { "grad_norm": 0.4210377037525177, "learning_rate": 7.61897494636595e-05, "loss": 0.0417, "step": 35830 }, { "grad_norm": 0.965909481048584, "learning_rate": 7.617566304253739e-05, "loss": 0.0437, "step": 35840 }, { "grad_norm": 0.43504437804222107, "learning_rate": 7.616157375888416e-05, "loss": 0.0447, "step": 35850 }, { "grad_norm": 0.4514963626861572, "learning_rate": 7.614748161424053e-05, "loss": 0.041, "step": 35860 }, { "grad_norm": 0.6891555190086365, "learning_rate": 7.613338661014763e-05, "loss": 0.0457, "step": 35870 }, { "grad_norm": 0.4702441096305847, "learning_rate": 7.611928874814686e-05, "loss": 0.0448, "step": 35880 }, { "grad_norm": 0.42191267013549805, "learning_rate": 7.610518802977996e-05, "loss": 0.0393, "step": 35890 }, { "grad_norm": 0.3873463273048401, "learning_rate": 7.609108445658893e-05, "loss": 0.0427, "step": 35900 }, { "grad_norm": 0.481886088848114, "learning_rate": 7.607697803011612e-05, "loss": 0.0538, "step": 35910 }, { "grad_norm": 0.31155651807785034, "learning_rate": 7.606286875190421e-05, "loss": 0.0436, "step": 35920 }, { "grad_norm": 1.1234698295593262, "learning_rate": 7.604875662349617e-05, "loss": 0.0371, "step": 35930 }, { "grad_norm": 0.34355273842811584, "learning_rate": 7.603464164643525e-05, "loss": 0.0486, "step": 35940 }, { "grad_norm": 0.4802764654159546, "learning_rate": 7.602052382226507e-05, "loss": 0.0467, "step": 35950 }, { "grad_norm": 0.4370604157447815, "learning_rate": 7.600640315252954e-05, "loss": 0.0453, "step": 35960 }, { "grad_norm": 0.6770819425582886, "learning_rate": 7.599227963877284e-05, "loss": 0.0553, "step": 35970 }, { "grad_norm": 0.6419390439987183, "learning_rate": 7.597815328253954e-05, "loss": 0.062, "step": 35980 }, { "grad_norm": 0.6919180750846863, "learning_rate": 7.596402408537443e-05, "loss": 0.0516, "step": 35990 }, { "grad_norm": 0.4612174928188324, "learning_rate": 7.59498920488227e-05, "loss": 0.057, "step": 36000 }, { "grad_norm": 0.3689418137073517, "learning_rate": 7.593575717442979e-05, "loss": 0.0383, "step": 36010 }, { "grad_norm": 0.32830071449279785, "learning_rate": 7.592161946374147e-05, "loss": 0.0442, "step": 36020 }, { "grad_norm": 0.3317253589630127, "learning_rate": 7.590747891830381e-05, "loss": 0.0493, "step": 36030 }, { "grad_norm": 0.5561344027519226, "learning_rate": 7.58933355396632e-05, "loss": 0.0428, "step": 36040 }, { "grad_norm": 0.4657820761203766, "learning_rate": 7.587918932936636e-05, "loss": 0.0498, "step": 36050 }, { "grad_norm": 0.3649466633796692, "learning_rate": 7.586504028896028e-05, "loss": 0.0507, "step": 36060 }, { "grad_norm": 0.4829563498497009, "learning_rate": 7.585088841999228e-05, "loss": 0.0458, "step": 36070 }, { "grad_norm": 0.5418773293495178, "learning_rate": 7.583673372400999e-05, "loss": 0.042, "step": 36080 }, { "grad_norm": 0.455058217048645, "learning_rate": 7.582257620256134e-05, "loss": 0.0445, "step": 36090 }, { "grad_norm": 0.6647145748138428, "learning_rate": 7.580841585719458e-05, "loss": 0.0446, "step": 36100 }, { "grad_norm": 0.8762398958206177, "learning_rate": 7.579425268945825e-05, "loss": 0.0437, "step": 36110 }, { "grad_norm": 0.30404970049858093, "learning_rate": 7.578008670090127e-05, "loss": 0.0332, "step": 36120 }, { "grad_norm": 0.3445846438407898, "learning_rate": 7.576591789307272e-05, "loss": 0.0439, "step": 36130 }, { "grad_norm": 0.28476962447166443, "learning_rate": 7.575174626752216e-05, "loss": 0.0363, "step": 36140 }, { "grad_norm": 0.366499125957489, "learning_rate": 7.573757182579934e-05, "loss": 0.0462, "step": 36150 }, { "grad_norm": 0.31217747926712036, "learning_rate": 7.572339456945435e-05, "loss": 0.0437, "step": 36160 }, { "grad_norm": 0.4757079482078552, "learning_rate": 7.570921450003762e-05, "loss": 0.0524, "step": 36170 }, { "grad_norm": 0.41766151785850525, "learning_rate": 7.569503161909983e-05, "loss": 0.0401, "step": 36180 }, { "grad_norm": 0.39752039313316345, "learning_rate": 7.568084592819202e-05, "loss": 0.0503, "step": 36190 }, { "grad_norm": 1.6192103624343872, "learning_rate": 7.566665742886551e-05, "loss": 0.0562, "step": 36200 }, { "grad_norm": 0.9200045466423035, "learning_rate": 7.565246612267194e-05, "loss": 0.0489, "step": 36210 }, { "grad_norm": 0.9103577136993408, "learning_rate": 7.563827201116325e-05, "loss": 0.0614, "step": 36220 }, { "grad_norm": 0.9670724272727966, "learning_rate": 7.562407509589164e-05, "loss": 0.0556, "step": 36230 }, { "grad_norm": 0.3984050452709198, "learning_rate": 7.560987537840973e-05, "loss": 0.0507, "step": 36240 }, { "grad_norm": 0.4433799982070923, "learning_rate": 7.559567286027036e-05, "loss": 0.0336, "step": 36250 }, { "grad_norm": 0.7794706225395203, "learning_rate": 7.558146754302668e-05, "loss": 0.0516, "step": 36260 }, { "grad_norm": 0.581031858921051, "learning_rate": 7.556725942823217e-05, "loss": 0.0447, "step": 36270 }, { "grad_norm": 0.36856165528297424, "learning_rate": 7.555304851744061e-05, "loss": 0.0459, "step": 36280 }, { "grad_norm": 0.3935468792915344, "learning_rate": 7.553883481220608e-05, "loss": 0.0622, "step": 36290 }, { "grad_norm": 0.2892267405986786, "learning_rate": 7.552461831408298e-05, "loss": 0.0399, "step": 36300 }, { "grad_norm": 0.4763151705265045, "learning_rate": 7.551039902462599e-05, "loss": 0.0415, "step": 36310 }, { "grad_norm": 0.3644315302371979, "learning_rate": 7.549617694539011e-05, "loss": 0.0321, "step": 36320 }, { "grad_norm": 0.4695347845554352, "learning_rate": 7.548195207793066e-05, "loss": 0.0466, "step": 36330 }, { "grad_norm": 0.5108466744422913, "learning_rate": 7.546772442380323e-05, "loss": 0.0409, "step": 36340 }, { "grad_norm": 0.5813077688217163, "learning_rate": 7.545349398456376e-05, "loss": 0.0549, "step": 36350 }, { "grad_norm": 0.3453129827976227, "learning_rate": 7.543926076176845e-05, "loss": 0.051, "step": 36360 }, { "grad_norm": 0.30370283126831055, "learning_rate": 7.542502475697385e-05, "loss": 0.0426, "step": 36370 }, { "grad_norm": 0.422250360250473, "learning_rate": 7.541078597173675e-05, "loss": 0.0537, "step": 36380 }, { "grad_norm": 0.6258126497268677, "learning_rate": 7.539654440761431e-05, "loss": 0.0499, "step": 36390 }, { "grad_norm": 0.4567083716392517, "learning_rate": 7.538230006616395e-05, "loss": 0.0463, "step": 36400 }, { "grad_norm": 0.2902889549732208, "learning_rate": 7.536805294894342e-05, "loss": 0.0597, "step": 36410 }, { "grad_norm": 0.5816670656204224, "learning_rate": 7.535380305751076e-05, "loss": 0.0415, "step": 36420 }, { "grad_norm": 0.36499136686325073, "learning_rate": 7.533955039342431e-05, "loss": 0.048, "step": 36430 }, { "grad_norm": 0.7687454223632812, "learning_rate": 7.532529495824274e-05, "loss": 0.0461, "step": 36440 }, { "grad_norm": 0.4940025210380554, "learning_rate": 7.531103675352497e-05, "loss": 0.0467, "step": 36450 }, { "grad_norm": 0.4251345694065094, "learning_rate": 7.529677578083029e-05, "loss": 0.0494, "step": 36460 }, { "grad_norm": 0.4977128803730011, "learning_rate": 7.528251204171823e-05, "loss": 0.0421, "step": 36470 }, { "grad_norm": 0.3160633146762848, "learning_rate": 7.526824553774867e-05, "loss": 0.0472, "step": 36480 }, { "grad_norm": 0.41500112414360046, "learning_rate": 7.525397627048178e-05, "loss": 0.0469, "step": 36490 }, { "grad_norm": 0.5283777117729187, "learning_rate": 7.523970424147802e-05, "loss": 0.0466, "step": 36500 }, { "grad_norm": 0.30057623982429504, "learning_rate": 7.522542945229813e-05, "loss": 0.0378, "step": 36510 }, { "grad_norm": 0.4217095971107483, "learning_rate": 7.521115190450321e-05, "loss": 0.0447, "step": 36520 }, { "grad_norm": 0.5598437190055847, "learning_rate": 7.519687159965462e-05, "loss": 0.0465, "step": 36530 }, { "grad_norm": 0.8915646076202393, "learning_rate": 7.518258853931403e-05, "loss": 0.0479, "step": 36540 }, { "grad_norm": 0.3992323577404022, "learning_rate": 7.516830272504342e-05, "loss": 0.0361, "step": 36550 }, { "grad_norm": 0.4340488314628601, "learning_rate": 7.515401415840509e-05, "loss": 0.0534, "step": 36560 }, { "grad_norm": 0.3221431076526642, "learning_rate": 7.513972284096155e-05, "loss": 0.0375, "step": 36570 }, { "grad_norm": 0.4314509630203247, "learning_rate": 7.512542877427574e-05, "loss": 0.0505, "step": 36580 }, { "grad_norm": 0.4394843578338623, "learning_rate": 7.51111319599108e-05, "loss": 0.0427, "step": 36590 }, { "grad_norm": 0.71127849817276, "learning_rate": 7.509683239943024e-05, "loss": 0.052, "step": 36600 }, { "grad_norm": 0.42825064063072205, "learning_rate": 7.50825300943978e-05, "loss": 0.0448, "step": 36610 }, { "grad_norm": 0.4733884632587433, "learning_rate": 7.50682250463776e-05, "loss": 0.0483, "step": 36620 }, { "grad_norm": 0.593928337097168, "learning_rate": 7.5053917256934e-05, "loss": 0.049, "step": 36630 }, { "grad_norm": 0.3086187243461609, "learning_rate": 7.503960672763166e-05, "loss": 0.0413, "step": 36640 }, { "grad_norm": 0.40053099393844604, "learning_rate": 7.502529346003559e-05, "loss": 0.0441, "step": 36650 }, { "grad_norm": 0.3264894485473633, "learning_rate": 7.501097745571106e-05, "loss": 0.05, "step": 36660 }, { "grad_norm": 0.34779390692710876, "learning_rate": 7.499665871622361e-05, "loss": 0.0518, "step": 36670 }, { "grad_norm": 0.5828514695167542, "learning_rate": 7.498233724313917e-05, "loss": 0.0511, "step": 36680 }, { "grad_norm": 0.35704389214515686, "learning_rate": 7.496801303802387e-05, "loss": 0.0559, "step": 36690 }, { "grad_norm": 0.46394386887550354, "learning_rate": 7.495368610244423e-05, "loss": 0.0452, "step": 36700 }, { "grad_norm": 0.41587045788764954, "learning_rate": 7.493935643796697e-05, "loss": 0.0515, "step": 36710 }, { "grad_norm": 0.5328904390335083, "learning_rate": 7.49250240461592e-05, "loss": 0.0475, "step": 36720 }, { "grad_norm": 0.4085903465747833, "learning_rate": 7.491068892858826e-05, "loss": 0.0445, "step": 36730 }, { "grad_norm": 0.3777308464050293, "learning_rate": 7.489635108682184e-05, "loss": 0.0392, "step": 36740 }, { "grad_norm": 0.48979538679122925, "learning_rate": 7.48820105224279e-05, "loss": 0.0389, "step": 36750 }, { "grad_norm": 0.5087663531303406, "learning_rate": 7.48676672369747e-05, "loss": 0.0449, "step": 36760 }, { "grad_norm": 0.5275149941444397, "learning_rate": 7.485332123203079e-05, "loss": 0.0454, "step": 36770 }, { "grad_norm": 0.4070615768432617, "learning_rate": 7.483897250916504e-05, "loss": 0.0498, "step": 36780 }, { "grad_norm": 0.5579041242599487, "learning_rate": 7.48246210699466e-05, "loss": 0.0473, "step": 36790 }, { "grad_norm": 0.5384296178817749, "learning_rate": 7.481026691594492e-05, "loss": 0.0426, "step": 36800 }, { "grad_norm": 0.5441387891769409, "learning_rate": 7.479591004872975e-05, "loss": 0.0383, "step": 36810 }, { "grad_norm": 0.3658669590950012, "learning_rate": 7.478155046987114e-05, "loss": 0.0486, "step": 36820 }, { "grad_norm": 0.48361736536026, "learning_rate": 7.476718818093944e-05, "loss": 0.0408, "step": 36830 }, { "grad_norm": 0.3037588596343994, "learning_rate": 7.475282318350527e-05, "loss": 0.0448, "step": 36840 }, { "grad_norm": 0.3651021122932434, "learning_rate": 7.473845547913957e-05, "loss": 0.0302, "step": 36850 }, { "grad_norm": 0.4382484257221222, "learning_rate": 7.472408506941357e-05, "loss": 0.053, "step": 36860 }, { "grad_norm": 0.3856504559516907, "learning_rate": 7.470971195589881e-05, "loss": 0.0402, "step": 36870 }, { "grad_norm": 0.7421706914901733, "learning_rate": 7.46953361401671e-05, "loss": 0.054, "step": 36880 }, { "grad_norm": 0.4605638384819031, "learning_rate": 7.468095762379055e-05, "loss": 0.0368, "step": 36890 }, { "grad_norm": 0.3104877769947052, "learning_rate": 7.466657640834158e-05, "loss": 0.0474, "step": 36900 }, { "grad_norm": 0.4154256582260132, "learning_rate": 7.46521924953929e-05, "loss": 0.046, "step": 36910 }, { "grad_norm": 0.4421171247959137, "learning_rate": 7.463780588651752e-05, "loss": 0.0419, "step": 36920 }, { "grad_norm": 0.7372615337371826, "learning_rate": 7.462341658328873e-05, "loss": 0.0507, "step": 36930 }, { "grad_norm": 0.3772355616092682, "learning_rate": 7.460902458728012e-05, "loss": 0.0501, "step": 36940 }, { "grad_norm": 0.3844568729400635, "learning_rate": 7.459462990006558e-05, "loss": 0.0435, "step": 36950 }, { "grad_norm": 0.4470570683479309, "learning_rate": 7.458023252321929e-05, "loss": 0.0448, "step": 36960 }, { "grad_norm": 0.4363507926464081, "learning_rate": 7.456583245831573e-05, "loss": 0.0366, "step": 36970 }, { "grad_norm": 0.5501168370246887, "learning_rate": 7.455142970692968e-05, "loss": 0.0553, "step": 36980 }, { "grad_norm": 0.4083021581172943, "learning_rate": 7.453702427063615e-05, "loss": 0.04, "step": 36990 }, { "grad_norm": 0.4334501028060913, "learning_rate": 7.452261615101057e-05, "loss": 0.0443, "step": 37000 }, { "grad_norm": 0.3200179636478424, "learning_rate": 7.450820534962856e-05, "loss": 0.0406, "step": 37010 }, { "grad_norm": 0.6563965082168579, "learning_rate": 7.449379186806604e-05, "loss": 0.0426, "step": 37020 }, { "grad_norm": 0.4683776795864105, "learning_rate": 7.44793757078993e-05, "loss": 0.049, "step": 37030 }, { "grad_norm": 0.49645712971687317, "learning_rate": 7.44649568707048e-05, "loss": 0.0449, "step": 37040 }, { "grad_norm": 0.38625362515449524, "learning_rate": 7.445053535805942e-05, "loss": 0.0535, "step": 37050 }, { "grad_norm": 0.5917053818702698, "learning_rate": 7.443611117154024e-05, "loss": 0.0477, "step": 37060 }, { "grad_norm": 1.012858510017395, "learning_rate": 7.442168431272469e-05, "loss": 0.0497, "step": 37070 }, { "grad_norm": 0.5419219732284546, "learning_rate": 7.440725478319044e-05, "loss": 0.0483, "step": 37080 }, { "grad_norm": 0.5641822814941406, "learning_rate": 7.439282258451553e-05, "loss": 0.0519, "step": 37090 }, { "grad_norm": 0.4069194197654724, "learning_rate": 7.43783877182782e-05, "loss": 0.0426, "step": 37100 }, { "grad_norm": 0.371276319026947, "learning_rate": 7.436395018605707e-05, "loss": 0.0475, "step": 37110 }, { "grad_norm": 0.4026225805282593, "learning_rate": 7.434950998943094e-05, "loss": 0.0401, "step": 37120 }, { "grad_norm": 0.410076379776001, "learning_rate": 7.433506712997902e-05, "loss": 0.0446, "step": 37130 }, { "grad_norm": 0.41592949628829956, "learning_rate": 7.432062160928077e-05, "loss": 0.0581, "step": 37140 }, { "grad_norm": 0.31833839416503906, "learning_rate": 7.430617342891588e-05, "loss": 0.0453, "step": 37150 }, { "grad_norm": 0.34640198945999146, "learning_rate": 7.429172259046442e-05, "loss": 0.0443, "step": 37160 }, { "grad_norm": 0.4921824634075165, "learning_rate": 7.42772690955067e-05, "loss": 0.0448, "step": 37170 }, { "grad_norm": 0.3952535092830658, "learning_rate": 7.426281294562332e-05, "loss": 0.0539, "step": 37180 }, { "grad_norm": 0.3521867096424103, "learning_rate": 7.424835414239519e-05, "loss": 0.042, "step": 37190 }, { "grad_norm": 0.4363027513027191, "learning_rate": 7.423389268740352e-05, "loss": 0.039, "step": 37200 }, { "grad_norm": 0.6592859625816345, "learning_rate": 7.42194285822298e-05, "loss": 0.0384, "step": 37210 }, { "grad_norm": 0.31984326243400574, "learning_rate": 7.420496182845576e-05, "loss": 0.0432, "step": 37220 }, { "grad_norm": 0.3717101216316223, "learning_rate": 7.419049242766348e-05, "loss": 0.0477, "step": 37230 }, { "grad_norm": 0.5513726472854614, "learning_rate": 7.417602038143533e-05, "loss": 0.0521, "step": 37240 }, { "grad_norm": 0.5600325465202332, "learning_rate": 7.416154569135393e-05, "loss": 0.0579, "step": 37250 }, { "grad_norm": 0.4848538637161255, "learning_rate": 7.414706835900222e-05, "loss": 0.0387, "step": 37260 }, { "grad_norm": 0.4292772114276886, "learning_rate": 7.413258838596345e-05, "loss": 0.0439, "step": 37270 }, { "grad_norm": 0.518083393573761, "learning_rate": 7.411810577382106e-05, "loss": 0.0477, "step": 37280 }, { "grad_norm": 0.4653034806251526, "learning_rate": 7.410362052415889e-05, "loss": 0.0454, "step": 37290 }, { "grad_norm": 0.3615429103374481, "learning_rate": 7.408913263856102e-05, "loss": 0.0381, "step": 37300 }, { "grad_norm": 0.4695446193218231, "learning_rate": 7.407464211861183e-05, "loss": 0.0468, "step": 37310 }, { "grad_norm": 0.37920013070106506, "learning_rate": 7.406014896589597e-05, "loss": 0.0436, "step": 37320 }, { "grad_norm": 0.3289685547351837, "learning_rate": 7.404565318199838e-05, "loss": 0.0383, "step": 37330 }, { "grad_norm": 0.36694157123565674, "learning_rate": 7.403115476850432e-05, "loss": 0.0646, "step": 37340 }, { "grad_norm": 0.3617364168167114, "learning_rate": 7.401665372699932e-05, "loss": 0.0464, "step": 37350 }, { "grad_norm": 0.5982478857040405, "learning_rate": 7.400215005906916e-05, "loss": 0.0423, "step": 37360 }, { "grad_norm": 0.4079420864582062, "learning_rate": 7.398764376629996e-05, "loss": 0.0446, "step": 37370 }, { "grad_norm": 0.40775808691978455, "learning_rate": 7.39731348502781e-05, "loss": 0.0427, "step": 37380 }, { "grad_norm": 0.5785626769065857, "learning_rate": 7.395862331259028e-05, "loss": 0.0411, "step": 37390 }, { "grad_norm": 0.5994827151298523, "learning_rate": 7.394410915482341e-05, "loss": 0.0491, "step": 37400 }, { "grad_norm": 0.38842010498046875, "learning_rate": 7.392959237856476e-05, "loss": 0.0459, "step": 37410 }, { "grad_norm": 0.426035076379776, "learning_rate": 7.391507298540187e-05, "loss": 0.0499, "step": 37420 }, { "grad_norm": 0.8785347938537598, "learning_rate": 7.390055097692257e-05, "loss": 0.0414, "step": 37430 }, { "grad_norm": 0.4430907070636749, "learning_rate": 7.388602635471494e-05, "loss": 0.0615, "step": 37440 }, { "grad_norm": 0.5206460356712341, "learning_rate": 7.387149912036738e-05, "loss": 0.0528, "step": 37450 }, { "grad_norm": 0.3666236996650696, "learning_rate": 7.385696927546854e-05, "loss": 0.0466, "step": 37460 }, { "grad_norm": 0.45027729868888855, "learning_rate": 7.384243682160743e-05, "loss": 0.071, "step": 37470 }, { "grad_norm": 0.5911464095115662, "learning_rate": 7.382790176037329e-05, "loss": 0.048, "step": 37480 }, { "grad_norm": 0.4288429319858551, "learning_rate": 7.381336409335562e-05, "loss": 0.0464, "step": 37490 }, { "grad_norm": 0.4974435865879059, "learning_rate": 7.379882382214426e-05, "loss": 0.0516, "step": 37500 }, { "grad_norm": 0.2986246347427368, "learning_rate": 7.378428094832931e-05, "loss": 0.0467, "step": 37510 }, { "grad_norm": 0.5338402390480042, "learning_rate": 7.376973547350114e-05, "loss": 0.0546, "step": 37520 }, { "grad_norm": 0.5814549326896667, "learning_rate": 7.375518739925046e-05, "loss": 0.0451, "step": 37530 }, { "grad_norm": 0.4337851405143738, "learning_rate": 7.374063672716818e-05, "loss": 0.0406, "step": 37540 }, { "grad_norm": 0.7453320026397705, "learning_rate": 7.372608345884558e-05, "loss": 0.0468, "step": 37550 }, { "grad_norm": 0.6201068162918091, "learning_rate": 7.371152759587415e-05, "loss": 0.0428, "step": 37560 }, { "grad_norm": 0.45263054966926575, "learning_rate": 7.369696913984572e-05, "loss": 0.0377, "step": 37570 }, { "grad_norm": 0.22348405420780182, "learning_rate": 7.368240809235237e-05, "loss": 0.0442, "step": 37580 }, { "grad_norm": 0.6154280304908752, "learning_rate": 7.366784445498649e-05, "loss": 0.0464, "step": 37590 }, { "grad_norm": 0.3400416374206543, "learning_rate": 7.36532782293407e-05, "loss": 0.0375, "step": 37600 }, { "grad_norm": 0.703110933303833, "learning_rate": 7.363870941700797e-05, "loss": 0.043, "step": 37610 }, { "grad_norm": 0.35794201493263245, "learning_rate": 7.362413801958152e-05, "loss": 0.0415, "step": 37620 }, { "grad_norm": 0.45819082856178284, "learning_rate": 7.360956403865486e-05, "loss": 0.0489, "step": 37630 }, { "grad_norm": 0.49413520097732544, "learning_rate": 7.359498747582177e-05, "loss": 0.0544, "step": 37640 }, { "grad_norm": 0.6117276549339294, "learning_rate": 7.358040833267634e-05, "loss": 0.0397, "step": 37650 }, { "grad_norm": 0.5000094175338745, "learning_rate": 7.356582661081288e-05, "loss": 0.0475, "step": 37660 }, { "grad_norm": 0.4784153699874878, "learning_rate": 7.355124231182607e-05, "loss": 0.035, "step": 37670 }, { "grad_norm": 0.4191991984844208, "learning_rate": 7.353665543731083e-05, "loss": 0.0405, "step": 37680 }, { "grad_norm": 0.335695743560791, "learning_rate": 7.352206598886229e-05, "loss": 0.0451, "step": 37690 }, { "grad_norm": 1.3107972145080566, "learning_rate": 7.350747396807601e-05, "loss": 0.0451, "step": 37700 }, { "grad_norm": 0.9391629695892334, "learning_rate": 7.349287937654772e-05, "loss": 0.0433, "step": 37710 }, { "grad_norm": 0.38168227672576904, "learning_rate": 7.347828221587345e-05, "loss": 0.0534, "step": 37720 }, { "grad_norm": 0.35487619042396545, "learning_rate": 7.346368248764955e-05, "loss": 0.0416, "step": 37730 }, { "grad_norm": 0.580294668674469, "learning_rate": 7.34490801934726e-05, "loss": 0.045, "step": 37740 }, { "grad_norm": 0.35617077350616455, "learning_rate": 7.343447533493947e-05, "loss": 0.0599, "step": 37750 }, { "grad_norm": 0.4789125919342041, "learning_rate": 7.341986791364739e-05, "loss": 0.0691, "step": 37760 }, { "grad_norm": 0.33154648542404175, "learning_rate": 7.340525793119374e-05, "loss": 0.0412, "step": 37770 }, { "grad_norm": 0.4692024886608124, "learning_rate": 7.33906453891763e-05, "loss": 0.0465, "step": 37780 }, { "grad_norm": 0.8053243160247803, "learning_rate": 7.337603028919303e-05, "loss": 0.0514, "step": 37790 }, { "grad_norm": 0.2716924548149109, "learning_rate": 7.336141263284225e-05, "loss": 0.0455, "step": 37800 }, { "grad_norm": 1.1312686204910278, "learning_rate": 7.334679242172249e-05, "loss": 0.0524, "step": 37810 }, { "grad_norm": 0.5044308304786682, "learning_rate": 7.33321696574326e-05, "loss": 0.048, "step": 37820 }, { "grad_norm": 0.3899661898612976, "learning_rate": 7.331754434157174e-05, "loss": 0.0444, "step": 37830 }, { "grad_norm": 0.36853086948394775, "learning_rate": 7.330291647573928e-05, "loss": 0.045, "step": 37840 }, { "grad_norm": 0.6696220636367798, "learning_rate": 7.32882860615349e-05, "loss": 0.0347, "step": 37850 }, { "grad_norm": 0.5148434638977051, "learning_rate": 7.327365310055858e-05, "loss": 0.0465, "step": 37860 }, { "grad_norm": 0.4446312487125397, "learning_rate": 7.325901759441055e-05, "loss": 0.0499, "step": 37870 }, { "grad_norm": 0.5725439786911011, "learning_rate": 7.324437954469132e-05, "loss": 0.0425, "step": 37880 }, { "grad_norm": 0.4586324691772461, "learning_rate": 7.32297389530017e-05, "loss": 0.039, "step": 37890 }, { "grad_norm": 0.30107539892196655, "learning_rate": 7.321509582094274e-05, "loss": 0.0416, "step": 37900 }, { "grad_norm": 0.30829286575317383, "learning_rate": 7.320045015011582e-05, "loss": 0.0469, "step": 37910 }, { "grad_norm": 0.3838062882423401, "learning_rate": 7.318580194212253e-05, "loss": 0.0413, "step": 37920 }, { "grad_norm": 0.3092803955078125, "learning_rate": 7.317115119856483e-05, "loss": 0.0402, "step": 37930 }, { "grad_norm": 0.5794987678527832, "learning_rate": 7.315649792104485e-05, "loss": 0.0396, "step": 37940 }, { "grad_norm": 0.6319827437400818, "learning_rate": 7.314184211116507e-05, "loss": 0.0504, "step": 37950 }, { "grad_norm": 0.373704731464386, "learning_rate": 7.312718377052823e-05, "loss": 0.0515, "step": 37960 }, { "grad_norm": 0.5036719441413879, "learning_rate": 7.311252290073737e-05, "loss": 0.0449, "step": 37970 }, { "grad_norm": 0.31165605783462524, "learning_rate": 7.309785950339572e-05, "loss": 0.0388, "step": 37980 }, { "grad_norm": 0.4063430428504944, "learning_rate": 7.308319358010692e-05, "loss": 0.0394, "step": 37990 }, { "grad_norm": 0.4977363348007202, "learning_rate": 7.306852513247474e-05, "loss": 0.064, "step": 38000 }, { "grad_norm": 0.5929417014122009, "learning_rate": 7.305385416210336e-05, "loss": 0.0529, "step": 38010 }, { "grad_norm": 0.50341796875, "learning_rate": 7.303918067059711e-05, "loss": 0.0375, "step": 38020 }, { "grad_norm": 0.3058265149593353, "learning_rate": 7.302450465956075e-05, "loss": 0.0449, "step": 38030 }, { "grad_norm": 0.2417764514684677, "learning_rate": 7.300982613059914e-05, "loss": 0.0406, "step": 38040 }, { "grad_norm": 0.6604036688804626, "learning_rate": 7.299514508531757e-05, "loss": 0.0584, "step": 38050 }, { "grad_norm": 0.4666403532028198, "learning_rate": 7.298046152532148e-05, "loss": 0.043, "step": 38060 }, { "grad_norm": 0.5900126099586487, "learning_rate": 7.296577545221668e-05, "loss": 0.0437, "step": 38070 }, { "grad_norm": 0.7391713261604309, "learning_rate": 7.29510868676092e-05, "loss": 0.0495, "step": 38080 }, { "grad_norm": 0.49398621916770935, "learning_rate": 7.293639577310538e-05, "loss": 0.0376, "step": 38090 }, { "grad_norm": 0.4831109046936035, "learning_rate": 7.292170217031179e-05, "loss": 0.0442, "step": 38100 }, { "grad_norm": 0.3536697030067444, "learning_rate": 7.290700606083532e-05, "loss": 0.0479, "step": 38110 }, { "grad_norm": 0.446317583322525, "learning_rate": 7.28923074462831e-05, "loss": 0.0444, "step": 38120 }, { "grad_norm": 0.3440845310688019, "learning_rate": 7.287760632826255e-05, "loss": 0.0425, "step": 38130 }, { "grad_norm": 0.5084020495414734, "learning_rate": 7.286290270838139e-05, "loss": 0.0542, "step": 38140 }, { "grad_norm": 0.47669342160224915, "learning_rate": 7.284819658824756e-05, "loss": 0.042, "step": 38150 }, { "grad_norm": 0.37687575817108154, "learning_rate": 7.283348796946929e-05, "loss": 0.05, "step": 38160 }, { "grad_norm": 0.4186123013496399, "learning_rate": 7.28187768536551e-05, "loss": 0.0418, "step": 38170 }, { "grad_norm": 0.630048394203186, "learning_rate": 7.280406324241379e-05, "loss": 0.0452, "step": 38180 }, { "grad_norm": 0.2958378493785858, "learning_rate": 7.278934713735441e-05, "loss": 0.0482, "step": 38190 }, { "grad_norm": 0.3511410057544708, "learning_rate": 7.277462854008629e-05, "loss": 0.0415, "step": 38200 }, { "grad_norm": 0.4702378809452057, "learning_rate": 7.275990745221904e-05, "loss": 0.0414, "step": 38210 }, { "grad_norm": 0.3700326383113861, "learning_rate": 7.274518387536252e-05, "loss": 0.0478, "step": 38220 }, { "grad_norm": 0.38496506214141846, "learning_rate": 7.273045781112689e-05, "loss": 0.0499, "step": 38230 }, { "grad_norm": 0.32455042004585266, "learning_rate": 7.271572926112257e-05, "loss": 0.0408, "step": 38240 }, { "grad_norm": 0.40495073795318604, "learning_rate": 7.270099822696024e-05, "loss": 0.0452, "step": 38250 }, { "grad_norm": 0.2968664765357971, "learning_rate": 7.26862647102509e-05, "loss": 0.045, "step": 38260 }, { "grad_norm": 0.31666988134384155, "learning_rate": 7.267152871260573e-05, "loss": 0.049, "step": 38270 }, { "grad_norm": 0.4320129156112671, "learning_rate": 7.265679023563629e-05, "loss": 0.0505, "step": 38280 }, { "grad_norm": 0.35772955417633057, "learning_rate": 7.264204928095433e-05, "loss": 0.0487, "step": 38290 }, { "grad_norm": 0.47284257411956787, "learning_rate": 7.262730585017188e-05, "loss": 0.0463, "step": 38300 }, { "grad_norm": 0.5207986831665039, "learning_rate": 7.261255994490131e-05, "loss": 0.0405, "step": 38310 }, { "grad_norm": 0.7006728053092957, "learning_rate": 7.259781156675517e-05, "loss": 0.0487, "step": 38320 }, { "grad_norm": 0.512993574142456, "learning_rate": 7.258306071734633e-05, "loss": 0.0381, "step": 38330 }, { "grad_norm": 0.40988245606422424, "learning_rate": 7.256830739828792e-05, "loss": 0.0449, "step": 38340 }, { "grad_norm": 0.5071316957473755, "learning_rate": 7.255355161119336e-05, "loss": 0.0341, "step": 38350 }, { "grad_norm": 0.9509270191192627, "learning_rate": 7.253879335767628e-05, "loss": 0.0394, "step": 38360 }, { "grad_norm": 0.3551611304283142, "learning_rate": 7.252403263935067e-05, "loss": 0.054, "step": 38370 }, { "grad_norm": 0.46374717354774475, "learning_rate": 7.25092694578307e-05, "loss": 0.0486, "step": 38380 }, { "grad_norm": 0.5041197538375854, "learning_rate": 7.249450381473087e-05, "loss": 0.0474, "step": 38390 }, { "grad_norm": 0.3176611065864563, "learning_rate": 7.247973571166593e-05, "loss": 0.0506, "step": 38400 }, { "grad_norm": 0.6948443055152893, "learning_rate": 7.246496515025089e-05, "loss": 0.0499, "step": 38410 }, { "grad_norm": 0.5648357272148132, "learning_rate": 7.245019213210101e-05, "loss": 0.0507, "step": 38420 }, { "grad_norm": 0.36591848731040955, "learning_rate": 7.24354166588319e-05, "loss": 0.0519, "step": 38430 }, { "grad_norm": 0.5294848680496216, "learning_rate": 7.242063873205935e-05, "loss": 0.0542, "step": 38440 }, { "grad_norm": 0.49745213985443115, "learning_rate": 7.240585835339946e-05, "loss": 0.0596, "step": 38450 }, { "grad_norm": 0.39577624201774597, "learning_rate": 7.239107552446857e-05, "loss": 0.0507, "step": 38460 }, { "grad_norm": 0.483630508184433, "learning_rate": 7.237629024688334e-05, "loss": 0.0418, "step": 38470 }, { "grad_norm": 0.3686661124229431, "learning_rate": 7.236150252226064e-05, "loss": 0.0475, "step": 38480 }, { "grad_norm": 0.2736222743988037, "learning_rate": 7.234671235221765e-05, "loss": 0.0514, "step": 38490 }, { "grad_norm": 0.6300815939903259, "learning_rate": 7.233191973837179e-05, "loss": 0.047, "step": 38500 }, { "grad_norm": 0.3482629954814911, "learning_rate": 7.231712468234078e-05, "loss": 0.0403, "step": 38510 }, { "grad_norm": 0.42335769534111023, "learning_rate": 7.230232718574254e-05, "loss": 0.0544, "step": 38520 }, { "grad_norm": 0.42478373646736145, "learning_rate": 7.228752725019535e-05, "loss": 0.0442, "step": 38530 }, { "grad_norm": 0.3383656144142151, "learning_rate": 7.227272487731769e-05, "loss": 0.0512, "step": 38540 }, { "grad_norm": 0.4620007574558258, "learning_rate": 7.225792006872831e-05, "loss": 0.0401, "step": 38550 }, { "grad_norm": 0.5323630571365356, "learning_rate": 7.224311282604628e-05, "loss": 0.0424, "step": 38560 }, { "grad_norm": 0.6138981580734253, "learning_rate": 7.222830315089085e-05, "loss": 0.0474, "step": 38570 }, { "grad_norm": 0.6843537092208862, "learning_rate": 7.22134910448816e-05, "loss": 0.046, "step": 38580 }, { "grad_norm": 1.2815994024276733, "learning_rate": 7.219867650963839e-05, "loss": 0.0435, "step": 38590 }, { "grad_norm": 0.38181984424591064, "learning_rate": 7.21838595467813e-05, "loss": 0.033, "step": 38600 }, { "grad_norm": 0.3564615249633789, "learning_rate": 7.216904015793067e-05, "loss": 0.0499, "step": 38610 }, { "grad_norm": 1.0066194534301758, "learning_rate": 7.215421834470713e-05, "loss": 0.0409, "step": 38620 }, { "grad_norm": 0.40756574273109436, "learning_rate": 7.213939410873158e-05, "loss": 0.0488, "step": 38630 }, { "grad_norm": 0.8178060054779053, "learning_rate": 7.212456745162519e-05, "loss": 0.05, "step": 38640 }, { "grad_norm": 0.48409923911094666, "learning_rate": 7.210973837500937e-05, "loss": 0.0496, "step": 38650 }, { "grad_norm": 0.3964793086051941, "learning_rate": 7.209490688050578e-05, "loss": 0.0402, "step": 38660 }, { "grad_norm": 0.42548006772994995, "learning_rate": 7.208007296973641e-05, "loss": 0.0411, "step": 38670 }, { "grad_norm": 0.47601252794265747, "learning_rate": 7.206523664432345e-05, "loss": 0.0567, "step": 38680 }, { "grad_norm": 0.5827255249023438, "learning_rate": 7.205039790588939e-05, "loss": 0.0402, "step": 38690 }, { "grad_norm": 0.49514859914779663, "learning_rate": 7.203555675605697e-05, "loss": 0.0425, "step": 38700 }, { "grad_norm": 0.42560455203056335, "learning_rate": 7.202071319644917e-05, "loss": 0.0455, "step": 38710 }, { "grad_norm": 0.4783395230770111, "learning_rate": 7.200586722868932e-05, "loss": 0.0489, "step": 38720 }, { "grad_norm": 0.3527579605579376, "learning_rate": 7.199101885440088e-05, "loss": 0.0571, "step": 38730 }, { "grad_norm": 0.3770718276500702, "learning_rate": 7.197616807520771e-05, "loss": 0.0492, "step": 38740 }, { "grad_norm": 0.4474594295024872, "learning_rate": 7.196131489273381e-05, "loss": 0.0395, "step": 38750 }, { "grad_norm": 0.3552720546722412, "learning_rate": 7.194645930860355e-05, "loss": 0.0388, "step": 38760 }, { "grad_norm": 0.8287341594696045, "learning_rate": 7.19316013244415e-05, "loss": 0.0403, "step": 38770 }, { "grad_norm": 0.44797244668006897, "learning_rate": 7.191674094187248e-05, "loss": 0.0447, "step": 38780 }, { "grad_norm": 0.6129283308982849, "learning_rate": 7.190187816252165e-05, "loss": 0.0393, "step": 38790 }, { "grad_norm": 0.40870681405067444, "learning_rate": 7.188701298801435e-05, "loss": 0.049, "step": 38800 }, { "grad_norm": 0.3719940781593323, "learning_rate": 7.18721454199762e-05, "loss": 0.0438, "step": 38810 }, { "grad_norm": 0.6085062026977539, "learning_rate": 7.185727546003312e-05, "loss": 0.0454, "step": 38820 }, { "grad_norm": 0.5018340945243835, "learning_rate": 7.184240310981126e-05, "loss": 0.0392, "step": 38830 }, { "grad_norm": 0.40380531549453735, "learning_rate": 7.182752837093704e-05, "loss": 0.0436, "step": 38840 }, { "grad_norm": 0.4289011061191559, "learning_rate": 7.181265124503711e-05, "loss": 0.0633, "step": 38850 }, { "grad_norm": 0.41776180267333984, "learning_rate": 7.179777173373847e-05, "loss": 0.0462, "step": 38860 }, { "grad_norm": 0.6629455089569092, "learning_rate": 7.178288983866826e-05, "loss": 0.0412, "step": 38870 }, { "grad_norm": 0.4517689049243927, "learning_rate": 7.176800556145397e-05, "loss": 0.052, "step": 38880 }, { "grad_norm": 0.44038617610931396, "learning_rate": 7.175311890372334e-05, "loss": 0.0509, "step": 38890 }, { "grad_norm": 0.5159400701522827, "learning_rate": 7.17382298671043e-05, "loss": 0.0471, "step": 38900 }, { "grad_norm": 0.4683067798614502, "learning_rate": 7.172333845322515e-05, "loss": 0.0401, "step": 38910 }, { "grad_norm": 0.2927684187889099, "learning_rate": 7.170844466371436e-05, "loss": 0.0524, "step": 38920 }, { "grad_norm": 0.4286436438560486, "learning_rate": 7.16935485002007e-05, "loss": 0.0433, "step": 38930 }, { "grad_norm": 0.5311933755874634, "learning_rate": 7.167864996431319e-05, "loss": 0.0504, "step": 38940 }, { "grad_norm": 0.5308822989463806, "learning_rate": 7.166374905768111e-05, "loss": 0.0409, "step": 38950 }, { "grad_norm": 0.366020530462265, "learning_rate": 7.164884578193401e-05, "loss": 0.0332, "step": 38960 }, { "grad_norm": 0.43876150250434875, "learning_rate": 7.163394013870168e-05, "loss": 0.044, "step": 38970 }, { "grad_norm": 0.6802164912223816, "learning_rate": 7.161903212961419e-05, "loss": 0.0505, "step": 38980 }, { "grad_norm": 0.5441470146179199, "learning_rate": 7.160412175630183e-05, "loss": 0.0418, "step": 38990 }, { "grad_norm": 0.36137324571609497, "learning_rate": 7.158920902039521e-05, "loss": 0.0334, "step": 39000 }, { "grad_norm": 0.41795146465301514, "learning_rate": 7.157429392352514e-05, "loss": 0.049, "step": 39010 }, { "grad_norm": 0.5002520084381104, "learning_rate": 7.155937646732275e-05, "loss": 0.0358, "step": 39020 }, { "grad_norm": 1.2245795726776123, "learning_rate": 7.154445665341933e-05, "loss": 0.0378, "step": 39030 }, { "grad_norm": 0.40801674127578735, "learning_rate": 7.152953448344654e-05, "loss": 0.0404, "step": 39040 }, { "grad_norm": 0.8190938234329224, "learning_rate": 7.151460995903624e-05, "loss": 0.0456, "step": 39050 }, { "grad_norm": 0.5105882287025452, "learning_rate": 7.149968308182052e-05, "loss": 0.045, "step": 39060 }, { "grad_norm": 0.5432350635528564, "learning_rate": 7.14847538534318e-05, "loss": 0.0437, "step": 39070 }, { "grad_norm": 0.44815683364868164, "learning_rate": 7.14698222755027e-05, "loss": 0.0377, "step": 39080 }, { "grad_norm": 0.4237990081310272, "learning_rate": 7.14548883496661e-05, "loss": 0.0417, "step": 39090 }, { "grad_norm": 0.5454065203666687, "learning_rate": 7.143995207755517e-05, "loss": 0.0475, "step": 39100 }, { "grad_norm": 0.5044477581977844, "learning_rate": 7.142501346080333e-05, "loss": 0.0375, "step": 39110 }, { "grad_norm": 0.6062731146812439, "learning_rate": 7.141007250104421e-05, "loss": 0.0456, "step": 39120 }, { "grad_norm": 0.38076111674308777, "learning_rate": 7.139512919991176e-05, "loss": 0.045, "step": 39130 }, { "grad_norm": 0.5863807201385498, "learning_rate": 7.138018355904014e-05, "loss": 0.0405, "step": 39140 }, { "grad_norm": 0.5244370698928833, "learning_rate": 7.13652355800638e-05, "loss": 0.0621, "step": 39150 }, { "grad_norm": 0.8249783515930176, "learning_rate": 7.135028526461743e-05, "loss": 0.0471, "step": 39160 }, { "grad_norm": 0.51309734582901, "learning_rate": 7.133533261433594e-05, "loss": 0.0364, "step": 39170 }, { "grad_norm": 0.4898369014263153, "learning_rate": 7.132037763085457e-05, "loss": 0.0427, "step": 39180 }, { "grad_norm": 0.42085000872612, "learning_rate": 7.130542031580875e-05, "loss": 0.0367, "step": 39190 }, { "grad_norm": 0.3617035746574402, "learning_rate": 7.12904606708342e-05, "loss": 0.0419, "step": 39200 }, { "grad_norm": 0.3456515669822693, "learning_rate": 7.127549869756687e-05, "loss": 0.051, "step": 39210 }, { "grad_norm": 0.34052133560180664, "learning_rate": 7.126053439764299e-05, "loss": 0.0419, "step": 39220 }, { "grad_norm": 0.3643095791339874, "learning_rate": 7.124556777269904e-05, "loss": 0.0484, "step": 39230 }, { "grad_norm": 0.6655486822128296, "learning_rate": 7.123059882437174e-05, "loss": 0.0524, "step": 39240 }, { "grad_norm": 0.43752211332321167, "learning_rate": 7.121562755429807e-05, "loss": 0.0399, "step": 39250 }, { "grad_norm": 0.5388575792312622, "learning_rate": 7.120065396411528e-05, "loss": 0.0389, "step": 39260 }, { "grad_norm": 0.45063912868499756, "learning_rate": 7.118567805546084e-05, "loss": 0.0385, "step": 39270 }, { "grad_norm": 0.5256432890892029, "learning_rate": 7.117069982997248e-05, "loss": 0.046, "step": 39280 }, { "grad_norm": 0.37460193037986755, "learning_rate": 7.115571928928825e-05, "loss": 0.0565, "step": 39290 }, { "grad_norm": 0.5695878267288208, "learning_rate": 7.114073643504635e-05, "loss": 0.0412, "step": 39300 }, { "grad_norm": 1.0268720388412476, "learning_rate": 7.11257512688853e-05, "loss": 0.0437, "step": 39310 }, { "grad_norm": 0.6848532557487488, "learning_rate": 7.111076379244384e-05, "loss": 0.0414, "step": 39320 }, { "grad_norm": 0.5316080451011658, "learning_rate": 7.109577400736101e-05, "loss": 0.051, "step": 39330 }, { "grad_norm": 0.7303645610809326, "learning_rate": 7.108078191527605e-05, "loss": 0.0471, "step": 39340 }, { "grad_norm": 0.7374826669692993, "learning_rate": 7.106578751782847e-05, "loss": 0.0536, "step": 39350 }, { "grad_norm": 0.3540191054344177, "learning_rate": 7.105079081665803e-05, "loss": 0.0469, "step": 39360 }, { "grad_norm": 0.4446394145488739, "learning_rate": 7.103579181340476e-05, "loss": 0.0516, "step": 39370 }, { "grad_norm": 0.9034110307693481, "learning_rate": 7.102079050970893e-05, "loss": 0.0423, "step": 39380 }, { "grad_norm": 0.33679965138435364, "learning_rate": 7.100578690721105e-05, "loss": 0.0464, "step": 39390 }, { "grad_norm": 0.7158055901527405, "learning_rate": 7.09907810075519e-05, "loss": 0.0549, "step": 39400 }, { "grad_norm": 0.41171544790267944, "learning_rate": 7.097577281237249e-05, "loss": 0.0553, "step": 39410 }, { "grad_norm": 0.42887455224990845, "learning_rate": 7.09607623233141e-05, "loss": 0.037, "step": 39420 }, { "grad_norm": 0.4815773665904999, "learning_rate": 7.094574954201827e-05, "loss": 0.0422, "step": 39430 }, { "grad_norm": 0.5348266959190369, "learning_rate": 7.093073447012675e-05, "loss": 0.0449, "step": 39440 }, { "grad_norm": 0.47471094131469727, "learning_rate": 7.09157171092816e-05, "loss": 0.037, "step": 39450 }, { "grad_norm": 0.36997655034065247, "learning_rate": 7.090069746112504e-05, "loss": 0.0423, "step": 39460 }, { "grad_norm": 0.7257765531539917, "learning_rate": 7.088567552729965e-05, "loss": 0.0391, "step": 39470 }, { "grad_norm": 0.28470802307128906, "learning_rate": 7.087065130944818e-05, "loss": 0.0466, "step": 39480 }, { "grad_norm": 0.4441598355770111, "learning_rate": 7.085562480921366e-05, "loss": 0.0488, "step": 39490 }, { "grad_norm": 0.3372531831264496, "learning_rate": 7.084059602823937e-05, "loss": 0.0395, "step": 39500 }, { "grad_norm": 0.44870537519454956, "learning_rate": 7.082556496816882e-05, "loss": 0.0393, "step": 39510 }, { "grad_norm": 0.4594212770462036, "learning_rate": 7.081053163064582e-05, "loss": 0.0554, "step": 39520 }, { "grad_norm": 0.4296182692050934, "learning_rate": 7.079549601731434e-05, "loss": 0.0383, "step": 39530 }, { "grad_norm": 0.40341854095458984, "learning_rate": 7.07804581298187e-05, "loss": 0.0471, "step": 39540 }, { "grad_norm": 0.41981521248817444, "learning_rate": 7.07654179698034e-05, "loss": 0.0574, "step": 39550 }, { "grad_norm": 0.5163803696632385, "learning_rate": 7.075037553891321e-05, "loss": 0.0476, "step": 39560 }, { "grad_norm": 0.4006626307964325, "learning_rate": 7.073533083879315e-05, "loss": 0.0365, "step": 39570 }, { "grad_norm": 0.42375046014785767, "learning_rate": 7.072028387108849e-05, "loss": 0.0421, "step": 39580 }, { "grad_norm": 0.4089753329753876, "learning_rate": 7.070523463744472e-05, "loss": 0.0469, "step": 39590 }, { "grad_norm": 0.3243926763534546, "learning_rate": 7.069018313950763e-05, "loss": 0.0399, "step": 39600 }, { "grad_norm": 0.36505308747291565, "learning_rate": 7.067512937892322e-05, "loss": 0.0364, "step": 39610 }, { "grad_norm": 0.29956167936325073, "learning_rate": 7.066007335733775e-05, "loss": 0.0308, "step": 39620 }, { "grad_norm": 0.4262751042842865, "learning_rate": 7.064501507639772e-05, "loss": 0.0479, "step": 39630 }, { "grad_norm": 0.4068503677845001, "learning_rate": 7.062995453774987e-05, "loss": 0.0502, "step": 39640 }, { "grad_norm": 0.372905969619751, "learning_rate": 7.061489174304121e-05, "loss": 0.0599, "step": 39650 }, { "grad_norm": 0.6945723295211792, "learning_rate": 7.0599826693919e-05, "loss": 0.0368, "step": 39660 }, { "grad_norm": 0.5573845505714417, "learning_rate": 7.05847593920307e-05, "loss": 0.0523, "step": 39670 }, { "grad_norm": 0.5135947465896606, "learning_rate": 7.056968983902406e-05, "loss": 0.0397, "step": 39680 }, { "grad_norm": 0.3130562901496887, "learning_rate": 7.055461803654706e-05, "loss": 0.0382, "step": 39690 }, { "grad_norm": 0.2570433020591736, "learning_rate": 7.053954398624794e-05, "loss": 0.0329, "step": 39700 }, { "grad_norm": 0.2701146602630615, "learning_rate": 7.052446768977518e-05, "loss": 0.0345, "step": 39710 }, { "grad_norm": 1.5875447988510132, "learning_rate": 7.050938914877748e-05, "loss": 0.0505, "step": 39720 }, { "grad_norm": 0.5999469757080078, "learning_rate": 7.049430836490382e-05, "loss": 0.0545, "step": 39730 }, { "grad_norm": 0.30388930439949036, "learning_rate": 7.047922533980342e-05, "loss": 0.045, "step": 39740 }, { "grad_norm": 0.4431889057159424, "learning_rate": 7.046414007512571e-05, "loss": 0.0418, "step": 39750 }, { "grad_norm": 0.8785716891288757, "learning_rate": 7.044905257252042e-05, "loss": 0.0371, "step": 39760 }, { "grad_norm": 0.5173503160476685, "learning_rate": 7.04339628336375e-05, "loss": 0.0494, "step": 39770 }, { "grad_norm": 0.5148320198059082, "learning_rate": 7.041887086012711e-05, "loss": 0.0378, "step": 39780 }, { "grad_norm": 0.6107016801834106, "learning_rate": 7.040377665363969e-05, "loss": 0.0448, "step": 39790 }, { "grad_norm": 0.3352833688259125, "learning_rate": 7.038868021582594e-05, "loss": 0.0428, "step": 39800 }, { "grad_norm": 0.29603111743927, "learning_rate": 7.037358154833679e-05, "loss": 0.0384, "step": 39810 }, { "grad_norm": 0.2686178684234619, "learning_rate": 7.035848065282339e-05, "loss": 0.045, "step": 39820 }, { "grad_norm": 0.46817415952682495, "learning_rate": 7.034337753093714e-05, "loss": 0.0349, "step": 39830 }, { "grad_norm": 0.36088132858276367, "learning_rate": 7.032827218432972e-05, "loss": 0.0366, "step": 39840 }, { "grad_norm": 0.30098286271095276, "learning_rate": 7.031316461465302e-05, "loss": 0.0356, "step": 39850 }, { "grad_norm": 0.4158289432525635, "learning_rate": 7.029805482355915e-05, "loss": 0.0389, "step": 39860 }, { "grad_norm": 0.4269011318683624, "learning_rate": 7.028294281270055e-05, "loss": 0.0351, "step": 39870 }, { "grad_norm": 0.626111626625061, "learning_rate": 7.026782858372982e-05, "loss": 0.0429, "step": 39880 }, { "grad_norm": 0.40711721777915955, "learning_rate": 7.025271213829982e-05, "loss": 0.0423, "step": 39890 }, { "grad_norm": 0.32522714138031006, "learning_rate": 7.023759347806366e-05, "loss": 0.0355, "step": 39900 }, { "grad_norm": 0.37110278010368347, "learning_rate": 7.022247260467472e-05, "loss": 0.0443, "step": 39910 }, { "grad_norm": 0.5484391450881958, "learning_rate": 7.020734951978658e-05, "loss": 0.0445, "step": 39920 }, { "grad_norm": 0.4509897828102112, "learning_rate": 7.019222422505307e-05, "loss": 0.0409, "step": 39930 }, { "grad_norm": 0.42847540974617004, "learning_rate": 7.017709672212827e-05, "loss": 0.0471, "step": 39940 }, { "grad_norm": 0.39136698842048645, "learning_rate": 7.016196701266652e-05, "loss": 0.0462, "step": 39950 }, { "grad_norm": 0.37136176228523254, "learning_rate": 7.014683509832238e-05, "loss": 0.0436, "step": 39960 }, { "grad_norm": 0.3566553592681885, "learning_rate": 7.013170098075063e-05, "loss": 0.0488, "step": 39970 }, { "grad_norm": 0.352069228887558, "learning_rate": 7.011656466160632e-05, "loss": 0.0384, "step": 39980 }, { "grad_norm": 0.3817393183708191, "learning_rate": 7.010142614254475e-05, "loss": 0.0359, "step": 39990 }, { "grad_norm": 0.3181687295436859, "learning_rate": 7.008628542522147e-05, "loss": 0.0495, "step": 40000 }, { "grad_norm": 0.28824904561042786, "learning_rate": 7.007114251129218e-05, "loss": 0.0399, "step": 40010 }, { "grad_norm": 0.5505950450897217, "learning_rate": 7.005599740241293e-05, "loss": 0.0331, "step": 40020 }, { "grad_norm": 0.4311603307723999, "learning_rate": 7.004085010023996e-05, "loss": 0.0462, "step": 40030 }, { "grad_norm": 0.4373287260532379, "learning_rate": 7.002570060642976e-05, "loss": 0.0364, "step": 40040 }, { "grad_norm": 0.5316632986068726, "learning_rate": 7.001054892263903e-05, "loss": 0.0342, "step": 40050 }, { "grad_norm": 0.4598081409931183, "learning_rate": 6.999539505052477e-05, "loss": 0.0527, "step": 40060 }, { "grad_norm": 0.3950376510620117, "learning_rate": 6.998023899174418e-05, "loss": 0.0403, "step": 40070 }, { "grad_norm": 0.38434022665023804, "learning_rate": 6.996508074795467e-05, "loss": 0.0414, "step": 40080 }, { "grad_norm": 0.4454289376735687, "learning_rate": 6.994992032081396e-05, "loss": 0.0535, "step": 40090 }, { "grad_norm": 0.3972344696521759, "learning_rate": 6.993475771197995e-05, "loss": 0.0443, "step": 40100 }, { "grad_norm": 0.37373045086860657, "learning_rate": 6.991959292311082e-05, "loss": 0.0399, "step": 40110 }, { "grad_norm": 0.5494242310523987, "learning_rate": 6.990442595586495e-05, "loss": 0.0473, "step": 40120 }, { "grad_norm": 0.5735034942626953, "learning_rate": 6.988925681190098e-05, "loss": 0.0455, "step": 40130 }, { "grad_norm": 0.5322055816650391, "learning_rate": 6.987408549287778e-05, "loss": 0.0457, "step": 40140 }, { "grad_norm": 0.5634461045265198, "learning_rate": 6.985891200045449e-05, "loss": 0.0516, "step": 40150 }, { "grad_norm": 0.4923756718635559, "learning_rate": 6.984373633629045e-05, "loss": 0.0511, "step": 40160 }, { "grad_norm": 0.4655517041683197, "learning_rate": 6.98285585020452e-05, "loss": 0.0367, "step": 40170 }, { "grad_norm": 0.3335195481777191, "learning_rate": 6.981337849937864e-05, "loss": 0.0416, "step": 40180 }, { "grad_norm": 0.37100479006767273, "learning_rate": 6.979819632995078e-05, "loss": 0.0426, "step": 40190 }, { "grad_norm": 0.5185335874557495, "learning_rate": 6.978301199542193e-05, "loss": 0.0334, "step": 40200 }, { "grad_norm": 0.38902318477630615, "learning_rate": 6.976782549745263e-05, "loss": 0.0422, "step": 40210 }, { "grad_norm": 0.3845829665660858, "learning_rate": 6.975263683770367e-05, "loss": 0.0436, "step": 40220 }, { "grad_norm": 0.5837946534156799, "learning_rate": 6.973744601783604e-05, "loss": 0.0382, "step": 40230 }, { "grad_norm": 0.3682866394519806, "learning_rate": 6.972225303951097e-05, "loss": 0.0422, "step": 40240 }, { "grad_norm": 0.34015849232673645, "learning_rate": 6.970705790438998e-05, "loss": 0.0322, "step": 40250 }, { "grad_norm": 0.45219653844833374, "learning_rate": 6.969186061413477e-05, "loss": 0.0357, "step": 40260 }, { "grad_norm": 0.2582547962665558, "learning_rate": 6.967666117040727e-05, "loss": 0.0512, "step": 40270 }, { "grad_norm": 0.4248785376548767, "learning_rate": 6.966145957486972e-05, "loss": 0.0432, "step": 40280 }, { "grad_norm": 0.5398623943328857, "learning_rate": 6.964625582918449e-05, "loss": 0.0479, "step": 40290 }, { "grad_norm": 0.5390434265136719, "learning_rate": 6.963104993501425e-05, "loss": 0.0431, "step": 40300 }, { "grad_norm": 0.6268911361694336, "learning_rate": 6.961584189402192e-05, "loss": 0.0428, "step": 40310 }, { "grad_norm": 0.4662528336048126, "learning_rate": 6.96006317078706e-05, "loss": 0.0339, "step": 40320 }, { "grad_norm": 0.584027111530304, "learning_rate": 6.958541937822367e-05, "loss": 0.0379, "step": 40330 }, { "grad_norm": 0.31191688776016235, "learning_rate": 6.957020490674473e-05, "loss": 0.0362, "step": 40340 }, { "grad_norm": 0.32332876324653625, "learning_rate": 6.95549882950976e-05, "loss": 0.0397, "step": 40350 }, { "grad_norm": 0.5188407897949219, "learning_rate": 6.953976954494635e-05, "loss": 0.0336, "step": 40360 }, { "grad_norm": 0.37150752544403076, "learning_rate": 6.952454865795528e-05, "loss": 0.0431, "step": 40370 }, { "grad_norm": 0.33460959792137146, "learning_rate": 6.950932563578892e-05, "loss": 0.043, "step": 40380 }, { "grad_norm": 0.31911584734916687, "learning_rate": 6.949410048011206e-05, "loss": 0.0404, "step": 40390 }, { "grad_norm": 0.32601693272590637, "learning_rate": 6.947887319258966e-05, "loss": 0.0413, "step": 40400 }, { "grad_norm": 0.2903432548046112, "learning_rate": 6.946364377488696e-05, "loss": 0.0457, "step": 40410 }, { "grad_norm": 0.36731937527656555, "learning_rate": 6.944841222866947e-05, "loss": 0.0494, "step": 40420 }, { "grad_norm": 0.4324439764022827, "learning_rate": 6.943317855560284e-05, "loss": 0.0505, "step": 40430 }, { "grad_norm": 0.5285816192626953, "learning_rate": 6.941794275735302e-05, "loss": 0.0379, "step": 40440 }, { "grad_norm": 0.29153335094451904, "learning_rate": 6.94027048355862e-05, "loss": 0.0327, "step": 40450 }, { "grad_norm": 0.321018785238266, "learning_rate": 6.938746479196871e-05, "loss": 0.0392, "step": 40460 }, { "grad_norm": 0.5217497944831848, "learning_rate": 6.937222262816724e-05, "loss": 0.0362, "step": 40470 }, { "grad_norm": 0.3953595459461212, "learning_rate": 6.935697834584865e-05, "loss": 0.0445, "step": 40480 }, { "grad_norm": 0.4166049659252167, "learning_rate": 6.934173194667998e-05, "loss": 0.0435, "step": 40490 }, { "grad_norm": 0.5364322662353516, "learning_rate": 6.93264834323286e-05, "loss": 0.0462, "step": 40500 }, { "grad_norm": 0.2467658519744873, "learning_rate": 6.931123280446206e-05, "loss": 0.0465, "step": 40510 }, { "grad_norm": 0.5829591751098633, "learning_rate": 6.929598006474811e-05, "loss": 0.0468, "step": 40520 }, { "grad_norm": 0.30040279030799866, "learning_rate": 6.92807252148548e-05, "loss": 0.0421, "step": 40530 }, { "grad_norm": 0.5059992671012878, "learning_rate": 6.926546825645037e-05, "loss": 0.0309, "step": 40540 }, { "grad_norm": 0.27616095542907715, "learning_rate": 6.92502091912033e-05, "loss": 0.0364, "step": 40550 }, { "grad_norm": 0.3779771625995636, "learning_rate": 6.923494802078229e-05, "loss": 0.0494, "step": 40560 }, { "grad_norm": 1.5861194133758545, "learning_rate": 6.921968474685628e-05, "loss": 0.0429, "step": 40570 }, { "grad_norm": 0.5566524863243103, "learning_rate": 6.920441937109445e-05, "loss": 0.0562, "step": 40580 }, { "grad_norm": 0.4148649275302887, "learning_rate": 6.91891518951662e-05, "loss": 0.0509, "step": 40590 }, { "grad_norm": 0.47660741209983826, "learning_rate": 6.917388232074114e-05, "loss": 0.0509, "step": 40600 }, { "grad_norm": 0.43758630752563477, "learning_rate": 6.915861064948914e-05, "loss": 0.0486, "step": 40610 }, { "grad_norm": 0.2620386779308319, "learning_rate": 6.914333688308025e-05, "loss": 0.036, "step": 40620 }, { "grad_norm": 0.30698615312576294, "learning_rate": 6.912806102318487e-05, "loss": 0.0374, "step": 40630 }, { "grad_norm": 0.2508881390094757, "learning_rate": 6.911278307147347e-05, "loss": 0.0392, "step": 40640 }, { "grad_norm": 0.33485937118530273, "learning_rate": 6.909750302961684e-05, "loss": 0.0415, "step": 40650 }, { "grad_norm": 0.4550580382347107, "learning_rate": 6.9082220899286e-05, "loss": 0.0427, "step": 40660 }, { "grad_norm": 0.4237055480480194, "learning_rate": 6.906693668215219e-05, "loss": 0.0532, "step": 40670 }, { "grad_norm": 0.4508135914802551, "learning_rate": 6.905165037988683e-05, "loss": 0.0431, "step": 40680 }, { "grad_norm": 0.2769908308982849, "learning_rate": 6.903636199416164e-05, "loss": 0.037, "step": 40690 }, { "grad_norm": 0.5586333274841309, "learning_rate": 6.902107152664851e-05, "loss": 0.0407, "step": 40700 }, { "grad_norm": 0.6955301761627197, "learning_rate": 6.90057789790196e-05, "loss": 0.0373, "step": 40710 }, { "grad_norm": 0.4903745949268341, "learning_rate": 6.899048435294728e-05, "loss": 0.046, "step": 40720 }, { "grad_norm": 0.38298743963241577, "learning_rate": 6.897518765010415e-05, "loss": 0.043, "step": 40730 }, { "grad_norm": 0.33936363458633423, "learning_rate": 6.895988887216302e-05, "loss": 0.0447, "step": 40740 }, { "grad_norm": 0.3819788098335266, "learning_rate": 6.894458802079694e-05, "loss": 0.0509, "step": 40750 }, { "grad_norm": 0.4877469837665558, "learning_rate": 6.892928509767922e-05, "loss": 0.0353, "step": 40760 }, { "grad_norm": 0.26040175557136536, "learning_rate": 6.891398010448333e-05, "loss": 0.0354, "step": 40770 }, { "grad_norm": 0.5010051131248474, "learning_rate": 6.8898673042883e-05, "loss": 0.057, "step": 40780 }, { "grad_norm": 0.42668986320495605, "learning_rate": 6.888336391455222e-05, "loss": 0.0441, "step": 40790 }, { "grad_norm": 0.30001533031463623, "learning_rate": 6.886805272116513e-05, "loss": 0.038, "step": 40800 }, { "grad_norm": 0.4521702826023102, "learning_rate": 6.885273946439617e-05, "loss": 0.044, "step": 40810 }, { "grad_norm": 0.8121956586837769, "learning_rate": 6.883742414591998e-05, "loss": 0.0524, "step": 40820 }, { "grad_norm": 0.36640074849128723, "learning_rate": 6.88221067674114e-05, "loss": 0.0391, "step": 40830 }, { "grad_norm": 1.811341404914856, "learning_rate": 6.88067873305455e-05, "loss": 0.0514, "step": 40840 }, { "grad_norm": 0.5874635577201843, "learning_rate": 6.879146583699765e-05, "loss": 0.0479, "step": 40850 }, { "grad_norm": 0.3301965296268463, "learning_rate": 6.877614228844334e-05, "loss": 0.0498, "step": 40860 }, { "grad_norm": 0.4149407744407654, "learning_rate": 6.876081668655832e-05, "loss": 0.0467, "step": 40870 }, { "grad_norm": 0.4332823157310486, "learning_rate": 6.874548903301861e-05, "loss": 0.0534, "step": 40880 }, { "grad_norm": 0.40890178084373474, "learning_rate": 6.873015932950039e-05, "loss": 0.0434, "step": 40890 }, { "grad_norm": 0.5114945769309998, "learning_rate": 6.871482757768012e-05, "loss": 0.049, "step": 40900 }, { "grad_norm": 0.41184738278388977, "learning_rate": 6.869949377923442e-05, "loss": 0.0457, "step": 40910 }, { "grad_norm": 0.4436991512775421, "learning_rate": 6.868415793584022e-05, "loss": 0.047, "step": 40920 }, { "grad_norm": 0.7096250653266907, "learning_rate": 6.86688200491746e-05, "loss": 0.043, "step": 40930 }, { "grad_norm": 0.2990979850292206, "learning_rate": 6.865348012091488e-05, "loss": 0.0367, "step": 40940 }, { "grad_norm": 0.37680014967918396, "learning_rate": 6.86381381527386e-05, "loss": 0.0457, "step": 40950 }, { "grad_norm": 0.42545828223228455, "learning_rate": 6.862279414632359e-05, "loss": 0.0389, "step": 40960 }, { "grad_norm": 0.41066423058509827, "learning_rate": 6.860744810334778e-05, "loss": 0.0383, "step": 40970 }, { "grad_norm": 0.3653920590877533, "learning_rate": 6.859210002548943e-05, "loss": 0.0395, "step": 40980 }, { "grad_norm": 0.4075745940208435, "learning_rate": 6.8576749914427e-05, "loss": 0.0386, "step": 40990 }, { "grad_norm": 0.5458601117134094, "learning_rate": 6.85613977718391e-05, "loss": 0.0416, "step": 41000 }, { "grad_norm": 0.9704867601394653, "learning_rate": 6.854604359940464e-05, "loss": 0.0369, "step": 41010 }, { "grad_norm": 0.42142829298973083, "learning_rate": 6.853068739880276e-05, "loss": 0.0403, "step": 41020 }, { "grad_norm": 0.564211368560791, "learning_rate": 6.851532917171276e-05, "loss": 0.0419, "step": 41030 }, { "grad_norm": 0.4984193444252014, "learning_rate": 6.849996891981418e-05, "loss": 0.0522, "step": 41040 }, { "grad_norm": 0.4681640863418579, "learning_rate": 6.848460664478684e-05, "loss": 0.036, "step": 41050 }, { "grad_norm": 0.3764421343803406, "learning_rate": 6.846924234831069e-05, "loss": 0.0299, "step": 41060 }, { "grad_norm": 0.3841967284679413, "learning_rate": 6.845387603206597e-05, "loss": 0.0481, "step": 41070 }, { "grad_norm": 0.3949703872203827, "learning_rate": 6.843850769773312e-05, "loss": 0.0409, "step": 41080 }, { "grad_norm": 0.3490547239780426, "learning_rate": 6.842313734699278e-05, "loss": 0.0479, "step": 41090 }, { "grad_norm": 0.44719409942626953, "learning_rate": 6.840776498152584e-05, "loss": 0.0367, "step": 41100 }, { "grad_norm": 0.33915194869041443, "learning_rate": 6.83923906030134e-05, "loss": 0.0346, "step": 41110 }, { "grad_norm": 0.6315507292747498, "learning_rate": 6.837701421313677e-05, "loss": 0.0602, "step": 41120 }, { "grad_norm": 0.5817111730575562, "learning_rate": 6.83616358135775e-05, "loss": 0.0425, "step": 41130 }, { "grad_norm": 0.8291776776313782, "learning_rate": 6.834625540601734e-05, "loss": 0.0529, "step": 41140 }, { "grad_norm": 0.4194899797439575, "learning_rate": 6.833087299213829e-05, "loss": 0.0375, "step": 41150 }, { "grad_norm": 0.770244300365448, "learning_rate": 6.83154885736225e-05, "loss": 0.0557, "step": 41160 }, { "grad_norm": 0.5232301354408264, "learning_rate": 6.830010215215244e-05, "loss": 0.0502, "step": 41170 }, { "grad_norm": 0.8686614632606506, "learning_rate": 6.828471372941072e-05, "loss": 0.0426, "step": 41180 }, { "grad_norm": 0.44700536131858826, "learning_rate": 6.826932330708018e-05, "loss": 0.0423, "step": 41190 }, { "grad_norm": 0.4320243000984192, "learning_rate": 6.825393088684393e-05, "loss": 0.0353, "step": 41200 }, { "grad_norm": 0.37886321544647217, "learning_rate": 6.823853647038524e-05, "loss": 0.0361, "step": 41210 }, { "grad_norm": 0.2519794702529907, "learning_rate": 6.822314005938762e-05, "loss": 0.0346, "step": 41220 }, { "grad_norm": 0.47460711002349854, "learning_rate": 6.820774165553479e-05, "loss": 0.0443, "step": 41230 }, { "grad_norm": 0.271837443113327, "learning_rate": 6.819234126051072e-05, "loss": 0.037, "step": 41240 }, { "grad_norm": 0.683458149433136, "learning_rate": 6.817693887599956e-05, "loss": 0.0465, "step": 41250 }, { "grad_norm": 0.3397873044013977, "learning_rate": 6.816153450368568e-05, "loss": 0.0508, "step": 41260 }, { "grad_norm": 0.24917270243167877, "learning_rate": 6.814612814525372e-05, "loss": 0.0483, "step": 41270 }, { "grad_norm": 0.5172823667526245, "learning_rate": 6.813071980238846e-05, "loss": 0.0369, "step": 41280 }, { "grad_norm": 0.8764515519142151, "learning_rate": 6.811530947677492e-05, "loss": 0.0413, "step": 41290 }, { "grad_norm": 0.3865261375904083, "learning_rate": 6.809989717009839e-05, "loss": 0.0335, "step": 41300 }, { "grad_norm": 0.5010880827903748, "learning_rate": 6.808448288404431e-05, "loss": 0.0577, "step": 41310 }, { "grad_norm": 0.4977051317691803, "learning_rate": 6.806906662029838e-05, "loss": 0.0487, "step": 41320 }, { "grad_norm": 0.5704377293586731, "learning_rate": 6.805364838054647e-05, "loss": 0.0433, "step": 41330 }, { "grad_norm": 0.3280700147151947, "learning_rate": 6.803822816647474e-05, "loss": 0.0308, "step": 41340 }, { "grad_norm": 0.5432111620903015, "learning_rate": 6.802280597976949e-05, "loss": 0.0399, "step": 41350 }, { "grad_norm": 0.5296091437339783, "learning_rate": 6.800738182211727e-05, "loss": 0.0352, "step": 41360 }, { "grad_norm": 0.3691132962703705, "learning_rate": 6.799195569520485e-05, "loss": 0.0398, "step": 41370 }, { "grad_norm": 0.8915606141090393, "learning_rate": 6.79765276007192e-05, "loss": 0.0424, "step": 41380 }, { "grad_norm": 0.36601313948631287, "learning_rate": 6.796109754034753e-05, "loss": 0.045, "step": 41390 }, { "grad_norm": 0.4225255846977234, "learning_rate": 6.794566551577724e-05, "loss": 0.0412, "step": 41400 }, { "grad_norm": 0.3754546642303467, "learning_rate": 6.793023152869594e-05, "loss": 0.0367, "step": 41410 }, { "grad_norm": 0.49809351563453674, "learning_rate": 6.791479558079147e-05, "loss": 0.0376, "step": 41420 }, { "grad_norm": 0.23194736242294312, "learning_rate": 6.789935767375191e-05, "loss": 0.0406, "step": 41430 }, { "grad_norm": 0.30725181102752686, "learning_rate": 6.78839178092655e-05, "loss": 0.0289, "step": 41440 }, { "grad_norm": 0.2975594997406006, "learning_rate": 6.786847598902072e-05, "loss": 0.0462, "step": 41450 }, { "grad_norm": 0.3985145092010498, "learning_rate": 6.785303221470627e-05, "loss": 0.0387, "step": 41460 }, { "grad_norm": 0.4486001431941986, "learning_rate": 6.783758648801108e-05, "loss": 0.0374, "step": 41470 }, { "grad_norm": 0.6712166666984558, "learning_rate": 6.782213881062422e-05, "loss": 0.0492, "step": 41480 }, { "grad_norm": 0.44811004400253296, "learning_rate": 6.780668918423508e-05, "loss": 0.0385, "step": 41490 }, { "grad_norm": 0.4842977225780487, "learning_rate": 6.779123761053317e-05, "loss": 0.0391, "step": 41500 }, { "grad_norm": 0.33183956146240234, "learning_rate": 6.777578409120826e-05, "loss": 0.0508, "step": 41510 }, { "grad_norm": 0.5703564882278442, "learning_rate": 6.776032862795033e-05, "loss": 0.0435, "step": 41520 }, { "grad_norm": 0.5570678114891052, "learning_rate": 6.774487122244956e-05, "loss": 0.0424, "step": 41530 }, { "grad_norm": 0.34712427854537964, "learning_rate": 6.772941187639637e-05, "loss": 0.0323, "step": 41540 }, { "grad_norm": 0.46244651079177856, "learning_rate": 6.771395059148134e-05, "loss": 0.0482, "step": 41550 }, { "grad_norm": 0.48347747325897217, "learning_rate": 6.76984873693953e-05, "loss": 0.0373, "step": 41560 }, { "grad_norm": 0.595879077911377, "learning_rate": 6.768302221182931e-05, "loss": 0.0523, "step": 41570 }, { "grad_norm": 0.4830242097377777, "learning_rate": 6.766755512047457e-05, "loss": 0.0505, "step": 41580 }, { "grad_norm": 0.31555306911468506, "learning_rate": 6.765208609702259e-05, "loss": 0.0385, "step": 41590 }, { "grad_norm": 0.38565289974212646, "learning_rate": 6.763661514316499e-05, "loss": 0.032, "step": 41600 }, { "grad_norm": 0.3104969561100006, "learning_rate": 6.76211422605937e-05, "loss": 0.0414, "step": 41610 }, { "grad_norm": 0.42042312026023865, "learning_rate": 6.760566745100076e-05, "loss": 0.0388, "step": 41620 }, { "grad_norm": 1.0705677270889282, "learning_rate": 6.75901907160785e-05, "loss": 0.0417, "step": 41630 }, { "grad_norm": 0.34873446822166443, "learning_rate": 6.757471205751943e-05, "loss": 0.0376, "step": 41640 }, { "grad_norm": 0.5178724527359009, "learning_rate": 6.75592314770163e-05, "loss": 0.0431, "step": 41650 }, { "grad_norm": 0.4755977988243103, "learning_rate": 6.7543748976262e-05, "loss": 0.0508, "step": 41660 }, { "grad_norm": 0.2543822228908539, "learning_rate": 6.752826455694968e-05, "loss": 0.0486, "step": 41670 }, { "grad_norm": 0.32211148738861084, "learning_rate": 6.751277822077271e-05, "loss": 0.0422, "step": 41680 }, { "grad_norm": 0.28131550550460815, "learning_rate": 6.749728996942465e-05, "loss": 0.0379, "step": 41690 }, { "grad_norm": 0.2788110077381134, "learning_rate": 6.748179980459924e-05, "loss": 0.0378, "step": 41700 }, { "grad_norm": 0.3739945590496063, "learning_rate": 6.746630772799052e-05, "loss": 0.0377, "step": 41710 }, { "grad_norm": 0.47200289368629456, "learning_rate": 6.745081374129262e-05, "loss": 0.042, "step": 41720 }, { "grad_norm": 0.4906158447265625, "learning_rate": 6.743531784619998e-05, "loss": 0.0514, "step": 41730 }, { "grad_norm": 0.567489743232727, "learning_rate": 6.741982004440719e-05, "loss": 0.0434, "step": 41740 }, { "grad_norm": 0.7613455653190613, "learning_rate": 6.740432033760907e-05, "loss": 0.0396, "step": 41750 }, { "grad_norm": 0.47532564401626587, "learning_rate": 6.738881872750066e-05, "loss": 0.04, "step": 41760 }, { "grad_norm": 0.6577401757240295, "learning_rate": 6.737331521577715e-05, "loss": 0.0459, "step": 41770 }, { "grad_norm": 0.48699015378952026, "learning_rate": 6.735780980413403e-05, "loss": 0.0434, "step": 41780 }, { "grad_norm": 0.5188448429107666, "learning_rate": 6.734230249426692e-05, "loss": 0.0468, "step": 41790 }, { "grad_norm": 0.7345014810562134, "learning_rate": 6.732679328787168e-05, "loss": 0.0442, "step": 41800 }, { "grad_norm": 0.26837822794914246, "learning_rate": 6.731128218664438e-05, "loss": 0.0339, "step": 41810 }, { "grad_norm": 0.6922622323036194, "learning_rate": 6.729576919228129e-05, "loss": 0.0343, "step": 41820 }, { "grad_norm": 0.392164409160614, "learning_rate": 6.728025430647888e-05, "loss": 0.0385, "step": 41830 }, { "grad_norm": 0.42425769567489624, "learning_rate": 6.726473753093383e-05, "loss": 0.034, "step": 41840 }, { "grad_norm": 0.37959685921669006, "learning_rate": 6.724921886734305e-05, "loss": 0.0384, "step": 41850 }, { "grad_norm": 0.444298654794693, "learning_rate": 6.723369831740363e-05, "loss": 0.05, "step": 41860 }, { "grad_norm": 0.2847403585910797, "learning_rate": 6.721817588281288e-05, "loss": 0.0367, "step": 41870 }, { "grad_norm": 0.4526239335536957, "learning_rate": 6.720265156526828e-05, "loss": 0.0396, "step": 41880 }, { "grad_norm": 0.7085995674133301, "learning_rate": 6.718712536646758e-05, "loss": 0.0439, "step": 41890 }, { "grad_norm": 0.48339343070983887, "learning_rate": 6.71715972881087e-05, "loss": 0.0459, "step": 41900 }, { "grad_norm": 0.3834795653820038, "learning_rate": 6.715606733188974e-05, "loss": 0.0388, "step": 41910 }, { "grad_norm": 0.47085243463516235, "learning_rate": 6.714053549950909e-05, "loss": 0.0396, "step": 41920 }, { "grad_norm": 1.0709798336029053, "learning_rate": 6.71250017926652e-05, "loss": 0.0403, "step": 41930 }, { "grad_norm": 0.3802865147590637, "learning_rate": 6.71094662130569e-05, "loss": 0.0327, "step": 41940 }, { "grad_norm": 0.3082069158554077, "learning_rate": 6.709392876238307e-05, "loss": 0.0465, "step": 41950 }, { "grad_norm": 0.3870506286621094, "learning_rate": 6.707838944234289e-05, "loss": 0.0484, "step": 41960 }, { "grad_norm": 0.3514629304409027, "learning_rate": 6.706284825463574e-05, "loss": 0.0366, "step": 41970 }, { "grad_norm": 0.48618531227111816, "learning_rate": 6.704730520096114e-05, "loss": 0.0444, "step": 41980 }, { "grad_norm": 0.4186389148235321, "learning_rate": 6.703176028301888e-05, "loss": 0.0383, "step": 41990 }, { "grad_norm": 0.3081035614013672, "learning_rate": 6.701621350250892e-05, "loss": 0.0364, "step": 42000 }, { "grad_norm": 0.7411990165710449, "learning_rate": 6.700066486113144e-05, "loss": 0.0358, "step": 42010 }, { "grad_norm": 0.3927650451660156, "learning_rate": 6.69851143605868e-05, "loss": 0.0408, "step": 42020 }, { "grad_norm": 0.32925650477409363, "learning_rate": 6.696956200257561e-05, "loss": 0.0374, "step": 42030 }, { "grad_norm": 0.4548770785331726, "learning_rate": 6.695400778879863e-05, "loss": 0.0507, "step": 42040 }, { "grad_norm": 0.8567088842391968, "learning_rate": 6.693845172095683e-05, "loss": 0.0539, "step": 42050 }, { "grad_norm": 0.36375823616981506, "learning_rate": 6.692289380075142e-05, "loss": 0.0447, "step": 42060 }, { "grad_norm": 0.46250614523887634, "learning_rate": 6.690733402988379e-05, "loss": 0.0357, "step": 42070 }, { "grad_norm": 0.49974536895751953, "learning_rate": 6.689177241005553e-05, "loss": 0.0466, "step": 42080 }, { "grad_norm": 0.3918460011482239, "learning_rate": 6.687620894296844e-05, "loss": 0.0393, "step": 42090 }, { "grad_norm": 0.5353372097015381, "learning_rate": 6.686064363032451e-05, "loss": 0.0433, "step": 42100 }, { "grad_norm": 0.45299291610717773, "learning_rate": 6.684507647382596e-05, "loss": 0.0455, "step": 42110 }, { "grad_norm": 0.5839159488677979, "learning_rate": 6.682950747517513e-05, "loss": 0.0492, "step": 42120 }, { "grad_norm": 0.5235230326652527, "learning_rate": 6.68139366360747e-05, "loss": 0.0436, "step": 42130 }, { "grad_norm": 0.45225945115089417, "learning_rate": 6.679836395822744e-05, "loss": 0.0385, "step": 42140 }, { "grad_norm": 0.4810148775577545, "learning_rate": 6.678278944333633e-05, "loss": 0.0365, "step": 42150 }, { "grad_norm": 0.4615866541862488, "learning_rate": 6.676721309310462e-05, "loss": 0.0304, "step": 42160 }, { "grad_norm": 0.7098641991615295, "learning_rate": 6.67516349092357e-05, "loss": 0.0421, "step": 42170 }, { "grad_norm": 0.30209416151046753, "learning_rate": 6.673605489343315e-05, "loss": 0.0327, "step": 42180 }, { "grad_norm": 0.5258479714393616, "learning_rate": 6.672047304740081e-05, "loss": 0.0353, "step": 42190 }, { "grad_norm": 0.3617745339870453, "learning_rate": 6.670488937284268e-05, "loss": 0.0375, "step": 42200 }, { "grad_norm": 0.3041459619998932, "learning_rate": 6.668930387146294e-05, "loss": 0.0391, "step": 42210 }, { "grad_norm": 0.3664485216140747, "learning_rate": 6.667371654496605e-05, "loss": 0.0359, "step": 42220 }, { "grad_norm": 0.28172925114631653, "learning_rate": 6.665812739505659e-05, "loss": 0.0417, "step": 42230 }, { "grad_norm": 0.32792261242866516, "learning_rate": 6.664253642343935e-05, "loss": 0.0442, "step": 42240 }, { "grad_norm": 0.6403568387031555, "learning_rate": 6.662694363181935e-05, "loss": 0.0412, "step": 42250 }, { "grad_norm": 0.414749413728714, "learning_rate": 6.66113490219018e-05, "loss": 0.0465, "step": 42260 }, { "grad_norm": 0.31918585300445557, "learning_rate": 6.659575259539207e-05, "loss": 0.0338, "step": 42270 }, { "grad_norm": 0.45974183082580566, "learning_rate": 6.65801543539958e-05, "loss": 0.041, "step": 42280 }, { "grad_norm": 0.531688392162323, "learning_rate": 6.65645542994188e-05, "loss": 0.0406, "step": 42290 }, { "grad_norm": 0.29360395669937134, "learning_rate": 6.654895243336702e-05, "loss": 0.0282, "step": 42300 }, { "grad_norm": 0.3536306917667389, "learning_rate": 6.65333487575467e-05, "loss": 0.0289, "step": 42310 }, { "grad_norm": 0.25961360335350037, "learning_rate": 6.651774327366422e-05, "loss": 0.0334, "step": 42320 }, { "grad_norm": 0.5215413570404053, "learning_rate": 6.650213598342613e-05, "loss": 0.0443, "step": 42330 }, { "grad_norm": 0.3245108723640442, "learning_rate": 6.648652688853931e-05, "loss": 0.0449, "step": 42340 }, { "grad_norm": 0.45611318945884705, "learning_rate": 6.647091599071066e-05, "loss": 0.0491, "step": 42350 }, { "grad_norm": 0.3137128949165344, "learning_rate": 6.645530329164743e-05, "loss": 0.0431, "step": 42360 }, { "grad_norm": 0.37834417819976807, "learning_rate": 6.643968879305694e-05, "loss": 0.0387, "step": 42370 }, { "grad_norm": 0.348228394985199, "learning_rate": 6.642407249664683e-05, "loss": 0.0349, "step": 42380 }, { "grad_norm": 0.5059852600097656, "learning_rate": 6.640845440412483e-05, "loss": 0.0424, "step": 42390 }, { "grad_norm": 0.4593043327331543, "learning_rate": 6.639283451719893e-05, "loss": 0.0408, "step": 42400 }, { "grad_norm": 0.406316876411438, "learning_rate": 6.637721283757729e-05, "loss": 0.0342, "step": 42410 }, { "grad_norm": 0.368382066488266, "learning_rate": 6.636158936696828e-05, "loss": 0.037, "step": 42420 }, { "grad_norm": 0.5561526417732239, "learning_rate": 6.634596410708047e-05, "loss": 0.0432, "step": 42430 }, { "grad_norm": 1.0541282892227173, "learning_rate": 6.633033705962257e-05, "loss": 0.0444, "step": 42440 }, { "grad_norm": 0.5004150867462158, "learning_rate": 6.631470822630359e-05, "loss": 0.0435, "step": 42450 }, { "grad_norm": 0.5743356943130493, "learning_rate": 6.629907760883263e-05, "loss": 0.0476, "step": 42460 }, { "grad_norm": 0.4339495599269867, "learning_rate": 6.628344520891905e-05, "loss": 0.0378, "step": 42470 }, { "grad_norm": 0.31045475602149963, "learning_rate": 6.62678110282724e-05, "loss": 0.0388, "step": 42480 }, { "grad_norm": 0.4568137228488922, "learning_rate": 6.625217506860237e-05, "loss": 0.05, "step": 42490 }, { "grad_norm": 0.29363423585891724, "learning_rate": 6.623653733161892e-05, "loss": 0.0467, "step": 42500 }, { "grad_norm": 0.2614351212978363, "learning_rate": 6.622089781903217e-05, "loss": 0.0414, "step": 42510 }, { "grad_norm": 0.4183875620365143, "learning_rate": 6.620525653255241e-05, "loss": 0.0382, "step": 42520 }, { "grad_norm": 0.4925920069217682, "learning_rate": 6.618961347389015e-05, "loss": 0.0386, "step": 42530 }, { "grad_norm": 0.41145166754722595, "learning_rate": 6.617396864475613e-05, "loss": 0.0342, "step": 42540 }, { "grad_norm": 0.36452534794807434, "learning_rate": 6.61583220468612e-05, "loss": 0.0335, "step": 42550 }, { "grad_norm": 0.28983041644096375, "learning_rate": 6.614267368191645e-05, "loss": 0.0355, "step": 42560 }, { "grad_norm": 0.47468864917755127, "learning_rate": 6.61270235516332e-05, "loss": 0.039, "step": 42570 }, { "grad_norm": 0.2568288743495941, "learning_rate": 6.611137165772287e-05, "loss": 0.0293, "step": 42580 }, { "grad_norm": 0.3417527675628662, "learning_rate": 6.60957180018972e-05, "loss": 0.0408, "step": 42590 }, { "grad_norm": 0.44803065061569214, "learning_rate": 6.608006258586797e-05, "loss": 0.0361, "step": 42600 }, { "grad_norm": 0.2980402410030365, "learning_rate": 6.606440541134731e-05, "loss": 0.0347, "step": 42610 }, { "grad_norm": 0.2351495772600174, "learning_rate": 6.60487464800474e-05, "loss": 0.0475, "step": 42620 }, { "grad_norm": 0.6210575699806213, "learning_rate": 6.603308579368071e-05, "loss": 0.046, "step": 42630 }, { "grad_norm": 0.47672200202941895, "learning_rate": 6.601742335395987e-05, "loss": 0.0389, "step": 42640 }, { "grad_norm": 0.3270323872566223, "learning_rate": 6.600175916259769e-05, "loss": 0.0409, "step": 42650 }, { "grad_norm": 0.386360764503479, "learning_rate": 6.598609322130718e-05, "loss": 0.0328, "step": 42660 }, { "grad_norm": 0.2742300033569336, "learning_rate": 6.597042553180155e-05, "loss": 0.0296, "step": 42670 }, { "grad_norm": 0.3782385587692261, "learning_rate": 6.59547560957942e-05, "loss": 0.04, "step": 42680 }, { "grad_norm": 0.3691030442714691, "learning_rate": 6.59390849149987e-05, "loss": 0.045, "step": 42690 }, { "grad_norm": 0.36507490277290344, "learning_rate": 6.592341199112886e-05, "loss": 0.0342, "step": 42700 }, { "grad_norm": 0.2978309690952301, "learning_rate": 6.590773732589861e-05, "loss": 0.0398, "step": 42710 }, { "grad_norm": 0.5065374374389648, "learning_rate": 6.589206092102213e-05, "loss": 0.049, "step": 42720 }, { "grad_norm": 0.33702078461647034, "learning_rate": 6.587638277821375e-05, "loss": 0.0316, "step": 42730 }, { "grad_norm": 0.3213277757167816, "learning_rate": 6.586070289918803e-05, "loss": 0.034, "step": 42740 }, { "grad_norm": 0.42504721879959106, "learning_rate": 6.584502128565968e-05, "loss": 0.0393, "step": 42750 }, { "grad_norm": 0.3637210726737976, "learning_rate": 6.582933793934363e-05, "loss": 0.0404, "step": 42760 }, { "grad_norm": 0.31627944111824036, "learning_rate": 6.581365286195498e-05, "loss": 0.0334, "step": 42770 }, { "grad_norm": 0.20223140716552734, "learning_rate": 6.579796605520903e-05, "loss": 0.0357, "step": 42780 }, { "grad_norm": 0.48283618688583374, "learning_rate": 6.578227752082127e-05, "loss": 0.0428, "step": 42790 }, { "grad_norm": 0.36647742986679077, "learning_rate": 6.576658726050735e-05, "loss": 0.0353, "step": 42800 }, { "grad_norm": 0.29712799191474915, "learning_rate": 6.575089527598316e-05, "loss": 0.0283, "step": 42810 }, { "grad_norm": 0.2925355136394501, "learning_rate": 6.573520156896476e-05, "loss": 0.0437, "step": 42820 }, { "grad_norm": 0.42725279927253723, "learning_rate": 6.571950614116835e-05, "loss": 0.0373, "step": 42830 }, { "grad_norm": 0.572291910648346, "learning_rate": 6.570380899431039e-05, "loss": 0.0543, "step": 42840 }, { "grad_norm": 0.42565757036209106, "learning_rate": 6.568811013010749e-05, "loss": 0.0389, "step": 42850 }, { "grad_norm": 0.36829861998558044, "learning_rate": 6.567240955027645e-05, "loss": 0.0435, "step": 42860 }, { "grad_norm": 0.41513603925704956, "learning_rate": 6.565670725653427e-05, "loss": 0.0301, "step": 42870 }, { "grad_norm": 0.35731741786003113, "learning_rate": 6.564100325059812e-05, "loss": 0.0301, "step": 42880 }, { "grad_norm": 0.3749620318412781, "learning_rate": 6.562529753418538e-05, "loss": 0.0359, "step": 42890 }, { "grad_norm": 0.2216123789548874, "learning_rate": 6.56095901090136e-05, "loss": 0.0337, "step": 42900 }, { "grad_norm": 0.6279293894767761, "learning_rate": 6.55938809768005e-05, "loss": 0.031, "step": 42910 }, { "grad_norm": 0.8592126965522766, "learning_rate": 6.557817013926402e-05, "loss": 0.0409, "step": 42920 }, { "grad_norm": 0.3782833218574524, "learning_rate": 6.556245759812227e-05, "loss": 0.0457, "step": 42930 }, { "grad_norm": 0.6469299793243408, "learning_rate": 6.554674335509357e-05, "loss": 0.047, "step": 42940 }, { "grad_norm": 0.34705424308776855, "learning_rate": 6.553102741189638e-05, "loss": 0.0387, "step": 42950 }, { "grad_norm": 0.3935950994491577, "learning_rate": 6.55153097702494e-05, "loss": 0.0498, "step": 42960 }, { "grad_norm": 0.3757975101470947, "learning_rate": 6.549959043187146e-05, "loss": 0.0496, "step": 42970 }, { "grad_norm": 0.8742631673812866, "learning_rate": 6.54838693984816e-05, "loss": 0.0407, "step": 42980 }, { "grad_norm": 0.5223940014839172, "learning_rate": 6.546814667179909e-05, "loss": 0.0388, "step": 42990 }, { "grad_norm": 0.40778788924217224, "learning_rate": 6.545242225354328e-05, "loss": 0.0474, "step": 43000 }, { "grad_norm": 0.5196084380149841, "learning_rate": 6.543669614543383e-05, "loss": 0.0454, "step": 43010 }, { "grad_norm": 0.45670902729034424, "learning_rate": 6.542096834919049e-05, "loss": 0.0439, "step": 43020 }, { "grad_norm": 0.3831748366355896, "learning_rate": 6.540523886653324e-05, "loss": 0.0408, "step": 43030 }, { "grad_norm": 0.38348427414894104, "learning_rate": 6.538950769918222e-05, "loss": 0.041, "step": 43040 }, { "grad_norm": 0.4434329867362976, "learning_rate": 6.537377484885779e-05, "loss": 0.0412, "step": 43050 }, { "grad_norm": 0.6056943535804749, "learning_rate": 6.535804031728045e-05, "loss": 0.0441, "step": 43060 }, { "grad_norm": 0.38888874650001526, "learning_rate": 6.534230410617092e-05, "loss": 0.0437, "step": 43070 }, { "grad_norm": 0.5153404474258423, "learning_rate": 6.532656621725007e-05, "loss": 0.0397, "step": 43080 }, { "grad_norm": 0.4323362112045288, "learning_rate": 6.531082665223898e-05, "loss": 0.0401, "step": 43090 }, { "grad_norm": 0.2853938937187195, "learning_rate": 6.529508541285889e-05, "loss": 0.0406, "step": 43100 }, { "grad_norm": 0.6554144620895386, "learning_rate": 6.527934250083128e-05, "loss": 0.046, "step": 43110 }, { "grad_norm": 0.3929232954978943, "learning_rate": 6.526359791787775e-05, "loss": 0.038, "step": 43120 }, { "grad_norm": 0.29611891508102417, "learning_rate": 6.524785166572009e-05, "loss": 0.0375, "step": 43130 }, { "grad_norm": 0.36993762850761414, "learning_rate": 6.523210374608029e-05, "loss": 0.0403, "step": 43140 }, { "grad_norm": 0.59474116563797, "learning_rate": 6.521635416068054e-05, "loss": 0.034, "step": 43150 }, { "grad_norm": 0.2707997262477875, "learning_rate": 6.520060291124317e-05, "loss": 0.0589, "step": 43160 }, { "grad_norm": 0.2785327136516571, "learning_rate": 6.518484999949072e-05, "loss": 0.0412, "step": 43170 }, { "grad_norm": 0.3392171859741211, "learning_rate": 6.516909542714591e-05, "loss": 0.0434, "step": 43180 }, { "grad_norm": 0.48391851782798767, "learning_rate": 6.515333919593161e-05, "loss": 0.0369, "step": 43190 }, { "grad_norm": 0.5143284797668457, "learning_rate": 6.513758130757094e-05, "loss": 0.0388, "step": 43200 }, { "grad_norm": 0.41524404287338257, "learning_rate": 6.512182176378713e-05, "loss": 0.0396, "step": 43210 }, { "grad_norm": 0.4084632396697998, "learning_rate": 6.510606056630362e-05, "loss": 0.0353, "step": 43220 }, { "grad_norm": 0.4391481280326843, "learning_rate": 6.509029771684405e-05, "loss": 0.035, "step": 43230 }, { "grad_norm": 0.3168967366218567, "learning_rate": 6.50745332171322e-05, "loss": 0.0373, "step": 43240 }, { "grad_norm": 0.4046175181865692, "learning_rate": 6.505876706889207e-05, "loss": 0.0504, "step": 43250 }, { "grad_norm": 0.34118279814720154, "learning_rate": 6.504299927384781e-05, "loss": 0.0425, "step": 43260 }, { "grad_norm": 0.49975359439849854, "learning_rate": 6.502722983372378e-05, "loss": 0.0365, "step": 43270 }, { "grad_norm": 0.26070982217788696, "learning_rate": 6.501145875024446e-05, "loss": 0.0453, "step": 43280 }, { "grad_norm": 0.39040717482566833, "learning_rate": 6.499568602513459e-05, "loss": 0.0415, "step": 43290 }, { "grad_norm": 0.3732597529888153, "learning_rate": 6.497991166011903e-05, "loss": 0.0341, "step": 43300 }, { "grad_norm": 0.3176945447921753, "learning_rate": 6.49641356569229e-05, "loss": 0.0336, "step": 43310 }, { "grad_norm": 0.296103298664093, "learning_rate": 6.494835801727135e-05, "loss": 0.0353, "step": 43320 }, { "grad_norm": 0.3746738135814667, "learning_rate": 6.493257874288986e-05, "loss": 0.0444, "step": 43330 }, { "grad_norm": 0.42841216921806335, "learning_rate": 6.491679783550399e-05, "loss": 0.0426, "step": 43340 }, { "grad_norm": 0.317505419254303, "learning_rate": 6.490101529683957e-05, "loss": 0.0365, "step": 43350 }, { "grad_norm": 0.2813608944416046, "learning_rate": 6.488523112862248e-05, "loss": 0.0373, "step": 43360 }, { "grad_norm": 0.338857501745224, "learning_rate": 6.486944533257891e-05, "loss": 0.0392, "step": 43370 }, { "grad_norm": 0.46268993616104126, "learning_rate": 6.485365791043517e-05, "loss": 0.0411, "step": 43380 }, { "grad_norm": 0.5034540295600891, "learning_rate": 6.483786886391772e-05, "loss": 0.041, "step": 43390 }, { "grad_norm": 0.32607755064964294, "learning_rate": 6.482207819475323e-05, "loss": 0.0306, "step": 43400 }, { "grad_norm": 0.4368366301059723, "learning_rate": 6.480628590466857e-05, "loss": 0.0286, "step": 43410 }, { "grad_norm": 0.5865004658699036, "learning_rate": 6.479049199539074e-05, "loss": 0.0369, "step": 43420 }, { "grad_norm": 0.3324977457523346, "learning_rate": 6.477469646864693e-05, "loss": 0.04, "step": 43430 }, { "grad_norm": 0.37310320138931274, "learning_rate": 6.475889932616454e-05, "loss": 0.0387, "step": 43440 }, { "grad_norm": 0.40156903862953186, "learning_rate": 6.474310056967111e-05, "loss": 0.0398, "step": 43450 }, { "grad_norm": 0.546958863735199, "learning_rate": 6.472730020089437e-05, "loss": 0.0517, "step": 43460 }, { "grad_norm": 0.34345775842666626, "learning_rate": 6.471149822156222e-05, "loss": 0.0382, "step": 43470 }, { "grad_norm": 1.1715377569198608, "learning_rate": 6.469569463340275e-05, "loss": 0.0346, "step": 43480 }, { "grad_norm": 0.4986702799797058, "learning_rate": 6.467988943814418e-05, "loss": 0.0407, "step": 43490 }, { "grad_norm": 0.4979822635650635, "learning_rate": 6.4664082637515e-05, "loss": 0.0389, "step": 43500 }, { "grad_norm": 0.46314537525177, "learning_rate": 6.464827423324377e-05, "loss": 0.0434, "step": 43510 }, { "grad_norm": 0.3193078935146332, "learning_rate": 6.463246422705931e-05, "loss": 0.0331, "step": 43520 }, { "grad_norm": 0.34550976753234863, "learning_rate": 6.461665262069055e-05, "loss": 0.0456, "step": 43530 }, { "grad_norm": 0.2294691950082779, "learning_rate": 6.460083941586665e-05, "loss": 0.0373, "step": 43540 }, { "grad_norm": 0.5522902607917786, "learning_rate": 6.45850246143169e-05, "loss": 0.0352, "step": 43550 }, { "grad_norm": 0.6081055402755737, "learning_rate": 6.456920821777077e-05, "loss": 0.0397, "step": 43560 }, { "grad_norm": 0.3973067104816437, "learning_rate": 6.455339022795796e-05, "loss": 0.0446, "step": 43570 }, { "grad_norm": 0.46775612235069275, "learning_rate": 6.453757064660825e-05, "loss": 0.035, "step": 43580 }, { "grad_norm": 0.48143091797828674, "learning_rate": 6.452174947545169e-05, "loss": 0.0264, "step": 43590 }, { "grad_norm": 0.3102063536643982, "learning_rate": 6.450592671621842e-05, "loss": 0.0357, "step": 43600 }, { "grad_norm": 0.5371299386024475, "learning_rate": 6.449010237063882e-05, "loss": 0.0411, "step": 43610 }, { "grad_norm": 0.34498172998428345, "learning_rate": 6.447427644044343e-05, "loss": 0.0353, "step": 43620 }, { "grad_norm": 0.3694096505641937, "learning_rate": 6.44584489273629e-05, "loss": 0.0339, "step": 43630 }, { "grad_norm": 0.3823331594467163, "learning_rate": 6.444261983312817e-05, "loss": 0.0367, "step": 43640 }, { "grad_norm": 0.4194476902484894, "learning_rate": 6.442678915947023e-05, "loss": 0.0394, "step": 43650 }, { "grad_norm": 0.46247705817222595, "learning_rate": 6.441095690812032e-05, "loss": 0.0345, "step": 43660 }, { "grad_norm": 0.4489839971065521, "learning_rate": 6.439512308080982e-05, "loss": 0.0363, "step": 43670 }, { "grad_norm": 0.24743907153606415, "learning_rate": 6.43792876792703e-05, "loss": 0.0377, "step": 43680 }, { "grad_norm": 0.35334131121635437, "learning_rate": 6.43634507052335e-05, "loss": 0.0388, "step": 43690 }, { "grad_norm": 0.7302053570747375, "learning_rate": 6.434761216043135e-05, "loss": 0.044, "step": 43700 }, { "grad_norm": 0.6839663982391357, "learning_rate": 6.433177204659589e-05, "loss": 0.0404, "step": 43710 }, { "grad_norm": 0.45585665106773376, "learning_rate": 6.431593036545938e-05, "loss": 0.0386, "step": 43720 }, { "grad_norm": 0.33037978410720825, "learning_rate": 6.430008711875426e-05, "loss": 0.0336, "step": 43730 }, { "grad_norm": 0.8325508236885071, "learning_rate": 6.42842423082131e-05, "loss": 0.0432, "step": 43740 }, { "grad_norm": 0.3164769411087036, "learning_rate": 6.426839593556868e-05, "loss": 0.0313, "step": 43750 }, { "grad_norm": 0.444618284702301, "learning_rate": 6.425254800255392e-05, "loss": 0.0409, "step": 43760 }, { "grad_norm": 0.4604942202568054, "learning_rate": 6.423669851090196e-05, "loss": 0.0428, "step": 43770 }, { "grad_norm": 0.36287161707878113, "learning_rate": 6.422084746234604e-05, "loss": 0.032, "step": 43780 }, { "grad_norm": 0.3204425573348999, "learning_rate": 6.420499485861965e-05, "loss": 0.0275, "step": 43790 }, { "grad_norm": 0.3478287160396576, "learning_rate": 6.418914070145636e-05, "loss": 0.0484, "step": 43800 }, { "grad_norm": 0.46091046929359436, "learning_rate": 6.417328499258996e-05, "loss": 0.0338, "step": 43810 }, { "grad_norm": 0.5571379065513611, "learning_rate": 6.415742773375446e-05, "loss": 0.0335, "step": 43820 }, { "grad_norm": 0.28909623622894287, "learning_rate": 6.414156892668393e-05, "loss": 0.0371, "step": 43830 }, { "grad_norm": 0.3773418664932251, "learning_rate": 6.412570857311267e-05, "loss": 0.0431, "step": 43840 }, { "grad_norm": 0.41067540645599365, "learning_rate": 6.410984667477518e-05, "loss": 0.0418, "step": 43850 }, { "grad_norm": 0.45062902569770813, "learning_rate": 6.409398323340607e-05, "loss": 0.0362, "step": 43860 }, { "grad_norm": 0.35254111886024475, "learning_rate": 6.407811825074012e-05, "loss": 0.0453, "step": 43870 }, { "grad_norm": 0.40608564019203186, "learning_rate": 6.406225172851234e-05, "loss": 0.0329, "step": 43880 }, { "grad_norm": 0.4936346709728241, "learning_rate": 6.404638366845786e-05, "loss": 0.0377, "step": 43890 }, { "grad_norm": 0.4836312532424927, "learning_rate": 6.403051407231196e-05, "loss": 0.0465, "step": 43900 }, { "grad_norm": 0.483038991689682, "learning_rate": 6.401464294181016e-05, "loss": 0.0409, "step": 43910 }, { "grad_norm": 0.3947182297706604, "learning_rate": 6.399877027868808e-05, "loss": 0.0349, "step": 43920 }, { "grad_norm": 0.4509202241897583, "learning_rate": 6.39828960846815e-05, "loss": 0.0402, "step": 43930 }, { "grad_norm": 0.47486013174057007, "learning_rate": 6.396702036152644e-05, "loss": 0.032, "step": 43940 }, { "grad_norm": 0.33321279287338257, "learning_rate": 6.395114311095905e-05, "loss": 0.0471, "step": 43950 }, { "grad_norm": 0.33768561482429504, "learning_rate": 6.393526433471559e-05, "loss": 0.0377, "step": 43960 }, { "grad_norm": 0.35558950901031494, "learning_rate": 6.391938403453262e-05, "loss": 0.0459, "step": 43970 }, { "grad_norm": 0.48477107286453247, "learning_rate": 6.390350221214671e-05, "loss": 0.0447, "step": 43980 }, { "grad_norm": 0.43200933933258057, "learning_rate": 6.38876188692947e-05, "loss": 0.05, "step": 43990 }, { "grad_norm": 0.7712181210517883, "learning_rate": 6.387173400771358e-05, "loss": 0.0396, "step": 44000 }, { "grad_norm": 0.3451445698738098, "learning_rate": 6.385584762914047e-05, "loss": 0.0346, "step": 44010 }, { "grad_norm": 0.24796582758426666, "learning_rate": 6.38399597353127e-05, "loss": 0.0363, "step": 44020 }, { "grad_norm": 0.34722548723220825, "learning_rate": 6.382407032796775e-05, "loss": 0.0279, "step": 44030 }, { "grad_norm": 0.3749784231185913, "learning_rate": 6.380817940884325e-05, "loss": 0.0342, "step": 44040 }, { "grad_norm": 0.7330859899520874, "learning_rate": 6.379228697967702e-05, "loss": 0.0458, "step": 44050 }, { "grad_norm": 0.2959139943122864, "learning_rate": 6.3776393042207e-05, "loss": 0.0374, "step": 44060 }, { "grad_norm": 0.3727472424507141, "learning_rate": 6.376049759817137e-05, "loss": 0.0414, "step": 44070 }, { "grad_norm": 0.2822067141532898, "learning_rate": 6.37446006493084e-05, "loss": 0.0433, "step": 44080 }, { "grad_norm": 0.49153345823287964, "learning_rate": 6.372870219735656e-05, "loss": 0.0487, "step": 44090 }, { "grad_norm": 0.4234342873096466, "learning_rate": 6.371280224405452e-05, "loss": 0.0422, "step": 44100 }, { "grad_norm": 0.4757918417453766, "learning_rate": 6.369690079114101e-05, "loss": 0.0375, "step": 44110 }, { "grad_norm": 0.417949914932251, "learning_rate": 6.368099784035504e-05, "loss": 0.0402, "step": 44120 }, { "grad_norm": 0.32550686597824097, "learning_rate": 6.366509339343572e-05, "loss": 0.035, "step": 44130 }, { "grad_norm": 0.4625767171382904, "learning_rate": 6.364918745212232e-05, "loss": 0.0395, "step": 44140 }, { "grad_norm": 0.552419900894165, "learning_rate": 6.363328001815431e-05, "loss": 0.0375, "step": 44150 }, { "grad_norm": 0.3641272783279419, "learning_rate": 6.361737109327128e-05, "loss": 0.0423, "step": 44160 }, { "grad_norm": 0.49730658531188965, "learning_rate": 6.360146067921301e-05, "loss": 0.0347, "step": 44170 }, { "grad_norm": 1.6267436742782593, "learning_rate": 6.35855487777195e-05, "loss": 0.0451, "step": 44180 }, { "grad_norm": 0.29682567715644836, "learning_rate": 6.356963539053076e-05, "loss": 0.0419, "step": 44190 }, { "grad_norm": 0.38557562232017517, "learning_rate": 6.355372051938711e-05, "loss": 0.038, "step": 44200 }, { "grad_norm": 0.39328086376190186, "learning_rate": 6.353780416602894e-05, "loss": 0.035, "step": 44210 }, { "grad_norm": 0.3284892439842224, "learning_rate": 6.352188633219689e-05, "loss": 0.0348, "step": 44220 }, { "grad_norm": 0.3170050084590912, "learning_rate": 6.350596701963166e-05, "loss": 0.0338, "step": 44230 }, { "grad_norm": 0.3132282793521881, "learning_rate": 6.349004623007419e-05, "loss": 0.0314, "step": 44240 }, { "grad_norm": 0.3772539794445038, "learning_rate": 6.347412396526555e-05, "loss": 0.0387, "step": 44250 }, { "grad_norm": 0.3416070342063904, "learning_rate": 6.345820022694696e-05, "loss": 0.0464, "step": 44260 }, { "grad_norm": 0.5456138253211975, "learning_rate": 6.344227501685984e-05, "loss": 0.036, "step": 44270 }, { "grad_norm": 0.4812580347061157, "learning_rate": 6.342634833674572e-05, "loss": 0.0395, "step": 44280 }, { "grad_norm": 0.49740248918533325, "learning_rate": 6.341042018834635e-05, "loss": 0.0415, "step": 44290 }, { "grad_norm": 0.6687355041503906, "learning_rate": 6.339449057340359e-05, "loss": 0.0398, "step": 44300 }, { "grad_norm": 0.27449649572372437, "learning_rate": 6.337855949365945e-05, "loss": 0.043, "step": 44310 }, { "grad_norm": 0.9031270146369934, "learning_rate": 6.336262695085619e-05, "loss": 0.0382, "step": 44320 }, { "grad_norm": 0.520115315914154, "learning_rate": 6.334669294673612e-05, "loss": 0.0355, "step": 44330 }, { "grad_norm": 0.6241172552108765, "learning_rate": 6.333075748304179e-05, "loss": 0.0448, "step": 44340 }, { "grad_norm": 0.575113832950592, "learning_rate": 6.331482056151585e-05, "loss": 0.0403, "step": 44350 }, { "grad_norm": 0.48698845505714417, "learning_rate": 6.329888218390117e-05, "loss": 0.0425, "step": 44360 }, { "grad_norm": 0.3532971143722534, "learning_rate": 6.328294235194072e-05, "loss": 0.0338, "step": 44370 }, { "grad_norm": 0.2940022349357605, "learning_rate": 6.326700106737765e-05, "loss": 0.0379, "step": 44380 }, { "grad_norm": 1.5594522953033447, "learning_rate": 6.325105833195531e-05, "loss": 0.0446, "step": 44390 }, { "grad_norm": 0.6038364171981812, "learning_rate": 6.323511414741715e-05, "loss": 0.0379, "step": 44400 }, { "grad_norm": 0.4734540283679962, "learning_rate": 6.321916851550678e-05, "loss": 0.0561, "step": 44410 }, { "grad_norm": 0.8588998913764954, "learning_rate": 6.320322143796806e-05, "loss": 0.0529, "step": 44420 }, { "grad_norm": 0.37006714940071106, "learning_rate": 6.318727291654488e-05, "loss": 0.0387, "step": 44430 }, { "grad_norm": 0.43521830439567566, "learning_rate": 6.317132295298134e-05, "loss": 0.0373, "step": 44440 }, { "grad_norm": 0.38437724113464355, "learning_rate": 6.315537154902173e-05, "loss": 0.0322, "step": 44450 }, { "grad_norm": 0.7222694158554077, "learning_rate": 6.313941870641048e-05, "loss": 0.0389, "step": 44460 }, { "grad_norm": 0.37159523367881775, "learning_rate": 6.312346442689214e-05, "loss": 0.0482, "step": 44470 }, { "grad_norm": 0.7004525065422058, "learning_rate": 6.310750871221147e-05, "loss": 0.0372, "step": 44480 }, { "grad_norm": 0.3572731912136078, "learning_rate": 6.309155156411335e-05, "loss": 0.0355, "step": 44490 }, { "grad_norm": 0.3256695866584778, "learning_rate": 6.307559298434284e-05, "loss": 0.0373, "step": 44500 }, { "grad_norm": 0.3940838575363159, "learning_rate": 6.305963297464512e-05, "loss": 0.037, "step": 44510 }, { "grad_norm": 0.3337126076221466, "learning_rate": 6.304367153676561e-05, "loss": 0.0359, "step": 44520 }, { "grad_norm": 2.8291783332824707, "learning_rate": 6.302770867244978e-05, "loss": 0.0364, "step": 44530 }, { "grad_norm": 0.6393617391586304, "learning_rate": 6.301174438344328e-05, "loss": 0.0409, "step": 44540 }, { "grad_norm": 0.4307904839515686, "learning_rate": 6.299577867149202e-05, "loss": 0.0504, "step": 44550 }, { "grad_norm": 0.3419238030910492, "learning_rate": 6.297981153834192e-05, "loss": 0.0357, "step": 44560 }, { "grad_norm": 0.26716145873069763, "learning_rate": 6.296384298573916e-05, "loss": 0.0492, "step": 44570 }, { "grad_norm": 0.3228764832019806, "learning_rate": 6.294787301543001e-05, "loss": 0.0564, "step": 44580 }, { "grad_norm": 0.3966570496559143, "learning_rate": 6.293190162916095e-05, "loss": 0.0372, "step": 44590 }, { "grad_norm": 0.3175210952758789, "learning_rate": 6.291592882867855e-05, "loss": 0.0379, "step": 44600 }, { "grad_norm": 0.5199688076972961, "learning_rate": 6.28999546157296e-05, "loss": 0.0406, "step": 44610 }, { "grad_norm": 0.30806201696395874, "learning_rate": 6.288397899206102e-05, "loss": 0.0385, "step": 44620 }, { "grad_norm": 0.6791965365409851, "learning_rate": 6.286800195941984e-05, "loss": 0.042, "step": 44630 }, { "grad_norm": 0.32255956530570984, "learning_rate": 6.285202351955334e-05, "loss": 0.0362, "step": 44640 }, { "grad_norm": 0.6508880853652954, "learning_rate": 6.283604367420887e-05, "loss": 0.0428, "step": 44650 }, { "grad_norm": 0.29810598492622375, "learning_rate": 6.282006242513394e-05, "loss": 0.0378, "step": 44660 }, { "grad_norm": 0.41658687591552734, "learning_rate": 6.280407977407628e-05, "loss": 0.0403, "step": 44670 }, { "grad_norm": 0.44288432598114014, "learning_rate": 6.27880957227837e-05, "loss": 0.0484, "step": 44680 }, { "grad_norm": 0.4080149233341217, "learning_rate": 6.27721102730042e-05, "loss": 0.043, "step": 44690 }, { "grad_norm": 0.4080697000026703, "learning_rate": 6.275612342648592e-05, "loss": 0.0435, "step": 44700 }, { "grad_norm": 0.5431415438652039, "learning_rate": 6.274013518497716e-05, "loss": 0.0474, "step": 44710 }, { "grad_norm": 0.4544559121131897, "learning_rate": 6.272414555022636e-05, "loss": 0.0368, "step": 44720 }, { "grad_norm": 0.3922833502292633, "learning_rate": 6.270815452398215e-05, "loss": 0.0352, "step": 44730 }, { "grad_norm": 0.7099936008453369, "learning_rate": 6.269216210799326e-05, "loss": 0.0396, "step": 44740 }, { "grad_norm": 0.4616732597351074, "learning_rate": 6.26761683040086e-05, "loss": 0.0502, "step": 44750 }, { "grad_norm": 0.36754029989242554, "learning_rate": 6.266017311377723e-05, "loss": 0.0432, "step": 44760 }, { "grad_norm": 0.35949140787124634, "learning_rate": 6.264417653904839e-05, "loss": 0.0469, "step": 44770 }, { "grad_norm": 0.4981078803539276, "learning_rate": 6.262817858157139e-05, "loss": 0.0438, "step": 44780 }, { "grad_norm": 0.4568711221218109, "learning_rate": 6.261217924309576e-05, "loss": 0.0548, "step": 44790 }, { "grad_norm": 0.4104995131492615, "learning_rate": 6.259617852537118e-05, "loss": 0.0442, "step": 44800 }, { "grad_norm": 0.4413048326969147, "learning_rate": 6.258017643014747e-05, "loss": 0.0365, "step": 44810 }, { "grad_norm": 0.40059491991996765, "learning_rate": 6.256417295917456e-05, "loss": 0.046, "step": 44820 }, { "grad_norm": 0.5001506209373474, "learning_rate": 6.254816811420258e-05, "loss": 0.0505, "step": 44830 }, { "grad_norm": 0.6178413033485413, "learning_rate": 6.253216189698183e-05, "loss": 0.0427, "step": 44840 }, { "grad_norm": 0.26278242468833923, "learning_rate": 6.251615430926267e-05, "loss": 0.0333, "step": 44850 }, { "grad_norm": 0.37823599576950073, "learning_rate": 6.25001453527957e-05, "loss": 0.0362, "step": 44860 }, { "grad_norm": 0.505957841873169, "learning_rate": 6.248413502933164e-05, "loss": 0.0378, "step": 44870 }, { "grad_norm": 0.3204711079597473, "learning_rate": 6.246812334062133e-05, "loss": 0.0324, "step": 44880 }, { "grad_norm": 0.4512268602848053, "learning_rate": 6.245211028841579e-05, "loss": 0.0448, "step": 44890 }, { "grad_norm": 0.31920310854911804, "learning_rate": 6.24360958744662e-05, "loss": 0.0381, "step": 44900 }, { "grad_norm": 0.2954276204109192, "learning_rate": 6.242008010052387e-05, "loss": 0.0398, "step": 44910 }, { "grad_norm": 0.4394480586051941, "learning_rate": 6.240406296834024e-05, "loss": 0.0453, "step": 44920 }, { "grad_norm": 0.3615569472312927, "learning_rate": 6.238804447966694e-05, "loss": 0.036, "step": 44930 }, { "grad_norm": 0.5204386115074158, "learning_rate": 6.237202463625573e-05, "loss": 0.0351, "step": 44940 }, { "grad_norm": 0.3712295889854431, "learning_rate": 6.235600343985848e-05, "loss": 0.043, "step": 44950 }, { "grad_norm": 0.37312695384025574, "learning_rate": 6.233998089222729e-05, "loss": 0.032, "step": 44960 }, { "grad_norm": 0.3050451874732971, "learning_rate": 6.232395699511433e-05, "loss": 0.0386, "step": 44970 }, { "grad_norm": 0.34557902812957764, "learning_rate": 6.230793175027199e-05, "loss": 0.0493, "step": 44980 }, { "grad_norm": 0.3771951198577881, "learning_rate": 6.22919051594527e-05, "loss": 0.0369, "step": 44990 }, { "grad_norm": 0.4263632595539093, "learning_rate": 6.227587722440917e-05, "loss": 0.0466, "step": 45000 }, { "grad_norm": 0.2740092873573303, "learning_rate": 6.225984794689414e-05, "loss": 0.0346, "step": 45010 }, { "grad_norm": 0.3311253786087036, "learning_rate": 6.22438173286606e-05, "loss": 0.034, "step": 45020 }, { "grad_norm": 0.4648277759552002, "learning_rate": 6.222778537146159e-05, "loss": 0.0366, "step": 45030 }, { "grad_norm": 0.390077143907547, "learning_rate": 6.221175207705037e-05, "loss": 0.0394, "step": 45040 }, { "grad_norm": 0.5229597091674805, "learning_rate": 6.219571744718027e-05, "loss": 0.0351, "step": 45050 }, { "grad_norm": 0.25275808572769165, "learning_rate": 6.217968148360487e-05, "loss": 0.0362, "step": 45060 }, { "grad_norm": 0.27316486835479736, "learning_rate": 6.216364418807782e-05, "loss": 0.0374, "step": 45070 }, { "grad_norm": 0.432984858751297, "learning_rate": 6.214760556235292e-05, "loss": 0.0337, "step": 45080 }, { "grad_norm": 0.39624708890914917, "learning_rate": 6.213156560818414e-05, "loss": 0.0365, "step": 45090 }, { "grad_norm": 0.43670955300331116, "learning_rate": 6.211552432732559e-05, "loss": 0.0372, "step": 45100 }, { "grad_norm": 0.4766484797000885, "learning_rate": 6.20994817215315e-05, "loss": 0.0391, "step": 45110 }, { "grad_norm": 0.32523685693740845, "learning_rate": 6.208343779255628e-05, "loss": 0.0376, "step": 45120 }, { "grad_norm": 0.5046015381813049, "learning_rate": 6.206739254215449e-05, "loss": 0.042, "step": 45130 }, { "grad_norm": 0.47157788276672363, "learning_rate": 6.205134597208077e-05, "loss": 0.0435, "step": 45140 }, { "grad_norm": 0.3157718777656555, "learning_rate": 6.203529808409e-05, "loss": 0.0392, "step": 45150 }, { "grad_norm": 0.5680859684944153, "learning_rate": 6.201924887993712e-05, "loss": 0.0429, "step": 45160 }, { "grad_norm": 0.44256967306137085, "learning_rate": 6.200319836137725e-05, "loss": 0.0459, "step": 45170 }, { "grad_norm": 0.2962413728237152, "learning_rate": 6.198714653016565e-05, "loss": 0.0451, "step": 45180 }, { "grad_norm": 0.3122231662273407, "learning_rate": 6.197109338805774e-05, "loss": 0.0406, "step": 45190 }, { "grad_norm": 0.4052460491657257, "learning_rate": 6.195503893680903e-05, "loss": 0.0428, "step": 45200 }, { "grad_norm": 0.3301441967487335, "learning_rate": 6.193898317817524e-05, "loss": 0.0376, "step": 45210 }, { "grad_norm": 0.5978981256484985, "learning_rate": 6.192292611391221e-05, "loss": 0.0297, "step": 45220 }, { "grad_norm": 0.24264825880527496, "learning_rate": 6.190686774577591e-05, "loss": 0.0323, "step": 45230 }, { "grad_norm": 0.2877923846244812, "learning_rate": 6.189080807552245e-05, "loss": 0.0432, "step": 45240 }, { "grad_norm": 0.3718695342540741, "learning_rate": 6.187474710490809e-05, "loss": 0.037, "step": 45250 }, { "grad_norm": 0.45526206493377686, "learning_rate": 6.185868483568926e-05, "loss": 0.0379, "step": 45260 }, { "grad_norm": 0.3883046805858612, "learning_rate": 6.184262126962245e-05, "loss": 0.0364, "step": 45270 }, { "grad_norm": 0.4077805280685425, "learning_rate": 6.182655640846442e-05, "loss": 0.0274, "step": 45280 }, { "grad_norm": 0.3397922217845917, "learning_rate": 6.181049025397196e-05, "loss": 0.0374, "step": 45290 }, { "grad_norm": 0.47001752257347107, "learning_rate": 6.179442280790202e-05, "loss": 0.0431, "step": 45300 }, { "grad_norm": 0.4395889341831207, "learning_rate": 6.177835407201174e-05, "loss": 0.0476, "step": 45310 }, { "grad_norm": 0.4184703230857849, "learning_rate": 6.176228404805839e-05, "loss": 0.0358, "step": 45320 }, { "grad_norm": 0.5060072541236877, "learning_rate": 6.174621273779932e-05, "loss": 0.0404, "step": 45330 }, { "grad_norm": 0.3997827470302582, "learning_rate": 6.17301401429921e-05, "loss": 0.0382, "step": 45340 }, { "grad_norm": 0.3172552287578583, "learning_rate": 6.171406626539436e-05, "loss": 0.0371, "step": 45350 }, { "grad_norm": 0.614387035369873, "learning_rate": 6.169799110676398e-05, "loss": 0.0537, "step": 45360 }, { "grad_norm": 0.44704195857048035, "learning_rate": 6.168191466885885e-05, "loss": 0.0372, "step": 45370 }, { "grad_norm": 0.31552067399024963, "learning_rate": 6.166583695343711e-05, "loss": 0.0348, "step": 45380 }, { "grad_norm": 0.3171527087688446, "learning_rate": 6.164975796225698e-05, "loss": 0.0414, "step": 45390 }, { "grad_norm": 0.505714476108551, "learning_rate": 6.163367769707683e-05, "loss": 0.0334, "step": 45400 }, { "grad_norm": 0.44864463806152344, "learning_rate": 6.161759615965519e-05, "loss": 0.039, "step": 45410 }, { "grad_norm": 0.3919982314109802, "learning_rate": 6.16015133517507e-05, "loss": 0.0333, "step": 45420 }, { "grad_norm": 0.33981040120124817, "learning_rate": 6.158542927512214e-05, "loss": 0.0288, "step": 45430 }, { "grad_norm": 0.3420933783054352, "learning_rate": 6.156934393152846e-05, "loss": 0.0347, "step": 45440 }, { "grad_norm": 0.2893991768360138, "learning_rate": 6.15532573227287e-05, "loss": 0.0449, "step": 45450 }, { "grad_norm": 0.6308883428573608, "learning_rate": 6.153716945048212e-05, "loss": 0.0574, "step": 45460 }, { "grad_norm": 0.4684719443321228, "learning_rate": 6.152108031654802e-05, "loss": 0.0471, "step": 45470 }, { "grad_norm": 0.329274445772171, "learning_rate": 6.15049899226859e-05, "loss": 0.0386, "step": 45480 }, { "grad_norm": 0.2817791700363159, "learning_rate": 6.148889827065537e-05, "loss": 0.0389, "step": 45490 }, { "grad_norm": 0.4412148594856262, "learning_rate": 6.147280536221622e-05, "loss": 0.0411, "step": 45500 }, { "grad_norm": 0.5228316783905029, "learning_rate": 6.145671119912832e-05, "loss": 0.0372, "step": 45510 }, { "grad_norm": 0.3700640797615051, "learning_rate": 6.144061578315169e-05, "loss": 0.0363, "step": 45520 }, { "grad_norm": 0.32119855284690857, "learning_rate": 6.142451911604654e-05, "loss": 0.0377, "step": 45530 }, { "grad_norm": 0.6177040338516235, "learning_rate": 6.140842119957315e-05, "loss": 0.0512, "step": 45540 }, { "grad_norm": 0.49751025438308716, "learning_rate": 6.139232203549197e-05, "loss": 0.0372, "step": 45550 }, { "grad_norm": 0.3841286301612854, "learning_rate": 6.137622162556357e-05, "loss": 0.0397, "step": 45560 }, { "grad_norm": 0.48038381338119507, "learning_rate": 6.13601199715487e-05, "loss": 0.0357, "step": 45570 }, { "grad_norm": 0.43722832202911377, "learning_rate": 6.134401707520816e-05, "loss": 0.045, "step": 45580 }, { "grad_norm": 0.2811677157878876, "learning_rate": 6.132791293830298e-05, "loss": 0.0324, "step": 45590 }, { "grad_norm": 0.31166383624076843, "learning_rate": 6.131180756259428e-05, "loss": 0.039, "step": 45600 }, { "grad_norm": 0.4353931248188019, "learning_rate": 6.129570094984331e-05, "loss": 0.0287, "step": 45610 }, { "grad_norm": 0.4676922559738159, "learning_rate": 6.127959310181145e-05, "loss": 0.0456, "step": 45620 }, { "grad_norm": 0.36887189745903015, "learning_rate": 6.126348402026026e-05, "loss": 0.0382, "step": 45630 }, { "grad_norm": 0.358590304851532, "learning_rate": 6.12473737069514e-05, "loss": 0.0306, "step": 45640 }, { "grad_norm": 0.38029876351356506, "learning_rate": 6.123126216364665e-05, "loss": 0.0454, "step": 45650 }, { "grad_norm": 0.4064992666244507, "learning_rate": 6.121514939210797e-05, "loss": 0.0346, "step": 45660 }, { "grad_norm": 0.6118716597557068, "learning_rate": 6.119903539409741e-05, "loss": 0.05, "step": 45670 }, { "grad_norm": 0.2984245717525482, "learning_rate": 6.118292017137716e-05, "loss": 0.0427, "step": 45680 }, { "grad_norm": 0.6233288645744324, "learning_rate": 6.116680372570959e-05, "loss": 0.0483, "step": 45690 }, { "grad_norm": 0.37252622842788696, "learning_rate": 6.115068605885713e-05, "loss": 0.0317, "step": 45700 }, { "grad_norm": 0.3857985734939575, "learning_rate": 6.113456717258243e-05, "loss": 0.0444, "step": 45710 }, { "grad_norm": 0.36646661162376404, "learning_rate": 6.11184470686482e-05, "loss": 0.0403, "step": 45720 }, { "grad_norm": 0.5530648827552795, "learning_rate": 6.110232574881731e-05, "loss": 0.036, "step": 45730 }, { "grad_norm": 0.42790132761001587, "learning_rate": 6.108620321485277e-05, "loss": 0.0455, "step": 45740 }, { "grad_norm": 0.38051638007164, "learning_rate": 6.107007946851773e-05, "loss": 0.0378, "step": 45750 }, { "grad_norm": 0.2521038353443146, "learning_rate": 6.105395451157542e-05, "loss": 0.04, "step": 45760 }, { "grad_norm": 0.40545642375946045, "learning_rate": 6.103782834578928e-05, "loss": 0.0349, "step": 45770 }, { "grad_norm": 0.34027954936027527, "learning_rate": 6.102170097292281e-05, "loss": 0.0383, "step": 45780 }, { "grad_norm": 0.44766461849212646, "learning_rate": 6.1005572394739716e-05, "loss": 0.046, "step": 45790 }, { "grad_norm": 1.0321601629257202, "learning_rate": 6.0989442613003765e-05, "loss": 0.0272, "step": 45800 }, { "grad_norm": 0.905096709728241, "learning_rate": 6.097331162947888e-05, "loss": 0.0464, "step": 45810 }, { "grad_norm": 0.4382127821445465, "learning_rate": 6.095717944592914e-05, "loss": 0.0451, "step": 45820 }, { "grad_norm": 0.9078747630119324, "learning_rate": 6.094104606411873e-05, "loss": 0.0494, "step": 45830 }, { "grad_norm": 0.49349164962768555, "learning_rate": 6.0924911485811966e-05, "loss": 0.0466, "step": 45840 }, { "grad_norm": 0.6704158186912537, "learning_rate": 6.090877571277331e-05, "loss": 0.046, "step": 45850 }, { "grad_norm": 0.30472853779792786, "learning_rate": 6.089263874676736e-05, "loss": 0.0388, "step": 45860 }, { "grad_norm": 0.7991026043891907, "learning_rate": 6.0876500589558796e-05, "loss": 0.0391, "step": 45870 }, { "grad_norm": 0.6572088599205017, "learning_rate": 6.086036124291248e-05, "loss": 0.0364, "step": 45880 }, { "grad_norm": 0.3459486961364746, "learning_rate": 6.084422070859339e-05, "loss": 0.0413, "step": 45890 }, { "grad_norm": 0.5437245965003967, "learning_rate": 6.082807898836663e-05, "loss": 0.0424, "step": 45900 }, { "grad_norm": 0.3242161273956299, "learning_rate": 6.081193608399742e-05, "loss": 0.0355, "step": 45910 }, { "grad_norm": 0.41960424184799194, "learning_rate": 6.0795791997251164e-05, "loss": 0.0389, "step": 45920 }, { "grad_norm": 0.3509552776813507, "learning_rate": 6.0779646729893294e-05, "loss": 0.0314, "step": 45930 }, { "grad_norm": 0.2789027988910675, "learning_rate": 6.0763500283689476e-05, "loss": 0.0299, "step": 45940 }, { "grad_norm": 0.2817000150680542, "learning_rate": 6.0747352660405455e-05, "loss": 0.0378, "step": 45950 }, { "grad_norm": 0.34069353342056274, "learning_rate": 6.073120386180709e-05, "loss": 0.0355, "step": 45960 }, { "grad_norm": 0.4730643630027771, "learning_rate": 6.0715053889660425e-05, "loss": 0.0331, "step": 45970 }, { "grad_norm": 0.3280617594718933, "learning_rate": 6.069890274573157e-05, "loss": 0.0361, "step": 45980 }, { "grad_norm": 0.38331297039985657, "learning_rate": 6.068275043178679e-05, "loss": 0.0451, "step": 45990 }, { "grad_norm": 0.7488261461257935, "learning_rate": 6.066659694959248e-05, "loss": 0.0321, "step": 46000 }, { "grad_norm": 0.27571776509284973, "learning_rate": 6.0650442300915176e-05, "loss": 0.0353, "step": 46010 }, { "grad_norm": 0.3117677867412567, "learning_rate": 6.063428648752152e-05, "loss": 0.0368, "step": 46020 }, { "grad_norm": 0.5305293202400208, "learning_rate": 6.0618129511178266e-05, "loss": 0.04, "step": 46030 }, { "grad_norm": 0.26989078521728516, "learning_rate": 6.060197137365234e-05, "loss": 0.0448, "step": 46040 }, { "grad_norm": 0.40517911314964294, "learning_rate": 6.058581207671077e-05, "loss": 0.0319, "step": 46050 }, { "grad_norm": 0.316719114780426, "learning_rate": 6.056965162212072e-05, "loss": 0.0403, "step": 46060 }, { "grad_norm": 0.2848305106163025, "learning_rate": 6.055349001164943e-05, "loss": 0.0377, "step": 46070 }, { "grad_norm": 0.3355713486671448, "learning_rate": 6.0537327247064347e-05, "loss": 0.04, "step": 46080 }, { "grad_norm": 0.6478321552276611, "learning_rate": 6.0521163330133014e-05, "loss": 0.0464, "step": 46090 }, { "grad_norm": 0.40181541442871094, "learning_rate": 6.050499826262306e-05, "loss": 0.0387, "step": 46100 }, { "grad_norm": 0.4586571156978607, "learning_rate": 6.0488832046302294e-05, "loss": 0.044, "step": 46110 }, { "grad_norm": 0.3311665952205658, "learning_rate": 6.0472664682938626e-05, "loss": 0.0408, "step": 46120 }, { "grad_norm": 0.24945399165153503, "learning_rate": 6.045649617430009e-05, "loss": 0.0399, "step": 46130 }, { "grad_norm": 0.32840269804000854, "learning_rate": 6.0440326522154866e-05, "loss": 0.0376, "step": 46140 }, { "grad_norm": 0.4209662675857544, "learning_rate": 6.0424155728271224e-05, "loss": 0.0424, "step": 46150 }, { "grad_norm": 0.26242896914482117, "learning_rate": 6.040798379441758e-05, "loss": 0.0421, "step": 46160 }, { "grad_norm": 0.27964672446250916, "learning_rate": 6.0391810722362485e-05, "loss": 0.0344, "step": 46170 }, { "grad_norm": 0.3434438109397888, "learning_rate": 6.037563651387458e-05, "loss": 0.0321, "step": 46180 }, { "grad_norm": 0.4800651967525482, "learning_rate": 6.0359461170722666e-05, "loss": 0.0361, "step": 46190 }, { "grad_norm": 0.3806573450565338, "learning_rate": 6.034328469467566e-05, "loss": 0.0362, "step": 46200 }, { "grad_norm": 0.37391528487205505, "learning_rate": 6.0327107087502596e-05, "loss": 0.0377, "step": 46210 }, { "grad_norm": 0.4539305567741394, "learning_rate": 6.031092835097262e-05, "loss": 0.0386, "step": 46220 }, { "grad_norm": 0.296528160572052, "learning_rate": 6.0294748486855024e-05, "loss": 0.0391, "step": 46230 }, { "grad_norm": 0.3382914364337921, "learning_rate": 6.0278567496919216e-05, "loss": 0.0438, "step": 46240 }, { "grad_norm": 0.33805790543556213, "learning_rate": 6.026238538293472e-05, "loss": 0.0394, "step": 46250 }, { "grad_norm": 0.34278082847595215, "learning_rate": 6.024620214667118e-05, "loss": 0.0417, "step": 46260 }, { "grad_norm": 0.31532129645347595, "learning_rate": 6.0230017789898384e-05, "loss": 0.038, "step": 46270 }, { "grad_norm": 0.28812554478645325, "learning_rate": 6.021383231438622e-05, "loss": 0.0372, "step": 46280 }, { "grad_norm": 0.36934927105903625, "learning_rate": 6.0197645721904704e-05, "loss": 0.0368, "step": 46290 }, { "grad_norm": 0.38165929913520813, "learning_rate": 6.0181458014224e-05, "loss": 0.0424, "step": 46300 }, { "grad_norm": 0.43138641119003296, "learning_rate": 6.016526919311435e-05, "loss": 0.0374, "step": 46310 }, { "grad_norm": 0.3012324571609497, "learning_rate": 6.014907926034613e-05, "loss": 0.0391, "step": 46320 }, { "grad_norm": 0.4796586036682129, "learning_rate": 6.0132888217689875e-05, "loss": 0.036, "step": 46330 }, { "grad_norm": 0.4809533655643463, "learning_rate": 6.01166960669162e-05, "loss": 0.0413, "step": 46340 }, { "grad_norm": 0.4231330156326294, "learning_rate": 6.0100502809795845e-05, "loss": 0.0368, "step": 46350 }, { "grad_norm": 0.2914831042289734, "learning_rate": 6.00843084480997e-05, "loss": 0.0363, "step": 46360 }, { "grad_norm": 0.27555954456329346, "learning_rate": 6.0068112983598736e-05, "loss": 0.0299, "step": 46370 }, { "grad_norm": 0.35403504967689514, "learning_rate": 6.005191641806407e-05, "loss": 0.036, "step": 46380 }, { "grad_norm": 0.4477801024913788, "learning_rate": 6.003571875326694e-05, "loss": 0.0381, "step": 46390 }, { "grad_norm": 0.3651118278503418, "learning_rate": 6.00195199909787e-05, "loss": 0.0398, "step": 46400 }, { "grad_norm": 0.3619047999382019, "learning_rate": 6.000332013297082e-05, "loss": 0.04, "step": 46410 }, { "grad_norm": 0.29453974962234497, "learning_rate": 5.998711918101487e-05, "loss": 0.0378, "step": 46420 }, { "grad_norm": 0.3554579019546509, "learning_rate": 5.997091713688261e-05, "loss": 0.0445, "step": 46430 }, { "grad_norm": 0.24380278587341309, "learning_rate": 5.9954714002345836e-05, "loss": 0.035, "step": 46440 }, { "grad_norm": 0.4532955288887024, "learning_rate": 5.993850977917649e-05, "loss": 0.042, "step": 46450 }, { "grad_norm": 0.3481871485710144, "learning_rate": 5.992230446914667e-05, "loss": 0.0375, "step": 46460 }, { "grad_norm": 0.3077324628829956, "learning_rate": 5.990609807402855e-05, "loss": 0.049, "step": 46470 }, { "grad_norm": 0.4163370728492737, "learning_rate": 5.988989059559443e-05, "loss": 0.0412, "step": 46480 }, { "grad_norm": 0.34904131293296814, "learning_rate": 5.987368203561675e-05, "loss": 0.0422, "step": 46490 }, { "grad_norm": 0.38000622391700745, "learning_rate": 5.9857472395868055e-05, "loss": 0.0388, "step": 46500 }, { "grad_norm": 0.4752963185310364, "learning_rate": 5.9841261678120983e-05, "loss": 0.0377, "step": 46510 }, { "grad_norm": 0.31896090507507324, "learning_rate": 5.982504988414834e-05, "loss": 0.0327, "step": 46520 }, { "grad_norm": 0.4893946647644043, "learning_rate": 5.9808837015723015e-05, "loss": 0.0519, "step": 46530 }, { "grad_norm": 0.3839085102081299, "learning_rate": 5.9792623074618016e-05, "loss": 0.0379, "step": 46540 }, { "grad_norm": 0.38073766231536865, "learning_rate": 5.977640806260648e-05, "loss": 0.0401, "step": 46550 }, { "grad_norm": 0.35577186942100525, "learning_rate": 5.976019198146166e-05, "loss": 0.0386, "step": 46560 }, { "grad_norm": 0.7733798623085022, "learning_rate": 5.974397483295692e-05, "loss": 0.0344, "step": 46570 }, { "grad_norm": 0.28609901666641235, "learning_rate": 5.972775661886575e-05, "loss": 0.0376, "step": 46580 }, { "grad_norm": 0.5803942084312439, "learning_rate": 5.971153734096173e-05, "loss": 0.0448, "step": 46590 }, { "grad_norm": 0.43578916788101196, "learning_rate": 5.969531700101859e-05, "loss": 0.0358, "step": 46600 }, { "grad_norm": 0.9977196455001831, "learning_rate": 5.9679095600810155e-05, "loss": 0.0462, "step": 46610 }, { "grad_norm": 0.512905478477478, "learning_rate": 5.9662873142110384e-05, "loss": 0.0361, "step": 46620 }, { "grad_norm": 0.8918631672859192, "learning_rate": 5.964664962669333e-05, "loss": 0.0403, "step": 46630 }, { "grad_norm": 0.3912805914878845, "learning_rate": 5.9630425056333186e-05, "loss": 0.0343, "step": 46640 }, { "grad_norm": 0.22248369455337524, "learning_rate": 5.961419943280422e-05, "loss": 0.0362, "step": 46650 }, { "grad_norm": 0.3455159664154053, "learning_rate": 5.959797275788087e-05, "loss": 0.0299, "step": 46660 }, { "grad_norm": 0.3310767412185669, "learning_rate": 5.958174503333765e-05, "loss": 0.0473, "step": 46670 }, { "grad_norm": 0.4309699535369873, "learning_rate": 5.9565516260949195e-05, "loss": 0.0473, "step": 46680 }, { "grad_norm": 0.4932928681373596, "learning_rate": 5.9549286442490273e-05, "loss": 0.0441, "step": 46690 }, { "grad_norm": 0.4130585789680481, "learning_rate": 5.953305557973572e-05, "loss": 0.0421, "step": 46700 }, { "grad_norm": 0.5548020601272583, "learning_rate": 5.9516823674460564e-05, "loss": 0.0421, "step": 46710 }, { "grad_norm": 0.5503128170967102, "learning_rate": 5.9500590728439875e-05, "loss": 0.0356, "step": 46720 }, { "grad_norm": 0.39106762409210205, "learning_rate": 5.948435674344886e-05, "loss": 0.0405, "step": 46730 }, { "grad_norm": 0.5509751439094543, "learning_rate": 5.946812172126285e-05, "loss": 0.035, "step": 46740 }, { "grad_norm": 0.4850744903087616, "learning_rate": 5.945188566365729e-05, "loss": 0.0292, "step": 46750 }, { "grad_norm": 0.3173723816871643, "learning_rate": 5.943564857240773e-05, "loss": 0.0369, "step": 46760 }, { "grad_norm": 0.7253363132476807, "learning_rate": 5.941941044928981e-05, "loss": 0.0383, "step": 46770 }, { "grad_norm": 0.46033400297164917, "learning_rate": 5.940317129607935e-05, "loss": 0.0374, "step": 46780 }, { "grad_norm": 0.36154234409332275, "learning_rate": 5.9386931114552204e-05, "loss": 0.043, "step": 46790 }, { "grad_norm": 0.3560892641544342, "learning_rate": 5.937068990648438e-05, "loss": 0.0363, "step": 46800 }, { "grad_norm": 0.6290460228919983, "learning_rate": 5.935444767365199e-05, "loss": 0.058, "step": 46810 }, { "grad_norm": 0.36108627915382385, "learning_rate": 5.933820441783129e-05, "loss": 0.0433, "step": 46820 }, { "grad_norm": 0.3285008668899536, "learning_rate": 5.932196014079857e-05, "loss": 0.0404, "step": 46830 }, { "grad_norm": 0.2806098759174347, "learning_rate": 5.930571484433032e-05, "loss": 0.0287, "step": 46840 }, { "grad_norm": 0.2896307706832886, "learning_rate": 5.928946853020309e-05, "loss": 0.0299, "step": 46850 }, { "grad_norm": 0.18984977900981903, "learning_rate": 5.927322120019352e-05, "loss": 0.0309, "step": 46860 }, { "grad_norm": 0.8163648247718811, "learning_rate": 5.925697285607844e-05, "loss": 0.0444, "step": 46870 }, { "grad_norm": 0.4931448996067047, "learning_rate": 5.924072349963472e-05, "loss": 0.039, "step": 46880 }, { "grad_norm": 0.4725797772407532, "learning_rate": 5.922447313263938e-05, "loss": 0.0326, "step": 46890 }, { "grad_norm": 0.3590245842933655, "learning_rate": 5.920822175686952e-05, "loss": 0.0412, "step": 46900 }, { "grad_norm": 0.5389986634254456, "learning_rate": 5.919196937410239e-05, "loss": 0.0435, "step": 46910 }, { "grad_norm": 0.40423667430877686, "learning_rate": 5.917571598611529e-05, "loss": 0.0396, "step": 46920 }, { "grad_norm": 0.5859786868095398, "learning_rate": 5.9159461594685686e-05, "loss": 0.0405, "step": 46930 }, { "grad_norm": 0.33666202425956726, "learning_rate": 5.914320620159114e-05, "loss": 0.0361, "step": 46940 }, { "grad_norm": 0.5241686701774597, "learning_rate": 5.912694980860931e-05, "loss": 0.0429, "step": 46950 }, { "grad_norm": 0.33527299761772156, "learning_rate": 5.9110692417517964e-05, "loss": 0.0298, "step": 46960 }, { "grad_norm": 0.38319799304008484, "learning_rate": 5.9094434030095e-05, "loss": 0.0405, "step": 46970 }, { "grad_norm": 0.36094924807548523, "learning_rate": 5.907817464811841e-05, "loss": 0.037, "step": 46980 }, { "grad_norm": 0.35558634996414185, "learning_rate": 5.906191427336627e-05, "loss": 0.0401, "step": 46990 }, { "grad_norm": 0.6315121650695801, "learning_rate": 5.9045652907616835e-05, "loss": 0.0422, "step": 47000 }, { "grad_norm": 0.4422016441822052, "learning_rate": 5.902939055264838e-05, "loss": 0.0405, "step": 47010 }, { "grad_norm": 1.3220840692520142, "learning_rate": 5.901312721023935e-05, "loss": 0.0373, "step": 47020 }, { "grad_norm": 0.23604097962379456, "learning_rate": 5.8996862882168294e-05, "loss": 0.0314, "step": 47030 }, { "grad_norm": 0.4707411825656891, "learning_rate": 5.8980597570213826e-05, "loss": 0.0423, "step": 47040 }, { "grad_norm": 0.6372202038764954, "learning_rate": 5.896433127615471e-05, "loss": 0.0499, "step": 47050 }, { "grad_norm": 0.40672287344932556, "learning_rate": 5.894806400176981e-05, "loss": 0.0341, "step": 47060 }, { "grad_norm": 0.5479748249053955, "learning_rate": 5.893179574883808e-05, "loss": 0.0348, "step": 47070 }, { "grad_norm": 0.3764338791370392, "learning_rate": 5.8915526519138585e-05, "loss": 0.0335, "step": 47080 }, { "grad_norm": 0.8083048462867737, "learning_rate": 5.889925631445053e-05, "loss": 0.043, "step": 47090 }, { "grad_norm": 0.4064640998840332, "learning_rate": 5.888298513655318e-05, "loss": 0.033, "step": 47100 }, { "grad_norm": 0.3342471420764923, "learning_rate": 5.886671298722592e-05, "loss": 0.0377, "step": 47110 }, { "grad_norm": 0.2363896369934082, "learning_rate": 5.885043986824828e-05, "loss": 0.0356, "step": 47120 }, { "grad_norm": 0.5563299059867859, "learning_rate": 5.883416578139982e-05, "loss": 0.032, "step": 47130 }, { "grad_norm": 0.27860313653945923, "learning_rate": 5.881789072846029e-05, "loss": 0.0376, "step": 47140 }, { "grad_norm": 0.38544318079948425, "learning_rate": 5.8801614711209474e-05, "loss": 0.0266, "step": 47150 }, { "grad_norm": 0.4224108159542084, "learning_rate": 5.878533773142732e-05, "loss": 0.039, "step": 47160 }, { "grad_norm": 0.5480400323867798, "learning_rate": 5.8769059790893834e-05, "loss": 0.0408, "step": 47170 }, { "grad_norm": 0.3699833154678345, "learning_rate": 5.8752780891389145e-05, "loss": 0.0393, "step": 47180 }, { "grad_norm": 0.30805703997612, "learning_rate": 5.873650103469351e-05, "loss": 0.0355, "step": 47190 }, { "grad_norm": 0.292294979095459, "learning_rate": 5.872022022258725e-05, "loss": 0.0266, "step": 47200 }, { "grad_norm": 0.3960845470428467, "learning_rate": 5.8703938456850814e-05, "loss": 0.0311, "step": 47210 }, { "grad_norm": 0.41739439964294434, "learning_rate": 5.868765573926477e-05, "loss": 0.0421, "step": 47220 }, { "grad_norm": 0.3802712559700012, "learning_rate": 5.867137207160974e-05, "loss": 0.0451, "step": 47230 }, { "grad_norm": 0.40698400139808655, "learning_rate": 5.86550874556665e-05, "loss": 0.0394, "step": 47240 }, { "grad_norm": 0.3504493236541748, "learning_rate": 5.863880189321592e-05, "loss": 0.0323, "step": 47250 }, { "grad_norm": 0.5599071383476257, "learning_rate": 5.862251538603894e-05, "loss": 0.0307, "step": 47260 }, { "grad_norm": 0.293271005153656, "learning_rate": 5.8606227935916656e-05, "loss": 0.0298, "step": 47270 }, { "grad_norm": 0.41317206621170044, "learning_rate": 5.858993954463021e-05, "loss": 0.0343, "step": 47280 }, { "grad_norm": 0.5547614097595215, "learning_rate": 5.8573650213960896e-05, "loss": 0.0437, "step": 47290 }, { "grad_norm": 0.365498423576355, "learning_rate": 5.8557359945690084e-05, "loss": 0.038, "step": 47300 }, { "grad_norm": 0.31830158829689026, "learning_rate": 5.8541068741599246e-05, "loss": 0.0345, "step": 47310 }, { "grad_norm": 0.3795385956764221, "learning_rate": 5.8524776603469985e-05, "loss": 0.0308, "step": 47320 }, { "grad_norm": 0.3461170196533203, "learning_rate": 5.850848353308397e-05, "loss": 0.0359, "step": 47330 }, { "grad_norm": 0.991365909576416, "learning_rate": 5.849218953222297e-05, "loss": 0.0329, "step": 47340 }, { "grad_norm": 0.3383646011352539, "learning_rate": 5.847589460266891e-05, "loss": 0.0399, "step": 47350 }, { "grad_norm": 0.34827616810798645, "learning_rate": 5.845959874620375e-05, "loss": 0.0358, "step": 47360 }, { "grad_norm": 0.5352479815483093, "learning_rate": 5.84433019646096e-05, "loss": 0.0413, "step": 47370 }, { "grad_norm": 0.36990582942962646, "learning_rate": 5.842700425966863e-05, "loss": 0.0386, "step": 47380 }, { "grad_norm": 0.4297504723072052, "learning_rate": 5.841070563316315e-05, "loss": 0.0381, "step": 47390 }, { "grad_norm": 0.35933032631874084, "learning_rate": 5.839440608687554e-05, "loss": 0.0435, "step": 47400 }, { "grad_norm": 0.36445650458335876, "learning_rate": 5.837810562258831e-05, "loss": 0.0311, "step": 47410 }, { "grad_norm": 0.4230445921421051, "learning_rate": 5.836180424208405e-05, "loss": 0.0516, "step": 47420 }, { "grad_norm": 0.2885145843029022, "learning_rate": 5.834550194714543e-05, "loss": 0.0385, "step": 47430 }, { "grad_norm": 0.4638286232948303, "learning_rate": 5.832919873955528e-05, "loss": 0.054, "step": 47440 }, { "grad_norm": 0.42417946457862854, "learning_rate": 5.831289462109648e-05, "loss": 0.0387, "step": 47450 }, { "grad_norm": 0.2838640511035919, "learning_rate": 5.829658959355202e-05, "loss": 0.0421, "step": 47460 }, { "grad_norm": 0.45628780126571655, "learning_rate": 5.8280283658704995e-05, "loss": 0.0491, "step": 47470 }, { "grad_norm": 0.46830883622169495, "learning_rate": 5.826397681833859e-05, "loss": 0.0383, "step": 47480 }, { "grad_norm": 0.4941192865371704, "learning_rate": 5.824766907423612e-05, "loss": 0.038, "step": 47490 }, { "grad_norm": 0.3417649269104004, "learning_rate": 5.8231360428180945e-05, "loss": 0.0408, "step": 47500 }, { "grad_norm": 0.27359116077423096, "learning_rate": 5.821505088195658e-05, "loss": 0.0345, "step": 47510 }, { "grad_norm": 0.34333083033561707, "learning_rate": 5.819874043734661e-05, "loss": 0.0413, "step": 47520 }, { "grad_norm": 0.38600751757621765, "learning_rate": 5.8182429096134695e-05, "loss": 0.0374, "step": 47530 }, { "grad_norm": 0.7362295389175415, "learning_rate": 5.816611686010465e-05, "loss": 0.0369, "step": 47540 }, { "grad_norm": 0.4044455587863922, "learning_rate": 5.814980373104033e-05, "loss": 0.0395, "step": 47550 }, { "grad_norm": 0.37704357504844666, "learning_rate": 5.813348971072572e-05, "loss": 0.0503, "step": 47560 }, { "grad_norm": 0.34802481532096863, "learning_rate": 5.811717480094492e-05, "loss": 0.0345, "step": 47570 }, { "grad_norm": 0.4431002140045166, "learning_rate": 5.810085900348209e-05, "loss": 0.0436, "step": 47580 }, { "grad_norm": 0.6760485172271729, "learning_rate": 5.8084542320121483e-05, "loss": 0.0358, "step": 47590 }, { "grad_norm": 0.3343296945095062, "learning_rate": 5.8068224752647497e-05, "loss": 0.0491, "step": 47600 }, { "grad_norm": 0.4188728928565979, "learning_rate": 5.805190630284457e-05, "loss": 0.0245, "step": 47610 }, { "grad_norm": 0.43057456612586975, "learning_rate": 5.803558697249729e-05, "loss": 0.0374, "step": 47620 }, { "grad_norm": 0.5418239831924438, "learning_rate": 5.8019266763390276e-05, "loss": 0.0404, "step": 47630 }, { "grad_norm": 0.2714531719684601, "learning_rate": 5.800294567730833e-05, "loss": 0.0302, "step": 47640 }, { "grad_norm": 0.2475275695323944, "learning_rate": 5.798662371603628e-05, "loss": 0.0403, "step": 47650 }, { "grad_norm": 0.4393156170845032, "learning_rate": 5.797030088135904e-05, "loss": 0.0432, "step": 47660 }, { "grad_norm": 0.3263935446739197, "learning_rate": 5.7953977175061705e-05, "loss": 0.0333, "step": 47670 }, { "grad_norm": 1.1288291215896606, "learning_rate": 5.793765259892938e-05, "loss": 0.0468, "step": 47680 }, { "grad_norm": 0.6034245491027832, "learning_rate": 5.792132715474729e-05, "loss": 0.0367, "step": 47690 }, { "grad_norm": 0.5928811430931091, "learning_rate": 5.790500084430078e-05, "loss": 0.0349, "step": 47700 }, { "grad_norm": 0.3085702657699585, "learning_rate": 5.7888673669375274e-05, "loss": 0.0344, "step": 47710 }, { "grad_norm": 0.301980197429657, "learning_rate": 5.787234563175625e-05, "loss": 0.0372, "step": 47720 }, { "grad_norm": 0.8501990437507629, "learning_rate": 5.7856016733229355e-05, "loss": 0.0428, "step": 47730 }, { "grad_norm": 0.4640389084815979, "learning_rate": 5.7839686975580297e-05, "loss": 0.0449, "step": 47740 }, { "grad_norm": 0.3295532464981079, "learning_rate": 5.782335636059484e-05, "loss": 0.0364, "step": 47750 }, { "grad_norm": 0.40213409066200256, "learning_rate": 5.780702489005889e-05, "loss": 0.0392, "step": 47760 }, { "grad_norm": 0.2900145649909973, "learning_rate": 5.779069256575845e-05, "loss": 0.0413, "step": 47770 }, { "grad_norm": 0.6113975048065186, "learning_rate": 5.7774359389479574e-05, "loss": 0.0316, "step": 47780 }, { "grad_norm": 0.47433194518089294, "learning_rate": 5.775802536300845e-05, "loss": 0.0376, "step": 47790 }, { "grad_norm": 0.3630164563655853, "learning_rate": 5.774169048813134e-05, "loss": 0.0315, "step": 47800 }, { "grad_norm": 0.31363293528556824, "learning_rate": 5.77253547666346e-05, "loss": 0.0303, "step": 47810 }, { "grad_norm": 0.5282543897628784, "learning_rate": 5.770901820030465e-05, "loss": 0.0403, "step": 47820 }, { "grad_norm": 0.44231879711151123, "learning_rate": 5.769268079092809e-05, "loss": 0.0376, "step": 47830 }, { "grad_norm": 0.28310880064964294, "learning_rate": 5.767634254029151e-05, "loss": 0.0335, "step": 47840 }, { "grad_norm": 0.5003359913825989, "learning_rate": 5.7660003450181655e-05, "loss": 0.0635, "step": 47850 }, { "grad_norm": 0.428730845451355, "learning_rate": 5.764366352238534e-05, "loss": 0.0405, "step": 47860 }, { "grad_norm": 0.48116177320480347, "learning_rate": 5.7627322758689474e-05, "loss": 0.0304, "step": 47870 }, { "grad_norm": 0.3877306282520294, "learning_rate": 5.761098116088105e-05, "loss": 0.0391, "step": 47880 }, { "grad_norm": 0.3886343240737915, "learning_rate": 5.759463873074717e-05, "loss": 0.0354, "step": 47890 }, { "grad_norm": 0.46838870644569397, "learning_rate": 5.757829547007504e-05, "loss": 0.0329, "step": 47900 }, { "grad_norm": 0.489653617143631, "learning_rate": 5.756195138065189e-05, "loss": 0.0354, "step": 47910 }, { "grad_norm": 0.4017762243747711, "learning_rate": 5.754560646426511e-05, "loss": 0.0385, "step": 47920 }, { "grad_norm": 0.2677384614944458, "learning_rate": 5.752926072270216e-05, "loss": 0.0314, "step": 47930 }, { "grad_norm": 0.44925686717033386, "learning_rate": 5.7512914157750563e-05, "loss": 0.0383, "step": 47940 }, { "grad_norm": 0.3275096118450165, "learning_rate": 5.749656677119798e-05, "loss": 0.0233, "step": 47950 }, { "grad_norm": 0.3935694098472595, "learning_rate": 5.7480218564832125e-05, "loss": 0.0416, "step": 47960 }, { "grad_norm": 0.29945722222328186, "learning_rate": 5.746386954044082e-05, "loss": 0.0271, "step": 47970 }, { "grad_norm": 0.47354617714881897, "learning_rate": 5.744751969981195e-05, "loss": 0.0313, "step": 47980 }, { "grad_norm": 0.4120311439037323, "learning_rate": 5.7431169044733526e-05, "loss": 0.0435, "step": 47990 }, { "grad_norm": 0.607267439365387, "learning_rate": 5.741481757699364e-05, "loss": 0.0358, "step": 48000 }, { "grad_norm": 0.4272235333919525, "learning_rate": 5.7398465298380434e-05, "loss": 0.0428, "step": 48010 }, { "grad_norm": 0.469107985496521, "learning_rate": 5.7382112210682193e-05, "loss": 0.0557, "step": 48020 }, { "grad_norm": 0.5069810748100281, "learning_rate": 5.7365758315687266e-05, "loss": 0.0432, "step": 48030 }, { "grad_norm": 0.4919806122779846, "learning_rate": 5.734940361518407e-05, "loss": 0.0463, "step": 48040 }, { "grad_norm": 1.1589919328689575, "learning_rate": 5.733304811096116e-05, "loss": 0.0426, "step": 48050 }, { "grad_norm": 0.2893611192703247, "learning_rate": 5.731669180480713e-05, "loss": 0.0464, "step": 48060 }, { "grad_norm": 0.6148503422737122, "learning_rate": 5.730033469851067e-05, "loss": 0.0476, "step": 48070 }, { "grad_norm": 0.5406756401062012, "learning_rate": 5.7283976793860607e-05, "loss": 0.0432, "step": 48080 }, { "grad_norm": 0.3875916302204132, "learning_rate": 5.726761809264577e-05, "loss": 0.0327, "step": 48090 }, { "grad_norm": 0.37072500586509705, "learning_rate": 5.7251258596655155e-05, "loss": 0.0315, "step": 48100 }, { "grad_norm": 0.6180645823478699, "learning_rate": 5.72348983076778e-05, "loss": 0.0397, "step": 48110 }, { "grad_norm": 0.3950877785682678, "learning_rate": 5.7218537227502854e-05, "loss": 0.0355, "step": 48120 }, { "grad_norm": 0.3270295560359955, "learning_rate": 5.720217535791951e-05, "loss": 0.0261, "step": 48130 }, { "grad_norm": 0.3717411458492279, "learning_rate": 5.718581270071711e-05, "loss": 0.0358, "step": 48140 }, { "grad_norm": 0.4711534082889557, "learning_rate": 5.716944925768505e-05, "loss": 0.0317, "step": 48150 }, { "grad_norm": 0.2948645055294037, "learning_rate": 5.7153085030612786e-05, "loss": 0.0343, "step": 48160 }, { "grad_norm": 0.4654587507247925, "learning_rate": 5.713672002128989e-05, "loss": 0.0466, "step": 48170 }, { "grad_norm": 0.3154057264328003, "learning_rate": 5.712035423150602e-05, "loss": 0.0396, "step": 48180 }, { "grad_norm": 0.35443365573883057, "learning_rate": 5.710398766305094e-05, "loss": 0.0413, "step": 48190 }, { "grad_norm": 0.49843457341194153, "learning_rate": 5.708762031771442e-05, "loss": 0.0364, "step": 48200 }, { "grad_norm": 0.32737240195274353, "learning_rate": 5.707125219728642e-05, "loss": 0.0291, "step": 48210 }, { "grad_norm": 0.5752713084220886, "learning_rate": 5.7054883303556905e-05, "loss": 0.0309, "step": 48220 }, { "grad_norm": 0.41850101947784424, "learning_rate": 5.703851363831595e-05, "loss": 0.0419, "step": 48230 }, { "grad_norm": 0.37954580783843994, "learning_rate": 5.702214320335374e-05, "loss": 0.0407, "step": 48240 }, { "grad_norm": 0.3143686056137085, "learning_rate": 5.7005772000460514e-05, "loss": 0.0401, "step": 48250 }, { "grad_norm": 0.4189356863498688, "learning_rate": 5.698940003142656e-05, "loss": 0.0351, "step": 48260 }, { "grad_norm": 0.4064207673072815, "learning_rate": 5.697302729804236e-05, "loss": 0.0359, "step": 48270 }, { "grad_norm": 0.48083388805389404, "learning_rate": 5.695665380209837e-05, "loss": 0.0385, "step": 48280 }, { "grad_norm": 0.5006412863731384, "learning_rate": 5.6940279545385165e-05, "loss": 0.0279, "step": 48290 }, { "grad_norm": 0.7753506302833557, "learning_rate": 5.692390452969344e-05, "loss": 0.0378, "step": 48300 }, { "grad_norm": 0.26909515261650085, "learning_rate": 5.690752875681392e-05, "loss": 0.0394, "step": 48310 }, { "grad_norm": 0.24197909235954285, "learning_rate": 5.6891152228537435e-05, "loss": 0.0382, "step": 48320 }, { "grad_norm": 0.59421706199646, "learning_rate": 5.687477494665492e-05, "loss": 0.0336, "step": 48330 }, { "grad_norm": 0.3517707288265228, "learning_rate": 5.685839691295733e-05, "loss": 0.039, "step": 48340 }, { "grad_norm": 0.3809163272380829, "learning_rate": 5.6842018129235786e-05, "loss": 0.0393, "step": 48350 }, { "grad_norm": 1.0776762962341309, "learning_rate": 5.6825638597281404e-05, "loss": 0.0363, "step": 48360 }, { "grad_norm": 0.3513786196708679, "learning_rate": 5.680925831888546e-05, "loss": 0.0504, "step": 48370 }, { "grad_norm": 0.3493987023830414, "learning_rate": 5.6792877295839274e-05, "loss": 0.0373, "step": 48380 }, { "grad_norm": 0.3984660804271698, "learning_rate": 5.6776495529934224e-05, "loss": 0.0372, "step": 48390 }, { "grad_norm": 0.27683156728744507, "learning_rate": 5.6760113022961824e-05, "loss": 0.0282, "step": 48400 }, { "grad_norm": 0.47643354535102844, "learning_rate": 5.6743729776713617e-05, "loss": 0.0347, "step": 48410 }, { "grad_norm": 0.38628894090652466, "learning_rate": 5.672734579298126e-05, "loss": 0.0274, "step": 48420 }, { "grad_norm": 0.5354347825050354, "learning_rate": 5.671096107355649e-05, "loss": 0.0365, "step": 48430 }, { "grad_norm": 0.28701457381248474, "learning_rate": 5.66945756202311e-05, "loss": 0.0332, "step": 48440 }, { "grad_norm": 0.46743401885032654, "learning_rate": 5.667818943479699e-05, "loss": 0.041, "step": 48450 }, { "grad_norm": 0.48425063490867615, "learning_rate": 5.666180251904612e-05, "loss": 0.04, "step": 48460 }, { "grad_norm": 0.4660603106021881, "learning_rate": 5.6645414874770555e-05, "loss": 0.0317, "step": 48470 }, { "grad_norm": 0.30430132150650024, "learning_rate": 5.66290265037624e-05, "loss": 0.0423, "step": 48480 }, { "grad_norm": 0.4710733890533447, "learning_rate": 5.661263740781386e-05, "loss": 0.0311, "step": 48490 }, { "grad_norm": 0.4897562861442566, "learning_rate": 5.6596247588717254e-05, "loss": 0.0426, "step": 48500 }, { "grad_norm": 0.6325657963752747, "learning_rate": 5.6579857048264926e-05, "loss": 0.0384, "step": 48510 }, { "grad_norm": 0.36141374707221985, "learning_rate": 5.6563465788249314e-05, "loss": 0.0312, "step": 48520 }, { "grad_norm": 0.2625887989997864, "learning_rate": 5.6547073810462956e-05, "loss": 0.0426, "step": 48530 }, { "grad_norm": 0.47106438875198364, "learning_rate": 5.653068111669846e-05, "loss": 0.0339, "step": 48540 }, { "grad_norm": 0.2792949378490448, "learning_rate": 5.651428770874848e-05, "loss": 0.0408, "step": 48550 }, { "grad_norm": 0.29859459400177, "learning_rate": 5.64978935884058e-05, "loss": 0.0326, "step": 48560 }, { "grad_norm": 0.30378037691116333, "learning_rate": 5.6481498757463244e-05, "loss": 0.0348, "step": 48570 }, { "grad_norm": 0.3355747163295746, "learning_rate": 5.646510321771373e-05, "loss": 0.039, "step": 48580 }, { "grad_norm": 0.4686979055404663, "learning_rate": 5.644870697095024e-05, "loss": 0.0364, "step": 48590 }, { "grad_norm": 0.33120885491371155, "learning_rate": 5.643231001896586e-05, "loss": 0.03, "step": 48600 }, { "grad_norm": 0.3471139669418335, "learning_rate": 5.6415912363553726e-05, "loss": 0.0522, "step": 48610 }, { "grad_norm": 0.3494327962398529, "learning_rate": 5.639951400650706e-05, "loss": 0.0357, "step": 48620 }, { "grad_norm": 0.33613425493240356, "learning_rate": 5.6383114949619165e-05, "loss": 0.0299, "step": 48630 }, { "grad_norm": 0.42999523878097534, "learning_rate": 5.636671519468342e-05, "loss": 0.0418, "step": 48640 }, { "grad_norm": 0.33877965807914734, "learning_rate": 5.635031474349327e-05, "loss": 0.0353, "step": 48650 }, { "grad_norm": 0.5767077207565308, "learning_rate": 5.6333913597842246e-05, "loss": 0.0326, "step": 48660 }, { "grad_norm": 0.747386634349823, "learning_rate": 5.6317511759523955e-05, "loss": 0.0454, "step": 48670 }, { "grad_norm": 0.527982234954834, "learning_rate": 5.630110923033207e-05, "loss": 0.0422, "step": 48680 }, { "grad_norm": 0.3961883783340454, "learning_rate": 5.628470601206036e-05, "loss": 0.0387, "step": 48690 }, { "grad_norm": 0.5002530217170715, "learning_rate": 5.626830210650266e-05, "loss": 0.0378, "step": 48700 }, { "grad_norm": 0.3406754434108734, "learning_rate": 5.625189751545285e-05, "loss": 0.0487, "step": 48710 }, { "grad_norm": 0.3966270089149475, "learning_rate": 5.6235492240704936e-05, "loss": 0.0335, "step": 48720 }, { "grad_norm": 0.4091041088104248, "learning_rate": 5.621908628405296e-05, "loss": 0.0414, "step": 48730 }, { "grad_norm": 0.8199350237846375, "learning_rate": 5.620267964729106e-05, "loss": 0.0335, "step": 48740 }, { "grad_norm": 0.3394327163696289, "learning_rate": 5.618627233221344e-05, "loss": 0.0314, "step": 48750 }, { "grad_norm": 0.5731573700904846, "learning_rate": 5.616986434061438e-05, "loss": 0.0461, "step": 48760 }, { "grad_norm": 0.33457764983177185, "learning_rate": 5.615345567428822e-05, "loss": 0.0337, "step": 48770 }, { "grad_norm": 0.5054206848144531, "learning_rate": 5.613704633502941e-05, "loss": 0.0279, "step": 48780 }, { "grad_norm": 0.4177418053150177, "learning_rate": 5.612063632463245e-05, "loss": 0.0357, "step": 48790 }, { "grad_norm": 0.39217323064804077, "learning_rate": 5.610422564489188e-05, "loss": 0.038, "step": 48800 }, { "grad_norm": 0.5016282200813293, "learning_rate": 5.608781429760239e-05, "loss": 0.0362, "step": 48810 }, { "grad_norm": 0.29497888684272766, "learning_rate": 5.607140228455866e-05, "loss": 0.0324, "step": 48820 }, { "grad_norm": 0.44477659463882446, "learning_rate": 5.605498960755553e-05, "loss": 0.0396, "step": 48830 }, { "grad_norm": 0.42020732164382935, "learning_rate": 5.603857626838782e-05, "loss": 0.0373, "step": 48840 }, { "grad_norm": 0.4908614158630371, "learning_rate": 5.60221622688505e-05, "loss": 0.0316, "step": 48850 }, { "grad_norm": 0.5480750203132629, "learning_rate": 5.6005747610738565e-05, "loss": 0.0304, "step": 48860 }, { "grad_norm": 0.32910048961639404, "learning_rate": 5.59893322958471e-05, "loss": 0.0348, "step": 48870 }, { "grad_norm": 0.46643418073654175, "learning_rate": 5.5972916325971256e-05, "loss": 0.0411, "step": 48880 }, { "grad_norm": 0.400104284286499, "learning_rate": 5.595649970290628e-05, "loss": 0.0401, "step": 48890 }, { "grad_norm": 0.2827267646789551, "learning_rate": 5.5940082428447426e-05, "loss": 0.0467, "step": 48900 }, { "grad_norm": 0.36357903480529785, "learning_rate": 5.592366450439012e-05, "loss": 0.0435, "step": 48910 }, { "grad_norm": 0.7086977362632751, "learning_rate": 5.590724593252975e-05, "loss": 0.042, "step": 48920 }, { "grad_norm": 0.2558269202709198, "learning_rate": 5.589082671466184e-05, "loss": 0.0347, "step": 48930 }, { "grad_norm": 0.43145111203193665, "learning_rate": 5.587440685258199e-05, "loss": 0.037, "step": 48940 }, { "grad_norm": 0.4303419291973114, "learning_rate": 5.585798634808583e-05, "loss": 0.0332, "step": 48950 }, { "grad_norm": 0.46863454580307007, "learning_rate": 5.584156520296909e-05, "loss": 0.034, "step": 48960 }, { "grad_norm": 0.5078457593917847, "learning_rate": 5.582514341902757e-05, "loss": 0.0406, "step": 48970 }, { "grad_norm": 0.3755258619785309, "learning_rate": 5.580872099805713e-05, "loss": 0.0386, "step": 48980 }, { "grad_norm": 0.2761898636817932, "learning_rate": 5.5792297941853674e-05, "loss": 0.0366, "step": 48990 }, { "grad_norm": 0.40082138776779175, "learning_rate": 5.5775874252213247e-05, "loss": 0.0412, "step": 49000 }, { "grad_norm": 0.6381046175956726, "learning_rate": 5.575944993093189e-05, "loss": 0.0395, "step": 49010 }, { "grad_norm": 0.5409019589424133, "learning_rate": 5.574302497980574e-05, "loss": 0.0332, "step": 49020 }, { "grad_norm": 0.38748428225517273, "learning_rate": 5.5726599400631005e-05, "loss": 0.0404, "step": 49030 }, { "grad_norm": 0.4417533576488495, "learning_rate": 5.571017319520399e-05, "loss": 0.0432, "step": 49040 }, { "grad_norm": 0.29230692982673645, "learning_rate": 5.5693746365320984e-05, "loss": 0.0324, "step": 49050 }, { "grad_norm": 0.5826665163040161, "learning_rate": 5.567731891277848e-05, "loss": 0.042, "step": 49060 }, { "grad_norm": 0.37805142998695374, "learning_rate": 5.566089083937288e-05, "loss": 0.0388, "step": 49070 }, { "grad_norm": 0.39587268233299255, "learning_rate": 5.564446214690079e-05, "loss": 0.0399, "step": 49080 }, { "grad_norm": 0.4226634204387665, "learning_rate": 5.56280328371588e-05, "loss": 0.0334, "step": 49090 }, { "grad_norm": 0.28142401576042175, "learning_rate": 5.5611602911943596e-05, "loss": 0.03, "step": 49100 }, { "grad_norm": 0.46160176396369934, "learning_rate": 5.559517237305194e-05, "loss": 0.041, "step": 49110 }, { "grad_norm": 0.41414594650268555, "learning_rate": 5.557874122228064e-05, "loss": 0.0414, "step": 49120 }, { "grad_norm": 0.4887986481189728, "learning_rate": 5.55623094614266e-05, "loss": 0.0312, "step": 49130 }, { "grad_norm": 0.5453243851661682, "learning_rate": 5.5545877092286743e-05, "loss": 0.0402, "step": 49140 }, { "grad_norm": 0.3848991394042969, "learning_rate": 5.552944411665811e-05, "loss": 0.0339, "step": 49150 }, { "grad_norm": 0.3220333158969879, "learning_rate": 5.5513010536337795e-05, "loss": 0.0417, "step": 49160 }, { "grad_norm": 0.49523717164993286, "learning_rate": 5.5496576353122934e-05, "loss": 0.0354, "step": 49170 }, { "grad_norm": 0.44336017966270447, "learning_rate": 5.548014156881074e-05, "loss": 0.0343, "step": 49180 }, { "grad_norm": 0.23961962759494781, "learning_rate": 5.5463706185198494e-05, "loss": 0.0406, "step": 49190 }, { "grad_norm": 0.26669248938560486, "learning_rate": 5.544727020408358e-05, "loss": 0.0355, "step": 49200 }, { "grad_norm": 0.6140438914299011, "learning_rate": 5.5430833627263366e-05, "loss": 0.0338, "step": 49210 }, { "grad_norm": 0.2899025082588196, "learning_rate": 5.5414396456535354e-05, "loss": 0.0325, "step": 49220 }, { "grad_norm": 0.3936218321323395, "learning_rate": 5.539795869369709e-05, "loss": 0.0329, "step": 49230 }, { "grad_norm": 0.338137149810791, "learning_rate": 5.538152034054618e-05, "loss": 0.0322, "step": 49240 }, { "grad_norm": 0.36421358585357666, "learning_rate": 5.536508139888028e-05, "loss": 0.035, "step": 49250 }, { "grad_norm": 0.48609355092048645, "learning_rate": 5.534864187049716e-05, "loss": 0.0283, "step": 49260 }, { "grad_norm": 0.18437637388706207, "learning_rate": 5.5332201757194615e-05, "loss": 0.0363, "step": 49270 }, { "grad_norm": 0.3804349899291992, "learning_rate": 5.531576106077049e-05, "loss": 0.0341, "step": 49280 }, { "grad_norm": 0.47853198647499084, "learning_rate": 5.529931978302272e-05, "loss": 0.0301, "step": 49290 }, { "grad_norm": 0.19674696028232574, "learning_rate": 5.528287792574932e-05, "loss": 0.0319, "step": 49300 }, { "grad_norm": 0.3288346529006958, "learning_rate": 5.5266435490748324e-05, "loss": 0.0297, "step": 49310 }, { "grad_norm": 0.5620580315589905, "learning_rate": 5.524999247981787e-05, "loss": 0.0359, "step": 49320 }, { "grad_norm": 0.22111806273460388, "learning_rate": 5.523354889475613e-05, "loss": 0.0382, "step": 49330 }, { "grad_norm": 0.4308231472969055, "learning_rate": 5.521710473736134e-05, "loss": 0.0378, "step": 49340 }, { "grad_norm": 0.4603953957557678, "learning_rate": 5.5200660009431826e-05, "loss": 0.0328, "step": 49350 }, { "grad_norm": 0.45609039068222046, "learning_rate": 5.518421471276596e-05, "loss": 0.0331, "step": 49360 }, { "grad_norm": 0.43947339057922363, "learning_rate": 5.5167768849162174e-05, "loss": 0.0377, "step": 49370 }, { "grad_norm": 0.36359813809394836, "learning_rate": 5.515132242041893e-05, "loss": 0.0352, "step": 49380 }, { "grad_norm": 0.27413272857666016, "learning_rate": 5.513487542833483e-05, "loss": 0.0315, "step": 49390 }, { "grad_norm": 0.5290615558624268, "learning_rate": 5.5118427874708466e-05, "loss": 0.0317, "step": 49400 }, { "grad_norm": 0.3041495084762573, "learning_rate": 5.510197976133853e-05, "loss": 0.0249, "step": 49410 }, { "grad_norm": 0.5173347592353821, "learning_rate": 5.508553109002376e-05, "loss": 0.0296, "step": 49420 }, { "grad_norm": 0.34888824820518494, "learning_rate": 5.5069081862562957e-05, "loss": 0.0254, "step": 49430 }, { "grad_norm": 0.268574059009552, "learning_rate": 5.5052632080754965e-05, "loss": 0.0284, "step": 49440 }, { "grad_norm": 0.3487589359283447, "learning_rate": 5.503618174639874e-05, "loss": 0.0342, "step": 49450 }, { "grad_norm": 0.5886706709861755, "learning_rate": 5.501973086129325e-05, "loss": 0.0381, "step": 49460 }, { "grad_norm": 0.2108292281627655, "learning_rate": 5.500327942723753e-05, "loss": 0.0288, "step": 49470 }, { "grad_norm": 0.3584587872028351, "learning_rate": 5.498682744603071e-05, "loss": 0.0314, "step": 49480 }, { "grad_norm": 0.31372323632240295, "learning_rate": 5.497037491947194e-05, "loss": 0.0316, "step": 49490 }, { "grad_norm": 0.28718388080596924, "learning_rate": 5.4953921849360424e-05, "loss": 0.0388, "step": 49500 }, { "grad_norm": 0.2852691411972046, "learning_rate": 5.493746823749547e-05, "loss": 0.0386, "step": 49510 }, { "grad_norm": 0.4255063235759735, "learning_rate": 5.49210140856764e-05, "loss": 0.0361, "step": 49520 }, { "grad_norm": 0.3411220610141754, "learning_rate": 5.4904559395702647e-05, "loss": 0.0408, "step": 49530 }, { "grad_norm": 0.2574109137058258, "learning_rate": 5.488810416937364e-05, "loss": 0.0284, "step": 49540 }, { "grad_norm": 0.34460315108299255, "learning_rate": 5.4871648408488926e-05, "loss": 0.0331, "step": 49550 }, { "grad_norm": 0.3221570551395416, "learning_rate": 5.485519211484807e-05, "loss": 0.0278, "step": 49560 }, { "grad_norm": 0.4579174816608429, "learning_rate": 5.4838735290250676e-05, "loss": 0.0356, "step": 49570 }, { "grad_norm": 0.39554429054260254, "learning_rate": 5.482227793649649e-05, "loss": 0.0371, "step": 49580 }, { "grad_norm": 0.2943195104598999, "learning_rate": 5.480582005538524e-05, "loss": 0.0342, "step": 49590 }, { "grad_norm": 0.4046507179737091, "learning_rate": 5.478936164871671e-05, "loss": 0.0392, "step": 49600 }, { "grad_norm": 0.4831428527832031, "learning_rate": 5.477290271829081e-05, "loss": 0.0386, "step": 49610 }, { "grad_norm": 0.318082332611084, "learning_rate": 5.475644326590744e-05, "loss": 0.0307, "step": 49620 }, { "grad_norm": 0.427328497171402, "learning_rate": 5.473998329336658e-05, "loss": 0.0359, "step": 49630 }, { "grad_norm": 0.3525753915309906, "learning_rate": 5.4723522802468286e-05, "loss": 0.0287, "step": 49640 }, { "grad_norm": 0.4709394574165344, "learning_rate": 5.470706179501264e-05, "loss": 0.0316, "step": 49650 }, { "grad_norm": 0.41723617911338806, "learning_rate": 5.4690600272799776e-05, "loss": 0.0388, "step": 49660 }, { "grad_norm": 0.526390552520752, "learning_rate": 5.467413823762993e-05, "loss": 0.0408, "step": 49670 }, { "grad_norm": 0.5560839176177979, "learning_rate": 5.465767569130335e-05, "loss": 0.0364, "step": 49680 }, { "grad_norm": 0.4176388382911682, "learning_rate": 5.464121263562036e-05, "loss": 0.0373, "step": 49690 }, { "grad_norm": 0.2717348635196686, "learning_rate": 5.4624749072381343e-05, "loss": 0.0356, "step": 49700 }, { "grad_norm": 0.27485203742980957, "learning_rate": 5.460828500338672e-05, "loss": 0.0342, "step": 49710 }, { "grad_norm": 0.3572239279747009, "learning_rate": 5.459182043043698e-05, "loss": 0.0422, "step": 49720 }, { "grad_norm": 0.24296221137046814, "learning_rate": 5.457535535533265e-05, "loss": 0.0329, "step": 49730 }, { "grad_norm": 0.4322992265224457, "learning_rate": 5.4558889779874334e-05, "loss": 0.0398, "step": 49740 }, { "grad_norm": 0.3774515986442566, "learning_rate": 5.454242370586269e-05, "loss": 0.0338, "step": 49750 }, { "grad_norm": 0.4479202330112457, "learning_rate": 5.452595713509843e-05, "loss": 0.0314, "step": 49760 }, { "grad_norm": 0.3627769947052002, "learning_rate": 5.4509490069382274e-05, "loss": 0.0358, "step": 49770 }, { "grad_norm": 0.2810981571674347, "learning_rate": 5.449302251051509e-05, "loss": 0.0249, "step": 49780 }, { "grad_norm": 0.4081282913684845, "learning_rate": 5.44765544602977e-05, "loss": 0.0388, "step": 49790 }, { "grad_norm": 0.27730101346969604, "learning_rate": 5.446008592053107e-05, "loss": 0.0303, "step": 49800 }, { "grad_norm": 0.36089277267456055, "learning_rate": 5.4443616893016135e-05, "loss": 0.0345, "step": 49810 }, { "grad_norm": 0.40276092290878296, "learning_rate": 5.4427147379553934e-05, "loss": 0.0351, "step": 49820 }, { "grad_norm": 0.47786688804626465, "learning_rate": 5.441067738194556e-05, "loss": 0.0331, "step": 49830 }, { "grad_norm": 0.35688915848731995, "learning_rate": 5.439420690199214e-05, "loss": 0.0405, "step": 49840 }, { "grad_norm": 0.35367074608802795, "learning_rate": 5.437773594149484e-05, "loss": 0.0341, "step": 49850 }, { "grad_norm": 0.370097279548645, "learning_rate": 5.436126450225495e-05, "loss": 0.0364, "step": 49860 }, { "grad_norm": 0.34849005937576294, "learning_rate": 5.434479258607371e-05, "loss": 0.0326, "step": 49870 }, { "grad_norm": 0.3332255780696869, "learning_rate": 5.43283201947525e-05, "loss": 0.032, "step": 49880 }, { "grad_norm": 0.43101513385772705, "learning_rate": 5.431184733009268e-05, "loss": 0.047, "step": 49890 }, { "grad_norm": 0.3446987569332123, "learning_rate": 5.4295373993895736e-05, "loss": 0.034, "step": 49900 }, { "grad_norm": 0.36515116691589355, "learning_rate": 5.4278900187963157e-05, "loss": 0.0389, "step": 49910 }, { "grad_norm": 0.3071427047252655, "learning_rate": 5.426242591409646e-05, "loss": 0.036, "step": 49920 }, { "grad_norm": 0.34220460057258606, "learning_rate": 5.42459511740973e-05, "loss": 0.0322, "step": 49930 }, { "grad_norm": 0.39182597398757935, "learning_rate": 5.422947596976729e-05, "loss": 0.0336, "step": 49940 }, { "grad_norm": 0.41972625255584717, "learning_rate": 5.4213000302908134e-05, "loss": 0.0375, "step": 49950 }, { "grad_norm": 0.3674035966396332, "learning_rate": 5.419652417532162e-05, "loss": 0.0309, "step": 49960 }, { "grad_norm": 0.4991086423397064, "learning_rate": 5.4180047588809534e-05, "loss": 0.0315, "step": 49970 }, { "grad_norm": 0.373740017414093, "learning_rate": 5.4163570545173704e-05, "loss": 0.0308, "step": 49980 }, { "grad_norm": 0.6772974729537964, "learning_rate": 5.414709304621608e-05, "loss": 0.0317, "step": 49990 }, { "grad_norm": 0.308400422334671, "learning_rate": 5.413061509373858e-05, "loss": 0.0456, "step": 50000 }, { "grad_norm": 0.4069851338863373, "learning_rate": 5.411413668954324e-05, "loss": 0.0345, "step": 50010 }, { "grad_norm": 0.3829937279224396, "learning_rate": 5.4097657835432083e-05, "loss": 0.0359, "step": 50020 }, { "grad_norm": 0.3013925850391388, "learning_rate": 5.408117853320723e-05, "loss": 0.0341, "step": 50030 }, { "grad_norm": 0.4019637703895569, "learning_rate": 5.406469878467084e-05, "loss": 0.0296, "step": 50040 }, { "grad_norm": 0.45233941078186035, "learning_rate": 5.404821859162509e-05, "loss": 0.035, "step": 50050 }, { "grad_norm": 0.3247220814228058, "learning_rate": 5.403173795587225e-05, "loss": 0.0317, "step": 50060 }, { "grad_norm": 0.3795664310455322, "learning_rate": 5.4015256879214606e-05, "loss": 0.0397, "step": 50070 }, { "grad_norm": 0.2546306550502777, "learning_rate": 5.39987753634545e-05, "loss": 0.0336, "step": 50080 }, { "grad_norm": 0.2637081444263458, "learning_rate": 5.3982293410394335e-05, "loss": 0.0385, "step": 50090 }, { "grad_norm": 0.38764259219169617, "learning_rate": 5.396581102183655e-05, "loss": 0.0328, "step": 50100 }, { "grad_norm": 0.44144490361213684, "learning_rate": 5.3949328199583634e-05, "loss": 0.0318, "step": 50110 }, { "grad_norm": 0.43359068036079407, "learning_rate": 5.393284494543813e-05, "loss": 0.0438, "step": 50120 }, { "grad_norm": 0.26431533694267273, "learning_rate": 5.391636126120262e-05, "loss": 0.0322, "step": 50130 }, { "grad_norm": 0.3575284481048584, "learning_rate": 5.389987714867971e-05, "loss": 0.0314, "step": 50140 }, { "grad_norm": 0.5363418459892273, "learning_rate": 5.3883392609672114e-05, "loss": 0.0438, "step": 50150 }, { "grad_norm": 0.2993781566619873, "learning_rate": 5.3866907645982545e-05, "loss": 0.0363, "step": 50160 }, { "grad_norm": 0.7228763103485107, "learning_rate": 5.385042225941375e-05, "loss": 0.0389, "step": 50170 }, { "grad_norm": 0.3527318239212036, "learning_rate": 5.3833936451768585e-05, "loss": 0.0359, "step": 50180 }, { "grad_norm": 0.4250335097312927, "learning_rate": 5.3817450224849884e-05, "loss": 0.0389, "step": 50190 }, { "grad_norm": 0.43833300471305847, "learning_rate": 5.380096358046056e-05, "loss": 0.0497, "step": 50200 }, { "grad_norm": 0.2845234274864197, "learning_rate": 5.378447652040359e-05, "loss": 0.0331, "step": 50210 }, { "grad_norm": 0.36887696385383606, "learning_rate": 5.376798904648195e-05, "loss": 0.0275, "step": 50220 }, { "grad_norm": 0.3485334813594818, "learning_rate": 5.375150116049869e-05, "loss": 0.032, "step": 50230 }, { "grad_norm": 0.33478429913520813, "learning_rate": 5.373501286425691e-05, "loss": 0.0256, "step": 50240 }, { "grad_norm": 0.3069491684436798, "learning_rate": 5.3718524159559726e-05, "loss": 0.0363, "step": 50250 }, { "grad_norm": 0.3878721594810486, "learning_rate": 5.370203504821034e-05, "loss": 0.0402, "step": 50260 }, { "grad_norm": 0.2837892472743988, "learning_rate": 5.368554553201196e-05, "loss": 0.0393, "step": 50270 }, { "grad_norm": 0.3577217757701874, "learning_rate": 5.3669055612767874e-05, "loss": 0.0447, "step": 50280 }, { "grad_norm": 0.38512861728668213, "learning_rate": 5.3652565292281376e-05, "loss": 0.0367, "step": 50290 }, { "grad_norm": 0.4164137840270996, "learning_rate": 5.3636074572355824e-05, "loss": 0.0332, "step": 50300 }, { "grad_norm": 0.40996024012565613, "learning_rate": 5.361958345479463e-05, "loss": 0.0353, "step": 50310 }, { "grad_norm": 0.3219619691371918, "learning_rate": 5.360309194140123e-05, "loss": 0.0401, "step": 50320 }, { "grad_norm": 0.3612028956413269, "learning_rate": 5.3586600033979106e-05, "loss": 0.031, "step": 50330 }, { "grad_norm": 0.6936739683151245, "learning_rate": 5.357010773433181e-05, "loss": 0.0376, "step": 50340 }, { "grad_norm": 0.39547279477119446, "learning_rate": 5.3553615044262895e-05, "loss": 0.0344, "step": 50350 }, { "grad_norm": 0.2531590461730957, "learning_rate": 5.3537121965575964e-05, "loss": 0.0375, "step": 50360 }, { "grad_norm": 0.38237056136131287, "learning_rate": 5.352062850007471e-05, "loss": 0.0428, "step": 50370 }, { "grad_norm": 0.20900079607963562, "learning_rate": 5.350413464956282e-05, "loss": 0.034, "step": 50380 }, { "grad_norm": 0.45406073331832886, "learning_rate": 5.348764041584403e-05, "loss": 0.0381, "step": 50390 }, { "grad_norm": 0.32412874698638916, "learning_rate": 5.347114580072212e-05, "loss": 0.0365, "step": 50400 }, { "grad_norm": 0.4247996509075165, "learning_rate": 5.345465080600093e-05, "loss": 0.041, "step": 50410 }, { "grad_norm": 0.3954009413719177, "learning_rate": 5.3438155433484326e-05, "loss": 0.0325, "step": 50420 }, { "grad_norm": 0.4235612154006958, "learning_rate": 5.3421659684976197e-05, "loss": 0.0339, "step": 50430 }, { "grad_norm": 0.5891638398170471, "learning_rate": 5.340516356228052e-05, "loss": 0.0324, "step": 50440 }, { "grad_norm": 0.5510687828063965, "learning_rate": 5.338866706720128e-05, "loss": 0.0383, "step": 50450 }, { "grad_norm": 0.2994423508644104, "learning_rate": 5.337217020154249e-05, "loss": 0.0329, "step": 50460 }, { "grad_norm": 0.6610921621322632, "learning_rate": 5.335567296710825e-05, "loss": 0.0351, "step": 50470 }, { "grad_norm": 0.4406626224517822, "learning_rate": 5.333917536570265e-05, "loss": 0.0347, "step": 50480 }, { "grad_norm": 0.346068412065506, "learning_rate": 5.332267739912986e-05, "loss": 0.0364, "step": 50490 }, { "grad_norm": 0.4121866524219513, "learning_rate": 5.330617906919405e-05, "loss": 0.0421, "step": 50500 }, { "grad_norm": 0.33504942059516907, "learning_rate": 5.328968037769949e-05, "loss": 0.0355, "step": 50510 }, { "grad_norm": 0.44271764159202576, "learning_rate": 5.327318132645042e-05, "loss": 0.0391, "step": 50520 }, { "grad_norm": 0.3815017640590668, "learning_rate": 5.325668191725116e-05, "loss": 0.0311, "step": 50530 }, { "grad_norm": 0.4103468656539917, "learning_rate": 5.324018215190606e-05, "loss": 0.0391, "step": 50540 }, { "grad_norm": 0.4247853755950928, "learning_rate": 5.3223682032219515e-05, "loss": 0.0369, "step": 50550 }, { "grad_norm": 0.2616156041622162, "learning_rate": 5.320718155999595e-05, "loss": 0.0362, "step": 50560 }, { "grad_norm": 0.35456013679504395, "learning_rate": 5.3190680737039835e-05, "loss": 0.0456, "step": 50570 }, { "grad_norm": 0.3497423529624939, "learning_rate": 5.317417956515567e-05, "loss": 0.0326, "step": 50580 }, { "grad_norm": 0.38930609822273254, "learning_rate": 5.315767804614798e-05, "loss": 0.0359, "step": 50590 }, { "grad_norm": 0.6931080222129822, "learning_rate": 5.3141176181821395e-05, "loss": 0.0372, "step": 50600 }, { "grad_norm": 0.271004319190979, "learning_rate": 5.31246739739805e-05, "loss": 0.0264, "step": 50610 }, { "grad_norm": 0.2888951897621155, "learning_rate": 5.310817142442994e-05, "loss": 0.0357, "step": 50620 }, { "grad_norm": 0.5115951895713806, "learning_rate": 5.309166853497445e-05, "loss": 0.0356, "step": 50630 }, { "grad_norm": 0.4406571686267853, "learning_rate": 5.307516530741873e-05, "loss": 0.0371, "step": 50640 }, { "grad_norm": 0.3060031235218048, "learning_rate": 5.305866174356754e-05, "loss": 0.0316, "step": 50650 }, { "grad_norm": 0.28257301449775696, "learning_rate": 5.304215784522571e-05, "loss": 0.0312, "step": 50660 }, { "grad_norm": 0.3158179223537445, "learning_rate": 5.302565361419808e-05, "loss": 0.0491, "step": 50670 }, { "grad_norm": 0.33418306708335876, "learning_rate": 5.3009149052289507e-05, "loss": 0.0337, "step": 50680 }, { "grad_norm": 0.37200692296028137, "learning_rate": 5.299264416130493e-05, "loss": 0.0325, "step": 50690 }, { "grad_norm": 0.30404800176620483, "learning_rate": 5.297613894304928e-05, "loss": 0.0344, "step": 50700 }, { "grad_norm": 0.36225518584251404, "learning_rate": 5.2959633399327534e-05, "loss": 0.0267, "step": 50710 }, { "grad_norm": 0.43609026074409485, "learning_rate": 5.294312753194476e-05, "loss": 0.0283, "step": 50720 }, { "grad_norm": 0.37744832038879395, "learning_rate": 5.292662134270596e-05, "loss": 0.0398, "step": 50730 }, { "grad_norm": 0.5004703402519226, "learning_rate": 5.291011483341626e-05, "loss": 0.0402, "step": 50740 }, { "grad_norm": 0.554520845413208, "learning_rate": 5.2893608005880767e-05, "loss": 0.046, "step": 50750 }, { "grad_norm": 0.31454089283943176, "learning_rate": 5.287710086190467e-05, "loss": 0.0365, "step": 50760 }, { "grad_norm": 0.3431588411331177, "learning_rate": 5.286059340329314e-05, "loss": 0.0331, "step": 50770 }, { "grad_norm": 0.6956056952476501, "learning_rate": 5.284408563185141e-05, "loss": 0.0432, "step": 50780 }, { "grad_norm": 0.46265652775764465, "learning_rate": 5.282757754938476e-05, "loss": 0.0383, "step": 50790 }, { "grad_norm": 0.38205426931381226, "learning_rate": 5.281106915769849e-05, "loss": 0.0404, "step": 50800 }, { "grad_norm": 0.4026148319244385, "learning_rate": 5.2794560458597897e-05, "loss": 0.0371, "step": 50810 }, { "grad_norm": 0.315339595079422, "learning_rate": 5.27780514538884e-05, "loss": 0.0308, "step": 50820 }, { "grad_norm": 0.3732113242149353, "learning_rate": 5.2761542145375365e-05, "loss": 0.0343, "step": 50830 }, { "grad_norm": 0.41492822766304016, "learning_rate": 5.274503253486421e-05, "loss": 0.0327, "step": 50840 }, { "grad_norm": 0.36198535561561584, "learning_rate": 5.272852262416046e-05, "loss": 0.034, "step": 50850 }, { "grad_norm": 0.5698140859603882, "learning_rate": 5.2712012415069555e-05, "loss": 0.0372, "step": 50860 }, { "grad_norm": 0.45671284198760986, "learning_rate": 5.269550190939705e-05, "loss": 0.037, "step": 50870 }, { "grad_norm": 0.5837885737419128, "learning_rate": 5.267899110894852e-05, "loss": 0.0374, "step": 50880 }, { "grad_norm": 0.2510669231414795, "learning_rate": 5.266248001552955e-05, "loss": 0.0357, "step": 50890 }, { "grad_norm": 0.33417317271232605, "learning_rate": 5.264596863094575e-05, "loss": 0.0375, "step": 50900 }, { "grad_norm": 1.1050735712051392, "learning_rate": 5.262945695700282e-05, "loss": 0.0365, "step": 50910 }, { "grad_norm": 0.2553815245628357, "learning_rate": 5.261294499550643e-05, "loss": 0.035, "step": 50920 }, { "grad_norm": 0.40352803468704224, "learning_rate": 5.25964327482623e-05, "loss": 0.0318, "step": 50930 }, { "grad_norm": 0.5165055394172668, "learning_rate": 5.257992021707617e-05, "loss": 0.032, "step": 50940 }, { "grad_norm": 0.35769739747047424, "learning_rate": 5.256340740375387e-05, "loss": 0.0395, "step": 50950 }, { "grad_norm": 0.2978961169719696, "learning_rate": 5.254689431010117e-05, "loss": 0.0416, "step": 50960 }, { "grad_norm": 0.30120691657066345, "learning_rate": 5.253038093792395e-05, "loss": 0.0288, "step": 50970 }, { "grad_norm": 0.3369518518447876, "learning_rate": 5.251386728902806e-05, "loss": 0.0369, "step": 50980 }, { "grad_norm": 0.4242005944252014, "learning_rate": 5.2497353365219446e-05, "loss": 0.0339, "step": 50990 }, { "grad_norm": 0.34384262561798096, "learning_rate": 5.2480839168304e-05, "loss": 0.0357, "step": 51000 }, { "grad_norm": 0.330090194940567, "learning_rate": 5.246432470008773e-05, "loss": 0.0345, "step": 51010 }, { "grad_norm": 0.3730686604976654, "learning_rate": 5.2447809962376616e-05, "loss": 0.0373, "step": 51020 }, { "grad_norm": 0.394798219203949, "learning_rate": 5.243129495697666e-05, "loss": 0.027, "step": 51030 }, { "grad_norm": 0.34922102093696594, "learning_rate": 5.241477968569397e-05, "loss": 0.0374, "step": 51040 }, { "grad_norm": 0.4104374349117279, "learning_rate": 5.23982641503346e-05, "loss": 0.0371, "step": 51050 }, { "grad_norm": 0.36500412225723267, "learning_rate": 5.238174835270465e-05, "loss": 0.0406, "step": 51060 }, { "grad_norm": 0.4204678237438202, "learning_rate": 5.23652322946103e-05, "loss": 0.0311, "step": 51070 }, { "grad_norm": 0.20979109406471252, "learning_rate": 5.234871597785771e-05, "loss": 0.025, "step": 51080 }, { "grad_norm": 0.6412791013717651, "learning_rate": 5.233219940425306e-05, "loss": 0.0453, "step": 51090 }, { "grad_norm": 0.5396692752838135, "learning_rate": 5.2315682575602576e-05, "loss": 0.0541, "step": 51100 }, { "grad_norm": 0.5238631963729858, "learning_rate": 5.229916549371254e-05, "loss": 0.0305, "step": 51110 }, { "grad_norm": 0.2996307611465454, "learning_rate": 5.228264816038924e-05, "loss": 0.0308, "step": 51120 }, { "grad_norm": 0.5224059224128723, "learning_rate": 5.2266130577438935e-05, "loss": 0.0321, "step": 51130 }, { "grad_norm": 0.32639917731285095, "learning_rate": 5.224961274666802e-05, "loss": 0.0358, "step": 51140 }, { "grad_norm": 0.7499276995658875, "learning_rate": 5.223309466988283e-05, "loss": 0.0423, "step": 51150 }, { "grad_norm": 0.347503662109375, "learning_rate": 5.221657634888976e-05, "loss": 0.0441, "step": 51160 }, { "grad_norm": 0.4186522960662842, "learning_rate": 5.2200057785495235e-05, "loss": 0.0559, "step": 51170 }, { "grad_norm": 0.3121837377548218, "learning_rate": 5.2183538981505706e-05, "loss": 0.0346, "step": 51180 }, { "grad_norm": 0.3040235638618469, "learning_rate": 5.216701993872762e-05, "loss": 0.0328, "step": 51190 }, { "grad_norm": 0.21859325468540192, "learning_rate": 5.215050065896752e-05, "loss": 0.0446, "step": 51200 }, { "grad_norm": 0.35789239406585693, "learning_rate": 5.2133981144031875e-05, "loss": 0.0548, "step": 51210 }, { "grad_norm": 0.35909339785575867, "learning_rate": 5.211746139572727e-05, "loss": 0.0395, "step": 51220 }, { "grad_norm": 0.32331424951553345, "learning_rate": 5.210094141586026e-05, "loss": 0.0314, "step": 51230 }, { "grad_norm": 0.378888875246048, "learning_rate": 5.2084421206237476e-05, "loss": 0.0413, "step": 51240 }, { "grad_norm": 0.3488468825817108, "learning_rate": 5.20679007686655e-05, "loss": 0.0407, "step": 51250 }, { "grad_norm": 0.33160892128944397, "learning_rate": 5.205138010495102e-05, "loss": 0.0402, "step": 51260 }, { "grad_norm": 0.42221254110336304, "learning_rate": 5.20348592169007e-05, "loss": 0.0368, "step": 51270 }, { "grad_norm": 0.5543147921562195, "learning_rate": 5.201833810632123e-05, "loss": 0.0327, "step": 51280 }, { "grad_norm": 0.5965652465820312, "learning_rate": 5.200181677501933e-05, "loss": 0.0383, "step": 51290 }, { "grad_norm": 0.38280701637268066, "learning_rate": 5.1985295224801777e-05, "loss": 0.0395, "step": 51300 }, { "grad_norm": 0.4184761047363281, "learning_rate": 5.196877345747531e-05, "loss": 0.0394, "step": 51310 }, { "grad_norm": 0.313572496175766, "learning_rate": 5.195225147484674e-05, "loss": 0.041, "step": 51320 }, { "grad_norm": 0.6002174019813538, "learning_rate": 5.193572927872288e-05, "loss": 0.0397, "step": 51330 }, { "grad_norm": 0.46801552176475525, "learning_rate": 5.191920687091059e-05, "loss": 0.0355, "step": 51340 }, { "grad_norm": 0.4307073652744293, "learning_rate": 5.1902684253216695e-05, "loss": 0.032, "step": 51350 }, { "grad_norm": 0.2736322581768036, "learning_rate": 5.1886161427448134e-05, "loss": 0.0361, "step": 51360 }, { "grad_norm": 0.2550186514854431, "learning_rate": 5.186963839541179e-05, "loss": 0.0309, "step": 51370 }, { "grad_norm": 0.41968223452568054, "learning_rate": 5.185311515891459e-05, "loss": 0.0244, "step": 51380 }, { "grad_norm": 0.3941729664802551, "learning_rate": 5.1836591719763516e-05, "loss": 0.0317, "step": 51390 }, { "grad_norm": 0.5546100735664368, "learning_rate": 5.182006807976554e-05, "loss": 0.0389, "step": 51400 }, { "grad_norm": 0.17458578944206238, "learning_rate": 5.180354424072763e-05, "loss": 0.0296, "step": 51410 }, { "grad_norm": 0.5236656665802002, "learning_rate": 5.178702020445686e-05, "loss": 0.0371, "step": 51420 }, { "grad_norm": 0.22040840983390808, "learning_rate": 5.177049597276024e-05, "loss": 0.0332, "step": 51430 }, { "grad_norm": 0.3819831311702728, "learning_rate": 5.175397154744482e-05, "loss": 0.0353, "step": 51440 }, { "grad_norm": 0.391254186630249, "learning_rate": 5.1737446930317736e-05, "loss": 0.0274, "step": 51450 }, { "grad_norm": 0.34035903215408325, "learning_rate": 5.172092212318604e-05, "loss": 0.0389, "step": 51460 }, { "grad_norm": 0.2616373896598816, "learning_rate": 5.170439712785692e-05, "loss": 0.0401, "step": 51470 }, { "grad_norm": 0.4397529065608978, "learning_rate": 5.1687871946137467e-05, "loss": 0.0307, "step": 51480 }, { "grad_norm": 0.37916702032089233, "learning_rate": 5.16713465798349e-05, "loss": 0.0332, "step": 51490 }, { "grad_norm": 0.3553253710269928, "learning_rate": 5.165482103075637e-05, "loss": 0.0362, "step": 51500 }, { "grad_norm": 0.39639756083488464, "learning_rate": 5.163829530070909e-05, "loss": 0.0289, "step": 51510 }, { "grad_norm": 0.5507827997207642, "learning_rate": 5.162176939150032e-05, "loss": 0.0393, "step": 51520 }, { "grad_norm": 0.4097708463668823, "learning_rate": 5.160524330493728e-05, "loss": 0.0322, "step": 51530 }, { "grad_norm": 0.27947676181793213, "learning_rate": 5.158871704282724e-05, "loss": 0.0292, "step": 51540 }, { "grad_norm": 0.3464164733886719, "learning_rate": 5.15721906069775e-05, "loss": 0.0393, "step": 51550 }, { "grad_norm": 0.3398383557796478, "learning_rate": 5.155566399919536e-05, "loss": 0.0356, "step": 51560 }, { "grad_norm": 0.43282631039619446, "learning_rate": 5.153913722128813e-05, "loss": 0.0366, "step": 51570 }, { "grad_norm": 0.5381327271461487, "learning_rate": 5.1522610275063176e-05, "loss": 0.034, "step": 51580 }, { "grad_norm": 0.40384289622306824, "learning_rate": 5.150608316232786e-05, "loss": 0.0286, "step": 51590 }, { "grad_norm": 0.36241668462753296, "learning_rate": 5.148955588488955e-05, "loss": 0.0286, "step": 51600 }, { "grad_norm": 0.3169262409210205, "learning_rate": 5.147302844455564e-05, "loss": 0.0339, "step": 51610 }, { "grad_norm": 0.3014815151691437, "learning_rate": 5.145650084313356e-05, "loss": 0.0385, "step": 51620 }, { "grad_norm": 0.4043462574481964, "learning_rate": 5.1439973082430734e-05, "loss": 0.0371, "step": 51630 }, { "grad_norm": 0.32199788093566895, "learning_rate": 5.142344516425461e-05, "loss": 0.0308, "step": 51640 }, { "grad_norm": 0.2885124385356903, "learning_rate": 5.1406917090412673e-05, "loss": 0.0399, "step": 51650 }, { "grad_norm": 0.3569343388080597, "learning_rate": 5.1390388862712404e-05, "loss": 0.034, "step": 51660 }, { "grad_norm": 0.584463894367218, "learning_rate": 5.137386048296129e-05, "loss": 0.0343, "step": 51670 }, { "grad_norm": 0.5035604238510132, "learning_rate": 5.135733195296686e-05, "loss": 0.0438, "step": 51680 }, { "grad_norm": 0.48767849802970886, "learning_rate": 5.134080327453664e-05, "loss": 0.0394, "step": 51690 }, { "grad_norm": 0.23323309421539307, "learning_rate": 5.1324274449478204e-05, "loss": 0.0338, "step": 51700 }, { "grad_norm": 0.27818557620048523, "learning_rate": 5.130774547959909e-05, "loss": 0.0403, "step": 51710 }, { "grad_norm": 0.3163018226623535, "learning_rate": 5.129121636670692e-05, "loss": 0.032, "step": 51720 }, { "grad_norm": 0.33890169858932495, "learning_rate": 5.1274687112609246e-05, "loss": 0.0244, "step": 51730 }, { "grad_norm": 0.4165394902229309, "learning_rate": 5.125815771911372e-05, "loss": 0.0382, "step": 51740 }, { "grad_norm": 0.5107395648956299, "learning_rate": 5.1241628188027956e-05, "loss": 0.0329, "step": 51750 }, { "grad_norm": 0.428524911403656, "learning_rate": 5.1225098521159585e-05, "loss": 0.0306, "step": 51760 }, { "grad_norm": 0.33598729968070984, "learning_rate": 5.12085687203163e-05, "loss": 0.0284, "step": 51770 }, { "grad_norm": 0.4747737646102905, "learning_rate": 5.119203878730575e-05, "loss": 0.0337, "step": 51780 }, { "grad_norm": 0.3517257571220398, "learning_rate": 5.117550872393564e-05, "loss": 0.0451, "step": 51790 }, { "grad_norm": 0.4019688367843628, "learning_rate": 5.1158978532013645e-05, "loss": 0.0375, "step": 51800 }, { "grad_norm": 0.28832077980041504, "learning_rate": 5.114244821334751e-05, "loss": 0.0365, "step": 51810 }, { "grad_norm": 0.34906038641929626, "learning_rate": 5.112591776974497e-05, "loss": 0.0304, "step": 51820 }, { "grad_norm": 0.34795644879341125, "learning_rate": 5.110938720301374e-05, "loss": 0.0299, "step": 51830 }, { "grad_norm": 0.18663249909877777, "learning_rate": 5.1092856514961595e-05, "loss": 0.0365, "step": 51840 }, { "grad_norm": 0.5334008932113647, "learning_rate": 5.107632570739631e-05, "loss": 0.0354, "step": 51850 }, { "grad_norm": 0.2387993335723877, "learning_rate": 5.1059794782125645e-05, "loss": 0.0445, "step": 51860 }, { "grad_norm": 0.5247085094451904, "learning_rate": 5.104326374095744e-05, "loss": 0.0433, "step": 51870 }, { "grad_norm": 0.3441837430000305, "learning_rate": 5.102673258569947e-05, "loss": 0.0337, "step": 51880 }, { "grad_norm": 0.4796609580516815, "learning_rate": 5.101020131815956e-05, "loss": 0.044, "step": 51890 }, { "grad_norm": 0.4379118084907532, "learning_rate": 5.099366994014556e-05, "loss": 0.0379, "step": 51900 }, { "grad_norm": 0.4614473581314087, "learning_rate": 5.097713845346532e-05, "loss": 0.0411, "step": 51910 }, { "grad_norm": 0.40323546528816223, "learning_rate": 5.096060685992668e-05, "loss": 0.0345, "step": 51920 }, { "grad_norm": 0.4850379228591919, "learning_rate": 5.0944075161337514e-05, "loss": 0.0452, "step": 51930 }, { "grad_norm": 0.4948246479034424, "learning_rate": 5.09275433595057e-05, "loss": 0.0361, "step": 51940 }, { "grad_norm": 0.4982747733592987, "learning_rate": 5.0911011456239157e-05, "loss": 0.0338, "step": 51950 }, { "grad_norm": 0.4522424638271332, "learning_rate": 5.089447945334574e-05, "loss": 0.0392, "step": 51960 }, { "grad_norm": 0.7708808183670044, "learning_rate": 5.087794735263343e-05, "loss": 0.0381, "step": 51970 }, { "grad_norm": 0.2964123785495758, "learning_rate": 5.086141515591011e-05, "loss": 0.0393, "step": 51980 }, { "grad_norm": 0.4427168369293213, "learning_rate": 5.084488286498371e-05, "loss": 0.0358, "step": 51990 }, { "grad_norm": 0.3347538113594055, "learning_rate": 5.08283504816622e-05, "loss": 0.0468, "step": 52000 }, { "grad_norm": 0.40078794956207275, "learning_rate": 5.081181800775353e-05, "loss": 0.0444, "step": 52010 }, { "grad_norm": 0.2510037124156952, "learning_rate": 5.079528544506565e-05, "loss": 0.0356, "step": 52020 }, { "grad_norm": 0.3948441743850708, "learning_rate": 5.077875279540657e-05, "loss": 0.032, "step": 52030 }, { "grad_norm": 0.35480886697769165, "learning_rate": 5.076222006058424e-05, "loss": 0.0409, "step": 52040 }, { "grad_norm": 0.3920900821685791, "learning_rate": 5.074568724240667e-05, "loss": 0.029, "step": 52050 }, { "grad_norm": 0.41890445351600647, "learning_rate": 5.072915434268187e-05, "loss": 0.0357, "step": 52060 }, { "grad_norm": 0.35875481367111206, "learning_rate": 5.071262136321785e-05, "loss": 0.037, "step": 52070 }, { "grad_norm": 0.5594493746757507, "learning_rate": 5.069608830582262e-05, "loss": 0.0424, "step": 52080 }, { "grad_norm": 0.2859925627708435, "learning_rate": 5.067955517230423e-05, "loss": 0.0371, "step": 52090 }, { "grad_norm": 0.3989155888557434, "learning_rate": 5.06630219644707e-05, "loss": 0.0445, "step": 52100 }, { "grad_norm": 0.4420238733291626, "learning_rate": 5.064648868413008e-05, "loss": 0.0349, "step": 52110 }, { "grad_norm": 0.5860395431518555, "learning_rate": 5.062995533309044e-05, "loss": 0.0368, "step": 52120 }, { "grad_norm": 0.21951007843017578, "learning_rate": 5.061342191315983e-05, "loss": 0.0371, "step": 52130 }, { "grad_norm": 0.45817914605140686, "learning_rate": 5.0596888426146325e-05, "loss": 0.0391, "step": 52140 }, { "grad_norm": 0.47392067313194275, "learning_rate": 5.058035487385798e-05, "loss": 0.0288, "step": 52150 }, { "grad_norm": 0.3627624809741974, "learning_rate": 5.0563821258102915e-05, "loss": 0.0339, "step": 52160 }, { "grad_norm": 0.6030815839767456, "learning_rate": 5.05472875806892e-05, "loss": 0.0319, "step": 52170 }, { "grad_norm": 0.338174968957901, "learning_rate": 5.053075384342494e-05, "loss": 0.0383, "step": 52180 }, { "grad_norm": 0.3283904492855072, "learning_rate": 5.0514220048118234e-05, "loss": 0.038, "step": 52190 }, { "grad_norm": 1.2147618532180786, "learning_rate": 5.049768619657721e-05, "loss": 0.0415, "step": 52200 }, { "grad_norm": 0.28615596890449524, "learning_rate": 5.048115229060996e-05, "loss": 0.0324, "step": 52210 }, { "grad_norm": 0.608102560043335, "learning_rate": 5.046461833202462e-05, "loss": 0.036, "step": 52220 }, { "grad_norm": 0.4912492334842682, "learning_rate": 5.044808432262932e-05, "loss": 0.0433, "step": 52230 }, { "grad_norm": 0.2679615616798401, "learning_rate": 5.043155026423219e-05, "loss": 0.0296, "step": 52240 }, { "grad_norm": 0.26254886388778687, "learning_rate": 5.041501615864137e-05, "loss": 0.0311, "step": 52250 }, { "grad_norm": 0.23602180182933807, "learning_rate": 5.039848200766502e-05, "loss": 0.0287, "step": 52260 }, { "grad_norm": 0.313018262386322, "learning_rate": 5.038194781311125e-05, "loss": 0.0329, "step": 52270 }, { "grad_norm": 0.23566187918186188, "learning_rate": 5.0365413576788264e-05, "loss": 0.0277, "step": 52280 }, { "grad_norm": 0.2882181704044342, "learning_rate": 5.0348879300504194e-05, "loss": 0.0307, "step": 52290 }, { "grad_norm": 0.30273109674453735, "learning_rate": 5.033234498606721e-05, "loss": 0.0246, "step": 52300 }, { "grad_norm": 0.3207676112651825, "learning_rate": 5.0315810635285476e-05, "loss": 0.0362, "step": 52310 }, { "grad_norm": 0.49218466877937317, "learning_rate": 5.0299276249967164e-05, "loss": 0.0314, "step": 52320 }, { "grad_norm": 0.28501877188682556, "learning_rate": 5.0282741831920454e-05, "loss": 0.0464, "step": 52330 }, { "grad_norm": 0.45520639419555664, "learning_rate": 5.026620738295351e-05, "loss": 0.0364, "step": 52340 }, { "grad_norm": 0.3173277974128723, "learning_rate": 5.024967290487455e-05, "loss": 0.0272, "step": 52350 }, { "grad_norm": 0.362511545419693, "learning_rate": 5.023313839949172e-05, "loss": 0.0325, "step": 52360 }, { "grad_norm": 0.3385850191116333, "learning_rate": 5.0216603868613224e-05, "loss": 0.0381, "step": 52370 }, { "grad_norm": 0.27992862462997437, "learning_rate": 5.020006931404726e-05, "loss": 0.0382, "step": 52380 }, { "grad_norm": 0.3090992569923401, "learning_rate": 5.0183534737602014e-05, "loss": 0.0232, "step": 52390 }, { "grad_norm": 0.27860841155052185, "learning_rate": 5.0167000141085686e-05, "loss": 0.0275, "step": 52400 }, { "grad_norm": 0.19635675847530365, "learning_rate": 5.015046552630647e-05, "loss": 0.0316, "step": 52410 }, { "grad_norm": 0.33850452303886414, "learning_rate": 5.013393089507258e-05, "loss": 0.0298, "step": 52420 }, { "grad_norm": 0.32349488139152527, "learning_rate": 5.011739624919219e-05, "loss": 0.0281, "step": 52430 }, { "grad_norm": 0.23633666336536407, "learning_rate": 5.0100861590473545e-05, "loss": 0.0383, "step": 52440 }, { "grad_norm": 0.358757346868515, "learning_rate": 5.008432692072482e-05, "loss": 0.0293, "step": 52450 }, { "grad_norm": 0.5711537599563599, "learning_rate": 5.006779224175423e-05, "loss": 0.0425, "step": 52460 }, { "grad_norm": 0.36365506052970886, "learning_rate": 5.005125755536998e-05, "loss": 0.0345, "step": 52470 }, { "grad_norm": 0.37938156723976135, "learning_rate": 5.003472286338029e-05, "loss": 0.0334, "step": 52480 }, { "grad_norm": 0.2752685546875, "learning_rate": 5.001818816759335e-05, "loss": 0.0475, "step": 52490 }, { "grad_norm": 0.4154776930809021, "learning_rate": 5.000165346981738e-05, "loss": 0.0319, "step": 52500 }, { "grad_norm": 0.34285587072372437, "learning_rate": 4.9985118771860595e-05, "loss": 0.0368, "step": 52510 }, { "grad_norm": 0.6245807409286499, "learning_rate": 4.996858407553119e-05, "loss": 0.0295, "step": 52520 }, { "grad_norm": 0.5591354966163635, "learning_rate": 4.99520493826374e-05, "loss": 0.0328, "step": 52530 }, { "grad_norm": 0.3234151303768158, "learning_rate": 4.9935514694987384e-05, "loss": 0.0385, "step": 52540 }, { "grad_norm": 0.6565362811088562, "learning_rate": 4.991898001438941e-05, "loss": 0.0453, "step": 52550 }, { "grad_norm": 0.43640705943107605, "learning_rate": 4.990244534265166e-05, "loss": 0.0364, "step": 52560 }, { "grad_norm": 0.4575926661491394, "learning_rate": 4.988591068158234e-05, "loss": 0.0319, "step": 52570 }, { "grad_norm": 0.39253056049346924, "learning_rate": 4.986937603298965e-05, "loss": 0.0344, "step": 52580 }, { "grad_norm": 0.5176910758018494, "learning_rate": 4.985284139868181e-05, "loss": 0.0297, "step": 52590 }, { "grad_norm": 0.3761419653892517, "learning_rate": 4.9836306780467e-05, "loss": 0.0355, "step": 52600 }, { "grad_norm": 0.4622974693775177, "learning_rate": 4.981977218015343e-05, "loss": 0.034, "step": 52610 }, { "grad_norm": 0.5587652325630188, "learning_rate": 4.980323759954932e-05, "loss": 0.0366, "step": 52620 }, { "grad_norm": 0.41468125581741333, "learning_rate": 4.978670304046284e-05, "loss": 0.0349, "step": 52630 }, { "grad_norm": 0.3367239236831665, "learning_rate": 4.9770168504702195e-05, "loss": 0.0378, "step": 52640 }, { "grad_norm": 0.3969710171222687, "learning_rate": 4.975363399407559e-05, "loss": 0.0329, "step": 52650 }, { "grad_norm": 0.4463009238243103, "learning_rate": 4.9737099510391194e-05, "loss": 0.0391, "step": 52660 }, { "grad_norm": 0.5990653038024902, "learning_rate": 4.972056505545719e-05, "loss": 0.0476, "step": 52670 }, { "grad_norm": 0.3298162817955017, "learning_rate": 4.9704030631081775e-05, "loss": 0.0329, "step": 52680 }, { "grad_norm": 0.22189828753471375, "learning_rate": 4.968749623907313e-05, "loss": 0.0308, "step": 52690 }, { "grad_norm": 0.542716383934021, "learning_rate": 4.967096188123943e-05, "loss": 0.0376, "step": 52700 }, { "grad_norm": 0.23452487587928772, "learning_rate": 4.965442755938884e-05, "loss": 0.0284, "step": 52710 }, { "grad_norm": 0.785999059677124, "learning_rate": 4.963789327532954e-05, "loss": 0.0338, "step": 52720 }, { "grad_norm": 0.4000718593597412, "learning_rate": 4.962135903086968e-05, "loss": 0.0345, "step": 52730 }, { "grad_norm": 0.40297168493270874, "learning_rate": 4.960482482781741e-05, "loss": 0.0331, "step": 52740 }, { "grad_norm": 0.37952905893325806, "learning_rate": 4.958829066798092e-05, "loss": 0.0317, "step": 52750 }, { "grad_norm": 0.3288222551345825, "learning_rate": 4.957175655316835e-05, "loss": 0.0311, "step": 52760 }, { "grad_norm": 0.4322054982185364, "learning_rate": 4.9555222485187834e-05, "loss": 0.0388, "step": 52770 }, { "grad_norm": 0.41150906682014465, "learning_rate": 4.953868846584752e-05, "loss": 0.031, "step": 52780 }, { "grad_norm": 0.4538201689720154, "learning_rate": 4.9522154496955544e-05, "loss": 0.0383, "step": 52790 }, { "grad_norm": 0.27552759647369385, "learning_rate": 4.950562058032001e-05, "loss": 0.0372, "step": 52800 }, { "grad_norm": 0.41936326026916504, "learning_rate": 4.948908671774909e-05, "loss": 0.0357, "step": 52810 }, { "grad_norm": 0.3025553524494171, "learning_rate": 4.947255291105087e-05, "loss": 0.0361, "step": 52820 }, { "grad_norm": 0.43100297451019287, "learning_rate": 4.945601916203348e-05, "loss": 0.0308, "step": 52830 }, { "grad_norm": 0.39775341749191284, "learning_rate": 4.9439485472505016e-05, "loss": 0.0304, "step": 52840 }, { "grad_norm": 0.24586156010627747, "learning_rate": 4.942295184427356e-05, "loss": 0.034, "step": 52850 }, { "grad_norm": 0.2944208085536957, "learning_rate": 4.9406418279147234e-05, "loss": 0.0399, "step": 52860 }, { "grad_norm": 0.8739572763442993, "learning_rate": 4.9389884778934116e-05, "loss": 0.0339, "step": 52870 }, { "grad_norm": 1.4056062698364258, "learning_rate": 4.937335134544228e-05, "loss": 0.0352, "step": 52880 }, { "grad_norm": 0.32498958706855774, "learning_rate": 4.935681798047979e-05, "loss": 0.0469, "step": 52890 }, { "grad_norm": 0.3094947934150696, "learning_rate": 4.934028468585473e-05, "loss": 0.0294, "step": 52900 }, { "grad_norm": 0.4246600866317749, "learning_rate": 4.932375146337514e-05, "loss": 0.0375, "step": 52910 }, { "grad_norm": 0.7356343865394592, "learning_rate": 4.930721831484908e-05, "loss": 0.0386, "step": 52920 }, { "grad_norm": 0.4462871849536896, "learning_rate": 4.929068524208456e-05, "loss": 0.0428, "step": 52930 }, { "grad_norm": 1.098880648612976, "learning_rate": 4.927415224688965e-05, "loss": 0.0443, "step": 52940 }, { "grad_norm": 0.36653614044189453, "learning_rate": 4.9257619331072375e-05, "loss": 0.0399, "step": 52950 }, { "grad_norm": 0.8259483575820923, "learning_rate": 4.924108649644072e-05, "loss": 0.0391, "step": 52960 }, { "grad_norm": 0.29122957587242126, "learning_rate": 4.922455374480271e-05, "loss": 0.0453, "step": 52970 }, { "grad_norm": 0.307866632938385, "learning_rate": 4.920802107796634e-05, "loss": 0.0318, "step": 52980 }, { "grad_norm": 0.4429660439491272, "learning_rate": 4.9191488497739564e-05, "loss": 0.0375, "step": 52990 }, { "grad_norm": 0.3162130117416382, "learning_rate": 4.917495600593042e-05, "loss": 0.0342, "step": 53000 }, { "grad_norm": 0.656091034412384, "learning_rate": 4.915842360434684e-05, "loss": 0.0358, "step": 53010 }, { "grad_norm": 2.5889363288879395, "learning_rate": 4.914189129479679e-05, "loss": 0.0474, "step": 53020 }, { "grad_norm": 0.2943163216114044, "learning_rate": 4.912535907908822e-05, "loss": 0.041, "step": 53030 }, { "grad_norm": 0.5482437014579773, "learning_rate": 4.910882695902907e-05, "loss": 0.0423, "step": 53040 }, { "grad_norm": 0.514850914478302, "learning_rate": 4.9092294936427234e-05, "loss": 0.032, "step": 53050 }, { "grad_norm": 0.36142101883888245, "learning_rate": 4.9075763013090685e-05, "loss": 0.0346, "step": 53060 }, { "grad_norm": 0.562315821647644, "learning_rate": 4.9059231190827305e-05, "loss": 0.034, "step": 53070 }, { "grad_norm": 0.32737529277801514, "learning_rate": 4.904269947144499e-05, "loss": 0.0452, "step": 53080 }, { "grad_norm": 0.4753693640232086, "learning_rate": 4.902616785675161e-05, "loss": 0.0428, "step": 53090 }, { "grad_norm": 0.47599416971206665, "learning_rate": 4.900963634855505e-05, "loss": 0.0361, "step": 53100 }, { "grad_norm": 0.44171011447906494, "learning_rate": 4.8993104948663173e-05, "loss": 0.0267, "step": 53110 }, { "grad_norm": 0.31261053681373596, "learning_rate": 4.8976573658883823e-05, "loss": 0.0283, "step": 53120 }, { "grad_norm": 0.3576388657093048, "learning_rate": 4.896004248102484e-05, "loss": 0.0348, "step": 53130 }, { "grad_norm": 0.35039907693862915, "learning_rate": 4.8943511416894045e-05, "loss": 0.034, "step": 53140 }, { "grad_norm": 0.5203590393066406, "learning_rate": 4.892698046829926e-05, "loss": 0.034, "step": 53150 }, { "grad_norm": 0.495601087808609, "learning_rate": 4.8910449637048275e-05, "loss": 0.0521, "step": 53160 }, { "grad_norm": 0.4595177173614502, "learning_rate": 4.8893918924948884e-05, "loss": 0.0371, "step": 53170 }, { "grad_norm": 0.2989412844181061, "learning_rate": 4.8877388333808835e-05, "loss": 0.0372, "step": 53180 }, { "grad_norm": 0.3054579198360443, "learning_rate": 4.8860857865435925e-05, "loss": 0.0394, "step": 53190 }, { "grad_norm": 0.4713558852672577, "learning_rate": 4.884432752163789e-05, "loss": 0.0297, "step": 53200 }, { "grad_norm": 0.5669779777526855, "learning_rate": 4.8827797304222466e-05, "loss": 0.038, "step": 53210 }, { "grad_norm": 0.6445514559745789, "learning_rate": 4.8811267214997366e-05, "loss": 0.0313, "step": 53220 }, { "grad_norm": 0.36411142349243164, "learning_rate": 4.879473725577029e-05, "loss": 0.0298, "step": 53230 }, { "grad_norm": 0.5397136807441711, "learning_rate": 4.877820742834891e-05, "loss": 0.0331, "step": 53240 }, { "grad_norm": 0.477557897567749, "learning_rate": 4.8761677734540956e-05, "loss": 0.0378, "step": 53250 }, { "grad_norm": 0.4026028513908386, "learning_rate": 4.874514817615407e-05, "loss": 0.0378, "step": 53260 }, { "grad_norm": 0.3424171805381775, "learning_rate": 4.872861875499588e-05, "loss": 0.0389, "step": 53270 }, { "grad_norm": 0.4143913984298706, "learning_rate": 4.871208947287404e-05, "loss": 0.0368, "step": 53280 }, { "grad_norm": 0.3628387153148651, "learning_rate": 4.869556033159615e-05, "loss": 0.0326, "step": 53290 }, { "grad_norm": 0.23162733018398285, "learning_rate": 4.8679031332969835e-05, "loss": 0.0369, "step": 53300 }, { "grad_norm": 0.2743053436279297, "learning_rate": 4.8662502478802635e-05, "loss": 0.041, "step": 53310 }, { "grad_norm": 0.3492341637611389, "learning_rate": 4.8645973770902176e-05, "loss": 0.0352, "step": 53320 }, { "grad_norm": 0.34084147214889526, "learning_rate": 4.862944521107599e-05, "loss": 0.0249, "step": 53330 }, { "grad_norm": 0.4636216461658478, "learning_rate": 4.861291680113162e-05, "loss": 0.0314, "step": 53340 }, { "grad_norm": 0.4407667815685272, "learning_rate": 4.8596388542876563e-05, "loss": 0.0349, "step": 53350 }, { "grad_norm": 0.45785626769065857, "learning_rate": 4.857986043811836e-05, "loss": 0.0359, "step": 53360 }, { "grad_norm": 0.42240723967552185, "learning_rate": 4.856333248866448e-05, "loss": 0.0316, "step": 53370 }, { "grad_norm": 0.3543103039264679, "learning_rate": 4.85468046963224e-05, "loss": 0.0296, "step": 53380 }, { "grad_norm": 0.5278281569480896, "learning_rate": 4.853027706289957e-05, "loss": 0.0288, "step": 53390 }, { "grad_norm": 0.49814414978027344, "learning_rate": 4.851374959020344e-05, "loss": 0.0407, "step": 53400 }, { "grad_norm": 0.4918445646762848, "learning_rate": 4.849722228004141e-05, "loss": 0.0332, "step": 53410 }, { "grad_norm": 0.4640592038631439, "learning_rate": 4.848069513422091e-05, "loss": 0.0356, "step": 53420 }, { "grad_norm": 0.32591500878334045, "learning_rate": 4.8464168154549294e-05, "loss": 0.0351, "step": 53430 }, { "grad_norm": 0.33904099464416504, "learning_rate": 4.844764134283392e-05, "loss": 0.0328, "step": 53440 }, { "grad_norm": 0.38878557085990906, "learning_rate": 4.843111470088218e-05, "loss": 0.0442, "step": 53450 }, { "grad_norm": 0.3689512014389038, "learning_rate": 4.841458823050137e-05, "loss": 0.0341, "step": 53460 }, { "grad_norm": 0.4591483473777771, "learning_rate": 4.8398061933498816e-05, "loss": 0.0335, "step": 53470 }, { "grad_norm": 0.5736628174781799, "learning_rate": 4.83815358116818e-05, "loss": 0.0324, "step": 53480 }, { "grad_norm": 0.34497490525245667, "learning_rate": 4.8365009866857584e-05, "loss": 0.0348, "step": 53490 }, { "grad_norm": 0.28303688764572144, "learning_rate": 4.834848410083342e-05, "loss": 0.0339, "step": 53500 }, { "grad_norm": 0.3454190492630005, "learning_rate": 4.833195851541657e-05, "loss": 0.0346, "step": 53510 }, { "grad_norm": 0.3139602839946747, "learning_rate": 4.831543311241423e-05, "loss": 0.0318, "step": 53520 }, { "grad_norm": 0.4701768457889557, "learning_rate": 4.829890789363359e-05, "loss": 0.0443, "step": 53530 }, { "grad_norm": 0.38655540347099304, "learning_rate": 4.828238286088182e-05, "loss": 0.0311, "step": 53540 }, { "grad_norm": 0.3362083435058594, "learning_rate": 4.826585801596609e-05, "loss": 0.0335, "step": 53550 }, { "grad_norm": 0.36210399866104126, "learning_rate": 4.8249333360693486e-05, "loss": 0.0306, "step": 53560 }, { "grad_norm": 0.36067458987236023, "learning_rate": 4.8232808896871185e-05, "loss": 0.0317, "step": 53570 }, { "grad_norm": 0.2938287556171417, "learning_rate": 4.821628462630624e-05, "loss": 0.0338, "step": 53580 }, { "grad_norm": 0.2499762624502182, "learning_rate": 4.8199760550805735e-05, "loss": 0.0326, "step": 53590 }, { "grad_norm": 0.4464861750602722, "learning_rate": 4.818323667217669e-05, "loss": 0.0344, "step": 53600 }, { "grad_norm": 0.3649928569793701, "learning_rate": 4.8166712992226165e-05, "loss": 0.0315, "step": 53610 }, { "grad_norm": 0.39339694380760193, "learning_rate": 4.8150189512761154e-05, "loss": 0.0402, "step": 53620 }, { "grad_norm": 0.3517951965332031, "learning_rate": 4.813366623558862e-05, "loss": 0.0344, "step": 53630 }, { "grad_norm": 0.4754473567008972, "learning_rate": 4.811714316251554e-05, "loss": 0.0376, "step": 53640 }, { "grad_norm": 0.31175997853279114, "learning_rate": 4.810062029534886e-05, "loss": 0.0324, "step": 53650 }, { "grad_norm": 0.28287750482559204, "learning_rate": 4.8084097635895505e-05, "loss": 0.0289, "step": 53660 }, { "grad_norm": 0.5015755295753479, "learning_rate": 4.8067575185962335e-05, "loss": 0.0405, "step": 53670 }, { "grad_norm": 0.6318278908729553, "learning_rate": 4.805105294735625e-05, "loss": 0.0298, "step": 53680 }, { "grad_norm": 0.39057886600494385, "learning_rate": 4.803453092188406e-05, "loss": 0.0359, "step": 53690 }, { "grad_norm": 0.42850497364997864, "learning_rate": 4.801800911135263e-05, "loss": 0.0316, "step": 53700 }, { "grad_norm": 0.2745426893234253, "learning_rate": 4.8001487517568744e-05, "loss": 0.0347, "step": 53710 }, { "grad_norm": 0.35601770877838135, "learning_rate": 4.798496614233918e-05, "loss": 0.0343, "step": 53720 }, { "grad_norm": 0.29801151156425476, "learning_rate": 4.796844498747069e-05, "loss": 0.0441, "step": 53730 }, { "grad_norm": 0.3093523383140564, "learning_rate": 4.795192405477e-05, "loss": 0.0249, "step": 53740 }, { "grad_norm": 0.4313792586326599, "learning_rate": 4.79354033460438e-05, "loss": 0.0369, "step": 53750 }, { "grad_norm": 1.5073111057281494, "learning_rate": 4.79188828630988e-05, "loss": 0.045, "step": 53760 }, { "grad_norm": 0.5415828227996826, "learning_rate": 4.790236260774165e-05, "loss": 0.0361, "step": 53770 }, { "grad_norm": 0.3537231683731079, "learning_rate": 4.788584258177896e-05, "loss": 0.036, "step": 53780 }, { "grad_norm": 0.3335283100605011, "learning_rate": 4.7869322787017356e-05, "loss": 0.0324, "step": 53790 }, { "grad_norm": 0.3582208454608917, "learning_rate": 4.78528032252634e-05, "loss": 0.0303, "step": 53800 }, { "grad_norm": 0.3345978558063507, "learning_rate": 4.783628389832366e-05, "loss": 0.0401, "step": 53810 }, { "grad_norm": 0.4285871386528015, "learning_rate": 4.7819764808004644e-05, "loss": 0.0431, "step": 53820 }, { "grad_norm": 0.489126980304718, "learning_rate": 4.780324595611289e-05, "loss": 0.0293, "step": 53830 }, { "grad_norm": 0.5362392067909241, "learning_rate": 4.7786727344454854e-05, "loss": 0.0433, "step": 53840 }, { "grad_norm": 0.497326135635376, "learning_rate": 4.777020897483697e-05, "loss": 0.0357, "step": 53850 }, { "grad_norm": 0.36812838912010193, "learning_rate": 4.775369084906569e-05, "loss": 0.0282, "step": 53860 }, { "grad_norm": 0.4515267610549927, "learning_rate": 4.773717296894742e-05, "loss": 0.0472, "step": 53870 }, { "grad_norm": 0.2981516718864441, "learning_rate": 4.772065533628848e-05, "loss": 0.0414, "step": 53880 }, { "grad_norm": 0.45100757479667664, "learning_rate": 4.7704137952895265e-05, "loss": 0.0303, "step": 53890 }, { "grad_norm": 0.5220188498497009, "learning_rate": 4.768762082057407e-05, "loss": 0.0246, "step": 53900 }, { "grad_norm": 0.4359012842178345, "learning_rate": 4.7671103941131186e-05, "loss": 0.0318, "step": 53910 }, { "grad_norm": 0.24858073890209198, "learning_rate": 4.765458731637287e-05, "loss": 0.0319, "step": 53920 }, { "grad_norm": 0.31472048163414, "learning_rate": 4.763807094810537e-05, "loss": 0.0393, "step": 53930 }, { "grad_norm": 0.5239877104759216, "learning_rate": 4.762155483813487e-05, "loss": 0.044, "step": 53940 }, { "grad_norm": 0.5119889378547668, "learning_rate": 4.760503898826754e-05, "loss": 0.0404, "step": 53950 }, { "grad_norm": 0.39901888370513916, "learning_rate": 4.758852340030957e-05, "loss": 0.0467, "step": 53960 }, { "grad_norm": 0.35015857219696045, "learning_rate": 4.757200807606705e-05, "loss": 0.0328, "step": 53970 }, { "grad_norm": 0.3556545078754425, "learning_rate": 4.755549301734607e-05, "loss": 0.0299, "step": 53980 }, { "grad_norm": 0.39335089921951294, "learning_rate": 4.753897822595271e-05, "loss": 0.0479, "step": 53990 }, { "grad_norm": 0.5673704147338867, "learning_rate": 4.7522463703692994e-05, "loss": 0.0296, "step": 54000 }, { "grad_norm": 0.32636529207229614, "learning_rate": 4.75059494523729e-05, "loss": 0.0326, "step": 54010 }, { "grad_norm": 0.38909661769866943, "learning_rate": 4.748943547379843e-05, "loss": 0.0328, "step": 54020 }, { "grad_norm": 0.25767138600349426, "learning_rate": 4.7472921769775535e-05, "loss": 0.0465, "step": 54030 }, { "grad_norm": 0.6314030289649963, "learning_rate": 4.745640834211011e-05, "loss": 0.0423, "step": 54040 }, { "grad_norm": 0.27049726247787476, "learning_rate": 4.7439895192608054e-05, "loss": 0.0317, "step": 54050 }, { "grad_norm": 0.4567681849002838, "learning_rate": 4.7423382323075186e-05, "loss": 0.0338, "step": 54060 }, { "grad_norm": 0.43628835678100586, "learning_rate": 4.740686973531737e-05, "loss": 0.0265, "step": 54070 }, { "grad_norm": 0.4829377830028534, "learning_rate": 4.7390357431140375e-05, "loss": 0.0345, "step": 54080 }, { "grad_norm": 1.6056571006774902, "learning_rate": 4.737384541234998e-05, "loss": 0.04, "step": 54090 }, { "grad_norm": 0.3310962915420532, "learning_rate": 4.735733368075189e-05, "loss": 0.0279, "step": 54100 }, { "grad_norm": 0.32589879631996155, "learning_rate": 4.734082223815182e-05, "loss": 0.0362, "step": 54110 }, { "grad_norm": 0.2333831638097763, "learning_rate": 4.7324311086355436e-05, "loss": 0.0242, "step": 54120 }, { "grad_norm": 0.7643181085586548, "learning_rate": 4.730780022716836e-05, "loss": 0.0369, "step": 54130 }, { "grad_norm": 0.3669266402721405, "learning_rate": 4.729128966239619e-05, "loss": 0.0284, "step": 54140 }, { "grad_norm": 0.38276129961013794, "learning_rate": 4.7274779393844536e-05, "loss": 0.0311, "step": 54150 }, { "grad_norm": 0.6067516207695007, "learning_rate": 4.725826942331891e-05, "loss": 0.0454, "step": 54160 }, { "grad_norm": 0.44928663969039917, "learning_rate": 4.724175975262481e-05, "loss": 0.034, "step": 54170 }, { "grad_norm": 0.4613122045993805, "learning_rate": 4.722525038356772e-05, "loss": 0.0368, "step": 54180 }, { "grad_norm": 0.3552738428115845, "learning_rate": 4.7208741317953075e-05, "loss": 0.0362, "step": 54190 }, { "grad_norm": 0.28810107707977295, "learning_rate": 4.7192232557586264e-05, "loss": 0.0365, "step": 54200 }, { "grad_norm": 0.44867509603500366, "learning_rate": 4.717572410427269e-05, "loss": 0.0304, "step": 54210 }, { "grad_norm": 0.3476015627384186, "learning_rate": 4.71592159598177e-05, "loss": 0.0404, "step": 54220 }, { "grad_norm": 0.3719271719455719, "learning_rate": 4.714270812602657e-05, "loss": 0.0349, "step": 54230 }, { "grad_norm": 0.31982913613319397, "learning_rate": 4.712620060470458e-05, "loss": 0.0441, "step": 54240 }, { "grad_norm": 0.33925238251686096, "learning_rate": 4.7109693397656986e-05, "loss": 0.0381, "step": 54250 }, { "grad_norm": 0.3114888370037079, "learning_rate": 4.709318650668894e-05, "loss": 0.0303, "step": 54260 }, { "grad_norm": 0.3522703945636749, "learning_rate": 4.7076679933605676e-05, "loss": 0.0323, "step": 54270 }, { "grad_norm": 0.28980425000190735, "learning_rate": 4.70601736802123e-05, "loss": 0.03, "step": 54280 }, { "grad_norm": 0.3027462959289551, "learning_rate": 4.704366774831391e-05, "loss": 0.0451, "step": 54290 }, { "grad_norm": 0.3713890612125397, "learning_rate": 4.702716213971557e-05, "loss": 0.0327, "step": 54300 }, { "grad_norm": 0.5713773965835571, "learning_rate": 4.70106568562223e-05, "loss": 0.0412, "step": 54310 }, { "grad_norm": 0.32314079999923706, "learning_rate": 4.6994151899639105e-05, "loss": 0.0315, "step": 54320 }, { "grad_norm": 0.37822243571281433, "learning_rate": 4.697764727177093e-05, "loss": 0.0499, "step": 54330 }, { "grad_norm": 0.4446588456630707, "learning_rate": 4.696114297442272e-05, "loss": 0.0359, "step": 54340 }, { "grad_norm": 0.3092571794986725, "learning_rate": 4.694463900939933e-05, "loss": 0.0328, "step": 54350 }, { "grad_norm": 0.31751924753189087, "learning_rate": 4.692813537850564e-05, "loss": 0.0326, "step": 54360 }, { "grad_norm": 0.3769831359386444, "learning_rate": 4.6911632083546445e-05, "loss": 0.0302, "step": 54370 }, { "grad_norm": 0.33636364340782166, "learning_rate": 4.6895129126326525e-05, "loss": 0.0371, "step": 54380 }, { "grad_norm": 0.5564724206924438, "learning_rate": 4.687862650865059e-05, "loss": 0.0325, "step": 54390 }, { "grad_norm": 0.3589833676815033, "learning_rate": 4.686212423232339e-05, "loss": 0.0294, "step": 54400 }, { "grad_norm": 0.325021892786026, "learning_rate": 4.6845622299149565e-05, "loss": 0.0334, "step": 54410 }, { "grad_norm": 0.3637297451496124, "learning_rate": 4.682912071093374e-05, "loss": 0.0341, "step": 54420 }, { "grad_norm": 0.3243727684020996, "learning_rate": 4.681261946948052e-05, "loss": 0.0356, "step": 54430 }, { "grad_norm": 0.3226001262664795, "learning_rate": 4.679611857659443e-05, "loss": 0.0412, "step": 54440 }, { "grad_norm": 0.2675354480743408, "learning_rate": 4.6779618034079975e-05, "loss": 0.0366, "step": 54450 }, { "grad_norm": 0.273786336183548, "learning_rate": 4.6763117843741675e-05, "loss": 0.0372, "step": 54460 }, { "grad_norm": 0.4242912530899048, "learning_rate": 4.674661800738394e-05, "loss": 0.0335, "step": 54470 }, { "grad_norm": 0.3284193277359009, "learning_rate": 4.673011852681117e-05, "loss": 0.0309, "step": 54480 }, { "grad_norm": 0.333099901676178, "learning_rate": 4.671361940382772e-05, "loss": 0.0395, "step": 54490 }, { "grad_norm": 0.3957419991493225, "learning_rate": 4.669712064023791e-05, "loss": 0.0387, "step": 54500 }, { "grad_norm": 0.4332602918148041, "learning_rate": 4.668062223784602e-05, "loss": 0.0356, "step": 54510 }, { "grad_norm": 0.25380295515060425, "learning_rate": 4.6664124198456274e-05, "loss": 0.0329, "step": 54520 }, { "grad_norm": 0.6203634142875671, "learning_rate": 4.6647626523872905e-05, "loss": 0.0353, "step": 54530 }, { "grad_norm": 0.24648542702198029, "learning_rate": 4.6631129215900066e-05, "loss": 0.0338, "step": 54540 }, { "grad_norm": 0.5609268546104431, "learning_rate": 4.661463227634187e-05, "loss": 0.0483, "step": 54550 }, { "grad_norm": 0.3140908181667328, "learning_rate": 4.659813570700238e-05, "loss": 0.0249, "step": 54560 }, { "grad_norm": 0.2656119763851166, "learning_rate": 4.658163950968568e-05, "loss": 0.0287, "step": 54570 }, { "grad_norm": 0.53029865026474, "learning_rate": 4.6565143686195726e-05, "loss": 0.0371, "step": 54580 }, { "grad_norm": 0.43268391489982605, "learning_rate": 4.654864823833649e-05, "loss": 0.0311, "step": 54590 }, { "grad_norm": 0.48856842517852783, "learning_rate": 4.65321531679119e-05, "loss": 0.034, "step": 54600 }, { "grad_norm": 0.35361748933792114, "learning_rate": 4.6515658476725834e-05, "loss": 0.0394, "step": 54610 }, { "grad_norm": 0.3447774350643158, "learning_rate": 4.649916416658212e-05, "loss": 0.0384, "step": 54620 }, { "grad_norm": 0.58280348777771, "learning_rate": 4.648267023928455e-05, "loss": 0.0298, "step": 54630 }, { "grad_norm": 1.1745661497116089, "learning_rate": 4.646617669663687e-05, "loss": 0.0374, "step": 54640 }, { "grad_norm": 0.4471575617790222, "learning_rate": 4.644968354044277e-05, "loss": 0.031, "step": 54650 }, { "grad_norm": 0.3422950506210327, "learning_rate": 4.643319077250596e-05, "loss": 0.0339, "step": 54660 }, { "grad_norm": 2.064133405685425, "learning_rate": 4.6416698394630055e-05, "loss": 0.0293, "step": 54670 }, { "grad_norm": 0.8791743516921997, "learning_rate": 4.640020640861862e-05, "loss": 0.0322, "step": 54680 }, { "grad_norm": 0.36860331892967224, "learning_rate": 4.6383714816275195e-05, "loss": 0.0405, "step": 54690 }, { "grad_norm": 0.2992926836013794, "learning_rate": 4.636722361940329e-05, "loss": 0.0315, "step": 54700 }, { "grad_norm": 0.3103684186935425, "learning_rate": 4.635073281980632e-05, "loss": 0.0375, "step": 54710 }, { "grad_norm": 0.36289164423942566, "learning_rate": 4.633424241928775e-05, "loss": 0.0516, "step": 54720 }, { "grad_norm": 0.37331897020339966, "learning_rate": 4.6317752419650914e-05, "loss": 0.0284, "step": 54730 }, { "grad_norm": 0.46541571617126465, "learning_rate": 4.630126282269914e-05, "loss": 0.0421, "step": 54740 }, { "grad_norm": 0.27418383955955505, "learning_rate": 4.628477363023569e-05, "loss": 0.0329, "step": 54750 }, { "grad_norm": 0.3719850182533264, "learning_rate": 4.626828484406383e-05, "loss": 0.0374, "step": 54760 }, { "grad_norm": 0.5391032099723816, "learning_rate": 4.625179646598669e-05, "loss": 0.0392, "step": 54770 }, { "grad_norm": 0.22808672487735748, "learning_rate": 4.6235308497807487e-05, "loss": 0.032, "step": 54780 }, { "grad_norm": 0.39835628867149353, "learning_rate": 4.6218820941329276e-05, "loss": 0.031, "step": 54790 }, { "grad_norm": 0.3582262098789215, "learning_rate": 4.620233379835513e-05, "loss": 0.0289, "step": 54800 }, { "grad_norm": 0.5152419209480286, "learning_rate": 4.618584707068804e-05, "loss": 0.0344, "step": 54810 }, { "grad_norm": 0.44809046387672424, "learning_rate": 4.616936076013099e-05, "loss": 0.0361, "step": 54820 }, { "grad_norm": 0.38475266098976135, "learning_rate": 4.615287486848688e-05, "loss": 0.0345, "step": 54830 }, { "grad_norm": 0.6459335684776306, "learning_rate": 4.6136389397558584e-05, "loss": 0.0434, "step": 54840 }, { "grad_norm": 0.3911345601081848, "learning_rate": 4.6119904349148935e-05, "loss": 0.0256, "step": 54850 }, { "grad_norm": 0.3470599949359894, "learning_rate": 4.6103419725060726e-05, "loss": 0.0299, "step": 54860 }, { "grad_norm": 0.39952683448791504, "learning_rate": 4.608693552709667e-05, "loss": 0.0251, "step": 54870 }, { "grad_norm": 0.4435216784477234, "learning_rate": 4.607045175705947e-05, "loss": 0.0372, "step": 54880 }, { "grad_norm": 0.41650107502937317, "learning_rate": 4.605396841675176e-05, "loss": 0.0322, "step": 54890 }, { "grad_norm": 0.39266952872276306, "learning_rate": 4.6037485507976094e-05, "loss": 0.0305, "step": 54900 }, { "grad_norm": 0.32039350271224976, "learning_rate": 4.602100303253509e-05, "loss": 0.0306, "step": 54910 }, { "grad_norm": 0.8049222230911255, "learning_rate": 4.600452099223122e-05, "loss": 0.0415, "step": 54920 }, { "grad_norm": 0.3257559835910797, "learning_rate": 4.598803938886691e-05, "loss": 0.0348, "step": 54930 }, { "grad_norm": 0.2668359577655792, "learning_rate": 4.597155822424459e-05, "loss": 0.0319, "step": 54940 }, { "grad_norm": 0.4806535840034485, "learning_rate": 4.59550775001666e-05, "loss": 0.0461, "step": 54950 }, { "grad_norm": 0.41261833906173706, "learning_rate": 4.5938597218435234e-05, "loss": 0.0331, "step": 54960 }, { "grad_norm": 0.3624096214771271, "learning_rate": 4.592211738085279e-05, "loss": 0.05, "step": 54970 }, { "grad_norm": 0.34818580746650696, "learning_rate": 4.590563798922146e-05, "loss": 0.0321, "step": 54980 }, { "grad_norm": 0.30073282122612, "learning_rate": 4.5889159045343404e-05, "loss": 0.0433, "step": 54990 }, { "grad_norm": 0.3182222843170166, "learning_rate": 4.5872680551020734e-05, "loss": 0.0333, "step": 55000 }, { "grad_norm": 0.433754026889801, "learning_rate": 4.585620250805552e-05, "loss": 0.0323, "step": 55010 }, { "grad_norm": 0.45758476853370667, "learning_rate": 4.583972491824975e-05, "loss": 0.0295, "step": 55020 }, { "grad_norm": 0.4445626735687256, "learning_rate": 4.582324778340541e-05, "loss": 0.0365, "step": 55030 }, { "grad_norm": 0.3196585178375244, "learning_rate": 4.5806771105324424e-05, "loss": 0.0379, "step": 55040 }, { "grad_norm": 0.5012878179550171, "learning_rate": 4.579029488580865e-05, "loss": 0.0378, "step": 55050 }, { "grad_norm": 0.41061604022979736, "learning_rate": 4.5773819126659876e-05, "loss": 0.0245, "step": 55060 }, { "grad_norm": 0.3356929123401642, "learning_rate": 4.575734382967991e-05, "loss": 0.0309, "step": 55070 }, { "grad_norm": 0.37664297223091125, "learning_rate": 4.574086899667044e-05, "loss": 0.0321, "step": 55080 }, { "grad_norm": 0.3752710521221161, "learning_rate": 4.5724394629433106e-05, "loss": 0.0313, "step": 55090 }, { "grad_norm": 0.25107690691947937, "learning_rate": 4.570792072976958e-05, "loss": 0.0353, "step": 55100 }, { "grad_norm": 0.34939754009246826, "learning_rate": 4.569144729948138e-05, "loss": 0.0365, "step": 55110 }, { "grad_norm": 0.33776265382766724, "learning_rate": 4.5674974340370033e-05, "loss": 0.0264, "step": 55120 }, { "grad_norm": 0.41116011142730713, "learning_rate": 4.565850185423699e-05, "loss": 0.037, "step": 55130 }, { "grad_norm": 0.4974774718284607, "learning_rate": 4.564202984288365e-05, "loss": 0.0477, "step": 55140 }, { "grad_norm": 0.3715623915195465, "learning_rate": 4.5625558308111356e-05, "loss": 0.0302, "step": 55150 }, { "grad_norm": 0.28703370690345764, "learning_rate": 4.560908725172145e-05, "loss": 0.0375, "step": 55160 }, { "grad_norm": 0.32920584082603455, "learning_rate": 4.559261667551516e-05, "loss": 0.027, "step": 55170 }, { "grad_norm": 0.37151721119880676, "learning_rate": 4.5576146581293685e-05, "loss": 0.0364, "step": 55180 }, { "grad_norm": 0.4789125919342041, "learning_rate": 4.5559676970858164e-05, "loss": 0.0342, "step": 55190 }, { "grad_norm": 0.6376321911811829, "learning_rate": 4.5543207846009705e-05, "loss": 0.0326, "step": 55200 }, { "grad_norm": 0.3022198975086212, "learning_rate": 4.552673920854933e-05, "loss": 0.0367, "step": 55210 }, { "grad_norm": 0.39916908740997314, "learning_rate": 4.5510271060278006e-05, "loss": 0.0299, "step": 55220 }, { "grad_norm": 0.3158646821975708, "learning_rate": 4.5493803402996704e-05, "loss": 0.0377, "step": 55230 }, { "grad_norm": 0.34226953983306885, "learning_rate": 4.5477336238506295e-05, "loss": 0.0314, "step": 55240 }, { "grad_norm": 0.30797079205513, "learning_rate": 4.546086956860759e-05, "loss": 0.0347, "step": 55250 }, { "grad_norm": 0.3036686182022095, "learning_rate": 4.5444403395101374e-05, "loss": 0.0248, "step": 55260 }, { "grad_norm": 0.3828297555446625, "learning_rate": 4.542793771978834e-05, "loss": 0.0274, "step": 55270 }, { "grad_norm": 0.44828203320503235, "learning_rate": 4.541147254446916e-05, "loss": 0.0367, "step": 55280 }, { "grad_norm": 0.3372291922569275, "learning_rate": 4.539500787094445e-05, "loss": 0.0329, "step": 55290 }, { "grad_norm": 0.4513230323791504, "learning_rate": 4.537854370101475e-05, "loss": 0.0357, "step": 55300 }, { "grad_norm": 0.30755913257598877, "learning_rate": 4.5362080036480566e-05, "loss": 0.026, "step": 55310 }, { "grad_norm": 0.2658262550830841, "learning_rate": 4.534561687914235e-05, "loss": 0.0267, "step": 55320 }, { "grad_norm": 0.42792460322380066, "learning_rate": 4.5329154230800466e-05, "loss": 0.0334, "step": 55330 }, { "grad_norm": 0.22256134450435638, "learning_rate": 4.531269209325525e-05, "loss": 0.0275, "step": 55340 }, { "grad_norm": 0.36083492636680603, "learning_rate": 4.529623046830696e-05, "loss": 0.0266, "step": 55350 }, { "grad_norm": 0.4407488703727722, "learning_rate": 4.5279769357755866e-05, "loss": 0.0373, "step": 55360 }, { "grad_norm": 0.432426393032074, "learning_rate": 4.5263308763402084e-05, "loss": 0.0287, "step": 55370 }, { "grad_norm": 0.827681303024292, "learning_rate": 4.524684868704574e-05, "loss": 0.0298, "step": 55380 }, { "grad_norm": 0.3279537260532379, "learning_rate": 4.523038913048688e-05, "loss": 0.0339, "step": 55390 }, { "grad_norm": 0.3148602843284607, "learning_rate": 4.521393009552549e-05, "loss": 0.0387, "step": 55400 }, { "grad_norm": 0.28537318110466003, "learning_rate": 4.5197471583961484e-05, "loss": 0.034, "step": 55410 }, { "grad_norm": 0.29376867413520813, "learning_rate": 4.51810135975948e-05, "loss": 0.0299, "step": 55420 }, { "grad_norm": 0.1720900982618332, "learning_rate": 4.516455613822521e-05, "loss": 0.03, "step": 55430 }, { "grad_norm": 0.2787381410598755, "learning_rate": 4.514809920765249e-05, "loss": 0.0335, "step": 55440 }, { "grad_norm": 0.43389445543289185, "learning_rate": 4.5131642807676356e-05, "loss": 0.0316, "step": 55450 }, { "grad_norm": 0.44747909903526306, "learning_rate": 4.511518694009644e-05, "loss": 0.0347, "step": 55460 }, { "grad_norm": 0.4102112948894501, "learning_rate": 4.509873160671231e-05, "loss": 0.0365, "step": 55470 }, { "grad_norm": 0.4436124861240387, "learning_rate": 4.508227680932356e-05, "loss": 0.0265, "step": 55480 }, { "grad_norm": 0.5771716833114624, "learning_rate": 4.5065822549729605e-05, "loss": 0.0366, "step": 55490 }, { "grad_norm": 0.3088866174221039, "learning_rate": 4.504936882972989e-05, "loss": 0.0276, "step": 55500 }, { "grad_norm": 0.39808058738708496, "learning_rate": 4.503291565112374e-05, "loss": 0.0241, "step": 55510 }, { "grad_norm": 0.3998807966709137, "learning_rate": 4.501646301571048e-05, "loss": 0.0298, "step": 55520 }, { "grad_norm": 0.3691508173942566, "learning_rate": 4.500001092528933e-05, "loss": 0.0353, "step": 55530 }, { "grad_norm": 0.23132532835006714, "learning_rate": 4.4983559381659455e-05, "loss": 0.031, "step": 55540 }, { "grad_norm": 0.402818500995636, "learning_rate": 4.496710838661999e-05, "loss": 0.0289, "step": 55550 }, { "grad_norm": 0.28729644417762756, "learning_rate": 4.495065794196999e-05, "loss": 0.0367, "step": 55560 }, { "grad_norm": 0.28582271933555603, "learning_rate": 4.4934208049508455e-05, "loss": 0.0294, "step": 55570 }, { "grad_norm": 0.3847346901893616, "learning_rate": 4.4917758711034303e-05, "loss": 0.0314, "step": 55580 }, { "grad_norm": 0.3677622079849243, "learning_rate": 4.490130992834642e-05, "loss": 0.0349, "step": 55590 }, { "grad_norm": 0.5459517240524292, "learning_rate": 4.488486170324359e-05, "loss": 0.0455, "step": 55600 }, { "grad_norm": 0.44477707147598267, "learning_rate": 4.486841403752463e-05, "loss": 0.0296, "step": 55610 }, { "grad_norm": 0.24832619726657867, "learning_rate": 4.4851966932988174e-05, "loss": 0.0252, "step": 55620 }, { "grad_norm": 0.40492919087409973, "learning_rate": 4.4835520391432894e-05, "loss": 0.0322, "step": 55630 }, { "grad_norm": 0.35649457573890686, "learning_rate": 4.481907441465733e-05, "loss": 0.0342, "step": 55640 }, { "grad_norm": 0.38460075855255127, "learning_rate": 4.480262900446e-05, "loss": 0.0343, "step": 55650 }, { "grad_norm": 0.3859862983226776, "learning_rate": 4.478618416263933e-05, "loss": 0.0337, "step": 55660 }, { "grad_norm": 0.33641716837882996, "learning_rate": 4.476973989099373e-05, "loss": 0.0331, "step": 55670 }, { "grad_norm": 0.3775642514228821, "learning_rate": 4.475329619132153e-05, "loss": 0.0451, "step": 55680 }, { "grad_norm": 0.43231436610221863, "learning_rate": 4.4736853065420956e-05, "loss": 0.045, "step": 55690 }, { "grad_norm": 0.2940789759159088, "learning_rate": 4.472041051509023e-05, "loss": 0.0336, "step": 55700 }, { "grad_norm": 0.29080143570899963, "learning_rate": 4.470396854212746e-05, "loss": 0.0368, "step": 55710 }, { "grad_norm": 0.6405948996543884, "learning_rate": 4.468752714833072e-05, "loss": 0.04, "step": 55720 }, { "grad_norm": 0.3355194330215454, "learning_rate": 4.467108633549803e-05, "loss": 0.0367, "step": 55730 }, { "grad_norm": 0.41790515184402466, "learning_rate": 4.4654646105427335e-05, "loss": 0.0505, "step": 55740 }, { "grad_norm": 0.3239718973636627, "learning_rate": 4.463820645991651e-05, "loss": 0.0295, "step": 55750 }, { "grad_norm": 0.4033745229244232, "learning_rate": 4.462176740076334e-05, "loss": 0.0289, "step": 55760 }, { "grad_norm": 0.4779888093471527, "learning_rate": 4.4605328929765624e-05, "loss": 0.0394, "step": 55770 }, { "grad_norm": 0.36054354906082153, "learning_rate": 4.458889104872102e-05, "loss": 0.0292, "step": 55780 }, { "grad_norm": 0.5488982200622559, "learning_rate": 4.457245375942715e-05, "loss": 0.0416, "step": 55790 }, { "grad_norm": 0.33097949624061584, "learning_rate": 4.455601706368157e-05, "loss": 0.0298, "step": 55800 }, { "grad_norm": 0.42628782987594604, "learning_rate": 4.453958096328179e-05, "loss": 0.0363, "step": 55810 }, { "grad_norm": 0.45131605863571167, "learning_rate": 4.452314546002523e-05, "loss": 0.0354, "step": 55820 }, { "grad_norm": 0.28529754281044006, "learning_rate": 4.4506710555709246e-05, "loss": 0.0357, "step": 55830 }, { "grad_norm": 0.4591175317764282, "learning_rate": 4.449027625213114e-05, "loss": 0.0293, "step": 55840 }, { "grad_norm": 0.2647915780544281, "learning_rate": 4.4473842551088135e-05, "loss": 0.0314, "step": 55850 }, { "grad_norm": 0.43916693329811096, "learning_rate": 4.445740945437738e-05, "loss": 0.0281, "step": 55860 }, { "grad_norm": 0.4744831621646881, "learning_rate": 4.444097696379602e-05, "loss": 0.0316, "step": 55870 }, { "grad_norm": 0.342800498008728, "learning_rate": 4.442454508114106e-05, "loss": 0.0311, "step": 55880 }, { "grad_norm": 0.34504663944244385, "learning_rate": 4.4408113808209465e-05, "loss": 0.035, "step": 55890 }, { "grad_norm": 0.31552034616470337, "learning_rate": 4.439168314679813e-05, "loss": 0.0297, "step": 55900 }, { "grad_norm": 0.30416518449783325, "learning_rate": 4.43752530987039e-05, "loss": 0.0293, "step": 55910 }, { "grad_norm": 0.6585611701011658, "learning_rate": 4.4358823665723515e-05, "loss": 0.0295, "step": 55920 }, { "grad_norm": 0.522229015827179, "learning_rate": 4.434239484965371e-05, "loss": 0.0316, "step": 55930 }, { "grad_norm": 0.30461615324020386, "learning_rate": 4.4325966652291103e-05, "loss": 0.031, "step": 55940 }, { "grad_norm": 0.34317049384117126, "learning_rate": 4.430953907543225e-05, "loss": 0.0252, "step": 55950 }, { "grad_norm": 0.30971214175224304, "learning_rate": 4.4293112120873645e-05, "loss": 0.0286, "step": 55960 }, { "grad_norm": 0.3698662519454956, "learning_rate": 4.427668579041172e-05, "loss": 0.027, "step": 55970 }, { "grad_norm": 0.33347681164741516, "learning_rate": 4.4260260085842816e-05, "loss": 0.0297, "step": 55980 }, { "grad_norm": 0.2878412902355194, "learning_rate": 4.4243835008963256e-05, "loss": 0.0279, "step": 55990 }, { "grad_norm": 0.39053359627723694, "learning_rate": 4.422741056156925e-05, "loss": 0.0264, "step": 56000 }, { "grad_norm": 0.24351496994495392, "learning_rate": 4.421098674545693e-05, "loss": 0.0236, "step": 56010 }, { "grad_norm": 0.30935001373291016, "learning_rate": 4.419456356242241e-05, "loss": 0.0331, "step": 56020 }, { "grad_norm": 0.4333782494068146, "learning_rate": 4.4178141014261685e-05, "loss": 0.0304, "step": 56030 }, { "grad_norm": 0.552070140838623, "learning_rate": 4.416171910277071e-05, "loss": 0.0348, "step": 56040 }, { "grad_norm": 0.37528377771377563, "learning_rate": 4.414529782974533e-05, "loss": 0.0288, "step": 56050 }, { "grad_norm": 0.2375643402338028, "learning_rate": 4.41288771969814e-05, "loss": 0.0274, "step": 56060 }, { "grad_norm": 0.47249260544776917, "learning_rate": 4.411245720627464e-05, "loss": 0.0291, "step": 56070 }, { "grad_norm": 0.7535814642906189, "learning_rate": 4.40960378594207e-05, "loss": 0.0355, "step": 56080 }, { "grad_norm": 0.2939859926700592, "learning_rate": 4.407961915821519e-05, "loss": 0.0322, "step": 56090 }, { "grad_norm": 0.2714914381504059, "learning_rate": 4.406320110445362e-05, "loss": 0.0289, "step": 56100 }, { "grad_norm": 0.3420696258544922, "learning_rate": 4.404678369993144e-05, "loss": 0.0279, "step": 56110 }, { "grad_norm": 0.3785597085952759, "learning_rate": 4.403036694644406e-05, "loss": 0.0345, "step": 56120 }, { "grad_norm": 0.4000625014305115, "learning_rate": 4.4013950845786764e-05, "loss": 0.0327, "step": 56130 }, { "grad_norm": 0.43181973695755005, "learning_rate": 4.399753539975482e-05, "loss": 0.0399, "step": 56140 }, { "grad_norm": 0.2668572962284088, "learning_rate": 4.398112061014337e-05, "loss": 0.0301, "step": 56150 }, { "grad_norm": 0.2831382751464844, "learning_rate": 4.396470647874753e-05, "loss": 0.0294, "step": 56160 }, { "grad_norm": 0.5134109854698181, "learning_rate": 4.394829300736229e-05, "loss": 0.0236, "step": 56170 }, { "grad_norm": 0.35937759280204773, "learning_rate": 4.393188019778265e-05, "loss": 0.0313, "step": 56180 }, { "grad_norm": 0.2627657353878021, "learning_rate": 4.391546805180347e-05, "loss": 0.0329, "step": 56190 }, { "grad_norm": 0.5065317749977112, "learning_rate": 4.389905657121955e-05, "loss": 0.0285, "step": 56200 }, { "grad_norm": 0.48735445737838745, "learning_rate": 4.3882645757825635e-05, "loss": 0.0321, "step": 56210 }, { "grad_norm": 0.2725120484828949, "learning_rate": 4.386623561341637e-05, "loss": 0.0266, "step": 56220 }, { "grad_norm": 0.3061589002609253, "learning_rate": 4.384982613978637e-05, "loss": 0.0349, "step": 56230 }, { "grad_norm": 0.4500388205051422, "learning_rate": 4.383341733873012e-05, "loss": 0.0278, "step": 56240 }, { "grad_norm": 0.3841620087623596, "learning_rate": 4.381700921204209e-05, "loss": 0.0359, "step": 56250 }, { "grad_norm": 0.4225820302963257, "learning_rate": 4.380060176151663e-05, "loss": 0.0297, "step": 56260 }, { "grad_norm": 0.419435054063797, "learning_rate": 4.378419498894805e-05, "loss": 0.0354, "step": 56270 }, { "grad_norm": 0.39168041944503784, "learning_rate": 4.376778889613056e-05, "loss": 0.0323, "step": 56280 }, { "grad_norm": 1.3187710046768188, "learning_rate": 4.375138348485831e-05, "loss": 0.0425, "step": 56290 }, { "grad_norm": 0.3771163821220398, "learning_rate": 4.3734978756925346e-05, "loss": 0.0264, "step": 56300 }, { "grad_norm": 0.3768254220485687, "learning_rate": 4.3718574714125706e-05, "loss": 0.0333, "step": 56310 }, { "grad_norm": 0.22572742402553558, "learning_rate": 4.370217135825329e-05, "loss": 0.0306, "step": 56320 }, { "grad_norm": 0.3700610101222992, "learning_rate": 4.368576869110194e-05, "loss": 0.0346, "step": 56330 }, { "grad_norm": 0.3603751063346863, "learning_rate": 4.366936671446544e-05, "loss": 0.0291, "step": 56340 }, { "grad_norm": 0.3030642867088318, "learning_rate": 4.3652965430137474e-05, "loss": 0.0361, "step": 56350 }, { "grad_norm": 0.21972133219242096, "learning_rate": 4.3636564839911646e-05, "loss": 0.0248, "step": 56360 }, { "grad_norm": 0.33683085441589355, "learning_rate": 4.3620164945581545e-05, "loss": 0.0303, "step": 56370 }, { "grad_norm": 0.46225684881210327, "learning_rate": 4.360376574894061e-05, "loss": 0.0361, "step": 56380 }, { "grad_norm": 0.264032781124115, "learning_rate": 4.358736725178224e-05, "loss": 0.0306, "step": 56390 }, { "grad_norm": 0.5028454065322876, "learning_rate": 4.357096945589974e-05, "loss": 0.0393, "step": 56400 }, { "grad_norm": 0.2768942713737488, "learning_rate": 4.355457236308636e-05, "loss": 0.0271, "step": 56410 }, { "grad_norm": 0.5080465078353882, "learning_rate": 4.353817597513526e-05, "loss": 0.033, "step": 56420 }, { "grad_norm": 0.40340694785118103, "learning_rate": 4.352178029383948e-05, "loss": 0.0356, "step": 56430 }, { "grad_norm": 0.39677920937538147, "learning_rate": 4.3505385320992105e-05, "loss": 0.0364, "step": 56440 }, { "grad_norm": 0.2445448935031891, "learning_rate": 4.348899105838602e-05, "loss": 0.0317, "step": 56450 }, { "grad_norm": 0.5830756425857544, "learning_rate": 4.3472597507814087e-05, "loss": 0.0312, "step": 56460 }, { "grad_norm": 0.33212336897850037, "learning_rate": 4.3456204671069066e-05, "loss": 0.0362, "step": 56470 }, { "grad_norm": 0.30354005098342896, "learning_rate": 4.343981254994367e-05, "loss": 0.025, "step": 56480 }, { "grad_norm": 0.3830585777759552, "learning_rate": 4.34234211462305e-05, "loss": 0.0355, "step": 56490 }, { "grad_norm": 0.3270963132381439, "learning_rate": 4.340703046172213e-05, "loss": 0.0285, "step": 56500 }, { "grad_norm": 0.27687302231788635, "learning_rate": 4.339064049821097e-05, "loss": 0.0362, "step": 56510 }, { "grad_norm": 0.6357318162918091, "learning_rate": 4.3374251257489446e-05, "loss": 0.0378, "step": 56520 }, { "grad_norm": 0.34811389446258545, "learning_rate": 4.335786274134984e-05, "loss": 0.0376, "step": 56530 }, { "grad_norm": 0.48899218440055847, "learning_rate": 4.334147495158438e-05, "loss": 0.0381, "step": 56540 }, { "grad_norm": 0.5068751573562622, "learning_rate": 4.33250878899852e-05, "loss": 0.0344, "step": 56550 }, { "grad_norm": 0.3769766390323639, "learning_rate": 4.330870155834437e-05, "loss": 0.0348, "step": 56560 }, { "grad_norm": 0.3099982738494873, "learning_rate": 4.3292315958453884e-05, "loss": 0.0301, "step": 56570 }, { "grad_norm": 0.29933902621269226, "learning_rate": 4.3275931092105646e-05, "loss": 0.0333, "step": 56580 }, { "grad_norm": 0.21906141936779022, "learning_rate": 4.325954696109147e-05, "loss": 0.0315, "step": 56590 }, { "grad_norm": 0.3917519450187683, "learning_rate": 4.32431635672031e-05, "loss": 0.0383, "step": 56600 }, { "grad_norm": 0.5268758535385132, "learning_rate": 4.3226780912232214e-05, "loss": 0.0348, "step": 56610 }, { "grad_norm": 0.37737053632736206, "learning_rate": 4.321039899797036e-05, "loss": 0.0296, "step": 56620 }, { "grad_norm": 0.27188223600387573, "learning_rate": 4.319401782620908e-05, "loss": 0.0267, "step": 56630 }, { "grad_norm": 0.2646706700325012, "learning_rate": 4.317763739873978e-05, "loss": 0.0295, "step": 56640 }, { "grad_norm": 0.461427241563797, "learning_rate": 4.31612577173538e-05, "loss": 0.0311, "step": 56650 }, { "grad_norm": 0.39448973536491394, "learning_rate": 4.314487878384239e-05, "loss": 0.038, "step": 56660 }, { "grad_norm": 0.5405974984169006, "learning_rate": 4.3128500599996736e-05, "loss": 0.0288, "step": 56670 }, { "grad_norm": 0.33933568000793457, "learning_rate": 4.31121231676079e-05, "loss": 0.0346, "step": 56680 }, { "grad_norm": 0.474416047334671, "learning_rate": 4.309574648846694e-05, "loss": 0.0334, "step": 56690 }, { "grad_norm": 0.3979555666446686, "learning_rate": 4.3079370564364755e-05, "loss": 0.0364, "step": 56700 }, { "grad_norm": 0.31400543451309204, "learning_rate": 4.306299539709221e-05, "loss": 0.0293, "step": 56710 }, { "grad_norm": 0.2772292494773865, "learning_rate": 4.3046620988440046e-05, "loss": 0.0378, "step": 56720 }, { "grad_norm": 0.41634103655815125, "learning_rate": 4.303024734019897e-05, "loss": 0.0266, "step": 56730 }, { "grad_norm": 0.27141472697257996, "learning_rate": 4.3013874454159565e-05, "loss": 0.0335, "step": 56740 }, { "grad_norm": 0.7743204832077026, "learning_rate": 4.299750233211233e-05, "loss": 0.0445, "step": 56750 }, { "grad_norm": 0.38908761739730835, "learning_rate": 4.2981130975847715e-05, "loss": 0.0298, "step": 56760 }, { "grad_norm": 0.43109047412872314, "learning_rate": 4.2964760387156076e-05, "loss": 0.0364, "step": 56770 }, { "grad_norm": 0.3401326537132263, "learning_rate": 4.294839056782767e-05, "loss": 0.0288, "step": 56780 }, { "grad_norm": 0.403852641582489, "learning_rate": 4.293202151965266e-05, "loss": 0.0378, "step": 56790 }, { "grad_norm": 0.3283177316188812, "learning_rate": 4.291565324442115e-05, "loss": 0.0279, "step": 56800 }, { "grad_norm": 0.2814786434173584, "learning_rate": 4.2899285743923135e-05, "loss": 0.0342, "step": 56810 }, { "grad_norm": 0.25375279784202576, "learning_rate": 4.288291901994857e-05, "loss": 0.0276, "step": 56820 }, { "grad_norm": 0.40368273854255676, "learning_rate": 4.286655307428729e-05, "loss": 0.0318, "step": 56830 }, { "grad_norm": 0.3471808433532715, "learning_rate": 4.285018790872903e-05, "loss": 0.0395, "step": 56840 }, { "grad_norm": 0.37387433648109436, "learning_rate": 4.283382352506347e-05, "loss": 0.0425, "step": 56850 }, { "grad_norm": 0.322929322719574, "learning_rate": 4.2817459925080205e-05, "loss": 0.0497, "step": 56860 }, { "grad_norm": 0.40435653924942017, "learning_rate": 4.28010971105687e-05, "loss": 0.0369, "step": 56870 }, { "grad_norm": 0.4271155893802643, "learning_rate": 4.278473508331841e-05, "loss": 0.0303, "step": 56880 }, { "grad_norm": 0.40893369913101196, "learning_rate": 4.276837384511864e-05, "loss": 0.0398, "step": 56890 }, { "grad_norm": 0.49924659729003906, "learning_rate": 4.275201339775864e-05, "loss": 0.0324, "step": 56900 }, { "grad_norm": 0.3262219727039337, "learning_rate": 4.273565374302756e-05, "loss": 0.0415, "step": 56910 }, { "grad_norm": 0.3791680634021759, "learning_rate": 4.2719294882714454e-05, "loss": 0.0273, "step": 56920 }, { "grad_norm": 0.20481650531291962, "learning_rate": 4.270293681860831e-05, "loss": 0.032, "step": 56930 }, { "grad_norm": 0.37354931235313416, "learning_rate": 4.268657955249802e-05, "loss": 0.0328, "step": 56940 }, { "grad_norm": 0.3299926221370697, "learning_rate": 4.267022308617241e-05, "loss": 0.0256, "step": 56950 }, { "grad_norm": 0.35877087712287903, "learning_rate": 4.265386742142018e-05, "loss": 0.034, "step": 56960 }, { "grad_norm": 0.37269923090934753, "learning_rate": 4.263751256002996e-05, "loss": 0.033, "step": 56970 }, { "grad_norm": 0.5288760662078857, "learning_rate": 4.26211585037903e-05, "loss": 0.0351, "step": 56980 }, { "grad_norm": 0.39471372961997986, "learning_rate": 4.2604805254489656e-05, "loss": 0.035, "step": 56990 }, { "grad_norm": 0.538078248500824, "learning_rate": 4.258845281391638e-05, "loss": 0.0321, "step": 57000 }, { "grad_norm": 0.22026561200618744, "learning_rate": 4.257210118385876e-05, "loss": 0.0332, "step": 57010 }, { "grad_norm": 0.41516461968421936, "learning_rate": 4.2555750366104994e-05, "loss": 0.0375, "step": 57020 }, { "grad_norm": 0.35429537296295166, "learning_rate": 4.2539400362443184e-05, "loss": 0.0334, "step": 57030 }, { "grad_norm": 0.549990713596344, "learning_rate": 4.252305117466133e-05, "loss": 0.0395, "step": 57040 }, { "grad_norm": 0.4360381066799164, "learning_rate": 4.2506702804547356e-05, "loss": 0.0312, "step": 57050 }, { "grad_norm": 0.3169628381729126, "learning_rate": 4.249035525388907e-05, "loss": 0.0297, "step": 57060 }, { "grad_norm": 0.38572415709495544, "learning_rate": 4.247400852447427e-05, "loss": 0.031, "step": 57070 }, { "grad_norm": 0.27510398626327515, "learning_rate": 4.245766261809059e-05, "loss": 0.0285, "step": 57080 }, { "grad_norm": 0.34481820464134216, "learning_rate": 4.244131753652558e-05, "loss": 0.0288, "step": 57090 }, { "grad_norm": 0.29412394762039185, "learning_rate": 4.242497328156672e-05, "loss": 0.0289, "step": 57100 }, { "grad_norm": 0.3983314335346222, "learning_rate": 4.240862985500139e-05, "loss": 0.0453, "step": 57110 }, { "grad_norm": 0.25414055585861206, "learning_rate": 4.2392287258616894e-05, "loss": 0.0253, "step": 57120 }, { "grad_norm": 0.23193109035491943, "learning_rate": 4.2375945494200396e-05, "loss": 0.0277, "step": 57130 }, { "grad_norm": 0.3101981580257416, "learning_rate": 4.2359604563539054e-05, "loss": 0.0378, "step": 57140 }, { "grad_norm": 0.2548518180847168, "learning_rate": 4.2343264468419885e-05, "loss": 0.0336, "step": 57150 }, { "grad_norm": 0.3874635398387909, "learning_rate": 4.232692521062979e-05, "loss": 0.0225, "step": 57160 }, { "grad_norm": 0.4844112992286682, "learning_rate": 4.231058679195562e-05, "loss": 0.0361, "step": 57170 }, { "grad_norm": 0.284244567155838, "learning_rate": 4.229424921418411e-05, "loss": 0.0355, "step": 57180 }, { "grad_norm": 0.579921543598175, "learning_rate": 4.2277912479101916e-05, "loss": 0.0366, "step": 57190 }, { "grad_norm": 0.620278000831604, "learning_rate": 4.226157658849562e-05, "loss": 0.035, "step": 57200 }, { "grad_norm": 0.29786089062690735, "learning_rate": 4.2245241544151664e-05, "loss": 0.0315, "step": 57210 }, { "grad_norm": 0.4203052520751953, "learning_rate": 4.2228907347856425e-05, "loss": 0.0452, "step": 57220 }, { "grad_norm": 0.30291175842285156, "learning_rate": 4.2212574001396214e-05, "loss": 0.0377, "step": 57230 }, { "grad_norm": 0.43002405762672424, "learning_rate": 4.21962415065572e-05, "loss": 0.0318, "step": 57240 }, { "grad_norm": 0.3122059404850006, "learning_rate": 4.217990986512548e-05, "loss": 0.0327, "step": 57250 }, { "grad_norm": 0.29308488965034485, "learning_rate": 4.216357907888704e-05, "loss": 0.0331, "step": 57260 }, { "grad_norm": 0.5589092969894409, "learning_rate": 4.2147249149627824e-05, "loss": 0.035, "step": 57270 }, { "grad_norm": 0.3765319585800171, "learning_rate": 4.213092007913364e-05, "loss": 0.0339, "step": 57280 }, { "grad_norm": 0.19299285113811493, "learning_rate": 4.2114591869190205e-05, "loss": 0.0368, "step": 57290 }, { "grad_norm": 0.2034359574317932, "learning_rate": 4.209826452158315e-05, "loss": 0.0363, "step": 57300 }, { "grad_norm": 0.4328862428665161, "learning_rate": 4.2081938038098e-05, "loss": 0.0322, "step": 57310 }, { "grad_norm": 0.45752084255218506, "learning_rate": 4.2065612420520185e-05, "loss": 0.0298, "step": 57320 }, { "grad_norm": 0.3646782338619232, "learning_rate": 4.204928767063509e-05, "loss": 0.0328, "step": 57330 }, { "grad_norm": 0.269787073135376, "learning_rate": 4.2032963790227944e-05, "loss": 0.0355, "step": 57340 }, { "grad_norm": 0.30688783526420593, "learning_rate": 4.2016640781083887e-05, "loss": 0.0313, "step": 57350 }, { "grad_norm": 0.37146511673927307, "learning_rate": 4.2000318644988004e-05, "loss": 0.0477, "step": 57360 }, { "grad_norm": 0.3569018840789795, "learning_rate": 4.198399738372524e-05, "loss": 0.0301, "step": 57370 }, { "grad_norm": 0.23956310749053955, "learning_rate": 4.196767699908045e-05, "loss": 0.031, "step": 57380 }, { "grad_norm": 0.3533906638622284, "learning_rate": 4.195135749283845e-05, "loss": 0.0335, "step": 57390 }, { "grad_norm": 0.42833054065704346, "learning_rate": 4.1935038866783894e-05, "loss": 0.0319, "step": 57400 }, { "grad_norm": 0.20493921637535095, "learning_rate": 4.1918721122701355e-05, "loss": 0.0258, "step": 57410 }, { "grad_norm": 0.5138921737670898, "learning_rate": 4.190240426237533e-05, "loss": 0.0386, "step": 57420 }, { "grad_norm": 0.23596902191638947, "learning_rate": 4.188608828759019e-05, "loss": 0.0287, "step": 57430 }, { "grad_norm": 0.34883737564086914, "learning_rate": 4.1869773200130246e-05, "loss": 0.0529, "step": 57440 }, { "grad_norm": 0.42009830474853516, "learning_rate": 4.185345900177967e-05, "loss": 0.0394, "step": 57450 }, { "grad_norm": 0.29359981417655945, "learning_rate": 4.183714569432258e-05, "loss": 0.0331, "step": 57460 }, { "grad_norm": 0.24103914201259613, "learning_rate": 4.182083327954295e-05, "loss": 0.0294, "step": 57470 }, { "grad_norm": 0.3828243315219879, "learning_rate": 4.1804521759224714e-05, "loss": 0.0321, "step": 57480 }, { "grad_norm": 0.34722045063972473, "learning_rate": 4.178821113515165e-05, "loss": 0.0375, "step": 57490 }, { "grad_norm": 0.3426055312156677, "learning_rate": 4.1771901409107476e-05, "loss": 0.0422, "step": 57500 }, { "grad_norm": 0.3710223138332367, "learning_rate": 4.175559258287577e-05, "loss": 0.0304, "step": 57510 }, { "grad_norm": 0.4773959815502167, "learning_rate": 4.173928465824008e-05, "loss": 0.0297, "step": 57520 }, { "grad_norm": 0.32162678241729736, "learning_rate": 4.17229776369838e-05, "loss": 0.0341, "step": 57530 }, { "grad_norm": 0.2807829976081848, "learning_rate": 4.170667152089025e-05, "loss": 0.0351, "step": 57540 }, { "grad_norm": 0.23046207427978516, "learning_rate": 4.1690366311742626e-05, "loss": 0.0219, "step": 57550 }, { "grad_norm": 0.33458176255226135, "learning_rate": 4.1674062011324055e-05, "loss": 0.0243, "step": 57560 }, { "grad_norm": 0.36801597476005554, "learning_rate": 4.165775862141752e-05, "loss": 0.0242, "step": 57570 }, { "grad_norm": 0.40476998686790466, "learning_rate": 4.164145614380598e-05, "loss": 0.0272, "step": 57580 }, { "grad_norm": 0.2778981626033783, "learning_rate": 4.162515458027223e-05, "loss": 0.0279, "step": 57590 }, { "grad_norm": 0.45651975274086, "learning_rate": 4.160885393259899e-05, "loss": 0.0309, "step": 57600 }, { "grad_norm": 0.46765634417533875, "learning_rate": 4.159255420256886e-05, "loss": 0.0383, "step": 57610 }, { "grad_norm": 0.3727419078350067, "learning_rate": 4.157625539196436e-05, "loss": 0.0307, "step": 57620 }, { "grad_norm": 0.3246033489704132, "learning_rate": 4.155995750256791e-05, "loss": 0.0393, "step": 57630 }, { "grad_norm": 0.40543806552886963, "learning_rate": 4.15436605361618e-05, "loss": 0.0298, "step": 57640 }, { "grad_norm": 0.3643743693828583, "learning_rate": 4.152736449452827e-05, "loss": 0.0419, "step": 57650 }, { "grad_norm": 0.291062593460083, "learning_rate": 4.151106937944943e-05, "loss": 0.0256, "step": 57660 }, { "grad_norm": 0.33566924929618835, "learning_rate": 4.149477519270727e-05, "loss": 0.0271, "step": 57670 }, { "grad_norm": 0.4240853488445282, "learning_rate": 4.147848193608371e-05, "loss": 0.039, "step": 57680 }, { "grad_norm": 0.8959119319915771, "learning_rate": 4.146218961136056e-05, "loss": 0.0323, "step": 57690 }, { "grad_norm": 0.44349610805511475, "learning_rate": 4.14458982203195e-05, "loss": 0.0405, "step": 57700 }, { "grad_norm": 0.24457360804080963, "learning_rate": 4.1429607764742165e-05, "loss": 0.0341, "step": 57710 }, { "grad_norm": 0.3811744749546051, "learning_rate": 4.141331824641003e-05, "loss": 0.0358, "step": 57720 }, { "grad_norm": 0.3095681369304657, "learning_rate": 4.139702966710452e-05, "loss": 0.0368, "step": 57730 }, { "grad_norm": 0.31055647134780884, "learning_rate": 4.13807420286069e-05, "loss": 0.0362, "step": 57740 }, { "grad_norm": 0.26602643728256226, "learning_rate": 4.136445533269837e-05, "loss": 0.0338, "step": 57750 }, { "grad_norm": 0.28262242674827576, "learning_rate": 4.134816958116003e-05, "loss": 0.0277, "step": 57760 }, { "grad_norm": 0.4434252977371216, "learning_rate": 4.133188477577283e-05, "loss": 0.0385, "step": 57770 }, { "grad_norm": 0.42905256152153015, "learning_rate": 4.131560091831771e-05, "loss": 0.028, "step": 57780 }, { "grad_norm": 0.5781198740005493, "learning_rate": 4.1299318010575405e-05, "loss": 0.0261, "step": 57790 }, { "grad_norm": 0.30639198422431946, "learning_rate": 4.128303605432661e-05, "loss": 0.0498, "step": 57800 }, { "grad_norm": 0.3839721381664276, "learning_rate": 4.126675505135188e-05, "loss": 0.0307, "step": 57810 }, { "grad_norm": 0.42406800389289856, "learning_rate": 4.125047500343169e-05, "loss": 0.0302, "step": 57820 }, { "grad_norm": 0.37889963388442993, "learning_rate": 4.1234195912346376e-05, "loss": 0.0316, "step": 57830 }, { "grad_norm": 0.7339319586753845, "learning_rate": 4.1217917779876235e-05, "loss": 0.0354, "step": 57840 }, { "grad_norm": 0.4011576771736145, "learning_rate": 4.120164060780141e-05, "loss": 0.0409, "step": 57850 }, { "grad_norm": 0.31458547711372375, "learning_rate": 4.1185364397901935e-05, "loss": 0.0383, "step": 57860 }, { "grad_norm": 0.30910542607307434, "learning_rate": 4.116908915195776e-05, "loss": 0.0286, "step": 57870 }, { "grad_norm": 0.44717055559158325, "learning_rate": 4.115281487174872e-05, "loss": 0.0307, "step": 57880 }, { "grad_norm": 0.6234701871871948, "learning_rate": 4.113654155905452e-05, "loss": 0.0345, "step": 57890 }, { "grad_norm": 0.3176528513431549, "learning_rate": 4.112026921565483e-05, "loss": 0.0305, "step": 57900 }, { "grad_norm": 0.41189339756965637, "learning_rate": 4.1103997843329147e-05, "loss": 0.0321, "step": 57910 }, { "grad_norm": 1.66728937625885, "learning_rate": 4.108772744385689e-05, "loss": 0.0472, "step": 57920 }, { "grad_norm": 0.28455498814582825, "learning_rate": 4.1071458019017346e-05, "loss": 0.03, "step": 57930 }, { "grad_norm": 0.34841111302375793, "learning_rate": 4.1055189570589734e-05, "loss": 0.0319, "step": 57940 }, { "grad_norm": 0.3343322277069092, "learning_rate": 4.103892210035315e-05, "loss": 0.0264, "step": 57950 }, { "grad_norm": 0.5293061137199402, "learning_rate": 4.102265561008656e-05, "loss": 0.0308, "step": 57960 }, { "grad_norm": 0.37575268745422363, "learning_rate": 4.100639010156885e-05, "loss": 0.0293, "step": 57970 }, { "grad_norm": 0.3640914559364319, "learning_rate": 4.099012557657882e-05, "loss": 0.0338, "step": 57980 }, { "grad_norm": 0.6058112382888794, "learning_rate": 4.09738620368951e-05, "loss": 0.0359, "step": 57990 }, { "grad_norm": 0.28318125009536743, "learning_rate": 4.095759948429626e-05, "loss": 0.0268, "step": 58000 }, { "grad_norm": 0.35305240750312805, "learning_rate": 4.094133792056075e-05, "loss": 0.0398, "step": 58010 }, { "grad_norm": 0.30276423692703247, "learning_rate": 4.092507734746688e-05, "loss": 0.0346, "step": 58020 }, { "grad_norm": 0.48116469383239746, "learning_rate": 4.090881776679293e-05, "loss": 0.0297, "step": 58030 }, { "grad_norm": 0.4064449369907379, "learning_rate": 4.0892559180316994e-05, "loss": 0.0336, "step": 58040 }, { "grad_norm": 0.4958784878253937, "learning_rate": 4.08763015898171e-05, "loss": 0.0293, "step": 58050 }, { "grad_norm": 0.3092362582683563, "learning_rate": 4.0860044997071145e-05, "loss": 0.0364, "step": 58060 }, { "grad_norm": 0.2728064954280853, "learning_rate": 4.084378940385693e-05, "loss": 0.0299, "step": 58070 }, { "grad_norm": 0.4309355914592743, "learning_rate": 4.082753481195211e-05, "loss": 0.0354, "step": 58080 }, { "grad_norm": 0.37630629539489746, "learning_rate": 4.0811281223134324e-05, "loss": 0.03, "step": 58090 }, { "grad_norm": 0.33100125193595886, "learning_rate": 4.0795028639181e-05, "loss": 0.039, "step": 58100 }, { "grad_norm": 0.3599523901939392, "learning_rate": 4.077877706186951e-05, "loss": 0.027, "step": 58110 }, { "grad_norm": 0.3254355788230896, "learning_rate": 4.076252649297709e-05, "loss": 0.0329, "step": 58120 }, { "grad_norm": 0.394921213388443, "learning_rate": 4.07462769342809e-05, "loss": 0.0327, "step": 58130 }, { "grad_norm": 0.4230670928955078, "learning_rate": 4.073002838755793e-05, "loss": 0.0319, "step": 58140 }, { "grad_norm": 0.4039466977119446, "learning_rate": 4.0713780854585115e-05, "loss": 0.027, "step": 58150 }, { "grad_norm": 0.41057831048965454, "learning_rate": 4.069753433713927e-05, "loss": 0.0377, "step": 58160 }, { "grad_norm": 0.454559862613678, "learning_rate": 4.068128883699709e-05, "loss": 0.0339, "step": 58170 }, { "grad_norm": 0.4594258964061737, "learning_rate": 4.0665044355935155e-05, "loss": 0.0337, "step": 58180 }, { "grad_norm": 0.2952105402946472, "learning_rate": 4.064880089572993e-05, "loss": 0.0287, "step": 58190 }, { "grad_norm": 0.33755266666412354, "learning_rate": 4.063255845815779e-05, "loss": 0.0381, "step": 58200 }, { "grad_norm": 0.2992144525051117, "learning_rate": 4.0616317044994945e-05, "loss": 0.0267, "step": 58210 }, { "grad_norm": 0.39545807242393494, "learning_rate": 4.0600076658017585e-05, "loss": 0.027, "step": 58220 }, { "grad_norm": 0.23792922496795654, "learning_rate": 4.0583837299001706e-05, "loss": 0.0261, "step": 58230 }, { "grad_norm": 0.1974550485610962, "learning_rate": 4.056759896972322e-05, "loss": 0.0261, "step": 58240 }, { "grad_norm": 0.2526409327983856, "learning_rate": 4.055136167195793e-05, "loss": 0.0328, "step": 58250 }, { "grad_norm": 0.28605616092681885, "learning_rate": 4.053512540748152e-05, "loss": 0.0295, "step": 58260 }, { "grad_norm": 0.5363861918449402, "learning_rate": 4.051889017806955e-05, "loss": 0.0389, "step": 58270 }, { "grad_norm": 0.568392813205719, "learning_rate": 4.0502655985497516e-05, "loss": 0.0299, "step": 58280 }, { "grad_norm": 0.3116656541824341, "learning_rate": 4.048642283154074e-05, "loss": 0.037, "step": 58290 }, { "grad_norm": 0.21177440881729126, "learning_rate": 4.0470190717974464e-05, "loss": 0.0403, "step": 58300 }, { "grad_norm": 0.3817324638366699, "learning_rate": 4.045395964657381e-05, "loss": 0.0281, "step": 58310 }, { "grad_norm": 0.40175381302833557, "learning_rate": 4.043772961911376e-05, "loss": 0.0347, "step": 58320 }, { "grad_norm": 0.36917585134506226, "learning_rate": 4.042150063736924e-05, "loss": 0.029, "step": 58330 }, { "grad_norm": 0.4879227578639984, "learning_rate": 4.0405272703114976e-05, "loss": 0.0296, "step": 58340 }, { "grad_norm": 0.39647358655929565, "learning_rate": 4.038904581812569e-05, "loss": 0.0326, "step": 58350 }, { "grad_norm": 0.2843039631843567, "learning_rate": 4.03728199841759e-05, "loss": 0.0247, "step": 58360 }, { "grad_norm": 0.31259992718696594, "learning_rate": 4.0356595203040044e-05, "loss": 0.033, "step": 58370 }, { "grad_norm": 0.3351058065891266, "learning_rate": 4.0340371476492426e-05, "loss": 0.0361, "step": 58380 }, { "grad_norm": 0.33973315358161926, "learning_rate": 4.032414880630727e-05, "loss": 0.0377, "step": 58390 }, { "grad_norm": 0.3761237561702728, "learning_rate": 4.0307927194258646e-05, "loss": 0.0295, "step": 58400 }, { "grad_norm": 0.24407218396663666, "learning_rate": 4.029170664212054e-05, "loss": 0.0288, "step": 58410 }, { "grad_norm": 0.32747700810432434, "learning_rate": 4.0275487151666794e-05, "loss": 0.0309, "step": 58420 }, { "grad_norm": 0.3830934464931488, "learning_rate": 4.025926872467116e-05, "loss": 0.0332, "step": 58430 }, { "grad_norm": 0.44083371758461, "learning_rate": 4.024305136290725e-05, "loss": 0.0303, "step": 58440 }, { "grad_norm": 0.32824191451072693, "learning_rate": 4.022683506814858e-05, "loss": 0.0336, "step": 58450 }, { "grad_norm": 0.3419942259788513, "learning_rate": 4.021061984216853e-05, "loss": 0.0377, "step": 58460 }, { "grad_norm": 0.24705404043197632, "learning_rate": 4.019440568674035e-05, "loss": 0.0259, "step": 58470 }, { "grad_norm": 0.9735421538352966, "learning_rate": 4.017819260363724e-05, "loss": 0.0312, "step": 58480 }, { "grad_norm": 0.34318333864212036, "learning_rate": 4.016198059463223e-05, "loss": 0.0424, "step": 58490 }, { "grad_norm": 0.2076016217470169, "learning_rate": 4.0145769661498233e-05, "loss": 0.0322, "step": 58500 }, { "grad_norm": 1.8478810787200928, "learning_rate": 4.012955980600804e-05, "loss": 0.0395, "step": 58510 }, { "grad_norm": 0.418906569480896, "learning_rate": 4.0113351029934355e-05, "loss": 0.0352, "step": 58520 }, { "grad_norm": 0.35594668984413147, "learning_rate": 4.0097143335049714e-05, "loss": 0.0326, "step": 58530 }, { "grad_norm": 0.3054889738559723, "learning_rate": 4.0080936723126606e-05, "loss": 0.0265, "step": 58540 }, { "grad_norm": 0.3115600645542145, "learning_rate": 4.006473119593735e-05, "loss": 0.0308, "step": 58550 }, { "grad_norm": 0.3777977228164673, "learning_rate": 4.004852675525415e-05, "loss": 0.0362, "step": 58560 }, { "grad_norm": 0.3241381347179413, "learning_rate": 4.00323234028491e-05, "loss": 0.0278, "step": 58570 }, { "grad_norm": 0.37122848629951477, "learning_rate": 4.0016121140494186e-05, "loss": 0.0351, "step": 58580 }, { "grad_norm": 0.30420494079589844, "learning_rate": 3.999991996996122e-05, "loss": 0.0298, "step": 58590 }, { "grad_norm": 0.3171132206916809, "learning_rate": 3.9983719893022e-05, "loss": 0.0265, "step": 58600 }, { "grad_norm": 0.5156972408294678, "learning_rate": 3.996752091144812e-05, "loss": 0.0317, "step": 58610 }, { "grad_norm": 0.3898029923439026, "learning_rate": 3.9951323027011055e-05, "loss": 0.0309, "step": 58620 }, { "grad_norm": 0.2902905344963074, "learning_rate": 3.993512624148219e-05, "loss": 0.0319, "step": 58630 }, { "grad_norm": 0.4589807093143463, "learning_rate": 3.99189305566328e-05, "loss": 0.0349, "step": 58640 }, { "grad_norm": 0.20999817550182343, "learning_rate": 3.990273597423401e-05, "loss": 0.0355, "step": 58650 }, { "grad_norm": 0.23950380086898804, "learning_rate": 3.988654249605681e-05, "loss": 0.0329, "step": 58660 }, { "grad_norm": 0.404308944940567, "learning_rate": 3.987035012387213e-05, "loss": 0.0379, "step": 58670 }, { "grad_norm": 0.4200836718082428, "learning_rate": 3.9854158859450735e-05, "loss": 0.0262, "step": 58680 }, { "grad_norm": 0.323851078748703, "learning_rate": 3.983796870456328e-05, "loss": 0.0335, "step": 58690 }, { "grad_norm": 0.13313573598861694, "learning_rate": 3.982177966098028e-05, "loss": 0.0231, "step": 58700 }, { "grad_norm": 0.338047593832016, "learning_rate": 3.980559173047216e-05, "loss": 0.0443, "step": 58710 }, { "grad_norm": 0.372195839881897, "learning_rate": 3.9789404914809176e-05, "loss": 0.0355, "step": 58720 }, { "grad_norm": 0.267127126455307, "learning_rate": 3.9773219215761544e-05, "loss": 0.03, "step": 58730 }, { "grad_norm": 0.26876670122146606, "learning_rate": 3.975703463509928e-05, "loss": 0.0311, "step": 58740 }, { "grad_norm": 0.5436235666275024, "learning_rate": 3.974085117459231e-05, "loss": 0.0314, "step": 58750 }, { "grad_norm": 0.4666292369365692, "learning_rate": 3.972466883601043e-05, "loss": 0.0371, "step": 58760 }, { "grad_norm": 0.20536258816719055, "learning_rate": 3.9708487621123316e-05, "loss": 0.0229, "step": 58770 }, { "grad_norm": 0.30505189299583435, "learning_rate": 3.969230753170051e-05, "loss": 0.0378, "step": 58780 }, { "grad_norm": 0.36261212825775146, "learning_rate": 3.967612856951146e-05, "loss": 0.028, "step": 58790 }, { "grad_norm": 0.5185979008674622, "learning_rate": 3.9659950736325476e-05, "loss": 0.0309, "step": 58800 }, { "grad_norm": 0.5805996060371399, "learning_rate": 3.964377403391174e-05, "loss": 0.0628, "step": 58810 }, { "grad_norm": 0.46044135093688965, "learning_rate": 3.9627598464039303e-05, "loss": 0.0517, "step": 58820 }, { "grad_norm": 0.2533714771270752, "learning_rate": 3.9611424028477096e-05, "loss": 0.0299, "step": 58830 }, { "grad_norm": 0.46112194657325745, "learning_rate": 3.959525072899394e-05, "loss": 0.0346, "step": 58840 }, { "grad_norm": 0.3761857748031616, "learning_rate": 3.9579078567358506e-05, "loss": 0.0278, "step": 58850 }, { "grad_norm": 0.38949161767959595, "learning_rate": 3.956290754533939e-05, "loss": 0.0315, "step": 58860 }, { "grad_norm": 0.353840172290802, "learning_rate": 3.954673766470501e-05, "loss": 0.0373, "step": 58870 }, { "grad_norm": 0.23918282985687256, "learning_rate": 3.9530568927223674e-05, "loss": 0.0323, "step": 58880 }, { "grad_norm": 0.26598966121673584, "learning_rate": 3.95144013346636e-05, "loss": 0.0374, "step": 58890 }, { "grad_norm": 0.2596043050289154, "learning_rate": 3.949823488879282e-05, "loss": 0.0298, "step": 58900 }, { "grad_norm": 0.4096105396747589, "learning_rate": 3.9482069591379276e-05, "loss": 0.038, "step": 58910 }, { "grad_norm": 0.40573492646217346, "learning_rate": 3.9465905444190784e-05, "loss": 0.0345, "step": 58920 }, { "grad_norm": 0.2914033830165863, "learning_rate": 3.944974244899505e-05, "loss": 0.0385, "step": 58930 }, { "grad_norm": 0.38981279730796814, "learning_rate": 3.9433580607559616e-05, "loss": 0.0347, "step": 58940 }, { "grad_norm": 0.25125405192375183, "learning_rate": 3.941741992165193e-05, "loss": 0.0391, "step": 58950 }, { "grad_norm": 0.4362783133983612, "learning_rate": 3.940126039303928e-05, "loss": 0.0328, "step": 58960 }, { "grad_norm": 0.3130943477153778, "learning_rate": 3.938510202348886e-05, "loss": 0.037, "step": 58970 }, { "grad_norm": 0.5376878380775452, "learning_rate": 3.93689448147677e-05, "loss": 0.0308, "step": 58980 }, { "grad_norm": 0.31776270270347595, "learning_rate": 3.935278876864278e-05, "loss": 0.03, "step": 58990 }, { "grad_norm": 0.3853279948234558, "learning_rate": 3.933663388688087e-05, "loss": 0.0388, "step": 59000 }, { "grad_norm": 0.5741214156150818, "learning_rate": 3.9320480171248645e-05, "loss": 0.0315, "step": 59010 }, { "grad_norm": 0.6888112425804138, "learning_rate": 3.930432762351265e-05, "loss": 0.0452, "step": 59020 }, { "grad_norm": 0.37750449776649475, "learning_rate": 3.928817624543931e-05, "loss": 0.0372, "step": 59030 }, { "grad_norm": 0.38321730494499207, "learning_rate": 3.9272026038794885e-05, "loss": 0.0281, "step": 59040 }, { "grad_norm": 0.45686760544776917, "learning_rate": 3.9255877005345586e-05, "loss": 0.0349, "step": 59050 }, { "grad_norm": 0.8916282653808594, "learning_rate": 3.923972914685743e-05, "loss": 0.0352, "step": 59060 }, { "grad_norm": 0.23642076551914215, "learning_rate": 3.922358246509631e-05, "loss": 0.0313, "step": 59070 }, { "grad_norm": 0.33019840717315674, "learning_rate": 3.920743696182801e-05, "loss": 0.0332, "step": 59080 }, { "grad_norm": 0.2782917320728302, "learning_rate": 3.9191292638818164e-05, "loss": 0.0334, "step": 59090 }, { "grad_norm": 0.5568386912345886, "learning_rate": 3.9175149497832316e-05, "loss": 0.0335, "step": 59100 }, { "grad_norm": 0.2804432213306427, "learning_rate": 3.915900754063584e-05, "loss": 0.0322, "step": 59110 }, { "grad_norm": 0.3470161259174347, "learning_rate": 3.9142866768994014e-05, "loss": 0.0392, "step": 59120 }, { "grad_norm": 0.2998741865158081, "learning_rate": 3.912672718467193e-05, "loss": 0.0354, "step": 59130 }, { "grad_norm": 0.40552496910095215, "learning_rate": 3.911058878943463e-05, "loss": 0.0311, "step": 59140 }, { "grad_norm": 0.32640451192855835, "learning_rate": 3.9094451585046974e-05, "loss": 0.0283, "step": 59150 }, { "grad_norm": 0.31839948892593384, "learning_rate": 3.90783155732737e-05, "loss": 0.0291, "step": 59160 }, { "grad_norm": 0.2951540946960449, "learning_rate": 3.9062180755879373e-05, "loss": 0.0268, "step": 59170 }, { "grad_norm": 0.37851449847221375, "learning_rate": 3.904604713462855e-05, "loss": 0.0317, "step": 59180 }, { "grad_norm": 0.2509203553199768, "learning_rate": 3.902991471128554e-05, "loss": 0.022, "step": 59190 }, { "grad_norm": 0.2553553283214569, "learning_rate": 3.901378348761456e-05, "loss": 0.0218, "step": 59200 }, { "grad_norm": 0.3559105098247528, "learning_rate": 3.899765346537971e-05, "loss": 0.0297, "step": 59210 }, { "grad_norm": 0.32638245820999146, "learning_rate": 3.898152464634493e-05, "loss": 0.0296, "step": 59220 }, { "grad_norm": 0.3289332687854767, "learning_rate": 3.896539703227402e-05, "loss": 0.0273, "step": 59230 }, { "grad_norm": 0.3673916459083557, "learning_rate": 3.894927062493073e-05, "loss": 0.0368, "step": 59240 }, { "grad_norm": 0.6709679961204529, "learning_rate": 3.893314542607859e-05, "loss": 0.0324, "step": 59250 }, { "grad_norm": 0.31357094645500183, "learning_rate": 3.891702143748103e-05, "loss": 0.0253, "step": 59260 }, { "grad_norm": 0.44498932361602783, "learning_rate": 3.890089866090133e-05, "loss": 0.033, "step": 59270 }, { "grad_norm": 0.30635136365890503, "learning_rate": 3.888477709810267e-05, "loss": 0.0288, "step": 59280 }, { "grad_norm": 0.4035798907279968, "learning_rate": 3.886865675084806e-05, "loss": 0.028, "step": 59290 }, { "grad_norm": 0.38742074370384216, "learning_rate": 3.885253762090043e-05, "loss": 0.0273, "step": 59300 }, { "grad_norm": 0.3016309142112732, "learning_rate": 3.8836419710022534e-05, "loss": 0.0345, "step": 59310 }, { "grad_norm": 0.5121853351593018, "learning_rate": 3.882030301997698e-05, "loss": 0.0327, "step": 59320 }, { "grad_norm": 0.4294992685317993, "learning_rate": 3.880418755252629e-05, "loss": 0.032, "step": 59330 }, { "grad_norm": 0.48737284541130066, "learning_rate": 3.87880733094328e-05, "loss": 0.0305, "step": 59340 }, { "grad_norm": 0.3328811228275299, "learning_rate": 3.877196029245877e-05, "loss": 0.0251, "step": 59350 }, { "grad_norm": 0.4078899621963501, "learning_rate": 3.875584850336627e-05, "loss": 0.038, "step": 59360 }, { "grad_norm": 0.2888662815093994, "learning_rate": 3.873973794391728e-05, "loss": 0.0282, "step": 59370 }, { "grad_norm": 0.4298751652240753, "learning_rate": 3.872362861587361e-05, "loss": 0.0255, "step": 59380 }, { "grad_norm": 0.24446327984333038, "learning_rate": 3.870752052099698e-05, "loss": 0.0294, "step": 59390 }, { "grad_norm": 0.6931892037391663, "learning_rate": 3.8691413661048916e-05, "loss": 0.0322, "step": 59400 }, { "grad_norm": 1.3321647644042969, "learning_rate": 3.867530803779085e-05, "loss": 0.0328, "step": 59410 }, { "grad_norm": 0.3775247633457184, "learning_rate": 3.8659203652984054e-05, "loss": 0.0369, "step": 59420 }, { "grad_norm": 0.29731887578964233, "learning_rate": 3.864310050838972e-05, "loss": 0.0329, "step": 59430 }, { "grad_norm": 0.22124674916267395, "learning_rate": 3.8626998605768835e-05, "loss": 0.0288, "step": 59440 }, { "grad_norm": 0.37951570749282837, "learning_rate": 3.8610897946882294e-05, "loss": 0.0384, "step": 59450 }, { "grad_norm": 0.3530534505844116, "learning_rate": 3.859479853349082e-05, "loss": 0.038, "step": 59460 }, { "grad_norm": 0.5145135521888733, "learning_rate": 3.857870036735503e-05, "loss": 0.0311, "step": 59470 }, { "grad_norm": 0.3427915871143341, "learning_rate": 3.856260345023538e-05, "loss": 0.0371, "step": 59480 }, { "grad_norm": 0.2843642830848694, "learning_rate": 3.854650778389224e-05, "loss": 0.0262, "step": 59490 }, { "grad_norm": 0.3312915861606598, "learning_rate": 3.8530413370085786e-05, "loss": 0.0298, "step": 59500 }, { "grad_norm": 0.24212050437927246, "learning_rate": 3.851432021057608e-05, "loss": 0.0269, "step": 59510 }, { "grad_norm": 0.3046671152114868, "learning_rate": 3.849822830712306e-05, "loss": 0.0333, "step": 59520 }, { "grad_norm": 0.4972369968891144, "learning_rate": 3.848213766148649e-05, "loss": 0.0291, "step": 59530 }, { "grad_norm": 0.4019722640514374, "learning_rate": 3.846604827542602e-05, "loss": 0.0248, "step": 59540 }, { "grad_norm": 0.37497103214263916, "learning_rate": 3.844996015070117e-05, "loss": 0.0328, "step": 59550 }, { "grad_norm": 0.3489859700202942, "learning_rate": 3.843387328907132e-05, "loss": 0.0255, "step": 59560 }, { "grad_norm": 0.3619144856929779, "learning_rate": 3.8417787692295696e-05, "loss": 0.04, "step": 59570 }, { "grad_norm": 0.30006125569343567, "learning_rate": 3.84017033621334e-05, "loss": 0.0309, "step": 59580 }, { "grad_norm": 0.27721381187438965, "learning_rate": 3.838562030034338e-05, "loss": 0.0271, "step": 59590 }, { "grad_norm": 0.37138158082962036, "learning_rate": 3.836953850868447e-05, "loss": 0.0343, "step": 59600 }, { "grad_norm": 0.5054904818534851, "learning_rate": 3.835345798891533e-05, "loss": 0.032, "step": 59610 }, { "grad_norm": 0.4075677692890167, "learning_rate": 3.833737874279452e-05, "loss": 0.0324, "step": 59620 }, { "grad_norm": 0.3711465895175934, "learning_rate": 3.8321300772080433e-05, "loss": 0.0241, "step": 59630 }, { "grad_norm": 0.39268529415130615, "learning_rate": 3.8305224078531345e-05, "loss": 0.0321, "step": 59640 }, { "grad_norm": 0.524467945098877, "learning_rate": 3.828914866390537e-05, "loss": 0.0372, "step": 59650 }, { "grad_norm": 0.35564735531806946, "learning_rate": 3.827307452996048e-05, "loss": 0.0349, "step": 59660 }, { "grad_norm": 0.36537811160087585, "learning_rate": 3.825700167845454e-05, "loss": 0.0416, "step": 59670 }, { "grad_norm": 0.4171640872955322, "learning_rate": 3.8240930111145214e-05, "loss": 0.0375, "step": 59680 }, { "grad_norm": 0.4452573359012604, "learning_rate": 3.8224859829790116e-05, "loss": 0.0273, "step": 59690 }, { "grad_norm": 0.3312571346759796, "learning_rate": 3.820879083614664e-05, "loss": 0.0401, "step": 59700 }, { "grad_norm": 0.4563349187374115, "learning_rate": 3.819272313197208e-05, "loss": 0.0373, "step": 59710 }, { "grad_norm": 0.6941428780555725, "learning_rate": 3.817665671902355e-05, "loss": 0.0266, "step": 59720 }, { "grad_norm": 0.30333182215690613, "learning_rate": 3.816059159905808e-05, "loss": 0.0353, "step": 59730 }, { "grad_norm": 0.38964852690696716, "learning_rate": 3.8144527773832476e-05, "loss": 0.036, "step": 59740 }, { "grad_norm": 0.316462904214859, "learning_rate": 3.812846524510352e-05, "loss": 0.0239, "step": 59750 }, { "grad_norm": 0.40691816806793213, "learning_rate": 3.811240401462775e-05, "loss": 0.0417, "step": 59760 }, { "grad_norm": 0.38097819685935974, "learning_rate": 3.809634408416162e-05, "loss": 0.0324, "step": 59770 }, { "grad_norm": 0.46053725481033325, "learning_rate": 3.8080285455461394e-05, "loss": 0.0332, "step": 59780 }, { "grad_norm": 0.29493239521980286, "learning_rate": 3.806422813028323e-05, "loss": 0.021, "step": 59790 }, { "grad_norm": 0.24497060477733612, "learning_rate": 3.8048172110383114e-05, "loss": 0.0242, "step": 59800 }, { "grad_norm": 0.6440892815589905, "learning_rate": 3.8032117397516944e-05, "loss": 0.0369, "step": 59810 }, { "grad_norm": 0.45094019174575806, "learning_rate": 3.801606399344042e-05, "loss": 0.0347, "step": 59820 }, { "grad_norm": 0.21605505049228668, "learning_rate": 3.800001189990913e-05, "loss": 0.0248, "step": 59830 }, { "grad_norm": 0.4038585424423218, "learning_rate": 3.798396111867847e-05, "loss": 0.0291, "step": 59840 }, { "grad_norm": 0.6160882115364075, "learning_rate": 3.796791165150378e-05, "loss": 0.0306, "step": 59850 }, { "grad_norm": 0.3761032521724701, "learning_rate": 3.795186350014018e-05, "loss": 0.0339, "step": 59860 }, { "grad_norm": 0.2833250164985657, "learning_rate": 3.793581666634266e-05, "loss": 0.0295, "step": 59870 }, { "grad_norm": 0.3783721327781677, "learning_rate": 3.791977115186609e-05, "loss": 0.0238, "step": 59880 }, { "grad_norm": 0.2502119243144989, "learning_rate": 3.79037269584652e-05, "loss": 0.0253, "step": 59890 }, { "grad_norm": 0.2913210093975067, "learning_rate": 3.7887684087894545e-05, "loss": 0.0305, "step": 59900 }, { "grad_norm": 0.39112624526023865, "learning_rate": 3.787164254190854e-05, "loss": 0.0295, "step": 59910 }, { "grad_norm": 0.3059837222099304, "learning_rate": 3.7855602322261473e-05, "loss": 0.0224, "step": 59920 }, { "grad_norm": 0.40360453724861145, "learning_rate": 3.783956343070746e-05, "loss": 0.0359, "step": 59930 }, { "grad_norm": 0.4221884310245514, "learning_rate": 3.7823525869000526e-05, "loss": 0.0281, "step": 59940 }, { "grad_norm": 0.4035794138908386, "learning_rate": 3.78074896388945e-05, "loss": 0.0285, "step": 59950 }, { "grad_norm": 0.3801821172237396, "learning_rate": 3.7791454742143075e-05, "loss": 0.0298, "step": 59960 }, { "grad_norm": 0.3241272270679474, "learning_rate": 3.7775421180499795e-05, "loss": 0.0301, "step": 59970 }, { "grad_norm": 0.2295122742652893, "learning_rate": 3.7759388955718086e-05, "loss": 0.0333, "step": 59980 }, { "grad_norm": 0.24388417601585388, "learning_rate": 3.774335806955117e-05, "loss": 0.0352, "step": 59990 }, { "grad_norm": 0.7443691492080688, "learning_rate": 3.7727328523752215e-05, "loss": 0.0289, "step": 60000 }, { "grad_norm": 0.27512243390083313, "learning_rate": 3.771130032007415e-05, "loss": 0.0375, "step": 60010 }, { "grad_norm": 0.26147136092185974, "learning_rate": 3.769527346026982e-05, "loss": 0.0235, "step": 60020 }, { "grad_norm": 0.32851123809814453, "learning_rate": 3.767924794609188e-05, "loss": 0.0276, "step": 60030 }, { "grad_norm": 0.5136595368385315, "learning_rate": 3.7663223779292854e-05, "loss": 0.0343, "step": 60040 }, { "grad_norm": 0.2873753011226654, "learning_rate": 3.764720096162512e-05, "loss": 0.0321, "step": 60050 }, { "grad_norm": 0.37451332807540894, "learning_rate": 3.763117949484092e-05, "loss": 0.0287, "step": 60060 }, { "grad_norm": 0.7774701714515686, "learning_rate": 3.7615159380692344e-05, "loss": 0.0351, "step": 60070 }, { "grad_norm": 0.3048303723335266, "learning_rate": 3.759914062093132e-05, "loss": 0.0247, "step": 60080 }, { "grad_norm": 0.30936765670776367, "learning_rate": 3.7583123217309615e-05, "loss": 0.0268, "step": 60090 }, { "grad_norm": 0.5713962912559509, "learning_rate": 3.7567107171578904e-05, "loss": 0.0382, "step": 60100 }, { "grad_norm": 0.6811699271202087, "learning_rate": 3.755109248549066e-05, "loss": 0.0329, "step": 60110 }, { "grad_norm": 0.4335627257823944, "learning_rate": 3.75350791607962e-05, "loss": 0.0286, "step": 60120 }, { "grad_norm": 0.5855022668838501, "learning_rate": 3.751906719924676e-05, "loss": 0.0359, "step": 60130 }, { "grad_norm": 0.3371928036212921, "learning_rate": 3.750305660259337e-05, "loss": 0.0298, "step": 60140 }, { "grad_norm": 0.3334648907184601, "learning_rate": 3.748704737258693e-05, "loss": 0.0278, "step": 60150 }, { "grad_norm": 0.24519668519496918, "learning_rate": 3.747103951097816e-05, "loss": 0.0447, "step": 60160 }, { "grad_norm": 0.3773103952407837, "learning_rate": 3.745503301951768e-05, "loss": 0.0301, "step": 60170 }, { "grad_norm": 0.5195199251174927, "learning_rate": 3.743902789995592e-05, "loss": 0.0331, "step": 60180 }, { "grad_norm": 0.4511295557022095, "learning_rate": 3.7423024154043183e-05, "loss": 0.0284, "step": 60190 }, { "grad_norm": 0.40160927176475525, "learning_rate": 3.740702178352963e-05, "loss": 0.0242, "step": 60200 }, { "grad_norm": 0.3638184070587158, "learning_rate": 3.739102079016523e-05, "loss": 0.0307, "step": 60210 }, { "grad_norm": 0.2682218849658966, "learning_rate": 3.7375021175699846e-05, "loss": 0.0218, "step": 60220 }, { "grad_norm": 0.307613730430603, "learning_rate": 3.735902294188317e-05, "loss": 0.028, "step": 60230 }, { "grad_norm": 0.451156347990036, "learning_rate": 3.734302609046473e-05, "loss": 0.0281, "step": 60240 }, { "grad_norm": 0.4045931100845337, "learning_rate": 3.732703062319391e-05, "loss": 0.0296, "step": 60250 }, { "grad_norm": 1.4491480588912964, "learning_rate": 3.731103654181999e-05, "loss": 0.0355, "step": 60260 }, { "grad_norm": 0.41247671842575073, "learning_rate": 3.729504384809203e-05, "loss": 0.0264, "step": 60270 }, { "grad_norm": 0.29606762528419495, "learning_rate": 3.727905254375898e-05, "loss": 0.0305, "step": 60280 }, { "grad_norm": 0.427462637424469, "learning_rate": 3.7263062630569614e-05, "loss": 0.0276, "step": 60290 }, { "grad_norm": 0.37428614497184753, "learning_rate": 3.724707411027256e-05, "loss": 0.029, "step": 60300 }, { "grad_norm": 0.3094009757041931, "learning_rate": 3.723108698461631e-05, "loss": 0.0246, "step": 60310 }, { "grad_norm": 0.7896850109100342, "learning_rate": 3.72151012553492e-05, "loss": 0.0349, "step": 60320 }, { "grad_norm": 0.33032292127609253, "learning_rate": 3.719911692421939e-05, "loss": 0.021, "step": 60330 }, { "grad_norm": 0.2615523934364319, "learning_rate": 3.71831339929749e-05, "loss": 0.0275, "step": 60340 }, { "grad_norm": 0.4756886065006256, "learning_rate": 3.716715246336362e-05, "loss": 0.0346, "step": 60350 }, { "grad_norm": 0.33028244972229004, "learning_rate": 3.715117233713324e-05, "loss": 0.0302, "step": 60360 }, { "grad_norm": 0.41186830401420593, "learning_rate": 3.713519361603135e-05, "loss": 0.0298, "step": 60370 }, { "grad_norm": 0.36878514289855957, "learning_rate": 3.711921630180532e-05, "loss": 0.0342, "step": 60380 }, { "grad_norm": 0.4725322723388672, "learning_rate": 3.710324039620245e-05, "loss": 0.0247, "step": 60390 }, { "grad_norm": 0.3542270362377167, "learning_rate": 3.708726590096982e-05, "loss": 0.03, "step": 60400 }, { "grad_norm": 0.2481231391429901, "learning_rate": 3.707129281785437e-05, "loss": 0.0268, "step": 60410 }, { "grad_norm": 0.3612735867500305, "learning_rate": 3.705532114860291e-05, "loss": 0.0281, "step": 60420 }, { "grad_norm": 0.5285367369651794, "learning_rate": 3.703935089496207e-05, "loss": 0.0487, "step": 60430 }, { "grad_norm": 0.3991299569606781, "learning_rate": 3.702338205867829e-05, "loss": 0.0299, "step": 60440 }, { "grad_norm": 0.33740341663360596, "learning_rate": 3.7007414641497976e-05, "loss": 0.0308, "step": 60450 }, { "grad_norm": 0.2735954821109772, "learning_rate": 3.699144864516726e-05, "loss": 0.0294, "step": 60460 }, { "grad_norm": 0.22529740631580353, "learning_rate": 3.6975484071432153e-05, "loss": 0.0272, "step": 60470 }, { "grad_norm": 0.2763127386569977, "learning_rate": 3.695952092203854e-05, "loss": 0.0354, "step": 60480 }, { "grad_norm": 0.4001387059688568, "learning_rate": 3.69435591987321e-05, "loss": 0.0341, "step": 60490 }, { "grad_norm": 0.3399592339992523, "learning_rate": 3.6927598903258374e-05, "loss": 0.0266, "step": 60500 }, { "grad_norm": 0.44360610842704773, "learning_rate": 3.691164003736281e-05, "loss": 0.022, "step": 60510 }, { "grad_norm": 0.3029923737049103, "learning_rate": 3.689568260279061e-05, "loss": 0.0316, "step": 60520 }, { "grad_norm": 0.3757285475730896, "learning_rate": 3.687972660128686e-05, "loss": 0.024, "step": 60530 }, { "grad_norm": 0.38684961199760437, "learning_rate": 3.686377203459648e-05, "loss": 0.0314, "step": 60540 }, { "grad_norm": 0.272937536239624, "learning_rate": 3.6847818904464226e-05, "loss": 0.0274, "step": 60550 }, { "grad_norm": 0.37087807059288025, "learning_rate": 3.683186721263474e-05, "loss": 0.0326, "step": 60560 }, { "grad_norm": 0.3971720337867737, "learning_rate": 3.6815916960852435e-05, "loss": 0.0312, "step": 60570 }, { "grad_norm": 0.26406222581863403, "learning_rate": 3.679996815086165e-05, "loss": 0.0277, "step": 60580 }, { "grad_norm": 0.33263617753982544, "learning_rate": 3.678402078440649e-05, "loss": 0.0244, "step": 60590 }, { "grad_norm": 0.4180505871772766, "learning_rate": 3.676807486323096e-05, "loss": 0.0311, "step": 60600 }, { "grad_norm": 0.33589598536491394, "learning_rate": 3.6752130389078864e-05, "loss": 0.0305, "step": 60610 }, { "grad_norm": 0.3371417224407196, "learning_rate": 3.673618736369388e-05, "loss": 0.0252, "step": 60620 }, { "grad_norm": 0.369499534368515, "learning_rate": 3.6720245788819474e-05, "loss": 0.0301, "step": 60630 }, { "grad_norm": 0.35920098423957825, "learning_rate": 3.6704305666199044e-05, "loss": 0.0302, "step": 60640 }, { "grad_norm": 0.39402544498443604, "learning_rate": 3.668836699757576e-05, "loss": 0.0263, "step": 60650 }, { "grad_norm": 0.450080931186676, "learning_rate": 3.667242978469265e-05, "loss": 0.0293, "step": 60660 }, { "grad_norm": 0.7937309741973877, "learning_rate": 3.6656494029292575e-05, "loss": 0.0368, "step": 60670 }, { "grad_norm": 0.4519682228565216, "learning_rate": 3.664055973311825e-05, "loss": 0.0298, "step": 60680 }, { "grad_norm": 0.37349459528923035, "learning_rate": 3.662462689791221e-05, "loss": 0.0256, "step": 60690 }, { "grad_norm": 0.26618677377700806, "learning_rate": 3.660869552541689e-05, "loss": 0.0281, "step": 60700 }, { "grad_norm": 0.5867949724197388, "learning_rate": 3.659276561737448e-05, "loss": 0.0303, "step": 60710 }, { "grad_norm": 0.2771640419960022, "learning_rate": 3.657683717552708e-05, "loss": 0.0318, "step": 60720 }, { "grad_norm": 0.9188153147697449, "learning_rate": 3.6560910201616574e-05, "loss": 0.0252, "step": 60730 }, { "grad_norm": 0.31644779443740845, "learning_rate": 3.654498469738472e-05, "loss": 0.0323, "step": 60740 }, { "grad_norm": 0.538810133934021, "learning_rate": 3.6529060664573126e-05, "loss": 0.0288, "step": 60750 }, { "grad_norm": 0.2822607159614563, "learning_rate": 3.6513138104923176e-05, "loss": 0.0266, "step": 60760 }, { "grad_norm": 0.4085228443145752, "learning_rate": 3.6497217020176176e-05, "loss": 0.028, "step": 60770 }, { "grad_norm": 0.43396976590156555, "learning_rate": 3.648129741207323e-05, "loss": 0.0363, "step": 60780 }, { "grad_norm": 0.28911370038986206, "learning_rate": 3.646537928235527e-05, "loss": 0.0268, "step": 60790 }, { "grad_norm": 0.42627018690109253, "learning_rate": 3.6449462632763075e-05, "loss": 0.0319, "step": 60800 }, { "grad_norm": 0.3415481150150299, "learning_rate": 3.643354746503729e-05, "loss": 0.0318, "step": 60810 }, { "grad_norm": 0.43814709782600403, "learning_rate": 3.6417633780918345e-05, "loss": 0.0296, "step": 60820 }, { "grad_norm": 0.3971923291683197, "learning_rate": 3.640172158214655e-05, "loss": 0.0301, "step": 60830 }, { "grad_norm": 0.624457836151123, "learning_rate": 3.6385810870462035e-05, "loss": 0.0238, "step": 60840 }, { "grad_norm": 0.3099433481693268, "learning_rate": 3.636990164760479e-05, "loss": 0.0264, "step": 60850 }, { "grad_norm": 0.3133166432380676, "learning_rate": 3.6353993915314596e-05, "loss": 0.0301, "step": 60860 }, { "grad_norm": 0.4463246464729309, "learning_rate": 3.6338087675331124e-05, "loss": 0.0254, "step": 60870 }, { "grad_norm": 0.2717677056789398, "learning_rate": 3.632218292939383e-05, "loss": 0.0233, "step": 60880 }, { "grad_norm": 0.3396094739437103, "learning_rate": 3.6306279679242033e-05, "loss": 0.0269, "step": 60890 }, { "grad_norm": 0.3076099157333374, "learning_rate": 3.629037792661492e-05, "loss": 0.0369, "step": 60900 }, { "grad_norm": 0.34711572527885437, "learning_rate": 3.627447767325147e-05, "loss": 0.0318, "step": 60910 }, { "grad_norm": 0.3690762221813202, "learning_rate": 3.62585789208905e-05, "loss": 0.0245, "step": 60920 }, { "grad_norm": 0.42922675609588623, "learning_rate": 3.624268167127068e-05, "loss": 0.0416, "step": 60930 }, { "grad_norm": 0.24985377490520477, "learning_rate": 3.622678592613051e-05, "loss": 0.0275, "step": 60940 }, { "grad_norm": 0.9559735655784607, "learning_rate": 3.6210891687208304e-05, "loss": 0.0324, "step": 60950 }, { "grad_norm": 0.39564692974090576, "learning_rate": 3.619499895624228e-05, "loss": 0.0279, "step": 60960 }, { "grad_norm": 0.3838694393634796, "learning_rate": 3.61791077349704e-05, "loss": 0.0351, "step": 60970 }, { "grad_norm": 0.4528544843196869, "learning_rate": 3.616321802513053e-05, "loss": 0.0314, "step": 60980 }, { "grad_norm": 0.4172457456588745, "learning_rate": 3.6147329828460325e-05, "loss": 0.0336, "step": 60990 }, { "grad_norm": 0.5978953838348389, "learning_rate": 3.6131443146697306e-05, "loss": 0.0332, "step": 61000 }, { "grad_norm": 0.357359379529953, "learning_rate": 3.611555798157879e-05, "loss": 0.0261, "step": 61010 }, { "grad_norm": 0.5649371147155762, "learning_rate": 3.6099674334842e-05, "loss": 0.033, "step": 61020 }, { "grad_norm": 0.36300069093704224, "learning_rate": 3.6083792208223934e-05, "loss": 0.0329, "step": 61030 }, { "grad_norm": 0.43591663241386414, "learning_rate": 3.606791160346142e-05, "loss": 0.0306, "step": 61040 }, { "grad_norm": 1.202975869178772, "learning_rate": 3.6052032522291134e-05, "loss": 0.0306, "step": 61050 }, { "grad_norm": 0.528786838054657, "learning_rate": 3.603615496644962e-05, "loss": 0.0467, "step": 61060 }, { "grad_norm": 0.6123021841049194, "learning_rate": 3.60202789376732e-05, "loss": 0.0331, "step": 61070 }, { "grad_norm": 0.3973851203918457, "learning_rate": 3.6004404437698036e-05, "loss": 0.0363, "step": 61080 }, { "grad_norm": 0.4246763586997986, "learning_rate": 3.5988531468260184e-05, "loss": 0.0297, "step": 61090 }, { "grad_norm": 0.4817545413970947, "learning_rate": 3.597266003109546e-05, "loss": 0.0281, "step": 61100 }, { "grad_norm": 0.2813129127025604, "learning_rate": 3.5956790127939557e-05, "loss": 0.0294, "step": 61110 }, { "grad_norm": 0.36888355016708374, "learning_rate": 3.594092176052796e-05, "loss": 0.0307, "step": 61120 }, { "grad_norm": 0.29600027203559875, "learning_rate": 3.592505493059603e-05, "loss": 0.027, "step": 61130 }, { "grad_norm": 0.2620311379432678, "learning_rate": 3.5909189639878905e-05, "loss": 0.037, "step": 61140 }, { "grad_norm": 0.26306408643722534, "learning_rate": 3.5893325890111626e-05, "loss": 0.0263, "step": 61150 }, { "grad_norm": 0.42190468311309814, "learning_rate": 3.5877463683029034e-05, "loss": 0.0371, "step": 61160 }, { "grad_norm": 0.23805499076843262, "learning_rate": 3.586160302036578e-05, "loss": 0.0325, "step": 61170 }, { "grad_norm": 0.3479901850223541, "learning_rate": 3.584574390385636e-05, "loss": 0.0259, "step": 61180 }, { "grad_norm": 0.21525822579860687, "learning_rate": 3.582988633523511e-05, "loss": 0.0349, "step": 61190 }, { "grad_norm": 0.23859164118766785, "learning_rate": 3.581403031623616e-05, "loss": 0.033, "step": 61200 }, { "grad_norm": 0.38795989751815796, "learning_rate": 3.579817584859354e-05, "loss": 0.0297, "step": 61210 }, { "grad_norm": 0.6578410863876343, "learning_rate": 3.578232293404107e-05, "loss": 0.0319, "step": 61220 }, { "grad_norm": 0.4391905665397644, "learning_rate": 3.5766471574312385e-05, "loss": 0.0275, "step": 61230 }, { "grad_norm": 0.3164002001285553, "learning_rate": 3.5750621771140964e-05, "loss": 0.0357, "step": 61240 }, { "grad_norm": 0.2932479977607727, "learning_rate": 3.5734773526260115e-05, "loss": 0.0273, "step": 61250 }, { "grad_norm": 0.3290444612503052, "learning_rate": 3.5718926841402996e-05, "loss": 0.0362, "step": 61260 }, { "grad_norm": 0.2657833993434906, "learning_rate": 3.5703081718302555e-05, "loss": 0.0338, "step": 61270 }, { "grad_norm": 0.3645694851875305, "learning_rate": 3.568723815869162e-05, "loss": 0.0382, "step": 61280 }, { "grad_norm": 0.3164746165275574, "learning_rate": 3.5671396164302777e-05, "loss": 0.0483, "step": 61290 }, { "grad_norm": 0.35324352979660034, "learning_rate": 3.565555573686853e-05, "loss": 0.0284, "step": 61300 }, { "grad_norm": 0.2932227551937103, "learning_rate": 3.563971687812113e-05, "loss": 0.0325, "step": 61310 }, { "grad_norm": 0.6193833351135254, "learning_rate": 3.5623879589792706e-05, "loss": 0.0298, "step": 61320 }, { "grad_norm": 0.24678130447864532, "learning_rate": 3.5608043873615185e-05, "loss": 0.041, "step": 61330 }, { "grad_norm": 0.44290706515312195, "learning_rate": 3.559220973132035e-05, "loss": 0.0264, "step": 61340 }, { "grad_norm": 0.3327983617782593, "learning_rate": 3.557637716463981e-05, "loss": 0.029, "step": 61350 }, { "grad_norm": 0.4558064639568329, "learning_rate": 3.556054617530497e-05, "loss": 0.031, "step": 61360 }, { "grad_norm": 0.7247751951217651, "learning_rate": 3.5544716765047106e-05, "loss": 0.025, "step": 61370 }, { "grad_norm": 0.5284976959228516, "learning_rate": 3.5528888935597274e-05, "loss": 0.0293, "step": 61380 }, { "grad_norm": 0.3601451516151428, "learning_rate": 3.551306268868636e-05, "loss": 0.0282, "step": 61390 }, { "grad_norm": 0.2074568122625351, "learning_rate": 3.5497238026045166e-05, "loss": 0.0296, "step": 61400 }, { "grad_norm": 0.32231974601745605, "learning_rate": 3.548141494940421e-05, "loss": 0.0295, "step": 61410 }, { "grad_norm": 0.5192675590515137, "learning_rate": 3.546559346049388e-05, "loss": 0.0266, "step": 61420 }, { "grad_norm": 0.2789355516433716, "learning_rate": 3.544977356104441e-05, "loss": 0.0264, "step": 61430 }, { "grad_norm": 0.30607667565345764, "learning_rate": 3.543395525278582e-05, "loss": 0.0299, "step": 61440 }, { "grad_norm": 0.4188045263290405, "learning_rate": 3.5418138537447985e-05, "loss": 0.0314, "step": 61450 }, { "grad_norm": 0.2114272117614746, "learning_rate": 3.5402323416760574e-05, "loss": 0.023, "step": 61460 }, { "grad_norm": 0.24180835485458374, "learning_rate": 3.538650989245316e-05, "loss": 0.0283, "step": 61470 }, { "grad_norm": 0.4408551752567291, "learning_rate": 3.537069796625504e-05, "loss": 0.0248, "step": 61480 }, { "grad_norm": 0.43426513671875, "learning_rate": 3.53548876398954e-05, "loss": 0.0303, "step": 61490 }, { "grad_norm": 0.6308358907699585, "learning_rate": 3.5339078915103214e-05, "loss": 0.0301, "step": 61500 }, { "grad_norm": 0.42664048075675964, "learning_rate": 3.532327179360733e-05, "loss": 0.0347, "step": 61510 }, { "grad_norm": 0.35777339339256287, "learning_rate": 3.530746627713636e-05, "loss": 0.0287, "step": 61520 }, { "grad_norm": 0.3746180832386017, "learning_rate": 3.5291662367418805e-05, "loss": 0.0264, "step": 61530 }, { "grad_norm": 0.430002897977829, "learning_rate": 3.527586006618293e-05, "loss": 0.0389, "step": 61540 }, { "grad_norm": 0.7097029685974121, "learning_rate": 3.526005937515687e-05, "loss": 0.0437, "step": 61550 }, { "grad_norm": 0.39541491866111755, "learning_rate": 3.524426029606856e-05, "loss": 0.0489, "step": 61560 }, { "grad_norm": 0.2667592167854309, "learning_rate": 3.5228462830645746e-05, "loss": 0.0319, "step": 61570 }, { "grad_norm": 0.47946876287460327, "learning_rate": 3.5212666980616036e-05, "loss": 0.0297, "step": 61580 }, { "grad_norm": 0.36700332164764404, "learning_rate": 3.519687274770681e-05, "loss": 0.0343, "step": 61590 }, { "grad_norm": 0.28237712383270264, "learning_rate": 3.518108013364535e-05, "loss": 0.0471, "step": 61600 }, { "grad_norm": 0.37903764843940735, "learning_rate": 3.516528914015868e-05, "loss": 0.0371, "step": 61610 }, { "grad_norm": 0.2891414761543274, "learning_rate": 3.5149499768973694e-05, "loss": 0.0375, "step": 61620 }, { "grad_norm": 0.1535412222146988, "learning_rate": 3.513371202181708e-05, "loss": 0.0288, "step": 61630 }, { "grad_norm": 0.5365455150604248, "learning_rate": 3.511792590041537e-05, "loss": 0.0279, "step": 61640 }, { "grad_norm": 0.41434618830680847, "learning_rate": 3.510214140649488e-05, "loss": 0.0365, "step": 61650 }, { "grad_norm": 0.2588222324848175, "learning_rate": 3.508635854178184e-05, "loss": 0.0354, "step": 61660 }, { "grad_norm": 0.47621288895606995, "learning_rate": 3.50705773080022e-05, "loss": 0.0342, "step": 61670 }, { "grad_norm": 0.3647913634777069, "learning_rate": 3.505479770688178e-05, "loss": 0.0274, "step": 61680 }, { "grad_norm": 0.2989189922809601, "learning_rate": 3.503901974014622e-05, "loss": 0.0214, "step": 61690 }, { "grad_norm": 0.2904120981693268, "learning_rate": 3.502324340952095e-05, "loss": 0.0307, "step": 61700 }, { "grad_norm": 0.45228707790374756, "learning_rate": 3.500746871673126e-05, "loss": 0.0294, "step": 61710 }, { "grad_norm": 0.3332561254501343, "learning_rate": 3.499169566350227e-05, "loss": 0.0379, "step": 61720 }, { "grad_norm": 0.31247928738594055, "learning_rate": 3.4975924251558876e-05, "loss": 0.0261, "step": 61730 }, { "grad_norm": 0.2214067131280899, "learning_rate": 3.496015448262582e-05, "loss": 0.0295, "step": 61740 }, { "grad_norm": 0.46560850739479065, "learning_rate": 3.494438635842764e-05, "loss": 0.0312, "step": 61750 }, { "grad_norm": 0.2326737344264984, "learning_rate": 3.492861988068874e-05, "loss": 0.0275, "step": 61760 }, { "grad_norm": 0.21329718828201294, "learning_rate": 3.491285505113332e-05, "loss": 0.0362, "step": 61770 }, { "grad_norm": 0.31004613637924194, "learning_rate": 3.4897091871485376e-05, "loss": 0.0439, "step": 61780 }, { "grad_norm": 0.2558082640171051, "learning_rate": 3.488133034346875e-05, "loss": 0.0255, "step": 61790 }, { "grad_norm": 0.6040802597999573, "learning_rate": 3.486557046880713e-05, "loss": 0.0299, "step": 61800 }, { "grad_norm": 0.4868335425853729, "learning_rate": 3.484981224922395e-05, "loss": 0.0359, "step": 61810 }, { "grad_norm": 0.3848229646682739, "learning_rate": 3.4834055686442544e-05, "loss": 0.0222, "step": 61820 }, { "grad_norm": 0.354021281003952, "learning_rate": 3.4818300782186e-05, "loss": 0.0271, "step": 61830 }, { "grad_norm": 0.47791948914527893, "learning_rate": 3.4802547538177236e-05, "loss": 0.0282, "step": 61840 }, { "grad_norm": 0.3503728210926056, "learning_rate": 3.478679595613905e-05, "loss": 0.0287, "step": 61850 }, { "grad_norm": 0.270025372505188, "learning_rate": 3.4771046037793975e-05, "loss": 0.0247, "step": 61860 }, { "grad_norm": 0.3142806887626648, "learning_rate": 3.4755297784864416e-05, "loss": 0.035, "step": 61870 }, { "grad_norm": 0.2984618842601776, "learning_rate": 3.473955119907257e-05, "loss": 0.0233, "step": 61880 }, { "grad_norm": 0.3411279618740082, "learning_rate": 3.472380628214046e-05, "loss": 0.0294, "step": 61890 }, { "grad_norm": 0.461853951215744, "learning_rate": 3.4708063035789906e-05, "loss": 0.0267, "step": 61900 }, { "grad_norm": 0.2851952910423279, "learning_rate": 3.4692321461742614e-05, "loss": 0.0316, "step": 61910 }, { "grad_norm": 0.2010648548603058, "learning_rate": 3.4676581561720023e-05, "loss": 0.0285, "step": 61920 }, { "grad_norm": 0.529572606086731, "learning_rate": 3.466084333744344e-05, "loss": 0.0268, "step": 61930 }, { "grad_norm": 0.2659015655517578, "learning_rate": 3.4645106790633964e-05, "loss": 0.0333, "step": 61940 }, { "grad_norm": 0.4013194441795349, "learning_rate": 3.4629371923012526e-05, "loss": 0.0382, "step": 61950 }, { "grad_norm": 0.2838525176048279, "learning_rate": 3.4613638736299856e-05, "loss": 0.0307, "step": 61960 }, { "grad_norm": 0.7040934562683105, "learning_rate": 3.459790723221652e-05, "loss": 0.022, "step": 61970 }, { "grad_norm": 0.6847468018531799, "learning_rate": 3.45821774124829e-05, "loss": 0.0329, "step": 61980 }, { "grad_norm": 0.37218987941741943, "learning_rate": 3.456644927881918e-05, "loss": 0.027, "step": 61990 }, { "grad_norm": 0.43997979164123535, "learning_rate": 3.4550722832945354e-05, "loss": 0.0304, "step": 62000 }, { "grad_norm": 0.30496519804000854, "learning_rate": 3.453499807658126e-05, "loss": 0.0266, "step": 62010 }, { "grad_norm": 0.37878143787384033, "learning_rate": 3.451927501144653e-05, "loss": 0.0265, "step": 62020 }, { "grad_norm": 0.509506344795227, "learning_rate": 3.450355363926059e-05, "loss": 0.0302, "step": 62030 }, { "grad_norm": 0.35650038719177246, "learning_rate": 3.448783396174273e-05, "loss": 0.028, "step": 62040 }, { "grad_norm": 0.4190034866333008, "learning_rate": 3.4472115980612034e-05, "loss": 0.0288, "step": 62050 }, { "grad_norm": 0.33071258664131165, "learning_rate": 3.4456399697587396e-05, "loss": 0.0298, "step": 62060 }, { "grad_norm": 0.5265462398529053, "learning_rate": 3.4440685114387506e-05, "loss": 0.0365, "step": 62070 }, { "grad_norm": 0.36122530698776245, "learning_rate": 3.442497223273091e-05, "loss": 0.0306, "step": 62080 }, { "grad_norm": 0.2830696702003479, "learning_rate": 3.4409261054335926e-05, "loss": 0.0361, "step": 62090 }, { "grad_norm": 0.3113544285297394, "learning_rate": 3.439355158092069e-05, "loss": 0.0406, "step": 62100 }, { "grad_norm": 0.3883553147315979, "learning_rate": 3.437784381420322e-05, "loss": 0.0365, "step": 62110 }, { "grad_norm": 0.3767012357711792, "learning_rate": 3.436213775590126e-05, "loss": 0.0246, "step": 62120 }, { "grad_norm": 0.24790863692760468, "learning_rate": 3.4346433407732405e-05, "loss": 0.0315, "step": 62130 }, { "grad_norm": 0.18005156517028809, "learning_rate": 3.433073077141405e-05, "loss": 0.0201, "step": 62140 }, { "grad_norm": 0.6886096000671387, "learning_rate": 3.431502984866343e-05, "loss": 0.0363, "step": 62150 }, { "grad_norm": 0.4174043536186218, "learning_rate": 3.429933064119753e-05, "loss": 0.0472, "step": 62160 }, { "grad_norm": 0.5597219467163086, "learning_rate": 3.4283633150733255e-05, "loss": 0.0306, "step": 62170 }, { "grad_norm": 0.3323746621608734, "learning_rate": 3.4267937378987225e-05, "loss": 0.0302, "step": 62180 }, { "grad_norm": 0.4401204586029053, "learning_rate": 3.425224332767592e-05, "loss": 0.0358, "step": 62190 }, { "grad_norm": 0.7737047672271729, "learning_rate": 3.4236550998515595e-05, "loss": 0.0413, "step": 62200 }, { "grad_norm": 0.37561461329460144, "learning_rate": 3.422086039322235e-05, "loss": 0.0288, "step": 62210 }, { "grad_norm": 0.3359679579734802, "learning_rate": 3.420517151351209e-05, "loss": 0.0283, "step": 62220 }, { "grad_norm": 0.30496785044670105, "learning_rate": 3.418948436110053e-05, "loss": 0.0374, "step": 62230 }, { "grad_norm": 0.35104700922966003, "learning_rate": 3.417379893770319e-05, "loss": 0.0372, "step": 62240 }, { "grad_norm": 0.40178871154785156, "learning_rate": 3.41581152450354e-05, "loss": 0.0225, "step": 62250 }, { "grad_norm": 0.20605269074440002, "learning_rate": 3.414243328481231e-05, "loss": 0.029, "step": 62260 }, { "grad_norm": 0.20853929221630096, "learning_rate": 3.412675305874888e-05, "loss": 0.0253, "step": 62270 }, { "grad_norm": 0.2722022533416748, "learning_rate": 3.4111074568559876e-05, "loss": 0.0269, "step": 62280 }, { "grad_norm": 0.32891732454299927, "learning_rate": 3.4095397815959846e-05, "loss": 0.0361, "step": 62290 }, { "grad_norm": 0.3012548089027405, "learning_rate": 3.407972280266322e-05, "loss": 0.031, "step": 62300 }, { "grad_norm": 0.7054838538169861, "learning_rate": 3.4064049530384166e-05, "loss": 0.0321, "step": 62310 }, { "grad_norm": 0.4154065251350403, "learning_rate": 3.40483780008367e-05, "loss": 0.0312, "step": 62320 }, { "grad_norm": 0.26556792855262756, "learning_rate": 3.4032708215734635e-05, "loss": 0.0325, "step": 62330 }, { "grad_norm": 0.42638054490089417, "learning_rate": 3.4017040176791584e-05, "loss": 0.0339, "step": 62340 }, { "grad_norm": 0.7655786275863647, "learning_rate": 3.4001373885720976e-05, "loss": 0.042, "step": 62350 }, { "grad_norm": 0.27994734048843384, "learning_rate": 3.398570934423608e-05, "loss": 0.0358, "step": 62360 }, { "grad_norm": 0.33447644114494324, "learning_rate": 3.397004655404994e-05, "loss": 0.0345, "step": 62370 }, { "grad_norm": 0.35852301120758057, "learning_rate": 3.395438551687541e-05, "loss": 0.0319, "step": 62380 }, { "grad_norm": 0.46289733052253723, "learning_rate": 3.393872623442515e-05, "loss": 0.0281, "step": 62390 }, { "grad_norm": 0.28656572103500366, "learning_rate": 3.392306870841164e-05, "loss": 0.0311, "step": 62400 }, { "grad_norm": 0.40709421038627625, "learning_rate": 3.390741294054715e-05, "loss": 0.0259, "step": 62410 }, { "grad_norm": 0.5506879091262817, "learning_rate": 3.38917589325438e-05, "loss": 0.0294, "step": 62420 }, { "grad_norm": 0.5691053867340088, "learning_rate": 3.387610668611349e-05, "loss": 0.0434, "step": 62430 }, { "grad_norm": 0.5236779451370239, "learning_rate": 3.38604562029679e-05, "loss": 0.0378, "step": 62440 }, { "grad_norm": 0.5574438571929932, "learning_rate": 3.384480748481855e-05, "loss": 0.0363, "step": 62450 }, { "grad_norm": 0.2883243262767792, "learning_rate": 3.382916053337676e-05, "loss": 0.0304, "step": 62460 }, { "grad_norm": 0.20543065667152405, "learning_rate": 3.381351535035366e-05, "loss": 0.0304, "step": 62470 }, { "grad_norm": 0.316316694021225, "learning_rate": 3.379787193746019e-05, "loss": 0.0265, "step": 62480 }, { "grad_norm": 0.3309361934661865, "learning_rate": 3.378223029640709e-05, "loss": 0.0295, "step": 62490 }, { "grad_norm": 0.31806471943855286, "learning_rate": 3.376659042890489e-05, "loss": 0.0262, "step": 62500 }, { "grad_norm": 0.741882860660553, "learning_rate": 3.375095233666397e-05, "loss": 0.0267, "step": 62510 }, { "grad_norm": 0.2662109434604645, "learning_rate": 3.373531602139447e-05, "loss": 0.0252, "step": 62520 }, { "grad_norm": 0.42600712180137634, "learning_rate": 3.371968148480635e-05, "loss": 0.0304, "step": 62530 }, { "grad_norm": 0.33348941802978516, "learning_rate": 3.370404872860937e-05, "loss": 0.031, "step": 62540 }, { "grad_norm": 0.21680834889411926, "learning_rate": 3.3688417754513145e-05, "loss": 0.0342, "step": 62550 }, { "grad_norm": 0.24568814039230347, "learning_rate": 3.367278856422703e-05, "loss": 0.0278, "step": 62560 }, { "grad_norm": 0.3852863609790802, "learning_rate": 3.365716115946021e-05, "loss": 0.0324, "step": 62570 }, { "grad_norm": 0.52874356508255, "learning_rate": 3.3641535541921666e-05, "loss": 0.0477, "step": 62580 }, { "grad_norm": 0.2893478274345398, "learning_rate": 3.36259117133202e-05, "loss": 0.0276, "step": 62590 }, { "grad_norm": 0.25564125180244446, "learning_rate": 3.36102896753644e-05, "loss": 0.0284, "step": 62600 }, { "grad_norm": 0.4688190817832947, "learning_rate": 3.35946694297627e-05, "loss": 0.0312, "step": 62610 }, { "grad_norm": 0.36128973960876465, "learning_rate": 3.357905097822328e-05, "loss": 0.0377, "step": 62620 }, { "grad_norm": 0.33192482590675354, "learning_rate": 3.356343432245416e-05, "loss": 0.0292, "step": 62630 }, { "grad_norm": 0.6210857629776001, "learning_rate": 3.354781946416315e-05, "loss": 0.0482, "step": 62640 }, { "grad_norm": 0.40937986969947815, "learning_rate": 3.3532206405057864e-05, "loss": 0.0324, "step": 62650 }, { "grad_norm": 0.48687508702278137, "learning_rate": 3.351659514684573e-05, "loss": 0.0299, "step": 62660 }, { "grad_norm": 0.36610689759254456, "learning_rate": 3.350098569123395e-05, "loss": 0.029, "step": 62670 }, { "grad_norm": 0.3007352650165558, "learning_rate": 3.348537803992958e-05, "loss": 0.0263, "step": 62680 }, { "grad_norm": 0.44848939776420593, "learning_rate": 3.346977219463945e-05, "loss": 0.0326, "step": 62690 }, { "grad_norm": 0.3344963490962982, "learning_rate": 3.345416815707017e-05, "loss": 0.03, "step": 62700 }, { "grad_norm": 0.25005125999450684, "learning_rate": 3.3438565928928185e-05, "loss": 0.0299, "step": 62710 }, { "grad_norm": 0.32782286405563354, "learning_rate": 3.3422965511919736e-05, "loss": 0.0245, "step": 62720 }, { "grad_norm": 0.3231363594532013, "learning_rate": 3.3407366907750846e-05, "loss": 0.0311, "step": 62730 }, { "grad_norm": 0.38678547739982605, "learning_rate": 3.339177011812737e-05, "loss": 0.039, "step": 62740 }, { "grad_norm": 0.30302321910858154, "learning_rate": 3.3376175144754935e-05, "loss": 0.0302, "step": 62750 }, { "grad_norm": 0.36360156536102295, "learning_rate": 3.3360581989339004e-05, "loss": 0.0336, "step": 62760 }, { "grad_norm": 0.1978919804096222, "learning_rate": 3.3344990653584804e-05, "loss": 0.0308, "step": 62770 }, { "grad_norm": 0.2702568769454956, "learning_rate": 3.332940113919739e-05, "loss": 0.0297, "step": 62780 }, { "grad_norm": 0.3566810190677643, "learning_rate": 3.331381344788161e-05, "loss": 0.0235, "step": 62790 }, { "grad_norm": 0.34928154945373535, "learning_rate": 3.3298227581342076e-05, "loss": 0.0253, "step": 62800 }, { "grad_norm": 0.39490973949432373, "learning_rate": 3.328264354128329e-05, "loss": 0.0382, "step": 62810 }, { "grad_norm": 0.5678836107254028, "learning_rate": 3.3267061329409465e-05, "loss": 0.0346, "step": 62820 }, { "grad_norm": 0.30854544043540955, "learning_rate": 3.325148094742466e-05, "loss": 0.0257, "step": 62830 }, { "grad_norm": 0.2221442312002182, "learning_rate": 3.323590239703271e-05, "loss": 0.0246, "step": 62840 }, { "grad_norm": 0.5170252919197083, "learning_rate": 3.322032567993728e-05, "loss": 0.0356, "step": 62850 }, { "grad_norm": 0.3469603657722473, "learning_rate": 3.3204750797841774e-05, "loss": 0.0325, "step": 62860 }, { "grad_norm": 0.20267511904239655, "learning_rate": 3.31891777524495e-05, "loss": 0.03, "step": 62870 }, { "grad_norm": 0.49002477526664734, "learning_rate": 3.317360654546347e-05, "loss": 0.0355, "step": 62880 }, { "grad_norm": 0.3606656789779663, "learning_rate": 3.3158037178586534e-05, "loss": 0.0345, "step": 62890 }, { "grad_norm": 0.254271924495697, "learning_rate": 3.3142469653521316e-05, "loss": 0.0326, "step": 62900 }, { "grad_norm": 0.3111288547515869, "learning_rate": 3.3126903971970284e-05, "loss": 0.0254, "step": 62910 }, { "grad_norm": 0.3806755244731903, "learning_rate": 3.311134013563564e-05, "loss": 0.0267, "step": 62920 }, { "grad_norm": 0.5685235261917114, "learning_rate": 3.309577814621947e-05, "loss": 0.0267, "step": 62930 }, { "grad_norm": 0.256089448928833, "learning_rate": 3.308021800542359e-05, "loss": 0.0271, "step": 62940 }, { "grad_norm": 0.3165889382362366, "learning_rate": 3.306465971494963e-05, "loss": 0.0312, "step": 62950 }, { "grad_norm": 0.5403958559036255, "learning_rate": 3.304910327649901e-05, "loss": 0.0314, "step": 62960 }, { "grad_norm": 0.2937171459197998, "learning_rate": 3.303354869177297e-05, "loss": 0.0262, "step": 62970 }, { "grad_norm": 0.38649046421051025, "learning_rate": 3.301799596247255e-05, "loss": 0.0259, "step": 62980 }, { "grad_norm": 0.2501843571662903, "learning_rate": 3.300244509029854e-05, "loss": 0.023, "step": 62990 }, { "grad_norm": 0.2875378131866455, "learning_rate": 3.2986896076951585e-05, "loss": 0.0234, "step": 63000 }, { "grad_norm": 0.3416661024093628, "learning_rate": 3.297134892413211e-05, "loss": 0.035, "step": 63010 }, { "grad_norm": 0.371381938457489, "learning_rate": 3.295580363354031e-05, "loss": 0.0324, "step": 63020 }, { "grad_norm": 0.39266452193260193, "learning_rate": 3.2940260206876194e-05, "loss": 0.0228, "step": 63030 }, { "grad_norm": 0.25637707114219666, "learning_rate": 3.292471864583958e-05, "loss": 0.0298, "step": 63040 }, { "grad_norm": 0.25622323155403137, "learning_rate": 3.290917895213004e-05, "loss": 0.0303, "step": 63050 }, { "grad_norm": 0.3862360119819641, "learning_rate": 3.289364112744701e-05, "loss": 0.036, "step": 63060 }, { "grad_norm": 0.2573249936103821, "learning_rate": 3.287810517348967e-05, "loss": 0.0343, "step": 63070 }, { "grad_norm": 0.3582223653793335, "learning_rate": 3.2862571091957014e-05, "loss": 0.0268, "step": 63080 }, { "grad_norm": 0.3164142072200775, "learning_rate": 3.284703888454781e-05, "loss": 0.0277, "step": 63090 }, { "grad_norm": 0.430757999420166, "learning_rate": 3.2831508552960646e-05, "loss": 0.031, "step": 63100 }, { "grad_norm": 0.6293476819992065, "learning_rate": 3.281598009889387e-05, "loss": 0.0316, "step": 63110 }, { "grad_norm": 0.3363226056098938, "learning_rate": 3.28004535240457e-05, "loss": 0.0391, "step": 63120 }, { "grad_norm": 0.7363793849945068, "learning_rate": 3.278492883011407e-05, "loss": 0.0296, "step": 63130 }, { "grad_norm": 0.21725717186927795, "learning_rate": 3.276940601879675e-05, "loss": 0.0302, "step": 63140 }, { "grad_norm": 0.32412806153297424, "learning_rate": 3.275388509179127e-05, "loss": 0.0321, "step": 63150 }, { "grad_norm": 0.39497196674346924, "learning_rate": 3.2738366050794985e-05, "loss": 0.0354, "step": 63160 }, { "grad_norm": 0.32576480507850647, "learning_rate": 3.2722848897505034e-05, "loss": 0.0319, "step": 63170 }, { "grad_norm": 0.38291043043136597, "learning_rate": 3.2707333633618356e-05, "loss": 0.0251, "step": 63180 }, { "grad_norm": 0.3578503727912903, "learning_rate": 3.269182026083167e-05, "loss": 0.0255, "step": 63190 }, { "grad_norm": 0.3556709289550781, "learning_rate": 3.2676308780841516e-05, "loss": 0.0304, "step": 63200 }, { "grad_norm": 0.24952025711536407, "learning_rate": 3.266079919534416e-05, "loss": 0.0255, "step": 63210 }, { "grad_norm": 0.284212201833725, "learning_rate": 3.264529150603575e-05, "loss": 0.0233, "step": 63220 }, { "grad_norm": 0.27978742122650146, "learning_rate": 3.262978571461217e-05, "loss": 0.0259, "step": 63230 }, { "grad_norm": 0.3309791088104248, "learning_rate": 3.26142818227691e-05, "loss": 0.0282, "step": 63240 }, { "grad_norm": 0.43790870904922485, "learning_rate": 3.259877983220203e-05, "loss": 0.0382, "step": 63250 }, { "grad_norm": 0.2568603754043579, "learning_rate": 3.2583279744606246e-05, "loss": 0.0293, "step": 63260 }, { "grad_norm": 0.2097286581993103, "learning_rate": 3.25677815616768e-05, "loss": 0.0253, "step": 63270 }, { "grad_norm": 0.3425317108631134, "learning_rate": 3.255228528510856e-05, "loss": 0.0414, "step": 63280 }, { "grad_norm": 0.25218257308006287, "learning_rate": 3.253679091659617e-05, "loss": 0.0239, "step": 63290 }, { "grad_norm": 0.5076910257339478, "learning_rate": 3.252129845783404e-05, "loss": 0.0237, "step": 63300 }, { "grad_norm": 0.29133132100105286, "learning_rate": 3.2505807910516456e-05, "loss": 0.0226, "step": 63310 }, { "grad_norm": 0.5063866376876831, "learning_rate": 3.249031927633742e-05, "loss": 0.0415, "step": 63320 }, { "grad_norm": 0.27773380279541016, "learning_rate": 3.2474832556990737e-05, "loss": 0.0305, "step": 63330 }, { "grad_norm": 0.36121174693107605, "learning_rate": 3.245934775417002e-05, "loss": 0.0237, "step": 63340 }, { "grad_norm": 0.33714818954467773, "learning_rate": 3.244386486956866e-05, "loss": 0.0318, "step": 63350 }, { "grad_norm": 0.5655787587165833, "learning_rate": 3.2428383904879846e-05, "loss": 0.0272, "step": 63360 }, { "grad_norm": 0.33898618817329407, "learning_rate": 3.241290486179652e-05, "loss": 0.0447, "step": 63370 }, { "grad_norm": 0.4140028655529022, "learning_rate": 3.239742774201151e-05, "loss": 0.0221, "step": 63380 }, { "grad_norm": 0.31108173727989197, "learning_rate": 3.238195254721733e-05, "loss": 0.0276, "step": 63390 }, { "grad_norm": 0.37390679121017456, "learning_rate": 3.236647927910633e-05, "loss": 0.0334, "step": 63400 }, { "grad_norm": 0.2645473778247833, "learning_rate": 3.2351007939370646e-05, "loss": 0.0306, "step": 63410 }, { "grad_norm": 0.46495887637138367, "learning_rate": 3.2335538529702184e-05, "loss": 0.0341, "step": 63420 }, { "grad_norm": 0.2913118302822113, "learning_rate": 3.232007105179268e-05, "loss": 0.0284, "step": 63430 }, { "grad_norm": 0.31921160221099854, "learning_rate": 3.230460550733363e-05, "loss": 0.0228, "step": 63440 }, { "grad_norm": 0.17631895840168, "learning_rate": 3.2289141898016315e-05, "loss": 0.0258, "step": 63450 }, { "grad_norm": 0.3125106394290924, "learning_rate": 3.227368022553181e-05, "loss": 0.0258, "step": 63460 }, { "grad_norm": 0.37501654028892517, "learning_rate": 3.225822049157099e-05, "loss": 0.0342, "step": 63470 }, { "grad_norm": 0.3877812922000885, "learning_rate": 3.2242762697824494e-05, "loss": 0.0242, "step": 63480 }, { "grad_norm": 0.2998082637786865, "learning_rate": 3.222730684598279e-05, "loss": 0.0362, "step": 63490 }, { "grad_norm": 0.3189321458339691, "learning_rate": 3.221185293773605e-05, "loss": 0.0262, "step": 63500 }, { "grad_norm": 0.5802356004714966, "learning_rate": 3.2196400974774354e-05, "loss": 0.0285, "step": 63510 }, { "grad_norm": 0.2682192027568817, "learning_rate": 3.2180950958787485e-05, "loss": 0.0387, "step": 63520 }, { "grad_norm": 0.22527384757995605, "learning_rate": 3.2165502891465024e-05, "loss": 0.0275, "step": 63530 }, { "grad_norm": 0.8180900812149048, "learning_rate": 3.215005677449636e-05, "loss": 0.034, "step": 63540 }, { "grad_norm": 0.431111603975296, "learning_rate": 3.2134612609570644e-05, "loss": 0.0212, "step": 63550 }, { "grad_norm": 0.3672720193862915, "learning_rate": 3.211917039837682e-05, "loss": 0.0248, "step": 63560 }, { "grad_norm": 0.26075613498687744, "learning_rate": 3.210373014260365e-05, "loss": 0.0373, "step": 63570 }, { "grad_norm": 0.36653780937194824, "learning_rate": 3.208829184393965e-05, "loss": 0.026, "step": 63580 }, { "grad_norm": 0.32095712423324585, "learning_rate": 3.207285550407312e-05, "loss": 0.0309, "step": 63590 }, { "grad_norm": 0.3002324402332306, "learning_rate": 3.205742112469217e-05, "loss": 0.0253, "step": 63600 }, { "grad_norm": 0.471965491771698, "learning_rate": 3.204198870748465e-05, "loss": 0.0399, "step": 63610 }, { "grad_norm": 0.5156078934669495, "learning_rate": 3.202655825413824e-05, "loss": 0.0321, "step": 63620 }, { "grad_norm": 0.3320641815662384, "learning_rate": 3.2011129766340416e-05, "loss": 0.0377, "step": 63630 }, { "grad_norm": 0.209800124168396, "learning_rate": 3.1995703245778384e-05, "loss": 0.0188, "step": 63640 }, { "grad_norm": 0.2964019775390625, "learning_rate": 3.1980278694139186e-05, "loss": 0.0276, "step": 63650 }, { "grad_norm": 0.27519696950912476, "learning_rate": 3.196485611310962e-05, "loss": 0.028, "step": 63660 }, { "grad_norm": 0.4669572412967682, "learning_rate": 3.1949435504376257e-05, "loss": 0.0303, "step": 63670 }, { "grad_norm": 0.472700297832489, "learning_rate": 3.1934016869625505e-05, "loss": 0.0242, "step": 63680 }, { "grad_norm": 0.5104279518127441, "learning_rate": 3.191860021054349e-05, "loss": 0.039, "step": 63690 }, { "grad_norm": 0.38019105792045593, "learning_rate": 3.190318552881618e-05, "loss": 0.0431, "step": 63700 }, { "grad_norm": 0.47321346402168274, "learning_rate": 3.1887772826129285e-05, "loss": 0.0268, "step": 63710 }, { "grad_norm": 0.2630893886089325, "learning_rate": 3.1872362104168335e-05, "loss": 0.0351, "step": 63720 }, { "grad_norm": 0.341655433177948, "learning_rate": 3.185695336461861e-05, "loss": 0.0292, "step": 63730 }, { "grad_norm": 0.30341866612434387, "learning_rate": 3.184154660916519e-05, "loss": 0.0216, "step": 63740 }, { "grad_norm": 0.35776135325431824, "learning_rate": 3.1826141839492906e-05, "loss": 0.0252, "step": 63750 }, { "grad_norm": 0.3964471220970154, "learning_rate": 3.181073905728644e-05, "loss": 0.0261, "step": 63760 }, { "grad_norm": 0.46983593702316284, "learning_rate": 3.1795338264230214e-05, "loss": 0.0218, "step": 63770 }, { "grad_norm": 0.3575960099697113, "learning_rate": 3.1779939462008426e-05, "loss": 0.0299, "step": 63780 }, { "grad_norm": 0.3862001895904541, "learning_rate": 3.1764542652305066e-05, "loss": 0.0238, "step": 63790 }, { "grad_norm": 0.32332491874694824, "learning_rate": 3.17491478368039e-05, "loss": 0.0297, "step": 63800 }, { "grad_norm": 0.44581711292266846, "learning_rate": 3.173375501718847e-05, "loss": 0.0308, "step": 63810 }, { "grad_norm": 0.42191776633262634, "learning_rate": 3.171836419514215e-05, "loss": 0.0236, "step": 63820 }, { "grad_norm": 0.4223612844944, "learning_rate": 3.170297537234803e-05, "loss": 0.0346, "step": 63830 }, { "grad_norm": 0.24901865422725677, "learning_rate": 3.168758855048901e-05, "loss": 0.0271, "step": 63840 }, { "grad_norm": 0.28720712661743164, "learning_rate": 3.1672203731247785e-05, "loss": 0.0227, "step": 63850 }, { "grad_norm": 0.3757151961326599, "learning_rate": 3.16568209163068e-05, "loss": 0.032, "step": 63860 }, { "grad_norm": 0.45148205757141113, "learning_rate": 3.1641440107348295e-05, "loss": 0.0303, "step": 63870 }, { "grad_norm": 0.3515283465385437, "learning_rate": 3.162606130605429e-05, "loss": 0.0245, "step": 63880 }, { "grad_norm": 0.26665258407592773, "learning_rate": 3.16106845141066e-05, "loss": 0.0236, "step": 63890 }, { "grad_norm": 0.3401682376861572, "learning_rate": 3.159530973318682e-05, "loss": 0.0227, "step": 63900 }, { "grad_norm": 0.26737573742866516, "learning_rate": 3.1579936964976286e-05, "loss": 0.0272, "step": 63910 }, { "grad_norm": 0.33238932490348816, "learning_rate": 3.156456621115615e-05, "loss": 0.0286, "step": 63920 }, { "grad_norm": 0.6517450213432312, "learning_rate": 3.154919747340734e-05, "loss": 0.0331, "step": 63930 }, { "grad_norm": 0.5690298080444336, "learning_rate": 3.153383075341054e-05, "loss": 0.0269, "step": 63940 }, { "grad_norm": 0.3524770438671112, "learning_rate": 3.1518466052846264e-05, "loss": 0.0297, "step": 63950 }, { "grad_norm": 0.25777411460876465, "learning_rate": 3.150310337339474e-05, "loss": 0.0271, "step": 63960 }, { "grad_norm": 0.31638097763061523, "learning_rate": 3.148774271673603e-05, "loss": 0.0298, "step": 63970 }, { "grad_norm": 0.2798644006252289, "learning_rate": 3.147238408454995e-05, "loss": 0.0377, "step": 63980 }, { "grad_norm": 0.2075885832309723, "learning_rate": 3.1457027478516086e-05, "loss": 0.023, "step": 63990 }, { "grad_norm": 0.2466551959514618, "learning_rate": 3.144167290031382e-05, "loss": 0.0275, "step": 64000 }, { "grad_norm": 0.3573835492134094, "learning_rate": 3.142632035162227e-05, "loss": 0.0334, "step": 64010 }, { "grad_norm": 0.28397634625434875, "learning_rate": 3.141096983412044e-05, "loss": 0.0232, "step": 64020 }, { "grad_norm": 0.18840327858924866, "learning_rate": 3.139562134948699e-05, "loss": 0.027, "step": 64030 }, { "grad_norm": 0.3912035822868347, "learning_rate": 3.1380274899400424e-05, "loss": 0.0268, "step": 64040 }, { "grad_norm": 0.3569907546043396, "learning_rate": 3.1364930485539e-05, "loss": 0.0214, "step": 64050 }, { "grad_norm": 0.25717124342918396, "learning_rate": 3.134958810958076e-05, "loss": 0.0234, "step": 64060 }, { "grad_norm": 0.307767391204834, "learning_rate": 3.13342477732035e-05, "loss": 0.0289, "step": 64070 }, { "grad_norm": 0.25882187485694885, "learning_rate": 3.131890947808487e-05, "loss": 0.0234, "step": 64080 }, { "grad_norm": 0.3581241965293884, "learning_rate": 3.13035732259022e-05, "loss": 0.0276, "step": 64090 }, { "grad_norm": 0.2699595093727112, "learning_rate": 3.1288239018332656e-05, "loss": 0.0243, "step": 64100 }, { "grad_norm": 0.33140474557876587, "learning_rate": 3.1272906857053164e-05, "loss": 0.023, "step": 64110 }, { "grad_norm": 0.4122059643268585, "learning_rate": 3.125757674374042e-05, "loss": 0.0282, "step": 64120 }, { "grad_norm": 0.3029000759124756, "learning_rate": 3.124224868007088e-05, "loss": 0.0316, "step": 64130 }, { "grad_norm": 0.2949310839176178, "learning_rate": 3.1226922667720854e-05, "loss": 0.0288, "step": 64140 }, { "grad_norm": 0.5673022270202637, "learning_rate": 3.121159870836633e-05, "loss": 0.0337, "step": 64150 }, { "grad_norm": 0.23336179554462433, "learning_rate": 3.119627680368312e-05, "loss": 0.0262, "step": 64160 }, { "grad_norm": 0.41896775364875793, "learning_rate": 3.1180956955346816e-05, "loss": 0.0291, "step": 64170 }, { "grad_norm": 0.25113704800605774, "learning_rate": 3.116563916503276e-05, "loss": 0.03, "step": 64180 }, { "grad_norm": 0.19352558255195618, "learning_rate": 3.1150323434416095e-05, "loss": 0.039, "step": 64190 }, { "grad_norm": 0.3015700578689575, "learning_rate": 3.1135009765171696e-05, "loss": 0.0259, "step": 64200 }, { "grad_norm": 0.4631514549255371, "learning_rate": 3.111969815897429e-05, "loss": 0.0256, "step": 64210 }, { "grad_norm": 0.2827182114124298, "learning_rate": 3.1104388617498295e-05, "loss": 0.0349, "step": 64220 }, { "grad_norm": 0.28223833441734314, "learning_rate": 3.1089081142417966e-05, "loss": 0.0316, "step": 64230 }, { "grad_norm": 0.270506888628006, "learning_rate": 3.1073775735407285e-05, "loss": 0.0259, "step": 64240 }, { "grad_norm": 0.38864997029304504, "learning_rate": 3.1058472398140035e-05, "loss": 0.0315, "step": 64250 }, { "grad_norm": 0.37090393900871277, "learning_rate": 3.1043171132289736e-05, "loss": 0.0268, "step": 64260 }, { "grad_norm": 0.2692926526069641, "learning_rate": 3.1027871939529774e-05, "loss": 0.0195, "step": 64270 }, { "grad_norm": 0.48363175988197327, "learning_rate": 3.1012574821533214e-05, "loss": 0.0447, "step": 64280 }, { "grad_norm": 0.5375329852104187, "learning_rate": 3.099727977997291e-05, "loss": 0.0337, "step": 64290 }, { "grad_norm": 0.20586957037448883, "learning_rate": 3.098198681652154e-05, "loss": 0.0386, "step": 64300 }, { "grad_norm": 0.27607494592666626, "learning_rate": 3.0966695932851475e-05, "loss": 0.021, "step": 64310 }, { "grad_norm": 0.3593248128890991, "learning_rate": 3.0951407130634926e-05, "loss": 0.0335, "step": 64320 }, { "grad_norm": 0.374803364276886, "learning_rate": 3.0936120411543864e-05, "loss": 0.0265, "step": 64330 }, { "grad_norm": 0.46622493863105774, "learning_rate": 3.0920835777250015e-05, "loss": 0.0287, "step": 64340 }, { "grad_norm": 0.41805312037467957, "learning_rate": 3.090555322942488e-05, "loss": 0.0251, "step": 64350 }, { "grad_norm": 0.2934727072715759, "learning_rate": 3.089027276973973e-05, "loss": 0.0313, "step": 64360 }, { "grad_norm": 0.2916317284107208, "learning_rate": 3.0874994399865615e-05, "loss": 0.0287, "step": 64370 }, { "grad_norm": 0.4196547865867615, "learning_rate": 3.085971812147337e-05, "loss": 0.0311, "step": 64380 }, { "grad_norm": 0.26221907138824463, "learning_rate": 3.0844443936233557e-05, "loss": 0.0335, "step": 64390 }, { "grad_norm": 0.42216137051582336, "learning_rate": 3.082917184581657e-05, "loss": 0.024, "step": 64400 }, { "grad_norm": 0.388926237821579, "learning_rate": 3.0813901851892513e-05, "loss": 0.0283, "step": 64410 }, { "grad_norm": 0.2273581624031067, "learning_rate": 3.0798633956131316e-05, "loss": 0.0272, "step": 64420 }, { "grad_norm": 0.13983802497386932, "learning_rate": 3.078336816020264e-05, "loss": 0.0227, "step": 64430 }, { "grad_norm": 0.2725905478000641, "learning_rate": 3.0768104465775935e-05, "loss": 0.0287, "step": 64440 }, { "grad_norm": 0.41892683506011963, "learning_rate": 3.0752842874520396e-05, "loss": 0.0254, "step": 64450 }, { "grad_norm": 0.40046054124832153, "learning_rate": 3.073758338810503e-05, "loss": 0.0236, "step": 64460 }, { "grad_norm": 0.36355748772621155, "learning_rate": 3.072232600819859e-05, "loss": 0.0312, "step": 64470 }, { "grad_norm": 0.2750340402126312, "learning_rate": 3.070707073646959e-05, "loss": 0.0352, "step": 64480 }, { "grad_norm": 0.32972073554992676, "learning_rate": 3.069181757458633e-05, "loss": 0.0348, "step": 64490 }, { "grad_norm": 0.819322943687439, "learning_rate": 3.067656652421687e-05, "loss": 0.0286, "step": 64500 }, { "grad_norm": 0.21806766092777252, "learning_rate": 3.0661317587029025e-05, "loss": 0.0276, "step": 64510 }, { "grad_norm": 0.3313930034637451, "learning_rate": 3.064607076469043e-05, "loss": 0.0195, "step": 64520 }, { "grad_norm": 0.3788222670555115, "learning_rate": 3.063082605886845e-05, "loss": 0.0314, "step": 64530 }, { "grad_norm": 0.264599472284317, "learning_rate": 3.061558347123019e-05, "loss": 0.035, "step": 64540 }, { "grad_norm": 0.3510153293609619, "learning_rate": 3.060034300344259e-05, "loss": 0.0264, "step": 64550 }, { "grad_norm": 1.238633155822754, "learning_rate": 3.058510465717231e-05, "loss": 0.037, "step": 64560 }, { "grad_norm": 0.27419763803482056, "learning_rate": 3.0569868434085804e-05, "loss": 0.024, "step": 64570 }, { "grad_norm": 0.29579246044158936, "learning_rate": 3.055463433584924e-05, "loss": 0.0199, "step": 64580 }, { "grad_norm": 0.36034953594207764, "learning_rate": 3.0539402364128656e-05, "loss": 0.0248, "step": 64590 }, { "grad_norm": 0.2574828267097473, "learning_rate": 3.052417252058977e-05, "loss": 0.034, "step": 64600 }, { "grad_norm": 0.4123998284339905, "learning_rate": 3.0508944806898092e-05, "loss": 0.0383, "step": 64610 }, { "grad_norm": 0.32162734866142273, "learning_rate": 3.04937192247189e-05, "loss": 0.0306, "step": 64620 }, { "grad_norm": 0.3285391926765442, "learning_rate": 3.047849577571725e-05, "loss": 0.0346, "step": 64630 }, { "grad_norm": 0.3742583692073822, "learning_rate": 3.046327446155795e-05, "loss": 0.0224, "step": 64640 }, { "grad_norm": 0.27263695001602173, "learning_rate": 3.0448055283905587e-05, "loss": 0.0224, "step": 64650 }, { "grad_norm": 0.2286088764667511, "learning_rate": 3.0432838244424494e-05, "loss": 0.0231, "step": 64660 }, { "grad_norm": 0.2204466611146927, "learning_rate": 3.04176233447788e-05, "loss": 0.0214, "step": 64670 }, { "grad_norm": 0.24769169092178345, "learning_rate": 3.040241058663238e-05, "loss": 0.0219, "step": 64680 }, { "grad_norm": 0.3839503824710846, "learning_rate": 3.038719997164887e-05, "loss": 0.0339, "step": 64690 }, { "grad_norm": 0.21516664326190948, "learning_rate": 3.0371991501491683e-05, "loss": 0.0291, "step": 64700 }, { "grad_norm": 0.34007084369659424, "learning_rate": 3.0356785177823975e-05, "loss": 0.0317, "step": 64710 }, { "grad_norm": 0.19802060723304749, "learning_rate": 3.0341581002308726e-05, "loss": 0.0296, "step": 64720 }, { "grad_norm": 0.4295423626899719, "learning_rate": 3.0326378976608627e-05, "loss": 0.0254, "step": 64730 }, { "grad_norm": 0.43423959612846375, "learning_rate": 3.0311179102386146e-05, "loss": 0.0368, "step": 64740 }, { "grad_norm": 0.4855179786682129, "learning_rate": 3.0295981381303513e-05, "loss": 0.0369, "step": 64750 }, { "grad_norm": 0.32598671317100525, "learning_rate": 3.0280785815022727e-05, "loss": 0.0265, "step": 64760 }, { "grad_norm": 0.28645622730255127, "learning_rate": 3.0265592405205544e-05, "loss": 0.029, "step": 64770 }, { "grad_norm": 0.44527679681777954, "learning_rate": 3.0250401153513517e-05, "loss": 0.0341, "step": 64780 }, { "grad_norm": 0.22451628744602203, "learning_rate": 3.0235212061607936e-05, "loss": 0.0365, "step": 64790 }, { "grad_norm": 0.17354638874530792, "learning_rate": 3.0220025131149842e-05, "loss": 0.0282, "step": 64800 }, { "grad_norm": 0.19717776775360107, "learning_rate": 3.020484036380007e-05, "loss": 0.0227, "step": 64810 }, { "grad_norm": 0.2562200129032135, "learning_rate": 3.0189657761219188e-05, "loss": 0.0289, "step": 64820 }, { "grad_norm": 0.1862948089838028, "learning_rate": 3.017447732506753e-05, "loss": 0.0267, "step": 64830 }, { "grad_norm": 0.3350410759449005, "learning_rate": 3.015929905700524e-05, "loss": 0.0211, "step": 64840 }, { "grad_norm": 0.2717316746711731, "learning_rate": 3.014412295869218e-05, "loss": 0.0258, "step": 64850 }, { "grad_norm": 0.4572960436344147, "learning_rate": 3.0128949031787977e-05, "loss": 0.0348, "step": 64860 }, { "grad_norm": 0.2052997201681137, "learning_rate": 3.0113777277952022e-05, "loss": 0.0377, "step": 64870 }, { "grad_norm": 0.4924026429653168, "learning_rate": 3.0098607698843496e-05, "loss": 0.0297, "step": 64880 }, { "grad_norm": 0.42907169461250305, "learning_rate": 3.0083440296121312e-05, "loss": 0.0299, "step": 64890 }, { "grad_norm": 0.29807138442993164, "learning_rate": 3.0068275071444142e-05, "loss": 0.0229, "step": 64900 }, { "grad_norm": 0.4155994951725006, "learning_rate": 3.0053112026470442e-05, "loss": 0.0338, "step": 64910 }, { "grad_norm": 0.4341164827346802, "learning_rate": 3.0037951162858434e-05, "loss": 0.0307, "step": 64920 }, { "grad_norm": 0.3196598291397095, "learning_rate": 3.002279248226607e-05, "loss": 0.0277, "step": 64930 }, { "grad_norm": 0.24482594430446625, "learning_rate": 3.000763598635109e-05, "loss": 0.0325, "step": 64940 }, { "grad_norm": 0.47497597336769104, "learning_rate": 2.999248167677098e-05, "loss": 0.0312, "step": 64950 }, { "grad_norm": 0.3296690881252289, "learning_rate": 2.9977329555182976e-05, "loss": 0.0361, "step": 64960 }, { "grad_norm": 0.3262042999267578, "learning_rate": 2.9962179623244124e-05, "loss": 0.0273, "step": 64970 }, { "grad_norm": 0.33438563346862793, "learning_rate": 2.9947031882611175e-05, "loss": 0.028, "step": 64980 }, { "grad_norm": 0.39391589164733887, "learning_rate": 2.9931886334940685e-05, "loss": 0.0262, "step": 64990 }, { "grad_norm": 0.17924346029758453, "learning_rate": 2.9916742981888923e-05, "loss": 0.025, "step": 65000 }, { "grad_norm": 0.3731049597263336, "learning_rate": 2.990160182511196e-05, "loss": 0.0377, "step": 65010 }, { "grad_norm": 0.3157525360584259, "learning_rate": 2.9886462866265574e-05, "loss": 0.0278, "step": 65020 }, { "grad_norm": 0.3496093451976776, "learning_rate": 2.9871326107005392e-05, "loss": 0.0292, "step": 65030 }, { "grad_norm": 0.3054784834384918, "learning_rate": 2.9856191548986735e-05, "loss": 0.0314, "step": 65040 }, { "grad_norm": 0.3015280067920685, "learning_rate": 2.9841059193864672e-05, "loss": 0.0264, "step": 65050 }, { "grad_norm": 0.3401598632335663, "learning_rate": 2.982592904329407e-05, "loss": 0.0262, "step": 65060 }, { "grad_norm": 0.17781580984592438, "learning_rate": 2.981080109892954e-05, "loss": 0.0202, "step": 65070 }, { "grad_norm": 0.23065930604934692, "learning_rate": 2.9795675362425434e-05, "loss": 0.0233, "step": 65080 }, { "grad_norm": 0.2601158320903778, "learning_rate": 2.97805518354359e-05, "loss": 0.0206, "step": 65090 }, { "grad_norm": 0.3059799075126648, "learning_rate": 2.9765430519614825e-05, "loss": 0.0235, "step": 65100 }, { "grad_norm": 0.3083411157131195, "learning_rate": 2.975031141661584e-05, "loss": 0.0293, "step": 65110 }, { "grad_norm": 0.27924612164497375, "learning_rate": 2.9735194528092347e-05, "loss": 0.026, "step": 65120 }, { "grad_norm": 0.3119371235370636, "learning_rate": 2.9720079855697525e-05, "loss": 0.0336, "step": 65130 }, { "grad_norm": 0.43941590189933777, "learning_rate": 2.970496740108427e-05, "loss": 0.0365, "step": 65140 }, { "grad_norm": 0.320499986410141, "learning_rate": 2.968985716590527e-05, "loss": 0.0251, "step": 65150 }, { "grad_norm": 0.4229027032852173, "learning_rate": 2.9674749151812953e-05, "loss": 0.0259, "step": 65160 }, { "grad_norm": 0.48822176456451416, "learning_rate": 2.965964336045951e-05, "loss": 0.0255, "step": 65170 }, { "grad_norm": 0.30647456645965576, "learning_rate": 2.9644539793496894e-05, "loss": 0.0263, "step": 65180 }, { "grad_norm": 0.5123183131217957, "learning_rate": 2.9629438452576807e-05, "loss": 0.0361, "step": 65190 }, { "grad_norm": 0.3890529274940491, "learning_rate": 2.9614339339350698e-05, "loss": 0.0268, "step": 65200 }, { "grad_norm": 0.2357991486787796, "learning_rate": 2.959924245546979e-05, "loss": 0.0252, "step": 65210 }, { "grad_norm": 0.4189581573009491, "learning_rate": 2.9584147802585038e-05, "loss": 0.0347, "step": 65220 }, { "grad_norm": 0.35476210713386536, "learning_rate": 2.956905538234721e-05, "loss": 0.0318, "step": 65230 }, { "grad_norm": 0.23682892322540283, "learning_rate": 2.9553965196406764e-05, "loss": 0.0256, "step": 65240 }, { "grad_norm": 0.2735477387905121, "learning_rate": 2.9538877246413943e-05, "loss": 0.0276, "step": 65250 }, { "grad_norm": 0.34788188338279724, "learning_rate": 2.952379153401874e-05, "loss": 0.0298, "step": 65260 }, { "grad_norm": 0.3520892262458801, "learning_rate": 2.9508708060870914e-05, "loss": 0.0277, "step": 65270 }, { "grad_norm": 0.2990235686302185, "learning_rate": 2.9493626828619948e-05, "loss": 0.028, "step": 65280 }, { "grad_norm": 0.29395386576652527, "learning_rate": 2.9478547838915137e-05, "loss": 0.0242, "step": 65290 }, { "grad_norm": 0.476325660943985, "learning_rate": 2.946347109340548e-05, "loss": 0.0297, "step": 65300 }, { "grad_norm": 0.4763106107711792, "learning_rate": 2.944839659373976e-05, "loss": 0.0272, "step": 65310 }, { "grad_norm": 0.46231991052627563, "learning_rate": 2.943332434156647e-05, "loss": 0.0315, "step": 65320 }, { "grad_norm": 0.4048781394958496, "learning_rate": 2.9418254338533912e-05, "loss": 0.033, "step": 65330 }, { "grad_norm": 0.36073148250579834, "learning_rate": 2.940318658629011e-05, "loss": 0.0282, "step": 65340 }, { "grad_norm": 0.4074938893318176, "learning_rate": 2.938812108648287e-05, "loss": 0.0325, "step": 65350 }, { "grad_norm": 0.3009926974773407, "learning_rate": 2.937305784075972e-05, "loss": 0.024, "step": 65360 }, { "grad_norm": 0.5099956393241882, "learning_rate": 2.9357996850767937e-05, "loss": 0.0242, "step": 65370 }, { "grad_norm": 0.33639997243881226, "learning_rate": 2.9342938118154595e-05, "loss": 0.0229, "step": 65380 }, { "grad_norm": 0.2822803556919098, "learning_rate": 2.9327881644566485e-05, "loss": 0.0424, "step": 65390 }, { "grad_norm": 0.35654696822166443, "learning_rate": 2.9312827431650157e-05, "loss": 0.0326, "step": 65400 }, { "grad_norm": 0.20195478200912476, "learning_rate": 2.92977754810519e-05, "loss": 0.0381, "step": 65410 }, { "grad_norm": 0.30786776542663574, "learning_rate": 2.928272579441781e-05, "loss": 0.0296, "step": 65420 }, { "grad_norm": 0.26301926374435425, "learning_rate": 2.9267678373393676e-05, "loss": 0.0285, "step": 65430 }, { "grad_norm": 0.2659391462802887, "learning_rate": 2.925263321962507e-05, "loss": 0.0311, "step": 65440 }, { "grad_norm": 0.5319299101829529, "learning_rate": 2.9237590334757292e-05, "loss": 0.0312, "step": 65450 }, { "grad_norm": 0.36321982741355896, "learning_rate": 2.922254972043543e-05, "loss": 0.0297, "step": 65460 }, { "grad_norm": 0.2633223235607147, "learning_rate": 2.9207511378304265e-05, "loss": 0.0271, "step": 65470 }, { "grad_norm": 0.40097203850746155, "learning_rate": 2.9192475310008428e-05, "loss": 0.0364, "step": 65480 }, { "grad_norm": 1.4654173851013184, "learning_rate": 2.917744151719219e-05, "loss": 0.0216, "step": 65490 }, { "grad_norm": 0.4592907428741455, "learning_rate": 2.916241000149964e-05, "loss": 0.0264, "step": 65500 }, { "grad_norm": 0.3407859206199646, "learning_rate": 2.9147380764574622e-05, "loss": 0.0262, "step": 65510 }, { "grad_norm": 0.39170950651168823, "learning_rate": 2.9132353808060674e-05, "loss": 0.0367, "step": 65520 }, { "grad_norm": 0.27624672651290894, "learning_rate": 2.9117329133601134e-05, "loss": 0.0274, "step": 65530 }, { "grad_norm": 0.3072448670864105, "learning_rate": 2.9102306742839093e-05, "loss": 0.0385, "step": 65540 }, { "grad_norm": 0.2816677391529083, "learning_rate": 2.9087286637417375e-05, "loss": 0.0273, "step": 65550 }, { "grad_norm": 0.2765074074268341, "learning_rate": 2.9072268818978537e-05, "loss": 0.0362, "step": 65560 }, { "grad_norm": 0.35091057419776917, "learning_rate": 2.905725328916493e-05, "loss": 0.0259, "step": 65570 }, { "grad_norm": 0.44280296564102173, "learning_rate": 2.9042240049618608e-05, "loss": 0.0331, "step": 65580 }, { "grad_norm": 0.46865004301071167, "learning_rate": 2.9027229101981413e-05, "loss": 0.0288, "step": 65590 }, { "grad_norm": 0.28134629130363464, "learning_rate": 2.901222044789488e-05, "loss": 0.0288, "step": 65600 }, { "grad_norm": 0.48651784658432007, "learning_rate": 2.89972140890004e-05, "loss": 0.0288, "step": 65610 }, { "grad_norm": 0.4148595333099365, "learning_rate": 2.8982210026938988e-05, "loss": 0.0258, "step": 65620 }, { "grad_norm": 0.558962881565094, "learning_rate": 2.896720826335151e-05, "loss": 0.0409, "step": 65630 }, { "grad_norm": 0.450744092464447, "learning_rate": 2.8952208799878493e-05, "loss": 0.0314, "step": 65640 }, { "grad_norm": 0.4497852325439453, "learning_rate": 2.89372116381603e-05, "loss": 0.0236, "step": 65650 }, { "grad_norm": 0.19686861336231232, "learning_rate": 2.8922216779836932e-05, "loss": 0.0248, "step": 65660 }, { "grad_norm": 0.370683491230011, "learning_rate": 2.8907224226548285e-05, "loss": 0.0205, "step": 65670 }, { "grad_norm": 0.45383256673812866, "learning_rate": 2.8892233979933868e-05, "loss": 0.0279, "step": 65680 }, { "grad_norm": 0.45532315969467163, "learning_rate": 2.8877246041633026e-05, "loss": 0.0336, "step": 65690 }, { "grad_norm": 0.388641357421875, "learning_rate": 2.886226041328478e-05, "loss": 0.0206, "step": 65700 }, { "grad_norm": 0.20317815244197845, "learning_rate": 2.8847277096527957e-05, "loss": 0.0261, "step": 65710 }, { "grad_norm": 0.30129674077033997, "learning_rate": 2.8832296093001105e-05, "loss": 0.0213, "step": 65720 }, { "grad_norm": 0.24224045872688293, "learning_rate": 2.8817317404342547e-05, "loss": 0.0252, "step": 65730 }, { "grad_norm": 0.33159247040748596, "learning_rate": 2.880234103219029e-05, "loss": 0.0308, "step": 65740 }, { "grad_norm": 0.3321278393268585, "learning_rate": 2.878736697818215e-05, "loss": 0.0291, "step": 65750 }, { "grad_norm": 0.8225778341293335, "learning_rate": 2.877239524395568e-05, "loss": 0.0274, "step": 65760 }, { "grad_norm": 0.21855928003787994, "learning_rate": 2.8757425831148134e-05, "loss": 0.027, "step": 65770 }, { "grad_norm": 0.3507097065448761, "learning_rate": 2.874245874139658e-05, "loss": 0.0279, "step": 65780 }, { "grad_norm": 0.2098967283964157, "learning_rate": 2.8727493976337737e-05, "loss": 0.0226, "step": 65790 }, { "grad_norm": 0.4175151586532593, "learning_rate": 2.8712531537608218e-05, "loss": 0.0241, "step": 65800 }, { "grad_norm": 0.2953610122203827, "learning_rate": 2.8697571426844217e-05, "loss": 0.034, "step": 65810 }, { "grad_norm": 0.38052406907081604, "learning_rate": 2.86826136456818e-05, "loss": 0.0286, "step": 65820 }, { "grad_norm": 0.3701174557209015, "learning_rate": 2.8667658195756685e-05, "loss": 0.0329, "step": 65830 }, { "grad_norm": 0.421495258808136, "learning_rate": 2.8652705078704412e-05, "loss": 0.0307, "step": 65840 }, { "grad_norm": 0.40847617387771606, "learning_rate": 2.863775429616018e-05, "loss": 0.0313, "step": 65850 }, { "grad_norm": 0.3436441719532013, "learning_rate": 2.862280584975906e-05, "loss": 0.0253, "step": 65860 }, { "grad_norm": 0.5330175161361694, "learning_rate": 2.8607859741135735e-05, "loss": 0.0341, "step": 65870 }, { "grad_norm": 0.24554775655269623, "learning_rate": 2.8592915971924723e-05, "loss": 0.0377, "step": 65880 }, { "grad_norm": 0.3884950280189514, "learning_rate": 2.8577974543760212e-05, "loss": 0.0276, "step": 65890 }, { "grad_norm": 0.5223835110664368, "learning_rate": 2.8563035458276222e-05, "loss": 0.0318, "step": 65900 }, { "grad_norm": 0.25340673327445984, "learning_rate": 2.8548098717106424e-05, "loss": 0.0294, "step": 65910 }, { "grad_norm": 0.27524638175964355, "learning_rate": 2.85331643218843e-05, "loss": 0.0268, "step": 65920 }, { "grad_norm": 0.33696770668029785, "learning_rate": 2.8518232274243046e-05, "loss": 0.0336, "step": 65930 }, { "grad_norm": 0.4125248193740845, "learning_rate": 2.8503302575815638e-05, "loss": 0.0397, "step": 65940 }, { "grad_norm": 0.5513928532600403, "learning_rate": 2.8488375228234716e-05, "loss": 0.0341, "step": 65950 }, { "grad_norm": 0.35587403178215027, "learning_rate": 2.8473450233132738e-05, "loss": 0.0346, "step": 65960 }, { "grad_norm": 0.2539452314376831, "learning_rate": 2.8458527592141898e-05, "loss": 0.03, "step": 65970 }, { "grad_norm": 0.2711867392063141, "learning_rate": 2.8443607306894073e-05, "loss": 0.0328, "step": 65980 }, { "grad_norm": 0.2180180698633194, "learning_rate": 2.8428689379020946e-05, "loss": 0.0269, "step": 65990 }, { "grad_norm": 0.33987948298454285, "learning_rate": 2.841377381015391e-05, "loss": 0.0288, "step": 66000 }, { "grad_norm": 0.28267955780029297, "learning_rate": 2.8398860601924144e-05, "loss": 0.0216, "step": 66010 }, { "grad_norm": 0.24514535069465637, "learning_rate": 2.838394975596248e-05, "loss": 0.0264, "step": 66020 }, { "grad_norm": 0.37663915753364563, "learning_rate": 2.83690412738996e-05, "loss": 0.0322, "step": 66030 }, { "grad_norm": 0.3151569962501526, "learning_rate": 2.835413515736579e-05, "loss": 0.0343, "step": 66040 }, { "grad_norm": 0.3147035539150238, "learning_rate": 2.8339231407991267e-05, "loss": 0.0256, "step": 66050 }, { "grad_norm": 0.54906165599823, "learning_rate": 2.832433002740581e-05, "loss": 0.0243, "step": 66060 }, { "grad_norm": 0.2653912603855133, "learning_rate": 2.8309431017239047e-05, "loss": 0.0236, "step": 66070 }, { "grad_norm": 0.2911567986011505, "learning_rate": 2.829453437912028e-05, "loss": 0.0318, "step": 66080 }, { "grad_norm": 0.484482079744339, "learning_rate": 2.8279640114678617e-05, "loss": 0.027, "step": 66090 }, { "grad_norm": 0.4838729202747345, "learning_rate": 2.8264748225542836e-05, "loss": 0.0343, "step": 66100 }, { "grad_norm": 0.23308111727237701, "learning_rate": 2.8249858713341516e-05, "loss": 0.0243, "step": 66110 }, { "grad_norm": 0.27365952730178833, "learning_rate": 2.823497157970294e-05, "loss": 0.0281, "step": 66120 }, { "grad_norm": 0.3814423680305481, "learning_rate": 2.8220086826255164e-05, "loss": 0.0238, "step": 66130 }, { "grad_norm": 0.3807651400566101, "learning_rate": 2.8205204454625934e-05, "loss": 0.0341, "step": 66140 }, { "grad_norm": 0.3025229573249817, "learning_rate": 2.8190324466442798e-05, "loss": 0.0282, "step": 66150 }, { "grad_norm": 0.3948068618774414, "learning_rate": 2.8175446863332965e-05, "loss": 0.0257, "step": 66160 }, { "grad_norm": 0.39258667826652527, "learning_rate": 2.816057164692345e-05, "loss": 0.0277, "step": 66170 }, { "grad_norm": 0.31374385952949524, "learning_rate": 2.8145698818840992e-05, "loss": 0.032, "step": 66180 }, { "grad_norm": 0.39443057775497437, "learning_rate": 2.8130828380712065e-05, "loss": 0.0303, "step": 66190 }, { "grad_norm": 0.30717387795448303, "learning_rate": 2.811596033416285e-05, "loss": 0.0252, "step": 66200 }, { "grad_norm": 0.2873283326625824, "learning_rate": 2.8101094680819307e-05, "loss": 0.028, "step": 66210 }, { "grad_norm": 0.683239758014679, "learning_rate": 2.808623142230715e-05, "loss": 0.025, "step": 66220 }, { "grad_norm": 0.6367064714431763, "learning_rate": 2.8071370560251757e-05, "loss": 0.0275, "step": 66230 }, { "grad_norm": 0.4653904438018799, "learning_rate": 2.805651209627831e-05, "loss": 0.0295, "step": 66240 }, { "grad_norm": 0.38558822870254517, "learning_rate": 2.804165603201171e-05, "loss": 0.0271, "step": 66250 }, { "grad_norm": 0.35058826208114624, "learning_rate": 2.802680236907661e-05, "loss": 0.0292, "step": 66260 }, { "grad_norm": 0.20303049683570862, "learning_rate": 2.8011951109097344e-05, "loss": 0.0274, "step": 66270 }, { "grad_norm": 0.34073615074157715, "learning_rate": 2.7997102253698072e-05, "loss": 0.0288, "step": 66280 }, { "grad_norm": 0.3610352575778961, "learning_rate": 2.7982255804502588e-05, "loss": 0.0271, "step": 66290 }, { "grad_norm": 0.2246992290019989, "learning_rate": 2.7967411763134516e-05, "loss": 0.0321, "step": 66300 }, { "grad_norm": 0.3969901204109192, "learning_rate": 2.7952570131217166e-05, "loss": 0.0291, "step": 66310 }, { "grad_norm": 0.3373399078845978, "learning_rate": 2.7937730910373618e-05, "loss": 0.0296, "step": 66320 }, { "grad_norm": 0.19013935327529907, "learning_rate": 2.7922894102226628e-05, "loss": 0.0256, "step": 66330 }, { "grad_norm": 0.257146418094635, "learning_rate": 2.790805970839876e-05, "loss": 0.0233, "step": 66340 }, { "grad_norm": 0.2133776992559433, "learning_rate": 2.789322773051225e-05, "loss": 0.0256, "step": 66350 }, { "grad_norm": 0.31724292039871216, "learning_rate": 2.7878398170189124e-05, "loss": 0.0268, "step": 66360 }, { "grad_norm": 0.2904154062271118, "learning_rate": 2.7863571029051112e-05, "loss": 0.0305, "step": 66370 }, { "grad_norm": 0.33658841252326965, "learning_rate": 2.784874630871971e-05, "loss": 0.0301, "step": 66380 }, { "grad_norm": 0.45395299792289734, "learning_rate": 2.7833924010816086e-05, "loss": 0.0276, "step": 66390 }, { "grad_norm": 0.16869927942752838, "learning_rate": 2.7819104136961226e-05, "loss": 0.0302, "step": 66400 }, { "grad_norm": 0.21724797785282135, "learning_rate": 2.780428668877577e-05, "loss": 0.0237, "step": 66410 }, { "grad_norm": 0.2712547779083252, "learning_rate": 2.7789471667880152e-05, "loss": 0.0256, "step": 66420 }, { "grad_norm": 0.43183717131614685, "learning_rate": 2.7774659075894523e-05, "loss": 0.0318, "step": 66430 }, { "grad_norm": 0.314285010099411, "learning_rate": 2.7759848914438778e-05, "loss": 0.0333, "step": 66440 }, { "grad_norm": 0.43383073806762695, "learning_rate": 2.7745041185132496e-05, "loss": 0.0305, "step": 66450 }, { "grad_norm": 0.4402233958244324, "learning_rate": 2.773023588959506e-05, "loss": 0.0346, "step": 66460 }, { "grad_norm": 0.3473542332649231, "learning_rate": 2.7715433029445555e-05, "loss": 0.0289, "step": 66470 }, { "grad_norm": 0.262710839509964, "learning_rate": 2.7700632606302766e-05, "loss": 0.0265, "step": 66480 }, { "grad_norm": 0.39130309224128723, "learning_rate": 2.768583462178528e-05, "loss": 0.0308, "step": 66490 }, { "grad_norm": 0.30787473917007446, "learning_rate": 2.767103907751136e-05, "loss": 0.0264, "step": 66500 }, { "grad_norm": 0.6524556279182434, "learning_rate": 2.7656245975099065e-05, "loss": 0.0253, "step": 66510 }, { "grad_norm": 0.23311249911785126, "learning_rate": 2.7641455316166093e-05, "loss": 0.0278, "step": 66520 }, { "grad_norm": 0.199382483959198, "learning_rate": 2.7626667102329973e-05, "loss": 0.0284, "step": 66530 }, { "grad_norm": 0.3093326687812805, "learning_rate": 2.761188133520788e-05, "loss": 0.0253, "step": 66540 }, { "grad_norm": 0.5580571889877319, "learning_rate": 2.7597098016416784e-05, "loss": 0.0297, "step": 66550 }, { "grad_norm": 0.4255402684211731, "learning_rate": 2.7582317147573368e-05, "loss": 0.0306, "step": 66560 }, { "grad_norm": 0.29113149642944336, "learning_rate": 2.7567538730294064e-05, "loss": 0.0304, "step": 66570 }, { "grad_norm": 1.052342414855957, "learning_rate": 2.7552762766194973e-05, "loss": 0.0297, "step": 66580 }, { "grad_norm": 0.24353614449501038, "learning_rate": 2.753798925689202e-05, "loss": 0.0285, "step": 66590 }, { "grad_norm": 0.6800748109817505, "learning_rate": 2.7523218204000774e-05, "loss": 0.0295, "step": 66600 }, { "grad_norm": 0.2909936308860779, "learning_rate": 2.7508449609136612e-05, "loss": 0.0257, "step": 66610 }, { "grad_norm": 0.35613375902175903, "learning_rate": 2.749368347391454e-05, "loss": 0.0309, "step": 66620 }, { "grad_norm": 0.3981339931488037, "learning_rate": 2.7478919799949443e-05, "loss": 0.0275, "step": 66630 }, { "grad_norm": 0.3271508812904358, "learning_rate": 2.7464158588855803e-05, "loss": 0.0245, "step": 66640 }, { "grad_norm": 0.458548903465271, "learning_rate": 2.7449399842247918e-05, "loss": 0.0246, "step": 66650 }, { "grad_norm": 0.34025660157203674, "learning_rate": 2.7434643561739736e-05, "loss": 0.0234, "step": 66660 }, { "grad_norm": 0.4380779266357422, "learning_rate": 2.7419889748945016e-05, "loss": 0.0225, "step": 66670 }, { "grad_norm": 0.25899145007133484, "learning_rate": 2.7405138405477203e-05, "loss": 0.0274, "step": 66680 }, { "grad_norm": 0.2588135302066803, "learning_rate": 2.73903895329495e-05, "loss": 0.0264, "step": 66690 }, { "grad_norm": 0.4315086901187897, "learning_rate": 2.737564313297478e-05, "loss": 0.0331, "step": 66700 }, { "grad_norm": 0.2585456371307373, "learning_rate": 2.736089920716572e-05, "loss": 0.0252, "step": 66710 }, { "grad_norm": 0.3188716769218445, "learning_rate": 2.7346157757134706e-05, "loss": 0.0303, "step": 66720 }, { "grad_norm": 0.8043875098228455, "learning_rate": 2.733141878449379e-05, "loss": 0.0297, "step": 66730 }, { "grad_norm": 0.20983874797821045, "learning_rate": 2.7316682290854844e-05, "loss": 0.029, "step": 66740 }, { "grad_norm": 0.39290231466293335, "learning_rate": 2.7301948277829414e-05, "loss": 0.0304, "step": 66750 }, { "grad_norm": 0.35114240646362305, "learning_rate": 2.7287216747028814e-05, "loss": 0.027, "step": 66760 }, { "grad_norm": 0.2655656337738037, "learning_rate": 2.7272487700064025e-05, "loss": 0.0255, "step": 66770 }, { "grad_norm": 0.28862759470939636, "learning_rate": 2.725776113854583e-05, "loss": 0.0319, "step": 66780 }, { "grad_norm": 0.2746874988079071, "learning_rate": 2.7243037064084664e-05, "loss": 0.0248, "step": 66790 }, { "grad_norm": 0.535975992679596, "learning_rate": 2.7228315478290766e-05, "loss": 0.0343, "step": 66800 }, { "grad_norm": 0.24518992006778717, "learning_rate": 2.721359638277402e-05, "loss": 0.026, "step": 66810 }, { "grad_norm": 0.2730979919433594, "learning_rate": 2.7198879779144148e-05, "loss": 0.0252, "step": 66820 }, { "grad_norm": 0.16725845634937286, "learning_rate": 2.7184165669010485e-05, "loss": 0.0268, "step": 66830 }, { "grad_norm": 0.3438543975353241, "learning_rate": 2.716945405398218e-05, "loss": 0.0301, "step": 66840 }, { "grad_norm": 0.3064032196998596, "learning_rate": 2.715474493566803e-05, "loss": 0.0244, "step": 66850 }, { "grad_norm": 0.2171788215637207, "learning_rate": 2.7140038315676652e-05, "loss": 0.0247, "step": 66860 }, { "grad_norm": 0.2881663739681244, "learning_rate": 2.7125334195616275e-05, "loss": 0.0454, "step": 66870 }, { "grad_norm": 0.33377882838249207, "learning_rate": 2.7110632577094997e-05, "loss": 0.0306, "step": 66880 }, { "grad_norm": 0.24251531064510345, "learning_rate": 2.709593346172051e-05, "loss": 0.0299, "step": 66890 }, { "grad_norm": 0.45819103717803955, "learning_rate": 2.708123685110032e-05, "loss": 0.0271, "step": 66900 }, { "grad_norm": 0.28354156017303467, "learning_rate": 2.7066542746841588e-05, "loss": 0.0172, "step": 66910 }, { "grad_norm": 0.6021929979324341, "learning_rate": 2.7051851150551267e-05, "loss": 0.0217, "step": 66920 }, { "grad_norm": 0.27719646692276, "learning_rate": 2.7037162063835996e-05, "loss": 0.0236, "step": 66930 }, { "grad_norm": 0.539064347743988, "learning_rate": 2.702247548830218e-05, "loss": 0.0314, "step": 66940 }, { "grad_norm": 0.2633081078529358, "learning_rate": 2.7007791425555883e-05, "loss": 0.0186, "step": 66950 }, { "grad_norm": 0.3085445165634155, "learning_rate": 2.6993109877202945e-05, "loss": 0.0279, "step": 66960 }, { "grad_norm": 0.3201863169670105, "learning_rate": 2.6978430844848935e-05, "loss": 0.0263, "step": 66970 }, { "grad_norm": 0.20195837318897247, "learning_rate": 2.69637543300991e-05, "loss": 0.022, "step": 66980 }, { "grad_norm": 0.2380787432193756, "learning_rate": 2.694908033455848e-05, "loss": 0.0261, "step": 66990 }, { "grad_norm": 0.25440552830696106, "learning_rate": 2.6934408859831733e-05, "loss": 0.0311, "step": 67000 }, { "grad_norm": 0.3636663854122162, "learning_rate": 2.69197399075234e-05, "loss": 0.0315, "step": 67010 }, { "grad_norm": 0.2887971103191376, "learning_rate": 2.6905073479237584e-05, "loss": 0.0276, "step": 67020 }, { "grad_norm": 0.36267125606536865, "learning_rate": 2.6890409576578235e-05, "loss": 0.0311, "step": 67030 }, { "grad_norm": 0.4829486608505249, "learning_rate": 2.687574820114892e-05, "loss": 0.0274, "step": 67040 }, { "grad_norm": 0.44376784563064575, "learning_rate": 2.686108935455305e-05, "loss": 0.0322, "step": 67050 }, { "grad_norm": 0.4160078465938568, "learning_rate": 2.6846433038393603e-05, "loss": 0.0321, "step": 67060 }, { "grad_norm": 0.428142786026001, "learning_rate": 2.6831779254273482e-05, "loss": 0.0264, "step": 67070 }, { "grad_norm": 0.3746844530105591, "learning_rate": 2.6817128003795122e-05, "loss": 0.0249, "step": 67080 }, { "grad_norm": 0.3343394696712494, "learning_rate": 2.680247928856081e-05, "loss": 0.0251, "step": 67090 }, { "grad_norm": 0.35477206110954285, "learning_rate": 2.6787833110172466e-05, "loss": 0.0321, "step": 67100 }, { "grad_norm": 0.7336992621421814, "learning_rate": 2.6773189470231807e-05, "loss": 0.0288, "step": 67110 }, { "grad_norm": 0.3528023660182953, "learning_rate": 2.675854837034022e-05, "loss": 0.0342, "step": 67120 }, { "grad_norm": 0.3131401836872101, "learning_rate": 2.6743909812098833e-05, "loss": 0.0252, "step": 67130 }, { "grad_norm": 0.28476521372795105, "learning_rate": 2.67292737971085e-05, "loss": 0.0303, "step": 67140 }, { "grad_norm": 0.5598470568656921, "learning_rate": 2.671464032696982e-05, "loss": 0.0304, "step": 67150 }, { "grad_norm": 0.7085975408554077, "learning_rate": 2.6700009403283055e-05, "loss": 0.0299, "step": 67160 }, { "grad_norm": 0.6126841902732849, "learning_rate": 2.6685381027648214e-05, "loss": 0.0326, "step": 67170 }, { "grad_norm": 0.41954293847084045, "learning_rate": 2.6670755201665086e-05, "loss": 0.0395, "step": 67180 }, { "grad_norm": 0.2928626835346222, "learning_rate": 2.6656131926933066e-05, "loss": 0.036, "step": 67190 }, { "grad_norm": 0.5534818172454834, "learning_rate": 2.664151120505136e-05, "loss": 0.0274, "step": 67200 }, { "grad_norm": 0.2232990264892578, "learning_rate": 2.6626893037618873e-05, "loss": 0.0264, "step": 67210 }, { "grad_norm": 0.39337193965911865, "learning_rate": 2.6612277426234243e-05, "loss": 0.0419, "step": 67220 }, { "grad_norm": 0.2804441452026367, "learning_rate": 2.659766437249577e-05, "loss": 0.0322, "step": 67230 }, { "grad_norm": 0.4730165898799896, "learning_rate": 2.6583053878001557e-05, "loss": 0.0252, "step": 67240 }, { "grad_norm": 0.3198249936103821, "learning_rate": 2.6568445944349323e-05, "loss": 0.0257, "step": 67250 }, { "grad_norm": 0.324298620223999, "learning_rate": 2.6553840573136657e-05, "loss": 0.025, "step": 67260 }, { "grad_norm": 0.5025755167007446, "learning_rate": 2.6539237765960706e-05, "loss": 0.0253, "step": 67270 }, { "grad_norm": 0.651524007320404, "learning_rate": 2.6524637524418465e-05, "loss": 0.0341, "step": 67280 }, { "grad_norm": 0.30975526571273804, "learning_rate": 2.651003985010655e-05, "loss": 0.0241, "step": 67290 }, { "grad_norm": 0.40357503294944763, "learning_rate": 2.649544474462138e-05, "loss": 0.0336, "step": 67300 }, { "grad_norm": 0.30381709337234497, "learning_rate": 2.6480852209559015e-05, "loss": 0.0239, "step": 67310 }, { "grad_norm": 0.40607699751853943, "learning_rate": 2.6466262246515296e-05, "loss": 0.0249, "step": 67320 }, { "grad_norm": 0.4181874394416809, "learning_rate": 2.6451674857085746e-05, "loss": 0.0257, "step": 67330 }, { "grad_norm": 0.2980571389198303, "learning_rate": 2.6437090042865655e-05, "loss": 0.0327, "step": 67340 }, { "grad_norm": 0.2179393768310547, "learning_rate": 2.6422507805449947e-05, "loss": 0.0245, "step": 67350 }, { "grad_norm": 0.30689334869384766, "learning_rate": 2.6407928146433357e-05, "loss": 0.0304, "step": 67360 }, { "grad_norm": 0.33179059624671936, "learning_rate": 2.639335106741026e-05, "loss": 0.028, "step": 67370 }, { "grad_norm": 0.32571351528167725, "learning_rate": 2.6378776569974796e-05, "loss": 0.0309, "step": 67380 }, { "grad_norm": 0.2617572546005249, "learning_rate": 2.6364204655720814e-05, "loss": 0.0258, "step": 67390 }, { "grad_norm": 0.34245729446411133, "learning_rate": 2.6349635326241894e-05, "loss": 0.0301, "step": 67400 }, { "grad_norm": 0.43249595165252686, "learning_rate": 2.633506858313127e-05, "loss": 0.0271, "step": 67410 }, { "grad_norm": 0.3617728054523468, "learning_rate": 2.6320504427981975e-05, "loss": 0.0252, "step": 67420 }, { "grad_norm": 0.25645890831947327, "learning_rate": 2.6305942862386734e-05, "loss": 0.0288, "step": 67430 }, { "grad_norm": 0.368278443813324, "learning_rate": 2.629138388793793e-05, "loss": 0.0237, "step": 67440 }, { "grad_norm": 0.3667079508304596, "learning_rate": 2.6276827506227743e-05, "loss": 0.029, "step": 67450 }, { "grad_norm": 0.2538674473762512, "learning_rate": 2.6262273718848037e-05, "loss": 0.0237, "step": 67460 }, { "grad_norm": 0.29937079548835754, "learning_rate": 2.6247722527390407e-05, "loss": 0.0237, "step": 67470 }, { "grad_norm": 0.24578090012073517, "learning_rate": 2.6233173933446104e-05, "loss": 0.0319, "step": 67480 }, { "grad_norm": 0.5890564918518066, "learning_rate": 2.6218627938606188e-05, "loss": 0.0297, "step": 67490 }, { "grad_norm": 0.2968057096004486, "learning_rate": 2.620408454446135e-05, "loss": 0.0245, "step": 67500 }, { "grad_norm": 0.3099187910556793, "learning_rate": 2.6189543752602046e-05, "loss": 0.0244, "step": 67510 }, { "grad_norm": 0.3952123522758484, "learning_rate": 2.6175005564618444e-05, "loss": 0.032, "step": 67520 }, { "grad_norm": 0.3743571937084198, "learning_rate": 2.6160469982100428e-05, "loss": 0.0298, "step": 67530 }, { "grad_norm": 0.5949870944023132, "learning_rate": 2.6145937006637555e-05, "loss": 0.0466, "step": 67540 }, { "grad_norm": 0.3340725898742676, "learning_rate": 2.6131406639819166e-05, "loss": 0.0293, "step": 67550 }, { "grad_norm": 0.2774398922920227, "learning_rate": 2.611687888323424e-05, "loss": 0.0269, "step": 67560 }, { "grad_norm": 0.38466864824295044, "learning_rate": 2.610235373847153e-05, "loss": 0.0358, "step": 67570 }, { "grad_norm": 0.41999703645706177, "learning_rate": 2.6087831207119494e-05, "loss": 0.0319, "step": 67580 }, { "grad_norm": 0.3675363063812256, "learning_rate": 2.60733112907663e-05, "loss": 0.0334, "step": 67590 }, { "grad_norm": 0.4043094217777252, "learning_rate": 2.6058793990999786e-05, "loss": 0.0302, "step": 67600 }, { "grad_norm": 0.45056331157684326, "learning_rate": 2.6044279309407593e-05, "loss": 0.0242, "step": 67610 }, { "grad_norm": 0.38892337679862976, "learning_rate": 2.6029767247576976e-05, "loss": 0.0355, "step": 67620 }, { "grad_norm": 0.3376346528530121, "learning_rate": 2.601525780709497e-05, "loss": 0.0251, "step": 67630 }, { "grad_norm": 0.3852459192276001, "learning_rate": 2.6000750989548318e-05, "loss": 0.0312, "step": 67640 }, { "grad_norm": 0.37547942996025085, "learning_rate": 2.5986246796523462e-05, "loss": 0.0253, "step": 67650 }, { "grad_norm": 0.24421750009059906, "learning_rate": 2.5971745229606546e-05, "loss": 0.0258, "step": 67660 }, { "grad_norm": 0.5342512726783752, "learning_rate": 2.5957246290383442e-05, "loss": 0.0357, "step": 67670 }, { "grad_norm": 0.22990833222866058, "learning_rate": 2.594274998043975e-05, "loss": 0.029, "step": 67680 }, { "grad_norm": 0.4207017719745636, "learning_rate": 2.5928256301360733e-05, "loss": 0.0248, "step": 67690 }, { "grad_norm": 0.24165712296962738, "learning_rate": 2.591376525473142e-05, "loss": 0.0223, "step": 67700 }, { "grad_norm": 0.36576181650161743, "learning_rate": 2.5899276842136523e-05, "loss": 0.0271, "step": 67710 }, { "grad_norm": 0.3056870698928833, "learning_rate": 2.5884791065160495e-05, "loss": 0.0262, "step": 67720 }, { "grad_norm": 0.32420384883880615, "learning_rate": 2.5870307925387436e-05, "loss": 0.0246, "step": 67730 }, { "grad_norm": 0.4080822169780731, "learning_rate": 2.585582742440125e-05, "loss": 0.0185, "step": 67740 }, { "grad_norm": 0.36250171065330505, "learning_rate": 2.5841349563785455e-05, "loss": 0.0251, "step": 67750 }, { "grad_norm": 0.25670790672302246, "learning_rate": 2.5826874345123355e-05, "loss": 0.0274, "step": 67760 }, { "grad_norm": 0.4050333797931671, "learning_rate": 2.581240176999792e-05, "loss": 0.0227, "step": 67770 }, { "grad_norm": 0.7775359153747559, "learning_rate": 2.579793183999189e-05, "loss": 0.0276, "step": 67780 }, { "grad_norm": 0.3478139638900757, "learning_rate": 2.5783464556687627e-05, "loss": 0.0299, "step": 67790 }, { "grad_norm": 0.3136236071586609, "learning_rate": 2.576899992166729e-05, "loss": 0.0295, "step": 67800 }, { "grad_norm": 0.3365972936153412, "learning_rate": 2.5754537936512668e-05, "loss": 0.0237, "step": 67810 }, { "grad_norm": 0.5264560580253601, "learning_rate": 2.574007860280535e-05, "loss": 0.0373, "step": 67820 }, { "grad_norm": 0.20751135051250458, "learning_rate": 2.5725621922126518e-05, "loss": 0.0243, "step": 67830 }, { "grad_norm": 0.3076118528842926, "learning_rate": 2.571116789605722e-05, "loss": 0.0273, "step": 67840 }, { "grad_norm": 0.3583265542984009, "learning_rate": 2.5696716526178054e-05, "loss": 0.0355, "step": 67850 }, { "grad_norm": 0.459124892950058, "learning_rate": 2.568226781406946e-05, "loss": 0.0387, "step": 67860 }, { "grad_norm": 0.432583212852478, "learning_rate": 2.5667821761311474e-05, "loss": 0.0342, "step": 67870 }, { "grad_norm": 0.22806261479854584, "learning_rate": 2.5653378369483917e-05, "loss": 0.0218, "step": 67880 }, { "grad_norm": 0.32827162742614746, "learning_rate": 2.5638937640166292e-05, "loss": 0.0226, "step": 67890 }, { "grad_norm": 0.27503350377082825, "learning_rate": 2.5624499574937842e-05, "loss": 0.0284, "step": 67900 }, { "grad_norm": 0.19375775754451752, "learning_rate": 2.5610064175377456e-05, "loss": 0.0272, "step": 67910 }, { "grad_norm": 0.2981224060058594, "learning_rate": 2.5595631443063782e-05, "loss": 0.0263, "step": 67920 }, { "grad_norm": 0.5635565519332886, "learning_rate": 2.5581201379575177e-05, "loss": 0.022, "step": 67930 }, { "grad_norm": 0.4232270121574402, "learning_rate": 2.5566773986489655e-05, "loss": 0.0289, "step": 67940 }, { "grad_norm": 0.3176587224006653, "learning_rate": 2.5552349265385e-05, "loss": 0.04, "step": 67950 }, { "grad_norm": 0.3120401203632355, "learning_rate": 2.5537927217838675e-05, "loss": 0.0233, "step": 67960 }, { "grad_norm": 0.2803388237953186, "learning_rate": 2.5523507845427864e-05, "loss": 0.0261, "step": 67970 }, { "grad_norm": 0.37864771485328674, "learning_rate": 2.5509091149729413e-05, "loss": 0.0266, "step": 67980 }, { "grad_norm": 0.23398526012897491, "learning_rate": 2.5494677132319962e-05, "loss": 0.028, "step": 67990 }, { "grad_norm": 0.39247411489486694, "learning_rate": 2.548026579477575e-05, "loss": 0.0312, "step": 68000 }, { "grad_norm": 0.4115184545516968, "learning_rate": 2.5465857138672834e-05, "loss": 0.0267, "step": 68010 }, { "grad_norm": 0.3042127788066864, "learning_rate": 2.5451451165586848e-05, "loss": 0.0309, "step": 68020 }, { "grad_norm": 0.3126039206981659, "learning_rate": 2.543704787709329e-05, "loss": 0.0247, "step": 68030 }, { "grad_norm": 0.7519035935401917, "learning_rate": 2.542264727476723e-05, "loss": 0.0274, "step": 68040 }, { "grad_norm": 0.2240159958600998, "learning_rate": 2.5408249360183535e-05, "loss": 0.0338, "step": 68050 }, { "grad_norm": 0.2585313320159912, "learning_rate": 2.5393854134916694e-05, "loss": 0.0232, "step": 68060 }, { "grad_norm": 0.37845900654792786, "learning_rate": 2.5379461600540993e-05, "loss": 0.0252, "step": 68070 }, { "grad_norm": 0.25865885615348816, "learning_rate": 2.5365071758630317e-05, "loss": 0.0355, "step": 68080 }, { "grad_norm": 0.32209330797195435, "learning_rate": 2.535068461075839e-05, "loss": 0.0265, "step": 68090 }, { "grad_norm": 0.27911898493766785, "learning_rate": 2.5336300158498516e-05, "loss": 0.0323, "step": 68100 }, { "grad_norm": 0.670493483543396, "learning_rate": 2.5321918403423793e-05, "loss": 0.0264, "step": 68110 }, { "grad_norm": 0.33571261167526245, "learning_rate": 2.5307539347106957e-05, "loss": 0.0264, "step": 68120 }, { "grad_norm": 0.32993996143341064, "learning_rate": 2.529316299112049e-05, "loss": 0.0256, "step": 68130 }, { "grad_norm": 0.5708624124526978, "learning_rate": 2.5278789337036567e-05, "loss": 0.0281, "step": 68140 }, { "grad_norm": 0.3947352468967438, "learning_rate": 2.52644183864271e-05, "loss": 0.0292, "step": 68150 }, { "grad_norm": 0.46549850702285767, "learning_rate": 2.525005014086363e-05, "loss": 0.0279, "step": 68160 }, { "grad_norm": 0.5310244560241699, "learning_rate": 2.5235684601917465e-05, "loss": 0.0324, "step": 68170 }, { "grad_norm": 0.3572040796279907, "learning_rate": 2.5221321771159618e-05, "loss": 0.0268, "step": 68180 }, { "grad_norm": 0.2942401170730591, "learning_rate": 2.5206961650160743e-05, "loss": 0.0265, "step": 68190 }, { "grad_norm": 0.2127646505832672, "learning_rate": 2.519260424049129e-05, "loss": 0.0317, "step": 68200 }, { "grad_norm": 0.23541532456874847, "learning_rate": 2.5178249543721304e-05, "loss": 0.0308, "step": 68210 }, { "grad_norm": 0.2567935585975647, "learning_rate": 2.5163897561420668e-05, "loss": 0.0243, "step": 68220 }, { "grad_norm": 0.3087853789329529, "learning_rate": 2.514954829515883e-05, "loss": 0.031, "step": 68230 }, { "grad_norm": 0.2820681631565094, "learning_rate": 2.513520174650505e-05, "loss": 0.0261, "step": 68240 }, { "grad_norm": 0.25464296340942383, "learning_rate": 2.5120857917028207e-05, "loss": 0.0229, "step": 68250 }, { "grad_norm": 0.3584957420825958, "learning_rate": 2.5106516808296953e-05, "loss": 0.0202, "step": 68260 }, { "grad_norm": 0.205290749669075, "learning_rate": 2.5092178421879552e-05, "loss": 0.0274, "step": 68270 }, { "grad_norm": 0.6208596229553223, "learning_rate": 2.5077842759344116e-05, "loss": 0.0363, "step": 68280 }, { "grad_norm": 0.31130385398864746, "learning_rate": 2.5063509822258304e-05, "loss": 0.0296, "step": 68290 }, { "grad_norm": 0.22679592669010162, "learning_rate": 2.5049179612189576e-05, "loss": 0.0309, "step": 68300 }, { "grad_norm": 0.3934241235256195, "learning_rate": 2.5034852130705043e-05, "loss": 0.0381, "step": 68310 }, { "grad_norm": 0.9763623476028442, "learning_rate": 2.5020527379371572e-05, "loss": 0.0334, "step": 68320 }, { "grad_norm": 0.3239240348339081, "learning_rate": 2.500620535975564e-05, "loss": 0.0233, "step": 68330 }, { "grad_norm": 0.41887304186820984, "learning_rate": 2.4991886073423515e-05, "loss": 0.0328, "step": 68340 }, { "grad_norm": 0.1970159411430359, "learning_rate": 2.4977569521941134e-05, "loss": 0.0236, "step": 68350 }, { "grad_norm": 0.4104607105255127, "learning_rate": 2.4963255706874143e-05, "loss": 0.0278, "step": 68360 }, { "grad_norm": 0.36425745487213135, "learning_rate": 2.4948944629787853e-05, "loss": 0.0313, "step": 68370 }, { "grad_norm": 0.4175354242324829, "learning_rate": 2.493463629224731e-05, "loss": 0.031, "step": 68380 }, { "grad_norm": 0.18823698163032532, "learning_rate": 2.492033069581728e-05, "loss": 0.0252, "step": 68390 }, { "grad_norm": 0.3312062621116638, "learning_rate": 2.4906027842062137e-05, "loss": 0.0309, "step": 68400 }, { "grad_norm": 0.23186667263507843, "learning_rate": 2.48917277325461e-05, "loss": 0.0257, "step": 68410 }, { "grad_norm": 0.2597653567790985, "learning_rate": 2.487743036883295e-05, "loss": 0.0268, "step": 68420 }, { "grad_norm": 0.297376811504364, "learning_rate": 2.486313575248626e-05, "loss": 0.0248, "step": 68430 }, { "grad_norm": 0.36011219024658203, "learning_rate": 2.4848843885069235e-05, "loss": 0.0334, "step": 68440 }, { "grad_norm": 0.2691580653190613, "learning_rate": 2.483455476814484e-05, "loss": 0.0234, "step": 68450 }, { "grad_norm": 0.36545488238334656, "learning_rate": 2.4820268403275672e-05, "loss": 0.0272, "step": 68460 }, { "grad_norm": 0.45930761098861694, "learning_rate": 2.480598479202413e-05, "loss": 0.0293, "step": 68470 }, { "grad_norm": 0.2472924143075943, "learning_rate": 2.4791703935952193e-05, "loss": 0.0256, "step": 68480 }, { "grad_norm": 0.3219233751296997, "learning_rate": 2.477742583662163e-05, "loss": 0.0295, "step": 68490 }, { "grad_norm": 0.298830509185791, "learning_rate": 2.476315049559384e-05, "loss": 0.0272, "step": 68500 }, { "grad_norm": 0.465725302696228, "learning_rate": 2.4748877914429987e-05, "loss": 0.0346, "step": 68510 }, { "grad_norm": 0.3103380799293518, "learning_rate": 2.4734608094690857e-05, "loss": 0.0314, "step": 68520 }, { "grad_norm": 0.22609823942184448, "learning_rate": 2.4720341037937006e-05, "loss": 0.0381, "step": 68530 }, { "grad_norm": 0.29584944248199463, "learning_rate": 2.4706076745728645e-05, "loss": 0.035, "step": 68540 }, { "grad_norm": 0.30124953389167786, "learning_rate": 2.469181521962573e-05, "loss": 0.0255, "step": 68550 }, { "grad_norm": 0.3211670517921448, "learning_rate": 2.467755646118783e-05, "loss": 0.0237, "step": 68560 }, { "grad_norm": 0.25103408098220825, "learning_rate": 2.46633004719743e-05, "loss": 0.027, "step": 68570 }, { "grad_norm": 0.2791064977645874, "learning_rate": 2.4649047253544123e-05, "loss": 0.0255, "step": 68580 }, { "grad_norm": 0.2343292385339737, "learning_rate": 2.4634796807456016e-05, "loss": 0.023, "step": 68590 }, { "grad_norm": 0.3184429109096527, "learning_rate": 2.46205491352684e-05, "loss": 0.0321, "step": 68600 }, { "grad_norm": 0.6623467803001404, "learning_rate": 2.4606304238539375e-05, "loss": 0.028, "step": 68610 }, { "grad_norm": 0.7924172878265381, "learning_rate": 2.4592062118826753e-05, "loss": 0.0296, "step": 68620 }, { "grad_norm": 0.38763701915740967, "learning_rate": 2.4577822777688004e-05, "loss": 0.0195, "step": 68630 }, { "grad_norm": 0.43251118063926697, "learning_rate": 2.4563586216680347e-05, "loss": 0.035, "step": 68640 }, { "grad_norm": 0.3134847581386566, "learning_rate": 2.4549352437360622e-05, "loss": 0.0235, "step": 68650 }, { "grad_norm": 0.4256132245063782, "learning_rate": 2.4535121441285493e-05, "loss": 0.0336, "step": 68660 }, { "grad_norm": 0.448042094707489, "learning_rate": 2.4520893230011172e-05, "loss": 0.0255, "step": 68670 }, { "grad_norm": 0.21310575306415558, "learning_rate": 2.450666780509368e-05, "loss": 0.0264, "step": 68680 }, { "grad_norm": 0.35672634840011597, "learning_rate": 2.4492445168088656e-05, "loss": 0.0347, "step": 68690 }, { "grad_norm": 0.27478304505348206, "learning_rate": 2.447822532055149e-05, "loss": 0.0299, "step": 68700 }, { "grad_norm": 0.4203876554965973, "learning_rate": 2.4464008264037212e-05, "loss": 0.0265, "step": 68710 }, { "grad_norm": 0.3533715605735779, "learning_rate": 2.4449794000100605e-05, "loss": 0.0292, "step": 68720 }, { "grad_norm": 0.3125239610671997, "learning_rate": 2.4435582530296108e-05, "loss": 0.0258, "step": 68730 }, { "grad_norm": 0.19294705986976624, "learning_rate": 2.4421373856177887e-05, "loss": 0.0237, "step": 68740 }, { "grad_norm": 0.2567295730113983, "learning_rate": 2.440716797929975e-05, "loss": 0.0227, "step": 68750 }, { "grad_norm": 0.28547170758247375, "learning_rate": 2.439296490121526e-05, "loss": 0.0181, "step": 68760 }, { "grad_norm": 0.42244911193847656, "learning_rate": 2.4378764623477608e-05, "loss": 0.0217, "step": 68770 }, { "grad_norm": 0.2874348759651184, "learning_rate": 2.436456714763974e-05, "loss": 0.0198, "step": 68780 }, { "grad_norm": 0.25762858986854553, "learning_rate": 2.4350372475254262e-05, "loss": 0.029, "step": 68790 }, { "grad_norm": 0.18858475983142853, "learning_rate": 2.433618060787351e-05, "loss": 0.0284, "step": 68800 }, { "grad_norm": 0.3038192391395569, "learning_rate": 2.432199154704944e-05, "loss": 0.0348, "step": 68810 }, { "grad_norm": 0.36032336950302124, "learning_rate": 2.430780529433377e-05, "loss": 0.0278, "step": 68820 }, { "grad_norm": 0.4327212870121002, "learning_rate": 2.42936218512779e-05, "loss": 0.0317, "step": 68830 }, { "grad_norm": 0.209786519408226, "learning_rate": 2.427944121943288e-05, "loss": 0.0181, "step": 68840 }, { "grad_norm": 0.3461360037326813, "learning_rate": 2.4265263400349497e-05, "loss": 0.0261, "step": 68850 }, { "grad_norm": 0.30033987760543823, "learning_rate": 2.4251088395578213e-05, "loss": 0.0262, "step": 68860 }, { "grad_norm": 0.2813161611557007, "learning_rate": 2.4236916206669214e-05, "loss": 0.0299, "step": 68870 }, { "grad_norm": 0.3364829123020172, "learning_rate": 2.4222746835172304e-05, "loss": 0.0236, "step": 68880 }, { "grad_norm": 0.37326598167419434, "learning_rate": 2.4208580282637065e-05, "loss": 0.0267, "step": 68890 }, { "grad_norm": 0.41319799423217773, "learning_rate": 2.4194416550612687e-05, "loss": 0.0249, "step": 68900 }, { "grad_norm": 0.19807963073253632, "learning_rate": 2.418025564064812e-05, "loss": 0.0213, "step": 68910 }, { "grad_norm": 0.5092809796333313, "learning_rate": 2.4166097554291978e-05, "loss": 0.0246, "step": 68920 }, { "grad_norm": 0.3213188648223877, "learning_rate": 2.4151942293092583e-05, "loss": 0.0238, "step": 68930 }, { "grad_norm": 0.09855655580759048, "learning_rate": 2.41377898585979e-05, "loss": 0.0215, "step": 68940 }, { "grad_norm": 0.3980536460876465, "learning_rate": 2.4123640252355652e-05, "loss": 0.0275, "step": 68950 }, { "grad_norm": 0.2710294723510742, "learning_rate": 2.410949347591318e-05, "loss": 0.0328, "step": 68960 }, { "grad_norm": 0.4545361399650574, "learning_rate": 2.4095349530817578e-05, "loss": 0.0425, "step": 68970 }, { "grad_norm": 0.35443198680877686, "learning_rate": 2.4081208418615603e-05, "loss": 0.0237, "step": 68980 }, { "grad_norm": 0.3642325699329376, "learning_rate": 2.4067070140853725e-05, "loss": 0.028, "step": 68990 }, { "grad_norm": 0.30346471071243286, "learning_rate": 2.405293469907805e-05, "loss": 0.0337, "step": 69000 }, { "grad_norm": 0.40208667516708374, "learning_rate": 2.4038802094834446e-05, "loss": 0.0265, "step": 69010 }, { "grad_norm": 0.22733917832374573, "learning_rate": 2.4024672329668395e-05, "loss": 0.0298, "step": 69020 }, { "grad_norm": 0.3054939806461334, "learning_rate": 2.4010545405125125e-05, "loss": 0.0187, "step": 69030 }, { "grad_norm": 0.3798740804195404, "learning_rate": 2.3996421322749528e-05, "loss": 0.0357, "step": 69040 }, { "grad_norm": 0.319704532623291, "learning_rate": 2.3982300084086224e-05, "loss": 0.0257, "step": 69050 }, { "grad_norm": 0.7980519533157349, "learning_rate": 2.396818169067944e-05, "loss": 0.0245, "step": 69060 }, { "grad_norm": 0.29003244638442993, "learning_rate": 2.3954066144073178e-05, "loss": 0.0225, "step": 69070 }, { "grad_norm": 0.5295034646987915, "learning_rate": 2.39399534458111e-05, "loss": 0.0263, "step": 69080 }, { "grad_norm": 0.30154910683631897, "learning_rate": 2.392584359743651e-05, "loss": 0.0256, "step": 69090 }, { "grad_norm": 0.28761786222457886, "learning_rate": 2.3911736600492463e-05, "loss": 0.0191, "step": 69100 }, { "grad_norm": 0.37805700302124023, "learning_rate": 2.3897632456521672e-05, "loss": 0.0238, "step": 69110 }, { "grad_norm": 0.373417466878891, "learning_rate": 2.388353116706658e-05, "loss": 0.0347, "step": 69120 }, { "grad_norm": 0.24023501574993134, "learning_rate": 2.3869432733669232e-05, "loss": 0.0248, "step": 69130 }, { "grad_norm": 0.21182101964950562, "learning_rate": 2.385533715787145e-05, "loss": 0.0283, "step": 69140 }, { "grad_norm": 0.2595479190349579, "learning_rate": 2.3841244441214673e-05, "loss": 0.0223, "step": 69150 }, { "grad_norm": 0.4260561764240265, "learning_rate": 2.3827154585240073e-05, "loss": 0.0233, "step": 69160 }, { "grad_norm": 0.41294869780540466, "learning_rate": 2.3813067591488498e-05, "loss": 0.0355, "step": 69170 }, { "grad_norm": 0.2694392502307892, "learning_rate": 2.3798983461500494e-05, "loss": 0.0353, "step": 69180 }, { "grad_norm": 0.5216948390007019, "learning_rate": 2.3784902196816256e-05, "loss": 0.0308, "step": 69190 }, { "grad_norm": 0.5039044618606567, "learning_rate": 2.3770823798975716e-05, "loss": 0.0283, "step": 69200 }, { "grad_norm": 0.4007428288459778, "learning_rate": 2.3756748269518424e-05, "loss": 0.0245, "step": 69210 }, { "grad_norm": 0.4019683003425598, "learning_rate": 2.374267560998371e-05, "loss": 0.0345, "step": 69220 }, { "grad_norm": 0.2977858781814575, "learning_rate": 2.3728605821910477e-05, "loss": 0.0303, "step": 69230 }, { "grad_norm": 0.5352983474731445, "learning_rate": 2.3714538906837453e-05, "loss": 0.039, "step": 69240 }, { "grad_norm": 0.3292132318019867, "learning_rate": 2.3700474866302913e-05, "loss": 0.0248, "step": 69250 }, { "grad_norm": 0.3211580514907837, "learning_rate": 2.3686413701844917e-05, "loss": 0.0237, "step": 69260 }, { "grad_norm": 0.2770039737224579, "learning_rate": 2.367235541500114e-05, "loss": 0.0241, "step": 69270 }, { "grad_norm": 0.20763631165027618, "learning_rate": 2.3658300007308993e-05, "loss": 0.0259, "step": 69280 }, { "grad_norm": 0.6577650308609009, "learning_rate": 2.364424748030555e-05, "loss": 0.0247, "step": 69290 }, { "grad_norm": 0.5996641516685486, "learning_rate": 2.3630197835527595e-05, "loss": 0.0309, "step": 69300 }, { "grad_norm": 0.2571612298488617, "learning_rate": 2.3616151074511546e-05, "loss": 0.0323, "step": 69310 }, { "grad_norm": 0.2503420114517212, "learning_rate": 2.3602107198793545e-05, "loss": 0.0356, "step": 69320 }, { "grad_norm": 0.314681738615036, "learning_rate": 2.3588066209909427e-05, "loss": 0.0237, "step": 69330 }, { "grad_norm": 0.4188089370727539, "learning_rate": 2.3574028109394662e-05, "loss": 0.0265, "step": 69340 }, { "grad_norm": 0.2294531762599945, "learning_rate": 2.3559992898784454e-05, "loss": 0.0221, "step": 69350 }, { "grad_norm": 0.2688858211040497, "learning_rate": 2.354596057961367e-05, "loss": 0.0222, "step": 69360 }, { "grad_norm": 0.28592658042907715, "learning_rate": 2.353193115341688e-05, "loss": 0.0344, "step": 69370 }, { "grad_norm": 0.2756097912788391, "learning_rate": 2.3517904621728294e-05, "loss": 0.0208, "step": 69380 }, { "grad_norm": 0.24423100054264069, "learning_rate": 2.3503880986081855e-05, "loss": 0.0269, "step": 69390 }, { "grad_norm": 0.40952640771865845, "learning_rate": 2.3489860248011136e-05, "loss": 0.0243, "step": 69400 }, { "grad_norm": 0.29977962374687195, "learning_rate": 2.3475842409049464e-05, "loss": 0.028, "step": 69410 }, { "grad_norm": 0.3462088704109192, "learning_rate": 2.346182747072975e-05, "loss": 0.0341, "step": 69420 }, { "grad_norm": 0.2631985545158386, "learning_rate": 2.3447815434584718e-05, "loss": 0.02, "step": 69430 }, { "grad_norm": 0.8293942809104919, "learning_rate": 2.3433806302146656e-05, "loss": 0.0288, "step": 69440 }, { "grad_norm": 0.4577449560165405, "learning_rate": 2.3419800074947612e-05, "loss": 0.0348, "step": 69450 }, { "grad_norm": 0.36996567249298096, "learning_rate": 2.3405796754519248e-05, "loss": 0.0305, "step": 69460 }, { "grad_norm": 0.49894025921821594, "learning_rate": 2.3391796342392992e-05, "loss": 0.0299, "step": 69470 }, { "grad_norm": 0.29797986149787903, "learning_rate": 2.337779884009984e-05, "loss": 0.0346, "step": 69480 }, { "grad_norm": 0.26432693004608154, "learning_rate": 2.3363804249170624e-05, "loss": 0.0265, "step": 69490 }, { "grad_norm": 0.3463691771030426, "learning_rate": 2.3349812571135703e-05, "loss": 0.0263, "step": 69500 }, { "grad_norm": 0.24941445887088776, "learning_rate": 2.333582380752523e-05, "loss": 0.0215, "step": 69510 }, { "grad_norm": 0.207955002784729, "learning_rate": 2.3321837959868963e-05, "loss": 0.0237, "step": 69520 }, { "grad_norm": 0.2663905620574951, "learning_rate": 2.3307855029696383e-05, "loss": 0.0238, "step": 69530 }, { "grad_norm": 0.3143073618412018, "learning_rate": 2.3293875018536648e-05, "loss": 0.0254, "step": 69540 }, { "grad_norm": 0.40581214427948, "learning_rate": 2.327989792791861e-05, "loss": 0.0269, "step": 69550 }, { "grad_norm": 0.4550561308860779, "learning_rate": 2.326592375937074e-05, "loss": 0.0252, "step": 69560 }, { "grad_norm": 0.44916343688964844, "learning_rate": 2.3251952514421248e-05, "loss": 0.0231, "step": 69570 }, { "grad_norm": 0.3311115503311157, "learning_rate": 2.323798419459804e-05, "loss": 0.0243, "step": 69580 }, { "grad_norm": 0.40654319524765015, "learning_rate": 2.3224018801428625e-05, "loss": 0.0259, "step": 69590 }, { "grad_norm": 0.38635021448135376, "learning_rate": 2.3210056336440273e-05, "loss": 0.0264, "step": 69600 }, { "grad_norm": 0.24174410104751587, "learning_rate": 2.3196096801159844e-05, "loss": 0.023, "step": 69610 }, { "grad_norm": 0.3440091609954834, "learning_rate": 2.3182140197114012e-05, "loss": 0.0275, "step": 69620 }, { "grad_norm": 0.5387662053108215, "learning_rate": 2.3168186525828985e-05, "loss": 0.0338, "step": 69630 }, { "grad_norm": 0.30044183135032654, "learning_rate": 2.3154235788830757e-05, "loss": 0.0181, "step": 69640 }, { "grad_norm": 0.5240611433982849, "learning_rate": 2.3140287987644927e-05, "loss": 0.0317, "step": 69650 }, { "grad_norm": 0.5132094621658325, "learning_rate": 2.3126343123796833e-05, "loss": 0.0337, "step": 69660 }, { "grad_norm": 0.3041937053203583, "learning_rate": 2.3112401198811418e-05, "loss": 0.0229, "step": 69670 }, { "grad_norm": 0.3324682116508484, "learning_rate": 2.3098462214213418e-05, "loss": 0.0314, "step": 69680 }, { "grad_norm": 0.24502438306808472, "learning_rate": 2.308452617152712e-05, "loss": 0.0247, "step": 69690 }, { "grad_norm": 0.3888723850250244, "learning_rate": 2.3070593072276603e-05, "loss": 0.0304, "step": 69700 }, { "grad_norm": 0.35946395993232727, "learning_rate": 2.3056662917985518e-05, "loss": 0.0226, "step": 69710 }, { "grad_norm": 0.3109568655490875, "learning_rate": 2.3042735710177283e-05, "loss": 0.0351, "step": 69720 }, { "grad_norm": 0.2375546544790268, "learning_rate": 2.302881145037492e-05, "loss": 0.0225, "step": 69730 }, { "grad_norm": 0.22165440022945404, "learning_rate": 2.3014890140101192e-05, "loss": 0.0228, "step": 69740 }, { "grad_norm": 0.25804391503334045, "learning_rate": 2.30009717808785e-05, "loss": 0.0213, "step": 69750 }, { "grad_norm": 0.37792983651161194, "learning_rate": 2.2987056374228967e-05, "loss": 0.028, "step": 69760 }, { "grad_norm": 0.40756914019584656, "learning_rate": 2.297314392167431e-05, "loss": 0.0249, "step": 69770 }, { "grad_norm": 0.4095822870731354, "learning_rate": 2.295923442473601e-05, "loss": 0.0249, "step": 69780 }, { "grad_norm": 0.26467829942703247, "learning_rate": 2.294532788493519e-05, "loss": 0.0263, "step": 69790 }, { "grad_norm": 0.33261433243751526, "learning_rate": 2.293142430379262e-05, "loss": 0.0275, "step": 69800 }, { "grad_norm": 0.7541510462760925, "learning_rate": 2.291752368282879e-05, "loss": 0.0314, "step": 69810 }, { "grad_norm": 0.26143771409988403, "learning_rate": 2.290362602356385e-05, "loss": 0.0251, "step": 69820 }, { "grad_norm": 0.24398072063922882, "learning_rate": 2.288973132751765e-05, "loss": 0.0258, "step": 69830 }, { "grad_norm": 0.31887146830558777, "learning_rate": 2.287583959620965e-05, "loss": 0.028, "step": 69840 }, { "grad_norm": 0.22752875089645386, "learning_rate": 2.2861950831159074e-05, "loss": 0.0281, "step": 69850 }, { "grad_norm": 0.28154879808425903, "learning_rate": 2.284806503388472e-05, "loss": 0.0266, "step": 69860 }, { "grad_norm": 0.30652302503585815, "learning_rate": 2.283418220590518e-05, "loss": 0.0224, "step": 69870 }, { "grad_norm": 0.5179729461669922, "learning_rate": 2.282030234873862e-05, "loss": 0.0319, "step": 69880 }, { "grad_norm": 0.7743839025497437, "learning_rate": 2.2806425463902948e-05, "loss": 0.0249, "step": 69890 }, { "grad_norm": 0.4031946361064911, "learning_rate": 2.279255155291568e-05, "loss": 0.0338, "step": 69900 }, { "grad_norm": 0.2868400812149048, "learning_rate": 2.277868061729409e-05, "loss": 0.0255, "step": 69910 }, { "grad_norm": 0.631254255771637, "learning_rate": 2.276481265855504e-05, "loss": 0.027, "step": 69920 }, { "grad_norm": 0.40750983357429504, "learning_rate": 2.2750947678215128e-05, "loss": 0.0287, "step": 69930 }, { "grad_norm": 0.2909184396266937, "learning_rate": 2.2737085677790616e-05, "loss": 0.0242, "step": 69940 }, { "grad_norm": 0.30274826288223267, "learning_rate": 2.2723226658797443e-05, "loss": 0.0264, "step": 69950 }, { "grad_norm": 0.2866874933242798, "learning_rate": 2.270937062275117e-05, "loss": 0.0267, "step": 69960 }, { "grad_norm": 0.26640790700912476, "learning_rate": 2.2695517571167113e-05, "loss": 0.0219, "step": 69970 }, { "grad_norm": 0.2754290699958801, "learning_rate": 2.268166750556019e-05, "loss": 0.0258, "step": 69980 }, { "grad_norm": 0.20145051181316376, "learning_rate": 2.266782042744503e-05, "loss": 0.0167, "step": 69990 }, { "grad_norm": 0.3914865255355835, "learning_rate": 2.2653976338335935e-05, "loss": 0.0274, "step": 70000 }, { "grad_norm": 0.5716065168380737, "learning_rate": 2.2640135239746895e-05, "loss": 0.0296, "step": 70010 }, { "grad_norm": 0.3220989406108856, "learning_rate": 2.2626297133191515e-05, "loss": 0.0231, "step": 70020 }, { "grad_norm": 0.41178497672080994, "learning_rate": 2.261246202018312e-05, "loss": 0.0241, "step": 70030 }, { "grad_norm": 0.16967304050922394, "learning_rate": 2.2598629902234718e-05, "loss": 0.0208, "step": 70040 }, { "grad_norm": 0.38028717041015625, "learning_rate": 2.258480078085894e-05, "loss": 0.0297, "step": 70050 }, { "grad_norm": 0.4679090976715088, "learning_rate": 2.2570974657568126e-05, "loss": 0.0321, "step": 70060 }, { "grad_norm": 0.18439704179763794, "learning_rate": 2.2557151533874286e-05, "loss": 0.0222, "step": 70070 }, { "grad_norm": 0.3209545314311981, "learning_rate": 2.2543331411289116e-05, "loss": 0.0227, "step": 70080 }, { "grad_norm": 0.7375590801239014, "learning_rate": 2.252951429132392e-05, "loss": 0.0322, "step": 70090 }, { "grad_norm": 0.32563042640686035, "learning_rate": 2.2515700175489762e-05, "loss": 0.0209, "step": 70100 }, { "grad_norm": 0.3115486800670624, "learning_rate": 2.2501889065297288e-05, "loss": 0.0185, "step": 70110 }, { "grad_norm": 0.2875961363315582, "learning_rate": 2.2488080962256887e-05, "loss": 0.0247, "step": 70120 }, { "grad_norm": 0.4001317322254181, "learning_rate": 2.2474275867878585e-05, "loss": 0.0378, "step": 70130 }, { "grad_norm": 0.4649440050125122, "learning_rate": 2.246047378367211e-05, "loss": 0.0409, "step": 70140 }, { "grad_norm": 0.2578555643558502, "learning_rate": 2.2446674711146798e-05, "loss": 0.0252, "step": 70150 }, { "grad_norm": 0.4320202171802521, "learning_rate": 2.2432878651811734e-05, "loss": 0.0291, "step": 70160 }, { "grad_norm": 0.2523268461227417, "learning_rate": 2.2419085607175594e-05, "loss": 0.033, "step": 70170 }, { "grad_norm": 0.30800995230674744, "learning_rate": 2.240529557874678e-05, "loss": 0.0262, "step": 70180 }, { "grad_norm": 0.3960154056549072, "learning_rate": 2.2391508568033358e-05, "loss": 0.0328, "step": 70190 }, { "grad_norm": 0.3680233657360077, "learning_rate": 2.237772457654307e-05, "loss": 0.0282, "step": 70200 }, { "grad_norm": 0.5765402913093567, "learning_rate": 2.236394360578327e-05, "loss": 0.027, "step": 70210 }, { "grad_norm": 0.449701189994812, "learning_rate": 2.2350165657261073e-05, "loss": 0.0199, "step": 70220 }, { "grad_norm": 0.3682335317134857, "learning_rate": 2.2336390732483164e-05, "loss": 0.0336, "step": 70230 }, { "grad_norm": 0.25826412439346313, "learning_rate": 2.2322618832955976e-05, "loss": 0.036, "step": 70240 }, { "grad_norm": 0.3170488774776459, "learning_rate": 2.2308849960185578e-05, "loss": 0.0316, "step": 70250 }, { "grad_norm": 0.21049684286117554, "learning_rate": 2.229508411567774e-05, "loss": 0.0241, "step": 70260 }, { "grad_norm": 0.30043143033981323, "learning_rate": 2.2281321300937823e-05, "loss": 0.0292, "step": 70270 }, { "grad_norm": 0.25321871042251587, "learning_rate": 2.2267561517470937e-05, "loss": 0.0288, "step": 70280 }, { "grad_norm": 0.40018415451049805, "learning_rate": 2.2253804766781845e-05, "loss": 0.0288, "step": 70290 }, { "grad_norm": 0.40127426385879517, "learning_rate": 2.2240051050374933e-05, "loss": 0.0232, "step": 70300 }, { "grad_norm": 0.3245426118373871, "learning_rate": 2.22263003697543e-05, "loss": 0.032, "step": 70310 }, { "grad_norm": 0.22796905040740967, "learning_rate": 2.2212552726423702e-05, "loss": 0.0321, "step": 70320 }, { "grad_norm": 0.2814663052558899, "learning_rate": 2.2198808121886578e-05, "loss": 0.0223, "step": 70330 }, { "grad_norm": 0.17773376405239105, "learning_rate": 2.218506655764598e-05, "loss": 0.027, "step": 70340 }, { "grad_norm": 0.2619202733039856, "learning_rate": 2.2171328035204704e-05, "loss": 0.0219, "step": 70350 }, { "grad_norm": 0.26217982172966003, "learning_rate": 2.2157592556065133e-05, "loss": 0.0214, "step": 70360 }, { "grad_norm": 0.3500540554523468, "learning_rate": 2.2143860121729382e-05, "loss": 0.0254, "step": 70370 }, { "grad_norm": 0.28702715039253235, "learning_rate": 2.2130130733699206e-05, "loss": 0.0222, "step": 70380 }, { "grad_norm": 0.35449784994125366, "learning_rate": 2.2116404393476052e-05, "loss": 0.0276, "step": 70390 }, { "grad_norm": 0.38820815086364746, "learning_rate": 2.2102681102560967e-05, "loss": 0.0222, "step": 70400 }, { "grad_norm": 0.5376306176185608, "learning_rate": 2.208896086245476e-05, "loss": 0.0245, "step": 70410 }, { "grad_norm": 0.4813723564147949, "learning_rate": 2.2075243674657808e-05, "loss": 0.0314, "step": 70420 }, { "grad_norm": 0.31006452441215515, "learning_rate": 2.206152954067024e-05, "loss": 0.0256, "step": 70430 }, { "grad_norm": 0.4021679162979126, "learning_rate": 2.2047818461991758e-05, "loss": 0.0294, "step": 70440 }, { "grad_norm": 0.38959386944770813, "learning_rate": 2.2034110440121864e-05, "loss": 0.0307, "step": 70450 }, { "grad_norm": 0.26987528800964355, "learning_rate": 2.2020405476559585e-05, "loss": 0.0234, "step": 70460 }, { "grad_norm": 0.28320077061653137, "learning_rate": 2.2006703572803722e-05, "loss": 0.0291, "step": 70470 }, { "grad_norm": 0.23231752216815948, "learning_rate": 2.1993004730352646e-05, "loss": 0.0251, "step": 70480 }, { "grad_norm": 0.2009846270084381, "learning_rate": 2.1979308950704464e-05, "loss": 0.0234, "step": 70490 }, { "grad_norm": 0.1995379626750946, "learning_rate": 2.1965616235356923e-05, "loss": 0.0285, "step": 70500 }, { "grad_norm": 0.2727203071117401, "learning_rate": 2.1951926585807465e-05, "loss": 0.0214, "step": 70510 }, { "grad_norm": 0.2074756771326065, "learning_rate": 2.193824000355313e-05, "loss": 0.0335, "step": 70520 }, { "grad_norm": 0.6195961236953735, "learning_rate": 2.192455649009067e-05, "loss": 0.0326, "step": 70530 }, { "grad_norm": 0.4046601951122284, "learning_rate": 2.1910876046916523e-05, "loss": 0.0281, "step": 70540 }, { "grad_norm": 0.23933075368404388, "learning_rate": 2.1897198675526725e-05, "loss": 0.0267, "step": 70550 }, { "grad_norm": 0.412640780210495, "learning_rate": 2.1883524377417024e-05, "loss": 0.0254, "step": 70560 }, { "grad_norm": 0.3401860296726227, "learning_rate": 2.1869853154082825e-05, "loss": 0.0253, "step": 70570 }, { "grad_norm": 0.2225552350282669, "learning_rate": 2.1856185007019215e-05, "loss": 0.0236, "step": 70580 }, { "grad_norm": 0.30336207151412964, "learning_rate": 2.1842519937720874e-05, "loss": 0.0325, "step": 70590 }, { "grad_norm": 0.34264060854911804, "learning_rate": 2.182885794768224e-05, "loss": 0.0294, "step": 70600 }, { "grad_norm": 0.31542977690696716, "learning_rate": 2.1815199038397326e-05, "loss": 0.0276, "step": 70610 }, { "grad_norm": 0.40085920691490173, "learning_rate": 2.1801543211359893e-05, "loss": 0.0291, "step": 70620 }, { "grad_norm": 0.37011975049972534, "learning_rate": 2.178789046806326e-05, "loss": 0.0264, "step": 70630 }, { "grad_norm": 0.7892542481422424, "learning_rate": 2.1774240810000547e-05, "loss": 0.0253, "step": 70640 }, { "grad_norm": 0.2738957107067108, "learning_rate": 2.17605942386644e-05, "loss": 0.0311, "step": 70650 }, { "grad_norm": 0.27909359335899353, "learning_rate": 2.174695075554723e-05, "loss": 0.0362, "step": 70660 }, { "grad_norm": 0.597409725189209, "learning_rate": 2.173331036214103e-05, "loss": 0.0351, "step": 70670 }, { "grad_norm": 0.2664053738117218, "learning_rate": 2.171967305993753e-05, "loss": 0.0255, "step": 70680 }, { "grad_norm": 0.26165226101875305, "learning_rate": 2.1706038850428024e-05, "loss": 0.0268, "step": 70690 }, { "grad_norm": 0.38660240173339844, "learning_rate": 2.169240773510361e-05, "loss": 0.023, "step": 70700 }, { "grad_norm": 0.3727162182331085, "learning_rate": 2.1678779715454906e-05, "loss": 0.0271, "step": 70710 }, { "grad_norm": 0.29694926738739014, "learning_rate": 2.166515479297229e-05, "loss": 0.0212, "step": 70720 }, { "grad_norm": 0.3277170956134796, "learning_rate": 2.1651532969145732e-05, "loss": 0.0277, "step": 70730 }, { "grad_norm": 0.21760118007659912, "learning_rate": 2.1637914245464903e-05, "loss": 0.0232, "step": 70740 }, { "grad_norm": 0.35853585600852966, "learning_rate": 2.1624298623419133e-05, "loss": 0.0218, "step": 70750 }, { "grad_norm": 0.6089755892753601, "learning_rate": 2.1610686104497413e-05, "loss": 0.0243, "step": 70760 }, { "grad_norm": 0.6037455201148987, "learning_rate": 2.1597076690188366e-05, "loss": 0.0354, "step": 70770 }, { "grad_norm": 0.4013507664203644, "learning_rate": 2.1583470381980315e-05, "loss": 0.0358, "step": 70780 }, { "grad_norm": 0.22052329778671265, "learning_rate": 2.156986718136123e-05, "loss": 0.0252, "step": 70790 }, { "grad_norm": 0.487627774477005, "learning_rate": 2.155626708981871e-05, "loss": 0.0259, "step": 70800 }, { "grad_norm": 0.23554904758930206, "learning_rate": 2.1542670108840075e-05, "loss": 0.0286, "step": 70810 }, { "grad_norm": 0.33753159642219543, "learning_rate": 2.152907623991221e-05, "loss": 0.0339, "step": 70820 }, { "grad_norm": 0.20937858521938324, "learning_rate": 2.1515485484521802e-05, "loss": 0.0246, "step": 70830 }, { "grad_norm": 0.2834346890449524, "learning_rate": 2.150189784415506e-05, "loss": 0.0288, "step": 70840 }, { "grad_norm": 0.2797103822231293, "learning_rate": 2.148831332029794e-05, "loss": 0.0216, "step": 70850 }, { "grad_norm": 0.24208810925483704, "learning_rate": 2.147473191443599e-05, "loss": 0.0284, "step": 70860 }, { "grad_norm": 0.21230590343475342, "learning_rate": 2.1461153628054492e-05, "loss": 0.0238, "step": 70870 }, { "grad_norm": 0.3838275372982025, "learning_rate": 2.1447578462638284e-05, "loss": 0.0248, "step": 70880 }, { "grad_norm": 0.2913326025009155, "learning_rate": 2.143400641967201e-05, "loss": 0.0215, "step": 70890 }, { "grad_norm": 0.22440291941165924, "learning_rate": 2.1420437500639823e-05, "loss": 0.0194, "step": 70900 }, { "grad_norm": 0.5563549995422363, "learning_rate": 2.1406871707025646e-05, "loss": 0.0268, "step": 70910 }, { "grad_norm": 0.5052012205123901, "learning_rate": 2.1393309040312963e-05, "loss": 0.0307, "step": 70920 }, { "grad_norm": 0.5636849403381348, "learning_rate": 2.137974950198502e-05, "loss": 0.031, "step": 70930 }, { "grad_norm": 0.27101606130599976, "learning_rate": 2.1366193093524617e-05, "loss": 0.0215, "step": 70940 }, { "grad_norm": 0.5325884222984314, "learning_rate": 2.1352639816414288e-05, "loss": 0.0283, "step": 70950 }, { "grad_norm": 0.2549327611923218, "learning_rate": 2.1339089672136202e-05, "loss": 0.0289, "step": 70960 }, { "grad_norm": 0.21596044301986694, "learning_rate": 2.1325542662172193e-05, "loss": 0.0309, "step": 70970 }, { "grad_norm": 0.2578366994857788, "learning_rate": 2.1311998788003712e-05, "loss": 0.0224, "step": 70980 }, { "grad_norm": 0.7501856684684753, "learning_rate": 2.1298458051111907e-05, "loss": 0.0334, "step": 70990 }, { "grad_norm": 0.22453249990940094, "learning_rate": 2.1284920452977602e-05, "loss": 0.022, "step": 71000 }, { "grad_norm": 0.1623517870903015, "learning_rate": 2.12713859950812e-05, "loss": 0.0225, "step": 71010 }, { "grad_norm": 0.28148359060287476, "learning_rate": 2.1257854678902832e-05, "loss": 0.0199, "step": 71020 }, { "grad_norm": 0.399578332901001, "learning_rate": 2.124432650592227e-05, "loss": 0.0313, "step": 71030 }, { "grad_norm": 0.37551698088645935, "learning_rate": 2.1230801477618937e-05, "loss": 0.0262, "step": 71040 }, { "grad_norm": 0.4127725660800934, "learning_rate": 2.121727959547189e-05, "loss": 0.0278, "step": 71050 }, { "grad_norm": 0.3885327875614166, "learning_rate": 2.1203760860959886e-05, "loss": 0.0272, "step": 71060 }, { "grad_norm": 0.3594990372657776, "learning_rate": 2.119024527556127e-05, "loss": 0.0363, "step": 71070 }, { "grad_norm": 0.3162837326526642, "learning_rate": 2.117673284075415e-05, "loss": 0.0199, "step": 71080 }, { "grad_norm": 0.3307225704193115, "learning_rate": 2.1163223558016165e-05, "loss": 0.0299, "step": 71090 }, { "grad_norm": 0.23140841722488403, "learning_rate": 2.1149717428824722e-05, "loss": 0.0271, "step": 71100 }, { "grad_norm": 0.44785696268081665, "learning_rate": 2.1136214454656784e-05, "loss": 0.0273, "step": 71110 }, { "grad_norm": 0.3994174003601074, "learning_rate": 2.1122714636989056e-05, "loss": 0.0335, "step": 71120 }, { "grad_norm": 0.46953827142715454, "learning_rate": 2.1109217977297823e-05, "loss": 0.0243, "step": 71130 }, { "grad_norm": 2.3163001537323, "learning_rate": 2.1095724477059077e-05, "loss": 0.0214, "step": 71140 }, { "grad_norm": 0.23586739599704742, "learning_rate": 2.1082234137748434e-05, "loss": 0.0189, "step": 71150 }, { "grad_norm": 0.37581679224967957, "learning_rate": 2.106874696084122e-05, "loss": 0.0289, "step": 71160 }, { "grad_norm": 0.3554871380329132, "learning_rate": 2.1055262947812315e-05, "loss": 0.026, "step": 71170 }, { "grad_norm": 0.29459837079048157, "learning_rate": 2.104178210013636e-05, "loss": 0.0339, "step": 71180 }, { "grad_norm": 0.4014396071434021, "learning_rate": 2.1028304419287554e-05, "loss": 0.0287, "step": 71190 }, { "grad_norm": 0.3590620458126068, "learning_rate": 2.101482990673982e-05, "loss": 0.0275, "step": 71200 }, { "grad_norm": 0.3699647784233093, "learning_rate": 2.1001358563966707e-05, "loss": 0.0286, "step": 71210 }, { "grad_norm": 0.28036800026893616, "learning_rate": 2.098789039244144e-05, "loss": 0.0215, "step": 71220 }, { "grad_norm": 0.29132041335105896, "learning_rate": 2.097442539363685e-05, "loss": 0.0246, "step": 71230 }, { "grad_norm": 0.21808312833309174, "learning_rate": 2.0960963569025448e-05, "loss": 0.0283, "step": 71240 }, { "grad_norm": 0.24694624543190002, "learning_rate": 2.0947504920079435e-05, "loss": 0.0208, "step": 71250 }, { "grad_norm": 0.21980728209018707, "learning_rate": 2.093404944827059e-05, "loss": 0.0276, "step": 71260 }, { "grad_norm": 0.31359246373176575, "learning_rate": 2.0920597155070394e-05, "loss": 0.0235, "step": 71270 }, { "grad_norm": 0.6621271967887878, "learning_rate": 2.090714804194997e-05, "loss": 0.0371, "step": 71280 }, { "grad_norm": 0.27112895250320435, "learning_rate": 2.0893702110380114e-05, "loss": 0.0272, "step": 71290 }, { "grad_norm": 0.36675164103507996, "learning_rate": 2.0880259361831216e-05, "loss": 0.0242, "step": 71300 }, { "grad_norm": 0.3186339735984802, "learning_rate": 2.0866819797773397e-05, "loss": 0.0239, "step": 71310 }, { "grad_norm": 0.3037383556365967, "learning_rate": 2.085338341967634e-05, "loss": 0.0253, "step": 71320 }, { "grad_norm": 0.2642776072025299, "learning_rate": 2.083995022900946e-05, "loss": 0.0258, "step": 71330 }, { "grad_norm": 0.36725887656211853, "learning_rate": 2.0826520227241768e-05, "loss": 0.0291, "step": 71340 }, { "grad_norm": 0.4305676817893982, "learning_rate": 2.081309341584199e-05, "loss": 0.0266, "step": 71350 }, { "grad_norm": 0.2755759656429291, "learning_rate": 2.0799669796278413e-05, "loss": 0.0231, "step": 71360 }, { "grad_norm": 0.2206800878047943, "learning_rate": 2.078624937001906e-05, "loss": 0.0335, "step": 71370 }, { "grad_norm": 0.39541471004486084, "learning_rate": 2.0772832138531544e-05, "loss": 0.0463, "step": 71380 }, { "grad_norm": 0.32739412784576416, "learning_rate": 2.0759418103283158e-05, "loss": 0.0272, "step": 71390 }, { "grad_norm": 0.6350943446159363, "learning_rate": 2.0746007265740846e-05, "loss": 0.0227, "step": 71400 }, { "grad_norm": 0.2664525508880615, "learning_rate": 2.0732599627371215e-05, "loss": 0.0297, "step": 71410 }, { "grad_norm": 0.28664055466651917, "learning_rate": 2.0719195189640468e-05, "loss": 0.0247, "step": 71420 }, { "grad_norm": 0.32767486572265625, "learning_rate": 2.0705793954014528e-05, "loss": 0.0223, "step": 71430 }, { "grad_norm": 0.2826199233531952, "learning_rate": 2.06923959219589e-05, "loss": 0.0224, "step": 71440 }, { "grad_norm": 0.324268102645874, "learning_rate": 2.0679001094938793e-05, "loss": 0.0223, "step": 71450 }, { "grad_norm": 0.2740802466869354, "learning_rate": 2.066560947441904e-05, "loss": 0.0298, "step": 71460 }, { "grad_norm": 0.2668878734111786, "learning_rate": 2.0652221061864152e-05, "loss": 0.0188, "step": 71470 }, { "grad_norm": 0.3813106119632721, "learning_rate": 2.0638835858738226e-05, "loss": 0.0294, "step": 71480 }, { "grad_norm": 0.3789593279361725, "learning_rate": 2.0625453866505068e-05, "loss": 0.0255, "step": 71490 }, { "grad_norm": 0.27249154448509216, "learning_rate": 2.0612075086628123e-05, "loss": 0.0228, "step": 71500 }, { "grad_norm": 0.2726310193538666, "learning_rate": 2.059869952057045e-05, "loss": 0.0237, "step": 71510 }, { "grad_norm": 0.5976571440696716, "learning_rate": 2.0585327169794793e-05, "loss": 0.0271, "step": 71520 }, { "grad_norm": 0.28640690445899963, "learning_rate": 2.0571958035763534e-05, "loss": 0.0246, "step": 71530 }, { "grad_norm": 0.47063684463500977, "learning_rate": 2.0558592119938715e-05, "loss": 0.0286, "step": 71540 }, { "grad_norm": 0.22970356047153473, "learning_rate": 2.054522942378198e-05, "loss": 0.0213, "step": 71550 }, { "grad_norm": 0.5762863159179688, "learning_rate": 2.053186994875469e-05, "loss": 0.0268, "step": 71560 }, { "grad_norm": 0.31842494010925293, "learning_rate": 2.0518513696317775e-05, "loss": 0.031, "step": 71570 }, { "grad_norm": 0.3194846510887146, "learning_rate": 2.050516066793188e-05, "loss": 0.0268, "step": 71580 }, { "grad_norm": 0.2976844310760498, "learning_rate": 2.0491810865057276e-05, "loss": 0.0215, "step": 71590 }, { "grad_norm": 0.36287370324134827, "learning_rate": 2.0478464289153882e-05, "loss": 0.0292, "step": 71600 }, { "grad_norm": 1.339074969291687, "learning_rate": 2.0465120941681232e-05, "loss": 0.0279, "step": 71610 }, { "grad_norm": 0.32295897603034973, "learning_rate": 2.0451780824098566e-05, "loss": 0.0221, "step": 71620 }, { "grad_norm": 0.22921374440193176, "learning_rate": 2.043844393786471e-05, "loss": 0.0228, "step": 71630 }, { "grad_norm": 0.548225998878479, "learning_rate": 2.0425110284438193e-05, "loss": 0.0251, "step": 71640 }, { "grad_norm": 0.37865957617759705, "learning_rate": 2.041177986527712e-05, "loss": 0.0268, "step": 71650 }, { "grad_norm": 0.31847476959228516, "learning_rate": 2.0398452681839347e-05, "loss": 0.0224, "step": 71660 }, { "grad_norm": 0.3700447678565979, "learning_rate": 2.0385128735582272e-05, "loss": 0.0371, "step": 71670 }, { "grad_norm": 0.2667224705219269, "learning_rate": 2.0371808027963003e-05, "loss": 0.02, "step": 71680 }, { "grad_norm": 0.43697452545166016, "learning_rate": 2.035849056043825e-05, "loss": 0.0344, "step": 71690 }, { "grad_norm": 0.23086495697498322, "learning_rate": 2.03451763344644e-05, "loss": 0.0207, "step": 71700 }, { "grad_norm": 0.3352591097354889, "learning_rate": 2.0331865351497482e-05, "loss": 0.0257, "step": 71710 }, { "grad_norm": 0.3367053270339966, "learning_rate": 2.0318557612993185e-05, "loss": 0.0239, "step": 71720 }, { "grad_norm": 0.394720196723938, "learning_rate": 2.0305253120406782e-05, "loss": 0.0272, "step": 71730 }, { "grad_norm": 0.22456909716129303, "learning_rate": 2.029195187519325e-05, "loss": 0.0222, "step": 71740 }, { "grad_norm": 0.18526257574558258, "learning_rate": 2.0278653878807217e-05, "loss": 0.026, "step": 71750 }, { "grad_norm": 0.21864528954029083, "learning_rate": 2.0265359132702892e-05, "loss": 0.0367, "step": 71760 }, { "grad_norm": 0.40484875440597534, "learning_rate": 2.0252067638334182e-05, "loss": 0.027, "step": 71770 }, { "grad_norm": 0.4125880300998688, "learning_rate": 2.023877939715464e-05, "loss": 0.0313, "step": 71780 }, { "grad_norm": 0.3440977931022644, "learning_rate": 2.0225494410617448e-05, "loss": 0.0259, "step": 71790 }, { "grad_norm": 0.26941296458244324, "learning_rate": 2.02122126801754e-05, "loss": 0.0281, "step": 71800 }, { "grad_norm": 0.23440705239772797, "learning_rate": 2.0198934207281018e-05, "loss": 0.028, "step": 71810 }, { "grad_norm": 0.23283697664737701, "learning_rate": 2.0185658993386357e-05, "loss": 0.0213, "step": 71820 }, { "grad_norm": 0.3405207693576813, "learning_rate": 2.0172387039943226e-05, "loss": 0.0235, "step": 71830 }, { "grad_norm": 0.25178101658821106, "learning_rate": 2.0159118348402968e-05, "loss": 0.0299, "step": 71840 }, { "grad_norm": 0.5313416123390198, "learning_rate": 2.0145852920216697e-05, "loss": 0.0249, "step": 71850 }, { "grad_norm": 0.5582094788551331, "learning_rate": 2.0132590756835045e-05, "loss": 0.0286, "step": 71860 }, { "grad_norm": 0.30485373735427856, "learning_rate": 2.0119331859708386e-05, "loss": 0.0366, "step": 71870 }, { "grad_norm": 0.2661415934562683, "learning_rate": 2.0106076230286654e-05, "loss": 0.0253, "step": 71880 }, { "grad_norm": 0.2974143326282501, "learning_rate": 2.0092823870019488e-05, "loss": 0.0251, "step": 71890 }, { "grad_norm": 0.3441869020462036, "learning_rate": 2.0079574780356115e-05, "loss": 0.0228, "step": 71900 }, { "grad_norm": 0.2865171432495117, "learning_rate": 2.0066328962745486e-05, "loss": 0.0262, "step": 71910 }, { "grad_norm": 0.5708495378494263, "learning_rate": 2.00530864186361e-05, "loss": 0.0217, "step": 71920 }, { "grad_norm": 0.2797876000404358, "learning_rate": 2.003984714947618e-05, "loss": 0.0311, "step": 71930 }, { "grad_norm": 0.30608445405960083, "learning_rate": 2.0026611156713515e-05, "loss": 0.0249, "step": 71940 }, { "grad_norm": 0.4214079678058624, "learning_rate": 2.0013378441795584e-05, "loss": 0.0244, "step": 71950 }, { "grad_norm": 0.3199009597301483, "learning_rate": 2.0000149006169506e-05, "loss": 0.0222, "step": 71960 }, { "grad_norm": 0.2851945161819458, "learning_rate": 1.9986922851282048e-05, "loss": 0.0281, "step": 71970 }, { "grad_norm": 0.2696289122104645, "learning_rate": 1.997369997857956e-05, "loss": 0.0268, "step": 71980 }, { "grad_norm": 0.5921023488044739, "learning_rate": 1.9960480389508103e-05, "loss": 0.027, "step": 71990 }, { "grad_norm": 0.33361414074897766, "learning_rate": 1.9947264085513363e-05, "loss": 0.0372, "step": 72000 }, { "grad_norm": 0.39034050703048706, "learning_rate": 1.993405106804062e-05, "loss": 0.0244, "step": 72010 }, { "grad_norm": 0.5146234035491943, "learning_rate": 1.9920841338534867e-05, "loss": 0.0321, "step": 72020 }, { "grad_norm": 0.30710262060165405, "learning_rate": 1.9907634898440646e-05, "loss": 0.0213, "step": 72030 }, { "grad_norm": 0.29144230484962463, "learning_rate": 1.989443174920227e-05, "loss": 0.0267, "step": 72040 }, { "grad_norm": 0.4174785315990448, "learning_rate": 1.988123189226355e-05, "loss": 0.0375, "step": 72050 }, { "grad_norm": 0.3068155348300934, "learning_rate": 1.986803532906805e-05, "loss": 0.0297, "step": 72060 }, { "grad_norm": 0.36535072326660156, "learning_rate": 1.985484206105889e-05, "loss": 0.0271, "step": 72070 }, { "grad_norm": 0.3553001880645752, "learning_rate": 1.9841652089678887e-05, "loss": 0.022, "step": 72080 }, { "grad_norm": 0.2887152433395386, "learning_rate": 1.9828465416370435e-05, "loss": 0.035, "step": 72090 }, { "grad_norm": 0.3546724021434784, "learning_rate": 1.9815282042575685e-05, "loss": 0.0299, "step": 72100 }, { "grad_norm": 0.1178593710064888, "learning_rate": 1.9802101969736285e-05, "loss": 0.0316, "step": 72110 }, { "grad_norm": 0.49377983808517456, "learning_rate": 1.9788925199293633e-05, "loss": 0.0312, "step": 72120 }, { "grad_norm": 0.2926473021507263, "learning_rate": 1.977575173268868e-05, "loss": 0.0259, "step": 72130 }, { "grad_norm": 0.35249295830726624, "learning_rate": 1.9762581571362094e-05, "loss": 0.0341, "step": 72140 }, { "grad_norm": 0.2645074129104614, "learning_rate": 1.9749414716754112e-05, "loss": 0.027, "step": 72150 }, { "grad_norm": 0.25013381242752075, "learning_rate": 1.973625117030465e-05, "loss": 0.0238, "step": 72160 }, { "grad_norm": 2.36200213432312, "learning_rate": 1.972309093345326e-05, "loss": 0.022, "step": 72170 }, { "grad_norm": 0.3473149836063385, "learning_rate": 1.970993400763915e-05, "loss": 0.0284, "step": 72180 }, { "grad_norm": 0.2876284718513489, "learning_rate": 1.969678039430109e-05, "loss": 0.0332, "step": 72190 }, { "grad_norm": 0.4157734811306, "learning_rate": 1.9683630094877564e-05, "loss": 0.0245, "step": 72200 }, { "grad_norm": 0.23464106023311615, "learning_rate": 1.9670483110806696e-05, "loss": 0.0236, "step": 72210 }, { "grad_norm": 0.4896998405456543, "learning_rate": 1.9657339443526172e-05, "loss": 0.0275, "step": 72220 }, { "grad_norm": 0.23185527324676514, "learning_rate": 1.964419909447339e-05, "loss": 0.03, "step": 72230 }, { "grad_norm": 0.438314288854599, "learning_rate": 1.963106206508536e-05, "loss": 0.0285, "step": 72240 }, { "grad_norm": 0.2041362076997757, "learning_rate": 1.9617928356798738e-05, "loss": 0.0237, "step": 72250 }, { "grad_norm": 0.17216487228870392, "learning_rate": 1.960479797104977e-05, "loss": 0.0237, "step": 72260 }, { "grad_norm": 0.6599195599555969, "learning_rate": 1.9591670909274418e-05, "loss": 0.0244, "step": 72270 }, { "grad_norm": 0.14786924421787262, "learning_rate": 1.9578547172908184e-05, "loss": 0.0283, "step": 72280 }, { "grad_norm": 0.30438151955604553, "learning_rate": 1.9565426763386325e-05, "loss": 0.024, "step": 72290 }, { "grad_norm": 0.36499497294425964, "learning_rate": 1.9552309682143622e-05, "loss": 0.0239, "step": 72300 }, { "grad_norm": 0.23288512229919434, "learning_rate": 1.9539195930614574e-05, "loss": 0.0232, "step": 72310 }, { "grad_norm": 0.677183985710144, "learning_rate": 1.952608551023324e-05, "loss": 0.0269, "step": 72320 }, { "grad_norm": 0.38983941078186035, "learning_rate": 1.9512978422433397e-05, "loss": 0.0251, "step": 72330 }, { "grad_norm": 0.23277322947978973, "learning_rate": 1.9499874668648376e-05, "loss": 0.0198, "step": 72340 }, { "grad_norm": 0.2068178653717041, "learning_rate": 1.9486774250311213e-05, "loss": 0.0229, "step": 72350 }, { "grad_norm": 0.29455527663230896, "learning_rate": 1.9473677168854537e-05, "loss": 0.0288, "step": 72360 }, { "grad_norm": 0.2656118869781494, "learning_rate": 1.946058342571065e-05, "loss": 0.0293, "step": 72370 }, { "grad_norm": 0.2986272871494293, "learning_rate": 1.9447493022311424e-05, "loss": 0.0257, "step": 72380 }, { "grad_norm": 0.37981754541397095, "learning_rate": 1.9434405960088442e-05, "loss": 0.0239, "step": 72390 }, { "grad_norm": 0.28115570545196533, "learning_rate": 1.9421322240472857e-05, "loss": 0.0223, "step": 72400 }, { "grad_norm": 0.29781538248062134, "learning_rate": 1.9408241864895494e-05, "loss": 0.0248, "step": 72410 }, { "grad_norm": 0.3581990599632263, "learning_rate": 1.939516483478681e-05, "loss": 0.0267, "step": 72420 }, { "grad_norm": 0.6126925349235535, "learning_rate": 1.9382091151576902e-05, "loss": 0.0225, "step": 72430 }, { "grad_norm": 0.2872713506221771, "learning_rate": 1.9369020816695454e-05, "loss": 0.0222, "step": 72440 }, { "grad_norm": 0.19419102370738983, "learning_rate": 1.935595383157184e-05, "loss": 0.0243, "step": 72450 }, { "grad_norm": 0.35894688963890076, "learning_rate": 1.9342890197635057e-05, "loss": 0.0191, "step": 72460 }, { "grad_norm": 0.24348412454128265, "learning_rate": 1.9329829916313685e-05, "loss": 0.0293, "step": 72470 }, { "grad_norm": 0.48942920565605164, "learning_rate": 1.9316772989036013e-05, "loss": 0.028, "step": 72480 }, { "grad_norm": 0.35071173310279846, "learning_rate": 1.930371941722991e-05, "loss": 0.0234, "step": 72490 }, { "grad_norm": 0.20559026300907135, "learning_rate": 1.929066920232292e-05, "loss": 0.0218, "step": 72500 }, { "grad_norm": 0.48702701926231384, "learning_rate": 1.9277622345742152e-05, "loss": 0.0308, "step": 72510 }, { "grad_norm": 0.3666938245296478, "learning_rate": 1.926457884891444e-05, "loss": 0.0263, "step": 72520 }, { "grad_norm": 0.34534189105033875, "learning_rate": 1.9251538713266147e-05, "loss": 0.022, "step": 72530 }, { "grad_norm": 0.27355340123176575, "learning_rate": 1.9238501940223358e-05, "loss": 0.0292, "step": 72540 }, { "grad_norm": 0.3806118965148926, "learning_rate": 1.922546853121175e-05, "loss": 0.0281, "step": 72550 }, { "grad_norm": 0.2425469607114792, "learning_rate": 1.9212438487656647e-05, "loss": 0.0223, "step": 72560 }, { "grad_norm": 0.19467920064926147, "learning_rate": 1.9199411810982965e-05, "loss": 0.0238, "step": 72570 }, { "grad_norm": 0.19461765885353088, "learning_rate": 1.918638850261532e-05, "loss": 0.0268, "step": 72580 }, { "grad_norm": 0.23949328064918518, "learning_rate": 1.9173368563977885e-05, "loss": 0.0251, "step": 72590 }, { "grad_norm": 0.3084748685359955, "learning_rate": 1.9160351996494513e-05, "loss": 0.0259, "step": 72600 }, { "grad_norm": 0.2775467038154602, "learning_rate": 1.9147338801588677e-05, "loss": 0.025, "step": 72610 }, { "grad_norm": 0.3804927468299866, "learning_rate": 1.913432898068351e-05, "loss": 0.0312, "step": 72620 }, { "grad_norm": 0.25017160177230835, "learning_rate": 1.912132253520169e-05, "loss": 0.0338, "step": 72630 }, { "grad_norm": 0.4215507209300995, "learning_rate": 1.910831946656564e-05, "loss": 0.0336, "step": 72640 }, { "grad_norm": 0.33993396162986755, "learning_rate": 1.909531977619731e-05, "loss": 0.0406, "step": 72650 }, { "grad_norm": 0.3777015507221222, "learning_rate": 1.908232346551834e-05, "loss": 0.0204, "step": 72660 }, { "grad_norm": 0.40140432119369507, "learning_rate": 1.9069330535949993e-05, "loss": 0.0324, "step": 72670 }, { "grad_norm": 0.3294384181499481, "learning_rate": 1.9056340988913167e-05, "loss": 0.0218, "step": 72680 }, { "grad_norm": 0.30525949597358704, "learning_rate": 1.9043354825828357e-05, "loss": 0.0263, "step": 72690 }, { "grad_norm": 0.21318940818309784, "learning_rate": 1.9030372048115724e-05, "loss": 0.0267, "step": 72700 }, { "grad_norm": 0.3478826582431793, "learning_rate": 1.9017392657195056e-05, "loss": 0.0312, "step": 72710 }, { "grad_norm": 0.4079189598560333, "learning_rate": 1.9004416654485724e-05, "loss": 0.0311, "step": 72720 }, { "grad_norm": 0.31521230936050415, "learning_rate": 1.8991444041406785e-05, "loss": 0.0292, "step": 72730 }, { "grad_norm": 0.6115574836730957, "learning_rate": 1.8978474819376912e-05, "loss": 0.0283, "step": 72740 }, { "grad_norm": 0.32842373847961426, "learning_rate": 1.8965508989814406e-05, "loss": 0.0259, "step": 72750 }, { "grad_norm": 0.41351747512817383, "learning_rate": 1.8952546554137158e-05, "loss": 0.0278, "step": 72760 }, { "grad_norm": 0.28703776001930237, "learning_rate": 1.8939587513762757e-05, "loss": 0.0215, "step": 72770 }, { "grad_norm": 0.34765148162841797, "learning_rate": 1.892663187010835e-05, "loss": 0.0212, "step": 72780 }, { "grad_norm": 0.28911542892456055, "learning_rate": 1.891367962459077e-05, "loss": 0.0235, "step": 72790 }, { "grad_norm": 0.4001156985759735, "learning_rate": 1.890073077862644e-05, "loss": 0.0273, "step": 72800 }, { "grad_norm": 0.32693010568618774, "learning_rate": 1.888778533363146e-05, "loss": 0.0231, "step": 72810 }, { "grad_norm": 0.2614266276359558, "learning_rate": 1.8874843291021478e-05, "loss": 0.0205, "step": 72820 }, { "grad_norm": 0.264546662569046, "learning_rate": 1.8861904652211853e-05, "loss": 0.0263, "step": 72830 }, { "grad_norm": 0.3668625056743622, "learning_rate": 1.8848969418617507e-05, "loss": 0.0248, "step": 72840 }, { "grad_norm": 0.33605459332466125, "learning_rate": 1.8836037591653044e-05, "loss": 0.0246, "step": 72850 }, { "grad_norm": 0.3925876319408417, "learning_rate": 1.882310917273262e-05, "loss": 0.0257, "step": 72860 }, { "grad_norm": 0.4872449040412903, "learning_rate": 1.881018416327014e-05, "loss": 0.032, "step": 72870 }, { "grad_norm": 0.26440533995628357, "learning_rate": 1.8797262564679014e-05, "loss": 0.0288, "step": 72880 }, { "grad_norm": 0.38439738750457764, "learning_rate": 1.8784344378372355e-05, "loss": 0.0236, "step": 72890 }, { "grad_norm": 0.2835499048233032, "learning_rate": 1.8771429605762843e-05, "loss": 0.0253, "step": 72900 }, { "grad_norm": 0.33419403433799744, "learning_rate": 1.8758518248262835e-05, "loss": 0.0271, "step": 72910 }, { "grad_norm": 0.32288020849227905, "learning_rate": 1.8745610307284305e-05, "loss": 0.0306, "step": 72920 }, { "grad_norm": 0.3993387222290039, "learning_rate": 1.8732705784238862e-05, "loss": 0.023, "step": 72930 }, { "grad_norm": 0.3543510437011719, "learning_rate": 1.8719804680537682e-05, "loss": 0.0286, "step": 72940 }, { "grad_norm": 0.42750757932662964, "learning_rate": 1.870690699759164e-05, "loss": 0.0268, "step": 72950 }, { "grad_norm": 0.22657598555088043, "learning_rate": 1.8694012736811213e-05, "loss": 0.026, "step": 72960 }, { "grad_norm": 0.36163976788520813, "learning_rate": 1.8681121899606463e-05, "loss": 0.0244, "step": 72970 }, { "grad_norm": 0.3858596682548523, "learning_rate": 1.8668234487387144e-05, "loss": 0.0292, "step": 72980 }, { "grad_norm": 0.1686481386423111, "learning_rate": 1.8655350501562592e-05, "loss": 0.0219, "step": 72990 }, { "grad_norm": 0.24978971481323242, "learning_rate": 1.8642469943541796e-05, "loss": 0.0337, "step": 73000 }, { "grad_norm": 0.2578376233577728, "learning_rate": 1.8629592814733328e-05, "loss": 0.0229, "step": 73010 }, { "grad_norm": 0.3411617875099182, "learning_rate": 1.8616719116545438e-05, "loss": 0.0323, "step": 73020 }, { "grad_norm": 0.27671489119529724, "learning_rate": 1.8603848850385952e-05, "loss": 0.0211, "step": 73030 }, { "grad_norm": 0.38444003462791443, "learning_rate": 1.859098201766236e-05, "loss": 0.0196, "step": 73040 }, { "grad_norm": 0.2982536256313324, "learning_rate": 1.8578118619781726e-05, "loss": 0.0224, "step": 73050 }, { "grad_norm": 0.2038395255804062, "learning_rate": 1.856525865815083e-05, "loss": 0.022, "step": 73060 }, { "grad_norm": 0.20900869369506836, "learning_rate": 1.8552402134175966e-05, "loss": 0.0209, "step": 73070 }, { "grad_norm": 0.30958253145217896, "learning_rate": 1.8539549049263143e-05, "loss": 0.0236, "step": 73080 }, { "grad_norm": 0.2013976126909256, "learning_rate": 1.8526699404817917e-05, "loss": 0.0227, "step": 73090 }, { "grad_norm": 0.4338489770889282, "learning_rate": 1.8513853202245546e-05, "loss": 0.0249, "step": 73100 }, { "grad_norm": 0.3370823264122009, "learning_rate": 1.850101044295081e-05, "loss": 0.0277, "step": 73110 }, { "grad_norm": 0.4466955363750458, "learning_rate": 1.848817112833825e-05, "loss": 0.0166, "step": 73120 }, { "grad_norm": 1.2551852464675903, "learning_rate": 1.8475335259811905e-05, "loss": 0.0272, "step": 73130 }, { "grad_norm": 0.35868847370147705, "learning_rate": 1.8462502838775508e-05, "loss": 0.0294, "step": 73140 }, { "grad_norm": 0.30871275067329407, "learning_rate": 1.8449673866632378e-05, "loss": 0.0287, "step": 73150 }, { "grad_norm": 0.25763002038002014, "learning_rate": 1.8436848344785473e-05, "loss": 0.0287, "step": 73160 }, { "grad_norm": 0.18330439925193787, "learning_rate": 1.8424026274637378e-05, "loss": 0.0266, "step": 73170 }, { "grad_norm": 0.3158337473869324, "learning_rate": 1.8411207657590312e-05, "loss": 0.0278, "step": 73180 }, { "grad_norm": 0.21031779050827026, "learning_rate": 1.839839249504607e-05, "loss": 0.0244, "step": 73190 }, { "grad_norm": 0.2629567086696625, "learning_rate": 1.8385580788406108e-05, "loss": 0.0285, "step": 73200 }, { "grad_norm": 0.3138394355773926, "learning_rate": 1.837277253907152e-05, "loss": 0.0296, "step": 73210 }, { "grad_norm": 0.3008146286010742, "learning_rate": 1.8359967748442957e-05, "loss": 0.025, "step": 73220 }, { "grad_norm": 0.32510650157928467, "learning_rate": 1.8347166417920774e-05, "loss": 0.02, "step": 73230 }, { "grad_norm": 0.38408398628234863, "learning_rate": 1.833436854890484e-05, "loss": 0.0327, "step": 73240 }, { "grad_norm": 0.3416590392589569, "learning_rate": 1.8321574142794796e-05, "loss": 0.0208, "step": 73250 }, { "grad_norm": 0.3599562644958496, "learning_rate": 1.8308783200989748e-05, "loss": 0.0214, "step": 73260 }, { "grad_norm": 0.2932867407798767, "learning_rate": 1.829599572488855e-05, "loss": 0.0259, "step": 73270 }, { "grad_norm": 0.1986427754163742, "learning_rate": 1.828321171588957e-05, "loss": 0.0182, "step": 73280 }, { "grad_norm": 0.5225970149040222, "learning_rate": 1.827043117539089e-05, "loss": 0.0277, "step": 73290 }, { "grad_norm": 0.302275687456131, "learning_rate": 1.8257654104790116e-05, "loss": 0.0235, "step": 73300 }, { "grad_norm": 0.28066638112068176, "learning_rate": 1.82448805054846e-05, "loss": 0.0303, "step": 73310 }, { "grad_norm": 0.33989042043685913, "learning_rate": 1.823211037887119e-05, "loss": 0.0308, "step": 73320 }, { "grad_norm": 0.3354913890361786, "learning_rate": 1.821934372634645e-05, "loss": 0.0231, "step": 73330 }, { "grad_norm": 0.3137936294078827, "learning_rate": 1.8206580549306472e-05, "loss": 0.0221, "step": 73340 }, { "grad_norm": 0.18802212178707123, "learning_rate": 1.819382084914707e-05, "loss": 0.0264, "step": 73350 }, { "grad_norm": 0.24252481758594513, "learning_rate": 1.8181064627263577e-05, "loss": 0.0218, "step": 73360 }, { "grad_norm": 0.45723602175712585, "learning_rate": 1.8168311885051022e-05, "loss": 0.029, "step": 73370 }, { "grad_norm": 0.2709207832813263, "learning_rate": 1.815556262390402e-05, "loss": 0.0187, "step": 73380 }, { "grad_norm": 0.23617522418498993, "learning_rate": 1.8142816845216826e-05, "loss": 0.0304, "step": 73390 }, { "grad_norm": 0.3412196636199951, "learning_rate": 1.8130074550383276e-05, "loss": 0.0223, "step": 73400 }, { "grad_norm": 0.41991081833839417, "learning_rate": 1.8117335740796848e-05, "loss": 0.0266, "step": 73410 }, { "grad_norm": 0.4639677107334137, "learning_rate": 1.810460041785067e-05, "loss": 0.0214, "step": 73420 }, { "grad_norm": 0.7286753058433533, "learning_rate": 1.809186858293742e-05, "loss": 0.0313, "step": 73430 }, { "grad_norm": 0.26952439546585083, "learning_rate": 1.8079140237449448e-05, "loss": 0.0201, "step": 73440 }, { "grad_norm": 0.27992820739746094, "learning_rate": 1.8066415382778707e-05, "loss": 0.0284, "step": 73450 }, { "grad_norm": 0.19816824793815613, "learning_rate": 1.8053694020316786e-05, "loss": 0.0191, "step": 73460 }, { "grad_norm": 0.27838000655174255, "learning_rate": 1.8040976151454837e-05, "loss": 0.0265, "step": 73470 }, { "grad_norm": 0.2562176287174225, "learning_rate": 1.8028261777583704e-05, "loss": 0.0375, "step": 73480 }, { "grad_norm": 0.47596481442451477, "learning_rate": 1.8015550900093758e-05, "loss": 0.0316, "step": 73490 }, { "grad_norm": 0.498580664396286, "learning_rate": 1.800284352037512e-05, "loss": 0.0214, "step": 73500 }, { "grad_norm": 0.3304491937160492, "learning_rate": 1.799013963981738e-05, "loss": 0.0335, "step": 73510 }, { "grad_norm": 0.45380762219429016, "learning_rate": 1.797743925980987e-05, "loss": 0.0253, "step": 73520 }, { "grad_norm": 0.3871804475784302, "learning_rate": 1.7964742381741435e-05, "loss": 0.0262, "step": 73530 }, { "grad_norm": 0.3479921817779541, "learning_rate": 1.7952049007000627e-05, "loss": 0.0315, "step": 73540 }, { "grad_norm": 0.4202686846256256, "learning_rate": 1.7939359136975535e-05, "loss": 0.0223, "step": 73550 }, { "grad_norm": 0.2587648928165436, "learning_rate": 1.7926672773053932e-05, "loss": 0.0243, "step": 73560 }, { "grad_norm": 0.4983285069465637, "learning_rate": 1.791398991662317e-05, "loss": 0.0245, "step": 73570 }, { "grad_norm": 0.1940310150384903, "learning_rate": 1.7901310569070244e-05, "loss": 0.0266, "step": 73580 }, { "grad_norm": 0.2855778932571411, "learning_rate": 1.7888634731781724e-05, "loss": 0.0212, "step": 73590 }, { "grad_norm": 0.6276835799217224, "learning_rate": 1.7875962406143844e-05, "loss": 0.0227, "step": 73600 }, { "grad_norm": 0.33928734064102173, "learning_rate": 1.7863293593542403e-05, "loss": 0.037, "step": 73610 }, { "grad_norm": 0.22394117712974548, "learning_rate": 1.7850628295362853e-05, "loss": 0.0248, "step": 73620 }, { "grad_norm": 0.803307056427002, "learning_rate": 1.7837966512990263e-05, "loss": 0.0304, "step": 73630 }, { "grad_norm": 0.32624173164367676, "learning_rate": 1.782530824780931e-05, "loss": 0.0226, "step": 73640 }, { "grad_norm": 0.33091455698013306, "learning_rate": 1.7812653501204263e-05, "loss": 0.0266, "step": 73650 }, { "grad_norm": 0.7312990427017212, "learning_rate": 1.780000227455903e-05, "loss": 0.0329, "step": 73660 }, { "grad_norm": 0.4326801002025604, "learning_rate": 1.778735456925716e-05, "loss": 0.0277, "step": 73670 }, { "grad_norm": 0.17504918575286865, "learning_rate": 1.7774710386681737e-05, "loss": 0.0217, "step": 73680 }, { "grad_norm": 0.22679416835308075, "learning_rate": 1.7762069728215545e-05, "loss": 0.0306, "step": 73690 }, { "grad_norm": 0.3317752182483673, "learning_rate": 1.7749432595240932e-05, "loss": 0.0232, "step": 73700 }, { "grad_norm": 0.24299444258213043, "learning_rate": 1.773679898913991e-05, "loss": 0.0319, "step": 73710 }, { "grad_norm": 0.38453230261802673, "learning_rate": 1.7724168911294015e-05, "loss": 0.0245, "step": 73720 }, { "grad_norm": 0.2803893983364105, "learning_rate": 1.7711542363084504e-05, "loss": 0.0199, "step": 73730 }, { "grad_norm": 0.2035234123468399, "learning_rate": 1.7698919345892158e-05, "loss": 0.02, "step": 73740 }, { "grad_norm": 0.2527289092540741, "learning_rate": 1.768629986109743e-05, "loss": 0.0291, "step": 73750 }, { "grad_norm": 0.4702320396900177, "learning_rate": 1.7673683910080357e-05, "loss": 0.0236, "step": 73760 }, { "grad_norm": 0.15380066633224487, "learning_rate": 1.7661071494220638e-05, "loss": 0.0224, "step": 73770 }, { "grad_norm": 0.19560055434703827, "learning_rate": 1.7648462614897498e-05, "loss": 0.0251, "step": 73780 }, { "grad_norm": 0.6268113851547241, "learning_rate": 1.7635857273489863e-05, "loss": 0.0214, "step": 73790 }, { "grad_norm": 0.2691386044025421, "learning_rate": 1.7623255471376198e-05, "loss": 0.0231, "step": 73800 }, { "grad_norm": 0.35824212431907654, "learning_rate": 1.7610657209934646e-05, "loss": 0.0324, "step": 73810 }, { "grad_norm": 0.2537190318107605, "learning_rate": 1.7598062490542915e-05, "loss": 0.0253, "step": 73820 }, { "grad_norm": 0.2984588146209717, "learning_rate": 1.7585471314578377e-05, "loss": 0.0328, "step": 73830 }, { "grad_norm": 0.3554581105709076, "learning_rate": 1.7572883683417945e-05, "loss": 0.0273, "step": 73840 }, { "grad_norm": 0.3532453775405884, "learning_rate": 1.7560299598438212e-05, "loss": 0.0257, "step": 73850 }, { "grad_norm": 0.263553649187088, "learning_rate": 1.7547719061015334e-05, "loss": 0.0238, "step": 73860 }, { "grad_norm": 0.40561729669570923, "learning_rate": 1.75351420725251e-05, "loss": 0.0212, "step": 73870 }, { "grad_norm": 0.3609612286090851, "learning_rate": 1.7522568634342923e-05, "loss": 0.0283, "step": 73880 }, { "grad_norm": 0.4916824400424957, "learning_rate": 1.7509998747843825e-05, "loss": 0.0218, "step": 73890 }, { "grad_norm": 0.34846773743629456, "learning_rate": 1.7497432414402403e-05, "loss": 0.026, "step": 73900 }, { "grad_norm": 0.32307010889053345, "learning_rate": 1.7484869635392903e-05, "loss": 0.0252, "step": 73910 }, { "grad_norm": 0.29872676730155945, "learning_rate": 1.747231041218919e-05, "loss": 0.0204, "step": 73920 }, { "grad_norm": 0.19291871786117554, "learning_rate": 1.7459754746164692e-05, "loss": 0.0236, "step": 73930 }, { "grad_norm": 0.3809531331062317, "learning_rate": 1.744720263869249e-05, "loss": 0.0302, "step": 73940 }, { "grad_norm": 0.31945037841796875, "learning_rate": 1.7434654091145263e-05, "loss": 0.0273, "step": 73950 }, { "grad_norm": 0.3173755407333374, "learning_rate": 1.742210910489532e-05, "loss": 0.0218, "step": 73960 }, { "grad_norm": 0.13520748913288116, "learning_rate": 1.7409567681314532e-05, "loss": 0.0218, "step": 73970 }, { "grad_norm": 0.21001531183719635, "learning_rate": 1.7397029821774434e-05, "loss": 0.0229, "step": 73980 }, { "grad_norm": 0.38954469561576843, "learning_rate": 1.7384495527646126e-05, "loss": 0.0286, "step": 73990 }, { "grad_norm": 0.3543959856033325, "learning_rate": 1.737196480030035e-05, "loss": 0.0184, "step": 74000 }, { "grad_norm": 0.2663611173629761, "learning_rate": 1.7359437641107447e-05, "loss": 0.0255, "step": 74010 }, { "grad_norm": 0.23848487436771393, "learning_rate": 1.7346914051437387e-05, "loss": 0.0246, "step": 74020 }, { "grad_norm": 0.2836967706680298, "learning_rate": 1.733439403265969e-05, "loss": 0.022, "step": 74030 }, { "grad_norm": 0.23803499341011047, "learning_rate": 1.732187758614357e-05, "loss": 0.034, "step": 74040 }, { "grad_norm": 0.28853392601013184, "learning_rate": 1.730936471325777e-05, "loss": 0.0217, "step": 74050 }, { "grad_norm": 0.20452381670475006, "learning_rate": 1.7296855415370718e-05, "loss": 0.0246, "step": 74060 }, { "grad_norm": 0.2751493453979492, "learning_rate": 1.728434969385035e-05, "loss": 0.0245, "step": 74070 }, { "grad_norm": 0.29174327850341797, "learning_rate": 1.7271847550064346e-05, "loss": 0.0239, "step": 74080 }, { "grad_norm": 0.4415053427219391, "learning_rate": 1.725934898537987e-05, "loss": 0.0293, "step": 74090 }, { "grad_norm": 0.26733890175819397, "learning_rate": 1.7246854001163783e-05, "loss": 0.0197, "step": 74100 }, { "grad_norm": 0.3992881178855896, "learning_rate": 1.7234362598782476e-05, "loss": 0.0246, "step": 74110 }, { "grad_norm": 0.7806869745254517, "learning_rate": 1.722187477960202e-05, "loss": 0.0256, "step": 74120 }, { "grad_norm": 0.49722281098365784, "learning_rate": 1.7209390544988057e-05, "loss": 0.0226, "step": 74130 }, { "grad_norm": 0.33195558190345764, "learning_rate": 1.7196909896305856e-05, "loss": 0.033, "step": 74140 }, { "grad_norm": 0.21452149748802185, "learning_rate": 1.718443283492026e-05, "loss": 0.0269, "step": 74150 }, { "grad_norm": 0.32268857955932617, "learning_rate": 1.7171959362195746e-05, "loss": 0.0249, "step": 74160 }, { "grad_norm": 0.21740181744098663, "learning_rate": 1.715948947949642e-05, "loss": 0.0282, "step": 74170 }, { "grad_norm": 0.28659501671791077, "learning_rate": 1.714702318818593e-05, "loss": 0.0223, "step": 74180 }, { "grad_norm": 0.3023329973220825, "learning_rate": 1.7134560489627592e-05, "loss": 0.0191, "step": 74190 }, { "grad_norm": 0.29063472151756287, "learning_rate": 1.712210138518431e-05, "loss": 0.0191, "step": 74200 }, { "grad_norm": 0.2869492769241333, "learning_rate": 1.71096458762186e-05, "loss": 0.034, "step": 74210 }, { "grad_norm": 0.33720308542251587, "learning_rate": 1.7097193964092552e-05, "loss": 0.0322, "step": 74220 }, { "grad_norm": 0.4741527736186981, "learning_rate": 1.708474565016792e-05, "loss": 0.0244, "step": 74230 }, { "grad_norm": 0.27676165103912354, "learning_rate": 1.7072300935806e-05, "loss": 0.0302, "step": 74240 }, { "grad_norm": 0.41453462839126587, "learning_rate": 1.705985982236776e-05, "loss": 0.0235, "step": 74250 }, { "grad_norm": 0.26961463689804077, "learning_rate": 1.7047422311213697e-05, "loss": 0.0224, "step": 74260 }, { "grad_norm": 0.37125998735427856, "learning_rate": 1.703498840370401e-05, "loss": 0.036, "step": 74270 }, { "grad_norm": 0.2877926528453827, "learning_rate": 1.702255810119842e-05, "loss": 0.0238, "step": 74280 }, { "grad_norm": 0.33932727575302124, "learning_rate": 1.7010131405056305e-05, "loss": 0.019, "step": 74290 }, { "grad_norm": 0.26258981227874756, "learning_rate": 1.6997708316636613e-05, "loss": 0.0237, "step": 74300 }, { "grad_norm": 0.2683219313621521, "learning_rate": 1.6985288837297932e-05, "loss": 0.0256, "step": 74310 }, { "grad_norm": 0.30771705508232117, "learning_rate": 1.6972872968398394e-05, "loss": 0.0246, "step": 74320 }, { "grad_norm": 0.24726589024066925, "learning_rate": 1.6960460711295845e-05, "loss": 0.0224, "step": 74330 }, { "grad_norm": 0.2591922879219055, "learning_rate": 1.6948052067347623e-05, "loss": 0.0224, "step": 74340 }, { "grad_norm": 0.37911269068717957, "learning_rate": 1.693564703791075e-05, "loss": 0.0274, "step": 74350 }, { "grad_norm": 0.23174485564231873, "learning_rate": 1.6923245624341793e-05, "loss": 0.0229, "step": 74360 }, { "grad_norm": 0.2647063732147217, "learning_rate": 1.6910847827996962e-05, "loss": 0.019, "step": 74370 }, { "grad_norm": 0.26492804288864136, "learning_rate": 1.689845365023206e-05, "loss": 0.0239, "step": 74380 }, { "grad_norm": 0.1448015570640564, "learning_rate": 1.6886063092402505e-05, "loss": 0.0214, "step": 74390 }, { "grad_norm": 0.9989524483680725, "learning_rate": 1.6873676155863317e-05, "loss": 0.0336, "step": 74400 }, { "grad_norm": 0.23344489932060242, "learning_rate": 1.6861292841969085e-05, "loss": 0.0273, "step": 74410 }, { "grad_norm": 0.28605860471725464, "learning_rate": 1.684891315207406e-05, "loss": 0.0244, "step": 74420 }, { "grad_norm": 0.24264679849147797, "learning_rate": 1.6836537087532034e-05, "loss": 0.0234, "step": 74430 }, { "grad_norm": 0.36772558093070984, "learning_rate": 1.6824164649696465e-05, "loss": 0.0325, "step": 74440 }, { "grad_norm": 0.509433925151825, "learning_rate": 1.681179583992034e-05, "loss": 0.0211, "step": 74450 }, { "grad_norm": 0.28361091017723083, "learning_rate": 1.679943065955636e-05, "loss": 0.0258, "step": 74460 }, { "grad_norm": 1.4759811162948608, "learning_rate": 1.678706910995671e-05, "loss": 0.0248, "step": 74470 }, { "grad_norm": 0.3780427575111389, "learning_rate": 1.6774711192473262e-05, "loss": 0.0281, "step": 74480 }, { "grad_norm": 0.2678402364253998, "learning_rate": 1.676235690845743e-05, "loss": 0.017, "step": 74490 }, { "grad_norm": 0.4781596064567566, "learning_rate": 1.6750006259260285e-05, "loss": 0.0301, "step": 74500 }, { "grad_norm": 0.3006407618522644, "learning_rate": 1.673765924623244e-05, "loss": 0.0264, "step": 74510 }, { "grad_norm": 0.3773621618747711, "learning_rate": 1.67253158707242e-05, "loss": 0.02, "step": 74520 }, { "grad_norm": 0.27690574526786804, "learning_rate": 1.671297613408537e-05, "loss": 0.0221, "step": 74530 }, { "grad_norm": 0.42829039692878723, "learning_rate": 1.6700640037665444e-05, "loss": 0.0209, "step": 74540 }, { "grad_norm": 0.20129571855068207, "learning_rate": 1.668830758281344e-05, "loss": 0.0244, "step": 74550 }, { "grad_norm": 0.30959072709083557, "learning_rate": 1.6675978770878052e-05, "loss": 0.0212, "step": 74560 }, { "grad_norm": 0.22342626750469208, "learning_rate": 1.6663653603207506e-05, "loss": 0.0194, "step": 74570 }, { "grad_norm": 0.21730181574821472, "learning_rate": 1.6651332081149685e-05, "loss": 0.0262, "step": 74580 }, { "grad_norm": 0.6860787868499756, "learning_rate": 1.6639014206052046e-05, "loss": 0.0208, "step": 74590 }, { "grad_norm": 0.27627408504486084, "learning_rate": 1.6626699979261657e-05, "loss": 0.0214, "step": 74600 }, { "grad_norm": 0.30346834659576416, "learning_rate": 1.6614389402125197e-05, "loss": 0.0302, "step": 74610 }, { "grad_norm": 0.2961379289627075, "learning_rate": 1.66020824759889e-05, "loss": 0.0208, "step": 74620 }, { "grad_norm": 0.2642354369163513, "learning_rate": 1.6589779202198675e-05, "loss": 0.0246, "step": 74630 }, { "grad_norm": 0.44228023290634155, "learning_rate": 1.6577479582099924e-05, "loss": 0.023, "step": 74640 }, { "grad_norm": 0.36779364943504333, "learning_rate": 1.6565183617037788e-05, "loss": 0.032, "step": 74650 }, { "grad_norm": 0.37367504835128784, "learning_rate": 1.6552891308356894e-05, "loss": 0.0261, "step": 74660 }, { "grad_norm": 0.3391002118587494, "learning_rate": 1.6540602657401526e-05, "loss": 0.0334, "step": 74670 }, { "grad_norm": 0.2585870623588562, "learning_rate": 1.6528317665515537e-05, "loss": 0.0326, "step": 74680 }, { "grad_norm": 0.2775113582611084, "learning_rate": 1.6516036334042416e-05, "loss": 0.0315, "step": 74690 }, { "grad_norm": 0.2174369990825653, "learning_rate": 1.650375866432519e-05, "loss": 0.0309, "step": 74700 }, { "grad_norm": 0.5622478127479553, "learning_rate": 1.649148465770659e-05, "loss": 0.0247, "step": 74710 }, { "grad_norm": 0.2754703760147095, "learning_rate": 1.647921431552883e-05, "loss": 0.0254, "step": 74720 }, { "grad_norm": 0.27064087986946106, "learning_rate": 1.646694763913381e-05, "loss": 0.0317, "step": 74730 }, { "grad_norm": 0.2596451938152313, "learning_rate": 1.6454684629862966e-05, "loss": 0.0258, "step": 74740 }, { "grad_norm": 0.2381836473941803, "learning_rate": 1.6442425289057388e-05, "loss": 0.0207, "step": 74750 }, { "grad_norm": 0.24773097038269043, "learning_rate": 1.6430169618057718e-05, "loss": 0.02, "step": 74760 }, { "grad_norm": 0.2302320897579193, "learning_rate": 1.641791761820422e-05, "loss": 0.0234, "step": 74770 }, { "grad_norm": 0.3343163728713989, "learning_rate": 1.640566929083676e-05, "loss": 0.0272, "step": 74780 }, { "grad_norm": 0.3294295072555542, "learning_rate": 1.6393424637294823e-05, "loss": 0.0329, "step": 74790 }, { "grad_norm": 0.25321897864341736, "learning_rate": 1.638118365891742e-05, "loss": 0.0222, "step": 74800 }, { "grad_norm": 0.36968597769737244, "learning_rate": 1.636894635704323e-05, "loss": 0.0292, "step": 74810 }, { "grad_norm": 0.2194240242242813, "learning_rate": 1.6356712733010517e-05, "loss": 0.0218, "step": 74820 }, { "grad_norm": 0.6455695629119873, "learning_rate": 1.6344482788157096e-05, "loss": 0.0237, "step": 74830 }, { "grad_norm": 0.20411914587020874, "learning_rate": 1.6332256523820444e-05, "loss": 0.0193, "step": 74840 }, { "grad_norm": 0.5347601175308228, "learning_rate": 1.6320033941337597e-05, "loss": 0.0306, "step": 74850 }, { "grad_norm": 0.2700745165348053, "learning_rate": 1.6307815042045216e-05, "loss": 0.0195, "step": 74860 }, { "grad_norm": 0.30433231592178345, "learning_rate": 1.6295599827279513e-05, "loss": 0.023, "step": 74870 }, { "grad_norm": 0.6188828945159912, "learning_rate": 1.6283388298376356e-05, "loss": 0.02, "step": 74880 }, { "grad_norm": 0.6791161894798279, "learning_rate": 1.6271180456671125e-05, "loss": 0.0286, "step": 74890 }, { "grad_norm": 0.3270478844642639, "learning_rate": 1.625897630349893e-05, "loss": 0.0203, "step": 74900 }, { "grad_norm": 0.2989868223667145, "learning_rate": 1.6246775840194335e-05, "loss": 0.0215, "step": 74910 }, { "grad_norm": 0.15811938047409058, "learning_rate": 1.6234579068091604e-05, "loss": 0.0191, "step": 74920 }, { "grad_norm": 0.2343270480632782, "learning_rate": 1.6222385988524524e-05, "loss": 0.0257, "step": 74930 }, { "grad_norm": 0.20489256083965302, "learning_rate": 1.6210196602826544e-05, "loss": 0.0258, "step": 74940 }, { "grad_norm": 0.33371803164482117, "learning_rate": 1.6198010912330646e-05, "loss": 0.0223, "step": 74950 }, { "grad_norm": 0.3027611970901489, "learning_rate": 1.6185828918369456e-05, "loss": 0.0272, "step": 74960 }, { "grad_norm": 0.21895091235637665, "learning_rate": 1.617365062227517e-05, "loss": 0.0199, "step": 74970 }, { "grad_norm": 0.19045887887477875, "learning_rate": 1.616147602537962e-05, "loss": 0.0312, "step": 74980 }, { "grad_norm": 0.38929760456085205, "learning_rate": 1.614930512901415e-05, "loss": 0.0249, "step": 74990 }, { "grad_norm": 0.7501670122146606, "learning_rate": 1.6137137934509804e-05, "loss": 0.0211, "step": 75000 }, { "grad_norm": 0.4520491063594818, "learning_rate": 1.6124974443197123e-05, "loss": 0.0221, "step": 75010 }, { "grad_norm": 0.3313966393470764, "learning_rate": 1.6112814656406306e-05, "loss": 0.0282, "step": 75020 }, { "grad_norm": 0.3058722913265228, "learning_rate": 1.6100658575467125e-05, "loss": 0.0224, "step": 75030 }, { "grad_norm": 0.2972303032875061, "learning_rate": 1.6088506201708974e-05, "loss": 0.0211, "step": 75040 }, { "grad_norm": 0.5004336833953857, "learning_rate": 1.6076357536460785e-05, "loss": 0.0225, "step": 75050 }, { "grad_norm": 0.2447364628314972, "learning_rate": 1.606421258105113e-05, "loss": 0.0226, "step": 75060 }, { "grad_norm": 0.3213229179382324, "learning_rate": 1.6052071336808183e-05, "loss": 0.026, "step": 75070 }, { "grad_norm": 0.3296554982662201, "learning_rate": 1.6039933805059658e-05, "loss": 0.0282, "step": 75080 }, { "grad_norm": 0.2498762160539627, "learning_rate": 1.6027799987132915e-05, "loss": 0.0314, "step": 75090 }, { "grad_norm": 0.3785558342933655, "learning_rate": 1.6015669884354888e-05, "loss": 0.0337, "step": 75100 }, { "grad_norm": 0.2509055435657501, "learning_rate": 1.6003543498052125e-05, "loss": 0.0271, "step": 75110 }, { "grad_norm": 0.3541279733181, "learning_rate": 1.5991420829550715e-05, "loss": 0.0289, "step": 75120 }, { "grad_norm": 0.18743698298931122, "learning_rate": 1.5979301880176405e-05, "loss": 0.0327, "step": 75130 }, { "grad_norm": 0.2466639280319214, "learning_rate": 1.5967186651254485e-05, "loss": 0.0257, "step": 75140 }, { "grad_norm": 0.7607827186584473, "learning_rate": 1.5955075144109853e-05, "loss": 0.0263, "step": 75150 }, { "grad_norm": 0.2709948718547821, "learning_rate": 1.594296736006703e-05, "loss": 0.0224, "step": 75160 }, { "grad_norm": 0.3198957145214081, "learning_rate": 1.59308633004501e-05, "loss": 0.031, "step": 75170 }, { "grad_norm": 0.2612261176109314, "learning_rate": 1.5918762966582724e-05, "loss": 0.026, "step": 75180 }, { "grad_norm": 0.459625244140625, "learning_rate": 1.5906666359788203e-05, "loss": 0.0244, "step": 75190 }, { "grad_norm": 0.19074538350105286, "learning_rate": 1.589457348138938e-05, "loss": 0.0244, "step": 75200 }, { "grad_norm": 0.5656468868255615, "learning_rate": 1.588248433270872e-05, "loss": 0.0251, "step": 75210 }, { "grad_norm": 0.40144023299217224, "learning_rate": 1.587039891506828e-05, "loss": 0.0242, "step": 75220 }, { "grad_norm": 0.2190934121608734, "learning_rate": 1.5858317229789715e-05, "loss": 0.0227, "step": 75230 }, { "grad_norm": 0.37755826115608215, "learning_rate": 1.584623927819423e-05, "loss": 0.0282, "step": 75240 }, { "grad_norm": 0.46023520827293396, "learning_rate": 1.583416506160269e-05, "loss": 0.0337, "step": 75250 }, { "grad_norm": 0.34379786252975464, "learning_rate": 1.5822094581335472e-05, "loss": 0.024, "step": 75260 }, { "grad_norm": 0.19288316369056702, "learning_rate": 1.5810027838712604e-05, "loss": 0.0243, "step": 75270 }, { "grad_norm": 0.23969903588294983, "learning_rate": 1.5797964835053687e-05, "loss": 0.0297, "step": 75280 }, { "grad_norm": 0.27248021960258484, "learning_rate": 1.5785905571677933e-05, "loss": 0.024, "step": 75290 }, { "grad_norm": 0.30202212929725647, "learning_rate": 1.5773850049904086e-05, "loss": 0.0226, "step": 75300 }, { "grad_norm": 0.205240860581398, "learning_rate": 1.5761798271050543e-05, "loss": 0.0215, "step": 75310 }, { "grad_norm": 0.28273382782936096, "learning_rate": 1.5749750236435277e-05, "loss": 0.0257, "step": 75320 }, { "grad_norm": 0.4279954135417938, "learning_rate": 1.5737705947375815e-05, "loss": 0.0221, "step": 75330 }, { "grad_norm": 0.3816946744918823, "learning_rate": 1.5725665405189322e-05, "loss": 0.0259, "step": 75340 }, { "grad_norm": 0.24210061132907867, "learning_rate": 1.571362861119253e-05, "loss": 0.0225, "step": 75350 }, { "grad_norm": 0.21481168270111084, "learning_rate": 1.570159556670177e-05, "loss": 0.02, "step": 75360 }, { "grad_norm": 0.1463865339756012, "learning_rate": 1.5689566273032946e-05, "loss": 0.0206, "step": 75370 }, { "grad_norm": 0.2865101993083954, "learning_rate": 1.5677540731501583e-05, "loss": 0.0232, "step": 75380 }, { "grad_norm": 0.29310598969459534, "learning_rate": 1.5665518943422746e-05, "loss": 0.0263, "step": 75390 }, { "grad_norm": 0.3874848186969757, "learning_rate": 1.5653500910111134e-05, "loss": 0.0292, "step": 75400 }, { "grad_norm": 0.3697910010814667, "learning_rate": 1.5641486632881023e-05, "loss": 0.0247, "step": 75410 }, { "grad_norm": 1.0450011491775513, "learning_rate": 1.56294761130463e-05, "loss": 0.0266, "step": 75420 }, { "grad_norm": 0.3111347556114197, "learning_rate": 1.5617469351920373e-05, "loss": 0.0274, "step": 75430 }, { "grad_norm": 0.357509046792984, "learning_rate": 1.5605466350816315e-05, "loss": 0.0307, "step": 75440 }, { "grad_norm": 0.3287336826324463, "learning_rate": 1.559346711104674e-05, "loss": 0.0234, "step": 75450 }, { "grad_norm": 0.3767964243888855, "learning_rate": 1.5581471633923883e-05, "loss": 0.029, "step": 75460 }, { "grad_norm": 0.4033772349357605, "learning_rate": 1.5569479920759505e-05, "loss": 0.0292, "step": 75470 }, { "grad_norm": 0.3122657835483551, "learning_rate": 1.5557491972865072e-05, "loss": 0.0257, "step": 75480 }, { "grad_norm": 0.28308671712875366, "learning_rate": 1.554550779155152e-05, "loss": 0.021, "step": 75490 }, { "grad_norm": 0.33856669068336487, "learning_rate": 1.5533527378129453e-05, "loss": 0.02, "step": 75500 }, { "grad_norm": 0.3053337335586548, "learning_rate": 1.552155073390899e-05, "loss": 0.0295, "step": 75510 }, { "grad_norm": 0.4901852011680603, "learning_rate": 1.5509577860199903e-05, "loss": 0.0327, "step": 75520 }, { "grad_norm": 0.1796761453151703, "learning_rate": 1.5497608758311525e-05, "loss": 0.0223, "step": 75530 }, { "grad_norm": 0.3207537829875946, "learning_rate": 1.5485643429552797e-05, "loss": 0.0303, "step": 75540 }, { "grad_norm": 0.25495806336402893, "learning_rate": 1.5473681875232197e-05, "loss": 0.0271, "step": 75550 }, { "grad_norm": 0.2771497964859009, "learning_rate": 1.5461724096657837e-05, "loss": 0.0227, "step": 75560 }, { "grad_norm": 0.34179821610450745, "learning_rate": 1.5449770095137416e-05, "loss": 0.0204, "step": 75570 }, { "grad_norm": 0.24955587089061737, "learning_rate": 1.5437819871978178e-05, "loss": 0.023, "step": 75580 }, { "grad_norm": 0.5595307350158691, "learning_rate": 1.5425873428486992e-05, "loss": 0.0266, "step": 75590 }, { "grad_norm": 0.33719736337661743, "learning_rate": 1.5413930765970312e-05, "loss": 0.025, "step": 75600 }, { "grad_norm": 0.37934592366218567, "learning_rate": 1.5401991885734175e-05, "loss": 0.0237, "step": 75610 }, { "grad_norm": 0.40950289368629456, "learning_rate": 1.5390056789084174e-05, "loss": 0.0206, "step": 75620 }, { "grad_norm": 0.469500869512558, "learning_rate": 1.5378125477325545e-05, "loss": 0.0207, "step": 75630 }, { "grad_norm": 0.2745424807071686, "learning_rate": 1.5366197951763046e-05, "loss": 0.0225, "step": 75640 }, { "grad_norm": 0.5827099084854126, "learning_rate": 1.5354274213701075e-05, "loss": 0.0303, "step": 75650 }, { "grad_norm": 0.35291486978530884, "learning_rate": 1.5342354264443566e-05, "loss": 0.0222, "step": 75660 }, { "grad_norm": 0.3114040195941925, "learning_rate": 1.533043810529411e-05, "loss": 0.0246, "step": 75670 }, { "grad_norm": 0.31682026386260986, "learning_rate": 1.5318525737555807e-05, "loss": 0.0275, "step": 75680 }, { "grad_norm": 0.27606531977653503, "learning_rate": 1.53066171625314e-05, "loss": 0.0242, "step": 75690 }, { "grad_norm": 0.7270466089248657, "learning_rate": 1.529471238152317e-05, "loss": 0.0241, "step": 75700 }, { "grad_norm": 0.35335394740104675, "learning_rate": 1.5282811395833023e-05, "loss": 0.0284, "step": 75710 }, { "grad_norm": 0.3828858435153961, "learning_rate": 1.5270914206762397e-05, "loss": 0.0242, "step": 75720 }, { "grad_norm": 0.38943520188331604, "learning_rate": 1.525902081561241e-05, "loss": 0.021, "step": 75730 }, { "grad_norm": 0.2567819058895111, "learning_rate": 1.5247131223683664e-05, "loss": 0.0277, "step": 75740 }, { "grad_norm": 0.28152772784233093, "learning_rate": 1.5235245432276408e-05, "loss": 0.026, "step": 75750 }, { "grad_norm": 0.15046913921833038, "learning_rate": 1.5223363442690424e-05, "loss": 0.0234, "step": 75760 }, { "grad_norm": 0.21547558903694153, "learning_rate": 1.5211485256225133e-05, "loss": 0.0216, "step": 75770 }, { "grad_norm": 0.3400561809539795, "learning_rate": 1.519961087417951e-05, "loss": 0.0215, "step": 75780 }, { "grad_norm": 0.3575708568096161, "learning_rate": 1.5187740297852137e-05, "loss": 0.0271, "step": 75790 }, { "grad_norm": 0.17089155316352844, "learning_rate": 1.517587352854113e-05, "loss": 0.0194, "step": 75800 }, { "grad_norm": 0.8500841856002808, "learning_rate": 1.516401056754423e-05, "loss": 0.0251, "step": 75810 }, { "grad_norm": 0.40390270948410034, "learning_rate": 1.5152151416158783e-05, "loss": 0.0324, "step": 75820 }, { "grad_norm": 0.3058927059173584, "learning_rate": 1.5140296075681642e-05, "loss": 0.0285, "step": 75830 }, { "grad_norm": 0.28490135073661804, "learning_rate": 1.5128444547409331e-05, "loss": 0.025, "step": 75840 }, { "grad_norm": 0.3806664049625397, "learning_rate": 1.5116596832637864e-05, "loss": 0.0207, "step": 75850 }, { "grad_norm": 0.2941676676273346, "learning_rate": 1.5104752932662952e-05, "loss": 0.0275, "step": 75860 }, { "grad_norm": 0.3525025546550751, "learning_rate": 1.5092912848779778e-05, "loss": 0.0275, "step": 75870 }, { "grad_norm": 0.24396854639053345, "learning_rate": 1.5081076582283194e-05, "loss": 0.021, "step": 75880 }, { "grad_norm": 0.2617489993572235, "learning_rate": 1.5069244134467553e-05, "loss": 0.0221, "step": 75890 }, { "grad_norm": 0.24432472884655, "learning_rate": 1.5057415506626876e-05, "loss": 0.0226, "step": 75900 }, { "grad_norm": 0.42647334933280945, "learning_rate": 1.5045590700054668e-05, "loss": 0.0236, "step": 75910 }, { "grad_norm": 0.2818884551525116, "learning_rate": 1.5033769716044144e-05, "loss": 0.0241, "step": 75920 }, { "grad_norm": 0.8604782223701477, "learning_rate": 1.5021952555887975e-05, "loss": 0.019, "step": 75930 }, { "grad_norm": 0.3477618098258972, "learning_rate": 1.50101392208785e-05, "loss": 0.0236, "step": 75940 }, { "grad_norm": 0.31545501947402954, "learning_rate": 1.4998329712307579e-05, "loss": 0.0196, "step": 75950 }, { "grad_norm": 0.46023398637771606, "learning_rate": 1.4986524031466708e-05, "loss": 0.0251, "step": 75960 }, { "grad_norm": 0.4396417737007141, "learning_rate": 1.4974722179646916e-05, "loss": 0.0172, "step": 75970 }, { "grad_norm": 0.3761748671531677, "learning_rate": 1.4962924158138847e-05, "loss": 0.027, "step": 75980 }, { "grad_norm": 0.254892498254776, "learning_rate": 1.4951129968232714e-05, "loss": 0.0199, "step": 75990 }, { "grad_norm": 0.4255969226360321, "learning_rate": 1.4939339611218329e-05, "loss": 0.0262, "step": 76000 }, { "grad_norm": 0.31703221797943115, "learning_rate": 1.4927553088385032e-05, "loss": 0.0312, "step": 76010 }, { "grad_norm": 0.33353477716445923, "learning_rate": 1.4915770401021801e-05, "loss": 0.0208, "step": 76020 }, { "grad_norm": 0.2887047529220581, "learning_rate": 1.4903991550417185e-05, "loss": 0.0264, "step": 76030 }, { "grad_norm": 0.44241639971733093, "learning_rate": 1.4892216537859277e-05, "loss": 0.0275, "step": 76040 }, { "grad_norm": 0.4035335183143616, "learning_rate": 1.4880445364635775e-05, "loss": 0.0249, "step": 76050 }, { "grad_norm": 0.3916884660720825, "learning_rate": 1.4868678032033973e-05, "loss": 0.0265, "step": 76060 }, { "grad_norm": 0.29525941610336304, "learning_rate": 1.4856914541340739e-05, "loss": 0.0276, "step": 76070 }, { "grad_norm": 0.21284057199954987, "learning_rate": 1.4845154893842472e-05, "loss": 0.0251, "step": 76080 }, { "grad_norm": 0.4258953332901001, "learning_rate": 1.4833399090825234e-05, "loss": 0.0314, "step": 76090 }, { "grad_norm": 0.43839505314826965, "learning_rate": 1.4821647133574562e-05, "loss": 0.0274, "step": 76100 }, { "grad_norm": 0.288499116897583, "learning_rate": 1.4809899023375706e-05, "loss": 0.0224, "step": 76110 }, { "grad_norm": 0.4223928451538086, "learning_rate": 1.4798154761513366e-05, "loss": 0.0272, "step": 76120 }, { "grad_norm": 0.3095605671405792, "learning_rate": 1.4786414349271916e-05, "loss": 0.0309, "step": 76130 }, { "grad_norm": 0.5486137270927429, "learning_rate": 1.4774677787935226e-05, "loss": 0.0308, "step": 76140 }, { "grad_norm": 0.30342280864715576, "learning_rate": 1.4762945078786838e-05, "loss": 0.0201, "step": 76150 }, { "grad_norm": 0.37442851066589355, "learning_rate": 1.4751216223109777e-05, "loss": 0.0249, "step": 76160 }, { "grad_norm": 0.31935909390449524, "learning_rate": 1.4739491222186714e-05, "loss": 0.0281, "step": 76170 }, { "grad_norm": 0.5066449046134949, "learning_rate": 1.4727770077299879e-05, "loss": 0.0248, "step": 76180 }, { "grad_norm": 0.21583549678325653, "learning_rate": 1.4716052789731094e-05, "loss": 0.0201, "step": 76190 }, { "grad_norm": 0.5886379480361938, "learning_rate": 1.470433936076171e-05, "loss": 0.0277, "step": 76200 }, { "grad_norm": 0.2354913204908371, "learning_rate": 1.4692629791672719e-05, "loss": 0.0227, "step": 76210 }, { "grad_norm": 0.39959126710891724, "learning_rate": 1.468092408374464e-05, "loss": 0.0212, "step": 76220 }, { "grad_norm": 0.35473671555519104, "learning_rate": 1.4669222238257602e-05, "loss": 0.0292, "step": 76230 }, { "grad_norm": 0.28600576519966125, "learning_rate": 1.4657524256491301e-05, "loss": 0.0198, "step": 76240 }, { "grad_norm": 0.35231468081474304, "learning_rate": 1.464583013972502e-05, "loss": 0.0243, "step": 76250 }, { "grad_norm": 0.28904032707214355, "learning_rate": 1.4634139889237591e-05, "loss": 0.0276, "step": 76260 }, { "grad_norm": 0.8181517124176025, "learning_rate": 1.4622453506307448e-05, "loss": 0.0261, "step": 76270 }, { "grad_norm": 0.41113975644111633, "learning_rate": 1.461077099221262e-05, "loss": 0.0278, "step": 76280 }, { "grad_norm": 0.22952301800251007, "learning_rate": 1.4599092348230652e-05, "loss": 0.0245, "step": 76290 }, { "grad_norm": 0.21636579930782318, "learning_rate": 1.4587417575638724e-05, "loss": 0.0224, "step": 76300 }, { "grad_norm": 0.8444675803184509, "learning_rate": 1.4575746675713563e-05, "loss": 0.0209, "step": 76310 }, { "grad_norm": 0.24539954960346222, "learning_rate": 1.4564079649731504e-05, "loss": 0.0271, "step": 76320 }, { "grad_norm": 0.7037820219993591, "learning_rate": 1.4552416498968402e-05, "loss": 0.0238, "step": 76330 }, { "grad_norm": 0.35262447595596313, "learning_rate": 1.4540757224699747e-05, "loss": 0.0223, "step": 76340 }, { "grad_norm": 0.719666063785553, "learning_rate": 1.452910182820056e-05, "loss": 0.028, "step": 76350 }, { "grad_norm": 0.36323317885398865, "learning_rate": 1.4517450310745468e-05, "loss": 0.0232, "step": 76360 }, { "grad_norm": 0.35986173152923584, "learning_rate": 1.450580267360866e-05, "loss": 0.0274, "step": 76370 }, { "grad_norm": 0.2544575333595276, "learning_rate": 1.4494158918063922e-05, "loss": 0.0223, "step": 76380 }, { "grad_norm": 0.3502855896949768, "learning_rate": 1.4482519045384568e-05, "loss": 0.0223, "step": 76390 }, { "grad_norm": 0.4023662805557251, "learning_rate": 1.4470883056843548e-05, "loss": 0.0232, "step": 76400 }, { "grad_norm": 0.3371494710445404, "learning_rate": 1.4459250953713322e-05, "loss": 0.0267, "step": 76410 }, { "grad_norm": 0.2969000041484833, "learning_rate": 1.444762273726598e-05, "loss": 0.0233, "step": 76420 }, { "grad_norm": 0.3305722177028656, "learning_rate": 1.443599840877316e-05, "loss": 0.0248, "step": 76430 }, { "grad_norm": 0.23136384785175323, "learning_rate": 1.4424377969506103e-05, "loss": 0.0218, "step": 76440 }, { "grad_norm": 0.3567984998226166, "learning_rate": 1.4412761420735565e-05, "loss": 0.0213, "step": 76450 }, { "grad_norm": 0.2276749610900879, "learning_rate": 1.4401148763731953e-05, "loss": 0.0263, "step": 76460 }, { "grad_norm": 0.23396559059619904, "learning_rate": 1.4389539999765173e-05, "loss": 0.0193, "step": 76470 }, { "grad_norm": 0.3513934016227722, "learning_rate": 1.4377935130104758e-05, "loss": 0.0205, "step": 76480 }, { "grad_norm": 0.2413686364889145, "learning_rate": 1.4366334156019801e-05, "loss": 0.0299, "step": 76490 }, { "grad_norm": 0.25872233510017395, "learning_rate": 1.435473707877898e-05, "loss": 0.0217, "step": 76500 }, { "grad_norm": 0.3058360517024994, "learning_rate": 1.4343143899650508e-05, "loss": 0.0253, "step": 76510 }, { "grad_norm": 0.35488325357437134, "learning_rate": 1.4331554619902204e-05, "loss": 0.0244, "step": 76520 }, { "grad_norm": 0.2736954689025879, "learning_rate": 1.4319969240801484e-05, "loss": 0.02, "step": 76530 }, { "grad_norm": 0.3370671570301056, "learning_rate": 1.430838776361526e-05, "loss": 0.0247, "step": 76540 }, { "grad_norm": 0.21352091431617737, "learning_rate": 1.4296810189610093e-05, "loss": 0.0241, "step": 76550 }, { "grad_norm": 0.21770596504211426, "learning_rate": 1.4285236520052087e-05, "loss": 0.0225, "step": 76560 }, { "grad_norm": 0.3370342552661896, "learning_rate": 1.4273666756206933e-05, "loss": 0.0198, "step": 76570 }, { "grad_norm": 0.1965879648923874, "learning_rate": 1.4262100899339848e-05, "loss": 0.0233, "step": 76580 }, { "grad_norm": 0.3998657763004303, "learning_rate": 1.4250538950715696e-05, "loss": 0.0227, "step": 76590 }, { "grad_norm": 0.3326070308685303, "learning_rate": 1.4238980911598842e-05, "loss": 0.0261, "step": 76600 }, { "grad_norm": 0.3960137963294983, "learning_rate": 1.4227426783253267e-05, "loss": 0.0222, "step": 76610 }, { "grad_norm": 0.39534991979599, "learning_rate": 1.4215876566942521e-05, "loss": 0.0242, "step": 76620 }, { "grad_norm": 0.21493099629878998, "learning_rate": 1.4204330263929721e-05, "loss": 0.0175, "step": 76630 }, { "grad_norm": 0.20046019554138184, "learning_rate": 1.4192787875477537e-05, "loss": 0.0295, "step": 76640 }, { "grad_norm": 0.2098478078842163, "learning_rate": 1.4181249402848246e-05, "loss": 0.0229, "step": 76650 }, { "grad_norm": 0.2697877287864685, "learning_rate": 1.4169714847303656e-05, "loss": 0.0247, "step": 76660 }, { "grad_norm": 0.49211645126342773, "learning_rate": 1.4158184210105197e-05, "loss": 0.0252, "step": 76670 }, { "grad_norm": 0.2500855028629303, "learning_rate": 1.4146657492513787e-05, "loss": 0.0232, "step": 76680 }, { "grad_norm": 0.22882919013500214, "learning_rate": 1.4135134695790047e-05, "loss": 0.019, "step": 76690 }, { "grad_norm": 0.4077426493167877, "learning_rate": 1.4123615821194037e-05, "loss": 0.0253, "step": 76700 }, { "grad_norm": 0.4052576720714569, "learning_rate": 1.4112100869985473e-05, "loss": 0.0226, "step": 76710 }, { "grad_norm": 0.19706983864307404, "learning_rate": 1.4100589843423589e-05, "loss": 0.0248, "step": 76720 }, { "grad_norm": 0.35292118787765503, "learning_rate": 1.408908274276723e-05, "loss": 0.0205, "step": 76730 }, { "grad_norm": 0.20048430562019348, "learning_rate": 1.4077579569274786e-05, "loss": 0.0255, "step": 76740 }, { "grad_norm": 0.29391181468963623, "learning_rate": 1.4066080324204245e-05, "loss": 0.0269, "step": 76750 }, { "grad_norm": 0.2644144892692566, "learning_rate": 1.405458500881312e-05, "loss": 0.0251, "step": 76760 }, { "grad_norm": 0.21128569543361664, "learning_rate": 1.4043093624358538e-05, "loss": 0.0223, "step": 76770 }, { "grad_norm": 0.29891717433929443, "learning_rate": 1.4031606172097189e-05, "loss": 0.035, "step": 76780 }, { "grad_norm": 0.2823665738105774, "learning_rate": 1.40201226532853e-05, "loss": 0.0277, "step": 76790 }, { "grad_norm": 0.2592581808567047, "learning_rate": 1.4008643069178701e-05, "loss": 0.0221, "step": 76800 }, { "grad_norm": 0.27002090215682983, "learning_rate": 1.399716742103278e-05, "loss": 0.0267, "step": 76810 }, { "grad_norm": 0.34393009543418884, "learning_rate": 1.398569571010252e-05, "loss": 0.0234, "step": 76820 }, { "grad_norm": 0.30568405985832214, "learning_rate": 1.3974227937642414e-05, "loss": 0.0293, "step": 76830 }, { "grad_norm": 0.21803165972232819, "learning_rate": 1.3962764104906595e-05, "loss": 0.0197, "step": 76840 }, { "grad_norm": 0.24471600353717804, "learning_rate": 1.3951304213148696e-05, "loss": 0.0291, "step": 76850 }, { "grad_norm": 0.21277371048927307, "learning_rate": 1.3939848263621979e-05, "loss": 0.0208, "step": 76860 }, { "grad_norm": 0.4557107985019684, "learning_rate": 1.3928396257579219e-05, "loss": 0.0288, "step": 76870 }, { "grad_norm": 0.20572195947170258, "learning_rate": 1.3916948196272833e-05, "loss": 0.0287, "step": 76880 }, { "grad_norm": 0.2776370346546173, "learning_rate": 1.3905504080954734e-05, "loss": 0.0341, "step": 76890 }, { "grad_norm": 0.3108106553554535, "learning_rate": 1.3894063912876454e-05, "loss": 0.0284, "step": 76900 }, { "grad_norm": 0.27594923973083496, "learning_rate": 1.3882627693289046e-05, "loss": 0.0234, "step": 76910 }, { "grad_norm": 0.48747846484184265, "learning_rate": 1.3871195423443183e-05, "loss": 0.0233, "step": 76920 }, { "grad_norm": 0.35022881627082825, "learning_rate": 1.3859767104589044e-05, "loss": 0.0198, "step": 76930 }, { "grad_norm": 0.38519349694252014, "learning_rate": 1.3848342737976467e-05, "loss": 0.0229, "step": 76940 }, { "grad_norm": 0.22130921483039856, "learning_rate": 1.3836922324854757e-05, "loss": 0.027, "step": 76950 }, { "grad_norm": 0.3496222496032715, "learning_rate": 1.382550586647287e-05, "loss": 0.0294, "step": 76960 }, { "grad_norm": 0.2326122373342514, "learning_rate": 1.3814093364079257e-05, "loss": 0.0222, "step": 76970 }, { "grad_norm": 0.3114522099494934, "learning_rate": 1.3802684818921991e-05, "loss": 0.03, "step": 76980 }, { "grad_norm": 0.35508590936660767, "learning_rate": 1.3791280232248687e-05, "loss": 0.0223, "step": 76990 }, { "grad_norm": 0.24755708873271942, "learning_rate": 1.3779879605306555e-05, "loss": 0.0241, "step": 77000 }, { "grad_norm": 0.3171817660331726, "learning_rate": 1.376848293934232e-05, "loss": 0.035, "step": 77010 }, { "grad_norm": 0.36634501814842224, "learning_rate": 1.3757090235602316e-05, "loss": 0.0234, "step": 77020 }, { "grad_norm": 0.33882153034210205, "learning_rate": 1.3745701495332447e-05, "loss": 0.0248, "step": 77030 }, { "grad_norm": 0.26398155093193054, "learning_rate": 1.3734316719778145e-05, "loss": 0.0212, "step": 77040 }, { "grad_norm": 0.36512595415115356, "learning_rate": 1.3722935910184454e-05, "loss": 0.0211, "step": 77050 }, { "grad_norm": 0.22439685463905334, "learning_rate": 1.3711559067795921e-05, "loss": 0.0215, "step": 77060 }, { "grad_norm": 0.24436995387077332, "learning_rate": 1.3700186193856756e-05, "loss": 0.025, "step": 77070 }, { "grad_norm": 0.5678732395172119, "learning_rate": 1.368881728961064e-05, "loss": 0.0229, "step": 77080 }, { "grad_norm": 0.21771900355815887, "learning_rate": 1.3677452356300885e-05, "loss": 0.0223, "step": 77090 }, { "grad_norm": 0.4264087677001953, "learning_rate": 1.3666091395170321e-05, "loss": 0.0204, "step": 77100 }, { "grad_norm": 0.3880610466003418, "learning_rate": 1.3654734407461384e-05, "loss": 0.0303, "step": 77110 }, { "grad_norm": 0.5380972623825073, "learning_rate": 1.3643381394416022e-05, "loss": 0.0228, "step": 77120 }, { "grad_norm": 0.255016952753067, "learning_rate": 1.3632032357275838e-05, "loss": 0.0296, "step": 77130 }, { "grad_norm": 0.24977955222129822, "learning_rate": 1.3620687297281898e-05, "loss": 0.0241, "step": 77140 }, { "grad_norm": 0.4807186424732208, "learning_rate": 1.3609346215674917e-05, "loss": 0.0243, "step": 77150 }, { "grad_norm": 0.46559447050094604, "learning_rate": 1.3598009113695109e-05, "loss": 0.037, "step": 77160 }, { "grad_norm": 0.21752485632896423, "learning_rate": 1.358667599258231e-05, "loss": 0.0212, "step": 77170 }, { "grad_norm": 0.23637942969799042, "learning_rate": 1.357534685357586e-05, "loss": 0.0243, "step": 77180 }, { "grad_norm": 0.34002095460891724, "learning_rate": 1.3564021697914715e-05, "loss": 0.0289, "step": 77190 }, { "grad_norm": 0.2960253655910492, "learning_rate": 1.3552700526837381e-05, "loss": 0.0252, "step": 77200 }, { "grad_norm": 0.2924986779689789, "learning_rate": 1.3541383341581925e-05, "loss": 0.0199, "step": 77210 }, { "grad_norm": 0.42221543192863464, "learning_rate": 1.3530070143385965e-05, "loss": 0.018, "step": 77220 }, { "grad_norm": 0.2631782591342926, "learning_rate": 1.3518760933486702e-05, "loss": 0.0197, "step": 77230 }, { "grad_norm": 0.21767456829547882, "learning_rate": 1.3507455713120908e-05, "loss": 0.0232, "step": 77240 }, { "grad_norm": 0.13775737583637238, "learning_rate": 1.3496154483524876e-05, "loss": 0.0243, "step": 77250 }, { "grad_norm": 0.3034544587135315, "learning_rate": 1.3484857245934506e-05, "loss": 0.0237, "step": 77260 }, { "grad_norm": 0.20416094362735748, "learning_rate": 1.3473564001585248e-05, "loss": 0.0235, "step": 77270 }, { "grad_norm": 0.2646677494049072, "learning_rate": 1.3462274751712129e-05, "loss": 0.0274, "step": 77280 }, { "grad_norm": 0.23910193145275116, "learning_rate": 1.3450989497549698e-05, "loss": 0.019, "step": 77290 }, { "grad_norm": 0.5327553749084473, "learning_rate": 1.3439708240332115e-05, "loss": 0.0287, "step": 77300 }, { "grad_norm": 0.27457860112190247, "learning_rate": 1.3428430981293043e-05, "loss": 0.0243, "step": 77310 }, { "grad_norm": 0.42697474360466003, "learning_rate": 1.3417157721665808e-05, "loss": 0.0264, "step": 77320 }, { "grad_norm": 0.424500048160553, "learning_rate": 1.3405888462683186e-05, "loss": 0.0232, "step": 77330 }, { "grad_norm": 0.2263607531785965, "learning_rate": 1.3394623205577595e-05, "loss": 0.0256, "step": 77340 }, { "grad_norm": 0.8264221549034119, "learning_rate": 1.3383361951580963e-05, "loss": 0.026, "step": 77350 }, { "grad_norm": 0.49537792801856995, "learning_rate": 1.337210470192483e-05, "loss": 0.024, "step": 77360 }, { "grad_norm": 0.2643928825855255, "learning_rate": 1.3360851457840245e-05, "loss": 0.0247, "step": 77370 }, { "grad_norm": 0.1866685152053833, "learning_rate": 1.3349602220557856e-05, "loss": 0.0213, "step": 77380 }, { "grad_norm": 0.3538166284561157, "learning_rate": 1.3338356991307865e-05, "loss": 0.0235, "step": 77390 }, { "grad_norm": 0.4696909487247467, "learning_rate": 1.3327115771320048e-05, "loss": 0.0273, "step": 77400 }, { "grad_norm": 0.16241614520549774, "learning_rate": 1.3315878561823697e-05, "loss": 0.0206, "step": 77410 }, { "grad_norm": 0.6795424222946167, "learning_rate": 1.330464536404773e-05, "loss": 0.0217, "step": 77420 }, { "grad_norm": 0.26250526309013367, "learning_rate": 1.3293416179220563e-05, "loss": 0.0225, "step": 77430 }, { "grad_norm": 0.3946840465068817, "learning_rate": 1.328219100857021e-05, "loss": 0.0213, "step": 77440 }, { "grad_norm": 0.31288260221481323, "learning_rate": 1.327096985332425e-05, "loss": 0.0215, "step": 77450 }, { "grad_norm": 0.300885409116745, "learning_rate": 1.3259752714709816e-05, "loss": 0.0224, "step": 77460 }, { "grad_norm": 0.5522231459617615, "learning_rate": 1.3248539593953569e-05, "loss": 0.0285, "step": 77470 }, { "grad_norm": 0.35348260402679443, "learning_rate": 1.323733049228178e-05, "loss": 0.0345, "step": 77480 }, { "grad_norm": 0.34031397104263306, "learning_rate": 1.3226125410920276e-05, "loss": 0.0281, "step": 77490 }, { "grad_norm": 0.2679109573364258, "learning_rate": 1.3214924351094388e-05, "loss": 0.0226, "step": 77500 }, { "grad_norm": 0.26682254672050476, "learning_rate": 1.3203727314029069e-05, "loss": 0.0229, "step": 77510 }, { "grad_norm": 0.3803226053714752, "learning_rate": 1.3192534300948811e-05, "loss": 0.0278, "step": 77520 }, { "grad_norm": 0.31850677728652954, "learning_rate": 1.3181345313077675e-05, "loss": 0.0227, "step": 77530 }, { "grad_norm": 0.3564557135105133, "learning_rate": 1.3170160351639243e-05, "loss": 0.0203, "step": 77540 }, { "grad_norm": 0.28141671419143677, "learning_rate": 1.3158979417856721e-05, "loss": 0.0306, "step": 77550 }, { "grad_norm": 0.40015217661857605, "learning_rate": 1.31478025129528e-05, "loss": 0.0307, "step": 77560 }, { "grad_norm": 1.1294704675674438, "learning_rate": 1.3136629638149794e-05, "loss": 0.027, "step": 77570 }, { "grad_norm": 0.2641761898994446, "learning_rate": 1.3125460794669547e-05, "loss": 0.0223, "step": 77580 }, { "grad_norm": 0.421258807182312, "learning_rate": 1.3114295983733483e-05, "loss": 0.02, "step": 77590 }, { "grad_norm": 0.27008888125419617, "learning_rate": 1.3103135206562534e-05, "loss": 0.0252, "step": 77600 }, { "grad_norm": 0.38454851508140564, "learning_rate": 1.3091978464377264e-05, "loss": 0.028, "step": 77610 }, { "grad_norm": 0.2335250824689865, "learning_rate": 1.3080825758397724e-05, "loss": 0.0245, "step": 77620 }, { "grad_norm": 0.34666669368743896, "learning_rate": 1.3069677089843568e-05, "loss": 0.0237, "step": 77630 }, { "grad_norm": 0.23453108966350555, "learning_rate": 1.305853245993401e-05, "loss": 0.0276, "step": 77640 }, { "grad_norm": 0.17318280041217804, "learning_rate": 1.3047391869887809e-05, "loss": 0.0257, "step": 77650 }, { "grad_norm": 0.2910684645175934, "learning_rate": 1.3036255320923263e-05, "loss": 0.0248, "step": 77660 }, { "grad_norm": 0.357707142829895, "learning_rate": 1.3025122814258273e-05, "loss": 0.0225, "step": 77670 }, { "grad_norm": 0.3616807758808136, "learning_rate": 1.3013994351110249e-05, "loss": 0.0257, "step": 77680 }, { "grad_norm": 0.30275124311447144, "learning_rate": 1.3002869932696194e-05, "loss": 0.0226, "step": 77690 }, { "grad_norm": 0.33607450127601624, "learning_rate": 1.2991749560232664e-05, "loss": 0.0302, "step": 77700 }, { "grad_norm": 0.2868140637874603, "learning_rate": 1.2980633234935768e-05, "loss": 0.0214, "step": 77710 }, { "grad_norm": 0.3420301377773285, "learning_rate": 1.2969520958021153e-05, "loss": 0.0238, "step": 77720 }, { "grad_norm": 0.31447380781173706, "learning_rate": 1.295841273070405e-05, "loss": 0.0258, "step": 77730 }, { "grad_norm": 0.38437071442604065, "learning_rate": 1.2947308554199255e-05, "loss": 0.0237, "step": 77740 }, { "grad_norm": 0.4120801091194153, "learning_rate": 1.2936208429721075e-05, "loss": 0.0291, "step": 77750 }, { "grad_norm": 0.3699136972427368, "learning_rate": 1.2925112358483416e-05, "loss": 0.0222, "step": 77760 }, { "grad_norm": 0.3210785984992981, "learning_rate": 1.2914020341699728e-05, "loss": 0.0234, "step": 77770 }, { "grad_norm": 0.32165253162384033, "learning_rate": 1.2902932380583032e-05, "loss": 0.0252, "step": 77780 }, { "grad_norm": 0.34950199723243713, "learning_rate": 1.2891848476345863e-05, "loss": 0.0248, "step": 77790 }, { "grad_norm": 0.41090700030326843, "learning_rate": 1.2880768630200368e-05, "loss": 0.0257, "step": 77800 }, { "grad_norm": 0.34885579347610474, "learning_rate": 1.286969284335819e-05, "loss": 0.0196, "step": 77810 }, { "grad_norm": 0.24393436312675476, "learning_rate": 1.2858621117030578e-05, "loss": 0.0224, "step": 77820 }, { "grad_norm": 0.3707699179649353, "learning_rate": 1.2847553452428324e-05, "loss": 0.017, "step": 77830 }, { "grad_norm": 0.23447628319263458, "learning_rate": 1.283648985076178e-05, "loss": 0.0172, "step": 77840 }, { "grad_norm": 0.24492889642715454, "learning_rate": 1.2825430313240817e-05, "loss": 0.0237, "step": 77850 }, { "grad_norm": 0.4857008755207062, "learning_rate": 1.281437484107491e-05, "loss": 0.026, "step": 77860 }, { "grad_norm": 0.3054066598415375, "learning_rate": 1.2803323435473058e-05, "loss": 0.0243, "step": 77870 }, { "grad_norm": 0.31062763929367065, "learning_rate": 1.2792276097643841e-05, "loss": 0.0234, "step": 77880 }, { "grad_norm": 0.253240704536438, "learning_rate": 1.2781232828795342e-05, "loss": 0.0228, "step": 77890 }, { "grad_norm": 0.24600060284137726, "learning_rate": 1.2770193630135292e-05, "loss": 0.0316, "step": 77900 }, { "grad_norm": 0.33863189816474915, "learning_rate": 1.2759158502870877e-05, "loss": 0.0268, "step": 77910 }, { "grad_norm": 0.36478477716445923, "learning_rate": 1.2748127448208913e-05, "loss": 0.0225, "step": 77920 }, { "grad_norm": 0.27630162239074707, "learning_rate": 1.2737100467355706e-05, "loss": 0.0281, "step": 77930 }, { "grad_norm": 0.2903989255428314, "learning_rate": 1.2726077561517169e-05, "loss": 0.0293, "step": 77940 }, { "grad_norm": 0.36745133996009827, "learning_rate": 1.2715058731898755e-05, "loss": 0.0187, "step": 77950 }, { "grad_norm": 0.31812089681625366, "learning_rate": 1.2704043979705471e-05, "loss": 0.0288, "step": 77960 }, { "grad_norm": 0.5519147515296936, "learning_rate": 1.269303330614185e-05, "loss": 0.0281, "step": 77970 }, { "grad_norm": 0.9568315744400024, "learning_rate": 1.2682026712412015e-05, "loss": 0.0223, "step": 77980 }, { "grad_norm": 0.2705079913139343, "learning_rate": 1.2671024199719644e-05, "loss": 0.0214, "step": 77990 }, { "grad_norm": 0.3699025809764862, "learning_rate": 1.2660025769267936e-05, "loss": 0.023, "step": 78000 }, { "grad_norm": 0.332468181848526, "learning_rate": 1.2649031422259667e-05, "loss": 0.0219, "step": 78010 }, { "grad_norm": 0.3136819303035736, "learning_rate": 1.2638041159897163e-05, "loss": 0.0222, "step": 78020 }, { "grad_norm": 0.3969601094722748, "learning_rate": 1.262705498338232e-05, "loss": 0.0206, "step": 78030 }, { "grad_norm": 0.28855040669441223, "learning_rate": 1.2616072893916542e-05, "loss": 0.0326, "step": 78040 }, { "grad_norm": 0.3423055112361908, "learning_rate": 1.2605094892700836e-05, "loss": 0.021, "step": 78050 }, { "grad_norm": 0.2947874963283539, "learning_rate": 1.259412098093572e-05, "loss": 0.0206, "step": 78060 }, { "grad_norm": 0.30683907866477966, "learning_rate": 1.258315115982131e-05, "loss": 0.0204, "step": 78070 }, { "grad_norm": 0.3785881996154785, "learning_rate": 1.2572185430557203e-05, "loss": 0.0206, "step": 78080 }, { "grad_norm": 0.4699423909187317, "learning_rate": 1.2561223794342659e-05, "loss": 0.0281, "step": 78090 }, { "grad_norm": 0.26013267040252686, "learning_rate": 1.255026625237637e-05, "loss": 0.028, "step": 78100 }, { "grad_norm": 0.28844010829925537, "learning_rate": 1.2539312805856678e-05, "loss": 0.0303, "step": 78110 }, { "grad_norm": 0.22752642631530762, "learning_rate": 1.25283634559814e-05, "loss": 0.0254, "step": 78120 }, { "grad_norm": 0.22268065810203552, "learning_rate": 1.2517418203947972e-05, "loss": 0.0206, "step": 78130 }, { "grad_norm": 0.2861645519733429, "learning_rate": 1.2506477050953308e-05, "loss": 0.0201, "step": 78140 }, { "grad_norm": 0.19956620037555695, "learning_rate": 1.2495539998193972e-05, "loss": 0.0256, "step": 78150 }, { "grad_norm": 0.2958853244781494, "learning_rate": 1.2484607046865975e-05, "loss": 0.0282, "step": 78160 }, { "grad_norm": 0.2528386116027832, "learning_rate": 1.2473678198164968e-05, "loss": 0.0207, "step": 78170 }, { "grad_norm": 0.5113276243209839, "learning_rate": 1.2462753453286075e-05, "loss": 0.0246, "step": 78180 }, { "grad_norm": 0.23703837394714355, "learning_rate": 1.2451832813424031e-05, "loss": 0.0375, "step": 78190 }, { "grad_norm": 0.4469413161277771, "learning_rate": 1.2440916279773095e-05, "loss": 0.0238, "step": 78200 }, { "grad_norm": 0.2835233211517334, "learning_rate": 1.2430003853527101e-05, "loss": 0.024, "step": 78210 }, { "grad_norm": 0.5485398173332214, "learning_rate": 1.2419095535879383e-05, "loss": 0.0309, "step": 78220 }, { "grad_norm": 0.19239389896392822, "learning_rate": 1.2408191328022878e-05, "loss": 0.0267, "step": 78230 }, { "grad_norm": 0.5369459390640259, "learning_rate": 1.2397291231150066e-05, "loss": 0.0249, "step": 78240 }, { "grad_norm": 0.33698928356170654, "learning_rate": 1.2386395246452937e-05, "loss": 0.0229, "step": 78250 }, { "grad_norm": 0.20384398102760315, "learning_rate": 1.2375503375123083e-05, "loss": 0.0219, "step": 78260 }, { "grad_norm": 0.1880306899547577, "learning_rate": 1.2364615618351583e-05, "loss": 0.0295, "step": 78270 }, { "grad_norm": 0.336712509393692, "learning_rate": 1.2353731977329169e-05, "loss": 0.0222, "step": 78280 }, { "grad_norm": 0.23451906442642212, "learning_rate": 1.2342852453246007e-05, "loss": 0.0195, "step": 78290 }, { "grad_norm": 0.25647592544555664, "learning_rate": 1.2331977047291898e-05, "loss": 0.0264, "step": 78300 }, { "grad_norm": 0.36053505539894104, "learning_rate": 1.2321105760656137e-05, "loss": 0.0243, "step": 78310 }, { "grad_norm": 0.20063284039497375, "learning_rate": 1.2310238594527613e-05, "loss": 0.0225, "step": 78320 }, { "grad_norm": 0.8607580065727234, "learning_rate": 1.2299375550094693e-05, "loss": 0.0241, "step": 78330 }, { "grad_norm": 0.4469710886478424, "learning_rate": 1.2288516628545421e-05, "loss": 0.0225, "step": 78340 }, { "grad_norm": 0.21102914214134216, "learning_rate": 1.2277661831067255e-05, "loss": 0.0177, "step": 78350 }, { "grad_norm": 0.23711448907852173, "learning_rate": 1.2266811158847286e-05, "loss": 0.0264, "step": 78360 }, { "grad_norm": 0.21556755900382996, "learning_rate": 1.2255964613072107e-05, "loss": 0.0176, "step": 78370 }, { "grad_norm": 0.278449684381485, "learning_rate": 1.2245122194927905e-05, "loss": 0.0173, "step": 78380 }, { "grad_norm": 0.2636014521121979, "learning_rate": 1.2234283905600364e-05, "loss": 0.0251, "step": 78390 }, { "grad_norm": 0.4570596516132355, "learning_rate": 1.222344974627475e-05, "loss": 0.0279, "step": 78400 }, { "grad_norm": 0.36886993050575256, "learning_rate": 1.2212619718135875e-05, "loss": 0.0241, "step": 78410 }, { "grad_norm": 0.23749801516532898, "learning_rate": 1.2201793822368118e-05, "loss": 0.0222, "step": 78420 }, { "grad_norm": 0.2975168228149414, "learning_rate": 1.2190972060155337e-05, "loss": 0.0189, "step": 78430 }, { "grad_norm": 0.29159116744995117, "learning_rate": 1.2180154432681013e-05, "loss": 0.0215, "step": 78440 }, { "grad_norm": 0.33292797207832336, "learning_rate": 1.2169340941128148e-05, "loss": 0.0226, "step": 78450 }, { "grad_norm": 0.3142009377479553, "learning_rate": 1.2158531586679267e-05, "loss": 0.0196, "step": 78460 }, { "grad_norm": 0.34539222717285156, "learning_rate": 1.2147726370516477e-05, "loss": 0.0234, "step": 78470 }, { "grad_norm": 0.15297135710716248, "learning_rate": 1.2136925293821422e-05, "loss": 0.028, "step": 78480 }, { "grad_norm": 0.4694608151912689, "learning_rate": 1.2126128357775306e-05, "loss": 0.0281, "step": 78490 }, { "grad_norm": 0.7154088616371155, "learning_rate": 1.2115335563558838e-05, "loss": 0.0238, "step": 78500 }, { "grad_norm": 0.1711389571428299, "learning_rate": 1.2104546912352327e-05, "loss": 0.0226, "step": 78510 }, { "grad_norm": 0.23397746682167053, "learning_rate": 1.2093762405335557e-05, "loss": 0.0214, "step": 78520 }, { "grad_norm": 0.4512442946434021, "learning_rate": 1.2082982043687974e-05, "loss": 0.0238, "step": 78530 }, { "grad_norm": 0.17365752160549164, "learning_rate": 1.207220582858845e-05, "loss": 0.0244, "step": 78540 }, { "grad_norm": 0.3958340883255005, "learning_rate": 1.206143376121549e-05, "loss": 0.0204, "step": 78550 }, { "grad_norm": 0.3567781150341034, "learning_rate": 1.2050665842747078e-05, "loss": 0.0215, "step": 78560 }, { "grad_norm": 0.40099918842315674, "learning_rate": 1.2039902074360804e-05, "loss": 0.0227, "step": 78570 }, { "grad_norm": 0.2693382203578949, "learning_rate": 1.2029142457233755e-05, "loss": 0.025, "step": 78580 }, { "grad_norm": 0.37033623456954956, "learning_rate": 1.2018386992542601e-05, "loss": 0.0304, "step": 78590 }, { "grad_norm": 0.14922739565372467, "learning_rate": 1.2007635681463542e-05, "loss": 0.0209, "step": 78600 }, { "grad_norm": 0.1915961056947708, "learning_rate": 1.1996888525172334e-05, "loss": 0.0214, "step": 78610 }, { "grad_norm": 0.32086116075515747, "learning_rate": 1.198614552484425e-05, "loss": 0.0192, "step": 78620 }, { "grad_norm": 0.26314952969551086, "learning_rate": 1.197540668165415e-05, "loss": 0.0224, "step": 78630 }, { "grad_norm": 0.24011844396591187, "learning_rate": 1.1964671996776395e-05, "loss": 0.026, "step": 78640 }, { "grad_norm": 0.26771125197410583, "learning_rate": 1.1953941471384922e-05, "loss": 0.0285, "step": 78650 }, { "grad_norm": 0.33130398392677307, "learning_rate": 1.1943215106653205e-05, "loss": 0.0216, "step": 78660 }, { "grad_norm": 0.31441858410835266, "learning_rate": 1.193249290375429e-05, "loss": 0.0206, "step": 78670 }, { "grad_norm": 0.2849322557449341, "learning_rate": 1.1921774863860696e-05, "loss": 0.0263, "step": 78680 }, { "grad_norm": 0.39260029792785645, "learning_rate": 1.1911060988144556e-05, "loss": 0.0231, "step": 78690 }, { "grad_norm": 0.298326313495636, "learning_rate": 1.1900351277777533e-05, "loss": 0.0207, "step": 78700 }, { "grad_norm": 0.44388848543167114, "learning_rate": 1.1889645733930798e-05, "loss": 0.0257, "step": 78710 }, { "grad_norm": 0.42560774087905884, "learning_rate": 1.1878944357775107e-05, "loss": 0.0257, "step": 78720 }, { "grad_norm": 0.2440839558839798, "learning_rate": 1.1868247150480749e-05, "loss": 0.0203, "step": 78730 }, { "grad_norm": 0.5627713799476624, "learning_rate": 1.1857554113217567e-05, "loss": 0.0251, "step": 78740 }, { "grad_norm": 0.25801408290863037, "learning_rate": 1.1846865247154903e-05, "loss": 0.0228, "step": 78750 }, { "grad_norm": 0.2103465050458908, "learning_rate": 1.1836180553461711e-05, "loss": 0.0252, "step": 78760 }, { "grad_norm": 0.21887333691120148, "learning_rate": 1.1825500033306419e-05, "loss": 0.0303, "step": 78770 }, { "grad_norm": 0.26567837595939636, "learning_rate": 1.1814823687857052e-05, "loss": 0.0284, "step": 78780 }, { "grad_norm": 0.2653264105319977, "learning_rate": 1.1804151518281153e-05, "loss": 0.0219, "step": 78790 }, { "grad_norm": 0.31617894768714905, "learning_rate": 1.179348352574583e-05, "loss": 0.0374, "step": 78800 }, { "grad_norm": 0.3127022087574005, "learning_rate": 1.1782819711417697e-05, "loss": 0.0226, "step": 78810 }, { "grad_norm": 0.2760635018348694, "learning_rate": 1.1772160076462951e-05, "loss": 0.0224, "step": 78820 }, { "grad_norm": 0.5590771436691284, "learning_rate": 1.1761504622047293e-05, "loss": 0.0317, "step": 78830 }, { "grad_norm": 0.2638667821884155, "learning_rate": 1.1750853349335999e-05, "loss": 0.0193, "step": 78840 }, { "grad_norm": 0.28559771180152893, "learning_rate": 1.1740206259493874e-05, "loss": 0.0246, "step": 78850 }, { "grad_norm": 0.3184579610824585, "learning_rate": 1.1729563353685286e-05, "loss": 0.0188, "step": 78860 }, { "grad_norm": 0.29405757784843445, "learning_rate": 1.1718924633074091e-05, "loss": 0.0249, "step": 78870 }, { "grad_norm": 0.16333916783332825, "learning_rate": 1.1708290098823765e-05, "loss": 0.0244, "step": 78880 }, { "grad_norm": 0.36263999342918396, "learning_rate": 1.1697659752097245e-05, "loss": 0.0247, "step": 78890 }, { "grad_norm": 0.35484105348587036, "learning_rate": 1.1687033594057068e-05, "loss": 0.0377, "step": 78900 }, { "grad_norm": 0.21733619272708893, "learning_rate": 1.1676411625865297e-05, "loss": 0.0325, "step": 78910 }, { "grad_norm": 0.3870640993118286, "learning_rate": 1.1665793848683542e-05, "loss": 0.0226, "step": 78920 }, { "grad_norm": 0.676992654800415, "learning_rate": 1.1655180263672927e-05, "loss": 0.0392, "step": 78930 }, { "grad_norm": 0.6806769371032715, "learning_rate": 1.1644570871994142e-05, "loss": 0.0237, "step": 78940 }, { "grad_norm": 0.46254533529281616, "learning_rate": 1.163396567480744e-05, "loss": 0.0249, "step": 78950 }, { "grad_norm": 0.42269545793533325, "learning_rate": 1.1623364673272552e-05, "loss": 0.0276, "step": 78960 }, { "grad_norm": 0.3360849916934967, "learning_rate": 1.1612767868548807e-05, "loss": 0.0248, "step": 78970 }, { "grad_norm": 0.3392977714538574, "learning_rate": 1.1602175261795057e-05, "loss": 0.0301, "step": 78980 }, { "grad_norm": 0.30736181139945984, "learning_rate": 1.1591586854169696e-05, "loss": 0.0253, "step": 78990 }, { "grad_norm": 0.21136948466300964, "learning_rate": 1.158100264683064e-05, "loss": 0.0192, "step": 79000 }, { "grad_norm": 0.49990496039390564, "learning_rate": 1.1570422640935386e-05, "loss": 0.0296, "step": 79010 }, { "grad_norm": 0.3673381805419922, "learning_rate": 1.155984683764092e-05, "loss": 0.0307, "step": 79020 }, { "grad_norm": 0.2859455347061157, "learning_rate": 1.1549275238103812e-05, "loss": 0.0283, "step": 79030 }, { "grad_norm": 0.2327902913093567, "learning_rate": 1.1538707843480152e-05, "loss": 0.0261, "step": 79040 }, { "grad_norm": 0.386881023645401, "learning_rate": 1.1528144654925593e-05, "loss": 0.0279, "step": 79050 }, { "grad_norm": 0.2713490426540375, "learning_rate": 1.1517585673595277e-05, "loss": 0.0236, "step": 79060 }, { "grad_norm": 0.268711119890213, "learning_rate": 1.150703090064395e-05, "loss": 0.0224, "step": 79070 }, { "grad_norm": 0.2512661814689636, "learning_rate": 1.1496480337225835e-05, "loss": 0.026, "step": 79080 }, { "grad_norm": 0.2258552610874176, "learning_rate": 1.1485933984494751e-05, "loss": 0.0277, "step": 79090 }, { "grad_norm": 0.23392604291439056, "learning_rate": 1.1475391843604e-05, "loss": 0.0236, "step": 79100 }, { "grad_norm": 0.20439918339252472, "learning_rate": 1.1464853915706503e-05, "loss": 0.03, "step": 79110 }, { "grad_norm": 0.5775834321975708, "learning_rate": 1.1454320201954626e-05, "loss": 0.0266, "step": 79120 }, { "grad_norm": 0.21193933486938477, "learning_rate": 1.1443790703500356e-05, "loss": 0.024, "step": 79130 }, { "grad_norm": 0.3841736912727356, "learning_rate": 1.1433265421495154e-05, "loss": 0.0249, "step": 79140 }, { "grad_norm": 0.2742003798484802, "learning_rate": 1.1422744357090065e-05, "loss": 0.0295, "step": 79150 }, { "grad_norm": 0.34406694769859314, "learning_rate": 1.1412227511435647e-05, "loss": 0.0257, "step": 79160 }, { "grad_norm": 0.280205100774765, "learning_rate": 1.1401714885682025e-05, "loss": 0.0197, "step": 79170 }, { "grad_norm": 0.46495354175567627, "learning_rate": 1.1391206480978823e-05, "loss": 0.0201, "step": 79180 }, { "grad_norm": 0.448126882314682, "learning_rate": 1.1380702298475227e-05, "loss": 0.0258, "step": 79190 }, { "grad_norm": 0.22164593636989594, "learning_rate": 1.1370202339319985e-05, "loss": 0.0221, "step": 79200 }, { "grad_norm": 0.46405428647994995, "learning_rate": 1.1359706604661319e-05, "loss": 0.0299, "step": 79210 }, { "grad_norm": 0.3977196216583252, "learning_rate": 1.1349215095647042e-05, "loss": 0.0269, "step": 79220 }, { "grad_norm": 0.23119211196899414, "learning_rate": 1.1338727813424493e-05, "loss": 0.0212, "step": 79230 }, { "grad_norm": 0.25331512093544006, "learning_rate": 1.132824475914056e-05, "loss": 0.0254, "step": 79240 }, { "grad_norm": 0.39036139845848083, "learning_rate": 1.131776593394162e-05, "loss": 0.02, "step": 79250 }, { "grad_norm": 0.9928498864173889, "learning_rate": 1.1307291338973657e-05, "loss": 0.0278, "step": 79260 }, { "grad_norm": 0.18407325446605682, "learning_rate": 1.1296820975382121e-05, "loss": 0.0244, "step": 79270 }, { "grad_norm": 0.16407419741153717, "learning_rate": 1.1286354844312074e-05, "loss": 0.0294, "step": 79280 }, { "grad_norm": 0.20176605880260468, "learning_rate": 1.1275892946908023e-05, "loss": 0.0231, "step": 79290 }, { "grad_norm": 0.33540603518486023, "learning_rate": 1.126543528431413e-05, "loss": 0.0368, "step": 79300 }, { "grad_norm": 0.2853577435016632, "learning_rate": 1.125498185767398e-05, "loss": 0.0226, "step": 79310 }, { "grad_norm": 0.19181910157203674, "learning_rate": 1.1244532668130781e-05, "loss": 0.0247, "step": 79320 }, { "grad_norm": 0.20782554149627686, "learning_rate": 1.1234087716827203e-05, "loss": 0.0262, "step": 79330 }, { "grad_norm": 0.2702842354774475, "learning_rate": 1.1223647004905524e-05, "loss": 0.0306, "step": 79340 }, { "grad_norm": 0.32423096895217896, "learning_rate": 1.1213210533507485e-05, "loss": 0.0282, "step": 79350 }, { "grad_norm": 0.2843908369541168, "learning_rate": 1.1202778303774447e-05, "loss": 0.0185, "step": 79360 }, { "grad_norm": 0.5822781324386597, "learning_rate": 1.1192350316847234e-05, "loss": 0.0259, "step": 79370 }, { "grad_norm": 0.258102685213089, "learning_rate": 1.1181926573866258e-05, "loss": 0.0249, "step": 79380 }, { "grad_norm": 0.28029459714889526, "learning_rate": 1.1171507075971416e-05, "loss": 0.0248, "step": 79390 }, { "grad_norm": 0.22588874399662018, "learning_rate": 1.1161091824302177e-05, "loss": 0.0231, "step": 79400 }, { "grad_norm": 0.31790441274642944, "learning_rate": 1.115068081999755e-05, "loss": 0.0242, "step": 79410 }, { "grad_norm": 0.33057260513305664, "learning_rate": 1.1140274064196071e-05, "loss": 0.0231, "step": 79420 }, { "grad_norm": 0.2907639741897583, "learning_rate": 1.112987155803578e-05, "loss": 0.0264, "step": 79430 }, { "grad_norm": 0.40142276883125305, "learning_rate": 1.1119473302654298e-05, "loss": 0.0246, "step": 79440 }, { "grad_norm": 0.4560007154941559, "learning_rate": 1.1109079299188769e-05, "loss": 0.024, "step": 79450 }, { "grad_norm": 0.27236509323120117, "learning_rate": 1.1098689548775847e-05, "loss": 0.0228, "step": 79460 }, { "grad_norm": 0.36068660020828247, "learning_rate": 1.1088304052551762e-05, "loss": 0.0208, "step": 79470 }, { "grad_norm": 0.24073728919029236, "learning_rate": 1.1077922811652215e-05, "loss": 0.0307, "step": 79480 }, { "grad_norm": 0.42913439869880676, "learning_rate": 1.1067545827212534e-05, "loss": 0.0246, "step": 79490 }, { "grad_norm": 0.4917539954185486, "learning_rate": 1.1057173100367496e-05, "loss": 0.019, "step": 79500 }, { "grad_norm": 0.2559230923652649, "learning_rate": 1.1046804632251474e-05, "loss": 0.0258, "step": 79510 }, { "grad_norm": 0.32795196771621704, "learning_rate": 1.1036440423998318e-05, "loss": 0.0167, "step": 79520 }, { "grad_norm": 0.31002017855644226, "learning_rate": 1.1026080476741469e-05, "loss": 0.0267, "step": 79530 }, { "grad_norm": 0.2798210382461548, "learning_rate": 1.1015724791613836e-05, "loss": 0.0187, "step": 79540 }, { "grad_norm": 0.3765052556991577, "learning_rate": 1.1005373369747951e-05, "loss": 0.0285, "step": 79550 }, { "grad_norm": 0.2588767409324646, "learning_rate": 1.09950262122758e-05, "loss": 0.0203, "step": 79560 }, { "grad_norm": 0.3029344081878662, "learning_rate": 1.0984683320328948e-05, "loss": 0.026, "step": 79570 }, { "grad_norm": 0.2128300815820694, "learning_rate": 1.0974344695038458e-05, "loss": 0.0219, "step": 79580 }, { "grad_norm": 0.21239566802978516, "learning_rate": 1.0964010337534968e-05, "loss": 0.0258, "step": 79590 }, { "grad_norm": 0.3643321692943573, "learning_rate": 1.0953680248948611e-05, "loss": 0.0204, "step": 79600 }, { "grad_norm": 0.13390399515628815, "learning_rate": 1.0943354430409069e-05, "loss": 0.017, "step": 79610 }, { "grad_norm": 0.9055701494216919, "learning_rate": 1.0933032883045574e-05, "loss": 0.0227, "step": 79620 }, { "grad_norm": 0.3884732127189636, "learning_rate": 1.0922715607986872e-05, "loss": 0.0304, "step": 79630 }, { "grad_norm": 0.35114526748657227, "learning_rate": 1.0912402606361234e-05, "loss": 0.0199, "step": 79640 }, { "grad_norm": 0.34671473503112793, "learning_rate": 1.090209387929647e-05, "loss": 0.0157, "step": 79650 }, { "grad_norm": 0.4513360857963562, "learning_rate": 1.0891789427919957e-05, "loss": 0.0218, "step": 79660 }, { "grad_norm": 0.19368822872638702, "learning_rate": 1.088148925335854e-05, "loss": 0.0188, "step": 79670 }, { "grad_norm": 0.34632501006126404, "learning_rate": 1.087119335673864e-05, "loss": 0.0272, "step": 79680 }, { "grad_norm": 0.2887931168079376, "learning_rate": 1.0860901739186208e-05, "loss": 0.0182, "step": 79690 }, { "grad_norm": 1.3053183555603027, "learning_rate": 1.085061440182673e-05, "loss": 0.023, "step": 79700 }, { "grad_norm": 0.2935989201068878, "learning_rate": 1.0840331345785193e-05, "loss": 0.0231, "step": 79710 }, { "grad_norm": 0.4817873239517212, "learning_rate": 1.083005257218615e-05, "loss": 0.0218, "step": 79720 }, { "grad_norm": 0.3169688284397125, "learning_rate": 1.0819778082153642e-05, "loss": 0.0195, "step": 79730 }, { "grad_norm": 0.3505612909793854, "learning_rate": 1.0809507876811325e-05, "loss": 0.0272, "step": 79740 }, { "grad_norm": 0.20935212075710297, "learning_rate": 1.0799241957282292e-05, "loss": 0.0276, "step": 79750 }, { "grad_norm": 0.2541973888874054, "learning_rate": 1.078898032468923e-05, "loss": 0.0225, "step": 79760 }, { "grad_norm": 0.3333427608013153, "learning_rate": 1.0778722980154315e-05, "loss": 0.0266, "step": 79770 }, { "grad_norm": 0.291550874710083, "learning_rate": 1.0768469924799302e-05, "loss": 0.0207, "step": 79780 }, { "grad_norm": 0.4615292251110077, "learning_rate": 1.0758221159745419e-05, "loss": 0.0326, "step": 79790 }, { "grad_norm": 0.2784978449344635, "learning_rate": 1.074797668611347e-05, "loss": 0.0196, "step": 79800 }, { "grad_norm": 0.16023559868335724, "learning_rate": 1.0737736505023777e-05, "loss": 0.018, "step": 79810 }, { "grad_norm": 0.9182378649711609, "learning_rate": 1.0727500617596203e-05, "loss": 0.0246, "step": 79820 }, { "grad_norm": 0.3950917720794678, "learning_rate": 1.0717269024950104e-05, "loss": 0.0221, "step": 79830 }, { "grad_norm": 0.15705715119838715, "learning_rate": 1.070704172820442e-05, "loss": 0.021, "step": 79840 }, { "grad_norm": 0.24027255177497864, "learning_rate": 1.0696818728477559e-05, "loss": 0.0244, "step": 79850 }, { "grad_norm": 0.23725716769695282, "learning_rate": 1.0686600026887511e-05, "loss": 0.0197, "step": 79860 }, { "grad_norm": 0.32072871923446655, "learning_rate": 1.0676385624551777e-05, "loss": 0.029, "step": 79870 }, { "grad_norm": 0.4986555576324463, "learning_rate": 1.0666175522587402e-05, "loss": 0.0293, "step": 79880 }, { "grad_norm": 0.26964014768600464, "learning_rate": 1.0655969722110925e-05, "loss": 0.0196, "step": 79890 }, { "grad_norm": 0.7244828343391418, "learning_rate": 1.0645768224238439e-05, "loss": 0.0186, "step": 79900 }, { "grad_norm": 0.29568731784820557, "learning_rate": 1.063557103008559e-05, "loss": 0.0226, "step": 79910 }, { "grad_norm": 0.1854134500026703, "learning_rate": 1.0625378140767494e-05, "loss": 0.0189, "step": 79920 }, { "grad_norm": 0.2641109526157379, "learning_rate": 1.0615189557398846e-05, "loss": 0.0191, "step": 79930 }, { "grad_norm": 0.6345382332801819, "learning_rate": 1.060500528109386e-05, "loss": 0.0245, "step": 79940 }, { "grad_norm": 0.2737581431865692, "learning_rate": 1.0594825312966272e-05, "loss": 0.0223, "step": 79950 }, { "grad_norm": 0.45383885502815247, "learning_rate": 1.0584649654129336e-05, "loss": 0.0265, "step": 79960 }, { "grad_norm": 0.3429426848888397, "learning_rate": 1.0574478305695863e-05, "loss": 0.0197, "step": 79970 }, { "grad_norm": 0.2309163361787796, "learning_rate": 1.0564311268778154e-05, "loss": 0.0205, "step": 79980 }, { "grad_norm": 0.3887251317501068, "learning_rate": 1.0554148544488073e-05, "loss": 0.0209, "step": 79990 }, { "grad_norm": 0.37791910767555237, "learning_rate": 1.0543990133937004e-05, "loss": 0.0207, "step": 80000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "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 }