{ "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": 0.7577008605003357, "learning_rate": 1.8e-07, "loss": 1.3836, "step": 10 }, { "grad_norm": 0.725820779800415, "learning_rate": 3.8e-07, "loss": 1.381, "step": 20 }, { "grad_norm": 0.7122315764427185, "learning_rate": 5.8e-07, "loss": 1.3837, "step": 30 }, { "grad_norm": 0.6758638620376587, "learning_rate": 7.8e-07, "loss": 1.3805, "step": 40 }, { "grad_norm": 0.6054528951644897, "learning_rate": 9.8e-07, "loss": 1.3704, "step": 50 }, { "grad_norm": 0.7525731325149536, "learning_rate": 1.18e-06, "loss": 1.3394, "step": 60 }, { "grad_norm": 0.8385981321334839, "learning_rate": 1.3800000000000001e-06, "loss": 1.3089, "step": 70 }, { "grad_norm": 0.7585828304290771, "learning_rate": 1.5800000000000003e-06, "loss": 1.2835, "step": 80 }, { "grad_norm": 0.7257124185562134, "learning_rate": 1.7800000000000001e-06, "loss": 1.286, "step": 90 }, { "grad_norm": 1.1745376586914062, "learning_rate": 1.98e-06, "loss": 1.2837, "step": 100 }, { "grad_norm": 0.7990751266479492, "learning_rate": 2.1800000000000003e-06, "loss": 1.2893, "step": 110 }, { "grad_norm": 0.6681017875671387, "learning_rate": 2.38e-06, "loss": 1.2874, "step": 120 }, { "grad_norm": 0.9213830232620239, "learning_rate": 2.5800000000000003e-06, "loss": 1.2794, "step": 130 }, { "grad_norm": 1.054428219795227, "learning_rate": 2.78e-06, "loss": 1.2715, "step": 140 }, { "grad_norm": 1.145557165145874, "learning_rate": 2.9800000000000003e-06, "loss": 1.2137, "step": 150 }, { "grad_norm": 1.0982874631881714, "learning_rate": 3.1800000000000005e-06, "loss": 1.2014, "step": 160 }, { "grad_norm": 2.3985159397125244, "learning_rate": 3.38e-06, "loss": 1.1816, "step": 170 }, { "grad_norm": 2.8492586612701416, "learning_rate": 3.58e-06, "loss": 1.1615, "step": 180 }, { "grad_norm": 1.1154073476791382, "learning_rate": 3.7800000000000002e-06, "loss": 1.1615, "step": 190 }, { "grad_norm": 1.436386227607727, "learning_rate": 3.98e-06, "loss": 1.1216, "step": 200 }, { "grad_norm": 1.4279000759124756, "learning_rate": 4.18e-06, "loss": 1.112, "step": 210 }, { "grad_norm": 1.9622182846069336, "learning_rate": 4.38e-06, "loss": 1.1113, "step": 220 }, { "grad_norm": 1.0753555297851562, "learning_rate": 4.58e-06, "loss": 1.105, "step": 230 }, { "grad_norm": 1.3205783367156982, "learning_rate": 4.780000000000001e-06, "loss": 1.0974, "step": 240 }, { "grad_norm": 1.2035267353057861, "learning_rate": 4.98e-06, "loss": 1.0942, "step": 250 }, { "grad_norm": 1.3880823850631714, "learning_rate": 5.18e-06, "loss": 1.0974, "step": 260 }, { "grad_norm": 1.041160225868225, "learning_rate": 5.38e-06, "loss": 1.0851, "step": 270 }, { "grad_norm": 1.7360436916351318, "learning_rate": 5.580000000000001e-06, "loss": 1.0842, "step": 280 }, { "grad_norm": 1.1658889055252075, "learning_rate": 5.78e-06, "loss": 1.0807, "step": 290 }, { "grad_norm": 1.0186794996261597, "learning_rate": 5.98e-06, "loss": 1.0829, "step": 300 }, { "grad_norm": 1.4655249118804932, "learning_rate": 6.18e-06, "loss": 1.0535, "step": 310 }, { "grad_norm": 1.3800466060638428, "learning_rate": 6.38e-06, "loss": 1.0506, "step": 320 }, { "grad_norm": 1.151153802871704, "learning_rate": 6.58e-06, "loss": 1.0297, "step": 330 }, { "grad_norm": 1.4121516942977905, "learning_rate": 6.78e-06, "loss": 1.0151, "step": 340 }, { "grad_norm": 1.453189492225647, "learning_rate": 6.98e-06, "loss": 0.9952, "step": 350 }, { "grad_norm": 1.8740990161895752, "learning_rate": 7.180000000000001e-06, "loss": 0.9809, "step": 360 }, { "grad_norm": 1.5014010667800903, "learning_rate": 7.3800000000000005e-06, "loss": 0.9406, "step": 370 }, { "grad_norm": 2.00141978263855, "learning_rate": 7.580000000000001e-06, "loss": 0.9162, "step": 380 }, { "grad_norm": 1.8078380823135376, "learning_rate": 7.78e-06, "loss": 0.8601, "step": 390 }, { "grad_norm": 1.8008062839508057, "learning_rate": 7.98e-06, "loss": 0.8323, "step": 400 }, { "grad_norm": 1.6568864583969116, "learning_rate": 8.18e-06, "loss": 0.797, "step": 410 }, { "grad_norm": 1.563050389289856, "learning_rate": 8.380000000000001e-06, "loss": 0.7487, "step": 420 }, { "grad_norm": 1.9962605237960815, "learning_rate": 8.580000000000001e-06, "loss": 0.7156, "step": 430 }, { "grad_norm": 2.1393113136291504, "learning_rate": 8.78e-06, "loss": 0.6992, "step": 440 }, { "grad_norm": 2.018634080886841, "learning_rate": 8.98e-06, "loss": 0.6441, "step": 450 }, { "grad_norm": 2.3975048065185547, "learning_rate": 9.180000000000002e-06, "loss": 0.6322, "step": 460 }, { "grad_norm": 2.4230804443359375, "learning_rate": 9.38e-06, "loss": 0.6113, "step": 470 }, { "grad_norm": 2.4458086490631104, "learning_rate": 9.58e-06, "loss": 0.5701, "step": 480 }, { "grad_norm": 2.255345344543457, "learning_rate": 9.78e-06, "loss": 0.5451, "step": 490 }, { "grad_norm": 2.2168757915496826, "learning_rate": 9.980000000000001e-06, "loss": 0.4815, "step": 500 }, { "grad_norm": 2.561781167984009, "learning_rate": 1.018e-05, "loss": 0.4627, "step": 510 }, { "grad_norm": 2.8580482006073, "learning_rate": 1.038e-05, "loss": 0.4328, "step": 520 }, { "grad_norm": 2.1386969089508057, "learning_rate": 1.058e-05, "loss": 0.407, "step": 530 }, { "grad_norm": 2.521651268005371, "learning_rate": 1.0780000000000002e-05, "loss": 0.3691, "step": 540 }, { "grad_norm": 3.0186214447021484, "learning_rate": 1.098e-05, "loss": 0.3746, "step": 550 }, { "grad_norm": 2.7694852352142334, "learning_rate": 1.118e-05, "loss": 0.3288, "step": 560 }, { "grad_norm": 2.954277753829956, "learning_rate": 1.1380000000000001e-05, "loss": 0.3315, "step": 570 }, { "grad_norm": 2.1366002559661865, "learning_rate": 1.1580000000000001e-05, "loss": 0.2995, "step": 580 }, { "grad_norm": 3.018644332885742, "learning_rate": 1.178e-05, "loss": 0.2878, "step": 590 }, { "grad_norm": 2.762058734893799, "learning_rate": 1.198e-05, "loss": 0.2813, "step": 600 }, { "grad_norm": 2.7196147441864014, "learning_rate": 1.2180000000000002e-05, "loss": 0.2643, "step": 610 }, { "grad_norm": 2.9618401527404785, "learning_rate": 1.238e-05, "loss": 0.2863, "step": 620 }, { "grad_norm": 2.7392685413360596, "learning_rate": 1.258e-05, "loss": 0.2288, "step": 630 }, { "grad_norm": 2.575446128845215, "learning_rate": 1.278e-05, "loss": 0.2498, "step": 640 }, { "grad_norm": 2.4438316822052, "learning_rate": 1.2980000000000001e-05, "loss": 0.2262, "step": 650 }, { "grad_norm": 2.0422070026397705, "learning_rate": 1.3180000000000001e-05, "loss": 0.2379, "step": 660 }, { "grad_norm": 2.635002851486206, "learning_rate": 1.338e-05, "loss": 0.2237, "step": 670 }, { "grad_norm": 2.4927940368652344, "learning_rate": 1.358e-05, "loss": 0.2072, "step": 680 }, { "grad_norm": 2.850599765777588, "learning_rate": 1.3780000000000002e-05, "loss": 0.2121, "step": 690 }, { "grad_norm": 2.3307151794433594, "learning_rate": 1.3980000000000002e-05, "loss": 0.209, "step": 700 }, { "grad_norm": 2.1824684143066406, "learning_rate": 1.4180000000000001e-05, "loss": 0.2169, "step": 710 }, { "grad_norm": 2.0482308864593506, "learning_rate": 1.4380000000000001e-05, "loss": 0.2028, "step": 720 }, { "grad_norm": 2.3030927181243896, "learning_rate": 1.4580000000000003e-05, "loss": 0.2011, "step": 730 }, { "grad_norm": 3.3859102725982666, "learning_rate": 1.4779999999999999e-05, "loss": 0.191, "step": 740 }, { "grad_norm": 2.1508827209472656, "learning_rate": 1.4979999999999999e-05, "loss": 0.1949, "step": 750 }, { "grad_norm": 2.243912935256958, "learning_rate": 1.518e-05, "loss": 0.1669, "step": 760 }, { "grad_norm": 2.1186673641204834, "learning_rate": 1.538e-05, "loss": 0.1877, "step": 770 }, { "grad_norm": 2.1485629081726074, "learning_rate": 1.558e-05, "loss": 0.176, "step": 780 }, { "grad_norm": 2.423243284225464, "learning_rate": 1.578e-05, "loss": 0.186, "step": 790 }, { "grad_norm": 1.9649564027786255, "learning_rate": 1.598e-05, "loss": 0.1816, "step": 800 }, { "grad_norm": 1.77261221408844, "learning_rate": 1.618e-05, "loss": 0.1835, "step": 810 }, { "grad_norm": 2.0704562664031982, "learning_rate": 1.6380000000000002e-05, "loss": 0.1743, "step": 820 }, { "grad_norm": 1.9605976343154907, "learning_rate": 1.658e-05, "loss": 0.1641, "step": 830 }, { "grad_norm": 2.289663076400757, "learning_rate": 1.6780000000000002e-05, "loss": 0.1545, "step": 840 }, { "grad_norm": 2.07043719291687, "learning_rate": 1.698e-05, "loss": 0.1628, "step": 850 }, { "grad_norm": 2.0452728271484375, "learning_rate": 1.718e-05, "loss": 0.1668, "step": 860 }, { "grad_norm": 2.693458318710327, "learning_rate": 1.7380000000000003e-05, "loss": 0.1672, "step": 870 }, { "grad_norm": 1.9206819534301758, "learning_rate": 1.758e-05, "loss": 0.1629, "step": 880 }, { "grad_norm": 2.1394786834716797, "learning_rate": 1.7780000000000003e-05, "loss": 0.1702, "step": 890 }, { "grad_norm": 2.394744634628296, "learning_rate": 1.798e-05, "loss": 0.1695, "step": 900 }, { "grad_norm": 2.0940020084381104, "learning_rate": 1.818e-05, "loss": 0.1619, "step": 910 }, { "grad_norm": 2.396169662475586, "learning_rate": 1.838e-05, "loss": 0.152, "step": 920 }, { "grad_norm": 2.0589957237243652, "learning_rate": 1.858e-05, "loss": 0.1613, "step": 930 }, { "grad_norm": 1.9179039001464844, "learning_rate": 1.878e-05, "loss": 0.1682, "step": 940 }, { "grad_norm": 1.7884573936462402, "learning_rate": 1.898e-05, "loss": 0.1595, "step": 950 }, { "grad_norm": 1.8957089185714722, "learning_rate": 1.918e-05, "loss": 0.1385, "step": 960 }, { "grad_norm": 1.9201585054397583, "learning_rate": 1.938e-05, "loss": 0.1466, "step": 970 }, { "grad_norm": 2.032327175140381, "learning_rate": 1.9580000000000002e-05, "loss": 0.131, "step": 980 }, { "grad_norm": 2.283958911895752, "learning_rate": 1.978e-05, "loss": 0.1287, "step": 990 }, { "grad_norm": 1.4311153888702393, "learning_rate": 1.9980000000000002e-05, "loss": 0.1532, "step": 1000 }, { "grad_norm": 1.6274725198745728, "learning_rate": 2.0180000000000003e-05, "loss": 0.1342, "step": 1010 }, { "grad_norm": 1.9203872680664062, "learning_rate": 2.038e-05, "loss": 0.1435, "step": 1020 }, { "grad_norm": 1.8719286918640137, "learning_rate": 2.0580000000000003e-05, "loss": 0.1454, "step": 1030 }, { "grad_norm": 1.4845267534255981, "learning_rate": 2.078e-05, "loss": 0.127, "step": 1040 }, { "grad_norm": 1.876242756843567, "learning_rate": 2.098e-05, "loss": 0.1279, "step": 1050 }, { "grad_norm": 1.698288083076477, "learning_rate": 2.118e-05, "loss": 0.1379, "step": 1060 }, { "grad_norm": 1.622426986694336, "learning_rate": 2.138e-05, "loss": 0.1342, "step": 1070 }, { "grad_norm": 1.4225844144821167, "learning_rate": 2.158e-05, "loss": 0.1507, "step": 1080 }, { "grad_norm": 2.0260159969329834, "learning_rate": 2.178e-05, "loss": 0.1404, "step": 1090 }, { "grad_norm": 2.072068929672241, "learning_rate": 2.198e-05, "loss": 0.1241, "step": 1100 }, { "grad_norm": 1.9827849864959717, "learning_rate": 2.218e-05, "loss": 0.1477, "step": 1110 }, { "grad_norm": 1.373953938484192, "learning_rate": 2.2380000000000003e-05, "loss": 0.1283, "step": 1120 }, { "grad_norm": 1.544819951057434, "learning_rate": 2.258e-05, "loss": 0.1421, "step": 1130 }, { "grad_norm": 1.5342422723770142, "learning_rate": 2.2780000000000002e-05, "loss": 0.1309, "step": 1140 }, { "grad_norm": 1.9834656715393066, "learning_rate": 2.298e-05, "loss": 0.135, "step": 1150 }, { "grad_norm": 1.5522745847702026, "learning_rate": 2.318e-05, "loss": 0.1334, "step": 1160 }, { "grad_norm": 1.5809719562530518, "learning_rate": 2.3380000000000003e-05, "loss": 0.1237, "step": 1170 }, { "grad_norm": 1.9126720428466797, "learning_rate": 2.358e-05, "loss": 0.1274, "step": 1180 }, { "grad_norm": 1.6281156539916992, "learning_rate": 2.3780000000000003e-05, "loss": 0.1267, "step": 1190 }, { "grad_norm": 1.6713403463363647, "learning_rate": 2.398e-05, "loss": 0.1403, "step": 1200 }, { "grad_norm": 1.7239394187927246, "learning_rate": 2.418e-05, "loss": 0.1217, "step": 1210 }, { "grad_norm": 1.6176154613494873, "learning_rate": 2.438e-05, "loss": 0.1237, "step": 1220 }, { "grad_norm": 1.2931221723556519, "learning_rate": 2.4580000000000002e-05, "loss": 0.1318, "step": 1230 }, { "grad_norm": 1.695386290550232, "learning_rate": 2.478e-05, "loss": 0.1192, "step": 1240 }, { "grad_norm": 1.5995484590530396, "learning_rate": 2.498e-05, "loss": 0.1205, "step": 1250 }, { "grad_norm": 1.6093312501907349, "learning_rate": 2.5180000000000003e-05, "loss": 0.122, "step": 1260 }, { "grad_norm": 1.8868176937103271, "learning_rate": 2.5380000000000004e-05, "loss": 0.1283, "step": 1270 }, { "grad_norm": 2.1105704307556152, "learning_rate": 2.5580000000000002e-05, "loss": 0.119, "step": 1280 }, { "grad_norm": 1.8880221843719482, "learning_rate": 2.5779999999999997e-05, "loss": 0.1186, "step": 1290 }, { "grad_norm": 1.325209379196167, "learning_rate": 2.598e-05, "loss": 0.1208, "step": 1300 }, { "grad_norm": 1.6902358531951904, "learning_rate": 2.618e-05, "loss": 0.1229, "step": 1310 }, { "grad_norm": 1.5872042179107666, "learning_rate": 2.6379999999999998e-05, "loss": 0.1345, "step": 1320 }, { "grad_norm": 1.9437546730041504, "learning_rate": 2.658e-05, "loss": 0.121, "step": 1330 }, { "grad_norm": 1.6549464464187622, "learning_rate": 2.678e-05, "loss": 0.1264, "step": 1340 }, { "grad_norm": 1.4572832584381104, "learning_rate": 2.698e-05, "loss": 0.1138, "step": 1350 }, { "grad_norm": 1.6916069984436035, "learning_rate": 2.718e-05, "loss": 0.1192, "step": 1360 }, { "grad_norm": 1.6987311840057373, "learning_rate": 2.738e-05, "loss": 0.1146, "step": 1370 }, { "grad_norm": 1.353557825088501, "learning_rate": 2.758e-05, "loss": 0.1155, "step": 1380 }, { "grad_norm": 1.1098440885543823, "learning_rate": 2.778e-05, "loss": 0.115, "step": 1390 }, { "grad_norm": 1.4133659601211548, "learning_rate": 2.798e-05, "loss": 0.1109, "step": 1400 }, { "grad_norm": 1.6969293355941772, "learning_rate": 2.818e-05, "loss": 0.1292, "step": 1410 }, { "grad_norm": 1.5177712440490723, "learning_rate": 2.8380000000000003e-05, "loss": 0.1282, "step": 1420 }, { "grad_norm": 1.5204286575317383, "learning_rate": 2.858e-05, "loss": 0.114, "step": 1430 }, { "grad_norm": 1.5514554977416992, "learning_rate": 2.8780000000000002e-05, "loss": 0.1217, "step": 1440 }, { "grad_norm": 1.0982764959335327, "learning_rate": 2.898e-05, "loss": 0.0986, "step": 1450 }, { "grad_norm": 1.4458794593811035, "learning_rate": 2.9180000000000002e-05, "loss": 0.1066, "step": 1460 }, { "grad_norm": 1.3319244384765625, "learning_rate": 2.9380000000000003e-05, "loss": 0.1122, "step": 1470 }, { "grad_norm": 1.3168091773986816, "learning_rate": 2.958e-05, "loss": 0.1203, "step": 1480 }, { "grad_norm": 1.647213101387024, "learning_rate": 2.9780000000000003e-05, "loss": 0.1157, "step": 1490 }, { "grad_norm": 1.3653044700622559, "learning_rate": 2.998e-05, "loss": 0.1165, "step": 1500 }, { "grad_norm": 1.2364357709884644, "learning_rate": 3.0180000000000002e-05, "loss": 0.1184, "step": 1510 }, { "grad_norm": 1.5575332641601562, "learning_rate": 3.0380000000000004e-05, "loss": 0.1227, "step": 1520 }, { "grad_norm": 1.5584442615509033, "learning_rate": 3.058e-05, "loss": 0.1196, "step": 1530 }, { "grad_norm": 1.249385118484497, "learning_rate": 3.078e-05, "loss": 0.108, "step": 1540 }, { "grad_norm": 1.411232829093933, "learning_rate": 3.0980000000000005e-05, "loss": 0.116, "step": 1550 }, { "grad_norm": 1.3306970596313477, "learning_rate": 3.118e-05, "loss": 0.1133, "step": 1560 }, { "grad_norm": 1.7256847620010376, "learning_rate": 3.138e-05, "loss": 0.1222, "step": 1570 }, { "grad_norm": 1.1204988956451416, "learning_rate": 3.1580000000000006e-05, "loss": 0.1162, "step": 1580 }, { "grad_norm": 1.553155779838562, "learning_rate": 3.1780000000000004e-05, "loss": 0.1119, "step": 1590 }, { "grad_norm": 1.2369136810302734, "learning_rate": 3.198e-05, "loss": 0.1155, "step": 1600 }, { "grad_norm": 1.6478238105773926, "learning_rate": 3.218e-05, "loss": 0.1168, "step": 1610 }, { "grad_norm": 1.4325453042984009, "learning_rate": 3.238e-05, "loss": 0.1074, "step": 1620 }, { "grad_norm": 1.2119832038879395, "learning_rate": 3.2579999999999996e-05, "loss": 0.1116, "step": 1630 }, { "grad_norm": 1.5327686071395874, "learning_rate": 3.278e-05, "loss": 0.109, "step": 1640 }, { "grad_norm": 1.4045544862747192, "learning_rate": 3.298e-05, "loss": 0.1112, "step": 1650 }, { "grad_norm": 1.365864872932434, "learning_rate": 3.318e-05, "loss": 0.1128, "step": 1660 }, { "grad_norm": 1.5497887134552002, "learning_rate": 3.338e-05, "loss": 0.1146, "step": 1670 }, { "grad_norm": 1.2988595962524414, "learning_rate": 3.358e-05, "loss": 0.1042, "step": 1680 }, { "grad_norm": 1.1818186044692993, "learning_rate": 3.378e-05, "loss": 0.1141, "step": 1690 }, { "grad_norm": 1.6048967838287354, "learning_rate": 3.398e-05, "loss": 0.1184, "step": 1700 }, { "grad_norm": 1.317836046218872, "learning_rate": 3.418e-05, "loss": 0.1247, "step": 1710 }, { "grad_norm": 1.3938318490982056, "learning_rate": 3.438e-05, "loss": 0.1182, "step": 1720 }, { "grad_norm": 1.400139331817627, "learning_rate": 3.4580000000000004e-05, "loss": 0.1115, "step": 1730 }, { "grad_norm": 1.1402385234832764, "learning_rate": 3.478e-05, "loss": 0.1052, "step": 1740 }, { "grad_norm": 1.2874343395233154, "learning_rate": 3.498e-05, "loss": 0.1095, "step": 1750 }, { "grad_norm": 0.9604700803756714, "learning_rate": 3.518e-05, "loss": 0.1164, "step": 1760 }, { "grad_norm": 1.0307679176330566, "learning_rate": 3.5380000000000003e-05, "loss": 0.1072, "step": 1770 }, { "grad_norm": 1.2435619831085205, "learning_rate": 3.558e-05, "loss": 0.1191, "step": 1780 }, { "grad_norm": 1.2745054960250854, "learning_rate": 3.578e-05, "loss": 0.1033, "step": 1790 }, { "grad_norm": 1.0953179597854614, "learning_rate": 3.5980000000000004e-05, "loss": 0.0999, "step": 1800 }, { "grad_norm": 1.4336488246917725, "learning_rate": 3.618e-05, "loss": 0.0961, "step": 1810 }, { "grad_norm": 1.080160140991211, "learning_rate": 3.638e-05, "loss": 0.1045, "step": 1820 }, { "grad_norm": 1.234252691268921, "learning_rate": 3.6580000000000006e-05, "loss": 0.0938, "step": 1830 }, { "grad_norm": 1.0085773468017578, "learning_rate": 3.6780000000000004e-05, "loss": 0.1081, "step": 1840 }, { "grad_norm": 1.1015454530715942, "learning_rate": 3.698e-05, "loss": 0.1, "step": 1850 }, { "grad_norm": 1.277449131011963, "learning_rate": 3.7180000000000007e-05, "loss": 0.1059, "step": 1860 }, { "grad_norm": 1.283392310142517, "learning_rate": 3.7380000000000005e-05, "loss": 0.1045, "step": 1870 }, { "grad_norm": 1.3553434610366821, "learning_rate": 3.758e-05, "loss": 0.0984, "step": 1880 }, { "grad_norm": 1.0926791429519653, "learning_rate": 3.778000000000001e-05, "loss": 0.1042, "step": 1890 }, { "grad_norm": 0.9056041836738586, "learning_rate": 3.7980000000000006e-05, "loss": 0.1056, "step": 1900 }, { "grad_norm": 1.0714775323867798, "learning_rate": 3.818e-05, "loss": 0.1041, "step": 1910 }, { "grad_norm": 1.289616346359253, "learning_rate": 3.838e-05, "loss": 0.1082, "step": 1920 }, { "grad_norm": 0.8689664006233215, "learning_rate": 3.858e-05, "loss": 0.112, "step": 1930 }, { "grad_norm": 0.9900644421577454, "learning_rate": 3.878e-05, "loss": 0.1133, "step": 1940 }, { "grad_norm": 1.085837721824646, "learning_rate": 3.898e-05, "loss": 0.0998, "step": 1950 }, { "grad_norm": 1.2198200225830078, "learning_rate": 3.918e-05, "loss": 0.117, "step": 1960 }, { "grad_norm": 0.949127733707428, "learning_rate": 3.938e-05, "loss": 0.1187, "step": 1970 }, { "grad_norm": 1.2700855731964111, "learning_rate": 3.958e-05, "loss": 0.099, "step": 1980 }, { "grad_norm": 1.0942388772964478, "learning_rate": 3.978e-05, "loss": 0.1, "step": 1990 }, { "grad_norm": 1.1452142000198364, "learning_rate": 3.998e-05, "loss": 0.1202, "step": 2000 }, { "grad_norm": 1.2151490449905396, "learning_rate": 4.018e-05, "loss": 0.1063, "step": 2010 }, { "grad_norm": 1.1422491073608398, "learning_rate": 4.038e-05, "loss": 0.1087, "step": 2020 }, { "grad_norm": 0.8965654373168945, "learning_rate": 4.058e-05, "loss": 0.1024, "step": 2030 }, { "grad_norm": 0.9135533571243286, "learning_rate": 4.078e-05, "loss": 0.0981, "step": 2040 }, { "grad_norm": 1.0375125408172607, "learning_rate": 4.0980000000000004e-05, "loss": 0.1014, "step": 2050 }, { "grad_norm": 1.1951789855957031, "learning_rate": 4.118e-05, "loss": 0.1131, "step": 2060 }, { "grad_norm": 0.9878064393997192, "learning_rate": 4.138e-05, "loss": 0.1043, "step": 2070 }, { "grad_norm": 1.168434977531433, "learning_rate": 4.1580000000000005e-05, "loss": 0.1106, "step": 2080 }, { "grad_norm": 1.1250540018081665, "learning_rate": 4.178e-05, "loss": 0.109, "step": 2090 }, { "grad_norm": 1.0184279680252075, "learning_rate": 4.198e-05, "loss": 0.0975, "step": 2100 }, { "grad_norm": 0.9802690744400024, "learning_rate": 4.2180000000000006e-05, "loss": 0.0897, "step": 2110 }, { "grad_norm": 0.9995561838150024, "learning_rate": 4.2380000000000004e-05, "loss": 0.1032, "step": 2120 }, { "grad_norm": 1.0819569826126099, "learning_rate": 4.258e-05, "loss": 0.1065, "step": 2130 }, { "grad_norm": 0.9770635366439819, "learning_rate": 4.278e-05, "loss": 0.0984, "step": 2140 }, { "grad_norm": 1.160718321800232, "learning_rate": 4.2980000000000005e-05, "loss": 0.0929, "step": 2150 }, { "grad_norm": 0.8626875281333923, "learning_rate": 4.318e-05, "loss": 0.0951, "step": 2160 }, { "grad_norm": 0.9788516163825989, "learning_rate": 4.338e-05, "loss": 0.0928, "step": 2170 }, { "grad_norm": 0.7156282663345337, "learning_rate": 4.3580000000000006e-05, "loss": 0.0926, "step": 2180 }, { "grad_norm": 1.0202832221984863, "learning_rate": 4.3780000000000004e-05, "loss": 0.1022, "step": 2190 }, { "grad_norm": 0.8410323262214661, "learning_rate": 4.398e-05, "loss": 0.1106, "step": 2200 }, { "grad_norm": 0.9159443378448486, "learning_rate": 4.418000000000001e-05, "loss": 0.0905, "step": 2210 }, { "grad_norm": 0.8739677667617798, "learning_rate": 4.438e-05, "loss": 0.0943, "step": 2220 }, { "grad_norm": 0.9752373695373535, "learning_rate": 4.458e-05, "loss": 0.0975, "step": 2230 }, { "grad_norm": 0.9145032167434692, "learning_rate": 4.478e-05, "loss": 0.1025, "step": 2240 }, { "grad_norm": 0.9973922371864319, "learning_rate": 4.498e-05, "loss": 0.1068, "step": 2250 }, { "grad_norm": 1.006311297416687, "learning_rate": 4.518e-05, "loss": 0.0947, "step": 2260 }, { "grad_norm": 0.8293299078941345, "learning_rate": 4.538e-05, "loss": 0.0965, "step": 2270 }, { "grad_norm": 1.0088499784469604, "learning_rate": 4.558e-05, "loss": 0.1046, "step": 2280 }, { "grad_norm": 1.0790984630584717, "learning_rate": 4.578e-05, "loss": 0.097, "step": 2290 }, { "grad_norm": 0.9250316023826599, "learning_rate": 4.5980000000000004e-05, "loss": 0.1067, "step": 2300 }, { "grad_norm": 0.9737200736999512, "learning_rate": 4.618e-05, "loss": 0.0952, "step": 2310 }, { "grad_norm": 1.0640029907226562, "learning_rate": 4.638e-05, "loss": 0.0859, "step": 2320 }, { "grad_norm": 1.0172910690307617, "learning_rate": 4.6580000000000005e-05, "loss": 0.0989, "step": 2330 }, { "grad_norm": 0.8128358721733093, "learning_rate": 4.678e-05, "loss": 0.1024, "step": 2340 }, { "grad_norm": 0.7860767245292664, "learning_rate": 4.698e-05, "loss": 0.1031, "step": 2350 }, { "grad_norm": 0.967225193977356, "learning_rate": 4.718e-05, "loss": 0.1007, "step": 2360 }, { "grad_norm": 0.8701141476631165, "learning_rate": 4.7380000000000004e-05, "loss": 0.1026, "step": 2370 }, { "grad_norm": 1.1585092544555664, "learning_rate": 4.758e-05, "loss": 0.102, "step": 2380 }, { "grad_norm": 0.8841970562934875, "learning_rate": 4.778e-05, "loss": 0.1123, "step": 2390 }, { "grad_norm": 0.9116512537002563, "learning_rate": 4.7980000000000005e-05, "loss": 0.0987, "step": 2400 }, { "grad_norm": 0.8868656158447266, "learning_rate": 4.818e-05, "loss": 0.0994, "step": 2410 }, { "grad_norm": 0.9593870639801025, "learning_rate": 4.838e-05, "loss": 0.0963, "step": 2420 }, { "grad_norm": 1.0329933166503906, "learning_rate": 4.8580000000000006e-05, "loss": 0.094, "step": 2430 }, { "grad_norm": 0.7542621493339539, "learning_rate": 4.8780000000000004e-05, "loss": 0.0941, "step": 2440 }, { "grad_norm": 0.8226548433303833, "learning_rate": 4.898e-05, "loss": 0.0886, "step": 2450 }, { "grad_norm": 0.7963999509811401, "learning_rate": 4.918000000000001e-05, "loss": 0.092, "step": 2460 }, { "grad_norm": 0.9379332661628723, "learning_rate": 4.9380000000000005e-05, "loss": 0.0924, "step": 2470 }, { "grad_norm": 0.9802100658416748, "learning_rate": 4.958e-05, "loss": 0.0958, "step": 2480 }, { "grad_norm": 0.8377652168273926, "learning_rate": 4.978e-05, "loss": 0.0916, "step": 2490 }, { "grad_norm": 0.8930109143257141, "learning_rate": 4.9980000000000006e-05, "loss": 0.0974, "step": 2500 }, { "grad_norm": 0.7837635278701782, "learning_rate": 5.0180000000000004e-05, "loss": 0.0817, "step": 2510 }, { "grad_norm": 0.6405776739120483, "learning_rate": 5.038e-05, "loss": 0.0847, "step": 2520 }, { "grad_norm": 1.0111613273620605, "learning_rate": 5.058000000000001e-05, "loss": 0.0937, "step": 2530 }, { "grad_norm": 0.8111575245857239, "learning_rate": 5.0780000000000005e-05, "loss": 0.1012, "step": 2540 }, { "grad_norm": 0.769325315952301, "learning_rate": 5.098e-05, "loss": 0.0906, "step": 2550 }, { "grad_norm": 0.9260642528533936, "learning_rate": 5.118000000000001e-05, "loss": 0.0943, "step": 2560 }, { "grad_norm": 0.9428144693374634, "learning_rate": 5.1380000000000006e-05, "loss": 0.0896, "step": 2570 }, { "grad_norm": 0.9631617665290833, "learning_rate": 5.1580000000000004e-05, "loss": 0.1012, "step": 2580 }, { "grad_norm": 0.9247748255729675, "learning_rate": 5.178000000000001e-05, "loss": 0.0922, "step": 2590 }, { "grad_norm": 0.8760806918144226, "learning_rate": 5.198000000000001e-05, "loss": 0.0989, "step": 2600 }, { "grad_norm": 0.7221530079841614, "learning_rate": 5.2180000000000005e-05, "loss": 0.0841, "step": 2610 }, { "grad_norm": 0.9759504795074463, "learning_rate": 5.238000000000001e-05, "loss": 0.0892, "step": 2620 }, { "grad_norm": 0.9073474407196045, "learning_rate": 5.258000000000001e-05, "loss": 0.0961, "step": 2630 }, { "grad_norm": 0.7806669473648071, "learning_rate": 5.2780000000000006e-05, "loss": 0.1045, "step": 2640 }, { "grad_norm": 0.9920849800109863, "learning_rate": 5.2980000000000004e-05, "loss": 0.0886, "step": 2650 }, { "grad_norm": 0.6801614761352539, "learning_rate": 5.318000000000001e-05, "loss": 0.084, "step": 2660 }, { "grad_norm": 0.7445484399795532, "learning_rate": 5.338000000000001e-05, "loss": 0.0819, "step": 2670 }, { "grad_norm": 0.6024136543273926, "learning_rate": 5.3580000000000005e-05, "loss": 0.0825, "step": 2680 }, { "grad_norm": 0.8049213290214539, "learning_rate": 5.378e-05, "loss": 0.0812, "step": 2690 }, { "grad_norm": 0.8662463426589966, "learning_rate": 5.3979999999999995e-05, "loss": 0.0948, "step": 2700 }, { "grad_norm": 0.9112851023674011, "learning_rate": 5.418e-05, "loss": 0.0952, "step": 2710 }, { "grad_norm": 0.8810095191001892, "learning_rate": 5.438e-05, "loss": 0.097, "step": 2720 }, { "grad_norm": 0.9731611609458923, "learning_rate": 5.4579999999999996e-05, "loss": 0.0904, "step": 2730 }, { "grad_norm": 0.8135035037994385, "learning_rate": 5.478e-05, "loss": 0.0891, "step": 2740 }, { "grad_norm": 0.7347433567047119, "learning_rate": 5.498e-05, "loss": 0.0859, "step": 2750 }, { "grad_norm": 0.8484992384910583, "learning_rate": 5.518e-05, "loss": 0.0977, "step": 2760 }, { "grad_norm": 0.7911871075630188, "learning_rate": 5.538e-05, "loss": 0.0908, "step": 2770 }, { "grad_norm": 0.7493025064468384, "learning_rate": 5.558e-05, "loss": 0.0764, "step": 2780 }, { "grad_norm": 0.8321363925933838, "learning_rate": 5.578e-05, "loss": 0.0871, "step": 2790 }, { "grad_norm": 0.687761664390564, "learning_rate": 5.5979999999999996e-05, "loss": 0.0848, "step": 2800 }, { "grad_norm": 0.6103227734565735, "learning_rate": 5.618e-05, "loss": 0.0822, "step": 2810 }, { "grad_norm": 0.7540189623832703, "learning_rate": 5.638e-05, "loss": 0.0832, "step": 2820 }, { "grad_norm": 0.8402074575424194, "learning_rate": 5.658e-05, "loss": 0.0866, "step": 2830 }, { "grad_norm": 0.6291273236274719, "learning_rate": 5.678e-05, "loss": 0.0824, "step": 2840 }, { "grad_norm": 0.8037564754486084, "learning_rate": 5.698e-05, "loss": 0.0913, "step": 2850 }, { "grad_norm": 1.0040680170059204, "learning_rate": 5.718e-05, "loss": 0.1023, "step": 2860 }, { "grad_norm": 0.6716125011444092, "learning_rate": 5.738e-05, "loss": 0.078, "step": 2870 }, { "grad_norm": 0.8848466873168945, "learning_rate": 5.758e-05, "loss": 0.0875, "step": 2880 }, { "grad_norm": 0.8355284929275513, "learning_rate": 5.778e-05, "loss": 0.0898, "step": 2890 }, { "grad_norm": 0.8599203824996948, "learning_rate": 5.7980000000000004e-05, "loss": 0.0817, "step": 2900 }, { "grad_norm": 0.6184887290000916, "learning_rate": 5.818e-05, "loss": 0.0908, "step": 2910 }, { "grad_norm": 0.6923024654388428, "learning_rate": 5.838e-05, "loss": 0.0789, "step": 2920 }, { "grad_norm": 0.6162668466567993, "learning_rate": 5.858e-05, "loss": 0.0835, "step": 2930 }, { "grad_norm": 0.8826581239700317, "learning_rate": 5.878e-05, "loss": 0.0799, "step": 2940 }, { "grad_norm": 0.721238374710083, "learning_rate": 5.898e-05, "loss": 0.0827, "step": 2950 }, { "grad_norm": 0.8496203422546387, "learning_rate": 5.918e-05, "loss": 0.0866, "step": 2960 }, { "grad_norm": 0.7243395447731018, "learning_rate": 5.9380000000000004e-05, "loss": 0.0803, "step": 2970 }, { "grad_norm": 0.6661104559898376, "learning_rate": 5.958e-05, "loss": 0.0893, "step": 2980 }, { "grad_norm": 0.7783999443054199, "learning_rate": 5.978e-05, "loss": 0.0813, "step": 2990 }, { "grad_norm": 0.6942806243896484, "learning_rate": 5.9980000000000005e-05, "loss": 0.0961, "step": 3000 }, { "grad_norm": 0.695567786693573, "learning_rate": 6.018e-05, "loss": 0.0822, "step": 3010 }, { "grad_norm": 0.6709478497505188, "learning_rate": 6.038e-05, "loss": 0.0901, "step": 3020 }, { "grad_norm": 0.6790230870246887, "learning_rate": 6.0580000000000006e-05, "loss": 0.0837, "step": 3030 }, { "grad_norm": 0.8424346446990967, "learning_rate": 6.0780000000000004e-05, "loss": 0.0957, "step": 3040 }, { "grad_norm": 0.7921832799911499, "learning_rate": 6.098e-05, "loss": 0.0871, "step": 3050 }, { "grad_norm": 0.6489101052284241, "learning_rate": 6.118000000000001e-05, "loss": 0.0758, "step": 3060 }, { "grad_norm": 0.7217493057250977, "learning_rate": 6.138e-05, "loss": 0.0861, "step": 3070 }, { "grad_norm": 0.5433980226516724, "learning_rate": 6.158e-05, "loss": 0.0835, "step": 3080 }, { "grad_norm": 0.7641897797584534, "learning_rate": 6.178000000000001e-05, "loss": 0.0753, "step": 3090 }, { "grad_norm": 0.6425055861473083, "learning_rate": 6.198e-05, "loss": 0.0912, "step": 3100 }, { "grad_norm": 0.7015649676322937, "learning_rate": 6.218e-05, "loss": 0.0891, "step": 3110 }, { "grad_norm": 0.8493577241897583, "learning_rate": 6.238000000000001e-05, "loss": 0.0826, "step": 3120 }, { "grad_norm": 0.7176684141159058, "learning_rate": 6.258e-05, "loss": 0.0844, "step": 3130 }, { "grad_norm": 0.6789489984512329, "learning_rate": 6.278e-05, "loss": 0.0819, "step": 3140 }, { "grad_norm": 0.6083881258964539, "learning_rate": 6.298000000000001e-05, "loss": 0.0736, "step": 3150 }, { "grad_norm": 0.5841996669769287, "learning_rate": 6.318e-05, "loss": 0.0838, "step": 3160 }, { "grad_norm": 0.6714922189712524, "learning_rate": 6.338e-05, "loss": 0.0748, "step": 3170 }, { "grad_norm": 0.735185444355011, "learning_rate": 6.358000000000001e-05, "loss": 0.0962, "step": 3180 }, { "grad_norm": 0.9876558184623718, "learning_rate": 6.378e-05, "loss": 0.0938, "step": 3190 }, { "grad_norm": 0.6920568346977234, "learning_rate": 6.398000000000001e-05, "loss": 0.0823, "step": 3200 }, { "grad_norm": 0.6249494552612305, "learning_rate": 6.418000000000001e-05, "loss": 0.0817, "step": 3210 }, { "grad_norm": 0.6611130237579346, "learning_rate": 6.438e-05, "loss": 0.0796, "step": 3220 }, { "grad_norm": 0.584447979927063, "learning_rate": 6.458000000000001e-05, "loss": 0.0797, "step": 3230 }, { "grad_norm": 0.7030606269836426, "learning_rate": 6.478000000000001e-05, "loss": 0.0733, "step": 3240 }, { "grad_norm": 0.603246808052063, "learning_rate": 6.498e-05, "loss": 0.083, "step": 3250 }, { "grad_norm": 0.5930961966514587, "learning_rate": 6.518000000000001e-05, "loss": 0.0758, "step": 3260 }, { "grad_norm": 0.663224458694458, "learning_rate": 6.538000000000001e-05, "loss": 0.0797, "step": 3270 }, { "grad_norm": 0.6641004085540771, "learning_rate": 6.558e-05, "loss": 0.0843, "step": 3280 }, { "grad_norm": 0.5905539393424988, "learning_rate": 6.578000000000001e-05, "loss": 0.0824, "step": 3290 }, { "grad_norm": 0.778709888458252, "learning_rate": 6.598e-05, "loss": 0.0772, "step": 3300 }, { "grad_norm": 0.5942777991294861, "learning_rate": 6.618e-05, "loss": 0.0785, "step": 3310 }, { "grad_norm": 0.5761550664901733, "learning_rate": 6.638e-05, "loss": 0.0716, "step": 3320 }, { "grad_norm": 0.6973931193351746, "learning_rate": 6.658e-05, "loss": 0.0792, "step": 3330 }, { "grad_norm": 0.6144301295280457, "learning_rate": 6.678e-05, "loss": 0.0804, "step": 3340 }, { "grad_norm": 0.6710376143455505, "learning_rate": 6.698e-05, "loss": 0.0803, "step": 3350 }, { "grad_norm": 0.6138147115707397, "learning_rate": 6.718e-05, "loss": 0.0814, "step": 3360 }, { "grad_norm": 0.6220517754554749, "learning_rate": 6.738e-05, "loss": 0.076, "step": 3370 }, { "grad_norm": 0.443404883146286, "learning_rate": 6.758e-05, "loss": 0.076, "step": 3380 }, { "grad_norm": 0.603587806224823, "learning_rate": 6.778e-05, "loss": 0.083, "step": 3390 }, { "grad_norm": 0.6581981778144836, "learning_rate": 6.798e-05, "loss": 0.0952, "step": 3400 }, { "grad_norm": 0.6721968054771423, "learning_rate": 6.818e-05, "loss": 0.0798, "step": 3410 }, { "grad_norm": 0.5638247728347778, "learning_rate": 6.838e-05, "loss": 0.0743, "step": 3420 }, { "grad_norm": 0.6646630167961121, "learning_rate": 6.858e-05, "loss": 0.0823, "step": 3430 }, { "grad_norm": 0.5956284999847412, "learning_rate": 6.878e-05, "loss": 0.0789, "step": 3440 }, { "grad_norm": 0.8856781125068665, "learning_rate": 6.898e-05, "loss": 0.0842, "step": 3450 }, { "grad_norm": 0.7279137969017029, "learning_rate": 6.918e-05, "loss": 0.0762, "step": 3460 }, { "grad_norm": 0.6834007501602173, "learning_rate": 6.938e-05, "loss": 0.0908, "step": 3470 }, { "grad_norm": 0.6556162238121033, "learning_rate": 6.958e-05, "loss": 0.0785, "step": 3480 }, { "grad_norm": 0.6993358731269836, "learning_rate": 6.978e-05, "loss": 0.0806, "step": 3490 }, { "grad_norm": 0.6976137161254883, "learning_rate": 6.998e-05, "loss": 0.0708, "step": 3500 }, { "grad_norm": 0.8373354077339172, "learning_rate": 7.018e-05, "loss": 0.0876, "step": 3510 }, { "grad_norm": 0.5539543032646179, "learning_rate": 7.038e-05, "loss": 0.0731, "step": 3520 }, { "grad_norm": 0.6831923723220825, "learning_rate": 7.058e-05, "loss": 0.0763, "step": 3530 }, { "grad_norm": 0.7051395773887634, "learning_rate": 7.078e-05, "loss": 0.0954, "step": 3540 }, { "grad_norm": 0.7387818098068237, "learning_rate": 7.098e-05, "loss": 0.0833, "step": 3550 }, { "grad_norm": 0.6240856647491455, "learning_rate": 7.118e-05, "loss": 0.0872, "step": 3560 }, { "grad_norm": 0.6894928812980652, "learning_rate": 7.138e-05, "loss": 0.0868, "step": 3570 }, { "grad_norm": 0.552067756652832, "learning_rate": 7.158e-05, "loss": 0.0751, "step": 3580 }, { "grad_norm": 0.5854730606079102, "learning_rate": 7.178000000000001e-05, "loss": 0.0891, "step": 3590 }, { "grad_norm": 0.6586112976074219, "learning_rate": 7.198e-05, "loss": 0.0852, "step": 3600 }, { "grad_norm": 0.6468860507011414, "learning_rate": 7.218e-05, "loss": 0.0858, "step": 3610 }, { "grad_norm": 0.6477048993110657, "learning_rate": 7.238000000000001e-05, "loss": 0.0834, "step": 3620 }, { "grad_norm": 0.4570399522781372, "learning_rate": 7.258e-05, "loss": 0.085, "step": 3630 }, { "grad_norm": 0.652307391166687, "learning_rate": 7.278e-05, "loss": 0.0821, "step": 3640 }, { "grad_norm": 0.5497347712516785, "learning_rate": 7.298000000000001e-05, "loss": 0.089, "step": 3650 }, { "grad_norm": 0.6096586585044861, "learning_rate": 7.318e-05, "loss": 0.0707, "step": 3660 }, { "grad_norm": 0.579817533493042, "learning_rate": 7.338e-05, "loss": 0.0766, "step": 3670 }, { "grad_norm": 0.5505943894386292, "learning_rate": 7.358000000000001e-05, "loss": 0.0812, "step": 3680 }, { "grad_norm": 0.5307539701461792, "learning_rate": 7.378e-05, "loss": 0.0785, "step": 3690 }, { "grad_norm": 0.5652599930763245, "learning_rate": 7.398e-05, "loss": 0.0754, "step": 3700 }, { "grad_norm": 0.5330557823181152, "learning_rate": 7.418000000000001e-05, "loss": 0.0752, "step": 3710 }, { "grad_norm": 0.5553550124168396, "learning_rate": 7.438e-05, "loss": 0.0766, "step": 3720 }, { "grad_norm": 0.631356954574585, "learning_rate": 7.458000000000001e-05, "loss": 0.0783, "step": 3730 }, { "grad_norm": 0.5068808794021606, "learning_rate": 7.478e-05, "loss": 0.0741, "step": 3740 }, { "grad_norm": 0.7191847562789917, "learning_rate": 7.498e-05, "loss": 0.0802, "step": 3750 }, { "grad_norm": 0.522840678691864, "learning_rate": 7.518000000000001e-05, "loss": 0.0764, "step": 3760 }, { "grad_norm": 0.5480353832244873, "learning_rate": 7.538e-05, "loss": 0.0727, "step": 3770 }, { "grad_norm": 0.6231732368469238, "learning_rate": 7.558e-05, "loss": 0.0698, "step": 3780 }, { "grad_norm": 0.6247300505638123, "learning_rate": 7.578000000000001e-05, "loss": 0.0769, "step": 3790 }, { "grad_norm": 0.7565157413482666, "learning_rate": 7.598e-05, "loss": 0.0773, "step": 3800 }, { "grad_norm": 0.6095043420791626, "learning_rate": 7.618e-05, "loss": 0.0769, "step": 3810 }, { "grad_norm": 0.5528314709663391, "learning_rate": 7.638000000000001e-05, "loss": 0.0716, "step": 3820 }, { "grad_norm": 0.6585263013839722, "learning_rate": 7.658e-05, "loss": 0.0708, "step": 3830 }, { "grad_norm": 0.5725743174552917, "learning_rate": 7.678000000000001e-05, "loss": 0.0732, "step": 3840 }, { "grad_norm": 0.5604068636894226, "learning_rate": 7.698000000000001e-05, "loss": 0.072, "step": 3850 }, { "grad_norm": 0.6420179605484009, "learning_rate": 7.718e-05, "loss": 0.0814, "step": 3860 }, { "grad_norm": 0.555986762046814, "learning_rate": 7.738000000000001e-05, "loss": 0.0834, "step": 3870 }, { "grad_norm": 0.5225560665130615, "learning_rate": 7.758000000000001e-05, "loss": 0.086, "step": 3880 }, { "grad_norm": 0.7061013579368591, "learning_rate": 7.778e-05, "loss": 0.074, "step": 3890 }, { "grad_norm": 0.5746482014656067, "learning_rate": 7.798000000000001e-05, "loss": 0.078, "step": 3900 }, { "grad_norm": 0.5633807182312012, "learning_rate": 7.818000000000001e-05, "loss": 0.0708, "step": 3910 }, { "grad_norm": 0.5678297877311707, "learning_rate": 7.838e-05, "loss": 0.0742, "step": 3920 }, { "grad_norm": 0.46359577775001526, "learning_rate": 7.858000000000001e-05, "loss": 0.0698, "step": 3930 }, { "grad_norm": 0.6172347068786621, "learning_rate": 7.878e-05, "loss": 0.0787, "step": 3940 }, { "grad_norm": 0.6319193243980408, "learning_rate": 7.897999999999999e-05, "loss": 0.0775, "step": 3950 }, { "grad_norm": 0.5995208621025085, "learning_rate": 7.918e-05, "loss": 0.0761, "step": 3960 }, { "grad_norm": 0.4680873155593872, "learning_rate": 7.938e-05, "loss": 0.0718, "step": 3970 }, { "grad_norm": 0.6596252918243408, "learning_rate": 7.958e-05, "loss": 0.0724, "step": 3980 }, { "grad_norm": 0.7049235105514526, "learning_rate": 7.978e-05, "loss": 0.0704, "step": 3990 }, { "grad_norm": 0.5252794623374939, "learning_rate": 7.998e-05, "loss": 0.0837, "step": 4000 }, { "grad_norm": 0.5718624591827393, "learning_rate": 8.018e-05, "loss": 0.0742, "step": 4010 }, { "grad_norm": 0.5465040802955627, "learning_rate": 8.038e-05, "loss": 0.0826, "step": 4020 }, { "grad_norm": 0.5386576652526855, "learning_rate": 8.058e-05, "loss": 0.085, "step": 4030 }, { "grad_norm": 0.49277645349502563, "learning_rate": 8.078e-05, "loss": 0.0792, "step": 4040 }, { "grad_norm": 0.5719993710517883, "learning_rate": 8.098e-05, "loss": 0.0715, "step": 4050 }, { "grad_norm": 0.6629863977432251, "learning_rate": 8.118e-05, "loss": 0.0834, "step": 4060 }, { "grad_norm": 0.6608491539955139, "learning_rate": 8.138e-05, "loss": 0.0867, "step": 4070 }, { "grad_norm": 0.4788576662540436, "learning_rate": 8.158e-05, "loss": 0.0757, "step": 4080 }, { "grad_norm": 0.5588198304176331, "learning_rate": 8.178e-05, "loss": 0.0851, "step": 4090 }, { "grad_norm": 0.6270920038223267, "learning_rate": 8.198e-05, "loss": 0.0753, "step": 4100 }, { "grad_norm": 0.501396656036377, "learning_rate": 8.218e-05, "loss": 0.075, "step": 4110 }, { "grad_norm": 0.5122924447059631, "learning_rate": 8.238000000000001e-05, "loss": 0.0816, "step": 4120 }, { "grad_norm": 0.7470740079879761, "learning_rate": 8.258e-05, "loss": 0.0761, "step": 4130 }, { "grad_norm": 0.5754150152206421, "learning_rate": 8.278e-05, "loss": 0.084, "step": 4140 }, { "grad_norm": 0.5191735029220581, "learning_rate": 8.298000000000001e-05, "loss": 0.0809, "step": 4150 }, { "grad_norm": 0.5129997134208679, "learning_rate": 8.318e-05, "loss": 0.0719, "step": 4160 }, { "grad_norm": 0.665488600730896, "learning_rate": 8.338e-05, "loss": 0.0808, "step": 4170 }, { "grad_norm": 0.6094200015068054, "learning_rate": 8.358e-05, "loss": 0.0708, "step": 4180 }, { "grad_norm": 0.5257768034934998, "learning_rate": 8.378e-05, "loss": 0.0692, "step": 4190 }, { "grad_norm": 0.5860055685043335, "learning_rate": 8.398e-05, "loss": 0.0779, "step": 4200 }, { "grad_norm": 0.6390093564987183, "learning_rate": 8.418e-05, "loss": 0.0741, "step": 4210 }, { "grad_norm": 0.7140728235244751, "learning_rate": 8.438e-05, "loss": 0.0872, "step": 4220 }, { "grad_norm": 0.5203526616096497, "learning_rate": 8.458e-05, "loss": 0.0769, "step": 4230 }, { "grad_norm": 0.4690493047237396, "learning_rate": 8.478e-05, "loss": 0.0727, "step": 4240 }, { "grad_norm": 0.5297150611877441, "learning_rate": 8.498e-05, "loss": 0.0774, "step": 4250 }, { "grad_norm": 0.5837594270706177, "learning_rate": 8.518000000000001e-05, "loss": 0.0855, "step": 4260 }, { "grad_norm": 0.42350733280181885, "learning_rate": 8.538e-05, "loss": 0.0804, "step": 4270 }, { "grad_norm": 0.4549275040626526, "learning_rate": 8.558e-05, "loss": 0.0803, "step": 4280 }, { "grad_norm": 0.521316647529602, "learning_rate": 8.578000000000001e-05, "loss": 0.0715, "step": 4290 }, { "grad_norm": 0.46921709179878235, "learning_rate": 8.598e-05, "loss": 0.0764, "step": 4300 }, { "grad_norm": 0.4841240346431732, "learning_rate": 8.618e-05, "loss": 0.0639, "step": 4310 }, { "grad_norm": 0.5227863788604736, "learning_rate": 8.638000000000001e-05, "loss": 0.0706, "step": 4320 }, { "grad_norm": 0.5152518153190613, "learning_rate": 8.658e-05, "loss": 0.066, "step": 4330 }, { "grad_norm": 0.5827736854553223, "learning_rate": 8.678e-05, "loss": 0.0823, "step": 4340 }, { "grad_norm": 0.4502127468585968, "learning_rate": 8.698000000000001e-05, "loss": 0.0643, "step": 4350 }, { "grad_norm": 0.4962633550167084, "learning_rate": 8.718e-05, "loss": 0.0697, "step": 4360 }, { "grad_norm": 0.4444354176521301, "learning_rate": 8.738000000000001e-05, "loss": 0.0684, "step": 4370 }, { "grad_norm": 0.4786420166492462, "learning_rate": 8.758000000000001e-05, "loss": 0.0706, "step": 4380 }, { "grad_norm": 0.4686032235622406, "learning_rate": 8.778e-05, "loss": 0.0634, "step": 4390 }, { "grad_norm": 0.4861849844455719, "learning_rate": 8.798000000000001e-05, "loss": 0.07, "step": 4400 }, { "grad_norm": 0.6844934821128845, "learning_rate": 8.818000000000001e-05, "loss": 0.0656, "step": 4410 }, { "grad_norm": 0.5949061512947083, "learning_rate": 8.838e-05, "loss": 0.0763, "step": 4420 }, { "grad_norm": 0.6039676070213318, "learning_rate": 8.858000000000001e-05, "loss": 0.0706, "step": 4430 }, { "grad_norm": 0.5893786549568176, "learning_rate": 8.878000000000001e-05, "loss": 0.0772, "step": 4440 }, { "grad_norm": 0.560758650302887, "learning_rate": 8.898e-05, "loss": 0.0731, "step": 4450 }, { "grad_norm": 0.6050992608070374, "learning_rate": 8.918000000000001e-05, "loss": 0.0746, "step": 4460 }, { "grad_norm": 0.4702772796154022, "learning_rate": 8.938e-05, "loss": 0.0855, "step": 4470 }, { "grad_norm": 0.5126912593841553, "learning_rate": 8.958e-05, "loss": 0.0774, "step": 4480 }, { "grad_norm": 0.39957940578460693, "learning_rate": 8.978000000000001e-05, "loss": 0.0853, "step": 4490 }, { "grad_norm": 0.503490149974823, "learning_rate": 8.998e-05, "loss": 0.0747, "step": 4500 }, { "grad_norm": 0.5185608267784119, "learning_rate": 9.018000000000001e-05, "loss": 0.0808, "step": 4510 }, { "grad_norm": 0.5931364297866821, "learning_rate": 9.038000000000001e-05, "loss": 0.0762, "step": 4520 }, { "grad_norm": 0.5321617126464844, "learning_rate": 9.058e-05, "loss": 0.0795, "step": 4530 }, { "grad_norm": 0.5799871683120728, "learning_rate": 9.078000000000001e-05, "loss": 0.0837, "step": 4540 }, { "grad_norm": 0.5718580484390259, "learning_rate": 9.098000000000001e-05, "loss": 0.0752, "step": 4550 }, { "grad_norm": 0.4874204397201538, "learning_rate": 9.118e-05, "loss": 0.0777, "step": 4560 }, { "grad_norm": 0.44963404536247253, "learning_rate": 9.138e-05, "loss": 0.0746, "step": 4570 }, { "grad_norm": 0.38120943307876587, "learning_rate": 9.158e-05, "loss": 0.0652, "step": 4580 }, { "grad_norm": 0.5028117299079895, "learning_rate": 9.178e-05, "loss": 0.0835, "step": 4590 }, { "grad_norm": 0.49912306666374207, "learning_rate": 9.198e-05, "loss": 0.072, "step": 4600 }, { "grad_norm": 0.5026575326919556, "learning_rate": 9.218e-05, "loss": 0.073, "step": 4610 }, { "grad_norm": 0.3852749764919281, "learning_rate": 9.238e-05, "loss": 0.0745, "step": 4620 }, { "grad_norm": 0.509943962097168, "learning_rate": 9.258e-05, "loss": 0.0741, "step": 4630 }, { "grad_norm": 0.7542542219161987, "learning_rate": 9.278e-05, "loss": 0.0699, "step": 4640 }, { "grad_norm": 0.5010205507278442, "learning_rate": 9.298e-05, "loss": 0.0747, "step": 4650 }, { "grad_norm": 0.636421799659729, "learning_rate": 9.318e-05, "loss": 0.0847, "step": 4660 }, { "grad_norm": 0.6421218514442444, "learning_rate": 9.338e-05, "loss": 0.0799, "step": 4670 }, { "grad_norm": 0.5433158874511719, "learning_rate": 9.358e-05, "loss": 0.0794, "step": 4680 }, { "grad_norm": 0.6316830515861511, "learning_rate": 9.378e-05, "loss": 0.0747, "step": 4690 }, { "grad_norm": 0.5000596046447754, "learning_rate": 9.398e-05, "loss": 0.076, "step": 4700 }, { "grad_norm": 0.7898378372192383, "learning_rate": 9.418e-05, "loss": 0.077, "step": 4710 }, { "grad_norm": 0.5334776639938354, "learning_rate": 9.438e-05, "loss": 0.0861, "step": 4720 }, { "grad_norm": 0.5763194561004639, "learning_rate": 9.458e-05, "loss": 0.0721, "step": 4730 }, { "grad_norm": 0.5576677322387695, "learning_rate": 9.478e-05, "loss": 0.0761, "step": 4740 }, { "grad_norm": 0.7949216365814209, "learning_rate": 9.498e-05, "loss": 0.0791, "step": 4750 }, { "grad_norm": 0.41263630986213684, "learning_rate": 9.518000000000001e-05, "loss": 0.0851, "step": 4760 }, { "grad_norm": 0.7013154625892639, "learning_rate": 9.538e-05, "loss": 0.0869, "step": 4770 }, { "grad_norm": 0.6281313300132751, "learning_rate": 9.558e-05, "loss": 0.0779, "step": 4780 }, { "grad_norm": 0.6667198538780212, "learning_rate": 9.578000000000001e-05, "loss": 0.0756, "step": 4790 }, { "grad_norm": 0.5127125978469849, "learning_rate": 9.598e-05, "loss": 0.0796, "step": 4800 }, { "grad_norm": 0.5824614763259888, "learning_rate": 9.618e-05, "loss": 0.0716, "step": 4810 }, { "grad_norm": 0.577195942401886, "learning_rate": 9.638000000000001e-05, "loss": 0.076, "step": 4820 }, { "grad_norm": 0.4470609426498413, "learning_rate": 9.658e-05, "loss": 0.0713, "step": 4830 }, { "grad_norm": 0.5846791863441467, "learning_rate": 9.678e-05, "loss": 0.0774, "step": 4840 }, { "grad_norm": 0.4643741250038147, "learning_rate": 9.698000000000001e-05, "loss": 0.0766, "step": 4850 }, { "grad_norm": 0.44236478209495544, "learning_rate": 9.718e-05, "loss": 0.0725, "step": 4860 }, { "grad_norm": 0.5053511261940002, "learning_rate": 9.738e-05, "loss": 0.0759, "step": 4870 }, { "grad_norm": 0.5472114682197571, "learning_rate": 9.758000000000001e-05, "loss": 0.0706, "step": 4880 }, { "grad_norm": 0.45788055658340454, "learning_rate": 9.778e-05, "loss": 0.0777, "step": 4890 }, { "grad_norm": 0.5401052236557007, "learning_rate": 9.798000000000001e-05, "loss": 0.0771, "step": 4900 }, { "grad_norm": 0.9601448774337769, "learning_rate": 9.818000000000001e-05, "loss": 0.0825, "step": 4910 }, { "grad_norm": 0.5003869533538818, "learning_rate": 9.838e-05, "loss": 0.0828, "step": 4920 }, { "grad_norm": 0.48957082629203796, "learning_rate": 9.858000000000001e-05, "loss": 0.0706, "step": 4930 }, { "grad_norm": 0.518464982509613, "learning_rate": 9.878e-05, "loss": 0.0817, "step": 4940 }, { "grad_norm": 0.6041037440299988, "learning_rate": 9.898e-05, "loss": 0.0866, "step": 4950 }, { "grad_norm": 0.4082501530647278, "learning_rate": 9.918000000000001e-05, "loss": 0.0721, "step": 4960 }, { "grad_norm": 0.6113001704216003, "learning_rate": 9.938e-05, "loss": 0.0729, "step": 4970 }, { "grad_norm": 0.5223151445388794, "learning_rate": 9.958e-05, "loss": 0.077, "step": 4980 }, { "grad_norm": 0.5143321752548218, "learning_rate": 9.978000000000001e-05, "loss": 0.0784, "step": 4990 }, { "grad_norm": 0.5067618489265442, "learning_rate": 9.998e-05, "loss": 0.0799, "step": 5000 }, { "grad_norm": 0.5139220952987671, "learning_rate": 9.999999778549045e-05, "loss": 0.0673, "step": 5010 }, { "grad_norm": 0.5879843831062317, "learning_rate": 9.999999013039593e-05, "loss": 0.091, "step": 5020 }, { "grad_norm": 0.45230966806411743, "learning_rate": 9.999997700737766e-05, "loss": 0.0792, "step": 5030 }, { "grad_norm": 0.42441296577453613, "learning_rate": 9.999995841643709e-05, "loss": 0.0708, "step": 5040 }, { "grad_norm": 0.43749600648880005, "learning_rate": 9.999993435757623e-05, "loss": 0.0653, "step": 5050 }, { "grad_norm": 0.5934156775474548, "learning_rate": 9.999990483079773e-05, "loss": 0.079, "step": 5060 }, { "grad_norm": 0.4716738164424896, "learning_rate": 9.999986983610481e-05, "loss": 0.0729, "step": 5070 }, { "grad_norm": 0.5297526121139526, "learning_rate": 9.99998293735013e-05, "loss": 0.0663, "step": 5080 }, { "grad_norm": 0.4154245853424072, "learning_rate": 9.999978344299161e-05, "loss": 0.0727, "step": 5090 }, { "grad_norm": 0.4640519320964813, "learning_rate": 9.99997320445808e-05, "loss": 0.078, "step": 5100 }, { "grad_norm": 0.5347883701324463, "learning_rate": 9.999967517827444e-05, "loss": 0.068, "step": 5110 }, { "grad_norm": 0.3761993646621704, "learning_rate": 9.999961284407879e-05, "loss": 0.0716, "step": 5120 }, { "grad_norm": 0.6252480745315552, "learning_rate": 9.999954504200067e-05, "loss": 0.0717, "step": 5130 }, { "grad_norm": 0.6337698101997375, "learning_rate": 9.999947177204744e-05, "loss": 0.0684, "step": 5140 }, { "grad_norm": 0.5586157441139221, "learning_rate": 9.999939303422718e-05, "loss": 0.0834, "step": 5150 }, { "grad_norm": 0.5340257287025452, "learning_rate": 9.999930882854847e-05, "loss": 0.0728, "step": 5160 }, { "grad_norm": 0.5399892330169678, "learning_rate": 9.999921915502051e-05, "loss": 0.0739, "step": 5170 }, { "grad_norm": 0.5400121808052063, "learning_rate": 9.99991240136531e-05, "loss": 0.0631, "step": 5180 }, { "grad_norm": 0.5483127236366272, "learning_rate": 9.999902340445668e-05, "loss": 0.0744, "step": 5190 }, { "grad_norm": 0.502492368221283, "learning_rate": 9.999891732744224e-05, "loss": 0.0722, "step": 5200 }, { "grad_norm": 0.44771289825439453, "learning_rate": 9.999880578262135e-05, "loss": 0.0726, "step": 5210 }, { "grad_norm": 0.6002525687217712, "learning_rate": 9.999868877000624e-05, "loss": 0.0852, "step": 5220 }, { "grad_norm": 0.4896456003189087, "learning_rate": 9.99985662896097e-05, "loss": 0.0679, "step": 5230 }, { "grad_norm": 0.5430349111557007, "learning_rate": 9.999843834144513e-05, "loss": 0.0707, "step": 5240 }, { "grad_norm": 0.46255001425743103, "learning_rate": 9.99983049255265e-05, "loss": 0.0783, "step": 5250 }, { "grad_norm": 0.3936828076839447, "learning_rate": 9.999816604186843e-05, "loss": 0.0723, "step": 5260 }, { "grad_norm": 0.4688170254230499, "learning_rate": 9.999802169048609e-05, "loss": 0.0724, "step": 5270 }, { "grad_norm": 0.5054097175598145, "learning_rate": 9.999787187139527e-05, "loss": 0.0663, "step": 5280 }, { "grad_norm": 0.5382629632949829, "learning_rate": 9.999771658461234e-05, "loss": 0.0693, "step": 5290 }, { "grad_norm": 0.4410962760448456, "learning_rate": 9.999755583015431e-05, "loss": 0.0673, "step": 5300 }, { "grad_norm": 0.4722372889518738, "learning_rate": 9.999738960803874e-05, "loss": 0.0707, "step": 5310 }, { "grad_norm": 0.5124234557151794, "learning_rate": 9.99972179182838e-05, "loss": 0.0624, "step": 5320 }, { "grad_norm": 0.5558239221572876, "learning_rate": 9.99970407609083e-05, "loss": 0.0682, "step": 5330 }, { "grad_norm": 0.4579834043979645, "learning_rate": 9.999685813593159e-05, "loss": 0.0638, "step": 5340 }, { "grad_norm": 0.43897026777267456, "learning_rate": 9.999667004337362e-05, "loss": 0.0688, "step": 5350 }, { "grad_norm": 0.4179498553276062, "learning_rate": 9.9996476483255e-05, "loss": 0.0644, "step": 5360 }, { "grad_norm": 0.43994081020355225, "learning_rate": 9.999627745559688e-05, "loss": 0.0682, "step": 5370 }, { "grad_norm": 0.5274957418441772, "learning_rate": 9.999607296042101e-05, "loss": 0.0714, "step": 5380 }, { "grad_norm": 0.4268817603588104, "learning_rate": 9.99958629977498e-05, "loss": 0.0725, "step": 5390 }, { "grad_norm": 0.4360606074333191, "learning_rate": 9.999564756760615e-05, "loss": 0.0628, "step": 5400 }, { "grad_norm": 0.423816055059433, "learning_rate": 9.999542667001366e-05, "loss": 0.0704, "step": 5410 }, { "grad_norm": 0.3717260956764221, "learning_rate": 9.999520030499647e-05, "loss": 0.0673, "step": 5420 }, { "grad_norm": 0.41009190678596497, "learning_rate": 9.999496847257936e-05, "loss": 0.0715, "step": 5430 }, { "grad_norm": 0.3671458661556244, "learning_rate": 9.999473117278764e-05, "loss": 0.066, "step": 5440 }, { "grad_norm": 0.4153347909450531, "learning_rate": 9.999448840564731e-05, "loss": 0.0684, "step": 5450 }, { "grad_norm": 0.476453959941864, "learning_rate": 9.999424017118488e-05, "loss": 0.0759, "step": 5460 }, { "grad_norm": 0.3801017999649048, "learning_rate": 9.999398646942751e-05, "loss": 0.0677, "step": 5470 }, { "grad_norm": 0.4445324242115021, "learning_rate": 9.999372730040296e-05, "loss": 0.0686, "step": 5480 }, { "grad_norm": 0.43568292260169983, "learning_rate": 9.999346266413953e-05, "loss": 0.0618, "step": 5490 }, { "grad_norm": 0.5252328515052795, "learning_rate": 9.99931925606662e-05, "loss": 0.0705, "step": 5500 }, { "grad_norm": 0.45931562781333923, "learning_rate": 9.99929169900125e-05, "loss": 0.0712, "step": 5510 }, { "grad_norm": 0.5012001395225525, "learning_rate": 9.999263595220855e-05, "loss": 0.0689, "step": 5520 }, { "grad_norm": 0.416574627161026, "learning_rate": 9.99923494472851e-05, "loss": 0.0685, "step": 5530 }, { "grad_norm": 0.3620631694793701, "learning_rate": 9.999205747527348e-05, "loss": 0.0591, "step": 5540 }, { "grad_norm": 0.41067618131637573, "learning_rate": 9.999176003620561e-05, "loss": 0.0625, "step": 5550 }, { "grad_norm": 0.5859541893005371, "learning_rate": 9.999145713011405e-05, "loss": 0.073, "step": 5560 }, { "grad_norm": 0.43763309717178345, "learning_rate": 9.999114875703186e-05, "loss": 0.0741, "step": 5570 }, { "grad_norm": 0.6057241559028625, "learning_rate": 9.999083491699281e-05, "loss": 0.0896, "step": 5580 }, { "grad_norm": 0.44714784622192383, "learning_rate": 9.999051561003123e-05, "loss": 0.0701, "step": 5590 }, { "grad_norm": 0.38291481137275696, "learning_rate": 9.999019083618202e-05, "loss": 0.0656, "step": 5600 }, { "grad_norm": 0.4049587547779083, "learning_rate": 9.99898605954807e-05, "loss": 0.0656, "step": 5610 }, { "grad_norm": 0.4798400104045868, "learning_rate": 9.998952488796338e-05, "loss": 0.0645, "step": 5620 }, { "grad_norm": 0.3955628275871277, "learning_rate": 9.998918371366676e-05, "loss": 0.0619, "step": 5630 }, { "grad_norm": 0.49062857031822205, "learning_rate": 9.99888370726282e-05, "loss": 0.063, "step": 5640 }, { "grad_norm": 0.529197096824646, "learning_rate": 9.998848496488556e-05, "loss": 0.0699, "step": 5650 }, { "grad_norm": 0.45932719111442566, "learning_rate": 9.998812739047736e-05, "loss": 0.0703, "step": 5660 }, { "grad_norm": 0.5676673054695129, "learning_rate": 9.99877643494427e-05, "loss": 0.0735, "step": 5670 }, { "grad_norm": 0.49665942788124084, "learning_rate": 9.998739584182128e-05, "loss": 0.0685, "step": 5680 }, { "grad_norm": 0.43805357813835144, "learning_rate": 9.998702186765342e-05, "loss": 0.071, "step": 5690 }, { "grad_norm": 0.5740752816200256, "learning_rate": 9.998664242698e-05, "loss": 0.0634, "step": 5700 }, { "grad_norm": 0.46155765652656555, "learning_rate": 9.998625751984251e-05, "loss": 0.0627, "step": 5710 }, { "grad_norm": 0.4992351830005646, "learning_rate": 9.998586714628307e-05, "loss": 0.0649, "step": 5720 }, { "grad_norm": 0.4454745054244995, "learning_rate": 9.998547130634432e-05, "loss": 0.0651, "step": 5730 }, { "grad_norm": 0.37209799885749817, "learning_rate": 9.99850700000696e-05, "loss": 0.0637, "step": 5740 }, { "grad_norm": 0.44835934042930603, "learning_rate": 9.998466322750278e-05, "loss": 0.0701, "step": 5750 }, { "grad_norm": 0.3839958906173706, "learning_rate": 9.998425098868834e-05, "loss": 0.0623, "step": 5760 }, { "grad_norm": 0.34093764424324036, "learning_rate": 9.998383328367136e-05, "loss": 0.0634, "step": 5770 }, { "grad_norm": 0.4694533050060272, "learning_rate": 9.99834101124975e-05, "loss": 0.0597, "step": 5780 }, { "grad_norm": 0.43764132261276245, "learning_rate": 9.998298147521309e-05, "loss": 0.067, "step": 5790 }, { "grad_norm": 0.4086277186870575, "learning_rate": 9.998254737186496e-05, "loss": 0.0645, "step": 5800 }, { "grad_norm": 0.3776073157787323, "learning_rate": 9.99821078025006e-05, "loss": 0.0602, "step": 5810 }, { "grad_norm": 0.4689655900001526, "learning_rate": 9.998166276716807e-05, "loss": 0.0651, "step": 5820 }, { "grad_norm": 0.39124995470046997, "learning_rate": 9.998121226591606e-05, "loss": 0.0593, "step": 5830 }, { "grad_norm": 0.4834187626838684, "learning_rate": 9.998075629879382e-05, "loss": 0.06, "step": 5840 }, { "grad_norm": 0.658681333065033, "learning_rate": 9.99802948658512e-05, "loss": 0.0641, "step": 5850 }, { "grad_norm": 0.48996639251708984, "learning_rate": 9.99798279671387e-05, "loss": 0.0709, "step": 5860 }, { "grad_norm": 0.49412021040916443, "learning_rate": 9.997935560270734e-05, "loss": 0.0741, "step": 5870 }, { "grad_norm": 0.4633129835128784, "learning_rate": 9.997887777260879e-05, "loss": 0.0659, "step": 5880 }, { "grad_norm": 0.3745655417442322, "learning_rate": 9.997839447689532e-05, "loss": 0.0661, "step": 5890 }, { "grad_norm": 0.46942517161369324, "learning_rate": 9.997790571561978e-05, "loss": 0.0645, "step": 5900 }, { "grad_norm": 0.31423303484916687, "learning_rate": 9.99774114888356e-05, "loss": 0.0615, "step": 5910 }, { "grad_norm": 0.4233357310295105, "learning_rate": 9.997691179659684e-05, "loss": 0.0743, "step": 5920 }, { "grad_norm": 0.37401002645492554, "learning_rate": 9.997640663895815e-05, "loss": 0.0624, "step": 5930 }, { "grad_norm": 0.44184204936027527, "learning_rate": 9.997589601597477e-05, "loss": 0.0605, "step": 5940 }, { "grad_norm": 0.37135618925094604, "learning_rate": 9.997537992770252e-05, "loss": 0.061, "step": 5950 }, { "grad_norm": 0.38284629583358765, "learning_rate": 9.997485837419788e-05, "loss": 0.0566, "step": 5960 }, { "grad_norm": 0.48000603914260864, "learning_rate": 9.997433135551786e-05, "loss": 0.0597, "step": 5970 }, { "grad_norm": 0.3971227705478668, "learning_rate": 9.997379887172009e-05, "loss": 0.0609, "step": 5980 }, { "grad_norm": 0.36790913343429565, "learning_rate": 9.997326092286281e-05, "loss": 0.061, "step": 5990 }, { "grad_norm": 0.5008909702301025, "learning_rate": 9.997271750900486e-05, "loss": 0.0653, "step": 6000 }, { "grad_norm": 0.4411888122558594, "learning_rate": 9.997216863020565e-05, "loss": 0.0681, "step": 6010 }, { "grad_norm": 0.4289947748184204, "learning_rate": 9.99716142865252e-05, "loss": 0.0663, "step": 6020 }, { "grad_norm": 0.5282012820243835, "learning_rate": 9.997105447802415e-05, "loss": 0.0599, "step": 6030 }, { "grad_norm": 0.32874828577041626, "learning_rate": 9.997048920476373e-05, "loss": 0.0664, "step": 6040 }, { "grad_norm": 0.4184695780277252, "learning_rate": 9.996991846680572e-05, "loss": 0.0627, "step": 6050 }, { "grad_norm": 0.372433066368103, "learning_rate": 9.996934226421257e-05, "loss": 0.0581, "step": 6060 }, { "grad_norm": 0.36019638180732727, "learning_rate": 9.996876059704726e-05, "loss": 0.0595, "step": 6070 }, { "grad_norm": 0.42039039731025696, "learning_rate": 9.996817346537343e-05, "loss": 0.0564, "step": 6080 }, { "grad_norm": 0.4089832901954651, "learning_rate": 9.996758086925526e-05, "loss": 0.0625, "step": 6090 }, { "grad_norm": 0.4162515699863434, "learning_rate": 9.996698280875759e-05, "loss": 0.0553, "step": 6100 }, { "grad_norm": 0.4194429814815521, "learning_rate": 9.99663792839458e-05, "loss": 0.0639, "step": 6110 }, { "grad_norm": 0.4462501108646393, "learning_rate": 9.99657702948859e-05, "loss": 0.0606, "step": 6120 }, { "grad_norm": 0.529503583908081, "learning_rate": 9.996515584164448e-05, "loss": 0.059, "step": 6130 }, { "grad_norm": 0.4573458135128021, "learning_rate": 9.996453592428873e-05, "loss": 0.0669, "step": 6140 }, { "grad_norm": 0.489671915769577, "learning_rate": 9.996391054288646e-05, "loss": 0.061, "step": 6150 }, { "grad_norm": 0.43633735179901123, "learning_rate": 9.996327969750605e-05, "loss": 0.0677, "step": 6160 }, { "grad_norm": 0.5040770173072815, "learning_rate": 9.996264338821649e-05, "loss": 0.0646, "step": 6170 }, { "grad_norm": 0.4780338406562805, "learning_rate": 9.996200161508735e-05, "loss": 0.0589, "step": 6180 }, { "grad_norm": 0.3810797929763794, "learning_rate": 9.996135437818885e-05, "loss": 0.0623, "step": 6190 }, { "grad_norm": 0.5644223093986511, "learning_rate": 9.996070167759175e-05, "loss": 0.0695, "step": 6200 }, { "grad_norm": 0.4651232957839966, "learning_rate": 9.996004351336743e-05, "loss": 0.0771, "step": 6210 }, { "grad_norm": 0.39103323221206665, "learning_rate": 9.995937988558785e-05, "loss": 0.0633, "step": 6220 }, { "grad_norm": 0.3625045716762543, "learning_rate": 9.995871079432561e-05, "loss": 0.0591, "step": 6230 }, { "grad_norm": 0.44025978446006775, "learning_rate": 9.995803623965389e-05, "loss": 0.061, "step": 6240 }, { "grad_norm": 0.4003317952156067, "learning_rate": 9.995735622164641e-05, "loss": 0.0666, "step": 6250 }, { "grad_norm": 0.4548111855983734, "learning_rate": 9.995667074037758e-05, "loss": 0.0728, "step": 6260 }, { "grad_norm": 0.42135369777679443, "learning_rate": 9.995597979592232e-05, "loss": 0.0582, "step": 6270 }, { "grad_norm": 0.3922084867954254, "learning_rate": 9.995528338835625e-05, "loss": 0.0613, "step": 6280 }, { "grad_norm": 0.5386683344841003, "learning_rate": 9.995458151775547e-05, "loss": 0.071, "step": 6290 }, { "grad_norm": 0.4027347266674042, "learning_rate": 9.995387418419677e-05, "loss": 0.0638, "step": 6300 }, { "grad_norm": 0.4193507134914398, "learning_rate": 9.99531613877575e-05, "loss": 0.0579, "step": 6310 }, { "grad_norm": 0.49114787578582764, "learning_rate": 9.995244312851559e-05, "loss": 0.0583, "step": 6320 }, { "grad_norm": 0.3663327097892761, "learning_rate": 9.995171940654961e-05, "loss": 0.0614, "step": 6330 }, { "grad_norm": 0.28641000390052795, "learning_rate": 9.995099022193871e-05, "loss": 0.0651, "step": 6340 }, { "grad_norm": 0.4367108643054962, "learning_rate": 9.995025557476261e-05, "loss": 0.0588, "step": 6350 }, { "grad_norm": 0.38948625326156616, "learning_rate": 9.994951546510165e-05, "loss": 0.066, "step": 6360 }, { "grad_norm": 0.3772992789745331, "learning_rate": 9.994876989303679e-05, "loss": 0.0699, "step": 6370 }, { "grad_norm": 0.40824148058891296, "learning_rate": 9.994801885864955e-05, "loss": 0.0606, "step": 6380 }, { "grad_norm": 0.3274983763694763, "learning_rate": 9.994726236202205e-05, "loss": 0.0556, "step": 6390 }, { "grad_norm": 0.41912007331848145, "learning_rate": 9.994650040323704e-05, "loss": 0.0639, "step": 6400 }, { "grad_norm": 0.421232670545578, "learning_rate": 9.994573298237784e-05, "loss": 0.0554, "step": 6410 }, { "grad_norm": 0.3889963626861572, "learning_rate": 9.994496009952837e-05, "loss": 0.064, "step": 6420 }, { "grad_norm": 0.39083436131477356, "learning_rate": 9.994418175477316e-05, "loss": 0.0679, "step": 6430 }, { "grad_norm": 0.4576594829559326, "learning_rate": 9.994339794819733e-05, "loss": 0.0643, "step": 6440 }, { "grad_norm": 0.5868600606918335, "learning_rate": 9.994260867988658e-05, "loss": 0.0817, "step": 6450 }, { "grad_norm": 0.3996567130088806, "learning_rate": 9.994181394992723e-05, "loss": 0.061, "step": 6460 }, { "grad_norm": 0.43284639716148376, "learning_rate": 9.994101375840618e-05, "loss": 0.0596, "step": 6470 }, { "grad_norm": 0.4197075366973877, "learning_rate": 9.994020810541098e-05, "loss": 0.0655, "step": 6480 }, { "grad_norm": 0.43407168984413147, "learning_rate": 9.99393969910297e-05, "loss": 0.066, "step": 6490 }, { "grad_norm": 0.32388022541999817, "learning_rate": 9.993858041535104e-05, "loss": 0.0632, "step": 6500 }, { "grad_norm": 0.429263174533844, "learning_rate": 9.99377583784643e-05, "loss": 0.0656, "step": 6510 }, { "grad_norm": 0.4705433249473572, "learning_rate": 9.993693088045939e-05, "loss": 0.0671, "step": 6520 }, { "grad_norm": 0.4611456096172333, "learning_rate": 9.99360979214268e-05, "loss": 0.0621, "step": 6530 }, { "grad_norm": 0.4936104118824005, "learning_rate": 9.99352595014576e-05, "loss": 0.0625, "step": 6540 }, { "grad_norm": 0.5055278539657593, "learning_rate": 9.993441562064354e-05, "loss": 0.066, "step": 6550 }, { "grad_norm": 0.44090336561203003, "learning_rate": 9.993356627907685e-05, "loss": 0.0585, "step": 6560 }, { "grad_norm": 0.3717265725135803, "learning_rate": 9.99327114768504e-05, "loss": 0.0675, "step": 6570 }, { "grad_norm": 0.5679259300231934, "learning_rate": 9.99318512140577e-05, "loss": 0.0627, "step": 6580 }, { "grad_norm": 0.35935190320014954, "learning_rate": 9.993098549079284e-05, "loss": 0.0634, "step": 6590 }, { "grad_norm": 0.3602069616317749, "learning_rate": 9.993011430715047e-05, "loss": 0.0668, "step": 6600 }, { "grad_norm": 0.4148220419883728, "learning_rate": 9.992923766322586e-05, "loss": 0.0614, "step": 6610 }, { "grad_norm": 0.3988305926322937, "learning_rate": 9.99283555591149e-05, "loss": 0.0661, "step": 6620 }, { "grad_norm": 0.4122796952724457, "learning_rate": 9.992746799491404e-05, "loss": 0.0667, "step": 6630 }, { "grad_norm": 0.41256919503211975, "learning_rate": 9.992657497072033e-05, "loss": 0.0656, "step": 6640 }, { "grad_norm": 0.48476362228393555, "learning_rate": 9.992567648663147e-05, "loss": 0.063, "step": 6650 }, { "grad_norm": 0.5337268710136414, "learning_rate": 9.992477254274568e-05, "loss": 0.0659, "step": 6660 }, { "grad_norm": 0.31555628776550293, "learning_rate": 9.992386313916183e-05, "loss": 0.0635, "step": 6670 }, { "grad_norm": 0.5259724855422974, "learning_rate": 9.992294827597934e-05, "loss": 0.0622, "step": 6680 }, { "grad_norm": 0.40001803636550903, "learning_rate": 9.992202795329831e-05, "loss": 0.0652, "step": 6690 }, { "grad_norm": 0.4760171175003052, "learning_rate": 9.992110217121936e-05, "loss": 0.0619, "step": 6700 }, { "grad_norm": 0.40129539370536804, "learning_rate": 9.992017092984372e-05, "loss": 0.0598, "step": 6710 }, { "grad_norm": 0.3645053803920746, "learning_rate": 9.991923422927326e-05, "loss": 0.0608, "step": 6720 }, { "grad_norm": 0.38147401809692383, "learning_rate": 9.991829206961037e-05, "loss": 0.0637, "step": 6730 }, { "grad_norm": 0.36863523721694946, "learning_rate": 9.991734445095813e-05, "loss": 0.0603, "step": 6740 }, { "grad_norm": 0.4895373284816742, "learning_rate": 9.991639137342015e-05, "loss": 0.0615, "step": 6750 }, { "grad_norm": 0.45460474491119385, "learning_rate": 9.991543283710064e-05, "loss": 0.0625, "step": 6760 }, { "grad_norm": 0.39817842841148376, "learning_rate": 9.991446884210445e-05, "loss": 0.0641, "step": 6770 }, { "grad_norm": 0.44588083028793335, "learning_rate": 9.9913499388537e-05, "loss": 0.0598, "step": 6780 }, { "grad_norm": 0.40845194458961487, "learning_rate": 9.99125244765043e-05, "loss": 0.0637, "step": 6790 }, { "grad_norm": 0.4782378673553467, "learning_rate": 9.991154410611296e-05, "loss": 0.0668, "step": 6800 }, { "grad_norm": 0.4596309959888458, "learning_rate": 9.99105582774702e-05, "loss": 0.0578, "step": 6810 }, { "grad_norm": 0.3839770257472992, "learning_rate": 9.990956699068384e-05, "loss": 0.0743, "step": 6820 }, { "grad_norm": 0.3816467225551605, "learning_rate": 9.990857024586224e-05, "loss": 0.0607, "step": 6830 }, { "grad_norm": 0.4536939263343811, "learning_rate": 9.990756804311446e-05, "loss": 0.0603, "step": 6840 }, { "grad_norm": 0.49158498644828796, "learning_rate": 9.990656038255006e-05, "loss": 0.0633, "step": 6850 }, { "grad_norm": 0.4835948050022125, "learning_rate": 9.990554726427926e-05, "loss": 0.056, "step": 6860 }, { "grad_norm": 0.39567163586616516, "learning_rate": 9.990452868841284e-05, "loss": 0.0637, "step": 6870 }, { "grad_norm": 0.37019622325897217, "learning_rate": 9.99035046550622e-05, "loss": 0.0606, "step": 6880 }, { "grad_norm": 0.35363200306892395, "learning_rate": 9.99024751643393e-05, "loss": 0.0479, "step": 6890 }, { "grad_norm": 0.43468692898750305, "learning_rate": 9.990144021635677e-05, "loss": 0.0599, "step": 6900 }, { "grad_norm": 0.3967653214931488, "learning_rate": 9.990039981122775e-05, "loss": 0.0671, "step": 6910 }, { "grad_norm": 0.4175512194633484, "learning_rate": 9.989935394906602e-05, "loss": 0.0624, "step": 6920 }, { "grad_norm": 0.31488215923309326, "learning_rate": 9.989830262998598e-05, "loss": 0.0628, "step": 6930 }, { "grad_norm": 0.44178804755210876, "learning_rate": 9.989724585410259e-05, "loss": 0.0683, "step": 6940 }, { "grad_norm": 0.47058746218681335, "learning_rate": 9.989618362153139e-05, "loss": 0.0589, "step": 6950 }, { "grad_norm": 0.4375351071357727, "learning_rate": 9.989511593238859e-05, "loss": 0.0596, "step": 6960 }, { "grad_norm": 0.44586050510406494, "learning_rate": 9.98940427867909e-05, "loss": 0.0607, "step": 6970 }, { "grad_norm": 0.3937840461730957, "learning_rate": 9.989296418485573e-05, "loss": 0.0557, "step": 6980 }, { "grad_norm": 0.392903596162796, "learning_rate": 9.989188012670101e-05, "loss": 0.0607, "step": 6990 }, { "grad_norm": 0.4971265196800232, "learning_rate": 9.989079061244528e-05, "loss": 0.0587, "step": 7000 }, { "grad_norm": 0.4158191978931427, "learning_rate": 9.988969564220769e-05, "loss": 0.0624, "step": 7010 }, { "grad_norm": 0.45910224318504333, "learning_rate": 9.988859521610801e-05, "loss": 0.0743, "step": 7020 }, { "grad_norm": 0.419558048248291, "learning_rate": 9.988748933426656e-05, "loss": 0.0616, "step": 7030 }, { "grad_norm": 0.4416235387325287, "learning_rate": 9.988637799680428e-05, "loss": 0.0661, "step": 7040 }, { "grad_norm": 0.3723473846912384, "learning_rate": 9.98852612038427e-05, "loss": 0.0659, "step": 7050 }, { "grad_norm": 0.3734408915042877, "learning_rate": 9.988413895550397e-05, "loss": 0.057, "step": 7060 }, { "grad_norm": 0.37401098012924194, "learning_rate": 9.98830112519108e-05, "loss": 0.0611, "step": 7070 }, { "grad_norm": 0.41399917006492615, "learning_rate": 9.98818780931865e-05, "loss": 0.0562, "step": 7080 }, { "grad_norm": 0.4289097785949707, "learning_rate": 9.988073947945502e-05, "loss": 0.063, "step": 7090 }, { "grad_norm": 0.4044332206249237, "learning_rate": 9.987959541084087e-05, "loss": 0.059, "step": 7100 }, { "grad_norm": 0.4669601321220398, "learning_rate": 9.987844588746915e-05, "loss": 0.0622, "step": 7110 }, { "grad_norm": 0.433123379945755, "learning_rate": 9.987729090946558e-05, "loss": 0.059, "step": 7120 }, { "grad_norm": 0.3678428530693054, "learning_rate": 9.987613047695647e-05, "loss": 0.0581, "step": 7130 }, { "grad_norm": 0.4898626506328583, "learning_rate": 9.987496459006871e-05, "loss": 0.0685, "step": 7140 }, { "grad_norm": 0.36681997776031494, "learning_rate": 9.987379324892982e-05, "loss": 0.068, "step": 7150 }, { "grad_norm": 0.3563937246799469, "learning_rate": 9.987261645366788e-05, "loss": 0.0645, "step": 7160 }, { "grad_norm": 0.39009222388267517, "learning_rate": 9.987143420441158e-05, "loss": 0.0599, "step": 7170 }, { "grad_norm": 0.3265001177787781, "learning_rate": 9.987024650129022e-05, "loss": 0.0556, "step": 7180 }, { "grad_norm": 0.5930823087692261, "learning_rate": 9.986905334443368e-05, "loss": 0.06, "step": 7190 }, { "grad_norm": 0.47412851452827454, "learning_rate": 9.986785473397245e-05, "loss": 0.0699, "step": 7200 }, { "grad_norm": 0.4470379948616028, "learning_rate": 9.98666506700376e-05, "loss": 0.0661, "step": 7210 }, { "grad_norm": 0.4803468585014343, "learning_rate": 9.986544115276081e-05, "loss": 0.0582, "step": 7220 }, { "grad_norm": 0.4237244725227356, "learning_rate": 9.986422618227433e-05, "loss": 0.0573, "step": 7230 }, { "grad_norm": 0.36857905983924866, "learning_rate": 9.986300575871106e-05, "loss": 0.0566, "step": 7240 }, { "grad_norm": 0.34338295459747314, "learning_rate": 9.986177988220444e-05, "loss": 0.053, "step": 7250 }, { "grad_norm": 0.3773908019065857, "learning_rate": 9.986054855288856e-05, "loss": 0.0692, "step": 7260 }, { "grad_norm": 0.44634464383125305, "learning_rate": 9.985931177089802e-05, "loss": 0.0535, "step": 7270 }, { "grad_norm": 0.49560433626174927, "learning_rate": 9.985806953636814e-05, "loss": 0.0605, "step": 7280 }, { "grad_norm": 0.46300116181373596, "learning_rate": 9.985682184943471e-05, "loss": 0.0643, "step": 7290 }, { "grad_norm": 0.4477812647819519, "learning_rate": 9.98555687102342e-05, "loss": 0.0553, "step": 7300 }, { "grad_norm": 0.3783385455608368, "learning_rate": 9.985431011890367e-05, "loss": 0.0567, "step": 7310 }, { "grad_norm": 0.36617136001586914, "learning_rate": 9.985304607558075e-05, "loss": 0.0558, "step": 7320 }, { "grad_norm": 0.35619083046913147, "learning_rate": 9.985177658040364e-05, "loss": 0.0517, "step": 7330 }, { "grad_norm": 0.3031013607978821, "learning_rate": 9.985050163351119e-05, "loss": 0.0541, "step": 7340 }, { "grad_norm": 0.379157155752182, "learning_rate": 9.984922123504286e-05, "loss": 0.0611, "step": 7350 }, { "grad_norm": 0.3369566798210144, "learning_rate": 9.984793538513862e-05, "loss": 0.053, "step": 7360 }, { "grad_norm": 0.47774171829223633, "learning_rate": 9.984664408393912e-05, "loss": 0.0571, "step": 7370 }, { "grad_norm": 0.3566886782646179, "learning_rate": 9.984534733158556e-05, "loss": 0.057, "step": 7380 }, { "grad_norm": 0.6400287747383118, "learning_rate": 9.984404512821977e-05, "loss": 0.0493, "step": 7390 }, { "grad_norm": 0.4516407549381256, "learning_rate": 9.984273747398411e-05, "loss": 0.0623, "step": 7400 }, { "grad_norm": 0.3866582214832306, "learning_rate": 9.984142436902165e-05, "loss": 0.0624, "step": 7410 }, { "grad_norm": 0.3809050917625427, "learning_rate": 9.984010581347596e-05, "loss": 0.058, "step": 7420 }, { "grad_norm": 0.47573357820510864, "learning_rate": 9.983878180749121e-05, "loss": 0.0548, "step": 7430 }, { "grad_norm": 0.4271972179412842, "learning_rate": 9.983745235121222e-05, "loss": 0.0557, "step": 7440 }, { "grad_norm": 0.31917282938957214, "learning_rate": 9.983611744478438e-05, "loss": 0.0593, "step": 7450 }, { "grad_norm": 0.3405366837978363, "learning_rate": 9.983477708835365e-05, "loss": 0.0591, "step": 7460 }, { "grad_norm": 0.41506457328796387, "learning_rate": 9.983343128206664e-05, "loss": 0.0643, "step": 7470 }, { "grad_norm": 0.45229607820510864, "learning_rate": 9.983208002607049e-05, "loss": 0.0592, "step": 7480 }, { "grad_norm": 0.4501022696495056, "learning_rate": 9.9830723320513e-05, "loss": 0.066, "step": 7490 }, { "grad_norm": 0.4163224995136261, "learning_rate": 9.982936116554254e-05, "loss": 0.0569, "step": 7500 }, { "grad_norm": 0.3684874176979065, "learning_rate": 9.982799356130803e-05, "loss": 0.0546, "step": 7510 }, { "grad_norm": 0.3686186373233795, "learning_rate": 9.982662050795908e-05, "loss": 0.0654, "step": 7520 }, { "grad_norm": 0.37556663155555725, "learning_rate": 9.982524200564583e-05, "loss": 0.0561, "step": 7530 }, { "grad_norm": 0.3914152681827545, "learning_rate": 9.982385805451901e-05, "loss": 0.0587, "step": 7540 }, { "grad_norm": 0.4327312707901001, "learning_rate": 9.982246865472998e-05, "loss": 0.0602, "step": 7550 }, { "grad_norm": 0.5125383138656616, "learning_rate": 9.982107380643069e-05, "loss": 0.0623, "step": 7560 }, { "grad_norm": 0.5124803781509399, "learning_rate": 9.981967350977368e-05, "loss": 0.062, "step": 7570 }, { "grad_norm": 0.38313454389572144, "learning_rate": 9.981826776491208e-05, "loss": 0.053, "step": 7580 }, { "grad_norm": 0.43065524101257324, "learning_rate": 9.98168565719996e-05, "loss": 0.0641, "step": 7590 }, { "grad_norm": 0.3735371232032776, "learning_rate": 9.98154399311906e-05, "loss": 0.0539, "step": 7600 }, { "grad_norm": 0.3520658314228058, "learning_rate": 9.981401784263997e-05, "loss": 0.0615, "step": 7610 }, { "grad_norm": 0.34766101837158203, "learning_rate": 9.981259030650326e-05, "loss": 0.0609, "step": 7620 }, { "grad_norm": 0.3086588382720947, "learning_rate": 9.981115732293655e-05, "loss": 0.0551, "step": 7630 }, { "grad_norm": 0.38403406739234924, "learning_rate": 9.980971889209659e-05, "loss": 0.0678, "step": 7640 }, { "grad_norm": 0.3975987732410431, "learning_rate": 9.980827501414064e-05, "loss": 0.0609, "step": 7650 }, { "grad_norm": 0.35783103108406067, "learning_rate": 9.980682568922663e-05, "loss": 0.056, "step": 7660 }, { "grad_norm": 0.3945101797580719, "learning_rate": 9.980537091751304e-05, "loss": 0.0524, "step": 7670 }, { "grad_norm": 0.27533429861068726, "learning_rate": 9.980391069915897e-05, "loss": 0.0558, "step": 7680 }, { "grad_norm": 0.5218346118927002, "learning_rate": 9.98024450343241e-05, "loss": 0.0675, "step": 7690 }, { "grad_norm": 0.31994062662124634, "learning_rate": 9.980097392316872e-05, "loss": 0.0525, "step": 7700 }, { "grad_norm": 0.29299983382225037, "learning_rate": 9.97994973658537e-05, "loss": 0.0543, "step": 7710 }, { "grad_norm": 0.4209241271018982, "learning_rate": 9.979801536254054e-05, "loss": 0.0597, "step": 7720 }, { "grad_norm": 0.3685998022556305, "learning_rate": 9.979652791339127e-05, "loss": 0.0567, "step": 7730 }, { "grad_norm": 0.42757943272590637, "learning_rate": 9.97950350185686e-05, "loss": 0.0504, "step": 7740 }, { "grad_norm": 0.3406164348125458, "learning_rate": 9.979353667823574e-05, "loss": 0.0564, "step": 7750 }, { "grad_norm": 0.3971841037273407, "learning_rate": 9.979203289255658e-05, "loss": 0.0536, "step": 7760 }, { "grad_norm": 0.32978370785713196, "learning_rate": 9.979052366169557e-05, "loss": 0.0618, "step": 7770 }, { "grad_norm": 0.433081716299057, "learning_rate": 9.978900898581775e-05, "loss": 0.0577, "step": 7780 }, { "grad_norm": 0.41083478927612305, "learning_rate": 9.978748886508875e-05, "loss": 0.0612, "step": 7790 }, { "grad_norm": 0.37839198112487793, "learning_rate": 9.978596329967484e-05, "loss": 0.0671, "step": 7800 }, { "grad_norm": 0.3202023506164551, "learning_rate": 9.978443228974284e-05, "loss": 0.0523, "step": 7810 }, { "grad_norm": 0.27494823932647705, "learning_rate": 9.978289583546015e-05, "loss": 0.053, "step": 7820 }, { "grad_norm": 0.347945898771286, "learning_rate": 9.978135393699484e-05, "loss": 0.0549, "step": 7830 }, { "grad_norm": 0.35923078656196594, "learning_rate": 9.977980659451548e-05, "loss": 0.0695, "step": 7840 }, { "grad_norm": 0.35722291469573975, "learning_rate": 9.977825380819135e-05, "loss": 0.0564, "step": 7850 }, { "grad_norm": 0.3554140329360962, "learning_rate": 9.97766955781922e-05, "loss": 0.0491, "step": 7860 }, { "grad_norm": 0.4632960259914398, "learning_rate": 9.977513190468848e-05, "loss": 0.0632, "step": 7870 }, { "grad_norm": 0.3726649880409241, "learning_rate": 9.977356278785116e-05, "loss": 0.0546, "step": 7880 }, { "grad_norm": 0.3954155445098877, "learning_rate": 9.977198822785184e-05, "loss": 0.0546, "step": 7890 }, { "grad_norm": 0.593625009059906, "learning_rate": 9.977040822486273e-05, "loss": 0.0608, "step": 7900 }, { "grad_norm": 0.470518559217453, "learning_rate": 9.97688227790566e-05, "loss": 0.0612, "step": 7910 }, { "grad_norm": 0.3751290738582611, "learning_rate": 9.976723189060684e-05, "loss": 0.0547, "step": 7920 }, { "grad_norm": 0.41525697708129883, "learning_rate": 9.976563555968742e-05, "loss": 0.0602, "step": 7930 }, { "grad_norm": 0.46641916036605835, "learning_rate": 9.976403378647292e-05, "loss": 0.0584, "step": 7940 }, { "grad_norm": 0.41880905628204346, "learning_rate": 9.97624265711385e-05, "loss": 0.0566, "step": 7950 }, { "grad_norm": 0.5123955607414246, "learning_rate": 9.976081391385993e-05, "loss": 0.0551, "step": 7960 }, { "grad_norm": 0.3429124653339386, "learning_rate": 9.975919581481356e-05, "loss": 0.0524, "step": 7970 }, { "grad_norm": 0.3612515926361084, "learning_rate": 9.975757227417634e-05, "loss": 0.0602, "step": 7980 }, { "grad_norm": 0.3666611909866333, "learning_rate": 9.975594329212586e-05, "loss": 0.0585, "step": 7990 }, { "grad_norm": 0.3346412181854248, "learning_rate": 9.97543088688402e-05, "loss": 0.0564, "step": 8000 }, { "grad_norm": 0.35814717411994934, "learning_rate": 9.975266900449814e-05, "loss": 0.048, "step": 8010 }, { "grad_norm": 0.4259287416934967, "learning_rate": 9.975102369927898e-05, "loss": 0.0547, "step": 8020 }, { "grad_norm": 0.3920745551586151, "learning_rate": 9.974937295336269e-05, "loss": 0.053, "step": 8030 }, { "grad_norm": 0.305461049079895, "learning_rate": 9.974771676692975e-05, "loss": 0.052, "step": 8040 }, { "grad_norm": 0.39729663729667664, "learning_rate": 9.974605514016131e-05, "loss": 0.0575, "step": 8050 }, { "grad_norm": 0.678360104560852, "learning_rate": 9.974438807323907e-05, "loss": 0.0533, "step": 8060 }, { "grad_norm": 0.3666832149028778, "learning_rate": 9.974271556634535e-05, "loss": 0.0601, "step": 8070 }, { "grad_norm": 0.43938538432121277, "learning_rate": 9.974103761966302e-05, "loss": 0.0582, "step": 8080 }, { "grad_norm": 0.4027506113052368, "learning_rate": 9.973935423337563e-05, "loss": 0.0567, "step": 8090 }, { "grad_norm": 0.32525983452796936, "learning_rate": 9.973766540766722e-05, "loss": 0.0622, "step": 8100 }, { "grad_norm": 0.29469457268714905, "learning_rate": 9.97359711427225e-05, "loss": 0.0524, "step": 8110 }, { "grad_norm": 0.3414384424686432, "learning_rate": 9.973427143872677e-05, "loss": 0.0577, "step": 8120 }, { "grad_norm": 0.383689820766449, "learning_rate": 9.973256629586589e-05, "loss": 0.0511, "step": 8130 }, { "grad_norm": 0.3772604167461395, "learning_rate": 9.973085571432632e-05, "loss": 0.0558, "step": 8140 }, { "grad_norm": 0.44888898730278015, "learning_rate": 9.972913969429513e-05, "loss": 0.0544, "step": 8150 }, { "grad_norm": 0.37680238485336304, "learning_rate": 9.972741823596e-05, "loss": 0.0571, "step": 8160 }, { "grad_norm": 0.37099596858024597, "learning_rate": 9.972569133950917e-05, "loss": 0.0598, "step": 8170 }, { "grad_norm": 0.4273243844509125, "learning_rate": 9.972395900513151e-05, "loss": 0.0573, "step": 8180 }, { "grad_norm": 0.4244578778743744, "learning_rate": 9.972222123301645e-05, "loss": 0.055, "step": 8190 }, { "grad_norm": 0.4193064272403717, "learning_rate": 9.972047802335403e-05, "loss": 0.0555, "step": 8200 }, { "grad_norm": 0.47116440534591675, "learning_rate": 9.971872937633488e-05, "loss": 0.0607, "step": 8210 }, { "grad_norm": 0.4272960424423218, "learning_rate": 9.971697529215024e-05, "loss": 0.061, "step": 8220 }, { "grad_norm": 0.3662150800228119, "learning_rate": 9.971521577099192e-05, "loss": 0.061, "step": 8230 }, { "grad_norm": 0.455067902803421, "learning_rate": 9.971345081305236e-05, "loss": 0.0652, "step": 8240 }, { "grad_norm": 0.4713378846645355, "learning_rate": 9.971168041852456e-05, "loss": 0.0611, "step": 8250 }, { "grad_norm": 0.4370878338813782, "learning_rate": 9.970990458760215e-05, "loss": 0.064, "step": 8260 }, { "grad_norm": 0.49261829257011414, "learning_rate": 9.970812332047929e-05, "loss": 0.0676, "step": 8270 }, { "grad_norm": 0.3708643615245819, "learning_rate": 9.97063366173508e-05, "loss": 0.0639, "step": 8280 }, { "grad_norm": 0.4174812436103821, "learning_rate": 9.970454447841207e-05, "loss": 0.0618, "step": 8290 }, { "grad_norm": 0.4190714955329895, "learning_rate": 9.970274690385909e-05, "loss": 0.0586, "step": 8300 }, { "grad_norm": 0.3659147024154663, "learning_rate": 9.970094389388844e-05, "loss": 0.0593, "step": 8310 }, { "grad_norm": 0.43824008107185364, "learning_rate": 9.969913544869728e-05, "loss": 0.0601, "step": 8320 }, { "grad_norm": 0.49778735637664795, "learning_rate": 9.96973215684834e-05, "loss": 0.0617, "step": 8330 }, { "grad_norm": 0.48293212056159973, "learning_rate": 9.969550225344513e-05, "loss": 0.061, "step": 8340 }, { "grad_norm": 0.3440205454826355, "learning_rate": 9.969367750378147e-05, "loss": 0.0591, "step": 8350 }, { "grad_norm": 0.3206910789012909, "learning_rate": 9.969184731969194e-05, "loss": 0.0552, "step": 8360 }, { "grad_norm": 0.36826393008232117, "learning_rate": 9.96900117013767e-05, "loss": 0.0576, "step": 8370 }, { "grad_norm": 0.40052634477615356, "learning_rate": 9.96881706490365e-05, "loss": 0.067, "step": 8380 }, { "grad_norm": 0.3707781732082367, "learning_rate": 9.968632416287265e-05, "loss": 0.0559, "step": 8390 }, { "grad_norm": 0.3988703191280365, "learning_rate": 9.96844722430871e-05, "loss": 0.0557, "step": 8400 }, { "grad_norm": 0.4019851088523865, "learning_rate": 9.968261488988235e-05, "loss": 0.0617, "step": 8410 }, { "grad_norm": 0.3890961706638336, "learning_rate": 9.968075210346155e-05, "loss": 0.0506, "step": 8420 }, { "grad_norm": 0.45266035199165344, "learning_rate": 9.967888388402839e-05, "loss": 0.0584, "step": 8430 }, { "grad_norm": 0.3738234341144562, "learning_rate": 9.967701023178717e-05, "loss": 0.0509, "step": 8440 }, { "grad_norm": 0.36645132303237915, "learning_rate": 9.967513114694282e-05, "loss": 0.0571, "step": 8450 }, { "grad_norm": 0.40540656447410583, "learning_rate": 9.967324662970079e-05, "loss": 0.0585, "step": 8460 }, { "grad_norm": 0.3499002456665039, "learning_rate": 9.96713566802672e-05, "loss": 0.0506, "step": 8470 }, { "grad_norm": 0.27401143312454224, "learning_rate": 9.966946129884873e-05, "loss": 0.0604, "step": 8480 }, { "grad_norm": 0.3800053596496582, "learning_rate": 9.966756048565265e-05, "loss": 0.061, "step": 8490 }, { "grad_norm": 0.3794548809528351, "learning_rate": 9.966565424088681e-05, "loss": 0.0554, "step": 8500 }, { "grad_norm": 0.39830565452575684, "learning_rate": 9.96637425647597e-05, "loss": 0.0561, "step": 8510 }, { "grad_norm": 0.3573337197303772, "learning_rate": 9.966182545748038e-05, "loss": 0.0553, "step": 8520 }, { "grad_norm": 0.3770023286342621, "learning_rate": 9.96599029192585e-05, "loss": 0.055, "step": 8530 }, { "grad_norm": 0.38386034965515137, "learning_rate": 9.965797495030428e-05, "loss": 0.0579, "step": 8540 }, { "grad_norm": 0.47824859619140625, "learning_rate": 9.96560415508286e-05, "loss": 0.0588, "step": 8550 }, { "grad_norm": 0.5034968256950378, "learning_rate": 9.965410272104286e-05, "loss": 0.0559, "step": 8560 }, { "grad_norm": 0.5682412385940552, "learning_rate": 9.96521584611591e-05, "loss": 0.0598, "step": 8570 }, { "grad_norm": 0.43288829922676086, "learning_rate": 9.965020877138994e-05, "loss": 0.059, "step": 8580 }, { "grad_norm": 0.4915772080421448, "learning_rate": 9.964825365194861e-05, "loss": 0.055, "step": 8590 }, { "grad_norm": 0.31247764825820923, "learning_rate": 9.96462931030489e-05, "loss": 0.0514, "step": 8600 }, { "grad_norm": 0.3338339924812317, "learning_rate": 9.96443271249052e-05, "loss": 0.0482, "step": 8610 }, { "grad_norm": 0.3738088309764862, "learning_rate": 9.964235571773255e-05, "loss": 0.0578, "step": 8620 }, { "grad_norm": 0.34887200593948364, "learning_rate": 9.96403788817465e-05, "loss": 0.0617, "step": 8630 }, { "grad_norm": 0.33100825548171997, "learning_rate": 9.963839661716325e-05, "loss": 0.0537, "step": 8640 }, { "grad_norm": 0.3721611201763153, "learning_rate": 9.963640892419958e-05, "loss": 0.0499, "step": 8650 }, { "grad_norm": 0.36353179812431335, "learning_rate": 9.963441580307286e-05, "loss": 0.0534, "step": 8660 }, { "grad_norm": 0.5003910660743713, "learning_rate": 9.963241725400104e-05, "loss": 0.0576, "step": 8670 }, { "grad_norm": 0.3745327591896057, "learning_rate": 9.963041327720271e-05, "loss": 0.0553, "step": 8680 }, { "grad_norm": 0.3048425316810608, "learning_rate": 9.962840387289697e-05, "loss": 0.0554, "step": 8690 }, { "grad_norm": 0.4611799716949463, "learning_rate": 9.962638904130363e-05, "loss": 0.0646, "step": 8700 }, { "grad_norm": 0.5745766758918762, "learning_rate": 9.962436878264298e-05, "loss": 0.0544, "step": 8710 }, { "grad_norm": 0.41731375455856323, "learning_rate": 9.962234309713598e-05, "loss": 0.0544, "step": 8720 }, { "grad_norm": 0.42722710967063904, "learning_rate": 9.962031198500414e-05, "loss": 0.058, "step": 8730 }, { "grad_norm": 0.44477957487106323, "learning_rate": 9.961827544646958e-05, "loss": 0.0582, "step": 8740 }, { "grad_norm": 0.3199083209037781, "learning_rate": 9.961623348175501e-05, "loss": 0.0571, "step": 8750 }, { "grad_norm": 0.3668220639228821, "learning_rate": 9.961418609108377e-05, "loss": 0.0607, "step": 8760 }, { "grad_norm": 0.48998314142227173, "learning_rate": 9.961213327467971e-05, "loss": 0.0642, "step": 8770 }, { "grad_norm": 0.3209971785545349, "learning_rate": 9.961007503276736e-05, "loss": 0.0518, "step": 8780 }, { "grad_norm": 0.3662101924419403, "learning_rate": 9.960801136557179e-05, "loss": 0.0545, "step": 8790 }, { "grad_norm": 0.402913898229599, "learning_rate": 9.960594227331866e-05, "loss": 0.0608, "step": 8800 }, { "grad_norm": 0.4033035933971405, "learning_rate": 9.960386775623429e-05, "loss": 0.0584, "step": 8810 }, { "grad_norm": 0.44941335916519165, "learning_rate": 9.96017878145455e-05, "loss": 0.061, "step": 8820 }, { "grad_norm": 0.42970672249794006, "learning_rate": 9.959970244847977e-05, "loss": 0.058, "step": 8830 }, { "grad_norm": 0.4022284150123596, "learning_rate": 9.959761165826518e-05, "loss": 0.0667, "step": 8840 }, { "grad_norm": 0.37976014614105225, "learning_rate": 9.959551544413033e-05, "loss": 0.0641, "step": 8850 }, { "grad_norm": 0.35427460074424744, "learning_rate": 9.959341380630448e-05, "loss": 0.0556, "step": 8860 }, { "grad_norm": 0.37301328778266907, "learning_rate": 9.959130674501746e-05, "loss": 0.058, "step": 8870 }, { "grad_norm": 0.7056081295013428, "learning_rate": 9.958919426049968e-05, "loss": 0.0586, "step": 8880 }, { "grad_norm": 0.3445463478565216, "learning_rate": 9.958707635298219e-05, "loss": 0.06, "step": 8890 }, { "grad_norm": 0.4139650762081146, "learning_rate": 9.958495302269657e-05, "loss": 0.0692, "step": 8900 }, { "grad_norm": 0.3309954106807709, "learning_rate": 9.958282426987503e-05, "loss": 0.0526, "step": 8910 }, { "grad_norm": 0.35861602425575256, "learning_rate": 9.95806900947504e-05, "loss": 0.0609, "step": 8920 }, { "grad_norm": 0.42458030581474304, "learning_rate": 9.957855049755604e-05, "loss": 0.0539, "step": 8930 }, { "grad_norm": 0.47998547554016113, "learning_rate": 9.957640547852593e-05, "loss": 0.0537, "step": 8940 }, { "grad_norm": 0.39820173382759094, "learning_rate": 9.957425503789466e-05, "loss": 0.0643, "step": 8950 }, { "grad_norm": 0.29651302099227905, "learning_rate": 9.957209917589738e-05, "loss": 0.0516, "step": 8960 }, { "grad_norm": 0.40633124113082886, "learning_rate": 9.956993789276987e-05, "loss": 0.0575, "step": 8970 }, { "grad_norm": 0.4324292838573456, "learning_rate": 9.956777118874847e-05, "loss": 0.0599, "step": 8980 }, { "grad_norm": 0.4308457672595978, "learning_rate": 9.956559906407016e-05, "loss": 0.0642, "step": 8990 }, { "grad_norm": 0.4097048044204712, "learning_rate": 9.956342151897245e-05, "loss": 0.0513, "step": 9000 }, { "grad_norm": 0.2915578782558441, "learning_rate": 9.956123855369346e-05, "loss": 0.0486, "step": 9010 }, { "grad_norm": 0.2899593114852905, "learning_rate": 9.955905016847196e-05, "loss": 0.0524, "step": 9020 }, { "grad_norm": 0.42518728971481323, "learning_rate": 9.955685636354723e-05, "loss": 0.051, "step": 9030 }, { "grad_norm": 0.3390202820301056, "learning_rate": 9.95546571391592e-05, "loss": 0.0513, "step": 9040 }, { "grad_norm": 0.2580470144748688, "learning_rate": 9.955245249554837e-05, "loss": 0.0493, "step": 9050 }, { "grad_norm": 0.45491811633110046, "learning_rate": 9.955024243295582e-05, "loss": 0.0588, "step": 9060 }, { "grad_norm": 0.32823479175567627, "learning_rate": 9.954802695162328e-05, "loss": 0.0616, "step": 9070 }, { "grad_norm": 0.3695860505104065, "learning_rate": 9.954580605179302e-05, "loss": 0.0561, "step": 9080 }, { "grad_norm": 0.2766895890235901, "learning_rate": 9.954357973370788e-05, "loss": 0.0514, "step": 9090 }, { "grad_norm": 0.396854966878891, "learning_rate": 9.954134799761135e-05, "loss": 0.0463, "step": 9100 }, { "grad_norm": 0.44419798254966736, "learning_rate": 9.953911084374748e-05, "loss": 0.0545, "step": 9110 }, { "grad_norm": 0.31203243136405945, "learning_rate": 9.953686827236093e-05, "loss": 0.0538, "step": 9120 }, { "grad_norm": 0.45329827070236206, "learning_rate": 9.953462028369695e-05, "loss": 0.0509, "step": 9130 }, { "grad_norm": 0.37141692638397217, "learning_rate": 9.953236687800136e-05, "loss": 0.0535, "step": 9140 }, { "grad_norm": 0.3378050625324249, "learning_rate": 9.95301080555206e-05, "loss": 0.0624, "step": 9150 }, { "grad_norm": 0.36904650926589966, "learning_rate": 9.952784381650171e-05, "loss": 0.0613, "step": 9160 }, { "grad_norm": 0.43846428394317627, "learning_rate": 9.952557416119226e-05, "loss": 0.0595, "step": 9170 }, { "grad_norm": 0.3311603367328644, "learning_rate": 9.95232990898405e-05, "loss": 0.0565, "step": 9180 }, { "grad_norm": 0.47956812381744385, "learning_rate": 9.95210186026952e-05, "loss": 0.0566, "step": 9190 }, { "grad_norm": 0.3884633481502533, "learning_rate": 9.951873270000576e-05, "loss": 0.0607, "step": 9200 }, { "grad_norm": 0.31769803166389465, "learning_rate": 9.951644138202216e-05, "loss": 0.0573, "step": 9210 }, { "grad_norm": 0.4280281364917755, "learning_rate": 9.951414464899498e-05, "loss": 0.0623, "step": 9220 }, { "grad_norm": 0.32132038474082947, "learning_rate": 9.951184250117538e-05, "loss": 0.0572, "step": 9230 }, { "grad_norm": 0.34684231877326965, "learning_rate": 9.950953493881513e-05, "loss": 0.0589, "step": 9240 }, { "grad_norm": 0.3631063401699066, "learning_rate": 9.950722196216658e-05, "loss": 0.0539, "step": 9250 }, { "grad_norm": 0.3086927831172943, "learning_rate": 9.950490357148265e-05, "loss": 0.056, "step": 9260 }, { "grad_norm": 0.37107551097869873, "learning_rate": 9.950257976701692e-05, "loss": 0.0599, "step": 9270 }, { "grad_norm": 0.46221569180488586, "learning_rate": 9.950025054902348e-05, "loss": 0.0558, "step": 9280 }, { "grad_norm": 0.2720668613910675, "learning_rate": 9.949791591775706e-05, "loss": 0.0459, "step": 9290 }, { "grad_norm": 0.3391346335411072, "learning_rate": 9.949557587347298e-05, "loss": 0.0583, "step": 9300 }, { "grad_norm": 0.3757137954235077, "learning_rate": 9.949323041642713e-05, "loss": 0.0496, "step": 9310 }, { "grad_norm": 0.3690979480743408, "learning_rate": 9.949087954687602e-05, "loss": 0.0508, "step": 9320 }, { "grad_norm": 0.2980894446372986, "learning_rate": 9.948852326507672e-05, "loss": 0.0478, "step": 9330 }, { "grad_norm": 0.3264578878879547, "learning_rate": 9.948616157128694e-05, "loss": 0.0531, "step": 9340 }, { "grad_norm": 0.3430921733379364, "learning_rate": 9.948379446576493e-05, "loss": 0.0472, "step": 9350 }, { "grad_norm": 0.3101007044315338, "learning_rate": 9.948142194876952e-05, "loss": 0.045, "step": 9360 }, { "grad_norm": 0.31674620509147644, "learning_rate": 9.947904402056024e-05, "loss": 0.048, "step": 9370 }, { "grad_norm": 0.27897730469703674, "learning_rate": 9.947666068139708e-05, "loss": 0.0462, "step": 9380 }, { "grad_norm": 0.24920164048671722, "learning_rate": 9.947427193154071e-05, "loss": 0.0509, "step": 9390 }, { "grad_norm": 0.3575505018234253, "learning_rate": 9.947187777125233e-05, "loss": 0.0627, "step": 9400 }, { "grad_norm": 0.3797573745250702, "learning_rate": 9.946947820079377e-05, "loss": 0.0476, "step": 9410 }, { "grad_norm": 0.3805687427520752, "learning_rate": 9.946707322042747e-05, "loss": 0.0543, "step": 9420 }, { "grad_norm": 0.41529375314712524, "learning_rate": 9.94646628304164e-05, "loss": 0.051, "step": 9430 }, { "grad_norm": 0.2982129156589508, "learning_rate": 9.946224703102418e-05, "loss": 0.0519, "step": 9440 }, { "grad_norm": 0.372713178396225, "learning_rate": 9.945982582251498e-05, "loss": 0.048, "step": 9450 }, { "grad_norm": 0.38133248686790466, "learning_rate": 9.94573992051536e-05, "loss": 0.0497, "step": 9460 }, { "grad_norm": 0.34590232372283936, "learning_rate": 9.94549671792054e-05, "loss": 0.0566, "step": 9470 }, { "grad_norm": 0.4507558345794678, "learning_rate": 9.945252974493635e-05, "loss": 0.0529, "step": 9480 }, { "grad_norm": 0.36260002851486206, "learning_rate": 9.9450086902613e-05, "loss": 0.0464, "step": 9490 }, { "grad_norm": 0.3651496171951294, "learning_rate": 9.944763865250248e-05, "loss": 0.0528, "step": 9500 }, { "grad_norm": 0.4608170688152313, "learning_rate": 9.944518499487254e-05, "loss": 0.0532, "step": 9510 }, { "grad_norm": 0.3711414337158203, "learning_rate": 9.944272592999151e-05, "loss": 0.0545, "step": 9520 }, { "grad_norm": 0.3034346401691437, "learning_rate": 9.94402614581283e-05, "loss": 0.0582, "step": 9530 }, { "grad_norm": 0.4247921407222748, "learning_rate": 9.943779157955244e-05, "loss": 0.0542, "step": 9540 }, { "grad_norm": 0.3623117506504059, "learning_rate": 9.943531629453403e-05, "loss": 0.0525, "step": 9550 }, { "grad_norm": 0.30367812514305115, "learning_rate": 9.943283560334375e-05, "loss": 0.0492, "step": 9560 }, { "grad_norm": 0.35213640332221985, "learning_rate": 9.943034950625288e-05, "loss": 0.0549, "step": 9570 }, { "grad_norm": 0.36782535910606384, "learning_rate": 9.942785800353332e-05, "loss": 0.0623, "step": 9580 }, { "grad_norm": 0.30240073800086975, "learning_rate": 9.942536109545751e-05, "loss": 0.0608, "step": 9590 }, { "grad_norm": 0.3894580006599426, "learning_rate": 9.942285878229853e-05, "loss": 0.0542, "step": 9600 }, { "grad_norm": 0.37638089060783386, "learning_rate": 9.942035106433001e-05, "loss": 0.0529, "step": 9610 }, { "grad_norm": 0.3524564206600189, "learning_rate": 9.94178379418262e-05, "loss": 0.0527, "step": 9620 }, { "grad_norm": 0.32617297768592834, "learning_rate": 9.941531941506194e-05, "loss": 0.0527, "step": 9630 }, { "grad_norm": 0.42130419611930847, "learning_rate": 9.941279548431263e-05, "loss": 0.0534, "step": 9640 }, { "grad_norm": 0.45266154408454895, "learning_rate": 9.941026614985431e-05, "loss": 0.0551, "step": 9650 }, { "grad_norm": 0.3494419753551483, "learning_rate": 9.940773141196357e-05, "loss": 0.0546, "step": 9660 }, { "grad_norm": 0.39885374903678894, "learning_rate": 9.94051912709176e-05, "loss": 0.0545, "step": 9670 }, { "grad_norm": 0.39373543858528137, "learning_rate": 9.940264572699421e-05, "loss": 0.0568, "step": 9680 }, { "grad_norm": 0.3519563674926758, "learning_rate": 9.940009478047174e-05, "loss": 0.0563, "step": 9690 }, { "grad_norm": 0.29237744212150574, "learning_rate": 9.939753843162918e-05, "loss": 0.0506, "step": 9700 }, { "grad_norm": 0.39676961302757263, "learning_rate": 9.939497668074609e-05, "loss": 0.0493, "step": 9710 }, { "grad_norm": 0.3654899001121521, "learning_rate": 9.93924095281026e-05, "loss": 0.0541, "step": 9720 }, { "grad_norm": 0.424706369638443, "learning_rate": 9.938983697397948e-05, "loss": 0.0608, "step": 9730 }, { "grad_norm": 0.45638301968574524, "learning_rate": 9.938725901865805e-05, "loss": 0.055, "step": 9740 }, { "grad_norm": 0.33807283639907837, "learning_rate": 9.93846756624202e-05, "loss": 0.0571, "step": 9750 }, { "grad_norm": 0.35274502635002136, "learning_rate": 9.938208690554849e-05, "loss": 0.0564, "step": 9760 }, { "grad_norm": 0.37883618474006653, "learning_rate": 9.9379492748326e-05, "loss": 0.0612, "step": 9770 }, { "grad_norm": 0.35023030638694763, "learning_rate": 9.937689319103641e-05, "loss": 0.0474, "step": 9780 }, { "grad_norm": 0.34377992153167725, "learning_rate": 9.937428823396404e-05, "loss": 0.0506, "step": 9790 }, { "grad_norm": 0.41970568895339966, "learning_rate": 9.937167787739372e-05, "loss": 0.0529, "step": 9800 }, { "grad_norm": 0.35720938444137573, "learning_rate": 9.936906212161095e-05, "loss": 0.0479, "step": 9810 }, { "grad_norm": 0.4446749687194824, "learning_rate": 9.936644096690176e-05, "loss": 0.0533, "step": 9820 }, { "grad_norm": 0.5719116926193237, "learning_rate": 9.936381441355282e-05, "loss": 0.0601, "step": 9830 }, { "grad_norm": 0.4850420653820038, "learning_rate": 9.936118246185136e-05, "loss": 0.0574, "step": 9840 }, { "grad_norm": 0.39803582429885864, "learning_rate": 9.935854511208518e-05, "loss": 0.0681, "step": 9850 }, { "grad_norm": 0.40126684308052063, "learning_rate": 9.935590236454272e-05, "loss": 0.0629, "step": 9860 }, { "grad_norm": 0.3018656373023987, "learning_rate": 9.935325421951298e-05, "loss": 0.0499, "step": 9870 }, { "grad_norm": 0.2811392545700073, "learning_rate": 9.935060067728557e-05, "loss": 0.0561, "step": 9880 }, { "grad_norm": 0.3654262125492096, "learning_rate": 9.934794173815067e-05, "loss": 0.0619, "step": 9890 }, { "grad_norm": 0.5153788924217224, "learning_rate": 9.934527740239906e-05, "loss": 0.0548, "step": 9900 }, { "grad_norm": 0.32451319694519043, "learning_rate": 9.934260767032209e-05, "loss": 0.0664, "step": 9910 }, { "grad_norm": 0.35001176595687866, "learning_rate": 9.933993254221172e-05, "loss": 0.0574, "step": 9920 }, { "grad_norm": 0.344211220741272, "learning_rate": 9.933725201836053e-05, "loss": 0.0543, "step": 9930 }, { "grad_norm": 0.3388674855232239, "learning_rate": 9.933456609906162e-05, "loss": 0.049, "step": 9940 }, { "grad_norm": 0.41353338956832886, "learning_rate": 9.933187478460875e-05, "loss": 0.0688, "step": 9950 }, { "grad_norm": 0.32781174778938293, "learning_rate": 9.93291780752962e-05, "loss": 0.0623, "step": 9960 }, { "grad_norm": 0.4088277518749237, "learning_rate": 9.932647597141893e-05, "loss": 0.0553, "step": 9970 }, { "grad_norm": 0.38477402925491333, "learning_rate": 9.932376847327239e-05, "loss": 0.0607, "step": 9980 }, { "grad_norm": 0.39728471636772156, "learning_rate": 9.932105558115268e-05, "loss": 0.0541, "step": 9990 }, { "grad_norm": 0.3069525957107544, "learning_rate": 9.931833729535651e-05, "loss": 0.0547, "step": 10000 }, { "grad_norm": 0.42256078124046326, "learning_rate": 9.931561361618111e-05, "loss": 0.055, "step": 10010 }, { "grad_norm": 0.33870309591293335, "learning_rate": 9.931288454392435e-05, "loss": 0.0535, "step": 10020 }, { "grad_norm": 0.2839166224002838, "learning_rate": 9.931015007888467e-05, "loss": 0.0484, "step": 10030 }, { "grad_norm": 0.3545072078704834, "learning_rate": 9.930741022136112e-05, "loss": 0.0595, "step": 10040 }, { "grad_norm": 0.28668826818466187, "learning_rate": 9.930466497165333e-05, "loss": 0.0554, "step": 10050 }, { "grad_norm": 0.34838226437568665, "learning_rate": 9.93019143300615e-05, "loss": 0.0498, "step": 10060 }, { "grad_norm": 0.3840208649635315, "learning_rate": 9.929915829688644e-05, "loss": 0.0466, "step": 10070 }, { "grad_norm": 0.4134078919887543, "learning_rate": 9.929639687242955e-05, "loss": 0.0462, "step": 10080 }, { "grad_norm": 0.29244905710220337, "learning_rate": 9.929363005699281e-05, "loss": 0.0463, "step": 10090 }, { "grad_norm": 0.31961244344711304, "learning_rate": 9.92908578508788e-05, "loss": 0.0556, "step": 10100 }, { "grad_norm": 0.42928436398506165, "learning_rate": 9.928808025439069e-05, "loss": 0.0608, "step": 10110 }, { "grad_norm": 0.34262439608573914, "learning_rate": 9.928529726783223e-05, "loss": 0.0534, "step": 10120 }, { "grad_norm": 0.3332134187221527, "learning_rate": 9.928250889150774e-05, "loss": 0.0567, "step": 10130 }, { "grad_norm": 0.3317621648311615, "learning_rate": 9.92797151257222e-05, "loss": 0.0502, "step": 10140 }, { "grad_norm": 0.402322381734848, "learning_rate": 9.927691597078108e-05, "loss": 0.0499, "step": 10150 }, { "grad_norm": 0.287572979927063, "learning_rate": 9.927411142699053e-05, "loss": 0.0474, "step": 10160 }, { "grad_norm": 0.3948141038417816, "learning_rate": 9.927130149465725e-05, "loss": 0.0537, "step": 10170 }, { "grad_norm": 0.3714626133441925, "learning_rate": 9.92684861740885e-05, "loss": 0.0528, "step": 10180 }, { "grad_norm": 0.33381161093711853, "learning_rate": 9.926566546559217e-05, "loss": 0.0505, "step": 10190 }, { "grad_norm": 0.4081650972366333, "learning_rate": 9.926283936947673e-05, "loss": 0.0507, "step": 10200 }, { "grad_norm": 0.41019636392593384, "learning_rate": 9.926000788605126e-05, "loss": 0.0538, "step": 10210 }, { "grad_norm": 0.4367842972278595, "learning_rate": 9.92571710156254e-05, "loss": 0.0525, "step": 10220 }, { "grad_norm": 0.29898446798324585, "learning_rate": 9.925432875850936e-05, "loss": 0.0502, "step": 10230 }, { "grad_norm": 0.38441476225852966, "learning_rate": 9.925148111501396e-05, "loss": 0.054, "step": 10240 }, { "grad_norm": 0.39198189973831177, "learning_rate": 9.924862808545066e-05, "loss": 0.0524, "step": 10250 }, { "grad_norm": 0.4317256212234497, "learning_rate": 9.924576967013141e-05, "loss": 0.0524, "step": 10260 }, { "grad_norm": 0.3898879885673523, "learning_rate": 9.924290586936887e-05, "loss": 0.0568, "step": 10270 }, { "grad_norm": 0.4188746213912964, "learning_rate": 9.924003668347614e-05, "loss": 0.0526, "step": 10280 }, { "grad_norm": 0.39831385016441345, "learning_rate": 9.923716211276704e-05, "loss": 0.05, "step": 10290 }, { "grad_norm": 0.3137796223163605, "learning_rate": 9.923428215755594e-05, "loss": 0.0535, "step": 10300 }, { "grad_norm": 0.3392252027988434, "learning_rate": 9.923139681815775e-05, "loss": 0.0517, "step": 10310 }, { "grad_norm": 0.46200159192085266, "learning_rate": 9.922850609488801e-05, "loss": 0.0492, "step": 10320 }, { "grad_norm": 0.4089395999908447, "learning_rate": 9.922560998806287e-05, "loss": 0.0534, "step": 10330 }, { "grad_norm": 0.3937968909740448, "learning_rate": 9.922270849799905e-05, "loss": 0.0516, "step": 10340 }, { "grad_norm": 0.4058781862258911, "learning_rate": 9.92198016250138e-05, "loss": 0.0604, "step": 10350 }, { "grad_norm": 0.384420782327652, "learning_rate": 9.921688936942506e-05, "loss": 0.0493, "step": 10360 }, { "grad_norm": 0.3488917350769043, "learning_rate": 9.921397173155129e-05, "loss": 0.0539, "step": 10370 }, { "grad_norm": 0.339858740568161, "learning_rate": 9.921104871171157e-05, "loss": 0.0481, "step": 10380 }, { "grad_norm": 0.3462877869606018, "learning_rate": 9.920812031022554e-05, "loss": 0.0428, "step": 10390 }, { "grad_norm": 0.323209285736084, "learning_rate": 9.920518652741348e-05, "loss": 0.0536, "step": 10400 }, { "grad_norm": 0.31863993406295776, "learning_rate": 9.920224736359618e-05, "loss": 0.0491, "step": 10410 }, { "grad_norm": 0.3517577350139618, "learning_rate": 9.91993028190951e-05, "loss": 0.0596, "step": 10420 }, { "grad_norm": 0.3281869888305664, "learning_rate": 9.919635289423222e-05, "loss": 0.0502, "step": 10430 }, { "grad_norm": 0.29334044456481934, "learning_rate": 9.919339758933015e-05, "loss": 0.0591, "step": 10440 }, { "grad_norm": 0.31704092025756836, "learning_rate": 9.919043690471209e-05, "loss": 0.0642, "step": 10450 }, { "grad_norm": 0.5896210670471191, "learning_rate": 9.91874708407018e-05, "loss": 0.0672, "step": 10460 }, { "grad_norm": 0.37488022446632385, "learning_rate": 9.918449939762367e-05, "loss": 0.0613, "step": 10470 }, { "grad_norm": 0.34268510341644287, "learning_rate": 9.91815225758026e-05, "loss": 0.0574, "step": 10480 }, { "grad_norm": 0.3918323218822479, "learning_rate": 9.917854037556419e-05, "loss": 0.0555, "step": 10490 }, { "grad_norm": 0.3346719741821289, "learning_rate": 9.917555279723454e-05, "loss": 0.0599, "step": 10500 }, { "grad_norm": 0.4215778112411499, "learning_rate": 9.917255984114036e-05, "loss": 0.0569, "step": 10510 }, { "grad_norm": 0.31806597113609314, "learning_rate": 9.916956150760896e-05, "loss": 0.0484, "step": 10520 }, { "grad_norm": 0.4018816351890564, "learning_rate": 9.916655779696826e-05, "loss": 0.0502, "step": 10530 }, { "grad_norm": 0.3285824656486511, "learning_rate": 9.916354870954671e-05, "loss": 0.0536, "step": 10540 }, { "grad_norm": 0.45901039242744446, "learning_rate": 9.91605342456734e-05, "loss": 0.0555, "step": 10550 }, { "grad_norm": 0.3990185856819153, "learning_rate": 9.915751440567795e-05, "loss": 0.0504, "step": 10560 }, { "grad_norm": 0.28498712182044983, "learning_rate": 9.915448918989066e-05, "loss": 0.0518, "step": 10570 }, { "grad_norm": 0.41969195008277893, "learning_rate": 9.915145859864232e-05, "loss": 0.0584, "step": 10580 }, { "grad_norm": 0.3174610733985901, "learning_rate": 9.914842263226437e-05, "loss": 0.0517, "step": 10590 }, { "grad_norm": 0.3294687271118164, "learning_rate": 9.914538129108882e-05, "loss": 0.0558, "step": 10600 }, { "grad_norm": 0.4225710332393646, "learning_rate": 9.914233457544825e-05, "loss": 0.063, "step": 10610 }, { "grad_norm": 0.3346511125564575, "learning_rate": 9.913928248567586e-05, "loss": 0.0574, "step": 10620 }, { "grad_norm": 0.3584263026714325, "learning_rate": 9.913622502210542e-05, "loss": 0.0568, "step": 10630 }, { "grad_norm": 0.4375753104686737, "learning_rate": 9.913316218507128e-05, "loss": 0.0622, "step": 10640 }, { "grad_norm": 0.36108195781707764, "learning_rate": 9.91300939749084e-05, "loss": 0.0541, "step": 10650 }, { "grad_norm": 0.31741929054260254, "learning_rate": 9.91270203919523e-05, "loss": 0.0558, "step": 10660 }, { "grad_norm": 0.3770129680633545, "learning_rate": 9.912394143653912e-05, "loss": 0.0568, "step": 10670 }, { "grad_norm": 0.42066699266433716, "learning_rate": 9.912085710900555e-05, "loss": 0.0555, "step": 10680 }, { "grad_norm": 0.3967154324054718, "learning_rate": 9.911776740968892e-05, "loss": 0.0543, "step": 10690 }, { "grad_norm": 0.27790766954421997, "learning_rate": 9.911467233892709e-05, "loss": 0.0492, "step": 10700 }, { "grad_norm": 0.3822956085205078, "learning_rate": 9.911157189705853e-05, "loss": 0.0502, "step": 10710 }, { "grad_norm": 0.33425039052963257, "learning_rate": 9.910846608442229e-05, "loss": 0.0469, "step": 10720 }, { "grad_norm": 0.321661114692688, "learning_rate": 9.910535490135805e-05, "loss": 0.0472, "step": 10730 }, { "grad_norm": 0.2592156231403351, "learning_rate": 9.910223834820603e-05, "loss": 0.0513, "step": 10740 }, { "grad_norm": 0.3242875039577484, "learning_rate": 9.909911642530703e-05, "loss": 0.0527, "step": 10750 }, { "grad_norm": 0.2828913927078247, "learning_rate": 9.909598913300249e-05, "loss": 0.0512, "step": 10760 }, { "grad_norm": 0.32557427883148193, "learning_rate": 9.909285647163438e-05, "loss": 0.0515, "step": 10770 }, { "grad_norm": 0.29603564739227295, "learning_rate": 9.908971844154531e-05, "loss": 0.0464, "step": 10780 }, { "grad_norm": 0.3461776673793793, "learning_rate": 9.908657504307843e-05, "loss": 0.0441, "step": 10790 }, { "grad_norm": 0.39344125986099243, "learning_rate": 9.908342627657751e-05, "loss": 0.0473, "step": 10800 }, { "grad_norm": 0.3830757737159729, "learning_rate": 9.908027214238689e-05, "loss": 0.0525, "step": 10810 }, { "grad_norm": 0.3400309383869171, "learning_rate": 9.90771126408515e-05, "loss": 0.0559, "step": 10820 }, { "grad_norm": 0.49821731448173523, "learning_rate": 9.907394777231685e-05, "loss": 0.0528, "step": 10830 }, { "grad_norm": 0.29055750370025635, "learning_rate": 9.907077753712905e-05, "loss": 0.0513, "step": 10840 }, { "grad_norm": 0.3964129686355591, "learning_rate": 9.906760193563482e-05, "loss": 0.0606, "step": 10850 }, { "grad_norm": 0.4092269837856293, "learning_rate": 9.906442096818139e-05, "loss": 0.0582, "step": 10860 }, { "grad_norm": 0.36253947019577026, "learning_rate": 9.906123463511665e-05, "loss": 0.0537, "step": 10870 }, { "grad_norm": 0.27545303106307983, "learning_rate": 9.905804293678907e-05, "loss": 0.0553, "step": 10880 }, { "grad_norm": 0.3584348261356354, "learning_rate": 9.905484587354766e-05, "loss": 0.0543, "step": 10890 }, { "grad_norm": 0.3225915729999542, "learning_rate": 9.905164344574205e-05, "loss": 0.0511, "step": 10900 }, { "grad_norm": 0.49193301796913147, "learning_rate": 9.904843565372248e-05, "loss": 0.0556, "step": 10910 }, { "grad_norm": 0.3745037019252777, "learning_rate": 9.904522249783972e-05, "loss": 0.0515, "step": 10920 }, { "grad_norm": 0.4350736439228058, "learning_rate": 9.904200397844517e-05, "loss": 0.0518, "step": 10930 }, { "grad_norm": 0.5238029956817627, "learning_rate": 9.903878009589078e-05, "loss": 0.0553, "step": 10940 }, { "grad_norm": 0.26271966099739075, "learning_rate": 9.903555085052915e-05, "loss": 0.0536, "step": 10950 }, { "grad_norm": 0.3954243063926697, "learning_rate": 9.903231624271338e-05, "loss": 0.0548, "step": 10960 }, { "grad_norm": 0.5309438109397888, "learning_rate": 9.902907627279724e-05, "loss": 0.056, "step": 10970 }, { "grad_norm": 0.3385268449783325, "learning_rate": 9.902583094113504e-05, "loss": 0.0513, "step": 10980 }, { "grad_norm": 0.3240995407104492, "learning_rate": 9.902258024808168e-05, "loss": 0.0516, "step": 10990 }, { "grad_norm": 0.2665518820285797, "learning_rate": 9.901932419399264e-05, "loss": 0.0505, "step": 11000 }, { "grad_norm": 0.40761831402778625, "learning_rate": 9.9016062779224e-05, "loss": 0.0496, "step": 11010 }, { "grad_norm": 0.3047691285610199, "learning_rate": 9.901279600413242e-05, "loss": 0.0551, "step": 11020 }, { "grad_norm": 0.2635865807533264, "learning_rate": 9.900952386907518e-05, "loss": 0.0538, "step": 11030 }, { "grad_norm": 0.2768528461456299, "learning_rate": 9.90062463744101e-05, "loss": 0.0547, "step": 11040 }, { "grad_norm": 0.4283389151096344, "learning_rate": 9.900296352049558e-05, "loss": 0.0543, "step": 11050 }, { "grad_norm": 0.3576904237270355, "learning_rate": 9.899967530769065e-05, "loss": 0.0565, "step": 11060 }, { "grad_norm": 0.31309422850608826, "learning_rate": 9.899638173635489e-05, "loss": 0.0481, "step": 11070 }, { "grad_norm": 0.3804836571216583, "learning_rate": 9.899308280684849e-05, "loss": 0.0596, "step": 11080 }, { "grad_norm": 0.45397090911865234, "learning_rate": 9.898977851953222e-05, "loss": 0.0603, "step": 11090 }, { "grad_norm": 0.26669085025787354, "learning_rate": 9.898646887476741e-05, "loss": 0.0501, "step": 11100 }, { "grad_norm": 0.4244576692581177, "learning_rate": 9.898315387291603e-05, "loss": 0.0542, "step": 11110 }, { "grad_norm": 0.32640600204467773, "learning_rate": 9.89798335143406e-05, "loss": 0.0516, "step": 11120 }, { "grad_norm": 0.42199328541755676, "learning_rate": 9.897650779940419e-05, "loss": 0.0597, "step": 11130 }, { "grad_norm": 0.343380868434906, "learning_rate": 9.897317672847054e-05, "loss": 0.0507, "step": 11140 }, { "grad_norm": 0.4310145676136017, "learning_rate": 9.89698403019039e-05, "loss": 0.0511, "step": 11150 }, { "grad_norm": 0.42337656021118164, "learning_rate": 9.896649852006917e-05, "loss": 0.051, "step": 11160 }, { "grad_norm": 0.41524869203567505, "learning_rate": 9.896315138333177e-05, "loss": 0.0555, "step": 11170 }, { "grad_norm": 0.3942844867706299, "learning_rate": 9.895979889205774e-05, "loss": 0.0496, "step": 11180 }, { "grad_norm": 0.3810170888900757, "learning_rate": 9.895644104661372e-05, "loss": 0.0583, "step": 11190 }, { "grad_norm": 0.3267385959625244, "learning_rate": 9.895307784736691e-05, "loss": 0.0448, "step": 11200 }, { "grad_norm": 0.3601128160953522, "learning_rate": 9.894970929468512e-05, "loss": 0.0499, "step": 11210 }, { "grad_norm": 0.29591405391693115, "learning_rate": 9.89463353889367e-05, "loss": 0.049, "step": 11220 }, { "grad_norm": 0.29303812980651855, "learning_rate": 9.894295613049065e-05, "loss": 0.0485, "step": 11230 }, { "grad_norm": 0.3124397099018097, "learning_rate": 9.893957151971649e-05, "loss": 0.0453, "step": 11240 }, { "grad_norm": 0.35347720980644226, "learning_rate": 9.893618155698436e-05, "loss": 0.0452, "step": 11250 }, { "grad_norm": 0.32865506410598755, "learning_rate": 9.8932786242665e-05, "loss": 0.0562, "step": 11260 }, { "grad_norm": 0.29997947812080383, "learning_rate": 9.89293855771297e-05, "loss": 0.0453, "step": 11270 }, { "grad_norm": 0.28656551241874695, "learning_rate": 9.892597956075036e-05, "loss": 0.0437, "step": 11280 }, { "grad_norm": 0.2580645978450775, "learning_rate": 9.892256819389947e-05, "loss": 0.043, "step": 11290 }, { "grad_norm": 0.3357199728488922, "learning_rate": 9.891915147695006e-05, "loss": 0.0543, "step": 11300 }, { "grad_norm": 0.4648645222187042, "learning_rate": 9.891572941027577e-05, "loss": 0.0597, "step": 11310 }, { "grad_norm": 0.2692905366420746, "learning_rate": 9.89123019942509e-05, "loss": 0.0571, "step": 11320 }, { "grad_norm": 0.3702443242073059, "learning_rate": 9.89088692292502e-05, "loss": 0.0479, "step": 11330 }, { "grad_norm": 0.3757365047931671, "learning_rate": 9.89054311156491e-05, "loss": 0.0557, "step": 11340 }, { "grad_norm": 0.33421245217323303, "learning_rate": 9.890198765382357e-05, "loss": 0.0538, "step": 11350 }, { "grad_norm": 0.30847084522247314, "learning_rate": 9.889853884415021e-05, "loss": 0.0494, "step": 11360 }, { "grad_norm": 0.3582276701927185, "learning_rate": 9.889508468700614e-05, "loss": 0.0493, "step": 11370 }, { "grad_norm": 0.32504770159721375, "learning_rate": 9.889162518276915e-05, "loss": 0.0461, "step": 11380 }, { "grad_norm": 0.3548200726509094, "learning_rate": 9.888816033181752e-05, "loss": 0.0479, "step": 11390 }, { "grad_norm": 0.3149525225162506, "learning_rate": 9.888469013453018e-05, "loss": 0.0492, "step": 11400 }, { "grad_norm": 0.3581077456474304, "learning_rate": 9.888121459128663e-05, "loss": 0.0502, "step": 11410 }, { "grad_norm": 0.4090034067630768, "learning_rate": 9.887773370246693e-05, "loss": 0.0514, "step": 11420 }, { "grad_norm": 0.3394065797328949, "learning_rate": 9.887424746845177e-05, "loss": 0.055, "step": 11430 }, { "grad_norm": 0.38740721344947815, "learning_rate": 9.887075588962239e-05, "loss": 0.0485, "step": 11440 }, { "grad_norm": 0.34173962473869324, "learning_rate": 9.88672589663606e-05, "loss": 0.0551, "step": 11450 }, { "grad_norm": 0.3246396481990814, "learning_rate": 9.886375669904886e-05, "loss": 0.0554, "step": 11460 }, { "grad_norm": 0.4355646073818207, "learning_rate": 9.886024908807014e-05, "loss": 0.0489, "step": 11470 }, { "grad_norm": 0.4097084701061249, "learning_rate": 9.885673613380806e-05, "loss": 0.053, "step": 11480 }, { "grad_norm": 0.3686857521533966, "learning_rate": 9.885321783664676e-05, "loss": 0.0541, "step": 11490 }, { "grad_norm": 0.2808628976345062, "learning_rate": 9.884969419697101e-05, "loss": 0.0504, "step": 11500 }, { "grad_norm": 0.34892919659614563, "learning_rate": 9.884616521516614e-05, "loss": 0.0521, "step": 11510 }, { "grad_norm": 0.2878177762031555, "learning_rate": 9.88426308916181e-05, "loss": 0.0412, "step": 11520 }, { "grad_norm": 0.29541677236557007, "learning_rate": 9.883909122671335e-05, "loss": 0.05, "step": 11530 }, { "grad_norm": 0.30991464853286743, "learning_rate": 9.883554622083904e-05, "loss": 0.0562, "step": 11540 }, { "grad_norm": 0.2695753872394562, "learning_rate": 9.88319958743828e-05, "loss": 0.0484, "step": 11550 }, { "grad_norm": 0.4385068714618683, "learning_rate": 9.882844018773291e-05, "loss": 0.0502, "step": 11560 }, { "grad_norm": 0.3004198372364044, "learning_rate": 9.882487916127823e-05, "loss": 0.0498, "step": 11570 }, { "grad_norm": 0.3392147719860077, "learning_rate": 9.882131279540815e-05, "loss": 0.053, "step": 11580 }, { "grad_norm": 0.22789599001407623, "learning_rate": 9.881774109051271e-05, "loss": 0.0485, "step": 11590 }, { "grad_norm": 0.2675340175628662, "learning_rate": 9.881416404698252e-05, "loss": 0.0487, "step": 11600 }, { "grad_norm": 0.3658326268196106, "learning_rate": 9.881058166520873e-05, "loss": 0.0452, "step": 11610 }, { "grad_norm": 0.30279552936553955, "learning_rate": 9.880699394558311e-05, "loss": 0.051, "step": 11620 }, { "grad_norm": 0.318928986787796, "learning_rate": 9.880340088849801e-05, "loss": 0.0525, "step": 11630 }, { "grad_norm": 0.29645901918411255, "learning_rate": 9.879980249434637e-05, "loss": 0.0455, "step": 11640 }, { "grad_norm": 0.3856724798679352, "learning_rate": 9.879619876352168e-05, "loss": 0.0475, "step": 11650 }, { "grad_norm": 0.4027184247970581, "learning_rate": 9.879258969641809e-05, "loss": 0.0567, "step": 11660 }, { "grad_norm": 0.41266706585884094, "learning_rate": 9.878897529343023e-05, "loss": 0.0569, "step": 11670 }, { "grad_norm": 0.2920987606048584, "learning_rate": 9.878535555495338e-05, "loss": 0.0504, "step": 11680 }, { "grad_norm": 0.3304232060909271, "learning_rate": 9.87817304813834e-05, "loss": 0.0505, "step": 11690 }, { "grad_norm": 0.38294070959091187, "learning_rate": 9.877810007311671e-05, "loss": 0.0563, "step": 11700 }, { "grad_norm": 0.31526196002960205, "learning_rate": 9.877446433055035e-05, "loss": 0.0503, "step": 11710 }, { "grad_norm": 0.3123669922351837, "learning_rate": 9.877082325408191e-05, "loss": 0.0519, "step": 11720 }, { "grad_norm": 0.3132535219192505, "learning_rate": 9.876717684410954e-05, "loss": 0.0456, "step": 11730 }, { "grad_norm": 0.34437328577041626, "learning_rate": 9.876352510103204e-05, "loss": 0.0564, "step": 11740 }, { "grad_norm": 0.2725062370300293, "learning_rate": 9.875986802524875e-05, "loss": 0.0468, "step": 11750 }, { "grad_norm": 0.33692172169685364, "learning_rate": 9.87562056171596e-05, "loss": 0.0533, "step": 11760 }, { "grad_norm": 0.365983247756958, "learning_rate": 9.875253787716511e-05, "loss": 0.0484, "step": 11770 }, { "grad_norm": 0.2905980050563812, "learning_rate": 9.874886480566637e-05, "loss": 0.0424, "step": 11780 }, { "grad_norm": 0.3177790343761444, "learning_rate": 9.874518640306507e-05, "loss": 0.052, "step": 11790 }, { "grad_norm": 0.2830192446708679, "learning_rate": 9.874150266976347e-05, "loss": 0.0556, "step": 11800 }, { "grad_norm": 0.23548652231693268, "learning_rate": 9.873781360616443e-05, "loss": 0.0506, "step": 11810 }, { "grad_norm": 0.3352868854999542, "learning_rate": 9.873411921267137e-05, "loss": 0.0467, "step": 11820 }, { "grad_norm": 0.2596540153026581, "learning_rate": 9.873041948968829e-05, "loss": 0.0401, "step": 11830 }, { "grad_norm": 0.3488447964191437, "learning_rate": 9.872671443761981e-05, "loss": 0.0456, "step": 11840 }, { "grad_norm": 0.43124890327453613, "learning_rate": 9.872300405687109e-05, "loss": 0.0469, "step": 11850 }, { "grad_norm": 0.32871943712234497, "learning_rate": 9.871928834784792e-05, "loss": 0.0466, "step": 11860 }, { "grad_norm": 0.4770122468471527, "learning_rate": 9.871556731095661e-05, "loss": 0.0552, "step": 11870 }, { "grad_norm": 0.31879928708076477, "learning_rate": 9.871184094660411e-05, "loss": 0.045, "step": 11880 }, { "grad_norm": 0.35866954922676086, "learning_rate": 9.870810925519791e-05, "loss": 0.0519, "step": 11890 }, { "grad_norm": 0.3539714515209198, "learning_rate": 9.870437223714612e-05, "loss": 0.0448, "step": 11900 }, { "grad_norm": 0.30755773186683655, "learning_rate": 9.87006298928574e-05, "loss": 0.047, "step": 11910 }, { "grad_norm": 0.3382813632488251, "learning_rate": 9.869688222274103e-05, "loss": 0.0478, "step": 11920 }, { "grad_norm": 0.4123079776763916, "learning_rate": 9.869312922720681e-05, "loss": 0.0487, "step": 11930 }, { "grad_norm": 0.3927604556083679, "learning_rate": 9.868937090666521e-05, "loss": 0.0498, "step": 11940 }, { "grad_norm": 0.3630758225917816, "learning_rate": 9.86856072615272e-05, "loss": 0.0479, "step": 11950 }, { "grad_norm": 0.39424312114715576, "learning_rate": 9.868183829220438e-05, "loss": 0.0477, "step": 11960 }, { "grad_norm": 0.35363683104515076, "learning_rate": 9.867806399910893e-05, "loss": 0.0507, "step": 11970 }, { "grad_norm": 0.37297189235687256, "learning_rate": 9.867428438265356e-05, "loss": 0.0497, "step": 11980 }, { "grad_norm": 0.320522278547287, "learning_rate": 9.867049944325165e-05, "loss": 0.0528, "step": 11990 }, { "grad_norm": 0.3523763120174408, "learning_rate": 9.86667091813171e-05, "loss": 0.0527, "step": 12000 }, { "grad_norm": 0.323015958070755, "learning_rate": 9.866291359726438e-05, "loss": 0.0494, "step": 12010 }, { "grad_norm": 0.37342119216918945, "learning_rate": 9.865911269150861e-05, "loss": 0.0467, "step": 12020 }, { "grad_norm": 0.2723780572414398, "learning_rate": 9.865530646446544e-05, "loss": 0.0464, "step": 12030 }, { "grad_norm": 0.36099398136138916, "learning_rate": 9.86514949165511e-05, "loss": 0.0463, "step": 12040 }, { "grad_norm": 0.3872695863246918, "learning_rate": 9.864767804818243e-05, "loss": 0.0454, "step": 12050 }, { "grad_norm": 0.42592528462409973, "learning_rate": 9.86438558597768e-05, "loss": 0.0488, "step": 12060 }, { "grad_norm": 0.34480467438697815, "learning_rate": 9.864002835175225e-05, "loss": 0.0517, "step": 12070 }, { "grad_norm": 0.3774372339248657, "learning_rate": 9.863619552452734e-05, "loss": 0.0518, "step": 12080 }, { "grad_norm": 0.3269802927970886, "learning_rate": 9.863235737852119e-05, "loss": 0.0478, "step": 12090 }, { "grad_norm": 0.3150104582309723, "learning_rate": 9.862851391415356e-05, "loss": 0.0489, "step": 12100 }, { "grad_norm": 0.306678831577301, "learning_rate": 9.862466513184477e-05, "loss": 0.0489, "step": 12110 }, { "grad_norm": 0.28279489278793335, "learning_rate": 9.86208110320157e-05, "loss": 0.0504, "step": 12120 }, { "grad_norm": 0.2899830937385559, "learning_rate": 9.861695161508784e-05, "loss": 0.0499, "step": 12130 }, { "grad_norm": 0.31508660316467285, "learning_rate": 9.861308688148324e-05, "loss": 0.0513, "step": 12140 }, { "grad_norm": 0.35474297404289246, "learning_rate": 9.860921683162455e-05, "loss": 0.053, "step": 12150 }, { "grad_norm": 0.36175337433815, "learning_rate": 9.860534146593499e-05, "loss": 0.0505, "step": 12160 }, { "grad_norm": 0.2900707721710205, "learning_rate": 9.860146078483836e-05, "loss": 0.0418, "step": 12170 }, { "grad_norm": 0.3718706965446472, "learning_rate": 9.859757478875905e-05, "loss": 0.0468, "step": 12180 }, { "grad_norm": 0.37820833921432495, "learning_rate": 9.859368347812204e-05, "loss": 0.0458, "step": 12190 }, { "grad_norm": 0.26614072918891907, "learning_rate": 9.858978685335285e-05, "loss": 0.0511, "step": 12200 }, { "grad_norm": 0.29652345180511475, "learning_rate": 9.858588491487763e-05, "loss": 0.0474, "step": 12210 }, { "grad_norm": 0.3641464114189148, "learning_rate": 9.858197766312308e-05, "loss": 0.0509, "step": 12220 }, { "grad_norm": 0.3611646890640259, "learning_rate": 9.857806509851649e-05, "loss": 0.0605, "step": 12230 }, { "grad_norm": 0.31926247477531433, "learning_rate": 9.857414722148574e-05, "loss": 0.0505, "step": 12240 }, { "grad_norm": 0.33594968914985657, "learning_rate": 9.857022403245928e-05, "loss": 0.0473, "step": 12250 }, { "grad_norm": 0.3573843836784363, "learning_rate": 9.856629553186615e-05, "loss": 0.0491, "step": 12260 }, { "grad_norm": 0.3607424795627594, "learning_rate": 9.856236172013595e-05, "loss": 0.0501, "step": 12270 }, { "grad_norm": 0.40187957882881165, "learning_rate": 9.85584225976989e-05, "loss": 0.0477, "step": 12280 }, { "grad_norm": 0.35601136088371277, "learning_rate": 9.855447816498575e-05, "loss": 0.0546, "step": 12290 }, { "grad_norm": 0.2891084551811218, "learning_rate": 9.855052842242787e-05, "loss": 0.053, "step": 12300 }, { "grad_norm": 0.33519771695137024, "learning_rate": 9.85465733704572e-05, "loss": 0.0487, "step": 12310 }, { "grad_norm": 0.2955290377140045, "learning_rate": 9.854261300950624e-05, "loss": 0.0447, "step": 12320 }, { "grad_norm": 0.28517839312553406, "learning_rate": 9.853864734000813e-05, "loss": 0.0551, "step": 12330 }, { "grad_norm": 0.3367580771446228, "learning_rate": 9.85346763623965e-05, "loss": 0.0457, "step": 12340 }, { "grad_norm": 0.28306102752685547, "learning_rate": 9.853070007710564e-05, "loss": 0.0432, "step": 12350 }, { "grad_norm": 0.32256922125816345, "learning_rate": 9.85267184845704e-05, "loss": 0.0451, "step": 12360 }, { "grad_norm": 0.2553161680698395, "learning_rate": 9.852273158522616e-05, "loss": 0.048, "step": 12370 }, { "grad_norm": 0.2751815617084503, "learning_rate": 9.851873937950896e-05, "loss": 0.0439, "step": 12380 }, { "grad_norm": 0.2865961492061615, "learning_rate": 9.851474186785537e-05, "loss": 0.0419, "step": 12390 }, { "grad_norm": 0.32302024960517883, "learning_rate": 9.851073905070254e-05, "loss": 0.0486, "step": 12400 }, { "grad_norm": 0.30479001998901367, "learning_rate": 9.850673092848824e-05, "loss": 0.0471, "step": 12410 }, { "grad_norm": 0.32079949975013733, "learning_rate": 9.850271750165077e-05, "loss": 0.0482, "step": 12420 }, { "grad_norm": 0.30527055263519287, "learning_rate": 9.849869877062902e-05, "loss": 0.0487, "step": 12430 }, { "grad_norm": 0.36184263229370117, "learning_rate": 9.849467473586252e-05, "loss": 0.05, "step": 12440 }, { "grad_norm": 0.35798224806785583, "learning_rate": 9.849064539779127e-05, "loss": 0.0499, "step": 12450 }, { "grad_norm": 0.4760807752609253, "learning_rate": 9.848661075685594e-05, "loss": 0.0534, "step": 12460 }, { "grad_norm": 0.36414894461631775, "learning_rate": 9.848257081349778e-05, "loss": 0.0478, "step": 12470 }, { "grad_norm": 0.3298071622848511, "learning_rate": 9.847852556815856e-05, "loss": 0.05, "step": 12480 }, { "grad_norm": 0.46286457777023315, "learning_rate": 9.847447502128067e-05, "loss": 0.0578, "step": 12490 }, { "grad_norm": 0.2871038615703583, "learning_rate": 9.847041917330708e-05, "loss": 0.0506, "step": 12500 }, { "grad_norm": 0.33409079909324646, "learning_rate": 9.846635802468132e-05, "loss": 0.053, "step": 12510 }, { "grad_norm": 0.4943197965621948, "learning_rate": 9.84622915758475e-05, "loss": 0.0545, "step": 12520 }, { "grad_norm": 0.3362365961074829, "learning_rate": 9.845821982725034e-05, "loss": 0.0465, "step": 12530 }, { "grad_norm": 0.3026328682899475, "learning_rate": 9.845414277933514e-05, "loss": 0.0454, "step": 12540 }, { "grad_norm": 0.24924205243587494, "learning_rate": 9.845006043254771e-05, "loss": 0.046, "step": 12550 }, { "grad_norm": 0.23700740933418274, "learning_rate": 9.844597278733451e-05, "loss": 0.0473, "step": 12560 }, { "grad_norm": 0.3978720009326935, "learning_rate": 9.844187984414259e-05, "loss": 0.043, "step": 12570 }, { "grad_norm": 0.34948378801345825, "learning_rate": 9.84377816034195e-05, "loss": 0.0439, "step": 12580 }, { "grad_norm": 0.2550783157348633, "learning_rate": 9.843367806561345e-05, "loss": 0.0534, "step": 12590 }, { "grad_norm": 0.31474897265434265, "learning_rate": 9.842956923117317e-05, "loss": 0.0469, "step": 12600 }, { "grad_norm": 0.30777695775032043, "learning_rate": 9.842545510054802e-05, "loss": 0.0493, "step": 12610 }, { "grad_norm": 0.3057761788368225, "learning_rate": 9.842133567418792e-05, "loss": 0.0436, "step": 12620 }, { "grad_norm": 0.4030630588531494, "learning_rate": 9.841721095254333e-05, "loss": 0.0474, "step": 12630 }, { "grad_norm": 0.33838674426078796, "learning_rate": 9.841308093606537e-05, "loss": 0.0466, "step": 12640 }, { "grad_norm": 0.3133462071418762, "learning_rate": 9.840894562520565e-05, "loss": 0.0498, "step": 12650 }, { "grad_norm": 0.4308626353740692, "learning_rate": 9.840480502041642e-05, "loss": 0.0471, "step": 12660 }, { "grad_norm": 0.2818009853363037, "learning_rate": 9.840065912215049e-05, "loss": 0.05, "step": 12670 }, { "grad_norm": 0.2600829303264618, "learning_rate": 9.839650793086124e-05, "loss": 0.0448, "step": 12680 }, { "grad_norm": 0.2983139455318451, "learning_rate": 9.839235144700265e-05, "loss": 0.0499, "step": 12690 }, { "grad_norm": 0.33348140120506287, "learning_rate": 9.838818967102926e-05, "loss": 0.0481, "step": 12700 }, { "grad_norm": 0.3430747091770172, "learning_rate": 9.83840226033962e-05, "loss": 0.0505, "step": 12710 }, { "grad_norm": 0.3702607750892639, "learning_rate": 9.837985024455918e-05, "loss": 0.0487, "step": 12720 }, { "grad_norm": 0.24751649796962738, "learning_rate": 9.837567259497447e-05, "loss": 0.0511, "step": 12730 }, { "grad_norm": 0.3068615198135376, "learning_rate": 9.837148965509894e-05, "loss": 0.0463, "step": 12740 }, { "grad_norm": 0.3918103277683258, "learning_rate": 9.836730142539001e-05, "loss": 0.0516, "step": 12750 }, { "grad_norm": 0.3545096218585968, "learning_rate": 9.836310790630574e-05, "loss": 0.0515, "step": 12760 }, { "grad_norm": 0.31353333592414856, "learning_rate": 9.83589090983047e-05, "loss": 0.047, "step": 12770 }, { "grad_norm": 0.42928481101989746, "learning_rate": 9.835470500184605e-05, "loss": 0.0473, "step": 12780 }, { "grad_norm": 0.3187423646450043, "learning_rate": 9.835049561738957e-05, "loss": 0.0414, "step": 12790 }, { "grad_norm": 0.29193809628486633, "learning_rate": 9.834628094539558e-05, "loss": 0.0506, "step": 12800 }, { "grad_norm": 0.37768691778182983, "learning_rate": 9.834206098632499e-05, "loss": 0.0549, "step": 12810 }, { "grad_norm": 0.4379011392593384, "learning_rate": 9.833783574063931e-05, "loss": 0.0508, "step": 12820 }, { "grad_norm": 0.30786609649658203, "learning_rate": 9.833360520880058e-05, "loss": 0.0548, "step": 12830 }, { "grad_norm": 0.3220038414001465, "learning_rate": 9.832936939127144e-05, "loss": 0.0495, "step": 12840 }, { "grad_norm": 0.3596915304660797, "learning_rate": 9.832512828851515e-05, "loss": 0.0487, "step": 12850 }, { "grad_norm": 0.42088595032691956, "learning_rate": 9.832088190099546e-05, "loss": 0.0499, "step": 12860 }, { "grad_norm": 0.4007439613342285, "learning_rate": 9.831663022917679e-05, "loss": 0.0501, "step": 12870 }, { "grad_norm": 0.3098336160182953, "learning_rate": 9.831237327352407e-05, "loss": 0.049, "step": 12880 }, { "grad_norm": 0.2919544577598572, "learning_rate": 9.830811103450286e-05, "loss": 0.0483, "step": 12890 }, { "grad_norm": 0.31587880849838257, "learning_rate": 9.830384351257924e-05, "loss": 0.0486, "step": 12900 }, { "grad_norm": 0.3614634573459625, "learning_rate": 9.829957070821993e-05, "loss": 0.0481, "step": 12910 }, { "grad_norm": 0.27257513999938965, "learning_rate": 9.829529262189218e-05, "loss": 0.0464, "step": 12920 }, { "grad_norm": 0.29772061109542847, "learning_rate": 9.829100925406385e-05, "loss": 0.0418, "step": 12930 }, { "grad_norm": 0.4577263295650482, "learning_rate": 9.828672060520333e-05, "loss": 0.0488, "step": 12940 }, { "grad_norm": 0.3537951409816742, "learning_rate": 9.828242667577966e-05, "loss": 0.0546, "step": 12950 }, { "grad_norm": 0.2788407504558563, "learning_rate": 9.82781274662624e-05, "loss": 0.0454, "step": 12960 }, { "grad_norm": 0.3361513614654541, "learning_rate": 9.82738229771217e-05, "loss": 0.0468, "step": 12970 }, { "grad_norm": 0.2404433935880661, "learning_rate": 9.826951320882829e-05, "loss": 0.0499, "step": 12980 }, { "grad_norm": 0.4146359860897064, "learning_rate": 9.826519816185351e-05, "loss": 0.0459, "step": 12990 }, { "grad_norm": 0.33333441615104675, "learning_rate": 9.826087783666921e-05, "loss": 0.0453, "step": 13000 }, { "grad_norm": 0.30728012323379517, "learning_rate": 9.825655223374787e-05, "loss": 0.0471, "step": 13010 }, { "grad_norm": 0.38800182938575745, "learning_rate": 9.825222135356253e-05, "loss": 0.051, "step": 13020 }, { "grad_norm": 0.36152490973472595, "learning_rate": 9.82478851965868e-05, "loss": 0.0489, "step": 13030 }, { "grad_norm": 0.3052184283733368, "learning_rate": 9.82435437632949e-05, "loss": 0.046, "step": 13040 }, { "grad_norm": 0.3989173471927643, "learning_rate": 9.823919705416158e-05, "loss": 0.052, "step": 13050 }, { "grad_norm": 0.3003065884113312, "learning_rate": 9.82348450696622e-05, "loss": 0.042, "step": 13060 }, { "grad_norm": 0.3062177300453186, "learning_rate": 9.823048781027268e-05, "loss": 0.0446, "step": 13070 }, { "grad_norm": 0.3229503631591797, "learning_rate": 9.822612527646953e-05, "loss": 0.0465, "step": 13080 }, { "grad_norm": 0.24630331993103027, "learning_rate": 9.822175746872984e-05, "loss": 0.0411, "step": 13090 }, { "grad_norm": 0.37411314249038696, "learning_rate": 9.821738438753123e-05, "loss": 0.0526, "step": 13100 }, { "grad_norm": 0.33638936281204224, "learning_rate": 9.821300603335196e-05, "loss": 0.0483, "step": 13110 }, { "grad_norm": 0.2853318750858307, "learning_rate": 9.820862240667085e-05, "loss": 0.0488, "step": 13120 }, { "grad_norm": 0.29516857862472534, "learning_rate": 9.820423350796726e-05, "loss": 0.0446, "step": 13130 }, { "grad_norm": 0.4135952889919281, "learning_rate": 9.819983933772118e-05, "loss": 0.0564, "step": 13140 }, { "grad_norm": 0.31507113575935364, "learning_rate": 9.819543989641314e-05, "loss": 0.0488, "step": 13150 }, { "grad_norm": 0.31467822194099426, "learning_rate": 9.819103518452423e-05, "loss": 0.0463, "step": 13160 }, { "grad_norm": 0.38095763325691223, "learning_rate": 9.818662520253618e-05, "loss": 0.0573, "step": 13170 }, { "grad_norm": 0.35210782289505005, "learning_rate": 9.818220995093126e-05, "loss": 0.0527, "step": 13180 }, { "grad_norm": 0.2887692153453827, "learning_rate": 9.817778943019228e-05, "loss": 0.0472, "step": 13190 }, { "grad_norm": 0.39099594950675964, "learning_rate": 9.81733636408027e-05, "loss": 0.0527, "step": 13200 }, { "grad_norm": 0.33575525879859924, "learning_rate": 9.816893258324649e-05, "loss": 0.0457, "step": 13210 }, { "grad_norm": 0.3095465302467346, "learning_rate": 9.816449625800823e-05, "loss": 0.0577, "step": 13220 }, { "grad_norm": 0.25718069076538086, "learning_rate": 9.816005466557308e-05, "loss": 0.0423, "step": 13230 }, { "grad_norm": 0.3346683382987976, "learning_rate": 9.815560780642674e-05, "loss": 0.0472, "step": 13240 }, { "grad_norm": 0.286268413066864, "learning_rate": 9.815115568105555e-05, "loss": 0.0511, "step": 13250 }, { "grad_norm": 0.3483783006668091, "learning_rate": 9.814669828994638e-05, "loss": 0.0502, "step": 13260 }, { "grad_norm": 0.30364423990249634, "learning_rate": 9.814223563358665e-05, "loss": 0.0528, "step": 13270 }, { "grad_norm": 0.3655754625797272, "learning_rate": 9.813776771246443e-05, "loss": 0.0512, "step": 13280 }, { "grad_norm": 0.34333938360214233, "learning_rate": 9.813329452706829e-05, "loss": 0.0454, "step": 13290 }, { "grad_norm": 0.3577309846878052, "learning_rate": 9.812881607788744e-05, "loss": 0.0494, "step": 13300 }, { "grad_norm": 0.2715643346309662, "learning_rate": 9.812433236541163e-05, "loss": 0.0502, "step": 13310 }, { "grad_norm": 0.3374332785606384, "learning_rate": 9.811984339013116e-05, "loss": 0.05, "step": 13320 }, { "grad_norm": 0.4166634678840637, "learning_rate": 9.811534915253698e-05, "loss": 0.0534, "step": 13330 }, { "grad_norm": 0.36839765310287476, "learning_rate": 9.811084965312056e-05, "loss": 0.0506, "step": 13340 }, { "grad_norm": 0.40521055459976196, "learning_rate": 9.810634489237396e-05, "loss": 0.0504, "step": 13350 }, { "grad_norm": 0.344232439994812, "learning_rate": 9.81018348707898e-05, "loss": 0.0552, "step": 13360 }, { "grad_norm": 0.34629616141319275, "learning_rate": 9.809731958886131e-05, "loss": 0.0494, "step": 13370 }, { "grad_norm": 0.34863603115081787, "learning_rate": 9.809279904708224e-05, "loss": 0.053, "step": 13380 }, { "grad_norm": 0.2674071192741394, "learning_rate": 9.808827324594699e-05, "loss": 0.048, "step": 13390 }, { "grad_norm": 0.3794316351413727, "learning_rate": 9.808374218595046e-05, "loss": 0.0441, "step": 13400 }, { "grad_norm": 0.3687169849872589, "learning_rate": 9.80792058675882e-05, "loss": 0.0486, "step": 13410 }, { "grad_norm": 0.33348339796066284, "learning_rate": 9.807466429135627e-05, "loss": 0.048, "step": 13420 }, { "grad_norm": 0.26267606019973755, "learning_rate": 9.807011745775132e-05, "loss": 0.0493, "step": 13430 }, { "grad_norm": 0.441738098859787, "learning_rate": 9.806556536727061e-05, "loss": 0.0454, "step": 13440 }, { "grad_norm": 0.32658302783966064, "learning_rate": 9.806100802041193e-05, "loss": 0.0453, "step": 13450 }, { "grad_norm": 0.3306415379047394, "learning_rate": 9.805644541767368e-05, "loss": 0.0456, "step": 13460 }, { "grad_norm": 0.3341897130012512, "learning_rate": 9.805187755955478e-05, "loss": 0.0458, "step": 13470 }, { "grad_norm": 0.35578298568725586, "learning_rate": 9.804730444655483e-05, "loss": 0.0508, "step": 13480 }, { "grad_norm": 0.2870798110961914, "learning_rate": 9.804272607917388e-05, "loss": 0.0416, "step": 13490 }, { "grad_norm": 0.3532654345035553, "learning_rate": 9.803814245791265e-05, "loss": 0.0441, "step": 13500 }, { "grad_norm": 0.30776870250701904, "learning_rate": 9.803355358327239e-05, "loss": 0.0483, "step": 13510 }, { "grad_norm": 0.3510238528251648, "learning_rate": 9.802895945575492e-05, "loss": 0.0431, "step": 13520 }, { "grad_norm": 0.3938954472541809, "learning_rate": 9.802436007586266e-05, "loss": 0.0437, "step": 13530 }, { "grad_norm": 0.40181729197502136, "learning_rate": 9.801975544409858e-05, "loss": 0.0467, "step": 13540 }, { "grad_norm": 0.3277469873428345, "learning_rate": 9.801514556096625e-05, "loss": 0.0453, "step": 13550 }, { "grad_norm": 0.35354942083358765, "learning_rate": 9.801053042696977e-05, "loss": 0.052, "step": 13560 }, { "grad_norm": 0.25917139649391174, "learning_rate": 9.800591004261388e-05, "loss": 0.047, "step": 13570 }, { "grad_norm": 0.32297950983047485, "learning_rate": 9.800128440840385e-05, "loss": 0.0472, "step": 13580 }, { "grad_norm": 0.48821061849594116, "learning_rate": 9.799665352484552e-05, "loss": 0.046, "step": 13590 }, { "grad_norm": 0.36839520931243896, "learning_rate": 9.799201739244532e-05, "loss": 0.0464, "step": 13600 }, { "grad_norm": 0.3233557641506195, "learning_rate": 9.798737601171025e-05, "loss": 0.0466, "step": 13610 }, { "grad_norm": 0.32160037755966187, "learning_rate": 9.79827293831479e-05, "loss": 0.041, "step": 13620 }, { "grad_norm": 0.40505149960517883, "learning_rate": 9.797807750726638e-05, "loss": 0.0538, "step": 13630 }, { "grad_norm": 0.33935075998306274, "learning_rate": 9.797342038457446e-05, "loss": 0.0475, "step": 13640 }, { "grad_norm": 0.35988545417785645, "learning_rate": 9.796875801558141e-05, "loss": 0.0438, "step": 13650 }, { "grad_norm": 0.3003961145877838, "learning_rate": 9.79640904007971e-05, "loss": 0.0463, "step": 13660 }, { "grad_norm": 0.41174906492233276, "learning_rate": 9.795941754073199e-05, "loss": 0.0438, "step": 13670 }, { "grad_norm": 0.37454381585121155, "learning_rate": 9.795473943589705e-05, "loss": 0.0496, "step": 13680 }, { "grad_norm": 0.36842548847198486, "learning_rate": 9.795005608680394e-05, "loss": 0.0508, "step": 13690 }, { "grad_norm": 0.2764884829521179, "learning_rate": 9.794536749396477e-05, "loss": 0.0491, "step": 13700 }, { "grad_norm": 0.440463125705719, "learning_rate": 9.79406736578923e-05, "loss": 0.0526, "step": 13710 }, { "grad_norm": 0.3320474624633789, "learning_rate": 9.793597457909984e-05, "loss": 0.0512, "step": 13720 }, { "grad_norm": 0.2751738429069519, "learning_rate": 9.793127025810127e-05, "loss": 0.0459, "step": 13730 }, { "grad_norm": 0.33384355902671814, "learning_rate": 9.792656069541104e-05, "loss": 0.039, "step": 13740 }, { "grad_norm": 0.35671043395996094, "learning_rate": 9.79218458915442e-05, "loss": 0.0526, "step": 13750 }, { "grad_norm": 0.2866700291633606, "learning_rate": 9.791712584701634e-05, "loss": 0.045, "step": 13760 }, { "grad_norm": 0.2703080475330353, "learning_rate": 9.791240056234364e-05, "loss": 0.0428, "step": 13770 }, { "grad_norm": 0.2870136797428131, "learning_rate": 9.790767003804283e-05, "loss": 0.0436, "step": 13780 }, { "grad_norm": 0.3164782226085663, "learning_rate": 9.790293427463126e-05, "loss": 0.0451, "step": 13790 }, { "grad_norm": 0.32408663630485535, "learning_rate": 9.789819327262684e-05, "loss": 0.0401, "step": 13800 }, { "grad_norm": 0.3544343113899231, "learning_rate": 9.7893447032548e-05, "loss": 0.0395, "step": 13810 }, { "grad_norm": 0.35057541728019714, "learning_rate": 9.78886955549138e-05, "loss": 0.0425, "step": 13820 }, { "grad_norm": 0.38948628306388855, "learning_rate": 9.788393884024387e-05, "loss": 0.0425, "step": 13830 }, { "grad_norm": 0.29353365302085876, "learning_rate": 9.787917688905836e-05, "loss": 0.0476, "step": 13840 }, { "grad_norm": 0.406169056892395, "learning_rate": 9.787440970187807e-05, "loss": 0.0434, "step": 13850 }, { "grad_norm": 0.3028606176376343, "learning_rate": 9.786963727922429e-05, "loss": 0.0506, "step": 13860 }, { "grad_norm": 0.31004565954208374, "learning_rate": 9.786485962161897e-05, "loss": 0.0388, "step": 13870 }, { "grad_norm": 0.2907721996307373, "learning_rate": 9.786007672958455e-05, "loss": 0.042, "step": 13880 }, { "grad_norm": 0.31260034441947937, "learning_rate": 9.78552886036441e-05, "loss": 0.0438, "step": 13890 }, { "grad_norm": 0.2363441437482834, "learning_rate": 9.785049524432124e-05, "loss": 0.0412, "step": 13900 }, { "grad_norm": 0.2965753674507141, "learning_rate": 9.784569665214016e-05, "loss": 0.0413, "step": 13910 }, { "grad_norm": 0.37589743733406067, "learning_rate": 9.784089282762563e-05, "loss": 0.0418, "step": 13920 }, { "grad_norm": 0.25861918926239014, "learning_rate": 9.7836083771303e-05, "loss": 0.0454, "step": 13930 }, { "grad_norm": 0.35295525193214417, "learning_rate": 9.783126948369817e-05, "loss": 0.0408, "step": 13940 }, { "grad_norm": 0.3085072338581085, "learning_rate": 9.78264499653376e-05, "loss": 0.049, "step": 13950 }, { "grad_norm": 0.34597522020339966, "learning_rate": 9.782162521674838e-05, "loss": 0.0434, "step": 13960 }, { "grad_norm": 0.3077165484428406, "learning_rate": 9.781679523845812e-05, "loss": 0.046, "step": 13970 }, { "grad_norm": 0.30466094613075256, "learning_rate": 9.781196003099502e-05, "loss": 0.0379, "step": 13980 }, { "grad_norm": 0.2920159697532654, "learning_rate": 9.780711959488786e-05, "loss": 0.0457, "step": 13990 }, { "grad_norm": 0.3415745496749878, "learning_rate": 9.780227393066599e-05, "loss": 0.0418, "step": 14000 }, { "grad_norm": 0.38001400232315063, "learning_rate": 9.77974230388593e-05, "loss": 0.0467, "step": 14010 }, { "grad_norm": 0.2957090735435486, "learning_rate": 9.779256691999829e-05, "loss": 0.0527, "step": 14020 }, { "grad_norm": 0.33138442039489746, "learning_rate": 9.778770557461403e-05, "loss": 0.0501, "step": 14030 }, { "grad_norm": 0.27471238374710083, "learning_rate": 9.778283900323812e-05, "loss": 0.0521, "step": 14040 }, { "grad_norm": 0.3573247492313385, "learning_rate": 9.777796720640277e-05, "loss": 0.0436, "step": 14050 }, { "grad_norm": 0.2985713481903076, "learning_rate": 9.777309018464078e-05, "loss": 0.0412, "step": 14060 }, { "grad_norm": 0.29401925206184387, "learning_rate": 9.776820793848547e-05, "loss": 0.04, "step": 14070 }, { "grad_norm": 0.3720160722732544, "learning_rate": 9.776332046847075e-05, "loss": 0.0452, "step": 14080 }, { "grad_norm": 0.3926321268081665, "learning_rate": 9.775842777513111e-05, "loss": 0.0444, "step": 14090 }, { "grad_norm": 0.3486253619194031, "learning_rate": 9.775352985900163e-05, "loss": 0.042, "step": 14100 }, { "grad_norm": 0.35012003779411316, "learning_rate": 9.774862672061791e-05, "loss": 0.0439, "step": 14110 }, { "grad_norm": 0.36270031332969666, "learning_rate": 9.774371836051616e-05, "loss": 0.0476, "step": 14120 }, { "grad_norm": 0.38845205307006836, "learning_rate": 9.773880477923315e-05, "loss": 0.0478, "step": 14130 }, { "grad_norm": 0.2758466601371765, "learning_rate": 9.773388597730623e-05, "loss": 0.0466, "step": 14140 }, { "grad_norm": 0.32505032420158386, "learning_rate": 9.77289619552733e-05, "loss": 0.0462, "step": 14150 }, { "grad_norm": 0.25244373083114624, "learning_rate": 9.772403271367285e-05, "loss": 0.042, "step": 14160 }, { "grad_norm": 0.3590511679649353, "learning_rate": 9.771909825304396e-05, "loss": 0.0536, "step": 14170 }, { "grad_norm": 0.3262268006801605, "learning_rate": 9.771415857392619e-05, "loss": 0.0493, "step": 14180 }, { "grad_norm": 0.3380194902420044, "learning_rate": 9.770921367685978e-05, "loss": 0.044, "step": 14190 }, { "grad_norm": 0.3094131350517273, "learning_rate": 9.770426356238551e-05, "loss": 0.0486, "step": 14200 }, { "grad_norm": 0.34609758853912354, "learning_rate": 9.769930823104469e-05, "loss": 0.048, "step": 14210 }, { "grad_norm": 0.39341816306114197, "learning_rate": 9.769434768337926e-05, "loss": 0.0469, "step": 14220 }, { "grad_norm": 0.4522303640842438, "learning_rate": 9.768938191993164e-05, "loss": 0.0455, "step": 14230 }, { "grad_norm": 0.3681087791919708, "learning_rate": 9.768441094124494e-05, "loss": 0.0509, "step": 14240 }, { "grad_norm": 0.3276671767234802, "learning_rate": 9.767943474786275e-05, "loss": 0.0418, "step": 14250 }, { "grad_norm": 0.37169158458709717, "learning_rate": 9.767445334032923e-05, "loss": 0.0438, "step": 14260 }, { "grad_norm": 0.3549058139324188, "learning_rate": 9.766946671918919e-05, "loss": 0.0456, "step": 14270 }, { "grad_norm": 0.3204484283924103, "learning_rate": 9.766447488498796e-05, "loss": 0.0483, "step": 14280 }, { "grad_norm": 0.2897203862667084, "learning_rate": 9.765947783827139e-05, "loss": 0.0398, "step": 14290 }, { "grad_norm": 0.36004355549812317, "learning_rate": 9.765447557958599e-05, "loss": 0.0445, "step": 14300 }, { "grad_norm": 0.32745644450187683, "learning_rate": 9.764946810947879e-05, "loss": 0.0439, "step": 14310 }, { "grad_norm": 0.3124435544013977, "learning_rate": 9.764445542849738e-05, "loss": 0.0456, "step": 14320 }, { "grad_norm": 0.40066033601760864, "learning_rate": 9.763943753718998e-05, "loss": 0.0476, "step": 14330 }, { "grad_norm": 0.28518539667129517, "learning_rate": 9.76344144361053e-05, "loss": 0.0456, "step": 14340 }, { "grad_norm": 0.33559921383857727, "learning_rate": 9.762938612579269e-05, "loss": 0.0501, "step": 14350 }, { "grad_norm": 0.3364734351634979, "learning_rate": 9.762435260680202e-05, "loss": 0.0457, "step": 14360 }, { "grad_norm": 0.32486602663993835, "learning_rate": 9.761931387968373e-05, "loss": 0.0473, "step": 14370 }, { "grad_norm": 0.4042028486728668, "learning_rate": 9.76142699449889e-05, "loss": 0.045, "step": 14380 }, { "grad_norm": 0.39334622025489807, "learning_rate": 9.760922080326908e-05, "loss": 0.0536, "step": 14390 }, { "grad_norm": 0.35524091124534607, "learning_rate": 9.760416645507644e-05, "loss": 0.0521, "step": 14400 }, { "grad_norm": 0.32214632630348206, "learning_rate": 9.759910690096375e-05, "loss": 0.0442, "step": 14410 }, { "grad_norm": 0.3594346344470978, "learning_rate": 9.759404214148429e-05, "loss": 0.0494, "step": 14420 }, { "grad_norm": 0.3762790560722351, "learning_rate": 9.758897217719191e-05, "loss": 0.0502, "step": 14430 }, { "grad_norm": 0.3714487552642822, "learning_rate": 9.758389700864113e-05, "loss": 0.0502, "step": 14440 }, { "grad_norm": 0.29588496685028076, "learning_rate": 9.757881663638688e-05, "loss": 0.0465, "step": 14450 }, { "grad_norm": 0.31265607476234436, "learning_rate": 9.757373106098478e-05, "loss": 0.0442, "step": 14460 }, { "grad_norm": 0.3411610722541809, "learning_rate": 9.756864028299097e-05, "loss": 0.043, "step": 14470 }, { "grad_norm": 0.2923526465892792, "learning_rate": 9.75635443029622e-05, "loss": 0.0497, "step": 14480 }, { "grad_norm": 0.3892192244529724, "learning_rate": 9.755844312145572e-05, "loss": 0.0463, "step": 14490 }, { "grad_norm": 0.3228699564933777, "learning_rate": 9.755333673902941e-05, "loss": 0.0416, "step": 14500 }, { "grad_norm": 0.2828015387058258, "learning_rate": 9.75482251562417e-05, "loss": 0.0457, "step": 14510 }, { "grad_norm": 0.3339994251728058, "learning_rate": 9.754310837365155e-05, "loss": 0.0427, "step": 14520 }, { "grad_norm": 0.3083797097206116, "learning_rate": 9.753798639181856e-05, "loss": 0.0422, "step": 14530 }, { "grad_norm": 0.31295502185821533, "learning_rate": 9.753285921130286e-05, "loss": 0.0455, "step": 14540 }, { "grad_norm": 0.29806196689605713, "learning_rate": 9.752772683266512e-05, "loss": 0.0476, "step": 14550 }, { "grad_norm": 0.3738783001899719, "learning_rate": 9.752258925646665e-05, "loss": 0.0487, "step": 14560 }, { "grad_norm": 0.3284691870212555, "learning_rate": 9.751744648326926e-05, "loss": 0.0549, "step": 14570 }, { "grad_norm": 0.3013484477996826, "learning_rate": 9.751229851363536e-05, "loss": 0.0477, "step": 14580 }, { "grad_norm": 0.3360784351825714, "learning_rate": 9.750714534812793e-05, "loss": 0.0552, "step": 14590 }, { "grad_norm": 0.25136396288871765, "learning_rate": 9.750198698731053e-05, "loss": 0.047, "step": 14600 }, { "grad_norm": 0.3690243065357208, "learning_rate": 9.749682343174722e-05, "loss": 0.0418, "step": 14610 }, { "grad_norm": 0.33122098445892334, "learning_rate": 9.749165468200272e-05, "loss": 0.0492, "step": 14620 }, { "grad_norm": 0.30175453424453735, "learning_rate": 9.748648073864229e-05, "loss": 0.0444, "step": 14630 }, { "grad_norm": 0.36585527658462524, "learning_rate": 9.748130160223168e-05, "loss": 0.0487, "step": 14640 }, { "grad_norm": 0.3245047628879547, "learning_rate": 9.747611727333734e-05, "loss": 0.0477, "step": 14650 }, { "grad_norm": 0.30815351009368896, "learning_rate": 9.74709277525262e-05, "loss": 0.0486, "step": 14660 }, { "grad_norm": 0.2963595390319824, "learning_rate": 9.746573304036576e-05, "loss": 0.045, "step": 14670 }, { "grad_norm": 0.2905140519142151, "learning_rate": 9.746053313742412e-05, "loss": 0.0393, "step": 14680 }, { "grad_norm": 0.38620609045028687, "learning_rate": 9.745532804426994e-05, "loss": 0.0485, "step": 14690 }, { "grad_norm": 0.41571998596191406, "learning_rate": 9.745011776147242e-05, "loss": 0.0528, "step": 14700 }, { "grad_norm": 0.269163578748703, "learning_rate": 9.744490228960138e-05, "loss": 0.0409, "step": 14710 }, { "grad_norm": 0.25411248207092285, "learning_rate": 9.743968162922713e-05, "loss": 0.041, "step": 14720 }, { "grad_norm": 0.38109007477760315, "learning_rate": 9.743445578092064e-05, "loss": 0.0429, "step": 14730 }, { "grad_norm": 0.35696813464164734, "learning_rate": 9.742922474525338e-05, "loss": 0.042, "step": 14740 }, { "grad_norm": 0.35835763812065125, "learning_rate": 9.742398852279741e-05, "loss": 0.0464, "step": 14750 }, { "grad_norm": 0.3842819631099701, "learning_rate": 9.741874711412535e-05, "loss": 0.0488, "step": 14760 }, { "grad_norm": 0.3816242218017578, "learning_rate": 9.741350051981042e-05, "loss": 0.0503, "step": 14770 }, { "grad_norm": 0.29834747314453125, "learning_rate": 9.740824874042633e-05, "loss": 0.0478, "step": 14780 }, { "grad_norm": 0.3455534279346466, "learning_rate": 9.740299177654746e-05, "loss": 0.0445, "step": 14790 }, { "grad_norm": 0.33854857087135315, "learning_rate": 9.739772962874867e-05, "loss": 0.0531, "step": 14800 }, { "grad_norm": 0.2720438838005066, "learning_rate": 9.739246229760541e-05, "loss": 0.0585, "step": 14810 }, { "grad_norm": 0.3417336046695709, "learning_rate": 9.738718978369376e-05, "loss": 0.0453, "step": 14820 }, { "grad_norm": 0.29328039288520813, "learning_rate": 9.738191208759025e-05, "loss": 0.0417, "step": 14830 }, { "grad_norm": 0.22284287214279175, "learning_rate": 9.73766292098721e-05, "loss": 0.0437, "step": 14840 }, { "grad_norm": 0.3493037223815918, "learning_rate": 9.737134115111699e-05, "loss": 0.0488, "step": 14850 }, { "grad_norm": 0.28382933139801025, "learning_rate": 9.736604791190323e-05, "loss": 0.0515, "step": 14860 }, { "grad_norm": 0.33378851413726807, "learning_rate": 9.73607494928097e-05, "loss": 0.0442, "step": 14870 }, { "grad_norm": 0.32439833879470825, "learning_rate": 9.735544589441581e-05, "loss": 0.0483, "step": 14880 }, { "grad_norm": 0.3889808654785156, "learning_rate": 9.735013711730154e-05, "loss": 0.0411, "step": 14890 }, { "grad_norm": 0.3918119966983795, "learning_rate": 9.734482316204747e-05, "loss": 0.046, "step": 14900 }, { "grad_norm": 0.3895469903945923, "learning_rate": 9.733950402923473e-05, "loss": 0.0444, "step": 14910 }, { "grad_norm": 0.3309823274612427, "learning_rate": 9.7334179719445e-05, "loss": 0.0538, "step": 14920 }, { "grad_norm": 0.47312131524086, "learning_rate": 9.732885023326053e-05, "loss": 0.0517, "step": 14930 }, { "grad_norm": 0.351066917181015, "learning_rate": 9.732351557126418e-05, "loss": 0.0489, "step": 14940 }, { "grad_norm": 0.3656689524650574, "learning_rate": 9.731817573403929e-05, "loss": 0.0484, "step": 14950 }, { "grad_norm": 0.3353118598461151, "learning_rate": 9.731283072216985e-05, "loss": 0.0504, "step": 14960 }, { "grad_norm": 0.27514445781707764, "learning_rate": 9.730748053624039e-05, "loss": 0.0441, "step": 14970 }, { "grad_norm": 0.3579701781272888, "learning_rate": 9.730212517683598e-05, "loss": 0.05, "step": 14980 }, { "grad_norm": 0.3378818929195404, "learning_rate": 9.729676464454228e-05, "loss": 0.0431, "step": 14990 }, { "grad_norm": 0.2910451292991638, "learning_rate": 9.72913989399455e-05, "loss": 0.0473, "step": 15000 }, { "grad_norm": 0.2947120666503906, "learning_rate": 9.728602806363242e-05, "loss": 0.0429, "step": 15010 }, { "grad_norm": 0.31144240498542786, "learning_rate": 9.728065201619043e-05, "loss": 0.0435, "step": 15020 }, { "grad_norm": 0.3259637951850891, "learning_rate": 9.727527079820742e-05, "loss": 0.0412, "step": 15030 }, { "grad_norm": 0.2389102727174759, "learning_rate": 9.726988441027186e-05, "loss": 0.0436, "step": 15040 }, { "grad_norm": 0.40138718485832214, "learning_rate": 9.726449285297281e-05, "loss": 0.0421, "step": 15050 }, { "grad_norm": 0.402533620595932, "learning_rate": 9.72590961268999e-05, "loss": 0.0516, "step": 15060 }, { "grad_norm": 0.3140102028846741, "learning_rate": 9.725369423264328e-05, "loss": 0.0417, "step": 15070 }, { "grad_norm": 0.2682567238807678, "learning_rate": 9.72482871707937e-05, "loss": 0.0408, "step": 15080 }, { "grad_norm": 0.4103451669216156, "learning_rate": 9.724287494194247e-05, "loss": 0.0473, "step": 15090 }, { "grad_norm": 0.4159606993198395, "learning_rate": 9.723745754668147e-05, "loss": 0.0498, "step": 15100 }, { "grad_norm": 0.30026155710220337, "learning_rate": 9.723203498560313e-05, "loss": 0.0442, "step": 15110 }, { "grad_norm": 0.3286173641681671, "learning_rate": 9.722660725930046e-05, "loss": 0.0473, "step": 15120 }, { "grad_norm": 0.4520411789417267, "learning_rate": 9.722117436836702e-05, "loss": 0.0426, "step": 15130 }, { "grad_norm": 0.2723243534564972, "learning_rate": 9.721573631339696e-05, "loss": 0.0394, "step": 15140 }, { "grad_norm": 0.3165961503982544, "learning_rate": 9.721029309498494e-05, "loss": 0.0431, "step": 15150 }, { "grad_norm": 0.30787885189056396, "learning_rate": 9.720484471372627e-05, "loss": 0.0472, "step": 15160 }, { "grad_norm": 0.3607669770717621, "learning_rate": 9.719939117021673e-05, "loss": 0.0417, "step": 15170 }, { "grad_norm": 0.4301949739456177, "learning_rate": 9.719393246505275e-05, "loss": 0.0401, "step": 15180 }, { "grad_norm": 0.27756327390670776, "learning_rate": 9.718846859883128e-05, "loss": 0.0514, "step": 15190 }, { "grad_norm": 0.3080368936061859, "learning_rate": 9.718299957214982e-05, "loss": 0.0435, "step": 15200 }, { "grad_norm": 0.3367936313152313, "learning_rate": 9.717752538560646e-05, "loss": 0.0445, "step": 15210 }, { "grad_norm": 0.3092474639415741, "learning_rate": 9.717204603979986e-05, "loss": 0.0418, "step": 15220 }, { "grad_norm": 0.3724108636379242, "learning_rate": 9.716656153532922e-05, "loss": 0.0442, "step": 15230 }, { "grad_norm": 0.31247586011886597, "learning_rate": 9.716107187279434e-05, "loss": 0.0441, "step": 15240 }, { "grad_norm": 0.3572485148906708, "learning_rate": 9.715557705279555e-05, "loss": 0.0432, "step": 15250 }, { "grad_norm": 0.2946728765964508, "learning_rate": 9.715007707593372e-05, "loss": 0.0374, "step": 15260 }, { "grad_norm": 0.30168354511260986, "learning_rate": 9.714457194281036e-05, "loss": 0.0504, "step": 15270 }, { "grad_norm": 0.3952465057373047, "learning_rate": 9.713906165402751e-05, "loss": 0.0532, "step": 15280 }, { "grad_norm": 0.29152369499206543, "learning_rate": 9.713354621018774e-05, "loss": 0.0456, "step": 15290 }, { "grad_norm": 0.2799842357635498, "learning_rate": 9.712802561189422e-05, "loss": 0.0475, "step": 15300 }, { "grad_norm": 0.3390808403491974, "learning_rate": 9.712249985975069e-05, "loss": 0.0404, "step": 15310 }, { "grad_norm": 0.2804727554321289, "learning_rate": 9.71169689543614e-05, "loss": 0.0374, "step": 15320 }, { "grad_norm": 0.28378525376319885, "learning_rate": 9.711143289633123e-05, "loss": 0.0465, "step": 15330 }, { "grad_norm": 0.35408949851989746, "learning_rate": 9.710589168626561e-05, "loss": 0.043, "step": 15340 }, { "grad_norm": 0.2621772885322571, "learning_rate": 9.710034532477048e-05, "loss": 0.0392, "step": 15350 }, { "grad_norm": 0.27282842993736267, "learning_rate": 9.709479381245239e-05, "loss": 0.0482, "step": 15360 }, { "grad_norm": 0.32468438148498535, "learning_rate": 9.708923714991847e-05, "loss": 0.0433, "step": 15370 }, { "grad_norm": 0.32101792097091675, "learning_rate": 9.708367533777638e-05, "loss": 0.0385, "step": 15380 }, { "grad_norm": 0.42214861512184143, "learning_rate": 9.707810837663431e-05, "loss": 0.049, "step": 15390 }, { "grad_norm": 0.3496937155723572, "learning_rate": 9.707253626710113e-05, "loss": 0.0441, "step": 15400 }, { "grad_norm": 0.2804862856864929, "learning_rate": 9.706695900978613e-05, "loss": 0.0449, "step": 15410 }, { "grad_norm": 0.3399381935596466, "learning_rate": 9.706137660529926e-05, "loss": 0.0459, "step": 15420 }, { "grad_norm": 0.28969550132751465, "learning_rate": 9.705578905425101e-05, "loss": 0.0396, "step": 15430 }, { "grad_norm": 0.32856640219688416, "learning_rate": 9.705019635725241e-05, "loss": 0.042, "step": 15440 }, { "grad_norm": 0.32885926961898804, "learning_rate": 9.704459851491508e-05, "loss": 0.0452, "step": 15450 }, { "grad_norm": 0.33575883507728577, "learning_rate": 9.703899552785118e-05, "loss": 0.04, "step": 15460 }, { "grad_norm": 0.23824596405029297, "learning_rate": 9.703338739667346e-05, "loss": 0.0409, "step": 15470 }, { "grad_norm": 0.24977383017539978, "learning_rate": 9.70277741219952e-05, "loss": 0.0405, "step": 15480 }, { "grad_norm": 0.394359827041626, "learning_rate": 9.702215570443027e-05, "loss": 0.0433, "step": 15490 }, { "grad_norm": 0.31121349334716797, "learning_rate": 9.701653214459309e-05, "loss": 0.0405, "step": 15500 }, { "grad_norm": 0.2739652395248413, "learning_rate": 9.701090344309865e-05, "loss": 0.0461, "step": 15510 }, { "grad_norm": 0.25030332803726196, "learning_rate": 9.700526960056247e-05, "loss": 0.0396, "step": 15520 }, { "grad_norm": 0.23285554349422455, "learning_rate": 9.699963061760068e-05, "loss": 0.0413, "step": 15530 }, { "grad_norm": 0.32779788970947266, "learning_rate": 9.699398649482997e-05, "loss": 0.0449, "step": 15540 }, { "grad_norm": 0.3232508897781372, "learning_rate": 9.698833723286753e-05, "loss": 0.0403, "step": 15550 }, { "grad_norm": 0.446063756942749, "learning_rate": 9.698268283233118e-05, "loss": 0.0424, "step": 15560 }, { "grad_norm": 0.3109360337257385, "learning_rate": 9.697702329383929e-05, "loss": 0.0434, "step": 15570 }, { "grad_norm": 0.4161677956581116, "learning_rate": 9.697135861801074e-05, "loss": 0.0453, "step": 15580 }, { "grad_norm": 0.4398047924041748, "learning_rate": 9.696568880546505e-05, "loss": 0.0471, "step": 15590 }, { "grad_norm": 0.36379674077033997, "learning_rate": 9.696001385682223e-05, "loss": 0.0458, "step": 15600 }, { "grad_norm": 0.2809068262577057, "learning_rate": 9.695433377270291e-05, "loss": 0.0507, "step": 15610 }, { "grad_norm": 0.32448655366897583, "learning_rate": 9.694864855372824e-05, "loss": 0.0376, "step": 15620 }, { "grad_norm": 0.24834498763084412, "learning_rate": 9.694295820051995e-05, "loss": 0.0403, "step": 15630 }, { "grad_norm": 0.44173550605773926, "learning_rate": 9.693726271370032e-05, "loss": 0.0452, "step": 15640 }, { "grad_norm": 0.29934296011924744, "learning_rate": 9.693156209389221e-05, "loss": 0.0434, "step": 15650 }, { "grad_norm": 0.2900875210762024, "learning_rate": 9.692585634171905e-05, "loss": 0.0416, "step": 15660 }, { "grad_norm": 0.31254106760025024, "learning_rate": 9.692014545780476e-05, "loss": 0.0399, "step": 15670 }, { "grad_norm": 0.27470818161964417, "learning_rate": 9.691442944277393e-05, "loss": 0.0399, "step": 15680 }, { "grad_norm": 0.2911321520805359, "learning_rate": 9.690870829725162e-05, "loss": 0.0392, "step": 15690 }, { "grad_norm": 0.25340989232063293, "learning_rate": 9.69029820218635e-05, "loss": 0.0402, "step": 15700 }, { "grad_norm": 0.25198227167129517, "learning_rate": 9.689725061723579e-05, "loss": 0.0397, "step": 15710 }, { "grad_norm": 0.32718127965927124, "learning_rate": 9.689151408399527e-05, "loss": 0.0444, "step": 15720 }, { "grad_norm": 0.2826775908470154, "learning_rate": 9.688577242276924e-05, "loss": 0.0407, "step": 15730 }, { "grad_norm": 0.3909599781036377, "learning_rate": 9.688002563418566e-05, "loss": 0.037, "step": 15740 }, { "grad_norm": 0.2967315912246704, "learning_rate": 9.687427371887293e-05, "loss": 0.0449, "step": 15750 }, { "grad_norm": 0.3029935359954834, "learning_rate": 9.686851667746012e-05, "loss": 0.0416, "step": 15760 }, { "grad_norm": 0.3712897002696991, "learning_rate": 9.686275451057677e-05, "loss": 0.0408, "step": 15770 }, { "grad_norm": 0.27583178877830505, "learning_rate": 9.685698721885308e-05, "loss": 0.0363, "step": 15780 }, { "grad_norm": 0.2798195779323578, "learning_rate": 9.68512148029197e-05, "loss": 0.0502, "step": 15790 }, { "grad_norm": 0.3004832863807678, "learning_rate": 9.684543726340791e-05, "loss": 0.0393, "step": 15800 }, { "grad_norm": 0.2651129961013794, "learning_rate": 9.683965460094952e-05, "loss": 0.0444, "step": 15810 }, { "grad_norm": 0.31254661083221436, "learning_rate": 9.683386681617694e-05, "loss": 0.0485, "step": 15820 }, { "grad_norm": 0.310168981552124, "learning_rate": 9.68280739097231e-05, "loss": 0.0436, "step": 15830 }, { "grad_norm": 0.28264063596725464, "learning_rate": 9.682227588222148e-05, "loss": 0.0433, "step": 15840 }, { "grad_norm": 0.26456934213638306, "learning_rate": 9.681647273430618e-05, "loss": 0.0395, "step": 15850 }, { "grad_norm": 0.33401933312416077, "learning_rate": 9.681066446661182e-05, "loss": 0.0494, "step": 15860 }, { "grad_norm": 0.286581814289093, "learning_rate": 9.680485107977357e-05, "loss": 0.0455, "step": 15870 }, { "grad_norm": 0.29972583055496216, "learning_rate": 9.679903257442716e-05, "loss": 0.043, "step": 15880 }, { "grad_norm": 0.41323786973953247, "learning_rate": 9.679320895120891e-05, "loss": 0.0401, "step": 15890 }, { "grad_norm": 0.3308199346065521, "learning_rate": 9.67873802107557e-05, "loss": 0.0416, "step": 15900 }, { "grad_norm": 0.3624911904335022, "learning_rate": 9.67815463537049e-05, "loss": 0.0491, "step": 15910 }, { "grad_norm": 0.3598167896270752, "learning_rate": 9.677570738069457e-05, "loss": 0.0436, "step": 15920 }, { "grad_norm": 0.30443042516708374, "learning_rate": 9.676986329236318e-05, "loss": 0.0409, "step": 15930 }, { "grad_norm": 0.3778388798236847, "learning_rate": 9.676401408934987e-05, "loss": 0.0433, "step": 15940 }, { "grad_norm": 0.31096193194389343, "learning_rate": 9.675815977229428e-05, "loss": 0.0458, "step": 15950 }, { "grad_norm": 0.35778284072875977, "learning_rate": 9.675230034183664e-05, "loss": 0.0385, "step": 15960 }, { "grad_norm": 0.3149619996547699, "learning_rate": 9.674643579861773e-05, "loss": 0.0422, "step": 15970 }, { "grad_norm": 0.34815582633018494, "learning_rate": 9.674056614327886e-05, "loss": 0.0382, "step": 15980 }, { "grad_norm": 0.28432512283325195, "learning_rate": 9.673469137646198e-05, "loss": 0.0395, "step": 15990 }, { "grad_norm": 0.37723222374916077, "learning_rate": 9.67288114988095e-05, "loss": 0.0436, "step": 16000 }, { "grad_norm": 0.3678649663925171, "learning_rate": 9.672292651096447e-05, "loss": 0.0458, "step": 16010 }, { "grad_norm": 0.36461248993873596, "learning_rate": 9.671703641357042e-05, "loss": 0.0412, "step": 16020 }, { "grad_norm": 0.3703886568546295, "learning_rate": 9.67111412072715e-05, "loss": 0.0471, "step": 16030 }, { "grad_norm": 0.34818050265312195, "learning_rate": 9.670524089271242e-05, "loss": 0.0417, "step": 16040 }, { "grad_norm": 0.33231985569000244, "learning_rate": 9.669933547053842e-05, "loss": 0.0412, "step": 16050 }, { "grad_norm": 0.36404839158058167, "learning_rate": 9.669342494139531e-05, "loss": 0.0464, "step": 16060 }, { "grad_norm": 0.30577218532562256, "learning_rate": 9.668750930592943e-05, "loss": 0.0425, "step": 16070 }, { "grad_norm": 0.3667142689228058, "learning_rate": 9.668158856478775e-05, "loss": 0.0509, "step": 16080 }, { "grad_norm": 0.37497591972351074, "learning_rate": 9.66756627186177e-05, "loss": 0.0438, "step": 16090 }, { "grad_norm": 0.26843154430389404, "learning_rate": 9.666973176806737e-05, "loss": 0.0357, "step": 16100 }, { "grad_norm": 0.3089504837989807, "learning_rate": 9.666379571378534e-05, "loss": 0.0378, "step": 16110 }, { "grad_norm": 0.4220036566257477, "learning_rate": 9.665785455642076e-05, "loss": 0.0388, "step": 16120 }, { "grad_norm": 0.31290191411972046, "learning_rate": 9.665190829662337e-05, "loss": 0.0474, "step": 16130 }, { "grad_norm": 0.33912473917007446, "learning_rate": 9.664595693504342e-05, "loss": 0.038, "step": 16140 }, { "grad_norm": 0.27295252680778503, "learning_rate": 9.664000047233175e-05, "loss": 0.0389, "step": 16150 }, { "grad_norm": 0.3448605239391327, "learning_rate": 9.663403890913976e-05, "loss": 0.0459, "step": 16160 }, { "grad_norm": 0.36205998063087463, "learning_rate": 9.662807224611938e-05, "loss": 0.0454, "step": 16170 }, { "grad_norm": 0.29304856061935425, "learning_rate": 9.662210048392311e-05, "loss": 0.0414, "step": 16180 }, { "grad_norm": 0.3001982867717743, "learning_rate": 9.661612362320405e-05, "loss": 0.0445, "step": 16190 }, { "grad_norm": 0.3060000240802765, "learning_rate": 9.661014166461579e-05, "loss": 0.0447, "step": 16200 }, { "grad_norm": 0.33609646558761597, "learning_rate": 9.66041546088125e-05, "loss": 0.0522, "step": 16210 }, { "grad_norm": 0.27222564816474915, "learning_rate": 9.659816245644895e-05, "loss": 0.0412, "step": 16220 }, { "grad_norm": 0.23174653947353363, "learning_rate": 9.65921652081804e-05, "loss": 0.0438, "step": 16230 }, { "grad_norm": 0.3359189033508301, "learning_rate": 9.658616286466271e-05, "loss": 0.0436, "step": 16240 }, { "grad_norm": 0.2563556730747223, "learning_rate": 9.65801554265523e-05, "loss": 0.0417, "step": 16250 }, { "grad_norm": 0.24515239894390106, "learning_rate": 9.657414289450612e-05, "loss": 0.0385, "step": 16260 }, { "grad_norm": 0.3354510962963104, "learning_rate": 9.656812526918171e-05, "loss": 0.0388, "step": 16270 }, { "grad_norm": 0.3740955889225006, "learning_rate": 9.656210255123712e-05, "loss": 0.0435, "step": 16280 }, { "grad_norm": 0.3210921883583069, "learning_rate": 9.6556074741331e-05, "loss": 0.0391, "step": 16290 }, { "grad_norm": 0.273158997297287, "learning_rate": 9.655004184012256e-05, "loss": 0.0448, "step": 16300 }, { "grad_norm": 0.3097323477268219, "learning_rate": 9.654400384827152e-05, "loss": 0.0455, "step": 16310 }, { "grad_norm": 0.3523668944835663, "learning_rate": 9.653796076643818e-05, "loss": 0.0429, "step": 16320 }, { "grad_norm": 0.2386063039302826, "learning_rate": 9.653191259528344e-05, "loss": 0.0445, "step": 16330 }, { "grad_norm": 0.3154197335243225, "learning_rate": 9.65258593354687e-05, "loss": 0.0437, "step": 16340 }, { "grad_norm": 0.2429541051387787, "learning_rate": 9.651980098765591e-05, "loss": 0.043, "step": 16350 }, { "grad_norm": 0.26738959550857544, "learning_rate": 9.651373755250765e-05, "loss": 0.0457, "step": 16360 }, { "grad_norm": 0.2637206017971039, "learning_rate": 9.650766903068697e-05, "loss": 0.0403, "step": 16370 }, { "grad_norm": 0.2940979599952698, "learning_rate": 9.650159542285753e-05, "loss": 0.0452, "step": 16380 }, { "grad_norm": 0.36707550287246704, "learning_rate": 9.649551672968353e-05, "loss": 0.0437, "step": 16390 }, { "grad_norm": 0.3337688744068146, "learning_rate": 9.648943295182973e-05, "loss": 0.0393, "step": 16400 }, { "grad_norm": 0.27663514018058777, "learning_rate": 9.648334408996144e-05, "loss": 0.0421, "step": 16410 }, { "grad_norm": 0.24658764898777008, "learning_rate": 9.647725014474452e-05, "loss": 0.0435, "step": 16420 }, { "grad_norm": 0.3212276101112366, "learning_rate": 9.64711511168454e-05, "loss": 0.0389, "step": 16430 }, { "grad_norm": 0.3392513692378998, "learning_rate": 9.646504700693108e-05, "loss": 0.0401, "step": 16440 }, { "grad_norm": 0.41584256291389465, "learning_rate": 9.645893781566907e-05, "loss": 0.041, "step": 16450 }, { "grad_norm": 0.3943125307559967, "learning_rate": 9.645282354372744e-05, "loss": 0.0428, "step": 16460 }, { "grad_norm": 0.31869906187057495, "learning_rate": 9.644670419177491e-05, "loss": 0.0406, "step": 16470 }, { "grad_norm": 0.398255318403244, "learning_rate": 9.644057976048062e-05, "loss": 0.0403, "step": 16480 }, { "grad_norm": 0.3808842897415161, "learning_rate": 9.643445025051435e-05, "loss": 0.0402, "step": 16490 }, { "grad_norm": 0.33736950159072876, "learning_rate": 9.642831566254641e-05, "loss": 0.044, "step": 16500 }, { "grad_norm": 0.26574525237083435, "learning_rate": 9.642217599724769e-05, "loss": 0.0477, "step": 16510 }, { "grad_norm": 0.5247750282287598, "learning_rate": 9.64160312552896e-05, "loss": 0.0399, "step": 16520 }, { "grad_norm": 0.3146195411682129, "learning_rate": 9.64098814373441e-05, "loss": 0.0484, "step": 16530 }, { "grad_norm": 0.3111524283885956, "learning_rate": 9.640372654408374e-05, "loss": 0.0426, "step": 16540 }, { "grad_norm": 0.2844304144382477, "learning_rate": 9.639756657618162e-05, "loss": 0.041, "step": 16550 }, { "grad_norm": 0.34984534978866577, "learning_rate": 9.639140153431138e-05, "loss": 0.0448, "step": 16560 }, { "grad_norm": 0.30989980697631836, "learning_rate": 9.638523141914721e-05, "loss": 0.0397, "step": 16570 }, { "grad_norm": 0.27076560258865356, "learning_rate": 9.637905623136388e-05, "loss": 0.0407, "step": 16580 }, { "grad_norm": 0.21530506014823914, "learning_rate": 9.637287597163669e-05, "loss": 0.0419, "step": 16590 }, { "grad_norm": 0.25302547216415405, "learning_rate": 9.63666906406415e-05, "loss": 0.0394, "step": 16600 }, { "grad_norm": 0.27485406398773193, "learning_rate": 9.636050023905473e-05, "loss": 0.045, "step": 16610 }, { "grad_norm": 0.33179694414138794, "learning_rate": 9.635430476755336e-05, "loss": 0.049, "step": 16620 }, { "grad_norm": 0.26728692650794983, "learning_rate": 9.63481042268149e-05, "loss": 0.044, "step": 16630 }, { "grad_norm": 0.27728089690208435, "learning_rate": 9.634189861751745e-05, "loss": 0.0438, "step": 16640 }, { "grad_norm": 0.3306518793106079, "learning_rate": 9.633568794033967e-05, "loss": 0.041, "step": 16650 }, { "grad_norm": 0.4234890937805176, "learning_rate": 9.63294721959607e-05, "loss": 0.0375, "step": 16660 }, { "grad_norm": 0.24428604543209076, "learning_rate": 9.63232513850603e-05, "loss": 0.0415, "step": 16670 }, { "grad_norm": 0.2979811429977417, "learning_rate": 9.631702550831878e-05, "loss": 0.0371, "step": 16680 }, { "grad_norm": 0.41324281692504883, "learning_rate": 9.631079456641698e-05, "loss": 0.0399, "step": 16690 }, { "grad_norm": 0.25200286507606506, "learning_rate": 9.630455856003632e-05, "loss": 0.0368, "step": 16700 }, { "grad_norm": 0.3774970769882202, "learning_rate": 9.629831748985876e-05, "loss": 0.041, "step": 16710 }, { "grad_norm": 0.26026520133018494, "learning_rate": 9.629207135656679e-05, "loss": 0.0383, "step": 16720 }, { "grad_norm": 0.29690057039260864, "learning_rate": 9.628582016084353e-05, "loss": 0.0383, "step": 16730 }, { "grad_norm": 0.3014800548553467, "learning_rate": 9.627956390337254e-05, "loss": 0.043, "step": 16740 }, { "grad_norm": 0.37475457787513733, "learning_rate": 9.627330258483802e-05, "loss": 0.0442, "step": 16750 }, { "grad_norm": 0.30186933279037476, "learning_rate": 9.62670362059247e-05, "loss": 0.0457, "step": 16760 }, { "grad_norm": 0.3371450901031494, "learning_rate": 9.626076476731786e-05, "loss": 0.0445, "step": 16770 }, { "grad_norm": 0.27747029066085815, "learning_rate": 9.625448826970336e-05, "loss": 0.0416, "step": 16780 }, { "grad_norm": 0.2824264466762543, "learning_rate": 9.624820671376755e-05, "loss": 0.0364, "step": 16790 }, { "grad_norm": 0.2496138960123062, "learning_rate": 9.62419201001974e-05, "loss": 0.0411, "step": 16800 }, { "grad_norm": 0.36249035596847534, "learning_rate": 9.623562842968037e-05, "loss": 0.0432, "step": 16810 }, { "grad_norm": 0.3191896080970764, "learning_rate": 9.622933170290454e-05, "loss": 0.0439, "step": 16820 }, { "grad_norm": 0.3708445131778717, "learning_rate": 9.622302992055849e-05, "loss": 0.0389, "step": 16830 }, { "grad_norm": 0.26082926988601685, "learning_rate": 9.62167230833314e-05, "loss": 0.0408, "step": 16840 }, { "grad_norm": 0.26341333985328674, "learning_rate": 9.621041119191295e-05, "loss": 0.0394, "step": 16850 }, { "grad_norm": 0.3167552649974823, "learning_rate": 9.620409424699342e-05, "loss": 0.0456, "step": 16860 }, { "grad_norm": 0.27426743507385254, "learning_rate": 9.619777224926359e-05, "loss": 0.0388, "step": 16870 }, { "grad_norm": 0.329929381608963, "learning_rate": 9.619144519941485e-05, "loss": 0.0433, "step": 16880 }, { "grad_norm": 0.24542711675167084, "learning_rate": 9.618511309813912e-05, "loss": 0.0412, "step": 16890 }, { "grad_norm": 0.34098580479621887, "learning_rate": 9.617877594612886e-05, "loss": 0.0412, "step": 16900 }, { "grad_norm": 0.31020689010620117, "learning_rate": 9.617243374407707e-05, "loss": 0.0394, "step": 16910 }, { "grad_norm": 0.2675580680370331, "learning_rate": 9.616608649267736e-05, "loss": 0.0463, "step": 16920 }, { "grad_norm": 0.17453977465629578, "learning_rate": 9.615973419262385e-05, "loss": 0.0425, "step": 16930 }, { "grad_norm": 0.28328201174736023, "learning_rate": 9.615337684461119e-05, "loss": 0.0462, "step": 16940 }, { "grad_norm": 0.300263911485672, "learning_rate": 9.614701444933465e-05, "loss": 0.0404, "step": 16950 }, { "grad_norm": 0.36560291051864624, "learning_rate": 9.614064700748997e-05, "loss": 0.0439, "step": 16960 }, { "grad_norm": 0.29796987771987915, "learning_rate": 9.613427451977352e-05, "loss": 0.0403, "step": 16970 }, { "grad_norm": 0.35999077558517456, "learning_rate": 9.612789698688216e-05, "loss": 0.0424, "step": 16980 }, { "grad_norm": 0.25568193197250366, "learning_rate": 9.612151440951334e-05, "loss": 0.0422, "step": 16990 }, { "grad_norm": 0.23859135806560516, "learning_rate": 9.611512678836506e-05, "loss": 0.0398, "step": 17000 }, { "grad_norm": 0.2925892174243927, "learning_rate": 9.610873412413584e-05, "loss": 0.0411, "step": 17010 }, { "grad_norm": 0.2874498963356018, "learning_rate": 9.610233641752476e-05, "loss": 0.0438, "step": 17020 }, { "grad_norm": 0.27673542499542236, "learning_rate": 9.609593366923151e-05, "loss": 0.0376, "step": 17030 }, { "grad_norm": 0.31578269600868225, "learning_rate": 9.608952587995625e-05, "loss": 0.0463, "step": 17040 }, { "grad_norm": 0.2937397062778473, "learning_rate": 9.608311305039972e-05, "loss": 0.0453, "step": 17050 }, { "grad_norm": 0.30353420972824097, "learning_rate": 9.607669518126326e-05, "loss": 0.0481, "step": 17060 }, { "grad_norm": 0.30723193287849426, "learning_rate": 9.607027227324866e-05, "loss": 0.0464, "step": 17070 }, { "grad_norm": 0.3082084059715271, "learning_rate": 9.606384432705837e-05, "loss": 0.0416, "step": 17080 }, { "grad_norm": 0.33428457379341125, "learning_rate": 9.60574113433953e-05, "loss": 0.0484, "step": 17090 }, { "grad_norm": 0.3124331533908844, "learning_rate": 9.6050973322963e-05, "loss": 0.0437, "step": 17100 }, { "grad_norm": 0.4215680658817291, "learning_rate": 9.604453026646547e-05, "loss": 0.056, "step": 17110 }, { "grad_norm": 0.28989553451538086, "learning_rate": 9.603808217460735e-05, "loss": 0.0492, "step": 17120 }, { "grad_norm": 0.27425163984298706, "learning_rate": 9.603162904809377e-05, "loss": 0.047, "step": 17130 }, { "grad_norm": 0.28322306275367737, "learning_rate": 9.602517088763045e-05, "loss": 0.0465, "step": 17140 }, { "grad_norm": 0.34516263008117676, "learning_rate": 9.601870769392365e-05, "loss": 0.0417, "step": 17150 }, { "grad_norm": 0.3403582274913788, "learning_rate": 9.601223946768017e-05, "loss": 0.0495, "step": 17160 }, { "grad_norm": 0.347642183303833, "learning_rate": 9.600576620960734e-05, "loss": 0.0446, "step": 17170 }, { "grad_norm": 0.34941500425338745, "learning_rate": 9.599928792041308e-05, "loss": 0.0441, "step": 17180 }, { "grad_norm": 0.37957337498664856, "learning_rate": 9.599280460080587e-05, "loss": 0.0519, "step": 17190 }, { "grad_norm": 0.3645027279853821, "learning_rate": 9.59863162514947e-05, "loss": 0.0378, "step": 17200 }, { "grad_norm": 0.28003156185150146, "learning_rate": 9.597982287318911e-05, "loss": 0.0375, "step": 17210 }, { "grad_norm": 0.2880697548389435, "learning_rate": 9.597332446659923e-05, "loss": 0.0423, "step": 17220 }, { "grad_norm": 0.34146347641944885, "learning_rate": 9.59668210324357e-05, "loss": 0.0425, "step": 17230 }, { "grad_norm": 0.2702723741531372, "learning_rate": 9.596031257140974e-05, "loss": 0.0415, "step": 17240 }, { "grad_norm": 0.28757187724113464, "learning_rate": 9.59537990842331e-05, "loss": 0.0341, "step": 17250 }, { "grad_norm": 0.45637020468711853, "learning_rate": 9.594728057161806e-05, "loss": 0.0409, "step": 17260 }, { "grad_norm": 0.3137635886669159, "learning_rate": 9.594075703427752e-05, "loss": 0.0398, "step": 17270 }, { "grad_norm": 0.2975374460220337, "learning_rate": 9.593422847292486e-05, "loss": 0.0406, "step": 17280 }, { "grad_norm": 0.30357804894447327, "learning_rate": 9.592769488827402e-05, "loss": 0.0419, "step": 17290 }, { "grad_norm": 0.29959139227867126, "learning_rate": 9.592115628103952e-05, "loss": 0.0424, "step": 17300 }, { "grad_norm": 0.3098236322402954, "learning_rate": 9.591461265193643e-05, "loss": 0.0457, "step": 17310 }, { "grad_norm": 0.30417507886886597, "learning_rate": 9.590806400168032e-05, "loss": 0.0431, "step": 17320 }, { "grad_norm": 0.38974329829216003, "learning_rate": 9.590151033098735e-05, "loss": 0.0442, "step": 17330 }, { "grad_norm": 0.26622599363327026, "learning_rate": 9.589495164057423e-05, "loss": 0.0457, "step": 17340 }, { "grad_norm": 0.31349921226501465, "learning_rate": 9.58883879311582e-05, "loss": 0.0425, "step": 17350 }, { "grad_norm": 0.2731916308403015, "learning_rate": 9.588181920345705e-05, "loss": 0.0442, "step": 17360 }, { "grad_norm": 0.2772630453109741, "learning_rate": 9.587524545818913e-05, "loss": 0.0466, "step": 17370 }, { "grad_norm": 0.2465173900127411, "learning_rate": 9.586866669607335e-05, "loss": 0.0406, "step": 17380 }, { "grad_norm": 0.37865009903907776, "learning_rate": 9.586208291782915e-05, "loss": 0.0339, "step": 17390 }, { "grad_norm": 0.3467321991920471, "learning_rate": 9.58554941241765e-05, "loss": 0.0412, "step": 17400 }, { "grad_norm": 0.22462882101535797, "learning_rate": 9.584890031583596e-05, "loss": 0.0387, "step": 17410 }, { "grad_norm": 0.2967197000980377, "learning_rate": 9.584230149352861e-05, "loss": 0.0311, "step": 17420 }, { "grad_norm": 0.3357115685939789, "learning_rate": 9.58356976579761e-05, "loss": 0.034, "step": 17430 }, { "grad_norm": 0.3953598439693451, "learning_rate": 9.58290888099006e-05, "loss": 0.0407, "step": 17440 }, { "grad_norm": 0.2909434139728546, "learning_rate": 9.582247495002486e-05, "loss": 0.044, "step": 17450 }, { "grad_norm": 0.28194257616996765, "learning_rate": 9.581585607907214e-05, "loss": 0.0476, "step": 17460 }, { "grad_norm": 0.3491169214248657, "learning_rate": 9.580923219776628e-05, "loss": 0.0417, "step": 17470 }, { "grad_norm": 0.2800208628177643, "learning_rate": 9.580260330683167e-05, "loss": 0.0424, "step": 17480 }, { "grad_norm": 0.25483956933021545, "learning_rate": 9.579596940699322e-05, "loss": 0.0448, "step": 17490 }, { "grad_norm": 0.33157065510749817, "learning_rate": 9.578933049897643e-05, "loss": 0.0469, "step": 17500 }, { "grad_norm": 0.3129747807979584, "learning_rate": 9.578268658350728e-05, "loss": 0.0418, "step": 17510 }, { "grad_norm": 0.2999844253063202, "learning_rate": 9.577603766131235e-05, "loss": 0.0404, "step": 17520 }, { "grad_norm": 0.3294738531112671, "learning_rate": 9.576938373311878e-05, "loss": 0.0423, "step": 17530 }, { "grad_norm": 0.2830516993999481, "learning_rate": 9.576272479965421e-05, "loss": 0.038, "step": 17540 }, { "grad_norm": 0.24196186661720276, "learning_rate": 9.575606086164687e-05, "loss": 0.0387, "step": 17550 }, { "grad_norm": 0.30509310960769653, "learning_rate": 9.57493919198255e-05, "loss": 0.0394, "step": 17560 }, { "grad_norm": 0.32510992884635925, "learning_rate": 9.57427179749194e-05, "loss": 0.0396, "step": 17570 }, { "grad_norm": 0.2571727931499481, "learning_rate": 9.573603902765846e-05, "loss": 0.0338, "step": 17580 }, { "grad_norm": 0.3126014173030853, "learning_rate": 9.572935507877304e-05, "loss": 0.0398, "step": 17590 }, { "grad_norm": 0.2741728723049164, "learning_rate": 9.57226661289941e-05, "loss": 0.0478, "step": 17600 }, { "grad_norm": 0.24495850503444672, "learning_rate": 9.571597217905315e-05, "loss": 0.0395, "step": 17610 }, { "grad_norm": 0.27567508816719055, "learning_rate": 9.57092732296822e-05, "loss": 0.0397, "step": 17620 }, { "grad_norm": 0.29123714566230774, "learning_rate": 9.570256928161385e-05, "loss": 0.0445, "step": 17630 }, { "grad_norm": 0.3126189112663269, "learning_rate": 9.569586033558126e-05, "loss": 0.0398, "step": 17640 }, { "grad_norm": 0.2972199022769928, "learning_rate": 9.568914639231807e-05, "loss": 0.0489, "step": 17650 }, { "grad_norm": 0.2243456095457077, "learning_rate": 9.568242745255852e-05, "loss": 0.0415, "step": 17660 }, { "grad_norm": 0.27293676137924194, "learning_rate": 9.567570351703739e-05, "loss": 0.0376, "step": 17670 }, { "grad_norm": 0.24774008989334106, "learning_rate": 9.566897458649001e-05, "loss": 0.0461, "step": 17680 }, { "grad_norm": 0.33632075786590576, "learning_rate": 9.566224066165221e-05, "loss": 0.0445, "step": 17690 }, { "grad_norm": 0.3553330898284912, "learning_rate": 9.565550174326043e-05, "loss": 0.0424, "step": 17700 }, { "grad_norm": 0.293307363986969, "learning_rate": 9.564875783205162e-05, "loss": 0.0371, "step": 17710 }, { "grad_norm": 0.33411768078804016, "learning_rate": 9.564200892876328e-05, "loss": 0.0451, "step": 17720 }, { "grad_norm": 0.35419073700904846, "learning_rate": 9.563525503413348e-05, "loss": 0.0445, "step": 17730 }, { "grad_norm": 0.33980098366737366, "learning_rate": 9.562849614890079e-05, "loss": 0.0388, "step": 17740 }, { "grad_norm": 0.2851468324661255, "learning_rate": 9.562173227380436e-05, "loss": 0.0381, "step": 17750 }, { "grad_norm": 0.31337305903434753, "learning_rate": 9.561496340958389e-05, "loss": 0.0423, "step": 17760 }, { "grad_norm": 0.7287373542785645, "learning_rate": 9.560818955697959e-05, "loss": 0.0422, "step": 17770 }, { "grad_norm": 0.3186013400554657, "learning_rate": 9.560141071673228e-05, "loss": 0.0412, "step": 17780 }, { "grad_norm": 0.3464910089969635, "learning_rate": 9.559462688958323e-05, "loss": 0.0412, "step": 17790 }, { "grad_norm": 0.34606826305389404, "learning_rate": 9.558783807627434e-05, "loss": 0.0381, "step": 17800 }, { "grad_norm": 0.27765601873397827, "learning_rate": 9.558104427754801e-05, "loss": 0.0444, "step": 17810 }, { "grad_norm": 0.330180287361145, "learning_rate": 9.557424549414722e-05, "loss": 0.0405, "step": 17820 }, { "grad_norm": 0.2588832378387451, "learning_rate": 9.556744172681546e-05, "loss": 0.0386, "step": 17830 }, { "grad_norm": 0.2580723762512207, "learning_rate": 9.556063297629677e-05, "loss": 0.041, "step": 17840 }, { "grad_norm": 0.2563159763813019, "learning_rate": 9.555381924333578e-05, "loss": 0.0415, "step": 17850 }, { "grad_norm": 0.37567687034606934, "learning_rate": 9.554700052867758e-05, "loss": 0.0371, "step": 17860 }, { "grad_norm": 0.2557414174079895, "learning_rate": 9.554017683306789e-05, "loss": 0.0387, "step": 17870 }, { "grad_norm": 0.2679833769798279, "learning_rate": 9.553334815725294e-05, "loss": 0.0432, "step": 17880 }, { "grad_norm": 0.2669524848461151, "learning_rate": 9.552651450197949e-05, "loss": 0.0405, "step": 17890 }, { "grad_norm": 0.4022902548313141, "learning_rate": 9.551967586799486e-05, "loss": 0.0456, "step": 17900 }, { "grad_norm": 0.285971075296402, "learning_rate": 9.551283225604692e-05, "loss": 0.0429, "step": 17910 }, { "grad_norm": 0.24196597933769226, "learning_rate": 9.550598366688406e-05, "loss": 0.0408, "step": 17920 }, { "grad_norm": 0.27677538990974426, "learning_rate": 9.549913010125526e-05, "loss": 0.038, "step": 17930 }, { "grad_norm": 0.32920971512794495, "learning_rate": 9.549227155990999e-05, "loss": 0.0377, "step": 17940 }, { "grad_norm": 0.2700376808643341, "learning_rate": 9.548540804359828e-05, "loss": 0.0375, "step": 17950 }, { "grad_norm": 0.24227701127529144, "learning_rate": 9.547853955307077e-05, "loss": 0.0378, "step": 17960 }, { "grad_norm": 0.30686068534851074, "learning_rate": 9.547166608907853e-05, "loss": 0.038, "step": 17970 }, { "grad_norm": 0.3231710195541382, "learning_rate": 9.546478765237326e-05, "loss": 0.0409, "step": 17980 }, { "grad_norm": 0.44651034474372864, "learning_rate": 9.545790424370715e-05, "loss": 0.0446, "step": 17990 }, { "grad_norm": 0.2475922554731369, "learning_rate": 9.5451015863833e-05, "loss": 0.0406, "step": 18000 }, { "grad_norm": 0.29244863986968994, "learning_rate": 9.544412251350408e-05, "loss": 0.0464, "step": 18010 }, { "grad_norm": 0.2606163024902344, "learning_rate": 9.543722419347422e-05, "loss": 0.0425, "step": 18020 }, { "grad_norm": 0.34266290068626404, "learning_rate": 9.543032090449788e-05, "loss": 0.036, "step": 18030 }, { "grad_norm": 0.33139124512672424, "learning_rate": 9.542341264732992e-05, "loss": 0.0378, "step": 18040 }, { "grad_norm": 0.2765955328941345, "learning_rate": 9.541649942272585e-05, "loss": 0.0413, "step": 18050 }, { "grad_norm": 0.288984090089798, "learning_rate": 9.54095812314417e-05, "loss": 0.0435, "step": 18060 }, { "grad_norm": 0.2979469895362854, "learning_rate": 9.540265807423401e-05, "loss": 0.0388, "step": 18070 }, { "grad_norm": 0.2917594611644745, "learning_rate": 9.53957299518599e-05, "loss": 0.0421, "step": 18080 }, { "grad_norm": 0.3076325058937073, "learning_rate": 9.5388796865077e-05, "loss": 0.0383, "step": 18090 }, { "grad_norm": 0.28375914692878723, "learning_rate": 9.538185881464353e-05, "loss": 0.0484, "step": 18100 }, { "grad_norm": 0.3129662275314331, "learning_rate": 9.537491580131821e-05, "loss": 0.0446, "step": 18110 }, { "grad_norm": 0.2768505811691284, "learning_rate": 9.53679678258603e-05, "loss": 0.0374, "step": 18120 }, { "grad_norm": 0.28951215744018555, "learning_rate": 9.536101488902966e-05, "loss": 0.0409, "step": 18130 }, { "grad_norm": 0.29802125692367554, "learning_rate": 9.535405699158663e-05, "loss": 0.0436, "step": 18140 }, { "grad_norm": 0.28442928194999695, "learning_rate": 9.53470941342921e-05, "loss": 0.0441, "step": 18150 }, { "grad_norm": 0.3183857202529907, "learning_rate": 9.534012631790756e-05, "loss": 0.0416, "step": 18160 }, { "grad_norm": 0.3495400547981262, "learning_rate": 9.533315354319494e-05, "loss": 0.0452, "step": 18170 }, { "grad_norm": 0.27167513966560364, "learning_rate": 9.532617581091682e-05, "loss": 0.0393, "step": 18180 }, { "grad_norm": 0.41342487931251526, "learning_rate": 9.531919312183629e-05, "loss": 0.0449, "step": 18190 }, { "grad_norm": 0.2130117118358612, "learning_rate": 9.531220547671688e-05, "loss": 0.0375, "step": 18200 }, { "grad_norm": 0.2556122839450836, "learning_rate": 9.530521287632285e-05, "loss": 0.0461, "step": 18210 }, { "grad_norm": 0.2223263531923294, "learning_rate": 9.529821532141884e-05, "loss": 0.0393, "step": 18220 }, { "grad_norm": 0.25439122319221497, "learning_rate": 9.52912128127701e-05, "loss": 0.0351, "step": 18230 }, { "grad_norm": 0.30413779616355896, "learning_rate": 9.528420535114244e-05, "loss": 0.0388, "step": 18240 }, { "grad_norm": 0.25840699672698975, "learning_rate": 9.527719293730215e-05, "loss": 0.0386, "step": 18250 }, { "grad_norm": 0.31339654326438904, "learning_rate": 9.527017557201611e-05, "loss": 0.0373, "step": 18260 }, { "grad_norm": 0.45870599150657654, "learning_rate": 9.526315325605176e-05, "loss": 0.0384, "step": 18270 }, { "grad_norm": 0.2612681984901428, "learning_rate": 9.525612599017699e-05, "loss": 0.0413, "step": 18280 }, { "grad_norm": 0.4031091034412384, "learning_rate": 9.524909377516033e-05, "loss": 0.039, "step": 18290 }, { "grad_norm": 0.31723737716674805, "learning_rate": 9.524205661177081e-05, "loss": 0.0445, "step": 18300 }, { "grad_norm": 0.29479897022247314, "learning_rate": 9.523501450077801e-05, "loss": 0.042, "step": 18310 }, { "grad_norm": 0.2996388077735901, "learning_rate": 9.522796744295202e-05, "loss": 0.0397, "step": 18320 }, { "grad_norm": 0.26283299922943115, "learning_rate": 9.522091543906352e-05, "loss": 0.0435, "step": 18330 }, { "grad_norm": 0.3436429798603058, "learning_rate": 9.521385848988369e-05, "loss": 0.0404, "step": 18340 }, { "grad_norm": 0.27270248532295227, "learning_rate": 9.520679659618428e-05, "loss": 0.037, "step": 18350 }, { "grad_norm": 0.32903122901916504, "learning_rate": 9.519972975873754e-05, "loss": 0.0461, "step": 18360 }, { "grad_norm": 0.2884679436683655, "learning_rate": 9.519265797831633e-05, "loss": 0.039, "step": 18370 }, { "grad_norm": 0.30101707577705383, "learning_rate": 9.518558125569399e-05, "loss": 0.0428, "step": 18380 }, { "grad_norm": 0.2733677625656128, "learning_rate": 9.517849959164442e-05, "loss": 0.0426, "step": 18390 }, { "grad_norm": 0.34474876523017883, "learning_rate": 9.517141298694205e-05, "loss": 0.0458, "step": 18400 }, { "grad_norm": 0.39893007278442383, "learning_rate": 9.516432144236188e-05, "loss": 0.0482, "step": 18410 }, { "grad_norm": 0.44775161147117615, "learning_rate": 9.515722495867941e-05, "loss": 0.0468, "step": 18420 }, { "grad_norm": 0.37069424986839294, "learning_rate": 9.515012353667072e-05, "loss": 0.0536, "step": 18430 }, { "grad_norm": 0.38823580741882324, "learning_rate": 9.51430171771124e-05, "loss": 0.0445, "step": 18440 }, { "grad_norm": 0.33926329016685486, "learning_rate": 9.513590588078159e-05, "loss": 0.0429, "step": 18450 }, { "grad_norm": 0.3252306580543518, "learning_rate": 9.512878964845597e-05, "loss": 0.0453, "step": 18460 }, { "grad_norm": 0.31953132152557373, "learning_rate": 9.512166848091377e-05, "loss": 0.0424, "step": 18470 }, { "grad_norm": 0.3354033827781677, "learning_rate": 9.511454237893376e-05, "loss": 0.0437, "step": 18480 }, { "grad_norm": 0.3420087993144989, "learning_rate": 9.51074113432952e-05, "loss": 0.0396, "step": 18490 }, { "grad_norm": 0.27666211128234863, "learning_rate": 9.510027537477797e-05, "loss": 0.0409, "step": 18500 }, { "grad_norm": 0.3046524226665497, "learning_rate": 9.509313447416242e-05, "loss": 0.0375, "step": 18510 }, { "grad_norm": 0.2593348026275635, "learning_rate": 9.508598864222949e-05, "loss": 0.0393, "step": 18520 }, { "grad_norm": 0.2670864164829254, "learning_rate": 9.507883787976062e-05, "loss": 0.0422, "step": 18530 }, { "grad_norm": 0.30834969878196716, "learning_rate": 9.507168218753781e-05, "loss": 0.0446, "step": 18540 }, { "grad_norm": 0.31990280747413635, "learning_rate": 9.506452156634362e-05, "loss": 0.0424, "step": 18550 }, { "grad_norm": 0.27586686611175537, "learning_rate": 9.505735601696109e-05, "loss": 0.0444, "step": 18560 }, { "grad_norm": 0.33491137623786926, "learning_rate": 9.505018554017385e-05, "loss": 0.0415, "step": 18570 }, { "grad_norm": 0.28059113025665283, "learning_rate": 9.504301013676604e-05, "loss": 0.0438, "step": 18580 }, { "grad_norm": 0.2587563395500183, "learning_rate": 9.503582980752238e-05, "loss": 0.0416, "step": 18590 }, { "grad_norm": 0.32043975591659546, "learning_rate": 9.502864455322809e-05, "loss": 0.041, "step": 18600 }, { "grad_norm": 0.3316322863101959, "learning_rate": 9.502145437466891e-05, "loss": 0.0381, "step": 18610 }, { "grad_norm": 0.24624072015285492, "learning_rate": 9.501425927263116e-05, "loss": 0.0347, "step": 18620 }, { "grad_norm": 0.30046120285987854, "learning_rate": 9.500705924790172e-05, "loss": 0.0429, "step": 18630 }, { "grad_norm": 0.307739794254303, "learning_rate": 9.499985430126794e-05, "loss": 0.0417, "step": 18640 }, { "grad_norm": 0.26824402809143066, "learning_rate": 9.499264443351775e-05, "loss": 0.0375, "step": 18650 }, { "grad_norm": 0.3105550706386566, "learning_rate": 9.498542964543961e-05, "loss": 0.0381, "step": 18660 }, { "grad_norm": 0.3388355076313019, "learning_rate": 9.497820993782252e-05, "loss": 0.0376, "step": 18670 }, { "grad_norm": 0.22770841419696808, "learning_rate": 9.497098531145601e-05, "loss": 0.0416, "step": 18680 }, { "grad_norm": 0.3943305015563965, "learning_rate": 9.496375576713017e-05, "loss": 0.0412, "step": 18690 }, { "grad_norm": 0.25705018639564514, "learning_rate": 9.49565213056356e-05, "loss": 0.0448, "step": 18700 }, { "grad_norm": 0.3657456040382385, "learning_rate": 9.494928192776342e-05, "loss": 0.0388, "step": 18710 }, { "grad_norm": 0.333088755607605, "learning_rate": 9.494203763430538e-05, "loss": 0.0401, "step": 18720 }, { "grad_norm": 0.29098764061927795, "learning_rate": 9.493478842605366e-05, "loss": 0.036, "step": 18730 }, { "grad_norm": 0.28773927688598633, "learning_rate": 9.492753430380105e-05, "loss": 0.0398, "step": 18740 }, { "grad_norm": 0.28541573882102966, "learning_rate": 9.492027526834083e-05, "loss": 0.0371, "step": 18750 }, { "grad_norm": 0.25188520550727844, "learning_rate": 9.491301132046684e-05, "loss": 0.0395, "step": 18760 }, { "grad_norm": 0.26351210474967957, "learning_rate": 9.490574246097345e-05, "loss": 0.0302, "step": 18770 }, { "grad_norm": 0.23832230269908905, "learning_rate": 9.48984686906556e-05, "loss": 0.0365, "step": 18780 }, { "grad_norm": 0.2046690136194229, "learning_rate": 9.489119001030871e-05, "loss": 0.0341, "step": 18790 }, { "grad_norm": 0.2574385106563568, "learning_rate": 9.488390642072878e-05, "loss": 0.0357, "step": 18800 }, { "grad_norm": 0.28547319769859314, "learning_rate": 9.48766179227123e-05, "loss": 0.0322, "step": 18810 }, { "grad_norm": 0.3232090175151825, "learning_rate": 9.486932451705636e-05, "loss": 0.0377, "step": 18820 }, { "grad_norm": 0.27693408727645874, "learning_rate": 9.486202620455857e-05, "loss": 0.034, "step": 18830 }, { "grad_norm": 0.26533111929893494, "learning_rate": 9.485472298601704e-05, "loss": 0.0355, "step": 18840 }, { "grad_norm": 0.37727540731430054, "learning_rate": 9.484741486223043e-05, "loss": 0.0409, "step": 18850 }, { "grad_norm": 0.3021012246608734, "learning_rate": 9.484010183399797e-05, "loss": 0.0373, "step": 18860 }, { "grad_norm": 0.2846033275127411, "learning_rate": 9.483278390211938e-05, "loss": 0.0375, "step": 18870 }, { "grad_norm": 0.27670493721961975, "learning_rate": 9.482546106739496e-05, "loss": 0.0374, "step": 18880 }, { "grad_norm": 0.27393415570259094, "learning_rate": 9.48181333306255e-05, "loss": 0.0405, "step": 18890 }, { "grad_norm": 0.23234710097312927, "learning_rate": 9.481080069261237e-05, "loss": 0.0378, "step": 18900 }, { "grad_norm": 0.2165234237909317, "learning_rate": 9.480346315415745e-05, "loss": 0.0347, "step": 18910 }, { "grad_norm": 0.32175853848457336, "learning_rate": 9.479612071606314e-05, "loss": 0.0336, "step": 18920 }, { "grad_norm": 0.2912549078464508, "learning_rate": 9.478877337913244e-05, "loss": 0.0347, "step": 18930 }, { "grad_norm": 0.29439041018486023, "learning_rate": 9.478142114416881e-05, "loss": 0.0337, "step": 18940 }, { "grad_norm": 0.3399583399295807, "learning_rate": 9.47740640119763e-05, "loss": 0.0386, "step": 18950 }, { "grad_norm": 0.33745118975639343, "learning_rate": 9.476670198335947e-05, "loss": 0.0372, "step": 18960 }, { "grad_norm": 0.3324277400970459, "learning_rate": 9.47593350591234e-05, "loss": 0.0399, "step": 18970 }, { "grad_norm": 0.315185010433197, "learning_rate": 9.475196324007376e-05, "loss": 0.0345, "step": 18980 }, { "grad_norm": 0.3069865107536316, "learning_rate": 9.474458652701669e-05, "loss": 0.041, "step": 18990 }, { "grad_norm": 0.3202873170375824, "learning_rate": 9.473720492075892e-05, "loss": 0.0372, "step": 19000 }, { "grad_norm": 0.293597012758255, "learning_rate": 9.472981842210768e-05, "loss": 0.0458, "step": 19010 }, { "grad_norm": 0.2470259666442871, "learning_rate": 9.472242703187074e-05, "loss": 0.0395, "step": 19020 }, { "grad_norm": 0.2698878049850464, "learning_rate": 9.471503075085643e-05, "loss": 0.0381, "step": 19030 }, { "grad_norm": 0.22980070114135742, "learning_rate": 9.470762957987359e-05, "loss": 0.0338, "step": 19040 }, { "grad_norm": 0.21686817705631256, "learning_rate": 9.470022351973158e-05, "loss": 0.0326, "step": 19050 }, { "grad_norm": 0.24532046914100647, "learning_rate": 9.469281257124034e-05, "loss": 0.0347, "step": 19060 }, { "grad_norm": 0.30601298809051514, "learning_rate": 9.46853967352103e-05, "loss": 0.0376, "step": 19070 }, { "grad_norm": 0.2373867928981781, "learning_rate": 9.467797601245246e-05, "loss": 0.0343, "step": 19080 }, { "grad_norm": 0.24964205920696259, "learning_rate": 9.467055040377834e-05, "loss": 0.0434, "step": 19090 }, { "grad_norm": 0.2909947335720062, "learning_rate": 9.466311990999999e-05, "loss": 0.0402, "step": 19100 }, { "grad_norm": 0.2612897753715515, "learning_rate": 9.465568453193e-05, "loss": 0.0387, "step": 19110 }, { "grad_norm": 0.31003692746162415, "learning_rate": 9.464824427038148e-05, "loss": 0.0463, "step": 19120 }, { "grad_norm": 0.2892453372478485, "learning_rate": 9.46407991261681e-05, "loss": 0.04, "step": 19130 }, { "grad_norm": 0.3196597695350647, "learning_rate": 9.463334910010404e-05, "loss": 0.041, "step": 19140 }, { "grad_norm": 0.23874130845069885, "learning_rate": 9.462589419300403e-05, "loss": 0.0402, "step": 19150 }, { "grad_norm": 0.26442885398864746, "learning_rate": 9.461843440568333e-05, "loss": 0.0368, "step": 19160 }, { "grad_norm": 0.2824859619140625, "learning_rate": 9.461096973895773e-05, "loss": 0.0396, "step": 19170 }, { "grad_norm": 0.3051084876060486, "learning_rate": 9.460350019364355e-05, "loss": 0.0416, "step": 19180 }, { "grad_norm": 0.31356748938560486, "learning_rate": 9.459602577055764e-05, "loss": 0.0356, "step": 19190 }, { "grad_norm": 0.29496827721595764, "learning_rate": 9.45885464705174e-05, "loss": 0.0418, "step": 19200 }, { "grad_norm": 0.32431647181510925, "learning_rate": 9.458106229434076e-05, "loss": 0.038, "step": 19210 }, { "grad_norm": 0.27538129687309265, "learning_rate": 9.457357324284617e-05, "loss": 0.037, "step": 19220 }, { "grad_norm": 0.25445234775543213, "learning_rate": 9.456607931685262e-05, "loss": 0.0339, "step": 19230 }, { "grad_norm": 0.2654205560684204, "learning_rate": 9.455858051717965e-05, "loss": 0.0319, "step": 19240 }, { "grad_norm": 0.27617335319519043, "learning_rate": 9.45510768446473e-05, "loss": 0.0325, "step": 19250 }, { "grad_norm": 0.27054375410079956, "learning_rate": 9.454356830007618e-05, "loss": 0.0368, "step": 19260 }, { "grad_norm": 0.2946556508541107, "learning_rate": 9.45360548842874e-05, "loss": 0.0458, "step": 19270 }, { "grad_norm": 0.2757743299007416, "learning_rate": 9.452853659810261e-05, "loss": 0.0371, "step": 19280 }, { "grad_norm": 0.2816832661628723, "learning_rate": 9.452101344234401e-05, "loss": 0.0401, "step": 19290 }, { "grad_norm": 0.20531560480594635, "learning_rate": 9.451348541783431e-05, "loss": 0.0416, "step": 19300 }, { "grad_norm": 0.3278072476387024, "learning_rate": 9.450595252539678e-05, "loss": 0.0454, "step": 19310 }, { "grad_norm": 0.26907727122306824, "learning_rate": 9.449841476585518e-05, "loss": 0.0388, "step": 19320 }, { "grad_norm": 0.289208322763443, "learning_rate": 9.449087214003384e-05, "loss": 0.0328, "step": 19330 }, { "grad_norm": 0.2931489944458008, "learning_rate": 9.448332464875765e-05, "loss": 0.0381, "step": 19340 }, { "grad_norm": 0.27656441926956177, "learning_rate": 9.447577229285192e-05, "loss": 0.0338, "step": 19350 }, { "grad_norm": 0.34551456570625305, "learning_rate": 9.446821507314261e-05, "loss": 0.0361, "step": 19360 }, { "grad_norm": 0.2750381529331207, "learning_rate": 9.446065299045617e-05, "loss": 0.037, "step": 19370 }, { "grad_norm": 0.3200207054615021, "learning_rate": 9.445308604561955e-05, "loss": 0.039, "step": 19380 }, { "grad_norm": 0.23488855361938477, "learning_rate": 9.444551423946028e-05, "loss": 0.0359, "step": 19390 }, { "grad_norm": 0.3115735650062561, "learning_rate": 9.443793757280638e-05, "loss": 0.0402, "step": 19400 }, { "grad_norm": 0.25990545749664307, "learning_rate": 9.443035604648646e-05, "loss": 0.0347, "step": 19410 }, { "grad_norm": 0.36313480138778687, "learning_rate": 9.44227696613296e-05, "loss": 0.0442, "step": 19420 }, { "grad_norm": 0.2828029692173004, "learning_rate": 9.441517841816542e-05, "loss": 0.0388, "step": 19430 }, { "grad_norm": 0.2889588177204132, "learning_rate": 9.440758231782413e-05, "loss": 0.035, "step": 19440 }, { "grad_norm": 0.28632253408432007, "learning_rate": 9.439998136113639e-05, "loss": 0.0353, "step": 19450 }, { "grad_norm": 0.24053709208965302, "learning_rate": 9.439237554893344e-05, "loss": 0.0361, "step": 19460 }, { "grad_norm": 0.33183062076568604, "learning_rate": 9.438476488204705e-05, "loss": 0.0385, "step": 19470 }, { "grad_norm": 0.2936590313911438, "learning_rate": 9.43771493613095e-05, "loss": 0.0403, "step": 19480 }, { "grad_norm": 0.3514171242713928, "learning_rate": 9.436952898755362e-05, "loss": 0.0379, "step": 19490 }, { "grad_norm": 0.31761783361434937, "learning_rate": 9.436190376161276e-05, "loss": 0.0418, "step": 19500 }, { "grad_norm": 0.35218924283981323, "learning_rate": 9.43542736843208e-05, "loss": 0.0397, "step": 19510 }, { "grad_norm": 0.275705486536026, "learning_rate": 9.434663875651216e-05, "loss": 0.0381, "step": 19520 }, { "grad_norm": 0.2796801030635834, "learning_rate": 9.433899897902177e-05, "loss": 0.0398, "step": 19530 }, { "grad_norm": 0.2635821998119354, "learning_rate": 9.433135435268511e-05, "loss": 0.038, "step": 19540 }, { "grad_norm": 0.3929353356361389, "learning_rate": 9.432370487833819e-05, "loss": 0.0379, "step": 19550 }, { "grad_norm": 0.22133208811283112, "learning_rate": 9.431605055681756e-05, "loss": 0.0375, "step": 19560 }, { "grad_norm": 0.2780284285545349, "learning_rate": 9.430839138896026e-05, "loss": 0.0395, "step": 19570 }, { "grad_norm": 0.3343154191970825, "learning_rate": 9.43007273756039e-05, "loss": 0.0311, "step": 19580 }, { "grad_norm": 0.27530691027641296, "learning_rate": 9.429305851758658e-05, "loss": 0.0369, "step": 19590 }, { "grad_norm": 0.2648106515407562, "learning_rate": 9.428538481574699e-05, "loss": 0.0378, "step": 19600 }, { "grad_norm": 0.31952208280563354, "learning_rate": 9.42777062709243e-05, "loss": 0.0495, "step": 19610 }, { "grad_norm": 0.29082632064819336, "learning_rate": 9.427002288395821e-05, "loss": 0.0371, "step": 19620 }, { "grad_norm": 0.2627028822898865, "learning_rate": 9.426233465568898e-05, "loss": 0.0403, "step": 19630 }, { "grad_norm": 0.2822083830833435, "learning_rate": 9.42546415869574e-05, "loss": 0.0387, "step": 19640 }, { "grad_norm": 0.3653187155723572, "learning_rate": 9.424694367860473e-05, "loss": 0.0467, "step": 19650 }, { "grad_norm": 0.2594050467014313, "learning_rate": 9.423924093147284e-05, "loss": 0.0398, "step": 19660 }, { "grad_norm": 0.24725407361984253, "learning_rate": 9.423153334640407e-05, "loss": 0.0411, "step": 19670 }, { "grad_norm": 0.33959147334098816, "learning_rate": 9.42238209242413e-05, "loss": 0.0433, "step": 19680 }, { "grad_norm": 0.327156662940979, "learning_rate": 9.421610366582798e-05, "loss": 0.0451, "step": 19690 }, { "grad_norm": 0.3106740415096283, "learning_rate": 9.420838157200803e-05, "loss": 0.0421, "step": 19700 }, { "grad_norm": 0.32612910866737366, "learning_rate": 9.420065464362594e-05, "loss": 0.0378, "step": 19710 }, { "grad_norm": 0.27762579917907715, "learning_rate": 9.419292288152673e-05, "loss": 0.0371, "step": 19720 }, { "grad_norm": 0.24844211339950562, "learning_rate": 9.418518628655588e-05, "loss": 0.0319, "step": 19730 }, { "grad_norm": 0.30013906955718994, "learning_rate": 9.417744485955951e-05, "loss": 0.0379, "step": 19740 }, { "grad_norm": 0.28103721141815186, "learning_rate": 9.41696986013842e-05, "loss": 0.0369, "step": 19750 }, { "grad_norm": 0.29499927163124084, "learning_rate": 9.416194751287705e-05, "loss": 0.0392, "step": 19760 }, { "grad_norm": 0.27008116245269775, "learning_rate": 9.415419159488572e-05, "loss": 0.0377, "step": 19770 }, { "grad_norm": 0.25202569365501404, "learning_rate": 9.414643084825837e-05, "loss": 0.0364, "step": 19780 }, { "grad_norm": 0.25471943616867065, "learning_rate": 9.413866527384372e-05, "loss": 0.0384, "step": 19790 }, { "grad_norm": 0.28578129410743713, "learning_rate": 9.4130894872491e-05, "loss": 0.0396, "step": 19800 }, { "grad_norm": 0.2486875206232071, "learning_rate": 9.412311964504998e-05, "loss": 0.0397, "step": 19810 }, { "grad_norm": 0.28260594606399536, "learning_rate": 9.411533959237091e-05, "loss": 0.0365, "step": 19820 }, { "grad_norm": 0.412209153175354, "learning_rate": 9.410755471530464e-05, "loss": 0.0413, "step": 19830 }, { "grad_norm": 0.40172672271728516, "learning_rate": 9.40997650147025e-05, "loss": 0.0416, "step": 19840 }, { "grad_norm": 0.3610338270664215, "learning_rate": 9.409197049141637e-05, "loss": 0.0421, "step": 19850 }, { "grad_norm": 0.4426870048046112, "learning_rate": 9.408417114629863e-05, "loss": 0.0364, "step": 19860 }, { "grad_norm": 0.2929072082042694, "learning_rate": 9.40763669802022e-05, "loss": 0.0328, "step": 19870 }, { "grad_norm": 0.32233554124832153, "learning_rate": 9.406855799398056e-05, "loss": 0.0396, "step": 19880 }, { "grad_norm": 0.26089048385620117, "learning_rate": 9.406074418848767e-05, "loss": 0.0446, "step": 19890 }, { "grad_norm": 0.3019491136074066, "learning_rate": 9.405292556457805e-05, "loss": 0.0405, "step": 19900 }, { "grad_norm": 0.22361546754837036, "learning_rate": 9.404510212310671e-05, "loss": 0.0346, "step": 19910 }, { "grad_norm": 0.3869791328907013, "learning_rate": 9.403727386492924e-05, "loss": 0.0404, "step": 19920 }, { "grad_norm": 0.3576466143131256, "learning_rate": 9.40294407909017e-05, "loss": 0.0383, "step": 19930 }, { "grad_norm": 0.30346959829330444, "learning_rate": 9.40216029018807e-05, "loss": 0.0364, "step": 19940 }, { "grad_norm": 0.27719640731811523, "learning_rate": 9.401376019872338e-05, "loss": 0.0352, "step": 19950 }, { "grad_norm": 0.2890178859233856, "learning_rate": 9.400591268228746e-05, "loss": 0.0335, "step": 19960 }, { "grad_norm": 0.3505420684814453, "learning_rate": 9.399806035343106e-05, "loss": 0.0379, "step": 19970 }, { "grad_norm": 0.2760930359363556, "learning_rate": 9.399020321301294e-05, "loss": 0.0404, "step": 19980 }, { "grad_norm": 0.29232272505760193, "learning_rate": 9.398234126189234e-05, "loss": 0.0381, "step": 19990 }, { "grad_norm": 0.32329511642456055, "learning_rate": 9.397447450092902e-05, "loss": 0.0375, "step": 20000 }, { "grad_norm": 0.3867575228214264, "learning_rate": 9.39666029309833e-05, "loss": 0.0429, "step": 20010 }, { "grad_norm": 0.3974667191505432, "learning_rate": 9.395872655291596e-05, "loss": 0.0402, "step": 20020 }, { "grad_norm": 0.2967197000980377, "learning_rate": 9.395084536758838e-05, "loss": 0.0323, "step": 20030 }, { "grad_norm": 0.26567134261131287, "learning_rate": 9.394295937586243e-05, "loss": 0.0401, "step": 20040 }, { "grad_norm": 0.36270710825920105, "learning_rate": 9.393506857860052e-05, "loss": 0.0361, "step": 20050 }, { "grad_norm": 0.2829955816268921, "learning_rate": 9.392717297666555e-05, "loss": 0.0421, "step": 20060 }, { "grad_norm": 0.34510430693626404, "learning_rate": 9.391927257092101e-05, "loss": 0.0395, "step": 20070 }, { "grad_norm": 0.24252592027187347, "learning_rate": 9.391136736223085e-05, "loss": 0.0379, "step": 20080 }, { "grad_norm": 0.3680776059627533, "learning_rate": 9.390345735145956e-05, "loss": 0.0398, "step": 20090 }, { "grad_norm": 0.35604697465896606, "learning_rate": 9.389554253947219e-05, "loss": 0.038, "step": 20100 }, { "grad_norm": 0.2927805781364441, "learning_rate": 9.388762292713428e-05, "loss": 0.0355, "step": 20110 }, { "grad_norm": 0.34969502687454224, "learning_rate": 9.38796985153119e-05, "loss": 0.041, "step": 20120 }, { "grad_norm": 0.28881409764289856, "learning_rate": 9.387176930487169e-05, "loss": 0.0358, "step": 20130 }, { "grad_norm": 0.33356809616088867, "learning_rate": 9.386383529668072e-05, "loss": 0.0361, "step": 20140 }, { "grad_norm": 0.4037269353866577, "learning_rate": 9.385589649160669e-05, "loss": 0.0487, "step": 20150 }, { "grad_norm": 0.2754477858543396, "learning_rate": 9.384795289051775e-05, "loss": 0.0408, "step": 20160 }, { "grad_norm": 0.26894304156303406, "learning_rate": 9.384000449428261e-05, "loss": 0.0384, "step": 20170 }, { "grad_norm": 0.29751187562942505, "learning_rate": 9.383205130377048e-05, "loss": 0.0399, "step": 20180 }, { "grad_norm": 0.25176674127578735, "learning_rate": 9.382409331985114e-05, "loss": 0.0418, "step": 20190 }, { "grad_norm": 0.226679265499115, "learning_rate": 9.381613054339482e-05, "loss": 0.0385, "step": 20200 }, { "grad_norm": 0.2658526599407196, "learning_rate": 9.380816297527235e-05, "loss": 0.0382, "step": 20210 }, { "grad_norm": 0.2943732440471649, "learning_rate": 9.380019061635506e-05, "loss": 0.034, "step": 20220 }, { "grad_norm": 0.3165264427661896, "learning_rate": 9.379221346751474e-05, "loss": 0.0402, "step": 20230 }, { "grad_norm": 0.29029351472854614, "learning_rate": 9.378423152962382e-05, "loss": 0.0375, "step": 20240 }, { "grad_norm": 0.26642417907714844, "learning_rate": 9.377624480355517e-05, "loss": 0.0372, "step": 20250 }, { "grad_norm": 0.3410193920135498, "learning_rate": 9.376825329018219e-05, "loss": 0.0406, "step": 20260 }, { "grad_norm": 0.3660009801387787, "learning_rate": 9.376025699037884e-05, "loss": 0.0385, "step": 20270 }, { "grad_norm": 0.27233344316482544, "learning_rate": 9.37522559050196e-05, "loss": 0.0346, "step": 20280 }, { "grad_norm": 0.3507392108440399, "learning_rate": 9.37442500349794e-05, "loss": 0.0369, "step": 20290 }, { "grad_norm": 0.29239189624786377, "learning_rate": 9.373623938113381e-05, "loss": 0.0366, "step": 20300 }, { "grad_norm": 0.24660751223564148, "learning_rate": 9.372822394435883e-05, "loss": 0.0381, "step": 20310 }, { "grad_norm": 0.27825459837913513, "learning_rate": 9.372020372553102e-05, "loss": 0.0386, "step": 20320 }, { "grad_norm": 0.3055013120174408, "learning_rate": 9.371217872552746e-05, "loss": 0.0449, "step": 20330 }, { "grad_norm": 0.2768404185771942, "learning_rate": 9.370414894522576e-05, "loss": 0.0352, "step": 20340 }, { "grad_norm": 0.3605550527572632, "learning_rate": 9.369611438550406e-05, "loss": 0.0397, "step": 20350 }, { "grad_norm": 0.2717949450016022, "learning_rate": 9.368807504724095e-05, "loss": 0.0368, "step": 20360 }, { "grad_norm": 0.24245135486125946, "learning_rate": 9.368003093131565e-05, "loss": 0.034, "step": 20370 }, { "grad_norm": 0.2679609954357147, "learning_rate": 9.367198203860785e-05, "loss": 0.0366, "step": 20380 }, { "grad_norm": 0.2650412321090698, "learning_rate": 9.366392836999774e-05, "loss": 0.0374, "step": 20390 }, { "grad_norm": 0.26520660519599915, "learning_rate": 9.365586992636607e-05, "loss": 0.0381, "step": 20400 }, { "grad_norm": 0.21747075021266937, "learning_rate": 9.364780670859412e-05, "loss": 0.0407, "step": 20410 }, { "grad_norm": 0.3178766965866089, "learning_rate": 9.363973871756364e-05, "loss": 0.0436, "step": 20420 }, { "grad_norm": 0.23866146802902222, "learning_rate": 9.363166595415696e-05, "loss": 0.0406, "step": 20430 }, { "grad_norm": 0.30555248260498047, "learning_rate": 9.362358841925686e-05, "loss": 0.0376, "step": 20440 }, { "grad_norm": 0.2082252949476242, "learning_rate": 9.361550611374674e-05, "loss": 0.0351, "step": 20450 }, { "grad_norm": 0.2360224425792694, "learning_rate": 9.360741903851043e-05, "loss": 0.034, "step": 20460 }, { "grad_norm": 0.357533723115921, "learning_rate": 9.359932719443236e-05, "loss": 0.0313, "step": 20470 }, { "grad_norm": 0.27811551094055176, "learning_rate": 9.35912305823974e-05, "loss": 0.0352, "step": 20480 }, { "grad_norm": 0.21218034625053406, "learning_rate": 9.358312920329101e-05, "loss": 0.0377, "step": 20490 }, { "grad_norm": 0.28852665424346924, "learning_rate": 9.357502305799914e-05, "loss": 0.0321, "step": 20500 }, { "grad_norm": 0.21259509027004242, "learning_rate": 9.356691214740824e-05, "loss": 0.0329, "step": 20510 }, { "grad_norm": 0.26131564378738403, "learning_rate": 9.355879647240535e-05, "loss": 0.0424, "step": 20520 }, { "grad_norm": 0.3304346799850464, "learning_rate": 9.355067603387798e-05, "loss": 0.0411, "step": 20530 }, { "grad_norm": 0.33373743295669556, "learning_rate": 9.354255083271412e-05, "loss": 0.0446, "step": 20540 }, { "grad_norm": 0.2628534436225891, "learning_rate": 9.353442086980239e-05, "loss": 0.0395, "step": 20550 }, { "grad_norm": 0.3122503459453583, "learning_rate": 9.352628614603185e-05, "loss": 0.0384, "step": 20560 }, { "grad_norm": 0.33395877480506897, "learning_rate": 9.351814666229209e-05, "loss": 0.0336, "step": 20570 }, { "grad_norm": 0.24642972648143768, "learning_rate": 9.351000241947324e-05, "loss": 0.0351, "step": 20580 }, { "grad_norm": 0.31559816002845764, "learning_rate": 9.350185341846594e-05, "loss": 0.0405, "step": 20590 }, { "grad_norm": 0.3201304078102112, "learning_rate": 9.349369966016134e-05, "loss": 0.0377, "step": 20600 }, { "grad_norm": 0.2549767792224884, "learning_rate": 9.348554114545117e-05, "loss": 0.0364, "step": 20610 }, { "grad_norm": 0.3292449712753296, "learning_rate": 9.347737787522758e-05, "loss": 0.0349, "step": 20620 }, { "grad_norm": 0.2544589936733246, "learning_rate": 9.346920985038332e-05, "loss": 0.0378, "step": 20630 }, { "grad_norm": 0.3193288743495941, "learning_rate": 9.346103707181162e-05, "loss": 0.0348, "step": 20640 }, { "grad_norm": 0.2995110750198364, "learning_rate": 9.345285954040626e-05, "loss": 0.041, "step": 20650 }, { "grad_norm": 0.3134433329105377, "learning_rate": 9.34446772570615e-05, "loss": 0.034, "step": 20660 }, { "grad_norm": 0.3285296559333801, "learning_rate": 9.343649022267214e-05, "loss": 0.0338, "step": 20670 }, { "grad_norm": 0.2764222323894501, "learning_rate": 9.342829843813353e-05, "loss": 0.0391, "step": 20680 }, { "grad_norm": 0.23656082153320312, "learning_rate": 9.342010190434149e-05, "loss": 0.0376, "step": 20690 }, { "grad_norm": 0.2951163947582245, "learning_rate": 9.34119006221924e-05, "loss": 0.0462, "step": 20700 }, { "grad_norm": 0.26443108916282654, "learning_rate": 9.340369459258313e-05, "loss": 0.036, "step": 20710 }, { "grad_norm": 0.2650337517261505, "learning_rate": 9.339548381641106e-05, "loss": 0.0409, "step": 20720 }, { "grad_norm": 0.26618149876594543, "learning_rate": 9.338726829457413e-05, "loss": 0.039, "step": 20730 }, { "grad_norm": 0.2722645699977875, "learning_rate": 9.337904802797078e-05, "loss": 0.0413, "step": 20740 }, { "grad_norm": 0.3009454905986786, "learning_rate": 9.337082301749993e-05, "loss": 0.0365, "step": 20750 }, { "grad_norm": 0.26694148778915405, "learning_rate": 9.336259326406109e-05, "loss": 0.0357, "step": 20760 }, { "grad_norm": 0.30775687098503113, "learning_rate": 9.335435876855427e-05, "loss": 0.0351, "step": 20770 }, { "grad_norm": 0.35907623171806335, "learning_rate": 9.334611953187994e-05, "loss": 0.0391, "step": 20780 }, { "grad_norm": 0.2757652997970581, "learning_rate": 9.333787555493914e-05, "loss": 0.0353, "step": 20790 }, { "grad_norm": 0.3288399279117584, "learning_rate": 9.332962683863345e-05, "loss": 0.0365, "step": 20800 }, { "grad_norm": 0.2733585238456726, "learning_rate": 9.332137338386489e-05, "loss": 0.0326, "step": 20810 }, { "grad_norm": 0.2552673816680908, "learning_rate": 9.33131151915361e-05, "loss": 0.038, "step": 20820 }, { "grad_norm": 0.2177245020866394, "learning_rate": 9.330485226255012e-05, "loss": 0.0313, "step": 20830 }, { "grad_norm": 0.255756139755249, "learning_rate": 9.329658459781061e-05, "loss": 0.0334, "step": 20840 }, { "grad_norm": 0.3581545650959015, "learning_rate": 9.328831219822172e-05, "loss": 0.0372, "step": 20850 }, { "grad_norm": 0.32050448656082153, "learning_rate": 9.328003506468808e-05, "loss": 0.0418, "step": 20860 }, { "grad_norm": 0.2663975954055786, "learning_rate": 9.327175319811488e-05, "loss": 0.0337, "step": 20870 }, { "grad_norm": 0.31789901852607727, "learning_rate": 9.326346659940781e-05, "loss": 0.0375, "step": 20880 }, { "grad_norm": 0.23921799659729004, "learning_rate": 9.325517526947308e-05, "loss": 0.0358, "step": 20890 }, { "grad_norm": 0.36372286081314087, "learning_rate": 9.32468792092174e-05, "loss": 0.0382, "step": 20900 }, { "grad_norm": 0.3173399269580841, "learning_rate": 9.323857841954803e-05, "loss": 0.0332, "step": 20910 }, { "grad_norm": 0.2734716236591339, "learning_rate": 9.323027290137276e-05, "loss": 0.0367, "step": 20920 }, { "grad_norm": 0.35170796513557434, "learning_rate": 9.322196265559981e-05, "loss": 0.035, "step": 20930 }, { "grad_norm": 0.2750101387500763, "learning_rate": 9.321364768313803e-05, "loss": 0.0384, "step": 20940 }, { "grad_norm": 0.25385963916778564, "learning_rate": 9.32053279848967e-05, "loss": 0.0365, "step": 20950 }, { "grad_norm": 0.3084172010421753, "learning_rate": 9.319700356178567e-05, "loss": 0.0377, "step": 20960 }, { "grad_norm": 0.26450589299201965, "learning_rate": 9.318867441471527e-05, "loss": 0.0355, "step": 20970 }, { "grad_norm": 0.2403160184621811, "learning_rate": 9.318034054459637e-05, "loss": 0.0371, "step": 20980 }, { "grad_norm": 0.26053255796432495, "learning_rate": 9.317200195234034e-05, "loss": 0.0394, "step": 20990 }, { "grad_norm": 0.29825758934020996, "learning_rate": 9.316365863885909e-05, "loss": 0.0353, "step": 21000 }, { "grad_norm": 0.3399063050746918, "learning_rate": 9.315531060506502e-05, "loss": 0.041, "step": 21010 }, { "grad_norm": 0.3064725399017334, "learning_rate": 9.314695785187108e-05, "loss": 0.0368, "step": 21020 }, { "grad_norm": 0.31965941190719604, "learning_rate": 9.313860038019069e-05, "loss": 0.0368, "step": 21030 }, { "grad_norm": 0.3122275173664093, "learning_rate": 9.313023819093782e-05, "loss": 0.0402, "step": 21040 }, { "grad_norm": 0.2971768379211426, "learning_rate": 9.312187128502695e-05, "loss": 0.0359, "step": 21050 }, { "grad_norm": 0.27696672081947327, "learning_rate": 9.311349966337307e-05, "loss": 0.0402, "step": 21060 }, { "grad_norm": 0.2771132290363312, "learning_rate": 9.310512332689169e-05, "loss": 0.0414, "step": 21070 }, { "grad_norm": 0.29646825790405273, "learning_rate": 9.309674227649883e-05, "loss": 0.0362, "step": 21080 }, { "grad_norm": 0.3174819350242615, "learning_rate": 9.308835651311103e-05, "loss": 0.0438, "step": 21090 }, { "grad_norm": 0.3175964653491974, "learning_rate": 9.307996603764533e-05, "loss": 0.0375, "step": 21100 }, { "grad_norm": 0.3340756893157959, "learning_rate": 9.307157085101932e-05, "loss": 0.0399, "step": 21110 }, { "grad_norm": 0.3241593539714813, "learning_rate": 9.306317095415109e-05, "loss": 0.04, "step": 21120 }, { "grad_norm": 0.30617257952690125, "learning_rate": 9.305476634795922e-05, "loss": 0.0418, "step": 21130 }, { "grad_norm": 0.2917558252811432, "learning_rate": 9.304635703336284e-05, "loss": 0.0399, "step": 21140 }, { "grad_norm": 0.2557065784931183, "learning_rate": 9.303794301128157e-05, "loss": 0.0351, "step": 21150 }, { "grad_norm": 0.26324617862701416, "learning_rate": 9.302952428263555e-05, "loss": 0.0356, "step": 21160 }, { "grad_norm": 0.24725636839866638, "learning_rate": 9.302110084834545e-05, "loss": 0.0389, "step": 21170 }, { "grad_norm": 0.23665794730186462, "learning_rate": 9.301267270933245e-05, "loss": 0.0364, "step": 21180 }, { "grad_norm": 0.2629730701446533, "learning_rate": 9.300423986651823e-05, "loss": 0.0389, "step": 21190 }, { "grad_norm": 0.3648085296154022, "learning_rate": 9.299580232082501e-05, "loss": 0.0405, "step": 21200 }, { "grad_norm": 0.31244075298309326, "learning_rate": 9.298736007317547e-05, "loss": 0.0469, "step": 21210 }, { "grad_norm": 0.2563059628009796, "learning_rate": 9.297891312449288e-05, "loss": 0.0411, "step": 21220 }, { "grad_norm": 0.2815606892108917, "learning_rate": 9.297046147570094e-05, "loss": 0.0442, "step": 21230 }, { "grad_norm": 0.27860361337661743, "learning_rate": 9.296200512772396e-05, "loss": 0.0445, "step": 21240 }, { "grad_norm": 0.28596046566963196, "learning_rate": 9.295354408148668e-05, "loss": 0.0357, "step": 21250 }, { "grad_norm": 0.2610618472099304, "learning_rate": 9.294507833791441e-05, "loss": 0.0402, "step": 21260 }, { "grad_norm": 0.398117333650589, "learning_rate": 9.293660789793295e-05, "loss": 0.0419, "step": 21270 }, { "grad_norm": 0.32503587007522583, "learning_rate": 9.292813276246858e-05, "loss": 0.0426, "step": 21280 }, { "grad_norm": 0.3044886291027069, "learning_rate": 9.291965293244816e-05, "loss": 0.0407, "step": 21290 }, { "grad_norm": 0.3172359764575958, "learning_rate": 9.291116840879904e-05, "loss": 0.0388, "step": 21300 }, { "grad_norm": 0.3862907588481903, "learning_rate": 9.290267919244904e-05, "loss": 0.0418, "step": 21310 }, { "grad_norm": 0.40224650502204895, "learning_rate": 9.289418528432655e-05, "loss": 0.0381, "step": 21320 }, { "grad_norm": 0.29731589555740356, "learning_rate": 9.288568668536045e-05, "loss": 0.0398, "step": 21330 }, { "grad_norm": 0.3628266155719757, "learning_rate": 9.287718339648013e-05, "loss": 0.0421, "step": 21340 }, { "grad_norm": 0.2501077353954315, "learning_rate": 9.28686754186155e-05, "loss": 0.0366, "step": 21350 }, { "grad_norm": 0.3357691466808319, "learning_rate": 9.286016275269698e-05, "loss": 0.0369, "step": 21360 }, { "grad_norm": 0.28886696696281433, "learning_rate": 9.285164539965551e-05, "loss": 0.032, "step": 21370 }, { "grad_norm": 0.32096347212791443, "learning_rate": 9.284312336042251e-05, "loss": 0.0383, "step": 21380 }, { "grad_norm": 0.23163039982318878, "learning_rate": 9.283459663592996e-05, "loss": 0.0371, "step": 21390 }, { "grad_norm": 0.3307441771030426, "learning_rate": 9.282606522711033e-05, "loss": 0.0399, "step": 21400 }, { "grad_norm": 0.27610790729522705, "learning_rate": 9.281752913489657e-05, "loss": 0.037, "step": 21410 }, { "grad_norm": 0.2840551435947418, "learning_rate": 9.280898836022222e-05, "loss": 0.033, "step": 21420 }, { "grad_norm": 0.3289940059185028, "learning_rate": 9.280044290402126e-05, "loss": 0.0384, "step": 21430 }, { "grad_norm": 0.28323301672935486, "learning_rate": 9.279189276722821e-05, "loss": 0.0358, "step": 21440 }, { "grad_norm": 0.3435569107532501, "learning_rate": 9.278333795077812e-05, "loss": 0.0355, "step": 21450 }, { "grad_norm": 0.33552446961402893, "learning_rate": 9.27747784556065e-05, "loss": 0.0392, "step": 21460 }, { "grad_norm": 0.26920828223228455, "learning_rate": 9.276621428264942e-05, "loss": 0.034, "step": 21470 }, { "grad_norm": 0.3044162392616272, "learning_rate": 9.275764543284345e-05, "loss": 0.0384, "step": 21480 }, { "grad_norm": 0.21926414966583252, "learning_rate": 9.274907190712566e-05, "loss": 0.032, "step": 21490 }, { "grad_norm": 0.2007804661989212, "learning_rate": 9.274049370643363e-05, "loss": 0.0341, "step": 21500 }, { "grad_norm": 0.2639056146144867, "learning_rate": 9.273191083170547e-05, "loss": 0.0372, "step": 21510 }, { "grad_norm": 0.48403313755989075, "learning_rate": 9.27233232838798e-05, "loss": 0.039, "step": 21520 }, { "grad_norm": 0.2488095909357071, "learning_rate": 9.27147310638957e-05, "loss": 0.037, "step": 21530 }, { "grad_norm": 0.31102123856544495, "learning_rate": 9.270613417269286e-05, "loss": 0.0427, "step": 21540 }, { "grad_norm": 0.24256375432014465, "learning_rate": 9.269753261121138e-05, "loss": 0.031, "step": 21550 }, { "grad_norm": 0.3626806437969208, "learning_rate": 9.268892638039194e-05, "loss": 0.0341, "step": 21560 }, { "grad_norm": 0.3209006190299988, "learning_rate": 9.268031548117569e-05, "loss": 0.0408, "step": 21570 }, { "grad_norm": 0.24347148835659027, "learning_rate": 9.26716999145043e-05, "loss": 0.0368, "step": 21580 }, { "grad_norm": 0.3576591908931732, "learning_rate": 9.266307968131998e-05, "loss": 0.0335, "step": 21590 }, { "grad_norm": 0.22786739468574524, "learning_rate": 9.26544547825654e-05, "loss": 0.0425, "step": 21600 }, { "grad_norm": 0.27855926752090454, "learning_rate": 9.264582521918376e-05, "loss": 0.0344, "step": 21610 }, { "grad_norm": 0.28774869441986084, "learning_rate": 9.263719099211881e-05, "loss": 0.0314, "step": 21620 }, { "grad_norm": 0.3152264356613159, "learning_rate": 9.262855210231476e-05, "loss": 0.038, "step": 21630 }, { "grad_norm": 0.3731269836425781, "learning_rate": 9.261990855071633e-05, "loss": 0.0368, "step": 21640 }, { "grad_norm": 0.3613412380218506, "learning_rate": 9.261126033826878e-05, "loss": 0.0345, "step": 21650 }, { "grad_norm": 0.31162580847740173, "learning_rate": 9.260260746591786e-05, "loss": 0.0398, "step": 21660 }, { "grad_norm": 0.348913311958313, "learning_rate": 9.259394993460985e-05, "loss": 0.0342, "step": 21670 }, { "grad_norm": 0.31521037220954895, "learning_rate": 9.258528774529151e-05, "loss": 0.0375, "step": 21680 }, { "grad_norm": 0.2670820951461792, "learning_rate": 9.257662089891013e-05, "loss": 0.0387, "step": 21690 }, { "grad_norm": 0.2264767438173294, "learning_rate": 9.25679493964135e-05, "loss": 0.0362, "step": 21700 }, { "grad_norm": 0.28595471382141113, "learning_rate": 9.255927323874994e-05, "loss": 0.0426, "step": 21710 }, { "grad_norm": 0.2137816697359085, "learning_rate": 9.255059242686822e-05, "loss": 0.0394, "step": 21720 }, { "grad_norm": 0.2498025745153427, "learning_rate": 9.254190696171769e-05, "loss": 0.0352, "step": 21730 }, { "grad_norm": 0.32028427720069885, "learning_rate": 9.25332168442482e-05, "loss": 0.0363, "step": 21740 }, { "grad_norm": 0.26765313744544983, "learning_rate": 9.252452207541004e-05, "loss": 0.0335, "step": 21750 }, { "grad_norm": 0.2356111854314804, "learning_rate": 9.251582265615409e-05, "loss": 0.0309, "step": 21760 }, { "grad_norm": 0.2705496549606323, "learning_rate": 9.250711858743169e-05, "loss": 0.0329, "step": 21770 }, { "grad_norm": 0.22975927591323853, "learning_rate": 9.24984098701947e-05, "loss": 0.0349, "step": 21780 }, { "grad_norm": 0.2718352973461151, "learning_rate": 9.248969650539552e-05, "loss": 0.0288, "step": 21790 }, { "grad_norm": 0.2341827154159546, "learning_rate": 9.2480978493987e-05, "loss": 0.0353, "step": 21800 }, { "grad_norm": 0.33352726697921753, "learning_rate": 9.247225583692256e-05, "loss": 0.0363, "step": 21810 }, { "grad_norm": 0.34556683897972107, "learning_rate": 9.246352853515607e-05, "loss": 0.036, "step": 21820 }, { "grad_norm": 0.283657968044281, "learning_rate": 9.245479658964194e-05, "loss": 0.0386, "step": 21830 }, { "grad_norm": 0.2976868450641632, "learning_rate": 9.244606000133507e-05, "loss": 0.0397, "step": 21840 }, { "grad_norm": 0.30201148986816406, "learning_rate": 9.24373187711909e-05, "loss": 0.0359, "step": 21850 }, { "grad_norm": 0.26989227533340454, "learning_rate": 9.242857290016537e-05, "loss": 0.0359, "step": 21860 }, { "grad_norm": 0.27564823627471924, "learning_rate": 9.241982238921488e-05, "loss": 0.0327, "step": 21870 }, { "grad_norm": 0.2225266844034195, "learning_rate": 9.24110672392964e-05, "loss": 0.0394, "step": 21880 }, { "grad_norm": 0.2795013189315796, "learning_rate": 9.240230745136737e-05, "loss": 0.0354, "step": 21890 }, { "grad_norm": 0.30024173855781555, "learning_rate": 9.239354302638575e-05, "loss": 0.035, "step": 21900 }, { "grad_norm": 0.2497883290052414, "learning_rate": 9.238477396531e-05, "loss": 0.0337, "step": 21910 }, { "grad_norm": 0.26290860772132874, "learning_rate": 9.23760002690991e-05, "loss": 0.0343, "step": 21920 }, { "grad_norm": 0.3114670515060425, "learning_rate": 9.236722193871252e-05, "loss": 0.0355, "step": 21930 }, { "grad_norm": 0.34820273518562317, "learning_rate": 9.235843897511023e-05, "loss": 0.0344, "step": 21940 }, { "grad_norm": 0.28888824582099915, "learning_rate": 9.234965137925276e-05, "loss": 0.0418, "step": 21950 }, { "grad_norm": 0.304705947637558, "learning_rate": 9.234085915210108e-05, "loss": 0.0355, "step": 21960 }, { "grad_norm": 0.35069093108177185, "learning_rate": 9.23320622946167e-05, "loss": 0.037, "step": 21970 }, { "grad_norm": 0.34910115599632263, "learning_rate": 9.232326080776163e-05, "loss": 0.0357, "step": 21980 }, { "grad_norm": 0.2613389194011688, "learning_rate": 9.23144546924984e-05, "loss": 0.0371, "step": 21990 }, { "grad_norm": 0.28565728664398193, "learning_rate": 9.230564394979e-05, "loss": 0.0361, "step": 22000 }, { "grad_norm": 0.25683295726776123, "learning_rate": 9.22968285806e-05, "loss": 0.0317, "step": 22010 }, { "grad_norm": 0.2505638599395752, "learning_rate": 9.228800858589242e-05, "loss": 0.0316, "step": 22020 }, { "grad_norm": 0.21845358610153198, "learning_rate": 9.227918396663179e-05, "loss": 0.0323, "step": 22030 }, { "grad_norm": 0.20539800822734833, "learning_rate": 9.227035472378319e-05, "loss": 0.0338, "step": 22040 }, { "grad_norm": 0.27941977977752686, "learning_rate": 9.226152085831213e-05, "loss": 0.0338, "step": 22050 }, { "grad_norm": 0.43380770087242126, "learning_rate": 9.22526823711847e-05, "loss": 0.0399, "step": 22060 }, { "grad_norm": 0.25813889503479004, "learning_rate": 9.224383926336745e-05, "loss": 0.0296, "step": 22070 }, { "grad_norm": 0.2584339678287506, "learning_rate": 9.223499153582744e-05, "loss": 0.0325, "step": 22080 }, { "grad_norm": 0.25666868686676025, "learning_rate": 9.222613918953226e-05, "loss": 0.0337, "step": 22090 }, { "grad_norm": 0.2131892591714859, "learning_rate": 9.221728222544999e-05, "loss": 0.0377, "step": 22100 }, { "grad_norm": 0.306027352809906, "learning_rate": 9.22084206445492e-05, "loss": 0.036, "step": 22110 }, { "grad_norm": 0.24785740673542023, "learning_rate": 9.2199554447799e-05, "loss": 0.0326, "step": 22120 }, { "grad_norm": 0.3126674294471741, "learning_rate": 9.219068363616897e-05, "loss": 0.0355, "step": 22130 }, { "grad_norm": 0.25532156229019165, "learning_rate": 9.218180821062919e-05, "loss": 0.0351, "step": 22140 }, { "grad_norm": 0.23832868039608002, "learning_rate": 9.21729281721503e-05, "loss": 0.0349, "step": 22150 }, { "grad_norm": 0.27764126658439636, "learning_rate": 9.216404352170339e-05, "loss": 0.0377, "step": 22160 }, { "grad_norm": 0.33294910192489624, "learning_rate": 9.215515426026007e-05, "loss": 0.0359, "step": 22170 }, { "grad_norm": 0.2123473584651947, "learning_rate": 9.214626038879246e-05, "loss": 0.0371, "step": 22180 }, { "grad_norm": 0.2546551525592804, "learning_rate": 9.21373619082732e-05, "loss": 0.037, "step": 22190 }, { "grad_norm": 0.2699865996837616, "learning_rate": 9.212845881967535e-05, "loss": 0.0369, "step": 22200 }, { "grad_norm": 0.3066254258155823, "learning_rate": 9.211955112397262e-05, "loss": 0.0426, "step": 22210 }, { "grad_norm": 0.24140451848506927, "learning_rate": 9.211063882213909e-05, "loss": 0.0342, "step": 22220 }, { "grad_norm": 0.2976453900337219, "learning_rate": 9.210172191514942e-05, "loss": 0.0335, "step": 22230 }, { "grad_norm": 0.26949992775917053, "learning_rate": 9.209280040397874e-05, "loss": 0.0338, "step": 22240 }, { "grad_norm": 0.47797703742980957, "learning_rate": 9.208387428960268e-05, "loss": 0.0332, "step": 22250 }, { "grad_norm": 0.31075093150138855, "learning_rate": 9.20749435729974e-05, "loss": 0.0354, "step": 22260 }, { "grad_norm": 0.2668318450450897, "learning_rate": 9.206600825513957e-05, "loss": 0.0409, "step": 22270 }, { "grad_norm": 0.22842589020729065, "learning_rate": 9.20570683370063e-05, "loss": 0.034, "step": 22280 }, { "grad_norm": 0.26511019468307495, "learning_rate": 9.204812381957528e-05, "loss": 0.0414, "step": 22290 }, { "grad_norm": 0.2642975449562073, "learning_rate": 9.203917470382465e-05, "loss": 0.0367, "step": 22300 }, { "grad_norm": 0.3307682275772095, "learning_rate": 9.203022099073309e-05, "loss": 0.0309, "step": 22310 }, { "grad_norm": 0.3530751168727875, "learning_rate": 9.202126268127976e-05, "loss": 0.0317, "step": 22320 }, { "grad_norm": 0.2154947817325592, "learning_rate": 9.20122997764443e-05, "loss": 0.0335, "step": 22330 }, { "grad_norm": 0.3441922962665558, "learning_rate": 9.200333227720692e-05, "loss": 0.0389, "step": 22340 }, { "grad_norm": 0.3039611876010895, "learning_rate": 9.199436018454826e-05, "loss": 0.0436, "step": 22350 }, { "grad_norm": 0.32697707414627075, "learning_rate": 9.198538349944952e-05, "loss": 0.0346, "step": 22360 }, { "grad_norm": 0.32814815640449524, "learning_rate": 9.197640222289234e-05, "loss": 0.0336, "step": 22370 }, { "grad_norm": 0.2827543318271637, "learning_rate": 9.196741635585895e-05, "loss": 0.0347, "step": 22380 }, { "grad_norm": 0.26244139671325684, "learning_rate": 9.195842589933199e-05, "loss": 0.0338, "step": 22390 }, { "grad_norm": 0.2732551097869873, "learning_rate": 9.194943085429466e-05, "loss": 0.0395, "step": 22400 }, { "grad_norm": 0.2570260167121887, "learning_rate": 9.194043122173065e-05, "loss": 0.0426, "step": 22410 }, { "grad_norm": 0.3127552568912506, "learning_rate": 9.193142700262413e-05, "loss": 0.0339, "step": 22420 }, { "grad_norm": 0.2962881922721863, "learning_rate": 9.192241819795979e-05, "loss": 0.0345, "step": 22430 }, { "grad_norm": 0.2790929675102234, "learning_rate": 9.191340480872284e-05, "loss": 0.0363, "step": 22440 }, { "grad_norm": 0.2479327768087387, "learning_rate": 9.190438683589895e-05, "loss": 0.0365, "step": 22450 }, { "grad_norm": 0.3092208504676819, "learning_rate": 9.189536428047432e-05, "loss": 0.0359, "step": 22460 }, { "grad_norm": 0.23847514390945435, "learning_rate": 9.188633714343564e-05, "loss": 0.0361, "step": 22470 }, { "grad_norm": 0.27218663692474365, "learning_rate": 9.18773054257701e-05, "loss": 0.0349, "step": 22480 }, { "grad_norm": 0.23198990523815155, "learning_rate": 9.18682691284654e-05, "loss": 0.0325, "step": 22490 }, { "grad_norm": 0.2571239471435547, "learning_rate": 9.185922825250974e-05, "loss": 0.0387, "step": 22500 }, { "grad_norm": 0.2956089377403259, "learning_rate": 9.185018279889181e-05, "loss": 0.037, "step": 22510 }, { "grad_norm": 0.2739623486995697, "learning_rate": 9.184113276860082e-05, "loss": 0.0426, "step": 22520 }, { "grad_norm": 0.24256311357021332, "learning_rate": 9.183207816262645e-05, "loss": 0.0401, "step": 22530 }, { "grad_norm": 0.28932327032089233, "learning_rate": 9.182301898195891e-05, "loss": 0.0413, "step": 22540 }, { "grad_norm": 0.30922064185142517, "learning_rate": 9.181395522758889e-05, "loss": 0.0353, "step": 22550 }, { "grad_norm": 0.2679431736469269, "learning_rate": 9.180488690050759e-05, "loss": 0.0345, "step": 22560 }, { "grad_norm": 0.2908359467983246, "learning_rate": 9.179581400170671e-05, "loss": 0.0335, "step": 22570 }, { "grad_norm": 0.2639276385307312, "learning_rate": 9.178673653217845e-05, "loss": 0.0336, "step": 22580 }, { "grad_norm": 0.20489342510700226, "learning_rate": 9.177765449291551e-05, "loss": 0.0382, "step": 22590 }, { "grad_norm": 0.22445745766162872, "learning_rate": 9.176856788491109e-05, "loss": 0.0338, "step": 22600 }, { "grad_norm": 0.23961450159549713, "learning_rate": 9.175947670915887e-05, "loss": 0.0414, "step": 22610 }, { "grad_norm": 0.3573387563228607, "learning_rate": 9.175038096665309e-05, "loss": 0.0414, "step": 22620 }, { "grad_norm": 0.20434758067131042, "learning_rate": 9.17412806583884e-05, "loss": 0.0354, "step": 22630 }, { "grad_norm": 0.2761479914188385, "learning_rate": 9.173217578536002e-05, "loss": 0.0339, "step": 22640 }, { "grad_norm": 0.27251309156417847, "learning_rate": 9.172306634856362e-05, "loss": 0.0405, "step": 22650 }, { "grad_norm": 0.3763740658760071, "learning_rate": 9.171395234899545e-05, "loss": 0.0372, "step": 22660 }, { "grad_norm": 0.2722419798374176, "learning_rate": 9.170483378765214e-05, "loss": 0.039, "step": 22670 }, { "grad_norm": 0.3454737067222595, "learning_rate": 9.169571066553091e-05, "loss": 0.0351, "step": 22680 }, { "grad_norm": 0.31155070662498474, "learning_rate": 9.168658298362946e-05, "loss": 0.0386, "step": 22690 }, { "grad_norm": 0.30869099497795105, "learning_rate": 9.167745074294598e-05, "loss": 0.0377, "step": 22700 }, { "grad_norm": 0.39945825934410095, "learning_rate": 9.166831394447913e-05, "loss": 0.0327, "step": 22710 }, { "grad_norm": 0.3515591025352478, "learning_rate": 9.165917258922812e-05, "loss": 0.0332, "step": 22720 }, { "grad_norm": 0.24185745418071747, "learning_rate": 9.165002667819262e-05, "loss": 0.0332, "step": 22730 }, { "grad_norm": 0.31839117407798767, "learning_rate": 9.164087621237282e-05, "loss": 0.0349, "step": 22740 }, { "grad_norm": 0.2819116413593292, "learning_rate": 9.163172119276942e-05, "loss": 0.0332, "step": 22750 }, { "grad_norm": 0.3112216889858246, "learning_rate": 9.162256162038358e-05, "loss": 0.04, "step": 22760 }, { "grad_norm": 0.3896408975124359, "learning_rate": 9.161339749621698e-05, "loss": 0.0396, "step": 22770 }, { "grad_norm": 0.18830141425132751, "learning_rate": 9.160422882127177e-05, "loss": 0.0283, "step": 22780 }, { "grad_norm": 0.22482021152973175, "learning_rate": 9.159505559655069e-05, "loss": 0.0342, "step": 22790 }, { "grad_norm": 0.265217661857605, "learning_rate": 9.158587782305684e-05, "loss": 0.0311, "step": 22800 }, { "grad_norm": 0.24046801030635834, "learning_rate": 9.157669550179391e-05, "loss": 0.0326, "step": 22810 }, { "grad_norm": 0.24340517818927765, "learning_rate": 9.156750863376609e-05, "loss": 0.0367, "step": 22820 }, { "grad_norm": 0.29421666264533997, "learning_rate": 9.155831721997801e-05, "loss": 0.0351, "step": 22830 }, { "grad_norm": 0.35755184292793274, "learning_rate": 9.154912126143484e-05, "loss": 0.0368, "step": 22840 }, { "grad_norm": 0.2286825031042099, "learning_rate": 9.153992075914224e-05, "loss": 0.0385, "step": 22850 }, { "grad_norm": 0.2300160825252533, "learning_rate": 9.153071571410635e-05, "loss": 0.0328, "step": 22860 }, { "grad_norm": 0.30237236618995667, "learning_rate": 9.152150612733384e-05, "loss": 0.0373, "step": 22870 }, { "grad_norm": 0.3186761736869812, "learning_rate": 9.151229199983184e-05, "loss": 0.0355, "step": 22880 }, { "grad_norm": 0.2653765082359314, "learning_rate": 9.150307333260802e-05, "loss": 0.0309, "step": 22890 }, { "grad_norm": 0.2923889458179474, "learning_rate": 9.149385012667048e-05, "loss": 0.033, "step": 22900 }, { "grad_norm": 0.2511974275112152, "learning_rate": 9.148462238302788e-05, "loss": 0.0386, "step": 22910 }, { "grad_norm": 0.41491958498954773, "learning_rate": 9.147539010268936e-05, "loss": 0.0346, "step": 22920 }, { "grad_norm": 0.38327062129974365, "learning_rate": 9.14661532866645e-05, "loss": 0.0346, "step": 22930 }, { "grad_norm": 0.25672855973243713, "learning_rate": 9.145691193596348e-05, "loss": 0.0377, "step": 22940 }, { "grad_norm": 0.2670483887195587, "learning_rate": 9.144766605159691e-05, "loss": 0.0366, "step": 22950 }, { "grad_norm": 0.3618052303791046, "learning_rate": 9.14384156345759e-05, "loss": 0.0357, "step": 22960 }, { "grad_norm": 0.27526047825813293, "learning_rate": 9.142916068591204e-05, "loss": 0.0349, "step": 22970 }, { "grad_norm": 0.28849077224731445, "learning_rate": 9.141990120661746e-05, "loss": 0.0348, "step": 22980 }, { "grad_norm": 0.28418347239494324, "learning_rate": 9.141063719770475e-05, "loss": 0.0442, "step": 22990 }, { "grad_norm": 0.2646365463733673, "learning_rate": 9.140136866018704e-05, "loss": 0.0316, "step": 23000 }, { "grad_norm": 0.26891228556632996, "learning_rate": 9.139209559507788e-05, "loss": 0.0325, "step": 23010 }, { "grad_norm": 0.24869142472743988, "learning_rate": 9.13828180033914e-05, "loss": 0.0334, "step": 23020 }, { "grad_norm": 0.269257515668869, "learning_rate": 9.137353588614212e-05, "loss": 0.0353, "step": 23030 }, { "grad_norm": 0.2394108921289444, "learning_rate": 9.136424924434519e-05, "loss": 0.0359, "step": 23040 }, { "grad_norm": 0.28257259726524353, "learning_rate": 9.135495807901615e-05, "loss": 0.0333, "step": 23050 }, { "grad_norm": 0.320427268743515, "learning_rate": 9.134566239117108e-05, "loss": 0.0368, "step": 23060 }, { "grad_norm": 0.2724214196205139, "learning_rate": 9.13363621818265e-05, "loss": 0.0342, "step": 23070 }, { "grad_norm": 0.2677893340587616, "learning_rate": 9.132705745199953e-05, "loss": 0.0381, "step": 23080 }, { "grad_norm": 0.3233020007610321, "learning_rate": 9.131774820270768e-05, "loss": 0.0297, "step": 23090 }, { "grad_norm": 0.26442936062812805, "learning_rate": 9.130843443496901e-05, "loss": 0.0321, "step": 23100 }, { "grad_norm": 0.3098238706588745, "learning_rate": 9.129911614980206e-05, "loss": 0.0383, "step": 23110 }, { "grad_norm": 0.27993258833885193, "learning_rate": 9.128979334822584e-05, "loss": 0.0341, "step": 23120 }, { "grad_norm": 0.28144899010658264, "learning_rate": 9.128046603125992e-05, "loss": 0.0383, "step": 23130 }, { "grad_norm": 0.34336918592453003, "learning_rate": 9.12711341999243e-05, "loss": 0.0381, "step": 23140 }, { "grad_norm": 0.26491203904151917, "learning_rate": 9.12617978552395e-05, "loss": 0.0334, "step": 23150 }, { "grad_norm": 0.30980920791625977, "learning_rate": 9.12524569982265e-05, "loss": 0.0355, "step": 23160 }, { "grad_norm": 0.22382552921772003, "learning_rate": 9.124311162990684e-05, "loss": 0.0315, "step": 23170 }, { "grad_norm": 0.2735011577606201, "learning_rate": 9.12337617513025e-05, "loss": 0.0334, "step": 23180 }, { "grad_norm": 0.2917586863040924, "learning_rate": 9.122440736343596e-05, "loss": 0.0294, "step": 23190 }, { "grad_norm": 0.31822511553764343, "learning_rate": 9.12150484673302e-05, "loss": 0.0348, "step": 23200 }, { "grad_norm": 0.2975533604621887, "learning_rate": 9.120568506400873e-05, "loss": 0.0377, "step": 23210 }, { "grad_norm": 0.31776732206344604, "learning_rate": 9.119631715449548e-05, "loss": 0.0352, "step": 23220 }, { "grad_norm": 0.2923530042171478, "learning_rate": 9.118694473981493e-05, "loss": 0.0389, "step": 23230 }, { "grad_norm": 0.27748772501945496, "learning_rate": 9.117756782099203e-05, "loss": 0.041, "step": 23240 }, { "grad_norm": 0.2245292067527771, "learning_rate": 9.11681863990522e-05, "loss": 0.0345, "step": 23250 }, { "grad_norm": 0.29065608978271484, "learning_rate": 9.115880047502142e-05, "loss": 0.0326, "step": 23260 }, { "grad_norm": 0.25329673290252686, "learning_rate": 9.114941004992609e-05, "loss": 0.0341, "step": 23270 }, { "grad_norm": 0.2591859996318817, "learning_rate": 9.114001512479317e-05, "loss": 0.0375, "step": 23280 }, { "grad_norm": 0.2627127170562744, "learning_rate": 9.113061570065003e-05, "loss": 0.0356, "step": 23290 }, { "grad_norm": 0.3747127652168274, "learning_rate": 9.112121177852459e-05, "loss": 0.0413, "step": 23300 }, { "grad_norm": 0.2435312569141388, "learning_rate": 9.111180335944527e-05, "loss": 0.033, "step": 23310 }, { "grad_norm": 0.2308550626039505, "learning_rate": 9.110239044444093e-05, "loss": 0.0311, "step": 23320 }, { "grad_norm": 0.2129334807395935, "learning_rate": 9.109297303454099e-05, "loss": 0.0317, "step": 23330 }, { "grad_norm": 0.2772226929664612, "learning_rate": 9.108355113077526e-05, "loss": 0.0341, "step": 23340 }, { "grad_norm": 0.3268973231315613, "learning_rate": 9.107412473417419e-05, "loss": 0.0314, "step": 23350 }, { "grad_norm": 0.3039814829826355, "learning_rate": 9.106469384576858e-05, "loss": 0.0351, "step": 23360 }, { "grad_norm": 0.37195059657096863, "learning_rate": 9.105525846658978e-05, "loss": 0.0341, "step": 23370 }, { "grad_norm": 0.25411906838417053, "learning_rate": 9.104581859766965e-05, "loss": 0.0394, "step": 23380 }, { "grad_norm": 0.27633166313171387, "learning_rate": 9.10363742400405e-05, "loss": 0.0349, "step": 23390 }, { "grad_norm": 0.2657143175601959, "learning_rate": 9.102692539473518e-05, "loss": 0.0346, "step": 23400 }, { "grad_norm": 0.30164483189582825, "learning_rate": 9.101747206278697e-05, "loss": 0.037, "step": 23410 }, { "grad_norm": 0.31344926357269287, "learning_rate": 9.100801424522968e-05, "loss": 0.0428, "step": 23420 }, { "grad_norm": 0.2850293815135956, "learning_rate": 9.099855194309762e-05, "loss": 0.0417, "step": 23430 }, { "grad_norm": 0.35793089866638184, "learning_rate": 9.098908515742554e-05, "loss": 0.0408, "step": 23440 }, { "grad_norm": 0.22372029721736908, "learning_rate": 9.097961388924873e-05, "loss": 0.0347, "step": 23450 }, { "grad_norm": 0.36788901686668396, "learning_rate": 9.097013813960298e-05, "loss": 0.0343, "step": 23460 }, { "grad_norm": 0.2393564134836197, "learning_rate": 9.09606579095245e-05, "loss": 0.0346, "step": 23470 }, { "grad_norm": 0.307892769575119, "learning_rate": 9.095117320005008e-05, "loss": 0.0372, "step": 23480 }, { "grad_norm": 0.30827075242996216, "learning_rate": 9.094168401221691e-05, "loss": 0.0352, "step": 23490 }, { "grad_norm": 0.3123425543308258, "learning_rate": 9.093219034706273e-05, "loss": 0.035, "step": 23500 }, { "grad_norm": 0.21852855384349823, "learning_rate": 9.092269220562577e-05, "loss": 0.034, "step": 23510 }, { "grad_norm": 0.30952882766723633, "learning_rate": 9.09131895889447e-05, "loss": 0.036, "step": 23520 }, { "grad_norm": 0.2956249415874481, "learning_rate": 9.090368249805873e-05, "loss": 0.0341, "step": 23530 }, { "grad_norm": 0.32586097717285156, "learning_rate": 9.089417093400754e-05, "loss": 0.0351, "step": 23540 }, { "grad_norm": 0.28997546434402466, "learning_rate": 9.088465489783131e-05, "loss": 0.0378, "step": 23550 }, { "grad_norm": 0.26156771183013916, "learning_rate": 9.087513439057068e-05, "loss": 0.0425, "step": 23560 }, { "grad_norm": 0.2535228133201599, "learning_rate": 9.08656094132668e-05, "loss": 0.0362, "step": 23570 }, { "grad_norm": 0.4779589772224426, "learning_rate": 9.085607996696134e-05, "loss": 0.0337, "step": 23580 }, { "grad_norm": 0.3786522448062897, "learning_rate": 9.084654605269639e-05, "loss": 0.0399, "step": 23590 }, { "grad_norm": 0.27496492862701416, "learning_rate": 9.083700767151457e-05, "loss": 0.0414, "step": 23600 }, { "grad_norm": 0.3536457121372223, "learning_rate": 9.082746482445898e-05, "loss": 0.0352, "step": 23610 }, { "grad_norm": 0.3218317925930023, "learning_rate": 9.081791751257325e-05, "loss": 0.0329, "step": 23620 }, { "grad_norm": 0.32189491391181946, "learning_rate": 9.080836573690142e-05, "loss": 0.0375, "step": 23630 }, { "grad_norm": 0.272858202457428, "learning_rate": 9.079880949848805e-05, "loss": 0.0364, "step": 23640 }, { "grad_norm": 0.3884052634239197, "learning_rate": 9.078924879837822e-05, "loss": 0.0345, "step": 23650 }, { "grad_norm": 0.24668669700622559, "learning_rate": 9.077968363761747e-05, "loss": 0.0351, "step": 23660 }, { "grad_norm": 0.2630050778388977, "learning_rate": 9.077011401725182e-05, "loss": 0.0336, "step": 23670 }, { "grad_norm": 0.30954673886299133, "learning_rate": 9.07605399383278e-05, "loss": 0.039, "step": 23680 }, { "grad_norm": 0.34846678376197815, "learning_rate": 9.075096140189243e-05, "loss": 0.0412, "step": 23690 }, { "grad_norm": 0.22040583193302155, "learning_rate": 9.074137840899318e-05, "loss": 0.0286, "step": 23700 }, { "grad_norm": 0.3098297715187073, "learning_rate": 9.073179096067804e-05, "loss": 0.0353, "step": 23710 }, { "grad_norm": 0.25621360540390015, "learning_rate": 9.072219905799549e-05, "loss": 0.0295, "step": 23720 }, { "grad_norm": 0.23707634210586548, "learning_rate": 9.071260270199447e-05, "loss": 0.0379, "step": 23730 }, { "grad_norm": 0.27912378311157227, "learning_rate": 9.070300189372441e-05, "loss": 0.0349, "step": 23740 }, { "grad_norm": 0.24953079223632812, "learning_rate": 9.069339663423528e-05, "loss": 0.0336, "step": 23750 }, { "grad_norm": 0.3451824188232422, "learning_rate": 9.068378692457747e-05, "loss": 0.0354, "step": 23760 }, { "grad_norm": 0.30327749252319336, "learning_rate": 9.067417276580189e-05, "loss": 0.0349, "step": 23770 }, { "grad_norm": 0.2932184934616089, "learning_rate": 9.066455415895993e-05, "loss": 0.0339, "step": 23780 }, { "grad_norm": 0.24632199108600616, "learning_rate": 9.065493110510346e-05, "loss": 0.0376, "step": 23790 }, { "grad_norm": 0.2715499699115753, "learning_rate": 9.064530360528484e-05, "loss": 0.0463, "step": 23800 }, { "grad_norm": 0.3019683063030243, "learning_rate": 9.063567166055695e-05, "loss": 0.0323, "step": 23810 }, { "grad_norm": 0.2492169588804245, "learning_rate": 9.062603527197308e-05, "loss": 0.0339, "step": 23820 }, { "grad_norm": 0.2520684599876404, "learning_rate": 9.06163944405871e-05, "loss": 0.034, "step": 23830 }, { "grad_norm": 0.23803949356079102, "learning_rate": 9.060674916745327e-05, "loss": 0.0347, "step": 23840 }, { "grad_norm": 0.29486632347106934, "learning_rate": 9.05970994536264e-05, "loss": 0.0332, "step": 23850 }, { "grad_norm": 0.3106437623500824, "learning_rate": 9.05874453001618e-05, "loss": 0.0367, "step": 23860 }, { "grad_norm": 0.28941410779953003, "learning_rate": 9.057778670811517e-05, "loss": 0.0316, "step": 23870 }, { "grad_norm": 0.25561991333961487, "learning_rate": 9.056812367854281e-05, "loss": 0.0341, "step": 23880 }, { "grad_norm": 0.3032684624195099, "learning_rate": 9.055845621250143e-05, "loss": 0.0415, "step": 23890 }, { "grad_norm": 0.21105246245861053, "learning_rate": 9.054878431104825e-05, "loss": 0.0352, "step": 23900 }, { "grad_norm": 0.28677383065223694, "learning_rate": 9.0539107975241e-05, "loss": 0.0393, "step": 23910 }, { "grad_norm": 0.3371962904930115, "learning_rate": 9.052942720613784e-05, "loss": 0.0347, "step": 23920 }, { "grad_norm": 0.29104480147361755, "learning_rate": 9.051974200479745e-05, "loss": 0.0383, "step": 23930 }, { "grad_norm": 0.31267669796943665, "learning_rate": 9.051005237227901e-05, "loss": 0.0355, "step": 23940 }, { "grad_norm": 0.26361799240112305, "learning_rate": 9.050035830964215e-05, "loss": 0.0322, "step": 23950 }, { "grad_norm": 0.29592496156692505, "learning_rate": 9.049065981794698e-05, "loss": 0.0403, "step": 23960 }, { "grad_norm": 0.26580557227134705, "learning_rate": 9.048095689825414e-05, "loss": 0.0348, "step": 23970 }, { "grad_norm": 0.3206099271774292, "learning_rate": 9.047124955162472e-05, "loss": 0.0369, "step": 23980 }, { "grad_norm": 0.2813839912414551, "learning_rate": 9.046153777912028e-05, "loss": 0.0382, "step": 23990 }, { "grad_norm": 0.27863243222236633, "learning_rate": 9.045182158180292e-05, "loss": 0.0377, "step": 24000 }, { "grad_norm": 0.264426052570343, "learning_rate": 9.044210096073516e-05, "loss": 0.0335, "step": 24010 }, { "grad_norm": 0.2538631558418274, "learning_rate": 9.043237591698004e-05, "loss": 0.0361, "step": 24020 }, { "grad_norm": 0.20141449570655823, "learning_rate": 9.04226464516011e-05, "loss": 0.0332, "step": 24030 }, { "grad_norm": 0.3050949275493622, "learning_rate": 9.041291256566229e-05, "loss": 0.0384, "step": 24040 }, { "grad_norm": 0.2631395757198334, "learning_rate": 9.040317426022814e-05, "loss": 0.0369, "step": 24050 }, { "grad_norm": 0.26693329215049744, "learning_rate": 9.03934315363636e-05, "loss": 0.0374, "step": 24060 }, { "grad_norm": 0.35140401124954224, "learning_rate": 9.038368439513409e-05, "loss": 0.0385, "step": 24070 }, { "grad_norm": 0.25243768095970154, "learning_rate": 9.03739328376056e-05, "loss": 0.0369, "step": 24080 }, { "grad_norm": 0.27145811915397644, "learning_rate": 9.036417686484451e-05, "loss": 0.0374, "step": 24090 }, { "grad_norm": 0.3657509684562683, "learning_rate": 9.035441647791773e-05, "loss": 0.0352, "step": 24100 }, { "grad_norm": 0.29196250438690186, "learning_rate": 9.034465167789263e-05, "loss": 0.0381, "step": 24110 }, { "grad_norm": 0.37270668148994446, "learning_rate": 9.033488246583706e-05, "loss": 0.0371, "step": 24120 }, { "grad_norm": 0.2142513394355774, "learning_rate": 9.032510884281941e-05, "loss": 0.0353, "step": 24130 }, { "grad_norm": 0.3128388226032257, "learning_rate": 9.031533080990848e-05, "loss": 0.037, "step": 24140 }, { "grad_norm": 0.2744039297103882, "learning_rate": 9.030554836817358e-05, "loss": 0.0317, "step": 24150 }, { "grad_norm": 0.3076283037662506, "learning_rate": 9.029576151868451e-05, "loss": 0.0363, "step": 24160 }, { "grad_norm": 0.30475378036499023, "learning_rate": 9.028597026251155e-05, "loss": 0.0337, "step": 24170 }, { "grad_norm": 0.23603107035160065, "learning_rate": 9.027617460072547e-05, "loss": 0.0356, "step": 24180 }, { "grad_norm": 0.2509973645210266, "learning_rate": 9.026637453439745e-05, "loss": 0.0401, "step": 24190 }, { "grad_norm": 0.2672114074230194, "learning_rate": 9.025657006459927e-05, "loss": 0.0335, "step": 24200 }, { "grad_norm": 0.24588194489479065, "learning_rate": 9.024676119240311e-05, "loss": 0.04, "step": 24210 }, { "grad_norm": 0.23513205349445343, "learning_rate": 9.023694791888166e-05, "loss": 0.0339, "step": 24220 }, { "grad_norm": 0.29455694556236267, "learning_rate": 9.022713024510808e-05, "loss": 0.0373, "step": 24230 }, { "grad_norm": 0.2838117182254791, "learning_rate": 9.021730817215601e-05, "loss": 0.0453, "step": 24240 }, { "grad_norm": 0.20912796258926392, "learning_rate": 9.02074817010996e-05, "loss": 0.0396, "step": 24250 }, { "grad_norm": 0.23490673303604126, "learning_rate": 9.019765083301342e-05, "loss": 0.0326, "step": 24260 }, { "grad_norm": 0.3284546136856079, "learning_rate": 9.01878155689726e-05, "loss": 0.0357, "step": 24270 }, { "grad_norm": 0.48297858238220215, "learning_rate": 9.017797591005268e-05, "loss": 0.0358, "step": 24280 }, { "grad_norm": 0.3384830951690674, "learning_rate": 9.016813185732972e-05, "loss": 0.0388, "step": 24290 }, { "grad_norm": 0.2906341552734375, "learning_rate": 9.015828341188027e-05, "loss": 0.0378, "step": 24300 }, { "grad_norm": 0.2938195466995239, "learning_rate": 9.01484305747813e-05, "loss": 0.0433, "step": 24310 }, { "grad_norm": 0.3156253397464752, "learning_rate": 9.013857334711033e-05, "loss": 0.0353, "step": 24320 }, { "grad_norm": 0.2746508717536926, "learning_rate": 9.012871172994534e-05, "loss": 0.036, "step": 24330 }, { "grad_norm": 0.2473682314157486, "learning_rate": 9.011884572436476e-05, "loss": 0.0394, "step": 24340 }, { "grad_norm": 0.28698304295539856, "learning_rate": 9.010897533144754e-05, "loss": 0.0289, "step": 24350 }, { "grad_norm": 0.3446674346923828, "learning_rate": 9.009910055227306e-05, "loss": 0.0323, "step": 24360 }, { "grad_norm": 0.3887897729873657, "learning_rate": 9.008922138792124e-05, "loss": 0.0329, "step": 24370 }, { "grad_norm": 0.2042372077703476, "learning_rate": 9.007933783947244e-05, "loss": 0.0336, "step": 24380 }, { "grad_norm": 0.24410726130008698, "learning_rate": 9.006944990800752e-05, "loss": 0.0387, "step": 24390 }, { "grad_norm": 0.3133048415184021, "learning_rate": 9.005955759460779e-05, "loss": 0.0347, "step": 24400 }, { "grad_norm": 0.2953587770462036, "learning_rate": 9.004966090035508e-05, "loss": 0.0337, "step": 24410 }, { "grad_norm": 0.2552272081375122, "learning_rate": 9.003975982633166e-05, "loss": 0.0319, "step": 24420 }, { "grad_norm": 0.3006243109703064, "learning_rate": 9.00298543736203e-05, "loss": 0.0334, "step": 24430 }, { "grad_norm": 0.3001505732536316, "learning_rate": 9.001994454330427e-05, "loss": 0.0358, "step": 24440 }, { "grad_norm": 0.4049713909626007, "learning_rate": 9.001003033646727e-05, "loss": 0.0433, "step": 24450 }, { "grad_norm": 0.31313666701316833, "learning_rate": 9.00001117541935e-05, "loss": 0.0327, "step": 24460 }, { "grad_norm": 0.3214746117591858, "learning_rate": 8.999018879756764e-05, "loss": 0.0316, "step": 24470 }, { "grad_norm": 0.25412362813949585, "learning_rate": 8.998026146767487e-05, "loss": 0.0407, "step": 24480 }, { "grad_norm": 0.29660555720329285, "learning_rate": 8.99703297656008e-05, "loss": 0.0356, "step": 24490 }, { "grad_norm": 0.2454189956188202, "learning_rate": 8.996039369243156e-05, "loss": 0.0397, "step": 24500 }, { "grad_norm": 0.30438271164894104, "learning_rate": 8.995045324925378e-05, "loss": 0.0327, "step": 24510 }, { "grad_norm": 0.24706289172172546, "learning_rate": 8.994050843715448e-05, "loss": 0.0327, "step": 24520 }, { "grad_norm": 0.43221360445022583, "learning_rate": 8.993055925722121e-05, "loss": 0.0418, "step": 24530 }, { "grad_norm": 0.3171767294406891, "learning_rate": 8.992060571054202e-05, "loss": 0.0383, "step": 24540 }, { "grad_norm": 0.31114661693573, "learning_rate": 8.991064779820542e-05, "loss": 0.0335, "step": 24550 }, { "grad_norm": 0.286291241645813, "learning_rate": 8.990068552130036e-05, "loss": 0.0355, "step": 24560 }, { "grad_norm": 0.2985469698905945, "learning_rate": 8.989071888091634e-05, "loss": 0.0353, "step": 24570 }, { "grad_norm": 0.3095471262931824, "learning_rate": 8.988074787814329e-05, "loss": 0.0336, "step": 24580 }, { "grad_norm": 0.34766149520874023, "learning_rate": 8.987077251407158e-05, "loss": 0.0356, "step": 24590 }, { "grad_norm": 0.20875360071659088, "learning_rate": 8.986079278979216e-05, "loss": 0.0319, "step": 24600 }, { "grad_norm": 0.262285441160202, "learning_rate": 8.985080870639635e-05, "loss": 0.0362, "step": 24610 }, { "grad_norm": 0.27511775493621826, "learning_rate": 8.984082026497603e-05, "loss": 0.0354, "step": 24620 }, { "grad_norm": 0.24613478779792786, "learning_rate": 8.98308274666235e-05, "loss": 0.0338, "step": 24630 }, { "grad_norm": 0.2817857563495636, "learning_rate": 8.982083031243155e-05, "loss": 0.0325, "step": 24640 }, { "grad_norm": 0.31656089425086975, "learning_rate": 8.98108288034935e-05, "loss": 0.0359, "step": 24650 }, { "grad_norm": 0.29334592819213867, "learning_rate": 8.980082294090305e-05, "loss": 0.0361, "step": 24660 }, { "grad_norm": 0.20554625988006592, "learning_rate": 8.979081272575443e-05, "loss": 0.0343, "step": 24670 }, { "grad_norm": 0.3712047338485718, "learning_rate": 8.978079815914236e-05, "loss": 0.0403, "step": 24680 }, { "grad_norm": 0.21351931989192963, "learning_rate": 8.977077924216202e-05, "loss": 0.0322, "step": 24690 }, { "grad_norm": 0.26299431920051575, "learning_rate": 8.976075597590905e-05, "loss": 0.0384, "step": 24700 }, { "grad_norm": 0.2845509350299835, "learning_rate": 8.975072836147958e-05, "loss": 0.0375, "step": 24710 }, { "grad_norm": 0.24246746301651, "learning_rate": 8.974069639997025e-05, "loss": 0.0412, "step": 24720 }, { "grad_norm": 0.33528852462768555, "learning_rate": 8.973066009247808e-05, "loss": 0.0353, "step": 24730 }, { "grad_norm": 0.2853644788265228, "learning_rate": 8.972061944010066e-05, "loss": 0.0377, "step": 24740 }, { "grad_norm": 0.2705329656600952, "learning_rate": 8.971057444393603e-05, "loss": 0.0354, "step": 24750 }, { "grad_norm": 0.22422903776168823, "learning_rate": 8.970052510508268e-05, "loss": 0.0357, "step": 24760 }, { "grad_norm": 0.29512351751327515, "learning_rate": 8.969047142463959e-05, "loss": 0.0308, "step": 24770 }, { "grad_norm": 0.32393455505371094, "learning_rate": 8.968041340370621e-05, "loss": 0.0381, "step": 24780 }, { "grad_norm": 0.3284186124801636, "learning_rate": 8.96703510433825e-05, "loss": 0.0374, "step": 24790 }, { "grad_norm": 0.3172527551651001, "learning_rate": 8.966028434476883e-05, "loss": 0.0354, "step": 24800 }, { "grad_norm": 0.21667179465293884, "learning_rate": 8.96502133089661e-05, "loss": 0.0311, "step": 24810 }, { "grad_norm": 0.26743096113204956, "learning_rate": 8.964013793707564e-05, "loss": 0.031, "step": 24820 }, { "grad_norm": 0.29261189699172974, "learning_rate": 8.963005823019932e-05, "loss": 0.0317, "step": 24830 }, { "grad_norm": 0.30072927474975586, "learning_rate": 8.961997418943939e-05, "loss": 0.0411, "step": 24840 }, { "grad_norm": 0.334221750497818, "learning_rate": 8.960988581589865e-05, "loss": 0.0312, "step": 24850 }, { "grad_norm": 0.28904053568840027, "learning_rate": 8.959979311068037e-05, "loss": 0.0321, "step": 24860 }, { "grad_norm": 0.25058063864707947, "learning_rate": 8.958969607488823e-05, "loss": 0.0325, "step": 24870 }, { "grad_norm": 0.26205626130104065, "learning_rate": 8.957959470962647e-05, "loss": 0.036, "step": 24880 }, { "grad_norm": 0.2722271978855133, "learning_rate": 8.956948901599971e-05, "loss": 0.0377, "step": 24890 }, { "grad_norm": 0.2163158804178238, "learning_rate": 8.955937899511315e-05, "loss": 0.0424, "step": 24900 }, { "grad_norm": 0.252270370721817, "learning_rate": 8.954926464807238e-05, "loss": 0.0321, "step": 24910 }, { "grad_norm": 0.30898332595825195, "learning_rate": 8.953914597598347e-05, "loss": 0.0351, "step": 24920 }, { "grad_norm": 0.2730039954185486, "learning_rate": 8.952902297995303e-05, "loss": 0.038, "step": 24930 }, { "grad_norm": 0.2634832262992859, "learning_rate": 8.951889566108804e-05, "loss": 0.0378, "step": 24940 }, { "grad_norm": 0.20318573713302612, "learning_rate": 8.950876402049606e-05, "loss": 0.0288, "step": 24950 }, { "grad_norm": 0.34180110692977905, "learning_rate": 8.949862805928504e-05, "loss": 0.0306, "step": 24960 }, { "grad_norm": 0.26392561197280884, "learning_rate": 8.948848777856343e-05, "loss": 0.0315, "step": 24970 }, { "grad_norm": 0.2348897010087967, "learning_rate": 8.947834317944017e-05, "loss": 0.0317, "step": 24980 }, { "grad_norm": 0.2726424038410187, "learning_rate": 8.946819426302466e-05, "loss": 0.0323, "step": 24990 }, { "grad_norm": 0.25079241394996643, "learning_rate": 8.945804103042676e-05, "loss": 0.029, "step": 25000 }, { "grad_norm": 0.3583299219608307, "learning_rate": 8.944788348275681e-05, "loss": 0.0339, "step": 25010 }, { "grad_norm": 0.5888344645500183, "learning_rate": 8.943772162112565e-05, "loss": 0.0343, "step": 25020 }, { "grad_norm": 0.2684769034385681, "learning_rate": 8.942755544664454e-05, "loss": 0.0325, "step": 25030 }, { "grad_norm": 0.2882938086986542, "learning_rate": 8.941738496042525e-05, "loss": 0.0373, "step": 25040 }, { "grad_norm": 0.24129068851470947, "learning_rate": 8.940721016357999e-05, "loss": 0.0367, "step": 25050 }, { "grad_norm": 0.30117267370224, "learning_rate": 8.939703105722148e-05, "loss": 0.0388, "step": 25060 }, { "grad_norm": 0.29586532711982727, "learning_rate": 8.93868476424629e-05, "loss": 0.0346, "step": 25070 }, { "grad_norm": 0.24778209626674652, "learning_rate": 8.937665992041786e-05, "loss": 0.0319, "step": 25080 }, { "grad_norm": 0.31122323870658875, "learning_rate": 8.93664678922005e-05, "loss": 0.0355, "step": 25090 }, { "grad_norm": 0.2572293281555176, "learning_rate": 8.93562715589254e-05, "loss": 0.0309, "step": 25100 }, { "grad_norm": 0.3153258264064789, "learning_rate": 8.934607092170762e-05, "loss": 0.0277, "step": 25110 }, { "grad_norm": 0.24745526909828186, "learning_rate": 8.933586598166266e-05, "loss": 0.0325, "step": 25120 }, { "grad_norm": 0.40596508979797363, "learning_rate": 8.932565673990655e-05, "loss": 0.0319, "step": 25130 }, { "grad_norm": 0.30728113651275635, "learning_rate": 8.931544319755574e-05, "loss": 0.0322, "step": 25140 }, { "grad_norm": 0.2743571698665619, "learning_rate": 8.930522535572718e-05, "loss": 0.0323, "step": 25150 }, { "grad_norm": 0.29707974195480347, "learning_rate": 8.929500321553826e-05, "loss": 0.0332, "step": 25160 }, { "grad_norm": 0.23782987892627716, "learning_rate": 8.928477677810686e-05, "loss": 0.0375, "step": 25170 }, { "grad_norm": 0.17859289050102234, "learning_rate": 8.927454604455137e-05, "loss": 0.0311, "step": 25180 }, { "grad_norm": 0.2847324013710022, "learning_rate": 8.926431101599053e-05, "loss": 0.0328, "step": 25190 }, { "grad_norm": 0.2668842673301697, "learning_rate": 8.925407169354369e-05, "loss": 0.0347, "step": 25200 }, { "grad_norm": 0.29303616285324097, "learning_rate": 8.92438280783306e-05, "loss": 0.0287, "step": 25210 }, { "grad_norm": 0.3048388659954071, "learning_rate": 8.923358017147146e-05, "loss": 0.0353, "step": 25220 }, { "grad_norm": 0.2178298681974411, "learning_rate": 8.922332797408697e-05, "loss": 0.0357, "step": 25230 }, { "grad_norm": 0.2857121527194977, "learning_rate": 8.921307148729831e-05, "loss": 0.0306, "step": 25240 }, { "grad_norm": 0.2049432098865509, "learning_rate": 8.920281071222712e-05, "loss": 0.0339, "step": 25250 }, { "grad_norm": 0.24789968132972717, "learning_rate": 8.919254564999548e-05, "loss": 0.0316, "step": 25260 }, { "grad_norm": 0.30238577723503113, "learning_rate": 8.918227630172598e-05, "loss": 0.0318, "step": 25270 }, { "grad_norm": 0.25432220101356506, "learning_rate": 8.917200266854165e-05, "loss": 0.0348, "step": 25280 }, { "grad_norm": 0.2756444215774536, "learning_rate": 8.9161724751566e-05, "loss": 0.033, "step": 25290 }, { "grad_norm": 0.2648950517177582, "learning_rate": 8.915144255192302e-05, "loss": 0.0297, "step": 25300 }, { "grad_norm": 0.29895204305648804, "learning_rate": 8.914115607073714e-05, "loss": 0.0317, "step": 25310 }, { "grad_norm": 0.3282690644264221, "learning_rate": 8.913086530913327e-05, "loss": 0.0312, "step": 25320 }, { "grad_norm": 0.22871585190296173, "learning_rate": 8.912057026823681e-05, "loss": 0.0336, "step": 25330 }, { "grad_norm": 0.26647499203681946, "learning_rate": 8.91102709491736e-05, "loss": 0.0316, "step": 25340 }, { "grad_norm": 0.43441861867904663, "learning_rate": 8.909996735306996e-05, "loss": 0.0336, "step": 25350 }, { "grad_norm": 0.31663477420806885, "learning_rate": 8.908965948105268e-05, "loss": 0.0365, "step": 25360 }, { "grad_norm": 0.3030276894569397, "learning_rate": 8.907934733424901e-05, "loss": 0.0352, "step": 25370 }, { "grad_norm": 0.21628262102603912, "learning_rate": 8.906903091378666e-05, "loss": 0.033, "step": 25380 }, { "grad_norm": 0.33389517664909363, "learning_rate": 8.905871022079384e-05, "loss": 0.0347, "step": 25390 }, { "grad_norm": 0.19872231781482697, "learning_rate": 8.90483852563992e-05, "loss": 0.0297, "step": 25400 }, { "grad_norm": 0.2994627356529236, "learning_rate": 8.903805602173185e-05, "loss": 0.0306, "step": 25410 }, { "grad_norm": 0.22480231523513794, "learning_rate": 8.902772251792137e-05, "loss": 0.0348, "step": 25420 }, { "grad_norm": 0.30225783586502075, "learning_rate": 8.901738474609786e-05, "loss": 0.0369, "step": 25430 }, { "grad_norm": 0.2349379062652588, "learning_rate": 8.900704270739179e-05, "loss": 0.0316, "step": 25440 }, { "grad_norm": 0.28212064504623413, "learning_rate": 8.89966964029342e-05, "loss": 0.0305, "step": 25450 }, { "grad_norm": 0.2876518964767456, "learning_rate": 8.898634583385652e-05, "loss": 0.0298, "step": 25460 }, { "grad_norm": 0.24893514811992645, "learning_rate": 8.897599100129065e-05, "loss": 0.031, "step": 25470 }, { "grad_norm": 0.23587782680988312, "learning_rate": 8.896563190636903e-05, "loss": 0.0308, "step": 25480 }, { "grad_norm": 0.20382951200008392, "learning_rate": 8.895526855022448e-05, "loss": 0.0313, "step": 25490 }, { "grad_norm": 0.22162111103534698, "learning_rate": 8.894490093399033e-05, "loss": 0.0315, "step": 25500 }, { "grad_norm": 0.2777242064476013, "learning_rate": 8.893452905880035e-05, "loss": 0.0323, "step": 25510 }, { "grad_norm": 0.21091394126415253, "learning_rate": 8.892415292578883e-05, "loss": 0.0299, "step": 25520 }, { "grad_norm": 0.35530972480773926, "learning_rate": 8.891377253609046e-05, "loss": 0.035, "step": 25530 }, { "grad_norm": 0.3145139217376709, "learning_rate": 8.890338789084043e-05, "loss": 0.0331, "step": 25540 }, { "grad_norm": 0.37156251072883606, "learning_rate": 8.88929989911744e-05, "loss": 0.034, "step": 25550 }, { "grad_norm": 0.32880330085754395, "learning_rate": 8.888260583822847e-05, "loss": 0.0365, "step": 25560 }, { "grad_norm": 0.2030463069677353, "learning_rate": 8.887220843313921e-05, "loss": 0.0289, "step": 25570 }, { "grad_norm": 0.22400709986686707, "learning_rate": 8.88618067770437e-05, "loss": 0.027, "step": 25580 }, { "grad_norm": 0.23655134439468384, "learning_rate": 8.885140087107942e-05, "loss": 0.0317, "step": 25590 }, { "grad_norm": 0.23974457383155823, "learning_rate": 8.884099071638436e-05, "loss": 0.0311, "step": 25600 }, { "grad_norm": 0.25007757544517517, "learning_rate": 8.883057631409695e-05, "loss": 0.0292, "step": 25610 }, { "grad_norm": 0.24051885306835175, "learning_rate": 8.882015766535608e-05, "loss": 0.0309, "step": 25620 }, { "grad_norm": 0.2879301905632019, "learning_rate": 8.880973477130115e-05, "loss": 0.0336, "step": 25630 }, { "grad_norm": 0.28109124302864075, "learning_rate": 8.879930763307197e-05, "loss": 0.038, "step": 25640 }, { "grad_norm": 0.2935623228549957, "learning_rate": 8.878887625180884e-05, "loss": 0.0294, "step": 25650 }, { "grad_norm": 0.2838490903377533, "learning_rate": 8.877844062865253e-05, "loss": 0.0361, "step": 25660 }, { "grad_norm": 0.2718854248523712, "learning_rate": 8.876800076474424e-05, "loss": 0.0288, "step": 25670 }, { "grad_norm": 0.26361149549484253, "learning_rate": 8.875755666122568e-05, "loss": 0.0317, "step": 25680 }, { "grad_norm": 0.2774748206138611, "learning_rate": 8.8747108319239e-05, "loss": 0.0318, "step": 25690 }, { "grad_norm": 0.24842435121536255, "learning_rate": 8.87366557399268e-05, "loss": 0.0336, "step": 25700 }, { "grad_norm": 0.28470495343208313, "learning_rate": 8.872619892443217e-05, "loss": 0.0346, "step": 25710 }, { "grad_norm": 0.3123997151851654, "learning_rate": 8.871573787389865e-05, "loss": 0.0317, "step": 25720 }, { "grad_norm": 0.39113929867744446, "learning_rate": 8.870527258947024e-05, "loss": 0.0305, "step": 25730 }, { "grad_norm": 0.25514543056488037, "learning_rate": 8.869480307229143e-05, "loss": 0.0301, "step": 25740 }, { "grad_norm": 0.20116733014583588, "learning_rate": 8.868432932350712e-05, "loss": 0.0319, "step": 25750 }, { "grad_norm": 0.17798534035682678, "learning_rate": 8.867385134426272e-05, "loss": 0.0307, "step": 25760 }, { "grad_norm": 0.2728956639766693, "learning_rate": 8.866336913570407e-05, "loss": 0.0361, "step": 25770 }, { "grad_norm": 0.205274760723114, "learning_rate": 8.865288269897751e-05, "loss": 0.0314, "step": 25780 }, { "grad_norm": 0.3008285462856293, "learning_rate": 8.864239203522981e-05, "loss": 0.0329, "step": 25790 }, { "grad_norm": 0.30818822979927063, "learning_rate": 8.863189714560822e-05, "loss": 0.0361, "step": 25800 }, { "grad_norm": 0.19410032033920288, "learning_rate": 8.862139803126043e-05, "loss": 0.0327, "step": 25810 }, { "grad_norm": 0.2562702000141144, "learning_rate": 8.861089469333463e-05, "loss": 0.0308, "step": 25820 }, { "grad_norm": 0.2647768259048462, "learning_rate": 8.860038713297944e-05, "loss": 0.0351, "step": 25830 }, { "grad_norm": 0.22354792058467865, "learning_rate": 8.858987535134394e-05, "loss": 0.034, "step": 25840 }, { "grad_norm": 0.2976824939250946, "learning_rate": 8.857935934957769e-05, "loss": 0.032, "step": 25850 }, { "grad_norm": 0.28365176916122437, "learning_rate": 8.856883912883071e-05, "loss": 0.0356, "step": 25860 }, { "grad_norm": 0.23978592455387115, "learning_rate": 8.855831469025346e-05, "loss": 0.0303, "step": 25870 }, { "grad_norm": 0.32249975204467773, "learning_rate": 8.854778603499689e-05, "loss": 0.0358, "step": 25880 }, { "grad_norm": 0.31497010588645935, "learning_rate": 8.85372531642124e-05, "loss": 0.0336, "step": 25890 }, { "grad_norm": 0.35456350445747375, "learning_rate": 8.852671607905185e-05, "loss": 0.035, "step": 25900 }, { "grad_norm": 0.22374066710472107, "learning_rate": 8.851617478066754e-05, "loss": 0.0329, "step": 25910 }, { "grad_norm": 0.27642619609832764, "learning_rate": 8.850562927021227e-05, "loss": 0.0368, "step": 25920 }, { "grad_norm": 0.27505579590797424, "learning_rate": 8.849507954883928e-05, "loss": 0.0306, "step": 25930 }, { "grad_norm": 0.2654416263103485, "learning_rate": 8.848452561770226e-05, "loss": 0.0288, "step": 25940 }, { "grad_norm": 0.24498039484024048, "learning_rate": 8.847396747795538e-05, "loss": 0.0344, "step": 25950 }, { "grad_norm": 0.221655011177063, "learning_rate": 8.846340513075327e-05, "loss": 0.0339, "step": 25960 }, { "grad_norm": 0.35126057267189026, "learning_rate": 8.845283857725099e-05, "loss": 0.0427, "step": 25970 }, { "grad_norm": 0.22768472135066986, "learning_rate": 8.844226781860409e-05, "loss": 0.0397, "step": 25980 }, { "grad_norm": 0.30838853120803833, "learning_rate": 8.84316928559686e-05, "loss": 0.0324, "step": 25990 }, { "grad_norm": 0.24339203536510468, "learning_rate": 8.842111369050094e-05, "loss": 0.0312, "step": 26000 }, { "grad_norm": 0.2196424901485443, "learning_rate": 8.841053032335808e-05, "loss": 0.03, "step": 26010 }, { "grad_norm": 0.2551206350326538, "learning_rate": 8.839994275569735e-05, "loss": 0.0352, "step": 26020 }, { "grad_norm": 0.275504469871521, "learning_rate": 8.838935098867662e-05, "loss": 0.0276, "step": 26030 }, { "grad_norm": 0.26076409220695496, "learning_rate": 8.837875502345418e-05, "loss": 0.0345, "step": 26040 }, { "grad_norm": 0.3193630278110504, "learning_rate": 8.83681548611888e-05, "loss": 0.0312, "step": 26050 }, { "grad_norm": 0.26447078585624695, "learning_rate": 8.835755050303969e-05, "loss": 0.034, "step": 26060 }, { "grad_norm": 0.305547833442688, "learning_rate": 8.834694195016653e-05, "loss": 0.0352, "step": 26070 }, { "grad_norm": 0.24812372028827667, "learning_rate": 8.833632920372942e-05, "loss": 0.0347, "step": 26080 }, { "grad_norm": 0.3205122649669647, "learning_rate": 8.832571226488903e-05, "loss": 0.0391, "step": 26090 }, { "grad_norm": 0.2578911781311035, "learning_rate": 8.831509113480634e-05, "loss": 0.0314, "step": 26100 }, { "grad_norm": 0.26034796237945557, "learning_rate": 8.83044658146429e-05, "loss": 0.0289, "step": 26110 }, { "grad_norm": 0.2876899540424347, "learning_rate": 8.829383630556067e-05, "loss": 0.0299, "step": 26120 }, { "grad_norm": 0.25172579288482666, "learning_rate": 8.828320260872207e-05, "loss": 0.0285, "step": 26130 }, { "grad_norm": 0.20887908339500427, "learning_rate": 8.827256472529e-05, "loss": 0.0304, "step": 26140 }, { "grad_norm": 0.2695079445838928, "learning_rate": 8.826192265642778e-05, "loss": 0.0266, "step": 26150 }, { "grad_norm": 0.22042959928512573, "learning_rate": 8.825127640329923e-05, "loss": 0.0288, "step": 26160 }, { "grad_norm": 0.243683323264122, "learning_rate": 8.824062596706861e-05, "loss": 0.0338, "step": 26170 }, { "grad_norm": 0.2821699380874634, "learning_rate": 8.822997134890062e-05, "loss": 0.0295, "step": 26180 }, { "grad_norm": 0.24535785615444183, "learning_rate": 8.821931254996044e-05, "loss": 0.0301, "step": 26190 }, { "grad_norm": 0.5683066844940186, "learning_rate": 8.82086495714137e-05, "loss": 0.0302, "step": 26200 }, { "grad_norm": 0.23369131982326508, "learning_rate": 8.81979824144265e-05, "loss": 0.0364, "step": 26210 }, { "grad_norm": 0.26108062267303467, "learning_rate": 8.818731108016536e-05, "loss": 0.036, "step": 26220 }, { "grad_norm": 0.331637442111969, "learning_rate": 8.81766355697973e-05, "loss": 0.0322, "step": 26230 }, { "grad_norm": 0.24491356313228607, "learning_rate": 8.816595588448977e-05, "loss": 0.0305, "step": 26240 }, { "grad_norm": 0.31158915162086487, "learning_rate": 8.81552720254107e-05, "loss": 0.0307, "step": 26250 }, { "grad_norm": 0.2864691913127899, "learning_rate": 8.814458399372842e-05, "loss": 0.0382, "step": 26260 }, { "grad_norm": 0.23198489844799042, "learning_rate": 8.813389179061181e-05, "loss": 0.0363, "step": 26270 }, { "grad_norm": 0.19381439685821533, "learning_rate": 8.812319541723012e-05, "loss": 0.0294, "step": 26280 }, { "grad_norm": 0.25635644793510437, "learning_rate": 8.811249487475309e-05, "loss": 0.0315, "step": 26290 }, { "grad_norm": 0.32539618015289307, "learning_rate": 8.810179016435092e-05, "loss": 0.035, "step": 26300 }, { "grad_norm": 0.30357110500335693, "learning_rate": 8.809108128719428e-05, "loss": 0.031, "step": 26310 }, { "grad_norm": 0.27250537276268005, "learning_rate": 8.808036824445424e-05, "loss": 0.0307, "step": 26320 }, { "grad_norm": 0.23637047410011292, "learning_rate": 8.806965103730238e-05, "loss": 0.0351, "step": 26330 }, { "grad_norm": 0.29495906829833984, "learning_rate": 8.805892966691074e-05, "loss": 0.034, "step": 26340 }, { "grad_norm": 0.2849141061306, "learning_rate": 8.804820413445175e-05, "loss": 0.0309, "step": 26350 }, { "grad_norm": 0.2527084946632385, "learning_rate": 8.803747444109837e-05, "loss": 0.0321, "step": 26360 }, { "grad_norm": 0.20111288130283356, "learning_rate": 8.802674058802399e-05, "loss": 0.0369, "step": 26370 }, { "grad_norm": 0.21939319372177124, "learning_rate": 8.801600257640241e-05, "loss": 0.028, "step": 26380 }, { "grad_norm": 0.2575414478778839, "learning_rate": 8.800526040740795e-05, "loss": 0.03, "step": 26390 }, { "grad_norm": 0.2633073627948761, "learning_rate": 8.799451408221535e-05, "loss": 0.0307, "step": 26400 }, { "grad_norm": 0.3798355162143707, "learning_rate": 8.798376360199982e-05, "loss": 0.032, "step": 26410 }, { "grad_norm": 0.1888285130262375, "learning_rate": 8.797300896793701e-05, "loss": 0.0311, "step": 26420 }, { "grad_norm": 0.29341983795166016, "learning_rate": 8.796225018120302e-05, "loss": 0.0318, "step": 26430 }, { "grad_norm": 0.2797953188419342, "learning_rate": 8.795148724297444e-05, "loss": 0.0319, "step": 26440 }, { "grad_norm": 0.2177926003932953, "learning_rate": 8.794072015442825e-05, "loss": 0.0342, "step": 26450 }, { "grad_norm": 0.24700815975666046, "learning_rate": 8.792994891674198e-05, "loss": 0.0296, "step": 26460 }, { "grad_norm": 0.3155190050601959, "learning_rate": 8.79191735310935e-05, "loss": 0.0379, "step": 26470 }, { "grad_norm": 0.24923697113990784, "learning_rate": 8.790839399866122e-05, "loss": 0.0356, "step": 26480 }, { "grad_norm": 0.3219169080257416, "learning_rate": 8.789761032062397e-05, "loss": 0.0332, "step": 26490 }, { "grad_norm": 0.3320319354534149, "learning_rate": 8.788682249816103e-05, "loss": 0.0325, "step": 26500 }, { "grad_norm": 0.19953490793704987, "learning_rate": 8.787603053245215e-05, "loss": 0.0368, "step": 26510 }, { "grad_norm": 0.23914195597171783, "learning_rate": 8.78652344246775e-05, "loss": 0.0308, "step": 26520 }, { "grad_norm": 0.26475340127944946, "learning_rate": 8.785443417601776e-05, "loss": 0.0319, "step": 26530 }, { "grad_norm": 0.2621425986289978, "learning_rate": 8.784362978765401e-05, "loss": 0.0297, "step": 26540 }, { "grad_norm": 0.2806166708469391, "learning_rate": 8.783282126076779e-05, "loss": 0.0344, "step": 26550 }, { "grad_norm": 0.2794418931007385, "learning_rate": 8.782200859654112e-05, "loss": 0.0348, "step": 26560 }, { "grad_norm": 0.23146659135818481, "learning_rate": 8.781119179615646e-05, "loss": 0.0284, "step": 26570 }, { "grad_norm": 0.2645247280597687, "learning_rate": 8.780037086079674e-05, "loss": 0.0277, "step": 26580 }, { "grad_norm": 0.3921785056591034, "learning_rate": 8.778954579164527e-05, "loss": 0.0352, "step": 26590 }, { "grad_norm": 0.295671671628952, "learning_rate": 8.777871658988588e-05, "loss": 0.0357, "step": 26600 }, { "grad_norm": 0.26442304253578186, "learning_rate": 8.776788325670285e-05, "loss": 0.0294, "step": 26610 }, { "grad_norm": 0.31399068236351013, "learning_rate": 8.775704579328089e-05, "loss": 0.0336, "step": 26620 }, { "grad_norm": 0.33061662316322327, "learning_rate": 8.774620420080517e-05, "loss": 0.0348, "step": 26630 }, { "grad_norm": 0.284458190202713, "learning_rate": 8.773535848046131e-05, "loss": 0.0298, "step": 26640 }, { "grad_norm": 0.233673557639122, "learning_rate": 8.772450863343538e-05, "loss": 0.031, "step": 26650 }, { "grad_norm": 0.30060601234436035, "learning_rate": 8.77136546609139e-05, "loss": 0.0328, "step": 26660 }, { "grad_norm": 0.28029945492744446, "learning_rate": 8.770279656408385e-05, "loss": 0.0285, "step": 26670 }, { "grad_norm": 0.3429965078830719, "learning_rate": 8.769193434413265e-05, "loss": 0.0363, "step": 26680 }, { "grad_norm": 0.22911083698272705, "learning_rate": 8.76810680022482e-05, "loss": 0.0304, "step": 26690 }, { "grad_norm": 0.34945064783096313, "learning_rate": 8.767019753961878e-05, "loss": 0.0293, "step": 26700 }, { "grad_norm": 0.39483916759490967, "learning_rate": 8.765932295743321e-05, "loss": 0.0334, "step": 26710 }, { "grad_norm": 0.3536365032196045, "learning_rate": 8.764844425688068e-05, "loss": 0.0319, "step": 26720 }, { "grad_norm": 0.2959522306919098, "learning_rate": 8.763756143915092e-05, "loss": 0.029, "step": 26730 }, { "grad_norm": 0.2954087555408478, "learning_rate": 8.7626674505434e-05, "loss": 0.0362, "step": 26740 }, { "grad_norm": 0.2583215832710266, "learning_rate": 8.761578345692053e-05, "loss": 0.0311, "step": 26750 }, { "grad_norm": 0.2934701144695282, "learning_rate": 8.760488829480156e-05, "loss": 0.0307, "step": 26760 }, { "grad_norm": 0.25193557143211365, "learning_rate": 8.759398902026854e-05, "loss": 0.0297, "step": 26770 }, { "grad_norm": 0.25806060433387756, "learning_rate": 8.758308563451339e-05, "loss": 0.0303, "step": 26780 }, { "grad_norm": 0.2028324007987976, "learning_rate": 8.75721781387285e-05, "loss": 0.0304, "step": 26790 }, { "grad_norm": 0.29793399572372437, "learning_rate": 8.75612665341067e-05, "loss": 0.0312, "step": 26800 }, { "grad_norm": 0.26868492364883423, "learning_rate": 8.755035082184126e-05, "loss": 0.0316, "step": 26810 }, { "grad_norm": 0.21711254119873047, "learning_rate": 8.753943100312592e-05, "loss": 0.0252, "step": 26820 }, { "grad_norm": 0.19377505779266357, "learning_rate": 8.752850707915484e-05, "loss": 0.024, "step": 26830 }, { "grad_norm": 0.2649829685688019, "learning_rate": 8.751757905112264e-05, "loss": 0.031, "step": 26840 }, { "grad_norm": 0.2590620517730713, "learning_rate": 8.75066469202244e-05, "loss": 0.0328, "step": 26850 }, { "grad_norm": 0.1995762139558792, "learning_rate": 8.749571068765567e-05, "loss": 0.0352, "step": 26860 }, { "grad_norm": 0.28173622488975525, "learning_rate": 8.748477035461238e-05, "loss": 0.0308, "step": 26870 }, { "grad_norm": 0.3331311047077179, "learning_rate": 8.747382592229095e-05, "loss": 0.041, "step": 26880 }, { "grad_norm": 0.3011120557785034, "learning_rate": 8.746287739188828e-05, "loss": 0.0312, "step": 26890 }, { "grad_norm": 0.27300581336021423, "learning_rate": 8.745192476460165e-05, "loss": 0.0328, "step": 26900 }, { "grad_norm": 0.2577802538871765, "learning_rate": 8.744096804162882e-05, "loss": 0.0349, "step": 26910 }, { "grad_norm": 0.28879454731941223, "learning_rate": 8.743000722416804e-05, "loss": 0.0332, "step": 26920 }, { "grad_norm": 0.29603129625320435, "learning_rate": 8.741904231341793e-05, "loss": 0.0371, "step": 26930 }, { "grad_norm": 0.24716421961784363, "learning_rate": 8.740807331057762e-05, "loss": 0.0346, "step": 26940 }, { "grad_norm": 0.28441372513771057, "learning_rate": 8.739710021684667e-05, "loss": 0.0342, "step": 26950 }, { "grad_norm": 0.3112241327762604, "learning_rate": 8.738612303342503e-05, "loss": 0.0345, "step": 26960 }, { "grad_norm": 0.3657880127429962, "learning_rate": 8.73751417615132e-05, "loss": 0.0396, "step": 26970 }, { "grad_norm": 0.2829769253730774, "learning_rate": 8.736415640231208e-05, "loss": 0.0335, "step": 26980 }, { "grad_norm": 0.2144586145877838, "learning_rate": 8.735316695702297e-05, "loss": 0.0328, "step": 26990 }, { "grad_norm": 0.23713324964046478, "learning_rate": 8.734217342684769e-05, "loss": 0.0301, "step": 27000 }, { "grad_norm": 0.37090426683425903, "learning_rate": 8.733117581298847e-05, "loss": 0.0354, "step": 27010 }, { "grad_norm": 0.29894694685935974, "learning_rate": 8.732017411664796e-05, "loss": 0.0352, "step": 27020 }, { "grad_norm": 0.28230366110801697, "learning_rate": 8.730916833902936e-05, "loss": 0.0315, "step": 27030 }, { "grad_norm": 0.2707149088382721, "learning_rate": 8.729815848133618e-05, "loss": 0.0302, "step": 27040 }, { "grad_norm": 0.27110719680786133, "learning_rate": 8.728714454477247e-05, "loss": 0.0319, "step": 27050 }, { "grad_norm": 0.4262988269329071, "learning_rate": 8.727612653054269e-05, "loss": 0.0355, "step": 27060 }, { "grad_norm": 0.3165813088417053, "learning_rate": 8.726510443985176e-05, "loss": 0.034, "step": 27070 }, { "grad_norm": 0.24349236488342285, "learning_rate": 8.725407827390503e-05, "loss": 0.0306, "step": 27080 }, { "grad_norm": 0.381091833114624, "learning_rate": 8.724304803390833e-05, "loss": 0.0337, "step": 27090 }, { "grad_norm": 0.2578577697277069, "learning_rate": 8.723201372106788e-05, "loss": 0.0324, "step": 27100 }, { "grad_norm": 0.24248425662517548, "learning_rate": 8.722097533659038e-05, "loss": 0.0317, "step": 27110 }, { "grad_norm": 0.2864668667316437, "learning_rate": 8.720993288168299e-05, "loss": 0.0381, "step": 27120 }, { "grad_norm": 0.25497689843177795, "learning_rate": 8.719888635755327e-05, "loss": 0.0284, "step": 27130 }, { "grad_norm": 0.2329937219619751, "learning_rate": 8.718783576540928e-05, "loss": 0.0297, "step": 27140 }, { "grad_norm": 0.2266053706407547, "learning_rate": 8.717678110645948e-05, "loss": 0.0303, "step": 27150 }, { "grad_norm": 0.22946278750896454, "learning_rate": 8.716572238191279e-05, "loss": 0.0304, "step": 27160 }, { "grad_norm": 0.27852046489715576, "learning_rate": 8.715465959297857e-05, "loss": 0.0253, "step": 27170 }, { "grad_norm": 0.34006500244140625, "learning_rate": 8.714359274086665e-05, "loss": 0.0332, "step": 27180 }, { "grad_norm": 0.21337474882602692, "learning_rate": 8.713252182678726e-05, "loss": 0.0288, "step": 27190 }, { "grad_norm": 0.2735951244831085, "learning_rate": 8.712144685195112e-05, "loss": 0.0292, "step": 27200 }, { "grad_norm": 0.24253275990486145, "learning_rate": 8.711036781756936e-05, "loss": 0.029, "step": 27210 }, { "grad_norm": 0.2618557810783386, "learning_rate": 8.709928472485357e-05, "loss": 0.0345, "step": 27220 }, { "grad_norm": 0.2855953872203827, "learning_rate": 8.708819757501579e-05, "loss": 0.0314, "step": 27230 }, { "grad_norm": 0.2361304759979248, "learning_rate": 8.707710636926846e-05, "loss": 0.0326, "step": 27240 }, { "grad_norm": 0.2637213170528412, "learning_rate": 8.706601110882455e-05, "loss": 0.0324, "step": 27250 }, { "grad_norm": 0.2235502451658249, "learning_rate": 8.705491179489738e-05, "loss": 0.028, "step": 27260 }, { "grad_norm": 0.3077320158481598, "learning_rate": 8.704380842870077e-05, "loss": 0.0327, "step": 27270 }, { "grad_norm": 0.2762828469276428, "learning_rate": 8.703270101144895e-05, "loss": 0.0317, "step": 27280 }, { "grad_norm": 0.3410525321960449, "learning_rate": 8.702158954435664e-05, "loss": 0.0336, "step": 27290 }, { "grad_norm": 0.3256855905056, "learning_rate": 8.701047402863896e-05, "loss": 0.0307, "step": 27300 }, { "grad_norm": 0.26226091384887695, "learning_rate": 8.699935446551148e-05, "loss": 0.029, "step": 27310 }, { "grad_norm": 0.23285186290740967, "learning_rate": 8.698823085619022e-05, "loss": 0.0287, "step": 27320 }, { "grad_norm": 0.3299570083618164, "learning_rate": 8.697710320189166e-05, "loss": 0.0293, "step": 27330 }, { "grad_norm": 0.18832939863204956, "learning_rate": 8.696597150383268e-05, "loss": 0.0369, "step": 27340 }, { "grad_norm": 0.2441069632768631, "learning_rate": 8.695483576323063e-05, "loss": 0.0307, "step": 27350 }, { "grad_norm": 0.2720819413661957, "learning_rate": 8.69436959813033e-05, "loss": 0.0318, "step": 27360 }, { "grad_norm": 0.27388864755630493, "learning_rate": 8.693255215926892e-05, "loss": 0.0329, "step": 27370 }, { "grad_norm": 0.22461706399917603, "learning_rate": 8.692140429834617e-05, "loss": 0.0314, "step": 27380 }, { "grad_norm": 0.2624846398830414, "learning_rate": 8.691025239975415e-05, "loss": 0.0347, "step": 27390 }, { "grad_norm": 0.20183420181274414, "learning_rate": 8.689909646471243e-05, "loss": 0.0328, "step": 27400 }, { "grad_norm": 0.3867409825325012, "learning_rate": 8.688793649444099e-05, "loss": 0.0379, "step": 27410 }, { "grad_norm": 0.2743644118309021, "learning_rate": 8.687677249016029e-05, "loss": 0.0319, "step": 27420 }, { "grad_norm": 0.2528073191642761, "learning_rate": 8.686560445309118e-05, "loss": 0.0359, "step": 27430 }, { "grad_norm": 0.2903256118297577, "learning_rate": 8.685443238445499e-05, "loss": 0.0332, "step": 27440 }, { "grad_norm": 0.21825505793094635, "learning_rate": 8.68432562854735e-05, "loss": 0.0313, "step": 27450 }, { "grad_norm": 0.2234235405921936, "learning_rate": 8.683207615736887e-05, "loss": 0.0344, "step": 27460 }, { "grad_norm": 0.30160069465637207, "learning_rate": 8.682089200136379e-05, "loss": 0.0373, "step": 27470 }, { "grad_norm": 0.27910158038139343, "learning_rate": 8.680970381868132e-05, "loss": 0.0289, "step": 27480 }, { "grad_norm": 0.24878476560115814, "learning_rate": 8.679851161054498e-05, "loss": 0.0277, "step": 27490 }, { "grad_norm": 0.3259486258029938, "learning_rate": 8.678731537817873e-05, "loss": 0.0292, "step": 27500 }, { "grad_norm": 0.25680163502693176, "learning_rate": 8.677611512280697e-05, "loss": 0.0297, "step": 27510 }, { "grad_norm": 0.3486025631427765, "learning_rate": 8.676491084565457e-05, "loss": 0.0331, "step": 27520 }, { "grad_norm": 0.23437157273292542, "learning_rate": 8.675370254794678e-05, "loss": 0.0301, "step": 27530 }, { "grad_norm": 0.31679269671440125, "learning_rate": 8.674249023090935e-05, "loss": 0.0346, "step": 27540 }, { "grad_norm": 0.28259357810020447, "learning_rate": 8.673127389576843e-05, "loss": 0.0332, "step": 27550 }, { "grad_norm": 0.3223806619644165, "learning_rate": 8.67200535437506e-05, "loss": 0.0343, "step": 27560 }, { "grad_norm": 0.2380538284778595, "learning_rate": 8.670882917608296e-05, "loss": 0.0315, "step": 27570 }, { "grad_norm": 0.2522450089454651, "learning_rate": 8.669760079399292e-05, "loss": 0.0335, "step": 27580 }, { "grad_norm": 0.3432752192020416, "learning_rate": 8.668636839870845e-05, "loss": 0.035, "step": 27590 }, { "grad_norm": 0.3127438724040985, "learning_rate": 8.667513199145789e-05, "loss": 0.0348, "step": 27600 }, { "grad_norm": 0.22100192308425903, "learning_rate": 8.666389157347002e-05, "loss": 0.0351, "step": 27610 }, { "grad_norm": 0.2671111822128296, "learning_rate": 8.66526471459741e-05, "loss": 0.037, "step": 27620 }, { "grad_norm": 0.3629077076911926, "learning_rate": 8.66413987101998e-05, "loss": 0.0337, "step": 27630 }, { "grad_norm": 0.28318145871162415, "learning_rate": 8.663014626737723e-05, "loss": 0.0338, "step": 27640 }, { "grad_norm": 0.2811373174190521, "learning_rate": 8.661888981873691e-05, "loss": 0.033, "step": 27650 }, { "grad_norm": 0.3362991511821747, "learning_rate": 8.660762936550988e-05, "loss": 0.0333, "step": 27660 }, { "grad_norm": 0.22550904750823975, "learning_rate": 8.659636490892753e-05, "loss": 0.0289, "step": 27670 }, { "grad_norm": 0.31222376227378845, "learning_rate": 8.658509645022174e-05, "loss": 0.0353, "step": 27680 }, { "grad_norm": 0.2966766953468323, "learning_rate": 8.657382399062481e-05, "loss": 0.0327, "step": 27690 }, { "grad_norm": 0.2582261562347412, "learning_rate": 8.656254753136946e-05, "loss": 0.0368, "step": 27700 }, { "grad_norm": 0.23848596215248108, "learning_rate": 8.655126707368891e-05, "loss": 0.0389, "step": 27710 }, { "grad_norm": 0.286754310131073, "learning_rate": 8.653998261881672e-05, "loss": 0.0385, "step": 27720 }, { "grad_norm": 0.2920667231082916, "learning_rate": 8.652869416798699e-05, "loss": 0.0378, "step": 27730 }, { "grad_norm": 0.2574578523635864, "learning_rate": 8.651740172243417e-05, "loss": 0.0291, "step": 27740 }, { "grad_norm": 0.19677841663360596, "learning_rate": 8.65061052833932e-05, "loss": 0.0335, "step": 27750 }, { "grad_norm": 0.3881799280643463, "learning_rate": 8.649480485209945e-05, "loss": 0.0281, "step": 27760 }, { "grad_norm": 0.3389708697795868, "learning_rate": 8.64835004297887e-05, "loss": 0.0286, "step": 27770 }, { "grad_norm": 0.2499667853116989, "learning_rate": 8.64721920176972e-05, "loss": 0.0327, "step": 27780 }, { "grad_norm": 0.35647356510162354, "learning_rate": 8.646087961706164e-05, "loss": 0.0413, "step": 27790 }, { "grad_norm": 0.28966426849365234, "learning_rate": 8.644956322911908e-05, "loss": 0.0331, "step": 27800 }, { "grad_norm": 0.2585265636444092, "learning_rate": 8.643824285510709e-05, "loss": 0.0326, "step": 27810 }, { "grad_norm": 0.273497074842453, "learning_rate": 8.642691849626364e-05, "loss": 0.0327, "step": 27820 }, { "grad_norm": 0.2072782665491104, "learning_rate": 8.641559015382717e-05, "loss": 0.0328, "step": 27830 }, { "grad_norm": 0.24554696679115295, "learning_rate": 8.640425782903649e-05, "loss": 0.0326, "step": 27840 }, { "grad_norm": 0.21862640976905823, "learning_rate": 8.639292152313091e-05, "loss": 0.0298, "step": 27850 }, { "grad_norm": 0.20832963287830353, "learning_rate": 8.638158123735015e-05, "loss": 0.0349, "step": 27860 }, { "grad_norm": 0.25898200273513794, "learning_rate": 8.637023697293436e-05, "loss": 0.0297, "step": 27870 }, { "grad_norm": 0.2667384743690491, "learning_rate": 8.635888873112414e-05, "loss": 0.0335, "step": 27880 }, { "grad_norm": 0.30192071199417114, "learning_rate": 8.634753651316052e-05, "loss": 0.0328, "step": 27890 }, { "grad_norm": 0.24700899422168732, "learning_rate": 8.633618032028496e-05, "loss": 0.0341, "step": 27900 }, { "grad_norm": 0.278747022151947, "learning_rate": 8.632482015373934e-05, "loss": 0.0294, "step": 27910 }, { "grad_norm": 0.28167447447776794, "learning_rate": 8.6313456014766e-05, "loss": 0.0326, "step": 27920 }, { "grad_norm": 0.28035035729408264, "learning_rate": 8.630208790460771e-05, "loss": 0.0297, "step": 27930 }, { "grad_norm": 0.27978649735450745, "learning_rate": 8.629071582450768e-05, "loss": 0.0314, "step": 27940 }, { "grad_norm": 0.2546902894973755, "learning_rate": 8.62793397757095e-05, "loss": 0.0318, "step": 27950 }, { "grad_norm": 0.22936201095581055, "learning_rate": 8.626795975945729e-05, "loss": 0.0347, "step": 27960 }, { "grad_norm": 0.3270769417285919, "learning_rate": 8.625657577699551e-05, "loss": 0.0352, "step": 27970 }, { "grad_norm": 0.2262604832649231, "learning_rate": 8.624518782956914e-05, "loss": 0.0299, "step": 27980 }, { "grad_norm": 0.21393388509750366, "learning_rate": 8.62337959184235e-05, "loss": 0.033, "step": 27990 }, { "grad_norm": 0.24439893662929535, "learning_rate": 8.622240004480441e-05, "loss": 0.0389, "step": 28000 }, { "grad_norm": 0.2785864472389221, "learning_rate": 8.621100020995814e-05, "loss": 0.0312, "step": 28010 }, { "grad_norm": 0.3763243556022644, "learning_rate": 8.619959641513132e-05, "loss": 0.0389, "step": 28020 }, { "grad_norm": 0.25696271657943726, "learning_rate": 8.618818866157105e-05, "loss": 0.033, "step": 28030 }, { "grad_norm": 0.23185275495052338, "learning_rate": 8.617677695052487e-05, "loss": 0.03, "step": 28040 }, { "grad_norm": 0.3430928885936737, "learning_rate": 8.616536128324078e-05, "loss": 0.032, "step": 28050 }, { "grad_norm": 0.28878796100616455, "learning_rate": 8.615394166096712e-05, "loss": 0.0357, "step": 28060 }, { "grad_norm": 0.3227677345275879, "learning_rate": 8.614251808495279e-05, "loss": 0.0305, "step": 28070 }, { "grad_norm": 0.32288414239883423, "learning_rate": 8.6131090556447e-05, "loss": 0.0348, "step": 28080 }, { "grad_norm": 0.2823508083820343, "learning_rate": 8.611965907669947e-05, "loss": 0.0326, "step": 28090 }, { "grad_norm": 0.2396944910287857, "learning_rate": 8.610822364696034e-05, "loss": 0.0425, "step": 28100 }, { "grad_norm": 0.2530900537967682, "learning_rate": 8.609678426848015e-05, "loss": 0.0342, "step": 28110 }, { "grad_norm": 0.27305901050567627, "learning_rate": 8.60853409425099e-05, "loss": 0.0301, "step": 28120 }, { "grad_norm": 0.2278807908296585, "learning_rate": 8.607389367030104e-05, "loss": 0.0308, "step": 28130 }, { "grad_norm": 0.3446587026119232, "learning_rate": 8.606244245310538e-05, "loss": 0.0359, "step": 28140 }, { "grad_norm": 0.2804109454154968, "learning_rate": 8.605098729217525e-05, "loss": 0.0329, "step": 28150 }, { "grad_norm": 0.2610727846622467, "learning_rate": 8.603952818876335e-05, "loss": 0.0301, "step": 28160 }, { "grad_norm": 0.32747316360473633, "learning_rate": 8.602806514412281e-05, "loss": 0.0316, "step": 28170 }, { "grad_norm": 0.24440018832683563, "learning_rate": 8.601659815950726e-05, "loss": 0.035, "step": 28180 }, { "grad_norm": 0.25330501794815063, "learning_rate": 8.600512723617067e-05, "loss": 0.0306, "step": 28190 }, { "grad_norm": 0.2912787199020386, "learning_rate": 8.59936523753675e-05, "loss": 0.0307, "step": 28200 }, { "grad_norm": 0.24501506984233856, "learning_rate": 8.598217357835264e-05, "loss": 0.0279, "step": 28210 }, { "grad_norm": 0.20931297540664673, "learning_rate": 8.597069084638135e-05, "loss": 0.0288, "step": 28220 }, { "grad_norm": 0.24848389625549316, "learning_rate": 8.595920418070939e-05, "loss": 0.0313, "step": 28230 }, { "grad_norm": 0.26637858152389526, "learning_rate": 8.594771358259295e-05, "loss": 0.0287, "step": 28240 }, { "grad_norm": 0.2678806781768799, "learning_rate": 8.593621905328858e-05, "loss": 0.0228, "step": 28250 }, { "grad_norm": 0.26651766896247864, "learning_rate": 8.592472059405333e-05, "loss": 0.0312, "step": 28260 }, { "grad_norm": 0.2398800104856491, "learning_rate": 8.591321820614464e-05, "loss": 0.0291, "step": 28270 }, { "grad_norm": 0.27052944898605347, "learning_rate": 8.590171189082041e-05, "loss": 0.03, "step": 28280 }, { "grad_norm": 0.40397244691848755, "learning_rate": 8.589020164933894e-05, "loss": 0.0318, "step": 28290 }, { "grad_norm": 0.4649145007133484, "learning_rate": 8.587868748295898e-05, "loss": 0.0308, "step": 28300 }, { "grad_norm": 0.2672227621078491, "learning_rate": 8.586716939293971e-05, "loss": 0.0341, "step": 28310 }, { "grad_norm": 0.27756568789482117, "learning_rate": 8.58556473805407e-05, "loss": 0.0335, "step": 28320 }, { "grad_norm": 0.2488502711057663, "learning_rate": 8.584412144702202e-05, "loss": 0.0328, "step": 28330 }, { "grad_norm": 0.23967841267585754, "learning_rate": 8.58325915936441e-05, "loss": 0.0317, "step": 28340 }, { "grad_norm": 0.2792705297470093, "learning_rate": 8.582105782166783e-05, "loss": 0.0361, "step": 28350 }, { "grad_norm": 0.27449050545692444, "learning_rate": 8.580952013235455e-05, "loss": 0.0345, "step": 28360 }, { "grad_norm": 0.24516959488391876, "learning_rate": 8.579797852696596e-05, "loss": 0.0339, "step": 28370 }, { "grad_norm": 0.29571348428726196, "learning_rate": 8.578643300676428e-05, "loss": 0.0318, "step": 28380 }, { "grad_norm": 0.31847983598709106, "learning_rate": 8.577488357301209e-05, "loss": 0.037, "step": 28390 }, { "grad_norm": 0.2934540808200836, "learning_rate": 8.576333022697242e-05, "loss": 0.0329, "step": 28400 }, { "grad_norm": 0.2795698940753937, "learning_rate": 8.575177296990873e-05, "loss": 0.031, "step": 28410 }, { "grad_norm": 0.22258010506629944, "learning_rate": 8.574021180308489e-05, "loss": 0.0277, "step": 28420 }, { "grad_norm": 0.3437902331352234, "learning_rate": 8.572864672776523e-05, "loss": 0.0344, "step": 28430 }, { "grad_norm": 0.2895053029060364, "learning_rate": 8.571707774521447e-05, "loss": 0.0304, "step": 28440 }, { "grad_norm": 0.3342839479446411, "learning_rate": 8.57055048566978e-05, "loss": 0.0347, "step": 28450 }, { "grad_norm": 0.27975988388061523, "learning_rate": 8.569392806348078e-05, "loss": 0.0312, "step": 28460 }, { "grad_norm": 0.24496467411518097, "learning_rate": 8.568234736682947e-05, "loss": 0.0308, "step": 28470 }, { "grad_norm": 0.2492228001356125, "learning_rate": 8.567076276801029e-05, "loss": 0.0349, "step": 28480 }, { "grad_norm": 0.21384268999099731, "learning_rate": 8.565917426829013e-05, "loss": 0.0328, "step": 28490 }, { "grad_norm": 0.227186381816864, "learning_rate": 8.564758186893628e-05, "loss": 0.0322, "step": 28500 }, { "grad_norm": 0.2532918453216553, "learning_rate": 8.563598557121649e-05, "loss": 0.0335, "step": 28510 }, { "grad_norm": 0.2795946002006531, "learning_rate": 8.562438537639888e-05, "loss": 0.0325, "step": 28520 }, { "grad_norm": 0.44494593143463135, "learning_rate": 8.561278128575206e-05, "loss": 0.0351, "step": 28530 }, { "grad_norm": 0.3023938238620758, "learning_rate": 8.5601173300545e-05, "loss": 0.0329, "step": 28540 }, { "grad_norm": 0.2757289409637451, "learning_rate": 8.558956142204717e-05, "loss": 0.0357, "step": 28550 }, { "grad_norm": 0.3260885775089264, "learning_rate": 8.55779456515284e-05, "loss": 0.0308, "step": 28560 }, { "grad_norm": 0.23717382550239563, "learning_rate": 8.556632599025898e-05, "loss": 0.0332, "step": 28570 }, { "grad_norm": 0.2924414277076721, "learning_rate": 8.555470243950964e-05, "loss": 0.03, "step": 28580 }, { "grad_norm": 0.2826656401157379, "learning_rate": 8.554307500055148e-05, "loss": 0.0337, "step": 28590 }, { "grad_norm": 0.2534666657447815, "learning_rate": 8.553144367465609e-05, "loss": 0.0353, "step": 28600 }, { "grad_norm": 0.24450407922267914, "learning_rate": 8.551980846309544e-05, "loss": 0.0289, "step": 28610 }, { "grad_norm": 0.21479777991771698, "learning_rate": 8.550816936714193e-05, "loss": 0.0315, "step": 28620 }, { "grad_norm": 0.34176069498062134, "learning_rate": 8.549652638806841e-05, "loss": 0.0331, "step": 28630 }, { "grad_norm": 0.20848138630390167, "learning_rate": 8.548487952714812e-05, "loss": 0.0325, "step": 28640 }, { "grad_norm": 0.21543844044208527, "learning_rate": 8.547322878565478e-05, "loss": 0.0302, "step": 28650 }, { "grad_norm": 0.31692901253700256, "learning_rate": 8.546157416486245e-05, "loss": 0.0331, "step": 28660 }, { "grad_norm": 0.3294679820537567, "learning_rate": 8.54499156660457e-05, "loss": 0.0305, "step": 28670 }, { "grad_norm": 0.2820577323436737, "learning_rate": 8.543825329047947e-05, "loss": 0.0321, "step": 28680 }, { "grad_norm": 0.256078839302063, "learning_rate": 8.542658703943913e-05, "loss": 0.0325, "step": 28690 }, { "grad_norm": 0.2426658570766449, "learning_rate": 8.541491691420051e-05, "loss": 0.0314, "step": 28700 }, { "grad_norm": 0.2855072021484375, "learning_rate": 8.54032429160398e-05, "loss": 0.029, "step": 28710 }, { "grad_norm": 0.20682239532470703, "learning_rate": 8.539156504623369e-05, "loss": 0.0298, "step": 28720 }, { "grad_norm": 0.3126106858253479, "learning_rate": 8.537988330605923e-05, "loss": 0.0293, "step": 28730 }, { "grad_norm": 0.2706621587276459, "learning_rate": 8.536819769679393e-05, "loss": 0.0305, "step": 28740 }, { "grad_norm": 0.22146378457546234, "learning_rate": 8.53565082197157e-05, "loss": 0.0311, "step": 28750 }, { "grad_norm": 0.2981277406215668, "learning_rate": 8.534481487610289e-05, "loss": 0.0305, "step": 28760 }, { "grad_norm": 0.2371606081724167, "learning_rate": 8.533311766723428e-05, "loss": 0.0274, "step": 28770 }, { "grad_norm": 0.22843468189239502, "learning_rate": 8.532141659438901e-05, "loss": 0.0342, "step": 28780 }, { "grad_norm": 0.3371422588825226, "learning_rate": 8.530971165884675e-05, "loss": 0.0342, "step": 28790 }, { "grad_norm": 0.2955073416233063, "learning_rate": 8.529800286188752e-05, "loss": 0.0273, "step": 28800 }, { "grad_norm": 0.2698720693588257, "learning_rate": 8.528629020479175e-05, "loss": 0.033, "step": 28810 }, { "grad_norm": 0.2664175033569336, "learning_rate": 8.527457368884033e-05, "loss": 0.0335, "step": 28820 }, { "grad_norm": 0.24332793056964874, "learning_rate": 8.526285331531458e-05, "loss": 0.031, "step": 28830 }, { "grad_norm": 0.2041701376438141, "learning_rate": 8.525112908549621e-05, "loss": 0.0288, "step": 28840 }, { "grad_norm": 0.285689115524292, "learning_rate": 8.523940100066735e-05, "loss": 0.0291, "step": 28850 }, { "grad_norm": 0.3134748637676239, "learning_rate": 8.52276690621106e-05, "loss": 0.0284, "step": 28860 }, { "grad_norm": 0.33145835995674133, "learning_rate": 8.521593327110889e-05, "loss": 0.0327, "step": 28870 }, { "grad_norm": 0.2737586498260498, "learning_rate": 8.520419362894569e-05, "loss": 0.0274, "step": 28880 }, { "grad_norm": 0.22529497742652893, "learning_rate": 8.51924501369048e-05, "loss": 0.0263, "step": 28890 }, { "grad_norm": 0.3044995367527008, "learning_rate": 8.518070279627047e-05, "loss": 0.0318, "step": 28900 }, { "grad_norm": 0.22555598616600037, "learning_rate": 8.516895160832737e-05, "loss": 0.0293, "step": 28910 }, { "grad_norm": 0.14800338447093964, "learning_rate": 8.515719657436061e-05, "loss": 0.0281, "step": 28920 }, { "grad_norm": 0.291343092918396, "learning_rate": 8.514543769565568e-05, "loss": 0.0307, "step": 28930 }, { "grad_norm": 0.23933616280555725, "learning_rate": 8.513367497349853e-05, "loss": 0.0289, "step": 28940 }, { "grad_norm": 0.25151658058166504, "learning_rate": 8.51219084091755e-05, "loss": 0.0306, "step": 28950 }, { "grad_norm": 0.2804187834262848, "learning_rate": 8.511013800397338e-05, "loss": 0.0295, "step": 28960 }, { "grad_norm": 0.22214095294475555, "learning_rate": 8.509836375917937e-05, "loss": 0.0279, "step": 28970 }, { "grad_norm": 0.2886159420013428, "learning_rate": 8.508658567608104e-05, "loss": 0.0328, "step": 28980 }, { "grad_norm": 0.24564361572265625, "learning_rate": 8.507480375596647e-05, "loss": 0.0269, "step": 28990 }, { "grad_norm": 0.30667248368263245, "learning_rate": 8.506301800012408e-05, "loss": 0.032, "step": 29000 }, { "grad_norm": 0.23483377695083618, "learning_rate": 8.505122840984278e-05, "loss": 0.0321, "step": 29010 }, { "grad_norm": 0.22142136096954346, "learning_rate": 8.503943498641182e-05, "loss": 0.0308, "step": 29020 }, { "grad_norm": 0.2671501934528351, "learning_rate": 8.502763773112095e-05, "loss": 0.0267, "step": 29030 }, { "grad_norm": 0.2717728614807129, "learning_rate": 8.501583664526026e-05, "loss": 0.028, "step": 29040 }, { "grad_norm": 0.2229035347700119, "learning_rate": 8.500403173012032e-05, "loss": 0.0301, "step": 29050 }, { "grad_norm": 0.2626475989818573, "learning_rate": 8.499222298699211e-05, "loss": 0.0267, "step": 29060 }, { "grad_norm": 0.28502821922302246, "learning_rate": 8.498041041716701e-05, "loss": 0.0309, "step": 29070 }, { "grad_norm": 0.23269999027252197, "learning_rate": 8.496859402193681e-05, "loss": 0.0332, "step": 29080 }, { "grad_norm": 0.2446773797273636, "learning_rate": 8.495677380259374e-05, "loss": 0.0278, "step": 29090 }, { "grad_norm": 0.23876947164535522, "learning_rate": 8.494494976043045e-05, "loss": 0.0316, "step": 29100 }, { "grad_norm": 0.24391363561153412, "learning_rate": 8.493312189673998e-05, "loss": 0.0313, "step": 29110 }, { "grad_norm": 0.26132598519325256, "learning_rate": 8.492129021281584e-05, "loss": 0.028, "step": 29120 }, { "grad_norm": 0.2390248328447342, "learning_rate": 8.490945470995188e-05, "loss": 0.0297, "step": 29130 }, { "grad_norm": 0.19004550576210022, "learning_rate": 8.489761538944247e-05, "loss": 0.0275, "step": 29140 }, { "grad_norm": 0.25588417053222656, "learning_rate": 8.48857722525823e-05, "loss": 0.0331, "step": 29150 }, { "grad_norm": 0.3398229479789734, "learning_rate": 8.487392530066652e-05, "loss": 0.0326, "step": 29160 }, { "grad_norm": 0.21335922181606293, "learning_rate": 8.486207453499069e-05, "loss": 0.0271, "step": 29170 }, { "grad_norm": 0.3073108494281769, "learning_rate": 8.485021995685082e-05, "loss": 0.0351, "step": 29180 }, { "grad_norm": 0.26445862650871277, "learning_rate": 8.483836156754328e-05, "loss": 0.0307, "step": 29190 }, { "grad_norm": 0.38994574546813965, "learning_rate": 8.482649936836491e-05, "loss": 0.0329, "step": 29200 }, { "grad_norm": 0.22312170267105103, "learning_rate": 8.481463336061293e-05, "loss": 0.0276, "step": 29210 }, { "grad_norm": 0.292822003364563, "learning_rate": 8.480276354558496e-05, "loss": 0.0273, "step": 29220 }, { "grad_norm": 0.3089216649532318, "learning_rate": 8.479088992457913e-05, "loss": 0.0297, "step": 29230 }, { "grad_norm": 0.2828824520111084, "learning_rate": 8.477901249889387e-05, "loss": 0.0328, "step": 29240 }, { "grad_norm": 0.2932935953140259, "learning_rate": 8.47671312698281e-05, "loss": 0.0343, "step": 29250 }, { "grad_norm": 0.30726927518844604, "learning_rate": 8.475524623868112e-05, "loss": 0.0335, "step": 29260 }, { "grad_norm": 0.22521932423114777, "learning_rate": 8.474335740675266e-05, "loss": 0.0285, "step": 29270 }, { "grad_norm": 0.224186509847641, "learning_rate": 8.473146477534289e-05, "loss": 0.0289, "step": 29280 }, { "grad_norm": 0.20257344841957092, "learning_rate": 8.471956834575232e-05, "loss": 0.0271, "step": 29290 }, { "grad_norm": 0.2892553210258484, "learning_rate": 8.470766811928197e-05, "loss": 0.0339, "step": 29300 }, { "grad_norm": 0.23444516956806183, "learning_rate": 8.469576409723323e-05, "loss": 0.0367, "step": 29310 }, { "grad_norm": 0.21870148181915283, "learning_rate": 8.468385628090788e-05, "loss": 0.0314, "step": 29320 }, { "grad_norm": 0.21861864626407623, "learning_rate": 8.467194467160815e-05, "loss": 0.0317, "step": 29330 }, { "grad_norm": 0.26734668016433716, "learning_rate": 8.466002927063667e-05, "loss": 0.0295, "step": 29340 }, { "grad_norm": 0.3004595935344696, "learning_rate": 8.464811007929651e-05, "loss": 0.03, "step": 29350 }, { "grad_norm": 0.20720170438289642, "learning_rate": 8.463618709889114e-05, "loss": 0.0313, "step": 29360 }, { "grad_norm": 0.39003312587738037, "learning_rate": 8.462426033072442e-05, "loss": 0.0298, "step": 29370 }, { "grad_norm": 0.26512810587882996, "learning_rate": 8.461232977610061e-05, "loss": 0.0352, "step": 29380 }, { "grad_norm": 0.2424694001674652, "learning_rate": 8.46003954363245e-05, "loss": 0.0332, "step": 29390 }, { "grad_norm": 0.229887455701828, "learning_rate": 8.458845731270115e-05, "loss": 0.0326, "step": 29400 }, { "grad_norm": 0.2076326161623001, "learning_rate": 8.45765154065361e-05, "loss": 0.0308, "step": 29410 }, { "grad_norm": 0.2896031439304352, "learning_rate": 8.456456971913532e-05, "loss": 0.0253, "step": 29420 }, { "grad_norm": 0.1731501817703247, "learning_rate": 8.455262025180517e-05, "loss": 0.0296, "step": 29430 }, { "grad_norm": 0.22917741537094116, "learning_rate": 8.454066700585242e-05, "loss": 0.0299, "step": 29440 }, { "grad_norm": 0.2413780689239502, "learning_rate": 8.452870998258423e-05, "loss": 0.0315, "step": 29450 }, { "grad_norm": 0.3223854899406433, "learning_rate": 8.451674918330825e-05, "loss": 0.0304, "step": 29460 }, { "grad_norm": 0.2745722234249115, "learning_rate": 8.450478460933246e-05, "loss": 0.0305, "step": 29470 }, { "grad_norm": 0.27845412492752075, "learning_rate": 8.449281626196532e-05, "loss": 0.0363, "step": 29480 }, { "grad_norm": 0.2814173400402069, "learning_rate": 8.448084414251564e-05, "loss": 0.0291, "step": 29490 }, { "grad_norm": 0.2528138756752014, "learning_rate": 8.446886825229271e-05, "loss": 0.031, "step": 29500 }, { "grad_norm": 0.2439904510974884, "learning_rate": 8.445688859260615e-05, "loss": 0.0269, "step": 29510 }, { "grad_norm": 0.29743215441703796, "learning_rate": 8.444490516476606e-05, "loss": 0.0309, "step": 29520 }, { "grad_norm": 0.29133734107017517, "learning_rate": 8.443291797008293e-05, "loss": 0.0315, "step": 29530 }, { "grad_norm": 0.33969005942344666, "learning_rate": 8.442092700986765e-05, "loss": 0.0325, "step": 29540 }, { "grad_norm": 0.27745166420936584, "learning_rate": 8.440893228543156e-05, "loss": 0.0313, "step": 29550 }, { "grad_norm": 0.24930553138256073, "learning_rate": 8.439693379808638e-05, "loss": 0.0321, "step": 29560 }, { "grad_norm": 0.31547582149505615, "learning_rate": 8.43849315491442e-05, "loss": 0.031, "step": 29570 }, { "grad_norm": 0.24278903007507324, "learning_rate": 8.437292553991763e-05, "loss": 0.0313, "step": 29580 }, { "grad_norm": 0.2999397814273834, "learning_rate": 8.436091577171959e-05, "loss": 0.0285, "step": 29590 }, { "grad_norm": 0.2559191882610321, "learning_rate": 8.434890224586347e-05, "loss": 0.034, "step": 29600 }, { "grad_norm": 0.263700395822525, "learning_rate": 8.433688496366303e-05, "loss": 0.0283, "step": 29610 }, { "grad_norm": 0.20381000638008118, "learning_rate": 8.432486392643248e-05, "loss": 0.0296, "step": 29620 }, { "grad_norm": 0.29367804527282715, "learning_rate": 8.431283913548643e-05, "loss": 0.0313, "step": 29630 }, { "grad_norm": 0.27848196029663086, "learning_rate": 8.430081059213985e-05, "loss": 0.0313, "step": 29640 }, { "grad_norm": 0.29716432094573975, "learning_rate": 8.428877829770823e-05, "loss": 0.029, "step": 29650 }, { "grad_norm": 0.2220153957605362, "learning_rate": 8.427674225350735e-05, "loss": 0.0305, "step": 29660 }, { "grad_norm": 0.2749600112438202, "learning_rate": 8.426470246085347e-05, "loss": 0.0318, "step": 29670 }, { "grad_norm": 0.2759235203266144, "learning_rate": 8.425265892106324e-05, "loss": 0.0309, "step": 29680 }, { "grad_norm": 0.25750452280044556, "learning_rate": 8.424061163545374e-05, "loss": 0.033, "step": 29690 }, { "grad_norm": 0.3033273220062256, "learning_rate": 8.422856060534243e-05, "loss": 0.0306, "step": 29700 }, { "grad_norm": 0.20890289545059204, "learning_rate": 8.421650583204718e-05, "loss": 0.0247, "step": 29710 }, { "grad_norm": 0.17172393202781677, "learning_rate": 8.420444731688633e-05, "loss": 0.0271, "step": 29720 }, { "grad_norm": 0.2787506580352783, "learning_rate": 8.419238506117852e-05, "loss": 0.0264, "step": 29730 }, { "grad_norm": 0.2283342033624649, "learning_rate": 8.418031906624289e-05, "loss": 0.0298, "step": 29740 }, { "grad_norm": 0.23428384959697723, "learning_rate": 8.416824933339898e-05, "loss": 0.0308, "step": 29750 }, { "grad_norm": 0.39986681938171387, "learning_rate": 8.415617586396667e-05, "loss": 0.0338, "step": 29760 }, { "grad_norm": 0.24992598593235016, "learning_rate": 8.414409865926632e-05, "loss": 0.0302, "step": 29770 }, { "grad_norm": 0.21004733443260193, "learning_rate": 8.413201772061867e-05, "loss": 0.0284, "step": 29780 }, { "grad_norm": 0.22172553837299347, "learning_rate": 8.411993304934488e-05, "loss": 0.0319, "step": 29790 }, { "grad_norm": 0.2683904469013214, "learning_rate": 8.410784464676654e-05, "loss": 0.0247, "step": 29800 }, { "grad_norm": 0.24377398192882538, "learning_rate": 8.409575251420556e-05, "loss": 0.0266, "step": 29810 }, { "grad_norm": 0.26926061511039734, "learning_rate": 8.408365665298435e-05, "loss": 0.0281, "step": 29820 }, { "grad_norm": 0.27900001406669617, "learning_rate": 8.40715570644257e-05, "loss": 0.0342, "step": 29830 }, { "grad_norm": 0.48499640822410583, "learning_rate": 8.40594537498528e-05, "loss": 0.0333, "step": 29840 }, { "grad_norm": 0.2119874209165573, "learning_rate": 8.404734671058924e-05, "loss": 0.028, "step": 29850 }, { "grad_norm": 0.2978315055370331, "learning_rate": 8.403523594795902e-05, "loss": 0.0333, "step": 29860 }, { "grad_norm": 0.3999640643596649, "learning_rate": 8.402312146328659e-05, "loss": 0.033, "step": 29870 }, { "grad_norm": 0.22438102960586548, "learning_rate": 8.401100325789675e-05, "loss": 0.0348, "step": 29880 }, { "grad_norm": 0.25436946749687195, "learning_rate": 8.399888133311472e-05, "loss": 0.0297, "step": 29890 }, { "grad_norm": 0.22483853995800018, "learning_rate": 8.398675569026613e-05, "loss": 0.0289, "step": 29900 }, { "grad_norm": 0.257575660943985, "learning_rate": 8.397462633067705e-05, "loss": 0.0308, "step": 29910 }, { "grad_norm": 0.2509446144104004, "learning_rate": 8.396249325567392e-05, "loss": 0.0268, "step": 29920 }, { "grad_norm": 0.2999871075153351, "learning_rate": 8.395035646658357e-05, "loss": 0.0339, "step": 29930 }, { "grad_norm": 0.33873721957206726, "learning_rate": 8.39382159647333e-05, "loss": 0.0323, "step": 29940 }, { "grad_norm": 0.2494763433933258, "learning_rate": 8.392607175145075e-05, "loss": 0.0284, "step": 29950 }, { "grad_norm": 0.24535004794597626, "learning_rate": 8.3913923828064e-05, "loss": 0.0285, "step": 29960 }, { "grad_norm": 0.28418534994125366, "learning_rate": 8.390177219590152e-05, "loss": 0.0305, "step": 29970 }, { "grad_norm": 0.25905898213386536, "learning_rate": 8.388961685629222e-05, "loss": 0.0273, "step": 29980 }, { "grad_norm": 0.2274399846792221, "learning_rate": 8.387745781056536e-05, "loss": 0.0329, "step": 29990 }, { "grad_norm": 0.24666529893875122, "learning_rate": 8.386529506005065e-05, "loss": 0.0303, "step": 30000 }, { "grad_norm": 0.4063149690628052, "learning_rate": 8.38531286060782e-05, "loss": 0.0266, "step": 30010 }, { "grad_norm": 0.2893071472644806, "learning_rate": 8.384095844997849e-05, "loss": 0.03, "step": 30020 }, { "grad_norm": 0.3155771493911743, "learning_rate": 8.382878459308245e-05, "loss": 0.0341, "step": 30030 }, { "grad_norm": 0.2997187077999115, "learning_rate": 8.381660703672138e-05, "loss": 0.0291, "step": 30040 }, { "grad_norm": 0.254285991191864, "learning_rate": 8.380442578222702e-05, "loss": 0.0272, "step": 30050 }, { "grad_norm": 0.2684168517589569, "learning_rate": 8.379224083093146e-05, "loss": 0.0317, "step": 30060 }, { "grad_norm": 0.23198585212230682, "learning_rate": 8.378005218416727e-05, "loss": 0.0288, "step": 30070 }, { "grad_norm": 0.22465215623378754, "learning_rate": 8.376785984326735e-05, "loss": 0.0266, "step": 30080 }, { "grad_norm": 0.21838340163230896, "learning_rate": 8.375566380956506e-05, "loss": 0.0295, "step": 30090 }, { "grad_norm": 0.40481331944465637, "learning_rate": 8.374346408439411e-05, "loss": 0.0316, "step": 30100 }, { "grad_norm": 0.30199939012527466, "learning_rate": 8.373126066908868e-05, "loss": 0.0296, "step": 30110 }, { "grad_norm": 0.2826845049858093, "learning_rate": 8.371905356498326e-05, "loss": 0.0275, "step": 30120 }, { "grad_norm": 0.396912544965744, "learning_rate": 8.370684277341288e-05, "loss": 0.0306, "step": 30130 }, { "grad_norm": 0.22614642977714539, "learning_rate": 8.369462829571282e-05, "loss": 0.0268, "step": 30140 }, { "grad_norm": 0.2532893121242523, "learning_rate": 8.36824101332189e-05, "loss": 0.0294, "step": 30150 }, { "grad_norm": 0.21263204514980316, "learning_rate": 8.367018828726721e-05, "loss": 0.0278, "step": 30160 }, { "grad_norm": 0.3130957782268524, "learning_rate": 8.365796275919438e-05, "loss": 0.0272, "step": 30170 }, { "grad_norm": 0.1586490422487259, "learning_rate": 8.364573355033734e-05, "loss": 0.0288, "step": 30180 }, { "grad_norm": 0.24578766524791718, "learning_rate": 8.363350066203346e-05, "loss": 0.0298, "step": 30190 }, { "grad_norm": 0.2982963025569916, "learning_rate": 8.362126409562053e-05, "loss": 0.0288, "step": 30200 }, { "grad_norm": 0.3220597207546234, "learning_rate": 8.360902385243667e-05, "loss": 0.0326, "step": 30210 }, { "grad_norm": 0.2601567506790161, "learning_rate": 8.359677993382052e-05, "loss": 0.0334, "step": 30220 }, { "grad_norm": 0.28115901350975037, "learning_rate": 8.358453234111103e-05, "loss": 0.029, "step": 30230 }, { "grad_norm": 0.22672900557518005, "learning_rate": 8.357228107564756e-05, "loss": 0.0298, "step": 30240 }, { "grad_norm": 0.26484742760658264, "learning_rate": 8.356002613876993e-05, "loss": 0.0314, "step": 30250 }, { "grad_norm": 0.2427399903535843, "learning_rate": 8.35477675318183e-05, "loss": 0.0299, "step": 30260 }, { "grad_norm": 0.28683823347091675, "learning_rate": 8.353550525613323e-05, "loss": 0.0331, "step": 30270 }, { "grad_norm": 0.2599339783191681, "learning_rate": 8.352323931305572e-05, "loss": 0.027, "step": 30280 }, { "grad_norm": 0.25476932525634766, "learning_rate": 8.351096970392717e-05, "loss": 0.0286, "step": 30290 }, { "grad_norm": 0.2807627022266388, "learning_rate": 8.349869643008937e-05, "loss": 0.0349, "step": 30300 }, { "grad_norm": 0.22570925951004028, "learning_rate": 8.348641949288449e-05, "loss": 0.0299, "step": 30310 }, { "grad_norm": 0.3033785820007324, "learning_rate": 8.34741388936551e-05, "loss": 0.033, "step": 30320 }, { "grad_norm": 0.22524285316467285, "learning_rate": 8.346185463374423e-05, "loss": 0.0308, "step": 30330 }, { "grad_norm": 0.2761530876159668, "learning_rate": 8.344956671449524e-05, "loss": 0.0321, "step": 30340 }, { "grad_norm": 0.2587639093399048, "learning_rate": 8.343727513725192e-05, "loss": 0.0296, "step": 30350 }, { "grad_norm": 0.24916011095046997, "learning_rate": 8.342497990335847e-05, "loss": 0.0263, "step": 30360 }, { "grad_norm": 0.30196413397789, "learning_rate": 8.341268101415946e-05, "loss": 0.0282, "step": 30370 }, { "grad_norm": 0.2796393930912018, "learning_rate": 8.34003784709999e-05, "loss": 0.036, "step": 30380 }, { "grad_norm": 0.23954588174819946, "learning_rate": 8.338807227522516e-05, "loss": 0.0255, "step": 30390 }, { "grad_norm": 0.44267743825912476, "learning_rate": 8.337576242818103e-05, "loss": 0.031, "step": 30400 }, { "grad_norm": 0.27421900629997253, "learning_rate": 8.336344893121372e-05, "loss": 0.0334, "step": 30410 }, { "grad_norm": 0.2055402398109436, "learning_rate": 8.335113178566977e-05, "loss": 0.0288, "step": 30420 }, { "grad_norm": 0.27586111426353455, "learning_rate": 8.33388109928962e-05, "loss": 0.0305, "step": 30430 }, { "grad_norm": 0.3571722209453583, "learning_rate": 8.33264865542404e-05, "loss": 0.0292, "step": 30440 }, { "grad_norm": 0.17804758250713348, "learning_rate": 8.331415847105013e-05, "loss": 0.028, "step": 30450 }, { "grad_norm": 0.1659199297428131, "learning_rate": 8.330182674467357e-05, "loss": 0.0272, "step": 30460 }, { "grad_norm": 0.21198736131191254, "learning_rate": 8.32894913764593e-05, "loss": 0.0291, "step": 30470 }, { "grad_norm": 0.2329607605934143, "learning_rate": 8.327715236775633e-05, "loss": 0.0326, "step": 30480 }, { "grad_norm": 0.4596140384674072, "learning_rate": 8.326480971991398e-05, "loss": 0.0276, "step": 30490 }, { "grad_norm": 0.28972235321998596, "learning_rate": 8.325246343428206e-05, "loss": 0.0283, "step": 30500 }, { "grad_norm": 0.29519131779670715, "learning_rate": 8.324011351221072e-05, "loss": 0.0289, "step": 30510 }, { "grad_norm": 0.2757754325866699, "learning_rate": 8.322775995505057e-05, "loss": 0.032, "step": 30520 }, { "grad_norm": 0.22799280285835266, "learning_rate": 8.321540276415254e-05, "loss": 0.0283, "step": 30530 }, { "grad_norm": 0.16531576216220856, "learning_rate": 8.320304194086798e-05, "loss": 0.0302, "step": 30540 }, { "grad_norm": 0.24091729521751404, "learning_rate": 8.31906774865487e-05, "loss": 0.0348, "step": 30550 }, { "grad_norm": 0.24157612025737762, "learning_rate": 8.317830940254682e-05, "loss": 0.0281, "step": 30560 }, { "grad_norm": 0.2767704725265503, "learning_rate": 8.316593769021491e-05, "loss": 0.0273, "step": 30570 }, { "grad_norm": 0.24214914441108704, "learning_rate": 8.315356235090592e-05, "loss": 0.0263, "step": 30580 }, { "grad_norm": 0.25207287073135376, "learning_rate": 8.314118338597319e-05, "loss": 0.0244, "step": 30590 }, { "grad_norm": 0.24467666447162628, "learning_rate": 8.312880079677048e-05, "loss": 0.0297, "step": 30600 }, { "grad_norm": 0.2892534136772156, "learning_rate": 8.311641458465191e-05, "loss": 0.0291, "step": 30610 }, { "grad_norm": 0.2911420464515686, "learning_rate": 8.310402475097205e-05, "loss": 0.0273, "step": 30620 }, { "grad_norm": 0.30606234073638916, "learning_rate": 8.309163129708581e-05, "loss": 0.03, "step": 30630 }, { "grad_norm": 0.24173404276371002, "learning_rate": 8.307923422434852e-05, "loss": 0.0293, "step": 30640 }, { "grad_norm": 0.3301555812358856, "learning_rate": 8.30668335341159e-05, "loss": 0.028, "step": 30650 }, { "grad_norm": 0.28205686807632446, "learning_rate": 8.305442922774408e-05, "loss": 0.026, "step": 30660 }, { "grad_norm": 0.2543947696685791, "learning_rate": 8.304202130658959e-05, "loss": 0.0269, "step": 30670 }, { "grad_norm": 0.23683176934719086, "learning_rate": 8.302960977200931e-05, "loss": 0.0328, "step": 30680 }, { "grad_norm": 0.18964719772338867, "learning_rate": 8.301719462536058e-05, "loss": 0.0277, "step": 30690 }, { "grad_norm": 0.2588675022125244, "learning_rate": 8.300477586800108e-05, "loss": 0.031, "step": 30700 }, { "grad_norm": 0.2253868281841278, "learning_rate": 8.299235350128892e-05, "loss": 0.0241, "step": 30710 }, { "grad_norm": 0.2864559292793274, "learning_rate": 8.297992752658258e-05, "loss": 0.0302, "step": 30720 }, { "grad_norm": 0.30650222301483154, "learning_rate": 8.296749794524094e-05, "loss": 0.0364, "step": 30730 }, { "grad_norm": 0.3304641544818878, "learning_rate": 8.295506475862332e-05, "loss": 0.0274, "step": 30740 }, { "grad_norm": 0.31130626797676086, "learning_rate": 8.294262796808933e-05, "loss": 0.0323, "step": 30750 }, { "grad_norm": 0.32716020941734314, "learning_rate": 8.293018757499909e-05, "loss": 0.0346, "step": 30760 }, { "grad_norm": 0.22346124053001404, "learning_rate": 8.291774358071305e-05, "loss": 0.0308, "step": 30770 }, { "grad_norm": 0.2727234959602356, "learning_rate": 8.290529598659205e-05, "loss": 0.0288, "step": 30780 }, { "grad_norm": 0.26781097054481506, "learning_rate": 8.289284479399738e-05, "loss": 0.0288, "step": 30790 }, { "grad_norm": 0.3824976086616516, "learning_rate": 8.288039000429064e-05, "loss": 0.0322, "step": 30800 }, { "grad_norm": 0.40170809626579285, "learning_rate": 8.286793161883388e-05, "loss": 0.0273, "step": 30810 }, { "grad_norm": 0.23443040251731873, "learning_rate": 8.285546963898954e-05, "loss": 0.0293, "step": 30820 }, { "grad_norm": 0.26452046632766724, "learning_rate": 8.284300406612044e-05, "loss": 0.0284, "step": 30830 }, { "grad_norm": 0.3401317000389099, "learning_rate": 8.283053490158978e-05, "loss": 0.0289, "step": 30840 }, { "grad_norm": 0.26362350583076477, "learning_rate": 8.28180621467612e-05, "loss": 0.0277, "step": 30850 }, { "grad_norm": 0.2647136449813843, "learning_rate": 8.280558580299867e-05, "loss": 0.0303, "step": 30860 }, { "grad_norm": 0.2436760514974594, "learning_rate": 8.279310587166661e-05, "loss": 0.0276, "step": 30870 }, { "grad_norm": 0.21014803647994995, "learning_rate": 8.278062235412978e-05, "loss": 0.0315, "step": 30880 }, { "grad_norm": 0.22225111722946167, "learning_rate": 8.276813525175339e-05, "loss": 0.029, "step": 30890 }, { "grad_norm": 0.24069683253765106, "learning_rate": 8.2755644565903e-05, "loss": 0.0268, "step": 30900 }, { "grad_norm": 0.22933071851730347, "learning_rate": 8.274315029794454e-05, "loss": 0.0307, "step": 30910 }, { "grad_norm": 0.23417414724826813, "learning_rate": 8.273065244924443e-05, "loss": 0.0239, "step": 30920 }, { "grad_norm": 0.25484997034072876, "learning_rate": 8.271815102116936e-05, "loss": 0.0285, "step": 30930 }, { "grad_norm": 0.2258586585521698, "learning_rate": 8.270564601508648e-05, "loss": 0.0286, "step": 30940 }, { "grad_norm": 0.2063283920288086, "learning_rate": 8.269313743236333e-05, "loss": 0.0282, "step": 30950 }, { "grad_norm": 0.24098023772239685, "learning_rate": 8.268062527436783e-05, "loss": 0.0296, "step": 30960 }, { "grad_norm": 0.26638123393058777, "learning_rate": 8.266810954246828e-05, "loss": 0.0332, "step": 30970 }, { "grad_norm": 0.2249659299850464, "learning_rate": 8.265559023803338e-05, "loss": 0.0289, "step": 30980 }, { "grad_norm": 0.30343717336654663, "learning_rate": 8.264306736243223e-05, "loss": 0.031, "step": 30990 }, { "grad_norm": 0.30590835213661194, "learning_rate": 8.263054091703432e-05, "loss": 0.0323, "step": 31000 }, { "grad_norm": 0.25962114334106445, "learning_rate": 8.26180109032095e-05, "loss": 0.0319, "step": 31010 }, { "grad_norm": 0.3230019807815552, "learning_rate": 8.260547732232807e-05, "loss": 0.0318, "step": 31020 }, { "grad_norm": 0.21906793117523193, "learning_rate": 8.259294017576063e-05, "loss": 0.028, "step": 31030 }, { "grad_norm": 0.24756604433059692, "learning_rate": 8.258039946487828e-05, "loss": 0.0304, "step": 31040 }, { "grad_norm": 0.27571040391921997, "learning_rate": 8.256785519105241e-05, "loss": 0.0302, "step": 31050 }, { "grad_norm": 0.22275303304195404, "learning_rate": 8.255530735565488e-05, "loss": 0.0272, "step": 31060 }, { "grad_norm": 0.2842899262905121, "learning_rate": 8.254275596005787e-05, "loss": 0.0278, "step": 31070 }, { "grad_norm": 0.24544450640678406, "learning_rate": 8.2530201005634e-05, "loss": 0.0262, "step": 31080 }, { "grad_norm": 0.24723252654075623, "learning_rate": 8.251764249375626e-05, "loss": 0.0292, "step": 31090 }, { "grad_norm": 0.347080796957016, "learning_rate": 8.250508042579803e-05, "loss": 0.0317, "step": 31100 }, { "grad_norm": 0.17764702439308167, "learning_rate": 8.249251480313307e-05, "loss": 0.0272, "step": 31110 }, { "grad_norm": 0.2806331217288971, "learning_rate": 8.247994562713555e-05, "loss": 0.0325, "step": 31120 }, { "grad_norm": 0.2688869535923004, "learning_rate": 8.246737289917998e-05, "loss": 0.0299, "step": 31130 }, { "grad_norm": 0.2516511380672455, "learning_rate": 8.245479662064135e-05, "loss": 0.0295, "step": 31140 }, { "grad_norm": 0.23999948799610138, "learning_rate": 8.244221679289496e-05, "loss": 0.0264, "step": 31150 }, { "grad_norm": 0.2075520157814026, "learning_rate": 8.242963341731652e-05, "loss": 0.0279, "step": 31160 }, { "grad_norm": 0.22710256278514862, "learning_rate": 8.24170464952821e-05, "loss": 0.0278, "step": 31170 }, { "grad_norm": 0.2527301013469696, "learning_rate": 8.240445602816824e-05, "loss": 0.0325, "step": 31180 }, { "grad_norm": 0.24703820049762726, "learning_rate": 8.239186201735179e-05, "loss": 0.0336, "step": 31190 }, { "grad_norm": 0.2432803064584732, "learning_rate": 8.237926446420998e-05, "loss": 0.028, "step": 31200 }, { "grad_norm": 0.20632870495319366, "learning_rate": 8.236666337012052e-05, "loss": 0.0277, "step": 31210 }, { "grad_norm": 0.2699194550514221, "learning_rate": 8.23540587364614e-05, "loss": 0.0316, "step": 31220 }, { "grad_norm": 0.198481485247612, "learning_rate": 8.234145056461107e-05, "loss": 0.0278, "step": 31230 }, { "grad_norm": 0.40718692541122437, "learning_rate": 8.232883885594832e-05, "loss": 0.0284, "step": 31240 }, { "grad_norm": 0.26544809341430664, "learning_rate": 8.231622361185236e-05, "loss": 0.0302, "step": 31250 }, { "grad_norm": 0.3054935932159424, "learning_rate": 8.230360483370278e-05, "loss": 0.0301, "step": 31260 }, { "grad_norm": 0.21688814461231232, "learning_rate": 8.229098252287953e-05, "loss": 0.0275, "step": 31270 }, { "grad_norm": 0.2770293951034546, "learning_rate": 8.2278356680763e-05, "loss": 0.0307, "step": 31280 }, { "grad_norm": 0.2543253004550934, "learning_rate": 8.22657273087339e-05, "loss": 0.0296, "step": 31290 }, { "grad_norm": 0.2294318974018097, "learning_rate": 8.225309440817336e-05, "loss": 0.0272, "step": 31300 }, { "grad_norm": 0.262507826089859, "learning_rate": 8.224045798046293e-05, "loss": 0.0327, "step": 31310 }, { "grad_norm": 0.3003593683242798, "learning_rate": 8.22278180269845e-05, "loss": 0.0317, "step": 31320 }, { "grad_norm": 0.3880457878112793, "learning_rate": 8.221517454912031e-05, "loss": 0.0291, "step": 31330 }, { "grad_norm": 0.2519112825393677, "learning_rate": 8.220252754825308e-05, "loss": 0.0286, "step": 31340 }, { "grad_norm": 0.34809496998786926, "learning_rate": 8.218987702576586e-05, "loss": 0.0274, "step": 31350 }, { "grad_norm": 0.23180019855499268, "learning_rate": 8.217722298304207e-05, "loss": 0.0298, "step": 31360 }, { "grad_norm": 0.24185074865818024, "learning_rate": 8.216456542146557e-05, "loss": 0.0238, "step": 31370 }, { "grad_norm": 0.17818395793437958, "learning_rate": 8.215190434242055e-05, "loss": 0.0283, "step": 31380 }, { "grad_norm": 0.23098188638687134, "learning_rate": 8.213923974729161e-05, "loss": 0.0266, "step": 31390 }, { "grad_norm": 0.19133077561855316, "learning_rate": 8.212657163746373e-05, "loss": 0.0233, "step": 31400 }, { "grad_norm": 0.2306385636329651, "learning_rate": 8.211390001432227e-05, "loss": 0.026, "step": 31410 }, { "grad_norm": 0.3114400804042816, "learning_rate": 8.210122487925297e-05, "loss": 0.03, "step": 31420 }, { "grad_norm": 0.35822373628616333, "learning_rate": 8.208854623364202e-05, "loss": 0.0304, "step": 31430 }, { "grad_norm": 0.2457152158021927, "learning_rate": 8.207586407887589e-05, "loss": 0.0268, "step": 31440 }, { "grad_norm": 0.254018098115921, "learning_rate": 8.206317841634148e-05, "loss": 0.0284, "step": 31450 }, { "grad_norm": 0.28396427631378174, "learning_rate": 8.205048924742609e-05, "loss": 0.0323, "step": 31460 }, { "grad_norm": 0.1986996978521347, "learning_rate": 8.203779657351738e-05, "loss": 0.028, "step": 31470 }, { "grad_norm": 0.21586200594902039, "learning_rate": 8.20251003960034e-05, "loss": 0.0294, "step": 31480 }, { "grad_norm": 0.29030841588974, "learning_rate": 8.201240071627258e-05, "loss": 0.0263, "step": 31490 }, { "grad_norm": 0.29492947459220886, "learning_rate": 8.199969753571377e-05, "loss": 0.0303, "step": 31500 }, { "grad_norm": 0.2373719960451126, "learning_rate": 8.198699085571615e-05, "loss": 0.0255, "step": 31510 }, { "grad_norm": 0.31844276189804077, "learning_rate": 8.197428067766928e-05, "loss": 0.0322, "step": 31520 }, { "grad_norm": 0.3507225215435028, "learning_rate": 8.196156700296316e-05, "loss": 0.0275, "step": 31530 }, { "grad_norm": 0.23421362042427063, "learning_rate": 8.194884983298814e-05, "loss": 0.034, "step": 31540 }, { "grad_norm": 0.28318825364112854, "learning_rate": 8.193612916913491e-05, "loss": 0.0275, "step": 31550 }, { "grad_norm": 0.24463504552841187, "learning_rate": 8.192340501279463e-05, "loss": 0.0305, "step": 31560 }, { "grad_norm": 0.21395041048526764, "learning_rate": 8.191067736535876e-05, "loss": 0.028, "step": 31570 }, { "grad_norm": 0.2372986525297165, "learning_rate": 8.18979462282192e-05, "loss": 0.0294, "step": 31580 }, { "grad_norm": 0.2825263738632202, "learning_rate": 8.188521160276819e-05, "loss": 0.0326, "step": 31590 }, { "grad_norm": 0.25210434198379517, "learning_rate": 8.187247349039837e-05, "loss": 0.025, "step": 31600 }, { "grad_norm": 0.2748572826385498, "learning_rate": 8.185973189250278e-05, "loss": 0.0288, "step": 31610 }, { "grad_norm": 0.18386144936084747, "learning_rate": 8.184698681047482e-05, "loss": 0.0263, "step": 31620 }, { "grad_norm": 0.23688752949237823, "learning_rate": 8.183423824570827e-05, "loss": 0.0376, "step": 31630 }, { "grad_norm": 0.20173926651477814, "learning_rate": 8.182148619959725e-05, "loss": 0.0262, "step": 31640 }, { "grad_norm": 0.2169017493724823, "learning_rate": 8.180873067353636e-05, "loss": 0.0287, "step": 31650 }, { "grad_norm": 0.25344574451446533, "learning_rate": 8.179597166892052e-05, "loss": 0.0282, "step": 31660 }, { "grad_norm": 0.2422330379486084, "learning_rate": 8.178320918714501e-05, "loss": 0.0321, "step": 31670 }, { "grad_norm": 0.25233030319213867, "learning_rate": 8.177044322960554e-05, "loss": 0.0259, "step": 31680 }, { "grad_norm": 0.30419570207595825, "learning_rate": 8.175767379769816e-05, "loss": 0.0287, "step": 31690 }, { "grad_norm": 0.3072051703929901, "learning_rate": 8.174490089281932e-05, "loss": 0.0259, "step": 31700 }, { "grad_norm": 0.249332994222641, "learning_rate": 8.173212451636584e-05, "loss": 0.0264, "step": 31710 }, { "grad_norm": 0.21614645421504974, "learning_rate": 8.171934466973493e-05, "loss": 0.0331, "step": 31720 }, { "grad_norm": 0.29404908418655396, "learning_rate": 8.170656135432418e-05, "loss": 0.0283, "step": 31730 }, { "grad_norm": 0.1812821626663208, "learning_rate": 8.169377457153155e-05, "loss": 0.0247, "step": 31740 }, { "grad_norm": 0.279453843832016, "learning_rate": 8.168098432275539e-05, "loss": 0.0302, "step": 31750 }, { "grad_norm": 0.2369987815618515, "learning_rate": 8.166819060939442e-05, "loss": 0.0308, "step": 31760 }, { "grad_norm": 0.2213619500398636, "learning_rate": 8.165539343284772e-05, "loss": 0.0329, "step": 31770 }, { "grad_norm": 0.39353883266448975, "learning_rate": 8.16425927945148e-05, "loss": 0.0264, "step": 31780 }, { "grad_norm": 0.21948063373565674, "learning_rate": 8.162978869579551e-05, "loss": 0.0303, "step": 31790 }, { "grad_norm": 0.26890408992767334, "learning_rate": 8.161698113809007e-05, "loss": 0.0342, "step": 31800 }, { "grad_norm": 0.2593308389186859, "learning_rate": 8.160417012279911e-05, "loss": 0.0311, "step": 31810 }, { "grad_norm": 0.22331593930721283, "learning_rate": 8.159135565132363e-05, "loss": 0.0295, "step": 31820 }, { "grad_norm": 0.23564830422401428, "learning_rate": 8.157853772506498e-05, "loss": 0.0328, "step": 31830 }, { "grad_norm": 0.22768086194992065, "learning_rate": 8.156571634542494e-05, "loss": 0.0338, "step": 31840 }, { "grad_norm": 0.2737153172492981, "learning_rate": 8.15528915138056e-05, "loss": 0.0276, "step": 31850 }, { "grad_norm": 0.24209614098072052, "learning_rate": 8.154006323160949e-05, "loss": 0.0283, "step": 31860 }, { "grad_norm": 0.23062960803508759, "learning_rate": 8.152723150023949e-05, "loss": 0.0272, "step": 31870 }, { "grad_norm": 0.25308191776275635, "learning_rate": 8.151439632109886e-05, "loss": 0.0308, "step": 31880 }, { "grad_norm": 0.252901554107666, "learning_rate": 8.150155769559122e-05, "loss": 0.029, "step": 31890 }, { "grad_norm": 0.25356563925743103, "learning_rate": 8.148871562512058e-05, "loss": 0.0316, "step": 31900 }, { "grad_norm": 0.26228946447372437, "learning_rate": 8.147587011109136e-05, "loss": 0.0292, "step": 31910 }, { "grad_norm": 0.29329201579093933, "learning_rate": 8.14630211549083e-05, "loss": 0.0248, "step": 31920 }, { "grad_norm": 0.2599695026874542, "learning_rate": 8.145016875797655e-05, "loss": 0.0275, "step": 31930 }, { "grad_norm": 0.24190981686115265, "learning_rate": 8.143731292170164e-05, "loss": 0.0277, "step": 31940 }, { "grad_norm": 0.22502732276916504, "learning_rate": 8.142445364748944e-05, "loss": 0.0281, "step": 31950 }, { "grad_norm": 0.23216819763183594, "learning_rate": 8.141159093674624e-05, "loss": 0.0269, "step": 31960 }, { "grad_norm": 0.19045746326446533, "learning_rate": 8.139872479087869e-05, "loss": 0.0278, "step": 31970 }, { "grad_norm": 0.2836018204689026, "learning_rate": 8.13858552112938e-05, "loss": 0.0297, "step": 31980 }, { "grad_norm": 0.21791790425777435, "learning_rate": 8.137298219939895e-05, "loss": 0.0316, "step": 31990 }, { "grad_norm": 0.22430147230625153, "learning_rate": 8.136010575660196e-05, "loss": 0.0297, "step": 32000 }, { "grad_norm": 0.2687208950519562, "learning_rate": 8.134722588431095e-05, "loss": 0.0292, "step": 32010 }, { "grad_norm": 0.26334962248802185, "learning_rate": 8.133434258393444e-05, "loss": 0.03, "step": 32020 }, { "grad_norm": 0.2700531482696533, "learning_rate": 8.132145585688134e-05, "loss": 0.0299, "step": 32030 }, { "grad_norm": 0.5601339340209961, "learning_rate": 8.130856570456093e-05, "loss": 0.0303, "step": 32040 }, { "grad_norm": 0.26591214537620544, "learning_rate": 8.129567212838283e-05, "loss": 0.0287, "step": 32050 }, { "grad_norm": 0.2564968168735504, "learning_rate": 8.128277512975708e-05, "loss": 0.0284, "step": 32060 }, { "grad_norm": 0.2954590320587158, "learning_rate": 8.126987471009408e-05, "loss": 0.0292, "step": 32070 }, { "grad_norm": 0.24040666222572327, "learning_rate": 8.125697087080459e-05, "loss": 0.0276, "step": 32080 }, { "grad_norm": 0.3046746551990509, "learning_rate": 8.124406361329976e-05, "loss": 0.0302, "step": 32090 }, { "grad_norm": 0.24515485763549805, "learning_rate": 8.123115293899111e-05, "loss": 0.0279, "step": 32100 }, { "grad_norm": 0.2788822054862976, "learning_rate": 8.121823884929055e-05, "loss": 0.0281, "step": 32110 }, { "grad_norm": 0.31691858172416687, "learning_rate": 8.12053213456103e-05, "loss": 0.0325, "step": 32120 }, { "grad_norm": 0.18601000308990479, "learning_rate": 8.119240042936303e-05, "loss": 0.0295, "step": 32130 }, { "grad_norm": 0.24179334938526154, "learning_rate": 8.117947610196175e-05, "loss": 0.0277, "step": 32140 }, { "grad_norm": 0.2546137571334839, "learning_rate": 8.116654836481982e-05, "loss": 0.0291, "step": 32150 }, { "grad_norm": 0.19925658404827118, "learning_rate": 8.115361721935106e-05, "loss": 0.0306, "step": 32160 }, { "grad_norm": 0.1802958846092224, "learning_rate": 8.114068266696953e-05, "loss": 0.0253, "step": 32170 }, { "grad_norm": 0.2570064067840576, "learning_rate": 8.112774470908978e-05, "loss": 0.0282, "step": 32180 }, { "grad_norm": 0.2760155200958252, "learning_rate": 8.111480334712665e-05, "loss": 0.0278, "step": 32190 }, { "grad_norm": 0.2798044979572296, "learning_rate": 8.110185858249542e-05, "loss": 0.0281, "step": 32200 }, { "grad_norm": 0.22216574847698212, "learning_rate": 8.108891041661168e-05, "loss": 0.0268, "step": 32210 }, { "grad_norm": 0.2875853180885315, "learning_rate": 8.107595885089146e-05, "loss": 0.0279, "step": 32220 }, { "grad_norm": 0.19773222506046295, "learning_rate": 8.106300388675108e-05, "loss": 0.0315, "step": 32230 }, { "grad_norm": 0.2240612953901291, "learning_rate": 8.105004552560731e-05, "loss": 0.0261, "step": 32240 }, { "grad_norm": 0.2833431363105774, "learning_rate": 8.103708376887724e-05, "loss": 0.0309, "step": 32250 }, { "grad_norm": 0.384967565536499, "learning_rate": 8.102411861797836e-05, "loss": 0.0313, "step": 32260 }, { "grad_norm": 0.31133735179901123, "learning_rate": 8.101115007432851e-05, "loss": 0.0311, "step": 32270 }, { "grad_norm": 0.2991024851799011, "learning_rate": 8.09981781393459e-05, "loss": 0.0274, "step": 32280 }, { "grad_norm": 0.20784388482570648, "learning_rate": 8.098520281444915e-05, "loss": 0.0276, "step": 32290 }, { "grad_norm": 0.3537066876888275, "learning_rate": 8.09722241010572e-05, "loss": 0.0307, "step": 32300 }, { "grad_norm": 0.2028840035200119, "learning_rate": 8.095924200058939e-05, "loss": 0.0291, "step": 32310 }, { "grad_norm": 0.27430540323257446, "learning_rate": 8.094625651446541e-05, "loss": 0.0302, "step": 32320 }, { "grad_norm": 0.2029874175786972, "learning_rate": 8.093326764410536e-05, "loss": 0.0252, "step": 32330 }, { "grad_norm": 0.23889531195163727, "learning_rate": 8.092027539092966e-05, "loss": 0.0277, "step": 32340 }, { "grad_norm": 0.21246632933616638, "learning_rate": 8.090727975635913e-05, "loss": 0.0287, "step": 32350 }, { "grad_norm": 0.26025572419166565, "learning_rate": 8.089428074181497e-05, "loss": 0.0284, "step": 32360 }, { "grad_norm": 0.20515136420726776, "learning_rate": 8.088127834871871e-05, "loss": 0.0253, "step": 32370 }, { "grad_norm": 0.31286895275115967, "learning_rate": 8.086827257849226e-05, "loss": 0.0286, "step": 32380 }, { "grad_norm": 0.20511676371097565, "learning_rate": 8.085526343255795e-05, "loss": 0.0251, "step": 32390 }, { "grad_norm": 0.27081990242004395, "learning_rate": 8.084225091233842e-05, "loss": 0.0286, "step": 32400 }, { "grad_norm": 0.27655258774757385, "learning_rate": 8.082923501925668e-05, "loss": 0.0273, "step": 32410 }, { "grad_norm": 0.21555398404598236, "learning_rate": 8.081621575473617e-05, "loss": 0.029, "step": 32420 }, { "grad_norm": 0.23464268445968628, "learning_rate": 8.080319312020064e-05, "loss": 0.0305, "step": 32430 }, { "grad_norm": 0.22130830585956573, "learning_rate": 8.079016711707421e-05, "loss": 0.028, "step": 32440 }, { "grad_norm": 0.24443599581718445, "learning_rate": 8.077713774678139e-05, "loss": 0.0267, "step": 32450 }, { "grad_norm": 0.30407363176345825, "learning_rate": 8.076410501074707e-05, "loss": 0.0295, "step": 32460 }, { "grad_norm": 0.22925132513046265, "learning_rate": 8.075106891039647e-05, "loss": 0.03, "step": 32470 }, { "grad_norm": 0.25686973333358765, "learning_rate": 8.073802944715523e-05, "loss": 0.0259, "step": 32480 }, { "grad_norm": 0.2667808532714844, "learning_rate": 8.072498662244929e-05, "loss": 0.0285, "step": 32490 }, { "grad_norm": 0.2100338339805603, "learning_rate": 8.0711940437705e-05, "loss": 0.0276, "step": 32500 }, { "grad_norm": 0.2442510426044464, "learning_rate": 8.06988908943491e-05, "loss": 0.0276, "step": 32510 }, { "grad_norm": 0.2561390697956085, "learning_rate": 8.068583799380863e-05, "loss": 0.0346, "step": 32520 }, { "grad_norm": 0.2694539427757263, "learning_rate": 8.067278173751104e-05, "loss": 0.0286, "step": 32530 }, { "grad_norm": 0.23241347074508667, "learning_rate": 8.065972212688417e-05, "loss": 0.0289, "step": 32540 }, { "grad_norm": 0.24601460993289948, "learning_rate": 8.064665916335618e-05, "loss": 0.025, "step": 32550 }, { "grad_norm": 0.2782860994338989, "learning_rate": 8.063359284835564e-05, "loss": 0.0366, "step": 32560 }, { "grad_norm": 0.2820891737937927, "learning_rate": 8.062052318331142e-05, "loss": 0.0256, "step": 32570 }, { "grad_norm": 0.1878107637166977, "learning_rate": 8.060745016965283e-05, "loss": 0.0211, "step": 32580 }, { "grad_norm": 0.2576199471950531, "learning_rate": 8.059437380880952e-05, "loss": 0.0288, "step": 32590 }, { "grad_norm": 0.195781871676445, "learning_rate": 8.058129410221146e-05, "loss": 0.0268, "step": 32600 }, { "grad_norm": 0.2179618775844574, "learning_rate": 8.056821105128908e-05, "loss": 0.0284, "step": 32610 }, { "grad_norm": 0.2813844084739685, "learning_rate": 8.05551246574731e-05, "loss": 0.0312, "step": 32620 }, { "grad_norm": 0.351678729057312, "learning_rate": 8.05420349221946e-05, "loss": 0.0299, "step": 32630 }, { "grad_norm": 0.28308290243148804, "learning_rate": 8.05289418468851e-05, "loss": 0.0271, "step": 32640 }, { "grad_norm": 0.23274850845336914, "learning_rate": 8.051584543297642e-05, "loss": 0.0269, "step": 32650 }, { "grad_norm": 0.29801493883132935, "learning_rate": 8.050274568190074e-05, "loss": 0.0266, "step": 32660 }, { "grad_norm": 0.25390851497650146, "learning_rate": 8.048964259509067e-05, "loss": 0.0242, "step": 32670 }, { "grad_norm": 0.24380671977996826, "learning_rate": 8.047653617397914e-05, "loss": 0.0253, "step": 32680 }, { "grad_norm": 0.2316248118877411, "learning_rate": 8.046342641999941e-05, "loss": 0.0265, "step": 32690 }, { "grad_norm": 0.24585942924022675, "learning_rate": 8.045031333458517e-05, "loss": 0.0305, "step": 32700 }, { "grad_norm": 0.23968222737312317, "learning_rate": 8.043719691917047e-05, "loss": 0.0305, "step": 32710 }, { "grad_norm": 0.31191858649253845, "learning_rate": 8.042407717518966e-05, "loss": 0.0281, "step": 32720 }, { "grad_norm": 0.26367640495300293, "learning_rate": 8.041095410407751e-05, "loss": 0.0269, "step": 32730 }, { "grad_norm": 0.23166930675506592, "learning_rate": 8.039782770726913e-05, "loss": 0.0227, "step": 32740 }, { "grad_norm": 0.3205086290836334, "learning_rate": 8.038469798620004e-05, "loss": 0.0272, "step": 32750 }, { "grad_norm": 0.2484693080186844, "learning_rate": 8.037156494230604e-05, "loss": 0.0281, "step": 32760 }, { "grad_norm": 0.15091022849082947, "learning_rate": 8.035842857702338e-05, "loss": 0.0236, "step": 32770 }, { "grad_norm": 0.34120991826057434, "learning_rate": 8.03452888917886e-05, "loss": 0.03, "step": 32780 }, { "grad_norm": 0.2848016321659088, "learning_rate": 8.033214588803866e-05, "loss": 0.0343, "step": 32790 }, { "grad_norm": 0.25190460681915283, "learning_rate": 8.031899956721083e-05, "loss": 0.0322, "step": 32800 }, { "grad_norm": 0.22240616381168365, "learning_rate": 8.030584993074282e-05, "loss": 0.0284, "step": 32810 }, { "grad_norm": 0.3493054211139679, "learning_rate": 8.02926969800726e-05, "loss": 0.0287, "step": 32820 }, { "grad_norm": 0.2156175673007965, "learning_rate": 8.027954071663859e-05, "loss": 0.0284, "step": 32830 }, { "grad_norm": 0.21712851524353027, "learning_rate": 8.026638114187954e-05, "loss": 0.0258, "step": 32840 }, { "grad_norm": 0.25082385540008545, "learning_rate": 8.025321825723456e-05, "loss": 0.0226, "step": 32850 }, { "grad_norm": 0.2481037974357605, "learning_rate": 8.02400520641431e-05, "loss": 0.0263, "step": 32860 }, { "grad_norm": 0.23332267999649048, "learning_rate": 8.022688256404501e-05, "loss": 0.028, "step": 32870 }, { "grad_norm": 0.2961257994174957, "learning_rate": 8.02137097583805e-05, "loss": 0.023, "step": 32880 }, { "grad_norm": 0.2620024085044861, "learning_rate": 8.02005336485901e-05, "loss": 0.029, "step": 32890 }, { "grad_norm": 0.22286807000637054, "learning_rate": 8.018735423611476e-05, "loss": 0.0264, "step": 32900 }, { "grad_norm": 0.387172669172287, "learning_rate": 8.017417152239574e-05, "loss": 0.0282, "step": 32910 }, { "grad_norm": 0.2779003381729126, "learning_rate": 8.01609855088747e-05, "loss": 0.029, "step": 32920 }, { "grad_norm": 0.25625574588775635, "learning_rate": 8.014779619699362e-05, "loss": 0.0297, "step": 32930 }, { "grad_norm": 0.2590630054473877, "learning_rate": 8.013460358819489e-05, "loss": 0.026, "step": 32940 }, { "grad_norm": 0.2428477555513382, "learning_rate": 8.01214076839212e-05, "loss": 0.0296, "step": 32950 }, { "grad_norm": 0.20444223284721375, "learning_rate": 8.010820848561565e-05, "loss": 0.025, "step": 32960 }, { "grad_norm": 0.3087635934352875, "learning_rate": 8.009500599472171e-05, "loss": 0.0271, "step": 32970 }, { "grad_norm": 0.2717059254646301, "learning_rate": 8.008180021268314e-05, "loss": 0.023, "step": 32980 }, { "grad_norm": 0.260881632566452, "learning_rate": 8.006859114094414e-05, "loss": 0.0243, "step": 32990 }, { "grad_norm": 0.19834278523921967, "learning_rate": 8.005537878094921e-05, "loss": 0.0289, "step": 33000 }, { "grad_norm": 0.240240678191185, "learning_rate": 8.004216313414323e-05, "loss": 0.023, "step": 33010 }, { "grad_norm": 0.28697916865348816, "learning_rate": 8.002894420197149e-05, "loss": 0.0244, "step": 33020 }, { "grad_norm": 0.23574188351631165, "learning_rate": 8.001572198587954e-05, "loss": 0.0231, "step": 33030 }, { "grad_norm": 0.2960185110569, "learning_rate": 8.000249648731338e-05, "loss": 0.0269, "step": 33040 }, { "grad_norm": 0.23912619054317474, "learning_rate": 7.998926770771928e-05, "loss": 0.0257, "step": 33050 }, { "grad_norm": 0.2270713895559311, "learning_rate": 7.997603564854397e-05, "loss": 0.0274, "step": 33060 }, { "grad_norm": 0.2508607804775238, "learning_rate": 7.996280031123448e-05, "loss": 0.0265, "step": 33070 }, { "grad_norm": 0.19532568752765656, "learning_rate": 7.994956169723818e-05, "loss": 0.0243, "step": 33080 }, { "grad_norm": 0.2228633463382721, "learning_rate": 7.993631980800285e-05, "loss": 0.0238, "step": 33090 }, { "grad_norm": 0.28640785813331604, "learning_rate": 7.992307464497659e-05, "loss": 0.0253, "step": 33100 }, { "grad_norm": 0.26056423783302307, "learning_rate": 7.990982620960787e-05, "loss": 0.0306, "step": 33110 }, { "grad_norm": 0.3487948179244995, "learning_rate": 7.989657450334554e-05, "loss": 0.0285, "step": 33120 }, { "grad_norm": 0.26015788316726685, "learning_rate": 7.988331952763877e-05, "loss": 0.0299, "step": 33130 }, { "grad_norm": 0.4918980896472931, "learning_rate": 7.987006128393709e-05, "loss": 0.0279, "step": 33140 }, { "grad_norm": 0.29793286323547363, "learning_rate": 7.985679977369043e-05, "loss": 0.0282, "step": 33150 }, { "grad_norm": 0.24412454664707184, "learning_rate": 7.984353499834902e-05, "loss": 0.0242, "step": 33160 }, { "grad_norm": 0.24445556104183197, "learning_rate": 7.983026695936351e-05, "loss": 0.0297, "step": 33170 }, { "grad_norm": 0.2963937819004059, "learning_rate": 7.981699565818486e-05, "loss": 0.0302, "step": 33180 }, { "grad_norm": 0.2683367133140564, "learning_rate": 7.980372109626437e-05, "loss": 0.0272, "step": 33190 }, { "grad_norm": 0.2452850043773651, "learning_rate": 7.979044327505375e-05, "loss": 0.0265, "step": 33200 }, { "grad_norm": 0.20841918885707855, "learning_rate": 7.977716219600506e-05, "loss": 0.027, "step": 33210 }, { "grad_norm": 0.23102477192878723, "learning_rate": 7.97638778605707e-05, "loss": 0.0254, "step": 33220 }, { "grad_norm": 0.24557746946811676, "learning_rate": 7.975059027020338e-05, "loss": 0.0257, "step": 33230 }, { "grad_norm": 0.29123735427856445, "learning_rate": 7.973729942635623e-05, "loss": 0.0257, "step": 33240 }, { "grad_norm": 0.2630877196788788, "learning_rate": 7.972400533048273e-05, "loss": 0.0299, "step": 33250 }, { "grad_norm": 0.2364790439605713, "learning_rate": 7.97107079840367e-05, "loss": 0.0266, "step": 33260 }, { "grad_norm": 0.21354690194129944, "learning_rate": 7.969740738847231e-05, "loss": 0.0267, "step": 33270 }, { "grad_norm": 0.19679395854473114, "learning_rate": 7.968410354524411e-05, "loss": 0.0244, "step": 33280 }, { "grad_norm": 0.2726512849330902, "learning_rate": 7.967079645580697e-05, "loss": 0.0308, "step": 33290 }, { "grad_norm": 0.29610708355903625, "learning_rate": 7.965748612161612e-05, "loss": 0.0361, "step": 33300 }, { "grad_norm": 0.26399821043014526, "learning_rate": 7.96441725441272e-05, "loss": 0.0279, "step": 33310 }, { "grad_norm": 0.24486960470676422, "learning_rate": 7.963085572479614e-05, "loss": 0.0263, "step": 33320 }, { "grad_norm": 0.27857494354248047, "learning_rate": 7.961753566507924e-05, "loss": 0.0256, "step": 33330 }, { "grad_norm": 0.35807129740715027, "learning_rate": 7.960421236643316e-05, "loss": 0.0402, "step": 33340 }, { "grad_norm": 0.2727981507778168, "learning_rate": 7.959088583031496e-05, "loss": 0.0271, "step": 33350 }, { "grad_norm": 0.22893580794334412, "learning_rate": 7.957755605818194e-05, "loss": 0.0329, "step": 33360 }, { "grad_norm": 0.21423983573913574, "learning_rate": 7.956422305149185e-05, "loss": 0.0254, "step": 33370 }, { "grad_norm": 0.45540061593055725, "learning_rate": 7.95508868117028e-05, "loss": 0.0281, "step": 33380 }, { "grad_norm": 0.3378423750400543, "learning_rate": 7.953754734027318e-05, "loss": 0.0266, "step": 33390 }, { "grad_norm": 0.2035924345254898, "learning_rate": 7.952420463866182e-05, "loss": 0.0278, "step": 33400 }, { "grad_norm": 0.28425154089927673, "learning_rate": 7.951085870832782e-05, "loss": 0.0273, "step": 33410 }, { "grad_norm": 0.43088024854660034, "learning_rate": 7.949750955073067e-05, "loss": 0.029, "step": 33420 }, { "grad_norm": 0.26267239451408386, "learning_rate": 7.948415716733022e-05, "loss": 0.0277, "step": 33430 }, { "grad_norm": 0.25630566477775574, "learning_rate": 7.947080155958669e-05, "loss": 0.0283, "step": 33440 }, { "grad_norm": 0.3169988989830017, "learning_rate": 7.94574427289606e-05, "loss": 0.0256, "step": 33450 }, { "grad_norm": 0.30572637915611267, "learning_rate": 7.944408067691284e-05, "loss": 0.0253, "step": 33460 }, { "grad_norm": 0.31735217571258545, "learning_rate": 7.943071540490473e-05, "loss": 0.0321, "step": 33470 }, { "grad_norm": 0.30143672227859497, "learning_rate": 7.94173469143978e-05, "loss": 0.0322, "step": 33480 }, { "grad_norm": 0.36679738759994507, "learning_rate": 7.940397520685406e-05, "loss": 0.035, "step": 33490 }, { "grad_norm": 0.21417686343193054, "learning_rate": 7.939060028373577e-05, "loss": 0.0267, "step": 33500 }, { "grad_norm": 0.24907377362251282, "learning_rate": 7.937722214650565e-05, "loss": 0.0275, "step": 33510 }, { "grad_norm": 0.2947236895561218, "learning_rate": 7.936384079662666e-05, "loss": 0.0291, "step": 33520 }, { "grad_norm": 0.23161381483078003, "learning_rate": 7.93504562355622e-05, "loss": 0.0319, "step": 33530 }, { "grad_norm": 0.31469589471817017, "learning_rate": 7.933706846477599e-05, "loss": 0.0305, "step": 33540 }, { "grad_norm": 0.3460402190685272, "learning_rate": 7.932367748573206e-05, "loss": 0.0333, "step": 33550 }, { "grad_norm": 0.223121777176857, "learning_rate": 7.931028329989485e-05, "loss": 0.0261, "step": 33560 }, { "grad_norm": 0.17454299330711365, "learning_rate": 7.929688590872913e-05, "loss": 0.0305, "step": 33570 }, { "grad_norm": 0.25162267684936523, "learning_rate": 7.928348531370003e-05, "loss": 0.0228, "step": 33580 }, { "grad_norm": 0.27953892946243286, "learning_rate": 7.927008151627297e-05, "loss": 0.0258, "step": 33590 }, { "grad_norm": 0.20806971192359924, "learning_rate": 7.925667451791383e-05, "loss": 0.0242, "step": 33600 }, { "grad_norm": 0.3585371971130371, "learning_rate": 7.924326432008874e-05, "loss": 0.024, "step": 33610 }, { "grad_norm": 0.19755956530570984, "learning_rate": 7.922985092426422e-05, "loss": 0.0251, "step": 33620 }, { "grad_norm": 0.25196000933647156, "learning_rate": 7.921643433190717e-05, "loss": 0.0309, "step": 33630 }, { "grad_norm": 0.19415564835071564, "learning_rate": 7.920301454448478e-05, "loss": 0.0299, "step": 33640 }, { "grad_norm": 0.24654385447502136, "learning_rate": 7.918959156346461e-05, "loss": 0.0256, "step": 33650 }, { "grad_norm": 0.2535514235496521, "learning_rate": 7.91761653903146e-05, "loss": 0.0295, "step": 33660 }, { "grad_norm": 0.30282706022262573, "learning_rate": 7.916273602650302e-05, "loss": 0.0287, "step": 33670 }, { "grad_norm": 0.2751908600330353, "learning_rate": 7.914930347349847e-05, "loss": 0.0242, "step": 33680 }, { "grad_norm": 0.2363734394311905, "learning_rate": 7.913586773276992e-05, "loss": 0.0302, "step": 33690 }, { "grad_norm": 0.28850361704826355, "learning_rate": 7.912242880578667e-05, "loss": 0.0329, "step": 33700 }, { "grad_norm": 0.2979849576950073, "learning_rate": 7.910898669401839e-05, "loss": 0.0287, "step": 33710 }, { "grad_norm": 0.20745901763439178, "learning_rate": 7.909554139893511e-05, "loss": 0.0267, "step": 33720 }, { "grad_norm": 0.1998852640390396, "learning_rate": 7.908209292200715e-05, "loss": 0.0265, "step": 33730 }, { "grad_norm": 0.2775636911392212, "learning_rate": 7.906864126470523e-05, "loss": 0.0301, "step": 33740 }, { "grad_norm": 0.31761226058006287, "learning_rate": 7.905518642850041e-05, "loss": 0.0263, "step": 33750 }, { "grad_norm": 0.22468553483486176, "learning_rate": 7.904172841486409e-05, "loss": 0.0279, "step": 33760 }, { "grad_norm": 0.2173033505678177, "learning_rate": 7.902826722526801e-05, "loss": 0.0299, "step": 33770 }, { "grad_norm": 0.2087104767560959, "learning_rate": 7.901480286118427e-05, "loss": 0.0285, "step": 33780 }, { "grad_norm": 0.24946659803390503, "learning_rate": 7.900133532408531e-05, "loss": 0.0264, "step": 33790 }, { "grad_norm": 0.17098303139209747, "learning_rate": 7.898786461544395e-05, "loss": 0.0273, "step": 33800 }, { "grad_norm": 0.219408318400383, "learning_rate": 7.897439073673325e-05, "loss": 0.0235, "step": 33810 }, { "grad_norm": 0.19266849756240845, "learning_rate": 7.896091368942677e-05, "loss": 0.0276, "step": 33820 }, { "grad_norm": 0.27883997559547424, "learning_rate": 7.894743347499832e-05, "loss": 0.0313, "step": 33830 }, { "grad_norm": 0.3353646397590637, "learning_rate": 7.893395009492203e-05, "loss": 0.031, "step": 33840 }, { "grad_norm": 0.3438062071800232, "learning_rate": 7.892046355067248e-05, "loss": 0.0278, "step": 33850 }, { "grad_norm": 0.33726927638053894, "learning_rate": 7.890697384372451e-05, "loss": 0.024, "step": 33860 }, { "grad_norm": 0.23359838128089905, "learning_rate": 7.889348097555336e-05, "loss": 0.026, "step": 33870 }, { "grad_norm": 0.38811972737312317, "learning_rate": 7.887998494763455e-05, "loss": 0.0275, "step": 33880 }, { "grad_norm": 0.22880983352661133, "learning_rate": 7.886648576144404e-05, "loss": 0.0271, "step": 33890 }, { "grad_norm": 0.4708065092563629, "learning_rate": 7.885298341845802e-05, "loss": 0.0288, "step": 33900 }, { "grad_norm": 0.2525084614753723, "learning_rate": 7.883947792015311e-05, "loss": 0.0276, "step": 33910 }, { "grad_norm": 0.2279070019721985, "learning_rate": 7.882596926800628e-05, "loss": 0.0335, "step": 33920 }, { "grad_norm": 0.2580150067806244, "learning_rate": 7.881245746349477e-05, "loss": 0.0311, "step": 33930 }, { "grad_norm": 0.30124765634536743, "learning_rate": 7.879894250809623e-05, "loss": 0.0316, "step": 33940 }, { "grad_norm": 1.2045992612838745, "learning_rate": 7.878542440328865e-05, "loss": 0.0306, "step": 33950 }, { "grad_norm": 0.23254750669002533, "learning_rate": 7.877190315055031e-05, "loss": 0.0274, "step": 33960 }, { "grad_norm": 0.32235729694366455, "learning_rate": 7.875837875135991e-05, "loss": 0.0268, "step": 33970 }, { "grad_norm": 0.287272185087204, "learning_rate": 7.874485120719646e-05, "loss": 0.0288, "step": 33980 }, { "grad_norm": 0.2521454095840454, "learning_rate": 7.873132051953928e-05, "loss": 0.0284, "step": 33990 }, { "grad_norm": 0.25716742873191833, "learning_rate": 7.87177866898681e-05, "loss": 0.0345, "step": 34000 }, { "grad_norm": 0.2838342785835266, "learning_rate": 7.870424971966294e-05, "loss": 0.0331, "step": 34010 }, { "grad_norm": 0.31785106658935547, "learning_rate": 7.869070961040419e-05, "loss": 0.0344, "step": 34020 }, { "grad_norm": 0.2265966385602951, "learning_rate": 7.867716636357257e-05, "loss": 0.0307, "step": 34030 }, { "grad_norm": 0.21650278568267822, "learning_rate": 7.866361998064915e-05, "loss": 0.0268, "step": 34040 }, { "grad_norm": 0.17994067072868347, "learning_rate": 7.865007046311534e-05, "loss": 0.0282, "step": 34050 }, { "grad_norm": 0.2898881733417511, "learning_rate": 7.86365178124529e-05, "loss": 0.0295, "step": 34060 }, { "grad_norm": 0.2664087414741516, "learning_rate": 7.862296203014394e-05, "loss": 0.0279, "step": 34070 }, { "grad_norm": 0.2991742491722107, "learning_rate": 7.860940311767088e-05, "loss": 0.0354, "step": 34080 }, { "grad_norm": 0.2211994081735611, "learning_rate": 7.85958410765165e-05, "loss": 0.0313, "step": 34090 }, { "grad_norm": 0.2923294007778168, "learning_rate": 7.858227590816394e-05, "loss": 0.028, "step": 34100 }, { "grad_norm": 0.2310945838689804, "learning_rate": 7.856870761409664e-05, "loss": 0.0268, "step": 34110 }, { "grad_norm": 0.29245275259017944, "learning_rate": 7.855513619579846e-05, "loss": 0.0278, "step": 34120 }, { "grad_norm": 0.30764028429985046, "learning_rate": 7.85415616547535e-05, "loss": 0.0321, "step": 34130 }, { "grad_norm": 0.23330409824848175, "learning_rate": 7.852798399244627e-05, "loss": 0.0299, "step": 34140 }, { "grad_norm": 0.30527162551879883, "learning_rate": 7.851440321036161e-05, "loss": 0.0277, "step": 34150 }, { "grad_norm": 0.22968310117721558, "learning_rate": 7.850081930998468e-05, "loss": 0.0276, "step": 34160 }, { "grad_norm": 0.2831929922103882, "learning_rate": 7.8487232292801e-05, "loss": 0.0321, "step": 34170 }, { "grad_norm": 0.3613317906856537, "learning_rate": 7.847364216029642e-05, "loss": 0.027, "step": 34180 }, { "grad_norm": 0.18847770988941193, "learning_rate": 7.846004891395716e-05, "loss": 0.0218, "step": 34190 }, { "grad_norm": 0.22754999995231628, "learning_rate": 7.844645255526972e-05, "loss": 0.0277, "step": 34200 }, { "grad_norm": 0.3214728534221649, "learning_rate": 7.843285308572101e-05, "loss": 0.0292, "step": 34210 }, { "grad_norm": 0.2832523286342621, "learning_rate": 7.841925050679823e-05, "loss": 0.0236, "step": 34220 }, { "grad_norm": 0.2934718132019043, "learning_rate": 7.840564481998895e-05, "loss": 0.024, "step": 34230 }, { "grad_norm": 0.247630774974823, "learning_rate": 7.839203602678105e-05, "loss": 0.029, "step": 34240 }, { "grad_norm": 0.2034345120191574, "learning_rate": 7.837842412866279e-05, "loss": 0.0238, "step": 34250 }, { "grad_norm": 0.20434434711933136, "learning_rate": 7.836480912712272e-05, "loss": 0.0259, "step": 34260 }, { "grad_norm": 0.22923094034194946, "learning_rate": 7.835119102364976e-05, "loss": 0.0288, "step": 34270 }, { "grad_norm": 0.2023334950208664, "learning_rate": 7.83375698197332e-05, "loss": 0.0275, "step": 34280 }, { "grad_norm": 0.2621515393257141, "learning_rate": 7.83239455168626e-05, "loss": 0.0246, "step": 34290 }, { "grad_norm": 0.26304683089256287, "learning_rate": 7.83103181165279e-05, "loss": 0.0251, "step": 34300 }, { "grad_norm": 0.24884387850761414, "learning_rate": 7.829668762021937e-05, "loss": 0.0303, "step": 34310 }, { "grad_norm": 0.2487470507621765, "learning_rate": 7.828305402942764e-05, "loss": 0.026, "step": 34320 }, { "grad_norm": 0.2624954581260681, "learning_rate": 7.826941734564363e-05, "loss": 0.0225, "step": 34330 }, { "grad_norm": 0.22207671403884888, "learning_rate": 7.825577757035865e-05, "loss": 0.0254, "step": 34340 }, { "grad_norm": 0.2927974462509155, "learning_rate": 7.824213470506431e-05, "loss": 0.0267, "step": 34350 }, { "grad_norm": 0.22664868831634521, "learning_rate": 7.822848875125257e-05, "loss": 0.023, "step": 34360 }, { "grad_norm": 0.2748177647590637, "learning_rate": 7.821483971041576e-05, "loss": 0.0258, "step": 34370 }, { "grad_norm": 0.24929380416870117, "learning_rate": 7.820118758404649e-05, "loss": 0.0297, "step": 34380 }, { "grad_norm": 0.26306065917015076, "learning_rate": 7.818753237363776e-05, "loss": 0.0245, "step": 34390 }, { "grad_norm": 0.22397735714912415, "learning_rate": 7.817387408068286e-05, "loss": 0.0245, "step": 34400 }, { "grad_norm": 0.21942868828773499, "learning_rate": 7.816021270667544e-05, "loss": 0.023, "step": 34410 }, { "grad_norm": 0.21271388232707977, "learning_rate": 7.81465482531095e-05, "loss": 0.0214, "step": 34420 }, { "grad_norm": 0.2772262990474701, "learning_rate": 7.813288072147938e-05, "loss": 0.0235, "step": 34430 }, { "grad_norm": 0.2573222815990448, "learning_rate": 7.811921011327972e-05, "loss": 0.0256, "step": 34440 }, { "grad_norm": 0.21141116321086884, "learning_rate": 7.810553643000549e-05, "loss": 0.0279, "step": 34450 }, { "grad_norm": 0.25456199049949646, "learning_rate": 7.809185967315206e-05, "loss": 0.03, "step": 34460 }, { "grad_norm": 0.262450248003006, "learning_rate": 7.80781798442151e-05, "loss": 0.0299, "step": 34470 }, { "grad_norm": 0.30953851342201233, "learning_rate": 7.806449694469061e-05, "loss": 0.0279, "step": 34480 }, { "grad_norm": 0.2534426152706146, "learning_rate": 7.805081097607492e-05, "loss": 0.0279, "step": 34490 }, { "grad_norm": 0.3066200911998749, "learning_rate": 7.803712193986474e-05, "loss": 0.0233, "step": 34500 }, { "grad_norm": 0.17927896976470947, "learning_rate": 7.802342983755702e-05, "loss": 0.0275, "step": 34510 }, { "grad_norm": 0.2673834562301636, "learning_rate": 7.800973467064918e-05, "loss": 0.0267, "step": 34520 }, { "grad_norm": 0.20746661722660065, "learning_rate": 7.799603644063884e-05, "loss": 0.0281, "step": 34530 }, { "grad_norm": 0.2586519718170166, "learning_rate": 7.798233514902405e-05, "loss": 0.0275, "step": 34540 }, { "grad_norm": 0.1926129162311554, "learning_rate": 7.796863079730318e-05, "loss": 0.0281, "step": 34550 }, { "grad_norm": 0.20057491958141327, "learning_rate": 7.795492338697488e-05, "loss": 0.0254, "step": 34560 }, { "grad_norm": 0.19290731847286224, "learning_rate": 7.794121291953819e-05, "loss": 0.0248, "step": 34570 }, { "grad_norm": 0.344383180141449, "learning_rate": 7.792749939649246e-05, "loss": 0.0235, "step": 34580 }, { "grad_norm": 0.2090982347726822, "learning_rate": 7.79137828193374e-05, "loss": 0.023, "step": 34590 }, { "grad_norm": 0.2155396044254303, "learning_rate": 7.790006318957301e-05, "loss": 0.0263, "step": 34600 }, { "grad_norm": 0.2599385380744934, "learning_rate": 7.788634050869965e-05, "loss": 0.0245, "step": 34610 }, { "grad_norm": 0.24533100426197052, "learning_rate": 7.787261477821803e-05, "loss": 0.0252, "step": 34620 }, { "grad_norm": 0.22837486863136292, "learning_rate": 7.785888599962916e-05, "loss": 0.03, "step": 34630 }, { "grad_norm": 0.29965412616729736, "learning_rate": 7.784515417443439e-05, "loss": 0.0233, "step": 34640 }, { "grad_norm": 0.2643623352050781, "learning_rate": 7.783141930413545e-05, "loss": 0.0285, "step": 34650 }, { "grad_norm": 0.31081724166870117, "learning_rate": 7.78176813902343e-05, "loss": 0.0274, "step": 34660 }, { "grad_norm": 0.25411200523376465, "learning_rate": 7.780394043423336e-05, "loss": 0.0274, "step": 34670 }, { "grad_norm": 0.26153460144996643, "learning_rate": 7.77901964376353e-05, "loss": 0.0251, "step": 34680 }, { "grad_norm": 0.20314380526542664, "learning_rate": 7.777644940194316e-05, "loss": 0.0264, "step": 34690 }, { "grad_norm": 0.19942311942577362, "learning_rate": 7.776269932866023e-05, "loss": 0.0261, "step": 34700 }, { "grad_norm": 0.22348786890506744, "learning_rate": 7.774894621929026e-05, "loss": 0.0271, "step": 34710 }, { "grad_norm": 0.2911343574523926, "learning_rate": 7.773519007533725e-05, "loss": 0.0247, "step": 34720 }, { "grad_norm": 0.2422582507133484, "learning_rate": 7.772143089830556e-05, "loss": 0.0283, "step": 34730 }, { "grad_norm": 0.17967069149017334, "learning_rate": 7.770766868969985e-05, "loss": 0.0246, "step": 34740 }, { "grad_norm": 0.2268131971359253, "learning_rate": 7.769390345102518e-05, "loss": 0.027, "step": 34750 }, { "grad_norm": 0.3064199388027191, "learning_rate": 7.768013518378683e-05, "loss": 0.0229, "step": 34760 }, { "grad_norm": 0.1968490332365036, "learning_rate": 7.766636388949053e-05, "loss": 0.0273, "step": 34770 }, { "grad_norm": 0.2817825973033905, "learning_rate": 7.765258956964229e-05, "loss": 0.0255, "step": 34780 }, { "grad_norm": 0.21893411874771118, "learning_rate": 7.76388122257484e-05, "loss": 0.0266, "step": 34790 }, { "grad_norm": 0.2469041794538498, "learning_rate": 7.762503185931558e-05, "loss": 0.032, "step": 34800 }, { "grad_norm": 0.34127193689346313, "learning_rate": 7.76112484718508e-05, "loss": 0.0267, "step": 34810 }, { "grad_norm": 0.2751559913158417, "learning_rate": 7.75974620648614e-05, "loss": 0.0323, "step": 34820 }, { "grad_norm": 0.18081046640872955, "learning_rate": 7.758367263985503e-05, "loss": 0.0274, "step": 34830 }, { "grad_norm": 0.17132535576820374, "learning_rate": 7.75698801983397e-05, "loss": 0.0255, "step": 34840 }, { "grad_norm": 0.19597609341144562, "learning_rate": 7.755608474182372e-05, "loss": 0.0214, "step": 34850 }, { "grad_norm": 0.29920047521591187, "learning_rate": 7.754228627181574e-05, "loss": 0.0297, "step": 34860 }, { "grad_norm": 0.18764494359493256, "learning_rate": 7.752848478982476e-05, "loss": 0.0261, "step": 34870 }, { "grad_norm": 0.2830583155155182, "learning_rate": 7.751468029736006e-05, "loss": 0.0295, "step": 34880 }, { "grad_norm": 0.20852790772914886, "learning_rate": 7.750087279593129e-05, "loss": 0.025, "step": 34890 }, { "grad_norm": 0.24424178898334503, "learning_rate": 7.748706228704843e-05, "loss": 0.0331, "step": 34900 }, { "grad_norm": 0.28328993916511536, "learning_rate": 7.747324877222176e-05, "loss": 0.0246, "step": 34910 }, { "grad_norm": 0.213545024394989, "learning_rate": 7.745943225296188e-05, "loss": 0.0249, "step": 34920 }, { "grad_norm": 0.25240829586982727, "learning_rate": 7.744561273077981e-05, "loss": 0.0271, "step": 34930 }, { "grad_norm": 0.23740297555923462, "learning_rate": 7.743179020718678e-05, "loss": 0.0277, "step": 34940 }, { "grad_norm": 0.2506614327430725, "learning_rate": 7.741796468369443e-05, "loss": 0.0277, "step": 34950 }, { "grad_norm": 0.26583877205848694, "learning_rate": 7.740413616181466e-05, "loss": 0.0252, "step": 34960 }, { "grad_norm": 0.3040630519390106, "learning_rate": 7.739030464305978e-05, "loss": 0.0292, "step": 34970 }, { "grad_norm": 0.32497552037239075, "learning_rate": 7.737647012894235e-05, "loss": 0.0253, "step": 34980 }, { "grad_norm": 0.24332040548324585, "learning_rate": 7.736263262097532e-05, "loss": 0.0242, "step": 34990 }, { "grad_norm": 0.28456911444664, "learning_rate": 7.734879212067192e-05, "loss": 0.03, "step": 35000 }, { "grad_norm": 0.22792570292949677, "learning_rate": 7.733494862954573e-05, "loss": 0.0254, "step": 35010 }, { "grad_norm": 0.19264312088489532, "learning_rate": 7.732110214911066e-05, "loss": 0.0265, "step": 35020 }, { "grad_norm": 0.16079114377498627, "learning_rate": 7.730725268088094e-05, "loss": 0.0228, "step": 35030 }, { "grad_norm": 0.23364445567131042, "learning_rate": 7.729340022637112e-05, "loss": 0.022, "step": 35040 }, { "grad_norm": 0.15285935997962952, "learning_rate": 7.727954478709607e-05, "loss": 0.0241, "step": 35050 }, { "grad_norm": 0.29813894629478455, "learning_rate": 7.726568636457103e-05, "loss": 0.0269, "step": 35060 }, { "grad_norm": 0.22363269329071045, "learning_rate": 7.725182496031153e-05, "loss": 0.0252, "step": 35070 }, { "grad_norm": 0.31649649143218994, "learning_rate": 7.723796057583341e-05, "loss": 0.026, "step": 35080 }, { "grad_norm": 0.32049301266670227, "learning_rate": 7.722409321265287e-05, "loss": 0.0293, "step": 35090 }, { "grad_norm": 0.24508391320705414, "learning_rate": 7.721022287228645e-05, "loss": 0.0262, "step": 35100 }, { "grad_norm": 0.22292225062847137, "learning_rate": 7.719634955625094e-05, "loss": 0.0286, "step": 35110 }, { "grad_norm": 0.1998397707939148, "learning_rate": 7.718247326606355e-05, "loss": 0.0252, "step": 35120 }, { "grad_norm": 0.29012301564216614, "learning_rate": 7.716859400324175e-05, "loss": 0.029, "step": 35130 }, { "grad_norm": 0.24607867002487183, "learning_rate": 7.715471176930335e-05, "loss": 0.0291, "step": 35140 }, { "grad_norm": 0.31493237614631653, "learning_rate": 7.714082656576651e-05, "loss": 0.0282, "step": 35150 }, { "grad_norm": 0.23958802223205566, "learning_rate": 7.712693839414968e-05, "loss": 0.0302, "step": 35160 }, { "grad_norm": 0.31868141889572144, "learning_rate": 7.711304725597164e-05, "loss": 0.033, "step": 35170 }, { "grad_norm": 0.20922915637493134, "learning_rate": 7.709915315275151e-05, "loss": 0.0277, "step": 35180 }, { "grad_norm": 0.2588924169540405, "learning_rate": 7.708525608600876e-05, "loss": 0.0255, "step": 35190 }, { "grad_norm": 0.2619532346725464, "learning_rate": 7.707135605726311e-05, "loss": 0.0259, "step": 35200 }, { "grad_norm": 0.26604577898979187, "learning_rate": 7.705745306803466e-05, "loss": 0.0313, "step": 35210 }, { "grad_norm": 0.7823427319526672, "learning_rate": 7.704354711984383e-05, "loss": 0.0297, "step": 35220 }, { "grad_norm": 0.2946148216724396, "learning_rate": 7.702963821421133e-05, "loss": 0.026, "step": 35230 }, { "grad_norm": 0.2237902134656906, "learning_rate": 7.701572635265826e-05, "loss": 0.0279, "step": 35240 }, { "grad_norm": 0.22742357850074768, "learning_rate": 7.700181153670596e-05, "loss": 0.0239, "step": 35250 }, { "grad_norm": 0.24917742609977722, "learning_rate": 7.698789376787616e-05, "loss": 0.029, "step": 35260 }, { "grad_norm": 0.2743645906448364, "learning_rate": 7.697397304769085e-05, "loss": 0.0306, "step": 35270 }, { "grad_norm": 0.2541390061378479, "learning_rate": 7.696004937767241e-05, "loss": 0.0273, "step": 35280 }, { "grad_norm": 0.20532874763011932, "learning_rate": 7.694612275934352e-05, "loss": 0.0255, "step": 35290 }, { "grad_norm": 0.28574836254119873, "learning_rate": 7.693219319422714e-05, "loss": 0.0241, "step": 35300 }, { "grad_norm": 0.26204296946525574, "learning_rate": 7.69182606838466e-05, "loss": 0.0242, "step": 35310 }, { "grad_norm": 0.2440166175365448, "learning_rate": 7.690432522972558e-05, "loss": 0.027, "step": 35320 }, { "grad_norm": 0.27614825963974, "learning_rate": 7.689038683338796e-05, "loss": 0.0304, "step": 35330 }, { "grad_norm": 0.19445131719112396, "learning_rate": 7.687644549635808e-05, "loss": 0.0252, "step": 35340 }, { "grad_norm": 0.20209190249443054, "learning_rate": 7.686250122016053e-05, "loss": 0.0259, "step": 35350 }, { "grad_norm": 0.2793757915496826, "learning_rate": 7.684855400632022e-05, "loss": 0.0306, "step": 35360 }, { "grad_norm": 0.2125096321105957, "learning_rate": 7.683460385636243e-05, "loss": 0.0228, "step": 35370 }, { "grad_norm": 0.3370799124240875, "learning_rate": 7.68206507718127e-05, "loss": 0.0296, "step": 35380 }, { "grad_norm": 0.26691654324531555, "learning_rate": 7.680669475419693e-05, "loss": 0.0263, "step": 35390 }, { "grad_norm": 0.24430453777313232, "learning_rate": 7.679273580504132e-05, "loss": 0.0295, "step": 35400 }, { "grad_norm": 0.25921696424484253, "learning_rate": 7.67787739258724e-05, "loss": 0.0244, "step": 35410 }, { "grad_norm": 0.2139459103345871, "learning_rate": 7.676480911821705e-05, "loss": 0.0246, "step": 35420 }, { "grad_norm": 0.24735210835933685, "learning_rate": 7.675084138360238e-05, "loss": 0.0273, "step": 35430 }, { "grad_norm": 0.2260286659002304, "learning_rate": 7.673687072355592e-05, "loss": 0.0285, "step": 35440 }, { "grad_norm": 0.20464923977851868, "learning_rate": 7.67228971396055e-05, "loss": 0.0249, "step": 35450 }, { "grad_norm": 0.26475799083709717, "learning_rate": 7.670892063327922e-05, "loss": 0.0256, "step": 35460 }, { "grad_norm": 0.2486627697944641, "learning_rate": 7.669494120610552e-05, "loss": 0.0275, "step": 35470 }, { "grad_norm": 0.29285311698913574, "learning_rate": 7.66809588596132e-05, "loss": 0.0237, "step": 35480 }, { "grad_norm": 0.2365093231201172, "learning_rate": 7.666697359533132e-05, "loss": 0.0209, "step": 35490 }, { "grad_norm": 0.30742210149765015, "learning_rate": 7.665298541478932e-05, "loss": 0.0291, "step": 35500 }, { "grad_norm": 0.2460704743862152, "learning_rate": 7.66389943195169e-05, "loss": 0.03, "step": 35510 }, { "grad_norm": 0.2614096403121948, "learning_rate": 7.66250003110441e-05, "loss": 0.0357, "step": 35520 }, { "grad_norm": 0.22634965181350708, "learning_rate": 7.66110033909013e-05, "loss": 0.0278, "step": 35530 }, { "grad_norm": 0.2399284690618515, "learning_rate": 7.659700356061918e-05, "loss": 0.0252, "step": 35540 }, { "grad_norm": 0.28589513897895813, "learning_rate": 7.658300082172875e-05, "loss": 0.0269, "step": 35550 }, { "grad_norm": 0.3117844760417938, "learning_rate": 7.65689951757613e-05, "loss": 0.0278, "step": 35560 }, { "grad_norm": 0.26613548398017883, "learning_rate": 7.655498662424849e-05, "loss": 0.0241, "step": 35570 }, { "grad_norm": 0.24491055309772491, "learning_rate": 7.654097516872227e-05, "loss": 0.0289, "step": 35580 }, { "grad_norm": 0.2576686441898346, "learning_rate": 7.65269608107149e-05, "loss": 0.0235, "step": 35590 }, { "grad_norm": 0.46305009722709656, "learning_rate": 7.6512943551759e-05, "loss": 0.0293, "step": 35600 }, { "grad_norm": 0.23480427265167236, "learning_rate": 7.649892339338743e-05, "loss": 0.0297, "step": 35610 }, { "grad_norm": 0.2644587457180023, "learning_rate": 7.648490033713344e-05, "loss": 0.0271, "step": 35620 }, { "grad_norm": 0.2735554575920105, "learning_rate": 7.647087438453058e-05, "loss": 0.0241, "step": 35630 }, { "grad_norm": 0.33592796325683594, "learning_rate": 7.64568455371127e-05, "loss": 0.0255, "step": 35640 }, { "grad_norm": 0.22001796960830688, "learning_rate": 7.644281379641396e-05, "loss": 0.0253, "step": 35650 }, { "grad_norm": 0.28185349702835083, "learning_rate": 7.642877916396887e-05, "loss": 0.0253, "step": 35660 }, { "grad_norm": 0.2161489874124527, "learning_rate": 7.641474164131221e-05, "loss": 0.0257, "step": 35670 }, { "grad_norm": 0.20849323272705078, "learning_rate": 7.640070122997913e-05, "loss": 0.0237, "step": 35680 }, { "grad_norm": 0.2638643980026245, "learning_rate": 7.638665793150505e-05, "loss": 0.0265, "step": 35690 }, { "grad_norm": 0.23726318776607513, "learning_rate": 7.637261174742574e-05, "loss": 0.0228, "step": 35700 }, { "grad_norm": 0.26295655965805054, "learning_rate": 7.635856267927725e-05, "loss": 0.0301, "step": 35710 }, { "grad_norm": 0.20864300429821014, "learning_rate": 7.634451072859597e-05, "loss": 0.026, "step": 35720 }, { "grad_norm": 0.2577241063117981, "learning_rate": 7.633045589691863e-05, "loss": 0.0282, "step": 35730 }, { "grad_norm": 0.30600032210350037, "learning_rate": 7.63163981857822e-05, "loss": 0.0295, "step": 35740 }, { "grad_norm": 0.3683592677116394, "learning_rate": 7.630233759672403e-05, "loss": 0.0263, "step": 35750 }, { "grad_norm": 0.24568386375904083, "learning_rate": 7.628827413128177e-05, "loss": 0.025, "step": 35760 }, { "grad_norm": 0.2374318391084671, "learning_rate": 7.627420779099338e-05, "loss": 0.0266, "step": 35770 }, { "grad_norm": 0.2748781442642212, "learning_rate": 7.626013857739711e-05, "loss": 0.0299, "step": 35780 }, { "grad_norm": 0.3229725658893585, "learning_rate": 7.624606649203158e-05, "loss": 0.0269, "step": 35790 }, { "grad_norm": 0.2933517396450043, "learning_rate": 7.623199153643569e-05, "loss": 0.0282, "step": 35800 }, { "grad_norm": 0.21779964864253998, "learning_rate": 7.621791371214863e-05, "loss": 0.0248, "step": 35810 }, { "grad_norm": 0.24468831717967987, "learning_rate": 7.620383302070995e-05, "loss": 0.0279, "step": 35820 }, { "grad_norm": 0.1970440149307251, "learning_rate": 7.61897494636595e-05, "loss": 0.0261, "step": 35830 }, { "grad_norm": 0.24574771523475647, "learning_rate": 7.617566304253739e-05, "loss": 0.0319, "step": 35840 }, { "grad_norm": 0.22902366518974304, "learning_rate": 7.616157375888416e-05, "loss": 0.0303, "step": 35850 }, { "grad_norm": 0.22223909199237823, "learning_rate": 7.614748161424053e-05, "loss": 0.0311, "step": 35860 }, { "grad_norm": 0.27080538868904114, "learning_rate": 7.613338661014763e-05, "loss": 0.0358, "step": 35870 }, { "grad_norm": 0.2056916505098343, "learning_rate": 7.611928874814686e-05, "loss": 0.0301, "step": 35880 }, { "grad_norm": 0.2551502287387848, "learning_rate": 7.610518802977996e-05, "loss": 0.0289, "step": 35890 }, { "grad_norm": 0.19925329089164734, "learning_rate": 7.609108445658893e-05, "loss": 0.0248, "step": 35900 }, { "grad_norm": 0.18789373338222504, "learning_rate": 7.607697803011612e-05, "loss": 0.0269, "step": 35910 }, { "grad_norm": 0.18451087176799774, "learning_rate": 7.606286875190421e-05, "loss": 0.0312, "step": 35920 }, { "grad_norm": 0.19569522142410278, "learning_rate": 7.604875662349617e-05, "loss": 0.03, "step": 35930 }, { "grad_norm": 0.19873091578483582, "learning_rate": 7.603464164643525e-05, "loss": 0.0271, "step": 35940 }, { "grad_norm": 0.28019028902053833, "learning_rate": 7.602052382226507e-05, "loss": 0.0268, "step": 35950 }, { "grad_norm": 0.36328789591789246, "learning_rate": 7.600640315252954e-05, "loss": 0.0222, "step": 35960 }, { "grad_norm": 0.29142701625823975, "learning_rate": 7.599227963877284e-05, "loss": 0.0265, "step": 35970 }, { "grad_norm": 0.2647651731967926, "learning_rate": 7.597815328253954e-05, "loss": 0.0285, "step": 35980 }, { "grad_norm": 0.2349255084991455, "learning_rate": 7.596402408537443e-05, "loss": 0.0305, "step": 35990 }, { "grad_norm": 0.2271953523159027, "learning_rate": 7.59498920488227e-05, "loss": 0.0267, "step": 36000 }, { "grad_norm": 0.2759001851081848, "learning_rate": 7.593575717442979e-05, "loss": 0.029, "step": 36010 }, { "grad_norm": 0.18702995777130127, "learning_rate": 7.592161946374147e-05, "loss": 0.025, "step": 36020 }, { "grad_norm": 0.22714506089687347, "learning_rate": 7.590747891830381e-05, "loss": 0.0259, "step": 36030 }, { "grad_norm": 0.1786549836397171, "learning_rate": 7.58933355396632e-05, "loss": 0.0272, "step": 36040 }, { "grad_norm": 0.24897035956382751, "learning_rate": 7.587918932936636e-05, "loss": 0.0249, "step": 36050 }, { "grad_norm": 0.18632078170776367, "learning_rate": 7.586504028896028e-05, "loss": 0.0283, "step": 36060 }, { "grad_norm": 0.23624186217784882, "learning_rate": 7.585088841999228e-05, "loss": 0.0252, "step": 36070 }, { "grad_norm": 0.27718979120254517, "learning_rate": 7.583673372400999e-05, "loss": 0.0265, "step": 36080 }, { "grad_norm": 0.26006343960762024, "learning_rate": 7.582257620256134e-05, "loss": 0.0265, "step": 36090 }, { "grad_norm": 0.1906326860189438, "learning_rate": 7.580841585719458e-05, "loss": 0.029, "step": 36100 }, { "grad_norm": 0.2595512568950653, "learning_rate": 7.579425268945825e-05, "loss": 0.0328, "step": 36110 }, { "grad_norm": 0.21527862548828125, "learning_rate": 7.578008670090127e-05, "loss": 0.0296, "step": 36120 }, { "grad_norm": 0.24158427119255066, "learning_rate": 7.576591789307272e-05, "loss": 0.0313, "step": 36130 }, { "grad_norm": 0.21147312223911285, "learning_rate": 7.575174626752216e-05, "loss": 0.0274, "step": 36140 }, { "grad_norm": 0.2063945084810257, "learning_rate": 7.573757182579934e-05, "loss": 0.0262, "step": 36150 }, { "grad_norm": 0.19520312547683716, "learning_rate": 7.572339456945435e-05, "loss": 0.0278, "step": 36160 }, { "grad_norm": 0.22759754955768585, "learning_rate": 7.570921450003762e-05, "loss": 0.0267, "step": 36170 }, { "grad_norm": 0.2558550536632538, "learning_rate": 7.569503161909983e-05, "loss": 0.0258, "step": 36180 }, { "grad_norm": 0.2373010814189911, "learning_rate": 7.568084592819202e-05, "loss": 0.0274, "step": 36190 }, { "grad_norm": 0.19977465271949768, "learning_rate": 7.566665742886551e-05, "loss": 0.0236, "step": 36200 }, { "grad_norm": 0.27130380272865295, "learning_rate": 7.565246612267194e-05, "loss": 0.0284, "step": 36210 }, { "grad_norm": 0.21711066365242004, "learning_rate": 7.563827201116325e-05, "loss": 0.0255, "step": 36220 }, { "grad_norm": 0.2715403139591217, "learning_rate": 7.562407509589164e-05, "loss": 0.028, "step": 36230 }, { "grad_norm": 0.25027140974998474, "learning_rate": 7.560987537840973e-05, "loss": 0.0299, "step": 36240 }, { "grad_norm": 0.26110145449638367, "learning_rate": 7.559567286027036e-05, "loss": 0.0263, "step": 36250 }, { "grad_norm": 0.26580825448036194, "learning_rate": 7.558146754302668e-05, "loss": 0.0271, "step": 36260 }, { "grad_norm": 0.2837343215942383, "learning_rate": 7.556725942823217e-05, "loss": 0.0255, "step": 36270 }, { "grad_norm": 0.2429528385400772, "learning_rate": 7.555304851744061e-05, "loss": 0.0245, "step": 36280 }, { "grad_norm": 0.27574101090431213, "learning_rate": 7.553883481220608e-05, "loss": 0.0225, "step": 36290 }, { "grad_norm": 0.23100171983242035, "learning_rate": 7.552461831408298e-05, "loss": 0.0239, "step": 36300 }, { "grad_norm": 0.2604421377182007, "learning_rate": 7.551039902462599e-05, "loss": 0.03, "step": 36310 }, { "grad_norm": 0.22778598964214325, "learning_rate": 7.549617694539011e-05, "loss": 0.0261, "step": 36320 }, { "grad_norm": 0.18372471630573273, "learning_rate": 7.548195207793066e-05, "loss": 0.0235, "step": 36330 }, { "grad_norm": 0.28153643012046814, "learning_rate": 7.546772442380323e-05, "loss": 0.0251, "step": 36340 }, { "grad_norm": 0.21212202310562134, "learning_rate": 7.545349398456376e-05, "loss": 0.0288, "step": 36350 }, { "grad_norm": 0.24929462373256683, "learning_rate": 7.543926076176845e-05, "loss": 0.0244, "step": 36360 }, { "grad_norm": 0.3193281292915344, "learning_rate": 7.542502475697385e-05, "loss": 0.0256, "step": 36370 }, { "grad_norm": 0.24831043183803558, "learning_rate": 7.541078597173675e-05, "loss": 0.028, "step": 36380 }, { "grad_norm": 0.23732809722423553, "learning_rate": 7.539654440761431e-05, "loss": 0.0239, "step": 36390 }, { "grad_norm": 0.26527994871139526, "learning_rate": 7.538230006616395e-05, "loss": 0.0282, "step": 36400 }, { "grad_norm": 0.24172553420066833, "learning_rate": 7.536805294894342e-05, "loss": 0.0227, "step": 36410 }, { "grad_norm": 0.19483280181884766, "learning_rate": 7.535380305751076e-05, "loss": 0.0252, "step": 36420 }, { "grad_norm": 0.22749292850494385, "learning_rate": 7.533955039342431e-05, "loss": 0.0274, "step": 36430 }, { "grad_norm": 0.18460658192634583, "learning_rate": 7.532529495824274e-05, "loss": 0.0244, "step": 36440 }, { "grad_norm": 0.19585683941841125, "learning_rate": 7.531103675352497e-05, "loss": 0.0261, "step": 36450 }, { "grad_norm": 0.20354020595550537, "learning_rate": 7.529677578083029e-05, "loss": 0.0288, "step": 36460 }, { "grad_norm": 0.31015148758888245, "learning_rate": 7.528251204171823e-05, "loss": 0.0292, "step": 36470 }, { "grad_norm": 0.19948165118694305, "learning_rate": 7.526824553774867e-05, "loss": 0.0286, "step": 36480 }, { "grad_norm": 0.3256489336490631, "learning_rate": 7.525397627048178e-05, "loss": 0.0322, "step": 36490 }, { "grad_norm": 0.1893988698720932, "learning_rate": 7.523970424147802e-05, "loss": 0.0233, "step": 36500 }, { "grad_norm": 0.3330492079257965, "learning_rate": 7.522542945229813e-05, "loss": 0.0269, "step": 36510 }, { "grad_norm": 0.2325344830751419, "learning_rate": 7.521115190450321e-05, "loss": 0.0252, "step": 36520 }, { "grad_norm": 0.4013141989707947, "learning_rate": 7.519687159965462e-05, "loss": 0.0304, "step": 36530 }, { "grad_norm": 0.22612065076828003, "learning_rate": 7.518258853931403e-05, "loss": 0.0258, "step": 36540 }, { "grad_norm": 0.18818333745002747, "learning_rate": 7.516830272504342e-05, "loss": 0.0234, "step": 36550 }, { "grad_norm": 0.24223293364048004, "learning_rate": 7.515401415840509e-05, "loss": 0.0296, "step": 36560 }, { "grad_norm": 0.18215686082839966, "learning_rate": 7.513972284096155e-05, "loss": 0.0244, "step": 36570 }, { "grad_norm": 0.15301333367824554, "learning_rate": 7.512542877427574e-05, "loss": 0.0229, "step": 36580 }, { "grad_norm": 0.2568630576133728, "learning_rate": 7.51111319599108e-05, "loss": 0.0249, "step": 36590 }, { "grad_norm": 0.3492150604724884, "learning_rate": 7.509683239943024e-05, "loss": 0.022, "step": 36600 }, { "grad_norm": 0.3238261342048645, "learning_rate": 7.50825300943978e-05, "loss": 0.0299, "step": 36610 }, { "grad_norm": 0.24066823720932007, "learning_rate": 7.50682250463776e-05, "loss": 0.0274, "step": 36620 }, { "grad_norm": 0.20768795907497406, "learning_rate": 7.5053917256934e-05, "loss": 0.0238, "step": 36630 }, { "grad_norm": 0.23065219819545746, "learning_rate": 7.503960672763166e-05, "loss": 0.0229, "step": 36640 }, { "grad_norm": 0.2496524602174759, "learning_rate": 7.502529346003559e-05, "loss": 0.0278, "step": 36650 }, { "grad_norm": 0.25873029232025146, "learning_rate": 7.501097745571106e-05, "loss": 0.0335, "step": 36660 }, { "grad_norm": 0.2514204978942871, "learning_rate": 7.499665871622361e-05, "loss": 0.024, "step": 36670 }, { "grad_norm": 0.30953043699264526, "learning_rate": 7.498233724313917e-05, "loss": 0.0263, "step": 36680 }, { "grad_norm": 0.20244325697422028, "learning_rate": 7.496801303802387e-05, "loss": 0.0295, "step": 36690 }, { "grad_norm": 0.24887122213840485, "learning_rate": 7.495368610244423e-05, "loss": 0.0237, "step": 36700 }, { "grad_norm": 0.2476940155029297, "learning_rate": 7.493935643796697e-05, "loss": 0.0225, "step": 36710 }, { "grad_norm": 0.19042685627937317, "learning_rate": 7.49250240461592e-05, "loss": 0.0222, "step": 36720 }, { "grad_norm": 0.23919770121574402, "learning_rate": 7.491068892858826e-05, "loss": 0.0227, "step": 36730 }, { "grad_norm": 0.24321693181991577, "learning_rate": 7.489635108682184e-05, "loss": 0.0244, "step": 36740 }, { "grad_norm": 0.20931130647659302, "learning_rate": 7.48820105224279e-05, "loss": 0.0254, "step": 36750 }, { "grad_norm": 0.22739355266094208, "learning_rate": 7.48676672369747e-05, "loss": 0.0247, "step": 36760 }, { "grad_norm": 0.20033790171146393, "learning_rate": 7.485332123203079e-05, "loss": 0.0197, "step": 36770 }, { "grad_norm": 0.1778850257396698, "learning_rate": 7.483897250916504e-05, "loss": 0.0246, "step": 36780 }, { "grad_norm": 0.1971290111541748, "learning_rate": 7.48246210699466e-05, "loss": 0.0248, "step": 36790 }, { "grad_norm": 0.22549082338809967, "learning_rate": 7.481026691594492e-05, "loss": 0.0265, "step": 36800 }, { "grad_norm": 0.22821645438671112, "learning_rate": 7.479591004872975e-05, "loss": 0.0272, "step": 36810 }, { "grad_norm": 0.2600482404232025, "learning_rate": 7.478155046987114e-05, "loss": 0.026, "step": 36820 }, { "grad_norm": 0.19890539348125458, "learning_rate": 7.476718818093944e-05, "loss": 0.0231, "step": 36830 }, { "grad_norm": 0.2420700490474701, "learning_rate": 7.475282318350527e-05, "loss": 0.0267, "step": 36840 }, { "grad_norm": 0.2427087277173996, "learning_rate": 7.473845547913957e-05, "loss": 0.0242, "step": 36850 }, { "grad_norm": 0.2534519135951996, "learning_rate": 7.472408506941357e-05, "loss": 0.0266, "step": 36860 }, { "grad_norm": 0.24675287306308746, "learning_rate": 7.470971195589881e-05, "loss": 0.028, "step": 36870 }, { "grad_norm": 0.2018173336982727, "learning_rate": 7.46953361401671e-05, "loss": 0.0236, "step": 36880 }, { "grad_norm": 0.21637572348117828, "learning_rate": 7.468095762379055e-05, "loss": 0.03, "step": 36890 }, { "grad_norm": 0.22519031167030334, "learning_rate": 7.466657640834158e-05, "loss": 0.0264, "step": 36900 }, { "grad_norm": 0.20389346778392792, "learning_rate": 7.46521924953929e-05, "loss": 0.0232, "step": 36910 }, { "grad_norm": 0.20818045735359192, "learning_rate": 7.463780588651752e-05, "loss": 0.021, "step": 36920 }, { "grad_norm": 0.4023255407810211, "learning_rate": 7.462341658328873e-05, "loss": 0.0255, "step": 36930 }, { "grad_norm": 0.20918631553649902, "learning_rate": 7.460902458728012e-05, "loss": 0.0234, "step": 36940 }, { "grad_norm": 0.205887109041214, "learning_rate": 7.459462990006558e-05, "loss": 0.0233, "step": 36950 }, { "grad_norm": 0.21102888882160187, "learning_rate": 7.458023252321929e-05, "loss": 0.0256, "step": 36960 }, { "grad_norm": 0.2524605095386505, "learning_rate": 7.456583245831573e-05, "loss": 0.0258, "step": 36970 }, { "grad_norm": 0.24666118621826172, "learning_rate": 7.455142970692968e-05, "loss": 0.0249, "step": 36980 }, { "grad_norm": 0.29588428139686584, "learning_rate": 7.453702427063615e-05, "loss": 0.0253, "step": 36990 }, { "grad_norm": 0.25850972533226013, "learning_rate": 7.452261615101057e-05, "loss": 0.0245, "step": 37000 }, { "grad_norm": 0.21698284149169922, "learning_rate": 7.450820534962856e-05, "loss": 0.0248, "step": 37010 }, { "grad_norm": 0.2450183928012848, "learning_rate": 7.449379186806604e-05, "loss": 0.0274, "step": 37020 }, { "grad_norm": 0.23351554572582245, "learning_rate": 7.44793757078993e-05, "loss": 0.0262, "step": 37030 }, { "grad_norm": 0.2474880814552307, "learning_rate": 7.44649568707048e-05, "loss": 0.0264, "step": 37040 }, { "grad_norm": 0.3158344626426697, "learning_rate": 7.445053535805942e-05, "loss": 0.0246, "step": 37050 }, { "grad_norm": 0.2703455090522766, "learning_rate": 7.443611117154024e-05, "loss": 0.0237, "step": 37060 }, { "grad_norm": 0.2505357563495636, "learning_rate": 7.442168431272469e-05, "loss": 0.0226, "step": 37070 }, { "grad_norm": 0.22385993599891663, "learning_rate": 7.440725478319044e-05, "loss": 0.0259, "step": 37080 }, { "grad_norm": 0.2407579869031906, "learning_rate": 7.439282258451553e-05, "loss": 0.0245, "step": 37090 }, { "grad_norm": 0.23338310420513153, "learning_rate": 7.43783877182782e-05, "loss": 0.0254, "step": 37100 }, { "grad_norm": 0.24678587913513184, "learning_rate": 7.436395018605707e-05, "loss": 0.0246, "step": 37110 }, { "grad_norm": 0.2516440451145172, "learning_rate": 7.434950998943094e-05, "loss": 0.024, "step": 37120 }, { "grad_norm": 0.15390080213546753, "learning_rate": 7.433506712997902e-05, "loss": 0.0207, "step": 37130 }, { "grad_norm": 0.22355736792087555, "learning_rate": 7.432062160928077e-05, "loss": 0.0253, "step": 37140 }, { "grad_norm": 0.2715085744857788, "learning_rate": 7.430617342891588e-05, "loss": 0.0241, "step": 37150 }, { "grad_norm": 0.21833044290542603, "learning_rate": 7.429172259046442e-05, "loss": 0.0246, "step": 37160 }, { "grad_norm": 0.21433471143245697, "learning_rate": 7.42772690955067e-05, "loss": 0.0238, "step": 37170 }, { "grad_norm": 0.21315345168113708, "learning_rate": 7.426281294562332e-05, "loss": 0.0277, "step": 37180 }, { "grad_norm": 0.23269562423229218, "learning_rate": 7.424835414239519e-05, "loss": 0.0253, "step": 37190 }, { "grad_norm": 0.17327004671096802, "learning_rate": 7.423389268740352e-05, "loss": 0.0241, "step": 37200 }, { "grad_norm": 0.2924984097480774, "learning_rate": 7.42194285822298e-05, "loss": 0.0285, "step": 37210 }, { "grad_norm": 0.21414412558078766, "learning_rate": 7.420496182845576e-05, "loss": 0.0243, "step": 37220 }, { "grad_norm": 0.22679102420806885, "learning_rate": 7.419049242766348e-05, "loss": 0.0294, "step": 37230 }, { "grad_norm": 0.29796475172042847, "learning_rate": 7.417602038143533e-05, "loss": 0.028, "step": 37240 }, { "grad_norm": 0.21682517230510712, "learning_rate": 7.416154569135393e-05, "loss": 0.0267, "step": 37250 }, { "grad_norm": 0.23468968272209167, "learning_rate": 7.414706835900222e-05, "loss": 0.0274, "step": 37260 }, { "grad_norm": 0.15233822166919708, "learning_rate": 7.413258838596345e-05, "loss": 0.0246, "step": 37270 }, { "grad_norm": 0.23144347965717316, "learning_rate": 7.411810577382106e-05, "loss": 0.0254, "step": 37280 }, { "grad_norm": 0.23104816675186157, "learning_rate": 7.410362052415889e-05, "loss": 0.0237, "step": 37290 }, { "grad_norm": 0.19235004484653473, "learning_rate": 7.408913263856102e-05, "loss": 0.0223, "step": 37300 }, { "grad_norm": 0.32165616750717163, "learning_rate": 7.407464211861183e-05, "loss": 0.0263, "step": 37310 }, { "grad_norm": 0.2162395864725113, "learning_rate": 7.406014896589597e-05, "loss": 0.0227, "step": 37320 }, { "grad_norm": 0.22505615651607513, "learning_rate": 7.404565318199838e-05, "loss": 0.0257, "step": 37330 }, { "grad_norm": 0.2268468141555786, "learning_rate": 7.403115476850432e-05, "loss": 0.0229, "step": 37340 }, { "grad_norm": 0.22163087129592896, "learning_rate": 7.401665372699932e-05, "loss": 0.0269, "step": 37350 }, { "grad_norm": 0.25425857305526733, "learning_rate": 7.400215005906916e-05, "loss": 0.0278, "step": 37360 }, { "grad_norm": 0.1928853988647461, "learning_rate": 7.398764376629996e-05, "loss": 0.0282, "step": 37370 }, { "grad_norm": 0.3387829065322876, "learning_rate": 7.39731348502781e-05, "loss": 0.0262, "step": 37380 }, { "grad_norm": 0.19137878715991974, "learning_rate": 7.395862331259028e-05, "loss": 0.0307, "step": 37390 }, { "grad_norm": 0.21814003586769104, "learning_rate": 7.394410915482341e-05, "loss": 0.0245, "step": 37400 }, { "grad_norm": 0.19806243479251862, "learning_rate": 7.392959237856476e-05, "loss": 0.0277, "step": 37410 }, { "grad_norm": 0.23483100533485413, "learning_rate": 7.391507298540187e-05, "loss": 0.0233, "step": 37420 }, { "grad_norm": 0.22632671892642975, "learning_rate": 7.390055097692257e-05, "loss": 0.026, "step": 37430 }, { "grad_norm": 0.14736407995224, "learning_rate": 7.388602635471494e-05, "loss": 0.0206, "step": 37440 }, { "grad_norm": 0.196986123919487, "learning_rate": 7.387149912036738e-05, "loss": 0.0219, "step": 37450 }, { "grad_norm": 0.27806347608566284, "learning_rate": 7.385696927546854e-05, "loss": 0.03, "step": 37460 }, { "grad_norm": 0.26716089248657227, "learning_rate": 7.384243682160743e-05, "loss": 0.0266, "step": 37470 }, { "grad_norm": 0.18865235149860382, "learning_rate": 7.382790176037329e-05, "loss": 0.0235, "step": 37480 }, { "grad_norm": 0.25734278559684753, "learning_rate": 7.381336409335562e-05, "loss": 0.0274, "step": 37490 }, { "grad_norm": 0.1672191619873047, "learning_rate": 7.379882382214426e-05, "loss": 0.0235, "step": 37500 }, { "grad_norm": 0.21018628776073456, "learning_rate": 7.378428094832931e-05, "loss": 0.021, "step": 37510 }, { "grad_norm": 0.24096864461898804, "learning_rate": 7.376973547350114e-05, "loss": 0.0252, "step": 37520 }, { "grad_norm": 0.2772797644138336, "learning_rate": 7.375518739925046e-05, "loss": 0.0236, "step": 37530 }, { "grad_norm": 0.23472930490970612, "learning_rate": 7.374063672716818e-05, "loss": 0.0205, "step": 37540 }, { "grad_norm": 0.2172943502664566, "learning_rate": 7.372608345884558e-05, "loss": 0.023, "step": 37550 }, { "grad_norm": 0.22856274247169495, "learning_rate": 7.371152759587415e-05, "loss": 0.0231, "step": 37560 }, { "grad_norm": 0.21718792617321014, "learning_rate": 7.369696913984572e-05, "loss": 0.0233, "step": 37570 }, { "grad_norm": 0.24428468942642212, "learning_rate": 7.368240809235237e-05, "loss": 0.0255, "step": 37580 }, { "grad_norm": 0.2693498432636261, "learning_rate": 7.366784445498649e-05, "loss": 0.0248, "step": 37590 }, { "grad_norm": 0.3310425579547882, "learning_rate": 7.36532782293407e-05, "loss": 0.0304, "step": 37600 }, { "grad_norm": 0.30834776163101196, "learning_rate": 7.363870941700797e-05, "loss": 0.0237, "step": 37610 }, { "grad_norm": 0.1984536051750183, "learning_rate": 7.362413801958152e-05, "loss": 0.0219, "step": 37620 }, { "grad_norm": 0.39566653966903687, "learning_rate": 7.360956403865486e-05, "loss": 0.032, "step": 37630 }, { "grad_norm": 0.2027982622385025, "learning_rate": 7.359498747582177e-05, "loss": 0.0292, "step": 37640 }, { "grad_norm": 0.28311166167259216, "learning_rate": 7.358040833267634e-05, "loss": 0.0256, "step": 37650 }, { "grad_norm": 0.305388867855072, "learning_rate": 7.356582661081288e-05, "loss": 0.0252, "step": 37660 }, { "grad_norm": 0.23515884578227997, "learning_rate": 7.355124231182607e-05, "loss": 0.0238, "step": 37670 }, { "grad_norm": 0.23787598311901093, "learning_rate": 7.353665543731083e-05, "loss": 0.0254, "step": 37680 }, { "grad_norm": 0.2099483460187912, "learning_rate": 7.352206598886229e-05, "loss": 0.0244, "step": 37690 }, { "grad_norm": 0.15834444761276245, "learning_rate": 7.350747396807601e-05, "loss": 0.0239, "step": 37700 }, { "grad_norm": 0.20094050467014313, "learning_rate": 7.349287937654772e-05, "loss": 0.0216, "step": 37710 }, { "grad_norm": 0.21435865759849548, "learning_rate": 7.347828221587345e-05, "loss": 0.0191, "step": 37720 }, { "grad_norm": 0.24482418596744537, "learning_rate": 7.346368248764955e-05, "loss": 0.0219, "step": 37730 }, { "grad_norm": 0.22152327001094818, "learning_rate": 7.34490801934726e-05, "loss": 0.0209, "step": 37740 }, { "grad_norm": 0.24728550016880035, "learning_rate": 7.343447533493947e-05, "loss": 0.0255, "step": 37750 }, { "grad_norm": 0.2835630476474762, "learning_rate": 7.341986791364739e-05, "loss": 0.0243, "step": 37760 }, { "grad_norm": 0.18982531130313873, "learning_rate": 7.340525793119374e-05, "loss": 0.0292, "step": 37770 }, { "grad_norm": 0.1861477643251419, "learning_rate": 7.33906453891763e-05, "loss": 0.0229, "step": 37780 }, { "grad_norm": 0.26229217648506165, "learning_rate": 7.337603028919303e-05, "loss": 0.0212, "step": 37790 }, { "grad_norm": 0.1802469789981842, "learning_rate": 7.336141263284225e-05, "loss": 0.0254, "step": 37800 }, { "grad_norm": 0.22733674943447113, "learning_rate": 7.334679242172249e-05, "loss": 0.0239, "step": 37810 }, { "grad_norm": 0.16503667831420898, "learning_rate": 7.33321696574326e-05, "loss": 0.0254, "step": 37820 }, { "grad_norm": 0.18014056980609894, "learning_rate": 7.331754434157174e-05, "loss": 0.0237, "step": 37830 }, { "grad_norm": 0.2505115270614624, "learning_rate": 7.330291647573928e-05, "loss": 0.0268, "step": 37840 }, { "grad_norm": 0.3853566348552704, "learning_rate": 7.32882860615349e-05, "loss": 0.0238, "step": 37850 }, { "grad_norm": 0.26230576634407043, "learning_rate": 7.327365310055858e-05, "loss": 0.0282, "step": 37860 }, { "grad_norm": 0.2208412140607834, "learning_rate": 7.325901759441055e-05, "loss": 0.0237, "step": 37870 }, { "grad_norm": 0.2051178365945816, "learning_rate": 7.324437954469132e-05, "loss": 0.0252, "step": 37880 }, { "grad_norm": 0.2227579653263092, "learning_rate": 7.32297389530017e-05, "loss": 0.0267, "step": 37890 }, { "grad_norm": 0.24331559240818024, "learning_rate": 7.321509582094274e-05, "loss": 0.0236, "step": 37900 }, { "grad_norm": 0.23446395993232727, "learning_rate": 7.320045015011582e-05, "loss": 0.0248, "step": 37910 }, { "grad_norm": 0.21765992045402527, "learning_rate": 7.318580194212253e-05, "loss": 0.0284, "step": 37920 }, { "grad_norm": 0.2249288409948349, "learning_rate": 7.317115119856483e-05, "loss": 0.0217, "step": 37930 }, { "grad_norm": 0.22330039739608765, "learning_rate": 7.315649792104485e-05, "loss": 0.0276, "step": 37940 }, { "grad_norm": 0.24721693992614746, "learning_rate": 7.314184211116507e-05, "loss": 0.0282, "step": 37950 }, { "grad_norm": 0.23727034032344818, "learning_rate": 7.312718377052823e-05, "loss": 0.0231, "step": 37960 }, { "grad_norm": 0.2597658932209015, "learning_rate": 7.311252290073737e-05, "loss": 0.0269, "step": 37970 }, { "grad_norm": 0.1714426428079605, "learning_rate": 7.309785950339572e-05, "loss": 0.0258, "step": 37980 }, { "grad_norm": 0.23931293189525604, "learning_rate": 7.308319358010692e-05, "loss": 0.0278, "step": 37990 }, { "grad_norm": 0.2702704668045044, "learning_rate": 7.306852513247474e-05, "loss": 0.0246, "step": 38000 }, { "grad_norm": 0.23878894746303558, "learning_rate": 7.305385416210336e-05, "loss": 0.0231, "step": 38010 }, { "grad_norm": 0.21319271624088287, "learning_rate": 7.303918067059711e-05, "loss": 0.0212, "step": 38020 }, { "grad_norm": 0.2734033167362213, "learning_rate": 7.302450465956075e-05, "loss": 0.0249, "step": 38030 }, { "grad_norm": 0.22262275218963623, "learning_rate": 7.300982613059914e-05, "loss": 0.0238, "step": 38040 }, { "grad_norm": 0.27282479405403137, "learning_rate": 7.299514508531757e-05, "loss": 0.0258, "step": 38050 }, { "grad_norm": 0.25704705715179443, "learning_rate": 7.298046152532148e-05, "loss": 0.0246, "step": 38060 }, { "grad_norm": 0.21366344392299652, "learning_rate": 7.296577545221668e-05, "loss": 0.0235, "step": 38070 }, { "grad_norm": 0.21074128150939941, "learning_rate": 7.29510868676092e-05, "loss": 0.0207, "step": 38080 }, { "grad_norm": 0.2067423313856125, "learning_rate": 7.293639577310538e-05, "loss": 0.0238, "step": 38090 }, { "grad_norm": 0.24994660913944244, "learning_rate": 7.292170217031179e-05, "loss": 0.0268, "step": 38100 }, { "grad_norm": 0.2286454141139984, "learning_rate": 7.290700606083532e-05, "loss": 0.0227, "step": 38110 }, { "grad_norm": 0.29406899213790894, "learning_rate": 7.28923074462831e-05, "loss": 0.0229, "step": 38120 }, { "grad_norm": 0.2500542998313904, "learning_rate": 7.287760632826255e-05, "loss": 0.0232, "step": 38130 }, { "grad_norm": 0.25835758447647095, "learning_rate": 7.286290270838139e-05, "loss": 0.0251, "step": 38140 }, { "grad_norm": 0.23301492631435394, "learning_rate": 7.284819658824756e-05, "loss": 0.029, "step": 38150 }, { "grad_norm": 0.22790946066379547, "learning_rate": 7.283348796946929e-05, "loss": 0.0242, "step": 38160 }, { "grad_norm": 0.22203759849071503, "learning_rate": 7.28187768536551e-05, "loss": 0.0271, "step": 38170 }, { "grad_norm": 0.22202880680561066, "learning_rate": 7.280406324241379e-05, "loss": 0.0236, "step": 38180 }, { "grad_norm": 0.33736422657966614, "learning_rate": 7.278934713735441e-05, "loss": 0.0258, "step": 38190 }, { "grad_norm": 0.2869080603122711, "learning_rate": 7.277462854008629e-05, "loss": 0.0228, "step": 38200 }, { "grad_norm": 0.2208193689584732, "learning_rate": 7.275990745221904e-05, "loss": 0.0206, "step": 38210 }, { "grad_norm": 0.2645217478275299, "learning_rate": 7.274518387536252e-05, "loss": 0.0273, "step": 38220 }, { "grad_norm": 0.36562225222587585, "learning_rate": 7.273045781112689e-05, "loss": 0.027, "step": 38230 }, { "grad_norm": 0.2279854714870453, "learning_rate": 7.271572926112257e-05, "loss": 0.0254, "step": 38240 }, { "grad_norm": 0.25885558128356934, "learning_rate": 7.270099822696024e-05, "loss": 0.0238, "step": 38250 }, { "grad_norm": 0.2896067202091217, "learning_rate": 7.26862647102509e-05, "loss": 0.0265, "step": 38260 }, { "grad_norm": 0.2705649435520172, "learning_rate": 7.267152871260573e-05, "loss": 0.0206, "step": 38270 }, { "grad_norm": 0.18397629261016846, "learning_rate": 7.265679023563629e-05, "loss": 0.0251, "step": 38280 }, { "grad_norm": 0.2910598814487457, "learning_rate": 7.264204928095433e-05, "loss": 0.0251, "step": 38290 }, { "grad_norm": 0.23972618579864502, "learning_rate": 7.262730585017188e-05, "loss": 0.0214, "step": 38300 }, { "grad_norm": 0.19637525081634521, "learning_rate": 7.261255994490131e-05, "loss": 0.0233, "step": 38310 }, { "grad_norm": 0.21979574859142303, "learning_rate": 7.259781156675517e-05, "loss": 0.0213, "step": 38320 }, { "grad_norm": 0.23711568117141724, "learning_rate": 7.258306071734633e-05, "loss": 0.0236, "step": 38330 }, { "grad_norm": 0.19582214951515198, "learning_rate": 7.256830739828792e-05, "loss": 0.0234, "step": 38340 }, { "grad_norm": 0.2404182106256485, "learning_rate": 7.255355161119336e-05, "loss": 0.0281, "step": 38350 }, { "grad_norm": 0.20925426483154297, "learning_rate": 7.253879335767628e-05, "loss": 0.0233, "step": 38360 }, { "grad_norm": 0.15583421289920807, "learning_rate": 7.252403263935067e-05, "loss": 0.0241, "step": 38370 }, { "grad_norm": 0.17230220139026642, "learning_rate": 7.25092694578307e-05, "loss": 0.0238, "step": 38380 }, { "grad_norm": 0.32972002029418945, "learning_rate": 7.249450381473087e-05, "loss": 0.0278, "step": 38390 }, { "grad_norm": 0.2104901820421219, "learning_rate": 7.247973571166593e-05, "loss": 0.0241, "step": 38400 }, { "grad_norm": 0.2487608939409256, "learning_rate": 7.246496515025089e-05, "loss": 0.0244, "step": 38410 }, { "grad_norm": 0.3127370774745941, "learning_rate": 7.245019213210101e-05, "loss": 0.0243, "step": 38420 }, { "grad_norm": 0.35183286666870117, "learning_rate": 7.24354166588319e-05, "loss": 0.0254, "step": 38430 }, { "grad_norm": 0.2151792198419571, "learning_rate": 7.242063873205935e-05, "loss": 0.0234, "step": 38440 }, { "grad_norm": 0.21579043567180634, "learning_rate": 7.240585835339946e-05, "loss": 0.0247, "step": 38450 }, { "grad_norm": 0.17990364134311676, "learning_rate": 7.239107552446857e-05, "loss": 0.0246, "step": 38460 }, { "grad_norm": 0.2873038947582245, "learning_rate": 7.237629024688334e-05, "loss": 0.0227, "step": 38470 }, { "grad_norm": 0.21081796288490295, "learning_rate": 7.236150252226064e-05, "loss": 0.0286, "step": 38480 }, { "grad_norm": 0.1951388120651245, "learning_rate": 7.234671235221765e-05, "loss": 0.0262, "step": 38490 }, { "grad_norm": 0.19438469409942627, "learning_rate": 7.233191973837179e-05, "loss": 0.0266, "step": 38500 }, { "grad_norm": 0.22425784170627594, "learning_rate": 7.231712468234078e-05, "loss": 0.0223, "step": 38510 }, { "grad_norm": 0.18882150948047638, "learning_rate": 7.230232718574254e-05, "loss": 0.0237, "step": 38520 }, { "grad_norm": 0.2761220633983612, "learning_rate": 7.228752725019535e-05, "loss": 0.0277, "step": 38530 }, { "grad_norm": 0.27971404790878296, "learning_rate": 7.227272487731769e-05, "loss": 0.0207, "step": 38540 }, { "grad_norm": 0.22728469967842102, "learning_rate": 7.225792006872831e-05, "loss": 0.0257, "step": 38550 }, { "grad_norm": 0.19601400196552277, "learning_rate": 7.224311282604628e-05, "loss": 0.0233, "step": 38560 }, { "grad_norm": 0.23449239134788513, "learning_rate": 7.222830315089085e-05, "loss": 0.0279, "step": 38570 }, { "grad_norm": 0.18137836456298828, "learning_rate": 7.22134910448816e-05, "loss": 0.0211, "step": 38580 }, { "grad_norm": 0.19782006740570068, "learning_rate": 7.219867650963839e-05, "loss": 0.021, "step": 38590 }, { "grad_norm": 0.24867452681064606, "learning_rate": 7.21838595467813e-05, "loss": 0.0242, "step": 38600 }, { "grad_norm": 0.20896345376968384, "learning_rate": 7.216904015793067e-05, "loss": 0.025, "step": 38610 }, { "grad_norm": 0.24366477131843567, "learning_rate": 7.215421834470713e-05, "loss": 0.0253, "step": 38620 }, { "grad_norm": 0.19957026839256287, "learning_rate": 7.213939410873158e-05, "loss": 0.0241, "step": 38630 }, { "grad_norm": 0.23892700672149658, "learning_rate": 7.212456745162519e-05, "loss": 0.0229, "step": 38640 }, { "grad_norm": 0.20567134022712708, "learning_rate": 7.210973837500937e-05, "loss": 0.0258, "step": 38650 }, { "grad_norm": 0.29157236218452454, "learning_rate": 7.209490688050578e-05, "loss": 0.0231, "step": 38660 }, { "grad_norm": 0.2596408724784851, "learning_rate": 7.208007296973641e-05, "loss": 0.0248, "step": 38670 }, { "grad_norm": 0.16842667758464813, "learning_rate": 7.206523664432345e-05, "loss": 0.0206, "step": 38680 }, { "grad_norm": 0.3448498249053955, "learning_rate": 7.205039790588939e-05, "loss": 0.0239, "step": 38690 }, { "grad_norm": 0.1785394847393036, "learning_rate": 7.203555675605697e-05, "loss": 0.0213, "step": 38700 }, { "grad_norm": 0.23387722671031952, "learning_rate": 7.202071319644917e-05, "loss": 0.0246, "step": 38710 }, { "grad_norm": 0.23051109910011292, "learning_rate": 7.200586722868932e-05, "loss": 0.0269, "step": 38720 }, { "grad_norm": 0.30353397130966187, "learning_rate": 7.199101885440088e-05, "loss": 0.0235, "step": 38730 }, { "grad_norm": 0.22952711582183838, "learning_rate": 7.197616807520771e-05, "loss": 0.0255, "step": 38740 }, { "grad_norm": 0.2596551775932312, "learning_rate": 7.196131489273381e-05, "loss": 0.028, "step": 38750 }, { "grad_norm": 0.2500953674316406, "learning_rate": 7.194645930860355e-05, "loss": 0.0208, "step": 38760 }, { "grad_norm": 0.3480193614959717, "learning_rate": 7.19316013244415e-05, "loss": 0.0238, "step": 38770 }, { "grad_norm": 0.24812239408493042, "learning_rate": 7.191674094187248e-05, "loss": 0.0271, "step": 38780 }, { "grad_norm": 0.21229173243045807, "learning_rate": 7.190187816252165e-05, "loss": 0.0259, "step": 38790 }, { "grad_norm": 0.1669473946094513, "learning_rate": 7.188701298801435e-05, "loss": 0.0232, "step": 38800 }, { "grad_norm": 0.21999377012252808, "learning_rate": 7.18721454199762e-05, "loss": 0.0296, "step": 38810 }, { "grad_norm": 0.2101113647222519, "learning_rate": 7.185727546003312e-05, "loss": 0.0219, "step": 38820 }, { "grad_norm": 0.25276514887809753, "learning_rate": 7.184240310981126e-05, "loss": 0.027, "step": 38830 }, { "grad_norm": 0.20507453382015228, "learning_rate": 7.182752837093704e-05, "loss": 0.0211, "step": 38840 }, { "grad_norm": 0.18559837341308594, "learning_rate": 7.181265124503711e-05, "loss": 0.0253, "step": 38850 }, { "grad_norm": 0.24360403418540955, "learning_rate": 7.179777173373847e-05, "loss": 0.0214, "step": 38860 }, { "grad_norm": 0.242562934756279, "learning_rate": 7.178288983866826e-05, "loss": 0.0211, "step": 38870 }, { "grad_norm": 0.17373861372470856, "learning_rate": 7.176800556145397e-05, "loss": 0.0229, "step": 38880 }, { "grad_norm": 0.17732003331184387, "learning_rate": 7.175311890372334e-05, "loss": 0.0209, "step": 38890 }, { "grad_norm": 0.29219645261764526, "learning_rate": 7.17382298671043e-05, "loss": 0.0284, "step": 38900 }, { "grad_norm": 0.2721913456916809, "learning_rate": 7.172333845322515e-05, "loss": 0.0228, "step": 38910 }, { "grad_norm": 0.1967894732952118, "learning_rate": 7.170844466371436e-05, "loss": 0.0245, "step": 38920 }, { "grad_norm": 0.21577584743499756, "learning_rate": 7.16935485002007e-05, "loss": 0.0285, "step": 38930 }, { "grad_norm": 0.247441828250885, "learning_rate": 7.167864996431319e-05, "loss": 0.0244, "step": 38940 }, { "grad_norm": 0.2528589963912964, "learning_rate": 7.166374905768111e-05, "loss": 0.0279, "step": 38950 }, { "grad_norm": 0.2288365215063095, "learning_rate": 7.164884578193401e-05, "loss": 0.0285, "step": 38960 }, { "grad_norm": 0.30300694704055786, "learning_rate": 7.163394013870168e-05, "loss": 0.0241, "step": 38970 }, { "grad_norm": 0.27553683519363403, "learning_rate": 7.161903212961419e-05, "loss": 0.0238, "step": 38980 }, { "grad_norm": 0.23671437799930573, "learning_rate": 7.160412175630183e-05, "loss": 0.0262, "step": 38990 }, { "grad_norm": 0.226639986038208, "learning_rate": 7.158920902039521e-05, "loss": 0.0214, "step": 39000 }, { "grad_norm": 0.17093831300735474, "learning_rate": 7.157429392352514e-05, "loss": 0.0268, "step": 39010 }, { "grad_norm": 0.21296001970767975, "learning_rate": 7.155937646732275e-05, "loss": 0.0271, "step": 39020 }, { "grad_norm": 0.2284618467092514, "learning_rate": 7.154445665341933e-05, "loss": 0.0226, "step": 39030 }, { "grad_norm": 0.2860146164894104, "learning_rate": 7.152953448344654e-05, "loss": 0.0247, "step": 39040 }, { "grad_norm": 0.21968887746334076, "learning_rate": 7.151460995903624e-05, "loss": 0.0252, "step": 39050 }, { "grad_norm": 0.2533744275569916, "learning_rate": 7.149968308182052e-05, "loss": 0.0226, "step": 39060 }, { "grad_norm": 0.20565173029899597, "learning_rate": 7.14847538534318e-05, "loss": 0.0197, "step": 39070 }, { "grad_norm": 0.2048238068819046, "learning_rate": 7.14698222755027e-05, "loss": 0.0259, "step": 39080 }, { "grad_norm": 0.24196255207061768, "learning_rate": 7.14548883496661e-05, "loss": 0.0239, "step": 39090 }, { "grad_norm": 0.24086061120033264, "learning_rate": 7.143995207755517e-05, "loss": 0.0282, "step": 39100 }, { "grad_norm": 0.2822672724723816, "learning_rate": 7.142501346080333e-05, "loss": 0.0244, "step": 39110 }, { "grad_norm": 0.1975717842578888, "learning_rate": 7.141007250104421e-05, "loss": 0.022, "step": 39120 }, { "grad_norm": 0.2239745855331421, "learning_rate": 7.139512919991176e-05, "loss": 0.0259, "step": 39130 }, { "grad_norm": 0.16720841825008392, "learning_rate": 7.138018355904014e-05, "loss": 0.0234, "step": 39140 }, { "grad_norm": 0.17322580516338348, "learning_rate": 7.13652355800638e-05, "loss": 0.0239, "step": 39150 }, { "grad_norm": 0.21780526638031006, "learning_rate": 7.135028526461743e-05, "loss": 0.0217, "step": 39160 }, { "grad_norm": 0.1625388264656067, "learning_rate": 7.133533261433594e-05, "loss": 0.0213, "step": 39170 }, { "grad_norm": 0.2634647488594055, "learning_rate": 7.132037763085457e-05, "loss": 0.0262, "step": 39180 }, { "grad_norm": 0.2093207985162735, "learning_rate": 7.130542031580875e-05, "loss": 0.0285, "step": 39190 }, { "grad_norm": 0.2429186999797821, "learning_rate": 7.12904606708342e-05, "loss": 0.0288, "step": 39200 }, { "grad_norm": 0.2854064106941223, "learning_rate": 7.127549869756687e-05, "loss": 0.0251, "step": 39210 }, { "grad_norm": 0.18013708293437958, "learning_rate": 7.126053439764299e-05, "loss": 0.0225, "step": 39220 }, { "grad_norm": 0.2256959080696106, "learning_rate": 7.124556777269904e-05, "loss": 0.0247, "step": 39230 }, { "grad_norm": 0.21784880757331848, "learning_rate": 7.123059882437174e-05, "loss": 0.0252, "step": 39240 }, { "grad_norm": 0.20987634360790253, "learning_rate": 7.121562755429807e-05, "loss": 0.0215, "step": 39250 }, { "grad_norm": 0.21403010189533234, "learning_rate": 7.120065396411528e-05, "loss": 0.0213, "step": 39260 }, { "grad_norm": 0.24804210662841797, "learning_rate": 7.118567805546084e-05, "loss": 0.0276, "step": 39270 }, { "grad_norm": 0.2485973984003067, "learning_rate": 7.117069982997248e-05, "loss": 0.0253, "step": 39280 }, { "grad_norm": 0.31445950269699097, "learning_rate": 7.115571928928825e-05, "loss": 0.0264, "step": 39290 }, { "grad_norm": 0.22802598774433136, "learning_rate": 7.114073643504635e-05, "loss": 0.0247, "step": 39300 }, { "grad_norm": 0.22179587185382843, "learning_rate": 7.11257512688853e-05, "loss": 0.0262, "step": 39310 }, { "grad_norm": 0.25634652376174927, "learning_rate": 7.111076379244384e-05, "loss": 0.023, "step": 39320 }, { "grad_norm": 0.21820323169231415, "learning_rate": 7.109577400736101e-05, "loss": 0.0192, "step": 39330 }, { "grad_norm": 0.30189546942710876, "learning_rate": 7.108078191527605e-05, "loss": 0.0236, "step": 39340 }, { "grad_norm": 0.15603455901145935, "learning_rate": 7.106578751782847e-05, "loss": 0.0242, "step": 39350 }, { "grad_norm": 0.1728152483701706, "learning_rate": 7.105079081665803e-05, "loss": 0.0244, "step": 39360 }, { "grad_norm": 0.17285794019699097, "learning_rate": 7.103579181340476e-05, "loss": 0.0236, "step": 39370 }, { "grad_norm": 0.2563321590423584, "learning_rate": 7.102079050970893e-05, "loss": 0.0235, "step": 39380 }, { "grad_norm": 0.4320881962776184, "learning_rate": 7.100578690721105e-05, "loss": 0.0202, "step": 39390 }, { "grad_norm": 0.22995375096797943, "learning_rate": 7.09907810075519e-05, "loss": 0.0215, "step": 39400 }, { "grad_norm": 0.1890963762998581, "learning_rate": 7.097577281237249e-05, "loss": 0.0237, "step": 39410 }, { "grad_norm": 0.3664630949497223, "learning_rate": 7.09607623233141e-05, "loss": 0.027, "step": 39420 }, { "grad_norm": 0.5916089415550232, "learning_rate": 7.094574954201827e-05, "loss": 0.0206, "step": 39430 }, { "grad_norm": 0.27101191878318787, "learning_rate": 7.093073447012675e-05, "loss": 0.0265, "step": 39440 }, { "grad_norm": 0.23995277285575867, "learning_rate": 7.09157171092816e-05, "loss": 0.0254, "step": 39450 }, { "grad_norm": 0.1823733150959015, "learning_rate": 7.090069746112504e-05, "loss": 0.0241, "step": 39460 }, { "grad_norm": 0.2080344259738922, "learning_rate": 7.088567552729965e-05, "loss": 0.0264, "step": 39470 }, { "grad_norm": 0.21220527589321136, "learning_rate": 7.087065130944818e-05, "loss": 0.0237, "step": 39480 }, { "grad_norm": 0.2578792870044708, "learning_rate": 7.085562480921366e-05, "loss": 0.0226, "step": 39490 }, { "grad_norm": 0.20966003835201263, "learning_rate": 7.084059602823937e-05, "loss": 0.0221, "step": 39500 }, { "grad_norm": 0.1615360528230667, "learning_rate": 7.082556496816882e-05, "loss": 0.02, "step": 39510 }, { "grad_norm": 0.17929205298423767, "learning_rate": 7.081053163064582e-05, "loss": 0.0227, "step": 39520 }, { "grad_norm": 0.20144590735435486, "learning_rate": 7.079549601731434e-05, "loss": 0.0249, "step": 39530 }, { "grad_norm": 0.322948694229126, "learning_rate": 7.07804581298187e-05, "loss": 0.0258, "step": 39540 }, { "grad_norm": 0.2736373841762543, "learning_rate": 7.07654179698034e-05, "loss": 0.022, "step": 39550 }, { "grad_norm": 0.15197592973709106, "learning_rate": 7.075037553891321e-05, "loss": 0.0229, "step": 39560 }, { "grad_norm": 0.1839665174484253, "learning_rate": 7.073533083879315e-05, "loss": 0.0233, "step": 39570 }, { "grad_norm": 0.18636880815029144, "learning_rate": 7.072028387108849e-05, "loss": 0.0228, "step": 39580 }, { "grad_norm": 0.3450333774089813, "learning_rate": 7.070523463744472e-05, "loss": 0.0212, "step": 39590 }, { "grad_norm": 0.2715063691139221, "learning_rate": 7.069018313950763e-05, "loss": 0.0208, "step": 39600 }, { "grad_norm": 0.1925961673259735, "learning_rate": 7.067512937892322e-05, "loss": 0.0229, "step": 39610 }, { "grad_norm": 0.21223048865795135, "learning_rate": 7.066007335733775e-05, "loss": 0.0218, "step": 39620 }, { "grad_norm": 0.21583575010299683, "learning_rate": 7.064501507639772e-05, "loss": 0.0245, "step": 39630 }, { "grad_norm": 0.19261977076530457, "learning_rate": 7.062995453774987e-05, "loss": 0.0241, "step": 39640 }, { "grad_norm": 0.30380508303642273, "learning_rate": 7.061489174304121e-05, "loss": 0.0216, "step": 39650 }, { "grad_norm": 0.21412868797779083, "learning_rate": 7.0599826693919e-05, "loss": 0.0274, "step": 39660 }, { "grad_norm": 0.1778029054403305, "learning_rate": 7.05847593920307e-05, "loss": 0.0211, "step": 39670 }, { "grad_norm": 0.30328503251075745, "learning_rate": 7.056968983902406e-05, "loss": 0.0259, "step": 39680 }, { "grad_norm": 0.21490038931369781, "learning_rate": 7.055461803654706e-05, "loss": 0.0219, "step": 39690 }, { "grad_norm": 0.2023099958896637, "learning_rate": 7.053954398624794e-05, "loss": 0.0218, "step": 39700 }, { "grad_norm": 0.24662066996097565, "learning_rate": 7.052446768977518e-05, "loss": 0.0266, "step": 39710 }, { "grad_norm": 0.27021560072898865, "learning_rate": 7.050938914877748e-05, "loss": 0.0262, "step": 39720 }, { "grad_norm": 0.2793995440006256, "learning_rate": 7.049430836490382e-05, "loss": 0.0258, "step": 39730 }, { "grad_norm": 0.22343727946281433, "learning_rate": 7.047922533980342e-05, "loss": 0.0269, "step": 39740 }, { "grad_norm": 0.3461609482765198, "learning_rate": 7.046414007512571e-05, "loss": 0.0253, "step": 39750 }, { "grad_norm": 0.2385825514793396, "learning_rate": 7.044905257252042e-05, "loss": 0.0246, "step": 39760 }, { "grad_norm": 0.20072488486766815, "learning_rate": 7.04339628336375e-05, "loss": 0.0205, "step": 39770 }, { "grad_norm": 0.23255391418933868, "learning_rate": 7.041887086012711e-05, "loss": 0.0281, "step": 39780 }, { "grad_norm": 0.2609987258911133, "learning_rate": 7.040377665363969e-05, "loss": 0.0243, "step": 39790 }, { "grad_norm": 0.252695769071579, "learning_rate": 7.038868021582594e-05, "loss": 0.0238, "step": 39800 }, { "grad_norm": 0.19461439549922943, "learning_rate": 7.037358154833679e-05, "loss": 0.0213, "step": 39810 }, { "grad_norm": 0.21585147082805634, "learning_rate": 7.035848065282339e-05, "loss": 0.0308, "step": 39820 }, { "grad_norm": 0.20250627398490906, "learning_rate": 7.034337753093714e-05, "loss": 0.0222, "step": 39830 }, { "grad_norm": 0.34976962208747864, "learning_rate": 7.032827218432972e-05, "loss": 0.0215, "step": 39840 }, { "grad_norm": 0.1969427615404129, "learning_rate": 7.031316461465302e-05, "loss": 0.0226, "step": 39850 }, { "grad_norm": 0.21878015995025635, "learning_rate": 7.029805482355915e-05, "loss": 0.0219, "step": 39860 }, { "grad_norm": 0.18576841056346893, "learning_rate": 7.028294281270055e-05, "loss": 0.0214, "step": 39870 }, { "grad_norm": 0.24157527089118958, "learning_rate": 7.026782858372982e-05, "loss": 0.0237, "step": 39880 }, { "grad_norm": 0.23142072558403015, "learning_rate": 7.025271213829982e-05, "loss": 0.0265, "step": 39890 }, { "grad_norm": 0.20834873616695404, "learning_rate": 7.023759347806366e-05, "loss": 0.0241, "step": 39900 }, { "grad_norm": 0.17807438969612122, "learning_rate": 7.022247260467472e-05, "loss": 0.0264, "step": 39910 }, { "grad_norm": 0.28107213973999023, "learning_rate": 7.020734951978658e-05, "loss": 0.0278, "step": 39920 }, { "grad_norm": 0.22673851251602173, "learning_rate": 7.019222422505307e-05, "loss": 0.0247, "step": 39930 }, { "grad_norm": 0.3679697811603546, "learning_rate": 7.017709672212827e-05, "loss": 0.0296, "step": 39940 }, { "grad_norm": 0.1996353566646576, "learning_rate": 7.016196701266652e-05, "loss": 0.0256, "step": 39950 }, { "grad_norm": 0.22771453857421875, "learning_rate": 7.014683509832238e-05, "loss": 0.0234, "step": 39960 }, { "grad_norm": 0.24236196279525757, "learning_rate": 7.013170098075063e-05, "loss": 0.0249, "step": 39970 }, { "grad_norm": 0.4212211072444916, "learning_rate": 7.011656466160632e-05, "loss": 0.0255, "step": 39980 }, { "grad_norm": 0.2239762842655182, "learning_rate": 7.010142614254475e-05, "loss": 0.0284, "step": 39990 }, { "grad_norm": 0.21146897971630096, "learning_rate": 7.008628542522147e-05, "loss": 0.0221, "step": 40000 }, { "grad_norm": 0.19032371044158936, "learning_rate": 7.007114251129218e-05, "loss": 0.0281, "step": 40010 }, { "grad_norm": 0.2358899563550949, "learning_rate": 7.005599740241293e-05, "loss": 0.0217, "step": 40020 }, { "grad_norm": 0.25761878490448, "learning_rate": 7.004085010023996e-05, "loss": 0.0238, "step": 40030 }, { "grad_norm": 0.1868135631084442, "learning_rate": 7.002570060642976e-05, "loss": 0.0255, "step": 40040 }, { "grad_norm": 0.20123626291751862, "learning_rate": 7.001054892263903e-05, "loss": 0.0261, "step": 40050 }, { "grad_norm": 0.2842618525028229, "learning_rate": 6.999539505052477e-05, "loss": 0.0242, "step": 40060 }, { "grad_norm": 0.2222726047039032, "learning_rate": 6.998023899174418e-05, "loss": 0.0242, "step": 40070 }, { "grad_norm": 0.24662524461746216, "learning_rate": 6.996508074795467e-05, "loss": 0.0243, "step": 40080 }, { "grad_norm": 0.23619592189788818, "learning_rate": 6.994992032081396e-05, "loss": 0.024, "step": 40090 }, { "grad_norm": 0.1555788516998291, "learning_rate": 6.993475771197995e-05, "loss": 0.0236, "step": 40100 }, { "grad_norm": 0.29173192381858826, "learning_rate": 6.991959292311082e-05, "loss": 0.0244, "step": 40110 }, { "grad_norm": 0.2072954624891281, "learning_rate": 6.990442595586495e-05, "loss": 0.0237, "step": 40120 }, { "grad_norm": 0.29148486256599426, "learning_rate": 6.988925681190098e-05, "loss": 0.0253, "step": 40130 }, { "grad_norm": 0.20005789399147034, "learning_rate": 6.987408549287778e-05, "loss": 0.0286, "step": 40140 }, { "grad_norm": 0.22202220559120178, "learning_rate": 6.985891200045449e-05, "loss": 0.0215, "step": 40150 }, { "grad_norm": 0.17496182024478912, "learning_rate": 6.984373633629045e-05, "loss": 0.0238, "step": 40160 }, { "grad_norm": 0.217220276594162, "learning_rate": 6.98285585020452e-05, "loss": 0.0258, "step": 40170 }, { "grad_norm": 0.2771781384944916, "learning_rate": 6.981337849937864e-05, "loss": 0.0277, "step": 40180 }, { "grad_norm": 0.26761019229888916, "learning_rate": 6.979819632995078e-05, "loss": 0.0276, "step": 40190 }, { "grad_norm": 0.24086397886276245, "learning_rate": 6.978301199542193e-05, "loss": 0.0302, "step": 40200 }, { "grad_norm": 0.31096407771110535, "learning_rate": 6.976782549745263e-05, "loss": 0.0296, "step": 40210 }, { "grad_norm": 0.24514365196228027, "learning_rate": 6.975263683770367e-05, "loss": 0.0259, "step": 40220 }, { "grad_norm": 0.25633010268211365, "learning_rate": 6.973744601783604e-05, "loss": 0.0243, "step": 40230 }, { "grad_norm": 0.24997855722904205, "learning_rate": 6.972225303951097e-05, "loss": 0.0278, "step": 40240 }, { "grad_norm": 0.23856675624847412, "learning_rate": 6.970705790438998e-05, "loss": 0.0246, "step": 40250 }, { "grad_norm": 0.20061616599559784, "learning_rate": 6.969186061413477e-05, "loss": 0.0267, "step": 40260 }, { "grad_norm": 0.26699191331863403, "learning_rate": 6.967666117040727e-05, "loss": 0.0256, "step": 40270 }, { "grad_norm": 0.2922615110874176, "learning_rate": 6.966145957486972e-05, "loss": 0.0213, "step": 40280 }, { "grad_norm": 0.25796204805374146, "learning_rate": 6.964625582918449e-05, "loss": 0.023, "step": 40290 }, { "grad_norm": 0.248200461268425, "learning_rate": 6.963104993501425e-05, "loss": 0.0354, "step": 40300 }, { "grad_norm": 0.19928519427776337, "learning_rate": 6.961584189402192e-05, "loss": 0.0227, "step": 40310 }, { "grad_norm": 0.20323419570922852, "learning_rate": 6.96006317078706e-05, "loss": 0.025, "step": 40320 }, { "grad_norm": 0.21084320545196533, "learning_rate": 6.958541937822367e-05, "loss": 0.021, "step": 40330 }, { "grad_norm": 0.21641166508197784, "learning_rate": 6.957020490674473e-05, "loss": 0.0212, "step": 40340 }, { "grad_norm": 0.27209025621414185, "learning_rate": 6.95549882950976e-05, "loss": 0.0213, "step": 40350 }, { "grad_norm": 0.2474859356880188, "learning_rate": 6.953976954494635e-05, "loss": 0.0221, "step": 40360 }, { "grad_norm": 0.19175514578819275, "learning_rate": 6.952454865795528e-05, "loss": 0.0254, "step": 40370 }, { "grad_norm": 0.22701048851013184, "learning_rate": 6.950932563578892e-05, "loss": 0.0207, "step": 40380 }, { "grad_norm": 0.22029590606689453, "learning_rate": 6.949410048011206e-05, "loss": 0.0218, "step": 40390 }, { "grad_norm": 0.2334885448217392, "learning_rate": 6.947887319258966e-05, "loss": 0.022, "step": 40400 }, { "grad_norm": 0.22966456413269043, "learning_rate": 6.946364377488696e-05, "loss": 0.0251, "step": 40410 }, { "grad_norm": 0.2817639410495758, "learning_rate": 6.944841222866947e-05, "loss": 0.0232, "step": 40420 }, { "grad_norm": 0.2721308767795563, "learning_rate": 6.943317855560284e-05, "loss": 0.0235, "step": 40430 }, { "grad_norm": 0.1844845563173294, "learning_rate": 6.941794275735302e-05, "loss": 0.0274, "step": 40440 }, { "grad_norm": 0.28063175082206726, "learning_rate": 6.94027048355862e-05, "loss": 0.0242, "step": 40450 }, { "grad_norm": 0.20570598542690277, "learning_rate": 6.938746479196871e-05, "loss": 0.0207, "step": 40460 }, { "grad_norm": 0.16871030628681183, "learning_rate": 6.937222262816724e-05, "loss": 0.0195, "step": 40470 }, { "grad_norm": 0.314848929643631, "learning_rate": 6.935697834584865e-05, "loss": 0.0241, "step": 40480 }, { "grad_norm": 0.1900281012058258, "learning_rate": 6.934173194667998e-05, "loss": 0.0237, "step": 40490 }, { "grad_norm": 0.17009775340557098, "learning_rate": 6.93264834323286e-05, "loss": 0.0273, "step": 40500 }, { "grad_norm": 0.19571220874786377, "learning_rate": 6.931123280446206e-05, "loss": 0.0236, "step": 40510 }, { "grad_norm": 0.2761209309101105, "learning_rate": 6.929598006474811e-05, "loss": 0.0237, "step": 40520 }, { "grad_norm": 0.2015313357114792, "learning_rate": 6.92807252148548e-05, "loss": 0.0229, "step": 40530 }, { "grad_norm": 0.21801702678203583, "learning_rate": 6.926546825645037e-05, "loss": 0.0209, "step": 40540 }, { "grad_norm": 0.25887632369995117, "learning_rate": 6.92502091912033e-05, "loss": 0.0206, "step": 40550 }, { "grad_norm": 0.24262785911560059, "learning_rate": 6.923494802078229e-05, "loss": 0.022, "step": 40560 }, { "grad_norm": 0.2472030520439148, "learning_rate": 6.921968474685628e-05, "loss": 0.024, "step": 40570 }, { "grad_norm": 0.2036551535129547, "learning_rate": 6.920441937109445e-05, "loss": 0.0233, "step": 40580 }, { "grad_norm": 0.17671546339988708, "learning_rate": 6.91891518951662e-05, "loss": 0.0217, "step": 40590 }, { "grad_norm": 0.2184869796037674, "learning_rate": 6.917388232074114e-05, "loss": 0.0228, "step": 40600 }, { "grad_norm": 0.26931512355804443, "learning_rate": 6.915861064948914e-05, "loss": 0.0229, "step": 40610 }, { "grad_norm": 0.1955084353685379, "learning_rate": 6.914333688308025e-05, "loss": 0.0259, "step": 40620 }, { "grad_norm": 0.18778233230113983, "learning_rate": 6.912806102318487e-05, "loss": 0.0196, "step": 40630 }, { "grad_norm": 0.23650841414928436, "learning_rate": 6.911278307147347e-05, "loss": 0.0224, "step": 40640 }, { "grad_norm": 0.20979829132556915, "learning_rate": 6.909750302961684e-05, "loss": 0.0215, "step": 40650 }, { "grad_norm": 0.22403733432292938, "learning_rate": 6.9082220899286e-05, "loss": 0.024, "step": 40660 }, { "grad_norm": 0.18675045669078827, "learning_rate": 6.906693668215219e-05, "loss": 0.0232, "step": 40670 }, { "grad_norm": 0.32059356570243835, "learning_rate": 6.905165037988683e-05, "loss": 0.0227, "step": 40680 }, { "grad_norm": 0.2641965448856354, "learning_rate": 6.903636199416164e-05, "loss": 0.0233, "step": 40690 }, { "grad_norm": 0.19324582815170288, "learning_rate": 6.902107152664851e-05, "loss": 0.0265, "step": 40700 }, { "grad_norm": 0.25027021765708923, "learning_rate": 6.90057789790196e-05, "loss": 0.0244, "step": 40710 }, { "grad_norm": 0.3234317898750305, "learning_rate": 6.899048435294728e-05, "loss": 0.0225, "step": 40720 }, { "grad_norm": 0.23649443686008453, "learning_rate": 6.897518765010415e-05, "loss": 0.0259, "step": 40730 }, { "grad_norm": 0.17729061841964722, "learning_rate": 6.895988887216302e-05, "loss": 0.0266, "step": 40740 }, { "grad_norm": 0.22842727601528168, "learning_rate": 6.894458802079694e-05, "loss": 0.0233, "step": 40750 }, { "grad_norm": 0.8565964102745056, "learning_rate": 6.892928509767922e-05, "loss": 0.0305, "step": 40760 }, { "grad_norm": 0.2512000799179077, "learning_rate": 6.891398010448333e-05, "loss": 0.0248, "step": 40770 }, { "grad_norm": 0.3166733384132385, "learning_rate": 6.8898673042883e-05, "loss": 0.0279, "step": 40780 }, { "grad_norm": 0.27894148230552673, "learning_rate": 6.888336391455222e-05, "loss": 0.0221, "step": 40790 }, { "grad_norm": 0.2014530450105667, "learning_rate": 6.886805272116513e-05, "loss": 0.0293, "step": 40800 }, { "grad_norm": 0.2354075163602829, "learning_rate": 6.885273946439617e-05, "loss": 0.0254, "step": 40810 }, { "grad_norm": 0.2859761416912079, "learning_rate": 6.883742414591998e-05, "loss": 0.0213, "step": 40820 }, { "grad_norm": 0.2748572528362274, "learning_rate": 6.88221067674114e-05, "loss": 0.02, "step": 40830 }, { "grad_norm": 0.26070451736450195, "learning_rate": 6.88067873305455e-05, "loss": 0.0201, "step": 40840 }, { "grad_norm": 0.2582380473613739, "learning_rate": 6.879146583699765e-05, "loss": 0.0216, "step": 40850 }, { "grad_norm": 0.19462266564369202, "learning_rate": 6.877614228844334e-05, "loss": 0.0202, "step": 40860 }, { "grad_norm": 0.22071586549282074, "learning_rate": 6.876081668655832e-05, "loss": 0.0245, "step": 40870 }, { "grad_norm": 0.21056291460990906, "learning_rate": 6.874548903301861e-05, "loss": 0.0275, "step": 40880 }, { "grad_norm": 0.1674399971961975, "learning_rate": 6.873015932950039e-05, "loss": 0.0241, "step": 40890 }, { "grad_norm": 0.2157321572303772, "learning_rate": 6.871482757768012e-05, "loss": 0.0207, "step": 40900 }, { "grad_norm": 0.14949913322925568, "learning_rate": 6.869949377923442e-05, "loss": 0.0257, "step": 40910 }, { "grad_norm": 0.168420672416687, "learning_rate": 6.868415793584022e-05, "loss": 0.0272, "step": 40920 }, { "grad_norm": 0.21180753409862518, "learning_rate": 6.86688200491746e-05, "loss": 0.0237, "step": 40930 }, { "grad_norm": 0.2698442041873932, "learning_rate": 6.865348012091488e-05, "loss": 0.0235, "step": 40940 }, { "grad_norm": 0.19871565699577332, "learning_rate": 6.86381381527386e-05, "loss": 0.0275, "step": 40950 }, { "grad_norm": 0.2131182849407196, "learning_rate": 6.862279414632359e-05, "loss": 0.0234, "step": 40960 }, { "grad_norm": 0.22011853754520416, "learning_rate": 6.860744810334778e-05, "loss": 0.0287, "step": 40970 }, { "grad_norm": 0.20092491805553436, "learning_rate": 6.859210002548943e-05, "loss": 0.0216, "step": 40980 }, { "grad_norm": 0.2158684879541397, "learning_rate": 6.8576749914427e-05, "loss": 0.0216, "step": 40990 }, { "grad_norm": 0.22681944072246552, "learning_rate": 6.85613977718391e-05, "loss": 0.0247, "step": 41000 }, { "grad_norm": 0.2187757045030594, "learning_rate": 6.854604359940464e-05, "loss": 0.0218, "step": 41010 }, { "grad_norm": 0.21178743243217468, "learning_rate": 6.853068739880276e-05, "loss": 0.0218, "step": 41020 }, { "grad_norm": 0.249098002910614, "learning_rate": 6.851532917171276e-05, "loss": 0.0289, "step": 41030 }, { "grad_norm": 0.2486528903245926, "learning_rate": 6.849996891981418e-05, "loss": 0.0206, "step": 41040 }, { "grad_norm": 0.33041372895240784, "learning_rate": 6.848460664478684e-05, "loss": 0.0283, "step": 41050 }, { "grad_norm": 0.24097900092601776, "learning_rate": 6.846924234831069e-05, "loss": 0.0237, "step": 41060 }, { "grad_norm": 0.2566114068031311, "learning_rate": 6.845387603206597e-05, "loss": 0.0214, "step": 41070 }, { "grad_norm": 0.26472046971321106, "learning_rate": 6.843850769773312e-05, "loss": 0.0211, "step": 41080 }, { "grad_norm": 0.23466211557388306, "learning_rate": 6.842313734699278e-05, "loss": 0.0227, "step": 41090 }, { "grad_norm": 0.17798738181591034, "learning_rate": 6.840776498152584e-05, "loss": 0.026, "step": 41100 }, { "grad_norm": 0.26221632957458496, "learning_rate": 6.83923906030134e-05, "loss": 0.0263, "step": 41110 }, { "grad_norm": 0.20951421558856964, "learning_rate": 6.837701421313677e-05, "loss": 0.0286, "step": 41120 }, { "grad_norm": 0.20225515961647034, "learning_rate": 6.83616358135775e-05, "loss": 0.0253, "step": 41130 }, { "grad_norm": 0.3174423277378082, "learning_rate": 6.834625540601734e-05, "loss": 0.0265, "step": 41140 }, { "grad_norm": 0.27753958106040955, "learning_rate": 6.833087299213829e-05, "loss": 0.0253, "step": 41150 }, { "grad_norm": 0.42729324102401733, "learning_rate": 6.83154885736225e-05, "loss": 0.0316, "step": 41160 }, { "grad_norm": 0.2485060691833496, "learning_rate": 6.830010215215244e-05, "loss": 0.0251, "step": 41170 }, { "grad_norm": 0.1977648138999939, "learning_rate": 6.828471372941072e-05, "loss": 0.0233, "step": 41180 }, { "grad_norm": 0.19773955643177032, "learning_rate": 6.826932330708018e-05, "loss": 0.0205, "step": 41190 }, { "grad_norm": 0.23903390765190125, "learning_rate": 6.825393088684393e-05, "loss": 0.0228, "step": 41200 }, { "grad_norm": 0.29134953022003174, "learning_rate": 6.823853647038524e-05, "loss": 0.0269, "step": 41210 }, { "grad_norm": 0.19716347754001617, "learning_rate": 6.822314005938762e-05, "loss": 0.0228, "step": 41220 }, { "grad_norm": 0.19907815754413605, "learning_rate": 6.820774165553479e-05, "loss": 0.0219, "step": 41230 }, { "grad_norm": 0.20597942173480988, "learning_rate": 6.819234126051072e-05, "loss": 0.0208, "step": 41240 }, { "grad_norm": 0.20395077764987946, "learning_rate": 6.817693887599956e-05, "loss": 0.021, "step": 41250 }, { "grad_norm": 0.21662472188472748, "learning_rate": 6.816153450368568e-05, "loss": 0.0256, "step": 41260 }, { "grad_norm": 0.20167379081249237, "learning_rate": 6.814612814525372e-05, "loss": 0.0261, "step": 41270 }, { "grad_norm": 0.33208969235420227, "learning_rate": 6.813071980238846e-05, "loss": 0.0242, "step": 41280 }, { "grad_norm": 0.17906178534030914, "learning_rate": 6.811530947677492e-05, "loss": 0.0226, "step": 41290 }, { "grad_norm": 0.22288201749324799, "learning_rate": 6.809989717009839e-05, "loss": 0.029, "step": 41300 }, { "grad_norm": 0.210755854845047, "learning_rate": 6.808448288404431e-05, "loss": 0.0247, "step": 41310 }, { "grad_norm": 0.20903438329696655, "learning_rate": 6.806906662029838e-05, "loss": 0.0311, "step": 41320 }, { "grad_norm": 0.29341983795166016, "learning_rate": 6.805364838054647e-05, "loss": 0.0257, "step": 41330 }, { "grad_norm": 0.19854266941547394, "learning_rate": 6.803822816647474e-05, "loss": 0.0226, "step": 41340 }, { "grad_norm": 0.2708945572376251, "learning_rate": 6.802280597976949e-05, "loss": 0.0243, "step": 41350 }, { "grad_norm": 0.24463097751140594, "learning_rate": 6.800738182211727e-05, "loss": 0.0333, "step": 41360 }, { "grad_norm": 0.3630925118923187, "learning_rate": 6.799195569520485e-05, "loss": 0.027, "step": 41370 }, { "grad_norm": 0.27364638447761536, "learning_rate": 6.79765276007192e-05, "loss": 0.0271, "step": 41380 }, { "grad_norm": 0.228692889213562, "learning_rate": 6.796109754034753e-05, "loss": 0.0303, "step": 41390 }, { "grad_norm": 0.2601059377193451, "learning_rate": 6.794566551577724e-05, "loss": 0.0236, "step": 41400 }, { "grad_norm": 0.1712338924407959, "learning_rate": 6.793023152869594e-05, "loss": 0.0222, "step": 41410 }, { "grad_norm": 0.22233334183692932, "learning_rate": 6.791479558079147e-05, "loss": 0.0242, "step": 41420 }, { "grad_norm": 0.24897733330726624, "learning_rate": 6.789935767375191e-05, "loss": 0.0261, "step": 41430 }, { "grad_norm": 0.2930731177330017, "learning_rate": 6.78839178092655e-05, "loss": 0.0233, "step": 41440 }, { "grad_norm": 0.22111856937408447, "learning_rate": 6.786847598902072e-05, "loss": 0.0277, "step": 41450 }, { "grad_norm": 0.22309353947639465, "learning_rate": 6.785303221470627e-05, "loss": 0.0234, "step": 41460 }, { "grad_norm": 0.25670167803764343, "learning_rate": 6.783758648801108e-05, "loss": 0.0256, "step": 41470 }, { "grad_norm": 0.21573351323604584, "learning_rate": 6.782213881062422e-05, "loss": 0.0285, "step": 41480 }, { "grad_norm": 0.22076091170310974, "learning_rate": 6.780668918423508e-05, "loss": 0.0238, "step": 41490 }, { "grad_norm": 0.21445752680301666, "learning_rate": 6.779123761053317e-05, "loss": 0.0227, "step": 41500 }, { "grad_norm": 0.26359736919403076, "learning_rate": 6.777578409120826e-05, "loss": 0.0275, "step": 41510 }, { "grad_norm": 0.18282996118068695, "learning_rate": 6.776032862795033e-05, "loss": 0.0229, "step": 41520 }, { "grad_norm": 0.18284669518470764, "learning_rate": 6.774487122244956e-05, "loss": 0.0266, "step": 41530 }, { "grad_norm": 0.24201616644859314, "learning_rate": 6.772941187639637e-05, "loss": 0.0241, "step": 41540 }, { "grad_norm": 0.23608067631721497, "learning_rate": 6.771395059148134e-05, "loss": 0.0213, "step": 41550 }, { "grad_norm": 0.17487750947475433, "learning_rate": 6.76984873693953e-05, "loss": 0.02, "step": 41560 }, { "grad_norm": 0.18465270102024078, "learning_rate": 6.768302221182931e-05, "loss": 0.0199, "step": 41570 }, { "grad_norm": 0.3125785291194916, "learning_rate": 6.766755512047457e-05, "loss": 0.0213, "step": 41580 }, { "grad_norm": 0.24724681675434113, "learning_rate": 6.765208609702259e-05, "loss": 0.0222, "step": 41590 }, { "grad_norm": 0.26769188046455383, "learning_rate": 6.763661514316499e-05, "loss": 0.022, "step": 41600 }, { "grad_norm": 0.2422814816236496, "learning_rate": 6.76211422605937e-05, "loss": 0.0225, "step": 41610 }, { "grad_norm": 0.21015165746212006, "learning_rate": 6.760566745100076e-05, "loss": 0.0205, "step": 41620 }, { "grad_norm": 0.35938185453414917, "learning_rate": 6.75901907160785e-05, "loss": 0.0264, "step": 41630 }, { "grad_norm": 0.17050200700759888, "learning_rate": 6.757471205751943e-05, "loss": 0.0199, "step": 41640 }, { "grad_norm": 0.23265261948108673, "learning_rate": 6.75592314770163e-05, "loss": 0.0245, "step": 41650 }, { "grad_norm": 0.20545408129692078, "learning_rate": 6.7543748976262e-05, "loss": 0.0224, "step": 41660 }, { "grad_norm": 0.26590701937675476, "learning_rate": 6.752826455694968e-05, "loss": 0.0233, "step": 41670 }, { "grad_norm": 0.26302868127822876, "learning_rate": 6.751277822077271e-05, "loss": 0.0221, "step": 41680 }, { "grad_norm": 0.21462856233119965, "learning_rate": 6.749728996942465e-05, "loss": 0.0208, "step": 41690 }, { "grad_norm": 0.20948174595832825, "learning_rate": 6.748179980459924e-05, "loss": 0.0192, "step": 41700 }, { "grad_norm": 0.17822611331939697, "learning_rate": 6.746630772799052e-05, "loss": 0.0222, "step": 41710 }, { "grad_norm": 0.240628182888031, "learning_rate": 6.745081374129262e-05, "loss": 0.022, "step": 41720 }, { "grad_norm": 0.19781538844108582, "learning_rate": 6.743531784619998e-05, "loss": 0.0185, "step": 41730 }, { "grad_norm": 0.20684272050857544, "learning_rate": 6.741982004440719e-05, "loss": 0.0221, "step": 41740 }, { "grad_norm": 0.19786953926086426, "learning_rate": 6.740432033760907e-05, "loss": 0.0216, "step": 41750 }, { "grad_norm": 0.3261364698410034, "learning_rate": 6.738881872750066e-05, "loss": 0.0245, "step": 41760 }, { "grad_norm": 0.1975441575050354, "learning_rate": 6.737331521577715e-05, "loss": 0.0221, "step": 41770 }, { "grad_norm": 0.19170492887496948, "learning_rate": 6.735780980413403e-05, "loss": 0.0212, "step": 41780 }, { "grad_norm": 0.2260657697916031, "learning_rate": 6.734230249426692e-05, "loss": 0.0184, "step": 41790 }, { "grad_norm": 0.21657946705818176, "learning_rate": 6.732679328787168e-05, "loss": 0.0197, "step": 41800 }, { "grad_norm": 0.19425146281719208, "learning_rate": 6.731128218664438e-05, "loss": 0.0256, "step": 41810 }, { "grad_norm": 0.2400985211133957, "learning_rate": 6.729576919228129e-05, "loss": 0.0209, "step": 41820 }, { "grad_norm": 0.22430746257305145, "learning_rate": 6.728025430647888e-05, "loss": 0.0208, "step": 41830 }, { "grad_norm": 0.2174525111913681, "learning_rate": 6.726473753093383e-05, "loss": 0.0208, "step": 41840 }, { "grad_norm": 0.278887540102005, "learning_rate": 6.724921886734305e-05, "loss": 0.0236, "step": 41850 }, { "grad_norm": 0.23528455197811127, "learning_rate": 6.723369831740363e-05, "loss": 0.0242, "step": 41860 }, { "grad_norm": 0.24678556621074677, "learning_rate": 6.721817588281288e-05, "loss": 0.0286, "step": 41870 }, { "grad_norm": 0.17241449654102325, "learning_rate": 6.720265156526828e-05, "loss": 0.0201, "step": 41880 }, { "grad_norm": 0.20419910550117493, "learning_rate": 6.718712536646758e-05, "loss": 0.0218, "step": 41890 }, { "grad_norm": 0.2290007621049881, "learning_rate": 6.71715972881087e-05, "loss": 0.0218, "step": 41900 }, { "grad_norm": 0.17152711749076843, "learning_rate": 6.715606733188974e-05, "loss": 0.0211, "step": 41910 }, { "grad_norm": 0.21691949665546417, "learning_rate": 6.714053549950909e-05, "loss": 0.0214, "step": 41920 }, { "grad_norm": 0.18126556277275085, "learning_rate": 6.71250017926652e-05, "loss": 0.0232, "step": 41930 }, { "grad_norm": 0.21111798286437988, "learning_rate": 6.71094662130569e-05, "loss": 0.0205, "step": 41940 }, { "grad_norm": 0.22324126958847046, "learning_rate": 6.709392876238307e-05, "loss": 0.022, "step": 41950 }, { "grad_norm": 0.2050296813249588, "learning_rate": 6.707838944234289e-05, "loss": 0.0252, "step": 41960 }, { "grad_norm": 0.19811323285102844, "learning_rate": 6.706284825463574e-05, "loss": 0.0196, "step": 41970 }, { "grad_norm": 0.22428566217422485, "learning_rate": 6.704730520096114e-05, "loss": 0.0224, "step": 41980 }, { "grad_norm": 0.19885312020778656, "learning_rate": 6.703176028301888e-05, "loss": 0.0278, "step": 41990 }, { "grad_norm": 0.2457142174243927, "learning_rate": 6.701621350250892e-05, "loss": 0.0239, "step": 42000 }, { "grad_norm": 0.23114697635173798, "learning_rate": 6.700066486113144e-05, "loss": 0.0254, "step": 42010 }, { "grad_norm": 0.21788400411605835, "learning_rate": 6.69851143605868e-05, "loss": 0.0217, "step": 42020 }, { "grad_norm": 0.1822504699230194, "learning_rate": 6.696956200257561e-05, "loss": 0.0206, "step": 42030 }, { "grad_norm": 0.18469974398612976, "learning_rate": 6.695400778879863e-05, "loss": 0.0246, "step": 42040 }, { "grad_norm": 0.18370778858661652, "learning_rate": 6.693845172095683e-05, "loss": 0.0247, "step": 42050 }, { "grad_norm": 0.18548673391342163, "learning_rate": 6.692289380075142e-05, "loss": 0.0194, "step": 42060 }, { "grad_norm": 0.22535574436187744, "learning_rate": 6.690733402988379e-05, "loss": 0.0218, "step": 42070 }, { "grad_norm": 0.23540961742401123, "learning_rate": 6.689177241005553e-05, "loss": 0.0239, "step": 42080 }, { "grad_norm": 0.30924192070961, "learning_rate": 6.687620894296844e-05, "loss": 0.0244, "step": 42090 }, { "grad_norm": 0.21222658455371857, "learning_rate": 6.686064363032451e-05, "loss": 0.0228, "step": 42100 }, { "grad_norm": 0.18549086153507233, "learning_rate": 6.684507647382596e-05, "loss": 0.0187, "step": 42110 }, { "grad_norm": 0.22897803783416748, "learning_rate": 6.682950747517513e-05, "loss": 0.0185, "step": 42120 }, { "grad_norm": 0.19651342928409576, "learning_rate": 6.68139366360747e-05, "loss": 0.0296, "step": 42130 }, { "grad_norm": 0.2272012084722519, "learning_rate": 6.679836395822744e-05, "loss": 0.0209, "step": 42140 }, { "grad_norm": 0.21333017945289612, "learning_rate": 6.678278944333633e-05, "loss": 0.0222, "step": 42150 }, { "grad_norm": 0.1632944494485855, "learning_rate": 6.676721309310462e-05, "loss": 0.0219, "step": 42160 }, { "grad_norm": 0.2565233111381531, "learning_rate": 6.67516349092357e-05, "loss": 0.0215, "step": 42170 }, { "grad_norm": 0.19363225996494293, "learning_rate": 6.673605489343315e-05, "loss": 0.0228, "step": 42180 }, { "grad_norm": 0.19114339351654053, "learning_rate": 6.672047304740081e-05, "loss": 0.0201, "step": 42190 }, { "grad_norm": 0.2133258730173111, "learning_rate": 6.670488937284268e-05, "loss": 0.022, "step": 42200 }, { "grad_norm": 0.13857045769691467, "learning_rate": 6.668930387146294e-05, "loss": 0.0213, "step": 42210 }, { "grad_norm": 0.2890845537185669, "learning_rate": 6.667371654496605e-05, "loss": 0.0255, "step": 42220 }, { "grad_norm": 0.23797732591629028, "learning_rate": 6.665812739505659e-05, "loss": 0.0211, "step": 42230 }, { "grad_norm": 0.23865702748298645, "learning_rate": 6.664253642343935e-05, "loss": 0.0205, "step": 42240 }, { "grad_norm": 0.21881645917892456, "learning_rate": 6.662694363181935e-05, "loss": 0.0289, "step": 42250 }, { "grad_norm": 0.22461265325546265, "learning_rate": 6.66113490219018e-05, "loss": 0.0242, "step": 42260 }, { "grad_norm": 0.1919666826725006, "learning_rate": 6.659575259539207e-05, "loss": 0.0211, "step": 42270 }, { "grad_norm": 0.20702332258224487, "learning_rate": 6.65801543539958e-05, "loss": 0.026, "step": 42280 }, { "grad_norm": 0.23094405233860016, "learning_rate": 6.65645542994188e-05, "loss": 0.0262, "step": 42290 }, { "grad_norm": 0.21541237831115723, "learning_rate": 6.654895243336702e-05, "loss": 0.0236, "step": 42300 }, { "grad_norm": 0.23261386156082153, "learning_rate": 6.65333487575467e-05, "loss": 0.0226, "step": 42310 }, { "grad_norm": 0.23384319245815277, "learning_rate": 6.651774327366422e-05, "loss": 0.0215, "step": 42320 }, { "grad_norm": 0.1906963735818863, "learning_rate": 6.650213598342613e-05, "loss": 0.0225, "step": 42330 }, { "grad_norm": 0.23858976364135742, "learning_rate": 6.648652688853931e-05, "loss": 0.0233, "step": 42340 }, { "grad_norm": 0.24952684342861176, "learning_rate": 6.647091599071066e-05, "loss": 0.0229, "step": 42350 }, { "grad_norm": 0.18001171946525574, "learning_rate": 6.645530329164743e-05, "loss": 0.0209, "step": 42360 }, { "grad_norm": 0.2569413185119629, "learning_rate": 6.643968879305694e-05, "loss": 0.0296, "step": 42370 }, { "grad_norm": 0.20620010793209076, "learning_rate": 6.642407249664683e-05, "loss": 0.0227, "step": 42380 }, { "grad_norm": 0.24905447661876678, "learning_rate": 6.640845440412483e-05, "loss": 0.0218, "step": 42390 }, { "grad_norm": 0.20384924113750458, "learning_rate": 6.639283451719893e-05, "loss": 0.0213, "step": 42400 }, { "grad_norm": 0.29338520765304565, "learning_rate": 6.637721283757729e-05, "loss": 0.0216, "step": 42410 }, { "grad_norm": 0.18383631110191345, "learning_rate": 6.636158936696828e-05, "loss": 0.0218, "step": 42420 }, { "grad_norm": 0.21860453486442566, "learning_rate": 6.634596410708047e-05, "loss": 0.0253, "step": 42430 }, { "grad_norm": 0.23164115846157074, "learning_rate": 6.633033705962257e-05, "loss": 0.0189, "step": 42440 }, { "grad_norm": 0.2074841409921646, "learning_rate": 6.631470822630359e-05, "loss": 0.0228, "step": 42450 }, { "grad_norm": 0.17822417616844177, "learning_rate": 6.629907760883263e-05, "loss": 0.0229, "step": 42460 }, { "grad_norm": 0.22959071397781372, "learning_rate": 6.628344520891905e-05, "loss": 0.0205, "step": 42470 }, { "grad_norm": 0.22813385725021362, "learning_rate": 6.62678110282724e-05, "loss": 0.0229, "step": 42480 }, { "grad_norm": 0.18025651574134827, "learning_rate": 6.625217506860237e-05, "loss": 0.0214, "step": 42490 }, { "grad_norm": 0.22602564096450806, "learning_rate": 6.623653733161892e-05, "loss": 0.0171, "step": 42500 }, { "grad_norm": 0.2518990635871887, "learning_rate": 6.622089781903217e-05, "loss": 0.0212, "step": 42510 }, { "grad_norm": 0.2300906479358673, "learning_rate": 6.620525653255241e-05, "loss": 0.0203, "step": 42520 }, { "grad_norm": 0.19781459867954254, "learning_rate": 6.618961347389015e-05, "loss": 0.0233, "step": 42530 }, { "grad_norm": 0.20003698766231537, "learning_rate": 6.617396864475613e-05, "loss": 0.0207, "step": 42540 }, { "grad_norm": 0.15013530850410461, "learning_rate": 6.61583220468612e-05, "loss": 0.0178, "step": 42550 }, { "grad_norm": 0.2886207401752472, "learning_rate": 6.614267368191645e-05, "loss": 0.0223, "step": 42560 }, { "grad_norm": 0.19922257959842682, "learning_rate": 6.61270235516332e-05, "loss": 0.0199, "step": 42570 }, { "grad_norm": 0.22875329852104187, "learning_rate": 6.611137165772287e-05, "loss": 0.025, "step": 42580 }, { "grad_norm": 0.16370582580566406, "learning_rate": 6.60957180018972e-05, "loss": 0.0226, "step": 42590 }, { "grad_norm": 0.256041556596756, "learning_rate": 6.608006258586797e-05, "loss": 0.0228, "step": 42600 }, { "grad_norm": 0.1605231761932373, "learning_rate": 6.606440541134731e-05, "loss": 0.0244, "step": 42610 }, { "grad_norm": 0.23881450295448303, "learning_rate": 6.60487464800474e-05, "loss": 0.0236, "step": 42620 }, { "grad_norm": 0.17236997187137604, "learning_rate": 6.603308579368071e-05, "loss": 0.0223, "step": 42630 }, { "grad_norm": 0.18951065838336945, "learning_rate": 6.601742335395987e-05, "loss": 0.0237, "step": 42640 }, { "grad_norm": 0.18212072551250458, "learning_rate": 6.600175916259769e-05, "loss": 0.0202, "step": 42650 }, { "grad_norm": 0.22357352077960968, "learning_rate": 6.598609322130718e-05, "loss": 0.0204, "step": 42660 }, { "grad_norm": 0.2815372943878174, "learning_rate": 6.597042553180155e-05, "loss": 0.0247, "step": 42670 }, { "grad_norm": 0.2447071373462677, "learning_rate": 6.59547560957942e-05, "loss": 0.0229, "step": 42680 }, { "grad_norm": 0.2246026247739792, "learning_rate": 6.59390849149987e-05, "loss": 0.024, "step": 42690 }, { "grad_norm": 0.23559851944446564, "learning_rate": 6.592341199112886e-05, "loss": 0.0273, "step": 42700 }, { "grad_norm": 0.2859216034412384, "learning_rate": 6.590773732589861e-05, "loss": 0.0231, "step": 42710 }, { "grad_norm": 0.2363184094429016, "learning_rate": 6.589206092102213e-05, "loss": 0.0258, "step": 42720 }, { "grad_norm": 0.3988356590270996, "learning_rate": 6.587638277821375e-05, "loss": 0.0243, "step": 42730 }, { "grad_norm": 0.2885708510875702, "learning_rate": 6.586070289918803e-05, "loss": 0.0282, "step": 42740 }, { "grad_norm": 0.21134693920612335, "learning_rate": 6.584502128565968e-05, "loss": 0.0247, "step": 42750 }, { "grad_norm": 0.16236615180969238, "learning_rate": 6.582933793934363e-05, "loss": 0.0234, "step": 42760 }, { "grad_norm": 0.20493757724761963, "learning_rate": 6.581365286195498e-05, "loss": 0.0247, "step": 42770 }, { "grad_norm": 0.23699812591075897, "learning_rate": 6.579796605520903e-05, "loss": 0.023, "step": 42780 }, { "grad_norm": 0.29025283455848694, "learning_rate": 6.578227752082127e-05, "loss": 0.0244, "step": 42790 }, { "grad_norm": 0.21307088434696198, "learning_rate": 6.576658726050735e-05, "loss": 0.0214, "step": 42800 }, { "grad_norm": 0.2364458441734314, "learning_rate": 6.575089527598316e-05, "loss": 0.022, "step": 42810 }, { "grad_norm": 0.17105752229690552, "learning_rate": 6.573520156896476e-05, "loss": 0.0218, "step": 42820 }, { "grad_norm": 0.2121848165988922, "learning_rate": 6.571950614116835e-05, "loss": 0.0258, "step": 42830 }, { "grad_norm": 0.1699030101299286, "learning_rate": 6.570380899431039e-05, "loss": 0.0228, "step": 42840 }, { "grad_norm": 0.2445462942123413, "learning_rate": 6.568811013010749e-05, "loss": 0.0213, "step": 42850 }, { "grad_norm": 0.24682405591011047, "learning_rate": 6.567240955027645e-05, "loss": 0.0239, "step": 42860 }, { "grad_norm": 0.2648177146911621, "learning_rate": 6.565670725653427e-05, "loss": 0.0265, "step": 42870 }, { "grad_norm": 0.24983267486095428, "learning_rate": 6.564100325059812e-05, "loss": 0.0228, "step": 42880 }, { "grad_norm": 0.2218683958053589, "learning_rate": 6.562529753418538e-05, "loss": 0.0257, "step": 42890 }, { "grad_norm": 0.1984541416168213, "learning_rate": 6.56095901090136e-05, "loss": 0.0247, "step": 42900 }, { "grad_norm": 0.2211386263370514, "learning_rate": 6.55938809768005e-05, "loss": 0.0224, "step": 42910 }, { "grad_norm": 0.17360684275627136, "learning_rate": 6.557817013926402e-05, "loss": 0.0218, "step": 42920 }, { "grad_norm": 0.17485913634300232, "learning_rate": 6.556245759812227e-05, "loss": 0.0205, "step": 42930 }, { "grad_norm": 0.2328832447528839, "learning_rate": 6.554674335509357e-05, "loss": 0.0218, "step": 42940 }, { "grad_norm": 0.23295602202415466, "learning_rate": 6.553102741189638e-05, "loss": 0.021, "step": 42950 }, { "grad_norm": 0.22932299971580505, "learning_rate": 6.55153097702494e-05, "loss": 0.025, "step": 42960 }, { "grad_norm": 0.22369436919689178, "learning_rate": 6.549959043187146e-05, "loss": 0.0212, "step": 42970 }, { "grad_norm": 0.2354477196931839, "learning_rate": 6.54838693984816e-05, "loss": 0.0215, "step": 42980 }, { "grad_norm": 0.2859239876270294, "learning_rate": 6.546814667179909e-05, "loss": 0.0257, "step": 42990 }, { "grad_norm": 0.2270577996969223, "learning_rate": 6.545242225354328e-05, "loss": 0.0207, "step": 43000 }, { "grad_norm": 0.27817147970199585, "learning_rate": 6.543669614543383e-05, "loss": 0.0189, "step": 43010 }, { "grad_norm": 0.23561322689056396, "learning_rate": 6.542096834919049e-05, "loss": 0.0232, "step": 43020 }, { "grad_norm": 0.19920112192630768, "learning_rate": 6.540523886653324e-05, "loss": 0.0206, "step": 43030 }, { "grad_norm": 0.18975505232810974, "learning_rate": 6.538950769918222e-05, "loss": 0.0217, "step": 43040 }, { "grad_norm": 0.21371160447597504, "learning_rate": 6.537377484885779e-05, "loss": 0.0234, "step": 43050 }, { "grad_norm": 0.22302615642547607, "learning_rate": 6.535804031728045e-05, "loss": 0.0191, "step": 43060 }, { "grad_norm": 0.24005474150180817, "learning_rate": 6.534230410617092e-05, "loss": 0.0191, "step": 43070 }, { "grad_norm": 0.21212001144886017, "learning_rate": 6.532656621725007e-05, "loss": 0.0229, "step": 43080 }, { "grad_norm": 0.21140995621681213, "learning_rate": 6.531082665223898e-05, "loss": 0.0205, "step": 43090 }, { "grad_norm": 0.24815307557582855, "learning_rate": 6.529508541285889e-05, "loss": 0.0206, "step": 43100 }, { "grad_norm": 0.20712675154209137, "learning_rate": 6.527934250083128e-05, "loss": 0.0196, "step": 43110 }, { "grad_norm": 0.19505901634693146, "learning_rate": 6.526359791787775e-05, "loss": 0.0199, "step": 43120 }, { "grad_norm": 0.20330052077770233, "learning_rate": 6.524785166572009e-05, "loss": 0.0221, "step": 43130 }, { "grad_norm": 0.26604193449020386, "learning_rate": 6.523210374608029e-05, "loss": 0.0172, "step": 43140 }, { "grad_norm": 0.20645588636398315, "learning_rate": 6.521635416068054e-05, "loss": 0.027, "step": 43150 }, { "grad_norm": 0.21924807131290436, "learning_rate": 6.520060291124317e-05, "loss": 0.0233, "step": 43160 }, { "grad_norm": 0.27683451771736145, "learning_rate": 6.518484999949072e-05, "loss": 0.0199, "step": 43170 }, { "grad_norm": 0.2845191955566406, "learning_rate": 6.516909542714591e-05, "loss": 0.0252, "step": 43180 }, { "grad_norm": 0.3895490765571594, "learning_rate": 6.515333919593161e-05, "loss": 0.0262, "step": 43190 }, { "grad_norm": 0.2789778709411621, "learning_rate": 6.513758130757094e-05, "loss": 0.0241, "step": 43200 }, { "grad_norm": 0.244865283370018, "learning_rate": 6.512182176378713e-05, "loss": 0.0226, "step": 43210 }, { "grad_norm": 0.2299611121416092, "learning_rate": 6.510606056630362e-05, "loss": 0.0225, "step": 43220 }, { "grad_norm": 0.20504125952720642, "learning_rate": 6.509029771684405e-05, "loss": 0.0217, "step": 43230 }, { "grad_norm": 0.20799259841442108, "learning_rate": 6.50745332171322e-05, "loss": 0.0253, "step": 43240 }, { "grad_norm": 0.19489622116088867, "learning_rate": 6.505876706889207e-05, "loss": 0.0231, "step": 43250 }, { "grad_norm": 0.24338188767433167, "learning_rate": 6.504299927384781e-05, "loss": 0.0227, "step": 43260 }, { "grad_norm": 0.19000419974327087, "learning_rate": 6.502722983372378e-05, "loss": 0.0221, "step": 43270 }, { "grad_norm": 0.20185859501361847, "learning_rate": 6.501145875024446e-05, "loss": 0.0188, "step": 43280 }, { "grad_norm": 0.21768353879451752, "learning_rate": 6.499568602513459e-05, "loss": 0.022, "step": 43290 }, { "grad_norm": 0.20874962210655212, "learning_rate": 6.497991166011903e-05, "loss": 0.019, "step": 43300 }, { "grad_norm": 0.23805636167526245, "learning_rate": 6.49641356569229e-05, "loss": 0.0216, "step": 43310 }, { "grad_norm": 0.2590212821960449, "learning_rate": 6.494835801727135e-05, "loss": 0.026, "step": 43320 }, { "grad_norm": 0.19464904069900513, "learning_rate": 6.493257874288986e-05, "loss": 0.022, "step": 43330 }, { "grad_norm": 0.23719348013401031, "learning_rate": 6.491679783550399e-05, "loss": 0.0222, "step": 43340 }, { "grad_norm": 0.26909032464027405, "learning_rate": 6.490101529683957e-05, "loss": 0.0248, "step": 43350 }, { "grad_norm": 0.22508400678634644, "learning_rate": 6.488523112862248e-05, "loss": 0.02, "step": 43360 }, { "grad_norm": 0.22256378829479218, "learning_rate": 6.486944533257891e-05, "loss": 0.0214, "step": 43370 }, { "grad_norm": 0.20186114311218262, "learning_rate": 6.485365791043517e-05, "loss": 0.0243, "step": 43380 }, { "grad_norm": 0.31101179122924805, "learning_rate": 6.483786886391772e-05, "loss": 0.0261, "step": 43390 }, { "grad_norm": 0.2353738397359848, "learning_rate": 6.482207819475323e-05, "loss": 0.0231, "step": 43400 }, { "grad_norm": 0.2363392561674118, "learning_rate": 6.480628590466857e-05, "loss": 0.0239, "step": 43410 }, { "grad_norm": 0.25734132528305054, "learning_rate": 6.479049199539074e-05, "loss": 0.024, "step": 43420 }, { "grad_norm": 0.34083250164985657, "learning_rate": 6.477469646864693e-05, "loss": 0.0248, "step": 43430 }, { "grad_norm": 0.16855230927467346, "learning_rate": 6.475889932616454e-05, "loss": 0.0272, "step": 43440 }, { "grad_norm": 0.20529603958129883, "learning_rate": 6.474310056967111e-05, "loss": 0.0224, "step": 43450 }, { "grad_norm": 0.2025841921567917, "learning_rate": 6.472730020089437e-05, "loss": 0.028, "step": 43460 }, { "grad_norm": 0.24803076684474945, "learning_rate": 6.471149822156222e-05, "loss": 0.0231, "step": 43470 }, { "grad_norm": 0.19283372163772583, "learning_rate": 6.469569463340275e-05, "loss": 0.0202, "step": 43480 }, { "grad_norm": 0.1418384611606598, "learning_rate": 6.467988943814418e-05, "loss": 0.023, "step": 43490 }, { "grad_norm": 0.28998252749443054, "learning_rate": 6.4664082637515e-05, "loss": 0.0239, "step": 43500 }, { "grad_norm": 0.22905303537845612, "learning_rate": 6.464827423324377e-05, "loss": 0.0222, "step": 43510 }, { "grad_norm": 0.2500750422477722, "learning_rate": 6.463246422705931e-05, "loss": 0.0257, "step": 43520 }, { "grad_norm": 0.24023699760437012, "learning_rate": 6.461665262069055e-05, "loss": 0.0276, "step": 43530 }, { "grad_norm": 0.24200373888015747, "learning_rate": 6.460083941586665e-05, "loss": 0.027, "step": 43540 }, { "grad_norm": 0.2417350858449936, "learning_rate": 6.45850246143169e-05, "loss": 0.0227, "step": 43550 }, { "grad_norm": 0.26713132858276367, "learning_rate": 6.456920821777077e-05, "loss": 0.0271, "step": 43560 }, { "grad_norm": 0.28871580958366394, "learning_rate": 6.455339022795796e-05, "loss": 0.0261, "step": 43570 }, { "grad_norm": 0.2944963872432709, "learning_rate": 6.453757064660825e-05, "loss": 0.0215, "step": 43580 }, { "grad_norm": 0.20810116827487946, "learning_rate": 6.452174947545169e-05, "loss": 0.0222, "step": 43590 }, { "grad_norm": 0.2256499081850052, "learning_rate": 6.450592671621842e-05, "loss": 0.0215, "step": 43600 }, { "grad_norm": 0.21496212482452393, "learning_rate": 6.449010237063882e-05, "loss": 0.0221, "step": 43610 }, { "grad_norm": 0.7378290891647339, "learning_rate": 6.447427644044343e-05, "loss": 0.0232, "step": 43620 }, { "grad_norm": 0.20618531107902527, "learning_rate": 6.44584489273629e-05, "loss": 0.0228, "step": 43630 }, { "grad_norm": 0.17109820246696472, "learning_rate": 6.444261983312817e-05, "loss": 0.0176, "step": 43640 }, { "grad_norm": 0.27974262833595276, "learning_rate": 6.442678915947023e-05, "loss": 0.0222, "step": 43650 }, { "grad_norm": 0.22269508242607117, "learning_rate": 6.441095690812032e-05, "loss": 0.0207, "step": 43660 }, { "grad_norm": 0.20912881195545197, "learning_rate": 6.439512308080982e-05, "loss": 0.0219, "step": 43670 }, { "grad_norm": 0.3004263639450073, "learning_rate": 6.43792876792703e-05, "loss": 0.0215, "step": 43680 }, { "grad_norm": 0.2323475182056427, "learning_rate": 6.43634507052335e-05, "loss": 0.024, "step": 43690 }, { "grad_norm": 0.19451653957366943, "learning_rate": 6.434761216043135e-05, "loss": 0.0214, "step": 43700 }, { "grad_norm": 0.21156799793243408, "learning_rate": 6.433177204659589e-05, "loss": 0.0233, "step": 43710 }, { "grad_norm": 0.23638074100017548, "learning_rate": 6.431593036545938e-05, "loss": 0.0202, "step": 43720 }, { "grad_norm": 0.16834169626235962, "learning_rate": 6.430008711875426e-05, "loss": 0.0217, "step": 43730 }, { "grad_norm": 0.2173835039138794, "learning_rate": 6.42842423082131e-05, "loss": 0.0236, "step": 43740 }, { "grad_norm": 0.19693127274513245, "learning_rate": 6.426839593556868e-05, "loss": 0.0201, "step": 43750 }, { "grad_norm": 0.2350001335144043, "learning_rate": 6.425254800255392e-05, "loss": 0.0203, "step": 43760 }, { "grad_norm": 0.19637620449066162, "learning_rate": 6.423669851090196e-05, "loss": 0.0199, "step": 43770 }, { "grad_norm": 0.20273205637931824, "learning_rate": 6.422084746234604e-05, "loss": 0.0233, "step": 43780 }, { "grad_norm": 0.2442035675048828, "learning_rate": 6.420499485861965e-05, "loss": 0.023, "step": 43790 }, { "grad_norm": 0.2491462528705597, "learning_rate": 6.418914070145636e-05, "loss": 0.0233, "step": 43800 }, { "grad_norm": 0.1859055459499359, "learning_rate": 6.417328499258996e-05, "loss": 0.0198, "step": 43810 }, { "grad_norm": 0.23876845836639404, "learning_rate": 6.415742773375446e-05, "loss": 0.0208, "step": 43820 }, { "grad_norm": 0.2796902656555176, "learning_rate": 6.414156892668393e-05, "loss": 0.0194, "step": 43830 }, { "grad_norm": 0.170126274228096, "learning_rate": 6.412570857311267e-05, "loss": 0.0195, "step": 43840 }, { "grad_norm": 0.23229025304317474, "learning_rate": 6.410984667477518e-05, "loss": 0.0215, "step": 43850 }, { "grad_norm": 0.21534138917922974, "learning_rate": 6.409398323340607e-05, "loss": 0.0216, "step": 43860 }, { "grad_norm": 0.21523064374923706, "learning_rate": 6.407811825074012e-05, "loss": 0.0245, "step": 43870 }, { "grad_norm": 0.21245962381362915, "learning_rate": 6.406225172851234e-05, "loss": 0.02, "step": 43880 }, { "grad_norm": 0.19405324757099152, "learning_rate": 6.404638366845786e-05, "loss": 0.018, "step": 43890 }, { "grad_norm": 0.21067413687705994, "learning_rate": 6.403051407231196e-05, "loss": 0.0176, "step": 43900 }, { "grad_norm": 0.20190368592739105, "learning_rate": 6.401464294181016e-05, "loss": 0.0199, "step": 43910 }, { "grad_norm": 0.18869489431381226, "learning_rate": 6.399877027868808e-05, "loss": 0.0227, "step": 43920 }, { "grad_norm": 0.1964644491672516, "learning_rate": 6.39828960846815e-05, "loss": 0.0245, "step": 43930 }, { "grad_norm": 0.16273345053195953, "learning_rate": 6.396702036152644e-05, "loss": 0.0211, "step": 43940 }, { "grad_norm": 0.21619249880313873, "learning_rate": 6.395114311095905e-05, "loss": 0.0239, "step": 43950 }, { "grad_norm": 0.2706255614757538, "learning_rate": 6.393526433471559e-05, "loss": 0.0252, "step": 43960 }, { "grad_norm": 0.26020851731300354, "learning_rate": 6.391938403453262e-05, "loss": 0.0222, "step": 43970 }, { "grad_norm": 0.2279399335384369, "learning_rate": 6.390350221214671e-05, "loss": 0.0217, "step": 43980 }, { "grad_norm": 0.2244960218667984, "learning_rate": 6.38876188692947e-05, "loss": 0.0189, "step": 43990 }, { "grad_norm": 0.21847590804100037, "learning_rate": 6.387173400771358e-05, "loss": 0.0196, "step": 44000 }, { "grad_norm": 0.26770150661468506, "learning_rate": 6.385584762914047e-05, "loss": 0.0205, "step": 44010 }, { "grad_norm": 0.19840405881404877, "learning_rate": 6.38399597353127e-05, "loss": 0.0188, "step": 44020 }, { "grad_norm": 0.23978592455387115, "learning_rate": 6.382407032796775e-05, "loss": 0.0213, "step": 44030 }, { "grad_norm": 0.22185273468494415, "learning_rate": 6.380817940884325e-05, "loss": 0.0211, "step": 44040 }, { "grad_norm": 0.2101445496082306, "learning_rate": 6.379228697967702e-05, "loss": 0.0203, "step": 44050 }, { "grad_norm": 0.2905212342739105, "learning_rate": 6.3776393042207e-05, "loss": 0.0194, "step": 44060 }, { "grad_norm": 0.2139645218849182, "learning_rate": 6.376049759817137e-05, "loss": 0.0218, "step": 44070 }, { "grad_norm": 0.19253621995449066, "learning_rate": 6.37446006493084e-05, "loss": 0.0189, "step": 44080 }, { "grad_norm": 0.26927515864372253, "learning_rate": 6.372870219735656e-05, "loss": 0.0208, "step": 44090 }, { "grad_norm": 0.24493388831615448, "learning_rate": 6.371280224405452e-05, "loss": 0.0202, "step": 44100 }, { "grad_norm": 0.26169657707214355, "learning_rate": 6.369690079114101e-05, "loss": 0.021, "step": 44110 }, { "grad_norm": 0.20403265953063965, "learning_rate": 6.368099784035504e-05, "loss": 0.0197, "step": 44120 }, { "grad_norm": 0.2304230034351349, "learning_rate": 6.366509339343572e-05, "loss": 0.0227, "step": 44130 }, { "grad_norm": 0.20578649640083313, "learning_rate": 6.364918745212232e-05, "loss": 0.0182, "step": 44140 }, { "grad_norm": 0.16543173789978027, "learning_rate": 6.363328001815431e-05, "loss": 0.0183, "step": 44150 }, { "grad_norm": 0.16957548260688782, "learning_rate": 6.361737109327128e-05, "loss": 0.0164, "step": 44160 }, { "grad_norm": 0.20637623965740204, "learning_rate": 6.360146067921301e-05, "loss": 0.019, "step": 44170 }, { "grad_norm": 0.3569530248641968, "learning_rate": 6.35855487777195e-05, "loss": 0.021, "step": 44180 }, { "grad_norm": 0.20266257226467133, "learning_rate": 6.356963539053076e-05, "loss": 0.0236, "step": 44190 }, { "grad_norm": 0.23963144421577454, "learning_rate": 6.355372051938711e-05, "loss": 0.0265, "step": 44200 }, { "grad_norm": 0.26134228706359863, "learning_rate": 6.353780416602894e-05, "loss": 0.0207, "step": 44210 }, { "grad_norm": 0.1949167102575302, "learning_rate": 6.352188633219689e-05, "loss": 0.0223, "step": 44220 }, { "grad_norm": 0.21511909365653992, "learning_rate": 6.350596701963166e-05, "loss": 0.0231, "step": 44230 }, { "grad_norm": 0.28178125619888306, "learning_rate": 6.349004623007419e-05, "loss": 0.0251, "step": 44240 }, { "grad_norm": 0.2442128211259842, "learning_rate": 6.347412396526555e-05, "loss": 0.0211, "step": 44250 }, { "grad_norm": 0.3017740845680237, "learning_rate": 6.345820022694696e-05, "loss": 0.0281, "step": 44260 }, { "grad_norm": 0.24132771790027618, "learning_rate": 6.344227501685984e-05, "loss": 0.0214, "step": 44270 }, { "grad_norm": 0.2235778570175171, "learning_rate": 6.342634833674572e-05, "loss": 0.0232, "step": 44280 }, { "grad_norm": 0.19806772470474243, "learning_rate": 6.341042018834635e-05, "loss": 0.0222, "step": 44290 }, { "grad_norm": 0.18154945969581604, "learning_rate": 6.339449057340359e-05, "loss": 0.02, "step": 44300 }, { "grad_norm": 0.1964799165725708, "learning_rate": 6.337855949365945e-05, "loss": 0.0194, "step": 44310 }, { "grad_norm": 0.21282140910625458, "learning_rate": 6.336262695085619e-05, "loss": 0.0216, "step": 44320 }, { "grad_norm": 0.2950844466686249, "learning_rate": 6.334669294673612e-05, "loss": 0.0222, "step": 44330 }, { "grad_norm": 0.1652977466583252, "learning_rate": 6.333075748304179e-05, "loss": 0.023, "step": 44340 }, { "grad_norm": 0.19313572347164154, "learning_rate": 6.331482056151585e-05, "loss": 0.0193, "step": 44350 }, { "grad_norm": 0.20206627249717712, "learning_rate": 6.329888218390117e-05, "loss": 0.0267, "step": 44360 }, { "grad_norm": 0.2150961458683014, "learning_rate": 6.328294235194072e-05, "loss": 0.0229, "step": 44370 }, { "grad_norm": 0.2440537065267563, "learning_rate": 6.326700106737765e-05, "loss": 0.0179, "step": 44380 }, { "grad_norm": 0.1826775223016739, "learning_rate": 6.325105833195531e-05, "loss": 0.0182, "step": 44390 }, { "grad_norm": 0.25578999519348145, "learning_rate": 6.323511414741715e-05, "loss": 0.0211, "step": 44400 }, { "grad_norm": 0.21364131569862366, "learning_rate": 6.321916851550678e-05, "loss": 0.0226, "step": 44410 }, { "grad_norm": 0.2547599673271179, "learning_rate": 6.320322143796806e-05, "loss": 0.0255, "step": 44420 }, { "grad_norm": 0.22947871685028076, "learning_rate": 6.318727291654488e-05, "loss": 0.0252, "step": 44430 }, { "grad_norm": 0.26378172636032104, "learning_rate": 6.317132295298134e-05, "loss": 0.0253, "step": 44440 }, { "grad_norm": 0.1903410702943802, "learning_rate": 6.315537154902173e-05, "loss": 0.022, "step": 44450 }, { "grad_norm": 0.20274320244789124, "learning_rate": 6.313941870641048e-05, "loss": 0.0219, "step": 44460 }, { "grad_norm": 0.16966880857944489, "learning_rate": 6.312346442689214e-05, "loss": 0.0181, "step": 44470 }, { "grad_norm": 0.22195132076740265, "learning_rate": 6.310750871221147e-05, "loss": 0.0251, "step": 44480 }, { "grad_norm": 0.2346438467502594, "learning_rate": 6.309155156411335e-05, "loss": 0.0222, "step": 44490 }, { "grad_norm": 0.2544190287590027, "learning_rate": 6.307559298434284e-05, "loss": 0.0315, "step": 44500 }, { "grad_norm": 0.2265568971633911, "learning_rate": 6.305963297464512e-05, "loss": 0.028, "step": 44510 }, { "grad_norm": 0.19513185322284698, "learning_rate": 6.304367153676561e-05, "loss": 0.0211, "step": 44520 }, { "grad_norm": 0.18348227441310883, "learning_rate": 6.302770867244978e-05, "loss": 0.0228, "step": 44530 }, { "grad_norm": 0.31609007716178894, "learning_rate": 6.301174438344328e-05, "loss": 0.0188, "step": 44540 }, { "grad_norm": 0.23251619935035706, "learning_rate": 6.299577867149202e-05, "loss": 0.0224, "step": 44550 }, { "grad_norm": 0.1961061954498291, "learning_rate": 6.297981153834192e-05, "loss": 0.022, "step": 44560 }, { "grad_norm": 0.23376135528087616, "learning_rate": 6.296384298573916e-05, "loss": 0.022, "step": 44570 }, { "grad_norm": 0.19595859944820404, "learning_rate": 6.294787301543001e-05, "loss": 0.0217, "step": 44580 }, { "grad_norm": 0.2588387429714203, "learning_rate": 6.293190162916095e-05, "loss": 0.0173, "step": 44590 }, { "grad_norm": 0.18029367923736572, "learning_rate": 6.291592882867855e-05, "loss": 0.0212, "step": 44600 }, { "grad_norm": 0.21761928498744965, "learning_rate": 6.28999546157296e-05, "loss": 0.0235, "step": 44610 }, { "grad_norm": 0.14689737558364868, "learning_rate": 6.288397899206102e-05, "loss": 0.0203, "step": 44620 }, { "grad_norm": 0.25254538655281067, "learning_rate": 6.286800195941984e-05, "loss": 0.0199, "step": 44630 }, { "grad_norm": 0.18213453888893127, "learning_rate": 6.285202351955334e-05, "loss": 0.0187, "step": 44640 }, { "grad_norm": 0.21003952622413635, "learning_rate": 6.283604367420887e-05, "loss": 0.0248, "step": 44650 }, { "grad_norm": 0.2842353582382202, "learning_rate": 6.282006242513394e-05, "loss": 0.0259, "step": 44660 }, { "grad_norm": 0.23235984146595, "learning_rate": 6.280407977407628e-05, "loss": 0.0208, "step": 44670 }, { "grad_norm": 0.19211728870868683, "learning_rate": 6.27880957227837e-05, "loss": 0.0235, "step": 44680 }, { "grad_norm": 0.2312222272157669, "learning_rate": 6.27721102730042e-05, "loss": 0.0226, "step": 44690 }, { "grad_norm": 0.22532854974269867, "learning_rate": 6.275612342648592e-05, "loss": 0.0218, "step": 44700 }, { "grad_norm": 0.16225098073482513, "learning_rate": 6.274013518497716e-05, "loss": 0.0214, "step": 44710 }, { "grad_norm": 0.1982930600643158, "learning_rate": 6.272414555022636e-05, "loss": 0.0255, "step": 44720 }, { "grad_norm": 0.21537919342517853, "learning_rate": 6.270815452398215e-05, "loss": 0.0212, "step": 44730 }, { "grad_norm": 0.18749253451824188, "learning_rate": 6.269216210799326e-05, "loss": 0.0194, "step": 44740 }, { "grad_norm": 0.24543507397174835, "learning_rate": 6.26761683040086e-05, "loss": 0.0216, "step": 44750 }, { "grad_norm": 0.14682312309741974, "learning_rate": 6.266017311377723e-05, "loss": 0.0208, "step": 44760 }, { "grad_norm": 0.18049894273281097, "learning_rate": 6.264417653904839e-05, "loss": 0.0256, "step": 44770 }, { "grad_norm": 0.2500104010105133, "learning_rate": 6.262817858157139e-05, "loss": 0.0232, "step": 44780 }, { "grad_norm": 0.21503935754299164, "learning_rate": 6.261217924309576e-05, "loss": 0.0218, "step": 44790 }, { "grad_norm": 0.19077832996845245, "learning_rate": 6.259617852537118e-05, "loss": 0.0233, "step": 44800 }, { "grad_norm": 0.1945551037788391, "learning_rate": 6.258017643014747e-05, "loss": 0.0211, "step": 44810 }, { "grad_norm": 0.23602621257305145, "learning_rate": 6.256417295917456e-05, "loss": 0.0218, "step": 44820 }, { "grad_norm": 0.17946091294288635, "learning_rate": 6.254816811420258e-05, "loss": 0.0266, "step": 44830 }, { "grad_norm": 0.18294890224933624, "learning_rate": 6.253216189698183e-05, "loss": 0.02, "step": 44840 }, { "grad_norm": 0.16467450559139252, "learning_rate": 6.251615430926267e-05, "loss": 0.0187, "step": 44850 }, { "grad_norm": 0.23946040868759155, "learning_rate": 6.25001453527957e-05, "loss": 0.0259, "step": 44860 }, { "grad_norm": 0.19628122448921204, "learning_rate": 6.248413502933164e-05, "loss": 0.0245, "step": 44870 }, { "grad_norm": 0.20069514214992523, "learning_rate": 6.246812334062133e-05, "loss": 0.0212, "step": 44880 }, { "grad_norm": 0.20464470982551575, "learning_rate": 6.245211028841579e-05, "loss": 0.0206, "step": 44890 }, { "grad_norm": 0.17487989366054535, "learning_rate": 6.24360958744662e-05, "loss": 0.0192, "step": 44900 }, { "grad_norm": 0.18467597663402557, "learning_rate": 6.242008010052387e-05, "loss": 0.0189, "step": 44910 }, { "grad_norm": 0.26913249492645264, "learning_rate": 6.240406296834024e-05, "loss": 0.0234, "step": 44920 }, { "grad_norm": 0.32138383388519287, "learning_rate": 6.238804447966694e-05, "loss": 0.021, "step": 44930 }, { "grad_norm": 0.22866177558898926, "learning_rate": 6.237202463625573e-05, "loss": 0.0194, "step": 44940 }, { "grad_norm": 0.18813712894916534, "learning_rate": 6.235600343985848e-05, "loss": 0.0201, "step": 44950 }, { "grad_norm": 0.21413469314575195, "learning_rate": 6.233998089222729e-05, "loss": 0.0221, "step": 44960 }, { "grad_norm": 0.1699851006269455, "learning_rate": 6.232395699511433e-05, "loss": 0.0215, "step": 44970 }, { "grad_norm": 0.24094924330711365, "learning_rate": 6.230793175027199e-05, "loss": 0.0229, "step": 44980 }, { "grad_norm": 0.23138633370399475, "learning_rate": 6.22919051594527e-05, "loss": 0.0229, "step": 44990 }, { "grad_norm": 0.3647454082965851, "learning_rate": 6.227587722440917e-05, "loss": 0.0203, "step": 45000 }, { "grad_norm": 0.27374884486198425, "learning_rate": 6.225984794689414e-05, "loss": 0.0227, "step": 45010 }, { "grad_norm": 0.22151732444763184, "learning_rate": 6.22438173286606e-05, "loss": 0.0215, "step": 45020 }, { "grad_norm": 0.20110636949539185, "learning_rate": 6.222778537146159e-05, "loss": 0.022, "step": 45030 }, { "grad_norm": 0.21035660803318024, "learning_rate": 6.221175207705037e-05, "loss": 0.0182, "step": 45040 }, { "grad_norm": 0.1665031909942627, "learning_rate": 6.219571744718027e-05, "loss": 0.02, "step": 45050 }, { "grad_norm": 0.21554984152317047, "learning_rate": 6.217968148360487e-05, "loss": 0.0215, "step": 45060 }, { "grad_norm": 0.24106915295124054, "learning_rate": 6.216364418807782e-05, "loss": 0.0228, "step": 45070 }, { "grad_norm": 0.21190622448921204, "learning_rate": 6.214760556235292e-05, "loss": 0.0182, "step": 45080 }, { "grad_norm": 0.18411582708358765, "learning_rate": 6.213156560818414e-05, "loss": 0.0213, "step": 45090 }, { "grad_norm": 0.2386268526315689, "learning_rate": 6.211552432732559e-05, "loss": 0.02, "step": 45100 }, { "grad_norm": 0.2281533181667328, "learning_rate": 6.20994817215315e-05, "loss": 0.0208, "step": 45110 }, { "grad_norm": 0.17502740025520325, "learning_rate": 6.208343779255628e-05, "loss": 0.02, "step": 45120 }, { "grad_norm": 0.2024284154176712, "learning_rate": 6.206739254215449e-05, "loss": 0.0213, "step": 45130 }, { "grad_norm": 0.2598832845687866, "learning_rate": 6.205134597208077e-05, "loss": 0.0204, "step": 45140 }, { "grad_norm": 0.2297862023115158, "learning_rate": 6.203529808409e-05, "loss": 0.0198, "step": 45150 }, { "grad_norm": 0.22818219661712646, "learning_rate": 6.201924887993712e-05, "loss": 0.0226, "step": 45160 }, { "grad_norm": 0.2042119801044464, "learning_rate": 6.200319836137725e-05, "loss": 0.0177, "step": 45170 }, { "grad_norm": 0.23268872499465942, "learning_rate": 6.198714653016565e-05, "loss": 0.0221, "step": 45180 }, { "grad_norm": 0.2048761397600174, "learning_rate": 6.197109338805774e-05, "loss": 0.0189, "step": 45190 }, { "grad_norm": 0.24358387291431427, "learning_rate": 6.195503893680903e-05, "loss": 0.0215, "step": 45200 }, { "grad_norm": 0.2971712052822113, "learning_rate": 6.193898317817524e-05, "loss": 0.0205, "step": 45210 }, { "grad_norm": 0.2206728458404541, "learning_rate": 6.192292611391221e-05, "loss": 0.0216, "step": 45220 }, { "grad_norm": 0.2171238511800766, "learning_rate": 6.190686774577591e-05, "loss": 0.0186, "step": 45230 }, { "grad_norm": 0.20307213068008423, "learning_rate": 6.189080807552245e-05, "loss": 0.0194, "step": 45240 }, { "grad_norm": 0.20438745617866516, "learning_rate": 6.187474710490809e-05, "loss": 0.0182, "step": 45250 }, { "grad_norm": 0.21534714102745056, "learning_rate": 6.185868483568926e-05, "loss": 0.0213, "step": 45260 }, { "grad_norm": 0.1729799211025238, "learning_rate": 6.184262126962245e-05, "loss": 0.0209, "step": 45270 }, { "grad_norm": 0.30931952595710754, "learning_rate": 6.182655640846442e-05, "loss": 0.0226, "step": 45280 }, { "grad_norm": 0.23449517786502838, "learning_rate": 6.181049025397196e-05, "loss": 0.0224, "step": 45290 }, { "grad_norm": 0.2052716165781021, "learning_rate": 6.179442280790202e-05, "loss": 0.021, "step": 45300 }, { "grad_norm": 0.19917038083076477, "learning_rate": 6.177835407201174e-05, "loss": 0.0199, "step": 45310 }, { "grad_norm": 0.17125272750854492, "learning_rate": 6.176228404805839e-05, "loss": 0.0189, "step": 45320 }, { "grad_norm": 0.23551692068576813, "learning_rate": 6.174621273779932e-05, "loss": 0.0255, "step": 45330 }, { "grad_norm": 0.1806517243385315, "learning_rate": 6.17301401429921e-05, "loss": 0.025, "step": 45340 }, { "grad_norm": 0.18327513337135315, "learning_rate": 6.171406626539436e-05, "loss": 0.017, "step": 45350 }, { "grad_norm": 0.22569027543067932, "learning_rate": 6.169799110676398e-05, "loss": 0.0236, "step": 45360 }, { "grad_norm": 0.15722650289535522, "learning_rate": 6.168191466885885e-05, "loss": 0.0219, "step": 45370 }, { "grad_norm": 0.2522951066493988, "learning_rate": 6.166583695343711e-05, "loss": 0.0199, "step": 45380 }, { "grad_norm": 0.19365699589252472, "learning_rate": 6.164975796225698e-05, "loss": 0.0204, "step": 45390 }, { "grad_norm": 0.235010027885437, "learning_rate": 6.163367769707683e-05, "loss": 0.0183, "step": 45400 }, { "grad_norm": 0.22017380595207214, "learning_rate": 6.161759615965519e-05, "loss": 0.0162, "step": 45410 }, { "grad_norm": 0.19309423863887787, "learning_rate": 6.16015133517507e-05, "loss": 0.0241, "step": 45420 }, { "grad_norm": 0.21348963677883148, "learning_rate": 6.158542927512214e-05, "loss": 0.0235, "step": 45430 }, { "grad_norm": 0.1742590218782425, "learning_rate": 6.156934393152846e-05, "loss": 0.0182, "step": 45440 }, { "grad_norm": 0.18566973507404327, "learning_rate": 6.15532573227287e-05, "loss": 0.0204, "step": 45450 }, { "grad_norm": 0.20703627169132233, "learning_rate": 6.153716945048212e-05, "loss": 0.0186, "step": 45460 }, { "grad_norm": 0.16674377024173737, "learning_rate": 6.152108031654802e-05, "loss": 0.0197, "step": 45470 }, { "grad_norm": 0.18460287153720856, "learning_rate": 6.15049899226859e-05, "loss": 0.0208, "step": 45480 }, { "grad_norm": 0.2054576277732849, "learning_rate": 6.148889827065537e-05, "loss": 0.0231, "step": 45490 }, { "grad_norm": 0.20931556820869446, "learning_rate": 6.147280536221622e-05, "loss": 0.0181, "step": 45500 }, { "grad_norm": 0.18942902982234955, "learning_rate": 6.145671119912832e-05, "loss": 0.0183, "step": 45510 }, { "grad_norm": 0.19011923670768738, "learning_rate": 6.144061578315169e-05, "loss": 0.019, "step": 45520 }, { "grad_norm": 0.2370053082704544, "learning_rate": 6.142451911604654e-05, "loss": 0.0241, "step": 45530 }, { "grad_norm": 0.21202349662780762, "learning_rate": 6.140842119957315e-05, "loss": 0.0193, "step": 45540 }, { "grad_norm": 0.1733171045780182, "learning_rate": 6.139232203549197e-05, "loss": 0.0198, "step": 45550 }, { "grad_norm": 0.21956360340118408, "learning_rate": 6.137622162556357e-05, "loss": 0.0195, "step": 45560 }, { "grad_norm": 0.2145630419254303, "learning_rate": 6.13601199715487e-05, "loss": 0.0218, "step": 45570 }, { "grad_norm": 0.21595460176467896, "learning_rate": 6.134401707520816e-05, "loss": 0.0196, "step": 45580 }, { "grad_norm": 0.22556072473526, "learning_rate": 6.132791293830298e-05, "loss": 0.0197, "step": 45590 }, { "grad_norm": 0.24508978426456451, "learning_rate": 6.131180756259428e-05, "loss": 0.0206, "step": 45600 }, { "grad_norm": 0.23566757142543793, "learning_rate": 6.129570094984331e-05, "loss": 0.0232, "step": 45610 }, { "grad_norm": 0.18170256912708282, "learning_rate": 6.127959310181145e-05, "loss": 0.0234, "step": 45620 }, { "grad_norm": 0.20926254987716675, "learning_rate": 6.126348402026026e-05, "loss": 0.0177, "step": 45630 }, { "grad_norm": 0.1962490826845169, "learning_rate": 6.12473737069514e-05, "loss": 0.0174, "step": 45640 }, { "grad_norm": 0.15308073163032532, "learning_rate": 6.123126216364665e-05, "loss": 0.0185, "step": 45650 }, { "grad_norm": 0.20857757329940796, "learning_rate": 6.121514939210797e-05, "loss": 0.0245, "step": 45660 }, { "grad_norm": 0.2515370845794678, "learning_rate": 6.119903539409741e-05, "loss": 0.0214, "step": 45670 }, { "grad_norm": 0.2344515025615692, "learning_rate": 6.118292017137716e-05, "loss": 0.021, "step": 45680 }, { "grad_norm": 0.20783166587352753, "learning_rate": 6.116680372570959e-05, "loss": 0.0224, "step": 45690 }, { "grad_norm": 0.28554651141166687, "learning_rate": 6.115068605885713e-05, "loss": 0.0238, "step": 45700 }, { "grad_norm": 0.25155192613601685, "learning_rate": 6.113456717258243e-05, "loss": 0.0227, "step": 45710 }, { "grad_norm": 0.20162688195705414, "learning_rate": 6.11184470686482e-05, "loss": 0.0229, "step": 45720 }, { "grad_norm": 0.14113934338092804, "learning_rate": 6.110232574881731e-05, "loss": 0.0194, "step": 45730 }, { "grad_norm": 0.3158957064151764, "learning_rate": 6.108620321485277e-05, "loss": 0.0239, "step": 45740 }, { "grad_norm": 0.16929170489311218, "learning_rate": 6.107007946851773e-05, "loss": 0.0201, "step": 45750 }, { "grad_norm": 0.14994682371616364, "learning_rate": 6.105395451157542e-05, "loss": 0.0199, "step": 45760 }, { "grad_norm": 0.20743583142757416, "learning_rate": 6.103782834578928e-05, "loss": 0.0223, "step": 45770 }, { "grad_norm": 0.20489849150180817, "learning_rate": 6.102170097292281e-05, "loss": 0.0183, "step": 45780 }, { "grad_norm": 0.19261308014392853, "learning_rate": 6.1005572394739716e-05, "loss": 0.0201, "step": 45790 }, { "grad_norm": 0.21800455451011658, "learning_rate": 6.0989442613003765e-05, "loss": 0.0216, "step": 45800 }, { "grad_norm": 0.21816447377204895, "learning_rate": 6.097331162947888e-05, "loss": 0.0224, "step": 45810 }, { "grad_norm": 0.1964154839515686, "learning_rate": 6.095717944592914e-05, "loss": 0.0206, "step": 45820 }, { "grad_norm": 0.1585221141576767, "learning_rate": 6.094104606411873e-05, "loss": 0.0198, "step": 45830 }, { "grad_norm": 0.1922139972448349, "learning_rate": 6.0924911485811966e-05, "loss": 0.0209, "step": 45840 }, { "grad_norm": 0.16068696975708008, "learning_rate": 6.090877571277331e-05, "loss": 0.0204, "step": 45850 }, { "grad_norm": 0.17548112571239471, "learning_rate": 6.089263874676736e-05, "loss": 0.0223, "step": 45860 }, { "grad_norm": 0.3631426990032196, "learning_rate": 6.0876500589558796e-05, "loss": 0.021, "step": 45870 }, { "grad_norm": 0.24186742305755615, "learning_rate": 6.086036124291248e-05, "loss": 0.0189, "step": 45880 }, { "grad_norm": 0.18584434688091278, "learning_rate": 6.084422070859339e-05, "loss": 0.0224, "step": 45890 }, { "grad_norm": 0.1941470503807068, "learning_rate": 6.082807898836663e-05, "loss": 0.0229, "step": 45900 }, { "grad_norm": 0.22792674601078033, "learning_rate": 6.081193608399742e-05, "loss": 0.0246, "step": 45910 }, { "grad_norm": 0.2245008796453476, "learning_rate": 6.0795791997251164e-05, "loss": 0.0208, "step": 45920 }, { "grad_norm": 0.20400822162628174, "learning_rate": 6.0779646729893294e-05, "loss": 0.0208, "step": 45930 }, { "grad_norm": 0.193977952003479, "learning_rate": 6.0763500283689476e-05, "loss": 0.0222, "step": 45940 }, { "grad_norm": 0.18198637664318085, "learning_rate": 6.0747352660405455e-05, "loss": 0.0207, "step": 45950 }, { "grad_norm": 0.24136079847812653, "learning_rate": 6.073120386180709e-05, "loss": 0.0203, "step": 45960 }, { "grad_norm": 0.2424173802137375, "learning_rate": 6.0715053889660425e-05, "loss": 0.0206, "step": 45970 }, { "grad_norm": 0.18056970834732056, "learning_rate": 6.069890274573157e-05, "loss": 0.0178, "step": 45980 }, { "grad_norm": 0.17033995687961578, "learning_rate": 6.068275043178679e-05, "loss": 0.023, "step": 45990 }, { "grad_norm": 0.22763696312904358, "learning_rate": 6.066659694959248e-05, "loss": 0.0201, "step": 46000 }, { "grad_norm": 0.21844516694545746, "learning_rate": 6.0650442300915176e-05, "loss": 0.0221, "step": 46010 }, { "grad_norm": 0.23597729206085205, "learning_rate": 6.063428648752152e-05, "loss": 0.0208, "step": 46020 }, { "grad_norm": 0.20344771444797516, "learning_rate": 6.0618129511178266e-05, "loss": 0.0202, "step": 46030 }, { "grad_norm": 0.1953102946281433, "learning_rate": 6.060197137365234e-05, "loss": 0.0175, "step": 46040 }, { "grad_norm": 0.18140798807144165, "learning_rate": 6.058581207671077e-05, "loss": 0.0194, "step": 46050 }, { "grad_norm": 0.2107851505279541, "learning_rate": 6.056965162212072e-05, "loss": 0.0186, "step": 46060 }, { "grad_norm": 0.2190256267786026, "learning_rate": 6.055349001164943e-05, "loss": 0.0211, "step": 46070 }, { "grad_norm": 0.16875311732292175, "learning_rate": 6.0537327247064347e-05, "loss": 0.018, "step": 46080 }, { "grad_norm": 0.2720407545566559, "learning_rate": 6.0521163330133014e-05, "loss": 0.0189, "step": 46090 }, { "grad_norm": 0.1794832944869995, "learning_rate": 6.050499826262306e-05, "loss": 0.0194, "step": 46100 }, { "grad_norm": 0.23740144073963165, "learning_rate": 6.0488832046302294e-05, "loss": 0.019, "step": 46110 }, { "grad_norm": 0.4792855978012085, "learning_rate": 6.0472664682938626e-05, "loss": 0.0268, "step": 46120 }, { "grad_norm": 0.2802939713001251, "learning_rate": 6.045649617430009e-05, "loss": 0.0214, "step": 46130 }, { "grad_norm": 0.25719499588012695, "learning_rate": 6.0440326522154866e-05, "loss": 0.0217, "step": 46140 }, { "grad_norm": 0.186235710978508, "learning_rate": 6.0424155728271224e-05, "loss": 0.0189, "step": 46150 }, { "grad_norm": 0.1750495284795761, "learning_rate": 6.040798379441758e-05, "loss": 0.0219, "step": 46160 }, { "grad_norm": 0.27698516845703125, "learning_rate": 6.0391810722362485e-05, "loss": 0.0251, "step": 46170 }, { "grad_norm": 0.2337825894355774, "learning_rate": 6.037563651387458e-05, "loss": 0.0205, "step": 46180 }, { "grad_norm": 0.1658116579055786, "learning_rate": 6.0359461170722666e-05, "loss": 0.0206, "step": 46190 }, { "grad_norm": 0.1928115040063858, "learning_rate": 6.034328469467566e-05, "loss": 0.0231, "step": 46200 }, { "grad_norm": 0.19923949241638184, "learning_rate": 6.0327107087502596e-05, "loss": 0.0209, "step": 46210 }, { "grad_norm": 0.2456951141357422, "learning_rate": 6.031092835097262e-05, "loss": 0.0207, "step": 46220 }, { "grad_norm": 0.1931282877922058, "learning_rate": 6.0294748486855024e-05, "loss": 0.019, "step": 46230 }, { "grad_norm": 0.26399821043014526, "learning_rate": 6.0278567496919216e-05, "loss": 0.02, "step": 46240 }, { "grad_norm": 0.23951250314712524, "learning_rate": 6.026238538293472e-05, "loss": 0.0194, "step": 46250 }, { "grad_norm": 0.1920233964920044, "learning_rate": 6.024620214667118e-05, "loss": 0.0188, "step": 46260 }, { "grad_norm": 0.21574139595031738, "learning_rate": 6.0230017789898384e-05, "loss": 0.0189, "step": 46270 }, { "grad_norm": 0.22900334000587463, "learning_rate": 6.021383231438622e-05, "loss": 0.0211, "step": 46280 }, { "grad_norm": 0.1883600354194641, "learning_rate": 6.0197645721904704e-05, "loss": 0.0192, "step": 46290 }, { "grad_norm": 0.21581827104091644, "learning_rate": 6.0181458014224e-05, "loss": 0.0261, "step": 46300 }, { "grad_norm": 0.19246652722358704, "learning_rate": 6.016526919311435e-05, "loss": 0.0191, "step": 46310 }, { "grad_norm": 0.19826780259609222, "learning_rate": 6.014907926034613e-05, "loss": 0.0181, "step": 46320 }, { "grad_norm": 0.22910477221012115, "learning_rate": 6.0132888217689875e-05, "loss": 0.0222, "step": 46330 }, { "grad_norm": 0.1697472482919693, "learning_rate": 6.01166960669162e-05, "loss": 0.0181, "step": 46340 }, { "grad_norm": 0.17233537137508392, "learning_rate": 6.0100502809795845e-05, "loss": 0.018, "step": 46350 }, { "grad_norm": 0.2638901174068451, "learning_rate": 6.00843084480997e-05, "loss": 0.0209, "step": 46360 }, { "grad_norm": 0.2019544392824173, "learning_rate": 6.0068112983598736e-05, "loss": 0.0222, "step": 46370 }, { "grad_norm": 0.26979419589042664, "learning_rate": 6.005191641806407e-05, "loss": 0.0224, "step": 46380 }, { "grad_norm": 0.2362063080072403, "learning_rate": 6.003571875326694e-05, "loss": 0.0232, "step": 46390 }, { "grad_norm": 0.18997681140899658, "learning_rate": 6.00195199909787e-05, "loss": 0.0228, "step": 46400 }, { "grad_norm": 0.22825951874256134, "learning_rate": 6.000332013297082e-05, "loss": 0.0192, "step": 46410 }, { "grad_norm": 0.25316423177719116, "learning_rate": 5.998711918101487e-05, "loss": 0.0196, "step": 46420 }, { "grad_norm": 0.16273006796836853, "learning_rate": 5.997091713688261e-05, "loss": 0.0186, "step": 46430 }, { "grad_norm": 0.24893833696842194, "learning_rate": 5.9954714002345836e-05, "loss": 0.0154, "step": 46440 }, { "grad_norm": 0.1849585920572281, "learning_rate": 5.993850977917649e-05, "loss": 0.0201, "step": 46450 }, { "grad_norm": 0.22005347907543182, "learning_rate": 5.992230446914667e-05, "loss": 0.0189, "step": 46460 }, { "grad_norm": 0.16398446261882782, "learning_rate": 5.990609807402855e-05, "loss": 0.018, "step": 46470 }, { "grad_norm": 0.18943244218826294, "learning_rate": 5.988989059559443e-05, "loss": 0.0187, "step": 46480 }, { "grad_norm": 0.20514877140522003, "learning_rate": 5.987368203561675e-05, "loss": 0.0179, "step": 46490 }, { "grad_norm": 0.17900308966636658, "learning_rate": 5.9857472395868055e-05, "loss": 0.0167, "step": 46500 }, { "grad_norm": 0.23099420964717865, "learning_rate": 5.9841261678120983e-05, "loss": 0.022, "step": 46510 }, { "grad_norm": 0.27317777276039124, "learning_rate": 5.982504988414834e-05, "loss": 0.0207, "step": 46520 }, { "grad_norm": 0.2721494436264038, "learning_rate": 5.9808837015723015e-05, "loss": 0.021, "step": 46530 }, { "grad_norm": 0.20230872929096222, "learning_rate": 5.9792623074618016e-05, "loss": 0.0235, "step": 46540 }, { "grad_norm": 0.19798126816749573, "learning_rate": 5.977640806260648e-05, "loss": 0.0236, "step": 46550 }, { "grad_norm": 0.20830024778842926, "learning_rate": 5.976019198146166e-05, "loss": 0.0218, "step": 46560 }, { "grad_norm": 0.2535995841026306, "learning_rate": 5.974397483295692e-05, "loss": 0.0217, "step": 46570 }, { "grad_norm": 0.20984888076782227, "learning_rate": 5.972775661886575e-05, "loss": 0.0197, "step": 46580 }, { "grad_norm": 0.14115707576274872, "learning_rate": 5.971153734096173e-05, "loss": 0.019, "step": 46590 }, { "grad_norm": 0.21011517941951752, "learning_rate": 5.969531700101859e-05, "loss": 0.0202, "step": 46600 }, { "grad_norm": 0.1938926726579666, "learning_rate": 5.9679095600810155e-05, "loss": 0.0179, "step": 46610 }, { "grad_norm": 0.5040748119354248, "learning_rate": 5.9662873142110384e-05, "loss": 0.0243, "step": 46620 }, { "grad_norm": 0.33830246329307556, "learning_rate": 5.964664962669333e-05, "loss": 0.0204, "step": 46630 }, { "grad_norm": 0.18028737604618073, "learning_rate": 5.9630425056333186e-05, "loss": 0.0188, "step": 46640 }, { "grad_norm": 0.18219126760959625, "learning_rate": 5.961419943280422e-05, "loss": 0.0175, "step": 46650 }, { "grad_norm": 0.23182347416877747, "learning_rate": 5.959797275788087e-05, "loss": 0.0219, "step": 46660 }, { "grad_norm": 0.25535136461257935, "learning_rate": 5.958174503333765e-05, "loss": 0.0222, "step": 46670 }, { "grad_norm": 0.20698733627796173, "learning_rate": 5.9565516260949195e-05, "loss": 0.0201, "step": 46680 }, { "grad_norm": 0.24976103007793427, "learning_rate": 5.9549286442490273e-05, "loss": 0.0215, "step": 46690 }, { "grad_norm": 0.2125736027956009, "learning_rate": 5.953305557973572e-05, "loss": 0.0212, "step": 46700 }, { "grad_norm": 0.5070973634719849, "learning_rate": 5.9516823674460564e-05, "loss": 0.025, "step": 46710 }, { "grad_norm": 0.16037599742412567, "learning_rate": 5.9500590728439875e-05, "loss": 0.0236, "step": 46720 }, { "grad_norm": 0.23520582914352417, "learning_rate": 5.948435674344886e-05, "loss": 0.0225, "step": 46730 }, { "grad_norm": 0.2939552664756775, "learning_rate": 5.946812172126285e-05, "loss": 0.0267, "step": 46740 }, { "grad_norm": 0.2618950605392456, "learning_rate": 5.945188566365729e-05, "loss": 0.0248, "step": 46750 }, { "grad_norm": 0.26621338725090027, "learning_rate": 5.943564857240773e-05, "loss": 0.0251, "step": 46760 }, { "grad_norm": 0.24847528338432312, "learning_rate": 5.941941044928981e-05, "loss": 0.0234, "step": 46770 }, { "grad_norm": 0.24909155070781708, "learning_rate": 5.940317129607935e-05, "loss": 0.02, "step": 46780 }, { "grad_norm": 0.1733478605747223, "learning_rate": 5.9386931114552204e-05, "loss": 0.0186, "step": 46790 }, { "grad_norm": 0.31479671597480774, "learning_rate": 5.937068990648438e-05, "loss": 0.0222, "step": 46800 }, { "grad_norm": 0.24462655186653137, "learning_rate": 5.935444767365199e-05, "loss": 0.023, "step": 46810 }, { "grad_norm": 0.2405378818511963, "learning_rate": 5.933820441783129e-05, "loss": 0.0219, "step": 46820 }, { "grad_norm": 0.20606569945812225, "learning_rate": 5.932196014079857e-05, "loss": 0.0216, "step": 46830 }, { "grad_norm": 0.1576017439365387, "learning_rate": 5.930571484433032e-05, "loss": 0.0213, "step": 46840 }, { "grad_norm": 0.22253024578094482, "learning_rate": 5.928946853020309e-05, "loss": 0.0222, "step": 46850 }, { "grad_norm": 0.25647205114364624, "learning_rate": 5.927322120019352e-05, "loss": 0.0214, "step": 46860 }, { "grad_norm": 0.16756801307201385, "learning_rate": 5.925697285607844e-05, "loss": 0.0193, "step": 46870 }, { "grad_norm": 0.2075852006673813, "learning_rate": 5.924072349963472e-05, "loss": 0.0213, "step": 46880 }, { "grad_norm": 0.19244204461574554, "learning_rate": 5.922447313263938e-05, "loss": 0.0267, "step": 46890 }, { "grad_norm": 0.20575886964797974, "learning_rate": 5.920822175686952e-05, "loss": 0.0188, "step": 46900 }, { "grad_norm": 0.1743752658367157, "learning_rate": 5.919196937410239e-05, "loss": 0.0193, "step": 46910 }, { "grad_norm": 0.18908926844596863, "learning_rate": 5.917571598611529e-05, "loss": 0.0217, "step": 46920 }, { "grad_norm": 0.24368131160736084, "learning_rate": 5.9159461594685686e-05, "loss": 0.0223, "step": 46930 }, { "grad_norm": 0.16928863525390625, "learning_rate": 5.914320620159114e-05, "loss": 0.0205, "step": 46940 }, { "grad_norm": 0.21092589199543, "learning_rate": 5.912694980860931e-05, "loss": 0.0206, "step": 46950 }, { "grad_norm": 0.22300530970096588, "learning_rate": 5.9110692417517964e-05, "loss": 0.0172, "step": 46960 }, { "grad_norm": 0.2353089153766632, "learning_rate": 5.9094434030095e-05, "loss": 0.0175, "step": 46970 }, { "grad_norm": 0.22682172060012817, "learning_rate": 5.907817464811841e-05, "loss": 0.0175, "step": 46980 }, { "grad_norm": 0.24663925170898438, "learning_rate": 5.906191427336627e-05, "loss": 0.0214, "step": 46990 }, { "grad_norm": 0.3106500208377838, "learning_rate": 5.9045652907616835e-05, "loss": 0.0204, "step": 47000 }, { "grad_norm": 0.19848176836967468, "learning_rate": 5.902939055264838e-05, "loss": 0.0248, "step": 47010 }, { "grad_norm": 0.20405720174312592, "learning_rate": 5.901312721023935e-05, "loss": 0.0222, "step": 47020 }, { "grad_norm": 0.19911842048168182, "learning_rate": 5.8996862882168294e-05, "loss": 0.0215, "step": 47030 }, { "grad_norm": 0.3012593686580658, "learning_rate": 5.8980597570213826e-05, "loss": 0.0219, "step": 47040 }, { "grad_norm": 0.17240580916404724, "learning_rate": 5.896433127615471e-05, "loss": 0.0212, "step": 47050 }, { "grad_norm": 0.21166758239269257, "learning_rate": 5.894806400176981e-05, "loss": 0.025, "step": 47060 }, { "grad_norm": 0.18442249298095703, "learning_rate": 5.893179574883808e-05, "loss": 0.0223, "step": 47070 }, { "grad_norm": 0.23143413662910461, "learning_rate": 5.8915526519138585e-05, "loss": 0.0232, "step": 47080 }, { "grad_norm": 0.2902586758136749, "learning_rate": 5.889925631445053e-05, "loss": 0.0198, "step": 47090 }, { "grad_norm": 0.24615277349948883, "learning_rate": 5.888298513655318e-05, "loss": 0.0214, "step": 47100 }, { "grad_norm": 0.2395818680524826, "learning_rate": 5.886671298722592e-05, "loss": 0.0196, "step": 47110 }, { "grad_norm": 0.18120428919792175, "learning_rate": 5.885043986824828e-05, "loss": 0.0197, "step": 47120 }, { "grad_norm": 0.2075040638446808, "learning_rate": 5.883416578139982e-05, "loss": 0.0192, "step": 47130 }, { "grad_norm": 0.23065046966075897, "learning_rate": 5.881789072846029e-05, "loss": 0.0204, "step": 47140 }, { "grad_norm": 0.21756911277770996, "learning_rate": 5.8801614711209474e-05, "loss": 0.0205, "step": 47150 }, { "grad_norm": 0.2515173554420471, "learning_rate": 5.878533773142732e-05, "loss": 0.0252, "step": 47160 }, { "grad_norm": 0.20500534772872925, "learning_rate": 5.8769059790893834e-05, "loss": 0.0196, "step": 47170 }, { "grad_norm": 0.1873360425233841, "learning_rate": 5.8752780891389145e-05, "loss": 0.0205, "step": 47180 }, { "grad_norm": 0.25603771209716797, "learning_rate": 5.873650103469351e-05, "loss": 0.0205, "step": 47190 }, { "grad_norm": 0.22113604843616486, "learning_rate": 5.872022022258725e-05, "loss": 0.0173, "step": 47200 }, { "grad_norm": 0.19511152803897858, "learning_rate": 5.8703938456850814e-05, "loss": 0.02, "step": 47210 }, { "grad_norm": 0.2238391935825348, "learning_rate": 5.868765573926477e-05, "loss": 0.0209, "step": 47220 }, { "grad_norm": 0.18058447539806366, "learning_rate": 5.867137207160974e-05, "loss": 0.0188, "step": 47230 }, { "grad_norm": 0.14805105328559875, "learning_rate": 5.86550874556665e-05, "loss": 0.0175, "step": 47240 }, { "grad_norm": 0.18190668523311615, "learning_rate": 5.863880189321592e-05, "loss": 0.0192, "step": 47250 }, { "grad_norm": 0.3990049660205841, "learning_rate": 5.862251538603894e-05, "loss": 0.0219, "step": 47260 }, { "grad_norm": 0.2545793056488037, "learning_rate": 5.8606227935916656e-05, "loss": 0.0208, "step": 47270 }, { "grad_norm": 0.22417056560516357, "learning_rate": 5.858993954463021e-05, "loss": 0.0167, "step": 47280 }, { "grad_norm": 0.2360944300889969, "learning_rate": 5.8573650213960896e-05, "loss": 0.0192, "step": 47290 }, { "grad_norm": 0.16020192205905914, "learning_rate": 5.8557359945690084e-05, "loss": 0.0198, "step": 47300 }, { "grad_norm": 0.2204781025648117, "learning_rate": 5.8541068741599246e-05, "loss": 0.0211, "step": 47310 }, { "grad_norm": 0.2485978752374649, "learning_rate": 5.8524776603469985e-05, "loss": 0.0197, "step": 47320 }, { "grad_norm": 0.20537984371185303, "learning_rate": 5.850848353308397e-05, "loss": 0.018, "step": 47330 }, { "grad_norm": 0.22784413397312164, "learning_rate": 5.849218953222297e-05, "loss": 0.0195, "step": 47340 }, { "grad_norm": 0.13053812086582184, "learning_rate": 5.847589460266891e-05, "loss": 0.0165, "step": 47350 }, { "grad_norm": 0.18764884769916534, "learning_rate": 5.845959874620375e-05, "loss": 0.0165, "step": 47360 }, { "grad_norm": 0.152369424700737, "learning_rate": 5.84433019646096e-05, "loss": 0.0172, "step": 47370 }, { "grad_norm": 0.21052229404449463, "learning_rate": 5.842700425966863e-05, "loss": 0.019, "step": 47380 }, { "grad_norm": 0.13962920010089874, "learning_rate": 5.841070563316315e-05, "loss": 0.0169, "step": 47390 }, { "grad_norm": 0.21747815608978271, "learning_rate": 5.839440608687554e-05, "loss": 0.0192, "step": 47400 }, { "grad_norm": 0.24254238605499268, "learning_rate": 5.837810562258831e-05, "loss": 0.0212, "step": 47410 }, { "grad_norm": 0.18382897973060608, "learning_rate": 5.836180424208405e-05, "loss": 0.0189, "step": 47420 }, { "grad_norm": 0.19662488996982574, "learning_rate": 5.834550194714543e-05, "loss": 0.0226, "step": 47430 }, { "grad_norm": 0.19814275205135345, "learning_rate": 5.832919873955528e-05, "loss": 0.0203, "step": 47440 }, { "grad_norm": 0.27385175228118896, "learning_rate": 5.831289462109648e-05, "loss": 0.0181, "step": 47450 }, { "grad_norm": 0.17098385095596313, "learning_rate": 5.829658959355202e-05, "loss": 0.0189, "step": 47460 }, { "grad_norm": 0.2332831472158432, "learning_rate": 5.8280283658704995e-05, "loss": 0.0205, "step": 47470 }, { "grad_norm": 0.27841684222221375, "learning_rate": 5.826397681833859e-05, "loss": 0.0233, "step": 47480 }, { "grad_norm": 0.16358980536460876, "learning_rate": 5.824766907423612e-05, "loss": 0.0185, "step": 47490 }, { "grad_norm": 0.21067968010902405, "learning_rate": 5.8231360428180945e-05, "loss": 0.0179, "step": 47500 }, { "grad_norm": 0.1989142894744873, "learning_rate": 5.821505088195658e-05, "loss": 0.0197, "step": 47510 }, { "grad_norm": 0.2276473343372345, "learning_rate": 5.819874043734661e-05, "loss": 0.0232, "step": 47520 }, { "grad_norm": 0.1727171391248703, "learning_rate": 5.8182429096134695e-05, "loss": 0.0188, "step": 47530 }, { "grad_norm": 0.23720085620880127, "learning_rate": 5.816611686010465e-05, "loss": 0.0187, "step": 47540 }, { "grad_norm": 0.2008066624403, "learning_rate": 5.814980373104033e-05, "loss": 0.0181, "step": 47550 }, { "grad_norm": 0.24431928992271423, "learning_rate": 5.813348971072572e-05, "loss": 0.0215, "step": 47560 }, { "grad_norm": 0.19309963285923004, "learning_rate": 5.811717480094492e-05, "loss": 0.0199, "step": 47570 }, { "grad_norm": 0.25356030464172363, "learning_rate": 5.810085900348209e-05, "loss": 0.0215, "step": 47580 }, { "grad_norm": 0.1919996589422226, "learning_rate": 5.8084542320121483e-05, "loss": 0.0224, "step": 47590 }, { "grad_norm": 0.24162907898426056, "learning_rate": 5.8068224752647497e-05, "loss": 0.0212, "step": 47600 }, { "grad_norm": 0.1741802841424942, "learning_rate": 5.805190630284457e-05, "loss": 0.0201, "step": 47610 }, { "grad_norm": 0.20303875207901, "learning_rate": 5.803558697249729e-05, "loss": 0.0211, "step": 47620 }, { "grad_norm": 0.19256576895713806, "learning_rate": 5.8019266763390276e-05, "loss": 0.0191, "step": 47630 }, { "grad_norm": 0.2522088289260864, "learning_rate": 5.800294567730833e-05, "loss": 0.0224, "step": 47640 }, { "grad_norm": 0.13865292072296143, "learning_rate": 5.798662371603628e-05, "loss": 0.0177, "step": 47650 }, { "grad_norm": 0.15390676259994507, "learning_rate": 5.797030088135904e-05, "loss": 0.0172, "step": 47660 }, { "grad_norm": 0.1574140191078186, "learning_rate": 5.7953977175061705e-05, "loss": 0.0172, "step": 47670 }, { "grad_norm": 0.2271893471479416, "learning_rate": 5.793765259892938e-05, "loss": 0.0247, "step": 47680 }, { "grad_norm": 0.23962940275669098, "learning_rate": 5.792132715474729e-05, "loss": 0.022, "step": 47690 }, { "grad_norm": 0.23514510691165924, "learning_rate": 5.790500084430078e-05, "loss": 0.0182, "step": 47700 }, { "grad_norm": 0.18030889332294464, "learning_rate": 5.7888673669375274e-05, "loss": 0.0188, "step": 47710 }, { "grad_norm": 0.31980258226394653, "learning_rate": 5.787234563175625e-05, "loss": 0.0202, "step": 47720 }, { "grad_norm": 0.14359350502490997, "learning_rate": 5.7856016733229355e-05, "loss": 0.0187, "step": 47730 }, { "grad_norm": 0.20120106637477875, "learning_rate": 5.7839686975580297e-05, "loss": 0.0201, "step": 47740 }, { "grad_norm": 0.20061108469963074, "learning_rate": 5.782335636059484e-05, "loss": 0.0189, "step": 47750 }, { "grad_norm": 0.2630338966846466, "learning_rate": 5.780702489005889e-05, "loss": 0.021, "step": 47760 }, { "grad_norm": 0.18471667170524597, "learning_rate": 5.779069256575845e-05, "loss": 0.0217, "step": 47770 }, { "grad_norm": 0.23552240431308746, "learning_rate": 5.7774359389479574e-05, "loss": 0.0224, "step": 47780 }, { "grad_norm": 0.25006648898124695, "learning_rate": 5.775802536300845e-05, "loss": 0.0226, "step": 47790 }, { "grad_norm": 0.1651162952184677, "learning_rate": 5.774169048813134e-05, "loss": 0.0233, "step": 47800 }, { "grad_norm": 0.23427967727184296, "learning_rate": 5.77253547666346e-05, "loss": 0.02, "step": 47810 }, { "grad_norm": 0.23769499361515045, "learning_rate": 5.770901820030465e-05, "loss": 0.0224, "step": 47820 }, { "grad_norm": 0.232705757021904, "learning_rate": 5.769268079092809e-05, "loss": 0.0163, "step": 47830 }, { "grad_norm": 0.18310290575027466, "learning_rate": 5.767634254029151e-05, "loss": 0.019, "step": 47840 }, { "grad_norm": 0.1850614994764328, "learning_rate": 5.7660003450181655e-05, "loss": 0.019, "step": 47850 }, { "grad_norm": 0.1896931380033493, "learning_rate": 5.764366352238534e-05, "loss": 0.0167, "step": 47860 }, { "grad_norm": 0.2826271057128906, "learning_rate": 5.7627322758689474e-05, "loss": 0.0209, "step": 47870 }, { "grad_norm": 0.2205754965543747, "learning_rate": 5.761098116088105e-05, "loss": 0.0192, "step": 47880 }, { "grad_norm": 0.20587590336799622, "learning_rate": 5.759463873074717e-05, "loss": 0.0213, "step": 47890 }, { "grad_norm": 0.19870340824127197, "learning_rate": 5.757829547007504e-05, "loss": 0.0217, "step": 47900 }, { "grad_norm": 0.1379522830247879, "learning_rate": 5.756195138065189e-05, "loss": 0.02, "step": 47910 }, { "grad_norm": 0.1781582236289978, "learning_rate": 5.754560646426511e-05, "loss": 0.0181, "step": 47920 }, { "grad_norm": 0.1947731226682663, "learning_rate": 5.752926072270216e-05, "loss": 0.0199, "step": 47930 }, { "grad_norm": 0.17776447534561157, "learning_rate": 5.7512914157750563e-05, "loss": 0.015, "step": 47940 }, { "grad_norm": 0.11743767559528351, "learning_rate": 5.749656677119798e-05, "loss": 0.0172, "step": 47950 }, { "grad_norm": 0.18807975947856903, "learning_rate": 5.7480218564832125e-05, "loss": 0.0187, "step": 47960 }, { "grad_norm": 0.1917911022901535, "learning_rate": 5.746386954044082e-05, "loss": 0.0225, "step": 47970 }, { "grad_norm": 0.20754791796207428, "learning_rate": 5.744751969981195e-05, "loss": 0.021, "step": 47980 }, { "grad_norm": 0.2112949639558792, "learning_rate": 5.7431169044733526e-05, "loss": 0.0197, "step": 47990 }, { "grad_norm": 0.24822449684143066, "learning_rate": 5.741481757699364e-05, "loss": 0.0224, "step": 48000 }, { "grad_norm": 0.2295006513595581, "learning_rate": 5.7398465298380434e-05, "loss": 0.0223, "step": 48010 }, { "grad_norm": 0.17916320264339447, "learning_rate": 5.7382112210682193e-05, "loss": 0.0154, "step": 48020 }, { "grad_norm": 0.2127985805273056, "learning_rate": 5.7365758315687266e-05, "loss": 0.021, "step": 48030 }, { "grad_norm": 0.2351708859205246, "learning_rate": 5.734940361518407e-05, "loss": 0.0209, "step": 48040 }, { "grad_norm": 0.1855027824640274, "learning_rate": 5.733304811096116e-05, "loss": 0.0195, "step": 48050 }, { "grad_norm": 0.23289939761161804, "learning_rate": 5.731669180480713e-05, "loss": 0.0177, "step": 48060 }, { "grad_norm": 0.26024335622787476, "learning_rate": 5.730033469851067e-05, "loss": 0.0182, "step": 48070 }, { "grad_norm": 0.24422864615917206, "learning_rate": 5.7283976793860607e-05, "loss": 0.0193, "step": 48080 }, { "grad_norm": 0.20918096601963043, "learning_rate": 5.726761809264577e-05, "loss": 0.0184, "step": 48090 }, { "grad_norm": 0.17953704297542572, "learning_rate": 5.7251258596655155e-05, "loss": 0.022, "step": 48100 }, { "grad_norm": 0.2596091330051422, "learning_rate": 5.72348983076778e-05, "loss": 0.0215, "step": 48110 }, { "grad_norm": 0.19315491616725922, "learning_rate": 5.7218537227502854e-05, "loss": 0.0252, "step": 48120 }, { "grad_norm": 0.23748469352722168, "learning_rate": 5.720217535791951e-05, "loss": 0.0204, "step": 48130 }, { "grad_norm": 0.21155837178230286, "learning_rate": 5.718581270071711e-05, "loss": 0.019, "step": 48140 }, { "grad_norm": 0.22520844638347626, "learning_rate": 5.716944925768505e-05, "loss": 0.0186, "step": 48150 }, { "grad_norm": 0.20119214057922363, "learning_rate": 5.7153085030612786e-05, "loss": 0.0193, "step": 48160 }, { "grad_norm": 0.24249513447284698, "learning_rate": 5.713672002128989e-05, "loss": 0.0231, "step": 48170 }, { "grad_norm": 0.212606742978096, "learning_rate": 5.712035423150602e-05, "loss": 0.0196, "step": 48180 }, { "grad_norm": 0.13851980865001678, "learning_rate": 5.710398766305094e-05, "loss": 0.0211, "step": 48190 }, { "grad_norm": 0.2012440413236618, "learning_rate": 5.708762031771442e-05, "loss": 0.0186, "step": 48200 }, { "grad_norm": 0.25801175832748413, "learning_rate": 5.707125219728642e-05, "loss": 0.0173, "step": 48210 }, { "grad_norm": 0.17352882027626038, "learning_rate": 5.7054883303556905e-05, "loss": 0.0203, "step": 48220 }, { "grad_norm": 0.2154560685157776, "learning_rate": 5.703851363831595e-05, "loss": 0.0176, "step": 48230 }, { "grad_norm": 0.17996440827846527, "learning_rate": 5.702214320335374e-05, "loss": 0.0243, "step": 48240 }, { "grad_norm": 0.17293782532215118, "learning_rate": 5.7005772000460514e-05, "loss": 0.0187, "step": 48250 }, { "grad_norm": 0.17898419499397278, "learning_rate": 5.698940003142656e-05, "loss": 0.0184, "step": 48260 }, { "grad_norm": 0.16265445947647095, "learning_rate": 5.697302729804236e-05, "loss": 0.0163, "step": 48270 }, { "grad_norm": 0.23998980224132538, "learning_rate": 5.695665380209837e-05, "loss": 0.0225, "step": 48280 }, { "grad_norm": 0.20938120782375336, "learning_rate": 5.6940279545385165e-05, "loss": 0.0186, "step": 48290 }, { "grad_norm": 0.16135022044181824, "learning_rate": 5.692390452969344e-05, "loss": 0.0191, "step": 48300 }, { "grad_norm": 0.21354907751083374, "learning_rate": 5.690752875681392e-05, "loss": 0.02, "step": 48310 }, { "grad_norm": 0.23008650541305542, "learning_rate": 5.6891152228537435e-05, "loss": 0.0197, "step": 48320 }, { "grad_norm": 0.25697439908981323, "learning_rate": 5.687477494665492e-05, "loss": 0.019, "step": 48330 }, { "grad_norm": 0.2805132567882538, "learning_rate": 5.685839691295733e-05, "loss": 0.0169, "step": 48340 }, { "grad_norm": 0.30507683753967285, "learning_rate": 5.6842018129235786e-05, "loss": 0.0219, "step": 48350 }, { "grad_norm": 0.17953647673130035, "learning_rate": 5.6825638597281404e-05, "loss": 0.0241, "step": 48360 }, { "grad_norm": 0.16648389399051666, "learning_rate": 5.680925831888546e-05, "loss": 0.0246, "step": 48370 }, { "grad_norm": 0.21485422551631927, "learning_rate": 5.6792877295839274e-05, "loss": 0.0176, "step": 48380 }, { "grad_norm": 0.16130119562149048, "learning_rate": 5.6776495529934224e-05, "loss": 0.02, "step": 48390 }, { "grad_norm": 0.205830380320549, "learning_rate": 5.6760113022961824e-05, "loss": 0.0176, "step": 48400 }, { "grad_norm": 0.18975621461868286, "learning_rate": 5.6743729776713617e-05, "loss": 0.0203, "step": 48410 }, { "grad_norm": 0.21899081766605377, "learning_rate": 5.672734579298126e-05, "loss": 0.0188, "step": 48420 }, { "grad_norm": 0.15241853892803192, "learning_rate": 5.671096107355649e-05, "loss": 0.018, "step": 48430 }, { "grad_norm": 0.20104584097862244, "learning_rate": 5.66945756202311e-05, "loss": 0.0265, "step": 48440 }, { "grad_norm": 0.19207677245140076, "learning_rate": 5.667818943479699e-05, "loss": 0.0202, "step": 48450 }, { "grad_norm": 0.17529310286045074, "learning_rate": 5.666180251904612e-05, "loss": 0.0199, "step": 48460 }, { "grad_norm": 0.1924264132976532, "learning_rate": 5.6645414874770555e-05, "loss": 0.0188, "step": 48470 }, { "grad_norm": 0.18242643773555756, "learning_rate": 5.66290265037624e-05, "loss": 0.0187, "step": 48480 }, { "grad_norm": 0.15101158618927002, "learning_rate": 5.661263740781386e-05, "loss": 0.0204, "step": 48490 }, { "grad_norm": 0.16418497264385223, "learning_rate": 5.6596247588717254e-05, "loss": 0.0208, "step": 48500 }, { "grad_norm": 0.18698984384536743, "learning_rate": 5.6579857048264926e-05, "loss": 0.0189, "step": 48510 }, { "grad_norm": 0.2020917683839798, "learning_rate": 5.6563465788249314e-05, "loss": 0.0188, "step": 48520 }, { "grad_norm": 0.22772501409053802, "learning_rate": 5.6547073810462956e-05, "loss": 0.0194, "step": 48530 }, { "grad_norm": 0.25147467851638794, "learning_rate": 5.653068111669846e-05, "loss": 0.0188, "step": 48540 }, { "grad_norm": 0.20780247449874878, "learning_rate": 5.651428770874848e-05, "loss": 0.018, "step": 48550 }, { "grad_norm": 0.2374182790517807, "learning_rate": 5.64978935884058e-05, "loss": 0.0176, "step": 48560 }, { "grad_norm": 0.17686550319194794, "learning_rate": 5.6481498757463244e-05, "loss": 0.0251, "step": 48570 }, { "grad_norm": 0.20464374125003815, "learning_rate": 5.646510321771373e-05, "loss": 0.0202, "step": 48580 }, { "grad_norm": 0.19939500093460083, "learning_rate": 5.644870697095024e-05, "loss": 0.0226, "step": 48590 }, { "grad_norm": 0.19647938013076782, "learning_rate": 5.643231001896586e-05, "loss": 0.0176, "step": 48600 }, { "grad_norm": 0.18613825738430023, "learning_rate": 5.6415912363553726e-05, "loss": 0.0178, "step": 48610 }, { "grad_norm": 0.16624195873737335, "learning_rate": 5.639951400650706e-05, "loss": 0.0189, "step": 48620 }, { "grad_norm": 0.1474734991788864, "learning_rate": 5.6383114949619165e-05, "loss": 0.0178, "step": 48630 }, { "grad_norm": 0.21778254210948944, "learning_rate": 5.636671519468342e-05, "loss": 0.0207, "step": 48640 }, { "grad_norm": 0.24109207093715668, "learning_rate": 5.635031474349327e-05, "loss": 0.0176, "step": 48650 }, { "grad_norm": 0.18262621760368347, "learning_rate": 5.6333913597842246e-05, "loss": 0.0189, "step": 48660 }, { "grad_norm": 0.1807336062192917, "learning_rate": 5.6317511759523955e-05, "loss": 0.0205, "step": 48670 }, { "grad_norm": 0.17159150540828705, "learning_rate": 5.630110923033207e-05, "loss": 0.0188, "step": 48680 }, { "grad_norm": 0.16279567778110504, "learning_rate": 5.628470601206036e-05, "loss": 0.0199, "step": 48690 }, { "grad_norm": 0.2137061059474945, "learning_rate": 5.626830210650266e-05, "loss": 0.0207, "step": 48700 }, { "grad_norm": 0.2313808649778366, "learning_rate": 5.625189751545285e-05, "loss": 0.0229, "step": 48710 }, { "grad_norm": 0.17095808684825897, "learning_rate": 5.6235492240704936e-05, "loss": 0.0203, "step": 48720 }, { "grad_norm": 0.16572768986225128, "learning_rate": 5.621908628405296e-05, "loss": 0.0192, "step": 48730 }, { "grad_norm": 0.2716091573238373, "learning_rate": 5.620267964729106e-05, "loss": 0.0275, "step": 48740 }, { "grad_norm": 0.2375939041376114, "learning_rate": 5.618627233221344e-05, "loss": 0.0185, "step": 48750 }, { "grad_norm": 0.2586715817451477, "learning_rate": 5.616986434061438e-05, "loss": 0.0195, "step": 48760 }, { "grad_norm": 0.31784912943840027, "learning_rate": 5.615345567428822e-05, "loss": 0.0196, "step": 48770 }, { "grad_norm": 0.1824372112751007, "learning_rate": 5.613704633502941e-05, "loss": 0.0182, "step": 48780 }, { "grad_norm": 0.20401670038700104, "learning_rate": 5.612063632463245e-05, "loss": 0.0224, "step": 48790 }, { "grad_norm": 0.24715439975261688, "learning_rate": 5.610422564489188e-05, "loss": 0.0188, "step": 48800 }, { "grad_norm": 0.21785539388656616, "learning_rate": 5.608781429760239e-05, "loss": 0.0213, "step": 48810 }, { "grad_norm": 0.21951550245285034, "learning_rate": 5.607140228455866e-05, "loss": 0.019, "step": 48820 }, { "grad_norm": 0.19534127414226532, "learning_rate": 5.605498960755553e-05, "loss": 0.0201, "step": 48830 }, { "grad_norm": 0.17475183308124542, "learning_rate": 5.603857626838782e-05, "loss": 0.0178, "step": 48840 }, { "grad_norm": 0.1933421939611435, "learning_rate": 5.60221622688505e-05, "loss": 0.0195, "step": 48850 }, { "grad_norm": 0.19634316861629486, "learning_rate": 5.6005747610738565e-05, "loss": 0.021, "step": 48860 }, { "grad_norm": 0.1917896717786789, "learning_rate": 5.59893322958471e-05, "loss": 0.0212, "step": 48870 }, { "grad_norm": 0.27095070481300354, "learning_rate": 5.5972916325971256e-05, "loss": 0.0232, "step": 48880 }, { "grad_norm": 0.17276066541671753, "learning_rate": 5.595649970290628e-05, "loss": 0.0167, "step": 48890 }, { "grad_norm": 0.2033999115228653, "learning_rate": 5.5940082428447426e-05, "loss": 0.0211, "step": 48900 }, { "grad_norm": 0.16608068346977234, "learning_rate": 5.592366450439012e-05, "loss": 0.0172, "step": 48910 }, { "grad_norm": 0.18286561965942383, "learning_rate": 5.590724593252975e-05, "loss": 0.0202, "step": 48920 }, { "grad_norm": 0.20334979891777039, "learning_rate": 5.589082671466184e-05, "loss": 0.0175, "step": 48930 }, { "grad_norm": 0.19467458128929138, "learning_rate": 5.587440685258199e-05, "loss": 0.0182, "step": 48940 }, { "grad_norm": 0.16140127182006836, "learning_rate": 5.585798634808583e-05, "loss": 0.0208, "step": 48950 }, { "grad_norm": 0.2469460815191269, "learning_rate": 5.584156520296909e-05, "loss": 0.0175, "step": 48960 }, { "grad_norm": 0.23516815900802612, "learning_rate": 5.582514341902757e-05, "loss": 0.0221, "step": 48970 }, { "grad_norm": 0.25455161929130554, "learning_rate": 5.580872099805713e-05, "loss": 0.0187, "step": 48980 }, { "grad_norm": 0.19623126089572906, "learning_rate": 5.5792297941853674e-05, "loss": 0.0195, "step": 48990 }, { "grad_norm": 0.18553070724010468, "learning_rate": 5.5775874252213247e-05, "loss": 0.0218, "step": 49000 }, { "grad_norm": 0.15991906821727753, "learning_rate": 5.575944993093189e-05, "loss": 0.0186, "step": 49010 }, { "grad_norm": 0.12698084115982056, "learning_rate": 5.574302497980574e-05, "loss": 0.0212, "step": 49020 }, { "grad_norm": 0.20478935539722443, "learning_rate": 5.5726599400631005e-05, "loss": 0.019, "step": 49030 }, { "grad_norm": 0.17465877532958984, "learning_rate": 5.571017319520399e-05, "loss": 0.0177, "step": 49040 }, { "grad_norm": 0.21690724790096283, "learning_rate": 5.5693746365320984e-05, "loss": 0.0191, "step": 49050 }, { "grad_norm": 0.19994370639324188, "learning_rate": 5.567731891277848e-05, "loss": 0.0181, "step": 49060 }, { "grad_norm": 0.22190117835998535, "learning_rate": 5.566089083937288e-05, "loss": 0.0207, "step": 49070 }, { "grad_norm": 0.23996062576770782, "learning_rate": 5.564446214690079e-05, "loss": 0.0225, "step": 49080 }, { "grad_norm": 0.24009424448013306, "learning_rate": 5.56280328371588e-05, "loss": 0.0221, "step": 49090 }, { "grad_norm": 0.22260348498821259, "learning_rate": 5.5611602911943596e-05, "loss": 0.0194, "step": 49100 }, { "grad_norm": 0.15678581595420837, "learning_rate": 5.559517237305194e-05, "loss": 0.0211, "step": 49110 }, { "grad_norm": 0.1931784600019455, "learning_rate": 5.557874122228064e-05, "loss": 0.0207, "step": 49120 }, { "grad_norm": 0.17316864430904388, "learning_rate": 5.55623094614266e-05, "loss": 0.021, "step": 49130 }, { "grad_norm": 0.2642395794391632, "learning_rate": 5.5545877092286743e-05, "loss": 0.0187, "step": 49140 }, { "grad_norm": 0.2080780267715454, "learning_rate": 5.552944411665811e-05, "loss": 0.0188, "step": 49150 }, { "grad_norm": 0.19085875153541565, "learning_rate": 5.5513010536337795e-05, "loss": 0.0163, "step": 49160 }, { "grad_norm": 0.19601596891880035, "learning_rate": 5.5496576353122934e-05, "loss": 0.0193, "step": 49170 }, { "grad_norm": 0.1787884384393692, "learning_rate": 5.548014156881074e-05, "loss": 0.0171, "step": 49180 }, { "grad_norm": 0.1854955106973648, "learning_rate": 5.5463706185198494e-05, "loss": 0.0258, "step": 49190 }, { "grad_norm": 0.23025088012218475, "learning_rate": 5.544727020408358e-05, "loss": 0.0177, "step": 49200 }, { "grad_norm": 0.1968647837638855, "learning_rate": 5.5430833627263366e-05, "loss": 0.0163, "step": 49210 }, { "grad_norm": 0.22929152846336365, "learning_rate": 5.5414396456535354e-05, "loss": 0.022, "step": 49220 }, { "grad_norm": 0.18890494108200073, "learning_rate": 5.539795869369709e-05, "loss": 0.0199, "step": 49230 }, { "grad_norm": 0.18234825134277344, "learning_rate": 5.538152034054618e-05, "loss": 0.0196, "step": 49240 }, { "grad_norm": 0.144931823015213, "learning_rate": 5.536508139888028e-05, "loss": 0.0165, "step": 49250 }, { "grad_norm": 0.2587563395500183, "learning_rate": 5.534864187049716e-05, "loss": 0.0197, "step": 49260 }, { "grad_norm": 0.20641553401947021, "learning_rate": 5.5332201757194615e-05, "loss": 0.0182, "step": 49270 }, { "grad_norm": 0.17737659811973572, "learning_rate": 5.531576106077049e-05, "loss": 0.0175, "step": 49280 }, { "grad_norm": 0.20132361352443695, "learning_rate": 5.529931978302272e-05, "loss": 0.0176, "step": 49290 }, { "grad_norm": 0.15603312849998474, "learning_rate": 5.528287792574932e-05, "loss": 0.0225, "step": 49300 }, { "grad_norm": 0.16670936346054077, "learning_rate": 5.5266435490748324e-05, "loss": 0.0215, "step": 49310 }, { "grad_norm": 0.25827690958976746, "learning_rate": 5.524999247981787e-05, "loss": 0.021, "step": 49320 }, { "grad_norm": 0.1858324408531189, "learning_rate": 5.523354889475613e-05, "loss": 0.0207, "step": 49330 }, { "grad_norm": 0.31264469027519226, "learning_rate": 5.521710473736134e-05, "loss": 0.0202, "step": 49340 }, { "grad_norm": 0.22365760803222656, "learning_rate": 5.5200660009431826e-05, "loss": 0.0207, "step": 49350 }, { "grad_norm": 0.23684662580490112, "learning_rate": 5.518421471276596e-05, "loss": 0.0211, "step": 49360 }, { "grad_norm": 0.1941947489976883, "learning_rate": 5.5167768849162174e-05, "loss": 0.026, "step": 49370 }, { "grad_norm": 0.26009854674339294, "learning_rate": 5.515132242041893e-05, "loss": 0.0183, "step": 49380 }, { "grad_norm": 0.25706347823143005, "learning_rate": 5.513487542833483e-05, "loss": 0.019, "step": 49390 }, { "grad_norm": 0.20842939615249634, "learning_rate": 5.5118427874708466e-05, "loss": 0.0199, "step": 49400 }, { "grad_norm": 0.20666785538196564, "learning_rate": 5.510197976133853e-05, "loss": 0.0183, "step": 49410 }, { "grad_norm": 0.25372374057769775, "learning_rate": 5.508553109002376e-05, "loss": 0.0209, "step": 49420 }, { "grad_norm": 0.21751298010349274, "learning_rate": 5.5069081862562957e-05, "loss": 0.0201, "step": 49430 }, { "grad_norm": 0.2037639170885086, "learning_rate": 5.5052632080754965e-05, "loss": 0.0224, "step": 49440 }, { "grad_norm": 0.20287249982357025, "learning_rate": 5.503618174639874e-05, "loss": 0.0196, "step": 49450 }, { "grad_norm": 0.2548051178455353, "learning_rate": 5.501973086129325e-05, "loss": 0.0185, "step": 49460 }, { "grad_norm": 0.220770925283432, "learning_rate": 5.500327942723753e-05, "loss": 0.0182, "step": 49470 }, { "grad_norm": 0.25481778383255005, "learning_rate": 5.498682744603071e-05, "loss": 0.0203, "step": 49480 }, { "grad_norm": 0.21696342527866364, "learning_rate": 5.497037491947194e-05, "loss": 0.0194, "step": 49490 }, { "grad_norm": 0.19789589941501617, "learning_rate": 5.4953921849360424e-05, "loss": 0.0165, "step": 49500 }, { "grad_norm": 0.2811076045036316, "learning_rate": 5.493746823749547e-05, "loss": 0.0182, "step": 49510 }, { "grad_norm": 0.23154719173908234, "learning_rate": 5.49210140856764e-05, "loss": 0.0245, "step": 49520 }, { "grad_norm": 0.15749596059322357, "learning_rate": 5.4904559395702647e-05, "loss": 0.0189, "step": 49530 }, { "grad_norm": 0.23058941960334778, "learning_rate": 5.488810416937364e-05, "loss": 0.0193, "step": 49540 }, { "grad_norm": 0.20277851819992065, "learning_rate": 5.4871648408488926e-05, "loss": 0.0162, "step": 49550 }, { "grad_norm": 0.20096729695796967, "learning_rate": 5.485519211484807e-05, "loss": 0.0189, "step": 49560 }, { "grad_norm": 0.23251977562904358, "learning_rate": 5.4838735290250676e-05, "loss": 0.0198, "step": 49570 }, { "grad_norm": 0.18583588302135468, "learning_rate": 5.482227793649649e-05, "loss": 0.0199, "step": 49580 }, { "grad_norm": 0.1995152086019516, "learning_rate": 5.480582005538524e-05, "loss": 0.0186, "step": 49590 }, { "grad_norm": 0.2458232045173645, "learning_rate": 5.478936164871671e-05, "loss": 0.0205, "step": 49600 }, { "grad_norm": 0.18801674246788025, "learning_rate": 5.477290271829081e-05, "loss": 0.0232, "step": 49610 }, { "grad_norm": 0.25075823068618774, "learning_rate": 5.475644326590744e-05, "loss": 0.0201, "step": 49620 }, { "grad_norm": 0.1790696233510971, "learning_rate": 5.473998329336658e-05, "loss": 0.0198, "step": 49630 }, { "grad_norm": 0.22619661688804626, "learning_rate": 5.4723522802468286e-05, "loss": 0.0194, "step": 49640 }, { "grad_norm": 0.22356751561164856, "learning_rate": 5.470706179501264e-05, "loss": 0.0199, "step": 49650 }, { "grad_norm": 0.23238518834114075, "learning_rate": 5.4690600272799776e-05, "loss": 0.0181, "step": 49660 }, { "grad_norm": 0.16288067400455475, "learning_rate": 5.467413823762993e-05, "loss": 0.0193, "step": 49670 }, { "grad_norm": 0.22198177874088287, "learning_rate": 5.465767569130335e-05, "loss": 0.02, "step": 49680 }, { "grad_norm": 0.2099529653787613, "learning_rate": 5.464121263562036e-05, "loss": 0.0153, "step": 49690 }, { "grad_norm": 0.18296678364276886, "learning_rate": 5.4624749072381343e-05, "loss": 0.0229, "step": 49700 }, { "grad_norm": 0.18021845817565918, "learning_rate": 5.460828500338672e-05, "loss": 0.0198, "step": 49710 }, { "grad_norm": 0.1420687437057495, "learning_rate": 5.459182043043698e-05, "loss": 0.0162, "step": 49720 }, { "grad_norm": 0.15656109154224396, "learning_rate": 5.457535535533265e-05, "loss": 0.0176, "step": 49730 }, { "grad_norm": 0.22650647163391113, "learning_rate": 5.4558889779874334e-05, "loss": 0.0191, "step": 49740 }, { "grad_norm": 0.2032853662967682, "learning_rate": 5.454242370586269e-05, "loss": 0.019, "step": 49750 }, { "grad_norm": 0.15626665949821472, "learning_rate": 5.452595713509843e-05, "loss": 0.0172, "step": 49760 }, { "grad_norm": 0.17162209749221802, "learning_rate": 5.4509490069382274e-05, "loss": 0.0171, "step": 49770 }, { "grad_norm": 0.1888427585363388, "learning_rate": 5.449302251051509e-05, "loss": 0.0173, "step": 49780 }, { "grad_norm": 0.15761199593544006, "learning_rate": 5.44765544602977e-05, "loss": 0.0155, "step": 49790 }, { "grad_norm": 0.20217305421829224, "learning_rate": 5.446008592053107e-05, "loss": 0.0188, "step": 49800 }, { "grad_norm": 0.17358003556728363, "learning_rate": 5.4443616893016135e-05, "loss": 0.0199, "step": 49810 }, { "grad_norm": 0.16976813971996307, "learning_rate": 5.4427147379553934e-05, "loss": 0.0173, "step": 49820 }, { "grad_norm": 0.19477955996990204, "learning_rate": 5.441067738194556e-05, "loss": 0.0197, "step": 49830 }, { "grad_norm": 0.1570097804069519, "learning_rate": 5.439420690199214e-05, "loss": 0.022, "step": 49840 }, { "grad_norm": 0.19091090559959412, "learning_rate": 5.437773594149484e-05, "loss": 0.016, "step": 49850 }, { "grad_norm": 0.38622647523880005, "learning_rate": 5.436126450225495e-05, "loss": 0.0198, "step": 49860 }, { "grad_norm": 0.16851256787776947, "learning_rate": 5.434479258607371e-05, "loss": 0.0166, "step": 49870 }, { "grad_norm": 0.2133115977048874, "learning_rate": 5.43283201947525e-05, "loss": 0.0178, "step": 49880 }, { "grad_norm": 0.22327616810798645, "learning_rate": 5.431184733009268e-05, "loss": 0.0204, "step": 49890 }, { "grad_norm": 0.1752210408449173, "learning_rate": 5.4295373993895736e-05, "loss": 0.0179, "step": 49900 }, { "grad_norm": 0.23839032649993896, "learning_rate": 5.4278900187963157e-05, "loss": 0.0196, "step": 49910 }, { "grad_norm": 0.2082325518131256, "learning_rate": 5.426242591409646e-05, "loss": 0.0182, "step": 49920 }, { "grad_norm": 0.17637045681476593, "learning_rate": 5.42459511740973e-05, "loss": 0.0191, "step": 49930 }, { "grad_norm": 0.17994408309459686, "learning_rate": 5.422947596976729e-05, "loss": 0.0157, "step": 49940 }, { "grad_norm": 0.19652847945690155, "learning_rate": 5.4213000302908134e-05, "loss": 0.0172, "step": 49950 }, { "grad_norm": 0.21011418104171753, "learning_rate": 5.419652417532162e-05, "loss": 0.0233, "step": 49960 }, { "grad_norm": 0.15342490375041962, "learning_rate": 5.4180047588809534e-05, "loss": 0.0188, "step": 49970 }, { "grad_norm": 0.18935421109199524, "learning_rate": 5.4163570545173704e-05, "loss": 0.0195, "step": 49980 }, { "grad_norm": 0.21124741435050964, "learning_rate": 5.414709304621608e-05, "loss": 0.0183, "step": 49990 }, { "grad_norm": 0.1818849742412567, "learning_rate": 5.413061509373858e-05, "loss": 0.0174, "step": 50000 }, { "grad_norm": 0.16073913872241974, "learning_rate": 5.411413668954324e-05, "loss": 0.0183, "step": 50010 }, { "grad_norm": 0.23096440732479095, "learning_rate": 5.4097657835432083e-05, "loss": 0.0197, "step": 50020 }, { "grad_norm": 0.18734410405158997, "learning_rate": 5.408117853320723e-05, "loss": 0.0167, "step": 50030 }, { "grad_norm": 0.17305541038513184, "learning_rate": 5.406469878467084e-05, "loss": 0.0188, "step": 50040 }, { "grad_norm": 0.22898150980472565, "learning_rate": 5.404821859162509e-05, "loss": 0.0184, "step": 50050 }, { "grad_norm": 0.18688532710075378, "learning_rate": 5.403173795587225e-05, "loss": 0.0208, "step": 50060 }, { "grad_norm": 0.19922126829624176, "learning_rate": 5.4015256879214606e-05, "loss": 0.0191, "step": 50070 }, { "grad_norm": 0.15658384561538696, "learning_rate": 5.39987753634545e-05, "loss": 0.0143, "step": 50080 }, { "grad_norm": 0.17840270698070526, "learning_rate": 5.3982293410394335e-05, "loss": 0.0199, "step": 50090 }, { "grad_norm": 0.2045721560716629, "learning_rate": 5.396581102183655e-05, "loss": 0.0155, "step": 50100 }, { "grad_norm": 0.2378220409154892, "learning_rate": 5.3949328199583634e-05, "loss": 0.0176, "step": 50110 }, { "grad_norm": 0.22796730697155, "learning_rate": 5.393284494543813e-05, "loss": 0.0182, "step": 50120 }, { "grad_norm": 0.21848906576633453, "learning_rate": 5.391636126120262e-05, "loss": 0.02, "step": 50130 }, { "grad_norm": 0.21640273928642273, "learning_rate": 5.389987714867971e-05, "loss": 0.0243, "step": 50140 }, { "grad_norm": 0.20120246708393097, "learning_rate": 5.3883392609672114e-05, "loss": 0.0169, "step": 50150 }, { "grad_norm": 0.1702539622783661, "learning_rate": 5.3866907645982545e-05, "loss": 0.0179, "step": 50160 }, { "grad_norm": 0.20504862070083618, "learning_rate": 5.385042225941375e-05, "loss": 0.0172, "step": 50170 }, { "grad_norm": 0.19199888408184052, "learning_rate": 5.3833936451768585e-05, "loss": 0.0172, "step": 50180 }, { "grad_norm": 0.2311856597661972, "learning_rate": 5.3817450224849884e-05, "loss": 0.0205, "step": 50190 }, { "grad_norm": 0.1871279776096344, "learning_rate": 5.380096358046056e-05, "loss": 0.0188, "step": 50200 }, { "grad_norm": 0.21379821002483368, "learning_rate": 5.378447652040359e-05, "loss": 0.0159, "step": 50210 }, { "grad_norm": 0.18750403821468353, "learning_rate": 5.376798904648195e-05, "loss": 0.0188, "step": 50220 }, { "grad_norm": 0.2220831960439682, "learning_rate": 5.375150116049869e-05, "loss": 0.0204, "step": 50230 }, { "grad_norm": 0.2060358226299286, "learning_rate": 5.373501286425691e-05, "loss": 0.0161, "step": 50240 }, { "grad_norm": 0.1781264692544937, "learning_rate": 5.3718524159559726e-05, "loss": 0.0183, "step": 50250 }, { "grad_norm": 0.16799676418304443, "learning_rate": 5.370203504821034e-05, "loss": 0.018, "step": 50260 }, { "grad_norm": 0.19358780980110168, "learning_rate": 5.368554553201196e-05, "loss": 0.0214, "step": 50270 }, { "grad_norm": 0.19305585324764252, "learning_rate": 5.3669055612767874e-05, "loss": 0.0211, "step": 50280 }, { "grad_norm": 0.2333652377128601, "learning_rate": 5.3652565292281376e-05, "loss": 0.0208, "step": 50290 }, { "grad_norm": 0.24552108347415924, "learning_rate": 5.3636074572355824e-05, "loss": 0.0171, "step": 50300 }, { "grad_norm": 0.2003379911184311, "learning_rate": 5.361958345479463e-05, "loss": 0.0182, "step": 50310 }, { "grad_norm": 0.24358636140823364, "learning_rate": 5.360309194140123e-05, "loss": 0.0185, "step": 50320 }, { "grad_norm": 0.25357329845428467, "learning_rate": 5.3586600033979106e-05, "loss": 0.0198, "step": 50330 }, { "grad_norm": 0.20659524202346802, "learning_rate": 5.357010773433181e-05, "loss": 0.0187, "step": 50340 }, { "grad_norm": 0.20839375257492065, "learning_rate": 5.3553615044262895e-05, "loss": 0.0186, "step": 50350 }, { "grad_norm": 0.1874624639749527, "learning_rate": 5.3537121965575964e-05, "loss": 0.0195, "step": 50360 }, { "grad_norm": 0.1891951858997345, "learning_rate": 5.352062850007471e-05, "loss": 0.0158, "step": 50370 }, { "grad_norm": 0.2073683738708496, "learning_rate": 5.350413464956282e-05, "loss": 0.0178, "step": 50380 }, { "grad_norm": 0.20244884490966797, "learning_rate": 5.348764041584403e-05, "loss": 0.0147, "step": 50390 }, { "grad_norm": 0.1810595691204071, "learning_rate": 5.347114580072212e-05, "loss": 0.0161, "step": 50400 }, { "grad_norm": 0.1938646286725998, "learning_rate": 5.345465080600093e-05, "loss": 0.0154, "step": 50410 }, { "grad_norm": 0.3731912672519684, "learning_rate": 5.3438155433484326e-05, "loss": 0.0148, "step": 50420 }, { "grad_norm": 0.2010578066110611, "learning_rate": 5.3421659684976197e-05, "loss": 0.0206, "step": 50430 }, { "grad_norm": 0.16968844830989838, "learning_rate": 5.340516356228052e-05, "loss": 0.0188, "step": 50440 }, { "grad_norm": 0.24389393627643585, "learning_rate": 5.338866706720128e-05, "loss": 0.0238, "step": 50450 }, { "grad_norm": 0.23799337446689606, "learning_rate": 5.337217020154249e-05, "loss": 0.0192, "step": 50460 }, { "grad_norm": 0.22105109691619873, "learning_rate": 5.335567296710825e-05, "loss": 0.0218, "step": 50470 }, { "grad_norm": 0.24054163694381714, "learning_rate": 5.333917536570265e-05, "loss": 0.0198, "step": 50480 }, { "grad_norm": 0.16632534563541412, "learning_rate": 5.332267739912986e-05, "loss": 0.0199, "step": 50490 }, { "grad_norm": 0.20101352035999298, "learning_rate": 5.330617906919405e-05, "loss": 0.0194, "step": 50500 }, { "grad_norm": 0.13760101795196533, "learning_rate": 5.328968037769949e-05, "loss": 0.0165, "step": 50510 }, { "grad_norm": 0.17222820222377777, "learning_rate": 5.327318132645042e-05, "loss": 0.0198, "step": 50520 }, { "grad_norm": 0.17748509347438812, "learning_rate": 5.325668191725116e-05, "loss": 0.0194, "step": 50530 }, { "grad_norm": 0.1805678904056549, "learning_rate": 5.324018215190606e-05, "loss": 0.0176, "step": 50540 }, { "grad_norm": 0.21690697968006134, "learning_rate": 5.3223682032219515e-05, "loss": 0.0205, "step": 50550 }, { "grad_norm": 0.16855517029762268, "learning_rate": 5.320718155999595e-05, "loss": 0.0196, "step": 50560 }, { "grad_norm": 0.2240501493215561, "learning_rate": 5.3190680737039835e-05, "loss": 0.0237, "step": 50570 }, { "grad_norm": 0.21504461765289307, "learning_rate": 5.317417956515567e-05, "loss": 0.0196, "step": 50580 }, { "grad_norm": 0.2551702857017517, "learning_rate": 5.315767804614798e-05, "loss": 0.0223, "step": 50590 }, { "grad_norm": 0.1730000376701355, "learning_rate": 5.3141176181821395e-05, "loss": 0.0163, "step": 50600 }, { "grad_norm": 0.15043827891349792, "learning_rate": 5.31246739739805e-05, "loss": 0.0244, "step": 50610 }, { "grad_norm": 0.20413269102573395, "learning_rate": 5.310817142442994e-05, "loss": 0.0181, "step": 50620 }, { "grad_norm": 0.19618389010429382, "learning_rate": 5.309166853497445e-05, "loss": 0.0194, "step": 50630 }, { "grad_norm": 0.23730595409870148, "learning_rate": 5.307516530741873e-05, "loss": 0.0208, "step": 50640 }, { "grad_norm": 0.16642577946186066, "learning_rate": 5.305866174356754e-05, "loss": 0.0177, "step": 50650 }, { "grad_norm": 0.18117794394493103, "learning_rate": 5.304215784522571e-05, "loss": 0.0208, "step": 50660 }, { "grad_norm": 0.2033066302537918, "learning_rate": 5.302565361419808e-05, "loss": 0.0187, "step": 50670 }, { "grad_norm": 0.16861222684383392, "learning_rate": 5.3009149052289507e-05, "loss": 0.0204, "step": 50680 }, { "grad_norm": 0.16098852455615997, "learning_rate": 5.299264416130493e-05, "loss": 0.0186, "step": 50690 }, { "grad_norm": 0.2233753651380539, "learning_rate": 5.297613894304928e-05, "loss": 0.0142, "step": 50700 }, { "grad_norm": 0.21601025760173798, "learning_rate": 5.2959633399327534e-05, "loss": 0.0186, "step": 50710 }, { "grad_norm": 0.23752200603485107, "learning_rate": 5.294312753194476e-05, "loss": 0.0186, "step": 50720 }, { "grad_norm": 0.1304495930671692, "learning_rate": 5.292662134270596e-05, "loss": 0.0178, "step": 50730 }, { "grad_norm": 0.20734666287899017, "learning_rate": 5.291011483341626e-05, "loss": 0.022, "step": 50740 }, { "grad_norm": 0.21229825913906097, "learning_rate": 5.2893608005880767e-05, "loss": 0.0179, "step": 50750 }, { "grad_norm": 0.1877831220626831, "learning_rate": 5.287710086190467e-05, "loss": 0.022, "step": 50760 }, { "grad_norm": 0.19565925002098083, "learning_rate": 5.286059340329314e-05, "loss": 0.0213, "step": 50770 }, { "grad_norm": 0.21392351388931274, "learning_rate": 5.284408563185141e-05, "loss": 0.0167, "step": 50780 }, { "grad_norm": 0.16295170783996582, "learning_rate": 5.282757754938476e-05, "loss": 0.0195, "step": 50790 }, { "grad_norm": 0.2188550978899002, "learning_rate": 5.281106915769849e-05, "loss": 0.0159, "step": 50800 }, { "grad_norm": 0.2235373854637146, "learning_rate": 5.2794560458597897e-05, "loss": 0.0225, "step": 50810 }, { "grad_norm": 0.16143003106117249, "learning_rate": 5.27780514538884e-05, "loss": 0.0204, "step": 50820 }, { "grad_norm": 0.2313777506351471, "learning_rate": 5.2761542145375365e-05, "loss": 0.0204, "step": 50830 }, { "grad_norm": 0.17491841316223145, "learning_rate": 5.274503253486421e-05, "loss": 0.0217, "step": 50840 }, { "grad_norm": 0.17933544516563416, "learning_rate": 5.272852262416046e-05, "loss": 0.0177, "step": 50850 }, { "grad_norm": 0.14183209836483002, "learning_rate": 5.2712012415069555e-05, "loss": 0.0182, "step": 50860 }, { "grad_norm": 0.17292892932891846, "learning_rate": 5.269550190939705e-05, "loss": 0.0203, "step": 50870 }, { "grad_norm": 0.25134751200675964, "learning_rate": 5.267899110894852e-05, "loss": 0.0213, "step": 50880 }, { "grad_norm": 0.2319619059562683, "learning_rate": 5.266248001552955e-05, "loss": 0.016, "step": 50890 }, { "grad_norm": 0.17954929172992706, "learning_rate": 5.264596863094575e-05, "loss": 0.0175, "step": 50900 }, { "grad_norm": 0.21825909614562988, "learning_rate": 5.262945695700282e-05, "loss": 0.0242, "step": 50910 }, { "grad_norm": 0.21866731345653534, "learning_rate": 5.261294499550643e-05, "loss": 0.0208, "step": 50920 }, { "grad_norm": 0.24035347998142242, "learning_rate": 5.25964327482623e-05, "loss": 0.0228, "step": 50930 }, { "grad_norm": 0.21064497530460358, "learning_rate": 5.257992021707617e-05, "loss": 0.0205, "step": 50940 }, { "grad_norm": 0.18543724715709686, "learning_rate": 5.256340740375387e-05, "loss": 0.0229, "step": 50950 }, { "grad_norm": 0.1548229306936264, "learning_rate": 5.254689431010117e-05, "loss": 0.0187, "step": 50960 }, { "grad_norm": 0.1599450409412384, "learning_rate": 5.253038093792395e-05, "loss": 0.0197, "step": 50970 }, { "grad_norm": 0.15205062925815582, "learning_rate": 5.251386728902806e-05, "loss": 0.0197, "step": 50980 }, { "grad_norm": 0.20417387783527374, "learning_rate": 5.2497353365219446e-05, "loss": 0.0166, "step": 50990 }, { "grad_norm": 0.20428574085235596, "learning_rate": 5.2480839168304e-05, "loss": 0.0215, "step": 51000 }, { "grad_norm": 0.25652429461479187, "learning_rate": 5.246432470008773e-05, "loss": 0.0203, "step": 51010 }, { "grad_norm": 0.16952821612358093, "learning_rate": 5.2447809962376616e-05, "loss": 0.0199, "step": 51020 }, { "grad_norm": 0.16451051831245422, "learning_rate": 5.243129495697666e-05, "loss": 0.0186, "step": 51030 }, { "grad_norm": 0.26837974786758423, "learning_rate": 5.241477968569397e-05, "loss": 0.0266, "step": 51040 }, { "grad_norm": 0.22673997282981873, "learning_rate": 5.23982641503346e-05, "loss": 0.0206, "step": 51050 }, { "grad_norm": 0.1715071201324463, "learning_rate": 5.238174835270465e-05, "loss": 0.0187, "step": 51060 }, { "grad_norm": 0.21671751141548157, "learning_rate": 5.23652322946103e-05, "loss": 0.0189, "step": 51070 }, { "grad_norm": 0.21819594502449036, "learning_rate": 5.234871597785771e-05, "loss": 0.0206, "step": 51080 }, { "grad_norm": 0.19959591329097748, "learning_rate": 5.233219940425306e-05, "loss": 0.0146, "step": 51090 }, { "grad_norm": 0.20351868867874146, "learning_rate": 5.2315682575602576e-05, "loss": 0.0172, "step": 51100 }, { "grad_norm": 0.20365801453590393, "learning_rate": 5.229916549371254e-05, "loss": 0.0186, "step": 51110 }, { "grad_norm": 0.14286698400974274, "learning_rate": 5.228264816038924e-05, "loss": 0.015, "step": 51120 }, { "grad_norm": 0.2114759087562561, "learning_rate": 5.2266130577438935e-05, "loss": 0.0179, "step": 51130 }, { "grad_norm": 0.2198518067598343, "learning_rate": 5.224961274666802e-05, "loss": 0.0183, "step": 51140 }, { "grad_norm": 0.2070714384317398, "learning_rate": 5.223309466988283e-05, "loss": 0.0185, "step": 51150 }, { "grad_norm": 0.25541165471076965, "learning_rate": 5.221657634888976e-05, "loss": 0.021, "step": 51160 }, { "grad_norm": 0.277004212141037, "learning_rate": 5.2200057785495235e-05, "loss": 0.018, "step": 51170 }, { "grad_norm": 0.17549993097782135, "learning_rate": 5.2183538981505706e-05, "loss": 0.0217, "step": 51180 }, { "grad_norm": 0.1556733101606369, "learning_rate": 5.216701993872762e-05, "loss": 0.0213, "step": 51190 }, { "grad_norm": 0.1751832813024521, "learning_rate": 5.215050065896752e-05, "loss": 0.0169, "step": 51200 }, { "grad_norm": 0.2246718406677246, "learning_rate": 5.2133981144031875e-05, "loss": 0.0138, "step": 51210 }, { "grad_norm": 0.1554124802350998, "learning_rate": 5.211746139572727e-05, "loss": 0.0166, "step": 51220 }, { "grad_norm": 0.14760948717594147, "learning_rate": 5.210094141586026e-05, "loss": 0.0175, "step": 51230 }, { "grad_norm": 0.15913496911525726, "learning_rate": 5.2084421206237476e-05, "loss": 0.0176, "step": 51240 }, { "grad_norm": 0.16434621810913086, "learning_rate": 5.20679007686655e-05, "loss": 0.0174, "step": 51250 }, { "grad_norm": 0.12448983639478683, "learning_rate": 5.205138010495102e-05, "loss": 0.0207, "step": 51260 }, { "grad_norm": 0.14941026270389557, "learning_rate": 5.20348592169007e-05, "loss": 0.0151, "step": 51270 }, { "grad_norm": 0.213431254029274, "learning_rate": 5.201833810632123e-05, "loss": 0.0185, "step": 51280 }, { "grad_norm": 0.20292846858501434, "learning_rate": 5.200181677501933e-05, "loss": 0.0167, "step": 51290 }, { "grad_norm": 0.2186279445886612, "learning_rate": 5.1985295224801777e-05, "loss": 0.02, "step": 51300 }, { "grad_norm": 0.1808885931968689, "learning_rate": 5.196877345747531e-05, "loss": 0.0207, "step": 51310 }, { "grad_norm": 0.2116571068763733, "learning_rate": 5.195225147484674e-05, "loss": 0.0175, "step": 51320 }, { "grad_norm": 0.27433791756629944, "learning_rate": 5.193572927872288e-05, "loss": 0.0214, "step": 51330 }, { "grad_norm": 0.2006361484527588, "learning_rate": 5.191920687091059e-05, "loss": 0.0214, "step": 51340 }, { "grad_norm": 0.15624257922172546, "learning_rate": 5.1902684253216695e-05, "loss": 0.0196, "step": 51350 }, { "grad_norm": 0.15953287482261658, "learning_rate": 5.1886161427448134e-05, "loss": 0.0163, "step": 51360 }, { "grad_norm": 0.19155515730381012, "learning_rate": 5.186963839541179e-05, "loss": 0.0154, "step": 51370 }, { "grad_norm": 0.16451965272426605, "learning_rate": 5.185311515891459e-05, "loss": 0.0142, "step": 51380 }, { "grad_norm": 0.21804915368556976, "learning_rate": 5.1836591719763516e-05, "loss": 0.0201, "step": 51390 }, { "grad_norm": 0.19137653708457947, "learning_rate": 5.182006807976554e-05, "loss": 0.0184, "step": 51400 }, { "grad_norm": 0.1752043068408966, "learning_rate": 5.180354424072763e-05, "loss": 0.0201, "step": 51410 }, { "grad_norm": 0.22909457981586456, "learning_rate": 5.178702020445686e-05, "loss": 0.0224, "step": 51420 }, { "grad_norm": 0.210322767496109, "learning_rate": 5.177049597276024e-05, "loss": 0.0177, "step": 51430 }, { "grad_norm": 0.1715833842754364, "learning_rate": 5.175397154744482e-05, "loss": 0.0183, "step": 51440 }, { "grad_norm": 0.2631772458553314, "learning_rate": 5.1737446930317736e-05, "loss": 0.0207, "step": 51450 }, { "grad_norm": 0.24432939291000366, "learning_rate": 5.172092212318604e-05, "loss": 0.0188, "step": 51460 }, { "grad_norm": 0.20336289703845978, "learning_rate": 5.170439712785692e-05, "loss": 0.0175, "step": 51470 }, { "grad_norm": 0.22288759052753448, "learning_rate": 5.1687871946137467e-05, "loss": 0.0176, "step": 51480 }, { "grad_norm": 0.20432904362678528, "learning_rate": 5.16713465798349e-05, "loss": 0.0177, "step": 51490 }, { "grad_norm": 0.17827962338924408, "learning_rate": 5.165482103075637e-05, "loss": 0.0182, "step": 51500 }, { "grad_norm": 0.20264199376106262, "learning_rate": 5.163829530070909e-05, "loss": 0.0176, "step": 51510 }, { "grad_norm": 0.24571552872657776, "learning_rate": 5.162176939150032e-05, "loss": 0.0196, "step": 51520 }, { "grad_norm": 0.2073485255241394, "learning_rate": 5.160524330493728e-05, "loss": 0.0167, "step": 51530 }, { "grad_norm": 0.1867312788963318, "learning_rate": 5.158871704282724e-05, "loss": 0.0156, "step": 51540 }, { "grad_norm": 0.20106881856918335, "learning_rate": 5.15721906069775e-05, "loss": 0.0192, "step": 51550 }, { "grad_norm": 0.22304558753967285, "learning_rate": 5.155566399919536e-05, "loss": 0.0193, "step": 51560 }, { "grad_norm": 0.19241660833358765, "learning_rate": 5.153913722128813e-05, "loss": 0.0181, "step": 51570 }, { "grad_norm": 0.18330176174640656, "learning_rate": 5.1522610275063176e-05, "loss": 0.0196, "step": 51580 }, { "grad_norm": 0.19284610450267792, "learning_rate": 5.150608316232786e-05, "loss": 0.0186, "step": 51590 }, { "grad_norm": 0.1323850303888321, "learning_rate": 5.148955588488955e-05, "loss": 0.0159, "step": 51600 }, { "grad_norm": 0.238554447889328, "learning_rate": 5.147302844455564e-05, "loss": 0.0187, "step": 51610 }, { "grad_norm": 0.5954830050468445, "learning_rate": 5.145650084313356e-05, "loss": 0.0165, "step": 51620 }, { "grad_norm": 0.15799088776111603, "learning_rate": 5.1439973082430734e-05, "loss": 0.0176, "step": 51630 }, { "grad_norm": 0.23278625309467316, "learning_rate": 5.142344516425461e-05, "loss": 0.0204, "step": 51640 }, { "grad_norm": 0.1600324660539627, "learning_rate": 5.1406917090412673e-05, "loss": 0.0231, "step": 51650 }, { "grad_norm": 0.20534564554691315, "learning_rate": 5.1390388862712404e-05, "loss": 0.0202, "step": 51660 }, { "grad_norm": 0.25278785824775696, "learning_rate": 5.137386048296129e-05, "loss": 0.019, "step": 51670 }, { "grad_norm": 0.1858716905117035, "learning_rate": 5.135733195296686e-05, "loss": 0.0184, "step": 51680 }, { "grad_norm": 0.1919567734003067, "learning_rate": 5.134080327453664e-05, "loss": 0.0177, "step": 51690 }, { "grad_norm": 0.1851094663143158, "learning_rate": 5.1324274449478204e-05, "loss": 0.0192, "step": 51700 }, { "grad_norm": 0.25974178314208984, "learning_rate": 5.130774547959909e-05, "loss": 0.0236, "step": 51710 }, { "grad_norm": 0.22006720304489136, "learning_rate": 5.129121636670692e-05, "loss": 0.0209, "step": 51720 }, { "grad_norm": 0.1619269698858261, "learning_rate": 5.1274687112609246e-05, "loss": 0.0166, "step": 51730 }, { "grad_norm": 0.14109821617603302, "learning_rate": 5.125815771911372e-05, "loss": 0.0177, "step": 51740 }, { "grad_norm": 0.19650165736675262, "learning_rate": 5.1241628188027956e-05, "loss": 0.0204, "step": 51750 }, { "grad_norm": 0.2374773770570755, "learning_rate": 5.1225098521159585e-05, "loss": 0.0185, "step": 51760 }, { "grad_norm": 0.15142741799354553, "learning_rate": 5.12085687203163e-05, "loss": 0.0169, "step": 51770 }, { "grad_norm": 0.1965264081954956, "learning_rate": 5.119203878730575e-05, "loss": 0.0204, "step": 51780 }, { "grad_norm": 0.2471356838941574, "learning_rate": 5.117550872393564e-05, "loss": 0.0188, "step": 51790 }, { "grad_norm": 0.23398885130882263, "learning_rate": 5.1158978532013645e-05, "loss": 0.0185, "step": 51800 }, { "grad_norm": 0.1293310970067978, "learning_rate": 5.114244821334751e-05, "loss": 0.0172, "step": 51810 }, { "grad_norm": 0.2753051817417145, "learning_rate": 5.112591776974497e-05, "loss": 0.0223, "step": 51820 }, { "grad_norm": 0.21889515221118927, "learning_rate": 5.110938720301374e-05, "loss": 0.0224, "step": 51830 }, { "grad_norm": 0.18590311706066132, "learning_rate": 5.1092856514961595e-05, "loss": 0.0181, "step": 51840 }, { "grad_norm": 0.2009938806295395, "learning_rate": 5.107632570739631e-05, "loss": 0.0177, "step": 51850 }, { "grad_norm": 0.19279620051383972, "learning_rate": 5.1059794782125645e-05, "loss": 0.0176, "step": 51860 }, { "grad_norm": 0.12465199828147888, "learning_rate": 5.104326374095744e-05, "loss": 0.0165, "step": 51870 }, { "grad_norm": 0.13497133553028107, "learning_rate": 5.102673258569947e-05, "loss": 0.0169, "step": 51880 }, { "grad_norm": 0.22754788398742676, "learning_rate": 5.101020131815956e-05, "loss": 0.0194, "step": 51890 }, { "grad_norm": 0.2220461666584015, "learning_rate": 5.099366994014556e-05, "loss": 0.02, "step": 51900 }, { "grad_norm": 0.17140692472457886, "learning_rate": 5.097713845346532e-05, "loss": 0.0148, "step": 51910 }, { "grad_norm": 0.18946805596351624, "learning_rate": 5.096060685992668e-05, "loss": 0.0182, "step": 51920 }, { "grad_norm": 0.15363775193691254, "learning_rate": 5.0944075161337514e-05, "loss": 0.0168, "step": 51930 }, { "grad_norm": 0.1622934341430664, "learning_rate": 5.09275433595057e-05, "loss": 0.0156, "step": 51940 }, { "grad_norm": 0.18799197673797607, "learning_rate": 5.0911011456239157e-05, "loss": 0.0185, "step": 51950 }, { "grad_norm": 0.2135840654373169, "learning_rate": 5.089447945334574e-05, "loss": 0.0188, "step": 51960 }, { "grad_norm": 0.1963774561882019, "learning_rate": 5.087794735263343e-05, "loss": 0.0201, "step": 51970 }, { "grad_norm": 0.23683495819568634, "learning_rate": 5.086141515591011e-05, "loss": 0.0182, "step": 51980 }, { "grad_norm": 0.20435982942581177, "learning_rate": 5.084488286498371e-05, "loss": 0.0185, "step": 51990 }, { "grad_norm": 0.13985443115234375, "learning_rate": 5.08283504816622e-05, "loss": 0.0157, "step": 52000 }, { "grad_norm": 0.174131840467453, "learning_rate": 5.081181800775353e-05, "loss": 0.0196, "step": 52010 }, { "grad_norm": 0.17359422147274017, "learning_rate": 5.079528544506565e-05, "loss": 0.0243, "step": 52020 }, { "grad_norm": 0.1695999801158905, "learning_rate": 5.077875279540657e-05, "loss": 0.0172, "step": 52030 }, { "grad_norm": 0.30109864473342896, "learning_rate": 5.076222006058424e-05, "loss": 0.0161, "step": 52040 }, { "grad_norm": 0.19942334294319153, "learning_rate": 5.074568724240667e-05, "loss": 0.0195, "step": 52050 }, { "grad_norm": 0.15619052946567535, "learning_rate": 5.072915434268187e-05, "loss": 0.0208, "step": 52060 }, { "grad_norm": 0.22153274714946747, "learning_rate": 5.071262136321785e-05, "loss": 0.0172, "step": 52070 }, { "grad_norm": 0.36316344141960144, "learning_rate": 5.069608830582262e-05, "loss": 0.0171, "step": 52080 }, { "grad_norm": 0.17498332262039185, "learning_rate": 5.067955517230423e-05, "loss": 0.0173, "step": 52090 }, { "grad_norm": 0.19557861983776093, "learning_rate": 5.06630219644707e-05, "loss": 0.0189, "step": 52100 }, { "grad_norm": 0.17736102640628815, "learning_rate": 5.064648868413008e-05, "loss": 0.016, "step": 52110 }, { "grad_norm": 0.18946205079555511, "learning_rate": 5.062995533309044e-05, "loss": 0.0209, "step": 52120 }, { "grad_norm": 0.3936944007873535, "learning_rate": 5.061342191315983e-05, "loss": 0.0187, "step": 52130 }, { "grad_norm": 0.18730421364307404, "learning_rate": 5.0596888426146325e-05, "loss": 0.0192, "step": 52140 }, { "grad_norm": 0.18400296568870544, "learning_rate": 5.058035487385798e-05, "loss": 0.0169, "step": 52150 }, { "grad_norm": 0.2176365703344345, "learning_rate": 5.0563821258102915e-05, "loss": 0.0161, "step": 52160 }, { "grad_norm": 0.15389859676361084, "learning_rate": 5.05472875806892e-05, "loss": 0.0148, "step": 52170 }, { "grad_norm": 0.21970456838607788, "learning_rate": 5.053075384342494e-05, "loss": 0.0199, "step": 52180 }, { "grad_norm": 0.23257744312286377, "learning_rate": 5.0514220048118234e-05, "loss": 0.0204, "step": 52190 }, { "grad_norm": 0.25878965854644775, "learning_rate": 5.049768619657721e-05, "loss": 0.0174, "step": 52200 }, { "grad_norm": 0.16129130125045776, "learning_rate": 5.048115229060996e-05, "loss": 0.0168, "step": 52210 }, { "grad_norm": 0.15214385092258453, "learning_rate": 5.046461833202462e-05, "loss": 0.0175, "step": 52220 }, { "grad_norm": 0.208980992436409, "learning_rate": 5.044808432262932e-05, "loss": 0.0176, "step": 52230 }, { "grad_norm": 0.19939200580120087, "learning_rate": 5.043155026423219e-05, "loss": 0.0168, "step": 52240 }, { "grad_norm": 0.23581242561340332, "learning_rate": 5.041501615864137e-05, "loss": 0.0165, "step": 52250 }, { "grad_norm": 0.18512295186519623, "learning_rate": 5.039848200766502e-05, "loss": 0.0161, "step": 52260 }, { "grad_norm": 0.22003766894340515, "learning_rate": 5.038194781311125e-05, "loss": 0.02, "step": 52270 }, { "grad_norm": 0.19159454107284546, "learning_rate": 5.0365413576788264e-05, "loss": 0.0157, "step": 52280 }, { "grad_norm": 0.2051037698984146, "learning_rate": 5.0348879300504194e-05, "loss": 0.0183, "step": 52290 }, { "grad_norm": 0.2025514841079712, "learning_rate": 5.033234498606721e-05, "loss": 0.0175, "step": 52300 }, { "grad_norm": 0.18603596091270447, "learning_rate": 5.0315810635285476e-05, "loss": 0.0177, "step": 52310 }, { "grad_norm": 0.20038791000843048, "learning_rate": 5.0299276249967164e-05, "loss": 0.015, "step": 52320 }, { "grad_norm": 0.14562474191188812, "learning_rate": 5.0282741831920454e-05, "loss": 0.0204, "step": 52330 }, { "grad_norm": 0.19297416508197784, "learning_rate": 5.026620738295351e-05, "loss": 0.018, "step": 52340 }, { "grad_norm": 0.19304992258548737, "learning_rate": 5.024967290487455e-05, "loss": 0.0142, "step": 52350 }, { "grad_norm": 0.18116751313209534, "learning_rate": 5.023313839949172e-05, "loss": 0.0156, "step": 52360 }, { "grad_norm": 0.1987074315547943, "learning_rate": 5.0216603868613224e-05, "loss": 0.0172, "step": 52370 }, { "grad_norm": 0.1637817770242691, "learning_rate": 5.020006931404726e-05, "loss": 0.0162, "step": 52380 }, { "grad_norm": 0.19406792521476746, "learning_rate": 5.0183534737602014e-05, "loss": 0.0204, "step": 52390 }, { "grad_norm": 0.1527727246284485, "learning_rate": 5.0167000141085686e-05, "loss": 0.0158, "step": 52400 }, { "grad_norm": 0.18313470482826233, "learning_rate": 5.015046552630647e-05, "loss": 0.0206, "step": 52410 }, { "grad_norm": 0.17936590313911438, "learning_rate": 5.013393089507258e-05, "loss": 0.0186, "step": 52420 }, { "grad_norm": 0.19794367253780365, "learning_rate": 5.011739624919219e-05, "loss": 0.0208, "step": 52430 }, { "grad_norm": 0.20742475986480713, "learning_rate": 5.0100861590473545e-05, "loss": 0.0173, "step": 52440 }, { "grad_norm": 0.22582729160785675, "learning_rate": 5.008432692072482e-05, "loss": 0.0246, "step": 52450 }, { "grad_norm": 0.2513780891895294, "learning_rate": 5.006779224175423e-05, "loss": 0.0207, "step": 52460 }, { "grad_norm": 0.23537635803222656, "learning_rate": 5.005125755536998e-05, "loss": 0.0183, "step": 52470 }, { "grad_norm": 0.3805379569530487, "learning_rate": 5.003472286338029e-05, "loss": 0.0189, "step": 52480 }, { "grad_norm": 0.1311885118484497, "learning_rate": 5.001818816759335e-05, "loss": 0.0142, "step": 52490 }, { "grad_norm": 0.18203550577163696, "learning_rate": 5.000165346981738e-05, "loss": 0.0186, "step": 52500 }, { "grad_norm": 0.1996927261352539, "learning_rate": 4.9985118771860595e-05, "loss": 0.0203, "step": 52510 }, { "grad_norm": 0.12340784072875977, "learning_rate": 4.996858407553119e-05, "loss": 0.0164, "step": 52520 }, { "grad_norm": 0.21025432646274567, "learning_rate": 4.99520493826374e-05, "loss": 0.0175, "step": 52530 }, { "grad_norm": 0.18875236809253693, "learning_rate": 4.9935514694987384e-05, "loss": 0.0193, "step": 52540 }, { "grad_norm": 0.1425473988056183, "learning_rate": 4.991898001438941e-05, "loss": 0.0171, "step": 52550 }, { "grad_norm": 0.17912833392620087, "learning_rate": 4.990244534265166e-05, "loss": 0.0146, "step": 52560 }, { "grad_norm": 0.1595565378665924, "learning_rate": 4.988591068158234e-05, "loss": 0.0153, "step": 52570 }, { "grad_norm": 0.2276812195777893, "learning_rate": 4.986937603298965e-05, "loss": 0.0155, "step": 52580 }, { "grad_norm": 0.16809843480587006, "learning_rate": 4.985284139868181e-05, "loss": 0.0199, "step": 52590 }, { "grad_norm": 0.15519796311855316, "learning_rate": 4.9836306780467e-05, "loss": 0.0157, "step": 52600 }, { "grad_norm": 0.19003179669380188, "learning_rate": 4.981977218015343e-05, "loss": 0.0182, "step": 52610 }, { "grad_norm": 0.18964901566505432, "learning_rate": 4.980323759954932e-05, "loss": 0.0172, "step": 52620 }, { "grad_norm": 0.19315475225448608, "learning_rate": 4.978670304046284e-05, "loss": 0.0189, "step": 52630 }, { "grad_norm": 0.1584247648715973, "learning_rate": 4.9770168504702195e-05, "loss": 0.017, "step": 52640 }, { "grad_norm": 0.2149994969367981, "learning_rate": 4.975363399407559e-05, "loss": 0.0181, "step": 52650 }, { "grad_norm": 0.16277824342250824, "learning_rate": 4.9737099510391194e-05, "loss": 0.0192, "step": 52660 }, { "grad_norm": 0.22801880538463593, "learning_rate": 4.972056505545719e-05, "loss": 0.0169, "step": 52670 }, { "grad_norm": 0.19312947988510132, "learning_rate": 4.9704030631081775e-05, "loss": 0.0179, "step": 52680 }, { "grad_norm": 0.20858223736286163, "learning_rate": 4.968749623907313e-05, "loss": 0.0238, "step": 52690 }, { "grad_norm": 0.15622378885746002, "learning_rate": 4.967096188123943e-05, "loss": 0.0177, "step": 52700 }, { "grad_norm": 0.22117839753627777, "learning_rate": 4.965442755938884e-05, "loss": 0.0167, "step": 52710 }, { "grad_norm": 0.22277210652828217, "learning_rate": 4.963789327532954e-05, "loss": 0.0212, "step": 52720 }, { "grad_norm": 0.19596490263938904, "learning_rate": 4.962135903086968e-05, "loss": 0.0169, "step": 52730 }, { "grad_norm": 0.22182311117649078, "learning_rate": 4.960482482781741e-05, "loss": 0.0163, "step": 52740 }, { "grad_norm": 0.18890665471553802, "learning_rate": 4.958829066798092e-05, "loss": 0.0175, "step": 52750 }, { "grad_norm": 0.2711225152015686, "learning_rate": 4.957175655316835e-05, "loss": 0.0214, "step": 52760 }, { "grad_norm": 0.1893531233072281, "learning_rate": 4.9555222485187834e-05, "loss": 0.0173, "step": 52770 }, { "grad_norm": 0.21860967576503754, "learning_rate": 4.953868846584752e-05, "loss": 0.0197, "step": 52780 }, { "grad_norm": 0.1322392076253891, "learning_rate": 4.9522154496955544e-05, "loss": 0.0159, "step": 52790 }, { "grad_norm": 0.17571789026260376, "learning_rate": 4.950562058032001e-05, "loss": 0.0165, "step": 52800 }, { "grad_norm": 0.18009084463119507, "learning_rate": 4.948908671774909e-05, "loss": 0.0191, "step": 52810 }, { "grad_norm": 0.17539577186107635, "learning_rate": 4.947255291105087e-05, "loss": 0.0173, "step": 52820 }, { "grad_norm": 0.26872971653938293, "learning_rate": 4.945601916203348e-05, "loss": 0.0184, "step": 52830 }, { "grad_norm": 0.2472536861896515, "learning_rate": 4.9439485472505016e-05, "loss": 0.0174, "step": 52840 }, { "grad_norm": 0.2398626059293747, "learning_rate": 4.942295184427356e-05, "loss": 0.0181, "step": 52850 }, { "grad_norm": 0.18870869278907776, "learning_rate": 4.9406418279147234e-05, "loss": 0.0151, "step": 52860 }, { "grad_norm": 0.1841411292552948, "learning_rate": 4.9389884778934116e-05, "loss": 0.0185, "step": 52870 }, { "grad_norm": 0.16302385926246643, "learning_rate": 4.937335134544228e-05, "loss": 0.0184, "step": 52880 }, { "grad_norm": 0.246585413813591, "learning_rate": 4.935681798047979e-05, "loss": 0.0177, "step": 52890 }, { "grad_norm": 0.19002754986286163, "learning_rate": 4.934028468585473e-05, "loss": 0.0169, "step": 52900 }, { "grad_norm": 0.18593814969062805, "learning_rate": 4.932375146337514e-05, "loss": 0.0194, "step": 52910 }, { "grad_norm": 0.23059649765491486, "learning_rate": 4.930721831484908e-05, "loss": 0.0172, "step": 52920 }, { "grad_norm": 0.19580736756324768, "learning_rate": 4.929068524208456e-05, "loss": 0.0171, "step": 52930 }, { "grad_norm": 0.2661670446395874, "learning_rate": 4.927415224688965e-05, "loss": 0.0174, "step": 52940 }, { "grad_norm": 0.14808055758476257, "learning_rate": 4.9257619331072375e-05, "loss": 0.0174, "step": 52950 }, { "grad_norm": 0.19824334979057312, "learning_rate": 4.924108649644072e-05, "loss": 0.0194, "step": 52960 }, { "grad_norm": 0.2166716307401657, "learning_rate": 4.922455374480271e-05, "loss": 0.0201, "step": 52970 }, { "grad_norm": 0.19680510461330414, "learning_rate": 4.920802107796634e-05, "loss": 0.0175, "step": 52980 }, { "grad_norm": 0.17833702266216278, "learning_rate": 4.9191488497739564e-05, "loss": 0.0172, "step": 52990 }, { "grad_norm": 0.22354286909103394, "learning_rate": 4.917495600593042e-05, "loss": 0.0166, "step": 53000 }, { "grad_norm": 0.20392993092536926, "learning_rate": 4.915842360434684e-05, "loss": 0.0166, "step": 53010 }, { "grad_norm": 0.17953236401081085, "learning_rate": 4.914189129479679e-05, "loss": 0.0179, "step": 53020 }, { "grad_norm": 0.15854115784168243, "learning_rate": 4.912535907908822e-05, "loss": 0.0185, "step": 53030 }, { "grad_norm": 0.17592114210128784, "learning_rate": 4.910882695902907e-05, "loss": 0.0176, "step": 53040 }, { "grad_norm": 0.13906192779541016, "learning_rate": 4.9092294936427234e-05, "loss": 0.0198, "step": 53050 }, { "grad_norm": 0.16591715812683105, "learning_rate": 4.9075763013090685e-05, "loss": 0.0147, "step": 53060 }, { "grad_norm": 0.20371443033218384, "learning_rate": 4.9059231190827305e-05, "loss": 0.0169, "step": 53070 }, { "grad_norm": 0.18217237293720245, "learning_rate": 4.904269947144499e-05, "loss": 0.0237, "step": 53080 }, { "grad_norm": 0.1480473428964615, "learning_rate": 4.902616785675161e-05, "loss": 0.0187, "step": 53090 }, { "grad_norm": 0.1676933914422989, "learning_rate": 4.900963634855505e-05, "loss": 0.0161, "step": 53100 }, { "grad_norm": 0.22995635867118835, "learning_rate": 4.8993104948663173e-05, "loss": 0.0189, "step": 53110 }, { "grad_norm": 0.1966099888086319, "learning_rate": 4.8976573658883823e-05, "loss": 0.0168, "step": 53120 }, { "grad_norm": 0.14911797642707825, "learning_rate": 4.896004248102484e-05, "loss": 0.0194, "step": 53130 }, { "grad_norm": 0.16331756114959717, "learning_rate": 4.8943511416894045e-05, "loss": 0.0231, "step": 53140 }, { "grad_norm": 0.1834547519683838, "learning_rate": 4.892698046829926e-05, "loss": 0.0199, "step": 53150 }, { "grad_norm": 0.17393052577972412, "learning_rate": 4.8910449637048275e-05, "loss": 0.0181, "step": 53160 }, { "grad_norm": 0.28941041231155396, "learning_rate": 4.8893918924948884e-05, "loss": 0.0189, "step": 53170 }, { "grad_norm": 0.20556098222732544, "learning_rate": 4.8877388333808835e-05, "loss": 0.0188, "step": 53180 }, { "grad_norm": 0.2104291468858719, "learning_rate": 4.8860857865435925e-05, "loss": 0.0152, "step": 53190 }, { "grad_norm": 0.21816790103912354, "learning_rate": 4.884432752163789e-05, "loss": 0.0177, "step": 53200 }, { "grad_norm": 0.18830563127994537, "learning_rate": 4.8827797304222466e-05, "loss": 0.022, "step": 53210 }, { "grad_norm": 0.1977497935295105, "learning_rate": 4.8811267214997366e-05, "loss": 0.0183, "step": 53220 }, { "grad_norm": 0.24616864323616028, "learning_rate": 4.879473725577029e-05, "loss": 0.0237, "step": 53230 }, { "grad_norm": 0.21882224082946777, "learning_rate": 4.877820742834891e-05, "loss": 0.0144, "step": 53240 }, { "grad_norm": 0.21175631880760193, "learning_rate": 4.8761677734540956e-05, "loss": 0.0188, "step": 53250 }, { "grad_norm": 0.21182215213775635, "learning_rate": 4.874514817615407e-05, "loss": 0.0189, "step": 53260 }, { "grad_norm": 0.1935834139585495, "learning_rate": 4.872861875499588e-05, "loss": 0.0165, "step": 53270 }, { "grad_norm": 0.19587168097496033, "learning_rate": 4.871208947287404e-05, "loss": 0.0175, "step": 53280 }, { "grad_norm": 0.1903376579284668, "learning_rate": 4.869556033159615e-05, "loss": 0.0149, "step": 53290 }, { "grad_norm": 0.1631786972284317, "learning_rate": 4.8679031332969835e-05, "loss": 0.0164, "step": 53300 }, { "grad_norm": 0.18715904653072357, "learning_rate": 4.8662502478802635e-05, "loss": 0.0174, "step": 53310 }, { "grad_norm": 0.20091862976551056, "learning_rate": 4.8645973770902176e-05, "loss": 0.0191, "step": 53320 }, { "grad_norm": 0.2626998722553253, "learning_rate": 4.862944521107599e-05, "loss": 0.0182, "step": 53330 }, { "grad_norm": 0.2603646218776703, "learning_rate": 4.861291680113162e-05, "loss": 0.0155, "step": 53340 }, { "grad_norm": 0.2261040210723877, "learning_rate": 4.8596388542876563e-05, "loss": 0.0157, "step": 53350 }, { "grad_norm": 0.2040701061487198, "learning_rate": 4.857986043811836e-05, "loss": 0.0158, "step": 53360 }, { "grad_norm": 0.1839853674173355, "learning_rate": 4.856333248866448e-05, "loss": 0.0179, "step": 53370 }, { "grad_norm": 0.21264582872390747, "learning_rate": 4.85468046963224e-05, "loss": 0.0162, "step": 53380 }, { "grad_norm": 0.17784732580184937, "learning_rate": 4.853027706289957e-05, "loss": 0.0189, "step": 53390 }, { "grad_norm": 0.19267891347408295, "learning_rate": 4.851374959020344e-05, "loss": 0.0172, "step": 53400 }, { "grad_norm": 0.27514591813087463, "learning_rate": 4.849722228004141e-05, "loss": 0.0226, "step": 53410 }, { "grad_norm": 0.2191624790430069, "learning_rate": 4.848069513422091e-05, "loss": 0.0186, "step": 53420 }, { "grad_norm": 0.14809425175189972, "learning_rate": 4.8464168154549294e-05, "loss": 0.0177, "step": 53430 }, { "grad_norm": 0.18292167782783508, "learning_rate": 4.844764134283392e-05, "loss": 0.016, "step": 53440 }, { "grad_norm": 0.12523864209651947, "learning_rate": 4.843111470088218e-05, "loss": 0.0171, "step": 53450 }, { "grad_norm": 0.13978737592697144, "learning_rate": 4.841458823050137e-05, "loss": 0.0151, "step": 53460 }, { "grad_norm": 0.18332667648792267, "learning_rate": 4.8398061933498816e-05, "loss": 0.015, "step": 53470 }, { "grad_norm": 0.17215608060359955, "learning_rate": 4.83815358116818e-05, "loss": 0.013, "step": 53480 }, { "grad_norm": 0.1302889883518219, "learning_rate": 4.8365009866857584e-05, "loss": 0.0157, "step": 53490 }, { "grad_norm": 0.1964782327413559, "learning_rate": 4.834848410083342e-05, "loss": 0.0163, "step": 53500 }, { "grad_norm": 0.15692070126533508, "learning_rate": 4.833195851541657e-05, "loss": 0.0202, "step": 53510 }, { "grad_norm": 0.1955311894416809, "learning_rate": 4.831543311241423e-05, "loss": 0.0177, "step": 53520 }, { "grad_norm": 0.16947141289710999, "learning_rate": 4.829890789363359e-05, "loss": 0.0173, "step": 53530 }, { "grad_norm": 0.28326231241226196, "learning_rate": 4.828238286088182e-05, "loss": 0.0174, "step": 53540 }, { "grad_norm": 0.24252715706825256, "learning_rate": 4.826585801596609e-05, "loss": 0.0218, "step": 53550 }, { "grad_norm": 0.2240571826696396, "learning_rate": 4.8249333360693486e-05, "loss": 0.0193, "step": 53560 }, { "grad_norm": 0.17785310745239258, "learning_rate": 4.8232808896871185e-05, "loss": 0.0183, "step": 53570 }, { "grad_norm": 0.17215517163276672, "learning_rate": 4.821628462630624e-05, "loss": 0.0145, "step": 53580 }, { "grad_norm": 0.13457448780536652, "learning_rate": 4.8199760550805735e-05, "loss": 0.0173, "step": 53590 }, { "grad_norm": 0.1576177179813385, "learning_rate": 4.818323667217669e-05, "loss": 0.0187, "step": 53600 }, { "grad_norm": 0.1851513832807541, "learning_rate": 4.8166712992226165e-05, "loss": 0.0167, "step": 53610 }, { "grad_norm": 0.15075673162937164, "learning_rate": 4.8150189512761154e-05, "loss": 0.0147, "step": 53620 }, { "grad_norm": 0.16996556520462036, "learning_rate": 4.813366623558862e-05, "loss": 0.0198, "step": 53630 }, { "grad_norm": 0.1754135638475418, "learning_rate": 4.811714316251554e-05, "loss": 0.0188, "step": 53640 }, { "grad_norm": 0.15926125645637512, "learning_rate": 4.810062029534886e-05, "loss": 0.0176, "step": 53650 }, { "grad_norm": 0.25397789478302, "learning_rate": 4.8084097635895505e-05, "loss": 0.0194, "step": 53660 }, { "grad_norm": 0.19201263785362244, "learning_rate": 4.8067575185962335e-05, "loss": 0.0185, "step": 53670 }, { "grad_norm": 0.1659916341304779, "learning_rate": 4.805105294735625e-05, "loss": 0.0195, "step": 53680 }, { "grad_norm": 0.24712440371513367, "learning_rate": 4.803453092188406e-05, "loss": 0.0189, "step": 53690 }, { "grad_norm": 0.17217500507831573, "learning_rate": 4.801800911135263e-05, "loss": 0.0161, "step": 53700 }, { "grad_norm": 0.2508419454097748, "learning_rate": 4.8001487517568744e-05, "loss": 0.0186, "step": 53710 }, { "grad_norm": 0.2043052762746811, "learning_rate": 4.798496614233918e-05, "loss": 0.0158, "step": 53720 }, { "grad_norm": 0.14731566607952118, "learning_rate": 4.796844498747069e-05, "loss": 0.0196, "step": 53730 }, { "grad_norm": 0.21357400715351105, "learning_rate": 4.795192405477e-05, "loss": 0.0188, "step": 53740 }, { "grad_norm": 0.18442344665527344, "learning_rate": 4.79354033460438e-05, "loss": 0.0216, "step": 53750 }, { "grad_norm": 0.18267808854579926, "learning_rate": 4.79188828630988e-05, "loss": 0.0174, "step": 53760 }, { "grad_norm": 0.15810444951057434, "learning_rate": 4.790236260774165e-05, "loss": 0.016, "step": 53770 }, { "grad_norm": 0.37291792035102844, "learning_rate": 4.788584258177896e-05, "loss": 0.0246, "step": 53780 }, { "grad_norm": 0.21580813825130463, "learning_rate": 4.7869322787017356e-05, "loss": 0.0185, "step": 53790 }, { "grad_norm": 0.1645013839006424, "learning_rate": 4.78528032252634e-05, "loss": 0.0178, "step": 53800 }, { "grad_norm": 0.21804121136665344, "learning_rate": 4.783628389832366e-05, "loss": 0.0155, "step": 53810 }, { "grad_norm": 0.26401281356811523, "learning_rate": 4.7819764808004644e-05, "loss": 0.019, "step": 53820 }, { "grad_norm": 0.23017175495624542, "learning_rate": 4.780324595611289e-05, "loss": 0.0176, "step": 53830 }, { "grad_norm": 0.2756394147872925, "learning_rate": 4.7786727344454854e-05, "loss": 0.0198, "step": 53840 }, { "grad_norm": 0.1501712054014206, "learning_rate": 4.777020897483697e-05, "loss": 0.0164, "step": 53850 }, { "grad_norm": 0.24660009145736694, "learning_rate": 4.775369084906569e-05, "loss": 0.0185, "step": 53860 }, { "grad_norm": 0.2548285722732544, "learning_rate": 4.773717296894742e-05, "loss": 0.0175, "step": 53870 }, { "grad_norm": 0.2240234911441803, "learning_rate": 4.772065533628848e-05, "loss": 0.018, "step": 53880 }, { "grad_norm": 0.2797035574913025, "learning_rate": 4.7704137952895265e-05, "loss": 0.0224, "step": 53890 }, { "grad_norm": 0.23049837350845337, "learning_rate": 4.768762082057407e-05, "loss": 0.0204, "step": 53900 }, { "grad_norm": 0.19567380845546722, "learning_rate": 4.7671103941131186e-05, "loss": 0.0195, "step": 53910 }, { "grad_norm": 0.2823897898197174, "learning_rate": 4.765458731637287e-05, "loss": 0.0203, "step": 53920 }, { "grad_norm": 0.19903579354286194, "learning_rate": 4.763807094810537e-05, "loss": 0.0176, "step": 53930 }, { "grad_norm": 0.24042914807796478, "learning_rate": 4.762155483813487e-05, "loss": 0.0181, "step": 53940 }, { "grad_norm": 0.19846314191818237, "learning_rate": 4.760503898826754e-05, "loss": 0.0178, "step": 53950 }, { "grad_norm": 0.18328604102134705, "learning_rate": 4.758852340030957e-05, "loss": 0.0181, "step": 53960 }, { "grad_norm": 0.15986384451389313, "learning_rate": 4.757200807606705e-05, "loss": 0.0167, "step": 53970 }, { "grad_norm": 0.25525397062301636, "learning_rate": 4.755549301734607e-05, "loss": 0.0185, "step": 53980 }, { "grad_norm": 0.18938681483268738, "learning_rate": 4.753897822595271e-05, "loss": 0.0171, "step": 53990 }, { "grad_norm": 0.1384282410144806, "learning_rate": 4.7522463703692994e-05, "loss": 0.0143, "step": 54000 }, { "grad_norm": 0.15474659204483032, "learning_rate": 4.75059494523729e-05, "loss": 0.0167, "step": 54010 }, { "grad_norm": 0.20636212825775146, "learning_rate": 4.748943547379843e-05, "loss": 0.0192, "step": 54020 }, { "grad_norm": 0.13804523646831512, "learning_rate": 4.7472921769775535e-05, "loss": 0.0158, "step": 54030 }, { "grad_norm": 0.21134616434574127, "learning_rate": 4.745640834211011e-05, "loss": 0.0179, "step": 54040 }, { "grad_norm": 0.16407904028892517, "learning_rate": 4.7439895192608054e-05, "loss": 0.0175, "step": 54050 }, { "grad_norm": 0.16398704051971436, "learning_rate": 4.7423382323075186e-05, "loss": 0.014, "step": 54060 }, { "grad_norm": 0.1805027723312378, "learning_rate": 4.740686973531737e-05, "loss": 0.0137, "step": 54070 }, { "grad_norm": 0.16742950677871704, "learning_rate": 4.7390357431140375e-05, "loss": 0.0161, "step": 54080 }, { "grad_norm": 0.1572987586259842, "learning_rate": 4.737384541234998e-05, "loss": 0.0161, "step": 54090 }, { "grad_norm": 0.17204560339450836, "learning_rate": 4.735733368075189e-05, "loss": 0.0169, "step": 54100 }, { "grad_norm": 0.25295424461364746, "learning_rate": 4.734082223815182e-05, "loss": 0.0176, "step": 54110 }, { "grad_norm": 0.17548520863056183, "learning_rate": 4.7324311086355436e-05, "loss": 0.018, "step": 54120 }, { "grad_norm": 0.18492260575294495, "learning_rate": 4.730780022716836e-05, "loss": 0.0199, "step": 54130 }, { "grad_norm": 0.16193103790283203, "learning_rate": 4.729128966239619e-05, "loss": 0.0152, "step": 54140 }, { "grad_norm": 0.19486385583877563, "learning_rate": 4.7274779393844536e-05, "loss": 0.0163, "step": 54150 }, { "grad_norm": 0.17699405550956726, "learning_rate": 4.725826942331891e-05, "loss": 0.0188, "step": 54160 }, { "grad_norm": 0.19149792194366455, "learning_rate": 4.724175975262481e-05, "loss": 0.019, "step": 54170 }, { "grad_norm": 0.204043909907341, "learning_rate": 4.722525038356772e-05, "loss": 0.0166, "step": 54180 }, { "grad_norm": 0.23159541189670563, "learning_rate": 4.7208741317953075e-05, "loss": 0.0166, "step": 54190 }, { "grad_norm": 0.2072705328464508, "learning_rate": 4.7192232557586264e-05, "loss": 0.0188, "step": 54200 }, { "grad_norm": 0.1530861109495163, "learning_rate": 4.717572410427269e-05, "loss": 0.0161, "step": 54210 }, { "grad_norm": 0.23433983325958252, "learning_rate": 4.71592159598177e-05, "loss": 0.0183, "step": 54220 }, { "grad_norm": 0.19046473503112793, "learning_rate": 4.714270812602657e-05, "loss": 0.0175, "step": 54230 }, { "grad_norm": 0.19208426773548126, "learning_rate": 4.712620060470458e-05, "loss": 0.016, "step": 54240 }, { "grad_norm": 0.1534605175256729, "learning_rate": 4.7109693397656986e-05, "loss": 0.0172, "step": 54250 }, { "grad_norm": 0.13185147941112518, "learning_rate": 4.709318650668894e-05, "loss": 0.0179, "step": 54260 }, { "grad_norm": 0.2193194031715393, "learning_rate": 4.7076679933605676e-05, "loss": 0.0162, "step": 54270 }, { "grad_norm": 0.15677432715892792, "learning_rate": 4.70601736802123e-05, "loss": 0.0165, "step": 54280 }, { "grad_norm": 0.2711458206176758, "learning_rate": 4.704366774831391e-05, "loss": 0.0229, "step": 54290 }, { "grad_norm": 0.17814037203788757, "learning_rate": 4.702716213971557e-05, "loss": 0.0145, "step": 54300 }, { "grad_norm": 0.23687930405139923, "learning_rate": 4.70106568562223e-05, "loss": 0.0196, "step": 54310 }, { "grad_norm": 0.17832134664058685, "learning_rate": 4.6994151899639105e-05, "loss": 0.0167, "step": 54320 }, { "grad_norm": 0.20976994931697845, "learning_rate": 4.697764727177093e-05, "loss": 0.021, "step": 54330 }, { "grad_norm": 0.2229590117931366, "learning_rate": 4.696114297442272e-05, "loss": 0.0185, "step": 54340 }, { "grad_norm": 0.17194204032421112, "learning_rate": 4.694463900939933e-05, "loss": 0.0206, "step": 54350 }, { "grad_norm": 0.20938391983509064, "learning_rate": 4.692813537850564e-05, "loss": 0.0145, "step": 54360 }, { "grad_norm": 0.18792837858200073, "learning_rate": 4.6911632083546445e-05, "loss": 0.0201, "step": 54370 }, { "grad_norm": 0.14598779380321503, "learning_rate": 4.6895129126326525e-05, "loss": 0.0179, "step": 54380 }, { "grad_norm": 0.1654892861843109, "learning_rate": 4.687862650865059e-05, "loss": 0.0184, "step": 54390 }, { "grad_norm": 0.21969817578792572, "learning_rate": 4.686212423232339e-05, "loss": 0.0167, "step": 54400 }, { "grad_norm": 0.20994551479816437, "learning_rate": 4.6845622299149565e-05, "loss": 0.0183, "step": 54410 }, { "grad_norm": 0.17144915461540222, "learning_rate": 4.682912071093374e-05, "loss": 0.0145, "step": 54420 }, { "grad_norm": 0.22722776234149933, "learning_rate": 4.681261946948052e-05, "loss": 0.0135, "step": 54430 }, { "grad_norm": 0.20288249850273132, "learning_rate": 4.679611857659443e-05, "loss": 0.019, "step": 54440 }, { "grad_norm": 0.2098500281572342, "learning_rate": 4.6779618034079975e-05, "loss": 0.0167, "step": 54450 }, { "grad_norm": 0.17377981543540955, "learning_rate": 4.6763117843741675e-05, "loss": 0.0155, "step": 54460 }, { "grad_norm": 0.16706527769565582, "learning_rate": 4.674661800738394e-05, "loss": 0.0155, "step": 54470 }, { "grad_norm": 0.2050708383321762, "learning_rate": 4.673011852681117e-05, "loss": 0.0179, "step": 54480 }, { "grad_norm": 0.24176129698753357, "learning_rate": 4.671361940382772e-05, "loss": 0.0202, "step": 54490 }, { "grad_norm": 0.14362700283527374, "learning_rate": 4.669712064023791e-05, "loss": 0.0175, "step": 54500 }, { "grad_norm": 0.2419232577085495, "learning_rate": 4.668062223784602e-05, "loss": 0.0162, "step": 54510 }, { "grad_norm": 0.15689992904663086, "learning_rate": 4.6664124198456274e-05, "loss": 0.0151, "step": 54520 }, { "grad_norm": 0.19811700284481049, "learning_rate": 4.6647626523872905e-05, "loss": 0.0184, "step": 54530 }, { "grad_norm": 0.19459670782089233, "learning_rate": 4.6631129215900066e-05, "loss": 0.0159, "step": 54540 }, { "grad_norm": 0.19389916956424713, "learning_rate": 4.661463227634187e-05, "loss": 0.0178, "step": 54550 }, { "grad_norm": 0.1384541094303131, "learning_rate": 4.659813570700238e-05, "loss": 0.0151, "step": 54560 }, { "grad_norm": 0.16253653168678284, "learning_rate": 4.658163950968568e-05, "loss": 0.0158, "step": 54570 }, { "grad_norm": 0.24583876132965088, "learning_rate": 4.6565143686195726e-05, "loss": 0.0188, "step": 54580 }, { "grad_norm": 0.21919883787631989, "learning_rate": 4.654864823833649e-05, "loss": 0.0192, "step": 54590 }, { "grad_norm": 0.1655336171388626, "learning_rate": 4.65321531679119e-05, "loss": 0.0194, "step": 54600 }, { "grad_norm": 0.16869014501571655, "learning_rate": 4.6515658476725834e-05, "loss": 0.0178, "step": 54610 }, { "grad_norm": 0.1386573761701584, "learning_rate": 4.649916416658212e-05, "loss": 0.0164, "step": 54620 }, { "grad_norm": 0.2021525800228119, "learning_rate": 4.648267023928455e-05, "loss": 0.0183, "step": 54630 }, { "grad_norm": 0.2233256697654724, "learning_rate": 4.646617669663687e-05, "loss": 0.0221, "step": 54640 }, { "grad_norm": 0.17939721047878265, "learning_rate": 4.644968354044277e-05, "loss": 0.0162, "step": 54650 }, { "grad_norm": 0.18977245688438416, "learning_rate": 4.643319077250596e-05, "loss": 0.0174, "step": 54660 }, { "grad_norm": 0.2378995716571808, "learning_rate": 4.6416698394630055e-05, "loss": 0.0154, "step": 54670 }, { "grad_norm": 0.21055251359939575, "learning_rate": 4.640020640861862e-05, "loss": 0.0166, "step": 54680 }, { "grad_norm": 0.19247905910015106, "learning_rate": 4.6383714816275195e-05, "loss": 0.0168, "step": 54690 }, { "grad_norm": 0.1972452998161316, "learning_rate": 4.636722361940329e-05, "loss": 0.0165, "step": 54700 }, { "grad_norm": 0.18833237886428833, "learning_rate": 4.635073281980632e-05, "loss": 0.0161, "step": 54710 }, { "grad_norm": 0.13660472631454468, "learning_rate": 4.633424241928775e-05, "loss": 0.0137, "step": 54720 }, { "grad_norm": 0.15663428604602814, "learning_rate": 4.6317752419650914e-05, "loss": 0.0172, "step": 54730 }, { "grad_norm": 0.24018952250480652, "learning_rate": 4.630126282269914e-05, "loss": 0.0167, "step": 54740 }, { "grad_norm": 0.2098516970872879, "learning_rate": 4.628477363023569e-05, "loss": 0.0185, "step": 54750 }, { "grad_norm": 0.16965439915657043, "learning_rate": 4.626828484406383e-05, "loss": 0.0186, "step": 54760 }, { "grad_norm": 0.23184078931808472, "learning_rate": 4.625179646598669e-05, "loss": 0.0165, "step": 54770 }, { "grad_norm": 0.17893096804618835, "learning_rate": 4.6235308497807487e-05, "loss": 0.0171, "step": 54780 }, { "grad_norm": 0.2501043379306793, "learning_rate": 4.6218820941329276e-05, "loss": 0.0161, "step": 54790 }, { "grad_norm": 0.16942782700061798, "learning_rate": 4.620233379835513e-05, "loss": 0.0177, "step": 54800 }, { "grad_norm": 0.1679028570652008, "learning_rate": 4.618584707068804e-05, "loss": 0.0162, "step": 54810 }, { "grad_norm": 0.19572724401950836, "learning_rate": 4.616936076013099e-05, "loss": 0.0172, "step": 54820 }, { "grad_norm": 0.17641428112983704, "learning_rate": 4.615287486848688e-05, "loss": 0.0172, "step": 54830 }, { "grad_norm": 0.1913377344608307, "learning_rate": 4.6136389397558584e-05, "loss": 0.0184, "step": 54840 }, { "grad_norm": 0.15614913403987885, "learning_rate": 4.6119904349148935e-05, "loss": 0.0152, "step": 54850 }, { "grad_norm": 0.22142848372459412, "learning_rate": 4.6103419725060726e-05, "loss": 0.0223, "step": 54860 }, { "grad_norm": 0.218649759888649, "learning_rate": 4.608693552709667e-05, "loss": 0.0166, "step": 54870 }, { "grad_norm": 0.1946866363286972, "learning_rate": 4.607045175705947e-05, "loss": 0.0175, "step": 54880 }, { "grad_norm": 0.1771465390920639, "learning_rate": 4.605396841675176e-05, "loss": 0.0175, "step": 54890 }, { "grad_norm": 0.21460875868797302, "learning_rate": 4.6037485507976094e-05, "loss": 0.0192, "step": 54900 }, { "grad_norm": 0.14350759983062744, "learning_rate": 4.602100303253509e-05, "loss": 0.0136, "step": 54910 }, { "grad_norm": 0.14571939408779144, "learning_rate": 4.600452099223122e-05, "loss": 0.0175, "step": 54920 }, { "grad_norm": 0.17375923693180084, "learning_rate": 4.598803938886691e-05, "loss": 0.0189, "step": 54930 }, { "grad_norm": 0.17809681594371796, "learning_rate": 4.597155822424459e-05, "loss": 0.0169, "step": 54940 }, { "grad_norm": 0.19650575518608093, "learning_rate": 4.59550775001666e-05, "loss": 0.0173, "step": 54950 }, { "grad_norm": 0.20310676097869873, "learning_rate": 4.5938597218435234e-05, "loss": 0.0188, "step": 54960 }, { "grad_norm": 0.18028244376182556, "learning_rate": 4.592211738085279e-05, "loss": 0.0153, "step": 54970 }, { "grad_norm": 0.1998988538980484, "learning_rate": 4.590563798922146e-05, "loss": 0.018, "step": 54980 }, { "grad_norm": 0.2413291335105896, "learning_rate": 4.5889159045343404e-05, "loss": 0.0194, "step": 54990 }, { "grad_norm": 0.1561877280473709, "learning_rate": 4.5872680551020734e-05, "loss": 0.0165, "step": 55000 }, { "grad_norm": 0.20083077251911163, "learning_rate": 4.585620250805552e-05, "loss": 0.0189, "step": 55010 }, { "grad_norm": 0.3116714358329773, "learning_rate": 4.583972491824975e-05, "loss": 0.014, "step": 55020 }, { "grad_norm": 0.1822778880596161, "learning_rate": 4.582324778340541e-05, "loss": 0.0161, "step": 55030 }, { "grad_norm": 0.14805833995342255, "learning_rate": 4.5806771105324424e-05, "loss": 0.0154, "step": 55040 }, { "grad_norm": 0.14770106971263885, "learning_rate": 4.579029488580865e-05, "loss": 0.0149, "step": 55050 }, { "grad_norm": 0.2110375612974167, "learning_rate": 4.5773819126659876e-05, "loss": 0.0167, "step": 55060 }, { "grad_norm": 0.16304366290569305, "learning_rate": 4.575734382967991e-05, "loss": 0.0165, "step": 55070 }, { "grad_norm": 0.2454967349767685, "learning_rate": 4.574086899667044e-05, "loss": 0.0191, "step": 55080 }, { "grad_norm": 0.21035712957382202, "learning_rate": 4.5724394629433106e-05, "loss": 0.0158, "step": 55090 }, { "grad_norm": 0.15167824923992157, "learning_rate": 4.570792072976958e-05, "loss": 0.0173, "step": 55100 }, { "grad_norm": 0.2445133477449417, "learning_rate": 4.569144729948138e-05, "loss": 0.016, "step": 55110 }, { "grad_norm": 0.17823609709739685, "learning_rate": 4.5674974340370033e-05, "loss": 0.0152, "step": 55120 }, { "grad_norm": 0.21436281502246857, "learning_rate": 4.565850185423699e-05, "loss": 0.0174, "step": 55130 }, { "grad_norm": 0.2132333517074585, "learning_rate": 4.564202984288365e-05, "loss": 0.0191, "step": 55140 }, { "grad_norm": 0.19020815193653107, "learning_rate": 4.5625558308111356e-05, "loss": 0.0142, "step": 55150 }, { "grad_norm": 0.17791207134723663, "learning_rate": 4.560908725172145e-05, "loss": 0.0173, "step": 55160 }, { "grad_norm": 0.15431292355060577, "learning_rate": 4.559261667551516e-05, "loss": 0.0192, "step": 55170 }, { "grad_norm": 0.1651323437690735, "learning_rate": 4.5576146581293685e-05, "loss": 0.0181, "step": 55180 }, { "grad_norm": 0.15123586356639862, "learning_rate": 4.5559676970858164e-05, "loss": 0.0144, "step": 55190 }, { "grad_norm": 0.14482073485851288, "learning_rate": 4.5543207846009705e-05, "loss": 0.0189, "step": 55200 }, { "grad_norm": 0.20349833369255066, "learning_rate": 4.552673920854933e-05, "loss": 0.0193, "step": 55210 }, { "grad_norm": 0.15557095408439636, "learning_rate": 4.5510271060278006e-05, "loss": 0.0168, "step": 55220 }, { "grad_norm": 0.17779091000556946, "learning_rate": 4.5493803402996704e-05, "loss": 0.0183, "step": 55230 }, { "grad_norm": 0.18510378897190094, "learning_rate": 4.5477336238506295e-05, "loss": 0.0192, "step": 55240 }, { "grad_norm": 0.19699832797050476, "learning_rate": 4.546086956860759e-05, "loss": 0.0158, "step": 55250 }, { "grad_norm": 0.2018374800682068, "learning_rate": 4.5444403395101374e-05, "loss": 0.0195, "step": 55260 }, { "grad_norm": 0.19699275493621826, "learning_rate": 4.542793771978834e-05, "loss": 0.0158, "step": 55270 }, { "grad_norm": 0.2312488853931427, "learning_rate": 4.541147254446916e-05, "loss": 0.0157, "step": 55280 }, { "grad_norm": 0.1934431493282318, "learning_rate": 4.539500787094445e-05, "loss": 0.0149, "step": 55290 }, { "grad_norm": 0.15572714805603027, "learning_rate": 4.537854370101475e-05, "loss": 0.0168, "step": 55300 }, { "grad_norm": 0.2531445026397705, "learning_rate": 4.5362080036480566e-05, "loss": 0.0153, "step": 55310 }, { "grad_norm": 0.1793600618839264, "learning_rate": 4.534561687914235e-05, "loss": 0.0157, "step": 55320 }, { "grad_norm": 0.21811680495738983, "learning_rate": 4.5329154230800466e-05, "loss": 0.0159, "step": 55330 }, { "grad_norm": 0.2109404057264328, "learning_rate": 4.531269209325525e-05, "loss": 0.0151, "step": 55340 }, { "grad_norm": 0.18569959700107574, "learning_rate": 4.529623046830696e-05, "loss": 0.0171, "step": 55350 }, { "grad_norm": 0.20770342648029327, "learning_rate": 4.5279769357755866e-05, "loss": 0.0181, "step": 55360 }, { "grad_norm": 0.20015868544578552, "learning_rate": 4.5263308763402084e-05, "loss": 0.0184, "step": 55370 }, { "grad_norm": 0.15844525396823883, "learning_rate": 4.524684868704574e-05, "loss": 0.0135, "step": 55380 }, { "grad_norm": 0.18513160943984985, "learning_rate": 4.523038913048688e-05, "loss": 0.0178, "step": 55390 }, { "grad_norm": 0.1924169361591339, "learning_rate": 4.521393009552549e-05, "loss": 0.0175, "step": 55400 }, { "grad_norm": 0.17282508313655853, "learning_rate": 4.5197471583961484e-05, "loss": 0.016, "step": 55410 }, { "grad_norm": 0.18753841519355774, "learning_rate": 4.51810135975948e-05, "loss": 0.0155, "step": 55420 }, { "grad_norm": 0.15274222195148468, "learning_rate": 4.516455613822521e-05, "loss": 0.0144, "step": 55430 }, { "grad_norm": 0.18156647682189941, "learning_rate": 4.514809920765249e-05, "loss": 0.0149, "step": 55440 }, { "grad_norm": 0.20398777723312378, "learning_rate": 4.5131642807676356e-05, "loss": 0.0148, "step": 55450 }, { "grad_norm": 0.17622718214988708, "learning_rate": 4.511518694009644e-05, "loss": 0.0149, "step": 55460 }, { "grad_norm": 0.17227745056152344, "learning_rate": 4.509873160671231e-05, "loss": 0.015, "step": 55470 }, { "grad_norm": 0.1481364667415619, "learning_rate": 4.508227680932356e-05, "loss": 0.0152, "step": 55480 }, { "grad_norm": 0.23949499428272247, "learning_rate": 4.5065822549729605e-05, "loss": 0.021, "step": 55490 }, { "grad_norm": 0.17521044611930847, "learning_rate": 4.504936882972989e-05, "loss": 0.0166, "step": 55500 }, { "grad_norm": 0.20516613125801086, "learning_rate": 4.503291565112374e-05, "loss": 0.0127, "step": 55510 }, { "grad_norm": 0.2030566930770874, "learning_rate": 4.501646301571048e-05, "loss": 0.0156, "step": 55520 }, { "grad_norm": 0.17748717963695526, "learning_rate": 4.500001092528933e-05, "loss": 0.0178, "step": 55530 }, { "grad_norm": 0.21619758009910583, "learning_rate": 4.4983559381659455e-05, "loss": 0.0193, "step": 55540 }, { "grad_norm": 0.1332009732723236, "learning_rate": 4.496710838661999e-05, "loss": 0.0148, "step": 55550 }, { "grad_norm": 0.191020667552948, "learning_rate": 4.495065794196999e-05, "loss": 0.0149, "step": 55560 }, { "grad_norm": 0.34279805421829224, "learning_rate": 4.4934208049508455e-05, "loss": 0.0148, "step": 55570 }, { "grad_norm": 0.15520580112934113, "learning_rate": 4.4917758711034303e-05, "loss": 0.0154, "step": 55580 }, { "grad_norm": 0.18260392546653748, "learning_rate": 4.490130992834642e-05, "loss": 0.0171, "step": 55590 }, { "grad_norm": 0.16489803791046143, "learning_rate": 4.488486170324359e-05, "loss": 0.0153, "step": 55600 }, { "grad_norm": 0.21870942413806915, "learning_rate": 4.486841403752463e-05, "loss": 0.0168, "step": 55610 }, { "grad_norm": 0.14751331508159637, "learning_rate": 4.4851966932988174e-05, "loss": 0.0159, "step": 55620 }, { "grad_norm": 0.18789979815483093, "learning_rate": 4.4835520391432894e-05, "loss": 0.0219, "step": 55630 }, { "grad_norm": 0.1586848795413971, "learning_rate": 4.481907441465733e-05, "loss": 0.016, "step": 55640 }, { "grad_norm": 0.17257095873355865, "learning_rate": 4.480262900446e-05, "loss": 0.0181, "step": 55650 }, { "grad_norm": 0.16669467091560364, "learning_rate": 4.478618416263933e-05, "loss": 0.0146, "step": 55660 }, { "grad_norm": 0.19270604848861694, "learning_rate": 4.476973989099373e-05, "loss": 0.0156, "step": 55670 }, { "grad_norm": 0.15507832169532776, "learning_rate": 4.475329619132153e-05, "loss": 0.0135, "step": 55680 }, { "grad_norm": 0.26080140471458435, "learning_rate": 4.4736853065420956e-05, "loss": 0.0168, "step": 55690 }, { "grad_norm": 0.18293434381484985, "learning_rate": 4.472041051509023e-05, "loss": 0.0137, "step": 55700 }, { "grad_norm": 0.1669398695230484, "learning_rate": 4.470396854212746e-05, "loss": 0.0176, "step": 55710 }, { "grad_norm": 0.17856663465499878, "learning_rate": 4.468752714833072e-05, "loss": 0.0153, "step": 55720 }, { "grad_norm": 0.23896557092666626, "learning_rate": 4.467108633549803e-05, "loss": 0.0205, "step": 55730 }, { "grad_norm": 0.1626485139131546, "learning_rate": 4.4654646105427335e-05, "loss": 0.0177, "step": 55740 }, { "grad_norm": 0.19374941289424896, "learning_rate": 4.463820645991651e-05, "loss": 0.0168, "step": 55750 }, { "grad_norm": 0.1801111102104187, "learning_rate": 4.462176740076334e-05, "loss": 0.0144, "step": 55760 }, { "grad_norm": 0.16184209287166595, "learning_rate": 4.4605328929765624e-05, "loss": 0.0149, "step": 55770 }, { "grad_norm": 0.16890551149845123, "learning_rate": 4.458889104872102e-05, "loss": 0.017, "step": 55780 }, { "grad_norm": 0.3170180916786194, "learning_rate": 4.457245375942715e-05, "loss": 0.0156, "step": 55790 }, { "grad_norm": 0.1955980658531189, "learning_rate": 4.455601706368157e-05, "loss": 0.0192, "step": 55800 }, { "grad_norm": 0.14682745933532715, "learning_rate": 4.453958096328179e-05, "loss": 0.0168, "step": 55810 }, { "grad_norm": 0.1934705227613449, "learning_rate": 4.452314546002523e-05, "loss": 0.02, "step": 55820 }, { "grad_norm": 0.21162578463554382, "learning_rate": 4.4506710555709246e-05, "loss": 0.0161, "step": 55830 }, { "grad_norm": 0.21711979806423187, "learning_rate": 4.449027625213114e-05, "loss": 0.0148, "step": 55840 }, { "grad_norm": 0.1576644629240036, "learning_rate": 4.4473842551088135e-05, "loss": 0.0161, "step": 55850 }, { "grad_norm": 0.19445665180683136, "learning_rate": 4.445740945437738e-05, "loss": 0.0182, "step": 55860 }, { "grad_norm": 0.23864686489105225, "learning_rate": 4.444097696379602e-05, "loss": 0.0154, "step": 55870 }, { "grad_norm": 0.17187219858169556, "learning_rate": 4.442454508114106e-05, "loss": 0.0166, "step": 55880 }, { "grad_norm": 0.18116147816181183, "learning_rate": 4.4408113808209465e-05, "loss": 0.0192, "step": 55890 }, { "grad_norm": 0.18298135697841644, "learning_rate": 4.439168314679813e-05, "loss": 0.0187, "step": 55900 }, { "grad_norm": 0.18802548944950104, "learning_rate": 4.43752530987039e-05, "loss": 0.0169, "step": 55910 }, { "grad_norm": 0.18206092715263367, "learning_rate": 4.4358823665723515e-05, "loss": 0.0157, "step": 55920 }, { "grad_norm": 0.21454671025276184, "learning_rate": 4.434239484965371e-05, "loss": 0.0181, "step": 55930 }, { "grad_norm": 0.13297981023788452, "learning_rate": 4.4325966652291103e-05, "loss": 0.0167, "step": 55940 }, { "grad_norm": 0.17815713584423065, "learning_rate": 4.430953907543225e-05, "loss": 0.0176, "step": 55950 }, { "grad_norm": 0.12893085181713104, "learning_rate": 4.4293112120873645e-05, "loss": 0.0145, "step": 55960 }, { "grad_norm": 0.16416890919208527, "learning_rate": 4.427668579041172e-05, "loss": 0.0157, "step": 55970 }, { "grad_norm": 0.20027676224708557, "learning_rate": 4.4260260085842816e-05, "loss": 0.0214, "step": 55980 }, { "grad_norm": 0.17876854538917542, "learning_rate": 4.4243835008963256e-05, "loss": 0.0173, "step": 55990 }, { "grad_norm": 0.18558788299560547, "learning_rate": 4.422741056156925e-05, "loss": 0.0174, "step": 56000 }, { "grad_norm": 0.15881562232971191, "learning_rate": 4.421098674545693e-05, "loss": 0.0162, "step": 56010 }, { "grad_norm": 0.20277062058448792, "learning_rate": 4.419456356242241e-05, "loss": 0.0185, "step": 56020 }, { "grad_norm": 0.22299611568450928, "learning_rate": 4.4178141014261685e-05, "loss": 0.0161, "step": 56030 }, { "grad_norm": 0.2221522033214569, "learning_rate": 4.416171910277071e-05, "loss": 0.017, "step": 56040 }, { "grad_norm": 0.18993504345417023, "learning_rate": 4.414529782974533e-05, "loss": 0.0204, "step": 56050 }, { "grad_norm": 0.12665528059005737, "learning_rate": 4.41288771969814e-05, "loss": 0.0148, "step": 56060 }, { "grad_norm": 0.17595362663269043, "learning_rate": 4.411245720627464e-05, "loss": 0.0172, "step": 56070 }, { "grad_norm": 0.19980685412883759, "learning_rate": 4.40960378594207e-05, "loss": 0.018, "step": 56080 }, { "grad_norm": 0.26004868745803833, "learning_rate": 4.407961915821519e-05, "loss": 0.0205, "step": 56090 }, { "grad_norm": 0.19267797470092773, "learning_rate": 4.406320110445362e-05, "loss": 0.0165, "step": 56100 }, { "grad_norm": 0.22019457817077637, "learning_rate": 4.404678369993144e-05, "loss": 0.0158, "step": 56110 }, { "grad_norm": 0.18750350177288055, "learning_rate": 4.403036694644406e-05, "loss": 0.0157, "step": 56120 }, { "grad_norm": 0.18179607391357422, "learning_rate": 4.4013950845786764e-05, "loss": 0.0136, "step": 56130 }, { "grad_norm": 0.15932713449001312, "learning_rate": 4.399753539975482e-05, "loss": 0.0171, "step": 56140 }, { "grad_norm": 0.15262271463871002, "learning_rate": 4.398112061014337e-05, "loss": 0.0109, "step": 56150 }, { "grad_norm": 0.16896963119506836, "learning_rate": 4.396470647874753e-05, "loss": 0.0184, "step": 56160 }, { "grad_norm": 0.1879718005657196, "learning_rate": 4.394829300736229e-05, "loss": 0.0133, "step": 56170 }, { "grad_norm": 0.16644689440727234, "learning_rate": 4.393188019778265e-05, "loss": 0.0116, "step": 56180 }, { "grad_norm": 0.16111084818840027, "learning_rate": 4.391546805180347e-05, "loss": 0.0149, "step": 56190 }, { "grad_norm": 0.19137470424175262, "learning_rate": 4.389905657121955e-05, "loss": 0.0202, "step": 56200 }, { "grad_norm": 0.15803895890712738, "learning_rate": 4.3882645757825635e-05, "loss": 0.0144, "step": 56210 }, { "grad_norm": 0.21412444114685059, "learning_rate": 4.386623561341637e-05, "loss": 0.0163, "step": 56220 }, { "grad_norm": 0.16154028475284576, "learning_rate": 4.384982613978637e-05, "loss": 0.013, "step": 56230 }, { "grad_norm": 0.22643712162971497, "learning_rate": 4.383341733873012e-05, "loss": 0.0156, "step": 56240 }, { "grad_norm": 0.13551685214042664, "learning_rate": 4.381700921204209e-05, "loss": 0.013, "step": 56250 }, { "grad_norm": 0.1611359566450119, "learning_rate": 4.380060176151663e-05, "loss": 0.0145, "step": 56260 }, { "grad_norm": 0.21447217464447021, "learning_rate": 4.378419498894805e-05, "loss": 0.0172, "step": 56270 }, { "grad_norm": 0.17114879190921783, "learning_rate": 4.376778889613056e-05, "loss": 0.0162, "step": 56280 }, { "grad_norm": 0.16046392917633057, "learning_rate": 4.375138348485831e-05, "loss": 0.014, "step": 56290 }, { "grad_norm": 0.2029641568660736, "learning_rate": 4.3734978756925346e-05, "loss": 0.014, "step": 56300 }, { "grad_norm": 0.18225271999835968, "learning_rate": 4.3718574714125706e-05, "loss": 0.0166, "step": 56310 }, { "grad_norm": 0.186872661113739, "learning_rate": 4.370217135825329e-05, "loss": 0.0159, "step": 56320 }, { "grad_norm": 0.2173527330160141, "learning_rate": 4.368576869110194e-05, "loss": 0.0175, "step": 56330 }, { "grad_norm": 0.18753787875175476, "learning_rate": 4.366936671446544e-05, "loss": 0.0176, "step": 56340 }, { "grad_norm": 0.16626954078674316, "learning_rate": 4.3652965430137474e-05, "loss": 0.0162, "step": 56350 }, { "grad_norm": 0.16870152950286865, "learning_rate": 4.3636564839911646e-05, "loss": 0.0154, "step": 56360 }, { "grad_norm": 0.17249442636966705, "learning_rate": 4.3620164945581545e-05, "loss": 0.0194, "step": 56370 }, { "grad_norm": 0.2412753701210022, "learning_rate": 4.360376574894061e-05, "loss": 0.0188, "step": 56380 }, { "grad_norm": 0.30130884051322937, "learning_rate": 4.358736725178224e-05, "loss": 0.0139, "step": 56390 }, { "grad_norm": 0.20764070749282837, "learning_rate": 4.357096945589974e-05, "loss": 0.017, "step": 56400 }, { "grad_norm": 0.14758910238742828, "learning_rate": 4.355457236308636e-05, "loss": 0.0146, "step": 56410 }, { "grad_norm": 0.1312585026025772, "learning_rate": 4.353817597513526e-05, "loss": 0.014, "step": 56420 }, { "grad_norm": 0.18109461665153503, "learning_rate": 4.352178029383948e-05, "loss": 0.017, "step": 56430 }, { "grad_norm": 0.1625305414199829, "learning_rate": 4.3505385320992105e-05, "loss": 0.0157, "step": 56440 }, { "grad_norm": 0.19075053930282593, "learning_rate": 4.348899105838602e-05, "loss": 0.0171, "step": 56450 }, { "grad_norm": 0.18653666973114014, "learning_rate": 4.3472597507814087e-05, "loss": 0.0146, "step": 56460 }, { "grad_norm": 0.1926196813583374, "learning_rate": 4.3456204671069066e-05, "loss": 0.0158, "step": 56470 }, { "grad_norm": 0.23422694206237793, "learning_rate": 4.343981254994367e-05, "loss": 0.02, "step": 56480 }, { "grad_norm": 0.2007894366979599, "learning_rate": 4.34234211462305e-05, "loss": 0.012, "step": 56490 }, { "grad_norm": 0.17163237929344177, "learning_rate": 4.340703046172213e-05, "loss": 0.0167, "step": 56500 }, { "grad_norm": 0.17199580371379852, "learning_rate": 4.339064049821097e-05, "loss": 0.0172, "step": 56510 }, { "grad_norm": 0.3194984495639801, "learning_rate": 4.3374251257489446e-05, "loss": 0.0146, "step": 56520 }, { "grad_norm": 0.17635370790958405, "learning_rate": 4.335786274134984e-05, "loss": 0.0149, "step": 56530 }, { "grad_norm": 0.2085781842470169, "learning_rate": 4.334147495158438e-05, "loss": 0.0155, "step": 56540 }, { "grad_norm": 0.1741902083158493, "learning_rate": 4.33250878899852e-05, "loss": 0.0164, "step": 56550 }, { "grad_norm": 0.1778208166360855, "learning_rate": 4.330870155834437e-05, "loss": 0.0164, "step": 56560 }, { "grad_norm": 0.22834093868732452, "learning_rate": 4.3292315958453884e-05, "loss": 0.015, "step": 56570 }, { "grad_norm": 0.18215259909629822, "learning_rate": 4.3275931092105646e-05, "loss": 0.0157, "step": 56580 }, { "grad_norm": 0.18366149067878723, "learning_rate": 4.325954696109147e-05, "loss": 0.016, "step": 56590 }, { "grad_norm": 0.19584928452968597, "learning_rate": 4.32431635672031e-05, "loss": 0.0154, "step": 56600 }, { "grad_norm": 0.26922905445098877, "learning_rate": 4.3226780912232214e-05, "loss": 0.018, "step": 56610 }, { "grad_norm": 0.23092757165431976, "learning_rate": 4.321039899797036e-05, "loss": 0.0173, "step": 56620 }, { "grad_norm": 0.13164210319519043, "learning_rate": 4.319401782620908e-05, "loss": 0.0155, "step": 56630 }, { "grad_norm": 0.23468756675720215, "learning_rate": 4.317763739873978e-05, "loss": 0.0217, "step": 56640 }, { "grad_norm": 0.15566396713256836, "learning_rate": 4.31612577173538e-05, "loss": 0.0153, "step": 56650 }, { "grad_norm": 0.13788703083992004, "learning_rate": 4.314487878384239e-05, "loss": 0.0163, "step": 56660 }, { "grad_norm": 0.16025207936763763, "learning_rate": 4.3128500599996736e-05, "loss": 0.0159, "step": 56670 }, { "grad_norm": 0.15757983922958374, "learning_rate": 4.31121231676079e-05, "loss": 0.0164, "step": 56680 }, { "grad_norm": 0.21537533402442932, "learning_rate": 4.309574648846694e-05, "loss": 0.0142, "step": 56690 }, { "grad_norm": 0.1923130601644516, "learning_rate": 4.3079370564364755e-05, "loss": 0.0182, "step": 56700 }, { "grad_norm": 0.14382705092430115, "learning_rate": 4.306299539709221e-05, "loss": 0.0166, "step": 56710 }, { "grad_norm": 0.14929544925689697, "learning_rate": 4.3046620988440046e-05, "loss": 0.0145, "step": 56720 }, { "grad_norm": 0.18655359745025635, "learning_rate": 4.303024734019897e-05, "loss": 0.0163, "step": 56730 }, { "grad_norm": 0.18324661254882812, "learning_rate": 4.3013874454159565e-05, "loss": 0.0147, "step": 56740 }, { "grad_norm": 0.1688399612903595, "learning_rate": 4.299750233211233e-05, "loss": 0.0174, "step": 56750 }, { "grad_norm": 0.20254656672477722, "learning_rate": 4.2981130975847715e-05, "loss": 0.016, "step": 56760 }, { "grad_norm": 0.19102874398231506, "learning_rate": 4.2964760387156076e-05, "loss": 0.0169, "step": 56770 }, { "grad_norm": 0.1995978057384491, "learning_rate": 4.294839056782767e-05, "loss": 0.0177, "step": 56780 }, { "grad_norm": 0.2619514763355255, "learning_rate": 4.293202151965266e-05, "loss": 0.0151, "step": 56790 }, { "grad_norm": 0.22809401154518127, "learning_rate": 4.291565324442115e-05, "loss": 0.0154, "step": 56800 }, { "grad_norm": 0.20055092871189117, "learning_rate": 4.2899285743923135e-05, "loss": 0.0209, "step": 56810 }, { "grad_norm": 0.1891431361436844, "learning_rate": 4.288291901994857e-05, "loss": 0.0159, "step": 56820 }, { "grad_norm": 0.20161138474941254, "learning_rate": 4.286655307428729e-05, "loss": 0.0174, "step": 56830 }, { "grad_norm": 0.1724945455789566, "learning_rate": 4.285018790872903e-05, "loss": 0.014, "step": 56840 }, { "grad_norm": 0.18902045488357544, "learning_rate": 4.283382352506347e-05, "loss": 0.0165, "step": 56850 }, { "grad_norm": 0.23775917291641235, "learning_rate": 4.2817459925080205e-05, "loss": 0.0166, "step": 56860 }, { "grad_norm": 0.17467163503170013, "learning_rate": 4.28010971105687e-05, "loss": 0.0167, "step": 56870 }, { "grad_norm": 0.17884166538715363, "learning_rate": 4.278473508331841e-05, "loss": 0.0169, "step": 56880 }, { "grad_norm": 0.2412363737821579, "learning_rate": 4.276837384511864e-05, "loss": 0.0146, "step": 56890 }, { "grad_norm": 0.22387638688087463, "learning_rate": 4.275201339775864e-05, "loss": 0.0161, "step": 56900 }, { "grad_norm": 0.16232310235500336, "learning_rate": 4.273565374302756e-05, "loss": 0.0131, "step": 56910 }, { "grad_norm": 0.2076660543680191, "learning_rate": 4.2719294882714454e-05, "loss": 0.0163, "step": 56920 }, { "grad_norm": 0.15681925415992737, "learning_rate": 4.270293681860831e-05, "loss": 0.0153, "step": 56930 }, { "grad_norm": 0.18189023435115814, "learning_rate": 4.268657955249802e-05, "loss": 0.0163, "step": 56940 }, { "grad_norm": 0.172752246260643, "learning_rate": 4.267022308617241e-05, "loss": 0.0155, "step": 56950 }, { "grad_norm": 0.1649588793516159, "learning_rate": 4.265386742142018e-05, "loss": 0.0178, "step": 56960 }, { "grad_norm": 0.2543222904205322, "learning_rate": 4.263751256002996e-05, "loss": 0.0143, "step": 56970 }, { "grad_norm": 0.23982460796833038, "learning_rate": 4.26211585037903e-05, "loss": 0.0179, "step": 56980 }, { "grad_norm": 0.15832000970840454, "learning_rate": 4.2604805254489656e-05, "loss": 0.0141, "step": 56990 }, { "grad_norm": 0.19227617979049683, "learning_rate": 4.258845281391638e-05, "loss": 0.0162, "step": 57000 }, { "grad_norm": 0.2081965208053589, "learning_rate": 4.257210118385876e-05, "loss": 0.0149, "step": 57010 }, { "grad_norm": 0.20450907945632935, "learning_rate": 4.2555750366104994e-05, "loss": 0.0147, "step": 57020 }, { "grad_norm": 0.22784964740276337, "learning_rate": 4.2539400362443184e-05, "loss": 0.0153, "step": 57030 }, { "grad_norm": 0.23117926716804504, "learning_rate": 4.252305117466133e-05, "loss": 0.0158, "step": 57040 }, { "grad_norm": 0.1927979588508606, "learning_rate": 4.2506702804547356e-05, "loss": 0.0165, "step": 57050 }, { "grad_norm": 0.20104199647903442, "learning_rate": 4.249035525388907e-05, "loss": 0.018, "step": 57060 }, { "grad_norm": 0.2271776646375656, "learning_rate": 4.247400852447427e-05, "loss": 0.0184, "step": 57070 }, { "grad_norm": 0.2425481081008911, "learning_rate": 4.245766261809059e-05, "loss": 0.0157, "step": 57080 }, { "grad_norm": 0.14797253906726837, "learning_rate": 4.244131753652558e-05, "loss": 0.0179, "step": 57090 }, { "grad_norm": 0.2649281322956085, "learning_rate": 4.242497328156672e-05, "loss": 0.0148, "step": 57100 }, { "grad_norm": 0.21578148007392883, "learning_rate": 4.240862985500139e-05, "loss": 0.0161, "step": 57110 }, { "grad_norm": 0.16328087449073792, "learning_rate": 4.2392287258616894e-05, "loss": 0.016, "step": 57120 }, { "grad_norm": 0.21940858662128448, "learning_rate": 4.2375945494200396e-05, "loss": 0.0138, "step": 57130 }, { "grad_norm": 0.16319099068641663, "learning_rate": 4.2359604563539054e-05, "loss": 0.0175, "step": 57140 }, { "grad_norm": 0.17614643275737762, "learning_rate": 4.2343264468419885e-05, "loss": 0.0155, "step": 57150 }, { "grad_norm": 0.1968332678079605, "learning_rate": 4.232692521062979e-05, "loss": 0.0157, "step": 57160 }, { "grad_norm": 0.44475623965263367, "learning_rate": 4.231058679195562e-05, "loss": 0.0181, "step": 57170 }, { "grad_norm": 0.18363316357135773, "learning_rate": 4.229424921418411e-05, "loss": 0.0163, "step": 57180 }, { "grad_norm": 0.19271306693553925, "learning_rate": 4.2277912479101916e-05, "loss": 0.0138, "step": 57190 }, { "grad_norm": 0.16372248530387878, "learning_rate": 4.226157658849562e-05, "loss": 0.0168, "step": 57200 }, { "grad_norm": 0.1323523223400116, "learning_rate": 4.2245241544151664e-05, "loss": 0.0126, "step": 57210 }, { "grad_norm": 0.18036609888076782, "learning_rate": 4.2228907347856425e-05, "loss": 0.0151, "step": 57220 }, { "grad_norm": 0.18230204284191132, "learning_rate": 4.2212574001396214e-05, "loss": 0.0172, "step": 57230 }, { "grad_norm": 0.174397274851799, "learning_rate": 4.21962415065572e-05, "loss": 0.012, "step": 57240 }, { "grad_norm": 0.21205444633960724, "learning_rate": 4.217990986512548e-05, "loss": 0.0204, "step": 57250 }, { "grad_norm": 0.1944502294063568, "learning_rate": 4.216357907888704e-05, "loss": 0.0162, "step": 57260 }, { "grad_norm": 0.14842118322849274, "learning_rate": 4.2147249149627824e-05, "loss": 0.0144, "step": 57270 }, { "grad_norm": 0.21798360347747803, "learning_rate": 4.213092007913364e-05, "loss": 0.0144, "step": 57280 }, { "grad_norm": 0.18761736154556274, "learning_rate": 4.2114591869190205e-05, "loss": 0.0173, "step": 57290 }, { "grad_norm": 0.13403740525245667, "learning_rate": 4.209826452158315e-05, "loss": 0.0136, "step": 57300 }, { "grad_norm": 0.16860266029834747, "learning_rate": 4.2081938038098e-05, "loss": 0.0131, "step": 57310 }, { "grad_norm": 0.2306678593158722, "learning_rate": 4.2065612420520185e-05, "loss": 0.0164, "step": 57320 }, { "grad_norm": 0.13638107478618622, "learning_rate": 4.204928767063509e-05, "loss": 0.0168, "step": 57330 }, { "grad_norm": 0.15583011507987976, "learning_rate": 4.2032963790227944e-05, "loss": 0.0135, "step": 57340 }, { "grad_norm": 0.2278028130531311, "learning_rate": 4.2016640781083887e-05, "loss": 0.0174, "step": 57350 }, { "grad_norm": 0.11647245287895203, "learning_rate": 4.2000318644988004e-05, "loss": 0.0164, "step": 57360 }, { "grad_norm": 0.17718741297721863, "learning_rate": 4.198399738372524e-05, "loss": 0.0148, "step": 57370 }, { "grad_norm": 0.20270605385303497, "learning_rate": 4.196767699908045e-05, "loss": 0.015, "step": 57380 }, { "grad_norm": 0.16459499299526215, "learning_rate": 4.195135749283845e-05, "loss": 0.0168, "step": 57390 }, { "grad_norm": 0.15340155363082886, "learning_rate": 4.1935038866783894e-05, "loss": 0.0145, "step": 57400 }, { "grad_norm": 0.2518381178379059, "learning_rate": 4.1918721122701355e-05, "loss": 0.0144, "step": 57410 }, { "grad_norm": 0.16075727343559265, "learning_rate": 4.190240426237533e-05, "loss": 0.0192, "step": 57420 }, { "grad_norm": 0.16255754232406616, "learning_rate": 4.188608828759019e-05, "loss": 0.013, "step": 57430 }, { "grad_norm": 0.11963466554880142, "learning_rate": 4.1869773200130246e-05, "loss": 0.0135, "step": 57440 }, { "grad_norm": 0.1734362244606018, "learning_rate": 4.185345900177967e-05, "loss": 0.017, "step": 57450 }, { "grad_norm": 0.16676901280879974, "learning_rate": 4.183714569432258e-05, "loss": 0.0149, "step": 57460 }, { "grad_norm": 0.16656647622585297, "learning_rate": 4.182083327954295e-05, "loss": 0.0151, "step": 57470 }, { "grad_norm": 0.13107509911060333, "learning_rate": 4.1804521759224714e-05, "loss": 0.0153, "step": 57480 }, { "grad_norm": 0.2415754795074463, "learning_rate": 4.178821113515165e-05, "loss": 0.0176, "step": 57490 }, { "grad_norm": 0.215780571103096, "learning_rate": 4.1771901409107476e-05, "loss": 0.0161, "step": 57500 }, { "grad_norm": 0.22515438497066498, "learning_rate": 4.175559258287577e-05, "loss": 0.0164, "step": 57510 }, { "grad_norm": 0.1533556431531906, "learning_rate": 4.173928465824008e-05, "loss": 0.0158, "step": 57520 }, { "grad_norm": 0.2367083579301834, "learning_rate": 4.17229776369838e-05, "loss": 0.0184, "step": 57530 }, { "grad_norm": 0.18484856188297272, "learning_rate": 4.170667152089025e-05, "loss": 0.0152, "step": 57540 }, { "grad_norm": 0.1727607250213623, "learning_rate": 4.1690366311742626e-05, "loss": 0.0167, "step": 57550 }, { "grad_norm": 0.15926474332809448, "learning_rate": 4.1674062011324055e-05, "loss": 0.017, "step": 57560 }, { "grad_norm": 0.1844191551208496, "learning_rate": 4.165775862141752e-05, "loss": 0.0148, "step": 57570 }, { "grad_norm": 0.2129463404417038, "learning_rate": 4.164145614380598e-05, "loss": 0.0156, "step": 57580 }, { "grad_norm": 0.15162771940231323, "learning_rate": 4.162515458027223e-05, "loss": 0.0131, "step": 57590 }, { "grad_norm": 0.15558339655399323, "learning_rate": 4.160885393259899e-05, "loss": 0.0145, "step": 57600 }, { "grad_norm": 0.20340265333652496, "learning_rate": 4.159255420256886e-05, "loss": 0.0137, "step": 57610 }, { "grad_norm": 0.196107417345047, "learning_rate": 4.157625539196436e-05, "loss": 0.015, "step": 57620 }, { "grad_norm": 0.22136470675468445, "learning_rate": 4.155995750256791e-05, "loss": 0.0201, "step": 57630 }, { "grad_norm": 0.16639603674411774, "learning_rate": 4.15436605361618e-05, "loss": 0.0188, "step": 57640 }, { "grad_norm": 0.19422627985477448, "learning_rate": 4.152736449452827e-05, "loss": 0.0134, "step": 57650 }, { "grad_norm": 0.18844576179981232, "learning_rate": 4.151106937944943e-05, "loss": 0.0138, "step": 57660 }, { "grad_norm": 0.15004709362983704, "learning_rate": 4.149477519270727e-05, "loss": 0.0203, "step": 57670 }, { "grad_norm": 0.19369909167289734, "learning_rate": 4.147848193608371e-05, "loss": 0.0143, "step": 57680 }, { "grad_norm": 0.12366102635860443, "learning_rate": 4.146218961136056e-05, "loss": 0.0139, "step": 57690 }, { "grad_norm": 0.14645123481750488, "learning_rate": 4.14458982203195e-05, "loss": 0.0157, "step": 57700 }, { "grad_norm": 0.17771518230438232, "learning_rate": 4.1429607764742165e-05, "loss": 0.0144, "step": 57710 }, { "grad_norm": 0.1903795450925827, "learning_rate": 4.141331824641003e-05, "loss": 0.0159, "step": 57720 }, { "grad_norm": 0.21789899468421936, "learning_rate": 4.139702966710452e-05, "loss": 0.0162, "step": 57730 }, { "grad_norm": 0.16324295103549957, "learning_rate": 4.13807420286069e-05, "loss": 0.0163, "step": 57740 }, { "grad_norm": 0.16502045094966888, "learning_rate": 4.136445533269837e-05, "loss": 0.0153, "step": 57750 }, { "grad_norm": 0.19516314566135406, "learning_rate": 4.134816958116003e-05, "loss": 0.0145, "step": 57760 }, { "grad_norm": 0.17096324265003204, "learning_rate": 4.133188477577283e-05, "loss": 0.0131, "step": 57770 }, { "grad_norm": 0.3964226543903351, "learning_rate": 4.131560091831771e-05, "loss": 0.0194, "step": 57780 }, { "grad_norm": 0.13731302320957184, "learning_rate": 4.1299318010575405e-05, "loss": 0.015, "step": 57790 }, { "grad_norm": 0.14661118388175964, "learning_rate": 4.128303605432661e-05, "loss": 0.0173, "step": 57800 }, { "grad_norm": 0.17812760174274445, "learning_rate": 4.126675505135188e-05, "loss": 0.0151, "step": 57810 }, { "grad_norm": 0.1978854387998581, "learning_rate": 4.125047500343169e-05, "loss": 0.0161, "step": 57820 }, { "grad_norm": 0.1998102217912674, "learning_rate": 4.1234195912346376e-05, "loss": 0.014, "step": 57830 }, { "grad_norm": 0.20271413028240204, "learning_rate": 4.1217917779876235e-05, "loss": 0.0154, "step": 57840 }, { "grad_norm": 0.13617202639579773, "learning_rate": 4.120164060780141e-05, "loss": 0.0155, "step": 57850 }, { "grad_norm": 0.1642276793718338, "learning_rate": 4.1185364397901935e-05, "loss": 0.0162, "step": 57860 }, { "grad_norm": 0.19712264835834503, "learning_rate": 4.116908915195776e-05, "loss": 0.0155, "step": 57870 }, { "grad_norm": 0.1743093580007553, "learning_rate": 4.115281487174872e-05, "loss": 0.0127, "step": 57880 }, { "grad_norm": 0.17018097639083862, "learning_rate": 4.113654155905452e-05, "loss": 0.0156, "step": 57890 }, { "grad_norm": 0.2560611367225647, "learning_rate": 4.112026921565483e-05, "loss": 0.0168, "step": 57900 }, { "grad_norm": 0.3050134479999542, "learning_rate": 4.1103997843329147e-05, "loss": 0.0196, "step": 57910 }, { "grad_norm": 0.13751006126403809, "learning_rate": 4.108772744385689e-05, "loss": 0.0147, "step": 57920 }, { "grad_norm": 0.1730329543352127, "learning_rate": 4.1071458019017346e-05, "loss": 0.0159, "step": 57930 }, { "grad_norm": 0.21574413776397705, "learning_rate": 4.1055189570589734e-05, "loss": 0.0158, "step": 57940 }, { "grad_norm": 0.19163845479488373, "learning_rate": 4.103892210035315e-05, "loss": 0.016, "step": 57950 }, { "grad_norm": 0.14781509339809418, "learning_rate": 4.102265561008656e-05, "loss": 0.0143, "step": 57960 }, { "grad_norm": 0.14826518297195435, "learning_rate": 4.100639010156885e-05, "loss": 0.0148, "step": 57970 }, { "grad_norm": 0.17784832417964935, "learning_rate": 4.099012557657882e-05, "loss": 0.0174, "step": 57980 }, { "grad_norm": 0.20874784886837006, "learning_rate": 4.09738620368951e-05, "loss": 0.016, "step": 57990 }, { "grad_norm": 0.19558757543563843, "learning_rate": 4.095759948429626e-05, "loss": 0.016, "step": 58000 }, { "grad_norm": 0.18421602249145508, "learning_rate": 4.094133792056075e-05, "loss": 0.0154, "step": 58010 }, { "grad_norm": 0.44777846336364746, "learning_rate": 4.092507734746688e-05, "loss": 0.0182, "step": 58020 }, { "grad_norm": 0.1511603146791458, "learning_rate": 4.090881776679293e-05, "loss": 0.0153, "step": 58030 }, { "grad_norm": 0.1900521218776703, "learning_rate": 4.0892559180316994e-05, "loss": 0.0152, "step": 58040 }, { "grad_norm": 0.1387837529182434, "learning_rate": 4.08763015898171e-05, "loss": 0.0159, "step": 58050 }, { "grad_norm": 0.33414870500564575, "learning_rate": 4.0860044997071145e-05, "loss": 0.0142, "step": 58060 }, { "grad_norm": 0.25450366735458374, "learning_rate": 4.084378940385693e-05, "loss": 0.0189, "step": 58070 }, { "grad_norm": 0.15848198533058167, "learning_rate": 4.082753481195211e-05, "loss": 0.0203, "step": 58080 }, { "grad_norm": 0.17762643098831177, "learning_rate": 4.0811281223134324e-05, "loss": 0.0139, "step": 58090 }, { "grad_norm": 0.14379680156707764, "learning_rate": 4.0795028639181e-05, "loss": 0.0131, "step": 58100 }, { "grad_norm": 0.20534642040729523, "learning_rate": 4.077877706186951e-05, "loss": 0.017, "step": 58110 }, { "grad_norm": 0.18771713972091675, "learning_rate": 4.076252649297709e-05, "loss": 0.0152, "step": 58120 }, { "grad_norm": 0.18253183364868164, "learning_rate": 4.07462769342809e-05, "loss": 0.0172, "step": 58130 }, { "grad_norm": 0.19624900817871094, "learning_rate": 4.073002838755793e-05, "loss": 0.0171, "step": 58140 }, { "grad_norm": 0.16990987956523895, "learning_rate": 4.0713780854585115e-05, "loss": 0.0181, "step": 58150 }, { "grad_norm": 0.1382807195186615, "learning_rate": 4.069753433713927e-05, "loss": 0.0162, "step": 58160 }, { "grad_norm": 0.1452997475862503, "learning_rate": 4.068128883699709e-05, "loss": 0.0164, "step": 58170 }, { "grad_norm": 0.17693838477134705, "learning_rate": 4.0665044355935155e-05, "loss": 0.0165, "step": 58180 }, { "grad_norm": 0.17552775144577026, "learning_rate": 4.064880089572993e-05, "loss": 0.0159, "step": 58190 }, { "grad_norm": 0.1883007138967514, "learning_rate": 4.063255845815779e-05, "loss": 0.0168, "step": 58200 }, { "grad_norm": 0.15182051062583923, "learning_rate": 4.0616317044994945e-05, "loss": 0.0149, "step": 58210 }, { "grad_norm": 0.14751915633678436, "learning_rate": 4.0600076658017585e-05, "loss": 0.015, "step": 58220 }, { "grad_norm": 0.21192365884780884, "learning_rate": 4.0583837299001706e-05, "loss": 0.0138, "step": 58230 }, { "grad_norm": 0.24380958080291748, "learning_rate": 4.056759896972322e-05, "loss": 0.0172, "step": 58240 }, { "grad_norm": 0.1835029423236847, "learning_rate": 4.055136167195793e-05, "loss": 0.0191, "step": 58250 }, { "grad_norm": 0.1883900761604309, "learning_rate": 4.053512540748152e-05, "loss": 0.0139, "step": 58260 }, { "grad_norm": 0.1787940114736557, "learning_rate": 4.051889017806955e-05, "loss": 0.0172, "step": 58270 }, { "grad_norm": 0.14464759826660156, "learning_rate": 4.0502655985497516e-05, "loss": 0.0173, "step": 58280 }, { "grad_norm": 0.19760124385356903, "learning_rate": 4.048642283154074e-05, "loss": 0.0178, "step": 58290 }, { "grad_norm": 0.23231858015060425, "learning_rate": 4.0470190717974464e-05, "loss": 0.021, "step": 58300 }, { "grad_norm": 0.16109788417816162, "learning_rate": 4.045395964657381e-05, "loss": 0.0181, "step": 58310 }, { "grad_norm": 0.15842604637145996, "learning_rate": 4.043772961911376e-05, "loss": 0.0167, "step": 58320 }, { "grad_norm": 0.18090292811393738, "learning_rate": 4.042150063736924e-05, "loss": 0.0147, "step": 58330 }, { "grad_norm": 0.19258205592632294, "learning_rate": 4.0405272703114976e-05, "loss": 0.0166, "step": 58340 }, { "grad_norm": 0.19990111887454987, "learning_rate": 4.038904581812569e-05, "loss": 0.0155, "step": 58350 }, { "grad_norm": 0.14737363159656525, "learning_rate": 4.03728199841759e-05, "loss": 0.0152, "step": 58360 }, { "grad_norm": 0.1885708123445511, "learning_rate": 4.0356595203040044e-05, "loss": 0.0161, "step": 58370 }, { "grad_norm": 0.16724947094917297, "learning_rate": 4.0340371476492426e-05, "loss": 0.0169, "step": 58380 }, { "grad_norm": 0.3076263964176178, "learning_rate": 4.032414880630727e-05, "loss": 0.0171, "step": 58390 }, { "grad_norm": 0.1288694143295288, "learning_rate": 4.0307927194258646e-05, "loss": 0.0132, "step": 58400 }, { "grad_norm": 0.21639402210712433, "learning_rate": 4.029170664212054e-05, "loss": 0.0164, "step": 58410 }, { "grad_norm": 0.13966278731822968, "learning_rate": 4.0275487151666794e-05, "loss": 0.0145, "step": 58420 }, { "grad_norm": 0.12860746681690216, "learning_rate": 4.025926872467116e-05, "loss": 0.0187, "step": 58430 }, { "grad_norm": 0.22494439780712128, "learning_rate": 4.024305136290725e-05, "loss": 0.0132, "step": 58440 }, { "grad_norm": 0.15665502846240997, "learning_rate": 4.022683506814858e-05, "loss": 0.0178, "step": 58450 }, { "grad_norm": 0.2005234956741333, "learning_rate": 4.021061984216853e-05, "loss": 0.0159, "step": 58460 }, { "grad_norm": 0.18474189937114716, "learning_rate": 4.019440568674035e-05, "loss": 0.0184, "step": 58470 }, { "grad_norm": 0.1672171652317047, "learning_rate": 4.017819260363724e-05, "loss": 0.0157, "step": 58480 }, { "grad_norm": 0.2324082851409912, "learning_rate": 4.016198059463223e-05, "loss": 0.018, "step": 58490 }, { "grad_norm": 0.25290048122406006, "learning_rate": 4.0145769661498233e-05, "loss": 0.0152, "step": 58500 }, { "grad_norm": 0.16255129873752594, "learning_rate": 4.012955980600804e-05, "loss": 0.0158, "step": 58510 }, { "grad_norm": 0.21748748421669006, "learning_rate": 4.0113351029934355e-05, "loss": 0.013, "step": 58520 }, { "grad_norm": 0.21602623164653778, "learning_rate": 4.0097143335049714e-05, "loss": 0.0152, "step": 58530 }, { "grad_norm": 0.20135746896266937, "learning_rate": 4.0080936723126606e-05, "loss": 0.0134, "step": 58540 }, { "grad_norm": 0.17405278980731964, "learning_rate": 4.006473119593735e-05, "loss": 0.0138, "step": 58550 }, { "grad_norm": 0.1323031634092331, "learning_rate": 4.004852675525415e-05, "loss": 0.0165, "step": 58560 }, { "grad_norm": 0.2425156533718109, "learning_rate": 4.00323234028491e-05, "loss": 0.0178, "step": 58570 }, { "grad_norm": 0.14171478152275085, "learning_rate": 4.0016121140494186e-05, "loss": 0.017, "step": 58580 }, { "grad_norm": 0.20816244184970856, "learning_rate": 3.999991996996122e-05, "loss": 0.0153, "step": 58590 }, { "grad_norm": 0.17660623788833618, "learning_rate": 3.9983719893022e-05, "loss": 0.0186, "step": 58600 }, { "grad_norm": 0.1914881467819214, "learning_rate": 3.996752091144812e-05, "loss": 0.0148, "step": 58610 }, { "grad_norm": 0.17735710740089417, "learning_rate": 3.9951323027011055e-05, "loss": 0.0152, "step": 58620 }, { "grad_norm": 0.19530121982097626, "learning_rate": 3.993512624148219e-05, "loss": 0.0212, "step": 58630 }, { "grad_norm": 0.19892165064811707, "learning_rate": 3.99189305566328e-05, "loss": 0.0157, "step": 58640 }, { "grad_norm": 0.15521620213985443, "learning_rate": 3.990273597423401e-05, "loss": 0.0178, "step": 58650 }, { "grad_norm": 0.09019490331411362, "learning_rate": 3.988654249605681e-05, "loss": 0.013, "step": 58660 }, { "grad_norm": 0.19626650214195251, "learning_rate": 3.987035012387213e-05, "loss": 0.0149, "step": 58670 }, { "grad_norm": 0.1502145677804947, "learning_rate": 3.9854158859450735e-05, "loss": 0.0117, "step": 58680 }, { "grad_norm": 0.13006554543972015, "learning_rate": 3.983796870456328e-05, "loss": 0.016, "step": 58690 }, { "grad_norm": 0.14583416283130646, "learning_rate": 3.982177966098028e-05, "loss": 0.0165, "step": 58700 }, { "grad_norm": 0.20338225364685059, "learning_rate": 3.980559173047216e-05, "loss": 0.0155, "step": 58710 }, { "grad_norm": 0.179691880941391, "learning_rate": 3.9789404914809176e-05, "loss": 0.0173, "step": 58720 }, { "grad_norm": 0.1851033717393875, "learning_rate": 3.9773219215761544e-05, "loss": 0.0147, "step": 58730 }, { "grad_norm": 0.16926448047161102, "learning_rate": 3.975703463509928e-05, "loss": 0.0171, "step": 58740 }, { "grad_norm": 0.18225906789302826, "learning_rate": 3.974085117459231e-05, "loss": 0.0146, "step": 58750 }, { "grad_norm": 0.16452454030513763, "learning_rate": 3.972466883601043e-05, "loss": 0.0148, "step": 58760 }, { "grad_norm": 0.15097573399543762, "learning_rate": 3.9708487621123316e-05, "loss": 0.0146, "step": 58770 }, { "grad_norm": 0.15858028829097748, "learning_rate": 3.969230753170051e-05, "loss": 0.0158, "step": 58780 }, { "grad_norm": 0.2030188888311386, "learning_rate": 3.967612856951146e-05, "loss": 0.016, "step": 58790 }, { "grad_norm": 0.13123124837875366, "learning_rate": 3.9659950736325476e-05, "loss": 0.0152, "step": 58800 }, { "grad_norm": 0.25008368492126465, "learning_rate": 3.964377403391174e-05, "loss": 0.0138, "step": 58810 }, { "grad_norm": 0.15194250643253326, "learning_rate": 3.9627598464039303e-05, "loss": 0.0146, "step": 58820 }, { "grad_norm": 0.15938127040863037, "learning_rate": 3.9611424028477096e-05, "loss": 0.0132, "step": 58830 }, { "grad_norm": 0.16448161005973816, "learning_rate": 3.959525072899394e-05, "loss": 0.0148, "step": 58840 }, { "grad_norm": 0.16653209924697876, "learning_rate": 3.9579078567358506e-05, "loss": 0.0155, "step": 58850 }, { "grad_norm": 0.15369071066379547, "learning_rate": 3.956290754533939e-05, "loss": 0.0167, "step": 58860 }, { "grad_norm": 0.253670334815979, "learning_rate": 3.954673766470501e-05, "loss": 0.0139, "step": 58870 }, { "grad_norm": 0.17783567309379578, "learning_rate": 3.9530568927223674e-05, "loss": 0.0158, "step": 58880 }, { "grad_norm": 0.18181535601615906, "learning_rate": 3.95144013346636e-05, "loss": 0.0177, "step": 58890 }, { "grad_norm": 0.14115265011787415, "learning_rate": 3.949823488879282e-05, "loss": 0.0146, "step": 58900 }, { "grad_norm": 0.28537604212760925, "learning_rate": 3.9482069591379276e-05, "loss": 0.0174, "step": 58910 }, { "grad_norm": 0.14290408790111542, "learning_rate": 3.9465905444190784e-05, "loss": 0.0142, "step": 58920 }, { "grad_norm": 0.1531551033258438, "learning_rate": 3.944974244899505e-05, "loss": 0.014, "step": 58930 }, { "grad_norm": 0.2180839329957962, "learning_rate": 3.9433580607559616e-05, "loss": 0.0153, "step": 58940 }, { "grad_norm": 0.1768973469734192, "learning_rate": 3.941741992165193e-05, "loss": 0.0184, "step": 58950 }, { "grad_norm": 0.19599975645542145, "learning_rate": 3.940126039303928e-05, "loss": 0.0153, "step": 58960 }, { "grad_norm": 0.17924578487873077, "learning_rate": 3.938510202348886e-05, "loss": 0.013, "step": 58970 }, { "grad_norm": 0.1446959227323532, "learning_rate": 3.93689448147677e-05, "loss": 0.0141, "step": 58980 }, { "grad_norm": 0.2320672571659088, "learning_rate": 3.935278876864278e-05, "loss": 0.0136, "step": 58990 }, { "grad_norm": 0.19934628903865814, "learning_rate": 3.933663388688087e-05, "loss": 0.0158, "step": 59000 }, { "grad_norm": 0.23256371915340424, "learning_rate": 3.9320480171248645e-05, "loss": 0.0176, "step": 59010 }, { "grad_norm": 0.14252258837223053, "learning_rate": 3.930432762351265e-05, "loss": 0.0138, "step": 59020 }, { "grad_norm": 0.20822474360466003, "learning_rate": 3.928817624543931e-05, "loss": 0.0171, "step": 59030 }, { "grad_norm": 0.16830846667289734, "learning_rate": 3.9272026038794885e-05, "loss": 0.0165, "step": 59040 }, { "grad_norm": 0.12745754420757294, "learning_rate": 3.9255877005345586e-05, "loss": 0.0178, "step": 59050 }, { "grad_norm": 0.18347157537937164, "learning_rate": 3.923972914685743e-05, "loss": 0.0133, "step": 59060 }, { "grad_norm": 0.17472095787525177, "learning_rate": 3.922358246509631e-05, "loss": 0.0176, "step": 59070 }, { "grad_norm": 0.28334420919418335, "learning_rate": 3.920743696182801e-05, "loss": 0.0198, "step": 59080 }, { "grad_norm": 0.17466624081134796, "learning_rate": 3.9191292638818164e-05, "loss": 0.0139, "step": 59090 }, { "grad_norm": 0.20132827758789062, "learning_rate": 3.9175149497832316e-05, "loss": 0.0159, "step": 59100 }, { "grad_norm": 0.14019091427326202, "learning_rate": 3.915900754063584e-05, "loss": 0.0158, "step": 59110 }, { "grad_norm": 0.1666397601366043, "learning_rate": 3.9142866768994014e-05, "loss": 0.0169, "step": 59120 }, { "grad_norm": 0.15586744248867035, "learning_rate": 3.912672718467193e-05, "loss": 0.0136, "step": 59130 }, { "grad_norm": 0.19734884798526764, "learning_rate": 3.911058878943463e-05, "loss": 0.0143, "step": 59140 }, { "grad_norm": 0.14617329835891724, "learning_rate": 3.9094451585046974e-05, "loss": 0.0135, "step": 59150 }, { "grad_norm": 0.17193110287189484, "learning_rate": 3.90783155732737e-05, "loss": 0.0146, "step": 59160 }, { "grad_norm": 0.2073686271905899, "learning_rate": 3.9062180755879373e-05, "loss": 0.0153, "step": 59170 }, { "grad_norm": 0.18613381683826447, "learning_rate": 3.904604713462855e-05, "loss": 0.0127, "step": 59180 }, { "grad_norm": 0.15728430449962616, "learning_rate": 3.902991471128554e-05, "loss": 0.0171, "step": 59190 }, { "grad_norm": 0.2090148776769638, "learning_rate": 3.901378348761456e-05, "loss": 0.0165, "step": 59200 }, { "grad_norm": 0.1774360090494156, "learning_rate": 3.899765346537971e-05, "loss": 0.014, "step": 59210 }, { "grad_norm": 0.1693728119134903, "learning_rate": 3.898152464634493e-05, "loss": 0.0213, "step": 59220 }, { "grad_norm": 0.22677026689052582, "learning_rate": 3.896539703227402e-05, "loss": 0.018, "step": 59230 }, { "grad_norm": 0.15997730195522308, "learning_rate": 3.894927062493073e-05, "loss": 0.015, "step": 59240 }, { "grad_norm": 0.15397508442401886, "learning_rate": 3.893314542607859e-05, "loss": 0.0124, "step": 59250 }, { "grad_norm": 0.19762521982192993, "learning_rate": 3.891702143748103e-05, "loss": 0.0162, "step": 59260 }, { "grad_norm": 0.1760539561510086, "learning_rate": 3.890089866090133e-05, "loss": 0.0132, "step": 59270 }, { "grad_norm": 0.16772399842739105, "learning_rate": 3.888477709810267e-05, "loss": 0.0144, "step": 59280 }, { "grad_norm": 0.15412423014640808, "learning_rate": 3.886865675084806e-05, "loss": 0.0155, "step": 59290 }, { "grad_norm": 0.20584993064403534, "learning_rate": 3.885253762090043e-05, "loss": 0.0165, "step": 59300 }, { "grad_norm": 0.16129785776138306, "learning_rate": 3.8836419710022534e-05, "loss": 0.0151, "step": 59310 }, { "grad_norm": 0.1948269009590149, "learning_rate": 3.882030301997698e-05, "loss": 0.0163, "step": 59320 }, { "grad_norm": 0.21589651703834534, "learning_rate": 3.880418755252629e-05, "loss": 0.0139, "step": 59330 }, { "grad_norm": 0.2074528932571411, "learning_rate": 3.87880733094328e-05, "loss": 0.0155, "step": 59340 }, { "grad_norm": 0.19709758460521698, "learning_rate": 3.877196029245877e-05, "loss": 0.015, "step": 59350 }, { "grad_norm": 0.16861486434936523, "learning_rate": 3.875584850336627e-05, "loss": 0.0146, "step": 59360 }, { "grad_norm": 0.18739944696426392, "learning_rate": 3.873973794391728e-05, "loss": 0.016, "step": 59370 }, { "grad_norm": 0.14704132080078125, "learning_rate": 3.872362861587361e-05, "loss": 0.0146, "step": 59380 }, { "grad_norm": 0.2194628119468689, "learning_rate": 3.870752052099698e-05, "loss": 0.015, "step": 59390 }, { "grad_norm": 0.17450110614299774, "learning_rate": 3.8691413661048916e-05, "loss": 0.0167, "step": 59400 }, { "grad_norm": 0.21379122138023376, "learning_rate": 3.867530803779085e-05, "loss": 0.0168, "step": 59410 }, { "grad_norm": 0.20307616889476776, "learning_rate": 3.8659203652984054e-05, "loss": 0.014, "step": 59420 }, { "grad_norm": 0.21762984991073608, "learning_rate": 3.864310050838972e-05, "loss": 0.0146, "step": 59430 }, { "grad_norm": 0.13297730684280396, "learning_rate": 3.8626998605768835e-05, "loss": 0.0159, "step": 59440 }, { "grad_norm": 0.17604054510593414, "learning_rate": 3.8610897946882294e-05, "loss": 0.025, "step": 59450 }, { "grad_norm": 0.14567552506923676, "learning_rate": 3.859479853349082e-05, "loss": 0.0133, "step": 59460 }, { "grad_norm": 0.20659957826137543, "learning_rate": 3.857870036735503e-05, "loss": 0.013, "step": 59470 }, { "grad_norm": 0.18688201904296875, "learning_rate": 3.856260345023538e-05, "loss": 0.0132, "step": 59480 }, { "grad_norm": 0.1280452162027359, "learning_rate": 3.854650778389224e-05, "loss": 0.0161, "step": 59490 }, { "grad_norm": 0.1761830449104309, "learning_rate": 3.8530413370085786e-05, "loss": 0.0153, "step": 59500 }, { "grad_norm": 0.20134364068508148, "learning_rate": 3.851432021057608e-05, "loss": 0.0153, "step": 59510 }, { "grad_norm": 0.1549859642982483, "learning_rate": 3.849822830712306e-05, "loss": 0.0187, "step": 59520 }, { "grad_norm": 0.22090381383895874, "learning_rate": 3.848213766148649e-05, "loss": 0.0183, "step": 59530 }, { "grad_norm": 0.1535910665988922, "learning_rate": 3.846604827542602e-05, "loss": 0.0171, "step": 59540 }, { "grad_norm": 0.15508918464183807, "learning_rate": 3.844996015070117e-05, "loss": 0.0138, "step": 59550 }, { "grad_norm": 0.17146387696266174, "learning_rate": 3.843387328907132e-05, "loss": 0.0156, "step": 59560 }, { "grad_norm": 0.23476018011569977, "learning_rate": 3.8417787692295696e-05, "loss": 0.0163, "step": 59570 }, { "grad_norm": 0.15520672500133514, "learning_rate": 3.84017033621334e-05, "loss": 0.0134, "step": 59580 }, { "grad_norm": 0.29750487208366394, "learning_rate": 3.838562030034338e-05, "loss": 0.0171, "step": 59590 }, { "grad_norm": 0.21131832897663116, "learning_rate": 3.836953850868447e-05, "loss": 0.0202, "step": 59600 }, { "grad_norm": 0.21839341521263123, "learning_rate": 3.835345798891533e-05, "loss": 0.0143, "step": 59610 }, { "grad_norm": 0.20022673904895782, "learning_rate": 3.833737874279452e-05, "loss": 0.0154, "step": 59620 }, { "grad_norm": 0.24693070352077484, "learning_rate": 3.8321300772080433e-05, "loss": 0.0149, "step": 59630 }, { "grad_norm": 0.16911043226718903, "learning_rate": 3.8305224078531345e-05, "loss": 0.0136, "step": 59640 }, { "grad_norm": 0.18306730687618256, "learning_rate": 3.828914866390537e-05, "loss": 0.0189, "step": 59650 }, { "grad_norm": 0.2508815824985504, "learning_rate": 3.827307452996048e-05, "loss": 0.0154, "step": 59660 }, { "grad_norm": 0.18273591995239258, "learning_rate": 3.825700167845454e-05, "loss": 0.0147, "step": 59670 }, { "grad_norm": 0.15564855933189392, "learning_rate": 3.8240930111145214e-05, "loss": 0.0159, "step": 59680 }, { "grad_norm": 0.17404063045978546, "learning_rate": 3.8224859829790116e-05, "loss": 0.0133, "step": 59690 }, { "grad_norm": 0.18935197591781616, "learning_rate": 3.820879083614664e-05, "loss": 0.0181, "step": 59700 }, { "grad_norm": 0.20379292964935303, "learning_rate": 3.819272313197208e-05, "loss": 0.0166, "step": 59710 }, { "grad_norm": 0.1698424071073532, "learning_rate": 3.817665671902355e-05, "loss": 0.0167, "step": 59720 }, { "grad_norm": 0.15801271796226501, "learning_rate": 3.816059159905808e-05, "loss": 0.0165, "step": 59730 }, { "grad_norm": 0.18441380560398102, "learning_rate": 3.8144527773832476e-05, "loss": 0.0154, "step": 59740 }, { "grad_norm": 0.26537561416625977, "learning_rate": 3.812846524510352e-05, "loss": 0.0184, "step": 59750 }, { "grad_norm": 0.1702812910079956, "learning_rate": 3.811240401462775e-05, "loss": 0.0159, "step": 59760 }, { "grad_norm": 0.15566974878311157, "learning_rate": 3.809634408416162e-05, "loss": 0.0185, "step": 59770 }, { "grad_norm": 0.1636994481086731, "learning_rate": 3.8080285455461394e-05, "loss": 0.015, "step": 59780 }, { "grad_norm": 0.1427096277475357, "learning_rate": 3.806422813028323e-05, "loss": 0.0161, "step": 59790 }, { "grad_norm": 0.18356139957904816, "learning_rate": 3.8048172110383114e-05, "loss": 0.0121, "step": 59800 }, { "grad_norm": 0.2145312875509262, "learning_rate": 3.8032117397516944e-05, "loss": 0.0149, "step": 59810 }, { "grad_norm": 0.20161661505699158, "learning_rate": 3.801606399344042e-05, "loss": 0.0182, "step": 59820 }, { "grad_norm": 0.17548847198486328, "learning_rate": 3.800001189990913e-05, "loss": 0.0137, "step": 59830 }, { "grad_norm": 0.15770260989665985, "learning_rate": 3.798396111867847e-05, "loss": 0.0141, "step": 59840 }, { "grad_norm": 0.1840313822031021, "learning_rate": 3.796791165150378e-05, "loss": 0.0193, "step": 59850 }, { "grad_norm": 0.169601172208786, "learning_rate": 3.795186350014018e-05, "loss": 0.0151, "step": 59860 }, { "grad_norm": 0.30122432112693787, "learning_rate": 3.793581666634266e-05, "loss": 0.0125, "step": 59870 }, { "grad_norm": 0.16737881302833557, "learning_rate": 3.791977115186609e-05, "loss": 0.0182, "step": 59880 }, { "grad_norm": 0.1507413536310196, "learning_rate": 3.79037269584652e-05, "loss": 0.017, "step": 59890 }, { "grad_norm": 0.22236691415309906, "learning_rate": 3.7887684087894545e-05, "loss": 0.0149, "step": 59900 }, { "grad_norm": 0.18682779371738434, "learning_rate": 3.787164254190854e-05, "loss": 0.0129, "step": 59910 }, { "grad_norm": 0.1637863963842392, "learning_rate": 3.7855602322261473e-05, "loss": 0.0154, "step": 59920 }, { "grad_norm": 0.13111920654773712, "learning_rate": 3.783956343070746e-05, "loss": 0.0141, "step": 59930 }, { "grad_norm": 0.174064040184021, "learning_rate": 3.7823525869000526e-05, "loss": 0.0122, "step": 59940 }, { "grad_norm": 0.17122890055179596, "learning_rate": 3.78074896388945e-05, "loss": 0.0137, "step": 59950 }, { "grad_norm": 0.16424179077148438, "learning_rate": 3.7791454742143075e-05, "loss": 0.0118, "step": 59960 }, { "grad_norm": 0.23120573163032532, "learning_rate": 3.7775421180499795e-05, "loss": 0.0106, "step": 59970 }, { "grad_norm": 0.21413205564022064, "learning_rate": 3.7759388955718086e-05, "loss": 0.0156, "step": 59980 }, { "grad_norm": 0.1930137723684311, "learning_rate": 3.774335806955117e-05, "loss": 0.0154, "step": 59990 }, { "grad_norm": 0.1473628729581833, "learning_rate": 3.7727328523752215e-05, "loss": 0.0147, "step": 60000 }, { "grad_norm": 0.11156860738992691, "learning_rate": 3.771130032007415e-05, "loss": 0.015, "step": 60010 }, { "grad_norm": 0.15562286972999573, "learning_rate": 3.769527346026982e-05, "loss": 0.0144, "step": 60020 }, { "grad_norm": 0.17367666959762573, "learning_rate": 3.767924794609188e-05, "loss": 0.018, "step": 60030 }, { "grad_norm": 0.14311204850673676, "learning_rate": 3.7663223779292854e-05, "loss": 0.0159, "step": 60040 }, { "grad_norm": 0.13963022828102112, "learning_rate": 3.764720096162512e-05, "loss": 0.0134, "step": 60050 }, { "grad_norm": 0.1868361234664917, "learning_rate": 3.763117949484092e-05, "loss": 0.015, "step": 60060 }, { "grad_norm": 0.157617449760437, "learning_rate": 3.7615159380692344e-05, "loss": 0.0131, "step": 60070 }, { "grad_norm": 0.1881200522184372, "learning_rate": 3.759914062093132e-05, "loss": 0.0137, "step": 60080 }, { "grad_norm": 0.20544230937957764, "learning_rate": 3.7583123217309615e-05, "loss": 0.016, "step": 60090 }, { "grad_norm": 0.15628184378147125, "learning_rate": 3.7567107171578904e-05, "loss": 0.0157, "step": 60100 }, { "grad_norm": 0.26496994495391846, "learning_rate": 3.755109248549066e-05, "loss": 0.0167, "step": 60110 }, { "grad_norm": 0.1624903827905655, "learning_rate": 3.75350791607962e-05, "loss": 0.0172, "step": 60120 }, { "grad_norm": 0.1826082020998001, "learning_rate": 3.751906719924676e-05, "loss": 0.0129, "step": 60130 }, { "grad_norm": 0.20146915316581726, "learning_rate": 3.750305660259337e-05, "loss": 0.0143, "step": 60140 }, { "grad_norm": 0.149277001619339, "learning_rate": 3.748704737258693e-05, "loss": 0.0138, "step": 60150 }, { "grad_norm": 0.1725972294807434, "learning_rate": 3.747103951097816e-05, "loss": 0.0139, "step": 60160 }, { "grad_norm": 0.17879408597946167, "learning_rate": 3.745503301951768e-05, "loss": 0.0174, "step": 60170 }, { "grad_norm": 0.27078747749328613, "learning_rate": 3.743902789995592e-05, "loss": 0.0153, "step": 60180 }, { "grad_norm": 0.1539285033941269, "learning_rate": 3.7423024154043183e-05, "loss": 0.0135, "step": 60190 }, { "grad_norm": 0.20229823887348175, "learning_rate": 3.740702178352963e-05, "loss": 0.0153, "step": 60200 }, { "grad_norm": 0.14950823783874512, "learning_rate": 3.739102079016523e-05, "loss": 0.015, "step": 60210 }, { "grad_norm": 0.25052058696746826, "learning_rate": 3.7375021175699846e-05, "loss": 0.0145, "step": 60220 }, { "grad_norm": 0.18095311522483826, "learning_rate": 3.735902294188317e-05, "loss": 0.0164, "step": 60230 }, { "grad_norm": 0.22924616932868958, "learning_rate": 3.734302609046473e-05, "loss": 0.0136, "step": 60240 }, { "grad_norm": 0.2476731836795807, "learning_rate": 3.732703062319391e-05, "loss": 0.0152, "step": 60250 }, { "grad_norm": 0.18355511128902435, "learning_rate": 3.731103654181999e-05, "loss": 0.0142, "step": 60260 }, { "grad_norm": 0.21972739696502686, "learning_rate": 3.729504384809203e-05, "loss": 0.0145, "step": 60270 }, { "grad_norm": 0.18856695294380188, "learning_rate": 3.727905254375898e-05, "loss": 0.0198, "step": 60280 }, { "grad_norm": 0.29869797825813293, "learning_rate": 3.7263062630569614e-05, "loss": 0.0148, "step": 60290 }, { "grad_norm": 0.23156313598155975, "learning_rate": 3.724707411027256e-05, "loss": 0.0172, "step": 60300 }, { "grad_norm": 0.15594497323036194, "learning_rate": 3.723108698461631e-05, "loss": 0.0143, "step": 60310 }, { "grad_norm": 0.18363171815872192, "learning_rate": 3.72151012553492e-05, "loss": 0.0178, "step": 60320 }, { "grad_norm": 0.22024933993816376, "learning_rate": 3.719911692421939e-05, "loss": 0.014, "step": 60330 }, { "grad_norm": 0.13793937861919403, "learning_rate": 3.71831339929749e-05, "loss": 0.0223, "step": 60340 }, { "grad_norm": 0.16528946161270142, "learning_rate": 3.716715246336362e-05, "loss": 0.0135, "step": 60350 }, { "grad_norm": 0.18057091534137726, "learning_rate": 3.715117233713324e-05, "loss": 0.0141, "step": 60360 }, { "grad_norm": 0.2143210619688034, "learning_rate": 3.713519361603135e-05, "loss": 0.0134, "step": 60370 }, { "grad_norm": 0.18117162585258484, "learning_rate": 3.711921630180532e-05, "loss": 0.0149, "step": 60380 }, { "grad_norm": 0.13349610567092896, "learning_rate": 3.710324039620245e-05, "loss": 0.0131, "step": 60390 }, { "grad_norm": 0.1490483283996582, "learning_rate": 3.708726590096982e-05, "loss": 0.0154, "step": 60400 }, { "grad_norm": 0.18259240686893463, "learning_rate": 3.707129281785437e-05, "loss": 0.017, "step": 60410 }, { "grad_norm": 0.1535792052745819, "learning_rate": 3.705532114860291e-05, "loss": 0.0146, "step": 60420 }, { "grad_norm": 0.1720525324344635, "learning_rate": 3.703935089496207e-05, "loss": 0.0143, "step": 60430 }, { "grad_norm": 0.18325059115886688, "learning_rate": 3.702338205867829e-05, "loss": 0.021, "step": 60440 }, { "grad_norm": 0.18612425029277802, "learning_rate": 3.7007414641497976e-05, "loss": 0.0149, "step": 60450 }, { "grad_norm": 0.2174122929573059, "learning_rate": 3.699144864516726e-05, "loss": 0.0129, "step": 60460 }, { "grad_norm": 0.19415053725242615, "learning_rate": 3.6975484071432153e-05, "loss": 0.0137, "step": 60470 }, { "grad_norm": 0.1959456503391266, "learning_rate": 3.695952092203854e-05, "loss": 0.0155, "step": 60480 }, { "grad_norm": 0.12955832481384277, "learning_rate": 3.69435591987321e-05, "loss": 0.0139, "step": 60490 }, { "grad_norm": 0.19412335753440857, "learning_rate": 3.6927598903258374e-05, "loss": 0.022, "step": 60500 }, { "grad_norm": 0.14048519730567932, "learning_rate": 3.691164003736281e-05, "loss": 0.0119, "step": 60510 }, { "grad_norm": 0.18335944414138794, "learning_rate": 3.689568260279061e-05, "loss": 0.0151, "step": 60520 }, { "grad_norm": 0.1400325447320938, "learning_rate": 3.687972660128686e-05, "loss": 0.0158, "step": 60530 }, { "grad_norm": 0.12743650376796722, "learning_rate": 3.686377203459648e-05, "loss": 0.0131, "step": 60540 }, { "grad_norm": 0.17184916138648987, "learning_rate": 3.6847818904464226e-05, "loss": 0.0174, "step": 60550 }, { "grad_norm": 0.17952504754066467, "learning_rate": 3.683186721263474e-05, "loss": 0.0154, "step": 60560 }, { "grad_norm": 0.18052245676517487, "learning_rate": 3.6815916960852435e-05, "loss": 0.0151, "step": 60570 }, { "grad_norm": 0.1393115222454071, "learning_rate": 3.679996815086165e-05, "loss": 0.0139, "step": 60580 }, { "grad_norm": 0.1584555208683014, "learning_rate": 3.678402078440649e-05, "loss": 0.0141, "step": 60590 }, { "grad_norm": 0.20403002202510834, "learning_rate": 3.676807486323096e-05, "loss": 0.0158, "step": 60600 }, { "grad_norm": 0.12649574875831604, "learning_rate": 3.6752130389078864e-05, "loss": 0.0135, "step": 60610 }, { "grad_norm": 0.15113018453121185, "learning_rate": 3.673618736369388e-05, "loss": 0.0139, "step": 60620 }, { "grad_norm": 0.17161059379577637, "learning_rate": 3.6720245788819474e-05, "loss": 0.0159, "step": 60630 }, { "grad_norm": 0.193317711353302, "learning_rate": 3.6704305666199044e-05, "loss": 0.0163, "step": 60640 }, { "grad_norm": 0.2207261025905609, "learning_rate": 3.668836699757576e-05, "loss": 0.0144, "step": 60650 }, { "grad_norm": 0.14885371923446655, "learning_rate": 3.667242978469265e-05, "loss": 0.0146, "step": 60660 }, { "grad_norm": 0.1458842009305954, "learning_rate": 3.6656494029292575e-05, "loss": 0.0146, "step": 60670 }, { "grad_norm": 0.1691426932811737, "learning_rate": 3.664055973311825e-05, "loss": 0.0173, "step": 60680 }, { "grad_norm": 0.17329250276088715, "learning_rate": 3.662462689791221e-05, "loss": 0.015, "step": 60690 }, { "grad_norm": 0.20833773910999298, "learning_rate": 3.660869552541689e-05, "loss": 0.0137, "step": 60700 }, { "grad_norm": 0.16339045763015747, "learning_rate": 3.659276561737448e-05, "loss": 0.0166, "step": 60710 }, { "grad_norm": 0.1724400520324707, "learning_rate": 3.657683717552708e-05, "loss": 0.0148, "step": 60720 }, { "grad_norm": 0.1556403487920761, "learning_rate": 3.6560910201616574e-05, "loss": 0.0169, "step": 60730 }, { "grad_norm": 0.15161804854869843, "learning_rate": 3.654498469738472e-05, "loss": 0.0138, "step": 60740 }, { "grad_norm": 0.1852388232946396, "learning_rate": 3.6529060664573126e-05, "loss": 0.0128, "step": 60750 }, { "grad_norm": 0.14551883935928345, "learning_rate": 3.6513138104923176e-05, "loss": 0.0137, "step": 60760 }, { "grad_norm": 0.17124545574188232, "learning_rate": 3.6497217020176176e-05, "loss": 0.0143, "step": 60770 }, { "grad_norm": 0.17323558032512665, "learning_rate": 3.648129741207323e-05, "loss": 0.015, "step": 60780 }, { "grad_norm": 0.1890445500612259, "learning_rate": 3.646537928235527e-05, "loss": 0.0123, "step": 60790 }, { "grad_norm": 0.20284000039100647, "learning_rate": 3.6449462632763075e-05, "loss": 0.0144, "step": 60800 }, { "grad_norm": 0.20221051573753357, "learning_rate": 3.643354746503729e-05, "loss": 0.016, "step": 60810 }, { "grad_norm": 0.1231871247291565, "learning_rate": 3.6417633780918345e-05, "loss": 0.0123, "step": 60820 }, { "grad_norm": 0.1772051900625229, "learning_rate": 3.640172158214655e-05, "loss": 0.0173, "step": 60830 }, { "grad_norm": 0.17579634487628937, "learning_rate": 3.6385810870462035e-05, "loss": 0.0124, "step": 60840 }, { "grad_norm": 0.155636727809906, "learning_rate": 3.636990164760479e-05, "loss": 0.0101, "step": 60850 }, { "grad_norm": 0.14846248924732208, "learning_rate": 3.6353993915314596e-05, "loss": 0.0102, "step": 60860 }, { "grad_norm": 0.21460583806037903, "learning_rate": 3.6338087675331124e-05, "loss": 0.0124, "step": 60870 }, { "grad_norm": 0.1482640951871872, "learning_rate": 3.632218292939383e-05, "loss": 0.0125, "step": 60880 }, { "grad_norm": 0.13169746100902557, "learning_rate": 3.6306279679242033e-05, "loss": 0.0135, "step": 60890 }, { "grad_norm": 0.23319652676582336, "learning_rate": 3.629037792661492e-05, "loss": 0.0151, "step": 60900 }, { "grad_norm": 0.13796202838420868, "learning_rate": 3.627447767325147e-05, "loss": 0.0126, "step": 60910 }, { "grad_norm": 0.14435254037380219, "learning_rate": 3.62585789208905e-05, "loss": 0.0143, "step": 60920 }, { "grad_norm": 0.13940797746181488, "learning_rate": 3.624268167127068e-05, "loss": 0.0138, "step": 60930 }, { "grad_norm": 0.23551417887210846, "learning_rate": 3.622678592613051e-05, "loss": 0.0163, "step": 60940 }, { "grad_norm": 0.17981413006782532, "learning_rate": 3.6210891687208304e-05, "loss": 0.0138, "step": 60950 }, { "grad_norm": 0.15636169910430908, "learning_rate": 3.619499895624228e-05, "loss": 0.011, "step": 60960 }, { "grad_norm": 0.1268191933631897, "learning_rate": 3.61791077349704e-05, "loss": 0.0163, "step": 60970 }, { "grad_norm": 0.19123922288417816, "learning_rate": 3.616321802513053e-05, "loss": 0.0148, "step": 60980 }, { "grad_norm": 0.15353454649448395, "learning_rate": 3.6147329828460325e-05, "loss": 0.0142, "step": 60990 }, { "grad_norm": 0.17237408459186554, "learning_rate": 3.6131443146697306e-05, "loss": 0.015, "step": 61000 }, { "grad_norm": 0.19732198119163513, "learning_rate": 3.611555798157879e-05, "loss": 0.0163, "step": 61010 }, { "grad_norm": 0.14693903923034668, "learning_rate": 3.6099674334842e-05, "loss": 0.0131, "step": 61020 }, { "grad_norm": 0.10510598868131638, "learning_rate": 3.6083792208223934e-05, "loss": 0.014, "step": 61030 }, { "grad_norm": 0.21686092019081116, "learning_rate": 3.606791160346142e-05, "loss": 0.013, "step": 61040 }, { "grad_norm": 0.11838739365339279, "learning_rate": 3.6052032522291134e-05, "loss": 0.0156, "step": 61050 }, { "grad_norm": 0.2603164315223694, "learning_rate": 3.603615496644962e-05, "loss": 0.0176, "step": 61060 }, { "grad_norm": 0.21434582769870758, "learning_rate": 3.60202789376732e-05, "loss": 0.0124, "step": 61070 }, { "grad_norm": 0.18400396406650543, "learning_rate": 3.6004404437698036e-05, "loss": 0.0169, "step": 61080 }, { "grad_norm": 0.16988441348075867, "learning_rate": 3.5988531468260184e-05, "loss": 0.0139, "step": 61090 }, { "grad_norm": 0.1834106296300888, "learning_rate": 3.597266003109546e-05, "loss": 0.0134, "step": 61100 }, { "grad_norm": 0.16466696560382843, "learning_rate": 3.5956790127939557e-05, "loss": 0.0171, "step": 61110 }, { "grad_norm": 0.14573204517364502, "learning_rate": 3.594092176052796e-05, "loss": 0.0151, "step": 61120 }, { "grad_norm": 0.259585976600647, "learning_rate": 3.592505493059603e-05, "loss": 0.015, "step": 61130 }, { "grad_norm": 0.2957206666469574, "learning_rate": 3.5909189639878905e-05, "loss": 0.0221, "step": 61140 }, { "grad_norm": 0.17908185720443726, "learning_rate": 3.5893325890111626e-05, "loss": 0.0135, "step": 61150 }, { "grad_norm": 0.1533159613609314, "learning_rate": 3.5877463683029034e-05, "loss": 0.0173, "step": 61160 }, { "grad_norm": 0.17291469871997833, "learning_rate": 3.586160302036578e-05, "loss": 0.0144, "step": 61170 }, { "grad_norm": 0.15532125532627106, "learning_rate": 3.584574390385636e-05, "loss": 0.0139, "step": 61180 }, { "grad_norm": 0.16457948088645935, "learning_rate": 3.582988633523511e-05, "loss": 0.0129, "step": 61190 }, { "grad_norm": 0.14074580371379852, "learning_rate": 3.581403031623616e-05, "loss": 0.0148, "step": 61200 }, { "grad_norm": 0.13637395203113556, "learning_rate": 3.579817584859354e-05, "loss": 0.0152, "step": 61210 }, { "grad_norm": 0.20935845375061035, "learning_rate": 3.578232293404107e-05, "loss": 0.015, "step": 61220 }, { "grad_norm": 0.15319982171058655, "learning_rate": 3.5766471574312385e-05, "loss": 0.0134, "step": 61230 }, { "grad_norm": 0.1604514718055725, "learning_rate": 3.5750621771140964e-05, "loss": 0.0134, "step": 61240 }, { "grad_norm": 0.12861034274101257, "learning_rate": 3.5734773526260115e-05, "loss": 0.014, "step": 61250 }, { "grad_norm": 0.21910090744495392, "learning_rate": 3.5718926841402996e-05, "loss": 0.0147, "step": 61260 }, { "grad_norm": 0.1496279090642929, "learning_rate": 3.5703081718302555e-05, "loss": 0.0171, "step": 61270 }, { "grad_norm": 0.19356639683246613, "learning_rate": 3.568723815869162e-05, "loss": 0.0178, "step": 61280 }, { "grad_norm": 0.1544259637594223, "learning_rate": 3.5671396164302777e-05, "loss": 0.0158, "step": 61290 }, { "grad_norm": 0.14467070996761322, "learning_rate": 3.565555573686853e-05, "loss": 0.011, "step": 61300 }, { "grad_norm": 0.1598178595304489, "learning_rate": 3.563971687812113e-05, "loss": 0.0169, "step": 61310 }, { "grad_norm": 0.1460193246603012, "learning_rate": 3.5623879589792706e-05, "loss": 0.0128, "step": 61320 }, { "grad_norm": 0.1721010059118271, "learning_rate": 3.5608043873615185e-05, "loss": 0.0139, "step": 61330 }, { "grad_norm": 0.16738903522491455, "learning_rate": 3.559220973132035e-05, "loss": 0.0129, "step": 61340 }, { "grad_norm": 0.15567074716091156, "learning_rate": 3.557637716463981e-05, "loss": 0.0139, "step": 61350 }, { "grad_norm": 0.15917059779167175, "learning_rate": 3.556054617530497e-05, "loss": 0.0146, "step": 61360 }, { "grad_norm": 0.22398647665977478, "learning_rate": 3.5544716765047106e-05, "loss": 0.0175, "step": 61370 }, { "grad_norm": 0.20631852746009827, "learning_rate": 3.5528888935597274e-05, "loss": 0.0128, "step": 61380 }, { "grad_norm": 0.15365158021450043, "learning_rate": 3.551306268868636e-05, "loss": 0.0137, "step": 61390 }, { "grad_norm": 0.1605353057384491, "learning_rate": 3.5497238026045166e-05, "loss": 0.014, "step": 61400 }, { "grad_norm": 0.16955327987670898, "learning_rate": 3.548141494940421e-05, "loss": 0.0155, "step": 61410 }, { "grad_norm": 0.20245225727558136, "learning_rate": 3.546559346049388e-05, "loss": 0.013, "step": 61420 }, { "grad_norm": 0.1533825546503067, "learning_rate": 3.544977356104441e-05, "loss": 0.0149, "step": 61430 }, { "grad_norm": 0.1729903519153595, "learning_rate": 3.543395525278582e-05, "loss": 0.017, "step": 61440 }, { "grad_norm": 0.2871035635471344, "learning_rate": 3.5418138537447985e-05, "loss": 0.0175, "step": 61450 }, { "grad_norm": 0.1625252515077591, "learning_rate": 3.5402323416760574e-05, "loss": 0.0152, "step": 61460 }, { "grad_norm": 0.15096458792686462, "learning_rate": 3.538650989245316e-05, "loss": 0.015, "step": 61470 }, { "grad_norm": 0.14332179725170135, "learning_rate": 3.537069796625504e-05, "loss": 0.0133, "step": 61480 }, { "grad_norm": 0.17065973579883575, "learning_rate": 3.53548876398954e-05, "loss": 0.0131, "step": 61490 }, { "grad_norm": 0.14761455357074738, "learning_rate": 3.5339078915103214e-05, "loss": 0.0126, "step": 61500 }, { "grad_norm": 0.17646567523479462, "learning_rate": 3.532327179360733e-05, "loss": 0.0146, "step": 61510 }, { "grad_norm": 0.20081636309623718, "learning_rate": 3.530746627713636e-05, "loss": 0.0145, "step": 61520 }, { "grad_norm": 0.15179452300071716, "learning_rate": 3.5291662367418805e-05, "loss": 0.0129, "step": 61530 }, { "grad_norm": 0.1576836258172989, "learning_rate": 3.527586006618293e-05, "loss": 0.017, "step": 61540 }, { "grad_norm": 0.21731384098529816, "learning_rate": 3.526005937515687e-05, "loss": 0.0173, "step": 61550 }, { "grad_norm": 0.15355490148067474, "learning_rate": 3.524426029606856e-05, "loss": 0.0139, "step": 61560 }, { "grad_norm": 0.18626873195171356, "learning_rate": 3.5228462830645746e-05, "loss": 0.0115, "step": 61570 }, { "grad_norm": 0.2591467797756195, "learning_rate": 3.5212666980616036e-05, "loss": 0.015, "step": 61580 }, { "grad_norm": 0.24669519066810608, "learning_rate": 3.519687274770681e-05, "loss": 0.0135, "step": 61590 }, { "grad_norm": 0.2051355242729187, "learning_rate": 3.518108013364535e-05, "loss": 0.0162, "step": 61600 }, { "grad_norm": 0.16631180047988892, "learning_rate": 3.516528914015868e-05, "loss": 0.0174, "step": 61610 }, { "grad_norm": 0.19864293932914734, "learning_rate": 3.5149499768973694e-05, "loss": 0.0149, "step": 61620 }, { "grad_norm": 0.1982591152191162, "learning_rate": 3.513371202181708e-05, "loss": 0.0145, "step": 61630 }, { "grad_norm": 0.1289958357810974, "learning_rate": 3.511792590041537e-05, "loss": 0.0165, "step": 61640 }, { "grad_norm": 0.1814105212688446, "learning_rate": 3.510214140649488e-05, "loss": 0.0142, "step": 61650 }, { "grad_norm": 0.18204206228256226, "learning_rate": 3.508635854178184e-05, "loss": 0.0157, "step": 61660 }, { "grad_norm": 0.17714239656925201, "learning_rate": 3.50705773080022e-05, "loss": 0.0141, "step": 61670 }, { "grad_norm": 0.2119138389825821, "learning_rate": 3.505479770688178e-05, "loss": 0.0113, "step": 61680 }, { "grad_norm": 0.17224468290805817, "learning_rate": 3.503901974014622e-05, "loss": 0.0123, "step": 61690 }, { "grad_norm": 0.18438492715358734, "learning_rate": 3.502324340952095e-05, "loss": 0.0155, "step": 61700 }, { "grad_norm": 0.14105796813964844, "learning_rate": 3.500746871673126e-05, "loss": 0.0128, "step": 61710 }, { "grad_norm": 0.14507035911083221, "learning_rate": 3.499169566350227e-05, "loss": 0.016, "step": 61720 }, { "grad_norm": 0.1855313777923584, "learning_rate": 3.4975924251558876e-05, "loss": 0.0141, "step": 61730 }, { "grad_norm": 0.1729043573141098, "learning_rate": 3.496015448262582e-05, "loss": 0.0196, "step": 61740 }, { "grad_norm": 0.19721783697605133, "learning_rate": 3.494438635842764e-05, "loss": 0.0165, "step": 61750 }, { "grad_norm": 0.14867080748081207, "learning_rate": 3.492861988068874e-05, "loss": 0.0149, "step": 61760 }, { "grad_norm": 0.17144069075584412, "learning_rate": 3.491285505113332e-05, "loss": 0.0145, "step": 61770 }, { "grad_norm": 0.17521730065345764, "learning_rate": 3.4897091871485376e-05, "loss": 0.014, "step": 61780 }, { "grad_norm": 0.24904312193393707, "learning_rate": 3.488133034346875e-05, "loss": 0.0149, "step": 61790 }, { "grad_norm": 0.23336274921894073, "learning_rate": 3.486557046880713e-05, "loss": 0.0155, "step": 61800 }, { "grad_norm": 0.3552899658679962, "learning_rate": 3.484981224922395e-05, "loss": 0.0165, "step": 61810 }, { "grad_norm": 0.2030661255121231, "learning_rate": 3.4834055686442544e-05, "loss": 0.0142, "step": 61820 }, { "grad_norm": 0.24418652057647705, "learning_rate": 3.4818300782186e-05, "loss": 0.0139, "step": 61830 }, { "grad_norm": 0.12967000901699066, "learning_rate": 3.4802547538177236e-05, "loss": 0.016, "step": 61840 }, { "grad_norm": 0.1759636551141739, "learning_rate": 3.478679595613905e-05, "loss": 0.0169, "step": 61850 }, { "grad_norm": 0.18723304569721222, "learning_rate": 3.4771046037793975e-05, "loss": 0.0118, "step": 61860 }, { "grad_norm": 0.1591520458459854, "learning_rate": 3.4755297784864416e-05, "loss": 0.0145, "step": 61870 }, { "grad_norm": 0.18792293965816498, "learning_rate": 3.473955119907257e-05, "loss": 0.0118, "step": 61880 }, { "grad_norm": 0.1516798585653305, "learning_rate": 3.472380628214046e-05, "loss": 0.0131, "step": 61890 }, { "grad_norm": 0.11454416811466217, "learning_rate": 3.4708063035789906e-05, "loss": 0.0156, "step": 61900 }, { "grad_norm": 0.1728958785533905, "learning_rate": 3.4692321461742614e-05, "loss": 0.0123, "step": 61910 }, { "grad_norm": 0.15963402390480042, "learning_rate": 3.4676581561720023e-05, "loss": 0.0148, "step": 61920 }, { "grad_norm": 0.14448900520801544, "learning_rate": 3.466084333744344e-05, "loss": 0.0119, "step": 61930 }, { "grad_norm": 0.1480613499879837, "learning_rate": 3.4645106790633964e-05, "loss": 0.0148, "step": 61940 }, { "grad_norm": 0.1531749665737152, "learning_rate": 3.4629371923012526e-05, "loss": 0.0123, "step": 61950 }, { "grad_norm": 0.1671329289674759, "learning_rate": 3.4613638736299856e-05, "loss": 0.0138, "step": 61960 }, { "grad_norm": 0.15830284357070923, "learning_rate": 3.459790723221652e-05, "loss": 0.0129, "step": 61970 }, { "grad_norm": 0.14057712256908417, "learning_rate": 3.45821774124829e-05, "loss": 0.0156, "step": 61980 }, { "grad_norm": 0.1905248761177063, "learning_rate": 3.456644927881918e-05, "loss": 0.0135, "step": 61990 }, { "grad_norm": 0.14729167520999908, "learning_rate": 3.4550722832945354e-05, "loss": 0.0127, "step": 62000 }, { "grad_norm": 0.1377001702785492, "learning_rate": 3.453499807658126e-05, "loss": 0.0146, "step": 62010 }, { "grad_norm": 0.11686280369758606, "learning_rate": 3.451927501144653e-05, "loss": 0.0146, "step": 62020 }, { "grad_norm": 0.16868272423744202, "learning_rate": 3.450355363926059e-05, "loss": 0.0149, "step": 62030 }, { "grad_norm": 0.20015831291675568, "learning_rate": 3.448783396174273e-05, "loss": 0.0171, "step": 62040 }, { "grad_norm": 0.1513453871011734, "learning_rate": 3.4472115980612034e-05, "loss": 0.0142, "step": 62050 }, { "grad_norm": 0.19501076638698578, "learning_rate": 3.4456399697587396e-05, "loss": 0.013, "step": 62060 }, { "grad_norm": 0.19138266146183014, "learning_rate": 3.4440685114387506e-05, "loss": 0.0158, "step": 62070 }, { "grad_norm": 0.18559962511062622, "learning_rate": 3.442497223273091e-05, "loss": 0.0156, "step": 62080 }, { "grad_norm": 0.11009247601032257, "learning_rate": 3.4409261054335926e-05, "loss": 0.0161, "step": 62090 }, { "grad_norm": 0.16278128325939178, "learning_rate": 3.439355158092069e-05, "loss": 0.0141, "step": 62100 }, { "grad_norm": 0.16498194634914398, "learning_rate": 3.437784381420322e-05, "loss": 0.0158, "step": 62110 }, { "grad_norm": 0.12597481906414032, "learning_rate": 3.436213775590126e-05, "loss": 0.0118, "step": 62120 }, { "grad_norm": 0.12111083418130875, "learning_rate": 3.4346433407732405e-05, "loss": 0.0146, "step": 62130 }, { "grad_norm": 0.1168365553021431, "learning_rate": 3.433073077141405e-05, "loss": 0.012, "step": 62140 }, { "grad_norm": 0.16196773946285248, "learning_rate": 3.431502984866343e-05, "loss": 0.0159, "step": 62150 }, { "grad_norm": 0.1459909826517105, "learning_rate": 3.429933064119753e-05, "loss": 0.0137, "step": 62160 }, { "grad_norm": 0.1415610909461975, "learning_rate": 3.4283633150733255e-05, "loss": 0.0138, "step": 62170 }, { "grad_norm": 0.15895724296569824, "learning_rate": 3.4267937378987225e-05, "loss": 0.0136, "step": 62180 }, { "grad_norm": 0.1694658249616623, "learning_rate": 3.425224332767592e-05, "loss": 0.019, "step": 62190 }, { "grad_norm": 0.13315384089946747, "learning_rate": 3.4236550998515595e-05, "loss": 0.0131, "step": 62200 }, { "grad_norm": 0.33038094639778137, "learning_rate": 3.422086039322235e-05, "loss": 0.0148, "step": 62210 }, { "grad_norm": 0.17050780355930328, "learning_rate": 3.420517151351209e-05, "loss": 0.015, "step": 62220 }, { "grad_norm": 0.25043797492980957, "learning_rate": 3.418948436110053e-05, "loss": 0.0161, "step": 62230 }, { "grad_norm": 0.24692751467227936, "learning_rate": 3.417379893770319e-05, "loss": 0.018, "step": 62240 }, { "grad_norm": 0.19138595461845398, "learning_rate": 3.41581152450354e-05, "loss": 0.0144, "step": 62250 }, { "grad_norm": 0.2168937623500824, "learning_rate": 3.414243328481231e-05, "loss": 0.0139, "step": 62260 }, { "grad_norm": 0.16541263461112976, "learning_rate": 3.412675305874888e-05, "loss": 0.0152, "step": 62270 }, { "grad_norm": 0.26106029748916626, "learning_rate": 3.4111074568559876e-05, "loss": 0.0152, "step": 62280 }, { "grad_norm": 0.12415573000907898, "learning_rate": 3.4095397815959846e-05, "loss": 0.0133, "step": 62290 }, { "grad_norm": 0.18306539952754974, "learning_rate": 3.407972280266322e-05, "loss": 0.0141, "step": 62300 }, { "grad_norm": 0.19246026873588562, "learning_rate": 3.4064049530384166e-05, "loss": 0.016, "step": 62310 }, { "grad_norm": 0.1593882292509079, "learning_rate": 3.40483780008367e-05, "loss": 0.0122, "step": 62320 }, { "grad_norm": 0.1568378508090973, "learning_rate": 3.4032708215734635e-05, "loss": 0.0148, "step": 62330 }, { "grad_norm": 0.13214951753616333, "learning_rate": 3.4017040176791584e-05, "loss": 0.0125, "step": 62340 }, { "grad_norm": 0.12920160591602325, "learning_rate": 3.4001373885720976e-05, "loss": 0.0122, "step": 62350 }, { "grad_norm": 0.1299949288368225, "learning_rate": 3.398570934423608e-05, "loss": 0.0155, "step": 62360 }, { "grad_norm": 0.16176187992095947, "learning_rate": 3.397004655404994e-05, "loss": 0.0117, "step": 62370 }, { "grad_norm": 0.18051356077194214, "learning_rate": 3.395438551687541e-05, "loss": 0.0153, "step": 62380 }, { "grad_norm": 0.145471453666687, "learning_rate": 3.393872623442515e-05, "loss": 0.0121, "step": 62390 }, { "grad_norm": 0.15375196933746338, "learning_rate": 3.392306870841164e-05, "loss": 0.0141, "step": 62400 }, { "grad_norm": 0.23949375748634338, "learning_rate": 3.390741294054715e-05, "loss": 0.0151, "step": 62410 }, { "grad_norm": 0.17458820343017578, "learning_rate": 3.38917589325438e-05, "loss": 0.0113, "step": 62420 }, { "grad_norm": 0.1946299970149994, "learning_rate": 3.387610668611349e-05, "loss": 0.0115, "step": 62430 }, { "grad_norm": 0.16830098628997803, "learning_rate": 3.38604562029679e-05, "loss": 0.0168, "step": 62440 }, { "grad_norm": 0.17065958678722382, "learning_rate": 3.384480748481855e-05, "loss": 0.016, "step": 62450 }, { "grad_norm": 0.1481030136346817, "learning_rate": 3.382916053337676e-05, "loss": 0.0121, "step": 62460 }, { "grad_norm": 0.13886483013629913, "learning_rate": 3.381351535035366e-05, "loss": 0.0167, "step": 62470 }, { "grad_norm": 0.1656406968832016, "learning_rate": 3.379787193746019e-05, "loss": 0.0153, "step": 62480 }, { "grad_norm": 0.10053835064172745, "learning_rate": 3.378223029640709e-05, "loss": 0.0138, "step": 62490 }, { "grad_norm": 0.16099107265472412, "learning_rate": 3.376659042890489e-05, "loss": 0.0128, "step": 62500 }, { "grad_norm": 0.20080114901065826, "learning_rate": 3.375095233666397e-05, "loss": 0.0143, "step": 62510 }, { "grad_norm": 0.1502000242471695, "learning_rate": 3.373531602139447e-05, "loss": 0.0117, "step": 62520 }, { "grad_norm": 0.18008697032928467, "learning_rate": 3.371968148480635e-05, "loss": 0.0113, "step": 62530 }, { "grad_norm": 0.22361594438552856, "learning_rate": 3.370404872860937e-05, "loss": 0.0171, "step": 62540 }, { "grad_norm": 0.2343161404132843, "learning_rate": 3.3688417754513145e-05, "loss": 0.0128, "step": 62550 }, { "grad_norm": 0.15336176753044128, "learning_rate": 3.367278856422703e-05, "loss": 0.0222, "step": 62560 }, { "grad_norm": 0.13109175860881805, "learning_rate": 3.365716115946021e-05, "loss": 0.0161, "step": 62570 }, { "grad_norm": 0.16262760758399963, "learning_rate": 3.3641535541921666e-05, "loss": 0.0138, "step": 62580 }, { "grad_norm": 0.19403935968875885, "learning_rate": 3.36259117133202e-05, "loss": 0.0179, "step": 62590 }, { "grad_norm": 0.21677623689174652, "learning_rate": 3.36102896753644e-05, "loss": 0.0169, "step": 62600 }, { "grad_norm": 0.16845734417438507, "learning_rate": 3.35946694297627e-05, "loss": 0.0149, "step": 62610 }, { "grad_norm": 0.22146689891815186, "learning_rate": 3.357905097822328e-05, "loss": 0.0115, "step": 62620 }, { "grad_norm": 0.1699027419090271, "learning_rate": 3.356343432245416e-05, "loss": 0.0133, "step": 62630 }, { "grad_norm": 0.1688254028558731, "learning_rate": 3.354781946416315e-05, "loss": 0.0143, "step": 62640 }, { "grad_norm": 0.20875659584999084, "learning_rate": 3.3532206405057864e-05, "loss": 0.0161, "step": 62650 }, { "grad_norm": 0.1812877357006073, "learning_rate": 3.351659514684573e-05, "loss": 0.0139, "step": 62660 }, { "grad_norm": 0.17772506177425385, "learning_rate": 3.350098569123395e-05, "loss": 0.0145, "step": 62670 }, { "grad_norm": 0.1705770492553711, "learning_rate": 3.348537803992958e-05, "loss": 0.013, "step": 62680 }, { "grad_norm": 0.12352770566940308, "learning_rate": 3.346977219463945e-05, "loss": 0.0145, "step": 62690 }, { "grad_norm": 0.15637093782424927, "learning_rate": 3.345416815707017e-05, "loss": 0.0131, "step": 62700 }, { "grad_norm": 0.16359449923038483, "learning_rate": 3.3438565928928185e-05, "loss": 0.0132, "step": 62710 }, { "grad_norm": 0.1941327452659607, "learning_rate": 3.3422965511919736e-05, "loss": 0.0155, "step": 62720 }, { "grad_norm": 0.1514066755771637, "learning_rate": 3.3407366907750846e-05, "loss": 0.0139, "step": 62730 }, { "grad_norm": 0.1690322458744049, "learning_rate": 3.339177011812737e-05, "loss": 0.0164, "step": 62740 }, { "grad_norm": 0.21963068842887878, "learning_rate": 3.3376175144754935e-05, "loss": 0.0147, "step": 62750 }, { "grad_norm": 0.12974770367145538, "learning_rate": 3.3360581989339004e-05, "loss": 0.0149, "step": 62760 }, { "grad_norm": 0.18250912427902222, "learning_rate": 3.3344990653584804e-05, "loss": 0.0148, "step": 62770 }, { "grad_norm": 0.14526143670082092, "learning_rate": 3.332940113919739e-05, "loss": 0.0158, "step": 62780 }, { "grad_norm": 0.198759987950325, "learning_rate": 3.331381344788161e-05, "loss": 0.0171, "step": 62790 }, { "grad_norm": 0.13955092430114746, "learning_rate": 3.3298227581342076e-05, "loss": 0.0111, "step": 62800 }, { "grad_norm": 0.18526354432106018, "learning_rate": 3.328264354128329e-05, "loss": 0.0137, "step": 62810 }, { "grad_norm": 0.15425284206867218, "learning_rate": 3.3267061329409465e-05, "loss": 0.0148, "step": 62820 }, { "grad_norm": 0.13151022791862488, "learning_rate": 3.325148094742466e-05, "loss": 0.0125, "step": 62830 }, { "grad_norm": 0.1431012898683548, "learning_rate": 3.323590239703271e-05, "loss": 0.0185, "step": 62840 }, { "grad_norm": 0.20215918123722076, "learning_rate": 3.322032567993728e-05, "loss": 0.0141, "step": 62850 }, { "grad_norm": 0.1241859644651413, "learning_rate": 3.3204750797841774e-05, "loss": 0.0135, "step": 62860 }, { "grad_norm": 0.16723603010177612, "learning_rate": 3.31891777524495e-05, "loss": 0.0141, "step": 62870 }, { "grad_norm": 0.22376860678195953, "learning_rate": 3.317360654546347e-05, "loss": 0.0154, "step": 62880 }, { "grad_norm": 0.12527711689472198, "learning_rate": 3.3158037178586534e-05, "loss": 0.0138, "step": 62890 }, { "grad_norm": 0.20469993352890015, "learning_rate": 3.3142469653521316e-05, "loss": 0.0136, "step": 62900 }, { "grad_norm": 0.1429566740989685, "learning_rate": 3.3126903971970284e-05, "loss": 0.0145, "step": 62910 }, { "grad_norm": 0.1343889981508255, "learning_rate": 3.311134013563564e-05, "loss": 0.0161, "step": 62920 }, { "grad_norm": 0.1961313784122467, "learning_rate": 3.309577814621947e-05, "loss": 0.0131, "step": 62930 }, { "grad_norm": 0.16310560703277588, "learning_rate": 3.308021800542359e-05, "loss": 0.0136, "step": 62940 }, { "grad_norm": 0.1851368546485901, "learning_rate": 3.306465971494963e-05, "loss": 0.0135, "step": 62950 }, { "grad_norm": 0.14488081634044647, "learning_rate": 3.304910327649901e-05, "loss": 0.0118, "step": 62960 }, { "grad_norm": 0.13069364428520203, "learning_rate": 3.303354869177297e-05, "loss": 0.015, "step": 62970 }, { "grad_norm": 0.16967430710792542, "learning_rate": 3.301799596247255e-05, "loss": 0.0135, "step": 62980 }, { "grad_norm": 0.1831936091184616, "learning_rate": 3.300244509029854e-05, "loss": 0.0132, "step": 62990 }, { "grad_norm": 0.3990750312805176, "learning_rate": 3.2986896076951585e-05, "loss": 0.0181, "step": 63000 }, { "grad_norm": 0.16610828042030334, "learning_rate": 3.297134892413211e-05, "loss": 0.0131, "step": 63010 }, { "grad_norm": 0.453707218170166, "learning_rate": 3.295580363354031e-05, "loss": 0.0136, "step": 63020 }, { "grad_norm": 0.17895755171775818, "learning_rate": 3.2940260206876194e-05, "loss": 0.0121, "step": 63030 }, { "grad_norm": 0.17876246571540833, "learning_rate": 3.292471864583958e-05, "loss": 0.0131, "step": 63040 }, { "grad_norm": 0.13308966159820557, "learning_rate": 3.290917895213004e-05, "loss": 0.0137, "step": 63050 }, { "grad_norm": 0.13829663395881653, "learning_rate": 3.289364112744701e-05, "loss": 0.0155, "step": 63060 }, { "grad_norm": 0.18325380980968475, "learning_rate": 3.287810517348967e-05, "loss": 0.0167, "step": 63070 }, { "grad_norm": 0.19093821942806244, "learning_rate": 3.2862571091957014e-05, "loss": 0.0142, "step": 63080 }, { "grad_norm": 0.24670599400997162, "learning_rate": 3.284703888454781e-05, "loss": 0.0179, "step": 63090 }, { "grad_norm": 0.18707656860351562, "learning_rate": 3.2831508552960646e-05, "loss": 0.0164, "step": 63100 }, { "grad_norm": 0.1548464596271515, "learning_rate": 3.281598009889387e-05, "loss": 0.0121, "step": 63110 }, { "grad_norm": 0.17784705758094788, "learning_rate": 3.28004535240457e-05, "loss": 0.0136, "step": 63120 }, { "grad_norm": 0.14193019270896912, "learning_rate": 3.278492883011407e-05, "loss": 0.0105, "step": 63130 }, { "grad_norm": 0.17597146332263947, "learning_rate": 3.276940601879675e-05, "loss": 0.0124, "step": 63140 }, { "grad_norm": 0.14347124099731445, "learning_rate": 3.275388509179127e-05, "loss": 0.0129, "step": 63150 }, { "grad_norm": 0.16109177470207214, "learning_rate": 3.2738366050794985e-05, "loss": 0.0118, "step": 63160 }, { "grad_norm": 0.1615189164876938, "learning_rate": 3.2722848897505034e-05, "loss": 0.0133, "step": 63170 }, { "grad_norm": 0.1775059998035431, "learning_rate": 3.2707333633618356e-05, "loss": 0.0125, "step": 63180 }, { "grad_norm": 0.18160299956798553, "learning_rate": 3.269182026083167e-05, "loss": 0.0131, "step": 63190 }, { "grad_norm": 0.16176408529281616, "learning_rate": 3.2676308780841516e-05, "loss": 0.0147, "step": 63200 }, { "grad_norm": 0.17180930078029633, "learning_rate": 3.266079919534416e-05, "loss": 0.0145, "step": 63210 }, { "grad_norm": 0.14458662271499634, "learning_rate": 3.264529150603575e-05, "loss": 0.0136, "step": 63220 }, { "grad_norm": 0.163847416639328, "learning_rate": 3.262978571461217e-05, "loss": 0.0152, "step": 63230 }, { "grad_norm": 0.15867860615253448, "learning_rate": 3.26142818227691e-05, "loss": 0.0177, "step": 63240 }, { "grad_norm": 0.12442400306463242, "learning_rate": 3.259877983220203e-05, "loss": 0.0127, "step": 63250 }, { "grad_norm": 0.22166788578033447, "learning_rate": 3.2583279744606246e-05, "loss": 0.0117, "step": 63260 }, { "grad_norm": 0.16434535384178162, "learning_rate": 3.25677815616768e-05, "loss": 0.0149, "step": 63270 }, { "grad_norm": 0.15019406378269196, "learning_rate": 3.255228528510856e-05, "loss": 0.0148, "step": 63280 }, { "grad_norm": 0.17123839259147644, "learning_rate": 3.253679091659617e-05, "loss": 0.0137, "step": 63290 }, { "grad_norm": 0.12816914916038513, "learning_rate": 3.252129845783404e-05, "loss": 0.0149, "step": 63300 }, { "grad_norm": 0.14772123098373413, "learning_rate": 3.2505807910516456e-05, "loss": 0.014, "step": 63310 }, { "grad_norm": 0.15952764451503754, "learning_rate": 3.249031927633742e-05, "loss": 0.0179, "step": 63320 }, { "grad_norm": 0.16744464635849, "learning_rate": 3.2474832556990737e-05, "loss": 0.0118, "step": 63330 }, { "grad_norm": 0.21569418907165527, "learning_rate": 3.245934775417002e-05, "loss": 0.017, "step": 63340 }, { "grad_norm": 0.16594654321670532, "learning_rate": 3.244386486956866e-05, "loss": 0.0113, "step": 63350 }, { "grad_norm": 0.13211597502231598, "learning_rate": 3.2428383904879846e-05, "loss": 0.0168, "step": 63360 }, { "grad_norm": 0.15754571557044983, "learning_rate": 3.241290486179652e-05, "loss": 0.0123, "step": 63370 }, { "grad_norm": 0.21204924583435059, "learning_rate": 3.239742774201151e-05, "loss": 0.0154, "step": 63380 }, { "grad_norm": 0.15458300709724426, "learning_rate": 3.238195254721733e-05, "loss": 0.0173, "step": 63390 }, { "grad_norm": 0.1667734831571579, "learning_rate": 3.236647927910633e-05, "loss": 0.0145, "step": 63400 }, { "grad_norm": 0.20089776813983917, "learning_rate": 3.2351007939370646e-05, "loss": 0.0148, "step": 63410 }, { "grad_norm": 0.18867681920528412, "learning_rate": 3.2335538529702184e-05, "loss": 0.0131, "step": 63420 }, { "grad_norm": 0.1650790125131607, "learning_rate": 3.232007105179268e-05, "loss": 0.0119, "step": 63430 }, { "grad_norm": 0.16567462682724, "learning_rate": 3.230460550733363e-05, "loss": 0.0122, "step": 63440 }, { "grad_norm": 0.19824619591236115, "learning_rate": 3.2289141898016315e-05, "loss": 0.012, "step": 63450 }, { "grad_norm": 0.17180030047893524, "learning_rate": 3.227368022553181e-05, "loss": 0.0135, "step": 63460 }, { "grad_norm": 0.20897091925144196, "learning_rate": 3.225822049157099e-05, "loss": 0.0149, "step": 63470 }, { "grad_norm": 0.17022690176963806, "learning_rate": 3.2242762697824494e-05, "loss": 0.0133, "step": 63480 }, { "grad_norm": 0.16628699004650116, "learning_rate": 3.222730684598279e-05, "loss": 0.0134, "step": 63490 }, { "grad_norm": 0.1735064536333084, "learning_rate": 3.221185293773605e-05, "loss": 0.0154, "step": 63500 }, { "grad_norm": 0.18808309733867645, "learning_rate": 3.2196400974774354e-05, "loss": 0.0137, "step": 63510 }, { "grad_norm": 0.19552715122699738, "learning_rate": 3.2180950958787485e-05, "loss": 0.016, "step": 63520 }, { "grad_norm": 0.22698120772838593, "learning_rate": 3.2165502891465024e-05, "loss": 0.0151, "step": 63530 }, { "grad_norm": 0.17045322060585022, "learning_rate": 3.215005677449636e-05, "loss": 0.0132, "step": 63540 }, { "grad_norm": 0.15297634899616241, "learning_rate": 3.2134612609570644e-05, "loss": 0.0111, "step": 63550 }, { "grad_norm": 0.14710040390491486, "learning_rate": 3.211917039837682e-05, "loss": 0.0129, "step": 63560 }, { "grad_norm": 0.14950242638587952, "learning_rate": 3.210373014260365e-05, "loss": 0.0125, "step": 63570 }, { "grad_norm": 0.38391685485839844, "learning_rate": 3.208829184393965e-05, "loss": 0.0154, "step": 63580 }, { "grad_norm": 0.12379343807697296, "learning_rate": 3.207285550407312e-05, "loss": 0.0128, "step": 63590 }, { "grad_norm": 0.20071549713611603, "learning_rate": 3.205742112469217e-05, "loss": 0.0145, "step": 63600 }, { "grad_norm": 0.21251274645328522, "learning_rate": 3.204198870748465e-05, "loss": 0.0149, "step": 63610 }, { "grad_norm": 0.15662410855293274, "learning_rate": 3.202655825413824e-05, "loss": 0.0124, "step": 63620 }, { "grad_norm": 0.14418113231658936, "learning_rate": 3.2011129766340416e-05, "loss": 0.0114, "step": 63630 }, { "grad_norm": 0.16937512159347534, "learning_rate": 3.1995703245778384e-05, "loss": 0.0122, "step": 63640 }, { "grad_norm": 0.3477242588996887, "learning_rate": 3.1980278694139186e-05, "loss": 0.0163, "step": 63650 }, { "grad_norm": 0.1814921647310257, "learning_rate": 3.196485611310962e-05, "loss": 0.012, "step": 63660 }, { "grad_norm": 0.27288681268692017, "learning_rate": 3.1949435504376257e-05, "loss": 0.0116, "step": 63670 }, { "grad_norm": 0.17737989127635956, "learning_rate": 3.1934016869625505e-05, "loss": 0.0122, "step": 63680 }, { "grad_norm": 0.10528405010700226, "learning_rate": 3.191860021054349e-05, "loss": 0.0143, "step": 63690 }, { "grad_norm": 0.14294104278087616, "learning_rate": 3.190318552881618e-05, "loss": 0.0126, "step": 63700 }, { "grad_norm": 0.16566061973571777, "learning_rate": 3.1887772826129285e-05, "loss": 0.0127, "step": 63710 }, { "grad_norm": 0.1410883069038391, "learning_rate": 3.1872362104168335e-05, "loss": 0.0147, "step": 63720 }, { "grad_norm": 0.1514044851064682, "learning_rate": 3.185695336461861e-05, "loss": 0.0135, "step": 63730 }, { "grad_norm": 0.16013199090957642, "learning_rate": 3.184154660916519e-05, "loss": 0.0145, "step": 63740 }, { "grad_norm": 0.13533970713615417, "learning_rate": 3.1826141839492906e-05, "loss": 0.0134, "step": 63750 }, { "grad_norm": 0.16600465774536133, "learning_rate": 3.181073905728644e-05, "loss": 0.0142, "step": 63760 }, { "grad_norm": 0.14939016103744507, "learning_rate": 3.1795338264230214e-05, "loss": 0.0111, "step": 63770 }, { "grad_norm": 0.15749581158161163, "learning_rate": 3.1779939462008426e-05, "loss": 0.0112, "step": 63780 }, { "grad_norm": 0.15782681107521057, "learning_rate": 3.1764542652305066e-05, "loss": 0.0117, "step": 63790 }, { "grad_norm": 0.1943274289369583, "learning_rate": 3.17491478368039e-05, "loss": 0.0133, "step": 63800 }, { "grad_norm": 0.16878393292427063, "learning_rate": 3.173375501718847e-05, "loss": 0.0137, "step": 63810 }, { "grad_norm": 0.11878763884305954, "learning_rate": 3.171836419514215e-05, "loss": 0.0151, "step": 63820 }, { "grad_norm": 0.20616789162158966, "learning_rate": 3.170297537234803e-05, "loss": 0.0129, "step": 63830 }, { "grad_norm": 0.16326551139354706, "learning_rate": 3.168758855048901e-05, "loss": 0.0128, "step": 63840 }, { "grad_norm": 0.2309321016073227, "learning_rate": 3.1672203731247785e-05, "loss": 0.0144, "step": 63850 }, { "grad_norm": 0.1850614696741104, "learning_rate": 3.16568209163068e-05, "loss": 0.0127, "step": 63860 }, { "grad_norm": 0.16276687383651733, "learning_rate": 3.1641440107348295e-05, "loss": 0.0146, "step": 63870 }, { "grad_norm": 0.19327421486377716, "learning_rate": 3.162606130605429e-05, "loss": 0.0174, "step": 63880 }, { "grad_norm": 0.17135433852672577, "learning_rate": 3.16106845141066e-05, "loss": 0.0126, "step": 63890 }, { "grad_norm": 0.22043351829051971, "learning_rate": 3.159530973318682e-05, "loss": 0.0162, "step": 63900 }, { "grad_norm": 0.1650921255350113, "learning_rate": 3.1579936964976286e-05, "loss": 0.0174, "step": 63910 }, { "grad_norm": 0.18391111493110657, "learning_rate": 3.156456621115615e-05, "loss": 0.0138, "step": 63920 }, { "grad_norm": 0.15971840918064117, "learning_rate": 3.154919747340734e-05, "loss": 0.0148, "step": 63930 }, { "grad_norm": 0.13486601412296295, "learning_rate": 3.153383075341054e-05, "loss": 0.0111, "step": 63940 }, { "grad_norm": 0.2142617553472519, "learning_rate": 3.1518466052846264e-05, "loss": 0.0133, "step": 63950 }, { "grad_norm": 0.14180868864059448, "learning_rate": 3.150310337339474e-05, "loss": 0.015, "step": 63960 }, { "grad_norm": 0.15887728333473206, "learning_rate": 3.148774271673603e-05, "loss": 0.0132, "step": 63970 }, { "grad_norm": 0.1515006721019745, "learning_rate": 3.147238408454995e-05, "loss": 0.0121, "step": 63980 }, { "grad_norm": 0.1461212933063507, "learning_rate": 3.1457027478516086e-05, "loss": 0.0109, "step": 63990 }, { "grad_norm": 0.1464787721633911, "learning_rate": 3.144167290031382e-05, "loss": 0.0147, "step": 64000 }, { "grad_norm": 0.14656545221805573, "learning_rate": 3.142632035162227e-05, "loss": 0.0141, "step": 64010 }, { "grad_norm": 0.14317519962787628, "learning_rate": 3.141096983412044e-05, "loss": 0.0167, "step": 64020 }, { "grad_norm": 0.15917930006980896, "learning_rate": 3.139562134948699e-05, "loss": 0.0131, "step": 64030 }, { "grad_norm": 0.20216457545757294, "learning_rate": 3.1380274899400424e-05, "loss": 0.013, "step": 64040 }, { "grad_norm": 0.19303086400032043, "learning_rate": 3.1364930485539e-05, "loss": 0.0138, "step": 64050 }, { "grad_norm": 0.18089225888252258, "learning_rate": 3.134958810958076e-05, "loss": 0.0148, "step": 64060 }, { "grad_norm": 0.19197681546211243, "learning_rate": 3.13342477732035e-05, "loss": 0.0112, "step": 64070 }, { "grad_norm": 0.13448290526866913, "learning_rate": 3.131890947808487e-05, "loss": 0.0122, "step": 64080 }, { "grad_norm": 0.20614534616470337, "learning_rate": 3.13035732259022e-05, "loss": 0.0146, "step": 64090 }, { "grad_norm": 0.1713762730360031, "learning_rate": 3.1288239018332656e-05, "loss": 0.0114, "step": 64100 }, { "grad_norm": 0.1313023418188095, "learning_rate": 3.1272906857053164e-05, "loss": 0.0128, "step": 64110 }, { "grad_norm": 0.16583803296089172, "learning_rate": 3.125757674374042e-05, "loss": 0.015, "step": 64120 }, { "grad_norm": 0.12279835343360901, "learning_rate": 3.124224868007088e-05, "loss": 0.0103, "step": 64130 }, { "grad_norm": 0.17849290370941162, "learning_rate": 3.1226922667720854e-05, "loss": 0.0116, "step": 64140 }, { "grad_norm": 0.16722208261489868, "learning_rate": 3.121159870836633e-05, "loss": 0.014, "step": 64150 }, { "grad_norm": 0.1938193440437317, "learning_rate": 3.119627680368312e-05, "loss": 0.0139, "step": 64160 }, { "grad_norm": 0.15684424340724945, "learning_rate": 3.1180956955346816e-05, "loss": 0.0154, "step": 64170 }, { "grad_norm": 0.09969140589237213, "learning_rate": 3.116563916503276e-05, "loss": 0.0116, "step": 64180 }, { "grad_norm": 0.1774163395166397, "learning_rate": 3.1150323434416095e-05, "loss": 0.0106, "step": 64190 }, { "grad_norm": 0.1958913505077362, "learning_rate": 3.1135009765171696e-05, "loss": 0.0124, "step": 64200 }, { "grad_norm": 0.17787224054336548, "learning_rate": 3.111969815897429e-05, "loss": 0.0132, "step": 64210 }, { "grad_norm": 0.1691119223833084, "learning_rate": 3.1104388617498295e-05, "loss": 0.0117, "step": 64220 }, { "grad_norm": 0.08364468812942505, "learning_rate": 3.1089081142417966e-05, "loss": 0.0135, "step": 64230 }, { "grad_norm": 0.11788904666900635, "learning_rate": 3.1073775735407285e-05, "loss": 0.013, "step": 64240 }, { "grad_norm": 0.16925007104873657, "learning_rate": 3.1058472398140035e-05, "loss": 0.0137, "step": 64250 }, { "grad_norm": 0.14944908022880554, "learning_rate": 3.1043171132289736e-05, "loss": 0.0127, "step": 64260 }, { "grad_norm": 0.16053886711597443, "learning_rate": 3.1027871939529774e-05, "loss": 0.0136, "step": 64270 }, { "grad_norm": 0.13091439008712769, "learning_rate": 3.1012574821533214e-05, "loss": 0.0117, "step": 64280 }, { "grad_norm": 0.14141909778118134, "learning_rate": 3.099727977997291e-05, "loss": 0.0175, "step": 64290 }, { "grad_norm": 0.16568678617477417, "learning_rate": 3.098198681652154e-05, "loss": 0.0107, "step": 64300 }, { "grad_norm": 0.18529826402664185, "learning_rate": 3.0966695932851475e-05, "loss": 0.0134, "step": 64310 }, { "grad_norm": 0.1617167890071869, "learning_rate": 3.0951407130634926e-05, "loss": 0.0134, "step": 64320 }, { "grad_norm": 0.1195993646979332, "learning_rate": 3.0936120411543864e-05, "loss": 0.0131, "step": 64330 }, { "grad_norm": 0.14983133971691132, "learning_rate": 3.0920835777250015e-05, "loss": 0.0141, "step": 64340 }, { "grad_norm": 0.20581376552581787, "learning_rate": 3.090555322942488e-05, "loss": 0.0136, "step": 64350 }, { "grad_norm": 0.11486692726612091, "learning_rate": 3.089027276973973e-05, "loss": 0.0133, "step": 64360 }, { "grad_norm": 0.13086792826652527, "learning_rate": 3.0874994399865615e-05, "loss": 0.012, "step": 64370 }, { "grad_norm": 0.13336077332496643, "learning_rate": 3.085971812147337e-05, "loss": 0.0125, "step": 64380 }, { "grad_norm": 0.16717271506786346, "learning_rate": 3.0844443936233557e-05, "loss": 0.0128, "step": 64390 }, { "grad_norm": 0.1243894174695015, "learning_rate": 3.082917184581657e-05, "loss": 0.0129, "step": 64400 }, { "grad_norm": 0.10867144167423248, "learning_rate": 3.0813901851892513e-05, "loss": 0.0157, "step": 64410 }, { "grad_norm": 0.13763481378555298, "learning_rate": 3.0798633956131316e-05, "loss": 0.015, "step": 64420 }, { "grad_norm": 0.15158243477344513, "learning_rate": 3.078336816020264e-05, "loss": 0.0131, "step": 64430 }, { "grad_norm": 0.162523090839386, "learning_rate": 3.0768104465775935e-05, "loss": 0.0141, "step": 64440 }, { "grad_norm": 0.13080503046512604, "learning_rate": 3.0752842874520396e-05, "loss": 0.0141, "step": 64450 }, { "grad_norm": 0.17385664582252502, "learning_rate": 3.073758338810503e-05, "loss": 0.016, "step": 64460 }, { "grad_norm": 0.13508890569210052, "learning_rate": 3.072232600819859e-05, "loss": 0.0154, "step": 64470 }, { "grad_norm": 0.17966429889202118, "learning_rate": 3.070707073646959e-05, "loss": 0.0163, "step": 64480 }, { "grad_norm": 0.12782929837703705, "learning_rate": 3.069181757458633e-05, "loss": 0.0112, "step": 64490 }, { "grad_norm": 0.15666157007217407, "learning_rate": 3.067656652421687e-05, "loss": 0.0119, "step": 64500 }, { "grad_norm": 0.19124214351177216, "learning_rate": 3.0661317587029025e-05, "loss": 0.0136, "step": 64510 }, { "grad_norm": 0.1692783534526825, "learning_rate": 3.064607076469043e-05, "loss": 0.0143, "step": 64520 }, { "grad_norm": 0.12785546481609344, "learning_rate": 3.063082605886845e-05, "loss": 0.0147, "step": 64530 }, { "grad_norm": 0.17412222921848297, "learning_rate": 3.061558347123019e-05, "loss": 0.012, "step": 64540 }, { "grad_norm": 0.19748318195343018, "learning_rate": 3.060034300344259e-05, "loss": 0.0146, "step": 64550 }, { "grad_norm": 0.1892365664243698, "learning_rate": 3.058510465717231e-05, "loss": 0.0146, "step": 64560 }, { "grad_norm": 0.15138332545757294, "learning_rate": 3.0569868434085804e-05, "loss": 0.0123, "step": 64570 }, { "grad_norm": 0.14289595186710358, "learning_rate": 3.055463433584924e-05, "loss": 0.0169, "step": 64580 }, { "grad_norm": 0.13890261948108673, "learning_rate": 3.0539402364128656e-05, "loss": 0.0129, "step": 64590 }, { "grad_norm": 0.18046315014362335, "learning_rate": 3.052417252058977e-05, "loss": 0.0151, "step": 64600 }, { "grad_norm": 0.1058596670627594, "learning_rate": 3.0508944806898092e-05, "loss": 0.0127, "step": 64610 }, { "grad_norm": 0.194846048951149, "learning_rate": 3.04937192247189e-05, "loss": 0.0137, "step": 64620 }, { "grad_norm": 0.12435431778430939, "learning_rate": 3.047849577571725e-05, "loss": 0.0116, "step": 64630 }, { "grad_norm": 0.1554846465587616, "learning_rate": 3.046327446155795e-05, "loss": 0.0119, "step": 64640 }, { "grad_norm": 0.2224925309419632, "learning_rate": 3.0448055283905587e-05, "loss": 0.0133, "step": 64650 }, { "grad_norm": 0.1818966269493103, "learning_rate": 3.0432838244424494e-05, "loss": 0.0149, "step": 64660 }, { "grad_norm": 0.1957448422908783, "learning_rate": 3.04176233447788e-05, "loss": 0.0163, "step": 64670 }, { "grad_norm": 0.18886734545230865, "learning_rate": 3.040241058663238e-05, "loss": 0.0137, "step": 64680 }, { "grad_norm": 0.19792303442955017, "learning_rate": 3.038719997164887e-05, "loss": 0.0151, "step": 64690 }, { "grad_norm": 0.14201503992080688, "learning_rate": 3.0371991501491683e-05, "loss": 0.0167, "step": 64700 }, { "grad_norm": 0.15116523206233978, "learning_rate": 3.0356785177823975e-05, "loss": 0.0146, "step": 64710 }, { "grad_norm": 0.13811707496643066, "learning_rate": 3.0341581002308726e-05, "loss": 0.0141, "step": 64720 }, { "grad_norm": 0.11262039840221405, "learning_rate": 3.0326378976608627e-05, "loss": 0.0139, "step": 64730 }, { "grad_norm": 0.19709669053554535, "learning_rate": 3.0311179102386146e-05, "loss": 0.0142, "step": 64740 }, { "grad_norm": 0.14443808794021606, "learning_rate": 3.0295981381303513e-05, "loss": 0.0153, "step": 64750 }, { "grad_norm": 0.16020341217517853, "learning_rate": 3.0280785815022727e-05, "loss": 0.0149, "step": 64760 }, { "grad_norm": 0.18724580109119415, "learning_rate": 3.0265592405205544e-05, "loss": 0.013, "step": 64770 }, { "grad_norm": 0.23120684921741486, "learning_rate": 3.0250401153513517e-05, "loss": 0.0172, "step": 64780 }, { "grad_norm": 0.18609078228473663, "learning_rate": 3.0235212061607936e-05, "loss": 0.0152, "step": 64790 }, { "grad_norm": 0.1759921759366989, "learning_rate": 3.0220025131149842e-05, "loss": 0.0146, "step": 64800 }, { "grad_norm": 0.18972069025039673, "learning_rate": 3.020484036380007e-05, "loss": 0.019, "step": 64810 }, { "grad_norm": 0.15630778670310974, "learning_rate": 3.0189657761219188e-05, "loss": 0.0133, "step": 64820 }, { "grad_norm": 0.18669627606868744, "learning_rate": 3.017447732506753e-05, "loss": 0.0131, "step": 64830 }, { "grad_norm": 0.183602973818779, "learning_rate": 3.015929905700524e-05, "loss": 0.0156, "step": 64840 }, { "grad_norm": 0.12057602405548096, "learning_rate": 3.014412295869218e-05, "loss": 0.0143, "step": 64850 }, { "grad_norm": 0.18696962296962738, "learning_rate": 3.0128949031787977e-05, "loss": 0.0128, "step": 64860 }, { "grad_norm": 0.12890543043613434, "learning_rate": 3.0113777277952022e-05, "loss": 0.0118, "step": 64870 }, { "grad_norm": 0.10220430046319962, "learning_rate": 3.0098607698843496e-05, "loss": 0.0125, "step": 64880 }, { "grad_norm": 0.11552568525075912, "learning_rate": 3.0083440296121312e-05, "loss": 0.0143, "step": 64890 }, { "grad_norm": 0.16275888681411743, "learning_rate": 3.0068275071444142e-05, "loss": 0.0131, "step": 64900 }, { "grad_norm": 0.3122219145298004, "learning_rate": 3.0053112026470442e-05, "loss": 0.016, "step": 64910 }, { "grad_norm": 0.14087025821208954, "learning_rate": 3.0037951162858434e-05, "loss": 0.0158, "step": 64920 }, { "grad_norm": 0.15824097394943237, "learning_rate": 3.002279248226607e-05, "loss": 0.0149, "step": 64930 }, { "grad_norm": 0.19562511146068573, "learning_rate": 3.000763598635109e-05, "loss": 0.0156, "step": 64940 }, { "grad_norm": 0.20846189558506012, "learning_rate": 2.999248167677098e-05, "loss": 0.0146, "step": 64950 }, { "grad_norm": 0.15373951196670532, "learning_rate": 2.9977329555182976e-05, "loss": 0.0145, "step": 64960 }, { "grad_norm": 0.1394064724445343, "learning_rate": 2.9962179623244124e-05, "loss": 0.013, "step": 64970 }, { "grad_norm": 0.1673436015844345, "learning_rate": 2.9947031882611175e-05, "loss": 0.0125, "step": 64980 }, { "grad_norm": 0.14847539365291595, "learning_rate": 2.9931886334940685e-05, "loss": 0.0133, "step": 64990 }, { "grad_norm": 0.16229791939258575, "learning_rate": 2.9916742981888923e-05, "loss": 0.0122, "step": 65000 }, { "grad_norm": 0.15305674076080322, "learning_rate": 2.990160182511196e-05, "loss": 0.0136, "step": 65010 }, { "grad_norm": 0.16733235120773315, "learning_rate": 2.9886462866265574e-05, "loss": 0.0138, "step": 65020 }, { "grad_norm": 0.17003104090690613, "learning_rate": 2.9871326107005392e-05, "loss": 0.0135, "step": 65030 }, { "grad_norm": 0.1990867406129837, "learning_rate": 2.9856191548986735e-05, "loss": 0.0167, "step": 65040 }, { "grad_norm": 0.21981187164783478, "learning_rate": 2.9841059193864672e-05, "loss": 0.0146, "step": 65050 }, { "grad_norm": 0.13497938215732574, "learning_rate": 2.982592904329407e-05, "loss": 0.0109, "step": 65060 }, { "grad_norm": 0.23492549359798431, "learning_rate": 2.981080109892954e-05, "loss": 0.0192, "step": 65070 }, { "grad_norm": 0.19638188183307648, "learning_rate": 2.9795675362425434e-05, "loss": 0.0152, "step": 65080 }, { "grad_norm": 0.20794767141342163, "learning_rate": 2.97805518354359e-05, "loss": 0.0139, "step": 65090 }, { "grad_norm": 0.1821189820766449, "learning_rate": 2.9765430519614825e-05, "loss": 0.0134, "step": 65100 }, { "grad_norm": 0.31127792596817017, "learning_rate": 2.975031141661584e-05, "loss": 0.0143, "step": 65110 }, { "grad_norm": 0.09403619170188904, "learning_rate": 2.9735194528092347e-05, "loss": 0.0119, "step": 65120 }, { "grad_norm": 0.2480572611093521, "learning_rate": 2.9720079855697525e-05, "loss": 0.0122, "step": 65130 }, { "grad_norm": 0.26211822032928467, "learning_rate": 2.970496740108427e-05, "loss": 0.0151, "step": 65140 }, { "grad_norm": 0.28617191314697266, "learning_rate": 2.968985716590527e-05, "loss": 0.016, "step": 65150 }, { "grad_norm": 0.20108020305633545, "learning_rate": 2.9674749151812953e-05, "loss": 0.0154, "step": 65160 }, { "grad_norm": 0.1497776210308075, "learning_rate": 2.965964336045951e-05, "loss": 0.0115, "step": 65170 }, { "grad_norm": 0.1326286494731903, "learning_rate": 2.9644539793496894e-05, "loss": 0.0154, "step": 65180 }, { "grad_norm": 0.14033246040344238, "learning_rate": 2.9629438452576807e-05, "loss": 0.0146, "step": 65190 }, { "grad_norm": 0.17427460849285126, "learning_rate": 2.9614339339350698e-05, "loss": 0.015, "step": 65200 }, { "grad_norm": 0.22759415209293365, "learning_rate": 2.959924245546979e-05, "loss": 0.0122, "step": 65210 }, { "grad_norm": 0.1358422487974167, "learning_rate": 2.9584147802585038e-05, "loss": 0.0129, "step": 65220 }, { "grad_norm": 0.16938741505146027, "learning_rate": 2.956905538234721e-05, "loss": 0.0126, "step": 65230 }, { "grad_norm": 0.1423901468515396, "learning_rate": 2.9553965196406764e-05, "loss": 0.0117, "step": 65240 }, { "grad_norm": 0.1514158844947815, "learning_rate": 2.9538877246413943e-05, "loss": 0.0167, "step": 65250 }, { "grad_norm": 0.09161584824323654, "learning_rate": 2.952379153401874e-05, "loss": 0.016, "step": 65260 }, { "grad_norm": 0.14806371927261353, "learning_rate": 2.9508708060870914e-05, "loss": 0.0124, "step": 65270 }, { "grad_norm": 0.2360396832227707, "learning_rate": 2.9493626828619948e-05, "loss": 0.013, "step": 65280 }, { "grad_norm": 0.16464534401893616, "learning_rate": 2.9478547838915137e-05, "loss": 0.0112, "step": 65290 }, { "grad_norm": 0.20866303145885468, "learning_rate": 2.946347109340548e-05, "loss": 0.012, "step": 65300 }, { "grad_norm": 0.14987051486968994, "learning_rate": 2.944839659373976e-05, "loss": 0.0122, "step": 65310 }, { "grad_norm": 0.2133064717054367, "learning_rate": 2.943332434156647e-05, "loss": 0.0157, "step": 65320 }, { "grad_norm": 0.16176027059555054, "learning_rate": 2.9418254338533912e-05, "loss": 0.0152, "step": 65330 }, { "grad_norm": 0.1710561364889145, "learning_rate": 2.940318658629011e-05, "loss": 0.0129, "step": 65340 }, { "grad_norm": 0.16863681375980377, "learning_rate": 2.938812108648287e-05, "loss": 0.0155, "step": 65350 }, { "grad_norm": 0.15179206430912018, "learning_rate": 2.937305784075972e-05, "loss": 0.0137, "step": 65360 }, { "grad_norm": 0.17528825998306274, "learning_rate": 2.9357996850767937e-05, "loss": 0.0147, "step": 65370 }, { "grad_norm": 0.1482807695865631, "learning_rate": 2.9342938118154595e-05, "loss": 0.0117, "step": 65380 }, { "grad_norm": 0.1212809830904007, "learning_rate": 2.9327881644566485e-05, "loss": 0.0121, "step": 65390 }, { "grad_norm": 0.1599014699459076, "learning_rate": 2.9312827431650157e-05, "loss": 0.0149, "step": 65400 }, { "grad_norm": 0.155538871884346, "learning_rate": 2.92977754810519e-05, "loss": 0.0123, "step": 65410 }, { "grad_norm": 0.15317554771900177, "learning_rate": 2.928272579441781e-05, "loss": 0.0134, "step": 65420 }, { "grad_norm": 0.16400015354156494, "learning_rate": 2.9267678373393676e-05, "loss": 0.0142, "step": 65430 }, { "grad_norm": 0.16414134204387665, "learning_rate": 2.925263321962507e-05, "loss": 0.0128, "step": 65440 }, { "grad_norm": 0.21177640557289124, "learning_rate": 2.9237590334757292e-05, "loss": 0.0133, "step": 65450 }, { "grad_norm": 0.17551100254058838, "learning_rate": 2.922254972043543e-05, "loss": 0.0134, "step": 65460 }, { "grad_norm": 0.14051571488380432, "learning_rate": 2.9207511378304265e-05, "loss": 0.011, "step": 65470 }, { "grad_norm": 0.3009869158267975, "learning_rate": 2.9192475310008428e-05, "loss": 0.0122, "step": 65480 }, { "grad_norm": 0.20530833303928375, "learning_rate": 2.917744151719219e-05, "loss": 0.0171, "step": 65490 }, { "grad_norm": 0.15153948962688446, "learning_rate": 2.916241000149964e-05, "loss": 0.0123, "step": 65500 }, { "grad_norm": 0.15117789804935455, "learning_rate": 2.9147380764574622e-05, "loss": 0.0121, "step": 65510 }, { "grad_norm": 0.15781418979167938, "learning_rate": 2.9132353808060674e-05, "loss": 0.0159, "step": 65520 }, { "grad_norm": 0.11536809802055359, "learning_rate": 2.9117329133601134e-05, "loss": 0.0151, "step": 65530 }, { "grad_norm": 0.19876669347286224, "learning_rate": 2.9102306742839093e-05, "loss": 0.0136, "step": 65540 }, { "grad_norm": 0.14298051595687866, "learning_rate": 2.9087286637417375e-05, "loss": 0.0137, "step": 65550 }, { "grad_norm": 0.2133745700120926, "learning_rate": 2.9072268818978537e-05, "loss": 0.0123, "step": 65560 }, { "grad_norm": 0.14992593228816986, "learning_rate": 2.905725328916493e-05, "loss": 0.012, "step": 65570 }, { "grad_norm": 0.12431662529706955, "learning_rate": 2.9042240049618608e-05, "loss": 0.0136, "step": 65580 }, { "grad_norm": 0.19760125875473022, "learning_rate": 2.9027229101981413e-05, "loss": 0.0141, "step": 65590 }, { "grad_norm": 0.18866200745105743, "learning_rate": 2.901222044789488e-05, "loss": 0.0122, "step": 65600 }, { "grad_norm": 0.24007274210453033, "learning_rate": 2.89972140890004e-05, "loss": 0.0101, "step": 65610 }, { "grad_norm": 0.1843998283147812, "learning_rate": 2.8982210026938988e-05, "loss": 0.0128, "step": 65620 }, { "grad_norm": 0.18659861385822296, "learning_rate": 2.896720826335151e-05, "loss": 0.0131, "step": 65630 }, { "grad_norm": 0.16105516254901886, "learning_rate": 2.8952208799878493e-05, "loss": 0.0103, "step": 65640 }, { "grad_norm": 0.20193131268024445, "learning_rate": 2.89372116381603e-05, "loss": 0.0147, "step": 65650 }, { "grad_norm": 0.16647164523601532, "learning_rate": 2.8922216779836932e-05, "loss": 0.0137, "step": 65660 }, { "grad_norm": 0.14113597571849823, "learning_rate": 2.8907224226548285e-05, "loss": 0.0118, "step": 65670 }, { "grad_norm": 0.14418774843215942, "learning_rate": 2.8892233979933868e-05, "loss": 0.0114, "step": 65680 }, { "grad_norm": 0.13190485537052155, "learning_rate": 2.8877246041633026e-05, "loss": 0.011, "step": 65690 }, { "grad_norm": 0.1355598419904709, "learning_rate": 2.886226041328478e-05, "loss": 0.0122, "step": 65700 }, { "grad_norm": 0.12205629795789719, "learning_rate": 2.8847277096527957e-05, "loss": 0.0122, "step": 65710 }, { "grad_norm": 0.18449950218200684, "learning_rate": 2.8832296093001105e-05, "loss": 0.0107, "step": 65720 }, { "grad_norm": 0.15264250338077545, "learning_rate": 2.8817317404342547e-05, "loss": 0.0133, "step": 65730 }, { "grad_norm": 0.1487533450126648, "learning_rate": 2.880234103219029e-05, "loss": 0.0121, "step": 65740 }, { "grad_norm": 0.17688432335853577, "learning_rate": 2.878736697818215e-05, "loss": 0.0129, "step": 65750 }, { "grad_norm": 0.2735940217971802, "learning_rate": 2.877239524395568e-05, "loss": 0.0123, "step": 65760 }, { "grad_norm": 0.220130056142807, "learning_rate": 2.8757425831148134e-05, "loss": 0.0136, "step": 65770 }, { "grad_norm": 0.17443376779556274, "learning_rate": 2.874245874139658e-05, "loss": 0.0122, "step": 65780 }, { "grad_norm": 0.1890283077955246, "learning_rate": 2.8727493976337737e-05, "loss": 0.0125, "step": 65790 }, { "grad_norm": 0.17709963023662567, "learning_rate": 2.8712531537608218e-05, "loss": 0.0143, "step": 65800 }, { "grad_norm": 0.09788314998149872, "learning_rate": 2.8697571426844217e-05, "loss": 0.0103, "step": 65810 }, { "grad_norm": 0.12689641118049622, "learning_rate": 2.86826136456818e-05, "loss": 0.0111, "step": 65820 }, { "grad_norm": 0.15117789804935455, "learning_rate": 2.8667658195756685e-05, "loss": 0.0146, "step": 65830 }, { "grad_norm": 0.19587625563144684, "learning_rate": 2.8652705078704412e-05, "loss": 0.0126, "step": 65840 }, { "grad_norm": 0.15492109954357147, "learning_rate": 2.863775429616018e-05, "loss": 0.0129, "step": 65850 }, { "grad_norm": 0.18497496843338013, "learning_rate": 2.862280584975906e-05, "loss": 0.0115, "step": 65860 }, { "grad_norm": 0.18948151171207428, "learning_rate": 2.8607859741135735e-05, "loss": 0.0129, "step": 65870 }, { "grad_norm": 0.20644602179527283, "learning_rate": 2.8592915971924723e-05, "loss": 0.0157, "step": 65880 }, { "grad_norm": 0.19122640788555145, "learning_rate": 2.8577974543760212e-05, "loss": 0.0111, "step": 65890 }, { "grad_norm": 0.16785795986652374, "learning_rate": 2.8563035458276222e-05, "loss": 0.0128, "step": 65900 }, { "grad_norm": 0.13208386301994324, "learning_rate": 2.8548098717106424e-05, "loss": 0.0126, "step": 65910 }, { "grad_norm": 0.4122697114944458, "learning_rate": 2.85331643218843e-05, "loss": 0.0122, "step": 65920 }, { "grad_norm": 0.1497766226530075, "learning_rate": 2.8518232274243046e-05, "loss": 0.012, "step": 65930 }, { "grad_norm": 0.16306884586811066, "learning_rate": 2.8503302575815638e-05, "loss": 0.0134, "step": 65940 }, { "grad_norm": 0.20417813956737518, "learning_rate": 2.8488375228234716e-05, "loss": 0.0154, "step": 65950 }, { "grad_norm": 0.1704529970884323, "learning_rate": 2.8473450233132738e-05, "loss": 0.0118, "step": 65960 }, { "grad_norm": 0.1848984956741333, "learning_rate": 2.8458527592141898e-05, "loss": 0.014, "step": 65970 }, { "grad_norm": 0.14387983083724976, "learning_rate": 2.8443607306894073e-05, "loss": 0.011, "step": 65980 }, { "grad_norm": 0.24712949991226196, "learning_rate": 2.8428689379020946e-05, "loss": 0.0148, "step": 65990 }, { "grad_norm": 0.16575519740581512, "learning_rate": 2.841377381015391e-05, "loss": 0.0118, "step": 66000 }, { "grad_norm": 0.16710227727890015, "learning_rate": 2.8398860601924144e-05, "loss": 0.0126, "step": 66010 }, { "grad_norm": 0.17427627742290497, "learning_rate": 2.838394975596248e-05, "loss": 0.0124, "step": 66020 }, { "grad_norm": 0.13647310435771942, "learning_rate": 2.83690412738996e-05, "loss": 0.0121, "step": 66030 }, { "grad_norm": 0.1884990930557251, "learning_rate": 2.835413515736579e-05, "loss": 0.0127, "step": 66040 }, { "grad_norm": 0.19014157354831696, "learning_rate": 2.8339231407991267e-05, "loss": 0.0162, "step": 66050 }, { "grad_norm": 0.11426199972629547, "learning_rate": 2.832433002740581e-05, "loss": 0.0112, "step": 66060 }, { "grad_norm": 0.1555793136358261, "learning_rate": 2.8309431017239047e-05, "loss": 0.0133, "step": 66070 }, { "grad_norm": 0.12404493987560272, "learning_rate": 2.829453437912028e-05, "loss": 0.0144, "step": 66080 }, { "grad_norm": 0.16796542704105377, "learning_rate": 2.8279640114678617e-05, "loss": 0.0127, "step": 66090 }, { "grad_norm": 0.15411916375160217, "learning_rate": 2.8264748225542836e-05, "loss": 0.0122, "step": 66100 }, { "grad_norm": 0.21563616394996643, "learning_rate": 2.8249858713341516e-05, "loss": 0.0142, "step": 66110 }, { "grad_norm": 0.1754365712404251, "learning_rate": 2.823497157970294e-05, "loss": 0.0122, "step": 66120 }, { "grad_norm": 0.5994186997413635, "learning_rate": 2.8220086826255164e-05, "loss": 0.0151, "step": 66130 }, { "grad_norm": 0.16315817832946777, "learning_rate": 2.8205204454625934e-05, "loss": 0.0123, "step": 66140 }, { "grad_norm": 0.14734621345996857, "learning_rate": 2.8190324466442798e-05, "loss": 0.0131, "step": 66150 }, { "grad_norm": 0.1469440907239914, "learning_rate": 2.8175446863332965e-05, "loss": 0.0156, "step": 66160 }, { "grad_norm": 0.17208640277385712, "learning_rate": 2.816057164692345e-05, "loss": 0.0138, "step": 66170 }, { "grad_norm": 0.15866714715957642, "learning_rate": 2.8145698818840992e-05, "loss": 0.0153, "step": 66180 }, { "grad_norm": 0.1788797527551651, "learning_rate": 2.8130828380712065e-05, "loss": 0.0145, "step": 66190 }, { "grad_norm": 0.16307349503040314, "learning_rate": 2.811596033416285e-05, "loss": 0.0124, "step": 66200 }, { "grad_norm": 0.16006141901016235, "learning_rate": 2.8101094680819307e-05, "loss": 0.0137, "step": 66210 }, { "grad_norm": 0.17710180580615997, "learning_rate": 2.808623142230715e-05, "loss": 0.0131, "step": 66220 }, { "grad_norm": 0.14566098153591156, "learning_rate": 2.8071370560251757e-05, "loss": 0.0115, "step": 66230 }, { "grad_norm": 0.14867743849754333, "learning_rate": 2.805651209627831e-05, "loss": 0.0131, "step": 66240 }, { "grad_norm": 0.16059713065624237, "learning_rate": 2.804165603201171e-05, "loss": 0.0127, "step": 66250 }, { "grad_norm": 0.2153378576040268, "learning_rate": 2.802680236907661e-05, "loss": 0.0109, "step": 66260 }, { "grad_norm": 0.2627370357513428, "learning_rate": 2.8011951109097344e-05, "loss": 0.014, "step": 66270 }, { "grad_norm": 0.1415649950504303, "learning_rate": 2.7997102253698072e-05, "loss": 0.0183, "step": 66280 }, { "grad_norm": 0.1458180993795395, "learning_rate": 2.7982255804502588e-05, "loss": 0.014, "step": 66290 }, { "grad_norm": 0.14285974204540253, "learning_rate": 2.7967411763134516e-05, "loss": 0.0129, "step": 66300 }, { "grad_norm": 0.14952103793621063, "learning_rate": 2.7952570131217166e-05, "loss": 0.0119, "step": 66310 }, { "grad_norm": 0.21348465979099274, "learning_rate": 2.7937730910373618e-05, "loss": 0.0152, "step": 66320 }, { "grad_norm": 0.16426178812980652, "learning_rate": 2.7922894102226628e-05, "loss": 0.0138, "step": 66330 }, { "grad_norm": 0.1972154825925827, "learning_rate": 2.790805970839876e-05, "loss": 0.017, "step": 66340 }, { "grad_norm": 0.1481200009584427, "learning_rate": 2.789322773051225e-05, "loss": 0.0115, "step": 66350 }, { "grad_norm": 0.15331314504146576, "learning_rate": 2.7878398170189124e-05, "loss": 0.0142, "step": 66360 }, { "grad_norm": 0.13549210131168365, "learning_rate": 2.7863571029051112e-05, "loss": 0.0116, "step": 66370 }, { "grad_norm": 0.1724197417497635, "learning_rate": 2.784874630871971e-05, "loss": 0.0126, "step": 66380 }, { "grad_norm": 0.17023910582065582, "learning_rate": 2.7833924010816086e-05, "loss": 0.0117, "step": 66390 }, { "grad_norm": 0.1660158932209015, "learning_rate": 2.7819104136961226e-05, "loss": 0.0134, "step": 66400 }, { "grad_norm": 0.146593376994133, "learning_rate": 2.780428668877577e-05, "loss": 0.0165, "step": 66410 }, { "grad_norm": 0.1491086632013321, "learning_rate": 2.7789471667880152e-05, "loss": 0.0144, "step": 66420 }, { "grad_norm": 0.167659729719162, "learning_rate": 2.7774659075894523e-05, "loss": 0.0132, "step": 66430 }, { "grad_norm": 0.1443862020969391, "learning_rate": 2.7759848914438778e-05, "loss": 0.0144, "step": 66440 }, { "grad_norm": 0.17906910181045532, "learning_rate": 2.7745041185132496e-05, "loss": 0.0131, "step": 66450 }, { "grad_norm": 0.15001365542411804, "learning_rate": 2.773023588959506e-05, "loss": 0.0126, "step": 66460 }, { "grad_norm": 0.15929411351680756, "learning_rate": 2.7715433029445555e-05, "loss": 0.0133, "step": 66470 }, { "grad_norm": 0.13926854729652405, "learning_rate": 2.7700632606302766e-05, "loss": 0.0123, "step": 66480 }, { "grad_norm": 0.15253542363643646, "learning_rate": 2.768583462178528e-05, "loss": 0.0114, "step": 66490 }, { "grad_norm": 0.16215547919273376, "learning_rate": 2.767103907751136e-05, "loss": 0.0141, "step": 66500 }, { "grad_norm": 0.1362875998020172, "learning_rate": 2.7656245975099065e-05, "loss": 0.0174, "step": 66510 }, { "grad_norm": 0.12334734946489334, "learning_rate": 2.7641455316166093e-05, "loss": 0.0114, "step": 66520 }, { "grad_norm": 0.10871999710798264, "learning_rate": 2.7626667102329973e-05, "loss": 0.0103, "step": 66530 }, { "grad_norm": 0.21508333086967468, "learning_rate": 2.761188133520788e-05, "loss": 0.0166, "step": 66540 }, { "grad_norm": 0.16290561854839325, "learning_rate": 2.7597098016416784e-05, "loss": 0.013, "step": 66550 }, { "grad_norm": 0.1378319412469864, "learning_rate": 2.7582317147573368e-05, "loss": 0.0135, "step": 66560 }, { "grad_norm": 0.11538195610046387, "learning_rate": 2.7567538730294064e-05, "loss": 0.0133, "step": 66570 }, { "grad_norm": 0.16145919263362885, "learning_rate": 2.7552762766194973e-05, "loss": 0.0146, "step": 66580 }, { "grad_norm": 0.14460699260234833, "learning_rate": 2.753798925689202e-05, "loss": 0.0129, "step": 66590 }, { "grad_norm": 0.11122231185436249, "learning_rate": 2.7523218204000774e-05, "loss": 0.0124, "step": 66600 }, { "grad_norm": 0.1926685869693756, "learning_rate": 2.7508449609136612e-05, "loss": 0.0142, "step": 66610 }, { "grad_norm": 0.15939952433109283, "learning_rate": 2.749368347391454e-05, "loss": 0.015, "step": 66620 }, { "grad_norm": 0.14714768528938293, "learning_rate": 2.7478919799949443e-05, "loss": 0.0113, "step": 66630 }, { "grad_norm": 0.22542330622673035, "learning_rate": 2.7464158588855803e-05, "loss": 0.0117, "step": 66640 }, { "grad_norm": 0.21047601103782654, "learning_rate": 2.7449399842247918e-05, "loss": 0.0174, "step": 66650 }, { "grad_norm": 0.22999289631843567, "learning_rate": 2.7434643561739736e-05, "loss": 0.0133, "step": 66660 }, { "grad_norm": 0.1569344401359558, "learning_rate": 2.7419889748945016e-05, "loss": 0.0121, "step": 66670 }, { "grad_norm": 0.1856958419084549, "learning_rate": 2.7405138405477203e-05, "loss": 0.0131, "step": 66680 }, { "grad_norm": 0.15863072872161865, "learning_rate": 2.73903895329495e-05, "loss": 0.0109, "step": 66690 }, { "grad_norm": 0.19878096878528595, "learning_rate": 2.737564313297478e-05, "loss": 0.0134, "step": 66700 }, { "grad_norm": 0.18886269629001617, "learning_rate": 2.736089920716572e-05, "loss": 0.0123, "step": 66710 }, { "grad_norm": 0.30835822224617004, "learning_rate": 2.7346157757134706e-05, "loss": 0.0138, "step": 66720 }, { "grad_norm": 0.19286739826202393, "learning_rate": 2.733141878449379e-05, "loss": 0.0133, "step": 66730 }, { "grad_norm": 0.22016100585460663, "learning_rate": 2.7316682290854844e-05, "loss": 0.0123, "step": 66740 }, { "grad_norm": 0.1088075339794159, "learning_rate": 2.7301948277829414e-05, "loss": 0.0121, "step": 66750 }, { "grad_norm": 0.16881956160068512, "learning_rate": 2.7287216747028814e-05, "loss": 0.0124, "step": 66760 }, { "grad_norm": 0.16339872777462006, "learning_rate": 2.7272487700064025e-05, "loss": 0.0152, "step": 66770 }, { "grad_norm": 0.14444048702716827, "learning_rate": 2.725776113854583e-05, "loss": 0.0102, "step": 66780 }, { "grad_norm": 0.11682432144880295, "learning_rate": 2.7243037064084664e-05, "loss": 0.012, "step": 66790 }, { "grad_norm": 0.1311737596988678, "learning_rate": 2.7228315478290766e-05, "loss": 0.0119, "step": 66800 }, { "grad_norm": 0.1391073763370514, "learning_rate": 2.721359638277402e-05, "loss": 0.0152, "step": 66810 }, { "grad_norm": 0.1748698353767395, "learning_rate": 2.7198879779144148e-05, "loss": 0.0133, "step": 66820 }, { "grad_norm": 0.1903877854347229, "learning_rate": 2.7184165669010485e-05, "loss": 0.0141, "step": 66830 }, { "grad_norm": 0.23607316613197327, "learning_rate": 2.716945405398218e-05, "loss": 0.0125, "step": 66840 }, { "grad_norm": 0.1837933510541916, "learning_rate": 2.715474493566803e-05, "loss": 0.0134, "step": 66850 }, { "grad_norm": 0.13558565080165863, "learning_rate": 2.7140038315676652e-05, "loss": 0.0119, "step": 66860 }, { "grad_norm": 0.1250043362379074, "learning_rate": 2.7125334195616275e-05, "loss": 0.0155, "step": 66870 }, { "grad_norm": 0.1646333783864975, "learning_rate": 2.7110632577094997e-05, "loss": 0.0144, "step": 66880 }, { "grad_norm": 0.15493081510066986, "learning_rate": 2.709593346172051e-05, "loss": 0.0113, "step": 66890 }, { "grad_norm": 0.12540102005004883, "learning_rate": 2.708123685110032e-05, "loss": 0.0116, "step": 66900 }, { "grad_norm": 0.119297556579113, "learning_rate": 2.7066542746841588e-05, "loss": 0.0128, "step": 66910 }, { "grad_norm": 0.1463104486465454, "learning_rate": 2.7051851150551267e-05, "loss": 0.0115, "step": 66920 }, { "grad_norm": 0.22856272757053375, "learning_rate": 2.7037162063835996e-05, "loss": 0.0153, "step": 66930 }, { "grad_norm": 0.14996345341205597, "learning_rate": 2.702247548830218e-05, "loss": 0.0127, "step": 66940 }, { "grad_norm": 0.24255438148975372, "learning_rate": 2.7007791425555883e-05, "loss": 0.0157, "step": 66950 }, { "grad_norm": 0.22831647098064423, "learning_rate": 2.6993109877202945e-05, "loss": 0.0128, "step": 66960 }, { "grad_norm": 0.14466920495033264, "learning_rate": 2.6978430844848935e-05, "loss": 0.0132, "step": 66970 }, { "grad_norm": 0.14466460049152374, "learning_rate": 2.69637543300991e-05, "loss": 0.0113, "step": 66980 }, { "grad_norm": 0.16278310120105743, "learning_rate": 2.694908033455848e-05, "loss": 0.011, "step": 66990 }, { "grad_norm": 0.14046460390090942, "learning_rate": 2.6934408859831733e-05, "loss": 0.0119, "step": 67000 }, { "grad_norm": 0.2019747644662857, "learning_rate": 2.69197399075234e-05, "loss": 0.0139, "step": 67010 }, { "grad_norm": 0.16885808110237122, "learning_rate": 2.6905073479237584e-05, "loss": 0.0163, "step": 67020 }, { "grad_norm": 0.18990960717201233, "learning_rate": 2.6890409576578235e-05, "loss": 0.0116, "step": 67030 }, { "grad_norm": 0.14275594055652618, "learning_rate": 2.687574820114892e-05, "loss": 0.013, "step": 67040 }, { "grad_norm": 0.14845626056194305, "learning_rate": 2.686108935455305e-05, "loss": 0.0141, "step": 67050 }, { "grad_norm": 0.17257820069789886, "learning_rate": 2.6846433038393603e-05, "loss": 0.0144, "step": 67060 }, { "grad_norm": 0.12550801038742065, "learning_rate": 2.6831779254273482e-05, "loss": 0.0117, "step": 67070 }, { "grad_norm": 0.11940398067235947, "learning_rate": 2.6817128003795122e-05, "loss": 0.0109, "step": 67080 }, { "grad_norm": 0.23457904160022736, "learning_rate": 2.680247928856081e-05, "loss": 0.0142, "step": 67090 }, { "grad_norm": 0.2226737141609192, "learning_rate": 2.6787833110172466e-05, "loss": 0.0139, "step": 67100 }, { "grad_norm": 0.17263582348823547, "learning_rate": 2.6773189470231807e-05, "loss": 0.0138, "step": 67110 }, { "grad_norm": 0.15914921462535858, "learning_rate": 2.675854837034022e-05, "loss": 0.012, "step": 67120 }, { "grad_norm": 0.15061143040657043, "learning_rate": 2.6743909812098833e-05, "loss": 0.0117, "step": 67130 }, { "grad_norm": 0.17152228951454163, "learning_rate": 2.67292737971085e-05, "loss": 0.0147, "step": 67140 }, { "grad_norm": 0.15750007331371307, "learning_rate": 2.671464032696982e-05, "loss": 0.0135, "step": 67150 }, { "grad_norm": 0.12588436901569366, "learning_rate": 2.6700009403283055e-05, "loss": 0.0137, "step": 67160 }, { "grad_norm": 0.2049497812986374, "learning_rate": 2.6685381027648214e-05, "loss": 0.0142, "step": 67170 }, { "grad_norm": 0.195059671998024, "learning_rate": 2.6670755201665086e-05, "loss": 0.0132, "step": 67180 }, { "grad_norm": 0.14218217134475708, "learning_rate": 2.6656131926933066e-05, "loss": 0.0112, "step": 67190 }, { "grad_norm": 0.1617368757724762, "learning_rate": 2.664151120505136e-05, "loss": 0.0104, "step": 67200 }, { "grad_norm": 0.15326586365699768, "learning_rate": 2.6626893037618873e-05, "loss": 0.0122, "step": 67210 }, { "grad_norm": 0.16468587517738342, "learning_rate": 2.6612277426234243e-05, "loss": 0.0172, "step": 67220 }, { "grad_norm": 0.12110356986522675, "learning_rate": 2.659766437249577e-05, "loss": 0.0127, "step": 67230 }, { "grad_norm": 0.1271793097257614, "learning_rate": 2.6583053878001557e-05, "loss": 0.0099, "step": 67240 }, { "grad_norm": 0.11943955719470978, "learning_rate": 2.6568445944349323e-05, "loss": 0.0136, "step": 67250 }, { "grad_norm": 0.16533449292182922, "learning_rate": 2.6553840573136657e-05, "loss": 0.0116, "step": 67260 }, { "grad_norm": 0.14000093936920166, "learning_rate": 2.6539237765960706e-05, "loss": 0.0132, "step": 67270 }, { "grad_norm": 0.13549675047397614, "learning_rate": 2.6524637524418465e-05, "loss": 0.0124, "step": 67280 }, { "grad_norm": 0.16796432435512543, "learning_rate": 2.651003985010655e-05, "loss": 0.0124, "step": 67290 }, { "grad_norm": 0.16601033508777618, "learning_rate": 2.649544474462138e-05, "loss": 0.0133, "step": 67300 }, { "grad_norm": 0.16010858118534088, "learning_rate": 2.6480852209559015e-05, "loss": 0.009, "step": 67310 }, { "grad_norm": 0.16126959025859833, "learning_rate": 2.6466262246515296e-05, "loss": 0.0154, "step": 67320 }, { "grad_norm": 0.13114771246910095, "learning_rate": 2.6451674857085746e-05, "loss": 0.0108, "step": 67330 }, { "grad_norm": 0.1488046795129776, "learning_rate": 2.6437090042865655e-05, "loss": 0.013, "step": 67340 }, { "grad_norm": 0.1497495025396347, "learning_rate": 2.6422507805449947e-05, "loss": 0.0119, "step": 67350 }, { "grad_norm": 0.13174358010292053, "learning_rate": 2.6407928146433357e-05, "loss": 0.0111, "step": 67360 }, { "grad_norm": 0.22875261306762695, "learning_rate": 2.639335106741026e-05, "loss": 0.0159, "step": 67370 }, { "grad_norm": 0.1614895612001419, "learning_rate": 2.6378776569974796e-05, "loss": 0.0117, "step": 67380 }, { "grad_norm": 0.23808714747428894, "learning_rate": 2.6364204655720814e-05, "loss": 0.0133, "step": 67390 }, { "grad_norm": 0.19098754227161407, "learning_rate": 2.6349635326241894e-05, "loss": 0.0107, "step": 67400 }, { "grad_norm": 0.19129765033721924, "learning_rate": 2.633506858313127e-05, "loss": 0.0138, "step": 67410 }, { "grad_norm": 0.13446608185768127, "learning_rate": 2.6320504427981975e-05, "loss": 0.013, "step": 67420 }, { "grad_norm": 0.14399772882461548, "learning_rate": 2.6305942862386734e-05, "loss": 0.0092, "step": 67430 }, { "grad_norm": 0.12352094799280167, "learning_rate": 2.629138388793793e-05, "loss": 0.0133, "step": 67440 }, { "grad_norm": 0.1699160933494568, "learning_rate": 2.6276827506227743e-05, "loss": 0.0126, "step": 67450 }, { "grad_norm": 0.14810428023338318, "learning_rate": 2.6262273718848037e-05, "loss": 0.0132, "step": 67460 }, { "grad_norm": 0.1685429960489273, "learning_rate": 2.6247722527390407e-05, "loss": 0.012, "step": 67470 }, { "grad_norm": 0.1323234587907791, "learning_rate": 2.6233173933446104e-05, "loss": 0.0119, "step": 67480 }, { "grad_norm": 0.12208054959774017, "learning_rate": 2.6218627938606188e-05, "loss": 0.0097, "step": 67490 }, { "grad_norm": 0.11831275373697281, "learning_rate": 2.620408454446135e-05, "loss": 0.0126, "step": 67500 }, { "grad_norm": 0.13515207171440125, "learning_rate": 2.6189543752602046e-05, "loss": 0.0144, "step": 67510 }, { "grad_norm": 0.20584121346473694, "learning_rate": 2.6175005564618444e-05, "loss": 0.0121, "step": 67520 }, { "grad_norm": 0.1490681916475296, "learning_rate": 2.6160469982100428e-05, "loss": 0.0135, "step": 67530 }, { "grad_norm": 0.1694234013557434, "learning_rate": 2.6145937006637555e-05, "loss": 0.0141, "step": 67540 }, { "grad_norm": 0.18228374421596527, "learning_rate": 2.6131406639819166e-05, "loss": 0.0144, "step": 67550 }, { "grad_norm": 0.10464922338724136, "learning_rate": 2.611687888323424e-05, "loss": 0.0137, "step": 67560 }, { "grad_norm": 0.13260909914970398, "learning_rate": 2.610235373847153e-05, "loss": 0.0118, "step": 67570 }, { "grad_norm": 0.12235908210277557, "learning_rate": 2.6087831207119494e-05, "loss": 0.0109, "step": 67580 }, { "grad_norm": 0.18913912773132324, "learning_rate": 2.60733112907663e-05, "loss": 0.0157, "step": 67590 }, { "grad_norm": 0.13449479639530182, "learning_rate": 2.6058793990999786e-05, "loss": 0.0104, "step": 67600 }, { "grad_norm": 0.18827751278877258, "learning_rate": 2.6044279309407593e-05, "loss": 0.0138, "step": 67610 }, { "grad_norm": 0.15938472747802734, "learning_rate": 2.6029767247576976e-05, "loss": 0.0118, "step": 67620 }, { "grad_norm": 0.1646559238433838, "learning_rate": 2.601525780709497e-05, "loss": 0.0108, "step": 67630 }, { "grad_norm": 0.15239614248275757, "learning_rate": 2.6000750989548318e-05, "loss": 0.0106, "step": 67640 }, { "grad_norm": 0.15247322618961334, "learning_rate": 2.5986246796523462e-05, "loss": 0.013, "step": 67650 }, { "grad_norm": 0.14828963577747345, "learning_rate": 2.5971745229606546e-05, "loss": 0.0153, "step": 67660 }, { "grad_norm": 0.132241353392601, "learning_rate": 2.5957246290383442e-05, "loss": 0.0128, "step": 67670 }, { "grad_norm": 0.14358685910701752, "learning_rate": 2.594274998043975e-05, "loss": 0.0111, "step": 67680 }, { "grad_norm": 0.1359923630952835, "learning_rate": 2.5928256301360733e-05, "loss": 0.0109, "step": 67690 }, { "grad_norm": 0.19937334954738617, "learning_rate": 2.591376525473142e-05, "loss": 0.0141, "step": 67700 }, { "grad_norm": 0.1391943097114563, "learning_rate": 2.5899276842136523e-05, "loss": 0.0104, "step": 67710 }, { "grad_norm": 0.12934686243534088, "learning_rate": 2.5884791065160495e-05, "loss": 0.011, "step": 67720 }, { "grad_norm": 0.13531798124313354, "learning_rate": 2.5870307925387436e-05, "loss": 0.0154, "step": 67730 }, { "grad_norm": 0.16115306317806244, "learning_rate": 2.585582742440125e-05, "loss": 0.0119, "step": 67740 }, { "grad_norm": 0.1572342962026596, "learning_rate": 2.5841349563785455e-05, "loss": 0.016, "step": 67750 }, { "grad_norm": 0.17916280031204224, "learning_rate": 2.5826874345123355e-05, "loss": 0.0129, "step": 67760 }, { "grad_norm": 0.18199893832206726, "learning_rate": 2.581240176999792e-05, "loss": 0.0142, "step": 67770 }, { "grad_norm": 0.16551955044269562, "learning_rate": 2.579793183999189e-05, "loss": 0.0136, "step": 67780 }, { "grad_norm": 0.15979646146297455, "learning_rate": 2.5783464556687627e-05, "loss": 0.0149, "step": 67790 }, { "grad_norm": 0.16340689361095428, "learning_rate": 2.576899992166729e-05, "loss": 0.0159, "step": 67800 }, { "grad_norm": 0.15623117983341217, "learning_rate": 2.5754537936512668e-05, "loss": 0.0107, "step": 67810 }, { "grad_norm": 0.17798787355422974, "learning_rate": 2.574007860280535e-05, "loss": 0.0101, "step": 67820 }, { "grad_norm": 0.3117430806159973, "learning_rate": 2.5725621922126518e-05, "loss": 0.014, "step": 67830 }, { "grad_norm": 0.14011409878730774, "learning_rate": 2.571116789605722e-05, "loss": 0.0128, "step": 67840 }, { "grad_norm": 0.17939312756061554, "learning_rate": 2.5696716526178054e-05, "loss": 0.0182, "step": 67850 }, { "grad_norm": 0.15568310022354126, "learning_rate": 2.568226781406946e-05, "loss": 0.0131, "step": 67860 }, { "grad_norm": 0.2272948920726776, "learning_rate": 2.5667821761311474e-05, "loss": 0.0124, "step": 67870 }, { "grad_norm": 0.11541013419628143, "learning_rate": 2.5653378369483917e-05, "loss": 0.0105, "step": 67880 }, { "grad_norm": 0.21151767671108246, "learning_rate": 2.5638937640166292e-05, "loss": 0.0107, "step": 67890 }, { "grad_norm": 0.14657379686832428, "learning_rate": 2.5624499574937842e-05, "loss": 0.011, "step": 67900 }, { "grad_norm": 0.20026373863220215, "learning_rate": 2.5610064175377456e-05, "loss": 0.0162, "step": 67910 }, { "grad_norm": 0.1368846297264099, "learning_rate": 2.5595631443063782e-05, "loss": 0.0126, "step": 67920 }, { "grad_norm": 0.19707374274730682, "learning_rate": 2.5581201379575177e-05, "loss": 0.0127, "step": 67930 }, { "grad_norm": 0.20164324343204498, "learning_rate": 2.5566773986489655e-05, "loss": 0.0124, "step": 67940 }, { "grad_norm": 0.17917566001415253, "learning_rate": 2.5552349265385e-05, "loss": 0.013, "step": 67950 }, { "grad_norm": 0.11833939701318741, "learning_rate": 2.5537927217838675e-05, "loss": 0.0137, "step": 67960 }, { "grad_norm": 0.18331670761108398, "learning_rate": 2.5523507845427864e-05, "loss": 0.0135, "step": 67970 }, { "grad_norm": 0.1622968465089798, "learning_rate": 2.5509091149729413e-05, "loss": 0.0119, "step": 67980 }, { "grad_norm": 0.12353993952274323, "learning_rate": 2.5494677132319962e-05, "loss": 0.0128, "step": 67990 }, { "grad_norm": 0.1874922662973404, "learning_rate": 2.548026579477575e-05, "loss": 0.015, "step": 68000 }, { "grad_norm": 0.1412018984556198, "learning_rate": 2.5465857138672834e-05, "loss": 0.0098, "step": 68010 }, { "grad_norm": 0.15673203766345978, "learning_rate": 2.5451451165586848e-05, "loss": 0.0143, "step": 68020 }, { "grad_norm": 0.15575499832630157, "learning_rate": 2.543704787709329e-05, "loss": 0.0121, "step": 68030 }, { "grad_norm": 0.13913553953170776, "learning_rate": 2.542264727476723e-05, "loss": 0.0137, "step": 68040 }, { "grad_norm": 0.14800086617469788, "learning_rate": 2.5408249360183535e-05, "loss": 0.0124, "step": 68050 }, { "grad_norm": 0.1510116457939148, "learning_rate": 2.5393854134916694e-05, "loss": 0.0173, "step": 68060 }, { "grad_norm": 0.1402467042207718, "learning_rate": 2.5379461600540993e-05, "loss": 0.0133, "step": 68070 }, { "grad_norm": 0.16185477375984192, "learning_rate": 2.5365071758630317e-05, "loss": 0.012, "step": 68080 }, { "grad_norm": 0.16161523759365082, "learning_rate": 2.535068461075839e-05, "loss": 0.0133, "step": 68090 }, { "grad_norm": 0.14309987425804138, "learning_rate": 2.5336300158498516e-05, "loss": 0.0107, "step": 68100 }, { "grad_norm": 0.12776075303554535, "learning_rate": 2.5321918403423793e-05, "loss": 0.0148, "step": 68110 }, { "grad_norm": 0.1979123204946518, "learning_rate": 2.5307539347106957e-05, "loss": 0.0124, "step": 68120 }, { "grad_norm": 0.1488964706659317, "learning_rate": 2.529316299112049e-05, "loss": 0.0116, "step": 68130 }, { "grad_norm": 0.16545969247817993, "learning_rate": 2.5278789337036567e-05, "loss": 0.0142, "step": 68140 }, { "grad_norm": 0.18609434366226196, "learning_rate": 2.52644183864271e-05, "loss": 0.0123, "step": 68150 }, { "grad_norm": 0.18399985134601593, "learning_rate": 2.525005014086363e-05, "loss": 0.0136, "step": 68160 }, { "grad_norm": 0.18318922817707062, "learning_rate": 2.5235684601917465e-05, "loss": 0.013, "step": 68170 }, { "grad_norm": 0.1532159298658371, "learning_rate": 2.5221321771159618e-05, "loss": 0.0133, "step": 68180 }, { "grad_norm": 0.16539278626441956, "learning_rate": 2.5206961650160743e-05, "loss": 0.0102, "step": 68190 }, { "grad_norm": 0.16205154359340668, "learning_rate": 2.519260424049129e-05, "loss": 0.0114, "step": 68200 }, { "grad_norm": 0.12893421947956085, "learning_rate": 2.5178249543721304e-05, "loss": 0.0103, "step": 68210 }, { "grad_norm": 0.14732502400875092, "learning_rate": 2.5163897561420668e-05, "loss": 0.0155, "step": 68220 }, { "grad_norm": 0.13201293349266052, "learning_rate": 2.514954829515883e-05, "loss": 0.0139, "step": 68230 }, { "grad_norm": 0.1311754584312439, "learning_rate": 2.513520174650505e-05, "loss": 0.0135, "step": 68240 }, { "grad_norm": 0.12743371725082397, "learning_rate": 2.5120857917028207e-05, "loss": 0.0156, "step": 68250 }, { "grad_norm": 0.1506301462650299, "learning_rate": 2.5106516808296953e-05, "loss": 0.0122, "step": 68260 }, { "grad_norm": 0.16955699026584625, "learning_rate": 2.5092178421879552e-05, "loss": 0.0131, "step": 68270 }, { "grad_norm": 0.16422756016254425, "learning_rate": 2.5077842759344116e-05, "loss": 0.0114, "step": 68280 }, { "grad_norm": 0.16096150875091553, "learning_rate": 2.5063509822258304e-05, "loss": 0.0153, "step": 68290 }, { "grad_norm": 0.1258980929851532, "learning_rate": 2.5049179612189576e-05, "loss": 0.0133, "step": 68300 }, { "grad_norm": 0.18713033199310303, "learning_rate": 2.5034852130705043e-05, "loss": 0.0111, "step": 68310 }, { "grad_norm": 0.24540255963802338, "learning_rate": 2.5020527379371572e-05, "loss": 0.0108, "step": 68320 }, { "grad_norm": 0.14095796644687653, "learning_rate": 2.500620535975564e-05, "loss": 0.0132, "step": 68330 }, { "grad_norm": 0.155714750289917, "learning_rate": 2.4991886073423515e-05, "loss": 0.0123, "step": 68340 }, { "grad_norm": 0.13846667110919952, "learning_rate": 2.4977569521941134e-05, "loss": 0.0123, "step": 68350 }, { "grad_norm": 0.1393996626138687, "learning_rate": 2.4963255706874143e-05, "loss": 0.013, "step": 68360 }, { "grad_norm": 0.17686663568019867, "learning_rate": 2.4948944629787853e-05, "loss": 0.013, "step": 68370 }, { "grad_norm": 0.1367473304271698, "learning_rate": 2.493463629224731e-05, "loss": 0.0124, "step": 68380 }, { "grad_norm": 0.1596098095178604, "learning_rate": 2.492033069581728e-05, "loss": 0.0116, "step": 68390 }, { "grad_norm": 0.2086050659418106, "learning_rate": 2.4906027842062137e-05, "loss": 0.0138, "step": 68400 }, { "grad_norm": 0.13141077756881714, "learning_rate": 2.48917277325461e-05, "loss": 0.0117, "step": 68410 }, { "grad_norm": 0.14814065396785736, "learning_rate": 2.487743036883295e-05, "loss": 0.0104, "step": 68420 }, { "grad_norm": 0.16881169378757477, "learning_rate": 2.486313575248626e-05, "loss": 0.0124, "step": 68430 }, { "grad_norm": 0.17164292931556702, "learning_rate": 2.4848843885069235e-05, "loss": 0.0133, "step": 68440 }, { "grad_norm": 0.1520436704158783, "learning_rate": 2.483455476814484e-05, "loss": 0.0111, "step": 68450 }, { "grad_norm": 0.1250312477350235, "learning_rate": 2.4820268403275672e-05, "loss": 0.0124, "step": 68460 }, { "grad_norm": 0.1590920090675354, "learning_rate": 2.480598479202413e-05, "loss": 0.0125, "step": 68470 }, { "grad_norm": 0.2201087772846222, "learning_rate": 2.4791703935952193e-05, "loss": 0.0143, "step": 68480 }, { "grad_norm": 0.1372992992401123, "learning_rate": 2.477742583662163e-05, "loss": 0.0123, "step": 68490 }, { "grad_norm": 0.1669394075870514, "learning_rate": 2.476315049559384e-05, "loss": 0.0127, "step": 68500 }, { "grad_norm": 0.14866647124290466, "learning_rate": 2.4748877914429987e-05, "loss": 0.0146, "step": 68510 }, { "grad_norm": 0.15210506319999695, "learning_rate": 2.4734608094690857e-05, "loss": 0.015, "step": 68520 }, { "grad_norm": 0.12922929227352142, "learning_rate": 2.4720341037937006e-05, "loss": 0.0129, "step": 68530 }, { "grad_norm": 0.16110318899154663, "learning_rate": 2.4706076745728645e-05, "loss": 0.013, "step": 68540 }, { "grad_norm": 0.18159076571464539, "learning_rate": 2.469181521962573e-05, "loss": 0.0129, "step": 68550 }, { "grad_norm": 0.1530631184577942, "learning_rate": 2.467755646118783e-05, "loss": 0.0176, "step": 68560 }, { "grad_norm": 0.23149336874485016, "learning_rate": 2.46633004719743e-05, "loss": 0.0109, "step": 68570 }, { "grad_norm": 0.42966797947883606, "learning_rate": 2.4649047253544123e-05, "loss": 0.0127, "step": 68580 }, { "grad_norm": 0.1663168966770172, "learning_rate": 2.4634796807456016e-05, "loss": 0.0101, "step": 68590 }, { "grad_norm": 0.16598881781101227, "learning_rate": 2.46205491352684e-05, "loss": 0.0125, "step": 68600 }, { "grad_norm": 0.1597094088792801, "learning_rate": 2.4606304238539375e-05, "loss": 0.0102, "step": 68610 }, { "grad_norm": 0.21712438762187958, "learning_rate": 2.4592062118826753e-05, "loss": 0.0124, "step": 68620 }, { "grad_norm": 0.15087108314037323, "learning_rate": 2.4577822777688004e-05, "loss": 0.0137, "step": 68630 }, { "grad_norm": 0.2101091593503952, "learning_rate": 2.4563586216680347e-05, "loss": 0.0133, "step": 68640 }, { "grad_norm": 0.1768660843372345, "learning_rate": 2.4549352437360622e-05, "loss": 0.0131, "step": 68650 }, { "grad_norm": 0.12429293245077133, "learning_rate": 2.4535121441285493e-05, "loss": 0.0115, "step": 68660 }, { "grad_norm": 0.14602577686309814, "learning_rate": 2.4520893230011172e-05, "loss": 0.011, "step": 68670 }, { "grad_norm": 0.11160058528184891, "learning_rate": 2.450666780509368e-05, "loss": 0.0116, "step": 68680 }, { "grad_norm": 0.12303557991981506, "learning_rate": 2.4492445168088656e-05, "loss": 0.0114, "step": 68690 }, { "grad_norm": 0.1276429146528244, "learning_rate": 2.447822532055149e-05, "loss": 0.0119, "step": 68700 }, { "grad_norm": 0.17536252737045288, "learning_rate": 2.4464008264037212e-05, "loss": 0.0111, "step": 68710 }, { "grad_norm": 0.16713422536849976, "learning_rate": 2.4449794000100605e-05, "loss": 0.0134, "step": 68720 }, { "grad_norm": 0.14340239763259888, "learning_rate": 2.4435582530296108e-05, "loss": 0.0117, "step": 68730 }, { "grad_norm": 0.13974082469940186, "learning_rate": 2.4421373856177887e-05, "loss": 0.0119, "step": 68740 }, { "grad_norm": 0.18992303311824799, "learning_rate": 2.440716797929975e-05, "loss": 0.0119, "step": 68750 }, { "grad_norm": 0.2017170488834381, "learning_rate": 2.439296490121526e-05, "loss": 0.0145, "step": 68760 }, { "grad_norm": 0.1830669790506363, "learning_rate": 2.4378764623477608e-05, "loss": 0.0179, "step": 68770 }, { "grad_norm": 0.19616860151290894, "learning_rate": 2.436456714763974e-05, "loss": 0.0143, "step": 68780 }, { "grad_norm": 0.11422611027956009, "learning_rate": 2.4350372475254262e-05, "loss": 0.0109, "step": 68790 }, { "grad_norm": 0.21891014277935028, "learning_rate": 2.433618060787351e-05, "loss": 0.0125, "step": 68800 }, { "grad_norm": 0.1465008705854416, "learning_rate": 2.432199154704944e-05, "loss": 0.0153, "step": 68810 }, { "grad_norm": 0.21948568522930145, "learning_rate": 2.430780529433377e-05, "loss": 0.0107, "step": 68820 }, { "grad_norm": 0.12722185254096985, "learning_rate": 2.42936218512779e-05, "loss": 0.0117, "step": 68830 }, { "grad_norm": 0.14569945633411407, "learning_rate": 2.427944121943288e-05, "loss": 0.011, "step": 68840 }, { "grad_norm": 0.12630783021450043, "learning_rate": 2.4265263400349497e-05, "loss": 0.0114, "step": 68850 }, { "grad_norm": 0.14863111078739166, "learning_rate": 2.4251088395578213e-05, "loss": 0.0105, "step": 68860 }, { "grad_norm": 0.12375764548778534, "learning_rate": 2.4236916206669214e-05, "loss": 0.011, "step": 68870 }, { "grad_norm": 0.1316535323858261, "learning_rate": 2.4222746835172304e-05, "loss": 0.0104, "step": 68880 }, { "grad_norm": 0.13746118545532227, "learning_rate": 2.4208580282637065e-05, "loss": 0.0142, "step": 68890 }, { "grad_norm": 0.20847781002521515, "learning_rate": 2.4194416550612687e-05, "loss": 0.0136, "step": 68900 }, { "grad_norm": 0.13194674253463745, "learning_rate": 2.418025564064812e-05, "loss": 0.0122, "step": 68910 }, { "grad_norm": 0.17954912781715393, "learning_rate": 2.4166097554291978e-05, "loss": 0.0118, "step": 68920 }, { "grad_norm": 0.17375236749649048, "learning_rate": 2.4151942293092583e-05, "loss": 0.0112, "step": 68930 }, { "grad_norm": 0.10724306106567383, "learning_rate": 2.41377898585979e-05, "loss": 0.0109, "step": 68940 }, { "grad_norm": 0.12841695547103882, "learning_rate": 2.4123640252355652e-05, "loss": 0.0115, "step": 68950 }, { "grad_norm": 0.1509581357240677, "learning_rate": 2.410949347591318e-05, "loss": 0.0116, "step": 68960 }, { "grad_norm": 0.20461367070674896, "learning_rate": 2.4095349530817578e-05, "loss": 0.0117, "step": 68970 }, { "grad_norm": 0.1707049310207367, "learning_rate": 2.4081208418615603e-05, "loss": 0.012, "step": 68980 }, { "grad_norm": 0.15527872741222382, "learning_rate": 2.4067070140853725e-05, "loss": 0.0142, "step": 68990 }, { "grad_norm": 0.18059037625789642, "learning_rate": 2.405293469907805e-05, "loss": 0.0126, "step": 69000 }, { "grad_norm": 0.1101338341832161, "learning_rate": 2.4038802094834446e-05, "loss": 0.012, "step": 69010 }, { "grad_norm": 0.16259564459323883, "learning_rate": 2.4024672329668395e-05, "loss": 0.0137, "step": 69020 }, { "grad_norm": 0.16177384555339813, "learning_rate": 2.4010545405125125e-05, "loss": 0.014, "step": 69030 }, { "grad_norm": 0.15067189931869507, "learning_rate": 2.3996421322749528e-05, "loss": 0.0123, "step": 69040 }, { "grad_norm": 0.24252113699913025, "learning_rate": 2.3982300084086224e-05, "loss": 0.0131, "step": 69050 }, { "grad_norm": 0.09404825419187546, "learning_rate": 2.396818169067944e-05, "loss": 0.0111, "step": 69060 }, { "grad_norm": 0.1498643159866333, "learning_rate": 2.3954066144073178e-05, "loss": 0.0108, "step": 69070 }, { "grad_norm": 0.11657769978046417, "learning_rate": 2.39399534458111e-05, "loss": 0.011, "step": 69080 }, { "grad_norm": 0.1713944375514984, "learning_rate": 2.392584359743651e-05, "loss": 0.0124, "step": 69090 }, { "grad_norm": 0.1185387447476387, "learning_rate": 2.3911736600492463e-05, "loss": 0.0093, "step": 69100 }, { "grad_norm": 0.1473686248064041, "learning_rate": 2.3897632456521672e-05, "loss": 0.0108, "step": 69110 }, { "grad_norm": 0.24934746325016022, "learning_rate": 2.388353116706658e-05, "loss": 0.0141, "step": 69120 }, { "grad_norm": 0.14388369023799896, "learning_rate": 2.3869432733669232e-05, "loss": 0.013, "step": 69130 }, { "grad_norm": 0.14720484614372253, "learning_rate": 2.385533715787145e-05, "loss": 0.0114, "step": 69140 }, { "grad_norm": 0.151225283741951, "learning_rate": 2.3841244441214673e-05, "loss": 0.0099, "step": 69150 }, { "grad_norm": 0.14811351895332336, "learning_rate": 2.3827154585240073e-05, "loss": 0.0134, "step": 69160 }, { "grad_norm": 0.13158999383449554, "learning_rate": 2.3813067591488498e-05, "loss": 0.0101, "step": 69170 }, { "grad_norm": 0.20419186353683472, "learning_rate": 2.3798983461500494e-05, "loss": 0.0123, "step": 69180 }, { "grad_norm": 0.09378557652235031, "learning_rate": 2.3784902196816256e-05, "loss": 0.0102, "step": 69190 }, { "grad_norm": 0.15659114718437195, "learning_rate": 2.3770823798975716e-05, "loss": 0.0114, "step": 69200 }, { "grad_norm": 0.1231360137462616, "learning_rate": 2.3756748269518424e-05, "loss": 0.012, "step": 69210 }, { "grad_norm": 0.09795364737510681, "learning_rate": 2.374267560998371e-05, "loss": 0.0089, "step": 69220 }, { "grad_norm": 0.17681793868541718, "learning_rate": 2.3728605821910477e-05, "loss": 0.0126, "step": 69230 }, { "grad_norm": 0.18298712372779846, "learning_rate": 2.3714538906837453e-05, "loss": 0.0119, "step": 69240 }, { "grad_norm": 0.12138663232326508, "learning_rate": 2.3700474866302913e-05, "loss": 0.0137, "step": 69250 }, { "grad_norm": 0.21885037422180176, "learning_rate": 2.3686413701844917e-05, "loss": 0.0124, "step": 69260 }, { "grad_norm": 0.19877155125141144, "learning_rate": 2.367235541500114e-05, "loss": 0.0143, "step": 69270 }, { "grad_norm": 0.21688997745513916, "learning_rate": 2.3658300007308993e-05, "loss": 0.015, "step": 69280 }, { "grad_norm": 0.15618035197257996, "learning_rate": 2.364424748030555e-05, "loss": 0.0153, "step": 69290 }, { "grad_norm": 0.15475185215473175, "learning_rate": 2.3630197835527595e-05, "loss": 0.0142, "step": 69300 }, { "grad_norm": 0.1384025514125824, "learning_rate": 2.3616151074511546e-05, "loss": 0.0101, "step": 69310 }, { "grad_norm": 0.12560155987739563, "learning_rate": 2.3602107198793545e-05, "loss": 0.0125, "step": 69320 }, { "grad_norm": 0.1803291290998459, "learning_rate": 2.3588066209909427e-05, "loss": 0.0104, "step": 69330 }, { "grad_norm": 0.276348739862442, "learning_rate": 2.3574028109394662e-05, "loss": 0.0102, "step": 69340 }, { "grad_norm": 0.13841289281845093, "learning_rate": 2.3559992898784454e-05, "loss": 0.0111, "step": 69350 }, { "grad_norm": 0.13274942338466644, "learning_rate": 2.354596057961367e-05, "loss": 0.0138, "step": 69360 }, { "grad_norm": 0.14430350065231323, "learning_rate": 2.353193115341688e-05, "loss": 0.0114, "step": 69370 }, { "grad_norm": 0.14641983807086945, "learning_rate": 2.3517904621728294e-05, "loss": 0.0106, "step": 69380 }, { "grad_norm": 0.1498260200023651, "learning_rate": 2.3503880986081855e-05, "loss": 0.0106, "step": 69390 }, { "grad_norm": 0.09959296137094498, "learning_rate": 2.3489860248011136e-05, "loss": 0.0133, "step": 69400 }, { "grad_norm": 0.20615379512310028, "learning_rate": 2.3475842409049464e-05, "loss": 0.0111, "step": 69410 }, { "grad_norm": 0.1552366316318512, "learning_rate": 2.346182747072975e-05, "loss": 0.0126, "step": 69420 }, { "grad_norm": 0.1235499307513237, "learning_rate": 2.3447815434584718e-05, "loss": 0.0123, "step": 69430 }, { "grad_norm": 0.1963042914867401, "learning_rate": 2.3433806302146656e-05, "loss": 0.0115, "step": 69440 }, { "grad_norm": 0.14545612037181854, "learning_rate": 2.3419800074947612e-05, "loss": 0.0111, "step": 69450 }, { "grad_norm": 0.16375422477722168, "learning_rate": 2.3405796754519248e-05, "loss": 0.0118, "step": 69460 }, { "grad_norm": 0.11702565848827362, "learning_rate": 2.3391796342392992e-05, "loss": 0.0109, "step": 69470 }, { "grad_norm": 0.1205805167555809, "learning_rate": 2.337779884009984e-05, "loss": 0.013, "step": 69480 }, { "grad_norm": 0.15293338894844055, "learning_rate": 2.3363804249170624e-05, "loss": 0.0135, "step": 69490 }, { "grad_norm": 0.16535457968711853, "learning_rate": 2.3349812571135703e-05, "loss": 0.0112, "step": 69500 }, { "grad_norm": 0.12671911716461182, "learning_rate": 2.333582380752523e-05, "loss": 0.0111, "step": 69510 }, { "grad_norm": 0.12694285809993744, "learning_rate": 2.3321837959868963e-05, "loss": 0.0114, "step": 69520 }, { "grad_norm": 0.12304341793060303, "learning_rate": 2.3307855029696383e-05, "loss": 0.0104, "step": 69530 }, { "grad_norm": 0.186976820230484, "learning_rate": 2.3293875018536648e-05, "loss": 0.0126, "step": 69540 }, { "grad_norm": 0.11133310198783875, "learning_rate": 2.327989792791861e-05, "loss": 0.0104, "step": 69550 }, { "grad_norm": 0.14701537787914276, "learning_rate": 2.326592375937074e-05, "loss": 0.0116, "step": 69560 }, { "grad_norm": 0.20843935012817383, "learning_rate": 2.3251952514421248e-05, "loss": 0.0124, "step": 69570 }, { "grad_norm": 0.18736693263053894, "learning_rate": 2.323798419459804e-05, "loss": 0.0111, "step": 69580 }, { "grad_norm": 0.14260204136371613, "learning_rate": 2.3224018801428625e-05, "loss": 0.0096, "step": 69590 }, { "grad_norm": 0.1726929098367691, "learning_rate": 2.3210056336440273e-05, "loss": 0.011, "step": 69600 }, { "grad_norm": 0.16629907488822937, "learning_rate": 2.3196096801159844e-05, "loss": 0.0144, "step": 69610 }, { "grad_norm": 0.15350379049777985, "learning_rate": 2.3182140197114012e-05, "loss": 0.0124, "step": 69620 }, { "grad_norm": 0.1676667034626007, "learning_rate": 2.3168186525828985e-05, "loss": 0.0114, "step": 69630 }, { "grad_norm": 0.19981445372104645, "learning_rate": 2.3154235788830757e-05, "loss": 0.0116, "step": 69640 }, { "grad_norm": 0.128595232963562, "learning_rate": 2.3140287987644927e-05, "loss": 0.015, "step": 69650 }, { "grad_norm": 0.2413952648639679, "learning_rate": 2.3126343123796833e-05, "loss": 0.0138, "step": 69660 }, { "grad_norm": 0.20371310412883759, "learning_rate": 2.3112401198811418e-05, "loss": 0.0114, "step": 69670 }, { "grad_norm": 0.16547244787216187, "learning_rate": 2.3098462214213418e-05, "loss": 0.0111, "step": 69680 }, { "grad_norm": 0.1582074761390686, "learning_rate": 2.308452617152712e-05, "loss": 0.0112, "step": 69690 }, { "grad_norm": 0.14267699420452118, "learning_rate": 2.3070593072276603e-05, "loss": 0.0104, "step": 69700 }, { "grad_norm": 0.1417090743780136, "learning_rate": 2.3056662917985518e-05, "loss": 0.0125, "step": 69710 }, { "grad_norm": 0.133194699883461, "learning_rate": 2.3042735710177283e-05, "loss": 0.0112, "step": 69720 }, { "grad_norm": 0.16961053013801575, "learning_rate": 2.302881145037492e-05, "loss": 0.0146, "step": 69730 }, { "grad_norm": 0.1722397357225418, "learning_rate": 2.3014890140101192e-05, "loss": 0.0101, "step": 69740 }, { "grad_norm": 0.1462298482656479, "learning_rate": 2.30009717808785e-05, "loss": 0.0118, "step": 69750 }, { "grad_norm": 0.22440454363822937, "learning_rate": 2.2987056374228967e-05, "loss": 0.01, "step": 69760 }, { "grad_norm": 0.18460892140865326, "learning_rate": 2.297314392167431e-05, "loss": 0.0128, "step": 69770 }, { "grad_norm": 0.1885300874710083, "learning_rate": 2.295923442473601e-05, "loss": 0.0108, "step": 69780 }, { "grad_norm": 0.14625877141952515, "learning_rate": 2.294532788493519e-05, "loss": 0.0113, "step": 69790 }, { "grad_norm": 0.15226469933986664, "learning_rate": 2.293142430379262e-05, "loss": 0.0122, "step": 69800 }, { "grad_norm": 0.11870567500591278, "learning_rate": 2.291752368282879e-05, "loss": 0.011, "step": 69810 }, { "grad_norm": 0.17597796022891998, "learning_rate": 2.290362602356385e-05, "loss": 0.0123, "step": 69820 }, { "grad_norm": 0.21035221219062805, "learning_rate": 2.288973132751765e-05, "loss": 0.0114, "step": 69830 }, { "grad_norm": 0.23409149050712585, "learning_rate": 2.287583959620965e-05, "loss": 0.0151, "step": 69840 }, { "grad_norm": 0.1606895476579666, "learning_rate": 2.2861950831159074e-05, "loss": 0.0119, "step": 69850 }, { "grad_norm": 0.13347825407981873, "learning_rate": 2.284806503388472e-05, "loss": 0.0103, "step": 69860 }, { "grad_norm": 0.12497754395008087, "learning_rate": 2.283418220590518e-05, "loss": 0.0121, "step": 69870 }, { "grad_norm": 0.15587469935417175, "learning_rate": 2.282030234873862e-05, "loss": 0.011, "step": 69880 }, { "grad_norm": 0.13582400977611542, "learning_rate": 2.2806425463902948e-05, "loss": 0.0119, "step": 69890 }, { "grad_norm": 0.1974179446697235, "learning_rate": 2.279255155291568e-05, "loss": 0.0122, "step": 69900 }, { "grad_norm": 0.15215447545051575, "learning_rate": 2.277868061729409e-05, "loss": 0.0098, "step": 69910 }, { "grad_norm": 0.10559707134962082, "learning_rate": 2.276481265855504e-05, "loss": 0.013, "step": 69920 }, { "grad_norm": 0.1445806473493576, "learning_rate": 2.2750947678215128e-05, "loss": 0.0113, "step": 69930 }, { "grad_norm": 0.11407401412725449, "learning_rate": 2.2737085677790616e-05, "loss": 0.0112, "step": 69940 }, { "grad_norm": 0.1423463374376297, "learning_rate": 2.2723226658797443e-05, "loss": 0.0136, "step": 69950 }, { "grad_norm": 0.19734704494476318, "learning_rate": 2.270937062275117e-05, "loss": 0.0136, "step": 69960 }, { "grad_norm": 0.1382170021533966, "learning_rate": 2.2695517571167113e-05, "loss": 0.0125, "step": 69970 }, { "grad_norm": 0.11951079219579697, "learning_rate": 2.268166750556019e-05, "loss": 0.0114, "step": 69980 }, { "grad_norm": 0.1283772736787796, "learning_rate": 2.266782042744503e-05, "loss": 0.0117, "step": 69990 }, { "grad_norm": 0.17268915474414825, "learning_rate": 2.2653976338335935e-05, "loss": 0.0109, "step": 70000 }, { "grad_norm": 0.166243314743042, "learning_rate": 2.2640135239746895e-05, "loss": 0.0163, "step": 70010 }, { "grad_norm": 0.19029445946216583, "learning_rate": 2.2626297133191515e-05, "loss": 0.0159, "step": 70020 }, { "grad_norm": 0.15728861093521118, "learning_rate": 2.261246202018312e-05, "loss": 0.0114, "step": 70030 }, { "grad_norm": 0.1866624653339386, "learning_rate": 2.2598629902234718e-05, "loss": 0.0122, "step": 70040 }, { "grad_norm": 0.12226861715316772, "learning_rate": 2.258480078085894e-05, "loss": 0.0135, "step": 70050 }, { "grad_norm": 0.15481685101985931, "learning_rate": 2.2570974657568126e-05, "loss": 0.0122, "step": 70060 }, { "grad_norm": 0.10905598849058151, "learning_rate": 2.2557151533874286e-05, "loss": 0.0117, "step": 70070 }, { "grad_norm": 0.14179585874080658, "learning_rate": 2.2543331411289116e-05, "loss": 0.0118, "step": 70080 }, { "grad_norm": 0.23855765163898468, "learning_rate": 2.252951429132392e-05, "loss": 0.0111, "step": 70090 }, { "grad_norm": 0.14873816072940826, "learning_rate": 2.2515700175489762e-05, "loss": 0.0115, "step": 70100 }, { "grad_norm": 0.12431712448596954, "learning_rate": 2.2501889065297288e-05, "loss": 0.011, "step": 70110 }, { "grad_norm": 0.18116527795791626, "learning_rate": 2.2488080962256887e-05, "loss": 0.0094, "step": 70120 }, { "grad_norm": 0.1770397573709488, "learning_rate": 2.2474275867878585e-05, "loss": 0.0129, "step": 70130 }, { "grad_norm": 0.14542171359062195, "learning_rate": 2.246047378367211e-05, "loss": 0.0119, "step": 70140 }, { "grad_norm": 0.16505776345729828, "learning_rate": 2.2446674711146798e-05, "loss": 0.0124, "step": 70150 }, { "grad_norm": 0.16194726526737213, "learning_rate": 2.2432878651811734e-05, "loss": 0.0146, "step": 70160 }, { "grad_norm": 0.15278057754039764, "learning_rate": 2.2419085607175594e-05, "loss": 0.0129, "step": 70170 }, { "grad_norm": 0.1695077270269394, "learning_rate": 2.240529557874678e-05, "loss": 0.0112, "step": 70180 }, { "grad_norm": 0.17874617874622345, "learning_rate": 2.2391508568033358e-05, "loss": 0.0116, "step": 70190 }, { "grad_norm": 0.1733674257993698, "learning_rate": 2.237772457654307e-05, "loss": 0.013, "step": 70200 }, { "grad_norm": 0.20595717430114746, "learning_rate": 2.236394360578327e-05, "loss": 0.0115, "step": 70210 }, { "grad_norm": 0.16271421313285828, "learning_rate": 2.2350165657261073e-05, "loss": 0.0171, "step": 70220 }, { "grad_norm": 0.15304872393608093, "learning_rate": 2.2336390732483164e-05, "loss": 0.0119, "step": 70230 }, { "grad_norm": 0.17501229047775269, "learning_rate": 2.2322618832955976e-05, "loss": 0.0117, "step": 70240 }, { "grad_norm": 0.16728058457374573, "learning_rate": 2.2308849960185578e-05, "loss": 0.0132, "step": 70250 }, { "grad_norm": 0.13098160922527313, "learning_rate": 2.229508411567774e-05, "loss": 0.0126, "step": 70260 }, { "grad_norm": 0.19923707842826843, "learning_rate": 2.2281321300937823e-05, "loss": 0.0116, "step": 70270 }, { "grad_norm": 0.15363262593746185, "learning_rate": 2.2267561517470937e-05, "loss": 0.0106, "step": 70280 }, { "grad_norm": 0.12832513451576233, "learning_rate": 2.2253804766781845e-05, "loss": 0.0106, "step": 70290 }, { "grad_norm": 0.15148302912712097, "learning_rate": 2.2240051050374933e-05, "loss": 0.0125, "step": 70300 }, { "grad_norm": 0.1618281602859497, "learning_rate": 2.22263003697543e-05, "loss": 0.01, "step": 70310 }, { "grad_norm": 0.14799274504184723, "learning_rate": 2.2212552726423702e-05, "loss": 0.014, "step": 70320 }, { "grad_norm": 0.13840767741203308, "learning_rate": 2.2198808121886578e-05, "loss": 0.0095, "step": 70330 }, { "grad_norm": 0.12285343557596207, "learning_rate": 2.218506655764598e-05, "loss": 0.0112, "step": 70340 }, { "grad_norm": 0.17041850090026855, "learning_rate": 2.2171328035204704e-05, "loss": 0.0128, "step": 70350 }, { "grad_norm": 0.13973894715309143, "learning_rate": 2.2157592556065133e-05, "loss": 0.0129, "step": 70360 }, { "grad_norm": 0.18557924032211304, "learning_rate": 2.2143860121729382e-05, "loss": 0.0129, "step": 70370 }, { "grad_norm": 0.156370148062706, "learning_rate": 2.2130130733699206e-05, "loss": 0.0115, "step": 70380 }, { "grad_norm": 0.14412374794483185, "learning_rate": 2.2116404393476052e-05, "loss": 0.011, "step": 70390 }, { "grad_norm": 0.26330429315567017, "learning_rate": 2.2102681102560967e-05, "loss": 0.0171, "step": 70400 }, { "grad_norm": 0.1689651608467102, "learning_rate": 2.208896086245476e-05, "loss": 0.0114, "step": 70410 }, { "grad_norm": 0.1312740296125412, "learning_rate": 2.2075243674657808e-05, "loss": 0.01, "step": 70420 }, { "grad_norm": 0.14018210768699646, "learning_rate": 2.206152954067024e-05, "loss": 0.0113, "step": 70430 }, { "grad_norm": 0.1856551170349121, "learning_rate": 2.2047818461991758e-05, "loss": 0.0101, "step": 70440 }, { "grad_norm": 0.16522863507270813, "learning_rate": 2.2034110440121864e-05, "loss": 0.0095, "step": 70450 }, { "grad_norm": 0.11330472677946091, "learning_rate": 2.2020405476559585e-05, "loss": 0.0115, "step": 70460 }, { "grad_norm": 0.1910075843334198, "learning_rate": 2.2006703572803722e-05, "loss": 0.0115, "step": 70470 }, { "grad_norm": 0.18039770424365997, "learning_rate": 2.1993004730352646e-05, "loss": 0.0116, "step": 70480 }, { "grad_norm": 0.12670038640499115, "learning_rate": 2.1979308950704464e-05, "loss": 0.0111, "step": 70490 }, { "grad_norm": 0.13772064447402954, "learning_rate": 2.1965616235356923e-05, "loss": 0.0108, "step": 70500 }, { "grad_norm": 0.18160611391067505, "learning_rate": 2.1951926585807465e-05, "loss": 0.0143, "step": 70510 }, { "grad_norm": 0.1039094626903534, "learning_rate": 2.193824000355313e-05, "loss": 0.0112, "step": 70520 }, { "grad_norm": 0.1296701729297638, "learning_rate": 2.192455649009067e-05, "loss": 0.0091, "step": 70530 }, { "grad_norm": 0.15256555378437042, "learning_rate": 2.1910876046916523e-05, "loss": 0.0138, "step": 70540 }, { "grad_norm": 0.10371634364128113, "learning_rate": 2.1897198675526725e-05, "loss": 0.0089, "step": 70550 }, { "grad_norm": 0.1442732959985733, "learning_rate": 2.1883524377417024e-05, "loss": 0.0099, "step": 70560 }, { "grad_norm": 0.15409712493419647, "learning_rate": 2.1869853154082825e-05, "loss": 0.0102, "step": 70570 }, { "grad_norm": 0.113578200340271, "learning_rate": 2.1856185007019215e-05, "loss": 0.0144, "step": 70580 }, { "grad_norm": 0.1761438101530075, "learning_rate": 2.1842519937720874e-05, "loss": 0.0132, "step": 70590 }, { "grad_norm": 0.12331131100654602, "learning_rate": 2.182885794768224e-05, "loss": 0.0121, "step": 70600 }, { "grad_norm": 0.15818475186824799, "learning_rate": 2.1815199038397326e-05, "loss": 0.0141, "step": 70610 }, { "grad_norm": 0.1535000503063202, "learning_rate": 2.1801543211359893e-05, "loss": 0.0135, "step": 70620 }, { "grad_norm": 0.15732146799564362, "learning_rate": 2.178789046806326e-05, "loss": 0.0108, "step": 70630 }, { "grad_norm": 0.12763914465904236, "learning_rate": 2.1774240810000547e-05, "loss": 0.0121, "step": 70640 }, { "grad_norm": 0.18998819589614868, "learning_rate": 2.17605942386644e-05, "loss": 0.015, "step": 70650 }, { "grad_norm": 0.1496821939945221, "learning_rate": 2.174695075554723e-05, "loss": 0.015, "step": 70660 }, { "grad_norm": 0.131727933883667, "learning_rate": 2.173331036214103e-05, "loss": 0.0126, "step": 70670 }, { "grad_norm": 0.18115602433681488, "learning_rate": 2.171967305993753e-05, "loss": 0.0115, "step": 70680 }, { "grad_norm": 0.15990184247493744, "learning_rate": 2.1706038850428024e-05, "loss": 0.0137, "step": 70690 }, { "grad_norm": 0.16654281318187714, "learning_rate": 2.169240773510361e-05, "loss": 0.0105, "step": 70700 }, { "grad_norm": 0.17845755815505981, "learning_rate": 2.1678779715454906e-05, "loss": 0.0123, "step": 70710 }, { "grad_norm": 0.1737886369228363, "learning_rate": 2.166515479297229e-05, "loss": 0.01, "step": 70720 }, { "grad_norm": 0.14927737414836884, "learning_rate": 2.1651532969145732e-05, "loss": 0.0154, "step": 70730 }, { "grad_norm": 0.11562678962945938, "learning_rate": 2.1637914245464903e-05, "loss": 0.011, "step": 70740 }, { "grad_norm": 0.11302164196968079, "learning_rate": 2.1624298623419133e-05, "loss": 0.0117, "step": 70750 }, { "grad_norm": 0.16633164882659912, "learning_rate": 2.1610686104497413e-05, "loss": 0.0122, "step": 70760 }, { "grad_norm": 0.12344206869602203, "learning_rate": 2.1597076690188366e-05, "loss": 0.0156, "step": 70770 }, { "grad_norm": 0.17282430827617645, "learning_rate": 2.1583470381980315e-05, "loss": 0.0118, "step": 70780 }, { "grad_norm": 0.166412815451622, "learning_rate": 2.156986718136123e-05, "loss": 0.0118, "step": 70790 }, { "grad_norm": 0.1597423255443573, "learning_rate": 2.155626708981871e-05, "loss": 0.015, "step": 70800 }, { "grad_norm": 0.1979968398809433, "learning_rate": 2.1542670108840075e-05, "loss": 0.013, "step": 70810 }, { "grad_norm": 0.16145089268684387, "learning_rate": 2.152907623991221e-05, "loss": 0.0114, "step": 70820 }, { "grad_norm": 0.16230016946792603, "learning_rate": 2.1515485484521802e-05, "loss": 0.0122, "step": 70830 }, { "grad_norm": 0.15107379853725433, "learning_rate": 2.150189784415506e-05, "loss": 0.0126, "step": 70840 }, { "grad_norm": 0.19797950983047485, "learning_rate": 2.148831332029794e-05, "loss": 0.0133, "step": 70850 }, { "grad_norm": 0.09718858450651169, "learning_rate": 2.147473191443599e-05, "loss": 0.0106, "step": 70860 }, { "grad_norm": 0.12935592234134674, "learning_rate": 2.1461153628054492e-05, "loss": 0.014, "step": 70870 }, { "grad_norm": 0.19709891080856323, "learning_rate": 2.1447578462638284e-05, "loss": 0.0121, "step": 70880 }, { "grad_norm": 0.15496821701526642, "learning_rate": 2.143400641967201e-05, "loss": 0.0137, "step": 70890 }, { "grad_norm": 0.17233623564243317, "learning_rate": 2.1420437500639823e-05, "loss": 0.0098, "step": 70900 }, { "grad_norm": 0.1296791136264801, "learning_rate": 2.1406871707025646e-05, "loss": 0.009, "step": 70910 }, { "grad_norm": 0.1874239444732666, "learning_rate": 2.1393309040312963e-05, "loss": 0.0154, "step": 70920 }, { "grad_norm": 0.2081429809331894, "learning_rate": 2.137974950198502e-05, "loss": 0.0127, "step": 70930 }, { "grad_norm": 0.19318018853664398, "learning_rate": 2.1366193093524617e-05, "loss": 0.0124, "step": 70940 }, { "grad_norm": 0.14610609412193298, "learning_rate": 2.1352639816414288e-05, "loss": 0.0124, "step": 70950 }, { "grad_norm": 0.23007094860076904, "learning_rate": 2.1339089672136202e-05, "loss": 0.0121, "step": 70960 }, { "grad_norm": 0.1395714432001114, "learning_rate": 2.1325542662172193e-05, "loss": 0.0108, "step": 70970 }, { "grad_norm": 0.14252519607543945, "learning_rate": 2.1311998788003712e-05, "loss": 0.0084, "step": 70980 }, { "grad_norm": 0.10861828923225403, "learning_rate": 2.1298458051111907e-05, "loss": 0.0121, "step": 70990 }, { "grad_norm": 0.14070507884025574, "learning_rate": 2.1284920452977602e-05, "loss": 0.0107, "step": 71000 }, { "grad_norm": 0.1552283763885498, "learning_rate": 2.12713859950812e-05, "loss": 0.0135, "step": 71010 }, { "grad_norm": 0.16717153787612915, "learning_rate": 2.1257854678902832e-05, "loss": 0.0132, "step": 71020 }, { "grad_norm": 0.19519557058811188, "learning_rate": 2.124432650592227e-05, "loss": 0.0142, "step": 71030 }, { "grad_norm": 0.15072715282440186, "learning_rate": 2.1230801477618937e-05, "loss": 0.0115, "step": 71040 }, { "grad_norm": 0.19926519691944122, "learning_rate": 2.121727959547189e-05, "loss": 0.0105, "step": 71050 }, { "grad_norm": 0.15687784552574158, "learning_rate": 2.1203760860959886e-05, "loss": 0.0113, "step": 71060 }, { "grad_norm": 0.16388161480426788, "learning_rate": 2.119024527556127e-05, "loss": 0.012, "step": 71070 }, { "grad_norm": 0.14789120852947235, "learning_rate": 2.117673284075415e-05, "loss": 0.0128, "step": 71080 }, { "grad_norm": 0.15437348186969757, "learning_rate": 2.1163223558016165e-05, "loss": 0.0117, "step": 71090 }, { "grad_norm": 0.15444622933864594, "learning_rate": 2.1149717428824722e-05, "loss": 0.0097, "step": 71100 }, { "grad_norm": 0.23039555549621582, "learning_rate": 2.1136214454656784e-05, "loss": 0.0137, "step": 71110 }, { "grad_norm": 0.4439053535461426, "learning_rate": 2.1122714636989056e-05, "loss": 0.01, "step": 71120 }, { "grad_norm": 0.18458795547485352, "learning_rate": 2.1109217977297823e-05, "loss": 0.0117, "step": 71130 }, { "grad_norm": 0.1890411376953125, "learning_rate": 2.1095724477059077e-05, "loss": 0.0117, "step": 71140 }, { "grad_norm": 0.13734374940395355, "learning_rate": 2.1082234137748434e-05, "loss": 0.0093, "step": 71150 }, { "grad_norm": 0.15396538376808167, "learning_rate": 2.106874696084122e-05, "loss": 0.0116, "step": 71160 }, { "grad_norm": 0.14054614305496216, "learning_rate": 2.1055262947812315e-05, "loss": 0.0113, "step": 71170 }, { "grad_norm": 0.14002250134944916, "learning_rate": 2.104178210013636e-05, "loss": 0.0119, "step": 71180 }, { "grad_norm": 0.17257757484912872, "learning_rate": 2.1028304419287554e-05, "loss": 0.0122, "step": 71190 }, { "grad_norm": 0.1489362269639969, "learning_rate": 2.101482990673982e-05, "loss": 0.0121, "step": 71200 }, { "grad_norm": 0.15222576260566711, "learning_rate": 2.1001358563966707e-05, "loss": 0.0106, "step": 71210 }, { "grad_norm": 0.19254592061042786, "learning_rate": 2.098789039244144e-05, "loss": 0.0124, "step": 71220 }, { "grad_norm": 0.1427304595708847, "learning_rate": 2.097442539363685e-05, "loss": 0.013, "step": 71230 }, { "grad_norm": 0.1219564899802208, "learning_rate": 2.0960963569025448e-05, "loss": 0.0114, "step": 71240 }, { "grad_norm": 0.15889045596122742, "learning_rate": 2.0947504920079435e-05, "loss": 0.011, "step": 71250 }, { "grad_norm": 0.12037011235952377, "learning_rate": 2.093404944827059e-05, "loss": 0.0129, "step": 71260 }, { "grad_norm": 0.17736411094665527, "learning_rate": 2.0920597155070394e-05, "loss": 0.013, "step": 71270 }, { "grad_norm": 0.22140279412269592, "learning_rate": 2.090714804194997e-05, "loss": 0.0169, "step": 71280 }, { "grad_norm": 0.16502605378627777, "learning_rate": 2.0893702110380114e-05, "loss": 0.0098, "step": 71290 }, { "grad_norm": 0.1588905304670334, "learning_rate": 2.0880259361831216e-05, "loss": 0.0105, "step": 71300 }, { "grad_norm": 0.17103028297424316, "learning_rate": 2.0866819797773397e-05, "loss": 0.012, "step": 71310 }, { "grad_norm": 0.17299693822860718, "learning_rate": 2.085338341967634e-05, "loss": 0.013, "step": 71320 }, { "grad_norm": 0.19648516178131104, "learning_rate": 2.083995022900946e-05, "loss": 0.0108, "step": 71330 }, { "grad_norm": 0.13698817789554596, "learning_rate": 2.0826520227241768e-05, "loss": 0.0132, "step": 71340 }, { "grad_norm": 0.158348947763443, "learning_rate": 2.081309341584199e-05, "loss": 0.0126, "step": 71350 }, { "grad_norm": 0.17678752541542053, "learning_rate": 2.0799669796278413e-05, "loss": 0.0112, "step": 71360 }, { "grad_norm": 0.19638817012310028, "learning_rate": 2.078624937001906e-05, "loss": 0.0139, "step": 71370 }, { "grad_norm": 0.1650267243385315, "learning_rate": 2.0772832138531544e-05, "loss": 0.0113, "step": 71380 }, { "grad_norm": 0.2652965784072876, "learning_rate": 2.0759418103283158e-05, "loss": 0.0111, "step": 71390 }, { "grad_norm": 0.132976233959198, "learning_rate": 2.0746007265740846e-05, "loss": 0.0122, "step": 71400 }, { "grad_norm": 0.16453298926353455, "learning_rate": 2.0732599627371215e-05, "loss": 0.0104, "step": 71410 }, { "grad_norm": 0.1240418329834938, "learning_rate": 2.0719195189640468e-05, "loss": 0.0136, "step": 71420 }, { "grad_norm": 0.15661980211734772, "learning_rate": 2.0705793954014528e-05, "loss": 0.018, "step": 71430 }, { "grad_norm": 0.1325514018535614, "learning_rate": 2.06923959219589e-05, "loss": 0.0116, "step": 71440 }, { "grad_norm": 0.14476309716701508, "learning_rate": 2.0679001094938793e-05, "loss": 0.0137, "step": 71450 }, { "grad_norm": 0.18939150869846344, "learning_rate": 2.066560947441904e-05, "loss": 0.0109, "step": 71460 }, { "grad_norm": 0.12830333411693573, "learning_rate": 2.0652221061864152e-05, "loss": 0.0109, "step": 71470 }, { "grad_norm": 0.27496063709259033, "learning_rate": 2.0638835858738226e-05, "loss": 0.0112, "step": 71480 }, { "grad_norm": 0.08442635834217072, "learning_rate": 2.0625453866505068e-05, "loss": 0.0102, "step": 71490 }, { "grad_norm": 0.14906705915927887, "learning_rate": 2.0612075086628123e-05, "loss": 0.0114, "step": 71500 }, { "grad_norm": 0.12800057232379913, "learning_rate": 2.059869952057045e-05, "loss": 0.0114, "step": 71510 }, { "grad_norm": 0.11723113805055618, "learning_rate": 2.0585327169794793e-05, "loss": 0.0106, "step": 71520 }, { "grad_norm": 0.13726186752319336, "learning_rate": 2.0571958035763534e-05, "loss": 0.0119, "step": 71530 }, { "grad_norm": 0.1483924686908722, "learning_rate": 2.0558592119938715e-05, "loss": 0.0138, "step": 71540 }, { "grad_norm": 0.11388277262449265, "learning_rate": 2.054522942378198e-05, "loss": 0.0098, "step": 71550 }, { "grad_norm": 0.1909266859292984, "learning_rate": 2.053186994875469e-05, "loss": 0.0102, "step": 71560 }, { "grad_norm": 0.1523391306400299, "learning_rate": 2.0518513696317775e-05, "loss": 0.0119, "step": 71570 }, { "grad_norm": 0.14886333048343658, "learning_rate": 2.050516066793188e-05, "loss": 0.0091, "step": 71580 }, { "grad_norm": 0.17234458029270172, "learning_rate": 2.0491810865057276e-05, "loss": 0.0117, "step": 71590 }, { "grad_norm": 0.18078409135341644, "learning_rate": 2.0478464289153882e-05, "loss": 0.0132, "step": 71600 }, { "grad_norm": 0.1644824743270874, "learning_rate": 2.0465120941681232e-05, "loss": 0.0123, "step": 71610 }, { "grad_norm": 0.14647114276885986, "learning_rate": 2.0451780824098566e-05, "loss": 0.0142, "step": 71620 }, { "grad_norm": 0.12179826200008392, "learning_rate": 2.043844393786471e-05, "loss": 0.0108, "step": 71630 }, { "grad_norm": 0.12838786840438843, "learning_rate": 2.0425110284438193e-05, "loss": 0.0113, "step": 71640 }, { "grad_norm": 0.1404702216386795, "learning_rate": 2.041177986527712e-05, "loss": 0.0124, "step": 71650 }, { "grad_norm": 0.14935196936130524, "learning_rate": 2.0398452681839347e-05, "loss": 0.0106, "step": 71660 }, { "grad_norm": 0.1368408352136612, "learning_rate": 2.0385128735582272e-05, "loss": 0.0097, "step": 71670 }, { "grad_norm": 0.24481208622455597, "learning_rate": 2.0371808027963003e-05, "loss": 0.014, "step": 71680 }, { "grad_norm": 0.13803179562091827, "learning_rate": 2.035849056043825e-05, "loss": 0.0134, "step": 71690 }, { "grad_norm": 0.17046093940734863, "learning_rate": 2.03451763344644e-05, "loss": 0.0101, "step": 71700 }, { "grad_norm": 0.12641191482543945, "learning_rate": 2.0331865351497482e-05, "loss": 0.0091, "step": 71710 }, { "grad_norm": 0.1875782608985901, "learning_rate": 2.0318557612993185e-05, "loss": 0.009, "step": 71720 }, { "grad_norm": 0.1980365812778473, "learning_rate": 2.0305253120406782e-05, "loss": 0.0129, "step": 71730 }, { "grad_norm": 0.17369498312473297, "learning_rate": 2.029195187519325e-05, "loss": 0.0127, "step": 71740 }, { "grad_norm": 0.1064990907907486, "learning_rate": 2.0278653878807217e-05, "loss": 0.0106, "step": 71750 }, { "grad_norm": 0.14184951782226562, "learning_rate": 2.0265359132702892e-05, "loss": 0.0093, "step": 71760 }, { "grad_norm": 0.13881279528141022, "learning_rate": 2.0252067638334182e-05, "loss": 0.0089, "step": 71770 }, { "grad_norm": 0.13809296488761902, "learning_rate": 2.023877939715464e-05, "loss": 0.0116, "step": 71780 }, { "grad_norm": 0.15396036207675934, "learning_rate": 2.0225494410617448e-05, "loss": 0.0095, "step": 71790 }, { "grad_norm": 0.13295255601406097, "learning_rate": 2.02122126801754e-05, "loss": 0.0099, "step": 71800 }, { "grad_norm": 0.1146969348192215, "learning_rate": 2.0198934207281018e-05, "loss": 0.0124, "step": 71810 }, { "grad_norm": 0.1380293220281601, "learning_rate": 2.0185658993386357e-05, "loss": 0.0118, "step": 71820 }, { "grad_norm": 0.11230699717998505, "learning_rate": 2.0172387039943226e-05, "loss": 0.0101, "step": 71830 }, { "grad_norm": 0.17066030204296112, "learning_rate": 2.0159118348402968e-05, "loss": 0.0099, "step": 71840 }, { "grad_norm": 0.15501470863819122, "learning_rate": 2.0145852920216697e-05, "loss": 0.0118, "step": 71850 }, { "grad_norm": 0.13172805309295654, "learning_rate": 2.0132590756835045e-05, "loss": 0.0083, "step": 71860 }, { "grad_norm": 0.1867201179265976, "learning_rate": 2.0119331859708386e-05, "loss": 0.0115, "step": 71870 }, { "grad_norm": 0.1716025173664093, "learning_rate": 2.0106076230286654e-05, "loss": 0.011, "step": 71880 }, { "grad_norm": 0.135774165391922, "learning_rate": 2.0092823870019488e-05, "loss": 0.0115, "step": 71890 }, { "grad_norm": 0.18203163146972656, "learning_rate": 2.0079574780356115e-05, "loss": 0.0115, "step": 71900 }, { "grad_norm": 0.0975155308842659, "learning_rate": 2.0066328962745486e-05, "loss": 0.0102, "step": 71910 }, { "grad_norm": 0.19921502470970154, "learning_rate": 2.00530864186361e-05, "loss": 0.0106, "step": 71920 }, { "grad_norm": 0.17277485132217407, "learning_rate": 2.003984714947618e-05, "loss": 0.0119, "step": 71930 }, { "grad_norm": 0.1766999214887619, "learning_rate": 2.0026611156713515e-05, "loss": 0.0117, "step": 71940 }, { "grad_norm": 0.17628738284111023, "learning_rate": 2.0013378441795584e-05, "loss": 0.0136, "step": 71950 }, { "grad_norm": 0.311171293258667, "learning_rate": 2.0000149006169506e-05, "loss": 0.0106, "step": 71960 }, { "grad_norm": 0.19827322661876678, "learning_rate": 1.9986922851282048e-05, "loss": 0.0135, "step": 71970 }, { "grad_norm": 0.16487297415733337, "learning_rate": 1.997369997857956e-05, "loss": 0.0124, "step": 71980 }, { "grad_norm": 0.17250069975852966, "learning_rate": 1.9960480389508103e-05, "loss": 0.0121, "step": 71990 }, { "grad_norm": 0.11106269061565399, "learning_rate": 1.9947264085513363e-05, "loss": 0.0118, "step": 72000 }, { "grad_norm": 0.1477181315422058, "learning_rate": 1.993405106804062e-05, "loss": 0.0113, "step": 72010 }, { "grad_norm": 0.2125018686056137, "learning_rate": 1.9920841338534867e-05, "loss": 0.0124, "step": 72020 }, { "grad_norm": 0.18219919502735138, "learning_rate": 1.9907634898440646e-05, "loss": 0.01, "step": 72030 }, { "grad_norm": 0.162958562374115, "learning_rate": 1.989443174920227e-05, "loss": 0.0126, "step": 72040 }, { "grad_norm": 0.18193064630031586, "learning_rate": 1.988123189226355e-05, "loss": 0.0117, "step": 72050 }, { "grad_norm": 0.17094051837921143, "learning_rate": 1.986803532906805e-05, "loss": 0.0132, "step": 72060 }, { "grad_norm": 0.1899312287569046, "learning_rate": 1.985484206105889e-05, "loss": 0.0124, "step": 72070 }, { "grad_norm": 0.11795463413000107, "learning_rate": 1.9841652089678887e-05, "loss": 0.0129, "step": 72080 }, { "grad_norm": 0.22416509687900543, "learning_rate": 1.9828465416370435e-05, "loss": 0.0138, "step": 72090 }, { "grad_norm": 0.1923329383134842, "learning_rate": 1.9815282042575685e-05, "loss": 0.0106, "step": 72100 }, { "grad_norm": 0.12917771935462952, "learning_rate": 1.9802101969736285e-05, "loss": 0.014, "step": 72110 }, { "grad_norm": 0.10227496922016144, "learning_rate": 1.9788925199293633e-05, "loss": 0.0097, "step": 72120 }, { "grad_norm": 0.11792305111885071, "learning_rate": 1.977575173268868e-05, "loss": 0.0128, "step": 72130 }, { "grad_norm": 0.14225715398788452, "learning_rate": 1.9762581571362094e-05, "loss": 0.013, "step": 72140 }, { "grad_norm": 0.17464569211006165, "learning_rate": 1.9749414716754112e-05, "loss": 0.0142, "step": 72150 }, { "grad_norm": 0.14434406161308289, "learning_rate": 1.973625117030465e-05, "loss": 0.0112, "step": 72160 }, { "grad_norm": 0.1521812081336975, "learning_rate": 1.972309093345326e-05, "loss": 0.0124, "step": 72170 }, { "grad_norm": 0.1512293666601181, "learning_rate": 1.970993400763915e-05, "loss": 0.0135, "step": 72180 }, { "grad_norm": 0.2134338766336441, "learning_rate": 1.969678039430109e-05, "loss": 0.013, "step": 72190 }, { "grad_norm": 0.15355293452739716, "learning_rate": 1.9683630094877564e-05, "loss": 0.0124, "step": 72200 }, { "grad_norm": 0.14220599830150604, "learning_rate": 1.9670483110806696e-05, "loss": 0.0129, "step": 72210 }, { "grad_norm": 0.13381928205490112, "learning_rate": 1.9657339443526172e-05, "loss": 0.0104, "step": 72220 }, { "grad_norm": 0.12360724806785583, "learning_rate": 1.964419909447339e-05, "loss": 0.0101, "step": 72230 }, { "grad_norm": 0.13968539237976074, "learning_rate": 1.963106206508536e-05, "loss": 0.0132, "step": 72240 }, { "grad_norm": 0.1807510256767273, "learning_rate": 1.9617928356798738e-05, "loss": 0.0122, "step": 72250 }, { "grad_norm": 0.15212568640708923, "learning_rate": 1.960479797104977e-05, "loss": 0.0118, "step": 72260 }, { "grad_norm": 0.17568306624889374, "learning_rate": 1.9591670909274418e-05, "loss": 0.0098, "step": 72270 }, { "grad_norm": 0.15324044227600098, "learning_rate": 1.9578547172908184e-05, "loss": 0.0102, "step": 72280 }, { "grad_norm": 0.13891620934009552, "learning_rate": 1.9565426763386325e-05, "loss": 0.0112, "step": 72290 }, { "grad_norm": 0.11509107053279877, "learning_rate": 1.9552309682143622e-05, "loss": 0.0113, "step": 72300 }, { "grad_norm": 0.18639390170574188, "learning_rate": 1.9539195930614574e-05, "loss": 0.0114, "step": 72310 }, { "grad_norm": 0.150421142578125, "learning_rate": 1.952608551023324e-05, "loss": 0.0079, "step": 72320 }, { "grad_norm": 0.17523358762264252, "learning_rate": 1.9512978422433397e-05, "loss": 0.0133, "step": 72330 }, { "grad_norm": 0.140696182847023, "learning_rate": 1.9499874668648376e-05, "loss": 0.0092, "step": 72340 }, { "grad_norm": 0.12103582918643951, "learning_rate": 1.9486774250311213e-05, "loss": 0.0091, "step": 72350 }, { "grad_norm": 0.10752540826797485, "learning_rate": 1.9473677168854537e-05, "loss": 0.0118, "step": 72360 }, { "grad_norm": 0.21983008086681366, "learning_rate": 1.946058342571065e-05, "loss": 0.0106, "step": 72370 }, { "grad_norm": 0.16493476927280426, "learning_rate": 1.9447493022311424e-05, "loss": 0.0099, "step": 72380 }, { "grad_norm": 0.11025695502758026, "learning_rate": 1.9434405960088442e-05, "loss": 0.0084, "step": 72390 }, { "grad_norm": 0.14973706007003784, "learning_rate": 1.9421322240472857e-05, "loss": 0.0121, "step": 72400 }, { "grad_norm": 0.1307673156261444, "learning_rate": 1.9408241864895494e-05, "loss": 0.0088, "step": 72410 }, { "grad_norm": 0.13542725145816803, "learning_rate": 1.939516483478681e-05, "loss": 0.0097, "step": 72420 }, { "grad_norm": 0.10669152438640594, "learning_rate": 1.9382091151576902e-05, "loss": 0.0093, "step": 72430 }, { "grad_norm": 0.16420871019363403, "learning_rate": 1.9369020816695454e-05, "loss": 0.0108, "step": 72440 }, { "grad_norm": 0.1883925348520279, "learning_rate": 1.935595383157184e-05, "loss": 0.0121, "step": 72450 }, { "grad_norm": 0.09681116044521332, "learning_rate": 1.9342890197635057e-05, "loss": 0.0142, "step": 72460 }, { "grad_norm": 0.158402681350708, "learning_rate": 1.9329829916313685e-05, "loss": 0.0097, "step": 72470 }, { "grad_norm": 0.1996186077594757, "learning_rate": 1.9316772989036013e-05, "loss": 0.0145, "step": 72480 }, { "grad_norm": 0.11890795826911926, "learning_rate": 1.930371941722991e-05, "loss": 0.0108, "step": 72490 }, { "grad_norm": 0.18681977689266205, "learning_rate": 1.929066920232292e-05, "loss": 0.0124, "step": 72500 }, { "grad_norm": 0.14603793621063232, "learning_rate": 1.9277622345742152e-05, "loss": 0.0098, "step": 72510 }, { "grad_norm": 0.19485236704349518, "learning_rate": 1.926457884891444e-05, "loss": 0.0115, "step": 72520 }, { "grad_norm": 0.15841881930828094, "learning_rate": 1.9251538713266147e-05, "loss": 0.009, "step": 72530 }, { "grad_norm": 0.1291409581899643, "learning_rate": 1.9238501940223358e-05, "loss": 0.0132, "step": 72540 }, { "grad_norm": 0.1294936239719391, "learning_rate": 1.922546853121175e-05, "loss": 0.0126, "step": 72550 }, { "grad_norm": 0.16354823112487793, "learning_rate": 1.9212438487656647e-05, "loss": 0.0114, "step": 72560 }, { "grad_norm": 0.12174840271472931, "learning_rate": 1.9199411810982965e-05, "loss": 0.0102, "step": 72570 }, { "grad_norm": 0.15988488495349884, "learning_rate": 1.918638850261532e-05, "loss": 0.0135, "step": 72580 }, { "grad_norm": 0.133721724152565, "learning_rate": 1.9173368563977885e-05, "loss": 0.0112, "step": 72590 }, { "grad_norm": 0.1350143849849701, "learning_rate": 1.9160351996494513e-05, "loss": 0.012, "step": 72600 }, { "grad_norm": 0.16547293961048126, "learning_rate": 1.9147338801588677e-05, "loss": 0.0114, "step": 72610 }, { "grad_norm": 0.19842684268951416, "learning_rate": 1.913432898068351e-05, "loss": 0.0118, "step": 72620 }, { "grad_norm": 0.13559933006763458, "learning_rate": 1.912132253520169e-05, "loss": 0.0127, "step": 72630 }, { "grad_norm": 0.1848496049642563, "learning_rate": 1.910831946656564e-05, "loss": 0.01, "step": 72640 }, { "grad_norm": 0.14768484234809875, "learning_rate": 1.909531977619731e-05, "loss": 0.0099, "step": 72650 }, { "grad_norm": 0.170668363571167, "learning_rate": 1.908232346551834e-05, "loss": 0.0113, "step": 72660 }, { "grad_norm": 0.19184309244155884, "learning_rate": 1.9069330535949993e-05, "loss": 0.0117, "step": 72670 }, { "grad_norm": 0.12653475999832153, "learning_rate": 1.9056340988913167e-05, "loss": 0.0112, "step": 72680 }, { "grad_norm": 0.13788552582263947, "learning_rate": 1.9043354825828357e-05, "loss": 0.0091, "step": 72690 }, { "grad_norm": 0.18510746955871582, "learning_rate": 1.9030372048115724e-05, "loss": 0.0122, "step": 72700 }, { "grad_norm": 0.1522907018661499, "learning_rate": 1.9017392657195056e-05, "loss": 0.0113, "step": 72710 }, { "grad_norm": 0.12978284060955048, "learning_rate": 1.9004416654485724e-05, "loss": 0.0088, "step": 72720 }, { "grad_norm": 0.1274413913488388, "learning_rate": 1.8991444041406785e-05, "loss": 0.0115, "step": 72730 }, { "grad_norm": 0.14235305786132812, "learning_rate": 1.8978474819376912e-05, "loss": 0.0109, "step": 72740 }, { "grad_norm": 0.13011212646961212, "learning_rate": 1.8965508989814406e-05, "loss": 0.0113, "step": 72750 }, { "grad_norm": 0.11770545691251755, "learning_rate": 1.8952546554137158e-05, "loss": 0.0132, "step": 72760 }, { "grad_norm": 0.1815837174654007, "learning_rate": 1.8939587513762757e-05, "loss": 0.0113, "step": 72770 }, { "grad_norm": 0.15356045961380005, "learning_rate": 1.892663187010835e-05, "loss": 0.0147, "step": 72780 }, { "grad_norm": 0.15713538229465485, "learning_rate": 1.891367962459077e-05, "loss": 0.0117, "step": 72790 }, { "grad_norm": 0.1803409606218338, "learning_rate": 1.890073077862644e-05, "loss": 0.0116, "step": 72800 }, { "grad_norm": 0.14029596745967865, "learning_rate": 1.888778533363146e-05, "loss": 0.0098, "step": 72810 }, { "grad_norm": 0.21834319829940796, "learning_rate": 1.8874843291021478e-05, "loss": 0.0139, "step": 72820 }, { "grad_norm": 0.1872924566268921, "learning_rate": 1.8861904652211853e-05, "loss": 0.0102, "step": 72830 }, { "grad_norm": 0.12114708125591278, "learning_rate": 1.8848969418617507e-05, "loss": 0.0102, "step": 72840 }, { "grad_norm": 0.09575014561414719, "learning_rate": 1.8836037591653044e-05, "loss": 0.0113, "step": 72850 }, { "grad_norm": 0.1497349888086319, "learning_rate": 1.882310917273262e-05, "loss": 0.0124, "step": 72860 }, { "grad_norm": 0.15172068774700165, "learning_rate": 1.881018416327014e-05, "loss": 0.0114, "step": 72870 }, { "grad_norm": 0.18262332677841187, "learning_rate": 1.8797262564679014e-05, "loss": 0.0131, "step": 72880 }, { "grad_norm": 0.12023599445819855, "learning_rate": 1.8784344378372355e-05, "loss": 0.0085, "step": 72890 }, { "grad_norm": 0.19544751942157745, "learning_rate": 1.8771429605762843e-05, "loss": 0.0142, "step": 72900 }, { "grad_norm": 0.13542664051055908, "learning_rate": 1.8758518248262835e-05, "loss": 0.0099, "step": 72910 }, { "grad_norm": 0.1286620944738388, "learning_rate": 1.8745610307284305e-05, "loss": 0.0099, "step": 72920 }, { "grad_norm": 0.16738663613796234, "learning_rate": 1.8732705784238862e-05, "loss": 0.0123, "step": 72930 }, { "grad_norm": 0.1600388139486313, "learning_rate": 1.8719804680537682e-05, "loss": 0.0128, "step": 72940 }, { "grad_norm": 0.20725609362125397, "learning_rate": 1.870690699759164e-05, "loss": 0.0085, "step": 72950 }, { "grad_norm": 0.10680598765611649, "learning_rate": 1.8694012736811213e-05, "loss": 0.013, "step": 72960 }, { "grad_norm": 0.16879470646381378, "learning_rate": 1.8681121899606463e-05, "loss": 0.0122, "step": 72970 }, { "grad_norm": 0.21936288475990295, "learning_rate": 1.8668234487387144e-05, "loss": 0.0102, "step": 72980 }, { "grad_norm": 0.19298335909843445, "learning_rate": 1.8655350501562592e-05, "loss": 0.0084, "step": 72990 }, { "grad_norm": 0.1760551482439041, "learning_rate": 1.8642469943541796e-05, "loss": 0.0112, "step": 73000 }, { "grad_norm": 0.17081114649772644, "learning_rate": 1.8629592814733328e-05, "loss": 0.0096, "step": 73010 }, { "grad_norm": 0.1397620141506195, "learning_rate": 1.8616719116545438e-05, "loss": 0.0109, "step": 73020 }, { "grad_norm": 0.11993630975484848, "learning_rate": 1.8603848850385952e-05, "loss": 0.0093, "step": 73030 }, { "grad_norm": 0.13328078389167786, "learning_rate": 1.859098201766236e-05, "loss": 0.0114, "step": 73040 }, { "grad_norm": 0.14456650614738464, "learning_rate": 1.8578118619781726e-05, "loss": 0.0102, "step": 73050 }, { "grad_norm": 0.10479272156953812, "learning_rate": 1.856525865815083e-05, "loss": 0.0119, "step": 73060 }, { "grad_norm": 0.626184344291687, "learning_rate": 1.8552402134175966e-05, "loss": 0.0129, "step": 73070 }, { "grad_norm": 0.2251596599817276, "learning_rate": 1.8539549049263143e-05, "loss": 0.0116, "step": 73080 }, { "grad_norm": 0.11824259907007217, "learning_rate": 1.8526699404817917e-05, "loss": 0.0103, "step": 73090 }, { "grad_norm": 0.19001248478889465, "learning_rate": 1.8513853202245546e-05, "loss": 0.0125, "step": 73100 }, { "grad_norm": 0.2154320478439331, "learning_rate": 1.850101044295081e-05, "loss": 0.0108, "step": 73110 }, { "grad_norm": 0.14282196760177612, "learning_rate": 1.848817112833825e-05, "loss": 0.0087, "step": 73120 }, { "grad_norm": 0.151325985789299, "learning_rate": 1.8475335259811905e-05, "loss": 0.0115, "step": 73130 }, { "grad_norm": 0.1265900880098343, "learning_rate": 1.8462502838775508e-05, "loss": 0.01, "step": 73140 }, { "grad_norm": 0.14387743175029755, "learning_rate": 1.8449673866632378e-05, "loss": 0.0145, "step": 73150 }, { "grad_norm": 0.14192523062229156, "learning_rate": 1.8436848344785473e-05, "loss": 0.0118, "step": 73160 }, { "grad_norm": 0.18596628308296204, "learning_rate": 1.8424026274637378e-05, "loss": 0.0126, "step": 73170 }, { "grad_norm": 0.12436063587665558, "learning_rate": 1.8411207657590312e-05, "loss": 0.0137, "step": 73180 }, { "grad_norm": 0.10435479879379272, "learning_rate": 1.839839249504607e-05, "loss": 0.0132, "step": 73190 }, { "grad_norm": 0.18639157712459564, "learning_rate": 1.8385580788406108e-05, "loss": 0.0099, "step": 73200 }, { "grad_norm": 0.12132000178098679, "learning_rate": 1.837277253907152e-05, "loss": 0.012, "step": 73210 }, { "grad_norm": 0.20048609375953674, "learning_rate": 1.8359967748442957e-05, "loss": 0.0141, "step": 73220 }, { "grad_norm": 0.23970168828964233, "learning_rate": 1.8347166417920774e-05, "loss": 0.0129, "step": 73230 }, { "grad_norm": 0.18206042051315308, "learning_rate": 1.833436854890484e-05, "loss": 0.0113, "step": 73240 }, { "grad_norm": 0.16124804317951202, "learning_rate": 1.8321574142794796e-05, "loss": 0.011, "step": 73250 }, { "grad_norm": 0.1287461370229721, "learning_rate": 1.8308783200989748e-05, "loss": 0.0104, "step": 73260 }, { "grad_norm": 0.13849791884422302, "learning_rate": 1.829599572488855e-05, "loss": 0.0108, "step": 73270 }, { "grad_norm": 0.1234239712357521, "learning_rate": 1.828321171588957e-05, "loss": 0.0097, "step": 73280 }, { "grad_norm": 0.14853708446025848, "learning_rate": 1.827043117539089e-05, "loss": 0.0119, "step": 73290 }, { "grad_norm": 0.16044040024280548, "learning_rate": 1.8257654104790116e-05, "loss": 0.0122, "step": 73300 }, { "grad_norm": 0.13896650075912476, "learning_rate": 1.82448805054846e-05, "loss": 0.0095, "step": 73310 }, { "grad_norm": 0.12674711644649506, "learning_rate": 1.823211037887119e-05, "loss": 0.0103, "step": 73320 }, { "grad_norm": 0.16314822435379028, "learning_rate": 1.821934372634645e-05, "loss": 0.0091, "step": 73330 }, { "grad_norm": 0.18262609839439392, "learning_rate": 1.8206580549306472e-05, "loss": 0.0116, "step": 73340 }, { "grad_norm": 0.1276446431875229, "learning_rate": 1.819382084914707e-05, "loss": 0.0125, "step": 73350 }, { "grad_norm": 0.17354215681552887, "learning_rate": 1.8181064627263577e-05, "loss": 0.0104, "step": 73360 }, { "grad_norm": 0.19094650447368622, "learning_rate": 1.8168311885051022e-05, "loss": 0.0128, "step": 73370 }, { "grad_norm": 0.16193880140781403, "learning_rate": 1.815556262390402e-05, "loss": 0.0123, "step": 73380 }, { "grad_norm": 0.15501637756824493, "learning_rate": 1.8142816845216826e-05, "loss": 0.011, "step": 73390 }, { "grad_norm": 0.12609760463237762, "learning_rate": 1.8130074550383276e-05, "loss": 0.0119, "step": 73400 }, { "grad_norm": 0.18175679445266724, "learning_rate": 1.8117335740796848e-05, "loss": 0.0088, "step": 73410 }, { "grad_norm": 0.17948196828365326, "learning_rate": 1.810460041785067e-05, "loss": 0.0132, "step": 73420 }, { "grad_norm": 0.1310252845287323, "learning_rate": 1.809186858293742e-05, "loss": 0.011, "step": 73430 }, { "grad_norm": 0.15807510912418365, "learning_rate": 1.8079140237449448e-05, "loss": 0.0135, "step": 73440 }, { "grad_norm": 0.17155742645263672, "learning_rate": 1.8066415382778707e-05, "loss": 0.0099, "step": 73450 }, { "grad_norm": 0.13158860802650452, "learning_rate": 1.8053694020316786e-05, "loss": 0.0102, "step": 73460 }, { "grad_norm": 0.15086260437965393, "learning_rate": 1.8040976151454837e-05, "loss": 0.0095, "step": 73470 }, { "grad_norm": 0.15332970023155212, "learning_rate": 1.8028261777583704e-05, "loss": 0.0108, "step": 73480 }, { "grad_norm": 0.15666289627552032, "learning_rate": 1.8015550900093758e-05, "loss": 0.0142, "step": 73490 }, { "grad_norm": 0.1374644786119461, "learning_rate": 1.800284352037512e-05, "loss": 0.0115, "step": 73500 }, { "grad_norm": 0.09460444003343582, "learning_rate": 1.799013963981738e-05, "loss": 0.0124, "step": 73510 }, { "grad_norm": 0.1962278187274933, "learning_rate": 1.797743925980987e-05, "loss": 0.0122, "step": 73520 }, { "grad_norm": 0.12416762858629227, "learning_rate": 1.7964742381741435e-05, "loss": 0.0082, "step": 73530 }, { "grad_norm": 0.10055333375930786, "learning_rate": 1.7952049007000627e-05, "loss": 0.0113, "step": 73540 }, { "grad_norm": 0.11406610906124115, "learning_rate": 1.7939359136975535e-05, "loss": 0.0103, "step": 73550 }, { "grad_norm": 0.1766965687274933, "learning_rate": 1.7926672773053932e-05, "loss": 0.013, "step": 73560 }, { "grad_norm": 0.14843404293060303, "learning_rate": 1.791398991662317e-05, "loss": 0.0107, "step": 73570 }, { "grad_norm": 0.1527218520641327, "learning_rate": 1.7901310569070244e-05, "loss": 0.0111, "step": 73580 }, { "grad_norm": 0.17828865349292755, "learning_rate": 1.7888634731781724e-05, "loss": 0.0115, "step": 73590 }, { "grad_norm": 0.25960204005241394, "learning_rate": 1.7875962406143844e-05, "loss": 0.0123, "step": 73600 }, { "grad_norm": 0.21620164811611176, "learning_rate": 1.7863293593542403e-05, "loss": 0.0116, "step": 73610 }, { "grad_norm": 0.11564359813928604, "learning_rate": 1.7850628295362853e-05, "loss": 0.0093, "step": 73620 }, { "grad_norm": 0.11560793966054916, "learning_rate": 1.7837966512990263e-05, "loss": 0.0117, "step": 73630 }, { "grad_norm": 0.16796058416366577, "learning_rate": 1.782530824780931e-05, "loss": 0.0131, "step": 73640 }, { "grad_norm": 0.10921087116003036, "learning_rate": 1.7812653501204263e-05, "loss": 0.0113, "step": 73650 }, { "grad_norm": 0.2814585268497467, "learning_rate": 1.780000227455903e-05, "loss": 0.0138, "step": 73660 }, { "grad_norm": 0.17445793747901917, "learning_rate": 1.778735456925716e-05, "loss": 0.011, "step": 73670 }, { "grad_norm": 0.1219799593091011, "learning_rate": 1.7774710386681737e-05, "loss": 0.0129, "step": 73680 }, { "grad_norm": 0.13077019155025482, "learning_rate": 1.7762069728215545e-05, "loss": 0.0116, "step": 73690 }, { "grad_norm": 0.17759069800376892, "learning_rate": 1.7749432595240932e-05, "loss": 0.0122, "step": 73700 }, { "grad_norm": 0.15163648128509521, "learning_rate": 1.773679898913991e-05, "loss": 0.0126, "step": 73710 }, { "grad_norm": 0.16171760857105255, "learning_rate": 1.7724168911294015e-05, "loss": 0.0105, "step": 73720 }, { "grad_norm": 0.186781108379364, "learning_rate": 1.7711542363084504e-05, "loss": 0.01, "step": 73730 }, { "grad_norm": 0.1407443732023239, "learning_rate": 1.7698919345892158e-05, "loss": 0.0108, "step": 73740 }, { "grad_norm": 0.15441669523715973, "learning_rate": 1.768629986109743e-05, "loss": 0.0114, "step": 73750 }, { "grad_norm": 0.1946934014558792, "learning_rate": 1.7673683910080357e-05, "loss": 0.0143, "step": 73760 }, { "grad_norm": 0.16275805234909058, "learning_rate": 1.7661071494220638e-05, "loss": 0.0086, "step": 73770 }, { "grad_norm": 0.14126358926296234, "learning_rate": 1.7648462614897498e-05, "loss": 0.01, "step": 73780 }, { "grad_norm": 0.16396522521972656, "learning_rate": 1.7635857273489863e-05, "loss": 0.0103, "step": 73790 }, { "grad_norm": 0.11933279037475586, "learning_rate": 1.7623255471376198e-05, "loss": 0.0104, "step": 73800 }, { "grad_norm": 0.14689739048480988, "learning_rate": 1.7610657209934646e-05, "loss": 0.0118, "step": 73810 }, { "grad_norm": 0.19882915914058685, "learning_rate": 1.7598062490542915e-05, "loss": 0.0145, "step": 73820 }, { "grad_norm": 0.12895070016384125, "learning_rate": 1.7585471314578377e-05, "loss": 0.0092, "step": 73830 }, { "grad_norm": 0.1268637329339981, "learning_rate": 1.7572883683417945e-05, "loss": 0.01, "step": 73840 }, { "grad_norm": 0.18780650198459625, "learning_rate": 1.7560299598438212e-05, "loss": 0.0105, "step": 73850 }, { "grad_norm": 0.17081479728221893, "learning_rate": 1.7547719061015334e-05, "loss": 0.0098, "step": 73860 }, { "grad_norm": 0.10768753290176392, "learning_rate": 1.75351420725251e-05, "loss": 0.0093, "step": 73870 }, { "grad_norm": 0.16197480261325836, "learning_rate": 1.7522568634342923e-05, "loss": 0.0139, "step": 73880 }, { "grad_norm": 0.23750756680965424, "learning_rate": 1.7509998747843825e-05, "loss": 0.0141, "step": 73890 }, { "grad_norm": 0.13992127776145935, "learning_rate": 1.7497432414402403e-05, "loss": 0.0116, "step": 73900 }, { "grad_norm": 0.15568602085113525, "learning_rate": 1.7484869635392903e-05, "loss": 0.0124, "step": 73910 }, { "grad_norm": 0.22038379311561584, "learning_rate": 1.747231041218919e-05, "loss": 0.0142, "step": 73920 }, { "grad_norm": 0.1136254221200943, "learning_rate": 1.7459754746164692e-05, "loss": 0.0109, "step": 73930 }, { "grad_norm": 0.15198983252048492, "learning_rate": 1.744720263869249e-05, "loss": 0.0117, "step": 73940 }, { "grad_norm": 0.21817027032375336, "learning_rate": 1.7434654091145263e-05, "loss": 0.0101, "step": 73950 }, { "grad_norm": 0.17607302963733673, "learning_rate": 1.742210910489532e-05, "loss": 0.0116, "step": 73960 }, { "grad_norm": 0.1834409087896347, "learning_rate": 1.7409567681314532e-05, "loss": 0.0102, "step": 73970 }, { "grad_norm": 0.15170158445835114, "learning_rate": 1.7397029821774434e-05, "loss": 0.0098, "step": 73980 }, { "grad_norm": 0.13334576785564423, "learning_rate": 1.7384495527646126e-05, "loss": 0.0146, "step": 73990 }, { "grad_norm": 0.1461310237646103, "learning_rate": 1.737196480030035e-05, "loss": 0.01, "step": 74000 }, { "grad_norm": 0.13197609782218933, "learning_rate": 1.7359437641107447e-05, "loss": 0.012, "step": 74010 }, { "grad_norm": 0.1806623488664627, "learning_rate": 1.7346914051437387e-05, "loss": 0.0126, "step": 74020 }, { "grad_norm": 0.18275347352027893, "learning_rate": 1.733439403265969e-05, "loss": 0.012, "step": 74030 }, { "grad_norm": 0.1521638035774231, "learning_rate": 1.732187758614357e-05, "loss": 0.0084, "step": 74040 }, { "grad_norm": 0.14589734375476837, "learning_rate": 1.730936471325777e-05, "loss": 0.0108, "step": 74050 }, { "grad_norm": 0.14519092440605164, "learning_rate": 1.7296855415370718e-05, "loss": 0.0089, "step": 74060 }, { "grad_norm": 0.2273293137550354, "learning_rate": 1.728434969385035e-05, "loss": 0.0163, "step": 74070 }, { "grad_norm": 0.11746219545602798, "learning_rate": 1.7271847550064346e-05, "loss": 0.0109, "step": 74080 }, { "grad_norm": 0.1544961929321289, "learning_rate": 1.725934898537987e-05, "loss": 0.0143, "step": 74090 }, { "grad_norm": 0.09508060663938522, "learning_rate": 1.7246854001163783e-05, "loss": 0.0114, "step": 74100 }, { "grad_norm": 0.15653382241725922, "learning_rate": 1.7234362598782476e-05, "loss": 0.0113, "step": 74110 }, { "grad_norm": 0.12097715586423874, "learning_rate": 1.722187477960202e-05, "loss": 0.0129, "step": 74120 }, { "grad_norm": 0.09444049000740051, "learning_rate": 1.7209390544988057e-05, "loss": 0.0113, "step": 74130 }, { "grad_norm": 0.12345530837774277, "learning_rate": 1.7196909896305856e-05, "loss": 0.0115, "step": 74140 }, { "grad_norm": 0.13794003427028656, "learning_rate": 1.718443283492026e-05, "loss": 0.0099, "step": 74150 }, { "grad_norm": 0.1725119799375534, "learning_rate": 1.7171959362195746e-05, "loss": 0.012, "step": 74160 }, { "grad_norm": 0.13169963657855988, "learning_rate": 1.715948947949642e-05, "loss": 0.0126, "step": 74170 }, { "grad_norm": 0.14895157516002655, "learning_rate": 1.714702318818593e-05, "loss": 0.0136, "step": 74180 }, { "grad_norm": 0.2050715982913971, "learning_rate": 1.7134560489627592e-05, "loss": 0.0114, "step": 74190 }, { "grad_norm": 0.11852686107158661, "learning_rate": 1.712210138518431e-05, "loss": 0.0125, "step": 74200 }, { "grad_norm": 0.15614457428455353, "learning_rate": 1.71096458762186e-05, "loss": 0.0118, "step": 74210 }, { "grad_norm": 0.21017619967460632, "learning_rate": 1.7097193964092552e-05, "loss": 0.0084, "step": 74220 }, { "grad_norm": 0.1465439796447754, "learning_rate": 1.708474565016792e-05, "loss": 0.0084, "step": 74230 }, { "grad_norm": 0.1885647177696228, "learning_rate": 1.7072300935806e-05, "loss": 0.0121, "step": 74240 }, { "grad_norm": 0.21576900780200958, "learning_rate": 1.705985982236776e-05, "loss": 0.0108, "step": 74250 }, { "grad_norm": 0.14451907575130463, "learning_rate": 1.7047422311213697e-05, "loss": 0.0105, "step": 74260 }, { "grad_norm": 0.14286740124225616, "learning_rate": 1.703498840370401e-05, "loss": 0.0104, "step": 74270 }, { "grad_norm": 0.12308509647846222, "learning_rate": 1.702255810119842e-05, "loss": 0.0081, "step": 74280 }, { "grad_norm": 0.17286230623722076, "learning_rate": 1.7010131405056305e-05, "loss": 0.0132, "step": 74290 }, { "grad_norm": 0.14609888195991516, "learning_rate": 1.6997708316636613e-05, "loss": 0.0113, "step": 74300 }, { "grad_norm": 0.13251091539859772, "learning_rate": 1.6985288837297932e-05, "loss": 0.0096, "step": 74310 }, { "grad_norm": 0.1387554258108139, "learning_rate": 1.6972872968398394e-05, "loss": 0.0104, "step": 74320 }, { "grad_norm": 0.11630468815565109, "learning_rate": 1.6960460711295845e-05, "loss": 0.0128, "step": 74330 }, { "grad_norm": 0.13066615164279938, "learning_rate": 1.6948052067347623e-05, "loss": 0.0113, "step": 74340 }, { "grad_norm": 0.14228878915309906, "learning_rate": 1.693564703791075e-05, "loss": 0.011, "step": 74350 }, { "grad_norm": 0.14244823157787323, "learning_rate": 1.6923245624341793e-05, "loss": 0.0105, "step": 74360 }, { "grad_norm": 0.1159539744257927, "learning_rate": 1.6910847827996962e-05, "loss": 0.0124, "step": 74370 }, { "grad_norm": 0.16677096486091614, "learning_rate": 1.689845365023206e-05, "loss": 0.0136, "step": 74380 }, { "grad_norm": 0.1176842749118805, "learning_rate": 1.6886063092402505e-05, "loss": 0.0112, "step": 74390 }, { "grad_norm": 0.14882627129554749, "learning_rate": 1.6873676155863317e-05, "loss": 0.0124, "step": 74400 }, { "grad_norm": 0.16535654664039612, "learning_rate": 1.6861292841969085e-05, "loss": 0.0127, "step": 74410 }, { "grad_norm": 0.14568662643432617, "learning_rate": 1.684891315207406e-05, "loss": 0.0137, "step": 74420 }, { "grad_norm": 0.18226352334022522, "learning_rate": 1.6836537087532034e-05, "loss": 0.0106, "step": 74430 }, { "grad_norm": 0.1331426054239273, "learning_rate": 1.6824164649696465e-05, "loss": 0.0115, "step": 74440 }, { "grad_norm": 0.17054764926433563, "learning_rate": 1.681179583992034e-05, "loss": 0.0148, "step": 74450 }, { "grad_norm": 0.16365709900856018, "learning_rate": 1.679943065955636e-05, "loss": 0.0103, "step": 74460 }, { "grad_norm": 0.1722044050693512, "learning_rate": 1.678706910995671e-05, "loss": 0.0102, "step": 74470 }, { "grad_norm": 0.15484417974948883, "learning_rate": 1.6774711192473262e-05, "loss": 0.0086, "step": 74480 }, { "grad_norm": 0.18284043669700623, "learning_rate": 1.676235690845743e-05, "loss": 0.0099, "step": 74490 }, { "grad_norm": 0.10911635309457779, "learning_rate": 1.6750006259260285e-05, "loss": 0.0106, "step": 74500 }, { "grad_norm": 0.15367667376995087, "learning_rate": 1.673765924623244e-05, "loss": 0.0108, "step": 74510 }, { "grad_norm": 0.1299264281988144, "learning_rate": 1.67253158707242e-05, "loss": 0.0113, "step": 74520 }, { "grad_norm": 0.15061213076114655, "learning_rate": 1.671297613408537e-05, "loss": 0.0108, "step": 74530 }, { "grad_norm": 0.13676437735557556, "learning_rate": 1.6700640037665444e-05, "loss": 0.0101, "step": 74540 }, { "grad_norm": 0.16793009638786316, "learning_rate": 1.668830758281344e-05, "loss": 0.0111, "step": 74550 }, { "grad_norm": 0.1514231115579605, "learning_rate": 1.6675978770878052e-05, "loss": 0.0116, "step": 74560 }, { "grad_norm": 0.11283528059720993, "learning_rate": 1.6663653603207506e-05, "loss": 0.0102, "step": 74570 }, { "grad_norm": 0.14678584039211273, "learning_rate": 1.6651332081149685e-05, "loss": 0.0096, "step": 74580 }, { "grad_norm": 0.0888381227850914, "learning_rate": 1.6639014206052046e-05, "loss": 0.0106, "step": 74590 }, { "grad_norm": 0.18589478731155396, "learning_rate": 1.6626699979261657e-05, "loss": 0.0112, "step": 74600 }, { "grad_norm": 0.14904433488845825, "learning_rate": 1.6614389402125197e-05, "loss": 0.0097, "step": 74610 }, { "grad_norm": 0.12613077461719513, "learning_rate": 1.66020824759889e-05, "loss": 0.0105, "step": 74620 }, { "grad_norm": 0.11794091016054153, "learning_rate": 1.6589779202198675e-05, "loss": 0.0105, "step": 74630 }, { "grad_norm": 0.28606027364730835, "learning_rate": 1.6577479582099924e-05, "loss": 0.0153, "step": 74640 }, { "grad_norm": 0.1786210685968399, "learning_rate": 1.6565183617037788e-05, "loss": 0.0137, "step": 74650 }, { "grad_norm": 0.16310729086399078, "learning_rate": 1.6552891308356894e-05, "loss": 0.0104, "step": 74660 }, { "grad_norm": 0.11312130838632584, "learning_rate": 1.6540602657401526e-05, "loss": 0.0125, "step": 74670 }, { "grad_norm": 0.17541824281215668, "learning_rate": 1.6528317665515537e-05, "loss": 0.0112, "step": 74680 }, { "grad_norm": 0.12818478047847748, "learning_rate": 1.6516036334042416e-05, "loss": 0.0115, "step": 74690 }, { "grad_norm": 0.1751355677843094, "learning_rate": 1.650375866432519e-05, "loss": 0.0114, "step": 74700 }, { "grad_norm": 0.1607017070055008, "learning_rate": 1.649148465770659e-05, "loss": 0.0104, "step": 74710 }, { "grad_norm": 0.3232595920562744, "learning_rate": 1.647921431552883e-05, "loss": 0.0121, "step": 74720 }, { "grad_norm": 0.15433290600776672, "learning_rate": 1.646694763913381e-05, "loss": 0.0114, "step": 74730 }, { "grad_norm": 0.11343295872211456, "learning_rate": 1.6454684629862966e-05, "loss": 0.0077, "step": 74740 }, { "grad_norm": 0.1793557107448578, "learning_rate": 1.6442425289057388e-05, "loss": 0.0102, "step": 74750 }, { "grad_norm": 0.12200698256492615, "learning_rate": 1.6430169618057718e-05, "loss": 0.0099, "step": 74760 }, { "grad_norm": 0.21149352192878723, "learning_rate": 1.641791761820422e-05, "loss": 0.0131, "step": 74770 }, { "grad_norm": 0.09094749391078949, "learning_rate": 1.640566929083676e-05, "loss": 0.01, "step": 74780 }, { "grad_norm": 0.09001312404870987, "learning_rate": 1.6393424637294823e-05, "loss": 0.01, "step": 74790 }, { "grad_norm": 0.12981344759464264, "learning_rate": 1.638118365891742e-05, "loss": 0.0133, "step": 74800 }, { "grad_norm": 0.13312003016471863, "learning_rate": 1.636894635704323e-05, "loss": 0.0123, "step": 74810 }, { "grad_norm": 0.12191469222307205, "learning_rate": 1.6356712733010517e-05, "loss": 0.012, "step": 74820 }, { "grad_norm": 0.1685560643672943, "learning_rate": 1.6344482788157096e-05, "loss": 0.0119, "step": 74830 }, { "grad_norm": 0.1725517362356186, "learning_rate": 1.6332256523820444e-05, "loss": 0.0101, "step": 74840 }, { "grad_norm": 0.15838423371315002, "learning_rate": 1.6320033941337597e-05, "loss": 0.01, "step": 74850 }, { "grad_norm": 0.12807123363018036, "learning_rate": 1.6307815042045216e-05, "loss": 0.0087, "step": 74860 }, { "grad_norm": 0.09410617500543594, "learning_rate": 1.6295599827279513e-05, "loss": 0.0159, "step": 74870 }, { "grad_norm": 0.14514502882957458, "learning_rate": 1.6283388298376356e-05, "loss": 0.0102, "step": 74880 }, { "grad_norm": 0.18499282002449036, "learning_rate": 1.6271180456671125e-05, "loss": 0.0116, "step": 74890 }, { "grad_norm": 0.14307653903961182, "learning_rate": 1.625897630349893e-05, "loss": 0.011, "step": 74900 }, { "grad_norm": 0.36543819308280945, "learning_rate": 1.6246775840194335e-05, "loss": 0.015, "step": 74910 }, { "grad_norm": 0.1718888133764267, "learning_rate": 1.6234579068091604e-05, "loss": 0.0101, "step": 74920 }, { "grad_norm": 0.11572884768247604, "learning_rate": 1.6222385988524524e-05, "loss": 0.0105, "step": 74930 }, { "grad_norm": 0.1454564779996872, "learning_rate": 1.6210196602826544e-05, "loss": 0.0136, "step": 74940 }, { "grad_norm": 0.26104286313056946, "learning_rate": 1.6198010912330646e-05, "loss": 0.0124, "step": 74950 }, { "grad_norm": 0.14266462624073029, "learning_rate": 1.6185828918369456e-05, "loss": 0.0105, "step": 74960 }, { "grad_norm": 0.13026726245880127, "learning_rate": 1.617365062227517e-05, "loss": 0.0101, "step": 74970 }, { "grad_norm": 0.14667022228240967, "learning_rate": 1.616147602537962e-05, "loss": 0.012, "step": 74980 }, { "grad_norm": 0.39487630128860474, "learning_rate": 1.614930512901415e-05, "loss": 0.0095, "step": 74990 }, { "grad_norm": 0.11558022350072861, "learning_rate": 1.6137137934509804e-05, "loss": 0.0117, "step": 75000 }, { "grad_norm": 0.17341956496238708, "learning_rate": 1.6124974443197123e-05, "loss": 0.0116, "step": 75010 }, { "grad_norm": 0.12767361104488373, "learning_rate": 1.6112814656406306e-05, "loss": 0.0104, "step": 75020 }, { "grad_norm": 0.16163945198059082, "learning_rate": 1.6100658575467125e-05, "loss": 0.0148, "step": 75030 }, { "grad_norm": 0.15159952640533447, "learning_rate": 1.6088506201708974e-05, "loss": 0.0101, "step": 75040 }, { "grad_norm": 0.15833742916584015, "learning_rate": 1.6076357536460785e-05, "loss": 0.0093, "step": 75050 }, { "grad_norm": 0.1657259613275528, "learning_rate": 1.606421258105113e-05, "loss": 0.0121, "step": 75060 }, { "grad_norm": 0.14827565848827362, "learning_rate": 1.6052071336808183e-05, "loss": 0.0131, "step": 75070 }, { "grad_norm": 0.13438190519809723, "learning_rate": 1.6039933805059658e-05, "loss": 0.01, "step": 75080 }, { "grad_norm": 0.26022055745124817, "learning_rate": 1.6027799987132915e-05, "loss": 0.0105, "step": 75090 }, { "grad_norm": 0.12122079730033875, "learning_rate": 1.6015669884354888e-05, "loss": 0.0098, "step": 75100 }, { "grad_norm": 0.13758201897144318, "learning_rate": 1.6003543498052125e-05, "loss": 0.01, "step": 75110 }, { "grad_norm": 0.14507777988910675, "learning_rate": 1.5991420829550715e-05, "loss": 0.01, "step": 75120 }, { "grad_norm": 0.14747664332389832, "learning_rate": 1.5979301880176405e-05, "loss": 0.011, "step": 75130 }, { "grad_norm": 0.153593972325325, "learning_rate": 1.5967186651254485e-05, "loss": 0.0118, "step": 75140 }, { "grad_norm": 0.16102713346481323, "learning_rate": 1.5955075144109853e-05, "loss": 0.0098, "step": 75150 }, { "grad_norm": 0.12659546732902527, "learning_rate": 1.594296736006703e-05, "loss": 0.01, "step": 75160 }, { "grad_norm": 0.13871246576309204, "learning_rate": 1.59308633004501e-05, "loss": 0.0102, "step": 75170 }, { "grad_norm": 0.13594935834407806, "learning_rate": 1.5918762966582724e-05, "loss": 0.0113, "step": 75180 }, { "grad_norm": 0.13333025574684143, "learning_rate": 1.5906666359788203e-05, "loss": 0.0073, "step": 75190 }, { "grad_norm": 0.14135326445102692, "learning_rate": 1.589457348138938e-05, "loss": 0.0093, "step": 75200 }, { "grad_norm": 0.1463976353406906, "learning_rate": 1.588248433270872e-05, "loss": 0.0093, "step": 75210 }, { "grad_norm": 0.17729930579662323, "learning_rate": 1.587039891506828e-05, "loss": 0.0123, "step": 75220 }, { "grad_norm": 0.13292859494686127, "learning_rate": 1.5858317229789715e-05, "loss": 0.009, "step": 75230 }, { "grad_norm": 0.20708045363426208, "learning_rate": 1.584623927819423e-05, "loss": 0.0115, "step": 75240 }, { "grad_norm": 0.13332900404930115, "learning_rate": 1.583416506160269e-05, "loss": 0.0088, "step": 75250 }, { "grad_norm": 0.13006730377674103, "learning_rate": 1.5822094581335472e-05, "loss": 0.0093, "step": 75260 }, { "grad_norm": 0.17806708812713623, "learning_rate": 1.5810027838712604e-05, "loss": 0.0119, "step": 75270 }, { "grad_norm": 0.1639556884765625, "learning_rate": 1.5797964835053687e-05, "loss": 0.0112, "step": 75280 }, { "grad_norm": 0.13930781185626984, "learning_rate": 1.5785905571677933e-05, "loss": 0.0108, "step": 75290 }, { "grad_norm": 0.16902880370616913, "learning_rate": 1.5773850049904086e-05, "loss": 0.0112, "step": 75300 }, { "grad_norm": 0.16774848103523254, "learning_rate": 1.5761798271050543e-05, "loss": 0.011, "step": 75310 }, { "grad_norm": 0.23887114226818085, "learning_rate": 1.5749750236435277e-05, "loss": 0.0118, "step": 75320 }, { "grad_norm": 0.18871517479419708, "learning_rate": 1.5737705947375815e-05, "loss": 0.0141, "step": 75330 }, { "grad_norm": 0.11193576455116272, "learning_rate": 1.5725665405189322e-05, "loss": 0.0107, "step": 75340 }, { "grad_norm": 0.14132839441299438, "learning_rate": 1.571362861119253e-05, "loss": 0.0113, "step": 75350 }, { "grad_norm": 0.1626904010772705, "learning_rate": 1.570159556670177e-05, "loss": 0.0124, "step": 75360 }, { "grad_norm": 0.13265468180179596, "learning_rate": 1.5689566273032946e-05, "loss": 0.0122, "step": 75370 }, { "grad_norm": 0.14221104979515076, "learning_rate": 1.5677540731501583e-05, "loss": 0.0123, "step": 75380 }, { "grad_norm": 0.17081831395626068, "learning_rate": 1.5665518943422746e-05, "loss": 0.0125, "step": 75390 }, { "grad_norm": 0.14209294319152832, "learning_rate": 1.5653500910111134e-05, "loss": 0.0092, "step": 75400 }, { "grad_norm": 0.12094906717538834, "learning_rate": 1.5641486632881023e-05, "loss": 0.0106, "step": 75410 }, { "grad_norm": 0.15266276895999908, "learning_rate": 1.56294761130463e-05, "loss": 0.011, "step": 75420 }, { "grad_norm": 0.1165120005607605, "learning_rate": 1.5617469351920373e-05, "loss": 0.0117, "step": 75430 }, { "grad_norm": 0.11178810894489288, "learning_rate": 1.5605466350816315e-05, "loss": 0.0117, "step": 75440 }, { "grad_norm": 0.19920586049556732, "learning_rate": 1.559346711104674e-05, "loss": 0.0091, "step": 75450 }, { "grad_norm": 0.12921901047229767, "learning_rate": 1.5581471633923883e-05, "loss": 0.0099, "step": 75460 }, { "grad_norm": 0.09528908133506775, "learning_rate": 1.5569479920759505e-05, "loss": 0.0105, "step": 75470 }, { "grad_norm": 0.20667977631092072, "learning_rate": 1.5557491972865072e-05, "loss": 0.0129, "step": 75480 }, { "grad_norm": 0.18669885396957397, "learning_rate": 1.554550779155152e-05, "loss": 0.0137, "step": 75490 }, { "grad_norm": 0.16867578029632568, "learning_rate": 1.5533527378129453e-05, "loss": 0.0098, "step": 75500 }, { "grad_norm": 0.22272025048732758, "learning_rate": 1.552155073390899e-05, "loss": 0.013, "step": 75510 }, { "grad_norm": 0.1252385824918747, "learning_rate": 1.5509577860199903e-05, "loss": 0.0097, "step": 75520 }, { "grad_norm": 0.09382633119821548, "learning_rate": 1.5497608758311525e-05, "loss": 0.0092, "step": 75530 }, { "grad_norm": 0.13709679245948792, "learning_rate": 1.5485643429552797e-05, "loss": 0.0087, "step": 75540 }, { "grad_norm": 0.18343029916286469, "learning_rate": 1.5473681875232197e-05, "loss": 0.0115, "step": 75550 }, { "grad_norm": 0.15361803770065308, "learning_rate": 1.5461724096657837e-05, "loss": 0.0119, "step": 75560 }, { "grad_norm": 0.14477133750915527, "learning_rate": 1.5449770095137416e-05, "loss": 0.0089, "step": 75570 }, { "grad_norm": 0.13331300020217896, "learning_rate": 1.5437819871978178e-05, "loss": 0.0084, "step": 75580 }, { "grad_norm": 0.1480286717414856, "learning_rate": 1.5425873428486992e-05, "loss": 0.0091, "step": 75590 }, { "grad_norm": 0.16562142968177795, "learning_rate": 1.5413930765970312e-05, "loss": 0.0112, "step": 75600 }, { "grad_norm": 0.17937614023685455, "learning_rate": 1.5401991885734175e-05, "loss": 0.0114, "step": 75610 }, { "grad_norm": 0.16974155604839325, "learning_rate": 1.5390056789084174e-05, "loss": 0.0106, "step": 75620 }, { "grad_norm": 0.14314477145671844, "learning_rate": 1.5378125477325545e-05, "loss": 0.011, "step": 75630 }, { "grad_norm": 0.10999521613121033, "learning_rate": 1.5366197951763046e-05, "loss": 0.0123, "step": 75640 }, { "grad_norm": 0.25148606300354004, "learning_rate": 1.5354274213701075e-05, "loss": 0.0096, "step": 75650 }, { "grad_norm": 0.1983870416879654, "learning_rate": 1.5342354264443566e-05, "loss": 0.0118, "step": 75660 }, { "grad_norm": 0.14063236117362976, "learning_rate": 1.533043810529411e-05, "loss": 0.0115, "step": 75670 }, { "grad_norm": 0.1101582944393158, "learning_rate": 1.5318525737555807e-05, "loss": 0.011, "step": 75680 }, { "grad_norm": 0.153220072388649, "learning_rate": 1.53066171625314e-05, "loss": 0.0131, "step": 75690 }, { "grad_norm": 0.13456867635250092, "learning_rate": 1.529471238152317e-05, "loss": 0.0116, "step": 75700 }, { "grad_norm": 0.11084534227848053, "learning_rate": 1.5282811395833023e-05, "loss": 0.0096, "step": 75710 }, { "grad_norm": 0.09440802782773972, "learning_rate": 1.5270914206762397e-05, "loss": 0.0093, "step": 75720 }, { "grad_norm": 0.10799554735422134, "learning_rate": 1.525902081561241e-05, "loss": 0.0095, "step": 75730 }, { "grad_norm": 0.1074100211262703, "learning_rate": 1.5247131223683664e-05, "loss": 0.0098, "step": 75740 }, { "grad_norm": 0.16321425139904022, "learning_rate": 1.5235245432276408e-05, "loss": 0.0095, "step": 75750 }, { "grad_norm": 0.14955034852027893, "learning_rate": 1.5223363442690424e-05, "loss": 0.0101, "step": 75760 }, { "grad_norm": 0.1923965960741043, "learning_rate": 1.5211485256225133e-05, "loss": 0.0108, "step": 75770 }, { "grad_norm": 0.17786410450935364, "learning_rate": 1.519961087417951e-05, "loss": 0.0103, "step": 75780 }, { "grad_norm": 0.112128347158432, "learning_rate": 1.5187740297852137e-05, "loss": 0.0096, "step": 75790 }, { "grad_norm": 0.14998990297317505, "learning_rate": 1.517587352854113e-05, "loss": 0.0117, "step": 75800 }, { "grad_norm": 0.16987605392932892, "learning_rate": 1.516401056754423e-05, "loss": 0.0114, "step": 75810 }, { "grad_norm": 0.09971641004085541, "learning_rate": 1.5152151416158783e-05, "loss": 0.0085, "step": 75820 }, { "grad_norm": 0.09793375432491302, "learning_rate": 1.5140296075681642e-05, "loss": 0.009, "step": 75830 }, { "grad_norm": 0.15305644273757935, "learning_rate": 1.5128444547409331e-05, "loss": 0.0116, "step": 75840 }, { "grad_norm": 0.2277720421552658, "learning_rate": 1.5116596832637864e-05, "loss": 0.0123, "step": 75850 }, { "grad_norm": 0.14252705872058868, "learning_rate": 1.5104752932662952e-05, "loss": 0.0098, "step": 75860 }, { "grad_norm": 0.12286607921123505, "learning_rate": 1.5092912848779778e-05, "loss": 0.0081, "step": 75870 }, { "grad_norm": 0.13556723296642303, "learning_rate": 1.5081076582283194e-05, "loss": 0.0099, "step": 75880 }, { "grad_norm": 0.10998541861772537, "learning_rate": 1.5069244134467553e-05, "loss": 0.009, "step": 75890 }, { "grad_norm": 0.15632230043411255, "learning_rate": 1.5057415506626876e-05, "loss": 0.0097, "step": 75900 }, { "grad_norm": 0.14562451839447021, "learning_rate": 1.5045590700054668e-05, "loss": 0.0105, "step": 75910 }, { "grad_norm": 0.08775673806667328, "learning_rate": 1.5033769716044144e-05, "loss": 0.0098, "step": 75920 }, { "grad_norm": 0.18604935705661774, "learning_rate": 1.5021952555887975e-05, "loss": 0.0109, "step": 75930 }, { "grad_norm": 0.08585941791534424, "learning_rate": 1.50101392208785e-05, "loss": 0.011, "step": 75940 }, { "grad_norm": 0.10300493985414505, "learning_rate": 1.4998329712307579e-05, "loss": 0.0122, "step": 75950 }, { "grad_norm": 0.18252073228359222, "learning_rate": 1.4986524031466708e-05, "loss": 0.0119, "step": 75960 }, { "grad_norm": 0.1446995884180069, "learning_rate": 1.4974722179646916e-05, "loss": 0.0086, "step": 75970 }, { "grad_norm": 0.09517841786146164, "learning_rate": 1.4962924158138847e-05, "loss": 0.0085, "step": 75980 }, { "grad_norm": 0.1665991097688675, "learning_rate": 1.4951129968232714e-05, "loss": 0.0125, "step": 75990 }, { "grad_norm": 0.16188104450702667, "learning_rate": 1.4939339611218329e-05, "loss": 0.0136, "step": 76000 }, { "grad_norm": 0.24411311745643616, "learning_rate": 1.4927553088385032e-05, "loss": 0.0115, "step": 76010 }, { "grad_norm": 0.13316352665424347, "learning_rate": 1.4915770401021801e-05, "loss": 0.0126, "step": 76020 }, { "grad_norm": 0.12945763766765594, "learning_rate": 1.4903991550417185e-05, "loss": 0.0092, "step": 76030 }, { "grad_norm": 0.16581197082996368, "learning_rate": 1.4892216537859277e-05, "loss": 0.0114, "step": 76040 }, { "grad_norm": 0.13314887881278992, "learning_rate": 1.4880445364635775e-05, "loss": 0.0092, "step": 76050 }, { "grad_norm": 0.14446696639060974, "learning_rate": 1.4868678032033973e-05, "loss": 0.0098, "step": 76060 }, { "grad_norm": 0.11525986343622208, "learning_rate": 1.4856914541340739e-05, "loss": 0.0098, "step": 76070 }, { "grad_norm": 0.10741753876209259, "learning_rate": 1.4845154893842472e-05, "loss": 0.0088, "step": 76080 }, { "grad_norm": 0.17237935960292816, "learning_rate": 1.4833399090825234e-05, "loss": 0.0086, "step": 76090 }, { "grad_norm": 0.21517769992351532, "learning_rate": 1.4821647133574562e-05, "loss": 0.0131, "step": 76100 }, { "grad_norm": 0.10734061151742935, "learning_rate": 1.4809899023375706e-05, "loss": 0.0099, "step": 76110 }, { "grad_norm": 0.1670222282409668, "learning_rate": 1.4798154761513366e-05, "loss": 0.0098, "step": 76120 }, { "grad_norm": 0.12276171147823334, "learning_rate": 1.4786414349271916e-05, "loss": 0.0071, "step": 76130 }, { "grad_norm": 0.13883380591869354, "learning_rate": 1.4774677787935226e-05, "loss": 0.0092, "step": 76140 }, { "grad_norm": 0.13068243861198425, "learning_rate": 1.4762945078786838e-05, "loss": 0.0117, "step": 76150 }, { "grad_norm": 0.14121781289577484, "learning_rate": 1.4751216223109777e-05, "loss": 0.012, "step": 76160 }, { "grad_norm": 0.18386204540729523, "learning_rate": 1.4739491222186714e-05, "loss": 0.0124, "step": 76170 }, { "grad_norm": 0.21942444145679474, "learning_rate": 1.4727770077299879e-05, "loss": 0.0124, "step": 76180 }, { "grad_norm": 0.18362358212471008, "learning_rate": 1.4716052789731094e-05, "loss": 0.0141, "step": 76190 }, { "grad_norm": 0.2137957662343979, "learning_rate": 1.470433936076171e-05, "loss": 0.0112, "step": 76200 }, { "grad_norm": 0.1962273269891739, "learning_rate": 1.4692629791672719e-05, "loss": 0.0119, "step": 76210 }, { "grad_norm": 0.14301511645317078, "learning_rate": 1.468092408374464e-05, "loss": 0.0101, "step": 76220 }, { "grad_norm": 0.1359729766845703, "learning_rate": 1.4669222238257602e-05, "loss": 0.0118, "step": 76230 }, { "grad_norm": 0.13591046631336212, "learning_rate": 1.4657524256491301e-05, "loss": 0.0099, "step": 76240 }, { "grad_norm": 0.1422303467988968, "learning_rate": 1.464583013972502e-05, "loss": 0.0121, "step": 76250 }, { "grad_norm": 0.14727771282196045, "learning_rate": 1.4634139889237591e-05, "loss": 0.0123, "step": 76260 }, { "grad_norm": 0.14604245126247406, "learning_rate": 1.4622453506307448e-05, "loss": 0.0087, "step": 76270 }, { "grad_norm": 0.15425778925418854, "learning_rate": 1.461077099221262e-05, "loss": 0.0097, "step": 76280 }, { "grad_norm": 0.15235283970832825, "learning_rate": 1.4599092348230652e-05, "loss": 0.0096, "step": 76290 }, { "grad_norm": 0.12638506293296814, "learning_rate": 1.4587417575638724e-05, "loss": 0.0088, "step": 76300 }, { "grad_norm": 0.23345303535461426, "learning_rate": 1.4575746675713563e-05, "loss": 0.0124, "step": 76310 }, { "grad_norm": 0.09829694032669067, "learning_rate": 1.4564079649731504e-05, "loss": 0.0114, "step": 76320 }, { "grad_norm": 0.11787879467010498, "learning_rate": 1.4552416498968402e-05, "loss": 0.0106, "step": 76330 }, { "grad_norm": 0.15199196338653564, "learning_rate": 1.4540757224699747e-05, "loss": 0.0127, "step": 76340 }, { "grad_norm": 0.11563418805599213, "learning_rate": 1.452910182820056e-05, "loss": 0.0098, "step": 76350 }, { "grad_norm": 0.1597272753715515, "learning_rate": 1.4517450310745468e-05, "loss": 0.0094, "step": 76360 }, { "grad_norm": 0.15744730830192566, "learning_rate": 1.450580267360866e-05, "loss": 0.0103, "step": 76370 }, { "grad_norm": 0.1646895557641983, "learning_rate": 1.4494158918063922e-05, "loss": 0.0114, "step": 76380 }, { "grad_norm": 0.13964436948299408, "learning_rate": 1.4482519045384568e-05, "loss": 0.009, "step": 76390 }, { "grad_norm": 0.23831859230995178, "learning_rate": 1.4470883056843548e-05, "loss": 0.0113, "step": 76400 }, { "grad_norm": 0.12670768797397614, "learning_rate": 1.4459250953713322e-05, "loss": 0.0093, "step": 76410 }, { "grad_norm": 0.19813892245292664, "learning_rate": 1.444762273726598e-05, "loss": 0.0109, "step": 76420 }, { "grad_norm": 0.10286182165145874, "learning_rate": 1.443599840877316e-05, "loss": 0.0078, "step": 76430 }, { "grad_norm": 0.10283380001783371, "learning_rate": 1.4424377969506103e-05, "loss": 0.0111, "step": 76440 }, { "grad_norm": 0.19977375864982605, "learning_rate": 1.4412761420735565e-05, "loss": 0.0139, "step": 76450 }, { "grad_norm": 0.1686490774154663, "learning_rate": 1.4401148763731953e-05, "loss": 0.0103, "step": 76460 }, { "grad_norm": 0.16369856894016266, "learning_rate": 1.4389539999765173e-05, "loss": 0.0101, "step": 76470 }, { "grad_norm": 0.15712019801139832, "learning_rate": 1.4377935130104758e-05, "loss": 0.0081, "step": 76480 }, { "grad_norm": 0.14294207096099854, "learning_rate": 1.4366334156019801e-05, "loss": 0.0121, "step": 76490 }, { "grad_norm": 0.17861436307430267, "learning_rate": 1.435473707877898e-05, "loss": 0.0109, "step": 76500 }, { "grad_norm": 0.1286231279373169, "learning_rate": 1.4343143899650508e-05, "loss": 0.0104, "step": 76510 }, { "grad_norm": 0.12635940313339233, "learning_rate": 1.4331554619902204e-05, "loss": 0.0089, "step": 76520 }, { "grad_norm": 0.1583511233329773, "learning_rate": 1.4319969240801484e-05, "loss": 0.0111, "step": 76530 }, { "grad_norm": 0.13168099522590637, "learning_rate": 1.430838776361526e-05, "loss": 0.0097, "step": 76540 }, { "grad_norm": 0.15681743621826172, "learning_rate": 1.4296810189610093e-05, "loss": 0.0101, "step": 76550 }, { "grad_norm": 0.15973533689975739, "learning_rate": 1.4285236520052087e-05, "loss": 0.0096, "step": 76560 }, { "grad_norm": 0.10375996679067612, "learning_rate": 1.4273666756206933e-05, "loss": 0.0095, "step": 76570 }, { "grad_norm": 0.1958736777305603, "learning_rate": 1.4262100899339848e-05, "loss": 0.0104, "step": 76580 }, { "grad_norm": 0.12267427146434784, "learning_rate": 1.4250538950715696e-05, "loss": 0.009, "step": 76590 }, { "grad_norm": 0.13274334371089935, "learning_rate": 1.4238980911598842e-05, "loss": 0.0092, "step": 76600 }, { "grad_norm": 0.1599535346031189, "learning_rate": 1.4227426783253267e-05, "loss": 0.009, "step": 76610 }, { "grad_norm": 0.1649726927280426, "learning_rate": 1.4215876566942521e-05, "loss": 0.0122, "step": 76620 }, { "grad_norm": 0.11763271689414978, "learning_rate": 1.4204330263929721e-05, "loss": 0.0094, "step": 76630 }, { "grad_norm": 0.13494586944580078, "learning_rate": 1.4192787875477537e-05, "loss": 0.009, "step": 76640 }, { "grad_norm": 0.10035501420497894, "learning_rate": 1.4181249402848246e-05, "loss": 0.0104, "step": 76650 }, { "grad_norm": 0.119723841547966, "learning_rate": 1.4169714847303656e-05, "loss": 0.0119, "step": 76660 }, { "grad_norm": 0.19468249380588531, "learning_rate": 1.4158184210105197e-05, "loss": 0.0142, "step": 76670 }, { "grad_norm": 0.11921036243438721, "learning_rate": 1.4146657492513787e-05, "loss": 0.0107, "step": 76680 }, { "grad_norm": 0.12818220257759094, "learning_rate": 1.4135134695790047e-05, "loss": 0.0111, "step": 76690 }, { "grad_norm": 0.10169262439012527, "learning_rate": 1.4123615821194037e-05, "loss": 0.0094, "step": 76700 }, { "grad_norm": 0.15009154379367828, "learning_rate": 1.4112100869985473e-05, "loss": 0.0108, "step": 76710 }, { "grad_norm": 0.19131678342819214, "learning_rate": 1.4100589843423589e-05, "loss": 0.0102, "step": 76720 }, { "grad_norm": 0.14646664261817932, "learning_rate": 1.408908274276723e-05, "loss": 0.0103, "step": 76730 }, { "grad_norm": 0.1907867193222046, "learning_rate": 1.4077579569274786e-05, "loss": 0.0112, "step": 76740 }, { "grad_norm": 0.12986066937446594, "learning_rate": 1.4066080324204245e-05, "loss": 0.0092, "step": 76750 }, { "grad_norm": 0.19045701622962952, "learning_rate": 1.405458500881312e-05, "loss": 0.0123, "step": 76760 }, { "grad_norm": 0.16456681489944458, "learning_rate": 1.4043093624358538e-05, "loss": 0.0084, "step": 76770 }, { "grad_norm": 0.14593034982681274, "learning_rate": 1.4031606172097189e-05, "loss": 0.0091, "step": 76780 }, { "grad_norm": 0.1258864402770996, "learning_rate": 1.40201226532853e-05, "loss": 0.0082, "step": 76790 }, { "grad_norm": 0.134210467338562, "learning_rate": 1.4008643069178701e-05, "loss": 0.0096, "step": 76800 }, { "grad_norm": 0.18593424558639526, "learning_rate": 1.399716742103278e-05, "loss": 0.0094, "step": 76810 }, { "grad_norm": 0.18152013421058655, "learning_rate": 1.398569571010252e-05, "loss": 0.0132, "step": 76820 }, { "grad_norm": 0.1312263011932373, "learning_rate": 1.3974227937642414e-05, "loss": 0.0108, "step": 76830 }, { "grad_norm": 0.16045361757278442, "learning_rate": 1.3962764104906595e-05, "loss": 0.0093, "step": 76840 }, { "grad_norm": 0.13563142716884613, "learning_rate": 1.3951304213148696e-05, "loss": 0.0123, "step": 76850 }, { "grad_norm": 0.1887037605047226, "learning_rate": 1.3939848263621979e-05, "loss": 0.0102, "step": 76860 }, { "grad_norm": 0.13480155169963837, "learning_rate": 1.3928396257579219e-05, "loss": 0.0117, "step": 76870 }, { "grad_norm": 0.1430443674325943, "learning_rate": 1.3916948196272833e-05, "loss": 0.0093, "step": 76880 }, { "grad_norm": 0.15217801928520203, "learning_rate": 1.3905504080954734e-05, "loss": 0.0099, "step": 76890 }, { "grad_norm": 0.17881731688976288, "learning_rate": 1.3894063912876454e-05, "loss": 0.0101, "step": 76900 }, { "grad_norm": 0.1340106576681137, "learning_rate": 1.3882627693289046e-05, "loss": 0.0118, "step": 76910 }, { "grad_norm": 0.1619601994752884, "learning_rate": 1.3871195423443183e-05, "loss": 0.0109, "step": 76920 }, { "grad_norm": 0.13932406902313232, "learning_rate": 1.3859767104589044e-05, "loss": 0.0103, "step": 76930 }, { "grad_norm": 0.22807106375694275, "learning_rate": 1.3848342737976467e-05, "loss": 0.0112, "step": 76940 }, { "grad_norm": 0.18201765418052673, "learning_rate": 1.3836922324854757e-05, "loss": 0.0113, "step": 76950 }, { "grad_norm": 0.1618746817111969, "learning_rate": 1.382550586647287e-05, "loss": 0.0103, "step": 76960 }, { "grad_norm": 0.1564808040857315, "learning_rate": 1.3814093364079257e-05, "loss": 0.0111, "step": 76970 }, { "grad_norm": 0.1498016119003296, "learning_rate": 1.3802684818921991e-05, "loss": 0.0125, "step": 76980 }, { "grad_norm": 0.18735167384147644, "learning_rate": 1.3791280232248687e-05, "loss": 0.0162, "step": 76990 }, { "grad_norm": 0.1150873526930809, "learning_rate": 1.3779879605306555e-05, "loss": 0.0086, "step": 77000 }, { "grad_norm": 0.18456153571605682, "learning_rate": 1.376848293934232e-05, "loss": 0.0126, "step": 77010 }, { "grad_norm": 0.10621514916419983, "learning_rate": 1.3757090235602316e-05, "loss": 0.0092, "step": 77020 }, { "grad_norm": 0.17601574957370758, "learning_rate": 1.3745701495332447e-05, "loss": 0.009, "step": 77030 }, { "grad_norm": 0.16071216762065887, "learning_rate": 1.3734316719778145e-05, "loss": 0.0108, "step": 77040 }, { "grad_norm": 0.20346008241176605, "learning_rate": 1.3722935910184454e-05, "loss": 0.0105, "step": 77050 }, { "grad_norm": 0.13996244966983795, "learning_rate": 1.3711559067795921e-05, "loss": 0.0101, "step": 77060 }, { "grad_norm": 0.16295568645000458, "learning_rate": 1.3700186193856756e-05, "loss": 0.0115, "step": 77070 }, { "grad_norm": 0.18308278918266296, "learning_rate": 1.368881728961064e-05, "loss": 0.0093, "step": 77080 }, { "grad_norm": 0.14653615653514862, "learning_rate": 1.3677452356300885e-05, "loss": 0.0133, "step": 77090 }, { "grad_norm": 0.13883142173290253, "learning_rate": 1.3666091395170321e-05, "loss": 0.0102, "step": 77100 }, { "grad_norm": 0.16015563905239105, "learning_rate": 1.3654734407461384e-05, "loss": 0.0109, "step": 77110 }, { "grad_norm": 0.14234286546707153, "learning_rate": 1.3643381394416022e-05, "loss": 0.011, "step": 77120 }, { "grad_norm": 0.12709490954875946, "learning_rate": 1.3632032357275838e-05, "loss": 0.0115, "step": 77130 }, { "grad_norm": 0.16440019011497498, "learning_rate": 1.3620687297281898e-05, "loss": 0.0119, "step": 77140 }, { "grad_norm": 0.18384507298469543, "learning_rate": 1.3609346215674917e-05, "loss": 0.0129, "step": 77150 }, { "grad_norm": 0.12502379715442657, "learning_rate": 1.3598009113695109e-05, "loss": 0.0096, "step": 77160 }, { "grad_norm": 0.15360915660858154, "learning_rate": 1.358667599258231e-05, "loss": 0.0078, "step": 77170 }, { "grad_norm": 0.15552490949630737, "learning_rate": 1.357534685357586e-05, "loss": 0.0139, "step": 77180 }, { "grad_norm": 0.11571948975324631, "learning_rate": 1.3564021697914715e-05, "loss": 0.0113, "step": 77190 }, { "grad_norm": 0.09853161126375198, "learning_rate": 1.3552700526837381e-05, "loss": 0.0102, "step": 77200 }, { "grad_norm": 0.12524795532226562, "learning_rate": 1.3541383341581925e-05, "loss": 0.0121, "step": 77210 }, { "grad_norm": 0.11674203723669052, "learning_rate": 1.3530070143385965e-05, "loss": 0.0094, "step": 77220 }, { "grad_norm": 0.1789042055606842, "learning_rate": 1.3518760933486702e-05, "loss": 0.0117, "step": 77230 }, { "grad_norm": 0.13245779275894165, "learning_rate": 1.3507455713120908e-05, "loss": 0.0114, "step": 77240 }, { "grad_norm": 0.12160411477088928, "learning_rate": 1.3496154483524876e-05, "loss": 0.0088, "step": 77250 }, { "grad_norm": 0.10174673795700073, "learning_rate": 1.3484857245934506e-05, "loss": 0.0096, "step": 77260 }, { "grad_norm": 0.18693870306015015, "learning_rate": 1.3473564001585248e-05, "loss": 0.0098, "step": 77270 }, { "grad_norm": 0.1858278512954712, "learning_rate": 1.3462274751712129e-05, "loss": 0.0111, "step": 77280 }, { "grad_norm": 0.13435284793376923, "learning_rate": 1.3450989497549698e-05, "loss": 0.0101, "step": 77290 }, { "grad_norm": 0.17444220185279846, "learning_rate": 1.3439708240332115e-05, "loss": 0.0087, "step": 77300 }, { "grad_norm": 0.13623741269111633, "learning_rate": 1.3428430981293043e-05, "loss": 0.0103, "step": 77310 }, { "grad_norm": 0.13256417214870453, "learning_rate": 1.3417157721665808e-05, "loss": 0.0107, "step": 77320 }, { "grad_norm": 0.17154042422771454, "learning_rate": 1.3405888462683186e-05, "loss": 0.0097, "step": 77330 }, { "grad_norm": 0.1347237080335617, "learning_rate": 1.3394623205577595e-05, "loss": 0.0092, "step": 77340 }, { "grad_norm": 0.13946795463562012, "learning_rate": 1.3383361951580963e-05, "loss": 0.0118, "step": 77350 }, { "grad_norm": 0.10330665111541748, "learning_rate": 1.337210470192483e-05, "loss": 0.0081, "step": 77360 }, { "grad_norm": 0.1138693168759346, "learning_rate": 1.3360851457840245e-05, "loss": 0.0093, "step": 77370 }, { "grad_norm": 0.146852508187294, "learning_rate": 1.3349602220557856e-05, "loss": 0.0101, "step": 77380 }, { "grad_norm": 0.20620301365852356, "learning_rate": 1.3338356991307865e-05, "loss": 0.0103, "step": 77390 }, { "grad_norm": 0.14154000580310822, "learning_rate": 1.3327115771320048e-05, "loss": 0.0093, "step": 77400 }, { "grad_norm": 0.10126787424087524, "learning_rate": 1.3315878561823697e-05, "loss": 0.0099, "step": 77410 }, { "grad_norm": 0.15147089958190918, "learning_rate": 1.330464536404773e-05, "loss": 0.0093, "step": 77420 }, { "grad_norm": 0.23645859956741333, "learning_rate": 1.3293416179220563e-05, "loss": 0.0096, "step": 77430 }, { "grad_norm": 0.147840678691864, "learning_rate": 1.328219100857021e-05, "loss": 0.0101, "step": 77440 }, { "grad_norm": 0.15433622896671295, "learning_rate": 1.327096985332425e-05, "loss": 0.0103, "step": 77450 }, { "grad_norm": 0.17309249937534332, "learning_rate": 1.3259752714709816e-05, "loss": 0.0122, "step": 77460 }, { "grad_norm": 0.1540309637784958, "learning_rate": 1.3248539593953569e-05, "loss": 0.0108, "step": 77470 }, { "grad_norm": 0.16212648153305054, "learning_rate": 1.323733049228178e-05, "loss": 0.0121, "step": 77480 }, { "grad_norm": 0.1734684556722641, "learning_rate": 1.3226125410920276e-05, "loss": 0.0113, "step": 77490 }, { "grad_norm": 0.15785230696201324, "learning_rate": 1.3214924351094388e-05, "loss": 0.0109, "step": 77500 }, { "grad_norm": 0.19660505652427673, "learning_rate": 1.3203727314029069e-05, "loss": 0.009, "step": 77510 }, { "grad_norm": 0.17933814227581024, "learning_rate": 1.3192534300948811e-05, "loss": 0.0092, "step": 77520 }, { "grad_norm": 0.11538124084472656, "learning_rate": 1.3181345313077675e-05, "loss": 0.0086, "step": 77530 }, { "grad_norm": 0.2215656191110611, "learning_rate": 1.3170160351639243e-05, "loss": 0.0123, "step": 77540 }, { "grad_norm": 0.1443634331226349, "learning_rate": 1.3158979417856721e-05, "loss": 0.0113, "step": 77550 }, { "grad_norm": 0.1137932613492012, "learning_rate": 1.31478025129528e-05, "loss": 0.0078, "step": 77560 }, { "grad_norm": 0.15165217220783234, "learning_rate": 1.3136629638149794e-05, "loss": 0.01, "step": 77570 }, { "grad_norm": 0.11426706612110138, "learning_rate": 1.3125460794669547e-05, "loss": 0.0098, "step": 77580 }, { "grad_norm": 0.1702757030725479, "learning_rate": 1.3114295983733483e-05, "loss": 0.0116, "step": 77590 }, { "grad_norm": 0.1659725159406662, "learning_rate": 1.3103135206562534e-05, "loss": 0.0108, "step": 77600 }, { "grad_norm": 0.1535559594631195, "learning_rate": 1.3091978464377264e-05, "loss": 0.0092, "step": 77610 }, { "grad_norm": 0.19154995679855347, "learning_rate": 1.3080825758397724e-05, "loss": 0.0114, "step": 77620 }, { "grad_norm": 0.17425015568733215, "learning_rate": 1.3069677089843568e-05, "loss": 0.0093, "step": 77630 }, { "grad_norm": 0.14015425741672516, "learning_rate": 1.305853245993401e-05, "loss": 0.011, "step": 77640 }, { "grad_norm": 0.15077504515647888, "learning_rate": 1.3047391869887809e-05, "loss": 0.0106, "step": 77650 }, { "grad_norm": 0.14535963535308838, "learning_rate": 1.3036255320923263e-05, "loss": 0.0119, "step": 77660 }, { "grad_norm": 0.18189674615859985, "learning_rate": 1.3025122814258273e-05, "loss": 0.011, "step": 77670 }, { "grad_norm": 0.1389508694410324, "learning_rate": 1.3013994351110249e-05, "loss": 0.0099, "step": 77680 }, { "grad_norm": 0.19755806028842926, "learning_rate": 1.3002869932696194e-05, "loss": 0.0098, "step": 77690 }, { "grad_norm": 0.14008986949920654, "learning_rate": 1.2991749560232664e-05, "loss": 0.0102, "step": 77700 }, { "grad_norm": 0.1122482568025589, "learning_rate": 1.2980633234935768e-05, "loss": 0.0126, "step": 77710 }, { "grad_norm": 0.1844431757926941, "learning_rate": 1.2969520958021153e-05, "loss": 0.0139, "step": 77720 }, { "grad_norm": 0.15062987804412842, "learning_rate": 1.295841273070405e-05, "loss": 0.0099, "step": 77730 }, { "grad_norm": 0.13221454620361328, "learning_rate": 1.2947308554199255e-05, "loss": 0.0108, "step": 77740 }, { "grad_norm": 0.18936258554458618, "learning_rate": 1.2936208429721075e-05, "loss": 0.0096, "step": 77750 }, { "grad_norm": 0.4303312301635742, "learning_rate": 1.2925112358483416e-05, "loss": 0.0113, "step": 77760 }, { "grad_norm": 0.09976965188980103, "learning_rate": 1.2914020341699728e-05, "loss": 0.0095, "step": 77770 }, { "grad_norm": 0.1471119374036789, "learning_rate": 1.2902932380583032e-05, "loss": 0.0096, "step": 77780 }, { "grad_norm": 0.15894833207130432, "learning_rate": 1.2891848476345863e-05, "loss": 0.0125, "step": 77790 }, { "grad_norm": 0.10586675256490707, "learning_rate": 1.2880768630200368e-05, "loss": 0.0115, "step": 77800 }, { "grad_norm": 0.15541626513004303, "learning_rate": 1.286969284335819e-05, "loss": 0.0112, "step": 77810 }, { "grad_norm": 0.13014723360538483, "learning_rate": 1.2858621117030578e-05, "loss": 0.01, "step": 77820 }, { "grad_norm": 0.37611761689186096, "learning_rate": 1.2847553452428324e-05, "loss": 0.0094, "step": 77830 }, { "grad_norm": 0.1397191286087036, "learning_rate": 1.283648985076178e-05, "loss": 0.0101, "step": 77840 }, { "grad_norm": 0.11232773959636688, "learning_rate": 1.2825430313240817e-05, "loss": 0.0082, "step": 77850 }, { "grad_norm": 0.13943736255168915, "learning_rate": 1.281437484107491e-05, "loss": 0.0088, "step": 77860 }, { "grad_norm": 0.13209645450115204, "learning_rate": 1.2803323435473058e-05, "loss": 0.0105, "step": 77870 }, { "grad_norm": 0.12692274153232574, "learning_rate": 1.2792276097643841e-05, "loss": 0.0098, "step": 77880 }, { "grad_norm": 0.15586118400096893, "learning_rate": 1.2781232828795342e-05, "loss": 0.011, "step": 77890 }, { "grad_norm": 0.14874009788036346, "learning_rate": 1.2770193630135292e-05, "loss": 0.0118, "step": 77900 }, { "grad_norm": 0.1757757067680359, "learning_rate": 1.2759158502870877e-05, "loss": 0.0093, "step": 77910 }, { "grad_norm": 0.16455544531345367, "learning_rate": 1.2748127448208913e-05, "loss": 0.0094, "step": 77920 }, { "grad_norm": 0.13890889286994934, "learning_rate": 1.2737100467355706e-05, "loss": 0.0076, "step": 77930 }, { "grad_norm": 0.16561701893806458, "learning_rate": 1.2726077561517169e-05, "loss": 0.013, "step": 77940 }, { "grad_norm": 0.11108626425266266, "learning_rate": 1.2715058731898755e-05, "loss": 0.0107, "step": 77950 }, { "grad_norm": 0.14930279552936554, "learning_rate": 1.2704043979705471e-05, "loss": 0.0089, "step": 77960 }, { "grad_norm": 0.18473047018051147, "learning_rate": 1.269303330614185e-05, "loss": 0.0146, "step": 77970 }, { "grad_norm": 0.11461587250232697, "learning_rate": 1.2682026712412015e-05, "loss": 0.0079, "step": 77980 }, { "grad_norm": 0.07573708146810532, "learning_rate": 1.2671024199719644e-05, "loss": 0.0098, "step": 77990 }, { "grad_norm": 0.17606203258037567, "learning_rate": 1.2660025769267936e-05, "loss": 0.0108, "step": 78000 }, { "grad_norm": 0.1573936939239502, "learning_rate": 1.2649031422259667e-05, "loss": 0.014, "step": 78010 }, { "grad_norm": 0.19177871942520142, "learning_rate": 1.2638041159897163e-05, "loss": 0.0103, "step": 78020 }, { "grad_norm": 0.08931169658899307, "learning_rate": 1.262705498338232e-05, "loss": 0.009, "step": 78030 }, { "grad_norm": 0.1343446522951126, "learning_rate": 1.2616072893916542e-05, "loss": 0.0117, "step": 78040 }, { "grad_norm": 0.09622412919998169, "learning_rate": 1.2605094892700836e-05, "loss": 0.0106, "step": 78050 }, { "grad_norm": 0.10043096542358398, "learning_rate": 1.259412098093572e-05, "loss": 0.0108, "step": 78060 }, { "grad_norm": 0.16047737002372742, "learning_rate": 1.258315115982131e-05, "loss": 0.0109, "step": 78070 }, { "grad_norm": 0.18219831585884094, "learning_rate": 1.2572185430557203e-05, "loss": 0.0111, "step": 78080 }, { "grad_norm": 0.26750797033309937, "learning_rate": 1.2561223794342659e-05, "loss": 0.0097, "step": 78090 }, { "grad_norm": 0.14229907095432281, "learning_rate": 1.255026625237637e-05, "loss": 0.0088, "step": 78100 }, { "grad_norm": 0.16151967644691467, "learning_rate": 1.2539312805856678e-05, "loss": 0.0131, "step": 78110 }, { "grad_norm": 0.1346127688884735, "learning_rate": 1.25283634559814e-05, "loss": 0.0084, "step": 78120 }, { "grad_norm": 0.22791266441345215, "learning_rate": 1.2517418203947972e-05, "loss": 0.0085, "step": 78130 }, { "grad_norm": 0.20592671632766724, "learning_rate": 1.2506477050953308e-05, "loss": 0.0124, "step": 78140 }, { "grad_norm": 0.12286077439785004, "learning_rate": 1.2495539998193972e-05, "loss": 0.0093, "step": 78150 }, { "grad_norm": 0.12294802069664001, "learning_rate": 1.2484607046865975e-05, "loss": 0.0088, "step": 78160 }, { "grad_norm": 0.1147468090057373, "learning_rate": 1.2473678198164968e-05, "loss": 0.0103, "step": 78170 }, { "grad_norm": 0.17860615253448486, "learning_rate": 1.2462753453286075e-05, "loss": 0.0107, "step": 78180 }, { "grad_norm": 0.13950850069522858, "learning_rate": 1.2451832813424031e-05, "loss": 0.0108, "step": 78190 }, { "grad_norm": 0.1526510864496231, "learning_rate": 1.2440916279773095e-05, "loss": 0.0097, "step": 78200 }, { "grad_norm": 0.10865077376365662, "learning_rate": 1.2430003853527101e-05, "loss": 0.0087, "step": 78210 }, { "grad_norm": 0.15681329369544983, "learning_rate": 1.2419095535879383e-05, "loss": 0.0085, "step": 78220 }, { "grad_norm": 0.12719881534576416, "learning_rate": 1.2408191328022878e-05, "loss": 0.0115, "step": 78230 }, { "grad_norm": 0.13843822479248047, "learning_rate": 1.2397291231150066e-05, "loss": 0.0097, "step": 78240 }, { "grad_norm": 0.1917797327041626, "learning_rate": 1.2386395246452937e-05, "loss": 0.0132, "step": 78250 }, { "grad_norm": 0.11043229699134827, "learning_rate": 1.2375503375123083e-05, "loss": 0.0104, "step": 78260 }, { "grad_norm": 0.18753394484519958, "learning_rate": 1.2364615618351583e-05, "loss": 0.0113, "step": 78270 }, { "grad_norm": 0.1676691621541977, "learning_rate": 1.2353731977329169e-05, "loss": 0.0087, "step": 78280 }, { "grad_norm": 0.1550932079553604, "learning_rate": 1.2342852453246007e-05, "loss": 0.0101, "step": 78290 }, { "grad_norm": 0.1155061200261116, "learning_rate": 1.2331977047291898e-05, "loss": 0.0108, "step": 78300 }, { "grad_norm": 0.10141342133283615, "learning_rate": 1.2321105760656137e-05, "loss": 0.0088, "step": 78310 }, { "grad_norm": 0.2091226726770401, "learning_rate": 1.2310238594527613e-05, "loss": 0.0107, "step": 78320 }, { "grad_norm": 0.13151413202285767, "learning_rate": 1.2299375550094693e-05, "loss": 0.0102, "step": 78330 }, { "grad_norm": 0.13656342029571533, "learning_rate": 1.2288516628545421e-05, "loss": 0.0071, "step": 78340 }, { "grad_norm": 0.07750924676656723, "learning_rate": 1.2277661831067255e-05, "loss": 0.0093, "step": 78350 }, { "grad_norm": 0.14225059747695923, "learning_rate": 1.2266811158847286e-05, "loss": 0.0092, "step": 78360 }, { "grad_norm": 0.22217239439487457, "learning_rate": 1.2255964613072107e-05, "loss": 0.0123, "step": 78370 }, { "grad_norm": 0.11083921045064926, "learning_rate": 1.2245122194927905e-05, "loss": 0.0089, "step": 78380 }, { "grad_norm": 0.18335945904254913, "learning_rate": 1.2234283905600364e-05, "loss": 0.0099, "step": 78390 }, { "grad_norm": 0.22477436065673828, "learning_rate": 1.222344974627475e-05, "loss": 0.0109, "step": 78400 }, { "grad_norm": 0.1672096997499466, "learning_rate": 1.2212619718135875e-05, "loss": 0.009, "step": 78410 }, { "grad_norm": 0.17298175394535065, "learning_rate": 1.2201793822368118e-05, "loss": 0.011, "step": 78420 }, { "grad_norm": 0.17042432725429535, "learning_rate": 1.2190972060155337e-05, "loss": 0.012, "step": 78430 }, { "grad_norm": 0.11476060003042221, "learning_rate": 1.2180154432681013e-05, "loss": 0.0107, "step": 78440 }, { "grad_norm": 0.18102407455444336, "learning_rate": 1.2169340941128148e-05, "loss": 0.0102, "step": 78450 }, { "grad_norm": 0.08619730919599533, "learning_rate": 1.2158531586679267e-05, "loss": 0.009, "step": 78460 }, { "grad_norm": 0.14444959163665771, "learning_rate": 1.2147726370516477e-05, "loss": 0.0116, "step": 78470 }, { "grad_norm": 0.15610411763191223, "learning_rate": 1.2136925293821422e-05, "loss": 0.0154, "step": 78480 }, { "grad_norm": 0.1395256519317627, "learning_rate": 1.2126128357775306e-05, "loss": 0.0085, "step": 78490 }, { "grad_norm": 0.08951389044523239, "learning_rate": 1.2115335563558838e-05, "loss": 0.008, "step": 78500 }, { "grad_norm": 0.15366610884666443, "learning_rate": 1.2104546912352327e-05, "loss": 0.0086, "step": 78510 }, { "grad_norm": 0.16092601418495178, "learning_rate": 1.2093762405335557e-05, "loss": 0.0098, "step": 78520 }, { "grad_norm": 0.09393416345119476, "learning_rate": 1.2082982043687974e-05, "loss": 0.0078, "step": 78530 }, { "grad_norm": 0.16226735711097717, "learning_rate": 1.207220582858845e-05, "loss": 0.0097, "step": 78540 }, { "grad_norm": 0.12083032727241516, "learning_rate": 1.206143376121549e-05, "loss": 0.0095, "step": 78550 }, { "grad_norm": 0.24058972299098969, "learning_rate": 1.2050665842747078e-05, "loss": 0.0142, "step": 78560 }, { "grad_norm": 0.13164903223514557, "learning_rate": 1.2039902074360804e-05, "loss": 0.0094, "step": 78570 }, { "grad_norm": 0.1741924285888672, "learning_rate": 1.2029142457233755e-05, "loss": 0.0137, "step": 78580 }, { "grad_norm": 0.13072623312473297, "learning_rate": 1.2018386992542601e-05, "loss": 0.0084, "step": 78590 }, { "grad_norm": 0.13846464455127716, "learning_rate": 1.2007635681463542e-05, "loss": 0.0094, "step": 78600 }, { "grad_norm": 0.1370706856250763, "learning_rate": 1.1996888525172334e-05, "loss": 0.0104, "step": 78610 }, { "grad_norm": 0.11105047911405563, "learning_rate": 1.198614552484425e-05, "loss": 0.0102, "step": 78620 }, { "grad_norm": 0.11397688835859299, "learning_rate": 1.197540668165415e-05, "loss": 0.0079, "step": 78630 }, { "grad_norm": 0.18238240480422974, "learning_rate": 1.1964671996776395e-05, "loss": 0.0114, "step": 78640 }, { "grad_norm": 0.08643367141485214, "learning_rate": 1.1953941471384922e-05, "loss": 0.0114, "step": 78650 }, { "grad_norm": 0.14685465395450592, "learning_rate": 1.1943215106653205e-05, "loss": 0.0098, "step": 78660 }, { "grad_norm": 0.12782108783721924, "learning_rate": 1.193249290375429e-05, "loss": 0.011, "step": 78670 }, { "grad_norm": 0.1980474442243576, "learning_rate": 1.1921774863860696e-05, "loss": 0.0076, "step": 78680 }, { "grad_norm": 0.2107195258140564, "learning_rate": 1.1911060988144556e-05, "loss": 0.0129, "step": 78690 }, { "grad_norm": 0.1594543755054474, "learning_rate": 1.1900351277777533e-05, "loss": 0.0128, "step": 78700 }, { "grad_norm": 0.2073819786310196, "learning_rate": 1.1889645733930798e-05, "loss": 0.0111, "step": 78710 }, { "grad_norm": 0.1901291161775589, "learning_rate": 1.1878944357775107e-05, "loss": 0.0101, "step": 78720 }, { "grad_norm": 0.14067962765693665, "learning_rate": 1.1868247150480749e-05, "loss": 0.0112, "step": 78730 }, { "grad_norm": 0.12389400601387024, "learning_rate": 1.1857554113217567e-05, "loss": 0.0089, "step": 78740 }, { "grad_norm": 0.1561223417520523, "learning_rate": 1.1846865247154903e-05, "loss": 0.0124, "step": 78750 }, { "grad_norm": 0.120414599776268, "learning_rate": 1.1836180553461711e-05, "loss": 0.0083, "step": 78760 }, { "grad_norm": 0.163938507437706, "learning_rate": 1.1825500033306419e-05, "loss": 0.0095, "step": 78770 }, { "grad_norm": 0.1458033174276352, "learning_rate": 1.1814823687857052e-05, "loss": 0.0099, "step": 78780 }, { "grad_norm": 0.1808008998632431, "learning_rate": 1.1804151518281153e-05, "loss": 0.0084, "step": 78790 }, { "grad_norm": 0.13649672269821167, "learning_rate": 1.179348352574583e-05, "loss": 0.0099, "step": 78800 }, { "grad_norm": 0.11519968509674072, "learning_rate": 1.1782819711417697e-05, "loss": 0.01, "step": 78810 }, { "grad_norm": 0.13935643434524536, "learning_rate": 1.1772160076462951e-05, "loss": 0.0118, "step": 78820 }, { "grad_norm": 0.1860647350549698, "learning_rate": 1.1761504622047293e-05, "loss": 0.0103, "step": 78830 }, { "grad_norm": 0.11914423853158951, "learning_rate": 1.1750853349335999e-05, "loss": 0.0104, "step": 78840 }, { "grad_norm": 0.20337915420532227, "learning_rate": 1.1740206259493874e-05, "loss": 0.0095, "step": 78850 }, { "grad_norm": 0.093773253262043, "learning_rate": 1.1729563353685286e-05, "loss": 0.0113, "step": 78860 }, { "grad_norm": 0.1575627326965332, "learning_rate": 1.1718924633074091e-05, "loss": 0.0145, "step": 78870 }, { "grad_norm": 0.14231941103935242, "learning_rate": 1.1708290098823765e-05, "loss": 0.0094, "step": 78880 }, { "grad_norm": 0.17001567780971527, "learning_rate": 1.1697659752097245e-05, "loss": 0.0125, "step": 78890 }, { "grad_norm": 0.1158253476023674, "learning_rate": 1.1687033594057068e-05, "loss": 0.0077, "step": 78900 }, { "grad_norm": 0.11643753200769424, "learning_rate": 1.1676411625865297e-05, "loss": 0.0086, "step": 78910 }, { "grad_norm": 0.22159211337566376, "learning_rate": 1.1665793848683542e-05, "loss": 0.0108, "step": 78920 }, { "grad_norm": 0.12345249205827713, "learning_rate": 1.1655180263672927e-05, "loss": 0.0104, "step": 78930 }, { "grad_norm": 0.12987515330314636, "learning_rate": 1.1644570871994142e-05, "loss": 0.0101, "step": 78940 }, { "grad_norm": 0.2397788017988205, "learning_rate": 1.163396567480744e-05, "loss": 0.0095, "step": 78950 }, { "grad_norm": 0.12084733694791794, "learning_rate": 1.1623364673272552e-05, "loss": 0.0088, "step": 78960 }, { "grad_norm": 0.323121577501297, "learning_rate": 1.1612767868548807e-05, "loss": 0.0092, "step": 78970 }, { "grad_norm": 0.11805281788110733, "learning_rate": 1.1602175261795057e-05, "loss": 0.0128, "step": 78980 }, { "grad_norm": 0.15068982541561127, "learning_rate": 1.1591586854169696e-05, "loss": 0.0102, "step": 78990 }, { "grad_norm": 0.08732733875513077, "learning_rate": 1.158100264683064e-05, "loss": 0.0092, "step": 79000 }, { "grad_norm": 0.13178811967372894, "learning_rate": 1.1570422640935386e-05, "loss": 0.0165, "step": 79010 }, { "grad_norm": 0.16523200273513794, "learning_rate": 1.155984683764092e-05, "loss": 0.0108, "step": 79020 }, { "grad_norm": 0.13775742053985596, "learning_rate": 1.1549275238103812e-05, "loss": 0.0114, "step": 79030 }, { "grad_norm": 0.1917218416929245, "learning_rate": 1.1538707843480152e-05, "loss": 0.011, "step": 79040 }, { "grad_norm": 0.11807599663734436, "learning_rate": 1.1528144654925593e-05, "loss": 0.0091, "step": 79050 }, { "grad_norm": 0.11883573979139328, "learning_rate": 1.1517585673595277e-05, "loss": 0.0096, "step": 79060 }, { "grad_norm": 0.1399165391921997, "learning_rate": 1.150703090064395e-05, "loss": 0.0112, "step": 79070 }, { "grad_norm": 0.16737525165081024, "learning_rate": 1.1496480337225835e-05, "loss": 0.0092, "step": 79080 }, { "grad_norm": 0.09077087044715881, "learning_rate": 1.1485933984494751e-05, "loss": 0.0087, "step": 79090 }, { "grad_norm": 0.15572887659072876, "learning_rate": 1.1475391843604e-05, "loss": 0.0097, "step": 79100 }, { "grad_norm": 0.1089714914560318, "learning_rate": 1.1464853915706503e-05, "loss": 0.0087, "step": 79110 }, { "grad_norm": 0.20984439551830292, "learning_rate": 1.1454320201954626e-05, "loss": 0.0092, "step": 79120 }, { "grad_norm": 0.17825229465961456, "learning_rate": 1.1443790703500356e-05, "loss": 0.009, "step": 79130 }, { "grad_norm": 0.1325809508562088, "learning_rate": 1.1433265421495154e-05, "loss": 0.0097, "step": 79140 }, { "grad_norm": 0.24008138477802277, "learning_rate": 1.1422744357090065e-05, "loss": 0.0108, "step": 79150 }, { "grad_norm": 0.12631262838840485, "learning_rate": 1.1412227511435647e-05, "loss": 0.0123, "step": 79160 }, { "grad_norm": 0.13034005463123322, "learning_rate": 1.1401714885682025e-05, "loss": 0.0082, "step": 79170 }, { "grad_norm": 0.1244312971830368, "learning_rate": 1.1391206480978823e-05, "loss": 0.0121, "step": 79180 }, { "grad_norm": 0.1663220375776291, "learning_rate": 1.1380702298475227e-05, "loss": 0.0093, "step": 79190 }, { "grad_norm": 0.12617571651935577, "learning_rate": 1.1370202339319985e-05, "loss": 0.0089, "step": 79200 }, { "grad_norm": 0.13084940612316132, "learning_rate": 1.1359706604661319e-05, "loss": 0.0129, "step": 79210 }, { "grad_norm": 0.16259537637233734, "learning_rate": 1.1349215095647042e-05, "loss": 0.0079, "step": 79220 }, { "grad_norm": 0.10800933092832565, "learning_rate": 1.1338727813424493e-05, "loss": 0.0079, "step": 79230 }, { "grad_norm": 0.14855165779590607, "learning_rate": 1.132824475914056e-05, "loss": 0.009, "step": 79240 }, { "grad_norm": 0.13195958733558655, "learning_rate": 1.131776593394162e-05, "loss": 0.0095, "step": 79250 }, { "grad_norm": 0.21285681426525116, "learning_rate": 1.1307291338973657e-05, "loss": 0.0096, "step": 79260 }, { "grad_norm": 0.25525468587875366, "learning_rate": 1.1296820975382121e-05, "loss": 0.0102, "step": 79270 }, { "grad_norm": 0.1535230576992035, "learning_rate": 1.1286354844312074e-05, "loss": 0.008, "step": 79280 }, { "grad_norm": 0.11210596561431885, "learning_rate": 1.1275892946908023e-05, "loss": 0.0083, "step": 79290 }, { "grad_norm": 0.14822137355804443, "learning_rate": 1.126543528431413e-05, "loss": 0.0084, "step": 79300 }, { "grad_norm": 0.10750871896743774, "learning_rate": 1.125498185767398e-05, "loss": 0.01, "step": 79310 }, { "grad_norm": 0.15108612179756165, "learning_rate": 1.1244532668130781e-05, "loss": 0.0101, "step": 79320 }, { "grad_norm": 0.1303851306438446, "learning_rate": 1.1234087716827203e-05, "loss": 0.008, "step": 79330 }, { "grad_norm": 0.1537071019411087, "learning_rate": 1.1223647004905524e-05, "loss": 0.0098, "step": 79340 }, { "grad_norm": 0.15696822106838226, "learning_rate": 1.1213210533507485e-05, "loss": 0.0118, "step": 79350 }, { "grad_norm": 0.1314118653535843, "learning_rate": 1.1202778303774447e-05, "loss": 0.0081, "step": 79360 }, { "grad_norm": 0.1769641637802124, "learning_rate": 1.1192350316847234e-05, "loss": 0.0097, "step": 79370 }, { "grad_norm": 0.1401410847902298, "learning_rate": 1.1181926573866258e-05, "loss": 0.0099, "step": 79380 }, { "grad_norm": 0.14274078607559204, "learning_rate": 1.1171507075971416e-05, "loss": 0.009, "step": 79390 }, { "grad_norm": 0.10966704040765762, "learning_rate": 1.1161091824302177e-05, "loss": 0.0088, "step": 79400 }, { "grad_norm": 0.10436579585075378, "learning_rate": 1.115068081999755e-05, "loss": 0.0102, "step": 79410 }, { "grad_norm": 0.1724313497543335, "learning_rate": 1.1140274064196071e-05, "loss": 0.0107, "step": 79420 }, { "grad_norm": 0.15211063623428345, "learning_rate": 1.112987155803578e-05, "loss": 0.0095, "step": 79430 }, { "grad_norm": 0.1277560591697693, "learning_rate": 1.1119473302654298e-05, "loss": 0.0094, "step": 79440 }, { "grad_norm": 0.17204543948173523, "learning_rate": 1.1109079299188769e-05, "loss": 0.0097, "step": 79450 }, { "grad_norm": 0.188472181558609, "learning_rate": 1.1098689548775847e-05, "loss": 0.0102, "step": 79460 }, { "grad_norm": 0.16150812804698944, "learning_rate": 1.1088304052551762e-05, "loss": 0.0113, "step": 79470 }, { "grad_norm": 0.11322782933712006, "learning_rate": 1.1077922811652215e-05, "loss": 0.0143, "step": 79480 }, { "grad_norm": 0.08517582714557648, "learning_rate": 1.1067545827212534e-05, "loss": 0.0079, "step": 79490 }, { "grad_norm": 0.11408987641334534, "learning_rate": 1.1057173100367496e-05, "loss": 0.0094, "step": 79500 }, { "grad_norm": 0.13408522307872772, "learning_rate": 1.1046804632251474e-05, "loss": 0.0088, "step": 79510 }, { "grad_norm": 0.20978152751922607, "learning_rate": 1.1036440423998318e-05, "loss": 0.0112, "step": 79520 }, { "grad_norm": 0.160264790058136, "learning_rate": 1.1026080476741469e-05, "loss": 0.0106, "step": 79530 }, { "grad_norm": 0.13172858953475952, "learning_rate": 1.1015724791613836e-05, "loss": 0.0091, "step": 79540 }, { "grad_norm": 0.17519710958003998, "learning_rate": 1.1005373369747951e-05, "loss": 0.0097, "step": 79550 }, { "grad_norm": 0.12965527176856995, "learning_rate": 1.09950262122758e-05, "loss": 0.0107, "step": 79560 }, { "grad_norm": 0.10479626804590225, "learning_rate": 1.0984683320328948e-05, "loss": 0.0113, "step": 79570 }, { "grad_norm": 0.10857352614402771, "learning_rate": 1.0974344695038458e-05, "loss": 0.0099, "step": 79580 }, { "grad_norm": 0.16697123646736145, "learning_rate": 1.0964010337534968e-05, "loss": 0.0119, "step": 79590 }, { "grad_norm": 0.21428944170475006, "learning_rate": 1.0953680248948611e-05, "loss": 0.0101, "step": 79600 }, { "grad_norm": 0.21638424694538116, "learning_rate": 1.0943354430409069e-05, "loss": 0.0122, "step": 79610 }, { "grad_norm": 0.17481893301010132, "learning_rate": 1.0933032883045574e-05, "loss": 0.0099, "step": 79620 }, { "grad_norm": 0.16169750690460205, "learning_rate": 1.0922715607986872e-05, "loss": 0.0111, "step": 79630 }, { "grad_norm": 0.14609745144844055, "learning_rate": 1.0912402606361234e-05, "loss": 0.0125, "step": 79640 }, { "grad_norm": 0.10949131101369858, "learning_rate": 1.090209387929647e-05, "loss": 0.0095, "step": 79650 }, { "grad_norm": 0.18099497258663177, "learning_rate": 1.0891789427919957e-05, "loss": 0.0108, "step": 79660 }, { "grad_norm": 0.10106644034385681, "learning_rate": 1.088148925335854e-05, "loss": 0.0113, "step": 79670 }, { "grad_norm": 0.1345289945602417, "learning_rate": 1.087119335673864e-05, "loss": 0.0103, "step": 79680 }, { "grad_norm": 0.14207038283348083, "learning_rate": 1.0860901739186208e-05, "loss": 0.0109, "step": 79690 }, { "grad_norm": 0.1820213794708252, "learning_rate": 1.085061440182673e-05, "loss": 0.012, "step": 79700 }, { "grad_norm": 0.12825258076190948, "learning_rate": 1.0840331345785193e-05, "loss": 0.0091, "step": 79710 }, { "grad_norm": 0.11504048854112625, "learning_rate": 1.083005257218615e-05, "loss": 0.0114, "step": 79720 }, { "grad_norm": 0.1322498917579651, "learning_rate": 1.0819778082153642e-05, "loss": 0.009, "step": 79730 }, { "grad_norm": 0.16463518142700195, "learning_rate": 1.0809507876811325e-05, "loss": 0.0084, "step": 79740 }, { "grad_norm": 0.10942863672971725, "learning_rate": 1.0799241957282292e-05, "loss": 0.0112, "step": 79750 }, { "grad_norm": 0.12429701536893845, "learning_rate": 1.078898032468923e-05, "loss": 0.0088, "step": 79760 }, { "grad_norm": 0.08829639106988907, "learning_rate": 1.0778722980154315e-05, "loss": 0.0103, "step": 79770 }, { "grad_norm": 0.1728365123271942, "learning_rate": 1.0768469924799302e-05, "loss": 0.0155, "step": 79780 }, { "grad_norm": 0.09425348788499832, "learning_rate": 1.0758221159745419e-05, "loss": 0.0113, "step": 79790 }, { "grad_norm": 0.1685355305671692, "learning_rate": 1.074797668611347e-05, "loss": 0.0101, "step": 79800 }, { "grad_norm": 0.12444866448640823, "learning_rate": 1.0737736505023777e-05, "loss": 0.0105, "step": 79810 }, { "grad_norm": 0.11329516023397446, "learning_rate": 1.0727500617596203e-05, "loss": 0.0077, "step": 79820 }, { "grad_norm": 0.16442438960075378, "learning_rate": 1.0717269024950104e-05, "loss": 0.0119, "step": 79830 }, { "grad_norm": 0.12170200794935226, "learning_rate": 1.070704172820442e-05, "loss": 0.012, "step": 79840 }, { "grad_norm": 0.12329870462417603, "learning_rate": 1.0696818728477559e-05, "loss": 0.011, "step": 79850 }, { "grad_norm": 0.11443974077701569, "learning_rate": 1.0686600026887511e-05, "loss": 0.0092, "step": 79860 }, { "grad_norm": 0.10489356517791748, "learning_rate": 1.0676385624551777e-05, "loss": 0.0095, "step": 79870 }, { "grad_norm": 0.1940058320760727, "learning_rate": 1.0666175522587402e-05, "loss": 0.0096, "step": 79880 }, { "grad_norm": 0.13473716378211975, "learning_rate": 1.0655969722110925e-05, "loss": 0.0088, "step": 79890 }, { "grad_norm": 0.16024865210056305, "learning_rate": 1.0645768224238439e-05, "loss": 0.0104, "step": 79900 }, { "grad_norm": 0.22383584082126617, "learning_rate": 1.063557103008559e-05, "loss": 0.0116, "step": 79910 }, { "grad_norm": 0.22114810347557068, "learning_rate": 1.0625378140767494e-05, "loss": 0.0103, "step": 79920 }, { "grad_norm": 0.1608397513628006, "learning_rate": 1.0615189557398846e-05, "loss": 0.0093, "step": 79930 }, { "grad_norm": 0.1049596518278122, "learning_rate": 1.060500528109386e-05, "loss": 0.0084, "step": 79940 }, { "grad_norm": 0.1412561684846878, "learning_rate": 1.0594825312966272e-05, "loss": 0.0108, "step": 79950 }, { "grad_norm": 0.15204963088035583, "learning_rate": 1.0584649654129336e-05, "loss": 0.0105, "step": 79960 }, { "grad_norm": 0.10766848176717758, "learning_rate": 1.0574478305695863e-05, "loss": 0.0089, "step": 79970 }, { "grad_norm": 0.20024262368679047, "learning_rate": 1.0564311268778154e-05, "loss": 0.0092, "step": 79980 }, { "grad_norm": 0.12085647881031036, "learning_rate": 1.0554148544488073e-05, "loss": 0.0101, "step": 79990 }, { "grad_norm": 0.13826067745685577, "learning_rate": 1.0543990133937004e-05, "loss": 0.009, "step": 80000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 24, "trial_name": null, "trial_params": null }